Fundamental and Molecular Mechanisms of Mutagenesis
Mutation Research 356 (1996) 229-235
Induction of SOS-independent mutations by benzo[ a]pyrene treatment in Escherichia coli cells deficient in MutY or MutM DNA glycosylases: possible role of oxidative lesions Amparo Urios, Manuel Blanc0 Instiruto de 1mwtigacione.r
Valenciana de hestigaciones Spain
Amadeo de Saboy
4. 46010 Valencia.
Received 19 December 1995: revised 19 March 1996: accepted 19 March 1996
Abstract The induction of SOS-independent mutations by exposure to benzo[a]pyrene (BaP) was screened in Eschcrichiu coii strains lacking SOS mutagenesis proteins and deficient in MutY or MutM glycosylases, which prevent mutations by 8-hydroxyguanine (GO lesion). Mutagenicity assays, performed in the presence of S9 mix, indicated a great increase in the reversion of the trpE65 ochre mutation in both m&Y and mutY mutM strains, whereas a lower increase was observed in a mu&i strain. This mutability by BaP was observed in either wr- or wr+ strains. Moreover. it was increased when strains carried a deletion of the oxyR gene that abolished the OxyR response to oxidative stress, and reduced in the presence of the o~R2 allele that rendered constitutive such response. It is suggested that SOS-independent mutations in cells treated with BaP arise from adenine/GO mispairs. The interaction of radical scavengers with BaP ultimate metabolites could cause oxidative stress capable of producing GO lesions. Strains lacking mutagenesis proteins and deficient in base excision repair systems, such as those dependent on MutY and MutM glycosylases, could be useful for screening the induction of SOS-independent mutations. Keyords:
mutagenesis: MutY; MutM; OxyR: Oxidative stress
1. Introduction Mutations caused by many genotoxic agents result from the bypass of replication blocking DNA lesions. The lesion bypass requires the activity of a mechanism called SOS mutagenesis which is in-
author. Tel.: ( + 34-6) 369-8500;
Fax: (+ 34-6)
duced as part of the SOS response to DNA damage. SOS mutagenesis is mediated by either UmuD/C proteins or their plasmid-encoded analogs MucA/B proteins [l]. One of the MucA/B-encoding plasmids, termed pKM 101, has played a major role in the success of the Ames test because of its ability to improve the sensitivity of the Salmonella typhimurium tester strains to the mutagenic effects of many chemicals . The mutagenicity of these chemicals has therefore been considered to be dependent on SOS mutagenesis. Since the S. typhimurium tester
1996 Elsevier Science B.V. All right!, reserved.
strains devoid of plasmid pKM 101 contain other SOS mutagenesis proteins, such as UmuD/C and SamA/B [3,4], the mutagenicity observed for some chemicals in strains lacking plasmid pKM 10 1 cannot be ascribed to an SOS-independent mechanism. Escherickid coli strains with a full deletion of the chromosomal urn& / C operon have proved to be useful to study the dependence of cell mutability on the SOS mutagenic mechanism . Using a AurrzuD /C strain transformed with a MucA/B-encoding plasmid, it was demonstrated that induction of base substitution mutations by benzo[ a]pyrene (BaP) requires the function of the matured MucA protein jointly with the MucB protein. The induction of these mutations was null when none of mutagenesis proteins was present in cells, suggesting that BaP did not cause SOS-independent mutations. In a recent study of mutagenesis induced by oxidants. we observed that the occurrence of SOS-independent mutations could easily be detected in strains lacking mutagenesis proteins if they were deficient in the MutY DNA glycosylase alone or in both MutY and MutM(Fpg) glycosylases [6,7]. These glycosylases are involved in the prevention of mutations resulting from the presence in the DNA of an oxidatively damaged form of guanine. X-hydroxy-
of the @
+ t t
guanine (also known as a GO lesion) [8,9]. GO lesions are mutagenic in an SOS-independent fashion due to their miscoding potential. The MutM glycosylase removes the oxidized base GO from DNA whereas MutY eliminates adenine misincorporated opposite the GO lesion. We reasoned that strains lacking mutagenesis proteins and deficient in MutY and MutM glycosylases could be useful for screening SOS-independent mutagenic effects of chemicals known to induce SOSdependent mutations. Moreover. in order to assess a role for reactive oxygen species in such mutagenic effects, the antioxidant defences of strains could be modulated by mutations affecting the OxyR-regulated response to oxidative stress. The induction of this response includes an enhanced expression of genes encoding catalase. alkyl hydroperoxide reductase and glutathione reductase [IO]. In this paper we show that the reversion of the 1171/35 ochre mutation could indeed be detected in cells containing no SOS mutagenesis proteins and deficient in MutY or MutM glycosylases after BaP exposure. This ceil mutability appeared to be promoted by oxidative DNA damage since it was influenced by the abolition as well as by the constitutive expression of the OxyR response to oxidative stress.
