Reactivation of PTEN tumor suppressor for cancer treatment through inhibition of a MYC-WWP1 inhibitory pathway

Abstract

Activation of tumor suppressors for the treatment of human cancer has been a long sought, yet elusive, strategy. PTEN is a critical tumor suppressive phosphatase that is active in its dimer configuration at the plasma membrane. Polyubiquitination by the ubiquitin E3 ligase WWP1 (WW domain-containing ubiquitin E3 ligase 1) suppressed the dimerization, membrane recruitment, and function of PTEN. Either genetic ablation or pharmacological inhibition of WWP1 triggered PTEN reactivation and unleashed tumor suppressive activity. WWP1 appears to be a direct MYC (MYC proto-oncogene) target gene and was critical for MYC-driven tumorigenesis. We identified indole-3-carbinol, a compound found in cruciferous vegetables, as a natural and potent WWP1 inhibitor. Thus, our findings unravel a potential therapeutic strategy for cancer prevention and treatment through PTEN reactivation.

Fig. 1.. Identification and characterization of the WWP1 E3 ligase for PTEN K27-linked polyubiquitination.

(A) Lysates from DU145 cells transfected with hemagglutinin (HA)–PTEN were immunoprecipitated (IP) with an anti-PTEN antibody followed by mass spectrometric peptide sequencing. WWP1 and PTEN were identified. IgG, immunoglobulin G. (B) WWP1 interacts with PTEN endogenously. DU145 cells were immunoprecipitated with either anti-WWP1 or anti-PTEN antibody and then analyzed by Western blot (WB). Input is 5% of the total lysates used in IP. (C) Analysis of endogenous PTEN ubiquitination (Ub) in Wwp1^−/− MEFs with stable reconstitution of either WWP1 WT or its catalytically inactive mutant (C890A, abbreviated CA). (D and E) Effects of the indicated ubiquitin KR (Lys to Arg) (D) or K-only (E) ubiquitin mutants on WWP1-mediated PTEN polyubiquitination. HEK293T cells were transfected with the indicated constructs, and PTEN ubiquitination was analyzed. The ubiquitinated proteins were pulled down under denaturing conditions by nickel–nitrilotriacetic acid (Ni-NTA) agarose and analyzed by Western blot. (F) Analysis of PTEN K27-linked polyubiquitination in PC3 cells expressing the indicated NEDD4 family ubiquitin ligases as in (D). EV, cells transfected with empty vector plasmids. (G) In vitro ubiquitination of PTEN by WWP1 E3 ligase. Flag-PTEN purified from HEK293 cells was subject to in vitro ubiquitination reaction in the presence of E1, E2, E3, and ubiquitin or various ubiquitin mutants and then examined by Western blot with anti-PTEN antibody. The input control of WWP1 was determined by Western blotting and is shown at the bottom. (H) Analysis of endogenous PTEN ubiquitination by small interfering RNA (siRNA)–mediated WWP1 knock down in U2OS ubiquitin replacement cells treated with 1 μg/ml doxycycline for 48 hours. shUb, shRNAs targeting ubiquitin. (I) Effects of the indicated ubiquitin KR mutants on WWP1-mediated PTEN polyubiquitination. PC3 cells were transfected with indicated constructs, and PTEN ubiquitination was analyzed as in (D). In (D), (F), (H), and (I), actin is used as a loading control.

Fig. 2.. K27-linked PTEN polyubiquitination suppresses PTEN dimerization, membrane recruitment, and function.

