CUT&Tag Chromatin Profiling Protocol

CUT&Tag library generation was performed as described previously (Kaya-Okur et al., 2020 (link)) with an altered nuclear extraction step. For the nuclear extraction, OE19 cells were initially lysed in Nuclei EZ lysis buffer (Sigma-Aldrich, NUC-101) at 4°C for 10 min followed by centrifugation at 500 × g for 5 min. The subsequent clean-up was performed in a buffer composed of 10 mM Tris–HCl pH 8.0, 10 mM NaCl and 0.2% NP40 followed by centrifugation at 1300 × g for 5 min. Nuclei were then lightly cross-linked in 0.1% formaldehyde for 2 min followed by quenching with 75 mM glycine followed by centrifugation at 500 × g for 5 min. Cross-linked nuclei were resuspended in 20 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES) pH 7.5, 150 mM NaCl, and 0.5 M spermidine at a concentration of 4–8 × 10³ / μl (2–4 × 10⁴ total). Subsequent stages were as previously described (Kaya-Okur et al., 2020 (link)). For 2–4 × 10⁴ nuclei, 0.5 μg of primary and secondary antibodies were used with 1 μl of pA-Tn5 (Epicypher, 15-1017). Antibodies used: anti-BRD4 (abcam, ab128874), anti-CTCF (Merck-Millipore, 07-729), anti-H3K27ac (abcam, ab4729), anti-H3K27me3 (Merck-Millipore, 07-449), anti-H3K4me1 (abcam, ab8895), anti-H3K4me2 (Diagenode, pAb-035-010), anti-H3K4me3 (abcam, ab8580), anti-H3K36me3 (Diagenode, pAb-058-010), anti-H4K20me1 (Diagenode, mAb-147-010), anti-PolII (abcam, ab817), anti-PolII-S2 (abcam, ab5095), anti-PolII-S5 (abcam, ab5131), and anti-Med1 (AntibodyOnline, A98044/10 UG). CUT&Tag libraries were pooled and sequenced on an Illumina HiSeq 4000 System (University of Manchester Genomic Technologies Core Facility). CUT&Tag data processing was performed as for ChIP-seq but with the MACS2 v2.1.1 (Zhang et al., 2008 (link)) but the --broad peak calling option was used for the H4K20me1, H3K27me3 and H3K36me3 marks. Fraction reads in peak (FRiP) scores for each mark were calculated using featureCounts and a stringent threshold of ≥2% was set to ensure quality of data for downstream analyses (Landt et al., 2012 (link); FRiP scores are listed in Supplementary file 12).
ChromHMM (Ernst and Kellis, 2012 (link)) was used to train an eight-state HMM using the CUT&Tag data for all marks assayed. The number of states was determined by running the model with increasing numbers of states until state separation was observed. Emission states were annotated in accordance with Roadmap Epigenomics Consortium Data (Kundaje et al., 2015 (link)).