Importance of Mitochondrial-Related Genes in Dilated Cardiomyopathy Based on Bioinformatics Analysis

Abstract

We designed this study to identify potential key protein interaction networks, genes, and correlated pathways in dilated cardiomyopathy (DCM) via bioinformatics methods. We selected the GSE3586 microarray dataset, consisting of 15 dilated cardiomyopathic heart biopsy samples and 13 nonfailing heart biopsy samples. Initially, the GSE3586 dataset was downloaded and was analyzed with the limma package to identify differentially expressed genes (DEGs). A total of 172 DEGs consisting of 162 upregulated genes and ten downregulated genes in DCM were selected by the criterion of adjusted Pvalues less than 0.01 and the log2-fold change of 0.6 or greater. Gene Ontology functional enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed to view the biological processes, cellular components, molecular function, and KEGG pathways of the DEGs. Next, protein-protein interactions were constructed, and the hub protein modules were identified. Then we selected the key genes DLD, UQCRC2, DLAT, SUCLA2, ATP5A1, PRDX3, FH, SDHD, and NDUFV1, which are involved in a wide range of biological activities, such as the citrate cycle, oxidation-reduction processes and cellular respiration, and energy derivation by oxidation of organic compounds in mitochondria. Finally, we found that currently there are no related gene-targeting drugs after exploring the predicted interactions between key genes and drugs, and transcription factors. In conclusion, our study provides greater understanding of the pathogenesis and underlying molecular mechanisms in DCM. This contributes to the exploration of potential gene therapy targets.

Main article text

Abbreviations: DCM, dilated cardiomyopathy; GEO, Gene Expression Omnibus; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; PPI, protein-protein interaction.

Introduction

Dilated cardiomyopathy (DCM) is a primary myocardial disease of unknown cause but frequently has a genetic background. The disease is characterized by left or right ventricular or bilateral ventricular enlargement with reduced ventricular systolic function, with or without congestive heart failure, and usually affects young patients with few comorbidities. Ventricular or atrial arrhythmia is more prevalent. The disease is progressive, and death can occur at any stage [1, 2]. The concept of DCM is as a dynamic disease requiring constant optimization of medical and nonpharmacological evidence-based treatments [3]. The advent of next-generation sequencing has contributed to the discovery of a large number of genes and alleles related to DCM, which makes it possible to intervene at the molecular level, especially with regard to patients in whom routine treatment is not suitable. Moreover, cascade genetic testing can identify those who are at risk or have early-stage disease, offering the opportunity for early intervention [4, 5]. Therefore, genetic testing and gene therapy in DCM will play an increasingly important role.

Although there have numerous efforts to understand the genetic mechanisms of initiation and progression of DCM, knowledge is still lacking regarding the key genes involved in both the occurrence and the progression of DCM and how to treat and control deterioration with DCM through targeted drugs. Therefore, it is important and urgent to uncover the mechanisms of DCM and develop novel therapeutic routes. Gene chip or expression profiling is a gene-level detection technique that makes it possible to detect expression of the entire genome in the same sample in a single experiment [6, 7]. A large amount of correlated genomic data from DCM patients has been deposited in public databases [8, 9]. Using the Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) public database, we identified differentially expressed genes (DEGs) of dilated cardiomyopathic heart and nonfailing heart biopsy samples and performed Gene Ontology (GO) and signaling pathway analyses.

We downloaded the GSE3586 soft formatted family file and series matrix file from the GEO. Using the RStudio-installed limma package, we identified DEGs between gene expression profiles of biopsy samples of dilated cardiomyopathic hearts, profiled at the time of transplantation, and those of biopsy samples of nonfailing hearts that were not transplanted because of palpable coronary calcifications [10, 11]. Then GO and pathway enrichment analyses were performed on the DAVID website (https://david.ncifcrf.gov) [12]. Through this work, we discovered the important biological functions and pathways in DCM, which can help to determine the hub genes and their roles in diagnosis, prognosis, and therapeutic targets.

Materials and Methods

Microarray Data

The GSE3586 microarray dataset [13] was downloaded from the GEO [14, 15]; these data are based on two genome-wide expression studies involving complementary DNA and short-oligonucleotide microarray platforms. The GSE3586 dataset contains data from 13 dilated cardiomyopathic hearts and 15 nonfailing donor hearts.

Data Preprocessing

After the GSE3586 dataset had been downloaded, we found that the expression matrix retrieved from the GEO had been standardized and probe identification numbers were transformed into gene symbols. If a gene’s symbols corresponded with multiple probe IDs, the expression level of that gene was represented by the mean of the probes.

Identification of DEGs

DEG analysis refers to identification of genes that are expressed at significantly different levels by multiple modes of analysis [16]. The txt and soft data GSE3586 files were analyzed with RStudio and the R package limma to identify the DEGs between dilated cardiomyopathic heart biopsy samples and nonfailing heart biopsy samples [16–18]. We identified the DEGs by using a threshold of the log2-fold change of 0.6 or greater and adjusted Pvalues less than 0.01 for subsequent analysis.

