SAMD9 Promotes Postoperative Recurrence of Esophageal Squamous Cell Carcinoma by Stimulating MYH9-Mediated GSK3β/β-Catenin Signaling

Abstract

Recurrence is a challenge to survival after the initial treatment of esophageal squamous cell carcinoma (ESCC). But, its mechanism remains elusive and there are currently no biomarkers to predict postoperative recurrence. Here, the possibility of sterile alpha motif domain-containing protein 9 (SAMD9) as a predictor of postoperative recurrence of ESCC is evaluated and the molecular mechanisms by which SAMD9 promotes ESCC recurrence are elucidated. The authors found that the high level of SAMD9 is correlated with postoperative recurrence and poor prognosis of ESCC. Overexpression of SAMD9 promotes tumor stemness, angiogenesis, and EMT, while downregulation of SAMD9 reduced these phenotypes. Mechanistically, it is found that SAMD9 stimulated ubiquitination-mediated glycogen synthase kinase-3 beta (GSK-3β) degradation by interaction with myosin-9 (MYH9) and TNF receptor-associated factor 6 (TRAF6), which in turn activated Wnt/β-catenin pathway. Further, the authors demonstrated that silencing SAMD9 inhibited lung metastasis and tumor formation in vivo. Finally, the authors found that silencing MYH9 or β-catenin, or overexpressing GSK-3β inhibited SAMD9-stimulated ESCC cell stemness, EMT, angiogenesis, metastasis, and tumorigenicity. Together, the findings indicate that the SAMD9/MYH9/GSK3β/β-catenin axis promotes ESCC postoperative recurrence and that SAMD9 is a crucial target for ESCC therapy. Additionally, SAMD9 has the potential as a predictor of postoperative recurrence in ESCC.

Keywords: esophageal squamous cell carcinoma recurrence; myosin-9; sterile alpha motif domain-containing protein 9; β-catenin signaling.

Figure 1

High level of SAMD9 is correlated with postoperative recurrence of ESCC. a) Heatmap showing differentially expressed genes between primary tumors of ESCC patients with postoperative metastatic recurrence and no recurrence. RNA sequencing was performed using primary tumors from ESCC patients with metastatic recurrence within 24 months after operation (n = 24) and no recurrence within 60 months after operation (n = 10). b) Genes that are increased in primary tumors from ESCC patients who had metastatic recurrence compared to those who had no recurrence. c) SAMD9 expression in primary tumors of ESCC patients was measured by immunohistochemistry (IHC) and the IHC staining was scored. Scores 0 and 1 were defined as low expression, and scores 2 and 3 were defined as high expression. d) The SAMD9 expression level is associated with postoperative recurrence of ESCC. Based on IHC scores, 123 ESCC cases were divided into high‐ and low‐ expression groups of SAMD9, and then, demonstrated the correlation between SAMD9 expression level and recurrence in ESCC patients. p‐value was calculated by the chi‐square test. e,f) Kaplan–Meier survival analysis shows the high level of SAMD9 was associated with lower recurrence‐free survival rate and a lower 5‐year overall survival rate of ESCC. p‐value was calculated by the log‐rank test. g) Multivariate survival analyses of clinicopathological characteristics indicate that SAMD9 is an independent poor prognostic factor of ESCC (n = 123). MTR, median time to recurrence; MST, median survival time. p‐values less than 0.05 were considered statistically significant.

Figure 2

SAMD9 promotes tumorigenicity and metastasis ability of ESCC cells. a) SAMD9 promotes soft agar colony formation, b) sphere formation, c) invasion and migration of ESCC cells. After 24 h of transfection with the indicated constructs, cells were subjected to analysis. Each bar represents the mean of three independent experiments. d) A subcutaneous xenograft model experiments showing that SAMD9 positively regulates ESCC cells tumorigenicity (n = 5/group) and e) tumor growth (n = 7/group). The indicated amounts of cells in 0.1 mL PBS were injected subcutaneously into the back of the mice. One month after cell injection, mice were sacrificed and tumors were collected. f) Western blot analysis of tumor tissues from subcutaneous xenograft models showing that SAMD9 positively regulates tumor stemness in vivo. g) A lung metastatic model showing that SAMD9 positively regulates ESCC cells lung metastasis (n = 5/group). 1.0 × 10⁶ cells in 0.1 mL of PBS were injected into mice via the tail vein and lungs were collected after one month of cell injection. h) Animal experiments show that SAMD9 promotes ESCC relapse. When the mean tumor volume reached ∼45 mm³, the nude mice were treated with 5 mg/kg CDDP every 2 days for 4 times. The tumor volume was measured on the indicated days (n = 5 per group). Significance between the control and treatment groups was determined using an unpaired two‐tailed Student t‐test, and a p‐value less than 0.05 was considered statistically significant. Error bar, SD. ^**, p < 0.01; ^***, p < 0.001.

