Analysis and validation of hub genes and immune cell infiltration characteristics in lens epithelium tissue with cataract surgical history

Analysis and validation of hub genes and immune cell infiltration characteristics in lens epithelium tissue with cataract surgical history | 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 Analysis and validation of hub genes and immune cell infiltration characteristics in lens epithelium tissue with cataract surgical history Shiming Jiang, Chao Gao This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8641774/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Purpose This study aims to employ transcriptomic analysis to investigate the characteristics of immune cell infiltration during Posterior capsular opacification (PCO) following cataract surgery, identify hub genes, elucidate the associated molecular pathways, and explore the potential interventional role of Cyclosporine A (CsA) in this process. Methods Data from lens capsules of non-cataract donors and those with cataract surgery history were obtained from the GEO database. Hub genes were identified through a systematic bioinformatic approach. The final hub genes were validated internally and with external datasets. Their correlations with immune profiles were assessed via CIBERSORT, and pathway enrichment analysis was conducted for immune-related genes. Cellular experiments simulated pathological conditions to validate hub gene expression, and molecular docking evaluated the binding affinity of CsA to the encoded proteins. Results Hub genes (FN1, SERPINE1, THBS1) were identified and validated. Their expression was associated with significant alterations in immune cell infiltration (resting CD4 + memory T cells, M0 macrophages, CD8 + T cells), which correlated with the genes and PCO-related pathways. CsA exhibited favorable binding affinity to proteins encoded by FN1 and SERPINE1. Conclusions Hub genes demonstrated strong discriminatory power in the gene expression profiles of lens capsules with a history of cataract surgery, occupying central positions within the co-expression network. Alterations in immune cell infiltration, closely linked to these hub genes, are significantly associated with pathways involved in PCO. CsA may inhibit PCO by targeting the proteins encoded by these hub genes and through its immunomodulatory functions. posterior capsular opacification inflammation RNA sequencing transcriptomics lens epithelial cells Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 1. Introduction Cataract, characterized by opacification of the crystalline lens, is the primary contributor to global blindness. This ocular structure maintains transparency through its unique composition of lens epithelial cells (LECs) and elongated fiber cell 1 , 2 . Epidemiological studies highlight cataract as a major cause of blindness in regions such as India, China, and sub-Saharan Africa, with an estimated 16 million people globally affected by blindness due to this condition 3 . Phacoemulsification combined with posterior chamber intraocular lens implantation is the primary surgical treatment for cataract 4 . However, a common complication following this procedure is PCO, which remains a significant challenge in ophthalmic surgery 5 . Following cataract surgery, LECs may undergo abnormal differentiation, leading to enhanced migratory capacity. These cells can relocate to the surface of the lens capsule and proliferate. This pathological process is likely triggered by direct surgical injury to LECs and alterations in the cellular microenvironment resulting from a breakdown of the blood–aqueous barrier 6 – 8 . Ultimately, when the aforementioned pathological changes progress to a sufficiently severe stage, they can lead to visual impairment. Research data indicate that the proportion of patients requiring laser capsulotomy within three years following standard cataract surgery ranges between 5% and 20% 5, 9, 10 . Despite being a well-recognized phenomenon, the precise molecular mechanisms and hub gene driving these cellular processes are not yet fully understood. The disruption of the blood-aqueous barrier following cataract surgery, combined with the implantation of pseudophakia, leads to the infiltration of inflammatory cells and immune mediators into the anterior chamber, a site of which was considered immunologically privileged 11 , 12 . Elevated expression of proinflammatory genes and proteins has been observed postoperatively in both clinical patients and murine cataract surgery models 13 , 14 . Studies analyzing aqueous humor from donors who underwent cataract surgery 3 to 12 years prior have revealed persistently elevated levels of Inflammatory cytokines, indicating that a chronic inflammatory state can persist long after the initial surgical intervention 15 . Inflammatory cytokines have been implicated in driving pathological changes in lens LECs 16 – 19 . While immunofluorescence studies have demonstrated the colocalization of inflammatory cell infiltration with phenotypic and metabolic changes in LECs, along with their correlation to fibrotic marker expression, the exact molecular pathways driving these interactions remain poorly understood 20 . CsA, a calcineurin inhibitor, exerts its immunosuppressive effects by selectively modulating T-cell infiltration and activation 21 . In ophthalmology, topical CsA is widely utilized to manage ocular surface inflammation, particularly in dry eye disease resulting from various etiologies such as cataract surgery, lacrimal dysfunction, and ocular surface impairment 22 , 23 .Evidence from animal studies has demonstrated that an intraocular lens modified with sustained-release CsA can significantly reduce anterior chamber inflammation and delay the onset of PCO following cataract surgery 24 , 25 . However, these findings remain predominantly at the phenomenological level, and a thorough investigation into the molecular mechanisms by which CsA inhibits PCO progression is still lacking. As a gene expression analysis method with high accuracy and a broad dynamic range, high-throughput sequencing, by integrating statistics, biological databases, and machine learning algorithms, has been widely applied across various disease domains to uncover underlying pathway alterations, key bioactive molecules, and functional targets in pathological tissues and cells 26 – 30 . Immune infiltration analysis further reveals the correlation between hub genes and immune cell 27 – 30 . However, there remains a lack of systematic bioinformatics studies, particularly immune infiltration analyses, on PCO following cataract surgery. 2. Methods 2.1. Data source The high-throughput sequencing data derived from lens capsules of non-cataract donors and donors with a history of cataract surgery were downloaded from the GEO database under the GSE295383 31 . This dataset comprised samples from 17 donors without cataracts and 17 donors with a history of cataract surgery. Lens capsules obtained from donors who had no documented history of cataracts, other ocular diseases, or diabetes were categorized as the non-cataract donors. Those acquired from donors with a prior cataract surgery but without diabetes or other ocular pathologies were classified into the post-cataract surgery group. All tissues were collected from donor globes within 2 to 10 hours postmortem; all lens capsule samples were selected for subsequent bioinformatic analysis. Then, we retrieved high-throughput sequencing data GSE111430 and GSE119879 from the GEO database as external datasets to validate the ultimately screened hub genes. 14, 20 . These datasets contained transcriptomic profiles from Mus musculus specimens, which included control groups (0h) and experimental groups at 24h and 48h following cataract surgery. 2.2. Methodology The key points of this research method are summarized in Fig. 1. 2.3. Differential expressed genes (DEGs) analysis A comparative analysis was conducted to examine differences between donors without cataracts and those who have undergone cataract surgery, utilizing the limma package (v3.60.6) within R software (v4.4.1) 32 . DEGs were determined using the criteria of |logFC| > 1 and P -value < 0.05. Visualization of the results from differential expression analysis was accomplished through the "ggplot2" package (v3.5.2) in R software, which produced volcano plots. 2.4. Enrichment analysis For functional enrichment analysis of gene sets, the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) annotation data from the org.Hs.eg.db package (v3.20.0) in R software were employed as the background reference 33 . Genes were mapped onto this background dataset, and enrichment procedures were performed using the clusterProfiler (v4.14.6) package in R software. 2.5. Weighted gene co-expression network analysis (WGCNA) We employed the WGCNA R package (v1.73) for WGCNA 34 , 35 . Initially, hierarchical clustering was carried out using the hclust function to assess sample outliers. The analysis utilized a soft-thresholding power of 8 for calculating biweight midcorrelations to construct signed networks. Additional parameters included TOMType = "unsigned", mergeCutHeight = 0.35, and minModuleSize = 50 to identify cohesive gene modules. Finally, the association between module eigengenes and status was evaluated using Pearson correlation analysis. 2.6. Protein-protein interaction (PPI ) network analysis The PPI network is a graph model comprising nodes (proteins) and edges (interactions between them) 36 . For network construction, a PPI network were retrieved from the STRING, for Homo sapiens , applying a confidence score cutoff set above 0.400. After removing unrelated protein entries, topological analysis was performed using the cytoHubba (v3.10.3) plugin, employing three distinct algorithms: Degree, Maximum Neighborhood Component (MNC), and Maximal Clique Centrality (MCC). The genes consistently ranked in the top 10 by all three algorithms were defined as candidate key genes for subsequent investigation. 2.7. Screening hub genes by machine learning Three distinct machine learning algorithms were implemented: least absolute shrinkage and selection operator (LASSO) regression, support vector machine-recursive feature elimination (SVM-RFE), and random forest (RF) 37 – 40 . A random seed (123456) was set at the outset to ensure the reproducibility of results. 2.7.1. LASSO regression The LASSO analysis was conducted using the glmnet package (v4.1.10) in R software. Model training was performed using the glmnet function, and the optimal regularization parameter was determined via 5-fold cross-validation. Features (genes expression) with non-zero coefficients under the optimal lambda value were retained as candidate feature genes. 2.7.2. SVM-RFE The SVM-RFE algorithm was utilized for its recursive feature ranking capability. Using the e1071package (v1.7.16) in R software, the process involved iteratively removing the least important genes as determined by the SVM model weights. The feature subset that yielded the highest classification accuracy during 5-fold cross-validation was selected as the final set of candidate genes. 2.7.3. RF The RF algorithm was implemented via the randomForest package (v4.7.1.2) in R software, with the number of trees set to 200. Feature importance was quantified using the Mean Decrease in Gini Index, which reflects the contribution of each feature to node impurity reduction across all trees. Gene expression characteristics with score greater than 1.5 were considered candidate feature genes. 2.8. Estimation of hub genes A multivariable logistic regression model was fitted to the finalized hub gene expression profiles using the R package rms (v8.0.0) in R software. A nomogram was constructed to visually represent the interactions among the selected genes 41 . The discriminative ability of the multivariable logistic regression model was assessed by the ROC (Receiver Operating Characteristic) curve. The optimal cutoff point on the ROC curve was determined by maximizing Youden’s index 42 . A confusion matrix was subsequently generated based on this threshold to evaluate classification performance, including metrics such as sensitivity, specificity, and accuracy 43 . The datasets GSE111430 and GSE119879 were utilized. After normalization processing. The data conforms to a normal distribution upon normality testing. independent two-sample t-tests were applied to assess whether the expression of hub genes exhibited statistically significant differences between comparative groups. 2.9. Immune infiltration analysis The relative abundances of human immune cell populations within the samples were deconvoluted using the CIBERSORT algorithm and its built-in LM22 signature matrix 44 . To identify distribution of immune cell infiltration between two predefined sample groups, statistical comparisons for each immune cell subset were carried out using the Wilcoxon signed-rank test. 2.10. Gene Set Enrichment Analysis (GSEA) Data were downloaded from the GSEA database 45 . Genes were ranked based on the correlation between their expression and the infiltration scores of specific immune cells, and perform pathway analysis on the ranked gene list using the clusterProfiler package (v4.14.6) in R software. 2.11. Cells culture and treatment The human lens epithelial cells (SRA01/04) were purchased from Shanghai Zhong Qiao Xin Zhou Biotechnology Co., Ltd. (product number: ZQ0612) and were cultured at 37℃ with 5% CO 2 in the Dulbecco's modified Eagle's medium with 10% fetal bovine serum (10091148, Gibco, USA), 100 µg/mL streptomycin, 100 U/mL penicillin (SV30010, Hyclone, USA). For TGF-β2 treatments, the culture medium was replaced on the second day after plating with fresh medium containing recombinant human TGF-β2 at concentrations of 0, 1, or 10 ng/ml, followed by a 48h incubation period. 2.12. Real-time quantitative reverse transcription PCR(qRT-PCR) Total RNA was isolated from SRA01/04 using TRIZOL (15596-018, Ambion,USA). cDNA synthesis was done by reverse transcription kit (R323-01, Vazyme, China). Quantitative PCR was subsequently performed on an ABI QuantStudio 5 system. The reaction protocol included an initial denaturation step, followed by 40 cycles of denaturation, and a combined annealing/extension step. Each sample had three technical replicates. The concentration of each transcript was then normalized to GAPDH and mRNA level using 2 - ΔΔCT method to analysis. Statistical analysis was conducted with two-way ANOVA in GraphPad Prism software (v10.6.0). 2.13. Wound-healing assay The migration capability of SRA01/04 cells was assessed using a wound-healing assay. Cells were seeded in plates and grown to a high confluence. Before creating the wound, the cells were serum-starved for 4h. A uniform scratch was then introduced into the cell monolayer using a sterile pipette tip. Images of the wound were captured at the initial time point (0h) and after a 20h incubation period at marked locations. The migration of cells into the wound area was observed and photographed using an inverted microscope (MF52-N, Mshot, China). The percentage of wound closure was quantified by comparing the remaining cell-free area at 20 hours to the initial wound area at 0 hours. 2.14. Molecular docking The protein structures encoded by hub genes, along with 2D chemical structures of the selected small molecules, were retrieved from the PDB and PubChem databases. The 2D structures of these molecules were then converted into 3D formats and energetically minimized using the Minimize Energy​ function in Chem3D (v14.0.0.117). Molecular docking was performed using AutoDock​ (v1.5.7). Finally, the docking results were visualized using PyMOL (v2.6.2). 2.15. Statistical analysis Unless otherwise specified, all the statistical data were analyzed using R software (v4.4.1) and GraphPad Prism(v10.6.0). Appropriate statistical methods were selected and applied according to the specific experimental design and data characteristics of each analysis. The normality of data distribution was assessed using established tests, such as the Shapiro–Wilk and Kolmogorov–Smirnov tests in SPSS(v27.0) software. The homogeneity of variances, a key assumption for many parametric tests, was evaluated using appropriate methods such as Levene's test or F-test, depending on the number of groups and data distribution. In instances where data violated the assumption of normality or variance homogeneity, non-parametric alternatives were employed. For all inferential analyses, statistical significance was defined using the following thresholds: P * < 0.05, ​​ P ** < 0.01, and ​​ P *** < 0.001. 3. Results 3.1. DEGs analysis and functional enrichment analysis Following preprocessing steps including removal of low-expression genes and deduplication of the gene expression profile. Subsequently, sample clustering analysis was conducted to identify and remove outlier samples, such as FXJ13 (Figure 2A). A total of 264 DEGs, comprising 199 upregulated and 65 downregulated genes, were identified between human lens capsules with cataract surgical histories and non-surgical histories. These DEGs were visualized using a volcano plot and pheatmap (Figure 2B,2C). Subsequent functional enrichment analysis of these genes (Figure 2D) revealed significant associations in GO related to inflammatory cells and their cytokines, as well as pathways associated with cell-substrate junction and ECM organization. KEGG pathway analysis further indicated significant enrichment of pathways involved in PI3K−Akt signaling pathway, TNF signaling pathway, AGE−RAGE signaling pathway and TGF−beta signaling pathway. Data processing and identification of DEGs from GSE295383 Figure 2. (A) Sample clustering analysis was performed to exclude potential outliers, ensuring the reliability of subsequent analyses. (B) Volcano plot visualizing the DEGs. (C) Functional enrichment analysis of identified DEGs from GSE295383 using GO terms and KEGG pathways. 