Prediction of LncRNA Subcellular Localization with Deep Learning from Sequence Features

A novel lincRNA identified in buffalo oocytes with protein binding characteristics could hold the key for oocyte competence

Studying the maternal oocyte-specific genes, in farm animals is a significant step towards delineating the underlying mechanisms that regulate oocyte quality, early embryonic development and survival. With the creation of buffalo oocyte-specific subtracted cDNA library, it has raised new questions which need to be answered. The present study has characterized one of the ESTs selected from the library and highlighted its importance in the oocyte quality. The selected EST was made full length by RLM-RACE and four transcript variants were identified. Bioinformatics analysis indicated the novelty of full-length transcript along with conserved intergenic nature. The largest transcript was identified as long intergenic noncoding RNA based upon coding potential calculator output. The expression analysis at different hours of oocyte maturation showed a significant variation in developmentally competent oocytes to that of incompetent ones. Along with this, the transcript was also found to have protein binding ability which was confirmed by RNA electrophoretic mobility shift assay. The protein used in the experiment was isolated from oocyte and cumulus cells via sonication. A novel lincRNA has been reported here that might have an important role in maturation of oocytes, inferred from its relative gene expression study and protein binding characteristics.

Cis-acting: A pattern of lncRNAs for gene regulation in induced pluripotent stem cells from patients with Down syndrome determined by integrative analysis of lncRNA and mRNA profiling data

Down syndrome (DS), caused by the trisomy of chromosome 21, is one of the common chromosomal disorders, the main clinical manifestations of which are delayed nervous development and intellectual disability. Long non-coding RNAs (lncRNAs) have critical roles in various biological processes, including cell growth, cell cycle regulation and differentiation. The roles of abnormally expressed lncRNAs have been previously reported; however, the biological functions and regulatory patterns of lncRNAs in DS have remained largely elusive. The aim of the present study was to perform a whole-genome-wide identification of lncRNAs and mRNAs associated with DS. In addition, global expression profiling analysis of DS-induced pluripotent stem cells was performed and differentially expressed (DE) lncRNAs and mRNAs were screened. Furthermore, the target genes and functions of the DE lncRNAs were predicted using Gene Ontology annotation and Kyoto Encyclopedia of Genes and Genomes signaling pathway enrichment analysis. The results revealed that the majority of the lncRNAs exerted functions in DS via cis-acting target genes. In addition, the results of the enrichment analysis indicated that these target genes were mainly involved in nervous and muscle development in DS. In conclusion, this integrative analysis using lncRNA and mRNA profiling provided novel insight into the pathogenesis of DS and it may promote the diagnosis and development of novel therapeutics for this disease.

PAX8-AS1 knockdown facilitates cell growth and inactivates autophagy in osteoblasts via the miR-1252-5p/GNB1 axis in osteoporosis

Osteoporosis (OP) is the most common systematic bone disorder among elderly individuals worldwide. Long noncoding RNAs (lncRNAs) are involved in biological processes in various human diseases. It has been previously revealed that PAX8 antisense RNA 1 (PAX8-AS1) is upregulated in OP. However, its molecular mechanism in OP remains unclear. Therefore, we specifically designed this study to determine the specific role of PAX8-AS1 in OP. We first established a rat model of OP and then detected PAX8-AS1 expression in the rats with RT-qPCR. Next, to explore the biological function of PAX8-AS1 in osteoblasts, in vitro experiments, such as Cell Counting Kit-8 (CCK-8) assays, flow cytometry, western blotting and immunofluorescence (IF) staining, were conducted. Subsequently, we performed bioinformatic analysis and luciferase reporter assays to predict and identify the relationships between microRNA 1252-5p (miR-1252-5p) and both PAX8-AS1 and G protein subunit beta 1 (GNB1). Additionally, rescue assays in osteoblasts clarified the regulatory network of the PAX8-AS1/miR-1252-5p/GNB1 axis. Finally, in vivo loss-of-function studies verified the role of PAX8-AS1 in OP progression. The results illustrated that PAX8-AS1 was upregulated in the proximal tibia of OP rats. PAX8-AS1 silencing promoted the viability and inhibited the apoptosis and autophagy of osteoblasts. PAX8-AS1 interacted with miR-1252-5p. GNB1 was negatively regulated by miR-1252-5p. In addition, the impacts of PAX8-AS1 knockdown on osteoblasts were counteracted by GNB1 overexpression. PAX8-AS1 depletion suppressed OP progression by inhibiting apoptosis and autophagy in osteoblasts. In summary, PAX8-AS1 suppressed the viability and activated the autophagy of osteoblasts via the miR-1252-5p/GNB1 axis in OP.

Expression-based prediction of human essential genes and candidate lncRNAs in cancer cells

MOTIVATION

Essential genes are required for the reproductive success at either cellular or organismal level. The identification of essential genes is important for understanding the core biological processes and identifying effective therapeutic drug targets. However, experimental identification of essential genes is costly, time consuming and labor intensive. Although several machine learning models have been developed to predict essential genes, these models are not readily applicable to lncRNAs. Moreover, the currently available models cannot be used to predict essential genes in a specific cancer type.

RESULTS

In this study, we have developed a new machine learning approach, XGEP (eXpression-based Gene Essentiality Prediction), to predict essential genes and candidate lncRNAs in cancer cells. The novelty of XGEP lies in the utilization of relevant features derived from the TCGA transcriptome dataset through collaborative embedding. When evaluated on the pan-cancer dataset, XGEP was able to accurately predict human essential genes and achieve significantly higher performance than previous models. Notably, several candidate lncRNAs selected by XGEP are reported to promote cell proliferation and inhibit cell apoptosis. Moreover, XGEP also demonstrated superior performance on cancer-type-specific datasets to identify essential genes. The comprehensive lists of candidate essential genes in specific cancer types may be used to guide experimental characterization and facilitate the discovery of drug targets for cancer therapy.

