PspA is a crucial component in pneumococcal virulence, contributing to immune evasion during different disease stages 25. Its surface exposure allows it to interact with different host components, protecting the bacterium. PspA can bind to CRP and complement proteins to prevent C3b deposition on the pneumococcal surface, limiting complement-mediated opsonization 26,27. It can also function as an adhesin, binding to GAPDH on dying lung cells during infection, which contribute to bacterial dissemination in the lungs 28.
PspA exhibits some structural and serological variability, being classified in two major families and four more prevalent clades 29. Since the N-terminal region includes the domains responsible for PspA activities – which correspond to the most variable regions of the protein – it is important to compare the effect of such sequence variations when investigating PspA’s activities.
One described virulence mechanism for PspA is the ability to prevent the bactericidal activity of apolactoferrin (the iron-free form of human lactoferrin) against pneumococci 12. The present study expands the previous data, by demonstrating that PspA can also interfere with indolicidin, another important cationic peptide with broad antimicrobial activity.
Previous studies have shown that indolicidin and synthetic peptides derived from it can reduce pneumococcal viability in vitro 30, while pneumococcal capsular polysaccharides may interfere with indolicidin activity 31,32. However, the role of PspA on pneumococcal interactions with indolicidin has not been explored.
To determine PspA’s role in the indolicidin action, we compared in vitro killing of wild-type and pspA-negative pneumococcal strains by indolicidin. The protective effect of PspA against indolicidin killing was partially dependent on the pneumococcal strain; the loss of PspA on D39, a serotype 2 strain, had a great impact on the bacterium resistance to the AMP, with the mutant displaying a marked reduction in viability at all concentrations of indolicidin tested. For the type 4 strain, TIGR4, on the other hand, the loss of PspA led to a significant decrease in viability only at a high concentration of indolicidin.
Since the polysaccharide capsules are known to interfere with the bactericidal activity of indolicidin over pneumococci, we hypothesized that the capsule differences are the main factor responsible for the observed variations between the mutant strains derived from D39 and TIGR4. However, other factors including surface proteins and variations in the bacterial envelope may also play a role in this mechanism.
PspA is the major component in a family of proteins anchored to the choline residues on the bacterial cell wall through non-covalent bounds. Choline-chloride wash can remove choline binding proteins. Therefore, we tested if PspA removal by choline-chloride (CC) wash would have a similar impact on bacterial resistance to the AMP as the genetic mutation. For the D39 strain, CC-wash rendered the bacterium more susceptible to indolicidin at lower concentrations, although the effect was much less pronounced when compared with the genetic mutants. Pneumococci also expresses other CBPs such as PspC, which could affect killing by the AMP. Therefore, we tested if this mutant would have lower resistance to indolicidin; no differences were observed between the mutant and the parental strain, indicating that PspC does not play a role in pneumococcal resistance to indolicidin. A similar result was obtained using lactoferrin 12, indicating that the main CBP involved in resistance to AMPs is PspA.
The combined effects of capsule and PspA were tested using Rx1, which does not express capsule, and its isogenic PspA-negative mutant. The double PspA and capsule mutant was more susceptible to indolicidin killing than Rx1, with a similar effect obtained through CBPs removal by CC-wash. Furthermore, CBPs removal of the D39 derived capsule negative strain, AM1000, rendered the bacteria much more susceptible to the AMP. The data suggests that capsule and PspA have an additive contribution to pneumococcal resistance to the AMP.
Next, we investigated if soluble rPspA fragments would interfere with indolicidin’s bactericidal efficacy. Pneumococcal killing by indolicidin was strongly reduced in the presence of free PspA fragments. The demonstration that PspA fragments including the exposed region of the protein, can rescue the bacteria from killing by indolicidin indicates that the N-terminal part of the molecule is responsible for interacting with the AMP. It also demonstrates that PspA does not need to be attached to the bacterial surface to exert its protective effect against indolicidin. Similar results have been obtained when our group and others analyzed the impact of PspA on lactoferrin 12,13. Moreover, free rPspA was equally effective at reducing killing of the PspA-negative mutant strain, further suggesting the ability of free PspA to prevent bacterial death by indolicidin.
PspA fragments of families 1 and 2 prevented killing by indolicidin, suggesting that the regions responsible for interacting with the AMP are conserved within PspAs of different families. A previous study has identified PspA fragments that could bind to lactoferricins and prevent their lytic action on pneumococci. One fragment, SM-1, demonstrated high efficacy in protecting the bacteria from being killed by lactoferricins 12. This fragment is present in both rPspAs used in the current work. By contrast, one study found that variations in pneumococcal susceptibility to apolactoferrin were strongly associated with the PspA type expressed by each strain, with little influence of the capsule type 33. Interestingly, the ability of a particular PspA to bind to lactoferrin positively correlated with resistance to killing by apolactoferrin.
The ability of rPspAs to block indolicidin also suggests that the two proteins may interact directly. Molecular docking studies have identified PspA as a potential target for indolicidin binding 30.
Such binding has also been shown to occur between PspA and lactoferrin. Binding experiments using different recombinant fragments of PspA determined a region in the C-terminal half of the α-helical domain of the protein as the primary binding site for lactoferrin 9. In a later study, the N-terminal domain of PspA has been co-crystalized with the N-lobe of lactoferrin, and the specific interaction moieties have been mapped in both molecules 34. Both studies demonstrated that the negatively charged surface-exposed helices in PspA are the main responsible for lactoferricin binding.
Pre-opsonization of pneumococci with anti-PspA antibodies enhanced the bactericidal action of indolicidin on pneumococci, further reinforcing PspA’s ability to interfere with indolicidin action. The effect was observed with bacteria of different serotypes, suggesting that antibodies against PspA can limit the protein’s ability to interfere with the bactericidal action of indolicidin regardless of the capsule type. Antibodies against PspAs of two different families exhibited a similar effect at promoting bacterial killing by indolicidin, suggesting that polyclonal mouse antisera against PspA can conceal the protein’s domains that protect the bacterium from indolicidin. This result follows previous studies demonstrating that serum from mice immunized with PspAs promotes killing of pneumococci by lactoferrin 12,13.
Based on the increased sensitivity of the pspA-negative strain to indolicidin and the protective role of rPspAs in this mechanism, we hypothesized that the ability of anti-PspA antibodies to enhance the bactericidal activity of indolicidin would involve blocking PspA domains responsible for interacting with the AMP; this would make indolicidin available to penetrate the bacterial membrane, leading to pneumococcal death. A direct interaction between PspA and indolicidin was observed by mass spectrometry, confirming that PspA binds chemically to indolicidin. This result corroborates our hypothesis and in silico analysis described in the literature and provides a possible mechanism of PspA mediated inhibition of killing by indolicidin.
One limitation of this study is the absence of complementation mutants for PspA. However, the mutant strains used in this work have been extensively characterized in previous publications, with no evidence of polar effects 9,12,27,35. Furthermore, we have confirmed that addition of rPspA increases resistance of the mutant strain to indolicidin and augments resistance in the wild-type bacteria, supporting its role in preventing AMP-mediated death.
Due to its high prevalence and contribution to bacterial virulence, PspA has been evaluated as a vaccine candidate in different infection models, with promising results 36. The present data further support the inclusion of PspA in future pneumococcal vaccines, with the induction of antibodies that can enhance pneumococcal clearance by AMPs.