A Highly Efficient, Clinically Applicable Transfection Method to Redirect the Specificity of Immune Cells and Enhance Their Anti-Tumor Capacity

Effective immunotherapies rely on triggering specific cytotoxic responses against target cells. Chimeric Antigen Receptors (CAR) composed of antigen recognition domains in the form of scFv portions of monoclonal antibodies and immune cell activation domains provide non-MHC restricted, KIR-independent, immune cellmediated cytotoxicity directed toward target cells expressing specific antigens. It is well demonstrated that CAR transfection into immune cells using viral vectors redirects their specificity and enhances their cytotoxicity. Translation of these findings into clinically feasible treatment strategies is challenged by the labor/costs associated with extended cell processing, and by safety concerns associated with viral vectors. We describe here the use of non-viral methods to deliver a nucleotide sequence encoding CAR into freshly isolated immune cells obtained from peripheral blood leukapheresis collection. Protocols for loading mRNA encoding eGFP (irrelevant control) and a previously described anti-CD19 CAR (Imai et al. Blood 2005) into peripheral blood mononuclear cells (PBMC) and cell subsets (T cells, NK cells, and monocyte-depleted PBMC) were developed and optimized using a clinically validated, automated, closed-system, flow electroporation technology. PBMC were obtained from 5 normal healthy donors and 1 patient with B-cell chronic lymphocytic leukemia (B-CLL). Expression levels of eGFP or anti-CD19-BB-z were analyzed by flow cytometry. Specific cytotoxicity of target cells after 4 hours was measured by flow cytometry using acetoxymethyl-calcein labeled CD19+ target cells which included acute lymphoblastic leukemia (ALL) cell lines, primary B-CLL cells and autologous B cells. Following loading of mRNA encoding eGFP, cell viability was consistently ≥80% for all cell populations. Most cells expressed eGFP: 97±1% (day 1) and 83±2% (day 7). Following loading of mRNA encoding anti-CD19-BB-z, cell viability was 96%±1% for resting T cells (n=2), 93%±10% for PBL (n=2), 94%±8% for PBMC (n=3), 80%±15% for NK cells (n=4). Expression of anti-CD19-BB-z protein receptor was 54%±11% (day 1) in PBMC (n=6). CAR expression decreased over time and was essentially similar to baseline on day 7 post-loading. Cells expressing anti-CD19-BB-z protein (but not those expressing eGFP only), lysed 60%–80% ALL or B-CLL target cells at an effector-to-target ratio (E:T) of 8:1. A shigher E:T ratio of 20:1 was required for 60%–80% lysis of autologous B-cell targets. There was no change in cytolytic activity of CAR mRNA loaded-PBMC during the first 3 days post-loading. Subsequently, cytolytic activity dropped and on day 7 (post-loading) was approximately half of the activity measured on day 3 (post-loading).

In vivo testing was performed in a sub-cutaneous tumor grafting model in Beige SCID mice. The mice (N = 5 per group) were transplanted subcutaneously with CD19+ tumor (HS Sultan) cells (1 × 10⁶) co-mixed with PBMC loaded with anti-CD19-BB-z mRNA (doses = 0.74, 2.22, 6.67 and 20 ×10⁶). Controls were tumor cells only or tumor cells co-mixed with unmodified PBMC. Tumors developed and grew in 3/5 animals in each control group and in 2/5 in lowest anti-CD19-BB-z mRNA-loaded dose. No tumor formation was observed in any animal in each of the 3 higher cell doses. More comprehensive animal studies and investigation of engineering additional constructs for chimeric signal receptor with tumor affinities other than anti-CD19 are in progress. These results indicate that CAR can be effectively transfected into immune cells with the methodology that we developed. Because of its simplicity and effectiveness, this methodology is well suited for clinical application and holds promise for improving the specificity and widening the feasibility of cell therapy.