You are viewing the site in preview mode

Skip to main content
Download PDF

Dysfunction in mitochondrial electron transport chain drives the pathogenesis of pulmonary arterial hypertension: insights from a multi-omics investigation

Respiratory Research volume 26, Article number: 29 (2025) Cite this article

Abstract

Background

Pulmonary arterial hypertension (PAH) is a progressive disorder that can lead to right ventricular failure and severe consequences. Despite extensive efforts, limited progress has been made in preventing the progression of PAH. Mitochondrial dysfunction is implicated in the development of PAH, but the key mitochondrial functional alterations in the pathogenesis have yet to be elucidated.

Methods

We integrated three microarray datasets from the Gene Expression Omnibus (GEO), including 222 lung samples (164 PAH, 58 controls), for differential expression and functional enrichment analyses. Machine learning identified key mitochondria-related signaling pathways. PAH and control lung tissue samples were collected, and transcriptomic and metabolomic profiling were performed. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis investigated shared pathways, and canonical correlation analysis assessed gene-metabolite relationships.

Results

In the GEO datasets, mitochondria-related signaling pathways were significantly enriched in PAH samples, in particular the electron transport chain (ETC) in mitochondrial oxidative phosphorylation system. Notably, the electron transport from cytochrome c to oxygen in ETC was identified as the most crucial mitochondria-related pathway, which was down-regulated in PAH samples. Transcriptomic profiling of the clinical lung tissue analysis identified 14 differentially expressed genes (DEGs) related to mitochondrial function. Metabolomic analysis revealed three differential metabolites in PAH samples: increased 3-phenyllactic acid and ADP, and decreased citric acid. Mitochondria-related genes highly correlated with these metabolites included KIT, OTC, CAMK2A, and CHRNA1.

Conclusions

Down-regulation of electron transport from cytochrome c to oxygen in mitochondrial ETC and disruption of the citric acid cycle homeostasis may contribute to PAH pathogenesis. 3-phenyllactic acid emerges as a potential novel diagnostic biomarker for PAH. These findings offer insights for developing novel PAH therapies and diagnostics.

Introduction

Pulmonary arterial hypertension (PAH) is a progressive disorder with vascular remodeling of pulmonary arteries and elevated pulmonary vascular resistance, posing a significant risk of right ventricular failure and potential mortality [1]. The estimated global prevalence of PAH is 15–50 cases per million individuals [2]. Despite the existence of therapeutic medications targeting prostacyclin and endothelin pathways for PAH, the prognosis of newly diagnosed PAH patients remains suboptimal, with a 5-year survival rate of 61.2% reported by the REVEAL registry, underscoring the need for improved outcomes in this population [3, 4]. Therefore, further elucidation of the molecular mechanisms underlying PAH pathophysiology could potentially advance therapeutic development and improve the prognosis of PAH patients.

As a multifactorial disease, the pathogenesis of PAH has been extensively explored. Endothelial dysfunction is hypothesized to play a key role in PAH, as long-term overexpression of vasoconstrictors such as endothelin-1 could exert effects on vascular tone and accelerate vascular remodeling [5]. Inflammation and immune dysregulation are critical factors in the pathogenic mechanisms of PAH as well, indicated by the growing evidence of accumulated inflammatory and immune cells in PAH lungs and walls of remodeled pulmonary vessels [6, 7]. The occurrence and development of PAH is also closely linked with mitochondrial dysfunction. In PAH, altered mitochondrial dynamics have been implicated in promoting cell proliferation/resistance to apoptosis phenotype in pulmonary arterial cells [8]. Furthermore, abnormal expression of reactive oxygen species (ROS), which are mainly produced by mitochondria, have been observed in PAH patients [9, 10]. Mitochondrial dysfunction is known to be associated with dysregulated metabolism, specifically characterized by a metabolic shift from oxidative phosphorylation to glycolysis [11]. However, the specific mitochondrial abnormalities that exert greater impact on the etiology of PAH have yet to be fully comprehended. The elucidation of the pivotal pathogenic mechanisms is crucial for developing more effective treatments for PAH.

The primary objectives of this study were to explore the key alterations in mitochondrial function and metabolic profiles in PAH, and to help identify potential diagnostic markers and therapeutic targets for PAH. We integrated three PAH microarray datasets from the Gene Expression Omnibus (GEO) and found significant enrichment of the electron transport chain (ETC)-related signaling pathways in mitochondrial oxidative phosphorylation system in PAH samples. Notably, the electron transport from cytochrome c to oxygen in ETC, which was down-regulated in PAH samples, was identified as the most crucial mitochondria-related pathway via machine learning. By further transcriptomic and metabolomic analyses, we identified 14 mitochondria-related genes and three differential metabolites (citric acid, 3-phenyllactic-acid, and ADP), suggesting that dysregulation of mitochondrial ETC and citric acid cycle might contribute to the development of PAH. These findings offer insights for the development of novel PAH diagnostics and therapies.

Materials and methods

GEO datasets analysis

The selection criteria of GEO datasets were as follows: (1) large sample size; (2) including PAH group and control group; (3) data interpretable. Three PAH datasets (GSE24988, GSE117261, and GSE53408) were retrieved from the GEO database (https://www.ncbi.nlm.nih.gov/geo/) for analysis. The basic information of these datasets is given in Tables 1 and 2. Samples in the GSE24988 dataset were obtained from the recipient organs of pulmonary fibrosis patients undergoing lung transplant, including 94 PAH samples and 22 controls without PAH. The GSE117261 dataset comprises 58 samples from PAH patients undergoing lung transplant and 25 control samples from failed donors (without matching recipients). The GSE53408 dataset includes 12 samples from end stage PAH patients who went through lung transplantation, and 11 control samples from normal tissue of cancer patients undergoing surgery (lobectomy).

Table 1 Basic information of the included datasets from GEO
Full size table
Table 2 Demographic characteristics of patients from the included GEO datasets
Full size table

In total, microarray gene expression data of 164 PAH and 58 control lung samples were obtained. The microarray probes were converted to corresponding gene symbols with the annotation file from the GPL6244 platform (Affymetrix Human Gene 1.0 ST Array) using R software (version 4.2.2, https://www.r-project.org/). The Robust Multichip Average (RMA) algorithm was utilized for gene expression normalization. Principal component analysis (PCA) was conducted after merging the gene expression profiles of the three datasets. The batch effects among them were assessed by the “sva” R package and removed by the “combat” function [12]. The three-dimensional (3D) PCA scatter plots before and after removing the batch effects were drawn to show the heterogeneity among the three datasets by the “scatterplot3d” R package. The difference in gene expression levels between PAH and control samples was estimated via “limma” package [13]. Genes with |log2 fold-change (FC)| ≥ 0.7 and Benjamini–Hochberg (B–H) false discovery rate (FDR) < 0.05 were identified as differentially expressed genes (DEGs) [14].

