Acetylsalicylic acid inhibits intravascular coagulation during Staphylococcus aureus–induced sepsis in mice

Sepsis is a devastating condition resulting from a dysregulated immune response to infection, organ damage, shock, and death in 15% to 25% of cases.^1,2  During this response, coagulation factors interact with immune cells and platelets, resulting in the formation of immunothrombi, complex structures of fibrin, platelets, and leukocytes.³  Immunothrombi are associated with critical complications of sepsis, including disseminated intravascular coagulation⁴  and aberrant coagulation resulting in microthrombi, organ dysfunction, and patient mortality.^5,6 

In addition to clotting, platelets also respond directly to infection,^7-9  demonstrating a binary role of hemostasis and immunity. Platelets modulate host immunity, including the release of neutrophil extracellular traps (NETs).^3,10-12  Although primarily involved in pathogen capture and clearance, NETs also activate thrombin, amplifying clot formation.^13-15 

Given the central role of platelets in inflammation and thrombosis, we assessed the impact of an antiplatelet drug on infection-induced coagulopathy. Acetylsalicylic acid (ASA; aspirin) is the most widely used drug globally; 20% of American adults and 49% of those older than 65 years of age take daily ASA prophylactically for thrombotic disorders.^16,17  The current study utilizes intravital microscopy (IVM) to assess the impact of ASA on immunothrombosis in sepsis.

C57Bl/6 mice were purchased from The Jackson Laboratory. Experiments were approved by University of Calgary Animal Care Committee in compliance with the Canadian Council for Animal Care.

Staphylococcus aureus (USA300-2406) was cultured to mid-log phase in brain-heart infusion–chloramphenicol broth, washed, and resuspended in saline.¹⁵  A total of 2 × 10⁷ colony-forming units (CFUs) were administered IV or intraperitoneally 6 hours prior to analysis. For treatment, 12.5 μg/g (low dose in mice¹⁸ ) of ASA (Sigma-Aldrich) was administered intraperitoneally 1 hour preinfection or 3 hours postinfection.

Fluorophore-labeled anti-Ly6G (clone 1A8), anti-CD49b, anti-CD11b (clone M1/70), and goat–anti-mouse neutrophil elastase (M18) were used. Other markers were Sytox Green (Life Technologies), 5-FAM/QXL 520 fluorescence resonance energy transfer substrate for thrombin (SensoLyte 520 Thrombin Activity Assay; AnaSpec, Inc), 1.00-µm Fluoresbrite YG microspheres (Polysciences, Inc), dihydroethidium (ThermoFisher), and fluorescein isothiocyanate–dextran.

Following anesthesia (10 mg/kg xylazine/ketamine hydrochloride), the liver was prepared as previously described¹⁵  and imaging was performed on an inverted Leica SP8 microscope (Leica Microsystems).

Analysis of NET staining, thrombin cleavage, and vessel occlusion was performed as described.¹⁵  Platelet aggregates were categorized by aggregate size. Neutrophil counts, bead capture, and reactive oxygen species (ROS) generation were manually counted.

Counting beads were added to red blood cell–lysed whole blood, and neutrophils were enumerated by flow cytometry. Platelet-neutrophil aggregates were identified as CD41⁺ neutrophils (Ly6G⁺).

Plasma alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels were measured by Calgary Laboratory Services. Thrombin-antithrombin (TAT) complexes were measured in plasma by enzyme-linked immunosorbent assay (Abcam).

Organs were harvested, weighed, homogenized (GentleMACS Tissue Dissociator), filtered through a 100-μm strainer, centrifuged (5 minutes, 4°C, 2500 rpm), and resuspended in 1 mL of phosphate-buffered saline. Ten-fold dilutions were performed in chloramphenicol (1 mg/mL); 10 μL of each was plated on brain-heart-infusion agar plates and incubated at 37°C overnight.

Plasma cytokine levels were determined using a MILLIPLEX MAP Mouse Cytokine/Chemokine Panel as per the manufacturer’s instructions (Millipore).

Data are presented as mean plus or minus standard error of the mean (SEM). One-way analysis of variance (ANOVA) with a post hoc Tukey test, the Student t test, or 2-way ANOVA with a post hoc Bonferroni test were used for statistical analysis as appropriate. Significance was set at P < .05.

Control and ASA-treated mice were challenged with S aureus IV and imaged by IVM 6 hours postinfection. Infection significantly increased neutrophil recruitment and platelet aggregation in the liver. Neutrophil recruitment was significantly inhibited in ASA-pretreated mice whereas platelet aggregation was attenuated by either pretreatment or, to a lesser extent, by administration of ASA 3 hours postinfection (Figure 1A-C). These results are in line with previous studies that suggest neutrophil recruitment to the liver in response to bacterial infection is platelet independent.¹⁹ 

Given that neutrophil-platelet interactions stimulate NET production,^11,15  and studies have indicated that ASA-treated platelets have a reduced capacity to induce NETs from human neutrophils,¹²  we examined the effect of ASA treatment on NETs in vivo. NETs were visualized by either anti–neutrophil elastase (NE) antibody or Sytox Green labeling of extracellular DNA. Infected mice had significantly increased NE and Sytox Green staining (Figure 1D-E) compared with controls. Treatment with ASA significantly reduced staining for NETs.

