Within their electron dense granule matrix, alpha-granules package a wide range of proteins important to hemostasis, thrombosis, inflammation, and angiogenesis. During the platelet release reaction, these proteins are secreted in a kinetically diverse manner that is proportional to agonist stimulation. The release reaction is accompanied by a large expansion in granule volume, matrix decondensation, and the solubilization of the included proteins. We have previously posited that some combination of differential protein proximity to granule exit sites and protein solubilization could contribute to the diversity of protein release.Here we have used thrombin stimulation and SNARE knockout mice to test two possible mechanisms for granule decondensation/matrix solubilization. In the first, granule swelling and decondensation precedes fusion with the plasma membrane and is regulated by aquaporin-mediated water influx presumably driving the volume change. In the second, matrix solubilization is plasma membrane fusion dependent with the extracellular influx of water through the fusion pore driving volume change and granule decondensation.In wild-type mouse platelets, we found a precursor-product relationship between the incidence of condensed and decondensed granules with the incidence of decondensed granules peaking at 90 sec post-stimulus (0.1 U/mL thrombin). The subsequent decrease in decondensed granules was due to an increasing incidence of compound granule fusion. By electron tomography, granules were connected to the plasma membrane by “necks” or short “pipes”. The granule membrane remained physically distinct during the entire granule decondensation process. Granule decondensation was marked by the conversion of the granule matrix into a diffuse fibrous structure that decreased progressively in electron density over the 5 min stimulation period. A residual electron-dense core was most obvious at 90 sec post stimulation. At the plasma membrane fused granule neck/pipe, solubilized granule matrix fibers were observed to be extruded extracellularly.Overall, the data suggest a progressive solubilization of granule matrix contents and their accompanying extrusion through the granule neck/pipe. In SNARE knockout mouse platelets, SNARE-selective inhibition of matrix solubilization/granule decondensation was observed. Both qualitatively and quantitatively, knockout of VAMP2/3 had little-to-no effect on 1) alpha-granule fusion with the plasma membrane, 2) granule decondensation, 3) matrix solubilization, or 4) compound granule fusion. In striking contrast, VAMP8 knockout initially produced a substantial inhibition of the incidence of all four steps in the granule releasate reaction while a VAMP2/3/8 knockout almost totally inhibited all four steps. In sum, our data strongly indicate that alpha-granule decondensation and matrix protein solubilization/extrusion are highly dependent on SNARE-mediated granule fusion with the plasma membrane. These data provide no support for the alternative mechanism that regulated aquaporin-dependent water influx mediates granule decondensation prior to granule fusion with the plasma membrane or canalicular system.This work was supported in part by NIH grant R01 HL 119393 to BS, NIH grant R01 HL56652 to SWW, and AHA grant 15PRE2555020 to SJ. DisclosuresNo relevant conflicts of interest to declare.