Attainment of cell type-specific cytotoxicity with minimal side effects is the ultimate goal of cancer therapy. By employing the prostate-specific antigen promoter (PSAP), we investigated (1) whether PSAP-driven antisense genetic constructs targeting DNA polymerase-alpha and topoisomerase II alpha (Top II alpha), designated PSAP-antipol and PSAP-antitop respectively, could induce death of prostate cancer cells, and (2) whether the cytotoxicity is restricted to cells of prostate origin. A PSAP-driven beta-galactosidase gene, PSAP-LacZ, was also used to estimate the expression of the PSAP-driven transcripts. Lipofection-mediated gene transfers were performed with these 3 constructs and a control plasmid, pCDNA3, in 3 human prostate cancer cell lines (LNCaP, DU-145, PC-3) and 5 other cell lines (Cos-1 [monkey kidney], HL-60 [human myeloid leukemia], Hep G2 [human hepatoma], NCI H460 [human lung cancer] and SW 480 [human colon cancer]). On transfection with PSAP-LacZ, LNCaP, DU-145, and PC-3 showed a 10.8, 1.8, and 1.6 fold increase in beta-galactosidase activity, respectively. The remaining 5 cell lines showed no changes after transfection. Corresponding to the levels of the induced beta-galactosidase activity, LNCaP showed the strongest growth inhibition by the antisense constructs: 36% by PSAP-antipol, 39% by PSAP-antitop and 80% by PSAP-antipol+PSAP-antitop. DU-145 and PC-3 had minimal growth inhibition with PSAP-antipol alone or PSAP-antitop alone. However, when cotransfected with PSAP-antipol and PSAP-antitop, DU-145 and PC-3 displayed 42% and 55% growth inhibition, respectively. In contrast, no cytotoxicity was observed in the remaining 5 cell lines when transfected with PSAP-antipol, PSAP-antitop or both. Therefore, PSAP-driven antisense gene therapy targeting DNA polymerase-alpha and Top II alpha inhibits the growth of human prostate cancer cells and the cytotoxic effect is restricted in cells of prostate origin.