http://orcid.org/0000-0002-6617-5634
Figures
Abstract
Ventricular Septal Defect (VSD), the most common congenital heart defect, is characterized by a hole in the septum between the right and left ventricles. The pathogenesis of VSD is unknown in most clinical cases. There is a paucity of data relevant to epigenetic changes in VSD. The placenta is a fetal tissue crucial in cardiac development and a potentially useful surrogate for evaluating the development of heart tissue. To understand epigenetic mechanisms that may play a role in the development of VSD, genome-wide DNA methylation assay on placentas of 8 term subjects with isolated VSD and no known or suspected genetic syndromes and 10 unaffected controls was performed using the Illumina HumanMethylation450 BeadChip assay. We identified a total of 80 highly accurate potential CpGs in 80 genes for detection of VSD; area under the receiver operating characteristic curve (AUC ROC) 1.0 with significant 95% CI (FDR) p-values < 0.05 for each individual locus. The biological processes and functions for many of these differentially methylated genes are previously known to be associated with heart development or disease, including cardiac ventricle development (HEY2, ISL1), heart looping (SRF), cardiac muscle cell differentiation (ACTC1, HEY2), cardiac septum development (ISL1), heart morphogenesis (SRF, HEY2, ISL1, HEYL), Notch signaling pathway (HEY2, HEYL), cardiac chamber development (ISL1), and cardiac muscle tissue development (ACTC1, ISL1). In addition, we identified 8 microRNAs that have the potential to be biomarkers for the detection of VSD including: miR-191, miR-548F1, miR-148A, miR-423, miR-92B, miR-611, miR-2110, and miR-548H4. To our knowledge this is the first report in which placental analysis has been used for determining the pathogenesis of and predicting VSD.
Citation: Radhakrishna U, Albayrak S, Zafra R, Baraa A, Vishweswaraiah S, Veerappa AM, et al. (2019) Placental epigenetics for evaluation of fetal congenital heart defects: Ventricular Septal Defect (VSD). PLoS ONE 14(3): e0200229. https://doi.org/10.1371/journal.pone.0200229
Editor: Nathaniel A. Hathaway, UNC Eshelman School of Pharmacy, UNITED STATES
Received: June 15, 2018; Accepted: March 11, 2019; Published: March 21, 2019
Copyright: © 2019 Radhakrishna et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This study was supported by Beaumont Health, Royal Oak, MI-48084. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Congenital Heart Defect (CHD) affects nearly 40,000 newborns per year in the United States [1,2]. Ventricular Septal Defect (VSD) is the most common congenital heart disease (CHD) and occurs in approximately 1 in 500 live births [3–5]. The frequency is even more common in prenatal life. Common risk factors for VSD include family history, ethnicity, and a few genetic disorders. Data indicate that genetic mutations of cardiac developmental genes such as NKX2-5, SMAD3, NTRK3, GATA6, TBX2, TBX18, ATA6, and TBX2, unique copy number variations [6], and chromosomal aneuploidy such as trisomy 13, 18, or 21 can play an important role in the etiology of VSD [7]. Targeted disruption of the CHF1/Hey2 locus in mice resulting in VSD has also been reported [8]. Despite these known genetic disruptions, in the clinical setting the cause of VSD remains unknown in the great majority of cases.
The heart and the placenta are the two first organs to develop in pregnancy. Their development is inter-related, a phenomenon known as the ‘placental-heart’ axis. This relationship is not widely understood, one consequence of which has been insufficient studies in humans to determine how early placental function contributes to the downstream development of CHD. Multiple lines of evidence: genetic [9], hemodynamic and vascular [10], metabolic, eg. folate deficiency [9], and others [11], do suggest a significant potential role of the placenta in the development of non-syndromic CHD.
DNA methylation is one of multiple different but interrelated epigenetic mechanisms and is currently the most widely studied. DNA methylation is a significant player in the transcriptional regulation of genes and aberrant methylation has been implicated in many complex and common diseases including cancer, diabetes, and psychiatric disorders. While insufficiently studied, emerging evidence suggests that DNA methylation may have an important role in both normal and pathologic heart development [12]. Methylation of CpG islands is classically thought to have its biological effect by displacing transcription binding factors, and instead attracting methyl binding factors that condense DNA and suppress gene expression. Environmental factors such as maternal diet, smoking and alcohol exposure profoundly affect DNA methylation. These factors are known to be risk modifiers for CHD development [1,13,14]. Further, DNA methylation changes have been demonstrated in the cardiac tissue DNA of CHD fetuses [15].
While methylation is known to be tissue-specific it is not tissue-exclusive. Several recent studies have shown correlation between CpG methylation-specific cytosine loci across different tissues [16,17]. This has raised the prospect of using the DNA methylation status in easily accessible tissues, such as blood leucocytes, to evaluate methylation status and by inference gene activity and disease status in inaccessible organs such as the brain [18]. Epigenetic markers in blood leucocytes have thus been used to detect psychiatric disorders such as schizophrenia [19]. The degree of correlation in cytosine methylation across different tissues is thought to vary based on the specific tissues involved.
Currently, there are no biomarkers available in clinical practice for the pre- or post-natal detection of congenital heart defects (CHDs). Given the significance of CHD and the frequency of missed or late diagnosis, this is a major deficiency. In our prior pilot data [20,21] cytosine methylation status of blood leucocytes was found to be a potentially useful molecular biomarker for the detection of multiple different categories of CHD. This is consistent with attempts to use epigenetic signatures in leucocytes to detect psychiatric disorders [19]. The placenta, like newborn blood that was used in our prior studies, is a fetal tissue. We therefore reasoned that although not identical, there was likely to be a significant minority of cytosine loci in which parallel or correlated epigenetic modifications could be identified in cardiac development genes in general and specific genes related to ventricular development in the placenta from VSD pregnancies. Our objective therefore in this proof-of-concept study was to evaluate the utility of determining cytosine methylation patterns in placental DNA to elucidate the pathogenesis of and for the detection of isolated non-syndromic VSD. Further, we used cytosine methylation to investigate the molecular pathogenesis of non-syndromic VSD based on the associated genes and gene pathways that were differentially methylated in the VSD cases.
MicroRNA (miRNA) is another important epigenetic mechanism that exerts control over DNA methylation and suppresses gene expression, among other functions. Recent data suggest an important role for miRNA in CHD development [22]. We therefore also evaluated methylation status of known microRNA genes in lieu of measuring actual miRNA levels. Given that DNA methylation status is known to correlate with gene expression, this approach can be used to identify miRNAs that are involved in cardiac development and thus further elucidate the mechanism of VSD pathogenesis. To our knowledge, there are no prior reports using placental molecular analysis for the detection of investigation of CHD pathogenesis.