ASA effects were not limited to liver. Although the number of circulating neutrophils in blood was not altered (Figure 1F), ASA treatment attenuated the observed decrease in circulating platelets following infection (Figure 1G). Platelet consumption in infected mice was correlated with increased platelet-neutrophil aggregates in blood; these aggregates were also reduced in ASA-treated animals (Figure 1H). Plasma levels of tumor necrosis factor α, MIP 1β, MCP-1, interleukin 10, MIP2, and IP-10 were elevated at 6 hours after infection, returning toward baseline by 24 hours. Levels of MIP-1α increased after 24 hours whereas RANTES elevated early and remained high through 24 hours. Administration of ASA 3 hours postinfection reduced plasma levels of tumor necrosis factor α and IP-10 (supplemental Figure 1, available on the Blood Web site).

NETs potentiate thrombin generation and drive intravascular coagulation.¹⁵  To understand this process in ASA-treated animals, thrombin generation was visualized using an activatable fluorescent peptide as previously described.¹⁵  IV-infected mice generated significantly more thrombin than controls (Figure 2A), and ASA pretreatment significantly reduced thrombin generation. These observations were supported by measuring plasma TAT complexes where ASA treatment significantly reduced TAT levels (Figure 2B).

Intravascular coagulation contributes to vessel occlusion and organ damage.¹⁵  Vascular occlusion was identified by introducing a fluorescent contrast agent (labeled dextran) into circulation. The percentage of occluded vessels in infected mice (IV) was significantly higher than control mice (Figure 2C); however, ASA treatment prevented this vascular occlusion. Moreover, ASA treatment decreased both ALT and AST (markers of organ dysfunction/damage) in the plasma of infected mice (Figure 2D-E). Histological examination of the liver identifies small, focal, necrotic patches that are prevented in ASA-treated animals (supplemental Figure 2). These results suggest that targeting the NET-platelet-thrombin axis can diminish intravascular coagulation and associated end-organ damage.

As platelets also contribute to the immune response, we next determined the impact of ASA on host immunity. Critically, ASA treatment did not alter bacterial clearance in the liver, lungs, or spleen (Figure 2F-H). This finding is in agreement with a recent report that NET formation and immunothrombi are dispensable for limiting dissemination of bloodstream bacterial infections.²⁰  It is known that platelets and Kupffer cells (KCs) work in concert to clear pathogens from the blood²¹ ; importantly, ASA did not impact the capacity of KCs to capture circulating microspheres from the bloodstream (Figure 2I). Production of ROS by phagocytes also contributes to pathogen destruction. Using dihydroethidium for direct assessment of ROS production in vivo,²²  we determined that ASA treatment did not alter ROS generation by KCs following phagocytosis of S aureus (Figure 2J). These results indicate that bacterial clearance from the bloodstream is not negatively impacted by ASA treatment.

As sepsis is often caused by an initially localized infection, such as peritonitis, we repeated experiments by injecting bacteria intraperitoneally. As with IV infection, ASA reduced neutrophil recruitment, platelet aggregation, and NET generation following intraperitoneal S aureus infection (supplemental Figure 3A-D). Additionally, ASA did not alter blood neutrophil counts, but partially reverted platelet depletion, reduced platelet-neutrophil aggregate formation, TAT complex generation, and vessel occlusion induced by intraperitoneal infection (supplemental Figure 3E-I). Most importantly, ASA did not alter bacterial clearance in liver, lungs, or spleen of infected animals (supplemental Figure 3J-K). More studies considering other models of localized infection, such as pneumonia, and with different bacterial species are needed to support the potential efficacy of ASA in clinical sepsis.

ASA is the most widely used drug in the world. Millions of patients take a low dose of ASA daily as a prophylactic therapy for thrombotic disorders. Our findings suggest that the presence of a low dose of ASA prior to bacterial infection may offer some protection to the patient without compromising the host immune response. Moreover, our findings support the potential use of ASA as a therapeutic for coagulopathy in sepsis and provide unique insights into the specific cellular mechanisms altered by ASA administration in the context of infection. As ASA targets both cyclooxygenase 1 (COX-1) and COX-2, the current study does not positively identify platelet COX-1 as the specific therapeutic target, though we believe this model is reflective of the overall physiological response of patients treated with ASA. Importantly, our results are in agreement with clinical studies showing that low-dose ASA usage reduces hospital mortality in sepsis,^23-25  potentially providing a therapeutic avenue to improve patient outcomes in this critical area.

The authors thank Braedon McDonald for helpful insights and comments regarding model development.

This work was supported by grants from the Canadian Foundation for Innovation, Alberta Innovates and Advanced Education, the Heart and Stroke Foundation of Canada (HSFC), and the Natural Sciences and Engineering Research Council (NSERC) (C.N.J.). C.N.J. was supported by the Canada Research Chairs Program. A.C. was supported by the Beverley Phillips Rising Star Postdoctoral Fellowship and the University of Calgary, Cumming School of Medicine Postdoctoral Scholar Program. R.P.D. was funded by the Canadian Liver Foundation and the University of Calgary Faculty of Graduate Studies Indigenous Graduate Award.

Contribution: A.C. performed experiments, analyzed the data, prepared figures, and wrote the manuscript; R.P.D., H.G., and M.W.L. performed experiments and contributed to the generation of the manuscript; and C.N.J. conceived, designed, and supervised the research, analyzed data, and wrote the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Craig N. Jenne, Department of Microbiology, Immunology, and Infectious Diseases, University of Calgary, 3330 Hospital Dr NW, Calgary, AB T2N 4N1, Canada; e-mail: cnjenne@ucalgary.ca.