3.2. Identification of trait-related module genes To identify trait-related module genes, the WGCNA was employed. A soft-thresholding power of 8 was selected for calculating biweight midcorrelations to build signed networks (Figure 3A, 3B, 3C). After generating a heatmap illustrating module-trait correlations, the "brown" module was chosen for further analysis due to its highest correlation with the grouping trait while remaining independent of variations in gender and age (Figure 3D). Hub genes within this module were ultimately identified based on the criteria of Module Membership > 0.8 and Gene Significance > 0.5 (Figure 3E). Construction of the co-expression network. Figure 3. (A) Mean connectivity of module eigengenes across a range of soft-thresholding powers. (B) Scale-free topology fit index (signed R²) for various soft-thresholding powers. The red line indicates the chosen power value of 8, at which the model fit exceeds 0.85, suggesting an approximate scale-free topology. (C) Hierarchical clustering dendrogram of genes. Highly similar modules were dynamically identified and merged (a representative subset is shown here; the full dendrogram is provided in Figure S1). (D) Heatmap depicting Pearson correlations between module eigengenes and external traits. (E) Scatter plot illustrating the relationship between Module Membership and Gene Significance for genes within the brown module. 3.3. PPI network analysis To identify genes with hub biological significance from those obtained through statistical methods, a total of 86 genes, representing the intersection of DEGs and key modules from the WGCNA analysis (Figure 4A). Candidate genes were selected for construction of the PPI network. This combined approach aimed to reduce false positives and focus on high-potential candidates. PPI network was subsequently constructed using these genes via the STRING database (Figure 4B). Hub genes within this network were identified by applying three topological analysis algorithms MCC, Degree, and MNC through the cytoHubba plugin. Seven genes(FN1, THBS1,SERPINE1, COL1A1,SPP1, TIMP1,and BGN)were consistently identified among the top ten most central nodes across all three algorithms (Figure 4C). These seven hub genes were selected for further downstream analysis. PPI network analysis . Figure 4. (A) Venn diagram of key module genes versus DEGs. (B) PPI network constructed using genes screened under specific criteria, depicting functional associations and molecular relationships among encoded proteins. Color saturation and pattern dimensions are configured in accordance with the Degree. (C) Venn diagram identifying seven consensus hub genes determined by the intersection of three distinct algorithms: MCC, Degree and MNC. 3.4. Screening hub genes by machine learning To identify robust hub genes associated with status, we employed three distinct machine-learning algorithms for hub selection. LASSO regression analysis was applied to refine the candidate pool derived from univariate analysis, ultimately identifying three predictor genes (Figure 5A, 5B). Subsequently, a RF algorithm integrated with hub importance evaluation was utilized to assess the relationship between the classification error rate and predictive performance, which was further quantified using ROC curve analysis. Genes achieving an importance score exceeding 1.5 were retained as candidate genes (Figure 5C, 5D, 5E). The SVM-RFE method was implemented to pinpoint optimal hub genes based on maximal accuracy (Figure 5F). The final set of hub genes, selected through this multi-step computational approach and demonstrating strong relevance to surgical history, was advanced for further validation. Machine learning-based hub selection and model evaluation. Figure 5. (A) Coefficient Distribution Plot of LASSO Regression Analysis.(B) Optimal log(λ) value selected through minimum mean squared error criteria. (C) Ideal number of trees determined via error rate convergence in the RF algorithm.(D) ROC curve illustrating the classification performance of the RF-based model.(E) Hub importance scores derived from the RF model, indicating the contribution of each gene.(F) Maximum accuracy achieved by the SVM-RFE algorithm. 3.5. Validation of hub genes in Internal and external datasets Three distinct machine learning algorithms consistently identified FN1, SERPINE1, and THBS1 as robust hub genes (Figure 6A). Based on these genes, a multivariable logistic regression model was constructed using the rms package. A nomogram was generated to visually represent the interactions among these selected genes (Figure 6B). The multivariable logistic regression model performance was evaluated using ROC curve analysis and a confusion matrix determined by the optimal cutoff point (Figure 6C, 6D). Results demonstrated that the model exhibited strong discriminative ability and high specificity. Validation of hub genes in internal datasets Figure 6. (A) Venn diagram illustrating the consensus hub genes identified by machine-learning algorithms. (B) Nomogram constructed based on a multivariable logistic regression model incorporating the hub genes FN1, SERPINE1, and THBS1. (C) ROC of the model, indicating the optimal cutoff point for classification. (D) Confusion matrix evaluating the model’s predictive performance against actual outcomes. For external validation, datasets GSE111430 and GSE119879, which contain expression profiles from remnant lens epithelial cells of Mus musculus at 0h, 24h, and 48h after simulated cataract surgery, were downloaded and processed. Following data normalization and normality testing, two-sample t-tests were applied to assess whether the expression levels of each signature gene showed statistically significant differences between comparative groups (Figure 6E, 6F). The analysis confirmed that all three hub genes were significantly differentially expressed in both external datasets. Validation of hub genes in external datasets. Figure 6. (E) Expression profiles of signature genes in the GSE111430 dataset, derived from high-throughput sequencing of remnant lens epithelial cells in Mus musculus at 0h and 24h post-surgery. (F) Expression patterns of signature genes in the GSE119879 dataset, obtained from high-throughput sequencing of remnant lens epithelial cells in Mus musculu s at 0h and 48h post-surgery. 3.6. Immune infiltration analysis and Its association with hub genes To investigate the immune cell infiltration between the two sample groups and explore intercellular relationships, as well as to determine whether the expression of hub genes is associated with these immune cells, the global expression matrix was deconvoluted using the CIBERSORT algorithm to estimate the relative abundance of various immune cells (Figure 7A). The analysis revealed that samples with a surgical history exhibited a significantly higher fraction of resting CD4+ memory T cells ( P < 0.01) and M0 macrophages ( P < 0.05), while a lower fraction of CD8+ T cells ( P < 0.01) was observed (Figure 7B). Correlation heatmap analysis further demonstrated a strong negative association (r = −0.8) between resting CD4+ memory T cells and CD8+ T cells, and a positive correlation (r = 0.59) between M0 macrophages and M2 Macrophages (Figure 7C). Furthermore, all hub genes exhibited significant associations with resting CD4+ memory T cells, M0 macrophages, and CD8+ T cells, which displayed notable infiltration differences. These correlations followed the same trend as the overall infiltration pattern. (Figure 7D). Immune infiltration analysis . Figure 7. (A) Histogram displaying the relative proportions of immune cell subsets across samples. (B) Violin plots comparing the infiltration levels of immune cell subsets between history of surgery group and non−surgery group, with statistical significance indicated. (C) Heatmap visualizing the pairwise correlations among various immune cell subsets based on their relative infiltration abundances. (D) Correlation analysis evaluating the relationships between the expression levels of identified hub genes and the infiltration abundance of specific immune cell subsets. 3.7. Identification of signaling pathways associated with Immune Infiltration via GSEA Along with the biological pathways enriched by genes correlated with specific immune cells, we performed Spearman correlation analysis between each gene in the expression matrix and immune infiltration scores across all samples. Genes were ranked based on the values of their correlation coefficients for subsequent GSEA. The GSEA results indicated that genes positively correlated with the high proportion of immune cells in samples with a surgical history were significantly enriched in pathways promoting epithelial-mesenchymal transition (EMT), cell junction disassembly, and multiple EMT-related processes such as ERK1/ERK2 cascade, MAPK signaling, TGF-β production, and cellular response to TGF-β (Figure 8A, 8B). In terms of inflammation-related pathways, genes involved in chemotaxis of various immune cells and inflammatory responses triggered by injury were also positively associated with M0 macrophages and resting CD4+ memory T cells (Figure 8C, 8D). In contrast, infiltration of CD8+ T cells was negatively correlated with the enrichment of genes in these pathways (Figure 8E, 8F). Enrichment plots from GSEA . Figure 8. (A) GSEA enrichment plot for EMT-related pathways correlation with M0 macrophages infiltration. (B) GSEA enrichment plot for EMT-related pathways correlation with resting CD4+ memory T cells infiltration. (C) GSEA enrichment plot for inflammation-related pathways correlation with M0 macrophages infiltration. (D) GSEA enrichment plot for inflammation-related pathways correlation with resting CD4+ memory T cells infiltration. (E) GSEA enrichment plot for EMT-related pathways correlation with CD8+ T cells infiltration. (F) GSEA enrichment plot for inflammation-related pathways correlation with CD8+ T cells infiltration. 2.8 Expression levels of EMT-associated markers and hub genes The high concentration of TGF-β2 in the aqueous humor resulting from surgical trauma induces pathological changes in LECs, which is considered a key pathological environment for the development and progression of PCO 16, 46, 47 Our previous findings demonstrated a close correlation between TGF-β signaling pathway activation, upregulation of EMT-related pathways, and the expression of hub genes (FN1, SERPINE1, and THBS1). To further elucidate this relationship under controlled conditions, we focused on a well-defined single cell line model. Specifically, we validated the effect of a high TGF-β2 microenvironment on the induction of EMT and the expression of these hub genes in lens epithelial cells (LECs). The results showed that the expression of CDH1 (encoding the epithelial marker E-cadherin) decreased in a concentration-dependent manner with TGF-β2 treatment. In contrast, the expression of VIM (encoding the mesenchymal marker Vimentin) significantly increased only in the 10 ng/ml TGF-β2 environment compared to the control group(Figure 9A). All hub genes exhibited a concentration-dependent upregulation in expression(Figure 9B). The results of qRT-PCR. The qRT-PCR analysis of SRA01/04 was conducted with three biological replicates per condition and three technical replicates per sample, and the data were analyzed using the 2 - ΔΔCT method. Figure 9. (A) Expression levels of VIM and CDH1 under different treatment conditions. (B) Expression levels of hub genes under different treatment conditions. 2.9 Assessment of migration capability To determine the effect of TGF-β2 environment on the migration capability of LECs, we performed a wound-healing assay and evaluated the cell mobility rate. The results demonstrated that treatment with both 1 ng/ml and 10 ng/ml TGF-β2 significantly enhanced the migration capability of LECs compared to the control group. Wound-healing assay. Figure 10. (A) Representative images of the wound area at 0 h and 20 h post-scratching under various treatments. (B) Quantitative analysis of the cell mobility rate. 3.10. Molecular docking of CsA To elucidate the potential molecular mechanisms of CsA intervention in PCO, molecular docking analysis was performed. The results showed that the proteins encoded by the hub genes strongly interact with CsA, with binding energies of -5.5 kcal/mol for FN1, -6.4 kcal/mol for SERPINE1 and -4.7 kcal/mol for THBS1. Among these, the proteins encoded by FN1 and SERPINE1 were found to form hydrogen bonds with CsA (Figure 11A-B). Molecular docking results. The green structure represents the small molecule compound bisphenol A, the blue structure denotes the protein encoded by the gene, and the orange indicates the residues forming hydrogen bonds between bisphenol A and the protein. Figure 11.(A) The protein encoded by FN1 establishes hydrogen bonds with CsA via the residues V15 and T104. (B) The protein encoded by SERPINE1 interacts with CsA through hydrogen bonding at the residue HIS-279. 4. Discussion In general, the inflammatory microenvironment and disruption of the Injury stimulus during cataract surgery can trigger significant adaptive and remodeling responses in remnant lens epithelial cells 48 – 50 . Unlike previous studies that provided the source data, our study extends beyond simple differential expression analysis and individual gene assessments. this study integrated multiple analytical approaches to conduct a systematic, multi-gene dimensional analysis of high-throughput sequencing data derived from human lens capsules with a surgical history compared to non-surgical controls. Our results reveal that surgical history activates biological processes and molecular pathways, and identify hub genes (including FN1, SERPINE1, and THBS1) which occupy statistically and biologically central positions within the related gene co-expression network. The expression of these hub genes in LECs under stimulation by TGF-β2, a classic factor in PCO, was further validated through cell experiments. Furthermore, we investigated immune cell infiltration landscape and demonstrated that immune subsets with significant infiltration differences are associated with hub genes and biological pathways. Ultimately, we investigated the potential of CsA to intervene in these changes by targeting the proteins encoded by the identified hub genes. DEGs analysis using the limma package identified DEGs. GO Biological Process enrichment analysis revealed that DEGs were significantly involved in pathways related to myeloid leukocyte activation, myeloid leukocyte differentiation, and regulation of mononuclear cell migration. These findings align with the inflammatory response triggered by cataract surgery, as discussed earlier. A pathological feature of PCO is the loss of LECs integrity, which is closely associated with abnormal cell proliferation and migration. During this process, LECs progressively deviate from their ectodermal epithelial origin and transform into mesenchymal-like cells with more midline characteristics, a process known as EMT 51 . The newly generated mesenchymal-like cells exhibit weakened cell-substrate junctions, enhanced migratory and invasive capabilities, and excessive production of ECM components. Consistent with this mechanism, enrichment analyses of GO and KEGG pathways among DEGs also indicated alterations related to collagen production, ECM organization, cell-substrate adhesion, and other metabolic changes characteristic of EMT. Multiple EMT-associated signaling pathways were significantly enriched. Specifically, Enhanced TGF-β signaling has been demonstrated to induce EMT in LECs 14 , 52 , 53 , contributing to a cataract-related phenotype. For instance, Shizuya Saika et al 54 revealed that EMT in primary LECs under in vitro conditions depends on TGF-β expression, whereas in vivo, physical injury-induced EMT in LECs requires the involvement of the Smad3 pathway downstream of TGF-β. In addition to the canonical Smad pathway, the PI3K–Akt pathway, which significantly enriched among the DEGs,serves as a non-canonical route capable of inducing EMT 55 , 56 . TNF, which interacts extensively with TGF-β in tumor studies, also plays an important role in EMT 57 , 58 . Furthermore, AGE–RAGE signaling pathway can lead to the production of various inflammatory mediators and directly inducing EMT 59 , 60 . Studies also indicate that the AGE–RAGE signaling pathway promotes the expression of TGF-β 61 . In summary, a history of cataract surgery leads to significant alterations in the expression of multiple genes associated with EMT. Through integrated bioinformatics analysis and validation across internal and external datasets, we identified FN1, SERPINE1, and THBS1 as hub genes. Among these, FN1 encodes fibronectin, a well-established marker of EMT and a critical intervention target 62 – 65 . Previous studies have shown that miR-142-3P can suppress cancer cell migration, invasion, and EMT by targeting FN1 and inactivating the FAK/ERK/PI3K signaling pathway 66 . Similarly, THBS1 and SERPINE1 have also been widely reported to participate in fibrosis and EMT processes, primarily through activation of pathways such as TGF-β/Smad, PI3K/AKT/ERK 67 , 68 , and ERK–MAPK signaling 69 , 70 , which consistent with the enrichment and GSEA results in this study. In the context of PCO, FN1 remains recognized as an EMT marker 71 . Notably, a recent study suggested that targeting the HOTAIR/WTAP/THBS1 axis may prevent or treat PCO 72 . In contrast, SERPINE1 has not been thoroughly investigated in PCO-related studies. To explore the potential relationship between these hub genes and the inflammatory microenvironment commonly associated with PCO, we further performed immune infiltration analysis. Our immune infiltration analysis revealed significantly increased infiltration of M0 macrophages in the History of surgery group. Both M0 and M2 macrophages, which showed a positive correlation, have been implicated as key immune regulators in fibrosis and TGF-β signaling 73 . Macrophages expressing Fn1, and Arg1 are known to mediate fibrotic processes 74 . Since TGF-β in the lens microenvironment requires activation to exert its effects, macrophages may contribute through integrin-mediated TGF-β activation 47 , 75 . Consistent with this, study reported that macrophage infiltration promotes TGF-β/Smad signaling activity 20 . GSEA of genes associated with M0 macrophages infiltration further highlighted significant enrichment in pathways promoting EMT, including ERK1/ERK2, TGF-β, MAPK, and integrin mediated signalling. Additionally, terms such as inflammatory cell chemotaxis and inflammatory response to wounding were enriched, suggesting a link between M0 macrophages infiltration and a sustained pro-inflammatory state in the anterior chamber following surgery. CD8 + T cells and Resting CD4 + memory T cells exhibited statistically significant differences in infiltration ratios between the two groups. Notably, a strong negative association (r = − 0.8) was observed between resting CD8 + T cells and Resting CD4 + memory T cells. Previous studies have indicated that EMT activity correlates with increased CD8 + T cells and reduced Resting CD4 + memory T cells 76 , 77 . Furthermore, activation of TGF-β type II receptor signaling can lead to T cell CD8 exhaustion, while CD8 + T cells themselves have been reported