Barstar Has a Highly Dynamic Hydrophobic Core:  Evidence from Molecular Dynamics Simulations and Nuclear Magnetic Resonance Relaxation Data

• Published in: Biochemistry (Easton) 1998-08, Vol.37 (32), p.11182-11192
• Main Authors: Wong, Kam-Bo, Daggett, Valerie
• Format: Article
• Language: English
• Subjects: Bacillus - enzymology; Bacterial Proteins - chemistry; Computer Simulation; Crystallography, X-Ray; Models, Molecular; Nuclear Magnetic Resonance, Biomolecular; Protein Conformation; Protein Structure, Secondary; Ribonucleases - antagonists & inhibitors; Software; Solutions; Temperature; Thermodynamics
• ISSN: 0006-2960; 1520-4995
• DOI: 10.1021/bi980552i

The dynamic behavior of the ribonuclease inhibitor barstar has been investigated by molecular dynamics (MD) simulations in explicit water. Two 2.5 ns MD simulations were performed, and an ensemble of 25 000 structures was generated. This ensemble reproduces the solution structures and is consistent with the experimental structural restraints from NMR spectroscopy. Reorientation of the backbone NH bond vectors and side chain methyl groups was monitored by calculation of autocorrelation functions and the generalized S 2 order parameters. Order parameters derived for motion in the ∼100 ps time scale were compared with those obtained from NMR relaxation measurements. Consistent with experiment, the backbone NH bond vectors were relatively rigid. In contrast, the side chain methyl groups exhibited a wide dynamic range, from restricted motion comparable to that of the backbone to rapid unrestricted motion. The order parameters for the methyl groups correlate well with their spatial separation from the backbone and are residue-type dependent. Smaller S 2 axis values were observed for leucine methyl groups, in part due to side chain hopping between two predominant rotamers (g+t and tg-). Motions such as the flipping of aromatic rings and the hopping of leucine side chains were prevalent within the hydrophobic core, suggesting that the core is fluid-like with low energy barriers between native conformational substates. Thus, our studies suggest that the entropy of the native state can be significant and should not be discounted in thermodynamic considerations of protein folding. On the basis of our results, the side chain motion represents the primary source of the residual entropy of the native state and entropic considerations based solely on backbone dynamics would be incomplete.