4 Establishing a tumor cell model to evaluate the bystander effect of antibody-drug conjugate (ADC) therapies

Background

Bystander killing is an important phenomenon in ADC therapeutics. Current bystander killing evaluation experiments often involve mixing two different tumor cells with positive or negative tumor specific targets, however tumors in patients typically contain the same type of tumor cells with varying expression levels of specific targets. To address this issue, we genetically edited a TROP2-positive non-fluorescent tumor cell line into a TROP2 knocked out fluorescent tumor cell line. The treatments of ADC drug candidates on the mixtures of these two kinds of cells can facilitate the analysis of how ADC drugs respond to heterogenous tumors with varying ratios of same-type tumor target positive cells to bystander tumor cells.

Methods

The full-length coding sequence of human TROP2 in human pancreatic cancer cell line BxPC-3 was replaced with luciferase and green fluorescent protein (GFP) to generate the B-Luc-GFP KI, TROP2 KO BxPC-3 tumor cell line. To validate the fluorescence of this tumor cell model, Bright-Glo™ luciferase assay and flow cytometry were used to detect the luciferase and GFP reporters, respectively. To confirm the human TROP2 knock-out in this cell model, the expression of human TROP2 was analyzed by flow cytometry. To evaluate the in vitro bystander killing effect of the TROP2-ADC drug, B-Luc-GFP KI, TROP2 KO BxPC-3 and BxPC-3 cells were co-cultured at ratios of 1:1, 1:0, and 0:1, and treated with the TROP2-ADC drug candidate for 3 days. Adherent cells were collected and counted by flow cytometry. To evaluate the in vivo bystander killing effect of the TROP2-ADC drug, cell-derived xenograft (CDX) model was established by inoculating B-Luc-GFP KI, TROP2 KO BxPC-3 and BxPC-3 cells at a specific ratio into immunodeficient mice, and tumor volumes post treatment were recorded by direct measurement and luminal signal.

Results

Human TROP2 protein was not detected, but a strong luminescence signal and the GFP fluorescence were detected in B-Luc-GFP KI, TROP2 KO BxPC-3 cells. The targeted killing of TROP2-positive BxPC-3 cells and the bystander killing of TROP2-negative B-Luc-GFP KI, TROP2 KO BxPC-3 cells by TROP2-ADC treatment were observed both in vitro and in vivo.

Conclusions

Our data demonstrates that the B-Luc-GFP KI, TROP2 KO BxPC-3 cell model is a reliable tool for the preclinical evaluation of the bystander killing effect of candidate ADC therapeutic drugs, both in vitro and in vivo.