Nono deletion impairs RNAPII activity upon DNA damage

NONO stimulates the formation of transcriptionally active bimolecular condensates in vitro and colocalises with RNAPII in vivo (Lewis et al, 2023; Zhang et al, 2023). To approach the role of murine Nono in KP lung cancer cells, we used the CRISPR/Cas9 technology and created the monoclonal Nono knockout cell line KPN. The absence of Nono expression was confirmed by immunoblotting and confocal imaging (Fig S1A and B). We wished to test whether Nono deletion alters RNAPII activity in response to DNA damage. We incubated KP or KPN cells in the presence or absence of the topoisomerase II inhibitor etoposide, and assessed RNA synthesis globally by confocal imaging of nascent transcripts that were pulse-labelled with 5-ethynyl uridine (EU) (Fig 1A). We observed a partial reduction in EU signals upon etoposide treatment in KP cells, which was more distinct in KPN cells. Next, we investigated RNAPII promoter occupancy and performed CUT&RUN-seq with an antibody that selectively recognises serine-5 phosphorylated residues of the RNAPII carboxy-terminal domain (CTD S5P). Indeed, combining Nono deletion with etoposide treatment significantly reduced CTD S5P occupancy at highly expressed genes, but had no impact in untreated cells (Fig 1B). This phenotype was confirmed by visual inspection of browser tracks at a subset of highly expressed genes like Actb (Fig 1C). To validate CUT&RUN-seq data, we performed manual chromatin immunoprecipitation (ChIP) and confirmed a partial loss of CTD S5P occupancy at the Actb promoter region upon combining Nono deletion with etoposide treatment (Fig 1D). Thus, Nono stimulates RNAPII activity and CTD S5P promoter occupancy upon etoposide treatment.

Nono deletion triggers genome instability and impairs DNA repair

We could recently show that the DDR triggers NONO nucleolar relocalisation to detain aberrant, intron-containing pre-mRNA transcripts, thereby promoting genome stability in U2OS cells upon etoposide treatment (Trifault et al, 2022, 2024). To assess whether the etoposide-induced defects in RNAPII activity and promoter occupancy in Nono-deficient KP cells correlate with genome instability, we compared the amount of broken chromatin in KP and KPN cells upon treatment with etoposide and other cytotoxic drugs by neutral comet assays (Fig 2A). Although etoposide treatment or Nono deletion alone induced modest appearance of DNA tails, we found more prominent DNA damage in a subset of etoposide-treated KPN cells. Nono-deficient KP cells were also hypersensitive to treatment with the radiomimetic drug bleomycin, but not to treatment with the DNA interstrand crosslinker cisplatin. Next, we applied DNA damage in situ ligation followed by the proximity ligation assay (DI-PLA), a method that uses ligation of a biotinylated DNA linker to blunted DSBs to detect persistent DNA damage as PLA foci in single cells (Galbiati et al, 2017). To perform DI-PLAs, we combined a biotin antibody with an antibody that recognises the DNA damage marker ser-139 phosphorylated histone H2A.X variant (γH2A.X), and quantified the number of dots per nucleus by confocal imaging (Fig 2B). Stratification revealed that the treatment with etoposide significantly increased the number of cells that were positive for DI-PLA signals, but reduced the number of KP cells without DI-PLA dots. Strikingly, this phenotype was elevated in Nono-deficient KP cells. We conclude that the deletion of Nono hypersensitises KP cells to DSB-inducing drugs.

Next, we wished to test the impact of Nono deletion on the recognition and repair of DSBs. We used etoposide incubation kinetics and compared the level of γH2A.X signals in KP and KPN cells (Fig 3A). We found that Nono deletion increased the amount of etoposide-induced γH2A.X levels twofold to threefold and also delayed the clearance of γH2A.X upon chase. The delayed clearance of γH2A.X levels could also be confirmed upon chasing bleomycin treatment (Fig 3B). Importantly, γH2A.X levels could partially be rescued by complementation with ectopically expressed mCherry-tagged human NONO (Fig 3C and D). Next, we used confocal imaging of the DSB marker 53BP1 to compare the formation of etoposide-induced DSB foci in KP and KPN cells. We observed that the number of KPN cells that displayed persistent 53BP1-positive staining was significantly increased in KPN cells upon chasing etoposide treatment (Fig 3E). A similar phenotype was detected upon staining for γH2A.X-positive foci, which were elevated in KPN cells both after etoposide incubation and chase (Fig 3F). We conclude that Nono promotes efficient DSBR in KP cells.

