SEARCH

SEARCH BY CITATION

Keywords:

  • XARP;
  • centrosome;
  • Xenopus;
  • DIX;
  • Dsh;
  • Axin;
  • Wnt

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

The Wnt/β-catenin signaling pathway regulates cell proliferation and cell fate determination in multiple systems. However, the subcellular localization of Wnt pathway components and the significance of this localization for the pathway regulation have not been extensively analyzed. Here we report that Xenopus Axin-related protein (XARP), a component of the β-catenin destruction complex, is localized to the centrosome. This localization of XARP requires the presence of the DIX domain and an adjacent region. Since other components of the Wnt pathway have also been shown to associate with the centrosome, we tested a hypothesis that the β-catenin destruction complex operates at the centrosome. However, XARP mutants with poor centrosomal localization revealed an enhanced rather than decreased ability to antagonize the Wnt/β-catenin pathway. Our data are consistent with the idea that the inactivation of XARP at the centrosome is an important regulatory point in Wnt signaling. Developmental Dynamics 239:261–270, 2010. © 2009 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

The canonical Wnt/β-catenin pathway is a major signaling pathway that regulates cell proliferation and differentiation at different times during embryonic development and adult tissue homeostasis (Peifer and Polakis,2000; Moon et al.,2002; Nusse,2005). The key signaling step involves stabilization of β-catenin, which, in the absence of pathway activation, is rapidly turned over by the ubiquitin-proteasome pathway. β-catenin is marked for proteasomal degradation by the β-catenin destruction complex that consists of glycogen synthase kinase 3β (GSK3β) and two scaffold proteins: Adenomatous polyposis coli (APC) and Axin (Ikeda et al.,1998; Itoh et al,1998; Kimelman and Xu,2006). When secreted Wnt ligands stimulate the pathway by binding to seven-transmembrane receptors Frizzled (Fz) and co-receptors LRP5/6, the cytoplasmic protein Disheveled (Dsh) inhibits β-catenin degradation. Accumulated β-catenin, together with LEF/TCF family transcription factors, stimulates target gene transcription (Arce et al.,2006). Although the basic mechanism of the Wnt/β-catenin pathway is rather well understood, the significance of subcellular localization of pathway components for pathway regulation is largely unknown.

The centrosome is the major microtubule-organizing center of the cell and a template for formation of basal bodies, which support growth of cilia and flagella (reviewed in Ou and Rattner,2004). The centrosome, basal bodies, and cilia have also been recognized to function in cell polarity determination and signaling (Badano et al.,2005; Bettencourt-Dias and Glover,2007; Gerdes and Katsanis,2008; Schatten,2008). Recent reports implicate basal bodies and cilia in the regulation of both canonical and non-canonical (β-catenin-independent) Wnt pathways (Simons et al.,2005; Gerdes et al.,2007; Kishimoto et al.,2008; Corbit et al.,2008; reviewed in Bisgrove and Yost,2006). In agreement, a number of Wnt pathway components: Dsh, Dsh-binding protein Inversin, APC, GSK3β, Axin1, Axin2, and β-catenin, have been reported to localize to the centrosome and its derivatives, where some of them regulate centrosomal functions (Kaplan et al.,2004; Louie et al.,2004; Fumoto et al.,2006; Hadjihannas et al.,2006; Huang et al.,2007; Corbit et al.,2008; Bahmanyar et al.,2008; Park et al.,2008; Kim et al,2009; Fumoto et al.,2009). Despite this evidence, the role of the centrosomal localization of Wnt pathway components for Wnt signaling remains to be clarified.

Xenopus axin–related protein (XARP) was originally isolated in a yeast two-hybrid screen for proteins interacting with Dsh (Itoh et al.,2000). XARP is homologous to two other vertebrate Axins: Axin1 (Zeng et al.,1997; Luo and Lin,2004) and Axin2/Conductin (Behrens et al.,1998; Yamamoto et al.,1998; Chia and Costantini,2005). As with other Axins, XARP contains the N-terminal RGS domain, which is known to bind APC in Axin1, GSK3β-, and β-catenin-binding domains in the middle, and the Dsh-interacting DIX domain at the C-terminus. XARP is present in X. laevis and X. tropicalis genomes, while other analyzed vertebrate genomes (human, mouse, chicken, and zebrafish) only contain Axin1 and 2 homologs (data not shown). Functionally, XARP is similar to Axin1 and 2 in its ability to inhibit Wnt/β-catenin pathway, while the deletion of its RGS domain results in a dominant-negative effect and pathway activation, similarly to Axin1ΔRGS (Zeng et al.,1997; Itoh et al.,1998,2000).

