Poster abstracts

Poster number 7 submitted by Nicholas Callahan

Mutual Information Guided Mutations in Adenylate Kinase LID Domain

Nicholas Callahan (Biophysics Program, Ohio State University), Dr. Deepa Perera (Department of Chemistry, Muskingum University), Deepti Mathur (Department of Chemistry, Ohio State University), Brandon Sullivan (Ohio State Biochemistry Program, Ohio State University), Venuka Durani (Department of Chemistry, Ohio State University)

Abstract:
Given the close packing of atoms in protein structures, it may be that protein structure is defined not only by single fortuitous mutations, but by clusters of complementary events. Statistical analysis of multiple sequence alignments has long been used to find mutations affecting stability and function, and new, more-sensitive calculations promise to expand the complexity of problems that can be addressed. We used Mutual Information analysis to find compensatory mutations in the LID domain of de-activated bacterial Adenylate Kinase (ADK), a model protein that regulates adenosine phosphate supplies, without significantly altering stability.

In our original constructs, we mutated the LID domain in gram-positive B. subtilis ADK to remove a zinc-finger-like chelating motif and replace it with the analogous residues from gram-negative E. coli, as identified by Pfam’s multiple sequence alignment. This dramatically reduced enzymatic activity. We then located positions with high mutual information to the altered motif and mutated the B. subtilis residues to their E. coli counterparts. These new mutations partially rescued activity, but did not show a change in melting temperature as determined by CD spectroscopy. Our interpretation of these data is that it shows the utility of Mutual Information in locating functionally relevant motifs and in discovering interactions whose complexity may make them difficult to notice by structural inspection.

Keywords: bioinformatics, protein structure