EMP1+ hepatic stellate cells drive hepatic fibrosis progression to hepatocellular carcinoma and predict prognosis

EMP1+ hepatic stellate cells drive hepatic fibrosis progression to hepatocellular carcinoma and predict prognosis | 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 Research Article EMP1+ hepatic stellate cells drive hepatic fibrosis progression to hepatocellular carcinoma and predict prognosis Tongyu Lu, Jie You, Yihuan Huang, Chenhao Jiang, Jiaqi Xiao, Jiebin Zhang, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7278297/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 01 Dec, 2025 Read the published version in Journal of Translational Medicine → Version 1 posted 4 You are reading this latest preprint version Abstract Background Hepatic fibrosis is a pathological response to chronic liver injury that results in accumulation of extracellular matrix proteins leading to fibrous scarring, which can further lead to liver failure and hepatocellular carcinoma (HCC) Although several clinical approaches have been applied to the diagnosis and treatment of hepatic fibrosis and HCC, the clinical prognosis and precision of targeted therapies still face great challenges. Methods In this study, we integrated single-cell sequencing analysis and bulk sequencing analysis to identify genes, cellular subpopulations, and signalling pathways that are closely related to and highly expressed in hepatic fibrosis and HCC. On this basis, clinical prediction models and prognostic genes were constructed and validated by combining single-cell analysis with bulk differential gene analysis in the TCGA database, using 101 machine learning approaches, combined with survival analysis tools, and making full use of clinical data. In addition, the expression heterogeneity of core prognostic genes and their correlation with prognostic outcomes were explored in depth, and new targeted therapeutic modalities were sought with the help of comprehensive and systematic network pharmacological analyses to identify drugs that can target core prognostic genes. Results We identified 45 HSC-associated pathogenic genes and an EMP1 + HSC subpopulation, along with their regulatory signaling pathways linked to energy metabolism, cell adhesion, and extracellular matrix organization. These pathways were found to contribute to hepatic fibrosis and HCC progression. Subsequently, we validated four core prognostic genes (NPY1R, CTHRC1, IGFBP3, and ADH1B) and analyzed the heterogeneity of their expression patterns, demonstrating their correlation with hepatic fibrosis progression and HCC prognosis. Finally, through a systematic screening of bioactive compounds from traditional Chinese medicine (TCM) with potential anti-liver disease effects, we determined that Salvia miltiorrhiza(Danshen) specifically interacts with these core prognostic targets, offering a novel therapeutic strategy for hepatic fibrosis and HCC. Conclusion This integrative study establishes EMP1 as a reliable biomarker for activated HSCs and identifies four core prognostic genes (NPY1R, CTHRC1, IGFBP3, and ADH1B) that play critical roles in the fibrosis-to-HCC progression and demonstrate significant clinical relevance to long-term patient outcomes. Our findings provide novel mechanistic insights into hepatic fibrogenesis and HCC development, while simultaneously revealing Salvia miltiorrhiza (Danshen) as a promising therapeutic agent targeting these key molecular pathways. These discoveries offer a dual advancement in both diagnostic precision and treatment strategy for hepatic fibrosis and HCC. EMP1 Hepatic fibrosis Hepatocellular carcinoma Prediction model Treatment Bioinformatics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 2 Introduction Hepatic fibrosis (HF) and hepatocellular carcinoma (HCC) represent two pathologically interconnected stages in the progression of chronic liver disease. Hepatic fibrosis is a wound-healing response to chronic liver injury characterized by excessive deposition of extracellular matrix (ECM) proteins 1 , 2 , which progressively disrupts hepatic architecture and function, culminating in cirrhosis and HCC development 3 – 5 . Globally, liver diseases account for approximately 2 million annual deaths, with cirrhosis responsible for 50% of these cases 6 ; As the fourth leading cause of cancer mortality worldwide, HCC demonstrates a dismal 5-year survival rate of merely 20% 7,8 . Notably, hepatitis B virus (HBV)-associated HCC predominates in high-prevalence regions such as China 9 , 10 ; where advanced hepatic fibrosis is prevalent among chronic HBV patients 11 . These clinical associations highlight the critical need to investigate HBV-induced hepatic fibrosis and its malignant transformation. Current diagnostic modalities for HCC face significant limitations. Serum biomarker tests, particularly alpha-fetoprotein (AFP), exhibit suboptimal sensitivity (40–60%). 12,13 . while conventional imaging techniques (including ultrasound and PET/CT) frequently fail to detect early-stage HCC within fibrotic liver tissue 14 , 15 . In addition, therapeutic challenges persist across disease stages. Although systemic therapies (encompassing targeted therapy, immunotherapy, and chemotherapy) are clinically established, they yield unsatisfactory outcomes in advanced HCC 16 , 17 . For cirrhotic patients without overt malignancy, conventional anti-fibrotic agents like pirfenidone demonstrate inadequate specificity for pathogenic cell subpopulations. Moreover, existing research has insufficiently addressed disease-driving genes and signaling pathways in specific cellular populations, hindering effective prevention of fibrotic progression to HCC 18 – 21 . Consequently, there is an urgent need to identify core molecular targets governing this malignant transformation. The activation of hepatic stellate cells (HSCs) constitutes the central pathogenic mechanism underlying hepatic fibrosis, as these cells serve as the predominant producers of extracellular matrix (ECM) 1 , 2 , 22 . Quiescent HSCs typically express desmin and glial fibrillary acidic protein (GFAP), whereas activated HSCs undergo profound phenotypic changes characterized by upregulated fibrogenic gene expression and enhanced proliferative capacity - the key drivers of excessive ECM deposition and pro-fibrotic cytokine release in hepatic fibrosis 23 – 25 . However, emerging evidence reveals remarkable heterogeneity within the HSC population, and the distinct contributions of its functional subpopulations to disease progression remain poorly defined. Notably, epithelial membrane protein 1 (EMP1) has been identified as a molecular hallmark of hepatic fibrosis 26 . As a membrane surface protein abundantly expressed in endothelial lineage cells, EMP1 plays pivotal roles in both fibrotic and malignant processes 27 , 28 . Intriguingly, EMP1 expression is also detected in HSCs, yet the precise biological functions of the EMP1 + HSC subpopulation and its mechanistic involvement in HCC pathogenesis - particularly its clinical relevance to patient prognosis - await systematic investigation. Therefore, elucidating the unique role of the EMP1 + HSCs subset in the progression from hepatic fibrosis to HCC is crucial for revealing the cellular mechanisms of disease progression and discovering new prognostic intervention targets. Clinical guidelines have indicated that the degree of hepatic fibrosis should be considered an independent risk factor for treatment decisions and prognosis in HCC. 29 , 30 Consequently, comprehensive exploration of EMP1-HSC interactions in chronic liver disease may yield innovative targets for predicting and treating hepatic fibrosis and HCC. The aim of this study was to employ single-cell transcriptomic analysis to systematically integrate disease spectrum data spanning from healthy controls (CON), grade 1 hepatitis B (G1), grade 2 hepatitis B with fibrosis (G2_HF), and HCC patients, with particular emphasis on the EMP1 + HSC subpopulation. We characterized the differentially expressed genes (DEGs) between EMP1 + HSCs and EMP1-HSCs, and elucidated their potential contributions to fibrogenesis through pathway enrichment and cell-cell communication analyses. Using machine learning approaches combined with survival analysis and clinical validation, we developed robust prognostic models that identified four core prognostic genes: NPY1R (Neuropeptide Y Receptor Type 1), CTHRC1 (Collagen Triple Helix Repeat-Containing Protein 1), IGFBP3 (Insulin-like Growth Factor Binding Protein 3), and ADH1B (Alcohol Dehydrogenase 1B). These analyses revealed significant expression heterogeneity between HCC tissues and EMP1 + HSCs, and established their clinical correlations with chronic liver disease progression. Through comprehensive network pharmacology screening of traditional Chinese medicine databases (including TCMSP), we evaluated numerous herbal formulations such as Paeonia lactiflora-Glycyrrhiza glabra soup and Fuzheng Water-Repelling Formula. This systematic approach identified Salvia miltiorrhiza (Danshen) as the most promising therapeutic candidate, demonstrating effective binding to all core prognostic targets (IGFBP3, CTHRC1, NPY1R, and ADH1B). Further investigation of its bioactive components, combined with in-depth analysis of HBV-related fibrotic progression, provides novel mechanistic insights and potential therapeutic targets for early intervention in hepatic fibrosis and HCC. 3 Materials and Methods In this study, we first constructed a single-cell transcriptional profile of the hepatic fibrotic progression continuum, focusing on the expression of HSC-related pathogenic genes, and performed intercellular communication analysis, trajectory analysis, and transcription factor regulatory network analysis. In addition, we utilized TCGA and GEO databases to validate and demonstrate that HSC-related pathogenic genes exhibit specificity in EMP1+HSCs and can serve as biological markers for both fibrotic progression to HCC and HCC prognosis. 3.1 Data sources and pre-processing Single-nucleus RNA sequencing data for hepatic fibrosis were obtained from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). We analyzed the hepatitis B progression continuum using the following datasets: Healthy controls (CON, n=5) from GSE136103; Hepatitis G1 stage (G1, n=1) from GSE186343; Hepatitis G2 stage with hepatic fibrosis (G2_HF, n=5) from GSE186343; Hepatocellular carcinoma patients (HCC, n=7) from GSE202642. The cell selection criteria were as follows: a) 300-10,000 cells per sample; b) >250 genes expressed per cell; c) Genes detected in ≥3 cells; d) <20% mitochondrial RNA content per cell 31 . Using the Seurat R package, we ultimately retained 75,143 high-quality cells for downstream analyses. Hepatocellular carcinoma bulk-RNA sequencing data and corresponding clinical data were obtained from the TCGA database (https://www.cancer.gov/ccg/research/genome-sequencing/tcga). The TCGA-LIHC dataset comprised 424 samples, including 50 normal samples and 374 tumor samples (368 with available clinical data).We identified differentially expressed genes (DEGs) using DESeq2 with thresholds of |log2FoldChange| >1.5 and adjusted p-value <0.05 31,32 . For validation, we obtained additional datasets (GSE16757 [n=100] and GSE43619 [n=88]) from GEO, totaling 188 HCC tumor samples with clinical information. 3.2 Data quality control After filtering the data for integration, we normalized the expression values using the "NormalizeData" function from the Seurat package. This normalization procedure involved: 1. Multiplying each gene's expression by the total gene expression per cell. 2.Scaling by a factor of 10,000. 3.Applying natural logarithmic transformation after adding a pseudocount of 1 to avoid taking the logarithm of zero. 3.3 Dimensionality Reduction and Cell Clustering Since each gene represents a distinct dimension in the sample, high-dimensional data visualization was challenging. We therefore employed dimensionality reduction techniques to represent the underlying data structure in reduced dimensions 31 . Using the Seurat package, we first applied the "RunPCA" function to reduce dimensionality based on highly variable genes. Subsequently, we identified 2,000 integration anchors through the "Find Integration Anchors" function (RunHarmony implementation), which effectively aligned corresponding cell types across datasets while mitigating batch effects. 3.4 Cell annotation, DEG and marker gene identification The SingleR package enables correlation analysis between single-cell gene expression profiles and reference cell type signatures at single-cell resolution. In our analysis, we employed the package to calculate expression correlations using highly variable genes (HVGs), iteratively eliminating the weakest correlations for each cell type to determine definitive cell identities 33 . For annotation, we utilized the "ref_Human_all" reference database within SingleR to facilitate manual cell type classification. We subsequently identified differentially expressed genes (DEGs) using the "FindAllMarkers" function in Seurat, applying thresholds of |log2FoldChange| > 2 and adjusted p-value < 0.05. The same function was employed to detect cell type-specific marker genes for downstream analyses. 3.5 GO and KEGG Enrichment Analysis Gene Ontology (GO) functional enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed using both the DAVID and Metascape databases. The enrichment results were subsequently visualized using the ggplot2 package in R. 3.6 Pseudotime Trajectory Analysis Cells undergo dynamic transitions between states, accompanied by concomitant changes in gene expression and functional phenotype 33 . Using the Monocle 2 package, we performed pseudotime trajectory analysis to reconstruct cellular differentiation processes by positioning cells along developmental trajectories based on their gene expression profiles 34 . The "plot_cell_trajectory" function was employed to order cells by pseudotemporal progression, while the "BEAM" function identified branch-dependent genes, with results visualized using the "plot_genes_branched_heatmap" function. 3.7 Cell-to-Cell Communication analysis Cell surface ligand-receptor interactions mediate crucial intercellular communication in biological processes. We employed the CellChat software package to construct cellular communication networks by integrating ligand-receptor interactions with their associated factors. This approach simulated intercellular communication patterns by leveraging gene expression profiles of ligands and receptors across different cell types, thereby inferring potential interactions and elucidating rich ligand-receptor interplay between cell population 35 . 3.8 GSVA and Metabolic Pathway Scoring Single-sample gene set enrichment analysis (ssGSEA) is a method that quantifies the enrichment score of specific gene sets in individual samples, where the ssGSEA score reflects the activation status of genomic systems. In this study, we obtained metabolic pathway scores for each cell subpopulation using ssGSEA implemented in the GSVA R package 36 . To identify metabolic differences between HSC subpopulations, we selected genes showing strong population specificity (adjusted p 1.5, pct.1 > 0.5, and pct.2 < 0.5) and calculated metabolic activity scores using the AUCell package 37 . The UMAP visualization colored by AUC scores revealed metabolically active cell subpopulations, while differential gene expression analysis between active and inactive subgroups was performed using the FindAllMarkers function with default parameters (avg_log2FC > 1). 3.9 Identification and Functional Analysis of HSC-Related Pathogenic Genes We analyzed differential gene expression between normal and tumor samples from the TCGA database using stringent thresholds (|logFC|>1.5 and p.adj < 0.05). By intersecting these bulk RNA-seq DEGs with (1) DEGs of HSCs from snRNA-seq data (G2_HF vs CON) and (2) HSC marker genes, we identified HSC-related pathogenic genes (HPGs) implicated in hepatic fibrosis across transcriptomic levels. Subsequent GO enrichment analysis revealed HPG functions, while pathway activity scores computed using the AUCell package enabled comparative assessment across clinical stages (CON, G1, G2_HF, HCC) and between EMP1+HSC and EMP1-HSC subpopulations 37 . 3.10 Construction of Machine Learning-Based Prognostic Models To validate the prognostic role of HSC-related pathogenic genes (HPGs), we constructed 101 machine learning prognostic models following established methodology 31 . The analysis proceeded through four key phases: 1. Dataset preparation and model construction: The TCGA-LIHC dataset (n=368) served as the training set, while GSE16757 (n=100) and GSE43619 (n=88) functioned as external validation sets. We implemented ten machine learning algorithms: Lasso, Ridge, Stepwise Cox, CoxBoost, Random Survival Forest (RSF), Elastic Net (Enet), Partial Least Squares Regression for Cox (plsRcox), Supervised Principal Components (SuperPC), Generalized Boosted Regression Modeling (GBM), and Survival Support Vector Machine (Survival-svm). Through ten-fold cross-validation, we developed 101 algorithm combinations for variable selection and model building. 2. Model evaluation and selection: All models were evaluated on both training and validation sets. We calculated concordance indices (C-indices) for each model across datasets, ranking them by average C-index to identify the most robust and clinically significant algorithm combination. 3. Prognostic validation: The optimal model (highest mean C-index) identified HCC-related prognostic genes (HCC-PGs) used to stratify patients into high- and low-risk groups based on median risk scores. We assessed prognostic significance through Kaplan-Meier analysis (survminer package; log-rank test, p<0.05) and evaluated predictive accuracy using time-dependent ROC curves (timeROC package). The HCC-PGs' predictive performance was compared against other clinical characteristics through AUC comparisons. 4. Biological validation: The HCC-PGs were analyzed using the AUCell package37 to calculate activity scores across HSC subpopulations (CON, G1, G2_HF, HCC) and between EMP1+HSC and EMP1-HSC subsets. 3.11 Cox Regression Analysis and Nomogram Construction To further evaluate the clinical relevance of HCC prognostic genes (HCC-PGs), we conducted comprehensive analyses through two main approaches: 1. Prognostic factor analysis: We examined the correlation between HCC-PG risk scores and key clinical characteristics (age, gender, stage, T/N/M classification, and tumor grade) in the TCGA-LIHC (n=368), GSE16757 (n=100), and GSE43619 (n=88) datasets. Univariate and multivariate Cox regression analyses were performed to determine whether these risk scores served as independent prognostic factors, with results visualized through forest plots. Genes demonstrating statistical significance (p < 0.05) in these analyses were designated as Core Prognostic Genes (CPGs). 2. Nomogram development: To enhance prognostic precision, we constructed a nomogram integrating HCC-PG risk scores with clinical characteristics using the "rms" R package. This predictive tool estimates 1- and 5-year survival probabilities by combining CPG profiles with standard clinicopathological parameters. 3.12 Drug Screening and Molecular Docking Analysis 1. Traditional Chinese medicine screening: We systematically screened herbal components from Paeoniae Alba-Glycyrrhiza glabra Tang and Fu Zheng Yi Shui Fang formulations through the TCMSP database (https://www.tcmsp-e.com/tcmsp.php; accessed January 24, 2025), ultimately identifying Salvia miltiorrhiza as the candidate therapeutic agent. Active compounds were selected based on pharmacokinetic parameters (oral bioavailability ≥30% and drug-likeness ≥0.18). Potential targets (n=7,727) were obtained from PharmMapper (http://lilab-ecust.cn/pharmmapper/index.html), CTD (http://ctdbase.org/) and Swiss Target Prediction (http://www.swisstargetprediction.ch/). 2. Network pharmacology analysis: Among the identified targets, 45 HSC-related pathogenic genes were recognized as potential therapeutic targets for hepatic fibrosis treatment. Using Cytoscape software (https://cytoscape.org/), we constructed a comprehensive drug-component-target-disease network, highlighting the top 10 active ingredients with the most target intersections. 3. Molecular docking validation: We analyzed interactions between Salvia miltiorrhiza bioactive compounds (Salvianolic acid B, Tanshinone IIA, etc.) and core prognostic genes (NPY1R [PDB:5ZBH], CTHRC1 [UniProt:Q96CG8], IGFBP3 [PDB:7WRQ], ADH1B [PDB:1HSZ]). Protein structures were obtained from RCSB PDB (https://www.rcsb.org/) and AlphaFold DB (https://alphafold.ebi.ac.uk/), while compound 3D structures were retrieved from TCMSP (accessed February 14, 2025). Molecular docking was performed using AutoDockTools 4.2.6 (accessed February 16, 2025) after preparing ligands (Chem3D 19.0) and receptors (PyMOL 2.6.0). Binding conformations with energies ≤-5 kcal/mol were considered significant and visualized using PyMOL. 3.13 Statistical Analysis All statistical analyses were performed using R software (version 4.1.1, accessed October 6, 2024). We employed the Student's t-test for comparing normally distributed continuous variables, while the Wilcoxon rank-sum test was used for non-parametric data comparisons. Survival differences between high-risk and low-risk groups were evaluated through Kaplan-Meier analysis with log-rank testing. A threshold of p < 0.05 was established for statistical significance throughout our study. 