Cells were harvested and processed for microarray and chromatin immunoprecipitation (ChIP) analyses as described below. MEFs were fixed ∼1% formaldehyde in cell growth media for 10 min followed by glycine fixation. Cells were then sonicated in ChIP–lysis buffer containing protease inhibitor cocktail after incubation on ice for 30 min. Upon sonication, the samples were spun at maximum speed for 12 min in a cold microcentrifuge. The supernatants were then addressed to as Input samples. A part of the input sample was precleared using 50% slurry of Protein A/G beads prepared in immunoprecipitation (IP) buffer for 2 h in cold, following which the beads were centrifuged out and supernatant was used for IP. Then the precleared samples were incubated with 2 µg of either of two distinct anti-Nrf2 antibodies that recognize different portions of Nrf2 protein. The rabbit-anti-Nrf2 monoclonal antibody from Epitomics Inc., CA, recognizes the C-terminal domain in all three cell lines, while the rabbit-anti-Nrf2 polyclonal antibody from Santa Cruz Biotech., CA, recognizes domains spanning the N-terminus in all three cell lines as well. For control, we used anti-rabbit IgG antibody for 1 h in cold. These antibodies have been used extensively in our laboratory for western and EMSA analysis [Figure 1A(i) and data not shown]. Protein A/G bead slurry was added to each reaction and was rotated overnight. Next day, beads were washed extensively, DNA was reverse crosslinked and eluted in elution buffer and stored at −80°C until used for sequencing or PCR validation assays. The eluted DNA was checked for 100- to 300-bp fragment enrichments using 8% polyacrylamide gel electrophoresis (data not shown). We generated two DNA libraries using two Nrf2-specific antibodies [one monoclonal, rabbit-anti-Nrf2 (Epitomics Inc., CA), and one polyclonal, rabbit-anti-Nrf2 (Santa Cruz Biotech., CA)] incubated with crosslinked and sonicated DNA from Keap1−/− MEFs. A third control library contains the input DNA from Keap1−/− MEFs.