Navigation and service

Probing Biological Water With First-Principle Simulations

Principal Investigator: Thomas Kühne, Institut für Physikalische Chemie, Johannes Gutenberg-Universität Mainz (Germany)
HPC Platform: JUQUEEN of JSC
JSC Project ID: hmz32

Proteins are the workhorse molecules of life, which is due to their participation in essentially every structure and activity of life. However, in the absence of water as a solvent they lose their function in biological systems. The collection of one to two layers of interfacial water molecules surrounding proteins is generally referred to as “biological water”. The surface of a protein with its hydrophobic and hydrophilic amino acids is very complex, which makes it notoriously difficult to directly study its hydration dynamics experimentally. Instead, large-scale Molecular Dynamics (MD) simulations are a powerful tool to untangle the contributions originating from the various aspects of protein hydration and to obtain atomic-scale information on different time scales.

Till now, most MD simulations devoted to this topic were based on so-called classical force fields, where the intermolecular interactions are modelled with springs and balls. However, such models are generally too simplistic to model the subtle quantum mechanical effects of hydrogen-bonding that conspire to make water unique, such as large permanent dipole and polarisability effects, the cooperativity of hydrogen-binding, as well as the recently found instantaneous asymmetry of liquid water (Fig. 1) [1,2,3]. Therefore, a first-principles based approach, such as density functional theory (DFT) based ab-initio MD, where the interatomic forces are calculated on-the-fly by accurate electronic structure calculations is very attractive since many of these limitations can be removed. Nevertheless, the high computational cost of ab-initio MD simulations has to date severely limited the attainable length and time scales.

With the general availability of large supercomputers such as those of the Gauss Centre for Supercomputing (GCS) and recent algorithmic developments suitable for such massively parallel supercomputers, such as the second-generation Car-Parrinello (CPMD) method [3,4], studying biological water from first-principles is made possible. Hence, a team of Mainz based scientists that consists of 5 senior researchers and 3 doctoral students, is currently investigating the structure, vibrational dynamics, and energetics of biological water at the surface of a mini-protein known as Anti-freeze protein (AFP1) (Fig. 2). AFP1 helps organisms to survive below zero degree Celsius by inhibiting ice growth. Specifically, the researchers are interested to investigate the underlying physical nature for AFP1’s preference of ice to water. The novelty of this study relies on providing new insights on the correlation between the electronic structure and vibrational spectroscopy [5], the coupling of biological water with AFP1, as well as the relevance to its biological function.

These endeavors would further shed light on the peculiar properties of interfacial water, its dynamical behavior and role in other important biological processes, such as protein folding.

Probing Biological Water With First-Principle SimulationsCopyright: Johannes Gutenberg-Universität Mainz, Institute for Physical Chemistry

Figure 1: Model of liquid water, which demonstrates the instantaneous asymmetry of the hydrogen-bond strengths.

Probing Biological Water With First-Principle SimulationsCopyright: Johannes Gutenberg-Universität Mainz, Institute for Physical Chemistry

Figure 2: First-principle simulations of AFP1 protein with explicit solvent, in total 3093 atoms. Biological water molecules and protein side chains are shown in CPK representation. Bulk molecules are represented as blue dots. Simulations were performed with CP2K suite of programs (

[1] T. D. Kühne and R. Z. Khaliullin, Nature Commun. 4, 1450 (2013).
[2] T. D. Kühne and R. Z. Khaliullin, J. Am. Chem. Soc. 136, 3395 (2014).
[3] R. Z. Khaliullin and T. D. Kühne, Phys. Chem. Chem. Phys. 15, 15746 (2013).
[4] T. D. Kühne, WIREs Comput. Mol. Sci. 4, 391 (2014)
[5] C. Zhang, R. Z. Khaliullin, D. Bovi, L. Guidoni and T. D. Kühne, J. Phys. Chem. Lett. 4, 3245 (2013).

Scientific Contact:

Prof. Dr. Thomas Kühne
Johannes Gutenberg-Universität Mainz
Institut für Physikalische Chemie
D-55099 Mainz/Germany

November 2014