A major limitation of adenovirus type 5-mediated cancer gene therapy is the inefficient infection of many cancer cells. Previously, we showed that treatment with low doses of the histone deacetylase inhibitor FK228 (FR901228, depsipeptide) increased coxsackie adenovirus receptor (CAR) levels, histone H3 acetylation, and adenovirus infection efficiencies as measured by viral transgene expression in cancer cell lines but not in cultured normal cells. To evaluate FK228 in vivo, the effects of FK228 therapy in athymic mice bearing LOX IMVI or UACC-62 human melanoma xenografts were examined. Groups of mice were treated with FK228 using several dosing schedules and the differences between treated and control animals were determined. In mice with LOX IMVI xenografts (n = 6), maximum CAR induction was observed 24 h following a single FK228 dose of 3.6 mg/kg with a 13.6 +/- 4.3-fold (mean +/- SD) increase in human CAR mRNA as determined by semiquantitative reverse transcription-PCR analysis. By comparison, mouse CAR levels in liver, kidney, and lung from the same animals showed little to no change. Maximum CAR protein induction of 9.2 +/- 4.8-fold was achieved with these treatment conditions and was associated with increased histone H3 acetylation. Adenovirus carrying a green fluorescent protein (GFP) transgene (2 x 10(9) viral particles) was injected into the xenografts and GFP mRNA levels were determined. A 7.4 +/- 5.2-fold increase in GFP mRNA was found 24 h following adenovirus injection into optimally FK228-treated mice (n = 10). A 4-fold increase in GFP protein-positive cells was found following FK228 treatment. These studies suggest that FK228 treatment prior to adenovirus infection could increase the efficiency of adenovirus gene therapy in xenograft model systems.