Comprehensive profiling of lactylation-associated genes in pulmonary hypertension through bulk and single-cell RNA sequencing integration

Abstract

Background

Pulmonary hypertension (PH) is a progressive disorder with high pulmonary arterial pressure, causing right ventricular dysfunction. Understanding PH’s molecular basis is vital for therapy development, but research on lactylation’s role in PH is limited and needs more investigation.

Methods

We integrated bulk RNA-seq datasets (GSE113439, GSE186996) and single-cell RNA-seq data (GSE210248) from PH patients and controls Differential expression analysis identified lactylation-associated hub genes. Functional enrichment (GO/KEGG), immune infiltration, and cell-cell communication analyses were performed. In vitro validation included RT-PCR and western blot on hypoxic pulmonary artery smooth muscle cells (PASMCs).

Results

We identified lymphocyte cytosolic protein 1 (LCP1) as a core lactylation-related hub gene, significantly upregulated in PH and validated in human/rat models. Single-cell profiling revealed elevated lactylation levels in vascular smooth muscle cells (VSMCs), T lymphocytes, and monocytes/macrophages within PH tissues. Cell communication analysis implicated TWEAK signaling from monocytes/macrophages to VSMCs as a driver of vascular remodeling. Metabolic pathways (e.g., glycolysis, bile acid metabolism) correlated strongly with lactylation activity. In vitro, hypoxia-induced LCP1 overexpression in PASMCs confirmed bioinformatic findings.

Conclusion

We analyzed lactylation-related genes in PH, identifying LCP1 as a key diagnostic marker, and described immune cell lactylation in PH patients, offering insights for future research.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12931-025-03446-9.

Introduction

Pulmonary hypertension (PH) is a serious global health issue marked by changes in the pulmonary vasculature, increasing vascular resistance and arterial pressure, potentially leading to right heart failure and death [1]. Despite advancements in diagnosis and treatment, PH continues to exhibit high prevalence and mortality rates. Based on pathological mechanisms and clinical manifestations, PH can be classified into five major groups, all sharing the common hemodynamic characteristic of a resting mean pulmonary artery pressure ≥ 20 mmHg [1, 2]. Characteristic manifestations of PH include respiratory dyspnea, cardiovascular chest pain and syncope, and circulatory signs like lower extremity edema and jugular venous distension [3]. Epidemiological studies indicate that this condition affects approximately 1% of the global adult population, with marked geographical disparities. According to an analysis of 17 studies, global 1-year, 3-year, and 5-year survival rates range from 85 to 99%, 65 to 95%, and 50 to 86%, respectively [4]. Current epidemiological studies indicate that PH affects approximately 25 individuals per million population, with an annual incidence of 5 cases per million, highlighting its rarity [5]. Therefore, it is of great clinical significance to explore the pathogenesis of PH and find new therapeutic targets.

Over the past decades, extensive research has demonstrated that immune-inflammatory factors play a critical role in the pathogenesis of PH and the related vascular remodeling. Specifically, the irreversible pulmonary vascular remodeling observed in PH leads to progressive luminal narrowing of blood vessels, a phenomenon driven by multifactorial pathogenic mechanisms [6]. Key contributors include aberrant proliferation of pulmonary arterial smooth muscle cells, endothelial cell dysfunction, and chronic inflammatory-immune dysregulation [1, 7–9]. Additionally, emerging evidence indicates that non-coding RNAs play a key role in the pathogenesis and progression of PH and may offer potential therapeutic targets [10–12]. At the same time, it has been reported that dysregulation of lactic acid homeostasis and lactic modification play an important role in the development of PH [13].

Lactate, traditionally recognized as a metabolic byproduct of uncoupled glycolysis, has emerged as a critical epigenetic modulator through its capacity to drive histone lysine lactylation. Current research has extensively characterized histone lactylation, with groundbreaking studies revealing its pathogenic significance in myocardial infarction [14], pulmonary fibrosis [15] and carcinoma [16], positioning it as a master regulator of disease pathogenesis. In recent years, lactylation-associated genes have emerged as potential biomarkers and therapeutic targets in various diseases [17, 18]. However, current research has not thoroughly investigated the expression profiles of lactate metabolism-related genes and their functional regulatory mechanisms in PH, while studies focusing on the expression characteristics and biological functions of key regulatory genes in this process remain relatively scarce.

In this research, a comprehensive approach is employed. To further elucidate the role of lactylation-related genes in PH, this study identified lymphocyte cytosolic protein 1 (LCP-1) as a differentially expressed gene through literature review and GEO dataset analysis. Subsequent single-cell transcriptomic profiling and immune infiltration analysis confirmed the critical function of LCP1 within the PH immune microenvironment, which was further validated through in vitro experiments. Collectively, this study provides a comprehensive cellular and molecular characterization of lactylation-associated mechanisms in PH pathogenesis, offering actionable molecular targets for therapeutic development and pharmacological research.