(A) Schematic description of the in vitro binding analysis of Flag-tagged unmodified or ubiquitinated PTEN with GST-PTEN from bacteria (left). In vitro pull-down assay with indicated Flag-tagged unmodified or ubiquitinated PTEN and GST-PTEN (right). Flagged untagged PTEN or ubiquitinated PTEN was purified from HEK293 cells transfected with Flag-PTEN or WWP1, along with individual His-ubiquitin variants, using M2 beads, whereas GST-PTEN was purified from bacteria. (B) Membrane and soluble fractions isolated from DU145 cells transfected with indicated constructs were analyzed by Western blot. Epidermal growth factor receptor (EGFR) serves as a membrane marker and actin as the internal control. (C) Analysis of AKT activation in DU145 cells. Total lysates were resolved by SDS–polyacrylamide gel electrophoresis (SDS-PAGE) and then probed with the indicated antibodies. (D) Evaluation of PTEN dimerization in Wwp1^−/− MEFs with stable reconstitution of either WWP1 WT or its catalytic dead mutant (C890A) by native gel electrophoresis. Total lysates from cells transduced with indicated constructs were immunoprecipitated with a rabbit anti-PTEN antibody, and then the immunocomplexes were natively eluted from the beads. The eluted samples were immediately run on the native gel. (E) Membrane and soluble fractions isolated from Wwp1^+/+ and Wwp1^−/− MEFs reconstituted with the indicated constructs were analyzed by Western blot. EGFR served as a membrane marker, and actin as the internal control. pS473-AKT and pT308-AKT are the markers of AKT activation. (F) Subcelluar localization of PTEN in Wwp1^+/+ or Wwp1^−/− MEFs. Confocal images of Wwp1^+/+ or Wwp1^−/− MEFs stained with 4′,6-diamidino-2-phenylindole (DAPI, blue) and PTEN (green) (top). The white arrows indicate PTEN plasma membrane localization. Scale bar, 20 μm. The percentage of cells displaying PTEN plasma membrane localization was quantified (bottom). Data are shown as mean ± SD for triplicate experiments, with 50 cells per group per experiment (***P < 0.0005, Student’s t test). (G) Evaluation of PTEN dimerization potential in DLP tissues derived from Wwp1 WT or Wwp1 knockout mice (n = 3). (H) Analysis of PI3K-AKT-mTOR signaling cascade in DLP tissues derived from Wwp1 WT or Wwp1 knockout mice (n = 3). Actin was used as a loading control. pS6 and S6 are markers to indicate the activation of the mTOR pathway. (I) Effects of the PTEN K342/K344R mutant on PTEN dimerization in PC3 cells. PC3 cells transfected with indicated constructs were serum starved for 6 hours and then treated with 100 ng/ml insulin for 10 min. Total lysates from cells were immunoprecipitated with a rabbit anti-Myc antibody, and then the immunocomplexes were natively eluted from the beads. The eluted samples were immediately run on a native gel, and Western blot using PTEN antibody showed the monomer and dimer of PTEN, as indicated by arrows. (J) Membrane localization of PTEN K342/K344R mutant in PC3 cells as in (I). Membrane and soluble fractions isolated from PC3 cells with indicated constructs were analyzed by Western blot. EGFR serves as the marker for the membrane fraction and actin as the internal control for the soluble fraction. Vector indicates cells transfected with empty vector plasmids. (K) Subcellular localization of the PTEN K342/K344R mutant in PC3 cells. Confocal images of PC3 cells stably expressing indicated PTEN WT or the K342/K344R mutant stained with DAPI and PTEN (green) (top). White arrows indicate PTEN plasma membrane localization. Scale bar, 20 μm. The percentage of cells displaying PTEN plasma membrane localization was quantified (bottom). Data are shown as mean ± SD for triplicate experiments, with 50 cells per group per experiment (***P < 0.0005, Student’s t test). (L) Effects of indicated PTEN KR mutants on PTEN lipid phosphatase activities in PC3 cells with indicated constructs. Data are shown as mean ± SD (***P < 0.0005, **P < 0.005, triplicate experiments, Student’s t test). OD620nm, absorbance at 620 nm. (M) Effects of indicated PTEN KR mutants on AKT activation in PC3 cells with indicated constructs. Total lysates were harvested and then probed with indicated antibodies. Actin was used as a loading control. (N) Effects of the PTEN K342/K344R mutant on tumor growth of PC3 cells as used in (K) in a xenograft mouse model. Error bars represent SEM (n = 5 mice per group).

Fig. 3.. MYC transactivates WWP1 gene expression toward PTEN suppression.

