The SWI/SNF subunit Dpf2 licenses Nrf2-dependent gene expression to enforce multilineage control of inflammation [ChIP-seq]

Purpose: Multi-omics characterization of the consequences of deleting the SWI/SNF subunit Dpf2 in the transcriptome (RNAseq), accessibility of the genome (ATACseq) and binding of SWI/SNF subunits (CUT&RUN) in hematopoietic stem/progenitor cells, resting and polarized bone-marrow derived macrophages and resting and activated CD4+ T cells obtained from Dpf2f/f and Dpf2D/D mice. Methods: Lin- cKit+ cells were sorted from bone marrow of 28-days old Dpf2f/f and Dpf2D/D mice. Bone-marrow derived macrophages were obtained from 28-days old Dpf2f/f and Dpf2D/D mice and polarized with IFNg and LPS, or IL-4 for 24hours. CD4+ splenic T helper cells were obtained from 28-days old Dpf2f/f and Dpf2D/D mice and activated by treatment with PMA and Ionomycin. Methods: For ChIPseq in SKNO1 cell lines (AML), 10 million SKNO1 cells carrying shLuciferase or shDpf2#2487 were grown in the presence of doxycycline for 7 days to induce Dpf2 depletion. At day 7 of culture, cells were crosslinked with 1% formaldehyde for 15 minutes, quenched with 125mM Glycine for 5 minutes, washed in PBS and flash frozen at -80C. Crosslinked cell pellets were processed for ChIPseq following a previously described protocol (Boulay G et al., Cell 2017). Results: For ChIPseq, reads were trimmed using Cutadapt (v1.15--nextseq-trim=20 -m 18), aligned to the Human h38 genome using BWA (v0.7.13) with parameters aln -q 5 -l 32 -k 2, and deduplicated using Picard Mark Duplicates (v2.10.6). Peaks were called using narrow peaks by macs2 (v2.1.1.20160309) with IgG used as background. Conclusions: The absence of Dpf2 in LK cells results in downregulation of anti-oxidative and anti-inflammatory gene expression programs in bone marrow LK cells, bone-marrow derived macrophages and CD4+ T cells. Samples were obtained from primary bone marrow or spleen cells isolated or sorted from Dpf2f/f and Dpf2D/D mice of approximately 28 days of age. For RNAseq and ATACseq datasets, each file corresponds to data from 1 independent mouse/genotype. At least 3 mice (independent biological replicates) per condition and genotype were used. CUT&RUN samples were obtained from a pool of LK cells. For ChIP-seq, antibodies used are as follows: Rabbit anti-ARID1A (BAF250A) (D2A8U, Cell Signaling Technology, Cat#12354, lot#1; RRID:AB_2637010) and Rabbit IgG (Diagenode, C15410206, lot#RIG001).