Streptococcus sanguinis Ecto-5'-nucleotidase modulates platelet aggregation.

Abstract

Streptococcus sanguinis, an oral commensal bacterium, is the leading cause of infective endocarditis (IE). In an animal model, the abilities of S. sanguinis to adhere to and activate platelets are correlated with the increased severity of IE. In response to S. sanguinis, platelet activation is associated with secretion of adenosine triphosphate (ATP) and adenosine diphosphate (ADP) from dense granules. The extracellular ADP is a potent platelet agonist and amplifies platelet aggregation induced by other pro-thrombotic agonists, whereas, the final product of hydrolysis of adenine nucleotides, adenosine, is a platelet aggregation antagonist. Here, we show that cell surface ecto-5'-nucleotidase (NT5E) of S. sanguinis can hydrolyze adenine nucleotides ATP to ADP, adenosine monophosphate (AMP) and finally adenosine. Therefore, we hypothesize that S. sanguinis ecto-5'-nucleotidase modulates platelet aggregation. A nt5e deletion mutant of S. sanguinis 133-79 showed significantly shorter lag time to onset of platelet aggregation than the wild-type strain (wt). However, the nt5e deletion mutant adhered to human platelets indistinguishably from the wild-type and complemented strains. By hydrolyzing the released ATP and ADP from dense granules of activated platelets, therefore, NT5E modulates S. sanguinis-induced platelet aggregation in vitro. In addition, strains of S. sanguinis showed different cell surface enzymatic activities for hydrolysis of adenine nucleotides, which may contribute to the determination of the platelet interactivity phenotypes. To further elucidate the mechanism, we distinguished the roles of ADP and adenosine receptors on streptococcal-platelet interactions using specific antagonists. We showed that the ADP receptors, P2Y1 and P2Y12, and the adenosine receptor A2a were all involved in S. sanguinis-induced platelet aggregation. Downstream of P2Y12, platelet activation involved two waves of Akt phosphorylation in response to S. sanguinis. NT5E also modulates platelet aggregation by indirectly signaling Rap1 activity. Through these pathways, S. sanguinis NT5E slows down platelet aggregation by removing ADP and generating adenosine. Using a rabbit endocarditis model, we found that in the absence of nt5e, the mass of the vegetations and recovered bacterial loads were greatly decreased, suggesting a contribution of NT5E to the virulence of S. sanguinis in vivo. Similar to the release of ADP, activated platelets secrete platelet microbicidal proteins (PMPs), which antagonize a broad range of pathogens. These data, therefore, indicate that NT5E-mediated inhibition of platelet aggregation might delay presentation of PMPs to infecting bacteria on heart valves. The delay would enable the infecting bacteria to colonize in the absence of this innate immune effector. Extracellular adenine nucleotides are also important signaling molecules that mediate both inflammatory and anti-inflammatory processes. By hydrolyzing ATP, a pro-inflammatory molecule, and generating adenosine, an immunosuppressive molecule, NT5E might inhibit phagocytic monocyte/macrophages associated with valvular vegetations, promoting the survival of infecting S. sanguinis. In conclusion, we now show for the first time that streptococcal NT5E modulates S. sanguinis-induced platelet aggregation and contributes to the virulence of streptococci in IE. These findings expand our knowledge of bacterial-host interactions and may suggest novel therapeutics for cardiovascular infectious diseases.