Inhibition of dimer excision in repeatedly UV-irradiated Escherichia coli: Its requirement for RecA protein and de novo protein synthesis

Inhibition of dimer excision in repeatedly UV-irradiated Escherichia coli: Its requirement for RecA protein and de novo protein synthesis

205 J. Photochem. Photobiol. B: BioL, 18 (1993) 205-210 Inhibition of dimer excision in repeatedly UV-irradiated Escherichia coli: its requirement f...

633KB Sizes 0 Downloads 22 Views


J. Photochem. Photobiol. B: BioL, 18 (1993) 205-210

Inhibition of dimer excision in repeatedly UV-irradiated Escherichia coli: its requirement for RecA protein and de nova protein synthesis I. Fridrichovfi,

K. Kleibl,

F. MaSek


M. Sedliakov$

Cancer Research Institute, Slovak Academy of Sciences, Department of Molecular Genetics, Spitrilska 21, 812 32 Brattilava (Slovak Republic) (Received


28, 1992; accepted


15, 1993)

Abstract In UV-irradiated Escherichia coli dimer excision was found to be inhibited by predamage (M. SedliakovB, F. MaSek and J. BrozmanovB, FEBS Lett., 23 (1972) 325-326) or overproduction of RecA protein, which suggests that the coating of the dimers by this protein may make them inaccessible to the excision nuclease (M. SedliakovB, K. Kleibl and F. MaSek, Mutut. Res., 191 (1987) 13-16). We measured the levels of RecA protein and dimer excision in cells irradiated with (i) a single dose of 50 J m-‘; (ii) two separate doses of 30 and 50 J m-‘, postincubated with chloramphenicol; (iii) two separate doses of 30 and 50 J m-‘, post-incubated without chloramphenicol. Dimer excision was complete in the first two cases, but in the latter it was inhibited by 40%. At the time of active dimer excision, there were marked differences in RecA protein content between the cells irradiated with a single dose and cells irradiated with two separate doses (both post-incubated without chloramphenicol), which might account for the differences in dimer excision. However, relatively small differences in RecA protein content were found in cells irradiated with two doses and post-incubated with or without chloramphenicol, which could therefore not account for the differences in dimer excision. The data suggest that the inhibition of dimer excision involves some short-lived component(s) other than RecA protein.

Keywords: Escherichia

coli, uvrABC, RecA


UV irradiation,

1. Introduction In damaged Escherichiu coli cells numerous genes are induced and their products are involved in DNA repair (for a review, see refs. 1 and 2). Among them the uvrAB gene products are involved in the uvrABC excision nuclease complex which releases a dimer in the form of a fragment of 12-13 nucleotides [3, 41. The operators of damage-inducible (din) genes have unequal affinities for the LexA repressor by which they are turned off and on gradually [l, 51. This may cause the repair processes to take place in a specific sequence. In repeatedly injured cells (if the second damage is inflicted before the response to the first damage is switched off), this fine tuning of induction of din genes may be impaired. It has been shown that, in UV-irradiated cells predamaged by thymine starvation, dimer excision is markedly inhibited [6-81. Since unexcised dimers ‘Author

to whom correspondence


should be addressed.

Es sites, Dimer


have been found to be less harmful for cells with intact uvr genes than for uvr-defective mutants, it has been suggested that, in addition to excision, UVUT genes are involved in another repair pathway [S-11]. This putative pathway is non-excisional, but is dependent on uvr as well as recA ZexA genes and de nova protein synthesis. We call it UVTdependent alternative repair. It has recently been shown that dimer excision is partially inhibited in cells which have been transformed with a multicopy plasmid carrying the recA gene. In such cells the amount of RecA protein is selectively increased, since its high level is independent of induction [12,13]. Because RecA protein binds to UV-irradiated DNA and coats pyrimidine dimers [14-161, it has been concluded that, in such cells, dimers are masked with RecA protein before excision can take place, which makes them inaccessible to the uvrABC excision nuclease and inhibits excision repair [12]. We have measured the level of RecA protein in cells irradiated with a single dose and in cells

0 1993 - Elsevier Sequoia. All rights reserved


I. Fridrichovd et al. / Dimer excision inhibition in UV-irradiated E. coli

irradiated repeatedly to determine the possible relationship between RecA protein content and dimer excision inhibition.

