Bordetella pertussis, molecular pathogenesis under multiple aspects

Recent studies, including those based on genomics, have demonstrated that besides toxins and adhesins, Bordetella pertussis uses many additional virulence determinants. Most of them are part of the BvgAS regulon, although some, in particular iron-uptake systems, are independent of BvgAS. They are regulated by iron, although in one case, the production of a siderophore receptor could be linked to the BvgAS regulon.     

Immunity to Bordetella pertussis.

Bordetella pertussis exploits extracellular and intracellular niches in the respiratory tract and a variety of immune evasion strategies to prolong its survival in the host. This article reviews evidence of complementary roles for cellular and humoral immunity in protection. It discusses the effector mechanisms of bacterial elimination, the strategies employed by the bacteria to subvert protective immune responses and the immunological basis for systemic and neurological responses to infection and vaccination.     

Structure of Bordetella pertussis peptidoglycan.

Bordetella pertussis Tohama phases I and III were grown to the late-exponential phase in liquid medium containing [3H]diaminopimelic acid and treated by a hot (96 degrees C) sodium dodecyl sulfate extraction procedure. Washed sodium dodecyl sulfate-insoluble residue from phases I and III consisted of complexes containing protein (ca. 40%) and peptidoglycan (60%). Subsequent treatment with proteinase K yielded purified peptidoglycan which contained N-acetylglucosamine, N-acetylmuramic acid, alanine, glutamic acid, and diaminopimelic acid in molar ratios of 1:1:2:1:1 and less than 2% protein. Radiochemical analyses indicated that 3H added in diaminopimelic acid was present in peptidoglycan-protein complexes and purified peptidoglycan as diaminopimelic acid exclusively and that pertussis peptidoglycan was not O acetylated, consistent with it being degraded completely by hen egg white lysozyme. Muramidase-derived disaccharide peptide monomers and peptide-cross-linked dimers and higher oligomers were isolated by molecular-sieve chromatography; from the distribution of these peptidoglycan fragments, the extent of peptide cross-linking of both phase I and III peptidoglycan was calculated to be ca. 48%. Unambiguous determination of the structure of muramidase-derived peptidoglycan fragments by fast atom bombardment-mass spectrometry and tandem mass spectrometry indicated that the pertussis peptidoglycan monomer fraction was surprisingly homogeneous, consisting of greater than 95% N-acetylglucosaminyl-N-acetylmuramyl-alanyl-glutamyl-diaminopimelyl++ +-alanine.     

Bordetella pertussis bacteriophage

For the first time Bordetella pertussis bacteriophage was isolated, and its presence was confirmed by electron microscopy and by agar layer titration. The lysogenic strains were activated by their treatment with mitomycin C in a dose of 4.5 mg/ml. The phage system of the Bordetella genus, heretofore unknown, has been revealed: Bordetella pertussis phage lyzed all the tested strains of Bordetella parapertussis (25 strains) and could be passaged in these strains. The phage formed turbid and transparent negative colonies 0.1 mm and 0.15 mm in size. The phage titer (e. g., in strain No. 3865) was 1 X 10(10). The lysogenic variants of Bordetella pertussis, capable of spontaneous release of the phage, were obtained. These variants were characterized by changes in some of their phenotypical properties, e.g., the increased content of certain toxic substances and increased virulence.     

Bordetella pertussis serotypes in Canada.

A study was done to determine the major antigenic factors of Bordetella pertussis strains isolated throughout Canada and whether these isolates have the same antigenic structure as the bacilli in the currently used vaccines. Testing for the major pertussis antigens, factors 1, 2 and 3, was conducted with 440 freshly isolated strains of B. pertussis received from seven canadian provinces between August 1976 and February 1978 and six batches of pertussis vaccine or immunizing agents containing pertussis vaccine. With the aid of specific antisera prepared in rabbits, the antigenic factors were detected by a slide agglutination technique. Almost all (98.9%) of the pertussis strains examined were serotype 1,3.All six batches of pertussis vaccine or immunizing agents containing pertussis vaccine proved to be rich in each of the three main pertussis agglutinogens.     

Helical structure of Bordetella pertussis fimbriae.

