Site-specific casein kinase 1e-dependent phosphorylation of Dishevelled modulates b-catenin signaling Laura K. Klimowski 1, *, Benjamin A. Garcia 2,† , Jeffrey Shabanowitz 2 , Donald F. Hunt 2,3 and David M. Virshup 4 1 Department of Pharmacology and Toxicology, University of Utah, Salt Lake City, UT, USA 2 Department of Chemistry, University of Virginia, Charlottesville, VA, USA 3 Department of Pathology, University of Virginia, Charlottesville, VA, USA 4 Departments of Pediatrics and the Center for Children, Huntsman Cancer Institute, University of Utah, Salt Lake City, UT, USA Wnt signaling is required for normal development and proliferation, and its correct regulation is critical to pre- vent cancer. Mutations in genes encoding proteins regu- lating the Wnt cascade, such as adenomatous polyposis coli (APC), b-catenin and axin, are common in dysregu- lated development and multiple cancers [1,2]. These mutations are thought to function in part by impeding the basal degradation of b-catenin, which is crucial for correct cell regulation. The Dishevelled protein (Dvl) plays a central role in Wnt-regulated signaling path- ways [3,4]. One function is to inhibit the degradation of b-catenin through inhibition of glycogen synthase kinase 3 activity. Dvl is also downstream of Wnt in pathways regulating cell migration and activation of Ca 2+ -dependent signaling [5,6]. The mechanism by which Dvl transduces signals to these diverse effector pathways has been the object of intense study. Wnt signaling generates hyperphosphorylated forms of Dvl that sometimes but not always correlate with increased b-catenin in the nucleus and Lef-1-induced transcription Keywords b-catenin; casein kinase 1; Dishevelled; phosphopeptides; Wnt Correspondence D. M. Virshup, Huntsman Cancer Institute, University of Utah, Salt Lake City, UT 84103, USA Fax: +1 1 587 9415 Tel: +1 1 801585 3408 E-mail: david.virshup@hci.utah.edu Present address *University of Virginia, Department of Emer- gency Medicine, Charlottesville, VA, USA † Institute for Genomic Biology, University of Illinois, Champaign-Urbana, Urbana, IL, USA (Received 29 June 2006, revised 10 August 2006, accepted 11 August 2006) doi:10.1111/j.1742-4658.2006.05462.x Careful regulation of the Wnt–B-catenin signaling pathway is critical to many aspects of development and cancer. Casein kinase Ie is a Wnt-activa- ted positive regulator of this pathway. Members of the Dishevelled family have been identified as key substrates of casein kinase I (CKI). However, the specific sites phosphorylated in vivo by CKI and their relative import- ance in the physiologic regulation of these proteins in the canonical Wnt– b-catenin signaling pathway remain unclear. To address this question, recombinant mouse Dishevelled (mDvl-1) was phosphorylated by CKI in vitro and phosphorylation sites were identified by MS. CKI phosphoryla- tion of mDvl-1 at two highly conserved residues, serines 139 and 142, was observed by MS and confirmed by phosphopeptide mapping of in vivo phosphorylated protein. Phosphorylation of these sites is dependent on casein kinase I epsilon activity in vivo. Phenotypic analysis of mutant mDvl-1 indicates that phosphorylation of these sites stimulates the Dvl- activated b-catenin-dependent Wnt signaling pathway in both cell culture and in Xenopus development. Casein kinase I epsilon is a Wnt-regulated kinase, and regulated phosphorylation of Dvl allows fine tuning of the Wnt–b-catenin signaling pathway. Abbreviations CAD, collisionally activated dissociation; CKIe, casein kinase I epsilon; Dvl, Dishevelled; IMAC, immobilized metal affinity chromatography; LC-MS ⁄ MS, HPLC ESI and tandem MS; MBP, maltose binding protein; MT, myc-tagged; mDvl-1, mouse Dishevelled 1; XDsh, Xenopus Dishevelled. 4594 FEBS Journal 273 (2006) 4594–4602 ª 2006 The Authors Journal compilation ª 2006 FEBS [7,8]. Dvl phosphorylation may therefore be a critical step in signal transduction [4,9–11]. However, there are limited data on specific sites on Dvl that are phosphor- ylated in vivo. Casein kinase I epsilon (CKIe) is a positive regulator of the Wnt signaling pathway downstream of Wnt and upstream of Dvl, axin and b-catenin. CKIe and Dvl physically interact in vivo, and CKIe phosphory- lates Dvl in response to Wnt signaling [9–12]. Low-level expression of CKIe and Dvl synergize to both stabilize b-catenin and induce ectopic secondary axis formation during Xenopus embryogenesis [11,13]. Inhibition or RNA interference-induced knockdown of CKIe (or its Drosophila ortholog Dbt) blocks the effect of Wnt over- expression on signaling in mammalian and