Abstract
Atrial fibrillation (AF) is one of the most common clinical cardiac arrhythmias. This study was done to screen differentially expressed circular RNAs (circRNAs) in human monocytes from patients with AF and healthy controls using microarray, and preliminarily explore the role of circRNAs in the development of AF. The expression of circRNAs in peripheral blood monocytes of 4 AF patients and 4 healthy donors was detected by chip technology and validated by qRT-PCR. Differentially expressed genes were screened out. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to identify the function of differentially expressed genes and related pathways. Potential connections between circRNAs and miRNAs were explored by using Cytoscape. 120 differentially expressed circRNAs (FC≥2, P<0.05) were preliminarily screened by circRNA microarray, of which 65 were up-regulated and 55 down-regulated. All of 4 upregulated circRNAs (circRNA_0031, circRNA_1837, circRNA_5901 and circRNA_7571) and 3 out of 4 downregulated circRNAs (circRNA_5801, circRNA_7386 and circRNA_7577) were randomly confirmed by RT-PCR. GO and KEGG analysis suggested that differentially expressed circRNA-related genes are mainly involved in inflammation, immunity, and signaling transduction. CircRNA_7571, circRNA_4648, circRNA_4631 and circRNA_2875 were the first 4 circRNAs with the most binding nodes in the co-expression network. In addition, hsa-miR-328 was the highest positively correlated miRNA in the networks. Our findings demonstrated that there were differentially expressed circRNAs in human monocytes from AF patients. circRNA_7571, circRNA_4648, circRNA_4631 and circRNA_2875 were the first 4 circRNAs with the most binding nodes in the co-expression network. hsa-miR-328 was the largest node that interacted with circRNAs in the co-expression network. circRNAs-hsa-miR-328 network may play a critical role in the pathophysiology and mechanism of AF.
Keywords: circRNAs, atrial fibrillation, chip, expression profile
Introduction
Atrial fibrillation (AF) is one of the most common clinical cardiac arrhythmias. Its incidence is high and increases with age. The age-adjusted prevalence of AF is 0.60% for men and 0.37% for women. Annual incidence of AF is 0.78% for men and 0.40% for women. The lifetime risk of AF in men 40 years and older is 26% for men and 23% for women, while the incidence rate for elderly 60-74 years is as high as 8.0%-11% [1,2]. At present, the pathogenesis of AF is mainly reflected in atrial structural remodeling, electrical remodeling, inflammation, and genes [3,4], but the specific pathogenesis has not been clarified. Moreover, there is a lack of new biomarkers with strong specificity for the diagnosis and screening of AF.
With the rapid development of genetic information technology, a class of non-coding RNAs (ncRNAs) that does not encode proteins after transcription and was once considered as “noise” has attracted more attention [5,6]. Research on ncRNAs associated with AF is currently focused on microRNA (miRNA) and long-chain ncRNA (lncRNA). It has been found that miRNA and lncRNA may play important roles in occurrence and development of AF by regulating the atrial structural remodeling, electrical remodeling, and neural remodeling, and can also be used as biomarkers of AF [7-9]. Circular RNAs (circRNAs) are a family of ncRNAs formed by a special splicing mechanism, which has a closed circular structure and is abundant in eukaryotic transcriptomes. Despite other putative regulatory functions, circRNAs perform as scavengers to capture other RNA molecules. Studies have shown that circRNAs can be used as a miRNA sponge to competitively bind miRNAs, deregulate the regulation of target molecules by these miRNAs, and play an important role in transcription, post-transcription, and translation [10,11].
Athough AF is associated with a high risk of stroke and death, few studies have performed to explore the role of circRNAs in AF. Whether circRNAs can be used as a diagnostic marker and therapeutic target has not been reported. Thus, in the current study, we evaluated the differential expression profiles of circRNAs in peripheral blood monocytes from AF patients and healthy donors by microarray. Quantitative reverse transcription PCR (qRT-PCR) for differentially expressed circRNAs was subsequently performed to validate the microarray results. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to identify the functions of differentially expressed genes and related pathways. Furthermore, the potential connections between circRNAs and miRNAs were explored by using Cytoscape.
Materials and methods
Study population and specimen collection
This study included 4 patients with AF diagnosed in the Department of Cardiovascular Medicine of Taizhou People’s Hospital in October 2019 (AF group). All of them had paroxysmal atrial fibrillation, and 4 healthy subjects who excluded AF were used as controls (control group). The diagnosis of AF was mainly based on criteria listed in the 2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS [12]. Patients with malignant tumor, acute infection, systemic immunity disease, thyroid disease, severe anemia, various organ transplantations and severe heart, lung, liver and renal insufficiency were excluded. About 10 ml of peripheral blood was drawn into ethylene diamine tetra-acetic acid (EDTA) anticoagulant tubes from each participant within 4 hours after admission. Monocytes were purified from PBMCs using Monocyte Isolation Kit II (Miltenyi Biotec, Tokyo, Japan) and frozen for analysis. This study was approved by the Ethics Committee of Taizhou People’s Hospital. Written informed consent was obtained from AF patients and controls before entering this experiment.
Main reagents and instruments
Monocyte Isolation Kit II (Miltenyi Biotec, Tokyo, Japan), Trizol reagent (Ambion, USA), RNA purification kit (QIGEN, Germany), PrimeScript RT reagent kit with gDNA Eraser reverse transcription kit (TaKaRa, Japan), quantitative PCR detection kit (TaKaRa, Japan); gene chip detection (Agilent human lncRNA Array V2.0), ND-2000 spectrophotometer (NanoDrop, USA); high-speed refrigerated centrifuge, -80°C refrigerator, and other conventional instruments.