Methods

Identification of differentially expressed genes (DEGs)

A total of 331 lactylation-associated genes were curated through systematic literature mining [19–21]. These candidate genes were subsequently cross-referenced with the PH-related transcriptomic dataset GSE113439 and GSE186966 from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/).Use the “GEOquery” software package to get the data and extract the expression matrix and sample information. Then the “limma” software package was used for difference analysis to screen the significant difference genes between PH patients and healthy controls. Intersection analysis was performed using adjusted p-value < 0.05 and |log2(fold change)| >0.263 (FC>1.2) as thresholds, ultimately identifying lactylation-related hub genes of PH.

Enrichment analyses

To investigate the functional implications of DEGs, we conducted comprehensive enrichment analyses utilizing the “clusterProfiler” package in R [22]. This included systematic evaluation of Gene Ontology (GO) categories and pathway enrichment through the Kyoto Encyclopedia of Genes and Genomes (KEGG) database [23]. The enrichment analysis results were visualized using the “ggplot2” package. A p-value < 0.05 was considered statistically significant.

GSVA

Gene Set Variation Analysis (GSVA, version 2.0.7) was employed to assess the enrichment of genes associated with differential expression of crotonylation. Following this, Single-sample Gene Set Enrichment Analysis (ssGSEA) was utilized to derive enrichment scores. The resulting score profiles were subsequently compared between samples from patients with PH and control groups.

Immune cell analyses

To comprehensively characterize the immune microenvironment in PH patients, we employed CIBERSORT - a well-validated deconvolution algorithm that quantifies relative proportions of 22 immune cell subsets based on their characteristic gene expression patterns [24]. Comparative analyses of immune cell distributions between PH patients and healthy controls were visualized using box plots and heatmap representations. Furthermore, we conducted systematic correlation studies to examine both intercellular relationships among immune populations and potential associations between DEG expression patterns and immune infiltration levels. These interactions were graphically represented to illustrate pairwise correlations between individual DEGs and specific immune cell types.

Single-cell RNA sequencing (scRNA-seq) analysis

Single-cell transcriptome data from PH patients and healthy controls were acquired from the GEO dataset GSE210248. Initial processing utilized Seurat (v5.2.1) [22]. To ensure data quality, cells were excluded if they expressed fewer than 300 genes, more than 3000 genes, had mitochondrial gene content exceeding 5%, or possessed total RNA counts either below 1000 or above the 97th percentile. Potential batch effects were mitigated using the R package ’harmony’. Following dimensionality reduction, principal components were analyzed (PCA) and subsequently used as input for non-linear embedding and cell clustering, implemented using the Uniform Manifold Approximation and Projection (UMAP) algorithm. Finally, cell type identities were assigned through reference-based annotation.

Cell-cell communication analysis

The CellChat package (v2.1.0) was employed to infer, analyze, and visualize intercellular communication networks. The workflow was executed through sequential computational phases: initial data preprocessing was performed via “identifyOverExpressedGenes”, “identifyOverExpressedInteractions”, and “projectData” to establish ligand-receptor mappings. Subsequently, the core analysis employed “computeCommunProb” to quantify interaction probabilities and “computeCommunProbPathway” to integrate signaling pathways. Network construction followed, involving “aggregateNet” to generate global networks and “netAnalysis_computeCentrality” to compute topological metrics. Finally, targeted interrogation was conducted through “subsetCommunication” to extract context-specific interactions, coupled with rigorous filtering via “filterCommunication”. We used the “GSVA” package to evaluate the lactylation enrichment scores of each cell in the scRNA-seq dataset and score each cell based on the HALLMARK pathway. Then, we analyzed the correlation between the pathway scores and the lactylation scores.

Cell culture

The human pulmonary artery smooth muscle (hPASMCs) was commercially obtained from the American Type Culture Collection (ATCC, Virginia, USA). These cells were maintained in smooth muscle cell-specific culture medium (SMCM, ScienCell, California, USA) supplemented with 10% fetal bovine serum, 1% smooth muscle cell growth supplement, and 1% penicillin-streptomycin antibiotic solution. The rat pulmonary artery smooth muscle cells (rPASMCs) were purchased from Aoruicell (Aoruicell, Shanghai, China) and were cultured in Dulbecco’s modified Eagle medium (DMEM) (Gibco, New York, USA). The culture medium was supplemented with 10% FBS and 1% penicillin/streptomycin. Cell cultures were incubated at 37 °C in a humidified atmosphere containing 21% oxygen and 5% carbon dioxide. For hypoxic exposure, cells were subjected to a 24-hour treatment in a specialized chamber maintaining a controlled atmosphere of 1% O₂, 5% CO₂, and balanced N₂ (94%). For drug-induced PAH modeling, Platelet-Derived Growth Factor-BB (PDGF-BB) (MCE, New Jersey, USA) was administered to hPASMCs at a concentration of 20 ng/mL.

