RNANetMotif: Identifying sequence-structure RNA network motifs in RNA-protein binding sites

RNA molecules can adopt stable secondary and tertiary structures, which are essential in mediating physical interactions with other partners such as RNA binding proteins (RBPs) and in carrying out their cellular functions. In vivo and in vitro experiments such as RNAcompete and eCLIP have revealed in vitro binding preferences of RBPs to RNA oligomers and in vivo binding sites in cells. Analysis of these binding data showed that the structure properties of the RNAs in these binding sites are important determinants of the binding events; however, it has been a challenge to incorporate the structure information into an interpretable model. Here we describe a new approach, RNANetMotif, which takes predicted secondary structure of thousands of RNA sequences bound by an RBP as input and uses a graph theory approach to recognize enriched subgraphs. These enriched subgraphs are in essence shared sequence-structure elements that are important in RBP-RNA binding. To validate our approach, we performed RNA structure modeling via coarse-grained molecular dynamics folding simulations for selected 4 RBPs, and RNA-protein docking for LIN28B. The simulation results, e.g., solvent accessibility and energetics, further support the biological relevance of the discovered network subgraphs.     

Presence of a poly(A) binding protein and two proteins with cell cycle-dependent phosphorylation in Crithidia fasciculata mRNA cycling sequence binding protein II

Crithidia fasciculata cycling sequence binding proteins (CSBP) have been shown to bind with high specificity to sequence elements present in several mRNAs that accumulate periodically during the cell cycle. The first described CSBP has subunits of 35.6 (CSBPA) and 42 kDa (CSBPB). A second distinct binding protein termed CSBP II has been purified from CSBPA null mutant cells, lacking both CSBPA and CSBPB proteins, and contains three major polypeptides with predicted molecular masses of 63, 44.5, and 33 kDa. Polypeptides of identical size were radiolabeled in UV cross-linking assays performed with purified CSBP II and 32P-labeled RNA probes containing six copies of the cycling sequence. The CSBP II binding activity was found to cycle in parallel with target mRNA levels during progression through the cell cycle. We have cloned genes encoding these three CSBP II proteins, termed RBP63, RBP45, and RBP33, and characterized their binding properties. The RBP63 protein is a member of the poly(A) binding protein family. Homologs of RBP45 and RBP33 proteins were found only among the kinetoplastids. Both RBP45 and RBP33 proteins and their homologs have a conserved carboxy-terminal half that contains a PSP1-like domain. All three CSBP II proteins show specificity for binding the wild-type cycling sequence in vitro. RBP45 and RBP33 are phosphoproteins, and RBP45 has been found to bind in vivo specifically to target mRNA containing cycling sequences. The levels of phosphorylation of both RBP45 and RBP33 were found to cycle during the cell cycle.     

The KH domains of Xenopus Vg1RBP mediate RNA binding and self-association

Xenopus Vg1 mRNA is localized to the vegetal cortex during oogenesis in a process involving microtubules and microfilaments and proteins that specifically recognize the vegetal localization element (VLE) within the 3' untranslated region. One of the best characterized VLE-binding proteins is Vg1RBP or Vera. Primary sequence analysis of Vg1RBP and its homologs suggests that most of its open reading frame is occupied by RNA-binding modules, including two RRMs and four KH domains, arranged as three pairs of didomains. In the first detailed domain analysis of Vg1RBP, we show that the interaction of Vg1RBP with the VLE requires both KH didomains, but not the RRM didomain, and moreover that the KH didomains contribute cooperatively to RNA binding. In the full-length protein, individual KH domains display significant redundancy, and their relative importance appears to vary with the RNA target. We also demonstrate that the KH34 didomain mediates Vg1RBP self-association, which is stabilized by RNA, and occurs in vivo as well as in vitro. Altogether, our findings highlight the importance of multiple KH domains in mediating RNA-protein and protein-protein interactions in the formation of a stable complex of Vg1RBP and Vg1 mRNA.     

PRIESSTESS: interpretable,high-performing models of the sequence and structure preferences of RNA-binding proteins

Modelling both primary sequence and secondary structure preferences for RNA binding proteins (RBPs) remains an ongoing challenge. Current models use varied RNA structure representations and can be difficult to interpret and evaluate. To address these issues, we present a universal RNA motif-finding/scanning strategy, termed PRIESSTESS (Predictive RBP-RNA InterpretablE Sequence-Structure moTif regrESSion), that can be applied to diverse RNA binding datasets. PRIESSTESS identifies dozens of enriched RNA sequence and/or structure motifs that are subsequently reduced to a set of core motifs by logistic regression with LASSO regularization. Importantly, these core motifs are easily visualized and interpreted, and provide a measure of RBP secondary structure specificity. We used PRIESSTESS to interrogate new HTR-SELEX data for 23 RBPs with diverse RNA binding modes and captured known primary sequence and secondary structure preferences for each. Moreover, when applying PRIESSTESS to 144 RBPs across 202 RNA binding datasets, 75% showed an RNA secondary structure preference but only 10% had a preference besides unpaired bases, suggesting that most RBPs simply recognize the accessibility of primary sequences.     