Figure 3

SAMD9 stimulates angiogenesis, EMT, and the Wnt/β‐catenin pathway. a) Differentially expressed genes between SAMD9‐overexpressed KYSE270 cells (SAMD9‐OV) and their control cells (vector). KYSE270 cells were transfected with a SAMD9‐expressing construct or empty vector for 72 h, followed by mRNA sequencing. b) The gene set enrichment analysis (GSEA) shows that SAMD9 is positively associated with EMT, angiogenesis, and the Wnt/β‐catenin pathway in ESCC cells. The GSEA was performed using mRNA sequencing data from SAMD9‐OV and their vector control cells. c) Encyclopedia of Genes and Genomes analysis of primary tumors mRNA sequencing dataset showing that angiogenesis, EMT, and Wnt/β‐catenin pathways are involved in metastatic recurrence of ESCC. The mRNA sequencing data from the primary tumor samples of ESCC patients with or without postoperative metastatic recurrence (Figure 1a). d) Immunohistochemistry analysis shows that expression of CD31 and VEGF is positively regulated by SAMD9 in tumors of animal models. Tumors from subcutaneous xenograft models (Figure 2e) and lung metastatic models (Figure 2g). e) Visualized tumor vasculature images were obtained from xenograft tumors perfused with FITC‐lectin by confocal fluorescence microscopy (n = 5/group). f) Immunofluorescence analysis of tumor tissues from subcutaneous xenograft models (Figure 2e) shows that SAMD9 positively regulates Vimentin and β‐catenin expression, while negatively regulating E‐cadherin expression. Blue: DAPI; Green: indicated proteins. g) TOP/FOP luciferase analysis shows that SAMD9 positively regulates β‐catenin signaling in ESCC cells. Each bar represents the mean of three independent experiments. Significance between the control and treatment groups was determined using an unpaired two‐tailed Student t‐test, and a p‐value less than 0.05 was considered statistically significant. Error bar, SD.

Figure 4

Silencing β‐catenin inhibits SAMD9‐induced oncogenic effects in ESCC cells. a) Western blot analysis showing that silencing β‐catenin inhibits SAMD9‐stimulated cancer stemness, angiogenesis, EMT, and upregulation of MYH9 in KYSE270 cells. After 72 h of transfection with the indicated constructs, cells were harvested and subjected to western blot analysis. b) Silencing β‐catenin inhibited SAMD9‐stimulated sphere formation of ECA109 and KYSE270 cells. After 24 h of transfection with the indicated constructs, cells were subjected to sphere formation assay. c) Flow cytometry analysis shows that silencing of β‐catenin inhibited SAMD9‐upregulated side population (SP) in ESCC cells. After 72 h of transfection with the indicated constructs, cells were subjected to flow cytometry analysis. d) In vitro vasculogenic mimicry tube formation assay showed that overexpression of SAMD9 promotes angiogenesis, whereas silencing of β‐catenin inhibits the pro‐angiogenic effect of SAMD9 in ESCC cells. After 24 h of transfection with the indicated constructs, cells were subjected to a vasculogenic mimicry tube formation assay. e) Immunofluorescence (IF) analysis showing that silencing of β‐catenin inhibited SAMD9‐stimulated EMT in ESCC cells. After 72 h of transfection with the indicated constructs, cells were subjected to IF analysis. f) Silencing β‐catenin inhibited SAMD9‐stimulated soft agar colony formation, g) invasion and migration. Sphere formation, colony formation, and transwell assays were performed after 24 h of transfection. Each bar represents the mean of three independent experiments. Significance between the control and treatment groups was determined using an unpaired two‐tailed Student t‐test. Error bar, SD. ^#, compared to SAMD9‐overexpression group;*, compared to control group. ^#, p < 0.05; ^##, p < 0.01; ^###, p < 0.001; *, p < 0.05; ^**, p < 0.01; ^***, p < 0.001.

