Culture Of Mobilized Rhesus Macaque CD34+ Cells In Hypoxic Conditions Does Not Improve Lentiviral Transduction Efficiency In Long-Term Repopulating Hematopoietic Stem Cells

Low oxygen levels have been proposed to play a fundamental role in the maintenance of normal hematopoietic stem cells (HSC) function. We investigated whether culture of rhesus macaque mobilized CD34+ cells under low oxygen conditions (1-5% O₂) could improve lentiviral transduction efficiency in HSCs compared to culture under atmospheric O₂ conditions (21%). Mobilized CD34+ cells from 4 animals were prestimulated for 12 hours in the presence of cytokines and subsequently divided in two equal fractions for transduction with SIN-lentiviral vectors (MOI=100) expressing GFP or TdTomato fluorescent proteins for 48 hours. In 3 animals, cells were transduced with GFP vectors in hypoxia and TdTomato vectors in normoxia. These conditions were reversed in one animal. A portion of the transduced cells was used for phenotypic, cell cycle, clonogenic and apoptosis assays. The remaining cells from both fractions were combined and injected intravenously in lethally irradiated rhesus macaques and in vivo gene marking was measured in peripheral blood granulocytes for up to 21 months after transplantation. The numbers of total live cells and CD34+ cells after 2.5 days in culture were not significantly different compared to baseline for both hypoxic and normoxic conditions (p=0.85 and p=0.84, respectively). Similarly, numbers of CD34+CD38- cells were stable compared to baseline in hypoxia and normoxia (p=0.94). After 2.5 days of culture, the total number of colony-forming cells increased 1.4-fold under both hypoxic and normoxic conditions (p=0.69) compared to freshly isolated CD34+ cells. At baseline, the majority (50.6 + 4.4%) of the CD34+ cells were in the G₀ phase of the cell cycle. After 2.5 days in culture under hypoxic or normoxic conditions, the percentages of cells in G₀ were 19.9 + 8.8% and 18.6 + 5.9%, respectively (p=0.82). The differences in percentages of cells in the G₁ and S/G₂/M phases of the cell cycle were also not statistically significant (p=0.175 and p=0.732, respectively). The pO₂ had no impact on cell death (18.2 + 5.3% in hypoxia and 16.8 + 4.8% in normoxia, p=0.69) or on the percentages of apoptotic cells (11.1 + 6.7% in hypoxia and 8.4 + 2.3% in normoxia, p=0.47). When considering the bulk of CD34+ cells after transduction with lentiviral vectors, the percentages of GFP-marked cells were consistently higher (range 1.8 to 2.2-fold, mean 2.1-fold) compared to TdTomato-marked cells independently of the transduction conditions used, consistent with a 2.1-fold intrinsic superiority of GFP-based lentiviral vectors compared to TdTomato vectors at equivalent MOI. Taking into account the inherent vector differences, transduction efficiencies were similar in normoxia (70.7 + 11.7%) and hypoxia (68.0 + 16.3%) (p=0.80). When adjusted for the intrinsic superiority of GFP-based vectors, the contribution to long-term in vivo gene marking measured by flow cytometry and quantitative PCR was comparable between cells transduced in 21% pO₂ (9.0 + 4.5%) and 5% pO₂ conditions (8.4 + 4.9%) (p=0.87). Consistent with these data, GFP and TdTomato-labeled hematopoietic cells were equally visualized, using a confocal/2-photon hybrid microscopy approach, in BM biopsy specimens collected at various time points after transplantation. Given recent evidence suggesting that lower oxygen concentrations (<1.5%) may be required to stabilize HIF-1a, a key sensor of hypoxic conditions, transduction was performed under extreme hypoxic conditions (1% O₂) in one animal. While the overall viability of the bulk CD34+ cells after 2.5 days of transduction under extreme hypoxia was comparable to cells cultured in normoxia, their contribution to long-term in vivo marking was negligible (0.3%) compared to cells transduced under normoxic conditions (2.4%), suggesting toxicity of very low oxygen levels on HSCs. Overall, when current methodologies used for the genetic manipulation of HSCs for gene therapy applications were performed under hypoxic conditions, susceptibility to lentiviral transduction of CD34+ cells was not ameliorated compared to cells cultured in 21% O₂ conditions in the rhesus macaque transplantation model. The short culture times (2-3 days) used in recent lentivirus-based gene therapy clinical trials and replicated in this study may not be sufficient to impact the phenotype of long-term repopulating HSCs and do not warrant incorporation of hypoxia in current gene therapy protocols.