Amplified UvrA protein can ameliorate the ultraviolet sensitivity of an Escherichia coli recA mutant

Amplified UvrA protein can ameliorate the ultraviolet sensitivity of an Escherichia coli recA mutant

Mutation Research 487 (2001) 149–156 Amplified UvrA protein can ameliorate the ultraviolet sensitivity of an Escherichia coli recA mutant Kazuhiro Ki...

120KB Sizes 0 Downloads 28 Views

Mutation Research 487 (2001) 149–156

Amplified UvrA protein can ameliorate the ultraviolet sensitivity of an Escherichia coli recA mutant Kazuhiro Kiyosawa a , Masashi Tanaka a,b , Tsukasa Matsunaga c , Osamu Nikaido c , Kazuo Yamamoto a,b,∗ a

c

Department of Biomolecular Sciences, Graduate School of Life Sciences, Tohoku University, Sendai 980-8578, Japan b Biotechnology Laboratory, Takasaki Radiation Chemistry Research Establishment, Japan Atomic Energy Research Institute, Takasaki 370-1292, Japan Division of Radiation Biology, Faculty of Pharmaceutical Sciences, Kanazawa University, Kanazawa 920-0934, Japan Received 17 May 2001; received in revised form 19 September 2001; accepted 19 September 2001

Abstract When a recA strain of Escherichia coli was transformed with the multicopy plasmid pSF11 carrying the uvrA gene of E. coli, its extreme ultraviolet (UV) sensitivity was decreased. The sensitivity of the lexA1 (Ind− ) strain to UV was also decreased by pSF11. The recA cells expressing Neurospora crassa UV damage endonuclease (UVDE), encoding UV-endonuclease, show UV resistance. On the other hand, only partial amelioration of UV sensitivity of the recA strain was observed in the presence of the plasmid pNP10 carrying the uvrB gene. Host cell reactivation of UV-irradiated ␭ phage in recA cells with pSF11 was as efficient as that in wild-type cells. Using an antibody to detect cyclobutane pyrimidine dimers, we found that UV-irradiated recA cells removed dimers from their DNA more rapidly if they carried pSF11 than if they carried a vacant control plasmid. Using anti-UvrA antibody, we observed that the expression level of UvrA protein was about 20-fold higher in the recA strain with pSF11 than in the recA strain without pSF11. Our results were consistent with the idea that constitutive level of UvrA protein in the recA cells results in constitutive levels of active UvrABC nuclease which is not enough to operate full nucleotide excision repair (NER), thus leading to extreme UV sensitivity. © 2001 Elsevier Science B.V. All rights reserved. Keywords: recA strain; uvrA gene; SOS response; Nucleotide excision repair

1. Introduction Irradiation of DNA with ultraviolet (UV) light produces a variety of photoproducts, of which the major species are cyclobutane pyrimidine dimers (CPDs) and pyrimidine-pyrimidone [6-4] adducts (6-4 adducts). Both lesions, if not repaired, cause mutagenesis and cell death. To survive in a UV-rich environment, E. coli is equipped with an inducible response known as ∗ Corresponding author. Tel.: +81-22-217-6706; fax: +81-22-217-6706. E-mail address: [email protected] (K. Yamamoto).

the SOS response regulated by the recA–lexA regulon [1,2]. The SOS response aids survival by combining increased expression of genes involved in nucleotide excision repair (NER) and recombinational repair mechanisms. The genes recA for the recombination enzyme, RecA, and uvrA and uvrB for subunits of the UvrABC NER enzymes, UvrA and UvrB, have SOS boxes [3] that are bound by the LexA repressor under physiological conditions. Upon UV irradiation, the constitutive amount of RecA protein binds single-stranded DNA resulting from replication blocks and acts as a coprotease for inactivation of LexA. The levels of RecA, UvrA and UvrB increase, as do

0921-8777/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 1 - 8 7 7 7 ( 0 1 ) 0 0 1 1 4 - 8

