Noncanonical Sonic Hedgehog signaling amplifies platelet reactivity and thrombogenicity

Abstract

Sonic Hedgehog (Shh) is a morphogen in vertebrate embryos that is also associated with organ homeostasis in adults. We report here that human platelets, though enucleate, synthesize Shh from preexisting mRNAs upon agonist stimulation, and mobilize it for surface expression and release on extracellular vesicles, thus alluding to its putative role in platelet activation. Shh, in turn, induced a wave of noncanonical signaling in platelets leading to activation of small GTPase Ras homolog family member A and phosphorylation of myosin light chain in activated protein kinase-dependent manner. Remarkably, agonist-induced thrombogenic responses in platelets, which include platelet aggregation, granule secretion, and spreading on immobilized fibrinogen, were significantly attenuated by inhibition of Hedgehog signaling, thus, implicating inputs from Shh in potentiation of agonist-mediated platelet activation. In consistence, inhibition of the Shh pathway significantly impaired arterial thrombosis in mice. Taken together, the above observations strongly support a feed-forward loop of platelet stimulation triggered locally by Shh, similar to ADP and thromboxane A2, that contributes significantly to the stability of occlusive arterial thrombus and that can be investigated as a potential therapeutic target in thrombotic disorders.

Graphical abstract

Figure 1.

Platelets synthesize the Shh ligand upon agonist stimulation. (A) Amplification chart representing Cq of mRNAs in platelets as indicated. (B) Synthesis of Shh in platelets treated with thrombin (Thr) (0.5 U/mL) for varying time points. PVDF membrane developed for Shh was stripped and reprobed for β-actin. (C) Corresponding densitometric analysis normalized against β-actin. (D-E) Surface externalization of Shh in thrombin (1 U/mL)-induced platelets. (F-G) Shh exposure on EVs released from thrombin (1 U/mL)-stimulated platelets. RP, resting platelets. Data are represented as mean ± standard error of the mean of 3 to 5 independent experiments. *P < .05 as compared with resting platelets; #P < .05 as compared with thrombin-stimulated platelets.

Figure 2.

Shh induces noncanonical signaling in platelets. (A) Expression of pMLC and pMYPT1 in platelets treated for 10 to 15 minutes with the following reagents (without stirring) as indicated in the figure: Shh, 3 μg/mL; thrombin (Thr), 0.5 U/mL; cyclopamine (Cyclo), 10 μM; compound C (CC), 50 μM; Y27632, 10 μM. PVDF membranes developed for pMLC and pMYPT1 were stripped and reprobed for total MLC and β-actin, respectively. (D) Expression of RhoA in platelets treated with Shh, thrombin, and cyclopamine (without stirring) as indicated in the figure. RhoA-GTP was pulled down from pretreated platelets employing a bead-based assay system and analyzed on Western blot. Whole cell lysate was separately prepared and probed for total RhoA. (F) Expression of pAMPK and pACC in platelets treated with different reagents for 15 minutes (without stirring) as indicated in the figure. PVDF membranes developed for pAMPK and pACC were stripped and reprobed for total AMPK and β-actin, respectively. (B-C,E,G-H) Corresponding bar diagrams after densitometric analyses of blots normalized against total MLC, RhoA, AMPK, or β-actin averaging 3 to 5 different experiments. (I) Intracellular calcium measurements in Fura-2–loaded platelets preincubated either with 10 μM cyclopamine (tracing 2) or vehicle (tracing 3), followed by the addition of thrombin (indicated by arrow). Tracing 1 represents unstimulated Fura-2–loaded platelets. (J) Represents intracellular calcium concentrations (mean ± standard error of the mean [SEM]) averaging 3 individual experiments. Figures are representative of ≥3 individual experiments (mean ± SEM). *P < .05 as compared with resting platelets; #P < .05 as compared with thrombin-stimulated platelets; and §P < .05 as compared with Shh-pretreated platelets.

Figure 3.

