15
Mutation Research, 59 (1979) 15-26 @ Elsevier/North-Holland Biomedical Press
EFFECT OF MUTAGENS, CHEMOTHERAPEUTIC IN DNA REPAIR GENES ON RECOMBINATION ESCHERZCHZA COLZ
AGENTS AND DEFECTS IN F’ PARTIAL DIPLOID
ALLEN J. NORIN a and EUGENE P. GOLDSCHMIDT University of Houston, Department
of Biology, Houston,
Texas 77004 (U.S.A.)
(Received 25 April 1978) (Revision received 18 August 1978) (Accepted 28 August 1978)
Summary The ability of mutagenic agents, nonmutagenic substances and defects in DNA repair to alter the genotype of F’ partial diploid (F30) Escherichiu coli was determined. The frequency of auxotrophic mutants and histidine requiring (His-) haploid colonies was increased by mutagen treatment but Hfr colonies were not detected in F30 E. coli even with specific selection techniques. Genotype changes due to nonreciprocal recombination were determined by measuring the frequency of His- homogenotes, eg. F’ hisC780, hisI’/hisC780, hisI’, arising from a His’ heterogenote, F’ hisC780 hisI’/hisC’, his1903. At least 75% of the recombinants were homozygous for histidine alleles which were present on the F’ plasmid (exogenote) of the parental hetergenote rather than for histidine alleles on the chromosome. Mutagens, chemotherapeutic agents which block DNA synthesis and a defective DNA polymerase I gene, polA1, were a Present address: Montefiore Hospital and Medical Center, C-416.111 East 210th Street, Bronx, N.Y. 10467 (U.S.A.)
Albert
Einstein
College
of Medicine,
and nomenclature: His-, phenotypic designation indicating a histidine requirement for growth; His+, growth is independent of histidine; his, genotypic designation indicating a defective gene for histidine biosynthesis; his+, genotypicslly His+; Hfr. an E. coli containing and F-factor which is integrated into the chromosome and is able to transfer the genetic material of this cell into an F- recipient; F’, an F factor that has a portion of the bacterial genome attached to it and coexists with but is not integrated into the chromosome: F-duce or F-duction. injection of the donor F-factor into a F- recipient via conjugation: F30, an F’ plasmid of E. coli which carries the genes for histidine biosynthesis; In describing partial diploids, the mutational content of the F’ will be given first, eg., strain 5059 is F’ hisC780/hisI903, pro signifying that the F’ plasmid (exogenote) carries the 780 mutation in the histidine operon C gene while the chromosome (endogenote) has the 903 mutation in the I gene as well as a second mutation in a gene involving proline biosynthesis. MNNG. N-methyl-N’-nitro-N-nitrosoguanidine; DES, diethyl sulfate; AO. a&dine orange; EMS, ethyl methanesulfonate; MMS, methyl methanesulfonate; 2AP, 2-aminopurine; MTX, methotrexate; UV. ultraviolet light; UVr, phenotypically resistant to UV-irradiation. wild-type; UVs, phenotypically sensitive to UV-irradiation. Abbreviations
16
found to increase the frequency of nonreciprocal recombination. A defect in the ability to excise thymine dimers, uurC34, did not increase spontaneous UV irradiation but not methyl nonreciprocal recombination. However, methanesulfonate (MMS) induced greater recombination in this excision-repair defective mutant than in DNA-repair-proficient strains. Mutagenic agents, with the exception of ethyl methanesulfonate (EMS), induced greater increases in recombination than the chemotherapeutic agents or the poZA1 mutation. EMS, which causes relatively little degradation of DNA, was more mutagenic but less recombinogenic than MMS, a homologous compound that induces rapid degradation of DNA. One explanation for these results is that inhibition of DNA synthesis increases recombination by prolonging the lifetime of naturally occurring single-stranded regions in replicative intermediates of the DNA. Mutagens which cause the rapid breakdown of DNA may, in addition, introduce lesions into the genome that increase the number of single-stranded regions thus inducing even higher frequencies of recombination.
Introduction Genotypic alterations regardless of their cause may have a profound effect on the functioning of cells. A number of reports indicate that both mutagenic and nonmutagenic substances can change the genotype of an organism by recombinational mechanisms. Mutagenic agents increase the frequency of recombination in micro-organisms [ 9,17,25,29,37]. Substances which inhibit DNA synthesis but do not react chemically with DNA also increase recombination [ 12,13,32]. Elevated frequencies of recombination are observed in mutants of T4 bacteriophage which are defective in intermediary DNA metabolism, DNA polymerase or DNA ligase activity [4,5,18]. The mechanisms by which the above agents and genetic defects increase recombination are not entirely clear. Presumably, mutagens induce a variety of perturbations in DNA which initiate recombinational events [21]. It is not surprising that chemotherapeutic agents and genetic defects that inhibit macromolecular DNA metabolism increase recombination since the processes of replication, repair, recombination and transcription partially overlap [8,25]. In addition, the extent of the overlap may vary depending on the organism [7,26]. Thus the molecular details of recombination may be quite different in one organism compared to another. No studies have been published comparing the effect of the above agents and genetic defects in DNA repair on recombination in one test organism. Consequently, it is difficult to determine which lesion is most effective in producing recombinant molecules in a given organism. This study was undertaken, therefore, to compare the relative effect of mutagens, chemotherapeutic agents which inhibit DNA synthesis and the effect of two different DNA-repair mutations (poZA1 and uurC34) on recombination in Escherichiu coli K12.
Recombination was studied in an F’ E. coli (F30) which is partially diploid for the histidine operon [2,14]. F’ strains are particularly advantageous for studies of this type since the effect of the experimental condition is on recombination rather than on chromosome transfer during conjugation. We observed
17
recombination between histidine alleles on the chromosome and those on the F’ plasmid. For example, recombination was determined by measuring the number of homozygous F’ hisC780 (plasmid genotype)@.&780 (chromosome genotype) colonies (His- homogenotes) which arose from a heterozygous F’ hisC’ISO/hisC’ parent (His’ heterogenote). Nonreciprocal recombination occurs more frequently in E. coli than reciprocal recombination [ 31 and is analogous to gene conversion in eukaryotic organisms. Our results indicate that mutagens which cause the rapid breakdown of DNA induce greater recombination than conditions which block DNA synthesis. Materials and methods Bacterial strains The E. coli K12 tion of the original dine operon in E. F’ was designated received from A.J. from J.C. Suit.
strains used in this study are described in Table 1. The isolaF’ plasmid and the identification of the genes of the histicoli have been described by Goldschmidt et al. [14]. This F30 by Bastarrachea and Willetts [2]. Strain AB 1157 was Clark and the poZA1 mutant [ll], strain 77 was received
Media Liquid cultures were grown in Difco antibiotic medium No. 3 (Penassay broth). Difco nutrient broth plus 1.5% Difco agar was the plating medium. The minimal medium of Vogel and Bonner [35] was used. Amino acids were added as needed to a final concentration of 20 pg/ml. Chemicals N-Methyl-N’-nitro-N-nitrosoguanidine (MNNG) was obtained from Aldrich Chemical Company. Diethyl sulfate (DES) and acridine orange (AO) were purchased from Fisher Scientific Company. Ethyl methanesulfonate (EMS) and methyl methanesulfonate (MMS) were obtained from Eastman Organic Chemicals Company. 2-Aminopurine (2AP) was purchased from Calbiochem
TABLE 1 ESCHERICHIA
COLI
University of Houston strain number
HETEROGENOTE
STRAINS
Genotype
source
or reference
UTH 4349
F’his+/hisG-E870,
UTH 5059
F’hisC780/hisI903,
pro
Cl41
UTH 5061
F’hisI903/hisC780.
pro
Recombinant isolated from UTH 5059
ile 2114
[I41
UTH 5231
F’hkG2743.
hisC780.
hisF504/AB1151
UTH 5232
F’hisG2743,
hisC180,
hisF504/thy,
UTH 6073
F’hisG2143.
