Poster abstracts
Poster number 36 submitted by Puyi Ma
Realistic conformational ensembles of active and inactive K-Ras by MD simulations validated against quantitative NMR data
Puyi Ma (Biophysics Graduate Program), Mamata Basnet (Department of Chemistry and Biochemistry), Eric Fagerberg (Department of Chemistry and Biochemistry), Xinyao Xiang, Lei Bruschweiler-Li (Department of Chemistry and Biochemistry)
Abstract:
K-Ras is one of the most oncogenic human proteins being present in 25% of all cancers. Once mutated, it remains stuck in the signaling-active GTP-bound state promoting cell proliferation and cancer development. The functional mechanism of K-Ras is poorly understood as the functionally critical Switch I and II regions could not be observed in native K-Ras·GTP by common experimental approaches due to the intrinsic hydrolysis activity. On the other hand, inactive K-Ras·GDP is highly dynamic on sub-microsecond timescales and is not well represented by static crystal structures. Molecular Dynamics (MD) simulations have the potential to provide a more complete picture of K-Ras function, but the lack of atomic-detail experimental structural dynamics information has prevented their validation. To fill in this critical gap, we performed extended 5 - 20 µs MD simulations with multiple replicas of both wild-type K-Ras·GTP and K-Ras·GDP which were rigorously validated against a large body of experimental NMR data, including 15N-1H S2 order parameters and residual dipolar 1H-15N couplings (RDCs). The MD results for inactive K-Ras·GDP shows heterogeneous behavior among different replicas, but the careful selection of a 10 µs sub-trajectory provides good agreement with experiment. For active K-Ras·GTP, an RDC-restrained simulation was performed first to identify a better MD starting structure subsequently producing a conformational ensemble consistent with experiment. The resulting conformational ensembles provide the first atomic-detail K-Ras models in solution consistent with experimental NMR data offering a realistic view of the structural dynamics of K-Ras. These conformational ensembles can reveal important information about potential protein binding modes and cryptic pockets as an important resource for in silico drug discovery of this key oncogenic protein.
Keywords: K-Ras, MD Simulation, NMR