Cell transfection

In the transfection process, lipofectamine 2000 (Thermo Fisher, CA, USA) was utilized. The overexpression of the target gene was achieved using plasmids that express LCP1 (Syngen-tech, Beijing, China).

Lactate assay

We used the lactate assay kit (Beyotime, Shanghai, China) to measure the lactate concentration in the samples. To detect the intracellular lactate concentration, the cultured adherent cells were lysed by adding lysis buffer according to the instructions, gently shaken, and then incubated on ice for 5–10 min to ensure complete lysis. The lysate was then centrifuged at 4 °C at 12,000×g for 3–5 min, and the supernatant was collected for subsequent analysis.

Cell Counting Kit-8 (CCK-8) assay

Cells were plated in 96-well plates at a density of 5 × 10³ cells per well and subsequently treated with the CCK-8 working solution (Servicebio, Wuhan, China) for a duration of 2 h at 37 °C. The absorbance was then quantified at a wavelength of 450 nm utilizing a microplate reader (Bio Tek, VT, USA).

5-ethynyl-2’ -deoxyuridine (EdU) assay

Proliferative cells were marked using the BeyoClick™ EdU Cell Proliferation Kit with AF555 (Beyotime, Shanghai, China), following the guidelines provided by the manufacturer. These cells were subsequently analyzed using the Celldiscoverer 7 (Carl Zeiss, Jena, Germany). The rate of EdU incorporation was quantified as the proportion of EdU-positive cells relative to the total number of DAPI-positive cells.

Apoptosis detection

We conducted the detection using the Annexin V-FITC apoptosis detection kit (Beyotime, Shanghai, China). The cells were inoculated into 6-well plates, and allowed to adhere to the surface. After cell attachment, each group received different intervention treatments. After the intervention treatments, the cells were digested with trypsin and resuspended; then, they were counted. A total of 10 × 10⁴ cells were collected from each group and washed twice with pre-cooled PBS. Then, the cells were centrifuged, and the supernatant was discarded. The cells were resuspended in binding buffer, and 5 µL of Annexin V-FITC, and 10 µL of PI solution were added. The cell suspension mixture was incubated at room temperature in the dark for 10 min. Then the cells were analyzed for apoptosis using a flow cytometer (Beckman, California, USA).

Reverse Transcription-Polymerase Chain Reaction (RT-PCR)

Total RNA was extracted (Genuine, Henan, China) from pulmonary artery smooth muscle cells (PASMCs) and then reverse-transcribed into cDNA (Genuine, Henan, China). The relative expression of LCP1 mRNA in each sample were determined by real-time PCR using SYBR™ green master mix (Thermo Fisher Scientific) in a PRISM^® 7500 real-time PCR machine (ABI) according to the assay kit and instrument manufacturers’ protocols [9, 10].The amplification protocol consisted of 40 cycles performed on a CFX96 PCR instrument (Bio-Rad, Hercules, CA, USA) with the following thermal profile: initial denaturation at 95°C for 60 seconds, followed by cyclic denaturation at 95°C for 20 seconds, primer annealing at 55°C for 20 seconds, and elongation at 72°C for 30 seconds. Oligonucleotide primers were computationally designed using the NCBI Primer-BLAST online tool to ensure specificity. The primer pairs used for human LCP1 amplification were as follows: LCP1-Forward: 5’-TTGCTTGCCTTTGCCTGGGTATAG-3’ and LCP1-Reverse: 5’-GCCAACGCTAGACTGCTCTGAAG-3’. Human GAPDH was amplified with the following primers: GAPDH-Forward: 5’-AGGTCGGTGTGAACGGATTTG-3’ and GAPDH-Reverse: 5’-TGTAGACCATGTAGTTGAGGTCA-3’.

Western blot

Protein extraction was performed by homogenizing cells and tissues in RIPA lysis buffer (Servicebio, Wuhan, China) containing protease inhibitor cocktail, followed by 30-minute incubation on ice and subsequent centrifugation to collect the soluble fraction. For protein separation, samples were resolved through 10% SDS-polyacrylamide gel electrophoresis (Servicebio, Wuhan, China) and then electrotransferred onto polyvinylidene difluoride membranes. The membranes were first blocked with 5% non-fat dry milk solution for one hour at room temperature before overnight incubation with primary antibodies at 4 °C. Immunodetection was completed by incubating the membranes with HRP-conjugated secondary antibodies (Servicebio, Wuhan, China) for one hour at ambient temperature, enabling subsequent chemiluminescent visualization of target proteins. Protein expression levels were quantitatively assessed through immunoblotting techniques employing a Tanon gel documentation system (Tanon, Shanghai, China). The following specific primary antibodies were utilized for target protein detection: LCP1 at 1:1000 (proteintech, Wuhan, China) and GAPDH at 1:5000 (Bioworld, MN, USA).

