Peroxisome-proliferator-activated receptors γ and β/δ mediate vascular endothelial growth factor production in colorectal tumor cells

Abstract

Background: Peroxisome-proliferator-activated receptors (PPARs) are nuclear receptors for fatty acids and their derivatives. PPAR subtypes PPARγ and PPARβ/δ are suspected to modulate cancer development in the colon, but their exact role is still discussed controversially.

Methods: The present study investigated the impact of PPARγ and PPARβ/δ on vascular endothelial growth factor (VEGF) and cyclooxygenase 2 (COX-2) expressions induced by synthetic and physiological agonists in the colorectal tumor cell lines SW480 and HT29 using reporter gene assays, qRT-PCR and ELISA.

Results: Activation of both PPARγ and PPARβ/δ induced expression of VEGF mRNA and protein in a PPAR-dependent way. The PPARγ agonists ciglitazone and PGJ(2) were the most effective inducers with up to ninefold and threefold increases in VEGF mRNA in SW480 and HT29 cultures, respectively. VEGF secretion was doubled in both cell lines. The PPARβ/δ agonists GW501516 and PGI(2) caused stimulations of only 1.5-fold in both cell lines. In addition, all PPAR agonists induced COX-2 mRNA and secretion of the COX-2 product PGE(2) in HT29 cells. However, this effect was not blocked by knock-down of PPAR expression nor was it essential for VEGF expression as shown by the lack of effect of the COX-2 inhibitor SC236.

Conclusion: In summary, our results identify both PPARγ and PPARβ/δ as an alternative COX-independent mechanism of VEGF induction in colorectal tumor cells.

Fig. 1

PPAR-dependent promoter activity in colorectal tumor cells. SW480 (a, b, e) and HT29 (c, d) cells were co-transfected with PPREγ (a, c, e) or PPREβ/δ (b, d, e) luciferase reporter constructs and a CMV control vector expressing renilla. Cultures were exposed to the indicated concentrations of the PPARγ agonists ciglitazone (a, c) and PGJ₂ (e) after 24 h. Alternatively, the PPARβ/δ agonists GW501516 (b, d) or PGI₂ (e) were used. Cell lysates were harvested 24 h after agonist addition, and PPAR-dependent promoter activity was determined by luciferase assay. The results shown represent the mean ± SD from at least three independent experiments. *, **, *** indicates an increase above control at P < 0.05, 0.01 and 0.001, respectively

Fig. 2

PPAR agonist induced VEGF expression. Ciglitazone (a–c) or GW501516 (d–f) was added to the culture medium of SW480 (a, c, d, f) and HT29 (b, c, e, f) cultures. RNA was isolated after 4, 6, and 8 h and VEGF expression determined by quantitative RT-PCR (a, b, d, e). Conditioned medium was harvested after 24 h for the quantification of secreted VEGF by ELISA (c, f). The results shown represent the mean ± SD from at least three independent experiments. * and ** indicates an increase above control at P < 0.05 and 0.01, respectively

Fig. 3

PGJ₂ and PGI₂ induced VEGF production. The prostaglandins were added to the medium of SW480 (open bars) and HT29 (closed bars) cultures. mRNA levels were determined after 8 h (a, b) and secreted VEGF after 24 h (c) as described in Fig. 2. The results shown represent the mean ± SD from at least three independent experiments. * and ** indicates an increase above control at P < 0.05 and 0.01, respectively

Fig. 4

PPAR dependency of VEGF induction. siRNA oligos were transfected into SW480 cells using siLentFect. PPREγ (a) and PPREβ/δ (e) reporter constructs were co-transfected, and 24 h later, cells were stimulated with 10 μM ciglitazone or 100nM GW501516 for determination of ligand-stimulated promoter activity. Parallel cultures were exposed to ciglitazone (b, c, d) or GW501516 (f, g, h) for isolation of RNA and the production of conditioned medium. The PPAR-target genes K20 (b) and ANGPTL4 (f) as well as VEGF mRNA (c, g) were analyzed by RT-PCR and secreted VEGF (d, h) were measured as described in Fig. 2. The results shown represent the mean ± SD from three independent experiments. * and ** indicate an increase above control at P < 0.05 and <0.01. § indicates a decrease when compared to the respective control knock-down group at P < 0.05

Fig. 5

PPAR-induced COX-2 expression. Ciglitazone (a, b) or GW501516 (d, e) were added to the culture medium. RNA was isolated after 1 and 2 h and COX-2 expression determined by quantitative RT-PCR (a, d). Conditioned medium was harvested after 4 and 8 h for the quantification of secreted PGE₂ by ELISA (b, e). In a parallel set of experiments, PGJ₂ (c) and PGI₂ (f) were used for induction. To determine the connection between COX-2 and VEGF expression, the COX-2 inhibitor SC236 (1 μM) was added simultaneously with the PPAR agonists (g, h) and VEGF expression and secretion measured as described. The results shown represent the mean ± SD from at least three independent experiments. * and ** indicates an increase above control at P < 0.05 and 0.01, respectively

Fig. 6

PPAR dependency of VEGF and COX-2 expression in HT29 cultures. siRNA oligos were transfected into HT29 cells using siLentFect. Cells were stimulated with 10 μM ciglitazone or 100nM GW501516 after 24 hours. RNA was isolated after 2 (PPARs, K20, ANGTPL4, and COX-2) and 8 (VEGF) hours, and the expression of PPARγ and β/δ (a), K20 (b), ANGPTL4 (c), VEGF (d), and COX-2 (e) was quantified by RT-PCR. The results shown represent the mean ± SD from three independent experiments. * indicates an increase above control and § a decrease when compared to the respective control knock-down group at P < 0.05