4-1BB Signaling Boosts the Anti-Tumor Activity of CD28-Incorporated 2nd Generation Chimeric Antigen Receptor-Modified T Cells

Abstract

While chimeric antigen receptor-modified T (CAR-T) cells have shown great success for the treatment of B cell leukemia, their efficacy appears to be compromised in B cell derived lymphoma and solid tumors. Optimization of the CAR design to improve persistence and cytotoxicity is a focus of the current CAR-T study. Herein, we established a novel CAR structure by adding a full length 4-1BB co-stimulatory receptor to a 28Z-based second generation CAR that targets CD20. Our data indicated that this new 2028Z-4-1BB CAR-T cell showed improved proliferation and cytotoxic ability. To further understand the mechanism of action, we found that constitutive 4-1BB sensing significantly reduced the apoptosis of CAR-T cells, enhanced proliferation, and increased NF-κB pathway activation. Consistent with the enhanced proliferation and cytotoxicity in vitro, this new structure of CAR-T cells exhibited robust persistence and anti-tumor activity in a mouse xenograft lymphoma model. This work provides evidence for a new strategy to optimize the function of CAR-T against lymphoma.

Keywords: CD20; 4-1BB; chimeric antigen receptor-modified T; immunotherapy; lymphoma.

Figure 1

Characterization of 2028Z-4-1BB CAR-T cells. (A, B) A schematic diagram of the CD20 targeting CAR constructs used in this study. An anti-human CD20 scFv was linked to CD28 and CD3ζ to generate a 2028Z construct. 4-1BB was linked to CD3ζ via a P2A sequence. (C, D) Flow cytometry analysis of CAR (C) and 4-1BB (D) expression on 2028Z and 2028Z-4-1BB CAR-T cells. (E) Western blot analysis of phosphorylated p65 in 2028Z and 2028Z-4-1BB CAR-T cells. GAPDH was used as a loading control. (F) Flow cytometry analysis of 4-1BB Ligand expression on 2028Z and 2028Z-4-1BB CAR-T cells. Representative results from one from three (C, D) and two (E) repeated experiments are shown. *P < 0.05, ***P < 0.001; NS: Not Significant.

Figure 2

4-1BB signal alters 2028Z-4-1BB CAR-T cell differentiation. (A) A schematic diagram of the long-term CAR-T expansion assay. T cells were infected with the indicated CAR lentivirus and stimulated with irradiated Raji cells every six days. Total cell number was recorded and CAR⁺ cell percentage was analyzed by flow cytometry. (B) The CD4⁺ and CD8⁺ CAR-T cell percentage in the long-term culture was analyzed by flow cytometry. (C) The cell surface expression of exhaustion markers PD-1, TIM-3, and LAG-3 was measured in CAR-T cells after stimulation with irradiated Raji cells at the indicated time point. (D) Relative proportion of naive (CD45RA⁺CD45RO^-CCR7⁺, TN), central memory (CD45RA^-CD45RO⁺CCR7⁺, TCM), effector memory (CD45RA^-CD45RO⁺CCR7^-, TEM), and effector (CD45RA⁺CD45RO^-CCR7^-, TEFF) T cells in 2028Z and 2028Z-4-1BB CAR-T cells at the indicated time point. Representative results from one from three (B, D) repeated experiments are shown. ***P < 0.001.

Figure 3

4-1BB signal improves in vitro proliferation of CAR-T cells. (A) Overall expansion of CAR-T cells in 2028Z and 2028Z-4-1BB CAR-T cell cultures. Relative cell proliferation was calculated by dividing the cell number by the cell number of Day 1. Experiments were repeated with three different donor-derived T cells (n = 3/group). Arrows indicate stimulation time points. (B) Apoptosis analysis of 2028Z and 2028Z-4-1BB CAR-T cells. Apoptosis was detected by active Caspase-3 staining. (C) Summarized active Caspase-3 staining data from three different donors. (D) Q-PCR analysis of the mRNA expression of Bcl2 in 2028Z and 2028Z-4-1BB CAR-T cells. Representative results of one from three (C, D) repeated experiments are shown. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 4

Profile of the cytotoxic ability of 2028Z-4-1BB CAR-T cells in vitro. (A, B) 2028Z and 2028Z-4-1BB CAR-T cells were co-cultured with NALM-6, NALM-6-hCD20 (A) or Raji (B) cells in triplicate at the different effector:target (E:T) ratios for 12-24 h. Relative cytotoxicity was calculated by analyzing the remaining tumor cells (CD3^-CD19⁺) by flow cytometry. (C) 2028Z and 2028Z-4-1BB CAR-T cells were co-cultured with Raji cells for 12 h. Cytokines secreted by 2028Z and 2028Z-4-1BB CAR-T cells were determined by CBA. (D) 2028Z and 2028Z-4-1BB CAR-T cells were co-cultured Raji cells for 12–24 h. The relative survival of CAR-T cells was calculated by analyzing CD3⁺CD19^- cells by flow cytometry. Representative results of one from three (A–D) repeated experiments are shown. *P < 0.05, **P < 0.01, ***P < 0.001; NS: Not Significant.

Figure 5

2028Z-4-1BB CAR-T cells show better persistence and anti-tumor effects in vivo. (A) A schematic diagram of the in vivo xenograft tumor model and CAR-T treatment protocol. (B, C) NSG mice were intravenously inoculated with 5x10⁵ Raji cells. Seven days later, tumor bearing mice were treated with 1x10⁷ 2028Z or 2028Z-4-1BB CAR-T cells. Nine days after the treatment, bone marrow, spleen (n = 3/group), and peripheral blood (n = 12/group) were collected for analysis of CAR-T cell (mCD45^-hCD45⁺hCD3⁺) persistence and Raji (mCD45^-hCD45⁺ hCD19⁺) tumor cell burden. (D) Kaplan-Meier analysis of the survival of mice (one point represents one mouse, n = 11/group). Representative results of one from three (B–D) repeated experiments are shown. *P < 0.05, **P < 0.01, ***P < 0.001; NS: Not Significant.

Figure 6

Proposed model of how 4-1BB signaling enhances the anti-tumor activity of 2028Z CAR-T cells. The NF-κB pathway is more active in 2028Z-4-1BB CAR-T cells than in 2028Z CAR-T cells. The expression of exhaustion marker PD-1 and pro-apoptotic active Caspase-3 is reduced in 2028Z-4-1BB CAR-T cells, compared with 2028Z CAR-T cells. These molecular mechanisms lead to enhanced proliferation, cytotoxicity ability, and anti-tumor activity of 2028Z-4-1BB CAR-T cells.