The helical structures of Bordetella pertussis fimbriae of serotypes 2 and 6 were determined by optical diffraction analysis of electron micrographs of negatively stained paracrystalline bundles of purified fimbriae. The fimbrial structure is based on an axial repeat of 13 nm that contains five repeating units in two complete turns of a single-start helix. This structure was confirmed by direct measurements of mass per unit length for individual fimbriae performed by dark-field scanning transmission electron microscopy of unstained specimens. These data further established that the helically repeating unit is a monomer of fimbrial protein (Mr congruent to 22,000 for type 2 and Mr congruent to 21,500 for type 6). Radial density profiles calculated from the scanning transmission electron micrographs showed that the fimbria has peak density at its center, i.e., no axial channel, consistent with the results of conventional negative-staining electron microscopy. The radial profile gives an outermost diameter of approximately 7.5 nm, although the peripheral density is, on average, diffuse, allowing sufficient intercalation between adjacent fimbriae to give a center-to-center spacing of approximately 5.5 nm in the paracrystals. Despite serological and biochemical differences between type 2 and type 6 fimbriae, the packing arrangements of their fimbrial subunits are identical. From this observation, we infer that the respective subunits may have in common conserved regions whose packing dictates the helical geometry of the fimbria. It is plausible that a similar mechanism may underlie the phenomenon of phase variations in other systems of bacterial fimbriae.     

Physico-chemical analysis of Bordetella pertussis antigens.

Physico-chemical methods are being developed for use in the control and standardization of acellular pertussis vaccines and their individual components. We have compared native and detoxified preparations of the B. pertussis antigens, pertussis toxin (PT), filamentous haemagglutinin (FHA), and the 69-kDa outer membrane protein (P69) using circular dichroism (CD), fluorescence spectroscopy, SDS-PAGE and FPLC gel filtration chromatography. Upon aldehyde detoxification, PT underwent a large change in its intrinsic fluorescence maximum (8-10 nm red-shift) and a large increase in its apparent size, detected by chromatography. Polyacrylamide gels showed individual subunits of the same apparent molecular weight (M(r)) as well as some polypeptides of higher M(r). FHA also changed conformation (5-nm red-shift in intrinsic fluorescence) upon aldehyde detoxification, with a resultant increase in the M(r)of its major constituent. The P69 protein appeared quite robust to formaldehyde treatment as measured by the same methods. Its near-UV CD spectrum contains a prominent tryptophan band; so this method may be more suitable for observing differences in conformation. We also examined an aluminium-desorbed DTaP preparation by these methods. When used in conjunction with immunochemical and toxicological assays, these methods are informative and useful in the characterization of candidate standards and should be valuable methods for ensuring the consistency of manufactured vaccines.     

Production of cell mass and pertussis toxin by Bordetella pertussis.

The cultivation of Bordetella pertussis affects production of pertussis toxin and biomass. Comparison of batch mode, chemostat operation and pHstat-turbidostatic control showed that productivities for the continuous process were greater than that for the batch operation. Continuous operation in balanced growth at the maximum specific growth rate, provided by the pHstat, resulted in the maximum specific production rate. Because of the strong association of pertussis toxin synthesis and cell growth, the concentration of toxin in the effluent of the continuous processes was greater than the maximum obtained in the batch bioprocess. An expanded Luedeking-Piret model of product formation kinetics fits the observed chemostat data and demonstrates that the production of pertussis toxin from the culture of B. pertussis is predominantly growth associated.     

Molecular and functional analysis of the lipopolysaccharide biosynthesis locus wlb from Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica

The Bordetella pertussis wlb locus (wlb_pe, formerly bpl ) is required for the biosynthesis of a trisaccharide that, when attached to the B. pertussis lipopolysaccharide (LPS) core (band B), generates band A LPS. The equivalent loci in Bordetella bronchiseptica (wlb_br) and Bordetella parapertussis (wlb_pa) were identified and cloned. The wlb_br and wlb_pa loci differ from wlb_pe in that they lack the insertion sequence that defines the right-hand terminus of wlb_pe. Deletion of 12 kb of DNA containing the whole wlb locus (Δwlb) by allelic exchange in each of the three bordetellae had no effect on band B biosynthesis, whereas band A biosynthesis was prevented in B. pertussis and B. bronchiseptica. In B. bronchiseptica and B. parapertussis, Δwlb mutants also lacked O-antigen. Reintroduction of the wlb_pe or wlb_br loci on a shuttle vector into the three Δwlb mutants restored the wild-type LPS phenotype in the B. pertussis and B. bronchiseptica mutants. In the case of B. parapertussis, which normally does not synthesize an apparent band A structure, introduction of the wlb_pe or wlb_br loci now enabled the generation of band A. This suggests that the attachment point for band A trisaccharide on the LPS core is present in B. parapertussis, and further suggests that the wild-type wlb_pa locus is not fully functional. Introduction of the wlb_pa locus into the Δwlb_pe, Δwlb_br and Δwlb_pa mutants had interesting consequences. The B. bronchiseptica and B. parapertussis recipients were now able to biosynthesize O-antigen, but no band A was generated. In the B. pertussis recipient, a truncated band A was expressed consistent with a mutation in the wlbH gene, but on Western blotting the expression of a small amount of full-length band A was also seen. Evidence that the wlbH_pa protein is not fully functional was provided by the failure of the wlb_pa locus to fully complement a B. pertussis wlbH (ΔwlbH_pe) mutant. This was supported by DNA sequence data showing that a single amino acid, conserved between homologous proteins from a range of bacteria, is altered in the B. parapertussis WlbH protein.     

