Identification of biomarkers associated with integrated stress response in pituitary adenomas based on bioinformatics

Identification of biomarkers associated with integrated stress response in pituitary adenomas based on bioinformatics | 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 Identification of biomarkers associated with integrated stress response in pituitary adenomas based on bioinformatics Yuan Sui, Zhiqian Yang, Wei Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8124800/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 12 You are reading this latest preprint version Abstract Background: Integrated stress response (ISR) participates in the development of tumor, however, the roles of ISR-related genes (ISR-RGs) in pituitary adenoma (PA) are unclear. The goal of this study was to identify the biomarkers related to ISR-RGs in PA and explore their action mechanism. Methods: In the present study, the public datasets from Gene Expression Omnibus database (GEO) database and ISR-RGs from literature were analyzed to obtain biomarkers for PA using differential expression analysis, machine learning algorithm, Boruta analysis and expression validation. Results: The results demonstrated that BCL2, ERN1, PMAIP1 and TWIST1, which all possessed significantly lower expression in PA samples in training and testing sets, were selected as biomarkers for PA. Subsequently, the nomogram showed good performance in predicting PA risk based on these biomarkers. BCL2 and PMAIP1 were both validated to locate in Chromosome 18, and BCL2 mainly functioned in cytoplasm and PMAIP1 worked in extracellular region. Additionally, the regulatory networks revealed that biomarkers might work through interacting with BBC3, AC004687, hsa-miR-142-5p, TP53 and STAT3, which provided significant insights into mechanism of biomarkers. Strong correlations were found to exist between activated dendritic cells and TWIST1, as well as between PMAIP1 and T follicular helper cells, which contributed to a more comprehensive understanding of the mechanism. Finally, we totally obtained 36 potential drugs interacting with biomarkers, including sodium arsenite, cyclosporine and valproic acid. Conclusions: This study identified BCL2, ERN1, PMAIP1 and TWIST1 as biomarkers for PA and revealed their action mechanism, which provided important references for treatment and study of PA. Pituitary adenoma Integrated stress response Regulatory network Biomarkers Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Highlight box Key findings This study identified four biomarkers for pituitary adenomas (PA), namely BCL2, ERN1, PMAIP1, and TWIST1. What is known and what is new? It is known that the integrated stress response (ISR) is involved in tumor development. However, the role of ISR - related genes in PA remains unclear. This study, for the first time, utilized bioinformatics approaches to clarify the above - mentioned four biomarkers, revealing their regulatory networks, signaling pathways, and relationships with immune cells. What is the implication, and what should change now? The study provides important references for PA treatment and research. In the future, experimental verification of the functions of these biomarkers and the drug prediction results should be conducted to promote the clinical translation of precise PA classification and targeted therapy. 1. Introduction PA is a common neuroendocrine tumor, a benign epithelial neoplasm originating from the anterior pituitary lobe. The incidence of PA ranks third among intracranial tumors, just behind glioma and meningioma[1]. The incidence of PA has been reported to account for 10-15% of all primary tumors in the central nervous system[1]. Approximately 10% of PAs exhibit active behavior, and among them, 0.2% develop metastases and are thus classified as pituitary carcinomas [1,2]. Pituitary adenomas include various hormone-producing subtypes such as Prolactinomas, ACTH-secreting adenomas associated with Cushing's syndrome, and neoplasms generating excessive growth hormone. This category also comprises adenomas that abnormally release TSH or gonadotrophin, along with rare mixed-type variants demonstrating combined hormonal secretion patterns. Each subtype exhibits distinct pathophysiological features requiring tailored diagnostic and therapeutic approaches[2]. Pituitary adenomas (PAs) exhibit diverse classification criteria. Based on size, they are categorized as microadenomas (4 cm). Classification based on radiological, biological characteristics, and invasiveness distinguishes functional from non-functional subtypes, as well as invasive from non-invasive subtypes. Invasiveness is commonly assessed using the Knosp classification system, which evaluates the tumor's relationship to the internal carotid artery[3,4]. Due to the critical physiological functions of the pituitary gland and its proximity to neurovascular structures, PAs can lead to significant morbidity and mortality[5]. Therapeutic approaches include surgery (e.g., transsphenoidal surgery), radiotherapy, chemotherapy, immunotherapy, and molecularly targeted therapies[6]. However, current treatments have significant limitations: pharmacological therapies may cause