to mediate anti-inflammatory responses, that consistent with our GSEA results 78 , 79 . Hub genes also exhibited correlations with immune infiltration that aligned with overall infiltration trends. Together, these findings underscore a profound interconnection between the immune microenvironment and multiple pathological processes in PCO, including the expression of hub genes central to disease progression. Subsequent molecular docking analyses of the proteins encoded by FN1, SERPINE1, and THBS1with CsA revealed that FN1 and SERPINE1exhibit favorable binding affinity with CsA and form stable hydrogen bonds. These results suggest that CsA may modulate their functions through these interactions, thereby affecting the EMT process associated with hub genes as well as the progression of TGF-β-related signaling pathways. These findings provide a mechanistic explanation for the previously reported inhibitory effect of CsA on PCO. Furthermore, the immunomodulatory properties of CsA, particularly its ability to reduce CD4 expression on T cells, suppress inflammatory cytokine release, and promote CD4 + T cell apoptosis, align with the modulation required for counteracting the aberrant inflammatory environment in PCO as identified in our analysis. Collectively, these findings indicate that CsA represents a promising therapeutic intervention for PCO. As discussed, a history of surgery creates an aqueous humor environment with elevated TGF-β2 concentrations for LECs. TGF-β2-activated pathways, recognized in previous studies and supported by our analysis, are key in stimulating pathological changes in LECs and contributing to PCO development. To model this critical pathological environment, we exposed LECs to varying concentrations of TGF-β2, aiming to investigate subsequent alterations in gene expression and cellular functions. Our results demonstrated a significant decrease in the epithelial marker CDH1 following TGF-β2 stimulation. In contrast, the mesenchymal marker VIM showed a significant increase only at a high concentration (10 ng/ml). This suggests that TGF-β2 can suppress epithelial marker expression and promote mesenchymal marker expression in LECs, with the regulation of epithelial markers appearing more sensitive. The loss of epithelial phenotype and acquisition of mesenchymal characteristics are hallmarks of EMT. The reduction of epithelial markers diminishes intercellular tight junctions, potentially enhancing migration capability and disrupting normal tissue architecture. Corroborating this, our wound-healing assay confirmed that TGF-β2 enhances the migration capability of LECs, indicating that TGF-β2 indeed stimulates EMT-like changes and alters cell motility. EMT involves complex physiological and pathological changes. It is not an absolute binary process. Cells undergoing functional changes can exhibit unexpected variations in marker expression, as observed in our experiments where epithelial and mesenchymal phenotypic shifts did not occur in a uniformly inverse manner 80 , 81 . In contrast, the genes FN1, SERPINE1, and THBS1, identified through our bioinformatic screening, showed significantly increased expression upon TGF-β2 stimulation. Their expression levels rose markedly with increasing TGF-β2 concentrations, demonstrating a strong concentration-dependent response. Consequently, FN1, SERPINE1, and THBS1 exhibit high sensitivity and correlation, positioning them as potential biomarkers that are more responsive to TGF-β2 exposure than classic EMT markers in LECs. Their association with immune cell infiltration and various pathological processes also suggests these genes may play critical roles in the pathological progression of LECs during PCO. However, this study has limitations regarding the sample conditions. As mentioned previously, the anterior chamber is an immune-privileged site; thus, the lens capsules from the non-surgery control group, without blood-aqueous barrier disruption caused by cataract surgery, should theoretically exhibit no immune cell infiltration. Nevertheless, CIBERSORT algorithm revealed the presence of immune cells in the Non-surgery group, with a notably higher proportion of CD8 + T cells compared to the History of surgery group. This observation may be attributed to the fact that all samples were taken from the donors 2 to 10 hours after their deaths. During this period, the postmortem breakdown of the blood-aqueous barrier likely leads to progressive loss of immune privilege, resulting in inflammatory cell infiltration even in normal anterior chambers 82 . Alternatively, the disruption of the blood-aqueous barrier may be attributed to the sample acquisition procedure itself. Furthermore, it is important to note that the CIBERSORT algorithm, which employs deconvolution analysis to infer the relative proportions of immune cells in a sample, cannot effectively distinguish whether the detected immune cells represent contamination at the time of extraction or are merely physically adjacent without physiological significance. To address this limitation, we integrated the immune cell proportions obtained from CIBERSORT with GSEA. This combined approach allowed us to examine the correlation between immune cell composition and the sample’s transcriptomic profile, thereby elucidating the impact on cellular pathways and clarifying the underlying biological significance. Taking this factor into account, and integrating findings from GSEA and immune infiltration analyses, a more refined interpretation can be drawn: the history of surgery group exhibits significant overexpression of pro-inflammatory genes, potentially promoting the chemotaxis and migration of M0 macrophages and resting CD4 + memory T cells. In contrast, the Non-surgery group may maintain a uniquely anti-inflammatory microenvironment, attenuating inflammatory responses while facilitating the infiltration of CD8 + T cells. This disparity in immune regulation likely underlies the differential activity of EMT-related pathways between the two groups. 5. Conclusion In conclusion, our findings demonstrate that FN1, SERPINE1, and THBS1 exhibit strong discriminatory power in the gene expression profiles of human lens epithelium tissue from patients with a history of cataract surgery. These hub genes occupy statistically and biologically central positions within the relevant gene co-expression network. Their expression is likely central to the pathological changes occurring in this tissue and may serve as potential biomarkers for identifying TGF-β2-stimulated LECs. Furthermore, the relative infiltration levels of M0 macrophages and resting CD4 + memory T cells within the tissue showed a significant positive correlation with both EMT and the activity of pro-inflammatory pathways. These findings suggest that their infiltration likely plays a critical role in driving postoperative pathological processes in LECs and is significantly associated with the expression of the identified hub genes. Ultimately, the potential and possible mechanism of action of CsA were revealed. Future research focused on elucidating the precise mechanisms through which these hub genes operate, and how specific immune cell populations are recruited and activated, will be essential for a deeper understanding of the molecular and cellular events that occur in LECs after surgery. Such insights also hold strong potential for developing targeted interventions aimed at modulating these key pathways or cell populations to effectively prevent the occurrence of PCO. Declarations Electronic supplementary material Supplementary charts and chart files are provided together. Acknowledgements We sincerely appreciate the GEO database for its platform support and the contributors who uploaded critical datasets. Author contributions SJ and CG conceived and designed the study, supervised data collection, and coordinated the research framework. SJ performed data processing and formal analysis. Both SJ and CG interpreted the results, drafted the initial manuscript, and revised it critically for intellectual content. All authors participated in substantive discussions, reviewed the final manuscript, approved its submission, and consented to publication. Availability of data The dataset used and/or analyzed in this study is available from the corresponding author upon reasonable request. Funding This work was supported by Bingtuan Science and Technology Program (2024ZD074) and the Project of TianChiYingCai of Xinjiang Uygur Autonomous Region (CZ001203). Clinical trial number: Not Applicable Consent for publication ： Not Applicable Ethics approval and consent to participate Our study utilizes data from the GEO database, which is publicly available. As such, our research complies with all applicable data use policies and presents no ethical issues Competing interests The authors declare no competing interests. References Bassnett, S., Shi, Y. & Vrensen, G.F. Biological glass: structural determinants of eye lens transparency. Philos Trans R Soc Lond B Biol Sci 366 , 1250-1264 (2011). Abdulhussein, D. & Abdul Hussein, M. WHO Vision 2020: Have We Done It? Ophthalmic Epidemiol 30 , 331-339 (2023). World Health Organization. Programme for the Prevention of, B. & Deafness (World Health Organization, 2000). Liu, Y.C., Wilkins, M., Kim, T., Malyugin, B. & Mehta, J.S. Cataracts. Lancet 390 , 600-612 (2017). Wormstone, I.M., Wormstone, Y.M., Smith, A.J.O. & Eldred, J.A. Posterior capsule opacification: What's in the bag? Prog Retin Eye Res 82 , 100905 (2021). Nibourg, L.M. et al. Prevention of posterior capsular opacification. Exp Eye Res 136 , 100-115 (2015). Way, C., Swampillai, A.J., Lim, K.S. & Nanavaty, M.A. Factors influencing aqueous flare after cataract surgery and its evaluation with laser flare photometry. Ther Adv Ophthalmol 15 , 25158414231204111 (2023). Wormstone, I.M. & Eldred, J.A. Experimental models for posterior capsule opacification research. Exp Eye Res 142 , 2-12 (2016). Leydolt, C. et al. Posterior Capsule Opacification With Two Hydrophobic Acrylic Intraocular Lenses: 3-Year Results of a Randomized Trial. Am J Ophthalmol 217 , 224-231 (2020). Ursell, P.G., Dhariwal, M., O'Boyle, D., Khan, J. & Venerus, A. 5 year incidence of YAG capsulotomy and PCO after cataract surgery with single-piece monofocal intraocular lenses: a real-world evidence study of 20,763 eyes. Eye (Lond) 34 , 960-968 (2020). Taravati, P., Lam, D.L., Leveque, T. & Van Gelder, R.N. Postcataract surgical inflammation. Curr Opin Ophthalmol 23 , 12-18 (2012). Jakobsson, G., Sundelin, K., Zetterberg, H. & Zetterberg, M. Increased levels of inflammatory immune mediators in vitreous from pseudophakic eyes. Invest Ophthalmol Vis Sci 56 , 3407-3414 (2015). Zhao, Y. et al. Expression Profiles of Inflammatory Cytokines in the Aqueous Humor of Children after Congenital Cataract Extraction. Transl Vis Sci Technol 9 , 3 (2020). Shihan, M.H. et al. Fibronectin has multifunctional roles in posterior capsular opacification (PCO). Matrix Biol 90 , 79-108 (2020). Hao, C. et al. Elevated Inflammatory Cytokines Persist in the Aqueous Humor Years After Cataract Surgery. Invest Ophthalmol Vis Sci 66 , 12 (2025). Wormstone, I.M., Tamiya, S., Anderson, I. & Duncan, G. TGF-beta2-induced matrix modification and cell transdifferentiation in the human lens capsular bag. Invest Ophthalmol Vis Sci 43 , 2301-2308 (2002). Lewis, A.C. Interleukin-6 in the pathogenesis of posterior capsule opacification and the potential role for interleukin-6 inhibition in the future of cataract surgery. Med Hypotheses 80 , 466-474 (2013). Flokis, M. & Lovicu, F.J. FGF-2 Differentially Regulates Lens Epithelial Cell Behaviour during TGF-β-Induced EMT. Cells 12 (2023). Si, W. et al. IL-8 promotes lens capsular residual cells migration by down-regulates expression of E-cadherin and ZO-1 via the CXCR1/2-NF-κB-RhoA signal pathway. Int Immunopharmacol 142 , 113074 (2024). Jiang, J., Shihan, M.H., Wang, Y. & Duncan, M.K. Lens Epithelial Cells Initiate an Inflammatory Response Following Cataract Surgery. Invest Ophthalmol Vis Sci 59 , 4986-4997 (2018). Gao, J. et al. Mitochondrial permeability transition pore in inflammatory apoptosis of human conjunctival epithelial cells and T cells: effect of cyclosporin A. Invest Ophthalmol Vis Sci 54 , 4717-4733 (2013). Periman, L.M., Mah, F.S. & Karpecki, P.M. A Review of the Mechanism of Action of Cyclosporine A: The Role of Cyclosporine A in Dry Eye Disease and Recent Formulation Developments. Clin Ophthalmol 14 , 4187-4200 (2020). Lu, M., Li, H. & Xiang, X. Efficacy of topical cyclosporine A in optimizing ocular surface and relieving dry eye symptoms after cataract surgery: a systematic review and meta-analysis of randomized controlled trials. Int Ophthalmol 45 , 201 (2025). Teng, H. et al. [The study of cyclosporin A modified intraocular lens preventing posterior capsular opacification in rabbit eyes]. Zhonghua Yan Ke Za Zhi 52 , 110-116 (2016). Teng, H., Sun, J., Wen, K., Han, G. & Tian, F. Observation of Cyclosporin A: Sustained Release Intraocular Lens Implantation in Rabbit Eyes. Curr Eye Res 47 , 1508-1515 (2022). Heid, C.A., Stevens, J., Livak, K.J. & Williams, P.M. Real time quantitative PCR. Genome Res 6 , 986-994 (1996). Peng, L. et al. Cellular and Molecular Network Characteristics of TARM1-Related Genes in Mycobacterium tuberculosis Infections. Int J Mol Sci 25 (2024). Wang, Z. et al. Identification of Key Biomarkers Associated with Immunogenic Cell Death and Their Regulatory Mechanisms in Severe Acute Pancreatitis Based on WGCNA and Machine Learning. Int J Mol Sci 24 (2023). Shen, F. et al. Integrative bioinformatics analysis of high-throughput sequencing and in vitro functional analysis leads to uncovering key hub genes in esophageal squamous cell carcinoma. Hereditas 162 , 38 (2025). Wang, H. et al. Potential of immune-related genes as promising biomarkers for premature coronary heart disease through high throughput sequencing and integrated bioinformatics analysis. Front Cardiovasc Med 9 , 893502 (2022). Hao, C. et al. Transcriptomic Analysis of Human Lens Epithelium Tissue With and Without Cataract Surgery: Uncovering Novel Pathways of Post-Surgical Lens Epithelium Remodeling. Invest Ophthalmol Vis Sci 66 , 28 (2025). Ritchie, M.E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 43 , e47 (2015). Kanehisa, M. & Goto, S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 28 , 27-30 (2000). Ke, Y. & Ge, W. Identification of prognostic biomarkers in neuroblastoma using WGCNA and multi-omics analysis. Discov Oncol 15 , 469 (2024). Langfelder, P. & Horvath, S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 9 , 559 (2008). Szklarczyk, D. et al. STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res 47 , D607-d613 (2019). Chen, K. et al. Identification of feature genes and key biological pathways in immune-mediated necrotizing myopathy: High-throughput sequencing and bioinformatics analysis. Comput Struct Biotechnol J 21 , 2228-2240 (2023). Geng, R. et al. Gene expression analysis in endometriosis: Immunopathology insights, transcription factors and therapeutic targets. Front Immunol 13 , 1037504 (2022). Zhang, W.Y. et al. Analysis and validation of diagnostic biomarkers and immune cell infiltration characteristics in pediatric sepsis by integrating bioinformatics and machine learning. World J Pediatr 19 , 1094-1103 (2023). Gong, T., Jia, B., Lv, H., Zeng, L. & Zhong, D. Computational Analyses Identified Three Diagnostic Biomarkers Associated With Programmed Cell Death for Lung Adenocarcinoma. Hum Mutat 2025 , 1743829 (2025). Shariat, S.F., Capitanio, U., Jeldres, C. & Karakiewicz, P.I. Can nomograms be superior to other prediction tools? BJU Int 103 , 492-495; discussion 495-497 (2009). Han, M. et al. Identification of hub genes and their correlation with immune infiltrating cells in membranous nephropathy: an integrated bioinformatics analysis. Eur J Med Res 28 , 525 (2023). Thilagar, S.S., Yadalam, P.K., Ronsivalle, V., Cicciù, M. & Minervini, G. Prediction of Interactomic HUB Genes in Periodontitis With Acute Myocardial Infarction. J Craniofac Surg 35 , 1292-1297 (2024). Le, T., Aronow, R.A., Kirshtein, A. & Shahriyari, L. A review of digital cytometry methods: estimating the relative abundance of cell types in a bulk of cells. Brief Bioinform 22 (2021). Liu, C.J. et al. GSCA: an integrated platform for gene set cancer analysis at genomic, pharmacogenomic and immunogenomic levels. Brief Bioinform 24 (2023). Kubo, E., Shibata, T., Singh, D.P. & Sasaki, H. Roles of TGF β and FGF Signals in the Lens: Tropomyosin Regulation for Posterior Capsule Opacity. Int J Mol Sci 19 (2018). Creyghton, W.M., Völker-Dieben, H.J. & Jager, M.J. Transforming growth factor β2 levels in aqueous humor during corneal transplantation and cataract surgery. Ocul Immunol Inflamm 3 , 29-36 (1995). DeDreu, J., Basta, M.D., Walker, J.L. & Menko, A.S. Immune Responses Induced at One Hour Post Cataract Surgery Wounding of the Chick Lens. Biomolecules 13 (2023). Novo, S.G. et al. The Immediate Early Response of Lens Epithelial Cells to Lens Injury. Cells 11 (2022). Kruger, A.J. et al. Postoperative inflammation after lens epithelial cell removal: 2 year results. J Cataract Refract Surg 27 , 1380-1385 (2001). Lovicu, F.J., Shin, E.H. & McAvoy, J.W. Fibrosis in the lens. Sprouty regulation of TGFβ-signaling prevents lens EMT leading to cataract. Exp Eye Res 142 , 92-101 (2016). Wang, X. et al. Roles of growth factors in eye development and ophthalmic diseases. Zhejiang Da Xue Xue Bao Yi Xue Ban 51 , 613-625 (2022). Hales, A.M., Chamberlain, C.G. & McAvoy, J.W. Cataract induction in lenses cultured with transforming growth factor-beta. Invest Ophthalmol Vis Sci 36 , 1709-1713 (1995). Saika, S. et al. Smad3 signaling is required for epithelial-mesenchymal transition of lens epithelium after injury. Am J Pathol 164 , 651-663 (2004). Luo, K. Signaling Cross Talk between TGF-β/Smad and Other Signaling Pathways. Cold Spring Harb Perspect Biol 9 (2017). Giarratana, A.O., Prendergast, C.M., Salvatore, M.M. & Capaccione, K.M. TGF-β signaling: critical nexus of fibrogenesis and cancer. J Transl Med 22 , 594 (2024). Yoshimatsu, Y. et al. TNF-α enhances TGF-β-induced endothelial-to-mesenchymal transition via TGF-β signal augmentation. Cancer Sci 111 , 2385-2399 (2020). Liu, Z.W. et al. Duality of Interactions Between TGF-β and TNF-α During Tumor Formation. Front Immunol 12 , 810286 (2021). Ramasamy, R., Shekhtman, A. & Schmidt, A.M. The multiple faces of RAGE--opportunities for therapeutic intervention in aging and chronic disease. Expert Opin Ther Targets 20 , 431-446 (2016). Wu, T.T. et al. AKR1B1-Induced Epithelial-Mesenchymal Transition Mediated