Two DNA recognition domains of the specificity polypeptides of a family of type I restriction enzymes

The hsd genes of Salmonella typhimurium and Salmonella potsdam encode related type I restriction and modification systems designated SB and SP, respectively; the polypeptide encoded by the hsdS gene dictates the DNA sequence recognized. The hsdS genes of the SB and SP systems have a conserved sequence of around 100 base pairs flanked by two nonhomologous (variable) regions of around 500 base pairs. Recombination between th hsdS genes of SB and SP generated a system (SQ) with a different recognition specificity. We have localized the position of the crossover in the central conserved region by analysis of nucleotide sequences. Concomitant with the generation of a new combination of flanking variable regions is the recombination of minor differences in the central conserved region. A polypeptide domain encoded on the 5' side of the crossover dictates recognition of the trinucleotide component of the target sequence, and a second domain, encoded on the 3' side of the crossover, similarly governs recognition of the tetra- or penta-nucleotide component. Our analysis implicates at least parts of the variable regions in the determination of the specificity of interaction between protein and DNA. Furthermore, the trinucleotide components of the recognition sequences of S. typhimurium and Escherichia coli K-12 are identical, and the 5' segments of their hsdS genes are strikingly homologous rather than variable.