ETM Experimental and Therapeutic Medicine 1792-0981 1792-1015 D.A. Spandidos 10.3892/etm.2017.4850 ETM-0-0-4850 Articles Bioinformatics method identifies potential biomarkers of dilated cardiomyopathy in a human induced pluripotent stem cell-derived cardiomyocyte model ZhuangYu 1 GongYu-Jia 2 ZhongBei-Fen 3 ZhouYi 1 GongLi 4 Department of Cardiovascular Surgery, Shanghai First People's Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200080, P.R. China Stomatology Faculty, School of Medicine, Nantong University, Nantong, Jiangsu 226000, P.R. China Department of Urology, Shanghai First People's Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200080, P.R. China Department of Cardiothoracic Surgery, The Affiliated Huai'an Hospital of Xuzhou Medical University, Huai'an, Jiangsu 223002, P.R. China Correspondence to: Dr Li Gong, Department of Cardiothoracic Surgery, The Affiliated Huai'an Hospital of Xuzhou Medical University, 62 South Huaihai Road, Huai'an, Jiangsu 223002, P.R. China, E-mail: 1740385802@qq.com 10 2017 28 07 2017 14 4 2771 2778 19022016 10022017 Copyright © 2017, Spandidos Publications 2017

Dilated cardiomyopathy (DCM) is the most common type of cardiomyopathy that account for the majority of heart failure cases. The present study aimed to reveal the underlying molecular mechanisms of DCM and provide potential biomarkers for detection of this condition. The public dataset of GSE35108 was downloaded, and 4 normal induced pluripotent stem cell (iPSC)-derived cardiomyocytes (N samples) and 4 DCM iPSC-derived cardiomyocytes (DCM samples) were utilized. Raw data were preprocessed, followed by identification of differentially expressed genes (DEGs) between N and DCM samples. Crucial functions and pathway enrichment analysis of DEGs were investigated, and protein-protein interaction (PPI) network analysis was conducted. Furthermore, a module network was extracted from the PPI network, followed by enrichment analysis. A set of 363 DEGs were identified, including 253 upregulated and 110 downregulated genes. Several biological processes (BPs), such as blood vessel development and vasculature development (FLT1 and MMP2), cell adhesion (CDH1, ITGB6, COL6A3, COL6A1 and LAMC2) and extracellular matrix (ECM)-receptor interaction pathway (CDH1, ITGB6, COL6A3, COL6A1 and LAMC2), were significantly enriched by these DEGs. Among them, MMP2, CDH1 and FLT1 were hub nodes in the PPI network, while COL6A3, COL6A1, LAMC2 and ITGB6 were highlighted in module 3 network. In addition, PENK and APLNR were two crucial nodes in module 2, which were linked to each other. In conclusion, several potential biomarkers for DCM were identified, such as MMP2, FLT1, CDH1, ITGB6, COL6A3, COL6A1, LAMC2, PENK and APLNR. These genes may serve significant roles in DCM via involvement of various BPs, such as blood vessel and vasculature development and cell adhesion, and the ECM-receptor interaction pathway.

 dilated cardiomyopathy cell adhesion extracellular matrix-receptor interaction induced pluripotent stem cell enrichment analysis protein-protein interaction bioinformatics Introduction

Cardiomyopathy is the leading cause of heart failure and mainly consists of four types: Dilated cardiomyopathy (DCM), hypertrophic cardiomyopathy (HCM), restrictive cardiomyopathy and arrhythmogenic right ventricular cardiomyopathy (ARVC) (1). Characterized by enlarged ventricular dimensions and systolic dysfunction (2), DCM is the most frequent type of cardiomyopathy, accounting for 10% of the newly diagnosed heart failure cases (3). Management of DCM is an economic burden worldwide (4). Therefore, it is urgent to identify the genetic etiologies of this disease.

Genetic mutations are known to influence the majority of cardiomyopathies, and DCM is triggered by mutations in giant protein titin (5). Reportedly, DCM is associated with mutations in several genes, such as DES, LMNA, SCN5A and TNNT2 (6–8). Among them, TNNT2 is the gene that encodes for cardiac troponin T, one of the subunits of the troponin complex that regulates muscle contraction and sarcomere assembly (9). Mouse models carrying TNNT2 mutations that denote the human DCM phenotype have been established and used to determine the underlying mechanisms of DCM (10). However, due to the differences between mouse and human models, the interpretation of such findings may be limited.