We report that XARP is localized to the centrosome, and this localization requires the C-terminal region of XARP that includes the DIX domain. When the centrosomal localization of XARP is compromised by mutagenesis, XARP's ability to inhibit Wnt signaling is enhanced, suggesting that XARP centrosomal localization may be critical for its function as a Wnt pathway antagonist.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

XARP Centrosomal Localization Requires the DIX Domain and Adjacent Sequences

To study the subcellular localization of XARP, we made GFP- and myc-XARP fusion proteins and examined their distribution in Xenopus embryos (Fig. 1). We found that both were present as one or two small puncta per cell, suggesting a centrosomal localization. This was confirmed by co-staining with γ-tubulin, a centrosomal marker. To test whether XARP localization is important for its function in the β-catenin destruction complex, we set to identify and delete the centrosomal localization signal in XARP. First, we analyzed the subcellular distribution of XARP constructs with large deletions. We found that removal of the N-terminus, including the RGS domain, does not affect XARP localization (Fig. 1A,B). In contrast, removal of both N- and C-termini or the C-terminus alone causes a significant displacement of XARP to the cytoplasm (Fig. 1C–E), indicating that the C-terminus is important for XARP centrosomal localization. Furthermore, we created a series of smaller deletions in the XARP C-terminus, K1–K6 (Figs. 1F; see Supp. Fig. S1, which is available online) and found that deletions K3 and K6 disrupt XARP centrosomal localization, but also affect XARP protein stability (Figs. S1C,E–G). In contrast, the K4 deletion, which almost precisely corresponds to the DIX domain, causes mislocalization of XARP without affecting its stability (Figs. 1F, S1F,G). Interestingly, the C-terminal region of XARP, XARP-C, is localized to both the centrosome and cytoplasmic microtubules (Fig. 1G,H). We conclude that the XARP-C region is necessary and sufficient for XARP centrosomal localization.

thumbnail image

Figure 1. DIX domain is necessary and sufficient for XARP centrosomal localization. A–H: Xenopus embryos were injected with RNAs encoding indicated XARP constructs, cryosectioned at gastrula stage 10, and co-stained with rabbit GFP or myc antibody to detect XARP (green), mouse γ-tubulin antibody to detect centrosomes (red), and the chromatin marker DAPI (blue), except that in H, XARP Ab1 to detect XARP (green) and mouse α-tubulin antibody to detect microtubules (red) were used. I: Schematics of constructs used in A–H showing corresponding tags and amino acid numbers and the names of the constructs that were used in subsequent experiments.

Download figure to PowerPoint

Since the region adjacent to the DIX domain is essential for XARP stability, we decided to analyze its possible contribution for XARP centrosomal localization by making point mutations in this region. We mutated amino acids conserved between XARP, Axin1, and Axin2, since, similar to XARP, Axin1 and Axin2/Conductin have been reported to localize to the centrosome (Hadjihannas et al.,2006; Kim et al,2009; Fumoto et al.,2009), suggesting a conserved centrosomal localization signal. In m5–m8 constructs, conserved charged amino acids were substituted for alanines, and protein stability was confirmed by Western blot analysis (Figs. 2A, S2B). We found that m5 and m7 retain centrosomal localization, while m6 and m8 are significantly displaced to the cytoplasm (Figs. 2B, S2A). To quantify the degree of mislocalization of m6, m8, and K4, we estimated, in multiple embryo sections, the ratio of cells with remaining centrosomal staining (one or two dots) to all stained cells, from “none or a few” to “half or more” (Fig. S2C–G). We found that m8 and K4 were least localized to the centrosome, while m6 was somewhat intermediate between m8/K4 and wild-type XARP (Fig. 2C,D). We conclude that, besides the DIX domain, the adjacent charged conserved amino acids are also essential for the centrosomal localization of XARP.

thumbnail image

Figure 2. Point mutations upstream of the DIX domain compromise XARP centrosomal localization. A: Alignment of vertebrate Axin1, Axin2, and XARP proteins and location of the DIX domain, point mutations m5–m8, and some C-terminal deletions used in this study. X. laevis and X. tropicalis Axin2 and X. tropicalis XARP protein sequences were deduced from GenBank ESTs. B: Subcellular localization of wild-type, m6, and m8 myc-XARP. Staining was done as in Figure 1. C,D: In five independent blind experiments, 30–50 embryo sections expressing indicated XARP constructs were analyzed and classified according to the legend in D and Figure S2C–G. Three representative experiments (C) and averages ± S.D. from five experiments (D) are shown. *P with wild-type <0.1, **P with wild-type <0.05.

