Antimutagenesis in Microbial Systems · fine a mutation as a stable, heritable change in the DNA....

BACTERIOLOGICAL REvIEws. Mar. 1975, p. 33-53> Copyright X 1975 American Society for Microbiology

Vol. 39, No. 1Printed in U.SA.

Antimutagenesis in Microbial SystemsCOLIN H. CLARKE AND DELBERT M. SHANKEL*

School ofBiological Sciences, University of East Anglia, Norwich, England, and Department of Microbiology,University of Kansas, Lawrence, Kansas 66045*

INTRODUCTION ................................................................. 33PHYSIOLOGICAL ANTIMUTAGENESIS ........ .................................. 34Purine Nucleosides (Adenosine, Guanosine, and Inosine) ..... ...................... 34Caffeine.35Manganous ions ................................................................. 37L-Methionine .................................................................. 38Histidine ................................................................. 38

Spermine and Other Polyamines, Quinacrine, and Acridines ..... .................. 38Phenothiazine Tranquilizers and Dibenzocycloheptene Antidepressants .... ........ 40Coumadin (Warfarin) .............................................................. 40Actinomycin D and Basic Fuchsin................................................... 40Chloramphenicol, 5-Hydroxyuridine, 6-Azauracil, and Pyronin B on UV-InducedMutations ................................................................. 41

Genie Derepression ................................................................ 41GENETIC ANTIMUTAGENESIS..................................................... 42Genetic Background Effects......................................................... 42Repair-Deficient Bacteria........................................................... 42Phage Lambda ................................................................. 44Antimutator Alleles ofPhage T4 Genes 32 and 43 ....... ............................. 44Rev, Umr, and Rad Mutants in S. cerevisiae ....... ................................. 45Radiation-Sensitive Mutants in Other Eukaryotic Microbes ..... ................... 46

CONCLUSIONS ................................................................. 47LITERATURE CITED................ 48

INTRODUCTION

Our reasons for writing this review are three-fold. First, we are currently working in this fieldand believe that this area of study has beencomparatively neglected, yet is intrinsically asinteresting and important as a study of muta-genesis. We hope this review will stimulateothers to solve some of the many remainingproblems in this area. Second, we know of onlytwo reviews dealing with this topic. One of thesedates from 1956 (119); the other, although morerecent, covers only antimutagenic effectsagainst spontaneously arising antibiotic-resist-ant mutants in bacteria (35). Third, one cannotunderstand the whole process of mutagenesis,including the regulation of mutation frequen-cies and mutagen specificities, without consid-ering antimutagenic effects. Of particular inter-est are the interrelationships that are emergingbetween antimutagenic action, repair systems,and other processes such as genetic recombina-tion. Furthermore, environmental an-timutagenesis may turn out to be a powerfultool to control environmental mutagenesis.We shall begin by defining some terms. A

pre-mutation is a lesion in the deoxyribonucleicacid (DNA) which is potentially able to give riseto a mutation. Whether this potential is realized

will depend on various factors such as theactivity of repair systems and successful pas-sage through the intricacies of the other aspectsof the mutational pathway (see below). We de-fine a mutation as a stable, heritable change inthe DNA. This will result from an alteredconfiguration of one or more base pairs. Amutation may exist in a cell that does not yetexhibit mutant phenotype. A mutant is anindividual organism that exhibits a mutantphenotype (i.e., in which mutational expressionhas been achieved). These distinctions are wellillustrated in the case of ultraviolet (UV)-induced mutations to streptomycin resistance.Here pyrimidine dimers are premutational le-sions; mutations in the DNA are subsequentlyformed by alterations in base pair sequences,but phenotypic expression delay, involving celldivisions and synthesis of altered ribosomalproteins, precedes final production of mutantswhich are phenotypically streptomycin resist-ant.

It is essential that we define clearly what wemean by antimutagenesis. True antimutagene-sis will involve any agent or effect that specifi-cally or preferentially reduces the yield of mu-tants. Any effects observed should not bemerely direct consequences of altered survival(e.g., cell death) or dose reduction (i.e., move-

33

on October 8, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

http://mmbr.asm.org/

CLARKE AND SHANKEL

ment back along a normal dose-response curvefor mutation induction). These latter effects wewould call apparent antimutagenesis. For ex-ample, decreases in absolute numbers of mu-tants resulting from cell death are not trueantimutagenesis, and more importantly, nor arereductions in the yield of mutants brought aboutby, for example, photoreactivation, if this is apredictable consequence of the observed rise insurvival levels. Naturally, in the case of UVmutagenesis occurring at 100% survival, onecould argue that photoreactivation-induced re-duction in mutational yield was truly antimuta-genic since no increase in survival is possible.Thus, a distinction between true and apparentmutagenesis may sometimes be difficult inpractice. This illustrates the vital importance ofdetermining complete dose-response curves be-fore concluding that any decreases in muta-tional yield which are observed are the result oftrue antimutagenesis (40).Mutations are generated, processed, and ex-

pressed via a complex mutational pathway (5,22, 26). Thus, mutations may arise, or premuta-tional lesions be generated, at least in theory, atthe levels of: (i) reaction between a mutagenand DNA; (ii) the utilization of mutagen-altered precursors or base-analogues in DNAreplication or repair; (iii) through errors in DNAreplication; (iv) through errors in DNA recom-bination; (v) through errors in DNA repair; (vi)indirectly via errors in transcription; (vii) in-directly via errors in translation. Mutationsarising by mechanisms vi and vii do so in-directly via mechanisms iii, iv, or v, that is tosay, the activity of error-prone enzyme mole-cules (98, 107, 150). Just as there is morethan one origin for mutations, so too one shouldexpect antimutagens to have many possiblemodes of action at various stages of the muta-tional pathway. For example, antimutageniceffects could arise through interference withthe reaction of a mutagen with DNA at thelevels of repair, replication, recombination,transcription, translation, phenotypic expres-sion including segregational delay, or, the finalstep in the mutational pathway, the growth of asingle mutant individual into a visible, scorea-ble, mutant clone. We shall not, therefore,confine our discussion of antimutagenesis toagents that are known or believed (often on thebasis of little experimental evidence) to act atany particular step in the mutational pathway.

In this review we shall deal with an-timutagenesis from two main standpoints: first,antimutagenesis at what we shall call the phys-iological level, by which we mean the reductionof mutational yields brought about by addition

of chemicals or alteration of cellular conditions;second, antimutagenesis at what we shall callthe genetic level, by which we mean antimuta-genic effects of alleles in replication genes,repair genes, or the general genetic background.We are well aware that this is essentially adescription of antimutagenesis at an observa-tional level and that well-proven biochemicalexplanations for either genetic or physiologicalantimutagenesis are presently exceptionallyrare. If our review serves to stimulate ex-perimentation into the biochemical bases ofantimutagenesis, it will, we believe, have justi-fied its existence.One of the main points we shall try to bring

out throughout this review is the occurrence ofantimutagenic specificity. Just as mutagensmay exhibit specificities with regard to thetypes of lesions they induce (transitions, trans-versions, frameshifts, deletions, hot-spot pat-terns within single genes, or differential muta-bility of various genes), so too are antimutagensoften observed to affect specific classes of muta-tions or to antagonize certain mutagens. Theanalysis of the modes of antimutagenic actionand specificity bids fair to become one of themore fascinating aspects of mutation research'in the next decade.

PHYSIOLOGICAL ANTIMUTAGENESIS

Purine Nucleosides (Adenosine, Guanosine,and Inosine)

The purine nucleosides adenosine, guanosine,and inosine were the first microbial antimuta-gens to be discovered (121). In these and subse-quent studies with tryptophan-limited Esche-richia coli B and B/r cells growing at relativelyslow rates in chemostats (119), forward muta-tions to phage T5 or T6 resistance were scored.Under these (but not other) conditions (88),mutation rates are proportional to time but notto the replication rate. Against the purine-induced, replication-independent mutationsstudied in these experiments, purine nucleo-sides are particularly effective antimutagens.Adenosine (0.4 ,ug/ml) or guanosine or inosine (2,ug/ml) gave a 50% decrease in the mutation rateinduced by theophylline (150 Ag/ml). Adenosineat 50 ,ug/ml completely abolished the mutagen-icities of the purine mutagens caffeine, theo-phylline, paraxanthine, theobromine, azagua-nine, 8-methoxycaffeine, and adenine (at 150,g/ml). In contrast, even at the higher concen-tration of 500 ;tg/ml, adenosine caused only apartial abolition of spontaneous mutations andof the mutagenicities of tetramethyluric acidand benzimidazole. Adenosine had no an-

34 BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from


ANTIMUTAGENESIS IN MICROBIAL SYSTEMS

timutagenic effect on UV- or gamma-inducedmutations in tryptophan-limited cells.

In contrast to the above results, tryptophan-limited cells in a rich medium, at short genera-tion times in chemostats, show both replication-dependent and replication-independent sponta-neous mutagenesis (52, 88). Adenosine orgrowth under anaerobic conditions (119) is re-ported to have no antimutagenic effect on thereplication-dependent spontaneous mutabilityin such tryptophan-limited cells.

Further confirmation of the antimutagenicityof adenosine was provided by Glass and Novick(57). They showed that adenosine (100 ,gg/ml)completely abolished the mutability by caffeineat 200 gg/ml in both growing and chloram-phenicol-inhibited E. coli B/r.The results of Kubitschek and Bendigkeit

(89) and Kubitschek (87) with glucose- or tryp-tophan-limited cells of E. coli in chemostatsshow that purine ribonucleosides are antimuta-genic against spontaneous and 2-aminopurine-induced mutagenesis (scoring T5-resistant mu-tants). There was, however, no reduction inUV-induced mutagenesis, suggesting that pu-rine ribonucleosides act only against mutationsarising duringDNA replication. However, in theabsence of additional information regarding themolecular nature of T5-resistant mutationsarising spontaneously or induced by 2-aminopu-rine or UV light, such a conclusion is necessarilytentative. Moreover, Nestmann (117), usingglucose-limited E. coli B/r cells in chemostats,where mutation rates are solely replicationdependent (88), showed that guanosine (50sg/ml) caused a 36% reduction in spontaneousand mutator Hi-induced T5-resistant mutationfrequencies.

