Identification and cloning of waaF (rfaF) from Bordetella pertussis and use to generate mutants of Bordetella spp. with deep rough lipopolysaccharide

Abstract

A DNA locus from Bordetella pertussis capable of reconstituting lipopolysaccharide (LPS) O-antigen biosynthesis in Salmonella typhimurium SL3789 (rfaF511) has been isolated, by using selection with the antibiotic novobiocin. DNA within the locus encodes a protein with amino acid sequence similarity to heptosyltransferase II, encoded by waaF (previously rfaF) in other gram-negative bacteria. Mutation of this gene in B. pertussis, Bordetella parapertussis, and Bordetella bronchiseptica by allelic exchange generated bacteria with deep rough LPS phenotypes consistent with the proposed function of the gene as an inner core heptosyltransferase. These are the first LPS mutants generated in B. parapertussis and B. bronchiseptica and the first deep rough mutants of any of the bordetellae.

FIG. 1

Silver-stained Tris-tricine polyacrylamide gel of S. typhimurium LPS isolated from wild type, waaF mutant, and waaF mutant complemented with B. pertussis waaF. Lane 1, S. typhimurium waaF mutant (SL3789) complemented by the BP536 waaF gene; lane 2, SL3789 alone; lane 3, S. typhimurium wild type (SL3770). The waaF mutant displays the deep rough LPS phenotype expected, while the wild-type control shows the ladder pattern expected for a full-length smooth LPS. The complemented mutant also has the O-antigen ladder, but a rough core molecule is also observed in the LPS preparation, suggesting that the complementation is not completely effective.

FIG. 2

BOXSHADE of a PILEUP performed in the GCG DNA analysis package with MsbA protein sequences from E. coli (Ecoli) and H. influenzae (Hin) and the proposed homolog from B. pertussis (BP536). The black shading surrounds blocks of amino acids which are identical, and the grey shading surrounds blocks with conservative substitutions. The B. pertussis sequence is shown as starting with a leucine residue since it has TTG as a start codon. This sequence is truncated at the position of the SacI site where the DNA sequence published here starts. Only the parts of the E. coli and H. influenzae sequences corresponding to the truncated B. pertussis sequence are shown.

FIG. 3

BOXSHADE of a PILEUP performed in the GCG DNA analysis package with WaaF protein sequences from N. gonorrhoeae (Ngon), Neisseria meningitidis (Nmen), E. coli (Ecoli), S. typhimurium (Salty), H. influenzae (Hin), and Pseudomonas aeruginosa (Psaer) and the proposed B. pertussis (BP536) WaaF protein. See Fig. 2 legend for an explanation of the shading.

FIG. 4

Silver-stained Tris-tricine polyacrylamide gel of wild type and waaF mutant allelic exchange mutants from B. pertussis, B. bronchiseptica, and B. parapertussis. Lane 1, B. pertussis (BP536) wild type; lane 2, B. pertussis waaF mutant; lane 3, B. bronchiseptica (CN 7635E) wild type; lane 4, B. bronchiseptica waaF mutant; lane 5, B. parapertussis (CN 2591) wild type; lane 6, B. parapertussis waaF mutant.