JOURNAL OF BACTERIOLOGY, June 1990,

p.

Vol. 172, No. 6

3009-3014

0021-9193/90/063009-06$02.00/0 Copyright © 1990, American Society for Microbiology

Specificity of the Mutator Effect Caused by Disruption of the RADI Excision Repair Gene of Saccharomyces cerevisiae BERNARD A. KUNZ,* LESTER KOHALMI, XIAOLIN KANG, AND KATHRYN A. MAGNUSSON Department of Microbiology, The University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2 Received 11 December 1989/Accepted 20 March 1990 - Disruption of RADI, a gene controlling excision repair in the yeast Saccharomyces cerevisiae, increased the frequency of spontaneous forward mutation in a plasmid-borne copy of the SUP4-o gene. To characterize this effect in detail, a collection of 249 SUP4-o mutations arising spontaneously in the radl strain was analyzed by DNA sequencing. The resulting mutational spectrum was compared with that derived from an examination of 322 spontaneous SUP4-o mutations selected in an isogenic wild-type (RADI) strain. This comparison revealed that the radl mutator phenotype was associated with increases in the frequencies of single-base-pair substitution, single-base-pair deletion, -and insertion of the yeast retrotransposon Ty. In the radl strain, the relative fractions of these events and their distributions within SUP4-o exhibited features similar to those for spontaneous mutagenesis in the isogenic.RADI background. The increase in the frequency of Ty insertion argues that Ty transposition can be activated by unrepaired spontaneous DNA damage, which normally would be removed by excision repair. We discuss the possibilities that either translesion synthesis, a reduced fidelity of DNA replication, or a deficiency in mismatch correction might be responsible for the majority of single-base-pair events in the radl strain.

mutation, perhaps by somehow reducing the fidelity. of DNA replication (43) or the efficiency of correcting replication errors (26). Alternatively, the products of some repair genes might play a regulatory role (7, 16) so that modulation of processes other than DNA repair could be responsible for the mutator phenotypes of certain repair-deficient mutants. Characterization of the locations and types of DNA sequence alteration occurring spontaneously within a single gene in a radl background would provide valuable infoi-mation about the specificity of the radl mutator effect. In turn, this could yield important clues about the mechanism(s) of enhanced spontaneous mutagenesis in excision repair-deficient strains. To examine mutational specificity, we previously developed a system for the DNA sequence analysis of forward mutations in the yeast tRNA suppressor gene SUP4-o (39). In this system, all types of base pair substitution as well as deletions, duplications, insertions, and more complex events can be identified (11, 20, 22). Here, we report our use of this system to characterize spontaneous SUP4-o mutations arising in a radl mutator strain. Comparison of the spectrum of mutations obtained in this strain with the corresponding spectrum for an isogenic wild-type strain revealed that the radl mutator phenotype was associated with enhanced frequencies of single-base-pair substitution, single-base-pair deletion, and insertion of the yeast retrotransposable element Ty. We consider the possible mechanisms that might give rise to the majority of single-base-pair events in the radl strain and suggest that spontaneous DNA damage, which is a substrate for excision repair, can activate Ty transposition.

The product of the RADI gene of the yeast Saccharomyces cerevisiae is believed to function at the incision step of excision repair of DNA damage (9). Thus, it is not surprising that defects in this gene increase sensitivity to a variety of DNA-damaging agents, including UV, the UV-mimetic chemical 4-nitroquinoline-1-oxide, mono- and bifunctional alkylating agents,. and photoactivated psoralens (6, 8, 9, 40, 41, 49). As well, the RADI gene product is required for the repair of NM-methyladenine (13). Not only do RADI deficiencies sensitize cells to the lethal effects of certain genotoxic agents, but also they enhance UV and 4-nitroquinoline1-oxide-induced mutagenesis (27, 40) and UV-induced mitotic interchromosomal recombination (45). Recent findings suggest that the RADI gene product also plays a role in mitotic intrachromosomal recombination and in integration of linear DNA molecules into homologous genomic sequences (2, 17, 44). In addition to these various properties, defects in RADl confer a mutator phenotype (35, 43, 47). Both enhanced locus reversion to prototrophy and forward mutation to suppression and canavanine resistance have been reported, but neither the precise mutational changes involved nor the specificity of the mutator effect has been elucidated. To account for the association between repair defects and enhanced spontaneous mutagenesis in yeast cells, von Borstel and colleagues (12, 42) have invoked the repair channeling hypothesis (5, 10). According to this hypothesis, impairment of a specific repair pathway results in channeling of damage normally repaired by that pathway along other, competitive pathways. In a similar fashion, shunting of spontaneous DNA lesions through mutagenic repair pathways might account for elevated spontaneous tnutation in repair-deficient strains. However, the nature of these errorprone pathways has remained obscure, and generally the magnitudes of the mutator effects in yeast cells are relatively small. Consequently, it has been suggested that repair defects might have a more indirect influence on spontaneous *

MATERIALS AND METHODS Strains, plasmid, and media. The haploid, repair-proficient yeast strain MKP-o (MATa canl-100 ade2-1 lys2-1 ura3-52 leu2-3,112, his3-A200, trpl-A901) has been described (39). KAM1, carrying a radl ::LEU2 insertion but otherwise isogenic to MKP-o, was derived by transforming MKP-o with a 6.96-kilobase (kb) Sall DNA fragment encompassing the RADI gene and flanking regions and having the internal

Corresponding author. 3009

J. BACTERIOL.

KUNZ ET AL.

