Proc. Natl. Acad. Sci. USA Vol. 84, pp. 4195-4199, June 1987 DNA polymerase III of Escherichia coli is required for UV and ethyl methanesulfonate mutagenesis (DNA replication/SOS repair) MICHAEL E. HAGENSEE, TERRY L. TIMME, SHARON K. BRYAN, AND ROBB E. MOSES Department of Cell Biology, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030 Communicated by Daniel Nathans, March 9, 1987 (receivedfor review January 7, 1987)

ABSTRACT Strains of Escherichia coli possessing the polC+ by transduction. Strain ERli is an E486 (polC486) (8) pebAl , a functional DNA polymerase I, and a derivative made by P1 transduction ofa TnJO linked topcbAl temperature-sensitive mutation in DNA polymerase m can from RM552. Strain RM552 is an ESli (8) derivative con- survive at the restrictive temperature (430C) for DNA poly- taining TnJO linked to pcbAl (zic-J: :TnJO) transduced from a merase m. The mutation rate of the bacterial genome of such CSM61 derivative (7). Strain SB229 was constructed by strains after exposure to either UV light or ethyl methanesul- transduction of recA56 srlJ300::TnJO into JM103. fonate was measured by its rifampicin resistance or amino acid Plasmid pDS4-26 was provided by C. McHenry. It contains requirements. In addition, Weigle mutagenesis of preirradi- the coding region for the a subunit of DNA polymerase III ated X phage was also measured. In all cases, no increase in (9). Plasmid pSB5 is a clone of the a subunit of DNA mutagenesis was noted at the restrictive temperature for DNA polymerase III derived in our laboratory. polymerase HI. Introduction of a cloned DNA polymerase HI Materials. Media were purchased from Difco. EtMes was gene returned the mutation rate of the bacterial genome as well obtained from Kodak, and rifampicin was purchased from as the Weigle mutagenesis to normal at 43C. Using a recA-LacZ Sigma. fusion, the SOS response after UV irradiation was measured Procedures. Phage X was grown and X lysogens were and found to be normal at the restrictive and permissive constructed using standard procedures. Transformation was temperature for DNA polymerase III, as was induction of X performed as described by Hanahan (10). prophage. Recombination was also normal at either tempera- To assess the mutation rate, cells were collected in ture. Our studies demonstrate that a functional DNA polymer- midlogarithmic phase and resuspended to half their original ase III is strictly required for mutagenesis at a step other than volume in 0.05 M K2HPO4 (pH 7.4). Cells were either SOS induction. UV-irradiated using a germicidal lamp (General Electric) with a UV flux of 1 J.m-2.sec-l at room temperature or The functional DNA polymerase III holoenzyme complex of exposed to EtMes at a concentration of 0.2 M at 32 or 43°C. Escherichia coli is composed ofat least seven proteins (1, 2). Samples were removed at time intervals to measure cell These components interact to replicate the E. coli genome at survival, concentrated, and plated on indicator plates at the high fidelity. The a subunit of this complex possesses the desired temperature (32 or 43°C). The indicator plates con- polymerizing activity and is encoded by the dnaE (polC) gene sisted of either L-agar with rifampicin at a concentration of (3). Temperature-sensitive at this locus prevent 100 ,ug/ml or M9 plates lacking leucine. Mutation rate was replication at the restrictive temperature of 43°C (4). Muta- calculated as colonies on indicator plates divided by survi- tions at the dnaE locus may produce strains with an increase vors on L-agar. D37 values (the doses necessary to reduce the in the spontaneous mutation rate (5) as well as UV-induced surviving fractions to e-1 or 0.37) were determined graphi- mutagenesis (6). cally. The pcbAl mutation allows DNA polymerase I-dependent The Weigle mutagenesis experiments were performed replication ofthe bacterial genome without a functional DNA essentially as described by Defais et al. (11) using wild-type polymerase III a subunit (7). Strains containing the pcbAl X. Reactivation was determined from the total number of mutation, a functional DNA polymerase I, and a tempera- surviving plaques, whereas mutagenesis was determined ture-sensitive DNA polymerase III are viable at the restric- from the number of clear plaques. tive temperature (43°C) for DNA polymerase III and, there- The SOS response was measured by assaying ,B-galacto- fore, can be used to assess the role of DNA polymerase III sidase activity from a X lysogen carrying a recA-lacZ fusion in mutagenesis. Using such strains, we find that DNA (XGE190) provided by G. Weinstock (12). Units of *- polymerase III is required for mutagenesis of the bacterial galactosidase activity were as given by Miller (13). genome by either UV irradiation or ethyl methanesulfonate (EtMes) exposure as well as for Weigle mutagenesis. DNA RESULTS polymerase III is not required for induction of the SOS response or DNA recombination. Mutagenesis Requires DNA Polymerase III a-Subunit Ac- tivity. The strain CSM61 contains the pcbAl mutation, a functional DNA polymerase I, and a temperature-sensitive MATERIALS AND METHODS mutation in the a subunit of DNA polymerase III. DNA Strains. The bacterial strains used are listed in Table 1. replication in this strain at 43°C (7) requires DNA polymerase CSM61 and CSM14 are spontaneous, temperature-resistant I. CSM61 was exposed to UV irradiation, and rifampicin- revertants from HS432 (polAl, polB100, polC1O26, pcbAl) as resistant mutants arising were measured at 32 and 43°C. At previously described (7). CSM98 is a CSM61 derivative made the permissive temperature for DNA polymerase III (32°C), the number of mutations increased with increasing exposure to At the for DNA The publication costs of this article were defrayed in part by page charge UV light (Fig. 1B). restrictive temperature payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. Abbreviation: EtMes, ethyl methanesulfonate.

