In vivo deamination of cytosine-containing cyclobutane pyrimidine dimers in E. coli: a feasible part of UV-mutagenesis.

We have estimated in vivo deamination rates for cytosines in cyclobutane pyrimidine dimers (CPD or PyPy) in UV-irradiated E. coli deficient in uracil DNA glycosylase. The protocol consisted of UV-irradiation, holding in buffer to allow for deamination of cytosines in CPDs and photoreversal (PR) to establish uracils where cytosines in CPD deaminated. The deamination rate at TC photoproducts targeting glutamine tRNA suppressor mutations was estimated from the increase in the mutation frequency after PR (MF(PR)) that developed as UV-irradiated cells were held before PR. Evidence suggested that an earlier study with this protocol under-estimated the deamination rate at sites producing the same mutations in an E. coli B/r strain. With a K12 strain, where the targeting apparently is principally by CPD and not (6-4) photoproducts, a larger rate of k = 0.0091 min(-1) at 42 degrees C resulted. The dark assay for MF also increased significantly with time for deamination consistent with a model for efficient mutation by translesion synthesis at uracil-containing CPD. In addition, we used a strain constructed by Cupples and Miller in which beta-galactosidase was inactive because -GGG- was at codon 461 and would revert to Lac(+) only when replaced by -GAG- or -GAA- for glutamate. CC photoproducts at this target site in the opposite DNA strand could reveal effects of first and second deaminations in the same CPD. MF(PR) for Lac(+) mutations increased and then decreased as a function of deamination time (at six temperatures 36-48 degrees C). Fitting an approximate model equation that distinguished two different deamination rates to these data suggested a first deamination producing Lac(+) at a rate about eight-fold less than a second deamination restoring the Lac(-) phenotype. We conclude that deamination, changing a cytosine-containing CPD to a uracil-containing CPD, could be an integral part of UV-induced C-to-T mutations.

DNA damage-processing in E. coli: on-going protein synthesis is required for fixation of UV-induced lethality and mutation.

UV irradiation of E. coli produces photoproducts in the DNA genome. In consequence, some bacteria lose viability (colony-forming ability) or remain viable as mutant cells. However, the end-points of viability inactivation (lethality) or mutation are determined by cellular processes that act on the UV-damaged DNA. We have investigated the in vivo time course for processes that deal with cyclobutane pyrimidine dimers (CPD) which can be specifically removed by photoreactivation (PR). At different times during post-UV incubation, samples were challenged with PR and assayed for viability or mutation. We used excision-defective E. coli B/r cells and worked under yellow light to avoid background PR. During post-UV incubation (0-100min) in fully supplemented defined medium, inactivation and mutation were initially significantly reversed by PR but the extent of this reversal decreased during continued incubation defining "fixation" of lethality or mutation, respectively. In contrast, if protein synthesis was restricted during the post-UV incubation, no fixation developed. When chloramphenicol was added to inhibit protein synthesis after 30min of supplemented post-UV incubation, at a time sufficient for expression of UV-induced protein(s), fixation of lethality or mutation was still annulled (no change in the effectiveness of PR developed). Lethality fixation did progress when protein synthesis was restricted and the cells were incubated in the presence of puromycin or were either clpP or clpX defective. We discuss these and related results to suggest (1) on-going protein synthesis is required in the fixation process for lethality and mutation to sustain an effective level of a hypothetical protein sensitive to ClpXP proteolysis and (2) this protein plays a critical role in the process leading to exchange between Pol III activity and alternative polymerase activities required as each cell deals with damage in template DNA.

Transcription-modulated repair in Escherichia coli evident with UV-induced mutation spectra in supF.

