Molecular Cell
Short Article R Loops Are Linked to Histone H3 S10 Phosphorylation and Chromatin Condensation Maikel Castellano-Pozo,1 Jose´ M. Santos-Pereira,1 Ana G. Rondo´n,1 Sonia Barroso,1 Eloisa Andu´jar,2 Mo´nica Pe´rez-Alegre,2 Tatiana Garcı´a-Muse,1 and Andre´s Aguilera1,2,* 1Department
of Molecular Biology Unit, Centro Andaluz de Biologı´a Molecular y Medicina Regenerativa CABIMER Universidad de Sevilla-CSIC, Avenida Ame´rico Vespucio, 41092 Seville, Spain *Correspondence: [email protected] http://dx.doi.org/10.1016/j.molcel.2013.10.006 2Genomic
SUMMARY
R loops are transcription byproducts that constitute a threat to genome integrity. Here we show that R loops are tightly linked to histone H3 S10 phosphorylation (H3S10P), a mark of chromatin condensation. Chromatin immunoprecipitation (ChIP)-on-chip (ChIP-chip) analyses reveal H3S10P accumulation at centromeres, pericentromeric chromatin, and a large number of active open reading frames (ORFs) in R-loop-accumulating yeast cells, better observed in G1. Histone H3S10 plays a key role in maintaining genome stability, as scored by ectopic recombination and plasmid loss, Rad52 foci, and Rad53 checkpoint activation. H3S10P coincides with the presence of DNA-RNA hybrids, is suppressed by ribonuclease H overexpression, and causes reduced accessibility of restriction endonucleases, implying a tight connection between R loops, H3S10P, and chromatin compaction. Such histone modifications were also observed in R-loop-accumulating Caenorhabditis elegans and HeLa cells. We therefore provide a role of RNA in chromatin structure essential to understand how R loops modulate genome dynamics. INTRODUCTION R loops are essential in replication initiation of mitochondrial DNA, in an alternative mechanism of replication initiation in bacteria, and in class-switch recombination in vertebrate B cells (Aguilera and Garcı´a-Muse, 2012) and have been suggested to play a role in transcription termination (Skourti-Stathaki et al., 2011). However, R loops are potentially harmful structures that promote genome instability by impairing replication fork progression (Gan et al., 2011; Wellinger et al., 2006). Accordingly, cells use factors to either remove R loops, such as ribonuclease H (RNase H) or the Sen1/SETX DNA-RNA helicase (Cerritelli and Crouch, 2009; Mischo et al., 2011), or prevent their formation, such as a number of mRNA biogenesis and processing proteins (Huertas and Aguilera, 2003; Li and Manley, 2005; Paulsen et al., 2009; Stirling et al., 2012; Wahba et al., 2011), the THO complex being a paradigmatic example conserved from yeast to humans
(Domı´nguez-Sa´nchez et al., 2011). Different mechanisms have been proposed to explain how R loops can contribute to genome instability. It is believed that R loops can either enhance the susceptibility of DNA to endogenous genotoxic agents or be an obstacle to the progression of replication fork that can lead to its collapse (Aguilera and Garcı´a-Muse 2012). However, we do not yet know the mechanism and factors responsible for genome instability mediated by R loops or whether additional elements, such as R-loop-mediated chromatin modifications, could modulate genome dynamics. Here we investigated the possible connections of R loops with histone modifications as a first step to explore additional elements by which R loops can contribute to genome instability. Strikingly, we found that R loops are linked to histone H3 S10 phosphorylation, a mark of chromatin condensation that opens unexpected views to understand the role of R loops in genome dynamics.
RESULTS High Levels of the H3S10P Chromatin Compaction Mark in R-Loop-Accumulating Strains during Mitosis and Meiosis To address whether R loops were connected to specific chromatin modifications in Saccharomyces cerevisiae, we assayed first by western blotting if posttranslational histone modifications were altered in R-loop-accumulating mutants. Notably, we found that histone H3 was highly phosphorylated at S10 (H3S10P), a marker of chromatin condensation (Hsu et al., 2000), both in mitotic and meiotic cells. In wild-type (WT) cells, H3S10P was either poorly or not detected in G1 during the mitotic cycle. Instead, H3S10 was clearly phosphorylated during G1 in Rloop-accumulating hpr1D mutants of the THO complex as well as at high levels in S/G2 (Figure 1A). In meiosis, H3S10P was observed only 1.5 hr after initiating synchronous meiosis in WT cells, whereas in hpr1D cells, H3S10P was clearly detected at all meiotic stages, as determined in synchronous meiosis of SK1 isogenic cells (Figure 1B), as the SK1 genetic background is able to undergo rapid and synchronized meiosis. Importantly, this mark is not specific to THO mutants, as we can also see a significant presence of H3S10P in G1-arrested cells of R-loopaccumulating rnh1D rnh2D strains (Figure S1A available online). Even though hpr1D cells remained longer in S/G2, the difference in H3S10P with respect to WT persists and even increases when cells are arrested in G1 for 3 hr (Figure S1B), confirming a true
Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc. 583
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
Figure 1. High Levels of Histone H3S10P in R-Loop-Accumulating THO Mutants and Genetic Instability in H3S10A and H3S10D Mutants (A) Immunoblotting with aH3S10P, aH3, aClb2, and ab-actin after G1 arrest with a factor and release of W303 wild-type and hpr1D cells. DNA content was analyzed by flow cytometry. (B) Immunoblotting with aH3S10P, aH3, and ab-actin after release in sporulation media of NKY1551 wild-type and hpr1D diploid cells. DNA content was analyzed by flow cytometry. (C) Recombination frequencies in wild-type, H3S10A, H3S10D, hpr1D, hpr1D H3S10A, and hpr1D H3S10D cells. (D) Rad52 foci in asynchronous cultures of the strains shown in (C). (E) Stability of centromeric plasmid pRS414 in wild-type, H3S10A, H3S10D, and hpr1D strains. (F) Immunoblotting with aRad53 and ab-actin of asynchronous cultures of the strains shown in (C). Data are represented as mean ± SD from three independent experiments. Mutant values were significantly different from WT values, as determined by Student’s t tests (*p < 0.05). See also Figure S1.
phosphorylation of H3S10 in G1 hpr1D cells that does not occur in WT cells. Other histone H3 modifications, such as H3K14 acetylation (H3K14ac) and H3K4 trimethylation (H3K4me3) known to be associated with transcription, presented the same pattern as the wild-type, as determined by western blotting (Figure S1C) and by chromatin immunoprecipitation (ChIP) in the GAL1FMP27 fusion construct (Figure S1D). Background levels of H3S10P were observed all over the FMP27 gene in wild-type and hpr1D cells (Figure S1D). The results indicate that the genome of R-loop-accumulating strains shows high levels of the H3S10P chromatin mark, known to associate with chromatin condensation, in mitosis and meiosis. Deregulation of Histone H3S10 Phosphorylation Results in Genome Instability We next determined whether such a chromatin mark had any effect on genome instability, a hallmark of R-loop-accumulating mutants. For this, we used specific histone H3 point mutations
at S10 that cannot be phosphorylated (H3S10A) or that mimic constitutive phosphorylation (H3S10D). A slight, but significant, increase in direct-repeat recombination together with an increase in Rad52 foci and plasmid loss was observed in both H3S10 mutants (Figures 1C–1E). In addition, the S phase checkpoint was activated, as determined by Rad53 phosphorylation in both H3S10 mutants (Figure 1F). The results indicate that H3S10 is important for genome integrity, consistent with our prediction. However, the similar phenotypes observed in both H3S10 mutants indicate that the phosphorylation state of this residue, or even the S10 residue itself, may control genome dynamics by itself, whether or not R loops are involved. Indeed, Rad52 foci accumulation and hyperrecombination of hpr1D cells are partially suppressed in the double mutants hpr1D H3S10A and hpr1D H3S10D (Figures 1C and 1D), confirming the role of the S10 residue in genome instability in hpr1D cells regardless of its R-loop-dependent phosphorylation stage. It is possible that H3S10 itself, or its phosphorylation state, controls different DNA metabolic processes, including replication and repair, by helping recruit specific factors, but this is an open question. A similar situation can be that of other histone modifications like H3K56 acetylation (Mun˜oz-Galva´n et al., 2013). Genome-wide Distribution of H3S10P in R-Loop-Accumulating Strains To get further insight into the physiological meaning of the H3S10P mark, we determined the distribution of this mark along
584 Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc.
