Dual-inhibitory domain iCARs improve the efficiency of the AND-NOT gate CAR T strategy

Abstract

CAR (chimeric antigen receptor) T cell therapy has shown clinical success in treating hematological malignancies, but its treatment of solid tumors has been limited. One major challenge is on-target, off-tumor toxicity, where CAR T cells also damage normal tissues that express the targeted antigen. To reduce this detrimental side-effect, Boolean-logic gates like AND-NOT gates have utilized an inhibitory CAR (iCAR) to specifically curb CAR T cell activity at selected nonmalignant tissue sites. However, the strategy seems inefficient, requiring high levels of iCAR and its target antigen for inhibition. Using a TROP2-targeting iCAR with a single PD1 inhibitory domain to inhibit a CEACAM5-targeting CAR (CEACAR), we observed that the inefficiency was due to a kinetic delay in iCAR inhibition of cytotoxicity. To improve iCAR efficiency, we modified three features of the iCAR-the avidity, the affinity, and the intracellular signaling domains. Increasing the avidity but not the affinity of the iCAR led to significant reductions in the delay. iCARs containing twelve different inhibitory signaling domains were screened for improved inhibition, and three domains (BTLA, LAIR-1, and SIGLEC-9) each suppressed CAR T function but did not enhance inhibitory kinetics. When inhibitory domains of LAIR-1 or SIGLEC-9 were combined with PD-1 into a single dual-inhibitory domain iCAR (DiCARs) and tested with the CEACAR, inhibition efficiency improved as evidenced by a significant reduction in the inhibitory delay. These data indicate that a delicate balance between CAR and iCAR signaling strength and kinetics must be achieved to regulate AND-NOT gate CAR T cell selectivity.

Keywords: AND-NOT logic gate; chimeric antigen receptor; immunotherapy; inhibitory CAR (iCAR); on-target, off-tumor toxicity.

Fig. 1.

Inhibition of CAR T cell cytotoxicity by the TROP2-PD1 iCAR is delayed. (A) The model illustrates the “AND-NOT”-gate CAR T strategy for specifically targeting CEA⁺ tumor cells. The CAR and iCAR target CEA and TROP2, respectively. The CEACAR consists of a FLAG tag, an scFv chain that recognizes CEA, an IgG 4/2NQ hinge, an IgG4 CH3 constant domain, a CD28 transmembrane domain (TM), a 41BB costimulatory (CS) domain, and a CD3ζ activation domain. The TROP2-PD1 iCAR consists of an HA tag, an scFv chain that recognizes TROP2, the IgG4 hinge, CH2, and CH3 constant domains, a CD28 TM, and a PD1 signaling domain. (B) CAR⁺/iCAR⁺ T cells can inhibit CAR T cell IFN-γ production as measured by ELISA 48 h after coculture of T cells with DU145 target cells that express CEA and/or TROP2. (C) CAR⁺/iCAR⁺ T cells that can specifically inhibit CAR T cell cytotoxicity after 48 h in coculture with DU145 target cells that express CEA and TROP2. Target cell presence was measured by total green object area (µm²/well) of DU145 target cells that express CEA and/or TROP2. (D) Inhibition of cytotoxicity is delayed in CAR⁺/iCAR⁺ T cells when cocultured with DU145 target cells. The cytotoxicity curve shown is a composite of three donors. Measurements of total green object area of GFP+ target cells were measured over ~140 h by Incucyte live cell image analysis at intervals of 2 h. Statistics are calculated based on the total green object area (µm²/well) at the last time point compared to the CAR⁺ only control. The data are reported as a mean ± SE (n = 3 donors). Statistics are performed using 1-way ANOVA analysis with Tukey multiple comparison correction. *P value ≤ 0.05, **P value ≤ 0.01, and ***P value ≤ 0.001.

Fig. 2.

Controlling the avidity of iCAR interactions reduces the delay in inhibitory signaling kinetics. (A) Engineered target cell lines have different surface level expression of CEA and TROP2 measured by flow cytometry. Histograms are representative images from one of three experiments comparing CEA and TROP2 expression of each target cell line. (B) Increasing the target antigen density reduces the delay in iCAR inhibition as measured by cytotoxicity over time. The cytotoxicity curves are representative images from one experiment measuring the total green object area of the target cells over time (μm²/well). (C) CAR⁺/iCAR⁺ T cells cocultured with target cells that express high levels of TROP2 have reduced delays in inhibition. The delay in inhibition was measured by calculating the area under each cytotoxicity curve. The AUC was normalized against the AUC calculated for untransduced T cells cocultured with target cells. The normalized AUC quantified is the mean ± SD (n = 3). (D) Representative histograms measuring the Mean Fluorescence Intensity (MFI) indicate the difference in CAR and iCAR surface expression between CAR T cell groups being tested. Groups have been transduced with CAR lentivirus at an MOI of 1 and iCAR lentivirus at an MOI of 1, 3, and 10, respectively. (E) Increasing the surface level expression of the iCAR in primary T cells reduces the delay in iCAR inhibition as measured by cytotoxicity over time. Representative cytotoxicity curves from one experiment are displayed comparing the killing ability of CAR T cells with different surface level expression of the iCAR when cocultured with DU145 target cells that express CEA or CEA and TROP2. (F) CAR⁺/iCAR⁺ T cells with higher iCAR surface expression have reduced delays in inhibition when cocultured with CEA^LO/TROP2^HI target cells. The delay in inhibition was measured by calculating the area under each cytotoxicity curve. The AUC was normalized against the AUC calculated for untransduced T cells cocultured with target cells. The normalized AUC quantified is the mean ± SD (n = 2). Statistics are performed using 1-way ANOVA analysis with Tukey multiple comparison correction. *P value ≤ 0.05, **P value ≤ 0.01, and ***P value ≤ 0.001.

