Prediction Tools And Databases Of miRNA Targets In Research

Written by : Mira Farzana Mohamad Mokhtar¹, Muhammad Redha Abdullah Zawawi¹ and Siti Aishah Sulaiman¹

Date Publish :  02 Disember 2022

INSTITUT BIOLOGI MOLEKUL PERUBATAN UKM

Prediction Tools And Databases Of miRNA Targets In Research

MicroRNA (miRNA) is a small, single-stranded non-coding RNA with 17-25 nucleotides in length that plays critical roles in numerous biological processes in living organisms, including humans, plants, and insects. Ambros and colleagues discovered the first miRNA called lin-4 in Caenorhabditis elegans in the year of 1993 [1]. Following that, many other miRNAs were discovered, and some have even been used in clinical trials as a molecular therapy for disease treatment [2, 3].

The biogenesis or production of miRNAs mostly follows the canonical pathway (Figure 1), whereby the RNA polymerases transcribe the miRNAs from their genes into primary miRNA with a stem-looped structure (pri-miRNA) [4]. The microprocessor, Drosha, cleaves the pri-miRNA to release a small hairpin-shaped RNA (pre-miRNA). Exportin 5 (XPO5), a karyopherin protein, will subsequently transport the pre-miRNA out of the nucleus to the cytoplasm, where the Dicer I, Ribonuclease III (DICER1) cleaves the terminal loop of pre-miRNA and producing the small RNA duplex [5]. One of the strands (mature miRNA) is incorporated into the AGO protein to form a protein complex known as the RNA-induced silencing complex (RISC) [6]. The mature miRNA with the RISC-complex will bind to the messenger RNA (mRNA) by complementary base pairing (seed sequence) to the target mRNA’s 3’-untranslated region (3’-UTR). An incomplete binding results in target mRNA translation suppression, whereas complete binding leads to mRNA degradation [4]. Due to complementary seed sequences, many miRNAs work in a network to regulate gene expression. Thus, one miRNA could have multiple targets, and one mRNA could be regulated by multiple miRNAs simultaneously [4]. Importantly, miRNAs are recognised as oncogenes or tumour suppressors [7], because they regulate the expression of crucial genes in cancer development.

Figure 1. A schematic diagram of the miRNA biogenesis pathway.

For two decades, the studies of miRNA follow the deciphering of the seed sequences and identification of the potential targets. The most commonly used strategy is to find a match between the seed sequence (position 2–7 nucleotides) and the 3’-UTR sequences of a candidate mRNA, followed by the validation of the binding in the laboratory [8]. Many cutting-edge bioinformatics tools have been developed for hosting and dissecting such miRNA-target relationships. Their importance is proven by the rapid increase in published articles on miRNA-target database subjects over the past ten years (Figure 2).

Figure 2. Graph showing an interest to miRNA target database based on the publications from 2012 to 2022. Source: PubMed

Since many potential target sites or binding between miRNA and mRNA could exist, a computational approach to predict the miRNA targets will help identify the targets by narrowing down the potential sites [8]. There are four commonly used features in the miRNA-target prediction tool to minimize false positives (high specificity): 1) seed sequence match, 2) conservation of the sequence, 3) thermodynamics, and 4) the accessibility of the sites for binding.

The seed sequence feature uses the complimentary binding based on the Watson-Crick (WC) nucleotide match between the miRNA and its given candidate target [9]. A base pairing match occurs when guanine (G) pairs with cytosine (C) and adenosine (A) pairs with uracil (U). Depending on the seed matches, several types are recognised, such as 1) 6mer: A perfect seed match for six nucleotides, 2) 7mer-m8: A perfect seed match from nucleotides 2–8 of the miRNA sequence, 3) 7mer-A1: A perfect seed match from nucleotides 2–7 of the miRNA sequence and with an A nucleotide at the position one nucleotide of the miRNA, and 4) 8mer: A perfect seed match from nucleotides 2–8 of the miRNA sequence and with an A nucleotide at the position one nucleotide of the miRNA [10].

