Comment on De Meyer et al, page 4728
In this issue of Blood, De Meyer and colleagues not only report the procurement of endothelial progenitor cells from dogs, but they also procure endothelial cells from a colony of dogs with severe von Willebrand disease that have the added potential to demonstrate genetic correction of the defect in VWF synthesis.
Because the cDNA for von Willebrand factor (VWF) is 8.4 kb, the potential for gene therapy using common viral vector delivery has been considered problematic. Type 3 or homozygous severe von Willebrand disease (VWD) is relatively rare and is estimated to affect 3 to 4 individuals per million.1 Although transfection of canine type 3 VWD endothelial cells has been reported previously2 in Blood, such transfection was done only with cultured endothelial cells using a plasmid containing the cDNA for VWF under the control of the cytomegalovirus (CMV) promoter. It would be difficult to procure a source of such endothelial cells in a practical manner that could be a potential strategy for in vivo gene transfer. Hebbel and coworkers3 have previously defined a mechanism for procuring circulating endothelial cells or endothelial progenitor cells from blood that permit the ex vivo expansion of endothelial cells that could be the target for gene delivery and potential subsequent re-introduction into the recipient. This had been achieved using human blood but had not been achieved in another species.FIG1
Delivery of the VWF cDNA using a retroviral vector has the potential for long-term expression of VWF synthesis, but the 8.4-kb cDNA has been thought to be a barrier to efficient viral gene transfer. DeMeyer and colleagues have been able to demonstrate efficient and effective gene transfer using a lentiviral packaging system in which VWF expression is under the control of the CMV promoter. Not only do they demonstrate that model endothelial cells can be isolated from peripheral blood of type 3 VWD dogs and that VWF synthesis can be transferred by lentiviral transduction, but the VWF that synthesized is fully multimerized and binds factor VIII (FVIII), binds to platelet glycoprotein Ib (GPIb) under expected agonist induction using ristocetin, and binds normally to collagen. Furthermore, the endothelial cells reestablish their Weibel-Palade body storage of VWF as demonstrated (see figure). Furthermore, not only can the cDNA for VWF using the CMV promoter be packaged in a lentiviral vector, but the titer of virus and the efficiency of cellular transduction suggests this is a viable approach.
This now provides an opportunity to study ex vivo gene transfer for the treatment of type 3 VWD. Whether ex vivo expansion and in vivo homing would provide therapeutic in vivo VWF levels will require additional study. Infusion of VWF in this canine VWD model has been previously demonstrated to be therapeutic.4 One observation by DeMeyer et al that could be a potential barrier to future success is the demonstrated decline in transgene expression and proliferative potential after long-term expansion. The authors point out that this phenomenon seems to be more of a problem with canine outgrowth endothelial cells than with human. It is not clear if this will be a stumbling block to long-term correction in vivo or whether the homing of these cells will be established in an environment that will permit long-term genetic correction.
There are several questions that will await further study. Optimally, does VWF need to be expressed in the endothelial cells lining our blood vessels, or is the provision of multimerized VWF to the plasma compartment from transduced cells sufficient? Is platelet expression of VWF important to optimal therapeutic benefit? The canine model of VWF synthesis varies from other species in that VWF synthesis and storage is limited to endothelial cells and does not occur in megakaryocytes (and their platelet progeny). Thus, correcting the endothelial compartment will normalize dogs, because they do not use platelets for VWF storage and release. In other animals, including humans, VWF has both endothelial and megakaryocytic synthesis and storage. Since VWF infusion alone is therapeutically beneficial in humans, the source of cellular released VWF may be more dependent on the quantity and quality of the VWF secreted into plasma. Blood-outgrowth endothelial-cell VWF appears to meet the need for quality of VWF, but whether adequate quantity will be achieved in vivo will only be answered by further study. ▪