S2 cells in culture and on Xenopus axis formation during early development [9,10,13]. CKIe activity is rapidly upregu- lated in response to Wnt ligand, and constitutively act- ive CKIe is a more potent activator of transcription from a b-catenin-responsive promoter [14]. These data strongly suggest that CKIe is a positive regulator of the Wnt–b-catenin signaling pathway and Dvl is an import- ant CKIe substrate. Here, we identify specific sites that are phosphorylated on Dvl-1 by CKIe. HPLC ESI and tandem MS (LC-MS ⁄ MS) and immobilized metal affin- ity chromatography (IMAC LC-MS ⁄ MS) analysis of in vitro phosphorylated protein identified multiple phos- phorylation sites. In vivo phosphopeptide mapping combined with the use of a CKIe inhibitor confirmed that a subset of sites, serines 139 and 142 of Dvl-1, were also phosphorylated in vivo by CKIe. Mutation of these highly conserved sites resulted in a significant decrease in the activity of Dvl-1 in both cell culture and Xenopus axis duplication assays. The results provide molecular insights into the mechanism by which Wnt signaling to CKIe regulates the b-catenin signaling pathway. Results and Discussion CKIe phosphorylates mouse Dvl-1 (mDvl-1) in vitro and in vivo CKIe-specific phosphorylation sites on mDvl-1 were first identified using in vitro phosphorylation and LC-MS ⁄ MS analyses. Stoichiometric analysis using constitutively active CKIeD319 phosphorylation of recombinant maltose binding protein (MBP)–mDvl-1 in vitro [4 pmol of MBP–mDvl-1, 200 lm ATP, 5 mm dithiothreitol, 10 mm MgCl 2 , 100 nm His 6 -CKIeD319, incubated at 37 °C for 5–30 min] demonstrated that at least three phosphates are incorporated per mole of MBP–mDvl-1 (data not shown). To identify these CKI phosphorylation sites, full-length MBP–mDvl-1 was phosphorylated in vitro, separated by SDS ⁄ PAGE, trypsin digested and gel extracted. Initial LC-MS ⁄ MS analysis yielded approximately 75% coverage of mDvl-1, with identification of four phosphorylated pep- tides. Phosphorylated protease-digested peptides were isolated by Fe +3 IMAC and analyzed by LC-MS ⁄ MS to enrich for extremely low-level phosphopeptides. IMAC LC-MS ⁄ MS yielded two additional phospho- peptides (SDMpSAIVR and YASpSMLK). The stoi- chiometry of phosphorylation varied for the six IMAC-isolated peptide fragments (Table 1). These val- ues were derived from the relative percentage phos- phorylation determined for each peptide analyzed by LC-MS ⁄ MS, as phosphopeptide and nonphosphor- ylated forms generally do not have exactly the same ionization efficiencies. To determine the relative per- centage phosphorylation, the areas under the curve (total mass spectrum ion current signal) for all charge states of a given phosphopeptide during an LC- MS ⁄ MS experiment were added together and divided by the sum of the areas under the curve for all charge states of the nonphosphorylated and phosphorylated forms of that peptide [15]. Each of the identified phos- phorylation sites was further analyzed for biological relevance (see below), leading us to focus on a single peptide. This peptide, producing a 998.37 m ⁄ z precur- sor ion, was identified as corresponding to a doubly phosphorylated tryptic fragment, DGMDNETG- TES*MVS*HR(129–144). The singly phosphorylated y4 and y5 and doubly phosphorylated y6 and y9 daughter ions revealed serines 139 and 142 (S139 ⁄ 142) as putative CKI phosphorylation sites (Fig. 1A). The Dvl family is encoded by three distinct genes in vertebrates, designated Dvl-1, Dvl-2 and Dvl-3. Amino acid sequence alignment of these peptide fragments indicated that serines 119, 139, 142, 384, 385, 416 and 473 were highly conserved between Dvl isoforms (Dvl-1, Dvl-2 and Dvl-3) and species (mouse, human and Xenopus). The phosphopeptide serine residues 632 and 635 were conserved between species, but only within the Dvl-1 isoform. CKI preferentially phosphorylates serine or threonine residues preceded by acidic or phosphorylated residues. The only in vitro phosphoryl- ated peptide containing highly conserved potential phosphorylation sites that are preceded by multiple acidic residues is DGMDNETGTES*MVS*HR(129– 144), where serines 139 and 142 are phosphorylated (Fig. 1B). Phosphorylation of serine 139 may be pro- moted by the presence of upstream highly conserved acidic residues. Phosphoserine 139 could then create a recognition site for phosphorylation of serine 142 in the motif pSXXS [16–18]. This conservation of CKI phosphorylation sites across species supports the L. K. Klimowski et al. Regulation of Dvl-1 by CKIe FEBS Journal 273 (2006) 4594–4602 ª 2006 