Doug Barrick, PhD
Professor
Thomas C. Jenkins Department of Biophysics
John Hopkins Krieger School of Arts and Sciences

Prof. Doug Barrick recently published “Biomolecular Thermodynamics: From Theory to Application”

“Using consensus sequences to learn about protein folding cooperativity, stability, and function”

The use of consensus sequences has recently been applied successfully to protein engineering studies by our lab and others.  In this seminar, I will discuss applications of consensus sequence information to linear repeat proteins and globular proteins to learn about protein folding and cooperativity, and to better understand the relationships between consensus sequence and higher-order sequence correlations in specifying protein structure, folding stability, functional properties, and evolution.

Linear repeat proteins provide an excellent framework for studying protein stability and cooperativity, simplifying structural complexity to a length scale of 20-50 residues due to translational symmetry.  Moreover, the linear architecture of repeat proteins permits the removal and addition of the repeated structural element, facilitating nearest-neighbor (1D-Ising) of folding, which provides a unique thermodynamic measure of cooperativity in folding.  To take advantage of these simplifying features we have designed repeat proteins with identical repeats, using consensus information from multiple sequence alignments.  Recent successes with this approach will be presented, along with generalities and variations in underlying cooperativity parameters.  In addition, Ising-analysis will be presented for a series of Rosetta-designed helical repeat proteins that bear no sequence or structural similarity to naturally occurring proteins.

In addition, we have been evaluating the applicability of consensus-based information to engineering globular proteins.  Starting with the homeodomain fold, we created a highly thermostabilized consensus variant that adopts the homeodomain fold, folds extremely fast, and to our surprise, binds cognate DNA sequences with nearly 100-fold greater affinity than the Drosophila engrailed homeodomain.  More recently, we have had success with other folds, including the a-helical spectrin domain and a mixed b/a and b-sheet proteins.  We have had a high success rate in creating folded and hyperstabilized proteins from consensus information.  Moreover, for three enzyme targets we have investigated, we obtain significant catalytic activity.  The implications of these findings to protein design, dynamics, and evolution will be discussed.