NOX2 is critical for heterotypic neutrophil-platelet interactions during vascular inflammation

Abstract

Platelet-leukocyte interactions on activated endothelial cells play an important role during microvascular occlusion under oxidative stress conditions. However, it remains poorly understood how neutrophil-platelet interactions are regulated during vascular inflammation. By using intravital microscopy with mice lacking nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 2 (NOX2) and their bone marrow chimera, we demonstrated that NOX2 from both hematopoietic and endothelial cells is crucial for neutrophil-platelet interactions during tumor necrosis factor alpha-induced venular inflammation. Platelet NOX2-produced reactive oxygen species (ROS) regulated P-selectin exposure upon agonist stimulation and the ligand-binding function of glycoprotein Ibα. Furthermore, neutrophil NOX2-generated ROS enhanced the activation and ligand-binding activity of αMβ2 integrin following N-formyl-methionyl-leucyl phenylalanine stimulation. Studies with isolated cells and a mouse model of hepatic ischemia/reperfusion injury revealed that NOX2 from both platelets and neutrophils is required for cell-cell interactions, which contribute to the pathology of hepatic ischemia/reperfusion injury. Platelet NOX2 modulated intracellular Ca(2+) release but not store-operated Ca(2+) entry (SOCE), whereas neutrophil NOX2 was crucial for SOCE but not intracellular Ca(2+) release. Different regulation of Ca(2+) signaling by platelet and neutrophil NOX2 correlated with differences in the phosphorylation of AKT, ERK, and p38MAPK. Our results indicate that platelet and neutrophil NOX2-produced ROS are critical for the function of surface receptors essential for neutrophil-platelet interactions during vascular inflammation.

Figure 1

Platelet and neutrophil NOX2 regulate heterotypic neutrophil-platelet interactions during TNF-α-induced venular inflammation. Intravital microscopy of WT and NOX2 KO mice was performed as described in the “Methods” section. Cremaster vascular inflammation was induced by intrascrotal injection of TNF-α. Three hours after TNF-α injection, neutrophils and platelets in inflamed venules were visualized by infusion of Alexa Fluor 647-conjugated anti-Gr-1 and Dylight 488-conjugated anti-CD42c antibodies, respectively, into the mice. (A) Representative images at various time points. Arrows show direction of blood flow. (B-C) Number of rolling and adherent neutrophils. (D-E) The integrated median fluorescence intensities of anti-CD42c antibodies (F platelets) were quantified, normalized by the number of adherent neutrophils and the vessel length, and plotted as a function of time. (E) F platelets in WT and NOX2 KO mice were compared at 60, 120, and 180 seconds after each capture (0 seconds). (F-G) WT and NOX2 KO platelets (F) or neutrophils (G) were isolated and labeled with calcein acetoxymethyl ester (calcein AM). The labeled cells (5 × 10⁷ platelets or 10⁶ neutrophils per mouse) were infused into TNF-α-inflamed mice. Endogenous neutrophils or platelets were visualized by infusion of Alexa Fluor 647-conjugated anti-Gr-1 or Dylight 649-conjugated anti-CD42c antibodies, respectively. Adhesion of the infused platelets to adherent neutrophils (F) and adhesion of endogenous platelets to infused neutrophils (G) were counted and normalized by the number of adherent neutrophils. Data represent the mean ± standard error of the mean (SEM) (n = 22-28 venules in 4 mice per group). *P < .05, **P < .01, or ***P < .001 vs WT control after Mann-Whitney test (E) or vs infusion of WT cells into WT mice (F-G) after Student t test. #P < .05 or ##P < .01 between two groups after Student t test.

Figure 2

Hematopoietic and EC NOX2 are important for neutrophil-platelet interactions during venular inflammation. Intravital microscopy in NOX2 bone marrow chimeric mice was performed as described in Figure 1. (A) Representative images at various time points. Arrows show direction of blood flow. (B-C) Number of rolling and adherent neutrophils. (D-E) The fluorescence intensities of anti-CD42c antibodies (F platelets) and the comparison of F platelets among groups. Data represent the mean ± SEM (n = 22-24 venules in 4 mice per group). *P < .05 or ***P < .001 vs WT control after Mann-Whitney test.

Figure 3

ROS produced from platelet NOX2 regulate P-selectin exposure and GPIbα function. (A) Mouse or (B) human platelets were pretreated with vehicle, catalase (1000 U/mL), dimethylsulfoxide (0.1%), or DPI (10 µM), and incubated with or without thrombin (0.025 U/mL in [B]). The surface expression of P-selectin (CD62P) was measured by flow cytometry. (C) Human platelets were treated with 1 μg/mL mouse immunoglobulin G1 (IgG1) or an anti-GPIbα antibody (VM16d), vehicle, or catalase followed by incubation with 10 μg/mL αMβ2 in the presence or absence of 0.5 mM MnCl₂. The bound αMβ2 was determined by flow cytometry using goat anti-human β2 antibodies. (D) αMβ2 binding was measured in WT and NOX2 KO platelets as described in (C). (E-F) Mouse platelets were pretreated with or without catalase or 1 μM H₂O₂ and then incubated with 10 µg/mL vWF and 10 mM EDTA in the presence or absence of 10 µg/mL botrocetin. vWF binding was analyzed by flow cytometry using anti-vWF antibodies. Flow cytometric data are shown as the geometric mean fluorescence intensity (MFI) value or fold increase obtained by normalization of the MFI of antibodies to that of control IgG. (G) Mouse platelets were incubated with 10 µg/mL vWF and 10 µg/mL botrocetin. In some experiments, NOX2 KO platelets were pretreated with 0.1 to 1 μM H₂O₂. Platelet agglutination was measured in an aggregometer. All data represent the mean ± standard deviation (SD) (n = 4-5). *P < .05, **P < .01, or ***P < .001 between two groups after Student t test.

