Phenotypic correction of von Willebrand disease type 3 blood-derived endothelial cells with lentiviral vectors expressing von Willebrand factor

Abstract

Von Willebrand disease (VWD) is an inherited bleeding disorder, caused by quantitative (type 1 and 3) or qualitative (type 2) defects in von Willebrand factor (VWF). Gene therapy is an appealing strategy for treatment of VWD because it is caused by a single gene defect and because VWF is secreted into the circulation, obviating the need for targeting specific organs or tissues. However, development of gene therapy for VWD has been hampered by the considerable length of the VWF cDNA (8.4 kb [kilobase]) and the inherent complexity of the VWF protein that requires extensive posttranslational processing. In this study, a gene-based approach for VWD was developed using lentiviral transduction of blood-outgrowth endothelial cells (BOECs) to express functional VWF. A lentiviral vector encoding complete human VWF was used to transduce BOECs isolated from type 3 VWD dogs resulting in high-transduction efficiencies (95.6% +/- 2.2%). Transduced VWD BOECs efficiently expressed functional vector-encoded VWF (4.6 +/- 0.4 U/24 hour per 10(6) cells), with normal binding to GPIbalpha and collagen and synthesis of a broad range of multimers resulting in phenotypic correction of these cells. These results indicate for the first time that gene therapy of type 3 VWD is feasible and that BOECs are attractive target cells for this purpose.

Figures

Figure 1.

Schematic representation of the Lenti-CMV-huVWF vector. This self-inactivating (SIN) lentiviral vector (11 kb) expresses the full-length human VWF cDNA (8.4 kb) from the human cytomegalovirus (CMV) promoter. The central polypurine tract (cPPT), which facilitates intranuclear import of the lentiviral preintegration complex, is depicted. The packaging signal (Ψ+) and viral long terminal repeats (LTRs) are indicated. The 3′ LTR contains a self-inactivating deletion which on transduction of the target cells is copied onto the 5′ LTR, rendering it inactive while preserving the activity of the internal CMV promoter.

Figure 2.

Phenotype of normal and VWD canine BOECs and normal human BOECs by phase contrast and fluorescence microscopy. Morphology (A), immunostaining for ac-LDL uptake (B), and VWF (C) on normal canine BOECs; morphology (D), immunostaining for ac-LDL uptake (E), and VWF (F) on canine VWD BOECs. Immunostaining for ac-LDL (G), VWF (H), CD31 (I), and CD144 (J) on normal human BOECs. Pictures are representatives of immunostainings of BOECs isolated from 4 different VWD type 3 dogs, from 1 normal dog, and from 3 healthy human control subjects. Magnification in panels B, C, E, and F is 5 times the magnification in panels A and D. (K) VWF in conditioned medium from normal canine BOECs, canine VWD BOECs, and normal human BOECs, harvested before passing the subconfluent monolayer of cells, and in conditioned medium alone. Data are mean ± SEM (n = 3) with concentration of bovine VWF (0.035 μg/mL) present in the conditioned medium not included.

Figure 3.

Assessment of functional quality of Lenti-CMV-huVWF expressed by COS-7 cells or by CHO-K1 cells. (A) COS-7 cells were transfected with Lenti-CMV-huVWF or pNUT-VWF, and serum-free expression medium was harvested, pooled, and concentrated. (B) CHO-K1 cells were transduced with Lenti-CMV-huVWF, and serum-containing medium was harvested. VWF:Ag (▪), VWF:RCo (□), and VWF: CBA () activities were determined in ELISA assays. VWF:RCo and VWF:CBA were determined by measuring the binding of VWF to rGPIbα in the presence of ristocetin and to human collagen type III respectively (A,B). VWF:RCo and VWF:CBA activities were absent in the serum-containing medium harvested from the nontransduced CHO-K1 cells. A human standard plasma pool was used as a reference (with VWF:Ag, WVF:RCo, and VWF:CBA all 1 U/mL). Data are mean ± SEM (n = 3). (C) Multimer analysis was conducted on 20 ng Lenti-CMV-huVWF present in the conditioned medium from transduced CHO-K1 cells (CHOK1-VWF) and on 60 ng VWF present in normal human plasma pool. Serum-containing medium (medium) was used as a negative control.

Figure 4.

Fluorescent microscopy analysis of the transduction efficiency of canine VWD BOECs. Canine VWD BOECs were subjected to a centrifugal transduction (900g, 30 minutes, 32°C at an MOI of 20) with Lenti-CMV-GFP (left) or empty vector (right), and 7 days after transduction, the presence of GFP was analyzed by using a fluorescence microscope. Cell nuclei were stained with DAPI.

Figure 5.

Analysis of human BOECs transduced with lentiviral and γ-retroviral vectors. Human BOECs were transduced with Lenti-CMV-GFP or Retro-CMV-GFP at an MOI of 6 and 60 and analyzed by fluorescent microscopy (A) and flow cytometry (B). Percentage of GFP+ cells within the M1 marker interval was indicated. Background fluorescence in negative controls corresponded to less than 0.5% positive cells. (C) A quantitative assessment of vector genome copies was determined by quantitative real-time PCR with primers specific for the WPRE element which is common between the 2 different vector types. Primers specific for the endogenous gene GAPDH were used as control for normalization. Data are mean ± SEM; n = 3.

Figure 6.

Characterization of huVWF expression by canine type 3 VWD BOECs transduced with Lenti-CMV-huVWF. (A) HuVWF levels in conditioned medium from transduced VWD BOECs as determined by ELISA (1 day after transduction) or from negative controls (ie, type 3 VWD BOECs transduced with empty vector or not transduced). (B) HuVWF immunostaining in cytoplasm and Weibel-Palade bodies (magnification) of VWD BOECs transduced with Lenti-CMV-huVWF or empty vector. (C) Multimer analysis was performed on 130 ng VWF present in normal human plasma pool and on expression medium harvested on 1, 14, and 27 days after transduction from canine VWD BOECs transduced with Lenti-CMV-huVWF or empty vector. (D) At several time points after transduction, 24-hour conditioned media from transduced VWD BOECs were tested for VWF:Ag levels by the VWF:Ag ELISA and for functional VWF using the VWF:RCo ELISA and the VWF:CBA assay. No VWF activity was found in conditioned medium from VWD BOECs transduced with empty vector. (E) Capacity of VWF expressed by transduced VWD BOECs to bind FVIII as determined by ELISA. Fresh medium was used as a negative control. Data are mean ± SEM; n = 3.