Quantitative Measurement of VWF Binding to the Platelet Integrin αIIbβ3 - Abnormalities Identified in Individuals with VWD

Most studies of VWF function have focused on VWF interactions with platelet GPIb, FVIII, and specific sites and types of collagen. Early studies carried out by Ruggeri and Montgomery demonstrated that in addition to binding GPIb on platelets, VWF also bound to the GPIIb-IIIa receptor (also known as integrin αIIbβ3), but only in the absence of fibrinogen. Ruggeri and other groups have demonstrated that under flow, VWF interaction with GPIIb-IIIa is also critical to VWF-platelet biology. A large systematic study of VWF binding to GPIIb-IIIa (αIIbβ3) has not been undertaken previously in a cohort of type 1 and 2 VWD subjects. Studies were carried out on 619 VWD index cases (both type 1 and type 2 VWD) and 84 normal controls enrolled in the Zimmerman Program for the Molecular and Clinical Biology of VWD (ZPMCB-VWD) using a novel assay of αIIbβ3 binding of VWF.

Zhu and coworkers expressed the soluble αIIbβ3 headpiece and previously characterized this in the open position when bound to ligand. This αIIbβ3 headpiece was used for our studies. An antigen capture ELISA was developed using the αIIbβ3 headpiece to capture VWF and to compare binding to the VWF:Ag as a ratio of VWF-αIIbβ3/VWF:Ag. ELISA plates were coated with the GPIIIa-specific monoclonal antibody (MAb), AP3 and used to capture the αIIbβ3 headpiece. Since fibrinogen blocks VWF binding to αIIbβ3, plasma samples were heated at 56°C to precipitate fibrinogen, and then centrifuged to remove the precipitate. The resulting heat defibrinated plasma was utilized in the assay. Bound VWF was quantified using biotin-labeled AVW15, a MAb to VWF. The reference was the ISTH SSC Plasma Standard. As specific positive and negative controls, we used wild-type recombinant VWF containing the normal RGDS sequence at position 2507-2510, and mutant VWF containing the RGDE sequence at the same position. As an additional control, the monoclonal antibody 7E3, known to block aggregation by binding to β3 was used to block αIIbβ3/VWF interactions. While binding of wt-VWF to αIIbβ3 was robust, no binding of the RGDE mutant was observed (<1% of wt). 7E3 completely inhibited binding of VWF to αIIbβ3. Interestingly, two other MAbs known to block the binding of Fg to αIIbβ3, AP2 and 10E5, did not inhibit VWF binding. The 84 normal controls had a mean VWF-αIIbβ3/VWF:Ag binding ratio of 0.87 with a 5% and 95% confidence limit of 0.53-1.39. 152 type 1 VWD (VWF <30IU/dL) samples had a modestly reduced binding ratio of 0.62 (0.36-0.85) with a significance of p<0.0001. The binding ratios of 55 type 2A and 41 2B VWD samples were markedly reduced with a ratio of 0.293 (0.12-0.67) and 0.40 (0.23-0.81) both with a significance of 0.0001. Since both of these latter VWF variants have abnormal VWF multimers, we tested recombinant D-pro VWF (no VWF propeptide present) and Y87S mutant VWF (absent VWF multimerization). No binding to αIIbβ3 was observed with VWF from either of these constructs. Among our ZPMCB-VWD index cases, we also had individuals with low VWF (VWF:Ag 30-50 IU/dL) that had an intermediate reduction between the normal controls and type 1 subjects. Type 2N VWD samples demonstrated no abnormal binding to αIIbβ3 and type 2M subjects had minimal differences.

While VWF binding to αIIbβ3 has been recognized previously, it has not been systematically studied in subjects with VWD. We identified a modest decrease in αIIbβ3 binding of VWF in type 1 VWD and in clinical subjects with low VWF. Furthermore, type 2A and 2B subjects have a much more profound reduction in αIIbβ3 binding and suggest an importance of normal multimeric VWF to the functional binding of VWF to αIIbβ3. Mutating the RGDS sequence in VWF, or adding 7E3 to block αIIbβ3, abrogate this VWF binding. To date we have not identified any of our enrolled ZPMCB-VWD subjects with mutation of the RGDS sequence or in the C4 domain of VWF, although mutations in this region recently have been reported by Legendre et al. at the ISTH2015 meeting. Such mutations are not identified with current VWF functional screening assays, but specific assays of αIIbβ3-VWF interaction, such as the one described here, can be used and suggest there are specific potential mutations and variants with abnormal multimers that have abnormal αIIbβ3 binding. The qualitative and quantitative assessment of VWF function continues to be complicated.