E. histolytica is the etiological agent of intestinal amebiasis and liver abscesses, which still poses public health threat globally. Metronidazole is the drug of choice against amebiasis. However, metronidazole-resistant amoebic clinical isolates and strains have been reported recently, challenging the efforts for amebiasis eradication. In search of alternative treatments, E. histolytica transcriptomes have shown the association of genes involved in RNA metabolism with the virulence of the parasite. Among the upregulated genes in amoebic liver abscesses are the splicing factors EhU2AF2 and a paralog of EhSF3B1. For this reason and because EhU2AF2 contains unusual KH-QUA2 (84KQ) motifs in its lengthened C-terminus domain, here we investigated how the role of EhU2AF2 in pre-mRNA processing impacts the virulence of the parasite. We found that 84KQ is involved in splicing inhibition/intron retention of several virulence and non-virulence-related genes. The 84KQ domain interacts with the same domain of the constitutive splicing factor SF1 (SF1KQ), both in solution and when SF1KQ is bound to branchpoint signal RNA probes. The 84KQ-SF1KQ interaction prevents splicing complex E to A transition, thus inhibiting splicing. Surprisingly, the deletion of the 84KQ domain in EhU2AF2 amoeba transformants increased splicing and enhanced the in vitro and in vivo virulence phenotypes. We conclude that the interaction of the 84KQ and SF1KQ domains, probably involving additional factors, tunes down Entamoeba virulence by favoring intron retention.

The Transformer2 (Tra2) proteins in humans are homologues of the Drosophila Tra2 protein. One of the two RNA-binding paralogs, Tra2β, has been very well-studied over the past decade, but not much is known about Tra2α. It was very recently shown that the two proteins demonstrate the phenomenon of paralog compensation. Here, we provide a structural basis for this genetic backup circuit, using molecular modelling and dynamics studies. We show that the two proteins display similar binding specificities, but differential affinities to a short GAA-rich RNA stretch. Starting from the 6-nucleotide RNA in the solution structure, close to 4000 virtual mutations were modelled on RNA and the domain-RNA interactions were studied after energy minimisation to convergence. Separately, another known 13-nucleotide stretch was docked and the domain-RNA interactions were observed through a 100-ns dynamics trajectory. We have also demonstrated the \'compensatory\' mechanism at the level of domains in one of the domain repeat-containing RNA-binding proteins.

Mitochondrial pre-mRNAs in trypanosomatids undergo RNA editing to be converted into translatable mRNAs. The reaction is characterized by the insertion and deletion of uridine residues and is catalyzed by a macromolecular protein complex called the editosome. Despite intensive research, structural information for the majority of editosome proteins is still missing and no high resolution structure for the editosome exists. Here we present a comprehensive structural bioinformatics analysis of all proteins of the Trypanosoma brucei editosome. We specifically focus on the interplay between intrinsic order and disorder. According to computational predictions, editosome proteins involved in the basal reaction steps of the processing cycle are mostly ordered. By contrast, thirty percent of the amino acid content of the editosome is intrinsically disordered, which includes most prominently proteins with OB-fold domains. Based on the data we suggest a functional model, in which the structurally disordered domains of the complex are correlated with the RNA binding and RNA unfolding activity of the T. brucei editosome.

Motivated by the pressing need to characterize protein-DNA and protein-RNA interactions on large scale, we review a comprehensive set of 30 computational methods for high-throughput prediction of RNA- or DNA-binding residues from protein sequences. We summarize these predictors from several significant perspectives including their design, outputs and availability. We perform empirical assessment of methods that offer web servers using a new benchmark data set characterized by a more complete annotation that includes binding residues transferred from the same or similar proteins. We show that predictors of DNA-binding (RNA-binding) residues offer relatively strong predictive performance but they are unable to properly separate DNA- from RNA-binding residues. We design and empirically assess several types of consensuses and demonstrate that machine learning (ML)-based approaches provide improved predictive performance when compared with the individual predictors of DNA-binding residues or RNA-binding residues. We also formulate and execute first-of-its-kind study that targets combined prediction of DNA- and RNA-binding residues. We design and test three types of consensuses for this prediction and conclude that this novel approach that relies on ML design provides better predictive quality than individual predictors when tested on prediction of DNA- and RNA-binding residues individually. It also substantially improves discrimination between these two types of nucleic acids. Our results suggest that development of a new generation of predictors would benefit from using training data sets that combine both RNA- and DNA-binding proteins, designing new inputs that specifically target either DNA- or RNA-binding residues and pursuing combined prediction of DNA- and RNA-binding residues.

Protein-RNA interactions play fundamental roles in many biological processes, such as regulation of gene expression, RNA splicing, and protein synthesis. The understanding of these processes improves as new structures of protein-RNA complexes are solved and the molecular details of interactions analyzed. However, experimental determination of protein-RNA complex structures by high-resolution methods is tedious and difficult. Therefore, studies on protein-RNA recognition and complex formation present major technical challenges for macromolecular structural biology. Alternatively, protein-RNA interactions can be predicted by computational methods. Although less accurate than experimental measurements, theoretical models of macromolecular structures can be sufficiently accurate to prompt functional hypotheses and guide e.g. identification of important amino acid or nucleotide residues. In this article we present an overview of strategies and methods for computational modeling of protein-RNA complexes, including software developed in our laboratory, and illustrate it with practical examples of structural predictions.