CCN family 2/connective tissue growth factor (CCN2/CTGF) regulates the expression of Vegf through Hif-1alpha expression in a chondrocytic cell line, HCS-2/8, under hypoxic condition

Abstract

Vascular endothelial growth factor (VEGF) is essential for establishing vascularization and regulating chondrocyte development and survival. We have demonstrated that VEGF regulates the expression of CCN2/connective tissue growth factor (CCN2/CTGF) an essential mediator of cartilage development and angiogenesis, suggesting that CCN2 functions in down-stream of VEGF, and that VEGF function is mediated in part by CCN2. On the other hand, the phenotype of Ccn2 mutant growth plates, which exhibit decreased expression of VEGF in the hypertrophic zone, indicates that Vegf expression is dependent on Ccn2 expression as well. Therefore, we investigated the molecular mechanisms underlying the induction of VEGF by CCN2 using a human chondrocytic cell line, HCS-2/8. Hypoxic stimulation (5% O(2)) of HCS-2/8 cells increased VEGF mRNA levels by approximately 8 fold within 6 h as compared with the cells cultured under normoxia. In addition, VEGF expression was further up-regulated under hypoxia in HCS-2/8 cells transfected with a Ccn2 expression plasmid. Hypoxia-inducible factor (HIF)-1alpha mRNA and protein levels were increased by stimulation with recombinant CCN2 (rCCN2). Furthermore, the activity of a VEGF promoter that contained a HIF-1 binding site was increased in HCS-2/8, when the cells were stimulated by rCCN2. These results suggest that CCN2 regulates the expression of VEGF at a transcriptional level by promoting HIF-1alpha activity. In fact, HIF-1alpha was detected in the nuclei of proliferative and pre-hypertrophic chondrocytes of wild-type mice, whereas it was not detected in Ccn2 mutant chondrocytes in vivo. This activation cascade from CCN2 to VEGF may therefore play a critical role in chondrocyte development and survival.

Fig. 1

Hypoxia-induced VEGF₁₆₅ expression in HCS-2/8 cells. (A) HCS-2/8 cells were exposed to normoxic (N) or hypoxic (H) conditions for indicated time periods. Total RNA was extracted and analyzed by Northern blotting for Vegf mRNA expression. (Upper panel) Hybridization signals for Vegf in the autoradiogram and signals for ribosomal RNA in the methylene blue-stained membrane are displayed. (Lower panel) Quantification of the fold-induction of Vegf mRNA signal at the indicated time periods. Relative fold-induction by hypoxic versus normoxic conditions is shown. Mean values of the results of two experiments are displayed with standard deviations. (B) Hypoxia-increased secretion of VEGF by HCS-2/8 cells. After the cells had reached subconfluence, the medium was replaced with fresh medium, and the cells were placed under the normoxic (open column) or hypoxic (closed column) conditions for indicated time periods. Then, the cell culture supernatant and cell layer fraction were harvested. Quantification of VEGF₁₆₅ was performed using an ELISA system. Asterisks indicate significant differences from normoxia (*p<0.05, **p<0.01). Results are presented as the mean±standard deviation of duplicates. (C) Effect of HIF-1α on the Vegf promoter activity under hypoxic conditions. HCS-2/8 cells were co-transfected with 1 μg of the Vegf promoter (−1910/+379)-driven firefly luciferase reporter plasmid, 0.1 μg of TK promoter-driven Renilla luciferase reporter plasmid (pRL-TK, internal control), and 1 μg of Hif-1α expression plasmid. After 24 h, the cells were exposed to normoxic or hypoxic conditions for 24 h, and the cells were assayed for luciferase activities. Relative luciferase activities are presented as relative values of the measured luminescence of firefly luciferase versus Renilla luciferase. Mean values of the results of duplicates are presented with standard deviations. Asterisks indicate significant differences from the cells transfected with the empty vector under hypoxia (*p<0.05).

Fig. 2

Effect of CCN2 on the expression of VEGF in chondrocytes. (A: left panel) Expression of Vegf₁₆₄ gene in wild-type and Ccn2-deficient chondrocytes. Wild-type (open column) and Ccn2-deficient chondrocytes (closed column) were cultured until reaching confluent. Total RNA was collected and quantitative real-time RT-PCR analysis was performed using mouse homologue; Vegf₁₆₄ and Gapdh specific primers. Data are presented as mean and standard deviation of duplicate cultures. (A: right panel) Western blot analysis of VEGF protein in lysates of wild-type and Ccn2-deficient chondrocytes. Levels of VEGF are decreased in Ccn2-deficient chondrocytes. (B) HCS-2/8 cells were transfected with a Ccn2-expression plasmid (Ccn2), a Hif-1α-expression plasmid (Hif-1α), or empty vector (Cont). After 24 h, the cells were exposed to normoxic or hypoxic conditions for 24 h, and total RNA was isolated using ISOGEN reagent. Northern blotting for Vegf₁₆₅ mRNA expression and methylene blue staining for 28S ribosomal RNA were performed.