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
Figure 2. Genome-wide Accumulation of H3S10P in R-Loop-Accumulating hpr1D Cells during G1 (A) Statistical analysis of H3S10P hits (p value < 0.01) obtained in the ChIP-chip analysis according to the different annotation categories of the Stanford Genome Database for wild-type and hpr1D strains. *Values significantly different from wildtype (p < 0.001). (B) ChIP-chip distribution of the H3S10P signal along chromosomes IV, VI, and VIII. The y axis shows the signal log2 ratio significantly enriched in the immunoprecipitated fraction of wild-type and hpr1D G1-arrested cells. (C) ChIP-chip profile of H3S10P enrichment at centromeric and 40 kb pericentromeric regions. The y axis shows the median of H3S10P-IP (log2 ratio) signal per segment of all the chromosomes for wild-type and hpr1D G1-arrested cells. (D) Profile of H3S10P enrichment across the ORFs of either the 16 pericentromeric regions of 40 kb or the whole genome plotted as the average signal (log2 ratio) significantly enriched (p % 0.01) in the immunoprecipitated fraction per segment in wildtype and hpr1D G1-arrested cells. The length of the ORFs was divided in 10 segments and analyzed together with the 500 bp immediately upstream and downstream of the segments, as previously described (Go´mez-Gonza´lez et al., 2011). See also Figures S2 and S3.
the yeast genome of the R-loop-accumulating hpr1D strain compared to WT cells in G1 and S phase by ChIP-on-chip (ChIP-chip) analyses. Notably, in contrast to WT cells, there is a much more abundant and spread distribution of H3S10P along all chromosomes in the hpr1D strain that is clearly manifested in G1 cells (Figures 2 and S2). The H3S10P signal was clearly overabundant at the centromeres and the pericentromeric regions as well as all over the genome in G1-arrested hpr1D cells (Figures 2A, 2B, and S2). The number of genes located in the 40 kb pericentromeric region and all over the genome that contained H3S10P signals was significantly higher (p < 0.001) in G1-arrested hpr1D cells than in wild-type cells (Figure 2A). Interestingly, a detailed statistical analysis of all 40 kb pericentromeric regions of the genome revealed a significantly high H3S10P accumulation in hpr1D cells not observed in the wild-type, while the global amount of histone H3 was unchanged (Figures 2C and S3A). The accumulation profile of H3S10P at the pericentromeric genes was clearly higher along the whole extension of the open reading frames (ORFs) analyzed in hpr1D cells as compared to WT cells, which show a reduced signal (Figure 2D). Importantly, even though this higher H3S10P accumulation is better observed in the pericen-
tromeric ORFs, it is also higher along the length of most ORFs, pericentromeric or not (Figures 2D and S3). Moreover, 92% of all H3S10P-positive ORFs (2,373 out of 2,578) show transcription at highly detectable levels in hpr1D cells (Go´mez-Gonza´lez et al., 2011), indicating that H3S10P is a feature of transcribed DNA regions. In general, a higher H3S10P signal could still be seen in S cells, in which H3S10P was clearly detected at centromeres (Figures 2A, S2, and S3), but at this phase the differences between hpr1D and WT cells diminished, consistent with the western analysis (Figure 1A). No significant differences could be seen in tRNAs and other DNA regions at any cell cycle stage, although the number of autonomously replicating sequences (ARSs) with H3S10P was higher in G1-arrested cells (Figure 2A). H3S10P Is Linked to R Loops and Chromatin Compaction The results suggest that the high levels of H3S10P are linked to R loops. To determine whether this was the case, we performed ChIP and DNA-RNA immunoprecipitation (DRIP) analysis of the pericentromeric region in chromosome VI with anti-H3S10P and anti-RNA-DNA hybrid (S9.6) antibodies, respectively (Figures 3A and 3B). G1 cells clearly accumulated the H3S10P signal in hpr1D, but not in WT cells, validating the ChIP-chip results. This signal was abolished by RNH1 overexpression (Figure 3A), indicating a tight link between R loops and H3S10P. Importantly, RNH1 overexpression does not affect cell-cycle progression in either WT or in hpr1D cells, excluding any possible effect of
Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc. 585
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
Figure 3. H3S10P Is Directly Linked to R Loop Accumulation (A) H3S10P ChIP analysis on wild-type and hpr1D G1-arrested cells with or without RNase H1 overexpression, showing that hpr1D centromeric and pericentromeric H3S10P accumulation is suppressed by RNase H1 overexpression. The y axis shows the H3S10P-IP/input average of three independent experiments measured at the pericentromeric region of CEN6. (B) hpr1D accumulates R loops at centromeres and pericentromeric regions. DRIP analyses of wild-type and hpr1D G1-arrested cells, with or without RNase H1 treatment, using the S9.6 antiRNA-DNA hybrid antibody. The y axis shows the S9.6-IP/input average of three independent experiments measured at the pericentromeric region of CEN6. (C) H3S10P ChIP and S9.6 DRIP assays of wildtype and hpr1D G1-arrested cells with or without RNase H1 overexpression or treatment analyzed at the indicated loci. S9.6 signals are normalized to a chromosome V (Chr V) intergenic region (9716-9863 Chr V). (D) hpr1D cells show chromatin compaction at centromeres as determined by restriction endonuclease accessibility. Restriction map of the CEN3 region. Southern of EcoRI-digested chromatin from wild-type, hpr1D, and hpr1-101 G1arrested cells incubated with increasing amounts of DraI restriction enzyme is shown. Data are represented as mean ± SEM from three independent experiments. See also Figure S4.
cell-cycle stage in the results (Figure S4A). Notably, RNA-DNA hybrids were detected in these regions in hpr1D, but not in WT cells (Figure 3B). Moreover, R loops accumulated in hpr1D were strongly reduced by RNH1 treatment, whereas WT cells with or without RNH1 treatment were unable to accumulate detectable R loops (Figure 3B). This strong correlation between H3S10P and R loops was also observed in genes located at different distances from the centromere (Figure 3C), implying that it occurs all over the genome and not only at centromeric and pericentromeric regions. Therefore, R loops are physically linked to H3S10P. To determine whether R-loop-dependent H3S10P behaved as a mark of chromatin compaction or condensation (Hsu et al., 2000; Wei et al., 1999), we assayed the accessibility of the restriction enzyme DraI to the pericentromeric chromatin by monitoring DNA cleavage in vivo. Clearly, the efficiency of DraI and EcoRI cleavage was significantly reduced in the R-loop-accumulating hpr1D cells, but not in WT cells (Figure 3D). Consistently, in the hpr1-101 mutant, which carries a defective hpr1 allele that does not enhance R loops (Go´mezGonza´lez and Aguilera, 2009), H3S10 phosphorylation (Figure S4B) and restriction enzyme accessibility were the same as in the wild-type (Figure 3D). Importantly, the difference in DraI accessibility is also observed in noncentromeric regions (Figure S4C), matching with the results of H3S10P distribution. These results are consistent with the conclusion that the H3S10P mark reflects a major compaction or condensation of chromatin.
The Link between R Loops and H3S10P Also Occurs in C. elegans and Human Cells In parallel with our yeast study, we wondered whether this phenomenon was conserved in multicellular eukaryotes. Given that R loops also accumulate in C. elegans thoc-2 null mutants of the THO complex, causing replication defects linked to genome instability (Castellano-Pozo et al., 2012), we determined by immunostaining whether chromatin compaction also occurred in these C. elegans strains. Notably, histone H3 was not only highly phosphorylated at S10 (H3S10P), but it was also dimethylated at K9 (H3K9me2), a marker of heterochromatin not conserved in yeast (Hsu et al., 2000; Reddy and Villeneuve, 2004), both at mitotic and pachytene meiotic regions of the germline of the R-loop-accumulating thoc-2 strain (Figure 4A). Such modifications are not found in N2(WT) worms during mitosis, whereas in meiosis they are observed only after pachytene, consistent with them being marks of condensation. Indeed, DAPI staining shows bright and condensed nuclei in the germline of thoc-2 worms, as compared to those of N2(WT) worms (Figure S4D). In agreement with the yeast data, we detect S9.6 foci in thoc-2 germlines, but not in N2(WT) worms (Figure S4E). No difference between N2(WT) and thoc-2 was observed for the H4 tetra-acetylation pattern used as a marker of euchromatin (Figure 4A) (Reddy and Villeneuve, 2004). Next, we asked whether this phenomenon was also conserved in human cells. For this, we assayed H3S10P accumulation in R-loop-accumulating THOC1- and senataxin (SETX)-depleted human cells (Domı´nguez-Sa´nchez et al., 2011; Skourti-Stathaki
586 Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc.
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
Figure 4. R-Loop-Accumulating C. elegans Mutants and HeLa Cells Show Histone Modifications of Chromatin Compaction or Condensation (A) C. elegans thoc-2 mutants show high levels of heterochromatin and condensation marks as compared with N2(WT). Representative images of mitotic or meiotic regions from fixed germlines of N2(WT) and thoc-2 mutant worms immunostained with aH4ac, aH3K9me2, and aH3S10P and counterstained with DAPI. (B) Immunofluorescence of H3S10P in THOC1- or SETX-depleted HeLa cells with or without RNase H1 overexpression. S phase replicating cells are labeled with thymidine analog EdU. (C) Percentage of G1 cells with >5 H3S10P foci. Data are represented as mean ± SEM from three independent experiments. *Values significantly different as determined by Mann-Whitney U test (p < 0.01). (D) Model showing how cotranscriptional R loops could trigger H3S10 phosphorylation and chromatin compaction or condensation. Such compaction would contribute to replication fork stalling or unfinished replication (drawn) responsible for genome instability, as well as transcription elongation impairment and silencing (not drawn). Nucleosomes around the DNA-RNA hybrid stretch appear transparent because we do not know, at this point, whether or not they would be there. See also Figure S4.
