Hinge and Transmembrane Domains of Chimeric Antigen Receptor Regulate Receptor Expression and Signaling Threshold

Abstract

Chimeric antigen receptor (CAR)-T cells have demonstrated significant clinical potential; however, their strong antitumor activity may cause severe adverse effects. To ensure efficacy and safe CAR-T cell therapy, it is important to understand CAR's structure-activity relationship. To clarify the role of hinge and transmembrane domains in CAR and CAR-T cell function, we generated different chimeras and analyzed their expression levels and antigen-specific activity on CAR-T cells. First, we created a basic CAR with hinge, transmembrane, and signal transduction domains derived from CD3ζ, then we generated six CAR variants whose hinge or hinge/transmembrane domains originated from CD4, CD8α, and CD28. CAR expression level and stability on the T cell were greatly affected by transmembrane rather than hinge domain. Antigen-specific functions of most CAR-T cells depended on their CAR expression levels. However, CARs with a CD8α- or CD28-derived hinge domain showed significant differences in CAR-T cell function, despite their equal expression levels. These results suggest that CAR signaling intensity into T cells was affected not only by CAR expression level, but also by the hinge domain. Our discoveries indicate that the hinge domain regulates the CAR signaling threshold and the transmembrane domain regulates the amount of CAR signaling via control of CAR expression level.

Keywords: CAR-T cell therapy; chimeric antigen receptor; hinge domain; structure-activity relationship; transmembrane domain.

Figure 1

Illustration of vascular endothelial growth factor receptor 2 (VEGFR2)-specific chimeric antigen receptors (CARs) with hinge domain (HD) or hinge/transmembrane domains (HD/TMD) modification. (A) Top panel, retroviral vector constructs containing the murine VEGFR2-specific first-generation CAR genes for mouse T cells; lower panel, representation of HD-modified or HD/TMD-modified CAR proteins. The figure shows the number of amino acids (AA) used to generate HD/TMD/signal-transduction-domain (STD) and the molecular weight (MW) of each CAR. (B) Top panel, human VEGFR2-specific second-generation CAR mRNA constructs for human T cells; lower panel, representation of HD/TMD-modified CAR proteins. (C) AA sequences of CD3ζ, CD4, CD8α, and CD28 used for constructing the HD/TMD in this study. Cysteines that form interchain disulfide bonds are shown in bold and glycosylation sites are shown in bold italic. For CD8α-HD/TMD and CD28-HD/TMD, sequence alignment between mouse and human is shown. An asterisk denotes an identical AA, whereas alignment gaps are indicated by a hyphen. In CD28-HD, the area that forms the β-sheet structure is underlined.

Figure 1

Illustration of vascular endothelial growth factor receptor 2 (VEGFR2)-specific chimeric antigen receptors (CARs) with hinge domain (HD) or hinge/transmembrane domains (HD/TMD) modification. (A) Top panel, retroviral vector constructs containing the murine VEGFR2-specific first-generation CAR genes for mouse T cells; lower panel, representation of HD-modified or HD/TMD-modified CAR proteins. The figure shows the number of amino acids (AA) used to generate HD/TMD/signal-transduction-domain (STD) and the molecular weight (MW) of each CAR. (B) Top panel, human VEGFR2-specific second-generation CAR mRNA constructs for human T cells; lower panel, representation of HD/TMD-modified CAR proteins. (C) AA sequences of CD3ζ, CD4, CD8α, and CD28 used for constructing the HD/TMD in this study. Cysteines that form interchain disulfide bonds are shown in bold and glycosylation sites are shown in bold italic. For CD8α-HD/TMD and CD28-HD/TMD, sequence alignment between mouse and human is shown. An asterisk denotes an identical AA, whereas alignment gaps are indicated by a hyphen. In CD28-HD, the area that forms the β-sheet structure is underlined.

