Mitochondrial dysfunction has been suggested to be the key factor in the development and progression of cardiac hypertrophy. The onset of mitochondrial dysfunction and the mechanisms underlying the development of cardiac hypertrophy (CH) are incompletely understood. The present study is based on the use of multiple bioinformatics analyses for the organization and analysis of scRNA-seq and microarray datasets from a transverse aortic constriction (TAC) model to examine the potential role of mitochondrial dysfunction in the pathophysiology of CH. The results showed that NADH:ubiquinone oxidoreductase core subunit S1- (Ndufs1-) dependent mitochondrial dysfunction plays a key role in pressure overload-induced CH. Furthermore, in vivo animal studies using a TAC mouse model of CH showed that Ndufs1 expression was significantly downregulated in hypertrophic heart tissue compared to that in normal controls. In an in vitro model of angiotensin II- (Ang II-) induced cardiomyocyte hypertrophy, Ang II treatment significantly downregulated the expression of Ndufs1 in cardiomyocytes. In vitro mechanistic studies showed that Ndufs1 knockdown induced CH; decreased the mitochondrial DNA content, mitochondrial membrane potential (MMP), and mitochondrial mass; and increased the production of mitochondrial reactive oxygen species (ROS) in cardiomyocytes. On the other hand, Ang II treatment upregulated the expression levels of atrial natriuretic peptide, brain natriuretic peptide, and myosin heavy chain beta; decreased the mitochondrial DNA content, MMP, and mitochondrial mass; and increased mitochondrial ROS production in cardiomyocytes. The Ang II-mediated effects were significantly attenuated by overexpression of Ndufs1 in rat cardiomyocytes. In conclusion, our results demonstrate downregulation of Ndufs1 in hypertrophic heart tissue, and the results of mechanistic studies suggest that Ndufs1 deficiency may cause mitochondrial dysfunction in cardiomyocytes, which may be associated with the development and progression of CH.
1. Introduction
Cardiac hypertrophy (CH) is a pathophysiological response characterized by increased thickness of the ventricular wall, greater myocardial cell volume, and enhanced myocardial contractility during the early stage of overload pressure [1]. Primarily, CH is the compensatory response for preservation of cardiac function; however, persistent CH is often associated with disturbed energy metabolism, deteriorated cardiac function, and interstitial fibrosis, which will eventually progress into heart failure [2, 3]. Heart failure caused by CH has been shown to be an independent risk factor for various cardiovascular diseases [4, 5]. To date, the pathophysiology underlying the progression of myocardial hypertrophy remains elusive. Thus, determination of potential molecular mechanisms is necessary for identification of novel and effective therapies to attenuate myocardial hypertrophy.
The contraction and relaxation of cardiomyocytes require a sufficient energy supply to meet the workload demand, and mitochondria are important primary organelles for energy production in cardiomyocytes [6–8]. Mitochondrial dysfunction was shown to be closely associated with the development of heart failure [9–11]. Under pathological conditions of CH, the activities of ATP synthase and mitochondrial oxidative phosphorylation complex are attenuated, which results in reduced production of ATP [12, 13]. Moreover, attenuated mitochondrial dynamics, reduced mitochondrial volume, and abnormal mitochondrial morphology were detected in cardiomyocytes in CH [14, 15]. Mitochondrial dysfunction was shown to increase the production of reactive oxygen species (ROS) via impaired electron transport chains, which can lead to increased oxidative stress and decreased energy production in cardiomyocytes [9, 16, 17]. Thus, restoration of impaired mitochondrial functions will provide novel strategies to attenuate the progression of CH. NADH:ubiquinone oxidoreductase core subunit S1 (Ndufs1) is one of the core subunits of mitochondrial complex I that regulates mitochondrial oxidative phosphorylation and ROS production [18–20]. However, the detailed role of Ndufs1 in the pathophysiology of CH is largely unknown.
In the present study, we initially demonstrated the deregulation of Ndufs1 in heart tissue of mice with CH by analyzing the GSE95140 scRNA-seq dataset. The expression of Ndusf1 was confirmed in heart tissue in a mouse model of CH. Furthermore, in vitro studies determined the molecular mechanisms of Ndusf1-mediated CH. The present study may provide novel insight into the role of Ndusf1 in the pathophysiology of CH.
