10月23日: Applied protein design: Engineering molecular interactious in comples systmes.
报告人：David F. Green
(Associate Professor & Graduate Program Director, Applied Mathematics and Statistics, Stony Brook University)
Interactions between two or more proteins form the basis of a great deal of biological function, from viral infectivity to embryonic development. While a significant amount of work has been done to elucidate the structural determinants of affinity in isolated protein complexes, the functional association of proteins in vivo involves many additional factors. For example, some proteins must be highly specific in their interactions, while others must bind more promiscuously. Additionally, many interactions involve glycosylated proteins, with poorly understood roles for the sugar moities. The design of novel, functional protein complexes requires that these issues are considered. We are working towards the development of robust protocols to engineer functional protein complexes to work in complex biological systems. Our work ranges from the detailed study of how affinity and specificity are modulated in natural systems to the engineering of novel complexes for specific applications.
Complex patterns of specificity are ideally highlighted in the heterotrimeric G-protein signaling pathways, where multiple closely-related proteins work to transduce a wide range of signals into a diverse set of biological responses. Using a combination of detailed molecular simulation and novel applications of protein design algorithms, we have begun to dissect the mechanisms underlying the specific interactions made by components of the G-protein trimer itself. Additionally, biochemical kinetic models of the signaling pathway have shown that multiple qualitative system responses can be achieved with modest variations in parameters of the system.
Few systems demonstrate the importance of glycosylation better than the recognition of target cells by human retroviruses. The cell targeting glycoprotein of HIV-1, for example, is nearly 50% carbohydrate, with conserved glycosylation patterns across viral clades. Lectins that bind the viral-surface carbohydrates have been shown to be potent inhibitors of infection and are pre-clinical candidates for use as topical virucides. We have developed robust modeling protocols that explain the oligosaccharide specificity of the best characterized of these lectins, Cyanovirin-N, and have engineered hyperstable variants with identical binding specificity. Preliminary results also suggest several avenues for the engineering of enhanced affinity and specificity for particular. oligosaccharide targets.