https://orcid.org/0000-0002-3354-8203
Figures
Abstract
Circular RNAs (circRNAs) are a new acknowledged class of RNAs that has been shown to play a major role in several biological functions both in physiological and pathological conditions, operating as critical part of regulatory processes, like competing endogenous RNA (ceRNA) networks. The ceRNA hypothesis is a recently discovered molecular mechanism that adds a new key layer of post-transcriptional regulation, whereby various types of RNAs can reciprocally influence each other’s expression competing for binding the same pool of microRNAs, even affecting disease development. In this study, we build a network of circRNA-miRNA-mRNA interactions in human breast cancer, called CERNOMA, that is a bipartite graph with one class of nodes corresponding to differentially expressed miRNAs (DEMs) and the other one corresponding to differentially expressed circRNAs (DEC) and mRNAs (DEGs). A link between a DEC (or DEG) and DEM is placed if it is predicted to be a target of the DEM and shows an opposite expression level trend with respect to the DEM. Within the CERNOMA, we highlighted an interesting deregulated circRNA-miRNA-mRNA triplet, including the up-regulated hsa_circRNA_102908 (BRCA1 associated RING domain 1), the down-regulated miR‐410-3p, and the up-regulated ESM1, whose overexpression has been already shown to promote tumor dissemination and metastasis in breast cancer.
Citation: Fiscon G, Funari A, Paci P (2023) Circular RNA mediated gene regulation in human breast cancer: A bioinformatics analysis. PLoS ONE 18(7): e0289051. https://doi.org/10.1371/journal.pone.0289051
Editor: Divijendra Natha Reddy Sirigiri, BMSCE: BMS College of Engineering, INDIA
Received: October 13, 2022; Accepted: July 11, 2023; Published: July 26, 2023
Copyright: © 2023 Fiscon et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files. The source code developed in R is freely available at https://github.com/giuliafiscon/CERNOMA.git
Funding: This work has been partially funded by BiBiNet project (grant number: H35F21000430002) within the POR-Lazio FESR 2014-2020, by PON Ricerca e Innovazione 2014-2020 – DM 1062/202, by Finalizzata Giovani Ricercatori 2021 (grant n.: GR-2021-12372614, CUP: B83C22007560001), and by Sapienza University of Rome grant - Progetto di ricerca di Ateneo 2022 (grant n: RP12218139131053).
Competing interests: The authors have declared that no competing interests exist.
Introduction
Circular RNAs (circRNAs) are a special class of non-coding RNAs that are generated by a process of non-canonical splicing that joins a 5’ splice site to an upstream 3’ splice site, resulting in a covalent closed loop [1–3]. CircRNAs are widely observed in both plants [4] and animals [5], and even if their biological functions remain broadly unknown, increasing evidence suggests them as crucial regulators of multiple biological processes, including the development and progression of human diseases such as cancers [6–13]. The high resistance to degradation of circRNAs, which is dependent on their circular structure, makes them different from other linear RNAs. This stability causes tissues such as blood and plasma to be especially enriched with circular RNAs compared to messenger RNAs (mRNAs) and other non-coding RNAs [14]. Thus, when released into the bloodstream by tumoral cells, circRNAs can be more easily detected with respect to other transcripts, revealing them as good potential biomarkers for early diagnosis, metastasis, and prognosis [15]. Several findings reported that circRNAs are aberrantly modulated in human cancer tissues, thus affecting carcinogenesis and metastatization, and can also be useful for predicting and monitoring treatment response [12, 16, 17]. Even though no circRNA have been effectively used as biomarkers in clinical trials yet, the impact of circRNA-mediated regulation on various cell transcriptome showed a great potential to be investigated especially in human diseases [14, 18]. Interestingly, recent studies have been focusing on the possibility that circRNAs can operate as part of competing endogenous RNA (ceRNA) regulatory networks, playing major roles in normal development and in pathologic conditions like human cancer [12, 15, 18–25].
The ceRNA mechanism is a recent discovery providing a possible explanation of fine-tuned post-transcriptional gene regulation orchestrated by the competing endogenous RNAs and microRNAs (miRNAs) [26–30]. microRNAs are small non-coding RNAs (∼ 20–22 nucleotides long) responsible for RNA silencing and post-transcriptional regulation of gene expression [31]. The ceRNA hypothesis states that various types of RNAs can reciprocally influence each other’s expression competing for binding the same pool of miRNAs, thus preventing mRNAs to be targeted [26]. This RNA-RNA crosstalk can add a new level to the understanding of complex regulatory networks that, when perturbed, could lead to disease development [19, 32–35].
Among several computation tools for ceRNAs discovery, we recently developed SPINNAKER [36], the R-implementation of the well-established model [37] that was acknowledged as the best one in terms of percentage of identified RNAs acting as ceRNA in breast cancer tissues [38]. By exploiting a multivariate statistical analysis, SPINNAKER first searches for highly correlated RNA pairs (i.e., co-expressed) and then evaluates the extent to which this correlation is direct or mediated by miRNAs, via the computation of the sensitivity correlation [37]. Finally, SPINNAKER selects only those RNA pairs whose interaction is mediated by some miRNAs (i.e., highest sensitivity correlation) and builds a ceRNA network where nodes are ceRNAs and links are miRNAs mediating their interactions. The ceRNA network can be optionally refined by considering only those triplets with ceRNAs showing a predicted binding site for the miRNA.
