Classifying RNA-binding proteins based on electrostatic properties

Protein structure can provide new insight into the biological function of a protein and can enable the design of better experiments to learn its biological roles. Moreover, deciphering the interactions of a protein with other molecules can contribute to the understanding of the protein's function within cellular processes. In this study, we apply a machine learning approach for classifying RNA-binding proteins based on their three-dimensional structures. The method is based on characterizing unique properties of electrostatic patches on the protein surface. Using an ensemble of general protein features and specific properties extracted from the electrostatic patches, we have trained a support vector machine (SVM) to distinguish RNA-binding proteins from other positively charged proteins that do not bind nucleic acids. Specifically, the method was applied on proteins possessing the RNA recognition motif (RRM) and successfully classified RNA-binding proteins from RRM domains involved in protein-protein interactions. Overall the method achieves 88% accuracy in classifying RNA-binding proteins, yet it cannot distinguish RNA from DNA binding proteins. Nevertheless, by applying a multiclass SVM approach we were able to classify the RNA-binding proteins based on their RNA targets, specifically, whether they bind a ribosomal RNA (rRNA), a transfer RNA (tRNA), or messenger RNA (mRNA). Finally, we present here an innovative approach that does not rely on sequence or structural homology and could be applied to identify novel RNA-binding proteins with unique folds and/or binding motifs.     

Components of the DNA methylation system of chromatin control are RNA-binding proteins

The view that autosomal gene expression is controlled exclusively by protein trans-acting factors has been challenged recently by the identification of RNA molecules that regulate chromatin. In the majority of cases where RNA molecules are implicated in DNA control, the molecular mechanisms are unknown, in large part because the RNA.protein complexes are uncharacterized. Here, we identify a novel set of RNA-binding proteins that are well known for their function in chromatin regulation. The RNA-interacting proteins are components of the mammalian DNA methylation system. Genomic methylation controls chromatin in the context of transposon silencing, imprinting, and X chromosome dosage compensation. DNA methyltransferases (DNMTs) catalyze methylation of cytosines in CGs. The methyl-CGs are recognized by methyl-DNA-binding domain (MBD) proteins, which recruit histone deacetylases and chromatin remodeling proteins to effect silencing. We show that a subset of the DNMTs and MBD proteins can form RNA.protein complexes. We characterize the MBD protein RNA-binding activity and show that it is distinct from the methyl-CG-binding domain and mediates a high affinity interaction with RNA. The RNA and methyl-CG binding properties of the MBD proteins are mutually exclusive. We speculate that DNMTs and MBD proteins allow RNA molecules to participate in DNA methylation-mediated chromatin control.     

Interaction of rat liver dexamethasone-receptor complexes with DNA]

Rat liver glucocorticoid-receptor complexes (GRC) acquire the ability to bind to DNA in a high affinity manner after activation by heating or precipitation with (NH)2SO4. DNA is practically non-saturable by GRC in low salt buffers as well as in 0.15 M NaCl-containing buffer, although in the latter case the binding decreases approximately 3--5 times. GRC bind to homo- and heterologous prokaryotic DNA in a similar way; in both cases an addition of KCl (up to 0.15 M) to the medium is followed by the same decrease of the binding. This data suggest that the association of GRC with DNA observed in vitro is not accompanied by "recognition" of any certain DNA site. Besides DNA, activated GRC can associate with other polymers, charged positively (DEAE-cellulose) or negatively (RNA, polyvinylsulfate). GRC interact very weakly with neutral compounds of the cellulose type but are strongly adsorbed on hydroxyapatite. Hence the activated GRC can be considered as an amphoteric protein. Salt solutions provoke dissociation of the GRC-DNA triple complexes: a complete dissociation is observed in the presence of 0,4 M NaCl or 0,4 M sodium phosphate buffer (pH 6,9). Sodium phosphate buffer also elutes GRC from other sorbents such as DEAE-cellulose or hydroxyapatite. No significant dissociation of the GRC-DNA complexes is observed at sucrose concentration up to 2 M. The data obtained are indicative of an essential role of electrostatic forces for the interaction of GRC with DNA. The non-ionic detergent Triton X-100 at a concentration as low as 0,05% completely destroys the GRC-DNA triple complexes. The models explicating the selectivity of the genome activation by GRC without their "recognition" of any specific DNA sequences are proposed.     