3010

2.1-kb StuI-Clal fragment ofRADl (70%o of the RADI coding sequence) replaced with a 3.2-kb BglII fragment containing the LEU2 gene. The 6.96-kb Sall fragment was obtained from plasmid pL962 (provided by R. L. Keil). MKP-o was transformed by using lithium acetate (15), and colonies that emerged on leucine omission medium were tested for sensitivity to 254-nm UV. Replacement of the RAD1 gene with a single copy of the 6.96-kb Sail fragment in the highly UV sensitive transformant KAM1 was confirmed by DNA hybridization analysis (data not shown). Plasmid YCpMP2 was then transformed into MKP-o and KAM1 to make strains MKP-op and KAM1-p, respectively. Escherichia coli JF1754 (Alac gal metB leuB hisB436 hsdR) was used for plasmid recovery from yeast cells. Plasmid YCpMP2 (39) is a centromere-containing vector having components that enable it to replicate autonomously in yeast and bacterial cells. In yeast cells, such plasmids exhibit typical chromatin organization, replicate once per cell cycle in S phase, and are maintained predominantly as single copies in haploid cells (37). YCpMP2 carries SUP4-o, an ochre suppressor allele of a yeast tyrosine tRNA gene. It also bears the yeast URA3, gene which permits selection for YCpMP2 via complementation of the ura3-52 allele present in MKP-o and KAML. Media for growth of yeast and bacterial cells and for mutant selection and characterization have been described (21, 39). Detection of SUP4-o mutants. Selection of forward mutations in the SUP4-o gene is based on loss of suppressor activity. MKP-op carries ochre-suppressible markers that

canavanine resistance (canl-100), red pigmentation (ade2-1), or lysine auxotrophy (Iys2-1). Normally, these confer

markers are suppressed by SUP4-o so that cells carrying YCpMP2 are sensitive to canavanine and form white, lysineindependent colonies. Reduction of suppressor activity leads to the formation of red, canavanine-resistant colonies that require lysine. This procedure can detect at least a 30% decrease in the production of functional suppressor tRNA (48). Scoring for the loss of suppression of all three markers is unlikely to bias the recovery of mutations significantly. Mutations that prevent the suppression of just two ochre mutations are rare (1% of mutations analyzed) at the chromosomal SUP4-o locus (24), and we have not detected DNA sequence alterations in the plasmid copy of the SUP4-o gene from mutants isolated only on the basis of canavanine resistance and red coloring (23). Presumably, such mutants are due to antisuppressors (30). Furthermore, we have recovered mutations at all but 6 of the 75

by the selection protocol used here (20; B. A. Kunz, unpublished data). In particular, we have detected single41 G

C

-*

*C pairs (GG and G

C

*C

A

T

T*A

transversions, 38 sites each) and at 29 of the 34 A T pairs (A

T-

G

C transitions, 23 sites; A

*T -C

G and A

T

-*T A transversions, 16 and 20 sites, respectively) within the SUP4-o exons. We have also found the three possible substitutions at position 51 (G C pair) within the 14-base-

A TT T- A transversion has been detected at position 43 of the chromosomal SUP4-o gene (25). Consequently, although SUP4-o is just 89 base pairs long, at least 172 substitutions at a minimum of 71 sites lead to detectable phenotypic changes. These findings argue against selection bias. Plasmid stability and spontaneous mutation. Cultures of MKP-op or KAMl-p were grown from low-titer inocula (33 cells per ml) in uracil or uracil-leucine omission broth, respectively, at 30°C with shaking to final titers of approxipair

intron

of SUP4-o.

In

addition,

an

No.

examined'

Strain Strain

Mutant

MKP-op (RADI)

Total White Red Red, Lys-

29,325 (100) 26,569 (90.6) 2,756 (9.4) 836 (2.8)

KAMl-p (radl)

Total White Red Red, Lys-

37,316 (100) 35,761 (96) 1,555 (4.2) 1,358 (3.6)

typesa

(% of total)

Frequencyb

(10-6)

72 65 6.7 2.01

360 346 15 13.1

a The total includes canavanine-resistant mutants that are (i) white or (ii) red and Lys' or Lys-. I Colonies were isolated from, and frequencies are the means for, 40 (MKP-op) or 22 (KAMl-p) independent cultures.