4195 Downloaded by guest on September 29, 2021 4196 Genetics: Hagensee et al. Proc. Natl. Acad. Sci. USA 84 (1987) Table 1. Bacterial strains Relevant genotype or Strain phenotype Source 55. CSM14 polA+, polB100, polClO26, 50 pcbAl, Leu-, His-, tr This laboratory CSM61 polAl, polB100, polCJO26, pcbAl, sup(PolI), 45 Leu-, His-, tr This laboratory CSM98 polA+, polB100, polC+, 40 pcbAl, Leu-, His-, tr This laboratory 0) ER11 E486 (polC486), pcbAl ° 35 zic-l::TnlO, tr This laboratory

SB229 JM103 recA56, 30 - srll300::TnlO This laboratory O 320C

W3110 Wild type J. Cairns 0 25 - tr, Temperature-resistant. , 20 - polymerase III (43°C), the background spontaneous mutation rate is the same as at 32°C, but there was no increase in the 15- number of mutations with increasing exposure to UV light. CSM61 was also exposed to a second , EtMes, and 10 the results were similar (Fig. 1A). CSM61 is phenotypically PolI+ because of a suppressor polAl amber and in 5 430C mutation (7) acting on the mutation, 0 extracts this strain has 10-15% ofthe wild-type level of DNA I I I I I polymerase I activity. This level is sufficient to allow normal 0 20 40 60 80 100 120 140 DNA repair at 32°C and to support replication at 43°C. seconds of UV at 50cm However, one might question if the failure of mutagenesis at a level DNA I FIG. 2. UV mutagenesis in ERll. Mutants were scored as 43°C is due to low of polymerase activity. rifampicin-resistant colonies. D37 values were 35 sec at 32 and 43°C. Strain CSM14 is an isogenic strain with an intragenic muta- UV flux was 1 Jm-2-sec-1. tion producing a polA+ allele giving DNA polymerase I with normal properties after a 1000-fold purification (S.K.B. and R.E.M., unpublished results). Strain CSM14 shows the same These results were also confirmed by using a second response to UV irradiation (Fig. 1C) and EtMes exposure marker, leucine prototrophy, to score for mutations. Strains (not shown) as CSM61. We conclude that the lack of CSM61 and ERll showed less than 1 Leu+ revertant per 108 mutagenesis at 43°C is not due to low levels of DNA survivors at 430C after EtMes exposure for 30 min but polymerase I. A nonisogenic polCts strain containing pcbAl, showed greater than 20 Leu+ revertants per 108 survivors at ERll (polC486), also showed no increase in the UV-induced 320C. mutagenesis at the restrictive temperature (Fig. 2). The same To rule out a simple temperature effect on mutagenesis, we result occurred with EtMes. This indicates that the deficit at constructed a polC+ strain, CSM98, which is isogenic to 43°C is not polC allele-specific. CSM61 and CSM14. As seen in Fig. 3 A and B, this strain

