# Computational Design of Broad-Spectrum Ebola Antibodies through Framework and Complementarity-Determining Region Synergistic Optimization

## TL;DR

Researchers designed improved antibodies against Ebola by combining computational predictions with lab experiments, enhancing their effectiveness against multiple virus strains.

## Contribution

A novel computational-experimental pipeline optimizes framework and CDR regions synergistically to improve antibody breadth and potency.

## Key findings

- FR grafting and CDR mutagenesis improved ADI-15878's neutralization against EBOV, BDBV, and SUDV by up to 17-fold.
- Docking-guided multi-site design boosted ADI-15946's SUDV neutralization by over 100-fold without losing EBOV/BDBV activity.
- Improved neutralization correlated with increased buried surface area and higher SPR-measured affinities.

## Abstract

Background: A major challenge in neutralizing antibody development lies in balancing improved potency with broad-spectrum efficacy, particularly for highly mutable pathogens such as the Ebola virus (EBOV). We developed a computational–experimental pipeline that combines multi-parameter in silico predictions with targeted wet-lab validation to enhance the breadth and potency of 2 previously reported pan-EBOV antibodies, ADI-15878 and ADI-15946. Methods: This pipeline employs multiple computational tools to prioritize affinity-enhancing mutations in complementarity-determining regions (CDRs) and to explore indirect effects through framework region (FR) modifications. These include mCSM-AB and mmCSM-AB for mutant screening, Foldseek for FR grafting, and ZDOCK for complex modeling. Candidate antibodies were expressed, purified, and evaluated using pseudovirus neutralization assays, flow cytometry binding assays, and surface plasmon resonance (SPR) affinity measurements. Result: For ADI-15878, ‌a combination of FR grafting and CDR mutagenesis generated the variant W32G-LC, which maintained structural integrity while improving neutralization against EBOV, Bundibugyo virus (BDBV), and Sudan virus (SUDV)—showing approximately 17-fold, 7-fold, and 2-fold enhancements over the parental antibody, respectively. For ADI-15946, ‌docking-guided deep mutation scanning‌ enabled the design of a multi-site light chain variant (H27Q/S52Y/G66R-LC and A50Y/S52Y/L54R-LC‌), which conferred greater than 40-fold and 100-fold increases in neutralization against SUDV, respectively, without compromising activity against EBOV/BDBV. Improved neutralization correlated with increased buried surface area (BSA) and enhanced SPR-measured affinities. Conclusions: The integrated FR–CDR optimization strategy combined with docking-guided multi-site design facilitates the rapid generation of antibody variants with improved breadth and potency. This modular pipeline provides a practical approach for updating therapeutic antibodies to combat diverse or emerging viral variants.