2. Materials and Methods
2.1. Analysis of scRNA-seq and Microarray Datasets
RNA sequencing data for single cardiomyocytes were downloaded from the GSE95140 dataset of the GEO database [21]. This dataset is based on the GPL17021 platform and contains 396 single-cardiomyocyte transcriptomes of mice after transverse aortic constriction (TAC) or sham operation assayed on day 3 (D3), week 1 (W1), week 2 (W2), week 4 (W4), and week 8 (W8). The expression in each cell was detected by using the “DropletUtils” package. Gene expression in the cells was calculated using the “QC-Metrics” function in the “scater” package [22]. and were used for subsequent filtering. After filtering, the expression matrix of each sample was normalized by using the “NormalizeData” function of the “Seurat” package (version 3.0) [23]. The genes with the most pronounced differences between the cells were selected using the “FindVariableFeatures” function of the “Seurat” package. The “ScaleData” function was used to convert the expression data to linear scale. Then, principal component analysis (PCA) was performed using the “RunPCA” function of the “Seurat” package. Principal components (PCs) with were selected. “RunUMAP” of the “Seurat” package was employed to perform UMAP dimensionality reduction analysis. The “FindAllMarkers” function of the “Seurat” package was used to define the criteria for identification of differentially expressed genes (DEGs) as follows: cell population expression , , and .
The differentially expressed genes were validated using the GSE24454 microarray dataset. In this dataset, mice were sacrificed 4 weeks after aortic banding (AB) or sham procedure (sAB) and subsequent debanding, including banding and subsequent debanding (DB3) or sham procedure and subsequent debanding (sDB3); the data were obtained at various time points up to day 3 [24]. Thus, the CEL raw data and corresponding annotation platform file were downloaded and preprocessed by background adjustment, normalization, probe summarization, and log2 transformation of the expression values using the “Affy” package in R.
2.2. Gene Ontology (GO) Term Enrichment Analysis
GO enrichment analysis was performed using the “clusterProfiler” package in R [25]. Notably, the major GO terms of DE genes in biological processes, molecular functions, cellular components, and pathways were evaluated. The Benjamini-Hochberg method was used to adjust the original values. The GO terms corresponding to the DE genes were enriched with the threshold of correction value < 0.05. Additionally, the enrichment analyses of the biological processes of the hub genes were carried out with the ToppGene tool (https://toppgene.cchmc.org/), which is a web-based analytic tool used for functional enrichment analysis of the gene lists [26]. Additionally, the cellular compartment-specific protein-protein interaction network was constructed by the ComPPI database (https://comppi.linkgroup.hu/) [27].
2.3. Gene Set Enrichment Analysis (GSEA)
GSEA was used to assess the Kyoto Encyclopedia of Genes and Genomes (KEGG) maps involved in TCA-induced CH development based on time series analysis [28]. Initially, the Kolmogorov-Smirnov method was used to determine the enrichment score (ES); then, the statistical significance of ES was assessed using the empirical phenotype replacement test procedure. The enrichment score (NES) was derived by normalization of ES for each gene set. The false discovery rate (FDR) of each NES was determined.
2.4. Gene Set Variation Analysis (GSVA)
The GSVA package of R was used to analyze the activation of the gene sets by unsupervised and nonparametric scoring calculations [29]. The hub pathway-related scores were calculated by the GSVA method in each cell based on the transcription expression matrix after assigning various groups in the TAC model. Significant differences in GSVA scores between various groups were assessed by one-way ANOVA.
2.5. Animals and Surgical Intervention
All animal experiments were approved by the Animal Ethics Committee of Sun Yat-sen University (SYSU-IACUC-2020-000469). Sixteen male C57BL/6 mice (8 weeks old) were purchased from Sun Yat-sen University, and the mice were randomly divided into two groups, including the sham () and TAC groups (). Before operation, the animals were anaesthetized by intraperitoneal injection with . After the animals reached general anesthesia, a small incision was made in the second intercostal space at the left upper sternal border to open the chest cavity, and the animals were subjected to respiratory ventilation. After exposure of the aortic arch, TAC was performed by tying a 7-0 nylon suture ligature against a 27-gauge needle between the left common carotid artery and the brachiocephalic artery. Then, the needle was quickly retracted to complete the partial constriction procedure. Sham-operated mice were subjected to the same surgical procedures without transverse aortic constriction. The chest was closed with 5-0 nonabsorbable sutures. Postoperatively, the animals were subcutaneously injected with 1.0 mg/kg buprenorphine to relieve postoperative pain every 12 h for 3 consecutive days. The mice were closely monitored every day for body weight and any signs of labored breathing or postoperative pain.
2.6. Echocardiography
Four weeks after ascending TAC operation, the animals from the sham and TAC groups were subjected to echocardiography examination. Briefly, the mice were anaesthetized by 3% isoflurane using an anesthesia machine. The hair on the left chest was carefully removed, and cardiac geometry was determined from the parasternal long axis view with a probe frequency of 30 MHz using a small animal color ultrasonic diagnostic apparatus (Vevo 2100, VisualSonics, Toronto, Canada). The images of the left ventricular area were captured using M-type echocardiography. The interventricular septum (IVS) thickness and left ventricular posterior wall (LVPW) thickness were measured.