To run SPINNAKER and build the ceRNA network, we need as input three matrices of RNA expression levels from the same cohort of tissues/cells (i.e., two matrices for the two classes of candidate competing RNAs and one for the miRNAs). Unfortunately, these types of data are not always available, especially for the recently acknowledged class of circRNAs. To tackle this issue, in this study we developed a new computational pipeline to unveil the regulatory role of circRNA in the miRNA-target interaction network (Fig 1), when we are unable to apply SPINNAKER.
The input data are: (i) RNA- and miRNA-sequencing expression data from breast invasive carcinoma (BRCA) and matched-normal tissues retrieved from TCGA, (ii) microarray data of circRNAs from breast cancer and adjacent no-tumor breast tissues retrieved from GEO (GSE182471). Expression data were processed to obtain differential expressed RNAs (DEGs), differential expressed miRNAs (DEMs), and differential expressed circRNAs (DECs) between normal and breast cancer tissues. Next, the miRNAs predicted to target DEGs and DECs were obtained from TargetScan and circFunBase, respectively. The predicted miRNAs and the DEMs were intersected and a miRNA-target regulatory network (MRN) was constructed. Then, mRNA-circRNA pairs sharing the same miRNA and showing an opposite expression level trend with respect to the miRNA were retained. The so-called CERNOMA network was thus obtained and released as final output of the analysis.
First, we build a miRNA-target regulatory network (MRN), where nodes are miRNAs and their target genes (in this case are circRNAs and mRNAs) being significantly differentially expressed between normal and cancer tissues. Then, we generate its mapping onto the space of ceRNA network, ending up with the here-defined CERNOMA. The CERNOMA network is obtained from the MRN: (i) by selecting only the circRNAs and mRNAs sharing the predicted binding site for the same miRNAs, and (ii) by narrowing the circRNA-miRNA-mRNA triplets to those ones with a specific expression pattern. Specifically, we selected those triplets whose mRNA and circRNA show the same expression level direction (significantly up- or down-regulated) and whose miRNA shows an opposite direction (significantly down- or up-regulated). This selection should mirror the action provided by SPINNAKER to retain only the highly correlated pairs with a highest sensitivity correlation, when the correlation and thus the sensitivity correlation cannot be computed.
By applying the pipeline to study breast invasive carcinoma, within the CERNOMA network we can identify some circRNAs modulated in breast cancer exhibiting a putative regulatory activity with respect to other RNAs.
Materials and methods
Expression data collection
High throughput RNA-sequencing and miRNA-sequencing expression data of breast invasive carcinoma (brca) were acquired from The Cancer Genome Atlas (TCGA) data portal on February 2022 [39]. RNA-sequencing data correspond to read counts calculated by HT-Seq and FPKM normalized. miRNA-sequencing data correspond to normalized counts in reads-per-million-miRNA-mapped. A total of 204 samples, 102 tumor and 102 matched-normal tissues (i.e., the matched-normal tissue is defined as the tissue that is adjacent to the tumor and taken from the same patient) for both RNA- and miRNA-sequencing experiments were retained for the subsequent analysis.
Microarray dataset providing circular RNA (circRNAs) expression profile data, detected with 074301 Arraystar Human CircRNA microarray V2 on August 2021, from 5 breast cancer tissues and 5 adjacent non-tumor breast tissues were acquired from Gene Expression Omnibus (GEO) [40] database (GSE182471).
Differential expression analysis
Collected expression data were first analyzed by performing the following two phases, following the same procedure implemented in [41–43]:
Pre-processing.
Expression data were first processed by applying a logarithmic (log2) transformation and then were filtered out those genes having too many missing values among the samples (i.e., we filtered out entries showing missing values for more than 75% of the samples) and those genes with a little variation—measured by the Inter Quartile Range (IQR) percentile—across the samples (i.e., we filtered out entries showing an IQR lower than the 10th percentile of the IQR distribution).
Filtering.
The logarithmic ratio of the average expression levels of tumor samples and matched-normal samples (log fold-change) was computed and those genes falling behind, in absolute value, a fixed cutoff on the log fold-change were removed. Then, according to the type of samples distribution, a parametric (Student’s t-test for RNAs and miRNAs) or non-parametric (Wilcoxon test for circRNAs) statistical test was performed. Finally, the obtained p-values were independently adjusted for each type of data set by using False Discovery Rate (FDR) method and those genes showing an FDR lower than a chosen cut-off were considered as statically significant.
At end of this step, the differentially expressed RNAs (DEGs), the differentially expressed miRNAs (DEMs), and the differentially expressed circRNAs (DECs) between tumor and normal samples were obtained.
miRNA-target regulatory network
Starting from DEGs, DECs, and DEMs, a miRNA-target regulatory network is built via the following two phases:
Prediction of miRNA-target interactions.
