The ectopic expression of proteins within platelets may offer a novel strategy for the targeted delivery of proteins of interest to a site of vascular injury. As a proof-of-principle, we transgenically expressed FVIII in the megakaryocytes of mice. The maximal level of platelet (p) FVIII of the 8 founder animals tested had the equivalent of a 3% plasma correction (Line #38) and was capable of correcting bleeding in FVIIInull animals. Confocal microscopy studies suggested that pFVIII co-localized with VWF in the alpha-granules. We were then interested in knowing whether VWF was necessary for pFVIII storage as previously shown for FVIII granular storage in neural cell lines. We wished to know this because this pFVIII level was 10% of platelet VWF level. Does the VWF level limit the maximal level of pFVIII that could be stored? Additionally, was our success in having FVIII stored in platelets and released at sites of injury unique because platelets coincidentally store and release FVIII’s carrier protein? To address the role of VWF in pFVIII storage and efficacy in correcting bleeding in the FVIIInull setting, we obtained VWF−/− mice (Denisa Wagner, Harvard U). When these were crossed with Line #38 mice, there was little change in pFVIII levels between VWF+/+ and VWF+/− platelets. The level of pFVIII was decreased in the VWF−/− animals to ~75% of that seen in the VWF+/+ animals (n=12 per arm, p= 0.05). Plasma FVIII levels in Line#38/FVIIInull/VWF−/− mice was undetectable, suggesting that the lost pFVIII was not leaking into the plasma. We do not have an explanation yet for the observed decrease in pFVIII level. Certainly more animals on a common strain background will need to be tested to see if this decrease remains. However, if this difference remains, it would suggest two pools of pFVIII, a small one that is VWF dependent and a larger one independent of VWF. We also began to test whether the pFVIII in the Line #38/FVIIInull/VWF−/− mice was as effective as FVIII from platelets from Line #38/FVIIInull/VWF+/+ animals in correcting bleeding in FVIIInull animals. A concern with the design of these studies was separating the effects of VWF deficiency from FVIII deficiency. We carried out whole blood clot timings (WBCT) and FeCl3 carotid artery injury studies with platelets infused into VWF+/+/FVIIInull mice. These studies assume that plasma VWF of the recipient animals will make up for the absence of platelet VWF release from the transfused platelets. The pFVIII in the VWF−/− platelets was able to correct the WBCT and the FeCl3 carotid injury thrombosis time equivalent to the Line #38/FVIIInull/VWF+/+ pFVIII. Confocal microscopy and immunoelectron microscopy studies to define where pFVIII is stored in the VWF−/− setting are underway, but the presented studies show that the majority of pFVIII does not require VWF to be stored in platelets in a biologically available form. These data suggest that the maximum level of pFVIII achievable may not have been reached with Line #38, that future FVIII constructs that do not bind VWF may still be stored in platelets and appropriately released, and that other proteins of interest that do not have an available carrier protein might also be stored and released at sites of vascular injury.

Author notes

Corresponding author