EFFECT OF PEROXISOME PROLIFERATORS ON BLADDER CANCER: INDUCTION OF APOPTOSIS IN CELL LINES THAT EXPRESS PEROXISOME PROLIFERATOR-ACTIVATED RECEPTORS

Peroxisome proliferator-activated receptors PPARs are members of steroid/thyroid nuclear receptor superfamily and function as ligand-responsive transcription factors upon heterodimerization with retinoid X receptors. During last years, it became evident that PPARs are involved in many processes important for cell and tissue homeostasis and contribute to various pathologic processes; they are key regulators not only in atherosclerosis and diabetes, but also in inflammation and, more recently, in cancer. Three genes code for three PPARs isoforms: PPARa, PPARb/dand PPARg. Recently, it was been demonstrated by immunoblotting that PPARg mRNA was highly expressed in eleven human transitional cell cancers and in five transitional cell cancer cell lines. The function of PPARg on urothelium is unknown and the discovery that it is expressed in urothelium tumoral cells raised questions about the role of PPARg and its ligands in bladder cancer. PPARg can be activated by a number of ligands, either natural, such as 15-deoxy-delta12,14-prostaglandin J2 (15d-PGJ2), or synthetic, such as antidiabetic thiazolidinediones (TZDs). Here we describe the pro-apoptotic effect of two PPARg agonists,namely ciglitazone (CGT), a representative of TZDs and 15d-PGJ2, on four human bladder tumor cell lines having different levels of expression of PPARg and showing different degrees of malignancy in nude mice. The cell lines were: TCCSUP (non tumorigenic), 639V (papillary superficial cancer), T24 (papillary invasive cancer) and SG65 (solid invasive cancer), having respectively 2.19, 10.06, 7.99 and 0.73% of PPARg positive cells. They were treated in vitro with 25 microM CGT or 25 microM 15 d-PGJ2 for 4, 24 and 48h; then the percentage of apoptotic index of the evaluated under the U.V. light microscope using the annexin V-FITC test. The apoptotic index of the untraited cells cultured for the same times in medium with or without fetal bovine serum was also evaluated as control experiment. The results indicate that CGT had a late pro-apoptotic effect, increasing the apoptotic index of all cell lines only after 48h treatment. This effect seemed to be not mediated by PPARg, being evident both in cells with low PPARg expression (TCCSUP and SG65) and in cells with high PPARgexpression (639V and T24). It seemed to be inversely correlated to the malignant characteristics of the cells, being more evident in not tumorigenic cells (TCCSUP) and less evident in cells causing undifferentiated metastatic tumors (SG65). 15d-PGJ2, on the contrary, showed an early pro-apoptotic effect probably mediated by PPARg expression: the apoptotic index of 639V cells, in fact, increased significantly soon after 4h treatment, after 24h treatment all cell lines exhibited augmented levels of apoptosis, which resulted higher in high PPARg-expressing cells (639V and T24) than in low PPARg-expressing cells (TCCSUP and SG65); after 48h treatment all cells were necrotic. These data are to be considered preliminary to in vivo studies, which will be done by treating nude mice with CGT or 15d-PGJ2 injection of the four cell lines