Intrinsically disordered electronegative clusters improve stability and binding specificity of RNA-binding proteins

RNA-binding proteins play crucial roles in various cellular functions and contain abundant disordered protein regions. The disordered regions in RNA-binding proteins are rich in repetitive sequences, such as poly-K/R, poly-N/Q, poly-A, and poly-G residues. Our bioinformatic analysis identified a largely neglected repetitive sequence family we define as electronegative clusters (ENCs) that contain acidic residues and/or phosphorylation sites. The abundance and length of ENCs exceed other known repetitive sequences. Despite their abundance, the functions of ENCs in RNA-binding proteins are still elusive. To investigate the impacts of ENCs on protein stability, RNA-binding affinity, and specificity, we selected one RNA-binding protein, the ribosomal biogenesis factor 15 (Nop15), as a model. We found that the Nop15 ENC increases protein stability and inhibits nonspecific RNA binding, but minimally interferes with specific RNA binding. To investigate the effect of ENCs on sequence specificity of RNA binding, we grafted an ENC to another RNA-binding protein, Ser/Arg-rich splicing factor 3. Using RNA Bind-n-Seq, we found that the engineered ENC inhibits disparate RNA motifs differently, instead of weakening all RNA motifs to the same extent. The motif site directly involved in electrostatic interaction is more susceptible to the ENC inhibition. These results suggest that one of functions of ENCs is to regulate RNA binding via electrostatic interaction. This is consistent with our finding that ENCs are also overrepresented in DNA-binding proteins, whereas underrepresented in halophiles, in which nonspecific nucleic acid binding is inhibited by high concentrations of salts.     

Structure-based prediction of RNA-binding domains and RNA-binding sites and application to structural genomics targets

Mechanistic understanding of many key cellular processes often involves identification of RNA binding proteins (RBPs) and RNA binding sites in two separate steps. Here, they are predicted simultaneously by structural alignment to known protein-RNA complex structures followed by binding assessment with a DFIRE-based statistical energy function. This method achieves 98% accuracy and 91% precision for predicting RBPs and 93% accuracy and 78% precision for predicting RNA-binding amino-acid residues for a large benchmark of 212 RNA binding and 6761 non-RNA binding domains (leave-one-out cross-validation). Additional tests revealed that the method makes no false positive prediction from 311 DNA binding domains but correctly detects six domains binding with both DNA and RNA. In addition, it correctly identified 31 of 75 unbound RNA-binding domains with 92% accuracy and 65% precision for predicted binding residues and achieved 86% success rate in its application to SCOP RNA binding domain superfamily (Structural Classification Of Proteins). It further predicts 25 targets as RBPs in 2076 structural genomics targets: 20 of 25 predicted ones (80%) are putatively RNA binding. The superior performance over existing methods indicates the importance of dividing structures into domains, using a Z-score to measure relative structural similarity, and a statistical energy function to measure protein-RNA binding affinity.     

An RNA-binding domain in the thyroid hormone receptor enhances transcriptional activation

Thyroid hormone plays important roles in development, differentiation, and metabolic homeostasis by binding to nuclear thyroid hormone receptors, which regulate target gene expression by interacting with DNA response elements and coregulatory proteins. We show that thyroid hormone receptors also are single-stranded RNA binding proteins and that this binding is functionally significant. By using a series of deletion mutants, a novel RNA-binding domain was localized to a 41-amino acid segment of thyroid hormone receptor alpha1 between the second zinc finger and the ligand-binding domain. This RNA-binding domain was necessary and sufficient for thyroid hormone receptor binding to the steroid receptor RNA activator (SRA). Although SRA does not bind directly to steroid receptors, it has been identified as a steroid receptor coactivator, and was thought not to be a coactivator for thyroid hormone receptors. However, transfection studies revealed that SRA enhances thyroid hormone induction of appropriate reporter genes and that the thyroid hormone receptor RNA-binding domain is important for this enhancement. We conclude that thyroid hormone receptors bind RNA through a novel domain and that the interaction of this domain with SRA, and perhaps other RNAs, enhances thyroid hormone receptor function.     

