Cancer therapeutic trispecific antibodies recruiting both T and natural killer cells to cancer cells

Abstract

T cells and natural killer (NK) cells are major effector cells recruited by cancer therapeutic bispecific antibodies; however, differences in the populations of these cells in individual tumors limit the general use of these antibodies. In the present study, trispecific antibodies were created, namely T cell and NK cell engagers (TaKEs), that recruit both T cells and NK cells. Notably, three Fc‑fused TaKEs were designed, TaKE1‑Fc, TaKE2‑Fc and TaKE3‑Fc, using variable fragments targeting the epidermal growth factor receptor on tumor cells, CD3 on T cells, and CD16 on NK cells. Among them, TaKE1‑Fc was predicted to form a circular tetrabody‑like configuration and exhibited the highest production and greatest cancer growth inhibitory effects. TaKE1 was prepared from TaKE1‑Fc by digesting the Fc region for further functional evaluation. The resulting TaKE1 exhibited trispecificity via its ability to bind cancer cells, T cells and NK cells, as well as comparable or greater cancer growth inhibitory effects to those of two bispecific antibodies that recruit T cells and NK cells, respectively. A functional trispecific antibody with the potential to exert strong therapeutic effects independent of T cell and NK cell populations was developed.

Keywords: CD16; CD3; EGFR; T cell and natural killer cell engager; cancer immunotherapy; therapeutic antibody; trispecific antibody.

Figure 1.

Construction of TaKEs. (A) Schematic diagrams of three types of TaKEs and their human IgG1 Fc fusion formats. (B) Schematic diagrams of the expression vectors for TaKEs with Fc. TaKE, T cell and natural killer (NK) cell engager; VH, variable heavy; VL, variable light; hOH, the VH regions of the humanized anti-CD3 antibody OKT3; hOL, the VL regions of the humanized anti-CD3 antibody OKT3; h5H, the VH regions of the humanized anti-EGFR antibody 528; h5L, the VL regions of the humanized anti-EGFR antibody 528; 3GH, the VH regions of the mouse anti-CD16 antibody 3G8; 3GL, the VL regions of the mouse anti-CD16 antibody 3G8; 2H, the VH regions of the mouse anti-EGFR antibody 225; 2L, the VL regions of the mouse anti-EGFR antibody 225.

Figure 2.

Preparation and evaluation of TaKE-Fcs. (A) Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of TaKE-Fcs after protein A purification. (B) Growth inhibition of EGFR-positive TFK-1 cells by each TaKE-Fc. Lymphokine-activated killer cells with T-cell phenotype cells were added to cancer cells at a ratio of 5:1. Data are presented as the mean ± 1 SD. TaKE-Fcs, Fc-fused T cell and natural killer cell engagers; EGFR, epidermal growth factor receptor.

Figure 3.

Preparation of TaKE from the Fc-fusion format. (A) Reducing SDS-PAGE analysis; lane 1, protein A chromatography-purified TaKE1-Fc; lane 2, after HRV3C protease digestion. (B) Gel filtration of TaKE1 after removal of HRV3C protease by glutathione Sepharose 4B chromatography followed by removal of Fc through protein A. The elution peaks were numbered as 1–3 (peak 1, TaKE1 dimer; peak 2, TaKE1 monomer; peak 3, impurities). (C) Reducing SDS-PAGE for each peak after gel filtration of TaKE1. Lanes 1–3 corresponded to the elution peaks in the chromatograph of B (lane 1, TaKE1 dimer; lane 2, TaKE1 monomer; lane 3, impurities). AU, absorbance unit; TaKE, T cell and natural killer cell engager.

Figure 4.

Binding activity of TaKE for EGFR-positive TFK-1 cells, CD3-positive T-LAK cells and NK-92/CD16A cells. Each cell was incubated with PBS as a negative control (shaded area) and with TaKE1 (open area), after which rabbit anti-Ex3 serum was added, followed by staining with FITC-labeled anti-rabbit IgG antibody. TaKE, T cell and natural killer cell engager; T-LAK, lymphokine-activated killer cells with the T-cell phenotype.

Figure 5.

Comparison of growth inhibitory effects of TaKE with each bispecific antibody for EGFR-positive TFK-1 cells. The purified antibodies were added to TFK-1 cells with CD3-positive T-LAK cells, NK-92/CD16A cells, or both cells, and growth inhibition was evaluated in an MTS assay. The ratio of each cell was indicated in each panel (A-E). Significant differences between TaKE1 and Ex16-scDb-HL (*P<0.05) or Ex3-scDb-LH (^†P<0.05) or effector cells alone (^‡P<0.05). Statistical analysis was conducted using the one-way ANOVA test. TaKE, T cell and natural killer cell engager; T-LAK, lymphokine-activated killer cells with the T-cell phenotype.

Figure 6.

Comparison of growth inhibitory effects of TaKE with the combination of two types of bispecific antibodies for EGFR-positive TFK-1 cells. The purified antibodies were added to TFK-1 cells with CD3-positive T-LAK cells, NK-92/CD16A cells, or both cells, and growth inhibition was evaluated in an MTS assay. The ratio of each cell and bispecific antibody is indicated in each panel (A-F). Concentrations for mixture of bsAbs are indicated as total concentration of bsAbs. Significant differences between TaKE1 and mixture of bsAbs (*P<0.05) or effector cells alone (^‡P<0.05). Statistical analysis was carried out using the one-way ANOVA test. TaKE, T cell and natural killer cell engager; T-LAK, lymphokine-activated killer cells with the T-cell phenotype; Ex3, Ex3-scDb-LH; Ex16, Ex16-scDb-HL.