Abstract

   Atrial fibrillation (AF) is a common arrhythmia in clinical practice,
   and its incidence is increasing year by year, which seriously affects
   the survival and prognosis of patients. In recent years, circRNAs has
   played an important role in the diagnosis and treatment of AF. The
   purpose of this study was to search for differentially expressed
   circRNAs(DEcircRNAs) in the serum of AF patients by analyzing the
   expression profile of existing chips, combining bioinformatics
   technology and in vitro experiments, and to explore the regulatory
   mechanism of circRNAs in the occurrence and development of AF. By using
   the AF datasets in the Gene expression omnibus (GEO) database, serum
   samples of patients with AF were collected, and the expression level of
   selected circRNAs was verified by qPCR. We found that the expression of
   four circRNAs was increased in the serum of patients with AF,
   suggesting that these four DEcircRNAs may be used as auxiliary
   diagnostic markers for AF. Bioinformatics predicts the related
   signaling pathways that differentially expressed genes may regulate in
   the occurrence and development of AF, providing a new theoretical basis
   for the molecular mechanism of the occurrence of atrial fibrillation
   and auxiliary diagnostic targets.

   Keywords: Atrial fibrillation, circRNAs, Biomarker, Diagnosis, ceRNA

Highlights

     * •
       4 DEcircRNAs were identified increased in the serum of patients
       with AF.
     * •
       Cystoscape was used to construct a ceRNA network for
       circRNAs-miRNAs-mRNAs in AF.
     * •
       Genetic elements are integral to AF, constituting an additional
       crucial mechanism.

1. Introduction

   Atrial fibrillation (AF) is one of the most common arrhythmias in
   clinical practice, and its incidence is increasing year by year, which
   has a higher risk of sudden death [[31]1,[32]2]. The risk factors of AF
   are closely related to cardiovascular diseases, among which organic or
   functional heart problems are more common. Furthermore, age, gender and
   genetic factors are also important factors leading to the occurrence of
   AF [[33]3,[34]4]. In addition, Coronary artery bypass grafting (CABG),
   mainly uses the body’s own blood vessels or vascular substitutes to
   connect the distal end of the narrowed or blocked coronary artery to
   the aorta, allowing blood to bypass the narrowed or blocked place and
   reach the ischemic site. Improve the blood supply of the myocardium,
   thereby alleviating the symptoms of angina pectoris and improving heart
   function [[35]5,[36]6]. However, AF is a relatively common arrhythmia
   in the early stage after coronary artery bypass grafting, and despite
   great advances in medical technology and postoperative care, the
   overall incidence of postoperative AF has not improved significantly
   [[37]7,[38]8]. Therefore, it is of great clinical significance to study
   the underlying mechanism of the occurrence and development of AF.

   CircRNAs are a class of non-coding RNAs, which are formed by laroses
   connecting the upstream 5′ and downstream 3’ ends to form a ring
   structure, and play an important regulatory role in many biological
   functions. Due to their unique circular structure, circRNAs are not
   susceptible to the influence of exonuclease and are stable in a variety
   of biological samples, making them potential biomarkers for auxiliary
   diagnosis and prognostic monitoring of a variety of diseases
   [[39]9,[40]10]. At present, the development of high-throughput
   sequencing technology and the advancement of bioinformatics analysis
   have laid the foundation for the research of circRNAs in AF
   [[41]11,[42]12]. Therefore, in-depth exploration of the regulatory
   pathways and mechanisms of circRNAs in AF provides a new basis and
   target for the diagnosis and treatment of AF, which has great clinical
   application value and prospects.

   Our hypothesis in this study was that the differentially expressed
   circRNAs in serum may be an influential factor in the occurrence of AF.
   We analyzed the circRNAs microarray expression profiles of serum
   samples from 15 patients with AF after CABG and myocardial tissue from
   3 patients with AF, and screened out 9 differential expressed circRNAs
   ([43]Table 1). Combined with real-time fluorescence quantitative PCR
   (RT-PCR), we determined 4 circRNAs with increased expression in serum
   of newly diagnosed AF patients. Considering that these DEcircRNAs are
   increased in the serum of patients with AF after CABG and also in the
   serum of patients with newly diagnosed AF, we hypothesized that these
   DEcircRNAs may be involved in the occurrence and progression of AF.
   Therefore, we used bioinformatics to construct competitive endogenous
   RNA (ceRNA) models of these DEcircRNAs to predict the potential
   regulatory mechanisms in the occurrence of AF, in order to provide new
   ideas for the diagnosis and treatment of AF.