Download figure to PowerPoint

Increased Inhibitory Activity and Decreased Dsh Binding of Mislocalized XARP Mutants

We made use of identified XARP mutant versions with defective centrosomal localization to ask whether XARP localization plays a role in its function as a Wnt/β-catenin pathway antagonist. When K4, m6, and m8 were used in SuperTOPFlash luciferase reporter assays (Moon et al.,2002), we found that they suppressed Wnt3a-induced reporter activation significantly stronger than wild-type XARP (Fig. 3A,B). Constructs K4 and m8 that had the least centrosomal localization were particularly active. In contrast, m6 was more active than wild-type XARP in only half of experiments and was comparable to wild-type in another half (Fig. 3B). We also compared activity of wild-type XARP, m8, and K4 in two other assays, inhibition of Wnt3a-induced and primary dorsal body axis (Figs. 3C, S3). Consistent with our reporter assays, m8 and K4 showed enhanced activity comparing to wild-type XARP. As an important control, all three assays were accompanied by Western blot analysis to ensure similar expression levels of wild-type and mutant XARP constructs. We conclude that the centrosomal localization of XARP correlates with its reduced activity as a Wnt pathway antagonist.

thumbnail image

Figure 3. XARP constructs with disrupted centrosomal localization have an increased activity as Wnt pathway antagonists. A: SuperTOPFlash reporter assay using 0.2 ng and about 0.3 ng of RNAs encoding wild-type XARP and K4, respectively, which is proportional (5 times less) to the doses used for Western blot analysis. B: SuperTOPFlash reporter assay using myc-wild-type XARP, m6, m8, or K4. Two representative experiments are shown. **P with wild-type <0.01. C: Secondary axis induction assay. Xenopus embryos were injected with RNAs encoding Wnt3a alone or together with indicated myc-XARP constructs and, at stage 27, classified as either without secondary axis (2°) or with incomplete or complete secondary axis, based on the absence or presence of the cement gland, respectively. Embryos representing each phenotype, along with injected constructs in parentheses, are shown on the right. Numbers above graph bars represent the number of analyzed embryos.

Download figure to PowerPoint

The mutants of XARP that have decreased centrosomal localization (m6, m8, and K4) all have alterations in or near the DIX domain (Fig. 2A). Since the DIX domain of Axin1 is known to be important for Dsh binding and Wnt pathway activation (Kishida et al.,1999; Smalley et al.,1999; Itoh et al.,2000), we tested whether XARP mutants m6, m8, and K4 have normal Dsh binding. We performed co-immunoprecipitation assays using HA-Dsh and wild-type or mutant myc-XARP proteins overexpressed in Xenopus embryos, as previously described (Itoh et al.,2000). We found that wild-type XARP, m5, and m7, all of which are localized to the centrosome, strongly bind Dsh. In contrast, m6, m8, and K4 have about 50% decreased Dsh binding (Fig. 4). This decrease is comparable to the negative control XARPΔC, a construct similar to previously described XARPΔDIX that does not bind Dsh (Itoh et al.,2000). We conclude that XARP mutants with decreased centrosomal localization also have decreased Dsh binding.

thumbnail image

Figure 4. XARP constructs with disrupted centrosomal localization have decreased Dsh binding. A: A representative co-immunoprecipitation experiment showing total (“Lysate”) and co-precipitated (“IP”) Dsh and XARP proteins. Intensity of XARP and Dsh bands in “IP” lanes was measured using ImageJ software. Ratios of XARP/Dsh band intensities are presented relative to wild-type XARP binding (100%). B: Average ± S.D. from five independent experiments. **P with wild-type <0.01. ΔC, XARPΔC construct.

