Large-scale analysis of single residue mutations on features of protein structure network
Loading...
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
IISER Mohali
Abstract
Single residue mutations in proteins can cause a wide range of functional and structural
affects that varies from no change to a complete loss of protein function/structure.
Consequently, many experimental studies of single residue mutations have contributed
immensely in deciphering function(s) of proteins. Insights into the pleiotropic affects of
amino acid mutations can assist in protein engineering as well as could shed light on
protein evolution. In the past, several large-scale systematic studies have been performed
to decipher consequences of single residue mutations on proteins structure and/or
function.
Most of these have largely been focused on predicting or exploring
thermostability of proteins or on determining the extent of conformational changes
between wild type (WT) and mutant (MT) structures. Interestingly, most previous studies
have ignored the subtle changes in the residue interactions in mutant proteins. The residue
interactions can be quantified using networks, which can be compared for WT-MT pairs to
analyse the local and global affect of single residue mutations. In such comparisons, it is
essential to distinguish network perturbation caused by mutation to that arising from
experimental conditions of crystallization.
In the present work, we have carefully
constructed two datasets, viz. WT-WT and WT-MT protein structure pairs primarily to
distinguish effect on protein networks due to experimental condition of crystallization
from mutations, respectively.
We constructed protein structure networks (PSNs) for
WT/MT proteins where residue interactions are represented as edges in PSNs. The edge
has been drawn between nodes (residues) based on distance, energy or atomic-contact
(Ohm).
Using these types of PSNs, we have analysed the impact of single residue
mutation on protein global/local network properties. The comparison of WT-MT pair of
proteins showed that global features of mutant proteins do not change significantly from
wild-type proteins except average Eigenvalue centrality showed correlation to
conformational changes. Importantly, local network property of Eigenvalue centrality
showed changes in Ohm based network. These changes were observed in both WT-WT
and WT-MT pairs, suggesting it is difficult to distinguish local network changes arising
from mutation to that from crystallization conditions.