scholarly journals A Deep Learning Approach for Learning Intrinsic Protein-RNA Binding Preferences

2018 ◽  
Author(s):  
Ilan Ben-Bassat ◽  
Benny Chor ◽  
Yaron Orenstein
Keyword(s):  
Neural Network ◽  
Rna Structure ◽  
Rna Binding ◽  
Gene Expression Regulation ◽  
Supplementary Information ◽  
Joint Analysis ◽  
Binding Prediction ◽  
Binding Models

AbstractMotivationThe complexes formed by binding of proteins to RNAs play key roles in many biological processes, such as splicing, gene expression regulation, translation, and viral replication. Understanding protein-RNA binding may thus provide important insights to the functionality and dynamics of many cellular processes. This has sparked substantial interest in exploring protein-RNA binding experimentally, and predicting it computationally. The key computational challenge is to efficiently and accurately infer RNA-binding models that will enable prediction of novel protein-RNA interactions to additional transcripts of interest.ResultsWe developed DLPRB, a new deep neural network (DNN) approach for learning protein-RNA binding preferences and predicting novel interactions. We present two different network architectures: a convolutional neural network (CNN), and a recurrent neural network (RNN). The novelty of our network hinges upon two key aspects: (i) the joint analysis of both RNA sequence and structure, which is represented as a probability vector of different RNA structural contexts; (ii) novel features in the architecture of the networks, such as the application of RNNs to RNA-binding prediction, and the combination of hundreds of variable-length filters in the CNN. Our results in inferring accurate RNA-binding models from high-throughput in vitro data exhibit substantial improvements, compared to all previous approaches for protein-RNA binding prediction (both DNN and non-DNN based). A highly significant improvement is achieved for in vitro binding prediction, and a more modest, yet statistically significant,improvement for in vivo binding prediction. When incorporating experimentally-measured RNA structure compared to predicted one, the improvement on in vivo data increases. By visualizing the binding specificities, we can gain novel biological insights underlying the mechanism of protein RNA-binding.AvailabilityThe source code is publicly available at https://github.com/ilanbb/[email protected] informationSupplementary data are available at Bioinformatics online.

scholarly journals RNA Binding Properties of the Ty1 LTR-Retrotransposon Gag Protein

2021 ◽  
Vol 22 (16) ◽  
pp. 9103
Author(s):  
Julita Gumna ◽  
Angelika Andrzejewska-Romanowska ◽  
David J. Garfinkel ◽  
Katarzyna Pachulska-Wieczorek
Keyword(s):  
Rna Structure ◽  
Rna Binding ◽  
Rna Packaging ◽  
Packaging Signal ◽  
Mobility Shift ◽  
Gag Protein ◽  
Rna Dimerization ◽  
Electrophoretic Mobility Shift Assays

A universal feature of retroelement propagation is the formation of distinct nucleoprotein complexes mediated by the Gag capsid protein. The Ty1 retrotransposon Gag protein from Saccharomyces cerevisiae lacks sequence homology with retroviral Gag, but is functionally related. In addition to capsid assembly functions, Ty1 Gag promotes Ty1 RNA dimerization and cyclization and initiation of reverse transcription. Direct interactions between Gag and retrotransposon genomic RNA (gRNA) are needed for Ty1 replication, and mutations in the RNA-binding domain disrupt nucleation of retrosomes and assembly of functional virus-like particles (VLPs). Unlike retroviral Gag, the specificity of Ty1 Gag-RNA interactions remain poorly understood. Here we use microscale thermophoresis (MST) and electrophoretic mobility shift assays (EMSA) to analyze interactions of immature and mature Ty1 Gag with RNAs. The salt-dependent experiments showed that Ty1 Gag binds with high and similar affinity to different RNAs. However, we observed a preferential interaction between Ty1 Gag and Ty1 RNA containing a packaging signal (Psi) in RNA competition analyses. We also uncover a relationship between Ty1 RNA structure and Gag binding involving the pseudoknot present on Ty1 gRNA. In all likelihood, the differences in Gag binding affinity detected in vitro only partially explain selective Ty1 RNA packaging into VLPs in vivo.

scholarly journals Multi-resBind: a residual network-based multi-label classifier for in vivo RNA binding prediction and preference visualization