We have determined several mutation spectra with the supF sequence after UV mutagenesis in Escherichia coli. The cells were either mfd(+) or mfd(-) and grown in defined or complex medium. The tRNA supF gene was expressed from the plasmid pZ189 or pLS1D (similar to pLS189, a variant of pZ189, but with a tac promoter for supF). Most of the mutations with either plasmid could be attributed to possible targeting photoproducts at dipyrimidine sites in the transcribed (TS) or non-transcribed (NTS) DNA strand with differential characteristics relevant to the repair process "mutation frequency decline" (MFD): (1) with pZ189, targeting sites in TS were favored over sites in NTS in all conditions except after an explicit MFD incubation with mfd(+) cells, when there was a majority in NTS; (2) with pLS1D (tac promoter), there was always a marked bias for targeting sites in TS and this was not altered by an MFD incubation; and (3) with pLS1D, spectra with mfd(-) cells vis-à-vis wild-type indicated a notable shift in the position of a hot-spot (both targeting sites in TS) and an increase in deletion mutations. The results support the Selby-Sancar idea that transcription-coupled nucleotide excision repair (TCR) at tRNA genes accounts for MFD and can be inhibited by rapid transcription. During interference of TCR by rapid transcription, however, the presence or absence of functional Mfd protein (transcription-repair coupling factor) can still influence the pattern of mutation, e.g. alter the position of a hot-spot in pLS1D. Only when a tRNA promoter is modulated by an MFD condition is transcription at a rate conducive to TCR. There were several deletion mutations with pLS1D between direct repeats (not present in pZ189) and a model for their production by UV damage is suggested. The spectra with pZ189 in E. coli had similarities with those published for UV mutagenesis in human cells, e.g. mutations at positions approximately 124 and 156.

Transcriptional mutagenesis and DNA strand asymmetrical mutations expressed in Escherichia coli under restrictive metabolic conditions.

Mutagenesis by C-to-U events specifically in either DNA strand assayed with ung uvrA defective Escherichia coli on a metabolically restrictive medium produces more glutamine tRNA suppressor mutations from U occurring in the non-transcribed DNA strand than from U in the transcribed (template) DNA strand. This bias is the reverse of what might be expected from transcriptional mutagenesis (mutation expression utilizing mutated RNA transcribed from damaged template strand DNA). The results and related ideas are discussed.

Photoreversal of UV-potentiated glutamine tRNA suppressor mutations in excision proficient Escherichia coli.

UV-irradiated excision proficient Escherichia coli were exposed to light for photoenzymatic reversal (PR) of cyclobutane pyrimidine dimers (CPD) and assayed for reversion mutation (glutamine tRNA suppressor mutations) on semi-enriched medium or on the same medium containing acriflavine to inhibit excision repair. The initial mutation frequency without PR was relatively greater when assayed with acriflavine, and this difference increased as larger UV fluences were used. The PR kinetics were first order and about the same or slightly faster when cells were assayed with acriflavine (after 15, 30 or 45 J/m2, respectively). The results indicated mutation targeting by CPD in excision proficient cells. These results and conclusion contrast sharply with the original study of this type done several years ago. PR kinetics were considerably slower with assays containing acriflavine, sustaining the idea that PR causes repair of non-dimer targeting lesions by enhancing excision repair. To explain this contrast we devised a fluence-decrement rate for estimating the effectiveness of PR and measured PR-dependent excision repair (PER) as the difference in the fluence-decrement rate with excision proficient and deficient cells. PER was more evident when cells were prepared as in the original study but was still an insufficient factor. More importantly, the original study included a component of indirect photoreactivation or photoprotection (using unfiltered PR light) which accentuated the role of excision repair. Taking these factors into account, the original data also are consistent with the model that glutamine tRNA suppressor mutations produced by UV-mutagenesis in excision proficient E. coli result from targeting by CPD just as in excision defective cells. Thus, with regards to a common UV mutation assay, there does not appear to be two types of targeting lesion depending on excision proficiency.

Mutation frequency decline in Escherichia coli. I. Effects of defects in mismatch repair.

Mutation frequency decline (MFD) in Escherichia coli was examined for effects associated with genetic defects in mismatch repair. The kinetics of MFD are slower when the B/r strain WU3610 carries the mutation mutS201::Tn5 or mutL::Tn10, both of which affect mismatch repair. Similar slow kinetics are produced by mutH34 but not by mutH471::Tn5; the latter has no apparent effect. Strain WU3610-45 (mfd-1) produces the slower kinetics if transcription is inhibited during the post-UV incubation, although it produces no decline in normal circumstances. The slower kinetics are therefore attributed to bulk excision repair that remains when rapid transcription-coupled repair (TCR) is eliminated by certain defects in mismatch repair. A model is proposed wherein mismatch repair defects are thought to slow the activity of TCR but, unlike an mfd defect, not to impede dissociation of stalled transcription complexes at lesions in the transcribed DNA strand.

