Global mapping of miRNA-target interactions in cattle (Bos taurus)

Benchmarking the negatives: Effect of negative data generation on the classification of miRNA-mRNA interactions

MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression post-transcriptionally. In animals, this regulation is achieved via base-pairing with partially complementary sequences on mainly 3' UTR region of messenger RNAs (mRNAs). Computational approaches that predict miRNA target interactions (MTIs) facilitate the process of narrowing down potential targets for experimental validation. The availability of new datasets of high-throughput, direct MTIs has led to the development of machine learning (ML) based methods for MTI prediction. To train an ML algorithm, it is beneficial to provide entries from all class labels (i.e., positive and negative). Currently, no high-throughput assays exist for capturing negative examples. Therefore, current ML approaches must rely on either artificially generated or inferred negative examples deduced from experimentally identified positive miRNA-target datasets. Moreover, the lack of uniform standards for generating such data leads to biased results and hampers comparisons between studies. In this comprehensive study, we collected methods for generating negative data for animal miRNAs-target interactions and investigated their impact on the classification of true human MTIs. Our study relies on training ML models on a fixed positive dataset in combination with different negative datasets and evaluating their intra- and cross-dataset performance. As a result, we were able to examine each method independently and evaluate ML models' sensitivity to the methodologies utilized in negative data generation. To achieve a deep understanding of the performance results, we analyzed unique features that distinguish between datasets. In addition, we examined whether one-class classification models that utilize solely positive interactions for training are suitable for the task of MTI classification. We demonstrate the importance of negative data in MTI classification, analyze specific methodological characteristics that differentiate negative datasets, and highlight the challenge of ML models generalizing interaction rules from training to testing sets derived from different approaches. This study provides valuable insights into the computational prediction of MTIs that can be further used to establish standards in the field.

Empowering prediction of miRNA-mRNA interactions in species with limited training data through transfer learning

MicroRNAs (miRNAs) play a crucial role in mRNA regulation. Identifying functionally important mRNA targets of a specific miRNA is essential for uncovering its biological function and assisting miRNA-based drug development. Datasets of high-throughput direct bona fide miRNA-target interactions (MTIs) exist only for a few model organisms, prompting the need for computational prediction. However, the scarcity of data poses a challenge in training accurate machine learning models for MTI prediction. In this study, we explored the potential of transfer learning technique (with ANN and XGB) to address the limited data challenge by leveraging the similarities in interaction rules between species. Furthermore, we introduced a novel approach called TransferSHAP for estimating the feature importance of transfer learning in tabular dataset tasks. We demonstrated that transfer learning improves MTI prediction accuracy for species with limited datasets and identified the specific interaction features the models employed to transfer information across different species.

Identification of the viral and cellular microRNA interactomes during SARS-CoV-2 infection

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic has had a tremendous impact worldwide. Mapping virus-host interactions is critical to understand disease progression. MicroRNAs (miRNAs) are important RNA regulators, but their interaction with SARS-CoV-2 RNA was not experimentally investigated. Here, using Argonaute (AGO) cross-linking immunoprecipitation combined with RNA proximity ligation (CLEAR-CLIP), we provide unbiased mapping of SARS-CoV-2/miRNA interactions. We identified six main regions on the viral RNA bound primarily by one specific miRNA. Targeted mutagenesis and AGO1-3 knockdown demonstrated that these interactions are not critical for virus production. Moreover, we identified perturbed regulation of cellular miRNA interactions during infection, including non-compensated viral sequestration of the miR-15 family. Transcriptome analysis further showed that mRNAs targeted by this miRNA family are derepressed. This work delineates the interphase between miRNA regulation and SARS-CoV-2 infection and further contributes to deciphering the full molecular interactome of this virus.

