Critical role of histone demethylase RBP2 in human gastric cancer angiogenesis

Abstract

Background: The molecular mechanisms responsible for angiogenesis and abnormal expression of angiogenic factors in gastric cancer, including vascular endothelial growth factor (VEGF), remain unclear. The histone demethylase retinoblastoma binding protein 2 (RBP2) is involved in gastric tumorgenesis by inhibiting the expression of cyclin-dependent kinase inhibitors (CDKIs).

Methods: The expression of RBP2, VEGF, CD31, CD34 and Ki67 was assessed in 30 human gastric cancer samples and normal control samples. We used quantitative RT-PCR, western blot analysis, ELISA, tube-formation assay and colony-formation assay to characterize the change in VEGF expression and associated biological activities induced by RBP2 silencing or overexpression. Luciferase assay and ChIP were used to explore the direct regulation of RBP2 on the promoter activity of VEGF. Nude mice and RBP2-targeted mutant mice were used to detect the role of RBP2 in VEGF expression and angiogenesis in vivo.

Results: RBP2 and VEGF were both overexpressed in human gastric cancer tissue, with greater microvessel density (MVD) and cell proliferation as compared with normal tissue. In gastric epithelial cell lines, RBP2 overexpression significantly promoted the expression of VEGF and the growth and angiogenesis of the cells, while RBP2 knockdown had the reverse effect. RBP2 directly bound to the promoter of VEGF to regulate its expression by histone H3K4 demethylation. The subcutis of nude mice transfected with BGC-823 cells with RBP2 knockdown showed reduced VEGF expression and MVD, with reduced carcinogenesis and cell proliferation. In addition, the gastric epithelia of RBP2 mutant mice with increased H3K4 trimethylation showed reduced VEGF expression and MVD.

Conclusions: The promotion of gastric tumorigenesis by RBP2 was significantly associated with transactivation of VEGF expression and elevated angiogenesis. Overexpression of RBP2 and activation of VEGF might play important roles in human gastric cancer development and progression.

Figures

Figure 1

Association of RBP2 overexpression and increased VEGF expression in human gastric cancer. QRT-PCR analyses of (A) RBP2 mRNA and (B) VEGF mRNA in normal and cancerous human gastric tissues. *P < 0.05. (C) Correlation of RBP2 and VEGF levels in human gastric cancer tissues after standardization with matched normal tissues. **P < 0.01. (D) Immunohistochemical staining of expression of RBP2, VEGF, CD31, CD34 and Ki67 in human normal (left panel) and cancerous (right panel) gastric tissues. Representative images are shown. (E) Percentage positive cells by immunohistochemistry for RBP2, VEGF and Ki67 in human normal and cancerous gastric tissues. **P < 0.01. (F) Percentage positive area by immunohistochemistry for CD31 and CD34 in human normal and cancerous gastric tissues. **P < 0.01. Data are mean ± SEM of 3 independent experiments.

Figure 2

The direct regulation of VEGF by RBP2 in human gastric cancer cells. QRT-PCR analyses of (A) RBP2 and (B) VEGF mRNA level in control and RBP2 siRNA (10 nM)-transfected BGC-823, SGC-7901 and GES-1 cell lines after 72 h. **P < 0.01. Western blot analyses of (C, D) RBP2 and VEGF protein levels and (E, F) tri- and di- methylation of H3K4 in cancer cells treated with control and RBP2 siRNAs. C and R: Control and RBP2 siRNA, respectively. **P < 0.01. QRT-PCR analyses of (G) RBP2 and (H) VEGF mRNA levels in cancer cells transfected with control and RBP2 expression vector (4 μg) after 48 h. **P < 0.01. (I, J) Western blot analyses of RBP2 and VEGF protein levels and (K, L) tri- and di- methylation of H3K4 in cancer cells treated with control and RBP2 expression vectors. Vector C and R: Control pcDNA3.1 and pcDNA3.1-RBP2 expression plasmids, respectively. **P < 0.01. Data are mean ± SEM of 3 independent experiments.