2021 ◽  
Vol 22 (1) ◽  
Author(s):  
Shitao Zhao ◽  
Michiaki Hamada
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Binding Sites ◽  
Prediction Accuracy ◽  
Large Scale ◽  
Rna Binding ◽  
Binding Prediction ◽  
Binding Preference ◽  
The Impact

Abstract Background Protein-RNA interactions play key roles in many processes regulating gene expression. To understand the underlying binding preference, ultraviolet cross-linking and immunoprecipitation (CLIP)-based methods have been used to identify the binding sites for hundreds of RNA-binding proteins (RBPs) in vivo. Using these large-scale experimental data to infer RNA binding preference and predict missing binding sites has become a great challenge. Some existing deep-learning models have demonstrated high prediction accuracy for individual RBPs. However, it remains difficult to avoid significant bias due to the experimental protocol. The DeepRiPe method was recently developed to solve this problem via introducing multi-task or multi-label learning into this field. However, this method has not reached an ideal level of prediction power due to the weak neural network architecture. Results Compared to the DeepRiPe approach, our Multi-resBind method demonstrated substantial improvements using the same large-scale PAR-CLIP dataset with respect to an increase in the area under the receiver operating characteristic curve and average precision. We conducted extensive experiments to evaluate the impact of various types of input data on the final prediction accuracy. The same approach was used to evaluate the effect of loss functions. Finally, a modified integrated gradient was employed to generate attribution maps. The patterns disentangled from relative contributions according to context offer biological insights into the underlying mechanism of protein-RNA interactions. Conclusions Here, we propose Multi-resBind as a new multi-label deep-learning approach to infer protein-RNA binding preferences and predict novel interactions. The results clearly demonstrate that Multi-resBind is a promising tool to predict unknown binding sites in vivo and gain biology insights into why the neural network makes a given prediction.

scholarly journals Enterococcal PrgU mitigates PrgB overexpression toxicity by binding to intergenic RNA downstream of the PQ promoter

2020 ◽  
Author(s):  
Lena Lassinantti ◽  
Martha I Camacho ◽  
Rebecca J B Erickson ◽  
Julia L E Willett ◽  
Nicholas R. De Lay ◽  
...  
Keyword(s):  
Enterococcus Faecalis ◽  
Rna Structure ◽  
Rna Binding ◽  
Fine Tuning ◽  
Conjugative Plasmid ◽  
Regulatory Control ◽  
Sequence Specificity ◽  
Secretion Systems

AbstractEfficient horizontal gene transfer of the conjugative plasmid pCF10 from Enterococcus faecalis depends on the sex pheromone cCF10, which induces the expression of the Type 4 Secretion System (T4SS) genes controlled by the PQ promoter. The pheromone responsive PQ promoter is strictly regulated to prevent overproduction of the prgQ operon, which contains the T4SS, and to limit the cell toxicity caused by overproduction of PrgB, a T4SS adhesin involved in cellular aggregation. PrgU plays an important role in regulating this toxicity by decreasing PrgB production. PrgU has an RNA-binding fold, prompting us to test whether PrgU exerts its regulatory control through binding of prgQ transcripts. With a combination of lacZ reporter fusion, northern blot, and RNAseq analyses, we provide evidence that PrgU binds a specific RNA sequence within the intergenic region (IGR), ca 400 bp downstream of the PQ promoter. PrgU-IGR binding reduces levels of downstream transcripts, with the strongest decrease seen for prgB messages. Consistent with these findings, we determined that pCF10-carrying cells expressing prgU decreased transcript levels more rapidly than isogenic cells deleted of prgU. Finally, purified PrgU bound RNA in vitro, but without sequence specificity, suggesting that PrgU requires a specific RNA structure or one or more host factors to bind its RNA target in vivo. Together, our results support a working model where PrgU binding to the IGR serves to recruit RNase(s) for targeted degradation of downstream transcripts.ImportanceBacteria utilize Type 4 Secretion Systems (T4SS) to efficiently transfer DNA from donor to recipient cells, thereby spreading genes encoding for antibiotic resistance as well as various virulence factors. The conjugative plasmid pCF10 from Enterococcus faecalis, originally isolated from clinical isolates, serves as a model system for these processes in Gram-positive bacteria. It is very important to strictly regulate the expression of the T4SS proteins for the bacteria, as some of these proteins are highly toxic to the cell. Here, we identify the mechanism by which PrgU performs its delicate fine tuning of the expression levels. As prgU genes are present in various conjugative plasmids and transposons, this provides an important new insight into the bacterial repertoire of regulation mechanisms of these clinically important systems.

