Ajuba negatively regulates the Wnt signaling pathway by ... - Nature

2 downloads 0 Views 289KB Size Report
Jul 9, 2007 - The Wnt signaling pathway is essential for embryonic development and carcinogenesis. Upon Wnt stimulation, b-catenin is stabilized and ...
Oncogene (2008) 27, 274–284

& 2008 Nature Publishing Group All rights reserved 0950-9232/08 $30.00 www.nature.com/onc

ORIGINAL ARTICLE

Ajuba negatively regulates the Wnt signaling pathway by promoting GSK-3b-mediated phosphorylation of b-catenin K Haraguchi1, M Ohsugi1, Y Abe1,4, K Semba2, T Akiyama3 and T Yamamoto1 1 Division of Oncology, Institute of Medical Science, University of Tokyo, Tokyo, Japan; 2Division of Cellular and Molecular Biology, Institute of Medical Science, University of Tokyo, Tokyo, Japan and 3Laboratory of Molecular and Genetic Information, Institute of Molecular and Cellular Bioscience, University of Tokyo, Tokyo, Japan

The Wnt signaling pathway is essential for embryonic development and carcinogenesis. Upon Wnt stimulation, b-catenin is stabilized and associates with T-cell factor or lymphoid enhancing factor, thereby activating transcription of target genes. In the absence of Wnt stimulation, the level of b-catenin is reduced via glycogen synthase kinase (GSK)-3b-mediated phosphorylation and subsequent proteasome-dependent degradation. Here, we report the identification of Ajuba as a negative regulator of the Wnt signaling pathway. Ajuba is a member of LIM domain-containing proteins that contribute to cell fate determination and regulate cell proliferation and differentiation. We found that enforced expression of Ajuba destabilized b-catenin and suppressed target gene expression. Ajuba promoted GSK-3b-mediated phosphorylation of b-catenin by reinforcing the association between b-catenin and GSK-3b. Furthermore, Wnt stimulation induced both accumulation of b-catenin and destabilization of Ajuba. Our findings suggest that Ajuba is important for regulation of the Wnt signaling pathway. Oncogene (2008) 27, 274–284; doi:10.1038/sj.onc.1210644; published online 9 July 2007 Keywords: Wnt; b-catenin; GSK-3b; Ajuba; phosphorylation

Introduction The Wnt family of secreted proteins is conserved among species from Caenorhabditis elegans to human and plays important roles in developmental processes by activating transcription of target genes that regulate cell proliferation and differentiation (Miller and Moon, 1996; Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003). Deregulation of the Wnt signaling pathway associates with development of various types of tumors (Miller and Moon, 1996; Correspondence: Professor T Yamamoto, Division of Oncology, Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo 108-8639, Japan. E-mail: [email protected] 4 Current address: Department of Molecular Biology, Institute of Gerontology, Nippon Medical School, Kawasaki 211-8533, Japan. Received 16 February 2007; revised 24 May 2007; accepted 1 June 2007; published online 9 July 2007

Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003). Regulation of the b-catenin level is a critical step in Wnt signaling. In the absence of Wnt stimulation, b-catenin is recruited to a multiprotein complex containing adenomatous polyposis coli (APC) tumor suppressor protein, Axin and glycogen synthase kinase (GSK)-3b. b-Catenin is phosphorylated by GSK-3b and then degraded through the ubiquitin-proteasome system (Miller and Moon, 1996; Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003). Upon Wnt stimulation, Dishevelled (Dsh) and frat-1 mediate the dissociation of GSK-3b and Axin (Li et al., 1999). In addition, Axin is recruited to the cytoplasmic domain of the Wnt co-receptor LRP5/6 (Mao et al., 2001; Tolwinski et al., 2003; Tamai et al., 2004). As a result, components of the Wnt multiprotein complex dissociate and b-catenin is stabilized. Accumulated b-catenin activates transcription of target genes through its interaction with T-cell factor/lymphoid enhancing factor (TCF/LEF) family of transcription factors (Miller and Moon, 1996; Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003). Importantly, mutations in GSK-3b phosphorylation sites of b-catenin result in unscheduled activation of the Wnt signaling pathway (Miller and Moon, 1996; Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003). Ajuba is an LIM protein and has proline-rich SH3 recognition motifs in the N-terminal region, a nuclear export signal in the middle region and three LIM domains in the C-terminal region (Goyal et al., 1999). LIM domains are characterized by two tandemly repeated zinc finger structures and are involved in many cellular events (Dawid et al., 1998). Previous reports showed that Ajuba is involved in the regulation of cell proliferation and differentiation. Ajuba associates with Grb2, and its overexpression induces meiotic maturation of Xenopus oocytes in a manner dependent on Grb2 and Ras (Goyal et al., 1999). Ajuba also induces endodermal differentiation of P19 embryonal carcinoma stem cells through activation of c-Jun kinase (Kanungo et al., 2000). There are reports suggesting that b-catenin associates with the LIM protein FHL2 (four-and-a-half LIM-only protein 2) (Martin et al., 2002; Wei et al.,

Ajuba and b-catenin phosphorylation in the Wnt pathway K Haraguchi et al

275

2003). FHL2 colocalizes with b-catenin in the nucleus and enhances TCF/LEF-dependent transcription in various cell lines through the promotion of b-catenin acetylation by p300, which enhances the interaction between b-catenin and TCF4 (Martin et al., 2002; Wei et al., 2003; Labalette et al., 2004; Levy et al., 2004). The data suggest that other LIM proteins may also be involved in the Wnt signaling pathway. In the present study, we investigated the role of Ajuba in the Wnt signaling pathway. We found that Ajuba associates with b-catenin and enhances GSK-3b-mediated phosphorylation of b-catenin. Our findings provide a novel insight in the regulation of the Wnt signaling pathway.

