Figure 3. SA proteins bind to RNA and localize to R-loops.

CLIP for (a) SA1, (b) SA2 and IgG controls. Autoradiograms of crosslinked ³²P-labeled RNA are shown at the top and the corresponding immunoblots, below. CLIP was performed with and without UV crosslinking and polynucleotide kinase (PNK) and with high (H; 1/50 dilution) or low (L; 1/500 dilution) concentrations of RNase I. (c) CLIP for (a) SA1, SA2 and IgG control in EtOH (-) or IAA-treated (+) Rad21^mAC cells. ³²P-labeled RNA and the corresponding immunoblots are shown as above. (d) Top, Representative confocal images of S9.6 and SA1 IF in RAD21^mAC cells untreated or overexpressing ppyCAG-RNaseH-v5. Expressing cells were identified with v5 staining. Nuclear outlines (white) are derived from DAPI counterstain. Bottom, Imaris quantification of the relative mean fluorescence Intensity (MFI) of S9.6 and SA1. Data are from two biological replicates with >75 cells counted/condition. Quantifications and statistical analysis were done as above. (e) STORM analysis of localization density for SA1 in S9.6 masks in EtOH and IAA. Ratio of SA1 localizations inside and outside S9.6 masks is shown. Ratio of above 1 represents an enrichment within the S9.6 domain. Mean and SD are plotted, statistics based on One-Way Anova test. Data are from two biological replicates. (f) -log10 transformed adjusted p-value (FDR) for enrichment of S9.6 interactome data from Cristini et al. and Wang et al., with the SA1^ΔCoh interactome. Overlap indicates the proteins identified in both of the S9.6 interactome datasets, representing a high confidence R-loop interactome list. (g) Chromatin coIP of S9.6 and IgG in RAD21^mAC cells treated with RNase H and immunoblotted with antibodies representing known R-loop proteins, as well as SA1. Input represents 1.25% of the material used for immunoprecipitation. Bottom, S9.6 dot blot of lysates used in coIP. (h) deepTools heatmap of DRIP-seq and ChIP-seq from RAD21^mAC cells. DRIP-seq was carried out in control (EtOH), RNase H (RNH), and IAA-treated cells. ChIP-seq was carried out for SA1, SA2, RAD21, CTCF, and IgG in EtOH and IAA-treated cells. Regions were selected based on DRIP-seq sensitivity to RNH and proximity with SA1 ChIP-seq. BEDTools identified regions of overlap or adjacent SA1 co-binding. (i) Summary plots showing mean DRIP-seq (top) or ChIP-seq (bottom) read density across the regions from (h), including sites of R-loop and SA ‘overlap’ (Left) or ‘adjacent’ (right) regions. Input samples are indicated with dotted lines.

Figure 3—figure supplement 1. SA proteins bind to RNA and localize to R-loops.

CLIP for (a) SA1 and (b) SA2 treated with various controls as shown. Autoradiograms of crosslinked ³²P-labeled RNA are shown at the top and the corresponding immunoblots, below. (c) CLIP for SA1, SA2, and IgG control in cells treated with siRNAs to scramble control (siscr), SA1 (siSA1), or SA2 (siSA2). ³²P-labeled RNA and the corresponding immunoblots are shown as above. Input material (WCL) is included on the blots. Note the loss of the RNA signal upon siRNA KD. (d) Immunoblot analysis of RAD21, SA1 and SA2 levels in whole cell lysates (WCL) from RAD21^mAC cells treated with EtOH or IAA. Actin serves as a loading control. (e) CLIP for SA1, SA2, and IgG control in EtOH (-) or IAA-treated (+) samples from (d). ³²P-labeled RNA and the corresponding immunoblots are shown as above. Note SA1 and SA2 samples were loaded to show proportional amounts of the proteins in the IAA condition. (f) Representative confocal images of S9.6 and AQR (leftmost panel), SA1 (middle panel), and SA2 (rightmost panel) IF in RAD21^mAC cells treated with scramble control siRNA (si Con) or siRNA to the protein of interest. Nuclear outlines (white) are derived from DAPI counterstain. Mean fluorescence Intensity (MFI) shown below images. Data are from three biological replicates with >50 cells counted/condition. Quantifications and statistical analysis were done as previously stated. (g) Imaris quantification of (left) mClover or (right) S9.6 signal by IF in RAD21^mAC cells treated with EtOH or IAA for 4 hr. Data are from three biological replicates with >50 cells counted/condition. NB We detected a change in S9.6 signal upon IAA treatment by IF; however, this did not reach statistical significance. Quantifications and statistical analysis were done as indicated previously. (h) Replicate chromatin coIP of S9.6 and IgG in RAD21^mAC cells treated with RNase H enzyme (RNH) and immunoblotted with antibodies representing known R-loop proteins and both SA1 and SA2. Input represents 1.25% of the material used for immunoprecipitation. Bottom, S9.6 dot blot of lysates used in coIP. (i) Representative S9.6 dot blot of two dilutions of chromatin samples used in the S9.6 coIPs and treated with enzymes as shown. A positive control for the digestion of RNA:DNA hybrids was included with a high concentration and temperature global RNase A digestion sample. (j) Summary plots showing mean ChIP-seq read density across the regions from Figure 3h including the SA1 ChIP-seq data.(k) ChromHMM analysis as in Figure 1—figure supplement 1i and now including our DRIP-seq data from HCT116 cells. The ‘R-loop_SA overlap’ and ‘R-loop_SA adjacent’ sites are marked in red. Marks enriched within a given state (left) and enriched genomic features (right) are shown. We note that ‘R-loop_SA overlap’ sites enrich genes and active marks such as Polr2a, H3K4me3, and H3K27ac in states 2/3/12 while the ‘R-loop_SA adjacent’ sites cluster separately with repressive chromatin marks such as Lamin B1 LADs in states 7/8.