von Willebrand factor in its native environment

In this issue of Blood, Starke et al¹  and Wang et al²  study endothelial cells cultured directly from the blood of patients with von Willebrand disease (VWD), and thereby overcome some limitation of studying VWD mutations in transfected cell systems that cannot reproduce key aspects of von Willebrand factor (VWF) expression in endothelium.

VWD is a relatively common bleeding disorder that is caused by mutations in the gene encoding VWF. VWF has binding sites for several physiological ligands, and mutations can impair VWF assembly, intracellular storage, secretion, and biological function in a bewildering number of ways to produce complex disease phenotypes. Furthermore, VWF is a multimeric protein assembled from the products of 2 alleles, which makes it possible for a single bad allele to disrupt the entire protein.

We have made a lot of progress understanding VWD by expressing recombinant mutant VWF in transfected cells and seeing what happens. For example, these studies have shown how defects in multimer assembly or stability lead to VWD type 2A, how gain-of-function mutations with exaggerated affinity for platelets produce VWD type 2B, and how defects in binding to factor VIII cause VWD type 2N, an autosomal recessive mimic of hemophilia A. In the last few years, we have also learned a great deal about the pathophysiology of VWD type 1, which is the most common type of VWD but in many ways is also the most perplexing variant.

However, several inherent features of heterologous expression systems limit what can be learned. Most nonendothelial cell types do not target recombinant VWF to Weibel-Palade-like organelles and cannot be used to assess VWF storage or regulated secretion. Patients are usually heterozygous, but modeling the heterozygous state in transfected cells is difficult. Transfected cDNA constructs usually are overexpressed with strong viral promoters, which can introduce artifacts. In addition, focusing exclusively on exons ignores mutations that may reduce transcription, alter splicing, or change mRNA stability. In a few fortuitous cases, these limitations have been overcome by characterizing the behavior of mutant VWF genes in human umbilical vein endothelial cells that were obtained as the patient was born,^3-5  but this approach is rarely feasible.

Alternatively, cells can be isolated from peripheral blood that have many properties of vascular endothelial cells.⁶  These blood outgrowth endothelial cells (BOEC) express endothelial cell surface antigens, store VWF in Weibel-Palade bodies, form capillary-like tubules when cultured in Matrigel, organize into blood vessels in vivo, and proliferate in culture for several passages.⁷  In a nice proof-of-principle study, BOECs from a compound heterozygous patient with VWD type 2N were used to identify abnormal VWF mRNA splicing, a premature stop codon, markedly impaired intracellular transport, and abnormal storage of VWF in Weibel-Palade bodies. None of these effects could be discovered by the expression of sequence variants in the exons of the patient’s VWF genes.⁸ 

Wang et al²  and Starke et al¹  have built on this foundation, studying BOECs from total of 8 patients with VWD type 1 and 5 patients with VWD type 2 variants. The phenotype of BOECs was stable for at least 8 passages in culture. For a single heterozygous patient, polymerase chain reaction demonstrated reduced transcription of the mutant VWF gene. Although patients with VWD type 1 had decreased levels of VWF mRNA, as might be expected, a striking finding was that many had complex phenotypes with increased intracellular retention in the endoplasmic reticulum, abnormal storage in misshapen Weibel-Palade bodies, and impaired secretion of functional VWF “strings” on the endothelial cell surface. The behavior of VWF in BOECs appears to correlate well with clinical features of VWD, including whether a patient responds satisfactorily to treatment with 1-desamino-8-D-arginine vasopressin (desmopressin).

These reports demonstrate that BOECs can provide unprecedented insight into the pathogenesis of VWD. However, additional experience will be needed to assess the reliability of the approach. The reproducibility of BOEC phenotypes needs to be established more completely, comparing different isolates from the same donor. Success in isolating BOECs seems to be donor-dependent. VWF expression varied significantly between healthy donors, and some donors never yielded BOECs. Nevertheless, BOECs enable studies of VWF phenotypes that are not otherwise accessible. For the first time, it appears possible to analyze the interactions between VWF alleles in a patient’s own endothelial cells to evaluate the interplay between alleles at the level of transcription, splicing, mRNA stability, multimer assembly, Weibel-Palade body morphology, and response to secretagogues. In this respect, the study of BOECs may represent a new standard for the characterization of mutations in VWD, and perhaps for other diseases that affect endothelial cells.