rtrutYhX( micA6X):: km mutation
C‘)::at lot11I .strlAl.and derive t’rwl strain [ 141: the nrrrtM Imutation is mutMI( fp,p-/)::km
r)~;\xR_ ’ [IO]: the MucA/B
i\ the low
: the rqK copy
LIo.\-vR_W ; the o\-vR’
A. Urios, M. Blanco/Mutation
356 (1996) 229-235
2. Materials and methods
2.1. Bacterial strains
The S9 metabolizing mixture (S9 mix) consisted, per assay, of: 50 pl of S9 fraction, 500 pl of 0.2 M phosphate buffer (pH 7.41, 40 ~1 0.1 M NADP, 5 ~1 1 M glucosed-phosphate, 20 ~1 0.4 M MgCl, and 20 ~1 1.65 M KCl. The S9 fraction was from rats that had received two injections of 20 and 10 mg of 3-methylcholanthrene in the 2 days before they were killed.
Bacterial strains are listed in Table 1. mutY68:: kan, mutM1:: kan and oxyR2 mutations were introduced in the recipient strains by Pl transduction. The oxyR2 allele, linked to btuB::TnlO, was from TA4110 [lo]. Double mutant mutY mutM strains were constructed as follows: mutM1:: kan x::mini-tet pyrE60 from strain IC3970  was introduced in a mutY68::kan strain by Pl transduction, selecting for Tc’, and then introducing mutM1:: kan by another Pl transduction, selecting for pyrE’ Pyr+ Tc”. The uur+ allele was introduced into uvrAZ.5.5 strains, which are 1amB (Mal-1, by transduction, selecting for Mal+. Deep-rough (Ifa) derivatives were isolated as resistant to phage C21 and sensitive to novobiocin and crystal violet [I 11. 2.2. Media Nutrient broth was Oxoid Nutrient Broth No. 2. Solid minimal B4 medium contained 15 g Difco agar and 4 g glucose per litre of Vogel-Bonner E medium . Solid ET4 medium was solid minimal E4 medium supplemented with 0.5 mg tryptophan per litre. Top agar contained 6 g Difco agar and 5 g NaCl per liter of distilled water.
Table 2 Mutagenicity
of t-butyl hydroperoxide
AoxyR30 mu+ o.qR+ mut+
of S9 mix
The experimental protocol was as follows. To 2 ml of molten top agar were added: 100 pl of a fresh overnight culture grown in nutrient broth at 37°C 100 pl of benzo[a]pyrene (Sigma) dissolved in DMSO, and 635 ~1 of the S9 mix. The complete mixture was poured on plates of minimal ET4 medium. Trp+ colonies were scored after 2 days at 37°C. t-Butyl hydroperoxide mutagenicity was measured similarly except that the S9 mix was omitted and the appropriate dilution of t-butyl hydroperoxide (Sigma) in distill e d water was added to the molten top agar. To evaluate the number of preplating mutants, the zero dose in the mutagenesis assay was screened on unsupplemented E4 plates. Experiments were carried out at least 3 times and 3 plates per dose were employed.
in strains lacking mutagenesis
- BuOOH 9*
A oqR30 mutM1 oxyR+ mutkll
IC3894 IC3793 IC5006
AoqR30 mutY68 oxyR+ mutY68 oxyR2 mutY68
37+ 39i 18!c
IC398 I IC3987 IC3996
AoxyR30 mutY68 m&Ml oxyR+ mutY68 mutM1 oxyR2 mutY68 mutMl
3 3 3
311+ 119k 61k
12 16 5
336 + 20 394 + 18 268 + 22
The number of Trp+ revertants is the average of 9 plates & SEM. The BuOOH strains, and 50 p,g/plate for mutY68 m&Ml double mutant strains.