Predictions of miRNAs targeting the differentially expressed mRNAs were obtained by querying TargetScan [44], which is the most up-to-date tool providing computationally predicted miRNA-mRNA interactions by searching for the exact matching between the seed region of a miRNA and the 3’ UTR of its targets.
Predictions of miRNAs targeting the differentially expressed circRNAs were obtained by querying circFunBase [45], which is a comprehensive database of functionally annotated circRNAs with more than 7000 functional circRNA entries regularly updated with newly published data, and including also computationally predicted miRNA-circRNA interactions.
Names and features for circRNAs and miRNAs refer to circBase [46] and miRBase [47] sources, respectively.
Network construction.
The miRNAs predicted to target DECs and DEGs were then intersected with DEMs and a miRNA-target regulatory network (MRN) was constructed as a bipartite network, where one class of node corresponds to DEMs and the other one corresponds to DECs or DEGs. A link between them is placed if a DEC or DEG is predicted to be target of the same DEM (Fig 2A).
A) miRNA-target Regulatory Network (MRN). It is a bipartite network, where one class of nodes corresponds to differential expressed miRNAs (diamond), the other one corresponds to differential expressed circRNAs (octagons) or differentially expressed mRNAs (circles), a link between them occurs if a circRNA or mRNA is predicted to be target of the miRNA. B) CERNOMA. It is the mapping of MRN onto a ceRNA space. It is a bipartite network with the same classes of MRN nodes, but a link between them occurs (light blue color) both if the circRNA and mRNA are predicted to be target of the same miRNA and they show an opposite expression level trend with respect to the miRNA targeting them (up-regulated circRNA and mRNA, down-regulated miRNA or viceversa). Yellow and blue nodes refer to up- and down-regulated genes in breast cancer tissues, respectively. Grey nodes refer to unselected nodes when mapping MRN onto CERNOMA. Grey links refer to links of MRN that are not in CERNOMA, light blue links refer to links of MRN that are mapped in the CERNOMA.
At the end of this step the miRNA-target regulatory network is released.
CERNOMA
The miRNA-target regulatory network is finally mapped onto a ceRNA space, where first the circRNAs and mRNAs sharing the same miRNAs were selected. Then, following the expression direction suggested by a ceRNA mechanism, the circRNA-miRNA-mRNA triplets were filtered in order to retain only those ones where the miRNA showed an opposite expression level trend with respect both to the circRNA and mRNA predicted to be its targets (i.e., up-regulated miRNA, down-regulated circRNA and mRNA or viceversa). Thus, the so-called CERNOMA was built as a bipartite network, where one class of nodes corresponds to DEMs and the other one corresponds to DECs and DEGs. A link between them is placed if both the DEC and DEG are predicted to be target of the DEM and show an opposite expression level trend with respect to the DEM (Fig 2B).
At the end of this step the CERNOMA is released and draw by using Cytoscape software [48].
Functional enrichment analysis
enrichR software [49] was used to perform Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis about the differentially expressed genes appearing in the CERNOMA network that were targets of at least one differentially expressed miRNAs. An adjusted p-value ≤ 0.05 was set as threshold to identify significantly enriched functional annotations amongst the selected gene list.
Results
Differential expression analysis
RNAs, miRNAs, and circRNAs expression data were first pre-processed and then analyzed by conducting a differential expressed analysis in order to extract genes that were significantly deregulated in breast cancer tissues (cf. Materials and Methods). According to the parameter settings defined in Table 1, we obtained a total of 562 DEGs, 265 DEMs, and 3267 DECs (S1 Table), whose expression profiles are able to clearly discriminate between breast cancer and non-tumoral adjacent breast tissues, as observed by the well-defined hierarchical clustering (Fig 3).
Heatmap and dendrogram of (A) differentially expressed RNAs (DEGs), (B) differentially expressed miRNAs (DEMs), and (C) differentially expressed circRNAs (DECs). The expression profiles of DEGs, DEMs, and DECs are clustered according to rows (genes) and columns (samples) by using as distance metrics 1-ρ, where ρ is the Pearson correlation and complete linkage algorithm as clustering method. Heatmap colors represent different expression levels (z-score normalized) that increase from blue to yellow. Red bars refer to breast cancer tissues, while green bars refer to matched-normal breast tissues.
DE (Differentially expressed), FC (Fold-change).
miRNA-target regulatory network
To investigate the ability of DECs and DEGs to bind miRNAs, we searched for predictions of miRNA-mRNA and miRNA-circRNA interactions by querying TargetScan and circFunBase database, respectively. We thus obtained a total of 39 miRNAs predicted to target 10 DECs and 231 DEGs. We then retained only those 17 miRNAs that were also differentially expressed in breast cancer and we constructed a miRNA-target regulatory network as a bipartite network composed of ten circRNAs (3 down-regulated and 7 up-regulated), 302 mRNAs (223 down-regulated, 79 up-regulated), 136 miRNAs (71 down-regulated, 65 up-regulated), and 2107 edges (S2 Table).