Statistical analysis

Statistical analyses were performed utilizing R statistical software (version 4.4.2). For comparisons involving multiple independent groups, the non-parametric Kruskal-Wallis test was employed. When analyzing differences between two independent groups, either Student’s t-test or the Wilcoxon rank-sum test was applied, depending on data distribution characteristics. The threshold for statistical significance was established at p < 0.05 for all hypothesis testing procedures.

Results

Identification of DEGs in PH

The study’s flow diagram is illustrated in Fig. 1. Differential expression analysis of human PH samples (GSE113439) revealed significant transcriptomic alterations (Fig. 2A), with subsequent heatmap visualization demonstrating clear segregation between PH and control cohorts alongside distinct co-expression modules (Fig. 2B). Complementing these findings, rat PH data (GSE186996) exhibited conserved dysregulation patterns characterized by predominant gene upregulation (Fig. 2C), while analogous heatmap clustering identified functionally congruent expression modules (Fig. 2D). Cross-species integration of DEGs identified 52 consistently upregulated and 9 downregulated targets (Fig. 2E-F). GO enrichment analysis demonstrated significant enrichment in cell stress response, muscle cell biology, and oxidative metabolism related pathways. Among them, muscle cell proliferation, smooth muscle cell proliferation, and reactive oxygen species metabolic process showed lower p values, suggesting they may play key roles (Fig. 2G). KEGG enrichment analysis further indicated predominant participation in cytokine-cytokine receptor interaction, TGF-βsignaling pathway, and phagosome pathways (Fig. 2H).

Fig. 1.

Flowchart illustrates the workflow of this study

Fig. 2.

Identification of Differentially Expressed Genes in Pulmonary Hypertension (PH). A Volcano plot of DEGs in human PH (GSE113439). B Heatmap of gene expression profiles in human PH (GSE113439). C Volcano plot of DEGs in rat PH (GSE186996) D. Heatmap of gene expression profiles in rat PH (GSE186996). E Cross-species analysis of upregulated DEGs in PH. F Cross-species analysis of downregulated DEGs in PH. G GO enrichment analysis of DEGs in PH. H KEGG pathway enrichment analysis of DEGs in PH

LCP1 drives PH pathogenesis as a key lactylation target

After systematically curating lactylation-related genes from the literature, gene set enrichment analysis (GSEA) was performed on the lactylation gene set in PH samples. The results showed a global enrichment trend of the lactylation gene set in PH (Fig. 3A), which suggests that the lactylation regulatory network is active during PH progression. Lactylation signaling intensity was quantified in the human cohort using ssGSEA. It was observed that the lactylation modification level in the PH group was significantly increased (Fig. 3B). Furthermore, by intersecting known lactylation-related genes with DEGs in PH, overlapping gene LCP1was obtained (Fig. 3C). Through validation analyses in human and rat GEO datasets, it was found that LCP1 expression in the PH group was significantly higher than that in the control group (Fig. 3D). To predict the potential functions of LCP1, we performed single-gene GSEA analysis. Based on the correlation between LCP1 and other genes, we obtained gene sets that are highly correlated with LCP1. Enrichment results for GO analysis revealed that LCP1 exerts its functions mainly through immune related pathways (Fig. 3E), while KEGG pathways pointed to pathways such as DNA replication licensing and spliceosome, implying that LCP1 may be involved in PH pathology by promoting DNA replication in vascular cells to accelerate their proliferation and regulating RNA splicing to reshape gene expression. The GO functional enrichment analysis of LCP1 reveals its core functions in PH from four dimensions: immune activation, extracellular matrix remodeling, cytoskeleton regulation, and regeneration and stress (Fig. 3F). Next, this study analyzed the immune cell infiltration pattern associated with LCP1. The heatmap robustly validated the existence of significant differential immune signatures between PH samples and control samples (Fig. 3G). To further quantify the association between LCP1 expression and immune cell infiltration, Pearson correlation analysis was performed based on immune cell infiltration scores and LCP1 expression levels, with results visualized via a bubble plot (Fig. 3H).

Fig. 3.

LCP1 Drives PH Pathogenesis as a Key Lactylation Target. A GSEA enrichment analysis of lactylation-related genes. B Lactylation signaling ssGSEA score: PH vs Control. C The intersection of PH differential genes and lactylation-related genes. D Verification of LCP1 expression in human and rat PH. E Single-gene GSEA analysis. F GO functional enrichment analysis of LCP1. G Heat maps of differences in 22 immunoinfiltrating cell types. Blue represents low-expression immune cells and red represents high-expression immune cells. H Bubble plot showing the correlation between LCP1 expression and immune cell infiltration

Finally, cross-species homology analysis was performed on the nucleic acid and protein sequences of LCP1 in humans, rats, and mice. The homology heatmap showed high similarity among the three species at both the nucleic acid and protein levels (Figure S1), providing a theoretical basis for the cross-species translation of LCP1 function research.