Table 1.

   Differentially expressed circRNAs.
   circRNA circBase ID Gene symbol Position Genomic Length
   CircTADA2A hsa_circ_0043278 TADA2A chr17:35797838–35800763 2925
   Circ RPPH1 hsa_circ_0000511 RPPH1 chr14:20811282–20811431 149
   Circ RPPH1 hsa_circ_0006853 RPPH1 chr14:20811282–20811360 78
   Circ RPPH1 hsa_circ_0000514 RPPH1 chr14:20811305–20811436 131
   Circ ZNF646 hsa_circ_0000691 ZNF646 chr16:31093273–31093358 85
   Circ RPPH1 hsa_circ_0000512 RPPH1 chr14:20811282–20811436 154
   Circ TADA2A hsa_circ_0006220 TADA2A chr17:35800605–35800763 158
   Circ RPPH1 hsa_circ_0000515 RPPH1 chr14:20811305–20811534 229
   Circ FAM120B hsa_circ_0001666 FAM120B chr6:170626457-170639638 13181
   [44]Open in a new tab

2. Materials and methods

2.1. Collection and processing of gene expression profile data

   The [45]GSE129409 and [46]GSE97455 datasets were obtained from the Gene
   Expression Omnibus (GEO, [47]http://www.ncbi.nlm.nih.gov/geo) database.
   The [48]GSE129409 dataset collected the heart tissues from three AF
   patients and three healthy controls and profiled their circRNA
   expressions with circRNA Microarray. The [49]GSE97455 dataset contains
   15 plasma samples with postoperative AF and 15 plasma samples without
   postoperative AF. The R software (version 3.6.3) was used to process
   and analyze the original data. The limma package was used to perform
   background correction on the original data. The DESeq package was
   employed to detect DEcircRNAs with thresholds of |log2FC|≥1 and
   P-value≤0.05. False discovery rate (FDR) was used to correct the
   statistical significance of all p-values.

2.2. Sample collection and storage

   20 serums of AF patients and 20 serums of healthy donors were obtained
   from the Aoyang Hospital affiliated to Jiangsu University. In this
   study, all patients were newly diagnosed with AF, and the control group
   was healthy people. Specimens were collected from January 2021 to May
   2021. Serum samples were divided into RNase-free tubes and store at
   −80 °C until use to extract total RNA.

2.3. RNA extraction, reverse transcription, and quantitative real-time
polymerase chain reaction

   Total RNA was extracted using the serum RNA extraction Kit (BioTeke),
   and complementary DNA (cDNA) was synthesized by reverse transcription
   Kit (Thermo Fisher Scientific) and specific primers. The relative
   expression level of circRNAs was determined by Roche LightCycler 480.
   The internal control GAPDH was used to standardized the expression of
   circRNAs, and the relative expression were calculated by the 2^−ΔΔCt
   methods. Primer sequences are provided in [50]Table 2.

Table 2.

   Primer sequences used for qRT-PCR analysis of circRNA levels.
      Gene name       Primer sequences (F:5′-3′)   Primer sequences (R:5′-3′)
   hsa_circ_0043278 AGCCATTCCATTTCACTACTTCA        TCCTGCCAATTTCCAAAGCC
   hsa_circ_0000511 GAACAGACTCACGGCCAGCGAAGTGAGTTC
   GAACTCACTTCGCTGGCCGTGAGTCTGTTC
   hsa_circ_0001666 GATGACCATTCCAGATCCTTTTTCAAGAGA
   AAAGGATCTGGAATGGTCATCTTTTT
   hsa_circ_0006220 CTACCCTGCTGAACCTGAAACA         TTCTCACACTCCTCCTTGGTCTT
   GAPDH            CCCTTCATTGACCTCAACTA           TGGAAGATGGTGATGGGATT
   [51]Open in a new tab

