PSO4: a novel gene involved in error-prone repair in Saccharomyces cerevisiae

The haploid xs9 mutant, originally selected for on the basis of a slight sensitivity to the lethal effect of X-rays, was found to be extremely sensitive to inactivation by 8-methoxypsoralen (8MOP) photoaddition, especially when cells are treated in the G2 phase of the cell cycle. As the xs9 mutation showed no allelism with any of the 3 known pso mutations, it was now given the name of pso4-1. Regarding inactivation, the pso4-1 mutant is also sensitive to mono- (HN1) or bi-functional (HN2) nitrogen mustards, it is slightly sensitive to 254 nm UV radiation (UV), and shows nearly normal sensitivity to 3-carbethoxypsoralen (3-CPs) photoaddition or methyl methanesulfonate (MMS). Regarding mutagenesis, the pso4-1 mutation completely blocks reverse and forward mutations induced by either 8MOP or 3CPs photoaddition, or by gamma-rays. In the cases of UV, HN1, HN2 or MMS treatments, while reversion induction is still completely abolished, forward mutagenesis is only partially inhibited for UV, HN1, or MMS, and it is unaffected for HN2. Besides severely inhibiting induced mutagenesis, the pso4-1 mutation was found to be semi-dominant, to block sporulation, to abolish the diploid resistance effect, and to block induced mitotic recombination, which indicates that the PSO4 gene is involved in a recombinational pathway of error-prone repair, comparable to the E. coli SOS repair pathway.     

Repair promoted by plasmid pKM101 is different from SOS repair.

In E. coli K12 bacteria carrying plasmid pKM101, prophage lambda was induced at UV doses higher than in plasmid-less parental bacteria. UV-induced reactivation per se was less effective. Bacteria with pKM101 showed no alteration in their division cycle. Plasmid pKM101 coded for a constitutive error-prone repair different from the inducible error-prone repair called SOS repair. Plasmid pKM101 protected E. coli bacteria from UV damage but slightly sensitized them to X-ray lesions. Protection against UV damage was effective in mutant bacteria deficient in DNA excision-repair provided that the recA, lexA and uvrE genes were functional. Survival of phages lambda and S13 after UV irradiation was enhanced in bacteria carrying plasmid pKM101; phage lambda mutagenesis was also increased. Plasmid pKM101 repaired potentially lethal DNA lesions, although wild-type DNA sequences may not necessarily be restored; hence the mutations observed are the traces of the original DNA lesions.     

Genetic control of multiple pathways of post-replicational repair in uvrB strains of Escherichia coli K-12.

The effect of the recA, uvrD, exrA, and recB mutations and of post-irradiation treatment with chloramphenicol on the survival and post-replication repair after ultraviolet irradiation of uvrB strains of Escherichia coli K-12 was examined. Each of these mutations or treatments was found to decrease survival and the extent of repair. The interactions of the inhibitory effects of the uvrD, exaA, and recB mutations and chloramphenicol treatment were determined by examining the survival and repair characteristics of the several multiple mutants. The survival results suggest that the post-replication repair process in uvrB strains may be subdivided into at least five different branches. These include three branches that are blocked by the exrA, recB, or uvrD mutation, a fourth branch that is blocked by any of these mutations and is also sensitive to chloramphenicol treatment, and at least one additional branch that is not sensitive to either of these mutations or to chloramphenicol treatment. The extent of post-replicational repair observed with each of the strains is in general agreement with the pathways postulated on the basis of the survival data, although there are several apparent exceptions to this correlation.     

Mutagenesis by normal metabolites in Escherichia coli: phenylalanine mutagenesis is dependent on error-prone DNA repair

In search of a model for the production of 'spontaneous' mutations induced by DNA damage produced during normal metabolism, 19 amino acids were tested for mutagenicity in Escherichia coli K-12 uvrB. Cystine, and, to a lesser extent, arginine and threonine were found to be antimutagenic; only phenylalanine was found to be mutagenic. At 2 mM, phenylalanine induced mutants at 1.5-2-fold above background [lacZ53(amber)----Lac+, rifampicin resistance (missense), and bacteriophage T6 resistance]. Tyrosine and, to a lesser extent, tryptophan (each at 2 mM) inhibited the mutagenicity of phenylalanine. Phenylalanine mutagenesis was detected in the uvrB strain, but not in the wild-type, uvrB umuC or uvrB lexA strains. Thus, phenylalanine seems to cause the production of excisable lesions ('UV-like'?) in DNA, which, if not excised, can induce mutations via error-prone DNA repair.     

