(A) Three different experimental strategies were employed to determine whether Keap1 regulates the ability of Nrf1 to transactivate ARE-driven gene expression. First (A, left panel), COS-1 cells were transfected, using Lipofectin® reagent (5 μl) in 1 ml of Opti-MEM1, with 0.6 μg of an expression vector for either Keap1 or empty pcDNA3.1/V5His B, along with 0.6 μg of an expression construct for either Nrf1 or Nrf2 or an empty pcDNA3.1/V5His B vector, and 0.6 μg of PSV40nqo1-ARE-Luc and 0.2 μg of pRL-TK Renilla. Secondly (A, middle panel), keap1+/+ and keap1−/− MEFs were transfected, using 10 μl of Lipofectamine™ 2000 and 2.5 ml of Opti-MEM1, with 2.6 μg of an expression construct for either Nrf2, Nrf1, Nrf1Δ2–120 or the empty pcDNA3.1/V5His B vector, together with 1.2 μg of PTKnqo1-ARE-Luc and 0.2 μg of β-gal reporter. Thirdly (A, right panel), keap1−/− MEFs were transfected, using Lipofectamine™ 2000, with a total of 3.8 μg of DNA comprising 1.3 μg of either Keap1 or empty pcDNA3.1/V5His B, along with 1.3 μg of an expression construct for either Nrf2, Nrf1, Nrf1Δ2–120, or empty pcDNA3.1/V5His B, and 1.0 μg of pGL-6×ARE-Luc and 0.2 μg of β-gal reporter. Luciferase activity was measured and normalized to activity of Renilla (left panel) or β-gal (middle and right panels). The data are presented as a fold change (mean±S.D.) from two independent experiments each performed in triplicate. (B) The keap1+/+ and keap1−/− MEFs were transfected with 2 μg of an expression construct either for wild-type Nrf1, its mutants Nrf1ΔETGE or Nrf1ΔDIDLID/DLG, or the empty pcDNA3.1/V5His B plasmid. These V5-tagged Nrf1 proteins were visualized using immunocytochemistry with a FITC-coated secondary antibody. Subsequently, related images were analysed by confocal microscopy as described in the legend to . The horizontal bar represents 20 μm.