AVAILABILITY AND IMPLEMENTATION

The source code and datasets used in this study are freely available at https://github.com/BioDataLearning/XGEP.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Identification and Functional Prediction of Long Non-Coding RNAs in Dilated Cardiomyopathy by Bioinformatics Analysis

Dilated cardiomyopathy (DCM) is a relatively common cause of heart failure and the leading cause of heart transplantation. Aberrant changes in long non-coding RNAs (lncRNAs) are involved in DCM disorder; however, the detailed mechanisms underlying DCM initiation and progression require further investigation, and new molecular targets are needed. Here, we obtained lncRNA-expression profiles associated with DCM and non-failing hearts through microarray probe-sequence re-annotation. Weighted gene co-expression network analysis revealed a module highly associated with DCM status. Then eight hub lncRNAs in this module (FGD5-AS1, AC009113.1, WDFY3-AS2, NIFK-AS1, ZNF571-AS1, MIR100HG, AC079089.1, and EIF3J-AS1) were identified. All hub lncRNAs except ZNF571-AS1 were predicted as localizing to the cytoplasm. As a possible mechanism of DCM pathogenesis, we predicted that these hub lncRNAs might exert functions by acting as competing endogenous RNAs (ceRNAs). Furthermore, we found that the above results can be essentially reproduced in an independent external dataset. We observed the localization of hub lncRNAs by RNA-FISH in human aortic smooth muscle cells and confirmed the upregulation of the hub lncRNAs in DCM patients through quantitative RT-PCR. In conclusion, these findings identified eight candidate lncRNAs associated with DCM disease and revealed their potential involvement in DCM partly through ceRNA crosstalk. Our results facilitate the discovery of therapeutic targets and enhance the understanding of DCM pathogenesis.

RNAs as Regulators of Cellular Matchmaking

RNA molecules are increasingly being identified as facilitating or impeding the interaction of proteins and nucleic acids, serving as so-called scaffolds or decoys. Long non-coding RNAs have been commonly implicated in such roles, particularly in the regulation of nuclear processes including chromosome topology, regulation of chromatin state and gene transcription, and assembly of nuclear biomolecular condensates such as paraspeckles. Recently, an increased awareness of cytoplasmic RNA scaffolds and decoys has begun to emerge, including the identification of non-coding regions of mRNAs that can also function in a scaffold-like manner to regulate interactions of nascently translated proteins. Collectively, cytoplasmic RNA scaffolds and decoys are now implicated in processes such as mRNA translation, decay, protein localization, protein degradation and assembly of cytoplasmic biomolecular condensates such as P-bodies. Here, we review examples of RNA scaffolds and decoys in both the nucleus and cytoplasm, illustrating common themes, the suitability of RNA to such roles, and future challenges in identifying and better understanding RNA scaffolding and decoy functions.

mRNALocater: Enhance the prediction accuracy of eukaryotic mRNA subcellular localization by using model fusion strategy

The functions of mRNAs are closely correlated with their locations in cells. Knowledge about the subcellular locations of mRNA is helpful to understand their biological functions. In recent years, it has become a hot topic to develop effective computational models to predict eukaryotic mRNA subcellular localizations. However, existing state-of-the-art models still have certain deficiencies in terms of prediction accuracy and generalization ability. Therefore, it is urgent to develop novel methods to accurately predict mRNA subcellular localizations. In this study, a novel method called mRNALocater was proposed to detect the subcellular localization of eukaryotic mRNA by adopting the model fusion strategy. To fully extract information from mRNA sequences, the electron-ion interaction pseudopotential and pseudo k-tuple nucleotide composition were used to encode the sequences. Moreover, the correlation coefficient filtering algorithm and feature forward search technology were used to mine hidden feature information, which guarantees that mRNALocater can be more effectively applied to new sequences. The results based on the independent dataset tests demonstrate that mRNALocater yields promising performances for predicting eukaryotic mRNA subcellular localizations and is a powerful tool in practical applications. A freely available online web server for mRNALocater has been established at http://bio-bigdata.cn/mRNALocater.

Global analysis of expression, maturation and subcellular localization of mouse liver transcriptome identifies novel sex-biased and TCPOBOP-responsive long non-coding RNAs

BACKGROUND

While nuclear transcription and RNA processing and localization are well established for protein coding genes (PCGs), these processes are poorly understood for long non-coding (lnc)RNAs. Here, we characterize global patterns of transcript expression, maturation and localization for mouse liver RNA, including more than 15,000 lncRNAs. PolyA-selected liver RNA was isolated and sequenced from four subcellular fractions (chromatin, nucleoplasm, total nucleus, and cytoplasm), and from the chromatin-bound fraction without polyA selection.

RESULTS

Transcript processing, determined from normalized intronic to exonic sequence read density ratios, progressively increased for PCG transcripts in going from the chromatin-bound fraction to the nucleoplasm and then on to the cytoplasm. Transcript maturation was similar for lncRNAs in the chromatin fraction, but was significantly lower in the nucleoplasm and cytoplasm. LncRNA transcripts were 11-fold more likely to be significantly enriched in the nucleus than cytoplasm, and 100-fold more likely to be significantly chromatin-bound than nucleoplasmic. Sequencing chromatin-bound RNA greatly increased the sensitivity for detecting lowly expressed lncRNAs and enabled us to discover and localize hundreds of novel regulated liver lncRNAs, including lncRNAs showing sex-biased expression or responsiveness to TCPOBOP a xenobiotic agonist ligand of constitutive androstane receptor (Nr1i3).

