The cellular response to DNA damage involves extensive interaction with and manipulation of chromatin. This includes the detection and repair of the DNA lesion, but there are also transcriptional responses to DNA damage, involving the up- or down-regulation of numerous genes. Therefore changes to chromatin structure, including covalent modification of histone proteins, are known to occur during DNA-damage responses. One of the most well characterized DNA-damage-responsive chromatin modification events is the phosphorylation of the SQ motif found in the C-terminal tail of histone H2A or the H2AX variant in higher eukaryotes. In the budding yeast, a number of additional residues in this region of histone H2A that contribute to the cellular response to DNA damage have been identified, providing an insight into the nature and complexity of the DNA-damage histone code.
- DNA damage
- DNA repair
- H2AX variant
- histone H2A
Chromatin is the term for DNA complexed with proteins that exists in the eukaryotic nucleus. The basic unit of chromatin is the nucleosome, made up of DNA wound around an octamer of four core histones (H2A, H2B, H3 and H4). The structure of chromatin and its behaviour can be altered by extensive covalent modification of histones such as acetylation, phosphorylation, methylation and ubiquitination. In addition, the incorporation of histone variants can create variability in chromatin.
When DNA is damaged, the cell must detect the DNA lesion and repair it within the context of chromatin. As with transcription and DNA replication, it is very clear that chromatin plays a significant and dynamic role during the process. It is also apparent that there is a great deal of complexity in how chromatin, and more specifically, modifications of the four core histones, contribute to this event. Here, we focus on a single region of histone H2A and its role in the cellular response to DNA damage.
The H2A C-terminal tail
Histone H2A is unique among the four core histones in that it has a significant C-terminal tail, which is flexible and protrudes from the nucleosome core particle . There are also more described variants of H2A than any of the other histones , and these vary significantly in the C-terminal region of the protein, suggesting that this region can provide multiple different functions.
One family of H2A variants is termed H2AX, and is characterized by an extended C-terminal tail containing an SQ motif (Figure 1a). This variant is not present in lower eukaryotes, but, interestingly, the SQ motif is found in the main H2A protein of many species in the same position relative to the stop codon (Figure 1). This motif is phosphorylated in response to DNA damage by the PIKKs [PI3K (phosphoinositide 3-kinase)-like kinases; Mec1 and Tel1 in the budding yeast]. It is currently the most extensively studied histone modification in DNA-damage responses and has been extensively reviewed elsewhere [3–6]. Interestingly, several reports have identified other residues in the H2A C-terminal tail of the budding yeast, which, in addition to Ser-129, contribute to cellular survival after DNA damage [7–10]. These additional residues in the H2A tail involved in DNA-damage responses will be the focus of the present review.
Phenotypes of H2A mutant strains
Several groups hypothesized that there may be other residues in histone H2A that contribute to the cellular response to DNA damage. Moreover, in addition to Ser-129, multiple residues in the C-terminal tail have the potential to be covalently modified (Figure 1), raising the possibility that other DNA-damage-dependent covalent modifications may exist. In order to investigate these possibilities, strains with mutations in the H2A tail were created and the phenotypes of the resulting strains were examined (summarized in Table 1).
Intriguingly, the four groups did not uncover identical phenotypes in these assays. One study found that H2A Thr-126 is critical for survival in the presence of phleomycin, and T126A mutant strains were also found to have a defect in an NHEJ (non-homologous end-joining) DNA DSB (double-strand break) repair assay . This strain was also found to have an Spt− phenotype as well as defects in telomeric silencing. In contrast, the H2A S129A mutant strain did not show these phenotypes in this study, and interestingly, a T126A/S129A mutant strain behaved as an S129A mutant strain , suggesting that there is some cross-talk between these residues.
