Proc. Natl. Acad. Sci. USA Vol. 88, pp. 6824-6828, August 1991 Biochemistry The C-terminal half of UvrC protein is sufficient to reconstitute (A)BC excinuclease (DNA repair/truncated protein/active site) JING-JER LIN AND Aziz SANCAR Department of Biochemistry and Biophysics, University of North Carolina School of Medicine, Chapel Hill, NC 27599 Communicated by Mary Ellen Jones, April 22, 1991 (received for review March 4, 1991) ABSTRACT The UvrC protein is one of three subunits of The uvrC gene of E. coli encodes a protein of 610 amino the Escherichia coli repair enzyme (A)BC excinuclease. This acids and a molecular weight of 68,510 (7, 11). To define the subunit is thought to have at least one of the active sites for role of UvrC in excision repair more specifically and to nucleophilic attack on the phosphodiester bonds of damaged identify the active site(s) within the protein, we mutagenized DNA. To localize the active site, mutant UvrC proteins were the gene by linker-scanning mutagenesis. The results ob- constructed by linker-scanning and deletion mutagenesis. In tained by this method led us to make plasmid constructs that vivo studies revealed that the C-terminal 314 amino acids ofthe expressed only parts of the gene from the N or C termini. We 610-amino acid UvrC protein were sufficient to confer UV found that a plasmid expressing the 314 C-terminal amino resistance to cells lacking the uvrC gene. The portion of the acids of the protein complemented uvrC- mutations in vivo. uvrC gene encoding the C-terminal halfofthe protein was fused A fusion protein (MBP-C314C) was constructed in which the to the 3' end of the E. coli malE gene (which encodes maltose sequence for the maltose binding protein (MBP) was fused to binding protein), and the fusion protein MBP-C314C was the sequence for the C-terminal 314 amino acids ofUvrC. The purified and characterized. The fusion protein, in combination fusion protein was purified and was found to produce the two with UvrA and UvrB subunits, reconstituted the excinuclease incisions typical of (A)BC excinuclease when mixed with activity that incised the eighth phosphodiester bond 5' and the UvrA and UvrB proteins. We conclude that the C-terminal fourth phosphodiester bond 3' to a psoralen-thymine adduct. halfofUvrC carries all the structural requirements necessary These results suggest that the C-terminal 314 amino acids of for (A)BC excinuclease activity. UvrC constitute a functional domain capable of interacting with the UvrB-damaged DNA complex and of inducing the two phosphodiester bond incisions characteristic of (A)BC excinu- MATERIALS AND METHODS clease. Materials. The E. coli K-12 DR1984 (uvrC34 recAl) was used as a host for initial testing of the complementing (A)BC excinuclease is the collective activity of UvrA, UvrB, activities of various plasmids carrying mutant uvrC genes and UvrC proteins of Escherichia coli (1, 2). The enzyme (11). The mutant genes that were studied in more detail were initiates nucleotide excision repair of DNA damage ranging also tested in E. coli N3024 (uvrC 279: :TnlO), a strain with a from covalently modified bases (2) to noncovalent drug- TnJO insertion in the 3' halfofuvrC (12). Finally, by "curing" nucleotide complexes (3, 4) to abasic sites generated by N3024 of the TnJO insertion, we obtained UNC3024(AuvrC) oxidative metabolism or ionizing radiation (5, 6). This type of in which at least 60o of the 5' terminal half of the uvrC gene enzymatic activity appears to be widespread in both pro- and has been deleted, as determined by Southern blot analysis eukaryotes (1, 7). Therefore, understanding the details of using the 1.2-kilobase-pair Bgl I-Nco I fragment of uvrC as nucleotide excision repair in E. coli is likely to contribute to the probe (11). The S1 nuclease, T4 DNA polymerase, and studies of DNA repair in other biological systems as well. restriction enzymes were purchased from Bethesda Research The reaction mechanism of (A)BC excinuclease has been Laboratories. E. coli endonuclease IV (13) was kindly pro- investigated in some detail (8, 9). These studies have revealed vided by B. Demple (Harvard University). UvrA, UvrB, and a unique mechanism for high-affinity binding to damaged UvrC proteins were purified according to Thomas et al. (14). DNA and specific removal of modified nucleotides. UvrA, The pMAL-c plasmid used for gene fusions was obtained which has a moderate affinity to damaged DNA compared to from New England Biolabs. unmodified DNA, makes a (UvrA)2(UvrB)1 complex and this Construction of Linker-Scanning and Deletion Mutants. complex delivers UvrB to the damaged