To validate our findings, we initially subjected hPASMCs to either normoxic or hypoxic conditions, designating the hypoxia-treated group as the HPH group. Separately, cells cultured under normoxia were stimulated with PDGF-BB to model PAH. Intracellular lactate levels were measured using a commercial assay kit, revealing a significant elevation in both the HPH (Fig. 4A) and PAH (Fig. 4B) groups compared with their respective controls. Subsequently, we assessed lactylation levels via WB in hPASMCs exposed to normoxia versus hypoxia. The results demonstrated that the HPH group exhibited significantly increased overall lysine lactylation as well as specific histone lactylation marks at H2BK16, H3K18, H4K5 and H4K12 (Fig. 4C). To further verify LCP1 expression in PASMCs, total RNA was isolated from hPASMCs grown under normoxia or hypoxia and analyzed (Fig. 4D). Similarly, under normoxic conditions, hPASMCs were either left untreated or treated with PDGF-BB, followed by total RNA extraction and RT-PCR (Fig. 4E). The results confirmed that LCP1 mRNA expression was markedly upregulated in hypoxia-exposed hPASMCs, corroborating our bioinformatics predictions. Given the high evolutionary conservation of LCP1 protein sequences between humans and rats, total protein was extracted from rPASMCs cultured under normoxia or hypoxia for Western blot analysis (Fig. 4F). Likewise, total protein was harvested from rPASMCs treated with PDGF-BB under normoxia (Fig. 4G). Consistently, LCP1 protein levels were significantly elevated in both the HPH and PAH groups. Next, we overexpressed LCP1 in hPASMCs under hypoxia and confirmed successful overexpression via Western blot (Supplementary Figure S2A–B). We next examined the proliferative influence of heightened LCP1 expression using CCK-8 (Supplementary Figure S2C) and EdU assays (Fig. 4H, 6D). Both CCK-8 and EdU assays revealed that cell proliferation was significantly enhanced in the hypoxia + oe-LCP1 group compared with the normoxia, hypoxia, and hypoxia + oe-NC groups. Flow cytometry analysis further showed that the apoptosis rate was markedly decreased in the hypoxia + oe-LCP1 group when compared with the normoxia, hypoxia alone, and hypoxia + oe-NC groups (Fig. 4I).

Fig. 4.

Single-cell RNA sequence analysis. A Lactate content in hPASMCs under hypoxic conditions. B Lactate content in hPASMCs under PDGF-BB stimulation conditions. C Hypoxia induces global and site-specific histone lactylation in hPASMCs. D Verification of mRNA expression of LCP1 in HPH. E Verification of mRNA expression of LCP1 in PAH F. Verification of protein expression of LCP1 in HPH. G Verification of protein expression of LCP1 in PAH. H Cell proliferation assessed by EdU assay. I Apoptpsis assessed by flow cytometry. *p < 0.05 vs normoxia; ** p < 0.01 vs normoxia; # p < 0.05 vs control; ## p < 0.01 vs control

Fig. 6.

Cell communication. A Cell - cell communication among the specified cell types. B Illustration of the interaction strengths for both incoming and outgoing signals among various cell types. C The incoming and outgoing signaling pathways of each cell type. D The cell-cell interaction network regarding the TWEAK pathway among different cell types. E The cell-cell interaction network related to the TWEAK signaling pathway among different cell types. F TWEAK mediated crosstalk between monocytes/macrophages (TNFSF12) and VSMCs (TNFRSF12A)

ScRNA-seq analysis

To further explore the cell specific regulatory mechanisms of lactylation in PH, we conducted single-cell transcriptome analysis. After completing data quality control (Figure S3), we first used an unsupervised clustering method based on the UMAP dimensionality reduction technique to partition cells into 16 transcriptionally heterogeneous clusters (Fig. 5A). Comparative analysis showed that there were significant differences in cell composition between PH and the control group (Fig. 5B). Through manual annotation, we annotated these 16 cell clusters into 10 cell types, including T cells, NK cells, B cells, cDC2 cells, monocytes/macrophages, mast cells, epithelial cells, endothelial cells, vascular smooth muscle cells (VSMCs), and Fibroblasts (Fig. 5C and D). After evaluating the lactylation level of single cells by detecting the expression of lactylation related genes (LRGs) in different cell types (Fig. 5E), we calculated the enrichment score of the LRGs for each cell and displayed it in the UMAP space with a color gradient. The result showed that the clusters where T lymphocytes, VSMCs, and monocytes/macrophages were located were densely distributed in dark blue, suggesting that these cells are hotspots with high lactylation activity. Subsequently, cells were classified into high and low LRG score groups based on lactylation levels. The differences in LRG scores among different cell types were analyzed together with the score difference (Fig. 5F). Focusing on the verification of VSMCs, it was found that the LRG score in the PH group was significantly higher than that in the control group (Fig. 5G), which directly confirmed that the lactylation activity of VSMCs was specifically enhanced in PH, and this is the core cellular level evidence driving pathological processes such as vascular remodeling in PH.

Fig. 5.