Prophage induction in Haemophilus influenzae and its relationship to mutation by chemical and physical agents

It is known that UV, X-rays, MMC and MMS are not mutagenic for H. influenzae, whereas HZ, EMS and MNNG are potent mutagens for this bacterium. All of these agents, however, are known to be both mutagenic and able to induce prophage in E. coli. We report here that all the agents except HZ induce prophage in H. influenzae, and EMS even induces in the recombination-defective recl mutant, which is non-inducible by UV, MMC, MNNG and MMS. MMS did not cause single-strand breaks or gaps in DNA synthesized after treatment of H. influenzae, but EMS and MNNG produced them. EMS caused more breaks in DNA synthesized before treatment than in that synthesized after treatment. On the other hand we did observe such breaks or gaps induced in E. coli in DNA synthesized posttreatment by EMS as well as by MMS and MNNG, at comparable survival levels.     

DNA repair defects channel interstrand DNA cross-links into alternate recombinational and error-prone repair pathways

The repair of psoralen interstrand cross-links in the yeast Saccharomyces cerevisiae involves the DNA repair groups nucleotide excision repair (NER), homologous recombination (HR), and post-replication repair (PRR). In repair-proficient yeast cells cross-links induce double-strand breaks, in an NER-dependent process; the double-strand breaks are then repaired by HR. An alternate error-prone repair pathway generates mutations at cross-link sites. We have characterized the repair of plasmid molecules carrying a single psoralen cross-link, psoralen monoadduct, or double-strand break in yeast cells with deficiencies in NER, HR, or PRR genes, measuring the repair efficiencies and the levels of gene conversions, crossing over, and mutations. Strains with deficiencies in the NER genes RAD1, RAD3, RAD4, and RAD10 had low levels of cross-link-induced recombination but higher mutation frequencies than repair-proficient cells. Deletion of the HR genes RAD51, RAD52, RAD54, RAD55, and RAD57 also decreased induced recombination and increased mutation frequencies above those of NER-deficient yeast. Strains lacking the PRR genes RAD5, RAD6, and RAD18 did not have any cross-link-induced mutations but showed increased levels of recombination; rad5 and rad6 cells also had altered patterns of cross-link-induced gene conversion in comparison with repair-proficient yeast. Our observations suggest that psoralen cross-links can be repaired by three pathways: an error-free recombinational pathway requiring NER and HR and two PRR-dependent error-prone pathways, one NER-dependent and one NER-independent.     

Comparative analysis of deletion and base-change mutabilities of Escherichia coli B strains differing in DNA repair capacity (wild-type, uvrA-, polA-, recA-) by various mutagens.

Dose-response curves were compared for deletions [ColBR (resistant to colicin B) mutations being more than 80% deletions] and base changes (reversion of argFam to prototrophy argplus) induced in the same set of E. coli strains (wild-type for DNA repair, uvrA-, polA- and recA-) by N-methyl-N'-nitro-N-nitrosoguanidine (NTG), ethyl methanesulfonate (EMS), hydroxylamine (HA), 4-nitroquinoline I-oxide (4NQO), mitomycin C (MTC, UV and X-rays. All these agents induced deletions as well as base changes in the wild-type strain. Thus chemical mutagenesis differed in E. coli and bacteriophages in vitro, for HA, NTG, EMS and perhaps UV produced only point mutations in phage Tr. The patterns of deletion and base-change mutability in E. coli were surprisingly similar. (I) The recombination less recA- strain was mutable by only three (NTG, EMS, HA) of the seven mutagens for either deletions or base changes. (2) The uvrA- strain, unable to excise pyrimidine dimers, was very highly mutable by 4NQO and UV but immutable by MTC for both deletions and base changes. (3) The polA- strain, defective in DNA polymerase I due to a non-suppressible mutation, was very highly mutable by HA and highly mutable by MTC and 4NQO for both deletions and base changes but was highly mutable only for deletions by UV and X-rays, remaining normally mutable by the other agents for both deletions and base changes despite its high sensitivity to their inactivating action. We conclude that errors in the recA-dependent repair of induced DNA damage (after 4NQO, MTC, UV and X-rays) or errors in replication enhanced by damage to the replication system or to the template strands (after NTG, EMS, and HA) give rise to deletions as well as to base changes. From a comparative analysis of 14 dose-response curves for deletions and base changes, we conclude that the order of mutagenic efficiency relative to killing is (EMS, NTG) greater than (UV, 4NQO) greater than HA greater than (X-rays, MTC), and that X-rays, 4NQO, HA and MTC induce more ColBR deletions than Argplus base changes, whereas UV and EMS induce ColBR deletions and Argplus base changes at nearly equal rates and the specificity of NTG is intermediate between these two types.     