RNA extraction and quality control
Trizol reagent (Ambion, USA) was used to extract the total RNA in monocytes, and QIAGEN Rneasy® Mini Kit (QIGEN, Germany) was used to purify the RNA. The high purity and concentrations of the extracted RNA were tested with NanoDrop nn-2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA). RNA integrity was tested by using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA).
circRNAs microarry analysis
Sample labeling and microarray hybridization were performed by Outdo Bio-tech (Shanghai, P.R. China) by the same method as previously described [13]. Thus, double-stranded cDNA was synthesized from the qualified RNA by reverse transcription, and Cy3 fluorescently labeled cRNA was further synthesized. The fluorescence intensity of Cy3 in each sample was scanned by Axon microarray 4000B microarray scanner and the samples with RNA integrity number ≥7 were subjected to the subsequent analysis on the Illumina sequencing platform (HiSeq 2500 or other platform) and 150/125 bp paired-end reads were generated. Junction reads of each sample were counted to evaluate the relative expression of circRNAs in different samples and normalized by DESeq software. The fold-change between different samples was calculated. The statistical significance was calculated by t test. circRNAs with fold-change >2 and P<0.05 were regarded as significant differential expression.
qRT-PCR validation of differentially expressed circRNAs
In order to confirm the results of microarray analysis, four upregulated circRNAs (circRNA_0031, circRNA_1837, circRNA_5901 and circRNA_7571) and four downregulated circRNAs (circRNA_2773, circRNA_5801, circRNA_7386 and circRNA_7577) were selected randomly for validation by qRT-PCR. Simply, 1 μl of cDNAs was added to 12.5 μl of SYBR-Green Gene Expression Master Mix (Applied Biosystems, Inc.), 10.5 μl of DEPC-treated water, and 0.5 μl of reverse and forward primers. The gene expression level of target circRNAs was normalized to the housekeeping gene GAPDH (Sangon Biotech, Shanghai, China) and calculated using the (2-ΔΔCt) method. The primer sequences for RT-PCR are shown in Table 1.
Table 1.
Primer sequences for reverse transcription polymerase chain reaction
| Gene name | circbase_id | Primer sequences | Fragment (bp) |
|---|---|---|---|
| GAPDH | - | F: 5’-TCTCTGCTCCTCCCTGTTCTA-3’ | 177 |
| R: 5’-ATGAAGGGGTCGTTGATGGC-3’ | |||
| circRNA_0031 | hsa_circ_0008737 | F: 5’-ACUGCCCUAAGUGCUCCUUCUGG-3’ | 179 |
| R: 5’-AGAGAAGGGGCCTGAGGGCAGA-3’ | |||
| circRNA_1837 | - | F: 5’-GCUGGGAUUACAGGCAUGAGCC-3’ | 192 |
| R: 5’-GGCTCACGCCTGTAATCCCAGG-3’ | |||
| circRNA_5901 | hsa_circ_0001240 | F: 5’-CAGUGGCCAGAGCCCUGACGUG-3’ | 159 |
| R: 5’-TGCTGCCGGGAGCATCGGCCACTG-3’ | |||
| circRNA_7571 | - | F: 5’-GGUCCAGAGGGCCGTCGT-3’ | 165 |
| R: 5’-ATCCCTGTCCATCTCTGGACC-3’ | |||
| circRNA_2773 | - | F: 5’-GGGGUUCCUGGGGAUGGGAUUU | 163 |
| R: 5’-TCAAAAAGAACCCTAGGAACCCc-3’ | |||
| circRNA_5801 | hsa_circ_0062426 | F: 5’-UGGGUAGAGAAGGAGCUCAGAGGA-3’ | 181 |
| R: 5’-CTCTCTGCAGCCCTTTGTCTACCCA-3’ | |||
| circRNA_7386 | - | F: 5’-UGAGGCCCUUGGGGCACAGUGG-3’ | 166 |
| R: 5’-ACACTTAGTGCTTACAAGGGCCTCA-3’ | |||
| circRNA_7577 | hsa_circ_0006109 | F: 5’-UGCCCCACCUGCUGACCACCCUC-3’ | 166 |
| R: 5’-CCCGGTGG-CGGCTTGTGGGGCT-3’ |
Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis
GO analysis of differentially expressed circRNAs, including the domains of biological processes, cellular components, and molecular function were further analyzed with the Database for Annotation, Visualization, and Integrated Discovery (DAVID) Bioinformatics Resource v6.8. KEGG analysis of differentially expressed circRNAs was performed to find the pathways they participated in by the KEGG Ontology‐Based Annotation System (KOBAS) 2.0.
Construction of circRNA-microRNA networks
Functional sponging activity of circRNAs over miRNAs was analyzed by the prediction of miRNA target binding sites over the sequences of differentially expressed circRNAs. Enrichment results of total differentially expressed circRNAs were sorted by p value, and the potential connections between circRNAs and miRNAs were further explored by using Cytoscape 3.4.0 (http://cytoscape.org/).
Results
Analysis of differentially expressed circRNA
In total, 9426 circRNAs were detected by microarray, and the relevant characteristics are shown in Figure 1. There were 120 differentially expressed circRNAs between AF patients and controls (fold-change >2 and P<0.05) (Figure 2). Among these, 65 circRNAs were up-regulated (Table 2) and circRNAs 55 were down-regulated (Table 3).
Figure 1.
Characteristics of 9426 circRNAs analyzed by microarray. A. Length distribution of circRNAs. B. Chromosome- or -scaffold distribution of circRNAs. C. Distribution of exon-per-circRNA. D. Category of 9426 circRNAs.
Figure 2.
