Peroxisome proliferator-activated receptor gamma (PPARgamma) agonists and PPARgamma/alpha dual agonists have been or are being developed for clinical use in the treatment of type 2 diabetes mellitus and hyperlipidemias. A common tumor finding in rodent carcinogenicity studies for these agonists is hemangioma/hemangiosarcoma in mice but not in rats. We hypothesized that increased endothelial cell proliferation may be involved in the mechanism of PPAR agonist-induced vascular tumors in mice, and we investigated the effects on endothelial cells utilizing troglitazone, the first clinically used PPARgamma agonist, in vivo and in vitro. Troglitazone (400 and 800 mg/kg/day) induced hemangiosarcomas in mice in a 2-year bioassay. We showed that troglitazone increased endothelial cell proliferation in brown and white adipose tissue and liver in mice at sarcomagenic doses after 4 weeks of treatment. Troglitazone was cytotoxic both to human dermal microvascular endothelial cells (HMEC1) and mouse mammary fat pad microvascular endothelial cells (MFP MVEC) at high concentrations. However, MFP MVEC were more resistant to the cytotoxic effects of troglitazone based on the much lower LC(50) in HMEC1 (17.4 muM) compared to MFP MVEC (92.2 muM). Troglitazone increased the proliferation and survival of MFP MVEC but not HMEC1 in growth factor reduced conditions. Our data demonstrate that troglitazone may induce hemangiosarcomas in mice, at least in part, through enhancement of survival and proliferation of microvascular endothelial cells. Such an effect does not occur with human cells, suggesting that human may react differently to exposure to PPAR agonists compared with mice.