Structure and sequence engineering approaches to improve in vivo expression of nucleic acid-delivered antibodies

Monoclonal antibodies are an important class of biologics with over 160 Food and Drug Administration/European Union-approved drugs. A significant bottleneck to global accessibility of recombinant monoclonal antibodies stems from complexities related to their production, storage, and distribution. Recently, gene-encoded approaches such as mRNA, DNA, or viral delivery have gained popularity, but ensuring biologically relevant levels of antibody expression in the host remains a critical issue. Using a synthetic DNA platform, we investigated the role of antibody structure and sequence toward in vivo expression. SARS-CoV-2 antibody 2196 was recently engineered as a DNA-encoded monoclonal antibody (DMAb-2196). Utilizing an immunoglobulin heavy and light chain “chain-swap” methodology, we interrogated features of DMAb-2196 that can modulate in vivo expression through rational design and structural modeling. Comparing these results to natural variation of antibody sequences resulted in development of an antibody frequency score that aids in the prediction of expression-improving mutations by leveraging antibody repertoire datasets. We demonstrate that a single amino acid mutation identified through this score increases in vivo expression up to 2-fold and that combinations of mutations can also enhance expression. This analysis has led to a generalized pipeline that can unlock the potential for in vivo delivery of therapeutic antibodies across many indications. [Display omitted] In vivo delivery of gene-encoded antibodies can broaden global accessibility to life-saving antibody therapeutics. For gene-encoded platforms, therapeutic efficacy relies on achieving biologically relevant levels of antibody expression. Here, Patel and colleagues develop and employ a predictive pipeline to improve in vivo expression with the potential to facilitate clinical translation.