(305b) Engineering Latency-Associated Peptide for Potent TGF-? Inhibition

Recent advances in cancer treatment, such as CAR T cell therapy and checkpoint inhibition, have revitalized the immunotherapy field. These therapeutic strategies go further than solely targeting cancer cells, addressing challenges in the tumor microenvironment that hinder endogenous immunity. Though successful in a subset of cases and conditions, many forms of cancer remain elusive to these treatments. One potential approach to augment treatment efficacy is combining multiple agents to simultaneously affect many disease-sustaining mechanisms. Though some intriguing biologics are available, there is an urgent need for potent, easily produced inhibitors for a variety of cancer microenvironment constituents. One intriguing alternative to developing novel inhibitors from scaffold proteins is the engineering of natural molecules to enhance their innate functions.

This talk will describe the application of this approach to engineering latency-associated peptide (LAP), the pro-peptide of transforming growth factor beta (TGF-B), to enhance its natural inhibitory function. TGF-B is a soluble signaling protein that has a variety of functions including regulation of the adaptive immune system and coordination of wound healing. Though biologically important, expression and dysregulation of TGF-B and the resulting signaling events have been implicated in tumorigenesis and progression of a variety of solid malignancies. Natural regulation of TGF-B activity involves binding to LAP, which sterically hinders TGF-B from interacting with its natural receptors. However, the TGF-B – LAP complex has approximately 100-fold weaker binding affinity than the TGF-B – TGF-B receptor II complex. To overcome this disparity, we are applying a combination of structure-guided protein engineering and high-throughput screening approaches to develop LAP variants with increased binding affinity for TGF-B. Using crystallographic data for the TGF-B – LAP complex, we performed saturation mutagenesis studies in silico to identify potentially beneficial mutations in interacting residues, resulting in the design of constrained variant libraries. Screening of these libraries using yeast surface display and directed evolution with inhibition as a selective pressure will be discussed. Overall, this approach presents a framework for natural molecule engineering that can be generally applied to generate a host of potent inhibitory proteins.