CONCLUSIONS

Integration of our findings with prior studies and lncRNA annotations identified candidate regulatory lncRNAs for a variety of hepatic functions based on gene co-localization within topologically associating domains or transcription divergent or antisense to PCGs associated with pathways linked to hepatic physiology and disease.

lncLocator 2.0: a cell-line-specific subcellular localization predictor for long non-coding RNAs with interpretable deep learning

MOTIVATION

Long non-coding RNAs (lncRNAs) are generally expressed in a tissue-specific way, and subcellular localizations of lncRNAs depend on the tissues or cell lines that they are expressed. Previous computational methods for predicting subcellular localizations of lncRNAs do not take this characteristic into account, they train a unified machine learning model for pooled lncRNAs from all available cell lines. It is of importance to develop a cell-line-specific computational method to predict lncRNA locations in different cell lines.

RESULTS

In this study, we present an updated cell-line-specific predictor lncLocator 2.0, which trains an end-to-end deep model per cell line, for predicting lncRNA subcellular localization from sequences.We first construct benchmark datasets of lncRNA subcellular localizations for 15 cell lines. Then we learn word embeddings using natural language models, and these learned embeddings are fed into convolutional neural network, long short-term memory and multilayer perceptron to classify subcellular localizations. lncLocator 2.0 achieves varying effectiveness for different cell lines and demonstrates the necessity of training cell-line-specific models. Furthermore, we adopt Integrated Gradients to explain the proposed model in lncLocator 2.0, and find some potential patterns that determine the subcellular localizations of lncRNAs, suggesting that the subcellular localization of lncRNAs is linked to some specific nucleotides.

AVAILABILITY

The lncLocator 2.0 is available at www.csbio.sjtu.edu.cn/bioinf/lncLocator2 and the source code can be found at https://github.com/Yang-J-LIN/lncLocator2.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Identification and analysis of consensus RNA motifs binding to the genome regulator CTCF

CCCTC-binding factor (CTCF) is a key regulator of 3D genome organization and gene expression. Recent studies suggest that RNA transcripts, mostly long non-coding RNAs (lncRNAs), can serve as locus-specific factors to bind and recruit CTCF to the chromatin. However, it remains unclear whether specific sequence patterns are shared by the CTCF-binding RNA sites, and no RNA motif has been reported so far for CTCF binding. In this study, we have developed DeepLncCTCF, a new deep learning model based on a convolutional neural network and a bidirectional long short-term memory network, to discover the RNA recognition patterns of CTCF and identify candidate lncRNAs binding to CTCF. When evaluated on two different datasets, human U2OS dataset and mouse ESC dataset, DeepLncCTCF was shown to be able to accurately predict CTCF-binding RNA sites from nucleotide sequence. By examining the sequence features learned by DeepLncCTCF, we discovered a novel RNA motif with the consensus sequence, AGAUNGGA, for potential CTCF binding in humans. Furthermore, the applicability of DeepLncCTCF was demonstrated by identifying nearly 5000 candidate lncRNAs that might bind to CTCF in the nucleus. Our results provide useful information for understanding the molecular mechanisms of CTCF function in 3D genome organization.

Subcellular Localization of uc.8+ as a Prognostic Biomarker in Bladder Cancer Tissue

Simple Summary

DNA regions having high sequence similarity among human, rat and mouse genomes are defined as Ultraconserved Regions. Non-coding RNA transcripts originating by these regions may play relevant roles in the onset and progression of multiple cancer types. We recently found that ultra-conserved-transcript-8+ (uc.8+) levels correlate with the grading and staging of bladder cancer. The aim of this study is to systematically evaluate the expression of ultra-conserved-transcript-8+ (uc.8+) in biopsies and assess its intracellular localization. Furthermore, we aimed to correlate uc.8+ levels with clinical parameters and patient survival. Our analysis indicates that uc.8+ can localize both in the cytoplasm and nucleus of bladder cells at early stages of tumorigenesis, while in tumors at advanced stages, uc.8+ has a prevalent cytoplasmic localization. These data provide relevant information about uc.8+ localization as a hallmark of tumor stage. Finally, using advanced computer-based techniques, we predicted the binding of uc.8+ to RNA-binding proteins. Our study overall suggests that uc.8+ localization can be used as a prognostic biomarker for bladder cancer.

Abstract

Non-coding RNA transcripts originating from Ultraconserved Regions (UCRs) have tissue-specific expression and play relevant roles in the pathophysiology of multiple cancer types. Among them, we recently identified and characterized the ultra-conserved-transcript-8+ (uc.8+), whose levels correlate with grading and staging of bladder cancer. Here, to validate uc.8+ as a potential biomarker in bladder cancer, we assessed its expression and subcellular localization by using tissue microarray on 73 human bladder cancer specimens. We quantified uc.8+ by in-situ hybridization and correlated its expression levels with clinical characteristics and patient survival. The analysis of subcellular localization indicated the simultaneous presence of uc.8+ in the cytoplasm and nucleus of cells from the Low-Grade group, whereas a prevalent cytoplasmic localization was observed in samples from the High-Grade group, supporting the hypothesis of uc.8+ nuclear-to-cytoplasmic translocation in most malignant tumor forms. Moreover, analysis of uc.8+ expression and subcellular localization in tumor-surrounding stroma revealed a marked down-regulation of uc.8+ levels compared to the paired (adjacent) tumor region. Finally, deep machine-learning approaches identified nucleotide sequences associated with uc.8+ localization in nucleus and/or cytoplasm, allowing to predict possible RNA binding proteins associated with uc.8+, recognizing also sequences involved in mRNA cytoplasm-translocation. Our model suggests uc.8+ subcellular localization as a potential prognostic biomarker for bladder cancer.

lncRNAs in development and differentiation: from sequence motifs to functional characterization

The number of long noncoding RNAs (lncRNAs) with characterized developmental and cellular functions continues to increase, but our understanding of the molecular mechanisms underlying lncRNA functions, and how they are dictated by RNA sequences, remains limited. Relatively short, conserved sequence motifs embedded in lncRNA transcripts are often important determinants of lncRNA localization, stability and interactions. Identifying such RNA motifs remains challenging due to the substantial length of lncRNA transcripts and the rapid evolutionary turnover of lncRNA sequences. Nevertheless, the recent discovery of specific RNA elements, together with their experimental interrogation, has enabled the first step in classifying heterogeneous lncRNAs into sub-groups with similar molecular mechanisms and functions. In this Review, we focus on lncRNAs with roles in development, cell differentiation and normal physiology in vertebrates, and we discuss the sequence elements defining their functions. We also summarize progress on the discovery of regulatory RNA sequence elements, as well as their molecular functions and interaction partners.

