Some mismatch repair activities in Escherichia coli

16 downloads 0 Views 1MB Size Report
plex DNA molecules of bacteriophage A has provided evi- dence for the mode of action oftwo mismatch repair systems in E. coli (5, 6). One of these, the adenineĀ ...
Proc. Nati. Acad. Sci. USA

Vol. 85, pp. %74-9678, December 1988

Genetics

Some mismatch repair activities in Escherichia coli (bacteriophage A/repair mutants/recombination)

J. PABLO RADICELLA*, ELIZABETH A. CLARK, AND MAURICE S. Foxt Department of Biology, Massachusetts Institute of Technology, Cambridge, MA 02139

Contributed by Maurice S. Fox, September 12, 1988

of repair capacities, one mismatch at a time. DNA is isolated from phage harboring conditional (amber) mutations in functions essential for DNA replication, the strands are separated, and complementary strands from each of two different mutants are hybridized. The heteroduplex molecules are packaged in vitro and the assembled phage can be plated under permissive conditions, to determine the phage yield, or on hosts that do not suppress the amber mutations. In the latter case, plaque formation requires repair of the mutant nucleotide on the transcribed strand, without co-correction of the neighboring wild-type nucleotide that is, in turn, mismatched with a mutant nucleotide on the untranscribed strand. Thus, plaque formation reflects repair of a particular mismatch. This strategy has revealed the presence of two mismatch repair capabilities present in E. coli. One of these functions removes the A of C-A and G-A mismatches and the other corrects a C in a C C mismatch. When the A removal function is disabled, the bacteria display a mutator phenotype. Evidence is provided that suggests that the A removal in C&A and G-A mismatches is the same activity as the transversion avoidance activity encoded by the recently reported mutY locus (12). The biochemical evidence reported for A removal in G-A mismatches (13, 14) appears to be an activity of the same function.

Heterozygous bacteriophage A DNA moleABSTRACT cules, whose replication requires mismatch correction of a mutant nucleotide in the transcribed strand, provide an assay for localized mismatch repair in Escherichia coli. We describe two systems: one removes the A in C-A or G-A mismatches and the other removes one or the other C in a C C mismatch. Mutations disabling the first system result in a mutator phenotype that may be identical to mutY.

The products of genetic crosses, from bacteriophage to fungi, reveal an unexpectedly high frequency of clustered exchanges among closely linked markers (1, 2). Investigation of the origin of these clustered exchanges, in bacteriophage A crosses (3), has provided evidence for the proposal that recombinants reflecting separation of closely linked markers arise by a mechanism different from the breakage and joining events that separate distant markers. The primary products of recombination harbor regions of heteroduplex DNA that are, on the average, =4 kilobases long (4). These regions, composed of one DNA strand from one of the parents and the complementary strand from the other, may join homoduplex regions from the same parent or from different parents. When markers reflecting base alterations that distinguish the two parents reside within the heteroduplex regions, base pair mismatches are present. The Escherichia coli host encodes an array of functions with various specificities capable of excising one or the other of the bases in the mismatch and repairing the molecule by using the complementary strand as a template. Investigation of the products present in infective centers of bacteria transfected with artificially constructed heteroduplex DNA molecules of bacteriophage A has provided evidence for the mode of action of two mismatch repair systems in E. coli (5, 6). One of these, the adenine methylationdirected mismatch repair system, has been shown to play an important role in replication fidelity by repairing errors in a newly replicated DNA strand, distinguishing that strand by the absence of methylation at GATC sequences. This repair process directs the replacement of long tracts of the undermethylated strand (several kilobases) when a base pair mismatch is present (7). Some of the functions repaired for this mismatch repair system are encoded by mutL, mutS, mutH, and mutU (uvrD) (8). A localized repair activity became evident when the methyl-directed mismatch repair system was disabled as a result of mutations in mutH or mutU (5). This localized repair results in separation of very closely linked markers, requires the functions mutL and mutS, is independent of adenine methylation, efficiently corrects the T of G-T mismatches resulting from a C to T transition mutation in the inner C in CC(A/T)G sites (6), and is similar to the very short patch repair described by Lieb (9, 10). Serendipitous evidence of additional localized mismatch repair (11) suggested a strategy that allows the investigation