Results Identification of Ajuba as a novel b-catenin-interacting protein We first investigated whether Ajuba associates with b-catenin. A pull-down assay with in vitro-translated b-catenin and glutathione S-transferase (GST)–Ajuba fusion protein purified from Escherichia coli showed that b-catenin was co-precipitated with GST–Ajuba but not with GST (Figure 1a). We next examined whether Ajuba and b-catenin interact with each other in a physiological condition. To detect endogenous Ajuba protein, we generated rabbit polyclonal antibodies against the N-terminal region of Ajuba. Immunoblot analysis of cell lysates revealed that the antibodies specifically recognized the 59-kDa endogenous Ajuba protein (Figure 1b). We then immunoprecipitated endogenous Ajuba with the antibodies. b-Catenin was detected in the Ajuba immunoprecipitates but not in control immunoglobulin precipitates (Figure 1c). Similarly, Ajuba was detected in the b-catenin immunoprecipitates but not in control immunoglobulin precipitates (Figure 1c). These results indicate that Ajuba associates directly with b-catenin in vivo. Mapping of the interacting region of b-catenin and Ajuba To determine Ajuba-binding region within b-catenin, we constructed a series of deletion mutants of b-catenin (Figure 2a). Co-immunoprecipitation analysis showed that Ajuba strongly interacted with armadillo repeats 7–9 (amino acids 400–530) of b-catenin (Figure 2a). This region is different from the binding site of Wntregulating factors such as GSK-3b and Axin (Ikeda et al., 1998; Peifer and Polakis, 2000), suggesting that the binding of Ajuba to b-catenin is not competitive with these molecules. Indeed co-immunoprecipitation assay using lysates of HEL293T cells transfected with b-catenin and Ajuba together with or without Axin showed that binding of Ajuba with b-catenin was not prevented by Axin (Supplementary Figure S3). We also searched for the b-catenin-binding region within Ajuba. Co-immunoprecipitation analysis with various deletion mutants of Ajuba showed that all LIM domains together with the N-terminally flanking region (amino acids 232–538) of Ajuba was necessary for its binding to

b-catenin (Figure 2b). Like transcriptional cofactor FHL2 (Wei et al., 2003), the LIM domains of Ajuba were important for the interaction with b-catenin, but the LIM domains alone were not sufficient for the interaction. Suppression of b-catenin-TCF-dependent transcription by Ajuba To clarify the biological significance of the association between Ajuba and b-catenin, we exogenously expressed green fluorescent protein (GFP)-fused Ajuba (GFPAjuba) in HeLa cells and analysed the subcellular localization or level of endogenous b-catenin by immunofluorescence microscopy. We found that the expression levels of b-catenin were remarkably reduced in cells expressing GFP-Ajuba, but not in cells expressing GFP alone (Figure 3a), suggesting that Ajuba promoted the degradation of b-catenin. The Ajubamediated decrease of b-catenin could result in inhibition of b-catenin-dependent gene expression. We then examined the effect of Ajuba on b-catenin-TCFmediated transcription using a luciferase reporter construct containing intact (TopFLASH) or mutated (FopFLASH) TCF-responsive elements. We transfected the reporter plasmid together with b-catenin and Ajuba expression plasmids into HEK293T cells. Whereas the reporter activity of TopFLASH was strongly activated by b-catenin, Ajuba suppressed b-catenin-TCF-activated transcription in a dose-dependent manner (Figure 3b). The effect of Ajuba overexpression was also monitored by cyclin D1 expression that is one of the well-established Wnt target genes (Shtutman et al., 1999; Tetsu and McCormick, 1999). Although cyclin D1 is activated by various transcription factors such as EtsB (Tetsu and McCormick, 1999), Ajuba repressed b-catenin-activated cyclin D1 promoter, but not EtsBactivated cyclin D1 promoter (Figure 3c). We also infected mouse fibroblast L cells with retroviruses that expressed GFP-Ajuba or GFP alone and prepared total RNA from the infected cells. Quantitative reverse transcription (RT)–PCR with cyclin D1 specific primers revealed that enforced expression of Ajuba suppressed expression of cyclin D1 mRNA (Figure 3d). To further establish the role of Ajuba in the control of b-cateninmediated gene expression, we performed Ajuba depletion experiment. We used small interfering RNAs (siRNAs; Elbashir et al., 2001) to inhibit the expression of Ajuba. In Ajuba-depleted cells, b-catenin-TCFdependent transcription was activated as compared to that in control siRNA-transfected cells (Figure 3e). In accordance with this, both b-catenin and cyclin D1 expression were increased in Ajuba-depleted cells (Figure 3e). By flow cytometric analysis, we observed that population of cells in G0/G1 phase was decreased and that of S and G2/M phase was increased in Ajubadepleted cells as compared to control cells (Figure 3f). Since cyclin D1 is a critical regulator for the progression of G1 phase in the cell cycle (Quelle et al., 1993), depletion of Ajuba resulted in stimulation of G0/G1 to S phase progression in the cell cycle. These results Oncogene

Ajuba and b-catenin phosphorylation in the Wnt pathway K Haraguchi et al

GST-Ajuba

b 1 Proline-rich

GST

+ GST-Ajuba

M.W. (kDa)