4 Results 4.1 Overview of single-cell mapping in the hepatitis B disease process We analyzed 18 single-nucleus RNA-seq samples from GSE136103 and GSE186343 datasets, encompassing healthy controls (CON, n = 5), hepatitis G1 stage (G1, n = 1), hepatitis G2 stage with hepatic fibrosis (G2_HF, n = 5), and hepatocellular carcinoma patients (HCC, n = 7). After quality control filtering, 75,143 high-quality cells were retained for analysis. Data integration using anchor-based alignment effectively removed batch effects, followed by normalization, pooling, and PCA dimensionality reduction (retaining top 20 PCs). UMAP visualization revealed 37 distinct clusters (Fig. 2A), which were annotated into 16 cell types (Fig. 2B): Kupffer cells (Kupffer, CD163+), macrophages (Macrophages, S100A9+), NK/T cells (NK_T, CD3E+), innate lymphoid cells (ILCs, CXCL8+), dendritic cells (DCs, CLEC10A+), NK cells (NK, KLRF1+), B cells (B_cells, CD79A+), plasma cells (Plasma, IGHM+), hepatic stellate cells (HSCs, ACTA2+), dividing cells (Dividing, TOP2A+), T cells (T_cells, BATF+), hepatocytes (Hepatocytes, ALDOB+), vascular endothelial cells (VECs, VWF+), cholangiocytes (Cholangiocytes, KRT7+), liver sinusoidal endothelial cells (LSECs, STAB2+), and hepatocellular carcinoma cells (HCCs, GPC3+). Figure 2C demonstrated the transcriptomic profiles across disease stages, while Fig. 2D displayed the cellular composition distribution. Marker gene expression patterns (Fig. 2E-F) and intercellular DEG analysis (Fig. 2G-H) validated annotation accuracy. Cellular proportion analysis (Fig. 2I) showed dominant populations of NK_T, LSECs and VECs, with relatively fewer HSCs, Hepatocytes, and Cholangiocytes. Quantitative analysis determined the percentage distribution of cell types across disease stages. 4.2 Single-cell Transcriptomic Profiling of Hepatic Stellate Cells Hepatic stellate cells (HSCs) are recognized as central mediators of hepatic fibrosis progression 1 . The snRNA-seq profiles showed the presence of a large number of hepatic stellate cells (HSCs) in all samples. Our snRNA-seq analysis identified 2,237 HSCs across all samples for subsequent clustering. Re-clustering revealed seven distinct HSC subpopulations (Fig. 3A), with disease-stage-specific transcriptomic profiles (Fig. 3B).Given EMP1's established role as a hepatic fibrosis biomarker 26 . we stratified HSCs into EMP1- HSCs (n = 1,971) and EMP1 + HSCs (n = 266) subsets based on EMP1 expression (Fig. 3C). Comparative analysis demonstrated differential distribution patterns across disease stages (CON, G1, G2_HF, HCC) (Fig. 3D). Differential gene expression analysis between EMP1 + and EMP1- HSCs identified distinct molecular signatures (Fig. 3E), visualized through UMAP embeddings (Fig. 3F) and bubble plots (Fig. 3G). Activation state marker analysis showed EMP1 + HSCs exhibited profiles resembling activated HSCs (COL1A1+, etc.), though paradoxically lower ACTA2 expression versus EMP1-HSCs (Fig. 3H) - aligning with recent evidence questioning ACTA2's reliability as an activation marker 38 . Finally, quantification revealed progressive expansion of EMP1 + HSCs during disease progression (CON, G1, G2_HF, HCC), significantly exceeding healthy control proportions (Fig. 3I), implicating this subset in HBV-related fibrogenesis. 4.3 Enrichment Analysis of Cell Subpopulations Figure 4A demonstrated significant upregulation of differential genes in hepatic stellate cells (HSCs) compared to other cell types, with higher logFC values indicating their predominant functional role in chronic hepatitis B-associated hepatic fibrosis. GO and KEGG enrichment analyses of HSC marker genes revealed key functional pathways (Fig. 4B-D): KEGG analysis identified enrichment in cytoskeletal regulation in muscle cells, focal adhesion, non-alcoholic fatty liver disease, chemical carcinogenesis-reactive oxygen species, and oxidative phosphorylation, while GO analysis showed extracellular structure organization, extracellular matrix (ECM) organization, and cell-substrate adhesion as the main biological processes (BP), collagen-containing extracellular matrix and cell-substrate junctions as primary cellular components (CC).The KEGG and GO enriched pathways described above represent the primary molecular mechanisms through which HSCs contribute to hepatic fibrosis 39–42 . And ECM structural constituents and NAD(P)H dehydrogenase (quinone) activity as major molecular functions (MF), suggesting that metabolic activity has an impact on HSC function. To further explore the role of EMP1 in HSCs and hepatic fibrosis during the course of chronic hepatitis B, we identified differential genes between the two types of HSCs (EMP1 + HSCs versus EMP1-HSCs) (Fig. 4E), which showed an overall increase in the gene expression level of EMP1 + HSCs relative to that of EMP1-HSCs, with 710 upregulated and 46 downregulated genes identified. GO enrichment of differentially expressed genes (Fig. 4F) showing similar extracellular organization pathway enrichment patterns to bulk HSCs, functionally validating EMP1 + HSCs as the primary effector subpopulation in HSC-mediated fibrotic processes. 4.4 Pseudotime Trajectory Analysis of HSC Subpopulations Pseudotime trajectory analysis, which reconstructs cellular developmental pathways based on temporal gene expression patterns, was performed on hepatic stellate cells (HSCs) from our snRNA-seq data. The analysis revealed three distinct branches of HSC subtypes (Fig. 5A) and three corresponding differentiation states (Fig. 5C). The developmental progression, illustrated in Figs. 5B-D, showed a clear transition from state1 to state3, characterized by a gradual shift from darker-colored to lighter-colored cells. Within this trajectory, EMP1 + HSCs were predominantly localized to intermediate developmental stages, while EMP1-HSCs were distributed throughout. Gene expression patterns were categorized into three temporally-regulated modules (Fig. 5E). The green module, associated with copper ion response and processing, suggested a potential role for copper detoxification in HSC activation and fibrotic responses. The red module demonstrated significant enrichment for extracellular matrix formation and maintenance pathways, consistent with the established role of HSCs in differentiating into collagen-producing myofibroblasts during hepatic fibrosis. The blue module, linked to energy metabolism regulation and cellular adhesion, indicated additional functional aspects of HSCs in liver injury response, development and regeneration through metabolic support and intercellular communication. Detailed examination of EMP1 and eight key developmentally-regulated genes (Fig. 5F) confirmed the mid-trajectory predominance of EMP1 + HSCs and pan-trajectory distribution of EMP1-HSCs, corroborating the findings in Fig. 5D. COL1A1, COL1A2, COL3A1 and TIMP1 expression progressively declined during development, whereas APOD, CFD, DCN and MFAP5 exhibited an initial increase followed by terminal stage decrease. Notably, all eight genes showed expression patterns tightly associated with EMP1 + HSCs and were nearly absent in terminal stage EMP1-HSCs, particularly evident for APOD, CFD, DCN and MFAP5. Importantly, while EMP1 expression served to distinguish HSC subpopulations, it did not appear directly involved in HSC proliferation or differentiation processes. And the identification of collagen genes (COL1A1, COL1A2, and COL3A1) - known critical mediators of HSC activation and function 42,41 , within this analysis further substantiates EMP1 + HSCs as the primary functional subpopulation in hepatic fibrogenesis. 4.5 Analysis of cell-to-cell communication in HSCs To further explore the effect of EMP1 + HSCs on other cells during chronic hepatitis B progression and their role in hepatic fibrosis, we performed cell-to-cell communication analysis by tracking ligand-receptor interactions using our snRNA-seq data. First, Figs. 6A-B demonstrated that in the chronic hepatitis B process, the number and strength of cellular communication between EMP1 + HSCs and other major cells were significantly stronger in the G2_HF group than in the CON group. Figure 6B further revealed the specific distribution of the cellular communication parameters across cell types in both the CON and G2_HF groups, showing that EMP1 + HSCs served as the most important source of ligand signaling in the G2_HF group and exhibited the most significant enhancement in cellular communication relative to the CON group. Figure 6C identified CD46, VTN, LAMININ and ANNEXIN as the predominant mediators of enhanced cellular communication in the G2_HF group (Fig. 6C), with LAMININ and ANNEXIN representing extracellular matrix-associated signaling pathways. Notably, analysis of the enhanced communication network showed that LSECs, VECs and Macrophages served as the primary signal-receiving cells, whereas EMP1 + HSCs emerged as the dominant signal-emitting population, with their contribution far exceeding that of EMP1-HSCs (Figs. 6D-E). Strikingly, Fig. 6F demonstrated that within the LAMININ signaling pathway, EMP1 + HSCs were the most active ligand signal emitters, where the LAMB2-CD44 pair constituted the single most influential receptor-ligand interaction (Fig. 6G). The specific communication patterns mediated by LAMB2-CD44 were visualized across cell types in both CON and G2_HF groups (Fig. 6H), while Fig. 6I provided a comprehensive comparison of the expression levels of all LAMININ pathway molecules between these two groups. 4.6 HSC metabolic pathway analysis In the pseudo-time analysis of HSCs (Fig. 5E), the processes in the blue module were found to be associated with the regulation of energy metabolism and cell adhesion, indicating that hepatic stellate cells not only participate in extracellular matrix organization but may also support their activation through energy metabolism and maintain intercellular connections via adhesion regulation. Given the liver's crucial role in systemic energy metabolism, we investigated the metabolic reprogramming of NPCs in hepatic disease states. Previous studies have shown that the hepatic immune response involves enhanced glucose metabolism in immunoreactive cells 43 . HSCs show a particularly high sensitivity, and they play an important role in immune metabolism by maintaining liver function and responding to injury 44 . To characterize metabolic alterations, we first selected the top 1,000 highly variable genes in each cell population and performed cell-to-cell similarity assessment. This analysis revealed that HSCs displayed the closest expression patterns to VECs and LSECs across all groups (CON, G1, G2_HF, HCC) (Fig. 7A). Subsequently, we quantified metabolic pathway activities to evaluate NPC reprogramming in different disease stages, focusing on carbohydrate, energy, and lipid metabolism. GSVA results demonstrated that HSCs were the more metabolically active population in all groups (Fig. 7B), with AUC analysis further identifying oxidative phosphorylation as the predominant metabolic pathway in both HSC subpopulations (Fig. 7C). Despite an overall reduction in HSC metabolic activities in the G2_HF group (Fig. 7B), we observed elevated oxidative phosphorylation activity in both HSC subpopulations compared to other groups, with EMP1 + HSCs exhibiting significantly higher activity than EMP1-HSCs (Fig. 7C). 4.7 Identification and Functional Analysis of HSC-Related Pathogenic Genes Differential gene analysis of the TCGA-LIHC dataset identified 821 DEGs (Fig. 8A), which were intersected with 5505 HSC marker genes in the single-cell samples (Fig. 4A) and 2882 intergroup DEGs (G2_HF vs. CON) in HSCs (Fig. 8B), yielding 45 overlapping genes (Fig. 8C). The three differential gene sets used to obtain the intersection genes are provided in Supplementary File S1. These genes were differentially expressed not only in HSCs of the G2_HF group compared to the CON group but also showed significant differences between hepatocellular carcinoma patients and healthy individuals. We hypothesized that these 45 genes might serve as key regulators in hepatocellular carcinoma prognosis through HSCs, with their altered expression contributing to the progression from hepatic fibrosis to hepatocellular carcinoma; hence, they were designated as HSC-related pathogenic genes (HPGs). To investigate the functional roles of HPGs, GO analysis was performed on these genes (Fig. 8D). The results revealed that epithelial cell proliferation was the most statistically significant term in the BP category, while collagen-containing extracellular matrix and extracellular matrix structural constituent were the most significant in the CC and MF categories, respectively (Fig. 8E). Epithelial cell proliferation can drive epithelial-derived cells (e.g., hepatocytes, cholangiocytes, or hepatocellular carcinoma cells) into the cell cycle. Activated HSCs secrete abundant HGF, EGF, and TGF-α, which bind to MET/EGFR on epithelial cell membranes, subsequently activating the PI3K-AKT or RAS-MAPK cascades and promoting hepatocyte/hepatocellular carcinoma cell proliferation 45 . Additionally, the collagen-containing extracellular matrix and extracellular matrix structural constituent are critical pathways in hepatic fibrosis, suggesting that HSCs may contribute not only to fibrogenesis but also to hepatocellular carcinoma progression by enhancing tumor cell proliferation and exacerbating fibrosis. To determine the predominant HSC subpopulation involved in hepatocellular carcinoma prognosis, we calculated the AUC activity scores of these three GO pathways in HSCs and visualized them using UMAP plots (Fig. 8F, 8H). Further analysis revealed that EMP1 + HSCs exhibited significantly higher activity in these pathways compared to EMP1-HSCs, indicating that EMP1 + HSCs represent the major HSC subset implicated in hepatocellular carcinoma prognosis (Fig. 8G). Finally, the three GO pathways displayed elevated activity in the G2_HF group compared to the CON group, with the highest AUC scores observed in the HCC group, further supporting the role of HSCs in the transition from hepatic fibrosis to hepatocellular carcinoma and their involvement in disease prognosis (Fig. 8I). 4.8 101 machine learning-based prognostic model construction and functional characterization of HCC prognostic genes The network diagram revealed the gene composition of the 45 HPGs (Fig. 9A). After normalizing the expression patterns of these genes, we employed 101 machine learning algorithms to calculate their risk scores, using the TCGA-LIHC dataset (n = 368) as the training set and the GSE16757 (n = 100) and GSE43619 (n = 88) datasets as validation sets (Fig. 9B). The StepCox[forward] + Ridge model demonstrated superior predictive performance, through which we identified 10 key genes designated as HCC Prognostic Genes (HCC-PGs), highlighted in black in the network diagram (Fig. 9A). Comparative analysis of HCC-PG expression in HSCs from G2_HF and HCC groups revealed that six genes (ABCA8, INMT, ANXA2, OLFML2B, ADH1B, and NPY1R) exhibited significant differential expression (Fig. 9C), suggesting their potential dual role in both HCC prognosis and the transition from hepatic fibrosis to hepatocellular carcinoma. Spatial distribution analysis of the 10 HCC-PGs across 16 cell subpopulations demonstrated their predominant expression in HSCs (Fig. 9D-E). Notably, EMP1 + HSCs showed significantly higher expression levels of these genes compared to EMP1-HSCs, indicating potential functional divergence between these HSC subtypes. Subsequent AUC activity scoring of the integrated 10-gene prognostic signature confirmed significantly enhanced activity in EMP1 + HSCs relative to EMP1-HSCs (Fig. 9F-G), further supporting the pivotal role of EMP1 + HSCs in HCC progression. 4.9 Validation of prognostic model predictive efficacy Among the 101 machine learning prognostic models evaluated, the StepCox[forward] + Ridge model demonstrated superior predictive performance (Fig. 9B). Kaplan-Meier survival analysis revealed significantly worse outcomes in high-risk groups across both training and test sets, with statistically significant separation of survival curves (Fig. 10A). Time-dependent ROC curve analysis confirmed the robust predictive capacity of our prognostic signature, with AUC values exceeding 0.7 for 1-year and 3-year survival predictions in both cohorts, while maintaining strong performance (AUC = 0.69) for 5-year survival prediction (Fig. 10B-D). Comprehensive Cox regression analyses incorporating clinical and molecular data identified several significant prognostic factors. Univariate analysis showed that ABCA8, INMT, GADD45B, GHR, CTHRC1, IGFBP3, ADH1B, along with risk scores, were significantly associated with overall survival (Fig. 10E). Subsequent multivariate analysis pinpointed four genes with particularly strong prognostic value: NPY1R (HR [95% CI]: 0.94 [0.90–0.98]), ADH1B (HR [95% CI]: 0.96 [0.92–0.99]), CTHRC1 (HR [95% CI]: 1.04 [1.01–1.08]), and IGFBP3 (HR [95% CI]: 1.09 [1.04–1.14]) (Fig. 10F). Based on these findings, we designated these four genes as core prognostic genes (CPGs), with NPY1R and ADH1B emerging as protective factors, while CTHRC1 and IGFBP3 were identified as risk factors. Further multivariate analysis incorporating clinical parameters revealed that tumor M1 stage and risk stratification maintained independent prognostic significance. Notably, the low-risk group exhibited substantially better outcomes compared to high-risk patients (HR [95% CI]: 0.36 [0.23–0.58]) (Fig. 10G). To facilitate clinical translation, we developed a comprehensive nomogram incorporating the four CPGs and risk scores, which provides individualized 1- to 5-year survival probability estimates, with risk scores emerging as the most dominant contributor to survival prediction (Fig. 10H). 4.10 Identification of the clinical significance of core prognostic genes The screening process of core prognostic genes (CPGs) from HSC-related pathogenic genes was illustrated in Fig. 11A. To validate the clinical relevance of these CPGs, we analyzed the expression patterns of NPY1R, CTHRC1, IGFBP3, and ADH1B in risk-stratified patients from three independent cohorts: TCGA-LIHC (n = 368), GSE16757 (n = 100), and GSE43619 (n = 88). To minimize batch effects, we presented the results from different databases separately (Fig. 11B-C). Our data showed that NPY1R exhibited significant differential expression only in TCGA samples (p < 0.01), but not in the GEO datasets, suggesting its context-dependent prognostic role. CTHRC1 and IGFBP3 demonstrated significantly higher expression in high-risk groups (p < 0.001), while ADH1B was preferentially expressed in low-risk patients (p < 0.01), consistent with their classification as risk and protective factors, respectively, in our previous analysis (Fig. 10F). Notably, ADH1B and NPY1R displayed marked expression differences between G2_HF and HCC groups in HSCs (Fig. 9B), indicating their dual involvement in HCC prognosis and fibrotic-to-HCC progression. 4.11 Drug screening and molecular docking analysis In order to find targeted drugs against HSC-related pathogenic genes (HPGs) to reverse the transformation of cirrhosis to hepatocellular carcinoma and improve prognosis, we systematically screened traditional Chinese medicine compounds from TCMSP and additional drug databases. After evaluating numerous candidates including Paeonia lactiflora, Glycyrrhiza glabra, and Fu Zheng Yi Shui Fang, Salvia miltiorrhiza was selected as the primary study drug. We obtained 7727 potential targets of Salvia miltiorrhiza through PharmMapper, CTD, and Swiss Target Prediction (These targets are listed in Supplementary File S2). The network diagram revealed the top 10 Salvia miltiorrhiza active components most strongly associated with HPGs (Fig. 12A): Salvianolic acid B, Isotanshinone II, Luteolin, Microstegiol, Miltirone II, Salvianolic acid G, Salvilenone I, Salviolone, Tanshinone IIA, and Tanshinone VI. Subsequent molecular docking analysis demonstrated strong binding potential (binding energy < -5 kcal/mol) between these components and proteins encoded by the four core prognostic genes (Fig. 12B). Particularly noteworthy were the exceptionally low binding energies observed between: 1) Salvilenone I and NPY1R, 2) Tanshinone IIA and CTHRC1, 3) Salvianolic acid G and IGFBP3, and 4) Isotanshinone II and ADH1B. Detailed structural analysis showed the molecular configurations of these four promising compounds (Fig. 12C) and their specific binding patterns with corresponding target proteins (Fig. 12D). The docking results illustrated precise amino acid interactions and chemical bond formations for each receptor-ligand pair, confirming the structural basis for their high-affinity binding. 