Fc receptor-mediated immunity against Bordetella pertussis.

The relevance of specific Abs for the induction of cellular effector functions against Bordetella pertussis was studied. IgG-opsonized B. pertussis was efficiently phagocytosed by human polymorphonuclear leukocytes (PMN). This process was mediated by the PMN IgG receptors, FcgammaRIIa (CD32) and FcgammaRIIIb (CD16), working synergistically. Furthermore, these FcgammaR triggered efficient PMN respiratory burst activity and mediated transfer of B. pertussis to lysosomal compartments, ultimately resulting in reduced bacterial viability. Bacteria opsonized with IgA triggered similar PMN activation via FcalphaR (CD89). Simultaneous engagement of FcalphaRI and FcgammaR by B. pertussis resulted in increased phagocytosis rates, compared with responses induced by either isotype alone. These data provide new insights into host immune mechanisms against B. pertussis and document a crucial role for Ig-FcR interactions in immunity to this human pathogen.     

Failure to convert Bordetella parapertussis to Bordetella pertussis with a pertussis phage

To analyze the described lysogenic conversion of Bordetella parapertussis to a Bordetella pertussis-like form we used the phage 134 to lysogenize a B. parapertussis strain. Southern blot analysis of the isolated ‘lysogens’ showed that they were not true lysogens, but rather chronically infected strains. These pseudo-lysogens did not show any changes in virulence properties compared with the parental strain. The only difference we could show was a change in the LPS-structure: the pseudolysogens had a rough LPS, like B. pertussis, whereas the parental B. parapertussis strain was smooth.     

Insertional mutagenesis of Bordetella pertussis adenylate cyclase.

We developed an improved method of linker insertion mutagenesis for introducing 2 or 16 codons into the Bordetella pertussis cyaA gene which encodes a calmodulin-dependent adenylate cyclase. A recombinant kanamycin resistance cassette, containing oligonucleotide linkers, was cloned in plasmids which carried a truncated cyaA gene, fused at its 3' end to the 5' end of the Escherichia coli lacZ gene, specifying the alpha-peptide. This construction permitted a double selection for in-frame insertions by using screening for kanamycin resistance and for lactose-positive phenotype, resulting from alpha-complementation. We showed that most of the two-amino acid insertions within the N-terminal moiety of the catalytic domain of adenylate cyclase abolished enzymatic activity and/or altered the stability of the protein. All two-amino acid insertions within the C-terminal part of adenylate cyclase resulted in fully stable and active enzymes. These results confirm the modular structure of the catalytic domain of adenylate cyclase, previously proposed on the basis of proteolytic studies. Two-amino acid insertions between residues 247-248 and 335-336 were shown to affect the calmodulin responsiveness of adenylate cyclase, suggesting that the corresponding region in the enzyme is involved in the binding of calmodulin or in the process of calmodulin activation. In addition, we have identified within the primary structure of adenylate cyclase several permissive sites which tolerate 16-amino acid insertions without interfering with the catalytic activity or calmodulin binding. By inserting foreign antigenic determinants into these permissive sites the resulting recombinant adenylate cyclase toxin could be used to deliver specific epitopes into antigen-presenting cells.     

Review of the biology of Bordetella pertussis.