dose was 100 pg/plate
3000 + 203 18OO* 74 970 * 32 for mut+, mutM1 and murY68
3. Results and discussion Table ?. summarizes the influence of a deficient\ in MutY or MutM DNA glycosylases on the mutability induced by the oxidant r-butyl hydroperoxide in strains lacking mutagenesis proteins [6.7]. A defect in MutY significantly increased the reversion of the trpE65 ochre mutation over the background observed in mut+ strains, whereas lower increaxeh wet-e promoted by a deficiency in MutM. Strains deficient in both MutY and MutM. exhibited high levels oi revertants induced by !-butyl hydroperoxide. In these double mutant strains. a significant enhancement in the frequency of spontaneous mutants was found. Also shown in Table r! is the increase in cell mutability occurring when the OxyR response to oxidative stress was abolished by the .lo.~-yR_W mutation, as well as the reduction in that mutability observed when the rqR2 allele allowed a constitutive induction of the OxyR-regulated genes. Derivatives of strains shown in Table ?. carrying rfil mutations, were used in the plate incorporation assay of BaP mutagenicity in the presence of S9 mix (Fig. 1). This mutagenicity was null in strain IC1870. having a mut+ genotype and containing no mutagenesis proteins. In contrast, after BaP exposure. strain IC5002 (mutM1) showed a low increase in the revertant number, this number being slightly enhanced in strain ICSOO9 (rmd4l A oxyR30 1, whereas strains deficient in MutY exhibited’a great increase in the number of revertants. At 2 p_g/plate of BaP. the fold-increases over the correspondinp control value\ were, respectively. IS. 30 and I I for strains 1C5OOI (m~tY68). IC3908 ( n o.vyR30 rmrtY6K) and ICSOO7 ( oxyR-7 tnut Y68 ).
lC3908, I’ t al ;;r a
IC5001 ,, 3
,, / 1001
I :’ 1
IC5009,_,_-- :- 4
In the ttlutY6A’ tnmhf/ double mutant strain IC.3991. a h-fold increase over background at I kg/platt: of BaP was found (Table 3). This mutabilit! was moderately improved in the presence of the
A. Urios, M. Blanc0 /Mutation
AoxyR30mutation, while it was about 30% reduced when cells carried the oxyR2 allele (data not shown). Note the high frequency of spontaneous mutagenesis exhibited by strain IC3991 when the S9 mix was present in the mutagenesis assay. In the absence of S9 mix, or when only the S9 fraction was present, that frequency was lower (about 300 revertants per plate), suggesting that addition of S9 mix may produce oxidative DNA damage. The results indicate an important role for the MutY glycosylase in the prevention of SOS-independent mutations in BaP-treated cells. The MutM glycosylase would have a minor role by itself, this role being more relevant when MutY is also defective. The MutY glycosylase excises adenine from an A/GO mispair whereas MutM removes the GO lesion from DNA [9,13]. Although MutY is also active on either A/G or A/C mispairs [8,14], and MutM can remove from DNA a variety of modified ring-opened purines , we suggest that A/GO mispairs, which are susceptible to be repaired by both MutY and mutM, constitute the promutagenic lesions generating SOS-independent mutations after BaP treatment. It would be consistent with the enhancement of the BaP mutagenicity observed in either mutY or mutM strains (Fig. l), as well as with the additional enhancement to this mutagenicity exhibited by the m&Y mu&l double mutant (Table 3). The involvement of A/GO mispairs in the SOSindependent mutagenicity promoted by BaP is also supported by the results of the characterization of extragenic ochre suppressors of trpE6.5 arising in mutY and mutA4 strains. The characterization was done, by testing growth of phage T4 mutants and by transductional mapping , in @z’ strains (see Table 1) because $r strains are resistant to both T4 and Pl phages. It was found that Trp’ revertants carried almost exclusively a supC suppressor resulting from G:C-T:A transversions which would be generated from A/GO mispairs (data not shown). The GO lesion presumably produces only a minor distortion of the DNA backbone and is not a good substrate for the nucleotide excision repair pathway [ 151. Therefore, this repair pathway might not influence the SOS-independent mutability of cells treated with BaP. This was confirmed by the fact that similar results were found in the uur+ strains IC500.5 and IC5013, lacking mutagenesis proteins and defi-
