Ezrin ubiquitylation by the E3 ubiquitin ligase, WWP1, and consequent regulation of hepatocyte growth factor receptor activity

Abstract

The membrane cytoskeleton linker ezrin participates in several functions downstream of the receptor Met in response to Hepatocyte Growth Factor (HGF) stimulation. Here we report a novel interaction of ezrin with a HECT E3 ubiquitin ligase, WWP1/Aip5/Tiul1, a potential oncogene that undergoes genomic amplification and overexpression in human breast and prostate cancers. We show that ezrin binds to the WW domains of WWP1 via the consensus motif PPVY(477) present in ezrin's C-terminus. This association results in the ubiquitylation of ezrin, a process that requires an intact PPVY(477) motif. Interestingly ezrin ubiquitylation does not target the protein for degradation by the proteasome. We find that ezrin ubiquitylation by WWP1 in epithelial cells leads to the upregulation of Met level in absence of HGF stimulation and increases the response of Met to HGF stimulation as measured by the ability of the cells to heal a wound. Interestingly this effect requires ubiquitylated ezrin since it can be rescued, after depletion of endogenous ezrin, by wild type ezrin but not by a mutant of ezrin that cannot be ubiquitylated. Taken together our data reveal a new role for ezrin in Met receptor stability and activity through its association with the E3 ubiquitin ligase WWP1. Given the role of Met in cell proliferation and tumorigenesis, our results may provide a mechanistic basis for understanding the role of ezrin in tumor progression.

Figure 1. Ezrin interacts with the E3 ubiquitin ligase WWP1.

(A) Schematic representation of the structural organization of ezrin and WWP1. (B) Lysates from 293T cells transfected with Flag-WWP1 were incubated with immobilized GST or GST fused to full length, N-ter or C-ter ezrin and the blot was performed with the anti-Flag antibody. Lower panel: the blot was stained with ponceau S to reveal the GST-tagged proteins. (C) An extract of 293T cells transfected with Flag-WWP1 was incubated with immobilized GST or GST fused to ezrin wild type or carrying mutations in the PPVY motif. Lower panel: Ponceau S staining of the blot. (D) Extract of cells transfected with Flag-WWP1 was incubated with immobilized GST or GST fused to the C-ter domain of ezrin either wild type or carrying the mutation in the PPVY motif. Lower panel: Coomassie staining of GST tagged proteins.

Figure 2. Mapping the interaction of ezrin with individual WW domain of WWP1.

The lysates of 293T cells transfected with the indicated VSV-G tagged ezrin constructs were incubated with immobilized GST or GST fused to WW domains 1 to 4. Immunoblotting was performed with the anti-VSVG antibody. Lower panel: GST tagged proteins were revealed with ponceau S.

Figure 3. Subcellular localization of ezrin and WWP1.

(A) Immunoblot with the anti WWP1 antibody on extracts from 293T cells expressing WWP1, 293T and HeLa cells. (B) Blot was performed with the anti-WWP1 antibody on lysates of 293T cells transfected with scramble siRNA (si scr) or a pool of siRNA targeting WWP1 (si WWP1). Tubulin serves as loading control. (C) Membrane and cytosolic fractions of 293T cells were blotted with ezrin and WWP1 antibodies. The receptor Met was used as a control for cell fractionation. (D) Immunofluorescence was performed on LLC-PK1 cells stably expressing WWP1-GFP with an anti- E-cadherin (red) or anti ezrin (red) antibodies and visualized by wide-field 3D sectioning microscope with Z step every 0.2 µm. Two different focal planes are shown. Scale bar: 10 µm.

Figure 4. Ezrin ubiquitylation by WWP1.

