Biological features between miRNAs and their targets are unveiled from deep learning models

MicroRNAs (miRNAs) are ~ 22 nucleotide ubiquitous gene regulators. They modulate a broad range of essential cellular processes linked to human health and diseases. Consequently, identifying miRNA targets and understanding how they function are critical for treating miRNA associated diseases. In our earlier work, a hybrid deep learning-based approach (miTAR) was developed for predicting miRNA targets. It performs substantially better than the existing methods. The approach integrates two major types of deep learning algorithms: convolutional neural networks (CNNs) and recurrent neural networks (RNNs). However, the features in miRNA:target interactions learned by miTAR have not been investigated. In the current study, we demonstrated that miTAR captures known features, including the involvement of seed region and the free energy, as well as multiple novel features, in the miRNA:target interactions. Interestingly, the CNN and RNN layers of the model perform differently at capturing the free energy feature: the units in RNN layer is more unique at capturing the feature but collectively the CNN layer is more efficient at capturing the feature. Although deep learning models are commonly thought “black-boxes”, our discoveries support that the biological features in miRNA:target can be unveiled from deep learning models, which will be beneficial to the understanding of the mechanisms in miRNA:target interactions.

Profiling and Characterization of microRNAs Responding to Sodium Butyrate Treatment in Gastric Cancer Cells

Background: Short-chain fatty acids exert anti-cancer effects on tumor cells. Objective: We aimed to reveal the signaling network altered by butyrate in Gastric Cancer (GC) using small RNA sequencing (sRNA-seq). Methods: The effects of butyrate on the biological behavior of NCI-N87 and KATO III cells in vitro were assessed by functional assays and half-maximal inhibitory concentrations (IC50) of butyrate in KATO III cells were calculated. sRNA-seq was performed on KATO III cells. Differentially expressed miRNAs (DE-miRNAs) were identified between butyrate treatment and control groups using DESeq2, and miRNA targets were predicted. A protein-protein interaction (PPI) network of DE-miRNA targets was created using Metascape. Key MCODE complexes were identified using the MCODE algorithm and cluster Profiler. The relationship between DE-miRNA and GC overall survival (OS) was evaluated using Kaplan-Meier curves. Results: Butyrate dose-dependently inhibited NCI-N87 and KATO III cell viability. KATO III cells were more sensitive to butyrate than NCI-N87 cells. Butyrate promoted apoptosis and inhibited KATO III cell migration. Total 324 DE-miRNAs were identified in KATO III cells, and 459 mRNAs were predicted as targets of 83 DE-miRNAs. Two key protein complexes were identified in a PPI network of the 459 targets. A key signaling network responding to butyrate were generated using targets in these key complexes and their miRNA regulators. The DE-miRNAs in the key signaling network were related to the OS of GC. Conclusion: Butyrate altered the biological behavior of GC cells, which may be achieved by regulating miRNAs and related oncogenic pathways.

The miRNA Profile of Inflammatory Colorectal Tumors Identify TGF-β as a Companion Target for Checkpoint Blockade Immunotherapy

Extrinsic factors such as expression of PD-L1 (programmed dealth-ligand 1) in the tumor microenvironment (TME) have been shown to correlate with responses to checkpoint blockade therapy. More recently two intrinsic factors related to tumor genetics, microsatellite instability (MSI), and tumor mutation burden (TMB), have been linked to high response rates to checkpoint blockade drugs. These response rates led to the first tissue-agnostic approval of any cancer therapy by the FDA for the treatment of metastatic, MSI-H tumors with anti-PD-1 immunotherapy. But there are still very few studies focusing on the association of miRNAs with immune therapy through checkpoint inhibitors. Our team sought to explore the biology of such tumors further and suggest potential companion therapeutics to current checkpoint inhibitors. Analysis by Pearson Correlation revealed 41 total miRNAs correlated with mutation burden, 62 miRNAs correlated with MSI, and 17 miRNAs correlated with PD-L1 expression. Three miRNAs were correlated with all three of these tumor features as well as M1 macrophage polarization. No miRNAs in any group were associated with overall survival. TGF-β was predicted to be influenced by these three miRNAs (p = 0.008). Exploring miRNA targets as companions to treatment by immune checkpoint blockade revealed three potential miRNA targets predicted to impact TGF-β. M1 macrophage polarization state was also associated with tumors predicted to respond to therapy by immune checkpoint blockade.

