Precision engineering of adalimumab fab variants enhances TNF-α neutralization and structural stability

The pro-inflammatory cytokine tumour necrosis factor-alpha (TNF-α) plays a substantial role in the advancement of various autoimmune diseases and is mostly suppressed by the monoclonal antibody adalimumab (Humira). This study utilized a rational in-silico method to design and optimize adalimumab Fab mutants with enhanced TNF-α binding affinity. A total of 512 mutants were produced with a Python automation script and FoldX. Eight mutants were selected for further analysed based on their stability (ΔG). The humanness ratings of the chosen variations were computed, demonstrating that they preserved similar human-like attributes. The anticipated expression in E. coli was elevated, with values between 0.93 and 0.94, as evidenced by the solubility analysis, which was analogous to the control (adalimumab) result of 0.94. Molecular docking revealed that all mutants displayed negligible RMSD values (≤ 0.001 Å) and uniform interface residues. FreeSASA and MM/GBSA studies were conducted to identify the top three candidates (Antibody-98, Antibody-354, and Antibody-374), which were subsequently confirmed by molecular dynamics simulations. Antibody-374 (yellow) and Antibody-98 (green) exhibit negligible variations (∼0.2-0.3 nm). Relative to the control, all three mutants demonstrated an elevation in total buried surface area (BSA), with Antibody-354 displaying the greatest measurement (5434.06 Å²). Antibody-98 exhibited 14-15 hydrogen bonds, which were sustained at about 8-10 bonds for the majority of the simulation. Antibody-354 exhibited a range of 6-8 number of hydrogen bonds, while Antibody-374 had a robust contact with 6-10 hydrogen bonds. In the binding free energy examination utilizing MM/GBSA, Antibody-374 exhibited the most advantageous ΔG_total (-36.67 kcal/mol), succeeded by Antibody-98 (-30.07 kcal/mol) and Antibody-354 (-25.47 kcal/mol). These antibodies surpassed the wild-type docked model (-24.80 kcal/mol) and the native crystal complex (-25.46 kcal/mol). The enhancements were ascribed to particular point mutations, namely G28S, R90E, Y96W (light chain), and W53Y/F, L102I/V (heavy chain), which augmented molecular interactions and complex stability. This study demonstrated the capability of computational antibody engineering to generate Fab variants with enhanced binding affinity and stability for TNF-α. These findings thus offer significant insights for the advancement of future biologic therapies.