CERNOMA
In order to identify a putative RNA-RNA cross-talk in breast cancer tissues, starting from the miRNA-target regulatory network (MRN), we generated the CERNOMA, i.e., the MRN mapping onto a ceRNA space, where circRNAs and mRNAs share the same miRNA and are characterized by opposite expression levels trend with respect to the miRNA predicted to target both of them. The CERNOMA shows a total 208 circRNA-miRNA-mRNA triplets, and it is composed of 218 miRNA-mRNA/circRNA interactions (edges), four circRNAs (2 up-regulated and 2 down-regulated in breast cancer tissues), ten miRNAs (8 up-regulated and 2 down-regulated in breast cancer tissues), and 103 mRNAs (12 up-regulated and 91 down-regulated in breast cancer tissues) (S2 Table).
The basic features of the four circRNAs modulated in breast cancer and appearing in the CERNOMA were summarized in Table 2.
Notably, the CERNOMA is marked by a clear segregation into two internally well-connected components (Figs 4 and 6), including genes involved in different pathways and biological processes (Figs 5 and 7). In particular, the first largest component (Fig 4) is composed of eight up-regulated miRNAs (hsa-miR-128-3p, hsa-miR-15a-5p, hsa-miR-15b-5p, hsa-miR-29a-3p, hsa-miR-200b-3p, hsa-miR-200c-3p, hsa-miR-301a-3p, hsa-miR-425-5p, hsa-miR-7-5p), two down-regulated circRNAs (hsa-circRNA-407041 and hsa-circRNA-104342), and 91 down-regulated mRNAs, mainly enriched in MAPK, PI3K, RAS, WTN, RAP1 signaling pathways, breast cancer pathway, cytokine-cytokine receptor interaction, and cytokine-related biological processes (S3 Table and Fig 5).
Network showing the circRNA-miRNA-mRNA interactions. Diamonds represents miRNAs, octagons represent circRNAs, circles represent mRNAs. Gradual changes in node color represent differences in the expression levels of different genes (increasing from blue to yellow).
Dot plot showing the top 20 KEGG pathways (A) and GO Biological Processes (B) (y axis) in which the mRNAs of the first connected component were enriched (adjusted p-value < 0.05) as function of the number of mRNAs found in each category (x axis). Nodes scale with the gene ratio (i.e., number of mRNAs over the total number of genes in that functional category) and are colored according to the adjusted p-value.
The smaller connected component (Fig 6) encompasses instead two down-regulated miRNAs (hsa-miR-410-3p and hsa-miR-29a-3p), two up-regulated circRNAs (hsa-circRNA-102908 and hsa-circRNA-403236), and involves 12 up-regulated mRNAs, mainly enriched in pathways and biological processes related to cell-cell communication, such as focal adhesion and extracellular matrix interaction (S3 Table and Fig 7).
Network showing the circRNA-miRNA-mRNA interactions. Diamonds represents miRNAs, octagons represent circRNAs, circles represent mRNAs. Gradual changes in node color represent differences in the expression levels of different genes (increasing from blue to yellow).
Dot plot showing the top 10 KEGG pathways (A) and the GO Biological Processes (B) (y axis) in which the mRNAs of the second connected component were enriched (adjusted p-value < 0.05) as function of the number of mRNAs found in each category (x axis). Nodes scale with the gene ratio (i.e., number of mRNAs over the total number of genes in that functional category) and are colored according to the adjusted p-value.
Discussion
CircRNAs are discovered as a special type of non-coding RNAs [3] likely plays a pivotal role in regulatory pathways controlling lineage determination, cell differentiation, and function of various cell types [20]. Due to their circular shape, circRNAs are resistant to degradation by exonuclease activity, making them more stable than linear RNAs and reliable biomarker. CircRNAs have been also revealed key players in diverse human cancers, functioning as regulator of the expression of their parental genes and exhibiting a ceRNA activity that may even affect disease [7, 8, 10, 12, 15, 22, 24]. Yet, the impact of circRNA-mediated regulation on various transcriptomes in cancer scenario still remains controversial and open-challenging field to explore [18].
In this study, we investigated the deregulation of circRNAs and their potential regulatory activity in human breast cancer via the development of a new computational pipeline, which first constructs a miRNA-target regulatory network composed of significantly deregulated circRNAs and mRNAs predicted to be target of significantly deregulated miRNAs; and then maps it onto a ceRNA space, ending up with the so-called CERNOMA, composed of circRNA-miRNA–mRNA triplets that could putatively exhibit RNA-RNA cross-talk activity. Within the released CERNOMA network, we can distinguish two connected components: the larger one including eight up-regulated miRNAs, two down-regulated circRNAs, and 91 down-regulated mRNAs (Fig 4); the smaller one including two down-regulated miRNAs, two up-regulated circRNAs, and 12 up-regulated mRNAs (Fig 6).
