The NuA4 complex promotes translesion synthesis (TLS)-mediated DNA damage tolerance

Abstract

Lesions in DNA can block replication fork progression, leading to its collapse and gross chromosomal rearrangements. To circumvent such outcomes, the DNA damage tolerance (DDT) pathway becomes engaged, allowing the replisome to bypass a lesion and complete S phase. Chromatin remodeling complexes have been implicated in the DDT pathways, and here we identify the NuA4 remodeler, which is a histone acetyltransferase, to function on the translesion synthesis (TLS) branch of DDT. Genetic analyses in Saccharomyces cerevisiae showed synergistic sensitivity to MMS when NuA4 alleles, esa1-L254P and yng2Δ, were combined with the error-free bypass mutant ubc13Δ. The loss of viability was less pronounced when NuA4 complex mutants were disrupted in combination with error-prone/TLS factors, such as rev3Δ, suggesting an epistatic relationship between NuA4 and error-prone bypass. Consistent with cellular viability measurements, replication profiles after exposure to MMS indicated that small regions of unreplicated DNA or damage were present to a greater extent in esa1-L254P/ubc13Δ mutants, which persist beyond the completion of bulk replication compared to esa1-L254P/rev3Δ. The critical role of NuA4 in error-prone bypass is functional even after the bulk of replication is complete. Underscoring this observation, when Yng2 expression is restricted specifically to G2/M of the cell cycle, viability and TLS-dependent mutagenesis rates were restored. Lastly, disruption of HTZ1, which is a target of NuA4, also resulted in mutagenic rates of reversion on level with esa1-L254P and yng2Δ mutants, indicating that the histone variant H2A.Z functions in vivo on the TLS branch of DDT.

Keywords: DNA damage tolerance; Esa1; H2A.Z; NuA4; Yng2.

Figure 1

The NuA4 complex genetically interacts with the DNA damage tolerance (DDT) pathway. (A) Schematic of the NuA4 complex. Esa1 and Yng2 are part of the smaller piccolo NuA4 complex that also includes Epl1 and Eaf6. The large NuA4 complex forms when Epl1 in piccolo interacts with Eaf1. (B) Cell survival was measured after transient exposure to increasing concentrations of MMS for 1 hr at 30° for wild type (JC470), esa1-L254P (JC2767), ubc13Δ (JC2291), esa1-L254P/ubc13Δ (JC2775), rev3Δ (JC2289), and esa1-L254P/rev3Δ (JC2771), (C) yng2Δ (JC2036), yng2Δ/ubc13Δ (JC2285) and yng2Δ/rev3Δ (JC2281), or (D) rad51Δ (JC1362), esa1-L254P/rad51Δ (JC3253) and yng2Δ/rad51Δ (JC2437). Multiple experiments (three or more) were averaged with standard deviation being reported.

Figure 2

Decreased cell survival after MMS exposure correlates with DNA damage in G2. (A) Cell survival was measured after transient exposure to increasing concentrations of MMS for 1 hr with wild type (JC470), ubc13Δ (JC2291), rev3Δ (JC2289), and rev3Δ/ubc13Δ (JC2777). (B) Cells were arrested in α-factor for 2 hr followed by release into YPAD media containing bromodeoxyuridine (BrdU; 400 µg/ml) and 0.01% MMS for 1 hr. Following MMS treatment, cells were released into YPAD + BrdU before samples were collected at the indicated time points. (C) Pulse field gel electrophoresis (PFGE) was performed, followed by a Southern transfer to nitrocellulose, and blotted with α-BrdU antibodies in wild type (JC604), rev3Δ (JC2978), ubc13Δ (JC2979), and rev3Δ/ubc13Δ (JC3009). The cell cycle stage was monitored by flow cytometry where G1 (black) and the 90-min time points after MMS release (red) are shown. (D) The BrdU signal at the 0 min (blue) and 60 min (green) time points were quantified by ImageJ with the migration distance of chromosomes vs. the intensity of BrdU plotted, giving a measure of newly synthesized chromosomes during one round of DNA replication. (E) Budding index was performed with samples to measure G1/S transition. (F) Quantitative densitometry was performed on the BrdU signal from chromosomes 7 and 15 after release from MMS using Bio-Rad Quantity One software at 0 min (blue), 60 min (green), and 90 min (red).

Figure 3

DNA lesions remain in G2 when the error-free pathway is disrupted in combination with the loss of NuA4 acetyltransferase activity. PFGE, cell cycle progression, and quantification were determined (as in Figure 2) for (A–D) wild type (JC604), esa1-L254P (JC3060), rev3Δ (JC2978), and esa1-L254P/rev3Δ (JC3053) and (E–H) ubc13Δ (JC2979) and esa1-L254P/ubc13Δ (JC3054).

Figure 4

Transcription of DDT factors is not dramatically altered in esa1-L254P mutants upon MMS treatment. (A) Viability assays as described in Figure 1 were performed for wild type (JC470), rev3Δ (JC2289), ubc13Δ (JC2291), hhf2^K→R (JC3178), hhf2^K→R/rev3Δ (JC3195), hhf2^K→R/ubc13Δ (JC3179) and (B) htz1Δ (JC2090), htz1Δ/rev3Δ (JC2762), and htz1Δ/ubc13Δ (JC2764). (C–G) qRT-PCR as described in Materials and Methods was performed on wild-type and esa1-L254P cells treated with α-factor for 2 hr followed by release into normal YPAD (S phase) or YPAD with 0.05% MMS (S + MMS) for 1 hr. Candidate genes in the DDT pathway were analyzed for expression with and without MMS. The qRT-PCR values from target genes were normalized to ALG9 (Table S1), as its transcript levels were most stable throughout the cell cycle.

Figure 5

Restricting expression of the NuA4 subunit Yng2 to G2 of the cell cycle rescues the cellular viability after MMS treatment. (A) Schematic of Clb2 (top panel) showing the nuclear-export signal (gray), D (blue), KEN boxes (orange), the nuclear localization signal (purple), and two cyclin domains (green). In G2–Yng2 (bottom panel) the YNG2 gene was fused to the G2 tag (Karras and Jentsch 2010), which includes the DNA sequence of the CLB2 promoter (pCLB2; activated in G2) and 180 aa of Clb2 with the nuclear export signal mutated, L26A, indicated by an asterisk (*) and the D- and KEN-box degrons. The G2 tag reacts with the Clb2 antibody and for the purposes of determining expression patterns by Western blot the G2–Yng2 fusion was further Flag tagged because it was difficult to differentiate G2–Yng2 from Clb2 due to their sizes. All subsequent analyses were performed with a G2–Yng2 fusion that did not have a Flag epitope tag. (B) G2–Yng2 and Clb2 are expressed only in G2/M. G2–Yng2–10Flag cells (JC3387) were arrested in G1 with α-factor for 3 hr, followed by release into YPAD with samples collected at the indicated time points prior to SDS–PAGE and Western blot analysis with α-Clb2, α-Flag (to visualize G2–Yng2–Flag), and α-Pgk1 as a loading control. (C and D) Drop assays (1:10 serial dilutions) from exponentially growing cultures were performed on YPAD ± media containing the indicated concentrations of MMS at 30° and viability assays as described in Figure 1 were performed for wild type (JC470), ubc13Δ (JC2291), yng2Δ (JC2036), G2–YNG2 (JC3255), yng2Δ/ubc13Δ (JC2285), and G2–YNG2/ubc13Δ (JC3257).