absence of THOC1 or SETX (Figures 4B and 4C). Moreover, RNase H1 overexpression decreases H3S10P foci formation, supporting the connection between H3S10P and R loops (Figures 4B and 4C). The results suggest that the link between R loops and chromatin condensation marks such as H3S10P is conserved in all eukaryotes, including yeast, nematodes, and humans. DISCUSSION
et al., 2011). We included senataxin-depleted cells to assay whether the R-loop-associated H3S10 phosphorylation occurred regardless of the system we used to enhance R loops. We performed immunofluorescence with anti-H3S10P antibody in HeLa cells grown in ethynyldeoxyuridine (EdU)-containing media to differentiate nonreplicating G1 cells from actively replicating S phase cells. Consistent with the yeast and worm results, G1 HeLa cells with H3S10P foci were only observed in the
The link between different histone modifications and RNA molecules has been shown in eukaryotes. A well-known example is the X-inactivation in mammals by RNA Xist (Lee and Bartolomei, 2013) or gene silencing linked to long noncoding RNAs (ncRNA) that, in most cases, is mediated by siRNAs (Bernstein and Allis, 2005) or via chromatin deacetylation by an antisense ncRNA, as shown in S. cerevisiae (Camblong et al., 2007). Interestingly, RNA-DNA hybrids prevent methylation and transcriptional silencing at CpG island promoters (Ginno et al., 2012) and induce transcriptional silencing, likely through siRNA, in S. pombe (Nakama et al., 2012). Furthermore, centromeres and pericentromeric regions are known to be transcribed, and transcription of pericentromeric DNA repeats is
Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc. 587
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
required for pericentromeric heterochromatin formation in S. pombe (Chan and Wong, 2012). Histone H3S10P and chromatin condensation have never been connected to RNA metabolism, even though H3S10P also controls transcription elongation by releasing the RNA polymerase II (RNAPII) from promoter-proximal pausing sites in Drosophila (Ivaldi et al., 2007). It is believed that H3S10P may modulate the access of different proteins to chromatin, establishing a hierarchy of subsequent events that affect chromatin structure and function (Liokatis et al., 2012). Our study shows that R loops are linked to H3S10P and chromatin compaction. The H3S10P mark is scattered at transcribed genes all over the genome, even though it is better detected at the pericentromeric and centromeric regions. This phenomenon is independent of siRNA, provided that S. cerevisiae does not have siRNA processes. According to our results, R loops can either trigger phosphorylation of H3S10 or inhibit its dephosphorylation. This can be achieved by either enhancement of the recruitment of the responsible kinase or inhibition of the activity of the responsible phosphatase. As more than one enzyme may be involved in this process (Baek, 2011), further extensive work will be needed to decipher the molecular mechanism of this control. Nevertheless, it has been shown that H3S10P levels in yeast depend on the Ipl1 kinase and the Glc7 phosphatase and that Ipl1 and Glc7 interact physically or genetically with factors, such as Set1 and Sen1 (Breitkreutz et al., 2010; Nedea et al., 2008; Zhang et al., 2005), that have been implicated in the metabolism of R loops (Mischo et al., 2011). These data open the possibility of a functional link between R loops and H3S10P that would need to be explored in the future. Although our study does not establish whether R loops help chromatin condensation or vice versa, we favor a model in which R loops trigger the formation of highly condensed chromatin patches that would appear throughout the genome, in particular at centromere and pericentromeric regions due to its higher capacity to nucleate chromatin condensation (Figure 4D). This model is supported by the reduction of H3S10P levels along the yeast genome of hpr1D cells by RNase H1 overexpression. Thus, it is possible that condensation spreads from the R loop. Importantly, such condensed regions would interfere with replication and/or transcription (Figure 4D), causing transcription-associated genome instability and gene silencing. This could contribute to the previously reported effect of R loops on both processes in eukaryotes (Huertas and Aguilera 2003). Interestingly, the observations that common fragile sites (CFSs) can be explained by retarded or incomplete replication at chromatin-condensed regions (Debatisse et al., 2012) and that CFSs can be related to cotranscriptional R loops (Helmrich et al., 2011) open the possibility that local chromatin compaction contributes to the replication fork progression impairment or to a delayed or less-efficient replication initiation as possible factors that cause genome instability. This unexpected link between R loops, histone H3S10P, and chromatin compaction or condensation therefore opens unexpected perspectives with which to understand the mechanism by which RNA controls genome dynamics and function.
EXPERIMENTAL PROCEDURES Yeast Strains and Growth The yeast strains used in this work are listed in Table S1. Cells were arrested in G1 with a factor (0.14 mg/ml) and in S phase with 200 mM hydroxyurea (HU). For meiotic analysis, synchronous meiotic cultures were prepared as previously described (Alani et al., 1990). Nematode Strains and Growth The nematode strain Bristol N2 used as wild-type was provided by the Caenorhabditis Genetics Center. The thoc-2 (tm1310) mutant was previously characterized (Castellano-Pozo et al., 2012). C. elegans strains were grown at 20 C, and standard methods were used for their maintenance and manipulation (Brenner, 1974). Human Cell Culture and Transfection HeLa cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; GIBCO) supplemented with 10% heat-inactivated fetal bovine serum at 37 C (5% CO2). Transient transfection of siRNA (10 nM) was performed using DharmaFECT (Dharmacon) according to the manufacturer’s instructions. Then, 96 hr after transfection, cells were incubated with 20 mM EdU (Click-iT EdU Imaging Kit, Invitrogen) for 20 min to label replicating cells. ChIP and DIP/DRIP Assays For ChIP experiments at the pericentromeric region of CEN6, cells harboring the empty vector pCM189 or the ptetRNH1 plasmid were grown in synthetic complete media without uracil (SC-ura), and samples were treated as described (Duch et al., 2013). Antibodies used are listed in Table S2. For real-time PCR, oligonucleotides for the CEN6 pericentromeric region previously described (Choy et al., 2011) and for other loci (Table S3) were used. Means ± SD of three independent experiments are shown. DRIP experiments were carried out essentially as originally described like DNA IP (DIP) (Mischo et al., 2011). Means ± SD of three independent experiments are shown. ChIP-Chip Assays Cells were grown in YEPD medium and arrested in G1 with a factor (0.14 mg/ml) for 2 hr (G1 phase) and then released in new rich medium with 200 mM HU for another hour (S phase). ChIP-chip was carried out as previously described (Bermejo et al., 2007; Bermejo et al., 2009; Katou et al., 2003; 2006). For details and bioinformatics analysis, see Supplemental Information. Chromatin Accessibility Assays Nuclei preparations from G1-arrested cells with a factor (0.14 mg/ml) were made as previously described (Crotti and Basrai, 2004). Similar quantities of nuclei were incubated with 0, 5, 15, 50, and 150 units/ml of DraI at 37 C for 30 min followed by digestion with 250 U/ml EcoRI. Southern blotting was performed, and filters were probed with a 32P-deoxycytidine triphosphate (dCTP)-labeled CEN3 0.9 kb HindIII-BamHI fragment following standard procedures. Miscellanea Immunofluorescence analyses using the indicated antibodies were performed in 1 day post-L4 C. elegans adult gonads treated as described (CastellanoPozo et al., 2012) and in human cells cultured on glass coverslips and fixed as described previously (Domı´nguez-Sa´nchez et al., 2011). EdU detection was performed using a Click-iT EdU Imaging Kit (Invitrogen) following manufacturer’s instructions. DNA was stained with DAPI. Recombination and plasmid-loss assays and Rad52 foci are described in Go´mez-Gonza´lez et al. (2009) and the Supplemental Information. ACCESSION NUMBERS The Gene Expression Omnibus accession number for the set of ChIP-chip data reported in this paper is GSE46627.
588 Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc.
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
SUPPLEMENTAL INFORMATION Supplemental Information includes Supplemental Experimental Procedures, four figures, and three tables and can be found with this article online at http://dx.doi.org/10.1016/j.molcel.2013.10.006. ACKNOWLEDGMENTS We thank J.C. Reyes and F. Monje-Casas for useful comments on the manuscript and D. Haun for style correction. Research was funded by grants from the Spanish Ministry of Economy and Competitiveness (Consolider Ingenio 2010 CSD2007-0015 and BFU2010-16372), Junta de Andalucı´a (CVI-4567), and the European Union (FEDER). M.C.-P. and J.M.S.-P. were recipients of predoctoral training grants from the Spanish National Institute of Health Carlos III and the Consejo Superior de Investigaciones Cientı´ficas (CSIC), respectively.
Domı´nguez-Sa´nchez, M.S., Barroso, S., Go´mez-Gonza´lez, B., Luna, R., and Aguilera, A. (2011). Genome instability and transcription elongation impairment in human cells depleted of THO/TREX. PLoS Genet. 7, e1002386. Duch, A., Felipe-Abrio, I., Barroso, S., Yaakov, G., Garcı´a-Rubio, M., Aguilera, A., de Nadal, E., and Posas, F. (2013). Coordinated control of replication and transcription by a SAPK protects genomic integrity. Nature 493, 116–119. Gan, W., Guan, Z., Liu, J., Gui, T., Shen, K., Manley, J.L., and Li, X. (2011). R-loop-mediated genomic instability is caused by impairment of replication fork progression. Genes Dev. 25, 2041–2056. Ginno, P.A., Lott, P.L., Christensen, H.C., Korf, I., and Che´din, F. (2012). R-loop formation is a distinctive characteristic of unmethylated human CpG island promoters. Mol. Cell 45, 814–825. Go´mez-Gonza´lez, B., and Aguilera, A. (2009). R-loops do not accumulate in transcription-defective hpr1-101 mutants: implications for the functional role of THO/TREX. Nucleic Acids Res. 37, 4315–4321. Go´mez-Gonza´lez, B., Felipe-Abrio, I., and Aguilera, A. (2009). The S-phase checkpoint is required to respond to R-loops accumulated in THO mutants. Mol. Cell. Biol. 29, 5203–5213.