The second feature is the conservation of the complimentary binding sequence across the species. The regions could be in the target mRNA 3′ UTR or 5′ UTR, the miRNA seed sequence, or a combination of the regions [9]. Another feature is the overall thermodynamics of the interaction between the miRNA and its target mRNA expressed in free energy (ΔG). The high and low free energy is calculated based on the miRNA and target mRNA hybridization, and the overall (ΔG) value is used to indicate how strong the binding is [11]. The last feature is the binding site accessibility, in which the likelihood of the miRNA could locate the target mRNA binding site. Since mRNA transcript does have a secondary structure, there is a chance of the binding being restricted [12]. A list of the reported miRNA prediction tools or databases is summarized in Table 1.

Table 1. A summary of the previously reported microRNA-target tools and databases.

Despite the numerous publications of miRNA prediction tools, each tool has some limitations. This limitation is due to incorporating or selecting which features into the algorithm to predict the targets. Among these miRNA prediction tools or databases, two outshine the rest based on their capacity, ease to use, input data, and software maintenance [9]. They are the miRPath that uses the DIANA-micro-T-CDS and TargetScan. Both of these tools are regularly updated, and miRPath can also predict the target sites in the gene’s coding region [23]. Notably, for miRPath, the DIANA-micro-T-CDS algorithm does not use conservation as one of the filters; thus, the prediction of the target mRNA could be lineage-specific [23]. Although TargetScan uses conservation as one of its features, the software allows for poorly conserved targets, increasing its capacity to predict more miRNAs [14].

As for others, the miRanda remains the most widely-used tool in miRNA-target prediction despite the software is not accessible online [19]. On the other hand, miRWalk is a freely accessible database providing experimentally validated and predicted miRNA-target interactions [22]. RNAhybrid requires an experienced user as the software predicts the miRNA-mRNA interaction based on the thermodynamics of the binding; thus, the users must supply the sequence input and adjusts multiple complex settings [13]. miRTarget tools predict thousands of miRNA-target interactions from high-throughput sequencing experiments in the miRDB database. Moreover, it also provides expression profiles of hundreds of cell lines that may be used to search for miRNA targets in cell lines [15].

Besides the miRNA-mRNA interaction, understanding this interaction on gene expression and biological activities is essential. The MiRSystem is the first analytical tool that simultaneously employs multiple algorithms to analyse miRNA target genes and predicts biological activities and canonical pathways for miRNAs and their targets (Figure 3) [16]. miRmap also provides more insights than regular prediction tools, as it combines multiple miRNA prediction tools to overcome the limitations of individual tools [21]. Interestingly, a disease-based miRNA prediction tool is also available. LeukmiR is a database for miRNAs and their target genes in Acute Lymphoblastic Leukaemia (ALL), a common haematological malignancy in children. LeukmiR stores all the predicted miRNAs involved in ALL, including their target genes, chromosomal locations, and deregulated signalling pathway [17].

Figure 3. Workflow for retrieving miRNA target gene data using miRSystem. Based on the list of miRNAs, target genes are retrieved via annotation from the predicted and validated databases.

In UMBI, many publications highlight the role of miRNAs in diseases. One example is the study published in 2018 that investigated the expression of miRNAs among papillary thyroid carcinoma (PTC) patients with lymph node metastasis (LNM) and without LNM and compared them to the adjacent normal tissues [25]. This study found 138 and 43 dysregulated miRNAs in PTC with LNM and without LNM, respectively, compared to normal tissues. The dysregulated miRNAs were also analysed using the miRPath for biological pathway enrichment analysis and showed that the dysregulated miRNAs in PTC with LNM were involved in fatty acid biosynthesis and metabolism, elongation, and degradation, whereas the pathways such as fatty acid biosynthesis and metabolism, proteoglycans in cancer, viral carcinogenesis, and pathways in cancer. The study also used the DIANA-micro-T database (currently incorporated in miRPath) to illustrate the relationship between the dysregulated miRNAs and the target genes in PTC, thus providing evidence of molecular interactions between miRNAs and genes in the PTC and LNM in Malaysia patients [25]. Besides that, several miRNAs-related studies from UMBI were published on cancers and non-cancers [26-30], indicating the critical roles of miRNAs in disease.

