Search

Pathogenesis by Bacillus anthracis requires coordination between two distinct activities: plasmid-encoded virulence factor expression (which protects vegetative cells from immune surveillance during outgrowth and replication) and chromosomally encoded sporulation (required only during the final stages of infection). Sporulation is regulated by at least five sensor histidine kinases that are activated in response to various environmental cues. One of these kinases, BA2291, harbors a sensor domain that has ∼35% sequence identity with two plasmid proteins, pXO1-118 and pXO2-61. Because overexpression of pXO2-61 (or pXO1-118) inhibits sporulation of B. anthracis in a BA2291-dependent manner, and pXO2-61 expression is strongly up-regulated by the major virulence gene regulator, AtxA, it was suggested that their function is to titrate out an environmental signal that would otherwise promote untimely sporulation. To explore this hypothesis, we determined crystal structures of both plasmid-encoded proteins. We found that they adopt a dimeric globin fold but, most unusually, do not bind heme. Instead, they house a hydrophobic tunnel and hydrophilic chamber that are occupied by fatty acid, which engages a conserved arginine and chloride ion via its carboxyl head group. In vivo, these domains may therefore recognize changes in fatty acid synthesis, chloride ion concentration, and/or pH. Structure-based comparisons with BA2291 suggest that it binds ligand and dimerizes in an analogous fashion, consistent with the titration hypothesis. Analysis of newly sequenced bacterial genomes points to the existence of a much broader family of non-heme, globin-based sensor domains, with related but distinct functionalities, that may have evolved from an ancestral heme-linked globin.

B. anthracis is the causative agent of anthrax. Pathogenesis is primarily mediated through the exotoxins lethal factor and edema factor, which bind protective antigen (PA) to gain entry into the host cell. The current anthrax vaccine (AVA, Biothrax) consists of aluminum-adsorbed cell-free filtrates of unencapsulated B. anthracis, wherein PA is thought to be the principle target of neutralization. In this study, we evaluated the efficacy of the natural adjuvant, C3d, versus alum in eliciting an anti-PA humoral response and found that C3d conjugation to PA and emulsion in incomplete Freund's adjuvant (IFA) imparted superior protection from anthrax challenge relative to PA in IFA or PA adsorbed to alum. Relative to alum-PA, immunization of mice with C3d-PA/IFA augmented both the onset and sustained production of PA-specific antibodies, including neutralizing antibodies to the receptor-binding portion (domain 4) of PA. C3d-PA/IFA was efficacious when administered either i.p. or s.c., and in adolescent mice lacking a fully mature B cell compartment. Induction of PA-specific antibodies by C3d-PA/IFA correlated with increased efficiency of germinal center formation and plasma cell generation. Importantly, C3d-PA immunization effectively protected mice from intranasal challenge with B. anthracis spores, and was approximately 10-fold more effective than alum-PA immunization or PA/IFA based on dose challenge. These data suggest that incorporation of C3d as an adjuvant may overcome shortcomings of the currently licensed aluminum-based vaccine, and may confer protection in the early days following acute anthrax exposure.

Anthrax toxin consists of the proteins protective antigen (PA), lethal factor (LF) and oedema factor (EF). The first step of toxin entry into host cells is the recognition by PA of a receptor on the surface of the target cell. Subsequent cleavage of receptor-bound PA enables EF and LF to bind and form a heptameric PA63 pre-pore, which triggers endocytosis. Upon acidification of the endosome, PA63 forms a pore that inserts into the membrane and translocates EF and LF into the cytosol. Two closely related host cell receptors, TEM8 and CMG2, have been identified. Both bind to PA with high affinity and are capable of mediating toxicity. Here, we report the crystal structure of the PA-CMG2 complex at 2.5 A resolution. The structure reveals an extensive receptor-pathogen interaction surface mimicking the non-pathogenic recognition of the extracellular matrix by integrins. The binding surface is closely conserved in the two receptors and across species, but is quite different in the integrin domains, explaining the specificity of the interaction. CMG2 engages two domains of PA, and modelling of the receptor-bound PA63 heptamer suggests that the receptor acts as a pH-sensitive brace to ensure accurate and timely membrane insertion. The structure provides new leads for the discovery of anthrax anti-toxins, and should aid the design of cancer therapeutics.