mately 107 cells per ml (determined by Coulter Counter). Then the cell suspensions were diluted when necessary and plated on uracil (MKP-op) or uracil-leucine (KAM1-p) omission medium to measure viability, on minimal medium supplemented with uracil and other nutrients to determine plasmid retention, and on uracil or uracil-leucine omission medium containing canavanine (30 ,ug/ml) to select canavanine-resistant mutants. All plates were incubated for 6 days at 30°C. Plasmid stability was expressed as the ratio [(number of colonies on minimal omission medium)/(number of colonies on minimal medium containing uracil)] x 100. Red, canavanine-resistant colonies were subcultured to uracil omission medium, incubated for 2 days at 30°C, replicated to uracil-lysine omission medium, and incubated at 30°C for 2 to 3 days. Lysine auxotrophs were scored as SUP4-o mutants. The mutation frequencies were then determined for each independent culture, and the rate of mutation per generation was calculated as described previously (22). DNA isolation, bacterial transformation, and DNA sequencing. YCpMP2 DNA for bacterial transformation was released from yeast cells by cellular disruption with glass beads (14). Plasmid DNA was isolated from E. coli by alkaline extraction (33). Bacterial cells were transformed by a modification of the CaCl2 method (39). Mutant SUP4-o genes were sequenced by dideoxynucleotide-mediated chain termination on linearized double-stranded YCpMP2 molecules (23).

sites in the

exon

gene and have identified a wide range of mutational classes

base-pair changes at 40 of the transitions, 34 sites; G *C

TABLE 1. Characterization of canavanine-resistant colonies

--

RESULTS Plasmid retention and isolation of canavanine-resistant mutants. Cultures of MKP-op (RADI) and KAM1-p (radl), grown from low-titer inocula, were plated to measure plasmid stability and to select canavanine-resistant mutants. To determine plasmid stability, the number of colonies that formed on medium selective for the plasmid was compared with the corresponding value for nonselective medium. Approximately 88% of the MKP-op (37,299/42,231 = 0.88) and KAM1-p (9,968/11,513 = 0.87) cells retained YCpMP2. Since this value is typical for maintenance of YCpMP2 in MKP-op (11, 22, 39), the results indicate that the excision repair deficiency did not influence plasmid maintenance. The frequencies of all three detectable types of canavanineresistant mutant were increased in the radl strain (Table 1). Presumably, the white or red and Lys' mutants were due to alterations at the CAN] locus or antisuppressor loci, respectively. Slightly more of the total canavanine-resistant mutants derived from KAMl-p than MKP-op were classified as

SPECIFICITY OF YEAST radl MUTATOR

VOL . 172, 1990 TABLE 2. Sequence alterations in SUP4-o mutants

KAM1-p (radl)

MKP-op (RADI) DNA sequence

alteration

No

detected (% of total)

Substitution 262 (81.4) Single change Tandem double 2 (0.6) Nontandem double Deletion 21 (6.5) 1 bp 7 (2.2) >1 bp Insertion 1 (0.3) 1 bp >1 bp 25 (7.8) Ty 1 (0.3) Duplication 3 (0.9) Complex event Total 322

Fr(10ency

No.

detected

(% of total)

(10)7n

0.1

211 (84.7) 1 (0.4) 1 (0.4)

1.4 0.4

16 (6.4) 2 (0.8)

8.5 1.1

1 (0.4) 17 (6.9)

0.5 8.9

16.4

111.0 0.5 0.5

0.06 1.6 0.06 0.2 20.12

249

131.0

having SUP4-o mutations (red and Lys-). The mean frequency of spontaneous mutation at the SUP4-o locus was 6.5-fold greater for KAMl-p than MKP-op, a significant increase (P < 0.001 for a chi-square test comparing the numbers of SUP4-o mutants and nonmutants among the total numbers of viable plasmid-bearing cells plated to select canavanine-resistant mutants for each strain). The corresponding mutation rates were calculated to be 9.1 x 10-7 and 1.3 x 10-7 events per generation. It is clear that the differences observed between the repair-proficient and -deficient strain reflect the radi mutator effect rather than variation in their genetic backgrounds, since the strains were isogenic except for the RADI locus. Classification of SUP4-o mutants. To ensure the independence of the spontaneous mutations characterized by DNA sequencing, each mutant was obtained from a separate culture. Collections of 249 mutants arising in the radl strain and 322 mutants selected in the RADl wild-type background were analyzed (the results for 100 mutants isolated here in MKP-op were pooled with the data for 222 mutants examined previously [22]). Between the two strains, 10 different mutational classes were identified (Table 2). With the exception of multiple-base-pair deletion, the relative proportions of those classes of mutation common to both strains (single and nontandem double-base-pair substitutions, single-basepair deletion, and insertion of the 6-kb yeast retrotransposon Ty) were similar. The increases in the frequencies of the single-base-pair events and Ty insertions in KAM1-p were all approximately sixfold and together accounted for >98% of the overall radl mutator effect. Analysis of base pair substitutions. The increment in the frequency of single-base-pair substitution was responsible for 85% of the mutation frequency increase in the radl strain (Table 2). Both types of transition and all four possible transversions were recovered in MKP-op and KAM1-p, although there was a small excess of transversions for each strain (Table 3). The relative fractions of the different base pair changes in the two strains were quite similar. The distributions, within SUP4-o, of the base pair substitutions arising in MKP-op and KAM1-p, including double events, are given in Fig. 1. A total of 67 sites were mutated, and some similarities were noted between the two distributions. For both strains, base pair substitutions occurred at only one position (site 51) within the tRNA intron, whereas the remaining changes were distributed throughout the

