Machine learning algorithms integrate bulk and single-cell RNA data to reveal the crosstalk and heterogeneity of Glycolysis and Lactylation activity following Pulmonary Arterial Hypertension

Machine learning algorithms integrate bulk and single-cell RNA data to reveal the crosstalk and heterogeneity of Glycolysis and Lactylation activity following Pulmonary Arterial Hypertension | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Machine learning algorithms integrate bulk and single-cell RNA data to reveal the crosstalk and heterogeneity of Glycolysis and Lactylation activity following Pulmonary Arterial Hypertension Qiuhong Chen, Qin Zheng, Hong Yang, Jinchen He, Yuyuan Wang, Tianqi Wu, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6229513/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 10 You are reading this latest preprint version Abstract Background: Glycolysis and lactylation activity significantly impact the pathogenesis of Pulmonary Arterial Hypertension (PAH); however, studies exploring their heterogeneity and potential correlation at the single-cell level are still lacking. Identifying the feature genes that are commonly regulated by both glycolysis and lactylation could significantly enhance our understanding of PAH. Methods: We employed single-cell RNA sequencing (scRNA-seq) to investigate the heterogeneity of glycolysis and lactylation activity across various cellular tiers following PAH, aiming to acquire comprehensive biological insights into PAH. We Utilized AUCell, Ucell, singscore, ssGSEA, and AddModuleScore algorithms to identify common positive and negative regulated glycolysis and lactylation activity in PAH cellular level. Furthermore, we employed three machine learning algorithms, Boruta, Random Forest, and SVM-RFE to identify the optimal feature genes related to PAH in BulkRNA-seq level. We further leveraged CellChat and pseudotime analysis to delve into the potential biological regulatory mechanisms of the characteristic genes. We used qPCR to detect the expression of ACTR2, CCDC88A, and MRC1 in the rat model of pulmonary hypertension. Results: For the first time at the cellular level, we discovered that glycolysis and lactylation activities exhibit heterogeneity across different cell layers following PAH. However, their activities show remarkable consistency, being highly active in macrophages, fibroblasts, monocytes, and epithelial cells, while displaying lower activity in lymphatic endothelial cells. This indicates a correlation between these two pathways in PAH. Consequently, we defined a set of genes that co-regulate both pathways at the PAH level. Using various machine learning algorithms, we further identified key predictive genes for PAH, namely ACTR2, CCDC88A, and MRC1. We used qPCR to verify the excessive expression of ACTR2, CCDC88A, and MRC1 in the rat model of pulmonary hypertension. Conclusions: Following PAH, ACTR2, CCDC88A, and MRC1 might simultaneously upregulating glycolysis and lactylation activities in macrophages and monocytes and further contribute PAH progression. Clinical trial Not applicable. Biological sciences/Computational biology and bioinformatics Biological sciences/Genetics Biological sciences/Microbiology scRNA-seq machine learning glycolysis PAH lactylation Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction Pulmonary arterial hypertension (PAH), a chronic cardiopulmonary disease with a prevalence of 10.6 per million adults in European and American countries, presents ineffective therapeutic options[ 1 ]. Vasoconstriction, obstructive pulmonary vasculopathy characterized by hyperproliferation and anti-apoptosis phenotypes of PASMCs, excessive fibrosis, inflammation, thrombosis, and altered mitochondrial metabolism were involved in PAH, a disorder for which the molecular underpinnings are still mostly unknown at the single-cell level[2]. Therefore, enhancing patient outcomes requires discovering trustworthy diagnostic biomarkers and comprehending the molecular mechanisms underlying the pathophysiology of PAH. The functions of the lactylation and glycolysis pathways in PAH have attracted a lot of interest from researchers lately[ 3 – 5 ]. These pathways may be directly linked to the pathological processes of PAH, since they play important roles in cell differentiation, proliferation, and death. A cancer-like increase in cell proliferation and resistance to apoptosis reflects acquired abnormalities of mitochondrial metabolism and dynamics. Aberrant protein signaling and epigenetic dysregulation in PAH promote cell proliferation in part through induction of a Warburg mitochondrial-metabolic state of uncoupled glycolysis. Complex changes in cytokines (interleukins and tumor necrosis factor), cellular immunity (T lymphocytes, natural killer cells, macrophages), and autoantibodies suggest that PAH is, in part, an autoimmune, inflammatory disease[2]. Recent experimental approaches aimed at the normalization of glucose oxidation or the modulation of the balance between fatty acid oxidation and glucose oxidation have shown promise as therapies for PAH, confirming the importance of glucose metabolic abnormalities in this disease[ 6 , 7 ]. Additionally, lactate is a crucial product of the glycolysis process and has been associated with the progression of a variety of diseases such as tumors, inflammation, and immunity. Lactate shifts metabolic reprogramming, shapes the acidic tissue microenvironment, and recruits immune cells[ 8 ]. Moreover, lactate serves as a donor for the lactylation of proteins. Lactylation is a novel modification that drastically affects cell fate and bridges metabolic reprogramming and epigenetics[ 9 ]. For example, Chen et al. found that lactylation of NUSAP1 maintains its stability and constitutes the NUSAP1-LDHA-glycolysis-lactate feedforward loop, thereby linking the Warburg effect to metastases of pancreatic ductal adenocarcinoma[ 10 ]. There is currently insufficient research on the relationship and interaction between lactylation and glycolysis in PAH. Because of this, more research into the specific processes of the glycolysis and lactylation pathways in PAH could lead to new ways of diagnosing and treating this condition. The biomedical field has seen a rise in the use of scRNA-seq due to the ongoing advancements in sequencing technology. This technology has emerged as a crucial tool for analyzing cell heterogeneity and has become indispensable in the occurrence and development of diseases. Additionally, machine learning algorithms, when combined with other bioinformatics methodologies, can be utilized to sort through a greater number of diagnostic biomarkers. The heterogeneity and interaction of glycolysis and lactylation activities were thus methodically investigated in this work by using a variety of algorithms at the cellular level in PAH. From that gene collection, we identified potential diagnostic biomarkers for PAH using a variety of algorithms for machine learning. We next investigated at how these biomarkers affect the processes behind PAH and upregulate lactylation and glycolysis from a variety of perspectives. By fostering a more thorough comprehension and reaction to glycolysis and lactylation in PAH, this initiative strives to provide a theoretical basis for the accurate treatment of PAH. Methods Data Acquisition A total of 15 scRNA-seq samples, including 6 PAH samples and 9 normal samples was procured from the GEO database under accession ID: GSE169471 and GSE210248[11, 12]. We additionally obtained BulkRNA-seq dataset from GEO including 15 PAH samples and 11 normal samples, with accession number: GSE113439, which were meticulously annotated using the official platform[13]. scRNA-Seq Data Processing In processing single-cell RNA sequencing data, we retained high-quality cells with less than 20% mitochondrial gene content and more than 200 genes expressed. We also focused on genes that were active in at least three cells and had expression levels between 200 and 7,000. A total of 44,647 eligible cells were retained for further analysis. Subsequently, we performed integration using the Seurat pipeline[14, 15]. The remaining cells were further scaled and normalized using a linear regression model with the "Log-normalization" method, and the top 3000 highly variable genes were detected using the "FindVariableFeatures" function. Principal Component Analysis (PCA) was then applied to reduce the dimensionality of the scRNA-seq data. To eliminate batch effects among samples, soft k-means clustering was executed using the "Harmony" package[16]. Cell clustering was conducted using the "FindClusters" function with the resolution parameter set to 0.8. The methodology for annotating cell clusters involved focusing on genes with elevated expression levels, genes exhibiting unique expression patterns, and documented canonical cellular markers. Evaluation of Glycolysis and Lactylation Activity In total 341 glycolysis-related genes were obtained from GSEA website with “HALLMARK_GLYCOLYSIS.v2023.2.Hs.gmt”,“GOBP_GLYCOLYTIC_PROCESS.v2023.2.Hs.gmt”,“c2.cp.kegg.v7.5.1.symbols.gmt”,“REACTOME_GLYCOLYSIS.v2023.2.Hs.gmt”,“BIOCARTA_GLYCOLYSIS_PATHWAY.v2023.2.Hs.gmt” and meta-analysis as well as literature[17, 18] ( Table S1 ). Meanwhile, a total of 332 lactylation-related genes were obtained from literature reports[19] ( Table S2 ). We further utilized AUCell, Ucell, singscore, ssGSEA, and AddModuleScore algorithms to evaluate the glycolysis and lactylation activity of each cell at the single-cell level, and calculate the overall glycolysis and lactylation activity based on above gene sets[20, 21]. Based on the quartile method: cells with scores below the 25th percentile were categorized into the low activity group, those between the 25th and 75th percentiles were classified as the intermediate activity group, and those above the 75th percentile were designated as the high activity group. Subsequently, the ‘FindMarkers’ function was used to perform differential gene expression analysis (DEGs) to identify genes that are upregulated or downregulated in the high and low activity groups. Further, we conducted the high dimensional weighted gene co-expression network analysis (hdWGCNA) to identify a gene module related to the glycolysis and lactylation activity, which mainly included data preprocessing, gene network construction, module identification, module preservation analysis, and functional enrichment analysis[22]. Eventually, based on the intersection of the two groups of differentially expressed genes and modules associated with the glycolysis and lactylation activity by hdWGCNA, we obtained a set of genes that regulate both glycolysis and lactylation at the PAH cells level. Functional Enrichment Analysis Using the “clusterProfiler” package in R[23], we performed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional enrichment analysis on the marker genes of core cells. Specifically, the GO classification was conducted, covering three aspects: biological processes (BP), cellular components (CC), and molecular functions (MF). Identification of PAH Feature Genes The application of a combination of machine learning algorithms including SVM-RFE, RF and Boruta has been employed to predict disease states and identify obviously prognostic variables. SVM is a supervised learning method used for regression and classification; RFE algorithm helps prevent overfitting while producing interpretable results[24, 25]. Consequently, the SVM-RFE algorithm is used to identify genomic sets with the highest discrimination, which are then used to identify the most suitable feature genes. RF is the most popular approaches in solving various prediction problems[26]. The optimal number of trees is determined by selecting trees with the lowest error rate and best stability among 1-500 trees. Subsequently, an RF model was constructed based on selected parameters, and important genes were chosen using the reduction in accuracy method (Gini coefficient). Twenty-three important genes (importance >0.2) were selected as key genes for PAH diagnosis. Finally, the genes shared among the intersections of several machine learning algorithms are the optimal feature genes. Boruta is an excellent feature selection method that can help us gain a more comprehensive understanding of the factors influencing the dependent variable[20, 27]. Expression and Diagnostic Relevance of Optimal Feature Genes The Wilcoxon rank-sum test was employed to assess the expression levels of the optimal feature genes in PAH samples compared to control samples. Additionally, the predictive value of these optimal feature genes was further validated using receiver operating characteristic (ROC) curves. Pseudotime analysis The Monocle2 was employed[28] to infer cell trajectories, providing a thorough examination of the cell differentiation process. Following dimensionality reduction and cell ordering, we utilized standard parameters to accurately and reliably deduce differentiation trajectories. Monocle2's robust dimensionality reduction algorithm effectively maps high-dimensional single-cell RNA sequencing data into a low-dimensional space, revealing dynamic changes in cell states and potential differentiation pathways. To further corroborate and enhance the results from Monocle2's trajectory analysis, we also implemented the CytoTRACE algorithm with default parameters[29, 30]. CytoTRACE is based on the robust observation that transcriptional diversity decreases during cell differentiation. By leveraging this principle, CytoTRACE predicts differentiation states from scRNA-seq data, offering an additional independent method to validate the trajectories inferred by Monocle2. Cell–cell interaction analysis Utilizing the CellChat R package[31], we performed a comprehensive analysis of intercellular communication networks on the annotated scRNA-seq dataset, meticulously examining ligand-receptor-mediated interactions among different cell types. Initially, we constructed detailed cell-cell communication network maps using the CellChat package. Subsequently, the netVisual_circle function was employed to visually represent the quantity and strength of communication between the target cell cluster and other cell clusters, thereby elucidating the interaction patterns among cells. Additionally, the netVisual_bubble function was utilized to generate bubble plots, highlighting significant ligand-receptor interactions between the target cell cluster and other cell clusters, and emphasizing the critical communication pathways between different cell types. Rats All animal experiments were conducted in the SPF animal room of the Central Laboratory of Chengdu Medical College. Eight male Sprague Dawley rats (Charles River Laboratories) at the age of 6 weeks (150 g to 160 g) were randomized for two groups and treated with monocrotaline (MCT) (subcutaneous, 32 mg/kg body weight) or saline and rested for 21 days. The animal feeding conditions were (22 ± 2) ℃, 21% 0, and a light cycle of 12h: 12h. General purpose rats were fed synthetic feed and allowed to drink water freely. Hematoxylin and eosin stain Pulmonary arteries were fixed, dehydrated and embedded in paraffin. Subsequently, 5 µm of sections were prepared and dewaxed in xylene. After rehydration, sections were stained with hematoxylin for 5 min, washed 5 times in water and processed to stained in eosin for 30 s. Then. QRT-PCR analysis Total RNA was once isolated from mouse with Zymo Research Quick-RNA Miniprep Kits with DNase I digestion. One microgram of RNA was once transcribed into cDNA the usage of the high-capability cDNA reverse transcription kits according to the manufacturer’s protocol. Quantitative RT-PCR evaluation was once performed on an QuantStudio three system with the Power Track SYBR Green Master package. Target mRNA was decided the usage of the comparative cycle threshold approach of relative quantitation. Cyclophilin was used as an inner