Integrated analysis of expression profiles with meat quality traits in cattle

MicroRNAs (miRNAs) play a vital role in improving meat quality by binding to messenger RNAs (mRNAs). We performed an integrated analysis of miRNA and mRNA expression profiling between bulls and steers based on the differences in meat quality traits. Fat and fatty acids are the major phenotypic indices of meat quality traits to estimate between-group variance. In the present study, 90 differentially expressed mRNAs (DEGs) and 18 differentially expressed miRNAs (DEMs) were identified. Eighty-three potential DEG targets and 18 DEMs were used to structure a negative interaction network, and 75 matching target genes were shown in this network. Twenty-six target genes were designated as intersection genes, screened from 18 DEMs, and overlapped with the DEGs. Seventeen of these genes enriched to 19 terms involved in lipid metabolism. Subsequently, 13 DEGs and nine DEMs were validated using quantitative real-time PCR, and seven critical genes were selected to explore the influence of fat and fatty acids through hub genes and predict functional association. A dual-luciferase reporter and Western blot assays confirmed a predicted miRNA target (bta-miR-409a and PLIN5). These findings provide substantial evidence for molecular genetic controls and interaction among genes in cattle.

Mammary epithelial cell transcriptome reveals potential roles of lncRNAs in regulating milk synthesis pathways in Jersey and Kashmiri cattle

Background

Long noncoding RNAs (lncRNAs) are now proven as essential regulatory elements, playing diverse roles in many biological processes including mammary gland development. However, little is known about their roles in the bovine lactation process.

Results

To identify and characterize the roles of lncRNAs in bovine lactation, high throughput RNA sequencing data from Jersey (high milk yield producer), and Kashmiri cattle (low milk yield producer) were utilized. Transcriptome data from three Kashmiri and three Jersey cattle throughout their lactation stages were utilized for differential expression analysis. At each stage (early, mid and late) three samples were taken from each breed. A total of 45 differentially expressed lncRNAs were identified between the three stages of lactation. The differentially expressed lncRNAs were found co-expressed with genes involved in the milk synthesis processes such as GPAM, LPL, and ABCG2 indicating their potential regulatory effects on milk quality genes. KEGG pathways analysis of potential cis and trans target genes of differentially expressed lncRNAs indicated that 27 and 48 pathways were significantly enriched between the three stages of lactation in Kashmiri and Jersey respectively, including mTOR signaling, PI3K-Akt signaling, and RAP1 signaling pathways. These pathways are known to play key roles in lactation biology and mammary gland development.

Conclusions

Expression profiles of lncRNAs across different lactation stages in Jersey and Kashmiri cattle provide a valuable resource for the study of the regulatory mechanisms involved in the lactation process as well as facilitate understanding of the role of lncRNAs in bovine lactation biology.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-022-08406-x.

Differentially Expressed Extracellular Vesicle, Exosome and Non-Exosome miRNA Profile in High and Low Tick-Resistant Beef Cattle

Heavy tick burden on beef cattle account for huge economic losses globally, with an estimated value of US$22-30 billion per annum. In Australia, ticks cost the northern beef industry approximately A$170-200 million. Methods to evaluate and predict tick resistance would therefore be of great value to the global cattle trade. Exosomes (EX) are small extracellular vesicles (EVs) of ~30-150nm diameter and have gained popularity for their diagnostic and prognostic potential. EX contain, among other biomolecules, various types of RNA including micro-RNA (miRNA) and long noncoding RNA (lncRNA). MiRNA specifically have been validated as therapeutic biomarkers as they perform regulatory functions at the post-transcriptional level and are differentially expressed between divergent groups. The objective of the present study was to evaluate the miRNA profiles of EV and fractionated exosomal samples of high and low tick-resistant beef cattle to highlight potential miRNA biomarkers of tick resistance. Cows (n = 3/group) were classified into high or low tick resistant groups according to a novel scoring system. EVs and EX were isolated and fractionated from the blood plasma of high and low tick resistant cattle using established isolation and enrichment protocols. The resultant EX and non-EX samples were processed for next generation miRNA sequencing. Offspring of the cows in each high and low tick resistant group underwent the same processing for blood plasma EX, non-EX and miRNA analysis to evaluate the heritability of miRNA associated with tick resistance. A total of 2631 miRNAs were identified in EX and non-EX fractionated samples from high and low tick-resistant beef cattle. MiR-449a was highly expressed in maternal high tick-resistant EX samples. Of these, 174 were novel miRNAs, and 10 were differentially expressed (DE) (FDR < 0.05). These 10 DE miRNAs were also present in EVs, and three miRNAs were highly expressed: miR-2419-3p, miR-7861-3p and miR-2372-5p. Although 196 novel miRNAs were identified in fractionated samples of offspring, no miRNA were differentially expressed in these animals.