Nono-dependent induction of Gadd45b promotes the DDR

Our data point towards defects in RNAPII transcription that are prevalent in DNA-damaged Nono-deficient KP cells and correlate with inefficient DSBR. To identify genes that may suppress such defects in Nono-proficient cells, we performed RNA-seq in KP and KPN cells in the absence or presence of etoposide. We first asked whether etoposide treatment induces gene sets in KP cells that are sensitive to Nono deletion. We identified gene sets involved in TNF-α signalling, the UV response, and the p53 pathway as highly significantly up-regulated upon etoposide treatment in both KP and KPN cells (Fig S2A and B). When scoring the relative induction of these three gene sets in etoposide-treated KPN cells relative to etoposide-treated KP cells, we found that the DNA damage–induced expression of the TNF-α gene set was selectively impaired in etoposide-treated KPN cells. Importantly, RNA-seq data were highly reproducible and consistent among replicates (Fig S2C). Next, we visualised the relative abundance of mRNA transcripts in KP and KPN cells in response to etoposide treatment on volcano plots (Fig 4A). In total, we identified 162 candidate mRNA transcripts that were significantly elevated in levels by etoposide in KP cells. Strikingly, 5 mRNA transcripts that belong to the TNF-α signalling gene set (Lif, Dusp1, Gadd45b, Sgk1, and Phlda1) were found among the top 20 induced candidates and all 5 transcripts failed to be induced by etoposide in KPN cells. Among the five candidates, Gadd45b caught our attention as it represents a prognostic marker in lung cancer and encodes a nuclear protein that modulates both RNAPII activity and the DDR (Niehrs & Schäfer, 2012; Lv et al, 2023). We validated the DNA damage–responsive onset of Gadd45b expression upon etoposide treatment in KP cells and its attenuated induction in KPN cells by RT–qPCR, confocal imaging, and immunoblotting (Figs 4B and C and S2D). Importantly, the attenuated Gadd45b transcript and protein expression in KPN cells could also be observed upon treatment with bleomycin (Figs 4D and S2E). Next, we asked whether the underrepresentation of Gadd45b mRNA transcripts in etoposide-treated KPN cells is reflected in CTD S5P promoter occupancy. We reassessed our CUT&RUN-seq data, which indeed displayed an about twofold reduction in CTD S5P occupancy both at the Gadd45b promoter and at the promoters of 3 of the other 4 TNF-α signalling genes upon combining etoposide treatment with Nono deletion (Fig S3A–C). Reassuringly, etoposide treatment combined with Nono deletion did not severely alter CTD S5P occupancy at the promoter region of at least five other RNA-seq candidates, which were induced by etoposide treatment irrespective of Nono deletion. To test whether the Nono-dependent expression of Gadd45b modulates the DDR, we preincubated KP or KPN cells with the selective Gadd45b inhibitor DTP3, which prevents Gadd45b from binding to the stress-responsive mitogen-activated protein kinase kinase (MAPKK) MKK7, thereby promoting efficient MAPKK signalling (Tornatore et al, 2014). We assessed the level of etoposide-induced γH2A.X signals by immunoblotting and confocal imaging (Fig 5A and B). Indeed, combining etoposide treatment with DTP3 elevated γH2A.X signals in both KP and KPN cells in a dose-dependent manner and appeared additive in KPN cells. We also assessed the CTD S5P phosphorylation status of RNAPII by immunoblotting and observed that neither Nono deletion nor DTP3 treatment altered the amount of total RNAPII or CTD S5P marks upon etoposide treatment (Fig S4A). Importantly, treatment with a high concentration of DTP3 (20 µM) elevated the levels of phosphorylated c-Jun N-terminal kinase (p-JNK) upon etoposide treatment, suggesting that the DTP3 inhibitor is active in both KP and KPN cells (Fig S4B). Finally, we wished to test the impact of the Gadd45b overexpression of genome stability in KP and KPN cells. Using a neutral comet assay and immunoblotting, we found that the overexpression of human HA-tagged GADD45 attenuates the excessive formation of DNA damage upon bleomycin treatment in Nono-deficient KP cells (Figs 5C and D and S4C). We conclude that Nono promotes the expression of Gadd45b to stimulate DSB signalling in KP cells.

