Histone acetylation, one of the most important and highly studied histone modifications, is the covalent addition of an acetyl group to the lysine residue within histones. It is mediated via histone acetyltransferases (HATs), which are divided into three families: GNAT (GCN5 related N-acetyltransferase), MYST (MOZ, YBF2/SAS3, SAS2, TIP60), and P300/CBP (Hodawadekar and Marmorstein, 2007). Histone deacetylation, on the other hand, is catalyzed by histone deacetylases (HDACs). Based on function and sequence homology to yeast original proteins, the 18 HDAC enzymes in humans are classified into four classes (Seto and Yoshida, 2014). The class I HDACs, homologous to yeast Rpd3 (reduced potassium dependency 3), include HDAC1, HDAC2, HDAC3, and HDAC8. The class II HDACs comprise HDAC4, HDAC5, HDAC6, HDAC7, HDAC9, and HDAC10 and share homology with yeast Hda1 (histone deacetylase 1). SIRT1, SIRT2, SIRT3, SIRT4, SIRT5, SIRT6, and SIRT7 belong to the class III and have homology to yeast Sir2 (silent information regulator 2), an NAD⁺-dependent enzyme. HDAC11 is assigned to the class IV as it is homologous to neither Rpd3 nor Hda1 yeast enzymes. Histone acetylation generally promotes open chromatin and gene activation because the addition of acetyl groups neutralizes the positive charge of histone lysine residues and thus reduces the strong electrostatic histone-DNA interaction, leading to a more open chromatin structure favorable for the transcriptional machinery. Histone acetylation can also be “read” by bromodomain-containing proteins which recruit positive elongation factors (Roth et al., 2001).

Seminal studies in the 1980s showed that histone hyperacetylation induced by sodium butyrate, an inhibitor of histone HDAC, stimulates NER in UV treated human fibroblasts (Smerdon et al., 1982; Ramanathan and Smerdon, 1989) and that a hyperacetylation phase, followed by a hypoacetylation phase, occurred immediately after UV irradiation (Ramanathan and Smerdon, 1986). Yeast Gcn5, a subunit of the SAGA (Spt-Ada-Gcn5 Acetyltransferase) transcriptional coactivator complex, was then identified to be responsible for H3 (K9 and K14) acetylation and global genomic repair at certain loci but not the whole genome (Teng et al., 2002; Yu et al., 2005). It was later found that Rad16, together with Rad7, mediate the UV-dependent H3 (K9 and K14) acetylation by increasing the occupancy of Gcn5 in chromatin after UV treatment (Teng et al., 2008; Yu et al., 2011). Intriguingly, histone variant Htz1 (H2A.Z in humans) promotes H3 acetylation for efficient NER by using the same way as Rad16 (Yu et al., 2013). Like H3 acetylation, histone H4 acetylation also increases in response to UV irradiation and NuA4 (nucleosome acetyltransferase of histone H4) HAT is important for efficient NER in yeast (Irizar et al., 2010; Hodges et al., 2019). Human GCN5, homologous to the yeast Gcn5, was also found to be involved in H3 acetylation and NER in UV-damaged chromatin by physically interacting with the transcription factor E2F1 (Guo et al., 2010; Guo et al., 2011). Another two transcription coactivators in mammalian cells, CBP (CREB-binding protein) and p300, were shown to participate in NER through their HAT activities and physical interactions with repair proteins such as p53, DDB2, CSB and PCNA (Datta et al., 2001; Hasan et al., 2001; Rubbi and Milner, 2003). H3K56 acetylation, catalyzed by P300/CBP in mammals and by Rtt109 (homolog of P300/CBP) in yeast, plays an important role in genome stability (Driscoll et al., 2007; Han et al., 2007). However, it is dispensable for NER and responsible for the restoration of chromatin structure after completion of NER (Battu et al., 2011). Similarly, H3K14 acetylation, on its own, was found to have no effect on the repair of UV damaged DNA in an elegant in vitro study. However, it could facilitate UV damage repair in the presence of RSC (Remodeling the Structure of Chromatin), a chromatin remodeler, by enhancing the interaction between RSC and nucleosome (Duan and Smerdon, 2014). Besides acetylating histones, P300/CBP also interacts and acetylates nonhistone NER factor XPG in a PCNA-p21 dependent manner, which makes the 3′ incision on the damaged DNA strand during dual incision (Tillhon et al., 2012). Thus, P300/CBP may promote NER by acetylating both histones and the essential NER factor XPG. Recently, it was reported that GCN5/PCAF (P300/CBP-associated factor) mediated acetylation of RPA1 (replication protein A1) and acetylation of XPF by TIP60 promote NER (He et al., 2017; Zhao et al., 2017; Wang et al., 2020). Like HATs, HDACs can deacetylate both histones and nonhistone NER factors. For example, HDAC3 deacetylates H3K14 after UV irradiation and this deacetylation of H3K14 facilitates the recruitment of XPC during NER (Kakumu et al., 2017; Nishimoto et al., 2020). Interestingly, HDAC4 itself was recently discovered to bind XPC directly for efficient NER (Li et al., 2020). A previous study has shown that XPA, a rate limiting NER factor, is deacetylated by SIRT1 and this deacetylation is required for optimal NER (Fan and Luo, 2010). However, it was later found that only a small fraction of XPA is acetylated and downregulation of SIRT1 has no measurable effect on overall rate of NER. Instead, XPA and hence NER is regulated by circadian clock at the transcriptional level (Kang et al., 2010; Kang et al., 2011).