3011

TABLE 3. Relative fractions and frequencies of base pair substitutions Substitution

Transitions G C-* A T A T- G *C Total Transversions G C -T A G C- C G A T- C *G A *T. T A Total

MKP-op (RADI) KAMl-p (radi) No. detected Frequency No. detected Frequency (% of total) (10-7) (% of total) (10-7)

69 (25.9) 41 (15.4) 110 (41.3)

4.3 2.5 6.8

53 (24.7) 14 (6.5) 67 (31.2)

27.7 7.3 35.0

88 (33.1) 52 (19.6) 5 (1.9) 11 (4.1) 156 (58.7)

5.5 3.2 0.3 0.7 9.7

74 (34.4) 56 (26.0) 7 (3.3) 11 (5.1) 148 (68.8)

38.5 29.1 3.7

5.7 77.0

SUP4-o gene with 47 sites in common. In addition, two of the most frequently mutated sites (18 and 88) coincided, and of the specific substitutions at the common sites, 75% of those detected in MKP-op were also found in KAM1-p. On the other hand, a number of differences were also observed. Whereas no changes were recovered 5' or 3' to the SUP4-o coding region in MKP-op, an A. T -- T. A transversion occurring as part of a nontandem double event was detected at position -2 in KAM1-p. This is the first substitution that we have identified outside of the region encoding the tRNA. However, since this change has not yet been detected as an individual event among more than 3,400 SUP4-o mutations characterized to date, it might not reduce SUP4-o expression or tRNA processing or function sufficiently to permit its recovery as a single substitution in our system. Base pair changes were detected at fewer sites for KAMl-p than for MKP-op (50 versus 64), and the substitution frequencies in the radl strain were from 3- to 52-fold greater at 42 of the 47 common sites. Finally, 25% of the specific substitutions detected in the radl background at sites common to both strains were missing from the corresponding sites in the RADI distribution. By using the Monte Carlo estimate of the P value of the hypergeometric test (1) to compare the two distributions statistically, we found that the probability of random sampling error being the cause of differences in the two distributions was less than 1% (with 1,500 simulated comparisons, the P value was estimated to be 0.002 and the 90% confidence interval ranged from 0.0005 to 0.0052). Single-base-pair deletions and Ty insertions. Deletions of single base pairs and insertions of the 6-kb retrotransposon Ty into SUP4-o each accounted for 6.5% of the mutation frequency increase attributable to the radl mutator effect (Table 2). In both strains, the majority (>93%) of deletions occurred in runs of two or more G C pairs and the number of deletions increased with the number of base pairs in the run (Table 4), suggesting that their formation may have involved strand slippage events (46). The run of five G. C pairs located at positions 79 through 83 constituted a deletion hot spot in each strain. Ty insertions were detected at four different locations in SUP4-o (Table 4). Two or more Ty elements integrated at two sites in each strain, and more than 80% of the total insertions were immediately 5' to position 38, the SUP4-o site that we previously reported to be a frequent target for transposition (11). Recently, it has been determined that transposition of Ty elements into the URA3 and LYS2 loci is also not random (36). Other mutational classes. A multiple-base-pair deletion, a single-base-pair insertion, and a duplication were recovered in MKP-op in this study. The deletion involved the loss of 27

J. BACTERIOL.

KUNZ ET AL.

3012

GTT

AAA AAA T T

G G

T T

A

T C A A A T

A A TAT CAA

c c c

c2 AA AA AA

AAAG T AGA AAAGCT

C

AA A A A A A A A A AT AA

A A A A A A A A A A A A A

c

C C C AGTAC T T TT

CAGCAC CCAGAACG

A T A T A T A C G AG GC GGCAG GCC AACAGTGCA

T T G T T C A T C A A C A A C A A CTACCAC

T T C T C G T C G GTT C GTACT TC CAGACA

G G G G AC AC

c c c

A CT CAGCC

TT T TT TT TT TT CG CG CG CA T

T T T T T T TC TC TC CA T TA AAGG TA AAGG AA

TTMGGAAA

I 1

-2 3' AT

T

A A A T T T T T

AAGC A AGGG T TG G T TG T

40

30

20

10

1

GAGAGCCATC

AATTCCGCGT

TCTCAAATTA

MA

A AAAAC

CA AC G C

AA CC G AA G AA AA AG AT

ACMC

GGTTCAACCA AC A AC

A

TTA C TTA C 2C

C

c c c

A

T

T2 T

A A A A A

G A G CA G TA C TA T T T

T

T T

T TT

50 AATAGTGATG

60 CTTTAGMCT A CTATA A TTT G A TT G T G A T G T G T T

80 89 70 CTAGCCCCCA AGCTGAGCGG GGGCCCTCT 5'

