Researcher(s)
- Tamar Peleg, Applied Molecular Biology & Biotechnology, University of Delaware
Faculty Mentor(s)
- Jason Gleghorn, Biomedical Engineering, University of Delaware
Abstract
Controlling cell-cell interactions through engineered adhesion molecules is fundamental to immunotherapy, tissue engineering, and cancer treatment. Leukocyte trafficking depends on sequential adhesion events mediated by L-selectin and LFA-1, making these receptors attractive models for engineering targeted cell interactions. Designing and validating targeting ligands is often slow and expensive, but computational protein design enables high-throughput design of potential binding ligands. We developed a modular framework that accelerates binder design by leveraging gradient-based sequence optimization. Using this framework, we computationally designed LFA-1 and L-selectin mimetics to reproduce the physiological receptor-mediated interactions. A Designed Ankyrin Repeat Proteins (DARPin) scaffold was chosen because of high-yield expression in E. coli, thermal stability, and low cost of production. In the generation phase, the DARPin variable residues were optimized for binding affinity, while in the critic phase that follows, sequences were evaluated based on specific computational oracle property predictors. State-of-the-art protein structure predictors, like ESMFold2, were leveraged to score the predicted interfaces of targets and designs. Selected candidate protein binders will then be experimentally validated by conjugation to a lymphocyte membrane-wrapped particle to mimic the physiological function of LFA-1 and L-selectin. Because receptor accessibility depends on ligand presentation above the membrane surface, it is hypothesized that varying PEG spacer lengths (2 kDa vs. 5 kDa) will optimize ligand presentation and promote rolling and firm adhesion. Methods for site-specific covalent anchoring that preserve ligand orientation are being evaluated for both computationally designed binders and recombinant controls. This integrated workflow enables rational engineering adhesion molecules to control cell-cell interactions through computational design and membrane modification.