2.3. Determination of pyrimidine dimers Pyrimidine dimers were estimated by two-dimensional paper radiochromatography as described in ref. 17.

2. Materials and methods

2.4. Determination of endonuclease-sensitive sites The endonuclease-sensitive site (Es site) assay [18] measures the pyrimidine dimer content sensitive to exogenous UV-endonuclease. A sample of DNA, treated with the extract from Micrococcus luteus, was layered onto an alkaline sucrose (5%-20%) gradient and sedimented as described previously [19].

2.1. Bacterial strains and growth conditions Escherichia coli D H 5 a cells, transformed with pBEU14 carrying the recA gene (the plasmid was kindly provided by Dr. Uhlin), were cultured in LB medium (bacto-tryptone 10 g, bacto-yeast extract 5 g, NaC1 10 g per liter) and served as a source of RecA protein. Escherichia coli B/r Hcr + thy- trp- strain, used for experiments, was cultured on a reciprocal shaker at 37 °C in a glucose-salt medium (containing Na2HPO4.12H20, 0.6 g, K H 2 P O 4 , 3 g; N a z S O 4 , 0.115 g; NaC1, 3 g; NHaCI, 2 g; MgClz.6H20, 0.083 g; glucose, 10 g; in 11), supplemented with thymine (2/zg ml-1) and tryptophan (14/zg ml-1). For DNA labelling thymine2-14C ( 0 . 5 /.t,C i m1-1) was added and cells were grown up to a density of 1 x 10s cells ml-1. They were then transferred into cold medium without tryptophan and treated as indicated in Table 1. After irradiation with 50 J m -2 UV, thymidine (4/zg ml-1) was added instead of thymine. After UV treatment, chloramphenicol was added at a final concentration of 40 /xg m1-1. 2.2. UV irradiation The cell suspension (density, 1 X 108 cells ml-1) was irradiated in a medium without thymine and tryptophan with a Philips TUV 15 W germicidal lamp emitting light of wavelength 253.7 nm at an incident exposure rate of 0.63 J m - 2 s- 1. Irradiated cells were kept in a sodium light environment to prevent photoreactivation.

2.5. Determination of surviving fraction Surviving fractions were determined as indicated previously [9]. Z6. Detection of RecA protein by enzyme-linked immunosorbent (ELISA ) The details of the technique used have been described previously [20]. Briefly, lysates of E. coli DH5odpBEU14 (overproducing RecA protein) were prepared as described in ref. 21; RecA protein isolated by three-step purification [22] was then purified by Western blotting. The corresponding region of nitrocellulose membrane was mechanically pulverized and the water suspension of a fine powder was injected (i.m. and s.c.) into a New Zealand rabbit. The total immunization dose (2.5 mg) of purified RecA protein was applied in decreasing portions of 500-100/xg monthly during a period of 11 months. For a rapid increase in the titre of polyclonal RecA-specific antibodies, the last four portions contained solubile antigen in Freund incomplete bacto adjuvant (Difco). The RecA-specific antibodies (serum diluted 1:120 000) together with the purified RecA protein

TABLE 1. Schedule of experiments and cell survival data Conditions


U V dose (J m -2)

AA + (min)

U V dose (J m -2)

Post-UV treatment

Cell survival

Symbols in Fig. 1

30+50 J m -z, C a m 30+50 J m -2, Cam + 50 J m -2, C a m -

90 90 90

30 30 -

60 60 60

50 50 50

CamCam + Cam-

4 x 10 -2 4 X 1 0 -2

O • -.-

2 × 10 -2

Twice irradiated cells were treated as follows. In the exponential phase of growth they were incubated for 9Q mi n wiLhout tryptophan ( A A - ) to enhance their U V resistance; they were then irradiated with the first dose (30 J m -2) which was non-lethal (induction fluence, IF). Cells were then incubated for 60 min in tryptophan-containing medium (AA +) to achieve protein synthesis and complete dimer excision. Subsequently, they were irradiated with a second dose (50 J m - z ) (lethal fluence, LF) and incubated for 120 rain with or without chloramphenicol, (Cam). Cells irradiated with a single dose were treated in the same way, but irradiated only with a dose of 50 J m -2 and then incubated for 120 min without Cam. Before irradiation with 50 J m -2 the density of each cell culture was adjusted to 1 x l0 s cells ml -~. Samples were taken as follows: for estimation of RecA protein and paper radiochromatography, during post-incubation with chloramphenicol; for endonuclease-sensitive (Es) site assay, after post-incubation with or without chloramphenicol; for determination of cell survival, immediately after irradiation with 50 J m -2 ( C a m - ) and after 120 min incubation with chloramphenicol (Cam+).