C AAAGCGC C AA GC CA T CA T TA TG G G G G G G G T

CGA CA T TGA TA T T TGC T TGC G

C CA A M C TT G AC C T G AC T G T T T T T T T T T T

FIG. 1. Distribution of spontaneous base pair substitutions in the SUP4-o gene. For simplicity, only the region of the transcribed strand encoding the tRNA is shown, along with the first two upstream bases. The anticodon is at 36 to 38 and the 14-base-pair intron extends from 40 through 53 (18). Mutations isolated in MKP-op (RADI) and KAMl-p (radl) are presented above and below the transcribed strand, respectively. Numbered substitutions indicate changes in double mutants.

base pairs from positions 63 through 89, the insertion occurred in the run of 5 G- C pairs at positions 79 through 83, and the duplication was a direct repeat of the 7-base-pair sequence encompassing positions 72 through 78. The other multiple-base-pair deletions and the complex events identified in MKP-op have been detailed elsewhere (11). Two triple-base-pair deletions (77 through 79) and a doubleTABLE 4. Characterization of single-base-pair deletions and Ty insertions KAM1-p (radl) MKP-op (RADI) Event

Sitea

Deletion

15 18 or 19 25 or 26 29 65 - 67 72 79 -83 84 -86

Total Insertion

Total

6 +7 17 + 18 37 + 38 43 4* 44

No. No. detected Frequency detected Frequency (1) (% of total) (% of total)

1 (4.8) 1 (4.8)

0.67 0.67

1 (4.8) 16 (76.2) 2 (9.4) 21

0.67 10.7 1.3 14.01

2 (8.0) 22 (88.0) 1 (4.0) 25

1.3 14.1 0.64 16.04

1 (6.2) 1 (6.2) 4 (25.0)

5.3 5.3 21.2

8 (50.0) 2 (12.6) 16

42.5 10.7 85.0

3 (17.6)

15.7

14 (82.4)

73.3

17

89.0

For deletions in runs of base pairs, the position of the run is given since the precise base pair eliminated cannot be determined. Ty insertions were identified as described previously (11); the symbol * indicates that Ty is presumed to have inserted between the two sites. a

base-pair insertion (5'-GC-3' found between positions 64 and 65) were detected in the radl strain. DISCUSSION Disruption of the RADI gene clearly resulted in a mutator phenotype. However, there has been controversy regarding the ability of radl defects to enhance spontaneous mutagenesis (43). Previously, detection of the mutator effect apparently depended on the assay system and the particular radl allele used, suggesting that the mutator phenotype might be obscured by its own specificity or by leakiness of individual radl alleles. Since the frequencies of all types of base pair substitution were increased in the radl strain, it is doubtful that the radl mutator effect could be concealed by its specificity alone. However, base pair changes were not detected at all sites in SUP4-o where substitutions occurred in the isogenic RADI strain. Thus, neighboring DNA sequences might modulate the radl mutator effect at specific base pairs. If so, this could explain why radl strains were not always found to be mutators when mutation was assayed only at specific sites. Leakiness of the radl alleles used might also have been a complicating factor. Increases in the frequencies of single-base-pair substitution, single-base-pair deletion, and Ty insertion accounted for almost the entire change in the SUP4-o mutation frequency in the radl strain. The relative fractions of these events were essentially the same in both strains, and their frequencies were increased to similar extents in KAM1-p, indicating that the three classes of mutation were coordinately enhanced by the radl mutator effect. Since RADI defects are believed to cause excision repair deficiency, it is reasonable to suppose that unexcised spontaneous damage, which normally would be removed by excision repair, could

VOL. 172, 1990)