Materials and methods
IRB approval to evaluate placental tissue and review maternal and pediatric medical records was provided by Wayne State University. The need to obtain patient consent was waived. CHD was diagnosed based on newborn echo, with cardiac catheterization as clinically dictated. Known or suspected genetic syndromic cases were excluded from the study.
Tissue samples for analysis
Genomic DNA was isolated from archived formalin-fixed, paraffin-embedded (FFPE) tissue blocks using Puregene DNA Purification kits (Gentra systems MN, USA) according to manufacturer’s protocol. The tissue samples were taken from the placenta of VSD subjects immediately after birth. DNA from archived paraffin-embedded tissue is a suitable template and has been used previously for genome-wide DNA methylation profiles using the Infinium HumanMethylation450 BeadChip assay. Importantly, it has been demonstrated that it is possible to achieve high-performance outcomes using FFPE-derived DNA in genome mapping arrays [23]. FFPE-derived DNA captures on average more than 99% of the CpG sites on the array.
Genome-wide methylation analysis using the HumanMethylation450
The HumanMethylation450 BeadChip (Illumina, Inc., California, USA) contains >485,000 CpGs per sample in enhancer regions, gene bodies, promoters and CpG islands at a single-nucleotide resolution and requires only 500 ng of genomic DNA. Many studies have previously used Illumina Infinium technology to assess DNA Methylation changes in the placenta associated within fetal disorders [24]. DNA was bisulfite converted using the EZ DNA Methylation-Direct Kit (Zymo Research, Orange, CA), and fluorescently-stained BeadChips imaged by the Illumina iScan. Data were analyzed with Illumina’s Genome Studio methylation analysis package program. The methodology has been previously detailed earlier [20].
As previously reported, to avoid potential confounding factors, probes associated with sex chromosomes and/or containing SNPs in the probe sequence (listing dbSNP entries near or within the probe sequence, i.e., within 10 bp of the CpG site) were excluded from further analysis [25–27]. Probes targeting CpG loci associated with SNPs near or within the probe sequence may influence corresponding methylated probes [28]. The remaining CpG sites were analyzed.
Statistical and bioinformatic analysis
Differential methylation was assessed by comparing the ß-values per individual nucleotide at each CpG site between VSD subjects and controls. The p-value for methylation differences between case and normal groups at each locus was calculated [29]. Filtering criteria for p-values was set at <0.05 and <0.01 to identify the most discriminating cytosines or the most important differentially methylated regions of the genome. P-values were calculated with False Discovery Rate (FDR) correction for multiple testing (Benjamini-Hochberg test). Further analysis of the differentially methylated genes was conducted for potential biological significance. Receiver Operating Characteristic (ROC) Area Under Curve (AUC) was calculated with R to determine the diagnostic accuracy of specific cytosine loci differentiating CHD from control groups. Data were normalized using the Controls Normalization Method.
The most significantly differentially methylated CpG sites were selected based on pre-set cutoff criteria of ≥2.0-fold increase or ≥2.0-fold decrease with Benjamini -Hochberg FDR p < 0.01. Multiple CpG sites within a gene were resolved by selection of the CpG with the highest fold-change ranking and the lowest FDR p-value. Fold changes in methylation variation were obtained by dividing the mean ß-value for the probes in each CpG site in cases by that of the normal controls. After this filtering, a threshold was set to select ROC curves based on sensitivity plotted against specificity using different ß-value threshold at each CPG locus for VSD prediction, and the corresponding AUC (95% CI) was calculated. In the case of multiple differentially methylated cytosine loci in the same gene we used the locus with the highest discriminating power as defined. Thus only one CpG locus per gene was considered. To avoid potential experimental confounding, various statistical modeling approaches were used. Markers with AUC ≥0.80 and significant 95% CI and FDR p-value < 0.005, i.e., markers with good diagnostic accuracy, were further used to generate heatmap and pathway analysis.
Gene ontology analysis and functional enrichment
Pathway analysis was carried out using Ingenuity Pathway Analysis (Ingenuity Systems, www.ingenuity.com) and Wikipathways (https://www.wikipathways.org/index.php/WikiPathways) using differentially methylated genes at FDR p-value < 0.01. Literature data mining for co-occurrence of gene names and keywords of interest was performed using Chilibot (www.chilibot.net). Only genes for which Entrez identifiers were available were used in the Pathway analysis. Over-represented canonical pathways, biological processes and molecular processes were identified. Differentially methylated genes based on FDR p-value, percentage methylation change in VSD cases and controls, and those having an AUC of ≥0.80 were selected for enrichment analysis. Enrichment analysis was conducted with an enrichment evaluation statistical analysis test. The p-values were adjusted with Benjamini & Hochberg correction. Information on this function is available at http://rweb.stat.umn.edu/R/library/stats/html/p.adjust.html. The top 10 pathways with the most significant p values were identified and included in the S1 Table. The categories and Q-values from a False Discovery Rate (FDR)-corrected hyper geometric test for enrichment are reported. Q-values are estimated using the Benjamini-Hochberg procedure with a Q-value cutoff of 0.1.
The identified differentially-methylated genes were used to generate a heatmap using the Complex Heatmap (v1.6.0) R package (v3.2.2). Ward distance was used for the hierarchical clustering of samples [30].
Logistic regression analysis
To select candidate markers for VSD prediction, logistic regression analysis was performed using a more stringent criterion (FDR p ≤5x10-8) using MetaboAnalyst 4.0 [31].
Quantitative pyrosequencing
DNA methylation variations were validated on bisulfite-converted genomic DNA by quantitative pyrosequencing. For validation and verification experiments, bisulfite conversion of 1 μg DNA was performed using the EZ DNA Methylation GoldTM kit (Zymo Research, Cambridge Bioscience, UK). DNA methylation variations were compared with the data obtained from conventional quantitative pyrosequencing.
Results
Genome-wide DNA methylation analysis of placenta-derived DNA was performed on 8 VSD cases and 10 unaffected controls. Comparison of clinical characteristics between cases and controls are shown in Table 1. There were no significant differences between groups.
Identification of differentially methylated CpG sites in VSD placenta
We identified 1328 statistically significant differentially methylated (increased or decreased) CpG regions in 1328 genes in VSD vs control placenta (FDR p-value <0.001). The heat map (Fig 1) visually displays the differential methylation between loci in VSD cases versus controls. Using the most significantly methylated CpG sites as biomarkers for VSD prediction, a total of 80 individual loci (in 80 genes) had an AUC = 1.00 with significant 95% CI for VSD detection. Further, total of 1248 loci (in 1248 genes) had good diagnostic accuracy defined as AUC ≥ 0.81 to 0.99 for VSD detection. In Table 2 we provide the top 55 markers with the stringent p-value<5x10-8. The remaining markers will be provided on request to the corresponding author. The ROC curves for four high performing CpG loci with AUC-ROC is depicted in S1 Fig.
