Coarse-Grained Model of Proteins Incorporating Atomistic Detail of the Active Site

We present a novel approach to explore the conformational space of globular proteins near their native state. It combines the advantages of coarse-grained models with those of all-atoms simulations, required to treat molecular recognition processes. The comparison between calculated structural properties with those obtained with all-atoms molecular dynamics simulations establishes the accuracy of the model. Our method has the potential to be extended to molecular recognition processes in systems whose characteristic size and time scale prevent an analysis based on all-atoms molecular dynamics.

Figure 1

Schematic representation of the regions described by the MM/CG model.

Figure 2

HIV-1 PR (a) and BACE (b) active sites. The balls and sticks representation depicts the aspartic dyad with the catalytic water, whereas the licorice representation depicts the substrate (MM region). Red lines represent the H-bond network. The line representation depicts a portion of I region. (c) Schematic representation of the catalytic site for aspartic proteases; ${\xi }_i$ indicates the distances between the ${\mathrm{C}}_{\alpha }$ atoms of the aspartic dyad and those of the closest two residues of the substrate.

Figure 3

RMSF of HIV-1 PR (a) and BACE (b), calculated by means of MD simulations, $\beta \mathrm{GM}$ model calculations, and MM/CG simulations.

Figure 4

Time evolution of the distance between the substrate and the active site for HIV-1 PR (a) and BACE (b). ${\xi }_i$ are defined as in Fig. 2c. Black and red lines refer to MD and MM/CG simulations, respectively.

Figure 5

H-bond number counts between the aspartic dyad along with the catalytic water and the substrate in BACE. Values were derived by MD and MM/CG simulations.