Received: June 17, 2013 Revised: September 4, 2013 Accepted: October 1, 2013 Published: November 7, 2013
Go´mez-Gonza´lez, B., Garcı´a-Rubio, M., Bermejo, R., Gaillard, H., Shirahige, K., Marı´n, A., Foiani, M., and Aguilera, A. (2011). Genome-wide function of THO/TREX in active genes prevents R-loop-dependent replication obstacles. EMBO J. 30, 3106–3119.
REFERENCES Aguilera, A., and Garcı´a-Muse, T. (2012). R loops: from transcription byproducts to threats to genome stability. Mol. Cell 46, 115–124. Alani, E., Padmore, R., and Kleckner, N. (1990). Analysis of wild-type and rad50 mutants of yeast suggests an intimate relationship between meiotic chromosome synapsis and recombination. Cell 61, 419–436. Baek, S.H. (2011). When signaling kinases meet histones and histone modifiers in the nucleus. Mol. Cell 42, 274–284. Bermejo, R., Doksani, Y., Capra, T., Katou, Y.M., Tanaka, H., Shirahige, K., and Foiani, M. (2007). Top1- and Top2-mediated topological transitions at replication forks ensure fork progression and stability and prevent DNA damage checkpoint activation. Genes Dev. 21, 1921–1936. Bermejo, R., Katou, Y.M., Shirahige, K., and Foiani, M. (2009). ChIP-on-chip analysis of DNA topoisomerases. Methods Mol. Biol. 582, 103–118.
Helmrich, A., Ballarino, M., and Tora, L. (2011). Collisions between replication and transcription complexes cause common fragile site instability at the longest human genes. Mol. Cell 44, 966–977. Hsu, J.Y., Sun, Z.W., Li, X., Reuben, M., Tatchell, K., Bishop, D.K., Grushcow, J.M., Brame, C.J., Caldwell, J.A., Hunt, D.F., et al. (2000). Mitotic phosphorylation of histone H3 is governed by Ipl1/aurora kinase and Glc7/PP1 phosphatase in budding yeast and nematodes. Cell 102, 279–291. Huertas, P., and Aguilera, A. (2003). Cotranscriptionally formed DNA:RNA hybrids mediate transcription elongation impairment and transcription-associated recombination. Mol. Cell 12, 711–721. Ivaldi, M.S., Karam, C.S., and Corces, V.G. (2007). Phosphorylation of histone H3 at Ser10 facilitates RNA polymerase II release from promoter-proximal pausing in Drosophila. Genes Dev. 21, 2818–2831.
Bernstein, E., and Allis, C.D. (2005). RNA meets chromatin. Genes Dev. 19, 1635–1655.
Katou, Y., Kanoh, Y., Bando, M., Noguchi, H., Tanaka, H., Ashikari, T., Sugimoto, K., and Shirahige, K. (2003). S-phase checkpoint proteins Tof1 and Mrc1 form a stable replication-pausing complex. Nature 424, 1078–1083.
Breitkreutz, A., Choi, H., Sharom, J.R., Boucher, L., Neduva, V., Larsen, B., Lin, Z.Y., Breitkreutz, B.J., Stark, C., Liu, G., et al. (2010). A global protein kinase and phosphatase interaction network in yeast. Science 328, 1043– 1046.
Katou, Y., Kaneshiro, K., Aburatani, H., and Shirahige, K. (2006). Genomic approach for the understanding of dynamic aspect of chromosome behavior. Methods Enzymol. 409, 389–410.
Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genetics 77, 71–94. Camblong, J., Iglesias, N., Fickentscher, C., Dieppois, G., and Stutz, F. (2007). Antisense RNA stabilization induces transcriptional gene silencing via histone deacetylation in S. cerevisiae. Cell 131, 706–717. Castellano-Pozo, M., Garcı´a-Muse, T., and Aguilera, A. (2012). R-loops cause replication impairment and genome instability during meiosis. EMBO Rep. 13, 923–929. Cerritelli, S.M., and Crouch, R.J. (2009). Ribonuclease H: the enzymes in eukaryotes. FEBS J. 276, 1494–1505. Chan, F.L., and Wong, L.H. (2012). Transcription in the maintenance of centromere chromatin identity. Nucleic Acids Res. 40, 11178–11188. Choy, J.S., Acun˜a, R., Au, W.C., and Basrai, M.A. (2011). A role for histone H4K16 hypoacetylation in Saccharomyces cerevisiae kinetochore function. Genetics 189, 11–21. Crotti, L.B., and Basrai, M.A. (2004). Functional roles for evolutionarily conserved Spt4p at centromeres and heterochromatin in Saccharomyces cerevisiae. EMBO J. 23, 1804–1814. Debatisse, M., Le Tallec, B., Letessier, A., Dutrillaux, B., and Brison, O. (2012). Common fragile sites: mechanisms of instability revisited. Trends Genet. 28, 22–32.
Lee, J.T., and Bartolomei, M.S. (2013). X-inactivation, imprinting, and long noncoding RNAs in health and disease. Cell 152, 1308–1323. Li, X., and Manley, J.L. (2005). Inactivation of the SR protein splicing factor ASF/SF2 results in genomic instability. Cell 122, 365–378. Liokatis, S., Stu¨tzer, A., Elsa¨sser, S.J., Theillet, F.X., Klingberg, R., van Rossum, B., Schwarzer, D., Allis, C.D., Fischle, W., and Selenko, P. (2012). Phosphorylation of histone H3 Ser10 establishes a hierarchy for subsequent intramolecular modification events. Nat. Struct. Mol. Biol. 19, 819–823. Mischo, H.E., Go´mez-Gonza´lez, B., Grzechnik, P., Rondo´n, A.G., Wei, W., Steinmetz, L., Aguilera, A., and Proudfoot, N.J. (2011). Yeast Sen1 helicase protects the genome from transcription-associated instability. Mol. Cell 41, 21–32. Mun˜oz-Galva´n, S., Jimeno, S., Rothstein, R., and Aguilera, A. (2013). Histone H3K56 acetylation, Rad52, and non-DNA repair factors control double-strand break repair choice with the sister chromatid. PLoS Genet. 9, e1003237. Nakama, M., Kawakami, K., Kajitani, T., Urano, T., and Murakami, Y. (2012). DNA-RNA hybrid formation mediates RNAi-directed heterochromatin formation. Genes Cells 17, 218–233. Nedea, E., Nalbant, D., Xia, D., Theoharis, N.T., Suter, B., Richardson, C.J., Tatchell, K., Kislinger, T., Greenblatt, J.F., and Nagy, P.L. (2008). The Glc7 phosphatase subunit of the cleavage and polyadenylation factor is essential for transcription termination on snoRNA genes. Mol. Cell 29, 577–587.
Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc. 589
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
Paulsen, R.D., Soni, D.V., Wollman, R., Hahn, A.T., Yee, M.C., Guan, A., Hesley, J.A., Miller, S.C., Cromwell, E.F., Solow-Cordero, D.E., et al. (2009). A genome-wide siRNA screen reveals diverse cellular processes and pathways that mediate genome stability. Mol. Cell 35, 228–239. Reddy, K.C., and Villeneuve, A.M. (2004). C. elegans HIM-17 links chromatin modification and competence for initiation of meiotic recombination. Cell 118, 439–452.
Wahba, L., Amon, J.D., Koshland, D., and Vuica-Ross, M. (2011). RNase H and multiple RNA biogenesis factors cooperate to prevent RNA:DNA hybrids from generating genome instability. Mol. Cell 44, 978–988. Wei, Y., Yu, L., Bowen, J., Gorovsky, M.A., and Allis, C.D. (1999). Phosphorylation of histone H3 is required for proper chromosome condensation and segregation. Cell 97, 99–109.
Skourti-Stathaki, K., Proudfoot, N.J., and Gromak, N. (2011). Human senataxin resolves RNA/DNA hybrids formed at transcriptional pause sites to promote Xrn2-dependent termination. Mol. Cell 42, 794–805.
Wellinger, R.E., Prado, F., and Aguilera, A. (2006). Replication fork progression is impaired by transcription in hyperrecombinant yeast cells lacking a functional THO complex. Mol. Cell. Biol. 26, 3327–3334.
Stirling, P.C., Chan, Y.A., Minaker, S.W., Aristizabal, M.J., Barrett, I., Sipahimalani, P., Kobor, M.S., and Hieter, P. (2012). R-loop-mediated genome instability in mRNA cleavage and polyadenylation mutants. Genes Dev. 26, 163–175.
Zhang, K., Lin, W., Latham, J.A., Riefler, G.M., Schumacher, J.M., Chan, C., Tatchell, K., Hawke, D.H., Kobayashi, R., and Dent, S.Y. (2005). The Set1 methyltransferase opposes Ipl1 aurora kinase functions in chromosome segregation. Cell 122, 723–734.
590 Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc.