In mice, four transcription factors, namely Oct3/4, Sox2, Klf4 and c-Myc, are essential to generate the induced pluripotent stem cells (iPSCs). By contrast, in humans, the four indispensable factors for iPSCs generation from fibroblasts are OCT4, SOX2, NANOG and LIN28 (11). This enables the establishment of disease-specific cellular models in humans, which contributes to the understanding of mechanisms and suitable drug screening for certain diseases. In humans, iPSCs derived from DCM patients have been used to obtain cardiomyocytes (12). A previous study created a human DCM iPSC-derived cardiomyocyte model, which was generated from DCM patients carrying the TNNT2 mutation and from healthy controls without this mutation, in order to investigate relevant DCM mechanisms and potential drug therapy (13). As a result, the study discovered that, in comparison with the healthy control group, the iPSC-derived cardiomyocytes from DCM patients showed altered regulation of Ca2+ and abnormal distribution of sarcomeric α-actin (13). Despite these valuable findings, gene alterations in DCM cardiomyocytes, as well as regulatory correlations of these genes were not considered.

Therefore, the present study re-analyzed the dataset GSE35108, mainly focusing on the gene alterations, as well as their biological functions and potential correlations in human DCM. Differentially expressed genes (DEGs) were identified between DCM iPSC-derived cardiomyocyte samples and healthy controls, followed by enrichment analysis. In addition, a protein-protein interaction (PPI) network was established to explore potential correlations between these DEGs, and three module networks were further extracted to explore detailed crucial genes and their functions. Through these comprehensive bioinformatics methods, the current study aimed to reveal underlying molecular mechanisms of DCM and to provide potential biomarkers for the detection and determination of the prognosis of the disease.

 Materials and methods <sec> <title>Data resource

Gene expression profile with the accession number of GSE35108 (13) was obtained from the public Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo/). The dataset contained a total of 17 samples; among them, 4 normal iPSC-derived cardiomyocytes (N samples) and 4 DCM iPSC-derived cardiomyocytes (DCM samples) were utilized in the present study. The platform for the gene expression profiles was Affymetrix Human Gene 1.0 ST Array (Affymetrix Inc., Santa Clara, CA, USA).

 Data preprocessing and differential analysis

The Bioconductor oligo package (http://www.bioconductor.org/packages/release/bioc/html/oligo.html) was used to perform the pretreatment analysis (14). Background correction and quantile normalization were performed for the raw data using the robust multichip average (RMA) algorithm (15). Next, expression values at probe level (probe ID) were converted to gene level (gene symbols). After removing the redundant probes, the gene expression matrix was finally obtained.

The Linear Models for Microarray Analysis (Limma) package (16) in R software (http://www.bioconductor.org/packages/release/bioc/html/limma.html) was applied to identify DEGs between the N and DCM samples, based on Student's t-test. Adjusted P-values were calculated using the Benjamini-Hochberg method (17). The thresholds for significant DEG selection were an adjusted P value of <0.05 and |log2 fold change (FC)|of >2.

 Functional enrichment analysis

To explore potential functions and pathways that may be altered by the DEGs, the online tool Database for Annotation, Visualization and Integration Discovery (DAVID; https://david.ncifcrf.gov/) (18) was selected to perform functional and pathway enrichment analyses of the identified DEGs. The Gene Ontology (GO; http://www.geneontology.org/) database was used to identify the biological processes (BPs) that the DEGs may be involved. In addition, the Kyoto Encyclopedia of Genes and Genomes (KEGG; http://www.genome.jp/kegg/pathway.html) database was used to perform pathway enrichment analysis for the identified DEGs, according to the modified Fisher exact test. The selection criteria for significant GO terms and pathways were P<0.05 and an enriched gene number (also known as count) of >2.

 PPI network construction

To further investigate potential associations of these DEGs at the protein level, the DEGs were mapped into the Search Tool for the Retrieval of Interacting Genes (STRING; http://string-db.org/) database. Subsequently, a PPI network was constructed, with the criterion of combined score of >0.4 between two interplayed proteins.

The Cytoscape software (http://cytoscape.org/) was used to perform topological structure analysis of the PPI network. A node in the PPI network was deemed as a protein product of the DEG. Based on the Connectivity Degree analysis (19), crucial nodes in the PPI network with high degree (where degree refers to the number of edges for a DEG in the PPI network) were highlighted and named as hub nodes.

 Module analysis of the PPI network

Module networks were extracted from the PPI network by the ClusterONE software (http://www.paccanarolab.org/clusterone/) in the Cytoscape, with the default parameter. The sub-modules network with P<0.05 was considered as significant functional sub-modules. Thereafter, functional enrichment analysis was performed for genes in the module networks, according to aforementioned method.

