Linker histone H1 prevents R-loop accumulation and genome instability in heterochromatin

Abstract

Linker histone H1 is an important structural component of chromatin that stabilizes the nucleosome and compacts the nucleofilament into higher-order structures. The biology of histone H1 remains, however, poorly understood. Here we show that Drosophila histone H1 (dH1) prevents genome instability as indicated by the increased γH2Av (H2AvS137P) content and the high incidence of DNA breaks and sister-chromatid exchanges observed in dH1-depleted cells. Increased γH2Av occurs preferentially at heterochromatic elements, which are upregulated upon dH1 depletion, and is due to the abnormal accumulation of DNA:RNA hybrids (R-loops). R-loops accumulation is readily detectable in G1-phase, whereas γH2Av increases mainly during DNA replication. These defects induce JNK-mediated apoptosis and are specific of dH1 depletion since they are not observed when heterochromatin silencing is relieved by HP1a depletion. Altogether, our results suggest that histone H1 prevents R-loops-induced DNA damage in heterochromatin and unveil its essential contribution to maintenance of genome stability.While structural importance of linker histone H1 in packaging eukaryotic genome into chromatin is well known, its biological function remains poorly understood. Here the authors reveal that Drosophila linker histone H1 prevents DNA:RNA hybrids accumulation and genome instability in heterochromatin.

Fig. 1

dH1 depletion induces DNA damage. a Immunostaining of dH1-depleted (siRNA^dH1) and control undepleted cells (siRNA^lacZ and untreated) with αdH1 and αγH2Av antibodies (both in green). DNA was stained with DAPI (blue). Insets show enlarged images of representative individual cells. Scale bars are 20 μm and 2 μm in the Insets. On the right, the number of γH2Av foci per cell is presented (n > 100 for each condition). Error bars are s.e.m. The p-value of siRNA^dH1 respect to siRNA^lacZ is indicated (***p<0.005; two-tailed Student’s t-test). b WB analyses with αdH1, αγH2Av and αH4 of increasing amounts of extracts (lanes 1–3) prepared from siRNA^dH1, siRNA^lacZ and untreated cells. The positions corresponding to molecular weight markers are indicated. On the right, quantitative analysis of the results (N = 3). Error bars are s.e.m. The p-value of siRNA^dH1 respect to siRNA^lacZ is indicated (***<0.005; two-tailed Student’s t-test). c Alkaline and neutral single-cell electrophoresis analyses of siRNA^dH1, siRNA^lacZ and untreated cells. Scale bar corresponds to 20 μm. On the right, relative comet-tail moments are presented (n > 100 for each condition). Error bars are s.e.m. The p-values of siRNA^dH1 respect to siRNA^lacZ are indicated (***<0.005; two-tailed Student’s t-test). d On the top, WB analysis with αγH2Av and αtubulin at different time points after X-ray irradiation (10 Gy) of siRNA^dH1 and untreated cells. The positions corresponding to molecular weight markers are indicated. On the bottom, quantitative analysis of the results (N = 3)

Fig. 2

DNA damage induced by dH1 depletion occurs preferentially in heterochromatin. a The proportion of base pairs (bp) matching to repetitive and non-repetitive elements, as determined by RepeatMasker analysis, are presented for regions showing specific γH2Av enrichment and depletion in dH1-depleted cells respect to control untreated cells. b Permutation experiments showing statistical significance of the enrichment in repeated DNA sequences of the regions showing specific γH2Av enrichment (right) and depletion (left) in dH1-depleted cells respect to control untreated cells. The frequency of the number of overlaps with repetitive DNA elements, as determined by the regioneR package using the UCSC dm3 RepeatMasker track (March 2017), is presented based on 5000 random permutations of the experimentally identified regions. The average expected number of overlaps (black) is compared with the observed number of overlaps (green). The α = 0.05 confidence interval is indicated (red). z-scores and permutation test p-values of the differences are also indicated. c γH2Av ChIP-qPCR analyses at the indicated repetitive and non-repetitive regions in siRNA^dH1, siRNA^lacZ and untreated cells (N = 2). For each position, the fold-change respect to the untreated condition at this position is presented. Error bars are s.e.m. The p-values of siRNA^dH1 respect to siRNA^lacZ are indicated (no asterisk >0.05, *<0.05, **<0.01, ***<0.005; two-tailed Student’s t-test)