The Authors Journal compilation ª 2006 FEBS 4595 possibility that mDvl-1 serines 139 and 142 may be important regulatory sites for Dvl-dependent Wnt sig- nal transduction. In vitro phosphorylation of purified proteins can be a useful method to determine phosphorylation sites, but their biological relevance must be validated in vivo. To determine how many sites on Dvl-1 were in fact phosphorylated in vivo by CKIe, we compared in vivo phosphopeptide maps of Dvl-1 from cells metabolically labeled in the presence or absence of the CKId ⁄ e inhibitor IC-261 [19]. Multiple phosphopeptides were generated by tryptic digest of the in vivo phosphorylat- ed mDvl-1, indicating multiple phosphorylation sites. Four of these sites appear to be CKId ⁄ e sites, as four major phosphopeptides (designated a–d in Fig. 2A,B) disappeared after CKId ⁄ e inhibition. We then system- atically compared phosphopeptide maps of 32 P-labeled mDvl-1 wild-type and the corresponding mutants, mDvl-1 S119A, S139 ⁄ 142A, S384 ⁄ 385A, S416A and S473A immunoprecipitated from metabolically labeled HEK293 cells. Peptides labeled a and b are present in maps from wild-type, but not mutant, mDvl-1 S139 ⁄ 142A, indicating that these sites are phosphoryl- ated in vivo. Several reasons may account for the pres- ence of two related phosphopeptides a and b. First, the trypsin-digested fragment, DGMDNETGTES- MVSHR(129–144), may exist in vivo as both mono- phosphorylated and diphosphorylated peptides. Second, trypsin cleavage is efficient, but after the ini- tial cleavage at the alternative site (between the two basic residues), no further cleavage occurs by the endo- protease [20]. Partial digestion between the C-terminal di-arginine, DGMDNETGTESMVSHR Ú R(129–145), may therefore result in two peptides differing only by one arginine residue. Methionine oxidation (+16 Da) was observed on both methionine residues on this pep- tide. This oxidation is a regular consequence of the in-gel digestion procedures, as formic acid (used in the in-gel peptide extraction protocol) is a strong acid and oxidizes Met residues. Nevertheless, both the missed cleavage and oxidized peptides containing residues 129–145 are minor species compared to the phospho- tryptic peptide shown in Fig. 1A. The IC-261-depend- ent loss of peptides c and d suggests that additional CKIe sites are present in vivo, as has been suggested by previous studies [9]. Unfortunately, peptides c and d did not correlate with the peptides listed in Table 1. Although extensive efforts were made to isolate sites a–d using LC-MS ⁄ MS and IMAC LC-MS ⁄ MS analy- ses of in vivo phosphorylated Dvl-1, recovery of these phosphopeptides was not achieved. Phosphoserines 139 and 142 are positive regulators of Dvl-dependent Wnt signal transduction We next assessed the functional importance of the phos- phoserine 139 and 142 sites on Dvl. Overexpression of Dvl-1 in HEK293 cells activates b-catenin-dependent signal transduction. If these CKIe-phosphorylation sites are important for Wnt signal transduction, we hypothesized that mutating serines 139 and 142 to ala- nines would diminish the intensity of signaling in HEK293 cells. b-Catenin ⁄ Lef-1-dependent transcription was assessed using the luciferase reporter plasmid pTOPFLASH. These experiments were performed three Table 1. Identification of in vitro casein kinase I (CKI) phosphorylation sites on Dishevelled 1 (Dvl-1). Phosphorylated serine residues are followed by an asterisk. Percentage peptide phosphorylated in vitro represents the sum of the area under the curve (total mass spectrum ion current signal) for all charge states of a given phosphopeptide during a normal LC-MS ⁄ MS experiment were divided by the sum of the area under the curve for all charge states of the nonphosphorylated and phosphorylated forms of that peptide. This should only be consid- ered a relative approximation of percentage phosphorylation, as phosphopeptides and nonphosphorylated forms generally do not have the exact same ionization efficiencies. Change in Dvl biological activity: mutants were tested for their ability to alter axis specification in Xenopus laevis (XL) and transactivation of a b-catenin-responsive promoter in HEK293 cells (293) as described in Materials and methods. Conserved species homology: indicates if the sites are conserved between human (H), mouse (M) and Xenopus (X). mDv1-1, mouse Dishevelled 1; Dvl, Dishevelled; IMAC, immobilized metal affinity chromatography. ND, not detected. mDvl-1 tryptic phosphopeptides phosphorylated by CKIe in vitro Residues phosphorylated Percentage of peptide phosphorylated in