by RAGE-Oxidative Stress in Diabetic Cataract Lens. Antioxidants (Basel) 9 (2020). Cheng, M. et al. HMGB1 Enhances the AGE-Induced Expression of CTGF and TGF-β via RAGE-Dependent Signaling in Renal Tubular Epithelial Cells. Am J Nephrol 41 , 257-266 (2015). Zheng, H. et al. PLA2R1 Inhibits Differentiated Thyroid Cancer Proliferation and Migration via the FN1-Mediated ITGB1/FAK Axis. Cancers (Basel) 15 (2023). Xiao, Z., Cai, Z., Deng, D., Tong, S. & Zu, X. An EMT-based risk score thoroughly predicts the clinical prognosis, tumor immune microenvironment and molecular subtypes of bladder cancer. Front Immunol 13 , 1000321 (2022). Li, B. et al. Fibronectin 1 promotes melanoma proliferation and metastasis by inhibiting apoptosis and regulating EMT. Onco Targets Ther 12 , 3207-3221 (2019). Zhou, J. et al. Targeting PRDX2 to inhibit tumor growth and metastasis in triple-negative breast cancer: the role of FN1 and the PI3K/AKT/SP1 pathway. J Transl Med 23 , 434 (2025). Jiang, Y. et al. MicroRNA-142-3P suppresses the progression of papillary thyroid carcinoma by targeting FN1 and inactivating FAK/ERK/PI3K signaling. Cell Signal 109 , 110792 (2023). Liao, F. et al. ECFC-derived exosomal THBS1 mediates angiogenesis and osteogenesis in distraction osteogenesis via the PI3K/AKT/ERK pathway. J Orthop Translat 37 , 12-22 (2022). Feng, L., Li, G., Li, D., Duan, G. & Liu, J. Cuproptosis-related gene SERPINE1 is a prognostic biomarker and correlated with immune infiltrates in gastric cancer. J Cancer Res Clin Oncol 149 , 10851-10865 (2023). Porreca, V. et al. Pigment Epithelium-Derived Factor Inhibits Cell Motility and p-ERK1/2 Signaling in Intrahepatic Cholangiocarcinoma Cell Lines. Biology (Basel) 14 (2025). Wang, D. et al. PAI-1 overexpression promotes invasion and migration of esophageal squamous carcinoma cells. Yi Chuan 42 , 287-295 (2020). Wernecke, L., Keckeis, S., Reichhart, N., Strauß, O. & Salchow, D.J. Epithelial-Mesenchymal Transdifferentiation in Pediatric Lens Epithelial Cells. Invest Ophthalmol Vis Sci 59 , 5785-5794 (2018). Chen, X. et al. LncRNA HOTAIR Interaction With WTAP Promotes m6A Methyltransferase Complex Assembly and Posterior Capsule Opacification Formation by Increasing THBS1. Invest Ophthalmol Vis Sci 66 , 20 (2025). Wynn, T.A. & Barron, L. Macrophages: master regulators of inflammation and fibrosis. Semin Liver Dis 30 , 245-257 (2010). Hoeft, K. et al. Platelet-instructed SPP1(+) macrophages drive myofibroblast activation in fibrosis in a CXCL4-dependent manner. Cell Rep 42 , 112131 (2023). Robertson, I.B. & Rifkin, D.B. Regulation of the Bioavailability of TGF-β and TGF-β-Related Proteins. Cold Spring Harb Perspect Biol 8 (2016). Wang, G. et al. The pan-cancer landscape of crosstalk between epithelial-mesenchymal transition and immune evasion relevant to prognosis and immunotherapy response. NPJ Precis Oncol 5 , 56 (2021). Liu, R. et al. Influence of Tumor Immune Infiltration on Immune Checkpoint Inhibitor Therapeutic Efficacy: A Computational Retrospective Study. Front Immunol 12 , 685370 (2021). Fan, J.N., Ho, H. & Chiang, B.L. Characterization of novel CD8(+) regulatory T cells and their modulatory effects in murine model of inflammatory bowel disease. Cell Mol Life Sci 81 , 327 (2024). Xie, F. et al. Breast cancer cell-derived extracellular vesicles promote CD8(+) T cell exhaustion via TGF-β type II receptor signaling. Nat Commun 13 , 4461 (2022). Lamouille, S., Xu, J. & Derynck, R. Molecular mechanisms of epithelial-mesenchymal transition. Nat Rev Mol Cell Biol 15 , 178-196 (2014). Zeisberg, M. & Neilson, E.G. Biomarkers for epithelial-mesenchymal transitions. J Clin Invest 119 , 1429-1437 (2009). Tripathi, R.C., Millard, C.B. & Tripathi, B.J. Protein composition of human aqueous humor: SDS-PAGE analysis of surgical and post-mortem samples. Exp Eye Res 48 , 117-130 (1989). Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted 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-8641774","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":581711203,"identity":"43767bf9-9710-4412-aee1-7a0a222e2d82","order_by":0,"name":"Shiming Jiang","email":"","orcid":"","institution":"First Affiliated Hospital of Shihezi University","correspondingAuthor":false,"prefix":"","firstName":"Shiming","middleName":"","lastName":"Jiang","suffix":""},{"id":581711204,"identity":"be570a91-763e-486f-9bd0-f90c6dc1a8e4","order_by":1,"name":"Chao Gao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5klEQVRIiWNgGAWjYJACZhDBxt584MAHAxs54rXw8RxLfDijIM2YeC1yEjnGxjwfDicSVC7vfvjg54IaGwY2iRwzaRsD5gQG9sNHN+DTYngmLVl6xrE0BjaeZ2XSOQZseQw8aWk38GppyDFj5mE7DPR+8jagFp5iBgkeM/xa+t8AtfwDamFIMJO2MJBIbCCkRR7oBWbeNqAWjhRjYwYDA8JaDCSeJUvz9oH8AgzkHoMEYzZCfpHvTz74meebDYN8OzAqf/z5L8fPfvgYflsOQOj6BpgIGz7lYFsaCKkYBaNgFIyCUQAAGGdDIQ3IpdQAAAAASUVORK5CYII=","orcid":"","institution":"First Affiliated Hospital of Shihezi University","correspondingAuthor":true,"prefix":"","firstName":"Chao","middleName":"","lastName":"Gao","suffix":""}],"badges":[],"createdAt":"2026-01-19 17:15:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8641774/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8641774/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":101517893,"identity":"cf0c3469-3dce-4852-9a06-0e1bdb4656dc","added_by":"auto","created_at":"2026-01-30 16:25:49","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":87274,"visible":true,"origin":"","legend":"\u003cp\u003eThe key points of this research method are summarized in Figure 1.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/18f51a23f26e2785e8fe45d2.png"},{"id":101518003,"identity":"951e7a08-7cba-4ddb-b3b7-709f8c122dc8","added_by":"auto","created_at":"2026-01-30 16:26:06","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":598328,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eData processing and identification of DEGs from GSE295383 \u003c/strong\u003eFigure 2. ​​(A)​​ Sample clustering analysis was performed to exclude potential outliers, ensuring the reliability of subsequent analyses. (B) Volcano plot visualizing the DEGs. (C) Functional enrichment analysis of identified DEGs from GSE295383 using GO terms and KEGG pathways.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/78c587351406c454147371a5.png"},{"id":101517934,"identity":"b1bd5464-52d6-4a56-8a36-7c1348fe8c67","added_by":"auto","created_at":"2026-01-30 16:25:54","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":934478,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConstruction of the co-expression network. \u003c/strong\u003eFigure 3. (A)​​ Mean connectivity of module eigengenes across a range of soft-thresholding powers. (B)​​ Scale-free topology fit index (signed R²) for various soft-thresholding powers. The red line indicates the chosen power value of 8, at which the model fit exceeds 0.85, suggesting an approximate scale-free topology. (C)​​ Hierarchical clustering dendrogram of genes. Highly similar modules were dynamically identified and merged (a representative subset is shown here; the full dendrogram is provided in Figure S1). (D)​​ Heatmap depicting Pearson correlations between module eigengenes and external traits. (E)​​ Scatter plot illustrating the relationship between Module Membership and Gene Significance for genes within the brown module.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/98bed46b03b8323888d2c6b3.png"},{"id":101517613,"identity":"39a5ed3e-a583-400f-abe7-2f46d92376e4","added_by":"auto","created_at":"2026-01-30 16:25:04","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1699416,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePPI network analysis. \u003c/strong\u003eFigure 4. (A) Venn diagram of key module genes versus DEGs. (B) PPI network constructed using genes screened under specific criteria, depicting functional associations and molecular relationships among encoded proteins. Color saturation and pattern dimensions are configured in accordance with the Degree. (C) Venn diagram identifying seven consensus hub genes determined by the intersection of three distinct algorithms: MCC, Degree and MNC.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/688fa35a1fbd3ea1212a2f9b.png"},{"id":101517821,"identity":"c469c99c-9050-4769-8888-56f253caf4bb","added_by":"auto","created_at":"2026-01-30 16:25:35","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":546567,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMachine learning-based hub selection and model evaluation​. \u003c/strong\u003eFigure 5. (A) Coefficient Distribution Plot of LASSO Regression Analysis.(B) Optimal log(λ) value selected through minimum mean squared error criteria. (C) Ideal number of trees determined via error rate convergence in the RF algorithm.(D) ROC curve illustrating the classification performance of the RF-based model.(E) Hub importance scores derived from the RF model, indicating the contribution of each gene.(F) Maximum accuracy achieved by the SVM-RFE algorithm.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/c7474fc2463d39a013c32597.png"},{"id":101517971,"identity":"3eb3555f-4d71-4326-9d70-ca410bfcb955","added_by":"auto","created_at":"2026-01-30 16:26:02","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1492603,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eValidation of hub genes in internal datasets \u003c/strong\u003eFigure 6. ​(A) Venn diagram illustrating the consensus hub genes identified by machine-learning algorithms. (B) Nomogram constructed based on a multivariable logistic regression model incorporating the hub genes FN1, SERPINE1, and THBS1. (C) ROC of the model, indicating the optimal cutoff point for classification. (D) Confusion matrix evaluating the model’s predictive performance against actual outcomes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eValidation of hub genes in external datasets. \u003c/strong\u003eFigure 6. (E) Expression profiles of signature genes in the GSE111430 dataset, derived from high-throughput sequencing of remnant lens epithelial cells in Mus musculus at 0h and 24h post-surgery. (F) Expression patterns of signature genes in the GSE119879 dataset, obtained from high-throughput sequencing of remnant lens epithelial cells in Mus musculu\u003cem\u003es\u003c/em\u003e at 0h and 48h post-surgery.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/d79e5507264aeec8c9386ab4.png"},{"id":101517876,"identity":"b49d0d24-1b4b-48a3-90fc-3f757741c94a","added_by":"auto","created_at":"2026-01-30 16:25:46","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":3258861,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eImmune infiltration analysis\u003c/strong\u003e. Figure 7. (A) Histogram displaying the relative proportions of immune cell subsets across samples. (B) Violin plots comparing the infiltration levels of immune cell subsets between history of surgery group and non−surgery group, with statistical significance indicated. (C) Heatmap visualizing the pairwise correlations among various immune cell subsets based on their relative infiltration abundances. (D) Correlation analysis evaluating the relationships between the expression levels of identified hub genes and the infiltration abundance of specific immune cell subsets.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/b1407314d2d3cedeeaad5fb7.png"},{"id":101517782,"identity":"4d064853-1cef-40b3-9e27-a2103dc22b7b","added_by":"auto","created_at":"2026-01-30 16:25:31","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1851986,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEnrichment plots from GSEA. \u003c/strong\u003eFigure 8. (A) GSEA enrichment plot for EMT-related pathways correlation with M0 macrophages infiltration. (B) GSEA enrichment plot for EMT-related pathways correlation with resting CD4+ memory T cells infiltration. (C) GSEA enrichment plot for inflammation-related pathways correlation with M0 macrophages infiltration. (D) GSEA enrichment plot for inflammation-related pathways correlation with resting CD4+ memory T cells infiltration. (E) GSEA enrichment plot for EMT-related pathways correlation with CD8+ T cells infiltration. (F) GSEA enrichment plot for inflammation-related pathways correlation with CD8+ T cells infiltration.\u003c/p\u003e","description":"","filename":"Figure8.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/38eb069dda32201692c5ea46.png"},{"id":101517884,"identity":"c2ff97c7-fd01-4e5e-8905-ea2d60db88f3","added_by":"auto","created_at":"2026-01-30 16:25:48","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":356660,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe results of qRT-PCR.\u003c/strong\u003e The qRT-PCR analysis of SRA01/04 was conducted with three biological replicates per condition and three technical replicates per sample, and the data were analyzed using the\u0026nbsp;2\u003csup\u003e- ΔΔCT\u003c/sup\u003e method. Figure 9. (A) Expression levels of VIM and CDH1 under different treatment conditions. (B) Expression levels of hub genes under different treatment conditions.\u003c/p\u003e","description":"","filename":"Figure9.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/dfab126571949bbf8b35ac99.png"},{"id":101517781,"identity":"9f9dbd07-cd11-4e26-9fe3-18a3d0887290","added_by":"auto","created_at":"2026-01-30 16:25:31","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":2529244,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWound-healing assay. \u003c/strong\u003eFigure 10. (A) Representative images of the wound area at 0 h and 20 h post-scratching under various treatments. (B) Quantitative analysis of the cell mobility rate.\u003c/p\u003e","description":"","filename":"Figure10.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/d9bd803fe8f32b2fb8cfb9ff.png"},{"id":101517739,"identity":"1111996b-530c-4c95-98d8-1ec78738f75d","added_by":"auto","created_at":"2026-01-30 16:25:22","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":1271394,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMolecular docking results.\u003c/strong\u003eThe green structure represents the small molecule compound bisphenol A, the blue structure denotes the protein encoded by the gene, and the orange indicates the residues forming hydrogen bonds between bisphenol A and the protein. Figure 11.(A) The protein encoded by FN1 establishes hydrogen bonds with CsA via the residues V15 and T104. (B) The protein encoded by SERPINE1 interacts with CsA through hydrogen bonding at the residue HIS-279.\u003c/p\u003e","description":"","filename":"Figure11.png","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/0bf84d3c61ad2e70f80f454d.png"},{"id":106403370,"identity":"74ec598c-540f-49fc-8628-76b7735929f4","added_by":"auto","created_at":"2026-04-08 09:14:10","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":16546359,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8641774/v1/be4dd699-2229-4105-b610-77abf78d01bc.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Analysis and validation of hub genes and immune cell infiltration characteristics in lens epithelium tissue with cataract surgical history","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eCataract, characterized by opacification of the crystalline lens, is the primary contributor to global blindness. This ocular structure maintains transparency through its unique composition of lens epithelial cells (LECs) and elongated fiber cell \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Epidemiological studies highlight cataract as a major cause of blindness in regions such as India, China, and sub-Saharan Africa, with an estimated 16\u0026nbsp;million people globally affected by blindness due to this condition \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003ePhacoemulsification combined with posterior chamber intraocular lens implantation is the primary surgical treatment for cataract \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. However, a common complication following this procedure is PCO, which remains a significant challenge in ophthalmic surgery \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Following cataract surgery, LECs may undergo abnormal differentiation, leading to enhanced migratory capacity. These cells can relocate to the surface of the lens capsule and proliferate. This pathological process is likely triggered by direct surgical injury to LECs and alterations in the cellular microenvironment resulting from a breakdown of the blood\u0026ndash;aqueous barrier \u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Ultimately, when the aforementioned pathological changes progress to a sufficiently severe stage, they can lead to visual impairment. Research data indicate that the proportion of patients requiring laser capsulotomy within three years following standard cataract surgery ranges between 5% and 20% \u003csup\u003e5, 9, 10\u003c/sup\u003e. Despite being a well-recognized phenomenon, the precise molecular mechanisms and hub gene driving these cellular processes are not yet fully understood.\u003c/p\u003e \u003cp\u003eThe disruption of the blood-aqueous barrier following cataract surgery, combined with the implantation of pseudophakia, leads to the infiltration of inflammatory cells and immune mediators into the anterior chamber, a site of which was considered immunologically privileged \u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Elevated expression of proinflammatory genes and proteins has been observed postoperatively in both clinical patients and murine cataract surgery models \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Studies analyzing aqueous humor from donors who underwent cataract surgery 3 to 12 years prior have revealed persistently elevated levels of Inflammatory cytokines, indicating that a chronic inflammatory state can persist long after the initial surgical intervention \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Inflammatory cytokines have been implicated in driving pathological changes in lens LECs \u003csup\u003e\u003cspan additionalcitationids=\"CR17 CR18\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. While immunofluorescence studies have demonstrated the colocalization of inflammatory cell infiltration with phenotypic and metabolic changes in LECs, along with their correlation to fibrotic marker expression, the exact molecular pathways driving these interactions remain poorly understood \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eCsA, a calcineurin inhibitor, exerts its immunosuppressive effects by selectively modulating T-cell infiltration and activation \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. In ophthalmology, topical CsA is widely utilized to manage ocular surface inflammation, particularly in dry eye disease resulting from various etiologies such as cataract surgery, lacrimal dysfunction, and ocular surface impairment \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e.Evidence from animal studies has demonstrated that an intraocular lens modified with sustained-release CsA can significantly reduce anterior chamber inflammation and delay the onset of PCO following cataract surgery \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. However, these findings remain predominantly at the phenomenological level, and a thorough investigation into the molecular mechanisms by which CsA inhibits PCO progression is still lacking.