hisC780,
hisF504/uvrC34
his+
polA1, his+
his+
F duction of E. coli AB1157 F duction of E. coli 77 [ill F duction of E. coli AB1884 Cl51 by T. Matney
18
Company. Methotrexate of Texas, M.D. Anderson Experimental
procedures:
(MTX) was furnished by Dr. T.L. Loo, the University Hospital and Tumor Institute. N-methyl-N’-nitro-N-nttrosoguanidine
Cultures were grown overnight in Penassay broth, diluted l/50 in the same medium and incubated for 2 h at 37°C on a New Brunswick rotary shaker. Inocula for overnight cultures were picked from single isolated colonies whose genotype had been determined previously. Stock solutions of MNNG (0.5 mg/ ml) in sterile broth were prepared for each experiment. Exponential phase cultures (4 X lo8 cells/ml) were treated with MNNG by addition of the appropriate amount of stock solution for 30 min at 37°C without agitation. Plate counts were made before and after treatment. Diethyl sulfate, ethyl methanesulfonate and methyl m,ethanesulfonate 5 ml of stationary phase culture (3 X lo9 cells/ml) were treated by addition
of either DES, EMS or MMS and allowed to incubate at 37°C without agitation. Treatment was stopped by a l/250 dilution into broth or by centrifugation and two saline washes. Treatment
with 2-aminopurine
and methotrexate
Exponential-phase cultures were washed and resuspended in minimal medium supplemented with adenine, methionine, proline and histidine and p20 mg/ml of either 2AP or MTX was added. After 9 h incubation at 37°C with agitation plate counts were done for comparison with untreated cultures. Ultraviolet
irradiation
Exponential-phase cells (8 X lo8 cells/ml) were diluted l/20 with saline into sterile glass petri dishes and irradiated with tandem GE germicidal lamps placed 75 cm from the target. Treatment
of DNA-repair-deficient
strains with mu tagens
Mutagen dosages were adjusted so that survival was 30 to 60% for comparison with DNA proficient strains (see Table 2). Elimination
of the F30 plasmid
by acridine
orange
Nutrient broth (2X) tubes containing 25 pug/ml of acridine orange were inoculated with 3 X lo4 F30 bacteria/ml and incubated in the dark at 37°C for 24 h. The cultures were serially diluted and spread onto the surface of nutrient agar plates. The loss of the F30 plasmid was detected by velveteen replicating the nutrient agar colonies to a minimally supplemented agar plate without histidine and to a minimal agar plate containing a lawn (0.1 ml of an overnight Pen assay broth culture spread to dryness) of a F- carrying a his mutation complemented by the his allele in the particular F30 strain [ 141. Less than 0.1% of the nutrient agar colonies from a F’ his’ strain (4349) grew on medium without histidine and were able to F-duce the F-His- recipient to His’ after A0 treatment. Expression
and analysis of treated
The technique
used to identify
bacteria
histidine
auxotrophic
recombinants
has been
19 TABLE
2
INDUCTION Treatment
OF RECOMBINANT.%
surviving fraction wo) W)
Control 2AP MTX EMS uv MNNG DES MMS
100 46 46 52 30 32 58 44
AUXOTROPHS,
AND HAPLOID
COLONIES
Total his homogenate colonies
Auxotrophic mutant colonies
Hishaploid colonies
@)
(%)
@Jo)
0.26 0.86 0.96 1.02 1.75 2.78 3.45 3.56
NI NI NI 2.35 NI 1.26 2.75 0.51
0.03 0.00 0.21 0.05 0.26 0.13 0.35 0.23
Total colonies examined
9120 818 942 1850 1142 1584 1710 1772
UTH 5059 was treated with, 2AP and MTX 200 I.tg/mleach for 9 h; EMS, 0.15 M for 30 min; UV-irradiation for 30 SW: MNNG, 10 pg/ml for 30 min; DES, 0.15 M for 30 min: and MMS. 0.15 M for 5 min. Homogenate recombinants, auxotrophic mutants, and haploid colonies are expressed as a percent of the total colonies examined. NI indicates that no derivatives were isolated.
described by us in detail [ 141. In brief, treated and untreated cultures were diluted l/250 in broth and grown overnight at 37°C with agitation to allow for phenotypic expression. The overnight cultures were diluted and plated on nutrient agar to yield isolated colonies. The nutrient agar colonies were replica plated to a minimally supplemented agar on which colonies with the parental genotype would grow. Colonies which did not grow were picked to nutrient agar master plates. His- derivatives were identified by velveteen replication to minimally supplemented agar with and without histidine. Colonies which failed to grow on the agar containing histidine were classified as having a new auxotropic mutation other than histidine biosynthesis. The mutational composition of the histidine-requiring derivatives was determined by replica plating to lawns of F- his strains which were spread onto minimal agar plates. Each F- strain used in this procedure contained one of the his alleles present in the original parental heterogenote. A colony which did not F-duce a given F- strain to His’ (i.e. no growth) indicated that the derivative was homozygous for that particular his allele (His- homogenote). His- homogenotes were distinguished from His- haploid colonies by their ability of Fduce to His’ (and thus act as a donor) a F- his strain with a mutation in a complementary histidine gene not present in the parental heterogenote. Re8dtS
Characterization of colonies isolated from F’ his’ heterogenotes of E. coli and the effect of mu tagens and chemotherapeutic agents Nearly 0.3% of the F’ His’ hetergenote colonies (UTH 5059, F’ hisC780/ hisI903, pro) grown on nutrient agar failed to grow on minimally supplemented agar. Non-replicating colonies were either His- homogenote recombinants, Hishaploids, or had an additional auxotrophic mutation. The genotype of these colonies was determined as described in Materials and Methods. Spontaneous and induced haploid colonies occurred at one tenth the
20
frequency of spontaneous and induced His- recombinant colonies, respectively (Table 2). The most frequent event in the F- partial diploid strain studied by Curtiss was haploidization [9]. These haploids were recombinant for exogenote allleles while haploid colonies arising in F30 partial diploid cultures are rarely recombinant [2]. Hfr colonies were not observed in cultures of F30 E. coli strains even when selective methods were employed. We looked for integration of an F’ his’ plasmid into an E. coli chromosome carrying a his allele after acridine orange treatment. No Hfr colonies were found in numerous experiments utilizing this selective technique. Most His- colonies, isolated from untreated cultures were His- homogenote recombinants either F’ hisC780/hisC780, pro or F’ hisI903/hisI903, pro. The spontaneous homogenote frequency of 0.26% for this F30 heterogenote is comparable with values obtained from other F’ plasmids [3,13]. When broth cultures were treated with chemotherapeutic agents or mutagens, we identified 0.85% to 3.55% of the surviving cells as His- homogenote recombinants. Representative experiments are shown in Table 2. All agents were tested at least twice and EMS and MMS were repeated 5 times. The variation in recombination frequency was less than 10%. Alkylating agents such as MNNG produced dose-dependent increases in recombination (Fig. 1) as reported for UV irradiation [9]. The most informative comparison of the effect of the mutagens on recombination can be made between EMS and MMS because they are homologous
-3.0
-2.0
g;
-1.5
2
-10
2 0, m m
a
-0.5
I-IS/ml
t
MNNG
Fig. 1. Effect of MNNG on growth and recombination. Log phase cells. UTH 5059, were treated with various concentrations of MNNG for 30 min at 37’C in Penassay broth. Viable cell counts were determined by serial dilution and plating on nutrient agar. The untreated cultures underwent two cell divisions during the 30 mm of treatment and the subsequent serial dilution procedure (15 mm). The viable cell count in the treated cultures were calculated as a percent of the viable cell count in the untreated culture. His- recombinants were determined after the 30-mm MNNG treatment as described in Materials and Methods.
21
compounds having a similar in vitro mode of action on oligomucleotides [27] and DNA [6,9,20,34]. In addition, the cells are treated in stationary phase for each of these agents thus eliminating differences in cellular physiology. EMS induced a higher frequency of auxotrophic mutants but a lower frequency of homogenote recombinant-s than MMS (Table 2). Origin of the recombinant homogeno te allele Greater than 75% of the His- recombinants isolated after treatment with all agents tested were homoiygous for the exogenic parental allele of UTH 5059, hisC780. In order to determine whether this was a position effect an experiment was done using an F30 heterogenote, UTH 5061, with transposed his alleles (i.e., F’ hisI903/hisC780). As before, most recombinants were homozygous for the exogenic parental allele i.e., F’ hisI903/hisI903 (Table 3). This position effect can account for the greater number of His- homogenotes recovered from strains which have 3 exogenic his alleles, UTH 5231, as compared to one exogenic his allele, UTH 5059 (see Tables 2 and 3). Recombinants, homozygous for hisC780 (an exogenic allele common to both heterogenote strains) were found at about the same frequency in UTH 5059 and UTH 5231. Thus an accurate comparison of nonreciprocal recombination between heterogenote strains carrying identical his alleles can only be accomplished if these alleles have the same location with respect to the F30 plasmid and the bacterial chromosome. Differential effect of mu tagens, chemotherapeutic agents and the polA1 allele on recombination The agents used in this study can be divided into two catagories based upon their effect on DNA metabolism and their recombinogenic effect. 2AP and MTX which block DNA synthesis by inhibiting precursor anabolism produced a 3- to 4-fold increase in recombination. Agents that are mutagenic by virtue of their ability to chemically modify DNA increased recombination 7- to 13-fold. A notable exception to this generalization is EMS which, as discussed above,
TABLE 3 EFFECT OF PARENTAL
strain
HIS ALLELE
LOCATION
ON MNNG INDUCED
RECOMBINATION Colonies examined
Total his homogenates
his C780 homogenates
his 1903 homogenates
(%b)
(S)
(%I
2.10
1.62
0.48
1050
2.94
0.59
2.36
1019
6059 F’ his I+ his C780 his 1903, his C+ 5061 F’ his 1903, his CT’ his I+. his C780 Log-phase bacterial cultures were treated with 10 Ccg/ml MNNG for 30 min and analyzed for recombinant6 as de&bed in the text.
