DUSP10 Is a Regulator of YAP1 Activity Promoting Cell Proliferation and Colorectal Cancer Progression

Abstract

Cell contact inhibition (CCI) is deregulated in cancer. Colorectal cancer (CRC) is the third most commonly diagnosed cancer worldwide. We found that dual-specificity phosphatase 10 (DUSP10) is involved in CRC. DUSP10 overexpression increased the growth of CRC cell lines and mouse xenografts, while the opposite phenotype was observed by DUSP10 silencing. High cell density (HD) induced DUSP10 expression in CRC cell lines, particularly within the nucleus. Yes-associated protein 1 (YAP1) is activated by dephosphorylation, controlling organ growth and CCI, both processes being deregulated in CRC. Expression levels and localization of DUSP10 matched with YAP1 levels in CRC cell lines. DUSP10 and YAP1 co-immunoprecipitated and their interaction was dependent on YAP1 Ser397. The existence of DUSP10 and YAP1 pathway in vivo was confirmed by using a transgenic Drosophila model. Finally, in CRC patients' samples, high levels of nuclear DUSP10 correlated with nuclear YAP1 in epithelial tumor tissue. Strong nuclear DUSP10 staining also correlated with high tumor stage and poor survival. Overall, these findings describe a DUSP10-YAP1 molecular link in CRC cell lines promoting cell growth in HD. We present evidence suggesting a pro-tumorigenic role of nuclear DUSP10 expression in CRC patients.

Keywords: cancer; cell contact inhibition; cell proliferation; prognosis; protein.

Figure 1

Dual-specificity phosphatase 10 (DUSP10) expression promotes higher colorectal cancer (CRC) cell proliferation and in vivo tumor growth. (a) Total cell number of HT29lucD6-DUSP10 was normalized to HT29lucD6-EV. Two-way ANOVA followed by Bonferroni’s post-test (mean ± standard error of mean (SEM); *** p < 0.001) and eight independent experiments were performed. (b) Growth curves of HT29lucD6-EV and HT29lucD6-DUSP10 for 42 h using real-time proliferation analysis by xCELLigence technology. Linear regression analysis was performed (*** p < 0.001). Representative graph of six independent experiments. (c) Total cell number of HT29lucD6-shDUSP10 cell lines was normalized to HT29lucD6-SCR. Two-way ANOVA followed by Bonferroni’s post-test (mean ± SEM; * p < 0.05, ** p < 0.01, *** p < 0.001) and seven independent experiments were performed. (d) Growth curves of HT29lucD6-shDUSP10 and HT29lucD6-SCR for 42 h using real-time proliferation analysis by xCELLigence technology. Linear regression analysis was performed (** p < 0.01, *** p < 0.001). Representative graph of three independent experiments. (e) Bioluminescence imaging (BLI) of mice xenoinjected with HT29lucD6-DUSP10 and HT29lucD6-EV. Data was normalized to first week post-inoculation for each cell line. Two-way ANOVA followed by Bonferroni’s multiple comparison and linear regression analysis were performed (mean ± SEM; p < 0.05; 7–8 mice per group). (f) Tumor volume of HT29lucD6-DUSP10 and HT29lucD6-EV xenografts was measured for seven weeks. Two-way ANOVA followed by Bonferroni’s multiple comparison tests were performed (mean ± SEM; p < 0.05; five mice per group). (g) BLI of mice xenoinjected with HT29lucD6-shDUSP10 and HT29lucD6-SCR. Two-way ANOVA with Bonferroni’s multiple comparison test and linear regression analysis were performed (mean ± SEM; *** p < 0.001; eight mice per group). (h) Tumor volume of HT29lucD6-shDUSP10 and HT29lucD6-SCR xenografts was measured for seven weeks. Two-way ANOVA and Bonferroni’s multiple comparison test were performed (mean ± SEM; *** p < 0.001; four mice per group).