Download figure to PowerPoint

Endogenous XARP Is Localized to the Centrosome and Basal Bodies

To examine whether endogenous XARP is a centrosomal protein, we generated two anti-XARP antibodies against two different regions of XARP, designated XARP Ab1 and Ab2. By Western blot analysis, we found that both antibodies recognize overexpressed XARP constructs at predicted molecular sizes (data not shown), but do not detect endogenous protein in Xenopus embryo lysates, likely due to low expression levels. Importantly, we found that both antibodies strongly stain basal bodies of ciliated cells on cryosections of Xenopus tailbud stage embryos (Fig. 5A,B,D,F, asterisks). Additionally, XARP Ab2, which stains basal bodies more strongly, also detects centrosomal staining in the epidermis, retina, and the neural tube (Fig. 5A,E, arrowheads, and data not shown). Neither centrosomal nor basal body staining was detected when unrelated control rabbit antibodies were used (Fig. 5C,G). These data establish the centrosomal localization of endogenous XARP protein.

thumbnail image

Figure 5. Localization of endogenous XARP in the centrosome and basal bodies. A,D,F: XARP Ab2 detects basal bodies of ciliated cells and apically localized centrosomes in the epidermis. B: XARP Ab1 detects basal bodies in the epidermis. C: Control rabbit antibody (anti-myc) does not detect either basal bodies or centrosomes in the epidermis. E,G: XARP Ab2, but not a control rabbit antibody (anti-GST), detects centrosomes in the retina. Insets show magnified images of the boxed areas. All images are cross-sections of stage 26 Xenopus embryos (except F, a surface view of a ciliated cell) co-stained with rabbit polyclonal XARP Ab1 or Ab2 (green), mouse γ-tubulin antibody (red) or mouse acetylated tubulin antibody (cilia marker, red), and DAPI (chromatin marker, blue). A–C,E,G: Bottom panels are merged images of the top two panels. Arrowheads mark centrosomes, asterisks mark basal bodies.

Download figure to PowerPoint

The Effect of Wnt Pathway Components on XARP Localization

To test if XARP localization changes in response to Wnt signaling, we analyzed distribution of GFP-XARP and CFP-XARP in the presence of different upstream Wnt pathway components. Similarly to Axin1, XARP was localized to cytoplasmic puncta in the presence of Dsh (Fagotto et al.,1999; Smalley et al.,1999; Itoh et al.,2000), and these puncta did not overlap with the centrosome (Fig. 6A–C). Furthermore, when XARP was co-expressed with Dsh and Fz, it acquired a uniform plasma membrane localization (Figs. 6D, S4), which became non-uniform in the presence of Wnt3a (Fig. 6E).

thumbnail image

Figure 6. XARP localization in the presence of upstream components of the Wnt pathway. A–F: Xenopus embryos were injected with indicated combinations of RNAs and stained as in Figure 1, except that CFP-XARP and YFP-Dsh were detected by autofluorescence, and mouse anti-HA antibody was used to detect HA-Dsh. G: SuperTOPFlash reporter activation by Wnt3a in the absence or presence of HA-XARP-C.

Download figure to PowerPoint

Additionally, we found that XARP-C construct (Fig. 1G,H) can partially displace full-length XARP from the centrosome to cytoplasmic microtubules (Figs. 6F, S5). We previously reported that XARP-C inhibits Wnt signaling in secondary axis induction assay (Itoh et al.,2000). Consistently, XARP-C inhibits Wnt3a-activated SuperTOPFlash reporter (Fig. 6G), suggesting that Wnt pathway inhibition by XARP-C may be due to enhanced activity of endogenous wild-type XARP that is displaced from the centrosome to microtubules.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Decreased Centrosomal Localization of XARP Correlates With Its Enhanced Activity in the Wnt/β-Catenin Pathway

In this study, we report that XARP is a novel centrosomal protein. We observe that overexpressed XARP proteins with different tags are localized to the centrosome and basal bodies in Xenopus embryos (Figs. 1A, 2B, 6A, and data not shown). Moreover, two rabbit polyclonal XARP antibodies stain centrosomes and basal bodies at tailbud stages (Fig. 5), although no staining was observed at blastula-gastrula stage, likely due to low levels of both XARP protein and RNA (data not shown). XARP centrosomal localization was not entirely surprising, since other components of the β-catenin destruction complex have been found at the centrosome as well (Kaplan et al.,2004; Louie et al.,2004; Hadjihannas et al.,2006; Huang et al.,2007; Bahmanyar et al.,2008; Corbit et al.,2008; Kim et al,2009; Fumoto et al.,2009). Based on our finding and the literature, we originally hypothesized that the β-catenin destruction complex operates at the centrosome. To test this hypothesis, we mutated XARP to prevent its centrosomal localization and analyzed the activity of mislocalized XARP to inhibit Wnt/β-catenin signaling. If the β-catenin destruction complex is associated with the centrosome, mislocalized XARP should lose its inhibitory activity and Wnt signaling would be upregulated.

