Abstract
Platelets may comprise an ideal vehicle for delivering FVIII in hemophilia A (HemA) as FVIII stored in platelet α-granules is protected from neutralization by inhibitory antibodies and, during bleeding, activated platelets locally excrete their contents to promote clot formation. In order to avoid specific challenges posed by ex vivo gene delivery including, in particular, the requirement to pre-condition the subject, we evaluated intraosseous (IO) infusion of self-inactivating lentiviral vectors (LV) for in situ gene transfer into bone marrow cells. In previous studies, we confirmed that hematopoietic stem cells (HSCs) can be efficiently transduced to express GFP after IO administration of LV driven by a MND promoter (M-GFP-LV).
In the current study, we aimed at limiting transgene expression to the megakaryocyte lineages using IO delivery of 20 µL LV containing either GFP (G-GFP-LV) or a B-domain variant human FVIII (G-F8-LV) gene under the control of megakaryocyte strictly specific promoter glycoprotein 1bα (Gp1bα).
In M-GFP-LV treated control mice, GFP was detected in 6.4% of HSCs, 3.4% of B220+, and 9.0% of CD11c+ bone marrow cells on day 29. In contrast, in G-GFP-LV (6.0E+08 TU/mL) treated mice, GFP was undetectable in bone marrow HSCs, B220+, CD11c+ or CD11b+ cells. GFP expression level in platelets of G-GFP-LV treated mice was ten folds of that in M-GFP-LV treated mice (0.1% vs 0.01%). It indicated that in platelets, the activity of Gp1bα was stronger than that of MND. More importantly, GFP expression levels were stable over 100 days, suggesting that platelets containing the transgene products did not elicit transgene-specific immune responses. Next, we treated HemA mice with G-F8-LV (6.0 E+07 TU/mL). There was no detectable hFVIII expression in bone marrow HSCs on day 8 or in blood cells (CD3ε+, B220+, CD11c+ or CD11b+) on day 35. However, up to 3% platelets express hFVIII on day 91. These results suggested that HSCs in HemA mice were successfully transduced by G-F8-LV after IO infusion, and in the long term, FVIII was synthesized in megakaryocytes and stored in platelet α-granules. In treated mice, the average percentage of platelets expressing hFVIII was stable at 1-2% from day 27 to day 160. The average FVIII antigen level in platelets on day 112 was 1 mU per 1 × 108 platelets, which was comparable with platelet FVIII in transgenic and ex vivo gene therapy treated mice. We also evaluated LV-treated HemA animals for phenotypic correction of bleeding diathesis by tail clip assay. The blood loss was 41% (n=7), 48% (n=5) and 33% (n=5) compared with control HemA (normalized to 100%), mock treated HemA (∼100%), and wild-type (2.5%) mice on days 35, 118 and 160, respectively. Additionally, there was neither detectable FVIII activity nor anti-FVIII antibodies in blood on day 160, which indicated that there was insignificant leaky expression of FVIII in other cells. Finally, we also infused G-F8-LV into HemA inhibitor mice. Inhibitors were induced by repeated injection of 3U recombinant hFVIII. The average antibody level was 80 Bethesda Unit before IO infusion of the vectors. In G-F8-LV treated mice, the average hFVIII antigen level on day 27 was 0.74 mU per 1 x 108 platelets (n=5). Bleeding assay was performed on day 160. The blood loss of treated mice was significantly reduced compared with untreated HemA mice, indicating that IO infusion of G-F8-LV can overcome anti-FVIII antibodies and correct hemophilia phenotype.
We have successfully transduced HSCs in situ by a single infusion of LVs into bone marrow to correct hemophilia A. Gp1bα promoter in lentiviral vectors can specifically direct the transgene expression in mouse platelets. Following IO infusion of G-F8-LV, FVIII stored in platelets can persistently and partially correct the HemA phenotype for at least five months (experimental duration) in mice with and without pre-existing inhibitors. Overall, direct transduction of bone marrow cells targeting platelet-specific FVIII expression may provide an effective therapy to treat severe hemophilia A patients with high-titer inhibitors.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.