Medicine Meets Mathematics: Measuring ADAM10 Activity Directly On a Forming Thrombus

Platelets respond rapidly to injury, infection and changes in blood shear stress where they engage subendothelial collagen and von Willebrand Factor via platelet-specific receptors glycoprotein (GP)VI and the GPIb-IX-V complex, respectively. Metalloproteolytic shedding of GPVI is one important consequence of platelet activation and occurs via activation of the receptor sheddase A Disintegrin and Metalloproteinase (ADAM)10. We demonstrated that exposure of platelets to brief, elevated shear was sufficient to activate ADAM10-mediated shedding of GPVI and that this activation did not require platelet receptor engagement, intracellular signalling or release of soluble mediators (Al Tamimi et al., Blood, 2012). A critical global question, however, is how ADAM10 activity towards vascular substrates can be so rapidly upregulated in a high shear environment? To directly examine shear-induced activation of a vascular metalloproteinase, we developed the first sensor capable of visualising ADAM10 activity on a forming thrombus, consisting of a GPVI sequence-based ADAM10-sensitive fluorescent peptide with an ADAM10-releasable quencher (GPVI-Cy3). Rapid recombinant (r) ADAM10 (K_m = 24.3 μM, k_cat = 0.27 s^-1) but not rADAM17 cleavage of GPVI-Cy3 permitted direct ADAM10 monitoring on platelets. First, suspensions of human washed platelets were exposed to variable rates of uniform shear in a cone-plate viscometer, or were treated with 10 μg/ml collagen-related peptide (CRP; a GPVI ligand), or 5 mM NEM (a potent generic activator of ADAMs) then mixed with 5-10 μM GPVI-Cy3. Real-time fluorescence was monitored in a fluorescence plate reader. Untreated platelets displayed basal (37% of maximal) levels of ADAM10 activity (2.3 pmol/min/10⁶ platelets; 100% denoting NEM-induced levels) consistent with active ADAM10, lacking an inhibitory pro-domain by western blot and flow cytometry, being present on the non-activated platelet surface. ADAM10 activity increased to 59% of maximal activity following CRP treatment and 86% of maximal activity after exposure to 10,000 s^-1 shear stress for 5 min (5.4 pmol/min/10⁶ platelets). Cleavage of GPVI-Cy3 was completely blocked by inclusion of 100 μM GM6001 or 2 μM GI254023 (specific ADAM10 inhibitor). Second, ADAM10 activity was visualised using fluorescence multi-channel confocal imaging (Nikon A1R Plus si, Piezo z-stage and perfect focus system) of thrombi formed by perfusion of collagen-coated capillaries with hirudinated whole blood (input wall shear rate = 1,800 s^-1). The channel profile was reconstructed digitally, and a Computational Fluid Dynamics package (OPEN-Foam) accurately simulated the shear forces acting on the flow throughout the domain as a function of time. Importantly, this technique permits shear rates to be precisely determined at the thrombi surface, and provides time-varying data on shear rates experienced by platelets traversing the domain in vitro. Highest ADAM10 activity was observed 5-10 minutes post thrombus formation and colocalized with areas of high (>6000 s^-1) shear on the surface of the thrombus. When GPVI-Cy3 was included for the duration of thrombus formation, serial z-stack thrombus cross section images revealed areas of high and low ADAM10 activity within the thrombus core, consistent with elevated ADAM10 activity on platelets involved in the initial stages of thrombus formation. Together, our findings (i) define a novel ADAM10-selective substrate GPVI-Cy3 which reports on ADAM10 activity in vitro and under hydrodynamic flow ex vivo, (ii) enable simultaneous quantitation of ADAM10 activity and local shear rates, and (iii) demonstrate correlation between ADAM10 activity and regions of elevated shear stress for the first time. Future studies will investigate how shear stress under pulsatile or continuous flow activates ADAM10 on vascular cell membranes, ultimately enabling the design of therapeutic agents that discretely target shear-mediated up-regulation of ADAMs activity.