5 Discussion While numerous clinical studies have established hepatic fibrosis as an independent risk factor for hepatocellular carcinoma 29 , 30 , the precise mechanisms linking fibrotic progression to hepatocarcinogenesis remain poorly characterized 46 – 48 . This persistent knowledge gap highlights the urgent need to delineate pathogenic genes and pathways at single-cell resolution, which could facilitate the development of precision therapies and improve clinical outcomes for fibrosis-associated HCC. Through systematic analysis of single-cell transcriptomes across the chronic hepatitis B disease continuum - including healthy controls (CON), grade 1 hepatitis B (G1), grade 2 hepatitis B with fibrosis (G2_HF), and HCC patients - we identified and characterized a novel EMP1-high hepatic stellate cell (HSC) subpopulation (EMP1 + HSCs). Comprehensive bioinformatics analysis revealed that EMP1-associated differentially expressed genes (DEGs) were predominantly enriched in extracellular matrix (ECM) organization pathways, directly implicating this cellular subset in fibrogenesis 9 . Of particular note, the transcriptional signature of EMP1 + HSCs showed striking similarity to activated HSCs (aHSCs) 49 , providing compelling evidence for EMP1 as a novel activation marker. The pseudo-time trajectory analysis revealed that EMP1 + HSCs were predominantly localized to intermediate and late activation stages of HSC development, while extracellular matrix (ECM) formation and maintenance emerged as a critical biological process during HSC differentiation. Through systematic screening, we identified eight key genes demonstrating the most significant expression changes during HSC activation, including COL1A1, COL1A2, and COL3A1, which are core structural components of the ECM. Strikingly, the upregulation of these ECM-related genes was temporally coupled with the emergence of EMP1 + HSCs, providing compelling evidence that EMP1 serves as a novel biomarker for activated HSCs (aHSCs). Single-cell communication analysis further demonstrated that EMP1 + HSCs acted as the dominant signaling hubs in intercellular communication networks, exhibiting significantly stronger signaling activity compared to EMP1-HSCs, particularly in fibrogenic pathways mediated by LAMININ interactions. Taken together, these findings establish that EMP1 + HSCs represent a functionally distinct HSC subpopulation with markedly different biological characteristics from EMP1-HSCs, and confirm their primary role in driving hepatic fibrosis progression. These results strongly support the potential clinical utility of EMP1 as a reliable molecular marker for identifying the activated state of HSCs. Interestingly, our pseudo-time trajectory analysis uncovered the involvement of energy metabolism pathways in HSC differentiation, revealing that HSCs contribute to liver homeostasis not only through extracellular matrix reorganization but also by participating in metabolic support and intercellular connectivity modulation via adhesion molecule regulation. These findings suggest a multifunctional role for HSCs in liver development, regeneration, and injury response. Metabolic pathway analysis demonstrated that HSCs exhibited heightened metabolic activity across all experimental groups compared to other cell types. However, HSCs in G2_HF group showed reduced overall metabolic activity relative to healthy controls (CON), while EMP1 + HSCs displayed significantly elevated oxidative phosphorylation capacity compared to EMP1- HSCs. These metabolic profiling results provide compelling evidence that distinct HSC subpopulations possess characteristic metabolic signatures corresponding to their functional states. As established in previous studies, hepatic fibrosis represents a chronic liver pathology whose malignant progression frequently culminates in cirrhosis and hepatocellular carcinoma (HCC) 5 , 50 , 51 . To elucidate the functional contribution of EMP1-associated differentially expressed genes (DEGs) in HSCs during fibrosis-induced hepatocarcinogenesis, we performed systematic intersection analysis between TCGA-LIHC bulk transcriptomic data and single-cell datasets, yielding 45 overlapping genes designated as HSC-related pathogenic genes (HPGs). Employing 101 machine learning algorithms, we constructed a robust prognostic model based on these HPGs, through which core prognostic genes were identified via optimal model selection and multivariate COX regression analysis. Notably, NPY1R and ADH1B emerged as significant protective factors, whereas IGFBP3 and CTHRC1 were characterized as detrimental factors in HCC progression. NPY1R, as a member of the G protein-coupled receptor (GPCR) family, is predominantly expressed in the nervous system and vascular smooth muscle cells, with relatively lower expression in HSCs, where it regulates cell proliferation and vasoconstriction 52 . Although NPY1R has been understudied in hepatocellular carcinoma, existing evidence demonstrates that it suppresses tumor cell growth by inhibiting the mitogen-activated protein kinase (MAPK) signaling pathway, with its elevated expression correlating positively with improved survival in advanced HCC patients 53 , 54 , which is in line with the results of this study. However, a recent contradictory study reported that NPY1R overexpression in HCC tissues might facilitate tumor immune evasion and metastasis through activation of pro-survival STAT3/AKT1 pathways 55 , highlighting the need for further investigations to clarify its context-dependent roles in hepatocarcinogenesis. In our single-cell analyses, CTHRC1 exhibited predominant expression in HSCs, particularly showing markedly higher levels in EMP1 + HSCs compared to EMP1-HSCs across both fibrotic and HCC stages. As a secreted glycoprotein primarily localized in the extracellular matrix, CTHRC1 directly contributes to hepatic fibrosis by activating HSCs through autocrine mechanisms and potentiating TGF-β signaling 56 , 57 . Although our study did not detect significant expression differences of CTHRC1 between HCC and G2_HF groups at the bulk level, its sustained high expression in EMP1 + HSCs suggests this subpopulation may drive fibrosis progression toward HCC through persistent ECM remodeling. Although our study did not detect significant expression differences of CTHRC1 between HCC and G2_HF groups at the bulk level, its sustained high expression in EMP1 + HSCs suggests this subpopulation may drive fibrosis progression toward HCC through persistent ECM remodeling 58 . While our TCGA-LIHC data confirmed its decreased expression at bulk level, emerging evidence indicates that in galectin-3-high HCC subtypes, PI3K-AKT-GSK-3β-β-catenin pathway activation can paradoxically upregulate IGFBP3 and vimentin, promoting angiogenesis and epithelial-mesenchymal transition (EMT)-mediated metastasis 59 , 60 . Thus, IGFBP3 may interact with various pathways at different stages of tumor development and lead to different outcomes. Intriguingly, in our single-cell data from G2_HF group, IGFBP3 was primarily expressed by HSCs rather than hepatocytes, consistent with reports that HSC-derived IGFBP3 promotes alcohol-induced steatohepatitis 61 , suggesting its complex, context-dependent roles in liver disease progression that may vary by etiology and cellular origin. ADH1B, the key ethanol-metabolizing enzyme, has well-documented tumor-suppressive effects, with its low expression associated with inflammatory pathway activation, metabolic reprogramming, and poorer HCC prognosis - findings corroborated by our current data showing its protective role 62 . Although mainly expressed in hepatocytes, we observed significant ADH1B expression differences between EMP1 + HSC and EMP1-HSC subpopulations. Given that acetaldehyde (the ethanol metabolite) directly stimulates HSC proliferation 63 , 64 , this differential ADH1B expression may functionally distinguish these HSC subsets. The observed ADH1B expression variations between HCC and G2_HF HSCs further suggest its potential involvement in modulating the fibrosis-to-HCC transition, possibly through immune microenvironment regulation rather than direct fibrogenic effects. Collectively, our findings demonstrate that the four core prognostic genes participate in hepatic fibrogenesis through distinct biological pathways and collectively contribute to the progression from hepatic fibrosis to hepatocellular carcinoma. The differential expression patterns of these genes in EMP1 + HSCs versus EMP1-HSCs not only support EMP1's utility as a reliable marker for HSC activation state, but also suggest its potential as a biomarker for monitoring fibrosis-to-HCC progression. To explore therapeutic strategies targeting HSC-related pathogenic genes (HPGs) for preventing cirrhosis-to-HCC progression, we conducted systematic screening of traditional herbal medicines, ultimately selecting Salvia miltiorrhiza as our primary candidate based on its documented therapeutic effects. Previous studies have established that Salvia miltiorrhiza ameliorates hepatic fibrosis by activating intrahepatic natural killer (NK) cells to eliminate activated HSCs, thereby reducing collagen deposition 65 ; furthermore, Tanshinone IIA, a key active component, has been shown to suppress HCC cell proliferation and invasion through inhibition of the TGF-β/SMAD7-YAP pathway 66 , confirming this herb's dual therapeutic potential for both hepatic fibrosis and HCC. Through comprehensive molecular docking analysis, we identified 10 Salvia miltiorrhiza components showing highest affinity with HPGs, with all binding energies <-5 kcal/mol. These strong molecular interactions provide a structural basis for the herb's potential efficacy in interrupting cirrhosis-to-HCC progression and improving HCC prognosis. 6 Conclusion Through integrative analysis combining single-cell and bulk transcriptomic data with advanced machine learning approaches, this study systematically elucidated the pivotal role of EMP1 + HSCs in hepatic fibrogenesis during chronic hepatitis B progression to hepatocellular carcinoma. By employing differential enrichment analysis, pseudo-time trajectory reconstruction, cell-cell communication profiling, and metabolic activity assessment, we identified 45 HSC-related pathogenic genes (HPGs) and associated pathways that collectively drive Hepatic fibrosis and HCC development, with EMP1 + HSCs emerging as the central cellular mediators of these pathogenic processes. Utilizing 101 machine learning algorithms, we established a robust prognostic model that pinpointed four core prognostic genes (NPY1R, CTHRC1, IGFBP3, and ADH1B), while comprehensively characterizing their expression heterogeneity and clinical relevance in HCC progression. These investigations not only confirmed EMP1 as a reliable biomarker for HSC activation states but also revealed its potential utility in monitoring fibrosis-to-HCC progression. Furthermore, through systematic drug screening and molecular docking analyses, we demonstrated that Salvia miltiorrhiza and its bioactive components exhibit strong binding affinities (<-5 kcal/mol) with the four core prognostic gene products, thereby providing molecular evidence for its therapeutic potential in interrupting cirrhosis-to-HCC progression and improving HCC outcomes. Declarations Ethics approval and consent to participate GEO is a public database. Since our study relies on open-source data, it encounters no ethical concerns or conflicts of interest. Consent for publication All authors agree to submit the article for publication. Competing interests The authors declare no competing interest in this work. Funding This work was supported by the grants from the National Natural Science Foundation of China (82270689, 82460139), the Natural Science Foundation of Guangdong Province (2025A03J3197), Guangzhou Basic and Applied Basic Research Project Co-funded by Municipal Schools (institutes) (2025A03J3197, 2024B03J1382), Guangzhou Key R&D Fields Project(Agricultural and Social Development Science and Technology Special Topic)(2024B03J1382), Natural Science Foundation of Xinjiang Uygur Autonomous Region(2024D01E21), Xinjiang Uygur Autonomous Region People’s Hospital Liver Transplantation Special Program(20240101). Authors' contributions Jie You and Yihuan Huang contributed equally to this study. Acknowledgements We express our gratitude to TCGA and the GEO database for their platforms and the contributors who uploaded significant datasets. Data availability The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. References Kisseleva T, Brenner D. Molecular and cellular mechanisms of liver fibrosis and its regression. Nat Rev Gastroenterol Hepatol. 2021;18(3):151–66. 10.1038/s41575-020-00372-7 . Roehlen N, Crouchet E, Baumert TF. Liver Fibrosis: Mechanistic Concepts and Therapeutic Perspectives. Cells. 2020;9(4). 10.3390/cells9040875 . Campos-Murguía A, Ruiz-Margáin A, González-Regueiro JA, Macías-Rodríguez RU. Clinical assessment and management of liver fibrosis in non-alcoholic fatty liver disease. World J Gastroenterol. 2020;26(39):5919–43. 10.3748/wjg.v26.i39.5919 . Aleknavičiūtė-Valienė G, Banys V. Clinical importance of laboratory biomarkers in liver fibrosis. Biochem Med (Zagreb). 2022;32(3):030501. 10.11613/BM.2022.030501 . Ginès P, Krag A, Abraldes JG, Solà E, Fabrellas N, Kamath PS. Liver cirrhosis. Lancet. 2021;398(10308):1359–76. 10.1016/S0140-6736(21)01374-X . Asrani SK, Devarbhavi H, Eaton J, Kamath PS. Burden of liver diseases in the world. J Hepatol. 2019;70(1):151–71. 10.1016/j.jhep.2018.09.014 . Tapper EB, Parikh ND. Diagnosis and Management of Cirrhosis and Its Complications: A Review. JAMA. 2023;329(18):1589–602. 10.1001/jama.2023.5997 . Chidambaranathan-Reghupaty S, Fisher PB, Sarkar D. Hepatocellular carcinoma (HCC): Epidemiology, etiology and molecular classification. Adv Cancer Res. 2021;149:1–61. 10.1016/bs.acr.2020.10.001 . Bataller R, Brenner DA. Liver fibrosis. J Clin Invest. 2005;115(2):209–18. 10.1172/JCI24282 . Zeng S, Liu Z, Ke B, Zhang Y, Wang Q, Tan S. The non-invasive serum biomarkers contributes to indicate liver fibrosis staging and evaluate the progress of chronic hepatitis B. BMC Infect Dis. 2024;24(1):638. 10.1186/s12879-024-09465-z . Jiang C, Liu J, Li R, et al. Clinicopathologic characteristics of liver inflammation and fibrosis in 310 patients with chronic hepatitis B. Zhong Nan Da Xue Xue Bao Yi Xue Ban. 2023;48(5):698–706. 10.11817/j.issn.1672-7347.2023.220622 . Luo P, Wu S, Yu Y, et al. Current Status and Perspective Biomarkers in AFP Negative HCC: Towards Screening for and Diagnosing Hepatocellular Carcinoma at an Earlier Stage. Pathol Oncol Res. 2020;26(2):599–603. 10.1007/s12253-019-00585-5 . Hanif H, Ali MJ, Susheela AT, et al. Update on the applications and limitations of alpha-fetoprotein for hepatocellular carcinoma. World J Gastroenterol. 2022;28(2):216–29. 10.3748/wjg.v28.i2.216 . Altinmakas E, Taouli B. Abbreviated Liver Magnetic Resonance Imaging Protocols and Applications. Radiol Clin North Am. 2022;60(5):695–703. 10.1016/j.rcl.2022.04.002 . Masokano IB, Liu W, Xie S, Marcellin DFH, Pei Y, Li W. The application of texture quantification in hepatocellular carcinoma using CT and MRI: a review of perspectives and challenges. Cancer Imaging. 2020;20(1):67. 10.1186/s40644-020-00341-y . Foerster F, Gairing SJ, Müller L, Galle PR. NAFLD-driven HCC: Safety and efficacy of current and emerging treatment options. J Hepatol. 2022;76(2):446–57. 10.1016/j.jhep.2021.09.007 . Llovet JM, Pinyol R, Kelley RK, et al. Molecular pathogenesis and systemic therapies for hepatocellular carcinoma. Nat Cancer. 2022;3(4):386–401. 10.1038/s43018-022-00357-2 . Bataller R, North KE, Brenner DA. Genetic polymorphisms and the progression of liver fibrosis: a critical appraisal. Hepatology. 2003;37(3):493–503. 10.1053/jhep.2003.50127 . Toh MR, Wong EYT, Wong SH, et al. Global Epidemiology and Genetics of Hepatocellular Carcinoma. Gastroenterology. 2023;164(5):766–82. 10.1053/j.gastro.2023.01.033 . Shah PA, Patil R, Harrison SA. NAFLD-related hepatocellular carcinoma: The growing challenge. Hepatology. 2023;77(1):323–38. 10.1002/hep.32542 . Babariya H, Gaidhane SA, Acharya S, Kumar S. Pirfenidone as a Cornerstone in the Management of Fibrotic Interstitial Lung Diseases and Its Emerging Applications: A Comprehensive Review. Cureus. 2024;16(9):e70497. 10.7759/cureus.70497 . Friedman SL. Liver fibrosis -- from bench to bedside. J Hepatol. 2003;38(Suppl 1):S38–53. 10.1016/s0168-8278(02)00429-4 . Mederacke I, Hsu CC, Troeger JS, et al. Fate tracing reveals hepatic stellate cells as dominant contributors to liver fibrosis independent of its aetiology. Nat Commun. 2013;4:2823. 10.1038/ncomms3823 . Trivedi P, Wang S, Friedman SL. The Power of Plasticity-Metabolic Regulation of Hepatic Stellate Cells. Cell Metab. 2021;33(2):242–57. 10.1016/j.cmet.2020.10.026 . Bates J, Vijayakumar A, Ghoshal S, et al. Acetyl-CoA carboxylase inhibition disrupts metabolic reprogramming during hepatic stellate cell activation. J Hepatol. 2020;73(4):896–905. 10.1016/j.jhep.2020.04.037 . Chen X, Lv X, Han M, et al. EMP1 as a Potential Biomarker in Liver Fibrosis: A Bioinformatics Analysis. Gastroenterol Res Pract. 2023;2023:2479192. 10.1155/2023/2479192 . Zhang N, Zhu HP, Huang W, et al. Unraveling the structures, functions and mechanisms of epithelial membrane protein family in human cancers. Exp Hematol Oncol. 2022;11(1):69. 10.1186/s40164-022-00321-x . Taylor V, Welcher AA, Program AE, Suter U. Epithelial membrane protein-1, peripheral myelin protein 22, and lens membrane protein 20 define a novel gene family. J Biol Chem. 1995;270(48):28824–33. 10.1074/jbc.270.48.28824 . Yasukawa K, Shimizu A, Kubota K, et al. Impact of Liver Fibrosis Severity on Oncological Prognosis in Hepatocellular Carcinoma. Liver Cancer. 2024;13(2):150–60. 10.1159/000533857 . Vilar-Gomez E, Calzadilla-Bertot L, Wai-Sun Wong V, et al. Fibrosis Severity as a Determinant of Cause-Specific Mortality in Patients With Advanced Nonalcoholic Fatty Liver Disease: A Multi-National Cohort Study. Gastroenterology. 2018;155(2):443–e45717. 10.1053/j.gastro.2018.04.034 . Li B, Hu J, Xu H. Integrated single-cell and bulk RNA sequencing reveals immune-related SPP1 + macrophages as a potential strategy for predicting the prognosis and treatment of liver fibrosis and hepatocellular carcinoma. Front Immunol. 2024;15:1455383. 10.3389/fimmu.2024.1455383 . Qin H, Abulaiti A, Maimaiti A, et al. Integrated machine learning survival framework develops a prognostic model based on inter-crosstalk definition of mitochondrial function and cell death patterns in a large multicenter cohort for lower-grade glioma. J Transl Med. 2023;21(1):588. 10.1186/s12967-023-04468-x . Wang Zyuan, Qian J, Lu X et al. A single-cell transcriptomic atlas characterizes liver non-parenchymal cells in healthy and diseased mice. bioRxiv . Published online 2021. https://api.semanticscholar.org/CorpusID:235779889 Chen J, Gao D, Chen J, et al. Pseudo-Temporal Analysis of Single-Cell RNA Sequencing Reveals Trans-Differentiation Potential of Greater Epithelial Ridge Cells Into Hair Cells During Postnatal Development of Cochlea in Rats. Front Mol Neurosci. 2022;15:832813. 10.3389/fnmol.2022.832813 . Cheng C, Chen W, Jin H, Chen X. A Review of Single-Cell RNA-Seq Annotation, Integration, and Cell-Cell Communication. Cells. 2023;12(15). 10.3390/cells12151970 . Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7. 10.1186/1471-2105-14-7 . Zheng M, Hu Y, Liu O, et al. Oxidative Stress Response Biomarkers of Ovarian Cancer Based on Single-Cell and Bulk RNA Sequencing. Oxid Med Cell Longev. 2023;2023:1261039. 10.1155/2023/1261039 . Cheng S, Zou Y, Zhang M, et al. Single-cell RNA sequencing reveals the heterogeneity and intercellular communication of hepatic stellate cells and macrophages during liver fibrosis. MedComm (2020). 2023;4(5):e378. 10.1002/mco2.378 . Iwaisako K, Jiang C, Zhang M, et al. Origin of myofibroblasts in the fibrotic liver in mice. Proc Natl Acad Sci U S A. 2014;111(32):E3297–3305. 10.1073/pnas.1400062111 . Koyama Y, Wang P, Liang S, et al. Mesothelin/mucin 16 signaling in activated portal fibroblasts regulates cholestatic liver fibrosis. J Clin Invest. 2017;127(4):1254–70. 10.1172/JCI88845 . Dooley S, Delvoux B, Streckert M, et al. Transforming growth factor beta signal transduction in hepatic stellate cells via Smad2/3 phosphorylation, a pathway that is abrogated during in vitro progression to myofibroblasts. TGFbeta signal transduction during transdifferentiation of hepatic stellate cells. FEBS Lett. 2001;502(1–2):4–10. 10.1016/s0014-5793(01)02656-4 . Chen H, Cai J, Wang J, et al. Targeting Nestin(+) hepatic stellate cells ameliorates liver fibrosis by facilitating TβRI degradation. J Hepatol. 2021;74(5):1176–87. 10.1016/j.jhep.2020.11.016 . Trivedi P, Wang S, Friedman SL. The Power of Plasticity-Metabolic Regulation of Hepatic Stellate Cells. Cell Metab. 2021;33(2):242–57. 10.1016/j.cmet.2020.10.026 . Mészáros K, Bojta J, Bautista AP, Lang CH, Spitzer JJ. Glucose utilization by Kupffer cells, endothelial cells, and granulocytes in endotoxemic rat liver. Am J Physiol. 1991;260(1 Pt 1):G7–12. 10.1152/ajpgi.1991.260.1.G7 . Quiroz Reyes AG, Lozano Sepulveda SA, Martinez-Acuña N, et al. Cancer Stem Cell and Hepatic Stellate Cells in Hepatocellular Carcinoma. Technol Cancer Res Treat. 2023;22:15330338231163677. 10.1177/15330338231163677 . Sangro B, Chan SL, Meyer T, Reig M, El-Khoueiry A, Galle PR. Diagnosis and management of toxicities of immune checkpoint inhibitors in hepatocellular carcinoma. J Hepatol. 2020;72(2):320–41. 10.1016/j.jhep.2019.10.021 . Anstee QM, Reeves HL, Kotsiliti E, Govaere O, Heikenwalder M. From NASH to HCC: current concepts and future challenges. Nat Rev Gastroenterol Hepatol. 2019;16(7):411–28. 10.1038/s41575-019-0145-7 . Huang DQ, Singal AG, Kanwal F, et al. Hepatocellular carcinoma surveillance - utilization, barriers and the impact of changing aetiology. Nat Rev Gastroenterol Hepatol. 2023;20(12):797–809. 10.1038/s41575-023-00818-8 . Cheng S, Zou Y, Zhang M, et al. Single-cell RNA sequencing reveals the heterogeneity and intercellular communication of hepatic stellate cells and macrophages during liver fibrosis. MedComm (2020). 2023;4(5):e378. 10.1002/mco2.378 . Wallace MC, Friedman SL. Hepatic fibrosis and the microenvironment: fertile soil for hepatocellular carcinoma development. Gene Expr. 2014;16(2):77–84. 10.3727/105221614X13919976902057 . Ebrahimi F, Hagström H, Sun J, et al. Familial coaggregation of MASLD with hepatocellular carcinoma and adverse liver outcomes: Nationwide multigenerational cohort study. J Hepatol. 2023;79(6):1374–84. 10.1016/j.jhep.2023.08.018 . Sigorski D, Sejda A, Abualsaud N, Krawczyk E, Izycka-Swieszewska E, Kitlinska J. Neuropeptide Y in cancer-biological functions and potential clinical implications. Cancer Metastasis Rev. 2025;44(1):21. 10.1007/s10555-024-10237-z . Lv X, Zhao F, Huo X, et al. Neuropeptide Y1 receptor inhibits cell growth through inactivating mitogen-activated protein kinase signal pathway in human hepatocellular carcinoma. Med Oncol. 2016;33(7):70. 10.1007/s12032-016-0785-1 . Lin T, Lin Z, Mai P, Zhang E, Peng L. Identification of prognostic biomarkers associated with the occurrence of portal vein tumor thrombus in hepatocellular carcinoma. Aging. 2021;13(8):11786–807. 10.18632/aging.202876 . Zhu Y, Zhang P. Gene expression profile of anoikis reveals new subtypes of liver cancer and discovery of therapeutic targets and biomarkers. Sci Rep. 2025;15(1):12740. 10.1038/s41598-025-96488-4 . Sial N, Ahmad M, Hussain MS, et al. CTHRC1 expression is a novel shared diagnostic and prognostic biomarker of survival in six different human cancer subtypes. Sci Rep. 2021;11(1):19873. 10.1038/s41598-021-99321-w . Li J, Wang Y, Ma M et al. Corrigendum to Autocrine CTHRC1 activates hepatic stellate cells and promotes liver fibrosis by activating TGF-β signaling [EBioMedicine 40 (2019) 43–55]. EBioMedicine . 2024;107:105303. 10.1016/j.ebiom.2024.105303 Li X, Zhang CC, Lin XT, et al. Elevated expression of WSB2 degrades p53 and activates the IGFBP3-AKT-mTOR-dependent pathway to drive hepatocellular carcinoma. Exp Mol Med. 2024;56(1):177–91. 10.1038/s12276-023-01142-6 . Song M, Pan Q, Yang J, et al. Galectin-3 favours tumour metastasis via the activation of β-catenin signalling in hepatocellular carcinoma. Br J Cancer. 2020;123(10):1521–34. 10.1038/s41416-020-1022-4 . Playford MP, Bicknell D, Bodmer WF, Macaulay VM. Insulin-like growth factor 1 regulates the location, stability, and transcriptional activity of beta-catenin. Proc Natl Acad Sci U S A. 2000;97(22):12103–8. 