Research 356 (1996) 229-235
cient in MutY, and in the corresponding uur- strains IC3991 and IC5001 (Table 3). In contrast, the BaPinduced mutability in the uur+ strain IC5008, containing MucA/B proteins, was greatly reduced with respect to that observed in the uur- strain IC5003 (Table 3) indicating that DNA adducts determining the SOS-dependent mutagenicity of BaP were susceptible to nucleotide excision repair. Attack of DNA by reactive oxygen species could cause the GO lesions generating SOS-independent mutations after BaP treatment. However, these mutations failed to be induced when the S9 mix metabolizing system was not included in the mutagenesis assay (data not shown), indicating that the occurrence of such mutations required the production of BaP ultimate metabolites, i.e., electrophiles able to react with DNA . It thus appeared that BaP metabolites would promote the formation of GO lesions, perhaps by means of unstable DNA adducts, as has been described for nitroquinolines [ 171. We think, however, that a role for oxygen radicals in the induction of SOS-independent revertants after BaP exposure is suggested by the influence, on this induction, of mutations which either abolished or rendered constitutive the OxyR response to oxidative stress (Fig. l), this influence being similar to that observed in cells treated with the oxidant t-butyl hydroperoxide (Table 2). The OxyR-regulated proteins include several antioxidant enzymes such as catalase, alkyl hydroperoxide reductase and glutathione reductase [10,18], which could affect the accumulation of reactive BaP metabolites in the cells. For example, it is known that reduced glutathione directly reacts with electrophiles and prevents them from binding to DNA [19,20]. We then propose that BaP metabolites could interact with radical scavengers, which constantly protect cells against endogenous oxygen radicals generated by the aerobic metabolism [21,22], and could thus originate oxidative stress capable of producing GO lesions. As suggested by the dose-response curves of Fig. 1, an enhanced mutability, which could result from an increase in the cellular pro-oxidant state, is correlated with the concentration of BaP, and therefore of BaP metabolites, in the mutagenesis assay. In conclusion, we have demonstrated that the occurrence of an SOS-independent mutability can be detected in cells treated with BaP. This mutability
would have not been observed in the standard mutagenesis assays since they use strains containing SOS mutagenesis proteins. Moreover. we have shown that utilization of strains lacking these mutagenesis proteins is not a sufficient condition to detect SOS-independent mutations. given that DNA lesions generating these mutations may be repaired by mechanisms such as the base excision repair mediated by MutY and MutM glycosylases. Abolition of these mechanisms in the tester strains is therefore required for screening SOS-independent mutagenic effects. In these effects, as suggested by our results. oxidative DNA damage could play a major role.
the OxyR or MutY function\: dependence on SOS mutagenesi\. Mutation Res., 332. 9-15. I71 Urios, A. and M. Blanco ( 1994)Specificity of spontnneou< and r-butyl hydroperoxide-induced mutations in htzq\R \tram\; of E.cchu/+hic~ co/i differing with respect to the SOS mutagenesis proficiency and to the MutY
tions. Mutation Res.. in press.
protecth organisms from the mutagenic effect of the spontaneouh lesion X-hydroxyguanine J. Bacterial..
T. and A.P. Grollman
ing the oxidatively
Repair of DNA contaitl-
damaged base, R-oxoguanine. Mutation
Res.. 299. X-287. [IO] Christman.
Ames (1985) agnin\t
Mot-gall. F.S. Jacobson and B.N.
of a regulon
stress and some heat-shock
.Str/rno,~r/lo rv~~iin~rn’um. Cell. J I. 753-762. [I I] Herrera, G.. A. Urios, V. Aleixandre
and M. Blanc0 (1993)
Mutability by polycyclic hydrocarbons is improved in derivatives of E.whrrichirr w/i
We thank G. Herrera for her collaboration on diverse aspects of this study and C. Navarro for technical assistance. We are also grateful to Dr. D. Touati for strain TA41 IO. This work was supported by Comision Interministerial para la Ciencia y Tecnologia Grant SAF93-0056.
1r13rAwith increased permc-
ability. Mutation Res.. 301, I-5.
