Proc. Nati Acad. Sci. USA Vol. 79, pp. 988-992, February 1982 Biochemistry

Purification and properties of the uvrA protein from Escherichia coli (UV excision repair/ATPase/DNA-binding protein) ERLING SEEBERG AND ANNE-LILL STEINUM Norwegian Defence Research Establishment, Division for Toxicology, P. 0. Box 25, N-2007 Kjeller, Norway Communicated by Richard B. Setlow, October 16, 1981

ABSTRACT The uvrA' gene product from Eacherichia coli It is shown that the uvrA' product is a DNA-independent was purified to apparent homogeneity; the assaymeasured its abil- ATPase that hydrolyzes ATP to ADP and phosphate. ity to restore repair activity in extracts from uvrA mutated cells. The uvrA protein is a 115,000 molecular weight DNA-binding protein having higher affinity for single-stranded MATERIALS AND METHODS than double-stranded DNA. It does not introduce single-strand breaks or alkali-labile bonds in native or UV-irradiated DNA, but Bacterial Strains. Strain DR1240 contains both plasmid it catalyzes hydrolysis ofATP to ADP and Pi. The ATPase activity pDR2000 and plasmid pdR1472, carrying uvrA' and uvrB+, is not DNA dependent and has a K. of 0.23 mM, which corre- respectively (14). This strain overproduces the uvrA protein sponds to the K. for the ATP requirement ofthe UV-endonuclease about 7-fold and was generously provided by A. Sancar and W. reaction catalyzed by the combined uvrA', uvrB+, and uvrC' D. Rupp ofYale University. Strain AB1885 is uvrB- and strain gene products. ADP and adenosine 5'-[y-thio]triphosphate both AB1886 is uvrA-. inhibit the uvrA ATPase as well as the uvrABC endonuclease and Purification Procedure. Cells ofstrain pDR1240 were grown also prevent specific binding ofthe uvrA protein to UV-irradiated in K medium (4) in 10-liter batches to a density of5 x 10 cells DNA. These results indicate that both the DNA-binding property per ml. Protein extracts (fraction 1) were prepared by a com- and the ATPase activity of the uvrA protein are essential for bination of sucrose plasmolysis and lysozyme treatment as de- uvrABC endonuclease activity and that the ATP requirement of scribed (11). The extract was applied directly to a DEAE-cel- the endonuclease reaction is determined by the uvrA ATPase. lulose column (1.5 X 16 cm) equilibrated with 0.1 M KCl in buffer A [50 mM 4-morpholinepropanesulfonic acid (Mops), pH Living organisms have developed several that specif- 7.5/1 mM EDTA/10 mM 2-mercaptoethanol]. The flow- ically recognize and initiate repair of damaged bases in DNA. through (fraction II) was collected and applied to a phospho- Some repair enzymes are highly specific and will act only at a cellulose column (1.5 x 16 cm) also equilibrated with 0.1 M KCl single type ofbase lesion. This seems characteristic ofthe DNA in buffer A. The column was washed with 120 ml of0.1 M KCl glycosylases, which cleave the glycosylic bond between the base in buffer B [buffer A containing 25% (vol/vol) glycerol] and and sugar moieties of the nucleotides and release the damaged eluted with a linear salt gradient of0.1-0.4 M KCl in buffer B. base in a free form (1). Other enzymes or complexes Fractions containing uvrA complementing activity eluting at have a broader substrate specificity and appear to represent a about 0.2 M KCl were pooled (fraction III) and applied to a general mechanism protecting against the lethal and mutagenic DNA-cellulose column [1 X 18 cm, made with heat-denatured effects ofagents causing major distortions in the DNA helix (2). calf thymus DNA as described by Alberts and Herrick (15)] Enzymes of this type are not well characterized in vitro, and equilibrated with 0.1 M KCl in buffer B. The column was their existence is indicated mainly from in vivo genetic studies, washed with 0.2 M KC1 in buffer B and eluted with 0.3 M KCl which show that bulky lesions are generally repaired by a com- in buffer B. The pool of fractions containing uvrA comple- mon pathway whose initial step is multigenic. menting activity (fraction IV) was concentrated 5-fold by ultra- The best-known example of a broad substrate specificity re- filtration (Amicon UM20 filter) and applied in 2 ml-portions to pair pathway is that controlled by the uvrA, uvrB, and uvrC an Ultrogel AcA 34 gel column (1.5 x 22 cm) equilibrated with genes in Escherichia coli (3-6). Mutants defective in any one 0.3 M KC1 in buffer B. of these genes are very sensitive to UV and certain chemical Assay of DNA Binding. DNA binding of the uvrA protein agents and cannot remove UV-induced pyrimidine dimers or was assayed by a slight modification ofthe procedure previously various chemically induced lesions from their DNA (7-10). Re- described (16). Binding mixtures contained 5 ,umol ofMops (pH cently, the gene products of uvrA', uvrB+, and uvrC+ were 7.5), 10 ,umol of KCl, 0.2 ,umol ofATP, 2 ,umol of MgSO4, 0.1 partially purified and shown to interact to produce strand cleav- ,mol of EDTA, 0.1 ,umol of dithiothreitol, 0.05 ,ug of ColEl age at pyrimidine dimers, base adducts, and interstrand cross- [3H]DNA (800 cpm), and uvrA protein as indicated in a total links in DNA (11, 12). The purification of the uvr+ gene prod- volume of120 ,ul. The mixtures were incubated for 5 min at 37°C ucts was monitored by means of an in vitro complementation and subsequently for 15 min at 0°C before addition of 2.5 ml assay (13), and the endonuclease was reconstituted from the ofice-cold dilution buffer (50 mM Mops, pH 7.5/100 mM KCl/ separated components. To further characterize the strand break 10 mM MgSOJl mM EDTA). The diluted samples were fil- reaction promoted by the concerted action of the uvr' gene tered slowly through nitrocellulose filters (Millipore HAWP) products, we have pursued the purification and characterization and the filters were washed twice with cold dilution buffer. The of the separated components. This communication describes a fraction of total radioactivity (DNA) retained on the filter was simple protocol for extensive purification of the uvrA protein. plotted as DNA-binding activity.