Gene regulation by long non-coding RNAs and its biological functions

Evidence accumulated over the past decade shows that long non-coding RNAs (lncRNAs) are widely expressed and have key roles in gene regulation. Recent studies have begun to unravel how the biogenesis of lncRNAs is distinct from that of mRNAs and is linked with their specific subcellular localizations and functions. Depending on their localization and their specific interactions with DNA, RNA and proteins, lncRNAs can modulate chromatin function, regulate the assembly and function of membraneless nuclear bodies, alter the stability and translation of cytoplasmic mRNAs and interfere with signalling pathways. Many of these functions ultimately affect gene expression in diverse biological and physiopathological contexts, such as in neuronal disorders, immune responses and cancer. Tissue-specific and condition-specific expression patterns suggest that lncRNAs are potential biomarkers and provide a rationale to target them clinically. In this Review, we discuss the mechanisms of lncRNA biogenesis, localization and functions in transcriptional, post-transcriptional and other modes of gene regulation, and their potential therapeutic applications.

The lncRNA Toolkit: Databases and In Silico Tools for lncRNA Analysis

Long non-coding RNAs (lncRNAs) are a rapidly expanding field of research, with many new transcripts identified each year. However, only a small subset of lncRNAs has been characterized functionally thus far. To aid investigating the mechanisms of action by which new lncRNAs act, bioinformatic tools and databases are invaluable. Here, we review a selection of computational tools and databases for the in silico analysis of lncRNAs, including tissue-specific expression, protein coding potential, subcellular localization, structural conformation, and interaction partners. The assembled lncRNA toolkit is aimed primarily at experimental researchers as a useful starting point to guide wet-lab experiments, mainly containing multi-functional, user-friendly interfaces. With more and more new lncRNA analysis tools available, it will be essential to provide continuous updates and maintain the availability of key software in the future.

Discoveries for Long Non-Coding RNA Dynamics in Traumatic Brain Injury

Simple Summary

The biomedical studies of traumatic brain injury (TBI) can lead to insight for treatment clinically. However, TBIs are occurred by various risk factors and showing heterogeneity that make difficult to accurate diagnosis for initiation treatment of patients. Therefore, identification of biomarkers requires to prediction and therapeutics for TBI treatment. The canonical function of the long non-coding RNAs (lncRNAs) have been recently shown to promote transcription, post-transcription, and protein activity in many different conditions. Therefore, understanding the molecular mechanisms that are altered by the expression of lncRNAs will allow the design of novel therapeutic strategies. Here, we review the molecular process of lncRNA as new targets and approaches in TBIs treatment.

Abstract

In recent years, our understanding of long non-coding RNAs (lncRNAs) has been challenged with advances in genome sequencing and the widespread use of high-throughput analysis for identifying novel lncRNAs. Since then, the characterization of lncRNAs has contributed to the establishment of their molecular roles and functions in transcriptional regulation. Although genetic studies have so far explored the sequence-based primary function of lncRNAs that guides the expression of target genes, recent insights have shed light on the potential of lncRNAs for widening the identification of biomarkers from non-degenerative to neurodegenerative diseases. Therefore, further advances in the genetic characteristics of lncRNAs are expected to lead to diagnostic accuracy during disease progression. In this review, we summarized the latest studies of lncRNAs in TBI as a non-degenerative disease and discussed their potential limitations for clinical treatment.

MirLocPredictor: A ConvNet-Based Multi-Label MicroRNA Subcellular Localization Predictor by Incorporating k-Mer Positional Information

MicroRNAs (miRNA) are small noncoding RNA sequences consisting of about 22 nucleotides that are involved in the regulation of almost 60% of mammalian genes. Presently, there are very limited approaches for the visualization of miRNA locations present inside cells to support the elucidation of pathways and mechanisms behind miRNA function, transport, and biogenesis. MIRLocator, a state-of-the-art tool for the prediction of subcellular localization of miRNAs makes use of a sequence-to-sequence model along with pretrained k-mer embeddings. Existing pretrained k-mer embedding generation methodologies focus on the extraction of semantics of k-mers. However, in RNA sequences, positional information of nucleotides is more important because distinct positions of the four nucleotides define the function of an RNA molecule. Considering the importance of the nucleotide position, we propose a novel approach (kmerPR2vec) which is a fusion of positional information of k-mers with randomly initialized neural k-mer embeddings. In contrast to existing k-mer-based representation, the proposed kmerPR2vec representation is much more rich in terms of semantic information and has more discriminative power. Using novel kmerPR2vec representation, we further present an end-to-end system (MirLocPredictor) which couples the discriminative power of kmerPR2vec with Convolutional Neural Networks (CNNs) for miRNA subcellular location prediction. The effectiveness of the proposed kmerPR2vec approach is evaluated with deep learning-based topologies (i.e., Convolutional Neural Networks (CNN) and Recurrent Neural Network (RNN)) and by using 9 different evaluation measures. Analysis of the results reveals that MirLocPredictor outperform state-of-the-art methods with a significant margin of 18% and 19% in terms of precision and recall.