22 TABLE 4 EFFECT OF THE POL Al AND UVR
strain
Treatment
C34 MUTATIONS
ON RECOMBINATION
His- haploid colonies
Auxotrophic colonies
Total his homogenates
(B)
(B)
WJ)
Colonies examined
5231 uvr
None uv MMS
NI 0.07 0.09
NI NI 0.47
0.85 2.89 (2.04) 6.17 (5.32)
2110 1316 1054
6073 u”r c34
None UV MMS
NI 0.11 NI
0.06 0.21 0.27
1.23 8.00 (6.77) 6.35 (5.12)
1625 1127 976
5232 poZA1
None uv MMS
0.05 0.22 0.13
0.11 0.33 NI
3.03 5.45 (2.42) 4.66 (1.63)
1882 934 1524
UTH 5231, 6073, 5232 were treated with UV irradiation for 30 sec. 3 set and 10 set, respectively and with MMS, 0.15 M. 0.15 M and 0.015 M, respectively each for 5 min. The numbers in parenthesis indicate the frequency of recombination minus the spontaneous frequency. NI indicates that no derivatives were isolated.
is highly mutagenic but produced only a 4-fold increase in recombination. If the recombinogenic effect of 2AP and MTX is due to inhibition of DNA synthesis, then the introduction of a defective DNA polymerase gene might, produce a similar increase in the frequency of spontaneous recombination as suggested in studies with bacteriophage polymerase mutants [4,5]. We therefore, constructed a F30 heterogenote with the polA1 mutation, UTH 5232. For comparative purposes, we constructed another UVs F30 strain (UTH 6073) which is unable to excise pyrimidine dimers due to the uurC34 mutation [ 161. The spontaneous recombination frequency of the polA1 heterogenote was nearly 4-fold greater than the spontaneous frequency of the ~021’ strain, UTH 5231 (Table 4) or much the same increase as observed with 2AP and MTX. UV-irradiation increased the frequency of recombination by nearly the same amount in the polA1 strain and in the UVr strain. MMS however, increased recombination in the polA1 mutant from 3% to 4.6%, a net increase of only 1.6%. The net increase in the uvrC34 mutant and the UV’ strain was about 5%. This result may be due to the sensitivity of polAl +ains to alkylating agents rather than indicating a primary role for polymerase I in MMSinduced recombination. The spontanous and MM&induced recombination frequency of the uurC34 mutant was similar respectively to that of the UV’ strain. UV induced recombination, however, was 3-fold greater in the uvrC34 strain as compared to the UV’ heterogenote. UV irradiation of the uurC34 strain did not significantly increase the frequency of auxotrophic mutants in comparison to unirradiated bacteria. Discussion We compared the effect of mutagens, chemotherapeutic agents which inhibit DNA synthesis, a DNA polymerase I mutation and a UV-excision-repair mutation on recombination in F’ E. coZi strains diploid for the histidine operon. The
23
assay involved determining the proportion of His- homogenote recombinants that arose from a His’ heterogenote. The technique of identifying His- recombinants also provided information on the proportion of haploid colonies, Hfr recombinants and colonies with an additional auxotrophic mutation resulting from the treatments. As expected the mutagenic agents increased the percentage of haploid and auxotrophic colonies. Hfr recombinants, however, were not observed. This result was unexpected and particularly interesting since the F30 plasmid can recombine in two ways. F30-bearing strains readily transfer the chromosome in bacterial matings, a process which requires the recA gene product [14] (T. Matney, personal communication). Recombination is also observed between histidine genes on the F factor and on the chromosome. The inability to isolate stable Hfr recombinants may be related to the arrangement of his genes on the F factor relative to the manner in which they were originally excised from the chromosome. Berg and Gallant [3] found 85% Hfr recombinants and 15% F’ homogenotes in a F’ heterogenote in which bacterial genes were attached to both sides of the F factor as it was excised from the chromosome. However, in another F’ strain, where the bacterial genes were attached to only one side of the F factor, they found 45% Hfr recombinants and 55% F’ homogenotes. Mutagens and the inhibitors of DNA synthesis, 2AP and MTX, increased the frequency of His- homogenote recombinants in F30 E. coli. Most of these recombinants were homozygous for parental exogenic alleles suggesting that exogenic homogenotes arise by a different pathway than endogenic homogenotes. Curtiss found a higher frequency of exogenic homogenates than endogenie homogenotes in a F- heterogenote and reached a similar conclusion [ 91. A defective DNA Polymerase I gene, polA1, was also found to increase the frequency of recombination. Increased spontaneous recombination was not observed in a F’ E. coli with another UV’ mutation, uvrC34. The above observations are consistent with results obtained with DNA polymerase mutants of T4 bacteriophage [4,5], and with yeast [12] and bacteriophage [32] treated with another inhibitor of DNA synthesis, 5-fluorodeoxyuridine. One explanation for these observations was suggested by Radding [26]. He proposed that inhibition of DNA synthesis increases the frequency of recombination by prolonging the lifetime of heteroduplex structures (e.g., gene conversion in a Bacillus subtilis phage) [ 321. We suggest that recombination induced by blockage of DNA synthesis occurs at the level of replicative intermediates which contain numerous single-stranded regions. Such single-stranded discontinuities may be expected to remain intact for longer periods when DNA synthesis is inhibited [ 181. Thus it is more likely that these regions will become sites for heteroduplex formation and subsequently substrates for recombinational enzymes. Single-stranded regions associated with replicating DNA have been demonstrated in bacteriophage [10,36], bacteria [22,23] and in mammalian tissue culture cells [ 24,28,31]. We found a greater increase in recombination with mutagens than with 2AP, MTX, and a polA1 mutation. As discussed above the chemotherapeutic agents and the polA1 allele probably increase recombination by reducing the DNA synthetic capacity of the bacteria. Mutagens may also increase the frequency of recombination by introducing lesions into the duplex which inhibit DNA syn-
24
thesis. In addition, strand breaks in DNA caused by mutagens or breaks which are widened into single-stranded discontinuities may provide additional sites for recombinational enzymes. This hypothesis is supported by the observation that MMS was more recombinogenic but was less mutagenic than EMS. Since methylated purines are eliminated from DNA more readily than ethylated purines [ 6,20,27], and apurinic sites are substrates for strand breakage [ 19,341, methylation produces more single-stranded discontinuities than ethylation [27, 331. Consequently, a smaller proportion of methylated bases remain attached to DNA than ethylated bases, as replication proceeds, so that methylation induces less mutation but more recombination [33]. Differences in specific alkylation products formed in DNA by these two compounds may also play a role in their relative mutagenic effects. We propose, therefore, that chemicals or mutations which inhibit DNA synthesis promote recombination by prolonging the lifetime of naturally occurring single-stranded regions in replicative intermediates. Mutagens may, in addition, produce lesions which increase the number of single-stranded regions thus inducing even higher frequencies of recombination. Carcinogens that are mutagenic can alter the genotype of an organism by inducing chemical modifications in DNA. We, (Table 2) and others [1,25,29, 371 have demonstrated that mutagens can also induce genotypic changes by nonreciprocal recombination. Inhibition of DNA replication by nonmutagenic substances [12,13,32] (Table 2) and genetic defects [4,5,18] (Table 4) that interfere with the metabolism of DNA may also alter the genotype of an organism. This latter alteration, however, presumably occurs by recombinational mechanisms alone rather than by a mutagenic pathway as well. In this study we demonstrate more UV-induced genotypic changes due to recombination in an E. coli strain (uurC34) defective in the excision of UV photoproducts than in an excision-proficient strain. UV-induced mutagenesis was not significantly increased in the uurC34 strain. This confirms and extends a previous study that showed elevated recombination of UV-irradiated bacteriophage grown in an E. coli strain with the uurC34 mutation [ 11. These observations are particularly interesting in that individuals with congenital UV-excision-repair deficiencies (xeroderma pigmentosa) are highly prone to develop cutaneous neoplasms after exposure to UV [28]. Cells which are phenotypioally DNA-repairdeficient such as peripheral blood lymphocytes may also be more likely to incur genotypic modification than DNA-repair-proficient cells [30]. In this regard it may be significant that a potent carcinogen, MNNG, has been shown to transform quiescent human lymphocytes, in vitro, into permanently proliferative cell lines which express viral genes [ 151. Further investigations will be required to determine if there are any relationships between genotypic alterations induced by recombinational mechanisms and the carcinogenic potential of mutagenic and nonmutagenic substances. Acknowledgement One of us (A.N.) appreciates the support, in part, of the James Hilton Manning and Emma Austin Manning Laboratory, Montefiore Hospital and Medical Center, New York, New York and USPHS Grant HL17417 during the preparation of this manuscript.
