Chimeric antigen receptor T cell-based targeting of CD317 as a novel immunotherapeutic strategy against glioblastoma

Abstract

Background: Chimeric antigen receptor (CAR) T cell therapy has proven to be successful against hematological malignancies. However, exploiting CAR T cells to treat solid tumors is more challenging for various reasons including the lack of suitable target antigens. Here, we identify the transmembrane protein CD317 as a novel target antigen for CAR T cell therapy against glioblastoma, one of the most aggressive solid tumors.

Methods: CD317-targeting CAR T cells were generated by lentivirally transducing human T cells from healthy donors. The anti-glioma activity of CD317-CAR T cells toward various glioma cells was assessed in vitro in cell lysis assays. Subsequently, we determined the efficacy of CD317-CAR T cells to control tumor growth in vivo in clinically relevant mouse glioma models.

Results: We generated CD317-specific CAR T cells and demonstrate strong anti-tumor activity against several glioma cell lines as well as primary patient-derived cells with varying CD317 expression levels in vitro. A CRISPR/Cas9-mediated knockout of CD317 protected glioma cells from CAR T cell lysis, demonstrating the target specificity of the approach. Silencing of CD317 expression in T cells by RNA interference reduced fratricide of engineered T cells and further improved their effector function. Using orthotopic glioma mouse models, we demonstrate the antigen-specific anti-tumor activity of CD317-CAR T cells, which resulted in prolonged survival and cure of a fraction of CAR T cell-treated animals.

Conclusions: These data reveal a promising role of CD317-CAR T cell therapy against glioblastoma, which warrants further evaluation to translate this immunotherapeutic strategy into clinical neuro-oncology.

Keywords: BST2; CAR T cell; adoptive immune cell transfer; glioma; immunotherapy.

Figure 1.

CD317 is expressed in glioma cells in vitro and in vivo. (A) Determination of CD317 expression on the surface of long-term glioma cells and GIC lines by flow cytometry. HEK293T cells were used as a negative control. SFI values are indicated on the right. (B) Analysis of CD317 expression ex vivo in LN-229 xenografts using immunohistochemistry (upper panel, scale bars, 100 μm) or fluorescence confocal microscopy (lower panel, scale bars, 10 μm). (C) Representative CD317 immunohistochemical stainings of glioblastoma tissue specimens and healthy brain sections are shown. Size bars correspond to 20 µm or 200 µm (left). CD317 expression was quantified in 56 glioblastoma samples and 13 normal brain tissue sections by H score (right). *** indicates significance P < .001.

Figure 2.

CD317-specific CAR T cells lyse glioma cells in an effector:target-dependent manner. (A) Scheme of the CD317-specific CAR construct composed of the elongation factor-1 promoter (EF1), a GM-CSF signal peptide, the CD317 targeting domain (CD317-scFv), a CD8 alpha hinge and transmembrane domain (CD8α TM), a 4-1BB cytoplasmic domain, a human CD3 zeta cytoplasmic domain (CD3ζ), a self-cleaving connecting peptide (T2A) and a marker suicide gene (RQR8). (B) Human T cells from healthy donors were transduced with the CD317-CAR construct or left untreated (non-transduced T cells). The expression of RQR8 was determined by flow cytometry (left panel). The presence of the CD317-specific CAR was determined by binding to a CD317-Fc fusion protein and detected by an anti-IgG antibody (right panel). Numbers indicate the percentage of RQR8-positive or CD317-CAR-positive T cells. (C) The human glioma cell lines A172, LN-18, LN-229, T98G, or the human glioma-initiating cell lines GS-2, GS-5, or T-325 were used as target cells in 17 h cytolysis assays. Human T cells were transduced with the CD317-CAR construct or not and used as effector cells at various effector:target (E:T) ratios. (D, and E) The human glioma cell line LN-229 was co-cultured with CD317-CAR or control T cells at an E:T ratio of 1:1. After 17 h of co-culture, IFN-γ levels were determined in CD4⁺ and CD8⁺ T cells by flow cytometry or in the co-culture supernatant by ELISA. Numbers indicate the percentage of IFN-γ -positive T cells (D) or the IFN-γ concentration in the supernatant (E). Data are presented as mean ± SD (*P < .05, **P < .01, and ***P < .001).