Intronic Determinants Coordinate Charme lncRNA Nuclear Activity through the Interaction with MATR3 and PTBP1

Chromatin architect of muscle expression (Charme) is a muscle-restricted long noncoding RNA (lncRNA) that plays an important role in myogenesis. Earlier evidence indicates that the nuclear Charme isoform, named pCharme, acts on the chromatin by assisting the formation of chromatin domains where myogenic transcription occurs. By combining RNA antisense purification (RAP) with mass spectrometry and loss-of-function analyses, we have now identified the proteins that assist these chromatin activities. These proteins—which include a sub-set of splicing regulators, principally PTBP1 and the multifunctional RNA/DNA binding protein MATR3—bind to sequences located within the alternatively spliced intron-1 to form nuclear aggregates. Consistent with the functional importance of pCharme interactome in vivo, a targeted deletion of the intron-1 by a CRISPR-Cas9 approach in mouse causes the release of pCharme from the chromatin and results in cardiac defects similar to what was observed upon knockout of the full-length transcript.

•

pCharme is the chromatin-retained isoform of the muscle-specific Charme lncRNA

•

Intronic signals coordinate the association of pCharme with MATR3 and PTBP1

•

The particle assembly prompts pCharme intron-1 chromatin retention

•

Deletion of the intron-1 by CRISPR-Cas9 leads to heart defects in mouse

Deep Learning in LncRNAome: Contribution, Challenges, and Perspectives

Long non-coding RNAs (lncRNA), the pervasively transcribed part of the mammalian genome, have played a significant role in changing our protein-centric view of genomes. The abundance of lncRNAs and their diverse roles across cell types have opened numerous avenues for the research community regarding lncRNAome. To discover and understand lncRNAome, many sophisticated computational techniques have been leveraged. Recently, deep learning (DL)-based modeling techniques have been successfully used in genomics due to their capacity to handle large amounts of data and produce relatively better results than traditional machine learning (ML) models. DL-based modeling techniques have now become a choice for many modeling tasks in the field of lncRNAome as well. In this review article, we summarized the contribution of DL-based methods in nine different lncRNAome research areas. We also outlined DL-based techniques leveraged in lncRNAome, highlighting the challenges computational scientists face while developing DL-based models for lncRNAome. To the best of our knowledge, this is the first review article that summarizes the role of DL-based techniques in multiple areas of lncRNAome.

Emerging Roles of Long Noncoding RNAs in the Cytoplasmic Milieu

While the important functions of long noncoding RNAs (lncRNAs) in nuclear organization are well documented, their orchestrating and architectural roles in the cytoplasmic environment have long been underestimated. However, recently developed fractionation and proximity labelling approaches have shown that a considerable proportion of cellular lncRNAs is exported into the cytoplasm and associates nonrandomly with proteins in the cytosol and organelles. The functions of these lncRNAs range from the control of translation and mitochondrial metabolism to the anchoring of cellular components on the cytoskeleton and regulation of protein degradation at the proteasome. In the present review, we provide an overview of the functions of lncRNAs in cytoplasmic structures and machineries und discuss their emerging roles in the coordination of the dense intracellular milieu. It is becoming apparent that further research into the functions of these lncRNAs will lead to an improved understanding of the spatiotemporal organization of cytoplasmic processes during homeostasis and disease.

Long non-coding RNAs in ovarian cancer: expression profile and functional spectrum

Long non-coding RNAs (lncRNAs), initially recognized as byproducts of the transcription process, have been proven to play crucial modulatory roles in preserving overall homoeostasis of cells and tissues. Furthermore, aberrant levels of these transcripts have been shown to contribute many diseases, including cancer. Among these, many aspects of ovarian cancer biology have been found to be regulated by lncRNAs, including cancer initiation, progression and dissemination. In this review, we summarize recent studies to highlight the various roles of lncRNAs in ovary in normal and pathological conditions, immune system, diagnosis, prognosis, and therapy. We address lncRNAs that have been extensively studied in ovarian cancer and their contribution to cellular dynamics.

LncRNA SNHG11 facilitates tumor metastasis by interacting with and stabilizing HIF-1α

Epigenetic alteration is one of the hallmarks of colorectal cancer (CRC). Many driver genes are regulated by DNA methylation in CRC. However, the role of DNA methylation regulating lncRNAs remain elusive. Here, we identify that SNHG11 (small nucleolar RNA host gene 11) is upregulated by promotor hypomethylation in CRC and is associated with poor prognosis in CRC patients. SNHG11 can promote CRC cell migration and metastasis under hypoxia. Interestingly, the DNA-binding motif of SNHG11 is similar to that of HIF-1α. In addition, SNHG11-associated genes are enriched with members of the HIF-1 signaling pathway in CRC. Mechanistically, SNHG11 binds to the pVHLrecognition sites on HIF-1α, thus blocking the interaction of pVHL with HIF-1α and preventing its ubiquitination and degradation. Moreover, SNHG11 upregulates the expression of HIF-1α target genes, i.e., AK4, ENO1, HK2, and Twist1. Notably, SNHG11 can bind to the HRE sites in the promoter of these genes and increase their transcription. In summary, these results identify a SNHG11/ HIF-1α axis that plays a pivotal role in tumor invasion and metastasis.

Competing Endogenous RNA Networks in the Epithelial to Mesenchymal Transition in Diffuse-Type of Gastric Cancer

Simple Summary

The diffuse-type of gastric cancer is associated with epithelial to mesenchymal transition. Loss of E-cadherin expression is the hallmark of this process and is largely due to the upregulation of the transcription factors ZEB1/2, Snail, Slug, and Twist1/2. However, miRNA and lncRNAs can also participate through these transcription factors which directly target E-cadherin. The competing endogenous RNA (ceRNA) network hypothesis state that lncRNA can sponge the miRNA pool that targets these transcripts. Based on the lack of said networks in the epithelial to mesenchymal transition, we performed a prediction analysis that resulted in novel ceRNA networks which will expand our knowledge of the molecular basis of the diffuse-type of gastric cancer.

