Bleeding Phenotype Observed in Murine Defect of VWF-Collagen 4 Interactions

VWF binding to type 4 collagen has recently been shown to occur via the VWF A1 domain, but the clinical implications of this interaction remain to be elucidated. In humans, the R1399H sequence variant (2% of the Caucasian population is heterozygous for this variant) abrogates VWF binding to collagen 4, while select other VWF A1 domain variants decrease VWF binding to collagen 4 to varying degrees. We generated a murine model of 1399H VWF to investigate the in vitro and in vivo effects of this variant more thoroughly.

Mice expressing the murine R1399H mutation were generated via gene targeting in ES cells, generation of germline-competent ES cell chimeras, and backcrossing the resultant mutant mice onto a C57Bl6 background. VWF antigen (VWF:Ag) and VWF collagen binding with murine type 4 collagen (VWF:CB4) were measured by ELISA (n≥5 for each group) on wild-type p.R1399 mice (WT), heterozygous p.R1399H mice (R1399H) and homozygous p.H1399 mice (1399HH) and compared to mice lacking VWF entirely (VWF -/-). Tail bleeding time was performed by clipping a 3 mm segment of tail, immersing in warm saline, and monitoring bleeding for a maximum of 10 minutes (n≥4 for each group). Blood loss was also measured during the tail bleeding time assay for each mouse. VWF-dependent platelet adhesion to collagen 4 was measured under flow using the Venaflux system. Channels were coated with murine type 4 collagen overnight. Blood was drawn from the inferior vena cava, anticoagulated with PPACK and heparin, and platelets labeled with mepacrine. Whole blood samples were run over the collagen 4-coated channel at a shear rate of 1111/sec (n≥3 for each group). Images at 180 seconds following the start of flow over the chip were analyzed using ImageJ.

Breeding of heterozygous and homozygous 1399H mice was observed to follow normal Mendelian ratios, indicating no deleterious effects of the 1399H variant on viability. No spontaneous bleeding was observed for any of the offspring. VWF expression as measured by VWF:Ag was similar for WT, R1399H, and 1399HH mice, but VWF:CB4/VWF:Ag ratios were decreased for R1399H mice, to an average of 55% of WT and 1399HH mice, to an average of 15% of WT. Blood loss during tail bleeding was increased compared to WT for both R1399H and 1399HH mice by approximately 1.8 fold and 2.8 fold respectively. Bleeding time also increased, with an average bleeding time of 92 seconds in the WT mice, 177 seconds in the R1399H mice, and 235 seconds in the 1399HH mice. Previous studies have demonstrated that platelet adhesion to collagen 4 under flow is dependent on VWF, with VWF -/- mice demonstrating absence of platelet adhesion. Platelet adhesion to collagen 4 was diminished in both heterozygous and homozygous 1399H mice, with maximum platelet binding intensity 61% of WT for the R1399H mice and 28% of WT for the 1388HH mice. These figures closely approximate the static collagen 4 binding as seen by comparable VWF:CB4/VWF:Ag ratios and platelet adhesion under flow for the R1399H heterozygous and 1399HH homozygous mice.

These results show that a decrease in VWF's ability to bind collagen 4 under static conditions corresponds to a decrease in binding under flow conditions, at least at the measured shear rate, and that this defect also translates to increased bleeding time and blood loss in a tail clip model. This study supports the potential for a bleeding phenotype in patients with aberrant VWF-collagen 4 binding. Since 2% of the US population is heterozygous for the p.R1399H sequence variation, defective VWF interaction with collagen 4 may be a relatively common cause of clinical bleeding and can be modeled in engineered mice.