Temperature-Accelerated Molecular Dynamics Reveals That Insulin Can Undergo Large-Scale Conformational Reorganization On Binding to Its Receptor

Insulin regulates blood glucose levels in higher organisms by
specifically binding to and activating a transmembrane glycoprotein
known as the insulin receptor (IR). Detailed photo-crosslinking
studies have suggested that insulin molecule reorganizes on binding to
IR due to the displacement of its flexible B-chain C-terminus in the
presence of a tandem hormone binding structural motif called
CT-peptide. Although a detailed understanding of how insulin binding
leads to receptor activation remains elusive, our recently proposed
all-atom structural models of insulin/IR complexes provide information
valuable for exploring structure-function repertoire of IR. However,
the large-scale conformational change in the insulin was not studied
in these structural models due to the absence of then-unresolved
CT-peptide, and also the lack of adequate simulation techniques. We
have proposed a new conformational sampling algorithm,
temperature-accelerated molecular dynamics (TAMD) (2010,  PNAS,
107, 4961-4966), that provides an immediate opportunity to use
all-atom simulation to directly address these important gaps in our
understanding of insulin binding to IR. Using a combination of
Metropolis Monte-Carlo (MC) and molecular dynamics (MD) simulations,
we have constructed all-atom structural models of insulin/IR complexes
in the presence of (now-resolved) CT-peptide. We further apply TAMD to
the C-terminus of IR-bound (T and R) insulin molecules at a fictitious
thermal energy of 6 kcal/mol, and observe that while insulin remains
stably-bound to IR, it can undergo large-scale conformational change
in the B-chain C-terminus at ~50-ns time-scale resulting in the
exposure of hidden hydrophobic core of insulin. These results are
significant because we demonstrate at the atomistic-scale for the
first time that insulin molecules can undergo large-scale
reorganization on binding to IR in presence of CT-peptide. Moreover,
we suggest TAMD as a viable simulation technique to study such
large-scale conformational changes in proteins.