Differentially expressed circRNAs (fold change >2, and P<0.05) between AF group and control group. A. M-A plot for negative binomially distributed simulation data. B. Volcano plots are displayed for visualizing the differential expression of circRNAs. The red and green points in the plot represent the differentially expressed circRNAs with statistical significance. C. Box plots show the distribution of circRNAs for the two compared samples. The distributions were nearly the same after normalization. D. Hierarchical cluster analysis of all the deregulated circRNAs.
Table 2.
Upregulation of circular RNA
| circRNA_id | circbase_id | circRNA_Chr | Type | gene | FoldChange | P-Value |
|---|---|---|---|---|---|---|
| circRNA_0031 | hsa_circ_0008737 | Chr1 | sense-overlapping | CAMTA1 | 3.34 | 0.031 |
| circRNA_0095 | - | Chr1 | intronic | CAPZB | 8.01 | 0.011 |
| circRNA_0161 | - | Chr1 | antisense | THEMIS2 | 4.14 | 0.001 |
| circRNA_0312 | hsa_circ_0004877 | Chr1 | sense-overlapping | EPS15 | 4.06 | 0.011 |
| circRNA_0544 | - | Chr1 | intergenic | 10.15 | 0.017 | |
| circRNA_0685 | hsa_circ_0000160 | Chr1 | sense-overlapping | SUCO | 2.49 | 0.014 |
| circRNA_1166 | - | Chr10 | intronic | JMJD1C | 8.73 | 0.042 |
| circRNA_1402 | - | Chr11 | sense-overlapping | IFITM2 | 5.78 | 0.049 |
| circRNA_1415 | hsa_circ_0000274 | Chr11 | sense-overlapping | NUP98 | 5.24 | 0.047 |
| circRNA_1417 | - | Chr11 | intronic | NUP98 | 3.84 | 0.015 |
| circRNA_1513 | hsa_circ_0000302 | Chr11 | sense-overlapping | SPI1 | 3.06 | 0.040 |
| circRNA_1741 | hsa_circ_0005589 | Chr11 | sense-overlapping | ARCN1 | 4.21 | 0.012 |
| circRNA_1837 | - | Chr12 | sense-overlapping | KLRC2 | 9.3 | 0.025 |
| circRNA_2116 | hsa_circ_0004901 | Chr12 | sense-overlapping | APAF1 | 3.88 | 0.037 |
| circRNA_2294 | hsa_circ_0007547 | Chr13 | sense-overlapping | SKA3 | 4.18 | 0.011 |
| circRNA_2371 | - | Chr13 | sense-overlapping | ELF1 | 10.23 | 0.029 |
| circRNA_2482 | - | Chr13 | sense-overlapping | SLAIN1 | 3.86 | 0.020 |
| circRNA_2551 | - | Chr14 | intergenic | 3.8 | 0.029 | |
| circRNA_2616 | hsa_circ_0008002 | Chr14 | sense-overlapping | POLE2 | 3.24 | 0.030 |
| circRNA_2681 | hsa_circ_0032109 | Chr14 | sense-overlapping | PPM1A | 3.54 | 0.020 |
| circRNA_3140 | hsa_circ_0003916 | Chr15 | sense-overlapping | PIAS1 | 5.52 | 0.002 |
| circRNA_3337 | hsa_circ_0000672 | Chr16 | sense-overlapping | CLEC16A | 3.08 | 0.040 |
| circRNA_3359 | hsa_circ_0002771 | Chr16 | sense-overlapping | PARN | 3.64 | 0.024 |
| circRNA_3421 | hsa_circ_0008223 | Chr16 | sense-overlapping | XPO6 | 2.91 | 0.048 |
| circRNA_3448 | hsa_circ_0039161 | Chr16 | sense-overlapping | ITGAX | 8.18 | 0.000 |
| circRNA_4003 | hsa_circ_0005347 | Chr17 | sense-overlapping | BPTF | 5.73 | 0.034 |
| circRNA_4284 | hsa_circ_0008699 | Chr18 | exonic | ZNF516 | 5.63 | 0.008 |
| circRNA_4314 | hsa_circ_0004891 | Chr19 | sense-overlapping | CNN2 | 4.06 | 0.040 |
| circRNA_4656 | hsa_circ_0008847 | Chr2 | sense-overlapping | MBOAT2 | 3.76 | 0.015 |
| circRNA_4657 | hsa_circ_0000972 | Chr2 | sense-overlapping | MBOAT2 | 2.45 | 0.010 |
| circRNA_4661 | - | Chr2 | sense-overlapping | MBOAT2 | 5.89 | 0.022 |
| circRNA_4864 | hsa_circ_0001006 | Chr2 | sense-overlapping | RTN4 | 3.43 | 0.029 |
| circRNA_4959 | - | Chr2 | sense-overlapping | DYSF | 3.69 | 0.026 |
| circRNA_5325 | - | Chr2 | antisense | NOP58 | 3.21 | 0.045 |
| circRNA_5335 | hsa_circ_0003493 | Chr2 | sense-overlapping | CARF | 3.55 | 0.026 |
| circRNA_5399 | hsa_circ_0058514 | Chr2 | sense-overlapping | AGFG1 | 3.89 | 0.014 |
| circRNA_5664 | - | Chr20 | intronic | CTSZ | 6.47 | 0.024 |
| circRNA_5691 | hsa_circ_0061286 | Chr21 | sense-overlapping | USP25 | 3.08 | 0.045 |