By deletion and mutagenesis analysis, we found that XARP constructs with mutated DIX domain and/or the adjacent region were significantly displaced from the centrosome to the cytoplasm. Contrary to our expectations, all three mutants of XARP with diminished centrosomal localization (m6, m8, and K4) revealed enhanced, rather than reduced, ability to antagonize the Wnt/β-catenin pathway. This suggests that XARP function is inhibited at the centrosome.

Contribution of Dsh Binding and Centrosomal Localization to XARP Activity

Since the centrosomal localization signal of XARP contains the DIX domain that has been shown to be critical for Dsh binding, one possible explanation for enhanced activity of XARP mutants is that they lose Dsh binding and become insensitive to negative regulation by upstream Wnt signaling (Kishida et al.,1999; Smalley et al.,1999; Itoh et al,2000). Indeed, co-immunoprecipitation experiments revealed that XARP mutants with poor centrosomal localization also have an approximately two-fold decrease in Dsh binding (Fig. 4B). Thus, m6, m8, and K4 may be more active than wild-type XARP, because (1) they are no longer inhibited by the centrosome; (2) they poorly bind Dsh, or both. We favor the last explanation, since all three mutants have a comparable, two-fold reduced Dsh binding (Fig. 4B), but in addition to that, m8 and K4 also have the most disrupted centrosomal localization (Fig. 2C,D), which correlates with their strongest activity (Fig. 3). This suggests that both Dsh binding and XARP subcellular localization probably contribute to XARP activity.

The Effect of Wnt Pathway Components on the Subcellular Localization of XARP

We find that localization of XARP changes in response to the status of Wnt signaling in the cells (Figs. 6, S4). For example, in the presence of Dsh, XARP is localized to cytoplasmic puncta that do not overlap with the centrosome. Furthermore, XARP is found at different plasma membrane locations in the presence of Fz and/or Wnt, which is similar to Axin1 (Axelrod et al.,1998; Cliffe et al.,2003; Schwarz-Romond et al.,2007; Zeng et al.,2008). At the plasma membrane, XARP likely participates in LRP5/6 phosphorylation and Wnt pathway activation, similarly to Axin1 and 2 (Zeng et al.,2005,2008; Bilic et al.,2007). Our study also suggests an intriguing possibility that by removing XARP from the centrosome, upstream Wnt signals relieve XARP inhibition and initiate a negative feedback loop to the Wnt pathway.

Regulation of Wnt Pathways by the Centrosome, Basal Bodies, and Cilia

While our data suggest that the centrosome positively regulates the Wnt/β-catenin pathway (by inhibiting XARP), several recent reports demonstrate that centrosome-derived structures, basal bodies, and cilia play a negative role in the Wnt/β-catenin pathway and/or a positive role in the non-canonical Wnt/planar cell polarity pathway (Simons et al.,2005; Gerdes et al.,2007; Kishimoto et al.,2008; Corbit et al.,2008; reviewed in Bisgrove and Yost,2006; Gerdes and Katsanis,2008). One possible explanation of these inconsistencies is that, despite being homologous structures, basal bodies/cilia and the centrosome may regulate Wnt pathways differently (Song et al.,2008). Besides that, we focused on a particular Wnt pathway component, XARP, while three other studies disrupted basal bodies and cilia as a whole (Gerdes et al.,2007; Kishimoto et al.,2008; Corbit et al.,2008), which might affect a number of Wnt pathway components present at the centrosome in different ways. Thus, the overall impact of the centrosome on the Wnt/β-catenin pathway may be either positive (which we propose here) or negative (consistent with previous reports), and may depend on the cell type.