2.4. Functional and pathway enrichment analysis

   The target genes of DEcircRNAs were predicted using the Cancer Specific
   CircRNA Database (CSCD, [52]http://gb.whu.edu.cn/CSCD/) based on the
   validated target module. The target genes of each circRNAs were
   subjected to Gene Ontology (GO) functional enrichment analysis and
   kyoto Encyclopedia of Genes and Genomes (KEGG) pathways enrichment
   analysis was performed using the ClusterProfiler package of R software.
   P-value<0.05 was considered significant.

2.5. Construction of ceRNA regulatory network

   Based on ceRNA theory, CircInteractome database was used to predict
   miRNAs that have regulatory relationships with circRNAs. miRDB database
   was used to predict mRNAs that have regulatory relationships with
   miRNAs. Meanwhile, differentially expressed miRNAs in [53]GSE68475 and
   mRNAs in [54]GSE64904 were analyzed. The predicted results were
   respectively intersected with the differentially expressed miRNAs and
   mRNAs screened from the databse, and the regulatory networks of
   differentially expressed circRNAs, miRNAs and mRNAs were visualized
   using Cytoscape software.

2.6. Statistical analysis

   SPSS statistics software version 20.0 was served as the statistical
   analysis. Mean ± standard deviation was used to represent measurement
   data with normal distribution. T-test was used to compare and analyze
   the data subject to normal distribution.

3. Results

3.1. Identification of differentially expressed genes (DEGs) in AF

   We obtained the circRNA Microarray from datasets [55]GSE129409 and
   [56]GSE97455 in the GEO database. The dataset [57]GSE12949 collected
   the heart tissues from three AF patients and three healthy controls,
   the dataset [58]GSE97455 contained 15 patients without postoperative
   (isolated off-pump coronary artery bypass grafting) AF and 15 patients
   with postoperative AF. With the cutoff criteria set at P value < 0.05
   and |log2FC|>1, we identified 1317 DEcircRNAs in dataset [59]GSE129409
   and 68 DEcircRNAs in dataset [60]GSE97455. The expression volcano maps
   and heat maps of the DEcircRNAs are shown in [61]Fig. 2A–D.
   Furthermore, we draw a Venn diagram to find the common changed circRNAs
   and 9 circRNAs were observed to be improved in the datasets
   [62]GSE129409 and [63]GSE97455 ([64]Fig. 1, [65]Table 1). Through
   comparative analysis of these datasets, we hypothesized that these 9
   DEcircRNAs may serve as auxiliary diagnostic markers for monitoring the
   occurrence and recurrence of AF.

Fig. 2.

   [66]Fig. 2
   [67]Open in a new tab

   4 DEcircRNAs in the serum of patients with AF. A: hsa_circ_0043278 was
   increased in AF; B: hsa_circ_0001666 was increased in AF; C:
   hsa_circ_0006220 was increased in AF; D: hsa_circ_0000511 was increased
   in AF.

Fig. 1.

   [68]Fig. 1
   [69]Open in a new tab

   DEcircRNAs in AF. A,C: volcano maps and heat maps of the DEcircRNAs in
   [70]GSE129409; B,D: volcano maps and heat maps of the DEcircRNAs in
   [71]GSE97455; E: The Venn diagram indicated that 9 circRNAs were found
   AF.

3.2. qRT-PCR validation of differentially expressed circRNAs

   qRT-PCR was used to verify the relative expression levels of 9
   differentially expressed circRNAs (hsa_circ_0043278, hsa_circ_0000511,
   hsa_circ_0006853, hsa_circ_0000514, hsa_circ_0000691, hsa_circ_0000512,
   hsa_circ_0006220, hsa_circ_0000515 and hsa_circ_0001666) in the serum
   of patients with AF, and the results showed that the expression levels
   of 4 circRNAs (hsa_circ_0043278, hsa_circ_0000511, hsa_circ_0006220 and
   hsa_circ_0001666) in the serum of patients with AF were higher than
   that of normal control group, and the differences were statistically
   significant. In the future, we will conduct in-depth studies on these 4
   differentially expressed circRNAs ([72]Fig. 2A–D).