Functional annotation and enrichment analysis

Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) [15, 16] enrichment analyses were performed with the expression profiles of DEGs by the “clusterProfiler” R package [17]. The GO gene sets are comprised of biological process (BP), cellular component (CC), and molecular function (MF) subsets. Gene set enrichment analysis (GSEA) was conducted with the “clusterProfiler” and “msigdbr” R packages [18]. Gene set variation analysis (GSVA) was used to calculate the enrichment scores of mitochondria-related pathways in each sample based on the expression profiles of samples by the “GSVA” R package [19]. The mitochondria-related pathway terms and genes were obtained from the Molecular Signatures Database (MSigDB; https://www.gsea-msigdb.org/gsea/index.jsp, v2023.1.Hs). The normalized enrichment score (NES) reflected the degree to which a gene set was overrepresented in the groups. A B–H adjusted P value <0.05 was considered statistically significant.

Screening crucial mitochondria-related pathways by machine learning

Machine learning algorithms including Naive Bayes, logistic regression, Gradient Boosting Machine (GBM), Adaptive Boosting (Adaboost), and eXtreme Gradient Boosting (XGBoost) were used to screen the most significant mitochondria-related pathways based on the GSVA scores by R [20,21,22,23,24]. Naive Bayes is based on Bayes’ theorem with the “naive” assumption that features are independent of each other, which simplifies calculations [20]. Logistic regression models the probability that a given input belongs to a particular class using the logistic function [21]. GBM works by building a series of decision trees sequentially, where each tree corrects the errors made by the previous ones [22]. Adaboost operates by combining multiple weak learners to create a strong classifier [23]. XGBoost is an advanced implementation of the gradient boosting algorithm, incorporating several features such as gradient boosting framework, regularization, sparsity awareness, and cross validation [24]. Different algorithms were used for modeling to make the analysis more robust. The risk of overfitting was mitigated by a tenfold cross-validation approach and the control of model complexity. A total of 222 samples (164 PAH and 58 controls) were randomly divided into training (75%) and validation (25%) sets. The SHapley Additive exPlanations (SHAP) algorithm was used to interpret the results of the XGBoost model by quantifying the contribution of each feature to the prediction outcome for individual data points [25]. Receiver operating characteristic (ROC) and precision-recall (PR) curves were plotted to evaluate the performance of the machine learning models in the validation set by the “pROC” and “PRROC” packages. The area under curve (AUC) values on ROC and PR curves represented the accuracy of the models. A decision tree was drawn to show the contributions of each mitochondrial pathway in the model by the “rpart” package. The decision tree models decisions based on a hierarchical structure resembling a tree, where each internal node represents a feature, each branch represents a decision based on that feature, and each leaf node represents the outcome or prediction.

Participant inclusion

The study was approved by the Ethics Committee of The First Affiliated Hospital, Sun Yat-sen University (No. [2023] 553). Lung tissue specimens were collected from eight adult patients with PAH and eight non-PAH adult patients as the control group undergoing pulmonary lobectomy or partial pulmonary lobectomy at the First Affiliated Hospital, Sun Yat-sen University between February 2023 and April 2023. The lung tissue samples in the control group were collected at a site remote from tumor foci of early-stage lung cancer patients. The diagnosis of PAH was defined as a pulmonary artery systolic pressure (PASP) > 35 mmHg detected by noninvasive echocardiography [26, 27]. Exclusion criteria were as follows: (1) severe cerebrovascular diseases; (2) chronic progressive nephropathy. The baseline information of the PAH and control patients, including body mass index (BMI), pulmonary artery systolic pressure (PASP), forced vital capacity (FVC), and total lung capacity (TLC), were recorded. All subjects provided informed consent for the study.

Transcriptomic analysis

Total RNA was extracted from the lung tissue samples of PAH and non-PAH patients recruited in our hospital using the Direct-zol RNA MiniPrep Plus Kit (Zymo Research) for RNA sequencing. RNA was quantified by a NanoDrop 8000 UV–Vis Spectrophotometer (Thermo Scientific) and quality was assessed with the Bioanalyzer RNA 6000 Nano Kit (Agilent). Samples with sufficient quantity and quality (RNA integrity number ≥ 6) were retained for sequencing. The purified RNA was used to prepare directional cDNA libraries for sequencing using the NEBNext Ultra Directional RNA Library Prep Kit for Illumina. Quality control (before and after library preparation) was performed with the Bioanalyzer HighSens DNA Kit (Agilent) and the molarity was examined by the Kapa qPCR quantification kit. Paired-end, 75 base pair sequence was generated using an Illumina Novaseq 6000, with the version 2 chemistry. After de-multiplexing and adapter removal, sequence quality was assessed with FastQC. Samples with sufficient quality were aligned to the human genome (version GRCh38) using STAR, which also produced transcript counts. Samples with aberrant alignment characteristics (e.g. low number of mapped reads, high number of reads mapped to multiple locations) were excluded from analysis. Genes with |log2 fold-change (FC)| ≥ 1 and B–H adjusted P value <0.05 were identified as differentially expressed genes (DEGs).

Mass spectrometry (MS)-based metabolomic profiling

The liquid chromatography-electrospray ionization-tandem mass spectrometry (LC–ESI–MS/MS, QTRAP® 6500+) system was used for non-targeted metabolomic profiling. The positive and negative ion modes were analyzed by Analyst 1.6.3 software (Sciex). A mixture of the sample extracts was used for quality control and was inserted into analytical samples. We subsequently obtained the total ion current (TIC) chromatogram and multiple reaction monitoring (MRM) multi-peak chromatogram. A triple quadrupole mass spectrometer was used to select the characteristic ion of each substance. The signal intensity (counts per second) of the characteristic ion was acquired in the detector. MultiQuant software was used to integrate and calibrate the chromatographic peaks in the sample output files by the mass spectrometer. The chromatographic peak area represents the relative content of the corresponding substance, and the chromatographic peak of each metabolite in different samples was corrected based on the retention time and peak type.