1197Identification of insulin signaling pathway-related circulating circRNAs as novel biomarkers of type 2 diabetes

Background The underlying molecular mechanism of type 2 diabetes (T2D) and insulin resistance is that abnormalities occur in the complex insulin signaling pathway. Circular RNAs (circRNAs) are involved in the development of diseases by regulating gene expression and become promising novel biomarkers for diseases. This study screened and validated the insulin signaling pathway-related circulating circRNAs, which are associated with T2D. Methods Based on circRNA microarray, candidate circRNAs involved in the insulin PI3K/Akt signaling pathway were selected and validated by RT-qPCR. The association between circRNAs and T2D and their clinical significance were further assessed by logistic regression model, correlation analysis and ROC curve in a large cohort. The miRNA targets of validated circRNAs was verified by dual-luciferase reporter assay. Results A total of 370 upregulated circRNAs and 180 downregulated circRNAs were differentially expressed between new T2D cases and controls. hsa_circ_0063425, hsa_circ_0056891 and hsa_circ_0104123 were selected as candidate circRNAs for validation. Low expressed circ_0063425 and hsa_circ_0056891 were independent predictors of T2D, impaired fasting glucose (IFG) and insulin resistance. The two-circRNA panel had a high diagnostic accuracy for discriminating T2D and IFG from healthy controls. miR-19a-3p and miR-1-3p were identified as the miRNA targets of hsa_circ_0063425 and hsa_circ_0056891, respectively. Significantly positive correlations were found between the expression levels of AKT and hsa_circ_0063425, PI3K and hsa_circ_0056891, in the total sample and subgroups stratified by glucose levels. Conclusion hsa_circ_0063425 and hsa_circ_0056891 are valuable circulating biomarkers for early detection of T2D, which may be involved in regulation of PI3K/AKT signaling. Key messages Insulin signaling pathway-related circulating circRNAs was identification as novel biomarkers of type 2 diabetes. Keywords circRNA; type 2 diabetes; insulin signaling; biomarker.

The regulatory roles of microRNAs toward pathogenesis and treatments in Huntington's disease

Huntington’s disease (HD) is one of neurodegenerative diseases, and is defined as a monogenetic disease due to the mutation of Huntingtin gene. This disease affects several cellular functions in neurons, and further influences motor and cognitive ability, leading to the suffering of devastating symptoms in HD patients. MicroRNA (miRNA) is a non-coding RNA, and is responsible for gene regulation at post-transcriptional levels in cells. Since one miRNA targets to several downstream genes, it may regulate different pathways simultaneously. As a result, it raises a potential therapy for different diseases using miRNAs, especially for inherited diseases. In this review, we will not only introduce the update information of HD and miRNA, but also discuss the development of potential miRNA-based therapy in HD. With the understanding toward the progression of miRNA studies in HD, we anticipate it may provide an insight to treat this devastating disease, even applying to other genetic diseases.

Biological features between miRNA and their targets are unveiled from deep learning models

MicroRNAs (miRNAs) are ~22 nucleotide ubiquitous gene regulators. They modulate a broad range of essential cellular processes linked to human health and diseases. Consequently, identifying miRNA targets and understanding how they function are critical for treating miRNA associated diseases. In our earlier work, we developed a hybrid deep learning-based approach (miTAR) for predicting miRNA targets at significantly higher accuracy compared to existing methods. It integrates two major types of deep learning algorithms: convolutional neural networks (CNNs) and recurrent neural networks (RNNs). However, the features in miRNA:target interactions learned by miTAR have not been investigated. In the current study, we demonstrated that miTAR captures known features, including the involvement of seed region and the free energy, as well as multiple novel features, in the miRNA:target interactions. Interestingly, the CNN and RNN layers of the model behave differently at capturing the free energy feature: the feature captured by the CNN layer units, but not the RNN layer units, is overlapped within and across feature maps. Although deep learning models are commonly thought “black-boxes”, our discoveries support that the biological features in miRNA:target can be unveiled from deep learning models, which will be beneficial to the understanding of the mechanisms in miRNA:target interactions.

Recent development of bioinformatics tools for microRNA target prediction

: MicroRNAs (miRNAs) are central players that regulate the post-transcriptional processes of gene expression. Binding of miRNAs to target mRNAs can repress their translation by inducing the degradation or by inhibiting the translation of the target mRNAs. High-throughput experimental approaches for miRNA target identification are costly and time-consuming, depending on various factors. It is vitally important to develop the bioinformatics methods for accurately predicting miRNA targets. With the increase of RNA sequences in the post-genomic era, bioinformatics methods are being developed for miRNA studies specially for miRNA target prediction. This review summarizes the current development of state-of-the-art bioinformatics tools for miRNA target prediction, points out the progress and limitations of the available miRNA databases, and their working principles. Finally, we discuss the caveat and perspectives of the next-generation algorithms for the prediction of miRNA targets.