3.3. GO and KEGG pathway analysis

   In order to study the regulatory mechanism of differentially expressed
   circRNAs in patients with AF and the occurrence of AF, we used the CSCD
   database to predict the downstream target genes with potential
   regulatory relationship with circRNAs, and predicted the regulatory
   pathways that circRNAs may participate in through GO and KEGG pathway
   analysis ([73]Fig. 3, [74]Fig. 4).

Fig. 3.

   [75]Fig. 3
   [76]Open in a new tab

   GO analysis of 4 DEcircRNAs. A: GO analysis of hsa_circ_0043278; B: GO
   analysis of hsa_circ_0001666; C: GO analysis of hsa_circ_0006220; D: GO
   analysis of hsa_circ_0000511.

Fig. 4.

   [77]Fig. 4
   [78]Open in a new tab

   KEGG analysis of 4 DEcircRNAs. A: KEGG analysis of hsa_circ_0043278; B:
   KEGG analysis of hsa_circ_0001666; C: KEGG analysis of
   hsa_circ_0006220; D: KEGG analysis of hsa_circ_0000511.

3.4. ceRNA regulatory network

   Most circRNAs are mainly located in cytoplasm, so circRNAs can compete
   with endogenous RNA as a miRNA molecular sponge to regulate miRNA
   target gene expression. Based on ceRNA theory, circRNA-miRNA-mRNA
   ternary interaction network is constructed and prognostic analysis is
   performed. Screening a new target for prognosis assessment after
   clinically assisted diagnosis of AF. First, 41 DEmiRNAs in the tissues
   of patients with AF were screened out using dataset [79]GSE68475, and
   132 DEmRNAs in the serum of patients with AF were screened out using
   dataset [80]GSE64904. Then, we used CircInteractome database to predict
   the miRNAs that have regulatory relationships with circRNAs, and
   intersected the differential miRNAs screened by [81]GSE68475 to obtain
   a total of 8 DEmiRNAs. Similarly, we used miRDB database to predict
   mRNAs that have regulatory relationships with miRNAs, and took
   intersection with differential mRNAs screened by [82]GSE64904. Finally,
   4 circRNAs, 8 miRNAs and 35 mRNAs were constructed for ceRNA network
   ([83]Fig. 5, [84]Fig. 6, [85]Table 3).

Fig. 5.

   [86]Fig. 5
   [87]Open in a new tab

   Venn diagram for circRNAs and miRNAs. A: CircInteractome database and
   [88]GSE68475 to predict the hsa_circ_0043278’s miRNAs; B:
   CircInteractome database and [89]GSE68475 to predict the
   hsa_circ_0000511’s miRNAs; C: CircInteractome database and [90]GSE68475
   to predict the hsa_circ_0006220’s miRNAs; D: CircInteractome database
   and [91]GSE68475 to predict the hsa_circ_0001666’s miRNAs.

Fig. 6.

   [92]Fig. 6
   [93]Open in a new tab

   The circRNA-miRNA-mRNA network of AF.

Table 3.

   The construction of ceRNA network through circRNAs-miRNAs-mRNAs.
   circRNA miRNA mRNA
   hsa_circ_0043278 miR-1207 OAS3,SGK1,FOSB, OASL,NUDT14,E2F2,SOCS3,GZF1
   miR-3192 STRADB,CD177,OCLN, OSBP2,ANK1,HES7,FOSB
   miR-3200 MAGED4B
   hsa_circ_0000511 miR-432 SIGLEC12
   hsa_circ_0006220 miR-187 SPIB
   miR-548 TSPAN16,PRKAR2B,OCLN, HEMGN,C2orf88,CLU,
   DNM3,PROS1,JPH1,ESX1,CD72,SLC4A10,IFI44L
   hsa_circ_0001666 miR-4254 OSBP2,SLC25A39,PCYT1B
   miR-345 RSAD2
   miR-548 TSPAN16,PRKAR2B,OCLN, HEMGN,C2orf88,CLU,
   DNM3,PROS1,JPH1,ESX1,CD72,SLC4A10,IFI44L
   [94]Open in a new tab