(A) Schematic description of the MYC responsive element on the WWP1 promoter (top). Chromatin level of MYC at the promoter of human WWP1 was measured in DU145 cells. Fold enrichment of MYC was determined by quantitative chromatin immunoprecipitation (qChIP) assays. JunB and RPL30 served as positive and negative controls, respectively. TSS, transcription start site. Data are shown as mean ± SD (***P < 0.0005, **P < 0.005, triplicate experiments, Student’s t test). (B) Reverse transcription–quantitative polymerase chain reaction (RT-qPCR) analysis of WWP1 in DU145 cells expressing the indicated constructs. The mRNA level of WWP1 was determined by RT-qPCR and is presented as a fold increase, as compared with the vector control (−). The HA-MYC levels are indicated by the triangle, from left to right. (C) Analysis of WWP1 and PTEN expression and AKT activation in DU145 cells expressing different amounts of HA-MYC. Total lysates were resolved by SDS-PAGE and then probed with indicated antibodies. The “−” indicates the vector control. (D) Tandem mass spectrum of a peptide derived from endogenous ubiquitinated PTEN in DU145 cells stably expressing the indicated constructs showed ubiquitin conjugation at the K27 residue of ubiquitin. (E) Ratio of indicated ubiquitin linkages detected by MS analysis of endogenous ubiquitinated PTEN purified from MYC overexpression or MYC overexpression–shWWP1 cells to that from control cells (without MYC overexpression). The abundance of each ubiquitin linkage was calculated as described in the materials and methods. (F) Analysis of PTEN K27-linked polyubiquitination in DU145 cells stably expressing MYC and/or WWP1 shRNAs. pCDH-puro-MYC, lentiviral expression of MYC. (G) Analysis of WWP1 and PTEN expression and AKT activation in DU145 cells expressing indicated siRNA SMARTpool. Total lysates were resolved by SDS-PAGE and then probed with indicated antibodies. (H) Analysis of WWP1 and PTEN expression and AKT activation in DU145 cells stably expressing MYC and/or WWP1 shRNAs. (I) Effects of WWP1 with or without PTEN on MYC-induced colony-forming activity in soft agar. The colony numbers are quantified and presented as mean ± SD (***P < 0.0005, **P < 0.005, *P < 0.05, triplicate experiments, Student’s t test). sgRNA, single-guide RNA. (J) Apoptosis assay of DU145 cells stably expressing MYC and/or WWP1 shRNAs. The percentage of apoptotic cells were stained with annexin V–PTEN and 7ADD and then quantified by fluorescence-activated cell sorting. Data are shown as mean ± SD (***P < 0.0005, **P < 0.005, triplicate experiments, Student’s t test). shCon, control. (K) Effects of WWP1 on Ras and MYC-induced colony-forming activity in soft agar or signaling pathway in cells. The colony numbers are quantified and presented as mean ± SD (***P < 0.0005, **P < 0.005, triplicate experiments, Student’s t test) (left). Total lysates were resolved by SDS-PAGE and then probed with indicated antibodies (right). ERK, extracellular signal–regulated kinase; pERK, phosphorylated ERK. (L) Effects of WWP1 on PTEN lipid phosphatase activities in MEFs with indicated constructs. Data are shown as mean ± SD (**P < 0.005, triplicate experiments, Student’s t test). In (C), (F), (H), and (K), actin was used as a loading control.

Fig. 4.. Characterization of WWP1 effects on Hi-Myc driven prostate tumors in mice.