Bovine extracellular vesicles contaminate human extracellular vesicles produced in cell culture conditioned medium when 'exosome-depleted serum' is utilised

Extracellular vesicles (EVs) are important intercellular communication messengers. Half of the published studies in the field are in vitro cell culture based in which bovine serum in various concentrations and forms is used to facilitate the production of extracellular vesicles. 'Exosome depleted serum' is the type of bovine serum most widely used in the production of human EVs. Herein, we demonstrate that, despite the initial caution raised in 2014 about the persistence of bovine EVs, 'exosome depleted serum' was still used in 46% of publications on human or rodent EVs between 2015 and 2019. Using nanoparticle tracking analysis combined with detergent lysis of vesicles as well as bovine CD9 ELISA, we show that there were approximately 5.33 x 107/mL of bovine EVs remaining in the 'exosome depleted serum'. Importantly, the 'exosome depleted serum' was relatively enriched in small EVs by approximately 2.7-fold relative to the large EVs compared to that in the original serum. Specifically, the percentage of small EVs in total vesicles had increased from the original 48% in the serum before ultracentrifugation to 92% in the 'exosome depleted serum'. Furthermore, the pervasive bovine EVs carried over by the 'exosome depleted serum', even when the lowest concentration (0.5%) was used in cell culture, resulted in a significant contamination of human EVs in cell culture conditioned medium. Our findings indicate that the use 'exosome depleted serum' in cell culture-based studies may introduce artefacts into research examining the function of human and rodent EVs, in particular those involving EV miRNA. Thus, we appeal to the researchers in the EV field to seriously reconsider the practice of using 'exosome depleted serum' in the production of human and other mammalian EVs in vitro.

Comprehensive machine-learning-based analysis of microRNA-target interactions reveals variable transferability of interaction rules across species

BACKGROUND

MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression post-transcriptionally via base-pairing with complementary sequences on messenger RNAs (mRNAs). Due to the technical challenges involved in the application of high-throughput experimental methods, datasets of direct bona fide miRNA targets exist only for a few model organisms. Machine learning (ML)-based target prediction models were successfully trained and tested on some of these datasets. There is a need to further apply the trained models to organisms in which experimental training data are unavailable. However, it is largely unknown how the features of miRNA-target interactions evolve and whether some features have remained fixed during evolution, raising questions regarding the general, cross-species applicability of currently available ML methods.

RESULTS

We examined the evolution of miRNA-target interaction rules and used data science and ML approaches to investigate whether these rules are transferable between species. We analyzed eight datasets of direct miRNA-target interactions in four species (human, mouse, worm, cattle). Using ML classifiers, we achieved high accuracy for intra-dataset classification and found that the most influential features of all datasets overlap significantly. To explore the relationships between datasets, we measured the divergence of their miRNA seed sequences and evaluated the performance of cross-dataset classification. We found that both measures coincide with the evolutionary distance between the compared species.

CONCLUSIONS

The transferability of miRNA-targeting rules between species depends on several factors, the most associated factors being the composition of seed families and evolutionary distance. Furthermore, our feature-importance results suggest that some miRNA-target features have evolved while others remained fixed during the evolution of the species. Our findings lay the foundation for the future development of target prediction tools that could be applied to "non-model" organisms for which minimal experimental data are available.

AVAILABILITY AND IMPLEMENTATION

The code is freely available at https://github.com/gbenor/TPVOD .

Argonaute-CLIP delineates versatile, functional RNAi networks in Aedes aegypti, a major vector of human viruses

Argonaute (AGO) proteins bind small RNAs to silence complementary RNA transcripts, and they are central to RNA interference (RNAi). RNAi is critical for regulation of gene expression and antiviral defense in Aedes aegypti mosquitoes, which transmit Zika, chikungunya, dengue, and yellow fever viruses. In mosquitoes, AGO1 mediates miRNA interactions, while AGO2 mediates siRNA interactions. We applied AGO-crosslinking immunoprecipitation (AGO-CLIP) for both AGO1 and AGO2, and we developed a universal software package for CLIP analysis (CLIPflexR), identifying 230 small RNAs and 5,447 small RNA targets that comprise a comprehensive RNAi network map in mosquitoes. RNAi network maps predicted expression levels of small RNA targets in specific tissues. Additionally, this resource identified unexpected, context-dependent AGO2 target preferences, including endogenous viral elements and 3'UTRs. Finally, contrary to current thinking, mosquito AGO2 repressed imperfect targets. These findings expand our understanding of small RNA networks and have broad implications for the study of antiviral RNAi.