Hsa-miR-4277 Decelerates the Metabolism or Clearance of Sorafenib in HCC Cells and Enhances the Sensitivity of HCC Cells to Sorafenib by Targeting cyp3a4

Increasing evidence has shown that the metabolism and clearance of molecular targeted agents, such as sorafenib, plays an important role in mediating the resistance of HCC cells to these agents. Metabolism of sorafenib is performed by oxidative metabolism, which is initially mediated by CYP3A4. Thus, targeting CYP3A4 is a promising approach to enhance the sensitivity of HCC cells to chemotherapeutic agents. In the present work, we examined the association between CYP3A4 and the prognosis of HCC patients receiving sorafenib. Using the online tool miRDB, we predicted that has-microRNA-4277 (miR-4277), an online miRNA targets the 3’UTR of the transcript of cyp3a4. Furthermore, overexpression of miR-4277 in HCC cells repressed the expression of CYP3A4 and reduced the elimination of sorafenib in HCC cells. Moreover, miR-4277 enhanced the sensitivity of HCC cells to sorafenib in vitro and in vivo. Therefore, our results not only expand our understanding of CYP3A4 regulation in HCC, but also provide evidence for the use of miR-4277 as a potential therapeutic in advanced HCC.

BTN3A2 Expression Is Connected With Favorable Prognosis and High Infiltrating Immune in Lung Adenocarcinoma

Background: butyrophilin subfamily 3 member A2 (BTN3A2) as a important mediator in immune activation, its reported to be linked to many cancers progress. However, the relation with infiltrating immune and prognosis of BTN3A2 in Lung adenocarcinoma are not clear.Methods: In our study, we checked the mRNA expression and protein expression profile of BTN3A2 in Lung adenocarcinoma (LUAD) and its relation to clinical outcomes using TIMER, UALCAN database. Besides, we analyzed the survival of BTN3A2 in LUAD by Kaplan-Meier Plotter and PrognoScan database. Moreover, we analyzed Gene Set Enrichment Analysis (GSEA) of the BTN3A2. Next, we explored the relation of BTN3A2 expression with the immune infiltration by TIMER. At last, In order to enrich the regulatory mechanism of BTN3A2, we used miRarbase, starbase and miRDB dadabase to look for miRNAs targets of BTN3A2.Results: the mRNA along with protein expression of BTN3A2 in the LUAD group was lower in contrast with that in the normal group. Besides, high BTN3A2 expression was connected with good first progression (FP) and overall survival (OS) in LUAD. Then, the GSEA analysis demonstrated that T cell receptor signaling cascade, B cell receptor signaling cascade, natural killer cell mediated cytotoxicity, immune receptor activity, immunological synapse, T cell activation were enriched differentially in BTN3A2 high expression phenotype of LUAD. Moreover, BTN3A2 expression is remarkable positively correlation with invading levels of tumor purity, B cells, neutrophil, CD4+ T cells, dendritic cells, macrophages, and CD8+ T cells in LUAD, and B cells and dendritic cells were linked with good prognosis of LUAD. To further enrich the possible regulatory mechanisms of BTN3A2, we analyzed the miRNA targets. The results showed that hsa-miR-17-5p may be miRNA targets of BTN3A2.Conclusion: Taking together, we provide evidence of BTN3A2 as possible prognosis biomarkers of LUAD. Besides, high BTN3A2 expression in LUAD may influence prognosis because of immune invasion. Moreover, our findings provide a potential mechanism that hsa-miR-17-5p may be miRNA targets of BTN3A2.

Nuclear and cytoplasmic RNA exosomes and PELOTA1 prevent miRNA-induced secondary siRNA production in Arabidopsis

Amplification of short interfering RNA (siRNAs) via RNA dependent RNA Polymerases (RdRPs) is of fundamental importance in RNA silencing. In plants, silencing by microRNAs (miRNAs) generally does not lead to engagement of RdRPs, in part thanks to an as yet poorly understood activity of the cytoplasmic exosome adaptor SKI2. Here, we show that mutation of the cytoplasmic exosome subunit RRP45B results in siRNA production very similar to what is observed in ski2 mutants. Furthermore, loss of the nuclear exosome adaptor HEN2 leads to secondary siRNA production from miRNA targets largely distinct from those producing siRNAs in ski2. Importantly, mutation of the Release Factor paralogue PELOTA1 required for subunit dissociation of stalled ribosomes causes siRNA production from miRNA targets overlapping with, but distinct from, those affected in ski2 and rrp45b mutants. We also show that miRNA-induced illicit secondary siRNA production correlates with miRNA levels rather than accumulation of stable 5'-cleavage fragments. We propose that stalled RNA-induced Silencing Complex (RISC) and ribosomes, but not stable target mRNA cleavage fragments released from RISC, trigger secondary siRNA production, and that the exosome limits siRNA amplification by reducing RISC dwell time on miRNA target mRNAs while PELOTA1 does so by reducing ribosome stalling.