The publication costs ofthis article were defrayed in part by page charge Abbreviations: Mops, 4-morpholinepropanesulfonic acid; ATP[-yS], payment. This article must therefore be hereby marked "advertise- adenosine 5'-[y-thio]triphosphate; uvrABC endonuclease, endonucle- ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact. ase activity of the concerted uvrA , uvrB+, and uvrC+ gene products. 988 Downloaded by guest on September 24, 2021 Biochemistry: Seeberg and Steinum Proc. Nati Acad. Sci. USA 79 (1982) 989

Assay of ATPase Activity. Standard reaction mixtures con- tained 1 Amol of Mops (pH 7.5), 4 Amol of KCl, 0.2 ,umol of MgSO4, 5 nmol of[3H]ATP (specific activity 0.8 Ci/mmol; 1 Ci = 3.7 X 1010 becquerels), 20 nmol of EDTA, 20 nmol of di- thiothreitol, and uvrA protein as indicated in a total volume of 20 A1. Samples were incubated for 30 min at 370C and subjected to polyethyleneimine thin-layer chromatography for separation of ATP, ADP, and AMP. The chromatograms were developed with 1 M HCOOH/0.5 M LiCl, and spots with ATP, ADP, and AMP were identified as UV-absorbing material by including nonradioactive compounds as internal markers in each sample. The amount ofradioactivity in each spot was calculated relative to the total on the chromatogram. Other Procedures. ColE 1 DNA was relaxed with unwinding protein (topoisomerase I) from mouse cells generously provided by I. Naes (Biochemistry Department, Bergen, Norway). The reaction mixtures contained 5 ,Amol of Tris-HCl (pH' 8.0), 100 /Lmol ofKCl, 5 pug ofDNA, and 10 A.l oftopoisomerase in a total volume of500 A.l and was incubated for 30 min at 370C. Linear DNA was made by incubating covalently closed circular ColE 1 DNA with EcoRI from Boehringer according to the supplier's instructions. The supercoiled, relaxed, and linear forms of ColE1 DNA were identified after electrophoresis in 0;8% agar- ose and staining with ethidium bromide at 5 Ag/ml. Single- DNA was by heating EcoRI-cleaved ColEl stranded prepared FIG. 1. Electrophoresis of uvrA' protein fractions in 0.1% at rapidly. DNA for 5 min 100'C and then cooling NaDodSOS/1O% polyacrylamide. The gels were stained with 0.2% Conditions for the in vitro complementation assay ofthe uvrA Coomassie blue. Tracks (numbered from the left): 1 and 7, marker pro- protein, partial purification ofuvrB' and uvrC' gene products, teins (from the top) myosin (Mr 200,000)t,&~galactosidase (Mr 116,000), and endonuclease reaction with mixtures ofpurified uvrA pro- P-phosphorylase. (Mr 94,000), bovine serum albumin (Mr 67,000), and tein and partially purified uvrB+/uvrC' gene products were ovalbuimin (Mr 46,000); 2, crude extract; 3, DEAE-cellulose fraction; 4, phosphocellulose fraction; 5, DNA-cellulose fraction, 10 ug of pro- as described (11). tein; 6, gel filtration fraction, -5 pg of protein. See Table 1 and the text for of fractions. RESULTS description

