Bordetella pertussis is a highly contagious, gram-negative bacteria associated with the respiratory disease known as whooping cough. The hallmark symptoms of a whooping cough infection include paroxysmal coughing with whooping and post-tussive vomiting [1]. The persistent cough can last from weeks to months after infection, and infection is often associated with pneumoniae, otitis media, seizures, encephalopathy, and hemorrhages [2]. B. pertussis infection is usually less severe in older populations, with infection manifesting as just a persistent cough with milder systemic symptoms [3]. Before widespread vaccination, B. pertussis was a leading cause of infant death with mortality in children under 10 as high as 13% [4]. However, despite a widely available vaccine, B. pertussis is still a threat to public health, with infection rates from 1-7% even in highly vaccinated populations [5,6]. B. pertussis’ unique staying power within vaccinated populations necessitates other measures of prevention and treatment to be more robust.
It has previously been shown that B. pertussis forms biofilms, both on abiotic surfaces and in vitro during infection [7,8,9,10,11]. A biofilm is formed when a consortium of bacteria aggregates together on a surface, creating a thin film, providing protection and more efficient cell to cell communication. Biofilm structures lower the efficacy of antibiotics by decreasing permeability to reach inner constituent cells and could result in an overuse of antibiotics, which is linked to an increase of antibiotic-resistant strains among human pathogens [12]. Finding alternative ways to treat disease caused by bacteria would help mitigate the global threat that resistant strains already pose to public health.
Comprising the B. pertussis biofilm matrix are a number of polysaccharides, extracellular DNA (eDNA), and proteins. While biofilm regulation in B. pertussis is not well studied, certain virulence factors have been shown to be associated with biofilm robustness. Filamentous haemagglutinin (FHA) is a surface expressed adhesion that has been shown to be required for robust biofilm formation in B. pertussis. Previous studies by Serra et al. showed anti-FHA antibodies were able to block biofilm formation, and FHA knockout mutants produced less biofilm in vivo compared to the wild type counterparts [10]. Adenylate cyclase toxin (ACT), a virulence factor secreted by B. pertussis, has also been shown to affect biofilm formation. The more well known function of ACT is as a toxin that actively targets cells by binding the αMß2 integrin [13]. However, ACT and FHA have been shown to directly interact [14], which, in tandem with evidence showing that the absence of ACT produces more robust biofilms in a related Bordetella species [15], suggests that ACT may play a role in B. pertussis biofilm regulation. The study done by Hoffman et al. delves deeper into the relationship between FHA and ACT in B. pertussis, and how this interaction can regulate biofilm formation [16].
Hoffman et al. were able to replicate the results obtained by Irie et al., showing ACT knockouts in Bordetella bronchiseptica formed more robust biofilms, in B. pertussis, which confirmed that this is a feature conserved between Bordetella species. ACT also directly inhibits biofilm formation in a concentration dependent manner. This was shown by growing the ACT knockouts in exogenous ACT and measuring biofilm formation. ACT is a relatively large protein with multiple domains, including the adenylate cyclase domain, a hydrophobic region, a repeat region, and a secretion signal. The researchers used strains of B. pertussis that truncated and deleted these portions of ACT to determine which domains are essential for the biofilm inhibition phenotype. The following six mutant proteins were tested: two mutants lacking the hydrophobic region, a mutant lacking the repeat region, a mutant lacking the AC domain, a mutant only containing the AC domain, and an enzymatically inactive form of ACT [17]. All ACT mutants resulted in biofilm inhibition except the ACT lacking the AC domain. This suggested that the AC domain was responsible for this form of biofilm regulation, and, interestingly, the AC domain’s enzymatic activity is not involved because the mutant lacking enzymatic activity was still able to inhibit biofilm formation to the same extent as the wild type ACT. To further confirm that the AC domain in particular inhibits biofilm formation, Hoffman et al. showed that a mutant B. pertussis that only lacks the AC domain shares the same robust biofilm growth phenotype that a full ACT knockout mutant has.
Not only does the AC domain inhibit biofilm from forming but adding the AC domain to already formed wild type B. pertussis biofilm will disperse the biofilm. This was shown by tracking biofilm biomass through its formation for 96 hours. The biofilm grew steadily for 72 hours before the exogenous AC domain was added, however, there was a significant drop in biomass observed after. To even further characterize how the AC domain is a regulator for biofilm formation, calmodulin (CaM), which is known to bind the AC domain, and an anti-ACT antibody were shown to inhibit the regulatory affect that ACT and the AC domain have. This suggested that the regulatory affect was being caused by a physical disruption of binding between B. pertussis cells. The reason FHA in particular was brought into the fold was because of previous findings that suggested ACT only remained surface attached on cells expressing FHA [18], which suggests the possibility that ACT could bind to FHA, disrupting FHA binding to form biofilms.
Hoffman et al. confirmed interactions between FHA and the AC domain using SPR kinetic analysis and ELISA binding assays. The SPR analysis once again confirmed the AC domain is the binding site by showing the binding between the AC domain and FHA is interrupted in the presence of high amounts of CaM. SPR analysis was also done between the AC domain and FHA44, a truncated form of FHA that lacks the C-terminal domain. To characterize the binding to the C-terminal domain, an ELISA was performed using anti-FHA antibodies that bind to the mature C-terminal domain (MCD). The ELISAs demonstrated that the AC domain interrupted the binding of the antibody, which confirms that the AC domain binds the MCD. Hoffman et al. were able to further characterize the binding between ACT and FHA using B. pertussis mutants, one of which produces truncated mutants lacking the MCD, and one that leaves the MCD unfolded. In both of these mutants, exogenous ACT was unable to inhibit biofilm formation, leading to the conclusion that the MCD in FHA must be present and in the correct confirmation in order for ACT to bind. Anti-MCD antibodies were also sufficient to inhibit biofilm formation. All of this evidence leads to the authors suggested model of B. pertussis biofilm inhibition: the AC domain on ACT binds FHA at the MCD, which blocks FHA-FHA binding and other surface interactions, resulting in biofilm dispersion.
How does this model of biofilm inhibition relate to the current understanding of B. pertussis pathogenesis? ACT is primarily produced during active infection, in which it suppresses the host immune system and has a cytotoxic effect upon entering the host cells. The authors suggest that, in ideal conditions, B. pertussis takes an offensive stance, releasing themselves into a planktonic state while actively infecting and killing surrounding host cells; then, in less favorable conditions, ACT production lessens, and forming biofilms may provide protection and nutrients while being less virulent.
This study presents a novel function for ACT, a protein that had previously been considered exclusively a host-directed protein bacterial toxin, for self directed biofilm regulation. It is still not clear how biofilm formation is involved in the pathogenesis of B. pertussis; however, this direct link between biofilm formation and ACT suggests that biofilm formation and dispersion is relevant to the stages of virulence and non-virulence in B. pertussis. This raises potential questions about how other virulence factors might play a role in biofilm regulation, or if it could be the other way around with the presence of biofilm affecting virulence factor expression. Understanding how B. pertussis controls its virulence could provide insight for treatment, potentially improve prevention efforts through vaccination. ACT is one of many protein candidates to be included in the acellular vaccine for B. pertussis in an effort to make the vaccine more effective. Its interaction with FHA, which is already included in the acellular cocktail, is especially important to understand for that endeavor. This revelation of a novel function for ACT once again illuminates just how much we don’t know about B. pertussis. The relationship between biofilm regulation and virulence in B. pertussis is one that needs to be explored further.