Low Platelet ₂β₁ Levels in Type I von Willebrand Disease Correlate With Impaired Platelet Function in a High Shear Stress System

PLATELET ADHESION TO collagen in damaged vascular subendothelium is of pivotal importance to both the initiation of platelet thrombus formation and procoagulant activity.1This first step in thrombus formation is mediated by both von Willebrand factor (vWF)–dependent recruitment of the glycoprotein Ib–IX receptor complex and collagen interaction with the integrin, α₂β₁.2 The integrin, α₂β₁, mediates platelet adhesion to both fibrillar (types I, III, and V) or nonfibrillar (types IV, VI, VII, and VIII) collagen.3-7 While the levels of the integrins α₅β₁ (fibronectin receptor) and α_IIbβ₃ (fibrinogen receptor) can vary from one individual to the next, this difference never exceeds a fraction of the mean population level. However α₂β₁levels can vary up to 10-fold in the general population, and this correlates with differences in adhesiveness to type I or type III collagens.8 This variation in α₂β₁ receptor density is associated with differential inheritance of three alleles of the human α2 gene that can be distinguished by multiple sequence differences in selected introns and silent sequence differences in a segment of the coding region.9,10 Because there is a correlation between platelet adhesion to collagen and α₂β₁ receptor levels, differences in α₂β₁ receptor levels may influence the risk for thrombosis or bleeding in vivo. While the risk is apparently negligible among healthy individuals, it can become an important factor in individuals who are otherwise predisposed toward thrombosis or bleeding by an unrelated genetic or acquired condition, such as von Willebrand disease (vWD).

Among inherited bleeding disorders, vWD is certainly the most common, with a reported prevalence ranging from 0.8% to 1.3%.11,12 The pathogenesis of vWD is based on quantitative and/or qualitative abnormalities of vWF,13 a large, multimeric protein, encoded by a gene in chromosome 12, which circulates in the blood plasma, but is also stored in granules of endothelial cells and platelets.14 The most common form of vWD, known as type 1, accounts for more than 70% of patients and is caused by a quantitative deficiency of vWF.15 Type 1 is very heterogeneous and several subtypes have been described according to platelet vWF content.16 Moreover, in type 1 vWD, platelet vWF cooperates with plasma vWF to determine the degree of abnormalities in platelet-vessel wall interaction.17 

Aside from the variation in vWF levels from one patient to the other, it is difficult to explain the significant variation in bleeding tendencies observed between patients with similar laboratory profiles and similar levels of vWF. Thus, genetic and nongenetic factors may augment or attenuate the likelihood of bleeding symptoms in these patients.17 We postulate that variation in the integrin α₂β₁ density on the platelet surface might be a major factor that influences bleeding in patients with type 1 vWD. To test this hypothesis, we determined the α₂ genotype in 148 patients with vWD by using restriction digest analysis of genomic DNA. To show correlation between collagen receptor density and platelet function, we also performed platelet adhesion and aggregation studies in the platelet function analyzer (PFA-100, Behring, Dade, FL), a high shear stress system that simulates platelet-based primary hemostasis in vitro.18 

We studied 148 symptomatic patients diagnosed with one of five types of vWD as follows: type 1, n = 78; type 2A, n = 25; type 2B, n = 14; type 2M, n = 11; and type 3, n = 20. The diagnosis of vWD was based on laboratory findings (plasma vWF levels, ristocetin cofactor activity [vWF:RCo], factor VIII coagulant activity, and multimer analysis), family and personal history of bleeding, and a prolonged bleeding time. We also studied 99 age-sex–matched Italians and 68 non-Italian whites without evidence of bleeding disorders and used them as controls. Patient blood samples were collected at the Angelo Bianchi Bonomi Hemophilia and Thrombosis Center, Milano, Italy and shipped to Children’s Hospital of Orange County (CHOC), Orange, CA. An additional 15 patients samples were harvested and processed at CHOC. Informed consent was obtained in all cases following institutional guidelines.

Genomic DNA was isolated from peripheral blood mononuclear cells by using the method of Miller et al.19 The desired segment of the α₂ gene (581 kb) was amplified by polymerase chain reaction (PCR) using a Perkin Elmer Cetus DNA Thermal Cycler (Norwalk, CT) and Taq polymerase (Pharmacia Biotech, Piscataway, NJ). The primers used in the PCR reaction were: (1) 5′ GATTTAACTTTCCCGACTGCCTTC 3′ and (2) 5′ CATAGGTTTTTGGGGAACAGGTGG 3′. The PCR cycling conditions were: 94°C × 10 minutes, one cycle, followed by 94°C × 1 minute, anneal 1 minute, 72°C × 1 minute, where the annealing was 69°C for two cycles, 67°C for two cycles, and 65° for 35 cycles. The PCR product is a 581-bp segment located within the intron separating two exons that encode the allelic polymorphisms at bp 807 and 873. Genomic DNA cloned from individuals homozygous for 807C or 807T served as positive controls.

Digestion of the PCR product was performed with the restriction enzyme Bgl II (Pharmacia Biotech). Briefly, for each 40-μL reaction, 15 μL of the PCR product, 4 μL of 10x restriction buffer (One-Phor-All buffer Plus, Pharmacia Biotech), and 1 μL of Bgl II enzyme (× U/mL) were used. The PCR product was digested at 37°c for at least 2 hours and analyzed on a 1.5% agarose gel in 1x tris acetate EDTA (TAE) buffer. The gel was stained with ethidium bromide and visualized under ultraviolet (UV) light.