adverse effects such as gastrointestinal reactions, while surgery and radiotherapy carry risks including hypopituitarism[7–9]. Consequently, elucidating the molecular mechanisms underlying PAs and developing targeted therapies are of paramount importance. Integrated Stress Response (ISR) is an important stress-support pathway. By regulating the rate of protein synthesis, it is increasingly regarded as a determinant in tumorigenesis[10]. Four stress-sensing kinases — EIF2AK1, EIF2AK2, EIF2AK3, and EIF2AK4 (Eukaryotic Translation Initiation Factor-2α Kinases 1-4) — mediate cellular responses to diverse stressors. These kinases converge on a shared molecular mechanism: phosphorylation of a conserved serine residue within the Eukaryotic Translation Initiation Factor 2 (eIF2) complex[11]. A key functional attribute of the ISR pathway is its ability to dynamically modulate intracellular levels of the Ternary Complex. The Ternary Complex is structurally defined by the heterotrimeric organization of eIF2, comprising three stoichiometrically equivalent subunits: α, β, and γ. The functional state of the eIF2 heterotrimer is modulated by phosphorylation of its α subunit. Within this complex, the γ subunit harbors a catalytic domain essential for GTPase-activating protein (GAP) activity and mediates the recruitment of the Ternary Complex to mRNA during translation initiation[12,13]. Dysregulation of the eIF2γ subunit is linked to developmental and functional abnormalities in the hypothalamic-pituitary axis, hypopituitarism, and metabolic disorders such as pancreatic insufficiency and disrupted glucose homeostasis[14]. Studies have revealed the two-sided nature of the ISR, exhibiting both cytoprotective functions and apoptosis-inducing capacities. Targeted modulation of ISR signaling elements has emerged as a viable therapeutic approach in oncology treatment strategies[15,16]. Despite widespread activation of the Integrated Stress Response across various malignancies, the mechanistic role of ISR in organ-specific tumorigenesis and malignant evolution continues to pose significant research challenges[17]. This study, based on public databases and ISR-related literature data, aims to identify ISR-related biomarkers in PA through bioinformatics methods such as the combination of the Least absolute shrinkage and selection operator (LASSO) algorithm and the Boruta algorithm. Employing immunological profiling, molecular interaction network architecture, and therapeutic compound screening, this investigation aims to elucidate groundbreaking perspectives for deciphering ISR-mediated pathways in PA pathogenesis and advancing patient-specific immunomodulatory interventions. 2 Materials and methods 2.1 Data acquisition In this study, GSE26966 (GPL570 platform; containing 14 pituitary adenoma (PA) tissue samples and 9 normal pituitary tissue samples) and GSE63357 (GPL570 platform; containing 20 PA tissue samples and 5 normal pituitary tissue samples) were both acquired from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/) database, among which GSE26966 was the training set and GSE63357 was the testing set[18,19]. Additionally, totally 529 integrated stress response related genes (ISR-RGs) were found from public paper[20] (Table S1). 2.2 Identification of candidate genes To explore the mechanism of PA development, the candidate genes were identified. The study firstly found the differentially expressed genes (DEGs) between PA and normal samples in GSE26966 using limma package [21](v 3.54.0) (p.adj 1), and the results, including top five up-regulated and down-regulated DEGs in PA samples, were labeled in volcano diagram plotted by ggplot2 package[22] (v 3.4.1). Additionally, top ten up-regulated and down-regulated DEGs in PA samples were displayed in heatmap plotted by ComplexHeatmap package[23] (v 2.14.0). Subsequently, the DEGs that belonged to ISR-RGs were further picked out to participate in following analyses as candidate genes using ggvenn package[24] (v 0.1.9), whose results were also plotted as a Venn diagram. 2.3 Functional analyses and construction of protein-protein interaction (PPI) network of candidate genes To further study the functions and pathways that candidate genes participated in, Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses (p < 0.05) were carried out using the clusterProfiler package[25] (v 4.2.2), in which top five significant terms in GO, comprising biological process (BP), cellular component (CC) and molecular function (MF), as well as KEGG were separately displayed using ggplot2 package (v 3.4.1). Moreover, all candidate genes were input into the Search Tool for the Retrieval of Interacting Genes (STRING, http://string-db.org) to establish the PPI network and analyze the protein interactions of candidate genes (confidence score ≥ 0.4), whose results were plotted using Cytoscape software[26] (v 3.8.2). To further lock the crucial DEGs that played significant roles in PA development, the CytoHubba within Cytoscape (v 3.8.2) was firstly used to compute the scores for all candidate genes using maximal neighborhood component (MNC), maximal clique centrality (MCC), edge-betweenness path centrality (EPC), respectively. Then the hub genes were obtained through overlapping the top 20 genes with the highest score in each algorithm, and the results were displayed in a Venn diagram plotted by VennDiagram package[27] (v 1.7.1). 