In conclusion, a summary of the miRNA-target prediction tools helps researchers to select the most appropriate tool for a specific objective. These tools have standard features to increase specificity and predict the actual targets. Nevertheless, they are also limited to the combinations of the features used, the version of the update, and the maintenance and support.

References

1. Lee, R.C., Feinbaum, R.L., and Ambros, V., The c. Elegans heterochronic gene lin-4 encodes small rnas with antisense complementarity to lin-14. Cell 1993, 75(5), 843-54.
2. Zhou, L.Y., Qin, Z., Zhu, Y.H., He, Z.Y., and Xu, T., Current rna-based therapeutics in clinical trials. Current gene therapy 2019, 19(3), 172-196.
3. Hong, D.S., Kang, Y.K., Borad, M., Sachdev, J., Ejadi, S., Lim, H.Y., Brenner, A.J., Park, K., Lee, J.L., Kim, T.Y., et al., Phase 1 study of mrx34, a liposomal mir-34a mimic, in patients with advanced solid tumours. British journal of cancer 2020, 122(11), 1630-1637.
4. Ergin, K. and Çetinkaya, R., Regulation of micrornas, in Mirnomics: Microrna biology and computational analysis, J Allmer and M Yousef, Editors. 2022, Springer US: New York, NY. p. 1-32.
5. Tsutsumi, A., Kawamata, T., Izumi, N., Seitz, H., and Tomari, Y., Recognition of the pre-mirna structure by drosophila dicer-1. Nature structural & molecular biology 2011, 18(10), 1153-8.
6. Hammond, S.M., Boettcher, S., Caudy, A.A., Kobayashi, R., and Hannon, G.J., Argonaute2, a link between genetic and biochemical analyses of rnai. Science (New York, N.Y.) 2001, 293(5532), 1146-50.
7. Otmani, K. and Lewalle, P., Tumor suppressor mirna in cancer cells and the tumor microenvironment: Mechanism of deregulation and clinical implications. Frontiers in oncology 2021, 11, 708765.
8. Peterson, S.M., Thompson, J.A., Ufkin, M.L., Sathyanarayana, P., Liaw, L., and Congdon, C.B., Common features of microrna target prediction tools. Frontiers in genetics 2014, 5, 23.
9. Lewis, B.P., Shih, I.H., Jones-Rhoades, M.W., Bartel, D.P., and Burge, C.B., Prediction of mammalian microrna targets. Cell 2003, 115(7), 787-98.
10. Brennecke, J., Stark, A., Russell, R.B., and Cohen, S.M., Principles of microrna-target recognition. PLoS biology 2005, 3(3), e85.
11. Yue, D., Liu, H., and Huang, Y., Survey of computational algorithms for microrna target prediction. Current genomics 2009, 10(7), 478-92.
12. Mahen, E.M., Watson, P.Y., Cottrell, J.W., and Fedor, M.J., Mrna secondary structures fold sequentially but exchange rapidly in vivo. PLoS biology 2010, 8(2), e1000307.
13. Rehmsmeier, M., Steffen, P., Hochsmann, M., and Giegerich, R., Fast and effective prediction of microrna/target duplexes. RNA (New York, N.Y.) 2004, 10(10), 1507-17.
14. Agarwal, V., Bell, G.W., Nam, J.W., and Bartel, D.P., Predicting effective microrna target sites in mammalian mrnas. eLife 2015, 4.
15. Chen, Y. and Wang, X., Mirdb: An online database for prediction of functional microrna targets. Nucleic acids research 2020, 48(D1), D127-d131.
16. Lu, T.P., Lee, C.Y., Tsai, M.H., Chiu, Y.C., Hsiao, C.K., Lai, L.C., and Chuang, E.Y., Mirsystem: An integrated system for characterizing enriched functions and pathways of microrna targets. PloS one 2012, 7(8), e42390.