 Results <sec> <title>DEGs between N and DCM samples

According to the criteria of adjusted P-value of <0.05 and |log2 FC| of >2, a set of 363 DEGs were identified in the DCM samples, in comparison with the N samples. Among them, 253 genes were upregulated, while 110 were downregulated.

 Altered function and pathway by these DEGs

As indicated in Table I, upregulated DEGs were significantly correlated with various BPs, including ectoderm or epidermis development (e.g. KLF4), blood vessel and vasculature development (e.g. FLT1, MMP2 and COL1A2), as well as cell and biological adhesion (e.g. CDH1, ITGB6, COL6A3, COL6A1 and LAMC2). By contrast, the downregulated DEGs were markedly enriched in BPs, such as muscle contraction and muscle system processes (e.g. FXYD1, SLC8A1, CRYAB), and muscle organ development (e.g. ITGA11) and heart development (e.g. XIRP1, MYL2, MYBPC3).

Based on the KEGG pathway enrichment analysis, five predominant pathways for upregulated DEGs were identified, as follows: Extracellular matrix (ECM)-receptor interaction (e.g. COL6A3, ITGB6, COL1A2, COL6A1), focal adhesion (e.g. FLT1, COL6A3, ITGB6, COL1A2, COL6A1), glutathione metabolism (e.g. GPX2, GSTA2, ANPEP), complement and coagulation cascades (e.g. PLAT, FGA, FGB) and O-Glycan biosynthesis (e.g. GCNT4, GALNT12, GCNT1). By contrast, only three significant pathways were enriched for the downregulated DEGs, including HCM, DCM and ARVC (e.g. ITGA11 in all these pathways; Table II).

 Predicted PPI network and module network analysis

According to the predefined criterion, a PPI network was constructed, containing 520 interplayed PPIs. As shown in Fig. 1, the top ten hub nodes in the PPI network were MMP2 (degree, 26), CDH1 (degree, 25), CAT (degree, 21), NANOG (degree, 18), PROM1 (degree, 18), KLF4 (degree, 15), FGF13 (degree, 13), PDE5A (degree, 12), FLT1 (degree, 12) and PDE2A (degree, 12).

Three significant module networks were extracted, including module 1 (P=5.80×10−4), module 2 (P=2.92×10−3) and module 3 (P=1.34×10−2). In module 1, nodes such as CAT, PDE5A, PDE2A and LRRN1 were highlighted; in module 2, PENK, APLNR, BDKRB1, OXGR1 and P2RY14 were predominant and linked to each other; and in module 3, the highlighted nodes were ITGB6, COL6A3, COL1A2, COL6A1 and ITGA11 (Fig. 2). No significant BP terms and pathways were enriched for the nodes in module 1. Enrichment analysis indicated that genes in module 2 were significantly associated with the following BPs (Table III): Cell surface receptor linked signal transduction (PLAT, APLNR, PENK, P2RY14, BDKRB1 and OXGR1), G-protein coupled receptor protein signaling pathway (APLNR, PENK, P2RY14, BDKRB1 and OXGR1), sensory perception of pain (PENK and BDKRB1), neuroactive ligand-receptor interaction (APLNR, P2RY14 and BDKRB1), and complement and coagulation cascades (PLAT and BDKRB1). Furthermore, as shown in Table IV, genes in module 3 were significantly associated with BPs such as cell adhesion (COL6A3, ITGB6, ITGA11, COL6A1 and LAMC2), biological adhesion (COL6A3, ITGB6, ITGA11, COL6A1 and LAMC2), integrin-mediated signaling pathway (ITGB6 and ITGA11), cell-matrix adhesion (ITGB6 and ITGA11), cell-substrate adhesion (ITGB6 and ITGA11), ECM-receptor interaction (e.g. COL6A3, ITGB6, COL1A2, ITGA11, COL6A1, LAMC2 and SV2C) and focal adhesion (e.g. COL6A3, ITGB6, COL1A2, ITGA11, COL6A1 and LAMC2). However, no significant BP terms were enriched based on the set criterion for enrichment analysis.

 Discussion

By comparing gene expression profiles in iPSC-derived cardiomyocytes from DCM patients and healthy individuals, the present study identified a set of critical DEGs, such as FLT1 and MMP2, which were upregulated and significantly enriched in the blood vessel development and vasculature development BPs. In addition, CDH1, ITGB6, COL6A3, COL6A1 and LAMC2 were upregulated and markedly correlated with the cell adhesion BP and the ECM-receptor interaction pathway. Notably, MMP2, CDH1 and FLT1 were three hub nodes in the PPI network, while COL6A3, COL6A1, LAMC2 and ITGB6 were the crucial nodes in module 3 network. Furthermore, other crucial nodes were identified in module 2, such as PENK and APLNR, which were linked to each other.

