Abstract
Background
Factor VIII (fVIII) functions as a co-factor for factor IXa on the membranes of stimulated platelets. Binding sites for fVIII(a) are expressed at two levels; thrombin induces 3,000 – 20,000 sites/platelet while the combination of collagen and thrombin or A28137 induce >50,000 sites/platelet.
Hypothesis
We hypothesized that binding sites for fVIII(a) on thrombin-stimulated platelets, are distinct from phosphatidylserine (PS), while those on maximally stimulated platelets are predominantly PS-containing sites. Corollaries were 1) that epitopes on fVIII interact with the non-PS sites and 2) that a macromolecule or a macromolecule complex comprises the binding sites on thrombin-stimulated platelets.
Methods
Platelets were purified on a density gradient and binding of fluorescein-labeled fVIII (fVIII-fluor) was measured by flow cytometry using a Becton Dickinson LSR-Fortessa flow cytometer. Factor VIII activity was measured in a discontinuous factor Xase assay using extruded phospholipid vesicles of composition PS:PE:PC 4:20:76 or platelets as the membrane source. Oligomeric fibrin was immobilized by incubating thrombin, 1 u/ml, with fibrinogen, 10 µg/ml for 10 min without mixing prior to addition of 59D8-Superose beads. Binding of fVIII-4 Ala to platelets was measured in complex with Alexa-488 labeled mAb GMA-8021, against the A2 domain. Polyphosphate was size-fractionated and recombinant PPX-MBD produced as previously described.
Results
Lactadherin, a phosphatidyl-L-serine-binding protein, competed for 97% of factor VIII-fluorescein (fVIII-fluor) binding sites on A23187-stimulated platelets but only 30% of binding sites on thrombin-stimulated platelets. Unlabeled fVIII competed with fVIII-fluor for all binding sites. A fVIII C2 domain mutant, with no measurable phospholipid binding - M2199A/F2200A/L2251A/L2252A (fVIII-4Ala) bound to only 3,000 – 5,000 sites on platelets stimulated with A23187 but to a similar number on thrombin-stimulated platelets with a KDof 7 nM. These data indicate that non-PS sites are dominant on thrombin-stimulated platelets but that PS-containing sites comprise at least 95% of sites on A23187-stimulated platelets.
We evaluated a panel of mAb’s against the fVIII-C2 domain for platelet-specific inhibition of binding and function. mAb’s ESH4 and I54, with overlapping epitopes, blocked binding of fVIII to thrombin-stimulated platelets but only decreased affinity for PS-containing membranes. In 1-stage and 2-stage commercial aPTT assays ESH4 inhibited 28-33% of fVIII activity. In contrast, ESH4 inhibited 80% of fVIII activity on thrombin-stimulated platelets. mAb’s ESH8 and G99, with partially overlapping epitopes, decreased the affinity of fVIII-fluor for thrombin-stimulated platelets approx. 70% but had no effect on phospholipid binding. ESH8 inhibited 58 ± 8% of fVIII activity on thrombin-stimulated platelets but did not decrease activity supported by phospholipid vesicles.
Because oligomeric fibrin is required for expression of most fVIII binding sites on thrombin-stimulated platelets (Phillips et al 2004; JTH 2:1806) we hypothesized that oligomeric, platelet-bound fibrin is a constituent of fVIII binding sites. fVIII-fluor bound to fibrin monomers and oligomers immobilized on mAb 59D8-Superose, detected in solution by flow cytometry. Binding was enhanced by mixing polyphosphate (polyP) with fibrinogen prior to thrombin, with a maximum gain in affinity at 0.1 µM elemental phosphorous. The apparent affinity of fibrin-polyP for fVIII-fluor was 2-12 nM, based on competition studies with unlabeled fVIII. Like binding to platelets, specific binding of fVIII to fibrin-polyP was blocked by mAb’s ESH4, I54 and diminished by ESH8, and G99. Thrombin-stimulated platelets, but not resting platelets, exhibited bound polyP, as detected by PPX-MBP, specific for polyP. Thus, bound polyP is present on thrombin-stimulated platelets under conditions that lead to binding of oligomeric fibrin.
Conclusions
These data indicate that thrombin-stimulated platelets bind fVIII via a non-PS binding site and that the binding is mediated by epitopes that have greater functional importance on platelets than on phospholipid vesicles. Platelet-bound oligomeric fibrin with polyP is a candidate for the non-PS binding site. These findings have clinical relevance to detection of inhibitory antibodies against fVIII.
Disclosures:
No relevant conflicts of interest to declare.
Specific phospholipid binding to Na,K-ATPase at two distinct sites