vitro Change in in vivo phosphopeptide map Change in Dvl biological activity Conserved species homology TGGIGDSRPPS*FHPNVASSR DGMDNETGTES*MVS*HR YGTSPC(SS)*AITR SDMS*AIVR YASS*MLK S119 S139, S142 S384, S385 S416 S473 49 19 27 IMAC IMAC ND Yes; see Fig. 2 ND ND ND ND (XL, 293) Yes (XL, 293) ND (XL, 293) ND (293) Not tested M, H, X M, H, X M, H, X M, H, X M, H, X S*QAS*AVAPGLPPLHPLTK S632, S635 10 Not tested ND (293) M, H, X Only Dvl-1 Regulation of Dvl-1 by CKIe L. K. Klimowski et al. 4596 FEBS Journal 273 (2006) 4594–4602 ª 2006 The Authors Journal compilation ª 2006 FEBS separate times with three replicates in each experiment (n ¼ 3). b-Catenin ⁄ Lef-1-dependent transcription acti- vation was reproducibly two-fold greater with expres- sion of mDvl-1 wild-type compared to mDvl-1 S139 ⁄ 142A (Fig. 3A). The diminished transactivation by the S139 ⁄ 142A mutant was not due to decreased mDvl-1 protein expression, because immunoblot analy- sis indicated that protein expression levels were equal if not greater in the mutant than in the wild-type at all concentrations tested (Fig. 3B). Dvl-1 undergoes a phosphorylation-dependent elec- trophoretic mobility shift in response to both Wnt signaling and CKIe expression [9,10,12]. Mutation of mDvl-1 serines 139 and 142 to alanine significantly altered the CKIe-dependent electrophoretic mobility shift, blocking the appearance of the slowest-migra- ting species. These data are consistent with the con- clusion that serines 139 and 142 are phosphorylated in vivo by CKIe. The incomplete abrogation of the CKIe-induced electrophoretic mobility shift (Fig. 3C, compare lanes 3 and 4) is also consistent with the phosphopeptide mapping data (Fig. 2A,C), which sug- gests that CKIe phosphorylates additional sites on Dvl-1. mDvl-1 serines 139 and 142 modulate secondary axis formation Previous studies have shown that CKIe synergizes with Xenopus Dishevelled (XDsh) in the development of the secondary axis phenotype in Xenopus. As mutation of the CKIe phosphorylation sites at mDvl-1 serines 139 Fig. 1. Identification of mouse Dishevelled 1 (mDvl-1) serines 139 and 142 as in vitro casein kinase I (CKI) phosphorylation sites. (A) IMAC LC-MS ⁄ MS spectrum of the phos- pho-mDvl-1 precursor ion demonstrates phosphorylation on mDvl-1 serines 139 and 142. IMAC LC-MS ⁄ MS identified a promin- ent species at m ⁄ z 998.37 (doubly charged  1995 m ⁄ 2z) corresponding to the indica- ted tryptic phosphopeptide. Internal and C-terminal carboxylates were converted to methyl esters prior to analysis. Subsequent fragmentation revealed multiple phosphoryl- ated daughter ions as indicated (bold under- lined), consistent with the structure DGMDNETGTEpSMVpSHR. (B) Serines 139 and 142 are conserved in vertebrate Dvl pro- teins. The relevant region of the indicated vertebrate Dvl1–Dvl13 were aligned using CLUSTALW. Invariant residues are shown in black boxes, similar residues are shown in gray boxes with white letters, and highly conserved residues are shown in gray boxes with black letters. Acidic residues N-terminal to the putative phosphorylation sites are also highly conserved. L. K. Klimowski et al. Regulation of Dvl-1 by CKIe FEBS Journal 273 (2006) 4594–4602 ª 2006 The Authors Journal compilation ª 2006 FEBS 4597 and 142 resulted in attenuated canonical Wnt signal transduction in cell culture, we hypothesized that the S139 ⁄ 142A mutation would partially inhibit secondary axis formation in Xenopus development. First, we examined the synergistic effect of coinjection of low levels of CKIe with wild-type or phosphorylation-site mutant XDsh into Xenopus ventral blastomeres at the four-cell stage (n ¼ 30). Injection of 100 pg of control green fluorescent protein mRNA produced no pheno- type. Fifty picograms of CKIe or myc-tagged (MT)XDsh mRNA alone produced a minimal pheno- type. However, coinjection of 50 pg each of CKIe and MTXDsh wild-type resulted in 81% secondary axis formation in Xenopus embryos at day 1. In contrast, coinjection of 50 pg each of CKIe and MTXDsh S139 ⁄ 142A resulted in 28% less secondary axis forma- tion (Fig. 4, top). Coinjection of 50 pg of CKIe K38A with MTXDsh or MTXDsh S139 ⁄ 142A did not signifi- cantly increase the secondary axis phenotype. Coinjec- tion experiments with IC-261 were not performed. Western blot analysis of embryo lysates at stage 21 indicated equivalent levels of MTXDsh wild-type and S139 ⁄ 142A protein, confirming that the difference was not due simply to changes in XDsh expression levels (Fig. 4, bottom). We next examined the ability of high ectopic expres- sion levels of XDsh wild-type and XDsh