CAR T Cells Redirected to CD44v6 Control Tumor Growth in Lung and Ovary Adenocarcinoma Bearing Mice

Abstract

The main challenge of adoptive therapy with Chimeric Antigen Receptor modified T cells (CAR T) is the application to the field of solid tumors, where the identification of a proper antigen has emerged as one of the major drawbacks to CAR T cell treatment success. CD44 is a glycoprotein involved in cell-cell and cell-matrix interactions. The isoform containing the variant domain 6 of CD44 gene (CD44v6) has been implicated in tumorigenesis, tumor cell invasion and metastasis and represents an attractive target for CAR T cell therapies. Targeting CD44v6 antigen has been shown to control tumor growth in acute myeloid leukemia and multiple myeloma mouse models. While CAR T approach for the treatment of B cell malignancies has shown great success, response rates among patients with solid cancer are less favorable. The purpose of our study was to test the efficacy of CD44v6.CAR T cells, produced in compliance with Good Manufacturing Practice (GMP), in adenocarcinoma tumor models. We generated a bicistronic retroviral vector containing the CD44v6 CAR and the HSV-TK Mut2 suicide gene to enhance the safety of the proposed CAR T cell therapy. CD44v6 transduced CAR T cells were homogeneously positive for ΔLNGFR selection marker, were enriched in T central memory (T_CM) and T memory stem cells (T_SCM) and displayed a highly activated phenotype. In vitro assays revealed antigen-specific activation and cytotoxicity of human CD44v6.CAR T cells against CD44v6 expressing tumor cell lines. When infused in immunodeficient tumor bearing mice, human CD44v6.CAR T cells were able to reach, infiltrate and proliferate at tumor sites, finally resulting in tumor growth control. Next, we checked if cells produced in compliance with GMP grade standards retained the same antitumor activity of those produced with research grade materials and protocols. Noteworthy, no differences in the potency of the CAR T obtained with the two manufacturing processes were observed. In conclusion, our preclinical results suggest that CD44v6.CAR T based adoptive therapy could be a promising strategy in solid cancer treatment.

Keywords: CAR T; CD44v6; GMP grade; adoptive therapy; solid tumor.

Figure 1

Composition and in vitro activity of CD44v6.CAR T cell product adoptively transferred into mice. T cell memory phenotypic analysis based on CD45RA and CD62L (A, left panel) or CD27 and CD28 (A, right panel) expression. T_N/SCM (CD45RA⁺/CD62L⁺ memory stem T cells), T_CM (CD45RA⁻/CD62L⁺ central memory T cells), T_EM (CD45RA⁻/CD62L⁻ effector memory T cells), T_EM−RA (CD45⁺/CD62L⁻ effector memory RA T cells). Data, collected at the infusion, are the mean ± SE obtained from three independent experiments. Representative flow cytometry plot demonstrating activation markers expression in CD44v6 transduced cells. Data shown are representative of three independent experiments (B). Cytotoxic activity of CAR T cells against IGROV-1 (C) and MR232 (D) cell lines. CD19 and CD44v6.CAR T cells were co-cultured for 48 h with luciferase expressing tumor cells at 1:5 and 1:1 E:T ratio. Percentage of viability was calculated measuring residual luciferase activity of the co-cultured cells respect to the target alone. The graphs show mean ± SE of a representative experiment (n = 3). Cell surface CD44v6 expression of target cells is shown on the left of each panel. Intracellular TNF-α and IFN-γ production (E,F, respectively), and cell surface expression of CD107a (G) were analyzed in CD19.CAR T or CD44v6.CAR T cells incubated with the indicated cell lines MR232 and IGROV-1 (CD44v6⁺ and CD19⁻) or BV173 (CD44v6⁻ and CD19⁺), at E:T ratio 1:1, for 5 h. Percentage of positive cells on gated viable CD3⁺ cells from different donors, is shown. *P < 0.05, **P < 0.01, ***P < 0.001 (Bonferroni's Multiple Comparison Test).