Abstract

The diffuse-type of gastric cancer (DGC), molecularly associated with epithelial to mesenchymal transition (EMT), is increasing in incidence. Loss of E-cadherin expression is the hallmark of the EMT process and is largely due to the upregulation of the EMT-inducing transcription factors ZEB1/2, Snail, Slug, and Twist1/2. However, ncRNA, such as miRNA and lncRNAs, can also participate in the EMT process through the direct targeting of E-cadherin and other EMT-inducing transcription factors. Additionally, lncRNA can sponge the miRNA pool that targets these transcripts through competing endogenous RNA (ceRNA) networks. In this review, we focus on the role of ncRNA in the direct deregulation of E-cadherin, as well as EMT-inducing transcription factors. Based on the relevance of the ceRNA network hypothesis, and the lack of said networks in EMT, we performed a prediction analysis for all miRNAs and lncRNAs that target E-cadherin, as well as EMT-inducing transcription factors. This analysis resulted in novel predicted ceRNA networks for E-cadherin and EMT-inducing transcription factors (EMT-TFs), as well as the expansion of the molecular basis of the DGC.

Back to the Future: Rethinking the Great Potential of lncRNAS for Optimizing Chemotherapeutic Response in Ovarian Cancer

Ovarian cancer (OC) is one of the most fatal cancers in women worldwide. Currently, platinum- and taxane-based chemotherapy is the mainstay for the treatment of OC. Yet, the emergence of chemoresistance results in therapeutic failure and significant relapse despite a consistent rate of primary response. Emerging evidence substantiates the potential role of lncRNAs in determining the response to standard chemotherapy in OC. The objective of this narrative review is to provide an integrated, synthesized overview of the current state of knowledge regarding the role of lncRNAs in the emergence of resistance to platinum- and taxane-based chemotherapy in OC. In addition, we sought to develop conceptual frameworks for harnessing the therapeutic potential of lncRNAs in strategies aimed at enhancing the chemotherapy response of OC. Furthermore, we offered significant new perspectives and insights on the interplay between lncRNAs and the molecular circuitries implicated in chemoresistance to determine their impacts on therapeutic response. Although this review summarizes robust data concerning the involvement of lncRNAs in the emergence of acquired resistance to platinum- and taxane-based chemotherapy in OC, effective approaches for translating these lncRNAs into clinical practice warrant further investigation.

Comprehensive Analysis of LincRNAs in Classical and Basal-Like Subtypes of Pancreatic Cancer

Pancreatic ductal adenocarcinomas (PDAC) belong to the deadliest malignancies in the western world. Mutations in TP53 and KRAS genes along with some other frequent polymorphisms occur almost universally and are major drivers of tumour initiation. However, these mutations cannot explain the heterogeneity in therapeutic responses and differences in overall survival observed in PDAC patients. Thus, recent classifications of PDAC tumour samples have leveraged transcriptome-wide gene expression data to account for epigenetic, transcriptional and post-transcriptional mechanisms that may contribute to this deadly disease. Intriguingly, long intervening RNAs (lincRNAs) are a special class of long non-coding RNAs (lncRNAs) that can control gene expression programs on multiple levels thereby contributing to cancer progression. However, their subtype-specific expression and function as well as molecular interactions in PDAC are not fully understood yet. In this study, we systematically investigated the expression of lincRNAs in pancreatic cancer and its molecular subtypes using publicly available data from large-scale studies. We identified 27 deregulated lincRNAs that showed a significant different expression pattern in PDAC subtypes suggesting context-dependent roles. We further analyzed these lincRNAs regarding their common expression patterns. Moreover, we inferred clues on their functions based on correlation analyses and predicted interactions with RNA-binding proteins, microRNAs, and mRNAs. In summary, we identified several PDAC-associated lincRNAs of prognostic relevance and potential context-dependent functions and molecular interactions. Hence, our study provides a valuable resource for future investigations to decipher the role of lincRNAs in pancreatic cancer.

CARMAL Is a Long Non-coding RNA Locus That Regulates MFGE8 Expression

Genome-wide association studies have identified several genetic loci linked to coronary artery disease (CAD) most of them located in non-protein coding regions of the genome. One such locus is the CAD Associated Region between MFGE8 and ABHD2 (CARMA), a ∼18 kb haplotype that was recently shown to regulate vicinal protein coding genes. Here, we further investigate the region by examining a long non-coding RNA gene locus (CARMAL/RP11-326A19.4/AC013565) abutting the CARMA region. Expression-genotype correlation analyses of public databases indicate that CARMAL levels are influenced by CAD associated variants suggesting that it might have cardioprotective functions. We found CARMAL to be stably expressed at relatively low levels and enriched in the cytosol. CARMAL function was investigated by several gene targeting approaches in HEK293T: inactive CRISPR fusion proteins, antisense, overexpression and inactivation by CRISPR-mediated knock-out. Modest increases in CARMAL (3–4×) obtained via CRISPRa using distinct single-guided RNAs did not result in consistent transcriptome effects. By contrast, CARMAL deletion or reduced CARMAL expression via CRISPRi increased MFGE8 levels, suggesting that CARMAL is contributing to reduce MFGE8 expression under basal conditions. While future investigations are required to clarify the mechanism(s) by which CARMAL acts on MFGE8, integrative bioinformatic analyses of the transcriptome of CARMAL deleted cells suggest that this locus may also be involved in leucine metabolism, splicing, transcriptional regulation and Shwachman-Bodian-Diamond syndrome protein function.

