Mice
Wild-type (WT) C57BL6/J mice (male, 6 weeks old; GemPharmatech Laboratory Animal, Shanghai, China) were used in this study. This study and all animal procedures conformed to the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health (NIH publication No. 85 − 23, revised 1996) and were approved by the Experimental Animal Welfare and Ethics Committee, Shanghai University of Chinese Medicine (PZSHUTCM210312008).
Models of HF were established by ligation of the left anterior descending (LAD) coronary artery
Models of HF were established by ligation of the LAD coronary artery. LAD operation was performed on 6-week-old WT mice as described previously17.Briefly, the mice were anesthetized with 4% chloral hydrate using intraperitoneal injection. Subsequently, we placed the endotracheal tube below the glottis and connected the tube with the ventilator (HX-101E, Techman Software, Chengdu, China) to assist the breathing mice. Then, a cut was made between the third and fourth intercostal space, and a 7 − 0 nylon suture was used for permanent ligation of the LAD. Animals, dead within the first 24 h after the surgery, were excluded from the analysis. Sham mice received the same operation without coronary artery ligation. At the specified time points, mice were euthanized by cervical dislocation, and tissues were then collected for analyses.
Transthoracic echocardiography
Transthoracic echocardiography assesses heart geometry, systolic and diastolic function, as described earlier18.The Vevo2100 ultrasound system (VisualSonics, Toronto, Canada) were used. Mice were lightly anesthetized with 0.5% isoflurane until the heart rate stabilized at 400–500 beats per minute, and placed on a ECG platform. Parasternal long-axis images were obtained in two-dimensional B mode at the appropriate scanning head position to determine the maximum LV length, and the sampling line was placed at the maximum cross-section of the left ventricle to guide the recording of continuous M-mode echocardiography. Automatic calculation of left ventricular ejection fraction (LVEF), fractional shortening (FS), left ventricular end-systolic dimension (LVESD), and left ventricular end-diastolic diameter (LVEDD) from at least three distinct frames for each mouse.
Aerobic exercise training program
HF mice were trained with a treadmill ( Slope 0°), with the speed of 12 m/ min, the training time of 30 min/time, five times per week, for 4 weeks. Meanwhile, WT mice received fed conventionally without specific exercise training.
10X sample processing and single nucleus RNA sequencing (SnRNA-Seq)
Single nucleus suspensions were mixed-extracted from three biological replicate mice at indicated time points after LAD with Nuclear Separation Kit (Shanghai Biotechnology Corporation) according to the manufacturer’s instructions. The concentration of nuclear suspension was adjusted with the corresponding NB solution according to subsequent experiments. About 50,000 nuclei were loaded into one channel of the Single Cell Chip M for each sample using the Single Cell 3’ HT kit (10X Genomics) for Gel bead Emulsion generation in the Chromium X system. Following capture and lysis, cDNA was synthesized and amplified for 14 cycles. 50 ng of the amplified cDNA were used for each sample to construct Illumina sequencing libraries. Sequencing was performed on the NextSeq500 Illumina sequencing platform following 10x Genomics instructions for reads generation.
Bioinformatic analyses
Raw data processing
Raw sequencing data (bcl-files) were converted to fastq files with the Illumina bcl2fastq tool, integrated into the CellRanger (10X Genomics) suite (version 2.1.1). The CellRanger analysis pipeline was used to generate a digital gene expression matrix starting from raw data. Pre-build mouse genome (version mm10-1.2.0) was used as genome reference. The CellRanger count module was used to map reads with default settings and sequence length set to r1-length = 26 and r2-length = 50. Global mapping statistics, such as Estimated Number of Cells, Mean Reads per Cell and Median Genes per Cell for each experimental condition are reported in Supplementary Table S1.