We report defects in RNA metabolism upon DNA damage in Nono-deficient KP cells and impaired DNA repair that are associated with defective Gadd45b expression in KPN cells (Fig 5E). How are Nono and Gadd45b linked? Nono associates with chromatin to stimulate both RNAPII activity and the formation of CTD S5P marks at a subset of protein-coding genes that are critical for the differentiation of mouse embryonic stem cells (Ma et al, 2016). Likewise, Gadd45b regulates the mitogen-activated protein kinase pathway, for instance, by activating the p38 kinase, which is also required for the modulation of the RNA metabolism upon DNA damage in human tissue culture cells (Yoo et al, 2003; Borisova et al, 2018). Perturbations in the activity of RNAPII can impact on pre-mRNA processing and, for instance, trigger the accumulation of aberrant transcripts such as DNA-RNA hybrids (R-loops) (de la Mata et al, 2003; Castillo-Guzman et al, 2020 Preprint). High levels of R-loops correlate with genome instability (García-Muse & Aguilera, 2019; Marnef & Legube, 2021). Interestingly, chromatin binding of Gadd45a is R-loop–dependent and the binding sites of human NONO to chromatin also strongly correlate with R-loop–prone sites (Arab et al, 2019; Wu et al, 2022). We postulate that murine Nono induces Gadd45b upon DNA damage to mitigate the formation of R-loops and that Nono and Gadd45b perhaps modulate both RNAPII activity and R-loop levels upon DNA damage in a shared pathway. However, the precise mechanism remains elusive. Of note, human NONO regulates gene expression also at the posttranscriptional level, for instance, by promoting the retention and editing of transcripts (Knott et al, 2016). Thus, we cannot exclude that the posttranscriptional roles of Nono are compromised in KPN cells and contribute to the phenotypes observed in this study.

Our data predict that the inactivation of Nono in KP cells may trigger genome instability upon chemotherapy and thus provide vulnerability in lung cancer. One option may be to interfere with the affinity of Nono to bind chromatin-associated nascent RNA. A panel of electrophilic small molecules has recently been described as direct modulators of human NONO RNA-binding affinity. (R)-SKBG-1 and other NONO ligands react with the cysteine-145 residue of NONO, which stabilises NONO RNA interactions and triggers NONO mislocalisation in nuclear foci that are accompanied by diminished oncogenic gene expression and growth defects in prostate cancer cells (Kathman et al, 2023). Interestingly, blockage of the protein arginine methyltransferase PRMT4/CARM1 with the small molecule inhibitor EZM-2302 has recently been established as novel vulnerability in tumours (Kumar et al, 2021). PRMT4/CARM1 methylates both the RNAPII CTD and the coiled-coil domain of NONO to stimulate RNAPII activity and attenuate the binding affinity of NONO to RNA, respectively (Sims et al, 2011; Hu et al, 2015). Thus, the growth defects observed upon EZM-2302 treatment may, at least in part, be caused by interference with NONO function. It is tempting to speculate that the combination of small molecule inhibitors with genotoxic stress inactivates Nono and triggers RNA-induced toxicity in KP cells. To understand the physiological relevance of the Nono-mediated DDR proposed in this study, it will be important to test the capability of KPN cells to form tumours in vivo and assess the sensitivity of such tumours to genotoxic stress in the future.

Tissue culture and transfections

Murine KP (KRas^G12D, Trp53^−/−) cells, KP-derived Nono knockout KPN (KRas^G12D, Trp53^−/−, Nono^−/−) cells, and human HEK293T cells were cultured in DMEM (Gibco) with 10% FBS (Capricorn), 100 U/ml penicillin–streptomycin (Gibco), and 2 mM L-glutamine (Gibco) at 37°C and 5% CO₂. Cells were incubated with etoposide (20 µM or as indicated; Sigma-Aldrich), bleomycin (100 µg/ml; Hycultec), and cisplatin (50 µM; Sigma-Aldrich) for 2–4 h or chase times or preincubated with the Gadd45b inhibitor DTP3 (Selleckchem) as indicated. Transfection of plasmids pmCherry-NONO (a kind gift from Ling-Ling Chen) and pcDNA3-HA-WT-GADD45 (Addgene) was performed using Lipofectamine 2000 (Invitrogen) and Opti-MEM (Gibco) following the manufacturer’s protocol.