2.4 Identification of biomarkers The crucial biomarkers played significant roles in PA development, and to lock crucial biomarkers, least absolute shrinkage and selection operator (LASSO) algorithm and Boruta analysis were separately conducted for all hub genes to perform this task. Firstly, LASSO characteristic genes were found from hub genes through 3-fold cross validation using glmnet package[28] (v 4.1.4), when the lambda approached the minimum. Secondly, Boruta characteristic genes were also obtained from all hub genes through Boruta analysis using Boruta package[29] (v 8.0.0) (pValue = 0.01, maxRuns = 50). The intersection genes of above two groups were then collected as potential candidate biomarkers using ggvenn package (v 0.1.9). Subsequently, expression validation for potential candidate biomarkers was successively carried out in GSE26966 and GSE63357 to find the biomarkers that existing consistent expression trends in two dataset using Wilcoxon test (p < 0.05). 2.5 Chromosome localization analysis and subcellular localization of biomarkers To deeply explore the functions of biomarkers, the distributions of biomarkers in chromosomes were identified using Ensembl database (https://useast.ensembl.org/Homo_sapiens/Info/Index), and the results were plotted using RCircos package[30] (v 1.2.2). Additionally, the amino acid sequences of biomarkers were found from the Protein database (https://ncbi.nlm.nih.gov/protein/), and then the localizations of biomarkers in intracellular were predicted using mRNALocater database (http://bio-bigdata.cn/mRNALocater/result/), whose results were displayed using ggplot2 package (v 3.4.1). 2.6 Construction and verification of line nomogram To explore whether biomarkers can predict the prevalence of PA, in all samples of GSE26966, the rms package (PMID: 30686944) (v 6.8-1) was used to construct a nomogram. Each biomarker was scored separately, with one score corresponding to each biomarker. The scores of all factors were summed up to obtain the total point, and then the incidence of PA was inferred according to the total point. To evaluate the predictive performance of the nomogram model, the calibration curve generated by the rms package (v 6.8-1) was used to intuitively show the relationship between the predicted probability values and the true probability values. The Hosmer-Lemeshow test (HL test) served as the model fitting index. Its principle lies in judging the dispersion between the predicted values and the true values. If the p-value is greater than 0.05, it indicates that the HL test is passed. The rmda package (PMID: 38855330) (v 1.6) was employed to draw the decision curve for the nomogram model. 2.7 Construction of gene-gene-interaction (GGI) and molecular regulatory networks To explore the interactions that could help biomarkers complete their functions, the GGI network was predicted using GeneMANIA database (https://genemania.org/). Furthermore, the lncRNA-miRNA-biomarker network for biomarkers was constructed using the miRNAs predicted from the TarBase (v 9.0) database (https://dianalab.e-ce.uth.gr/tarbasev9) and long noncoding RNAs (lncRNAs) targeted to miRNAs predicted from the StarBase database (https://rnasysu.com/encori/) (clipExpNum>30). Additionally, transcription factors (TFs) that were targeting biomarkers were predicted from the JASPAR (https://jaspar.elixir.no/) to further perfect the TF-miRNA-mRNA network. The above networks were all plotted by Cytoscape (v 3.8.2). 2.8 Gene set enrichment analysis (GSEA) Moreover, to further study the functions that biomarkers participated in to affect the development of PA, GSEA was performed. Firstly, the Spearman correlations between biomarkers and the remaining genes in GSE26966 were computed using psych package[31] (v 2.1.6), and then the remaining genes were ranked based on the correlations. Secondly, GSEA was performed based on the “c2.cp.kegg.v2023.1.Hs.symbols.gmt” from the Molecular Signatures Database (MSigDB, http://software.broadinstitute.org/gsea/msigdb) using the clusterProfiler package (v 4.2.2) (p.adj 1), and the results were then visualized in ridge diagrams using enrichplot package[32] (v 1.18.3). 2.9 Immune infiltration analysis To investigate the functional dynamics of biological indicators within immunological niches, the study implemented a methodological framework employing single-sample Gene Set Enrichment Analysis (ssGSEA) to systematically quantify infiltration indices across 28 distinct immune cell populations[33] using GSVA package[34] (v 1.42.0). Then the enrichment score differences of each immune cell between PA and normal samples were identified to find differential immune cells using Wilcoxon test (p 0.3, p < 0.05) revealed the correlations between differential immune cells, as well as between differential immune cells and biomarkers using the cor function within psych package (v 2.1.6). Different immune cells perform distinct functions. The expressions of biomarkers vary among different immune cells, which is likely to have an impact on the functions of immune cells. Therefore, the Human Protein Atlas (HPA) database (https://www.proteinatlas.org/) is utilized to analyze the expression profiles of biomarkers in eighteen different types of immune cells. 