Thrombin-induced platelet functional responses are inhibited by Shh antagonists. Binding of Alexa fluor 488-fibrinogen (A-B) and FITC-labeled PAC-1 (C-D) to platelets treated with different reagents (thrombin, 0.1 U/mL for fibrinogen binding and 0.25 U/mL for PAC-1 binding; cyclopamine, 10 μM; and vismodegib, 25 μM) as indicated. (E-F) P-selectin externalization in platelets induced by 0.25 U/mL thrombin in the presence of vehicle, cyclopamine (10 μM), or vismodegib (25 μM), as indicated. (G-H) EV release from platelets pretreated with vehicle, cyclopamine, or vismodegib upon stimulation with thrombin (0.5 U/mL) as indicated. Figures are representative of ≥5 individual experiments (mean ± standard error of the mean). *P < .05 as compared with resting platelets and #P < .05 as compared with thrombin-stimulated platelets.

Figure 4.

Agonist-induced platelet aggregation is inhibited by Shh antagonists. Aggregation of platelets in PRP induced by collagen (0.5 μg/mL) (A), TRAP (2.5 μM) (C), or ADP (2.5 μM) (E) in the presence of vehicle (control), cyclopamine (10 μM), or vismodegib (25 μM), as indicated. Aggregation of washed human (G) and murine platelets (I) induced by 0.1 U/mL thrombin in the presence of vehicle (control), cyclopamine (10 μM), or vismodegib (25 μM) as indicated. (B,D,F,H,J) Represent corresponding bar diagrams showing percent aggregation (mean ± standard error of the mean) averaging ≥5 individual experiments. *P < .05 as compared with control.

Figure 5.

Platelet spreading on immobilized matrices is restrained by Shh antagonists. Phalloidin FITC-labeled platelets were pretreated with vehicle (control) cyclopamine (10 μM) or vismodegib (25 μM) as indicated. (A) Platelets treated with thrombin (0.5 U/mL) were allowed to spread over an immobilized fibrinogen matrix for 15 minutes. Images are representative of 5 different fields, each from 3 independent experiments. (B) Thrombin (0.5 U/mL)-treated platelets were allowed to spread over immobilized collagen matrix for 15 minutes. Images are representative of 10 different fields, each from 3 independent experiments. (C-D) Corresponding quantification of platelet spreading on fibrinogen and collagen matrices, respectively. Figures are representative of ≥3 individual experiments (mean ± standard error of the mean). *P < .05 as compared with vehicle-treated control platelets.

Figure 6.

Platelet adhesion on immobilized collagen matrix under arterial shear is restricted by Shh antagonists. Washed human or murine platelets were perfused over an immobilized collagen matrix for 5 minutes in a microfluidic flow chamber at a shear rate of 1500s⁻¹. (A,C,E,G) Representative images of platelet accumulation after 5 minutes of perfusion of human (A,C) or murine (E,G) platelets. Total thrombi area after 5 minutes of perfusion of human (B,D) or murine (F,H) platelets calculated as the average surface area covered by platelets in 5 representative fields. Figures are representative of ≥3 individual experiments (mean ± standard error of the mean). *P < .05 as compared with vehicle-treated control platelets.

Figure 7.

Inhibitors of Hedgehog signaling impair thrombus formation in mice. (A) Representative timelapse images exhibiting thrombus formation in mice, preadministered either with vehicle (control), cyclopamine, or vismodegib, captured 5, 15, or 25 minutes after injury of mesenteric arterioles of >100 μm diameter with 10% ferric chloride. (B-E) Scatter dot plots representing (B) time to first thrombus formation, (C) time to occlusion, (D) thrombus growth rate, and (E) tail-bleeding time in mice preadministered with vehicle, cyclopamine, or vismodegib. Each dot in scatter plots represents an independent observation. (F) Kaplan-Meier curve exhibiting proportion of occluded arteries at various time points of observation in mice preadministered with vehicle, cyclopamine, or vismodegib. Data are expressed as the mean ± standard error of the mean. *P < .05 with respect to vehicle-treated mice.

Figure 8.

Scheme depicting the role of Shh signaling in potentiating agonist-induced platelet stimulation. Short-range signaling by Shh, ADP, and thromboxane A₂ amplifies thrombin-induced platelet activation. Shh antagonists (cyclopamine/vismodegib), cyclooxygenase inhibitors (aspirin), and P2Y12 antagonists (clopidogrel) target these feed-forward loops to limit platelet activity.