Fig. 3.

Increasing the affinity of the TROP2-targeting iCAR does not increase inhibition efficiency. (A) Comparison of binding affinities between three different antibodies targeting Trop2. Binding affinity kinetics were measured by BLI using biosensors precoated with recombinant TROP2 protein. Antibodies were serially diluted in concentrations ranging from 400 to 12.5 nM. The binding values were obtained and plotted against concentrations of antibody (nM). (B) Representative histograms from one experiment that show CAR and iCAR surface expression is similar between all T cell groups being tested. CAR and iCAR expression are measured using flow cytometry with antibodies against the FLAG- and HA-tags on the engineered receptors respectively. (C) CAR⁺/iCAR⁺ T cells with the C3, B11, and H11 scFv show similar levels of cytotoxicity to each other. Representative cytotoxicity curves are displayed from one experiment where the total green object area (µm/well) of GFP⁺ DU145 target cells that express CEACAM5 and/or TROP2 were measured over approximately 160 h. (D) Area under the curve analysis of cytotoxicity curves. The delay in inhibition was measured by calculating the area under each cytotoxicity curve. The AUC was normalized against the AUC calculated for untransduced T cells cocultured with target cells. The normalized AUC quantified is the mean ± SD (n = 2) from two independent experiments. The significance values shown are comparisons between a control group. For the CEA^–/TROP2^– cell line, values are compared to the untransduced control. For the CEA⁺/TROP2^– cell line, values are compared to the CAR control. For the CEA^LO/TROP2^HI or CEA^HI/TROP2^LO cell lines, values are compared to the CAR⁺/iCAR⁺ (H11-Long) group. Statistics are performed using 1-way ANOVA analysis with Tukey multiple comparison correction. *P value ≤ 0.05, **P value ≤ 0.01, and ***P value ≤ 0.001.

Fig. 4.

iCARs containing signaling domains with ITIM motifs can reduce CAR T cell cytotoxicity. (A) The five different populations (untransduced, CAR, CAR + TROP2-PD1 iCAR, CAR + TROP2-Short iCAR, CAR + TROP2-Long iCAR) tested for each inhibitory signaling domain are plotted in a flow cytometry plot that corresponds to each color in the legend. The flow cytometry plot is a representative from one experiment. (B) CAR⁺/iCAR⁺ T cells that were engineered with inhibitory signaling domains with an ITIM motif can inhibit CAR T cell cytotoxicity as measured by total green object area (µm/well) of DU145 target cells that express CEA and/or TROP2 over 150 h. Each curve represents a coculture with the CAR T cell population represented by the color in the legend. These cytotoxicity curves are representative images from one experiment. (C) CAR⁺/iCAR⁺ T cells that were engineered with inhibitory signaling domains with an ITIM motif can inhibit CAR T cell cytotoxicity with a similar efficiency as the TROP2-PD1 iCAR as measured by area under the cytotoxicity curve. AUC was normalized to the untransduced population cocultured with the target cells. The normalized AUC quantified is the mean ± SD of at least two independent experiments (BTLA—n = 2; LAIR1—n = 2, SIGLEC9—n = 3, VISTA—n = 2). The significance values shown are comparisons between a control group. For the CEA^–/TROP2^– cell line, values are compared to the untransduced control. For the CEA⁺/TROP2^– cell line, values are compared to the CAR control. For the CEA^LO/TROP2^HI or CEA^HI/TROP2^LO cell lines, values are compared to the CAR + PD1 iCAR group. Statistics are performed using 1-way ANOVA analysis with Tukey multiple comparison correction. *P value ≤ 0.05, **P value ≤ 0.01, and ***P value ≤ 0.001.

Fig. 5.

DiCARs increases the efficiency of inhibition in the AND-NOT-gate CAR T strategy. (A) The models represent the structure of each DiCAR tested. The DiCARs are composed of a TROP2 scFv, the IgG4 Hinge, CH2, and CH3 constant domains, a CD28 TM, the PD-1 inhibitory signaling domain, and the additional inhibitory signaling domains PD-1, BTLA, SIGLEC-9, or LAIR-1. (B) The representative histogram indicates that iCAR/DiCAR surface expression level is similar between the groups of DiCARs being compared. The DiCAR surface expression was determined by flow cytometry for an HA-tag located on the N terminus of the iCAR/DiCAR. (C) Representative cytotoxicity curves of each CAR T cell population demonstrate that CAR T cell populations with a DiCAR have a reduced delay in inhibition compared to the TROP2-PD1 iCAR. CAR T cells were cocultured with DU145 target cells that express GFP and CEACAM5 and/or TROP2. Presence of target cells was measured by total green object area (µm²/well) over time as measured by Incucyte live cell image analysis over 150 h. (D) The delay in inhibition of the iCAR was measured by area under the cytotoxicity curve analysis of each cytotoxicity curve and normalized to the coculture with the untransduced T cell group. This AUC is a representative of one experiment in which triplicate wells were analyzed. Three biological replicates were performed and reported in SI Appendix, Figs. S6 and S7. The significance values shown are comparisons between a control group. For the CEA^–/TROP2^– cell line, values are compared to the untransduced control. For the CEA⁺/TROP2^– cell line, values are compared to the CAR control. For the CEA^LO/TROP2^HI or CEA^HI/TROP2^LO cell lines, values are compared to the CAR + PD1 iCAR group. Statistics are performed using 1-way ANOVA analysis with Tukey multiple comparison correction. *P value ≤ 0.05, **P value ≤ 0.01, and ***P value ≤ 0.001.