Self-Assembly and Crystallization of Proteins

The goal of the project has been the study of various self-assembly processes – the spontaneous and reversible organization of matter into ordered arrangements without any external direction. The structures arising from self-assembly are diverse and we have studied different self-assembly mechanisms using various techniques based on the principles of statistical mechanics. On the one hand we have focused on the self-assembly of simple building blocks like macromolecules into bulk thermodynamic phases such as fluids or vapours, i.e. we investigated phase transitions in these systems. Studying the phase behaviour of macromolecules such as proteins or colloids has turned out to be a big challenge both for computer simulations and for theoretical models since the range of the effective interaction between the molecules is significantly smaller than the size of the building blocks and an accurate localization of the coexistence curve and the critical point becomes extremely difficult. Using simple model systems and a rather sophisticated theoretical approach we have been able to accurately predict the liquid-vapour transition of protein and colloidal suspensions and we confirmed Noro-Frenkel's extended law of scaling according to which the properties of a short-ranged system at a given temperature and density are independent of the detailed form of the interaction, but just depend on the value of the second virial coefficient – that can be regarded as an integral over the interaction. In the second part of the project we have studied the self-assembly using computer simulations. In particular, we have been using molecular dynamics simulations to study the self-assembly of liposomes. These systems have become established as one of the most reliable drug delivery systems. Lipsomes are self-assembled spherical vesicles, that are typically 100nm in diameter, consisting of a lipid bilayer that is surrounding an aqueous core. The great advantage of liposomes which renders them particularly attractive for drug delivery systems is their ability to both encapsulate hydrophobic cargo in the lipid membrane interior and hydrophilic cargo in the aqueous lumen or attached to the surface of the vesicle. In computer simulations we have seen both the formation of (metastable) spherical micelles and the formation of (stable) liposomes. It turned out that the final assembly is highly sensitive to the starting configuration of the simulations and depends strongly on the water concentration. In particular, we have observed the formation of a hemifused vesicle, i.e. a vesicle obtained by fusion of lipid vesicles, where the aqueous interiors of the two initial vesicles are left unmixed.