A B C 22 32°C

18 U) 0 :314 32°C

Cu

0 10 20 30 0 40 80 120 0 20 60 100 140 minutes with .2M EMS seconds of UV at 50cm seconds of UV at 50cm FIG. 1. Mutagenesis in CSM61 after EtMes (EMS) (A) or UV (B) exposure and also in CSM14 after UV exposure (C). Mutants were scored as rifampicin-resistant colonies. D37 values at 32 and 43°C were 16 min in A and 32 sec in both B and C. UV flux was 1 J'm-2-sec-1. Cells from either strain held at 32 or 43°C in phosphate buffer did not show an increase in rifampicin-resistant mutants after 30 min. Survival rate was greater than 90%o in buffer at both temperatures. Downloaded by guest on September 29, 2021 Genetics: Hagensee et al. Proc. Natl. Acad. Sci. USA 84 (1987) 4197

A B

> 30 -430C

s25 - 20c 20-

15-

10

5

I I I I I _I I _I 0 20 40 60 80 100 120 140 0 10 20 30 UV-seconds at 50cm minutes with .2M EMS FIG. 3. Mutagenesis in CSM98. Mutants were scored as rifampicin-resistant colonies after UV (A) or EtMes (EMS) (B) exposure. D37 values were 35 sec with UV irradiation and 16 min with EtMes exposure at 32 and 43°C. UV flux was 1 J m-2 sec-'.

