nai

Project Progress

We refined the structure of the artificial RNA ligase enzyme, which was isolated from a randomized non-catalytic library of polypeptides. We showed that this enzyme exhibits a fold and catalyses a reaction not existing in nature. Even though the protein is catalytically active, its structured part is quite small and the rest of the protein consists of a flexible loop. This architecture is at variance with the basic paradigm developed for modern enzymes, which requires that functional proteins are compact. This result demonstrates that primordial enzymes, which did not undergo fine-tuning through billions of years of evolution, were different from contemporary enzymes. To probe the proposed zinc-binding sites in our enzyme (Fig. 1a), we replaced individual amino acids that could potentially coordinate zinc and tested the mutants for activity and secondary structure changes. Data on 30 mutants and titration of the enzyme with the substrate combined with the measurements of NMR loss of signal (Fig. 1b) helped to identify the potential substrate-binding site.

We are also investigating the possibility that the same catalytic ligase activity could emerge from entirely random sequences. We have optimized the ligase selection procedure for a protein library containing 80 random amino acids and we are poised to begin the selection.

To explore how ancestral proteins evolved to acquire new functions we redesigned computationally an adenosine triphosphate (ATP)-binding protein that had evolved from random sequences of amino acids, to bind guanosine triphosphate (GTP). The redesign involved not only changing the orientation of the loop that hydrogen bonds to the nucleobase but also modifying the structures that stabilize the binding pocket and prevent water from accessing the binding site (Fig. 2)

The activity of the redesigned protein will be studied experimentally. We will also explore whether protein evolution from binding ATP to binding GTP can proceed without loss of function.