4. Discussion

   AF is the most common persistent arrhythmia. The incidence of AF
   increases with age. When AF occurs, blood is easy to stagnate in the
   heart chamber and form thrombus. Once the thrombus falls off, it can
   flow to the whole body with the blood, resulting in vascular embolism
   and even life-threatening. At present, AF is mainly treated by drug
   therapy, electrical cardioversion and surgery, but the prognosis of
   patients is still not ideal. With the rapid development of
   high-throughput sequencing technology and bioinformatics analysis,
   circRNAs played an important regulatory role in the occurrence and
   development of human diseases. The ceRNA hypothesis, originally
   proposed by Salmena et al., suggests that miRNAs can affect RNA
   expression at the post-transcriptional level by binding to target
   genes, and conversely, target genes can affect miRNAs [[95]13]. At
   present, a variety of ceRNAs have been found, including mRNA, circRNA,
   and lncRNA. All of them contained miRNA response elements (MREs).
   circRNA, as a new research hotspot in the ceRNA family, also has MREs,
   which can competitively inhibit the function of miRNA and play the role
   of RNA sponge, blocking and inhibiting the binding of miRNA and target
   genes, thereby leading to mutual regulation. CircRNAs can be stably
   expressed in all kinds of samples due to their unique ring structure,
   and there are differences in expression among different patients, which
   is expected to be a new strategy for the diagnosis and treatment of AF
   ([96]Fig. 7). Liu et al. using bioinformatics tools as well as
   literature review, constructed the circRNA-miRNA-mRNA ceRNA network
   related to AF fibrosis. qPCR valiated the expression of
   hsa_circ_0000672 in peripheral blood monocytes of patients with
   persistent AF was significantly higher than that of controls. The dual
   luciferase activity experiment confirmed that hsa_circ_0000672
   functioned as a sponge for miR-516a-5p and regulated the expression of
   its target gene TRAF6. hsa_circ_0000672 may indirectly regulate TRAF6
   by absorbing miR-516a-5p as ceRNA to participate in atrial fibrosis
   [[97]14]. Wu et al. found that hsa_circ_0099734 was significantly less
   expressed in atrial tissue of patients with AF than in normal tissue.
   Mechanistic studies have shown that has_circ_0099734 can competitively
   bind miR-499-5p to promote the expression of Kcnd1, Kcnd3, Scn5a and
   Kcnn3, thereby playing a protective role in the development of AF
   [[98]15]. In addition, circRNA was associated with cardiac fibrosis and
   plays an important role in the malignant progression of AF. By
   analyzing the differentially expressed circRNA in the plasma of 6
   patients with persistent AF and 6 patients with paroxysmal AF,
   hsa_circ_0004104 was confirmed to be down-regulated in patients with
   persistent AF. Mechanism study found that hsa_circ_0004104 was
   negatively correlated with TGF-β1. While TGF-β1 is a known marker of
   atrial fibrosis, low expression of has_circ_0004104 may affect
   myocardial fibrosis by regulation, possibly by targeting the MAPK and
   TGF-β pathways. These studies suggest that has_circ_0004104 is of great
   significance in predicting the malignant progression of AF [[99]16].
   Zhang et al. found that circCAMTA1 was upregulated in atrial muscle
   tissue from patients with AF and in angiotensin-II (Ang-II) -treated
   human atrial fibroblasts (HAFs). In vitro and in vivo experiments
   showed that knockdown of circCAMTA1 significantly inhibited
   Ang–II–induced HAFs proliferation and decreased the expression of
   atrial fibrosis-related genes, while reducing the incidence and
   duration of AF. The mechanism study showed that circCAMTA1 promoted
   Ang–II–induced atrial fibrosis by down-regulating the inhibitory effect
   of miR-214-3p on the expression of transforming growth factor β
   receptor 1 (TGFBR1). circCAMTA1/miR-214-3p/TGFBR1 axis may be a new
   target for the clinical treatment of AF [[100]17].

