Computational Modeling of Bispecific Antibodies for CD30+ Hodgkin Lymphoma & Anaplastic Large Cell Lymphoma

Abstract

Lymphoma remains one of the most prevalent forms of cancer, with over 80,000 cases diagnosed annually in the U.S. alone. The CD30 receptor, overexpressed in lymphoma cells, is essential for cancer cell proliferation and survival, making it a critical therapeutic target. The CD3ε receptor subunit is found on immunocompetent lymphocytes and promotes activity in response to specific antigens. Bispecific antibodies (bsAbs) can bind to two antigens simultaneously, facilitating immune cell-mediated cytotoxicity. This research aims to identify Fab regions in silico that can bind to CD3e and CD30 receptors while also providing a model bsAb discovery pipeline. To formally investigate how scientists may consider using these technologies for bsAb development, this study hypothesizes that using computational analysis of various Fab fragments and the CD30 and CD3ε receptors will aid in recognizing antibodies that bind to CD8+ cells and lymphoma cells, respectively, resulting in the identification of potential treatments. The receptor and antibody structures were obtained from the AlphaFold 3 webserver and Protein Data Bank (PDB), respectively. Molecular docking was performed using HDOCK to model receptor-antibody interactions, and resulting complexes were evaluated using HADDOCK PRODIGY webserver, a computational program to compute binding energy. Key criteria of Fab fragment selection included binding energy, hydrogen bond count, and visual inspection of docking conformations. Notably, Fab fragment 1IQW demonstrated the highest binding affinity (-25.6 kcal/mol) and the highest hydrogen bond count of 28 for the CD3ε receptor. 1A5F had the highest binding affinity (-23.0 kcal/mol), though a low hydrogen bond count of 7, for the CD30 receptor. This research developed a model pathway for producing effective bispecific antibodies targeting the CD30 and CD3ε receptor chains while demonstrating the potential of computational techniques for bsAb development.