acid), on pancreatic cancer cell proliferation and survival. AGN193198 treatment reduces BxPC-3 cell proliferation more efficiently than high-affinity retinoid acid receptor (RAR)-or retinoid X receptor (RXR)-selective retinoids. Moreover, AGN193198 does not activate transcription from RAR or RXR response elements and its effects on cell survival are not reversed by treatment with RAR-or RXR receptorselective antagonists. These results suggest that the AGN193198-dependent inhibition of BxPC-3 cell function is not mediated via activation of the classical retinoid receptors. Cell cycle analysis of AGN193198-treated BxPC-3 cells indicates that AGN193198 causes accumulation of cells in G2/M. This change is associated with a marked reduction in regulators of S (cyclin A, cyclindependent kinase (cdk)2), G2/M (cyclin B1, cdk1, cdc25c) and G1 (cyclin D1, cyclin E, cdk2, cdk4) phase, and an increase in p21 and p27 level. Kinases assays reveal that cdk1, cdk2 and cdk4 activity are suppressed in AGN193198-treated cells. In addition, reduced cell proliferation is associated with enhanced procaspase (3, 8 and 9) and PARP cleavage. Z-VAD-FMK, a pancaspase inhibitor, inhibits AGN193198-dependent caspase activation and attenuates cell death. Z-VAD-FMK inhibits PARP cleavage, but does not alter the AGN193198-dependent reduction in cell cycle regulatory protein expression and activity, suggesting that caspase activation and suppression of cell cycle regulatory protein levels are independent processes. AGN193198 produces similar responses in other pancreatic cancer cell lines including AsPC-1 and MIA PaCa-2. These studies suggest that AGN193198 may be useful for the treatment of pancreatic cancer.