Mutation frequency decline in Escherichia coli. II. Kinetics support the involvement of transcription-coupled excision repair.

Mutation frequency decline (MFD) in Escherichia coli was examined to demonstrate repair of targeting photoproducts during the post-UV incubation required in this process. Repair of mutation-targeting cyclobutane pyrimidine dimers (T < > C) was demonstrated when a correlation was established between the mutation frequency normally associated with these lesions and the rate of mutation production at these lesions by spontaneous deamination of cytosines and photoreversal in ung-defective cells. An incubation producing a decline in mutation frequency, i.e., MFD, also produces lower rates of mutation increase via the deamination mechanism. Since the latter assay involves processes entirely within the post-UV incubation period, the lower rates are attributed to rapid transcription-coupled nucleotide excision repair (TCR) that reduces the number of relevant T < > C dimers during this period. Rediscovery of the neglected fact that MFD can be stimulated by post-UV incubation in buffer alone is part of the analysis. Results presented here and a variety of others are discussed to support a model of MFD as a particular example of TCR: effective repair of photoproducts in the transcribed DNA strand that target glutamine tRNA suppressor mutations occurs during the appropriate post-UV incubation and is responsible for MFD.

Benefit of transcription-coupled nucleotide excision repair for gene expression in u.v.-damaged Escherichia coli.

Expression of the lactose operon upon induction by IPTG was studied with Escherichia coli B/r and K-12 strains as a function of exposure to ultraviolet light. Patterns of expression inactivation were compared in cells with wild-type UvrABC nucleotide excision repair, with transcription-coupled excision repair (TCR) specifically defective because of a defect at mfd, or with excision repair (ER) and TCR eliminated by defects at uvrA or uvrC. Sets of inactivation patterns were also determined for cells expressing the lactose operon via the "UV5' promoter, an alternative to the wild-type promoter that eliminates dependence of expression on negative DNA supercoiling. The results demonstrated a major contribution by TCR to successful gene expression. Gene expression was more sensitive to u.v. inactivation when TCR was defective and similarly more sensitive when both ER and TCR were defective. Thus, TCR may be the only means of repairing transcription-blocking damage at active genes. Contrasting results with wild-type and UV5 promoters suggested that relaxed supercoiling might accompany repair and reduce expression even though a template lesion is removed. A test of mismatch repair defects on ultraviolet inactivation of gene expression found only limited interference with TCR as it benefits gene expression.

Photolyase-dimer-DNA complexes and exclusion stimulation in Escherichia coli: depolarization of the plasma membrane.

Using cells that overproduce DNA photolyase, we found that UV irradiation (3 J/m2) efficiently inactivates accumulation of methylthiogalactoside (TMG) when RexAB proteins of phage lambda are present. The effect requires both formation of photolyase-dimer-DNA (PDD) complexes and expression of the RexAB proteins. It is reversed completely by a flash of visible light if given immediately after UV and becomes irreversible after post-UV incubation for about 15 min. Inactivation is significant after only 5 min of post-UV incubation, is accompanied by a loss of previously accumulated TMG, and does not require de novo protein synthesis. Passive transport of O-nitrophenylgalactoside by inactivated cells is typical of energy-depleted membranes. We suggest that PDD complexes mimic a developmental intermediate of phage superinfection and stimulate formation of the RexB membrane channel recently proposed by others to explain classical "exclusion". This suggestion is supported by additional data showing an inactivation of colony-forming ability by exclusion stimulation and an inability of PDD complexes to inactivate accumulation of TMG if RexB is present in larger relative amounts than RexA (a detail characteristic of exclusion stimulated by phage superinfection).

Chemically altered apurinic sites in phi X174 DNA give increased mutagenesis in SOS-induced E. coli.

