It is generally accepted that the evolution of protein families involves duplication of genetic material. Symmetrical protein architectures containing repeated domains are strong evidence of duplication and fusion events, and beta-propeller proteins are a prime example.
In order to investigate the evolutionary processes that led to a wide diversity of such structures, we have computationally designed a 6-fold symmetrical beta-propeller protein, and experimentally validated the model using crystallography. Given its circular shape, built from wedge shaped domains, the protein was named "pizza6". Smaller proteins with only 2 or 3 domains were also purified (pizza2 and pizza3). These assemble into a trimer and dimer respectively, recreating the 6-fold symmetrical architecture. Higher order repeat proteins (pizza4, 5, 7, 8, 9 and 10) were found to reassemble according to the lowest common multiplier of the number of repeats and 6. In the case of pizza7 protein, as well as the expected higher order complex, a smaller amount was found of the monomer carrying 7 repeats.
Our results indicate that gene duplication and fusion can create proteins of similar shapes or new, larger complexes. Our fully symmetric designer proteins prove to be very stable and molecular dynamics / normal mode simulation approach was utilized to analyze the origin of stability due to symmetry.
Given their high stability and ability to self-associate in set patterns, our approach to computationally design self-assembling proteins is useful to create novel protein building blocks for bio-nanotechnology.