In the first connected component, we observe the down-regulated circular RNA hsa_circRNA_407041 showing a predicted binding site for: (i) two members of miR-200 family (hsa-miR-200b-3p and has-miR-200c-3p), whose deregulation have been already associated to human breast cancer development and progression [50–54]; (ii) hsa-miR-425-5p, whose overexpression has been recently observed to significantly promote breast cancer cell growth and predicted a poor prognosis for breast cancer patients [55]; (iii) hsa-miR-301a-3p, an oncogenic miRNA whose expression is associated with tumor development, metastases, and overall poor prognosis in breast cancer [56]; and (iv) hsa-miR-7-5p, which was already known to be inhibited by ciRS-7 [22] and whose over expression was found to be associated to poor prognosis in other cancers such as lung carcinomas [57]. We identified also the down-regulated hsa_circRNA_104342 showing a predicted binding for two members of miR-15 family (hsa-miR-15a-5p and hsa-miR-15b-5p), recently associated to breast cancer metastasis [58], and for hsa-miR-128-3p that has been shown to function as oncomiR in breast cancer tissues and cell lines, by increasing cell invasion, proliferation, and reducing apoptosis [59]. Downregulation of circRNAs is frequent in cancer cells, as observed in hepatocellular carcinoma, colorectal adenocarcinoma, prostate and ovarian cancer, lung adenocarcinoma [12]. In an attempt to better understand the potential role of molecular players in the disease development and progression, we studied the pathways and functional gene ontology (GO) processes in which they were involved. The KEGG pathway analysis indicated that the down-regulated DEGs of the first connected component were mainly associated with MAPK, PI3K, RAP1, WNT signaling pathways, breast cancer, cytokine-cytokine receptor interaction pathway (Fig 5A); and the GO analysis revealed that they were involved in cellular response to cytokine stimulus, positive regulation of transcription, regulation of Mitogen-Activated Protein Kinase (MAPK) cascades (Fig 5B). MAPK pathway is evolutionarily conserved kinase module, which participates in several intracellular signaling pathways and plays an important role in controlling a wide spectrum of cellular processes, including proliferation, growth, migration, differentiation, and apoptosis. Abnormal functioning of MAPK signaling pathways can play a crucial role in cancer development and progression [60, 61].
In the second connected component, we can observe the up-regulated circular RNA hsa_circRNA_102908 showing a predicted binding for hsa-miR-410-3p, which has been predicted to bind also the up-regulated ESM1 gene. Increased expression level of ESM1 has been shown to exhibited significantly enhanced proliferation, migration, and invasion in breast cancer cells [10], as well as an aberrant expression of hsa-miR-410-3p is common in a variety of cancers including breast cancer, suggesting that miR-410-3p may play an important role in cancer development and progression [21]. The circular RNA hsa_circRNA_102908 is originated from the BRCA1 associated RING domain 1 and has been already found significantly up-regulated in human radioresistant esophageal cancer cell line KYSE-150R when compared with the parental cell line KYSE-150, suggesting its possible involvement in the development of radiation resistance and treatment failure [62]. We also found the up-regulated circular RNA hsa_circRNA_403236 predicted to bind hsa-miR-29a-3p, which in turn could target several genes encoding for the collagens family of proteins that strengthen and support many tissues. Cell-cell adhesion is well-known to be a fundamental process for tissue architecture and morphogenesis, and its alteration can disrupt important cellular processes and lead to a variety of diseases, including cancer [63]. Both KEGG and GO functional analyses confirmed that the up-regulated DEGs of the second connected component were mainly enriched in cell communication processes, such as extracellular matrix interaction and focal adhesion pathways (Fig 7A), as well as extracellular matrix organization and structure biological processes (Fig 7B).
The analysis conducted in this study can be generalized to investigate other pathologies and could offer potential insights for the disease understanding that are worthy of further investigation.
Supporting information
S1 Table. Differentially expressed genes.
This table lists, in three separated sheets, the differentially expressed RNAs (DEGs), miRNAs (DEMs), and circRNAs (DECs), respectively.
https://doi.org/10.1371/journal.pone.0289051.s001
(XLSX)
S2 Table. ceRNA network.
This table includes the miRNA-target regulatory network (MRN), the CERNOMA, and the list of all the circRNA-miRNA-mRNA triplets forming the CERNOMA together with their statistics obtained for breast cancer dataset.
https://doi.org/10.1371/journal.pone.0289051.s002
(XLSX)
S3 Table. Functional enrichment analysis.
This table includes the results of functional enrichment analysis for the DEGs included in the first and second component of the CERNOMA.