A structural feature of DCM is chamber dilation. Dilation requires myocyte and ECM alteration, which are demonstrated as causative factors for cardiac enlargement (20). Matrix metalloproteinases (MMPs) are family members of Ca-dependent and Zn-containing endopeptidases that are essential for ECM degradation and remodeling (21). MMP2 is a member of the MMP family that has been identified as a crucial protease in myocardial ischemia (22). Furthermore, MMP2 serves significant roles in infarction, which is associated with fibrous tissue enlargement (23). Reportedly, normal blood vessel development is regulated by signals between ECM and endothelial cells, while MMP2 is mainly secreted by endothelial cells (24). Additionally, MMP2 can enter the vasculature (25). The FLT1 encoded protein is a family member of the vascular endothelial growth factor receptor. During blood vessel morphogenesis, FLT1 is identified as a crucial molecule for the control of endothelial cell activity (26). Alteration of FLT1 has been detected in cardiomyopathies by the NHLBI-Go Exome Sequencing Project (27); however, the association between this gene and HCM or DCM needs to be re-evaluated (28). Blood vessels serve important roles in the heart during cardiac disease and remodeling, and several vasculature-associate genes are involved in cardiac hypertrophy (29). DCM has the characteristics of cardiac enlargement and hypertrophy. Therefore, it is not surprising that several upregulated genes in DCM, such as MMP2 and FLT1, were enriched in the blood vessel and vasculature development BPs in the present study. These collectively suggest that the two genes may have significant roles in human DCM, via regulation of the blood vessel and vasculature development.

Two hallmarks of DCM are abnormalities of cellular function and immune system. Overexpression of cell adhesion molecules is a common event in DCM (30). CDH1 encodes an E-cadherin protein, which is a cell-cell adhesion molecule. Epithelial-to-mesenchymal transition (EMT) is proposed as the key mechanism for the production of cardiovascular progenitor cells, which finally differentiate into endothelial cells and cardiomyocytes, and CDH1 is one of the prominent regulators of EMT (31). Increased expression of CDH1 in DCM samples with the enrichment in cell adhesion BP in the current study suggests that CDH1 may be a crucial gene in DCM that controls the regulation of cell adhesion. ITGB6, LAMC2, as well as collagens COL6A1 and COL6A3, are other cell adhesion-associated molecules (32). All these genes were enriched in the cell adhesion BP in the present study, suggesting that altered cell adhesion may serve a significant role in DCM.

The ITGB6 encoded protein is a family member of adhesion receptors that have significant roles in signaling from the ECM to the cell. Intergrins are known as the major receptors for substantial ECM-mediated cellular activities, including cell adhesion, cell proliferation and differentiation (33). Enrichment analysis in a mouse model has indicated that integrins, such as ITGB6, can participate in the DCM pathway (34). The findings of the present study indicated that the DEG ITGB6 in DCM was significantly enriched in the ECM-receptor interaction pathway. Notably, ECM-receptor interaction is also involved in DCM pathway (35). Thus, it can be inferred that ECM-receptor interaction may be an important pathway for DCM, which may be altered by the integrin gene ITGB6.

PENK encodes for the proenkephalin protein, and has been reported as one of the DEGs identified in skeletal muscle, while the muscle development-associated DEGs are implicated in the DCM pathway (36). APLNR encodes a family member of the G protein-coupled receptors. Several studies have investigated polymorphisms of APLNR and the susceptibility of DCM (37–39). In addition, the 212A allele of the APLNR G212A polymorphism is considered to be significantly associated with reduced risk of idiopathic DCM (39), suggesting that alteration of this gene may influence the DCM. In the present study, PENK and APLNR were identified as two crucial genes in module 2, suggesting that they may exert more important functions in DCM than other genes. In addition, the linkage of these two genes in the module 2 network implies potential co-regulation of DCM.

Although a relatively comprehensive bioinformatics analysis was conducted in the present study, there remained several limitations. As the dataset was obtained from the public database, the sample size was small. Furthermore, there lacked validations of DEG expressions, as well as correlations between them. However, the obtained predictive results are also of great value to provide novel insights into the underlying DCM mechanisms and potential biomarkers for DCM prediction.

In conclusion, several potential biomarkers for DCM were identified based on the bioinformatics results of the current study, including MMP2, FLT1, CDH1, ITGB6, COL6A3, COL6A1, LAMC2, PENK and APLNR. These genes may serve significant roles in DCM via the involvement of various BPs, including blood vessel and vasculature development, and cell adhesion, as well as through the pathway of ECM-receptor interaction. However, further experimental validations are required to confirm these findings.

