A Novel VWF A3 Domain Mutation with Absent Binding to Type I and Type III Collagen.

Abstract 541

Von Willebrand factor (VWF) plays a key role in coagulation by tethering platelets to injured subendothelium through binding sites for collagen and platelet GPIb. Laboratory testing for von Willebrand disease (VWD) utilizes various assays of VWF protein and function, but collagen binding (VWF:CB) has not always been performed. Samples obtained from a large, multi-center US study of von Willebrand disease (VWD), the Zimmerman Program for the Molecular and Clinical Biology of VWD (ZPMCB-VWD), were evaluated for VWF:CB. Whole blood was collected for DNA sequencing and plasma samples were collected for VWF testing. VWF antigen (VWF:Ag), ristocetin cofactor activity (VWF:RCo), and VWF:CB were performed in the BloodCenter of Wisconsin hemostasis laboratory. A normal range for the VWF:CB was established using 233 healthy controls enrolled in the ZPMCB-VWD. The mean VWF:Ag was 105 IU/dL and the mean VWF:CB was 123 U/dL, with a mean VWF:CB/VWF:Ag ratio of 1.19. This yields a normal range of 0.86-1.51 (mean ± 2 SD). No sequence variations in the VWF A3 domain affecting VWF:CB were observed in the healthy controls. Collagen binding was then examined in subjects enrolled with a diagnosis of VWD. The index case in one family was noted to have a low VWF:CB/VWF:Ag ratio of 0.54. The VWF:Ag was 41 IU/dL, VWF:RCo was 44 IU/dL, VWF:CB was 22 U/dL, and multimer distribution was normal. Decreased VWF:CB/VWF:Ag was found in the father and sister of the index case, while the mother had normal VWF:CB/VWF:Ag. The index case initially presented with epistaxis and easy bruising. Her bleeding score was 8, as determined by the revised European Union questionnaire. DNA sequencing of exons 28-32 was performed and revealed a heterozygous mutation (5356C>G) predicting replacement of wild-type histidine 1786 by aspartic acid in all affected family members. This amino acid is located in a critical face of the A3 loop. The 1786D mutation was cloned into a full-length VWF expression vector and expressed, as was a previously reported collagen-binding mutation, 1731T. Collagen binding assays were performed using type I human placental collagen at 5 μg/mL or type III human placental collagen at 1 μg/mL to capture either plasma or recombinant VWF (rVWF). Both mutations had normal expression, at 97% and 87% of the expression seen with wild-type VWF, respectively. Multimer distribution was also normal, including presence of high molecular weight multimers. Plasma from family members heterozygous for the H1786D mutation showed a mean VWF:CB/VWF:Ag ratio of 0.23 ± 0.05 (mean ± SD) for type I collagen and 0.50 ± 0.11 for type III collagen. When the rVWF constructs were examined, the novel 1786D mutation showed essentially no binding to either type I or type III collagen, at <1% of wild-type rVWF. In contrast, the previously reported collagen binding mutation 1731T showed VWF:CB/VWF:Ag ratios of 45% for type I and 50% for type III collagen when compared to wild-type rVWF. As the proband had a minimally decreased VWF:Ag, we then examined collagen binding in ZPMCB-VWD subjects with VWF:Ag between 30 and 50 IU/dL. The mean VWF:CB/VWF:Ag ratio in this group was 1.02, with a range of 0.62 to 1.28. These results, taken together with the data in the healthy controls, suggest that decreased VWF:CB/VWF:Ag ratios may identify a specific defect in VWF function. In particular, the H1786D mutation appears to result in a complete loss of VWF's ability to bind either type I or type III collagen. The subjects with this mutation were all heterozygous, which may account for reduction, rather than absence, of in vitro VWF:CB. The VWF:CB may be useful in diagnosis of VWD, and a decreased VWF:CB/VWF:Ag ratio in the face of normal VWF multimer distribution may reflect specific loss of collagen binding ability leading to bleeding symptoms.