In the absence of knowledge about the molec-ular nature of the T5 Rand T6R mutants studiedand of the comparative mechanisms of replica-tion-dependent and -independent mutagenesis,it is very difficult to speculate meaningfullyabout purine nucleoside antimutagenesis. Itmay be significant, however, that adenosine is amore effective antimutagen than guanosine orinosine at the same molar concentration (119).One is tempted to believe that replication-dependent mutagenesis probably solely reflectserrors in DNA replication and that replication-independent mutagenesis may reflect the activ-ity of error-prone repair systems, perhaps actingupon spontaneously arising single-strand nicks.It would be extremely interesting to determinethe effects of known repair deficiencies, e.g.,Uvr-, Rec-, Lex- (Exr-), Pol-, upon replica-tion-dependent mutagenesis, replication-independent spontaneous mutagenesis, purine-

induced mutagenesis, and purine nucleosideantimutagenesis in glucose- and tryptophan-limited chemostat populations, respectively.One very speculative possibility is that purine

nucleosides act to prevent the induction of anerror-prone, and hence mutagenic, repair sys-tem which is induced in response to lesions inthe DNA. Thus, Witkin (165) has recentlyreported a special case of antimutagenesis bycytidine plus guanosine (100 ,ug/ml each) or bypantoyl lactone. Cytidine plus guanosine (orpantoyl lactone) prevented the enhanced UVmutability exhibited at low UV dosages in a tif-E. coli strain at 42 C. In this strain a set offunctions is induced at 42 C (14, 83).

CaffeineCaffeine is not only a mutagen (57, 120) and

repair inhibitor (108, 157) under some condi-tions in some bacterial systems, but also, underother circumstances, an antimutagen. Griggand Stuckey (64) studied spontaneously arisingHis- _ His+ revertants in an E. coli 15 auxo-troph when cells were held in liquid minimalmedium over extended time periods. Unfortu-nately, the molecular nature of this particularHis- mutant and the His+ revertants is notknown, i.e., whether base-pair substitutions orframe-shifts, and their suppressors, were in-volved. In such nonreplicating stationary-phasecells caffeine, at a concentration of 1,150 Ag/ml,caused a greater than 90% reduction in thespontaneous rate of His- -_ His+ mutation.However, this same concentration of caffeinehad no effect on the phenotypically similar,although not necessarily genetically identical,His+ mutations arising during growth and repli-cation of such His- cells. These differences incaffeine effect possibly could be due to repres-sion-derepression differences at the his or sup-pressor loci between stationary and growingcells. Alternatively, and more likely, differentmechanisms of mutagenesis might operate instationary and growing cells and, hence, pre-sumably on nonreplicating or replicating DNA.Of particular interest would be a comparison ofcaffeine antimutagenesis in this stationary-phase system using pairs of otherwise isogenicstrains, differing only in single repair functionsspecified by exr, pol, rec, and uvr loci. It may besignificant that, whereas caffeine concentra-tions of 350 to 500 ,ug/ml cause maximal inhibi-tion of excision repair (139, 143), in their studyGrigg and Stuckey (64) observed that caffeineantimutagenesis was totally absent at a concen-tration of 260 ,ug/ml. Thus, it may be some typeof error-prone repair, other than normal exci-sion repair, that is causing stationary-phase

VOL. 39, 1975 35


http://mm

br.asm.org/

Dow

nloaded from


CLARKE AND SHANKEL

mutations and is inhibited by higher concentra-tions of caffeine.

Caffeine antimutagenesis against UV-induced mutations in various E. coli strains hasbeen observed in several studies. Horneck-Wittand Kaplan (77) showed that addition of caf-feine, at a concentration of 1,000 ,ug/ml orhigher, to the post-irradiation broth incubationmedium caused a decrease in the yield ofUV-induced mutations conferring low-level (3,gg/ml) resistance to streptomycin. These exper-iments were performed in an E. coli B deriva-tive, B/phr-/MC2.

Clarke (21) using UV-irradiated, broth-grownstationary-phase cells of an Hcr- (uvrA) deriva-tive of E. coli B/r trp- WWP-2 showed that 500gg of caffeine per ml added to the aminoacid-enriched minimal plating medium causeda small, but real, decrease in the yield ofUV-induced Trp- Trp+ reversions. Many ofthese Trp+ revertants result from ochre sup-pressor mutations. Results similar to thosedescribed above but in which there was scoringof UV-induced mutations to high-level (1,000zg/ml) dihydrostreptomycin resistance (StrR)were obtained by Sideropoulos and Shankel(144) in E. coli B/r WWP-2 Hcr-. Their datarevealed a consistent diminution of inducedmutational yields in the presence of 500 Mug ofcaffeine per ml.Witkin and Farquharson (166) made an ex-

tensive study of caffeine effects upon UV-induced mutational yield in repair-proficientand -deficient strains of E. coli. Mutationsstudied were those from streptomycin sensitiv-ity to resistance (StrR) in an excision repair-defective (Hcr-) strain and mutations fromauxotrophy to prototrophy in both Hcr+ andHcr- strains. Broth-grown late-lag-phase cellswere used in these experiments, and caffeineconcentrations of 1,000 and 2,000 ,ug/ml wereemployed. The high concentration of caffeineconsistently depressed the yield of inducedmutations of both types in all Hcr- strains,particularly at low doses of UV. In an Hcr+strain, caffeine at 2,000 g/iml caused a smallerenhancement of UV-induced Trp- - Trp+ re-versions than did a 1,000 g/ml concentration.This result is consistent with the idea that in anHcr+ strain the enhancing effect of caffeine(resulting from inhibition of excision repair) isto some extent counteracted by an opposingantimutagenic effect of caffeine. Unfortunatelythe generality of this result cannot be assumedsince in their paper Witkin and Farquharsondid not report on results with high concentra-tions of caffeine in an Hcr+ strain, scoringmutations from streptomycin sensitivity to re-

sistance. However, they showed that caffeinecan exert its antimutagenic action on the UV-induced yield of prototrophic revertants in anHcr- strain up to the time when the firstpostirradiation replication occurs. Furthermore,in matings between E. coli K-12 Hfr CslO1 met-str+ strain and three different E. coli B Hcr-StrR auxotrophic strains, a concentration of2,000 Mg of caffeine per ml in the final platingmedium, after 100 min of mating, caused a 50 to70% diminution in recombinant yield. Althoughthese results may well indicate that high con-centrations of caffeine interfere with geneticrecombination, one cannot rule out the possibil-ity that the observed reductions in recombinantyields were due to other causes. For example, inthe above matings, which involve K-12 x Bcrosses, restriction phenomena will certainlyoperate. Thus, these interesting results need tobe confirmed in Hfr x F- crosses involvingstrains that do not differ in restriction-modifi-cation patterns. The fact that in Hcr-Exr-strains there was no enhancement of UV-induced killing by even high concentrations ofcaffeine lead Witkin and Farquharson to postu-late that the effect of caffeine is to inhibit boththe activity and error-proneness (and hencemutagenicity) of the Exr system. They sug-gested that this effect could be either direct orindirect by an inhibition of nuclease action.They further predicted (results of such experi-ments have not yet been reported) that highconcentrations of caffeine should be antimuta-genic in Hcr- strains against gamma-ray andthymineless mutagenesis, since the mutabilityof these agents, as with UV, is supposed todepend on errors in recombinational postrepli-cation repair (12). This topic will be dealt within more detail when the effects of exr and recmutations on mutability are discussed below.With lag- and exponential-phase cells of Sal-

monella typhimurium grown before irradiationin liquid minimal medium plus tryptophan (anessential auxotrophic requirement), Williamsand Clarke (155) were able to demonstrate anantimutagenic effect of caffeine (500 ,g/ml).Caffeine, when added to a post-irradiation plat-ing medium devoid of an amino acid mixture(acid-hydrolyzed casein), reduced the yield ofUV-induced Trp- Trp+ reversions. Recon-struction experiments excluded a selective ac-tion against growth of established Trp+ cellsinto scoreable colonies. Caffeine antimutagene-sis resulted neither when a plating medium sup-plemented with casein hydrolysate was used,nor when cells grown before irradiation underother physiological conditions were employed.

There is evidence that caffeine is antimuta-

36 BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



genic in E. coli not only against UV-inducedmutations, but also against those induced bynitrous acid. In the excision repair-deficientuvrA-trp- E. coli B/r strain WWP-2 Hcr-,Clarke (25) showed that, regardless of whethercomparisons were made at equal survival orequal dose levels, caffeine at 500 Mg/ml had anantimutagenic action. This was in contrast toan enhancing effect of caffeine on the Trp+revertant yield in the Hcr+ excision-proficientstrain. Tentatively this result may indicate aninhibition by caffeine of DNA replication orpostreplication error-prone processes that playa positive role in nitrous acid mutagenesis.There is evidence from transformation andtransfection experiments with Bacillus subtilis(106) and phage KBl-mediated transduction inS. typhimurium (Clarke, unpublished data)indicating that caffeine inhibits recombinationprocesses.

Caffeine, at a concentration of 1,000 gg/ml,has been found to be an antimutagen againstboth spontaneous and 2-aminopurine-inducedTrp- -- Trp+ reversions, resulting from bothback-mutations of an ochre codon and ochresuppressors, in E. coli strains WWP-2 Hcr+ andHcr- and also trpA46 (missense) and trpA96(ochre) revertants (27). Spontaneous reversionfrequencies were decreased approximately 90%,as were 2-aminopurine-induced frequencies. Atconcentrations of 100 to 150 gg/ml, caffeine isan antimutagen against a variety of spontane-ously arising Trp+ revertants developing instationary-phase cells of Trp- ochre, missense,and frameshift mutants of E. coli B/r and K-12(59). The antimutagenic effect of caffeine onTrp- ochre -_ Trp+ revertants was present inboth uvrA+ and uvrA- genetic backgrounds,implying that the antimutagenic mechanism isnot an inhibition of excision repair by caffeine.

In the fission yeast Schizosaccharomycespombe, Clarke (23) showed that 1,000 gsg ofcaffeine per ml in the plating medium had anantimutagenic effect on UV-induced Met-Met' reverse mutations. This antimutageniceffect operated only at lower doses of UV anddid not apply to Met+ reversions induced bynitrous acid or nitrosomethyl urethane (nitroso-methyl ethylcarbamate). Subsequently, Lo-prieno and Schupbach (101) showed that caf-feine, at 2,000 Mg/ml, reduced the yield of UV-and nitrosoguanidine-induced Adn+ -- Adn-and UV-induced His- -b His+ mutations in S.pombe. There was no detectable effect uponspontaneous mutations in the same systems.Intergenic recombination frenquencies betweena his-2 and a his-7 mutant were also reducedabout threefold when the crosses were per-

formed in the presence of 2,000 Mg of caffeine perml. These results and those of Fabre (51) andLoprieno et al. (100) are consistent with, but donot prove, the hypothesis that caffeine inhibitsthe activity of a recombinational repair systemwhose activity is necessary for at least a part ofUV and nitrosoguanidine mutagenesis in S.pombe, but which is not required for spontane-ous, nitrous acid, or nitrosomethyl urethanemutagenesis.Sarachek et al. (131) have reported an appar-

ent case of caffeine antimutagenesis in theyeast Candida albicans. Yields of UV-inducedHis- - His+ reversions were reduced by caffeineunder the special circumstances of incubationat 37 C and high UV doses. The effect was foundin both caffeine-sensitive and -resistant strains,and revertant yields at 37 C, even in the ab-sence of caffeine, were consistently below thoseobtained at 25 C.