Fig. 7.

   [101]Fig. 7
   [102]Open in a new tab

   DEcircRNAs may be involved in the occurrence and progression of AF.

   Our study mainly aims to screen out circRNAs related to the occurrence
   and development of AF, and provide new ideas and directions for the
   diagnosis and treatment. By analyzing the circRNAs expression levels in
   serum of patients with AF after CABG, we found 68 differentially
   expressed circRNAs, and we speculated that these DEcircRNAs might be
   involved in the occurrence of AF. Then, we analyzed the expression
   level of circRANs in cardiac tissue of patients with AF, and screened
   out more than 1317 DEcircRNAs. By comprehensive analysis of the two
   datasets, a total of 9 circRNAs were screened out with the same
   expression level in the two databases. RT-PCR was used to verify the
   expression levels of these circRNAs in serum of newly diagnosed
   patients with AF, and the results showed that the expression levels of
   4 circRNAs were consistent with the sequencing results. Due to the
   particularity of the samples in these two datasets, the four
   differentially expressed circRNAs we screened and verified may be
   involved in the occurrence of AF, and can also be used to indicate the
   possibility of AF recurrence after CABG and guide clinical preventive
   measures to improve the prognosis of patients. In order to further
   investigate the potential regulatory mechanism of these 4 DEcircRNAs in
   the occurrence and development of AF, we performed the GO and KEGG
   pathway analyses of DEmRNAs in the ceRNA network. Using dataset
   [103]GSE68475 and [104]GSE64904 to screen out DEmiRNAs and DEmRNAs in
   the tissues and plasma of AF patients. DEcircRNAs may competitively
   bind miRNAs, thereby affecting the expression of downstream target
   genes, thereby regulating certain signaling pathways and affecting the
   prognosis of patients with AF. By analyzing the interaction between
   circRNA-miRNA-mRNA, and constructing the regulation model of ceRNA, it
   is expected to provide theoretical basis for the early diagnosis and
   prognosis monitoring of AF. Furthermore, as a drug carrier,
   nanomaterials have been widely studied in cancer therapy [[105]18].
   These materials not only have the advantages of stability and
   biocompatibility, but also have certain tumor targeting. In future
   studies, we will further expand the sample size and explore the
   reliability of these regulatory pathways through in vitro and in vivo
   experiments. At the same time, nanocarriers combined with circRNAs
   interference plasmids were used to target and efficiently transport
   plasmids to tumor sites, and it is expected that the circRNAs we
   screened can be better applied in clinical treatment and prognosis
   assessment.

5. Conclusions

   In this study, we identified four circRNAs that are up-regulated in the
   plasma of AF patients by PCR, and the highly expressed circRNAs may
   have a certain role in the diagnosis and prognosis evaluation of AF.
   Combined with bioinformatics analysis, the ceRNA regulatory network of
   differentially expressed circRNAs (DEcircRNA) was constructed. GO
   functional enrichment and KEGG pathway showed that these DEcircRNAs
   affected the occurrence and development of AF mainly by participating
   in PI3K-AKT signaling pathway and MAPK signaling pathway. These
   findings provide a theoretical basis for us to further investigate
   circRNA as an adjuvant therapy in AF.

Ethics statement

   Approval was obtained from the Medical Ethics Committee of Aoyang
   Hospital affiliated to Jiangsu University.

Consent to participate

   Informed consent was obtained from all individual participants included
   in the study.

Author contributions

   D- JJ and X-ZQdesigned the study and wrote the manuscript. Z-JB, LJ and
   W-LL performed the experiments and analyzed the data. All authors
   contribute to this study and approved the manuscript.

Funding

   This work was supported by the Science and Technology Project of
   Zhangjiagang City (ZKS2043), Zhangjiagang City Health Youth Science and
   Technology Project (ZJGQNKJ202113) and Suzhou Science and Technology
   Development Plan (SKJYD2021006).

Declaration of competing interest

   The authors declare that they have no known competing financial
   interests or personal relationships that could have appeared to
   influence the work reported in this paper.

Data availability

   Data will be made available on request.

References