play a key part in this enhancement. Given that singlebase-pair events and transposition of Ty elements were specifically promoted in the radl strain, it seems probable that if unrepaired spontaneous DNA lesion are involved, then at least two different mechanisms would have to be responsible for the radl mutator phenotype. Like radl defects, rad3 mutations that increase UV sensitivity appear to eliminate the incision step of excision repair (9). It has been shown that the mutator phenotype associated with the rad3-12 allele, which markedly enhances UV sensitivity (42) and so is unlikely to be a remi allele of RAD3 (9), can be offset by the antimutator effect conferred by defective alleles of the REV3 gene (42). In fact, rad3-12 rev3 double mutants were found to have reduced spontaneous mutation rates similar to those of rev3 antimutator strains alone (42). Recently, DNA sequencing of the REV3 gene has led to the prediction that it encodes a novel, nonessential DNA polymerase (34). Since REV3 is required for mutagenesis by UV, 4-nitroquinoline-1-oxide and gamma rays but is distinct from the genes encoding the three known yeast nuclear DNA polymerases, it was concluded that the REV3 polymerase likely functions only in translesion synthesis (34). If so, then translesion synthesis should be capable of producing single-base-pair substitutions and deletions because UV induction of these events is substantially decreased by defects in REV3 (28, 29). Here we have determined that the radl mutator enhanced the frequencies of spontaneous single-base-pair substitutions and deletions. Taken collectively, these findings suggest that the radl mutator phenotype in yeast cells could be explained, in large part, by error-prone translesion synthesis across spontaneous damage, which is a substrate for excision repair. The notion of translesion synthesis past unrepaired spontaneous damage is reminiscent of the hypothesis that spontaneous lesions are channeled through mutagenic repair pathways in strains deficient in error-free repair (12, 42). The difference is that in the case of translesion synthesis, elevated spontaneous mutagenesis would not reflect mutagenic repair of DNA damage but rather tolerance of spontaneous DNA lesions via an error-prone mechanism. Insertion of the yeast retrotransposon Ty into SUP4-o was also increased in the radl background. Elsewhere, we have demonstrated that Ty transposition is severely reduced in a radS2 strain (22) but is not influenced by disruption of RAD18 (B. A. Kunz et al., submitted for publication). It has been found that treatment with UV or 4-nitroquinoline1-oxide, mutagens that produce DNA damage which is subject to excision repair, stimulates transcription of Ty elements and activates their transposition (4, 31). Although this activation is mutagen dose dependent and so presumably correlates with the production of DNA damage, the precise mechanism(s) at work has not yet been elucidated. Nevertheless, these results plus our findings argue that activation of Ty transposition by spontaneous DNA damage, which would usually be repaired in a RADI strain, contributes to the radl mutator phenotype. After this report was submitted for publication, it was reported that mutations in RAD6, a yeast gene whose product polyubiquitinates histones in vitro, increase Ty transposition into the CAN] and URA3 genes (38). It should also be considered that the radl mutator effect could be unrelated to spontaneous damage (26, 43). One possible explanation would be that defects in the RADI gene might somehow reduce the accuracy of DNA synthesis. Since Ty retrotransposition is increased in the radl strain, it seems improbable that altered replicational fidelity alone

SPECIFICITY OF YEAST radl MUTATOR

3013

could be responsible for the mutator phenotype. Still, it could be that a decrease in the fidelity of replicative DNA synthesis is responsible for the enhanced frequencies of single-base-pair substitution and deletion in the radl strain. However, this would be difficult to reconcile with the ability of rev3 alleles to eliminate the rad3-12 mutator phenotype, since the predicted REV3 polymerase is not required for replicative DNA synthesis (34). Alternatively, it has been suggested that excision repair functions might recognize a subset of DNA replication errors, presumably mismatches (32). Accordingly, the increase in single-base-pair deletions and substitutions in the radl strain might reflect a deficiency in mismatch correction. In yeast cells, a frameshift mismatch (single-nucleotide loop opposite a missing nucleotide) can be repaired to restore the missing information or fix the deletion (3, 19). Furthermore, in one study (19), reductions in mismatch repair were found to enhance the incidence of deletion. Thus, a defect in repair of frameshift mismatches could account for the increased deletion frequency in the radl background. On the other hand, it seems less likely that disruption of the RADI gene enhanced substitution mutagenesis by perturbing mismatch correction. There was no significant increase in the fraction of a particular type(s) of base-pair change in the radl strain, arguing that the specificity of mismatch correction was not altered. If instead the RADI disruption reduced the overall efficiency of mismatch correction, an increase in transitions would be expected in the radl background, since yeast mismatch repair corrects transition mismatches more efficiently than most transversion mismatches (3, 19), yet the ratio of transversions to transitions was slightly greater in KAM1-p than in MKP-op. In conclusion, the radl mutator effect in yeast cells specifically increases the frequencies of single-base-pair substitution, single-base-pair deletion, and Ty insertion. We cannot unequivocally discount the possibilities that disruption of RADI enhanced single-base-pair events by decreasing the fidelity of replicative DNA synthesis or the efficiency of mismatch correction. However, the simplest explanation would be that the mutator phenotype involves error-prone translesion synthesis past unrepaired spontaneous damage and activation of Ty transposition by such damage. ACKNOWLEDGMENTS We thank W. T. Adams for running the computer analysis of the base pair distributions, R. L. Klein for supplying plasmid pL%2, and M. Gabriel, T. Kolodka, and F. Yadao for technical assistance. T.K. was supported by a Natural Sciences and Engineering Research Council of Canada undergraduate scholarship, and M.G. and F.Y. were supported in part by a subsidy from the Manitoba Careerstart '89 Program. This work was supported by the Natural Sciences and Engineering Research Council of Canada. LITERATURE CITED 1. Adams, W. T., and T. R. Skopek. 1987. Statistical test for the comparison of samples from mutational spectra. J. Mol. Biol.

194:391-396. 2. Aguilera, A., and H. L. Klein. 1989. Yeast intrachromosomal recombination: long gene conversion tracts are preferentially associated with reciprocal exchange and require the RADI and RAD3 gene products. Genetics 123:683-694. 3. Bishop, D. K., J. Andersen, and R. D. Kolodner. 1989. Specificity of mismatch repair following transformation of Saccharomyces cerevisiae with heteroduplex plasmid DNA. Proc. Natl. Acad. Sci. USA 86:3713-3717. 4. Bradshaw, V. A., and K. McEntee. 1989. DNA damage activates transcription and transposition of yeast Ty retrotransposons.