2.7. Evaluation of Cardiac Index
After assessment by echocardiography, the animals were sacrificed by an overdose of 5% isoflurane. The heart was immediately dissected and rinsed with ice-cold saline to remove blood clots. After draining the heart tissue on sterile paper, the whole weight of the heart was measured using a digital balance. The left ventricular weight (LVW) was determined by removing the atrium and right ventricle from the whole heart. The heart mass index (HMI) and left ventricular mass index (LVMI) were calculated as follows: ; . The length of the medial malleolar distance on the right hindlimb to the tibial plateau edge was defined as the tibia length (TL). The ratios of LVW to TL were used as an index of cardiac hypertrophy.
2.8. Hematoxylin and Eosin (H&E) Staining
After animals were sacrificed by an overdose of 5% isoflurane, a part of the heart tissue was fixed with 4% paraformaldehyde and embedded in paraffin. The paraffin-embedded heart tissue was sectioned into 5 μm sections and stained by hematoxylin and eosin. The stained sections were examined under a light microscope (Nikon, Tokyo, Japan).
2.9. Transmission Electron Microscopy (TEM)
The mitochondria in the heart tissue were evaluated by TEM. Briefly, the heart tissue was sectioned into 1 mm³ pieces, which were fixed with 4% glutaraldehyde and 1% osmic acid. Then, the tissue was dehydrated with acetone, embedded in Epon 821, and cut into 70 nm sections. Then, the sections were double stained with uranyl acetate and lead citrate. The mitochondria were examined using TEM (JEM-1230, Tokyo, Japan). Mitochondrial volume and mitochondrial number were evaluated based on the TEM images.
2.10. Rat Cardiomyocyte Culture
Neonatal Sprague-Dawley rats (1-2 days old) were sacrificed by cervical dislocation, and the heart was immediately dissected under sterile conditions. Ventricular tissue was isolated from the atria and digested in Hanks balanced salt solution containing 0.25% trypsin (Sigma-Aldrich, St. Louis, USA) at 37°C for 5 min, and the digestion cycle was repeated 10 times. After digestion, the supernatants were pooled and mixed with an equal volume of DMEM supplemented with 10% fetal bovine serum (FBS; Thermo Fisher Scientific, Waltham, USA). After centrifugation at for 5 min, the supernatant was discarded, and the cell pellet was resuspended in DMEM supplemented with 10% FBS. After incubation for 4 h at 37°C in a humidified 5% CO2 incubator, cardiomyocytes were collected from the medium. Cardiac fibroblasts adhered to the walls of the dishes. Cardiomyocytes were cultured in 6-well plates for 24 h and in fresh DMEM supplemented with 10% FBS for 2-3 days before in vitro assays.
2.11. Construction of the Ndusf1 siRNA and Overexpression Vectors and Ang II Treatment
The siRNAs targeting Ndusf1 (si-Ndusf1) and the corresponding scrambled siRNAs were designed and synthesized by RiboBio (Guangzhou, China). The vector for Ndusf1 overexpression was constructed by cloning the full-length Ndusf1 sequence into the pcDNA3.1 vector, and the empty pcDNA3.1 vector was used as the corresponding negative control. All plasmids were purchased from RiboBio. For transfections, rat cardiomyocytes were seeded in 12-well plates and cultured for 24 h; then, cardiomyocytes were transfected with various plasmids or siRNAs by using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, USA) according to the manufacturer’s protocol. Cardiomyocytes were collected for the experiments 24 h after the transfection. For angiotensin II (Ang II; Sigma-Aldrich) treatment, cardiomyocytes were seeded in 12-well plates and cultured for 24 h; then, cardiomyocytes were treated with 100 nM Ang II for 24 h and harvested for subsequent experiments.
2.12. Quantitative Real-Time PCR (qRT-PCR)
Total RNA from cardiomyocytes and heart tissue was extracted using TRIzol reagent (Invitrogen, Carlsbad, USA) according to the manufacturer’s protocol. RNA was reverse transcribed using a PrimerScript RT kit with gDNA eraser (Takara, Dalian, USA). Real-time PCR was performed using a SYBR Premix Ex Taq II kit (Takara) on an ABI7900 instrument (Applied Biosystems, Foster City, USA). The parameters for thermal cycling were as follows: 95°C for 15 s, 55°C for 15 s, and 72°C for 15 s for 40 cycles. The relative mRNA expression levels were determined by the comparative Ct method, and β-actin was used as the internal control.
