Protein Disulphide Isomerase and NADPH Oxidase 1 Cooperate to Control Platelet Function and Are Associated with Cardiometabolic Disease Risk Factors

Abstract

Background: Protein disulphide isomerase (PDI) and NADPH oxidase 1 (Nox-1) regulate platelet function and reactive oxygen species (ROS) generation, suggesting potentially interdependent roles. Increased platelet reactivity and ROS production have been correlated with cardiometabolic disease risk factors.

Objectives: To establish whether PDI and Nox-1 cooperate to control platelet function.

Methods: Immunofluorescence microscopy was utilised to determine expression and localisation of PDI and Nox-1. Platelet aggregation, fibrinogen binding, P-selectin exposure, spreading and calcium mobilization were measured as markers of platelet function. A cross-sectional population study (n = 136) was conducted to assess the relationship between platelet PDI and Nox-1 levels and cardiometabolic risk factors.

Results: PDI and Nox-1 co-localized upon activation induced by the collagen receptor GPVI. Co-inhibition of PDI and Nox-1 led to additive inhibition of GPVI-mediated platelet aggregation, activation and calcium flux. This was confirmed in murine Nox-1^-/- platelets treated with PDI inhibitor bepristat, without affecting bleeding. PDI and Nox-1 together contributed to GPVI signalling that involved the phosphorylation of p38 MAPK, p47phox, PKC and Akt. Platelet PDI and Nox-1 levels were upregulated in obesity, with platelet Nox-1 also elevated in hypertensive individuals.

Conclusions: We show that PDI and Nox-1 cooperate to control platelet function and are associated with cardiometabolic risk factors.

Keywords: NADPH oxidase; metabolic syndrome; platelets; protein disulphide isomerase; redox biology.

Figure 1

PDI and Nox-1 cellular localization in resting and CRP-activated platelets. Immunofluorescence of platelet-rich plasma (PRP) not stimulated (resting) (A) or stimulated with 1 μg/mL CRP (B) for 3 min in the presence of 4 μg/mL integrilin. (C) Pearson’s correlation of at least 3 different fields for 3 independent experiments represents the degree of co-localization of Nox-1 and PDI (i), PDI and p47phox (ii) and Nox-1 and p47phox (iii). Bar graphs show mean ± SEM. Data analysed by paired Student t-test. Exact p-value are presented in each bar graph. Pink colour is the colocalisation of p47phox (red) and PDI (blue), yellow is the colocalisation of p47phox and Nox-1 (green) and white is the colocalisation of all three proteins.

Figure 2

Bepristat and ML171 exert an additive inhibitory effect on plate-based platelet aggregation induced by GPVI agonists. Human WP at 4 × 10⁸ platelets/mL were pre-treated with 3 µM ML171 and/or 15 µM Bepristat for 10 min prior to addition of agonists: 1 µg/mL CRP (A), 2 µg/mL Collagen (B), 10 µM TRAP-6 (C) or 500 nM PMA (D) (n = 4–5). Platelet aggregation was measured using a plate-based assay. Data on graphs show mean ± SEM. Data analysed by paired one-way ANOVA and Tukey’s multiple comparisons test. * p < 0.05 and ** p < 0.01. ns: non-significant.

Figure 3

Bepristat and ML171 display additive inhibitory effects on CRP-induced fibrinogen binding, P-selectin exposure and calcium mobilization. (A,B): Human WP (4 × 10⁷ platelets/mL) were incubated with 3 µM ML171 and/or 15 µM bepristat for 10 min, then 1 µg/mL CRP was added. FITC-fibrinogen and PE/Cy5-anti-P-selectin were incubated for 30 min and events acquired using a flow cytometer. Data expressed as median fluorescence intensity (MFI). (C): Representative curve of WP pre-incubated with calcium dye Fura-2 AM and pre-treated with 3 µM ML171 and/or 3.75 µM bepristat for 10 min prior to activation with 1 µg/mL CRP. Fluorescence acquired over 5 min using a plate reader. (D) Summary statistics for data in (C) activated with CRP. Peak Fura-2 was determined as the highest fluorescence value subtracted from baseline before agonist addition. n = 6 for (A,B), while n = 4 for (C,D). Bar graphs show mean ± SEM. Data analysed by paired one-way ANOVA and Tukey’s post-test. * p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001.

Figure 4

PDI and Nox-1 co-inhibition decreases phosphorylation of PKC substrates, Akt, p38MAPK and ERK. WP at 4 × 10⁸ platelets/mL were incubated with 3 µM ML171 and/or 15 µM bepristat for 10 min prior to adding 3 µg/mL Collagen. Platelets were lysed after 90 s and immunoblots performed. Samples were tested for: putative PKC substrate phosphorylation (A), AktS473 (B), p38 MAPKT180/Y182 (C) and ERKT202/Y204 (D). GAPDH was used as a control for equal loading. Representative blot is presented above of bar graphs with summary statistics. Each lane represents the condition in graph below. A basal unstimulated condition is presented (Rest). Data are representative of 3–4 independent experiments. Bar graphs show mean ± SEM and were analysed by paired one-way ANOVA and Tukey’s post-test. * p < 0.05, ** p < 0.01.

Figure 5

Nox-1^−/− mice treated with bepristat showed additive anti-platelet effects with no repercussion on bleeding time. (A) Platelet aggregation curves of mouse wildtype (WT) and Nox-1^−/− platelets pre-treated with 7.5 µM bepristat or vehicle for 10 min and activated with 5 µg/mL Collagen. (B) Summary statistics of aggregation curves. Immunoblots were performed in platelets pre-treated with 7.5 µM bepristat or vehicle for 10 min and activated with 5 µg/mL Collagen for 90 s. (C) PKC substrate. (D) 4G10 total Tyr phosphorylation. Representative blots are presented on top of bar graphs with summary statistics. Each lane represents the condition in graph below. A basal unstimulated condition is presented (Rest). (E) Tail bleeding time in mice injected with 50 µM bepristat or vehicle control. n = 5–6 for (A–D), while n = 5–8 for (E). Data on graphs expressed as mean ± SEM analysed by unpaired two-way ANOVA and Sidak’s post-test. * p < 0.05.

Figure 6

Platelet PDI and Nox-1 levels are independent of one another and upregulated in conditions of increased cardiovascular risk. Washed platelets (WP) from 136 volunteers were lysed and immunoblots performed for PDI, Nox-1 and loading control GAPDH. Anthropometric and metabolic characteristics were also collected. Value cut-offs of stratifications can be found on Methods and were all performed according to international guidelines. (A) Linear regression of platelet PDI and Nox-1 protein levels. (B–G) PDI and Nox-1 expression were stratified by: BMI (B,E) in healthy weight (18.5–24.9 kg/m²), overweight (25–29 kg/m²), class 1 obesity (30–34.9 kg/m²) and class 2 obesity (35–39.9 kg/m²); blood pressure (C,F) in normal (<120/80 mmHg), elevated (120–129/80 mmHg), high blood pressure (HBP) stage 1 (130–139/80–89 mmHg), HBP stage 2 (140/90 or above mmHg); waist circumference (D,G) in normal (Caucasian men < 94 cm; men of other ethnicities < 90 cm; women < 80 cm) and central obesity (Caucasian men ≥ 94 cm; men of other ethnicities ≥ 90 cm; women ≥ 80 cm). Data in graph show box and whiskers depicting median, range and 25th and 75th percentiles analysed by one-way ANOVA and Tukey’s post-test. Overall p-value of one-way ANOVA is shown. ns: non-significant.