scholarly journals The methyltransferase SETD2 couples transcription and splicing by engaging mRNA processing factors through its SHI domain

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Saikat Bhattacharya ◽  
Michaella J. Levy ◽  
Ning Zhang ◽  
Hua Li ◽  
Laurence Florens ◽  
...  
Keyword(s):  
Rna Splicing ◽  
Rna Binding ◽  
Mrna Processing ◽  
Key Factors ◽  
Pol Ii ◽  
Mrna Transcripts ◽  
Hnrnp Protein ◽  
Processing Factors

AbstractHeterogeneous ribonucleoproteins (hnRNPs) are RNA binding molecules that are involved in key processes such as RNA splicing and transcription. One such hnRNP protein, hnRNP L, regulates alternative splicing (AS) by binding to pre-mRNA transcripts. However, it is unclear what factors contribute to hnRNP L-regulated AS events. Using proteomic approaches, we identified several key factors that co-purify with hnRNP L. We demonstrate that one such factor, the histone methyltransferase SETD2, specifically interacts with hnRNP L in vitro and in vivo. This interaction occurs through a previously uncharacterized domain in SETD2, the SETD2-hnRNP Interaction (SHI) domain, the deletion of which, leads to a reduced H3K36me3 deposition. Functionally, SETD2 regulates a subset of hnRNP L-targeted AS events. Our findings demonstrate that SETD2, by interacting with Pol II as well as hnRNP L, can mediate the crosstalk between the transcription and the splicing machinery.

scholarly journals Oncogenic lncRNA ZNF561-AS1 is essential for colorectal cancer proliferation and survival through regulation of miR-26a-3p/miR-128-5p-SRSF6 axis

2021 ◽  
Vol 40 (1) ◽  
Author(s):  
Zizhen Si ◽  
Lei Yu ◽  
Haoyu Jing ◽  
Lun Wu ◽  
Xidi Wang
Keyword(s):  
Colorectal Cancer ◽  
Rna Binding ◽  
Xenograft Model ◽  
Luciferase Reporter ◽  
Cancer Proliferation ◽  
Mouse Xenograft ◽  
Mouse Xenograft Model

Abstract Background Long non-coding RNAs (lncRNA) are reported to influence colorectal cancer (CRC) progression. Currently, the functions of the lncRNA ZNF561 antisense RNA 1 (ZNF561-AS1) in CRC are unknown. Methods ZNF561-AS1 and SRSF6 expression in CRC patient samples and CRC cell lines was evaluated through TCGA database analysis, western blot along with real-time PCR. SRSF6 expression in CRC cells was also examined upon ZNF561-AS1 depletion or overexpression. Interaction between miR-26a-3p, miR-128-5p, ZNF561-AS1, and SRSF6 was examined by dual luciferase reporter assay, as well as RNA binding protein immunoprecipitation (RIP) assay. Small interfering RNA (siRNA) mediated knockdown experiments were performed to assess the role of ZNF561-AS1 and SRSF6 in the proliferative actives and apoptosis rate of CRC cells. A mouse xenograft model was employed to assess tumor growth upon ZNF561-AS1 knockdown and SRSF6 rescue. Results We find that ZNF561-AS1 and SRSF6 were upregulated in CRC patient tissues. ZNF561-AS1 expression was reduced in tissues from treated CRC patients but upregulated in CRC tissues from relapsed patients. SRSF6 expression was suppressed and enhanced by ZNF561-AS1 depletion and overexpression, respectively. Mechanistically, ZNF561-AS1 regulated SRSF6 expression by sponging miR-26a-3p and miR-128-5p. ZNF561-AS1-miR-26a-3p/miR-128-5p-SRSF6 axis was required for CRC proliferation and survival. ZNF561-AS1 knockdown suppressed CRC cell proliferation and triggered apoptosis. ZNF561-AS1 depletion suppressed the growth of tumors in a model of a nude mouse xenograft. Similar observations were made upon SRSF6 depletion. SRSF6 overexpression reversed the inhibitory activities of ZNF561-AS1 in vivo, as well as in vitro. Conclusion In summary, we find that ZNF561-AS1 promotes CRC progression via the miR-26a-3p/miR-128-5p-SRSF6 axis. This study reveals new perspectives into the role of ZNF561-AS1 in CRC.