2.13. Western Blot Assay
Proteins from cardiomyocytes or heart tissue were isolated using RIPA buffer supplemented with proteinase inhibitors (Sigma-Aldrich). The concentrations of the protein samples were measured by the BCA method. Equal amounts of proteins (50 μg) were resolved by gel electrophoresis and transferred to a polyvinylidene difluoride (PVDF) membrane. After blocking with 5% nonfat milk at room temperature for 1 h, the membranes were incubated with primary antibodies against NDUFS1 (1 : 1,000; CST, Danvers, USA), atrial natriuretic peptide (ANP; 1 : 1,000; CST), brain natriuretic peptide (BNP; 1 : 1,000; CST), myosin heavy chain beta (β-MHC; 1 : 1,000; CST), and β-actin (1 : 2,000; CST) at 4°C overnight. Then, the membrane was incubated with horseradish peroxidase-conjugated secondary antibodies (1 : 2,000; CST) for 2 h at room temperature. The immunoreactive bands were analyzed by using a chemiluminescence system (Bio-Rad).
2.14. Assessment of mtDNA Copy Number
The mtDNA/nDNA ratio was evaluated by using the qRT-PCR assay as described previously. The primers were designed to target mtDNA (NADH dehydrogenase: 1,5-AAACGCCCTAACAACCAT-3 and 5-GGATAGGATGC TCGGATT-3) and nDNA (β-actin: 5-ATGGTGGGAATGGGTCAGAA-3 and 5-CTTTTCACG GTTGGCCTTAG-3). The relative mtDNA copy number was calculated by normalizing the mtDNA content to the expression of the β-actin gene.
2.15. Assessment of Mitochondrial Membrane Potential (MMP)
MMP of cardiomyocytes was evaluated using a JC-1 mitochondria staining kit (Thermo Fisher Scientific). Briefly, cardiomyocytes ( cells/well) were plated in 96-well plates, treated for 24, and incubated with JC-1 fluorescent dye for 20 min at room temperature in the dark. The fluorescent staining by JC-1 was evaluated by fluorescence microscopy. JC-1 monomers were imaged at excitation and emission wavelengths of 490 nm and 530 nm, respectively; JC-1 aggregates were imaged at excitation and emission wavelengths of 525 nm and 590 nm, respectively.
2.16. Detection of Mitochondrial ROS
The production of mitochondrial ROS was determined by using a MitoSOX fluorescent staining kit (Thermo Fisher Scientific, Waltham, USA) according to the manufacturer’s protocol. Confocal laser scanning microscopy was used to capture fluorescent images, which were further analyzed using ImageJ software.
2.17. Flow Cytometry Analysis of ROS-Positive Cells
ROS production in cardiomyocytes was evaluated using the 2,7-dichlorofluorescein diacetate (DCF-DA) staining assay (Thermo Fisher Scientific). Briefly, the cells were incubated with DCF-DA for 30 min at 37°C in the dark, washed, resuspended in PBS, and maintained on ice for immediate assay by flow cytometry (BD Biosciences). The data were analyzed using FACSDiva software (BD) to calculate the number of ROS-positive cardiomyocytes.
2.18. Mitochondrial Mass Analysis Using MitoTracker Red Staining
MitoTracker Red staining was performed to assess mitochondrial mass. Briefly, cardiomyocytes were incubated with 100 nM MitoTracker Red for 30 min at 37°C. Fluorescence was detected at excitation and emission wavelengths of 490 and 516 nm, respectively, using an ELx-800 microplate reader (BioTek; Winooski, VT, USA).
2.19. Statistical Analysis
The data are presented as the . All data analyses were performed using GraphPad Prism software (version 8; GraphPad Software, La Jolla, USA). Statistical significance of differences between various treatment groups was assessed using unpaired Student’s -test or one-way ANOVA followed by the Bonferroni multiple comparison test. indicated statistical significance.
3. Results
3.1. scRNA-seq Clustering by the Seurat Package and Functional Enrichment Analysis
The number of principal components was set as 12, and cardiomyocytes were classified into six clusters based on UMAP visualization after batch correction (Figure 1(a)). A total of 3,408 highly variable genes were detected after normalization. Consequently, a total of 288 markers were identified by the Wilcoxon signed-rank test. These markers were then subjected to GO enrichment analysis. As shown in Figure 1(b) and Table S1, the genes were significantly enriched in “ribonucleotide metabolic process” (, ), “mitochondrion organization” (, ), “muscle cell development” (, ), “heart contraction” (, ), and “proton transmembrane transport” (, ) in the biological process category. Additionally, DE genes were significantly enriched in “mitochondrial protein complex” (, ), “myelin sheath” (, ), “respiratory chain” (, ), “intercalated disc” (, ), and “chaperone complex” (, ) in the cellular component category. For molecular function, the terms “structural constituent of ribosome” (, ), “electron transfer activity” (, ), “proton transmembrane transporter activity” (, ), “coenzyme binding” (, ), and “ubiquitin protein ligase binding” (, ) were also enriched.
(a)