Fig. 3

Effect of CCN2 on the expression of Hif-1α in HCS-2/8 cells. (A) HCS-2/8 cells were cultured until reaching sub-confluent in DMEM containing 10% FBS. Then, rCCN2 was added to the experimental cultures at the concentration of 50 ng/ml. PBS of the same volume was added to the control cultures. After 24 h, these cells were exposed to normoxia (N) or hypoxia (H). Total RNA was isolated 24 h later, and Northern blot analysis was performed using specific probe for Hif-1α. Hybridization signals of Hif-1α in the autoradiogram and methylene blue-stained rRNA on the same membrane are displayed. The arrow indicates the signal for Hif-1α. (B) Western blot analysis of HIF-1α protein in cell lysate from HCS-2/8 cells stimulated by rCCN2. HCS-2/8 cells pre-treated with 50 or 100 ng/ml rCCN2 were exposed to normoxic (N) or hypoxic (H) conditions for 24 h, and the cell lysate was corrected. Western blot analysis was performed as described in “Materials and methods”. The arrow indicates the signal for HIF-1α or β-actin. Levels of HIF-1α are increased in cells treated with rCCN2 both normoxic and hypoxic conditions. (C) Degradation of Hif-1α mRNA in HCS-2/8 cells in the presence or absence of 50 ng/ml rCCN2. HCS-2/8 cells were cultured until being sub-confluent in DMEM containing 10% FBS. Then, rCCN2 was added to the experimental cultures at the concentration of 50 ng/ml. After 24 h, the cells were treated with 10 μg/ml actinomycin D (ActD) for 6 h. The relative amount of Hif-1α mRNA remaining after the treatment with ActD was quantified by real-time RT-PCR analysis and shown as a percentage versus the control sample at time 0. Data presented are mean and standard deviation relative to 18S ribosomal RNA level of two separate reactions using mRNA from different cultures. (D) Effect of rCCN2 on rate of Hif-1α transcription in HCS-2/8 cells. Nuclei from PBS- or rCCN2-treated cells were isolated and used for run-on transcription assays. Autoradiograms of labeled transcripts representing Hif-1α and β-actin as an internal control are shown.

Fig. 4

Effect of CCN2 on the activation of HRE at −1006/−954 of the human VEGF promoter. (A) HCS-2/8 cells pre-treated with 100 ng/ml rCCN2 (closed column) or the same volume of PBS (open column) were transiently transfected either with the firefly luciferase reporter construct, 1HRE/WT-LUC containing the HRE at −1006/−954 of VEGF promoter, or its mutant, 1HRE/HM-LUC and thymidine kinase promoter-Renilla luciferase reporter plasmid (pRL-TK, internal control). After 24 h, cells were exposed to hypoxia (5% oxygen), and the cells were then assayed for luciferase activities. The ordinate shows the mean and standard error of relative value standardized against control culture of the measured luminescence of firefly versus Renilla luciferase of duplicate cultures. Relative luciferase activity of control culture is represented as 1.0. (B) HCS-2/8 cells pre-treated with 50 or 100 ng/ml rCCN2 (closed column) or same volume of PBS (open column) were transiently transfected with the reporter construct, 2HRE-LUC. After 24 h, the cells were exposed to normoxia or hypoxia (5% oxygen) for 24 h, and the cells were then assayed for luciferase activities. The ordinate shows the mean and standard error of relative value of the measured luminescence of firefly versus Renilla luciferase of duplicate cultures.

Fig. 5

Immunolocalization of VEGF and HIF-1α in sections of the growth plate cartilage of wild-type and Ccn2 mutant mice. Sections of the growth plates of E18.5 femurs in wild-type (A–D) and Ccn2 mutant (E–H) littermates were stained with safranin-O (A, E), and immunostained with anti-VEGF antibody (B, C, F and G), and with anti-HIF-1α antibody (D, H). The primary antibodies were visualized by immunoperoxidase, and then the sections were counterstained with methyl green. Images in C and G are magnifications of the boxed regions in B and F respectively. In the wild type, VEGF was localized in proliferative and pre-hypertrophic zone of the growth plate (C; arrows). In the Ccn2 mutant, it was slightly detected in proliferative and pre-hypertrophic zones (G). On the other hand, HIF-1α was localized in the nuclei of epiphyseal cartilage cells in the wild type (D: arrows), but in Ccn2 mutant, low level of expression was seen in the nuclei of epiphyseal cartilage cells (H). Bars in A, B, E, F and C, D, G, H represent 100 μm, and 50 μm, respectively.