+ GST

a

IVT: β-catenin

276 NES

538 aa LIM1

LIM2

LIM3

Antigen (1-110 aa) M.W. (kDa) 250

150

150

100

100 75 75

Ajuba 50

50

37

37

25

25 IB: Ajuba

c

Ly

Cont IgG

IP

Ajuba

Ly

Cont IgG

IP

β-catenin

IB: β-catenin

CBB

IB: β-catenin

IB: Ajuba

*

*

Figure 1 Ajuba interacts directly with b-catenin. (a) Interaction between Ajuba and b-catenin in vitro. In vitro-translated b-catenin was incubated with glutathione S-transferase (GST) or GST–Ajuba. GST or GST–Ajuba bound to glutathione sepharose was separated by 10% SDS–PAGE and immunoblotted with anti-b-catenin antibody. CBB: Coomassie Brilliant Blue staining. (b) Preparation of polyclonal anti-Ajuba antibodies. Lysates from HeLa cells were subjected to immunoblot assay with affinity-purified anti-Ajuba antibodies. (c) Association of Ajuba with b-catenin in vivo. Lysates prepared from HEK 293T or HeLa cells were immunoprecipitated by anti-Ajuba antibodies or anti-b-catenin antibody and immunoblotted with the indicated antibodies. Asterisk indicates immunoglobulin G (IgG) heavy chain.

suggested that Ajuba negatively regulates the b-cateninTCF-mediated transcription and cell cycle progression. Wnt stimulation induces destabilization of Ajuba When we incubated L cells with Wnt-3A-conditioned medium (CM), we observed accumulation of b-catenin as reported previously (Shibamoto et al., 1998) (Figure 4a). On the contrary, we found that the levels of Ajuba decreased after Wnt-3A stimulation (Figure 4a). Quantitative RT–PCR with Ajuba-specific primers revealed that mRNA levels of Ajuba did not change in the presence or the absence of Wnt-3A-CM (Figure 4b), suggesting that the Ajuba is not regulated at the transcriptional level following Wnt-3A-CM stimulation. On the other hand, treatment of cells with proteasome inhibitor, MG132, inhibited the Wnt-3Ainduced reduction of Ajuba (Figure 4c), indicating that Ajuba is degraded through the ubiquitin–proteasome Oncogene

system. We also found that upon Wnt-3A stimulation, migration of Ajuba protein in sodium dodecyl sulfate polyacrylamide gel (SDS–PAGE) became faster in the presence of MG132 than its absence (Figure 4c). The data suggested that the fast-migrating form of Ajuba was more susceptible to proteasome-dependent degradation than the slow-migrating form. Since alkaline phosphatase treatment of the Ajuba-immunoprecipitates resulted in disappearance of the slow-migrating form of Ajuba (Figure 5a), we concluded that the slow-migrating form of Ajuba was generated by phosphorylation and Wnt-3A stimulation promoted dephosphorylation of Ajuba. To further investigate the molecular mechanism by which Wnt-3A stimulation promotes the Ajuba degradation, we examined whether GSK-3b phosphorylates Ajuba because Wnt stimulation inactivates GSK-3b (Miller and Moon, 1996; Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003). We performed in vitro kinase assay using recombinant Ajuba

Ajuba and b-catenin phosphorylation in the Wnt pathway K Haraguchi et al

277

a

667-781 aa

Arm10-12

Arm7-9

Arm4-6

Arm1-3

1-150 aa

WT

GFP

GFP-β-catenin

β-catenin Arm repeat 1

781 aa

Bind

WT

+

1-150 aa

-

Arm1-3

-

Arm4-6

-

Arm7-9

+

IB: GFP IP: GFP

*

*

IB: FLAG

Arm10-12

-

667-781 aa

-

Ly

b

LIM3

LIM 3

PreLIM

LIM

231 aa

98 aa

WT

FLAG

WT

FLAG-Ajuba

Ajuba 1 Proline-rich

IB: GFP

NES

538 aa

Bind WT

*

IP: GFP

Ly IB: FLAG

Ly

+

98 aa

+

231 aa

+

LIM

-

PreLIM

-

LIM2 3

-

LIM3

-

Figure 2 Mapping of the interacting regions of b-catenin and Ajuba. (a) Mapping of the Ajuba-binding region of b-catenin. FLAG– Ajuba and a series of deletion constructs of green fluorescent protein (GFP)-b-catenin were transfected into HEK293T cells. Lysates were immunoprecipitated with anti-GFP antibodies and immunoblotted with the indicated antibodies. Asterisk indicates IgG heavy chain. (b) Mapping of the b-catenin-binding region of Ajuba. GFP-b-catenin and a series of deletion constructs of FLAG–Ajuba were transfected into HEK-293T cells. Lysates were immunoprecipitated with anti-GFP antibodies and immunoblotted with the indicated antibodies. Asterisk indicates IgG heavy chain.

and GSK-3b. As shown in Figure 5b, Ajuba was phosphorylated by GSK-3b. We next examined the effect of LiCl, an inhibitor of GSK-3b, on Ajuba phosphorylation state and its stability. Treatment of cells with LiCl resulted in disappearance of phosphorylated form of Ajuba as well as the decreases the amount of Ajuba (Figure 5c). These results indicated that GSK-3b-phosphorylated Ajuba is stable and Wnt stimulation-dependent inactivation of GSK-3b results in dephosphorylation and subsequent degradation of Ajuba. Ajuba inhibits Wnt-activated gene transcription We next examined the effect of Ajuba on Wnt-3A stimulation-dependent cellular response. When L cells were treated with Wnt-3A-CM, b-catenin remarkably accumulated both in the cytoplasm and in the nucleus

(Shibamoto et al., 1998). When GFP-Ajuba was exogenously expressed in L cells, b-catenin did not accumulate even after Wnt-3A-CM stimulation (Figure 6). The accumulation of b-catenin was not suppressed by GFP expression (Figure 6). In contrast, depletion of Ajuba did not have any more effect on TopFLASH activity in the presence of Wnt-3A stimulation (Supplementary Figure S1). Therefore, we assume that effect of Ajuba depletion is, at least in part, overlapping with that of Wnt stimulation. Taken together, we concluded that Ajuba inhibits the Wnt-stimulation-dependent cellular response. Ajuba promotes GSK-3b-mediated b-catenin phosphorylation and b-catenin degradation Since the reduction of the b-catenin level is regulated through its phosphorylation by GSK-3b (Miller and Oncogene