(A) Gross anatomy of representative urogenital tracts from Hi-Myc; Wwp1^+/+ or Hi-Myc; Wwp1^−/− mice. The mice were analyzed at 5 months of age. (B) Analysis of the weight of the prostate lobes derived from Hi-Myc; Wwp1^+/+ or Hi-Myc; Wwp1^−/− mice. The mice were analyzed at 5 months of age (n = 5 mice per genotype). Data are shown as mean ± SD (***P < 0.0005, **P < 0.005, Student’s t test). KO, knockout. (C) Hematoxylin and eosin (H&E) and IHC staining of DLPs from Hi-Myc; Wwp1^+/+ or Hi-Myc; Wwp1^−/− mice with indicated antibodies. The mice were analyzed at 5 months of age. Scale bar, 50 μm. (D) Western blot analysis of DLP lysates from Hi-Myc; Wwp1^+/+ or Hi-Myc; Wwp1^−/− mice. The mice were analyzed at 5 months of age (n = 3 mice per genotype). Actin was used as a loading control. CC3, cleaved caspase-3. (E) Confocal analysis of DLPs from indicated mice stained with PTEN (green) and DAPI (blue). The white-outlined areas in the top row are enlarged 2.6-fold (bottom row) to show the subcellular localization of PTEN. Representative pictures are shown (n = 2 mice per genotype). Scale bars, 50 μm. (F) GSEA of RNA-seq data from the DLPs of Wwp1 knockout versus Wwp1 WT mice using the PI3K-Akt signaling pathway gene set annotated in the KEGG. Mice from all four groups contain the same Hi-Myc genetic background unless otherwise noted. NES, normalized enrichment score.

Fig. 5.. Therapeutic targeting of WWP1 E3 ligase in vitro and in vivo.

(A) In silico modeling of predicted interactions with the HECT domain of WWP1. (B) MST analysis determined the K_d of I3C toward His-WWP1 HECT WT or the His-WWP1 HECT F577/Y656A mutant. Concentration is reported in nanomolar. (C) Analysis of the effects of I3C on prostate organoid–forming ability from WT or Hi-Myc mice treated with or without 10 or 20 μM I3C for 3 days. Scale bar, 100 μm. DMSO, dimethyl sulfoxide. (D) Gross anatomy of representative urogenital tracts from Hi-Myc mice treated with vehicle or I3C. The mice, at 5 months of age, were treated intraperitoneally with I3C (20 mg/kg), three times a week for 1 month starting on day 0. Representative pictures are shown in the pane (n = 9 mice per group). (E) H&E and IHC staining of DLPs from Hi-Myc mice treated with vehicle or I3C with indicated antibodies. Scale bar, 50 μm. (F) In vivo ubiquitination analysis of endogenous PTEN ubiquitination levels of DLPs from Hi-Myc mice treated with vehicle or I3C. (G) Western blot analysis of AP, DLP, and VP lysates from Hi-Myc mice, as shown in (D), with the indicated antibodies. Actin was used as a loading control. (H) Confocal analysis of DLPs from indicated mice stained with PTEN (green) and DAPI (blue). The white-outlined areas in the top row are enlarged 2.6-fold (bottom row) to show the subcellular localization of PTEN. Representative pictures are shown (n = 2 mice per genotype). Scale bars, 50 μm. (I) GSEA of RNA-seq data from the DLPs of (i) I3C-treated mice versus vehicle (VEH)–treated mice and (ii) BKM120-treated mice versus vehicle-treated mice using the PI3K-Akt signaling pathway gene set annotated in the KEGG. Mice from all four groups contain the same Hi-Myc genetic background unless otherwise noted. (J) H&E staining of APs from Pten^+/− mice treated with vehicle or I3C. At 7.5 months of age, the mice were treated intraperitoneally with I3C (20 mg/kg), three times a week for 1 month starting on day 0. Representative pictures are shown (n = 3 mice per group). Scale bar, 100 μm. (K) Western blot analysis of AP lysates from Pten^+/− mice, as shown in (J), with the indicated antibodies. Actin was used as a loading control. (L) Model for WWP1-mediated PTEN K27-linked polyubiquitination in cell growth, tumor development, and progression. Deregulated MYC overexpression or MYC amplification promotes WWP1 expression and, in turn, triggers PTEN K27-linked polyubiquitination. Aberrant K27-linked polyubiquitination suppresses PTEN dimerization, plasma membrane recruitment, and tumor-suppressive function, leading to the tumor initiation and progression. PIP₂, phosphatidylinositol 4,5-bisphosphate; u, ubiquitin.