Construction and characterization of two lexA mutants of Salmonella typhimurium with different UV sensitivities and UV mutabilities.

Salmonella typhimurium has a SOS regulon which resembles that of Escherichia coli. recA mutants of S. typhimurium have already been isolated, but no mutations in lexA have been described yet. In this work, two different lexA mutants of S. typhimurium LT2 have been constructed on a sulA background to prevent cell death and further characterized. The lexA552 and lexA11 alleles contain an insertion of the kanamycin resistance fragment into the carboxy- and amino-terminal regions of the lexA gene, respectively. SOS induction assays indicated that both lexA mutants exhibited a LexA(Def) phenotype, although SOS genes were apparently more derepressed in the lexA11 mutant than in the lexA552 mutant. Like lexA(Def) of E. coli, both lexA mutations only moderately increased the UV survival of S. typhimurium, and the lexA552 strain was as mutable as the lexA+ strain by UV in the presence of plasmids encoding MucAB or E. coli UmuDC (UmuDCEc). In contrast, a lexA11 strain carrying any of these plasmids was nonmutable by UV. This unexpected behavior was abolished when the lexA11 mutation was complemented in trans by the lexA gene of S. typhimurium. The results of UV mutagenesis correlated well with those of survival to UV irradiation, indicating that MucAB and UmuDCEc proteins participate in the error-prone repair of UV damage in lexA552 but not in lexA11. These intriguing differences between the mutagenic responses of lexA552 and lexA11 mutants to UV irradiation are discussed, taking into account the different degrees to which the SOS response is derepressed in these mutants.     

Inducible error-prone repair in yeast. Suppression by heat shock

The production of reversion mutations in wild-type, diploid Saccharomyces cerevisiae by the alkylating agents N-methyl-N'-nitro- N-nitrosoguanidine (MNNG) and methylnitrosourea (MNU) was suppressed in cells previously treated with a heat shock, or the protein synthesis inhibitor, cycloheximide. The same cells previously treated with a heat shock, or the protein synthesis inhibitor, cycloheximide. The same treatment after mutagen exposure did not lower the induced mutation frequency. In split-dose experiments, a first MNNG exposure prevented subsequent heat (or cycloheximide) treatment from blocking mutation by a second, later mutagen exposure. These data suggest that, in yeast, MNNG or MNU induces an error-prone DNA-repair system, and that this induction is blocked by protein-synthesis inhibitors. The specificity of this system for different types of DNA damage was investigated using a variety of other mutagenic agents. A prior heat shock did not suppress mutation produced by exposure to ethyl methanesulfonate, ethylnitrosourea, 8-methoxypsoralen + UVA, or gamma-radiation. Partial suppression was observed in cells exposed to methyl methanesulfonate or to 254-nm ultraviolet light. These results indicate that, unlike the SOS system of E. coli, this inducible error-prone process of yeast is responsive to only certain mutagens. Heat shock suppression of mutation produced by MNNG exposure was also demonstrated in wild-type haploid cells, as well as haploid strains mutant in representative genes of the RAD52 epistasis group (rad52, rad53, rad54), the RAD3 epistasis group (rad1, rad2, rad3) and the RAD6 epistasis group (rad9, rad18). The rad6 mutant itself was immutable with MNNG and therefore untestable by these techniques. These data indicate that this error-prone repair system is not absolutely dependent on the integrity of the RAD52 (recombination) or the RAD3 (excision) systems, or on at least some parts of the RAD6 system.     

