Fig. 3. The effect of BRG1 on the CIITA promoter is chromatin dependent. (A) STAT1, IRF-1 and USF-1 expression in SW13 cells. Cells were treated with IFN-γ for 0, 6, 12 or 24 h, and lysate analyzed by western blot using antibodies to STAT1, IRF-1 and USF-1. (B) IFN-γ activation of a transiently transfected CIITA reporter vector does not require BRG1. SW13 cells were transfected in duplicate with 0.8 µg of phCIITAPIV-LUC reporter plasmid (shown schematically) and increasing amounts (1, 3 and 5 µg) of a vector expressing BRG1 (pBJ5-BRG1). The empty expression vector pBJ5 was used to fill samples to ensure equimolar amounts of total plasmid. Samples were left untreated, or exposed to IFN-γ for 24 h. Baseline activity (relative activity = 1) is that obtained in untreated cells transfected with pBJ5. The results shown are the average and range of two experiments each performed in duplicate. (C) BRG1 has opposite effects on the endogenous versus transiently transfected CIITA promoter. SW13 cells were infected with AdtTA plus control AdGFP (lanes 1 and 2), AdBRG1 (lanes 3 and 4) or AdK798R (lanes 5 and 6) adenoviruses, transfected the next day with 1 µg of phCIITAPIV-LUC and then left untreated or exposed to IFN-γ for 24 h. Protein and RNA were prepared for luciferase (graph) and RT–PCR (lower panel) assays, respectively. The luciferase assays are the average and range of two experiments, each performed in duplicate. The RT–PCR is representative of two separate experiments (and reproduces the conclusions from Figure 2).