RBP16 stimulates trypanosome RNA editing in vitro at an early step in the editing reaction

RBP16 is an abundant RNA binding protein from Trypanosoma brucei mitochondria that affects both RNA editing and stability. We report here experiments aimed at elucidating the mechanism of RBP16 function in RNA editing. In in vitro RNA editing assays, recombinant RBP16 is able to significantly stimulate insertion editing of both CYb and A6 pre-mRNAs. Enhancement of in vitro editing activity occurs at, or prior to, the step of pre-mRNA cleavage, as evidenced by increased accumulation of pre-mRNA 3' cleavage products in the presence of RBP16. Mutated RBP16 that is severely compromised in cold shock domain (CSD)-mediated RNA binding was able to enhance editing to levels comparable to the wild-type protein in some assays at the highest RBP16 levels tested. However, at low RBP16 concentrations or in assays with native, oligo(U)-tail-bearing gRNAs, editing stimulation by mutant RBP16 was somewhat compromised. Together, these results indicate that both the N-terminal CSD and C-terminal RGG RNA binding domains of RBP16 are required for maximal editing stimulation. Finally, the relaxed specificity of RBP16 for stimulation of both CYb and A6 editing in vitro implicates additional specificity factors that account for the strict CYb specificity of RBP16 action in editing in vivo. Our results constitute the first report of any putative RNA editing accessory factor eliciting an effect on editing in vitro. Overall, these results support a novel accessory role for RBP16 in U insertion editing.     

Cross-domain retinopathy classification with optical coherence tomography images via a novel deep domain adaptation method

Deep learning based retinopathy classification with optical coherence tomography (OCT) images has recently attracted great attention. However, existing deep learning methods fail to work well when training and testing datasets are different due to the general issue of domain shift between datasets caused by different collection devices, subjects, imaging parameters, etc. To address this practical and challenging issue, we propose a novel deep domain adaptation (DDA) method to train a model on a labeled dataset and adapt it to an unlabelled dataset (collected under different conditions). It consists of two modules for domain alignment, that is, adversarial learning and entropy minimization. We conduct extensive experiments on three public datasets to evaluate the performance of the proposed method. The results indicate that there are large domain shifts between datasets, resulting a poor performance for conventional deep learning methods. The proposed DDA method can significantly outperform existing methods for retinopathy classification with OCT images. It achieves retinopathy classification accuracies of 0.915, 0.959 and 0.990 under three cross-domain (cross-dataset) scenarios. Moreover, it obtains a comparable performance with human experts on a dataset where no labeled data in this dataset have been used to train the proposed DDA method. We have also visualized the learnt features by using the t-distributed stochastic neighbor embedding (t-SNE) technique. The results demonstrate that the proposed method can learn discriminative features for retinopathy classification.     

gRNA/pre-mRNA annealing and RNA chaperone activities of RBP16

Editing in trypanosomes involves the addition or deletion of uridines at specific sites to produce translatable mitochondrial mRNAs. RBP16 is an accessory factor from Trypanosoma brucei that affects mitochondrial RNA editing in vivo and also stimulates editing in vitro. We report here experiments aimed at elucidating the biochemical activities of RBP16 involved in modulating RNA editing. In vitro RNA annealing assays demonstrate that RBP16 significantly stimulates the annealing of gRNAs to cognate pre-mRNAs. In addition, RBP16 also facilitates hybridization of partially complementary RNAs unrelated to the editing process. The RNA annealing activity of RBP16 is independent of its high-affinity binding to gRNA oligo(U) tails, consistent with the previously reported in vitro editing stimulatory properties of the protein. In vivo studies expressing recombinant RBP16 in mutant Escherichia coli strains demonstrate that RBP16 is an RNA chaperone and that in addition to RNA annealing activity, it contains RNA unwinding activity. Our data suggest that the mechanism by which RBP16 facilitates RNA editing involves its capacity to modulate RNA secondary structure and promote gRNA/pre-mRNA annealing.     

Quantitative dissection of color patterning in the foliar ornamental coleus

Coleus (Coleus scutellarioides) is a popular ornamental plant that exhibits a diverse array of foliar color patterns. New cultivars are currently hand selected by both amateur and experienced plant breeders. In this study, we reimagine breeding for color patterning using a quantitative color analysis framework. Despite impressive advances in high-throughput data collection and processing, complex color patterns remain challenging to extract from image datasets. Using a phenotyping approach called “ColourQuant,” we extract and analyze pigmentation patterns from one of the largest coleus breeding populations in the world. Working with this massive dataset, we can analyze quantitative relationships between maternal plants and their progeny, identify features that underlie breeder-selections, and collect and compare public input on trait preferences. This study is one of the most comprehensive explorations into complex color patterning in plant biology and provides insights and tools for exploring the color pallet of the plant kingdom.