Platelet adhesion and aggregation was evaluated in 32 patients with vWD type1 in the PFA-100, as previously described by Fressinaud et al.20 The PFA–100 assay was performed with two different types of cartridges (collagen/epinephrine and collagen/adenine diphosphate [ADP]), and values shown represent the mean of duplicate measurements. The statistical analysis of the results was performed using t-test.

We have previously shown that there are at least three alleles of the human α₂ gene distinguishable by a cluster of silent and/or noncoding sequence differences within a 4-kb segment of the gene (Fig 1).10 A naturally occurring restriction site for the enzyme, Bgl II, was found in intron G of the 807T allele. Thus, PCR products amplified from the 807T allele and spanning the Bgl II site can be digested with Bgl II to generate two fragments of 343 and 238 bp, while product amplified from the 807C allele remains undigested. Bgl II digestion of the 581 bp PCR product amplified from genomic DNA of an individual heterozygous for these alleles yields three fragments (581 bp, 343 bp, 238 bp) (Fig 2). The gene frequencies of the 807C and 807T alleles in the control group (n = 167) were .58 and .42, respectively.

Genomic DNA from 148 individuals diagnosed with vWD was analyzed, and these results were compared with the control group as seen in Table 1. By χ² analysis, it was determined that the frequency of 807C allele associated with low collagen receptor density in type 1 vWD patients is higher than that of the normal population (P = .007), the frequencies of these alleles being .71 for 807C and .29 for 807T. In the type 2M patients, the P value was significant at P = .050, albeit not as strong an association as that seen with type 1 vWD patients. On the other hand, there is not a significant difference in allele frequencies among patients with vWD type 2A (P = .246), 2B (P = .353), or 3 (P = .351), compared with the normal group.

Platelet adhesion was analyzed in 32 patients with vWD type 1 in the PFA-100 and results are shown in Fig 3. At critically low vWF:RCo levels (<30 U/dL), the deficiency of vWF overrides the allele effect on closure time in the PFA-100. When vWF:RCo levels reach values between 40 to 80 U/dL, as seen in many type 1 patients, the effect of the allele on platelet adhesion to collagen is evident. As seen in Fig 3, patients with similar or identical vWF:RCo levels in the 40 to 80 U/dL range, the closure time is significantly longer in those patients who are homozygous for the 807C allele (open circles). By t-test, the differences in closure times between the patients homozygous for the 807C allele and the patients homozygous for the 807T are statistically significant, whether all the subjects with vWF:RCo greater than 40 (P = .018) or just the ones between 40 and 80 (P = .003) are included.

Multiple advances in the molecular characterization, diagnosis, and treatment of vWD have occurred over the last few years. However, it is unclear why certain vWD type 1 patients develop more severe bleeding symptoms than other patients with equivalent vWF levels.21,22 Differences in severity of bleeding can even be detected in individuals with equivalent antigen level in the same immediate family. Our findings suggest that the level of α₂β₁ on platelets may represent one of the additional factors that determines severity of bleeding in this particular disease.

It is possible that genetic factors directly related to the vWF molecule and its mode of expression could also influence variability of clinical presentation. One such factor is multiple heterozygosity for mutations in the vWF gene.23 Furthermore, there is an association of O blood type with vWF levels in the lower range of normal.24 In addition, D’Alessio et al25 have emphasized the important contribution of platelet vWF to platelet adhesion to collagen in vWD type 1.

Other genetic factors, unrelated to the vWF gene, could also be involved. Our results are consistent with the hypothesis that a low level of the α₂β₁ receptor may result in delayed primary platelet adhesion in patients with vWD type 1. Studies have shown that the primary mechanism of platelet attachment to the subendothelium is dependent on GPIb-IX–vWF interactions. A secondary mechanism involving α₂β₁ binding to the subendothelial collagen also plays an important role.10,26-28The importance of α₂β₁ as a primary receptor that mediates platelet adhesion to collagen under flow conditions is underscored by the report of a patient lacking glycoprotein Ia (α₂) whose platelets failed to respond to collagen.7,27-29 Thus, a lower platelet α₂β₁ density, which has been shown to correlate with decreased adhesiveness to collagen, together with a low vWF level is likely to further impair the adhesion of the platelet to damaged endothelium.

Our studies on individuals with type 1 vWD show a significantly higher prevalence of the 807C allele, which is associated with low α₂β₁ density. As the patients with more bleeding are more likely to be diagnosed, this increased association of the low receptor phenotype in type 1 vWD suggests that the 807C homozygous state may account for the variability in bleeding in vWD. The studies in the PFA-100 show the influence of the α₂β₁ receptor density in platelet adhesion in vWD type 1 patients who have a vWF:RCo activity in the 40 to 80 U/dL range. In patients with vWD type 1 whose vWF:Rco activity is less than 30 U/dL, the collagen receptor density does not seem to play a major role in closure time due to the fact that these patients’ platelet adhesion is extremely impaired. This may explain why the group of patients with type 3 vWD did not show a significant difference of allele frequencies when compared with the general population.

In most of the cases with type 2 and 3 vWD, a genetic defect has been correlated with moderate to severe bleeding symptoms, while no clear definition of the molecular basis of such defects is available for mild-moderate type 1 vWD. Therefore, additional information about the nature of bleeding in the latter group of patients will be very useful. We are currently performing expanded studies in families with vWD type 1 to evaluate platelet α₂β₁ receptor density variation as a possible mechanism for disparity in bleeding symptoms. This newly described genetic marker must be included in larger patient studies to determine its true role as a risk factor for increased bleeding.

The authors would like to thank Peggy Nakagawa, MS and Diana Rozenshteyn, BS for their excellent technical assistance and Dr Shirley Williams for her help and suggestions.