However, two other studies found that a T126A mutant strain showed no significant hypersensitivity to either phleomycin or bleomycin or any defect in NHEJ assays [8,9]. Nor was the strain hypersensitive to a wide range of other DNA-damaging agents [8–10], with the possible exception of a weak decrease in survival after exposure to MMS (methyl methanesulfonate) . Moreover, we found that the hypersensitivity to MMS that we identified in our S129A mutant strain was not altered in an S129A/T126A mutant strain .
Although the reason for this phenotypic variability is not known, one possibility is that there may be genetic differences in the strains used which are significant, and that Thr-126 only makes an appreciable contribution to mediating cellular survival after DNA damage under certain conditions. It will clearly be very important to work out why a H2A T126A mutation can show such variable phenotypes in different studies.
In addition to T126A phenotypes, there were a few differences between the conclusions reached by the different groups. For example, a weak hypersensitivity after UV irradiation was seen in one study , while we detected no difference from wt (wild-type) in a H2A K120A mutant strain . In addition, H2A S129A mutant strains were found to have defects in NHEJ assays in two studies [8,9], but not in a third .
When taken together, however, a number of conclusions about the contribution of individual amino acid residues to DNA-damage responses emerge from the studies. First, the lysine residues in the H2A tail do not individually play a very dramatic role in mediating cellular survival in the presence of DNA damage [8–10], although, notably, Lys-120, Lys-121, Lys-124 and Lys-127 all appear to contribute to NHEJ activity . Interestingly, while a K127A mutant strain has a relatively pronounced end-joining defect, a K127Q mutant strain is not significantly different from wt, suggesting that Lys-127 is acetylated and that this modification contributes to NHEJ . Combining K127A with either S122A or S129A did not increase the severity of the defect of either single mutant strain, yet the S122A/S129A mutant was more defective in end-joining than either S122A or S129A, making it difficult to determine precisely how Lys-127 is contributing to NHEJ activity (discussed in more detail below).
Secondly, Ser-129 clearly plays an important role, but this is specific to a subset of DNA-damaging agents. In particular, significant hypersensitivity is consistently seen in the presence of MMS, suggesting a potential role in dealing with replicative stress.
Thirdly, and most compellingly, H2A Ser-122 contributes to cellular survival in response to a wide range of DNA-damaging agents [8–10]. This suggests a very central role in responding to DNA damage for H2A Ser-122.
The interplay between Ser-122 and Ser-129 was also examined in two of the studies, and these residues were found to provide separate functions in mediating viability in the presence of damage [8,9]. Indeed, a strain containing an S122A mutation can be ‘rescued’ by introducing a second H2A-encoding gene with an S129A mutation , indicating that the two residues contribute independently and that not every H2A molecule needs to have both serine residues in order to function at wt levels.
Phosphorylation of the H2A tail
As mentioned above, H2A Ser-129 is known to be phosphorylated in response to DNA damage in vivo by the DNA-damage-dependent kinases Mec1 and Tel1 . Importantly, work from Lustig and colleagues demonstrated that all three of the phosphorylatable residues (Ser-122, Thr-126 and Ser-129) are phosphorylated in vivo .
Using antibodies against the specific phosphorylated motifs, it was found that phosphorylation at all three residues is modulated in response to DNA damage [9,11,12]. H2A Ser-129 phosphorylation levels clearly increase to varying degrees in response to a variety of DNA-damaging agents [9,11,12]. In contrast, Ser-122 and Thr-126 seem to show a more complex phosphorylation pattern. Phosphorylation at both sites appears to increase after treatment with phleomycin, MMS and menadione, but not after UV irradiation where the signal may in fact be even lower than in untreated cells .
Finally, mutation of Ser-122 results in a significant increase in H2A phosphorylation elsewhere , although the physiological consequence of this is not known. In contrast, an increase in H2A phosphorylation levels is not seen when either Thr-126 or Ser-129 is mutated .
How do these residues contribute to DNA-damage responses?