Short Article R Loops Are Linked to Histone H3 S10 Phosphorylation and Chromatin Condensation Maikel Castellano-Pozo,1 Jose´ M. Santos-Pereira,1 Ana G. Rondo´n,1 Sonia Barroso,1 Eloisa Andu´jar,2 Mo´nica Pe´rez-Alegre,2 Tatiana Garcı´a-Muse,1 and Andre´s Aguilera1,2,* 1Department
of Molecular Biology Unit, Centro Andaluz de Biologı´a Molecular y Medicina Regenerativa CABIMER Universidad de Sevilla-CSIC, Avenida Ame´rico Vespucio, 41092 Seville, Spain *Correspondence: [email protected] http://dx.doi.org/10.1016/j.molcel.2013.10.006 2Genomic
SUMMARY
R loops are transcription byproducts that constitute a threat to genome integrity. Here we show that R loops are tightly linked to histone H3 S10 phosphorylation (H3S10P), a mark of chromatin condensation. Chromatin immunoprecipitation (ChIP)-on-chip (ChIP-chip) analyses reveal H3S10P accumulation at centromeres, pericentromeric chromatin, and a large number of active open reading frames (ORFs) in R-loop-accumulating yeast cells, better observed in G1. Histone H3S10 plays a key role in maintaining genome stability, as scored by ectopic recombination and plasmid loss, Rad52 foci, and Rad53 checkpoint activation. H3S10P coincides with the presence of DNA-RNA hybrids, is suppressed by ribonuclease H overexpression, and causes reduced accessibility of restriction endonucleases, implying a tight connection between R loops, H3S10P, and chromatin compaction. Such histone modifications were also observed in R-loop-accumulating Caenorhabditis elegans and HeLa cells. We therefore provide a role of RNA in chromatin structure essential to understand how R loops modulate genome dynamics. INTRODUCTION R loops are essential in replication initiation of mitochondrial DNA, in an alternative mechanism of replication initiation in bacteria, and in class-switch recombination in vertebrate B cells (Aguilera and Garcı´a-Muse, 2012) and have been suggested to play a role in transcription termination (Skourti-Stathaki et al., 2011). However, R loops are potentially harmful structures that promote genome instability by impairing replication fork progression (Gan et al., 2011; Wellinger et al., 2006). Accordingly, cells use factors to either remove R loops, such as ribonuclease H (RNase H) or the Sen1/SETX DNA-RNA helicase (Cerritelli and Crouch, 2009; Mischo et al., 2011), or prevent their formation, such as a number of mRNA biogenesis and processing proteins (Huertas and Aguilera, 2003; Li and Manley, 2005; Paulsen et al., 2009; Stirling et al., 2012; Wahba et al., 2011), the THO complex being a paradigmatic example conserved from yeast to humans
(Domı´nguez-Sa´nchez et al., 2011). Different mechanisms have been proposed to explain how R loops can contribute to genome instability. It is believed that R loops can either enhance the susceptibility of DNA to endogenous genotoxic agents or be an obstacle to the progression of replication fork that can lead to its collapse (Aguilera and Garcı´a-Muse 2012). However, we do not yet know the mechanism and factors responsible for genome instability mediated by R loops or whether additional elements, such as R-loop-mediated chromatin modifications, could modulate genome dynamics. Here we investigated the possible connections of R loops with histone modifications as a first step to explore additional elements by which R loops can contribute to genome instability. Strikingly, we found that R loops are linked to histone H3 S10 phosphorylation, a mark of chromatin condensation that opens unexpected views to understand the role of R loops in genome dynamics.
RESULTS High Levels of the H3S10P Chromatin Compaction Mark in R-Loop-Accumulating Strains during Mitosis and Meiosis To address whether R loops were connected to specific chromatin modifications in Saccharomyces cerevisiae, we assayed first by western blotting if posttranslational histone modifications were altered in R-loop-accumulating mutants. Notably, we found that histone H3 was highly phosphorylated at S10 (H3S10P), a marker of chromatin condensation (Hsu et al., 2000), both in mitotic and meiotic cells. In wild-type (WT) cells, H3S10P was either poorly or not detected in G1 during the mitotic cycle. Instead, H3S10 was clearly phosphorylated during G1 in Rloop-accumulating hpr1D mutants of the THO complex as well as at high levels in S/G2 (Figure 1A). In meiosis, H3S10P was observed only 1.5 hr after initiating synchronous meiosis in WT cells, whereas in hpr1D cells, H3S10P was clearly detected at all meiotic stages, as determined in synchronous meiosis of SK1 isogenic cells (Figure 1B), as the SK1 genetic background is able to undergo rapid and synchronized meiosis. Importantly, this mark is not specific to THO mutants, as we can also see a significant presence of H3S10P in G1-arrested cells of R-loopaccumulating rnh1D rnh2D strains (Figure S1A available online). Even though hpr1D cells remained longer in S/G2, the difference in H3S10P with respect to WT persists and even increases when cells are arrested in G1 for 3 hr (Figure S1B), confirming a true
Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc. 583
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
Figure 1. High Levels of Histone H3S10P in R-Loop-Accumulating THO Mutants and Genetic Instability in H3S10A and H3S10D Mutants (A) Immunoblotting with aH3S10P, aH3, aClb2, and ab-actin after G1 arrest with a factor and release of W303 wild-type and hpr1D cells. DNA content was analyzed by flow cytometry. (B) Immunoblotting with aH3S10P, aH3, and ab-actin after release in sporulation media of NKY1551 wild-type and hpr1D diploid cells. DNA content was analyzed by flow cytometry. (C) Recombination frequencies in wild-type, H3S10A, H3S10D, hpr1D, hpr1D H3S10A, and hpr1D H3S10D cells. (D) Rad52 foci in asynchronous cultures of the strains shown in (C). (E) Stability of centromeric plasmid pRS414 in wild-type, H3S10A, H3S10D, and hpr1D strains. (F) Immunoblotting with aRad53 and ab-actin of asynchronous cultures of the strains shown in (C). Data are represented as mean ± SD from three independent experiments. Mutant values were significantly different from WT values, as determined by Student’s t tests (*p < 0.05). See also Figure S1.
phosphorylation of H3S10 in G1 hpr1D cells that does not occur in WT cells. Other histone H3 modifications, such as H3K14 acetylation (H3K14ac) and H3K4 trimethylation (H3K4me3) known to be associated with transcription, presented the same pattern as the wild-type, as determined by western blotting (Figure S1C) and by chromatin immunoprecipitation (ChIP) in the GAL1FMP27 fusion construct (Figure S1D). Background levels of H3S10P were observed all over the FMP27 gene in wild-type and hpr1D cells (Figure S1D). The results indicate that the genome of R-loop-accumulating strains shows high levels of the H3S10P chromatin mark, known to associate with chromatin condensation, in mitosis and meiosis. Deregulation of Histone H3S10 Phosphorylation Results in Genome Instability We next determined whether such a chromatin mark had any effect on genome instability, a hallmark of R-loop-accumulating mutants. For this, we used specific histone H3 point mutations
at S10 that cannot be phosphorylated (H3S10A) or that mimic constitutive phosphorylation (H3S10D). A slight, but significant, increase in direct-repeat recombination together with an increase in Rad52 foci and plasmid loss was observed in both H3S10 mutants (Figures 1C–1E). In addition, the S phase checkpoint was activated, as determined by Rad53 phosphorylation in both H3S10 mutants (Figure 1F). The results indicate that H3S10 is important for genome integrity, consistent with our prediction. However, the similar phenotypes observed in both H3S10 mutants indicate that the phosphorylation state of this residue, or even the S10 residue itself, may control genome dynamics by itself, whether or not R loops are involved. Indeed, Rad52 foci accumulation and hyperrecombination of hpr1D cells are partially suppressed in the double mutants hpr1D H3S10A and hpr1D H3S10D (Figures 1C and 1D), confirming the role of the S10 residue in genome instability in hpr1D cells regardless of its R-loop-dependent phosphorylation stage. It is possible that H3S10 itself, or its phosphorylation state, controls different DNA metabolic processes, including replication and repair, by helping recruit specific factors, but this is an open question. A similar situation can be that of other histone modifications like H3K56 acetylation (Mun˜oz-Galva´n et al., 2013). Genome-wide Distribution of H3S10P in R-Loop-Accumulating Strains To get further insight into the physiological meaning of the H3S10P mark, we determined the distribution of this mark along
584 Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc.