Single-strand DNA from bacteriophage phi X174 am3 is treated with mild acid and heat to produce increasing numbers of apurinic sites per molecule. Samples are assayed, either directly or after additional chemical reactions, by electroporation into the recipient E. coli strain HF4714(su-1+). Modified apurinic sites are produced by reactions with O-methyl- or O-benzyl-hydroxylamine, and reduced apurinic sites by reactions with sodium borohydride. Reversion mutation frequencies are significant only if the recipient strain is SOS-induced (by growth after UV irradiation). A simple apurinic site at the target gives rise to mutation (a transversion) with a probability of 0.07, while the modified or reduced apurinic site has a mutagenic efficiency of 0.22-0.27 or 0.29, respectively. The open ring form of deoxyribose may account for the 3-4-fold increased mutagenicity with altered apurinic lesions. Also considered are effects by temperature and cyclobutane pyrimidine dimers on mutagenicity and the relatively invariant survival curves that obtain regardless of chemical alterations at the apurinic sites and/or SOS induction.

Thymine ring saturation and fragmentation products: lesion bypass, misinsertion and implications for mutagenesis.

We have used thymine glycol and dihydrothymine as representative ring saturation products resulting from free-radical interaction with DNA pyrimidines, and urea glycosides and beta-ureidoisobutyric acid (UBA) as models for pyrimidine-ring fragmentation products. We have shown that thymine glycol and the ring-fragmentation products urea and beta-ureidoisobutyric acid, as well as abasic sites, are strong blocks to DNA polymerases in vitro. In contrast, dihydrothymine is not a block to any of the polymerases tested. For thymine glycol, termination sites were observed opposite the putative lesions, whereas for the ring-fragmentation products, the termination sites were primarily one base prior to the lesion. These and other data have suggested that thymine glycol codes for an A, and that a base is stably inserted opposite the damage, whereas when a base is inserted opposite the non-coding lesions, it is removed by the 3-->5 exonuclease activity of DNA polymerase I. Despite their efficiency as blocking lesions, thymine glycol, urea and UBA can be bypassed at low frequency in certain specific sequence contexts. When the model lesions were introduced individually into single-stranded biologically active DNA, we found that thymine glycol, urea, beta-ureidoisobutyric acid, and abasic sites were all lethal lesions having an activation efficiency of 1, whereas dihydrothymine was not. Thus the in vitro studies predicted the in vivo results. When the survival of biologically active single-stranded DNA was examined in UV-induced Escherichia coli cells where the block to replication was released, no increase in survival was observed for DNA containing urea or abasic sites, suggesting inefficient bypass of these lesions. In contrast, beta-ureidoisobutyric acid survival was slightly enhanced, and transfecting DNA containing thymine glycols was significantly reactivated. When mutation induction by unique lesions was measured using f1-K12 hybrid DNA containing an E. coli target gene, thymine glycols and dihydrothymine were found to be inefficient as premutagenic lesions, suggesting that in vivo, as in vitro, they primarily code for A. In contrast, urea and beta-ureidoisobutyric acid were efficient premutagenic lesions, with beta-ureidoisobutyric acid being about 4-5-fold more effective than urea glycosides, which have approximately the same rate of mutation induction as abasic sites from purines. Sequence analysis of the mutations resulting from these ring-fragmentation products shows that the mutations produced are both lesion and sequence context dependent. The possible roles that bypass efficiency and lesion-directed misinsertion might play in mutagenesis are discussed.

RexAB proteins of bacteriophage lambda enhance the effect of photolyase-dimer complexes on lacZ gene expression in Escherichia coli.