Figure 3

VEGF expression and angiogenesis in RBP2-targeted mutant mice. (A) QRT-PCR analysis of VEGF mRNA in the wild type (the number is 6), heterozygote (4) and mutant mice (3). **P < 0.01 vs wild type group, ##P < 0.01 vs wild type group. (B, C) Trimethylation of H3K4 in wild-type, heterozygote and mutant mice. Two representatives of each group are shown. **P < 0.01 vs wild type group, ##P < 0.01 vs wild type group. (D) Hematoxylin and eosin (HE) staining and immunohistochemical staining of VEGF, CD31, CD34 and Ki67 expression in wild-type (left panel), heterozygote (middle panel) and mutant mice (right panel). Representative images are shown. Percentage positive cells for (E) VEGF and Ki67 and (F) percentage positive staining area for CD31 and CD34 in wild-type, heterozygote and mutant mice determined immunohistochemically. **P < 0.01 vs wild type group, ##P < 0.01 vs wild type group. Data are mean ± SEM of 3 independent experiments.

Figure 4

Transactivation of VEGF promoter induced by RBP2. (A) Schematic structure of pGL3-VEGF promoter and pGL3-VEGF promoter mutant. (B) After transfection of RBP2 expression vector for 48 h, luciferase activity of VEGF promoter reporters as compared with the control vector and RBP2 siRNA cotransfection for 72 h as compared with control siRNA. *P < 0.05 vs control vector. #P < 0.05 vs control siRNA. (C) The luciferase activity of VEGF promoter mutant reporters with RBP2 expression vector and RBP2 siRNA cotransfection. (D) RBP2 occupancy and trimethylation status of H3K4 at the promoter of the VEGF gene in BGC-823 cells. C and R: Control siRNA and RBP2 siRNA, respectively. Data are mean ± SEM of 3 independent experiments.

Figure 5

Promotion of tumorgenesis and angiogenesis by RBP2-induced VEGF expression in vitro. (A, B) Western blot analysis of RBP2 and VEGF protein levels in cancer cells treated with control and RBP2 siRNA or cotransfection of RBP2 siRNA and VEGF expression vector. **P < 0.01 vs control, ##P < 0.01 vs RBP2 siRNA + pEGEP. (C) Foci formation of human gastric cancer cells after treatment with control and RBP2 siRNA or cotransfection of RBP2 siRNA and VEGF expression vector. (D) The foci numbers of clones. **P < 0.01 vs control. #P < 0.05 vs RBP2 siRNA + pEGEP. (E) ELISA results for VEGF concentration in cell cultures treated with control vector and RBP2 expression vector. **P < 0.01. (F) Tube formation in HUVEC cells treated with the supernatants from the cells transfected with control vector and RBP2 expression vector. (G) Tube formation calculated as the percentage of cell surface area to the total surface area. The group incubated with the control cell cultures were set to 100%. **P < 0.01. Data are mean ± SEM of 3 independent experiments.

Figure 6

Inhibition of VEGF expression and angiogenesis by RBP2 knockdown in nude mice. (A) Tumorgenesis after injection of BGC-823 cells stably expressing RBP2 shRNA and control shRNA. (B) Tumor volume with RBP2 shRNA stable expression and controls. Data are median, with box edges as interquartile range and whiskers as the minimum and maximum values. **P < 0.01. (C) Growth curve with RBP2 shRNA stable expression and controls. **P < 0.01. (D) QRT-PCR analysis of RBP2 mRNA expression with RBP2 shRNA stable expression and controls. **P < 0.01 (E, F) Western blot analysis of RBP2 protein expression with RBP2 shRNA stable expression and the control. **P < 0.01. (G) Immunohistochemical staining for VEGF, CD31, CD34 and Ki67 in the control group (left panel) and RBP2 shRNA stable expression group (right panel). Representative images are shown. (H) Percentage positive area for CD31 and CD34 expression with the control and RBP2 shRNA stable expression determined immunohistochemically. **P < 0.01. (I) Percentage positive cells for VEGF and Ki67 with the control and RBP2 shRNA stable expression determined immunohistochemically. **P < 0.01.