GO and Pathway Enrichment Analysis

To further understand the biofunctions of the DEGs, GO analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were performed through multiple online tools to annotate genes and gene products [19], identify the biological roles of the DEGs [20], and interconnect the genomic information with high-level functional information [21]. DAVID, among them, is a website with gene annotation and visualization, and provides gene attributes. P<0.05 was considered statistically significant and was used as the cutoff criterion [21, 22].

Integration of the Protein-Protein Interaction Network and Module Analysis

Afterward, we constructed a connection diagram to show DEG-encoded proteins and protein-protein interaction (PPI) information to understand the potential functions of a certain protein (STRING, https://string-db.org) [23]. We chose a minimum required interaction score of 0.400 (medium confidence) as the significance limitation. Then the PPI network acquired from the STRING database was imported into Cytoscape with the help of the Molecular Complex Detection (MCODE) plug-in in the software to mark the pivotal modules in huge complex PPI networks [24]. Except for a K-core of 4, we used the default parameters of MCODE.

Drug-Gene Interaction and Prediction of Transcription Factors

DGIdb (the drug-gene interaction database; https://dgidb.org) is an online database that integrates several databases, including ChEMBL and Ensembl, to search for drug-gene interactions by gene symbols. We imported the nine hub genes of the pivotal modules into DGIdb [25] to get the drug-gene interaction results. Furthermore, the prediction of related transcription factors was performed with PASTAA [26] (a program for predicting transcription factors that regulate a user-defined set of genes.)

Results

Identification of DEGs

The GSE3586 dataset was analyzed with RStudio to identify the DEGs. By comparing dilated cardiomyopathic heart biopsy samples versus nonfailing heart biopsy samples, we identified a total of 172 genes, including 162 upregulated DEGs and ten downregulated DEGs (Table 1 and Figure 1), using adjusted Pvalues less than 0.01 and a modulus of the log₂ fold change of 0.6 or greater as criteria.

Figure 1

Volcano Plot for Differentially Expressed Genes.

Table 1

Differentially Expressed Genes were Identified From the GSE3586 Dataset, Including 162 Upregulated Genes and Ten Downregulated Genes in Dilated Cardiomyopathic Heart Biopsy Samples Compared with Nonfailing Heart Biopsy Samples.

Status	Gene
Upregulated	DLAT, GJA5, CDC42EP3, SEC31L1, PAIP2, UNR, PRSS11, ZNFN1A5, YWHA, SUCLA2, ATP6V1D, ZNF83, ECHDC1, GHITM, CPT1B, Ells1, ASAH1, HNRPC, CRYM, FLJ14054, ANKMY2, PRDX3, ACO2, TTC12, FBXO8, TP53INP2, DKFZp434I1610, NFE2L2, ZNF567, DLD, HIST1H4E, VDAC3, MLYCD, EIF4A2, FH, RPL22, TOP2B, MGC2714, NDUFA5, KIF5B, FLJ13710, ZNF23, ZNF600, TAX1BP1, C2orf25, LDB3, APLP2, E2-230K, SCHIP1, EXT1, COPS8, ZNF165, SGCG, UQCRC2, DNM3, HSPA2, STK38L, PSMC6, NDUFV1, HSPCB, SSPN, TARS, UBB, NFIA, P29, CRYZ, MPP4, CASQ2, NNT, MCC, CDH2, FLJ26175, AKAP13, BCAR1, ARIH2, GJA1, ARGBP2, EIF4G2, MGC7036, SDHD, MORF4L2, SGCD, SMARCC1, KLHDC2, CANX, NIPSNAP3A, CAP2, PTPN11, MFN2, TXNIP, MATR3, SLC40A1, HBP1, HDLBP, NQO3A2, ZNF7, GNAS, DAP3, AK3, LAPTM4B, ODC1, ZNF267, HMGN2, HIST2H2BE, KIAA1618, CGI-51, STAT1, BRP44L, PPARA, MAOA, ATP5A1, NOL5A, MB, MAP2K6, KIAA0774, SH3BP2, PSME1, LUM, PPP1R14C, LMOD2, RBBP7, GUP1, FLJ22662, OCIL, ZFP106, SLC8A1, NFE2L1, ERBP, ELOVL6, INSIG1, PROS1, DNCL1, LTBP2, SLC16A1, HMGB3, PKP2, NPM1, FHOD3, DSP, FLJ21127, MGC15619, HSPA1L, AF1Q, CD36, FYCO1, HSPA8, KIF5C, NDRG4, LOC339924, FGF1, FLJ20152, EEF1A1, LOC92235, PDE4DIP, NRG1, C6orf142, TNNI3, CSDA, TNNC2, SFRP1, PAM, ACPP, ABLIM1, ZNF364, ACSL1, CA3, MYOZ2, FHL2
Downregulated	HDAC9, C18orf22, PAPD5, PRAME, DEPC-1, IFITM3, ARFIP1, IFITM2, C3, KCNIP2, CCL2