10.1073/pnas.210394297 . Arab JP, Cabrera D, Sehrawat TS, et al. Hepatic stellate cell activation promotes alcohol-induced steatohepatitis through Igfbp3 and SerpinA12. J Hepatol. 2020;73(1):149–60. 10.1016/j.jhep.2020.02.005 . Liu T, Zhang F, Feng Y, Han P, Gao Y. Alcohol-Metabolizing Enzymes, Liver Diseases and Cancer. Semin Liver Dis. 2025;45(1):99–113. 10.1055/a-2551-3320 . Mello T, Ceni E, Surrenti C, Galli A. Alcohol induced hepatic fibrosis: role of acetaldehyde. Mol Aspects Med. 2008;29(1–2):17–21. 10.1016/j.mam.2007.10.001 . Friedman SL. Stellate cell activation in alcoholic fibrosis–an overview. Alcohol Clin Exp Res. 1999;23(5):904–10. Peng Y, Yang T, Huang K, Shen L, Tao Y, Liu C. Salvia Miltiorrhiza Ameliorates Liver Fibrosis by Activating Hepatic Natural Killer Cells in Vivo and in Vitro. Front Pharmacol. 2018;9:762. 10.3389/fphar.2018.00762 . Zhang P, Liu W, Wang Y. The mechanisms of tanshinone in the treatment of tumors. Front Pharmacol. 2023;14:1282203. 10.3389/fphar.2023.1282203 . Supplementary Files CellMarkersdeHSC.csv DanshenDrugTargetAfterTreatment.xlsx HBVCONvsG2HFpbmc.csv TCGADEGupdown.csv venn.csv Cite Share Download PDF Status: Published Journal Publication published 01 Dec, 2025 Read the published version in Journal of Translational Medicine → Version 1 posted Reviewers agreed at journal 15 Aug, 2025 Reviewers invited by journal 15 Aug, 2025 Editor assigned by journal 05 Aug, 2025 First submitted to journal 02 Aug, 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-7278297","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":501142993,"identity":"6b672950-3acd-462f-bb72-3d1db3e80c9c","order_by":0,"name":"Tongyu Lu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7klEQVRIiWNgGAWjYBACPiBm/lPBwMDGDhdLwK+FjYGZgYHnDJDBTJIW3jaQXURrkcg/JiE5b5s8HzMD82eeP4cZ+NlzDBh+7sCnJZlNwnDbbcM2ZgY2ad62wwySPW8MGHvPENCSuO02I0gLM2/DYQaDGzkGzIxtBLQcnHPbvg3mMHtitEg2NtxOBGphkOZhA9oiQUgLz2Nja4Zjt5PbgMok57al80iceVZwsBePFn72xIe3GWpu285vbz784c0fazn+9uSND37i0QIELBIQmrGBiQcYrSDmAbwagHH4AcZi/EFA6SgYBaNgFIxMAAAoF0Qc+5O/cAAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0009-0000-9812-5415","institution":"Third Affiliated Hospital of Sun Yat-Sen University","correspondingAuthor":true,"prefix":"","firstName":"Tongyu","middleName":"","lastName":"Lu","suffix":""},{"id":501142994,"identity":"ed907b75-89ff-4138-b970-86059aacdcfc","order_by":1,"name":"Jie You","email":"","orcid":"","institution":"Third Affiliated Hospital of Sun Yat-Sen University","correspondingAuthor":false,"prefix":"","firstName":"Jie","middleName":"","lastName":"You","suffix":""},{"id":501142995,"identity":"bb6d9d85-5eba-47e3-bed6-a4942c8f4ab6","order_by":2,"name":"Yihuan Huang","email":"","orcid":"","institution":"Third Affiliated Hospital of Sun Yat-Sen University","correspondingAuthor":false,"prefix":"","firstName":"Yihuan","middleName":"","lastName":"Huang","suffix":""},{"id":501142996,"identity":"bd10bbde-83ec-4858-942a-9cab2af60853","order_by":3,"name":"Chenhao Jiang","email":"","orcid":"","institution":"Third Affiliated Hospital of Sun Yat-Sen University","correspondingAuthor":false,"prefix":"","firstName":"Chenhao","middleName":"","lastName":"Jiang","suffix":""},{"id":501142997,"identity":"703ec1ac-6150-453c-b1ba-3a562c3b5a51","order_by":4,"name":"Jiaqi Xiao","email":"","orcid":"","institution":"Third Affiliated Hospital of Sun Yat-Sen University","correspondingAuthor":false,"prefix":"","firstName":"Jiaqi","middleName":"","lastName":"Xiao","suffix":""},{"id":501142998,"identity":"ac62498a-e832-4164-8bb5-087d3b90ff54","order_by":5,"name":"Jiebin Zhang","email":"","orcid":"","institution":"Third Affiliated Hospital of Sun Yat-Sen University","correspondingAuthor":false,"prefix":"","firstName":"Jiebin","middleName":"","lastName":"Zhang","suffix":""},{"id":501142999,"identity":"18cf16e0-650b-490b-81a7-833690682718","order_by":6,"name":"Yasong Liu","email":"","orcid":"","institution":"Third Affiliated Hospital of Sun Yat-Sen University","correspondingAuthor":false,"prefix":"","firstName":"Yasong","middleName":"","lastName":"Liu","suffix":""},{"id":501143000,"identity":"e0c349e5-ef41-4cc0-8bd7-9d4933dfbd68","order_by":7,"name":"Xin Sui","email":"","orcid":"","institution":"Third Affiliated Hospital of Sun Yat-Sen University","correspondingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"Sui","suffix":""},{"id":501143001,"identity":"8a947903-ae49-45ac-95fd-eb4960593d2d","order_by":8,"name":"Yingcai Zhang","email":"","orcid":"https://orcid.org/0000-0002-3086-2440","institution":"Third Affiliated Hospital of Sun Yat-Sen University","correspondingAuthor":false,"prefix":"","firstName":"Yingcai","middleName":"","lastName":"Zhang","suffix":""},{"id":501143002,"identity":"22d3d231-2249-47ff-b1a1-d64da1cd4a59","order_by":9,"name":"Jia Yao","email":"","orcid":"","institution":"Third Affiliated Hospital of Sun Yat-Sen University","correspondingAuthor":false,"prefix":"","firstName":"Jia","middleName":"","lastName":"Yao","suffix":""}],"badges":[],"createdAt":"2025-08-02 12:39:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7278297/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7278297/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12967-025-07454-7","type":"published","date":"2025-12-01T15:57:38+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":89975058,"identity":"de019446-6c5b-4fd9-97cb-ba7ab2f2a3c7","added_by":"auto","created_at":"2025-08-27 05:56:32","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":655626,"visible":true,"origin":"","legend":"\u003cp\u003eThe flowchart of this study.\u003c/p\u003e","description":"","filename":"Picture1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/0948c50727d4696f07a98dbf.jpg"},{"id":89976570,"identity":"463db80b-9160-4990-8ec2-ee4db9945cc3","added_by":"auto","created_at":"2025-08-27 06:04:32","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":637322,"visible":true,"origin":"","legend":"\u003cp\u003eSingle-cell mapping of hepatitis B disease progression\u003c/p\u003e\n\u003cp\u003e(A) UMAP visualization revealed 37 distinct cell clusters across all samples.\u003c/p\u003e\n\u003cp\u003e(B) Sixteen cell types were identified based on canonical marker expression.\u003c/p\u003e\n\u003cp\u003e(C) Comparative UMAP projections demonstrated clustering patterns across disease stages (CON, G1, G2_HF, HCC).\u003c/p\u003e\n\u003cp\u003e(D) Cellular composition analysis identified the distribution of 16 cell types in each disease stage.\u003c/p\u003e\n\u003cp\u003e(E) Bubble plots illustrated cell type-specific marker gene expression profiles.\u003c/p\u003e\n\u003cp\u003e(F) UMAP feature plots highlighted spatial expression patterns of marker genes.\u003c/p\u003e\n\u003cp\u003e(G) Heatmap analysis showed differential gene expression across cell types.\u003c/p\u003e\n\u003cp\u003e(H) Violin plots displayed gene expression variation among cell populations.\u003c/p\u003e\n\u003cp\u003e(I) Quantitative analysis determined the percentage distribution of cell types across disease stages.\u003c/p\u003e","description":"","filename":"Picture2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/0d135fac997ef169e57ba6dc.jpg"},{"id":89975080,"identity":"8f8d7e7c-1a06-49ff-b003-e009904d8450","added_by":"auto","created_at":"2025-08-27 05:56:33","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":443075,"visible":true,"origin":"","legend":"\u003cp\u003eSingle-nucleus transcriptomic characterization of HSCs\u003c/p\u003e\n\u003cp\u003e(A) UMAP projection of HSCs subclustered into seven distinct populations.\u003c/p\u003e\n\u003cp\u003e(B) Disease-stage-specific HSC clustering (CON, G1, G2_HF, HCC).\u003c/p\u003e\n\u003cp\u003e(C) EMP1 expression-based stratification of HSC subsets.\u003c/p\u003e\n\u003cp\u003e(D) Distribution of EMP1+ and EMP1- HSCs across disease stages.\u003c/p\u003e\n\u003cp\u003e(E) Heatmap of differentially expressed genes between HSC subsets.\u003c/p\u003e\n\u003cp\u003e(F) UMAP visualization of subset-specific gene expression patterns.\u003c/p\u003e\n\u003cp\u003e(G) Bubble plot representation of differential gene expression.\u003c/p\u003e\n\u003cp\u003e(H) Activation/resting state marker expression profiles.\u003c/p\u003e\n\u003cp\u003e(I) Proportional abundance of HSC subsets across disease stages.\u003c/p\u003e","description":"","filename":"Picture3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/73102a6af57984aa17a2790e.jpg"},{"id":89977856,"identity":"99c429e9-ad6a-4bde-9215-e6f7baf18532","added_by":"auto","created_at":"2025-08-27 06:12:32","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":757464,"visible":true,"origin":"","legend":"\u003cp\u003eEnrichment analysis of cell subpopulations\u003c/p\u003e\n\u003cp\u003e(A) Volcano plot of intercellular differential gene expression.\u003c/p\u003e\n\u003cp\u003e(B) KEGG pathway enrichment of HSC-specific genes.\u003c/p\u003e\n\u003cp\u003e(C-D) GO functional enrichment of HSC markers (BP/MF/CC).\u003c/p\u003e\n\u003cp\u003e(E) Differential gene expression (EMP1+HSCs vs EMP1-HSCs).\u003c/p\u003e\n\u003cp\u003e(F) GO enrichment of EMP1+HSC signature genes.\u003c/p\u003e","description":"","filename":"Picture4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/29916264e80332200df9868c.jpg"},{"id":89976572,"identity":"19503f82-d6bc-482b-bb3f-372f36a3f8bc","added_by":"auto","created_at":"2025-08-27 06:04:32","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":595294,"visible":true,"origin":"","legend":"\u003cp\u003ePseudo-time analysis of HSC subpopulations\u003c/p\u003e\n\u003cp\u003e(A) Distribution of EMP1+ and EMP1- HSCs along pseudotime trajectory.\u003c/p\u003e\n\u003cp\u003e(B) HSC differentiation states across pseudotime progression.\u003c/p\u003e\n\u003cp\u003e(C) Developmental stage classification of HSCs.\u003c/p\u003e\n\u003cp\u003e(D) Proportional representation of HSC subpopulations at each pseudotime stage.\u003c/p\u003e\n\u003cp\u003e(E) Heatmap visualization of temporally-regulated gene modules and their functional enrichment.\u003c/p\u003e\n\u003cp\u003e(F) Expression dynamics of EMP1 and eight key developmental genes across pseudotime.\u003c/p\u003e","description":"","filename":"Picture5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/9d115edaa2f6f75438bc6dfe.jpg"},{"id":89977858,"identity":"07ec4eef-9ab4-483c-a72e-fd42dfee47ba","added_by":"auto","created_at":"2025-08-27 06:12:33","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":857398,"visible":true,"origin":"","legend":"\u003cp\u003eCell communication analysis in snRNA-seq\u003c/p\u003e\n\u003cp\u003e(A) Comparison of the number and strength of cellular communication between groups(CON, G1 ,G2_HF,HCC).\u003c/p\u003e\n\u003cp\u003e(B) Circle plots of the number and strength of intercellular interactions between the CON and G2_HF groups.\u003c/p\u003e\n\u003cp\u003e(C) Differences in the number and strength of signalling pathway-related cellular communication between the G2_HF group and the CON group.\u003c/p\u003e\n\u003cp\u003e(D) Circle plots comparing the number and strength of intercellular interactions in the G2_HF group compared to the CON group.\u003c/p\u003e\n\u003cp\u003e(E) Circle plots of the number number and strength comparison of intercellular interactions in the G2_HF group compared to the CON group.\u003c/p\u003e\n\u003cp\u003e(F) Heatmaps of ligand-receptor interaction patterns in the LAMININ signalling pathway in different cell types in each group \u0026nbsp;(CON, G1 ,G2_HF,HCC).\u003c/p\u003e\n\u003cp\u003e(G) Contribution ranking of receptor-ligand pairs within the LAMININ pathway in G2_HF.\u003c/p\u003e\n\u003cp\u003e(H) Cell-cell communication networks mediated by individual ligand-receptor pairs of the LAMB2-CD44 ligand in the G2_HF and CON groups.\u003c/p\u003e\n\u003cp\u003e(I) Differential expression of LAMININ signalling pathway molecules in G2_HF and CON groups.\u003c/p\u003e","description":"","filename":"Picture6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/1d66b8cb00e78ec18ffbc710.jpg"},{"id":89976577,"identity":"41bd9aa0-1af0-4fed-b0d2-3b7ba6fe4117","added_by":"auto","created_at":"2025-08-27 06:04:33","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":1035577,"visible":true,"origin":"","legend":"\u003cp\u003eMetabolic changes in each cell type during the course of hepatitis B disease\u003c/p\u003e\n\u003cp\u003e(A) Correlation of GSVA scores of highly variable genes (1000 genes with the largest standard deviation) among each cell type in each group (CON, \u0026nbsp;G1 , G2_HF, HCC).\u003c/p\u003e\n\u003cp\u003e(B) Metabolic pathway activity of each cell type in each group (CON, \u0026nbsp;G1 , G2_HF, HCC).For each metabolic pathway, pathway activity scores greater than 0.5 or less than 0.5 imply significant up- or down-regulation.\u003c/p\u003e\n\u003cp\u003e(C) Comparison of metabolic pathway activity of EMP1+ HSCs and EMP1-HSCs in each group (CON, \u0026nbsp;G1 , G2_HF, HCC); (ns P＞0.05,P\u0026lt;0.05, **P\u0026lt;0.01, ***P\u0026lt;0.001, ****P\u0026lt;0.0001).\u003c/p\u003e","description":"","filename":"Picture7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/95eba0c2ab077d46581527f0.jpg"},{"id":89975070,"identity":"0d34e817-ec84-4ed0-a1b1-9856c38d8d4c","added_by":"auto","created_at":"2025-08-27 05:56:33","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":591185,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification and Functional Analysis of HSC-Related Pathogenic Genes\u003c/p\u003e\n\u003cp\u003e(A) Volcano plot of differentially expressed genes between TCGA-LIHC tumors and normal liver tissues.\u003c/p\u003e\n\u003cp\u003e(B) Volcano plot of differentially expressed genes in HSCs (G2_HF vs CON groups).\u003c/p\u003e\n\u003cp\u003e(C) Venn diagram showing intersection of HSC marker genes, single-cell DEGs (G2_HF vs CON), and TCGA-LIHC DEGs.\u003c/p\u003e\n\u003cp\u003e(D) GO enrichment analysis of 45 HPGs (sorted by p-value).\u003c/p\u003e\n\u003cp\u003e(E) GO enrichment map showing BP, MF and CC terms.\u003c/p\u003e\n\u003cp\u003e(F) UMAP visualization of GO pathway activity (AUC scores) in EMP1+ vs EMP1- HSCs (purple: low; yellow: high activity).\u003c/p\u003e\n\u003cp\u003e(G) Violin plots comparing GO pathway activities between EMP1+ and EMP1- HSCs (ns P＞0.05, *P\u0026lt;0.05, **P\u0026lt;0.01, ***P\u0026lt;0.001, ****P\u0026lt;0.0001).\u003c/p\u003e\n\u003cp\u003e(H) UMAP visualization of GO pathway activities across experimental groups (CON, G1, G2_HF, HCC).\u003c/p\u003e\n\u003cp\u003e(I) Violin plots comparing GO pathway activities among experimental groups (ns P＞0.05, P\u0026lt;0.05, **P\u0026lt;0.01, ***P\u0026lt;0.001, ****P\u0026lt;0.0001).\u003c/p\u003e","description":"","filename":"Picture8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/8a0b67948c5ffa028f154d45.jpg"},{"id":89975104,"identity":"36d0f76a-abb4-45c7-a512-3c6e318d1310","added_by":"auto","created_at":"2025-08-27 05:56:34","extension":"jpg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":862572,"visible":true,"origin":"","legend":"\u003cp\u003e101 machine learning-based prognostic model construction and functional characterization of HCC prognostic genes\u003c/p\u003e\n\u003cp\u003e(A) Network plot of intersecting genes among HSC marker genes in single cell samples, intergroup DEGs of HSCs and DEGs of TCGA-LIHC samples.\u003c/p\u003e\n\u003cp\u003e(B) Construction of 1101 machine learning models trained on TCGA-LIHC and validated on GSE16757 and GSE43619 datasets.\u003c/p\u003e\n\u003cp\u003e(C) Boxplot of the expression differences of HCC prognostic genes in the G2_HF and HCC groups of HSCs.\u003c/p\u003e\n\u003cp\u003e(D) Bubble plots showing the expression of HCC prognostic genes screened by the optimal models in 16 cell types across cells in the G2_HF group and HCC group.\u003c/p\u003e\n\u003cp\u003e(E) Bubble plots showing the expression of HCC prognostic genes screened by the optimal model in EMP1+HSCs versus EMP1-HSCs.\u003c/p\u003e\n\u003cp\u003e(F) Umap plot demonstrating the AUC activity score of the prognostic gene set in EMP1+HSCs versus EMP1-HSCs; (bright blue: low; red: high activity).\u003c/p\u003e\n\u003cp\u003e(G) Quantitative comparison of prognostic gene set activities between EMP1+ and EMP1- HSCs (ns P＞0.05, *P\u0026lt;0.05, **P\u0026lt;0.01, ***P\u0026lt;0.001, ****P\u0026lt;0.0001).\u003c/p\u003e","description":"","filename":"Picture9.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/c98207383751ce7f886d31a9.jpg"},{"id":89975093,"identity":"d7408f4e-e78d-4278-b1fd-79dd8656e11d","added_by":"auto","created_at":"2025-08-27 05:56:34","extension":"jpg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":751735,"visible":true,"origin":"","legend":"\u003cp\u003eValidation of prognostic model predictive efficacy\u003c/p\u003e\n\u003cp\u003e(A) Kaplan-Meier survival curves based on patient risk scores and survival status in the training and test sets.\u003c/p\u003e\n\u003cp\u003e(B) ROC curves predicting the probability of patient death over time for 1, 3, and 5 years in the training and test sets.\u003c/p\u003e\n\u003cp\u003e(C) Histogram showing the overall AUC values in the training and test sets.\u003c/p\u003e\n\u003cp\u003e(D) Histogram demonstrating AUC values for 1, 3, and year 5 in the prediction training set and test set.\u003c/p\u003e\n\u003cp\u003e(E) Univariate Cox regression analyses of patient HCC-related prognostic genes, risk scores, and risk groupings.\u003c/p\u003e\n\u003cp\u003e(F) Multivariable Cox regression analysis of HCC-related prognostic genes in patients.\u003c/p\u003e\n\u003cp\u003e(G) Multivariate analysis incorporating clinical parameters and risk stratification.\u003c/p\u003e\n\u003cp\u003e(H) Clinical nomogram integrating CPGs and risk scores for survival prediction.\u003c/p\u003e","description":"","filename":"Picture10.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/7a5bfe9499d8c7cb3e923876.jpg"},{"id":89975076,"identity":"8e69756b-7db4-41c3-8c1c-508d7fa58c1d","added_by":"auto","created_at":"2025-08-27 05:56:33","extension":"jpg","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":279391,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of the clinical significance of core prognostic genes\u003c/p\u003e\n\u003cp\u003e(A) Flowchart of screening core prognostic genes from HSC-related pathogenic genes.\u003c/p\u003e\n\u003cp\u003e(B) Heatmap showing the distribution of the four core prognostic genes in high and low risk groups.\u003c/p\u003e\n\u003cp\u003e(C) Box plots showing comparative expression of CPGs between high- and low-risk subgroups (*P\u0026lt;0.05, **P\u0026lt;0.01, ***P\u0026lt;0.001).\u003c/p\u003e","description":"","filename":"Picture11.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/39fb84f7acc7972828e01dc4.jpg"},{"id":89975075,"identity":"4a943062-9d27-4e12-acbe-ef381d144e20","added_by":"auto","created_at":"2025-08-27 05:56:33","extension":"jpg","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":832403,"visible":true,"origin":"","legend":"\u003cp\u003eDrug screening and molecular docking results\u003c/p\u003e\n\u003cp\u003e(A) Network diagram of the prognostic gene set - HSC-related pathogenic genes - drug active ingredient - Salvia divinorum(Danshen).\u003c/p\u003e\n\u003cp\u003e(B) Heatmap of binding energies between active components and core prognostic gene products.\u003c/p\u003e\n\u003cp\u003e(C) Molecular structures of four key active components (Salvilenone_I、Tanshinone_IIA、Salvianolic_acid_g、Isotanshinone_II).\u003c/p\u003e\n\u003cp\u003e(D) Molecular docking results of Danshen active ingredients with core prognostic gene products.\u003c/p\u003e","description":"","filename":"Picture12.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/c39dc439b4c0fa1df5474191.jpg"},{"id":97724026,"identity":"3a3fe56a-7687-4754-b365-60bd99dfeaed","added_by":"auto","created_at":"2025-12-08 16:11:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9599437,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/37d6558a-993c-4a13-92bd-a0fe06d5295f.pdf"},{"id":89975092,"identity":"4081f954-ed95-4597-9108-1366c3456538","added_by":"auto","created_at":"2025-08-27 05:56:34","extension":"csv","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":5937062,"visible":true,"origin":"","legend":"","description":"","filename":"CellMarkersdeHSC.csv","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/898b6967b6067284c9825888.csv"},{"id":89976576,"identity":"5389ecec-df4e-44b0-a22a-68204e5e2ad3","added_by":"auto","created_at":"2025-08-27 06:04:33","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":87617,"visible":true,"origin":"","legend":"","description":"","filename":"DanshenDrugTargetAfterTreatment.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/c35bc291bedc8a65189d70fa.xlsx"},{"id":89975079,"identity":"816738a2-d098-42a1-a7d6-e06088c6f3b9","added_by":"auto","created_at":"2025-08-27 05:56:33","extension":"csv","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":672674,"visible":true,"origin":"","legend":"","description":"","filename":"HBVCONvsG2HFpbmc.csv","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/c507007aa802bcd3a4296493.csv"},{"id":89975072,"identity":"ad3fbdba-d6d2-45a4-bf19-f231a520ddf5","added_by":"auto","created_at":"2025-08-27 05:56:33","extension":"csv","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":2546492,"visible":true,"origin":"","legend":"","description":"","filename":"TCGADEGupdown.csv","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/a67591faddd7d325e5941601.csv"},{"id":89975078,"identity":"ca137cc6-1699-4e53-8d8d-fac42b6cb325","added_by":"auto","created_at":"2025-08-27 05:56:33","extension":"csv","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":367,"visible":true,"origin":"","legend":"","description":"","filename":"venn.csv","url":"https://assets-eu.researchsquare.com/files/rs-7278297/v1/d3bed23512690208f1115591.csv"}],"financialInterests":"","formattedTitle":"EMP1+ hepatic stellate cells drive hepatic fibrosis progression to hepatocellular carcinoma and predict prognosis","fulltext":[{"header":"2 Introduction","content":"\u003cp\u003eHepatic fibrosis (HF) and hepatocellular carcinoma (HCC) represent two pathologically interconnected stages in the progression of chronic liver disease.