Single-cell RNA sequence analysis. A UMAP classifies cells into 16 distinct clusters. B Differentiation of the control group and PH cell types based on UMAP. C UMAP - based spatial distribution of major cell types in PH samples. D Expression of genes within the lactylation gene set across cells. E The distribution of the scores of the lactylation-related gene set in the single-cell UMAP. F Number and percentage of cells of each cell type between the high and low expression groups of LRGs score. G Comparison of LRG scores of vascular smooth muscle cells between the control group and the PH group

Quality control and validation

During the preprocessing and quality control phase of scRNA-seq data analysis, rigorous quality assessment procedures are implemented to ensure data integrity (Figure S3A). The initial cell selection process involves evaluating multiple key parameters:

1. Gene detection threshold (rFeature_RNA): Cells expressing fewer than 200 distinct genes are excluded from subsequent analysis, as this may indicate compromised cellular integrity due to technical artifacts during sample preparation or sequencing.

2. Transcript abundance (sCount_RNA): Cells demonstrating either extremely low or high unique molecular identifier (UMI) counts are removed. This filtering step eliminates potentially inactive cells (low UMI counts) and potential multiplets (high UMI counts) that could distort downstream analyses.

3. Mitochondrial content (mt_percent): Cells exhibiting mitochondrial gene expression exceeding 15% of total transcripts are discarded, as elevated mitochondrial RNA levels often serve as biomarkers for cellular stress responses or apoptotic processes, including those induced by ischemic conditions.

PCA was performed to visualize the data distribution before and after batch effect correction (Figure S3B). The initial PCA plot displayed significant batch-related clustering, which was subsequently eliminated following harmonization using the Harmony algorithm. This comparative analysis demonstrated the effectiveness of batch effect removal in reducing technical variation while preserving biological signals. The post-correction PCA revealed improved sample clustering patterns that better reflected the underlying biological relationships. Furthermore, the assessment of cell cycle progression scores through correlation analysis demonstrated satisfactory data integrity (Figure S3C).

Inference of cell–cell communication in PH

To investigate cell-cell interactions in PH, we performed cell communication analysis using the CellChat tool. Interestingly, compared to the control group, the PH group exhibited higher signaling pathway activity (Figure S4A-S4B). We then inferred the number and strength of interactions between different cell populations in PH samples (Fig. 6A-B). Figure 6C details the inferred incoming and outgoing interaction strengths, identifying T cells, monocytes, and macrophages as the major cell types expressing ligands and receptors involved in these interactions. To identify signaling pathways contributing to this complex communication, we calculated the outgoing and incoming interaction strength for each pathway (Fig. 6D). Notably, TWEAK signaling emerged as the most predominant incoming signal in VSMCs, while monocytes and macrophages showed strong outgoing TWEAK signaling (Fig. 6E). Within the TWEAK pathway, TNFSF12 (TWEAK) was enriched in monocytes and macrophages, and TNFRSF12A (TWEAK receptor) showed higher expression levels in VSMCs (Fig. 6F). In summary, these findings indicate that monocytes/macrophages likely activate VSMCs via TWEAK signaling to drive vascular remodeling, representing the molecular basis of the observed VSMCs-centered microenvironment in PH vascular tissues.

Functional clustering analysis reveals critical pathways associated with lactylation

To deeply dissect the core signaling pathways associated with lactylation modification, this study utilized the GSVA algorithm to quantify pathway activity at the single - cell level. By integrating with the MSigDB database, we systematically analyzed the correlation between pathway scores and lactylation levels. Through the analysis of the HALLMARK pathway collection, it was found that the “SPERMATOGENESIS”, “BILE ACID METABOLISM”, “ALLOGRAFT REJECTION” and “HEDGEHOG SIGNALING” pathways showed a strong correlation with lactylation levels in VSMCs. The pathways “SPERMATOGENESIS”, “E2F TARGETS”, “MYC TARGETS V1”, “G2M CHECKPOINT”, and “HEDGEHOG_SIGNALING” showed strong correlations with lactylation levels in T cells. And the pathways “SPERMATOGENESIS”, “APICAL JUNCTION”, “MYC TARGETS V1”, “G2M CHECKPOINT”, and “GLYCOLYSIS” showed strong correlations with lactylation levels in monocytes and macrophages (Fig. 7).

Fig. 7.

Heatmap of correlation between LRGs scores and the HALLMARK pathway in each cell type

Discussion

PH is a complex and progressive disease that poses a significant threat to human health [1]. Currently, pulmonary vascular remodeling is recognized as a crucial pathological basis for PH. The abnormal proliferation of PASMCs is not only one of the key causes of pulmonary vascular remodeling but also one of the critical pathological bases for irreversible pulmonary hypertension [25]. Despite progress in diagnostic techniques and therapeutic interventions, pulmonary hypertension remains characterized by elevated prevalence and significant mortality rates. The current study aims to explore the underlying molecular mechanisms of PH. This study systematically investigated the role of lactylation-associated genes in the pathogenesis of PH by integrating multi-omics analytical approaches, including bulk RNA sequencing (GSE113439, GSE186996), single-cell transcriptomic analysis (GSE210248), and cell-cell communication network construction. While it is true that the datasets used in this study have been previously analyzed by others, our work is distinguished by its unique lactylation-centric perspective and integrative methodology. Prior studies predominantly studies predominantly focused on traditional mechanisms such as immune genes [26, 27], autophagy [28], or specific non-coding RNAs [29], whereas our study is the first to systematically explore lactylation-associated gene networks in PH across both bulk and single-cell resolutions.