showed normal mutagenesis at 32 and 430C for UV and soned that transformation ofCSM61 by apolC plasmid would EtMes. These results are similar to those with other wild-type produce unambiguous complementation of the DNA poly- (polA+, polC+) strains. A temperature difference is, there- merase III a-subunit defect and provide a specific test for the fore, not responsible for the results in Fig. 1. We conclude basis of the mutagenesis defect at 43°C. that an active a subunit ofDNA polymerase III is required for The strain CSM61 was made polC+ by introduction of a mutagenesis of the bacterial genome after exposure to UV cloned DNA polymerase III a-subunit gene on the plasmid light or EtMes. The results also rule out that the pcbAl allele pDS4-26 (9). Normal mutagenesis of the bacterial genome blocks mutagenesis, since that mutation is present in CSM98. was noted in such cells at 43°C (Fig. 4). The plasmid pSB5 Weigle Mutagenesis Requires DNA Polymerase m Activity. also restored the Weigle mutagenesis to wild-type levels in Weigle mutagenesis of bacteriophage X was measured as a ERli (Table 2). We conclude that the defect in mutagenesis function of induced mutagenesis (Table 2). At 430C, UV- is specifically due to a nonfuntional a subunit of DNA irradiated W3110 showed an increase in the number of polymerase III. mutants after UV treatment to the phage, which is similar to SOS Induction Is Normal When DNA Polymerase m Is the results obtained by Defais et al. (11). However, thepolCts Inactive. The defect in mutagenesis seen in polCts strains strain, ER11, showed no increase in the number of phage could be due to either a defect acting directly in the mutagenic mutants under conditions where DNA polymerase III is process or indirectly due to inhibition of the SOS response. inactive. Thus, the a subunit of DNA polymerase III is also The X construct XGE190 contains a recA-lacZ fusion (12). X required for Weigle mutagenesis. lysogens were made of W3110 (wild type), SB229 (recA), and DNA Polymerase m on a Plasmid Restores Mutagenesis. ERil (polCts) strains, and the SOS response was measured in The results above (Fig. 3) indicate that conversion of CSM14 these strains after exposure to UV light. The polCt, strain to a polC+ strain restores normal mutagenesis at 430C. The ERli showed a profile similar to that ofwild-type cells (Table polC gene has been cloned and characterized (9). We rea- 3); it had a 3- to 5-fold increase in ,B-galactosidase activity under SOS-inducing conditions where the recA control Table 2. Weigle mutagenesis showed no induction. Thus, it appears that a functional DNA polymerase III a subunit is not required for the induction of UV Mutagenesis ratet the SOS response. This conclusion is supported by the Reacti- dose to after irradiation of Weigle reactivation data, which show a wild-type level of vation cells, phage reactivation under conditions where DNA polymerase III is Strain factor* J/m2 0 J/m2 240 J/m2 inactive (Table 2). The role of DNA polymerase III in W3110 (polA+, polC+) 17 0 4.6 3.7 mutagenesis is, therefore, at a different step from induction. 30 5.8 30 DNA Recombination Is Normal When DNA Polymerase m ERll (polA+, polC486, 16 0 2.2 4.0 Is Inactive. The inducible mutagenic response in E. coli is pcbAl) 30 2.4 4.6 interdependent with functions involved in recombination. ER11(pSB5) (polA+, 11 0 2.6 1.5 Recombination was measured utilizing two plasmids that are poIC+/polC486, pcbAl) 30 5.5 45 derivatives ofpSV2neo (14) having a different deletion in the kanamycin-resistance gene. Plasmid A5 has a 248-base-pair *Reactivation factor = (number of plaques on UV-irradiated cells)/ deletion between the two Nar I sites. Plasmid A3x8 was (number ofplaques on non-UV-irradiated cells) at a UV dose of240 Nae I J/m2 to the phage. The plating efficiency ofX was within 1.2-fold on constructed by deleting 283 base pairs between the the three strains and was the same at 43°C in each case. Burst sizes sites and inserting an Xho I linker. Approximately 900 base of the strains were within 1.5-fold of each other. pairs are available between the two deletions for recombi- tMutagenesis was scored as the number of clear plaque mutants per nation. Similar rates of recombination were obtained at 32 104 wild-type plaques. and 43°C in CSM61 (Table 4) on selection for kanamycin Downloaded by guest on September 29, 2021 4198 Genetics: Hagensee et al. Proc. Natl. Acad. Sci. USA 84 (1987)

