The genus Bordetella contains eight species, of which only three have been studied in detail — Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. On the basis of similarities between cloned genes, DNA hybridization studies, multi-locus enzyme electrophoresis (MLEE), identification of insertion sequence (IS) element POLYMORPHISMS and metabolic characteristics, it was proposed that these species were extremely closely related. These studies indicated that the genetic diversity amongst these organisms was restricted when compared with other pathogenic bacterial species, including Helicobacter pylori, Haemophilus influenzae and Streptococcus pyogenes. On the basis of this lack of diversity, Musser and colleagues proposed that B. pertussis and B. parapertussis should be classified as subspecies of B. bronchiseptica rather than as species in their own right. These studies also revealed a clustering of Bordetella clones. Genomic analyses showed that ovine clones of B. parapertussis were clustered together but were quite separate from human B. parapertussis isolates. Furthermore, whereas ovine isolates showed genetic variability, using these experimental approaches the human isolates were identical, regardless of the geographical location and period of time during which they were isolated. In addition, B. pertussis isolates clustered separately from other Bordetella isolates.

General features of the three genomes are shown as follows. They differ considerably in size and this correlates with their gene-coding capacities; B. bronchiseptica encodes 1,191 more genes than B. pertussis, a difference that is compounded because the B. pertussis genome has 358 pseudogenes whereas the B. bronchiseptica genome has only 18 pseudogenes.



Preston Andrew, The bordetellae: lessons from genomics. [J] .Nat. Rev. Microbiol..

Comparison of the gross structures of the three genomes confirmed earlier studies which indicated that B. pertussis and B. parapertussis evolved from a B. bronchiseptica ancestor. B. bronchiseptica RB50 lacks IS elements (although other strains of B. bronchiseptica do contain IS elements), whereas the B. pertussis and B. parapertussis genomes have 261 and 112 copies of different IS elements, respectively. It seems likely that the acquisition or expansion of IS elements was important in the evolution of B. pertussis and B. parapertussis from independent B. bronchiseptica like ancestors. There is considerable co-linearity between the B. bronchiseptica and B. parapertussis genomes, but several rearrangements have clearly taken place in B. parapertussis, most of which are bordered by copies of IS1001 or IS1002. This indicates that recombination between identical copies of the IS elements is an important mechanism of chromosomal rearrangements in these species. Similar, but much more extensive IS-element related recombination is observed in B. pertussis, adjacent to IS481, with nearly 150 rearrangements observed. Recombination between IS elements has also resulted in deletions that are detectable in the B. pertussis and B. parapertussis genomes, resulting in different genome sizes between the three bordetellae that have been sequenced. Speciation of B. pertussis and B. parapertussis seems to be primarily through loss of DNA rather than acquisition of foreign DNA. Inactivation of genes by insertion of IS elements is observed frequently in the B. pertussis and B. parapertussis genomes, which provides a further mechanism for the generation of the large number of pseudogenes that are present in these species.

Analysis of the gene complements of the three genomes also supports the hypothesis that B. pertussis and B. parapertussis are independent derivatives of B. bronchiseptica, and that they evolved primarily due to a loss of gene function. Genome analysis identified only 114 genes that are unique to B. pertussis, excluding IS elements. However, this gene set could merely reflect differences between B. bronchiseptica strain RB50 and the B. bronchiseptica strain from which B. pertussis evolved. Microarray- based GENOMOTYPING of 42 Bordetella strains revealed that 103 of the 114 genes are present in other strains of B. bronchiseptica, and that only 11 genes are specific to B. pertussis. Functions for these 11 genes could not be assigned using bio-informatic approaches, but the authors note that 9 out of the set of 11 genes were probably acquired horizontally as a single contiguous locus. The same study showed that all of the 50 B. parapertussis genes that were identified by the genome project as being unique to this species were present in strains of B. bronchiseptica other than RB50. So, there are no genes that have been identified so far as being unique to B. parapertussis.

Bordetella pathogenesis

Despite the differences in host range of the bordetellae, and the different pathologies that each cause in different hosts, the steps in the pathogenesis of the different bordetellae are thought to be similar. Bordetella pathogenesis has been comprehensively reviewed. In most cases, the host clears B. pertussis and B. parapertussis infections, whereas B. bronchiseptica infections can be chronic. Acquired immunity to B. pertussis in humans develops after natural infection, although the emergence of B. pertussis infections in adults indicates that this protection, in common with the protection obtained after vaccination, might not be life-long. The immunology of Bordetella infection has been reviewed elsewhere. Numerous bacterial components have been implicated in the pathogenesis of Bordetella, for example, adhesins, including toxins and fimbriae, and filamentous haemagluttinin (FHA), including pertussis toxin (PT) and adenylate cyclase.

Lipid A is the hydrophobic anchor of lipopolysaccharide (LPS), which forms the outer leaflet of the outer membrane (OM) of Gram-negative bacteria Lipid A is an essential structural component of the Gram-negative bacterial cell envelope and constitutes a target for the development of antibacterial agents. Some non-essential modifications to the structure of lipid A have proven necessary for bacterial virulence and, thus, provide targets for anti-infective agents. Lipid A is also the active component of LPS endotoxin, which can promote septic shock when shed from the bacterial surface during systemic infection. This inflammatory reaction is mediated through the Toll-like receptor 4 (TLR4) signal transduction pathway, and leads to the production of cytokines, chemokines and antimicrobial agents needed to mount an effective innate immune response to bacterial infection. Additionally, activation of TLR4 by lipid A initiates the expression on antigen-presenting cell surfaces of the costimulatory molecules needed to mount an adaptive immune response to infection. The inflammatory response is a double-edged sword because it is necessary to eliminate most infections and, at the same time, responsible for some of the main pathophysiological symptoms associated with persistent infections. Bacterial pathogens and symbionts can coexist with their hosts in part because they modify the structure of lipid A to attenuate the inflammatory response and evade immune recognition.

To be continued in Part Two…