Analysis of the spatial and temporal characteristics of platelet-delivered factor VIII-based clots

Abstract

Normally factor (F) VIII is not expressed in megakaryocytes, but when human FVIII was transgenically expressed in murine megakaryocytes, it was stored in platelet alpha-granules and released at sites of injury. This platelet FVIII (pFVIII) is effective in correcting hemostasis, even in the presence of circulating inhibitors, so it offers a potential gene therapy strategy for hemophilia A. To understand clot development by pFVIII, we have examined clot response to laser injury in both cremaster arterioles and venules in FVIII(null) mice either infused with FVIII or transgenic for pFVIII. In both sets of vessels, pFVIII is at least as effective as infused FVIII. However, there are temporal and spatial differences in fibrin and platelet accumulation within clots depending on how FVIII is delivered. These differences may be related to the temporal and spatial distribution of the alpha-granular-released FVIII within the developing clot, and may explain the increased frequency and size of embolic events seen with pFVIII. These observations may not only have implications for the use of pFVIII in gene therapy for hemophilia A, but may also have physiologic consequences, explaining why many procoagulant factors are delivered both in the plasma and in platelet alpha-granules.

Figure 1

Platelet and fibrin accumulation in laser-induced arteriole and venule injuries in wildtype and untreated FVIII^null mice. Analysis of in situ laser injury in cremaster arterioles and venules. Average platelet accumulation was measured using an Alexa⁶⁴⁷-labeled anti-CD41 Fab fragment (RVU), and fibrin clot accumulation using an Alexa⁴⁸⁸-labeled anti-fibrin antibody (RVU). Studies of wildtype (WT) mice are shown on the top row and FVIII^null mice on the bottom row, with arterioles on the left and venules on the right for both rows. The number of individual injuries and total mice used are noted at the top of each graph. Measurements are in relative value units and held constant throughout these studies.

Figure 2

Platelet and fibrin accumulation in laser-induced arteriole and venule injuries in treated FVIII^null mice. Same as in Figure 1, but for FVIII^null mice receiving a human FVIII infusion 5 minutes before the first injury. All studies were completed within 1 hour of infusion. The estimated percentage of human FVIII antigenic correction is indicated at the top of each graph.

Figure 3

Platelet and fibrin accumulation in laser-induced arteriole and venule injuries in pFVIII/FVIII^null mice. Same as in Figure 1, but studies were done in pFVIII/FVIII^null mice that were littermates of the FVIII^null mice in Figures 1 and 2.

Figure 4

CD62P exposure in WT and pFVIII/FVIII^null mice. Studies as in Figures 1 through 3 with an Alexa⁶⁴⁷-labeled anti-fibrin antibody, but instead of measuring platelet accumulation using an anti-CD41 Fab fragment, CD62P exposure was measured using an Alexa⁶⁴⁷-labeled anti-CD62P antibody. (A) Top row shows representative frames from specific points in Videos S1 and S2 of fibrin (yellow) and p-selectin (blue) accumulation with overlap between fibrin and CD62P exposure in white. The left 2 panels present the growth of a clot from a WT arteriole injury shown at an early (20 seconds) and a late (150 seconds) time point after injury. The right panel presents growth of a clot from a pFVIII/FVIII^null arteriole injury at a late (105 seconds) timepoint. The bottom row is a set of cartoons depicting the details of the frames in the top row. Here, the whole platelet plug is shown in gray and the accumulation of fibrin in yellow, with CD62P exposure in blue and overlap in white. The entire Video for the WT injury is available as Video S1 and for the pFVIII/FVIII^null injury, Video S2. (B) Average data analysis for CD62P exposure and fibrin accumulation in WT arterioles (left) and pFVIII/FVIII^null arterioles (right) were carried out as in Figures 1 through 3 except that an Alexa⁶⁴⁷-labeled anti-CD62P antibody was used. Fibrin is depicted as □ and CD62P as ♦.

Figure 5

Embolization from WT, untreated and treated FVIII^null, and pFVIII/FVIII^null mice. (A) This panel from a venular WT injury illustrates how data were collected for these studies. Mice were injected with Alexa⁶⁴⁷-labeled anti-CD41 Fab fragments. A laser injury was made as indicated with a growing clot in red at that site, but data were collected at the indicated mask, where an embolus in red is passing by. The large arrow refers to the direction of blood flow. (B) Individual measurement of number of detected events for WT, untreated and treated FVIII^null, and pFVIII/FVIII^null mice are shown as well as the mean (red bar). Twenty separate injuries were analyzed for each study. Studies of arteriolar injuries are shown as filled symbols, while venular injuries are shown as empty symbols. Panels C and D are the same as panel B but show relative size of emboli and total amount of embolization, respectively.

Figure 6

Model of thrombus development. The top row shows the initial accumulation of platelets at a site of injury in a vessel. Subsequent rows show early fibrin clot development on the left (at ∼20 seconds), later events (at ∼150 seconds) in the middle, and degree of embolization on the right in that vessel. From top to bottom, the rows are for WT, untreated and treated FVIII^null, and pFVIII/FVIII^null mice as indicated. WT mice rapidly develop a fibrin scaffold that leads to vigorous platelet plug growth and little embolization. Untreated FVIII^null mice barely develop a fibrin clot and any accumulated platelet plug is unstable and rapidly breaks off. Treated FVIII^null mice develop a spatially and temporally correct clot, but of a limited size so that platelet plug development remains limited, but embolization approaches that seen in WT mice. The pFVIII/FVIII^null mice develop a rapid fibrin clot at the base, but it may be spatially more upstream than normal. Furthermore the core and top of the clot remain deficient in available FVIII, and the developing platelet plug is not well scaffolded. Frequent and relative large emboli then occur.