S139 ⁄ 142A mRNA to induce secondary axis formation. For this experiment, we injected either 1 ng of MTXDsh wild- type or 1 ng of MTXDsh S139 ⁄ 142A mRNA into the ventral blastomere at the four-cell stage (n ¼ 30). Embryos were scored for phenotypes after stage 21 (day 1). Secondary axis development was 8% higher in embryos injected with MTXDsh wild-type mRNA than in the mutant, and incomplete axis development was 10% higher in MTXDsh wild-type embyros than in the mutant (data not shown; P-value < 0.3). These results were not statistically significant, and suggest that high levels of ectopic XDsh expression compen- sate for the activity change regulated by phosphoryla- tion of serines 139 and 142. Taken together, our data suggest that CKIe phosphorylation of Dvl at S139 and S142 positively regulates Dvl activity in both the mam- malian and Xenopus systems. A recent study examined mutation of candidate CKI phosphorylation sites in Dvl-1 by replacement with acidic residues [13]. In particular, a multiple mutant with replacement of threonines 135 and 137, as well as serines 139 and 142, by negatively charged aspartic acid residues did not induce Wnt–b-catenin ⁄ Lef-1 tran- scription activation. Our phosphopeptide maps suggest that threonines 135 and 137 are not phosphorylated in vivo, at least under basal conditions. If the CKIe- phosphorylated tryptic peptides had additionally been phosphorylated on theonines 135 and 137, we would have expected a shift in the location of these AB C Fig. 2. Mouse Dishevelled 1 (mDvl-1) serines 139 and 142 are phosphorylated in vivo by casein kinase I (CKI). Two-dimensional tryptic phos- phopeptide maps of Myc epitope (MT)–mDvl-1 labeled in vivo with [ 32 P]orthophosphate. (A) 0.5 lg of MT mouse Dishevelled 1 (mDvl-1) wild-type; (B) 0.5 lg of MT–mDvl-1 S139 ⁄ 142A; and (C) 0.5 lg of MT–mDvl-1 wild-type in the presence of the casein kinase I (CKI) inhibitor IC-261 (20 l M for the last hour of labeling). Phosphopeptides a and b are absent in both MT–mDvl 1S139 ⁄ 142A and in the MT–mDvl-1 wild- type sample treated with the CKI inhibitor IC-261. IC-261 treatment also leads to loss of the phosphopeptides indicated by boxes c and d in Fig. 2B, suggesting the presence of additional CKI sites on Dvl-1. Regulation of Dvl-1 by CKIe L. K. Klimowski et al. 4598 FEBS Journal 273 (2006) 4594–4602 ª 2006 The Authors Journal compilation ª 2006 FEBS 32 P-labeled peptides rather than their complete dis- appearance after mutation of serines 139 and 142 (Fig. 2A,B). However, consistent with this report [13], we also found that mutation of mDvl-1 serines 139 and 142 to aspartic acids did not induce activation of Wnt–b-catenin ⁄ Lef-1 transcription (data not shown). Therefore, negatively charged residues surrounding and including serines 139 and 142 are not sufficient to promote Dvl activity in b-catenin ⁄ Lef-1 transcription. Taken together, our data suggest that the primary function of CKIe phosphorylation of Dvl at serines 139 and 142 is to regulate the intensity of Wnt signal transduction. The LC-MS ⁄ MS approach was successful in identify- ing several CKI-specific in vitro phosphorylation sites on mDvl-1. The results of further analysis using peptide mapping and biological assays suggest that only two of these sites were physiologically relevant, given the limi- tations of the assays. In vitro phosphorylation events may not reflect in vivo phosphorylation, due to multiple factors, including differences in concentrations of the substrate and kinase, the presence of associated regula- tory proteins in the cell, activity of protein phosphatas- es, and the need in some cases for prior phosphorylation by ‘priming’ kinases that generate CKI recognition sites. Furthermore, the IMAC MS method may enrich subsets of phosphopeptides and fail to isolate others. CKIe and Dvl play multiple roles as regulators of Wnt signaling pathways. Our data suggest that one function of CKIe is to phosphorylate positive regula- tory sites on Dvl-1. The consequence of these specific phosphorylation events is significantly enhanced b-catenin-dependent signaling. Wnt signaling, by acti- A 0 2 4 6 8 10 12 14 16 0.00 0.25 0.50 1.00 Fold increase Lef-1:Luciferase µg transfected MT-mDvl-1 C + - + - - + - MT-mDvl-1 wt - + - + - + - MT-mDvl-1 S139/142A CKIε K38A - - - - + + - CKIε wt + + - - - - - MT-mDvl-1 β-catenin 1 2 3 4 5 6 7 B actin MT-mDvl-1 S139/142A actin MT-mDvl-1 wt 1 2 3 4 1 2 3 4 mDvl-1 wt mDvl-1 S139/142A Fig. 3. Casein kinase I (CKI) phosphorylation at mouse Dishevelled 1 (mDvl-1) serines 139 and 142 contributes to Wnt signal transduc- tion. (A) Mutation