150

K. Kiyosawa et al. / Mutation Research 487 (2001) 149–156

the cell’s recombinational repair and NER activities. Upon completion of repair, the inducing signal disappears and the cells return to the preinduction state. Mutations in the uvrA or uvrB gene inhibit the NER and cause sensitivity to UV radiation. Mutations in the recA genes result in inhibition of the SOS response. Thus recA-defective cells are extremely sensitive to UV radiation due to defects in recombinational repair and reduced or constitutive levels of NER activity. NER in E. coli is accomplished by UvrABCD proteins. It involves recognition of DNA damage by UvrA and UvrB in the form of a UvrA2 UvrB1 complex, incision of the damaged DNA strand by UvrB and UvrC, removal of the damaged region by UvrD helicase, followed by repair synthesis, which fills the gap using the intact strand as a template, and is completed by ligation of the repaired section to the undamaged DNA [4]. Certain classes of mutations in the lexA gene, called lexA (Ind− ), result in loss of the SOS response [5]. Thus, lexA (Ind− ) cells are UV-sensitive because they only have constitutive levels of RecA protein for recombinational repair and UvrABC enzyme for NER. It was demonstrated that a recA-deficient lexA41 double mutant, the latter of which encodes temperature-labile LexA protein [6], showed higher survival after UV radiation than a recA strain with the wildtype lexA gene [7–9]. When the strains were excision deficient, for example when carrying an uvrA6 mutation, there was no difference between the UV-survival

of lexA41 and lexA+ [9]. The LexA41 protein is defective as a repressor and thus permits derepression of the LexA regulon genes including uvrA and uvrB in recA cells and leads to increased survival due to more efficient NER. Extreme UV sensitivity of the recA strains is thought to be due to their inability to induce UvrA and UvrB proteins. As mentioned above, UvrA protein dimerizes and forms an A2 B1 complex with UvrB protein. Since the numbers of UvrA and UvrB proteins in recA cell are estimated to be 20 and 140, respectively [10,11], about 10 UvrA2 B complexes will be formed per cell. In other words, 130 unbound UvrB molecules will remain in the recA cell. If this estimation is correct, amplification of UvrA protein but not UvrB protein in recA strains would be sufficient to obtain UV resistance. In this study, the effects of amplification of UvrA, UvrB or both UvrA and UvrB proteins in the recA cells on survival following UV exposure and on dimer excision were studied. Under such conditions, UV sensitivity and dimer excision of the recA cells were alleviated by amplification of UvrA protein but not UvrB protein.

2. Materials and methods 2.1. Bacterial strains, plasmids and phage The bacterial strains and plasmids used are listed in Table 1. FS03 was constructed by eliminating the Tn10 trasposon from KY1225 according to Bochner et al.

Table 1 E. coli strains and plasmids Strains

Relevant genotype

Origin or source

AB1157

[27]

KY1220 KY1221 KY1225 KY1226 FS03 DH5␣MCR

thr-1 his-4 argE3 proA2 thi-1 lacY1 galK2 xyl-5 mtl-1 tsx-33 rpsL31 supE44 As AB1157 but phr-36::Cmr As KY1220 but lexA1 malE::Tn10 As KY1220 but recA56 srlC::Tn10 As KY1220 but uvrA6 malE::Tn10 As KY1225 but Tets recA1

[13] [13] [13] [13] This study [28]

Plasmids

Relevant marker

Origin or source

pUVDE pJA61 pNP10 pSF11

Ampr ,

Neurospora crassa UVDE gene, uvrA gene, Ampr , Tetr , pBR322 ori uvrB gene, Tetr , pBR322 ori As pJA61 but Tets , pBR322 ori

pBR322 ori

[19] [29] [30] This study

K. Kiyosawa et al. / Mutation Research 487 (2001) 149–156

151

[12]. To rule out uncontrolled CPD photoreactivation, all the strains used carried phr-36 allele [13]. Plasmid pSF11 was constructed by deleting the BamHI-HindIII fragment from pJA61, preserving the uvrA gene but deleting the Tetr gene. Lambda phage used is clear plaque mutant ␭CI− .

various UV influences at a rate of 1.0 J m−2 s−1 . The samples were diluted and pour-plated on LB-plates with stationary cultures of appropriate indicator bacteria for the measurement of host cell reactivation (HCR), and the plates were incubated overnight at 30 ◦ C.