https://doi.org/10.1371/journal.pone.0289051.s003
(XLSX)
References
- 1. Wilusz J.E. A 360° View of Circular RNAs: From Biogenesis to Functions. WIREs RNA 2018, 9, e1478, pmid:29655315
- 2. Capel B.; Swain A.; Nicolis S.; Hacker A.; Walter M.; Koopman P.; et al. Circular Transcripts of the Testis-Determining Gene Sry in Adult Mouse Testis. Cell 1993, 73, 1019–1030. pmid:7684656
- 3. Ivanov A.; Memczak S.; Wyler E.; Torti F.; Porath H.T.; Orejuela M.R.; et al. Analysis of Intron Sequences Reveals Hallmarks of Circular RNA Biogenesis in Animals. Cell Rep. 2015, 10, 170–177, pmid:25558066
- 4. Ye C.-Y.; Chen L.; Liu C.; Zhu Q.-H.; Fan L. Widespread Noncoding Circular RNAs in Plants. New Phytol. 2015, 208, 88–95, pmid:26204923
- 5. Salzman J.; Gawad C.; Wang P.L.; Lacayo N.; Brown P.O. Circular RNAs Are the Predominant Transcript Isoform from Hundreds of Human Genes in Diverse Cell Types. PLOS ONE 2012, 7, e30733, pmid:22319583
- 6. Jeck W.R.; Sharpless N.E. Detecting and Characterizing Circular RNAs. Nat. Biotechnol. 2014, 32, 453. pmid:24811520
- 7. Wang Y.; Mo Y.; Gong Z.; Yang X.; Yang M.; Zhang S.; et al. Circular RNAs in Human Cancer. Mol. Cancer 2017, 16, 25, pmid:28143578
- 8. Meng X.; Li X.; Zhang P.; Wang J.; Zhou Y.; Chen M. Circular RNA: An Emerging Key Player in RNA World. Brief. Bioinform. 2017, 18, 547–557, pmid:27255916
- 9. Zhao Z.-J.; Shen J. Circular RNA Participates in the Carcinogenesis and the Malignant Behavior of Cancer. RNA Biol. 2017, 14, 514–521, pmid:26649774
- 10. Sarver A.L.; Subramanian S. Competing Endogenous RNA Database. Bioinformation 2012, 8, 731–733. pmid:23055620
- 11. Han B.; Chao J.; Yao H. Circular RNA and Its Mechanisms in Disease: From the Bench to the Clinic. Pharmacol. Ther. 2018, 187, 31–44, pmid:29406246
- 12. Kristensen L.S.; Jakobsen T.; Hager H.; Kjems J. The Emerging Roles of CircRNAs in Cancer and Oncology. Nat. Rev. Clin. Oncol. 2022, 19, 188–206, pmid:34912049
- 13. De Palma F.D.E.; Salvatore F.; Pol J.G.; Kroemer G.; Maiuri M.C. Circular RNAs as Potential Biomarkers in Breast Cancer. Biomedicines 2022, 10, 725, pmid:35327527
- 14. Patop I.L.; Kadener S. CircRNAs in Cancer. Curr. Opin. Genet. Dev. 2018, 48, 121–127, pmid:29245064
- 15. Fontemaggi G.; Turco C.; Esposito G.; Di Agostino S. New Molecular Mechanisms and Clinical Impact of CircRNAs in Human Cancer. Cancers 2021, 13, 3154, pmid:34202482
- 16. Smid M.; Wilting S.M.; Uhr K.; Rodríguez-González F.G.; de Weerd V.; Prager-Van der Smissen W.J.C.; et al. The Circular RNome of Primary Breast Cancer. Genome Res. 2019, 29, 356–366, pmid:30692147
- 17. Sang Y.; Chen B.; Song X.; Li Y.; Liang Y.; Han D.; et al. CircRNA_0025202 Regulates Tamoxifen Sensitivity and Tumor Progression via Regulating the MiR-182-5p/FOXO3a Axis in Breast Cancer. Mol. Ther. J. Am. Soc. Gene Ther. 2019, 27, 1638–1652, pmid:31153828
- 18. Yu C.-Y.; Kuo H.-C. The Emerging Roles and Functions of Circular RNAs and Their Generation. J. Biomed. Sci. 2019, 26, 29, pmid:31027496
- 19. Cheng D.-L.; Xiang Y.-Y.; Ji L.; Lu X.-J. Competing Endogenous RNA Interplay in Cancer: Mechanism, Methodology, and Perspectives. Tumor Biol. 2015, 36, 479–488, pmid:25604144
- 20. Memczak S.; Jens M.; Elefsinioti A.; Torti F.; Krueger J.; Rybak A.; et al. Circular RNAs Are a Large Class of Animal RNAs with Regulatory Potency. Nature 2013, 495, 333–338. pmid:23446348
- 21. Hansen T.B.; Jensen T.I.; Clausen B.H.; Bramsen J.B.; Finsen B.; Damgaard C.K.; et al. Natural RNA Circles Function as Efficient MicroRNA Sponges. Nature 2013, 495, 384–388. pmid:23446346
- 22. Hansen T.B.; Kjems J.; Damgaard C.K. Circular RNA and MiR-7 in Cancer. Cancer Res. 2013, 73, 5609–5612. pmid:24014594
- 23. He Y.; Huang H.; Jin L.; Zhang F.; Zeng M.; Wei L.; et al. CircZNF609 Enhances Hepatocellular Carcinoma Cell Proliferation, Metastasis, and Stemness by Activating the Hedgehog Pathway through the Regulation of MiR-15a-5p/15b-5p and GLI2 Expressions. Cell Death Dis. 2020, 11, 1–12, pmid:32398664
- 24. Zhong Y.; Du Y.; Yang X.; Mo Y.; Fan C.; Xiong F.; et al. Circular RNAs Function as CeRNAs to Regulate and Control Human Cancer Progression. Mol. Cancer 2018, 17, 79, pmid:29626935
- 25. Vo J.N.; Cieslik M.; Zhang Y.; Shukla S.; Xiao L.; Zhang Y.; et al. The Landscape of Circular RNA in Cancer. Cell 2019, 176, 869–881.e13, pmid:30735636
- 26. Salmena L.; Poliseno L.; Tay Y.; Kats L.; Pandolfi P.P. A CeRNA Hypothesis: The Rosetta Stone of a Hidden RNA Language? Cell 2011, 146, 353–358. pmid:21802130
- 27. Poliseno L.; Salmena L.; Zhang J.; Carver B.; Haveman W.J.; Pandolfi P.P. A Coding-Independent Function of Gene and Pseudogene MRNAs Regulates Tumour Biology. Nature 2010, 465, 1033–1038. pmid:20577206
- 28.