manage for evaluation of expression of mouse genes. The primer sequences were supplied in Extended Data Table 1 Table 1 Primer sequences of biomarkers in qRT-PCR Primer Sequences ACTR2-F GCAACAGCAGCAGCTTGATAA ACTR2-R GGTGTCACTCACATTTGCCC CCDC88A-F ACCCAGGCTAGACATTCACG CCDC88A-R CAGTAAATGCCCCATGGCTG MRC1-F AAATTGCCGTGAGTCCAAGAG MRC1-R TGGACTAAGCCAAGGGGCAA Statistical Analysis All data processing, statistical analysis, and plotting were conducted in R software (version 4.1.3). Wilcoxon rank-sum test or Student’s t-test was utilized for analyzing the difference between the two groups. The correlation between the variables was determined using Pearson’s or Spearman’s correlation test. All statistical P -values were two-side, and P < 0.05 was regarded as statistical significance. Results scRNA‑seq analysis quantified the diversity of major cell populations in PAH To understand gene-expression perturbations and generate a comprehensive map of the immune landscape of PAH at single-cell resolution, six PAH samples and nine PAH control standards described in the method section were kept for further exploration ( Figure S1A ). Following the application of the harmony algorithm, the cellular distribution within each sample showed uniformity, indicating that there were no significant batch effects across the samples, and this consistency allows for their use in downstream analyses ( Figure 1A ). Total cells were grouped into 27 clusters using the Seurat pipeline ( Figure 1B ) and were annotated with major cell types based on classical marker genes, including macrophages, fibroblasts, T/NK cells, monocytes, endothelial cells, smooth muscle cells (SMCs), lymphatic endothelial cells (LECs), epithelial cells, ciliated cells, myeloid dendritic cells (MDCs), innate lymphoid cells (ILCs), B cells, and club/Goblet/ Basal cells ( Figure 1C ). The marker genes corresponding to each cell population exhibited pronounced distinctions, thereby underscoring the precision of the annotations ( Figure 1D, 1E ). Analyzing the complexity and heterogeneity of glycolysis and lactylation activity To investigate the glycolysis activity at single-cell level after PAH, we used AUCell, Ucell, singscore, ssgsea, and AddModuleScore algorithms to calculate the glycolysis activity for each cell. The results of these algorithms showed that lactylation activity displayed heterogeneity across distinct cellular strata following PAH ( Figure 2A-2C) . macrophages, MDCs, and ciliated cells had highest glycolysis activity, while B cells and LEC had relatively lower lactylation activity. Subsequently, we used violin plots to illustrate the distribution of lactylation activity across different cell types in PAH and control tissue origins. For lactylation activity, by comparing the PAH group (red) with the control group (green), we observed significant heterogeneity in lactylation activity among different cell types in PAH ( Figure 2D) . Based on the quartile method described in the methods section, we defined cells with lactylation scores below 1.721 (25%) as the low lactylation score group, those with scores above 2.74 (75%) as the high lactylation score group, and those in between as the median lactylation score group ( Figure S1B, Figure 2E ). Likewise, we used same method to calculate the glycolysis activity for each cell. We found that fibroblasts, macrophages, and ciliated cells had highest glycolysis activity, while T/NK cells, B cells, LECs, endothelial cells, and ILCs had relatively lower glycolysis activity ( Figure 2F-2I) . We also used the same method, based on the 25% and 75% percentiles of total glycolysis activity, to define cells with glycolysis activity below the 25% percentile as the low glycolysis group, those above the 75% percentile as the high glycolysis group, and those in between as the intermediate glycolysis group ( Figure S1C, Figure 2J ). Compared with the high and low activity groups of lactylation, we found that the distribution of the high and low groups of the two were highly consistent in the UMAP map, which suggested the crosstalk of the two pathways. hdWGCNA revealed that high glycolysis and lactylation cells-related genes Considering the high concordance of the two pathway activities at the PAH cell level, we integrated the results using the following strategy: cells that were both in the high lactylation group and the high glycolysis group were defined as the "high_Lactylation_Glycolysis" group, cells that were both in the low lactylation group and the low glycolysis group were defined as the "low_Lactylation_Glycolysis" group, and cells that did not meet these criteria were collectively named the "median" group ( Figure 3A ). We further next applied hdWGCNA to further investigate the hub regulated genes of high_Lactylation_Glycolysis group cells. To ensure the accuracy and reliability of our analysis, we employed a rigorously optimized soft threshold, which was ultimately set at 7, as depicted in Figure 3B , and 5 modules were generated accordingly ( Figure 3C ). The top 10 hub genes of each module are displayed in Figure 3D , respectively. Within these modules, genes in the turquoise module exhibit a positive correlation with high_Lactylation_Glycolysis group cells ( Figure 3E ) and specific expression in high_Lactylation_Glycolysis group cells was observed ( Figure 3F ), demonstrating high specificity and significant expression patterns. Consequently, we identified 233 hub genes using kME (eigengene connectivity) > 0.4 from turquoise module were identified as hub genes related with high_Lactylation_Glycolysis group cells ( Table S3 ) A set of genes that simultaneously regulate glycolysis and lactylation activities at the cellular level in PAH. The transcriptional regulation of differential genes plays a critical role in the high and low lactylation and glycolysis groups. We further performed “Findmarkers” function to identified DEGs (logFC > 0.25 and adj P < 0.05 ) between high and low group at the lactylation activity ( Figure 4A, Table S4 ) and glycolysis activity ( Figure 4B, Table S5 ), subsequently. We further intersected 208 genes that separately upregulated glycolysis and lactylation activities as well as hub genes form turquoise module via hdWGCNA ( Figure 4C, Table S6 ). We further utilized the PAH bulk RNA-seq dataset GSE113439, which includes 15 PAH and 11 control samples. Using the pathway gene sets collected as described in the methods section, we calculated pathway activities at the bulk level using the ssGSEA algorithm. We found that both lactylation ( Figure 4D ) and glycolysis ( Figure 4E ) activities were significantly upregulated in PAH samples compared to normal samples. The results similarly indicate that the activities of glycolysis and lactylation are highly correlated at the bulk RNA-seq level ( Figure 4F, r = 0.83, p < 0.01 ), again suggesting high concordance of the two pathway activities in PAH. Furthermore, we found that most of the newly identified genes that can simultaneously regulate glycolysis and lactylation activities are significantly differentially expressed at the PAH bulk RNA-seq level between normal and PAH samples ( Figure 4G ). We further performed Go enrichment analysis on a total of 208 genes ( Figure S1D, Table S7 ). In Biological Processes, terms such as neutrophil activation (GO:0042119) and neutrophil mediated immunity (GO:0002446) are prominently enriched. These processes are related to immune response and protein folding repair, which may play roles in the pathological mechanisms of PAH. In Cellular Components, a significant enrichment of genes related to vacuolar membrane (GO:0005774) and lytic vacuole membrane (GO:0098852) suggests the importance of alterations in the cell membrane in PAH. Moreover, Disease Ontology (DO) enrichment analysis revealed that these 208 genes are closely related to diseases associated with pulmonary arterial hypertension (PAH), particularly those impacting the cardiovascular and respiratory systems. Notably, lung diseases and vascular conditions such as arteriosclerosis and atherosclerosis, which can contribute to altered blood flow and vessel structure, were enriched. Cardiovascular disorders like coronary artery disease and myocardial infarction, which share common pathological pathways with PAH, such as endothelial dysfunction and increased vascular stiffness, were also identified. Additionally, diseases such as HIV infection, a known risk factor for PAH, were highlighted, further emphasizing the potential involvement of these genes in the underlying mechanisms of PAH ( Figure 4H, Table S8 ). Identification of feature genes by integrating multiple machine learning algorithms We further screened the disease-specific genes for PAH from the genes that simultaneously regulate glycolysis and lactylatio by using various machine learning algorithms. Specifically, we used Boruta algorithm analysis to identify 45 diagnostic fearure genes for PAH ( Figure 5A,5B, Table S9 ). For the RF algorithm, we determined 6 top feature genes with importance greater than 0.5 ( Figure 5C, 5D, Table S10 ). Additionally, using the SVM-RFE algorithm, we selected 40 feature genes after performing 5-fold cross-validation ( Figure 5E, Table S11 ). Finally, the intersection of the feature genes obtained from the above three machine learning algorithms was taken and a total of three optimal feature genes were identified, including ACTR2, CCDC88A, and MRC1 ( Figure 5F, 5G ). Assessment of the expression and diagnosis significance of feature genes We further validated the expression levels of the 3 feature genes in PAH samples and normal samples. Additionally, the expression levels of the 3 genes were significantly upregulated in the PAH samples ( Figure 6A-6C ). Besides, to quantitatively assess the diagnostic and predictive value of the optimal feature genes, we conducted a ROC curve analysis ( Figure 6D ). The AUC values of ROC curves were ACTR2 of 0.994, CCDC88A of 1, and MRC1 of 0.982, demonstrating that these optimal feature genes had a high diagnostic value for PAH. Furthermore, we leveraged scRNA-seq to valid the expression patterns of 3 feature genes, regardless of expression ( Figures 6E, 6F ) or density ( Figures 6G ), these three genes are clearly located on macrophages, fibroblasts, and monocytes. Furthermore, three feature genes were significantly upregulated in high “high_Lactylation_Glycolysis” group cells, suggesting a possible metabolic adaptation within this group ( Figures 6H ). Additionally, we observed a strong positive correlation between three key feature genes, lactylation, and glycolysis activity at the Bulk RNA-seq level, suggesting potential crosstalk and joint involvement in the regulation of lactylation and glycolysis activity ( Figures 6I ). The Critical Role of MRC1 in Macrophage-Mediated Cellular Communication, Remodeling, and Development in PAH Pathogenesis Based on the presence or absence of MRC1 gene expression, we segregated macrophage into MRC1+ macrophage group and MRC1- macrophage group ( Figure 7A ). To better explore the mechanism by which MRC1 regulates macrophage, we utilized Monocle2 ( Figure 7B ) and CytoTRACE ( Figure 7C, 7E, 7F ) for trajectory analysis to assess transcriptional heterogeneity in macrophage. Combining the results of CytoTRACE, we identified the lower right corner as the developmental starting point of macrophage ( Figure 7D ) We found that MRC1+macrophage is primarily located in the early and mid-stages of macrophage development. Subsequently, we further meticulously constructed a co-expression network based on CellChat. We discovered that the receptors of MRC1+macrophage exhibit higher responsiveness in both signal reception and transmission compared to MRC1- macrophage, suggesting that MRC1 lays a critical role in macrophage biological processes ( Figure 7G ). Compared to the MRC1- macrophage, the MRC1+macrophage exhibited specific pathways of intercellular communication, like GALECTIN, GRN , and TWEAK pathways ( Figure 7H ). Afterwards, the scMetabolism pipeline [33], focusing on metabolic changes, revealed a completely different activation of metabolic pathways between MRC1+macrophage and MRC1- macrophage, suggesting that MRC1 in macrophages drives the reprogramming of numerous metabolic pathways ( Figure 7I ). The expression levels of ACTR2, CCDC88A and MRC1 were found to be elevated in models of pulmonary arterial hypertension. Histological samples of rat heart and lung tissue were observed under light microscope. The right ventricle of rats in the PAH group was remodeled, and the pulmonary arterioles were narrow to varying extents, with some approaching complete occlusion. Compared with the control group, the right ventricular thickness and RVHI of rats in PAH group significantly increased[(0.32±0.01)% vs. (0.58±0.07), P <0.01]. Compared with the control group, the pulmonary artery wall thickness/outer diameter (WT%) of rats in PAH group was significantly increased [(79.36±4.39)% vs. (39.14±9.49)%, P <0.01], the pulmonary artery wall area/total tube area (WA%) of rats in MCT group was significantly increased[(95.61±1.76)% vs. (54.21±3.09)%, P <0.01], and the pulmonary artery lumen area/total tube area (LA%) of rats in PAH group was significantly reduced [(4.39±1.76)% vs. (45.79±3.09)%, P <0.01] ( Figure 8A ), pulmonary hypertension model was established successfully. Compared with the control group, The mRNA expression levels of ACTR2 [(2.31±0.11) vs. (1.00±0.12), P <0.01], CCDC88A[(2.52±0.07) vs. (1.00±0.06), P <0.01], and MRC1[(2.11±0.08) vs. (1.00±0.10), P <0.01] in PAH group were significantly increased ( Figure 8B ). Discussion PAH is a fatal and progressive disease characterized by elevated pulmonary artery pressure and right ventricular hypertrophy[ 1 , 32 ]. Despite advancements in clinical treatments, PAH remains incurable, and current research is focused on identifying novel therapeutic targets at the molecular level[ 33 ]. These include gene editing to correct familial BMPR2 mutations[ 34 ], stem cell therapies to restore endothelial function[ 35 ], and immune-targeted drug interventions[ 36 ]. However, the complex pathogenesis and high heterogeneity of PAH pose significant challenges to the development of effective individualized treatments[ 37 ]. It is increasingly recognized that metabolic abnormalities are not only present in the heart and lungs of PAH patients but are also evident in animal models of the disease[38–40]. Alterations in metabolism and bioenergetics have emerged as common features in both PAH patients and animal models[ 41 ]. In this study, we focus on the critical role of glycolysis and lactate in the pathogenesis and progression of PAH. Current research indicates that glycolysis serves as a key energy source for vascular endothelial cells and smooth muscle cells[42], a finding that aligns with our scRNA scoring results. Notably, in PAH patients, as well as in cell and animal models, there is a shift in glucose metabolism from complete mitochondrial oxidative phosphorylation to increased cytoplasmic glycolysis, leading to the conversion of glucose to pyruvate and ultimately to lactate[ 43 , 44 ]. Historically, lactate was considered a metabolic byproduct, but recent studies have highlighted its role as a donor for the lactylation of histone lysine residues[45]. This lactylation can influence epigenetic regulation by promoting chromatin-associated gene transcription, a process known as lactate modification[ 46 ]. PAH shares biological processes with tumors, such as unchecked cell proliferation and resistance to regulatory mechanisms, suggesting that epigenetic dysregulation plays a significant role in PAH development and progression[ 47 ]. For instance, recent research by Jian Chen and colleagues demonstrated that mitochondrial reactive oxygen species can drive the switch to glycolysis and lactate accumulation in pulmonary artery smooth muscle cells (PASMCs). This, in turn, leads to