RNA-GPS predicts high-resolution RNA subcellular localization and highlights the role of splicing

Subcellular localization is essential to RNA biogenesis, processing, and function across the gene expression life cycle. However, the specific nucleotide sequence motifs that direct RNA localization are incompletely understood. Fortunately, new sequencing technologies have provided transcriptome-wide atlases of RNA localization, creating an opportunity to leverage computational modeling. Here we present RNA-GPS, a new machine learning model that uses nucleotide-level features to predict RNA localization across eight different subcellular locations-the first to provide such a wide range of predictions. RNA-GPS's design enables high-throughput sequence ablation and feature importance analyses to probe the sequence motifs that drive localization prediction. We find localization informative motifs to be concentrated on 3'-UTRs and scattered along the coding sequence, and motifs related to splicing to be important drivers of predicted localization, even for cytotopic distinctions for membraneless bodies within the nucleus or for organelles within the cytoplasm. Overall, our results suggest transcript splicing is one of many elements influencing RNA subcellular localization.

Prediction of mRNA subcellular localization using deep recurrent neural networks

MOTIVATION

Messenger RNA subcellular localization mechanisms play a crucial role in post-transcriptional gene regulation. This trafficking is mediated by trans-acting RNA-binding proteins interacting with cis-regulatory elements called zipcodes. While new sequencing-based technologies allow the high-throughput identification of RNAs localized to specific subcellular compartments, the precise mechanisms at play, and their dependency on specific sequence elements, remain poorly understood.

RESULTS

We introduce RNATracker, a novel deep neural network built to predict, from their sequence alone, the distributions of mRNA transcripts over a predefined set of subcellular compartments. RNATracker integrates several state-of-the-art deep learning techniques (e.g. CNN, LSTM and attention layers) and can make use of both sequence and secondary structure information. We report on a variety of evaluations showing RNATracker's strong predictive power, which is significantly superior to a variety of baseline predictors. Despite its complexity, several aspects of the model can be isolated to yield valuable, testable mechanistic hypotheses, and to locate candidate zipcode sequences within transcripts.

AVAILABILITY AND IMPLEMENTATION

Code and data can be accessed at https://www.github.com/HarveyYan/RNATracker.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Long non-coding RNAs: Key regulators in oxaliplatin resistance of colorectal cancer

Colorectal cancer (CRC) is one of the most commonly diagnosed malignancies in the world with high relapse and mortality rates. Although oxaliplatin (OXA), a platinum-based anticancer drug, is widely used in CRC treatment, the resulting chemoresistance dramatically attenuates the drug efficacy and increases the failure rate of this therapy. Thus, the study on OXA-induced chemoresistance is extremely urgent. In recent years, emerging evidence has shown that lncRNAs play irreplaceable roles in drug resistance. However, we only have a limited knowledge of the lncRNAs that are closely related to oxaliplatin resistance in CRC. In present study, we identify and characterize these lncRNAs, including their functions, underlying mechanisms and possible applications.

Subcellular Spatial Transcriptomes: Emerging Frontier for Understanding Gene Regulation

RNAs are trafficked and localized with exquisite precision inside the cell. Studies of candidate messenger RNAs have shown the vital importance of RNA subcellular location in development and cellular function. New sequencing- and imaging-based methods are providing complementary insights into subcellular localization of RNAs transcriptome-wide. APEX-seq and ribosome profiling as well as proximity-labeling approaches have revealed thousands of transcript isoforms are localized to distinct cytotopic locations, including locations that defy biochemical fractionation and hence were missed by prior studies. Sequences in the 3' and 5' untranslated regions (UTRs) serve as "zip codes" to direct transcripts to particular locales, and it is clear that intronic and retrotransposable sequences within transcripts have been co-opted by cells to control localization. Molecular motors, nuclear-to-cytosol RNA export, liquid-liquid phase separation, RNA modifications, and RNA structure dynamically shape the subcellular transcriptome. Location-based RNA regulation continues to pose new mysteries for the field, yet promises to reveal insights into fundamental cell biology and disease mechanisms.

Localization elements and zip codes in the intracellular transport and localization of messenger RNAs in Saccharomyces cerevisiae

Intracellular trafficking and localization of mRNAs provide a mechanism of regulation of expression of genes with excellent spatial control. mRNA localization followed by localized translation appears to be a mechanism of targeted protein sorting to a specific cell-compartment, which is linked to the establishment of cell polarity, cell asymmetry, embryonic axis determination, and neuronal plasticity in metazoans. However, the complexity of the mechanism and the components of mRNA localization in higher organisms prompted the use of the unicellular organism Saccharomyces cerevisiae as a simplified model organism to study this vital process. Current knowledge indicates that a variety of mRNAs are asymmetrically and selectively localized to the tip of the bud of the daughter cells, to the vicinity of endoplasmic reticulum, mitochondria, and nucleus in this organism, which are connected to diverse cellular processes. Interestingly, specific cis-acting RNA localization elements (LEs) or RNA zip codes play a crucial role in the localization and trafficking of these localized mRNAs by providing critical binding sites for the specific RNA-binding proteins (RBPs). In this review, we present a comprehensive account of mRNA localization in S. cerevisiae, various types of localization elements influencing the mRNA localization, and the RBPs, which bind to these LEs to implement a number of vital physiological processes. Finally, we emphasize the significance of this process by highlighting their connection to several neuropathological disorders and cancers. This article is categorized under: RNA Export and Localization > RNA Localization.

