Monobodies (Mb) are small, synthetic antibodies based on the human fibronectin type III domain that exhibit high thermodynamic stability without disulfide bonds. To explore suitability of these domains as cancer-targeting molecules, we apply computational and experimental methods to develop a Mb that targets programmed-death-ligand 1 (PDL1). As some tumors overexpress PDL1 to inhibit T-cell responses, a PDL1-binding Mb could be used to promote successful immunotherapy. More broadly, computational methods applied to developing a PDL1-binding Mb can also be applied to developing new Mbs that target cancer antigens. Starting from the G9 Mb (PDB: 1ttg), we grafted the CDR3 loop of a known PDL1-binder (PDB: 5jds) into the FG loop of the Mb. This design was displayed on the surface of yeast and shown to bind PDL1 better than the G9 Mb, though not as well as a high affinity variant of PD1, the cognate receptor of PDL1. Using molecular dynamics simulations, we evaluated structural features that are unlikely to contribute to PDL1-binding. Using this approach, we selected amino acid positions for site-saturated mutagenesis and constructed a yeast library displaying Mb isoforms. Through iterative rounds of selection, we identified a Mb isoform with further improved PDL1-binding. Citation Format: Reed E. Harrison, Chi-Wei Man, Yingxiao Wang. Integrated computational and experimental design of a monobody targeting PDL1 [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 598.

No abstract