Effect of tsl (thermosensitive suppressor of lex) mutation on postreplication repair in Escherichia coli K-12.

Cells of Escherichia coli K-12 carrying lexA or recA mutations are more sensitive to UV radiation than corresponding wild-type cells and are defective in postreplication repair. Supressor mutations (tsl) have been described previously which increase the UV resistance of lexA uvr+, lexA uvrA, and recAI uvr+ strains, but not the resistance of recA1 uvrA strains. We have studied the effect of the tsl-1 mutation on postreplication repair and find that the enhanced survival conferred by this mutation is correlated with an increased capacity for postreplication repair.     

Phenotypes of lexA mutations in Salmonella enterica: evidence for a lethal lexA null phenotype due to the Fels-2 prophage

The LexA protein of Escherichia coli represses the damage-inducible SOS regulon, which includes genes for repair of DNA. Surprisingly, lexA null mutations in Salmonella enterica are lethal even with a sulA mutation, which corrects lexA lethality in E. coli. Nine suppressors of lethality isolated in a sulA mutant of S. enterica had lost the Fels-2 prophage, and seven of these (which grew better) had also lost the Gifsy-1 and Gifsy-2 prophages. All three phage genomes included a homologue of the tum gene of coliphage 186, which encodes a LexA-repressed cI antirepressor. The tum homologue of Fels-2 was responsible for lexA lethality and had a LexA-repressed promoter. This basis of lexA lethality was unexpected because the four prophages of S. enterica LT2 are not strongly UV inducible and do not sensitize strains to UV killing. In S. enterica, lexA(Ind(-)) mutants have the same phenotypes as their E. coli counterparts. Although lexA null mutants express their error-prone DinB polymerase constitutively, they are not mutators in either S. enterica or E. coli.     

Involvement of separate pathways in the repair of mutational and lethal lesions induced by a monofunctional sulfur mustard.

The mutagenic and lethal effects of a monofunctional sulfur mustard, 2-chloro-ethylethylsulfide (CEES), have been studied in a number of repair deficient variants of Escherichia coli K12, B/r and B. The results indicate that CEES induces a (pre)mutational lesion which is subject to Uvr+-excision-repair. Extensive CEES-induced mutagenesis can occur in exrA- uvrA- and recA- uvrB- variants suggesting that the majority of the mutations in Uvr-bacteria do not arise from error-prone repair. These findings are similar to results previously reported with a volatile degradation product of captan and with ethyl methanesulfonate (EMS) but differ from those reported with methyl methanesulfonate (MMS). It is hypothesized that CEES alkylates guanine at the O-6 position (R-O-6-G) and that this R-O-6-G which is Uvr+-excisable is directly mutagenic by producing G-C to A-T transitions during replication. Reduced levels of induced mutation frequencies observed in an endonuclease II-deficient variant lead us to postulate that, in constrast to Uvr- bacteria, CEES-induced mutation in wild-type cells arise from error-prone repair of apurinic sites. Analysis of the lethal actions of CEES indicates that the lesion produced is largely unexcisable by the Uvr+ system. Host-cell reactivation of CEES-treated TI bacteriophage shows that the production of the (pre)ethal lesion is dependent on both the initial dose and post-treatment incubation. The efficient repair of the (pre)ethal lesion requires both endonuclease II and polymerase I. Moreover, deficiencies of these two enzymes rendered bacteria more sensitive to the cytotoxic action of CEES. It is postulated that the lethal mechanism of CEES involves: (I) alkylation at the N-3 position of adenine and the N-7 position of guanine; (2) spontaneous depurination of these alkylated bases; and (3) production of apurinic sites which are lethal unless repaired by the endonuclease II-polymerase I excision-repair system.     

