Tetrameric INTS6-SOSS1 complex facilitates DNA:RNA hybrid autoregulation at double-strand breaks

Abstract

DNA double-strand breaks (DSBs) represent a lethal form of DNA damage that can trigger cell death or initiate oncogenesis. The activity of RNA polymerase II (RNAPII) at the break site is required for efficient DSB repair. However, the regulatory mechanisms governing the transcription cycle at DSBs are not well understood. Here, we show that Integrator complex subunit 6 (INTS6) associates with the heterotrimeric sensor of ssDNA (SOSS1) complex (comprising INTS3, INIP and hSSB1) to form the tetrameric SOSS1 complex. INTS6 binds to DNA:RNA hybrids and promotes Protein Phosphatase 2A (PP2A) recruitment to DSBs, facilitating the dephosphorylation of RNAPII. Furthermore, INTS6 prevents the accumulation of damage-associated RNA transcripts (DARTs) and the stabilization of DNA:RNA hybrids at DSB sites. INTS6 interacts with and promotes the recruitment of senataxin (SETX) to DSBs, facilitating the resolution of DNA:RNA hybrids/R-loops. Our results underscore the significance of the tetrameric SOSS1 complex in the autoregulation of DNA:RNA hybrids and efficient DNA repair.

Graphical Abstract

Figure 1.

INTS6 localizes to DSBs in a DNA:RNA hybrid-dependent manner. (A) Left: Scans of representative EMSA experiments of INTS6 with 61- or 21-mer ssDNA, 61-mer dsDNA, R-loops and DNA:RNA hybrids. Right: Graph representing quantification of EMSA experiments (n = 3). (B) Left: Scans of representative EMSA experiments of the tetrameric SOSS1 complex with 61- or 21-mer ssDNA, 61-mer dsDNA, R-loops and DNA:RNA hybrids. Right: Graph representing quantification of EMSA experiments (n = 3). (C) Laser striping of cells transiently transfected with INTS6-GFP plasmid with or without RNase H1-RFP plasmid. White lines indicate laser stripes. Representative confocal microscopy images and quantification (n ≥ 20) show GFP and RFP signals at the indicated time points. Error bars: mean ± SEM. (D) PLA of INTS6 and S9.6 in cells without IR, with IR or with IR and RNaseH1 treatment. IR = 10 Gy, samples were collected 10 min post-IR. Here, we used modified PLA protocol (55), which includes RNaseT1 (digest ssRNA) and RNase III (digest dsRNA) treatment during slide preparation process. Top: Representative confocal microscopy images; bottom: quantification of top, error bar = mean ± SD, significance was determined using non-parametric Mann–Whitney test. ****P ≤ 0.0001. Scale bar = 10 μm. Single antibodies were used as a negative control. (E) PLA of INTS6 and γH2AX in cells transiently transfected with RNAseH1^wt-GFP or RNAseH1^WKKD-GFP (binding and catalytic) or RNAseH1^D210N−GFP (catalytic) mutants with or without IR. IR = 10 Gy, samples were collected 10 min post-IR. Left: representative confocal microscopy images; right: quantification of left, error bar = mean ± SD, significance was determined using non-parametric Mann–Whitney test. ****P ≤ 0.0001, *P ≤ 0.05. Scale bar = 10 μm. Single antibodies were used as a negative control.

Figure 2.