Figure 3.

CD317-CAR T cells are specific for their target antigen. (A) LN-18, LN-229, or ZH-161 glioma cells were exposed to 50–100 IU/ml IFN-β for 48 h or left untreated. The expression of CD317 on the cell surface was determined by flow cytometry. (B) Glioma cells were exposed to IFN-β or not and used as target cells in 17 h cytolysis assays. CD317-CAR or non-transduced T cells were used as effector cells at various E:T ratios. (C and D) CD317 was overexpressed in LN-229 cells (OE) and the expression was determined by flow cytometry (C). LN-229 CD317 OE cells were used as target cells and co-cultured with CD317-CAR or non-transduced T cells at various E:T ratios for 17 h (D). (E and F) CD317 was knocked-out using CRISPR/Cas9 gene editing in LN-229 cells. CD317 expression was determined by flow cytometry (E). CD317 knockout (KO) or control LN-229 cells were used as target cells in 17 h cytolysis assays at varying E:T ratios with CD317-CAR or non-transduced T cells as effectors (F). Data are presented as mean ± SD (*P < .05, **P < .01, and ***P < .001).

Figure 4.

CD317 gene silencing in T cells reduces fratricide and increases viability, expansion, and cytotoxic function of CD317-CAR T cells. (A) The percentage of viable T cells transduced with the different constructs was determined by flow cytometry (left panel) and the total numbers of T cells were identified manually (right panel) at the indicated days following transduction. (B) Non-transduced T cells, CD317-CAR, or CD317-shRNA-CAR T cells were co-cultured or not with LN-229 glioma cells at an E:T ratio of 1:1. After 17 h of co-culture, IFN-γ levels were determined in CD4⁺ (left panel) and CD8⁺ (right panel) T cells by flow cytometry. Numbers indicate the percentage of IFN-γ -positive T cells. (C) CAR T cells or non-transduced T cells were co-cultured or not with LN-229 cells at a E:T ratio of 1:1 and the expression of the exhaustion markers PD-1, TIM-3, and LAG-3 was determined by flow cytometry after 17 h of co-culture. Numbers indicate the percentage of positive T cells. (D) Re-challenge assay of CD317-CAR and CD317-shRNA-CAR T cells. Two thousand T cells were co-cultured with 2000 LN-229 cells and repetitively challenged with 3000 tumor cells every 48 h (arrows at day (D) 2, 4, 6, and 8). Numbers of remaining viable tumor cells were determined at indicated time points by flow cytometry (day 1, 3, 5, 7, and 9). Data are presented as mean ± SD (*P < .05, **P < .01, ***P < .001).

Figure 5.

Treatment with CD317-CAR T cells prolongs the survival and cures a fraction of orthotopic glioma-bearing mice. (A) LN-229 tumor-bearing mice were treated with a single intratumoral injection of 10⁶ CD317-shRNA-CAR T cells, 10⁶ non-transduced T cells, or PBS control at day 7 after tumor implantation. (B) LN-229 wild-type (WT) or CD317 KO cells were implanted into nude mice. Seven days later, mice were treated intratumorally with 10⁶ CD317-shRNA-CAR T cells, non-transduced T cells, or PBS control. (C) ZH-161 WT or CD317 OE tumor-bearing mice were treated intratumorally with 10⁶ CD317-shRNA-CAR T cells or 10⁶ non-transduced T cells 5 days after tumor implantation. Survival data are presented as Kaplan–Meier plots and significance was determined with the log-rank test (*P < .05, **P < 0.01, and ***P < 0.001).