I. Fridrichovri et al. I Dimer excision inhibition in UV-irradiated E. coli

were used for the detection of RecA protein levels of UV-irradiated cells by a competitive enzymelinked immunosorbent assay (ELISA). This and the mathematical processing of the data obtained were performed as described in ref. 23.


incubation with chloramphenicol. The data on the stability of RecA protein and its relatively slow increase are in agreement with those reported previously [24, 251. 3.2. Excision of dimers

3. Results To determine the possible relationship between the inhibition of dimer excision and the levels of RecA protein, the cells were irradiated either with two separate doses (30 J m-‘) and after 1 h by 50 J me2 (IF+LF) and then incubated with or without chloramphenicol (Cam+, Cam-) or were irradiated with a single dose (50 J m-“) and then incubated without chloramphenicol (LF, Cam-). In all cases the levels of RecA protein, dimer excision and cell survival were measured (for schedule of experiments see Table 1).

3.1. Level of RecA protein As shown in Table 2, a single dose of 50 J mm2 (LF, Cam-) stimulated a relatively slow enhancement of RecA protein during the period of 10-20 min after irradiation (i.e. during the time of most active dimer excision). The level of RecA protein was increased only about 12-fold when measured 30 min after irradiation. In contrast, the initial level was increased 17-fold in repeatedly irradiated cells (IF+LF, Cam-) and was doubled after 30 min. Thus the two types of cells markedly differed in their RecA protein contents during the period of most active dimer excision. However, there were relatively small differences between repeatedly irradiated cells post-incubated with or without chloramphenicol (IF + LF, Cam+; IF + LF, Cam-). This was due to the stability of RecA protein which did not decrease but slightly increased during the

The fate of dimers was measured by paper radiochromatography and by detection of sites sensitive to the endonuclease from Micrococcus Zuteus (Es sites). Dimer excision followed by paper radiochromatography is illustrated in the inset of Fig. 1. As shown, it was incomplete when the repeatedly irradiated cells were post-incubated without chloramphenicol (IF + LF, Cam-); however, it was complete when protein synthesis was inhibited (IF + LF, Cam +). Data obtained for the detection of Es sites are presented in Fig. 1. They indicate that the DNA


TABLE 2. RecA protein level in UV-irradiated cells expressed as amplification factors Cells

Time after UV treatment (min) 0

IF+LF (Cam-) IF + LF (Cam+) LF (Cam-)

17 17 1







33 19 12

34 23 25



The amount of RecA protein in unirradiated cells was 0.728 f 0.069 fig mgg’ of the whole protein and 0.712*0.077 pg mg-’ of the whole protein in exponentially growing cells and cells starved for 90 min without tryptophan (AA-) respectively. The amplification factors were calculated as (RecA in irradiated cells)/ (RecA in unirradiated cells) [25]. For conditions, see Table 1. The values are averages of at least four experiments.



20 Fraction number



Fig. 1. Endonuclease-sensitive sites. Exponentially growing cells, prelabelled with “‘C-thymine, were treated and W irradiated as indicated in Table 1. Samples were taken 2 h after irradiation with 50 J m-* and treated with the extract from M. luteus; DNA was analysed in alkaline sucrose gradients. The arrow indicates sedimentation of DNA untreated with W-endonuclease. Inset. Pyrimidine dimer content. At the given time after irradiation with 50 J m-*, samples were taken and pyrimidine dimer contents (% PyPy) were determined radiochromatographically: - . -, 50 J m-*, Cam-; 0, 30+50 J m-* Cam-; l ,30+50 J m-*, Cam+.


