Distribution and dynamics of chromatin modification induced by a defined DNA double-strand break

Abstract

Background: In response to DNA double-strand breaks (DSBs), eukaryotic cells rapidly phosphorylate histone H2A isoform H2AX at a C-terminal serine (to form gamma-H2AX) and accumulate repair proteins at or near DSBs. To date, these events have been defined primarily at the resolution of light microscopes, and the relationship between gamma-H2AX formation and repair protein recruitment remains to be defined.

Results: We report here the first molecular-level characterization of regional chromatin changes that accompany a DSB formed by the HO endonuclease in Saccharomyces cerevisiae. Break induction provoked rapid gamma-H2AX formation and equally rapid recruitment of the Mre11 repair protein. gamma-H2AX formation was efficiently promoted by both Tel1p and Mec1p, the yeast ATM and ATR homologs; in G1-arrested cells, most gamma-H2AX formation was dependent on Tel1 and Mre11. gamma-H2AX formed in a large (ca. 50 kb) region surrounding the DSB. Remarkably, very little gamma-H2AX could be detected in chromatin within 1-2 kb of the break. In contrast, this region contains almost all the Mre11p and other repair proteins that bind as a result of the break.

Conclusions: Both Mec1p and Tel1p can respond to a DSB, with distinct roles for these checkpoint kinases at different phases of the cell cycle. Part of this response involves histone phosphorylation over large chromosomal domains; however, the distinct distributions of gamma-H2AX and repair proteins near DSBs indicate that localization of repair proteins to breaks is not likely to be the main function of this histone modification.

Figure 1. Single-Strand DNA Formation, γ-H2AX Formation, and Mre11p/Rad51p Binding in Response to a DSB at the MAT Locus

HO cutting at MAT was induced either in asynchronously growing cells (“cycling,” H1069) or in cells arrested by exposure to α-mating pheromone (“G1,” H1072), and samples were collected at the appropriate time points (see Experimental Procedures). (A) Map of the region immediately centromere-distal to the MAT DSB site. The map shows the method used to detect DSB formation and 5′-to-3′ resection. DNA was digested with SspI and separated on alkaline agarose gels, and gel blots were hybridized with a single-strand probe specific to the unresected strand (ss probe). HO-cut and uncut chromosomes produce 0.9 kb and 1.2 kb fragments, respectively; 5′-to-3′ resection past SspI sites eliminates cutting at these sites and thus produces larger SspI fragments detected by the probe. (B) Southern blot illustrating this analysis. Bands labeled r1 through r6 are the products of resection through SspI sites 0.9, 1.6, 3.5, 4.7, 5.9, and 6.5 kb from the HO-cut site; * denotes a cross-hybridizing band that was used as a loading control. (C) Quantitative analysis of the blot in (B) and of one replicate. (Filled squares, filled circles) Percent of chromosomes with DSBs; (open squares, open circles) percent of DSB-containing chromosomes that have had at least 0.9 kb resected. (filled squares, open squares) Data from cycling cells; (filled circles, open circles) data from G1-arrested cultures. (D) Examples of multiplex PCR products of ChIP reactions used to determine relative amounts of γ-H2AX formed and of Mre11p bound near the MAT DSB. All reactions contained primer pairs for a sequence 66 kb from the break (“con”) and experimental sequences (“dsb”) 5.1 kb to the left for γ-H2AX and 0.02 kb to the right for Mre11p. (+) ChIP prepared with the indicated primary antibody; (−) treated identically but without primary antibody; (DNA) PCR reactions for which genomic DNA was used. (E) Timing of γ-H2AX formation (filled squares, open squares) and Mre11p binding (filled circles, open circles) in cycling (filled symbols) and G1-arrested (open symbols) cultures. Data are from panels in (D) and 3–4 replicate experiments. For normalizing DSB-specific relative ChIP values (dsb/control band intensity ratios, see Supplemental Experimental Procedures), the maximum value obtained for each time course was set to unity. Symbols and error bars report the average and standard deviation for these normalized values.

Figure 2. γ-H2AX Formation and Repair Protein Binding Induced by the MAT DSB

(A) Distribution of γ-H2AX in cycling cells (H1069). DSB-specific relative ChIP values (band intensity ratios, indicated experimental locus/control locus 66 kb from the DSB; see Supplemental Experimental Procedures) are for samples taken 15 min (dotted line), 30 min (dashed line), and 60 min (solid line) after HO induction. (B) Distribution of γ-H2AX in G1-arrested cells (H1072). DSB-specific relative ChIP values are for samples taken 30 min (dashed line) and 60 min (solid line) after HO induction, as in (A). (C) Distribution of Mre11p (this work) and Rad51p (reproduced from Sugawara et al. [18] for comparative purposes) bound in the region around the MAT DSB. DSB-specific relative ChIP values for Mre11p (left-hand Y axis; filled circles, cycling cells; open circles, G1-arrested cells) and for Rad51p (closed triangles, right-hand Y axis) are from samples taken 60 min after HO induction.

Figure 3. DSB Formation is Not Accompanied by Histone Depletion

Relative histone H2B ChIP values (band intensity ratios, indicated experimental locus/ control locus 66 kb from the DSB; see Supplemental Experimental Procedures) were measured in samples taken from cycling cells (H1069) before (0 min, open circles) or 60 min after (closed circles) HO induction; a strain in which about half of all histones H2B are tagged with a single HA epitope (H1069) was used. For comparative purposes, relative γ-H2AX ChIP levels in a 60 min sample are reproduced from Figure 2A. (Inset) Representative multiplex PCR reactions with H2B ChIP samples: +, ChIP samples; −, primary antibody omitted; in, 100-fold dilution of input lysate. DSB-specific PCR primers were for sequences 0.02 kb from the DSB.

Figure 4. Checkpoint Proteins Required for γ-H2AX Formation

(A) Schematic diagram of functional relationships among DNA damage response kinases. PI3KKs and their binding partners are shown in the top line; target downstream kinases are shown in the second. Arrows indicate the direction in which activating phosphorylation signals can be transmitted. (B) Both Mec1 and Tel1 kinases can promote γ-H2AX formation in cycling cells. Panels contain products of multiplex PCR reactions on samples taken 60 min after galactose addition; primer pairs 66 (con) and 5.1 (dsb) kb from the break were used. (DNA) PCR with purified genomic DNA. Values below each panel are the dsb/control band intensity ratio. All strains are isogenic to JKM179 (wild-type) with the indicated additional mutations (details in supplementary Table S1). (C) Most γ-H2AX formation in G1-arrested cells is TEL1 and MRE11 dependent. Strains isogenic to yXW1 (wild-type) were arrested in G1, and HO was induced as described in the Experimental Procedures. All other details are as in panel (B). (D) Mre11p binding is TEL1 and MEC1 independent. Mre11-ChIP reactions from G1-arrested cells were analyzed as described above with primer pairs 66 (con) and 0.02 (dsb) kb from the break.