GO Enrichment Analysis

We uploaded the DEGs to the DAVID website to identify the significant GO terms and classified them into three functional categories: biological processes, molecular functions, and cell components (Figure 2). In the biological process group, the upregulated DEGs were enriched mainly in genes associated with the heart development and oxidation-reduction processes. The downregulated DEGs were enriched in genes associated with the regulation of vasculature development response to biotic stimulus and positive regulation of apoptotic cell clearance. In the cell component group, the upregulated DEGs were enriched mainly in genes associated with the actin cytoskeleton and intercalated disk, but as for the downregulated DEGs, we did not find any useful information. In the molecular function group, the upregulated DEGs were enriched mainly in genes associated with oxidoreductase activity, acting on NAD(P)H, and wide pore channel activity. For the downregulated DEGs, there were no available significant molecular functions (P<0.05) (Table 2).

Figure 2

Gene Ontology Terms and Kyoto Encyclopedia of Genes and Genomes Pathways of Differentially Expressed Genes Significantly Enriched in Dilated Cardiomyopathy.

TCA, tricarboxylic acid.

Table 2

Gene Ontology Analysis of Differentially Expressed Genes (DEGs) Associated with Dilated Cardiomyopathy.

Category	Term	Count	P
Upregulated DEGs
GOTERM_BP_FAT	GO:0060047 heart contraction	10	7.07×10−7
GOTERM_BP_FAT	GO:0003015 heart process	10	8.08×10−7
GOTERM_BP_FAT	GO:0055114 oxidation-reduction process	18	3.25×10−6
GOTERM_BP_FAT	GO:0009060 aerobic respiration	7	3.79×10−6
GOTERM_CC_FAT	GO:0015629 actin cytoskeleton	14	2.28×10−5
GOTERM_CC_FAT	GO:0014704 intercalated disc	5	1.70×10−4
GOTERM_CC_FAT	GO:0043292 contractile fiber	8	4.71×10−4
GOTERM_MF_FAT	GO:0016651 oxidoreductase activity, acting on NAD(P)H	5	0.003076304
GOTERM_MF_FAT	GO:0022829 wide pore channel activity	3	0.005179203
GOTERM_MF_FAT	GO:0019899 enzyme binding	11	0.008728806
Downregulated DEGs
GOTERM_BP_FAT	GO:2000427 positive regulation of apoptotic cell clearance	2	3.69×10−3
GOTERM_BP_FAT	GO:2000425 regulation of apoptotic cell clearance	2	5.89×10−3
GOTERM_BP_FAT	GO:0045765 regulation of angiogenesis	3	6.16×10−3
GOTERM_BP_FAT	GO:1901342 regulation of vasculature development	3	7.37×10−3
GOTERM_BP_FAT	GO:0009607 response to biotic stimulus	4	1.07×10−2
GOTERM_BP_FAT	GO:0043277 apoptotic cell clearance	2	1.76×10−2
GOTERM_BP_FAT	GO:0010574 regulation of vascular endothelial growth factor production	2	1.83×10−2
GOTERM_BP_FAT	GO:0001525 angiogenesis	3	1.98×10−2
GOTERM_BP_FAT	GO:0050766 positive regulation of phagocytosis	2	2.27×10−2
GOTERM_BP_FAT	GO:0048514 blood vessel morphogenesis	3	2.93×10−2

KEGG Pathway Analysis

The associated KEGG pathways for the most significantly enriched DEGs were identified by means of the DAVID website and are listed in Figure 2. In the KEGG pathway group, the most enriched signaling pathways for the upregulated DEGs were in the citrate cycle (also known as the tricarboxylic acid cycle) and arrhythmogenic right ventricular cardiomyopathy. The downregulated DEGs were enriched only in Chagas disease (also known as American trypanosomiasis), at a cutoff of P _adjusted<0.05 (Table 3).

Table 3

Pathway Enrichment Analysis of Differentially Expressed Genes (DEGs).

Pathway	Term	Count	P	Genes
Upregulated DEGs
KEGG_PATHWAY	bta01100: metabolic pathways	18	0.038370751	UQCRC2, ODC1, NDUFA5, ACO2, MAOA, DLAT, ATP6V1D, ASAH1, ACSL1, NNT, MLYCD, FH
KEGG_PATHWAY	bta01130: biosynthesis of antibiotic	8	0.002208011	NDUFV1, SDHD, DLD, ATP5A1, SUCLA2, EXT1
KEGG_PATHWAY	bta05012: Parkinson’s disease	7	0.002189488	ODC1, ACO2, DLD, SDHD, AK3, DLAT, SUCLA2, FH
KEGG_PATHWAY	bta00020: citrate cycle (TCA cycle)	6	5.68×10−6	UQCRC2, NDUFA5, NDUFV1, SDHD, UBB, ATP5A1, VDAC3
KEGG_PATHWAY	bta05412: arrhythmogenic right ventricular cardiomyopathy	6	2.59×10−4	ACO2, DLD, SDHD, DLAT, SUCLA2, FH
KEGG_PATHWAY	bta01200: carbon metabolism	6	0.002714139	SGCG, PKP2, GJA1, DSP, SGCD, CDH2, DLD ACO2,SDHD, DLAT, SUCLA2, FH
Downregulated DEGs
KEGG_PATHWAY	bta05142: Chagas disease	2	0.044623307	CCL2, C3

TCA, tricarboxylic acid.