References 1 Baker, R.M., and R.H. Haynes, UV-induced enhancement of recombination among lambada bacteriophages in UV-sensitive host bacteria, Mol. Gem Genet.. 100 (1967) 166-177. 2 Bastarrachea. F., and N. WilIetts. The elimination by acridine orange of F30 from recombination deficient strains of Escherichia co2i K-12. Genetics, 59 (1968) 153-166. 3 Berg. D.E.. and J.A. Gallant. Test of reciprocality in crossing over in partially diploid strains of Escherichia coli, Genetics, 68 (1971) 457472. 4 Berger. H., A.J. Warren and K.E. Fry, Variations in genetic recombination due to amber mutations in TqD bacteriophage, J. Viral., 3 (1969) 171-175. 5 Bernstein, H., Repair and recombination in phage T4, Genes affecting recombination, Cold Spring Harbor Symp. Quant. Biol.. 33 (1969) 325-331. 6 Brooks, P., and P.D. Lawley, Effects of alkylating agents on T2 and T4 bacteriophages, Biochem. J., 89 (1963) 138-144. 7 Clark. A.J.. Recombination deficient mutants of E. coli and other bacteria, Annu. Rev. Genet., 7 (1973) 67-86. 8 Clark, A.J., Toward a metabolic interpretation of genetic recombination of E. coli and its phages, Annu. Rev. Microbial., 25 (1971) 437464. 9 Curtiss. R.. UV-induced recombination in a partially diploid strain of Escherichia coli, Genetics, 58 (1968) 9-54. 10 Delius, H., C. Howe and A. Koyensky, Structure of the replicating DNA from bacteriophage T4, Proc. Natl. Acad. Sci. (U.S.A.), 68 (1971) 3049-3053. 11 DeLucia, P., and J. Cairns. Isolation of an E. coli strain with a mutation affecting DNA polymerase, Nature (London), 224 (1969) 1164-1166. 12 Esposito, R.E., and R. Holliday, The effect of 5-fluorodeoxyuridine or genetic replication and mitotic crossing-over in synchronized cultures of Ustilago maydis, Genetics, 50 (1964) 1009-1017. 13 Gallant. J., and T. Spottswood, The recombinogenic effect of thymidylate starvation in Escherichia coli, Genetics, 52 (1965) 107-118. 14 Goldschmidt, E.P.. MS. Cater. T.S. Matney, M.A. Butler and A. Greene, Genetic analysis of the histidine operon in Escherichia coli K-12, Genetics, 66 (1970) 219-229. 15 Henderson. E.E., A.J. Norin and B.S. Strauss. Induction of permanently proliferating human lymphoblastoid lines by N-methyl-N’-nitro-N-nitrosoguanidine, Cancer Res., 35 (1975) 358-363. 16 Howard-Flanders, P., R.P. Boyce and L. Theriot, Three loci in Escherichia coli K-12 that control the excision of pyrimidine dimers and certain other mutagen products from DNA, Genetics, 53 (1966) 1119-1136. 17 Jacob, F., and E.L. Wollman, Etude g&&ique d’un bact&iophage tempPre d%scherichia coli, Effet du rayonnement ultraviolet sur la recombination &n&ique. Ann. Inst. Pasteur, 88 (1955) 724-749. 18 Krisch, H.M., N.V. Hamlett and H. Berger. Polynucleotide Iigase in bacteriophage T4D recombination, Genetics, 72 (1972) 187-203. 19 Lindahl. T.. and A. Anderson, Rate of chain breakage at apurinic sites in double-stranded deoxyribonucleic acid, Biochemistry, 11 (1972) 3618-3623. ,20 Lawley, P.D.. and P. Brookes, Further studies on the alkylation of nucleic acids and their constituent nucleotides, Biochem. J., 89 (1963) 127-138. 21 Meselson. M.S., and C.M. Radding, A general model for genetic recombination, Proc. Natl. Acad. Sci. (U.S.A.), 72 (1975) 358361. 22 Oishi, M., Studies of DNA replication in viva, 1. Isolation of the first intermediate of DNA replication in bacteria as single-stranded DNA, Proc. Natl. Acad. Sci. (U.S.A.), 60 (1968) 329-336. 23 Okazaki. R.. T. Okayaki, K. Sakabe, K. Sugimoto and A. Sugino, Mechanism of DNA chain growth. 1. Possible discontinuity and unusual secondary structure of newly synthesized chains. Proc. Natl. Acad. Sci. (U.S.A.), 59 (1968) 598-605. 24 Painter, R., and A. Schaeffer, State of newly synthesized Hela DNA. Nature (London), 221 (1969) 1215-1217. 25 Perry, J.M., Comparison of the effects of ultraviolet light and ethyl methanesulfonate upon the frequency of mitotic recombination in yeast. Mol. Gen. Genet., 106 (1969) 62-72. 26 Radding. C.M., Molecular mechanisms in genetic recombination, Annu. Rev. Genet.. 7 (1973) 87-111. 27 Rhaese. H.J., and E Freese, Chemical analvsis of DNA alterations, IV. Reactions of oligodeoxynucleotides with monofunctional alkylating agents leading to backbone breakage. Biochim. Biophys. Acta. 190 (1969) 418-433. 28 Robbins, J.H., K. Kraemer, M. Lutzner, B. Festoff and H. Coon, Xeroderma pigmentosum, An inherited disease with sun sensitivity, multiple cutaneous neoplasms. and abnormal DNA repair, Annu. Int. Med., 80 (1974) 221-248.
26 29 30 31 32 33 34 35 36 37
Schanfield. B., and E. Kafer, Chemical induction of mitotic recombination in Aspergillus nidulans, Genetics, 67 (1971) 209-219. Scudiero, D.. A.J. Norin, P. Karran and B. Strauss, DNA excision-repair deficiency of human peripheral blood lymphocytes treated with chemical carcinogens, Cancer Res.. 36 (1976) 1397-1403. Scudiero. D., and B. Strauss, Accumulation of single-stranded regions in DNA and the block to replication in a human cell line alkylated with methyl methanesulfonate, J. Mol. Biol., 83 (1974) 17-34. Spatz, H.C., and T.A. Trautner. One way to do experiments on gene conversion. Transfection with heteroduplex Sppl DNA, Mol. Gen. Genet., 109 (1970) 84-106. Strauss, B.. M. Coyle and M. Robbins, Alkylation damage and its repair, Cold Spring Harbor Symp. Quant. Biol., 33 (1968) 277-287. Strauss, B.. and T. Hill, The intermediate in the degradation of DNA alkylated with a monofunctional alkylating agent, Biochim. Biophys. Acta. 213 (1970) 14-25. Vogel, H.J., and D.M. Banner, Acetylornithinase of Escherichia coli; partial purification and some properties, J. Biol. Chem., 218 (1956) 97-106. Wolfson, J., and D. Dressier. Regions of single-stranded DNA in teh growing points of replicating bacteriophage T7 crhomosome. Proc. Natl. Acad. Sci. (U.S.A.), 69 (1972) 2682-2686. Zimmerman, F.K., and R. Schwa&. Induction of mitotic gene conversion with nitrous acid, l-methyl3-nitro-1-nitrosoguanidine, and other alkylating agents in Saccharomyces cereuisiae, Mol. Gen. Genet., 100 (1967) 63-76.