Partner-specific prediction of RNA-binding residues in proteins: A critical assessment

RNA-protein interactions play essential roles in regulating gene expression. While some RNA-protein interactions are “specific”, that is, the RNA-binding proteins preferentially bind to particular RNA sequence or structural motifs, others are “non-RNA specific.” Deciphering the protein-RNA recognition code is essential for comprehending the functional implications of these interactions and for developing new therapies for many diseases. Because of the high cost of experimental determination of protein-RNA interfaces, there is a need for computational methods to identify RNA-binding residues in proteins. While most of the existing computational methods for predicting RNA-binding residues in RNA-binding proteins are oblivious to the characteristics of the partner RNA, there is growing interest in methods for partner-specific prediction of RNA binding sites in proteins. In this work, we assess the performance of two recently published partner-specific protein-RNA interface prediction tools, PS-PRIP, and PRIdictor, along with our own new tools. Specifically, we introduce a novel metric, RNA-specificity metric (RSM), for quantifying the RNA-specificity of the RNA binding residues predicted by such tools. Our results show that the RNA-binding residues predicted by previously published methods are oblivious to the characteristics of the putative RNA binding partner. Moreover, when evaluated using partner-agnostic metrics, RNA partner-specific methods are outperformed by the state-of-the-art partner-agnostic methods. We conjecture that either (a) the protein-RNA complexes in PDB are not representative of the protein-RNA interactions in nature, or (b) the current methods for partner-specific prediction of RNA-binding residues in proteins fail to account for the differences in RNA partner-specific versus partner-agnostic protein-RNA interactions, or both.     

RNA-binding proteins and RNA metabolism: a new scenario in the pathogenesis of Amyotrophic lateral sclerosis

Several RNA-processing genes have been implicated in the pathogenesis of Amyotrophic lateral sclerosis (ALS). In particular, causative mutations in the genes encoding for two DNA/RNA binding proteins, TAR DNA binding protein-43 (TDP-43) and fused in sarcoma/translocated in liposarcoma (FUS/TLS), were recently identified in ALS patients. These genetic findings and the presence of abnormal aggregates of these two RNA-binding proteins in ALS affected tissues suggest that molecular mechanisms regulating RNA metabolism are implicated in ALS pathogenesis through common pathways. In this review similarities and differences between TDP-43 and FUS/TLS proteins and their activities in physiological and pathological conditions will be discussed.     

Design of RNA-binding proteins and ligands

The rapidly expanding database of RNA structures and protein complexes is beginning to lead to the successful design of specific RNA-binding molecules. Recent combinatorial and structure-based approaches have utilized known nucleic-acid-binding scaffolds from both proteins and small molecules to display a relatively small set of functional groups often used in protein--RNA recognition. Several studies have shown that the tethering of multiple binding modules can enhance RNA-binding affinity and specificity, a strategy also commonly used in DNA recognition.     

RNA-binding activity of hepatitis delta antigen involves two arginine-rich motifs and is required for hepatitis delta virus RNA replication.