| circRNA_5774 | hsa_circ_0008021 | Chr21 | sense-overlapping | PDXK | 13.23 | 0.004 |
| circRNA_5897 | hsa_circ_0008806 | Chr22 | sense-overlapping | CCDC134 | 5.19 | 0.022 |
| circRNA_5901 | hsa_circ_0001240 | Chr22 | exonic | NFAM1 | 6.34 | 0.033 |
| circRNA_5988 | hsa_circ_0001274 | Chr3 | sense-overlapping | PLCL2 | 8.66 | 0.046 |
| circRNA_6087 | hsa_circ_0001289 | Chr3 | sense-overlapping | SETD2 | 3.18 | 0.032 |
| circRNA_6264 | hsa_circ_0066959 | Chr3 | sense-overlapping | HCLS1 | 3.62 | 0.028 |
| circRNA_6360 | - | Chr3 | sense-overlapping | PLOD2 | 3.69 | 0.015 |
| circRNA_6574 | hsa_circ_0001394 | Chr4 | exonic | TBC1D14 | 4.04 | 0.004 |
| circRNA_6624 | - | Chr4 | exonic | TLR6 | 3.43 | 0.033 |
| circRNA_6644 | - | Chr4 | sense-overlapping | RBM47 | 3.13 | 0.050 |
| circRNA_6903 | hsa_circ_0071174 | Chr4 | sense-overlapping | LRBA | 3.18 | 0.032 |
| circRNA_6955 | hsa_circ_0001460 | Chr4 | sense-overlapping | NEIL3 | 3.25 | 0.044 |
| circRNA_6991 | - | Chr5 | intergenic | 5.86 | 0.002 | |
| circRNA_7097 | hsa_circ_0072697 | Chr5 | sense-overlapping | PPWD1 | 6.69 | 0.008 |
| circRNA_7571 | - | Chr6 | sense-overlapping | HLA-A | 28.22 | 0.005 |
| circRNA_7672 | hsa_circ_0003700 | Chr6 | sense-overlapping | FBXO9 | 6.12 | 0.030 |
| circRNA_7952 | hsa_circ_0004662 | Chr6 | sense-overlapping | SOD2 | 5.68 | 0.011 |
| circRNA_7964 | hsa_circ_0078665 | Chr6 | sense-overlapping | RNASET2 | 3.43 | 0.033 |
| circRNA_8132 | hsa_circ_0001707 | Chr7 | intronic | ABCA13 | 15.44 | 0.010 |
| circRNA_8233 | - | Chr7 | sense-overlapping | ANKIB1 | 3.43 | 0.037 |
| circRNA_8255 | hsa_circ_0007940 | Chr7 | sense-overlapping | ARPC1B | 3.62 | 0.028 |
| circRNA_8317 | hsa_circ_0082096 | Chr7 | sense-overlapping | ZNF800 | 4.88 | 0.031 |
| circRNA_8548 | hsa_circ_0006376 | Chr8 | sense-overlapping | HOOK3 | 3.31 | 0.043 |
| circRNA_8895 | hsa_circ_0003945 | Chr9 | sense-overlapping | UBAP2 | 3.37 | 0.015 |
| circRNA_9098 | hsa_circ_0008192 | Chr9 | sense-overlapping | PTBP3 | 4.22 | 0.014 |
| circRNA_9396 | hsa_circ_0001947 | ChrX | exonic | AFF2 | 7.79 | 0.001 |
| circRNA_9422 | hsa_circ_0008297 | ChrY | sense-overlapping | DDX3Y | 5.27 | 0.037 |
Table 3.
Downregulation of circRNA
| circRNA_id | circbase_id | circRNA_Chr | Type | gene | Fold Change | P-Value |
|---|---|---|---|---|---|---|
| circRNA_0259 | hsa_circ_0009142 | Chr1 | sense-overlapping | CAP1 | 3.41 | 0.029 |
| circRNA_0323 | hsa_circ_0012553 | Chr1 | sense-overlapping | ZCCHC11 | 2.88 | 0.014 |
| circRNA_0831 | - | Chr1 | sense-overlapping | LYPLAL1 | 4.38 | 0.024 |
| circRNA_0835 | hsa_circ_0004417 | Chr1 | sense-overlapping | LYPLAL1 | 9.69 | 0.023 |
| circRNA_0947 | hsa_circ_0002802 | Chr1 | sense-overlapping | ZNF124 | 6.37 | 0.042 |
| circRNA_0995 | hsa_circ_0000211 | Chr10 | sense-overlapping | SFMBT2 | 4.55 | 0.024 |
| circRNA_1111 | - | Chr10 | sense-overlapping | CCDC7 | 2.94 | 0.028 |
| circRNA_1292 | - | Chr10 | sense-overlapping | EXOSC1 | 3.23 | 0.015 |
| circRNA_1335 | hsa_circ_0000260 | Chr10 | sense-overlapping | SMC3 | 4.44 | 0.037 |
| circRNA_1450 | - | Chr11 | sense-overlapping | SERGEF | 3.47 | 0.010 |
| circRNA_1496 | - | Chr11 | sense-overlapping | PRR5L | 3.79 | 0.011 |
| circRNA_1693 | hsa_circ_0006208 | Chr11 | sense-overlapping | NPAT | 7.11 | 0.003 |
| circRNA_1786 | hsa_circ_0002881 | Chr12 | sense-overlapping | KDM5A | 3.08 | 0.019 |
| circRNA_1787 | hsa_circ_0024946 | Chr12 | sense-overlapping | KDM5A | 3.82 | 0.009 |
| circRNA_1800 | - | Chr12 | antisense | CACNA1C | 5.31 | 0.005 |
| circRNA_1834 | - | Chr12 | sense-overlapping | KLRC4-KLRK1 | 2.95 | 0.000 |
| circRNA_2370 | - | Chr13 | exonic | ELF1 | 3.09 | 0.021 |
| circRNA_2527 | hsa_circ_0004096 | Chr13 | sense-overlapping | RASA3 | 4.44 | 0.001 |