In summary, this study establishes a negative correlation between XARP centrosomal localization and its function in the Wnt pathway. The question remains whether XARP also regulates the function of the centrosome as a microtubule-organizing center, similar to what has been proposed for Axin1 and other Wnt pathway components (Kaplan et al.,2004; Fumoto et al.,2006; Hadjihannas et al.,2006; Huang et al.,2007; Bahmanyar et al.,2008; Kim et al,2009; Fumoto et al.,2009). Future XARP loss-of-function studies that are focused on the organization and function of the centrosome, basal bodies, and cilia are needed to address this issue. It also remains to be established whether and how centrosomal localization of other Wnt pathway components is critical for Wnt/β-catenin pathway regulation.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

DNA Constructs, In Vitro Transcription

Myc-tagged Drosophila melanogaster Fz1 (Boutros et al.,2000), CFP-XARP, myc-XARP, myc-XARPΔRGS, HA-XARP-C, YFP-Dsh, and HA-Dsh (Itoh et al.,2000) were previously described. The other GFP- and myc-tagged XARP constructs and point mutations were made using convenient restriction enzyme sites and standard PCR-based mutagenesis techniques and verified by sequencing (details are available upon request). For amino acids left in XARP deletion mutants, see Figures 1I and S1. Myc-XARP point mutants were as follows: m5, K(613, 616, 617)A; m6, R643A/K645A; m7, E621A; m8, E636A. RNAs were in vitro synthesized from linearized plasmids using mMessage mMachine kits (Ambion).

Embryo Manipulations, Microinjections, Cryosectioning, Immunostaining

Xenopus laevis embryos were obtained by standard in vitro fertilization, de-jellied in 2% cysteine, pH 8.5, microinjected with 10 nl solution per cell, and incubated for several hours in 3% Ficoll+0.5×Marc's modified Ringers (MMR)+10 μg/ml gentamicin, followed by 0.1×MMR+10 μg/ml gentamicin. Immunofluorescence staining was done as previously described, using stage-9–11 embryos injected animally into both cells at the 2-cell stage and cryosectioned (Fagotto and Gumbiner,1994). For immunofluorescence experiments, 1 ng RNA for Wnt3a and XARP-C or 2–3 ng RNA for the other XARP constructs, myc-dFz1, YFP-Dsh, and HA-Dsh was used per embryo. Staining was viewed and photographed using a Zeiss Axiophot microscope. For secondary axis induction assay, 30 pg Wnt3a RNA alone or together with 20 pg RNA encoding myc-wild-type XARP, m8, or K4 was injected into one vegetal ventral cell at the 8-cell stage; 200 pg RNA from the same stocks was injected for accompanying Western blot analysis. All stages are according to Nieuwkoop and Faber (1967).

Luciferase Reporter Assays

For reporter assays, 20 pg SuperTOPFlash DNA with or without 30 pg Wnt3a RNA were co-injected with 50 pg RNA for myc-XARP constructs, unless indicated otherwise, or 100 pg RNA for HA-XARP-C, into the animal pole of both cells at the 2-cell stage. For each experimental group, the measurement of the luciferase activity was carried out by standard methods on stage-10–10.5 embryos divided into 4 or 5 replicates (Ossipova et al.,2005), using Turner BioSystem luminometer and Veritas software. For accompanying Western blot analysis, embryos injected with 200 pg RNA from the same stocks were used.

Immunoprecipitation, Statistical Analysis

For co-immunoprecipitation experiments, 30 embryos injected four times at the 4-cell stage with RNA encoding myc-tagged wild-type XARP, m5-8, K4, or XARPΔC with or without HA-Dsh RNA were used at stage 10–11, as previously described (Itoh et al.,2000). Four nanograms of each RNA was injected per embryo. The bands of interest were quantified using ImageJ software, and average ± S.D. ratios of XARP to Dsh in “IP” fractions was plotted. For statistical analysis, unpaired Student's t-test was performed to find P values (p).

Antibodies

Rabbit polyclonal XARP Ab1 and Ab2 were obtained by immunizing rabbits with GST-fused bacterially-produced peptides encoding XARP amino acids 495–706 and 341–444, respectively (Cocalico). The immune sera were affinity-purified against respective antigens and GST-specific antibodies were removed by incubation with membrane-bound GST. For immunostaining, the following antibodies were used at indicated dilutions: XARP Ab1 and Ab2 1:100, mouse γ-tubulin (GTU-88, Sigma) 1:300, mouse α-tubulin (B-512, Sigma, St. Louis, MO) 1:300, mouse acetylated tubulin (6-11B-1, Sigma) 1:300, rabbit anti-myc (Cell Signalling, Danvers, MA) 1:200, mouse anti-myc (hybridoma supernatant 9E10) 1:30, mouse anti-HA (hybridoma supernatant 12CA5) 1:200, rabbit GFP (Clontech, Palo Alto, CA) 1:300, followed by Cy3-anti-mouse IgG, Cy3-anti-rabbit IgG (Jackson ImmunoResearch, West Grove, PA), or Alexa Fluor 488-anti-rabbit IgG (Molecular Probes, Eugene, OR) at 1:200 dilution.