Purification of the uvrA Protein. By assaying for the ability The uvrA' Product Is a DNA.;Binding Protein with High to restore UV endonuclease activity in extracts from uvrA mu- Affinity for Single-Stranded DNA. -It was previously shown tant cells we purified a protein of Mr 115,000 from a strain that the uvrA' gene product is associated with a DNA-binding carrying the uvrA' gene on a multicopy plasmid. At each stage activity with higher affinity for UV-irradiated than nonirra- in the purification the active fractions (Table 1) were examined diated DNA (16). Fig. 2 shows that the DNA-binding activity by electrophoresis in NaDodSOjpolyacrylamide gels (Fig. 1). cochromatographed with the uvrA protein to its final stage of After DNA-cellulose chromatography more than 90% ofthe re- purification and thus is an intrinsic property of the uvrA protein maining protein was in the Mr 115,000 band as judged from itself. UV irradiation of the DNA increases binding of uvrA densitometer tracings of the stained gel. protein to supercoiled, relaxed, and linear double-stranded A minor contaminant of Mr 20,000 was sometimes present DNA (Fig. 3). However, binding to single-stranded DNA is.not in the DNA-cellulose fraction, but this was removed by gel fil- affected by UV and is similar to that observed with UV-irra- tration. The molecular weight obtained for uvrA complement- diated double-stranded DNA. This indicates that the uvrA pro- ing activity in gel filtration experiments corresponds to that tein does not recognize damaged bases per se but rather binds observed for the remaining purified protein by electrophoresis. to regions in DNA where base pairing has been impaired-for We conclude that the Mr 115,000 protein is the product of the instance, to sites where pyrimidine- dimers have been intro- uvrA' gene. This agrees with the radiochemical identification duced. Nevertheless, the uvrA protein will also bind to pure made by Sancar et al (18) by specific labeling of plasmid coded double-stranded DNA as judged from the binding observed to proteins in maxi-cells. unirradiated relaxed circular DNA. The slightly higher binding

Table 1. Purification of the uvrA protein from E. coli Specific activity, Purifi- Volume, Protein, units/mg cation, Recovery, Fraction ml mg protein fold % I Extract 300 840 200 (1) (100) II DEAE-cellulose 310. 620 380 1.9 140 m Phosphocellulose 43 6.5 25,000 125 96 IV DNA-cellulose 22 0.22 220,000 1100 29 V Gel filtration 50 - - - - One-unit of complementation activity is defined as the amount of'protein required to cause 0.4 break permolecule of complementation activity in 0.05 pg of DNA exposed to 180J/m2 of UV. Amountof protein was determined by the method of Bradford (17). Downloaded by guest on September 24, 2021 990 Biochemistry: Seeberg and Steinum Proc. Nad Acad. Sci. USA 79 (1982)

1..5 2.0- +AB1856 extract *1.0 -- 0

o X 1.5 - U A 1..0 TP ± DNA be 0 E b. S 0 "6 M;0K1.0 - -0.5 . 0 1-0E _ cd '-, *15 e g3 5/- z +AB1885 extract GTP X