3014

KUNZ ET AL.

Mol. Gen. Genet. 218:465-474. 5. Brendel, M., N. A. Khan, and R. H. Haynes. 1970. Common steps in the repair of alkylation and radiation damage in yeast. Mol. Gen. Genet. 106:289-295. 6. Chanet, R., C. Izard, and E. Moustacchi. 1976. Genetic effects of formaldehyde in yeast. III. Influence of ploidy and mutations affecting radiosensitivity on its lethal effects. Mutat. Res. 35:29-38. 7. Chanet, R., N. Magana-Schwencke, and F. Fabre. 1988. Potential DNA-binding domains in the RAD18 gene product of Saccharomyces cerevisiae. Gene 74:543-547. 8. Cooper, A. J., and R. Waters. 1987. A complex pattern of sensitivity to simple monofunctional alkylating agents amongst the rad mutants. Mol. Gen. Genet. 209:142-148. 9. Friedberg, E. C. 1988. Deoxyribonucleic acid repair in Saccharomyces cerevisiae. Microbiol. Rev. 52:70-102. 10. Game, J. C., and B. S. Cox. 1973. Synergistic interactions between rad mutations in yeast. Mutat. Res. 20:35-44. 11. Giroux, C. N., J. R. A. Mis, M. K. Pierce, S. E. Kohalmi, and B. A. Kunz. 1988. DNA sequence analysis of spontaneous mutations in the SUP4-o gene of Saccharomyces cerevisiae. Mol. Cell. Biol. 8:978-981. 12. Hastings, P. J., S.-K. Quah, and R. C. von Borstel. 1976. Spontaneous mutation by mutagenic repair of spontaneous lesions in DNA. Nature (London) 264:719-722. 13. Hoekstra, M. F., and R. E. Malone. 1986. Excision repair functions in Saccharomyces cerevisiae recognize and repair methylation of adenine by the Escherichia coli dam gene. Mol. Cell. Biol. 6:3555-3558. 14. Hoffman, C. S., and F. Winston. 1987. A ten-minute DNA preparation from yeast efficiently releases autonomous plasmids for transformation of Escherichia coli. Gene 57:267-272. 15. Ito, Y., Y. Fukada, K. Murata, and A. Kimura. 1983. Transformation of intact yeast cells treated with alkali ions. J. Bacteriol. 153:163-168. 16. Jones, J. S., S. Weber, and L. Prakash. 1988. The Saccharomyces cerevisiae RAD18 gene encodes a protein that contains potential zinc finger domains for nucleic acid binding and a putative nucleotide binding sequence. Nucleic Acids Res. 16:7119-7131. 17. Klein, H. 1988. Different types of recombination events are controlled by the RADI and RADS2 genes of Saccharomyces cerevisiae. Genetics 120:367-377. 18. Knapp, G., J. S. Beckwith, P. F. Johnson, S. A. Fuhrman, and J. Abelson. 1978. Transcription and processing of intervening sequences in yeast tRNA genes. Cell 14:221-236. 19. Kramer, B., W. Kramer, M. S. Williamson, and S. Fogel. 1989. Heteroduplex DNA correction in Saccharomyces cerevisiae is mismatch specific and requires functional PMS genes. Mol. Cell. Biol. 9:4432-4440. 20. Kunz, B. A., J. D. Armstrong, M. Glattke, S. E. Kohahmi, and J. R. A. Mis. 1990. The SUP4-o system for analysis of mutational specificity in yeast, p. 337-346. In M. L. Mendelsohn and R. J. Albertini (ed.), Fifth international conference on environmental mutagens. Alan R. Liss, Inc., New York. 21. Kunz, B. A., F. Eckardt, and R. H. Haynes. 1985. Analysis of non-linearities in frequency curves for UV-induced mitotic recombination in wild-type and excision-repair-deficient strains of yeast. Mutat. Res. 151:235-242. 22. Kunz, B. A., M. G. Peters, S. E. Kohlami, J. D. Armstrong, M. Glattke, and K. Badiani. 1989. Disruption of the RAD52 gene alters the spectrum of spontaneous SUP4-o mutations in Saccharomyces cerevisiae. Genetics 122:535-542. 23. Kunz, B. A., M. K. Pierce, J. R. A. Mis, and C. N. Giroux. 1987. DNA sequence analysis of the mutational specificity of u.v. light in the SUP4-o gene of yeast. Mutagenesis 2:445-453. 24. Kurjan, J., and B. D. Hail. 1982. Mutations at the Saccharomyces cerevisiae SUP4 tRNATYr locus: isolation, genetic fine structure mapping, and correlation with physical structure. Mol. Cell. Biol. 2:1501-1513. 25. Kurjan, J., B. D. Hall, S. Gilliam, and M. Smith. 1980. Mutations at the yeast SUP4 tRNATYr locus: DNA sequence changes in mutants lacking suppressor activity. Cell 20:701-709. 26. Lawrence, C. W. 1982. Mutagenesis in Saccharomyces cerevisiae. Adv. Genet. 21:173-254.