of serines 139 and 142 attenuates the ability of mDvl-1 to activate transcription from a Wnt-responsive promoter. Increasing concentrations of Myc epitope (MT)–mDvl-1 wild-type or MT–mDvl-1 S139 ⁄ 142A were transiently expressed in HEK293 cells and assayed for transactivation of a Lef-1-responsive reporter. Loss of phosphorylation sites S139 and S142 caused approximately a two-fold decrease in transcriptional activation compared with MT–mDvl-1 wild-type (P-values < 0.01). (B) Mutation of serines 139 and 142 does not alter the abundance of mDvl-1 in transfected cells. HEK293 lysates were probed with anti-Myc antibody and b-actin. The figure shows samples from the experiment in Fig. 3A, with all lanes from the same immunoblot. Lanes 1–3, 0.25–1.0 lg of MT– mDvl-1; lane 4, 1.0 lg of MT empty vector. (C) Mutation of mDvl-1 serines 139 and 142 decreases the CKIe-induced electrophoretic mobility shift. MT–mDvl-1 wild-type (0.1 lg) and MT–mDvl-1 S139 ⁄ 142A (0.1 lg) were transiently expressed in HEK293 cells along with wild-type CKIe (0.5 lg) or dominant negative CKIe (K38A) as indicated. XDsh S139/142A + CKIε XDsh wild type + CKIε XDsh S139/142A XDsh wild type GFP CKIε percent phenotype 020406080100 complete secondary axis incomplete secondary axis MT-mDvl-1 Actin Fig. 4. Dsh S139 ⁄ 142A mRNA attenuates double axis formation in Xenopus day 1 embryos. Top: Xenopus embryos at the four-cell stage were injected ventrally with low doses (50 pg each) of casein kinase I (CKIe) and MTXDsh mRNA that individually do not induce secondary axis formation. Where necessary, mRNA for green fluor- escent protein (GFP) was coinjected so that 100 pg of mRNA was always used. Coinjection of CKIe with MTXDsh wild-type mRNA synergistically induces secondary axis formation (Fisher’s exact test, P-values < 0.00005). The percentage of secondary axis pheno- type is attenuated in Xenopus embryos coinjected with CKIe (50 pg) and MTXDsh S139 ⁄ 142A (50 pg) mRNA [Fisher’s exact test (one-sided), P-values < 0.023]. Bottom: Western blot analysis of 25 lg of total protein from embryo lysates at stage 21; anti-Myc, top; anti-b-actin, below. Lane 1, uninjected; lane 2, CKIe; lane 3, MTXDsh wild-type; lane 4, MTXDsh S139 ⁄ 142A; lane 5, MTXDsh wild-type and CKIe; lane 6, MTXDsh S139 ⁄ 142A and CKIe. L. K. Klimowski et al. Regulation of Dvl-1 by CKIe FEBS Journal 273 (2006) 4594–4602 ª 2006 The Authors Journal compilation ª 2006 FEBS 4599 vating CKIe and facilitating Dvl-1 phosphorylation at serines 139 and 142, may be one of several mechanisms to modulate and reinforce signal strength. This com- plex regulation of Dvl-1 function may be critical to obtain the diverse effects seen from Wnt signaling in different tissues. The quantitative changes in Dvl-1 function and resultant b-catenin signal intensity dem- onstrated here may regulate the shape and extent of signaling gradients and are therefore likely to be crit- ical in the correct establishment of proliferative zones and developmental boundaries. Materials and methods Plasmids and antibodies The expression plasmid pCS2-Myc-mDvl-1 encodes six cop- ies of the Myc epitope (MT) fused at the N-terminus to mDvl-1 (MT-mDvl-1). Point mutants were constructed by site-directed mutagenesis, and all mutants were resequenced to confirm the presence of the desired mutation and absence of unplanned mutations [12]. TOPFLASH, pEV3S-Lef-1 and pRL-SV40 were kindly provided by H. Clevers (Robert- Roessle-Klinik, Berlin, Germany), M. Waterman (University of California, Irvine, CA, USA), and D. Ayer (University of Utah, Salt Lake City, UT, USA), respectively. Anti-Myc monoclonal antibody (9E10) was obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). In vitro phosphorylation of mDvl-1 MBP–mDvl-1 and constitutively active CKI were purified essentially as previously described [12,21]. In vitro kinase reactions were performed using 4 pmol of MBP–mDvl-1, 200 lm ATP, 5 mm dithiothreitol, 10 mm MgCl 2 , and 100 nm His 6 -CKID319, and incubated at 37 °C for 5– 30 min. The reactions were terminated by addition of SDS loading buffer and heating at 100 °C for 5 min. Samples were analyzed by SDS ⁄ PAGE. Sample preparation and MS analysis SDS ⁄ PAGE gel bands of phosphorylated MBP–mDvl-1 were sliced into 1 mm cubes and washed with 100 lLof 0.1 m ammonium bicarbonate buffer (pH 8.0). The diced cubes were reduced and alkylated using 10 mm dithiothrei- tol and 50 mm iodoacetamide, respectively. The gel pieces were then dehydrated and reswollen in a minimal volume of 0.1 m ammonium bicarbonate buffer containing 12.5 ngÆlL )1 trypsin (Roche, Indianapolis, IN) and