Patients
This study was approved by the Ethics Committee of Fujian Provincial Hospital, and all subjects provided written informed consent following the Declaration of Helsinki. The clinical diagnostic criteria of HCM were based on the 2014 European Society of Cardiology Guidelines: HCM in adults was defined by a wall thickness of ≥ 15 mm in one or more LV myocardial segments as measured by any imaging technique (echocardiography, cardiac magnetic resonance imaging, or computed tomography), which is not explained solely by loading conditions [3]. Controls consisted of healthy subjects matched by sex and age without cardiac diseases. The exclusion criteria were as follows: history of hypertension for over 10 years, rheumatic disease, aortic stenosis, congenital heart and metabolic diseases (such as myocardial amyloidosis, Danon disease, and Pompe disease), cardiac hypertrophy of athletes, other organic heart diseases, trauma within 6 months, diabetes, surgery, cancer, or renal dysfunction.
Plasma collection and RNA isolation
The venous blood of patients with HCM and healthy controls was collected into BD Vacutainer venous blood collection tubes containing ethylenediaminetetraacetic acid. The plasma was separated, and total RNA from plasma was isolated using TRIzol LS Reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. The quantity and integrity of RNA were measured using a NanoDrop ND-1000 spectrophotometer (OD 260 nm, NanoDrop Technologies, Wilmington, DE, USA) and standard denaturing agarose gel electrophoresis, respectively.
lncRNA and mRNA microarrays and data analysis
Human lncRNAArraystar V4.0 was manufactured by Arraystar Inc. (MD, USA), and it covered more than 40,000 lncRNAs and more than 20,000 mRNAs in human genome. Transcript data were collected from authoritative sources, including NCBI RefSeq, Ensembl database, UCSC Genome Browser, and other sources from related literature. Every transcript was represented by 1–5 probesto improve confidence of statistical results. Microarray hybridization and collection of expression data were performed by KangChen Bio-tech, Shanghai, China.
Total RNAs from plasma were transcribed into complementary RNAs (cRNAs) and then labeled with Cyanine-3-CTP (Cy3) using the One-Color Quick Amp Labeling Kit (p/n 5190–0442, Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer’s instruction. Labeled cRNAs were purified, and the concentration and specific activity were measured using a NanoDrop ND-1000 spectrophotometer and hybridized onto the microarray using the Agilent Gene Expression Hybridization Kit (p/n 5188–5242). The hybridized arrays were washed and scanned using the Agilent Microarray Scanner (p/n G2565BA). Raw data were collected from scanned images (TIF format) by Agilent Feature Extraction (version 11.0.1.1). After the quantile normalization of raw data by GeneSpringGX v12.0 (Agilent Technologies), lncRNAs and mRNAs with fold change (FC) cut-offs of ≥ 1.5 and significant P values of < 0.05, which were identified by paired t test, were selected for further analysis. Differentially expressed lncRNAs(DElncRNAs) and differentially expressed mRNAs (DEGs) derived from t test were identified through Volcano Plot filtering. Hierarchical clustering was performed by R package, including gplots and function heatmap 2.
MiRNA sequencing and data analysis
MiRNA sequencing was carried out by KangChen Bio-tech using the Illumina Small RNA Sequencing Platform (San Diego, CA, USA) following the manufacturer’s instructions. Total RNAs from plasma were used to construct the miRNA sequencing library through the following steps: 3′ and 5′ adaptor ligation, cDNA synthesis followed by PCR amplification, and size selection (~ 135–155 bp PCR fragment, corresponding to ~ 15–35 nt small RNAs). Then single-stranded DNA denatured from the libraries were captured on Illumina flow cell, further amplified in situ as clusters, and sequenced for 51 cycles on an Illumina NextSeq 500 sequencer. The raw data were analyzed using routine algorithms (KangChen Biotech, Shanghai, China). Furthermore, differentially expressed miRNAs (DEmiRNAs) among groups were identified using edgeR (version 3.18.1) package with FC of ≥ 1.5 and P values of < 0.05.
Construction of the ceRNA network
The experimental procedure for microarray data acquisition, bioinformatic analysis, and network construction was performed following the flow chart shown in Fig. 1. In the ceRNA network, lncRNA and mRNA had the same expression trend. Thus, the correlation coefficient between DElncRNAs and DEGs was calculated using the Pearson correlation coefficient (PCC), and the pairs with PCC ≥ 0.9 and P < 0.05 were selected. The targeted miRNAsby DElncRNAs were searched using miRcode and StarBase, and only the DEmiRNA list was used to determine lncRNA–miRNA pairs. The targeted mRNAs by DEmiRNAs were searched using miRDB and TargetScan, and only the DEG list was used to determine miRNA–mRNA pairs. Then, theceRNA network was constructed, which showed that the lncRNA and mRNA were targeted, negatively co-expressed with the same miRNA among the selected lncRNA–mRNApairs, and visualized using Cytoscape V3.6.1. Then, all node degrees of the genes in ceRNA were calculated simultaneously using plugin CytoHubba.
Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis
GO analysis and KEGG databases were carried out to analyze differentially expressed genes. GO project provides a vocabulary to describe gene functions (http://geneontology.org/), including three domains (biological process, cellular component, and molecular function). KEGG analysis allows genes to be mapped to pathways in metabolism, various cellular processes, and many human diseases. The Database for Annotation, Visualization and Integrated Discovery (DAVID) was used to find the potential function and underlying mechanism of differentially expressed genes [13, 14].
Quantitative real-time polymerase chain reaction (qRT-PCR) validation
qRT-PCR was performed to validate hub genes. In brief, total RNA was reversetranscribed to cDNA using GoScript™ Reverse Transcription Mix (Promega), and then qRT-PCR was performed using GoTaq qPCR Master Mix (Promega) in the StepOne™ Real-Time PCR System (Thermo Scientific). The relative expression level of RNAs was normalized to the internal control β-actin, and all data were calculated using the 2−ΔΔCt method. The primers for qRT-PCR were as follows: LINC00310 (GenBank Accession No. NR_027266) F: 5′-CAGCTTCAGAGAGTTCGAGTA-3′, R: 5′-CTCACAGAAACACCCAGAATA-3′; β-actin F: 5′-GTGGCCGAGGACTTTGATTG-3′, R: 5′-CCTGTAACAACGCATCTCATATT-3′.
Statistical analyses
Data with a normal distribution were presented as mean ± SD. No normal data were compared with Mann–Whitney U test and expressed as median (P25–P75). Statistical analyses were carried out using SPSS v23.0 and R v3.6. P < 0.05 was considered significant.