J. Mol. Biol. (1992) 223, 823-829
Genetical and Biochemical Evidence for the Involvement of the Coprotease Domain of Escherichia coli RecA Protein in Recombination Christophe Cazaux and Martine Defais Laboratoire de Pharmacologic et de Toxicobgie Fondamentales C.N.R.S., 205 route de Narbonne, 31077 Toulouse Cedex, France (Received 15 May
1991; accepted 4 November 1991)
RecA amino acid residue 204 is involved in the coprotease domain of the protein responsible for the induction of mutagenic repair. Two mutations were created at this site leading to the addition of either a methyl or an isopropyl group on the original glycine. Analyses of bath the in vivo and the in vitro properties of these mutated proteins demonstrated that this residue 204 is involved in many RecA activities, suggesting that this site could allosterically direct conformational changes in the protein or could be situated in a region interacting with many RecA cofactors. Keywords: RecA; recombination;
SOS Repair; strand-exchange
Escherichia coli RecA protein carries out two main functions: it is required for homologous recombination and post-replication repair (Clark, 1973), and it acts as a protease cofactor facilitating LexA-repressor cleavage, which leads to the induction of the SOS repair response when E. coli is exposed to DNA-damaging agents (Walker, 1984). RecA protein is involved in mutagenesis not only by controlling the posttranscriptional proteolytic activation of UmuD protein (Burckardt et al., 1988; Shinagawa et al., 1988) but also by exerting another as yet unknown role (Nohmi et al., 1988; Dutreix et al., 1989; Sweasy et aZ., 1990). In addition RecA protein participates in a few SOS functions, such as stable replication (Witkin & Kogoma, 1984) and induced replisome recovery (Khidhir et al., 1985). Thus, RecA protein performs multiple roles in E. coEi (Walker, 1984; Cox & Lehman, 1987). which suggests the presence of several functional domains in the protein. Many mutants of RecA protein have been characterized (Clark & Margulies, 1965; Blanc0 et al., 1975) and recently new mutants have been isolated, some of them identifying separate potential domains for coprotease and recombinase activities (Kawashima et al., 1984; Yarranton & Sedgwick, 1982; Larminat & Defais, 1989; Dutreix et aE., 1989; Cazaux et al., 1991) and others indicating that both domains may overlap in the vicinity of amino acid residue 204 (Wang & Tessman, 1986). Among the former, the recA430 mutation, where glycine residue 204 is replaced by a serine residue (Kawashima et aZ., 1984), is of particular interest since it reduces protease activity, as OO22-2836/92/0400823-07
$03.OWO
reaction; mutagenesis
shown by a decreased induction of both SOS response and prophage (Blanc0 et al., 1975; Devoret et al., 1983), and an inefficient UmuD cleavage (Nohmi et al., 1988) while it preserves recombination activity (Morand et al., 1977). The molecular bases of these effects have been investigated in vitro. It was first suggested that the primary defect of RecA430 protein was an impaired formation of the complex between the protein and single-stranded (sst) DNA in the presence of ATP (Wabiko et al., 1983). Then Ikawa et al. (1989) demonstrated that the binding itself of RecA430 protein to ssDNA was normal, but that the defect came from an inability of the protein to form a ternary complex by homology-independent conjunction of single and doublestranded (ds) DNA molecules. Recently it has been demonstrated that the steady-state affinity and the rate of association of RecA430 protein to ssDNA was reduced (Menetski & Kowalczykowski, 1990). ln order to understand better the functional role of the RecA region around residue 204 in both SOS repair and recombination, minor changes have been created three and seven residues downstream from GIy204 (Cazaux et al., 1991). These mutations strongly affect. all the in vivo and in vitro activities of RecA protein. In addition, the mutation recA142, which is mostly affected in recombination, has been located at residue 225 (Dutreix et al., 1985, 1989;
t Abbreviations used: ss, single-stranded; ds. doublestranded; u.v.. ultra-violet light; RecAwt, wild type RecA protein: SSB, single-stranded-DXA4 binding. 0
1992 Academic
Press Limited
824
C. Cazaux
and M. Defais
-
Table 1 In vivo
Allele
phenotypes
Survivalt (%)
of recA
mutant
SOS induction1
wtreeA recA430
100 0.7
50 1.2
AWA
< 0.0 1
I.0
wcA604 recA605
1.1 0.3
2.5 12
alleles Intrachromosomal recombination$ (mating cross-efficiency) 100 (100) 27 (100) Null
24 (92) 03 (6)
