The extra domain A of fibronectin increases VEGF-C expression in colorectal carcinoma involving the PI3K/AKT signaling pathway

Abstract

The extra domain A (EDA)-containing fibronectin (EDA-FN), an alternatively spliced form of the extracellular matrix protein fibronectin, is predominantly expressed in various malignancies but not in normal tissues. In the present study, we investigated the potential pro-lymphangiogenesis effects of extra domain A (EDA)-mediated vascular endothelial growth factor-C (VEGF-C) secretion in colorectal carcinoma (CRC). We detected the expressions of EDA and VEGF-C in 52 human colorectal tumor tissues and their surrounding mucosae by immunohistochemical analysis, and further tested the correlation between the expressions of these two proteins in aforementioned CRC tissues. Both EDA and VEGF-C were abundantly expressed in the specimens of human CRC tissues. And VEGF-C was associated with increased expression of EDA in human CRC according to linear regression analysis. Besides, EDA expression was significantly correlated with lymph node metastasis, tumor differentiation and clinical stage by clinicopathological analysis of tissue microarrays containing tumor tissues of 115 CRC patients. Then, human CRC cell SW480 was transfected with lentivectors to elicit expression of shRNA against EDA (shRNA-EDA), and SW620 was transfected with a lentiviral vector to overexpress EDA (pGC-FU-EDA), respectively. We confirmed that VEGF-C was upregulated in EDA-overexpressed cells, and downregulated in shRNA-EDA cells. Moreover, a PI3K-dependent signaling pathway was found to be involved in EDA-mediated VEGF-C secretion. The in vivo result demonstrated that EDA could promote tumor growth and tumor-induced lymphangiogenesis in mouse xenograft models. Our findings provide evidence that EDA could play a role in tumor-induced lymphangiogenesis via upregulating autocrine secretion of VEGF-C in colorectal cancer, which is associated with the PI3K/Akt-dependent pathway.

Figure 1. Immunohistochemical staining for EDA and VEGF-C in human colorectal carcinoma.

Representative images of EDA expression in human colorectal carcinoma tissues (B) and in normal colorectal mucosae (A) are shown. Representative images of VEGF-C expression in colorectal carcinoma tissues (E) and in normal mucosae (D) are shown. Inset shows higher magnification of EDA(C) and VEGF-C (F) staining (brown), distinct from the colorectal benign mucosa. (G) Linear regression of EDA and VEGF-C of 52 human colorectal cancer samples was performed. (H) Representative images of EDA expression in tissue microarrays containing tumor samples from 115 CRC patients are shown. The bottom panel shows higher magnification of EDA staining. The slides were counterstained with hematoxylin. Scale bar = 50 µm.

Figure 2. Expression of cellular and secreted VEGF-C protein in transfected cells and control cells.

(A) shRNA-EDA SW480, mock SW480, pGC-FU-EDA SW620 and mock SW620 were confirmed by fluorescent microscopy. (B) Expression of EDA and VEGF-C protein was determined by western blotting. GAPDH was a loading control. VEGF-C protein concentration in transfected SW480 group(C) and transfected SW620 group (D) was detected by ELISA. This experiment was repeated in duplicate more than three times. The error bars are the means ± SEM. and ** p < 0.01 is considered as statistically significant. Scale bar  = 50 µm.

Figure 3. Activation of PI3K/Akt signaling pathway in transfected cells and control cells.

(A) p-Akt and Akt proteins of SW620 group and SW480 group were performed by western blotting analysis. (B) Quantitative analysis shows the ratio of p-Akt/GAPDH. The error bars are standard deviations of three independent experimental results. * p < 0.05 and ** p < 0.01 are considered statistically significant and highly significant, respectively. (C) VEGF-C, p-Akt and Akt proteins of EDA-overexpressed SW620 cells were performed by western blotting when the EDA-overexpressed cells were preincubated in medium containing 0–20 µM LY294002 for 24 h. Wild type SW620 was set up as the control group. The results shown are the average of three experiments.

Figure 4. BALB/c nude mice were subcutaneouly injected with transfected cells and negative control cells.

Xenografts were excised and sized 42 days later. (A) Effect of EDA on tumor proliferation. (B) Tumor volumes were measured when mice were sacrificed and the data are presented as mean determinants (±SEM). * p < 0.05. (C) Immunohistochemical staining of EDA and VEGF-C was performed in nude mouse xenografts. Scale bar  = 50 µm.

Figure 5. Orthotopic xenografts derived from transfected cells and control cells.

Seven to 8 wk old male BALB/c nude mice were orthotopically implanted with 6 groups of tumor xenografts. (A) Representative images of xenografts. Orthotopic images of xenograft tumors derived from 6 groups of transfected cells and control cells. Scale bar = 5 mm. (B) Expression of LYVE-1 was used to count LMVD in xenografts by immunohistochemical staining. The distribution of lymph vessels was stained brown under the light microscope. Scale bar  = 50 µm. (C) Quantification of lymph microvessel density (LMVD) is shown. The results shown are the average of three experiments. ** p < 0.01, as compared with control group. The values are displayed as the mean ± SEM.