Towards First-Principles Understanding of Biomolecular Adsorption

Doctoral thesis, 2006

A fundamental understanding of the interactions of biomolecules, such as proteins and DNA, with surfaces is of immense importance in numerous applications and poses a true challenge for theory. Density functional theory (DFT) is a quantum-mechanical tool well established for its ability to balance large system size with good accuracy in first-principles studies. However, traditional implementations cannot describe the nonlocal van der Waals interactions, that are essential in many soft matter systems, including biomolecule-surface interactions. Recently, a van der Waals density functional (vdW-DF) has been developed, making DFT applicable to soft matter. Since the functional is new, it is important to evaluate its performance. Here two versions of vdW-DF are applied to study dimers of benzene, naphthalene, anthracene and pyrene (molecules from the polycyclic aromatic hydrocarbon (PAH) family), as well as to interacting graphite sheets. Further, we consider several key systems of molecular adsorption, in particular, adsorption of benzene, naphthalene, phenol, and adenine on a graphite surface. These systems have aromatic rings, molecular units that also appear in the side chains of amino acids in proteins, and are also important in DNA interactions, where adenine is one of the four DNA base-pairs. They are ideal models for van der Waals bonded complexes and thus very suitable for vdW-DF performance tests. We calculate values for binding distances and energies and find these consistent with those of experiments and other theoretical studies using accurate wave-function based methods. However, the latter methods find even our small-dimer systems laborious, the larger dimers nearly impossible to treat, and cannot treat molecular adsorption, in which respect vdW-DF is more or less unique. Based on the comparison with experimental data, our results show great promise for broad application of the vdW-DF to soft matter systems. Simulation of a system with size close to that of complex biomolecules, such as proteins, requires a multi-scale approach. We present a model study of protein unfolding upon adsorption that applies a simple but exactly solvable two-dimensional lattice model, describing amino acids based on their hydrophobicities. This approach indicates the effect of hydrophobic and polar surfaces, as well as the factors influencing protein stability, and we propose experiments to test the predicted correlations. In summary, we treat biomolecule-surface interactions on two scales. The prospect of obtaining a deeper insight by bridging them to a complete description is judged very promising, given our steps taken towards a first-principles understanding of biomolecular adsorption.