Orthogonal partial least squares‑discriminant analysis (OPLS-DA)

Differential metabolites between PAH and control samples were identified by OPLS-DA, a method that combines orthogonal signal correction and partial least squares‑discriminant analysis to decompose the matrix information of independent variables into related- and unrelated-information of dependent variables [28]. Differential metabolites between groups were selected using variable importance in projection (VIP) of the OPLS-DA model by the “MetaboAnalystR” R package [29].

Metabolites with VIP ≥ 1.0 and |log2 fold-change (FC)| ≥ 1 were defined as significantly differential metabolites. The heatmap of differential metabolites was drawn by the “pheatmap” R package.

KEGG analysis of differential metabolites

Metabolites were annotated and mapped to potential metabolic pathways based on the KEGG compound and pathway database (http://www.kegg.jp/kegg/compound/), including carbohydrate, nucleotide, amino acid metabolism and organic substance biodegradation [30]. The hypergeometric test’s P value <0.05 was considered statistically significant for pathway enrichment.

Clustering and correlation analyses of DEGs and differential metabolites

DEGs were assigned into subgroups by k-means clustering. Correlations between gene expression were assessed by Pearson’s correlation coefficients. Relationships between DEGs and differential metabolites were analyzed by Pearson’s correlation analysis and canonical correlation analysis (CCA). A Pearson’s correlation coefficient >0.7 was considered as strong correlation.

Results

Identification of DEGs in the PAH datasets from GEO

Three microarray datasets from GEO with a total of 222 lung samples (164 PAH and 58 controls) were integrated for the analysis. Certain heterogeneity existed among the three datasets in the initial PCA graph (Fig. 1A). After removing batch effects, the gene expression data in the three datasets exhibit similar distributions, suitable for further analysis (Fig. 1B). The differential expression analysis identified 36 DEGs between PAH and control samples (Fig. 1C; Additional file 1). The genes displaying the most significantly up-regulation in PAH included peptidase inhibitor 15 (PI15), asporin (ASPN), collagen type xiv alpha 1 chain (COL14A1), and secreted frizzled related protein 2 (SFRP2), whereas the genes with the most remarkable down-regulation in PAH were S100 calcium binding protein A8 (S100A8), S100 calcium binding protein A12 (S100A12), and ficolin 3 (FCN3).

Fig. 1
figure 1

Identification of DEGs and KEGG pathways in PAH datasets from GEO. A The 3D scatter plot illustrating the heterogeneity among the three datasets through PCA analysis before batch effects were removed. Comp. for Component. B The 3D scatter plot of the three datasets post-batch effects removal, demonstrating reduced heterogeneity. Comp. for Component. C The volcano plot displaying DEGs between the PAH and control groups. Red and green dots indicate up-regulated and down-regulated DEGs in PAH, respectively. D KEGG analysis revealing six significantly enriched pathways among the DEGs

Full size image

Significant alterations of mitochondria-related pathways in PAH

Following the identification of DEGs in the PAH datasets, KEGG enrichment analysis was performed and six significantly enriched pathways were identified, including mitochondria-related pathways (calcium and cAMP signaling pathways) as well as cardiac disease-related pathways (arrhythmogenic right ventricular cardiomyopathy and hypertrophic cardiomyopathy) (Fig. 1D). The GO enrichment analysis demonstrated notable enrichment of pathways related to transmembrane transport, ion gated channel activity, and metabolic processes (Fig. 2A). As mitochondrial dysfunction plays an important role in PAH, we subsequently evaluated the mitochondria-related pathways using the GSVA scoring method. A total of 30 mitochondria-related pathways showed significant differences between PAH and control samples (Fig. 2B). The GSEA revealed substantial up-regulation of ATP dependent activity (NES = 1.896, adjusted P value = 6.85E−08) and down-regulation of the ETC in mitochondrial oxidative phosphorylation system (NES = −2.158, adjusted P value = 0.001) in PAH (Fig. 2C, D). Collectively, the development of PAH is linked with a variety of cell signaling pathways, particularly those that involve disruption in mitochondrial functionality.

Fig. 2
figure 2

Functional enrichment analysis of the DEGs in PAH datasets from GEO. A Pathways significantly enriched in DEGs through GO analysis, covering biological processes (BP), cellular components (CC), and molecular functions (MF). B The heatmap displaying the differences in GSVA scores for mitochondria-related pathways between the PAH and control groups. C,D GSEA (gene set enrichment analysis) revealing notable up-regulation of DEGs associated with ATP-dependent activity and concurrent down-regulation of DEGs linked to the electron transport chain in mitochondrial oxidative phosphorylation (OXPHOS) system in the PAH group

Full size image

Electron transport from cytochrome c to oxygen in ETC as the most critical mitochondria-related pathway in PAH

To determine the most crucial mitochondria-related pathway in PAH, machine learning algorithms including Naïve Bayes, logistic regression, GBM, Adaboost, XGBoost, and decision tree were employed. The AUCs on ROC and PR curves of the Naïve Bayes model were 0.631 and 0.819, respectively (Fig. 3A). The models of logistic regression (ROC-AUC: 0.635, PR-AUC: 0.838; Fig. 3A), GBM (ROC-AUC: 0.631, PR-AUC: 0.803; Fig. 3B), and Adaboost (ROC-AUC: 0.666, PR-AUC: 0.830; Fig. 3B) exhibited similar accuracy. In the decision tree model, the pathway of mitochondrial electron transport (cytochrome c to oxygen) accounted for the largest part, followed by the regulation of mitophagy and the activation of NOXA and translocation to mitochondria (Fig. 3C).

Fig. 3
figure 3

Screening crucial mitochondria-related pathways through machine learning in PAH datasets from GEO. A Receiver operating characteristic (ROC) and precision-recall (PR) curves illustrating the performance of the Naïve Bayes and logistic regression models. B ROC and PR curves of the GBM and Adaboost models. C The decision tree model identifying the five most significant mitochondria-related pathways in the PAH group

Full size image

The SHAP algorithm for XGBoost model interpretation identified five mitochondria-related pathways with the greatest influence on the model (Fig. 4A). The SHAP force plot and summary plot indicated that the pathway of mitochondrial electron transport (cytochrome c to oxygen) was of the greatest impact on the model (Fig. 4B, C), implying its optimal performance in distinguishing PAH from control samples. The GSEA showed the pathway of mitochondrial electron transport from cytochrome c to oxygen was significantly down-regulated in the PAH group (NES = −1.851, adjusted P value = 0.015; Fig. 4D). Taken together, the electron transport from cytochrome c to oxygen in mitochondria is postulated to be the most significant mitochondria-related pathway implicated in the pathogenesis of PAH.