Ajuba and b-catenin phosphorylation in the Wnt pathway K Haraguchi et al

278

a

β-catenin

GFP

b Luciferase activity (arbitaryunit)

15

β-catenin

GFP-Ajub a

10

5

1

c

β-catenin Ajuba

3

+

+

+

+

d 2

1

EtsB β-catenin Ajuba

+

+



+

e ∗

1

0.5

+ +

FopFLASH

Luciferase activity (arbitary unit)

Relative cyclin D1 mRNA levels

Luciferase activity (arbitary unit)

TopFLASH

+ +

+ Virus

∗∗

2

1

GFP GFP-Ajuba

-1745 Cyclin D1 reporter RNAi

f G1

S

G2/M

77.3 ± 0.47

12.2 ± 0.42

10.3 ± 0.35

Ajuba RNAi 1 65.0 ± 0.32

17.0 ± 0.23

17.8 ± 0.44

Control

Cont Ajuba 1 Ajuba 2

IB: Ajuba IB: β-catenin IB: Cyclin D1

Figure 3 Ajuba inhibits b-catenin-TCF-activated gene transcription. (a) Immunofluorescence analysis. HeLa cells were transfected with green fluorescent protein (GFP) or GFP–Ajuba expression plasmid. After 48 h, immunostaining with anti-GFP and anti-b-catenin antibodies was performed. Scale bar, 10 mm. (b) Ajuba suppressed b-catenin-TCF-mediated transcriptional activity. HEK293T cells were transfected with reporter plasmid (0.5 mg), pRL-TK (0.05 mg), b-catenin (1 mg) and Ajuba expression plasmids (0.5, 3 mg). Reporter plasmids, TopFLASH and FopFLASH were obtained from V Korinek and H Clevers. (c) Ajuba decreased cyclin D1 mRNA levels. Total RNA was isolated from L cells expressing GFP or GFP–Ajuba and subjected to RT–PCR with real-time quantification. Cyclin D1 mRNA levels were normalized to those of HPRT mRNA and shown as the mean7s.d. of triplicate samples. Asterisk indicates P ¼ 0.003. (d) Ajuba represses b-catenin-activated cyclin D1 promoter. HEK-293T cells were transfected with EtsB (0.5 mg) or b-catenin (0.5 mg) expression plasmid with 1745 cyclin D1 reporter plasmid (0.5 mg) and Ajuba expression plasmids (0.5, 3 mg). (e) Depletion of Ajuba resulted in activation of b-catenin-TCF-dependent transcription. HeLa cells were transfected with control or Ajuba siRNA. After 24 h, cells were further transfected with reporter plasmid and measured luciferase activity after an additional 24 h. Asterisk indicates P ¼ 0.019. Double asterisk indicates P ¼ 0.013. Lysates from Ajuba-depleted cells were used for western blot analysis. (f) Flow cytometry analysis. Cells were transfected with control or Ajuba siRNA. After 48 h, cells were stained with propidium iodide and analysed by flow cytometer. The cell cycle population was determined by analysis of 10 000 cells. Data are presented as the mean7s.d. of triplicate samples.

Oncogene

Ajuba and b-catenin phosphorylation in the Wnt pathway K Haraguchi et al

279

a

Control CM 0

1

2

6

Wnt-3A CM 24

48

0

1

2

6

24

48

h

IB: β-catenin IB: Ajuba IB: α-tubulin

c

b

MG132 + Wnt-3A CM 0 3 6 24 48 (h)

Relative Ajuba mRNA levels

IB: Ajuba 1 IB: α-tubulin

0

6 Control

48

0

6

48

h

Wnt-3A CM

Figure 4 Wnt-3A-conditioned medium (CM) induces destabilization of Ajuba. (a) Effect of Wnt-3A-CM on Ajuba. L cells were treated with control or Wnt-3A-CM. Cell lysates at different time points were separated by SDS–PAGE and subjected to immunoblot assay with the indicated antibodies. (b) The levels of Ajuba mRNA in response to Wnt-3A-CM. Total RNA was isolated from L cells treated with control medium or Wnt-3A-CM and subjected to RT–PCR with real-time quantification. Ajuba mRNA levels were normalized to those of HPRT mRNA and shown as the mean7s.d. of triplicate samples. (c). Effect of proteasome inhibitor, MG132on Ajuba expression. L cells stimulated with Wnt-3A CM for different time points and treated with 50 mM MG132 for 3 h. Lysates were subjected to western blot analysis.

Moon, 1996; Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003), we examined the effect of Ajuba expression on GSK-3b-mediated b-catenin phosphorylation. In vitro kinase assay revealed that purified GSK-3b alone was unable to phosphorylate b-catenin efficiently (Figure 7a). Interestingly, phosphorylation of b-catenin by GSK-3b was strongly enhanced by the addition of purified Ajuba protein (Figure 7a). We also observed the direct interaction of Ajuba with b-catenin and GSK-3b (Figure 1a, Supplementary Figure S2a). These data suggested that Ajuba is responsible for GSK-3bmediated phosphorylation of b-catenin by promoting interaction between b-catenin and GSK-3b. To confirm this, we exogenously expressed Ajuba in HEK293T cells and examined the phosphorylation state of b-catenin with antibodies recognizing phosphorylated b-catenin at Ser33, which is one of the phosphorylation sites by GSK-3b (Miller and Moon, 1996; Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003). As Ajuba expression increased, the levels of phosphorylated b-catenin increased, whereas