PPI Analysis and Gene Module Analysis

With the aid of the STRING online database and the Cytoscape software platform, we formed a complex PPI network for genes, consisting of 85 nodes and 162 edges (Figure 3A) and accounting for 49.42% of the DEGs. According to the filtering of node degree 8 criteria, the top ten genes were UBB, ATP5A1, FH, UQCRC2, DLD, DLAT, SDHD, NDUFV1, PRDX3, and HSP90AB1. Then we identified the key module (nine nodes and 34 edges, Figure 3B) from the PPI network with use of the MCODE plug-in. The functional annotation of the hub genes is listed in Table 4 and visualized in Figure 4. Our results show the functions of nine hub genes and their probable role in DCM via use of DAVID.

Figure 3

Protein-Protein Interaction (PPI) Networks of Differentially Expressed Genes (DEGs).

(A) By means of the STRING online database, 172 genes/nodes were filtered into a DEG-PPI network. (B) The key module from the PPI network.

Figure 4

Functional Enrichment Analysis of Nine Candidate Hepatocellular Dilated Cardiomyopathy–Related Genes (green indicates that the gene is enriched in the item, black indicates that the gene is not enriched in the item).

Color images are available online. TCA, tricarboxylic acid.

Table 4

Functional and Pathway Enrichment of Nine Hub Genes.

Category	Term	Genes
GOTERM_BP_FAT	GO:0055114 oxidation-reduction process	UQCRC2, NDUFV1, DLD, PRDX3, FH
GOTERM_BP_FAT	GO:0045333 cellular respiration	UQCRC2, NDUFV1, DLD, FH
GOTERM_BP_FAT	GO:0015980 energy derivation by oxidation of organic compounds	UQCRC2, NDUFV1, DLD, FH
GOTERM_CC_FAT	GO:0005739 mitochondrion	UQCRC2, NDUFV1, DLD, SDHD, PRDX3, DLAT, ATP5A1, SUCLA2, FH
GOTERM_CC_FAT	GO:1990204 oxidoreductase complex	UQCRC2, NDUFV1, DLD, DLAT
GOTERM_MF_FAT	GO:0030523 dihydrolipoamide S-acyltransferase activity	DLAT
GOTERM_MF_FAT	GO:0016651 oxidoreductase activity, acting on NAD(P)H	NDUFV1, DLD
KEGG_PATHWAY	mcc00020: citrate cycle (TCA cycle)	DLD, SDHD, DLAT, SUCLA2, FH
KEGG_PATHWAY	mcc01200: carbon metabolism	DLD, SDHD, DLAT, SUCLA2, FH
KEGG_PATHWAY	mcc01100: metabolic pathways	UQCRC2, NDUFV1, DLD, SDHD, DLAT, ATP5A1, SUCLA2, FH

TCA, tricarboxylic acid.

Drug-Gene Interactions and Prediction of Transcription Factors

The drugs with the highest scores for each gene were identified with use of DGIdb. Among them, the SDHD and NDUFV1 genes correspond to the drugs succinic acid and CHEMBL1161866, respectively. However, no interactions were found for PRDX3 (Table 5). Meanwhile the top 20 transcription factors predicted by hub gene expression are listed in Table 6, and on the basis of the Pvalue and association score that were used to evaluate the correlation between the disease and transcription factors, we identified the transcription factor–binding site of the highest-scoring transcription factor (Figure 5). Neurofibromin1 could be a potential target for gene therapy.

Figure 5

Predicted Transcription Factor–Binding Site of Neurofibromin 1.

Table 5

Gene-Drug Interactions.

Gene	Drug	Interaction types	Sources	Score
DLD	CHEMBL1161866	NA	DrugBank	4
UQCRC2	Coenzyme Q2	NA	DrugBank	2
DLAT	CHEMBL1161866	NA	DrugBank	4
SUCLA2	Succinic acid	NA	DrugBank	4
ATP5A1	Quercetin	NA	DrugBank	2
FH	Ditiocarb	NA	NCI	2
SDHD	Succinic acid	NA	DrugBank	6
NDUFV1	CHEMBL1161866	NA	DrugBank	6
PRDX3	No interactions found

NA, not available; NCI, National Cancer Institute.

Table 6

Top 20 transcription Factors Predicted by Hub Gene Expression.