\u003c/p\u003e \u003cp\u003eAs a gene expression analysis method with high accuracy and a broad dynamic range, high-throughput sequencing, by integrating statistics, biological databases, and machine learning algorithms, has been widely applied across various disease domains to uncover underlying pathway alterations, key bioactive molecules, and functional targets in pathological tissues and cells \u003csup\u003e\u003cspan additionalcitationids=\"CR27 CR28 CR29\" citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Immune infiltration analysis further reveals the correlation between hub genes and immune cell \u003csup\u003e\u003cspan additionalcitationids=\"CR28 CR29\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. However, there remains a lack of systematic bioinformatics studies, particularly immune infiltration analyses, on PCO following cataract surgery.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Data source\u003c/h2\u003e \u003cp\u003eThe high-throughput sequencing data derived from lens capsules of non-cataract donors and donors with a history of cataract surgery were downloaded from the GEO database under the GSE295383 \u003csup\u003e31\u003c/sup\u003e. This dataset comprised samples from 17 donors without cataracts and 17 donors with a history of cataract surgery. Lens capsules obtained from donors who had no documented history of cataracts, other ocular diseases, or diabetes were categorized as the non-cataract donors. Those acquired from donors with a prior cataract surgery but without diabetes or other ocular pathologies were classified into the post-cataract surgery group. All tissues were collected from donor globes within 2 to 10 hours postmortem; all lens capsule samples were selected for subsequent bioinformatic analysis. Then, we retrieved high-throughput sequencing data GSE111430 and GSE119879 from the GEO database as external datasets to validate the ultimately screened hub genes. \u003csup\u003e14, 20\u003c/sup\u003e. These datasets contained transcriptomic profiles from Mus musculus specimens, which included control groups (0h) and experimental groups at 24h and 48h following cataract surgery.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Methodology\u003c/h2\u003e \u003cp\u003eThe key points of this research method are summarized in Fig.\u0026nbsp;1.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Differential expressed genes (DEGs) analysis\u003c/h2\u003e \u003cp\u003eA comparative analysis was conducted to examine differences between donors without cataracts and those who have undergone cataract surgery, utilizing the limma package (v3.60.6) within R software (v4.4.1) \u003csup\u003e32\u003c/sup\u003e. DEGs were determined using the criteria of |logFC| \u0026gt; 1 and \u003cem\u003eP\u003c/em\u003e-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05. Visualization of the results from differential expression analysis was accomplished through the \"ggplot2\" package (v3.5.2) in R software, which produced volcano plots.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Enrichment analysis\u003c/h2\u003e \u003cp\u003eFor functional enrichment analysis of gene sets, the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) annotation data from the org.Hs.eg.db package (v3.20.0) in R software were employed as the background reference \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Genes were mapped onto this background dataset, and enrichment procedures were performed using the clusterProfiler (v4.14.6) package in R software.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5. Weighted gene co-expression network analysis (WGCNA)\u003c/h2\u003e \u003cp\u003eWe employed the WGCNA R package (v1.73) for WGCNA \u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Initially, hierarchical clustering was carried out using the hclust function to assess sample outliers. The analysis utilized a soft-thresholding power of 8 for calculating biweight midcorrelations to construct signed networks. Additional parameters included TOMType = \"unsigned\", mergeCutHeight\u0026thinsp;=\u0026thinsp;0.35, and minModuleSize\u0026thinsp;=\u0026thinsp;50 to identify cohesive gene modules. Finally, the association between module eigengenes and status was evaluated using Pearson correlation analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e\u003cb\u003e2.6. Protein-protein interaction (PPI\u003c/b\u003e) \u003cb\u003enetwork analysis\u003c/b\u003e\u003c/h2\u003e \u003cp\u003eThe PPI network is a graph model comprising nodes (proteins) and edges (interactions between them) \u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. For network construction, a PPI network were retrieved from the STRING, for \u003cem\u003eHomo sapiens\u003c/em\u003e, applying a confidence score cutoff set above 0.400. After removing unrelated protein entries, topological analysis was performed using the cytoHubba (v3.10.3) plugin, employing three distinct algorithms: Degree, Maximum Neighborhood Component (MNC), and Maximal Clique Centrality (MCC). The genes consistently ranked in the top 10 by all three algorithms were defined as candidate key genes for subsequent investigation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7. Screening hub genes by machine learning\u003c/h2\u003e \u003cp\u003eThree distinct machine learning algorithms were implemented: least absolute shrinkage and selection operator (LASSO) regression, support vector machine-recursive feature elimination (SVM-RFE), and random forest (RF) \u003csup\u003e\u003cspan additionalcitationids=\"CR38 CR39\" citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. A random seed (123456) was set at the outset to ensure the reproducibility of results.\u003c/p\u003e \u003cdiv id=\"Sec10\" class=\"Section3\"\u003e \u003ch2\u003e2.7.1. LASSO regression\u003c/h2\u003e \u003cp\u003eThe LASSO analysis was conducted using the glmnet package (v4.1.10) in R software. Model training was performed using the glmnet function, and the optimal regularization parameter was determined via 5-fold cross-validation. Features (genes expression) with non-zero coefficients under the optimal lambda value were retained as candidate feature genes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e2.7.2. SVM-RFE\u003c/h2\u003e \u003cp\u003eThe SVM-RFE algorithm was utilized for its recursive feature ranking capability. Using the e1071package (v1.7.16) in R software, the process involved iteratively removing the least important genes as determined by the SVM model weights. The feature subset that yielded the highest classification accuracy during 5-fold cross-validation was selected as the final set of candidate genes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e2.7.3. RF\u003c/h2\u003e \u003cp\u003eThe RF algorithm was implemented via the randomForest package (v4.7.1.2) in R software, with the number of trees set to 200. Feature importance was quantified using the Mean Decrease in Gini Index, which reflects the contribution of each feature to node impurity reduction across all trees. Gene expression characteristics with score greater than 1.5 were considered candidate feature genes.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e2.8. Estimation of hub genes\u003c/h2\u003e \u003cp\u003eA multivariable logistic regression model was fitted to the finalized hub gene expression profiles using the R package rms (v8.0.0) in R software. A nomogram was constructed to visually represent the interactions among the selected genes \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. The discriminative ability of the multivariable logistic regression model was assessed by the ROC (Receiver Operating Characteristic) curve. The optimal cutoff point on the ROC curve was determined by maximizing Youden\u0026rsquo;s index \u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. A confusion matrix was subsequently generated based on this threshold to evaluate classification performance, including metrics such as sensitivity, specificity, and accuracy \u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. The datasets GSE111430 and GSE119879 were utilized. After normalization processing. The data conforms to a normal distribution upon normality testing. independent two-sample t-tests were applied to assess whether the expression of hub genes exhibited statistically significant differences between comparative groups.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e2.9. Immune infiltration analysis\u003c/h2\u003e \u003cp\u003eThe relative abundances of human immune cell populations within the samples were deconvoluted using the CIBERSORT algorithm and its built-in LM22 signature matrix \u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. To identify distribution of immune cell infiltration between two predefined sample groups, statistical comparisons for each immune cell subset were carried out using the Wilcoxon signed-rank test.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e2.10. Gene Set Enrichment Analysis (GSEA)\u003c/h2\u003e \u003cp\u003eData were downloaded from the GSEA database \u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Genes were ranked based on the correlation between their expression and the infiltration scores of specific immune cells, and perform pathway analysis on the ranked gene list using the clusterProfiler package (v4.14.6) in R software.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e\u003cb\u003e2.11. Cells culture and treatment\u003c/b\u003e\u003c/h2\u003e \u003cp\u003eThe human lens epithelial cells (SRA01/04) were purchased from Shanghai Zhong Qiao Xin Zhou Biotechnology Co., Ltd. (product number: ZQ0612) and were cultured at 37℃ with 5% CO\u003csub\u003e2\u003c/sub\u003e in the Dulbecco's modified Eagle's medium with 10% fetal bovine serum (10091148, Gibco, USA), 100 \u0026micro;g/mL streptomycin, 100 U/mL penicillin (SV30010, Hyclone, USA). For TGF-β2 treatments, the culture medium was replaced on the second day after plating with fresh medium containing recombinant human TGF-β2 at concentrations of 0, 1, or 10 ng/ml, followed by a 48h incubation period.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e\u003cb\u003e2.12. Real-time quantitative reverse transcription PCR(qRT-PCR)\u003c/b\u003e\u003c/h2\u003e \u003cp\u003eTotal RNA was isolated from SRA01/04 using TRIZOL (15596-018, Ambion,USA). cDNA synthesis was done by reverse transcription kit (R323-01, Vazyme, China). Quantitative PCR was subsequently performed on an ABI QuantStudio 5 system. The reaction protocol included an initial denaturation step, followed by 40 cycles of denaturation, and a combined annealing/extension step. Each sample had three technical replicates. The concentration of each transcript was then normalized to GAPDH and mRNA level using 2\u003csup\u003e- ΔΔCT\u003c/sup\u003e method to analysis. Statistical analysis was conducted with two-way ANOVA in GraphPad Prism software (v10.6.0).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e2.13. Wound-healing assay\u003c/h2\u003e \u003cp\u003eThe migration capability of SRA01/04 cells was assessed using a wound-healing assay. Cells were seeded in plates and grown to a high confluence. Before creating the wound, the cells were serum-starved for 4h. A uniform scratch was then introduced into the cell monolayer using a sterile pipette tip. Images of the wound were captured at the initial time point (0h) and after a 20h incubation period at marked locations. The migration of cells into the wound area was observed and photographed using an inverted microscope (MF52-N, Mshot, China). The percentage of wound closure was quantified by comparing the remaining cell-free area at 20 hours to the initial wound area at 0 hours.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e2.14. Molecular docking\u003c/h2\u003e \u003cp\u003eThe protein structures encoded by hub genes, along with 2D chemical structures of the selected small molecules, were retrieved from the PDB and PubChem databases. The 2D structures of these molecules were then converted into 3D formats and energetically minimized using the Minimize Energy​ function in Chem3D (v14.0.0.117). Molecular docking was performed using AutoDock​ (v1.5.7). Finally, the docking results were visualized using PyMOL (v2.6.2).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e2.15. Statistical analysis\u003c/h2\u003e \u003cp\u003eUnless otherwise specified, all the statistical data were analyzed using R software (v4.4.1) and GraphPad Prism(v10.6.0). Appropriate statistical methods were selected and applied according to the specific experimental design and data characteristics of each analysis. The normality of data distribution was assessed using established tests, such as the Shapiro\u0026ndash;Wilk and Kolmogorov\u0026ndash;Smirnov tests in SPSS(v27.0) software. The homogeneity of variances, a key assumption for many parametric tests, was evaluated using appropriate methods such as Levene's test or F-test, depending on the number of groups and data distribution. In instances where data violated the assumption of normality or variance homogeneity, non-parametric alternatives were employed. For all inferential analyses, statistical significance was defined using the following thresholds: \u003cem\u003eP\u003c/em\u003e* \u0026lt; 0.05, ​​\u003cem\u003eP\u003c/em\u003e** \u0026lt; 0.01, and ​​\u003cem\u003eP\u003c/em\u003e*** \u0026lt; 0.001.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cp\u003e\u003cstrong\u003e3.1. DEGs\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eanalysis\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;and functional enrichment analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFollowing preprocessing steps including removal of low-expression genes and deduplication of the gene expression profile. Subsequently, sample clustering analysis was conducted to identify and remove outlier samples, such as FXJ13 (Figure 2A). A total of 264 DEGs, comprising 199 upregulated and 65 downregulated genes, were identified between human lens capsules with cataract surgical histories and non-surgical histories. These DEGs were visualized using a volcano plot and pheatmap (Figure 2B,2C). Subsequent functional enrichment analysis of these genes (Figure 2D) revealed significant associations in GO related to inflammatory cells and their cytokines, as well as pathways associated with cell-substrate junction and ECM organization. KEGG pathway analysis further indicated significant enrichment of pathways involved in PI3K\u0026minus;Akt signaling pathway, TNF signaling pathway, AGE\u0026minus;RAGE signaling pathway and TGF\u0026minus;beta signaling pathway.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;Data processing and identification of DEGs from GSE295383\u0026nbsp;\u003c/strong\u003eFigure 2. (A) Sample clustering analysis was performed to exclude potential outliers, ensuring the reliability of subsequent analyses. (B) Volcano plot visualizing the DEGs. (C) Functional enrichment analysis of identified DEGs from GSE295383 using GO terms and KEGG pathways.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.2. Identification of trait-related module genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify trait-related module genes, the WGCNA was employed. A soft-thresholding power of 8 was selected for calculating biweight midcorrelations to build signed networks (Figure 3A, 3B, 3C). After generating a heatmap illustrating module-trait correlations, the \u0026quot;brown\u0026quot; module was chosen for further analysis due to its highest correlation with the grouping trait while remaining independent of variations in gender and age (Figure 3D). Hub genes within this module were ultimately identified based on the criteria of Module Membership \u0026gt; 0.8 and Gene Significance \u0026gt; 0.5 (Figure 3E).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConstruction of the co-expression network.\u0026nbsp;\u003c/strong\u003eFigure 3. (A) Mean connectivity of module eigengenes across a range of soft-thresholding powers. (B) Scale-free topology fit index (signed R\u0026sup2;) for various soft-thresholding powers. The red line indicates the chosen power value of 8, at which the model fit exceeds 0.85, suggesting an approximate scale-free topology. (C) Hierarchical clustering dendrogram of genes. Highly similar modules were dynamically identified and merged (a representative subset is shown here; the full dendrogram is provided in Figure S1). (D) Heatmap depicting Pearson correlations between module eigengenes and external traits. (E) Scatter plot illustrating the relationship between Module Membership and Gene Significance for genes within the brown module.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.3. PPI network analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify genes with hub biological significance from those obtained through statistical methods, a total of 86 genes, representing the intersection of DEGs and key modules from the WGCNA analysis (Figure 4A). Candidate genes were selected for construction of the PPI network. This combined approach aimed to reduce false positives and focus on high-potential candidates. PPI network was subsequently constructed using these genes via the STRING database (Figure 4B). Hub genes within this network were identified by applying three topological analysis algorithms MCC, Degree, and MNC through the cytoHubba plugin. Seven genes(FN1, THBS1,SERPINE1, COL1A1,SPP1, TIMP1,and BGN)were consistently identified among the top ten most central nodes across all three algorithms (Figure 4C). These seven hub genes were selected for further downstream analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePPI network analysis\u003c/strong\u003e\u003cstrong\u003e.