\u003c/p\u003e\u003cp\u003eHepatic fibrosis is a wound-healing response to chronic liver injury characterized by excessive deposition of extracellular matrix (ECM) proteins\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, which progressively disrupts hepatic architecture and function, culminating in cirrhosis and HCC development\u003csup\u003e\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eGlobally, liver diseases account for approximately 2\u0026nbsp;million annual deaths, with cirrhosis responsible for 50% of these cases\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e; As the fourth leading cause of cancer mortality worldwide, HCC demonstrates a dismal 5-year survival rate of merely 20%\u003csup\u003e7,8\u003c/sup\u003e. Notably, hepatitis B virus (HBV)-associated HCC predominates in high-prevalence regions such as China\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e; where advanced hepatic fibrosis is prevalent among chronic HBV patients\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. These clinical associations highlight the critical need to investigate HBV-induced hepatic fibrosis and its malignant transformation.\u003c/p\u003e\u003cp\u003eCurrent diagnostic modalities for HCC face significant limitations. Serum biomarker tests, particularly alpha-fetoprotein (AFP), exhibit suboptimal sensitivity (40\u0026ndash;60%).\u003csup\u003e12,13\u003c/sup\u003e. while conventional imaging techniques (including ultrasound and PET/CT) frequently fail to detect early-stage HCC within fibrotic liver tissue\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eIn addition, therapeutic challenges persist across disease stages. Although systemic therapies (encompassing targeted therapy, immunotherapy, and chemotherapy) are clinically established, they yield unsatisfactory outcomes in advanced HCC\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. For cirrhotic patients without overt malignancy, conventional anti-fibrotic agents like pirfenidone demonstrate inadequate specificity for pathogenic cell subpopulations. Moreover, existing research has insufficiently addressed disease-driving genes and signaling pathways in specific cellular populations, hindering effective prevention of fibrotic progression to HCC\u003csup\u003e\u003cspan additionalcitationids=\"CR19 CR20\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Consequently, there is an urgent need to identify core molecular targets governing this malignant transformation.\u003c/p\u003e\u003cp\u003eThe activation of hepatic stellate cells (HSCs) constitutes the central pathogenic mechanism underlying hepatic fibrosis, as these cells serve as the predominant producers of extracellular matrix (ECM)\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. Quiescent HSCs typically express desmin and glial fibrillary acidic protein (GFAP), whereas activated HSCs undergo profound phenotypic changes characterized by upregulated fibrogenic gene expression and enhanced proliferative capacity - the key drivers of excessive ECM deposition and pro-fibrotic cytokine release in hepatic fibrosis\u003csup\u003e\u003cspan additionalcitationids=\"CR24\" citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. However, emerging evidence reveals remarkable heterogeneity within the HSC population, and the distinct contributions of its functional subpopulations to disease progression remain poorly defined.\u003c/p\u003e\u003cp\u003eNotably, epithelial membrane protein 1 (EMP1) has been identified as a molecular hallmark of hepatic fibrosis\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. As a membrane surface protein abundantly expressed in endothelial lineage cells, EMP1 plays pivotal roles in both fibrotic and malignant processes\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Intriguingly, EMP1 expression is also detected in HSCs, yet the precise biological functions of the EMP1\u0026thinsp;+\u0026thinsp;HSC subpopulation and its mechanistic involvement in HCC pathogenesis - particularly its clinical relevance to patient prognosis - await systematic investigation.\u003c/p\u003e\u003cp\u003eTherefore, elucidating the unique role of the EMP1\u0026thinsp;+\u0026thinsp;HSCs subset in the progression from hepatic fibrosis to HCC is crucial for revealing the cellular mechanisms of disease progression and discovering new prognostic intervention targets. Clinical guidelines have indicated that the degree of hepatic fibrosis should be considered an independent risk factor for treatment decisions and prognosis in HCC.\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e Consequently, comprehensive exploration of EMP1-HSC interactions in chronic liver disease may yield innovative targets for predicting and treating hepatic fibrosis and HCC.\u003c/p\u003e\u003cp\u003eThe aim of this study was to employ single-cell transcriptomic analysis to systematically integrate disease spectrum data spanning from healthy controls (CON), grade 1 hepatitis B (G1), grade 2 hepatitis B with fibrosis (G2_HF), and HCC patients, with particular emphasis on the EMP1\u0026thinsp;+\u0026thinsp;HSC subpopulation. We characterized the differentially expressed genes (DEGs) between EMP1\u0026thinsp;+\u0026thinsp;HSCs and EMP1-HSCs, and elucidated their potential contributions to fibrogenesis through pathway enrichment and cell-cell communication analyses. Using machine learning approaches combined with survival analysis and clinical validation, we developed robust prognostic models that identified four core prognostic genes: NPY1R (Neuropeptide Y Receptor Type 1), CTHRC1 (Collagen Triple Helix Repeat-Containing Protein 1), IGFBP3 (Insulin-like Growth Factor Binding Protein 3), and ADH1B (Alcohol Dehydrogenase 1B). These analyses revealed significant expression heterogeneity between HCC tissues and EMP1\u0026thinsp;+\u0026thinsp;HSCs, and established their clinical correlations with chronic liver disease progression. Through comprehensive network pharmacology screening of traditional Chinese medicine databases (including TCMSP), we evaluated numerous herbal formulations such as Paeonia lactiflora-Glycyrrhiza glabra soup and Fuzheng Water-Repelling Formula. This systematic approach identified Salvia miltiorrhiza (Danshen) as the most promising therapeutic candidate, demonstrating effective binding to all core prognostic targets (IGFBP3, CTHRC1, NPY1R, and ADH1B). Further investigation of its bioactive components, combined with in-depth analysis of HBV-related fibrotic progression, provides novel mechanistic insights and potential therapeutic targets for early intervention in hepatic fibrosis and HCC.\u003c/p\u003e"},{"header":"3 Materials and Methods","content":"\u003cp\u003eIn this study, we first constructed a single-cell transcriptional profile of the hepatic fibrotic progression continuum, focusing on the expression of HSC-related pathogenic genes, and performed intercellular communication analysis, trajectory analysis, and transcription factor regulatory network analysis. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn addition, we utilized TCGA and GEO databases to validate and demonstrate that HSC-related pathogenic genes exhibit specificity in EMP1+HSCs and can serve as biological markers for both fibrotic progression to HCC and HCC prognosis.\u003c/p\u003e\n\u003ch2\u003e3.1\u0026nbsp; \u0026nbsp;\u0026nbsp; Data sources\u0026nbsp;and pre-processing\u003c/h2\u003e\n\u003cp\u003eSingle-nucleus RNA sequencing data for hepatic fibrosis were obtained from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). We analyzed the hepatitis B progression continuum using the following datasets: Healthy controls (CON, n=5) from GSE136103;\u0026nbsp;Hepatitis G1 stage (G1, n=1) from GSE186343; Hepatitis G2 stage with hepatic fibrosis (G2_HF, n=5) from GSE186343;\u0026nbsp;Hepatocellular carcinoma patients (HCC, n=7) from GSE202642.\u0026nbsp;The cell selection criteria were as follows:\u0026nbsp;a) 300-10,000 cells per sample; b) \u0026gt;250 genes expressed per cell; c) Genes detected in \u0026ge;3 cells; d) \u0026lt;20% mitochondrial RNA content per cell\u003csup\u003e31\u003c/sup\u003e. Using the Seurat R package, we ultimately retained 75,143 high-quality cells for downstream analyses.\u003c/p\u003e\n\u003cp\u003eHepatocellular carcinoma bulk-RNA sequencing data and corresponding clinical data were obtained from the TCGA database (https://www.cancer.gov/ccg/research/genome-sequencing/tcga). The TCGA-LIHC dataset comprised 424 samples, including 50 normal samples and 374 tumor samples (368 with available clinical data).We identified differentially expressed genes (DEGs) using DESeq2 with thresholds of |log2FoldChange| \u0026gt;1.5 and adjusted p-value \u0026lt;0.05\u003csup\u003e31,32\u003c/sup\u003e.\u0026nbsp;For validation, we obtained additional datasets (GSE16757 [n=100] and GSE43619 [n=88]) from GEO, totaling 188 HCC tumor samples with clinical information.\u003c/p\u003e\n\u003ch2\u003e3.2\u0026nbsp; \u0026nbsp;\u0026nbsp; Data quality control\u003c/h2\u003e\n\u003cp\u003eAfter filtering the data for integration, we normalized the expression values using the \u0026quot;NormalizeData\u0026quot; function from the Seurat package. This normalization procedure involved:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e1. Multiplying each gene\u0026apos;s expression by the total gene expression per cell. 2.Scaling by a factor of 10,000. 3.Applying natural logarithmic transformation after adding a pseudocount of 1 to avoid taking the logarithm of zero.\u003c/p\u003e\n\u003ch2\u003e3.3\u0026nbsp; \u0026nbsp;\u0026nbsp; Dimensionality Reduction and Cell Clustering\u003c/h2\u003e\n\u003cp\u003eSince each gene represents a distinct dimension in the sample, high-dimensional data visualization was challenging. We therefore employed dimensionality reduction techniques to represent the underlying data structure in reduced dimensions\u003csup\u003e31\u003c/sup\u003e. Using the Seurat package, we first applied the \u0026quot;RunPCA\u0026quot; function to reduce dimensionality based on highly variable genes. Subsequently, we identified 2,000 integration anchors through the \u0026quot;Find Integration Anchors\u0026quot; function (RunHarmony implementation), which effectively aligned corresponding cell types across datasets while mitigating batch effects.\u003c/p\u003e\n\u003ch2\u003e3.4\u0026nbsp; \u0026nbsp;\u0026nbsp; Cell annotation, DEG and marker gene identification\u003c/h2\u003e\n\u003cp\u003eThe SingleR package enables correlation analysis between single-cell gene expression profiles and reference cell type signatures at single-cell resolution. In our analysis, we employed the package to calculate expression correlations using highly variable genes (HVGs), iteratively eliminating the weakest correlations for each cell type to determine definitive cell identities\u003csup\u003e33\u003c/sup\u003e. For annotation, we utilized the \u0026quot;ref_Human_all\u0026quot; reference database within SingleR to facilitate manual cell type classification. We subsequently identified differentially expressed genes (DEGs) using the \u0026quot;FindAllMarkers\u0026quot; function in Seurat, applying thresholds of |log2FoldChange| \u0026gt; 2 and adjusted p-value \u0026lt; 0.05. The same function was employed to detect cell type-specific marker genes for downstream analyses.\u003c/p\u003e\n\u003ch2\u003e3.5\u0026nbsp; \u0026nbsp;\u0026nbsp; GO and KEGG Enrichment Analysis\u003c/h2\u003e\n\u003cp\u003eGene Ontology (GO) functional enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed using both the DAVID and Metascape databases. The enrichment results were subsequently visualized using the ggplot2 package in R.\u003c/p\u003e\n\u003ch2\u003e3.6\u0026nbsp; \u0026nbsp;\u0026nbsp; Pseudotime Trajectory Analysis\u003c/h2\u003e\n\u003cp\u003eCells undergo dynamic transitions between states, accompanied by concomitant changes in gene expression and functional phenotype\u0026nbsp;\u003csup\u003e33\u003c/sup\u003e.\u0026nbsp;Using the Monocle 2 package, we performed pseudotime trajectory analysis to reconstruct cellular differentiation processes by positioning cells along developmental trajectories based on their gene expression profiles\u003csup\u003e34\u003c/sup\u003e.\u0026nbsp;The \u0026quot;plot_cell_trajectory\u0026quot; function was employed to order cells by pseudotemporal progression, while the \u0026quot;BEAM\u0026quot; function identified branch-dependent genes, with results visualized using the \u0026quot;plot_genes_branched_heatmap\u0026quot; function.\u003c/p\u003e\n\u003ch2\u003e3.7\u0026nbsp; \u0026nbsp;\u0026nbsp; Cell-to-Cell Communication analysis\u003c/h2\u003e\n\u003cp\u003eCell surface ligand-receptor interactions mediate crucial intercellular communication in biological processes. We employed the CellChat software package to construct cellular communication networks by integrating ligand-receptor interactions with their associated factors. This approach simulated intercellular communication patterns by leveraging gene expression profiles of ligands and receptors across different cell types, thereby inferring potential interactions and elucidating rich ligand-receptor interplay between cell population\u0026nbsp;\u003csup\u003e35\u003c/sup\u003e.\u003c/p\u003e\n\u003ch2\u003e3.8\u0026nbsp; \u0026nbsp;\u0026nbsp; GSVA and Metabolic Pathway Scoring\u003c/h2\u003e\n\u003cp\u003eSingle-sample gene set enrichment analysis (ssGSEA) is a method that quantifies the enrichment score of specific gene sets in individual samples, where the ssGSEA score reflects the activation status of genomic systems. In this study, we obtained metabolic pathway scores for each cell subpopulation using ssGSEA implemented in the GSVA R package\u0026nbsp;\u003csup\u003e36\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eTo identify metabolic differences between HSC subpopulations, we selected genes showing strong population specificity (adjusted p \u0026lt; 0.05, |log2FoldChange| \u0026gt; 1.5, pct.1 \u0026gt; 0.5, and pct.2 \u0026lt; 0.5) and calculated metabolic activity scores using the AUCell package\u003csup\u003e37\u003c/sup\u003e. The UMAP visualization colored by AUC scores revealed metabolically active cell subpopulations, while differential gene expression analysis between active and inactive subgroups was performed using the FindAllMarkers function with default parameters (avg_log2FC \u0026gt; 1).\u003c/p\u003e\n\u003ch2\u003e3.9\u0026nbsp; \u0026nbsp;\u0026nbsp; Identification and Functional Analysis of HSC-Related Pathogenic Genes\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eWe analyzed differential gene expression between normal and tumor samples from the TCGA database using stringent thresholds (|logFC|\u0026gt;1.5 and p.adj \u0026lt; 0.05). By intersecting these bulk RNA-seq DEGs with (1) DEGs of HSCs from snRNA-seq data (G2_HF vs CON) and (2) HSC marker genes, we identified HSC-related pathogenic genes (HPGs) implicated in hepatic fibrosis across transcriptomic levels.\u003c/p\u003e\n\u003cp\u003eSubsequent GO enrichment analysis revealed HPG functions, while pathway activity scores computed using the AUCell package\u0026nbsp;enabled comparative assessment across clinical stages (CON, G1, G2_HF, HCC) and between EMP1+HSC and EMP1-HSC subpopulations\u003csup\u003e37\u003c/sup\u003e.\u003c/p\u003e\n\u003ch2\u003e3.10\u0026nbsp; Construction of Machine Learning-Based Prognostic Models\u003c/h2\u003e\n\u003cp\u003eTo validate the prognostic role of HSC-related pathogenic genes (HPGs), we constructed 101 machine learning prognostic models following established methodology\u0026nbsp;\u003csup\u003e31\u003c/sup\u003e.\u0026nbsp;The analysis proceeded through four key phases:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e1. Dataset preparation and model construction: The TCGA-LIHC dataset (n=368) served as the training set, while GSE16757 (n=100) and GSE43619 (n=88) functioned as external validation sets. We implemented ten machine learning algorithms: Lasso, Ridge, Stepwise Cox, CoxBoost, Random Survival Forest (RSF), Elastic Net (Enet), Partial Least Squares Regression for Cox (plsRcox), Supervised Principal Components (SuperPC), Generalized Boosted Regression Modeling (GBM), and Survival Support Vector Machine (Survival-svm). Through ten-fold cross-validation, we developed 101 algorithm combinations for variable selection and model building.\u003c/p\u003e\n\u003cp\u003e2. Model evaluation and selection: All models were evaluated on both training and validation sets. We calculated concordance indices (C-indices) for each model across datasets, ranking them by average C-index to identify the most robust and clinically significant algorithm combination.\u003c/p\u003e\n\u003cp\u003e3.\u0026nbsp;Prognostic validation: The optimal model (highest mean C-index) identified HCC-related prognostic genes (HCC-PGs) used to stratify patients into high- and low-risk groups based on median risk scores. We assessed prognostic significance through Kaplan-Meier analysis (survminer package; log-rank test, p\u0026lt;0.05) and evaluated predictive accuracy using time-dependent ROC curves (timeROC package). The HCC-PGs\u0026apos; predictive performance was compared against other clinical characteristics through AUC comparisons.\u003c/p\u003e\n\u003cp\u003e4.\u0026nbsp;Biological validation: The HCC-PGs were analyzed using the AUCell package37 to calculate activity scores across HSC subpopulations (CON, G1, G2_HF, HCC) and between EMP1+HSC and EMP1-HSC subsets.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e3.11\u0026nbsp;\u0026nbsp;Cox Regression Analysis and Nomogram Construction\u003c/h2\u003e\n\u003cp\u003eTo further evaluate the clinical relevance of HCC prognostic genes (HCC-PGs), we conducted comprehensive analyses through two main approaches:\u003c/p\u003e\n\u003cp\u003e1.\u0026nbsp;Prognostic factor analysis: We examined the correlation between HCC-PG risk scores and key clinical characteristics (age, gender, stage, T/N/M classification, and tumor grade) in the TCGA-LIHC (n=368), GSE16757 (n=100), and GSE43619 (n=88) datasets. Univariate and multivariate Cox regression analyses were performed to determine whether these risk scores served as independent prognostic factors, with results visualized through forest plots. Genes demonstrating statistical significance (p \u0026lt; 0.05) in these analyses were designated as Core Prognostic Genes (CPGs).\u003c/p\u003e\n\u003cp\u003e2.\u0026nbsp;Nomogram development: To enhance prognostic precision, we constructed a nomogram integrating HCC-PG risk scores with clinical characteristics using the \u0026quot;rms\u0026quot; R package. This predictive tool estimates 1- and 5-year survival probabilities by combining CPG profiles with standard clinicopathological parameters.\u003c/p\u003e\n\u003ch2\u003e3.12\u0026nbsp;Drug Screening and Molecular Docking Analysis\u003c/h2\u003e\n\u003cp\u003e1. Traditional Chinese medicine screening: We systematically screened herbal components from Paeoniae Alba-Glycyrrhiza glabra Tang and Fu Zheng Yi Shui Fang formulations through the TCMSP database (https://www.tcmsp-e.com/tcmsp.php; accessed January 24, 2025), ultimately identifying Salvia miltiorrhiza as the candidate therapeutic agent. Active compounds were selected based on pharmacokinetic parameters (oral bioavailability \u0026ge;30% and drug-likeness \u0026ge;0.18). Potential targets (n=7,727) were obtained from PharmMapper (http://lilab-ecust.cn/pharmmapper/index.html), CTD (http://ctdbase.org/) and Swiss Target Prediction (http://www.swisstargetprediction.ch/).\u003c/p\u003e\n\u003cp\u003e2. Network pharmacology analysis: Among the identified targets, 45 HSC-related pathogenic genes were recognized as potential therapeutic targets for hepatic fibrosis treatment. Using Cytoscape software (https://cytoscape.org/), we constructed a comprehensive drug-component-target-disease network, highlighting the top 10 active ingredients with the most target intersections.\u003c/p\u003e\n\u003cp\u003e3. Molecular docking validation: We analyzed interactions between Salvia miltiorrhiza bioactive compounds (Salvianolic acid B, Tanshinone IIA, etc.) and core prognostic genes (NPY1R [PDB:5ZBH], CTHRC1 [UniProt:Q96CG8], IGFBP3 [PDB:7WRQ], ADH1B [PDB:1HSZ]). Protein structures were obtained from RCSB PDB (https://www.rcsb.org/) and AlphaFold DB (https://alphafold.ebi.ac.uk/), while compound 3D structures were retrieved from TCMSP (accessed February 14, 2025). Molecular docking was performed using AutoDockTools 4.2.6 (accessed February 16, 2025) after preparing ligands (Chem3D 19.0) and receptors (PyMOL 2.6.0). Binding conformations with energies \u0026le;-5 kcal/mol were considered significant and visualized using PyMOL.\u003c/p\u003e\n\u003ch2\u003e3.13\u0026nbsp;Statistical Analysis\u003c/h2\u003e\n\u003cp\u003eAll statistical analyses were performed using R software (version 4.1.1, accessed October 6, 2024). We employed the Student\u0026apos;s t-test for comparing normally distributed continuous variables, while the Wilcoxon rank-sum test was used for non-parametric data comparisons. Survival differences between high-risk and low-risk groups were evaluated through Kaplan-Meier analysis with log-rank testing. A threshold of p \u0026lt; 0.05 was established for statistical significance throughout our study.\u0026nbsp;\u003c/p\u003e"},{"header":"4 Results","content":"\u003cdiv id=\"Sec16\"\u003e\n \u003ch2\u003e4.1 Overview of single-cell mapping in the hepatitis B disease process\u003c/h2\u003e\n \u003cp\u003eWe analyzed 18 single-nucleus RNA-seq samples from GSE136103 and GSE186343 datasets, encompassing healthy controls (CON, n\u0026thinsp;=\u0026thinsp;5), hepatitis G1 stage (G1, n\u0026thinsp;=\u0026thinsp;1), hepatitis G2 stage with hepatic fibrosis (G2_HF, n\u0026thinsp;=\u0026thinsp;5), and hepatocellular carcinoma patients (HCC, n\u0026thinsp;=\u0026thinsp;7). After quality control filtering, 75,143 high-quality cells were retained for analysis. Data integration using anchor-based alignment effectively removed batch effects, followed by normalization, pooling, and PCA dimensionality reduction (retaining top 20 PCs). UMAP visualization revealed 37 distinct clusters (Fig.\u0026nbsp;2A), which were annotated into 16 cell types (Fig.