| circRNA_2683 | hsa_circ_0032116 | Chr14 | sense-overlapping | MNAT1 | 3.67 | 0.007 |
| circRNA_2773 | - | Chr14 | intergenic | 12.02 | 0.043 | |
| circRNA_2875 | - | Chr14 | intergenic | 3.06 | 0.030 | |
| circRNA_3138 | - | Chr15 | intronic | PIAS1 | 4.33 | 0.036 |
| circRNA_3307 | hsa_circ_0007788 | Chr16 | sense-overlapping | NMRAL1 | 10.03 | 0.023 |
| circRNA_3807 | - | Chr17 | sense-overlapping | CCL3L3 | 7.42 | 0.016 |
| circRNA_3830 | - | Chr17 | sense-overlapping | ERBB2 | 3.01 | 0.004 |
| circRNA_4184 | - | Chr18 | sense-overlapping | RNF138 | 6.13 | 0.000 |
| circRNA_4402 | - | Chr19 | sense-overlapping | ZNF564 | 3.51 | 0.014 |
| circRNA_4581 | hsa_circ_0003912 | Chr19 | exonic | DBP | 4.63 | 0.005 |
| circRNA_4624 | - | Chr19 | sense-overlapping | LILRA1 | 7.92 | 0.002 |
| circRNA_4631 | - | Chr19 | sense-overlapping | KIR2DL1 | 8.77 | 0.009 |
| circRNA_4648 | - | Chr2 | intergenic | 4.41 | 0.007 | |
| circRNA_4737 | - | Chr2 | exonic | GTF3C2 | 4.23 | 0.011 |
| circRNA_5440 | hsa_circ_0001112 | Chr2 | sense-overlapping | DGKD | 2.13 | 0.050 |
| circRNA_5625 | hsa_circ_0003998 | Chr20 | sense-overlapping | ARFGEF2 | 6.95 | 0.037 |
| circRNA_5801 | hsa_circ_0062426 | Chr22 | sense-overlapping | PPIL2 | 4.82 | 0.043 |
| circRNA_5996 | - | Chr3 | intergenic | 4.12 | 0.021 | |
| circRNA_6086 | - | Chr3 | sense-overlapping | SETD2 | 4.63 | 0.005 |
| circRNA_6610 | hsa_circ_0069397 | Chr4 | sense-overlapping | ARAP2 | 7.28 | 0.043 |
| circRNA_6775 | hsa_circ_0002782 | Chr4 | sense-overlapping | SLC39A8 | 5.38 | 0.019 |
| circRNA_6810 | hsa_circ_0007477 | Chr4 | sense-overlapping | PPA2 | 5.64 | 0.030 |
| circRNA_7032 | hsa_circ_0072380 | Chr5 | exonic | ZNF131 | 4.18 | 0.009 |
| circRNA_7335 | hsa_circ_0006716 | Chr5 | sense-overlapping | UBE2D2 | 3.66 | 0.032 |
| circRNA_7386 | - | Chr5 | sense-overlapping | SGCD | 4.37 | 0.007 |
| circRNA_7577 | hsa_circ_0006109 | Chr6 | sense-overlapping | C6orf136 | 2.29 | 0.028 |
| circRNA_7599 | - | Chr6 | sense-overlapping | HLA-DRB1 | 3.16 | 0.042 |
| circRNA_7797 | hsa_circ_0001638 | Chr6 | sense-overlapping | MFSD4B | 3.21 | 0.031 |
| circRNA_8031 | hsa_circ_0005519 | Chr7 | sense-overlapping | SNX13 | 8.57 | 0.045 |
| circRNA_8108 | - | Chr7 | sense-overlapping | TARP | 6.28 | 0.001 |
| circRNA_8280 | hsa_circ_0007395 | Chr7 | sense-overlapping | KMT2E | 12.57 | 0.033 |
| circRNA_8455 | - | Chr8 | intronic | ERI1 | 9.61 | 0.023 |
| circRNA_8731 | hsa_circ_0085438 | Chr8 | sense-overlapping | TBC1D31 | 5.03 | 0.002 |
| circRNA_8841 | - | Chr9 | sense-overlapping | KIAA2026 | 3.34 | 0.025 |
| circRNA_8857 | hsa_circ_0008732 | Chr9 | sense-overlapping | BNC2 | 3.62 | 0.022 |
| circRNA_9064 | - | Chr9 | sense-overlapping | NIPSNAP3A | 7.75 | 0.000 |
| circRNA_9326 | hsa_circ_0091175 | ChrX | sense-overlapping | BRWD3 | 3.69 | 0.020 |
qRT-PCR validation of differentially expressed circRNAs
Four upregulated circRNAs (circRNA_0031, circRNA_1837, circRNA_5901 and circRNA_7571) and four downregulated circRNAs (circRNA_2773, circRNA_5801, circRNA_7386 and circRNA_7577) were selected randomly for qRT-PCR validation to confirm the microarray results. As a result, all of 4 upregulated circRNAs (P<0.05 or P<0.01 for circRNA_0031, circRNA_1837, circRNA_5901 and circRNA_7571, respectively) and 3 out of 4 downregulated circRNAs (P<0.05 or P<0.01 for circRNA_5801, circRNA_7386 and circRNA_7577, respectively) showed a significantly different expression (Figure 3), which was consistent with microarray results.
Figure 3.
Quantitative reverse transcription-polymerase chain reaction analysis for validation of differentially expressed circRNAs. Compared with control group, *P<0.05 and **P<0.01.
