The Factor VIIII C2 Domain Shows Stereospecificity for Phosphatidyl-L-Serine and Increases Factor IXa Activity

Background: Factor VIII is an essential cofactor in the blood coagulation cascade and shows greatly increased activity when bound to phospholipid membranes. While high proportions of various negatively-charged phospholipids support binding of factor VIII, only phosphatidyl-L-serine (Ptd-L-Ser) supports stereospecific affinity when present at physiologically relevant proportions. Factor VIII binds to phospholipid membranes primarily through its C2 domain, but binding is also mediated by the C1 domain (Meems et al., abstract ASH 2008). However, the membrane-binding properties of the isolated C2 domain and the relationship of the C2 domain to factor IXa and factor X have not been fully studied.

Methods: The factor VIII C2 domain (fVIII-C2) and a mutant in which residue Val2223 was replaced by Cys (fVIII-C2_C) were produced in E. coli and purified from cytosol by metal ion affinity chromatography followed by cation exchange chromatography. fVIII-C2_C was labeled with fluorescein maleimide (fVIII-C2_C-fluor). Binding studies were performed by flow cytometry using membranes supported by glass microspheres (lipospheres) and by fluorescence resonance energy transfer between fVIII-C2 and dansyl-labeled vesicles. The factor Xase assay was utilized to infer the capacity of fVIII-C2 to influence membranebinding and protein-protein interactions.

Results: fVIII-C2_C-fluor bound to liposphere membranes containing at least 10% Ptd-LSer. The K_D for binding to lipospheres was 150 ± 40 nM. The K_D for fVIII-C2 binding to sonicated vesicles of composition Ptd-L-Ser:PE-dansyl:PC 20:5:75 was 230 ± 30 nM indicating that fVIII-C2 and fVIII-C2_C-fluor bind with similar affinities. Binding was measurable at pH 6.0 but was at least 10-fold lower affinity at pH 7.8. Change of pH from 6.0 to 7.8 was associated with a change in intrinsic fluorescence and was not associated with increased light scatter, suggesting conformational change rather than formation of dimers or aggregates. Sonicated vesicles with 15% Ptd-L-Ser competed with lipospheres for binding of fVIII-C2_C-fluor with 3-fold higher affinity than vesicles with 15% Ptd-D-Ser. Phosphatidylinositol or phosphatidic acid-containing vesicles bound fVIII-C2_C-fluor with at least 4-fold lower affinity than Ptd-L-Ser. fVIII-C2 did not compete with factor VIII-fluor for binding to lipospheres at pH 6 or 7.8. At pH 6.0 the factor Xase assay was inhibited by addition of fVIII-C2 with a plateau of 30% inhibition at 1.5 μM. In the absence of intact factor VIII, fVIII-C2 enhanced the activity of factor IXa by 2-fold. The enhanced activity correlated to a reduced K_M but not an altered apparent affinity for phospholipid vesicles.

Conclusions: These results indicate that fVIII-C2 binds membranes containing Ptd-L-Ser in a stereospecific manner with approximately 30-fold lower affinity than intact factor VIII. Membrane binding is pH-dependent, likely requiring a conformational change in fVIII-C2. Lack of competition with intact factor VIII implies that fVIII-C2 does not recognize the initial membrane contact site of the full protein. Partial inhibition of the factor Xase complex and enhancement of factor IXa activity in the absence of intact factor VIIIa implies that fVIII-C2 binds either factor IXa or factor X. We speculate that the conformational change enabling membrane binding is caused by an acidic microenvironment in intact factor VIII produced by proximity to the C1 domain and/or the phospholipid membrane.