the lactylation of HIF-1α target proteins and histones, promoting PASMC proliferation and vascular remodeling[ 48 ]. Although research on lactylation in PAH is still in its early stages, findings from Jian Chen's team and our own work suggest that targeting glycolysis and lactylation holds significant therapeutic potential for PAH. In our subsequent machine learning analysis, we identified three genes closely associated with the progression of PAH: ACTR2, CCDC88A, and MRC1. ACTR2 (Actin-Related Protein 2) is a crucial component of the Arp2/3 complex, playing a vital role in the dynamic regulation of the cytoskeleton, particularly in the formation of new actin filament branches[49]. While direct research on ACTR2’s role in PAH is currently lacking, both Hui Zhao[ 50 ] and Xu He[ 51 ] identified it as a key gene in their analyses of PAH sequencing data. Notably, studies have shown that the cytoskeleton formed by ACTR2 is important for preventing nuclear distortion and fragility caused by inflammatory signals triggered by cGAS[ 52 , 53 ]. Given that inflammation plays a significant role in PAH progression[ 54 , 55 ], future research could delve deeper into this connection. Similarly, CCDC88A (Coiled-Coil Domain Containing 88A), a signal transduction protein, also contributes to cytoskeletal regulation[56, 57]. Although CCDC88A has not yet been directly studied in the context of PAH, cytoskeleton remodeling is a critical step in cell migration and vascular remodeling—both key processes in PAH[58]. Additionally, CCDC88A is known to interact with the Akt signaling pathway[56], which has been extensively studied for its pathogenic role in PAH[ 59 ]. Some studies have even explored preclinical drug research targeting Akt[ 60 ]. Therefore, CCDC88A may also represent a valuable therapeutic target for PAH. This refined perspective underscores the potential significance of ACTR2 and CCDC88A in PAH and suggests avenues for further research into their roles as therapeutic targets. Among the three genes identified, MRC1 captured our greatest attention. MRC1 (Mannose Receptor C-Type 1), also known as CD206, is a transmembrane glycoprotein receptor expressed on the surface of macrophages, dendritic cells, and some endothelial cells[ 61 ]. During our scRNA scoring, we observed that a subpopulation of macrophages exhibited high scores for glycolysis and lactylation. Numerous studies have highlighted the critical role of macrophages in vascular remodeling and the pathogenesis of PAH. To delve deeper, we isolated macrophages for further subpopulation analysis and discovered an MRC1-negative macrophage subpopulation in PAH. MRC1 + macrophages are typically associated with the M2 phenotype[ 62 ], which is anti-inflammatory, whereas the MRC1- subpopulation is more aligned with the pro-inflammatory M1 phenotype. In our pseudo-time trajectory analysis, we observed that MRC1 expression in macrophages decreased as the cell fate progressed. Similarly, Gou et al[ 63 ]. demonstrated that the immune microenvironment of PAH rats significantly changed after exposure to lipopolysaccharide (LPS), leading to increased polarization of M1 macrophages. This shift exacerbates the inflammatory cascade, further damaging the pulmonary artery and heart through various inflammatory mediators. Additionally, Feng-Jin Shao et al. reported that knocking down the key inflammatory gene Nrf2 in PAH mice resulted in the upregulation of MRC1 mRNA expression[ 64 ]. Both studies underscore the complex interplay between the inflammatory environment in PAH and macrophage polarization. Importantly, recent research has increasingly focused on the metabolic reprogramming of macrophages in PAH progression[ 65 ]. It has been established that LPS and interferon-γ can jointly enhance glycolytic activity and remodel the tricarboxylic acid (TCA) cycle in macrophages. The high levels of succinate produced by the reprogrammed TCA cycle further amplify the release of inflammatory factors from macrophages by increasing ROS formation[ 66 ]. Based on this evidence, we propose that regulating macrophage homeostasis through modulation of glycolysis and lactylation represents a promising therapeutic target for PAH. Declarations Acknowledgements The authors sincerely thank related contributors for uploading their datasets and acknowledge the GEO database for providing their platforms. Author information Qiuhong Chen and Qin Zheng contributed equally to this work. Authors and Affiliations Department of Cardiovascular Medicine, The Second Affiliated Hospital of Chengdu Medical College,Nuclear Industry 416 Hospital, Chengdu, 610000 Sichuan, China.Qiuhong Chen, Hong Yang, Jinchen He,Yuyuan Wang, Tianqi Wu, Qi Wu 2Department of Geriatrics, The First Affiliated Hospital of Chengdu Medical College, Chengdu, 610000 Sichuan, China.Qin Zheng Author contributions Qh. C: Data curation, Investigation, Methodology, Software, Supervision, Writing original draft, Writing–review & editing, animal experiment. Q Z: Methodology, Software, Writing–original draft, animal experiment. H Y: Investigation, Writing–original draft, Software. Jc. H: Methodology, Writing–original draft. Yy. W: Investigation, Writing–review. Tq. W: Investigation, Methodology. Q W: Supervision, Writing–original draft, Writing – review & editing, fundin. All authors contributed to the article and approved the submitted version. All authors read and approved the final manuscript. Corresponding authors Correspondence to Qi Wu, E-mail: [email protected] Funding The authors declare that they received financial support for the research, authorship, and/or publication of this article.This study was supported by Key Medical Specialty Project of Chengdu,(No.CDS2022Z076); Chengdu Key Clinical Specialty Project (CDS2023ZD002), The College-level project of Chengdu Medical College(CYZYB23-19), Scientific research project of Sichuan Administration of Traditional Chinese Medicine(2023MS177) and Scientific Research Project of Chengdu Science and Technology Bureau (2022-YF05-01459-SN). Data availability The datasets generated and analysed during the current study are available from the NCBI Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) database. And the data used to support the findings of this study are available from the corresponding author upon request. Ethics approval and consent to participate. This study involving animol tissues strictly adhered to ethical regulations as approved by the Medical Ethics Committee of Chengdu Medical College. All animal experiments should comply with the ARRIVE guidelines.All methods were carried out in accordance with relevant guidelines and regulations. Consent for publication Applicable. Competing interests The authors declare no competing interests. Publisher ’ s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Consent to Publish declarations Applicable. 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/. References Ruopp, N.F. and B.A. Cockrill, Diagnosis and Treatment of Pulmonary Arterial i2. Thenappan, T., et al., Pulmonary arterial hypertension: pathogenesis and clinical management. Bmj, 2018. 360 : p. j5492. Mathew, R., et al., Metabolic Deregulation in Pulmonary Hypertension. Curr Issues Mol Biol, 2023. 45 (6): p. 4850-4874. Wang, J., et al., Functions and novel regulatory mechanisms of key glycolytic enzymes in pulmonary arterial hypertension. Eur J Pharmacol, 2024. 970 : p. 176492. Zhao, S.S., et al., Role of histone lactylation interference RNA m(6)A modification and immune microenvironment homeostasis in pulmonary arterial hypertension. Front Cell Dev Biol, 2023. 11 : p. 1268646. Fang, Y.H., et al., Therapeutic inhibition of fatty acid oxidation in right ventricular hypertrophy: exploiting Randle's cycle. J Mol Med (Berl), 2012. 90 (1): p. 31-43. Michelakis, E.D., et al., Dichloroacetate, a metabolic modulator, prevents and reverses chronic hypoxic pulmonary hypertension in rats: role of increased expression and activity of voltage-gated potassium channels. Circulation, 2002. 105 (2): p. 244-50. Zhang, Y., et al., The function and mechanism of lactate and lactylation in tumor metabolism and microenvironment. Genes Dis, 2023. 10 (5): p. 2029-2037. Wang, T., et al., Lactate-induced protein lactylation: A bridge between epigenetics and metabolic reprogramming in cancer. Cell Prolif, 2023. 56 (10): p. e13478. Chen, M., et al., NUSAP1-LDHA-Glycolysis-Lactate feedforward loop promotes Warburg effect and metastasis in pancreatic ductal adenocarcinoma. Cancer Lett, 2023. 567 : p. 216285. Crnkovic, S., et al., Single-cell transcriptomics reveals skewed cellular communication and phenotypic shift in pulmonary artery remodeling. JCI Insight, 2022. 7 (20). Saygin, D., et al., Transcriptional profiling of lung cell populations in idiopathic pulmonary arterial hypertension. Pulm Circ, 2020. 10 (1). Mura, M., et al., Osteopontin lung gene expression is a marker of disease severity in pulmonary arterial hypertension. Respirology, 2019. 24 (11): p. 1104-1110. Butler, A., et al., Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat Biotechnol, 2018. 36 (5): p. 411-420. Sun, X., et al., Architecting the metabolic reprogramming survival risk framework in LUAD through single-cell landscape analysis: three-stage ensemble learning with genetic algorithm optimization. J Transl Med, 2024. 22 (1): p. 353. Korsunsky, I., et al., Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods, 2019. 16 (12): p. 1289-1296. Nie, Q., et al., Integration RNA bulk and single cell RNA sequencing to explore the change of glycolysis-related immune microenvironment and construct prognostic signature in head and neck squamous cell carcinoma. Transl Oncol, 2024. 46 : p. 102021. Shou, Y., et al., A prediction model identifying glycolysis signature as therapeutic target for psoriasis. Front Immunol, 2023. 14 : p. 1188745. Huang, H., et al., A multi-dimensional approach to unravel the intricacies of lactylation related signature for prognostic and therapeutic insight in colorectal cancer. J Transl Med, 2024. 22 (1): p. 211. Du, C., et al., Machine learning algorithms integrate bulk and single-cell RNA data to unveil oxidative stress following intracerebral hemorrhage. Int Immunopharmacol, 2024. 137 : p. 112449. Aibar, S., et al., SCENIC: single-cell regulatory network inference and clustering. Nat Methods, 2017. 14 (11): p. 1083-1086. Han, J., et al., Intratumoral immune heterogeneity of prostate cancer characterized by typing and hub genes. J Cell Mol Med, 2023. 27 (1): p. 101-112. Yu, G., et al., clusterProfiler: an R package for comparing biological themes among gene clusters. Omics, 2012. 16 (5): p. 284-7. Sanz, H., et al., SVM-RFE: selection and visualization of the most relevant features through non-linear kernels. BMC Bioinformatics, 2018. 19 (1): p. 432. Huang, S., et al., Applications of Support Vector Machine (SVM) Learning in Cancer Genomics. Cancer Genomics Proteomics, 2018. 15 (1): p. 41-51. Walker, A.M., et al., Evaluating the performance of random forest and iterative random forest based methods when applied to gene expression data. Comput Struct Biotechnol J, 2022. 20 : p. 3372-3386. Huang, R.H., et al., A machine learning framework develops a DNA replication stress model for predicting clinical outcomes and therapeutic vulnerability in primary prostate cancer. J Transl Med, 2023. 21 (1): p. 20. Qiu, X., et al., Single-cell mRNA quantification and differential analysis with Census. Nat Methods, 2017. 14 (3): p. 309-315. Cheng, S., et al., A pan-cancer single-cell transcriptional atlas of tumor infiltrating myeloid cells. Cell, 2021. 184 (3): p. 792-809.e23. Gulati, G.S., et al., Single-cell transcriptional diversity is a hallmark of developmental potential. Science, 2020. 367 (6476): p. 405-411. Jin, S., et al., Inference and analysis of cell-cell communication using CellChat. Nat Commun, 2021. 12 (1): p. 1088. Hassoun, P.M., Pulmonary Arterial Hypertension. N Engl J Med, 2021. 385 (25): p. 2361-2376. Maron, B.A., et al., Pulmonary Arterial Hypertension: Diagnosis, Treatment, and Novel Advances. Am J Respir Crit Care Med, 2021. 203 (12): p. 1472-1487. Bisserier, M., et al., Regulation of the Methylation and Expression Levels of the BMPR2 Gene by SIN3a as a Novel Therapeutic Mechanism in Pulmonary Arterial Hypertension. Circulation, 2021. 144 (1): p. 52-73. Chi, P.L., et al., Induced pluripotent stem cell-derived exosomes attenuate vascular remodelling in pulmonary arterial hypertension by targeting HIF-1α and Runx2. Cardiovasc Res, 2024. 120 (2): p. 203-214. Frantz, R.P., et al., Seralutinib in adults with pulmonary arterial hypertension (TORREY): a randomised, double-blind, placebo-controlled phase 2 trial. Lancet Respir Med, 2024. 12 (7): p. 523-534. Humbert, M., et al., Treatment of pulmonary arterial hypertension: recent progress and a look to the future. Lancet Respir Med, 2023. 11 (9): p. 804-819. Xu, W. and S.C. Erzurum, Endothelial cell energy metabolism, proliferation, and apoptosis in pulmonary hypertension. Compr Physiol, 2011. 1 (1): p. 357-72. Xu, W., et al., Integrative proteomics and phosphoproteomics in pulmonary arterial hypertension. Sci Rep, 2019. 9 (1): p. 18623. Hernandez-Saavedra, D., et al., Stable isotope metabolomics of pulmonary artery smooth muscle and endothelial cells in pulmonary hypertension and with TGF-beta treatment. Sci Rep, 2020. 10 (1): p. 413. Xu, W., A.J. Janocha, and S.C. Erzurum, Metabolism in Pulmonary Hypertension. Annu Rev Physiol, 2021. 83 : p. 551-576. Potente, M. and P. Carmeliet, The Link Between Angiogenesis and Endothelial Metabolism. Annu Rev Physiol, 2017. 79 : p. 43-66. Drake, J.I., et al., Molecular signature of a right heart failure program in chronic severe pulmonary hypertension. Am J Respir Cell Mol Biol, 2011. 45 (6): p. 1239-47. Zhao, L., et al., Heterogeneity in lung (18)FDG uptake in pulmonary arterial hypertension: potential of dynamic (18)FDG positron emission tomography with kinetic analysis as a bridging biomarker for pulmonary vascular remodeling targeted treatments. Circulation, 2013. 128 (11): p. 1214-24. Morgan, M.A.J. and A. Shilatifard, Reevaluating the roles of histone-modifying enzymes and their associated chromatin modifications in transcriptional regulation. Nat Genet, 2020. 52 (12): p. 1271-1281. Zhang, D., et al., Metabolic regulation of gene expression by histone lactylation. Nature, 2019. 574 (7779): p. 575-580. Galiè, N., et al., An overview of the 6th World Symposium on Pulmonary Hypertension. Eur Respir J, 2019. 53 (1). Chen, J., et al., Histone lactylation driven by mROS-mediated glycolytic shift promotes hypoxic pulmonary hypertension. J Mol Cell Biol, 2023. 14 (12). Schafer, D.A. and T.A. Schroer, Actin-related proteins. Annu Rev Cell Dev Biol, 1999. 15 : p. 341-63. Zhao, H., et al., Identification of the shared gene signatures between pulmonary fibrosis and pulmonary hypertension using bioinformatics analysis. Front Immunol, 2023. 14 : p. 1197752. He, X., et al., Identification of immune-associated signatures and potential therapeutic targets for pulmonary arterial hypertension. J Cell Mol Med, 2023. 27 (23): p. 3864-3877. Sladitschek-Martens, H.L., et al., YAP/TAZ activity in stromal cells prevents ageing by controlling cGAS-STING. Nature, 2022. 607 (7920): p. 790-798. Gao, P., et al., Cyclic [G(2',5')pA(3',5')p] is the metazoan second messenger produced by DNA-activated cyclic GMP-AMP synthase. Cell, 2013. 153 (5): p. 1094-107. Wang, R.R., et al., Immunity and inflammation in pulmonary arterial hypertension: From pathophysiology mechanisms to treatment perspective. Pharmacol Res, 2022. 180 : p. 106238. Liu, S.F., et al., Pulmonary hypertension: Linking inflammation and pulmonary arterial stiffening. Front Immunol, 2022. 13 : p. 959209. Enomoto, A., J. Ping, and M. Takahashi, Girdin, a novel actin-binding protein, and its family of proteins possess versatile functions in the Akt and Wnt signaling pathways. Ann N Y Acad Sci, 2006. 