For Western blot analysis, the following antibodies were used at indicated dilutions: rabbit anti-myc (Cell Signalling) 1:1,500, mouse anti-myc (9E10) 1:300, mouse anti-HA (12CA5) 1:1,000, mouse β-tubulin (Biogenex) 1:500, followed by HRP-anti-mouse IgG or HRP-anti-rabbit IgG (Jackson ImmunoResearch) at 1:1,500 dilution.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

We are thankful to A. Antipova, K. Itoh, and V. Krupnik for help with XARP Ab1 production, K. Itoh and M. Ratcliffe for myc-XARPΔC, GFP-XARP and related constructs, M. Mlodzik for pCS2-myc-dFz1 plasmid, A. Sproul and K. Itoh for critical reading of the manuscript. This study was supported by NIH grants to S.S.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
DVDY_22125_sm_suppinfofigure1.tif2209KFig. S1. Characterization of small deletions in the XARP C-terminus. A–E: Unlike deletion K4 (Fig. 1F), small deletions K1, K2, and K5 in the context of full-length GFP-XARP do not abrogate its centrosomal localization. Deletions K3 and K6 just upstream of the DIX domain appear to localize to the cytoplasm. Staining was done as in Figure 1. F,G: By Western blot analysis, all small deletions, except for K3 and K6, are well expressed after injecting 2 ng mRNA. GFP-K3 and GFP-K6 cannot be detected at any dose (F), while myc-K3 and myc-K6 after injecting 15 ng mRNA are detected as a series of degradation products (G), which likely explains their abnormal localization (C,E). β-tubulin served as a loading control. The following amino acids were removed by deletions: K1, 540–559; K2, 562–573; K3, 631–648; K4, stop-codon after aa651; K5, 533–593; K6, 610–645.
DVDY_22125_sm_suppinfofigure2.tif4497KFig. S2. A: Subcellular localization of myc-wild-type XARP and m5-m8 detected by mouse anti-myc antibody, followed by goat Cy3-anti-mouse IgG antibody. B: Western blot analysis confirming stability of constructs m5-m8 (β-tubulin served as a loading control). C–G: Examples of sections that have indicated ratios of cells with centrosomal staining. In parentheses, injected constructs are indicated. Staining was done as in Figure 1; the presence of the centrosomes was confirmed by γ-tubulin co-staining (not shown).
DVDY_22125_sm_suppinfofigure3.tif6470KFig. S3. Embryo ventralization assay. Xenopus embryos were injected into the dorsal marginal zone at the 4-cell stage with 0.8 ng RNA encoding myc-wild-type XARP, m8, or K4, and, at stage 28, classified as either normal, with short axis, or without axis. Some embryos were used for accompanying Western blot analysis. Embryos representing each phenotype, along with injected constructs in parentheses, are shown on the right. Numbers above graph bars represent total number of analyzed embryos from two experiments, except for m8 (one experiment).
DVDY_22125_sm_suppinfofigure4.tif1232KFig. S4. XARP localization changes in response to Fz protein accumulation. A,B: Xenopus embryos were injected with GFP-XARP and Dsh RNAs and pCS2-myc-dFz1 DNA and stained at stages 9 and 11 with rabbit GFP antibody (green) and mouse anti-myc antibody (red). pCS2-myc-dFz1 plasmid encodes myc-dFz1 under the eukaryotic promoter CMV and is silent until midblastula transition (Heasman, 2006). Concomitantly with accumulation of Fz protein between stages 9 and 11, GFP-XARP appears at the plasma membrane. Since no detectable novel synthesis of GFP-XARP occurs between stages 9 and 11 (C, β-tubulin is the loading control), it is likely that previously made GFP-XARP protein dynamically changed its localization. Note that cells in B are smaller than in A because of several additional rounds of cell division.
DVDY_22125_sm_suppinfofigure5.tif2731KFig. S5. Subcellular localization of GFP-XARP in the absence or presence of HA-XARP-C (controls for Fig. 6F). After injection of indicated RNA combinations, embryos were cryosectioned at gastrula stage 10 and stained as described in Figure1, except that mouse-HA antibody was used in E. Bottom panels are merged images of the top two panels.
DVDY_22125_sm_suppinfofigure6.tif4396KSupporting Information Figure 6

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.