Fig. 4
figure 4

Electron transport chain emerges as the most significant mitochondria-related pathway in PAH. A The XGBoost and SHAP models pinpointing the top five crucial mitochondria-related pathways in the PAH group. B The SHAP force plot providing a visual representation of the contributions of the five mitochondrial pathways to the base value in the SHAP model. C The SHAP summary plot showcasing the SHAP values, which indicate the relative impacts of the five mitochondrial pathways on the model. D The GSEA plot of the mitochondrial electron transport pathway (cytochrome c to oxygen) underlining its significance in PAH

Full size image

Transcriptomic analysis of clinical PAH lung tissue samples

To further investigate the findings, lung tissue samples from eight PAH and eight non-PAH patients were obtained for transcriptomic profiling. Both the PAH and control groups were Han Chinese. The PAH group included 4 males and 4 females, aged 65.50 ± 6.41 years old, with BMI 22.03 ± 2.48 kg/m2, PASP 43.13 ± 3.14 mmHg, FVC (%) 95.63 ± 14.90, and TLC (%) 92.50 ± 5.25. The control group included 5 males and 3 females, aged 60.63 ± 10.39 years old, with BMI 23.02 ± 3.01 kg/m2, PASP 20.38 ± 2.83 mmHg, FVC (%) 97.38 ± 6.52, and TLC (%) 97.50 ± 18.25.

Through the transcriptomic analysis, a total of 361 DEGs were classified into three subgroups by K-means clustering (Fig. 5A; Additional file 2). Through KEGG analysis of the DEGs, six major categories of pathways were significantly enriched, including the metabolism pathway (Fig. 5B). Among the 361 DEGs, 204 genes displayed up-regulated expression patterns and 157 genes were down-regulated in the PAH group compared with the controls (Fig. 5C). Subsequently, a total of 630 mitochondria-related genes (mito-genes) were acquired from the MSigDB (Additional file 3). Analysis of the Venn diagram revealed a shared set of 14 genes between the DEGs and the mito-genes, the majority of which exhibited elevated levels of expression in PAH (Fig. 5D, E). The interrelationships of the 14 genes were estimated by Pearson’s correlation coefficients, which demonstrated a close association between mitochondrial-related genes, including mitochondrially encoded ATP synthase membrane subunit 6 (MT-ATP6), MT-ATP8, mitochondrially encoded NADH dehydrogenase 4L (MT-ND4L), and mitochondrially encoded NADH: ubiquinone oxidoreductase core subunit 5 (MT-ND5) (Fig. 5F). Strong correlations were also observed between metallothionein 1E (MT1E), MT1F, and MT1G (Fig. 5F). These results indicate the involvement of altered metabolism and mitochondrial function in PAH at the transcriptional level.

Fig. 5
figure 5

Transcriptomic analysis of clinical samples from PAH patients and non-PAH controls. A The classification of DEGs between PAH samples and controls into three subgroups. B KEGG annotation and enrichment analysis of the DEGs. C 361 identified DEGs, comprising 204 up-regulated and 157 down-regulated genes. D The Venn diagram demonstrating 14 shared genes between DEGs and mitochondria-related genes (mito-genes). E The heatmap displaying the expression patterns of these 14 shared genes. F The correlation heatmap showing the interrelationships between the 14 shared genes based on Pearson’s correlation coefficients, including CHRNA1 (cholinergic receptor nicotinic alpha 1 subunit), CAMK2A (calcium/calmodulin-dependent protein kinase II alpha), CYP26B1 (cytochrome P450 family 26 subfamily b member 1), MMP9 (matrix metallopeptidase 9), MT-ND4L (mitochondrially encoded NADH dehydrogenase 4L), MT-ND5 (mitochondrially encoded NADH: ubiquinone oxidoreductase core subunit 5), MT-ATP6 (mitochondrially encoded ATP synthase membrane subunit 6), MT-ATP8 (mitochondrially encoded ATP synthase membrane subunit 8), OTC (ornithine transcarbamylase), KIT (kit proto-oncogene), LMNA (lamin A/C), MT1F (metallothionein 1F), MT1E (metallothionein 1E), and MT1G (metallothionein 1G)

Full size image

Metabolomic profiling of clinical PAH and control lung tissue samples

Metabolomic analysis was performed to investigate the metabolic alterations in PAH. The levels of 63 metabolites were detected in the lung tissue samples, and three of them (citric acid, 3-phenyllactic-acid, and ADP) exhibited significant alterations in PAH samples compared with controls (Fig. 6A; Additional file 4). Specifically, the PAH samples appeared to have decreased levels of citric acid and elevated levels of 3-phenyllactic-acid as well as ADP. To further identify key genes related to the differential metabolites, Pearson’s correlation analysis was performed between the expression levels of the DEGs and the three metabolites, and a correlation clustering heatmap was drawn based on the Pearson’s correlation coefficients between the DEGs and metabolites (Fig. 6B). KEGG pathway analysis was performed on the identified differential metabolites, and the results were integrated with the enriched KEGG pathways of DEGs identified through transcriptomic analysis to obtain the shared pathways. CCA analysis was performed between the DEGs and differential metabolites in the shared KEGG pathways, and DEGs were closely linked with citric acid (Fig. 6C). Four mito-genes were highly correlated with the metabolites (citric acid and ADP) as well, including cholinergic receptor nicotinic alpha 1 subunit (CHRNA1), ornithine transcarbamylase (OTC), kit proto-oncogene (KIT), and calcium/calmodulin dependent protein kinase II alpha (CAMK2A) (Fig. 6D). Through transcriptomic and metabolomic analyses, we identified three key metabolites (citric acid, 3-phenyllactic-acid, and ADP) and four related mito-genes (CHRNA1, OTC, KIT, and CAMK2A) in PAH pathogenesis.