I. Fridrichova’ et al. I Dimer excision inhibition in UV-irradiated E. coli

of repeatedly irradiated cells incubated for 2 h without chloramphenicol (IF + LF, Cam-) consisted of two fractions, one containing Es sites and the other not, whereas the DNA of repeatedly irradiated cells incubated for 2 h with chloramphenicol (IF+ LF, Cam+) was free of Es sites. Earlier we provided evidence that the DNA free of Es sites corresponds to the fraction which replicates after the second UV dose, whereas the DNA with Es sites corresponds to the fraction which remains unreplicated [19]. Data indicating that dimers once replicated are not distinguished by dimer-specific endonucleases from Micrococcus Zuteusor T4 bacteriophage have also been obtained by others [26, 271. Thus the structure of dimers (or the structure of DNA regions containing dimers) should be modified on replication such that it is not distinguished by the specific endonucleases above. Cell survival data are given in Table 1. The activity of dimer excision does not affect cell survival. 4. Discussion The observation that an increased recA gene dosage inhibits, to some degree, excision of dimers in all E. coli strains investigated so far, namely K12AB2497, B/r Her+, 15 555-7 [28], suggests that RecA protein is involved in the inhibition of dimer excision observed in predamaged cells [8, 9, 11, 191. However, the degree of inhibition in UVirradiated cells, transformed with a multicopy plasmid containing the red gene, has been found to be relatively low, although the levels of RecA protein were enhanced several tens of times [28]. The pretreatment of cells without thymine and amino acids inhibits dimer excision almost completely [8, 301, while RecA protein levels are increased by only about five times [31]. This indicates that the inhibition of dimer excision is not proportional to the intracellular levels of RecA protein. The data presented here show that in repeatedly irradiated cells dimer excision is inhibited by about 50% if the cells are post-incubated without chloramphenicol, but no inhibition is observed when the cells are post-incubated with chloramphenicol (Fig. 1). Since the levels of RecA protein are similar in both cases, this indicates that other factors (besides RecA protein) are important for the inhibition of excision repair. In vitro studies suggest that these may be nucleotide cofactors of RecA protein which influence its state of binding

affinity to DNA [32, 331, as well as conditions influencing the state of RecA self-assembly [34-361. Recent studies have indicated that UmuC protein also contributes to the inhibition of dimer excision [29, 301. UmuC is a damage-inducible protein and its level is undetectable in intact cells [37]. If this protein is unstable, concomitant de no210protein synthesis may be necessary to maintain it at the level needed for the inhibition of dimer excision. In agreement with previous results, the data presented here seem to indicate that it is irrelevant for cell survival whether dimers are excised or left in situ (Table 1, Fig. 1). The putative mode of repair of unexcised dimers is highly speculative. Data indicate that it is efficient, error free [ll] and uvr dependent [8, 381. Such a repair may be secured by the interaction of two sister duplexes, mediated by RecA (and further cooperative proteins) and involving a non-excisional step produced by the uvr system. A uvr-dependent, recombinational system has already been described by Howard-Flanders and coworkers [39-41], who observed cutting of intact DNA strands homologous to strands containing cross-links. It has been postulated that this uvr- and red-dependent step (called “cutting in trans”) may provide 3’ ends of intact DNA needed for an efficient recombinational repair of the DNA containing cross-links. By analogy, cutting in trans may be needed for an efficient recombinational repair of dimers masked by RecA protein. Another mode of repair secured by interaction of sister duplexes may be replicational bypass of lesions. This putative mode suggests that the DNA polymerase, whose motion is interrupted by a lesion, should replicate the daughter strand of a sister duplex instead of the parental template containing the lesion [42-44]. Cutting in trans may be needed for switching off the newly synthesized strand from the template temporarily provided by the sister duplex. The putative mode of toleration of lesions is rather insignificant for cells which, at the time of injury, have a basal level of inducible proteins, since in such cells dimers are almost completely excised. It may be important, however, for the cells whose fine tuning of the SOS response is impaired (see Section 1) and which, at the time of injury, contain high levels of RecA protein (as well as component(s) requiring concomitant protein synthesis) that may mask single strands containing dimers before the dimers are excised.