With the development of NGS-based methods for genome-wide mapping of DNA damage and repair (Sloan et al., 2018; Li and Sancar, 2020), our understanding of DNA damage formation and repair in diverse chromatin environments has increased significantly. With the aid of XR-seq (Hu et al., 2015; Hu et al., 2019), which purely measures ongoing repair by isolating and sequencing the excision products released during NER, a genome-wide comparison of chromatin states, histone modifications, gene expression and UV damage repair kinetics was performed (Adar et al., 2016). It was found that the earliest repair occurs in active and open chromatin regions, while repair in repressed heterochromatin regions is relatively slow. Regions with H3K27 acetylation, which is associated with active promoters and enhancers, have high levels of early repair. Indeed, UV damage repair super hotspots, which are defined as the earliest-repair sites in the genome, are significantly enriched in both frequently interacting regions (FIREs) and super enhancers (Jiang et al., 2021).

Histone methylation, catalyzed by histone methyltransferases (HMTs), is the process of adding one, two, or three methyl groups to lysine and arginine residues in histone proteins. Methyl groups can also be removed from histone residues by histone demethylases. Depending on which residues are methylated and how many methyl groups are added, histone methylation can either increase or decrease gene transcription activity (Greer and Shi, 2012). The most extensively studied methylation of H3K9, H3K27 and H4K20 is associated with repressed transcription, whereas methylation of H3K4, H3K36, and H3K79 is linked to active transcription. Like histone acetylation, methylation events, which can weaken electrostatic attractions between histone and DNA, lead to unwinding of the DNA followed by recruitment of transcriptional machinery, and thus increase transcription activity. Some histone methylations, however, trigger chromatin compaction and inhibit the access of transcriptional machinery to DNA. For example, H3K9 methylation can recruit HP1 (heterochromatin protein 1) and bind to the highly conserved chromodomain of HP1, resulting in chromatin compaction and gene silencing (Jacobs and Khorasanizadeh, 2002). In Drosophila melanogaster, a global decrease of H3K9 trimethylation after UV irradiation was discovered in salivary gland cells (Palomera-Sanchez et al., 2010).

In contrast to histone acetylation, which has been known to function in NER for a long time, the involvement of histone methylation in NER was only recently revealed, and one histone methylation, H3K79, was found to function in NER (Bostelman et al., 2007; Chaudhuri et al., 2009; Tatum and Li, 2011; Zhu et al., 2018). In yeast, the H3K79 methylation is catalyzed by Dot1 with its HMT activity. Both Dot1 and H3K79 methylation are required for efficient global genomic repair but not for TCR (Tatum and Li, 2011). It was found that the role of H3K79 methylation in global genomic repair was unlikely achieved through activating cell cycle checkpoints or regulating the expression of NER genes. Rather, it may serve as a docking site for the recruitment of the NER machinery required for global genomic repair (Gsell et al., 2020). It was later found that DOT1L mediated H3K79 methylation is indeed essential for XPC recruitment and efficient NER after UV irradiation in mammalian cells (Zhu et al., 2018). Nevertheless, another study in mouse embryonic fibroblasts suggested that DOT1L is not implicated in global genomic repair, and that it mainly facilitates the reactivation of RNAP II transcription initiation by securing an open chromatin structure (Oksenych et al., 2013). The difference may be explained by the fact that global genomic repair in rodents is significantly slower than that in humans (van der Horst et al., 1997) and that embryonic stem cells primarily eliminate cells containing massively damaged DNA through apoptosis rather than NER (Li et al., 2019). Besides its role in NER, DOT1L is also critical for transcription elongation, DNA damage response, normal development, and formation of heterochromatin (Wysocki et al., 2005; Wood et al., 2018; Ljungman et al., 2019).