Genomic data mining for functional annotation of human long noncoding RNAs

Life may have begun in an RNA world, which is supported by increasing evidence of the vital role that RNAs perform in biological systems. In the human genome, most genes actually do not encode proteins; they are noncoding RNA genes. The largest class of noncoding genes is known as long noncoding RNAs (lncRNAs), which are transcripts greater in length than 200 nucleotides, but with no protein-coding capacity. While some lncRNAs have been demonstrated to be key regulators of gene expression and 3D genome organization, most lncRNAs are still uncharacterized. We thus propose several data mining and machine learning approaches for the functional annotation of human lncRNAs by leveraging the vast amount of data from genetic and genomic studies. Recent results from our studies and those of other groups indicate that genomic data mining can give insights into lncRNA functions and provide valuable information for experimental studies of candidate lncRNAs associated with human disease.

Nuclear sorting of RNA

The majority of the mammalian genome is transcribed by RNA polymerase II, yielding a vast amount of noncoding RNA (ncRNA) in addition to the standard production of mRNA. The typical nuclear biogenesis of mRNA relies on the tightly controlled coupling of co- and post-transcriptional processing events, which ultimately results in the export of transcripts into the cytoplasm. These processes are subject to surveillance by nuclear RNA decay pathways to prevent the export of aberrant, or otherwise "non-optimal," transcripts. However, unlike mRNA, many long ncRNAs are nuclear retained and those that maintain enduring functions must employ precautions to evade decay. Proper sorting and localization of RNA is therefore an essential activity in eukaryotic cells and the formation of ribonucleoprotein complexes during early stages of RNA synthesis is central to deciding such transcript fate. This review details our current understanding of the pathways and factors that direct RNAs towards a particular destiny and how transcripts combat the adverse conditions of the nucleus. This article is categorized under: RNA Export and Localization > Nuclear Export/Import RNA Turnover and Surveillance > Turnover/Surveillance Mechanisms RNA Interactions with Proteins and Other Molecules > Protein-RNA Interactions: Functional Implications.

Diagnostic and prognostic potential of tissue and circulating long non-coding RNAs in colorectal tumors

Long non-coding RNAs (lncRNAs) are members of the non-protein coding RNA family longer than 200 nucleotides. They participate in the regulation of gene and protein expression influencing apoptosis, cell proliferation and immune responses, thereby playing a critical role in the development and progression of various cancers, including colorectal cancer (CRC). As CRC is one of the most frequently diagnosed malignancies worldwide with high mortality, its screening and early detection are crucial, so the identification of disease-specific biomarkers is necessary. LncRNAs are promising candidates as they are involved in carcinogenesis, and certain lncRNAs (e.g., CCAT1, CRNDE, CRCAL1-4) show altered expression in adenomas, making them potential early diagnostic markers. In addition to being useful as tissue-specific markers, analysis of circulating lncRNAs (e.g., CCAT1, CCAT2, BLACAT1, CRNDE, NEAT1, UCA1) in peripheral blood offers the possibility to establish minimally invasive, liquid biopsy-based diagnostic tests. This review article aims to describe the origin, structure, and functions of lncRNAs and to discuss their contribution to CRC development. Moreover, our purpose is to summarise lncRNAs showing altered expression levels during tumor formation in both colon tissue and plasma/serum samples and to demonstrate their clinical implications as diagnostic or prognostic biomarkers for CRC.

Nuclear retention element recruits U1 snRNP components to restrain spliced lncRNAs in the nucleus

In contrast to cytoplasmic localization of spliced mRNAs, many spliced lncRNAs are localized in the nucleus. To investigate the mechanism, we used lncRNA MEG3 as a reporter and mapped a potent nuclear retention element (NRE), deletion of this element led to striking export of MEG3 from the nucleus to the cytoplasm. Insertion of the NRE resulted in nuclear retention of spliced lncRNA as well as spliced mRNA. We further purified RNP assembled on the NRE in vitro and identified the proteins by mass spectrometry. Screen using siRNA revealed depletion of U1 snRNP components SNRPA, SNRNP70 or SNRPD2 caused significant cytoplasmic localization of MEG3 reporter transcripts. Co-knockdown these factors in HFF1 cells resulted in an increased cytoplasmic distribution of endogenous lncRNAs. Together, these data support a model that U1 snRNP components restrain spliced lncRNAs in the nucleus via the interaction with nuclear retention element.

Predictive models of subcellular localization of long RNAs

Export to the cytoplasm is a key regulatory junction for both protein-coding mRNAs and long noncoding RNAs (lncRNAs), and cytoplasmic enrichment varies dramatically both within and between those groups. We used a new computational approach and RNA-seq data from human and mouse cells to quantify the genome-wide association between cytoplasmic/nuclear ratios of both gene groups and various factors, including expression levels, splicing efficiency, gene architecture, chromatin marks, and sequence elements. Splicing efficiency emerged as the main predictive factor, explaining up to a third of the variability in localization. Combination with other features allowed predictive models that could explain up to 45% of the variance for protein-coding genes and up to 34% for lncRNAs. Factors associated with localization were similar between lncRNAs and mRNAs with some important differences. Readily accessible features can thus be used to predict RNA localization.

Bioinformatics Approaches to Gain Insights into cis-Regulatory Motifs Involved in mRNA Localization

Messenger RNA (mRNA) is a fundamental intermediate in the expression of proteins. As an integral part of this important process, protein production can be localized by the targeting of mRNA to a specific subcellular compartment. The subcellular destination of mRNA is suggested to be governed by a region of its primary sequence or secondary structure, which consequently dictates the recruitment of trans-acting factors, such as RNA-binding proteins or regulatory RNAs, to form a messenger ribonucleoprotein particle. This molecular ensemble is requisite for precise and spatiotemporal control of gene expression. In the context of RNA localization, the description of the binding preferences of an RNA-binding protein defines a motif, and one, or more, instance of a given motif is defined as a localization element (zip code). In this chapter, we first discuss the cis-regulatory motifs previously identified as mRNA localization elements. We then describe motif representation in terms of entropy and information content and offer an overview of motif databases and search algorithms. Finally, we provide an outline of the motif topology of asymmetrically localized mRNA molecules.