Mutation Research, 59 (1979) 15-26 @ Elsevier/North-Holland Biomedical Press
EFFECT OF MUTAGENS, CHEMOTHERAPEUTIC IN DNA REPAIR GENES ON RECOMBINATION ESCHERZCHZA COLZ
AGENTS AND DEFECTS IN F’ PARTIAL DIPLOID
ALLEN J. NORIN a and EUGENE P. GOLDSCHMIDT University of Houston, Department
of Biology, Houston,
Texas 77004 (U.S.A.)
(Received 25 April 1978) (Revision received 18 August 1978) (Accepted 28 August 1978)
Summary The ability of mutagenic agents, nonmutagenic substances and defects in DNA repair to alter the genotype of F’ partial diploid (F30) Escherichiu coli was determined. The frequency of auxotrophic mutants and histidine requiring (His-) haploid colonies was increased by mutagen treatment but Hfr colonies were not detected in F30 E. coli even with specific selection techniques. Genotype changes due to nonreciprocal recombination were determined by measuring the frequency of His- homogenotes, eg. F’ hisC780, hisI’/hisC780, hisI’, arising from a His’ heterogenote, F’ hisC780 hisI’/hisC’, his1903. At least 75% of the recombinants were homozygous for histidine alleles which were present on the F’ plasmid (exogenote) of the parental hetergenote rather than for histidine alleles on the chromosome. Mutagens, chemotherapeutic agents which block DNA synthesis and a defective DNA polymerase I gene, polA1, were a Present address: Montefiore Hospital and Medical Center, C-416.111 East 210th Street, Bronx, N.Y. 10467 (U.S.A.)
Albert
Einstein
College
of Medicine,
and nomenclature: His-, phenotypic designation indicating a histidine requirement for growth; His+, growth is independent of histidine; his, genotypic designation indicating a defective gene for histidine biosynthesis; his+, genotypicslly His+; Hfr. an E. coli containing and F-factor which is integrated into the chromosome and is able to transfer the genetic material of this cell into an F- recipient; F’, an F factor that has a portion of the bacterial genome attached to it and coexists with but is not integrated into the chromosome: F-duce or F-duction. injection of the donor F-factor into a F- recipient via conjugation: F30, an F’ plasmid of E. coli which carries the genes for histidine biosynthesis; In describing partial diploids, the mutational content of the F’ will be given first, eg., strain 5059 is F’ hisC780/hisI903, pro signifying that the F’ plasmid (exogenote) carries the 780 mutation in the histidine operon C gene while the chromosome (endogenote) has the 903 mutation in the I gene as well as a second mutation in a gene involving proline biosynthesis. MNNG. N-methyl-N’-nitro-N-nitrosoguanidine; DES, diethyl sulfate; AO. a&dine orange; EMS, ethyl methanesulfonate; MMS, methyl methanesulfonate; 2AP, 2-aminopurine; MTX, methotrexate; UV. ultraviolet light; UVr, phenotypically resistant to UV-irradiation. wild-type; UVs, phenotypically sensitive to UV-irradiation. Abbreviations
16
found to increase the frequency of nonreciprocal recombination. A defect in the ability to excise thymine dimers, uurC34, did not increase spontaneous UV irradiation but not methyl nonreciprocal recombination. However, methanesulfonate (MMS) induced greater recombination in this excision-repair defective mutant than in DNA-repair-proficient strains. Mutagenic agents, with the exception of ethyl methanesulfonate (EMS), induced greater increases in recombination than the chemotherapeutic agents or the poZA1 mutation. EMS, which causes relatively little degradation of DNA, was more mutagenic but less recombinogenic than MMS, a homologous compound that induces rapid degradation of DNA. One explanation for these results is that inhibition of DNA synthesis increases recombination by prolonging the lifetime of naturally occurring single-stranded regions in replicative intermediates of the DNA. Mutagens which cause the rapid breakdown of DNA may, in addition, introduce lesions into the genome that increase the number of single-stranded regions thus inducing even higher frequencies of recombination.
Introduction Genotypic alterations regardless of their cause may have a profound effect on the functioning of cells. A number of reports indicate that both mutagenic and nonmutagenic substances can change the genotype of an organism by recombinational mechanisms. Mutagenic agents increase the frequency of recombination in micro-organisms [ 9,17,25,29,37]. Substances which inhibit DNA synthesis but do not react chemically with DNA also increase recombination [ 12,13,32]. Elevated frequencies of recombination are observed in mutants of T4 bacteriophage which are defective in intermediary DNA metabolism, DNA polymerase or DNA ligase activity [4,5,18]. The mechanisms by which the above agents and genetic defects increase recombination are not entirely clear. Presumably, mutagens induce a variety of perturbations in DNA which initiate recombinational events [21]. It is not surprising that chemotherapeutic agents and genetic defects that inhibit macromolecular DNA metabolism increase recombination since the processes of replication, repair, recombination and transcription partially overlap [8,25]. In addition, the extent of the overlap may vary depending on the organism [7,26]. Thus the molecular details of recombination may be quite different in one organism compared to another. No studies have been published comparing the effect of the above agents and genetic defects in DNA repair on recombination in one test organism. Consequently, it is difficult to determine which lesion is most effective in producing recombinant molecules in a given organism. This study was undertaken, therefore, to compare the relative effect of mutagens, chemotherapeutic agents which inhibit DNA synthesis and the effect of two different DNA-repair mutations (poZA1 and uurC34) on recombination in Escherichiu coli K12.
Recombination was studied in an F’ E. coli (F30) which is partially diploid for the histidine operon [2,14]. F’ strains are particularly advantageous for studies of this type since the effect of the experimental condition is on recombination rather than on chromosome transfer during conjugation. We observed
17
recombination between histidine alleles on the chromosome and those on the F’ plasmid. For example, recombination was determined by measuring the number of homozygous F’ hisC780 (plasmid genotype)@.&780 (chromosome genotype) colonies (His- homogenotes) which arose from a heterozygous F’ hisC’ISO/hisC’ parent (His’ heterogenote). Nonreciprocal recombination occurs more frequently in E. coli than reciprocal recombination [ 31 and is analogous to gene conversion in eukaryotic organisms. Our results indicate that mutagens which cause the rapid breakdown of DNA induce greater recombination than conditions which block DNA synthesis. Materials and methods Bacterial strains The E. coli K12 tion of the original dine operon in E. F’ was designated received from A.J. from J.C. Suit.
strains used in this study are described in Table 1. The isolaF’ plasmid and the identification of the genes of the histicoli have been described by Goldschmidt et al. [14]. This F30 by Bastarrachea and Willetts [2]. Strain AB 1157 was Clark and the poZA1 mutant [ll], strain 77 was received
Media Liquid cultures were grown in Difco antibiotic medium No. 3 (Penassay broth). Difco nutrient broth plus 1.5% Difco agar was the plating medium. The minimal medium of Vogel and Bonner [35] was used. Amino acids were added as needed to a final concentration of 20 pg/ml. Chemicals N-Methyl-N’-nitro-N-nitrosoguanidine (MNNG) was obtained from Aldrich Chemical Company. Diethyl sulfate (DES) and acridine orange (AO) were purchased from Fisher Scientific Company. Ethyl methanesulfonate (EMS) and methyl methanesulfonate (MMS) were obtained from Eastman Organic Chemicals Company. 2-Aminopurine (2AP) was purchased from Calbiochem
TABLE 1 ESCHERICHIA
COLI
University of Houston strain number
HETEROGENOTE
STRAINS
Genotype
source
or reference
UTH 4349
F’his+/hisG-E870,
UTH 5059
F’hisC780/hisI903,
pro
Cl41
UTH 5061
F’hisI903/hisC780.
pro
Recombinant isolated from UTH 5059
ile 2114
[I41
UTH 5231
F’hkG2743.
hisC780.
hisF504/AB1151
UTH 5232
F’hisG2743,
hisC180,
hisF504/thy,
UTH 6073
F’hisG2143.