Expression of the lacZ gene in Escherichia coli is inactivated by exposure to ultraviolet light (UV). Inactivation is exceptionally effective when cells contain amplified levels of DNA photolyase (which forms complexes with pyrimidine dimers in the absence of light for actual photoreversal) and a lambda prophage. Without amplified photolyase, the lambda prophage or both, inactivation rates are similar and much lower. UV-inactivation of lacZ gene expression in the presence of both amplified photolyase and lambda is even more effective if lambda cI857 is used in place of the wildtype prophage but is wholly unexceptional if the prophage carries defects in the lambda genes rexA or rexB. When Rex AB proteins are provided by expression from a plasmid and the cell also contains amplified photolyase, exceptional inactivation rates again obtain; in fact inactivation is most effective under these conditions. The data are considered to reveal a role for Rex AB proteins, which mediate superinfection exclusion, in the exceptional inactivation of gene expression by photolyase bound to pyrimidine dimers in DNA. Photolyase-dimer complexes may mimic the structure of certain complexes that arise during phage development and thus influence Rex A and/or B proteins, thereby shutting down cell metabolism.

Inactivation of lacZ gene expression by UV light and bound DNA photolyase implies formation of extended complexes in the genomes of specific Escherichia coli strains.

In Escherichia coli strains WU and CS101, UV inactivation of lacZ gene expression is more effective when the cells contain amplified DNA photolyase, and flash photoreactivation (fPR) after 15 min of metabolism does not reverse inactivation by the photolyase-dimer complexes. In other strains, also studied with or without amplified DNA photolyase, there is no differential UV inactivation and fPR reverses inactivation by the complexes regardless of continued metabolism. The irreparable condition in strain WU is not due to dysfunction of photolyase: during post-UV metabolism, fPR still restores viability and dimers are removed from the region of the lac operon. When the wild-type lac promoter is replaced by the UV5 promoter, making expression insensitive to relaxed supercoiling and catabolite repression, inactivation by dimers alone becomes more resistant, i.e. requires higher fluences, but inactivation in WU and CS101 is still exceptionally sensitive to photolyase-dimer complexes. This indicates that dimers external to the wild-type lac operon may inhibit expression by altering supercoiling but that complexes must involve some other mechanism for their special effect in WU and CS101. The exceptionally efficient inactivation and irreparable condition are consistent with the idea that, in two specific laboratory strains, photolyase bound to dimers at a considerable distance from the lac operon may initiate an aggregation of DNA with other cellular molecules that extends to, and inactivates expression from, the operon.

Diverse backmutations at an ochre defect in the tyrA gene sequence of E. coli B/r.

A DNA fragment including most of the tyrA gene from E. coli B/r strain WU (Tyr-, Leu-) was amplified in vitro by polymerase chain reaction. The sequence was determined, first, for essentially all of the fragment to locate an ochre nonsense defect, and second, repeatedly for a region of the fragment from several independent isolates containing backmutations at the ochre codon (spontaneous and UV-induced). There were 20 single base differences in the tyrA gene region from the analogous wild-type E. coli K12 sequence: an ochre codon at amino acid position 161, 18 silent changes (1 at the first codon base and 17 at the third) and one replacement of valine by alanine. Different backmutations at the ochre codon encoded lysine, glutamine, glutamic acid, leucine, cysteine, phenylalanine, serine or tyrosine. The diversities of base substitutions at the ochre codon after UV mutagenesis or after mutagenesis where targeting by dimers was reduced or eliminated (after photoreversal of irradiated cells treated with nalidixic acid to induce SOS functions or after UV mutagenesis of cells containing amplified DNA photolyase) were similar (with two notable exceptions). The overall differences between the gene sequences for E. coli K12 or B/r seemed consistent with the neutral theory of molecular evolution.

DNA photolyase in E. coli: effects on UV mutagenesis by plasmids expressing the phr gene.

Derivatives of an E. coli plasmid pKY33 are described having specific insertions or deletions that effect or do not effect the phr gene (for DNA photolyase) carried in this plasmid. The various plasmids are tested to determine which cause an inhibition of UV mutagenesis producing glutamine tRNA ochre suppressor mutations. The inhibition is found to require a functional phr gene, which substantiates our earlier report that amplified DNA photolyase interferes specifically with a category of mutagenesis involving targeting by a pyrimidine dimer.

On the possible role of cytosine deamination in delayed photoreversal mutagenesis targeted at thymine-cytosine dimers in E. coli.