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
Figure 2. Genome-wide Accumulation of H3S10P in R-Loop-Accumulating hpr1D Cells during G1 (A) Statistical analysis of H3S10P hits (p value < 0.01) obtained in the ChIP-chip analysis according to the different annotation categories of the Stanford Genome Database for wild-type and hpr1D strains. *Values significantly different from wildtype (p < 0.001). (B) ChIP-chip distribution of the H3S10P signal along chromosomes IV, VI, and VIII. The y axis shows the signal log2 ratio significantly enriched in the immunoprecipitated fraction of wild-type and hpr1D G1-arrested cells. (C) ChIP-chip profile of H3S10P enrichment at centromeric and 40 kb pericentromeric regions. The y axis shows the median of H3S10P-IP (log2 ratio) signal per segment of all the chromosomes for wild-type and hpr1D G1-arrested cells. (D) Profile of H3S10P enrichment across the ORFs of either the 16 pericentromeric regions of 40 kb or the whole genome plotted as the average signal (log2 ratio) significantly enriched (p % 0.01) in the immunoprecipitated fraction per segment in wildtype and hpr1D G1-arrested cells. The length of the ORFs was divided in 10 segments and analyzed together with the 500 bp immediately upstream and downstream of the segments, as previously described (Go´mez-Gonza´lez et al., 2011). See also Figures S2 and S3.
the yeast genome of the R-loop-accumulating hpr1D strain compared to WT cells in G1 and S phase by ChIP-on-chip (ChIP-chip) analyses. Notably, in contrast to WT cells, there is a much more abundant and spread distribution of H3S10P along all chromosomes in the hpr1D strain that is clearly manifested in G1 cells (Figures 2 and S2). The H3S10P signal was clearly overabundant at the centromeres and the pericentromeric regions as well as all over the genome in G1-arrested hpr1D cells (Figures 2A, 2B, and S2). The number of genes located in the 40 kb pericentromeric region and all over the genome that contained H3S10P signals was significantly higher (p < 0.001) in G1-arrested hpr1D cells than in wild-type cells (Figure 2A). Interestingly, a detailed statistical analysis of all 40 kb pericentromeric regions of the genome revealed a significantly high H3S10P accumulation in hpr1D cells not observed in the wild-type, while the global amount of histone H3 was unchanged (Figures 2C and S3A). The accumulation profile of H3S10P at the pericentromeric genes was clearly higher along the whole extension of the open reading frames (ORFs) analyzed in hpr1D cells as compared to WT cells, which show a reduced signal (Figure 2D). Importantly, even though this higher H3S10P accumulation is better observed in the pericen-
tromeric ORFs, it is also higher along the length of most ORFs, pericentromeric or not (Figures 2D and S3). Moreover, 92% of all H3S10P-positive ORFs (2,373 out of 2,578) show transcription at highly detectable levels in hpr1D cells (Go´mez-Gonza´lez et al., 2011), indicating that H3S10P is a feature of transcribed DNA regions. In general, a higher H3S10P signal could still be seen in S cells, in which H3S10P was clearly detected at centromeres (Figures 2A, S2, and S3), but at this phase the differences between hpr1D and WT cells diminished, consistent with the western analysis (Figure 1A). No significant differences could be seen in tRNAs and other DNA regions at any cell cycle stage, although the number of autonomously replicating sequences (ARSs) with H3S10P was higher in G1-arrested cells (Figure 2A). H3S10P Is Linked to R Loops and Chromatin Compaction The results suggest that the high levels of H3S10P are linked to R loops. To determine whether this was the case, we performed ChIP and DNA-RNA immunoprecipitation (DRIP) analysis of the pericentromeric region in chromosome VI with anti-H3S10P and anti-RNA-DNA hybrid (S9.6) antibodies, respectively (Figures 3A and 3B). G1 cells clearly accumulated the H3S10P signal in hpr1D, but not in WT cells, validating the ChIP-chip results. This signal was abolished by RNH1 overexpression (Figure 3A), indicating a tight link between R loops and H3S10P. Importantly, RNH1 overexpression does not affect cell-cycle progression in either WT or in hpr1D cells, excluding any possible effect of
Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc. 585
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
Figure 3. H3S10P Is Directly Linked to R Loop Accumulation (A) H3S10P ChIP analysis on wild-type and hpr1D G1-arrested cells with or without RNase H1 overexpression, showing that hpr1D centromeric and pericentromeric H3S10P accumulation is suppressed by RNase H1 overexpression. The y axis shows the H3S10P-IP/input average of three independent experiments measured at the pericentromeric region of CEN6. (B) hpr1D accumulates R loops at centromeres and pericentromeric regions. DRIP analyses of wild-type and hpr1D G1-arrested cells, with or without RNase H1 treatment, using the S9.6 antiRNA-DNA hybrid antibody. The y axis shows the S9.6-IP/input average of three independent experiments measured at the pericentromeric region of CEN6. (C) H3S10P ChIP and S9.6 DRIP assays of wildtype and hpr1D G1-arrested cells with or without RNase H1 overexpression or treatment analyzed at the indicated loci. S9.6 signals are normalized to a chromosome V (Chr V) intergenic region (9716-9863 Chr V). (D) hpr1D cells show chromatin compaction at centromeres as determined by restriction endonuclease accessibility. Restriction map of the CEN3 region. Southern of EcoRI-digested chromatin from wild-type, hpr1D, and hpr1-101 G1arrested cells incubated with increasing amounts of DraI restriction enzyme is shown. Data are represented as mean ± SEM from three independent experiments. See also Figure S4.
cell-cycle stage in the results (Figure S4A). Notably, RNA-DNA hybrids were detected in these regions in hpr1D, but not in WT cells (Figure 3B). Moreover, R loops accumulated in hpr1D were strongly reduced by RNH1 treatment, whereas WT cells with or without RNH1 treatment were unable to accumulate detectable R loops (Figure 3B). This strong correlation between H3S10P and R loops was also observed in genes located at different distances from the centromere (Figure 3C), implying that it occurs all over the genome and not only at centromeric and pericentromeric regions. Therefore, R loops are physically linked to H3S10P. To determine whether R-loop-dependent H3S10P behaved as a mark of chromatin compaction or condensation (Hsu et al., 2000; Wei et al., 1999), we assayed the accessibility of the restriction enzyme DraI to the pericentromeric chromatin by monitoring DNA cleavage in vivo. Clearly, the efficiency of DraI and EcoRI cleavage was significantly reduced in the R-loop-accumulating hpr1D cells, but not in WT cells (Figure 3D). Consistently, in the hpr1-101 mutant, which carries a defective hpr1 allele that does not enhance R loops (Go´mezGonza´lez and Aguilera, 2009), H3S10 phosphorylation (Figure S4B) and restriction enzyme accessibility were the same as in the wild-type (Figure 3D). Importantly, the difference in DraI accessibility is also observed in noncentromeric regions (Figure S4C), matching with the results of H3S10P distribution. These results are consistent with the conclusion that the H3S10P mark reflects a major compaction or condensation of chromatin.
The Link between R Loops and H3S10P Also Occurs in C. elegans and Human Cells In parallel with our yeast study, we wondered whether this phenomenon was conserved in multicellular eukaryotes. Given that R loops also accumulate in C. elegans thoc-2 null mutants of the THO complex, causing replication defects linked to genome instability (Castellano-Pozo et al., 2012), we determined by immunostaining whether chromatin compaction also occurred in these C. elegans strains. Notably, histone H3 was not only highly phosphorylated at S10 (H3S10P), but it was also dimethylated at K9 (H3K9me2), a marker of heterochromatin not conserved in yeast (Hsu et al., 2000; Reddy and Villeneuve, 2004), both at mitotic and pachytene meiotic regions of the germline of the R-loop-accumulating thoc-2 strain (Figure 4A). Such modifications are not found in N2(WT) worms during mitosis, whereas in meiosis they are observed only after pachytene, consistent with them being marks of condensation. Indeed, DAPI staining shows bright and condensed nuclei in the germline of thoc-2 worms, as compared to those of N2(WT) worms (Figure S4D). In agreement with the yeast data, we detect S9.6 foci in thoc-2 germlines, but not in N2(WT) worms (Figure S4E). No difference between N2(WT) and thoc-2 was observed for the H4 tetra-acetylation pattern used as a marker of euchromatin (Figure 4A) (Reddy and Villeneuve, 2004). Next, we asked whether this phenomenon was also conserved in human cells. For this, we assayed H3S10P accumulation in R-loop-accumulating THOC1- and senataxin (SETX)-depleted human cells (Domı´nguez-Sa´nchez et al., 2011; Skourti-Stathaki
586 Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc.
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
Figure 4. R-Loop-Accumulating C. elegans Mutants and HeLa Cells Show Histone Modifications of Chromatin Compaction or Condensation (A) C. elegans thoc-2 mutants show high levels of heterochromatin and condensation marks as compared with N2(WT). Representative images of mitotic or meiotic regions from fixed germlines of N2(WT) and thoc-2 mutant worms immunostained with aH4ac, aH3K9me2, and aH3S10P and counterstained with DAPI. (B) Immunofluorescence of H3S10P in THOC1- or SETX-depleted HeLa cells with or without RNase H1 overexpression. S phase replicating cells are labeled with thymidine analog EdU. (C) Percentage of G1 cells with >5 H3S10P foci. Data are represented as mean ± SEM from three independent experiments. *Values significantly different as determined by Mann-Whitney U test (p < 0.01). (D) Model showing how cotranscriptional R loops could trigger H3S10 phosphorylation and chromatin compaction or condensation. Such compaction would contribute to replication fork stalling or unfinished replication (drawn) responsible for genome instability, as well as transcription elongation impairment and silencing (not drawn). Nucleosomes around the DNA-RNA hybrid stretch appear transparent because we do not know, at this point, whether or not they would be there. See also Figure S4.