Figure 4

ROS generated from neutrophil NOX2 are crucial for the activation and ligand-binding activity of αMβ2 integrin. (A-B) Mouse neutrophils were treated with or without fMLF, and the surface expression of PSGL-1 and αMβ2 integrin was measured by flow cytometry. (C) WT and NOX2 KO neutrophils were pretreated with or without 1 μM H₂O₂ and stimulated with fMLF. Lysates were immunoprecipitated (IP) with control IgG or anti-β2 antibodies, followed by immunoblotting and densitometry. (D) Human neutrophils pretreated with 1000 U/mL catalase were stimulated with fMLF, followed by flow cytometry using antibodies against total (ICRF44) and activated αMβ2 (CBRM1/5). Data are shown as a fold increase obtained by normalization of the geometric MFI value of antibodies to that of control IgG. (E-F) Mouse (E) or human (F) neutrophils were pretreated with or without H₂O₂ (1 µM) or catalase and then incubated with Alexa Fluor 488-conjugated FG in the absence or presence of fMLF. Untreated cells are shown as 100% (white bar). Data represent the mean ± SD (n = 3-4). *P < .05, **P < .01, or ***P < .001 vs WT or vehicle control after analysis of variance (ANOVA) and Dunnett’s test, and #P < .05 or ##P < .01 vs WT or vehicle control after Student t test.

Figure 5

NOX2 is important for neutrophil-platelet aggregation under stirring conditions and for hepatic ischemia/reperfusion injury in vivo. (A-B) In vitro neutrophil-platelet aggregation assay was performed as described in the “Methods” section. Neutrophils and platelets isolated from WT and NOX2 KO mice were labeled with Alexa Fluor 647-conjugated anti-Gr-1 and Dylight 488-conjugated anti-CD42c antibodies, respectively. Thrombin-activated platelets were mixed with TNF-α-treated neutrophils under stirring conditions (1000 rpm). Cells were analyzed by flow cytometry. R1 designates leukocyte-platelet aggregates and R2 designates total population of neutrophils. Neutrophil-platelet aggregation was measured by the fluorescence intensity of anti-CD42c antibodies (B) in the R1 gate. Data represent the mean ± SD (n = 3). SS, side scatter; FS, forward scatter. (C-F) Hepatic ischemia/reperfusion injury was induced as described in the “Methods” section. (C) Representative pictures of liver sections stained with triphenyltetrazolium chloride. (D) Infarct sizes (white areas in C) were measured by Image J. (E) Serum levels of aspartate aminotransferase (AST). (F) Neutrophils (arrow heads) and platelets (arrows) in liver sections were stained with an esterase kit (pink) and anti-CD41 antibodies (brown), respectively. Nucleated cells were stained with hematoxylin (purple). Bar = 20 μm. (G) The number of neutrophils was counted inside (gray) and outside (white) the hepatic vessels. (H) The number of platelets that interact with adherent neutrophils to the vessels was counted. Data represent the mean ± SD (n = 5 mice and 15 sections in 5 mice for G-H). *P < .05, **P < .01, or ***P < .001 vs WT control or sham after ANOVA and Dunnett’s test. In (G), **P < .01 was obtained after comparison of the total number of neutrophils. #P < .05, ##P < .01, and ###P < .001 between 2 groups after Student t test.

Figure 6

Platelet and neutrophil NOX2 differentially regulate Ca²⁺ release and influx. Mouse platelets (A-B) and neutrophils (C-D) were pretreated with Ca²⁺ dye and then incubated with thrombin (0.025 U/mL [A]), fMLF (C), or thapsigargin (200 nM [B,D]) for 3-4 minutes and 2 mM CaCl₂ was then added. Intracellular Ca²⁺ release and SOCE were measured and quantified by the AUC (arbitrary units). Quantitative data represent the mean ± SD (n = 4-5). *P < .05 or **P < .01 vs WT after Student t test.

Figure 7

Platelet and neutrophil NOX2 are important for regulating the phosphorylation of AKT, ERK, and p38MAPK. (A-B) Mouse platelets and (C-D) neutrophils were treated with thrombin (0.025 U/mL) and fMLF, respectively, for 0.5, 1, and 2 minutes. (E-F) Platelets and (G-H) neutrophils were treated with 200 nM TG for 3 minutes, and 2 mM CaCl₂ was then added and incubated for 1 minute. Cells were lysed at different time points, followed by immunoblotting with equal amounts of proteins and densitometry. (A,C,E,G) Representative blots. (B,D,F,H) Quantitative graphs. Data represent the mean ± SD (n = 4-6). *P < .05, **P < .01, or ***P < .001 vs WT after Student t test.