Platelet factor 4 (PF4) modulates endothelial cell antimicrobial activity to enhance bacterial clearance and improve sepsis outcomes

Abstract

Abstract Introduction: Upon sensing pathogens, platelets (plt) release high concentrations of the positively charged chemokine PF4 from their ⍺-granules. PF4 rapidly binds to endothelial chemokine receptor, CXCR3B, and forms electrostatic interactions with glycosaminoglycans (GAGs) on the endothelial cell (EC) glycocalyx. PF4 similarly forms complexes with anionic polymers in the bacterial cell wall, bridging bacteria to ECs. This study aims to investigate mechanisms by which ECs and plt collaborate to promote bacterial clearance, paving the way for translational studies on PF4-based therapeutics for sepsis. Methods: PF4-mediated bacterial aggregation: Escherichia (E) coli (K-12 strain) (1x108 colony forming unit (CFU)/mL) were incubated with PF4 (0-100µg/mL). Bacterial cluster size was quantified with scanning electron microscope (SEM). PF4 effects on in vitro EC-mediated bacterial clearance: Human umbilical vein ECs (HUVECs) were incubated with TNF⍺ (20ng/mL) to induce inflammation. A subset was then pre-treated with AMG-487 (20µg/mL), Dynasore (100µM), Pitstop (30µM), heparinases (1U/mL), or hydroxychloroquine (HCQ; 50µM). Prior to exposure to E coli at a multiplicity of infection of 10 ± PF4 (25µg/mL) for 1h at 37°C. ECs were then washed with PBS to remove unbound bacteria and incubated for up to 5h. HUVEC culture supernatant was collected every hour to quantify lactate dehydrogenase (LDH) release. Cells were treated with 1% penicillin-streptomycin for 1h at 37°C to kill extracellular bacteria and lysed for internalized CFU enumeration. PF4 effects on in vivo bacterial clearance: Twelve-to-16-week-old WT and PF4-/- mice on a C57BL/6J background received intravenous (I.V) injection of cecal slurry (CS; 20mg) ± PF4 (100µg). Sepsis severity and mortality were assessed up to 96h. Blood and organ homogenates were subjected to CFU enumeration at 24h. Visualizing PF4-mediated bacterial clearance in vivo: WT and PF4-/-mice received I.V infusions of GFP-expressing E coli (1x109 CFU) and underwent lung and liver intravital microscopy. Lung and liver CFUs were quantified at 15min and 2h. A subset of mice received intraperitoneal injections of clodronate liposomes to deplete Kupffer cells (KC) 3 days prior to imaging. Results: E. coli cluster size rose from 3.09µm2 at baseline to 4.06µm2 and 5.24µm2with 20 and 100µg/mL of PF4, respectively. Two hours post E coli exposure, HUVECs incubated with bacteria and PF4 exhibited higher intracellular CFUs, indicating enhanced uptake. LDH levels did not differ between groups suggesting increased bacterial uptake was not associated with EC injury. Surprisingly, from 3 to 5h, intracellular CFUs declined in PF4-treated group, falling below levels observed in untreated cells, suggesting accelerated intracellular bacterial clearance. CXCR3 inhibition, removing GAGs, and blocking endocytosis decreased E coli internalization. Lysosomal alkalinization resulted in a marked increase in intracellular E coli, suggesting that PF4 promotes bacterial uptake through interaction with endothelial CXCR3 and GAGs, leading to clathrin-mediated endocytosis, followed by trafficking to lysosomes for killing. CS-challenged PF4-/-mice had reduced survival compared to WT animals (100% vs 33% mortality). All of WT mice and 62.5% of PF4-/- mice that received CS+PF4 were protected from mortality. In both genotypes, PF4 significantly reduced CFUs in blood, lung, and liver compared to genotype-matched mice given CS alone. Intravital imaging studies revealed that PF4-/-mice exhibited higher E coli count, increased residual CFUs, and more pronounced plt aggregation in the pulmonary microvasculature. In the liver microcirculation of both genotypes, E coli was similarly sequestered by KC, suggesting that PF4 does not influence KC-mediated bacterial trapping despite higher residual CFUs in PF4-/- mice. Post KC-depletion, PF4-/- mice exhibitedlarger E coli aggregates and increased plt accumulation inthe hepatic sinusoids as compared to WT mice, suggesting that PF4 facilitates bacterial clearance in liver through a mechanism partly independent of KC phagocytic activity. Conclusion: This study uncovers a critical role for plt PF4 in enhancing host defense against bloodstream infections by promoting EC uptake and clearance of bacteria. Beyond expanding the known functions of PF4, our findings support a broader concept that ECs, long recognized as a passive barrier, actively contribute to innate immune defense during systemic infection.