GO and KEGG analysis of dysregulated circRNAs
To explore the functions of dysregulated circRNAs in AF patients compared with normal controls, GO gene enrichment analysis and KEGG pathway analysis were underwent by using DAVID and KOBAS. GO analysis indicated that the largest enriched biologic processes include phosphatidylethanolamine acyl chain remodeling, phosphatidylcholine acyl chain remodeling and phospholipid biosynthesis. The largest enriched cell composition includes special particles, cell surfaces, and extracellular areas. The largest enriched molecular functions include: 2-acyl glycerol-3-phosphate acyltransferase activity, transmembrane signal receptor activity, 1-acyl glycerol-3-phosphate acyltransferase activity (Table 4; Figure 4). The bubble map of top 20 pathway terms in KEGG enrichment analysis is shown in Figure 5. Differentially expressed circRNAs and the comparison of all genes at KEGG Level 2 (including cell growth and death, transcription and development) are shown in Figure 6. According to KEGG results, differentially expressed circRNA-related signaling pathways are mainly involved in inflammation and the immune system.
Table 4.
GO analysis of dysregulated circRNAs
| id | Term | Category | ListHits | P-Value | Enrichment_score |
|---|---|---|---|---|---|
| GO:0036152 | phosphatidylethanolamine acyl-chain remodeling | biological_process | 3 | 1.93E-05 | 17.4619 |
| GO:0036151 | phosphatidylcholine acyl-chain remodeling | biological_process | 3 | 3.45E-05 | 15.27917 |
| GO:0008654 | phospholipid biosynthetic process | biological_process | 3 | 0.000187 | 10.18611 |
| GO:0032436 | positive regulation of proteasomal ubiquitin-dependent protein catabolic process | biological_process | 4 | 0.000739 | 5.52467 |
| GO:0006955 | immune response | biological_process | 4 | 0.000799 | 5.432593 |
| GO:0010468 | regulation of gene expression | biological_process | 3 | 0.001816 | 5.685271 |
| GO:0006364 | rRNA processing | biological_process | 3 | 0.002949 | 4.989116 |
| GO:0045444 | fat cell differentiation | biological_process | 3 | 0.002949 | 4.989116 |
| GO:0050776 | regulation of immune response | biological_process | 3 | 0.004489 | 4.444848 |
| GO:0006954 | inflammatory response | biological_process | 4 | 0.004822 | 3.621728 |
| GO:0035580 | specific granule lumen | cellular_component | 3 | 8.84E-05 | 12.22333 |
| GO:0009986 | cell surface | cellular_component | 5 | 0.001757 | 3.880423 |
| GO:0005576 | extracellular region | cellular_component | 6 | 0.013994 | 2.339394 |
| GO:0031410 | cytoplasmic vesicle | cellular_component | 4 | 0.014215 | 2.785945 |
| GO:0000151 | ubiquitin ligase complex | cellular_component | 3 | 0.02707 | 2.628674 |
| GO:0005887 | integral component of plasma membrane | cellular_component | 5 | 0.062825 | 1.771498 |
| GO:0005634 | nucleus | cellular_component | 38 | 0.064788 | 1.201621 |
| GO:0005694 | chromosome | cellular_component | 3 | 0.071951 | 1.909896 |
| GO:0005730 | nucleolus | cellular_component | 8 | 0.081944 | 1.509053 |
| GO:0005886 | plasma membrane | cellular_component | 20 | 0.111773 | 1.245056 |
| GO:0047144 | 2-acylglycerol-3-phosphate O-acyltransferase activity | molecular_function | 3 | 1.39E-05 | 18.80513 |
| GO:0004888 | transmembrane signaling receptor activity | molecular_function | 4 | 3.29E-05 | 10.5147 |
| GO:0003841 | 1-acylglycerol-3-phosphate O-acyltransferase activity | molecular_function | 3 | 4.47E-05 | 14.38039 |
| GO:0001077 | transcriptional activator activity, RNA polymerase II proximal promoter sequence-specific DNA binding | molecular_function | 4 | 0.006333 | 3.39537 |
| GO:0046872 | metal ion binding | molecular_function | 21 | 0.011359 | 1.564229 |
| GO:0004872 | receptor activity | molecular_function | 3 | 0.013905 | 3.216667 |
| GO:0018024 | histone-lysine N-methyltransferase activity | molecular_function | 3 | 0.015178 | 3.134188 |
| GO:0003779 | actin binding | molecular_function | 5 | 0.019732 | 2.341635 |
| GO:0008022 | protein C-terminus binding | molecular_function | 3 | 0.036288 | 2.396732 |
| GO:0003714 | transcription corepressor activity | molecular_function | 3 | 0.058097 | 2.054342 |
Figure 4.
The results of Gene Ontology analysis. A. The top 10 neighbor coding genes of GO enrichment correspond to the upregulated circRNAs. B. The top 10 neighbor coding genes of GO enrichment correspond to the downregulated circRNAs. C. The top 10 neighbor coding genes of GO enrichment correspond to the total differentially expressed circRNAs.
Figure 5.
Bubble map of top 20 pathway terms in KEGG enrichment analysis of dysregulated circRNAs. A. The top 20 pathway terms of the upregulated circRNAs. B. The top 20 pathway terms of the downregulated circRNAs. C. Top 20 pathway terms of the total differentially expressed circRNAs.
Figure 6.
Results of KEGG enrichment analysis. A. The top 30 neighbor coding genes of KEGG enrichment correspond to the upregulated circRNAs. B. The top 30 neighbor coding genes of KEGG enrichment correspond to the downregulated circRNAs. C. The top 30 neighbor coding genes of KEGG enrichment correspond to the total differentially expressed circRNAs.
Construction of circRNA-miRNA co-expression network
According to the significantly altered circRNAs, the miRNA binding sites for each circRNA were explored by using Cytoscape and the significance of shared miRNAs for each circRNA-miRNA pair was estimated to construct the circRNA-miRNA co-expression network. Finally, the top 300 networks of circRNA-miRNA were extracted and drawn according to the p-value of the enrichment results of the total dysregulated circRNAs (Figure 7). As shown in Figure 7, circRNA_7571, circRNA_4648, circRNA_4631, and circRNA_2875 are the first four circRNAS with the most binding nodes in the co-expression network, which suggests that the four circRNAs may play important roles in AF, and could be regarded as key circRNAs. In addition, hsa-miR-328 was the largest node that interacted with circRNAs in the co-expression network, which suggested that circRNA-hsa-miR-123p may play a key role in the pathogenesis of AF.