Fig. 6
figure 6

Metabolomic profiling of clinical samples from PAH patients and non-PAH controls. A The heatmap displaying the top ten metabolites with the most significant logFC (logarithm of fold change) between PAH samples and controls. Metabolites with significant up-regulation in PAH are marked with red stars, while those with down-regulation are marked with blue stars. B The correlation clustering heatmap illustrating the relationships between DEGs (genes) and differential metabolites. meta for metabolites. C The CCA (canonical correlation analysis) graph illustrating the correlations between the DEGs (depicted as red dots) and metabolites (ADP and citric acid) in the shared KEGG pathways. D The correlation network showing four mitochondria-related genes highly correlated with ADP or citric acid, including CHRNA1, OTC, KIT, and CAMK2A

Full size image

Discussion

Mitochondrial dysfunction has been implicated in the pathogenesis and progression in a wide range of diseases, such as Alzheimer’s disease, chronic lung diseases, and various malignancies [31,32,33]. It also plays an important role in PAH pathophysiology [8]. Mitochondrial dynamic encompass the process of fusion and fission, which plays critical roles in maintaining the functional integrity of mitochondria [34]. Specifically, the expression of dynamin-related protein 1 (DRP1), a central mediator of mitochondrial fission, was markedly up-regulated in pulmonary artery smooth muscle cells (PASMCs) of PAH patients and could induce the proliferation of PASMCs [35]. Increased levels of citric acid cycle intermediates have been found in PAH patients compared with healthy subjects, suggesting possible dysfunction of mitochondrial citric acid cycle in PAH [36]. Nevertheless, the key mitochondrial functional changes in the development of PAH remain to be fully understood. Further elucidation of PAH pathogenesis would be beneficial for therapy development. Herein, we employed integrated transcriptomic and metabolomic analyses to explore the key genes and metabolites involved in the development of PAH.

In our analysis of the GEO datasets, KEGG pathway enrichment indicated that the DEGs in calcium signaling and cAMP signaling pathways were significantly enriched in the PAH group. Abnormalities in calcium signaling have been linked with pathological modulation of pulmonary vascular tone and pulmonary artery smooth muscle cell proliferation [37, 38]. Moreover, mitochondria served as sensors and regulators of calcium signaling have been shown to affect the Ca2+ feedback regulation of channel activity, and an increased uptake of mitochondrial Ca2+ was observed in pulmonary arterial smooth muscle cells from patients with PAH [39, 40]. The cAMP signaling pathway was also implicated in PAH. A recent study revealed that cAMP signaling pathway played an important role in PAH pathogenesis [41]. In addition, the cAMP-protein kinase A (PKA) signaling can regulate mitochondrial functions, and alterations of mitochondrial cAMP-PKA signaling have implications in the pathogenesis of various diseases [42]. Given the close connections between mitochondria and calcium signaling as well as cAMP signaling pathways, and the growing evidence of the role of mitochondrial dysfunction in PAH, we further examined the alterations of mitochondria-related pathways in the PAH group, and sifted the most critical signals by machine learning. Results revealed that the ETC in mitochondrial oxidative phosphorylation system was significantly down-regulated in PAH, especially the process of electron transport from cytochrome c to oxygen. It was suggested that the main proteins of the ETC involved in mitochondrial dysfunction might contribute to PAH [43]. The mitochondrial ETC (also known as respiratory chain) as a major part of mitochondrial oxidative phosphorylation system contains four enzymatic complexes (complex I to IV), and complex IV (cytochrome c oxidases, COX) receives electrons transferred by cytochrome c to reduce oxygen to water for ATP production [44]. COX4I2, one of the major gene participating in the process of electron transport from cytochrome c to oxygen, is essential for acute pulmonary oxygen sensing and involved in hypoxia-induced pulmonary hypertension [45, 46]. Our findings suggest that dysfunction of the electron transport from cytochrome c to oxygen in mitochondria might contribute to PAH.

To further investigate the findings derived from public datasets, we obtained PAH and non-PAH lung tissue samples from patients for transcriptomic and metabolomic analyses. In the transcriptomic analysis, 14 DEGs were mitochondria-related genes, four of which (KIT, OTC, CAMK2A, and CHRNA1) were highly linked with the differential metabolites identified through subsequent metabolomic profiling. Similar to our findings, quantitative immunohistochemistry in human lung PAH tissues demonstrated an elevation in c-KIT level [47], and c-KIT-positive cells were reported to participate in vascular remodeling in PAH [48]. OTC, which had higher gene expression levels in PAH lung tissue samples in the present study, also exhibited a significant increase in PAH gut bacteria [49]. As for CAMK2A, it has been reported that alpha1A-adrenoceptor is involved in the proliferation of pulmonary artery smooth muscle cells via CaMKII signaling [50]. In addition, variation in CHRNA1 genetics was highly associated with diastolic blood pressure [51]. These results indicate that mitochondria-related genes are involved in the development of PAH.

The regulation of gene expression is intricately linked to cell metabolism. Metabolic pathways provide precursor molecules and ATP that are necessary for gene expression [52]. Mitochondria play a pivotal role in cellular metabolism, and mitochondrial dysfunction has been implicated in the pathogenesis of various metabolic diseases [53]. Hence, we conducted metabolomic profiling to examine the metabolic alterations in PAH, and elevated levels of ADP were observed. It has been proposed that purinergic signaling activated by an alteration of nucleotides contributes to the pathogenesis of PAH [54]. Moreover, activation of purinergic P2Y1-receptor (P2Y1R) and P2Y12R mediated by ADP plays an important role in pulmonary vascular remodeling and inflammation in PAH [54]. Our metabolomic analysis also revealed a significant decrease in the levels of citric acid in the PAH samples. Conversely, increased levels of citric acid or citrate (the conjugate base of citric acid) in PAH patients were observed in other studies [55, 56]. The discrepancy should be further explored. Citrate is a significant intermediate produced in the initial step of the citric acid cycle (also known as the tricarboxylic acid cycle), subsequently undergoing conversion into isocitrate [57]. The citric acid cycle, a vital metabolic pathway that takes play in mitochondria, plays a pivotal role in regulating cellular physiology and maintaining homeostasis, and the metabolites generated during the citric acid cycle exert significant control over various cellular processes, influencing both physiology function and disease development [57]. Dysfunction of the citric acid cycle has been reported to occur in PAH and might potentially contribute to the pathogenesis of PAH [8]. In addition, increased levels of 3-phenyllactic-acid were found in the PAH samples in the current study. 3-phenyllactic acid is a product of phenylalanine catabolism generated by human host and intestinal bacteria [58]. It regulates human immune system and acts as a mediator of bacterial-host interactions [58]. Phenyllactic acid can be accumulated in the blood and tissues of patients with phenylketonuria [59]. It also promotes cell migration and invasion in cervical cancer by up-regulating the expression of the mitochondrial protease MMP9 [60]. In fungi, 3-phenyllactic acid could disrupt cell membrane and interfere with mitochondrial energy metabolism [61, 62]. Our findings suggest that 3-phenyllactic-acid might potentially be a novel biomarker for PAH.