A B

aU 0

cn

0CO

0 20 40 60 80 100 120 140 seconds of UV at 50cm minutes with .2M EMS

FIG. 4. Mutagenesis in CSM61(pSD4-26). Mutants were scored as rifampicin-resistant colonies after UV (A) or EtMes (EMS) (B) exposure. D37 values were 35 sec at 32 and 43°C for UV irradiation and 20 min for EtMes exposure at 32 and 430C. UV flux was 1 J m-2 sec-'. resistance. Thus, DNA polymerase III is not required for DNA polymerase III. The defect in mutagenesis was com- plasmid recombination. It should be noted that plasmid plemented by the introduction of a chromosomal or a cloned recombination may be due to only one of several pathways DNA polymerase III gene, verifying that the defect in available (15). mutagenesis is specifically due to a nonfunctional a subunit Pl-mediated transduction oftetracycline drug resistance at of DNA polymerase III. DNA polymerase III is not needed 430C was within 2-fold ofthat at 320C. This also demonstrates for induction of SOS, P1, or plasmid recombination; thus, normal recombination when DNA polymerase III is inactive. SOS induction and mutagenesis can be fractionated. This It is not clear from these results if pcbAl allows DNA model is summarized in Fig. 5. polymerase I to fulfill a function normally performed by DNA The model shown should be viewed in terms of both UV polymerase III. and EtMes mutagenesis. UV mutagenesis we would suppose to be dependent on induction. However, our results indicate DISCUSSION EtMes, a small alkylating agent, gives the same results. EtMes mutagenesis is apparently not dependent on an The results presented here show that functional DNA poly- active merase III is required for mutagenesis of the bacterial recA or umuDC gene product (16-18). It may be that genome as well as Weigle mutagenesis. This is in accord with induction of another gene product is required. Alternatively, the conclusions of Bridges and Mottershead (16), who EtMes mutagenesis may operate independent of any induc- showed a decrease in the nonphotoreactivating (mutagenic) tion process. The common element in mutagenesis, accord- damage produced by UV irradiation in polCts cells under ing to our results, is DNA polymerase III, whether the conditions where DNA polymerase III was inactivated. Our damage results from a bulky or small adduct. experiments differ in that we have used strains that are viable A possible role for DNA polymerase III in mutagenesis is at the restrictive temperature for a temperature-sensitive that this enzyme misincorporates bases opposite DNA dam- age and, thereby, fixes the mutation in the genome. Recent Table 3. SOS /3-galactosidase assay Table 4. Plasmid recombination /-galactosidase activity, units per mg of extract 320C 430C No UV UV Strain RR PY RR PY Temperature, 0 30 0 30 W3110 Strain °C min min min min A5 5 10 7 3 A3x8 2 4 6 3 W3110 (polA+, polC+) 32 332 463 624 1690 A55+A3x8 220 4 170 2 43 291 376 533 1269 ER1l SB229 (recA56) 32 125 171 144 140 A5 4 2 1 1 43 108 121 88 82 A3x8 8 2 7 2 ER11 (polA+, polC486, 32 292 356 528 1304 A5+A3x8 313 3 160 2 pcbAl) 43 302 379 437 1661 RR, recombination rate [106 x (number of kanamycin-resistant 3-galactosidase activity is given as units per mg of extract from colonies)/(number of ampicillin-resistant colonies)]; PY, plasmid lysogens containing recA-lacZ fusions. yield [10-5 X (ampicillin-resistant colonies)/(,ug of input DNA)]. Downloaded by guest on September 29, 2021 Genetics: Hagensee et al. Proc. Natl. Acad. Sci. USA 84 (1987) 4199

II ! Mutation I -D~~e 0o Fixation umrecA rec A? > *I Commitment (Induction) Excision to 0 rec A Repair Mutagenesis

4i,&'ov

FIG. 5. Model of mutagenesis.