Survival after DNA damage depends on multiple events. First, the cell must recognize the presence of the damage and signal this to the cell-cycle machinery, resulting in a DNA-damage-induced checkpoint . The signal is also transduced to the transcriptional machinery, and genes involved in the DNA-damage response are up-regulated . DNA repair activity is clearly critical for survival, but, additionally, strains lacking the ability to re-enter the cell cycle, even in the presence of persistent damage, termed adaptation, can show decreased survival after exposure to DNA-damaging agents .
Histone modifications could function either directly at the lesion, aiding the detection and/or repair of damaged DNA, or could contribute to full levels of DNA-damage-induced transcriptional up-regulation. Moreover, indirect effects from elsewhere in the genome, such as the telomeres, might affect survival after DNA damage. For example, NHEJ proteins are enriched at telomeres and are mobilized after DNA damage [16–18]. Histone modifications could play a role in allowing the factors to be appropriately released from the telomeric chromatin.
H2A Ser-129 has been shown to function directly at the site of damage (for a review, see ). One very straightforward explanation for the independent contributions of Ser-122 and Ser-129 is that they function at different locations in the genome, and a role for Ser-122 in transcription is one possibility. Notably, this possibility has not yet been thoroughly examined, but Ser-122 was found to contribute to transcriptional regulation of the CUP1 gene .
Regardless of which step(s) of DNA-damage responses are being affected by the H2A tail residues, the immediate consequence of modification of these residues will be to either create novel binding platforms for proteins to be recruited to chromatin or to affect higher-order chromatin structure.
H2A modification and novel binding platforms
Because of their proximity to each other, different combinations of H2A tail modifications may have profound consequences on their binding partners. However, the genetic data suggest that Ser-122 and Ser-129 modifications do not have to be on the same molecule to function, and thus are unlikely to work together to provide a binding site for a protein involved in DNA-damage responses.
Additionally, Thr-126, which has been shown to be phosphorylated in vivo, does not appear to affect survival either alone or in combination with S122A or S129A mutations, at least in some strain backgrounds [8–10]. This suggests that, if Thr-126 does function in combination with these sites to provide a binding platform, it is not uniformly critical for survival after DNA damage.
Finally, Lys-127 is likely to be acetylated in vivo and mediate NHEJ activity . Because Lys-127 appears to function on the same pathways as both Ser-122 and Ser-129 , it is tempting to speculate that Ser-122 and Ser-129 do indeed provide functions at physically distinct regions of the genome, which would allow Lys-127 to then function independently in combination with each of them.
Clearly, it will be important to work out, in the first instance, where in the genome each of these modifications occurs and under what circumstances, and secondly, whether they co-exist on the same molecule. In this regard, it is worth noting that Ser-129 phosphorylation has not been detected by ChIP (chromatin immunoprecipitation) on nucleosomes immediately flanking a DNA DSB [20,21]. In addition, we find that the histone acetyltransferase NuA4 can specifically associate with phosphorylated H2A in vitro, and yet NuA4 accumulates in vivo at the regions flanking a DNA break where we do not detect H2A Ser-129 phosphorylation . One highly speculative possibility is that NuA4 may acetylate Lys-127 on these nucleosomes and disrupt the ability of the antibody to detect phosphorylated Ser-129.
The H2A tail and higher-order chromatin structure
It is well established that in addition to creating novel binding sites, modification of histones can impinge on the structure of chromatin fibres. This is particularly true of acetylation, which reduces the positive charge of the lysine residues and can therefore dramatically alter the ability of the histone tails to interact with negatively charged DNA.
In thinking about the potential affects of modification of the H2A C-terminal tail, it is crucial to consider the location of the tail in the nucleosome structure. The tail exits in the nucleosome close to the point where DNA enters and exits the structure (Figure 2a). In vitro, the flexible tail is in contact with DNA within the nucleosome . By adding multiple negatively charged phosphate groups to the tail, the affinity of the tail for negatively charged DNA is likely to be altered. Moreover, this is the region bound by linker histones as well as some HMG (high-mobility group) proteins , which could be influenced by phosphorylation of the H2A tail.
