Specific Correction of the Intron-22 Inverted Factor VIII Gene in Autologous Blood Outgrowth Endothelial Cells from Patients with Severe Hemophilia A

A heterogeneous collection of >1500 distinct causative factor (F) VIII gene (F8) mutations have been identified thus far in unrelated severe Hemophilia A (HA) patients, who have less than 1% of normal FVIII activity in their plasmas and experience recurrent bleeding that often results in crippling arthropathies and can be life threatening. Curative gene therapies are being pursued intensely as the current standard of care, which involves 2-3 infusions/week of therapeutic FVIII (tFVIII) proteins throughout a patient's life, is extremely expensive and very demanding. Moreover, ~25% of patients with severe HA (PSHA) develop anti-tFVIII antibodies that neutralize the efficacy of their tFVIII proteins. Despite remarkable progress in clinical trials of various adeno-associated viruses (AAVs) as vectors for in vivo delivery of therapeutic F8 genes, it is not clear how widespread viral-mediated gene replacement therapy (GRT) will become due to current limitations that include the: (1) presence of existing immunity to the AAV capsid protein (CP) in ~30-70% of PSHA for whom GRT is contraindicated; (2) immunity to AAV-CP induced in all PSHA during the initial GRT that precludes subsequent dosing; (3) use of heterologous promoters which drive F8 expression in non-physiologic cells that may increase the encoded tFVIII protein's immunogenicity; and 4) episomal location of AAV-genome replication, which, together with "(2)" and "(3)", precludes GRT in children. These important unmet needs require new gene-based therapeutic strategies for HA. Our goal was to develop a virus-free, ex vivo personalized gene repair therapy that minimally manipulates the mutant F8 in autologous patient-derived blood outgrowth endothelial cells (BOECs)-the physiologically relevant cell type for FVIII production in vivo-followed by their expansion and reinfusion into the same individual patient. We chose to focus initially on the intron (I) 22 inversion (I22I) mutation initially as it is causative in >40% of all PSHA. CRISPR/Cas9 guide RNAs were designed to target the 3' end of F8 exon (E) 22. For initial experiments in K562 cells, a donor plasmid containing a restriction enzyme site was nucleofected with the CRISPR system encoding plasmid, which triggered successful homology-directed repair (HDR) at the target site (efficiency of 18.2%, n=2). Subsequently, a cDNA-based therapeutic donor plasmid was constructed containing all F8 coding sequences in E23-E26 followed by a bGH polyA signal. After appropriate informed consent was obtained, BOECs were cultured from the blood of 3 severe pediatric HA patients with the I22I (ages 6 years, 7 years, and 13 years). After nucleofecting the BOECs with the F8-specific CRISPR and HDR constructs, site-specific knock-in of the cDNA at the 3' end of E22 was confirmed via PCR (n=3). Direct Sanger sequencing of the resultant amplicons from the repaired I22-inverted F8 locus in treated BOECs from one representative patient confirmed complete and seamless knock-in of the therapeutic cDNA at the endogenous site of the mutant F8. Current efforts are to isolate clonal populations of the repaired BOECs and characterize their ability to secrete active FVIII in vitro. Similar experiments are underway using canine BOECs in the canine model of HA. The use of clonal populations of gene corrected autologous BOECs as the infused therapy allows whole genome sequencing analysis to be performed to confirm that no off-target cutting or integration occurred in the therapeutic cell preparation prior to infusion into the recipient patient, further strengthening the safety profile of this proposed autologous cell therapy. Overall, these current results lay promising proof-of-concept data for a potential new curative therapeutic alternative approach for HA which should overcome drawbacks of the current generations of AAV-based treatments.