Unsupervised hierarchical clustering of differentially methylated loci (rows) in 8 affected and 10 control samples (columns). These 15 CpG loci are at least either 2.0-fold hypomethylated or 2.0-fold hypermethylated with False Detection Rate < 0.00001 in the disease (VSD) condition (red colored squares) compared to normal subjects (blue colored squares). The figure also displays direction, fold change in disease, probe relationship and probe annotation of differentially methylated CpG sites. The red color in the heatmap indicates hyper- DNA-methylation and blue, hypo-DNA-methylation values.
CpG sites with significant p-value of 5X10-8 indicating methylation status and area under the receiving operator characteristic curve ≥0.80 appear to have strong potential as diagnostic biomarkers for VSD.
Gene ontology and pathway analysis revealed an over-representation of pathways known to be involved in cardiac and ventricular chamber development (S1 Table). These genes included some known to be responsible for cardiac ventricle development (HEY2, FDR p = 0.002; ISL1, FDR p = 2.61E-5), cardiac septum formation (ISL1, FDR p = 2.61E-5), cardiac muscle development (ACTC1, FDR p = 7.15E-12) and others. In addition, Fig 2 displays pathways previously thought to be involved in cardiac development (Wikipathways). The associated legends indicate how to identify constituent genes found to be differentially methylated in the current study. Among the top differentially methylated genes were BMP4, TGFB1, PDGFRA, TBX2, EGR1, SRF, CXCR4, TLX3, DPP6, and FGFR1, known to be related to heart development or anomalies. In addition, the study also identified 8 microRNAs: miR-191, miR-548F1, miR-148A, miR-423, miR-92B, miR-611, miR-2110, and miR-548H4 (S2 Table). Significant differential methylation in the genes that coded for 52 Zinc finger proteins (S3 Table) were detected followed by 62 open reading frames (ORFs, S4 Table) and 28 LOCs (genes without a name and of unknown function, S5 Table) were detected. Additionally, differential gene methylation for 8 Small Nucleolar RNAs (SNOR, S6 Table) were identified and two non-coding RNAs (NCRNAs, S7 Table) were also identified. All CpG loci had individual AUC ROC ≥0.80 (at FDR p-value <0.001) for the detection of isolated VSD.
Previous studies identified several genetic components related to cardiac development and progression. Genes identified in the present study are shown are highlighted with yellow background.
We also used conventional logistic regression to identify a parsimonious combination of genes for prediction of VSD. The combination of cg20739526 (DPP6) and cg19393008 (KRT82) markers had an AUC value of 0.997 with 95% CI (1.0, 1.0) (S2 Fig).
We did not perform gene expression studies correlated with CpG methylation levels. We did however perform GDAC FIREHOSE database analysis which showed that these methylation differences are correlated with altered gene expression in other tissues such as cancer cells (S8 and S9 Tables, S3 Fig). S9 Table shows the CpG targets, H3k27Ac modification, transcription factors and location of binding sites along with the statistical significance.
Discussion
The placenta is now being recognized to play a “gate-keeper” role in cardiac development by way of multiple mechanisms [32]. There are hundreds of genes whose mutation produces both placental morphological and congenital heart defects [32]. The developing heart derives its supply of nutrients, including folate, from the placenta. Folate is the main source of the methyl group involved in the methylation of DNA. Folate is known to be protective of the placental-heart axis against environmental insults [9]. Several environmental insults are in turn known to play a critical role in the development of non-syndromic CHD. Folate deficiency has been linked to the development of CHD, primarily septal defects [33]. Other more striking examples of the role of the placenta in CHD development are commonly encountered in clinical obstetrics. There are two major arteries connecting the placenta to the fetal cardiovascular system, namely the umbilical arteries. Failure of development of one of the umbilical arteries (‘single umbilical artery’) is well-known to be associated with increased risk of several CHDs in the fetus [10]. Finally, blood flow, volume, and intra- cardiac pressures are known to play a decisive role in the development of CHD. In this context, the most dramatic example of the relationship of placental vascular abnormality and CHD is in twin-to-twin transfusion syndrome. This is a condition of monozygotic identical twins with monochorionic-diamniotic placentation. In monochorionic-diamniotic twins there is a single placenta in which there are extensive vascular connections between placental blood vessels, both arterial and venous, of each twin. In twin-to-twin transfusion syndrome, there is a gross transfer of blood volume through these connections favoring the “recipient” twin over the “donor” twin during monochorionic diamniotic twin gestation. As a consequence, there is a greater than 9-fold increase in CHD incidence in identical twins with monochorionic-diamniotic placentation compared to the general population, and the risk of CHD is more than two fold greater when monochorionic diamniotic twins develop twin-twin-transfusion syndrome compared to monochorionic diamniotic twins without twin to twin transfusion syndrome [34]. In twin-twin transfusion syndrome, treatment of placental vascular disorder with prenatal laser therapy is known to reverse some developing structural abnormalities of the heart. This confirms the important role of placental vascular development in the CHD pathogenesis. Further altered methylation of angiogenesis genes such as VEGF has been reported to regular placental angiogenesis [35]. Thus, providing a very credible link and potential mechanism by which the placenta can induce the development of CHD. For these reasons, we chose to evaluate placental epigenetics as a tool for understanding the pathogenesis of and prediction of non-symptomatic CHD, the dominant group of congenital heart defects.
We identified 1488 unique loci, one per gene, that were significantly differentially methylated in the placental DNA in cases of non-syndromic isolated VSD compared to controls. The methylation status of the CpG sites had high predictive accuracy for VSD detection with AUC ≥0.80. Indeed 80 of the cytosine loci in 80 genes had AUC = 1.0 at the FDR p-value <0.005. In addition, significant differential methylation was noted in genes coding for many ORFs, NCRAs and LOCs. These also demonstrated good to excellent diagnostic accuracy for VSD detection. We also identified methylation changes in 8 miRNAs that modulate the activity of 160 genes identified.
Cardiac progenitors and their role in heart development
Pathway analysis identified multiple genes known to be involved in cardiac development overall, and ventricular development and cardiac septum formation. These associations give biological plausibility to our findings (S1 Table). Recent work indicates that a common multipotent stem cell differentiates into cardiomyocyte, smooth muscle, and endothelial lineages, the major differentiated cell types of the heart [36–38]. These precursor cells are characterized by expression of the transcription factors Isl1 and Nkx2-5. Lineage tracing studies indicate that Isl1 and Nkx2-5 expression contributes to differentiation of the cardiomyocytes and a subset of the smooth muscle and endothelial cells of the heart [38]. Both genes were differentially methylated in CHD cases compared to controls in our study. We will now briefly discuss some of the genes and pathways found to be epigenetically dysregulated in our study and possible biological significance in CHD and VSD development.
