Molecular characterization of a large Borrelia burgdorferi motility operon which is initiated by a consensus sigma70 promoter.

A large motility operon, referred to as the flgB operon, was identified, characterized, and mapped at 310 to 320 kb on the linear chromosome of the spirochete Borrelia burgdorferi. This is the first report that a sigma70-like promoter rather than a sigma28-like promoter is involved in the transcription of a major motility operon in bacteria. From these results in conjunction with results from a previous study (Y. Ge and N. W. Charon, Gene, in press), we have identified 26 genes in this operon that are relevant to motility and flagellar synthesis. With few exceptions, the gene order and deduced gene products were most similar to those of other spirochetes and Bacillus subtilis. Primer extension analysis indicated that transcription initiated from a conserved sigma70-like promoter immediately upstream of flgB; this promoter mapped within the heat-shock-induced protease gene hslU. Reverse transcriptase PCR analysis indicated that a single transcript of 21 kb initiated at this promoter and extended through flgE and (with our previous results) onto the putative motility gene flbE. The flgB promoter element had strong activity in both Escherichia coli and Salmonella typhimurium. As expected, a mutant of S. typhimurium with an inactivated flagellum-specific sigma28 factor did not affect the function of this promoter. Western blot analysis indicated that B. burgdorferi recombinant FliG and FliI were antigenically similar to those of E. coli and other spirochetes. Although complementation of E. coli or S. typhimurium fliG or fliI mutants with the B. burgdorferi genes was unsuccessful, B. burgdorferi recombinant FliI completely inhibited flagellar synthesis and motility of wild-type E. coli and S. typhimurium. These results show that spirochete motility genes can influence flagellar synthesis in other species of bacteria. Finally, Western blot analysis with sera from infected humans and animals indicated a weak or nondetectable response to recombinant FliG and FliI. These results indicate that these antigens are not favorable candidate reagents to be used in the diagnosis of Lyme disease.