Reconstruction of Primordial P-loop NTPase Precursors

Abstract

In collaboration with: Agnes Toth-Petroczy, Alexander Goncearenco and Igor Berezovsky (inference of the ancestral P-loop sequence) and Lin Yu-Ru and David Baker (computational protein design).Short functional peptides are likely to have served as crucial intermediates between a primordial RNA world and the extant protein world. Our working hypothesis is that relics of these ancestral peptides still exist in the form of key motifs in active sites of present-day proteins. One such motif, probably the most obvious one, is the P-loop (Walker A motif) that typically binds the transferred phosphate moiety of ATP. The P-loop's extended motif comprises a β-strand and an α-helix connected by the P-loop – a glycine-rich phosphate-binding loop. The P-loop containing nucleoside triphosphate hydrolase fold includes this motif, and is one of the most ancient folds and by far the most abundant (10-18% of all ORFs). In the present work, the sequence of a primordial β-Ploop-α motif was inferred. We further showed that this β-Ploop-α motif (∼25 amino acids) could have served as a building block for the creation of ancestral globular domains by means of duplication and fusion. Computational designed proteins were generated consisting of 2 repetitions of the ancestral motif grafted into an ideal β/α- repeat protein. Our results indicate that these simple, repetitive proteins avidly bind ssDNA and RNA (possibly a crucial step in the transition from the RNA world to a protein world). Further, a β-α-β fragment of this protein (< 40 amino acids) was found to self-assemble and thereby confer ssDNA binding. Overall, our results show that P-loop NTPases could have emerged from a relatively short P-loop containing peptide and that self-assembly played a key role in endowing biochemical function despite limited size and complexity.