|Home | About | Journals | Submit | Contact Us | Français|
We investigated the differential expression of circular RNAs (circRNAs) in plasma samples from three coronary artery disease (CAD) patients to identify putative therapeutic targets. We identified 24 differentially expressed circRNAs (18 up-regulated and 6 down-regulated) and 7 differentially expressed mRNAs (6 up-regulated and 1 down-regulated) in CAD patients based on competing endogenous RNA (ceRNA) microarray analysis. MiR-221(p = 0.001), miR-155(p = 0.049), and miR-130a (p = 0.001) were downregulated in CAD patients based on qRT-PCR analysis of another independent population of 932 study subjects (648 CAD subjects and 284 controls). We constructed a hsa-miR-130a-3p-mediated circRNA-mRNA ceRNA network using the miRanda database. This included 9 circRNAs (hsa_circ_0089378, hsa_circ_0083357, hsa_circ_0082824, hsa_circ_0068942, hsa_circ_0057576, hsa_circ_0054537, hsa_circ_0051172, hsa_circ_0032970, and hsa_circ_0006323) and 1 mRNA (transient receptor potential cation channel subfamily M member 3 [TRPM3]). We have shown that 9 circRNAs promote TRPM3 expression by inhibiting hsa-miR-130a-3p in CAD patients.
Cardiovascular disease (CVD) is the leading cause of human morbidity and mortality worldwide. According to World Health Organization (WHO), approximately 17.5 million people die annually from CVD . This represents 31% of all global deaths. Stroke and coronary artery disease (CAD) are the major causes of CVD related deaths . In China, CVD accounts for 300 deaths out of every 100,000 individuals , while, nearly a quarter of the Western population suffers from CAD and stroke [3–4]. Therefore, it is essential to identify risk factors associated with CVD that could be used for diagnosis and treatment at an early stage. The role of non-coding RNAs (ncRNAs) has been recognized in various human pathologies and they represent great potential as CAD biomarkers.
Human genome sequencing shows that although only 3% of the human genome codes for proteins, 80% of the human genome are transcribed . The numbers of non-coding transcripts greatly exceed protein-coding mRNAs and represent species complexity [6–7]. Long noncoding RNAs (lncRNAs) are noncoding RNAs that are longer than 200 nucleotides that regulate diverse cellular functions. Recently, lncRNAs have been detected in human plasma, which can be used as disease biomarkers . Aberrant levels of lncRNAs have also been reported in plasma from CAD patients . Circular RNAs (circRNAs) are another class of non-coding RNAs that regulate gene expression in eukaryotes . The circRNAs are competing endogenous RNAs (ceRNAs) that act as a sponge for microRNAs (miRNAs) by complementary base paring and therefore regulate gene transcription [11–12]. However, the expression and biological functions of ceRNAs in CAD are unknown. Therefore, in the present study, we generated plasma ceRNA expression profiles of three pairs of CAD and control samples to investigate their potential role as diagnostic markers in CAD.
The circRNA expression profiles in human CAD and control plasma were compared by the scatter plot (Figure (Figure1)1) and volcano plot filtering (Figure (Figure2)2) to identify differentially expressed circRNAs. We identified 24 differentially expressed circRNAs (fold change ≥ 1.5 and P < 0.05) between CAD and control plasma (Table (Table1).1). Figure Figure33 shows the heat map of these 24 differentially expressed circRNAs. Among these, 18 circRNAs were up-regulated and 6 circRNAs were down-regulated in CAD plasma. Among the up-regulated circRNAs, 17 were exonic, and 1 intragenic. Among the down-regulated circRNAs, 6 were exonic and 1 intronic. The circRNA, hsa_circ_0051686 was both intronic and exonic. SBC Human (4*180 K) ceRNA microarray profiling showed 6 up-regulated and 1 down-regulated mRNAs (fold change ≥ 1.5 and P < 0.05) in CAD patients (Table (Table22).
To establish a circRNA-miRNA-mRNA ceRNA network, we compared plasma miRNA expression profiles of CAD and control subjects in another independent population that was previously reported . We identified 9 CAD-related miRNAs including miR-122, miR-133b, miR-214, miR-21, miR-106a, miR-130a, miR-155, miR-221, and miR-125b. Among these, miR-221 (p = 0.001), miR-155 (p = 0.049), and miR-130a (p = 0.001) were downregulated in CAD subjects than in non-CAD subjects (Table (Table33).