overall b-catenin levels decreased (Figure 7b). In Ajubadepleted cells, the level of b-catenin increased as compared to that in control siRNA-transfected cells (Figure 7c). In addition, when the same amounts of b-catenin were immunoprecipitated from the lysates of both control and Ajuba-depleted cells, the degree of b-catenin phosphorylation was reduced in Ajuba-depleted cells (Figure 7c). We further examined the effect of Ajuba on TCF-dependent reporter activity enhanced by the b-catenin mutant in which Serine37 is substituted with Alanine (S37A b-catenin). S37A b-catenin is resistant to proteasome-dependent degradation in comparison with wild-type b-catenin because phosphorylation of b-catenin by GSK-3b was insufficient (Filali et al., 2002). We transfected wild-type or S37A b-catenin expression plasmid into HEK293T cells together with TopFLASH reporter plasmid to activate the reporter at the same extent. Ajuba effectively suppressed the wildtype b-catenin-induced, but not S37A b-catenin-induced TopFLASH activity (Figure 7d). To confirm that Ajuba promotes GSK-3b-mediated phosphorylation and subsequent degradation of b-catenin, we transfected Oncogene

Ajuba and b-catenin phosphorylation in the Wnt pathway K Haraguchi et al

280

FLAG-Ajuba expression plasmid together with Mycb-catenin and HA-GSK-3b expression plasmids into HeLa cells. Co-immunoprecipitation analysis revealed that the interaction between b-catenin and GSK-3b was strongly enhanced by expression of Ajuba (Figure 7e). Taken together, these data suggest that Ajuba facilitates GSK-3b-mediated phosphorylation and degradation of b-catenin by strengthening the association between b-catenin and GSK-3b.

Ajuba and b-catenin resulted in the promotion of GSK-3b-mediated phosphorylation of b-catenin and its subsequent degradation. Using a series of deletion mutants, we found that similar to another LIM protein FHL2 (Wei et al., 2003), LIM domains of Ajuba were necessary for the binding to b-catenin (Figure 2b). We also showed that armadillo repeats 7–9 in b-catenin participated in the association with Ajuba (Figure 2a). To date, many factors have been identified to be important for the regulation of the Wnt signaling

Discussion In the present study, we found that Ajuba was a novel b-catenin-binding protein. The interaction between

a

β-catenin

IP: FLAG-Ajuba

BAP NaF+Na3VO4

+

Control CM

+ + p-FLAG-Ajuba FLAG-Ajuba

c

b 32P

Ajuba

GFP

LiCl -

+

IB: β-catenin CBB

Wnt-3A CM

IB: Ajuba GFP-Ajuba

GSTGSK-3β

1 WT

KN

GSK-3β β

0.43

IB: α-tubulin

Figure 5 Ajuba is phosphorylated by glycogen synthase kinase (GSK)-3b. (a) Ajuba is phosphorylated in vivo. HEK293T cells were transfected with FLAG–Ajuba expression plasmid. Lysates were subjected to immunoprecipitation with anti-FLAG antibody. Immunoprecipitates were treated with bacterial alkaline phosphatase (BAP) in the presence or absence of NaF and Na3VO4. (b) Phosphorylation of Ajuba by GSK-3b. In vitro kinase assay was performed using recombinant Ajuba and GST-GSK-3b. (c). Effect of GSK-3b inhibitor, LiCl on Ajuba expression. HEK293T cells were treated with 30 mM LiCl for 6 h. Lysates were subjected to western blot analysis.

Figure 6 Ajuba inhibits Wnt-activated b-catenin activity. Inhibition of Wnt-3A-CM-induced b-catenin accumulation by Ajuba overexpression. L cells were transfected with green fluorescent protein (GFP)–Ajuba or GFP expression plasmid. After 24 h, these cells were treated with Wnt-3A-CM for 3 h. Immunostaining was performed with anti-GFP and anti-b-catenin antibodies. Scale bar, 10 mm.

Figure 7 Ajuba enhances glycogen synthase kinase (GSK)-3b-mediated phosphorylation of b-catenin and reduction of b-catenin levels. (a) In vitro kinase assay. Glutathione S-transferase (GST)-b-catenin was treated with precision protease to generate GST-free b-catenin. Kinase reaction with b-catenin and GST-GSK-3b was performed in the presence or absence of His-Ajuba protein. CBB: Coomassie Brilliant Blue staining. (b) Promotion of GSK-3b-mediated phosphorylation of b-catenin and reduction of b-catenin levels by Ajuba expression. HEK293T cells were transfected with FLAG–Ajuba expression plasmid increasingly. Cell lysates were subjected to immunoprecipitation assay with anti-b-catenin antibody and immunoblot assay with the indicated antibodies. (c) Decrease of GSK3b-mediated phosphorylation of b-catenin and increase of b-catenin levels by Ajuba RNAi. HeLa cells were transfected with control or Ajuba siRNA. After 72 h, immunoprecipitation assay was performed with anti-b-catenin antibody and immunoblot assay with the indicated antibodies. Similar results were obtained with Ajuba 2 siRNA (data not shown). (d) Effect of Ajuba on TCF-mediated transcriptional activity enhanced by mutant b-catenin. Wild type (0.5 mg) or S37A b-catenin (0.4 mg) expression plasmid was transfected into HEK293T cells to activate TopFLASH reporter activity approximately as the same level. (e) Association of b-catenin and GSK-3b in Ajuba-overexpressed cells. Myc-b-catenin and HA-GSK-3b expression plasmids were transfected into HeLa cells with or without FLAG–Ajuba expression plasmid. Immunoprecipitation assay was performed with anti-Myc antibody and immunoblot assay with the indicated antibodies. Asterisk indicates IgG heavy chain. Oncogene