\u0026nbsp;\u003c/strong\u003eFigure 4. (A) Venn diagram of key module genes versus DEGs. (B) PPI network constructed using genes screened under specific criteria, depicting functional associations and molecular relationships among encoded proteins. Color saturation and pattern dimensions are configured in accordance with the Degree. (C) Venn diagram identifying seven consensus hub genes determined by the intersection of three distinct algorithms: MCC, Degree and MNC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.4. Screening\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003ehub genes\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;by machine learning\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify robust hub genes associated with status, we employed three distinct machine-learning algorithms for hub selection. LASSO regression analysis was applied to refine the candidate pool derived from univariate analysis, ultimately identifying three predictor genes (Figure 5A, 5B). Subsequently, a RF algorithm integrated with hub importance evaluation was utilized to assess the relationship between the classification error rate and predictive performance, which was further quantified using ROC curve analysis. Genes achieving an importance score exceeding 1.5 were retained as candidate genes (Figure 5C, 5D, 5E). The SVM-RFE method was implemented to pinpoint optimal hub genes based on maximal accuracy (Figure 5F). The final set of hub genes, selected through this multi-step computational approach and demonstrating strong relevance to surgical history, was advanced for further validation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;Machine learning-based hub selection and model evaluation.\u0026nbsp;\u003c/strong\u003eFigure 5. (A) Coefficient Distribution Plot of LASSO Regression Analysis.(B) Optimal log(\u0026lambda;) value selected through minimum mean squared error criteria. (C) Ideal number of trees determined via error rate convergence in the RF algorithm.(D) ROC curve illustrating the classification performance of the RF-based model.(E) Hub importance scores derived from the RF model, indicating the contribution of each gene.(F) Maximum accuracy achieved by the SVM-RFE algorithm.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.5. Validation of hub genes in Internal and external datasets\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThree distinct machine learning algorithms consistently identified FN1, SERPINE1, and THBS1 as robust hub genes (Figure 6A). Based on these genes, a multivariable logistic regression model was constructed using the rms package. A nomogram was generated to visually represent the interactions among these selected genes (Figure 6B). The multivariable logistic regression model performance was evaluated using ROC curve analysis and a confusion matrix determined by the optimal cutoff point (Figure 6C, 6D). Results demonstrated that the model exhibited strong discriminative ability and high specificity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eValidation of hub genes in internal datasets\u0026nbsp;\u003c/strong\u003eFigure 6. (A) Venn diagram illustrating the consensus hub genes identified by machine-learning algorithms. (B) Nomogram constructed based on a multivariable logistic regression model incorporating the hub genes FN1, SERPINE1, and THBS1. (C) ROC of the model, indicating the optimal cutoff point for classification. (D) Confusion matrix evaluating the model\u0026rsquo;s predictive performance against actual outcomes.\u003c/p\u003e\n\u003cp\u003eFor external validation, datasets GSE111430 and GSE119879, which contain expression profiles from remnant lens epithelial cells of Mus musculus at 0h, 24h, and 48h after simulated cataract surgery, were downloaded and processed. Following data normalization and normality testing, two-sample t-tests were applied to assess whether the expression levels of each signature gene showed statistically significant differences between comparative groups (Figure 6E, 6F). The analysis confirmed that all three hub genes were significantly differentially expressed in both external datasets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eValidation of hub genes in external datasets.\u0026nbsp;\u003c/strong\u003eFigure 6. (E) Expression profiles of signature genes in the GSE111430 dataset, derived from high-throughput sequencing of remnant lens epithelial cells in Mus musculus at 0h and 24h post-surgery. (F) Expression patterns of signature genes in the GSE119879 dataset, obtained from high-throughput sequencing of remnant lens epithelial cells in Mus musculu\u003cem\u003es\u003c/em\u003e at 0h and 48h post-surgery.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.6. Immune infiltration analysis and Its association with hub genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo investigate the immune cell infiltration between the two sample groups and explore intercellular relationships, as well as to determine whether the expression of hub genes is associated with these immune cells, the global expression matrix was deconvoluted using the CIBERSORT algorithm to estimate the relative abundance of various immune cells (Figure 7A). The analysis revealed that samples with a surgical history exhibited a significantly higher fraction of resting CD4+ memory T cells (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01) and M0 macrophages (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05), while a lower fraction of CD8+ T cells (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01) was observed (Figure 7B). Correlation heatmap analysis further demonstrated a strong negative association (r = \u0026minus;0.8) between resting CD4+ memory T cells and CD8+ T cells, and a positive correlation (r = 0.59) between M0 macrophages and M2 Macrophages (Figure 7C). Furthermore, all hub genes exhibited significant associations with resting CD4+ memory T cells, M0 macrophages, and CD8+ T cells, which displayed notable infiltration differences. These correlations followed the same trend as the overall infiltration pattern. (Figure 7D).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImmune infiltration analysis\u003c/strong\u003e. Figure 7. (A) Histogram displaying the relative proportions of immune cell subsets across samples. (B) Violin plots comparing the infiltration levels of immune cell subsets between history of surgery group and non\u0026minus;surgery group, with statistical significance indicated. (C) Heatmap visualizing the pairwise correlations among various immune cell subsets based on their relative infiltration abundances. (D) Correlation analysis evaluating the relationships between the expression levels of identified hub genes and the infiltration abundance of specific immune cell subsets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.7. Identification of signaling pathways associated with Immune Infiltration via GSEA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAlong with the biological pathways enriched by genes correlated with specific immune cells, we performed Spearman correlation analysis between each gene in the expression matrix and immune infiltration scores across all samples. Genes were ranked based on the values of their correlation coefficients for subsequent GSEA. The GSEA results indicated that genes positively correlated with the high proportion of immune cells in samples with a surgical history were significantly enriched in pathways promoting epithelial-mesenchymal transition (EMT), cell junction disassembly, and multiple EMT-related processes such as ERK1/ERK2 cascade, MAPK signaling, TGF-\u0026beta; production, and cellular response to TGF-\u0026beta; (Figure 8A, 8B). In terms of inflammation-related pathways, genes involved in chemotaxis of various immune cells and inflammatory responses triggered by injury were also positively associated with M0 macrophages and resting CD4+ memory T cells (Figure 8C, 8D). In contrast, infiltration of CD8+ T cells was negatively correlated with the enrichment of genes in these pathways (Figure 8E, 8F).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEnrichment plots from GSEA\u003c/strong\u003e\u003cstrong\u003e.\u0026nbsp;\u003c/strong\u003eFigure 8. (A) GSEA enrichment plot for EMT-related pathways correlation with M0 macrophages infiltration. (B) GSEA enrichment plot for EMT-related pathways correlation with resting CD4+ memory T cells infiltration. (C) GSEA enrichment plot for inflammation-related pathways correlation with M0 macrophages infiltration. (D) GSEA enrichment plot for inflammation-related pathways correlation with resting CD4+ memory T cells infiltration. (E) GSEA enrichment plot for EMT-related pathways correlation with CD8+ T cells infiltration. (F) GSEA enrichment plot for inflammation-related pathways correlation with CD8+ T cells infiltration.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.8 Expression levels of EMT-associated markers and hub genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe high concentration of TGF-\u0026beta;2 in the aqueous humor resulting from surgical trauma induces pathological changes in LECs, which is considered a key pathological environment for the development and progression of PCO \u003csup\u003e16, 46, 47\u003c/sup\u003e Our previous findings demonstrated a close correlation between TGF-\u0026beta; signaling pathway activation, upregulation of EMT-related pathways, and the expression of hub genes (FN1, SERPINE1, and THBS1). To further elucidate this relationship under controlled conditions, we focused on a well-defined single cell line model. Specifically, we validated the effect of a high TGF-\u0026beta;2 microenvironment on the induction of EMT and the expression of these hub genes in lens epithelial cells (LECs). The results showed that the expression of CDH1 (encoding the epithelial marker E-cadherin) decreased in a concentration-dependent manner with TGF-\u0026beta;2 treatment. In contrast, the expression of VIM (encoding the mesenchymal marker Vimentin) significantly increased only in the 10 ng/ml TGF-\u0026beta;2 environment compared to the control group(Figure 9A). All hub genes exhibited a concentration-dependent upregulation in expression(Figure 9B).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe results of qRT-PCR.\u003c/strong\u003e The qRT-PCR analysis of SRA01/04 was conducted with three biological replicates per condition and three technical replicates per sample, and the data were analyzed using the 2\u003csup\u003e- \u0026Delta;\u0026Delta;CT\u003c/sup\u003e method. Figure 9. (A) Expression levels of VIM and CDH1 under different treatment conditions. (B) Expression levels of hub genes under different treatment conditions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.9 Assessment of migration capability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo determine the effect of TGF-\u0026beta;2 environment on the migration capability of LECs, we performed a wound-healing assay and evaluated the cell mobility rate. The results demonstrated that treatment with both 1 ng/ml and 10 ng/ml TGF-\u0026beta;2 significantly enhanced the migration capability of LECs compared to the control group.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eWound-healing assay.\u0026nbsp;\u003c/strong\u003eFigure 10. (A) Representative images of the wound area at 0 h and 20 h post-scratching under various treatments. (B) Quantitative analysis of the cell mobility rate.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.10. Molecular docking of CsA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo elucidate the potential molecular mechanisms of CsA intervention in PCO, molecular docking analysis was performed. The results showed that the proteins encoded by the hub genes strongly interact with CsA, with binding energies of -5.5 kcal/mol for FN1, -6.4 kcal/mol for SERPINE1 and -4.7 kcal/mol for THBS1. Among these, the proteins encoded by FN1 and SERPINE1 were found to form hydrogen bonds with CsA (Figure 11A-B).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMolecular docking results.\u003c/strong\u003eThe green structure represents the small molecule compound bisphenol A, the blue structure denotes the protein encoded by the gene, and the orange indicates the residues forming hydrogen bonds between bisphenol A and the protein. Figure 11.(A) The protein encoded by FN1 establishes hydrogen bonds with CsA via the residues V15 and T104. (B) The protein encoded by SERPINE1 interacts with CsA through hydrogen bonding at the residue HIS-279.\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eIn general, the inflammatory microenvironment and disruption of the Injury stimulus during cataract surgery can trigger significant adaptive and remodeling responses in remnant lens epithelial cells \u003csup\u003e\u003cspan additionalcitationids=\"CR49\" citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. Unlike previous studies that provided the source data, our study extends beyond simple differential expression analysis and individual gene assessments. this study integrated multiple analytical approaches to conduct a systematic, multi-gene dimensional analysis of high-throughput sequencing data derived from human lens capsules with a surgical history compared to non-surgical controls. Our results reveal that surgical history activates biological processes and molecular pathways, and identify hub genes (including FN1, SERPINE1, and THBS1) which occupy statistically and biologically central positions within the related gene co-expression network. The expression of these hub genes in LECs under stimulation by TGF-β2, a classic factor in PCO, was further validated through cell experiments. Furthermore, we investigated immune cell infiltration landscape and demonstrated that immune subsets with significant infiltration differences are associated with hub genes and biological pathways. Ultimately, we investigated the potential of CsA to intervene in these changes by targeting the proteins encoded by the identified hub genes.\u003c/p\u003e \u003cp\u003eDEGs analysis using the limma package identified DEGs. GO Biological Process enrichment analysis revealed that DEGs were significantly involved in pathways related to myeloid leukocyte activation, myeloid leukocyte differentiation, and regulation of mononuclear cell migration. These findings align with the inflammatory response triggered by cataract surgery, as discussed earlier.\u003c/p\u003e \u003cp\u003eA pathological feature of PCO is the loss of LECs integrity, which is closely associated with abnormal cell proliferation and migration. During this process, LECs progressively deviate from their ectodermal epithelial origin and transform into mesenchymal-like cells with more midline characteristics, a process known as EMT \u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. The newly generated mesenchymal-like cells exhibit weakened cell-substrate junctions, enhanced migratory and invasive capabilities, and excessive production of ECM components. Consistent with this mechanism, enrichment analyses of GO and KEGG pathways among DEGs also indicated alterations related to collagen production, ECM organization, cell-substrate adhesion, and other metabolic changes characteristic of EMT. Multiple EMT-associated signaling pathways were significantly enriched. Specifically, Enhanced TGF-β signaling has been demonstrated to induce EMT in LECs \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e, contributing to a cataract-related phenotype. For instance, Shizuya Saika et al \u003csup\u003e54\u003c/sup\u003erevealed that EMT in primary LECs under in vitro conditions depends on TGF-β expression, whereas in vivo, physical injury-induced EMT in LECs requires the involvement of the Smad3 pathway downstream of TGF-β. In addition to the canonical Smad pathway, the PI3K\u0026ndash;Akt pathway, which significantly enriched among the DEGs,serves as a non-canonical route capable of inducing EMT \u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e. TNF, which interacts extensively with TGF-β in tumor studies, also plays an important role in EMT \u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. Furthermore, AGE\u0026ndash;RAGE signaling pathway can lead to the production of various inflammatory mediators and directly inducing EMT \u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e. Studies also indicate that the AGE\u0026ndash;RAGE signaling pathway promotes the expression of TGF-β\u003csup\u003e61\u003c/sup\u003e. In summary, a history of cataract surgery leads to significant alterations in the expression of multiple genes associated with EMT.\u003c/p\u003e \u003cp\u003eThrough integrated bioinformatics analysis and validation across internal and external datasets, we identified FN1, SERPINE1, and THBS1 as hub genes. Among these, FN1 encodes fibronectin, a well-established marker of EMT and a critical intervention target \u003csup\u003e\u003cspan additionalcitationids=\"CR63 CR64\" citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e. Previous studies have shown that miR-142-3P can suppress cancer cell migration, invasion, and EMT by targeting FN1 and inactivating the FAK/ERK/PI3K signaling pathway \u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e. Similarly, THBS1 and SERPINE1 have also been widely reported to participate in fibrosis and EMT processes, primarily through activation of pathways such as TGF-β/Smad, PI3K/AKT/ERK \u003csup\u003e\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e, and ERK\u0026ndash;MAPK signaling \u003csup\u003e\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e\u003c/sup\u003e, which consistent with the enrichment and GSEA results in this study. In the context of PCO, FN1 remains recognized as an EMT marker \u003csup\u003e\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e\u003c/sup\u003e. Notably, a recent study suggested that targeting the HOTAIR/WTAP/THBS1 axis may prevent or treat PCO \u003csup\u003e\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e. In contrast, SERPINE1 has not been thoroughly investigated in PCO-related studies. To explore the potential relationship between these hub genes and the inflammatory microenvironment commonly associated with PCO, we further performed immune infiltration analysis.