Conformation-Specific Statistical Coupling Analysis of the α7 Acetylcholine Receptor

Abstract

It is well known that information contained in a protein sequence is what allows it to fold into its three-dimensional shape, which performs a specific function. It has been possible for some time to search for proteins with similar sequences, using bioinformatics tools such as BLAST. But it is also known that proteins with similar, or even the same sequence can adopt different structures and vice-versa. With this in mind, we look to use a method called Rosetta-HMMER to perform conformationally specific sequence searches in order to exploit this property of proteins. This method involves the use of Rosetta to redesign protein structures to fit a specified α-carbon backbone, and then uses HMMER to generate a sequence profile. This profile can then be used to query for sequences able to adopt the specified backbone structure. These collected sequences can then be aligned for the purpose of performing statistical coupling analysis. We have used this Rosetta-HMMER method in conjunction with available structures of the α7 acetylcholine receptor to show that distinct sequence profiles generated from different conformations of the same protein are capable of retrieving unique sets of natural sequences when used as a query. We have also shown that when these unique sets of natural sequences are used to perform statistical coupling analysis, different residues are identified as statistically coupled, potentially generating insight into residues that have more potential importance for one backbone conformation over another.