The blood samples of LVNC patients and healthy subjects were obtained from May 2015 to October 2015 in the Department of Cardiology in the First Affiliated Hospital of Nanchang University. Six patients with LVNC were selected for the case-study group and six healthy subjects were selected as the control group. Clinical characteristics of LVNC patients and control subjects are presented in Table 1. Written informed consents were obtained from all patients and the study was approved by the Institutional Review Board (IRB) of the First Affiliated Hospital of Nanchang University. We followed all ethical compliance for the study.

Blood samples from each patient and healthy subject were centrifuged at 1,000 rpm for 10 min to separate plasma from cells. They were then stored at −80°C. Total RNAs from each sample were prepared from blood cells with the RNA isolation kit (Qiagen, Canada) after removal of rRNA from a Eukaryotic mRNA Isolation Kit, Epicenter. Then, each sample was amplified and transcribed into fluorescent cRNA along the entire length of the transcripts, without 3' bias, utilizing a random priming method (Arraystar Flash RNA Labeling Kit, Arraystar). The labeled cRNAs were purified by RNeasy Mini Kit (Qiagen). The concentration and specific activity of the labeled cRNAs (pmol Cy3/μg cRNA) were measured by NanoDrop ND-1000 (Table 2). One microgram and each labeled cRNA was fragmented by adding 5 μl 10 × Blocking Agent and 1 μl of 25 × Fragmentation Buffer. The mixture was then heated at 60°C for 30 min. Finally, 25 μl 2 × GE Hybridization buffer was added to dilute the labeled cRNA. Fifty microliter of hybridization solution was dispensed into the gasket slide and assembled to the lncRNA expression microarray slide (Agilent lncRNA microarray chip (catalog no.: G4851C, Agilent technology) that contained 26,803 Entrez unique genes and 30,606 unique lncRNAs. Microarray hybridization was performed according to the Agilent One-Color Microarray-Based Gene Expression Analysis protocol (Agilent Technology) with minor modifications. The slides were incubated for 17 h at 65°C in an Agilent Hybridization Oven. The hybridized arrays were washed, fixed, and scanned using the Agilent DNA Microarray Scanner (part number G2505C).

RT-QPCR was used to validate the microarray lncRNA expression among the six LVNC and six healthy patients. The primer sequences of β-actin (H, ENST00000423402, uc021ybu.1, uc002nqf.1, ENST00000550301, ENST00000417844, and ENST00000554022 were shown in Table 2. The primers used for RT-QPCR were designed and synthesized by Invitrogen (Shanghai, China). β-actin was used as an internal control (housekeeping gene). RT-QPCR was performed using GeneAmp PCR System 9700 (Applied Biosystems). The PCR reaction conditions were an initial denaturation at 95°C for 10 min, followed by 40 PCR cycles at 95°C for 10 s, and 60°C for 60 s. The products were amplified for 10 min by slowly heating them from 60 to 99°C after the extension. The target (lncRNA or gene) and housekeeping genes for each sample were used for Real-time PCR reactions. The target gene/lncRNA concentration of each sample was divided by the concentration of the housekeeping gene to obtain the relative concentration of the target gene for each sample. The amount of target sequence for the sample was derived from the standard curve throughout the experiment.

Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images. Quantile normalization and subsequent data processing were performed by using the GeneSpring GX v12.1 software package (Agilent Technologies). After quantile normalization of the raw data, lncRNAs and mRNAs which had flags in at least six out of 12 samples in Present or Marginal (“All Targets Value”) were chosen for further data analysis. Differentially expressed lncRNAs and mRNAs with statistical significance between the two groups of healthy and LVNC patients were identified through p-value or FDR filtering. Differentially expressed lncRNAs and mRNAs between the two samples were identified through Fold Change filtering. Hierarchical Clustering and combined analysis were performed and the differences between mRNAs were analyzed to infer their role in cellular pathway using Gene Ontology (GO) and KEGG.