Figure 5

SAMD9 stimulates EMT, angiogenesis, cancer stemness, and the Wnt/β‐catenin pathway by binding to MYH9 in ESCC. a) Silver staining of co‐immunoprecipitation (Co‐IP) shows that SAMD9 interacts with several proteins in TE1 cells, including MYH9. Co‐IP was performed using antibodies against SAMD9. b) The physical interaction between SAMD9 and MYH9 was demonstrated by the GST pull‐down assay. c) The binding of SAMD9 and MYH9 was detected by Co‐IP in TE1 cells and Flag‐SAMD9 overexpressed ECA‐109 cells. d) Cytosolic colocalization of SAMD9 and MYH9 was detected in TE1 cells by immunofluorescence analysis. Red, SAMD9; green, MYH9, blue, DAPI. e) The positive regulation of MYH9 by SAMD9 in ESCC cells is demonstrated by WB. f) WB shows that inhibition of MYH9 blocks SAMD9‐stimulated EMT, angiogenesis, cancer stemness, and β‐catenin upregulation in ESCC cells. g) Silencing MYH9 inhibited tumor growth, angiogenesis, EMT, and β‐catenin expression in the subcutaneous xenograft model generated using SAMD9 overexpressing ECA109 cells (n = 5/group). h) Predicted interaction residues between SAMD9 and MYH9 using HDOCK server. i) The mutation of amino acid residues of SAMD9 (red). j) The Co‐IP analysis shows that SAMD9 mutation reduces the interaction between SAMD9 and MYH9 in ECA109 cells. k) WB analysis shows that mutation of SAMD9 did not affect EMT, angiogenesis, cancer stemness, and β‐catenin expression in SAMD9 overexpressing ESCC cells. After 72 h of transfection with the indicated constructs, cells were subjected to Co‐IP and WB analysis. l) TOP/FOP reporter analysis shows that mutation of SAMD9 did not activate the Wnt/β‐catenin pathway in ESCC cells. After 36 h of transfection with the indicated constructs, cells were subjected to luciferase assay. All in vitro data was from three independent experiments. Significance between the control and treatment groups was determined using an unpaired two‐tailed Student t‐test, and a p‐value less than 0.05 was considered statistically significant. Error bar, SD.

Figure 6

SAMD9 stimulates Wnt/β‐catenin pathway by stimulating MYH9‐mediated GSK3β degradation in ESCC cells. a) Co‐IP experiments show that SAMD9 form a complex with MYH9, GSK3β, and TRAF6 in ECA109 cells. b) Western blot analysis shows that SAMD9 negatively regulates GSK3β expression in ESCC cells. c) Co‐IP experiment show that SAMD9 increases the interaction of MYH9, TRAF6, and GSK3β in ESCC cells. d) Ubiquitination analysis shows that SAMD9 promotes GSK3β ubiquitination in ESCC cells. e) The Co‐IP experiment shows that silencing of SAMD9 reduced the interaction of MYH9, TRAF6, and GSK3β in ESCC cells. f) Ubiquitination analysis shows that silencing SAMD9 inhibits GSK3β ubiquitination in ESCC cells. g) Western blot analysis shows that overexpression of GSK3β inhibits SAMD9‐stimulated angiogenesis, EMT, and upregulation of c‐Jun, and MYH9 in ESCC cells. All in vitro experiments were performed after 72 h of cell transfection. h) GSK3β‐overexpression inhibits tumor growth, EMT, angiogenesis, and expression of MYH9 and β‐catenin in a subcutaneous xenograft model overexpressing SAMD9. 2 × 10⁶ cells in 0.1 mL PBS were injected subcutaneously into the back of mice (n = 7). One month after cell injection, mice were sacrificed and collected tumors. Significance was determined using an unpaired two‐tailed Student t‐test, and a p‐value less than 0.05 was considered statistically significant.

Figure 7

Expression correlation between SAMD9 and its target genes in ESCC clinical samples. A,b) Expression correlation between SAMD9 and its target genes was examined in 123 primary tumors of ESCC patients. Significance was determined using the chi‐square test. c) Expression correlation between MYH9 and VEGF, E‐cadherin, and β‐catenin was determined in 123 primary tumors of ESCC patients. Significance was determined using the chi‐square test. d) The correlation between the expression of SAMD9 and MYH9 in primary tumors and the recurrence‐free survival rate of ESCC patients was examined by Kaplan–Meier survival analysis. e,f) Expression of SAMD9 and its target genes were detected by IHC in primary and recurrent tumors of nine ESCC patients. Significance was determined using a paired two‐tailed Student t‐test. MVD, microvessel (CD31‐positive vessel) density.

Figure 8

The working model of SAMD9 in ESCC.