Screening for Key Structural Differences in Thrombosis Versus Hemostasis through Single Platelet Analysis

INTRODUCTION: Taking a single platelet analysis (SPA) approach to score platelet activation states distribution within a thrombus provides a unique opportunity to examine cellular difference in the process of thrombotic versus hemostatic clot formation. Potentially, it may reveal platelet activation states are dispensable for hemostasis but critical for thrombosis, e.g., platelet infill within the interior cavities of thrombus. Hence, our goal of determining the distribution of platelet activation states and of a-granule content release in an occlusive clot compared to a hemostatic thrombus.

METHODS: We used a femoral artery model in mice in which occlusive clots were induced by ferric chloride and hemostatic processes occurred within a puncture wound. Images were produced by wide area transmission electron microscopy of full thrombi cross sections, 3.185 nm XY pixel size, at various depths. We mapped and analyzed 2D images of ferric chloride induced clotting using automated SPA with locally run Zeiss Arivis Pro software. Through automated SPA, we segmented and mapped the distribution of tens of thousands of platelets and trapped red blood cells versus the vessel wall and associated vascular damage. The incidence of condensed and decondensed a-granules, an indicator of platelet secretion activity, was scored through Cloud implemented APEER models. Occlusive clot cross sections were analyzed at multiple scales to reveal the distribution of platelet activation states based on platelet shape or to reveal a-granule release state (condensed versus decondensed granules). For the simpler problem of SPA of puncture wound hemostasis, we manually scored platelet activation states across tens of thousands of platelets.

RESULTS AND CONCLUSIONS: With respect to occlusive clotting, several significant results and conclusions emerged from the platelet shape analysis. In the occlusive clot, we observed that activated platelets adhering directly to the damaged area of the vessel wall, exposed intima, lost their shape, becoming oval with extended pseudopods to more compressed, compacted tightly adherent shape compared to circulating, discoid platelets. Tightly adherent platelets formed column-like projections that structured the interior of the clot. Alpha-granule secretion was apparent. Most interestingly, the interior space between columns was filled with loosely adherent elongated, discoid-shaped platelets that frequently contained decondensed a-granules indicative of granule release (Figure 1). To the best of our knowledge, these data are the first indications that discoid platelets can secrete at least partially, their a-granule contents. In addition to platelets in various activation states, ~7% of the clot cross-sectional area was occupied by shape distorted red blood cells trapped in clusters. Any morphological effect of the red blood cells on platelet activation was restricted to a short distance of ~5 microns.

Our SPA analysis of the femoral artery, puncture wound thrombi, support at least two conclusions: 1) Decondensed a-granules were abundant in discoid platelets coating the intravascular “crown” of the 20 min post puncture thrombus, i.e., a-granule secretion by discoid platelet is found in both hemostasis and thrombosis. 2) Circular folding of the hemostatic thrombus to give complete infill of the artery would place discoid platelets within central interior portions of an occlusive clot. We suggest that there may well be shared principles of assembly between the two situations. Importantly, the intravascular accumulation of discoid platelets in hemostasis occurs post-bleeding cessation. Therefore, drugs selectively affecting this state would target preferentially thrombotic clots in which clot infill results in the accumulation of discoid platelets.

Figure 1: Distribution of platelets within a cross section through the middle of a ferric chloride induced, occlusive femoral clot. A)Platelet rich clot showing extensive platelet adhesion to the intima, white line encircling the platelet rich clot, inset blowup of the centrally located discoid platelets B) Distribution of red blood cells (red) and elongated, discoid shaped platelets (yellow). C) Cloud based SPA of red blood cells, platelets (yellow), condensed a-granules (light blue) and decondensed a-granules (magenta). D) Global distributions of granules (light blue and magenta) relative to platelets.