TGF-β signaling in cardiomyocyte differentiation
The TGF-β superfamily members critically regulate many different processes within the cardiovascular system (CVS), including heart development and angiogenesis. The predominant TGF-β superfamily ligands expressed in the CVS are TGF-β1, TGF-β2, TGF-β3, BMP-2, BMP-4, BMP-6, and BMP-7 [39]. The importance of the TGF-β superfamily in cardiovascular development is demonstrated by the significant phenotypic changes in knock-out mice models. Disturbances in TGF-β signaling in mice causes embryonic lethality and the development of VSDs, myocardial thinning, and double outlet right ventricle (DORV), a cardiac anomaly of which VSD is a feature [40]. Moreover, BMP-dependent activation of transcription factors including the cardiac zinc finger transcription factor GATA4, homeobox protein NKX2.5, and myocyte enhancer factor 2C (MEF2C) augment differentiation mediated by the SMAD signaling pathways that regulate cardiomyocyte proliferation [41].
MicroRNAs and cardiac diseases
MicroRNAs (miRNAs) are endogenous, small, evolutionary conserved, non-coding RNAs molecules about 22 nucleotides in length that function in the regulation of gene expression [42]. They bind to and inactivate mRNA thus inhibiting protein production. miRNAs are commonly found in clusters through many different regions of the genome, most frequently within intergenic regions and the introns of protein-coding genes [43]. Altered miRNAs expression is known to occur in various cardiovascular disorders such as hypertension, congestive heart failure, CHD, coronary artery disease and stroke. In the present study, significant changes in the methylation of several miRNA genes: miR-191, miR-548f1, miR-148a, miR-423, miR-92b, miR-611, miR-2110, and miR-548h4 were observed in the placenta of VSD subjects as compared to controls. Our findings are supported by previous publications indicating these microRNAs are associated with cardiac disorders including VSD and heart failure [44–46].
Chen et al [47] suggested that miR-92b is involved in visceral and cardiac muscle differentiation and regeneration. Additionally, significant changes in miR-92b expression levels have been reported in neointima formation in a rat model of vascular injury [48], in individuals with heart failure [49], and during cardiac rehabilitation following surgical coronary revascularization [50], suggesting its involvement in these processes. In the present study, a significant variation in placental gene methylation was observed in VSD cases compared to controls.
Fetal / Embryonic cardiac muscle formation
The ACTC1 (Actin, Alpha, Cardiac Muscle 1) gene on chromosome 15q14 (MIM 102540) is the only actin expressed in embryonic heart muscle [51]. It has been suggested that the lack of ACTC1 may induce apoptosis, leading to disrupted cardiac differentiation. Apoptosis plays a crucial role in embryological development and excessive absorption/apoptosis of the primary septum is thought to be a cause of atrioventricular septal defects [52].
Jak / stat signaling in cardiac remodeling and proliferation:
One study found that Jak1/Stat3 downstream mediators are induced upon cardiac injury. Inhibition of Stat3 signaling in cardiomyocytes using a dominant-negative transgene led to fibrotic scarring and reduced cardiomyocyte proliferation [53].
Other significant genes involved in cardiac development:
Recent evidence has pointed to IGF2 involvement in cardiac development. IGF2 is expressed by epicardial cells during mid-gestation murine heart development, and IGF2-deficient embryos show reduced cardiomyocyte proliferation in the ventricular wall. Importantly, the cardiomyocyte-specific deletion of the genes encoding the IGF2 receptors IGF1R and Insr replicated this effect [54].
The inactivation of Hey2, a significant Notch signaling molecule, in mice resulted in a spectrum of cardiac malformations that resembled those associated with mutations of human JAG1 [55], associated with Tetralogy of Fallot (TOF), VSD, and tricuspid atresia. Sakata et al [8] reported that targeted disruption of the CHF1/Hey2 locus in mice resulted in VSDs.
Bone Morphogenetic Protein 4 (BMP4) on 14q22.2 region is a member of the TGF-β signaling family. The effect of BMPs are exerted through receptor-mediated activation of the transcription factor Smad by binding to heterotetrameric receptor complexes [56]. BMPs can induce formation of bone and cartilage, and play a major role in embryogenesis [57]. BMP4 plays a key role in cardiac development, with expression in ventral splanchnic arteries, branchial-arch mesoderm and outflow tract myocardium underlying the cushion-forming regions of the heart [58,59]. BMP4 alterations are known to be associated with defects in cardiac septation such as atrial septal defect (ASD), VSD, and atrioventricular septal defects (AVSD). The inactivation of BMP4 results in neonatal lethality [60,61].
Isl Lim Homeobox 1 (ISL1) on chromosome 5q11.1 is a transcription factor important in multipotent cardiovascular cell lineages [62]. The lineages of Isl1-expressing cells include neural crest, endocardium, endothelium, myocardium and smooth muscle cells [63]. Genetic mutations of ISL1 have been found to be associated with CHD and VSD [62,64].
Platelet-Derived Growth Factor Receptor, Alpha (PDGFR-A) proteins are receptor tyrosine kinases (RTKs) coded on chromosome 4q12 and are important in PDGF signaling. They specifically localize in the primary cilium of the heart where the downstream effectors PI3K–AKT and MEK1/2–ERK1/2 are active and regulate processes like cell cycle control and cell migration in fibroblasts [65,66]. Cilia play a major role in signaling processes which contribute to the left-right organ asymmetry, differentiation, morphogenesis and the maturation of the heart [66].
During heart development, T-BOX 2 (TBX2) on chromosome 17q23.2 is expressed in the outflow and inflow tracts, inner curvature and atrioventricular canal [67]. Sequence variants at the promoter sequences of TBX2 have been identified in VSD patients [6].
Early Growth Response 1 (EGR1) factor gene on chromosome 5q31.2 controls proliferation, differentiation, and apoptosis of cells, and is also involved in hypoxic and ischemic cardiovascular injuries [68]. EGR1 is a transcription factor for NAB1 and is involved in the regulation of physiological and pathological hypertrophy of the heart. Egr1 may not be required for normal development of muscle mass, but it acquires functional importance during the response to cardiac stress resulting from mechanical overload and hypertrophic signaling [69].
DPP6 is an essential component of the native cardiac channel complex (Kv4.3, KCND3) which regulates cardiac conduction [70]. DPP6 may serve as an additional beta-subunit responsible for the transient outward current generated during repolarization of the nodal cells in the human heart. Kv4.3, an outward flux potassium ion channel, is expressed extensively in ventricular myocytes.
