Engineering safe anti-CD19-CD28ζ CAR T cells with CD8a hinge domain in serum-free media for adoptive immunotherapy

Abstract

Background: Despite the curative potential, high cost of manufacturing and the toxicities limits the wider access of Chimeric Antigen Receptor (CAR) T cell therapy in global medicine. CARs are modular synthetic antigen receptors integrating the single-chain variable fragment (scFv) of an immunoglobulin molecule to the TCR signaling. CARs allow HLA independent, T cell mediated destruction of tumor cells independent of tumor associated-HLA downregulation and survive within the patient as 'living drug.' Here we report a safer approach for engineering alpha beta T cells with anti- CD19-CD28ζ CAR using self-inactivating (SIN) lentiviral vectors for adoptive immunotherapy.

Method: αβ T cells from the peripheral blood (PB) were lentivirally transduced with CAR construct containing hinge domain from CD8α, transmembrane and co-stimulatory domain from CD28 along with signaling domain from CD3ζ and driven by human UBC promoter. The cells were pre-stimulated through CD3/CD28 beads before lentiviral transduction. Transduction efficiency, fold expansion and phenotype were monitored for the CAR T cells expanded for 10-12 days. The antigen-specific tumor-killing capacity of CD19 CAR T cells was assessed against a standard CD19 expressing NALM6 cell lines with a flow cytometry-based assay optimized in the lab.

Results and conclusion: We have generated high titer lentiviral vectors of CAR with a titer of 9.85 ± 2.2×10⁷ TU/ml (mean ± SEM; n=9) generating a transduction efficiency of 27.57 ± 2.4%. (n=7) at an MOI of 10 in total T cells. The product got higher CD8+ to CD4+ CAR T cell ratio with preponderance of an effector memory phenotype on day 07 and day 12. The CAR-T cells expanded (148.4 ± 29 fold; n=7) in serum free media with very high viability (87.8 ± 2.2%; n=7) on day 12. The antitumor functions of CD19 CAR T cells as gauged against percentage lysis of NALM6 cells at a 1:1 ratio is 27.68 ± 6.87% drawing up to the release criteria. CAR T cells produced IFNγ (11.23 ± 1.5%; n=6) and degranulation marker CD107α (34.82 ± 2.08%; n=5) in an antigen-specific manner. Furthermore, the sequences of WPRE, GFP, and P2A were removed from the CAR construct to enhance safety. These CAR T cells expanded up to 21.7 ± 5.53 fold with 82.7±5.43% viability (n=4).

Conclusion: We have generated, validated, and characterized a reproducible indigenous workflow for generating anti-CD19 CAR T cells in vitro. This approach can be used for targeting cancer and autoimmune diseases in which CD19+ B lineage cells cause host damage.

Keywords: B lineage malignancies; CD19 CAR T cells; NALM-6; SIN vector; serum free media.

Figure 1

Expansion dynamics of CAR-T cells. (A) Graph representing the number of untransduced T cells and CAR transduced T cells at the indicated days after initiating the T cell expansion culture. (B) Fold expansion of the untransduced and CAR-transduced T cells on day 12. (C) Viability of the untransduced and CAR-transduced cells in the expansion culture. (D–F) T cells were transduced with CAR LV vectors, and GFP expression was monitored on day 7 and day 12 as a surrogate of CAR expression in CD3+ (D), CD4+ (E), and CD8+ (F) compartments. The data represented as mean ± SEM from seven individual donors (n=7). The statistical significance is estimated by the Student’s T-test. ns, not significant.

Figure 2

Antigen-specific stimulation of CAR-T cells. (A) Experimental design of antigen-specific expansion, Nalm-6 cells were co-cultured with CAR-T cells and CAR without extracellular domain (scFv) at 1:1 ratio and expanded for 7 days. (B) Expression of CAR was measured before and after the 7 days of co-culture by flowcytometry. (C, D) The CAR expression (C) and cell number (D) differences were measured before and after stimulation. (E) The fold change of antigen-specific proliferation for CAR-T cells and CAR without extracellular domain. The data represented as mean ± SEM from three individual donors (n=3). The statistical significance is estimated by using Student’s T-test. * p<0.05, **p<0.005, ns, not significant.

Figure 3

Cytotoxicity of CAR T cells. Nalm-6 cells co-cultured with UT and CAR-T cells at a 1:1 ratio. (A) Cells were gated from CD3+ and GFP+ to exclude the CAR-T cells, and from the negative SSC and 7-AAD to measure the lysis of Nalm-6 cells. (B, C) Cells were gated from 7-AAD negative and CAR-T cell populations, which were excluded by CD3 and GFP staining. The negative population is the survival of target cells. (D, E) Lysis (%CD3neg7AAD+) (D) and survival percentage (E) of target cells (n=7). The data represented as mean ± SEM from seven individual donors (n=7). The statistical significance is estimated by using Student’s T-test. **p<0.005.

Figure 4

Antigen-specific IFNγ secretion and CD107 degranulation of CD19 CAR-T cells. (A) Schematic representation of antigen-specific IFNγ secretion and CD107 degranulation of CD19 CAR-T cells and mock CAR-T cells (without extracellular domain). Effector cells were co-cultured with an equal number of target cells for 5 hours. (B) FACS plots representing the percentage of intracellular IFN-γ, and the degranulation was assessed by measuring the surface expression of CD107a. (C, D) Bar graph representing the Interferon-gamma (C) percentage and CD107a (D). The data are represented as mean ± SEM from six individual donors (n=6) for Interferon-gamma and five individual donors (n=5) for CD107a. The statistical significance is estimated by using the Student’s T-test. ***p<0.001, ***p<0.0001.

Figure 5

Construction of CD19 CAR construct (without WPRE/GFP/P2A). (A) Experimental plan for generating and validating the cGMP grade CAR construct. (B) A map of the modified cGMP grade CAR constructs. (C, D) The surface expression of CAR was measured by CD19-PE staining across the six different donors (n=6). Data is represented as mean ± SEM.