The role of Nox1 and Nox2 in GPVI-dependent platelet activation and thrombus formation

Abstract

Background: Activation of the platelet-specific collagen receptor, glycoprotein (GP) VI, induces intracellular reactive oxygen species (ROS) production; however the relevance of ROS to GPVI-mediated platelet responses remains unclear.

Objective: The objective of this study was to explore the role of the ROS-producing NADPH oxidase (Nox)1 and 2 complexes in GPVI-dependent platelet activation and collagen-induced thrombus formation.

Methods and results: ROS production was measured by quantitating changes in the oxidation-sensitive dye, H2DCF-DA, following platelet activation with the GPVI-specific agonist, collagen related peptide (CRP). Using a pharmacological inhibitor specific for Nox1, 2-acetylphenothiazine (ML171), and Nox2 deficient mice, we show that Nox1 is the key Nox homolog regulating GPVI-dependent ROS production. Nox1, but not Nox2, was essential for CRP-dependent thromboxane (Tx)A2 production, which was mediated in part through p38 MAPK signaling; while neither Nox1 nor Nox2 was significantly involved in regulating CRP-induced platelet aggregation/integrin αIIbβ3 activation, platelet spreading, or dense granule and α-granule release (ATP release and P-selectin surface expression, respectively). Ex-vivo perfusion analysis of mouse whole blood revealed that both Nox1 and Nox2 were involved in collagen-mediated thrombus formation at arterial shear.

Conclusion: Together these results demonstrate a novel role for Nox1 in regulating GPVI-induced ROS production, which is essential for optimal p38 activation and subsequent TxA2 production, providing an explanation for reduced thrombus formation following Nox1 inhibition.

Keywords: Glycoprotein VI; NAPDH oxidase; Platelets; Reactive oxygen species; Thrombus.

Graphical abstract

Fig. 1

GPVI-specific ROS production does not involve Nox2. (A) Washed platelets from wildtype (Wt) or Nox2-knockout (KO) mice were pre-loaded with 10 µM H₂DCFDA, then stimulated with 1 µg/mL CRP for 2 min and monitored for ROS production by flow cytometry. Data are mean±SEM, n=8, non-significant (ns). (B) H₂DCFDA loaded neutrophils from Wt or Nox2-KO mice were stimulated with 5 µM PMA for 5 min and analyzed by flow cytometer to assess ROS production. Data are mean±SEM, n=4, ^⁎⁎⁎P<0.001. Two-way ANOVA with Bonferroni post-hoc analysis were performed for A and B.

Fig. 2

Nox1 is the key Nox homolog regulating GPVI-dependent ROS production. Platelet aggregation is independent of Nox1/2. (A) Wildtype (Wt) and Nox2-KO H₂DCFDA-loaded platelets were pretreated with vehicle control (0.1% DMSO) or the Nox1 specific inhibitor (5 µM ML171) and monitored for ROS production using 0.25 µg/mL (left panel) and 1 µg/mL (right panel) CRP with 2 min stimulation. Data are mean±SEM, n=3 (0.25 µg/mL), n=4 (1 µg/mL); ^⁎P<0.05, ^⁎⁎P<0.01. (B) Washed platelets from Wt or Nox2-KO platelets containing 0.1% DMSO or 5 µM ML171 were monitored for CRP-dependent platelet aggregation (left panels) with area under the curve analysis (right panels) using 0.25 µg/mL and 1 µg/mL CRP. Data are mean±SEM, n=7, non-significant (ns). Two-way ANOVA with Bonferroni analysis were performed for A and B.

Fig. 3

GPVI-mediated platelet α and dense granule release is independent of Nox1/2, but thromboxane A₂ generation requires Nox1-derived ROS. Washed platelets from wildtype or Nox2-KO mice were pre-incubated with vehicle control (0.1% DMSO) or Nox1 inhibitor (5 µM ML171) and assessed following stimulation with 1 µg/mL CRP for (A) α-granule release by measuring P-selectin (CD62P) surface expression, (B) dense body secretion by ATP release and (C) thromboxane A₂ production using a TxB₂ ELISA. Data are mean±SEM, n=4 (n=3 for TxA₂ assay), ^⁎P<0.05, ^⁎⁎P<0.01 by two-way ANOVA with Bonferroni correction.

Fig. 4

Collagen-induced platelet aggregation and ROS production requires Nox1. Wildtype and Nox2-KO washed platelets were pretreated with vehicle control (0.1% DMSO) or the Nox1 specific inhibitor (5 µM ML171), then stimulated with 5 µg/mL collagen and monitored for (A) platelet aggregation (left panel) with area under the curve analysis (right panel) and (B) ROS production. Data are mean±SEM, n=4, ^⁎P<0.05, ^⁎⁎P<0.01 by two-way ANOVA with Bonferroni analysis.

Fig. 5

CRP-mediated p38 activation requires Nox1-derived ROS. (A) Washed platelets from wildtype (WT) mice were pre-treated with vehicle control (0.1% DMSO), Nox1 inhibitor (5 µM ML171) or ROS scavenger apocynin (250 µM) and stimulated for up to 3 min with 1 µg/mL CRP under non-stirring conditions. Platelets were lysed, separated by SDS-PAGE and immunoblotted using phospho-specific antibodies for active ERK1/2 (upper panel), Akt (middle panel) and p38 (lower panel), with subsequent loading controls for each protein. Blots are representative of three independent experiments. (B) Densitometric analysis of p38 phosphorylation expressed as fold change relative to unstimulated control. Data are mean±SEM, n=3, ^⁎P<0.05, ^⁎⁎P<0.01 by two-way ANOVA with Bonferroni correction. (C) Washed platelets from WT mice were pre-incubated with 0.1% DMSO or p38 inhibitor (10 µM SB202190), stimulated with 1 µg/mL CRP and monitored for thromboxane A₂ production using a TxB₂ ELISA. Data are mean±SEM, n=3 ^⁎⁎P<0.01 by two-way ANOVA with Bonferroni correction.

Fig. 6

Collagen-dependent thrombus formation at arterial shear requires both Nox1 and Nox2. Hirudin-anticoagulated whole blood from wildtype (Wt) or Nox2 knockout (KO) mice was diluted 1:2 with modified HEPES-Tyrode's buffer, then pre-incubated with vehicle control (0.5% DMSO), Nox1 inhibitor (5 µM ML171) or aspirin (500 µM). Samples were perfused over a collagen-coated microfluidic channel at a constant shear of 1500 s⁻¹ for 6 min, followed by washing in modified HEPES-Tyrode's buffer for 5 min with subsequent fixation in 3.7% formaldehyde. (A) Representative fluorescent, end point images of platelet thrombi from Wt (upper panel) and Nox2 KO (lower panel) mice following vehicle control/inhibitor treatment. (B) Platelet surface coverage (left panel) and thrombus volume (right panel) are expressed as mean±SEM, n=5 (n=3 for aspirin), ^⁎P<0.05, ^⁎⁎P<0.01 by two-way ANOVA with Bonferroni analysis.