 References McNallyEMGolbusJRPuckelwartzMJGenetic mutations and mechanisms in dilated cardiomyopathyJ Clin Invest1231926201310.1172/JCI6286223281406 MaronBJTowbinJAThieneGAntzelevitchCCorradoDArnettDMossAJSeidmanCEYoungJBAmerican Heart Association; Council on Clinical Cardiology, Heart Failure and Transplantation Committee, et al: Contemporary definitions and classification of the cardiomyopathies: An American heart association scientific statement from the council on clinical cardiology, heart failure and transplantation committee; quality of care and outcomes research and functional genomics and translational biology interdisciplinary working groups; and council on epidemiology and preventionCirculation11318071816200610.1161/CIRCULATIONAHA.106.17428716567565 LiXLuoRHuaWLiLKwongJSWChanCPCMYCardiac resynchronization therapy for dilated cardiomyopathyCochrane Libr2013 VergaroGDel FrancoAGiannoniAPronteraCRipoliABarisonAMasciPGAquaroGDCohenSolal APadelettiLGalectin-3 and myocardial fibrosis in nonischemic dilated cardiomyopathyInt J Cardiol18496100201510.1016/j.ijcard.2015.02.00825697876 McNallyEMBarefieldDYPuckelwartzMJThe genetic landscape of cardiomyopathy and its role in heart failureCell Metab21174182201510.1016/j.cmet.2015.01.01325651172 HershbergerREMoralesALMNA-related dilated cardiomyopathyGeneReviews®PagonRAAdamMPArdingerHHWallaceSEAmemiyaABeanLJHBirdTDLedbetterNMeffordHCSmithRJHStephensKUniversity of WashingtonSeattle2008[updated July 7, 2016] RobynsTNuyensDVan CasterenLCorveleynADe RavelTHeidbuchelHWillemsRReduced penetrance and variable expression of SCN5A mutations and the importance of co-inherited genetic variants: Case report and review of the literatureIndian Pacing Electrophysiol J14133149201410.1016/S0972-6292(16)30754-9 HershbergerREPintoJRParksSBKushnerJDLiDLudwigsenSCowanJMoralesAParvatiyarMSPotterJDClinical and functional characterization of TNNT2 mutations identified in patients with dilated cardiomyopathyCirc Cardiovasc Genet2306313200910.1161/CIRCGENETICS.108.846733 SehnertAJHuqAWeinsteinBMWalkerCFishmanMStainierDYCardiac troponin T is essential in sarcomere assembly and cardiac contractilityNat Genet31106110200210.1038/ng875 AhmadFBanerjeeSKLageMLHuangXNSmithSHSabaSRagerJConnerDAJanczewskiAMTobitaKThe role of cardiac troponin T quantity and function in cardiac development and dilated cardiomyopathyPLoS One3e2642200810.1371/journal.pone.0002642 YuJVodyanikMASmuga-OttoKAntosiewicz-BourgetJFraneJLTianSNieJJonsdottirGARuottiVStewartRInduced pluripotent stem cell lines derived from human somatic cellsScience31819171920200710.1126/science.1151526 ZhangJWilsonGFSoerensAGKoonceCHYuJPalecekSPThomsonJAKampTJFunctional cardiomyocytes derived from human induced pluripotent stem cellsCirc Res104e30e41200910.1161/CIRCRESAHA.108.192237 SunNYazawaMLiuJHanLSanchez-FreireVAbilezOJNavarreteEGHuSWangLLeeAPatient-specific induced pluripotent stem cells as a model for familial dilated cardiomyopathySci Transl Med4130ra47201210.1126/scitranslmed.3003552 SchaferEIrizarryRNegiSMcIntyreESmallDFigueroaMEMelnickABrownPPromoter hypermethylation in MLL-r infant acute lymphoblastic leukemia: Biology and therapeutic targetingBlood11547984809201010.1182/blood-2009-09-243634 Charafe-JauffretEGinestierCIovinoFWicinskiJCerveraNFinettiPHurMHDiebelMEMonvilleFDutcherJBreast cancer cell lines contain functional cancer stem cells with metastatic capacity and a distinct molecular signatureCancer Res6913021313200910.1158/0008-5472.CAN-08-2741 SmythGKLimma: Linear models for microarray dataBioinformatics and computational biology solutions using R and BioconductorSpringer397420200510.1007/0-387-29362-0_23 FerreiraJAZwindermanAHOn the Benjamini-Hochberg methodAnn Stat3418271849200610.1214/009053606000000425 DennisGJrShermanBTHosackDAYangJGaoWLaneHCLempickiRADAVID: Database for annotation, visualization, and integrated discoveryGenome Biol4P3200310.1186/gb-2003-4-5-p3 GreiciusMDKrasnowBReissALMenonVFunctional connectivity in the resting brain: A network analysis