A significant amount of inactive CaMKII forms a molecular complex with Kv4.3 potassium ion channel proteins in cardiomyocytes as a CaMKII reservoir [71]. CaMKII is also a positive regulator of T-Type calcium ion channels, one of which is CACNA1H [72]. This channel is responsible for transient calcium ion influx current (ICa,T) into myocardial cells following depolarization [72]. In ventricular myocardium, the T-type Ca2+ current channel CACNA1H, which is temporarily observed during fetal and neonatal periods, has been shown to reappear in failing/remodeling hearts. These findings suggest that CACNA1H is related to cell growth, proliferation, and development [73].
LOC gene(s) are predicted genes whose functions are uncharacterized, and that are currently unnamed. They are assigned to a generic gene and protein category “uncharacterized LOC” plus the GeneID [45]. In the present analysis, we have identified 28 LOC genes on 14 chromosomal regions (S5 Table). Each of these 26 CpG different targets have a ROC AUC ≥0.81 and the FDR p-value <0.005 for VSD prediction. Among these 28 candidates, LOC374443 [74], LOC255512 [75], LOC348926 [76], LOC96610 [77], LOC25845 [78] and LOC220930 [79] expression are associated with different heart ailments. LOC150381 was previously identified under rare Copy Number Variation (CNV) deletion associated with aortic valve stenosis [80].
As previously noted, though DNA methylation patterns are largely tissue specific recent data suggests correlation between disease-induced methylation changes across tissues [81]. For example, a high degree of correlation in CpG methylation was demonstrated between blood leucocyte and tissue DNA from the lip in cleft lip and palate patients [17]. This phenomenon of cross tissue correlation of methylation markers is the most likely explanation for the observed epigenetic modification in cardiac development genes in the placental of non-syndromic VSD cases.
Our study does have some limitations, one of which is the small sample size. We are now unable to evaluate the role of potential confounders (e.g. diet, obesity, and ethnicity) on methylation patterns. Our study, however, is a proof-of-concept study whose purpose is to establish the plausibility of a concept. The next step will be to explore this phenomenon in a larger study population.
While it is widely recognized that DNA methylation correlates with gene expression, we did not evaluate this correlation in our study. This is clearly an area for future study. While we did not perform gene expression studies, GDAC FIREHOSE database analysis suggested that methylation levels in many of our loci correlate with gene expression at least in other tissues (S8 and S9 Tables, S3 Fig).
Conclusions
To the author’s knowledge, this is the first study reporting the use of placental molecular markers for CHD detection and for elucidating the mechanism of CHD. We report significant epigenetic dysregulation in the placental DNA of VSD cases. Multiple genes and gene pathways including some known or suspected to be involved in heart and ventricular development were affected. These findings could have scientific and clinical significance in furthering understanding the pathogenesis of CHD and in the development of accurate biomarkers for CHD using surrogate tissues such as placenta. Placental tissue is clinically accessible in both the first and early second trimester when diagnosis of CHD is of optimal value. Further, placental tissue obtained at birth could also be used for the prediction of newborn CHD. Newborn screening for critical CHD is now a standard of care.
Supporting information
S1 Fig. (Online only) Receiver operating characteristic (ROC) curve analysis of methylation profiles for four specific markers associated with VSD.
The study identified 1328 CpG sites in 1328 genes with significantly differentially-methylated genes that have an area under the ROC curve ≥0.80. At each locus, the False Detection Rate p-value for the methylation difference between VSD subjects and controls was highly significantly different. Due to figure resolution concerns, we have included only four markers (chr 21; cg21161649) (chr 17; cg04245057) (chr 4; cg07809452) (chr 12; cg22129822). AUC: Area Under the Receiver Operating Characteristics Curve; 95% CI: 95% Confidence Interval. Lower and upper confidence intervals are given in parentheses.
https://doi.org/10.1371/journal.pone.0200229.s001
(PDF)
S2 Fig. (Online only). Logistic regression analysis of VSD cases and controls.
https://doi.org/10.1371/journal.pone.0200229.s002
(PNG)
S3 Fig. (Online only) Open chromatin and transcription factor occupancy in functional CpG sites.
Pie charts depicting H3K27Ac histone mark layering, location of CpG sites and transcription factor binding sites.
https://doi.org/10.1371/journal.pone.0200229.s003
(PDF)
S1 Table. (Online only) Gene Ontology (GO) terms enriched among the genes in the network displayed by GeneMANIA.
https://doi.org/10.1371/journal.pone.0200229.s004
(PDF)
S2 Table. (Online only) Differentially methylated CpG sites with Target ID, Gene ID, chromosome location, FDR p-value, and % methylation change in VSD cases and controls for each target methylated for MicroRNA.
https://doi.org/10.1371/journal.pone.0200229.s005
(PDF)
S3 Table. (Online only) Differentially methylated CpG sites with Target ID, Gene ID, chromosome location, FDR p-value, and % methylation change in VSD cases and controls for each target methylated for Zinc finger genes.
https://doi.org/10.1371/journal.pone.0200229.s006
(PDF)
S4 Table. (Online only) Differentially methylated CpG sites with Target ID, Gene ID, chromosome location, FDR p-value, and % methylation change in VSD cases and controls for each target methylated for open reading frame genes (ORFs).
https://doi.org/10.1371/journal.pone.0200229.s007
(PDF)
S5 Table. (Online only) Differentially methylated CpG sites with Target ID, Gene ID, chromosome location, FDR p-value, and % methylation change in VSD cases and controls for each target methylated for LOC (uncharacterized) genes.
https://doi.org/10.1371/journal.pone.0200229.s008
(PDF)
S6 Table. (Online only) Differentially methylated CpG sites with Target ID, Gene ID, chromosome location, FDR p-value, and % methylation change in VSD cases and controls for each target methylated for small non-coding RNA (SNOR) genes.
https://doi.org/10.1371/journal.pone.0200229.s009
(PDF)
S7 Table. (Online only) Differentially methylated CpG sites with Target ID, Gene ID, chromosome location, FDR p-value, and % methylation change in VSD cases and controls for each target methylated for non-coding RNA genes (NCRNAs).
https://doi.org/10.1371/journal.pone.0200229.s010
(PDF)
S8 Table. (Online only) Correlation of methylation mean with expression mean in various human tissues from GDAC data.
180 differentially methylated CpG targets were correlated with expression (RNA-seq) data. A bar chart was generated for each CpG target showing the proportion of methylation and mean of expression of the gene in which the CpG target resided.
https://doi.org/10.1371/journal.pone.0200229.s011
(PDF)
S9 Table. (Online only) Open chromatin conformation and transcription factor binding in differentially methylated CpG sites indicated their role in transcription initiation.