Fig. 3

dH1-depleted cells show a high frequency of SCE and chromosome segregation defects. a SCE-assay in siRNA^dH1 cells. Sister chromatids are identified on the basis of their differential Giemsa staining due to their twofold difference in BrdU incorporation. Enlarged images of representative examples are shown on the right. Scale bars are 6 μm and 3 μm in the enlarged images. Red arrowheads indicate crossovers. On the bottom, the percentages of chromosomes with crossovers are presented for siRNA^dH1, siRNA^lacZ and untreated cells expressing human RNH1 (+) or not (−) (n > 100 for each condition). The p-values of siRNA^dH1 respect to siRNA^lacZ are indicated (no asterisk >0.05, ***<0.005; two-tailed Fisher’s exact F-test). b Anaphase figures for siRNA^dH1, siRNA^lacZ and untreated cells stained with DAPI (in white) are presented. Scale bars are 5 μm. Red arrowheads indicate chromatin bridges. On the right, the percentages of mitoses showing segregation defects are presented for siRNA^dH1, siRNA^lacZ and untreated cells (n > 40 for each condition). The p-value of siRNA^dH1 respect to siRNA^lacZ is indicated (***<0.005; two-tailed Fisher’s exact F-test)

Fig. 4

DNA damage induced by dH1 depletion associates with R-loops accumulation. a Immunostainings with αγH2Av (green) and S9.6 (red) antibodies of siRNA^dH1, siRNA^lacZ and untreated cells overexpressing human RNH1 (+) or not (−). Scale bar corresponds to 10 μm. b Quantitative analysis of the results shown in a. S9.6 (top) and γH2Av (center) reactivities determined as the proportion of DAPI area stained with S9.6 antibodies and the number of γH2Av foci per cell are presented (n > 50 for each condition). On the bottom, the extent of γH2Av/S9.6 colocalization is presented as the proportion of γH2Av area overlapping with S9.6 reactivity (n > 50 for each condition). Error bars are s.e.m. The p-values of siRNA^dH1 respect to siRNA^lacZ are indicated (no asterisk >0.05, ***<0.005; two-tailed Student’s t-test). c αγH2Av (top) and S9.6 (bottom) reactivities of G1-, S- and G2/M-phase sorted siRNA^dH1, siRNA^lacZ and untreated cells overexpressing RNH1 (+) or not (−) are presented as in b (n > 50 for each condition). Error bars are s.e.m. The p-values of siRNA^dH1 respect to siRNA^lacZ are indicated (no asterisk >0.05, ***<0.005; two-tailed Student’s t-test). d WB analyses with αdH1, αγH2Av and αH4 antibodies of siRNA^dH1 and untreated cells sorted at G1-, S and G2/M-phase. The positions corresponding to molecular weight markers are indicated. Quantitative analysis is shown on the bottom (N = 2). Error bars are s.e.m. The p-values respect to untreated are indicated (no asterisk >0.05, **<0.01, ***<0.005; two-tailed Student’s t-test)

Fig. 5

R-loops induced by dH1 depletion accumulate in heterochromatin. a The proportion of base pairs (bp) matching to repetitive and non-repetitive elements, as determined by RepeatMasker analysis, are presented for regions showing specific R-loops enrichment and depletion in dH1-depleted cells respect to control untreated cells. b Permutation experiments showing statistical significance of the enrichment in repeated DNA sequences of the regions showing specific R-loops enrichment (right) and depletion (left) in dH1-depleted cells respect to control untreated cells. The frequency of the number of overlaps with repetitive DNA elements, as determined by the regioneR package using the UCSC dm3 RepeatMasker track (March 2017), is presented based on 5000 random permutations of the experimentally identified regions. The average expected number of overlaps (black) is compared with the observed number of overlaps (green). The α = 0.05 confidence interval is indicated (red). z-scores and permutation test p-values of the differences are also indicated. c DRIP-qPCR analyses at the indicated regions in siRNA^dH1 cells, siRNA^lacZ and untreated cells. Before immunoprecipitation samples were treated with bacterial RNH (+) or not (−) (N = 2). For each position, the fold-change respect to the untreated condition at this position is presented. Error bars are s.e.m. The p-values of siRNA^dH1 with respect to siRNA^lacZ are indicated (no asterisk >0.05, *<0.05, **<0.01, ***<0.005; two-tailed Student’s t-test)

Fig. 6

dH1 depletion affects DNA replication. a On the top, DNA fibers labeled with a double pulse of IdU (1^st label, red) and CldU (2^nd label, green). On the bottom, box-plots showing the length of the 2^nd label (left) and the asymmetry index (right) of DNA fibers obtained from siRNA^dH1, siRNA^lacZ and untreated cells overexpressing RNH1 ( + ) or not (−) (n > 100 for each condition). Error bars are s.e.m. The p-values of siRNA^dH1 respect to siRNA^lacZ are indicated (no asterisk >0.05, ***<0.005; two-tailed Wilcoxon test). b Staining for EdU (red) and immunostaining with αHP1a (green) of siRNA^dH1, siRNA^lacZ and untreated cells sorted at early, mid and late S-phase. DNA was stained with DAPI (blue). Scale bar corresponds to 4 μm. c The extent of EdU/HP1a colocalization determined as the normalized proportion of total αHP1a area that shows EdU incorporation is presented (n > 50 for each condition). Error bars are s.e.m. The p-values of siRNA^dH1 respect to siRNA^lacZ are indicated (no asterisk >0.05, ***<0.005; two-tailed Student’s t-test)

