Defining the RNaseH2 enzyme-initiated ribonucleotide excision repair pathway in Archaea

Abstract

Incorporation of ribonucleotides during DNA replication has severe consequences for genome stability. Although eukaryotes possess a number of redundancies for initiating and completing repair of misincorporated ribonucleotides, archaea such as Thermococcus rely only upon RNaseH2 to initiate the pathway. Because Thermococcus DNA polymerases incorporate as many as 1,000 ribonucleotides per genome, RNaseH2 must be efficient at recognizing and nicking at embedded ribonucleotides to ensure genome integrity. Here, we show that ribonucleotides are incorporated by the hyperthermophilic archaeon Thermococcus kodakarensis both in vitro and in vivo and a robust ribonucleotide excision repair pathway is critical to keeping incorporation levels low in wild-type cells. Using pre-steady-state and steady-state kinetics experiments, we also show that archaeal RNaseH2 rapidly cleaves at embedded ribonucleotides (200-450 s^-1), but exhibits an ∼1,000-fold slower turnover rate (0.06-0.17 s^-1), suggesting a potential role for RNaseH2 in protecting or marking nicked sites for further processing. We found that following RNaseH2 cleavage, the combined activities of polymerase B (PolB), flap endonuclease (Fen1), and DNA ligase are required to complete ribonucleotide processing. PolB formed a ribonucleotide-containing flap by strand displacement synthesis that was cleaved by Fen1, and DNA ligase sealed the nick for complete repair. Our study reveals conservation of the overall mechanism of ribonucleotide excision repair across domains of life. The lack of redundancies in ribonucleotide repair in archaea perhaps suggests a more ancestral form of ribonucleotide excision repair compared with the eukaryotic pathway.

Keywords: DNA damage; DNA repair; DNA replication; DNA synthesis; RNaseH2; archaea; enzyme kinetics; pre-steady-state kinetics; ribonucleotide excision repair.

Figure 1.

Archaea require an RNaseH2-initiated RER pathway. A, genomic DNA from wild-type and ΔRNaseH2 T. kodakarensis (Tko) cells was treated with either 0.3 m NaCl or 0.3 m NaOH, separated on a 1% alkaline-agarose gel, and visualized using SYBR Gold staining. The fluorescence intensity distribution was quantified using ImageQuant software. B, primed M13mp18 ssDNA was fully extended by either Thermococcus sp. 9°N PolB or PolD with dNTPs or dNTPs/excess rNTPs (see “Experimental procedures”) with [³²P]dCTP in place of dCTP for phosphorimaging. Full extension products were visualized by neutral agarose and rNMP incorporation was assessed by 0.3 m NaOH treatment and alkaline-agarose electrophoresis. C, RNaseH2 activity in Tko extracts was monitored by CE using a 50-bp dsDNA substrate with an embedded rGMP nucleotide, a 5′-FAM label, and a 3′-MAX label. Reactions were carried out over a time course from 15 s to 20 min at 60 °C. A subset of representative CE traces are shown indicating the formation of 21- and 29-nt RNaseH2 products. 3′-MAX-labeled Fen1 flap cleavage products (<29 nt) are also observed. D, the formation of the 29-nt MAX product was quantified over time for both wild-type (red) and ΔRNaseH2 (blue) extracts. Data are the average of three biological replicates and error bars indicate standard deviation.

Figure 2.

RNaseH2 cleavage is not the rate-limiting step in RER. A, for steady-state kinetics, a 5-fold excess of 50 bp, 5′-FAM, 3′-MAX-labeled rGMP RER substrate was rapidly mixed with purified Thermococcus sp. 9°N RNaseH2 in an RQF instrument at 60 °C. The reaction was quenched with 50 mm EDTA. B, the conversion of the 50-nt, dual-labeled rG substrate to 5′-FAM and 3′-MAX cleavage products was monitored over time using CE. A subset of representative traces from a time course are shown. C, the yield of 5′-FAM product was graphed as a function of time and fit to Equation 1 to obtain k_ss = 0.06 s⁻¹ and the active enzyme concentration (A) = 5.5 nm. A representative plot for the rG substrate is shown and replicates for rG and the other ribonucleotide substrates are shown in supplemental Fig. S1. D, pre-steady state kinetics were performed with purified Thermococcus sp. 9°N RNaseH2 chelated for metal ion, pre-bound to the 50-bp RER substrates with different embedded rNMPs. The enzyme was in 3-fold excess to the DNA substrate. The pre-bound DNA-RNaseH2 complex was rapidly mixed with buffer containing MgCl₂ in an RQF at 60 °C and quenched with 50 mm EDTA. Cleavage products were visualized by CE. E, the yield of 5′-FAM product was graphed as a function of time and fit to Equation 2 to obtain k_cleavage for rG, rA, rC, and rU (222, 196, 447, and 426 s⁻¹). A representative plot for the rG substrate is shown. Technical replicates of the pre-steady-state experiments for the rG and other ribonucleotide substrates are shown in supplemental Fig. S2.