\u0026nbsp;2B): Kupffer cells (Kupffer, CD163+), macrophages (Macrophages, S100A9+), NK/T cells (NK_T, CD3E+), innate lymphoid cells (ILCs, CXCL8+), dendritic cells (DCs, CLEC10A+), NK cells (NK, KLRF1+), B cells (B_cells, CD79A+), plasma cells (Plasma, IGHM+), hepatic stellate cells (HSCs, ACTA2+), dividing cells (Dividing, TOP2A+), T cells (T_cells, BATF+), hepatocytes (Hepatocytes, ALDOB+), vascular endothelial cells (VECs, VWF+), cholangiocytes (Cholangiocytes, KRT7+), liver sinusoidal endothelial cells (LSECs, STAB2+), and hepatocellular carcinoma cells (HCCs, GPC3+).\u003c/p\u003e\n \u003cp\u003eFigure 2C demonstrated the transcriptomic profiles across disease stages, while Fig. 2D displayed the cellular composition distribution. Marker gene expression patterns (Fig. 2E-F) and intercellular DEG analysis (Fig. 2G-H) validated annotation accuracy. Cellular proportion analysis (Fig. 2I) showed dominant populations of NK_T, LSECs and VECs, with relatively fewer HSCs, Hepatocytes, and Cholangiocytes.\u003c/p\u003e\n \u003cp\u003eQuantitative analysis determined the percentage distribution of cell types across disease stages.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\"\u003e\n \u003ch2\u003e4.2 Single-cell Transcriptomic Profiling of Hepatic Stellate Cells\u003c/h2\u003e\n \u003cp\u003eHepatic stellate cells (HSCs) are recognized as central mediators of hepatic fibrosis progression \u003csup\u003e1\u003c/sup\u003e. The snRNA-seq profiles showed the presence of a large number of hepatic stellate cells (HSCs) in all samples. Our snRNA-seq analysis identified 2,237 HSCs across all samples for subsequent clustering. Re-clustering revealed seven distinct HSC subpopulations (Fig.\u0026nbsp;3A), with disease-stage-specific transcriptomic profiles (Fig.\u0026nbsp;3B).Given EMP1\u0026apos;s established role as a hepatic fibrosis biomarker\u003csup\u003e26\u003c/sup\u003e. we stratified HSCs into EMP1- HSCs (n\u0026thinsp;=\u0026thinsp;1,971) and EMP1\u0026thinsp;+\u0026thinsp;HSCs (n\u0026thinsp;=\u0026thinsp;266) subsets based on EMP1 expression (Fig.\u0026nbsp;3C). Comparative analysis demonstrated differential distribution patterns across disease stages (CON, G1, G2_HF, HCC) (Fig.\u0026nbsp;3D).\u003c/p\u003e\n \u003cp\u003eDifferential gene expression analysis between EMP1\u0026thinsp;+\u0026thinsp;and EMP1- HSCs identified distinct molecular signatures (Fig.\u0026nbsp;3E), visualized through UMAP embeddings (Fig.\u0026nbsp;3F) and bubble plots (Fig.\u0026nbsp;3G). Activation state marker analysis showed EMP1\u0026thinsp;+\u0026thinsp;HSCs exhibited profiles resembling activated HSCs (COL1A1+, etc.), though paradoxically lower ACTA2 expression versus EMP1-HSCs (Fig.\u0026nbsp;3H) - aligning with recent evidence questioning ACTA2\u0026apos;s reliability as an activation marker\u003csup\u003e38\u003c/sup\u003e. Finally, quantification revealed progressive expansion of EMP1\u0026thinsp;+\u0026thinsp;HSCs during disease progression (CON, G1, G2_HF, HCC), significantly exceeding healthy control proportions (Fig. 3I), implicating this subset in HBV-related fibrogenesis.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\"\u003e\n \u003ch2\u003e4.3 Enrichment Analysis of Cell Subpopulations\u003c/h2\u003e\n \u003cp\u003eFigure 4A demonstrated significant upregulation of differential genes in hepatic stellate cells (HSCs) compared to other cell types, with higher logFC values indicating their predominant functional role in chronic hepatitis B-associated hepatic fibrosis. GO and KEGG enrichment analyses of HSC marker genes revealed key functional pathways (Fig. 4B-D): KEGG analysis identified enrichment in cytoskeletal regulation in muscle cells, focal adhesion, non-alcoholic fatty liver disease, chemical carcinogenesis-reactive oxygen species, and oxidative phosphorylation, while GO analysis showed extracellular structure organization, extracellular matrix (ECM) organization, and cell-substrate adhesion as the main biological processes (BP), collagen-containing extracellular matrix and cell-substrate junctions as primary cellular components (CC).The KEGG and GO enriched pathways described above represent the primary molecular mechanisms through which HSCs contribute to hepatic fibrosis \u003csup\u003e39\u0026ndash;42\u003c/sup\u003e. And ECM structural constituents and NAD(P)H dehydrogenase (quinone) activity as major molecular functions (MF), suggesting that metabolic activity has an impact on HSC function.\u003c/p\u003e\n \u003cp\u003eTo further explore the role of EMP1 in HSCs and hepatic fibrosis during the course of chronic hepatitis B, we identified differential genes between the two types of HSCs (EMP1\u0026thinsp;+\u0026thinsp;HSCs versus EMP1-HSCs) (Fig. 4E), which showed an overall increase in the gene expression level of EMP1\u0026thinsp;+\u0026thinsp;HSCs relative to that of EMP1-HSCs, with 710 upregulated and 46 downregulated genes identified. GO enrichment of differentially expressed genes (Fig. 4F) showing similar extracellular organization pathway enrichment patterns to bulk HSCs, functionally validating EMP1\u0026thinsp;+\u0026thinsp;HSCs as the primary effector subpopulation in HSC-mediated fibrotic processes.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec19\"\u003e\n \u003ch2\u003e4.4 Pseudotime Trajectory Analysis of HSC Subpopulations\u003c/h2\u003e\n \u003cp\u003ePseudotime trajectory analysis, which reconstructs cellular developmental pathways based on temporal gene expression patterns, was performed on hepatic stellate cells (HSCs) from our snRNA-seq data. The analysis revealed three distinct branches of HSC subtypes (Fig.\u0026nbsp;5A) and three corresponding differentiation states (Fig.\u0026nbsp;5C). The developmental progression, illustrated in Figs.\u0026nbsp;5B-D, showed a clear transition from state1 to state3, characterized by a gradual shift from darker-colored to lighter-colored cells. Within this trajectory, EMP1\u0026thinsp;+\u0026thinsp;HSCs were predominantly localized to intermediate developmental stages, while EMP1-HSCs were distributed throughout.\u003c/p\u003e\n \u003cp\u003eGene expression patterns were categorized into three temporally-regulated modules (Fig.\u0026nbsp;5E). The green module, associated with copper ion response and processing, suggested a potential role for copper detoxification in HSC activation and fibrotic responses. The red module demonstrated significant enrichment for extracellular matrix formation and maintenance pathways, consistent with the established role of HSCs in differentiating into collagen-producing myofibroblasts during hepatic fibrosis. The blue module, linked to energy metabolism regulation and cellular adhesion, indicated additional functional aspects of HSCs in liver injury response, development and regeneration through metabolic support and intercellular communication.\u003c/p\u003e\n \u003cp\u003eDetailed examination of EMP1 and eight key developmentally-regulated genes (Fig.\u0026nbsp;5F) confirmed the mid-trajectory predominance of EMP1\u0026thinsp;+\u0026thinsp;HSCs and pan-trajectory distribution of EMP1-HSCs, corroborating the findings in Fig.\u0026nbsp;5D. COL1A1, COL1A2, COL3A1 and TIMP1 expression progressively declined during development, whereas APOD, CFD, DCN and MFAP5 exhibited an initial increase followed by terminal stage decrease. Notably, all eight genes showed expression patterns tightly associated with EMP1\u0026thinsp;+\u0026thinsp;HSCs and were nearly absent in terminal stage EMP1-HSCs, particularly evident for APOD, CFD, DCN and MFAP5. Importantly, while EMP1 expression served to distinguish HSC subpopulations, it did not appear directly involved in HSC proliferation or differentiation processes. And the identification of collagen genes (COL1A1, COL1A2, and COL3A1) - known critical mediators of HSC activation and function\u003csup\u003e42,41\u003c/sup\u003e, within this analysis further substantiates EMP1\u0026thinsp;+\u0026thinsp;HSCs as the primary functional subpopulation in hepatic fibrogenesis.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec20\"\u003e\n \u003ch2\u003e4.5 Analysis of cell-to-cell communication in HSCs\u003c/h2\u003e\n \u003cp\u003eTo further explore the effect of EMP1\u0026thinsp;+\u0026thinsp;HSCs on other cells during chronic hepatitis B progression and their role in hepatic fibrosis, we performed cell-to-cell communication analysis by tracking ligand-receptor interactions using our snRNA-seq data. First, Figs. 6A-B demonstrated that in the chronic hepatitis B process, the number and strength of cellular communication between EMP1\u0026thinsp;+\u0026thinsp;HSCs and other major cells were significantly stronger in the G2_HF group than in the CON group. Figure 6B further revealed the specific distribution of the cellular communication parameters across cell types in both the CON and G2_HF groups, showing that EMP1\u0026thinsp;+\u0026thinsp;HSCs served as the most important source of ligand signaling in the G2_HF group and exhibited the most significant enhancement in cellular communication relative to the CON group. Figure 6C identified CD46, VTN, LAMININ and ANNEXIN as the predominant mediators of enhanced cellular communication in the G2_HF group (Fig. 6C), with LAMININ and ANNEXIN representing extracellular matrix-associated signaling pathways. Notably, analysis of the enhanced communication network showed that LSECs, VECs and Macrophages served as the primary signal-receiving cells, whereas EMP1\u0026thinsp;+\u0026thinsp;HSCs emerged as the dominant signal-emitting population, with their contribution far exceeding that of EMP1-HSCs (Figs. 6D-E). Strikingly, Fig. 6F demonstrated that within the LAMININ signaling pathway, EMP1\u0026thinsp;+\u0026thinsp;HSCs were the most active ligand signal emitters, where the LAMB2-CD44 pair constituted the single most influential receptor-ligand interaction (Fig. 6G). The specific communication patterns mediated by LAMB2-CD44 were visualized across cell types in both CON and G2_HF groups (Fig. 6H), while Fig. 6I provided a comprehensive comparison of the expression levels of all LAMININ pathway molecules between these two groups.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec21\"\u003e\n \u003ch2\u003e4.6 HSC metabolic pathway analysis\u003c/h2\u003e\n \u003cp\u003eIn the pseudo-time analysis of HSCs (Fig. 5E), the processes in the blue module were found to be associated with the regulation of energy metabolism and cell adhesion, indicating that hepatic stellate cells not only participate in extracellular matrix organization but may also support their activation through energy metabolism and maintain intercellular connections via adhesion regulation. Given the liver\u0026apos;s crucial role in systemic energy metabolism, we investigated the metabolic reprogramming of NPCs in hepatic disease states. Previous studies have shown that the hepatic immune response involves enhanced glucose metabolism in immunoreactive cells \u003csup\u003e43\u003c/sup\u003e. HSCs show a particularly high sensitivity, and they play an important role in immune metabolism by maintaining liver function and responding to injury\u003csup\u003e44\u003c/sup\u003e.\u003c/p\u003e\n \u003cp\u003eTo characterize metabolic alterations, we first selected the top 1,000 highly variable genes in each cell population and performed cell-to-cell similarity assessment. This analysis revealed that HSCs displayed the closest expression patterns to VECs and LSECs across all groups (CON, G1, G2_HF, HCC) (Fig. 7A). Subsequently, we quantified metabolic pathway activities to evaluate NPC reprogramming in different disease stages, focusing on carbohydrate, energy, and lipid metabolism. GSVA results demonstrated that HSCs were the more metabolically active population in all groups (Fig. 7B), with AUC analysis further identifying oxidative phosphorylation as the predominant metabolic pathway in both HSC subpopulations (Fig. 7C). Despite an overall reduction in HSC metabolic activities in the G2_HF group (Fig. 7B), we observed elevated oxidative phosphorylation activity in both HSC subpopulations compared to other groups, with EMP1\u0026thinsp;+\u0026thinsp;HSCs exhibiting significantly higher activity than EMP1-HSCs (Fig. 7C).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec22\"\u003e\n \u003ch2\u003e4.7 Identification and Functional Analysis of HSC-Related Pathogenic Genes\u003c/h2\u003e\n \u003cp\u003eDifferential gene analysis of the TCGA-LIHC dataset identified 821 DEGs (Fig.\u0026nbsp;8A), which were intersected with 5505 HSC marker genes in the single-cell samples (Fig.\u0026nbsp;4A) and 2882 intergroup DEGs (G2_HF vs. CON) in HSCs (Fig.\u0026nbsp;8B), yielding 45 overlapping genes (Fig.\u0026nbsp;8C). The three differential gene sets used to obtain the intersection genes are provided in Supplementary File S1. These genes were differentially expressed not only in HSCs of the G2_HF group compared to the CON group but also showed significant differences between hepatocellular carcinoma patients and healthy individuals. We hypothesized that these 45 genes might serve as key regulators in hepatocellular carcinoma prognosis through HSCs, with their altered expression contributing to the progression from hepatic fibrosis to hepatocellular carcinoma; hence, they were designated as HSC-related pathogenic genes (HPGs).\u003c/p\u003e\n \u003cp\u003eTo investigate the functional roles of HPGs, GO analysis was performed on these genes (Fig. 8D). The results revealed that epithelial cell proliferation was the most statistically significant term in the BP category, while collagen-containing extracellular matrix and extracellular matrix structural constituent were the most significant in the CC and MF categories, respectively (Fig. 8E). Epithelial cell proliferation can drive epithelial-derived cells (e.g., hepatocytes, cholangiocytes, or hepatocellular carcinoma cells) into the cell cycle. Activated HSCs secrete abundant HGF, EGF, and TGF-\u0026alpha;, which bind to MET/EGFR on epithelial cell membranes, subsequently activating the PI3K-AKT or RAS-MAPK cascades and promoting hepatocyte/hepatocellular carcinoma cell proliferation \u003csup\u003e45\u003c/sup\u003e. Additionally, the collagen-containing extracellular matrix and extracellular matrix structural constituent are critical pathways in hepatic fibrosis, suggesting that HSCs may contribute not only to fibrogenesis but also to hepatocellular carcinoma progression by enhancing tumor cell proliferation and exacerbating fibrosis.\u003c/p\u003e\n \u003cp\u003eTo determine the predominant HSC subpopulation involved in hepatocellular carcinoma prognosis, we calculated the AUC activity scores of these three GO pathways in HSCs and visualized them using UMAP plots (Fig. 8F, 8H). Further analysis revealed that EMP1\u0026thinsp;+\u0026thinsp;HSCs exhibited significantly higher activity in these pathways compared to EMP1-HSCs, indicating that EMP1\u0026thinsp;+\u0026thinsp;HSCs represent the major HSC subset implicated in hepatocellular carcinoma prognosis (Fig. 8G). Finally, the three GO pathways displayed elevated activity in the G2_HF group compared to the CON group, with the highest AUC scores observed in the HCC group, further supporting the role of HSCs in the transition from hepatic fibrosis to hepatocellular carcinoma and their involvement in disease prognosis (Fig. 8I).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec23\"\u003e\n \u003ch2\u003e4.8 101 machine learning-based prognostic model construction and functional characterization of HCC prognostic genes\u003c/h2\u003e\n \u003cp\u003eThe network diagram revealed the gene composition of the 45 HPGs (Fig. 9A). After normalizing the expression patterns of these genes, we employed 101 machine learning algorithms to calculate their risk scores, using the TCGA-LIHC dataset (n\u0026thinsp;=\u0026thinsp;368) as the training set and the GSE16757 (n\u0026thinsp;=\u0026thinsp;100) and GSE43619 (n\u0026thinsp;=\u0026thinsp;88) datasets as validation sets (Fig. 9B). The StepCox[forward]\u0026thinsp;+\u0026thinsp;Ridge model demonstrated superior predictive performance, through which we identified 10 key genes designated as HCC Prognostic Genes (HCC-PGs), highlighted in black in the network diagram (Fig. 9A). Comparative analysis of HCC-PG expression in HSCs from G2_HF and HCC groups revealed that six genes (ABCA8, INMT, ANXA2, OLFML2B, ADH1B, and NPY1R) exhibited significant differential expression (Fig. 9C), suggesting their potential dual role in both HCC prognosis and the transition from hepatic fibrosis to hepatocellular carcinoma. Spatial distribution analysis of the 10 HCC-PGs across 16 cell subpopulations demonstrated their predominant expression in HSCs (Fig. 9D-E). Notably, EMP1\u0026thinsp;+\u0026thinsp;HSCs showed significantly higher expression levels of these genes compared to EMP1-HSCs, indicating potential functional divergence between these HSC subtypes. Subsequent AUC activity scoring of the integrated 10-gene prognostic signature confirmed significantly enhanced activity in EMP1\u0026thinsp;+\u0026thinsp;HSCs relative to EMP1-HSCs (Fig. 9F-G), further supporting the pivotal role of EMP1\u0026thinsp;+\u0026thinsp;HSCs in HCC progression.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec24\"\u003e\n \u003ch2\u003e4.9 Validation of prognostic model predictive efficacy\u003c/h2\u003e\n \u003cp\u003eAmong the 101 machine learning prognostic models evaluated, the StepCox[forward]\u0026thinsp;+\u0026thinsp;Ridge model demonstrated superior predictive performance (Fig. 9B). Kaplan-Meier survival analysis revealed significantly worse outcomes in high-risk groups across both training and test sets, with statistically significant separation of survival curves (Fig. 10A). Time-dependent ROC curve analysis confirmed the robust predictive capacity of our prognostic signature, with AUC values exceeding 0.7 for 1-year and 3-year survival predictions in both cohorts, while maintaining strong performance (AUC\u0026thinsp;=\u0026thinsp;0.69) for 5-year survival prediction (Fig. 10B-D). Comprehensive Cox regression analyses incorporating clinical and molecular data identified several significant prognostic factors. Univariate analysis showed that ABCA8, INMT, GADD45B, GHR, CTHRC1, IGFBP3, ADH1B, along with risk scores, were significantly associated with overall survival (Fig. 10E). Subsequent multivariate analysis pinpointed four genes with particularly strong prognostic value: NPY1R (HR [95% CI]: 0.94 [0.90\u0026ndash;0.98]), ADH1B (HR [95% CI]: 0.96 [0.92\u0026ndash;0.99]), CTHRC1 (HR [95% CI]: 1.04 [1.01\u0026ndash;1.08]), and IGFBP3 (HR [95% CI]: 1.09 [1.04\u0026ndash;1.14]) (Fig. 10F). Based on these findings, we designated these four genes as core prognostic genes (CPGs), with NPY1R and ADH1B emerging as protective factors, while CTHRC1 and IGFBP3 were identified as risk factors. Further multivariate analysis incorporating clinical parameters revealed that tumor M1 stage and risk stratification maintained independent prognostic significance. Notably, the low-risk group exhibited substantially better outcomes compared to high-risk patients (HR [95% CI]: 0.36 [0.23\u0026ndash;0.58]) (Fig. 10G). To facilitate clinical translation, we developed a comprehensive nomogram incorporating the four CPGs and risk scores, which provides individualized 1- to 5-year survival probability estimates, with risk scores emerging as the most dominant contributor to survival prediction (Fig. 10H).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec25\"\u003e\n \u003ch2\u003e4.10 Identification of the clinical significance of core prognostic genes\u003c/h2\u003e\n \u003cp\u003eThe screening process of core prognostic genes (CPGs) from HSC-related pathogenic genes was illustrated in Fig.\u0026nbsp;11A. To validate the clinical relevance of these CPGs, we analyzed the expression patterns of NPY1R, CTHRC1, IGFBP3, and ADH1B in risk-stratified patients from three independent cohorts: TCGA-LIHC (n\u0026thinsp;=\u0026thinsp;368), GSE16757 (n\u0026thinsp;=\u0026thinsp;100), and GSE43619 (n\u0026thinsp;=\u0026thinsp;88). To minimize batch effects, we presented the results from different databases separately (Fig.\u0026nbsp;11B-C).\u003c/p\u003e\n \u003cp\u003eOur data showed that NPY1R exhibited significant differential expression only in TCGA samples (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), but not in the GEO datasets, suggesting its context-dependent prognostic role. CTHRC1 and IGFBP3 demonstrated significantly higher expression in high-risk groups (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), while ADH1B was preferentially expressed in low-risk patients (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01), consistent with their classification as risk and protective factors, respectively, in our previous analysis (Fig. 10F). Notably, ADH1B and NPY1R displayed marked expression differences between G2_HF and HCC groups in HSCs (Fig. 9B), indicating their dual involvement in HCC prognosis and fibrotic-to-HCC progression.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec26\"\u003e\n \u003ch2\u003e4.11 Drug screening and molecular docking analysis\u003c/h2\u003e\n \u003cp\u003eIn order to find targeted drugs against HSC-related pathogenic genes (HPGs) to reverse the transformation of cirrhosis to hepatocellular carcinoma and improve prognosis, we systematically screened traditional Chinese medicine compounds from TCMSP and additional drug databases. After evaluating numerous candidates including Paeonia lactiflora, Glycyrrhiza glabra, and Fu Zheng Yi Shui Fang, Salvia miltiorrhiza was selected as the primary study drug. We obtained 7727 potential targets of Salvia miltiorrhiza through PharmMapper, CTD, and Swiss Target Prediction (These targets are listed in Supplementary File S2).\u003c/p\u003e\n \u003cp\u003eThe network diagram revealed the top 10 Salvia miltiorrhiza active components most strongly associated with HPGs (Fig. 12A): Salvianolic acid B, Isotanshinone II, Luteolin, Microstegiol, Miltirone II, Salvianolic acid G, Salvilenone I, Salviolone, Tanshinone IIA, and Tanshinone VI. Subsequent molecular docking analysis demonstrated strong binding potential (binding energy \u0026lt; -5 kcal/mol) between these components and proteins encoded by the four core prognostic genes (Fig. 12B). Particularly noteworthy were the exceptionally low binding energies observed between: 1) Salvilenone I and NPY1R, 2) Tanshinone IIA and CTHRC1, 3) Salvianolic acid G and IGFBP3, and 4) Isotanshinone II and ADH1B. Detailed structural analysis showed the molecular configurations of these four promising compounds (Fig. 12C) and their specific binding patterns with corresponding target proteins (Fig. 12D). The docking results illustrated precise amino acid interactions and chemical bond formations for each receptor-ligand pair, confirming the structural basis for their high-affinity binding.