scholarly journals RBMX suppresses tumorigenicity and progression of bladder cancer by interacting with the hnRNP A1 protein to regulate PKM alternative splicing

Oncogene ◽  
2021 ◽  
Author(s):  
Qiuxia Yan ◽  
Peng Zeng ◽  
Xiuqin Zhou ◽  
Xiaoying Zhao ◽  
Runqiang Chen ◽  
...  
Keyword(s):  
Bladder Cancer ◽  
Alternative Splicing ◽  
Rna Binding ◽  
Aerobic Glycolysis ◽  
Histological Grade ◽  
Migration And Invasion ◽  
Binding Motif ◽  
Hnrnp A1

AbstractThe prognosis for patients with metastatic bladder cancer (BCa) is poor, and it is not improved by current treatments. RNA-binding motif protein X-linked (RBMX) are involved in the regulation of the malignant progression of various tumors. However, the role of RBMX in BCa tumorigenicity and progression remains unclear. In this study, we found that RBMX was significantly downregulated in BCa tissues, especially in muscle-invasive BCa tissues. RBMX expression was negatively correlated with tumor stage, histological grade and poor patient prognosis. Functional assays demonstrated that RBMX inhibited BCa cell proliferation, colony formation, migration, and invasion in vitro and suppressed tumor growth and metastasis in vivo. Mechanistic investigations revealed that hnRNP A1 was an RBMX-binding protein. RBMX competitively inhibited the combination of the RGG motif in hnRNP A1 and the sequences flanking PKM exon 9, leading to the formation of lower PKM2 and higher PKM1 levels, which attenuated the tumorigenicity and progression of BCa. Moreover, RBMX inhibited aerobic glycolysis through hnRNP A1-dependent PKM alternative splicing and counteracted the PKM2 overexpression-induced aggressive phenotype of the BCa cells. In conclusion, our findings indicate that RBMX suppresses BCa tumorigenicity and progression via an hnRNP A1-mediated PKM alternative splicing mechanism. RBMX may serve as a novel prognostic biomarker for clinical intervention in BCa.

scholarly journals Andrographis paniculata (Burm. F.) Wall. Ex Nees, Andrographolide, and Andrographolide Analogues as SARS-CoV-2 Antivirals? A Rapid Review

2021 ◽  
Vol 16 (5) ◽  
pp. 1934578X2110166
Author(s):  
Xin Yi Lim ◽  
Janice Sue Wen Chan ◽  
Terence Yew Chin Tan ◽  
Bee Ping Teh ◽  
Mohd Ridzuan Mohd Abd Razak ◽  
...  
Keyword(s):  
Inhibitory Activity ◽  
In Silico ◽  
Rna Binding ◽  
Symptomatic Relief ◽  
Andrographis Paniculata ◽  
Spike Protein ◽  
Rapid Review ◽  
In Silico Studies

Drug repurposing is commonly employed in the search for potential therapeutic agents. Andrographis paniculata, a medicinal plant commonly used for symptomatic relief of the common cold, and its phytoconstituent andrographolide, have been repeatedly identified as potential antivirals against SARS-CoV-2. In light of new evidence emerging since the onset of the COVID-19 pandemic, this rapid review was conducted to identify and evaluate the current SARS-CoV-2 antiviral evidence for A. paniculata, andrographolide, and andrographolide analogs. A systematic search and screen strategy of electronic databases and gray literature was undertaken to identify relevant primary articles. One target-based in vitro study reported the 3CLpro inhibitory activity of andrographolide as being no better than disulfiram. Another Vero cell-based study reported potential SARS-CoV-2 inhibitory activity for both andrographolide and A. paniculata extract. Eleven in silico studies predicted the binding of andrographolide and its analogs to several key antiviral targets of SARS-CoV-2 including the spike protein-ACE-2 receptor complex, spike protein, ACE-2 receptor, RdRp, 3CLpro, PLpro, and N-protein RNA-binding domain. In conclusion, in silico and in vitro studies collectively suggest multi-pathway targeting SARS-CoV-2 antiviral properties of andrographolide and its analogs, but in vivo data are needed to support these predictions.