Quality control (QC) and normalization of SnRNA-seq data
For further data pre-processing, filtering and QC were done using the Seurat package (v3.1.2). For each specimen, genes with counts in fewer than 10 nuclei were discarded to exclude random noise. Nuclei were filtered for genes (500 < nFeature_RNA < 5000), the proportion of mitochondrial genes (percent.mt < 0.15) and the proportion of ribosomal genes (percent.ribo < 0.05) to remove poor-quality nuclei potentially ascribed to doublets or other technical noise. For normalization, unique molecular identifier (UMI) counts for all nuclei were scaled by library size (total UMI counts), multiplied by 10,000 and transformed to a log scale.
Unsupervised dimensional reduction and single cell clustering
Highly variable genes (HVGs) were identified using the function FindVariableFeatures in Seurat, and 2000 HVGs were selected for each sample. Nuclei of all samples were integrated via canonical correlation analysis implemented in Seurat to correct potential batch effects and identify shared cell states across samples. Subsequently, the scaled data were subjected to linear dimensional reduction through principal component analysis (PCA). Using the first 40 PCA components, a shared nearest neighbor graph of the nuclei was calculated, followed by clustering and visualization with t-distributed stochastic neighbor embedding (t-SNE), which were performed using the FindClusters and RunTSNE functions, respectively. The clustering was done using a resolution of 0.4.
Marker identification and Cluster annotation
In this study, a targeted clustering analysis was performed within the most abundant cell types. The annotations of cell identity on each cluster were defined by the expression of known marker genes. The DotPlot function (Seurat package) was used to visualize the average expression of the known markers related to specific cell types, based on the overall expression profile of the nuclei, regardless of dropout events. On this basis, we employed Pecam1, Cdh5, Emcn, and Flt1 to classify “Endothelial cell”; Ttn, Myh6, Tnnt2, and Atp2a2 for “Cardiomyocyte”; Col1a2, Col8a1, Pdgfra, and Fstl1 for “Fibroblast”; Rgs5, Abcc9, and Pdgfrb for “Pericyte”; Myh11, Acta2, and Synpo2 for “Myofibroblast”; Cd163, Mrc1, and C1qa for “Macrophage”; Cd3e, Cd8a, Il2ra, and Tcf7 for “T cell”; Pax5, Cd22, Cd19, and Blnk for “B cell”. We performed the function check markers (Garnett package) to evaluate the ambiguity score and the relative number of cells for each cell type.
Differentially expressed gene (DEG) analysis
For the identification of DEGs in AET-HF and HF groups, we calculated the log2 fold change (log2FC) between these groups using the Seurat FindMarkers function. The significance of the difference was determined using Wilcoxon test with the statistical threshold of adjusted P-value < 0.05 and log2|FC| > 0.25.
Functional enrichment analyses
The upregulated and downregulated genes identified for each group (AET-HF or HF group) were used in subsequent functional enrichment analyses using R-based application enrichR, respectively. EnrichGO was applied to identify genes enriched in biological process (BP), molecular function (MF), and cellular component (CC). Meanwhile, enrichKEGG was applied to detect genes enriched in pathway maps, such as metabolic pathways, signal transduction, protein interaction, and other network-related pathways. All terms with a P-value (Benjamini or Benjamini– Hochberg adjusted) less than 0.05 were considered significant and ranked by adjusted P-value or the number of genes identified in the group.
Cell-cell communication analysis
CellCall was performed to infer inter- and intracellular communication pathways by integrating paired ligand-receptor and transcription factor (TF) activity among all cell types. The L (ligand) -R (receptor) -TF axis dataset was extracted from the KEGG pathway analysis. This method infers the potential interaction strength between two cell subsets based on gene expression level, and provides the significance through permutation test. Only those with a P- value < 0.01 were used for the prediction of cell-cell communication between any two cell types. In addition, CellCall embeds a pathway activity analysis method to help explore the main pathways involved in communication between certain cell types. In brief, the activity of pathway i was quantified according to the pathway activity score nPASi, based on Jaccard similarity coefficient, and the significance of pathway activity was estimated by hypergeometric testing.
