Abstract: Objective  To study the cytotoxicity of bi-chimeric antigen receptors modified T- lymphocytes (BiCAR-T) on the human acute myeloid leukemia (AML) cell line HL60 in vitro and the antitumor effects of BiCAR-T on the NOD SCID mouse model of AML in vivo. Methods  The BiCAR-T were prepared and the expression of chimeric antigen receptor (CAR) of prepared BiCAR-T was analyzed by flow cytometry. In vitro study was divided into two groups: the experiment group (BiCAR-T) and the control group (T lymphocyte). The killing rate of BiCAR-T in vitro on HL60 cells was determined by CCK8 assay and the level of interferon-γ (IFN-γ) secreted from BiCAR-T co-culturing with HL60 cells for 48 hours was detected by enzyme linked immunosorbent assay (ELISA) at different effect/target ratios (5∶1, 10∶1, 20∶1). The NOD SCID mice AML model was established by the injection of HL60 cells through tail vein and  used to assess the antitumor effects in vivo. The mice were randomly divided into three groups according to the random number table: the blank control group receiving 0.9% NaCl 0.2 ml through tail vein， the model group and the treatment group receiving 1×10⁷ HL60 cells in 0.2 ml phosphate buffer saline (PBS). After 20 days, the treatment group was injected with 2×10⁷ BiCAR-T in 0.2 ml PBS 3 times a week for 2 weeks, while the other two groups received 0.9% NaCl 0.2 ml. The pathological changes in the mice livers and spleens were observed after 2 weeks of treatment. Results  The CAR expression rates of BiCAR-T were more than 50.00%. In vitro experiments proved that the killing rates of BiCAR-T in the experimental group and T lymphocytes in the control group on HL60 cells were (25.43±1.32)% vs. (16.18±0.75)%, (50.33±3.11)%  vs. (25.47±1.27)%, and (85.89±3.96)% vs. (49.45±2.77)% at different effect/target ratios (5∶1, 10∶1, 20∶1). The killing efficiency of BiCAR-T and T lymphocytes on HL60 cells was significantly different (F=404.17, P＜0.001); the killing efficiency of BiCAR-T and T lymphocytes on HL60 cells was significantly different at different effect/target ratios (F=548.09, P＜0.001); and the killing efficiency on HL60 cells in the experimental group (BiCAR-T) was significantly higher than that in the control group  (T lymphocytes) at different effect/target ratios (F=45.36, P＜0.001). The IFN-γ levels secreted from BiCAR-T in the experiment group and T lymphocytes in the control group coculturing with HL60 cells after 48 h were (435.65±20.44)pg/ml vs. (356.75±19.87)pg/ml, (1 639.98±95.75)pg/ml vs. (1 109.37±80.98)pg/ml, and (3 467.43±187.54)pg/ml vs. (2 245.52±112.66)pg/ml.  The IFN-γ level in the experiment group (BiCAR-T)  and the control group  (T lymphocytes) was significantly different (F=156.24, P＜0.001); the IFN-γ level was significantly different at different effect/target ratios (F=857.67, P＜0.001); the IFN-γ level in the experimental group  (BiCAR-T) was significantly higher than that in the control group  (T lymphocytes) at different effect/target ratios of 5∶1, 10∶1, 20∶1, respectively (F=46.31, P＜0.001). The result of hematoxylineosin staining (HE) staining showed that leukocyte infiltration in the treatment group was significantly decreased compared with the model group. Conclusion  The experimental results showed that BiCAR-T is a kind of efficient targeted immunocyte modified by gene engineering, and it can significantly inhibit leukocyte infiltration of AML in vivo and in vitro.

Key words: Leukemia, myeloid, acute, T-lymphocytes, Chimeric antigen receptor