Integrative Network Analysis of Predicted miRNA-Targets Regulating Expression of Immune Response Genes in Bovine Coronavirus Infection

Bovine coronavirus (BCoV) infection that causes disease outbreaks among farm animals, resulting in significant economic losses particularly in the cattle industry, has the potential to become zoonotic. miRNAs, which are short non-coding segments of RNA that inhibits the expression of their target genes, have been identified as potential biomarkers and drug targets, though this potential in BCoV remains largely unknown. We hypothesize that certain miRNAs could simultaneously target multiple genes, are significantly conserved across many species, thereby demonstrating the potential to serve as diagnostic or therapeutic tools for bovine coronavirus infection. To this end, we utilized different existing and publicly available computational tools to conduct system analysis predicting important miRNAs that could affect BCoV pathogenesis. Eleven genes including CEBPD, IRF1, TLR9, SRC, and RHOA, significantly indicated in immune-related pathways, were identified to be associated with BCoV, and implicated in other coronaviruses. Of the 70 miRNAs predicted to target the identified genes, four concomitant miRNAs (bta-miR-11975, bta-miR-11976, bta-miR-22-3p, and bta-miR-2325c) were found. Examining the gene interaction network suggests IL-6, IRF1, and TP53 as key drivers. Phylogenetic analysis revealed that miR-22 was completely conserved across all 14 species it was searched against, suggesting a shared and important functional role. Functional annotation and associated pathways of target genes, such as positive regulation of cytokine production, IL-6 signaling pathway, and regulation of leukocyte differentiation, indicate the miRNAs are major participants in multiple aspects of both innate and adaptive immune response. Examination of variants evinced a potentially deleterious SNP in bta-miR-22-3p and an advantageous SNP in bta-miR-2325c. Conclusively, this study provides new insight into miRNAs regulating genes responding to BCoV infection, with bta-miR-22-3p particularly indicated as a potential drug target or diagnostic marker for bovine coronavirus.

Regulation of circadian rhythm and sleep by miR-375-timeless interaction in Drosophila

MicroRNAs are important coordinators of circadian regulation that mediate the fine-tuning of gene expression. Although many studies have shown the effects of individual miRNAs on the circadian clock, the global functional miRNA-mRNA interaction network involved in the circadian system remains poorly understood. Here, we used CLEAR (Covalent Ligation of Endogenous Argonaute-bound RNAs)-CLIP (Cross-Linking and Immuno-Precipitation) to explore the regulatory functions of miRNAs in the circadian system by comparing the miRNA-mRNA interactions between Drosophila wild-type strain W¹¹¹⁸ and a mutant of the key circadian transcriptional regulator Clock (Clk^jrk ). This experimental approach unambiguously identified tens of thousands of miRNA-mRNA interactions in both the head and body. The miRNA-mRNA interactome showed dramatic changes in the Clk^jrk flies. Particularly, among ~300 miRNA-mRNA circadian relevant interactions, multiple interactions involving core clock genes pdp1, tim, and vri displayed distinct changes as a result of the Clk mutation. Based on the CLEAR-CLIP analysis, we found a novel regulation of the circadian rhythm and sleep by the miR-375-timeless interaction. The results indicated that Clk disruption abolished normal rhythmic expression of miR-375 and the functional regulation occurred in the l-LNv neurons, where miR-375 modulated the circadian rhythm and sleep via targeting timeless. This work provides the first global view of miRNA regulation in the circadian rhythm.