H3K4 trimethylation (H3K4me3) and H3K4 monomethylation (H3K4me1) are associated with actively transcribed promoters and active enhancers, respectively. It was found that DDB2, a DNA damage-binding protein, interacts with and recruits the histone methyltransferase ASH1L (Absent, Small or Homeotic discs 1) to UV lesions leading to deposition of H3K4me3. Similar to H3K79 methylation, H3K4me3 facilitates the stable docking of XPC to DNA damage sites enabling the recruitment of downstream NER proteins by XPC (Balbo Pogliano et al., 2017). Interestingly, another study in yeast showed that H4H75E mutation decreases global genomic repair by impairing the recruitment of Rad4 (XPC in humans) to chromatin after UV irradiation (Selvam et al., 2019). As DDB2 preferentially binds to unmethylated nucleosomes, it is likely that H3K4me3 may promote the DDB2-XPC handover at DNA damage sites (Apelt et al., 2021). Like H3K27 acetylation, excision repair in regions marked by H3K4me3 and H3K4me1 occurs much earlier than those regions marked by repressive histone methylations (Adar et al., 2016). H3K4 methylation was also found to promote the repair of DNA double-strand breaks by the non-homologous end joining pathway (Wei et al., 2018). Whether and how H3K79 methylation and H3K4me3 are coordinated during the NER process needs to be elucidated in future studies.

Ubiquitylation is the addition of ubiquitin, which consists of 76 amino acids and exists in all eukaryotes, to a substrate protein. This process is through a reversible three-step enzymatic reaction requiring ubiquitin-activating enzymes (E1), ubiquitin-conjugating enzymes (E2), and ubiquitin ligases (E3). During the ubiquitylation, one single ubiquitin (monoubiquitylation) or a chain of ubiquitin (polyubiquitylation) can be added to the substrate protein. For polyubiquitylation, the seven lysine residues (K6, K11, K27, K29, K33, K48, K63) and N-terminal methionine (M1) in ubiquitin serve as linkage points of the ubiquitin chain (Komander and Rape, 2012). It is associated with a wide range of cellular processes such as protein degradation, DNA replication, gene transcription and DNA repair (Welchman et al., 2005). Nevertheless, only polyubiquitylation, mostly K48- and K29-linked polyubiquitylation, marks the substrate protein for degradation by the proteasome. Other polyubiquitylations and monoubiquitylations may regulate other cellular processes including NER (Miranda and Sorkin, 2007; Chauhan et al., 2021). Histone ubiquitylation occurs primarily on H2A (K119) and H2B (K20 in humans and K123 in yeast). Both H2A and H2B are mainly monoubiquitylated and involved in crosstalk with other histone modifications such as histone methylations in a variety of cellular processes (Weake and Workman, 2008). H2A ubiquitylation is generally linked to gene silencing, while H2B ubiquitylation plays a role in both repression and activation of transcription. Ubiquitylation of H2A and H2B has been shown to participate in the process of NER. Meanwhile, a subset of NER-related proteins, such as XPC (Rad4 in yeast), DDB2, CSB, UVSSA and RPB1 (the largest subunit of RNAP II), are also ubiquitylated during NER (Gillette et al., 2006; Borsos et al., 2020; Apelt et al., 2021). For example, recent studies revealed that ELOF1 is required for the ubiquitylation of RPB1 K1268, a key signal for the recruitment of downstream repair factors including UVSSA and TFIIH (van der Weegen et al., 2021). However, how ubiquitylation of NER proteins functions in NER will not be discussed in detail in this review.

In yeast, the monoubiquitylation of H2B K123, catalyzed by Rad6/Bre1 complex, is partially required for global genomic repair. The Paf1 complex, a transcription elongation factor containing five subunits, is required for the catalytic activity of Rad6/Bre1 complex (Wood et al., 2003; Tatum et al., 2011; Tatum and Li, 2011). Interestingly, ubiquitylation of H2B K123 is essential for H3K79 methylation, which is catalyzed by Dot1 and required for global genomic repair but not TCR (Tatum and Li, 2011). Thus, it is likely that H2B ubiquitylation promotes global genomic repair indirectly by enabling the recruitment of Dot1 and the subsequent methylation of H3K79. Indeed, it was revealed that H2B ubiquitylation can regulate chromatin dynamics by enhancing nucleosome stability (Chandrasekharan et al., 2009; Chandrasekharan et al., 2010). Ubiquitylation of H2B is also necessary for H3K4 methylation which is mediated by the methyltransferase Set1-COMPASS and promotes NER in a similar way to H3K79 methylation.

