Sequence of theProvidencia rettgeri lexAgene and its control region

In Escherichia coli, DNA damage induces a set of at least 20 genes, called the SOS network, whose products directly repair DNA or allow the cell to tolerate the lesion until the repair may be accomplished (1). The SOS response is controlled by the RecA and LexA proteins. LexA is the common repressor of the SOS genes, which include both recA and lexA (2). The basal RecA protein is reversibly activated by an inducing signal after DNA damage (3). Activated RecA protein has apoprotease activity which facilitates autocatalytic cleavage of the LexA repressor (4). Similar DNA damage-inducible responses in other bacterial species have been reported (5, 6). A system to directly isolate lexA genes of Gram-negative bacteria has recently been developed (7). Through this system, the lexA genes of Salmonella typhimurium, Erwinia carotovora, Pseudomonas aeruginosa and Pseudomonas putida have been characterized (8). Providencia is phylogenetically the genus of Enterobacteria which is the most distant from E. coli (9). It has been suggested that the regulation of the SOS response of Providencia rettgeri could be different from that of E.coli (10). To clarify this point, and to analyze the variability of the lexA gene in the Enterobacteriaceae family, we have cloned and sequenced the lexA gene of P.rettgeri ATCC29944. Five lexA genes sequenced so far have two tandem consensus LexA binding sites 5' to the lexA gene which are used for autocontrol of LexA protein levels (8, 11). P. rettgeri also presented two such binding sites, the distance between them being 5 base pairs, as happens in all the other enterobacterial lexA genes. The first SOS box of P. rettgeri showed the same sequence as that of E.coli, S. typhimurium and E. carotovora (CTGTATATACTCACAG). The downstream P. rettgeri binding site differs by 1 residue from that of E. carotovora and by 4 from those of E. coli and S.typhimurium. These data are in agreement with the hypothesis that the first SOS box is the most important one in the process of LexA repressor binding to the lexA gene (12). The open reading frame of the lexA gene of P. rettgeri presents an insertion of 9 base pairs at position 221 in comparison with that of E.coli. This fact makes P. rettgeri have the longest lexA gene known at this moment (618 bp long). This insertion is in the hinge region which links two functional domains of the LexA repressor (13). The predicted P. rettgeri LexA protein is 86% identical to the E.coli, S.typhimurium and E.carotovora, and 62% to the P. aeruginosa and P.putida. It is known that the aminoterminal portion of E.coli LexA is involved in DNA binding (14). This domain contains three a-helices comprised of amino acids Gln8 to Ser20, Arg28 to Leu35, and Asn41 to Gly54 (15). There is not any difference in the composition of the a3> helix of the LexA repressor of P. rettgeri, whereas in the a l helix there are three substitutions at positions 10, 12 and 15, and the ot2 helix showed two changes at positions 33 and 34. The hydrophobic pocket formed by Ala32, Ala42, and Ala43 of the LexA protein of E.coli is perfectly conserved in P.rettgeri, including the Phe37 residue. The fragment that spans positions 61—67, which has been proposed to be involved in DNA-binding ability (16), does not present any substitution. The carboxy-terminal domain of E.coli LexA contains the amino acid residues necessary for cleavage reaction, interaction with RecA protein (4), and dimerization (17). Cleavage of E.coli LexA takes place between amino acids Ala84 and Gly85 with the participation of residues Serll9 and Lysl56 (18). As expected, P. rettgeri LexA also contains the Ala-Gly bond, but at positions 87 — 88 as a consequence of the 9 base pairs insertion mentioned above. The residues Serl 19 and Lysl56 of E.coli are displaced three positions in P. rettgeri. Thus, regardless of the precise position of the Ala-Gly bond and of the Lys and Ser residues, the distance between these three points remains unaltered in the P. rettgeri LexA repressor. Finally, repression of the E.coli sulA promoter-operator in vivo by LexA protein of P. rettgeri is about two-fold stronger than that carried out by the LexA repressor of E.coli. E.coli RecA coprotease can support LexA repressor cleavage of P. rettgeri in vivo, although with lower efficiency than that shown with its own LexA repressor.