hisC780,
hisF504/uvrC34
his+
polA1, his+
his+
F duction of E. coli AB1157 F duction of E. coli 77 [ill F duction of E. coli AB1884 Cl51 by T. Matney
18
Company. Methotrexate of Texas, M.D. Anderson Experimental
procedures:
(MTX) was furnished by Dr. T.L. Loo, the University Hospital and Tumor Institute. N-methyl-N’-nitro-N-nttrosoguanidine
Cultures were grown overnight in Penassay broth, diluted l/50 in the same medium and incubated for 2 h at 37°C on a New Brunswick rotary shaker. Inocula for overnight cultures were picked from single isolated colonies whose genotype had been determined previously. Stock solutions of MNNG (0.5 mg/ ml) in sterile broth were prepared for each experiment. Exponential phase cultures (4 X lo8 cells/ml) were treated with MNNG by addition of the appropriate amount of stock solution for 30 min at 37°C without agitation. Plate counts were made before and after treatment. Diethyl sulfate, ethyl methanesulfonate and methyl m,ethanesulfonate 5 ml of stationary phase culture (3 X lo9 cells/ml) were treated by addition
of either DES, EMS or MMS and allowed to incubate at 37°C without agitation. Treatment was stopped by a l/250 dilution into broth or by centrifugation and two saline washes. Treatment
with 2-aminopurine
and methotrexate
Exponential-phase cultures were washed and resuspended in minimal medium supplemented with adenine, methionine, proline and histidine and p20 mg/ml of either 2AP or MTX was added. After 9 h incubation at 37°C with agitation plate counts were done for comparison with untreated cultures. Ultraviolet
irradiation
Exponential-phase cells (8 X lo8 cells/ml) were diluted l/20 with saline into sterile glass petri dishes and irradiated with tandem GE germicidal lamps placed 75 cm from the target. Treatment
of DNA-repair-deficient
strains with mu tagens
Mutagen dosages were adjusted so that survival was 30 to 60% for comparison with DNA proficient strains (see Table 2). Elimination
of the F30 plasmid
by acridine
orange
Nutrient broth (2X) tubes containing 25 pug/ml of acridine orange were inoculated with 3 X lo4 F30 bacteria/ml and incubated in the dark at 37°C for 24 h. The cultures were serially diluted and spread onto the surface of nutrient agar plates. The loss of the F30 plasmid was detected by velveteen replicating the nutrient agar colonies to a minimally supplemented agar plate without histidine and to a minimal agar plate containing a lawn (0.1 ml of an overnight Pen assay broth culture spread to dryness) of a F- carrying a his mutation complemented by the his allele in the particular F30 strain [ 141. Less than 0.1% of the nutrient agar colonies from a F’ his’ strain (4349) grew on medium without histidine and were able to F-duce the F-His- recipient to His’ after A0 treatment. Expression
and analysis of treated
The technique
used to identify
bacteria
histidine
auxotrophic
recombinants
has been
19 TABLE
2
INDUCTION Treatment
OF RECOMBINANT.%
surviving fraction wo) W)
Control 2AP MTX EMS uv MNNG DES MMS
100 46 46 52 30 32 58 44
AUXOTROPHS,
AND HAPLOID
COLONIES
Total his homogenate colonies
Auxotrophic mutant colonies
Hishaploid colonies
@)
(%)
@Jo)
0.26 0.86 0.96 1.02 1.75 2.78 3.45 3.56
NI NI NI 2.35 NI 1.26 2.75 0.51
0.03 0.00 0.21 0.05 0.26 0.13 0.35 0.23
Total colonies examined
9120 818 942 1850 1142 1584 1710 1772
UTH 5059 was treated with, 2AP and MTX 200 I.tg/mleach for 9 h; EMS, 0.15 M for 30 min; UV-irradiation for 30 SW: MNNG, 10 pg/ml for 30 min; DES, 0.15 M for 30 min: and MMS. 0.15 M for 5 min. Homogenate recombinants, auxotrophic mutants, and haploid colonies are expressed as a percent of the total colonies examined. NI indicates that no derivatives were isolated.
described by us in detail [ 141. In brief, treated and untreated cultures were diluted l/250 in broth and grown overnight at 37°C with agitation to allow for phenotypic expression. The overnight cultures were diluted and plated on nutrient agar to yield isolated colonies. The nutrient agar colonies were replica plated to a minimally supplemented agar on which colonies with the parental genotype would grow. Colonies which did not grow were picked to nutrient agar master plates. His- derivatives were identified by velveteen replication to minimally supplemented agar with and without histidine. Colonies which failed to grow on the agar containing histidine were classified as having a new auxotropic mutation other than histidine biosynthesis. The mutational composition of the histidine-requiring derivatives was determined by replica plating to lawns of F- his strains which were spread onto minimal agar plates. Each F- strain used in this procedure contained one of the his alleles present in the original parental heterogenote. A colony which did not F-duce a given F- strain to His’ (i.e. no growth) indicated that the derivative was homozygous for that particular his allele (His- homogenote). His- homogenotes were distinguished from His- haploid colonies by their ability of Fduce to His’ (and thus act as a donor) a F- his strain with a mutation in a complementary histidine gene not present in the parental heterogenote. Re8dtS
Characterization of colonies isolated from F’ his’ heterogenotes of E. coli and the effect of mu tagens and chemotherapeutic agents Nearly 0.3% of the F’ His’ hetergenote colonies (UTH 5059, F’ hisC780/ hisI903, pro) grown on nutrient agar failed to grow on minimally supplemented agar. Non-replicating colonies were either His- homogenote recombinants, Hishaploids, or had an additional auxotrophic mutation. The genotype of these colonies was determined as described in Materials and Methods. Spontaneous and induced haploid colonies occurred at one tenth the
20
frequency of spontaneous and induced His- recombinant colonies, respectively (Table 2). The most frequent event in the F- partial diploid strain studied by Curtiss was haploidization [9]. These haploids were recombinant for exogenote allleles while haploid colonies arising in F30 partial diploid cultures are rarely recombinant [2]. Hfr colonies were not observed in cultures of F30 E. coli strains even when selective methods were employed. We looked for integration of an F’ his’ plasmid into an E. coli chromosome carrying a his allele after acridine orange treatment. No Hfr colonies were found in numerous experiments utilizing this selective technique. Most His- colonies, isolated from untreated cultures were His- homogenote recombinants either F’ hisC780/hisC780, pro or F’ hisI903/hisI903, pro. The spontaneous homogenote frequency of 0.26% for this F30 heterogenote is comparable with values obtained from other F’ plasmids [3,13]. When broth cultures were treated with chemotherapeutic agents or mutagens, we identified 0.85% to 3.55% of the surviving cells as His- homogenote recombinants. Representative experiments are shown in Table 2. All agents were tested at least twice and EMS and MMS were repeated 5 times. The variation in recombination frequency was less than 10%. Alkylating agents such as MNNG produced dose-dependent increases in recombination (Fig. 1) as reported for UV irradiation [9]. The most informative comparison of the effect of the mutagens on recombination can be made between EMS and MMS because they are homologous
-3.0
-2.0
g;
-1.5
2
-10
2 0, m m
a
-0.5
I-IS/ml
t
MNNG
Fig. 1. Effect of MNNG on growth and recombination. Log phase cells. UTH 5059, were treated with various concentrations of MNNG for 30 min at 37’C in Penassay broth. Viable cell counts were determined by serial dilution and plating on nutrient agar. The untreated cultures underwent two cell divisions during the 30 mm of treatment and the subsequent serial dilution procedure (15 mm). The viable cell count in the treated cultures were calculated as a percent of the viable cell count in the untreated culture. His- recombinants were determined after the 30-mm MNNG treatment as described in Materials and Methods.
21
compounds having a similar in vitro mode of action on oligomucleotides [27] and DNA [6,9,20,34]. In addition, the cells are treated in stationary phase for each of these agents thus eliminating differences in cellular physiology. EMS induced a higher frequency of auxotrophic mutants but a lower frequency of homogenote recombinant-s than MMS (Table 2). Origin of the recombinant homogeno te allele Greater than 75% of the His- recombinants isolated after treatment with all agents tested were homoiygous for the exogenic parental allele of UTH 5059, hisC780. In order to determine whether this was a position effect an experiment was done using an F30 heterogenote, UTH 5061, with transposed his alleles (i.e., F’ hisI903/hisC780). As before, most recombinants were homozygous for the exogenic parental allele i.e., F’ hisI903/hisI903 (Table 3). This position effect can account for the greater number of His- homogenotes recovered from strains which have 3 exogenic his alleles, UTH 5231, as compared to one exogenic his allele, UTH 5059 (see Tables 2 and 3). Recombinants, homozygous for hisC780 (an exogenic allele common to both heterogenote strains) were found at about the same frequency in UTH 5059 and UTH 5231. Thus an accurate comparison of nonreciprocal recombination between heterogenote strains carrying identical his alleles can only be accomplished if these alleles have the same location with respect to the F30 plasmid and the bacterial chromosome. Differential effect of mu tagens, chemotherapeutic agents and the polA1 allele on recombination The agents used in this study can be divided into two catagories based upon their effect on DNA metabolism and their recombinogenic effect. 2AP and MTX which block DNA synthesis by inhibiting precursor anabolism produced a 3- to 4-fold increase in recombination. Agents that are mutagenic by virtue of their ability to chemically modify DNA increased recombination 7- to 13-fold. A notable exception to this generalization is EMS which, as discussed above,
TABLE 3 EFFECT OF PARENTAL
strain
HIS ALLELE
LOCATION
ON MNNG INDUCED
RECOMBINATION Colonies examined
Total his homogenates
his C780 homogenates
his 1903 homogenates
(%b)
(S)
(%I
2.10
1.62
0.48
1050
2.94
0.59
2.36
1019
6059 F’ his I+ his C780 his 1903, his C+ 5061 F’ his 1903, his CT’ his I+. his C780 Log-phase bacterial cultures were treated with 10 Ccg/ml MNNG for 30 min and analyzed for recombinant6 as de&bed in the text.