of the default mode hypothesisProc Natl Acad Sci USA100253258200310.1073/pnas.0135058100 ChegeniSKhakiZShiraniDVajhiATaheriMTamrchiYRostamiAInvestigation of MMP-2 and MMP-9 activities in canine sera with dilated cardiomyopathyIran J Vet Res161821872015 StamenkovicIExtracellular matrix remodelling: The role of matrix metalloproteinasesJ Pathol200448464200310.1002/path.1400 ChengYZhangCMicroRNA-21 in cardiovascular diseaseJ Cardiovasc Transl Res3251255201010.1007/s12265-010-9169-7 HayashidaniSTsutsuiHIkeuchiMShiomiTMatsusakaHKubotaTImanaka-YoshidaKItohTTakeshitaATargeted deletion of MMP-2 attenuates early LV rupture and late remodeling after experimental myocardial infarctionAm J Physiol Heart Circ Physiol285H1229H1235200310.1152/ajpheart.00207.2003 AliSSaikJEGouldDJDickinsonMEWestJLImmobilization of cell-adhesive laminin peptides in degradable PEGDA hydrogels influences endothelial cell tubulogenesisBiores Open Access2241249201310.1089/biores.2013.0021 LebelRLepageMA comprehensive review on controls in molecular imaging: Lessons from MMP-2 imagingContrast Media Mol Imaging9187210201410.1002/cmmi.1555 HerbertSPStainierDYMolecular control of endothelial cell behaviour during blood vessel morphogenesisNat Rev Mol Cell Biol12551564201110.1038/nrm3176 PunethaJHoffmanEPShort read (next-generation) sequencing: A tutorial with cardiomyopathy diagnostics as an exemplarCirc Cardiovasc Genet6427434201310.1161/CIRCGENETICS.113.000085 AndreasenCNielsenJBRefsgaardLHolstAGChristensenAHAndreasenLSajadiehAHaunsøSSvendsenJHOlesenMSNew population-based exome data are questioning the pathogenicity of previously cardiomyopathy-associated genetic variantsEur J Hum Genet21918928201310.1038/ejhg.2012.283 RajanSPenaJRJeggaAGAronowBJWolskaBMWieczorekDFMicroarray analysis of active cardiac remodeling genes in a familial hypertrophic cardiomyopathy mouse model rescued by a phospholamban knockoutPhysiol Genomics45764773201310.1152/physiolgenomics.00023.2013 FelixSBStaudtANon-specific immunoadsorption in patients with dilated cardiomyopathy: Mechanisms and clinical effectsInt J Cardiol1123033200610.1016/j.ijcard.2006.05.014 TangYLWangYJChenLJPanYHZhangLWeintraubNLCardiac-derived stem cell-based therapy for heart failure: Progress and clinical applicationsExp Biol Med (Maywood)238294300201310.1177/1535370213477982 HuangYDe ReynièsADe LevalLGhaziBMartin-GarciaNTravertMBosqJBrièreJPetitBThomasEGene expression profiling identifies emerging oncogenic pathways operating in extranodal NK/T-cell lymphoma, nasal typeBlood11512261237201010.1182/blood-2009-05-221275 KimTHShinSWParkJSParkCSGenome wide identification and expression profile in epithelial cells exposed to TiO2 particlesEnviron Toxicol30293300201510.1002/tox.21906 DiamantiDMoriEIncarnatoDMalusaFFondelliCMagnoniLPollioGWhole gene expression profile in blood reveals multiple pathways deregulation in R6/2 mouse modelBiomark Res128201310.1186/2050-7771-1-28 LeeHJJangMKimHKwakWParkWHwangJYLeeCKJangGWParkMNKimHCComparative transcriptome analysis of adipose tissues reveals that ECM-receptor interaction is involved in the depot-specific adipogenesis in cattlePLoS One8e66267201310.1371/journal.pone.0066267 ZhangRPLiuHHLiuJYHuJWYanXPWangDMLiLWangJWTranscriptional profiling identifies location-specific and breed-specific differentially expressed genes in embryonic myogenesis in anas platyrhynchosPLoS One10e0143378201510.1371/journal.pone.0143378 StarkKCandidate gene approach on DCM causing or susceptibility genes2010 Esteban-MartínezRLPérez-RazoJCVargas-AlarcónGMartínez-RodríguezNCano-MartínezLJLópez-HernándezLBRojano-MejíaDCantoPCoral-VazquezRMPolymorphisms of APLN-APLNR system are associated with essential hypertension in Mexican-Mestizo individualsExp Mol Pathol101105109201610.1016/j.yexmp.2016.07.007 KotanidouEPKalinderiKKyrgiosIEfraimidouSFidaniLPapadopoulou-AlatakiEEboriadou-PetikopoulouMGalli-TsinopoulouAApelin and G212A apelin receptor gene polymorphism in obese and diabese youthPediatr Obes10213219201510.1111/ijpo.251