In view of the findings (119) that caffeine isone of the most effective mutagens and dark-repair inhibitors (42) of several methylatedpurines tested, it would be extremely interest-ing to compare the relative antimutagenic ac-tivities of a series of methylated purines. If allthree activities of methylated purines-mutagenesis of replicating bacterial DNA(57), repair inhibition (65, 144), and an-timutagenesis-depend on weak binding toDNA (41), then the same rank order for thevarious purines might be expected for the abovethree types of activities. Any deviation from thisexpectation might well provide clues as to amore specific mode of antimutagenic action ofmethylated purines.

Manganous IonsUsing a haploid prototrophic strain of Penicil-

lium chrysogenum, Arditti and Sermonti (3)showed that the presence of 5 mM manganouschloride in the post-treatment minimal platingmedium prevented the appearance of essen-tially all 8-azaguanine-resistant mutants in-duced by treatment with the nitrogen mustardmethyl-bis(,B-chloroethyl)amine. The level ofresistance was to 1.5 mM 8-azaguanine, with adose of the nitrogen mustard giving about 10%survival. Post-treatment incubation of the co-nidia in a complete medium for 6 h or longer at24 C, before plating in 8-azaguanine-containingminimal medium on the presence of MnCl2, ledto a complete disappearance of the MnCl2 effecton mutational yield. The MnCl2 antimutagene-sis did not apply to 8-azaguanine resistancemutations induced by UV light, X rays, diethylsulfate, or DL-p-N-di(chloroethyl)phenylala-nine. Furthermore, there was no inhibitory

VOL. 39, 1975 37


http://mm

br.asm.org/

Dow

nloaded from


CLARKE AND SHANKEL

effect of MnCl2 in the plating medium on theyield of nitrogen mustard-induced mutants con-ferring resistance to 7-azaindole (0.2 to 0.8mM). The above effects of MnCl2 were possiblyrelated, in some unspecified way, to the earlierfindings of Morpurgo and Sermonti (111) thatMnCl2 diminished somatic segregation andhaploidization and increased the survival ofdiploid condidia of P. chrysogenum treated withnitrogen mustard. Bohme (11) has obtainedresults in Proteus mirabilis indicating thatMn2+ ions, in this organism but not in E. coli(where Mn2+ ions act as a mutagen), causeinhibition of the repair of ethyl methane sulfo-nate lethal damage. This is probably the resultof an effect on the Exr dark-repair system (76),although Mn 2+ ions caused an enhanced yield ofethyl methane sulfonate-induced StrR muta-tions in the P. mirabilis system. If so, one mightpredict that pretreatment of P. mirabilis withMn2+ would be antimutagenic against subse-quent UV mutagenesis, since UV mutagenesisseems to be totally dependent on the Exrfunction.

L-Methionine

In the haploid, fission yeast S. pombe Clarke(18-20) showed that L-methionine added to theminimal plating medium decreased the yield ofspontaneous and HNO2- or UV-induced Adn-

Adn+ reverse mutations. The antimutageniceffect was optimal with L-methionine concen-trations of 20 ,ug/ml or higher, specific for Adn+reversions, and most pronounced for those Adn+revertants that were due to suppressor muta-tions. The inhibitory effect of L-methionine onAdn+ reversions was paralleled by inhibition ofthe residual divisions undergone by adn- cellswhen deprived of extraneously supplied ade-nine, by an inhibition of complementation be-tween pairs of adn-1 mutants, and by a decreasein growth of leaky adn- mutants. All of theseeffects were tentatively attributed to accumula-tion of S-adenosylmethionine in adn- cells inthe presence of excess L-methionine. Accordingto this hypothesis, such cells would becomedepleted of adenosine 5'-triphosphate and thusbe unable to express induced mutations. Ofsome interest would be a demonstration thatsimilar antimutagenic effects of L-methionineoccur with adn- mutants of S. cerevisiae, anddirect measurements of intracellular S-adeno-syl-methionine concentrations. However, an ap-parent antimutagenic effect of L-methionine onHNO2-induced fast-growing Leu+ revertants ofa Leu- strain, leu-3, 241 (20), cannot be ex-plained by this postulated mechanism, al-though it is possible that it results from a

contaminant present in commercially availablemethionine.

HistidineQueiroz (126), working with a polyauxo-

trophic strain of Saccharomyces cerevisiae,showed that histidine in the plating mediumhad a specific antimutagenic effect on tyrosine-inserting class I ochre suppressor (super-sup-pressor) mutations of UV-induced origin. Thishistidine effect was relieved by the addition ofadenine and could be explained at a biochemi-cal level by the known interactions involvingAICAR and feedback inhi ition by histidinebetween adenine and histidine-isynthesis(102). This represents one of the very few casesin which a convincing biochemical explanationexists for a case of antimutagenesis.

Spermine and Other Polyamines,Quinacrine, and Acridines

In a series of papers published over the pastdecade Sevag, DeCourcy, and their co-workershave presented evidence that the presence ofspermine, spermidine, quinacrine (atabrine),and acridine derivatives reduce spontaneousmutability to many different types of antibioticresistance in a variety of bacteria (35-38,135-138). These results have, in our opinion,been largely overlooked by geneticists becauseof the strongly Lamarckian overtones of theexplanations provided, including perhaps someconfusion over the distinction between a muta-tional event in the DNA and a mutant cellshowing full phenotypic expression. In addition,most of their experiments were based on turbid-ity measurements in liquid cultures and didnot, therefore, measure quantitatively mutationfrequencies or rates. Since we do not presentlyunderstand the mechanism(s) that generatesspontaneous mutations, and since the possiblerole of repair systems in the antimutageniceffects observed was not examined, it seemsimportant to extend testing of the above-namedcompounds to mutagenic systems other thanspontaneous antibiotic resistance in wild-typebacteria.To some extent this has already been

achieved. Johnson and Bach (80, 81) showedthat spermine tetrahydrochloride, at a concen-tration of 150 mg/ml, caused a small but signifi-cant (two- to fourfold) lowering of spontaneousmutation rates. The systems studied were StrS

StrRin E. coli and Staphylococcus aureus andTrp- Trp+ in E. coli UC707. Furthermore,StrS StrR mutations induced in E. coli bygrowth in the presence of 150 jig of caffeine perml were reduced about 11.5-fold by the presence

38 BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



of spermine. These authors also studied UV-induced Strs - Str-R mutations in cells grownprior to irradiation with or without spermine.With a single UV dose of 150 ergs/mm, givingthe surprisingly low survival level of approxi-mately 1%, they obtained an approximately8.5-fold decrease in mutational yield by pre-growing the cells in spermine. This result,however, deserves reinvestigation since the au-thors did not provide details of the conditionsallowed for phenotypic expression of the UV-induced StrR mutations prior to challenge withstreptomycin. Such conditions can dramati-cally alter the mutant numbers. Caffeine- orUV-induced Trpj _ Trp+ reversions also werenot investigated in this particular study.Johnson and Bach demonstrated later (81)

that mutations to streptomycin resistance in-duced in E. coli through the action of theTreffers mutator gene (now known to cause A-T-- C-G transversions during DNA replication)(33) and 2-aminopurine (causing A-T - G-Ctransitions) were susceptible to spermine andquinacrine antimutagenesis. These compoundsat the concentrations used (150 gg/ml forspermine; 2 to 20 sg/ml for quinacrine) causedapproximately 30 and 80% reductions, respec-tively, in induced mutation rates. Spermine at aconcentration of 50 ,gg/ml had no antimutageniceffect. In the published results, details ofspermine antimutagenesis against 2-aminopu-rine-induced StrR mutations are not given.

In a later paper, Zamenhof (171), using E. coliK-12, showed that spermine tetrahydrochloride(150 ig/ml) and quinacrine hydrochloride (6zggml) depressed spontaneous mutation ratesfrom azide sensitivity to azide resistance andMet- - Met+ reversions about fourfold each.However, these two compounds did not reducethe mutability (in the same genetic back-ground) of the ast mutator which had beenshown to cause bidirectional transition muta-genesis. Spermine and quinacrine therefore donot appear to be totally nonspecific antimuta-gens. A detailed knowledge of the mode ofaction of the ast mutator should be helpful ininterpreting the failure of spermine and quina-crine to counteract ast mutagenesis.