I. Fridrichovd et al. / Dimer excision inhibition in W-irradiated


1 J. W. Little and D. W. Mount,


















The SOS regulatory system of Escherichia coli, Cell, 29 (1982) 11-22. G. C. Walker, Mutagenesis and inducible responses to deoxyribonucleic acid damage in Escherichia co& Microbial. Rev., 48 (1984) 60-93. A. Sancar and W. D. Rupp, A novel repair enzyme: UvrABC excision nuclease of Escherichia coli cuts a DNA strand on both sides of the damaged region, Cell, 33 (1983) 249-260. E. Seeberg and A. L. Steinum, Properties of the uvrABC endonuclease from E. coli, in E. C. Friedberg and B. A. Bridges (eds.), Cellular Responses to DNA Damage, A. R. Liss, New York, 1983, pp. 39-l9. R. Brent and M. Ptashne, Mechanism of action of the levi gene product, Proc. NatL Acad. Sci. USA, 28 (1981) 4204-4208. M. Sedliakova, F. MaSek and J. Brozmanova, Thymine dimer excision after the preirradiation inhibition of DNA synthesis, FEBS Lett., 23 (1972) 325-326. J. Brozmanova, L. MaSkova and M. Sedliakova, The inhibition of thymine dimer excision after the preirradiation inhibition of DNA synthesis by cytidine, Stud. Biophys., 36/37 (1973) 245-252. V. Slezarikova and M. Sedliakova, uvrB-dependent, recFindependent post-replication (or replication) repair in Escherichia co& 1. Photochem. Photobiol. B: Biol., 10 (1991) 329-337. M. Sedliakova, V. Slezarikova, F. MaSek and J. Brozmanova, UV-inducible repair: influence on survival, dimer excision, DNA replication and breakdown in Escherichia coli B/r Her+ cells, Mol. Gen. Genet., 160 (1978a) 81-87. M. Sedliakova, V. Slezlrikova and M. PirSel, UV-inducible repair II. Its role in various defective mutants of E. coli K12, Mol. Gen. Genet., 167 (1978b) 209-215. M. PirSel, M. Bencovi, F. MaSek and M. Sedliakovb, Errorfree uvr+-dependent inducible DNA repair in E. coli B/r Her+ cells, Int. J. Radiat. Biol., 45 (1984) 389-397. M. Sedliakova, K. Kleibl and F. MaSek, Inhibition of pyrimidine dimer excision in ultraviolet-irradiated Escherichia colt’ overproducing RecA protein, Mutat. Res., 191 (1987) 13-16. M. Sedliakova, F. MaSek and K. Kleibl, In UV-irradiated Escherichia coli PQ35 overproducing the RecAprotein, expression of the sfiA gene and dimer excision are alleviated, Mol. Gen. Genet., 217 (1989) 427-429. L. J. Gudas and A. B. Pardee, DNA synthesis inhibition and the induction of protein X in Escherkhia coli, J. Mol. Biol., 101 (1976) 459-477. T. Shibata, R. P. Cunningham, C. Dasgupta and C. M. Radding, Homologous pairing in genetic recombination: complexes of RecA protein and DNA, Proc. Natl. Acad. Sci. USA, 76 (1979) 5100-5104. C. Lu, R. H. Schuermann and H. Echols, Capacity of RecA protein to bind preferentially to UV lesions and inhibit the editing subunit of DNA polymerase III: a possible mechanism for SOS-induced targeted mutagenesis, Proc. Natl. Acad. Sci. USA, 83 (1986) 619-623. W. L. Carrier and R. B. Setlow, The excision of pyrimidine dimers, in L. Grossman and K. Moldave (eds.), Methoak in Enzymology, Vol. 21, Academic Press, New York, 1971, pp. 230-237. R. J. Wilkins, Endonuclease-sensitive sites in the DNA of irradiated bacteria: a rapid and sensitive assay, B&him. Biophys. Acta, 312 (1973) 33-37.

