Protein adaptability involved in self-assembled icosahedral capsids

Interactions between viral coat proteins determine the size and shape of the virus capsid. Therefore, the plain closed shell of a spherical virus provides an excellent resource for studying and testing protein-protein interactions so as to reveal the properties that lead to symmetrical capsids or other shapes. Icosahedral architecture similar to that found in viruses has also been adopted by some giant enzymes. In both, cases the biological activity and the shape of the particle are parameters that help to identify the accuracy of the small units forming the assembled giant molecule. Vice versa, the small units and their linkers may be engineered to guide the assembly into useful particles. In this thesis, I focus on the protein framework in icosahedral virus-like particles of the human polyomavirus BK and the enzyme lumazine synthase from Aquifex aeolicus. My aim was to explore how a single type of building unit, in this case a pentamer, can assemble into particles with different size and surface morphology. This work includes the structure determination and comparison of two different virus-like particles (26.4 and 50.0 nm in diameter) of the human polyomavirus BK protein VP1. They both have icosahedral symmetry and their pentamic capsomers establish a T=1 and T=7d surface lattice, respectively. Likewise, the structure of a Lumazine synthase particle from Aquifex aeolicus was solved and compared to a previously solved structure of the native enzyme. The enzyme particle has a sequence insertion of four amino acids. This influences the shape of the assembly and the particles formed are approximately 14 nm larger in diameter than the native structure. The two systems, the viruslike particles and the enzyme particles, differ in their strategy for assembly: the BKV VP1 protein forms pentamers which are linked together by long (approximately 60 amino acids) flexible arms, while the enzyme protein is globular without protruding domains and uses relatively flat contact surfaces within pentamers and surrounding subunits. The BK virus-like particles primarily use the two loops, before and after the Cterminal helix, to adjust the position of the flexible arm so that the intercapsomer interactions can be kept as similar as possible even at different symmetrical environments and particle curvatures. In the case of the enzyme, the flexibility is not the result of different folding of a contact arm. Instead, alternative areas of the protein surface are used to form the interaction. In all the particles studied, the hydrophobic effect seems to be a main stabilizing force. However, assembly of pentamers (or connecting subunits) also utilizes more specific interactions provided by patterns of complementary charges. Assembly of the "correct" quaternary structure in both systems may be driven by prosthetic groups. In the enzyme case flexibility is required for catalytic reasons - the active site is located at the subunit interface within the pentamer. Here the prosthetic group would control the needed motion for binding substrates and releasing products. In the case of the BK virus-like particles, the main flexibility is found within the C-terminal arm, forming the main interpentameric contacts. Here we find the assembly-controlling calcium ion. The electrostatic interactions and the prosthetic groups are placed in the structure for two functional purposes: to guide the assembly pathway and to maintain the local flexibility needed for assigned functions