Rank	Matrix	Transcription factor	Full name	Association score	P
1	NF1_Q6	Nf-1	Nuclear factor 1	4.48	2.19×10−4
2	NF1_Q6_01	Ctf-1, Ctf-2	Cardiotrophin factors 1, Cardiotrophin factors 2	4.48	2.19×10−4
3	HES1_Q2	Hes-1	Hairy-related factors 1	3.255	2.52×10−3
4	IK3_01	N/A	N/A	3.252	2.52×10−3
5	ZIC2_01	Zic2, Zic2	Zinc finger factors 2	2.792	6.74×10−3
6	MAX_01	Max1	MYC associated factor X 1	2.577	1.07×10−2
7	HFH4_01	Foxf1, Foxj1	Forkhead box (FOX) factors f1, Forkhead box (FOX) factors j1	2.521	1.33×10−2
8	MYCMAX_03	Max, C-myc	MYC associated factor X, MYC proto-oncogene	2.414	1.70×10−2
9	CMYB_01	N/A	N/A	2.372	1.80×10−2
10	MYCMAX_02	Max1, C-myc	MYC associated factor X 1, MYC proto-oncogene,	2.346	1.83×10−2
11	ZBRK1_01	N/A	N/A	2.34	1.83×10−2
12	ARNT_02	Arnt	Aryl hydrocarbon receptor nuclear translocator	2.332	1.85×10−2
13	USF_01	Usf1	upstream transcription factor 1	2.332	1.85×10−2
14	USF_02	Usf1	upstream transcription factor 1	2.332	1.85×10−2
15	USF_Q6_01	Usf-1, Usf1	upstream transcription factor 1	2.332	1.85×10−2
16	USF_Q6	Usf1, Usf2a	upstream transcription factor 1, upstream transcription factor 2a	2.331	1.85×10−2
17	COUP_01	Coup-tf1, Hnf-4alpha1	Coup-Transcription factor 1, hepatocyte nuclear factor 4 alpha1	2.221	2.49×10−2
18	HFH1_01	Hfh-1	forkhead box Q1	2.22	2.49×10−2
19	STRA13_01	Stra13	stimulated by retinoic acid 13	2.196	2.49×10−2
20	P300_01	P300	zinc finger protein 300	2.125	3.12×10−2

NA, not available.

Discussion

DCM is a heart muscle disease characterized by left ventricular or biventricular dilatation and systolic dysfunction [27]. The implementation of optimal pharmacological and nonpharmacological treatment has dramatically improved the prognosis of DCM [28], but despite this therapeutic success, emerging evidence suggests that some patients remain vulnerable to sudden cardiac death and refractory heart failure requiring heart transplantation or mechanical circulatory support [29]. Moreover, genetic screening suggests that up to 40% of DCM is genetically determined [29]. Familial screening should be systematically performed to obtain an early diagnosis in relatives as this facilitates prompt prophylactic therapy in early or preclinical disease [30] but gene diagnosis and therapy of DCM are not currently sufficiently advanced in clinical practice.

In this study, we screened potentially related pathogenic genes by comparing dilated cardiomyopathic heart biopsy samples with nonfailing heart biopsy samples during transplantation. Among 162 upregulated DEGs and ten downregulated DEGs identified, we selected nine key candidate genes, DLD, UQCRC2, DLAT, SUCLA2, ATP5A1, PRDX3, FH, SDHD, and NDUFV1, for characterization and found connections with these lay in the oxidation-reduction process, cellular respiration, the citrate cycle, and carbon metabolism of metabolic pathways by using GO and KEGG annotation and PPI network analysis.

Dihydrolipoamide dehydrogenase (encoded by DLD) is essential for energy metabolism across eukaryotes. The enzyme modulates production of reactive oxygen [31], and also affects branched-chain amino acid metabolism, resulting in maple syrup urine disease [32]. Mitochondrial complex III deficiency caused by UQCRC2 mutation can lead to neonatal-onset recurrent hepatocellular insufficiency, lactic acidosis, hypoglycemia, ketosis, and hyperammonemia [33, 34], and alterations in gene-related mitochondrial proteins and their quantities have been confirmed to affect prognosis and progression in many cancers [35, 36]. The DLAT gene encodes component E2 of the multienzyme pyruvate dehydrogenase complex, and E2 is the main mitochondrial autoantigen that plays a vital role of the innate immune recognition system in the pathogenesis of primary biliary cirrhosis [37]. SUCLA2 encodes a subunit of the enzyme succinate-CoA ligase and is at the crossroads of several biochemical pathways encompassing the citrate cycle, heme synthesis, propionyl-CoA metabolism, and ketone body utilization, to name a few [38]. ATP5A1 encodes a subunit of mitochondrial ATP synthase and plays a critical role in catalyzing ATP synthesis. Not only is it highly expressed in glioblastoma tumor cells and endothelial cells of microvascular proliferation [39], it is also involved in progression of clear cell carcinoma and mitochondrial encephalopathy, which could cause death in neonates [40, 41]. PRDX3 encodes a mitochondrial protein that has antioxidant functions and has physiological and pathological significance, with the initiation and progression of various cancer types, such as ovarian cancer, endometrial cancer, and prostate cancer [42–45]. FH encodes a homotetrameric mitochondrial protein that is part of the citrate cycle [46]. Current data show that recessive mutations in FH cause severe encephalopathies and early death, whereas dominant FH mutations predispose to tumor formation [47], but fumarate hydratase (which is encoded by this gene) can be used as a multifaceted tumor suppressor in cancer [48]. SDHD encodes one of four subunits of the citrate cycle enzyme succinate dehydrogenase (SDH), which is responsible for the oxidation of succinate [49]. The most important SDH-related mutation is in the SDHD gene, which has a key function both in familial and nonfamilial hereditary paraganglioma/pheochromocytoma syndrome [50]. NDUFV1 encodes a 51 kDa subunit of NADH:ubiquinone oxidoreductase complex I. Complex I deficiency is one of the most common mitochondrial respiratory chain defects. Mutations in NDUFV1 are involved in autosomal recessive mitochondrial complex I deficiency, and account for a constellation of clinical features, including encephalopathies, myopathies, and Leigh syndrome [51, 52].