Bordetella pertussis produces a complex array of adhesins, aggressins and toxins that are presumed to be important in the colonisation of its human host and in ensuring its survival and propagation. The organism also has highly sophisticated mechanisms for regulating virulence factor expression, in response to environmental signals or by reversible mutations. Despite the rapidly increasing knowledge of these aspects of the biology of B. pertussis, our understanding of the pathogenesis of whooping cough is still far from clear. In defining the role of individual factors, reliance has to be placed on in vitro assays or animal models of the human infection, particularly in the mouse, where different conditions may prevail. Some clues to pathogenic mechanisms may be provided by considering other bordetellae, especially B. parapertussis, B. bronchiseptica and B. avium, their similar, but not identical, range of virulence factors and the common features of the diseases caused by these species in their respective hosts. The bordetellae are usually defined as obligate, non-invasive parasites of the respiratory tracts of warm-blooded animals, including birds, with a predilection for the respiratory ciliated epithelium. This definition has been challenged by a number of recent observations. For example, the ability of Bordetella spp. to regulate virulence factor expression in response to external signals strongly suggests that they have alternative habitats where such regulation would be an advantage. These habitats may be intracellular, since it has been shown that B. pertussis, B. parapertussis and B. bronchiseptica can invade and survive within host cells, or they may be in other sites within the same or different hosts. Recent DNA fingerprinting studies of B. pertussis have revealed hitherto unsuspected heterogeneity amongst isolates which could be reflected in antigenic differences between strains. Some of these new perspectives on Bordetella pathogenicity may have implications for pertussis vaccine development.     

Construction of Bordetella pertussis strains that overproduce genetically inactivated pertussis toxin.

Nontoxic analogs of pertussis toxin (PT), produced by in vitro mutagenesis of the tox operon, are immunogenic and protective against infection by Bordetella pertussis. The moderate levels of PT production by B. pertussis, however, make it the limiting antigen in the formulation of multicomponent, acellular, recombinant whooping cough vaccines. To increase production of the highly detoxified Lys9Gly129 PT analog by B. pertussis, additional copies of the mutated tox operon were integrated into the bacterial chromosome at the tox or fha locus by unmarked allelic exchange. Recombinant strains produced in this way secreted elevated levels of the PT analog proportional to gene dosage. The strains were stable during 10-liter fermentations, and yields of up to 80 mg of PT analog per liter were obtained under production-scale conditions. The nontoxic analog was purified and shown to be indistinguishable from material obtained from a B. pertussis strain that contained only a single copy of the toxLys9Gly129 operon. Such strains are therefore suitable for large-scale, industrial production of an acellular whooping cough vaccine containing a genetically detoxified PT analog.     

Construction of Bordetella pertussis strains that overproduce genetically inactivated pertussis toxin.

Nontoxic analogs of pertussis toxin (PT), produced by in vitro mutagenesis of the tox operon, are immunogenic and protective against infection by Bordetella pertussis. The moderate levels of PT production by B. pertussis, however, make it the limiting antigen in the formulation of multicomponent, acellular, recombinant whooping cough vaccines. To increase production of the highly detoxified Lys9Gly129 PT analog by B. pertussis, additional copies of the mutated tox operon were integrated into the bacterial chromosome at the tox or fha locus by unmarked allelic exchange. Recombinant strains produced in this way secreted elevated levels of the PT analog proportional to gene dosage. The strains were stable during 10-liter fermentations, and yields of up to 80 mg of PT analog per liter were obtained under production-scale conditions. The nontoxic analog was purified and shown to be indistinguishable from material obtained from a B. pertussis strain that contained only a single copy of the toxLys9Gly129 operon. Such strains are therefore suitable for large-scale, industrial production of an acellular whooping cough vaccine containing a genetically detoxified PT analog.     

Virulence factors of Bordetella pertussis

Clearly, B. pertussis has evolved very elaborate mechanisms to maintain itself in the human host. Three different proteins (FHA, pertussis toxin and fimbriae) have been implicated in adherence. Furthermore, a number of toxins are produced (pertussis toxin, adenylate cyclase, dermonecrotic toxin, and tracheal cytotoxin) which destroy the clearance mechanisms of the respiratory tract, or suppress the immune response. There is evidence that B. pertussis may survive intracellularly, and the possibility that it is a facultative intracellular parasite should certainly be explored. The availability of a large number of cloned virulence genes, and a system to construct well defined mutants by allelic exchange (Stibbitz et al. 1986) will greatly facilitate the study of Bordetella virulence factors at the molecular level. It opens the possibility to construct avirulent strains, which are still able to colonize and stimulate the local immune response. Such strains may be used as live, oral vaccines, to present (heterologous) antigens to the mucosal immune system of the respiratory tract.