17. Rawoof, A., Swaminathan, G., Tiwari, S., Nair, R.A., and Dinesh Kumar, L., Leukmir: A database for mirnas and their targets in acute lymphoblastic leukemia. Database : the journal of biological databases and curation 2020, 2020.
18. Rennie, W., Liu, C., Carmack, C.S., Wolenc, A., Kanoria, S., Lu, J., Long, D., and Ding, Y., Starmir: A web server for prediction of microrna binding sites. Nucleic acids research 2014, 42(Web Server issue), W114-8.
19. Enright, A.J., John, B., Gaul, U., Tuschl, T., Sander, C., and Marks, D.S., Microrna targets in drosophila. 2003, 5(1), R1.
20. Xiao, F., Zuo, Z., Cai, G., Kang, S., Gao, X., and Li, T., Mirecords: An integrated resource for microrna-target interactions. Nucleic acids research 2009, 37(Database issue), D105-10.
21. Vejnar, C.E. and Zdobnov, E.M., Mirmap: Comprehensive prediction of microrna target repression strength. Nucleic acids research 2012, 40(22), 11673-83.
22. Sticht, C., De La Torre, C., Parveen, A., and Gretz, N., Mirwalk: An online resource for prediction of microrna binding sites. PloS one 2018, 13(10), e0206239.
23. Vlachos, I.S., Zagganas, K., Paraskevopoulou, M.D., Georgakilas, G., Karagkouni, D., Vergoulis, T., Dalamagas, T., and Hatzigeorgiou, A.G., Diana-mirpath v3.0: Deciphering microrna function with experimental support. Nucleic acids research 2015, 43(W1), W460-6.
24. Vlachos, I.S., Paraskevopoulou, M.D., Karagkouni, D., Georgakilas, G., Vergoulis, T., Kanellos, I., Anastasopoulos, I.L., Maniou, S., Karathanou, K., Kalfakakou, D., et al., Diana-tarbase v7.0: Indexing more than half a million experimentally supported mirna:Mrna interactions. Nucleic acids research 2015, 43(Database issue), D153-9.
25. Mohamad Yusof, A., Jamal, R., Muhammad, R., Abdullah Suhaimi, S.N., Mohamed Rose, I., Saidin, S., and Ab Mutalib, N.-S., Integrated characterization of microrna and mrna transcriptome in papillary thyroid carcinoma. 2018, 9.
26. Yusof, A.M., Jamal, R., Saidin, S., Muhammad, R., Suhaimi, S.N.A., Rose, I.M., Nazarie, W.F.W., Fu, F.T.Y., and Mutalib, N.-S.A., Deep small rna sequencing of braf v600e mutated papillary thyroid carcinoma with lymph node metastasis. 2019, 10.
27. Mohd Shahadin, S., Ab Mutalib, N., Latif, M., Abdul Hamid, M., Hassan, T., and Jamal, R., Elucidating the role of microrna expression as a post-transcriptional mechanism in the carcinogenesis of air pollution-related lung cancer. . 2019. 10.3389/conf.fphar.2018.63.00033.
28. Ab Mutalib, N.S., Othman, S.N., Mohamad Yusof, A., Abdullah Suhaimi, S.N., Muhammad, R., and Jamal, R., Integrated microrna, gene expression and transcription factors signature in papillary thyroid cancer with lymph node metastasis. PeerJ 2016, 4, e2119.
29. Sulaiman, S.A., Abu, N., Ab-Mutalib, N.-S., Low, T.Y., and Jamal, R., Signatures of gene expression, DNA methylation and micrornas of hepatocellular carcinoma with vascular invasion. 2019, 15(22), 2603-2617.
30. Sulaiman, S.A., Muhsin, N.I.A., Arshad, A.R., Wan Mohamad Nazarie, W.F., Jamal, R., Mohamed Ibrahim, N., and Abdul Murad, N.A., Differential expression of circulating mirnas in parkinson’s disease patients: Potential early biomarker? 2020, 25(3), 319-329.