Figure 7.
circRNA-miRNA coexpression network explored by using Cytoscape. The size of each node represents functional connectivity of each circRNA. The network consists of 37 circRNAs and 90 miRNAs. The red node represents circRNA and the green node represents miRNA. circRNA_7571, circRNA_4648, circRNA_4631 and circRNA_2875 were the four largest nodes in the network. hsa-miR-328 was the highest positive correlated miRNA in the networks.
Discussion
circRNA has received attention in the field of non-coding RNAs in recent years, and was first discovered in RNA viruses. circRNA is different from linear RNA in that it does not have a 5’ end cap and a 3’ end tail structure. It has a covalent closed loop structure and is not easily degraded by exonuclease, so it can stably exist in tissue or body fluids. In addition, the circRNA sequence is highly conserved, with a certain timing and tissue specificity [14], which suggests that circRNA may be highly conserved among different species and play an important role in the occurrence and development of disease. It also has potential as a biomarker of disease.
AF is the arrhythmia with the highest clinical incidence, which can induce heart failure, stroke, peripheral vascular embolism and other fatal cardiovascular and cerebrovascular diseases. So far, the pathogenesis of AF is still unclear, and specific biochemical diagnostic markers are also lacking [15,16]. In recent years, the important role of ncRNAs in ontogenetic development and the occurrence and development of various diseases has become increasingly apparent. More and more studies have shown that ncRNAs are closely related to the occurrence and maintenance of AF [17,18]. Further studies of circRNAs in AF patients can help clarify the pathogenesis of AF and find more stable AF markers. However, the relationship between circRNAs and AF is still unclear.
In this study, microarray was used to screen the differentially expressed circRNAs in peripheral blood monocytes between 4 AF patients and 4 healthy individuals. The results showed that 9426 circRNAs were abnormally expressed in AF patients, and 120 circRNAs were differentially expressed (fold-change >2 and P<0.05) compared with the healthy control group, of which 65 circRNAs were up-regulated and 55 circRNAs were down-regulated, indicating that these differentially expressed circRNAs may play an important role in the pathogenesis of AF. All of 4 upregulated circRNAs (circRNA_0031, circRNA_1837, circRNA_5901 and circRNA_7571) and 3 out of 4 downregulated circRNAs (circRNA_5801, circRNA_7386 and circRNA_7577) selected randomly for qRT-PCR validation showed a significantly different expression, which was consistent with microarray results. In order to further understand the roles of differentially expressed circRNAs in AF, the bioinformatics analysis of genes involved in the differentially expressed circRNAs was conducted. GO analysis of differentially expressed circRNAs suggests that the largest enriched biological process was mainly involved in phosphatidylethanolamine acyl chain remodeling, phosphatidylcholine acyl chain remodeling, and phospholipid biosynthesis. The largest enriched cell composition was mainly involved in special particles, cell surfaces, and extracellular areas. The largest enriched molecular functions were mainly involved in 2-acyl glycerol-3-phosphate acyltransferase activity, transmembrane signal receptor activity and 1-acyl glycerol-3-phosphate acyltransferase activity, etc. According to KEGG results, differentially expressed circRNAs-related signaling pathways were mainly involved in inflammation and the immune system. circRNA_7571, circRNA_4648, circRNA_4631 and circRNA_2875 were the first four circRNAs with the most combined nodes in the circRNAs-miRNA co-expression network. In addition, hsa-miR-328 was the largest node that interacts with circRNAs in the circRNAs-miRNA co-expression network. This suggests that circRNA-hsa-miR-328 may play a key role in the pathogenesis of AF. It has been reported that hsa-miR-328 plays a role in the proliferation and collagen production of atrial fibroblasts and is involved in the formation and maintenance of AF [19,20], which is consistent with our findings.
At present, studies on the circRNAs of AF are just beginning; further experiments are needed to confirm these circRNAs and related pathways and biologic processes discovered. In this study, for the first time, the relationship between circRNAs and AF was explored. Specific circRNAs were found to be expressed in AF peripheral blood monocytes, which not only provided a reliable standard basis for the diagnosis of AF at the gene level, but also provided a new research direction for studying the pathogenesis and prognosis of AF at the molecular level. However, the numbers of samples in this study were small; large-sample clinical validation of RT-PCR would be required and further analysis of differentially expressed circRNAs in peripheral monocytes had not been further explored in the present study. Further analysis and related experimental studies will further clarify the role of these circRNAs in the occurrence and development of AF.
Acknowledgements
The study was supported by Jiangsu Provincial Medical Innovation Team (grant No. CXTD2017015). Jiangsu Commission of Health, China (grant No. H201665), the Talent Foundation of Jiangsu Province, China (grant No. WSN-20) and Taizhou science and technology support plan (grant No. TS201729). The authors are thankful to Hai-Hui Sheng for technical assistance.
Disclosure of conflict of interest
None.