Conte F.; Fiscon G.; Sibilio P.; Licursi V.; Paci P. An Overview of the Computational Models Dealing with the Regulatory CeRNACeRNAsMechanism and CeRNACeRNAsDeregulation in Cancer. In Pseudogenes: Functions and Protocols; Poliseno, L., Ed.; Methods in Molecular Biology; Springer US: New York, NY, 2021; pp. 149–164 ISBN 978-1-07-161503-4.
- 29. Poliseno L.; Pandolfi P.P. PTEN CeRNA Networks in Human Cancer. Methods 2015, 77, 41–50, pmid:25644446
- 30. Russo F.; Fiscon G.; Conte F.; Rizzo M.; Paci P.; Pellegrini M. Interplay Between Long Noncoding RNAs and MicroRNAs in Cancer. Methods Mol. Biol. Clifton NJ 2018, 1819, 75–92, pmid:30421400
- 31. Friedman R.C.; Farh K.K.-H.; Burge C.B.; Bartel D.P. Most Mammalian MRNAs Are Conserved Targets of MicroRNAs. Genome Res. 2009, 19, 92–105, pmid:18955434
- 32. Zheng Q.; Bao C.; Guo W.; Li S.; Chen J.; Chen B.; et al. Circular RNA Profiling Reveals an Abundant CircHIPK3 That Regulates Cell Growth by Sponging Multiple MiRNAs. Nat. Commun. 2016, 7, 11215, pmid:27050392
- 33. Li F.; Zhang L.; Li W.; Deng J.; Zheng J.; An M.; et al Circular RNA ITCH Has Inhibitory Effect on ESCC by Suppressing the Wnt/β-Catenin Pathway. Oncotarget 2015, 6, 6001–6013.
- 34. Ergun S.; Oztuzcu S. Oncocers: CeRNA-Mediated Cross-Talk by Sponging MiRNAs in Oncogenic Pathways. Tumor Biol. 2015, 36, 3129–3136, pmid:25809705
- 35. Karreth F.A.; Tay Y.; Perna D.; Ala U.; Tan S.M.; Rust A.G.; et al. In Vivo Identification of Tumor- Suppressive PTEN CeRNAs in an Oncogenic BRAF-Induced Mouse Model of Melanoma. Cell 2011, 147, 382–395, pmid:22000016
- 36. Paci P.; Fiscon G. SPINNAKER: An R-Based Tool to Highlight Key RNA Interactions in Complex Biological Networks. BMC Bioinformatics 2022, 23, 166, pmid:35524174
- 37. Paci P.; Colombo T.; Farina L. Computational Analysis Identifies a Sponge Interaction Network between Long Non-Coding RNAs and Messenger RNAs in Human Breast Cancer. BMC Syst. Biol. 2014, 8, 83, pmid:25033876
- 38. Le T.D.; Zhang J.; Liu L.; Li J. Computational Methods for Identifying MiRNA Sponge Interactions. Brief. Bioinform. 2017, 18, 577–590, pmid:27273287
- 39. Tomczak K.; Czerwinska P.; Wiznerowicz M.; others The Cancer Genome Atlas (TCGA): An Immeasurable Source of Knowledge. Contemp Oncol Pozn 2015, 19, A68–A77.