absence of THOC1 or SETX (Figures 4B and 4C). Moreover, RNase H1 overexpression decreases H3S10P foci formation, supporting the connection between H3S10P and R loops (Figures 4B and 4C). The results suggest that the link between R loops and chromatin condensation marks such as H3S10P is conserved in all eukaryotes, including yeast, nematodes, and humans. DISCUSSION
et al., 2011). We included senataxin-depleted cells to assay whether the R-loop-associated H3S10 phosphorylation occurred regardless of the system we used to enhance R loops. We performed immunofluorescence with anti-H3S10P antibody in HeLa cells grown in ethynyldeoxyuridine (EdU)-containing media to differentiate nonreplicating G1 cells from actively replicating S phase cells. Consistent with the yeast and worm results, G1 HeLa cells with H3S10P foci were only observed in the
The link between different histone modifications and RNA molecules has been shown in eukaryotes. A well-known example is the X-inactivation in mammals by RNA Xist (Lee and Bartolomei, 2013) or gene silencing linked to long noncoding RNAs (ncRNA) that, in most cases, is mediated by siRNAs (Bernstein and Allis, 2005) or via chromatin deacetylation by an antisense ncRNA, as shown in S. cerevisiae (Camblong et al., 2007). Interestingly, RNA-DNA hybrids prevent methylation and transcriptional silencing at CpG island promoters (Ginno et al., 2012) and induce transcriptional silencing, likely through siRNA, in S. pombe (Nakama et al., 2012). Furthermore, centromeres and pericentromeric regions are known to be transcribed, and transcription of pericentromeric DNA repeats is
Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc. 587
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
required for pericentromeric heterochromatin formation in S. pombe (Chan and Wong, 2012). Histone H3S10P and chromatin condensation have never been connected to RNA metabolism, even though H3S10P also controls transcription elongation by releasing the RNA polymerase II (RNAPII) from promoter-proximal pausing sites in Drosophila (Ivaldi et al., 2007). It is believed that H3S10P may modulate the access of different proteins to chromatin, establishing a hierarchy of subsequent events that affect chromatin structure and function (Liokatis et al., 2012). Our study shows that R loops are linked to H3S10P and chromatin compaction. The H3S10P mark is scattered at transcribed genes all over the genome, even though it is better detected at the pericentromeric and centromeric regions. This phenomenon is independent of siRNA, provided that S. cerevisiae does not have siRNA processes. According to our results, R loops can either trigger phosphorylation of H3S10 or inhibit its dephosphorylation. This can be achieved by either enhancement of the recruitment of the responsible kinase or inhibition of the activity of the responsible phosphatase. As more than one enzyme may be involved in this process (Baek, 2011), further extensive work will be needed to decipher the molecular mechanism of this control. Nevertheless, it has been shown that H3S10P levels in yeast depend on the Ipl1 kinase and the Glc7 phosphatase and that Ipl1 and Glc7 interact physically or genetically with factors, such as Set1 and Sen1 (Breitkreutz et al., 2010; Nedea et al., 2008; Zhang et al., 2005), that have been implicated in the metabolism of R loops (Mischo et al., 2011). These data open the possibility of a functional link between R loops and H3S10P that would need to be explored in the future. Although our study does not establish whether R loops help chromatin condensation or vice versa, we favor a model in which R loops trigger the formation of highly condensed chromatin patches that would appear throughout the genome, in particular at centromere and pericentromeric regions due to its higher capacity to nucleate chromatin condensation (Figure 4D). This model is supported by the reduction of H3S10P levels along the yeast genome of hpr1D cells by RNase H1 overexpression. Thus, it is possible that condensation spreads from the R loop. Importantly, such condensed regions would interfere with replication and/or transcription (Figure 4D), causing transcription-associated genome instability and gene silencing. This could contribute to the previously reported effect of R loops on both processes in eukaryotes (Huertas and Aguilera 2003). Interestingly, the observations that common fragile sites (CFSs) can be explained by retarded or incomplete replication at chromatin-condensed regions (Debatisse et al., 2012) and that CFSs can be related to cotranscriptional R loops (Helmrich et al., 2011) open the possibility that local chromatin compaction contributes to the replication fork progression impairment or to a delayed or less-efficient replication initiation as possible factors that cause genome instability. This unexpected link between R loops, histone H3S10P, and chromatin compaction or condensation therefore opens unexpected perspectives with which to understand the mechanism by which RNA controls genome dynamics and function.
EXPERIMENTAL PROCEDURES Yeast Strains and Growth The yeast strains used in this work are listed in Table S1. Cells were arrested in G1 with a factor (0.14 mg/ml) and in S phase with 200 mM hydroxyurea (HU). For meiotic analysis, synchronous meiotic cultures were prepared as previously described (Alani et al., 1990). Nematode Strains and Growth The nematode strain Bristol N2 used as wild-type was provided by the Caenorhabditis Genetics Center. The thoc-2 (tm1310) mutant was previously characterized (Castellano-Pozo et al., 2012). C. elegans strains were grown at 20 C, and standard methods were used for their maintenance and manipulation (Brenner, 1974). Human Cell Culture and Transfection HeLa cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; GIBCO) supplemented with 10% heat-inactivated fetal bovine serum at 37 C (5% CO2). Transient transfection of siRNA (10 nM) was performed using DharmaFECT (Dharmacon) according to the manufacturer’s instructions. Then, 96 hr after transfection, cells were incubated with 20 mM EdU (Click-iT EdU Imaging Kit, Invitrogen) for 20 min to label replicating cells. ChIP and DIP/DRIP Assays For ChIP experiments at the pericentromeric region of CEN6, cells harboring the empty vector pCM189 or the ptetRNH1 plasmid were grown in synthetic complete media without uracil (SC-ura), and samples were treated as described (Duch et al., 2013). Antibodies used are listed in Table S2. For real-time PCR, oligonucleotides for the CEN6 pericentromeric region previously described (Choy et al., 2011) and for other loci (Table S3) were used. Means ± SD of three independent experiments are shown. DRIP experiments were carried out essentially as originally described like DNA IP (DIP) (Mischo et al., 2011). Means ± SD of three independent experiments are shown. ChIP-Chip Assays Cells were grown in YEPD medium and arrested in G1 with a factor (0.14 mg/ml) for 2 hr (G1 phase) and then released in new rich medium with 200 mM HU for another hour (S phase). ChIP-chip was carried out as previously described (Bermejo et al., 2007; Bermejo et al., 2009; Katou et al., 2003; 2006). For details and bioinformatics analysis, see Supplemental Information. Chromatin Accessibility Assays Nuclei preparations from G1-arrested cells with a factor (0.14 mg/ml) were made as previously described (Crotti and Basrai, 2004). Similar quantities of nuclei were incubated with 0, 5, 15, 50, and 150 units/ml of DraI at 37 C for 30 min followed by digestion with 250 U/ml EcoRI. Southern blotting was performed, and filters were probed with a 32P-deoxycytidine triphosphate (dCTP)-labeled CEN3 0.9 kb HindIII-BamHI fragment following standard procedures. Miscellanea Immunofluorescence analyses using the indicated antibodies were performed in 1 day post-L4 C. elegans adult gonads treated as described (CastellanoPozo et al., 2012) and in human cells cultured on glass coverslips and fixed as described previously (Domı´nguez-Sa´nchez et al., 2011). EdU detection was performed using a Click-iT EdU Imaging Kit (Invitrogen) following manufacturer’s instructions. DNA was stained with DAPI. Recombination and plasmid-loss assays and Rad52 foci are described in Go´mez-Gonza´lez et al. (2009) and the Supplemental Information. ACCESSION NUMBERS The Gene Expression Omnibus accession number for the set of ChIP-chip data reported in this paper is GSE46627.
588 Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc.
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
SUPPLEMENTAL INFORMATION Supplemental Information includes Supplemental Experimental Procedures, four figures, and three tables and can be found with this article online at http://dx.doi.org/10.1016/j.molcel.2013.10.006. ACKNOWLEDGMENTS We thank J.C. Reyes and F. Monje-Casas for useful comments on the manuscript and D. Haun for style correction. Research was funded by grants from the Spanish Ministry of Economy and Competitiveness (Consolider Ingenio 2010 CSD2007-0015 and BFU2010-16372), Junta de Andalucı´a (CVI-4567), and the European Union (FEDER). M.C.-P. and J.M.S.-P. were recipients of predoctoral training grants from the Spanish National Institute of Health Carlos III and the Consejo Superior de Investigaciones Cientı´ficas (CSIC), respectively.
Domı´nguez-Sa´nchez, M.S., Barroso, S., Go´mez-Gonza´lez, B., Luna, R., and Aguilera, A. (2011). Genome instability and transcription elongation impairment in human cells depleted of THO/TREX. PLoS Genet. 7, e1002386. Duch, A., Felipe-Abrio, I., Barroso, S., Yaakov, G., Garcı´a-Rubio, M., Aguilera, A., de Nadal, E., and Posas, F. (2013). Coordinated control of replication and transcription by a SAPK protects genomic integrity. Nature 493, 116–119. Gan, W., Guan, Z., Liu, J., Gui, T., Shen, K., Manley, J.L., and Li, X. (2011). R-loop-mediated genomic instability is caused by impairment of replication fork progression. Genes Dev. 25, 2041–2056. Ginno, P.A., Lott, P.L., Christensen, H.C., Korf, I., and Che´din, F. (2012). R-loop formation is a distinctive characteristic of unmethylated human CpG island promoters. Mol. Cell 45, 814–825. Go´mez-Gonza´lez, B., and Aguilera, A. (2009). R-loops do not accumulate in transcription-defective hpr1-101 mutants: implications for the functional role of THO/TREX. Nucleic Acids Res. 37, 4315–4321. Go´mez-Gonza´lez, B., Felipe-Abrio, I., and Aguilera, A. (2009). The S-phase checkpoint is required to respond to R-loops accumulated in THO mutants. Mol. Cell. Biol. 29, 5203–5213.