site; a very stable Linker-scanning mutants of uvrC (Fig. 1) were generated in UvrB-DNA complex is formed and UvrA dissociates from the UvrC-overproducing plasmid pDR3274 (11) as described the site. UvrC, which has no affinity for free UvrB, binds to (15). Briefly, an average of one apurinic site per plasmid was the UvrB-DNA complex with high affinity and specificity introduced by heating pDR3274 at 65°C at pH 5.2 for 30 min and induces the cleavage of the eighth phosphodiester bond (5). The plasmid was then treated with endonuclease IV and 5' and the fourth or fifth phosphodiester bond 3' to the S1 nuclease to induce a double-strand break. The linearized damaged nucleotide. Since both UvrB and UvrC are present DNA was then ligated to the 8-base-pair (bp) Sst I linker in the incision complex, it is unclear at present which of the CGAGCTCG. The plasmids were transformed into DR1984 two subunits is responsible for the dual nucleophilic attack on and tested for phenotypic complementation by UV survival the phosphodiester backbone. It has been reported that UvrC and the sites of linker insertion were mapped by restriction has a weak nonspecific nuclease activity (10), which would analysis. The exact locations of the inserts in plasmids of suggest that at least one of the specific incisions is made by interest were determined by double-stranded DNA sequenc- UvrC. ing. Plasmids expressing the N-terminal fragments of UvrC The publication costs of this article were defrayed in part by page charge (1-182 and 1-404 amino acids with internal deletions of payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. Abbreviation: MBP, maltose binding protein. 6824 Downloaded by guest on October 2, 2021 Biochemistry: Lin and Sancar Proc. NatL Acad. Sci. USA 88 (1991) 6825 uvrC UV survival Ncol BgIll (5 erg/mm2 I WtIto- -1~ ~~~~~11.4 xx 10-0 0 300 600 900 1200 1500 1801l0 bp A (187- * 9.2 X 1i0 270) A(423- . 5.8 x 10 832) A (487- 5.1 X 106 515)+ A (532- 9.1 623)+ X 106 A(546- 1.8 X 10 885) FIG. 1. Schematic representation of uvrC A (636- _-3 linker-scanning mutants. All 11 mutants shown 707) 3_0__3.0 x 10 had deletions (A) of the indicated base pairs and an insertion of an 8-bp Sst I linker. Thick lines, A(952- 40 X i02 coding sequences; thin lines, internal deletions. 1014) On the right are the survivals ofDR1984 (uvrC34 recAlI) carrying the appropriate constructs after A (1097 ___________________________________________ 5.0 x 10-7 254-nm irradiation (5 erg/mm2;1 erg = 0.1 1160) from a germicidal lamp. Because ofthe extremepJ) 4 sensitivity of this strain to UV, minor differ- A (1453- 3.8 x 10 ences in the UV dose delivered can cause sur- 1465) vival differences of up to 10-fold with a given - 6 strain. Therefore, we do not consider the sur- 4.1 A(1583- ._. X 10 vivals from 4 x 10-6to 4 x 10-7to be signifi- 1652) cantly different. The higher survivals 10-2_10-5 A( 1713- 6- 6 have been obtained reproducibly. +, In-frame 1785) 7.2 x 10 deletion fusion. These experiments were re- 7 peated in a uvrC null background and identical uvrC strain: 3.9 x 10 results were obtained (data not shown). various lengths) were constructed (Fig. 2) by digesting the Nco I linkers of appropriate lengths (for in-frame fusion with appropriate linker-scanning mutants with EcoRP and Sst I or ATG at the Nco I site) were ligated to the blunt termini to Nco I and inserting the resulting uvrC fragment into pBR328 generate plasmids in which the Sst I sites were replaced with (Boehringer Mannheim). Plasmids expressing C-terminal Nco I sites. These latter plasmids were then digested with fragments of UvrC were constructed as follows. First, the Nco I and Pst I to obtain fragments carrying the 3'-terminal linker-scanning plasmids were digested with Sst I and the regions of uvrC. These fragments were inserted into the sticky ends were "blunted" with T4 DNA polymerase. Then, expression vector pKK233-2 (Pharmacia LKB) to obtain UV survival UvrC (5 erg/mm2) Wt. N C 1.4 x 10 0 100 200 300 400 500 600 a.a. C182N 2.5 x 106 FIG. 2. UvrC C- and N-terminal deletions. The C404NA1 4.1 x 1 6 wild-type protein is 610 amino acids long. The deletion mutants are indicated by the number and relative locations of the amino acids they are pre- C404NA2 1.7 x 1 6 dicted to contain. The constructs that encode N-ter- minal peptides C4O4NA&1 and C40NA2 carry the first 1212-bp coding sequences ofuvrC with internal C375C 1.1 x id'1deletions of 340 bp (Al) and 72 bp (A2), respec- tively, corresponding to deletions in constructs C31 4C 3.4 >~ 3 Both4(546-885)oftheseanddeletionsA(636-707),shift therespectively,reading frame.in Fig.The1.