Host: John Hunt
Title: “Computational modeling and controlling immune recognition”
Abstract: The immune system manifests a powerful ability to specifically recognize and mount responses against non-"self" macromolecules. Immune responses are beneficial in natural infection and are the goal of vaccination, but are detrimental when directed against therapeutic proteins. This talk will present the development and application of integrated computational-experimental methods to better understand and direct immune recognition and response. In the context of vaccines, we have been investigating antibody-directed activities in both naturally infected human subjects and nonhuman primate vaccinees. We have identified associations among antibody features, immune functions, and protection from infection, and demonstrated that computational models can effectively make use of antibody properties to robustly predict outcomes. In the context of therapeutic proteins, we have been integrating models of immunogenicity with models of protein stability and function, in order to design variants that maintain therapeutic activity while reducing detrimental immune responses. In this manner, we have reengineered several therapeutic candidates including lysostaphin, a highly potent but immunogenic anti-staphylococcal enzyme, for which our variants maintain low antibody titers and are able to repeatedly rescue humanized mice from challenges with methicillin resistant Staphylococcus aureus (MRSA).
Chris Bailey-Kellogg is Professor of Computer Science at Dartmouth. His lab develops and applies computational methods for experiment optimization and analysis in studies of protein sequence-structure-function relationships. He is currently pursuing computationally-driven development of both protein therapeutics, redesigning non-human proteins so as to render them acceptable to the immune system, as well as vaccines, modeling and predicting immune responses to infection and vaccination. Chris earned a BS/MS with Sandy Pentland at MIT and a PhD with Feng Zhao at Ohio State and Xerox PARC, conducted postdoctoral research with Bruce Donald at Dartmouth, and started his faculty career at Purdue. His work has been recognized with an NSF Career award and an Alfred P. Sloan Foundation fellowship.