Immunofluorescence
For immunofluorescence, hearts were collected and encased in optimal cutting temperature (OCT) compound (Sakura, Torrance, CA, USA). Frozen slices (7-µm thickness) were obtained and sealed with diluted donkey serum for 1 h at 25°C, followed by overnight incubation with primary antibody at 4°C. Visualized signal of secondary antibody (Invitrogen, Carlsbad, CA, USA) conjugated with Alexa. The following primary antibodies were used in these experiments: CD68 (AB53444, 1:300; Abcam, Cambridge, England), TNNI3 (#SAB2502170, 1:300; Sigma-Aldrich, St. Louis, MO, USA), Lgmn (sc-133234, 1:300; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and Myoglobin (AB77232, 1:300; Abcam, Cambridge, England). Sections were viewed and photographed using a confocal laser scanning microscope (LEICA TCS SP8, Oberkochen, Germany).
Terminal deoxynucleotidyl transferase mediated dUTP-biotin nick-end labeling (TUNEL) assay
TUNEL staining was performed with an In Situ Apoptosis Detection kit (Yeasen, Shanghai, China), according to the manufacturer’s instructions. Images were obtained using a laser scanning confocal microscope (LEICA TCS SP8, Oberkochen, Germany).
Histological Analysis
Hearts collected after mice were perfused with cold PBS, fixed with 4% paraformaldehyde for 24 hours, paraffin or OCT embedding, paraffin embedding interval of 6µm, OCT embedding interval of 8µm, and continuous sections were performed. Continuous sections were stained with Masson's trichrome19 for detection of myocardial fibrosis. The images were captured by a Leica microscope (DM6000B, Leica, Germany). In order to quantify myocardial fibrosis, 10 visual fields were randomly selected from 3 heart sections to calculate the percentage of Masson’s trichromatic positive staining area in the total myocardial area.
Western Blot Analysis
Cell particles and heart tissues samples were homogenized in RIPA Lysis Buffer (Beyotime, Shanghai, China) containing proteinase and phosphatase inhibitor cocktail. Protein concentrations were determined by a BCA kit (23225, Pierce, USA). Protein (30 to 50 mg total protein) was isolated by SDS–polyacrylamide gel electrophoresis (SDS-PAGE; Biorad, Hercules, California, USA) gels, and then transferred to polyvinylidene difluoride (PVDF) membranes. Subsequently, in TBS-T, the membranes were blocked in 5% non-fat dried milk at room temperature for 2 hours and incubated with indicated primary antibodies at 4°C overnight according to the requirements of each experiment. The primary antibodies used in the study were: TNNI3 (#SAB2502170, 1:2000; Sigma-Aldrich, St. Louis, MO, USA), Lgmn (sc-133234, 1:2000; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and Myoglobin (AB77232, 1:2000; Abcam, Cambridge, England) and GAPDH (HRP60004,1:10000; Proteintech,San Diego, MO, USA). The second day, it was washed three times with TBS-T for 10 min, and then incubated with horseradish peroxidase-conjugated secondary antibodies (1:3000) for 2 hours at room temperature. After washing in TBS-T for another three times, all membranes were detected using a chemiluminescent system, and signal intensities were analyzed with an Amersham Imager 600 (GE Healthcare, USA). Experiments were repeated three times and the target protein level was quantified by Image J and normalized to internal control.
Statistical Analysis
Results are presented as mean ± SEM for at least 3 independent assays unless otherwise noted. Data normality was determined by the Shapiro-Wilk test. The Student t test was used for 2-sample comparisons; 1-way ANOVA with Turkey post hoc tests was used for comparisons between multiple groups; and 2-way ANOVA was used for comparisons between multiple groups when there were 2 experimental factors. Nonnormal data were analyzed by Mann-Whitney U test or Kruskal-Wallis test with Dunn multiple comparisons. A value of P < 0.05 was considered statistically different. GraphPad Prism 8 and SPSS 22 for Windows was used to perform all statistical analyses.