2.2. Media and reagents

2.6. Enzyme-linked immunosorbent assay (ELISA) for cyclobutane pyrimidine dimers (CPD)

Luria–Bertani (LB) broth, LB-plates, M9 buffer and phosphate buffer were described as previously [14]. Ampicillin (Amp), chloramphenicol (Cm) and tetracycline (Tet) were included, if necessary, in LB-broth and LB-plate at concentrations of 50, 30 and 10 ␮g ml−1 , respectively. Minimal medium used was M9 buffer with 1% glucose and 1% casamino acids. 2.3. UV irradiation Cells to be irradiated with UV were grown overnight in LB-broth. They were centrifuged, washed in phosphate buffer and resuspended to their original volume in phosphate buffer or M9 buffer. Cells in glass dishes 90 mm in diameter were exposed to 254 nm UV radiation from a germicidal lamp. 2.4. Western blot analysis For determination of the cellular levels of UvrA protein, extract of FS03/pSF11 were made from over night culture of cells. Cells were pelleted by centrifugation and lysed using FastPROTEINTM BLUE kit (BIO 101, Inc., USA) following manufacturer’s instructions. The extracts were boiled in loading buffer containing 2.5% SDS for 10 min. Equal amounts of the resulting extracts were loaded onto SDS-7.5% polyacrylamide gels and electrophoresed, and the proteins were transferred to PVDF membranes. The blots were probed with anti-UvrA protein monoclonal antibodies [15] and developed using Western-LightTM and Western-StarTM kit (TROPIX, Inc., USA), and their amounts were quantified by phosphorimager analysis and associated software. 2.5. Host cell reactivation Lambda phage suspensions were diluted to 2 × 105 PFU ml−1 in 0.01 M MgSO4 and irradiated with

Binding of the antibody to CPD was measured by ELISA. Details of the procedure were described previously [16,17]. Briefly, 40 J m−2 UV-irradiated cells were resuspended in minimal medium and incubated with shaking for various periods, followed by DNA extraction. ELISA was performed in 96-well poly microtiter plates precoated with 1% protamine sulfate. For CPD detection with TDM-2 antibodies [18], 15 ng of DNA was used. After adding biotinylated F (ab )2 goat anti-mouse IgG fragments and streptavidinperoxidase, the absorbance (OD) of colored products from o-phenylene-diamine was measured at 492 nm.

3. Results 3.1. Effects of uvr-plasmid on UV sensitivity of recA strains Ganesan and Hanawalt [9] reported that the efficiency of UV damage excision repair increased more rapidly in the recA lexA41 (Def) cells than in the recA lexA+ cells. Derepression of the uvr genes due to defective LexA41 protein leads to high levels of UvrABC protein complexes in the recA cells, and thus leads to increased survival due to more efficient NER. We were interested in whether increased production of UvrA and/or UvrB proteins is involved in UV resistance in recA lexA41 cells. When E. coli FS03 (recA56) cells transformed with the multicopy plasmid pSF11 carrying the uvrA gene were irradiated with UV, their extreme UV sensitivity was decreased (Fig. 1). We also found the same potentiation effect of pSF11 on DH5␣MCR carrying a different recA allele, recA1 (data not shown). We digested pSF11 with EcoRV to make 210 bp deletion of the uvrA gene but leaving 5 and 3 flanking regions intact. Transformation of FS03 with this uvrA

152

K. Kiyosawa et al. / Mutation Research 487 (2001) 149–156

Fig. 1. UV survival of E. coli cells in the presence of uvr- or UVDE-plasmid (filled symbols) or vacant plasmid (open symbols). Experiments were repeated several times and highly reproducible, and representative results from a single experiment are shown. (䊊) KY1220 (phr36); (䊐) FS03 (recA56 phr36); (䉲) FS03/uvrA; (䉬) FS03/uvrB; (䉱) FS03/uvrA uvrB; (䊏) FS03/UVDE; (夽) KY1221 (lexA1 phr36); (夹) KY1221/uvrA.

deletion plasmid indicated that the plasmid did not show the potentiation effect (data not shown). Thus, UV resistance of FS03/pSF11 was inferred to be due to the existence of the uvrA gene but not 5 and 3 flanking regions. FS03 transformed with the multicopy plasmid pNP10 carrying the uvrB gene showed no significant changes in their extreme UV sensitivity (Fig. 1). When uvrB deficient strains were transformed with pNP10, their extreme UV sensitivi-

ties were decreased (data not shown). Thus, the uvrB gene in the pNP10 is active. FS03 cells carrying both pSF11 and pNP10 plasmids were as resistant to UV as FS03 transformed with pUVDE plasmid, which carries the Neurospora crassa UV damage endonuclease, a distinct repair system from NER [19]. The uvrA-plasmid was also able to complement the UV sensitivity of KY1221 (lexA1). Thus, amplification of UvrA protein but not UvrB protein was sufficient to

K. Kiyosawa et al. / Mutation Research 487 (2001) 149–156

confer increased UV-resistance on the recA strain as well as lexA(Ind− ) strain. 3.2. Host cell reactivation of FS03 carrying uvr-plasmids HCR is due to excision of CPD from phage DNA [20], and was shown to be less efficient in recA mutants than in recA+ cells [21–23]. The results shown in Fig. 2 indicate that FS03 carrying uvrA-plasmid, both uvrA- and uvrB-plasmids or UVDE-plasmid