CRISPR cloning and lentiviral transduction

For cloning of the Nono-targeting CRISPR vector, the parental lentiCRISPR_v2 plasmid (a kind gift from Markus Diefenbacher) was restricted with BsmBI-v2 (NEB) and ligated with either of the two preannealed sgRNA-encoding DNA oligonucleotide inserts, which were designed using the CHOPCHOP online tool (Table S1). The correct integration of sgRNA sequences into pCRISPR-V2-sgRNA-Nono-targeting vectors was validated by Sanger sequencing. To generate monoclonal Nono knockout cells, 10 µg pCRISPR-V2-sgRNA-Nono plasmid was pooled with psPAX2 and pMD7.G (kind gifts from Elmar Wolf), mixed with 30 µl polyethylenimine (Calbiochem), diluted in 500 µl Opti-MEM, vortexed, incubated (25 min, RT), added to HEK293 cells that were preincubated in 5 ml DMEM/2% FBS, and transfected (8 h). The virus was harvested two times every 12 h, sterile-filtered, and frozen. For infection, KP cells were cultured (24 h) in viral mixture (1.5 ml DMEM, 1.5 ml viral harvest, 6 µl polybrene; Invitrogen). The mixture was replaced by DMEM with 2.5 µg/ml puromycin (Invitrogen) for polyclonal selection (10 d). Individual Nono knockout candidate clones were yielded by single-cell suspension and validated by downstream assays.

Immunoblotting

Proteins were assessed as whole-cell extracts that were directly lysed, boiled, and sonicated in 4x sample buffer (250 mM Tris–HCl, pH 6.8, 8% SDS, 40% glycerol, 0.8% β-mercaptoethanol, 0.02% bromophenol blue). Samples were separated by SDS–PAGE and transferred to nitrocellulose membranes (Cytiva), blocked and washed in PBS/0.1% Triton x-100/5% milk (PBST), probed with selective antibodies (Table S2), and visualised with an ECL kit (Cytiva) and an imaging station (LAS-4000, Fuji; Fusion FX, Vilber). Signals were quantified by ImageJ (NIH).

Indirect immunofluorescence

Cells were grown on coverslips (Roth), washed in PBS, fixed (10 min) in 3% paraformaldehyde (Sigma-Aldrich), washed in PBS (three times, 5 min), permeabilised with PBS/0.1% Triton x-100 (10 min), and blocked with PBS/10% FBS (2 h, 4°C). Primary and secondary antibodies (Table S2) were diluted in PBS/0.15% FBS and incubated in a humidified chamber (overnight, 4°C, or 2 h, RT), respectively. Cells were washed between incubations with PBS/0.1% Triton x-100 (three times, 5 min), sealed in DAPI-containing mounting medium (Vectashield), and imaged by confocal microscopy (CLSM; Leica SP2, 1,024 × 1,024 resolution, 63x, airy = 1). Channels were acquired as single snapshots, sequentially, between frames, with equal exposure times. Data were analysed using ImageJ (NIH). At least 80 cells per condition were quantified. For imaging of nascent RNA, cells were incubated with 1 ml DMEM containing 1 mM EU (Invitrogen) at 37°C for 1 h, washed, fixed, and permeabilised as described above, and incubated with 500 µl Click-iT reaction cocktail (428 µl Click-iT RNA reaction buffer, 20 µl CuSO₄, 1.6 µl Alexa Fluor 488 azide, 50 µl Click-iT reaction buffer additive; Invitrogen) at RT for 30 min in the dark. Cells were washed once in 1 ml Click-iT reaction rinse buffer, sealed, and imaged as described above.

DNA damage in situ ligation–proximity ligation assay (DI-PLA)

For the DI-PLA, cells were grown, fixed, and permeabilised as above. For DSB blunting, samples were washed twice in 1x CutSmart buffer (2 min, RT; NEB), and incubated (1 h, RT) in 50 µl blunting mix (38.5 µl ddH₂O, 5 µl 10x blunting buffer; NEB, 5 µl 1 mM dNTPs, 0.5 µl 20 mg/ml BSA, 1 µl blunting enzyme mix from Quick Blunting Kit; NEB). For ligation, samples were washed twice in 1x CutSmart buffer (2 min, RT; NEB), preincubated (5 min, RT) in 1x T4 ligase buffer (NEB), and incubated (18 h, 16°C with gentle shaking) in 100 µl ligation buffer (83.5 µl ddH₂O, 10 µl 10x T4 ligase buffer; NEB, 1 µl 20 mg/ml BSA, 2.5 µl 10 µM biotinylated linker pair [Table S3], 2 µl 2,000 U/μL T4 ligase; NEB, 1 µl 100 mM ATP solution; NEB). For removal of the excess linker, samples were washed twice in PBS. Proximity ligation assays were performed with a Duolink in situ PLA kit (Sigma-Aldrich) following the manufacturer’s protocol and assessed by confocal microscopy as described above. For analysis, data were plotted as % of cells from six individual acquisitions that contain ∼50 cells each and were taken as technical replicates.

