Incremental Innovation of Ex Vivo Hematopoietic Stem Cell Engineering to Expand Clinical Gene Therapy Applications

Transplantation of genetically engineered, autologous hematopoietic stem and progenitor cells (HSPC) is becoming a promising alternative to allogeneic stem cell transplantation for curing genetic diseases, avoiding the risks of graft versus host disease and prolonged immunosuppression. Most clinical gene therapy protocols are based on CD34+ HSPC engineered during >2 days of ex vivo culture. By xenotransplanting mobilized peripheral blood (mPB) CD34+ HSPC, which were lentivirally (LV) marked with different fluorescent proteins according to CD38/CD90 expression levels allowing quantitative assessment of the contribution of CD38/CD90 subpopulations to hematopoietic reconstitution (n=48 NSG mice, 3 experiments), we identified 2 distinct waves of reconstitution: (1) short term repopulation (up to 2 months) mostly driven by CD34+CD38intCD90+/- cells and (2) long-term repopulation driven by CD34+CD38-CD90+ (70%) and CD34+CD38-CD90- cells (30%). Notably, an intermediate wave extending from 2 to 4 months driven by CD34+CD38low cells was selectively eliminated by prolonged ex vivo culture and could be rescued when culture time was reduced to 1 day. We therefore developed a novel LV transduction protocol able to provide curative levels of gene transfer during a single day of ex vivo culture. Stimulating CD34+ cells or CD34+CD38- cells with Prostaglandin E2 (PGE2) increased gene transfer with VSVg-pseudotyped LVs by 1.5-2 fold acting on early steps of transduction, an effect that was further potentiated by the late-acting compound Cyclosporin A. Using large-scale vector preparations for gene therapy of mucopolysaccharidosis type 1, chronic granulomatous disease or beta-thalassemia, we show by in vitro and xenotransplantation assays that a 1-day PGE2 protocol achieved similar transduction efficiencies into BM or MPB HSPC from healthy donors and patients as our 62h benchmark protocol. PGE2 treatment did not result in toxicity or skewed multi-lineage differentiation. However, shortening ex vivo culture increased engraftment levels in the NSG mouse model. To entirely avoid culturing progenitor cells, we explored the feasibility to limit ex vivo manipulation to HSC-enriched CD34+CD38- cells that may be co-transplanted with unmanipulated CD34+ progenitor cells devoid of long-term engraftment potential. This could further improve hematopoietic reconstitution, increase safety by reducing the LV integration load infused into the patient and downscale ex vivo manipulation making the process more efficient and economically sustainable. To this end, we optimized a sequential bead-based, GMP-compatible selection procedure to separate mPB into a CD34+CD38- stem and CD38+ progenitor cell fraction. We reached high purity (87+/-6.6% CD34+) and recovery of CD34+CD38- cells (37.3+/-8.7%), making their isolation clinically viable. Bead-selected CD34+CD38- cells showed higher engraftment potential than equivalent numbers of FACS-sorted cells. Co-infusion of unmanipulated (culture-sensitive) CD38+ supporter cells with genetically-engineered CD34+CD38- cells into NSG mice resulted in rapid engraftment followed by near-complete replacement of untransduced short-term repopulating progenitors by gene-marked HSPC deriving from CD34+CD38- cells after the 3rd month post-transplant. Finally, we explored ex vivo expansion of mPB CD34+CD38- cells with arylhydrocarbon receptor antagonists and/or pyrimido-indole-derivatives. These cells expanded 3-10 fold in a 7-14 d time-window, far less than seen for total CD34+ cells, thereby facilitating culture handling and reducing cost. Unlike CD34+ cells, expanded mPB CD34+CD38- cells largely maintained their SCID-repopulating potential providing proof-of-concept for the expansion of gene-modified HSC. This clinically applicable platform will improve the efficacy, safety and sustainability of ex vivo gene addition and open up new opportunities in the field of gene editing.