1086 : p. 169-84. Aznar, N., et al., Heterotrimeric G protein signaling via GIV/Girdin: Breaking the rules of engagement, space, and time. Bioessays, 2016. 38 (4): p. 379-93. Karki, P. and A.A. Birukova, Microtubules as Major Regulators of Endothelial Function: Implication for Lung Injury. Front Physiol, 2021. 12 : p. 758313. Tang, H., et al., Deficiency of the Deubiquitinase UCHL1 Attenuates Pulmonary Arterial Hypertension. Circulation, 2024. 150 (4): p. 302-316. Mirhadi, E., et al., Resveratrol: Mechanistic and therapeutic perspectives in pulmonary arterial hypertension. Pharmacol Res, 2021. 163 : p. 105287. Revel, M., et al., C1q+ macrophages: passengers or drivers of cancer progression. Trends Cancer, 2022. 8 (7): p. 517-526. Wu, Y., et al., Spatiotemporal Immune Landscape of Colorectal Cancer Liver Metastasis at Single-Cell Level. Cancer Discov, 2022. 12 (1): p. 134-153. Guo, L., et al., Regulation of the Immune Microenvironment by an NLRP3 Inhibitor Contributes to Attenuation of Acute Right Ventricular Failure in Rats with Pulmonary Arterial Hypertension. J Inflamm Res, 2021. 14 : p. 5699-5711. Shao, F.J., et al., Potential contribution of early endothelial progenitor cell (eEPC)-to-macrophage switching in the development of pulmonary plexogenic lesion. Respir Res, 2022. 23 (1): p. 290. D'Alessandro, A., et al., Hallmarks of Pulmonary Hypertension: Mesenchymal and Inflammatory Cell Metabolic Reprogramming. Antioxid Redox Signal, 2018. 28 (3): p. 230-250. Jha, A.K., et al., Network integration of parallel metabolic and transcriptional data reveals metabolic modules that regulate macrophage polarization. Immunity, 2015. 42 (3): p. 419-30. Additional Declarations No competing interests reported. Supplementary Files FigureS1.tif SupplementaryTable.xlsx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 28 Aug, 2025 Reviews received at journal 27 Aug, 2025 Reviewers agreed at journal 13 Aug, 2025 Reviews received at journal 03 May, 2025 Reviewers agreed at journal 09 Apr, 2025 Reviewers invited by journal 04 Apr, 2025 Editor assigned by journal 04 Apr, 2025 Editor invited by journal 27 Mar, 2025 Submission checks completed at journal 25 Mar, 2025 First submitted to journal 14 Mar, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6229513","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":440856792,"identity":"56a8b003-89b1-42e6-b8b6-8297fbc73a4c","order_by":0,"name":"Qiuhong Chen","email":"","orcid":"","institution":"The Second Affiliated Hospital of Chengdu Medical College,Nuclear Industry 416 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Qiuhong","middleName":"","lastName":"Chen","suffix":""},{"id":440856794,"identity":"430745f3-5acc-479a-a261-8e060432c190","order_by":1,"name":"Qin Zheng","email":"","orcid":"","institution":"The First Affiliated Hospital of Chengdu Medical College","correspondingAuthor":false,"prefix":"","firstName":"Qin","middleName":"","lastName":"Zheng","suffix":""},{"id":440856796,"identity":"46334ad1-e938-44e2-a02d-12924ccc2a65","order_by":2,"name":"Hong Yang","email":"","orcid":"","institution":"The Second Affiliated Hospital of Chengdu Medical College,Nuclear Industry 416 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Hong","middleName":"","lastName":"Yang","suffix":""},{"id":440856797,"identity":"ede24703-506b-4f5a-a6b6-553bab78afc8","order_by":3,"name":"Jinchen He","email":"","orcid":"","institution":"The Second Affiliated Hospital of Chengdu Medical College,Nuclear Industry 416 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Jinchen","middleName":"","lastName":"He","suffix":""},{"id":440856798,"identity":"cc311c7f-7a25-4f40-b4bb-fc7c135448e0","order_by":4,"name":"Yuyuan Wang","email":"","orcid":"","institution":"The Second Affiliated Hospital of Chengdu Medical College,Nuclear Industry 416 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yuyuan","middleName":"","lastName":"Wang","suffix":""},{"id":440856799,"identity":"55b971ab-079a-41f2-acad-7bf4cf338e5e","order_by":5,"name":"Tianqi Wu","email":"","orcid":"","institution":"The Second Affiliated Hospital of Chengdu Medical College,Nuclear Industry 416 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Tianqi","middleName":"","lastName":"Wu","suffix":""},{"id":440856800,"identity":"3e40e39e-03e9-4452-a17b-4da1ada2c320","order_by":6,"name":"Qi Wu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzklEQVRIie2QvQrCMBCADwLB4aSbpCj0FU4Ef6DQV6kInTr0FZxcdK84+AqKLxAJOPUBBBdBcM7opjcKBZPRId/0Be7jyAEEAn+Ktq83ZvuHZvFMztu1GBAUcxbPRKAUKUE5Mh3pMU7XxfFco8QJNNYAQhL1tCspKm2niLPl5mCqKQy3u/x3Mr6WxFsUgukeTI2Q080jMSgJ4YJ3Fv8kR2p4kVeSNc+Kb6sxriWxKPdf4tXiZO1LZ5ESD5Y0ifqOhKHvh3KOt5JAIBAItPkAa9FKIh3w9bwAAAAASUVORK5CYII=","orcid":"","institution":"The Second Affiliated Hospital of Chengdu Medical College,Nuclear Industry 416 Hospital","correspondingAuthor":true,"prefix":"","firstName":"Qi","middleName":"","lastName":"Wu","suffix":""}],"badges":[],"createdAt":"2025-03-14 23:38:07","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6229513/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6229513/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":81392482,"identity":"7e88cefd-511f-469c-8cef-bf8bb110d970","added_by":"auto","created_at":"2025-04-25 14:58:59","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":583790,"visible":true,"origin":"","legend":"\u003cp\u003e(\u003cstrong\u003eA\u003c/strong\u003e) The cell distribution of the samples showed no significant batch effect. (\u003cstrong\u003eB\u003c/strong\u003e) umap plot colored by 27 cluster of cells. (\u003cstrong\u003eC\u003c/strong\u003e) The umap of cells from all scRNA-seq samples, colored by cell-type annotation. (\u003cstrong\u003eD\u003c/strong\u003e) Dot plot showing representative marker genes for each cell type. (\u003cstrong\u003eE\u003c/strong\u003e) The umap showing representative marker genes for each cell type.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/3e0d0ac6218e786dd1aa3943.png"},{"id":81392916,"identity":"5fb91152-0cc5-4cde-949b-cc7764e875ab","added_by":"auto","created_at":"2025-04-25 15:06:59","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":603692,"visible":true,"origin":"","legend":"\u003cp\u003e(\u003cstrong\u003eA, B\u003c/strong\u003e) The results of AUCell, Ucell, singscore, ssGSEA, and AddModuleScore algorithms showed the lactylation activity using Dot plots (\u003cstrong\u003eA\u003c/strong\u003e) and violin plots (\u003cstrong\u003eB\u003c/strong\u003e). (\u003cstrong\u003eC\u003c/strong\u003e) The results of UMAP displayed lactylation activity distribution. (\u003cstrong\u003eD\u003c/strong\u003e) The difference in lactylation scoring score in each cell type in normal and PAH samples. (\u003cstrong\u003eE\u003c/strong\u003e) All cells were classified as high, median, and low lactylation group cells. (\u003cstrong\u003eF, G\u003c/strong\u003e) The results of AUCell, Ucell, singscore, ssGSEA, and AddModuleScore algorithms showed the glycolysis activity using Dot plots (\u003cstrong\u003eF\u003c/strong\u003e) and violin plots (\u003cstrong\u003eG\u003c/strong\u003e). (\u003cstrong\u003eH\u003c/strong\u003e) The results of UMAP displayed glycolysis activity distribution. (\u003cstrong\u003eI\u003c/strong\u003e) The difference in glycolysis scoring score in each cell type in normal and PAH samples. (\u003cstrong\u003eJ\u003c/strong\u003e) All cells were classified as high, median, and low glycolysis group cells.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/b73ffd51d8efb982f9587423.png"},{"id":81392485,"identity":"e7728d5c-9c4c-4822-9b51-ed433a61a671","added_by":"auto","created_at":"2025-04-25 14:58:59","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":364059,"visible":true,"origin":"","legend":"\u003cp\u003e(\u003cstrong\u003eA\u003c/strong\u003e) The results of UMAP displayed meta-activity distribution. (\u003cstrong\u003eB\u003c/strong\u003e) Top left panel depicted the soft power threshold for choosing a scale-free topology model ft greater than 0.8. The other three panels showed the mean, median, and max connectivity of the topological network respectively when different minimum soft thresholds are chosen, reflecting the connectivity of the network. The average connectivity of the topological network is most stable at the lowest soft threshold equals. (\u003cstrong\u003eC\u003c/strong\u003e) 5 modules were identified as shown in the hdWGCNA dendrogram. (\u003cstrong\u003eD\u003c/strong\u003e) Hub genes in each module were identifed and ranked by kME (eigengene-based connectivity). (\u003cstrong\u003eE\u003c/strong\u003e) Feather plots depicted the corresponding module scores in mac cells. (\u003cstrong\u003eF\u003c/strong\u003e) The bubble plot displayed the scores obtained by 5 modules in different groups.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/a30d8c1a6e64b57fb69b67d4.png"},{"id":81392486,"identity":"6ce160e8-34c9-4713-b00b-57585a950386","added_by":"auto","created_at":"2025-04-25 14:58:59","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":421099,"visible":true,"origin":"","legend":"\u003cp\u003e(\u003cstrong\u003eA\u003c/strong\u003e) The results of DEG analysis of lactylation groups. (\u003cstrong\u003eB\u003c/strong\u003e) The results of DEG analysis of glycolysis groups. (C) Venn diagram shows the source of intersecting genes. (D) The ssGSEA scores based on the lactylation gene set were compared between normal and PAH samples. (E) The ssGSEA scores based on the glycolysis gene set were compared between normal and PAH samples. (\u003cstrong\u003eF\u003c/strong\u003e) The results of Pearson correlation coefficient analysis indicated that a notable high correlation between the glycolysis and lactylatio activity scores in BulkRNA-seq level. (\u003cstrong\u003eG\u003c/strong\u003e) The volcano map shows the expression analysis of newly identified genes between PAH and normal samples. (\u003cstrong\u003eH\u003c/strong\u003e) GO enrichment analysis.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/7afd27110e168ccfafc116fa.png"},{"id":81392489,"identity":"34e2612b-15ed-4cc8-b912-ccc65e5fa231","added_by":"auto","created_at":"2025-04-25 14:58:59","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":569969,"visible":true,"origin":"","legend":"\u003cp\u003e(\u003cstrong\u003eA\u003c/strong\u003e) The history of the Boruta algorithm. (\u003cstrong\u003eB\u003c/strong\u003e) The Boruta algorithm identified feature genes associated with PAH diagnostic. Yellow represents confirmed features, while other colors denote shadow attributes. (\u003cstrong\u003eC\u003c/strong\u003e) Random forest for the relationships between the number of trees and error rate. The x-axis represents the number of decision trees and the y-axis is the error rate. (\u003cstrong\u003eD\u003c/strong\u003e) Relative importance of feature genes calculated in random forest. (\u003cstrong\u003eE\u003c/strong\u003e) The SVM-RFE algorithm was used to further select candidate optimal feature genes with the highest accuracy and lowest error obtained in the curves. The x-axis shows the number of feature selections, and the y-axis shows the prediction accuracy (left) and error (right). (\u003cstrong\u003eF-G\u003c/strong\u003e) Venn (\u003cstrong\u003eF\u003c/strong\u003e) and upset (\u003cstrong\u003eG\u003c/strong\u003e) diagrams showing the three feature genes shared by the above algorithms.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/217482cadebbf7c8bd3e00ce.png"},{"id":81392917,"identity":"64975217-e40f-45e0-ac69-a4f161993baa","added_by":"auto","created_at":"2025-04-25 15:06:59","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":445141,"visible":true,"origin":"","legend":"\u003cp\u003e(\u003cstrong\u003eA-C\u003c/strong\u003e) Box plots showing the expression of ACTR2 (\u003cstrong\u003eA\u003c/strong\u003e), CCDC88A (\u003cstrong\u003eB\u003c/strong\u003e), and MRC1 (\u003cstrong\u003eC\u003c/strong\u003e) between CT (n=11) and PAH (n=15) samples. (\u003cstrong\u003eD\u003c/strong\u003e) Roc curves estimating the diagnostic performance of ACTR2, CCDC88A, and MRC1. (\u003cstrong\u003eE\u003c/strong\u003e) Dot plot showing the expression of ACTR2, CCDC88A, and MRC1 in scRNA-seq level. (\u003cstrong\u003eF\u003c/strong\u003e) Violin plot showing the expression of ACTR2, CCDC88A, and MRC1 in scRNA-seq level. (\u003cstrong\u003eG\u003c/strong\u003e) UMAP plot showing the density of ACTR2, CCDC88A, and MRC1 in scRNA-seq level. (\u003cstrong\u003eH\u003c/strong\u003e) Violin plot showing the expression of CCDC88A, MRC1, and ACTR2 in meta groups. (\u003cstrong\u003eI\u003c/strong\u003e) The correlation among feature genes, lactylation, and glycolysis scores in BulkRNA-seq level.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/a334ecb5b3cc387e87a10c03.png"},{"id":81392918,"identity":"1d9c45bb-94e5-41ea-bb4c-495f429452e2","added_by":"auto","created_at":"2025-04-25 15:06:59","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":629972,"visible":true,"origin":"","legend":"\u003cp\u003e(\u003cstrong\u003eA)\u003c/strong\u003e The UMAP plot displayed the distribution of MRC1+ macrophage group and MRC1- macrophage group. (\u003cstrong\u003eB\u003c/strong\u003e) The development trajectory of macrophage inferred by Monocle2. (\u003cstrong\u003eC\u003c/strong\u003e) Differentiation states of macrophage predicted by cytoTRACE. (\u003cstrong\u003eD, E\u003c/strong\u003e) MRC1+ macrophage is primarily located in the early macrophage development. (\u003cstrong\u003eF\u003c/strong\u003e) The scatter plot of the inferred roles of celltypes considering their ingoing and outgoing interaction strength. (\u003cstrong\u003eG\u003c/strong\u003e) Cellchat analysis of MRC1+ macrophage group and MRC1- macrophage group, and all other cell types. Both interaction numbers and interaction strengths were showed. (\u003cstrong\u003eH\u003c/strong\u003e) Signaling role analysis on the cell–cell communication network from all signaling pathways between all cell types. (\u003cstrong\u003eI\u003c/strong\u003e) The difference of metabolic pathways activation between MRC1+ macrophage group and MRC1- macrophage group.\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/1ba94d9aeb04bccdccea6d10.png"},{"id":81392495,"identity":"6e97018b-4611-4a82-a37e-e2ad35b47123","added_by":"auto","created_at":"2025-04-25 14:58:59","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":368260,"visible":true,"origin":"","legend":"\u003cp\u003e(\u003cstrong\u003eA)\u003c/strong\u003e Pulmonary hypertension model was established successfully. (\u003cstrong\u003eB\u003c/strong\u003e) Expression of ACTR2, CCDC88A and MRC1 in lung tissue.\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/ea6a8bdc1ebeb9fdd0a22f71.png"},{"id":81394261,"identity":"65743ff6-d419-44b5-a6a1-450de3cff869","added_by":"auto","created_at":"2025-04-25 15:23:02","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5303341,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/63f9761c-206c-42cd-a766-5b57672509fa.pdf"},{"id":81392497,"identity":"51f8d3e8-4c35-4da6-abea-f1c4970d4b1c","added_by":"auto","created_at":"2025-04-25 14:58:59","extension":"tif","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":19639208,"visible":true,"origin":"","legend":"","description":"","filename":"FigureS1.tif","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/b1d60dec57a676f29e1fa067.tif"},{"id":81392487,"identity":"b23d3d03-9362-48cf-bb61-81dec2d2ca35","added_by":"auto","created_at":"2025-04-25 14:58:59","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":317371,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6229513/v1/0fd1d34a2a4a76a63436fb29.