(A) Lysates from 293T cells transfected with the indicated combinations of wild type flag-WWP1, flag-WWP1 C890A (C/A) and Myc-ubiquitin were immunoprecipitated with an anti-ezrin antibody followed by a blot with a Myc antibody. (B) 293T cells expressing Flag-WWP1, ezrin-VSVG and Myc-His-ubiquitin were lysed either in RIPA buffer or RIPA buffer containing 1% SDS and the ubiquitylated proteins were purified on TALON beads. The immunoblot was performed with an anti-VSVG antibody. (C) In vitro ubiquitylation was performed in presence of the indicated proteins. Left panel shows a Ponceau staining. Right panel: the blot was performed with an anti-Myc antibody to detect ubiquitylated protein. The brackets indicate the extent ubiquitylation of ezrin Δ29 and WWP1 (D) Ezrin is multi-ubiquitylated. 293T cells were transfected with the indicated plasmids. Upper panels: Immunoprecipitation of ezrin-Myc was followed by immunoblotswith the anti-VSVG (ubiquitin) and anti-ezrin antibodies. The asterisk indicates the position of non ubiquitylated ezrin. Lower panels: the immunoblots were performed on cell lysates with anti-Flag (WWP1) and anti-ezrin antibodies. (E) Ezrin ubiquitylation requires the PPXY motif. Lysates from 293T cells transfected with the indicated combinations of wild type flag-WWP1, flag-WWP1 C890A, wild type ezrin-VSVG, P474A, P475A, Y477F ezrin-VSVG and Myc-ubiquitin were immunoprecipitated with an anti-ezrin antibody followed by immunoblot with anti-Myc and -VSVG antibodies. (F) Ezrin steady state levels are not affected by WWP1 expression. 293T cells were transfected with the indicated combinations of plasmids and treated or not with the proteasomal inhibitor MG132. Immunoblots were performed with the indicated antibodies. Tubulin serves as a loading control. (G) Ezrin ubiquitylation with MG132 treatment. 293T cells were transfected with the indicated plasmids and treated or not with MG132. Immunoprecipitation with anti-VSVG antibody was followed by immunoblotting with anti-Myc and -VSVG antibodies. (H) Expression of WWP1 does not affect the half-life of ezrin. Stable Flp-In™ T-REx™ 293 cell lines expressing wild type PC- tagged WWP1 or vector alone were pulse-labeled for 15 min with ^35Smethionine/cysteine and chased for 0, 2, 24, 48, and 72 hours. Autoradiography of ezrin immunoprecipitates is shown.

Figure 5. Ezrin ubiquitylation by WWP1 is required for Met upregulation.

(A) Met levels in LLC-PK1 cells expressing GFP-WWP1 or GFP alone. Left panel: Blots were performed with anti-Met and anti-tubulin antibodies. Right panel: The quantification of Met level normalized to tubulin results from five independent experiments. Data are expressed as mean ± SEM, *P<0.05 (paired t-test). (B) Ezrin depletion prevents Met upregulation. Stable GFP-WWP1 cells were transfected with plasmids coding for shRNA targeting ezrin (shEz) or a scramble sequence (Scr). Left panel: Cell lysates were blotted with the indicated antibodies. Right panel: the quantification of Met level normalized to tubulin results from five independent experiments. Data are expressed as mean ± SEM, *P<0.05 (paired t-test). (C) Cell speed measured in a wound healing assay. In each graph, the bars represent the increase in the speed of cells treated with HGF as compared to the speed of untreated cells. Left panel: LLC-PK1 cells expressing either GFP or GFP-WWP1. Middle panel: LLC-PK1 cells expressing GFP-WWP1 and transfected with plasmids coding for scramble shRNA (Scr) or shRNA targeting ezrin. Right panel: LLC-PK1 cells expressing GFP-WWP1 and shRNA targeting ezrin were transfected with either an empty vector (vector) or plasmids coding for ezrin wild type or ezrin P475A. The data correspond to three independent experiments. Significance was tested using a two-way ANOVA model with interaction (SigmaStat). The pairwise comparisons were performed with the Student-Newman-Keuls method using an α risk of 0.05.