Fig. 7

HP1a depletion does not induce R-loops accumulation. a Immunostainings with αγH2Av (red) and S9.6 (green) antibodies of siRNA^HP1a and control untreated cells. Scale bar corresponds to 20 μm. On the right, S9.6 (left) and γH2Av (right) reactivities determined as the proportion of DAPI area stained with S9.6 antibodies and the number of γH2Av foci per cell are presented (n > 50 for each condition). The values obtained for siRNA^dH1 cells are included for comparison. Error bars are s.e.m. The p-values of siRNA^dH1 and siRNA^HP1a respect to untreated are indicated (no asterisk >0.05, ***<0.005; two-tailed Student’s t-test). b WB analyses with αHP1a, αγH2Av and αH4 antibodies of increasing amounts of extracts (lanes 1–2) prepared from siRNA^HP1a and untreated cells. The positions corresponding to molecular weight markers are indicated. On the right, quantitative analysis of the results (N = 2). The values obtained for siRNA^dH1 cells are included for comparison. Error bars are s.e.m. The p-values of siRNA^dH1 and siRNA^HP1a respect to untreated are indicated (no asterisk >0.05, ***<0.005; two-tailed Student’s t-test). c DRIP-qPCR analyses at the indicated repetitive elements in siRNA^HP1a, siRNA^HP1a+dH1, siRNA^lacZ and untreated cells. Before immunoprecipitation samples were treated with bacterial RNH (+) or not (−) (N = 2). For each repetitive element, the fold-change respect to the untreated condition at this repetitive element is presented. Error bars are s.e.m. The p-values of siRNA^HP1a and siRNA^HP1a+dH1 respect to siRNA^lacZ are indicated (no asterisk >0.05, **<0.01, ***<0.005; two-tailed Student’s t-test). d HP1a ChIP-qPCR analysis at the indicated repetitive elements in siRNA^HP1a, siRNA^dH1 and untreated cells. For each repetitive element, the fold-change respect to the untreated condition at this repetitive element is presented. Error bars are s.e.m. The p-values of siRNA^dH1 respect to siRNA^lacZ are indicated (no asterisk >0.05, ***<0.005; two-tailed Student’s t-test). e as in d but for dH1 ChIP-qPCR analysis

Fig. 8

R-loops induced by dH1 depletion activate JNK-dependent apoptosis. a Wings from dH1-depleted his1 ^RNAi flies of the indicated genotypes where dH1 depletion was induced in the pouch region of the wing imaginal disc. Scale bar corresponds to 500 μm. b Quantitative analysis of the wing area of dH1-depleted his1 ^RNAi flies of the indicated genotypes. Data is expressed as fold change respect to control dH1-depleted his1 ^RNAi ; GFP ^RNAi (n > 20 for each condition). Error bars are s.e.m. The p-values respect to control his1 ^RNAi ; GFP ^RNAi are indicated (**<0.01, ***<0.005; two-tailed Student’s t-test). c Immunostaining with αγH2Av (green) and αCaspase 3 (red) of wing imaginal discs from dH1-depleted his1 ^RNAi flies upon dATR ^RNAi co-depletion (bottom) or not (top). DNA was stained with DAPI (blue). d Immunostaining with αCaspase 3 (red) of wing imaginal discs from dH1-depleted his1 ^RNAi flies upon p53 ^RNAi co-depletion (bottom) or p53 ^H159N overexpression (top). DNA was stained with DAPI (blue). e As in d but upon bsk ^RNAi co-depletion (top) or puc ^2A overexpression (bottom). Scale bars in c–e are 200 μm

Fig. 9

Model for the contribution of dH1 to maintenance of genome integrity. In the absence of dH1, expression of heterochromatic elements is upregulated a, inducing the accumulation of R-loops b that could generate SSBs through targeting of the ssDNA regions by DNA-damaging agents or ssDNA gaps through processing of the DNA:RNA hybrids by the NER system b and c. During DNA replication, SSBs and ssDNA gaps are converted into DSBs d and e. Collisions with the replication machinery could also stall replication fork progression and generate DSBs d and e