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Machine learning algorithms integrate bulk and single-cell RNA data to reveal the crosstalk and heterogeneity of Glycolysis and Lactylation activity following Pulmonary Arterial Hypertension","fulltext":[{"header":"Introduction","content":"\u003cp\u003ePulmonary arterial hypertension (PAH), a chronic cardiopulmonary disease with a prevalence of 10.6 per million adults in European and American countries, presents ineffective therapeutic options[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Vasoconstriction, obstructive pulmonary vasculopathy characterized by hyperproliferation and anti-apoptosis phenotypes of PASMCs, excessive fibrosis, inflammation, thrombosis, and altered mitochondrial metabolism were involved in PAH, a disorder for which the molecular underpinnings are still mostly unknown at the single-cell level[2]. Therefore, enhancing patient outcomes requires discovering trustworthy diagnostic biomarkers and comprehending the molecular mechanisms underlying the pathophysiology of PAH.\u003c/p\u003e \u003cp\u003eThe functions of the lactylation and glycolysis pathways in PAH have attracted a lot of interest from researchers lately[\u003cspan additionalcitationids=\"CR4\" citationid=\"CR2\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. These pathways may be directly linked to the pathological processes of PAH, since they play important roles in cell differentiation, proliferation, and death. A cancer-like increase in cell proliferation and resistance to apoptosis reflects acquired abnormalities of mitochondrial metabolism and dynamics. Aberrant protein signaling and epigenetic dysregulation in PAH promote cell proliferation in part through induction of a Warburg mitochondrial-metabolic state of uncoupled glycolysis. Complex changes in cytokines (interleukins and tumor necrosis factor), cellular immunity (T lymphocytes, natural killer cells, macrophages), and autoantibodies suggest that PAH is, in part, an autoimmune, inflammatory disease[2]. Recent experimental approaches aimed at the normalization of glucose oxidation or the modulation of the balance between fatty acid oxidation and glucose oxidation have shown promise as therapies for PAH, confirming the importance of glucose metabolic abnormalities in this disease[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Additionally, lactate is a crucial product of the glycolysis process and has been associated with the progression of a variety of diseases such as tumors, inflammation, and immunity.\u003c/p\u003e \u003cp\u003eLactate shifts metabolic reprogramming, shapes the acidic tissue microenvironment, and recruits immune cells[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Moreover, lactate serves as a donor for the lactylation of proteins. Lactylation is a novel modification that drastically affects cell fate and bridges metabolic reprogramming and epigenetics[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. For example, Chen et al. found that lactylation of NUSAP1 maintains its stability and constitutes the NUSAP1-LDHA-glycolysis-lactate feedforward loop, thereby linking the Warburg effect to metastases of pancreatic ductal adenocarcinoma[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. There is currently insufficient research on the relationship and interaction between lactylation and glycolysis in PAH. Because of this, more research into the specific processes of the glycolysis and lactylation pathways in PAH could lead to new ways of diagnosing and treating this condition.\u003c/p\u003e \u003cp\u003eThe biomedical field has seen a rise in the use of scRNA-seq due to the ongoing advancements in sequencing technology. This technology has emerged as a crucial tool for analyzing cell heterogeneity and has become indispensable in the occurrence and development of diseases. Additionally, machine learning algorithms, when combined with other bioinformatics methodologies, can be utilized to sort through a greater number of diagnostic biomarkers. The heterogeneity and interaction of glycolysis and lactylation activities were thus methodically investigated in this work by using a variety of algorithms at the cellular level in PAH. From that gene collection, we identified potential diagnostic biomarkers for PAH using a variety of algorithms for machine learning. We next investigated at how these biomarkers affect the processes behind PAH and upregulate lactylation and glycolysis from a variety of perspectives. By fostering a more thorough comprehension and reaction to glycolysis and lactylation in PAH, this initiative strives to provide a theoretical basis for the accurate treatment of PAH.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eData Acquisition\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 15 scRNA-seq samples, including 6 PAH samples and 9 normal samples was procured from the GEO database under accession ID: GSE169471 and GSE210248[11, 12]. We additionally obtained BulkRNA-seq dataset from GEO including 15 PAH samples and 11 normal samples, with accession number: GSE113439, which were meticulously annotated using the official platform[13].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003escRNA-Seq Data Processing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn processing single-cell RNA sequencing data, we retained high-quality cells with less than 20% mitochondrial gene content and more than 200 genes expressed. We also focused on genes that were active in at least three cells and had expression levels between 200 and 7,000. A total of 44,647 eligible cells were retained for further analysis. Subsequently, we performed integration using the Seurat pipeline[14, 15]. The remaining cells were further scaled and normalized using a linear regression model with the \u0026quot;Log-normalization\u0026quot; method, and the top 3000 highly variable genes were detected using the \u0026quot;FindVariableFeatures\u0026quot; function. Principal Component Analysis (PCA) was then applied to reduce the dimensionality of the scRNA-seq data. To eliminate batch effects among samples, soft k-means clustering was executed using the \u0026quot;Harmony\u0026quot; package[16]. Cell clustering was conducted using the \u0026quot;FindClusters\u0026quot; function with the resolution parameter set to 0.8. The methodology for annotating cell clusters involved focusing on genes with elevated expression levels, genes exhibiting unique expression patterns, and documented canonical cellular markers.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEvaluation of Glycolysis and Lactylation Activity\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn total 341 glycolysis-related genes were obtained from GSEA website with \u0026ldquo;HALLMARK_GLYCOLYSIS.v2023.2.Hs.gmt\u0026rdquo;,\u0026ldquo;GOBP_GLYCOLYTIC_PROCESS.v2023.2.Hs.gmt\u0026rdquo;,\u0026ldquo;c2.cp.kegg.v7.5.1.symbols.gmt\u0026rdquo;,\u0026ldquo;REACTOME_GLYCOLYSIS.v2023.2.Hs.gmt\u0026rdquo;,\u0026ldquo;BIOCARTA_GLYCOLYSIS_PATHWAY.v2023.2.Hs.gmt\u0026rdquo; and meta-analysis as well as literature[17, 18] (\u003cstrong\u003eTable S1\u003c/strong\u003e). Meanwhile, a total of 332 lactylation-related genes were obtained from literature reports[19] (\u003cstrong\u003eTable S2\u003c/strong\u003e). We further utilized AUCell, Ucell, singscore, ssGSEA, and AddModuleScore algorithms to evaluate the glycolysis and lactylation activity of each cell at the single-cell level, and calculate the overall glycolysis and lactylation activity based on above gene sets[20, 21]. Based on the quartile method: cells with scores below the 25th percentile were categorized into the low activity group, those between the 25th and 75th percentiles were classified as the intermediate activity group, and those above the 75th percentile were designated as the high activity group. Subsequently, the \u0026lsquo;FindMarkers\u0026rsquo; function was used to perform differential gene expression analysis (DEGs) to identify genes that are upregulated or downregulated in the high and low activity groups. Further, we conducted the\u0026nbsp;high dimensional weighted gene co-expression network analysis (hdWGCNA) to identify a gene module related to the glycolysis and lactylation activity, which mainly included data preprocessing, gene network construction, module identification, module preservation analysis, and functional enrichment analysis[22]. Eventually, based on the intersection of the two groups of differentially expressed genes and modules associated with the glycolysis and lactylation activity by hdWGCNA, we obtained a set of genes that regulate both glycolysis and lactylation at the PAH cells level.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunctional Enrichment Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUsing the \u0026ldquo;clusterProfiler\u0026rdquo; package in R[23], we performed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional enrichment analysis on the marker genes of core cells. Specifically, the GO classification was conducted, covering three aspects: biological processes (BP), cellular components (CC), and molecular functions (MF).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of PAH Feature Genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe application of a combination of machine learning algorithms including SVM-RFE, RF and Boruta has been employed to predict disease states and identify obviously prognostic variables. SVM is a supervised learning method used for regression and classification; RFE algorithm helps prevent overfitting while producing interpretable results[24, 25]. Consequently, the SVM-RFE algorithm is used to identify genomic sets with the highest discrimination, which are then used to identify the most suitable feature genes. RF is the most popular approaches in solving various prediction problems[26]. The optimal number of trees is determined by selecting trees with the lowest error rate and best stability among 1-500 trees. Subsequently, an RF model was constructed based on selected parameters, and important genes were chosen using the reduction in accuracy method (Gini coefficient). Twenty-three important genes (importance \u0026gt;0.2) were selected as key genes for PAH diagnosis. Finally, the genes shared among the intersections of several machine learning algorithms are the optimal feature genes. Boruta is an excellent feature selection method that can help us gain a more comprehensive understanding of the factors influencing the dependent variable[20, 27].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eExpression and Diagnostic Relevance of Optimal Feature Genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Wilcoxon rank-sum test was employed to assess the expression levels of the optimal feature genes in PAH samples compared to control samples. Additionally, the predictive value of these optimal feature genes was further validated using receiver operating characteristic (ROC) curves.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePseudotime analysis\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Monocle2 was employed[28] to infer cell trajectories, providing a thorough examination of the cell differentiation process. Following dimensionality reduction and cell ordering, we utilized standard parameters to accurately and reliably deduce differentiation trajectories. Monocle2\u0026apos;s robust dimensionality reduction algorithm effectively maps high-dimensional single-cell RNA sequencing data into a low-dimensional space, revealing dynamic changes in cell states and potential differentiation pathways. To further corroborate and enhance the results from Monocle2\u0026apos;s trajectory analysis, we also implemented the CytoTRACE algorithm with default parameters[29, 30]. CytoTRACE is based on the robust observation that transcriptional diversity decreases during cell differentiation. By leveraging this principle, CytoTRACE predicts differentiation states from scRNA-seq data, offering an additional independent method to validate the trajectories inferred by Monocle2.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCell\u0026ndash;cell interaction analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUtilizing the CellChat R package[31], we performed a comprehensive analysis of intercellular communication networks on the annotated scRNA-seq dataset, meticulously examining ligand-receptor-mediated interactions among different cell types. Initially, we constructed detailed cell-cell communication network maps using the CellChat package. Subsequently, the netVisual_circle function was employed to visually represent the quantity and strength of communication between the target cell cluster and other cell clusters, thereby elucidating the interaction patterns among cells. Additionally, the netVisual_bubble function was utilized to generate bubble plots, highlighting significant ligand-receptor interactions between the target cell cluster and other cell clusters, and emphasizing the critical communication pathways between different cell types. \u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRats\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll animal experiments were conducted in the SPF animal room of the Central Laboratory of Chengdu Medical College. Eight male Sprague Dawley rats (Charles River Laboratories) at the age of 6 weeks (150 g to 160 g) were randomized for two groups and treated with monocrotaline (MCT) (subcutaneous, 32 mg/kg body weight) or saline and rested for 21 days. The animal feeding conditions were (22\u0026nbsp;\u0026plusmn;\u0026nbsp;2)\u0026nbsp;℃, 21% 0, and a light cycle of 12h: 12h. General purpose rats were fed synthetic feed and allowed to drink water freely.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHematoxylin and eosin stain\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePulmonary arteries were fixed, dehydrated and embedded in paraffin. Subsequently, 5 \u0026micro;m of sections were prepared and dewaxed in xylene. After rehydration, sections were stained with hematoxylin for 5 min, washed 5 times in water and processed to stained in eosin for 30 s. Then.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eQRT-PCR analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTotal RNA\u0026nbsp;was once\u0026nbsp;isolated\u0026nbsp;from mouse with Zymo Research Quick-RNA Miniprep Kits with DNase I digestion. One microgram of RNA\u0026nbsp;was once\u0026nbsp;transcribed into cDNA\u0026nbsp;the usage of\u0026nbsp;the high-capability\u0026nbsp;cDNA reverse transcription kits according\u0026nbsp;to the manufacturer\u0026rsquo;s protocol. Quantitative RT-PCR\u0026nbsp;evaluation\u0026nbsp;was once\u0026nbsp;performed\u0026nbsp;on an QuantStudio\u0026nbsp;three\u0026nbsp;system\u0026nbsp;with the Power Track SYBR Green Master\u0026nbsp;package. Target mRNA\u0026nbsp;was\u0026nbsp;decided\u0026nbsp;the usage of\u0026nbsp;the comparative cycle threshold\u0026nbsp;approach\u0026nbsp;of relative quantitation. Cyclophilin\u0026nbsp;was\u0026nbsp;used as an\u0026nbsp;inner\u0026nbsp;manage\u0026nbsp;for\u0026nbsp;evaluation\u0026nbsp;of expression of mouse genes.\u0026nbsp;The primer sequences\u0026nbsp;were\u0026nbsp;supplied\u0026nbsp;in Extended Data Table 1\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1 Primer sequences of biomarkers in qRT-PCR\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 184px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePrimer\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 184px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSequences\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eACTR2-F\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eGCAACAGCAGCAGCTTGATAA\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eACTR2-R\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eGGTGTCACTCACATTTGCCC\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eCCDC88A-F\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eACCCAGGCTAGACATTCACG\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eCCDC88A-R\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eCAGTAAATGCCCCATGGCTG\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eMRC1-F\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eAAATTGCCGTGAGTCCAAGAG\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eMRC1-R\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 184px;\"\u003e\n \u003cp\u003eTGGACTAAGCCAAGGGGCAA\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data processing, statistical analysis, and plotting were conducted in R software (version 4.1.3). Wilcoxon rank-sum test or Student\u0026rsquo;s t-test was utilized for analyzing the difference between the two groups. The correlation between the variables was determined using Pearson\u0026rsquo;s or Spearman\u0026rsquo;s correlation test. All statistical \u003cem\u003eP\u003c/em\u003e-values were two-side, and \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05 was regarded as statistical significance.