Ajuba and b-catenin phosphorylation in the Wnt pathway K Haraguchi et al

281

a β-catenin +

+

His-Ajuba

-

+

GST-GSK-3β

+

+

b

FLAG-Ajuba 0

0.05 0.1 1.5

5

(µg)

IB: FLAG β-catenin

32P

IB: β-catenin

β-catenin

Ly

IB: α-tubulin

CBB

His-Ajuba IB: β-catenin

GST-GSK3β

c

RNAi

IP: β-catenin

Cont

Ajuba 1

IB: p-β-catenin

IB: FLAG

IB: Ajuba

e

FLAG-Ajuba -

IB: β-catenin

+

Ly IB: Myc (β-catenin)

IB: α-tubulin

Ly

IB: FLAG (Ajuba) IB: β-catenin

IB: HA (GSK-3β)

IB: p-β-catenin

IP: β-catenin 1

IB: Myc (β-catenin)

0.59 IB: FLAG (Ajuba)

IB: Ajuba

*

IP: Myc

IB: HA (GSK-3β)

Luciferase activity (arbitrary unit)

d 10

5

1 β-catenin WT β-catenin S37A Ajuba

+

+

+

1 0.58 0.37

+

+

+

1 0.79 0.67

Oncogene

Ajuba and b-catenin phosphorylation in the Wnt pathway K Haraguchi et al

282

pathway (Miller and Moon, 1996; Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003). One of them, Axin, facilitates GSK3b-dependent phosphorylation of b-catenin by forming a complex with these molecules (Ikeda et al., 1998). Since the Ajuba-binding region in b-catenin is different from the Axin-binding region in b-catenin (Ikeda et al., 1998) and Axin did not prevent Ajuba from binding to b-catenin (Supplementary Figure S3), Ajuba might cooperate with Axin in b-catenin phosphorylation by forming a multi-protein complex. Indeed, we observed that Ajuba interacted with both b-catenin and GSK-3b through different regions, and that Axin was included in the anti-Ajuba immunoprecipitates (Figure 1a, Supplementary Figures S2 and S3). We found that exogenous expression of Ajuba suppressed b-catenin-TCF-dependent transcriptional activation and expression of cyclin D1 gene (Figure 3c–e). Cyclin D1 is one of the representative Wnt target genes, which is responsible for Wnt-dependent cell growth regulation (Shtutman et al., 1999; Tetsu and McCormick, 1999). By flow cytometric analysis, we observed that population of cells in G0/G1 phase was decreased and that of cells in S phase was increased in the absence of Ajuba (Figure 3e). These data suggested that Ajuba is involved in cell growth regulation by inhibiting the Wnt signaling pathway. Ajuba is expressed in embryonal cells and its expression increases during endodermal differentiation (Kanungo et al., 2000). In mouse tissues, Ajuba expresses ubiquitously in early embryonic stage and its expression becomes restricted to skin, uterus, kidney, brain (post-E12.5) as development is progressed (Goyal et al., 1999). Thus, we hypothesize that Ajuba may function in prematuring embryo and be involved in cell differentiation by regulating the Wnt signaling pathway. Although Ajubanull mice showed no obvious defect in development (Pratt et al., 2005), it is possible that other proteins such as Axin could compensate for the function of Ajuba in developmental stage. Wnt-3A stimulation stabilized b-catenin by inhibiting GSK-3b (Miller and Moon, 1996; Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003). In present data, we found that Ajuba was dephosphorylated and subsequently degraded through proteasome-dependent system after Wnt-3A stimulation (Figure 4). The data suggested that both activation of b-catenin and inactivation of Ajuba are important in the Wnt signaling pathway. Indeed, enforced expression of Ajuba inhibited the Wnt-3Ainduced activation of b-catenin (Figure 6), whereas depletion of Ajuba activated b-catenin–TCF reporter activity (Figure 3e). It should be noted that depletion of Ajuba did not induce further the activation of Wnt-induced reporter activity in the presence of Wnt stimulation (Supplementary Figure S1). In this case, depletion of Ajuba might have little additional effect on the Wnt signaling due to degradation of Ajuba (Figures 4a and c). Because Ajuba shuttles between the nucleus and the cytoplasm, and because we detected endogenous Ajuba Oncogene

both in the cytoplasm and the nucleus (Supplementary Figure S4; Kanungo et al., 2000), Ajuba might be involved in the transport of b-catenin from the nucleus to the cytoplasm, where b-catenin undergoes phosphorylation and degradation. Thus, not only cytoplasmic but also nuclear b-catenin levels might have decreased in Ajuba-overexpressed cells (Figures 3a and 6). Further studies are necessary to elucidate the mechanisms by which Ajuba transports b-catenin. Mutations of the Wnt pathway components such as APC, Axin and b-catenin were discovered in various types of tumors. Mutations in these genes are linked to inactivation of APC and Axin, or unscheduled activation of b-catenin (Miller and Moon, 1996; Cadigan and Nusse, 1997; Bienz and Clevers, 2000; Peifer and Polakis, 2000; Giles et al., 2003). Abnormal expression of the Wnt signaling components is also related to tumorigenesis. For instance, b-catenin-binding protein B9L, an activator of b-catenin-TCF-mediated transcription, was highly expressed in colorectal tumors (Adachi et al., 2004). Moreover, enhanced expression of LIM protein FHL2 was detected in hepatoblastoma cancer cells (Wei et al., 2003). Search for mutation or deregulation of Ajuba in tumors would be important in obtaining valuable insights into the mechanism of cancer development.