The recA430 gene is carried by plasmid pREU46 derived from pBEU2 (Uhlin et al., 1983) and prepared from E. coli strain JC10289 (kindly provided by A. J. Clark). Mutations recA604 and recA605 were constructed as previously described (Cazaux et al., 1991). The Ml3 phage DNA BamHI fragments carrying the recA604, recA605 or wild-type gene were cloned into the unique BarnHI site in pRR322 (Bolivar et al., 1977) to produce, respectively, pCC604. pCC605 and pFL352 plasmids. Plasmid-harbouring cells were grown with 100 pg ampicillin/ml in Luria broth (Difco) except for the SOS induction, for which M63 minimal medium, supplemented with 0,4e/, (w/v) glucose, 65’j/, (w/v) casamino acid and 1 pg thiamine/ml, was used (Miller, 1972). t u.v.-irradiation (30 J/m’) and cell survival assays were previously described (Calsou t Defais, 1985). Strain used was ,JC10289 that had been deleted for the chromosomal recA gene (Csonka et al., 1979). $ SOS induction was determined as previously described (Larminat & Defais, 1989). The induction factor is the ratio of /?-galactosidase activities after and before irradiation (Miller, 1972). Cells were exposed to 10 J/m2 and incubated at 37°C for 90 min after which tea : : la& induction was measured. 5 For the genetic recombination assay, 2 different methods were used. Intrachomosomal recombination was measured as described (Dutreix et al., 1989) using stain FL7023, which is derived from GY7023 and carries the deletion A(recA-srlR)306. Conjugational recombination (values in parentheses) was measured by mixing exponential cultures of about 2 x 10s cells/ml in the ratio of 1 Hfr to 10 F- recipients. Cells were incubated at 37°C for 30 min, after which mating was interrupted. Recombinants were selected on media containing 15 pg chloramphenicol/ml and 125 pg tetracycline/ ml. Hfr donor was GY7236, (HfrJ2 Zeu: : Tn!?: generously provided by R. Devoret) and recipient cells were plasmid-harbouring ,JC10289 strains. The data in the Table are the recombination frequencies relative to the wild-typetaken to be 100.
Kowalczykowski et al., 1989). However, another mutation that was produced upstream from Gly204, at residue 199, does not seem to decrease recombination (Cazaux et al., 1991). All these data suggest that RecA protein possesses a site between residues 204 and 225 involved in the recombination pathway. In this paper we describe the in viva and in vitro properties of two mutants leading to the replacement of the original Gly204 with either alanine or valine. These changes were expected to modify the interactions that possibly involve this site by progressively increasing steric hindrance and hydrophobic properties of residue 204. Our data demonstrate that residue 204 is not only essential for RecA protease activity but also participates actively in recombination. The recA430 allele displays a minor change of amino acid residue (Gly+Ser) leading to a decreased interaction of RecA protein with the repressors without greatly affecting recombination (Devoret et al., 1983). In order t,o analyse the role of residue 204 in the various RecA functions, we introduced new, minor modifications to the orginal glycine residue by adding first one methyl group (Ala), then an ispropyl group (Val), thus providing an increase in the steric hindrance and the hydrophobic feature of the protein. In this case, the last t’wo bases of codon 204 were changed (Cazaux et al., 1991). Both mutations led to new restriction sites that allow easy screening. recA604 and recA605 mutated genes were then cloned into the BamHI site of pBR322.
Two-dimensional electrophoresis showed that the purified mutated proteins have the same structural properties as wild-type RecA (data not shown). Table 1 shows that both recA604 and recA605 mutants, like recA430, are sensitive to ultra-violet light (u.v.), recA605 being twice as sensitive as the other two, without reaching the null phenotype. This confirms that neither of the chemical groups added at position 204 inactivated the RecA protein. Induction of the SOS response after u.v.-irradiation demonstrated that the recA605 mutant, was as affected as recA430 in LexA-repressor cleavage, whereas the recA604 mutant had an intermediate capability to induce the SOS response (Table 1). Thus the presence of one methyl group on amino acid residue 204 appears to affect only moderately the coprotease activity of RecA protein. However. when residue 204 carries either a hydroxyl (RecA430) or an isopropyl group (RecA605), the interaction between RecA protein and LexA repressor seems to be strongly hindered. This result’ was confirmed by measuring the in vitro cleavage of LexA repressor (generous gift from Dr M. Schnarr), in the presence of u.v.-irradiated supercoiled DNA (data not shown). In addition, neither recA604 nor recA605 mutants were able to induce Weiglereactivation, and no mutants could be detected in the phage progeny produced by these mutants (data not shown). This absence of error-prone repair in the recA604 mutant while the SOS response is still induced could be explained by an inability of the
825
Communications
0
5
IO 2040
60 00 120
0
5
IO 20 40 6080
120
Time (mm) RecA605
RecA604
Time (min)
603 4