To address potential batch effects across the bulk RNA-seq datasets (GSE113439, GSE186996) and the single-cell RNA-seq data (GSE210248), we applied standardized preprocessing pipelines tailored to each data type. Specifically, for the bulk RNA-seq datasets, we performed TPM normalization (GSE113439 and GSE186996), while for the microarray data, we applied log2 transformation and quantile normalization. For the single-cell data (GSE210248), we employed the Harmony algorithm to effectively correct for technical variations while preserving biologically meaningful signals. This correction was demonstrated by UMAP and PCA before and after integration (Supplementary Revised Figure S3), showing a clear reduction of batch effect-induced clustering without obscuring disease-specific transcriptional patterns. All normalized expression matrices were then pooled for downstream comparative and integrative analyses, thereby ensuring robust and reproducible findings across platforms and cohorts.

Analysis of dual datasets from humans and rats demonstrates cross-species conservation of dysregulated gene expression in PH. GO/KEGG enrichment analyses decipher the functional roles and pathway associations of these genes, revealing that DEGs primarily participate in hypoxic adaptation, vascular cell proliferation, and related processes. This aligns with PH’s pathological mechanism wherein pulmonary vascular hypoxia induces excessive PASMC proliferation, subsequently leading to luminal narrowing [30]. KEGG enrichment analysis demonstrated that differentially expressed genes were enriched in pathways such as metabolic reprogramming, immune-inflammatory responses, pro-fibrotic signaling, and cell survival regulation, thereby elucidating PH’s core pathological triad—vascular remodeling, metabolic dysregulation, and inflammation-driven pathology—at the molecular pathway level [31].

Our integrated analysis revealed that lactylation serves as a crucial link connecting metabolic reprogramming, epigenetic modulation, and phenotypic transition. Specifically, by integrating lactylation-related genes from the literature and conducting GSEA on the lactylation gene set in PH samples, we demonstrated widespread enrichment of lactylation-modified genes in PH. Measurement of intracellular lactate levels showed that hypoxia and PDGF-BB stimulation significantly increased lactate accumulation. Additionally, hypoxia induced widespread histone lactylation at sites including H2BK16, H3K18, H4K5, and H4K12. These findings suggest an association between glycolytic reprogramming and histone lactylation in the pathogenesis of PH. Building on this foundation, we identified LCP1 as a core hub gene associated with lactylation, whose upregulation was rigorously validated across human and rodent. Homology assessments further confirmed substantial sequence conservation of LCP1 across species, supporting its potential for translational research. Functionally, LCP1 overexpression substantially enhanced PASMC proliferative capacity while concurrently suppressing apoptosis, both of which are central drivers of obstructive vascular remodeling. This chain of evidence is interlinked, not only proving that LCP1 is a core target of lactylation regulation in PH but also providing support for cross-species mechanistic research through sequence conservation. Combined with single-gene GSEA and GO functional enrichment analysis, LCP1 was shown to have functional dimensions such as immune activation, extracellular matrix remodeling, and cytoskeleton regulation, which highly align with the known pathogenesis of PH. One study demonstrated that NEDD9 is involved in the pathogenesis of PH by regulating extracellular matrix remodeling [32]. Additionally, another study found that the monocrotaline-induced PH model better reflects the increase in conventional dendritic cells and interstitial macrophages in PH compared to the Sugen-5416/hypoxia model [33]. Immune infiltration profiling provides further support that LCP1 drives PH pathogenesis by modulating immune-inflammatory responses.

ScRNA-seq analysis has revealed significant differences in cellular composition between PH patients and healthy controls, particularly highlighting the abnormal enrichment of T lymphocytes, VSMCs and monocytes/macrophages. These findings suggest that the accumulation and interaction of immune cells may exacerbate the vascular remodeling process characteristic of PH. The role of T lymphocytes in modulating immune responses and the involvement of macrophages in tissue remodeling are critical in understanding the pathophysiology of PH. The abnormal immune cell profiles observed in PH patients may contribute to a pro-inflammatory environment, leading to further vascular damage and disease progression [34].