Hepatitis delta antigen (HDAg) is an RNA-binding protein with binding specificity for hepatitis delta virus (HDV) RNA (J. H. Lin, M. F. Chang, S. C. Baker, S. Govindarajan, and M. M. C. Lai, J. Virol. 64:4051-4058, 1990). By amino acid sequence homology search, we have identified within its RNA-binding domain two stretches of an arginine-rich motif (ARM), which is present in many prokaryotic and eukaryotic RNA-binding proteins. The first one is KERQDHRRRKA and the second is EDEKRERRIAG, and they are separated by 29 amino acids. Deletion of either one of these ARM sequences resulted in the total loss of the in vitro RNA-binding activity of HDAg. Thus, HDAg is different from other RNA-binding proteins in that it requires two ARM-like sequences for its RNA-binding activity. Replacement of the spacer sequence between the two ARMs with a shorter stretch of sequence also reduced RNA binding in vitro. Furthermore, site-specific mutations of the basic amino acid residues in both ARMs resulted in the total loss or reduction of RNA-binding activity. The biological significance of the RNA-binding activity was studied by examining the trans-activating activity of the RNA-binding mutants. The plasmids expressing HDAgs with various mutations in the RNA-binding motifs were cotransfected with a replication-defective HDV dimer cDNA construct into COS cells. It was found that all the HDAg mutants which had lost the in vitro RNA-binding activity also lost the ability to complement the defect of HDV RNA replication. We conclude that the trans-activating function of HDAg requires its binding to HDV RNA.     

Identification and characterization of proteins binding A + U-rich elements

A + U-Rich elements (AREs) have been extensively investigated as cis-acting determinants of rapid mRNA turnover. Recently, a number of RNA-binding proteins interacting with AREs have been described. This article presents strategies and techniques used by our laboratory to identify and characterize a family of ARE-binding proteins collectively termed AUF1. However, these techniques may be applied to the study of any protein displaying sequence-specific RNA binding activity. The techniques described here include the purification of native AUF1 from cultured cells as well as the preparation of recombinant AUF1 proteins using a bacterial expression system. Analyses of RNA-protein interactions are also described, including the use of gel mobility shift assays with synthetic RNA probes to monitor specific RNA binding activity in cell extracts or with recombinant proteins. Variations of this technique are also described to evaluate the RNA binding affinity of recombinant proteins and the use of specific RNA competitors to assess RNA determinants of protein binding specificity. Other techniques presented include the identification of specific proteins in RNA:protein complexes using antibody supershifts and the estimation of molecular weights of RNA-binding proteins by UV crosslinking. Results of individual experiments are presented as examples of some techniques. Throughout the article, suggestions are included to avoid commonly encountered problems and to assist in the optimization of these techniques for the study of other RNA-binding proteins.     

Identification of Sam68 arginine glycine-rich sequences capable of conferring nonspecific RNA binding to the GSG domain

Sam68 is an RNA-binding protein that contains a heterogeneous nuclear ribonucleoprotein K homology domain embedded in a larger RNA binding domain called the GSG (GRP33, Sam68, GLD-1) domain. This family of proteins is often referred to as the STAR (signal transduction and activators of RNA metabolism) proteins. It is not known whether Sam68 is a general nonspecific RNA-binding protein or whether it recognizes specific response elements in mRNAs with high affinity. Sam68 has been shown to bind homopolymeric RNA and a synthetic RNA sequence called G8-5 that has a core UAAA motif. Here we performed a structure function analysis of Sam68 and identified two arginine glycine (RG)-rich regions that confer nonspecific RNA binding to the Sam68 GSG domain. In addition, by using chimeric proteins between Sam68 and QKI-7, we demonstrated that one of the Sam68 RG-rich sequences of 26 amino acids was sufficient to confer homopolymeric RNA binding to the GSG domain of QKI-7, another STAR protein. Furthermore, that minimal sequence can also give QKI-7 the ability (as Sam68) to functionally substitute for HIV-1 REV to facilitate the nuclear export of RNAs. Our studies suggest that neighboring RG-rich sequences may impose nonspecific RNA binding to GSG domains. Because the Sam68 RNA binding activity is negatively regulated by tyrosine phosphorylation, our data lead us to propose that Sam68 might be a specific RNA-binding protein when tyrosine phosphorylated.     