ENCODE data showing the H3K27Ac layering on each CpG site presenting an open chromatin conformation. These CpG targets were also occupied with various transcription initiation factors, mostly PolR2A. The position of each CpG site was also noted in respect to the gene in which it resided. Some of the differentially methylated CpG sites that were resided in intronic or 1st exonic regions, signifying their essential function in modulating transcription.
https://doi.org/10.1371/journal.pone.0200229.s012
(PDF)
Acknowledgments
We thank Rose Callahan, Beaumont Health, for critical manuscript review and editing.
References
- 1. Hoffman JI, Kaplan S (2002) The incidence of congenital heart disease. J Am Coll Cardiol 39: 1890–1900. pmid:12084585
- 2. Reller MD, Strickland MJ, Riehle-Colarusso T, Mahle WT, Correa A (2008) Prevalence of congenital heart defects in metropolitan Atlanta, 1998–2005. J Pediatr 153: 807–813. pmid:18657826
- 3. Botto LD, Correa A, Erickson JD (2001) Racial and temporal variations in the prevalence of heart defects. Pediatrics 107: E32. pmid:11230613
- 4. Bjornard K, Riehle-Colarusso T, Gilboa SM, Correa A (2013) Patterns in the prevalence of congenital heart defects, metropolitan Atlanta, 1978 to 2005. Birth Defects Res A Clin Mol Teratol 97: 87–94. pmid:23404870
- 5.
Park MK (2008) Pediatric cardiology for Practitioners. 5th ed. Philadelphia: Mosby Elsevier,. PLoS One: pp 166–175.
- 6. Pang S, Liu Y, Zhao Z, Huang W, Chen D, et al. (2013) Novel and functional sequence variants within the TBX2 gene promoter in ventricular septal defects. Biochimie 95: 1807–1809. pmid:23727221
- 7.
Prasad S (2013) Ventricular Septal Defect. In: Diagnosis and Management of Adult Congenital Heart Disease, Gatzoulis MA WG, Daubeney PEF (Eds),. Churchill Livingstone, Philadelphia 2003 p171.
- 8. Sakata Y, Kamei CN, Nakagami H, Bronson R, Liao JK, et al. (2002) Ventricular septal defect and cardiomyopathy in mice lacking the transcription factor CHF1/Hey2. Proc Natl Acad Sci U S A 99: 16197–16202. pmid:12454287
- 9.
Linask KK (2013) The Heart-Placenta Axis in the First Month of Pregnancy: Induction and Prevention of Cardiovascular Birth Defects. 11 p.
- 10. Araujo Junior E, Palma-Dias R, Martins WP, Reidy K, da Silva Costa F (2015) Congenital heart disease and adverse perinatal outcome in fetuses with confirmed isolated single functioning umbilical artery. J Obstet Gynaecol 35: 85–87. pmid:25020205
- 11. Maslen CL (2018) Recent Advances in Placenta-Heart Interactions. Front Physiol 9: 735. pmid:29962966
- 12. Serra-Juhe C, Cusco I, Homs A, Flores R, Toran N, et al. (2015) DNA methylation abnormalities in congenital heart disease. Epigenetics 10: 167–177. pmid:25587870
- 13. Blue GM, Kirk EP, Sholler GF, Harvey RP, Winlaw DS (2012) Congenital heart disease: current knowledge about causes and inheritance. Med J Aust 197: 155–159. pmid:22860792
- 14. Zaidi S, Choi M, Wakimoto H, Ma L, Jiang J, et al. (2013) De novo mutations in histone-modifying genes in congenital heart disease. Nature 498: 220–223. pmid:23665959
- 15. Wamstad JA, Alexander JM, Truty RM, Shrikumar A, Li F, et al. (2012) Dynamic and coordinated epigenetic regulation of developmental transitions in the cardiac lineage. Cell 151: 206–220. pmid:22981692
- 16. Fan S, Zhang X (2009) CpG island methylation pattern in different human tissues and its correlation with gene expression. Biochem Biophys Res Commun 383: 421–425. pmid:19364493
- 17. Alvizi L, Ke X, Brito LA, Seselgyte R, Moore GE, et al. (2017) Differential methylation is associated with non-syndromic cleft lip and palate and contributes to penetrance effects. Sci Rep 7: 2441. pmid:28550290
- 18. Lin X, Teh AL, Chen L, Lim IY, Tan PF, et al. (2017) Choice of surrogate tissue influences neonatal EWAS findings. BMC Med 15: 211. pmid:29202839
- 19. Melas PA, Rogdaki M, Osby U, Schalling M, Lavebratt C, et al. (2012) Epigenetic aberrations in leukocytes of patients with schizophrenia: association of global DNA methylation with antipsychotic drug treatment and disease onset. FASEB J 26: 2712–2718. pmid:22426120
- 20. Radhakrishna U, Albayrak S, Alpay-Savasan Z, Zeb A, Turkoglu O, et al. (2016) Genome-Wide DNA Methylation Analysis and Epigenetic Variations Associated with Congenital Aortic Valve Stenosis (AVS). PLoS One 11: e0154010. pmid:27152866
- 21. Bahado-Singh RO, Zaffra R, Albayarak S, Chelliah A, Bolinjkar R, et al. (2016) Epigenetic markers for newborn congenital heart defect (CHD). J Matern Fetal Neonatal Med 29: 1881–1887. pmid:26429603
- 22. Smith T, Rajakaruna C, Caputo M, Emanueli C (2015) MicroRNAs in congenital heart disease. Ann Transl Med 3: 333. pmid:26734643
- 23. Dumenil TD, Wockner LF, Bettington M, McKeone DM, Klein K, et al. (2014) Genome-wide DNA methylation analysis of formalin-fixed paraffin embedded colorectal cancer tissue. Genes Chromosomes Cancer 53: 537–548. pmid:24677610
- 24. Maccani JZ, Koestler DC, Houseman EA, Armstrong DA, Marsit CJ, et al. (2015) DNA methylation changes in the placenta are associated with fetal manganese exposure. Reprod Toxicol 57: 43–49. pmid:25982381
- 25. Liu Y, Aryee MJ, Padyukov L, Fallin MD, Hesselberg E, et al. (2013) Epigenome-wide association data implicate DNA methylation as an intermediary of genetic risk in rheumatoid arthritis. Nat Biotechnol 31: 142–147. pmid:23334450
- 26. Chen YA, Lemire M, Choufani S, Butcher DT, Grafodatskaya D, et al. (2013) Discovery of cross-reactive probes and polymorphic CpGs in the Illumina Infinium HumanMethylation450 microarray. Epigenetics 8: 203–209. pmid:23314698
- 27. Wilhelm-Benartzi CS, Koestler DC, Karagas MR, Flanagan JM, Christensen BC, et al. (2013) Review of processing and analysis methods for DNA methylation array data. Br J Cancer 109: 1394–1402. pmid:23982603
- 28. Daca-Roszak P, Pfeifer A, Zebracka-Gala J, Rusinek D, Szybinska A, et al. (2015) Impact of SNPs on methylation readouts by Illumina Infinium HumanMethylation450 BeadChip Array: implications for comparative population studies. BMC Genomics 16: 1003. pmid:26607064
- 29. Altorok N, Tsou PS, Coit P, Khanna D, Sawalha AH (2014) Genome-wide DNA methylation analysis in dermal fibroblasts from patients with diffuse and limited systemic sclerosis reveals common and subset-specific DNA methylation aberrancies. Ann Rheum Dis.