J. BACTERIOL.

27. Lawrence, C. W., and R. B. Christensen. 1976. UV mutagenesis in radiation sensitive strains of yeast. Genetics 82:207-232. 28. Lawrence, C. W., and R. B. Christensen. 1979. Ultravioletinduced reversion of cycl alleles in radiation-sensitive strains of yeast. Genetics 92:397-408. 29. Lawrence, C. W., T. O'Brien, and J. Bond. 1984. Ultravioletinduced reversion of his4 frameshift mutations in rad6, revi, and rev3 mutants of yeast. Mol. Gen. Genet. 195:487-490. 30. McCready, S. J., and B. S. Cox. 1973. Antisuppressors in yeast. Mol. Gen. Genet. 124:305-320. 31. McEntee, K., and V. A. Bradshaw. 1988. Effects of DNA damage on transcription and transposition of Ty retrotransposons of yeast, p. 245-253. In M. E. Lambert, J. F. McDonald, and I. B. Weinstein (ed.), Eukaryotic transposable elements as mutagenic agents. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 32. Monteleone, B. A., M. F. Hoekstra, and R. E. Malone. 1988. Spontaneous mitotic recombination in yeast: the hyper-recombinational remi mutations are alleles of the RAD3 gene. Genetics 119:289-301. 33. Moreile, G. 1989. A plasmid extraction procedure on a miniprep scale. Focus 11:7-8. 34. Morrison, A., R. B. Christensen, J. Alley, A. K. Beck, E. G. Bernstein, J. F. Lemontt, and C. W. Lawrence. 1989. REV3, a Saccharomyces cerevisiae gene whose function is required for induced mutagenesis, is predicted to encode a nonessential DNA polymerase. J. Bacteriol. 171:5659-5667. 35. Moustacchi, E. 1969. Cytoplasmic and nuclear genetic events induced by UV light in strains of Saccharomyces cerevisiae with different UV sensitivities. Mutat. Res. 7:171-185. 36. Natsoulis, G., W. Thomas, M.-C. Roghmann, F. Winston, and J. D. Boeke. 1989. Tyl transposition in Saccharomyces cerevisiae is nonrandom. Genetics 123:269-279. 37. Newlon, C. S. 1988. Yeast chromosome replication and segregation. Microbiol. Rev. 52:568-601. 38. Picologlou, S., N. Brown, and S. W. Liebman. 1990. Mutations in RAD6, a yeast gene encoding a ubiquitin-conjugating enzyme, stimulate retrotransposition. Mol. Cell. Biol. 10:1017-1022. 39. Pierce, M. K., C. N. Giroux, and B. A. Kunz. 1987. Development of a yeast system to assay mutational specificity. Mutat. Res. 182:65-74. 40. Prakash, L. 1976. Effects of genes controlling radiation sensitivity on chemically induced mutations in Saccharomyces cerevisiae. Genetics 83:285-301. 41. Prakash, L., and. S. Prakash. 1977. Isolation and characterization of mms-sensitive mutants of Saccharomyces cerevisiae. Genetics 86:33-55. 42. Quah, S.-K., R. C. von Borstel, and P. J. Hastings. 1980. The origin of spontaneous mutation in Saccharomyces cerevisiae. Genetics 96:819-839. 43. Sargentini, N. J., and K. C. Smith. 1985. Spontaneous mutagenesis: the roles of DNA repair, replication, and recombination. Mutat. Res. 154:1-27. 44. Schiestl, R. H., and S. Prakash. 1988. RADI, an excision repair gene of Saccharomyces cerevisiae, is also involved in recombination. Mol. Cell. Biol. 8:3619-3626. 45. Snow, R. 1968. Recombination in ultraviolet sensitive strains of Saccharomyces cerevisiae. Mutat. Res. 6:409-418. 46. Streisinger, G., Y. Okada, J. Emrich, J. Newton, A. Tsugita, E. Terzaghi, and M. Inouye. 1966. Frameshift mutations and the genetic code. Cold Spring Harbor Symp. Quant. Biol. 31:157-179. 47. von Borstel, R. C., and P. J. Hastings. 1980. DNA repair and mutagen interactions in Saccharomyces: theoretical considerations, p. 159-167. In W. M. Generoso, M. D. Shelby, and F. J. de Serres (ed.), DNA repair and mutagenesis in eukaryotes. Plenum Publishing Corp., New York. 48. Wang, S. S., and A. K. Hopper. 1988. Isolation of a yeast gene involved in species-specific tRNA processing. Mol. Cell. Biol. 8:5140-5149. 49. Zuk, J., D. Zabrowska, and Z. Swietlinska. 1979. Comparison of sensitivity and liquid holding recovery in rad mutants of Saccharomyces cerevisiae inactivated by UV and DEB. Mol. Gen. Genet. 166:91-96.