Mutations in Haemophilus influenzae mismatch repair genes increase mutation rates of dinucleotide repeat tracts but not dinucleotide repeat-driven pilin phase variation rates

High-frequency, reversible switches in expression of surface antigens, referred to as phase variation (PV), are characteristic of Haemophilus influenzae. PV enables this bacterial species, an obligate commensal and pathogen of the human upper respiratory tract, to adapt to changes in the host environment. Phase-variable hemagglutinating pili are expressed by many H. influenzae isolates. PV involves alterations in the number of 5' TA repeats located between the -10 and -35 promoter elements of the overlapping, divergently orientated promoters of hifA and hifBCDE, whose products mediate biosynthesis and assembly of pili. Dinucleotide repeat tracts are destabilized by mismatch repair (MMR) mutations in Escherichia coli. The influence of mutations in MMR genes of H. influenzae strain Rd on dinucleotide repeat-mediated PV rates was investigated by using reporter constructs containing 20 5' AT repeats. Mutations in mutS, mutL, and mutH elevated rates approximately 30-fold, while rates in dam and uvrD mutants were increased 14- and 3-fold, respectively. PV rates of constructs containing 10 to 12 5' AT repeats were significantly elevated in mutS mutants of H. influenzae strains Rd and Eagan. An intact hif locus was found in 14 and 12% of representative nontypeable H. influenzae isolates associated with either otitis media or carriage, respectively. Nine or more tandem 5' TA repeats were present in the promoter region. Surprisingly, inactivation of mutS in two serotype b H. influenzae strains did not alter pilin PV rates. Thus, although functionally analogous to the E. coli MMR pathway and active on dinucleotide repeat tracts, defects in H. influenzae MMR do not affect 5' TA-mediated pilin PV.     

Comparison of the Escherichia coli umu+-encoded function with plasmid R46-mediated error-prone repair in DNA-damaged cells

Escherichia coli strain TK701 umu+ was more resistant than strain TK702 umu when tested against bleomycin (BLM), cis-platinum(II) diamminodichloride (PDD), ultraviolet light and methyl methanesulphonate (MMS), which produce single-strand DNA damage. However, the umu mutant was no more sensitive to mitomycin C (MTC) or proflavine (PF), which cause double-strand DNA binding. Strain TK702 umu was nonmutable by any of the agents, whereas mutations were induced in the wild-type strain by PDD, UV, MMS and MTC. The E. coli umu+ function therefore mimics plasmid R46-mediated error-prone repair in protecting only against single-strand DNA damage, whilst enhancing mutagenesis by both single- and double-strand damaging agents. Comparison of plasmid R46-mediated protection and mutagenesis in umu+ and umu strains indicated that the plasmid confers a greater error-prone DNA-repair activity in the mutant. Results are discussed in terms of analogy between host umu+ and plasmid muc+ functions.     

Absence of interaction between X-rays and UV light in inducing ouabain- and thioguanine-resistant mutants in Chinese hamster cells

Chinese hamster ovary cells were irradiated with X-rays at times from 0 to 17 h before being irradiated with ultraviolet (UV) light. No synergism was observed between the two radiations for the production of mutants resistant to either ouabain or 6-thioguanine. These experiments were designed to test whether X-rays induced an error-prone repair system that would increase the frequency of mutations produced by UV light, but no such system was detected.     

Mutagenic DNA repair in escherichia coli. V. Mutation frequency decline and error-free post-replication repair in an excision-proficient strain.

Mutation frequency decline (MFD) is an irreversible loss of newly-induced suppressor mutations occurring in excision-proficient Escherichia coli during a short period of incubation in minimal medium before plating on broth- or Casamino acids-enriched selective agar. It is known that MFD of UV-induced mutations may occur before DNA containing pre-mutagenic lesions is replicated, but we conclude that MFD can also occur after the damaged DNA has been replicated on the basis of the following evidence. (1) Mutation fixation in rich medium (i.e., loss of susceptibility to mutation frequency decline) with ethyl methanesulphonate mutagenesis begins immediately, whereas with UV it is delayed for 20--30 min. (2) The delay in mutation fixation after UV can be explained neither by inhibition of DNA replication nor by a delay in the appearance of error-prone repair activity in the irradiated population. (3) MFD at later times after UV irradiation is more rapid and is less strongly inhibited by caffeine than is MFD immediately after irradiation. (4) Excision is virtually complete 20 min after 3 J m-2 UV but at that time virtually all mutations are still susceptible to MFD. We have presented evidence elsewhere that in bacteria there is an alternative error-free excision-dependent type of post-replication repair of potentially mutagenic daughter strand gaps. We suggest that this process is inhibited at tRNA loci in the presence of nutrient broth or Casamino acids, possibly because of a broth-dependent change in the structure of the single-stranded region including the tRNA locus.     