studies support the notion of an inducible factor interacting GM19122 and GM24711 and Robert A. Welch Foundation grant with the replication apparatus to increase mutagenesis. The Q-1027. M.E.H. is the recipient of an Achievement Reward for umuDC gene product is SOS inducible (19, 20) and may be College Scientists. required for replication beyond the DNA damage. The recA protein may also participate directly in this process (21-24). 1. McHenry, C. & Kornberg, A. (1977) J. Biol. Chem. 252, 6478-6484. Thus, DNA polymerase III-dependent synthesis could pro- 2. McHenry, C. S. & Crow, W. (1979) J. Biol. Chem. 254, 1748-1753. ceed to a site of damage and, under the influence of induced 3. McHenry, C. & Kornberg, A. (1981) in The Enzymes, ed. Boyer, proteins, misincorporate. A prediction of this view is that P. D. (Academic, New York), pp. 39-50. DNA polymerase III can be converted to a more error-prone 4. Gefter, M. L., Hirota, Y., Kornberg, T., Wechsler, J. & Barnoux, form by interaction with proteins not normally present in the C. (1971) Proc. Natl. Acad. Sci. USA 68, 3150-3153. Acad. or an induced processing the 5. Sevastopoulos, C. G. & Glaser, D. A. (1977) Proc. Natl. holoenzyme complex by protein Sci. USA 74, 3947-3950. a subunit to an altered form. This would predict that induced 6. Brotcorne-Lannoye, A., Maenhaut-Michel, G. & Radman, M. mutagenesis is expressed by way of the function of an altered (1985) Mol. Gen. Genet. 199, 64-69. complex containing DNA polymerase III and occurs during 7. Bryan, S. K. & Moses, R. E. (1984) J. Bacteriol. 158, 216-221. replication. UV irradiation inhibits replication, and it may be 8. Weschler, J. & Gross, J. (1971) Mol. Gen. Genet. 113, 273-284. 9. Shepard, D., Oberfelder, R., Welch, M. & McHenry, C. (1984) J. that the DNA polymerase III complex is altered by SOS- Bacteriol. 158, 455-459. induced proteins during the inhibition phase. This sequence 10. Hanahan, D. (1983) J. Mol. Biol. 166, 557-580. of events would commit the cell to mutagenesis with resump- 11. Defais, M., Farquet, P., Radman, M. & Errera, M. (1971) Virology tion of replication. 43, 495-503. 12. Weisemann, J. M., Funk, C. & Weinstock, G. M. (1984) J. Bacte- Lackey et al. (25, 26) have reported the isolation of an riol. 160, 112-121. altered form ofDNA polymerase I from cells induced for SOS 13. Miller, J. H. (1972) Experiments in Molecular Genetics (Cold functions. This polymerase, termed DNA polymerase I*, has Spring Harbor Laboratory, Cold Spring Harbor, NY), pp. 403. decreased fidelity as compared to DNA polymerase I in an in 14. Southern, P. J. & Berg, P. (1982) J. Mol. Appl. Genet. 1, 327-341. 15. James, A. A., Morrison, P. T. & Kolodner, R. (1982) J. Mol. Biol. vitro fidelity assay, although it appears to have normal 3' to 160, 411-430. 5' exonuclease activity. The results presented here do not 16. Bridges, B. A. & Mottershead, R. P. (1976) Mol. Gen. Genet. 144, demonstrate an in vivo role for DNA polymerase I*. Fur- 53-58. thermore, cells deficient in DNA polymerase I but containing 17. Schendel, P., Defais, M., Jeggo, P., Sampson, L. & Cairns, J. (1978) J. Pacteriol. 135, 466-475. DNA polymerase III show normal or increased mutagenesis 18. Kondo, S., Ichikawa, H., Iwo, K. & Kato, T. (1970) Genetics 66, (27). Our experiments do not rule out a participation by DNA 187-217. polymerase I, or an altered form of it, in mutagenesis, but 19. Bagg, A., Kenyon, C. & Walker, G. (1981) Proc. Natl. Acad. Sci. they do indicate that this enzyme alone is not sufficient. USA 78, 5749-5753. 60-93. our be to argue that 20. Walker, G. (1984) Microbiol. Rev. 48, Another way of viewing results would 21. Blanco, M., Herrera, G., Collado, P., Rabollo, J. & Botella, L. at 430C DNA polymerase I is responsible for replication in (1982) Biochimie 64, 633-636. our mutants and that the replicative complex containing DNA 22. Ennis, D., Fisher, B., Edmiston, S. & Mount, D. (1985) Proc. Natl. polymerase I is not converted to an error-prone mode under Acad. Sci. USA 82, 3325-3329. 23. Bridges, B. A. & Woodgate, R. (1985) Proc. Nati. Acad. Sci. USA inducing conditions. Whether this is due to greater accuracy 82, 4193-4197. of incorporation by DNA polymerase I, failure to interact 24. Lu, C., Scheuermann, R. H. & Echols, H. (1986) Proc. Natl. Acad. with an induced mutation-promoting factor, or failure to be Sci. USA 83, 619-623. 25. Lackey, D., Krauss, S. & Linn, S. (1982) Proc. Natl. Acad. Sci. acted upon to produce an altered form allowing mutation, we USA 79, 330-334. cannot yet determine. 26. Lackey, D., Krauss, S. & Linn, S. (1985) J. Biol. Chem. 260, 3178-3184. This work was supported by Public Health Service Grants 27. Witkin, E. M. & George, D. L. (1973) Genetics 73, Suppl., 91-108. Downloaded by guest on September 29, 2021