allowed to digest overnight at room temperature. Peptides were extracted with 100 lL of 50% acetonitrile containing 5.0% formic acid for 15 min (two cycles), followed by a further extraction with 100% acetonitrile. The extracted peptide solutions were then concentrated to a volume of about 1.0 lL. The samples were brought up to 20 lL with 0.1% acetic acid for MS analysis using conventional C18 liquid chromatography. Each gel band solution sample was loaded onto a 360 lm outside diameter (o.d.) · 75 lm inside diameter (i.d.) microcapillary fused silica column packed with C18 irregular 5–20 lm resin (Polymicro Technologies, Phoenix, AZ). After sample loading, the precolumn was washed with 0.1% acetic acid for 15 min to remove any buffer salts. The precolumn was then connected to a 360 lm o.d. · 50 lm i.d. analytical column (Polymicro Technologies) packed with C18 regular 5 lm resin constructed with an integrated electrospray emitter tip [22]. Additionally, IMAC was util- ized to enrich for phosphopeptides [23]. Samples were first prepared for IMAC by converting peptides to their corres- ponding methyl esters by addition of 100 lL of methyl ester reagent (160 lL of acetyl chloride in 1 mL of MeOH) and incubation at room temperature for 1 h. Methanolic HCl was used to reduce nonspecific binding of acidic pep- tides. Peptide samples were then dried and resuspended in a 1 : 1 : 1 mixture of MeOH ⁄ MeCN ⁄ 0.1% acetic acid. IMAC columns were constructed by packing capillary col- umns (360 o.d. · 100 i.d.) with 8 cm POROS 20 MC (Applied Biosystems, Framingham, MA). IMAC columns were first activated with a 100 mm FeCl 3 solution (Aldrich, Milwaukee, WI). Samples were loaded onto the IMAC col- umns and washed with several column volumes of 0.01% acetic acid. After phosphopeptides were bound to the columns, they were eluted with 250 mm Na 2 HPO 4 (pH 6) (Aldrich) onto C18 packed capillary precolumns. Samples were then gradient eluted (Agilent 1100 Series; Santa Clara, CA) directly into a Finnigan LCQ quadrupole ion trap mass spectrometer (Thermo Electron, San Jose, CA) at a flow rate of 60 nLÆmin )1 . The nano-flow HPLC gradient used was 0–60% acetonitrile in 0.1% acetic acid in 90 min. The ion trap mass spectrometer was operated in the data- dependent mode, where an initial MS scan recorded the m ⁄ z ratios of parent ions over the mass range 300– 2000 Da. The five most abundant ions were then selected for subsequent collisionally activated dissociation (CAD) and an MS ⁄ MS spectrum was recorded. All MS ⁄ MS data were searched using the sequest program and validated manually. In vivo assays LEF1-luciferase reporter assays were performed as previ- ously described [14], with pCS2-MTmDvl-1 wild-type, MTmDvl-1 S139 ⁄ 142A ⁄ D ⁄ E, or MT empty vector (0.025– 1 lg) transfected into HEK293 cells in 35 mm dishes along with TOPFLASH (500 ng), pEV3S-Lef-1 (100 ng), and pRL-SV40 (100 ng). Experiments were performed three sep- arate times with three replicates (n ¼ 3) in each experiment. Regulation of Dvl-1 by CKIe L. K. Klimowski et al. 4600 FEBS Journal 273 (2006) 4594–4602 ª 2006 The Authors Journal compilation ª 2006 FEBS Data are presented as mean ± SD from three separate wells. Phosphopeptide maps of immunoprecipitated 32 P-labeled MT–mDvl-1 wild-type and MT–mDvl-1 S139 ⁄ 142A were performed in HEK293 cells as previously described [12]. For the inhibitor studies, 20 lm IC-261, a CKIe ⁄ d-selective inhib- itor [19], was added for the final hour of labeling. Xenopus embryo injections and analysis of phenotypes were performed as previously described [14]. MTXDsh cDNA was obtained from S Sokol (Department of Micro- biology and Molecular Gentics, Harvard Medical School). The XDsh sequence is most closely related to other verteb- rate Dvl-2 sequences (see Fig. 1B). Embryos were scored for secondary axis formation at stage 20, approximately 22 h postinjection. Acknowledgements We thank Kenneth Boucher for statistical analysis, and members of the Virshup and Yost laboratories for technical advice and helpful discussions. These studies were supported by The Huntsman Cancer Foundation, the Willard Snow Hansen Chair in Cancer Research (DMV), NIH R01 CA-71074 and P01 CA73992 (DMV) and GM 37537 (DFH). References 1 Moon RT, Kohn AD, De Ferrari GV & Kaykas A (2004) WNT and beta-catenin signalling: diseases and therapies. Nat Rev Genet 5, 691–701. 2 Polakis P (2000) Wnt signaling and cancer. Genes Dev 14, 1837–1851. 3 Sussman DJ, Klingensmith J, Salinas P, Adams PS, Nusse R & Perrimon N (1994) Isolation and characteri- zation of a mouse homolog of the Drosophila segment polarity gene dishevelled. Dev Biol 166, 73–86. 