DNA damage triggers nucleotide excision repair-dependent monoubiquitylation of histone H2A

Abstract

Chromatin changes within the context of DNA repair remain largely obscure. Here we show that DNA damage induces monoubiquitylation of histone H2A in the vicinity of DNA lesions. Ultraviolet (UV)-induced monoubiquitylation of H2A is dependent on functional nucleotide excision repair and occurs after incision of the damaged strand. The ubiquitin ligase Ring2 is required for the DNA damage-induced H2A ubiquitylation. UV-induced ubiquitylation of H2A is dependent on the DNA damage signaling kinase ATR (ATM- and Rad3-related) but not the related kinase ATM (ataxia telangiectasia-mutated). Although the response coincides with phosphorylation of variant histone H2AX, H2AX was not required for H2A ubiquitylation. Together our data show that monoubiquitylation of H2A forms part of the cellular response to UV damage and suggest a role of this modification in DNA repair-induced chromatin remodeling.

Figure 1.

DNA damage induces accumulation of ubiquitin. (A) Mel JuSo cells stably expressing GFP-Ub were locally irradiated in the nucleus with a high-intensity 405-nm laser beam. Accumulation of GFP-Ub appeared at the local laser-irradiated spot, indicated by the arrow. (B) Mel JuSo cells stably expressing GFP-Ub were locally irradiated with UV-C light. A clear accumulation of GFP-Ub is visible at the local UV-irradiated areas. The arrows indicate the local UV DNA-damaged spot visualized with either the presence of DNA damage by CPD counterstaining (top middle panel) or the presence of the NER protein XPC (bottom middle panel). Colocalization of GFP-Ub with the DNA damage (top right panel) and XPC protein (bottom right panel) is shown. (C) An antibody recognizing conjugated endogenous ubiquitin revealed accumulation of ubiquitin at local irradiated spots, marked by the presence of XPC protein, in primary human fibroblasts (C5RO). Staining with the ubiquitinspecific antibody (right panel), the XPC-specific antibody (middle panel), and the merged images (left panel) are shown. (D) Ubiquitin accumulation at local DNA damage is dependent on conjugation. The nonfunctional mutant GFP-Ub^K0,G76V (left panel) did not accumulate in Mel JuSo cells at the DNA damage, as detected by CPD (top middle image) or the presence of XPC (bottom middle image).

Figure 2.

DNA damage causes monoubiquitylation without proteasome recruitment. (A) Cells were transfected with either GFP-Ub or the mutant GFP-Ub^k0, which can be efficiently monoubiquitylated but hinders polyubiquitylation. Both GFP-Ub and GFP-Ub^K0 localized at the laser-irradiated areas, which are indicated by the arrows. (B) Quantification of fluorescence recovery in the bleached area circled in A. Accumulation of GFP-Ub (closed circles) and GFP-Ub^K0 (open circles) did not differ, indicating that the observed GFP-Ub accumulation is primarily caused by monoubiquitylation. (C) Proteasomes tagged with α3-GFP fusion subunit were followed in Mel JuSo cells after high-intensity laser bleaching as in A. The circle denotes the bleached spot. The proteasomes did not accumulate at the high laser-irradiated spots. (D) The mRFP-Ub fusion accumulates on local laser-induced damage in G1 or G2 cells as was observed by the GFP-PCNA staining. (E) Similarly, cells with a typical S-phase-like GFP-PCNA staining were locally damaged using the 405 laser; mRFP-Ub accumulated on these sites.

Figure 3.

UV-induced local monoubiquitylation is dependent on functional NER. (A) A schematic drawing of the NER pathway is shown. NER factor deficiencies that have been tested in this study are indicated in red. NER lesion recognition is predominantly performed by the XPC complex, after which the TFIIH complex and XPA and RPA are recruited, probably proceeded by XPG. The DNA helix is unwound by the TFIIH complex in an ATP-dependent fashion. The two endonucleases, XPG and XPF-ERCC1, excise the lesion. After incision, the GAP is filled in by DNA polymerase δ/ε, PCNA, and RFC and is sealed by ligase I (Hoeijmakers 2001). (B) In normal fibroblasts (C5RO-hTERT), GFP-Ub (left panel) accumulated at local damaged spots, indicated by the arrow. (C) In XP-C human fibroblast (XP21RO-Sv), GFP-Ub (left panel) did not accumulate at local damaged spots as detected by CPD (middle panel). (D) GFP-Ub (left panel) did not accumulate in XP-A cells (XP12RO-Sv) at NER active sites as detected by XPC (middle panel). (E) In XP-G (XP3BR-Sv), GFP-Ub (left panel) did not accumulate at local UV-irradiated spots as indicated by the presence of XPC protein (middle panel). (F) In primary NER proficient XP-V cells (XP1RO), deficient in TLS polymerase η, ubiquitin accumulated at the UV-induced lesions.

