Engineering Naturally Occurring CD7 Negative Cells for the Immunotherapy of CD7 Positive Leukemia

Background:

CD7 has emerged as a promising target for the adoptive immunotherapy with T-cells expressing chimeric antigen receptors (CAR T-cells) of CD7⁺ T-cell acute lymphoblastic leukemia (T-ALL) and acute myeloid leukemia (AML). However, expressing CD7 CARs in T-cells results in fratricide due to high expression of CD7 in most T-cells. While investigators have developed strategies to overcome this limitation by additional genetic modifications of CD7 CAR T-cells, the goal of this project was to explore the feasibility of selecting and genetically modifying naturally occurring CD7 negative (CD7^-) T cells for the adoptive immunotherapy of CD7⁺ leukemia.

Methods:

CD7^- T-cells were isolated from PBMCs using a 2-step magnetic bead depletion/selection procedure (CD7 depletion followed by selection of CD3⁺ T cells from the CD7^- fraction). Non-selected T-cells (bulk T-cells), CD7⁺ and CD7^- T cells were activated and transduced with a retroviral vector encoding a second-generation CD7 CAR with a CD28 costimulatory endodomain, and expanded with IL7 and IL15. The effector function of CD7^- T-cells expressing CD7 CARs (CD7 CAR^CD7- T cells) was assessed in vitro as well as in xenograft models.

Results:

To assess the feasibility of our approach, we first determined the frequency of CD7^- T-cells in PBMCs. On average, 4.7 % of T cells were CD7^- (range: 2% - 12.3%; N=22), and we successfully selected these cells from bulk PBMCs with a combined CD7 depletion/CD3 selection procedure. We genetically modified CD7^-, CD7⁺ and bulk T cellsto express CD7 CARs (CD7 CAR^CD7-, CD7 CAR^CD7+, CD7 CAR^Bulk). Transduction efficiencies ranged from 31% to 75% (± 5%) for each T-cell population. Post transduction, CD7 CAR^CD7- T-cells did not undergo fratricide and had similar expansion kinetics (N=6, p=ns) in comparison to non-transduced (NT) T-cell cultures (NT CD7^-, NT CD7⁺, NT bulk). In contrast, CD7 CAR^CD7+or CD7 CAR^Bulk T-cells underwent fratricide and did not expand (N=6, p<0.0001). CD7^- T-cells (NT and CD7 CAR^CD7-) had a predominantly CD4⁺ effector memory phenotype at day 7 and 14 of culture. To assess the effector function of CD7 CAR^CD7- T-cells, we co-cultured them with CD7⁺ T-ALL cell lines (CCRF, MOLT3). CD7 CAR^CD7- T-cells recognized CD7⁺ targets in contrast to CD7^- targets (BV173, Daudi) as evidenced by significant (N=6, p<0.0001) IFN-γ and IL-2 production. Control CAR T-cells (CD19 CAR^CD7-) did not recognize CD7⁺ target cells, confirming specificity. CD7 CAR^CD7- T-cells also had potent cytolytic activity against CD7⁺ targets in cytotoxicity assays. To assess in vivo the anti-tumor activity of CD7 CAR^CD7- T-cells, we used a NSG mouse xenograft model with CCRF cells, genetically modified to express firefly luciferase (CCRF.ffluc) to allow for serial bioluminescence imaging. A single infusion of CD7 CAR^CD7- T-cells had potent anti-leukemia activity as judged by serial imaging resulting in a significant survival (p<0.003) advantage in comparison to control mice.

Conclusion:

We have successfully generated CD7 CAR^CD7- T-cells from peripheral blood CD7^- T-cells. CD7 CAR^CD7- T-cells had a predominantly CD4⁺ effector memory phenotype, and potent anti-leukemia activity in vitro and in vivo. Thus, naturally occurring CD7^- T cells may present a promising T-cell source for the cellular immunotherapy of CD7⁺ leukemia.