Clarke (27) investigated the influence ofspermine (250 Ag/ml) on spontaneous and 2-aminopurine-induced Trp- - Trp+ reversionsin E. coli B/r strain WWP-2 (ochre) in uvrA+and uvrA- genetic backgrounds and in E. coliK-12 trp- strains A-46 (missense) and A-96(ochre). Spermine was not found to be an-timutagenic against spontaneously arisingrevertants in any of the four strains. There was,however, a 25 to 55% reduction in 2-aminopu-

rine-induced Trp+ revertant frequencies, in-volving both back mutations and suppressorrevertants. Subsequently, Godsell and Clarke(59) investigated the influence of spermine, at50 to 200 ig/ml concentrations, on spontaneousTrp- - Trp+ reversions occurring in growingand stationary-phase cells. The Trp- strainsused were in E. coli B/r and K-12 backgroundsand involved ochre, missense, and frameshiftmutations. There was only a weak antimuta-genic effect against frameshift revertants in theK-12 strains and against ochre revertants, in-cluding suppressors, in the B/r strains. In fur-ther tests in this same series of experiments,employing quinacrine at 5 to 30 Ag/ml concen-trations, Godsell and Clarke found an apprecia-ble antimutagenic effect against Trp- frame-shift revertants in stationary-phase cells, butnot against Trp+ reversions of the other Trp-strains. Reconstruction experiments were per-formed and appeared to exclude selectionagainst Trp+ cells as an explanation for theobserved antimutagenesis. Nestmann (117),studying spontaneous and mutator Hi-inducedT5-resistant mutations in glucose-limited cellsof E. coli B/r growing in chemostats, foundadditional evidence for spermine antimutagene-sis. In such cells, where mutations arise in ageneration time-dependent manner, spermine,at a concentration of 50 Ag/ml, caused a 62%reduction in both spontaneous and mutatorHi-induced mutation rates.Nestmann (117) has hypothesized, on the

basis of the available biochemical evidence,that spermine antimutagenesis might operatevia a stimulation of the exonucleolytic (editing,error-correcting) function of a DNA polymerase(16, 60, 70, 114). For example, Chiu and Sung(16) demonstrated that spermidine stronglystimulates the activity of DNA polymerase Bfrom rat brain but that it does not stimulate theactivity of DNA polymerase A from the samesource. To the extent that DNA polymerases areinvolved in repair processes or in error-correct-ing mechanisms, this type of observation mayhelp to explain the antimutagenic activity ofspermidine. Unfortunately, polyamines exertsuch a diversity of biochemical effects (7, 31,147, 149) that it may prove extremely difficultto identify the site(s) of polyamine an-timutagenesis.Of obvious interest in the context of poly-

amine antimutagenesis would be a determina-tion of spontaneous and induced mutation fre-quencies in bacterial strains that are polyamineauxotrophs (103, 104) when such cells are grownat a range of polyamine concentrations. Suchstudies might reveal whether or not naturally

39VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


CLARKE AND SHANKEL

occurring polyamines in wild-type microbialcells act as major regulators of mutagenic proc-esses.Using glucose-limited cells of E. coli B/r/1

trp- growing in a chemostat, Webb and Kubit-schek (154) showed that in the dark acridineorange at concentrations of 10-6 to 5 x 10-1 Mcaused an approximately threefold reduction inthe rate of spontaneous mutations to phage T5resistance. Similarly, mutations induced by 450Ag of caffeine per ml were reduced to about thesame extent by acridine orange. It should benoted that in glucose-limited cultures spontane-ous mutation rates are directly proportional tothe generation time. (This is in contrast to thesituation with slowly growing tryptophan-limited cultures under chemostat conditions[88, 119].)Magni et al. (105), while studying spontane-

ous mutations in haploid strains of S. cerevis-iae, showed that 5-aminoacridine (10lOg/ml)was antimutagenic during vegetative growth.The mutations studied were from canavaninesensitivity to resistance (30-fold reduction) andreversions were from His- to His+ (sixfoldreduction). Unfortunately, no results are re-ported for the effects of similar concentrationsof the agent on spontaneously arising mutationsin mitotically dividing diploid cells. In a furtherstudy, employing the same canavanine resist-ance mutation system, Puglisi (124) comparedthe antimutagenic activity of acridine and threemethyl acridines (at 10 gg/ml) on growinghaploid cells. Whereas acridine had only a weakantimutagenic effect (50% reduction), themethyl acridines caused up to 60-fold reduc-tions in spontaneous mutation rates. An-timutagenesis by the methyl acridines washighly dependent upon conditions of pH andaeration during growth, but the relatively weakeffect of the acridine was not dependent uponthese factors.Ethidium bromide at 2 to 5 ,ug/ml and hycan-

thone methane sulfonate at 20 to 40 ,ug/ml haverecently been shown to act antimutagenically inan excision repair-deficient strain of E. coli (30,141). The mutations studied were UV-inducedmutations to intermediate- or high-level strep-tomycin resistance and Trp- ochre Trp+reversions. The StrR system was chosen sincehycanthone is a frameshift muittagen (30). Itsantimutagenic activity could be revealed sinceframeshift mutations are lethal in the strep-tomycin locus (62, 148).

In view of Riva's (128) demonstration that theframeshift mutagenicity of acridines is not sat-isfactorily explained on the basis of their inter-calating ability, it might be useful to determine

whether acridine antimutagenesis is, likewise,unrelated to intercalation. Since reported casesof acridine and spermine antimutagenesis are inreplicating systems, it may well be that an-timutagenesis results from binding to the DNApolymerase or replication fork complex. Acri-dines and quinacrine are known to be effectiveinhibitors of excision and probably other dark-repair processes (30, 53, 158).

Phenothiazine Tranquilizers andDibenzocycloheptene AntidepressantsAn interesting bridge has been created be-

tween human pharmacology and microbial an-timutagenesis by the findings of Heller andSevag (66) that tranquilizers and antidepres-sants can reduce spontaneous mutation fre-quencies in bacteria. These studies were carriedout with S. aureus and E. coli B and theirmutants resistant to streptomycin, sulfa-thiazole, or chloramphenicol. It will be of con-siderable interest to determine whether thesedrugs act antimutagenically against biochemi-cally defined types of induced bacterial muta-tions and in eukaryotic organisms.

Coumadin (Warfarin)DeCourcy et al. (36) demonstrated that cou-

madin inhibited the occurrence of spontaneousstreptomycin-resistant mutants of S. aureusand E. coli and polymyxin-resistant Pseudomo-nas aeruginosa. These results obviously deserverepetition in genetically identified mutationalsystems involving known types of base-pairchanges, and determination of the dependenceof the antimutagenic action on DNA replica-tion, repair activities, etc. It may be relevantthat coumadin intercalates with DNA.

Actinomycin D and Basic FuchsinPuglisi (125) scored spontaneously arising

forward mutations to canavanine resistance in ahaploid strain of S. cerevisiae growing in mini-mal medium. In the presence of sublethalconcentrations of actinomycin D (16 to 48,g/ml), spontaneous mutation rates were re-duced 2.5- to 30-fold. Repeat experiments gaveresults which agreed within a factor of 2. Like-wise, addition of basic fuchsin (10 to 30 ,ug/ml)to the growth medium caused a marked depres-sion of spontaneous mutation rates to canavan-ine resistance. However, in the two reportedexperiments reproducibility of results was verylow: 10 ,g of basic fuchsin per ml giving in onecase a 36.5-fold decrease, but in the other caseonly a 1.5-fold decrease in mutation rate belowthe spontaneous rate. In the absence of informa-

40 BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



tion regarding the mechanism by which thespontaneous mutations are generated in thissystem and of their molecular nature, anyspeculation as to the mode of action of the twoantimutagens would be premature. However, aspointed out by Puglisi, it is probably significantthat both actinomycin D and basic fuchsin bindstrongly and weakly, respectively, to DNA andcould, therefore, inhibit the production or con-sequences of single-strand breaks in the DNA.

Chloramphenicol, 5-Hydroxyuridine,6-Azauracil, Pyronin B on UV-Induced

MutationsUV-irradiated cells of Salmonella typhimu-

rium and E. coli B/r auxotrophs often, but notinvariably, give markedly higher yields of proto-trophic revertants when plated, immediatelyafter irradiation, on an amino acid-enrichedmedium rather than on a minimal mediumcontaining only a low concentration of all spe-cifically required nutrients (155, 156). Thisenhancement of UV-induced mutational yieldby a nonspecific amino acid mixture, alsoknown as the broth effect, is especially pro-nounced for those mutations to a prototrophicphenotype that result from some classes ofochre or amber suppressor mutations. Thebroth effect is very dependent on the physiologi-cal conditions under which the cells were grownbefore UV irradiation (155). The broth effectwas believed to result from a stimulation ofpost-irradiation protein synthesis (cf. 10). Thisin turn caused UV-induced lesions to be re-moved, not by the error-free excision repairprocess (called in this context mutation fre-quency decline (43, 1691) but instead by anerror-prone, and hence mutagenic, replicationalor post-replicational repair system. Thus, thebroth effect may in fact be a compositephenomenon-an inhibition of excision repair(mutation frequency decline) plus stimulationof an error-prone process (10, 22, 55, 63, 109).Whatever the detailed explanation for the

broth effect on UV-induced mutational yield,the enhancing effect of amino acids is abolishedby chloramphenicol (168), 5-hydroxyuridine(43), 6-azauracil (44), and perhaps by pyronin B(158). That is, these compounds probably exerttheir antimutagenic influence because they areacting as inhibitors of protein synthesis. It is ofinterest to note that 5-hydroxyuridine causesmutation frequency decline in the presence ofamino acids only after an initial delay (43). Onepossibility is that 5-hydroxyuridine must firstbe incorporated into newly synthesized ribonu-cleic acid before it can exert its action andinhibit protein synthesis. Although Witkin

(158) provided good evidence that pyronin Babolishes the broth effect, this was true onlywith some batches of the dye. Clarke (26) wasunable to show any diminution at all of thebroth effect with pyronin B. On the contrary, inthese later experiments pyronin B acted as atypical excision repair inhibitor, causing de-creased survival of UV-irradiated cells and anincrease in mutational yield of a dose-enhance-ment type. Thus, there may well have been aprotein synthesis inhibitor present as a contam-inant in some commercial batches of the dyewhich gave the antimutagenesis (broth effect-abolition) reported by Witkin (158).An obvious extension of these bacterial stud-

ies could be into mutational systems in eukary-otic microorganisms. There is already evidencefrom the work of Sarachek and Bish (130) thattemporary inhibition of protein synthesis afterUV irradiation of yeast cells can have an an-timutagenic effect, at least when one is dealingwith cells in which intracellular amino acidpools have been depleted by prior starvation.Experiments with eukaryotic protein synthesisinhibitors, such as cycloheximide, might wellreveal these agents to exert an antimutageniceffect if applied transiently after, for example,UN irradiation of yeast cells.

Genic DerepressionStates of genic repression or derepression

might be expected to affect observed mutabilityby a variety of mechanisms (99). First, theremight be altered accessibility of the geneticmaterial to direct reactivity with a mutagen.Second, there might be an indirect effect viaaltered accessibility to repair systems (84) andto mutagenic recombinational processes. Conse-quently of interest, as examples of physiologicalantimutagenesis, are several reports thatstates of genic repression and derepressionhave marked effects upon mutability and ge-netic recombination. Brock (13) tested mutabil-ity at the fl-galactosidase locus when the genewas borne on an F' episome of E. coli K-12 inthe presence or absence of the gratuitous in-ducer isopropyl-0-thiogalactoside. Further-more, in some of the experiments, lactose i-(lacI3) mutants were tested which were notexpected to respond to inducer. The resultsshowed that back-mutations within the lac,(lacZ4) locus were much more frequent in thepresence of inducer than in its absence. Thiswas true of back-mutations induced by diethyl-sulfate or nitrosoguanidine, but not for thoseinduced by gamma. irradiation. Although itwas claimed that mutagenesis by 2-amino-purine and 5-bromodeoxyuridine did not re-

VOL. 39, 1975 41


http://mm

br.asm.org/

Dow

nloaded from


CLARKE AND SHANKEL

spond to gene induction, the actual resultsindicate that under the conditions used (non-replicating cells) neither of these base analoguesis mutagenic. Thus, transcription, or its initia-tion, was believed in some way to facilitatemutagenesis by alkylating agents. Conversely,one might say that states of genic repression areantimutagenic.