__ mm 0- 10 . 0 o-I * X X _X_ . 0 10 20 .30 0 100 200 20 5 z uvrA protein, ng "_ 15* *n ^OA FIG. 4. ATP hydrolysis by uvrA protein. (Left) Various amounts XE 110 uv p40. of uvrA protein were tested for uvrA complementing activity (i) and C:~ 5,- uvrB complementing activity (W) by using extracts of deficient strains. (Right) The uvrA protein was tested for ATPase activity in the absence 0 of DNA (o) or in the presence of UV-irradiated DNA (0.05 jig, 180 J/ 15 18 21 24 27 30 in2) (a) or single-stranded DNA (0.05 j.g, EcoRI-cleaved and boiled) (A). GTPase activity (x) was assayed in the absence of DNA. Eluate, ml FIG. 2. Gel filtration of uvrA protein. Ten micrograms of protein 5'-[,f, y-imido]triphosphate that normally are nothydrolyzed by of fraction IV was applied to an Ultrogel AcA 34 (LKB) column (1.5 ATPases cannot substitute for ATP in the reaction mixture (20), x 22 cm), and eluting' fractions were tested for uvrA complementing suggesting that ATP hydrolysis may be required for the endo- activity (o), DNA-binding activity (o), and ATPase activity (A). Stan- Fig. 2 shows that the uvrA protein, in ad- dard assay conditions were employed except in the ATPase assay, in reaction. which only 0.1 nmol of ATP was used in the reaction mixtures. DNA dition to its DNA-binding property, also has ATPase activity polymerase I (Boehringer; Mr 109,000) was applied in a separate ex- which could be responsible for such hydrolysis. The ATPase periment and assayed according to the supplier's instructions. Ultrogel activity is not increased by adding UV-irradiated DNA to the AcA 34 separates protein molecules in the range of 20,000-350,000 reaction mixture, and GTP is not hydrolyzed by the uvrA pro- Mr. V., void volume; Vi, inner volume. tein (Fig. 4). The possible relationship between the ATPase activity of the uvrA protein and the ATP requirement of the observed to the unirradiated supercoiled DNA can be ascribed uvrABC endonuclease was further investigated by measuring to the presence of single-strand-like regions in such DNA. Km values of ATP for both reactions. Similar Km values were The uvrA Protein Is an ATPase that Hydrolyzes ATP to ADP observed (0.23 mM ATP for the uvrA ATPase activity and 0.22 and Pi. The strand cleavage reaction catalyzed by the combined mM ATP for the uvrABC endonuclease) as would be expected uvr+ gene products is ATP dependent (19). ATP analogues such if the ATP requirement of the uvrABC endonuclease is deter- as adenosine 5'-[y-thio]triphosphate (ATP [yS]) and adenosine mined by the ATPase activity of the uvrA protein (Fig. 5). Both ADP and ATP[yS] are competitive inhibitors of the ATPase activity, with KY = 21 and 53 jLM, respectively (Fig. 5). 81>0 Linear DNA Single-stranded DNA Effect ofATP and ATPase hxihibitors on the Binding ofuvrA lW Protein to DNA. The presence ofMg2" is essential for the DNA- 61 binding activity of the uvrA protein (16). ATP stimulates bind- ing, at least to UV-irradiated DNA (Table 2). In contrast, ADP 4.0 inhibits binding to both nonirradiated and UV-irradiated DNA. 2 No binding at all is observed when ADP replaces ATP, and an Native intermediate response is obtained by adding both these com- 0 pounds together. ATP[yS] has opposite effects and stimulates Superhelical DNA Relaxed DNA :C 6 aS- Table 2. Effect of ATP and uvrA' ATPase inhibitors on binding of uvrA protein to DNA DNA bound, % 10 /inhibitor Control UV-irradiated 10- None 14 20 + ATP 23 55 A 0 Y + ADP 4 4 0 100 0 100 200 + ATP[iS] 99 96 uvrA protein, ng + ATP[-y]* 44 41 + ATP and ADP 17 30 FIG. 3. Bindingof uvrA protein to different forms of nonirradiated + ATP and ATP[-SI 97 100 or UV-irradiated ColEl DNA. Closed symbols refer to binding assays with DNA exposed to 180 J/m2 of 254-nm light and open symbols refer Concentrations of cofactor/inhibitor were 0.5 mM. Results are ex- to nonirradiated DNA. The circular forms of the DNA (supercoiled and pressed as percent DNA retained on filter after incubation with 75 ng relaxed DNA) were repurified by CsCl/ethidium bromide centrifu- of uvrA protein. Irradiated DNA was exposed to 180 J/m2. None refers gation immediately prior to the binding experiment to ensure that es- to standard conditions without ATP in the binding mixture. sentially all (>90%) of the DNA was in the covalently closed form. * Only 15 ng of uvrA protein. Downloaded by guest on September 24, 2021 Biochemistry: Seeberg and Steinum Proc. NatL Acad. Sci. USA 79 (1982) 991