Folding of DNA into nucleosomes represents the first level of chromatin compaction. Recently, insights into the nature of the next level of packing have been generated, showing that nucleosome arrays fold into two-start helices [24,25]. Notably, the C-terminal tail of H2A is in the centre of this folded structure (Figure 2b). This has two significant implications. First, phosphorylation of the histone H2A C-terminal tail may be able to influence the nature of these structures. Extensive phosphorylation may even disrupt or impair the ability of these structures to form. Secondly, this raises questions about the ability of the relevant kinases to access their target sites in the first instance. Because DNA damage can occur anywhere in the genome, if phosphorylation occurs at the site of damage, then there must be a mechanism allowing the DNA-damage-dependent kinases to access the tails when they are buried in very compact higher-order structures. It is possible that the kinases or proteins that recruit the kinases to chromatin are capable of disrupting higher-order structure. Alternatively, the activity of ATP-dependent chromatin remodelling complexes may be required. Consistent with the latter possibility, we and others have found that the RSC chromatin remodelling complex in yeast is required for efficient H2A phosphorylation [26,27].
Beyond the budding yeast
We know that the role of budding yeast H2A Ser-129 is conserved and the analogous residue is phosphorylated in response to DNA damage in a wide range of organisms [3–6]. While Thr-126 is phosphorylated in vivo, its contribution to DNA-damage responses is variable among different studies, and may only be detectable under certain conditions. Moreover, while there is a phosphorylatable residue in some H2AX variants (Figure 1a), this is not universally conserved, and it is clearly not present in other organisms in the core H2A (Figure 1b).
The lysine residues in the H2A tail were not shown to play a significant role in mediating survival after DNA damage, but K120A, K121A, K124A and K127A mutant strains all show some degree of reduced NHEJ activity . Interestingly, Lys-120 and Lys-121 are very highly conserved (Figure 1), raising the possibility that they may provide similar functions in other organisms.
We and others found that H2A Ser-122 contributes to the ability of cells to survive in the presence of many different types of DNA damage [8–10] and has a more pronounced hypersensitivity relative to Ser-129, suggesting that this residue is playing a more central role in DNA-damage responses. Moreover, Ser-122 is in a highly conserved region of H2A (Figure 1), and the position is almost always occupied by a phosphorylatable residue. This strongly suggests that the role of Ser-122 in DNA-damage responses may be conserved among eukaryotic organisms.
Interestingly, in this regard, the analogous residue in Drosophila H2A (Thr-119) has been shown to be phosphorylated in vivo and this occurs mainly during mitosis . A kinase able to phosphorylate this residue was purified and termed NHK-1 (nucleosomal histone kinase 1). It will be interesting to determine whether H2A Thr-119 or NHK-1 contributes to the cellular response to DNA damage in Drosophila.
Clearly, multiple residues in the H2A C-terminal tail contribute to DNA-damage responses in the budding yeast and there is evidence that at least some of these are covalently modified. Many residues elsewhere in the nucleosome are also involved (e.g. [10,20,29]). Not only is there likely to be cross-talk and interplay between these residues and modifications, but also their roles will vary depending on the phase of the cell cycle as well as the repair pathway utilized. Further studies systematically examining histone residues and their modification during DNA-damage responses will be required in order to build up a base of knowledge on which to develop an understanding of the DNA-damage histone code.
British Yeast Group Meeting 2007: Independent Meeting held at the Paramount Palace Hotel, Buxton, U.K., 26–28 March 2007. Organized and Edited by A. Goldman (Sheffield, U.K.).
Abbreviations: DSB, double-strand break; HMG, high-mobility group; MMS, methyl methanesulfonate; NHEJ, non-homologous end-joining; NHK-1, nucleosomal histone kinase 1; PI3K, phosphoinositide 3-kinase; PIKK, PI3K-like kinase; wt, wild-type
- © The Authors Journal compilation © 2007 Biochemical Society