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003escRNA‑seq analysis quantified the\u0026nbsp;diversity of\u0026nbsp;major cell populations in\u0026nbsp;PAH\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo understand gene-expression perturbations and generate a comprehensive map of the immune landscape of PAH at single-cell resolution, six PAH samples and nine PAH control standards described in the method section were kept for further exploration (\u003cstrong\u003eFigure S1A\u003c/strong\u003e). Following the application of the harmony algorithm, the cellular distribution within each sample showed uniformity, indicating that there were no significant batch effects across the samples, and this consistency allows for their use in downstream analyses (\u003cstrong\u003eFigure 1A\u003c/strong\u003e). Total cells were grouped into 27 clusters using the Seurat pipeline (\u003cstrong\u003eFigure 1B\u003c/strong\u003e) and were annotated with major cell types based on classical marker genes, including macrophages, fibroblasts, T/NK cells, monocytes, endothelial cells, smooth muscle cells (SMCs), lymphatic endothelial cells (LECs), epithelial cells, ciliated cells, myeloid dendritic cells (MDCs), innate lymphoid cells (ILCs), B cells, and club/Goblet/ Basal cells (\u003cstrong\u003eFigure 1C\u003c/strong\u003e). The marker genes corresponding to each cell population exhibited pronounced distinctions, thereby underscoring the precision of the annotations (\u003cstrong\u003eFigure 1D, 1E\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAnalyzing the complexity and heterogeneity of glycolysis and lactylation activity\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo investigate the glycolysis activity at single-cell level after PAH, we used AUCell, Ucell, singscore, ssgsea, and AddModuleScore algorithms to calculate the glycolysis activity for each cell. The results of these algorithms showed that lactylation activity displayed heterogeneity across distinct cellular strata following PAH (\u003cstrong\u003eFigure 2A-2C)\u003c/strong\u003e. macrophages, MDCs, and ciliated cells had highest glycolysis activity, while B cells and LEC had relatively lower lactylation activity. Subsequently, we used violin plots to illustrate the distribution of lactylation activity across different cell types in PAH and control tissue origins. For lactylation activity, by comparing the PAH group (red) with the control group (green), we observed significant heterogeneity in lactylation activity among different cell types in PAH (\u003cstrong\u003eFigure 2D)\u003c/strong\u003e. Based on the quartile method described in the methods section, we defined cells with lactylation scores below 1.721 (25%) as the low lactylation score group, those with scores above 2.74 (75%) as the high lactylation score group, and those in between as the median lactylation score group (\u003cstrong\u003eFigure S1B, Figure 2E\u003c/strong\u003e). Likewise, we used same method to calculate the glycolysis activity for each cell. We found that fibroblasts, macrophages, and ciliated cells had highest glycolysis activity, while T/NK cells, B cells, LECs, endothelial cells, and ILCs had relatively lower glycolysis activity (\u003cstrong\u003eFigure 2F-2I)\u003c/strong\u003e. We also used the same method, based on the 25% and 75% percentiles of total glycolysis activity, to define cells with glycolysis activity below the 25% percentile as the low glycolysis group, those above the 75% percentile as the high glycolysis group, and those in between as the intermediate glycolysis group (\u003cstrong\u003eFigure S1C, Figure 2J\u003c/strong\u003e). Compared with the high and low activity groups of lactylation, we found that the distribution of the high and low groups of the two were highly consistent in the UMAP map, which suggested the crosstalk of the two pathways.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ehdWGCNA revealed that high glycolysis and lactylation cells-related genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConsidering the high concordance of the two pathway activities at the PAH cell level, we integrated the results using the following strategy: cells that were both in the high lactylation group and the high glycolysis group were defined as the \u0026quot;high_Lactylation_Glycolysis\u0026quot; group, cells that were both in the low lactylation group and the low glycolysis group were defined as the \u0026quot;low_Lactylation_Glycolysis\u0026quot; group, and cells that did not meet these criteria were collectively named the \u0026quot;median\u0026quot; group (\u003cstrong\u003eFigure 3A\u003c/strong\u003e). We further next applied hdWGCNA to further investigate the hub regulated genes of high_Lactylation_Glycolysis group cells. To ensure the accuracy and reliability of our analysis, we employed a rigorously optimized soft threshold, which was ultimately set at 7, as depicted in \u003cstrong\u003eFigure 3B\u003c/strong\u003e, and 5 modules were generated accordingly (\u003cstrong\u003eFigure 3C\u003c/strong\u003e). The top 10 hub genes of each module are displayed in \u003cstrong\u003eFigure 3D\u003c/strong\u003e, respectively. Within these modules, genes in the turquoise module exhibit a positive correlation with high_Lactylation_Glycolysis group cells (\u003cstrong\u003eFigure 3E\u003c/strong\u003e) and specific expression in high_Lactylation_Glycolysis group cells was observed (\u003cstrong\u003eFigure 3F\u003c/strong\u003e), demonstrating high specificity and significant expression patterns. Consequently, we identified 233 hub genes using kME (eigengene connectivity) \u0026gt; 0.4 from turquoise module were identified as hub genes related with high_Lactylation_Glycolysis group cells (\u003cstrong\u003eTable S3\u003c/strong\u003e)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eA set of genes that simultaneously regulate glycolysis and lactylation activities at the cellular level in PAH.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe transcriptional regulation of differential genes plays a critical role in the high and low lactylation and glycolysis groups. We further performed \u0026ldquo;Findmarkers\u0026rdquo; function to identified DEGs (logFC \u0026gt; 0.25 and adj\u003cem\u003eP \u0026lt; 0.05\u003c/em\u003e) between high and low group at the lactylation activity (\u003cstrong\u003eFigure 4A, Table S4\u003c/strong\u003e) and glycolysis activity (\u003cstrong\u003eFigure 4B, Table S5\u003c/strong\u003e), subsequently. We further intersected 208 genes that separately upregulated glycolysis and lactylation activities as well as hub genes form turquoise module via hdWGCNA (\u003cstrong\u003eFigure 4C, Table S6\u003c/strong\u003e). We further utilized the PAH bulk RNA-seq dataset GSE113439, which includes 15 PAH and 11 control samples. Using the pathway gene sets collected as described in the methods section, we calculated pathway activities at the bulk level using the ssGSEA algorithm. We found that both lactylation (\u003cstrong\u003eFigure 4D\u003c/strong\u003e) and glycolysis (\u003cstrong\u003eFigure 4E\u003c/strong\u003e) activities were significantly upregulated in PAH samples compared to normal samples. The results similarly indicate that the activities of glycolysis and lactylation are highly correlated at the bulk RNA-seq level (\u003cstrong\u003eFigure 4F, \u003cem\u003er\u003c/em\u003e = 0.83, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01\u003c/strong\u003e), again suggesting high concordance of the two pathway activities in PAH. Furthermore, we found that most of the newly identified genes that can simultaneously regulate glycolysis and lactylation activities are significantly differentially expressed at the PAH bulk RNA-seq level between normal and PAH samples (\u003cstrong\u003eFigure 4G\u003c/strong\u003e). We further performed Go enrichment analysis on a total of 208 genes (\u003cstrong\u003eFigure S1D, Table S7\u003c/strong\u003e). In Biological Processes, terms such as neutrophil activation (GO:0042119) and neutrophil mediated immunity (GO:0002446) are prominently enriched. These processes are related to immune response and protein folding repair, which may play roles in the pathological mechanisms of PAH. In Cellular Components, a significant enrichment of genes related to vacuolar membrane (GO:0005774) and lytic vacuole membrane (GO:0098852) suggests the importance of alterations in the cell membrane in PAH. Moreover, Disease Ontology (DO) enrichment analysis revealed that these 208 genes are closely related to diseases associated with pulmonary arterial hypertension (PAH), particularly those impacting the cardiovascular and respiratory systems. Notably, lung diseases and vascular conditions such as arteriosclerosis and atherosclerosis, which can contribute to altered blood flow and vessel structure, were enriched. Cardiovascular disorders like coronary artery disease and myocardial infarction, which share common pathological pathways with PAH, such as endothelial dysfunction and increased vascular stiffness, were also identified. Additionally, diseases such as HIV infection, a known risk factor for PAH, were highlighted, further emphasizing the potential involvement of these genes in the underlying mechanisms of PAH (\u003cstrong\u003eFigure 4H, Table S8\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of feature genes by integrating multiple machine learning algorithms\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe further screened the disease-specific genes for PAH from the genes that simultaneously regulate glycolysis and lactylatio by using various machine learning algorithms. Specifically, we used Boruta algorithm analysis to identify 45 diagnostic fearure genes for PAH (\u003cstrong\u003eFigure 5A,5B, Table S9\u003c/strong\u003e). For the RF algorithm, we determined 6 top feature genes with importance greater than 0.5 (\u003cstrong\u003eFigure 5C, 5D, Table S10\u003c/strong\u003e). Additionally, using the SVM-RFE algorithm, we selected 40 feature genes after performing 5-fold cross-validation (\u003cstrong\u003eFigure 5E, Table S11\u003c/strong\u003e). Finally, the intersection of the feature genes obtained from the above three machine learning algorithms was taken and a total of three optimal feature genes were identified, including ACTR2, CCDC88A, and MRC1 (\u003cstrong\u003eFigure 5F, 5G\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of the expression and diagnosis significance of feature genes\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe further validated the expression levels of the 3 feature genes in PAH samples and normal samples. Additionally, the expression levels of the 3 genes were significantly upregulated in the PAH samples (\u003cstrong\u003eFigure 6A-6C\u003c/strong\u003e). Besides, to quantitatively assess the diagnostic and predictive value of the optimal feature genes, we conducted a ROC curve analysis (\u003cstrong\u003eFigure 6D\u003c/strong\u003e). The AUC values of ROC curves were ACTR2 of 0.994, CCDC88A of 1, and MRC1 of 0.982, demonstrating that these optimal feature genes had a high diagnostic value for PAH. Furthermore, we leveraged scRNA-seq to valid the expression patterns of 3 feature genes, regardless of expression (\u003cstrong\u003eFigures 6E, 6F\u003c/strong\u003e) or density (\u003cstrong\u003eFigures 6G\u003c/strong\u003e), these three genes are clearly located on macrophages, fibroblasts, and monocytes. Furthermore, three feature genes were significantly upregulated in high \u0026ldquo;high_Lactylation_Glycolysis\u0026rdquo; group cells, suggesting a possible metabolic adaptation within this group (\u003cstrong\u003eFigures 6H\u003c/strong\u003e). Additionally, we observed a strong positive correlation between three key feature genes, lactylation, and glycolysis activity at the Bulk RNA-seq level, suggesting potential crosstalk and joint involvement in the regulation of lactylation and glycolysis activity (\u003cstrong\u003eFigures 6I\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe Critical Role of MRC1 in Macrophage-Mediated Cellular Communication, Remodeling, and Development in PAH Pathogenesis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on the presence or absence of MRC1 gene expression, we segregated macrophage into MRC1+ macrophage group and MRC1- macrophage group (\u003cstrong\u003eFigure 7A\u003c/strong\u003e). To better explore the mechanism by which MRC1 regulates macrophage, we utilized Monocle2 (\u003cstrong\u003eFigure 7B\u003c/strong\u003e) and CytoTRACE (\u003cstrong\u003eFigure 7C, 7E, 7F\u003c/strong\u003e) for trajectory analysis to assess transcriptional heterogeneity in macrophage. Combining the results of CytoTRACE, we identified the lower right corner as the developmental starting point of macrophage (\u003cstrong\u003eFigure 7D\u003c/strong\u003e) We found that MRC1+macrophage is primarily located in the early and mid-stages of macrophage development. Subsequently, we further meticulously constructed a co-expression network based on CellChat. We discovered that the receptors of MRC1+macrophage exhibit higher responsiveness in both signal reception and transmission compared to MRC1- macrophage, suggesting that MRC1 lays a critical role in macrophage biological processes (\u003cstrong\u003eFigure 7G\u003c/strong\u003e). Compared to the MRC1- macrophage, the MRC1+macrophage exhibited specific pathways of intercellular communication, like GALECTIN, GRN , and TWEAK pathways (\u003cstrong\u003eFigure 7H\u003c/strong\u003e). Afterwards, the scMetabolism pipeline [33], focusing on metabolic changes, revealed a completely different activation of metabolic pathways between MRC1+macrophage and MRC1- macrophage, suggesting that MRC1 in macrophages drives the reprogramming of numerous metabolic pathways (\u003cstrong\u003eFigure 7I\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe expression levels of ACTR2, CCDC88A and MRC1 were found to be elevated in models of pulmonary arterial hypertension.