\u003c/p\u003e \u003cp\u003eOur immune infiltration analysis revealed significantly increased infiltration of M0 macrophages in the History of surgery group. Both M0 and M2 macrophages, which showed a positive correlation, have been implicated as key immune regulators in fibrosis and TGF-β signaling \u003csup\u003e\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e\u003c/sup\u003e. Macrophages expressing Fn1, and Arg1 are known to mediate fibrotic processes \u003csup\u003e\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e\u003c/sup\u003e. Since TGF-β in the lens microenvironment requires activation to exert its effects, macrophages may contribute through integrin-mediated TGF-β activation \u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e\u003c/sup\u003e. Consistent with this, study reported that macrophage infiltration promotes TGF-β/Smad signaling activity \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. GSEA of genes associated with M0 macrophages infiltration further highlighted significant enrichment in pathways promoting EMT, including ERK1/ERK2, TGF-β, MAPK, and integrin mediated signalling. Additionally, terms such as inflammatory cell chemotaxis and inflammatory response to wounding were enriched, suggesting a link between M0 macrophages infiltration and a sustained pro-inflammatory state in the anterior chamber following surgery. CD8\u0026thinsp;+\u0026thinsp;T cells and Resting CD4\u0026thinsp;+\u0026thinsp;memory T cells exhibited statistically significant differences in infiltration ratios between the two groups. Notably, a strong negative association (r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.8) was observed between resting CD8\u0026thinsp;+\u0026thinsp;T cells and Resting CD4\u0026thinsp;+\u0026thinsp;memory T cells. Previous studies have indicated that EMT activity correlates with increased CD8\u0026thinsp;+\u0026thinsp;T cells and reduced Resting CD4\u0026thinsp;+\u0026thinsp;memory T cells \u003csup\u003e\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e, \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e\u003c/sup\u003e. Furthermore, activation of TGF-β type II receptor signaling can lead to T cell CD8 exhaustion, while CD8\u0026thinsp;+\u0026thinsp;T cells themselves have been reported to mediate anti-inflammatory responses, that consistent with our GSEA results \u003csup\u003e\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e\u003c/sup\u003e. Hub genes also exhibited correlations with immune infiltration that aligned with overall infiltration trends. Together, these findings underscore a profound interconnection between the immune microenvironment and multiple pathological processes in PCO, including the expression of hub genes central to disease progression.\u003c/p\u003e \u003cp\u003eSubsequent molecular docking analyses of the proteins encoded by FN1, SERPINE1, and THBS1with CsA revealed that FN1 and SERPINE1exhibit favorable binding affinity with CsA and form stable hydrogen bonds. These results suggest that CsA may modulate their functions through these interactions, thereby affecting the EMT process associated with hub genes as well as the progression of TGF-β-related signaling pathways. These findings provide a mechanistic explanation for the previously reported inhibitory effect of CsA on PCO. Furthermore, the immunomodulatory properties of CsA, particularly its ability to reduce CD4 expression on T cells, suppress inflammatory cytokine release, and promote CD4\u0026thinsp;+\u0026thinsp;T cell apoptosis, align with the modulation required for counteracting the aberrant inflammatory environment in PCO as identified in our analysis. Collectively, these findings indicate that CsA represents a promising therapeutic intervention for PCO.\u003c/p\u003e \u003cp\u003eAs discussed, a history of surgery creates an aqueous humor environment with elevated TGF-β2 concentrations for LECs. TGF-β2-activated pathways, recognized in previous studies and supported by our analysis, are key in stimulating pathological changes in LECs and contributing to PCO development. To model this critical pathological environment, we exposed LECs to varying concentrations of TGF-β2, aiming to investigate subsequent alterations in gene expression and cellular functions. Our results demonstrated a significant decrease in the epithelial marker CDH1 following TGF-β2 stimulation. In contrast, the mesenchymal marker VIM showed a significant increase only at a high concentration (10 ng/ml). This suggests that TGF-β2 can suppress epithelial marker expression and promote mesenchymal marker expression in LECs, with the regulation of epithelial markers appearing more sensitive. The loss of epithelial phenotype and acquisition of mesenchymal characteristics are hallmarks of EMT. The reduction of epithelial markers diminishes intercellular tight junctions, potentially enhancing migration capability and disrupting normal tissue architecture. Corroborating this, our wound-healing assay confirmed that TGF-β2 enhances the migration capability of LECs, indicating that TGF-β2 indeed stimulates EMT-like changes and alters cell motility.\u003c/p\u003e \u003cp\u003eEMT involves complex physiological and pathological changes. It is not an absolute binary process. Cells undergoing functional changes can exhibit unexpected variations in marker expression, as observed in our experiments where epithelial and mesenchymal phenotypic shifts did not occur in a uniformly inverse manner\u003csup\u003e\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e, \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e\u003c/sup\u003e. In contrast, the genes FN1, SERPINE1, and THBS1, identified through our bioinformatic screening, showed significantly increased expression upon TGF-β2 stimulation. Their expression levels rose markedly with increasing TGF-β2 concentrations, demonstrating a strong concentration-dependent response. Consequently, FN1, SERPINE1, and THBS1 exhibit high sensitivity and correlation, positioning them as potential biomarkers that are more responsive to TGF-β2 exposure than classic EMT markers in LECs. Their association with immune cell infiltration and various pathological processes also suggests these genes may play critical roles in the pathological progression of LECs during PCO.\u003c/p\u003e \u003cp\u003eHowever, this study has limitations regarding the sample conditions. As mentioned previously, the anterior chamber is an immune-privileged site; thus, the lens capsules from the non-surgery control group, without blood-aqueous barrier disruption caused by cataract surgery, should theoretically exhibit no immune cell infiltration. Nevertheless, CIBERSORT algorithm revealed the presence of immune cells in the Non-surgery group, with a notably higher proportion of CD8\u0026thinsp;+\u0026thinsp;T cells compared to the History of surgery group. This observation may be attributed to the fact that all samples were taken from the donors 2 to 10 hours after their deaths. During this period, the postmortem breakdown of the blood-aqueous barrier likely leads to progressive loss of immune privilege, resulting in inflammatory cell infiltration even in normal anterior chambers \u003csup\u003e\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e\u003c/sup\u003e. Alternatively, the disruption of the blood-aqueous barrier may be attributed to the sample acquisition procedure itself. Furthermore, it is important to note that the CIBERSORT algorithm, which employs deconvolution analysis to infer the relative proportions of immune cells in a sample, cannot effectively distinguish whether the detected immune cells represent contamination at the time of extraction or are merely physically adjacent without physiological significance. To address this limitation, we integrated the immune cell proportions obtained from CIBERSORT with GSEA. This combined approach allowed us to examine the correlation between immune cell composition and the sample\u0026rsquo;s transcriptomic profile, thereby elucidating the impact on cellular pathways and clarifying the underlying biological significance. Taking this factor into account, and integrating findings from GSEA and immune infiltration analyses, a more refined interpretation can be drawn: the history of surgery group exhibits significant overexpression of pro-inflammatory genes, potentially promoting the chemotaxis and migration of M0 macrophages and resting CD4\u0026thinsp;+\u0026thinsp;memory T cells. In contrast, the Non-surgery group may maintain a uniquely anti-inflammatory microenvironment, attenuating inflammatory responses while facilitating the infiltration of CD8\u0026thinsp;+\u0026thinsp;T cells. This disparity in immune regulation likely underlies the differential activity of EMT-related pathways between the two groups.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eIn conclusion, our findings demonstrate that FN1, SERPINE1, and THBS1 exhibit strong discriminatory power in the gene expression profiles of human lens epithelium tissue from patients with a history of cataract surgery. These hub genes occupy statistically and biologically central positions within the relevant gene co-expression network. Their expression is likely central to the pathological changes occurring in this tissue and may serve as potential biomarkers for identifying TGF-β2-stimulated LECs. Furthermore, the relative infiltration levels of M0 macrophages and resting CD4\u0026thinsp;+\u0026thinsp;memory T cells within the tissue showed a significant positive correlation with both EMT and the activity of pro-inflammatory pathways. These findings suggest that their infiltration likely plays a critical role in driving postoperative pathological processes in LECs and is significantly associated with the expression of the identified hub genes. Ultimately, the potential and possible mechanism of action of CsA were revealed. Future research focused on elucidating the precise mechanisms through which these hub genes operate, and how specific immune cell populations are recruited and activated, will be essential for a deeper understanding of the molecular and cellular events that occur in LECs after surgery. Such insights also hold strong potential for developing targeted interventions aimed at modulating these key pathways or cell populations to effectively prevent the occurrence of PCO.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eElectronic supplementary material\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupplementary charts and chart files are provided together.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe sincerely appreciate the GEO database for its platform support and the contributors who uploaded critical datasets.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSJ and CG conceived and designed the study, supervised data collection, and coordinated the research framework. SJ performed data processing and formal analysis. Both SJ and CG interpreted the results, drafted the initial manuscript, and revised it critically for intellectual content. All authors participated in substantive discussions, reviewed the final manuscript, approved its submission, and consented to publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe dataset used and/or analyzed in this study is available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by Bingtuan Science and Technology Program (2024ZD074) and the Project of TianChiYingCai of Xinjiang Uygur Autonomous Region (CZ001203).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial number: Not Applicable\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003cstrong\u003e：\u003c/strong\u003e\u003cstrong\u003eNot Applicable\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur study utilizes data from the GEO database, which is publicly available. As such, our research complies with all applicable data use policies and presents no ethical issues\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBassnett, S., Shi, Y. \u0026amp; Vrensen, G.F. Biological glass: structural determinants of eye lens transparency. \u003cem\u003ePhilos Trans R Soc Lond B Biol Sci\u003c/em\u003e \u003cstrong\u003e366\u003c/strong\u003e, 1250-1264 (2011).\u003c/li\u003e\n\u003cli\u003eAbdulhussein, D. \u0026amp; Abdul Hussein, M. WHO Vision 2020: Have We Done It? \u003cem\u003eOphthalmic Epidemiol\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 331-339 (2023).\u003c/li\u003e\n\u003cli\u003eWorld Health Organization. Programme for the Prevention of, B. \u0026amp; Deafness (World Health Organization, 2000).\u003c/li\u003e\n\u003cli\u003eLiu, Y.C., Wilkins, M., Kim, T., Malyugin, B. \u0026amp; Mehta, J.S. Cataracts. \u003cem\u003eLancet\u003c/em\u003e \u003cstrong\u003e390\u003c/strong\u003e, 600-612 (2017).\u003c/li\u003e\n\u003cli\u003eWormstone, I.M., Wormstone, Y.M., Smith, A.J.O. \u0026amp; Eldred, J.A. Posterior capsule opacification: What\u0026apos;s in the bag? \u003cem\u003eProg Retin Eye Res\u003c/em\u003e \u003cstrong\u003e82\u003c/strong\u003e, 100905 (2021).\u003c/li\u003e\n\u003cli\u003eNibourg, L.M.\u003cem\u003e et al.\u003c/em\u003e Prevention of posterior capsular opacification. \u003cem\u003eExp Eye Res\u003c/em\u003e \u003cstrong\u003e136\u003c/strong\u003e, 100-115 (2015).\u003c/li\u003e\n\u003cli\u003eWay, C., Swampillai, A.J., Lim, K.S. \u0026amp; Nanavaty, M.A. Factors influencing aqueous flare after cataract surgery and its evaluation with laser flare photometry. \u003cem\u003eTher Adv Ophthalmol\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 25158414231204111 (2023).\u003c/li\u003e\n\u003cli\u003eWormstone, I.M. \u0026amp; Eldred, J.A. Experimental models for posterior capsule opacification research. \u003cem\u003eExp Eye Res\u003c/em\u003e \u003cstrong\u003e142\u003c/strong\u003e, 2-12 (2016).\u003c/li\u003e\n\u003cli\u003eLeydolt, C.\u003cem\u003e et al.\u003c/em\u003e Posterior Capsule Opacification With Two Hydrophobic Acrylic Intraocular Lenses: 3-Year Results of a Randomized Trial. \u003cem\u003eAm J Ophthalmol\u003c/em\u003e \u003cstrong\u003e217\u003c/strong\u003e, 224-231 (2020).\u003c/li\u003e\n\u003cli\u003eUrsell, P.G., Dhariwal, M., O\u0026apos;Boyle, D., Khan, J. \u0026amp; Venerus, A. 5 year incidence of YAG capsulotomy and PCO after cataract surgery with single-piece monofocal intraocular lenses: a real-world evidence study of 20,763 eyes. \u003cem\u003eEye (Lond)\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 960-968 (2020).\u003c/li\u003e\n\u003cli\u003eTaravati, P., Lam, D.L., Leveque, T. \u0026amp; Van Gelder, R.N. Postcataract surgical inflammation. \u003cem\u003eCurr Opin Ophthalmol\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 12-18 (2012).\u003c/li\u003e\n\u003cli\u003eJakobsson, G., Sundelin, K., Zetterberg, H. \u0026amp; Zetterberg, M. Increased levels of inflammatory immune mediators in vitreous from pseudophakic eyes. \u003cem\u003eInvest Ophthalmol Vis Sci\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e, 3407-3414 (2015).\u003c/li\u003e\n\u003cli\u003eZhao, Y.\u003cem\u003e et al.\u003c/em\u003e Expression Profiles of Inflammatory Cytokines in the Aqueous Humor of Children after Congenital Cataract Extraction. \u003cem\u003eTransl Vis Sci Technol\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 3 (2020).\u003c/li\u003e\n\u003cli\u003eShihan, M.H.\u003cem\u003e et al.\u003c/em\u003e Fibronectin has multifunctional roles in posterior capsular opacification (PCO). \u003cem\u003eMatrix Biol\u003c/em\u003e \u003cstrong\u003e90\u003c/strong\u003e, 79-108 (2020).\u003c/li\u003e\n\u003cli\u003eHao, C.\u003cem\u003e et al.\u003c/em\u003e Elevated Inflammatory Cytokines Persist in the Aqueous Humor Years After Cataract Surgery. \u003cem\u003eInvest Ophthalmol Vis Sci\u003c/em\u003e \u003cstrong\u003e66\u003c/strong\u003e, 12 (2025).\u003c/li\u003e\n\u003cli\u003eWormstone, I.M., Tamiya, S., Anderson, I. \u0026amp; Duncan, G. TGF-beta2-induced matrix modification and cell transdifferentiation in the human lens capsular bag. \u003cem\u003eInvest Ophthalmol Vis Sci\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, 2301-2308 (2002).\u003c/li\u003e\n\u003cli\u003eLewis, A.C. Interleukin-6 in the pathogenesis of posterior capsule opacification and the potential role for interleukin-6 inhibition in the future of cataract surgery. \u003cem\u003eMed Hypotheses\u003c/em\u003e \u003cstrong\u003e80\u003c/strong\u003e, 466-474 (2013).\u003c/li\u003e\n\u003cli\u003eFlokis, M. \u0026amp; Lovicu, F.J. FGF-2 Differentially Regulates Lens Epithelial Cell Behaviour during TGF-\u0026beta;-Induced EMT. \u003cem\u003eCells\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e (2023).\u003c/li\u003e\n\u003cli\u003eSi, W.\u003cem\u003e et al.\u003c/em\u003e IL-8 promotes lens capsular residual cells migration by down-regulates expression of E-cadherin and ZO-1 via the CXCR1/2-NF-\u0026kappa;B-RhoA signal pathway. \u003cem\u003eInt Immunopharmacol\u003c/em\u003e \u003cstrong\u003e142\u003c/strong\u003e, 113074 (2024).\u003c/li\u003e\n\u003cli\u003eJiang, J., Shihan, M.H., Wang, Y. \u0026amp; Duncan, M.K. Lens Epithelial Cells Initiate an Inflammatory Response Following Cataract Surgery. \u003cem\u003eInvest Ophthalmol Vis Sci\u003c/em\u003e \u003cstrong\u003e59\u003c/strong\u003e, 4986-4997 (2018).\u003c/li\u003e\n\u003cli\u003eGao, J.\u003cem\u003e et al.\u003c/em\u003e Mitochondrial permeability transition pore in inflammatory apoptosis of human conjunctival epithelial cells and T cells: effect of cyclosporin A. \u003cem\u003eInvest Ophthalmol Vis Sci\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, 4717-4733 (2013).\u003c/li\u003e\n\u003cli\u003ePeriman, L.M., Mah, F.S. \u0026amp; Karpecki, P.M. A Review of the Mechanism of Action of Cyclosporine A: The Role of Cyclosporine A in Dry Eye Disease and Recent Formulation Developments. \u003cem\u003eClin Ophthalmol\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 4187-4200 (2020).\u003c/li\u003e\n\u003cli\u003eLu, M., Li, H. \u0026amp; Xiang, X. Efficacy of topical cyclosporine A in optimizing ocular surface and relieving dry eye symptoms after cataract surgery: a systematic review and meta-analysis of randomized controlled trials. \u003cem\u003eInt Ophthalmol\u003c/em\u003e \u003cstrong\u003e45\u003c/strong\u003e, 201 (2025).\u003c/li\u003e\n\u003cli\u003eTeng, H.\u003cem\u003e et al.\u003c/em\u003e [The study of cyclosporin A modified intraocular lens preventing posterior capsular opacification in rabbit eyes]. \u003cem\u003eZhonghua Yan Ke Za Zhi\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, 110-116 (2016).\u003c/li\u003e\n\u003cli\u003eTeng, H., Sun, J., Wen, K., Han, G. \u0026amp; Tian, F. Observation of Cyclosporin A: Sustained Release Intraocular Lens Implantation in Rabbit Eyes. \u003cem\u003eCurr Eye Res\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, 1508-1515 (2022).\u003c/li\u003e\n\u003cli\u003eHeid, C.A., Stevens, J., Livak, K.J. \u0026amp; Williams, P.M. Real time quantitative PCR. \u003cem\u003eGenome Res\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 986-994 (1996).\u003c/li\u003e\n\u003cli\u003ePeng, L.\u003cem\u003e et al.\u003c/em\u003e Cellular and Molecular Network Characteristics of TARM1-Related Genes in Mycobacterium tuberculosis Infections. