br To determine the in vitro significance of
To determine the in vitro significance of the above clinical data, we generated adenoviruses expressing Bach1 and infected the epithelial ovarian cancer cell line A2780. The wound healing and cell migration abilities were significantly greater in AdBach1 cells than in AdGFP cells (Fig. 2A–C) as well as in Con-siRNA cells vs Bach1-siRNA cells
(Fig. 2B–C). EMT is known to be involved in the progression and me-tastasis of various cancers. We then evaluated the expression of EMT-related proteins in A2780 cells. Higher levels of Bach1 expression were associated with significant increases in Vimentin, Slug, Snail, Twist, MMP1 and CXCR4 expression as well as with a significant reduction in E-cadherin expression. In contrast, silencing endogenous Bach1 in A2780 cells significantly reduced the expression of Vimentin, Slug, Snail, Twist, MMP1 and CXCR4 and increased E-cadherin expression
Fig. 3. Bach1 cooperates with HMGA2 in regulating EMT in human ovarian cancer cells. A-B. In vitro co-IP experiments. Endogenous Bach1 or HMGA2 was immunoprecipitated from lysates of A2780 cells (A) using a Bach1 or HMGA2-specific antibody, respectively, and immunoglobulin G (IgG) was used as the control. Co-IP proteins were analyzed by western blotting (WB) using the indicated antibodies. Input, 5% of material used for IP. Lysates from 293 cells (B) cotransfected with the Bach1-Flag and HMGA2-HA plasmids were immunoprecipitated with an anti-HA antibody. Total lysates and immunoprecipitates were then analyzed by WB. C-D. ChIP-qPCR was performed in A2780 cells with anti-Bach1 or anti-HMGA2 12141-67-2 (C) and in HMGA2 siRNA-transfected A2780 cells with an anti-Bach1 antibody
(D) to determine the enrichment of Vimentin, Twist, Snail and Slug promoter region sequences in the obtained ChIP DNA.##P < 0.01 vs IgG, **P < 0.01, *P < 0.05 vs Con-siRNA. t-test. E. Analysis of Slug promoter activity. A2780 cells were transfected with the Bach1-FLAG or control plasmid for 24 h; the cells were then transfected with a Slug promoter plasmid or a plasmid containing a mutated Slug promoter sequence for an additional 24 h. Cells were lysed, and luciferase activity was measured. **P < 0.01 vs control. t-test. NS, not significant. F. Assessments of EMT-associated protein expression by western blotting after HMGA2 knockdown in A2780 cells. The data are representative of three independent experiments. *P < 0.05, **P < 0.01 vs Ad-GFP; ##P < 0.01 vs Ad-Bach1. NS, not significant. ANOVA with post hoc test. G. Assessments of cell migration ability after HMGA2 knockdown in A2780 cells. HMGA2 downregulation reversed the eﬀect of Bach1 overexpression, inhibiting EMT in A2780 cells. Scale bars, 200 μm n = 3. *P < 0.05 vs Ad-GFP, ##P < 0.01 vs Ad-Bach1. ANOVA with post hoc test.
3.3. Bach1 promotes EMT and cell migration by recruiting HMGA2 in human ovarian cancer cells
HMGA2 was shown to directly bind to the Snail1 promoter, acting as a transcriptional regulator of Snail1 expression during the induction of EMT . Therefore, we determined whether HMGA2 is involved in the regulation of Bach1-induced EMT in ovarian cancer. We found that Bach1 coimmunoprecipitated with HMGA2 in A2780 cells (Fig. 3A). Bach1 and HMGA2 also coprecipitated from the lysate of HEK293T cells engineered to express Flag-tagged Bach1 and HA-tagged HMGA2 (Fig. 3B). ChIP-qPCR assays revealed that Bach1 and HMGA2 co-occupy the promoters of EMT-related genes (Fig. 3C). Furthermore, down-regulation of HMGA2 in A2780 cells impaired the binding of Bach1 to the promoter region of EMT-related genes (Fig. 3D). Bach1 over-expression significantly increased the luciferase activity of the Slug promoter but did not increase the luciferase activity of the Slug pro-moter when the Bach1 binding site was mutated (Fig. 3E). Importantly, the expression of these EMT-related genes and the cell migration in-duced by Bach1 overexpression were partially abolished by knockdown of HMGA2 (Fig. 3F–G), indicating that HMGA2 is required for the in-duction of EMT by Bach1. Interestingly, the enhancement of CXCR4 expression by Bach1 did not seem to be mediated by HMGA2 (Fig. 3F).
3.4. Bach1 promotes the metastasis of the epithelial ovarian cancer cell
We sought to determine whether Bach1 could promote the metas-tasis of EOC in an established mouse model of metastatic dissemination. We then generated Bach1 knockout A2780 cells (Bach1-KO A2780) using CRISPR–Cas9 genome editing (Fig. 4A–B) and generated a Bach1-overexpressing cell line by lentiviral infection (Fig. 4B). Accordingly, we injected A2780 cells into the peritoneal cavity of nude mice to mimic the route of dissemination of human ovarian cancer. We ob-served that the number and volume of metastatic liver tumors and the tumor volumes on the diaphragm were lower in the Bach1-KO group than in the WT group and higher in the LV-Bach1 group than in the LV-Con group (Fig. 4C–D); higher levels of Bach1 in metastatic tumors of the liver were associated with higher levels of Slug (Fig. 4E), indicating that Bach1 promotes the metastasis of EOC in vivo.