153

reactivated UV-irradiated ␭ phage as efficiently as KY1220 (uvr+ ). On the other hand, HCR in FS03 carrying the uvrB-plasmid was similar to that of FS03 strain. Thus, amplification of UvrA protein alone is enough to sustain sufficient HCR capacity in recA strain. 3.3. Excision of CPDs in FS03 carrying uvr-plasmids To determine more precisely the extent of UV damage repair in FS03, we assayed the amounts of CPDs

Fig. 2. Host cell reactivation of bacteriophage ␭ in E. coli cells in the presence of uvr- or UVDE-plasmid (filled symbols) or vacant vector (open symbols). Experiments were repeated several times and were reproducible. The results from one representative experiment done in parallel are shown. (䊊) KY1220 (phr36); (䉭) KY1226 (uvrA6 phr36); (䊐) FS03 (recA56 phr36); (䉲) FS03/uvrA; (䉬) FS03/uvrB; (䉱) FS03/uvrA uvrB; (䊏) FS03/UVDE.

154

K. Kiyosawa et al. / Mutation Research 487 (2001) 149–156

Fig. 3. Removal of CPD from DNA in UV-irradiated E. coli cells. The amounts of CPDs remaining are shown for cells in the presence of uvr-plasmid (filled symbols) or vacant vector (open symbols). The results represent the mean values from experiments repeated twice. (䊊) KY1220; (䉭) KY1226; (䊐) FS03; (䉲) FS03/uvrA; (䉬) FS03/uvrB; (䉱) FS03/uvrA uvrB.

remaining in the DNA using monoclonal antibodies against CPDs. As shown in Fig. 3, the removal of CPD from FS03 carrying pSF11 was more rapid than that from FS03 carrying vacant plasmid and was almost equivalent to that of KY1220. On the other hand, the rate of removal of CPD from FS03 carrying the uvrB-plasmid was essentially the same as that from FS03. These results support the assumption that UvrA protein is the limiting factor for NER in uninduced cells and induction of uvrA is required for full repair.

3.4. Amounts of UvrA protein in the recA cells We determined the amounts of UvrA protein in FS03 cells carrying pSF11 plasmid using anti-UvrA protein antibody. As shown in Fig. 4, a band corresponding 104 kDa UvrA protein was detected in extracts of FS03 and FS03/pSF11. Relative amount of UvrA protein was estimated by scanning the image; the value for FS03/pSF11 relative to the FS03 was 20, probably reflecting the number of the uvrA genes.

K. Kiyosawa et al. / Mutation Research 487 (2001) 149–156

Fig. 4. Immunodetection of UvrA protein. Crude extract from FS03 (lane 1) and FS03/pSF11 (lane 2) were subjected to SDS–PAGE and proteins were transferred to PVDF membrane. The membrane was reacted with anti-UvrA antibody. An arrow indicates the signal for 104 kDa protein.

4. Discussion We observed increases in the rate of UV survival, HCR of UV-irradiated ␭ phage and CPD excision in the recA cells transformed with uvrA-plasmid but not uvrB-plasmid (Figs. 1–3). The simplest explanation is that increased NER activity due to 20-fold amplification of UvrA protein (Fig. 4) is responsible for these observations. This interpretation was strengthened by the finding that uvrA-plasmid increased the UV resistance in the lexA1 (Ind− ) strain (Fig. 1). Expression of uvrA and uvrB genes is increased by various SOS-inducing treatments [24,25], and the recA and lexA1 strains used in this study were defective in SOS-induction. Since recA as well as lexA1 strains are sensitive to UV radiation and amplification of UvrA protein can alleviate the UV sensitivities of these strains, constitutive levels of UvrA protein in these strains are not enough to sustain full repair. Finally, addition of UVDE-plasmid into the recA strain increased UV survival (Fig. 1) and fully reactivated UV-irradiated ␭ phage (HCR) (Fig. 2). N. crassa UVDE is a distinct excision repair system from NER [19]. Addition of UVDE-plasmid to recA strains can fully complement their NER defect (Fig. 1, [19]). These results agreed with those of Ganesan and Hanawalt [9] who used SOS-inducible lexA41 bacteria and observed that DNA damage produced by UV irradiation can be excised more efficiently in recA