Figure 4.

Global UV irradiation causes NER-dependent immobilization of nuclear ubiquitin. (A) FRAP analysis of GFP and GFP-Ub in nuclei of wild-type C5RO-hTERT cells. The immobile pool of GFP-Ub is indicated. (B) FRAP analysis of GFP-Ub in nuclei of wild-type C5RO-hTERT cells. Cells were left untreated or exposed to 4, 8, and 16 J/m² UV light. Approximately 21% of the total nuclear ubiquitin pool was immobilized in wild-type C5RO-hTERT human fibroblasts. UV increased the immobile fraction in a dose-dependent manner to 24%, 27%, and 30% immobilization after 4, 8, and 16 J/m², respectively (n > 24). Note that the FRAP curve of untreated GFP-Ub cells is the same data set as seen in A. (C) FRAP analysis of nuclear GFP-Ub in the nuclei of NER-deficient XP-A cells (XP12RO-Sv). Note that the UV damage-induced immobilization was not observed in XP-A cells (mean of 25 cells). (D) FRAP analysis of GFP-Ub^K0,G76V and GFP in the nuclei of wild-type C5RO-hTERT cells. The dynamics of the mutant GFP-Ub^K0,G76V and GFP (mean of 16 cells) in C5RO-hTERT cells were similar. No difference was detected in mobility before and after UV exposure (mean of 16 cells).

Figure 5.

UV irradiation causes NER-dependent ubiquitylation of histone H2A. (A) Core histones were purified before and after UV (16 J/m²) exposure from NER-proficient (MRC5-Sv) and probed with an antibody directed against ubiquitin. An increase in a 23-kDa protein reacting with antibodies against ubiquitin was observed. (B) The samples shown in A probed with an antibody specific for uH2A. Blots in the bottom panels of A and B were stripped and reprobed with H2B-specific antibody. (C) Peptides identified by liquid chromatography tandem mass spectroscopy from the top band shown in A. MRC5-Sv (NER-proficient) (D) and XP12RO-Sv (XP-A, NER-deficient) (E) were locally UV-irradiated and stained with uH2A- and XPC-specific antibodies. uH2A accumulated in NER-proficient cells (top panels) at the damaged spot but not in XP-A cells (bottom panels). (F) Core histones were purified before and after UV (16 J/m²) treatment from cells ectopically expressing ^FlagH2A or ^FlagH2A^K119R. Western blots probed with anti-Flag. An increase in the band corresponding to ubiquitylated ^FlagH2A was detected after UV treatment, while no ubiquitylated ^FlagH2A^K119R before or after UV treatment was detected.

Figure 6.

The ubiquitin ligase Ring2 is required for DNA damage-induced H2A ubiquitylation. GFP-Ub expressing Mel JuSo cells were transfected with a plasmid encoding a Ring2-specific RNAi and mRFP. (Top left panel) The transfected cells could be identified by mRFP expression (grayscale). This picture shows four representative transfected cells (asterisks) and two untransfected cells. (Top right panel) Ring2 RNAi transfected cells had reduced nuclear GFP-Ub levels (LUT [Look Up Table] color). (Bottom left panel) Laser damage was induced in two untransfected cells and two RNAi transfected cells (arrows) (LUT color). (Bottom right panel) The untransfected cells showed a clear accumulation of GFP-Ub in the laser exposed region whereas the response was very low in both Ring2 RNAi transfected cells (LUT color). The LUT for the color coding is shown at the right. Note that blue refers to overexposed on the LUT scale.

Figure 7.

H2A ubiquitylation does not require variant histone H2AX but is dependent on ATR. (A) Local DNA damage was caused by UV light in CRO5 cells. Phosphorylated H2AX and accumulation of ubiquitin colocalized on the lesion. (B) Wild-type and H2AX^−/− MEF cells were transfected with GFP-Ub. After 405-nm laser damage, GFP-Ub localized at the laser-irradiated areas, which are indicated by the arrows. Local DNA damage was inflicted in C5RO (NER-proficient) cells and ATR mutant cells. In C5RO cells (C), but not in ATR mutant cells (D), endogenous ubiquitin accumulated on the lesions. ATR (E) and ATM (F) mutant cell lines were transiently transfected with GFP-Ub. The lesions were identified by XPC staining. In the ATM cells, but not in the ATR cells, GFP-Ub was found enriched on the DNA damage.