Figure 2

Upon repeated antigen stimulation, CD44v6.CAR T cells proliferate, maintain an indifferentiated phenotype and retain effector cytokine secretion and cytotoxicity in vitro. CD19 and CD44v6.CAR T cells were exposed to CD44v6⁺ irradiated cells (IGROV-1) every 7 days for a total of three stimulations. Graph shows cumulative cell counts, mean ± SE (n = 3 Donors) (A). CAR.CD44v6 expression at cell surface (ΔLNGFR⁺ cells) (B), CD4/CD8 ratio (C), T cell memory phenotypic analysis based on CD45RA and CD62L, T_CM (CD45RA⁻/CD62L⁺ central memory T cells) (D). T_EM (CD45RA⁻/CD62L⁻ effector memory T cells) (E) at day 0 and at day 21 upon weekly stimulation. After the third antigen stimulation CD44v6.CAR T cells functionality was assessed against IGROV-1 cell line in vitro. Intracellular TNF-α, IFN-γ and IL2 production, and cell surface expression of CD107a (F) were analyzed in CD44v6.CAR T cells incubated with IGROV-1 (CD44v6⁺) at E:T ratio 1:1, for 5 h. CD44v6.CAR T cells were co-cultured for 72 h with luciferase expressing tumor cells at 1:1 and 1:5 E:T ratio, before and after repeated stimuli. Percentage of viability was calculated measuring residual luciferase activity of the co-cultured cells respect to the target alone. The graph shows mean ± SE of a representative experiment (G).

Figure 3

CD44v6.CAR T cells, infused in mice bearing human ovarian carcinoma, infiltrate and proliferate at tumor site. NSG mice were subcutaneously injected with CD44v6⁺ IGROV-1 tumor cells. Seven days later, mice were infused via tail vein injection with 7 × 10⁶ CD19 or CD44v6.CAR T cells and re-infused 4 days later with 12 × 10⁶ CD19 or CD44v6.CAR T cells (A). Mice were sacrificed at different time points and tumors were excised and analyzed by cytofluorimetric analysis. A progressive infiltration of CD44v6.CAR T cells in tumors was evident compared to control (B). CD44v6.CAR T cells isolated from tumors were still proliferating 53 days after infusion (n = 3) (C). Quantification of the percentage of CD19 or CD44v6.CAR T and tumor cells after cytofluorimetric analysis (D). Tumor cells isolated by tumors excised from Ctrl group (n = 4) and from CD44v6 group (n = 4) were CD44v6⁺ (E). Infiltrating T cells displayed a polyclonal TCR Vβ repertoire (F). At later time points, TILs lost CD62L and showed higher CD45RA expression compatible with a differentiated phenotype (G).

Figure 4

Systemic treatment of mice bearing human ovarian and lung carcinoma with CD44v6.CAR T cells leads to tumor growth control and enhances survival. NSG mice were subcutaneously injected with 3 × 10⁵ CD44v6⁺ IGROV-1 tumor cells, which stably express Firefly Luciferase. Seven days later, mice were infused via tail vein injection with 4.5 × 10⁶ CD19 or CD44v6.CAR T cells and tumor growth quantified by in vivo imaging of luciferase expressing tumor cells (A) and by measuring tumor size (B). Mice treated with CD44v6.CAR T cells showed an enhancement of survival several weeks after treatment compared to Control mice, n ≥ 17 from four independent experiments (C). NSG mice were subcutaneously injected with 3 × 10⁵ CD44v6⁺ MR232 tumor cells. Three days later mice were infused via tail vein injection with 4.5 × 10⁶ CD19 or CD44v6.CAR T cells and tumor growth quantified by measuring tumor size (D). Mice were sacrificed at day 27, tumors were harvested and digested, cells suspension was analyzed by FACS for CD3 and CD45 (E, left panel), CD45RA and CD62L (E, middle panel), and HLA DR (E, right panel) expression. TILs still had cytotoxic activity on CD44v6⁺ MOLT-4 (F). NSG mice were subcutaneously injected with 3 × 10⁵ CD44v6⁺ MR232 tumor cells. Three days later mice were infused via tail vein injection with 10 × 10⁶ CD19 or CD44v6.CAR T cells and tumor growth quantified by measuring tumor size (G). Mice treated with CD44v6.CAR T cells showed an enhancement of survival several weeks after treatment compared to control mice n ≥ 13 from three independent experiments (H); *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 5

Comparison of phenotype and in vivo functionality of CAR T cells produced according to research or manufacturing processes. T cell memory phenotypic analysis based on CD45RA and CD62L expression. T_N/SCM (CD45RA⁺/CD62L⁺ memory stem T cells); T_CM (CD45RA⁻/CD62L⁺ central memory T cells); T_EM (CD45RA⁻/CD62L⁻ effector memory T cells); T_EM−RA (CD45RA⁺/CD62L⁻effector memory RA T cells) (A). HLA DR and CD25 expression on Research Grade Product (light gray) and GMP grade Product (dark gray) (B). All data are representative of two independent CAR T preparations. CAR T cells produced according to both Research Grade and GMP Grade Processes control tumor growth in mice bearing lung carcinoma (C) and delay mice survival (D). Number of human T cells per microliter in peripheral blood of mice infused with Research Grade and GMP Grade CAR T cells at different time points (E).