- 30.
Gu. Z (2015) ComplexHeatmap: Making Complex Heatmaps. R package version 1.6.0. https://githubcom/jokergoo/ComplexHeatmap.
- 31. Chong J, Soufan O, Li C, Caraus I, Li S, et al. (2018) MetaboAnalyst 4.0: towards more transparent and integrative metabolomics analysis. Nucleic Acids Research 46: W486–W494. pmid:29762782
- 32. Camm EJ, Botting KJ, Sferruzzi-Perri AN (2018) Near to One’s Heart: The Intimate Relationship Between the Placenta and Fetal Heart. Front Physiol 9: 629. pmid:29997513
- 33. Mamasoula C, Prentice RR, Pierscionek T, Pangilinan F, Mills JL, et al. (2013) Association between C677T polymorphism of methylene tetrahydrofolate reductase and congenital heart disease: meta-analysis of 7697 cases and 13,125 controls. Circ Cardiovasc Genet 6: 347–353. pmid:23876493
- 34. Bahtiyar MO, Dulay AT, Weeks BP, Friedman AH, Copel JA (2007) Prevalence of congenital heart defects in monochorionic/diamniotic twin gestations: a systematic literature review. J Ultrasound Med 26: 1491–1498. pmid:17957043
- 35. Sundrani DP, Reddy US, Chavan-Gautam PM, Mehendale SS, Chandak GR, et al. (2014) Altered methylation and expression patterns of genes regulating placental angiogenesis in preterm pregnancy. Reprod Sci 21: 1508–1517. pmid:24803508
- 36. Wu SM, Fujiwara Y, Cibulsky SM, Clapham DE, Lien CL, et al. (2006) Developmental origin of a bipotential myocardial and smooth muscle cell precursor in the mammalian heart. Cell 127: 1137–1150. pmid:17123591
- 37. Kattman SJ, Huber TL, Keller GM (2006) Multipotent flk-1+ cardiovascular progenitor cells give rise to the cardiomyocyte, endothelial, and vascular smooth muscle lineages. Dev Cell 11: 723–732. pmid:17084363
- 38. Moretti A, Caron L, Nakano A, Lam JT, Bernshausen A, et al. (2006) Multipotent embryonic isl1+ progenitor cells lead to cardiac, smooth muscle, and endothelial cell diversification. Cell 127: 1151–1165. pmid:17123592
- 39. Bobik A (2006) Transforming growth factor-betas and vascular disorders. Arterioscler Thromb Vasc Biol 26: 1712–1720. pmid:16675726
- 40. Sanford LP, Ormsby I, Gittenberger-de Groot AC, Sariola H, Friedman R, et al. (1997) TGFbeta2 knockout mice have multiple developmental defects that are non-overlapping with other TGFbeta knockout phenotypes. Development 124: 2659–2670. pmid:9217007
- 41. Pal R, Khanna A (2006) Role of smad- and wnt-dependent pathways in embryonic cardiac development. Stem Cells Dev 15: 29–39. pmid:16522160
- 42. Ambros V, Lee RC, Lavanway A, Williams PT, Jewell D (2003) MicroRNAs and other tiny endogenous RNAs in C. elegans. Curr Biol 13: 807–818. pmid:12747828
- 43. Macfarlane LA, Murphy PR (2010) MicroRNA: Biogenesis, Function and Role in Cancer. Curr Genomics 11: 537–561. pmid:21532838
- 44. Li D, Ji L, Liu L, Liu Y, Hou H, et al. (2014) Characterization of circulating microRNA expression in patients with a ventricular septal defect. PLoS One 9: e106318. pmid:25165856
- 45.
Thibaud-Nissen F, Souvorov A, Murphy T, DiCuccio M, Kitts P (2013) Eukaryotic genome annotation pipeline. Bethesda: National Center for Biotechnology Information. 10: e0119646.
- 46. Tijsen AJ, Creemers EE, Moerland PD, de Windt LJ, van der Wal AC, et al. (2010) MiR423-5p as a circulating biomarker for heart failure. Circ Res 106: 1035–1039. pmid:20185794
- 47. Chen Z, Liang S, Zhao Y, Han Z (2012) miR-92b regulates Mef2 levels through a negative-feedback circuit during Drosophila muscle development. Development 139: 3543–3552. pmid:22899845
- 48. Ji R, Cheng Y, Yue J, Yang J, Liu X, et al. (2007) MicroRNA expression signature and antisense-mediated depletion reveal an essential role of MicroRNA in vascular neointimal lesion formation. Circ Res 100: 1579–1588. pmid:17478730
- 49. Goren Y, Kushnir M, Zafrir B, Tabak S, Lewis BS, et al. (2012) Serum levels of microRNAs in patients with heart failure. Eur J Heart Fail 14: 147–154. pmid:22120965
- 50. Taurino C, Miller WH, McBride MW, McClure JD, Khanin R, et al. (2010) Gene expression profiling in whole blood of patients with coronary artery disease. Clin Sci (Lond) 119: 335–343.