Neutral comet assay

Glass slides were covered in ddH₂O containing 0.01% poly-L-lysine solution (Sigma-Aldrich) and 1% agarose (Roth) and incubated in a hybridisation oven (UVP) at 70°C overnight. Cells were trypsinised, harvested in 1x PBS, and diluted to 10⁵ cells/ml. The cell suspension was mixed with an equal volume of 1.5% low melting agarose gel (Biozym) in 1x PBS at 37°C, pipetted on preincubated glass slides, and flattened with a coverslip immediately. The slides were cooled down for 10 min at 4°C. After removal of coverslips, lysis buffer (2.5 M NaCl, 0.1 M EDTA, 0.1 M Tris–HCl, pH 10, 1% Triton x-100) was added directly on the slide, covered with parafilm, and incubated (1 h, 4°C). Slides were washed twice in 1x PBS and analysed by electrophoresis (1 V/cm, 15 min) in neutral comet buffer (100 mM Tris base, pH 8.5, 300 mM sodium acetate) at 4°C. Slides were fixed in 70% EtOH and dried at RT overnight, stained with SYBR Gold (Invitrogen), imaged by confocal microscopy, and quantified using CometScore software.

ChIP

Cells were fixed with 1% formaldehyde (10 min, 37°C), quenched in 0.125 M glycine (10 min, 37°C), washed in PBS, and centrifuged (320g, 5 min). Pellets were resuspended in 500 µl cold cell lysis buffer (5 mM PIPES, pH 8.0, 85 mM KCl, 0.5% NP-40, 1x protease/phosphatase inhibitor) and lysed (10 min on ice). Nuclei were centrifuged (720g, 5 min) and resuspended in 300 µl cold nuclear lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris–HCl, pH 8.0, 1x protease/phosphatase inhibitor) and lysed (10 min on ice). Lysates were sonicated five times (5 min, 30 s on/off) with a Bioruptor (Diagenode) and pelleted (13,600g, 10 min). The supernatant was mixed with 2 ml dilution buffer (0.01% SDS, 1.1% Triton x-100, 1.2 mM EDTA, 16.7 mM Tris–HCl, pH 8.0, 167 mM NaCl, 1x protease/phosphatase inhibitor). Diluted samples were aliquoted, and 5 µg antibodies were added (IP sample) or not (input) and incubated overnight (4°C with rotation). For pull-down, 20 µl of Protein G Dynabeads was added to IP samples, incubated (1.5 h with rotation), immobilised on a magnet, and washed with 800 µl wash buffer A (0.1% SDS, 1% Triton x-100, 2 mM EDTA, 20 mM Tris–HCl, pH 8.0, 150 mM NaCl), buffer B (0.1% SDS, 1% Triton x-100, 2 mM EDTA, 20 mM Tris–HCl, pH 8.0, 500 mM NaCl), buffer C (0.25 M LiCl, 1% NP-40, 1% sodium deoxycholate, 1 mM EDTA, and 10 mM Tris–HCl, pH 8.0), and twice with buffer D (10 mM Tris–HCl, pH 8.0, 1 mM EDTA). For elution, samples were incubated with 500 µl elution buffer (1% SDS, 0.1 M NaHCO₃) for 30 min with rotation. Reversal of crosslinks was performed at 65°C overnight after adding 30 µl 5 M NaCl, 1 µl 10 mg/ml RNase A (Sigma-Aldrich), 10 µl 0.5 M EDTA, 20 µl 1 M Tris–HCl, pH 6.8, 2 µl 10 mg/ml proteinase K (Sigma-Aldrich) to input and IP samples. DNA was purified by phenol/chloroform extraction, recovered in ddH₂O, and assessed by quantitative PCR with selective primers (Table S4).