Figure 2

Expression of hinge domain (HD)-modified and hinge/transmembrane domain (HD/TMD)-modified chimeric antigen receptors (CARs) in mouse T cells. (A) CAR mRNAs were analyzed by reverse transcription and quantitative polymerase chain reaction, and their transcription levels were calculated relative to GAPDH mRNA as an endogenous control. (B) CAR expression on T cells were analyzed by flow cytometry using anti-HA-tag mAb (solid color histograms) or isotype control antibody (dashed white histograms). Each CAR expression level was calculated from the ratio of geometric mean fluorescence intensity (GMFI) when stained with the anti-HA-tag mAb to GMFI when stained with the isotype control antibody. (C) The mouse VEGFR2-binding of CAR-T cell on day 0 was analyzed by flow cytometry using an mVEGFR2-Fc chimera, and the relationship between the CAR expression level and mVEGFR2-binding of CAR-T cells was evaluated. (D) CAR proteins in the whole lysate were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and western blot analysis using anti-HA-tag mAb on day 0. The data were obtained from three independent tests. Statistical analysis was performed using Dunnett’s test against mV/3z/3z/3z, * p < 0.05; ** p < 0.01.

Figure 3

Functional characteristics of mouse T cells expressing CARs with different hinge domain (HD) and transmembrane domain (TMD). (A) Top panel, cytotoxic activity of HD-modified and HD/TMD-modified CAR-T cells four days after Rv transduction against mVEGFR2⁺ EL4 cells; lower panel, relationship between cytotoxicity at E/T ratio = 20 and CAR expression level. (B) Top panel, proliferation activity of HD/TMD-modified CAR-T cells following mVEGFR2-stimulation; lower panel, relationship between proliferation activity upon stimulation with mVEGFR2-Fc 20 ng/mL and CAR expression level. (C) Top panel, cytokine-producing ability of the above cells: interferon-γ, tumor necrosis factor-α, and interleukin-2; lower panel, relationship between cytokine-producing ability upon stimulation with mVEGFR2-Fc 200 or 2000 ng/mL and CAR expression level. The data are representative of at least two independent tests. Statistical analysis was performed using Dunnett’s test against mV/3z/3z/3z; * p < 0.05, ** p < 0.01; or Student’s t-test; ^† p < 0.05, ^†† p < 0.01.

Figure 4

Expression and function of hinge/transmembrane domains (HD/TMD)-modified CARs in human T cells introduced CAR mRNA by electroporation. (A) CAR mRNAs were analyzed by reverse transcription and quantitative polymerase chain reaction, and their amount were calculated relative to GAPDH mRNA. (B) CAR expression on T cells were analyzed by flow cytometry using anti-HA-tag mAbs (solid color histograms) or isotype control antibody (dashed white histograms). Each CAR expression level was calculated from the ratio of geometric mean fluorescence intensity (GMFI) when stained with the anti-HA-tag mAb to GMFI when stained with the isotype antibody. (C) Top panel, cytotoxic activity of human CAR-T cells cultured for 24 h after electroporation against hVEGFR2⁺ L1.2 cells at E/T ratio = 10; lower panel, cytotoxicity per CAR expression level. The data were obtained from three independent tests. Statistical analysis was performed using Student’s t-test; ^†† p < 0.01.

Figure 5

Role of hinge domain (HD) and transmembrane domain (TMD) in CAR. (A) Role of HD in the expression modality and membrane transport efficiency as well as CAR signaling threshold. (B) Role of TMD in CAR intracellular dynamics and surface expression stability. (C) Strategy to adjust CAR-T cell function by modifying CAR HD/TMD.

Figure 5

Role of hinge domain (HD) and transmembrane domain (TMD) in CAR. (A) Role of HD in the expression modality and membrane transport efficiency as well as CAR signaling threshold. (B) Role of TMD in CAR intracellular dynamics and surface expression stability. (C) Strategy to adjust CAR-T cell function by modifying CAR HD/TMD.