Protein-protein interaction network of the differentially expressed genes. Red nodes indicate upregulated genes, and green nodes represent downregulated genes.

Three module networks of the differentially expressed genes. Red nodes indicate upregulated genes, and green nodes represent downregulated genes.

Significantly enriched biological processes of the differentially expressed genes.

A. Upregulated genes
Description Count Genes P-value
GO:0007398-ectoderm development 15 KRT6A, KRT6B, PTGS2, SPRR2F, GRHL3, SPRR2E, SCEL, PTHLH, FLG, COL1A2, SPRR3, LAMC2, KLK14, KLF4, EMP1 7.18×10−7
GO:0008544-epidermis development 13 PTGS2, SPRR2F, GRHL3, SPRR2E, SCEL, PTHLH, FLG, COL1A2, SPRR3, LAMC2, KLF4, EMP1, KLK14 9.95×10−6
GO:0001568-blood vessel development 13 PLAT, APOB, FLT1, EPAS1, HAND1, TDGF1, COL1A2, SEMA3C, TGFBR3, ANPEP, PLXND1, MMP2, CEACAM1 1.62×10−4
GO:0001944-vasculature development 13 PLAT, APOB, FLT1, EPAS1, HAND1, TDGF1, COL1A2, SEMA3C, TGFBR3, ANPEP, PLXND1, MMP2, CEACAM1 2.03×10−4
GO:0010035-response to inorganic substance 11 APOB, PTGS2, FGA, OLR1, FGB, DUOX2, MGP, CA2, CAT, ITPR3, AQP3 5.90×10−4
GO:0007155-cell adhesion 22 CLDN7, CLDN18, CLCA2, MPZL2, TNXB, OLR1, MGP, CDH1, VTN, PCDH18, CD9, NLGN4Y, CD34, DSG3, TEK, ITGB6, COL6A3, DSC3, COL6A1, LAMC2, SPON2, CEACAM1 7.33×10−4
GO:0022610-biological adhesion 22 CLDN7, CLDN18, CLCA2, MPZL2, TNXB, OLR1, MGP, CDH1, VTN, PCDH18, CD9, NLGN4Y, CD34, DSG3, TEK, ITGB6, COL6A3, DSC3, COL6A1, LAMC2, SPON2, CEACAM1 7.46×10−4
GO:0060541-respiratory system development 8 WNT2, PTHLH, RBP4, LIPA, EPAS1, CRISPLD2, ANO1, MGP 7.66×10−4
GO:0035295-tube development 11 WNT2, PTHLH, RBP4, FLT1, LIPA, GPC3, HAND1, EPAS1, CRISPLD2, MGP, PLXND1 1.02×10−3
GO:0016337-cell-cell adhesion 12 CLDN7, CLDN18, MPZL2, OLR1, CD34, DSG3, TEK, DSC3, MGP, CDH1, CEACAM1, PCDH18 1.67×10−3
B. Downregulated genes
Description Count Genes P-value
GO:0006936-muscle contraction 11 FXYD1, SLC8A1, CRYAB, CKMT2, MYBPC3, RYR2, PGAM2, SMPX, CHRNB1, CASQ2, MB 1.46×10−8
GO:0003012-muscle system process 11 FXYD1, SLC8A1, CRYAB, CKMT2, MYBPC3, RYR2, PGAM2, SMPX, CHRNB1, CASQ2, MB 3.57×10−8
GO:0006941-striated muscle contraction 6 MYBPC3, RYR2, PGAM2, SMPX, CASQ2, MB 5.98×10−6
GO:0007517-muscle organ development 8 XIRP1, MYL2, CRYAB, DMD, MYBPC3, ITGB1BP2, ITGA11, CSRP3 1.94×10−4
GO:0007507-heart development 8 XIRP1, MYL2, MYBPC3, PDLIM3, SMYD1, CSRP3, CASQ2, MB 2.18×10−4
GO:0044057-regulation of system process 9 SLC8A1, CYP2J2, MYL2, BCHE, MYBPC3, NLGN1, RYR2, CNN1, CSRP3 3.63×10−4
GO:0031032-actomyosin structure organization 4 XIRP1, MYL2, MYOZ1, CNN1 5.24×10−4
GO:0030036-actin cytoskeleton organization 7 XIRP1, MYL2, PDLIM3, FHOD3, OPHN1, MYOZ1, CNN1 1.80×10−3
GO:0030029-actin filament-based process 7 XIRP1, MYL2, PDLIM3, FHOD3, OPHN1, MYOZ1, CNN1 2.48×10−3
GO:0006575-cellular amino acid derivative metabolic process 6 COLQ, BCHE, CKMT2, GSTT1, MOXD1, GHR 2.54×10−3