Figure 3.

Dual-label fluorescence assay to monitor post-RNaseH2 RER by capillary electrophoresis. A, the RER substrate was generated by annealing a 44-nt (5′-MAX labeled) oligonucleotide and a 90-nt (3′-FAM labeled with a 5′-phosphate-rGMP) oligonucleotide to ssM13 DNA. Annealed oligos form a DNA substrate containing a nick with 3′-OH and 5′-phosphate-rG termini to mimic DNA nicked by RNaseH2. On CE, the double-stranded DNA is denatured and ssDNA oligonucleotides can be visualized individually. On the right is a hypothetical CE trace representing the expected result with two individual MAX and FAM oligonucleotide peaks. B, when a strand-displacing DNA polymerase is added, 5′-MAX-labeled strand displacement products larger than 44-nt can be observed. C, when a flap endonuclease is added, 3′-FAM products smaller than 90-nt (predominantly by 1–2 nt) are observed. D, full sealing and repair by ligation results in a FAM- and MAX-labeled 134-nt product.

Figure 4.

Thermococcus extracts reveal critical roles for PolB and Fen1 in RER. A, the substrate depicted in Fig. 3 was incubated with cell extracts prepared from Tko wild-type, Tko ΔPolB, and Tko ΔFen1 strains over a time course from 0 to 60 min at 60 °C (subset of representative traces shown). Over time the overall signal of the substrate peaks (44-nt MAX and 90-nt FAM) decrease due to substrate conversion to repaired product (134-nt FAM/MAX) as well as competing enzymatic activities. B, the amount of repaired product at 60 min was quantified for each extract. The data shown are the average of four independent experiments with S.D.

Figure 5.

Archaeal RER reconstituted in vitro. A, reaction schematic. The substrate depicted in Fig. 3 is incubated with purified Thermococcus sp. 9°N proteins including PCNA, RFC, and different combinations of PolB, PolD, Fen1, and DNA ligase. Reactions were incubated over a time course from 0 to 30 min at 60 °C and repair was monitored by appearance of the 134-nt FAM/MAX-labeled DNA product by CE. B, representative CE traces for each reaction condition at the 30-min time point are shown. C, quantification of the conversion of 90-nt FAM substrate to 134-nt FAM/MAX product at the 30-min time point. Data are the average of three independent experiments with S.D.

Figure 6.

Simplified models of ribonucleotide excision repair in Eukarya (A), Archaea (B), and Bacteria (C). A, rNMPs are incorporated into eukaryotic genomic DNA by any of the replicative polymerases (Polϵ, Polδ, or Polα) with the incorporation frequencies shown (6, 9, 12). The RER pathway begins with incision by the heterotrimeric RNaseH2 followed by strand displacement synthesis by either of the replicative polymerases Polϵ or Polδ, flap cleavage by Fen1 or Exo1, and sealing by DNA ligase I. B, archaeal genomic DNA acquires rNMPs primarily through incorporation by the leading and lagging strand polymerase PolD at a rate of 1 rN in 1,500 nucleotides synthesized. Monomeric RNaseH2 nicks DNA at rNMP sites. Following cleavage, strand displacement synthesis by PolB creates a flap that is cleaved by Fen1 and DNA ligase seals the resulting nick. C, in bacteria, the replicative polymerase PolIII incorporates rNMPs in genomic DNA and monomeric RNaseHII nicks at these sites. PolI then fulfills two functions in RER by performing both strand displacement synthesis and flap cleavage. DNA ligase then seals the nick.