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"5 Discussion","content":"\u003cp\u003eWhile numerous clinical studies have established hepatic fibrosis as an independent risk factor for hepatocellular carcinoma\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, the precise mechanisms linking fibrotic progression to hepatocarcinogenesis remain poorly characterized \u003csup\u003e\u003cspan additionalcitationids=\"CR47\" citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. This persistent knowledge gap highlights the urgent need to delineate pathogenic genes and pathways at single-cell resolution, which could facilitate the development of precision therapies and improve clinical outcomes for fibrosis-associated HCC.\u003c/p\u003e\u003cp\u003eThrough systematic analysis of single-cell transcriptomes across the chronic hepatitis B disease continuum - including healthy controls (CON), grade 1 hepatitis B (G1), grade 2 hepatitis B with fibrosis (G2_HF), and HCC patients - we identified and characterized a novel EMP1-high hepatic stellate cell (HSC) subpopulation (EMP1\u0026thinsp;+\u0026thinsp;HSCs). Comprehensive bioinformatics analysis revealed that EMP1-associated differentially expressed genes (DEGs) were predominantly enriched in extracellular matrix (ECM) organization pathways, directly implicating this cellular subset in fibrogenesis\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Of particular note, the transcriptional signature of EMP1\u0026thinsp;+\u0026thinsp;HSCs showed striking similarity to activated HSCs (aHSCs)\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e, providing compelling evidence for EMP1 as a novel activation marker.\u003c/p\u003e\u003cp\u003eThe pseudo-time trajectory analysis revealed that EMP1\u0026thinsp;+\u0026thinsp;HSCs were predominantly localized to intermediate and late activation stages of HSC development, while extracellular matrix (ECM) formation and maintenance emerged as a critical biological process during HSC differentiation. Through systematic screening, we identified eight key genes demonstrating the most significant expression changes during HSC activation, including COL1A1, COL1A2, and COL3A1, which are core structural components of the ECM. Strikingly, the upregulation of these ECM-related genes was temporally coupled with the emergence of EMP1\u0026thinsp;+\u0026thinsp;HSCs, providing compelling evidence that EMP1 serves as a novel biomarker for activated HSCs (aHSCs). Single-cell communication analysis further demonstrated that EMP1\u0026thinsp;+\u0026thinsp;HSCs acted as the dominant signaling hubs in intercellular communication networks, exhibiting significantly stronger signaling activity compared to EMP1-HSCs, particularly in fibrogenic pathways mediated by LAMININ interactions. Taken together, these findings establish that EMP1\u0026thinsp;+\u0026thinsp;HSCs represent a functionally distinct HSC subpopulation with markedly different biological characteristics from EMP1-HSCs, and confirm their primary role in driving hepatic fibrosis progression. These results strongly support the potential clinical utility of EMP1 as a reliable molecular marker for identifying the activated state of HSCs.\u003c/p\u003e\u003cp\u003eInterestingly, our pseudo-time trajectory analysis uncovered the involvement of energy metabolism pathways in HSC differentiation, revealing that HSCs contribute to liver homeostasis not only through extracellular matrix reorganization but also by participating in metabolic support and intercellular connectivity modulation via adhesion molecule regulation. These findings suggest a multifunctional role for HSCs in liver development, regeneration, and injury response. Metabolic pathway analysis demonstrated that HSCs exhibited heightened metabolic activity across all experimental groups compared to other cell types. However, HSCs in G2_HF group showed reduced overall metabolic activity relative to healthy controls (CON), while EMP1\u0026thinsp;+\u0026thinsp;HSCs displayed significantly elevated oxidative phosphorylation capacity compared to EMP1- HSCs. These metabolic profiling results provide compelling evidence that distinct HSC subpopulations possess characteristic metabolic signatures corresponding to their functional states.\u003c/p\u003e\u003cp\u003eAs established in previous studies, hepatic fibrosis represents a chronic liver pathology whose malignant progression frequently culminates in cirrhosis and hepatocellular carcinoma (HCC) \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e,\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. To elucidate the functional contribution of EMP1-associated differentially expressed genes (DEGs) in HSCs during fibrosis-induced hepatocarcinogenesis, we performed systematic intersection analysis between TCGA-LIHC bulk transcriptomic data and single-cell datasets, yielding 45 overlapping genes designated as HSC-related pathogenic genes (HPGs). Employing 101 machine learning algorithms, we constructed a robust prognostic model based on these HPGs, through which core prognostic genes were identified via optimal model selection and multivariate COX regression analysis. Notably, NPY1R and ADH1B emerged as significant protective factors, whereas IGFBP3 and CTHRC1 were characterized as detrimental factors in HCC progression.\u003c/p\u003e\u003cp\u003eNPY1R, as a member of the G protein-coupled receptor (GPCR) family, is predominantly expressed in the nervous system and vascular smooth muscle cells, with relatively lower expression in HSCs, where it regulates cell proliferation and vasoconstriction\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Although NPY1R has been understudied in hepatocellular carcinoma, existing evidence demonstrates that it suppresses tumor cell growth by inhibiting the mitogen-activated protein kinase (MAPK) signaling pathway, with its elevated expression correlating positively with improved survival in advanced HCC patients\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e,\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e, which is in line with the results of this study. However, a recent contradictory study reported that NPY1R overexpression in HCC tissues might facilitate tumor immune evasion and metastasis through activation of pro-survival STAT3/AKT1 pathways\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e, highlighting the need for further investigations to clarify its context-dependent roles in hepatocarcinogenesis.\u003c/p\u003e\u003cp\u003eIn our single-cell analyses, CTHRC1 exhibited predominant expression in HSCs, particularly showing markedly higher levels in EMP1\u0026thinsp;+\u0026thinsp;HSCs compared to EMP1-HSCs across both fibrotic and HCC stages. As a secreted glycoprotein primarily localized in the extracellular matrix, CTHRC1 directly contributes to hepatic fibrosis by activating HSCs through autocrine mechanisms and potentiating TGF-β signaling\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e,\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e. Although our study did not detect significant expression differences of CTHRC1 between HCC and G2_HF groups at the bulk level, its sustained high expression in EMP1\u0026thinsp;+\u0026thinsp;HSCs suggests this subpopulation may drive fibrosis progression toward HCC through persistent ECM remodeling.\u003c/p\u003e\u003cp\u003eAlthough our study did not detect significant expression differences of CTHRC1 between HCC and G2_HF groups at the bulk level, its sustained high expression in EMP1\u0026thinsp;+\u0026thinsp;HSCs suggests this subpopulation may drive fibrosis progression toward HCC through persistent ECM remodeling \u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. While our TCGA-LIHC data confirmed its decreased expression at bulk level, emerging evidence indicates that in galectin-3-high HCC subtypes, PI3K-AKT-GSK-3β-β-catenin pathway activation can paradoxically upregulate IGFBP3 and vimentin, promoting angiogenesis and epithelial-mesenchymal transition (EMT)-mediated metastasis \u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e,\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e. Thus, IGFBP3 may interact with various pathways at different stages of tumor development and lead to different outcomes. Intriguingly, in our single-cell data from G2_HF group, IGFBP3 was primarily expressed by HSCs rather than hepatocytes, consistent with reports that HSC-derived IGFBP3 promotes alcohol-induced steatohepatitis\u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e, suggesting its complex, context-dependent roles in liver disease progression that may vary by etiology and cellular origin.\u003c/p\u003e\u003cp\u003eADH1B, the key ethanol-metabolizing enzyme, has well-documented tumor-suppressive effects, with its low expression associated with inflammatory pathway activation, metabolic reprogramming, and poorer HCC prognosis - findings corroborated by our current data showing its protective role \u003csup\u003e\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e. Although mainly expressed in hepatocytes, we observed significant ADH1B expression differences between EMP1\u0026thinsp;+\u0026thinsp;HSC and EMP1-HSC subpopulations. Given that acetaldehyde (the ethanol metabolite) directly stimulates HSC proliferation\u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e,\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e, this differential ADH1B expression may functionally distinguish these HSC subsets. The observed ADH1B expression variations between HCC and G2_HF HSCs further suggest its potential involvement in modulating the fibrosis-to-HCC transition, possibly through immune microenvironment regulation rather than direct fibrogenic effects.\u003c/p\u003e\u003cp\u003eCollectively, our findings demonstrate that the four core prognostic genes participate in hepatic fibrogenesis through distinct biological pathways and collectively contribute to the progression from hepatic fibrosis to hepatocellular carcinoma. The differential expression patterns of these genes in EMP1\u0026thinsp;+\u0026thinsp;HSCs versus EMP1-HSCs not only support EMP1's utility as a reliable marker for HSC activation state, but also suggest its potential as a biomarker for monitoring fibrosis-to-HCC progression.\u003c/p\u003e\u003cp\u003eTo explore therapeutic strategies targeting HSC-related pathogenic genes (HPGs) for preventing cirrhosis-to-HCC progression, we conducted systematic screening of traditional herbal medicines, ultimately selecting Salvia miltiorrhiza as our primary candidate based on its documented therapeutic effects. Previous studies have established that Salvia miltiorrhiza ameliorates hepatic fibrosis by activating intrahepatic natural killer (NK) cells to eliminate activated HSCs, thereby reducing collagen deposition\u003csup\u003e\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e; furthermore, Tanshinone IIA, a key active component, has been shown to suppress HCC cell proliferation and invasion through inhibition of the TGF-β/SMAD7-YAP pathway\u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e, confirming this herb's dual therapeutic potential for both hepatic fibrosis and HCC. Through comprehensive molecular docking analysis, we identified 10 Salvia miltiorrhiza components showing highest affinity with HPGs, with all binding energies \u0026lt;-5 kcal/mol. These strong molecular interactions provide a structural basis for the herb's potential efficacy in interrupting cirrhosis-to-HCC progression and improving HCC prognosis.\u003c/p\u003e"},{"header":"6 Conclusion","content":"\u003cp\u003eThrough integrative analysis combining single-cell and bulk transcriptomic data with advanced machine learning approaches, this study systematically elucidated the pivotal role of EMP1\u0026thinsp;+\u0026thinsp;HSCs in hepatic fibrogenesis during chronic hepatitis B progression to hepatocellular carcinoma. By employing differential enrichment analysis, pseudo-time trajectory reconstruction, cell-cell communication profiling, and metabolic activity assessment, we identified 45 HSC-related pathogenic genes (HPGs) and associated pathways that collectively drive Hepatic fibrosis and HCC development, with EMP1\u0026thinsp;+\u0026thinsp;HSCs emerging as the central cellular mediators of these pathogenic processes. Utilizing 101 machine learning algorithms, we established a robust prognostic model that pinpointed four core prognostic genes (NPY1R, CTHRC1, IGFBP3, and ADH1B), while comprehensively characterizing their expression heterogeneity and clinical relevance in HCC progression. These investigations not only confirmed EMP1 as a reliable biomarker for HSC activation states but also revealed its potential utility in monitoring fibrosis-to-HCC progression. Furthermore, through systematic drug screening and molecular docking analyses, we demonstrated that Salvia miltiorrhiza and its bioactive components exhibit strong binding affinities (\u0026lt;-5 kcal/mol) with the four core prognostic gene products, thereby providing molecular evidence for its therapeutic potential in interrupting cirrhosis-to-HCC progression and improving HCC outcomes.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003cp\u003eGEO is a public database. Since our study relies on open-source data, it encounters no ethical concerns or conflicts of interest.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003cp\u003eAll authors agree to submit the article for publication.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003cp\u003eThe authors declare no competing interest in this work.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eThis work was supported by the grants from the National Natural Science Foundation of China (82270689, 82460139), the Natural Science Foundation of Guangdong Province (2025A03J3197), Guangzhou Basic and Applied Basic Research Project Co-funded by Municipal Schools (institutes) (2025A03J3197, 2024B03J1382), Guangzhou Key R\u0026amp;D Fields Project(Agricultural and Social Development Science and Technology Special Topic)(2024B03J1382), Natural Science Foundation of Xinjiang Uygur Autonomous Region(2024D01E21), Xinjiang Uygur Autonomous Region People\u0026rsquo;s Hospital Liver Transplantation Special Program(20240101).\u003c/p\u003e\u003ch2\u003eAuthors' contributions\u003c/h2\u003e\u003cp\u003eJie You and Yihuan Huang contributed equally to this study.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e\u003cp\u003eWe express our gratitude to TCGA and the GEO database for their platforms and the contributors who uploaded significant datasets.\u003c/p\u003e\u003ch2\u003eData availability\u003c/h2\u003e\u003cp\u003eThe datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eKisseleva T, Brenner D. Molecular and cellular mechanisms of liver fibrosis and its regression. Nat Rev Gastroenterol Hepatol. 2021;18(3):151\u0026ndash;66. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41575-020-00372-7\u003c/span\u003e\u003cspan address=\"10.1038/s41575-020-00372-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRoehlen N, Crouchet E, Baumert TF. Liver Fibrosis: Mechanistic Concepts and Therapeutic Perspectives. Cells. 2020;9(4). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/cells9040875\u003c/span\u003e\u003cspan address=\"10.3390/cells9040875\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCampos-Murgu\u0026iacute;a A, Ruiz-Marg\u0026aacute;in A, Gonz\u0026aacute;lez-Regueiro JA, Mac\u0026iacute;as-Rodr\u0026iacute;guez RU. Clinical assessment and management of liver fibrosis in non-alcoholic fatty liver disease. World J Gastroenterol. 2020;26(39):5919\u0026ndash;43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3748/wjg.v26.i39.5919\u003c/span\u003e\u003cspan address=\"10.3748/wjg.v26.i39.5919\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAleknavičiūtė-Valienė G, Banys V. Clinical importance of laboratory biomarkers in liver fibrosis. Biochem Med (Zagreb). 2022;32(3):030501. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.11613/BM.2022.030501\u003c/span\u003e\u003cspan address=\"10.11613/BM.2022.030501\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGin\u0026egrave;s P, Krag A, Abraldes JG, Sol\u0026agrave; E, Fabrellas N, Kamath PS. Liver cirrhosis. Lancet. 2021;398(10308):1359\u0026ndash;76. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S0140-6736(21)01374-X\u003c/span\u003e\u003cspan address=\"10.1016/S0140-6736(21)01374-X\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAsrani SK, Devarbhavi H, Eaton J, Kamath PS. Burden of liver diseases in the world. J Hepatol. 2019;70(1):151\u0026ndash;71. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jhep.2018.09.014\u003c/span\u003e\u003cspan address=\"10.1016/j.jhep.2018.09.014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTapper EB, Parikh ND. Diagnosis and Management of Cirrhosis and Its Complications: A Review. JAMA. 2023;329(18):1589\u0026ndash;602. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jama.2023.5997\u003c/span\u003e\u003cspan address=\"10.1001/jama.2023.5997\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChidambaranathan-Reghupaty S, Fisher PB, Sarkar D. Hepatocellular carcinoma (HCC): Epidemiology, etiology and molecular classification. Adv Cancer Res. 2021;149:1\u0026ndash;61. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/bs.acr.2020.10.001\u003c/span\u003e\u003cspan address=\"10.1016/bs.acr.2020.10.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBataller R, Brenner DA. Liver fibrosis. J Clin Invest. 2005;115(2):209\u0026ndash;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1172/JCI24282\u003c/span\u003e\u003cspan address=\"10.1172/JCI24282\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZeng S, Liu Z, Ke B, Zhang Y, Wang Q, Tan S. The non-invasive serum biomarkers contributes to indicate liver fibrosis staging and evaluate the progress of chronic hepatitis B. BMC Infect Dis. 2024;24(1):638. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12879-024-09465-z\u003c/span\u003e\u003cspan address=\"10.1186/s12879-024-09465-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJiang C, Liu J, Li R, et al. Clinicopathologic characteristics of liver inflammation and fibrosis in 310 patients with chronic hepatitis B. Zhong Nan Da Xue Xue Bao Yi Xue Ban. 2023;48(5):698\u0026ndash;706. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.11817/j.issn.1672-7347.2023.220622\u003c/span\u003e\u003cspan address=\"10.11817/j.issn.1672-7347.2023.220622\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLuo P, Wu S, Yu Y, et al. Current Status and Perspective Biomarkers in AFP Negative HCC: Towards Screening for and Diagnosing Hepatocellular Carcinoma at an Earlier Stage. Pathol Oncol Res. 2020;26(2):599\u0026ndash;603. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s12253-019-00585-5\u003c/span\u003e\u003cspan address=\"10.1007/s12253-019-00585-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHanif H, Ali MJ, Susheela AT, et al. Update on the applications and limitations of alpha-fetoprotein for hepatocellular carcinoma. World J Gastroenterol. 2022;28(2):216\u0026ndash;29. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3748/wjg.v28.i2.216\u003c/span\u003e\u003cspan address=\"10.3748/wjg.v28.i2.216\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAltinmakas E, Taouli B. Abbreviated Liver Magnetic Resonance Imaging Protocols and Applications. Radiol Clin North Am. 2022;60(5):695\u0026ndash;703. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.rcl.2022.04.002\u003c/span\u003e\u003cspan address=\"10.1016/j.rcl.2022.04.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMasokano IB, Liu W, Xie S, Marcellin DFH, Pei Y, Li W. The application of texture quantification in hepatocellular carcinoma using CT and MRI: a review of perspectives and challenges. Cancer Imaging. 2020;20(1):67. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s40644-020-00341-y\u003c/span\u003e\u003cspan address=\"10.1186/s40644-020-00341-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFoerster F, Gairing SJ, M\u0026uuml;ller L, Galle PR. NAFLD-driven HCC: Safety and efficacy of current and emerging treatment options. J Hepatol. 2022;76(2):446\u0026ndash;57. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jhep.2021.09.007\u003c/span\u003e\u003cspan address=\"10.1016/j.jhep.2021.09.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLlovet JM, Pinyol R, Kelley RK, et al. Molecular pathogenesis and systemic therapies for hepatocellular carcinoma. Nat Cancer. 2022;3(4):386\u0026ndash;401. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s43018-022-00357-2\u003c/span\u003e\u003cspan address=\"10.1038/s43018-022-00357-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBataller R, North KE, Brenner DA. Genetic polymorphisms and the progression of liver fibrosis: a critical appraisal. Hepatology. 2003;37(3):493\u0026ndash;503. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1053/jhep.2003.50127\u003c/span\u003e\u003cspan address=\"10.1053/jhep.2003.50127\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eToh MR, Wong EYT, Wong SH, et al. Global Epidemiology and Genetics of Hepatocellular Carcinoma. Gastroenterology. 2023;164(5):766\u0026ndash;82. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1053/j.gastro.2023.01.033\u003c/span\u003e\u003cspan address=\"10.1053/j.gastro.2023.01.033\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShah PA, Patil R, Harrison SA. NAFLD-related hepatocellular carcinoma: The growing challenge. Hepatology. 2023;77(1):323\u0026ndash;38. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hep.32542\u003c/span\u003e\u003cspan address=\"10.1002/hep.32542\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBabariya H, Gaidhane SA, Acharya S, Kumar S. Pirfenidone as a Cornerstone in the Management of Fibrotic Interstitial Lung Diseases and Its Emerging Applications: A Comprehensive Review. Cureus. 