Certain limitations existed in the current study. Limited lung tissue samples from Han Chinese patients were available for analyses, which might affect the broader applicability of the results. Access to larger sample sizes and samples from different ethnicities will allow further exploration of the present results. Moreover, future in vitro experiments would provide further insights into the potential pathogenic mechanisms of PAH.

Conclusions

The present study revealed a significant down-regulation of mitochondrial ETC in PAH, and the process of electron transport from cytochrome c to oxygen was identified as the most crucial mitochondria-related pathway in PAH. Metabolomic profiling demonstrated decreased levels of citric acid and elevated levels of 3-phenyllactic-acid as well as ADP in PAH lung tissues. The integration of transcriptomic and metabolomic analyses suggested that disruption of mitochondrial ETC and citric acid cycle homeostasis may contribute to the pathogenesis of PAH, and 3-phenyllactic acid is a potential emerging biomarker for the diagnosis of PAH. These findings could provide insights into further understanding of PAH pathogenesis and presenting potential diagnostic biomarkers and therapeutic targets of PAH.

Availability of data and materials

No datasets were generated or analysed during the current study. The datasets included in this article (GSE24988, GSE117261, and GSE53408) are available in the GEO database (https://www.ncbi.nlm.nih.gov/geo/). The transcriptomic and metabolomic data of PAH and control patient samples can be accessed on the Science Data Bank (https://www.scidb.cn/en/s/ZnUrue). The R-scripts used in this study can be viewed on the GitHub (http://github.com/JL000888/R-scripts/tree/main).

References

  1. Hassoun PM. Pulmonary arterial hypertension. N Engl J Med. 2021;385(25):2361–76.

    Article  CAS  PubMed  Google Scholar 

  2. Lau EMT, Giannoulatou E, Celermajer DS, Humbert M. Epidemiology and treatment of pulmonary arterial hypertension. Nat Rev Cardiol. 2017;14(10):603–14.

    Article  CAS  PubMed  Google Scholar 

  3. Farber HW, Miller DP, Poms AD, Badesch DB, Frost AE, Muros-Le Rouzic E, et al. Five-year outcomes of patients enrolled in the REVEAL registry. Chest. 2015;148(4):1043–54.

    Article  PubMed  Google Scholar 

  4. Ruopp NF, Cockrill BA. Diagnosis and treatment of pulmonary arterial hypertension: a review. JAMA. 2022;327(14):1379–91.

    Article  CAS  PubMed  Google Scholar 

  5. Humbert M, Sitbon O, Simonneau G. Treatment of pulmonary arterial hypertension. N Engl J Med. 2004;351(14):1425–36.

    Article  CAS  PubMed  Google Scholar 

  6. Wang R-R, Yuan T-Y, Wang J-M, Chen Y-C, Zhao J-L, Li M-T, et al. Immunity and inflammation in pulmonary arterial hypertension: from pathophysiology mechanisms to treatment perspective. Pharmacol Res. 2022;180: 106238.

    Article  CAS  PubMed  Google Scholar 

  7. Kuebler WM, Bonnet S, Tabuchi A. Inflammation and autoimmunity in pulmonary hypertension: is there a role for endothelial adhesion molecules? (2017 Grover Conference Series). Pulm Circ. 2018;8(2):2045893218757596.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Riou M, Enache I, Sauer F, Charles A-L, Geny B. Targeting mitochondrial metabolic dysfunction in pulmonary hypertension: toward new therapeutic approaches? Int J Mol Sci. 2023;24(11):9572.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Aggarwal S, Gross CM, Sharma S, Fineman JR, Black SM. Reactive oxygen species in pulmonary vascular remodeling. Compr Physiol. 2013;3(3):1011–34.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Archer SL, Gomberg-Maitland M, Maitland ML, Rich S, Garcia JG, Weir EK. Mitochondrial metabolism, redox signaling, and fusion: a mitochondria-ROS-HIF-1alpha-Kv1.5 O2-sensing pathway at the intersection of pulmonary hypertension and cancer. Am J Physiol Heart Circ Physiol. 2008;294(2):H570–8.

    Article  CAS  PubMed  Google Scholar 

  11. Dromparis P, Sutendra G, Michelakis ED. The role of mitochondria in pulmonary vascular remodeling. J Mol Med. 2010;88(10):1003–10.

    Article  CAS  PubMed  Google Scholar 

  12. Leek JT, Johnson WE, Parker HS, Jaffe AE, Storey JD. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics. 2012;28(6):882–3.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7): e47.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Glickman ME, Rao SR, Schultz MR. False discovery rate control is a recommended alternative to Bonferroni-type adjustments in health studies. J Clin Epidemiol. 2014;67(8):850–7.

    Article  PubMed  Google Scholar 

  15. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet. 2000;25(1):25–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Kanehisa M, Goto S, Sato Y, Furumichi M, Tanabe M. KEGG for integration and interpretation of large-scale molecular data sets. Nucleic Acids Res. 2012;40(Database issue):D109–14.

    Article  CAS  PubMed  Google Scholar 

  17. Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA. 2005;102(43):15545–50.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Hanzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinform. 2013;14:7.

    Article  Google Scholar 

  20. Friedman N, Geiger D, Goldszmidt M. Bayesian network classifiers. Mach Learn. 1997;29(2–3):131–63.

    Article  Google Scholar 

  21. Mood C. Logistic regression: why we cannot do what we think we can do, and what we can do about it. Eur Sociol Rev. 2010;26(1):67–82.

    Article  Google Scholar 

  22. Friedman JH. Greedy function approximation: a gradient boosting machine. Ann Stat. 2001;29(5):1189–232.

    Article  Google Scholar 

  23. Freund Y, Schapire R, Abe N. A short introduction to boosting. J Jpn Soc Artif Intell. 1999;14(5):771–80.

    Google Scholar 

  24. Chen TQ, Guestrin C. XGBoost: a scalable tree boosting system. In: 22nd ACM SIGKDD international conference on knowledge discovery and data mining (KDD), San Francisco, CA, 13–17 Aug 2016.