4 Yanagawa S, van Leeuwen F, Wodarz A, Klingensmith J & Nusse R (1995) The dishevelled protein is modified by wingless signaling in Drosophila. Genes Dev 9, 1087– 1097. 5 Kuhl M, Sheldahl LC, Park M, Miller JR & Moon RT (2000) The Wnt ⁄ Ca 2+ pathway: a new vertebrate Wnt signaling pathway takes shape. Trends Genet 16, 279–283. 6 Endo Y, Wolf V, Muraiso K, Kamijo K, Soon L, Uren A, Barshishat-Kupper M & Rubin JS (2005) Wnt-3a- dependent cell motility involves RhoA activation and is specifically regulated by dishevelled-2. J Biol Chem 280 , 777–786. 7 Sokol SY (1996) Analysis of Dishevelled signalling path- ways during Xenopus development. Curr Biol 6, 1456– 1467. 8 Gonzalez-Sancho JM, Brennan KR, Castelo-Soccio LA & Brown AM (2004) Wnt proteins induce dishevelled phosphorylation via an LRP5 ⁄ 6-independent mechan- ism, irrespective of their ability to stabilize beta-catenin. Mol Cell Biol 24, 4757–4768. 9 Hino S, Michiue T, Asashima M & Kikuchi A (2003) Casein kinase I epsilon enhances the binding of Dvl-1 to Frat-1 and is essential for Wnt-3a- induced accumulation of beta-catenin. J Biol Chem 278, 14066–14073. 10 Peters JM, McKay RM, McKay JP & Graff JM (1999) Casein kinase I transduces Wnt signals. Nature 401, 345–350. 11 Kishida M, Hino Si S, Michiue T, Yamamoto H, Kishida S, Fukui A, Asashima M & Kikuchi A (2001) Synergistic activation of the Wnt signaling pathway by Dvl and casein kinase Iepsilon. J Biol Chem 276 , 33147– 33155. 12 Gao Z-H, Seeling JM, Hill V, Yochum A & Virshup DM (2002) Casein kinase I phosphorylates and destabi- lizes the b-catenin degradation complex. Proc Natl Acad Sci USA 99, 1182–1187. 13 Cong F, Schweizer L & Varmus H (2004) Casein kinase Iepsilon modulates the signaling specificities of dishev- elled. Mol Cell Biol 24, 2000–2011. 14 Swiatek W, Tsai IC, Klimowski L, Pepler A, Barnette J, Yost HJ & Virshup DM (2004) Regulation of casein kinase I epsilon activity by Wnt signaling. J Biol Chem 279, 13011–13017. 15 Chrestensen CA, Schroeder MJ, Shabanowitz J, Hunt DF, Pelo JW, Worthington MT & Sturgill TW (2004) MAPKAP kinase 2 phosphorylates tristetraprolin on in vivo sites including Ser178, a site required for 14–3)3 binding. J Biol Chem 279, 10176–10184. 16 Marin O, Bustos VH, Cesaro L, Meggio F, Pagano MA, Antonelli M, Allende CC, Pinna LA & Allende JE (2003) A noncanonical sequence phosphorylated by casein kinase 1 in beta-catenin may play a role in casein kinase 1 targeting of important signaling proteins. Proc Natl Acad Sci USA 100, 10193–10200. 17 Kennelly PJ & Krebs EG (1991) Consensus sequences as substrate specificity determinants for protein kinases and protein phosphatases. J Biol Chem 266, 15555– 15558. 18 Vielhaber EL & Virshup DM (2001) Casein kinase I: from obscurity to center stage. IUBMB Life 51, 273–278. 19 Mashhoon N, DeMaggio AJ, Tereshko V, Bergmeier SC, Egli M, Hoekstra MF & Kuret J (2000) Crystal structure of a conformation-selective casein kinase-1 inhibitor. J Biol Chem 275, 20052–20060. 20 Hardie DG (1999) Protein Phosphorylation: a Practical Approach, 2nd edn. Oxford University Press, Oxford. 21 Salic A, Lee E, Mayer L & Kirschner MW (2000) Con- trol of b-catenin stability: reconstitution of the cytoplas- mic steps of the Wnt pathway in Xenopus egg extracts. Mol Cell 5, 523–532. L. K. Klimowski et al. Regulation of Dvl-1 by CKIe FEBS Journal 273 (2006) 4594–4602 ª 2006 The Authors Journal compilation ª 2006 FEBS 4601 22 Martin SE, Shabanowitz J, Hunt DF & Marto JA (2000) Subfemtomole MS and MS ⁄ MS peptide sequence analysis using nano-HPLC micro-ESI fourier transform ion cyclotron resonance mass spectrometry. Anal Chem 72, 4266–4274. 23 Ficarro SB, McCleland ML, Stukenberg PT, Burke DJ, Ross MM, Shabanowitz J, Hunt DF & White FM (2002) Phosphoproteome analysis by mass spectrometry and its application to Saccharomyces cerevisiae. Nat Biotechnol 20, 301–305. 4602 FEBS Journal 273 (2006) 4594–4602 ª 2006 The Authors Journal compilation ª 2006 FEBS Regulation of Dvl-1 by CKIe L. K. Klimowski et al. . Site-specific casein kinase 1e-dependent phosphorylation of Dishevelled modulates b-catenin signaling Laura K. Klimowski 1, *, Benjamin A. Garcia 2,† ,. regulation of the Wnt B-catenin signaling pathway is critical to many aspects of development and cancer. Casein kinase Ie is a Wnt-activa- ted positive regulator of this pathway. Members of the Dishevelled. culture and in Xenopus development. Casein kinase I epsilon is a Wnt-regulated kinase, and regulated phosphorylation of Dvl allows fine tuning of the Wnt b-catenin signaling pathway. Abbreviations CAD,