The hsa-miR-221-3p, hsa-miR-155-5p, and hsa-miR-130a-3p were selected for integrated analysis of miRNA and mRNA profiling data. The miRNA target predictions were based on miRanda (release August 2010; http://www.microrna.org/). We performed integrated analysis of the inverse relations of expressed miRNAs and mRNAs in conjunction with their predicted targets and identified transient receptor potential cation channel subfamily M member 3 (TRPM3) as a target gene for hsa-miR-130a-3p.
Recently, circRNAs have been identified as miRNA sponges that regulate gene expression. Therefore, we used miRanda database to investigate potential miRNAs that bind to circRNAs in CAD patients. We observed that hsa-miR-221-3p, hsa-miR-155-5p, and hsa-miR-130a-3p were bound by 10, 12, and 9 circRNAs, respectively (Table (Table44).
Since circRNAs interact with miRNAs through miRNA response elements (MREs), we searched for putative hsa-miR-130a-3p MREs in the circRNAs using miRanda. We identified 9 circRNAs that had hsa-miR-130a-3p binding sites and negatively associated with hsa-miR-130a-3p. These included hsa_circ_0089378, hsa_circ_0083357, hsa_circ_0082824, hsa_circ_0068942, hsa_circ_0057576, hsa_circ_0054537, hsa_circ_0051172, hsa_circ_0032970, and hsa_circ_0006323. Based on these data, we constructed a hsa-miR-130a-3p-mediated circRNA-mRNA ceRNA network with 9 circRNAs and 1 mRNA (Figure (Figure44 and Table Table55).
The ceRNAs play a key role in post-transcriptional regulation and have been implicated in cardiovascular disease [12, 14–15]. In our previous study, we reported several miRNAs associated with CAD, but their actions were unknown . In order to deciphere the ceRNA mechanisms related to CAD, we constructed a global ceRNA-miRNA-mRNA triple network using the miRanda database. We identified 9 circRNAs and 1 mRNA that formed a network with hsa-miR-130a-3p. This study sheds new insights to exploring the complex post-transcriptional regulatory networks via ceRNA interactions and identifying pathways that are altered in CAD. Subsequently, specific ceRNAs can be therapeutically used to modulate specific pathways that are involved in CAD pathology.
The miR-130 precursor is a small non-coding RNA that has been identified in mice (MI0000156, MI0000408), humans (MI0000448, MI0000748) and a range of vertebrate species (MIPF0000034). Mature miR-130 is generated by the dicer enzyme upon excision from the 3′ arm of the hairpin. MiR-130a-3p is inversely associated with coronary atherosclerosis with its down-regulation contributing to endothelial progenitor cell dysfunction in subjects suffering from coronary artery disease [16–18]. Therefore, miR-130a-3p has therapeutic potential for the prevention and treatment of CAD. Its regulation by ceRNAs as shown in this study represents one possible mode of regulation that is clinically applicable.
TRPM3 belongs to the family of transient receptor potential (TRP) channels that regulate cellular calcium homeostasis. TRPM3 mediates calcium entry potentiated by calcium store depletion. Alternatively spliced transcript variants encoding different isoforms of TRPM3 have also been identified. TRPM3 regulates proliferation and contractility of vascular smooth muscle cells in co-ordination with cholesterol and is potentially involved in therapeutic vascular modulation . In the present study, we identified TRPM3 mRNA as a hsa-miR-130a-3p target and is upregulated in CAD subjects.
Circular RNAs (circRNAs) are new members of ceRNAs that are involved in regulating gene expression . However, their role in CAD pathogenesis has not been reported. In the present study, we identified 24 aberrantly expressed circRNAs (18 up-regulated and 6 down-regulated) in CAD patients. These included 9 circRNAs,namely, hsa_circ_0089378, hsa_circ_0083357, hsa_circ_0082824, hsa_circ_0068942, hsa_circ_0057576, hsa_circ_0054537, hsa_circ_0051172, hsa_circ_0032970, and hsa_circ_0006323, which sponge hsa-miR-130a-3p that regulates TRPM3. This suggests that the cohort of circRNAs negatively regulate miR-130A-3p, thereby resulting in upregulation of TRPM3.
This study has several limitations that need to be addressed while interpreting our results. First, the ceRNA microarray was based on a small group of 3 CAD and 3 control subjects. Since the expression of circRNA and ceRNA can vary in individuals due to a number of factors, the circRNA and mRNA profiles were verified in another independent cohort of 932 subjects by qRT-PCR. The in vivo relevance of our findings requires further comprehensive investigation. Second, the mechanism of circRNA regulation of hsa-miR-130a-3p and TRPM3 is based on the bioinformatics analysis and needs to be tested in in vitro and in vivo models.