INTS6 facilitates PP2A recruitment to DSBs to dephosphorylate RNAPII. (A) PLA of PP2A and γH2AX in WT or INTS6 knockdown cells with or without IR. IR = 10 Gy, samples were collected 10 min post-IR. Left: Representative confocal microscopy images; right: quantification of left, error bar = mean ± SD, significance was determined using non-parametric Mann–Whitney test., ****P ≤ 0.0001, ***P ≤ 0.001, *P ≤ 0.05. Scale bar = 10 μm. Single antibodies were used as a negative control. (B) PLA of PP2A and γH2AX in cells with or without transient RNaseH1-GFP expression, with or without IR. IR = 10Gy, samples were collected 10 min post-IR. Left: Representative confocal microscopy images; right: quantification of left, error bar = mean ± SD, significance was determined using non-parametric Mann–Whitney test. ****P ≤ 0.0001, *P ≤ 0.05. Scale bar = 10 μm. Single antibodies were used as a negative control. (C) PLA of S5P and γH2AX in wildtype or INTS6 knockdown cells with or without IR. IR = 10 Gy, samples were collected 10 min post-IR. Left: Representative confocal microscopy images; right: quantification of left, error bar = mean ± SD, significance was determined using non-parametric Mann–Whitney test. ****P ≤ 0.0001, *P ≤ 0.05. Scale bar = 10 μm. Single antibodies were used as a negative control. (D) PLA of S5P and γH2AX with or without IR in the presence or absence of PP2A inhibitor (LB-100, 2.5 μM, 2 h). IR = 10 Gy, samples were collected 10 min post-IR. Left: Representative confocal microscopy images; right: quantification of left, error bar = mean ± SD, significance was determined using non-parametric Mann–Whitney test. ****P ≤ 0.0001, **P ≤ 0.01. Scale bar = 10 μm. Single antibodies were used as a negative control. (E) Left: Drawing of ChIP probes positions around DS1. Right: drawing showing position of ChIP probes in GAPDH gene. (F) Bar charts showing RNAPII, S2P and S5P ChIP signals at DS1 in the absence or presence of PP2A inhibitor (LB-100, 2.5 μM, 4 h). n ≥ 3. Error bar = mean ± SD, significance was determined using unpaired Student’s t-test, ****P ≤ 0.0001, ***P ≤ 0.001,**P ≤ 0.01*P ≤ 0.05.

Figure 3.

INTS6 depletion leads to the accumulation of DARTs. (A) Metagene plot shows chrRNA-seq sense and antisense coverage in no damage (-4OHT), damage (siCtrl + 4OHT) and INTS6 knockdown (siINTS6 + 4OHT) cells around 2.5 kb flank region of cut AsiSI sites (details in Supplementary Table S4). The reference genome is human hg19. (B) Metagene plot shows chrRNA-seq sense and antisense coverage in no damage (-4OHT), damage (siCtrl + 4OHT) and INTS6 knockdown (siINTS6 + 4OHT) cells around 2.5 kb flank region of uncut AsiSI sites (details in Supplementary Table S4). The reference genome is human hg19. (C) Box plots show log2 fold change of chrRNA-seq coverage of sense reads and antisense reads upon INTS6 knockdown with damage induction compared to control with damage induction for cut AsiSI (± 500 bp). Two-sample Wilcoxon test is used for statistical testing of medians between sense and antisense log2 fold change distribution. (D) Heatmaps show antisense nascent RNA (chrRNA-seq) read coverage across annotated AsiSI sites (details in Supplementary Table S4) sorted based on their cleavage efficiency. (E) Box plot shows chrRNA-seq antisense coverage in ± 500 bp flank of AsiSI cut sites (details in Supplementary Table S4) in -4OHT, siCtrl + 4OHT and siINTS6 + 4OHT samples. Two-sample Wilcoxon test is used with medians test. Error bar = mean ± SD. (F, G) Representative snapshots of individual genes showing sense and antisense chrRNA-seq coverage in INTS6 knockdown and control with damage induction around 2.5 kb flank region of AsiSI cut. The specific loci information is listed on top of the snapshots. The reference genome is human hg19.

Figure 4.

INTS6 associates with SETX. (A) Silver stain of affinity-purified Integrator complex. The integrator complex was purified from nuclear lysate of HEK293Tcells, stably overexpressing FLAG-INTS6. Mock indicates the same FLAG-IP purification steps from parental HEK293T cells. The indicated Integrator subunits were assigned as identified by Baillat et al. (31). (B) Affinity-purified Integrator complex mass spectrometry analyses were performed on nuclear lysate of HEK293T cells stably overexpressing FLAG-INTS6 or mock FLAG-IP purification steps from parental HEK293T cells. The values represent intensity-based absolute quantification (iBAQ) intensities and the unique peptides. (C) Affinity-purified Integrator complex followed by western blot showing indicated proteins. (D) PLA of SETX and INTS6 with or without IR treatment. IR = 10 Gy, samples were collected 10 min post-IR. Left: Representative confocal microscopy images; right: quantification of left, error bar = mean ± SD, significance was determined using non-parametric Mann–Whitney test. ****P ≤ 0.0001. Scale bar = 10 μm. Single antibodies were used as a negative control. (E) Immunoprecipitation of SETX from cells with or without IR treatment (IR = 10 Gy, samples were collected 10 min post-IR), followed by Western blot showing signals for SETX, INTS6, INTS3 and hSSB1. KDa indicates size of the proteins. Bar charts show quantifications of three independent blots, error bar = mean ± SD, significance was determined using unpaired Student’s t-test, *P ≤ 0.05.

Figure 5.