Figure 2

Nuclear DUSP10 and Yes-associated protein 1 (YAP1) are increased in high-density conditions. (a) DUSP10 mRNA was quantified by HT29 in low density (LD) and high density (HD). Student’s t-test (mean ± SEM; *** p < 0.001) and four independent experiments were performed. (b) Expression of DUSP10, p-p38, and p38 of HT29 in LD and HD. (Left) A representative image of five independent experiments. (Right) Quantification of blots performed (mean ± SEM; Student’s t-test; ** p < 0.01, *** p < 0.001). (c) YAP1 mRNA was quantified from HT29 in LD and HD. Student’s t-test (mean ± SEM; * p < 0.05) and three independent experiments were performed. (d) Expression of YAP1 and p-YAP^Ser127 of HT29 in LD and HD. (Left) A representative image of five independent experiments. (Right) Quantification of blots performed (mean ± SEM; Student’s t-test; ** p < 0.01, *** p < 0.001). (e) Expression of DUSP10, YAP1, and p-YAP^Ser127 proteins from nuclear and cytoplasmic extracts from HT29lucD6-DUSP10 and HT29lucD6-EV in HD. LAMIN A/C and GAPDH (glyceraldehyde 3-phosphate dehydrogenase) are nuclear and cytoplasm control proteins, respectively. (Left) A representative image of three independent experiments. (Right) Quantification of all blots performed (mean ± SEM; Student’s t-test; p < 0.05). (f) Expression of DUSP10, YAP1, and p-YAP^Ser127 proteins from nuclear and cytoplasmic extracts of HT29lucD6-shDUSP10 and HT29lucD6-SCR in HD. LAMIN A/C and GAPDH/ACTIN are nuclear and cytoplasm control proteins, respectively. (Left) A representative image of three independent experiments. (Right) Quantification of all blots performed (mean ± SEM; Student’s t-test; * p < 0.05, ** p < 0.01, *** p < 0.001). Completed immunoblots of Figure 2b,d,e,f are in Figures S8 and S9, respectively.

Figure 3

A YAP inhibitor prevents DUSP10-enhanced proliferation. (a) Growth curves of HT29 proliferative response to SB239063 (SB 1 µM) and verteporfin (VP 1 µM) for 40 h after 2 h seeding (▼) using real-time proliferation analysis by xCELLigence technology. Linear regression analysis was performed (*** p < 0.001). Representative graph of four independent experiments. (b) Expression of DUSP10, YAP1, p-p38, and p38 protein levels in HT29 treated with SB (1 µM) and VP (1 µM) for 24 h (dotted line within Figure 3a). (Left) A representative image of three independent experiments. (Right) Quantification of all blots performed (mean ± SEM; Student’s t-test; * p < 0.05, ** p < 0.01). (c) Growth curves of HT29-EV and HT29-DUSP10 proliferative response to SB 1 µM and DMSO for 40 h after 2 h seeding (▼) using real-time proliferation analysis by xCELLigence technology. Linear regression analysis was performed (*** p < 0.001). Representative graph of three independent experiments. Completed immunoblots are in Figure S10. (d) Growth curves of HT29-EV and HT29-DUSP10 proliferative response treated with VP 1 µM and DMSO for 40 h after 2 h seeding (▼) using real-time proliferation analysis by xCELLigence technology. Linear regression analysis was performed (*** p < 0.001). Representative graph of three independent experiments.