UV-DDB, a heterodimeric complex containing DDB1 and DDB2, is part of a big ubiquitin E3 ligase complex that recognizes damaged chromatin and ubiquitylates core histones at damaged sites (Hannah and Zhou, 2009; Sugasawa, 2009). After UV irradiation, H2A (K118 and K119) and H2B are ubiquitylated by this ubiquitin E3 ligase complex (Kapetanaki et al., 2006; Lan et al., 2012). In addition, ubiquitin ligases Ring1B (Ring2) (Wang et al., 2004; Gracheva et al., 2016) and RNF8 (Marteijn et al., 2009) can also catalyze the H2A ubiquitylation. Ring1B was found to interact with UV-DDB-CUL4 and form a stable complex to ubiquitylate H2A at an early step of damage recognition after UV irradiation (Gracheva et al., 2016). Then, ZRF1, a H2A-ubiquitin binding protein, recognizes and remodels the UV-DDB-CUL4-RING1B complex causing the assembly of the canonical UV-DDB-CUL4 complex. XPC is, then, ubiquitylated by UV-DDB-CUL4 (Sugasawa et al., 2005). In this process, ZRF1 works as a switch protein that regulates XPC ubiquitylation through remodeling of the UV-DDB-CUL4-RING1B complex (Gracheva et al., 2016). It was later discovered that NER involves chromatin reorganization and ZRF1, in combination with XPC, facilitates the relocalization of damaged chromatin to the nucleolus for repair (Chitale and Richly, 2017). In the case of RNF8-mediated H2A ubiquitylation, it occurs after the incision step of NER and RNF8 is essential for the recruitment of downstream factors 53BP1 and BRCA1 (Marteijn et al., 2009; Zhu et al., 2009). Interestingly, RNAP II stalling caused by UV-induced DNA damage triggers H2B deubiquitylation at global level through histone deubiquitylases (Ubp8 and Ubp10) and H2B ubiquitylation level restores gradually during NER in yeast and humans (Mao et al., 2014). Without H2B deubiquitylation, TCR is decreased and RNAP II degradation is increased, suggesting deubiquitylation of H2B can facilitate rescue of RNAP II stalled at UV damage sites through TCR in chromatin (Meas and Mao, 2015).

Histones H3 and H4 can also be ubiquitylated by UV-DDB-CUL4 after UV irradiation, thus destabilizing the nucleosomes and facilitating the recruitment of XPC to DNA damage sites (Wang et al., 2006).

Histone phosphorylation is the addition of a phosphate group to histone residues (serine, threonine and tyrosine) by protein kinases and dephosphorylation is the removal of the phosphate group by phosphatases. The most well studied histone phosphorylation is that of the histone H2AX variant (γ-H2AX) on S139 in mammals (Rogakou et al., 1998) or S129 in yeast (Downs et al., 2000). Upon induction of DNA double strand breaks, histone H2AX is rapidly phosphorylated by ATM (ataxia telangiectasia mutated) and DNA-PKcs (Burma et al., 2001; Fernandez-Capetillo et al., 2004). In the case of UV induced γ-H2AX, the phosphorylation is mainly mediated by the kinase ATR (ataxia telangiectasia and Rad3-related) in the presence of DNA single-strand breaks (Ward and Chen, 2001; Hanasoge and Ljungman, 2007; Matsumoto et al., 2007). Histone phosphorylation reduces the positive charge of the histones, leading to a more open chromatin conformation. Unlike histone acetylation and methylation, histone phosphorylation interplays with other histone modifications and serves as a platform for recruiting factors for downstream cascade of events, such as DNA damage checkpoint activation (Fernandez-Capetillo et al., 2002).

It was shown that γ-H2AX has no significant effect on NER in mammals and yeast (Moore et al., 2007; Revet et al., 2011). Instead, NER plays a key role in the induction of γ-H2AX by generating single-stranded DNA during the repair process (Marti et al., 2006; Hanasoge and Ljungman, 2007; Matsumoto et al., 2007). Single-stranded DNA gap intermediates produced during NER can be extended by EXO1 (exonuclease 1) and coated by RPA (replication protein A), the major eukaryotic single stranded DNA-binding protein, which then recruits ATR and other ATR signaling proteins for ATR activation (Giannattasio et al., 2010; Sertic et al., 2011; Kemp, 2019). It is also likely that the single-stranded excision products released during NER are bound by RPA and involved in ATR kinase signaling pathway (Kemp and Sancar, 2012). Moreover, R-loop formation caused by the stalling of elongating RNAP II at a DNA lesion and DNA replication fork stalling induced by UV damage can activate H2AX phosphorylation (Halicka et al., 2005; Marti et al., 2006; Tresini et al., 2015). Interestingly, other phosphorylated histones, such as H3 (S10 and T11) in mammals and H2A (S122 and T126) in yeast, are dephosphorylated after UV irradiation (Sen and De Benedetti, 2006; Moore et al., 2007; Shimada et al., 2008). Phosphorylation on other histones, including H2B (T129) in yeast, H3 (T45) in humans and H3.3 (S31) in mouse embryonic stem cells, have been identified (Lee et al., 2014; Lee et al., 2015; Martire et al., 2019). However, whether and how histone phosphorylation on these residues is involved in NER remain to be investigated.