22 TABLE 4 EFFECT OF THE POL Al AND UVR
strain
Treatment
C34 MUTATIONS
ON RECOMBINATION
His- haploid colonies
Auxotrophic colonies
Total his homogenates
(B)
(B)
WJ)
Colonies examined
5231 uvr
None uv MMS
NI 0.07 0.09
NI NI 0.47
0.85 2.89 (2.04) 6.17 (5.32)
2110 1316 1054
6073 u”r c34
None UV MMS
NI 0.11 NI
0.06 0.21 0.27
1.23 8.00 (6.77) 6.35 (5.12)
1625 1127 976
5232 poZA1
None uv MMS
0.05 0.22 0.13
0.11 0.33 NI
3.03 5.45 (2.42) 4.66 (1.63)
1882 934 1524
UTH 5231, 6073, 5232 were treated with UV irradiation for 30 sec. 3 set and 10 set, respectively and with MMS, 0.15 M. 0.15 M and 0.015 M, respectively each for 5 min. The numbers in parenthesis indicate the frequency of recombination minus the spontaneous frequency. NI indicates that no derivatives were isolated.
is highly mutagenic but produced only a 4-fold increase in recombination. If the recombinogenic effect of 2AP and MTX is due to inhibition of DNA synthesis, then the introduction of a defective DNA polymerase gene might, produce a similar increase in the frequency of spontaneous recombination as suggested in studies with bacteriophage polymerase mutants [4,5]. We therefore, constructed a F30 heterogenote with the polA1 mutation, UTH 5232. For comparative purposes, we constructed another UVs F30 strain (UTH 6073) which is unable to excise pyrimidine dimers due to the uurC34 mutation [ 161. The spontaneous recombination frequency of the polA1 heterogenote was nearly 4-fold greater than the spontaneous frequency of the ~021’ strain, UTH 5231 (Table 4) or much the same increase as observed with 2AP and MTX. UV-irradiation increased the frequency of recombination by nearly the same amount in the polA1 strain and in the UVr strain. MMS however, increased recombination in the polA1 mutant from 3% to 4.6%, a net increase of only 1.6%. The net increase in the uvrC34 mutant and the UV’ strain was about 5%. This result may be due to the sensitivity of polAl +ains to alkylating agents rather than indicating a primary role for polymerase I in MMSinduced recombination. The spontanous and MM&induced recombination frequency of the uurC34 mutant was similar respectively to that of the UV’ strain. UV induced recombination, however, was 3-fold greater in the uvrC34 strain as compared to the UV’ heterogenote. UV irradiation of the uurC34 strain did not significantly increase the frequency of auxotrophic mutants in comparison to unirradiated bacteria. Discussion We compared the effect of mutagens, chemotherapeutic agents which inhibit DNA synthesis, a DNA polymerase I mutation and a UV-excision-repair mutation on recombination in F’ E. coZi strains diploid for the histidine operon. The
23
assay involved determining the proportion of His- homogenote recombinants that arose from a His’ heterogenote. The technique of identifying His- recombinants also provided information on the proportion of haploid colonies, Hfr recombinants and colonies with an additional auxotrophic mutation resulting from the treatments. As expected the mutagenic agents increased the percentage of haploid and auxotrophic colonies. Hfr recombinants, however, were not observed. This result was unexpected and particularly interesting since the F30 plasmid can recombine in two ways. F30-bearing strains readily transfer the chromosome in bacterial matings, a process which requires the recA gene product [14] (T. Matney, personal communication). Recombination is also observed between histidine genes on the F factor and on the chromosome. The inability to isolate stable Hfr recombinants may be related to the arrangement of his genes on the F factor relative to the manner in which they were originally excised from the chromosome. Berg and Gallant [3] found 85% Hfr recombinants and 15% F’ homogenotes in a F’ heterogenote in which bacterial genes were attached to both sides of the F factor as it was excised from the chromosome. However, in another F’ strain, where the bacterial genes were attached to only one side of the F factor, they found 45% Hfr recombinants and 55% F’ homogenotes. Mutagens and the inhibitors of DNA synthesis, 2AP and MTX, increased the frequency of His- homogenote recombinants in F30 E. coli. Most of these recombinants were homozygous for parental exogenic alleles suggesting that exogenic homogenotes arise by a different pathway than endogenic homogenotes. Curtiss found a higher frequency of exogenic homogenates than endogenie homogenotes in a F- heterogenote and reached a similar conclusion [ 91. A defective DNA Polymerase I gene, polA1, was also found to increase the frequency of recombination. Increased spontaneous recombination was not observed in a F’ E. coli with another UV’ mutation, uvrC34. The above observations are consistent with results obtained with DNA polymerase mutants of T4 bacteriophage [4,5], and with yeast [12] and bacteriophage [32] treated with another inhibitor of DNA synthesis, 5-fluorodeoxyuridine. One explanation for these observations was suggested by Radding [26]. He proposed that inhibition of DNA synthesis increases the frequency of recombination by prolonging the lifetime of heteroduplex structures (e.g., gene conversion in a Bacillus subtilis phage) [ 321. We suggest that recombination induced by blockage of DNA synthesis occurs at the level of replicative intermediates which contain numerous single-stranded regions. Such single-stranded discontinuities may be expected to remain intact for longer periods when DNA synthesis is inhibited [ 181. Thus it is more likely that these regions will become sites for heteroduplex formation and subsequently substrates for recombinational enzymes. Single-stranded regions associated with replicating DNA have been demonstrated in bacteriophage [10,36], bacteria [22,23] and in mammalian tissue culture cells [ 24,28,31]. We found a greater increase in recombination with mutagens than with 2AP, MTX, and a polA1 mutation. As discussed above the chemotherapeutic agents and the polA1 allele probably increase recombination by reducing the DNA synthetic capacity of the bacteria. Mutagens may also increase the frequency of recombination by introducing lesions into the duplex which inhibit DNA syn-
24
thesis. In addition, strand breaks in DNA caused by mutagens or breaks which are widened into single-stranded discontinuities may provide additional sites for recombinational enzymes. This hypothesis is supported by the observation that MMS was more recombinogenic but was less mutagenic than EMS. Since methylated purines are eliminated from DNA more readily than ethylated purines [ 6,20,27], and apurinic sites are substrates for strand breakage [ 19,341, methylation produces more single-stranded discontinuities than ethylation [27, 331. Consequently, a smaller proportion of methylated bases remain attached to DNA than ethylated bases, as replication proceeds, so that methylation induces less mutation but more recombination [33]. Differences in specific alkylation products formed in DNA by these two compounds may also play a role in their relative mutagenic effects. We propose, therefore, that chemicals or mutations which inhibit DNA synthesis promote recombination by prolonging the lifetime of naturally occurring single-stranded regions in replicative intermediates. Mutagens may, in addition, produce lesions which increase the number of single-stranded regions thus inducing even higher frequencies of recombination. Carcinogens that are mutagenic can alter the genotype of an organism by inducing chemical modifications in DNA. We, (Table 2) and others [1,25,29, 371 have demonstrated that mutagens can also induce genotypic changes by nonreciprocal recombination. Inhibition of DNA replication by nonmutagenic substances [12,13,32] (Table 2) and genetic defects [4,5,18] (Table 4) that interfere with the metabolism of DNA may also alter the genotype of an organism. This latter alteration, however, presumably occurs by recombinational mechanisms alone rather than by a mutagenic pathway as well. In this study we demonstrate more UV-induced genotypic changes due to recombination in an E. coli strain (uurC34) defective in the excision of UV photoproducts than in an excision-proficient strain. UV-induced mutagenesis was not significantly increased in the uurC34 strain. This confirms and extends a previous study that showed elevated recombination of UV-irradiated bacteriophage grown in an E. coli strain with the uurC34 mutation [ 11. These observations are particularly interesting in that individuals with congenital UV-excision-repair deficiencies (xeroderma pigmentosa) are highly prone to develop cutaneous neoplasms after exposure to UV [28]. Cells which are phenotypioally DNA-repairdeficient such as peripheral blood lymphocytes may also be more likely to incur genotypic modification than DNA-repair-proficient cells [30]. In this regard it may be significant that a potent carcinogen, MNNG, has been shown to transform quiescent human lymphocytes, in vitro, into permanently proliferative cell lines which express viral genes [ 151. Further investigations will be required to determine if there are any relationships between genotypic alterations induced by recombinational mechanisms and the carcinogenic potential of mutagenic and nonmutagenic substances. Acknowledgement One of us (A.N.) appreciates the support, in part, of the James Hilton Manning and Emma Austin Manning Laboratory, Montefiore Hospital and Medical Center, New York, New York and USPHS Grant HL17417 during the preparation of this manuscript.
