A defect in homologous recombination leads to increased translesion synthesis in E. coli

Abstract

DNA damage tolerance pathways allow cells to duplicate their genomes despite the presence of replication blocking lesions. Cells possess two major tolerance strategies, namely translesion synthesis (TLS) and homology directed gap repair (HDGR). TLS pathways involve specialized DNA polymerases that are able to synthesize past DNA lesions with an intrinsic risk of causing point mutations. In contrast, HDGR pathways are essentially error-free as they rely on the recovery of missing information from the sister chromatid by RecA-mediated homologous recombination. We have investigated the genetic control of pathway choice between TLS and HDGR in vivo in Escherichia coli In a strain with wild type RecA activity, the extent of TLS across replication blocking lesions is generally low while HDGR is used extensively. Interestingly, recA alleles that are partially impaired in D-loop formation confer a decrease in HDGR and a concomitant increase in TLS. Thus, partial defect of RecA's capacity to invade the homologous sister chromatid increases the lifetime of the ssDNA.RecA filament, i.e. the 'SOS signal'. This increase favors TLS by increasing both the TLS polymerase concentration and the lifetime of the TLS substrate, before it becomes sequestered by homologous recombination. In conclusion, the pathway choice between error-prone TLS and error-free HDGR is controlled by the efficiency of homologous recombination.

Figure 1.

(A) Gel retardation assay demonstrating the interaction between ssDNA oligonucleotide and increasing quantity of his-tagged forms of RecA(wt), RecA-G288Y, RecA-Q300A, RecA-N304D and non-tagged RecA(wt). 0.05μM ssDNA (60mer) was incubated with 0; 0.25; 0.5; 1 and 2 μM RecA-WT or RecA allele. RecA interaction with oligonucleotide is affected neither by the specific recA mutations nor by the presence of polyhistidine tag. (B) Kinetic of strand exchange (D-loop formation) promoted by his-tagged forms of RecA(wt), RecA-G288Y, RecA-Q300A, RecA-N304D and non-tagged RecA(wt). The D-loop formation was analyzed 0, 5, 10, 20 and 40 min after initiation of dsDNA uptake into recA filament. The graph represents the average and standard deviation of three independent experiments.

Figure 2.

Expression of RecA (A) cleavage of UmuD (B) and of LexA (C) in vivo. All cell extracts are made from cultures grown under standard conditions without genotoxic stress. (A) The level of expression of RecA in recA-Q300A and N304D extracts is intermediate between wt recA+ and lexA(Def) strains, illustrating partial SOS induction in Q300A and N304D strains. No induction is seen in allele G288Y. A similar conclusion can be reached from experiments monitoring UmuD cleavage into UmuD’ (B) or the disappearance of the intact form of LexA (C). All three assays suggest a relatively modest SOS induction in strain Q300A and a more robust one in strain N304D. o^c-umuD'C is a strain that expresses chromosomal umuD'C under the control of a constitutive promoter.

Figure 3.

DNA Damage tolerance of typical replication blocking lesions in partially recombination defective strains: partition between TLS and damage avoidance (DA): (A) major UV lesions, TT-CPD and TT(6-4) and (B) G-AAF. Tolerance events (%) are determined as the efficiency of site-specific integration of a single lesion-containing construct relative to the same amount of damage-free control construct. DA and TLS events are determined as specified in Materials and Methods. DDT profiles in ΔrecA strain were published previously in (6) and found to be mainly due to bacteria that could divide and proliferate following replication of the non-damaged chromatid only (1). Tolerance events for G-AAF in wt was previously published in (7). The data represent the average value and standard deviation of at least 3 independent experiments. T-test was performed to compare values from the different mutants to the recA(wt) strain. For TLS0: *P < 0.05; **P < 0.005. For TLS-2: ⁺P < 0.05; ⁺⁺P < 0.005; for DA: ^•P < 0.05; ^••P < 0.005.

Figure 4.

DNA damage tolerance of (A) G-AAF and (B) BaP(−) in the strains delayed for recombination: partition between TLS and DA. Tolerance events (%) are determined as the efficiency of site-specific integration of a single lesion-containing construct relative to the same amount of damage-free control construct. DA and TLS events are determined as specified in Materials and Methods section. The data represent the average value and standard deviation of at least three independent experiments. t-test was performed to compare values from the different mutants to the recA(wt) strain. For TLS0: *P < 0.05; **P < 0.005. For TLS-2: ⁺P < 0.05; ⁺⁺P < 0.005; for DA: ^•P < 0.05; ^••P < 0.005.

Figure 5.

Crosstalk between TLS and HDGR: When the fork encounters a replication-blocking lesion in one of the template strands, it skips over the lesion via downstream re-priming leaving a single-stranded gap. These single-stranded DNA gaps become converted into ssDNA.RecA filaments, the so-called SOS signal; loading of RecA on single-stranded DNA gaps is catalyzed by RecFOR recombination mediator proteins that help displace SSB from ssDNA. Recent evidence suggests that RecBCD partially contributes to ssDNA.RecA filament formation as well (47). Once formed, the ssDNA.RecA filament plays several important roles: (i) SOS induction leading to increased TLS polymerase expression, (ii) activation of Pol V, i(ii) stability of the fork and (iv) initiation of homologous recombination. In the present paper we show that TLS is increased in strains carrying recA alleles that are partially defective in D-loop formation. This increase in TLS is due to increased persistence of the ssDNA.RecA filament that in turn induces both TLS polymerase expression and the time during which the TLS substrate persists. The role of increased persistence of the TLS substrate on the extent of TLS, was also directly evidenced in recF (or recO) strains (see text). As soon as the ssDNA.RecA filament invades the homologous sister chromatid and forms a D-loop, i.e. an early HDGR intermediate, the TLS reaction is turned off by mere substrate sequestration. This model defines a chronological competition between TLS and HDGR, the time window for TLS closes upon initiation of recombination.