Speaker: Doug Barrick, Thomas C. Jenkins Professor - Department of Biophysics, Johns Hopkins University

Title: Using sequence statistics to explore protein cooperativity, evolution and design

Abstract: 

As a result of the large amounts of sequence data that has been obtained in the last few decades, many protein families are represented by as many as 100,000 different sequences. This high sequence density allows statistical features of these families to be extracted with high accuracy. We have been using information from multiple sequence alignments to answer questions related to protein stability, folding, and function, and evolution. In this seminar, I will describe three different approaches to answer such questions.

First, I will describe experiments using repeat proteins to learn about cooperativity and energy distributions in proteins. In these studies, we have used consensus sequence information from multiple sequence alignments to design protein arrays composed of identical repeats. This allows us to analyze folding using a nearest neighbor "Ising" model, providing an experimental dissection of cooperativity into local stability and long-range coupling.

Second, I will describe experiments that extend our consensus approach to non-repetitive globular proteins. A primary goal of these studies is to determine whether the very high stabilities of consensus repeat proteins can also be achieved in globular proteins. We find consensus globular proteins of various sizes and structures to be well folded, and in most cases, to be highly stable. Moreover, most of these proteins retain at least some level of biological function.

Third, I will describe recent efforts to identify specific sequence features of consensus design lead to enhanced stability. Using the homeodomain as a test case, we have used point substitutions to test the contributions of various sequence attributes (e.g., conservation, charge, burial) to stability. I will also present recent work that attempts to quantify and test the importance of sequence covariance between residue pairs in enhancing protein stability. These results have implications for protein design, bioinformatics, and protein evolution.