4)

0

4) 04 S.,

-4 -2 0 2 4 6 8 10 -4 -2 0 2 4 6 8 10 1/ATP, mM-1 1/ATP, mM-1 FIG. 5. Double reciprocal plots of uvrA ATPase activity (Left) and uvrABC endonuclease activity (Right) versus ATP concentration. In the ATPase assays 0.1 ,ug of uvrA protein was used and in the endonuclease assays 0.1 pg of uvrA protein and 10 pg of protein of a partially purified combination of uvrB+ and uvrC' gene products (12) were used. (Left) ATPase, in the presence of: only ATP (o), 0.05 mM ADP (u), 0.1 mM ADP (A), 0.15 mM ATP[-y] (o), or 0.4 mM ATP[KI (A). (Right) uvrABC endonuclease in the presence of only ATP. binding to give maximal DNA retention on nitrocellulose filters does not result in strand breaks or alkali-labile phosphodiester irrespective ofthe presence ofADP. When the amount ofpro- bonds in the DNA (Table 3), implying that the uvrA protein tein was reduced to give less than maximal retention, no dif- does not have nuclease properties by itself. Addition ofpartially ference was observed in the extent ofbinding to UV-irradiated purified uvrB+ and uvrC' gene products, however, results in and non-irradiated DNA, showing that ATP[yS] abolishes the strand break formation in essentially all ofthe DNA molecules. preferential binding to UV-irradiated DNA. Further addition of either of the ATPase inhibitors ADP or The uvrA Protein Is Not a UV-DNA Endonuclease or Gly- ATP[yS] results in inhibition ofthe strand break reaction, fur- cosylase. Incubation of UV-irradiated DNA with uvrA protein ther indicating an essential role for the uvrA ATPase activity in the repair endonuclease reaction catalyzed by the combined Table 3. Strand cleavage of UV-irradiated ColEl DNA by the uvrA', uvrB+, and uvrC' gene products. uvrABC endonuclease: Effect of uvrA ATPase inhibitors % DNA DISCUSSION remaining Breaks covalently per The uvrA+ gene product from E. coli has been purified to ap- Proteins added Inhibitor closed circular molecule parent homogeneity from a strain carrying the uvrA gene and a multicopy plasmid. Taking into account a 7-fold overproduc- None 100 0 tion in the plasmid-carrying strain, it can be calculated from uvrA protein, 1 ug - 101 0 Table 1 that the product exists in uvrA protein, 1 pg - 100* 0 uvrA' normally approximately uvrB+/uvrC' gene 100 copies per cell. However, because the uvrA + gene pro- products, 6 Ag of moter appears to be under recA/lexA control, the level ofuvrA+ protein 88 0.12 product will be increased by exposing the cells to UV (21). We uvrA protein, 0.03 have observed a 4-fold increase in the amount ofuvrA+ protein lug, + uvrBE/ by UV induction (unpublished data), which means that-the copy uvrC' gene number can be increased to' 400 per cell during repair in UV- products, 6 pug of irradiated cells. protein - 5 >3 Two distinct features of the uvrA protein have been ob- uvrA + uvrB+/ served, a DNA-binding property and an ATPase activity. Both uvrC', as above ADP, 0.1 mM 27 1.0 ofthese appear to have an essential role in the strand cleavage uvrA + uvrB+/ reaction promoted by the uvrABC endonuclease as indicated uvrC+, as above ADP, 0.2 mM 53 0.63 from several observations. First, the Km value for the ATPase uvrA + uvrB+/ is the same as that observed for the ATP requirement of the