Among the predicted transcription factors that modulate gene expression in DCM, we chose the NFI/CTF family of mammalian transcription factors. The NFI/CTF family is a family of transcription-replication factors comprising polypeptides encoded by four paralogous genes located on different chromosomes in mammals (NFIA, NFIB, NFIC, and NFIX). The NFI family of proteins displays the unusual property of not only regulating the initiation of transcription but also mediating DNA replication. NFI recognition sequences are found in the promoter sequences of many cellular genes, where they may act as an activator or a repressor of transcription. Recently, it was proposed that NFI transcription-replication factors may be involved in long-range regulation of gene expression through the formation of chromatin barriers and by blocking the propagation of a heterochromatic structure [53, 54].

By the preceding identification of hub genes from a GO, KEGG, and transcription factor perspective, we know that proteins encoded by these genes are involved in mitochondrial activity. Mitochondria play a key role in providing ATP for the energy-consuming cardiac tissues. Mitochondrial cardiomyopathy is a myocardial condition characterized by abnormal heart structure and/or function secondary to genetic defects involving the mitochondrial respiratory chain. Mitochondrial DNA deletion mutations are common in a variety of heart lesions, such as DCM and hypertrophic cardiomyopathy [55, 56]. In conclusion, our study provides greater understanding of the pathogenesis and underlying molecular mechanisms in DCM. This contributes to the exploration of potential gene therapy targets. However, further molecular biological experiments are required to confirm the function of these identified genes in DCM.

Acknowledgment

We thank ShengXinZhuShou (official WeChat account SCIPhD).

Availability of data and materials

The datasets generated and/or analyzed during the current study are available in the GEO repository at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE3586.

Author contributions

Data curation: Y. Chen, X. Wu, Formal analysis: X. Wu, D. Hu, Writing, original draft: Y. Chen, D. Hu, Writing, review & editing: W. Wang.

Conflicts of interests

The authors declare that they have no conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

1. RosenbaumAN, AgreKE, PereiraNL. Genetics of dilated cardiomyopathy: practical implications for heart failure management. Nat Rev Cardiol 2020;17:286–97.

2. MerloM, CannatàA, GobboM, StolfoD, ElliotPM, SinagraG. Evolving concepts in dilated cardiomyopathy. Eur J Heart Fail 2018;20(2):228–39.

3. McNallyEM, MestroniL. Dilated cardiomyopathy: genetic determinants and mechanisms. Circ Res 2017;121(7):731–48.

4. JúniorAL, FerrariF, MaxR, RittLEF, SteinR. Importance of genetic testing in dilated cardiomyopathy: applications and challenges in clinical practice. Arq Bras Cardiol 2019;113(2):274–81.

5. DomínguezF, CuencaS, BilińskaZ, ToroR, VillardE, Barriales-VillaR, et al. Dilated cardiomyopathy due to BLC2-associated athanogene 3 (BAG3) mutations. J Am Coll Cardiol 2018;72(20):2471–81.

6. KirbyJ, HeathPR, ShawPJ, HamdyFC. Gene expression assays. Adv Clin Chem 2007;44:247–92.

7. VogelsteinB, PapadopoulosN, VelculescuVE, ZhouS, DiazLA, KinzlerKW. Cancer genome landscapes. Science 2013;339(6127):1546–58.

8. VerdonschotJAJ, HazebroekMR, WareJS, PrasadSK, HeymansSRB. Role of targeted therapy in dilated cardiomyopathy: the challenging road toward a personalized approach. J Am Heart Assoc 2019;8(11):e012514.

9. BondueA, ArbustiniE, BiancoA, CiccarelliM, DawsonD, De RosaM, et al. Complex roads from genotype to phenotype in dilated cardiomyopathy: scientific update from the Working Group of Myocardial Function of the European Society of Cardiology. Cardiovasc Res 2018;114(10):1287–303.