is 40 E t :
30t
A
Genetical and Biochemical Evidence for the Involvement of the Coprotease Domain of Escherichia coli RecA Protein in Recombination Christophe Cazaux and Martine Defais Laboratoire de Pharmacologic et de Toxicobgie Fondamentales C.N.R.S., 205 route de Narbonne, 31077 Toulouse Cedex, France (Received 15 May
1991; accepted 4 November 1991)
RecA amino acid residue 204 is involved in the coprotease domain of the protein responsible for the induction of mutagenic repair. Two mutations were created at this site leading to the addition of either a methyl or an isopropyl group on the original glycine. Analyses of bath the in vivo and the in vitro properties of these mutated proteins demonstrated that this residue 204 is involved in many RecA activities, suggesting that this site could allosterically direct conformational changes in the protein or could be situated in a region interacting with many RecA cofactors. Keywords: RecA; recombination;
SOS Repair; strand-exchange
Escherichia coli RecA protein carries out two main functions: it is required for homologous recombination and post-replication repair (Clark, 1973), and it acts as a protease cofactor facilitating LexA-repressor cleavage, which leads to the induction of the SOS repair response when E. coli is exposed to DNA-damaging agents (Walker, 1984). RecA protein is involved in mutagenesis not only by controlling the posttranscriptional proteolytic activation of UmuD protein (Burckardt et al., 1988; Shinagawa et al., 1988) but also by exerting another as yet unknown role (Nohmi et al., 1988; Dutreix et al., 1989; Sweasy et aZ., 1990). In addition RecA protein participates in a few SOS functions, such as stable replication (Witkin & Kogoma, 1984) and induced replisome recovery (Khidhir et al., 1985). Thus, RecA protein performs multiple roles in E. coEi (Walker, 1984; Cox & Lehman, 1987). which suggests the presence of several functional domains in the protein. Many mutants of RecA protein have been characterized (Clark & Margulies, 1965; Blanc0 et al., 1975) and recently new mutants have been isolated, some of them identifying separate potential domains for coprotease and recombinase activities (Kawashima et al., 1984; Yarranton & Sedgwick, 1982; Larminat & Defais, 1989; Dutreix et aE., 1989; Cazaux et al., 1991) and others indicating that both domains may overlap in the vicinity of amino acid residue 204 (Wang & Tessman, 1986). Among the former, the recA430 mutation, where glycine residue 204 is replaced by a serine residue (Kawashima et aZ., 1984), is of particular interest since it reduces protease activity, as OO22-2836/92/0400823-07
$03.OWO
reaction; mutagenesis
shown by a decreased induction of both SOS response and prophage (Blanc0 et al., 1975; Devoret et al., 1983), and an inefficient UmuD cleavage (Nohmi et al., 1988) while it preserves recombination activity (Morand et al., 1977). The molecular bases of these effects have been investigated in vitro. It was first suggested that the primary defect of RecA430 protein was an impaired formation of the complex between the protein and single-stranded (sst) DNA in the presence of ATP (Wabiko et al., 1983). Then Ikawa et al. (1989) demonstrated that the binding itself of RecA430 protein to ssDNA was normal, but that the defect came from an inability of the protein to form a ternary complex by homology-independent conjunction of single and doublestranded (ds) DNA molecules. Recently it has been demonstrated that the steady-state affinity and the rate of association of RecA430 protein to ssDNA was reduced (Menetski & Kowalczykowski, 1990). ln order to understand better the functional role of the RecA region around residue 204 in both SOS repair and recombination, minor changes have been created three and seven residues downstream from GIy204 (Cazaux et al., 1991). These mutations strongly affect. all the in vivo and in vitro activities of RecA protein. In addition, the mutation recA142, which is mostly affected in recombination, has been located at residue 225 (Dutreix et al., 1985, 1989;
t Abbreviations used: ss, single-stranded; ds. doublestranded; u.v.. ultra-violet light; RecAwt, wild type RecA protein: SSB, single-stranded-DXA4 binding. 0
1992 Academic
Press Limited
824
C. Cazaux
and M. Defais
-
Table 1 In vivo
Allele
phenotypes
Survivalt (%)
of recA
mutant
SOS induction1
wtreeA recA430
100 0.7
50 1.2
AWA
< 0.0 1
I.0
wcA604 recA605
1.1 0.3
2.5 12
alleles Intrachromosomal recombination$ (mating cross-efficiency) 100 (100) 27 (100) Null
24 (92) 03 (6)