Materials and methods Cell culture HEK293T, HeLa, L and Plat-E cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Nissui, Tokyo, Japan) supplemented with 10% (v/v) fetal bovine serum (FBS). Control or Wnt-3A-CM was prepared from L or L-Wnt-3A cells following American Type Culture Collection’s (Manassas, VA, USA) protocol. Antibodies Polyclonal antibodies against Ajuba were prepared by immunizing rabbit with peptides containing amino acids 1–110 of Ajuba. Antibodies were purified by affinity chromatography. Monoclonal antibodies against b-catenin (BD, NJ, USA), a-tubulin (Sigma, St Louis, MO, USA), cyclin D1 (sc-450, Santa Cruz Biotechnology, CA, USA), FLAG (M2, Sigma), Myc (sc-40, Santa Cruz Biotechnology) and HA (Cell Signaling, Beverly, MA, USA), and polyclonal antibodies against FLAG (F.7425, Sigma), phosphorylated b-catenin (Ser 33) (sc-16743-R, Santa Cruz Biotechnology) and GFP (MBL, Nagoya, Japan) were also used. In vitro pull-down assay b-Catenin was synthesized with the TNT-Coupled Reticulocyte Lysate System (Promega, Tokyo, Japan). In vitrotranslated b-catenin and GST or GST–Ajuba purified from E. coli was incubated in buffer A (50 mM Tris-HCl, pH 7.4, 0.5% Triton X-100, 10% glycerol, 20 mM KCl, 2 mM MgCl2, 0.1 mM CaCl2) for 1 h at 41C. The protein complex was adsorbed to glutathione-Sepharose (Amersham Pharmacia Biotech, Little Chalfont, Buchinghamshire, UK) for 1 h at 41C. After the sepharose was washed with buffer A, samples were resolved by 10% SDS–PAGE.

Ajuba and b-catenin phosphorylation in the Wnt pathway K Haraguchi et al

283 Immunoprecipitation assay Cells were lysed in buffer B (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM ethylenediamine tetraacetic acid (EDTA), pH 7.5, 2 mM Na3VO4, 10 mM NaF, 0.1 mM phenylmethylsulphonyl fluoride (PMSF)) containing 1% Triton X-100. Lysates were incubated with antibodies for 1 h at 41C. Immunocomplexes were adsorbed to protein G Sepharose 4B (Amersham Pharmacia Biotech) for 1 h at 41C. After the sepharose was washed with buffer B containing 0.1% Triton X-100, samples were resolved by SDS–PAGE. In vitro kinase assay Kinase reactions were carried out at 301C for 10 min in buffer C (50 mM Tris-HCl, pH 7.5, 10 mM MgCl2, 1 mM dithiothreitol (DTT), 20 mM ATP) containing 10 mCi [g-32P] ATP (Amersham Pharmacia Biotech). Samples were resolved by SDS–PAGE and autoradiography. RNAi Small interfering RNAs were transfected with oligofectamine (Invitrogen Japan, Tokyo, Japan). Target sequences were Ajuba 1, 50 -GGACCGGGAUUAUCACUUUTT-30 and Ajuba 2, 50 -CCA AGUAUACUGUGUCACCTT-30 . Control siRNAs were from Euglena gracilis chloroplast DNA and target sequence was control 50 -UUCUCCGAACGUGUCACGUTT-30 . Luciferase assay Experiments were performed with a Dual-Luciferase Assay System (Promega) as previously described (Tago et al., 2000). All data are presented as the mean7s.d. of triplicate samples. Retroviral infection Retroviral vector was transiently transfected into Plat-E packaging cells with Lipofectamine Plus (Invitrogen Japan). After 24 h, cells were changed into a fresh medium and after an additional 24 h, the culture medium was collected and centrifuged at 2000 g for 30 min. The supernatant with 10 mg/

ml polybrene (Sigma) was added to L cells. The infected cells were selected by 2 mg/ml puromycin (Sigma) for 3 days. RT–PCR with real-time quantification Cells were suspended in ISOGEN (Nippon Gene, Tokyo, Japan) and total RNA was isolated. Reverse transcription was performed with oligo-dT primer and reverse transcriptase Superscript II (Invitrogen) to generate cDNA. Quantitative real-time PCR was carried out with SYBR Premix Ex Taq (Takara, Kyoto, Japan) and the indicated primers. Primer sequences are cyclin D1-FW, 50 -AAGGAGATTGTGCC ATCCAT-30 ; cyclin D1-RV, 50 -GCACTTCTGCTCCTCAC AGA-30 ; Ajuba-FW, 50 -TGGCTCTGTCTACTGTGAGG AA-30 ; Ajuba-RV, 50 -GCCACAGACACAGCACTTCT-30 ; HPRT-FW, 50 -GTAATGATCAGTCAACGGGGGAC-30 ; and HPRT-RV, 50 -CCAGCAAGCTTGCAACCTTAACCA-30 .

Abbreviations APC, adenomatous polyposis coli; GFP, green fluorescent protein; GSK-3b, glycogen synthase kinase-3b; GST, glutathione S-transferase; HA, hemagglutinin; LEF, lymphoid enhancing factor; PBS, phosphate-buffered saline; TCF, T-cell factor. Acknowledgements We thank T Nakamura for b-catenin plasmids, A Kikuchi for GSK-3b plasmids and T Kitamura for Plat-E cells. We also thank N Tokai-Nishizumi, T Suzuki, K Yokoyama and M Delawary for experimental suggestions. This work was supported by grants-in-aid from the Japan Society for the Promotion of Science and from the Ministry of Education, Cultures, Sports, Science and Technology, Japan.