- 51. Suurmeijer AJ, Clement S, Francesconi A, Bocchi L, Angelini A, et al. (2003) Alpha-actin isoform distribution in normal and failing human heart: a morphological, morphometric, and biochemical study. J Pathol 199: 387–397. pmid:12579541
- 52. Matsson H, Eason J, Bookwalter CS, Klar J, Gustavsson P, et al. (2008) Alpha-cardiac actin mutations produce atrial septal defects. Hum Mol Genet 17: 256–265. pmid:17947298
- 53. Fang Y, Gupta V, Karra R, Holdway JE, Kikuchi K, et al. (2013) Translational profiling of cardiomyocytes identifies an early Jak1/Stat3 injury response required for zebrafish heart regeneration. Proc Natl Acad Sci U S A 110: 13416–13421. pmid:23901114
- 54. Li P, Cavallero S, Gu Y, Chen TH, Hughes J, et al. (2011) IGF signaling directs ventricular cardiomyocyte proliferation during embryonic heart development. Development 138: 1795–1805. pmid:21429986
- 55. Donovan S, Shannon KM, Bollag G (2002) GTPase activating proteins: critical regulators of intracellular signaling. Biochim Biophys Acta 1602: 23–45. pmid:11960693
- 56. Derynck R, Zhang YE (2003) Smad-dependent and Smad-independent pathways in TGF-beta family signalling. Nature 425: 577–584. pmid:14534577
- 57. Wang J, Greene SB, Martin JF (2011) BMP signaling in congenital heart disease: new developments and future directions. Birth Defects Res A Clin Mol Teratol 91: 441–448. pmid:21384533
- 58. Delot EC, Bahamonde ME, Zhao M, Lyons KM (2003) BMP signaling is required for septation of the outflow tract of the mammalian heart. Development 130: 209–220. pmid:12441304
- 59. Lin CJ, Lin CY, Chen CH, Zhou B, Chang CP (2012) Partitioning the heart: mechanisms of cardiac septation and valve development. Development 139: 3277–3299. pmid:22912411
- 60. Qian B, Mo R, Da M, Peng W, Hu Y, et al. (2014) Common variations in BMP4 confer genetic susceptibility to sporadic congenital heart disease in a Han Chinese population. Pediatr Cardiol 35: 1442–1447. pmid:25022354
- 61. Posch MG, Perrot A, Berger F, Ozcelik C (2010) Molecular genetics of congenital atrial septal defects. Clin Res Cardiol 99: 137–147. pmid:20012542
- 62. Stevens KN, Hakonarson H, Kim CE, Doevendans PA, Koeleman BP, et al. (2010) Common variation in ISL1 confers genetic susceptibility for human congenital heart disease. PLoS One 5: e10855. pmid:20520780
- 63. Zhang T, Liu J, Zhang J, Thekkethottiyil EB, Macatee TL, et al. (2013) Jun is required in Isl1-expressing progenitor cells for cardiovascular development. PLoS One 8: e57032. pmid:23437302
- 64. Lang J, Tian W, Sun X (2013) Association between ISL1 variants and susceptibility to ventricular septal defect in a Chinese cohort. Molecular diagnosis & therapy 17: 101–106.
- 65. Clement DL, Mally S, Stock C, Lethan M, Satir P, et al. (2013) PDGFRalpha signaling in the primary cilium regulates NHE1-dependent fibroblast migration via coordinated differential activity of MEK1/2-ERK1/2-p90RSK and AKT signaling pathways. J Cell Sci 126: 953–965. pmid:23264740
- 66. Koefoed K, Veland IR, Pedersen LB, Larsen LA, Christensen ST (2014) Cilia and coordination of signaling networks during heart development. Organogenesis 10: 108–125. pmid:24345806
- 67. Harrelson Z, Kelly RG, Goldin SN, Gibson-Brown JJ, Bollag RJ, et al. (2004) Tbx2 is essential for patterning the atrioventricular canal and for morphogenesis of the outflow tract during heart development. Development 131: 5041–5052. pmid:15459098
- 68. Zins K, Pomyje J, Hofer E, Abraham D, Lucas T, et al. (2014) Egr-1 upregulates Siva-1 expression and induces cardiac fibroblast apoptosis. Int J Mol Sci 15: 1538–1553. pmid:24451137
- 69. Buitrago M, Lorenz K, Maass AH, Oberdorf-Maass S, Keller U, et al. (2005) The transcriptional repressor Nab1 is a specific regulator of pathological cardiac hypertrophy. Nat Med 11: 837–844. pmid:16025126
- 70. Radicke S, Cotella D, Graf EM, Ravens U, Wettwer E (2005) Expression and function of dipeptidyl-aminopeptidase-like protein 6 as a putative beta-subunit of human cardiac transient outward current encoded by Kv4.3. J Physiol 565: 751–756. pmid:15890703
- 71. Keskanokwong T, Lim HJ, Zhang P, Cheng J, Xu L, et al. (2011) Dynamic Kv4.3-CaMKII unit in heart: an intrinsic negative regulator for CaMKII activation. Eur Heart J 32: 305–315. pmid:21148163
- 72. Zhang Y, Jiang X, Snutch TP, Tao J (2013) Modulation of low-voltage-activated T-type Ca(2)(+) channels. Biochim Biophys Acta 1828: 1550–1559. pmid:22975282
- 73. Izumi T, Kihara Y, Sarai N, Yoneda T, Iwanaga Y, et al. (2003) Reinduction of T-type calcium channels by endothelin-1 in failing hearts in vivo and in adult rat ventricular myocytes in vitro. Circulation 108: 2530–2535. pmid:14581409
- 74. He C, Hu H, Wilson KD, Wu H, Feng J, et al. (2016) Systematic Characterization of Long Noncoding RNAs Reveals the Contrasting Coordination of Cis- and Trans-Molecular Regulation in Human Fetal and Adult Hearts. Circ Cardiovasc Genet 9: 110–118. pmid:26896382
- 75. Coles JG, Boscarino C, Takahashi M, Grant D, Chang A, et al. (2005) Cardioprotective stress response in the human fetal heart. J Thorac Cardiovasc Surg 129: 1128–1136. pmid:15867790
- 76. Kertai MD, Li YW, Li YJ, Shah SH, Kraus WE, et al. (2014) G protein-coupled receptor kinase 5 gene polymorphisms are associated with postoperative atrial fibrillation after coronary artery bypass grafting in patients receiving beta-blockers. Circ Cardiovasc Genet 7: 625–633. pmid:25049040
- 77. Deng DX, Spin JM, Tsalenko A, Vailaya A, Ben-Dor A, et al. (2006) Molecular signatures determining coronary artery and saphenous vein smooth muscle cell phenotypes: distinct responses to stimuli. Arterioscler Thromb Vasc Biol 26: 1058–1065. pmid:16456091
- 78. Thalji NM, Hagler MA, Zhang H, Casaclang-Verzosa G, Nair AA, et al. (2015) Nonbiased Molecular Screening Identifies Novel Molecular Regulators of Fibrogenic and Proliferative Signaling in Myxomatous Mitral Valve Disease. Circ Cardiovasc Genet 8: 516–528. pmid:25814644
- 79. Puig O, Yuan J, Stepaniants S, Zieba R, Zycband E, et al. (2011) A gene expression signature that classifies human atherosclerotic plaque by relative inflammation status. Circ Cardiovasc Genet 4: 595–604. pmid:22010137
- 80. Soemedi R, Wilson IJ, Bentham J, Darlay R, Topf A, et al. (2012) Contribution of global rare copy-number variants to the risk of sporadic congenital heart disease. Am J Hum Genet 91: 489–501. pmid:22939634
- 81. Song F, Mahmood S, Ghosh S, Liang P, Smiraglia DJ, et al. (2009) Tissue specific differentially methylated regions (TDMR): Changes in DNA methylation during development. Genomics 93: 130–139. pmid:18952162