- 40. Barrett T.; Wilhite S.E.; Ledoux P.; Evangelista C.; Kim I.F.; Tomashevsky M.; et al. NCBI GEO: Archive for Functional Genomics Data Sets—Update. Nucleic Acids Res. 2013, 41, D991–D995. pmid:23193258
- 41. Paci P.; Colombo T.; Fiscon G.; Gurtner A.; Pavesi G.; Farina L. SWIM: A Computational Tool to Unveiling Crucial Nodes in Complex Biological Networks. Sci. Rep. 2017, 7, srep44797, pmid:28317894
- 42. Paci P.; Fiscon G. SWIMmeR: An R-Based Software to Unveiling Crucial Nodes in Complex Biological Networks. Bioinformatics 2022, 38, 586–588, pmid:34524429
- 43. Fiscon G.; Conte F.; Paci P. SWIM Tool Application to Expression Data of Glioblastoma Stem-like Cell Lines, Corresponding Primary Tumors and Conventional Glioma Cell Lines. BMC Bioinformatics 2018, 19, 436, pmid:30497369
- 44. Agarwal V.; Bell G.W.; Nam J.-W.; Bartel D.P. Predicting Effective MicroRNA Target Sites in Mammalian MRNAs. eLife 2015, 4, e05005, pmid:26267216
- 45. Meng X.; Hu D.; Zhang P.; Chen Q.; Chen M. CircFunBase: A Database for Functional Circular RNAs. Database 2019, 2019, baz003, pmid:30715276
- 46. Glažar P.; Papavasileiou P.; Rajewsky N. CircBase: A Database for Circular RNAs. RNA 2014, pmid:25234927
- 47. Kozomara A.; Griffiths-Jones S. MiRBase: Annotating High Confidence MicroRNAs Using Deep Sequencing Data. Nucleic Acids Res. 2014, 42, D68–D73, pmid:24275495
- 48. Shannon P.; Markiel A.; Ozier O.; Baliga N.S.; Wang J.T.; Ramage D.; et al. Cytoscape: A Software Environment for Integrated Models of Biomolecular Interaction Networks. Genome Res. 2003, 13, 2498–2504, pmid:14597658
- 49. Kuleshov M.V.; Jones M.R.; Rouillard A.D.; Fernandez N.F.; Duan Q.; Wang Z.; et al. Enrichr: A Comprehensive Gene Set Enrichment Analysis Web Server 2016 Update. Nucleic Acids Res. 2016, 44, W90–97, pmid:27141961
- 50. Yao C.-X.; Wei Q.-X.; Zhang Y.-Y.; Wang W.-P.; Xue L.-X.; Yang F.; et al. MiR-200b Targets GATA-4 during Cell Growth and Differentiation. RNA Biol. 2013, 10, 465–480, pmid:23558708
- 51. Conte F.; Fiscon G.; Chiara M.; Colombo T.; Farina L.; Paci P. Role of the Long Non-Coding RNA PVT1 in the Dysregulation of the CeRNA-CeRNA Network in Human Breast Cancer. PLoS ONE 2017, 12, pmid:28187158
- 52. Wang G.; Guo X.; Hong W.; Liu Q.; Wei T.; Lu C.; et al. Critical Regulation of MiR-200/ZEB2 Pathway in Oct4/Sox2-Induced Mesenchymal-to-Epithelial Transition and Induced Pluripotent Stem Cell Generation. Proc. Natl. Acad. Sci. 2013, 110, 2858–2863, pmid:23386720
- 53. Mongroo P.S.; Rustgi A.K. The Role of the MiR-200 Family in Epithelial-Mesenchymal Transition. Cancer Biol. Ther. 2010, 10, 219–222. pmid:20592490
- 54. Yuan J.; Xiao C.; Lu H.; Yu H.; Hong H.; Guo C.; et al. MiR‑200b Regulates Breast Cancer Cell Proliferation and Invasion by Targeting Radixin. Exp. Ther. Med. 2020, 19, 2741–2750, pmid:32256756
- 55. Zhang L.-F.; Zhang J.-G.; Zhou H.; Dai T.-T.; Guo F.-B.; Xu S.-Y.; et al. MicroRNA-425-5p Promotes Breast Cancer Cell Growth by Inducing PI3K/AKT Signaling. Kaohsiung J. Med. Sci. 2020, 36, 250–256, pmid:31688991
- 56. Lettlova S.; Brynychova V.; Blecha J.; Vrana D.; Vondrusova M.; Soucek P.; et al. MiR-301a-3p Suppresses Estrogen Signaling by Directly Inhibiting ESR1 in ERα Positive Breast Cancer. Cell. Physiol. Biochem. Int. J. Exp. Cell. Physiol. Biochem. Pharmacol. 2018, 46, 2601–2615, pmid:29763890
- 57. Chou Y.-T.; Lin H.-H.; Lien Y.-C.; Wang Y.-H.; Hong C.-F.; Kao Y.-R.; et al. EGFR Promotes Lung Tumorigenesis by Activating MiR-7 through a Ras/ERK/Myc Pathway That Targets the Ets2 Transcriptional Repressor ERF. Cancer Res. 2010, 70, 8822–8831, pmid:20978205
- 58. Wu B.; Liu G.; Jin Y.; Yang T.; Zhang D.; Ding L.; et al. MiR-15b-5p Promotes Growth and Metastasis in Breast Cancer by Targeting HPSE2. Front. Oncol. 2020, 10. pmid:32175269
- 59. Li Y.; Wang Y.; Shen X.; Han X. MiR-128 Functions as an OncomiR for the Downregulation of HIC1 in Breast Cancer. Front. Pharmacol. 2019, 10. pmid:31680974
- 60. Dhillon A.S.; Hagan S.; Rath O.; Kolch W. MAP Kinase Signalling Pathways in Cancer. Oncogene 2007, 26, 3279–3290, pmid:17496922
- 61. Munshi A.; Ramesh R. Mitogen-Activated Protein Kinases and Their Role in Radiation Response. Genes Cancer 2013, 4, 401–408, pmid:24349638
- 62. Su H.; Lin F.; Deng X.; Shen L.; Fang Y.; Fei Z.; et al. Profiling and Bioinformatics Analyses Reveal Differential Circular RNA Expression in Radioresistant Esophageal Cancer Cells. J. Transl. Med. 2016, 14, 225, pmid:27465405
- 63. Hirohashi S.; Kanai Y. Cell Adhesion System and Human Cancer Morphogenesis. Cancer Sci. 2003, 94, 575–581, pmid:12841864