References Adachi S, Jigami T, Yasui T, Nakano T, Ohwada S, Omori Y et al. (2004). Role of a BCL9-related b-catenin-binding protein, B9L, in tumorigenesis induced by aberrant activation of Wnt signaling. Cancer Res 64: 8496–8501. Bienz M, Clevers H. (2000). Linking colorectal cancer to Wnt signaling. Cell 103: 311–320. Cadigan KM, Nusse R. (1997). Wnt signaling: a common theme in animal development. Genes Dev 11: 3286–3305. Dawid IB, Breen JJ, Toyama R. (1998). LIM domains: multiple roles as adapters and functional modifiers in protein interactions. Trends Genet 14: 156–162. Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K, Tuschln T. (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411: 494–498. Filali M, Cheng N, Abbott D, Leontiev V, Engelhardt JF. (2002). Wnt-3A/b-catenin signaling induces transcription from the LEF-1 promoter. J Biol Chem 277: 33398–33410. Giles RH, van Es JH, Clevers H. (2003). Caught up in a Wnt storm: Wnt signaling in cancer. Biochim Biophys Acta 1653: 1–24. Goyal RK, Lin P, Kanungo J, Payne AS, Muslin AJ, Longmore GD. (1999). Ajuba, a novel LIM protein, interacts with Grb2, augments mitogen-activated protein kinase activity in fibroblasts, and promotes meiotic maturation of Xenopus oocytes in a Grb2- and Ras-dependent manner. Mol Cell Biol 19: 4379–4389. Ikeda S, Kishida S, Yamamoto H, Murai H, Koyama S, Kikuchi A. (1998). Axin, a negative regulator of the Wnt signaling pathway, forms a complex with GSK-3b and b-catenin and promotes

GSK-3b-dependent phosphorylation of b-catenin. EMBO J 17: 1371–1384. Kanungo J, Pratt SJ, Marie H, Longmore GD. (2000). Ajuba, a cytosolic LIM protein, shuttles into the nucleus and affects embryonal cell proliferation and fate decisions. Mol Biol Cell 11: 3299–3313. Labalette C, Renard CA, Neuveut C, Buendia MA, Wei Y. (2004). Interaction and functional cooperation between the LIM protein FHL2, CBP/p300, and b-catenin. Mol Cell Biol 24: 10689–10702. Levy L, Wei Y, Labalette C, Wu Y, Renard CA, Buendia MA et al. (2004). Acetylation of b-catenin by p300 regulates b-catenin–Tcf4 interaction. Mol Cell Biol 24: 3404–3414. Li L, Yuan H, Weaver CD, Mao J, Farr III GH, Sussman DJ et al. (1999). Axin and Frat1 interact with dvl and GSK, bridging Dvl to GSK in Wnt-mediated regulation of LEF-1. EMBO J 18: 4233–4240. Mao J, Wang J, Liu B, Pan W, Farr III GH, Flynn C et al. (2001). Low-density lipoprotein receptor-related protein-5 binds to Axin and regulates the canonical Wnt signaling pathway. Mol Cell 7: 801–809. Martin B, Schneider R, Janetzky S, Waibler Z, Pandur P, Kuhl M et al. (2002). The LIM-only protein FHL2 interacts with b-catenin and promotes differentiation of mouse myoblasts. J Cell Biol 159: 113–122. Miller JR, Moon RT. (1996). Signal transduction through b-catenin and specification of cell fate during embryogenesis. Genes Dev 10: 2527–2539. Oncogene

Ajuba and b-catenin phosphorylation in the Wnt pathway K Haraguchi et al

284 Peifer M, Polakis P. (2000). Wnt signaling in oncogenesis and embryogenesis – a look outside the nucleus. Science 287: 1606–1609. Pratt SJ, Epple H, Ward M, Feng Y, Braga VM, Longmore GD. (2005). The LIM protein Ajuba influences p130Cas localization and Rac1 activity during cell migration. J Cell Biol 168: 813–824. Quelle DE, Ashmun RA, Shurtleff SA, Kato JY, Bar-Sagi D, Roussel MF et al. (1993). Overexpression of mouse D-type cyclins accelerates G1 phase in rodent fibroblasts. Genes Dev 7: 1559–1571. Shibamoto S, Higano K, Takada R, Ito F, Takeichi M, Takada S. (1998). Cytoskeletal reorganization by soluble Wnt-3a protein signalling. Genes Cells 3: 659–670. Shtutman M, Zhurinsky J, Simcha I, Albanese C, D’Amico M, Pestell R et al. (1999). The cyclin D1 gene is a target of the b-catenin/LEF-1 pathway. Proc Natl Acad Sci USA 96: 5522–5527.

Tago K, Nakamura T, Nishita M, Hyodo J, Nagai S, Murata Y et al. (2000). Inhibition of Wnt signaling by ICAT, a novel b-catenininteracting protein. Genes Dev 14: 1741–1749. Tamai K, Zeng X, Liu C, Zhang X, Harada Y, Chang Z et al. (2004). A mechanism for Wnt coreceptor activation. Mol Cell 13: 149–156. Tetsu O, McCormick F. (1999). b-catenin regulates expression of cyclin D1 in colon carcinoma cells. Nature 398: 422–426. Tolwinski NS, Wehrli M, Rives A, Erdeniz N, DiNardo S, Wieschaus E. (2003). Wg/Wnt signal can be transmitted through arrow/ LRP5, 6 and Axin independently of Zw3/Gsk3b activity. Dev Cell 4: 407–418. Wei Y, Renard CA, Labalette C, Wu Y, Levy L, Neuveut C et al. (2003). Identification of the LIM protein FHL2 as a coactivator of b-catenin. J Biol Chem 278: 5188–5194.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc).

Oncogene