155

lexA41 than recA lexA+ strains. However, it could not be determined whether expression of the uvrA gene and/or uvrB gene was involved in their observations. It is evident from our results that expression of the uvrA gene but not the uvrB gene is important for the NER in the recA cells. In recA cells, the number of UvrA protein molecules expressed constitutively per cell is about 20 [10] and that of UvrB is about 140 [11]. It was also estimated that uninduced wild type cells possessed approximately 200 UvrA and 400 UvrB molecules [26]. We further estimated the number of UvrA protein in FS03/pSF11 as about 400 (Fig. 4, i.e. 20 proteins in FS03 multiply 20-fold in FS03/pSF11). In the first step of the NER, UvrA protein dimerizes and forms an UvrA2 UvrB1 complex [4]. Thus, we estimated that about 10 active UvrA2 UvrB1 complexes are formed per recA cell. The number of active UvrA2 UvrB1 complexes remains 10 even after UV irradiation because the strain is recA and therefore cannot induce the uvrA gene. In other words, 130 unbound UvrB proteins exist per recA cell. Amplification of UvrB protein cannot result in any increase in number of active UvrA2 UvrB1 complexes, while amplification of UvrA protein can. In FS03/pSF11, about 400 UvrA proteins and 140 UvrB proteins can make 140 UvrA2 UvrB1 complexes. This may be one reason why amplification of UvrA protein but not UvrB protein ameliorates the UV sensitivity of recA strains. This interpretation is supported by our observation that D37 value of FS03 was 0.5 Jm−2 and that of FS03/pSF11 was 5 Jm−2 (Fig. 1). There is about 10-fold difference of D37 value between FS03 and FS03/pSF11 which is corresponding to the difference of estimated amount of active UvrA2 B1 complexes, these are 10 for FS03 and 140 for FS03/pSF11. Since CPD excision in recA strains is not as efficient as that in wild-type strains (Fig. 3), and these strains cannot fully perform HCR (Fig. 2) and are sensitive to UV radiation (Fig. 1), we concluded that the constitutive level of UvrA2 UvrB1 complex is not enough for efficient NER. Supplementation of UvrA protein in recA cells by amplification of UvrA protein using lexA41 mutant [9] or using a multicopy uvrA-plasmid (this study) can almost fully restore the NER. As a conclusion we argue that UvrA protein is limiting factor for NER in uninduced cells and induction of UvrA is required for full NER in E. coli.

156

K. Kiyosawa et al. / Mutation Research 487 (2001) 149–156

Acknowledgements We thank Drs. H. Ikeda and A. Yasui for plasmids, and Dr. O.I. Kovalsky for anti-UvrA monoclonal antibody. This research was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan. References [1] E.M. Witkin, Ultraviolet mutagenesis and inducible DNA repair in Escherichia coli, Bacteriol. Rev. 40 (1976) 869–907. [2] G.C. Walker, Inducible DNA repair systems, Annu. Rev. Biochem. 54 (1985) 425–457. [3] A. Sancar, G.B. Sancar, DNA repair enzymes, Annu. Rev. Biochem. 57 (1988) 29–67. [4] B. Van Houten, Nucleotide excision repair in Escherichia coli, Microbiol. Rev. 54 (1990) 18–51. [5] J.W. Little, S.H. Edmiston, L.Z. Pacelli, D.W. Mount, Cleavage of the Escherichia coli lexA protein by the recA protease, Proc. Natl. Acad. Sci. U.S.A. 77 (1980) 3225–3229. [6] D.W. Mount, A.C. Walker, C. Kosel, Suppression of lex mutations affecting deoxyribonucleic acid repair in Escherichia coli K-12 by closely linked thermolabile mutations, J. Bacteriol. 116 (1973) 950–956. [7] D.W. Mount, A.C. Walker, C. Kosel, Effect of tsl mutations in decreasing radiation sensitivity of a recA− strain of Escherichia coli K-12, J. Bacteriol. 121 (1975) 1203–1207. [8] A.K. Ganesan, P.C. Seawell, D.W. Mount, Effect of thermosensitive suppressor of lex (tsl) mutation on postreplication repair in Escherichia coli K-12, J. Bacteriol. 135 (1978) 935–942. [9] A.K. Ganesan, P.C. Hanawalt, Effect ofa lexA41(Ts) mutation on DNA repair in recA(Def) derivatives of Escherichia coli K-12, Mol. Gen. Genet. 201 (1985) 387–392. [10] A. Sancar, R.P. Wharton, S. Seltzer, B.M. Kacinski, N.D. Clarke, W.D. Rupp, Identification of the uvrA gene product, J. Mol. Biol. 148 (1981) 45–62. [11] A. Sancar, N.D. Clarke, J. Griswold, W.J. Kennedy, W.D. Rupp, Identification of the uvrB gene product, J. Mol. Biol. 148 (1981) 63–76. [12] B.R. Bochner, H.C. Huang, G.L. Schieven, B.N. Ames, Positive selection for loss of tetracycline resistance, J Bacteriol. 143 (1980) 926–933. [13] S. Akasaka, K. Yamamoto, Construction of Escherichia coli K-12 phr deletion and insertion mutants by gene replacement, Mutation Res. 254 (1991) 27–35. [14] S. Akasaka, K. Yamamoto, Hydrogen peroxide induces G:C to T:A and G:C to C:G transversions in the supFgene of Escherichia coli, Mol. Gen. Genet. 243 (1994) 500–505. [15] O.I. Kovalsky, L. Grossman, The use of monoclonal antibodies for studying intermediates in DNA repair by the