E. coli


19 M. Sedliakova, J. Brozmanova, F. MaSek and K. Kleibl, Evidence that dimers remaining in preinduced Escherichia coli B/r Her+ become insensitive after DNA replication to the extract from Micrococcus luteus, Biophys. J., 36 (1981) 429-441. 20 I. Fridrichova, A. Kovaiik and 0. Rosskopfovl, Immunological quantification of the RecA protein in E. coli cell extracts after treatment with different chemical mutagens or UV radiation, Folia Microbial., 37 (1992) 24-30. 21 M. Cox, K. McEntee and I. R. Lehman, A simple and rapid procedure for the large scale purification of the RecA protein of Escherichia coli, J. Biol. Chem., 256 (1981) 467w678. 22 T. Shibata, R. P. Cunningham and C. M. Radding, Homologous pairing in genetic recombination, purification and characterization of Escherichia coli RecA protein, J. Biol. Chem., 256 (1981) 7557-7564. 23 A. E. Kant and E. D. Belk, Induction of E. coli RecA protein via recBC and alternate pathways: quantitation by enzymelinked immunosorbent assay (ELISA), Mol. Gen. Genet., 185 (1982) 275-282. 24 A. Pierre, B. Salles and C. Paoletti, Measurement of RecA protein induction in Salmonella typhymurium: a possible biochemical test for the detection of DNA damaging agents, Biochimie, 64 (1982) 775-781. 25 B. Salles and C. Paoletti, Control of W induction of RecA protein, Proc. Natl. Acad. Sci. USA, 80 (1983) 65-69. 26 J. M. Vos and J. Rommelaere, Are pyrimidine dimers tolerated during DNA replication of UV-irradiated parvovirus MinuteVirus-of-Mice in mouse fibroblasts?, Biochimie, 64 (1982) 839-844. 21 Z. Livneh, Replication of UV-irradiated single-stranded DNA by DNA polymerase III holoenzyme of Escherichia coli: evidence for bypass of pyrimidine photodimers, Proc. Natl. Acad. Sci. USA, 83 (1986) 4599-4603. 28 M. Sedliakova, in preparation. 29 V. SledrikovB, M. Sedliakova, I. V. Andreeva, U. Yu. Rusina and A. G. Skavronskaga, Effect of plasmid pKMlO1 in ultraviolet irradiated uvr+ and uvr- Escherichia colt, Mutation Res., 270 (1992) 145-149. 30 F. MaSek and M. Sedliakova, UmuC product contributes to the inhibition of dimer excision produced by thyminelessaminoacidless pretreatment in UV irradiated Escherichia coli, J. Photochem. Photobiol. B: Biol., 17 (1993) 57-61. 31 I. Fridrichovi in preparation. 32 J. P. Menetski, A. Varghese and S. C. Kowalczykowski, Properties of the high-affinity single-stranded DNA binding state of the Escherichia coli RecA protein, Biochemistry 27 (1988) 1205-1212. 33 J. P. Menetski and S. C. Kowalczykowski, Enhancement of Escherichia coli RecA protein enzymatic function by dATP, Biochemistry, 28 (1989) 5871-5881. 34 S. L. Brenner, A. Zlotnick and J. D. Griffith, RecA protein self-assembly multiple discrete aggregation states,J. Mol. Biol., 204 (1988) 959-972. 35 R. W. H. Ruigrok and E. DiCapua, On the polymerization state of recA in the absence of DNA, Biochemie, 73 (1991) 191-197. 36 R. M. Story, I. T. Weber and T. A. Steitz, The structure of the E. coli RecA protein monomer and polymer, Nature, 355 (1992) 318-325. 37 R. Woodgate and D. G. Ennis, Levels of chromosomally encoded Umu protein and requirements for in vitro UmuD cleavage, Mol. Gen. Genet., 229 (1991) 10-16.


I. Fridrikhovn’ et al. / Dimer srcision inhibition in W-irradiated

38 M. Sedliakovl, J. Brozmanova, F. MaGek and V. Slezarikovl, Interaction of restoration processes in UV irradiated Escherichia coli cells, Phofochem. PhotobioL, 25 (1977) 259-264. 39 P. Ross and P. Howard-Flanders, Initiation of recA+-dependent recombination in Escherichia coli I. Undamaged covalent circular lambda DNA molecules in uvrA+ recA+ lysogenic host cells are cut following superinfection with psoralendamaged phages, J. Mol. Biol., 117 (1977) 137-158. 40 E. Cassuto, J. Mursalim and P. Howard-Flanders, Homology dependent cutting in trans of DNA in extracts of Escherichia coli: an approach to the enzymology of genetic recombination, Froc. Natl. Acad. Sci. USA, 75 (1978) 620-624.

E. coli

41 P. Ross and P. Howard-Flanders, Effect of lig-7 on strand joining in repair of damaged DNA and on cutting of intact homologous DNA (cutting in trans) in Escherichia coli, J. Mol. BioZ., 144 (1980) 117-131. 42 H. Echols, Mutation rate: some biological and biochemical considerations, Biochimie, 64 (1982) 571-575. 43 H. Echols and M. Goodman, Mutation induced by DNA damage: a many protein affair, Mutat. Rex, 236 (1990) 301-311. 44 R. Woodgate, M. Rajagopalan, C. Lu and H. Echols, UmuC mutagenesis protein of Escherichia colk purification and interaction with UmuD and UmuD’, Proc. Natl. Acad. Sci. USA, 86 (1989) 7301-7305.