Human DND1-RRM2 forms a non-canonical domain-swapped dimer.

Abstract

RNA Recognition Motifs (RRMs) being the most abundant RNA binding domain in eukaryotes, is a major player in cellular regulation. Human DND1 (dead end protein homolog1) also known as microRNA-mediated repression inhibitor 1 is an RNA binding protein containing two RRMs in tandem. It is essential for migration and viability of primordial germ cells during development, its complete loss causes lethality, and mutations in the gene are associated with perturbed cellular regulation and eventually cancer. Using x-ray crystallography we have determined 2.3 Å structure of the human DND1 RRM2 domain. The structure revealed an interesting non-canonical RRM fold. This is maintained by the formation of a 3D domain-swapped dimer between β1 and β4 strands across protomers. We have delineated the structural basis of the stable domain-swapped dimer formation using the residue level dynamics of protein explored by NMR spectroscopy and MD simulations. Our results indicate that monomer to dimer switch is governed by several determinants such as hinge loop residues, its length, hydrophobicity, concentration of protein, disulphide bond, and so forth. Several variations in the canonical βαββαβ topology have been observed so far. However, this is the first report of domain swapping in the RRM domain, which may increase the binding surface area and allow this protein to exhibit multifaceted role during post-transcriptional regulation. Our structural and dynamics studies substantiate major determinants and molecular basis for domain-swapped dimerization observed in the RRM domain that could allow varied roles for its RRMs.