Interactions between Bordetella pertussis and the complement inhibitor factor H

Abstract

Bordetella pertussis causes whooping cough in humans, a highly contagious disease of the upper respiratory tract. An increase in cases of whooping cough in adolescents and adults in many countries has been reported, despite high immunization rates in children. To efficiently colonize the host the bacteria have to resist complement, the first defence line of innate immunity. B. pertussis has previously been shown to bind the classical pathway inhibitors C4b-binding protein and C1-inhibitor being thereby able to escape the classical pathway of complement. In this study recent clinical isolates of B. pertussis and B. parapertussis were found to survive alternative pathway attack in fresh non-immune serum better than the reference B. pertussis strain, Tohama I. By using adsorption assays, flow cytometry and a radioligand binding assay we observed that both B. pertussis and B. parapertussis bound the alternative pathway inhibitor factor H (FH) from normal human serum. The surface attached FH maintained its complement regulatory activity and promoted factor I-mediated cleavage of C3b. The main binding region was located to the C-terminal part of FH, into short consensus repeat domains 19–20. In contrast, the avian pathogen B. avium did not bind FH and was sensitive to the alternative pathway of human complement. In conclusion, the human pathogens B. pertussis and B. parapertussis are able to evade the alternative complement pathway by surface acquisition of the host complement regulator FH.

Introduction

Bordetella pertussis is a Gram-negative pathogen that colonizes ciliated respiratory epithelial surfaces of humans. Jules Bordet, a Belgian doctor and scientist is credited for the discovery of the causative agent of whooping cough as well as the complement (C) system that he originally called as alexin. Despite a wide coverage with vaccines, whooping cough is one of the major vaccine-preventable diseases in the world (He and Mertsola, 2008). Its reemergence among adolescents and adults has been reported in several countries (Wood and McIntyre, 2008). An infection with B. pertussis is initiated by its attachment to ciliated epithelial cells by several adhesins (Smith et al., 2001). After apparently escaping the host's immune defence it causes local tissue damage and systemic manifestations via its toxins. The pathogenesis of B. pertussis infection is complex. While many potential virulence factors have been described the mechanisms of immune escape and basis of protective immunity against B. pertussis are still incompletely understood.

In addition to B. pertussis, the genus Bordetella comprises eight other species. B. parapertussis can cause milder pertussis-like symptoms in human. B. bronchiseptica (Goodnow, 1980), B. holmesii (Weyant et al., 1995), B. hinzii (Cookson et al., 1994), B. trematum (Vandamme et al., 1996) and B. ansorpii (Ko et al., 2005) have been isolated from humans, although only rarely. B. avium is thought to be strictly an animal pathogen that causes tracheobronchitis in birds, but recently it has been isolated from humans with respiratory disease (Harrington et al., 2009). B. petrii is the only environmental species found among the Bordetella, whose genome sequence has recently been published (Gross et al., 2008).

In order to survive in the host, bacteria must be able to evade killing and clearance by several host defence mechanisms, including the C system. The C system consists of approximately 40 soluble or membrane-bound proteins (Walport, 2001). It can be activated through three different pathways: the classical, lectin or the alternative pathway. The classical pathway (CP) is primarily triggered by surface-bound antibodies or direct binding of its first component C1q. The lectin pathway is activated by mannan binding lectin or ficolins that bind to arrays of sugar residues or acetylated molecules on targets. The alternative pathway (AP) is activated at a low rate in plasma as its main component, C3, is continuously hydrolyzed to a form referred to as C3(H₂O). C3(H₂O), unlike C3, can bind factor B to generate the initial C3 convertase enzyme after cleavage of B to Bb by factor D. The actual C3 convertase, C3bBb, constitutes the basis of AP amplification by cleaving further C3 molecules to C3b, which subsequently become subunits of new C3 cleaving enzymes.

Specific cell surface molecules that regulate the C system protect viable human tissues from C attack. In addition, soluble regulators, such as C4b-binding protein (C4bp) and C1 inhibitor of the CP, control C activation in the fluid phase. The main regulator of the AP is the 150 kDa soluble protein, factor H (FH). FH is an elongated protein consisting of 20 domains, called short consensus repeats (SCRs). Factor H-like protein (FHL-1) is a 42 kDa protein that has the first seven SCRs of FH and an additional extension of four amino acid residues at its C-terminus (Zipfel et al., 1999). FH and FHL-1 regulate C activity by binding to C3b and thereby accelerating the decay of the AP convertase, C3bBb (Wu et al., 2009). In addition, FH blocks the interaction of C3b with factor B and acts as a cofactor in the factor I-mediated cleavage of C3b. Thus, FH prevents C3 activation in the fluid phase and on host cell surfaces to which it can bind if they contain polyanions, like glycosaminoglycans (Meri and Pangburn, 1990).

Both FH and FHL-1 are members of the factor H protein family, which also includes five factor H-related proteins (Jozsi et al., 2005). Factor H-related protein 1 (FHR-1) is composed of five SCRs and is present in plasma in two forms, 37 kDa FHR-1α and the 43 kDa FHR-1β (Skerka et al., 1991). The function of FHR-1 is still unresolved, although SCRs 3–5 show a high homology to SCRs 18–20 of FH, suggesting similar and conserved function(s) for these regions of molecules (Zipfel et al., 2002). Both FH and FHR-1 bind to C3d and polyanions via these sites (Jokiranta et al., 2000).

A variety of microbes have evolved mechanisms to escape C attack (Lambris et al., 2008). Among these are pathogens that facilitate the recruitment of FH and/or FHL-1 such as Neisseria meningitidis (Madico et al., 2006), N. gonorrhoeae (Ram et al., 1998), Streptococcus pyogenes (Horstmann et al., 1988), S. pneumoniae (Jarva et al., 2003), Fusobacterium necrophorum (Friberg et al., 2008), Borrelia burgdorferi (Hellwage et al., 2001), the fungal pathogen Candida albicans (Meri et al., 2002a) and microfilariae of the nematode Onchocerca volvulus (Meri et al., 2002b). Staphylococcus aureus has multiple mechanisms to resist C attack, of which the binding of a secreted protein SCIN to C3bBb and blocking of its activity is one of the best characterized ones (Rooijakkers et al., 2009).

In this study we demonstrate binding of the complement regulator FH by B. pertussis and the closely related B. parapertussis. Binding occurred when B. pertussis and B. parapertussis were incubated with the purified factor H protein or in whole human plasma. Bound FH remained functionally active and inhibited complement activation on the surface of the bacteria, thus contributing to their serum resistance. The major binding site within FH to Bordetella was located to the C-terminal domains 19–20. Furthermore, we observed that strains recently isolated from patients survived better in normal human serum (NHS) than the reference strain Tohama I, a strain used for production of acellular vaccine against whooping cough, or B. avium, which is primarily an avian pathogen.