OST-HTH: a novel predicted RNA-binding domain

Abstract

Background: The mechanism by which the arthropod Oskar and vertebrate TDRD5/TDRD7 proteins nucleate or organize structurally related ribonucleoprotein (RNP) complexes, the polar granule and nuage, is poorly understood. Using sequence profile searches we identify a novel domain in these proteins that is widely conserved across eukaryotes and bacteria.

Results: Using contextual information from domain architectures, sequence-structure superpositions and available functional information we predict that this domain is likely to adopt the winged helix-turn-helix fold and bind RNA with a potential specificity for dsRNA. We show that in eukaryotes this domain is often combined in the same polypeptide with protein-protein- or lipid- interaction domains that might play a role in anchoring these proteins to specific cytoskeletal structures.

Conclusions: Thus, proteins with this domain might have a key role in the recognition and localization of dsRNA, including miRNAs, rasiRNAs and piRNAs hybridized to their targets. In other cases, this domain is fused to ubiquitin-binding, E3 ligase and ubiquitin-like domains indicating a previously under-appreciated role for ubiquitination in regulating the assembly and stability of nuage-like RNP complexes. Both bacteria and eukaryotes encode a conserved family of proteins that combines this predicted RNA-binding domain with a previously uncharacterized domain (DUF88). We present evidence that it is an RNAse belonging to the superfamily that includes the 5'->3' nucleases, PIN and NYN domains and might be recruited to degrade certain RNAs.

Figure 1

A. Multiple alignment of selected representatives of the OST-HTH superfamily. The consensus shown below was computed on a complete alignment of the superfamily which included 256 non-redundant versions of domain and was used to color columns of the alignment. The consensus positions are labeled thus: h- hydrophobic, l- aliphatic, s-small and p-polar. The proteins are labeled with the protein name followed by the organism abbreviation and Genbank Gi number delimited by underscores. The organism abbreviations are: Anid: Aspergillus nidulans; Atha: Arabidopsis thaliana; Bjap: Bradyrhizobium japonicum; Bthe: Bacteroides thetaiotaomicron; Ccur: Cryptobacterium curtum; Cele: Caenorhabditis elegans; Chut: Cytophaga hutchinsonii; Cjej: Campylobacter jejuni; Cpar: Cryptosporidium parvum; Daro: Dechloromonas aromatica; Dmel: Drosophila melanogaster; Hsap: Homo sapiens; Mbur: Methanococcoides burtonii; Neur: nitrosomonas europaea; Nvec: Nematostella vectensis; Oter: Opitutus terrae; Pfal: Plasmodium falciparum; Pmar: Perkinsus marinus; Psyr: Pseudomonas syringae; Ptet: Paramecium tetraurelia; Rpal: Rhodopseudomonas palustris; Rsph: Rhodobacter sphaeroides; Ssp: Synechocystis sp.; Tann: Theileria annulata; Tpal: Treponema pallidum; Tpar: Theileria parva; Tthe: Tetrahymena thermophila; Xaxo: Xanthomonas axonopodis; Osat: Oryza sativa; Vvin: Vitis vinifera; Jant: Jonquetella anthropi; Pinf: Phytophthora infestans. B. The average NMR structure of the OST-HTH domain from the protein NE0665 (2 kpm) is rendered as a cartoon with the helices and strands labeled as per the secondary structure progression. The regions inferred to potentially contact dsRNA based on the superposition to the archaeal CDC6-DNA co-crystal structure are indicated. The region of the OST-HTH that is circled is the unique insert that distinguishes it from other wHTH domains.

Figure 2

A. Domain architectures of OST-HTH proteins. The domains are shown approximately scale but the intervening non-globular regions are not shown. The phyletic pattern typical of each architecture is shown below the illustrated representative. The species abbreviations and protein labels are as in figure 1. We observed that the DUF2384 domain fused to certain bacterial LK-nucleases is another HTH domain distinct from the OST-HTH described in this article. A key is provided for the domains whose names have been abbreviated in a non-standard fashion. B. Architectural network for the OSK-HTH domain. The network is centered on the OST-HTH with the other domains grouped as per their function. The arrows indicating the directionality of the connections have been omitted for easier examination.