Allelic diversity and recombination in Campylobacter jejuni.

TLDR: These analyses showed that intraspecific recombination is frequent in C. jejuni and has generated extensive diversity of allelic profiles from a small number of polymorphic nucleotides.

Abstract: Campylobacter jejuni infection is one of the most frequent causes of bacterial food-borne diarrheal disease all over the world. While in most cases the disease is self-limiting, C. jejuni infections can give rise to the debilitating and potentially fatal Guillain-Barre syndrome, a progressive neuromuscular paralysis (for a review, see reference 32). Over the last decade, numerous genotypical typing methods for Campylobacter species have been described, including pulsed-field gel electrophoresis (12, 35), flagellin gene typing (4, 24), randomly amplified polymorphic DNA (RAPD) PCR analysis (11, 23), and, most recently, amplified fragment length polymorphism (8). (For a review of typing methods for C. jejuni, see reference 34.) However, although all of these methods ultimately depend on sequence variation, up to now, there have been very few systematic analyses of nucleotide sequence variability in C. jejuni. In several other species of pathogenic bacteria, analysis of nucleotide sequence variation at multiple gene loci has permitted us to gain further understanding of the population structure of these pathogens. Such analyses have shown that different pathogens differ widely in the extent of sequence variation, their population structure, the relative roles of mutation and recombination, and the existence of clonal groupings with distinct geographic distribution patterns (2, 3, 30). Multilocus sequence typing (MLST), a method that is based on partial nucleotide sequences of multiple (usually around seven) housekeeping genes, has recently been shown to be a powerful technique for bacterial typing (3, 9, 19). Housekeeping genes are preferred over virulence-associated genes, because an analysis of mutations (most of which are usually synonymous, given the strong selection against changes of the amino acid sequence in genes coding for proteins required for growth) in such genes is more likely to adequately reflect the phylogeny of strains. For a more extensive discussion of these arguments, see reference 1. While still relatively expensive, major advantages of this technique are the easy portability of both the method and results and the possibility of building up global databases by using the internet. An additional advantage of MLST approaches is that results can be used to perform phylogenetic and population genetic analyses. The feasibility of sequence-based typing depends on the identification of genes that have sufficiently high sequence variability. Since there was very little information about the extent of sequence variability in C. jejuni, we have determined nucleotide sequences of seven housekeeping genes that were selected from the recently published whole genomic sequence of C. jejuni (25). Nucleotide sequences of 423- to 660-bp fragments from these housekeeping genes were obtained for a collection of 32 C. jejuni strains from Germany, Hungary, Thailand, and the United States and analyzed for their variability. We have also assessed the frequency of recombination with the homoplasy test (21) and studied the population structure of C. jejuni by split decomposition. The data show that the population structure of C. jejuni is characterized by a low degree of sequence diversity, a relatively small pool of alleles in the housekeeping genes tested, and high rates of intraspecies recombination. Recombination is frequent enough to generate a large number of unique combinations of alleles (sequence types), implying that MLST approaches could be valuable for future studies of the molecular epidemiology of C. jejuni.