Maron. D.M. and B.N. Ames
Revised methods for the
Snlmr~r~ellrrmutagenicity test. Mutation Res., 113. 173-215. Il.31 Michaels.
( 1992) Evidence that MutY
combine to prevent
mutations by an oxidatively damaged form of guanine. Proc. Natl. Acad. Sci. USA. 89. 70X-702.5.
mismatch repair activities in E.cclreri&icl r,o/i. Proc. Natl. Acad. Sci. USA. 85. 9673-9678.
A.P. and M. Moriyn (1993)
oxoguanine: an enemy within. Trends Genet.. 9, 246-349.
R. and S.G. Sedgwick
duced by bacterial UmuDC protein\ and their plasmid homologues. Mol. Microbial..
J.. N.E. Spingarn. J. Kobori and B.N. Ames
residues in DNA
Detection of carcinogens ah mutagens: bacterial tester ptratn\
treated with 4-hydroxyaminoquinoline
with R fLator plasmid>. Proc. Natl. Acad. SC!. USA.
compounds in the presence of heryl-AMP,
phyh. Req. Commun.. C.M.
and E. Ei\enstudt
~rmuL)C locus in So/rn~~rrr//r/ f~phimwiurtr LT?. J. Bactrriol..
rayama and T. Sofuni
( I Y9 I ) Srr/rwm~/lrr r~~~/~rm~~~iw~ ha>
new [rrn[mC-like operon
in ;t 60.megadal-
ton cryptic plasmid of .S. r~pl~irnr~~irr,n.J. Bacterial..
B.N. Ames. T. Kada and L.W. of mutagenesis and their rele-
cance to carcinogenesia, Mutation Res., 168. 47-65.
Waters, M.D.. (I990)
Brady. H.F. Stack and H.E. Brockman
pounds. Mutation Res., 138, 57-85.
A., G. Herrcra. V. Aleixandre
and M. Blanco
Processing of MucA protein is t-equired for \pontaneoua and benr;o[a]pyrene-induced
frpA_7.?missenhe mutation hy G.C-T.A in the MutY
tran\versions: effect glycoayla\e.
Re\.. 3 I I. 7-57-263. Urio,.
S.B. Farr and B.N. Ames
Ratnel. C., U.K. Alekperov. Wattenberg (1986)
two homologous but different ~rfnf~L)Coperons: cloning of ;I
of ;I deficiency
G.. L.A. Tartaglia.
Nohmi. T., A. Hakul-a. Y. Nakal. M. Wutanabc. S.L.
I -oxide and its related
Bacterial defenses against oxidative stres\. Trends Genet.. 6.
I7 I ( 3860-3865.
Fifty years of benzo[a]pyrene, Nature.
K.. M. Tada. A. Hakurd, H. Kasai and Y. Kawaroe
Ames. B.N. and L.S. Gold
( 1995) Drtcction O&rri~~h,tr co/i deficient
( IY9 I)
Endogenous mutagens and
the cauhes of aging and cancer. Mutation Res.. 250. 3- 16.
B. and 0.1.
species. Its mechanism and tneasurement in
mammalian systems. FEBS Lett.. 28 I. 9- 19.
A.. G. Herrera and M. Blanco
oxidative mutagens in strain% of
E.M.. J.O. McCall. M.R. Volkert and I.E. Wermund-
modulation ot. mutagenesis rehulting from resolution of ge-
expression of SOS functions and
A. Urios. M. Blanco /Mutation Research 356 (1996) 229-235 netic instability at or near the recA locus of Escherichiu coli, Mol. Gen. Genet., 185, 43-50.  Boiteux, S. and 0. Huisman (1989) Isolation of a formamidopyrimidine-DNA glycosylase (fig) mutant of Escherichia coli K12. Mol. Gen. Genet., 215, 300-305.  Blanco. M.. G. Herrera and A. Urios (1995) Increased mutability by oxidative stress in OxyR-deficient Escherichia coli and Salmonella typhimurium cells: clonal occurrence of the
mutants during growth on nonselective media. Mutation Res., 346, 215-220. I261 Ho. C., 0.1. Kulaeva, AS. Levine and R. Woodgate (1993) A rapid method for cloning mutagenic DNA repair genes: isolation of umu-complementing genes from multidrug resistance plasmids R391, R446b. and R47la, J. Bacterial.. 175, 5411-5419.