Lentiviral vector gene therapy for hematopoietic disorders caused by single gene mutation or deficiency has demonstrated success in clinical trials. However, efficient gene delivery is still challenging, requiring high multiplicity of infection (MOI) to achieve average vector copy numbers of at least 1 per cell. Conditions utilizing high MOI result in improved transduction rates and adequate copy number but these benefits can coincide with potential insertional mutagenesis. Attempting to achieve optimal gene transfer at reduced MOI, we directly compared self-inactivating (SIN) lentiviral vector particles encoding eGFP under control of a murine stem cell virus (MSCV) promoter/enhancer pseudotyped with envelope glycoproteins for BaEV or vesicular stomatitis virus (VSV). BaEV proteins were either full-length (BaWT) or lacked the terminal R-peptide (BaR-less), a version previously reported to improve infectious titer (TU, transducing units per mL) using transient production procedures. Using a standardized four-plasmid transient production protocol, cells transfected with R-less expression plasmids produced syncitia; an issue that resolved using BaWT. Titers were significantly higher for virus packaged with BaWT compared to R-less (BaR-less: 3.6 × 105 ± 6.6 × 104 N=12; BaWT 2.4 × 106 ± 4.5 × 105; n=10; p<0.001) determined by flow cytometry analysis of GFP expression in transduced HEK-293T cells. However, these values were 10-to 20-fold less than those observed for VSV-G (3.5 × 107 ± 4.5 × 106). Human CD34+ hematopoietic stem/progenitor cells isolated from bone marrow (BM) or cytokine-mobilized peripheral blood (mPB) of healthy donors were prestimulated and transduced overnight on retronectin-coated plates with BaWT particles at an MOI of 0.5, 1, 2, or 4. The following day, cells were either plated in methylcellulose to assess colony-formation (CFU), maintained in liquid culture (Bulk), or transplanted into NOD-scid IL2Rgnull (NSG) mice pre-conditioned with busulfan (n=10 mice/1×106 cells each). Gene transfer efficiency was gauged by expression of GFP following flow cytometry analysis of cells maintained in culture or visual inspection of colonies growing in methylcellulose using an inverted fluorescence microscope after 6 or 12 days, respectively. Transduction efficiency increased with viral MOI reaching peak levels using an MOI of 2 (Bulk, 93% GFP+; CFU, 98% GFP+). Six NSG mice survived to 14 weeks post-transplant and demonstrated engraftment of CD45+ cells ranging from 19 to 54% (39 ± 13%; mean ± SD) and GFP marking ranging from 32 to 59% (41 ± 10%) determined by flow cytometry. Finally, we compared GFP-encoding particles pseudotyped with VSV-G or BaWT normalizing for levels of p24 quantified by ELISA (BaWT, 1155.0 pg/ml, 3.3 × 106 TU; VSV-G, 1164.9 pg/ml, 5.1 × 107 TU) to account for differences in calculated titer. Prestimulated mPB CD34+ cells were transduced overnight and placed into methylcellulose or transplanted into 15 NSG mice each. Methylcellulose colonies were analyzed for GFP after 12 days of growth (BaWT, 95.6%; VSV-G, 76.6%). Transplanted mice will be analyzed at 16-18 weeks post-transplant to measure engraftment and GFP expression. We believe the stability of producer cells expressing BaWT, the high efficiency of transduction at low MOI, and the ability to transduce adequate numbers of cells with unconcentrated virus supernatant warrant continued optimization of transient production methods and development of a stable producer cell line utilizing BaWT envelope as an alternative to VSV-g.
Read full abstract