Systematic Comparison of the EF-1 Alpha Short (EFS) and Viral Promoters for Gene Modification of Human Primary Cells for Clinical Applications

Optimization of transgene expression is paramount for successful gene modification of primary cells for clinical applications, and careful selection of the viral vector construct is a critical part of this process. Viral promoters based on the U3 region of the Moloney murine leukemia virus (such as MNDU3 and MSCV) are currently the most commonly used for gene transfer in human primary cells. These viral promoter-containing vectors, however, can activate nearby genes, potentially causing toxicity and/or neoplastic transformation. EF1alpha (or its short, intron-less form, EFS) is a promoter that has been recently used in many clinical trials. It is a cellular-derived enhancer/promoter with decreased cross-activation of nearby promoters, therefore hypothetically decreasing the risk of genotoxicity. We have produced vector constructs carrying the internal enhancer/promoters MNDU3, MSCV, or EFS driving clinically relevant transgenes for modification of primary human T lymphocytes and hematopoietic stem cells.

Lentiviral vectors containing either the MNDU3 or EFS promoters driving the EGFP reporter gene were used to transduce Jurkat cells and primary human T cells. In Jurkat cells, MNDU3-driven vectors provided 2-3 times higher vector copy integrations with a corresponding higher percentage of EGFP expression, across a wide range of multiplicity of infection (MOI). In primary T cells, however, there was no significant increase in vector copy numbers per cell, but a significant increase in transduction efficiency and geometric mean fluorescence intensity of EGFP expression in cells transduced with MNDU3-driven vectors at all MOI studied, even when corrected for vector copy number.

Lentiviral vectors containing either a MNDU3 or EFS promoter driving a first-generation anti-CD19 chimeric antigen receptor (CAR) were used to transduce primary human T cells. We found that integrated vector copy numbers per cell were 0.8 with MNDU3 and 0.5 with EFS, and resultant transgene expression in the transduced populations was 45% with MNDU3 and 22% with EFS. Primary human T cells were also transduced with a lentivirus carrying MSCV or EFS driving a codon-optimized MART-1-specific T cell receptor (TCR) and then analyzed by tetramer staining. MSCV promoter-driven vectors resulted in 33.76%, 33.1%, and 29% higher transgene expression at 5 ng, 10 ng, and 25 ng p24 equivalents compared with T cells transduced with vectors driven by the EFS promoter using the same amount of p24. After correction for integrated vector copy numbers, T cells had more than 2-fold increase in transgene expression when using the MSCV promoter.

CD34+ hematopoietic stem cells isolated from human cord blood were transduced using the same high-titer MSCV- or EFS-driven MART-1-specific TCR expression vectors; MSCV-driven lentiviral vectors provided an average vector copy number of 0.5 copies per cell compared to 0.7 copies per cell with the similar EFS-containing vectors. These gene-modified cells were then injected into NOD-scid-IL2rγ^null mice, with peripheral blood analyzed by flow cytometry after 8 weeks. HuCD45⁺/huCD3⁺/huCD4⁺ and huCD45⁺/huCD3⁺/huCD8⁺cells had mean transgene expression of 18% and 16% in the MSCV group, compared to 0% and 0% in the EFS group.

Together, these results demonstrate more efficient transgene expression is conveyed by the virally-derived MSCV and MNDU3 promoters versus the cellular EFS promoter in gene-modified primary human hematopoietic cells. Higher transgene expression relative to integrated vector copies is consistent with higher promoter function, and transgene expression may be significantly decreased when using the EFS promoter in lentiviral vectors for clinical applications. Further studies are needed to carefully evaluate genotoxic effects of the MNDU3 and MSCV promoters in comparison to the EFS promoter for safe and efficient clinical translation.