The recA430 gene is carried by plasmid pREU46 derived from pBEU2 (Uhlin et al., 1983) and prepared from E. coli strain JC10289 (kindly provided by A. J. Clark). Mutations recA604 and recA605 were constructed as previously described (Cazaux et al., 1991). The Ml3 phage DNA BamHI fragments carrying the recA604, recA605 or wild-type gene were cloned into the unique BarnHI site in pRR322 (Bolivar et al., 1977) to produce, respectively, pCC604. pCC605 and pFL352 plasmids. Plasmid-harbouring cells were grown with 100 pg ampicillin/ml in Luria broth (Difco) except for the SOS induction, for which M63 minimal medium, supplemented with 0,4e/, (w/v) glucose, 65’j/, (w/v) casamino acid and 1 pg thiamine/ml, was used (Miller, 1972). t u.v.-irradiation (30 J/m’) and cell survival assays were previously described (Calsou t Defais, 1985). Strain used was ,JC10289 that had been deleted for the chromosomal recA gene (Csonka et al., 1979). $ SOS induction was determined as previously described (Larminat & Defais, 1989). The induction factor is the ratio of /?-galactosidase activities after and before irradiation (Miller, 1972). Cells were exposed to 10 J/m2 and incubated at 37°C for 90 min after which tea : : la& induction was measured. 5 For the genetic recombination assay, 2 different methods were used. Intrachomosomal recombination was measured as described (Dutreix et al., 1989) using stain FL7023, which is derived from GY7023 and carries the deletion A(recA-srlR)306. Conjugational recombination (values in parentheses) was measured by mixing exponential cultures of about 2 x 10s cells/ml in the ratio of 1 Hfr to 10 F- recipients. Cells were incubated at 37°C for 30 min, after which mating was interrupted. Recombinants were selected on media containing 15 pg chloramphenicol/ml and 125 pg tetracycline/ ml. Hfr donor was GY7236, (HfrJ2 Zeu: : Tn!?: generously provided by R. Devoret) and recipient cells were plasmid-harbouring ,JC10289 strains. The data in the Table are the recombination frequencies relative to the wild-typetaken to be 100.
Kowalczykowski et al., 1989). However, another mutation that was produced upstream from Gly204, at residue 199, does not seem to decrease recombination (Cazaux et al., 1991). All these data suggest that RecA protein possesses a site between residues 204 and 225 involved in the recombination pathway. In this paper we describe the in viva and in vitro properties of two mutants leading to the replacement of the original Gly204 with either alanine or valine. These changes were expected to modify the interactions that possibly involve this site by progressively increasing steric hindrance and hydrophobic properties of residue 204. Our data demonstrate that residue 204 is not only essential for RecA protease activity but also participates actively in recombination. The recA430 allele displays a minor change of amino acid residue (Gly+Ser) leading to a decreased interaction of RecA protein with the repressors without greatly affecting recombination (Devoret et al., 1983). In order t,o analyse the role of residue 204 in the various RecA functions, we introduced new, minor modifications to the orginal glycine residue by adding first one methyl group (Ala), then an ispropyl group (Val), thus providing an increase in the steric hindrance and the hydrophobic feature of the protein. In this case, the last t’wo bases of codon 204 were changed (Cazaux et al., 1991). Both mutations led to new restriction sites that allow easy screening. recA604 and recA605 mutated genes were then cloned into the BamHI site of pBR322.
Two-dimensional electrophoresis showed that the purified mutated proteins have the same structural properties as wild-type RecA (data not shown). Table 1 shows that both recA604 and recA605 mutants, like recA430, are sensitive to ultra-violet light (u.v.), recA605 being twice as sensitive as the other two, without reaching the null phenotype. This confirms that neither of the chemical groups added at position 204 inactivated the RecA protein. Induction of the SOS response after u.v.-irradiation demonstrated that the recA605 mutant, was as affected as recA430 in LexA-repressor cleavage, whereas the recA604 mutant had an intermediate capability to induce the SOS response (Table 1). Thus the presence of one methyl group on amino acid residue 204 appears to affect only moderately the coprotease activity of RecA protein. However. when residue 204 carries either a hydroxyl (RecA430) or an isopropyl group (RecA605), the interaction between RecA protein and LexA repressor seems to be strongly hindered. This result’ was confirmed by measuring the in vitro cleavage of LexA repressor (generous gift from Dr M. Schnarr), in the presence of u.v.-irradiated supercoiled DNA (data not shown). In addition, neither recA604 nor recA605 mutants were able to induce Weiglereactivation, and no mutants could be detected in the phage progeny produced by these mutants (data not shown). This absence of error-prone repair in the recA604 mutant while the SOS response is still induced could be explained by an inability of the
825
Communications
0
5
IO 2040
60 00 120
0
5
IO 20 40 6080
120
Time (mm) RecA605
RecA604
Time (min)
603 4
is 40 E t :
30t
A