Repair of UV-induced damage in wild-type and mutant strains of Schizophyllum commune

The basidiomycete fungus Schizophyllum commune was found to have both photo-repair and dark-repair systems for UV-induced damage. Three UV-sensitive mutants were isolated and characterized for ability to repair UV-induced damage in light and dark, and for cross-sensitivity to caffeine and methyl methanesulfonate. Two of the mutants were damaged, to different extents, in their capacity for excision repair; one of these mutants was also probably damaged in post-replication repair. The third mutant was damaged only in post-replication repair.     

Pathways of mutagenesis and repair in Escherichia coli exposed to low levels of simple alkylating agents.

Mutagenesis by simple alkylating agents is thought to occur by either a lexA+-dependent process called error-prone repair or a lex-independent process often attributed to mispairing during replication. We show here that error-prone repair is responsible for the majority of mutants formed after a large dose of alkylating agent, but it is unlikely that it contributes significantly to mutagenesis during exposure to low concentrations of these chemicals. The mutagenicity of these low doses of alkylating agent is reduced by a repair system constitutively present in lexA+ cells but absent in lexA mutants. This system reduces mutagenesis until a second error-free system, called the adaptive responses, can be induced [P. Jeggo, M. Defais, L. Samson, and P. Schendel, Mol. Gen. Genet, 157:1-9, 1977; L. Samson and J. Cairns, Nature (London) 267:281-283, 1977]. The adaptive response is capable of dealing with a much larger amount of alkylation damage than the constitutive system and, when induced, appears to be able to reduce mutagenesis by both decreasing the number of sites available for mutagenesis and delaying the induction of error-prone repair enzymes. Finally, we discuss a model of chemically induced mutagenesis based on these findings which maintains that the observed mutation frequency is dependent on a "race" between these two error-free systems and the two mutagenic pathways.     

Distribution and evolutionary significance of mitochondrial plasmids in Neurospora spp.

The lethal and mutagenic effects of various mutagens on Neisseria gonorrhoeae were investigated. Lethality studies demonstrated that N. gonorrhoeae was relatively sensitive to ethyl methanesulfonate, UV light, and methyl methanesulfonate. Although N. gonorrhoeae was readily mutated by ethyl methanesulfonate and N-methyl-N'-nitro-N-nitrosoguanidine for the three genetic markers assayed, no increase in the mutation frequency was observed for any of the selective markers after UV irradiation or methyl methanesulfonate treatment. These results suggest that N. gonorrhoeae lacks an error-prone repair mechanism.     

Two Lactococcus lactis genes, including lacX, cooperate to trigger an SOS response in a recA-negative background.

A 4.3-kb EcoRI fragment from a Lactococcus lactis genomic library alleviates the methyl methanesulfonate, mitomycin C, and UV sensitivities of an Escherichia coli recA mutant (M. Novel, X. F. Huang, and G. Novel, FEMS Microbiol. Lett. 72:309-314, 1990). It complements recA1 and delta recA mutations but not recA13. Three proteins (with molecular masses of 20, 35, and 23 kDa) were produced from this fragment in a T7-directed system, and three corresponding genes were detected by DNA sequencing, namely, ISS1CH;lacX, which is the distal gene of the lac operon; and a third open reading frame, named lacN, which encodes 211 amino acids. Mutations produced in either lacX or in lacN resulted in the loss of the resistance to DNA-damaging agents. Thus, these two genes appeared to be involved in this activity. Introduction of pUCB214 carrying the 4.3-kb fragment into a lexA+ delta recA306 sfiA::lacZ strain resulted in UV-inducible synthesis of beta-galactosidase. A uvrA strain or a lexA (Ind-) strain containing pUCB214 did not support any DNA repair. However, a lexA (Def-) strain carrying pUCB214 could partly repair UV damage. We discuss possible targets for LacX and LacN products, and we speculate that LacX and LacN may constitute a two-component regulatory system that is able to respond to SOS signals, and then to act in the SOS response, bypassing the RecA-activated function.