References
- 1.Koziolova NA, Polyanskaya EA, Chernyavina AI, Mironova SV. Atrial fibrillation in patients on dialysis therapy: epidemiology, prognosis and choice of anticoagulant therapy. Kardiologiia. 2019;59:72–83. doi: 10.18087/cardio.2019.12.n733. [DOI] [PubMed] [Google Scholar]
- 2.Komal S, Yin JJ, Wang SH, Huang CZ, Tao HL, Dong JZ, Han SN, Zhang LR. MicroRNAs: emerging biomarkers for atrial fibrillation. J Cardiol. 2019;74:475–482. doi: 10.1016/j.jjcc.2019.05.018. [DOI] [PubMed] [Google Scholar]
- 3.Xue XD, Huang JH, Wang HS. Angiotensin II activates signal transducers and activators of transcription 3 via rac1 in the atrial tissue in permanent atrial fibrillation patients with rheumatic heart disease. Cell Biochem Biophys. 2015;71:205–213. doi: 10.1007/s12013-014-0186-z. [DOI] [PubMed] [Google Scholar]
- 4.Zhang J, Youn JY, Kim AY, Ramirez RJ, Gao L, Ngo D, Chen P, Scovotti J, Mahajan A, Cai H. NOX4-dependent hydrogen peroxide overproduction in human atrial fibrillation and HL-1 atrial cells: relationship to hypertension. Front Physiol. 2012;3:140. doi: 10.3389/fphys.2012.00140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kaikkonen MU, Lam MT, Glass CK. Non-coding RNAs as regulators of gene expression and epigenetics. Cardiovasc Res. 2011;90:430–440. doi: 10.1093/cvr/cvr097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Mathy NW, Chen XM. Long non-coding RNAs (lncRNAs) and their transcriptional control of inflammatory responses. J Biol Chem. 2017;292:12375–12382. doi: 10.1074/jbc.R116.760884. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Boon RA, Jae N, Holdt L, Dimmeler S. Long noncoding RNAs: from clinical genetics to therapeutic targets? J Am Coll Cardiol. 2016;67:1214–1226. doi: 10.1016/j.jacc.2015.12.051. [DOI] [PubMed] [Google Scholar]
- 8.Beermann J, Piccoli MT, Viereck J, Thum T. Non-coding RNAs in development and disease: background, mechanisms, and therapeutic approaches. Physiol Rev. 2016;96:1297–1325. doi: 10.1152/physrev.00041.2015. [DOI] [PubMed] [Google Scholar]
- 9.Ruan Z, Sun X, Sheng H, Zhu L. Long non-coding RNA expression profile in atrial fibrillation. Int J Clin Exp Pathol. 2015;8:8402–8410. [PMC free article] [PubMed] [Google Scholar]
- 10.Stępień E, Costa MC, Kurc S, Drożdż A, Cortez-Dias N, Enguita FJ. The circulating non-coding RNA landscape for biomarker research: lessons and prospects from cardiovascular diseases. Acta Pharmacol Sin. 2018;39:1085–1099. doi: 10.1038/aps.2018.35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Holdt LM, Kohlmaier A, Teupser D. Molecular functions and specific roles of circRNAs in the cardiovascular system. Noncoding RNA Res. 2018;3:75–98. doi: 10.1016/j.ncrna.2018.05.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B, Castella M, Diener HC, Heidbuchel H, Hendriks J, Hindricks G, Manolis AS, Oldgren J, Popescu BA, Schotten U, Van Putte B, Vardas P. 2016 ESC guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Eur Heart J. 2016;37:2893–2962. doi: 10.1093/eurheartj/ehw210. [DOI] [PubMed] [Google Scholar]
- 13.Ruan ZB, Chen GC, Zhang R, Zhu L. Circular RNA expression profiles during the differentiation of human umbilical cord-derived mesenchymal stem cells into cardiomyocyte-like cells. J Cell Physiol. 2019;2019:1–12. doi: 10.1002/jcp.28310. [DOI] [PubMed] [Google Scholar]
- 14.Conn SJ, Pillman KA, Toubia J, Conn VM, Salmanidis M, Phillips CA, Roslan S, Schreiber AW, Gregory PA, Goodall GJ. The RNAbinding protein quaking regulates formation of circRNAs. Cell. 2015;160:1125–1134. doi: 10.1016/j.cell.2015.02.014. [DOI] [PubMed] [Google Scholar]
- 15.Ardhianto P, Yuniadi Y. Biomarkers of atrial fibrillation: which one is a true marker? Cardiol Res Pract. 2019;2019:8302326. doi: 10.1155/2019/8302326. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Severino P, Mariani MV, Maraone A, Piro A, Ceccacci A, Tarsitani L, Maestrini V, Mancone M, Lavalle C, Pasquini M, Fedele F. Triggers for atrial fibrillation: the role of anxiety. Cardiol Res Pract. 2019;2019:1208505. doi: 10.1155/2019/1208505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Savelieva I, Camm J. Update on atrial fibrillation: part II. Clin Cardiol. 2008;31:102–108. doi: 10.1002/clc.20136. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mikhailov AT, Torrado M. Interplay between cardiac transcription factors and non-coding RNAs in predisposing to atrial fibrillation. J Mol Med (Berl) 2018;96:601–610. doi: 10.1007/s00109-018-1647-4. [DOI] [PubMed] [Google Scholar]
- 19.da Silva AMG, de Araújo JNG, de Oliveira KM, Novaes AEM, Lopes MB, de Sousa JCV, Filho AAA, Luchessi AD, de Rezende AA, Hirata MH, Silbiger VN. Circulating miRNAs in acute new-onset atrial fibrillation and their target mRNA network. J Cardiovasc Electrophysiol. 2018;29:1159–1166. doi: 10.1111/jce.13612. [DOI] [PubMed] [Google Scholar]
- 20.Masè M, Grasso M, Avogaro L, Nicolussi Giacomaz M, D’Amato E, Tessarolo F, Graffigna A, Denti MA, Ravelli F. Upregulation of miR-133b and miR-328 in patients with atrial dilatation: implications for stretch-induced atrial fibrillation. Front Physiol. 2019;10:1133. doi: 10.3389/fphys.2019.01133. [DOI] [PMC free article] [PubMed] [Google Scholar]