MATERIALS AND METHODS Phage and Bacteria. Bacteriophage A with Pam mutations are from the laboratory collection. Oam mutants were obtained from Mark Furth (15). All the Su- E. coli strains are derivatives of M182 that is Su- lacZ X74 galU galK Sm' (E. Signer, MIT). The Su' strain is BNN45 (met gal lacY r-m suit su"') (D. Botstein, MIT). MC109 (Su- mutY) was obtained from J. Miller (12). mutS201::TnS and mutL218::TnJO (E. Siegel, Tufts University, Medford, MA) mutations were introduced into M182, BNN45, and MC109 by P1 transductions following the procedures described by Miller (16). The presence of mutator activity was determined by monitoring the frequency of rifampicin-resistant mutants. The set of Hfr strains harboring a TnlO insertion close to the origin of transfer as well as strains carrying the markers used for P1 mapping were obtained from C. Gross (University of Wisconsin-Madison). The relevant mutation in each strain is zdg2JO::TnJO (CG18585) and nupG51::TnJO (CG18472). CGSC6902 (JM2071) carrying the gaiP::TnJO marker came from J. Miller (12). Plasmid pTP166 for overproduction of the adenine methylase was from M. Marinus (17). Media and buffers used and the conditions for titering phage have already been described (18). To determine the nucleotide substitutions created by the Oam mutations, DNA was isolated from the phages (19), and

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Ā§1734 solely to indicate this fact.

*Present address: Institute of Molecular Biology, University of Oregon, Eugene, OR 97403. tTo whom reprint requests should be addressed. 9674

Genetics: Radicella et al. the relevant portion of the 0 gene for each of the mutants was sequenced by the dideoxy chain-termination method (20). Tetracycline-sensitive derivatives of TnJO-containing strains were obtained by selection with fusaric acid (21). Preparation of Heteroduplex-Containing Phage. Phage stocks were grown in BNN45 except when hypermethylated DNA was required; in that case, the strain used was BNN45 carrying the plasmid pTP166 (17). Phage stocks were concentrated as described by Lichten and Fox (22). Mixed random heteroduplex molecules were prepared as described by Pearson and Fox (11). DNA denaturation and strand separation with poly(U, G) were carried out by the method of Davis et al. (23). Strand annealing was performed as described by Raposa and Fox (6), with hybridization at 500C for 1 hr. Self hybridization gave 50% in one case and 2-3 times lower in the other (Oam905), suggesting the possibility that the nucleotide context of the

mismatch influences the correction efficiency. Jones et al.

(30) have reported a context-dependent mismatch correction efficiency for

the

methyl-directed system,

and Lu and

Chang

(13) have reported evidence for a context-dependent biochemical activity for

C-A

moval from

or

G-A G-A

mismatch

correction.

mismatches does

depend on adenine methylation in GATC coli

The

A

appear

re-

to

sequences.

G-A to G-C repair activity has been identified in extracts (13, 14) that is independent of adenine methyl-

Recently,

E.

not

a

ation and of mutL, mutS, and mutHgene products. Although no CUA correction was detected in vitro, this biochemical activity appears to display properties very similar to MicA. The presence of a mutant cytosine (Oam205) in the

transcribed strand mismatched with a cytosine in the complementary strand can also result in a high yield of infective centers. In this case, however, the yield appears to depend on the relative location of the mutant nucleotide on the complementary strand. When the mutant nucleotide on the 1 strand is to the right of the C C mismatch, in the R direction on the A map, about half of the phage form plaques. When the mutant nucleotide on the l strand is located nearby and to the left of the mismatch, very few of the phage form plaques but the yield increases with the nucleotide distance separating the

9678

Genetics: Radicella et al.

two mutations. These observations are consistent with the view that the repair process that removes the cytosines in a CC mismatch acts equally frequently on either cytosine, initiating the correction at or near to the C-C mismatch, with nucleolytic excision in the 5' to 3' direction of =900 nucleotides, The C'C repair does not appear to depend on the degree of adenine methylation in GATC sequences or on the functions of mutH or uvrD. The other mismatches that have been examined, removal of T in GOT mismatches, removal of an A in ALA mismatches, and removal of AA in CAAA mismatches, all largely escape correction, and the heteroduplexes harboring them fail to give rise to significant yields of plaques. We have succeeded in isolating mutants, micA, defective in the correction of adenines in COA and G-A mismatches but remaining competent in a correction of a C in CC mismatches. When transferred out of the mutL background in which they were isolated, these mutants display a mutator phenotype. The mutations are linked by transduction to nupGSJJ, mapping at -64 minutes in the E. coli chromosome. Nghiem et al. (12) have reported a mutator function, mut Y, that appears to map in the same position as micA. Bacteria that are mutY are enhanced for G-C to TEA transversions, and we have shown that a mutY mutS strain is defective in correction of adenines in C-A or G-A mismatches, mutY and micA seem to define the same function, a function playing a role in maintenance of replication fidelity by acting on the product of DNA replication. The complementary function mutT (31), which plays a role in T-A to G-C mutagenesis, must therefore intervene in the replication process itself. The role of the C C correction system remains to be elaborated. Examination of 5 of the 12 possible mismatches whose correction we can detect has revealed the presence of two mismatch correction systems that can separate closely linked markers. It will be interesting to see whether the future examination of the other 7 possible mismatches will reveal the presence of other systems. We thank Min Zhu for her help in the sequencing, and Karyn Baum for her help in examining the activity of mutL and mutS packaging mixes. This work has been supported by the National Institutes of Health Grant A105388 to M.S.F. 1. Amati, P. & Meselson, M. (1965) Genetics 51, 369-379. 2. Pritchard, R. H. (1960) Genet. Res. 1, 1-24. 3. White, R. L. & Fox, M. S. (1974) Proc. Natl. Acad. Sci. USA 71, 1544-1548.