ScRNA-seq analysis revealed remarkable differences in cellular composition between PH patients and healthy controls, with a particular emphasis on the abnormal enrichment of T lymphocytes, VSMCs, and monocytes/macrophages. This indicates that these cells are core participants in the pathological microenvironment of PH, and the accumulation and interaction of immune cells may exacerbate the vascular remodeling process characteristic of PH. To quantify lactylation activity at the cellular level, we detected the expression characteristics of LRGs in different cell types. Integrating the inter-group comparison of VSMCs directly confirms that VSMCs are core effector cells with specifically enhanced lactylation modification in PH, and their abnormal lactylation activity is the key cellular-level evidence driving pathological processes such as pulmonary vascular remodeling. In the future, it is necessary to deeply explore the dynamic changes of lactylation-related enzymes and substrates to clarify their upstream regulatory mechanisms.

Analysis of cellular communication revealed significant activation of the TWEAK signaling axis from monocytes/macrophages to VSMCs in PH. The ligand TNFSF12 (TWEAK) is highly expressed in monocytes/macrophages, while receptor TNFRSF12A shows significantly elevated expression in VSMCs. This pathway directly mediates immune cell activation of vascular cells, providing a molecular basis for understanding the perivascular inflammatory microenvironment in PH.

TWEAK signaling has known roles in tissue repair and fibrosis. This study is the first to position it as a key driver pathway in PH vascular remodeling, offering new directions for targeted interventions (e.g., blocking the TWEAK receptor). The role of T lymphocytes in regulating immune responses and macrophage involvement in tissue remodeling are crucial for understanding PH pathophysiology. The abnormal immune cell profiles observed in PH patients may contribute to a pro-inflammatory environment, leading to further vascular damage and disease progression. Therefore, the TWEAK signaling pathway may represent a central mediator of immune-vascular crosstalk in PH [34].

Furthermore, analysis of the HALLMARK pathway collection identified key pathways significantly correlated with lactylation levels, such as bile acid metabolism and glycolysis. Of particular note, glycolysis—a central pathway in energy metabolism—exhibited a strong association with lactylation levels, especially in monocytes and macrophages. This connection aligns with the metabolic origin of lactylation modification, which depends on lactate as its substrate. These findings provide novel molecular insights into how lactylation may regulate critical biological processes under varying cellular states, such as metabolic reprogramming and immune responses, while highlighting potential mechanistic avenues for further investigation.

In summary, our study underscores the importance of immune cell dynamics in PH and highlights the potential for targeting these pathways in therapeutic strategies. However, it is essential to acknowledge the limitations of our analysis, including the need for larger sample sizes and longitudinal studies to validate these findings. Future research should focus on elucidating the precise mechanisms by which immune cell interactions contribute to vascular remodeling in PH, as well as exploring novel therapeutic avenues that may arise from targeting these immune pathways. Overall, the identification of distinct immune cell populations in PH patients not only enhances our understanding of the disease but also opens new avenues for intervention that could improve patient outcomes in the future.

Conclusion

We conducted a comprehensive analysis of the effects of lactylation-related genes on PH, and identified LCP1 as a key gene that can serve as a hub gene for the new diagnostic marker of PH patients. Meanwhile, based on single-cell data, we described the lactylation of immune cells in PH patients, providing new insights for future research on the relationship between lactylation and PH.

Abbreviations

PH

Pulmonary hypertension

DEGs

Differentially expressed genes

LCP1

Lymphocyte cytosolic protein 1

GO

Gene Ontology

KEGG

Kyoto Encyclopedia of Genes and Genomes

GSVA

Gene set variation analysis

ssGSEA

Single-sample gene set enrichment analysis

scRNA-seq

Single-cell RNA sequencing

RT-PCR

Reverse transcription-polymerase chain reaction

HVGs

Highly variable genes

PCA

Principal component analysis

UMAP

Uniform manifold approximation and projection

hPASMCs

Human pulmonary artery smooth muscle

rPASMCs

Rat pulmonary artery smooth muscle cells

SMCM

Smooth muscle cell-specific culture medium

DMEM

Dulbecco’s modified Eagle medium

PDGF-BB

Platelet-Derived Growth Factor-BB

CCK-8

Cell counting Kit-8

EdU

5-ethynyl-2’ -deoxyuridine

GSEA

Gene set enrichment analysis

PASMCs

Pulmonary artery smooth muscle cells

VSMCs

Vascular smooth muscle cells

LRGs

Lactylation related genes

Authors’ contributions

RH wrote the manuscript; XL and YW analyzed data; WL, WLW and WRW designed research; DZ and RW approved the final version to be published. All authors reviewed and approved the manuscript.

Funding

This research was supported by the fund for Natural Science Foundation of China (No.81970051), Anhui Provincial Natural Science Foundation (2508085MH191), A sub-project of the AnhuiMedical University National first-class undergraduate specialty construction program (2020SJJXSFK1364), Young Jianghuai famous medical training project, Research Fund project of Anhui Medical University (2023xkj144), Excellent physician Training program of Anhui Medical University.

Data availability

The datasets supporting the conclusions of this article are available in the GEO repository. The accession numbers are GSE113439, GSE186996 and GSE210248.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.