While delayed photoreversal (PR) mutagenesis has been interpreted as a measure of misincorporation step in targeted mutagenesis, the specificity to produce glutamine tRNA suppressor mutations (C to T transitions) at sites in DNA where a thymine-cytosine dimer (T = C) may target mutation suggests a deamination model: deamination T = C to T = U and trans-U DNA replication after PR. We describe here two enquires that did not support the latter model: (a) Uracil DNA glycosylase activity as estimated from the restricted plating efficiency of phage T5 containing uracil-substituted DNA showed no variation that might allow an exceptional opportunity for mutation at U in DNA, and (b). The kinetics of delayed PR mutagenesis were unaltered if UV-irradiated cells were held in buffer suspension for 2 h at 41 degrees C (a procedure known to allow deamination T = C to T = U) and then assayed. Other results with cells containing both umuC and ung (uracil DNA glycosylase) defects showed the magnitude of T = C deamination sufficient to provide T = U at the critical site of mutation to an extent greater than the mutation frequencies produced by delayed PR mutagenesis, and considerations of the kinetics led to the suggestion that the deamination model could apply if there were an optimum period 30-130 min post-UV for efficient recovery of DNA replication after PR. The results underscored the feasibility of delayed PR mutagenesis by deamination and trans-U replication, but a selection between the two models could not be determined.

An in vivo complex with DNA photolyase blocks UV mutagenesis targeted at a thymine-cytosine dimer in Escherichia coli.

UV mutation frequency responses for two types of Escherichia coli prototrophic mutant were measured. Only the response associated with a mutation targeted by a thymine-cytosine pyrimidine dimer was reduced in the dark in cells with amplified DNA photolyase. This specific reduction is attributed to the interruption of mutational DNA synthesis by a photolyase complex at the targeting dimer.

Anti-mutagenic effect of ultraviolet light on spontaneous tyrosine tRNA ochre suppressor mutations in Escherichia coli.

The numbers of tyrosine tRNA ochre suppressor mutations arising spontaneously or after UV irradiation in different strains of Escherichia coli K12 are considered. The DNA sequence change requisite for this type of mutation would be a transversion at a cytosine between two purines, where pyrimidine-pyrimidine photoproducts could not form. We find that UV mutagenesis does not produce these tyrosine tRNA ochre suppressor mutations. With lexA51 recA441 defective cells, the spontaneous yield of these mutations is elevated and UV irradiation produces a significant decrease in the numbers of this particular mutation. As explanation we suggest that the spontaneous appearance of these mutations reflects mutation at apurinic sites, the efficiency of which is elevated in lexA51 recA441 cells (with derepressed SOS functions and an activated form of RecA protein). The addition of UV damage in the DNA of these cells cannot further stimulate the positive functions that are required for the production of these mutations and are typically associated with UV mutagenesis (induction of SOS functions, activation of RecA protein and introduction of a targeting photoproduct) but apparently can have a negative effect on mutagenesis, hitherto not realized.

UV mutagenesis in E. coli with excision repair initiated by uvrABC or denV gene products.

Mutation frequency responses produced by ultraviolet light are compared in 4 closely related strains of E. coli B/r having the same tyr(Oc) allele and different excision-repair capabilities: uvr+ (excision repair initiated by wild-type UvrABC activity), uvrA (excision repair defective), uvrA/pdenV-7 (excision repair initiated by endonuclease V of bacteriophage T4, DenV activity), and uvr+/pdenV-7 (excision repair initiated by UvrABC and DenV activities). The production of Tyr+ prototrophic mutants is classified into back-mutations and de novo or converted glutamine tRNA suppressor mutations to indicate different mutation events. Cells transformed with the plasmid pdenV-7 require larger exposures than the parent strains to produce comparable mutation frequency responses, indicating that DenV activity can repair mutagenic photoproducts. When damage reduction by UvrABC or DenV is compared for each of the specific categories of mutation, the results are consistent with the idea that pyrimidine dimers infrequently or never target back-mutations of this allele, frequently target the de novo suppressor mutations, and extensively or exclusively target the converted suppressor mutations. This analysis is based on the distinction that UvrABC-initiated excision repair recognizes dimer and non-dimer (pyrimidine (6-4) pyrimidone) photoproducts but that DenV-initiated repair recognizes only pyrimidine dimers.