Proc. Natl. Acad. Sci. USA 85 (1988) 4. Huisman, 0. & Fox, M. S. (1986) Genetics 112, 409-420. 5. Fox, M. S. & Raposa, S. (1983) in Cellular Responses to DNA Damage, eds. Friedberg, E. C. & Bridges, B. A. (Liss, New York), pp. 333-335. 6. Raposa, S. & Fox, M. S. (1987) Genetics 117, 381-390. 7. Meselson, M. (1988) in The Recombination of Genetic Material, ed. Low, K. B. (Academic, San Diego), pp. 91-113. 8. Lu, A. L., Clark, S. & Modrich, P. (1983) Proc. Natl. Acad. Sci. USA 80, 4639-4643. 9. Lieb, M. (1963) Mol. Gen. Genet. 191, 118-125. 10. Lieb, M., Allen, E. & Read, D. (1986) Genetics 114, 1041-1060. 11. Pearson, R. & Fox, M. S. (1988) Genetics 118, 5-12. 12. Nghiem, Y., Cabrera, M., Cupples, C. G. & Miller, J. H. (1988) Proc. Natl. Acad. Sci. USA 85, 2709-2713. 13. Lu, A. L. & Chang, D. Y. (1988) Genetics 118, 593-600. 14. Su, S. S., Lahue, R. S., Au, K. G. & Modrich, P. (1988) J. Biol. Chem. 263, 6829-6835. 15. Furth, M. E., McLeester, C. & Dove, W. F. (1978) J. Mol. Biol. 126, 195-225. 16. Miller, J. H. (1972) Experiments in Molecular Genetics (Cold Spring Harbor Lab., Cold Spring Harbor, NY). 17. Marinus, M. G., Poteete, A. & Arraj, J. A. (1984) Gene 28, 123-125. 18. Pearson, R. & Fox, M. S. (1988) Genetics 118, 13-19. 19. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Lab., Cold Spring Harbor, NY). 20. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467. 21. Bochner, B. R., Huang, H. C., Schieven, G. L. & Ames,13. N. (1980) J. Bacteriol. 143, 926-933. 22. Lichten, M. & Fox, M. S. (1983) Nucleic Acids Res. 11, 39593971. 23. Davis, R. W., Botstein, D. & Roth, J. R. (1980) Advanced Bacterial Genetics (Cold Spring Harbor Lab., Cold Spring Harbor, NY). 24. Rosenberg, S. M., Stahl, M. M., Kobayashi, I. & Stahl, F. W. (1985) Gene 38, 165-176. 25. Rosenberg, S. M. (1987) Cell 48, 855-865. 26. Way, J. C., Davis, M. A., Morisato, D., Roberts, D. E. & Kleckner, N. (1984) Gene 32, 369-379. 27. Roberts, D. E. (1987) Dissertation (Harvard Univ., Cambridge, MA). 28. Reiser, W. (1983) Dissertation (Univ. of Heidelberg, Heidelberg). 29. Gellert, M. & Bullock, M. L. (1970) Proc. Natl. Acad. Sci. USA 67, 1580-1587. 30. Jones, M., Wagner, R. & Radman, M. (1987) Genetics 115, 605610. 31. Cox, E. C. (1973) Genetics 73 (Suppl.), 67-80. 32. Sanger, F., Coulson, A. R., Hong, G. F., Hill, D. F. & Petersen, G. B. (1982) J. Mol. Biol. 162, 729-773.