In conclusion, we identified a network of 9 circRNAs that regulate TRPM3 expression by inhibiting hsa-miR-130a-3p in CAD patients.
In the first group, we enrolled 3 CAD and 3 control (4 males and 2 females) subjects at the Yancheng TCM Hospital Affiliated to Nanjing University of Chinese Medicine in China. The study was performed as approved by the ethics committee of the Yancheng TCM Hospital Affiliated to Nanjing University of Chinese Medicine and the First Affiliated Hospital of Nanjing Medical University. All subjects provided written informed consent. The characteristics of the study subjects are shown in Table Table66.
In the second group, 932 consecutive adult subjects (681 males and 251 females, 648 CAD subjects and 284 controls) aged 32–84 years that underwent coronary angiography for suspected or known coronary atherosclerosis was analyzed. They were part of a previous study on CAD. Subjects with spastic angina pectoris, infectious processes within 2 weeks, heart failure, adrenal dysfunction, and thyroid dysfunction were excluded from this study. The plasma miRNA levels were confirmed by qRT-PCR analysis. Circulating levels of miRNAs were quantified using the 2−Δct method.
Total RNA was extracted from the plasma of the subjects and purified using the mirVanaTM PARISTM kit (Cat#AM1556, Ambion, Austin, TX, USA) according to the manufacturer's instructions. The RNA integration number (RIN) was determined by an Agilent Bioanalyzer 2100 (Agilent technologies, Santa Clara, CA, USA). Total RNA (1μg) was amplified and labeled by Low Input Quick Amp WT Labeling Kit (Cat. # 5190-2943, Agilent technologies, Santa Clara, CA, USA) according to the manufacturer's instructions. Labeled cRNA were purified by RNeasy mini kit (Cat.# 74106, QIAGEN, GmBH, Germany).
The cDNA was labeled and hybridized to the human SBC-ceRNA (4×180k) Array, which can detect 88,371 circRNAs, and 18,853 coding transcripts (Circbase (88371), GENCODE v21 /Ensembl (18,100), LNCipedia v3.1 (40,621), Lncrnadb (28), Noncode v4 (2,608), UCSC (25,919) databases). We hybridized 1μg Cy3-labeled cRNA onto each slide using Gene Expression Hybridization Kit (Cat.# 5188-5242, Agilent technologies, Santa Clara, CA, US) in an hybridization oven (Cat.# G2545A, Agilent technologies, Santa Clara, CA, US) according to the manufacturer's instructions for 17 h. Then, slides were washed in staining dishes (Cat.# 121, Thermo Shandon, Waltham, MA, USA) with Gene Expression Wash Buffer Kit (Cat.# 5188-5327, Agilent technologies, Santa Clara, CA, USA) according to the manufacturer's instructions.
The slides were scanned by Agilent Microarray Scanner (Cat#G2565CA, Agilent technologies, Santa Clara, CA, USA) with default settings (Dye channel: Green; Scan resolution = 3 μm; PMT 100%; 20 bit). Raw data was extracted with Feature Extraction software 10.7 (Agilent technologies, Santa Clara, CA, USA) and normalized by Quantile algorithm and LIMMA packages in R. The microarray work was performed by Shanghai Biotechnology Cooperation, Shanghai, P.R. China.
Figure Figure55 shows the methodology used to identify the ceRNA interacting genes. The miRanda method, which is based on dynamic programming (SW algorithm) and computing free energy was used to identify the target genes of miRNAs and ceRNAs [21–25]. Based on this analysis, we built a microRNA-cirRNA-mRNA interaction network. The relationship between the target and the microRNA was based on the adjacency matrix of microRNA and target A = [ai,j], where ai,j represents the weight of the relationship between the target (i) and its microRNA (j). In the microRNA-cirRNA-mRNA network, the circle represents one edge, whereas the center of the network represents a degree. The degree denotes the contribution of a microRNA to the target gene around it or the contribution of a target to the microRNAs around it. The key microRNAs and their targets in the network always had the largest degrees.
This study was supported by grants from the National Natural Science Foundations of China (grant numbers 81170180, 30400173 and 30971257) and the Priority Academic Program Development of Jiangsu Higher Education Institutions.
CONFLICTS OF INTEREST
The authors declare that no conflicts of interest exist.