10. DessauR, PipperCB. [“R”--project for statistical computing]. Ugeskr Laeger 2008;170(5):328–30.

11. WettenhallJM, SmythGK. limmaGUI: a graphical user interface for linear modeling of microarray data. Bioinformatics 2004;20(18):3705–6.

12. HuangDW, ShermanBT, LempickiRA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009;4(1):44–57.

13. BarthAS, KunerR, BunessA, RuschhauptM, MerkS, ZwermannL, et al. Identification of a common gene expression signature in dilated cardiomyopathy across independent microarray studies. J Am Coll Cardiol 2006;48(8):1610–7.

14. BarrettT, TroupDB, WilhiteSE, LedouxP, RudnevD, EvangelistaC, et al. NCBI GEO: archive for high-throughput functional genomic data. Nucleic Acids Res 2009;37:D885–90.

15. EdgarR, DomrachevM, LashAE. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res 2002;30(1):207–10.

16. EloLL, HiissaJ, TuimalaJ, KallioA, KorpelainenE, AittokallioT. Optimized detection of differential expression in global profiling experiments: case studies in clinical transcriptomic and quantitative proteomic datasets. Brief Bioinform 2009;10(5):547–55.

17. ServantN, GravierE, GestraudP, LaurentC, PaccardC, BitonA, et al. EMA – a R package for easy microarray data analysis. BMC Rese Notes 2010;3:277.

18. SmythGK. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol 2004;3:Article 3.

19. BlakeJA, HarrisMA. The Gene Ontology (GO) project: structured vocabularies for molecular biology and their application to genome and expression analysis. Curr Protoc Bioinformatics 2008:Unit 7.2.

20. AshburnerM, BallCA, BlakeJA, BotsteinD, ButlerH, CherryJM, et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 2000;25(1):25–9.

21. KanehisaaM, GotoS. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 2000;28(1):27–30.

22. LebrecJJP, HuizingaTWJ, ToesREM, Houwing-DuistermaatJJ, van HouwelingenHC. Integration of gene ontology pathways with North American Rheumatoid Arthritis Consortium genome-wide association data via linear modeling. BMC Proc 2009;3(Suppl 7):S94.

23. SzklarczykD, MorrisJH, CookH, KuhnM, WyderS, SimonovicM, et al. The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res 2017;45:D362–8.

24. ChinCH, ChenSH, HoCW, KoMT, LinCY. A hub-attachment based method to detect functional modules from confidence-scored protein interactions and expression profiles. BMC Bioinformatics 2010;11(Suppl 1):S25.

25. GriffithM, GriffithOL, CoffmanAC, WeibleJV, McMichaelJF, SpiesNC, et al. DGIdb: mining the druggable genome. Nat Methods 2013;10(12):1209–10.

26. MinF, GaoF, LiuZ. Screening and further analyzing differentially expressed genes in acute idiopathic pulmonary fibrosis with DNA microarray. Eur Rev Med Pharmacol Sci 2013;17(20):2784–90.

27. ElliottP, AnderssonB, ArbustiniE, BilinskaZ, CecchiF, CharronP, et al. Classification of the cardiomyopathies: a position statement from the European Society of Cardiology Working Group on Myocardial and Pericardial Diseases. Eur Heart J 2008;29(2):270–6.

28. MerloM, PivettaA, PinamontiB, StolfoD, ZecchinM, BarbatiG, et al. Long-term prognostic impact of therapeutic strategies in patients with idiopathic dilated cardiomyopathy: changing mortality over the last 30 years. Eur J Heart Fail 2014;16(3):317–24.

29. ZecchinM, MerloM, PivettaA, BarbatiG, LutmanC, GregoriD, et al. How can optimization of medical treatment avoid unnecessary implantable cardioverter-defibrillator implantations in patients with idiopathic dilated cardiomyopathy presenting with “SCD-HeFT criteria?”. Am J Cardiol 2012;109(5):729–35.

30. MorettiM, MerloM, BarbatiG, Di LenardaA, BrunF, PinamontiB, et al. Prognostic impact of familial screening in dilated cardiomyopathy. Eur J Heart Fail 2010;12(9):922–7.

31. DayanA, FlemingerG, Ashur-FabianO. Targeting the Achilles’ heel of cancer cells via integrin-mediated delivery of ROS-generating dihydrolipoamide dehydrogenase. Oncogene 2019;38(25):5050–61.

32. AbiriM, SaeiH, EghbaliM, KaramzadehR, ShirzadehT, SharifiZ, et al. Maple syrup urine disease mutation spectrum in a cohort of 40 consanguineous patients and insilico analysis of novel mutations. Metab Brain Dis 2019;34(4):1145–56.

