Thrombocytopenia Impairs Host Defense Against Burkholderia pseudomallei (Melioidosis)

Abstract

Background: Infection with the gram-negative bacillus Burkholderia pseudomallei (melioidosis) is an important cause of pneumosepsis in Southeast Asia and has a mortality of up to 40%. We aimed to assess the role of platelets in the host response against B. pseudomallei infection.

Methods: Association between platelet counts and mortality was determined in 1160 patients with culture-proven melioidosis. Mice treated with (low- or high-dose) platelet-depleting antibody were inoculated intranasally with B. pseudomallei and killed. Additional studies using functional glycoprotein Ibα-deficient mice were conducted.

Results: Thrombocytopenia was present in 31% of patients at admission and predicted mortality in melioidosis patients even after adjustment for confounders. In our murine-melioidosis model, platelet counts decreased, and mice treated with a platelet-depleting antibody showed enhanced mortality and higher bacterial loads compared to mice with normal platelet counts. Low platelet counts had a modest impact on early-pulmonary neutrophil influx. Reminiscent of their role in hemostasis, platelet depletion impaired vascular integrity, resulting in early lung bleeding. Glycoprotein Ibα-deficient mice had reduced platelet counts during B. pseudomallei infection together with an impaired local host defense in the lung.

Conclusions: Thrombocytopenia predicts mortality in melioidosis patients and, during experimental melioidosis, platelets play a protective role in both innate immunity and vascular integrity.

Figure 1.

Thrombocytopenia is associated with increased mortality in melioidosis patients. Kaplan–Meier survival curves of 1160 patients with melioidosis stratified according to platelet counts on admission. Patient were stratified in groups with low platelet count (<100 × 10⁹/L) in light gray, intermediate-low platelet count (100–149 × 10⁹/L) in gray, or normal platelet count in black (≥150 × 10⁹/L). P < .001 for comparison between 3 groups for survival analysis.

Figure 2.

Experimental melioidosis is associated with thrombocytopenia and effect of anti-GPIbα on platelet counts. Mice were treated with anti-GPIbα (platelet depletion) or immunoglobulin G (IgG) control (both 0.4 µg/g) and infected with Burkholderia pseudomallei via the airway and killed after 24, 48, or 72 hours or killed uninfected (t = 0 hours). A, Platelet counts before (t = 0 hours) and after (t = 24, 48, or 72 hours) infection. B, Representative log-scale scatterplots of CD61-positive platelets in blood of uninfected control and anti–GPIbα-treated mice. Data are represented as bars (median with interquartile range). n = 8 mice per group. *P < .05, ***P < .001 vs IgG control or vs uninfected mice. Abbreviations: FSC-H, forward scatter height; SSC-H, side scatter height.

Figure 3.

Thrombocytopenia impairs survival and enhances bacterial growth during Burkholderia pseudomallei pneumonia-derived sepsis. Mice were treated with anti-GPIbα (platelet depletion) or immunoglobulin G (IgG) control (both 0.4 µg/g) and infected with B. pseudomallei via the airway and killed after 24, 48, or 72 hours or were observed in a survival experiment. Survival (A) and clinical observation score (B). C, Bacterial quantification of indicated organs. Data are expressed as box-and-whisker plots depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. n = 20 per group for survival experiment and n = 8 mice per group for bacterial quantification. *P < .05, **P < .01, ***P < .001 vs IgG control. Abbreviations: BALF, bronchoalveolar lavage; CFU, colony-forming units.

Figure 4.

Platelet depletion affects early neutrophil recruitment but not neutrophil extracellular trap formation. Mice were treated with anti-GPIbα (platelet depletion) or immunoglobulin G (IgG) control (both 0.4 µg/g) and infected with Burkholderia pseudomallei via the airway and killed after 24, 48, or 72 hours or killed uninfected. A, Representative images of Ly6G staining lung sections. B, Neutrophil influx; measured by Ly6G staining quantification, myeloperoxidase levels in lung, and neutrophil counts in bronchoalveolar lavage fluid (BALF). C, Cell-free DNA and citrullinated histone 3 levels in BALF. Data are expressed as box-and-whisker plots depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. n = 8 mice per group. *P < .05 vs IgG control. Abbreviations: BALF, bronchoalveolar lavage fluid; BPS, B. pseudomallei; cfDNA, cell-free DNA; CitH3, citrullinated histone 3; IgG, immunoglobulin G; MPO, myeloperoxidase.

Figure 5.

Thrombocytopenia results in lung bleeding at the site of infection. Mice were treated with anti-GPIbα (platelet depletion) or immunoglobulin G (IgG) control (both 0.4 µg/g) and infected with Burkholderia pseudomallei via the airway and killed after 24, 48, or 72 hours or killed uninfected. A, Representative photographs of naive or infected lungs and bronchoalveolar lavage fluid (BALF). B, Representative microphotographs of hematoxylin and eosin (H&E)–stained tissue sections (original magnification ×40), bleeding indicated by arrow. C, Quantification of lung bleeding; both scored on H&E-stained tissue sections by a pathologist blinded for groups and hemoglobin measurement in 50-fold diluted lung homogenates or BALF. Data are represented as bars (mean with standard error of the mean). n = 8 mice per group. *P < .05, ***P < .001 vs IgG control. Abbreviations: BALF, bronchoalveolar lavage fluid; BPS, B. pseudomallei; OD, optical density.

Figure 6.

Platelet GPIb deficiency decreases platelet counts and leads to increased bacterial growth in the lung during experimental melioidosis. IL4R/GPIbα or control mice were infected with Burkholderia pseudomallei via the airway and killed after 72 hours. A, GPIb expression on platelets. B, Platelet counts, P-selectin expression, and platelet–neutrophil complex formation. C, Bacterial quantification in indicated organs. Data are expressed as box-and-whisker plots depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. n = 8 mice per group. *P < .05, **P < .01, ***P < .001 vs IgG control. Abbreviations: CFU, colony-forming units; SSC-H, side scatter height.