Repeatability of protein structural evolution following convergent gene fusions

Abstract

Convergent evolution of proteins provides insights into repeatability of genetic adaptation. While local convergence of proteins at residue or domain level has been characterized, global structural convergence by inter-domain/molecular interactions remains largely unknown. Here we present structural convergent evolution on fusion enzymes of aldehyde dehydrogenases (ALDHs) and alcohol dehydrogenases (ADHs). We discovered BdhE (bifunctional dehydrogenase E), an enzyme clade that emerged independently from the previously known AdhE family through distinct gene fusion events. AdhE and BdhE showed shared enzymatic activities and non-overlapping phylogenetic distribution, suggesting common functions in different species. Cryo-electron microscopy revealed BdhEs form donut-like homotetramers, contrasting AdhEs helical homopolymers. Intriguingly, despite distinct quaternary structures and >70% unshared amino acids, both enzymes form resembled dimeric structure units by ALDH-ADH interactions via convergently elongated loop structures. These findings suggest convergent gene fusions recurrently led to substrate channeling evolution to enhance two-step reaction efficiency. Our study unveils structural convergence at inter-domain/molecular level, expanding our knowledges on patterns behind molecular evolution exploring protein structural universe.