Quantitative analysis of color patterning in a large coleus breeding population reveals color features that are associated with aesthetic value.     

Protein domain-based prediction of drug/compound–target interactions and experimental validation on LIM kinases

Predictive approaches such as virtual screening have been used in drug discovery with the objective of reducing developmental time and costs. Current machine learning and network-based approaches have issues related to generalization, usability, or model interpretability, especially due to the complexity of target proteins’ structure/function, and bias in system training datasets. Here, we propose a new method “DRUIDom” (DRUg Interacting Domain prediction) to identify bio-interactions between drug candidate compounds and targets by utilizing the domain modularity of proteins, to overcome problems associated with current approaches. DRUIDom is composed of two methodological steps. First, ligands/compounds are statistically mapped to structural domains of their target proteins, with the aim of identifying their interactions. As such, other proteins containing the same mapped domain or domain pair become new candidate targets for the corresponding compounds. Next, a million-scale dataset of small molecule compounds, including those mapped to domains in the previous step, are clustered based on their molecular similarities, and their domain associations are propagated to other compounds within the same clusters. Experimentally verified bioactivity data points, obtained from public databases, are meticulously filtered to construct datasets of active/interacting and inactive/non-interacting drug/compound–target pairs (~2.9M data points), and used as training data for calculating parameters of compound–domain mappings, which led to 27,032 high-confidence associations between 250 domains and 8,165 compounds, and a finalized output of ~5 million new compound–protein interactions. DRUIDom is experimentally validated by syntheses and bioactivity analyses of compounds predicted to target LIM-kinase proteins, which play critical roles in the regulation of cell motility, cell cycle progression, and differentiation through actin filament dynamics. We showed that LIMK-inhibitor-2 and its derivatives significantly block the cancer cell migration through inhibition of LIMK phosphorylation and the downstream protein cofilin. One of the derivative compounds (LIMKi-2d) was identified as a promising candidate due to its action on resistant Mahlavu liver cancer cells. The results demonstrated that DRUIDom can be exploited to identify drug candidate compounds for intended targets and to predict new target proteins based on the defined compound–domain relationships. Datasets, results, and the source code of DRUIDom are fully-available at: https://github.com/cansyl/DRUIDom.     

An amino acid in the central catalytic domain of three retroviral integrases that affects target site selection in nonviral DNA

Integrase can insert retroviral DNA into almost any site in cellular DNA; however, target site preferences are noted in vitro and in vivo. We recently demonstrated that amino acid 119, in the alpha2 helix of the central domain of the human immunodeficiency virus type 1 integrase, affected the choice of nonviral target DNA sites. We have now extended these findings to the integrases of a nonprimate lentivirus and a more distantly related alpharetrovirus. We found that substitutions at the analogous positions in visna virus integrase and Rous sarcoma virus integrase changed the target site preferences in five assays that monitor insertion into nonviral DNA. Thus, the importance of this protein residue in the selection of nonviral target DNA sites is likely to be a general property of retroviral integrases. Moreover, this amino acid might be part of the cellular DNA binding site on integrase proteins.     

An essential ligand-binding domain in the membrane receptor for retinol-binding protein revealed by large-scale mutagenesis and a human polymorphism

Plasma retinol-binding protein (RBP), the principal carrier of vitamin A in the blood, delivers vitamin A from liver, the site of storage, to distant organs that need vitamin A, such as the eye, brain, placenta, and testis. STRA6 is a high-affinity membrane receptor for RBP and mediates vitamin A uptake in these target organs. STRA6 is a 74-kDa multi-transmembrane domain protein that represents a new class of membrane transport protein. In this study, we used an unbiased strategy by analyzing >900 random mutants of STRA6 to study its structure and function, and we identified an essential RBP-binding domain in STRA6. Mutations in any of the three essential residues in this domain can almost completely abolish binding of STRA6 to RBP and its vitamin A uptake activity from holo-RBP without affecting its cell surface expression. We have also functionally characterized the mutations in human STRA6 that cause severe birth defects as well as several human polymorphisms. All STRA6 mutants associated with severe birth defects have largely abolished vitamin A uptake activity, consistent with the severe clinical phenotypes. In addition, we have identified a human polymorphism that significantly reduces the vitamin A uptake activity of STRA6. Interestingly, the residue affected by this polymorphism is located in the RBP-binding domain we identified, and the polymorphism causes decreased vitamin A uptake by reducing RBP binding. This study identifies an essential functional domain in STRA6 and a human polymorphism in this domain that leads to reduced vitamin A uptake activity.