Figure 4

DUSP10 interacts with YAP1 through Ser397 residue. (a) Expression of DUSP10, YAP1, p-YAP^Ser397, and p-YAP^Ser127 proteins in DUSP10-wild type (DUSP10-WT), phosphatase catalytic site mutant (DUSP10-C408S), and p38 binding site mutant (DUSP10-AA) HT29 cell line in LD and HD. TUBULIN is the control protein. (Top) A representative image of three independent experiments. (Bottom) Quantification of all blots performed (mean ± SEM; Student’s t-test; * p < 0.05, ** p < 0.01, *** p < 0.001). (b) Analysis of DUSP10, YAP1, and p-YAP^Ser127 proteins from nuclei and cytoplasm extracts of HT29 DUSP10 mutant cell lines in HD. LAMIN A/C and TUBULIN are used as nuclear and cytoplasm control proteins, respectively. (Left) A representative image of three independent experiments. (Right) Quantification of all blots performed (mean ± SEM; Student’s t-test; * p < 0.05, ** p < 0.01). (c) Immunoprecipitation of YAP-FLAG and detection of DUSP10 and p38 in HCT116. DUSP10-V5 and YAP-FLAG plasmids were co-transfected into the cell line and detected by anti-FLAG and anti-V5 antibodies, respectively. YAP1 wild type (YAP1-FLAG) and mutant (S381A-FLAG, S127A-FLAG) plasmids were immunoprecipitated with anti-FLAG. Representative images of three independent experiments. (d) Relative luciferase activity of the 8xGTII-luc (YAP/TEAD binding element reporter) was measured in HD, responding to DUSP10 overexpression and DUSP10 mutant constructs. HT29 (Top graph) and HCT116 (Bottom graph) were transiently transfected with the indicated plasmids and its control constructs. Student’s t-test (mean ± SEM; ** p < 0.01, *** p < 0.001) and three independent experiments were performed. Completed immunoblots of Figure 4a–c are in Figures S11 and S12, respectively.

Figure 5

Effect of DUSP10 expression on Hippo-Salvador-Warts pathway in Drosophila melanogaster. (a) Subcellular localization of DUSP10-Myc (red) in salivary gland cells in sal^EPv-Gal4 UAS-GFP/+; UAS-DUSP10-Myc/+ flies. ToPro3 (blue) was used as DNA marker and nuclei were detected by GFP (green) expression. Bars, 50 µm. (b) Higher magnification of a salivary gland cell from sal^EPv-Gal4 UAS-GFP/+;UAS-DUSP10-Myc/+ individuals. Transversal section was represented. Bars, 5 µm. (c) Quantification of wing surfaces from anterior margin to L5 vein in wild type sal^EPv-Gal4 UAS-GFP/+ (GFP) and sal^EPv-Gal4/+;UAS-DUSP10-Myc/+ (DUSP10-Myc) individuals. Representative images of wing surfaces of both genotypes. Some cell clusters were detected between the two epithelial wings’ surfaces (red arrow). (d) Quantification of wing surfaces from margin to L5 vein in Ste-20 protein kinase hippo overexpressing alone (sal^EPv-Gal4 UAS-GFP/UAS-hpo, hpo/GFP) and combined to human DUSP10-Myc (sal^EPv-Gal4/UAS-hpo;UAS-DUSP10-Myc/+, hpo/DUSP10-Myc) individuals. Representative images of wing surfaces for both genotypes. (e) Quantification of wing surfaces from margin to L5 vein in Merlin/Nf2 knockdown alone (sal^EPv-Gal4 UAS-GFP/+;UAS-Mer-RNAi, Mer-RNAi/GFP) and combined to human DUSP10-Myc (sal^EPv-Gal4/UAS-Mer-RNAi;UAS-DUSP10-Myc/+, Mer-RNAi;DUSP10-Myc) individuals. Representative images of wing surfaces for both genotypes. All graphs were represented by the media ± standard deviation (SD) (n = 10). Student’s t-test was performed to compare genotypes (** p < 0.01; *** p < 0.001).

Figure 6

DUSP10 is expressed in colon cancer patients and its nuclear expression is associated with a poor prognosis. (a) Representative images of DUSP10, p-p38, and YAP1 staining on a tissue microarray (TMA) in normal epithelial and tumor colon tissue from 73 human patients by immunohistochemistry (IHC). (Left) Image taken at 10× magnification (Bars, 100 μm) and (Right) magnified 400% by zoom. (b) Kaplan–Meier survival curve (p < 0.05) of 999 colon cancer patients with weak (blue line) and strong (green line) nuclear DUSP10 expression in tumor epithelial tissue for six years.