Received: June 17, 2013 Revised: September 4, 2013 Accepted: October 1, 2013 Published: November 7, 2013
Go´mez-Gonza´lez, B., Garcı´a-Rubio, M., Bermejo, R., Gaillard, H., Shirahige, K., Marı´n, A., Foiani, M., and Aguilera, A. (2011). Genome-wide function of THO/TREX in active genes prevents R-loop-dependent replication obstacles. EMBO J. 30, 3106–3119.
REFERENCES Aguilera, A., and Garcı´a-Muse, T. (2012). R loops: from transcription byproducts to threats to genome stability. Mol. Cell 46, 115–124. Alani, E., Padmore, R., and Kleckner, N. (1990). Analysis of wild-type and rad50 mutants of yeast suggests an intimate relationship between meiotic chromosome synapsis and recombination. Cell 61, 419–436. Baek, S.H. (2011). When signaling kinases meet histones and histone modifiers in the nucleus. Mol. Cell 42, 274–284. Bermejo, R., Doksani, Y., Capra, T., Katou, Y.M., Tanaka, H., Shirahige, K., and Foiani, M. (2007). Top1- and Top2-mediated topological transitions at replication forks ensure fork progression and stability and prevent DNA damage checkpoint activation. Genes Dev. 21, 1921–1936. Bermejo, R., Katou, Y.M., Shirahige, K., and Foiani, M. (2009). ChIP-on-chip analysis of DNA topoisomerases. Methods Mol. Biol. 582, 103–118.
Helmrich, A., Ballarino, M., and Tora, L. (2011). Collisions between replication and transcription complexes cause common fragile site instability at the longest human genes. Mol. Cell 44, 966–977. Hsu, J.Y., Sun, Z.W., Li, X., Reuben, M., Tatchell, K., Bishop, D.K., Grushcow, J.M., Brame, C.J., Caldwell, J.A., Hunt, D.F., et al. (2000). Mitotic phosphorylation of histone H3 is governed by Ipl1/aurora kinase and Glc7/PP1 phosphatase in budding yeast and nematodes. Cell 102, 279–291. Huertas, P., and Aguilera, A. (2003). Cotranscriptionally formed DNA:RNA hybrids mediate transcription elongation impairment and transcription-associated recombination. Mol. Cell 12, 711–721. Ivaldi, M.S., Karam, C.S., and Corces, V.G. (2007). Phosphorylation of histone H3 at Ser10 facilitates RNA polymerase II release from promoter-proximal pausing in Drosophila. Genes Dev. 21, 2818–2831.
Bernstein, E., and Allis, C.D. (2005). RNA meets chromatin. Genes Dev. 19, 1635–1655.
Katou, Y., Kanoh, Y., Bando, M., Noguchi, H., Tanaka, H., Ashikari, T., Sugimoto, K., and Shirahige, K. (2003). S-phase checkpoint proteins Tof1 and Mrc1 form a stable replication-pausing complex. Nature 424, 1078–1083.
Breitkreutz, A., Choi, H., Sharom, J.R., Boucher, L., Neduva, V., Larsen, B., Lin, Z.Y., Breitkreutz, B.J., Stark, C., Liu, G., et al. (2010). A global protein kinase and phosphatase interaction network in yeast. Science 328, 1043– 1046.
Katou, Y., Kaneshiro, K., Aburatani, H., and Shirahige, K. (2006). Genomic approach for the understanding of dynamic aspect of chromosome behavior. Methods Enzymol. 409, 389–410.
Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genetics 77, 71–94. Camblong, J., Iglesias, N., Fickentscher, C., Dieppois, G., and Stutz, F. (2007). Antisense RNA stabilization induces transcriptional gene silencing via histone deacetylation in S. cerevisiae. Cell 131, 706–717. Castellano-Pozo, M., Garcı´a-Muse, T., and Aguilera, A. (2012). R-loops cause replication impairment and genome instability during meiosis. EMBO Rep. 13, 923–929. Cerritelli, S.M., and Crouch, R.J. (2009). Ribonuclease H: the enzymes in eukaryotes. FEBS J. 276, 1494–1505. Chan, F.L., and Wong, L.H. (2012). Transcription in the maintenance of centromere chromatin identity. Nucleic Acids Res. 40, 11178–11188. Choy, J.S., Acun˜a, R., Au, W.C., and Basrai, M.A. (2011). A role for histone H4K16 hypoacetylation in Saccharomyces cerevisiae kinetochore function. Genetics 189, 11–21. Crotti, L.B., and Basrai, M.A. (2004). Functional roles for evolutionarily conserved Spt4p at centromeres and heterochromatin in Saccharomyces cerevisiae. EMBO J. 23, 1804–1814. Debatisse, M., Le Tallec, B., Letessier, A., Dutrillaux, B., and Brison, O. (2012). Common fragile sites: mechanisms of instability revisited. Trends Genet. 28, 22–32.
Lee, J.T., and Bartolomei, M.S. (2013). X-inactivation, imprinting, and long noncoding RNAs in health and disease. Cell 152, 1308–1323. Li, X., and Manley, J.L. (2005). Inactivation of the SR protein splicing factor ASF/SF2 results in genomic instability. Cell 122, 365–378. Liokatis, S., Stu¨tzer, A., Elsa¨sser, S.J., Theillet, F.X., Klingberg, R., van Rossum, B., Schwarzer, D., Allis, C.D., Fischle, W., and Selenko, P. (2012). Phosphorylation of histone H3 Ser10 establishes a hierarchy for subsequent intramolecular modification events. Nat. Struct. Mol. Biol. 19, 819–823. Mischo, H.E., Go´mez-Gonza´lez, B., Grzechnik, P., Rondo´n, A.G., Wei, W., Steinmetz, L., Aguilera, A., and Proudfoot, N.J. (2011). Yeast Sen1 helicase protects the genome from transcription-associated instability. Mol. Cell 41, 21–32. Mun˜oz-Galva´n, S., Jimeno, S., Rothstein, R., and Aguilera, A. (2013). Histone H3K56 acetylation, Rad52, and non-DNA repair factors control double-strand break repair choice with the sister chromatid. PLoS Genet. 9, e1003237. Nakama, M., Kawakami, K., Kajitani, T., Urano, T., and Murakami, Y. (2012). DNA-RNA hybrid formation mediates RNAi-directed heterochromatin formation. Genes Cells 17, 218–233. Nedea, E., Nalbant, D., Xia, D., Theoharis, N.T., Suter, B., Richardson, C.J., Tatchell, K., Kislinger, T., Greenblatt, J.F., and Nagy, P.L. (2008). The Glc7 phosphatase subunit of the cleavage and polyadenylation factor is essential for transcription termination on snoRNA genes. Mol. Cell 29, 577–587.
Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc. 589
Molecular Cell R Loops Are Linked to the H3S10P Condensation Mark
Paulsen, R.D., Soni, D.V., Wollman, R., Hahn, A.T., Yee, M.C., Guan, A., Hesley, J.A., Miller, S.C., Cromwell, E.F., Solow-Cordero, D.E., et al. (2009). A genome-wide siRNA screen reveals diverse cellular processes and pathways that mediate genome stability. Mol. Cell 35, 228–239. Reddy, K.C., and Villeneuve, A.M. (2004). C. elegans HIM-17 links chromatin modification and competence for initiation of meiotic recombination. Cell 118, 439–452.
Wahba, L., Amon, J.D., Koshland, D., and Vuica-Ross, M. (2011). RNase H and multiple RNA biogenesis factors cooperate to prevent RNA:DNA hybrids from generating genome instability. Mol. Cell 44, 978–988. Wei, Y., Yu, L., Bowen, J., Gorovsky, M.A., and Allis, C.D. (1999). Phosphorylation of histone H3 is required for proper chromosome condensation and segregation. Cell 97, 99–109.
Skourti-Stathaki, K., Proudfoot, N.J., and Gromak, N. (2011). Human senataxin resolves RNA/DNA hybrids formed at transcriptional pause sites to promote Xrn2-dependent termination. Mol. Cell 42, 794–805.
Wellinger, R.E., Prado, F., and Aguilera, A. (2006). Replication fork progression is impaired by transcription in hyperrecombinant yeast cells lacking a functional THO complex. Mol. Cell. Biol. 26, 3327–3334.
Stirling, P.C., Chan, Y.A., Minaker, S.W., Aristizabal, M.J., Barrett, I., Sipahimalani, P., Kobor, M.S., and Hieter, P. (2012). R-loop-mediated genome instability in mRNA cleavage and polyadenylation mutants. Genes Dev. 26, 163–175.
Zhang, K., Lin, W., Latham, J.A., Riefler, G.M., Schumacher, J.M., Chan, C., Tatchell, K., Hawke, D.H., Kobayashi, R., and Dent, S.Y. (2005). The Set1 methyltransferase opposes Ipl1 aurora kinase functions in chromosome segregation. Cell 122, 723–734.
590 Molecular Cell 52, 583–590, November 21, 2013 ª2013 Elsevier Inc.