CUT&RUN sequencing

Cells were harvested with Accutase (Sigma-Aldrich), centrifuged (500g, 3 min), and washed three times in 1.5 ml wash buffer (20 mM Hepes, pH 7.5, 150 mM NaCl, 0.5 mM spermidine). Cells were incubated (10 min, RT) with 10 µl concanavalin A–coated magnetic beads (BioMag) resuspended in an equal volume of binding buffer (20 mM Hepes, pH 7.5, 10 mM KCl, 1 mM CaCl₂, 1 mM MnCl₂), immobilised on a magnet, permeabilised with 150 µl antibody buffer (20 mM Hepes, pH 7.5, 150 mM NaCl, 0.5 mM spermidine, 0.05% digitonin, 2 mM EDTA), and incubated with 1 µg primary antibody (4°C, overnight with rotation). Samples were placed on a magnet, washed two times with 1 ml dig-wash buffer (20 mM Hepes, pH 7.5, 150 mM NaCl, 0.5 mM spermidine, 0.05% digitonin), and incubated (1 h, 4°C with rotation) with 150 µl protein A/G–micrococcal nuclease fusion protein (1 µg/ml; CST). Reactions were placed on a magnet, and washed two times with 1 ml dig-wash buffer and once with 1 ml rinse buffer (20 mM Hepes, pH 7.5, 0.05% digitonin, 0.5 mM spermidine). For chromatin digestion and release, samples were incubated (30 min, on ice) in ice-cold digestion buffer (3.5 mM Hepes, pH 7.5, 10 mM CaCl₂, 0.05% digitonin). The reaction was stopped by the addition of 200 µl stop buffer (170 mM NaCl, 20 mM EGTA, 0.05% digitonin, 50 µg/ml RNase A, 25 µg/ml glycogen), and fragments were released by incubation (30 min, 37°C). The supernatant was incubated (1 h, 50°C) with 2 µl 10% SDS and 5 µl proteinase K (10 mg/ml; Sigma-Aldrich). Chromatin was recovered by phenol/chloroform extraction and resuspended in 30 µl TE (1 mM Tris–HCl, pH 8.0, 0.1 mM EDTA). For sequencing, replicates were quantified with a fragment analyser (Advanced Analytical) and subjected to library preparation. Libraries for small DNA fragments (25–75 bp) were prepared based on NEBNext Ultra II DNA Library Prep Kit for Illumina (NEB).

RNA analytics

Total RNA was isolated using TRIzol (Invitrogen) following the manufacturer’s protocol. cDNA was synthesised using SuperScript III reverse transcriptase (Invitrogen) with gene-specific primers (Table S4) and quantified upon reverse transcription–quantitative PCR (RT–qPCR) in a thermocycler (Applied) with PowerUp SYBR Green Master Mix (Applied Biosystems) following the protocols of the manufacturers. For RNA sequencing, cells were directly lysed in 2.1 ml QIAzol (Qiagen). Total RNA was extracted with miRNeasy Kit (Qiagen) using the manufacturer’s protocol. 750 ng of total RNA was diluted in 50 µl ddH₂O. Individual samples were subjected to a poly(A) mRNA magnetic isolation module (NEB), and libraries were prepared with NEBNext Ultra II Directional RNA Library Prep Kit (NEB) using the manufacturer’s protocol. The top 1,000 highly expressed genes in untreated KP cells were used for CUT&RUN-seq metagene plots. Correlation analysis and visualisation of RNA-seq data were performed using the package “ggplot2” in R software using Spearman’s method.

Generation of FASTQ, BAM, and bedGraph files

For CUT&RUN-seq and RNA-seq, base calling was performed using Illumina’s FASTQ generation software v1.0.0 and sequencing quality was tested using FastQC. Reads were mapped with STAR v2.7.10a (RNA-seq) or Bowtie2 v2.3.5.1 (CUT&RUN-seq) to the murine genome mm10, and normalised to the number of mapped reads. BAM files obtained after normalisation were sorted and indexed using SAMtools v1.9. bedGraph files were generated using the genomecov function from BEDTools v2.26.0 (Quinlan & Hall, 2010). CUT&RUN-seq metagene plots of the top 1,000 highly expressed genes were generated using the R package “metagene” with the assay parameters “ChIPseq” and 100 bins. The Integrated Genome Browser was used to visualise these density files. The CUT&RUN-seq signal sum count was performed using BEDTools intersect (Quinlan & Hall, 2010) and visualised with RStudio.

Generation of density and volcano plots

For RNA-seq, gene expression was assessed with featureCounts v2.0.3 using bam files and similarity between replicates of each condition was verified by PCA plot v2.2.1 (Galaxy). Differential gene expression was assessed with edgeR v3.24.1 (Galaxy) using the Benjamini–Hochberg adjusted P-value < 0.05, and an expression filter excluding non- and weakly expressed genes. Genes were plotted with Volcano plot tool v0.05 (Galaxy). Genes with an FDR P-value < 0.05 and log₂FC > 1 or less than −1 were considered significantly up-regulated or down-regulated, respectively. GSEA between conditions was performed with GSEA-Broad tool v4.1.0 (Galaxy), and only gene sets with an FDR < 25% were considered.