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHistological samples of rat heart and lung tissue were observed under light microscope. The right ventricle of rats in the PAH group was remodeled, and the pulmonary arterioles were narrow to varying extents, with some approaching complete occlusion. Compared with the control group, the right ventricular thickness and RVHI of rats in PAH group significantly increased[(0.32\u0026plusmn;0.01)% vs. (0.58\u0026plusmn;0.07), \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01]. Compared with the control group, the pulmonary artery wall thickness/outer diameter (WT%) of rats in PAH group was significantly increased [(79.36\u0026plusmn;4.39)% vs. (39.14\u0026plusmn;9.49)%, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01], the pulmonary artery wall area/total tube area (WA%) of rats in MCT group was significantly increased[(95.61\u0026plusmn;1.76)% vs. (54.21\u0026plusmn;3.09)%, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01], and the pulmonary artery lumen area/total tube area (LA%) of rats in PAH group was significantly reduced [(4.39\u0026plusmn;1.76)% vs. (45.79\u0026plusmn;3.09)%, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01]\u0026nbsp;(\u003cstrong\u003eFigure 8A\u003c/strong\u003e), pulmonary hypertension model was established successfully. Compared with the control group, The mRNA expression levels of ACTR2 [(2.31\u0026plusmn;0.11) vs. (1.00\u0026plusmn;0.12), \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01], CCDC88A[(2.52\u0026plusmn;0.07) vs. (1.00\u0026plusmn;0.06), \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01], and MRC1[(2.11\u0026plusmn;0.08) vs. (1.00\u0026plusmn;0.10), \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01] in PAH group were significantly increased (\u003cstrong\u003eFigure 8B\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003ePAH is a fatal and progressive disease characterized by elevated pulmonary artery pressure and right ventricular hypertrophy[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Despite advancements in clinical treatments, PAH remains incurable, and current research is focused on identifying novel therapeutic targets at the molecular level[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. These include gene editing to correct familial BMPR2 mutations[\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e34\u003c/span\u003e], stem cell therapies to restore endothelial function[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e35\u003c/span\u003e], and immune-targeted drug interventions[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. However, the complex pathogenesis and high heterogeneity of PAH pose significant challenges to the development of effective individualized treatments[\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIt is increasingly recognized that metabolic abnormalities are not only present in the heart and lungs of PAH patients but are also evident in animal models of the disease[38\u0026ndash;40]. Alterations in metabolism and bioenergetics have emerged as common features in both PAH patients and animal models[\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. In this study, we focus on the critical role of glycolysis and lactate in the pathogenesis and progression of PAH. Current research indicates that glycolysis serves as a key energy source for vascular endothelial cells and smooth muscle cells[42], a finding that aligns with our scRNA scoring results. Notably, in PAH patients, as well as in cell and animal models, there is a shift in glucose metabolism from complete mitochondrial oxidative phosphorylation to increased cytoplasmic glycolysis, leading to the conversion of glucose to pyruvate and ultimately to lactate[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. Historically, lactate was considered a metabolic byproduct, but recent studies have highlighted its role as a donor for the lactylation of histone lysine residues[45]. This lactylation can influence epigenetic regulation by promoting chromatin-associated gene transcription, a process known as lactate modification[\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. PAH shares biological processes with tumors, such as unchecked cell proliferation and resistance to regulatory mechanisms, suggesting that epigenetic dysregulation plays a significant role in PAH development and progression[\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. For instance, recent research by Jian Chen and colleagues demonstrated that mitochondrial reactive oxygen species can drive the switch to glycolysis and lactate accumulation in pulmonary artery smooth muscle cells (PASMCs). This, in turn, leads to the lactylation of HIF-1α target proteins and histones, promoting PASMC proliferation and vascular remodeling[\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. Although research on lactylation in PAH is still in its early stages, findings from Jian Chen's team and our own work suggest that targeting glycolysis and lactylation holds significant therapeutic potential for PAH.\u003c/p\u003e \u003cp\u003eIn our subsequent machine learning analysis, we identified three genes closely associated with the progression of PAH: ACTR2, CCDC88A, and MRC1. ACTR2 (Actin-Related Protein 2) is a crucial component of the Arp2/3 complex, playing a vital role in the dynamic regulation of the cytoskeleton, particularly in the formation of new actin filament branches[49]. While direct research on ACTR2\u0026rsquo;s role in PAH is currently lacking, both Hui Zhao[\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e50\u003c/span\u003e] and Xu He[\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e51\u003c/span\u003e] identified it as a key gene in their analyses of PAH sequencing data. Notably, studies have shown that the cytoskeleton formed by ACTR2 is important for preventing nuclear distortion and fragility caused by inflammatory signals triggered by cGAS[\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e53\u003c/span\u003e]. Given that inflammation plays a significant role in PAH progression[\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e54\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e], future research could delve deeper into this connection. Similarly, CCDC88A (Coiled-Coil Domain Containing 88A), a signal transduction protein, also contributes to cytoskeletal regulation[56, 57]. Although CCDC88A has not yet been directly studied in the context of PAH, cytoskeleton remodeling is a critical step in cell migration and vascular remodeling\u0026mdash;both key processes in PAH[58]. Additionally, CCDC88A is known to interact with the Akt signaling pathway[56], which has been extensively studied for its pathogenic role in PAH[\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e59\u003c/span\u003e]. Some studies have even explored preclinical drug research targeting Akt[\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e60\u003c/span\u003e]. Therefore, CCDC88A may also represent a valuable therapeutic target for PAH. This refined perspective underscores the potential significance of ACTR2 and CCDC88A in PAH and suggests avenues for further research into their roles as therapeutic targets.\u003c/p\u003e \u003cp\u003eAmong the three genes identified, MRC1 captured our greatest attention. MRC1 (Mannose Receptor C-Type 1), also known as CD206, is a transmembrane glycoprotein receptor expressed on the surface of macrophages, dendritic cells, and some endothelial cells[\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e61\u003c/span\u003e]. During our scRNA scoring, we observed that a subpopulation of macrophages exhibited high scores for glycolysis and lactylation. Numerous studies have highlighted the critical role of macrophages in vascular remodeling and the pathogenesis of PAH. To delve deeper, we isolated macrophages for further subpopulation analysis and discovered an MRC1-negative macrophage subpopulation in PAH. MRC1\u0026thinsp;+\u0026thinsp;macrophages are typically associated with the M2 phenotype[\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e62\u003c/span\u003e], which is anti-inflammatory, whereas the MRC1- subpopulation is more aligned with the pro-inflammatory M1 phenotype. In our pseudo-time trajectory analysis, we observed that MRC1 expression in macrophages decreased as the cell fate progressed. Similarly, Gou et al[\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e63\u003c/span\u003e]. demonstrated that the immune microenvironment of PAH rats significantly changed after exposure to lipopolysaccharide (LPS), leading to increased polarization of M1 macrophages. This shift exacerbates the inflammatory cascade, further damaging the pulmonary artery and heart through various inflammatory mediators. Additionally, Feng-Jin Shao et al. reported that knocking down the key inflammatory gene Nrf2 in PAH mice resulted in the upregulation of MRC1 mRNA expression[\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. Both studies underscore the complex interplay between the inflammatory environment in PAH and macrophage polarization. Importantly, recent research has increasingly focused on the metabolic reprogramming of macrophages in PAH progression[\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e65\u003c/span\u003e]. It has been established that LPS and interferon-γ can jointly enhance glycolytic activity and remodel the tricarboxylic acid (TCA) cycle in macrophages. The high levels of succinate produced by the reprogrammed TCA cycle further amplify the release of inflammatory factors from macrophages by increasing ROS formation[\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e66\u003c/span\u003e]. Based on this evidence, we propose that regulating macrophage homeostasis through modulation of glycolysis and lactylation represents a promising therapeutic target for PAH.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors sincerely thank related contributors for uploading their datasets and acknowledge the GEO database for providing their platforms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eQiuhong Chen and Qin Zheng contributed equally to this work.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors and Affiliations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDepartment of Cardiovascular Medicine, The Second Affiliated Hospital of Chengdu Medical College,Nuclear Industry 416 Hospital, Chengdu, 610000 Sichuan, China.Qiuhong Chen, Hong Yang, Jinchen He,Yuyuan Wang, Tianqi Wu, Qi Wu\u003c/p\u003e\n\u003cp\u003e2Department of Geriatrics, The First Affiliated Hospital of Chengdu Medical College, Chengdu, 610000 Sichuan, China.Qin Zheng\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eQh. C: Data curation, Investigation, Methodology, Software, Supervision, Writing original draft, Writing\u0026ndash;review \u0026amp; editing, animal experiment. Q Z: Methodology, Software, Writing\u0026ndash;original draft, animal experiment. H Y: Investigation, Writing\u0026ndash;original draft, Software. Jc. H: Methodology, Writing\u0026ndash;original draft. Yy. W: Investigation, Writing\u0026ndash;review. Tq. W: Investigation, Methodology. Q W: Supervision, Writing\u0026ndash;original draft, Writing \u0026ndash; review \u0026amp; editing, fundin. All authors contributed to the article and approved the submitted version. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorresponding authors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCorrespondence to Qi Wu, E-mail: [email protected]\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they received financial support for the research, authorship, and/or publication of this article.This study was supported by Key Medical Specialty Project of Chengdu,(No.CDS2022Z076); Chengdu Key Clinical Specialty Project (CDS2023ZD002), The College-level project of Chengdu Medical College(CYZYB23-19), Scientific research project of Sichuan Administration of Traditional Chinese Medicine(2023MS177) and Scientific Research Project of Chengdu Science and Technology Bureau (2022-YF05-01459-SN).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and analysed during the current study are available from the NCBI Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) database. And the data used to support the findings of this study are available from the corresponding author upon request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study involving animol tissues strictly adhered to ethical regulations as approved by the Medical Ethics Committee of Chengdu Medical College. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAll animal experiments should comply with the ARRIVE guidelines.All methods were carried out in accordance with relevant guidelines and regulations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eApplicable.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePublisher\u003c/strong\u003e\u003cstrong\u003e\u0026rsquo;\u003c/strong\u003e\u003cstrong\u003es note\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSpringer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to Publish declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eApplicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRights and permissions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOpen Access:\u0026nbsp;\u003c/strong\u003eThis 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\u0026rsquo;s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article\u0026rsquo;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/.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eRuopp, N.F. and B.A. Cockrill, Diagnosis and Treatment of Pulmonary Arterial i2. Thenappan, T., et al., Pulmonary arterial hypertension: pathogenesis and clinical management. Bmj, 2018. \u003cstrong\u003e360\u003c/strong\u003e: p. j5492.\u003c/li\u003e\n\u003cli\u003eMathew, R., et al., Metabolic Deregulation in Pulmonary Hypertension. Curr Issues Mol Biol, 2023. \u003cstrong\u003e45\u003c/strong\u003e(6): p. 4850-4874.\u003c/li\u003e\n\u003cli\u003eWang, J., et al., Functions and novel regulatory mechanisms of key glycolytic enzymes in pulmonary arterial hypertension. Eur J Pharmacol, 2024. \u003cstrong\u003e970\u003c/strong\u003e: p. 176492.\u003c/li\u003e\n\u003cli\u003eZhao, S.S., et al., Role of histone lactylation interference RNA m(6)A modification and immune microenvironment homeostasis in pulmonary arterial hypertension. Front Cell Dev Biol, 2023. \u003cstrong\u003e11\u003c/strong\u003e: p. 1268646.\u003c/li\u003e\n\u003cli\u003eFang, Y.H., et al., Therapeutic inhibition of fatty acid oxidation in right ventricular hypertrophy: exploiting Randle\u0026apos;s cycle. J Mol Med (Berl), 2012. \u003cstrong\u003e90\u003c/strong\u003e(1): p. 31-43.\u003c/li\u003e\n\u003cli\u003eMichelakis, E.D., et al., Dichloroacetate, a metabolic modulator, prevents and reverses chronic hypoxic pulmonary hypertension in rats: role of increased expression and activity of voltage-gated potassium channels. Circulation, 2002. \u003cstrong\u003e105\u003c/strong\u003e(2): p. 244-50.\u003c/li\u003e\n\u003cli\u003eZhang, Y., et al., The function and mechanism of lactate and lactylation in tumor metabolism and microenvironment. Genes Dis, 2023. \u003cstrong\u003e10\u003c/strong\u003e(5): p. 2029-2037.\u003c/li\u003e\n\u003cli\u003eWang, T., et al., Lactate-induced protein lactylation: A bridge between epigenetics and metabolic reprogramming in cancer. Cell Prolif, 2023. \u003cstrong\u003e56\u003c/strong\u003e(10): p. e13478.\u003c/li\u003e\n\u003cli\u003eChen, M., et al., NUSAP1-LDHA-Glycolysis-Lactate feedforward loop promotes Warburg effect and metastasis in pancreatic ductal adenocarcinoma. Cancer Lett, 2023. \u003cstrong\u003e567\u003c/strong\u003e: p. 216285.\u003c/li\u003e\n\u003cli\u003eCrnkovic, S., et al., Single-cell transcriptomics reveals skewed cellular communication and phenotypic shift in pulmonary artery remodeling. JCI Insight, 2022. \u003cstrong\u003e7\u003c/strong\u003e(20).\u003c/li\u003e\n\u003cli\u003eSaygin, D., et al., Transcriptional profiling of lung cell populations in idiopathic pulmonary arterial hypertension. Pulm Circ, 2020. \u003cstrong\u003e10\u003c/strong\u003e(1).\u003c/li\u003e\n\u003cli\u003eMura, M., et al., Osteopontin lung gene expression is a marker of disease severity in pulmonary arterial hypertension. Respirology, 2019. \u003cstrong\u003e24\u003c/strong\u003e(11): p. 1104-1110.\u003c/li\u003e\n\u003cli\u003eButler, A., et al., Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat Biotechnol, 2018. \u003cstrong\u003e36\u003c/strong\u003e(5): p. 411-420.\u003c/li\u003e\n\u003cli\u003eSun, X., et al., Architecting the metabolic reprogramming survival risk framework in LUAD through single-cell landscape analysis: three-stage ensemble learning with genetic algorithm optimization. J Transl Med, 2024. \u003cstrong\u003e22\u003c/strong\u003e(1): p. 353.\u003c/li\u003e\n\u003cli\u003eKorsunsky, I., et al., Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods, 2019. \u003cstrong\u003e16\u003c/strong\u003e(12): p. 1289-1296.