Intrinsic Disorder in Human RNA-Binding Proteins

Although RNA-binding proteins (RBPs) are known to be enriched in intrinsic disorder, no previous analysis focused on RBPs interacting with specific RNA types. We fill this gap with a comprehensive analysis of the putative disorder in RBPs binding to six common RNA types: messenger RNA (mRNA), transfer RNA (tRNA), small nuclear RNA (snRNA), non-coding RNA (ncRNA), ribosomal RNA (rRNA), and internal ribosome RNA (irRNA). We also analyze the amount of putative intrinsic disorder in the RNA-binding domains (RBDs) and non-RNA-binding-domain regions (non-RBD regions). Consistent with previous studies, we show that in comparison with human proteome, RBPs are significantly enriched in disorder. However, closer examination finds significant enrichment in predicted disorder for the mRNA-, rRNA- and snRNA-binding proteins, while the proteins that interact with ncRNA and irRNA are not enriched in disorder, and the tRNA-binding proteins are significantly depleted in disorder. We show a consistent pattern of significant disorder enrichment in the non-RBD regions coupled with low levels of disorder in RBDs, which suggests that disorder is relatively rarely utilized in the RNA-binding regions. Our analysis of the non-RBD regions suggests that disorder harbors posttranslational modification sites and is involved in the putative interactions with DNA. Importantly, we utilize experimental data from DisProt and independent data from Pfam to validate the above observations that rely on the disorder predictions. This study provides new insights into the distribution of disorder across proteins that bind different RNA types and the functional role of disorder in the regions where it is enriched.     

Functional assignment to maize group 1 LEA protein EMB564 within the cell nucleus using computational analysis

Maize protein EMB564 is a member of group 1 LEA (late embryogenesis abundant) proteins. Currently, the molecular functions of group 1 LEA proteins remain largely unclear. We here report on the functional assignment to EMB564 by computational analysis. EMB564 is predicted as nuclear localization by five different predictors including CELLO, Plant-mPLoc, WoLF PSORT, Predotar and TargetP. EMB564 is found to be remote homologous with DNA/RNA helicases and single-stranded DNA-binding proteins, and their sequences contains similar DNA/RNA binding sites. Furthermore, the three-dimensional (3D) model of EMB564 structurally resembles a variety of nuclear and DNA/RNA-binding proteins, especially those involving in the regulation of cell division, chromosomal replication and DNA unwinding or repairing. Our results reveal that EMB564 protein is most likely to function within the cell nucleus.     

Both Sm-domain and C-terminal extension of Lsm1 are important for the RNA-binding activity of the Lsm1-7-Pat1 complex

Lsm proteins are a ubiquitous family of proteins characterized by the Sm-domain. They exist as hexa- or heptameric RNA-binding complexes and carry out RNA-related functions. The Sm-domain is thought to be sufficient for the RNA-binding activity of these proteins. The highly conserved eukaryotic Lsm1 through Lsm7 proteins are part of the cytoplasmic Lsm1-7-Pat1 complex, which is an activator of decapping in the conserved 5'-3' mRNA decay pathway. This complex also protects mRNA 3'-ends from trimming in vivo. Purified Lsm1-7-Pat1 complex is able to bind RNA in vitro and exhibits a unique binding preference for oligoadenylated RNA (over polyadenylated and unadenylated RNA). Lsm1 is a key subunit that determines the RNA-binding properties of this complex. The normal RNA-binding activity of this complex is crucial for mRNA decay and 3'-end protection in vivo and requires the intact Sm-domain of Lsm1. Here, we show that though necessary, the Sm-domain of Lsm1 is not sufficient for the normal RNA-binding ability of the Lsm1-7-Pat1 complex. Deletion of the C-terminal domain (CTD) of Lsm1 (while keeping the Sm-domain intact) impairs mRNA decay in vivo and results in Lsm1-7-Pat1 complexes that are severely impaired in RNA binding in vitro. Interestingly, the mRNA decay and 3'-end protection defects of such CTD-truncated lsm1 mutants could be suppressed in trans by overexpression of the CTD polypeptide. Thus, unlike most Sm-like proteins, Lsm1 uniquely requires both its Sm-domain and CTD for its normal RNA-binding function.