  25. Lundberg SM, Lee SI. A unified approach to interpreting model predictions. In: 31st annual conference on neural information processing systems (NIPS), Long Beach, CA, 4–9 Dec 2017.

  26. Barst RJ, McGoon M, Torbicki A, Sitbon O, Krowka MJ, Olschewski H, et al. Diagnosis and differential assessment of pulmonary arterial hypertension. J Am Coll Cardiol. 2004;43(12 Suppl S):40S–47S.

    Article  PubMed  Google Scholar 

  27. Thienemann F, Dzudie A, Mocumbi AO, Blauwet L, Sani MU, Karaye KM, et al. Rationale and design of the Pan African Pulmonary hypertension Cohort (PAPUCO) study: implementing a contemporary registry on pulmonary hypertension in Africa. BMJ Open. 2014;4(10): e005950.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Vereyken L, Dillen L, Vreeken RJ, Cuyckens F. High-resolution mass spectrometry quantification: impact of differences in data processing of centroid and continuum data. J Am Soc Mass Spectrom. 2019;30(2):203–12.

    Article  CAS  PubMed  Google Scholar 

  29. Pang Z, Chong J, Li S, Xia J. MetaboAnalystR 3.0: toward an optimized workflow for global metabolomics. Metabolites. 2020;10(5):186.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Kanehisa M, Furumichi M, Tanabe M, Sato Y, Morishima K. KEGG: new perspectives on genomes, pathways, diseases and drugs. Nucleic Acids Res. 2017;45(D1):D353–61.

    Article  CAS  PubMed  Google Scholar 

  31. Deng P, Haynes CM. Mitochondrial dysfunction in cancer: potential roles of ATF5 and the mitochondrial UPR. Semin Cancer Biol. 2017;47:43–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Pradeepkiran JA, Reddy PH. Defective mitophagy in Alzheimer’s disease. Ageing Res Rev. 2020;64: 101191.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Cloonan SM, Kim K, Esteves P, Trian T, Barnes PJ. Mitochondrial dysfunction in lung ageing and disease. Eur Respir Rev. 2020;29(157): 200165.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Chan DC. Mitochondrial dynamics and its involvement in disease. Annu Rev Pathol. 2020;15:235–59.

    Article  CAS  PubMed  Google Scholar 

  35. Marsboom G, Toth PT, Ryan JJ, Hong Z, Wu X, Fang Y-H, et al. Dynamin-related protein 1-mediated mitochondrial mitotic fission permits hyperproliferation of vascular smooth muscle cells and offers a novel therapeutic target in pulmonary hypertension. Circ Res. 2012;110(11):1484–97.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Rhodes CJ, Ghataorhe P, Wharton J, Rue-Albrecht KC, Hadinnapola C, Watson G, et al. Plasma metabolomics implicates modified transfer RNAs and altered bioenergetics in the outcomes of pulmonary arterial hypertension. Circulation. 2017;135(5):460–75.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Leopold JA, Maron BA. Molecular mechanisms of pulmonary vascular remodeling in pulmonary arterial hypertension. Int J Mol Sci. 2016;17(5):761.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Schermuly RT, Ghofrani HA, Wilkins MR, Grimminger F. Mechanisms of disease: pulmonary arterial hypertension. Nat Rev Cardiol. 2011;8(8):443–55.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Masson B, Le Ribeuz H, Sabourin J, Laubry L, Woodhouse E, Foster R, et al. Orai1 inhibitors as potential treatments for pulmonary arterial hypertension. Circ Res. 2022;131(9):e102–19.

    Article  CAS  PubMed  Google Scholar 

  40. Rizzuto R, De Stefani D, Raffaello A, Mammucari C. Mitochondria as sensors and regulators of calcium signalling. Nat Rev Mol Cell Biol. 2012;13(9):566–78.

    Article  CAS  PubMed  Google Scholar 

  41. Yuan C, Chen HX, Hou HT, Wang J, Yang Q, He GW. Protein biomarkers and risk scores in pulmonary arterial hypertension associated with ventricular septal defect: integration of multi-omics and validation. Am J Physiol Lung Cell Mol Physiol. 2020;319(5):L810–22.

    Article  CAS  PubMed  Google Scholar 

  42. Ould Amer Y, Hebert-Chatelain E. Mitochondrial cAMP-PKA signaling: what do we really know? Biochim Biophys Acta Bioenerg. 2018;1859(9):868–77.

    Article  CAS  PubMed  Google Scholar 

  43. Marshall JD, Bazan I, Zhang Y, Fares WH, Lee PJ. Mitochondrial dysfunction and pulmonary hypertension: cause, effect, or both. Am J Physiol Lung Cell Mol Physiol. 2018;314(5):L782–96.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Vercellino I, Sazanov LA. The assembly, regulation and function of the mitochondrial respiratory chain. Nat Rev Mol Cell Biol. 2022;23(2):141–61.

    Article  CAS  PubMed  Google Scholar 

  45. Sommer N, Huttemann M, Pak O, Scheibe S, Knoepp F, Sinkler C, et al. Mitochondrial complex IV subunit 4 isoform 2 is essential for acute pulmonary oxygen sensing. Circ Res. 2017;121(4):424–38.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Crnkovic S, Valzano F, Fliesser E, Gindlhuber J, Thekkekara Puthenparampil H, Basil M, et al. Single-cell transcriptomics reveals skewed cellular communication and phenotypic shift in pulmonary artery remodeling. JCI Insight. 2022;7(20): e153471.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Galkin A, Sitapara R, Clemons B, Garcia E, Kennedy M, Guimond D, et al. Inhaled seralutinib exhibits potent efficacy in models of pulmonary arterial hypertension. Eur Respir J. 2022;60(6):2102356.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Montani D, Perros F, Gambaryan N, Girerd B, Dorfmuller P, Price LC, et al. C-kit-positive cells accumulate in remodeled vessels of idiopathic pulmonary arterial hypertension. Am J Respir Crit Care Med. 2011;184(1):116–23.