GO, Gene Ontology; count, gene numbers enriched in a specific GO term.

Significantly enriched pathways of the differentially expressed genes.

A. Upregulated genes
Description Count Genes P-value
hsa04512: ECM-receptor interaction   7 TNXB, COL6A3, ITGB6, COL1A2, COL6A1, LAMC2, VTN 1.47×10−3
hsa04510: Focal adhesion 10 FLT1, TNXB, COL6A3, ITGB6, COL1A2, COL6A1, LAMC2, VTN, HGF, FLNB 2.82×10−3
hsa00512: O-Glycan biosynthesis 4 GCNT4, GALNT12, GCNT1, ST6GALNAC1 9.75×10−3
hsa04610: Complement and coagulation cascades 5 PLAT, FGA, FGB, F3, BDKRB1 1.92×10−2
hsa00480: Glutathione metabolism 4 GPX2, GSTA2, ANPEP, GSTO1 3.81×10−2
B. Downregulated genes
Description Count Genes P-value
hsa05410: Hypertrophic cardiomyopathy 7 SLC8A1, MYL2, DMD, MYBPC3, ITGA11, RYR2, PRKAA2 3.22×10−6
hsa05414: Dilated cardiomyopathy 6 SLC8A1, MYL2, DMD, MYBPC3, ITGA11, RYR2 8.47×10−5
hsa05412: Arrhythmogenic right ventricular cardiomyopathy 4 SLC8A1, DMD, ITGA11, RYR2 6.52×10−3

Count, gene numbers enriched in a specific pathway term; ECM, extracellular matrix.

Significantly enriched biological process and pathway for genes in Module 2.

Category Description Count Genes P-value
GOBP GO:0007166-cell surface receptor linked signal transduction 6 PLAT, APLNR, PENK, P2RY14, BDKRB1, OXGR1 4.84×10−5
GOBP GO:0007186-G-protein coupled receptor protein signaling pathway 5 APLNR, PENK, P2RY14, BDKRB1, OXGR1 2.21×10−4
GOBP GO:0019233-sensory perception of pain 2 PENK, BDKRB1 1.18×10−2
KEGG hsa04080:Neuroactive ligand-receptor interaction 3 APLNR, P2RY14, BDKRB1 7.32×10−3
KEGG hsa04610:Complement and coagulation cascades 2 PLAT, BDKRB1 4.02×10−2

GO, Gene Ontology; BP, biological process; KEGG, Kyoto Encyclopedia of Genes and Genomes; Count, gene numbers enriched in a specific GO term or pathway term.

Significantly enriched biological process and pathway for genes in Module 3.

Category Description Count Genes P-value
GOBP GO:0007155-cell adhesion 5 COL6A3, ITGB6, ITGA11, COL6A1, LAMC2 9.81×10−5
GOBP GO:0022610-biological adhesion 5 COL6A3, ITGB6, ITGA11, COL6A1, LAMC2 9.86×100
GOBP GO:0007229-integrin-mediated signaling pathway 2 ITGB6, ITGA11 3.07×10−2
GOBP GO:0007160-cell-matrix adhesion 2 ITGB6, ITGA11 3.88×10−2
GOBP GO:0031589-cell-substrate adhesion 2 ITGB6, ITGA11 4.27×10−2
KEGG hsa04512:ECM-receptor interaction 7 COL6A3, ITGB6, COL1A2, ITGA11, COL6A1, LAMC2, SV2C 1.70×10−11
KEGG hsa04510:Focal adhesion 6 COL6A3, ITGB6, COL1A2, ITGA11, COL6A1, LAMC2 5.34×10−7

GO, Gene Ontology; BP, biological process; KEGG, Kyoto Encyclopedia of Genes and Genomes; Count, gene numbers enriched in a specific GO term or pathway term; ECM, extracellular matrix.