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24] [25] [26]

[27] [28]

[29]

[30]

Escherichia coli Uvr(A)BC endonuclease, J. Biol. Chem. 269 (1994) 27421–27426. T. Matsunaga, T. Mori, O. Nikaido, Base sequence specificity of a monoclonal antibody binding to (6-4) photoproducts, Mutation Res. 235 (1990) 187–194. M. Tanaka, S. Nakajima, M. Ihara, T. Matsunaga, O. Nikaido, K. Yamamoto, Effect of photoreactivation for cyclobutane pyrimidine dimers and pyrimidine (6-4) pyrimidone photoproducts on ultraviolet mutagenesis in SOS-induced Escherichia coli, Mutagenesis 16 (2001) 1–6. T. Mizuno, T. Matsunaga, M. Ihara, O. Nikaido, Establishment of a monoclonal antibody recognizing cyclobutane-type thymine dimers in DNA: a comparative study with 64M-1 antibody specific for (6-4) photoproducts, Mutation Res. 254 (1991) 175–184. H. Yajima, M. Takao, S. Yasuhira, J.H. Zhao, C. Ishii, H. Inoue, A. Yasui, A eukaryotic gene encoding an endonuclease that specifically repairs DNA damaged by ultraviolet light, EMBO J. 14 (1995) 2393–2399. J.M. Boyle, R.B. Setlow, Correlations between host-cell reactivation, ultraviolet reactivation and pyrimidine dimer excision in the DNA of bacteriophage lambda, J. Mol. Biol. 51 (1970) 131–144. H. Kneser, Relationship between K-reactivation and UV-reactivation of bacteriophage lambda, Virology 36 (1968) 303–305. H. Echols, R. Gingery, Mutants of bacteriophage ␭ defective in vegetative genetic recombination, J. Mol. Biol. 34 (1968) 239–249. K. Yamamoto, M. Satake, H. Shinagawa, A multicopy phr-plasmid increases the ultraviolet resistance of a recA strain of Escherichia coli, Mutation Res. 131 (1984) 11–18. M. Fogliano, P.F. Schendel, Evidence for the inducibility of the uvrB operon, Nature 289 (1981) 196–198. C.J. Kenyon, G.C. Walker, Expression of the E. coli uvrA gene is inducible, Nature 289 (1981) 808–810. D.J. Crowley, P.C. Hanawalt, Induction of the SOS response increases the efficiency of global nucleotide excision repair of cyclobutane pyrimidine dimers, but not 6-4 photoproducts, in UV-irradiated Escherichia coli, J. Bacteriol. 180 (1998) 3345–3352. B.J. Bachmann, Pedigrees of some mutant strains of Escherichia coli K-12, Bacteriol. Rev. 36 (1972) 525–557. S.G. Grant, J. Jessee, F.R. Bloom, D. Hanahan, Differential plasmid rescue from transgenic mouse DNAs into Escherichia coli methylation-restriction mutants, Proc. Natl. Acad. Sci. U.S.A. 87 (1990) 4645–4649. J.A. Brandsma, C.A. van Sluis, P. van de Putte, Use of transposons in cloning poorly selectable genes of Escherichia coli: cloning of uvrA and adjacent genes, J. Bacteriol. 147 (1981) 682–684. E. van den Berg, J. Zwetsloot, I. Noordermeer, H. Pannekoek, B. Dekker, R. Dijkema, H. van Ormondt, The structure and function of the regulatory elements of the Escherichia coli uvrB gene, Nucleic Acids Res. 9 (1981) 5623–5643.