    Article  PubMed  Google Scholar 

  49. Kim S, Rigatto K, Gazzana MB, Knorst MM, Richards EM, Pepine CJ, et al. Altered gut microbiome profile in patients with pulmonary arterial hypertension. Hypertension. 2020;75(4):1063–71.

    Article  CAS  PubMed  Google Scholar 

  50. Luo Q, Wang X, Liu R, Qiao H, Wang P, Jiang C, et al. alpha1A-adrenoceptor is involved in norepinephrine-induced proliferation of pulmonary artery smooth muscle cells via CaMKII signaling. J Cell Biochem. 2019;120(6):9345–55.

    Article  CAS  PubMed  Google Scholar 

  51. McArdle PF, Rutherford S, Mitchell BD, Damcott CM, Wang Y, Ramachandran V, et al. Nicotinic acetylcholine receptor subunit variants are associated with blood pressure; findings in the Old Order Amish and replication in the Framingham Heart Study. BMC Med Genet. 2008;9:67.

    Article  PubMed  PubMed Central  Google Scholar 

  52. Carthew RW. Gene regulation and cellular metabolism: an essential partnership. Trends Genet. 2021;37(4):389–400.

    Article  CAS  PubMed  Google Scholar 

  53. Chen W, Zhao H, Li Y. Mitochondrial dynamics in health and disease: mechanisms and potential targets. Signal Transduct Target Ther. 2023;8(1):333.

    Article  PubMed  PubMed Central  Google Scholar 

  54. Cai Z, Tu L, Guignabert C, Merkus D, Zhou Z. Purinergic dysfunction in pulmonary arterial hypertension. J Am Heart Assoc. 2020;9(18): e017404.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  55. Zhao Y, Peng J, Lu C, Hsin M, Mura M, Wu L, et al. Metabolomic heterogeneity of pulmonary arterial hypertension. PLoS ONE. 2014;9(2): e88727.

    Article  PubMed  PubMed Central  Google Scholar 

  56. Chen C, Luo F, Wu P, Huang Y, Das A, Chen S, et al. Metabolomics reveals metabolite changes of patients with pulmonary arterial hypertension in China. J Cell Mol Med. 2020;24(4):2484–96.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Martínez-Reyes I, Chandel NS. Mitochondrial TCA cycle metabolites control physiology and disease. Nat Commun. 2020;11(1):102.

    Article  PubMed  PubMed Central  Google Scholar 

  58. Oezguen N, Yılmaz V, Horvath TD, Akbayir E, Haidacher SJ, Hoch KM, et al. Serum 3-phenyllactic acid level is reduced in benign multiple sclerosis and is associated with effector B cell ratios. Mult Scler Relat Disord. 2022;68: 104239.

    Article  CAS  PubMed  Google Scholar 

  59. Rosa AP, Jacques CED, Moraes TB, Wannmacher CMD, Dutra AM, Dutra-Filho CS. Phenylpyruvic acid decreases glucose-6-phosphate dehydrogenase activity in rat brain. Cell Mol Neurobiol. 2012;32(7):1113–8.

    Article  CAS  PubMed  Google Scholar 

  60. Li C, Li Y, Sui L, Wang J, Li F. Phenyllactic acid promotes cell migration and invasion in cervical cancer via IKK/NF-κB-mediated MMP-9 activation. Cancer Cell Int. 2019;19:241.

    Article  PubMed  PubMed Central  Google Scholar 

  61. Li M, Yao B, Meng X. Inhibitory effect and possible mechanism of phenyllactic acid on Aspergillus flavus spore germination. J Basic Microbiol. 2022;62(12):1457–66.

    Article  CAS  PubMed  Google Scholar 

  62. Fan W, Li B, Du N, Hui T, Cao Y, Li X, et al. Energy metabolism as the target of 3-phenyllactic acid against Rhizopus oryzae. Int J Food Microbiol. 2022;369: 109606.

    Article  CAS  PubMed  Google Scholar 

Download references

Funding

This research was funded by the National Natural Science Foundation of China (82104291).

Author information

Author notes

  1. Jieling Li and Minyi Fu have contributed equally to this work.

Authors and Affiliations

  1. Department of Thoracic Surgery, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China

    Xin Zhang, Junjie Hu, Jianyong Zou & Bo Zeng

  2. Department of Pharmacy, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China

    Jieling Li, Ke-Jing Tang, Pan Chen & Xiaoman Liu

  3. Surgical and Anesthesia Center, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China

    Minyi Fu & Xijie Geng

  4. School of Pharmaceutical Sciences, Sun Yat-Sen University, Guangzhou, China

    Jieling Li

  5. Division of Pulmonary and Critical Care Medicine, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China

    Ke-Jing Tang

Authors
  1. Xin Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Jieling Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Minyi Fu
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Xijie Geng
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Junjie Hu
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Ke-Jing Tang
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Pan Chen
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Jianyong Zou
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Xiaoman Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Bo Zeng
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

BZ, XML and JYZ designed the study. XZ and JLL take responsibility for collecting patient samples, performing the analyses, and writing the manuscript. MYF and PC made contributions to preparing the figures. XJG, JJH, and KJT revised the manuscript. All authors reviewed the manuscript and approved the final version.

Corresponding authors

Correspondence to Jianyong Zou, Xiaoman Liu or Bo Zeng.

Ethics declarations

Ethics approval and consent to participate

In accordance with the Declaration of Helsinki, the research protocol for this study was approved by the Ethics Committee of the First Affiliated Hospital, Sun Yat-sen University (No. [2023] 553). All subjects provided informed consent for the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Table S1

The 36 DEGs in the GEO dataset analysis of 164 PAH and 58 control samples, with log2 (fold change), P value and adjusted P value.

Additional file 2: Table S2

The 361 DEGs in the transcriptomic analysis of 8 PAH and 8 control samples, with log2 (fold change), P value and adjusted P value.

Additional file 3: Table S3

The 630 mitochondria-related genes summarized through the mitochondria-related pathways retrieved from the MSigDB.

Additional file 4: Table S4

The 63 metabolites detected by metabolomic profiling, with VIP, log2 (fold change), and P value.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhang, X., Li, J., Fu, M. et al. Dysfunction in mitochondrial electron transport chain drives the pathogenesis of pulmonary arterial hypertension: insights from a multi-omics investigation. Respir Res 26, 29 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12931-025-03099-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12931-025-03099-8

Keywords

Respiratory Research

ISSN: 1465-993X

Contact us