References 1 Baker, R.M., and R.H. Haynes, UV-induced enhancement of recombination among lambada bacteriophages in UV-sensitive host bacteria, Mol. Gem Genet.. 100 (1967) 166-177. 2 Bastarrachea. F., and N. WilIetts. The elimination by acridine orange of F30 from recombination deficient strains of Escherichia co2i K-12. Genetics, 59 (1968) 153-166. 3 Berg. D.E.. and J.A. Gallant. Test of reciprocality in crossing over in partially diploid strains of Escherichia coli, Genetics, 68 (1971) 457472. 4 Berger. H., A.J. Warren and K.E. Fry, Variations in genetic recombination due to amber mutations in TqD bacteriophage, J. Viral., 3 (1969) 171-175. 5 Bernstein, H., Repair and recombination in phage T4, Genes affecting recombination, Cold Spring Harbor Symp. Quant. Biol.. 33 (1969) 325-331. 6 Brooks, P., and P.D. Lawley, Effects of alkylating agents on T2 and T4 bacteriophages, Biochem. J., 89 (1963) 138-144. 7 Clark. A.J.. Recombination deficient mutants of E. coli and other bacteria, Annu. Rev. Genet., 7 (1973) 67-86. 8 Clark, A.J., Toward a metabolic interpretation of genetic recombination of E. coli and its phages, Annu. Rev. Microbial., 25 (1971) 437464. 9 Curtiss. R.. UV-induced recombination in a partially diploid strain of Escherichia coli, Genetics, 58 (1968) 9-54. 10 Delius, H., C. Howe and A. Koyensky, Structure of the replicating DNA from bacteriophage T4, Proc. Natl. Acad. Sci. (U.S.A.), 68 (1971) 3049-3053. 11 DeLucia, P., and J. Cairns. Isolation of an E. coli strain with a mutation affecting DNA polymerase, Nature (London), 224 (1969) 1164-1166. 12 Esposito, R.E., and R. Holliday, The effect of 5-fluorodeoxyuridine or genetic replication and mitotic crossing-over in synchronized cultures of Ustilago maydis, Genetics, 50 (1964) 1009-1017. 13 Gallant. J., and T. Spottswood, The recombinogenic effect of thymidylate starvation in Escherichia coli, Genetics, 52 (1965) 107-118. 14 Goldschmidt, E.P.. MS. Cater. T.S. Matney, M.A. Butler and A. Greene, Genetic analysis of the histidine operon in Escherichia coli K-12, Genetics, 66 (1970) 219-229. 15 Henderson. E.E., A.J. Norin and B.S. Strauss. Induction of permanently proliferating human lymphoblastoid lines by N-methyl-N’-nitro-N-nitrosoguanidine, Cancer Res., 35 (1975) 358-363. 16 Howard-Flanders, P., R.P. Boyce and L. Theriot, Three loci in Escherichia coli K-12 that control the excision of pyrimidine dimers and certain other mutagen products from DNA, Genetics, 53 (1966) 1119-1136. 17 Jacob, F., and E.L. Wollman, Etude g&&ique d’un bact&iophage tempPre d%scherichia coli, Effet du rayonnement ultraviolet sur la recombination &n&ique. Ann. Inst. Pasteur, 88 (1955) 724-749. 18 Krisch, H.M., N.V. Hamlett and H. Berger. Polynucleotide Iigase in bacteriophage T4D recombination, Genetics, 72 (1972) 187-203. 19 Lindahl. T.. and A. Anderson, Rate of chain breakage at apurinic sites in double-stranded deoxyribonucleic acid, Biochemistry, 11 (1972) 3618-3623. ,20 Lawley, P.D.. and P. Brookes, Further studies on the alkylation of nucleic acids and their constituent nucleotides, Biochem. J., 89 (1963) 127-138. 21 Meselson. M.S., and C.M. Radding, A general model for genetic recombination, Proc. Natl. Acad. Sci. (U.S.A.), 72 (1975) 358361. 22 Oishi, M., Studies of DNA replication in viva, 1. Isolation of the first intermediate of DNA replication in bacteria as single-stranded DNA, Proc. Natl. Acad. Sci. (U.S.A.), 60 (1968) 329-336. 23 Okazaki. R.. T. Okayaki, K. Sakabe, K. Sugimoto and A. Sugino, Mechanism of DNA chain growth. 1. Possible discontinuity and unusual secondary structure of newly synthesized chains. Proc. Natl. Acad. Sci. (U.S.A.), 59 (1968) 598-605. 24 Painter, R., and A. Schaeffer, State of newly synthesized Hela DNA. Nature (London), 221 (1969) 1215-1217. 25 Perry, J.M., Comparison of the effects of ultraviolet light and ethyl methanesulfonate upon the frequency of mitotic recombination in yeast. Mol. Gen. Genet., 106 (1969) 62-72. 26 Radding. C.M., Molecular mechanisms in genetic recombination, Annu. Rev. Genet.. 7 (1973) 87-111. 27 Rhaese. H.J., and E Freese, Chemical analvsis of DNA alterations, IV. Reactions of oligodeoxynucleotides with monofunctional alkylating agents leading to backbone breakage. Biochim. Biophys. Acta. 190 (1969) 418-433. 28 Robbins, J.H., K. Kraemer, M. Lutzner, B. Festoff and H. Coon, Xeroderma pigmentosum, An inherited disease with sun sensitivity, multiple cutaneous neoplasms. and abnormal DNA repair, Annu. Int. Med., 80 (1974) 221-248.
26 29 30 31 32 33 34 35 36 37
Schanfield. B., and E. Kafer, Chemical induction of mitotic recombination in Aspergillus nidulans, Genetics, 67 (1971) 209-219. Scudiero, D.. A.J. Norin, P. Karran and B. Strauss, DNA excision-repair deficiency of human peripheral blood lymphocytes treated with chemical carcinogens, Cancer Res.. 36 (1976) 1397-1403. Scudiero. D., and B. Strauss, Accumulation of single-stranded regions in DNA and the block to replication in a human cell line alkylated with methyl methanesulfonate, J. Mol. Biol., 83 (1974) 17-34. Spatz, H.C., and T.A. Trautner. One way to do experiments on gene conversion. Transfection with heteroduplex Sppl DNA, Mol. Gen. Genet., 109 (1970) 84-106. Strauss, B.. M. Coyle and M. Robbins, Alkylation damage and its repair, Cold Spring Harbor Symp. Quant. Biol., 33 (1968) 277-287. Strauss, B.. and T. Hill, The intermediate in the degradation of DNA alkylated with a monofunctional alkylating agent, Biochim. Biophys. Acta. 213 (1970) 14-25. Vogel, H.J., and D.M. Banner, Acetylornithinase of Escherichia coli; partial purification and some properties, J. Biol. Chem., 218 (1956) 97-106. Wolfson, J., and D. Dressier. Regions of single-stranded DNA in teh growing points of replicating bacteriophage T7 crhomosome. Proc. Natl. Acad. Sci. (U.S.A.), 69 (1972) 2682-2686. Zimmerman, F.K., and R. Schwa&. Induction of mitotic gene conversion with nitrous acid, l-methyl3-nitro-1-nitrosoguanidine, and other alkylating agents in Saccharomyces cereuisiae, Mol. Gen. Genet., 100 (1967) 63-76.