Similarly, Herman and Dworkin (69) studiedLac-. Lac+ reversions in E. coli K-12, in thiscase those induced by the frameshift mutagenICR-191. They showed that induction of geneexpression and specifically transcription and/ortranslation caused an approximately twofoldstimulation of ICR-induced mutagenesis. Thiswas true for several different Lac - mutants, butone exceptional mutant reverted to Lac+ ap-proximately sevenfold less frequently when in-duced. Thus, in agreement with Brock's find-ings (above), a state of genic repression isusually antimutagenic, except in the one mu-tant described by Martin and Dworkin.

Savici and Kanazir (133), using SalmonellahisC and hisF mutants in the presence orabsence of an additional histidine operator-con-stitutive mutation that causes 15-fold derepres-sion of the operon, showed that UV mutabilitywas under repression-derepression control. His -, His+ revertant frequencies were shown to be

five- to eightfold higher in derepressed than inrepressed genes. Thus, in agreement with theprevious studies in E. coli, here too a state ofgenic repression is antimutagenic. One puzzlingfeature of this study is that a proportion of theHis+ revertants would be expected to be due tosuppressor mutations distant from the His op-eron whose expression was being experimentallymanipulated. Equally importantly, Savici andKanazir showed that the same differential mu-tagenesis is observed in an excision-repair-defective strain. They have therefore suggestedthat states of genic repression-derepression af-fect mutagenesis through altered accessibilityto recombinational repair and not to excisionrepair. This is in line with the finding by Savici(132) in the same Salmonella system, that anoperator-constitutive mutation (i.e., derepres-sion) caused reduced recombination within thehistidine operon. Similarly, Herman (68)showed that gene induction caused locally de-creased recombination within the lactose op-eron of E. coli. However, if this were so, then onewould have to postulate a rather complex(though not necessarily incorrect) hypothesis toexplain why decreased recombination led toincreased mutagenesis. (It should be noted that,in the arabinose operon of E. coli, Helling [67]reported that occurrence of transcription/trans-lation [derepression ] led to increased recombi-

nation.) Moreover, Shestakov and Barbour(142) found no influence of genic derepressionon recombination in the lac genes of E. coliK-12.

GENETIC ANTIMUTAGENESIS

Genetic Background EffectsThere have long been known examples in

microbial systems where alleles in one gene hada profound influence on the mutability of an-other locus (17-19, 58, 168, 172). Some of thesecases have been analyzed, particularly where anantimutagenic effect of a second marker wasevident. For example, Corran (32) showed thatin a His- Thr- strain of B. subtilis the greatlyreduced yield of His+ revertants, compared tothat in the His- Thr+ strain, was due to thephenomenon of "auxotrophic pre-emption."The excess of His- Thr- cells removed threo-nine from the plating medium and thus pre-vented the development of His+ Thr- cells intocolonies. Such an effect could be relieved byexcess threonine. In the case of an apparentantimutagenic effect of a Met- allele uponAdn-* Adn+ reversion frequencies in S.pombe, this was shown by Clarke to be dueexclusively to the necessary addition of L-meth-ionine to the plating medium (18-20). Similarplating medium effects were excluded by Zet-terberg (173) in the case of an antimutageniceffect of a Met- marker on UV-, dimethylsulfate-, X-ray-, and nitrosomethylurethane-induced reversion frequencies of Ura- - Ura+mutations in Ophiostoma multiannulatum, andChopra (17) in the case of an antimutageniceffect of an adn- marker on Trp- - Trp+reversions of spontaneous origin in E. coli B/r.

In the case of an absolute antimutageniceffect of a streptomycin dependence allele uponTrp- - Trp+ reversions induced by UV in E.coli (168), this is probably a consequence ofreduced efficiency of ochre suppressors (respon-sible for the Trp+ phenotype) in the presence ofstreptomycin-dependent ribosomes (62). A sim-ilar but less pronounced antimutagenic effect ofstreptomycin resistance markers upon UV-induced Trp- - Trp+ reversions (8, 28, 146) hasbeen shown to depend on restricted efficiency ofochre suppressors brought about by the ribo-somal mutations conferring streptomycin resist-ance (2, 56, 75, 148). In these cases the revertantyield could be restored to near normality in theStrR (suppressor restrictive) strain by the addi-tion of streptomycin to the plating medium (28,92, 146).

Repair-Deficient BacteriaThere are real difficulties of interpretation in

42 BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



studies with radiation-sensitive microbial mu-tants in which the sensitive allele in found tocause depression or elimination of induced mu-tation frequencies. Such an apparent antimuta-genic effect of a radiation-sensitive allele shouldtants in which the sensitive allele is found toimply that the corresponding wild-type allelenecessarily has a positive role to play in muta-genesis by specifying some repair enzyme thathas an error-prone, and hence mutagenic, rolein some repair process. First, some radiation-sensitive mutations may not be localized instructural loci for repair enzymes. Secondly,absence of mutations may be due to activedestruction of premutational lesions, for exam-ple, through uncontrolled nuclease activity andexcessive DNA degradation, such as occurs inrecA - strains after irradiation. With these cau-tions in mind we shall now outline those caseswhere mutant alleles, often in known or sus-pected repair genes, have been found to exertantimutagenic effects.Witkin (159, 160, 162) has presented good

evidence that Exr- mutants in E. coli B or B/rgenetic backgrounds are nonmutable by UVlight. These are probably equivalent to the Lex-mutations in K-12, which show similar proper-ties (112). Similarly recA- mutants of E. colihave been shown to be totally nonmutable byUV (86, 109, 161). This seems to be true also forgamma rays and thymine deprivation muta-genesis (12). However, both recA - and Exr(Lex) mutants undergo extensive DNA degra-dation after UV irradiation. It is thereforepossible that the reason for the apparentnonmutability of such strains is that "muta-tions" occur only in dead cells, i.e., that survi-vors are actually cells that received no irradia-tion dosage, or at least none that was notrapidly removed by excision repair.

Igali et al. (78) showed that mutability by9-methoxypsoralen plus 365-nm (long wave-length) UV light was abolished in a recA - or anExr- genetic background. The mutations theystudied were Trp-* Trp+ reversions due to amixture of ochre codon back-mutations andochre suppressor mutations. Witkin (161, 164)has presented evidence that recC- and uvrA-recB - or uvrA - recC- strains show reduced, butnot zero, mutability by UV. However Hill andNestmann (74) showed subsequently that Lac+-. Lac- mutations induced by UV are not lessfrequent in a recC- than in a Rec+ strain.Furthermore, they showed that lethal sectoringled to underestimation of other classes of muta-tions in recC- strains. It is important, therefore,to decide whether the reduced yields of UV-induced Trp- - Trp+ and StrS StrR muta-tions reported by Witkin in uvrA- recB- and

uvrA - recC- strains, compared with uvrA - rec +strains, are due to genuine antimutagenesis orto the lethal sectoring artifact. It may besignificant that in Witkin's data the frequenciesof different phenotypic classes of mutations areapparently reduced to different degrees by thepresence of a recC- allele (161). Furthermore,whereas recB73 and recC mutations causedapproximately the same degree of apparentantimutagenesis against UV-induced StrsStr5 mutations, another recB allele, recB21,caused a more pronounced diminution in muta-tional yield (164).A polA - mutation has been stated to be with-

out any effect upon UV mutability (74, 86, 163,167). The Kornberg DNA polymerase specifiedby the polA locus is believed to be an integralpart of the excision repair process. Neverthe-less, this apparent absence of a polA effect onUV mutability depends on how the results areexpressed. If one plots UV-induced mutationfrequencies not against dose but against percentsurvival (i.e., strains are compared at equalsurvival levels), then polA - strains show greatlyreduced UV mutability. (It should be pointedout that comparison of a Uvr- (Hcr-) strainwith a Uvr+ at equivalent doses would show theUvr- to have an apparently greatly enhancedUV mutability (71), whereas when the compari-son is made at equivalent survivals the strainsshow approximately equal mutability (71,159).) If valid, this comparison at equal survivallevels would imply that excision and resynthesisis error prone, whereas the incision step inexcision repair is error free (29, 118). Thisall-important point, as to whether one canvalidly compare mutation frequencies at equiv-alent doses of mutagen in pairs of strainsshowing very different sensitivities to the lethalaction of the mutagen, or whether one shouldrather use comparisons at equivalent survivallevels (after different doses of mutagen to thetwo strains), is not easily resolved (170). Itcannot be overemphasized that these two waysof treating the data can give very differentconclusions regarding the antimutagenic actionof any particular repair-deficient mutant (29).Among factors to be considered in decidingwhich type of comparison to use, or accept, onemight include whether or not mutations andlethal events arise from identical lesions suscep-tible to the same repair systems and whethermutations occur at random in the population oroccur rather in a selected fraction (e.g., survi-vors, or cells receiving more or fewer thanaverage numbers of lesions) of the population(71, 72). In this connection, one may note caseswhere survival after mutagenic treatment is notaffected but mutations are decreased (e.g., T4

VOL. 39, 1975 43


http://mm

br.asm.org/

Dow

nloaded from


CLARKE AND SHANKEL

antimutator alleles and HNO2 or ethyl methanesulfonate mutagenesis) or increased (e.g., UV-induced Tyr-. Tyr+ and Leu- Leu+ rever-sions in E. coli strain 36-10-45 [24]). Further-more, there are cases where survival is de-creased but mutagenesis is unchanged (e.g.,polA- with UV, when compared with pol+ atequivalent doses), decreased (e.g., recA- withnitrosoguanidine), increased (e.g., Uvr- withUV or HNO2, when compared with Uvr+ atequivalent doses), or totally abolished (e.g.,exrA or recA with UV).

In their extensive study of the effect of Uvr-,polA-, and recA- mutations on killing and Arg-

Arg+ mutation induction by a variety ofmutagens, Kondo et al. (85, 86) made all of theircomparisons of mutational yield at equivalentdoses. In Table 1 we compare their results withthose obtained if one uses comparisons at equiv-alent survival levels.

