Differential regulation of cyclic adenosine monophosphate by phosphodiesterase 3A discriminates thrombin-induced protease activated receptor 1– and 4–dependent platelet activation

Figures

• 

  FIGURE 1. Protease activated receptor (PAR)-specific inhibition reduces PAR1- or PAR4-dependent platelet activation. (A) Schematic representation of the strategy to inhibit PARs on platelets (top row) or CD34+-derived megakaryocytes (MKs) (bottom row). To examine the effect of thrombin-induced PAR1- or PAR4-dependent platelet activation, platelets were treated with vorapaxar, a PAR1-specific inhibitor, or BMS-986120, a PAR4-specific inhibitor. To complement inhibitor studies, genetic knockout of PAR1 or PAR4 expression was performed in CD34+-derived MKs using CRISPR/Cas9. (B–E) Washed platelets were stimulated with thrombin (B) or with PAR1- or PAR4-activating peptides (APs) (C–E) in the presence of vorapaxar, BMS, or vorapaxar and BMS. Dimethyl sulfoxide (DMSO) was used as a control. Platelet aggregation (C, D) and P-selectin expression (E) was then measured. A 1-way anova with Tukey multiple comparisons test for (C) and (D); n = 4 per group in (C) and n = 7 per group in (D). A 2-way anova with Tukey multiple comparisons test for (E); n = 5 per group. (F–K). PAR1 (PAR1 knockout [KO]) or PAR4 (PAR4 KO) were genetically disrupted in CD34+-derived MKs using CRISPR/Cas9. A scramble (scr)…

• 

  FIGURE 2. Protease activated receptor (PAR)4 induced platelet aggregation is delayed compared with PAR1. (A–C) Washed platelets were stimulated with the indicated thrombin concentration alone (black) or in the presence of vorapaxar (red) or BMS-986120 (blue) and platelet aggregation was measured. Tracings show mean ± SEM for each condition are shown. (D–I). Platelet aggregation tracings were quantified measuring lag time (D, F, H) and maximum aggregation (E, G, I). Mean ± SEM is shown. A Kruskal–Wallis with a Dunn multiple comparisons test was used in D, E, F, and H; n = 8–11 per group. A 1-way anova with a Tukey multiple comparisons test was used in (G) and (I); n = 8 per group for (G) and n = 11 per group for (I).

• 

  FIGURE 3. Calcium flux induced by protease activated receptor (PAR)4 activation is slower than that by PAR1. (A–C) Washed platelets were stimulated with the indicated thrombin concentrations alone (black) or in the presence of vorapaxar (red) or BMS-986120 (blue) and calcium flux based on the ratio of Fura-2AM at 340 to 380 nm was measured. Mean ± SEM of tracings for each concentration of thrombin are shown. Mean ± SEM of tracings are shown. (D–I) The time to peak calcium flux (D, F, H) and the maximum calcium flux (E, G, I) were quantified for all tracings. Mean ± SEM is shown. A Friedman test with a Dunn multiple comparisons test was used in (D); n = 6–7 per group. A mixed-effect analysis with a Tukey multiple comparisons test was used in (E), (F), and (G); n = 5–7 per group. A repeated-measures 1-way anova with Tukey multiple comparisons test was used in (H) and (I); n = 6–7 per group.

• 

  FIGURE 4. Protease activated receptor (PAR)1 activation leads to decreased cyclic adenosine monophosphate (cAMP) levels and more potent calcium flux. (A–C) Washed platelets were pretreated with vorapaxar (red), BMS-986120 (blue), or vehicle control (dimethyl sulfoxide [DMSO], black). One minute before activation (t = −1 minute), platelets were treated with 10 nM prostaglandin (PG)I2, followed by thrombin stimulation at the indicated concentrations at t = 0 minutes. At the specified time points, platelets were harvested and lysed with 0.1 N HCl to measure cAMP levels. cAMP levels were normalized 0% to 100% based on cAMP levels at t = −1 and t = 0 minutes, respectively. A mixed-effects analysis with Tukey multiple comparisons test for (A) and (B); n = 4–12. A 2-way anova with Tukey multiple comparisons test for (C); n = 6–10 per group. P values showed in (A–C) were calculated for comparing activation through PAR4 alone vs through PAR1 alone (blue) or through PAR4 alone vs both PAR1 and PAR4 (black). (D–F). Washed platelets were treated with 10 nM PGI2, followed by stimulated with the indicated thrombin concentrations alone (black) or in the presence of vorapaxar (red) or BMS-986120 (blue)…

• 

  FIGURE 5. Protease activated receptor (PAR)1 activation downregulates cyclic adenosine monophosphate (cAMP) levels through PDE3 in platelets. (A, B) Washed platelets were pretreated with vorapaxar (A, red) or BMS-986120 (B, blue) in the presence or absence of 20 μM Milrinone, phosphodiesterase (PD)E3 specific inhibitor. One minute before activation (t = −1 minute), platelets were treated with 10 nM prostaglandin (PG)I2 followed by thrombin stimulation at the indicated concentrations at t = 0 minutes. At the specified time points, platelets were lysed with 0.1N HCl to measure cAMP levels. cAMP levels were normalized 0% to 100% based on cAMP levels at t = −1 minute and t = 0 minutes, respectively. A 2-way anova with a Sidak multiple comparisons test was performed in (A) and (B); n = 7 per group in (A) and n = 6–7 per group in (B).

• 

  FIGURE 6. Protease activated receptor (PAR)1 activation downregulates cyclic adenosine monophosphate (cAMP) levels through phosphodiesterase (PD)E3A in CD34+-derived megakaryocytes (MKs). (A) PAR1 (red) and PAR4 (blue) were genetically disrupted in CD34+-derived MKs, and calcium flux was measured after stimulation with thrombin. CD34+-derived MKs treated with negative control guides CRISPR RNA (crRNA) were used as a control (black). Mean ± SEM of tracings are shown. (B, C) The time to peak calcium flux (B) and the maximum calcium flux (C) were quantified for all tracings. Mean ± SEM is shown. A repeated-measures 1-way anova with Tukey multiple comparisons test was used in (B) and (C); n = 9 per group. (D) PAR1 (red) and PAR4 (blue) were genetically disrupted in CD34+-derived MKs. CD34+-derived MKs treated with negative control guides crRNA were used as a control (black). CD34+-derived MKs were treated with 50 nM PGI2 at t = −1 minute followed by thrombin stimulation at the indicated concentration at t = 0 minutes. CD34+-derived MKs were harvested at the indicated time points to measure cAMP. cAMP levels were normalized 0% to 100% based on cAMP levels at t = −1 minute and t = 0 minutes,…