uvrC', as above ATP[;S], 0.2 mM 71 0.34 uvrABC endonuclease. Second, ADP and ATP[yS] both inhibit uvrA + uvrBt/ the uvrABC endonuclease as well as the ATPase. Third, both uvrC', as above ATP[ySI, 0.4 mM 77 0.26 these compounds also impair the DNA-binding property; ADP Reaction mixtures contained 0.05 ,ug of ColE [3H]DNA exposed to is inhibitory and ATP[yS], although stimulatory, precludes the 50 J/m2 of UV, 5 Amol of Mops (pH 7.5), 10 ,umol of KCl, 2 ,umol of preferential binding to damaged DNA. These results suggest MgSO4, 0.1 ,umol ofEDTA, 0.1 ,mol of dithiothreitol, 50 nmol of ATP, that the role ofthe ATPase is to promote specific binding ofthe and protein as indicated in a total volume of 130 ul. Incubation was uvrA protein to the lesion. The uvrA protein may be a recog- for 15 min at 37°C. Cleaved DNA was separated from covalently closed nition factor that promotes and DNA by sedimentation in alkaline sucrose gradients as described (12). assembly ofthe uvrA+, uvrB+, Results are normalized to the percent of covalently closed DNA in con- uvrC+ proteins at the damaged site for strand cleavage to occur. trol samples incubated in buffer (65%). It can further be speculated that the uvrA protein forms a high- * Treated with 0.3 M NaOHfor 30 min at20°C afterenzyme incubation affinity DNA-binding complex with ATP that is converted to and before sedimentation in alkali. a low-affinity complex when ATP is hydrolyzed to ADP. The Downloaded by guest on September 24, 2021 992 Biochemistry: Seeberg and Steinum Proc. Nad Acad. Sci. USA 79 (1982) cycle of ADP-ATP exchange and ATP hydrolysis could then 1. Lindahl, T., (1979) Prog. Nucleic Acid Res. MoL Biol 22, 1-192. allow the uvrA protein to move along the DNA helix until the 2. Hanawalt, P. C., Cooper, P. K., Ganesan, A. K. & Smith, C. A. This model is suggested by (1979) Annu. Rev. Biochem. 48, 783-836. damaged site has been identified. 3. Setlow, R. B, & Carrier, W. L. (1964) Proc. Natl Acad. Sci. USA the observed increased binding with the nonhydrolyzable ATP 51, 226-231. analogue ATP[-yS]* and the reduced binding with ADP. 4. Boyce, R. P. & Howard-Flanders, P. (1964) Proc. Natl Acad. Sci. The combination of being a DNA-binding protein and an USA 51,'293-300. ATPase makes the uvrA protein similar to some other proteins 5. Howard-Flanders, P., Boyce, R. T. & Theriot, L. (1966) involved in'DNA metabolism-i.e., the dnaB gene product, 53, 1119-1136. is required for DNA replication (22); the recA gene prod- 6. van de Putte, P., van Sluis, C. A., van Dillewijn, J. & Rbrsch, A. which (1965) Mutat. Res. 2, 97-100. uct, which is essential for DNA recombination (2325); and 7. Lawley, P. D. & Brookes, P. (1965) Nature (London) 206, DNA gyrase, which introduces negative superhelical turns into .480-483. double-stranded DNA (26). In these cases the ATPase activity 8. Ikenaga, M., Ishikawa-Ryo, H. & Kondo, S. (1975) J. Mot Biol seems DNA dependent. The uvrA ATPase appears to be DNA 92, 341-356. independent and there is an idling consumption of ATP in the 9. Cole, R. S., Levitan, D. & Sinden, R. R. (1976)J' Mol Biol 103, 39-59. absence ofexogenous DNA. It is calculated that each uvrA pro- 10. Venitt, S. & Tarmy, E. M. (1972) Biochim. Biophys. Acta 287, tein molecule