Computational and Experimental Evidence for the Evolution of a (βα)8-Barrel Protein from an Ancestral Quarter-Barrel Stabilised by Disulfide Bonds

Markus Richter, Manal Bosnali, Linn Carstensen, Tobias Seitz, Helmut Durchschlag, Samuel Blanquart, Rainer Merkl, Reinhard Sterner

The evolution of the prototypical (beta alpha)(8)-barrel protein imidazole glycerol phosphate synthase (HisF) was studied by complementary computational and experimental approaches. The 4-fold symmetry of HisF suggested that its constituting (beta alpha)(2) quarter-barrels have a common evolutionary origin. This conclusion was supported by the computational reconstruction of the HisF sequence of the last common ancestor, which showed that its quarter-barrels were more similar to each other than are those of extant HisF proteins. A comprehensive sequence analysis identified HisF-N1 [corresponding to (beta alpha)(1-2)] as the slowest evolving quarter-barrel. This finding indicated that it is the closest relative of the common (beta alpha)(2) predecessor, which must have been a stable and presumably tetrameric protein. In accordance with this prediction, a recombinantly produced HisF-N1 protein was properly folded and formed a tetramer being stabilised by disulfide bonds. The introduction of a disulfide bond in HisF-C1 [corresponding to (beta alpha)(5-6)] also resulted in the formation of a stable tetramer. The fusion of two identical HisF-N1 quarter-barrels yielded the stable dimeric half-barrel HisF-N1N1. Our findings suggest a two-step evolutionary pathway in which a HisF-N1-like predecessor was duplicated and fused twice to yield HisF. Most likely, the (beta alpha)(2) quarter-barrel and (beta alpha)(4) half-barrel intermediates on this pathway were stabilised by disulfide bonds that became dispensable upon consolidation of the (beta alpha)(8)-barrel.