Construction of microRNA and transcription factor regulatory network based on gene expression data in cardiomyopathy

Cardiomyopathy is a progressive myocardial disorder, usually leading to cardiovascular death or heart failure-related disability [1]. It develops at any age, in either sex, and in any population [2],[3]. The etiology of cardiomyopathy is highly complex and improving its treatments has become a research hotspot.

Using bioinformatics combined with gene expression data to identify potential therapeutic targets has shown great application prospects. The majority of the previous studies mainly focused on the analyses of differentially expressed genes (DEGs), without considering microRNAs (miRNAs) and transcription factors (TFs) that regulate the expression of DEGs. miRNAs are small non-coding RNAs that control various biological processes through affecting the stability and translation of target mRNAs. Previous studies have proposed several miRNAs as being involved in the pathogenesis of cardiomyopathy, such as miR-1 [4] and miR-21 [5]. TFs can regulate gene expression through binding to the cis-elements in target genes' promoter regions. TFs, such as MEF2, have been reported to be associated with cardiomyopathy [6]. Considering the important regulatory roles of miRNAs and TFs in the pathogenesis of cardiomyopathy, identification of miRNAs and TFs that enriched with target DEGs and construction of a double-factor regulatory network may provide new understanding of the molecular mechanism of cardiomyopathy.

In the current study, based on gene expression data from the Gene Expression Omnibus (GEO) database, we acquired DEGs, miRNAs and TFs that enriched with target DEGs and constructed a double-factor regulatory network. Our results may reveal the possible mechanism of cardiomyopathy at the transcription level.

A total of 1,680 DEGs were identified, including 963 down-regulated genes and 717 up-regulated ones in advanced cardiomyopathy patients.

As shown in Figure 1, the IPA results of canonical pathways showed that the DEGs were enriched for various pathways, among which glucocorticoid receptor signaling was the most significant. Molecular and cellular functions analysis revealed that `Cell Death and Survival' (P =2.53 × 10<sup>–24</sup> to 1.36 × 10<sup>–3</sup>) was the top molecular function affected by DEGs followed by `Cellular Growth and Proliferation' (P =3.37 × 10^–24 to 1.36 × 10^–3) (Table 1). `Cardiovascular System Development and Function' (P =4.17 × 10^–12 to 1.20 × 10^–3) was the top physiological function mediated by DEGs (Table 1).

The results of miRNA and TF enrichment analysis are listed in Table 2. A network was constructed to illustrate the regulatory relationships (Figure 2). Considering the important role of SRF and MEF2C in the progression of cardiomyopathy, a subnetwork under their regulation was constructed and the biological functions they may affect were also indicated (Figure 3).

In the current study, based on GSE5406 from the GEO database, 1,680 DEGs were identified between advanced cardiomyopathy patients and healthy controls. DEGs were mapped into the Ingenuity Knowledge Base and several canonical pathways were screened out, with glucocorticoid receptor signaling as the most significant. Glucocorticoid receptors are involved in cardiovascular homeostasis [12] and glucocorticoid could protect rodent hearts from ischemia/reperfusion injury [13]. Our results confirmed the essential roles of this pathway in the pathogenesis of cardiomyopathy. Molecular and cellular functions analysis revealed that `Cell Death and Survival' (P =2.53 × 10^–24 to 1.36 × 10^–3) was the top molecular function affected by DEGs, suggesting that dysregulation of the cell function may contribute to the progressive heart failure induced by cardiomyopathy.

TFs or miRNAs that DEGs significantly enriched were identified and a double-factor regulatory network was constructed, including seven TFs and two miRNAs (Figure 2). All these miRNAs and TFs have previously been implicated in cardiomyogenesis or cardiac function. Considering their important regulatory roles in controlling various biological processes, these miRNAs and TFs may provide new avenues for the therapeutic strategies of cardiomyopathy. In addition, since all these miRNAs and TFs were detected in patients suffered from New York Heart Association class 3 to 4 symptoms, they may also be considered as prognosis markers in clinical practice. A subnetwork under the regulation of SRF and MEF2C was also constructed and the functions they may affect were also indicated (Figure 3) considering the considerable evidence in support of their involvement in cardiomyopathy [14]-[17]. As shown in Figure 3, both SRF and MEF2C may inhibit the expression of MYH6 and CNN1. MYH6 is a famous cardiac muscle myosin gene and mutations of this gene are implicated in cardiomyopathy in various studies [18]. Inhibition of CNN1 expression may disrupt its suppression of cardiomyopathy through the εPKC pathway [19]. In addition, both SRF and MEF2C may regulate the expression of NPPA, which was overexpressed in patients in our results. This gene is involved in many cardiac dysfunctions (Figure 3), including myocardial infarction, heart failure, and high ventricular pressure [20],[21].

In summary, with DEGs identified by gene chip data generated by high-throughput technologies from the GEO database, we constructed a cardiomyopathy double-factor regulatory network and revealed the possible mechanism on regulation level. A subnetwork under the regulation of SRF and MEF2C was also constructed to illustrate their possible regulatory roles. Our results may facilitate further therapeutic studies of cardiomyopathy.

LW conceived of the study and drafted the manuscript. JH, HX and MS performed the statistical analysis. JW, QJ and HY revised the manuscript. All authors read and approved the final manuscript.