Phage LambdaThere have been a number of reports that

repair deficiencies in the host cell exert anantimutagenic action on lambda phage UVmutagenesis, which is normally dependent onirradiation of both phage and host cell. Thus,Miura and Tomizawa (109, 110) showed thatUV-induced c mutations of X did not occur in

TABLE 1. Comparison of effects of repair genes atequivalent doses or survivalsa

Compari- Effect of repair gene onMutagen son made arg- - Arg+ frequencies

at equiv-alent uvr- polA- recA

UV Dose Increased Same AbolishedaSurvival Same Reduceda Abolisheda

NQO Dose Increased Same AbolishedaSurvival Same Reduced" Abolisheda

MMC Dose Abolished" Same AbolishedaSurvival Abolisheda Same Abolisheda

X rays Dose Same Same Abolished"Survival Same Reduceda Abolisheda

NTG Dose Same Same ReducedaSurvival Same Reduceda Greatly re-

duceda

MMS Dose Same -Same AbolishedaSurvival Same Reduceda Abolisheda

EMS Dose -Same -Same -SameSurvival Same Reduced" Reduceda

a Reduction or abolition of Arg- _ Arg+ frequenciesindicates an antimutagenic action of the Uvr-, polA-, orrecA- allele. Abbreviations: NQO, nitroquinoline oxide;MMC, mitomycin C; NTG, nitrosoguanidine; MMS, methylmethane sulfonate; EMS, ethyl methane sulfonate.

recA- hosts, but that recB- and recC- hostsgave normal yields of such mutants. PhageRed- and Int- mutations also failed to impairUV mutagenesis of X. Subsequently, Defais etal. (39) showed that the Lex+ function of the E.coli K-12 host is necessary for production ofUV-induced c mutations (i.e., Lex- acts an-timutagenically [Lex in K-12 = Exr in B]).Furthermore, in contrast to the earlier report ofMiura and Tomizawa (110), they showed thatthe Red+ function had a small effect on X UVmutagenesis but a large effect on X spontaneousmutagenesis. Kerr and Hart (82) also showedthat mutagenesis of X by UV or nitrous acid didnot occur in recA or Exr- hosts and were able toseparate phage mutagenesis from the UV-reac-tivation process.There is clear evidence from the work of

Castellazzi et al. (14) that mutations in thebacterial recA, lexA, and zab (? = lexB) lociblock UV mutagenesis of phage X in a tif-1 hostat 42 C. Under these conditions in recA+, lexA +,Zab+ hosts, Tif-controlled functions, includinghost functions necessary for mutagenesis ofUV-irradiated phage X, are normally induced.In this connection, it would be extremely inter-esting to determine whether bacterial recA,lexA,2 and zab mutations also had an an-timutagenic influence on UV-induced muta-tions taking place in a host cell mul- geneticbackground (151). It should also be of someinterest to determine whether host recA and exr(lex) functions affect UV mutagenesis of otherphages, such as T1 or T3, and mutagenesis of Xand other phages by mutagens other than UVand nitrous acid. Pietrzykowska (122), workingwith 5-bromouracil-induced am - am+ muta-tions in phage X C17amO8, showed that muta-genesis by this base-analogue depended on hostrepair functions. Thus, bromouracil mutagene-sis of A was reduced or abolished in the absenceof a A red or host recA function and in a lex-host.

Tests might also be made in T4 phages of thepossible involvement of v, x, and y genes in mu-tagenesis. To a limited extent this has beendone in that Drake (46) found that the px com-ponent of Harm's x mutant (which is actuallya px, hm double mutant) slightly reduces spon-taneous mutation rates and decreases T4 UVand methyl methane sulfonate mutagenesisabout fourfold. Gene px is believed to be in-volved in generalized error-prone repair.

Antimutator Alleles of Phage T4 Genes 32and 43

One of the major advances in the understand-ing of the mechanisms of spontaneous muta-

44 BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



genesis has been the work by Drake and hisco-workers on those alleles in T4 genes 43 and 32which possess mutator or antimutator activity.Gene 43 is the structural locus for the phage-specified DNA polymerase and gene 32 codes forAlberts' protein (46). Mutations leading to totalloss of these functions are lethal, but tempera-ture-sensitive (ts) mutant alleles can be studied(1, 45-49). The antimutator ts gene 43 mutantsof phage T4 were shown to depress spontaneousmutations, in a variety of genes, arising by AT-- GC transitions. There was no antimutatoreffect, however, on transversion or frameshiftspontaneous mutagenesis. Indeed, some T4gene 43 ts alleles that were antimutators fortransitions actually increased transversion fre-quencies.With regard to antimutator effects on in-

duced mutations, transition mutagenesis by2-aminopurine, 5-bromouracil, or thymine dep-rivation was strongly suppressed. There was amoderate reduction in mutagenesis by ethylmethane sulfonate at G:C sites and by HNO2 atA:T sites. In contrast, these particular gene 43ts alleles had no antimutator effect on hydroxyl-amine mutagenesis or HNO2 mutagenesis at G:C sites. It is of some interest that two differentantimutator alleles in T4 gene 43, e.g., ts CB87(= L141) and ts CB120, had in some casesequal, and in other cases unequal, antimuta-genic effects upon the spontaneous or inducedreversions of an nIl or amber mutant in anothergene. This probably indicates that the mutabil-ity of a particular base pair is influenced byneighboring base pairs and that such differencesare perceived in different ways by the variousmutant DNA polymerase molecules (145). Fur-thermore ts+ revertants of ts antimutator allelesin gene 43, with restoration of normal burst sizesat 42.5 C, showed, in some cases, retention ofantimutator activity (46). This suggests thatsome missense mutations in the T4 DNA po-lymerase allow almost wild-type activity interms of DNA replication, yet have reducedlevels of error-proneness during replication. Inthe case of T4 gene 32 ts mutants, some causeddecreased spontaneous mutation rates at G:Cbase pairs and in some cases frameshift muta-tions were also reduced. A temperature-sensi-tive mutation in the T4 ligase gene has beenfound by Bernstein and Wilson (9) to exert anantimutagenic effect, in this case against 2-aminopurine- and 5-bromouracil-induced rever-sions of various amber mutants of T4.One of the few areas in which antimutagene-

sis is becoming a biochemically respectablestudy is in the investigation of mutant T4 DNApolymerases, including those exhibiting an-

timutator activity. Thus Muzyczka et al. (114)showed that partially purified DNA polymer-ases from two T4 gene 43 mutator, two an-timutator, and one neutral mutant could bedistinguished from each other and from wild-type polymerases by their ratios of polymer-ase:exonuclease:deoxynucleotide turnover ac-tivities. Schnaar et al. (134) further showed thatpolymerases of the antimutator type exhibit ahigher exonuclease to polymerase activity ratiothan do the wild-type enzymes. Whereas alltypes of T4 polymerase could discriminate be-tween 2-aminopurine and adenine during invitro DNA synthesis, the antimutator polymer-ase of ts L141 incorporated only one molecule of2-amino purine per 50 molecules of adenine, ascompared with values of 1 molecule of 2-aminopurine per 14 molecules of adenine for thewild-type polymerase and 1 per 10 for the muta-tor polymerase of ts L98. More recently Good-man et al. (61) have shown that antimutatorDNA polymerases of T4 are inhibited, in nu-cleotide incorporation studies, to a far greaterextent than wild-type or mutator enzymes bythe anticancer drugs adriamycin and daunoru-bicin and by ethidium bromide and 9-aminoac-ridine. In contrast, these compounds cause nodifferential inhibition of polymerase-associated3'-exonuclease activities of wild-type or mu-tants.

Rev, Umr, and Rad Mutants in S. cerevisiaeAn important advance in understanding the

control of induced mutagenesis in eukaryoticmicrobes has been the isolation and character-ization by Lemontt (93-96) of rev mutants.These were isolated as showing reduced UV-induced revertability of the arg4-1 7 ochre allele.Twenty rev isolates could be assigned to threeloci, namely revl, rev2, and rev3. Interestinglythe three different rev3 alleles showed verysimilar UV survival curves (all rev mutants aremore UV sensitive than is the wild-type, andrev2 is synonymous, according to Game and Cox[54], with the rad5 locus) yet had very differenteffects on arg4-17 - Arg+ revertant yields.Likewise, revl-1 and rev3-1 mutants had amuch more pronounced antimutagenic effect onarg4-17 UV-induced true back-mutationalyields than did the rev2-1 allele. Both revi -1and rev3-1 greatly reduced UV reversion of theochre alleles arg4-17 and lysl-l as well as themissense allele arg4-6. However, rev2-1 hadmuch less antimutagenic effect on arg4-17 orlysl-l reversion and no detectable effect onarg4-6 reversion. One surprising result was that,whereas rev3-2 and rev3-3 reduced arg4-17 trueback-mutations induced by UV, these same two

45VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


CLARKE AND SHANKEL

rev3 alleles resulted in apparently increasedyields of Arg+ revertants due to suppressormutations. However, all of these comparisonsby Lemontt (93) were made at equivalent UVdoses. If one replots his data so as to allowcomparisons at equal survivals, then rev3-1 hasthe most pronounced antimutagenic effectagainst UV-induced arg+ true back mutations,with revl-1, rev3-2, and rev3-3 having some-what lesser effects and rev2-1 being least an-timutagenic. Similar comparisons for UV-induced Arg+ revertants due to suppressor mu-tations show revi-1 and rev3-1 to have thestrongest antimutagenic effects, rev2-1 andrev3-3 to have intermediate effects, and rev3-2to have the least antimutagenic effect. Thus,the degree of antimutagenesis exerted by anygiven rev allele depends upon the particularUV-induced mutation being scored.Further experiments by Lemontt (96) demon-

strated that revi-1 and rev3-1 strongly, andrev2-1 weakly, suppressed UV-induced forwardmutation frequencies from prototrophy to auxo-trophy involving many loci throughout the ge-nome. Considering specific forward mutationsat the ade-1 and ade-2 loci, the same patternwas observed. The rather weak net antimuta-genic effect of the rev2-1 allele was probably dueto a counteracting influence of a mutator actionof rev2-1. Neither revl-1, rev2-1, nor rev3-1 hadmarked antimutator effects against ethylmethane sulfonate-induced forward mutationsfrom prototrophy to auxotrophy, or specificforward mutations at the ade-1 and ade-2 loci.Further studies by Lemontt (97) resulted in

the isolation of 40 Umr- mutants characterizedby their failure to yield UV-induced forwardmutations to canavanine resistance. Some ofthese Umr- mutants were found to be allelicwith rev mutants. One awaits with interestfurther details of the properties of these rev andumr mutants.

