Development of a Novel 'Imbalanced' CD20/CD3 Bispecific Antibody by Computational Design

摘要

Bispecific antibodies(BsAb) with a CD3 binding arm to recruit and activate T cells have been widely studied for cancer therapy. However, many of these BsAb formats eliminated antibody effector functions considering safety concerns. To "safely" maintain the effector functions and expand the applications of these BsAbs, we developed a novel "imbalanced bispecific antibody technology platform" based on computational antibody design. A CD20/CD3 "imbalanced bispecific antibody" as first proof of concept, was designed to contain the following features: 1) CD3 binding affinity was significantly reduced to increase safety. 2) ADCC/CDC effector functions were maintained in order to broaden clinical implementation. 3) Biochemical and biophysical features of the CD20 binding arm and the CD3 binding arm of this "KIH + common light chain" version BsAb were differentiated to enable better isolation of BsAb during downstream purification process. Because of these designs, using in vitro assay system, this antibody showed better CD20+ Raji cell killing efficacy than that of CD20 homodimer and Rituxan in the presence of human PBMCs. Meanwhile, this antibody failed to kill CD3+ Jurkat cells or deplete normal T cell under the same condition. Therefore, we inferred that this antibody may illustrate promising wider clinical applications than current anti-CD20 cancer therapies: 1) Compared to Rituxan and it's biobetters, this antibody has T cell recruiting function; 2) Compared to CAR-T/ other T cell recruiting therapies, this antibody maintains functional effector function; 3)this antibody didn't show any safety concern in vitro. Taken together, the CD20/CD3 antibody and novel platform we created may contribute to the fulfillment of the unmet needs in the field of targeted cancer therapies.