Gene set enrichment analysis (GSEA) is essentially done as described by Subramanian et al. (20). The top 1,500 differentially expressed genes (DEG) are used with the Hallmark Gene Set database. Analysis was done with a “weighted” scoring scheme with “no collapse” as the main parameters. This analysis picked 13 enriched pathways. Among these pathways, the G2M transition, estrogen response, and inflammatory pathways had the highest enrichment score and were studied further. Individual gene expression of GSEA significant genes in these pathways across the LVNC and healthy subjects are performed in R package with “plot” and “bar plot” function.

Various biological functions and interactome analysis of DEG are determined by the use of FUNRICH software (21). The top 500 DEGs are used as input to carry out the analysis.

Information about biological pathways and target miRNA for these DEGs are deduced by using EnrichR software (22–24). The top 500 DEGs are also used as input to identify pathways and miRNA.

Microarray expression analysis identified numerous lncRNAs and mRNAs whose expression varies between LVNC and the control group. In this study, we used a t-test to identify differentially expressed RNAs. According to the Benjamini Hochberg FDR method and the corrected p-value (FDR), differentially expressed lncRNAs and mRNAs were screened with Fold change ≥ 2.0, FDR <0.05. Compared with the p-value filter, this method can better reduce the false positive rate. Two thousand six hundred eighty-two lncRNA (1,568 up and, 1,114 downregulated) and 2,083 mRNA (1,066 up and, 1,017 down) transcripts were identified in serum compared with the control group (log fold-change > 2.0). The top 10 upregulated lncRNA that were significantly downregulated were ENST00000550301, ENST00000433152, uc002nqf.1, ENST00000415393, ENST00000584722, TCONS_00029778, ENST00000443008, TCONS_00024700, NR_003256, ENST00000485225, TCONS_00006918, ENST00000423402, TCONS_00006919, uc022bqs.1, uc021ybu.1, ENST00000537478, and ENST00000447277. The differentially expressed top twenty lncRNA and mRNA were listed in Tables 3, 4. A Scatterplot is used for assessing the lncRNA expression variation between the two groups of LVNC and the control (Figures 1A,B). Hierarchical Cluster shows lncRNA expression patterns (Figure 1C).

By using quantitative real-time reverse-transcription PCR (RT-QPCR), the top three up-regulated lncRNAs and top three downregulated lncRNAs with log fold-changes > 2 were selected to test and verify the microarray data in different blood samples of LVNC patients and paired healthy people. Due to the failure of TCONS_00006918 and TCONS_00006919 primer design, we selected the downregulated expression ENST00000423402, uc022bqs.1, and uc021ybu.1 for PCR verification. The downregulated expression of ENST00000423402, uc021ybu.1, and the upregulated expression of uc002nqf.1 and ENST00000550301 were statistically significant (p <0.05). Two lncRNAs (ENST00000417844 and, ENST00000554022) were randomly selected to prove the consistency of microarray and qPCR. The expression of lncRNA uc002nqf.1, ENST00000550301, ENST00000417844 and, ENST00000554022 was up-regulated, and ENST00000423402 and uc021ybu.1 was downregulated in the LVNC samples vs. the control samples (Figure 2), consistent with the microarray results. Thus, microarray data profiling indicated a series of lncRNAs that were constantly differentially expressed between LVNC blood and healthy blood.

To establish the involvement of these RNAs in cellular pathways, we used GO databases to analyze their potential biological functions. As previously reported, GO analysis was performed to determine the gene and gene product enrichment, which covered three domains: biological processes (BP), cellular components (CC), and molecular functions (MF) (25).