INTS6 is required for SETX recruitment to DSBs and clearance of DNA:RNA hybrids. (A) PLA of SETX and γH2AX with or without IR in mock or siINTS6 cells. IR = 10 Gy, samples were collected 10 min post-IR. Left: Representative confocal microscopy images; right: quantification of left, error bar = mean ± SD, significance was determined using non-parametric Mann–Whitney test. ****P ≤ 0.0001. Scale bar = 10 μm. Single antibodies were used as a negative control. (B, C) Bar charts showing SETX ChIP signals over DS1 (B) and DS2 (C) loci in the presence or absence of INTS6. n = 3. Error bar = mean ± SD, significance was determined using Student’s t-test, unpaired, **P ≤ 0.01. (D) PLA of S9.6 and γH2AX with or without IR in mock or siINTS6 cells. IR = 10 Gy, samples were collected 10 min post-IR. Left: Representative confocal microscopy images; right: quantification of top, error bar = mean ± SD, significance was determined using non-parametric Mann–Whitney test. ****P ≤ 0.0001. Scale bar = 10 μm. Single antibodies were used as a negative control. (E, F) Bar charts showing DRIP signals over DS1 (E) and DS2 (F) loci in the presence or absence of INTS6 and RNaseH1 treatment. n = 3. Error bar = mean ± SD, significance was determined using unpaired Student’s t-test, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05.

Figure 6.

INTS6-dependent accumulation of DARTs correlates with DNA:RNA hybrids at DSBs. (A) Heatmaps show the chrRNA-seq coverage in siINTS6 + 4OHT, siCtrl + 4OHT, -4OHT, SETX + 4OHT ChIP-seq coverage and S9.6 + 4OHT DRIP-seq coverage across cut AsiSI sites (details in Supplementary Table S4) sorted by their cleavage efficiency. The reference genome is human hg19. (B) Box plots show log2 fold change of total chrRNA-seq coverage of INTS6 knockdown with damage induction compared to control with damage induction in 500 bp bins centered at DSB, 1, 2 and 3 kb away from DSB (cut AsiSI sites, details in Supplementary Table S4). The reference genome is human hg19. (C) Metagene plots show S9.6 DRIP-seq coverage and SETX coverage upon damage induction along with chrRNA-seq sense and antisense coverage around 2.5 kb flank region of cut AsiSI sites (details in Supplementary Table S4). The reference genome is human hg19. (D) Metagene plots show S9.6 DRIP-seq coverage and SETX coverage upon damage induction along with chrRNA-seq sense and antisense coverage around 2.5 kb flank region of uncut AsiSI sites (n = 20). The reference genome is human hg19. (E, F) Representative snapshots of individual genes showing DRIP-seq coverage and SETX coverage upon damage induction along with sense and antisense chrRNA-seq coverage in siINTS6 and control with damage induction around 2.5 kb flank region of AsiSI cut. The specific loci information is listed on top of the snapshot respectively. The reference genome is human hg19.

Figure 7.

INTS6 is required for efficient DNA damage repair. (A) Left: Representative images of the clonogenic assay in control and INTS6 knockdown cells. The cells were stained and counted after 10 days of growing. Right: Quantification of left. * P ≤ 0.05, **P ≤ 0.01. (B) MTT assay to show cell viability (%) at indicated time points in siCtrl and siINTS6 cells with or without 2 Gy IR treatment. Error bar = mean ± SD, significance was determined using unpaired Student’s t-test, *P ≤ 0.05. (C) Left: Drawing of DR-GFP HR reporter strategy. Right: Bar chart shows the efficiency of HR repair in DR-GFP HeLa reporter cells, as measured by FACS. BRCA1 knockdown was used as the positive control. ****P ≤ 0.0001, ***P ≤ 0.001. (D) Top: Representative images of comet assay in siCtrl, siINTS6, siRAD51 and si53BP1 cells. siRAD51 and si53BP1 cells were used as positive controls. IR = 5 Gy. Error bar = 100 μm. Bottom: Quantification of top. ****P ≤ 0.0001, **P ≤ 0.01. (E) Model: INTS6 as part of tetrameric SOSS1 complex binds to DNA:RNA hybrids at DSBs and recruits PP2A to dephosphorylate RNAPII. Depletion of INTS6 results in increased levels of DARTs and DNA:RNA hybrids. INTS6 interacts with SETX and is required for its recruitment to damaged sites. SETX, in turn, resolves DNA:RNA hybrids at DSBs facilitating their INTS6-dependent autoregulation. Image was created with Biorender.