hydrolyzes 500 molecules of ATP at 37TC in 30 38-51. min under normal conditions. We have been worried that minor 11. Seeberg, E. (1978) Proc. Natl Acad. Sci. USA 75, 2569-2573. quantities ofDNA from the DNA cellulose column, which could 12. Seeberg, E. (1981) Mutat. Res. 82, 11-22. be present in the uvrA protein preparation, may give a false 13. Seeberg, E., Nissen-Meyer, J. & Strike, P. (1976) Nature (Lon- indication of DNA independence. However, attempts to re- don) 263, 524-526. DNA DEAE-cellulose or 14. Rupp, W. D., Sancar, A., Kennedy, W. J., Ayers, J. & Griswold, move such either by chromatography J. (1978) in DNA Repair Mechanisms, eds. Hanawalt, P. C., by treatment with DNase did not reduce the ATPase activity Friedberg, E. C. & Fox, C. F. (Academic, New York), pp. of the uvrA protein. Although we cannot exclude that the pu- 229-235. rified uvrA protein still has tightly bound pieces of DNA that 15. Alberts, B. &.Herrick, G. (1970) Methods Enzymol 21, 198-217. are not removed by these treatments, it seems more likely that 16. Seeberg, E. (1978) in DNA Repair Mechanisms, eds. Hanawalt, the uvrA ATPase activity is not coupled to a specific reaction P. C., Friedberg, E. C. & Fox, C. F. (Academic, New York), pp. the uvrA is 225-228. with the DNA, at least not when protein separated 17. Bradford, M. M. (1976) AnaL Biochem. 72, 248-253. from the other uvr+ gene products. 18. Sancar, A., Wharton, R. P., Seltzer, S., Kacinski, B. M., Clarke, In conclusion, a simple protocol is described for extensive N. D. & Rupp, W. D. (1981) J. Mol Biol 148, 45-62. purification of the uvrA+ protein from E. coli, and the availa- 19. Waldstein, E. A., Sharon, R. & Ben-Ishai, R. (1974) Proc. Natl bility of this protein in a homogeneous form should help to Acad. Sci. USA 71, 2651-2654. elucidate early steps of excision repair in E. coli. 20. Seeberg, E. (1981) Prog. Nucleic Acid Res. Mol. Biol. 26, 217-226. 21. Kenyon, C. J. & Walker, G. C. (1980) Proc. Nati. Acad. Sci. USA 77, 2819-2823. Note Added in Proof. Since the submission ofthis manuscript, Kacinski 22. Wickner, S. & Hurwitz, J. (1975) Proc. Nati Acad. Sci. USA 72, et aL (27) have reported the purification ofthe uvrA gene product from 921-925. radiolabeled maxicells and shown that the uvrA protein is an ATPase, 23. Ogawa, T., Wabiko, H., Tsujimoto, T., Horii, T., Masuka, H. in agreement with the results reported here. & Ogawa, H. (1978) Cold Spring Harbor Symp. Quant. Biol 43, 909-915. 24. Roberts, J. W., Roberts, C. W., Craig, N. L. & Phizicky, E. M. (1978) Cold Spring Harbor Symp. Quant. Biol 43, 917-920. E.S. acknowledges travel support from the Norwegian Research 25. Weinstock, G. M., McEntee, K. & Lehman, I. R. (1979) Proc. Council for Science and the Humanities. Nati Acad. Sci. USA 76, 126-130. 26. Sugino, A., Higgins, N. P., Brown, P. Q., Peebles, C. L. & Coz- * ThatATP[-yS] is not beinghydrolyzed by the uvrA' ATPase was tested zarelli, N. R. (1978) Proc. Natl Acad. Sci. USA 75, 4838-4842. in experiments with ATP[7S]. Less than 5 pmol of 'S was released 27. Kacinski, B. M., Sancar, A. & Rupp, W. D. (1981) Nucleic Acids in 30 min at 370C by 200 ng ofuvrA protein under standard conditions. Res. 9, 4495-4508. Downloaded by guest on September 24, 2021