33. MiyakeN, YanoS, SakaiC, HatakeyamaH, MatsushimaY, ShiinaM, et al. Mitochondrial complex III deficiency caused by a homozygous UQCRC2 mutation presenting with neonatal-onset recurrent metabolic decompensation. Hum Mutat 2013;34(3):446–52.

34. GaignardP, EyerD, LebigotE, OliveiraC, TherondP, BoutronA, et al. UQCRC2 mutation in a patient with mitochondrial complex III deficiency causing recurrent liver failure, lactic acidosis and hypoglycemia. J Hum Genet 2017;62(7):729–31.

35. WeinbergSE, ChandelNS. Targeting mitochondria metabolism for cancer therapy. Nat Chem Biol 2015;11(1):9–15.

36. PutignaniL, RaffaS, PescosolidoR, RizzaT, Del ChiericoF, LeoneL, et al. Preliminary evidences on mitochondrial injury and impaired oxidative metabolism in breast cancer. Mitochondrion 2012;12(3):363–9.

37. BergPA. The role of the innate immune recognition system in the pathogenesis of primary biliary cirrhosis: a conceptual view. Liver Int 2011;31(7):920–31.

38. TretterL, PatocsA, ChinopoulosC. Succinate, an intermediate in metabolism, signal transduction, ROS, hypoxia, and tumorigenesis. Biochim Biophys Acta 2016;1857(8):1086–101.

39. XuG, LiJY. ATP5A1 and ATP5B are highly expressed in glioblastoma tumor cells and endothelial cells of microvascular proliferation. J Neurooncol 2016;126(3):405–13.

40. JonckheereAI, RenkemaGH, BrasM, van den HeuvelLP, HoischenA, GilissenC, et al. A complex V ATP5A1 defect causes fatal neonatal mitochondrial encephalopathy. Brain 2013;136:1544–54.

41. YuanL, ChenL, QianK, WangG, LuM, QianG, et al. A novel correlation between ATP5A1 gene expression and progression of human clear cell renal cell carcinoma identified by coexpression analysis. Oncol Rep 2018;39(2):525–36.

42. IsmailT, KimY, LeeH, LeeDS, LeeHS. Interplay between mitochondrial peroxiredoxins and ROS in cancer development and progression. Int J Mol Sci 2019;20(18):4407.

43. LiS, HuX, YeM, ZhuX. The prognostic values of the peroxiredoxins family in ovarian cancer. Biosci Rep 2018;38(5):BSR20180667.

44. ByunJM, KimSS, KimKT, KangMS, JeongDH, LeeDS, et al. Overexpression of peroxiredoxin-3 and -5 is a potential biomarker for prognosis in endometrial cancer. Oncol Lett 2018;15(4):5111–8.

45. WhitakerHC, PatelD, HowatWJ, WarrenAY, KayJD, SanganT, et al. Peroxiredoxin-3 is overexpressed in prostate cancer and promotes cancer cell survival by protecting cells from oxidative stress. Br J Cancer 2013;109(4):983–93.

46. FrezzaC. Mitochondrial metabolites: undercover signalling molecules. Interface Focus 2017;7(2):20160100.

47. TomlinsonIPM, AlamNA, RowanAJ, BarclayE, JaegerEEM, KelsellD, et al. Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer. Nat Genet 2002;30(4):406–10.

48. SchmidtC, SciacovelliM, FrezzaC. Fumarate hydratase in cancer: a multifaceted tumour suppressor. Semin Cell Dev Biol 2020;98:15–25.

49. BaysalBE, FerrellRE, Willett-BrozickJE, LawrenceEC, MyssiorekD, BoschA, et al. Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma. Science 2000;287(5454):848–51.

50. NazarE, KhatamiF, SaffarH, TavangarSM. The emerging role of succinate dehyrogenase genes (SDHx) in tumorigenesis. Int J Hematol Oncol and Stem Cell Res 2019;13(2):72–82.

51. SrivastavaA, SrivastavaKR, HebbarM, GaladaC, KadavigrereR, SuF, et al. Genetic diversity of NDUFV1-dependent mitochondrial complex I deficiency. Eur J Hum Genet 2018;26(11):1582–7.

52. WadhwaY, RohillaS, KaushikJS. Cystic Leucoencephalopathy in NDUFV1 mutation. Indian J Pediatr 2018;85(12):1128–31.

53. GronostajskiRM. Roles of the NFI/CTF gene family in transcription and development. Gene 2000;249:31–45.

54. PjanicM, PjanicP, SchmidC, AmbrosiniG, GaussinA, PlasariG, et al. Nuclear factor I revealed as family of promoter binding transcription activators. BMC Genomics 2011;12:181.

55. ZhugeR, ZhouR, NiX. Progress in molecular genetic study of mitochondrial cardiomyopathy. Zhongguo Yi Xue Ke Xue Yuan Xue Bao 2017;39(3):438–44.

56. DadsonK, HauckL, BilliaF. Molecular mechanisms in cardiomyopathy. Clin Sci (Lond) 2017;131(13):1375–92.