Evolutionary selection of pestivirus variants with altered or no microRNA dependency

Host microRNA (miRNA) dependency is a hallmark of the human pathogen hepatitis C virus (HCV) and was also described for the related pestiviruses, which are important livestock pathogens. The liver-specific miR-122 binds within the HCV 5′ untranslated region (UTR), whereas the broadly expressed let-7 and miR-17 families bind two sites (S1 and S2, respectively) in the pestiviral 3′ UTR. Here, we dissected the mechanism of miRNA dependency of the pestivirus bovine viral diarrhea virus (BVDV). Argonaute 2 (AGO2) and miR-17 binding were essential for viral replication, whereas let-7 binding was mainly required for full translational efficiency. Furthermore, using seed site randomized genomes and evolutionary selection experiments, we found that tropism could be redirected to different miRNAs. AGO cross-linking and immunoprecipitation (CLIP) experiments and miRNA antagonism demonstrated that these alternative variants bound and depended on the corresponding miRNAs. Interestingly, we also identified miRNA-independent variants that were obtained through acquisition of compensatory mutations near the genomic 3′ terminus. Rescue experiments demonstrated that miRNA binding and 3′ mutagenesis contribute to replication through mutually exclusive mechanisms. Altogether, our findings suggest that pestiviruses, although capable of miRNA-independent replication, took advantage of miRNAs as essential host factors, suggesting a favorable path during evolutionary adaptation.

Circulating microRNA trafficking and regulation: computational principles and practice

Rapid advances in genomics discovery tools and a growing realization of microRNA's implication in intercellular communication have led to a proliferation of studies of circulating microRNA sorting and regulation across cells and different species. Although sometimes, reaching controversial scientific discoveries and conclusions, these studies have yielded new insights in the functional roles of circulating microRNA and a plethora of analytical methods and tools. Here, we consider this body of work in light of key computational principles underpinning discovery of circulating microRNAs in terms of their sorting and targeting, with the goal of providing practical guidance for applications that is focused on the design and analysis of circulating microRNAs and their context-dependent regulation. We survey a broad range of informatics methods and tools that are available to the researcher, discuss their key features, applications and various unsolved problems and close this review with prospects and broader implication of this field.

Comprehensive analysis of differentially expressed microRNAs and mRNAs in MDBK cells expressing bovine papillomavirus E5 oncogene

Delta bovine papillomaviruses (δBPVs) causes fibropapillomas or bladder cancer in cattle. E5 is the major oncogene of δBPVs; however, the influence that E5 oncogene has on host microRNA (miRNA) and mRNA expression profiles remains little elucidated. In the present study, small RNA sequencing and RNA sequencing were used to explore alterations in miRNAs and mRNAs in E5 over-expressing Madin-Darby bovine kidney (MDBK) cells compared with controls. In total, 77 miRNAs (including 30 bovine-derived miRNAs) and 223 genes were differentially expressed (DE) following E5 overexpression. The dysregulated genes were mainly involved in metabolic and biosynthetic processes. We constructed a potential miRNA-gene regulatory network from the differentially expressed genes (DEGs) and DE miRNAs. Finally, 22 DEGs and nine DE miRNAs were selected for RT-qPCR validation. Of these, downregulation of six miRNAs, bta-miR-34c, bta-miR-122, bta-miR-195, bta-miR-449b, bta-miR-2425-5p, and bta-miR-2428-3p were confirmed; In addition, upregulation of 16 genes, ACSS2, DDIT4, INHBE, INSIG1, PNRC1, PSAT1, PSPH, PYCR1, SC4MOL, SLC34A2, SCD, SPARC, IDI1, PCK2, HMGCS1, and SMIM14 and downregulation of two genes, BATF3 and WFDC2 were confirmed. Specially, bta-miR-34c and bta-miR-449b potentially regulated PYCR1 and DDIT4, which were involved in cancer progression and angiogenesis. Our study presented for the first time the comprehensive miRNA and mRNA alterations in MDBK cells expressing the BPV E5 oncogene, providing new insights into the tumorigenesis induced by BPV E5.