We found that the highest GO classification targeted by underregulated transcripts were intracellular signal transduction (BP), plasma membrane (CC), and enzyme binding (MF) (Figures 3A–C). Meanwhile, the highest GO classification targeted by the overregulated transcripts were protein modification process (BP), an integral component of the plasma membrane (CC), and protein binding (MF), (Figures 3D–F). In addition, the KEGG pathway analysis demonstrated that the “bile secretion” and “thyroid hormone synthesis” were the most significant pathways containing differentially upregulated mRNAs. Increased bile secretion is shown to have an adverse effect and promotes heart failure in the long run and thus, may be relevant for developing LVNC (26, 27). Thyroid hormone disbalance has long been known to be a risk factor for heart diseases although its role in LVNC is not demonstrated (28, 29). The “neurotrophin signaling pathways” and “GnRH signaling pathways” were also important pathways containing differentially downregulated mRNAs (Figures 4A,B).

The Hallmark gene set with which GSEA has analyzed covers a wide range of pathways that would capture a set of genes represented in DEG. Our analysis with the top 1,500 genes identified 13 pathways that are enriched significantly. The rank order of genes with top enrichment score and their heatmap is shown in Supplementary Table 1 and Supplementary Figure 1. Among the pathways with the highest enrichment scores are the G2M transition pathway, Estrogen response pathway, and inflammatory pathway (Figure 5).

The significance of the G2M pathway is relevant to the LVNC since many stress-responsive genes often play a role in heart development and pathogenesis (30). The glutathione-S-Transferase pi 1 (GSTP1) gene is highly polymorphic and involved in many diseases including cancer and heart failure (31). Its role in LVNC could be further investigated. HSPA8 (Hsc-70) is reported to be a titin interacting protein in an LVNC gene-edited mice model. Therefore, this gene could have a critical role in pathogenesis (32). Trophinin associated protein (TROAP) is shown to be involved in myocardial infarction (33), thus, its role in LVNC could also be important. Dystrophin (DMD) is known to play a role in LVNC development and pathogenesis (34), justifying its higher expression in patients (Figure 6). MYB is a nuclear protein and acts as a transcription factor for many cellular pathways in cardiac development (35). Adenylate cyclase (ADCY1) is involved in cardiac development and cardiomyopathy (36). HSPB8 is a heat shock protein and is also involved in cardiac development and cardiomyopathy (37). IFITM1 is an interferon-induced transmembrane protein that helps with the proliferation of neonatal cardiomyocytes in rats (38). TNF-inducing gene (TNFSF15) plays a critical role in neovascularization and inflammation in the heart. Thus, it could also play a critical role in developing LVNC (39).

The biological functions of these DEGs are deduced as components, such as biological processes, transcription factors, cellular components, and molecular functions. Major biological processes include ion transport, transcription, purine cleavage, neurogenesis, and cell migration (Figures 7A–D). Interactome analysis identified a major pathway (Figure 8A) that includes CTLA4, CDKN2A, and HSPA8 which denotes cell cycle, apoptosis, autophagy, and cellular remodeling (40).

Analysis with EnrichR identified several important pathways (Supplementary Figure 2) such as transcription factor (TF) perturbation, epigenetic genes, and more importantly, miRNA. It used the miRtargetBase 2017 and targetScanmicroRNA2017 database and identified miRNAs that are linked to these DEGs (Tables 5, 6). miRTargetBase miRNAs identified CISD2, CDKN2A, etc. (Figure 8B) as target genes, and miRTargetScan identified RUNXT1 and KSR2 (Figure 8C) as target genes. Their role in developing LVNC could be further established.

The top significant miRNAs, hsa-miR-3197 and hsa-miR-6780-3p are identified by miRTargetbase and miRTargetScan, respectively. The function of Has-miR-3197 has not been researched extensively though it has been reported to be a potential biomarker for diabetic retinopathy (41). However, its role in cardiovascular disease and LVNC pathogenesis could be investigated. The has-miR-6780-3p is shown to be downregulated in the plasma of patients with increased low-density lipoprotein (LDL) which could have an impact on developing cardiac dysfunction (42).