Although the cytotoxic T-lymphocyte (CTL) response has long been recognized as the major limitation for successful adenovirus (Ad) gene therapy, it has been difficult to determine the major mechanisms of Ad-specific CTL generation and effector functions. To address this problem, we utilized an MHC I tetramer, Db-E1Bp, and an in vivo CTL assay to examine the role of CD28, perforin, FasL, and TNFa in generation and/or function of Ad-specific CTLs. Eight days after Ad infection, 1% of CD8+ cells in spleen were Db-E1Bp+ in CD28KO mice, compared to 9.3% in WT mice. The target cell killing efficiencies are also reduced by 77% and 80% in the spleen and liver of CD28KO mice, respectively, compared to those of WT mice. To distinguish between the defects in CTL generation and effector function in CD28KO mice, a plot of the number of Db-E1Bp+ CTLs and the E1Bp-specific target cell lysis was compared to that of WT mice. The results indicate that two out of the three CD28KO mice exhibited lower killing efficiency per cell compared with that of WT mice (Figure 3). The expression of CD28 on CD8+ T cells exhibited a direct effect on CTL activation since there was no defect in generation or function of Db-E1Bp+CD8+ T cells in CD4+ depleted mice. Perforin-deficient (Prf-/-)mice exhibited 70% target cell lysis compared to 98% in WT mice, whereas the number of Ad-specific CTLs remained the same as in WT mice. In contrast, the generation of Ad-specific CD8 T cells in FasL-deficient mice was significantly reduced by 62% and 50% in the spleen and liver, respectively, but the specific lysis were equivalent to that of WT mice. The killing efficiency in Prf2/2 mice were further reduced to 37% in killing of Fas-deficient cells but remained the same in killing of cells from TNFR12/2TNFR22/2 mice. These results indicate that CD28 is necessary for an optimal CTL response. In WT mice, the CTL activity is mediated mainly by perforin. In the absence of perforin, FasL, but not TNFa, produced by the CTLs appears to mediate the Ad-specific killing of target cells (Figure 4).