Computational Network Analysis Identifies Evolutionarily Conserved miRNA Gene Interactions Potentially Regulating Immune Response in Bovine Trypanosomosis

Bovine trypanosomosis is a devastating disease that causes huge economic loss to the global cattle industry on a yearly basis. Selection of accurate biomarkers are important in early disease diagnosis and treatment. Of late, micro-RNAs (miRNAs) are becoming the most useful biomarkers for both infectious and non-infectious diseases in humans, but this is not the case in animals. miRNAs are non-coding RNAs that regulate gene expression through binding to the 3'-, 5'-untranslated regions (UTR) or coding sequence (CDS) region of one or more target genes. The molecular identification of miRNAs that regulates the expression of immune genes responding to bovine trypanosomosis is poorly defined, as is the possibility that these miRNAs could serve as potential biomarkers for disease diagnosis and treatment currently unknown. To this end, we utilized in silico tools to elucidate conserved miRNAs regulating immune response genes during infection, in addition to cataloging significant genes. Based on the p value of 1.77E-32, we selected 25 significantly expressed immune genes. Using prediction analysis, we identified a total of 4,251 bovine miRNAs targeting these selected genes across the 3'UTR, 5'UTR and CDS regions. Thereafter, we identified candidate miRNAs based on the number of gene targets and their abundance at the three regions. In all, we found the top 13 miRNAs that are significantly conserved targeting 7 innate immune response genes, including bta-mir-2460, bta-mir-193a, bta-mir-2316, and bta-mir-2456. Our gene ontology analysis suggests that these miRNAs are involved in gene silencing, cellular protein modification process, RNA-induced silencing complex, regulation of humoral immune response mediated by circulating immunoglobulin and negative regulation of chronic inflammatory response, among others. In conclusion, this study identifies specific miRNAs that may be involved in the regulation of gene expression during bovine trypanosomosis. These miRNAs have the potential to be used as biomarkers in the animal and veterinary research community to facilitate the development of tools for early disease diagnosis/detection, drug targeting, and the rational design of drugs to facilitate disease treatment.

The Diverse Roles of microRNAs at the Host⁻Virus Interface

MicroRNAs (miRNAs) are small, non-coding RNAs that regulate gene expression at the post-transcriptional level. Through this activity, they are implicated in almost every cellular process investigated to date. Hence, it is not surprising that miRNAs play diverse roles in regulation of viral infections and antiviral responses. Diverse families of DNA and RNA viruses have been shown to take advantage of cellular miRNAs or produce virally encoded miRNAs that alter host or viral gene expression. MiRNA-mediated changes in gene expression have been demonstrated to modulate viral replication, antiviral immune responses, viral latency, and pathogenesis. Interestingly, viruses mediate both canonical and non-canonical interactions with miRNAs to downregulate specific targets or to promote viral genome stability, translation, and/or RNA accumulation. In this review, we focus on recent findings elucidating several key mechanisms employed by diverse virus families, with a focus on miRNAs at the host–virus interface during herpesvirus, polyomavirus, retroviruses, pestivirus, and hepacivirus infections.

MiR-27a-5p Increases Steer Fat Deposition Partly by Targeting Calcium-sensing Receptor (CASR)

Castration increases fat deposition, improving beef quality in cattle. Here, the steer group exhibited a significantly higher intramuscular fat (IMF) content than the bull group. To determine the potential roles of microRNAs (miRNAs) in castration-induced fat deposition, differential expression patterns of miRNA in liver tissue were investigated in bulls and steers. A total of 7,827,294 clean reads were obtained from the bull liver library, and 8,312,483 were obtained from the steer liver library; 452 conserved bovine miRNAs and 20 novel miRNAs were identified. The results showed that the expression profiles of miRNA in liver tissue were changed by castration, and 12 miRNAs that were differentially expressed between bulls and steers were identified. Their target genes were majorly involved in the metabolic, PI3K-Akt, and MAPK signaling pathways. Furthermore, six differentially expressed miRNAs were validated by quantitative real-time PCR, and luciferase reporter assays verified that calcium-sensing receptor (CASR) was the direct target of miR-27a-5p. Meantime, we found that the expression level of CASR was significantly higher in steers than in bulls, and revealed that CASR gene silencing in bovine hepatocytes significantly inhibited triacylglycerol (TAG) accumulation and reduced secretion of very low density lipoprotein (VLDL). These results obtained in the liver indicate that miR-27a-5p may increase fat deposition partly by targeting CASR in steers.