2024;16(9):e70497. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7759/cureus.70497\u003c/span\u003e\u003cspan address=\"10.7759/cureus.70497\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFriedman SL. Liver fibrosis -- from bench to bedside. J Hepatol. 2003;38(Suppl 1):S38\u0026ndash;53. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/s0168-8278(02)00429-4\u003c/span\u003e\u003cspan address=\"10.1016/s0168-8278(02)00429-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMederacke I, Hsu CC, Troeger JS, et al. Fate tracing reveals hepatic stellate cells as dominant contributors to liver fibrosis independent of its aetiology. Nat Commun. 2013;4:2823. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/ncomms3823\u003c/span\u003e\u003cspan address=\"10.1038/ncomms3823\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTrivedi P, Wang S, Friedman SL. The Power of Plasticity-Metabolic Regulation of Hepatic Stellate Cells. Cell Metab. 2021;33(2):242\u0026ndash;57. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cmet.2020.10.026\u003c/span\u003e\u003cspan address=\"10.1016/j.cmet.2020.10.026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBates J, Vijayakumar A, Ghoshal S, et al. Acetyl-CoA carboxylase inhibition disrupts metabolic reprogramming during hepatic stellate cell activation. J Hepatol. 2020;73(4):896\u0026ndash;905. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jhep.2020.04.037\u003c/span\u003e\u003cspan address=\"10.1016/j.jhep.2020.04.037\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen X, Lv X, Han M, et al. EMP1 as a Potential Biomarker in Liver Fibrosis: A Bioinformatics Analysis. Gastroenterol Res Pract. 2023;2023:2479192. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1155/2023/2479192\u003c/span\u003e\u003cspan address=\"10.1155/2023/2479192\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang N, Zhu HP, Huang W, et al. Unraveling the structures, functions and mechanisms of epithelial membrane protein family in human cancers. Exp Hematol Oncol. 2022;11(1):69. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s40164-022-00321-x\u003c/span\u003e\u003cspan address=\"10.1186/s40164-022-00321-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTaylor V, Welcher AA, Program AE, Suter U. Epithelial membrane protein-1, peripheral myelin protein 22, and lens membrane protein 20 define a novel gene family. J Biol Chem. 1995;270(48):28824\u0026ndash;33. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1074/jbc.270.48.28824\u003c/span\u003e\u003cspan address=\"10.1074/jbc.270.48.28824\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYasukawa K, Shimizu A, Kubota K, et al. Impact of Liver Fibrosis Severity on Oncological Prognosis in Hepatocellular Carcinoma. Liver Cancer. 2024;13(2):150\u0026ndash;60. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1159/000533857\u003c/span\u003e\u003cspan address=\"10.1159/000533857\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVilar-Gomez E, Calzadilla-Bertot L, Wai-Sun Wong V, et al. Fibrosis Severity as a Determinant of Cause-Specific Mortality in Patients With Advanced Nonalcoholic Fatty Liver Disease: A Multi-National Cohort Study. Gastroenterology. 2018;155(2):443\u0026ndash;e45717. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1053/j.gastro.2018.04.034\u003c/span\u003e\u003cspan address=\"10.1053/j.gastro.2018.04.034\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi B, Hu J, Xu H. Integrated single-cell and bulk RNA sequencing reveals immune-related SPP1\u0026thinsp;+\u0026thinsp;macrophages as a potential strategy for predicting the prognosis and treatment of liver fibrosis and hepatocellular carcinoma. Front Immunol. 2024;15:1455383. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fimmu.2024.1455383\u003c/span\u003e\u003cspan address=\"10.3389/fimmu.2024.1455383\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQin H, Abulaiti A, Maimaiti A, et al. Integrated machine learning survival framework develops a prognostic model based on inter-crosstalk definition of mitochondrial function and cell death patterns in a large multicenter cohort for lower-grade glioma. J Transl Med. 2023;21(1):588. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12967-023-04468-x\u003c/span\u003e\u003cspan address=\"10.1186/s12967-023-04468-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang Zyuan, Qian J, Lu X et al. A single-cell transcriptomic atlas characterizes liver non-parenchymal cells in healthy and diseased mice. \u003cem\u003ebioRxiv\u003c/em\u003e. Published online 2021. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://api.semanticscholar.org/CorpusID:235779889\u003c/span\u003e\u003cspan address=\"https://api.semanticscholar.org/CorpusID:235779889\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen J, Gao D, Chen J, et al. Pseudo-Temporal Analysis of Single-Cell RNA Sequencing Reveals Trans-Differentiation Potential of Greater Epithelial Ridge Cells Into Hair Cells During Postnatal Development of Cochlea in Rats. Front Mol Neurosci. 2022;15:832813. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fnmol.2022.832813\u003c/span\u003e\u003cspan address=\"10.3389/fnmol.2022.832813\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCheng C, Chen W, Jin H, Chen X. A Review of Single-Cell RNA-Seq Annotation, Integration, and Cell-Cell Communication. Cells. 2023;12(15). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/cells12151970\u003c/span\u003e\u003cspan address=\"10.3390/cells12151970\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eH\u0026auml;nzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/1471-2105-14-7\u003c/span\u003e\u003cspan address=\"10.1186/1471-2105-14-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZheng M, Hu Y, Liu O, et al. Oxidative Stress Response Biomarkers of Ovarian Cancer Based on Single-Cell and Bulk RNA Sequencing. Oxid Med Cell Longev. 2023;2023:1261039. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1155/2023/1261039\u003c/span\u003e\u003cspan address=\"10.1155/2023/1261039\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCheng S, Zou Y, Zhang M, et al. Single-cell RNA sequencing reveals the heterogeneity and intercellular communication of hepatic stellate cells and macrophages during liver fibrosis. MedComm (2020). 2023;4(5):e378. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/mco2.378\u003c/span\u003e\u003cspan address=\"10.1002/mco2.378\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIwaisako K, Jiang C, Zhang M, et al. Origin of myofibroblasts in the fibrotic liver in mice. Proc Natl Acad Sci U S A. 2014;111(32):E3297\u0026ndash;3305. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1073/pnas.1400062111\u003c/span\u003e\u003cspan address=\"10.1073/pnas.1400062111\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKoyama Y, Wang P, Liang S, et al. Mesothelin/mucin 16 signaling in activated portal fibroblasts regulates cholestatic liver fibrosis. J Clin Invest. 2017;127(4):1254\u0026ndash;70. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1172/JCI88845\u003c/span\u003e\u003cspan address=\"10.1172/JCI88845\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDooley S, Delvoux B, Streckert M, et al. Transforming growth factor beta signal transduction in hepatic stellate cells via Smad2/3 phosphorylation, a pathway that is abrogated during in vitro progression to myofibroblasts. TGFbeta signal transduction during transdifferentiation of hepatic stellate cells. FEBS Lett. 2001;502(1\u0026ndash;2):4\u0026ndash;10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/s0014-5793(01)02656-4\u003c/span\u003e\u003cspan address=\"10.1016/s0014-5793(01)02656-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen H, Cai J, Wang J, et al. Targeting Nestin(+) hepatic stellate cells ameliorates liver fibrosis by facilitating TβRI degradation. J Hepatol. 2021;74(5):1176\u0026ndash;87. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jhep.2020.11.016\u003c/span\u003e\u003cspan address=\"10.1016/j.jhep.2020.11.016\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTrivedi P, Wang S, Friedman SL. The Power of Plasticity-Metabolic Regulation of Hepatic Stellate Cells. Cell Metab. 2021;33(2):242\u0026ndash;57. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cmet.2020.10.026\u003c/span\u003e\u003cspan address=\"10.1016/j.cmet.2020.10.026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eM\u0026eacute;sz\u0026aacute;ros K, Bojta J, Bautista AP, Lang CH, Spitzer JJ. Glucose utilization by Kupffer cells, endothelial cells, and granulocytes in endotoxemic rat liver. Am J Physiol. 1991;260(1 Pt 1):G7\u0026ndash;12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1152/ajpgi.1991.260.1.G7\u003c/span\u003e\u003cspan address=\"10.1152/ajpgi.1991.260.1.G7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQuiroz Reyes AG, Lozano Sepulveda SA, Martinez-Acu\u0026ntilde;a N, et al. Cancer Stem Cell and Hepatic Stellate Cells in Hepatocellular Carcinoma. Technol Cancer Res Treat. 2023;22:15330338231163677. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/15330338231163677\u003c/span\u003e\u003cspan address=\"10.1177/15330338231163677\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSangro B, Chan SL, Meyer T, Reig M, El-Khoueiry A, Galle PR. Diagnosis and management of toxicities of immune checkpoint inhibitors in hepatocellular carcinoma. J Hepatol. 2020;72(2):320\u0026ndash;41. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jhep.2019.10.021\u003c/span\u003e\u003cspan address=\"10.1016/j.jhep.2019.10.021\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAnstee QM, Reeves HL, Kotsiliti E, Govaere O, Heikenwalder M. From NASH to HCC: current concepts and future challenges. Nat Rev Gastroenterol Hepatol. 2019;16(7):411\u0026ndash;28. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41575-019-0145-7\u003c/span\u003e\u003cspan address=\"10.1038/s41575-019-0145-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHuang DQ, Singal AG, Kanwal F, et al. Hepatocellular carcinoma surveillance - utilization, barriers and the impact of changing aetiology. Nat Rev Gastroenterol Hepatol. 2023;20(12):797\u0026ndash;809. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41575-023-00818-8\u003c/span\u003e\u003cspan address=\"10.1038/s41575-023-00818-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCheng S, Zou Y, Zhang M, et al. Single-cell RNA sequencing reveals the heterogeneity and intercellular communication of hepatic stellate cells and macrophages during liver fibrosis. MedComm (2020). 2023;4(5):e378. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/mco2.378\u003c/span\u003e\u003cspan address=\"10.1002/mco2.378\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWallace MC, Friedman SL. Hepatic fibrosis and the microenvironment: fertile soil for hepatocellular carcinoma development. Gene Expr. 2014;16(2):77\u0026ndash;84. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3727/105221614X13919976902057\u003c/span\u003e\u003cspan address=\"10.3727/105221614X13919976902057\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEbrahimi F, Hagstr\u0026ouml;m H, Sun J, et al. Familial coaggregation of MASLD with hepatocellular carcinoma and adverse liver outcomes: Nationwide multigenerational cohort study. J Hepatol. 2023;79(6):1374\u0026ndash;84. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jhep.2023.08.018\u003c/span\u003e\u003cspan address=\"10.1016/j.jhep.2023.08.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSigorski D, Sejda A, Abualsaud N, Krawczyk E, Izycka-Swieszewska E, Kitlinska J. Neuropeptide Y in cancer-biological functions and potential clinical implications. Cancer Metastasis Rev. 2025;44(1):21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10555-024-10237-z\u003c/span\u003e\u003cspan address=\"10.1007/s10555-024-10237-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLv X, Zhao F, Huo X, et al. Neuropeptide Y1 receptor inhibits cell growth through inactivating mitogen-activated protein kinase signal pathway in human hepatocellular carcinoma. Med Oncol. 2016;33(7):70. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s12032-016-0785-1\u003c/span\u003e\u003cspan address=\"10.1007/s12032-016-0785-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLin T, Lin Z, Mai P, Zhang E, Peng L. Identification of prognostic biomarkers associated with the occurrence of portal vein tumor thrombus in hepatocellular carcinoma. Aging. 2021;13(8):11786\u0026ndash;807. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18632/aging.202876\u003c/span\u003e\u003cspan address=\"10.18632/aging.202876\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhu Y, Zhang P. Gene expression profile of anoikis reveals new subtypes of liver cancer and discovery of therapeutic targets and biomarkers. Sci Rep. 2025;15(1):12740. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-025-96488-4\u003c/span\u003e\u003cspan address=\"10.1038/s41598-025-96488-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSial N, Ahmad M, Hussain MS, et al. CTHRC1 expression is a novel shared diagnostic and prognostic biomarker of survival in six different human cancer subtypes. Sci Rep. 2021;11(1):19873. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-021-99321-w\u003c/span\u003e\u003cspan address=\"10.1038/s41598-021-99321-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi J, Wang Y, Ma M et al. Corrigendum to Autocrine CTHRC1 activates hepatic stellate cells and promotes liver fibrosis by activating TGF-β signaling [EBioMedicine 40 (2019) 43\u0026ndash;55]. \u003cem\u003eEBioMedicine\u003c/em\u003e. 2024;107:105303. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ebiom.2024.105303\u003c/span\u003e\u003cspan address=\"10.1016/j.ebiom.2024.105303\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi X, Zhang CC, Lin XT, et al. Elevated expression of WSB2 degrades p53 and activates the IGFBP3-AKT-mTOR-dependent pathway to drive hepatocellular carcinoma. Exp Mol Med. 2024;56(1):177\u0026ndash;91. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s12276-023-01142-6\u003c/span\u003e\u003cspan address=\"10.1038/s12276-023-01142-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSong M, Pan Q, Yang J, et al. Galectin-3 favours tumour metastasis via the activation of β-catenin signalling in hepatocellular carcinoma. Br J Cancer. 2020;123(10):1521\u0026ndash;34. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41416-020-1022-4\u003c/span\u003e\u003cspan address=\"10.1038/s41416-020-1022-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePlayford MP, Bicknell D, Bodmer WF, Macaulay VM. Insulin-like growth factor 1 regulates the location, stability, and transcriptional activity of beta-catenin. Proc Natl Acad Sci U S A. 2000;97(22):12103\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1073/pnas.210394297\u003c/span\u003e\u003cspan address=\"10.1073/pnas.210394297\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eArab JP, Cabrera D, Sehrawat TS, et al. Hepatic stellate cell activation promotes alcohol-induced steatohepatitis through Igfbp3 and SerpinA12. J Hepatol. 2020;73(1):149\u0026ndash;60. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jhep.2020.02.005\u003c/span\u003e\u003cspan address=\"10.1016/j.jhep.2020.02.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu T, Zhang F, Feng Y, Han P, Gao Y. Alcohol-Metabolizing Enzymes, Liver Diseases and Cancer. Semin Liver Dis. 2025;45(1):99\u0026ndash;113. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1055/a-2551-3320\u003c/span\u003e\u003cspan address=\"10.1055/a-2551-3320\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMello T, Ceni E, Surrenti C, Galli A. Alcohol induced hepatic fibrosis: role of acetaldehyde. Mol Aspects Med. 2008;29(1\u0026ndash;2):17\u0026ndash;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.mam.2007.10.001\u003c/span\u003e\u003cspan address=\"10.1016/j.mam.2007.10.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFriedman SL. Stellate cell activation in alcoholic fibrosis\u0026ndash;an overview. Alcohol Clin Exp Res. 1999;23(5):904\u0026ndash;10.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePeng Y, Yang T, Huang K, Shen L, Tao Y, Liu C. Salvia Miltiorrhiza Ameliorates Liver Fibrosis by Activating Hepatic Natural Killer Cells in Vivo and in Vitro. Front Pharmacol. 2018;9:762. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fphar.2018.00762\u003c/span\u003e\u003cspan address=\"10.3389/fphar.2018.00762\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang P, Liu W, Wang Y. The mechanisms of tanshinone in the treatment of tumors. Front Pharmacol. 2023;14:1282203. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fphar.2023.1282203\u003c/span\u003e\u003cspan address=\"10.3389/fphar.2023.1282203\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"journal-of-translational-medicine","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jtrm","sideBox":"Learn more about [Journal of Translational Medicine](http://translational-medicine.biomedcentral.com)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/jtrm/default.aspx","title":"Journal of Translational Medicine","twitterHandle":"@BioMedCentral","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"EMP1, Hepatic fibrosis, Hepatocellular carcinoma, Prediction model, Treatment Bioinformatics","lastPublishedDoi":"10.21203/rs.3.rs-7278297/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7278297/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eHepatic fibrosis is a pathological response to chronic liver injury that results in accumulation of extracellular matrix proteins leading to fibrous scarring, which can further lead to liver failure and hepatocellular carcinoma (HCC) Although several clinical approaches have been applied to the diagnosis and treatment of hepatic fibrosis and HCC, the clinical prognosis and precision of targeted therapies still face great challenges.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eIn this study, we integrated single-cell sequencing analysis and bulk sequencing analysis to identify genes, cellular subpopulations, and signalling pathways that are closely related to and highly expressed in hepatic fibrosis and HCC. On this basis, clinical prediction models and prognostic genes were constructed and validated by combining single-cell analysis with bulk differential gene analysis in the TCGA database, using 101 machine learning approaches, combined with survival analysis tools, and making full use of clinical data. In addition, the expression heterogeneity of core prognostic genes and their correlation with prognostic outcomes were explored in depth, and new targeted therapeutic modalities were sought with the help of comprehensive and systematic network pharmacological analyses to identify drugs that can target core prognostic genes.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eWe identified 45 HSC-associated pathogenic genes and an EMP1\u0026thinsp;+\u0026thinsp;HSC subpopulation, along with their regulatory signaling pathways linked to energy metabolism, cell adhesion, and extracellular matrix organization. These pathways were found to contribute to hepatic fibrosis and HCC progression. Subsequently, we validated four core prognostic genes (NPY1R, CTHRC1, IGFBP3, and ADH1B) and analyzed the heterogeneity of their expression patterns, demonstrating their correlation with hepatic fibrosis progression and HCC prognosis. Finally, through a systematic screening of bioactive compounds from traditional Chinese medicine (TCM) with potential anti-liver disease effects, we determined that Salvia miltiorrhiza(Danshen) specifically interacts with these core prognostic targets, offering a novel therapeutic strategy for hepatic fibrosis and HCC.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e\u003cp\u003eThis integrative study establishes EMP1 as a reliable biomarker for activated HSCs and identifies four core prognostic genes (NPY1R, CTHRC1, IGFBP3, and ADH1B) that play critical roles in the fibrosis-to-HCC progression and demonstrate significant clinical relevance to long-term patient outcomes. Our findings provide novel mechanistic insights into hepatic fibrogenesis and HCC development, while simultaneously revealing Salvia miltiorrhiza (Danshen) as a promising therapeutic agent targeting these key molecular pathways. These discoveries offer a dual advancement in both diagnostic precision and treatment strategy for hepatic fibrosis and HCC.\u003c/p\u003e","manuscriptTitle":"EMP1+ hepatic stellate cells drive hepatic fibrosis progression to hepatocellular carcinoma and predict prognosis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-27 05:56:27","doi":"10.21203/rs.3.rs-7278297/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2025-08-15T22:44:18+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-15T22:39:21+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-05T14:36:36+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of Translational Medicine","date":"2025-08-02T08:39:15+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-translational-medicine","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jtrm","sideBox":"Learn more about [Journal of Translational Medicine](http://translational-medicine.biomedcentral.com)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/jtrm/default.aspx","title":"Journal of Translational Medicine","twitterHandle":"@BioMedCentral","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"689cc04a-aee1-466c-b7b5-3d0f4679a179","owner":[],"postedDate":"August 27th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-12-08T16:05:26+00:00","versionOfRecord":{"articleIdentity":"rs-7278297","link":"https://doi.org/10.1186/s12967-025-07454-7","journal":{"identity":"journal-of-translational-medicine","isVorOnly":false,"title":"Journal of Translational Medicine"},"publishedOn":"2025-12-01 15:57:38","publishedOnDateReadable":"December 1st, 2025"},"versionCreatedAt":"2025-08-27 05:56:27","video":"","vorDoi":"10.1186/s12967-025-07454-7","vorDoiUrl":"https://doi.org/10.1186/s12967-025-07454-7","workflowStages":[]},"version":"v1","identity":"rs-7278297","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7278297","identity":"rs-7278297","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}