In addition to Lemontt's studies with mu-tants that were isolated primarily on the basisof their reduced UV mutability, Averbeck et al.(6) have reported on reduced UV mutability instrains selected on the basis of their radiationsensitivity. Mutants r2s and r%3 l showed greatlyreduced yields of Ilva- m Ilva+ reversions afterUV irradiation, compared with wild-type andsome other radiation-sensitive strains. Mutantr83%, has been shown subsequently (54) to belocated at the rad-2 locus of S. cerevisiae.Nakai and Yamaguchi (115) studied the in-

duction of a variety of types of mutations inwild-type (with regard to repair) S. cerevisiaeand the uvs-1 strain (rad-1 locus). When com-parisons were made on the basis of equal

survivals rather than UV dose, arg4-17 backmutations, Arg+ revertants due to suppressors,and reversions of the hisi -1 frameshift mutationwere all induced at lower frequencies in theuvs-1 than in the wild-type strain (115, Fig. 7).Resnick (127) showed that in haploid strains

of S. cerevisiae the uvs9-3 allele (= rad-2 locus)(54) resulted in reduced yields of UV-inducedArg+, Lys+, and His+ revertants when compari-sons were made between a Uvs+ and the Uvs-strain at equal survival levels. The reversionsstudied resulted from presumed addition-dele-tion (frameshift) and base-pair substitution(transitions, transversions) mutations, with thelatter class being represented by back-muta-tions in structural loci and suppressor muta-tions. Similarly, Moustacchi (113) studied UV-induced His- His+ reversions in Uvs+ anduVs2 mutants (= radl allele) (54). She showedthat when a comparison was made at equalsurvival levels above about 10%, the uvs2 strainwas less UV-mutable than its Uvs+ parentstrain, i.e., that the effect of uVs2 was antimuta-genic. Waters (153) demonstrated that variousdifferent rad-3 alleles result in lowered UV-induced mutation frequencies compared withthe rad+ strain when comparisons are made atequal survival levels. In this case, the mutationsstudied were Adn-. Adn+ reversions andforward mutations conferring resistance to acti-dione.Lawrence et al. (91) compared the specificity

and frequency of UV-induced reversion of aniso-1-cytochrome c ochre mutant in wild-typeand radiation-sensitive strains of S. cerevisiae.Their results show that the rad6-1 allele resultsnot only in reduced UV reversion frequencies,but there is also a loss of a high degree ofmutagenic specificity, in that the revertants aredue to a variety of base-pair substitutionsrather than, as in wild-type, predominantly AT- GC transitions in the first position of theochre codon.Prakash (123) has also demonstrated that

rad6 and rad9 mutants greatly reduce chemicaland radiation mutagenesis in S. cerevisiae. Thetest system involved reversions of an iso-1-cyto-chrome c mutation. Of numerous mutagenstested, only nitrous acid and nitrosoimidazoli-done exhibited apparently normal mutability atlow doses in rad6 and rad9 strains.

Radiation-Sensitive Mutants in OtherEukaryotic Microbes

Chang et al. (15) and Wohlrab and Tuveson(170) showed that in the uvs-1 mutant ofAspergillus nidulans yields of UV-induced Adn-- Adn+ reversions were reduced, compared

46 BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



with those in the Uvs+ strain. This was true Iwhether comparisons were made at equal UVdoses or equal survival levels (170). In thesesame strains, Met- - Met+ reversions werereduced in the uvs-1 mutant at doses of UVgiving above approximately 3% survival, but athigher doses (lower survivals) the uvs-1 straingave apparently greater Met+ yields than thewild-type strain. Similarly, in the uvs-1 mutantof Neurospora crassa, Chang et al. (15) showedthat there was a reduced yield of UV-inducedmutations to acriflavine or caffeine resistance.Again this conclusion did not depend uponwhether one compared the uvs-1 mutant of N.crassa with the wild-type at equal UV doses orequal survival levels. de Serres (40), studying aseries of UV-sensitive mutants of N. crassa andscoring UV-induced forward mutations at theadn3A and adn3B loci, showed that uvs-3 anduvs-4 gave markedly reduced mutational yieldscompared with wild-type, and that uvs-1, upr-i,and uvs-5 gave smaller antimutagenic effects.In the fission yeast S. pombe, Nasim (116)found that the uvs-1 mutant showed greatlyreduced UV mutability. Forward mutations in aseries of genes concerned with purine biosynthe-sis, which converted the phenotype of an adn-7mutant from red to white colonies, were scored.Davies and Levin (34) studied UV-induced

mutations from acetate requirement to in-dependence in haploid cells of Chlamydomonasreinhardi. They used a wild-type (Uvs+ withrespect to dark-repair capacity) and two UV-sensitive (Uvs-) strains in their experiments.Comparisons of mutation frequencies betweenthe Uvs+ and the two Uvs- strains were made atvarious doses of UV light. Strain uvs-1 showedreduced and abnormal revertant frequencies,compared with the Uvs+ strain, giving a near-normal revertant frequency after low doses ofUV but a decrease in revertant frequency athigher UV dosages. Strain uvs-6 showed arevertant frequency that increased with increas-ing dosages of UV and reached a plateau level.The maximum revertant frequencies given bythe two Uvs- strains were at least 10-fold lowerthan those given by the Uvs+ strain. Thus, onemay conclude that the effect of both Uvsmutations is antimutagenic on UV-induced ace-tate revertants. This conclusion is confirmed ifone replots the results of Davies and Levin (34)so as to allow comparisons to be made of themutabilities of the Uvs+ and the two Uvs-strains at equivalent survival levels.

Finally, Arlett (4) provided an intriguingexample of a cytoplasmic antimutagenesis. Inthe red cytoplasmic variant of A. nidulans,which showed increased resistance to UV killing

but not to gamma radiation lethality, there wastotal abolition of both UV- and gamma-inducedmutations to 8-azaguanine resistance. Suchmutations were induced readily in the wild-typecytoplasm. More surprisingly, radiation-induced mutations to 6-methylpurine resistancewere induced about normally in the red cyto-plasm. Thus, the red cytoplasm exhibits a highdegree of antimutagenic specificity. This isreminiscent of the specific antimutagenesis alsoexerted in A. nidulans against 8-azaguanine-resistant mutations by manganous ions (3).

CONCLUSIONSIn concluding this review, there are several

points we wish to emphasize. First, it is obviousthat knowledge of antimutagenesis is at presentvery fragmentary. There is clear need to repeatseveral studies employing mutations involvingknown types of base-pair changes. Second, itwill be of obvious advantage in some instancesto test for antimutagenic effects of chemicals inpairs of strains having and lacking a singlewell-defined repair function. Third, the exten-sion of antimutagenesis studies from phage andbacterial systems into eukaryotic systems isalready taking place and will surely increase.Fourth, a word of caution is needed in that theknown, or suspected, biochemical effects ofmany antimutagenic chemicals, e.g., spermineor caffeine, are legion and it will be extremelydifficult to pinpoint the one, if indeed it exists,effect which is responsible for their antimuta-genic action. Fifth, caution is also necessary ininterpreting the antimutagenic effects of radia-tion-sensitive mutations. Reduction or abolitionof mutability in a radiation-sensitive strainneed not imply that an error-prone repair genefunction is directly implicated in mutagenesis.Known antimutagens are, like known muta-

gens, so chemically diverse that a search for newtypes cannot logically be confined only to a fewclasses of compounds. If naturally occurringantimutagens are one of the normal regulatorsof spontaneous and induced mutation frequen-cies, then one might expect to be able to testthis hypothesis by a biochemical analysis ofsome classes of mutators and mutagen hyper-mutable mutants. Mutators of several types arewell known in several microorganisms (79), andUV hypermutable (hm) mutants have beendescribed by Drake (46) in phage T4, and2-aminopurine hypermutable mutants havebeen isolated in E. coli B/r by Burr and Clarke(unpublished data). To date none of these hasbeen investigated for, or shown to be lacking in,an endogenous antimutagen. At a pragmaticlevel it would be most useful if antimutagens

47VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


CLARKE AND SHANKEL

should be found which strongly counteract themutagenic activities of at least some environ-mental mutagens, particularly those havingdesirable pharmacological uses.Antimutagens have already been used, ap-

parently successfully, as adjuvants in chemo-therapy of patients in an attempt to reduce thefrequency of occurrence of spontaneous antibi-otic-resistant bacterial variants (50). It istempting to speculate regarding whether or notthey might also have some practical use incancer chemotherapy or prevention.

ACKNOWLEDGMENTSOur collaboration in this review has been made

possible by a NATO Research Grant for ScientificCooperation.We wish to thank Earl R. Nestmann for making

available details of his results prior to publication andfor extremely useful correspondence and discussionswith one of us (C.H.C.). We thank D. A. Hopwoodand H. 0. Stone for their constructive comments onour manuscript.

LITERATURE CITED1. Allen, E. F., I. Albrecht, and J. W. Drake. 1970.

Properties of bacteriophage T4 mutants defec-tive in DNA polymerase. Genetics 65:187-200.

2. Apirion, D., and D. Schlessinger. 1967. The lossof phenotypic suppression in streptomycinresistant mutants of Escherichia coli. Genetics56:542.

3. Arditti, R. R., and G. Sermonti. 1962. Modifica-tion by manganous chloride of the frequencyof mutation induced by nitrogen mustard.Genetics 47:761-768.

4. Arlett, C. F. 1966. The influence of the cyto-plasm on mutation in Aspergillus nidulans.Mutat. Res. 3:410-419.

5. Auerbach, C. 1970. Mutagen specificity. Trans.Kans. Acad. Sci. 72:273-294.

6. Averbeck, D., W. Laskowski, F. Eckardt, and E.Lehmann-Brauns. 1970. Four radiation sensi-tive mutants of Saccharomyces. Survival afterUV- and X-ray-irradiation as well as UV-induced reversion rates from isoleucine-valinedependence to independence. Mol. Gen. Ge-net. 107:117-127.

7. Bachrach, U. 1970. Metabolism and function ofspermine and related polyamines. Annu. Rev.Microbiol. 24:109-130.

8. Barfknecht, T. R., and D. M. Shankel. 1973.Antimutagenesis of suppressors of a trypto-phane requiring mutant in Escherichia coli.Abstr. Annu. Meet. Am. Soc. Microbiol. 1973,G7, p. 27.

9. Bernstein, C. and L. B. Wilson. 1973. The effectof a temperature-sensitive ligase on base ana-logue mutagenesis in bacteriophage T4. Ge-netics 74:s22.