Tissue specificity of type I collagen gene expression is determined at both transcriptional and post-transcriptional levels

1984 ◽  
Vol 4 (9) ◽  
pp. 1843-1852
Author(s):  
R J Focht ◽  
S L Adams
Keyword(s):  
Gene Expression ◽  
Rna Structure ◽  
Type I Collagen ◽  
Translational Efficiency ◽  
Type I ◽  
Collagen Gene ◽  
Differentiated Cells ◽  
Collagen Gene Expression

We analyzed the control of type I collagen synthesis in four kinds of differentiated cells from chicken embryos which synthesize very different amounts of the protein. Tendon, skin, and smooth muscle cells were found to have identical amounts of type I collagen RNAs; however, the RNAs had inherently different translatabilities, which were observed both in vivo and in vitro. Chondrocytes also had substantial amounts of type I collagen RNAs, even though they directed no detectable synthesis of the protein either in vivo or in vitro. Type I collagen RNAs in chondrocytes display altered electrophoretic mobilities, suggesting that in these cells the reduction in translational efficiency may be mediated in part by changes in the RNA structure. These data indicate that control of type I collagen gene expression is a complex process which is exerted at both transcriptional and post-transcriptional levels.

scholarly journals Identification of mRNAs Associated with αCP2-Containing RNP Complexes

2003 ◽  
Vol 23 (19) ◽  
pp. 7055-7067 ◽  
Author(s):  
Shelly A. Waggoner ◽  
Stephen A. Liebhaber
Keyword(s):  
Binding Proteins ◽  
Rna Binding ◽  
Cell Function ◽  
Rna Binding Proteins ◽  
Translation Control ◽  
Viral Mrnas ◽  
Posttranscriptional Gene Regulation ◽  
Direct Binding

ABSTRACT Posttranscriptional controls in higher eukaryotes are central to cell differentiation and developmental programs. These controls reflect sequence-specific interactions of mRNAs with one or more RNA binding proteins. The α-globin poly(C) binding proteins (αCPs) comprise a highly abundant subset of K homology (KH) domain RNA binding proteins and have a characteristic preference for binding single-stranded C-rich motifs. αCPs have been implicated in translation control and stabilization of multiple cellular and viral mRNAs. To explore the full contribution of αCPs to cell function, we have identified a set of mRNAs that associate in vivo with the major αCP2 isoforms. One hundred sixty mRNA species were consistently identified in three independent analyses of αCP2-RNP complexes immunopurified from a human hematopoietic cell line (K562). These mRNAs could be grouped into subsets encoding cytoskeletal components, transcription factors, proto-oncogenes, and cell signaling factors. Two mRNAs were linked to ceroid lipofuscinosis, indicating a potential role for αCP2 in this infantile neurodegenerative disease. Surprisingly, αCP2 mRNA itself was represented in αCP2-RNP complexes, suggesting autoregulatory control of αCP2 expression. In vitro analyses of representative target mRNAs confirmed direct binding of αCP2 within their 3′ untranslated regions. These data expand the list of mRNAs that associate with αCP2 in vivo and establish a foundation for modeling its role in coordinating pathways of posttranscriptional gene regulation.

scholarly journals In vivo reconstitution finds multivalent RNA-RNA interactions as drivers of mesh-like condensates

eLife ◽  
2021 ◽  
Vol 10 ◽  
Author(s):  
Weirui Ma ◽  
Gang Zheng ◽  
Wei Xie ◽  
Christine Mayr
Keyword(s):  
Surface Area ◽  
Protein Trafficking ◽  
Rna Binding ◽  
Interaction Network ◽  
Large Surface Area ◽  
High Propensity ◽  
Membrane Protein Trafficking ◽  
Protein Components

Liquid-like condensates have been thought to be sphere-like. Recently, various condensates with filamentous morphology have been observed in cells. One such condensate is the TIS granule network that shares a large surface area with the rough endoplasmic reticulum and is important for membrane protein trafficking. It has been unclear how condensates with mesh-like shapes, but dynamic protein components are formed. In vitro and in vivo reconstitution experiments revealed that the minimal components are a multivalent RNA-binding protein that concentrates RNAs that are able to form extensive intermolecular mRNA-mRNA interactions. mRNAs with large unstructured regions have a high propensity to form a pervasive intermolecular interaction network that acts as condensate skeleton. The underlying RNA matrix prevents full fusion of spherical liquid-like condensates, thus driving the formation of irregularly shaped membraneless organelles. The resulting large surface area may promote interactions at the condensate surface and at the interface with other organelles.

Sign in / Sign up

Export Citation Format

Share Document