\u003c/li\u003e\n\u003cli\u003eNie, Q., et al., Integration RNA bulk and single cell RNA sequencing to explore the change of glycolysis-related immune microenvironment and construct prognostic signature in head and neck squamous cell carcinoma. Transl Oncol, 2024. \u003cstrong\u003e46\u003c/strong\u003e: p. 102021.\u003c/li\u003e\n\u003cli\u003eShou, Y., et al., A prediction model identifying glycolysis signature as therapeutic target for psoriasis. Front Immunol, 2023. \u003cstrong\u003e14\u003c/strong\u003e: p. 1188745.\u003c/li\u003e\n\u003cli\u003eHuang, H., et al., A multi-dimensional approach to unravel the intricacies of lactylation related signature for prognostic and therapeutic insight in colorectal cancer. J Transl Med, 2024. \u003cstrong\u003e22\u003c/strong\u003e(1): p. 211.\u003c/li\u003e\n\u003cli\u003eDu, C., et al., Machine learning algorithms integrate bulk and single-cell RNA data to unveil oxidative stress following intracerebral hemorrhage. Int Immunopharmacol, 2024. \u003cstrong\u003e137\u003c/strong\u003e: p. 112449.\u003c/li\u003e\n\u003cli\u003eAibar, S., et al., SCENIC: single-cell regulatory network inference and clustering. Nat Methods, 2017. \u003cstrong\u003e14\u003c/strong\u003e(11): p. 1083-1086.\u003c/li\u003e\n\u003cli\u003eHan, J., et al., Intratumoral immune heterogeneity of prostate cancer characterized by typing and hub genes. J Cell Mol Med, 2023. \u003cstrong\u003e27\u003c/strong\u003e(1): p. 101-112.\u003c/li\u003e\n\u003cli\u003eYu, G., et al., clusterProfiler: an R package for comparing biological themes among gene clusters. Omics, 2012. \u003cstrong\u003e16\u003c/strong\u003e(5): p. 284-7.\u003c/li\u003e\n\u003cli\u003eSanz, H., et al., SVM-RFE: selection and visualization of the most relevant features through non-linear kernels. BMC Bioinformatics, 2018. \u003cstrong\u003e19\u003c/strong\u003e(1): p. 432.\u003c/li\u003e\n\u003cli\u003eHuang, S., et al., Applications of Support Vector Machine (SVM) Learning in Cancer Genomics. Cancer Genomics Proteomics, 2018. \u003cstrong\u003e15\u003c/strong\u003e(1): p. 41-51.\u003c/li\u003e\n\u003cli\u003eWalker, A.M., et al., Evaluating the performance of random forest and iterative random forest based methods when applied to gene expression data. Comput Struct Biotechnol J, 2022. \u003cstrong\u003e20\u003c/strong\u003e: p. 3372-3386.\u003c/li\u003e\n\u003cli\u003eHuang, R.H., et al., A machine learning framework develops a DNA replication stress model for predicting clinical outcomes and therapeutic vulnerability in primary prostate cancer. J Transl Med, 2023. \u003cstrong\u003e21\u003c/strong\u003e(1): p. 20.\u003c/li\u003e\n\u003cli\u003eQiu, X., et al., Single-cell mRNA quantification and differential analysis with Census. Nat Methods, 2017. \u003cstrong\u003e14\u003c/strong\u003e(3): p. 309-315.\u003c/li\u003e\n\u003cli\u003eCheng, S., et al., A pan-cancer single-cell transcriptional atlas of tumor infiltrating myeloid cells. Cell, 2021. \u003cstrong\u003e184\u003c/strong\u003e(3): p. 792-809.e23.\u003c/li\u003e\n\u003cli\u003eGulati, G.S., et al., Single-cell transcriptional diversity is a hallmark of developmental potential. Science, 2020. \u003cstrong\u003e367\u003c/strong\u003e(6476): p. 405-411.\u003c/li\u003e\n\u003cli\u003eJin, S., et al., Inference and analysis of cell-cell communication using CellChat. Nat Commun, 2021. \u003cstrong\u003e12\u003c/strong\u003e(1): p. 1088.\u003c/li\u003e\n\u003cli\u003eHassoun, P.M., Pulmonary Arterial Hypertension. N Engl J Med, 2021. \u003cstrong\u003e385\u003c/strong\u003e(25): p. 2361-2376.\u003c/li\u003e\n\u003cli\u003eMaron, B.A., et al., Pulmonary Arterial Hypertension: Diagnosis, Treatment, and Novel Advances. Am J Respir Crit Care Med, 2021. \u003cstrong\u003e203\u003c/strong\u003e(12): p. 1472-1487.\u003c/li\u003e\n\u003cli\u003eBisserier, M., et al., Regulation of the Methylation and Expression Levels of the BMPR2 Gene by SIN3a as a Novel Therapeutic Mechanism in Pulmonary Arterial Hypertension. Circulation, 2021. \u003cstrong\u003e144\u003c/strong\u003e(1): p. 52-73.\u003c/li\u003e\n\u003cli\u003eChi, P.L., et al., Induced pluripotent stem cell-derived exosomes attenuate vascular remodelling in pulmonary arterial hypertension by targeting HIF-1\u0026alpha; and Runx2. Cardiovasc Res, 2024. \u003cstrong\u003e120\u003c/strong\u003e(2): p. 203-214.\u003c/li\u003e\n\u003cli\u003eFrantz, R.P., et al., Seralutinib in adults with pulmonary arterial hypertension (TORREY): a randomised, double-blind, placebo-controlled phase 2 trial. Lancet Respir Med, 2024. \u003cstrong\u003e12\u003c/strong\u003e(7): p. 523-534.\u003c/li\u003e\n\u003cli\u003eHumbert, M., et al., Treatment of pulmonary arterial hypertension: recent progress and a look to the future. Lancet Respir Med, 2023. \u003cstrong\u003e11\u003c/strong\u003e(9): p. 804-819.\u003c/li\u003e\n\u003cli\u003eXu, W. and S.C. Erzurum, Endothelial cell energy metabolism, proliferation, and apoptosis in pulmonary hypertension. Compr Physiol, 2011. \u003cstrong\u003e1\u003c/strong\u003e(1): p. 357-72.\u003c/li\u003e\n\u003cli\u003eXu, W., et al., Integrative proteomics and phosphoproteomics in pulmonary arterial hypertension. Sci Rep, 2019. \u003cstrong\u003e9\u003c/strong\u003e(1): p. 18623.\u003c/li\u003e\n\u003cli\u003eHernandez-Saavedra, D., et al., Stable isotope metabolomics of pulmonary artery smooth muscle and endothelial cells in pulmonary hypertension and with TGF-beta treatment. Sci Rep, 2020. \u003cstrong\u003e10\u003c/strong\u003e(1): p. 413.\u003c/li\u003e\n\u003cli\u003eXu, W., A.J. Janocha, and S.C. Erzurum, Metabolism in Pulmonary Hypertension. Annu Rev Physiol, 2021. \u003cstrong\u003e83\u003c/strong\u003e: p. 551-576.\u003c/li\u003e\n\u003cli\u003ePotente, M. and P. Carmeliet, The Link Between Angiogenesis and Endothelial Metabolism. Annu Rev Physiol, 2017. \u003cstrong\u003e79\u003c/strong\u003e: p. 43-66.\u003c/li\u003e\n\u003cli\u003eDrake, J.I., et al., Molecular signature of a right heart failure program in chronic severe pulmonary hypertension. Am J Respir Cell Mol Biol, 2011. \u003cstrong\u003e45\u003c/strong\u003e(6): p. 1239-47.\u003c/li\u003e\n\u003cli\u003eZhao, L., et al., Heterogeneity in lung (18)FDG uptake in pulmonary arterial hypertension: potential of dynamic (18)FDG positron emission tomography with kinetic analysis as a bridging biomarker for pulmonary vascular remodeling targeted treatments. Circulation, 2013. \u003cstrong\u003e128\u003c/strong\u003e(11): p. 1214-24.\u003c/li\u003e\n\u003cli\u003eMorgan, M.A.J. and A. Shilatifard, Reevaluating the roles of histone-modifying enzymes and their associated chromatin modifications in transcriptional regulation. Nat Genet, 2020. \u003cstrong\u003e52\u003c/strong\u003e(12): p. 1271-1281.\u003c/li\u003e\n\u003cli\u003eZhang, D., et al., Metabolic regulation of gene expression by histone lactylation. Nature, 2019. \u003cstrong\u003e574\u003c/strong\u003e(7779): p. 575-580.\u003c/li\u003e\n\u003cli\u003eGali\u0026egrave;, N., et al., An overview of the 6th World Symposium on Pulmonary Hypertension. Eur Respir J, 2019. \u003cstrong\u003e53\u003c/strong\u003e(1).\u003c/li\u003e\n\u003cli\u003eChen, J., et al., Histone lactylation driven by mROS-mediated glycolytic shift promotes hypoxic pulmonary hypertension. J Mol Cell Biol, 2023. \u003cstrong\u003e14\u003c/strong\u003e(12).\u003c/li\u003e\n\u003cli\u003eSchafer, D.A. and T.A. Schroer, Actin-related proteins. Annu Rev Cell Dev Biol, 1999. \u003cstrong\u003e15\u003c/strong\u003e: p. 341-63.\u003c/li\u003e\n\u003cli\u003eZhao, H., et al., Identification of the shared gene signatures between pulmonary fibrosis and pulmonary hypertension using bioinformatics analysis. Front Immunol, 2023. \u003cstrong\u003e14\u003c/strong\u003e: p. 1197752.\u003c/li\u003e\n\u003cli\u003eHe, X., et al., Identification of immune-associated signatures and potential therapeutic targets for pulmonary arterial hypertension. J Cell Mol Med, 2023. \u003cstrong\u003e27\u003c/strong\u003e(23): p. 3864-3877.\u003c/li\u003e\n\u003cli\u003eSladitschek-Martens, H.L., et al., YAP/TAZ activity in stromal cells prevents ageing by controlling cGAS-STING. Nature, 2022. \u003cstrong\u003e607\u003c/strong\u003e(7920): p. 790-798.\u003c/li\u003e\n\u003cli\u003eGao, P., et al., Cyclic [G(2\u0026apos;,5\u0026apos;)pA(3\u0026apos;,5\u0026apos;)p] is the metazoan second messenger produced by DNA-activated cyclic GMP-AMP synthase. Cell, 2013. \u003cstrong\u003e153\u003c/strong\u003e(5): p. 1094-107.\u003c/li\u003e\n\u003cli\u003eWang, R.R., et al., Immunity and inflammation in pulmonary arterial hypertension: From pathophysiology mechanisms to treatment perspective. Pharmacol Res, 2022. \u003cstrong\u003e180\u003c/strong\u003e: p. 106238.\u003c/li\u003e\n\u003cli\u003eLiu, S.F., et al., Pulmonary hypertension: Linking inflammation and pulmonary arterial stiffening. Front Immunol, 2022. \u003cstrong\u003e13\u003c/strong\u003e: p. 959209.\u003c/li\u003e\n\u003cli\u003eEnomoto, A., J. Ping, and M. Takahashi, Girdin, a novel actin-binding protein, and its family of proteins possess versatile functions in the Akt and Wnt signaling pathways. Ann N Y Acad Sci, 2006. \u003cstrong\u003e1086\u003c/strong\u003e: p. 169-84.\u003c/li\u003e\n\u003cli\u003eAznar, N., et al., Heterotrimeric G protein signaling via GIV/Girdin: Breaking the rules of engagement, space, and time. Bioessays, 2016. \u003cstrong\u003e38\u003c/strong\u003e(4): p. 379-93.\u003c/li\u003e\n\u003cli\u003eKarki, P. and A.A. Birukova, Microtubules as Major Regulators of Endothelial Function: Implication for Lung Injury. Front Physiol, 2021. \u003cstrong\u003e12\u003c/strong\u003e: p. 758313.\u003c/li\u003e\n\u003cli\u003eTang, H., et al., Deficiency of the Deubiquitinase UCHL1 Attenuates Pulmonary Arterial Hypertension. Circulation, 2024. \u003cstrong\u003e150\u003c/strong\u003e(4): p. 302-316.\u003c/li\u003e\n\u003cli\u003eMirhadi, E., et al., Resveratrol: Mechanistic and therapeutic perspectives in pulmonary arterial hypertension. Pharmacol Res, 2021. \u003cstrong\u003e163\u003c/strong\u003e: p. 105287.\u003c/li\u003e\n\u003cli\u003eRevel, M., et al., C1q+ macrophages: passengers or drivers of cancer progression. Trends Cancer, 2022. \u003cstrong\u003e8\u003c/strong\u003e(7): p. 517-526.\u003c/li\u003e\n\u003cli\u003eWu, Y., et al., Spatiotemporal Immune Landscape of Colorectal Cancer Liver Metastasis at Single-Cell Level. Cancer Discov, 2022. \u003cstrong\u003e12\u003c/strong\u003e(1): p. 134-153.\u003c/li\u003e\n\u003cli\u003eGuo, L., et al., Regulation of the Immune Microenvironment by an NLRP3 Inhibitor Contributes to Attenuation of Acute Right Ventricular Failure in Rats with Pulmonary Arterial Hypertension. J Inflamm Res, 2021. \u003cstrong\u003e14\u003c/strong\u003e: p. 5699-5711.\u003c/li\u003e\n\u003cli\u003eShao, F.J., et al., Potential contribution of early endothelial progenitor cell (eEPC)-to-macrophage switching in the development of pulmonary plexogenic lesion. Respir Res, 2022. \u003cstrong\u003e23\u003c/strong\u003e(1): p. 290.\u003c/li\u003e\n\u003cli\u003eD\u0026apos;Alessandro, A., et al., Hallmarks of Pulmonary Hypertension: Mesenchymal and Inflammatory Cell Metabolic Reprogramming. Antioxid Redox Signal, 2018. \u003cstrong\u003e28\u003c/strong\u003e(3): p. 230-250.\u003c/li\u003e\n\u003cli\u003eJha, A.K., et al., Network integration of parallel metabolic and transcriptional data reveals metabolic modules that regulate macrophage polarization. Immunity, 2015. \u003cstrong\u003e42\u003c/strong\u003e(3): p. 419-30.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"scRNA-seq, machine learning, glycolysis, PAH, lactylation","lastPublishedDoi":"10.21203/rs.3.rs-6229513/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6229513/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground: \u003c/strong\u003eGlycolysis and lactylation activity significantly impact the pathogenesis of Pulmonary Arterial Hypertension (PAH); however, studies exploring their heterogeneity and potential correlation at the single-cell level are still lacking. Identifying the feature genes that are commonly regulated by both glycolysis and lactylation could significantly enhance our understanding of PAH.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods: \u003c/strong\u003eWe employed single-cell RNA sequencing (scRNA-seq) to investigate the heterogeneity of glycolysis and lactylation activity across various cellular tiers following PAH, aiming to acquire comprehensive biological insights into PAH. We Utilized AUCell, Ucell, singscore, ssGSEA, and AddModuleScore algorithms to identify common positive and negative regulated glycolysis and lactylation activity in PAH cellular level. Furthermore, we employed three machine learning algorithms, Boruta, Random Forest, and SVM-RFE to identify the optimal feature genes related to PAH in BulkRNA-seq level. We further leveraged CellChat and pseudotime analysis to delve into the potential biological regulatory mechanisms of the characteristic genes. We used qPCR to detect the expression of ACTR2, CCDC88A, and MRC1 in the rat model of pulmonary hypertension.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e For the first time at the cellular level, we discovered that glycolysis and lactylation activities exhibit heterogeneity across different cell layers following PAH. However, their activities show remarkable consistency, being highly active in macrophages, fibroblasts, monocytes, and epithelial cells, while displaying lower activity in lymphatic endothelial cells. This indicates a correlation between these two pathways in PAH. Consequently, we defined a set of genes that co-regulate both pathways at the PAH level. Using various machine learning algorithms, we further identified key predictive genes for PAH, namely ACTR2, CCDC88A, and MRC1. We used qPCR to verify the excessive expression of ACTR2, CCDC88A, and MRC1 in the rat model of pulmonary hypertension.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions:\u003c/strong\u003e Following PAH, ACTR2, CCDC88A, and MRC1 might simultaneously upregulating glycolysis and lactylation activities in macrophages and monocytes and further contribute PAH progression.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e","manuscriptTitle":"Machine learning algorithms integrate bulk and single-cell RNA data to reveal the crosstalk and heterogeneity of Glycolysis and Lactylation activity following Pulmonary Arterial Hypertension","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-25 14:58:54","doi":"10.21203/rs.3.rs-6229513/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-08-28T09:00:05+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-27T07:56:22+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"143845771530164996056189106574766546353","date":"2025-08-13T10:50:03+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-03T16:30:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"287484371739133569719908376979451463350","date":"2025-04-10T00:37:59+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-04-04T12:19:22+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-04-04T12:16:14+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-03-27T04:31:41+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-03-25T11:36:36+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-03-14T23:35:02+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"83b214bb-6e81-4490-8eec-6965ef3358d8","owner":[],"postedDate":"April 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":46939317,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":46939318,"name":"Biological sciences/Genetics"},{"id":46939319,"name":"Biological sciences/Microbiology"}],"tags":[],"updatedAt":"2026-05-20T06:41:03+00:00","versionOfRecord":[],"versionCreatedAt":"2025-04-25 14:58:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6229513","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6229513","identity":"rs-6229513","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}