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e (2024).\u003c/li\u003e\n\u003cli\u003eWang, Z.\u003cem\u003e et al.\u003c/em\u003e Identification of Key Biomarkers Associated with Immunogenic Cell Death and Their Regulatory Mechanisms in Severe Acute Pancreatitis Based on WGCNA and Machine Learning. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e (2023).\u003c/li\u003e\n\u003cli\u003eShen, F.\u003cem\u003e et al.\u003c/em\u003e Integrative bioinformatics analysis of high-throughput sequencing and in vitro functional analysis leads to uncovering key hub genes in esophageal squamous cell carcinoma. \u003cem\u003eHereditas\u003c/em\u003e \u003cstrong\u003e162\u003c/strong\u003e, 38 (2025).\u003c/li\u003e\n\u003cli\u003eWang, H.\u003cem\u003e et al.\u003c/em\u003e Potential of immune-related genes as promising biomarkers for premature coronary heart disease through high throughput sequencing and integrated bioinformatics analysis. \u003cem\u003eFront Cardiovasc Med\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 893502 (2022).\u003c/li\u003e\n\u003cli\u003eHao, C.\u003cem\u003e et al.\u003c/em\u003e Transcriptomic Analysis of Human Lens Epithelium Tissue With and Without Cataract Surgery: Uncovering Novel Pathways of Post-Surgical Lens Epithelium Remodeling. \u003cem\u003eInvest Ophthalmol Vis Sci\u003c/em\u003e \u003cstrong\u003e66\u003c/strong\u003e, 28 (2025).\u003c/li\u003e\n\u003cli\u003eRitchie, M.E.\u003cem\u003e et al.\u003c/em\u003e limma powers differential expression analyses for RNA-sequencing and microarray studies. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, e47 (2015).\u003c/li\u003e\n\u003cli\u003eKanehisa, M. \u0026amp; Goto, S. KEGG: kyoto encyclopedia of genes and genomes. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 27-30 (2000).\u003c/li\u003e\n\u003cli\u003eKe, Y. \u0026amp; Ge, W. Identification of prognostic biomarkers in neuroblastoma using WGCNA and multi-omics analysis. \u003cem\u003eDiscov Oncol\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 469 (2024).\u003c/li\u003e\n\u003cli\u003eLangfelder, P. \u0026amp; Horvath, S. WGCNA: an R package for weighted correlation network analysis. \u003cem\u003eBMC Bioinformatics\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 559 (2008).\u003c/li\u003e\n\u003cli\u003eSzklarczyk, D.\u003cem\u003e et al.\u003c/em\u003e STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. \u003cem\u003eNucleic Acids Res\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, D607-d613 (2019).\u003c/li\u003e\n\u003cli\u003eChen, K.\u003cem\u003e et al.\u003c/em\u003e Identification of feature genes and key biological pathways in immune-mediated necrotizing myopathy: High-throughput sequencing and bioinformatics analysis. \u003cem\u003eComput Struct Biotechnol J\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 2228-2240 (2023).\u003c/li\u003e\n\u003cli\u003eGeng, R.\u003cem\u003e et al.\u003c/em\u003e Gene expression analysis in endometriosis: Immunopathology insights, transcription factors and therapeutic targets. \u003cem\u003eFront Immunol\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 1037504 (2022).\u003c/li\u003e\n\u003cli\u003eZhang, W.Y.\u003cem\u003e et al.\u003c/em\u003e Analysis and validation of diagnostic biomarkers and immune cell infiltration characteristics in pediatric sepsis by integrating bioinformatics and machine learning. \u003cem\u003eWorld J Pediatr\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 1094-1103 (2023).\u003c/li\u003e\n\u003cli\u003eGong, T., Jia, B., Lv, H., Zeng, L. \u0026amp; Zhong, D. Computational Analyses Identified Three Diagnostic Biomarkers Associated With Programmed Cell Death for Lung Adenocarcinoma. \u003cem\u003eHum Mutat\u003c/em\u003e \u003cstrong\u003e2025\u003c/strong\u003e, 1743829 (2025).\u003c/li\u003e\n\u003cli\u003eShariat, S.F., Capitanio, U., Jeldres, C. \u0026amp; Karakiewicz, P.I. Can nomograms be superior to other prediction tools? \u003cem\u003eBJU Int\u003c/em\u003e \u003cstrong\u003e103\u003c/strong\u003e, 492-495; discussion 495-497 (2009).\u003c/li\u003e\n\u003cli\u003eHan, M.\u003cem\u003e et al.\u003c/em\u003e Identification of hub genes and their correlation with immune infiltrating cells in membranous nephropathy: an integrated bioinformatics analysis. \u003cem\u003eEur J Med Res\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 525 (2023).\u003c/li\u003e\n\u003cli\u003eThilagar, S.S., Yadalam, P.K., Ronsivalle, V., Cicci\u0026ugrave;, M. \u0026amp; Minervini, G. Prediction of Interactomic HUB Genes in Periodontitis With Acute Myocardial Infarction. \u003cem\u003eJ Craniofac Surg\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 1292-1297 (2024).\u003c/li\u003e\n\u003cli\u003eLe, T., Aronow, R.A., Kirshtein, A. \u0026amp; Shahriyari, L. A review of digital cytometry methods: estimating the relative abundance of cell types in a bulk of cells. \u003cem\u003eBrief Bioinform\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e (2021).\u003c/li\u003e\n\u003cli\u003eLiu, C.J.\u003cem\u003e et al.\u003c/em\u003e GSCA: an integrated platform for gene set cancer analysis at genomic, pharmacogenomic and immunogenomic levels. \u003cem\u003eBrief Bioinform\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e (2023).\u003c/li\u003e\n\u003cli\u003eKubo, E., Shibata, T., Singh, D.P. \u0026amp; Sasaki, H. Roles of TGF \u0026beta; and FGF Signals in the Lens: Tropomyosin Regulation for Posterior Capsule Opacity. \u003cem\u003eInt J Mol Sci\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e (2018).\u003c/li\u003e\n\u003cli\u003eCreyghton, W.M., V\u0026ouml;lker-Dieben, H.J. \u0026amp; Jager, M.J. Transforming growth factor \u0026beta;2 levels in aqueous humor during corneal transplantation and cataract surgery. \u003cem\u003eOcul Immunol Inflamm\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 29-36 (1995).\u003c/li\u003e\n\u003cli\u003eDeDreu, J., Basta, M.D., Walker, J.L. \u0026amp; Menko, A.S. Immune Responses Induced at One Hour Post Cataract Surgery Wounding of the Chick Lens. \u003cem\u003eBiomolecules\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e (2023).\u003c/li\u003e\n\u003cli\u003eNovo, S.G.\u003cem\u003e et al.\u003c/em\u003e The Immediate Early Response of Lens Epithelial Cells to Lens Injury. \u003cem\u003eCells\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e (2022).\u003c/li\u003e\n\u003cli\u003eKruger, A.J.\u003cem\u003e et al.\u003c/em\u003e Postoperative inflammation after lens epithelial cell removal: 2 year results. \u003cem\u003eJ Cataract Refract Surg\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1380-1385 (2001).\u003c/li\u003e\n\u003cli\u003eLovicu, F.J., Shin, E.H. \u0026amp; McAvoy, J.W. Fibrosis in the lens. Sprouty regulation of TGF\u0026beta;-signaling prevents lens EMT leading to cataract. \u003cem\u003eExp Eye Res\u003c/em\u003e \u003cstrong\u003e142\u003c/strong\u003e, 92-101 (2016).\u003c/li\u003e\n\u003cli\u003eWang, X.\u003cem\u003e et al.\u003c/em\u003e Roles of growth factors in eye development and ophthalmic diseases. \u003cem\u003eZhejiang Da Xue Xue Bao Yi Xue Ban\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, 613-625 (2022).\u003c/li\u003e\n\u003cli\u003eHales, A.M., Chamberlain, C.G. \u0026amp; McAvoy, J.W. Cataract induction in lenses cultured with transforming growth factor-beta. \u003cem\u003eInvest Ophthalmol Vis Sci\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 1709-1713 (1995).\u003c/li\u003e\n\u003cli\u003eSaika, S.\u003cem\u003e et al.\u003c/em\u003e Smad3 signaling is required for epithelial-mesenchymal transition of lens epithelium after injury. \u003cem\u003eAm J Pathol\u003c/em\u003e \u003cstrong\u003e164\u003c/strong\u003e, 651-663 (2004).\u003c/li\u003e\n\u003cli\u003eLuo, K. Signaling Cross Talk between TGF-\u0026beta;/Smad and Other Signaling Pathways. \u003cem\u003eCold Spring Harb Perspect Biol\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e (2017).\u003c/li\u003e\n\u003cli\u003eGiarratana, A.O., Prendergast, C.M., Salvatore, M.M. \u0026amp; Capaccione, K.M. TGF-\u0026beta; signaling: critical nexus of fibrogenesis and cancer. \u003cem\u003eJ Transl Med\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 594 (2024).\u003c/li\u003e\n\u003cli\u003eYoshimatsu, Y.\u003cem\u003e et al.\u003c/em\u003e TNF-\u0026alpha; enhances TGF-\u0026beta;-induced endothelial-to-mesenchymal transition via TGF-\u0026beta; signal augmentation. \u003cem\u003eCancer Sci\u003c/em\u003e \u003cstrong\u003e111\u003c/strong\u003e, 2385-2399 (2020).\u003c/li\u003e\n\u003cli\u003eLiu, Z.W.\u003cem\u003e et al.\u003c/em\u003e Duality of Interactions Between TGF-\u0026beta; and TNF-\u0026alpha; During Tumor Formation. \u003cem\u003eFront Immunol\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 810286 (2021).\u003c/li\u003e\n\u003cli\u003eRamasamy, R., Shekhtman, A. \u0026amp; Schmidt, A.M. The multiple faces of RAGE--opportunities for therapeutic intervention in aging and chronic disease. \u003cem\u003eExpert Opin Ther Targets\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 431-446 (2016).\u003c/li\u003e\n\u003cli\u003eWu, T.T.\u003cem\u003e et al.\u003c/em\u003e AKR1B1-Induced Epithelial-Mesenchymal Transition Mediated by RAGE-Oxidative Stress in Diabetic Cataract Lens. \u003cem\u003eAntioxidants (Basel)\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e (2020).\u003c/li\u003e\n\u003cli\u003eCheng, M.\u003cem\u003e et al.\u003c/em\u003e HMGB1 Enhances the AGE-Induced Expression of CTGF and TGF-\u0026beta; via RAGE-Dependent Signaling in Renal Tubular Epithelial Cells. \u003cem\u003eAm J Nephrol\u003c/em\u003e \u003cstrong\u003e41\u003c/strong\u003e, 257-266 (2015).\u003c/li\u003e\n\u003cli\u003eZheng, H.\u003cem\u003e et al.\u003c/em\u003e PLA2R1 Inhibits Differentiated Thyroid Cancer Proliferation and Migration via the FN1-Mediated ITGB1/FAK Axis. \u003cem\u003eCancers (Basel)\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e (2023).\u003c/li\u003e\n\u003cli\u003eXiao, Z., Cai, Z., Deng, D., Tong, S. \u0026amp; Zu, X. An EMT-based risk score thoroughly predicts the clinical prognosis, tumor immune microenvironment and molecular subtypes of bladder cancer. \u003cem\u003eFront Immunol\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 1000321 (2022).\u003c/li\u003e\n\u003cli\u003eLi, B.\u003cem\u003e et al.\u003c/em\u003e Fibronectin 1 promotes melanoma proliferation and metastasis by inhibiting apoptosis and regulating EMT. \u003cem\u003eOnco Targets Ther\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 3207-3221 (2019).\u003c/li\u003e\n\u003cli\u003eZhou, J.\u003cem\u003e et al.\u003c/em\u003e Targeting PRDX2 to inhibit tumor growth and metastasis in triple-negative breast cancer: the role of FN1 and the PI3K/AKT/SP1 pathway. \u003cem\u003eJ Transl Med\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 434 (2025).\u003c/li\u003e\n\u003cli\u003eJiang, Y.\u003cem\u003e et al.\u003c/em\u003e MicroRNA-142-3P suppresses the progression of papillary thyroid carcinoma by targeting FN1 and inactivating FAK/ERK/PI3K signaling. \u003cem\u003eCell Signal\u003c/em\u003e \u003cstrong\u003e109\u003c/strong\u003e, 110792 (2023).\u003c/li\u003e\n\u003cli\u003eLiao, F.\u003cem\u003e et al.\u003c/em\u003e ECFC-derived exosomal THBS1 mediates angiogenesis and osteogenesis in distraction osteogenesis via the PI3K/AKT/ERK pathway. \u003cem\u003eJ Orthop Translat\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e, 12-22 (2022).\u003c/li\u003e\n\u003cli\u003eFeng, L., Li, G., Li, D., Duan, G. \u0026amp; Liu, J. Cuproptosis-related gene SERPINE1 is a prognostic biomarker and correlated with immune infiltrates in gastric cancer. \u003cem\u003eJ Cancer Res Clin Oncol\u003c/em\u003e \u003cstrong\u003e149\u003c/strong\u003e, 10851-10865 (2023).\u003c/li\u003e\n\u003cli\u003ePorreca, V.\u003cem\u003e et al.\u003c/em\u003e Pigment Epithelium-Derived Factor Inhibits Cell Motility and p-ERK1/2 Signaling in Intrahepatic Cholangiocarcinoma Cell Lines. \u003cem\u003eBiology (Basel)\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e (2025).\u003c/li\u003e\n\u003cli\u003eWang, D.\u003cem\u003e et al.\u003c/em\u003e PAI-1 overexpression promotes invasion and migration of esophageal squamous carcinoma cells. \u003cem\u003eYi Chuan\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e, 287-295 (2020).\u003c/li\u003e\n\u003cli\u003eWernecke, L., Keckeis, S., Reichhart, N., Strau\u0026szlig;, O. \u0026amp; Salchow, D.J. Epithelial-Mesenchymal Transdifferentiation in Pediatric Lens Epithelial Cells. \u003cem\u003eInvest Ophthalmol Vis Sci\u003c/em\u003e \u003cstrong\u003e59\u003c/strong\u003e, 5785-5794 (2018).\u003c/li\u003e\n\u003cli\u003eChen, X.\u003cem\u003e et al.\u003c/em\u003e LncRNA HOTAIR Interaction With WTAP Promotes m6A Methyltransferase Complex Assembly and Posterior Capsule Opacification Formation by Increasing THBS1. \u003cem\u003eInvest Ophthalmol Vis Sci\u003c/em\u003e \u003cstrong\u003e66\u003c/strong\u003e, 20 (2025).\u003c/li\u003e\n\u003cli\u003eWynn, T.A. \u0026amp; Barron, L. Macrophages: master regulators of inflammation and fibrosis. \u003cem\u003eSemin Liver Dis\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 245-257 (2010).\u003c/li\u003e\n\u003cli\u003eHoeft, K.\u003cem\u003e et al.\u003c/em\u003e Platelet-instructed SPP1(+) macrophages drive myofibroblast activation in fibrosis in a CXCL4-dependent manner. \u003cem\u003eCell Rep\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e, 112131 (2023).\u003c/li\u003e\n\u003cli\u003eRobertson, I.B. \u0026amp; Rifkin, D.B. Regulation of the Bioavailability of TGF-\u0026beta; and TGF-\u0026beta;-Related Proteins. \u003cem\u003eCold Spring Harb Perspect Biol\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e (2016).\u003c/li\u003e\n\u003cli\u003eWang, G.\u003cem\u003e et al.\u003c/em\u003e The pan-cancer landscape of crosstalk between epithelial-mesenchymal transition and immune evasion relevant to prognosis and immunotherapy response. \u003cem\u003eNPJ Precis Oncol\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 56 (2021).\u003c/li\u003e\n\u003cli\u003eLiu, R.\u003cem\u003e et al.\u003c/em\u003e Influence of Tumor Immune Infiltration on Immune Checkpoint Inhibitor Therapeutic Efficacy: A Computational Retrospective Study. \u003cem\u003eFront Immunol\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 685370 (2021).\u003c/li\u003e\n\u003cli\u003eFan, J.N., Ho, H. \u0026amp; Chiang, B.L. Characterization of novel CD8(+) regulatory T cells and their modulatory effects in murine model of inflammatory bowel disease. \u003cem\u003eCell Mol Life Sci\u003c/em\u003e \u003cstrong\u003e81\u003c/strong\u003e, 327 (2024).\u003c/li\u003e\n\u003cli\u003eXie, F.\u003cem\u003e et al.\u003c/em\u003e Breast cancer cell-derived extracellular vesicles promote CD8(+)\u0026thinsp;T cell exhaustion via TGF-\u0026beta; type II receptor signaling. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 4461 (2022).\u003c/li\u003e\n\u003cli\u003eLamouille, S., Xu, J. \u0026amp; Derynck, R. Molecular mechanisms of epithelial-mesenchymal transition. \u003cem\u003eNat Rev Mol Cell Biol\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 178-196 (2014).\u003c/li\u003e\n\u003cli\u003eZeisberg, M. \u0026amp; Neilson, E.G. Biomarkers for epithelial-mesenchymal transitions. \u003cem\u003eJ Clin Invest\u003c/em\u003e \u003cstrong\u003e119\u003c/strong\u003e, 1429-1437 (2009).\u003c/li\u003e\n\u003cli\u003eTripathi, R.C., Millard, C.B. \u0026amp; Tripathi, B.J. Protein composition of human aqueous humor: SDS-PAGE analysis of surgical and post-mortem samples. \u003cem\u003eExp Eye Res\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 117-130 (1989). \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"posterior capsular opacification, inflammation, RNA sequencing, transcriptomics, lens epithelial cells","lastPublishedDoi":"10.21203/rs.3.rs-8641774/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8641774/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003ePurpose\u003c/h2\u003e \u003cp\u003eThis study aims to employ transcriptomic analysis to investigate the characteristics of immune cell infiltration during Posterior capsular opacification (PCO) following cataract surgery, identify hub genes, elucidate the associated molecular pathways, and explore the potential interventional role of Cyclosporine A (CsA) in this process.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eData from lens capsules of non-cataract donors and those with cataract surgery history were obtained from the GEO database. Hub genes were identified through a systematic bioinformatic approach. The final hub genes were validated internally and with external datasets. Their correlations with immune profiles were assessed via CIBERSORT, and pathway enrichment analysis was conducted for immune-related genes. Cellular experiments simulated pathological conditions to validate hub gene expression, and molecular docking evaluated the binding affinity of CsA to the encoded proteins.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eHub genes (FN1, SERPINE1, THBS1) were identified and validated. Their expression was associated with significant alterations in immune cell infiltration (resting CD4\u0026thinsp;+\u0026thinsp;memory T cells, M0 macrophages, CD8\u0026thinsp;+\u0026thinsp;T cells), which correlated with the genes and PCO-related pathways. CsA exhibited favorable binding affinity to proteins encoded by FN1 and SERPINE1.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eHub genes demonstrated strong discriminatory power in the gene expression profiles of lens capsules with a history of cataract surgery, occupying central positions within the co-expression network. Alterations in immune cell infiltration, closely linked to these hub genes, are significantly associated with pathways involved in PCO. CsA may inhibit PCO by targeting the proteins encoded by these hub genes and through its immunomodulatory functions.\u003c/p\u003e","manuscriptTitle":"Analysis and validation of hub genes and immune cell infiltration characteristics in lens epithelium tissue with cataract surgical history","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-30 16:23:09","doi":"10.21203/rs.3.rs-8641774/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f88b36dc-eb6a-4fab-a8b3-e18abef65093","owner":[],"postedDate":"January 30th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-04-06T10:56:04+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-30 16:23:09","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8641774","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8641774","identity":"rs-8641774","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}