10. Bockrath, R., and M. K. Cheung. 1973. The roleof nutrient broth supplementation in UV mu-tagenesis of E. coli. Mutat. Res. 19:23-32.

11. Bohme. H. 1962. Die Beeinflussung der inaktivi-erenden und mutagenen Wirkung von Athyl-Methansulfonat durch Mangan-und Magnesi-umchlorid. Biolog. Zententralbl. 81:267-276.

12. Bridges, B. A., J. Law, and R. J. Munson. 1968.Mutagenesis in E. coli. II. Evidence for acommon pathway for mutagenesis by ultravio-let light, ionising radiation and thymine depri-vation. Mol. Gen. Genet. 103:266-273.

13. Brock, R. D. 1971. Differential mutation of the,B-galactosidase gene of Escherichia coli. Mu-tat. Res. 11:181-186.

14. Castellazzi, M., J. George, and G. Buttin. 1972.Prophage induction and cell division in E.coli. II. Linked (rec A, zab) and unlinked (lex)suppressors of tif-1 mediated induction andfilamentation. Mol. Gen. Genet. 119:153-174.

15. Chang, L. T., J. E. Lennox, and R. W. Tuveson.1968. Induced mutation in UV-sensitive mu-tants of Aspergillus nidulans and Neurosporacrassa. Mutat. Res. 5:217-224.

16. Chiu, J. F., and S. C. Sung. 1972. Effect ofspermidine on the activity of DNA polymer-ases. Biochim. Biophys. Acta 281:535-542.

17. Chopra, V. L. 1967. Gene-controlled change inmutational stability of a tryptophanless mu-tant of E. coli WP2. Mutat. Res. 4:382-384.

18. Clarke, C. H. 1962. A case of mutagen specificityattributable to a plating medium effect. Z.Vererbungsl. 93:435-440.

19. Clarke, C. H. 1963. Suppression by methionineof reversions to adenine independence inSchizosaccharomyces pombe. J. Gen. Micro-biol. 31:353-363.

20. Clarke, C. H. 1965. Methionine as an antimuta-gen in Schizosaccharomyces pombe. J. Gen.Microbiol. 39:21-31.

21. Clarke, C. H. 1967. Caffeine- and amino acid-effects upon try+ revertant yield in UV-irradiated hcr+ and hcr- mutants of E. coliB/r. Mol. Gen. Genet. 99:97-108.

22. Clarke, C. H. 1967. Mutational and dark-repairspecificities in UV-irradiated di-auxotrophs ofE. coli B/r. Mol. Gen. Genet. 100:225-241.

23. Clarke, C. H. 1968. Differential effects of caf-feine in mutagen-treated Schizosac-charomyces pombe. Mutat. Res. 5:33-40.

24. Clarke, C. H. 1969. Influence of cellular physiol-ogy on the realization of mutations-resultsand prospects, p. 17-28. In G. E. W. Wolsten-holme and M. O'Connor (ed.), Ciba Founda-tion Symposium on Mutation as CellularProcess. J. and A. Churchill Ltd., London.

25. Clarke, C. H. 1970. Repair systems and nitrousacid mutagenesis in E. coli B/r. Mutat. Res.9:359-368.

26. Clarke, C. H. 1972. Studies with antimutagensin an ochre suppressor mutation system inEscherichia coli, p. 41-65. In N. P. Dubininand D. Goldfarb (ed.), Molecular mechanismsin molecular genetics (Proc. First Symp. Mol.Genet.) Nauka Publishing House, Moscow.

27. Clarke, C. H. 1972. Spermine antimutagenesis inE. coli. Heredity 29:124.

28. Clarke, C. H. 1973. Influence of streptomycin on

48 BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



UV-induced Trp+ reversions in a strep-tomycin-resistant strain of E. coli B/r. Mutat.Res. 19:43-47.

29. Clarke, C. H., and D. M. Shankel. 1973. Areinterpretation of UV mutagenesis. Genetics74:733.

30. Clarke, C. H., and D. M. Shankel. 1974. Effectsof ethidium, quinacrine and hycanthone onsurvival and mutagenesis of UV-irradiatedHcr+ and Hcr- strains of E. coli B/r. Mutat.Res. 26:473-481.

31. Cohen, S. S. 1971. Introduction to the poly-amines. Prentice-Hall, Englewood Cliffs, N.J.

32. Corran, J. 1969. Analysis of an apparent case of"gene-controlled mutational stability": theauxotrophic preemption of a specific growthrequirement. Mutat. Res. 7:287-295.

33. Cox, E. C. 1973. Mutatortgene studies in Esche-richia coli: The mut T gene. Genetics73(Suppl.) :67-80.

34. Davies, D. R., and S. Levin. 1968. UV-inducedreversion to prototrophy in three strains ofChiamydomonas reinhardi differing in darkrepair capacity. Mutat. Res. 5:231-236.

35. DeCourcy, S. J., Jr. 1971. Antimutagens andantimicrobial drug resistance. Prog. Mol. Sub-cell. Biol. 2:316-327.

36. DeCourcy, S. J., Jr., M. M. Barr, W. S. Blake-more, and S. Mudd. 1971. Prevention of anti-biotic resistance in vitro in Staphylococcusaureus, Escherichia coli, and Pseudomonasaeruginosa by coumadin. J. Infect. Dis.123:11-15.

37. DeCourcy, S. J., Jr., and S. Mudd. 1969. Effectof the nutritional environment on the develop-ment of resistance to polymyxin B in Pseu-domonas aensginosa and its prevention byAtabrine, p. 72-76. Antimicrob. AgentsChemother. 1968.

38. DeCourcy, S. J., Jr., and M. G. Sevag. 1967.Population dynamics and results of fluctua-tion tests in a study of the role of Atabrine asan antimutagen in preventing streptomycinresistance in Staphylococcus aureus, p.235-243. Antimicrob. Agents Chemother.1966.

39. Defais, M., P. Fauquet, M. Radman, and M.Errera. 1971. Ultraviolet reactivation and ul-traviolet mutagenesis of X in different geneticsystems. Virology 43:495-503.

40. de Serres, F. J. 1971. Mutability of UV-sensitivestrains of Neurospora crassa. Genetics 68:s14.

41. Domon, M., B. Barton, A. Porte, and A. M.Rauth. 1970. The interaction of caffeine withultraviolet-light-irradiated DNA. Int. J. Ra-diat. Biol. 17:395-399.

42. Doneson, I., and D. M. Shankel. 1963. Muta-tional synergism between radiations andmethylated purines in Escherichia coli. J.Bacteriol. 87:61/67.

43. Doudney, C. O., and F. L. Haas. 1959. Effect ofpurine and pyrimidine analogues on reversionof a tryptophan-requiring strain of Esche-richia coli. Nature (London) 184:114-115.

44. Doudney, C. O., and F. L. Haas. 1960. Some

biochemical aspects of the postirradiationmodification of ultraviolet induced mutationfrequency in bacteria. Genetics 45:1481-1502.

45. Drake, J. W. 1969. Mutagenic mechanisms.Annu. Rev. Genet. 3:247-268.

46. Drake, J. W. 1973. The genetic control of sponta-neous and induced mutation rates in bacterio-phage T4. Genetics 73(Suppl.):45-64.

47. Drake, J. W., and E. F. Allen. 1968. Antitnuta-genic DNA polymerases of bacteriophage T4.Cold Spring Harbor Symp. Quant. Biol.33:339-344.

48. Drake, J. W., E. F. Allen, S. A. Forsberg, R. M.Preparata, and E. 0. Greening. 1969. Geneticcontrol of mutation rates in bacteriophage T4.Nature (London) 221:1128-1132.

49. Drake, J. W., and E. 0. Greening. 1970. Sup-pression of chemical mutagenesis in bacterio-phage T4 by genetically modified DNA polym-erases. Proc. Nat. Acad. Sci. U.S.A.66:823-829.

50. Eshelman, J. L., M. R. Horwitz, S. J. DeCourcy,Jr., S. Mudd, and W. S. Blakemore. 1970.Atabrine as an adjuvant in chemotherapy ofurinary tract infections. J. Urol. 104:902-907.

51. Fabre, F. 1972. Relation between repair mech-anisms and induced mitotic recombinationafter UV irradiation in the yeast Schizosac-charomyces pombe. Mol. Gen. Genet.117:153-166.

52. Fox, M. S. 1955. Mutation rates of bacteria insteady state populations. J. Gen. Physiol.39:267-278.

53. Fuks, Z., and K. C. Smith. 1971. Effect ofquinacrine on X-ray sensitivity and the repairof damaged DNA in Escherichia coli K-12.Radiat. Res. 48:63-73.

54. Game, J. C., and B. S. Cox. 1971. Allelism testsof mutants affecting sensitivity to radiation inyeast and a proposed nomenclature. Mutat.Res. 12:328-331.

55. Ganesan, A. K., and K. C. Smith. 1972. Require-ment for protein synthesis in rec-dependentrepair of deoxyribonucleic acid in Escherichiacoli after ultraviolet or X irradiation. J. Bacte-riol. 111:575-585.

56. Gartner, T. K., and E. Orias. 1966. Effects ofmutations to streptomycin resistance on therate of translation of mutant genetic informa-tion. J. Bacteriol. 91:1021-1028.

57. Glass, E., and A. Novick. 1959. Induction ofmutation in chloramphenicol-inhibited bacte-ria. J. Bacteriol. 77:10-16.

58. Glover, S. W. 1956. A comparative study ofinduced reversions in Escherichia coli. Carne-gie Inst. Washington Publ. 612:121-136.

59. Godsell, A. M., and C. H. Clarke. 1975. Caffeineand quinacrine (Atabrin) as mutagens andantimutagens for spontaneous revertants ofochre, missense and frameshift trp- mutationsin E. coli. In N. P. Dubinin and D. Goldfarb(ed.), Proc. Second Symp. Mol. Genet., inpress. Nauka Publishing House, Moscow.

60. Goodman, M. F., and M. J. Bessman. 1973.Purification and properties of two growth

49VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


50 CLARKE ANI

stage-dependent deoxyribonucleic acid po-lymerases from Lactobacillus acidophilus. J.Biol. Chem. 248:815-821.

61. Goodman, M. F., M. J. Bessman, and N. R.Bachur. 1974. Adriamycin and daunorubicininhibition of muta

Antimutagenesis in Microbial Systems · fine a mutation as a stable, heritable change in the DNA....

Documents

Transcript of Antimutagenesis in Microbial Systems · fine a mutation as a stable, heritable change in the DNA....