nai

Project Progress

We have been characterizing an artificial enzyme that joins two RNA fragments. This enzyme was previously generated de novo from a large library of proteins based on a zinc finger protein scaffold with randomized loops. Structural investigations by Nuclear Magnetic Resonance (NMR) have generated sufficient data to model an ensemble of its structures (Fig. 1a). The protein has a highly compact core composed of two regions connected by a flexible loop and very flexible ends. NMR data suggest that two zinc-binding sites are located in the compact core. The protein structure is unrelated to the original scaffold and has no homology to other known structures. This indicates that (1) protein evolution through a modest number of mutations can yield both novel functions and novel structures, and (2) structures and functions not found in contemporary organisms are possible. We have generated over 20 mutant proteins (Fig. 1b) to identify the zinc-binding sites and the amino acids involved in catalysis.

We investigated two models of protobiological protein ion channels, antiamoebin and trichotoxin. They facilitate ion transport across cell walls needed for establishing osmotic equilibria, energy transduction and stability of macromolecules in protocells. Peptides forming the channels consist of only 14-16 residues and contain non-standard amino acids that were common on the early earth. By comparing calculated and measured conductance we determined the channel structures. Despite their simplicity the channels are quite efficient in increasing rates of ion transport through membranes, a property that is essential for preventing the buildup of osmotic stresses. The specific identity of many amino acids in these channels is not important, as long as their polarity is conserved, a convenient property at the origins of life. Our results support a view that the first ion channels were similar to the channels studied here.