2.10 Drug prediction To provide more references for the treatment of PA, the Comparative Toxicogenomics Database (CTD; http://ctdbase.org/) was used to predict the potential drugs that interacted with biomarkers (interaction count > 5), whose results were shown in a drug-biomarker network plotted by Cytoscape (v 3.8.2). 2.11 Statistical analysis R language (v 4.2.3) was leveraged in this study for bioinformatics analyses. The Wilcoxon test was used to perform statistical analysis, and a p-value less than 0.05 was considered significant. 3 Results 3.1 Identification of 18 hub genes To explore the roles that ISR-RGs played in PA, totally 2,654 DEGs, comprising 1,058 up-regulated and 1,596 down-regulated genes in PA samples, were firstly identified from the GSE26966 (Fig.1a-b, Table S2). Subsequently, a sum of 91 ISR-RGs were further proved to exhibit significant differences in PA and normal samples (p.adj < 0.05), and then were renamed as candidate genes for further analyses (Fig.1c, Table S3). In the following analyses, multiple pathways were found that candidate genes might participate in, comprising 1,121 BP terms, 63 CC terms, 82 MF terms, as well as 92 terms from KEGG (Fig.1d-e, Table S4-5), which suggested all candidate genes might affect PA development through multiple response ways to oxygen, as well as indicated they might function in membrane structures and cause multiple complications, such as colorectal cancer and hepatitis B. Moreover, PPI network indicated that there were 243 interactions between proteins of 75 candidate genes (Fig.2a, Table S6), in which BCL2 was validated to interact with BTG2 and BRCA1, and TWIST1 had interactions with ERN1 and EPHA2. Subsequently, through overlapping the top 20 genes with the highest scores of three algorithm, totally 18 genes were identified as hub genes to further screen biomarkers for PA (Fig.2b, Table S7). 3.2 BCL2, ERN1, PMAIP1 and TWIST1 as biomarkers for PA In the following analyses, LASSO algorithm firstly locked seven hub genes, including BCL2, ATF3, ATF4, ERN1, PMAIP1, BRCA1 and TWIST1, when log(lambda.min) equaled -5.4173 (Fig.3a-b). Meanwhile, a total of 17 hub genes except MYC from 18 hub genes were assessed as confirmed through Boruta analysis (Fig.3c). Furthermore, the intersection genes of above two groups, including BCL2, ATF3, ATF4, ERN1, PMAIP1, BRCA1 and TWIST1, were renamed as potential candidate biomarkers (Fig.3d). According to the expression validation, BCL2, ERN1, PMAIP1 and TWIST1 were all further validated to have the significant lower expression in PA samples in GSE26966 and GSE63357 (p < 0.05), and then selected as biomarkers to explore the mechanism of PA development (Fig.3e-f). 3.3 Establishment and assessment of a nomogram for PA diagnosis Based on the biomarkers (BCL2, ERN1, PMAIP1, TWIST1), a diagnostic model for PA was established. The higher the total score, the greater the probability of having PA (Fig.4a). The slope of the calibration curve is close to 1, and with a p-value of 0.668, it indicates that the HL test is passed. In other words, there is no significant difference between the predicted values and the true values, suggesting that the model has a good degree of fitting and prediction accuracy. Decision Curve Analysis (DCA) is a method for evaluating clinical prediction models, diagnostic tests, and molecular markers (Fig.4b). The DCA results indicated that patients can benefit from the diagnostic model developed using the four biomarkers with a threshold probability ranging from 0 to 1 (Fig.4c). Overall, it shows that our diagnostic model has a strong ability to distinguish between PA patients and healthy controls. 3.4 Chromosome localization and sub-cellular localization of biomarkers According to chromosome localization, BCL2 and PMAIP1 were both validated to locate in chromosome 18, ERN1 was detected in chromosome 17, and TWIST1 was found in chromosome 7 (Fig.5a), which was helpful to further study the function of biomarkers and the genetic patterns of related diseases. Additionally, BCL2, ERN1, and TWIST1 were proved to be able to affect the development of PA through functioning in cytoplasm and nucleus, and PMAIP1 mainly worked in extracellular region (Fig.5b). 3.5 Regulatory networks of biomarkers Based on GeneMANIA database, GGI network was established for all biomarkers, which suggested that PMAIP1 and BCL2 both could affect the development of PA through multiple pathways, such as regulating the membrane permeability and functioning in mitochondrial membrane, as well as ERN1 could interact with BBC3 and BID to participate in such pathways (Fig.6a). Additionally, the lncRNA-miRNA-biomarker network uncovered that AC004687.1 could regulate PMAIP1 and ERN1 through hsa-miR-142-5p, as well as BCL2 and TWIST1 could be regulated by AC021078.1 through hsa-miR-106a-5p (Fig.6b). Furthermore, more regulatory pathways of biomarkers were revealed by the TF-miRNA-mRNA network, in which STAT3 could regulate BCL2 through hsa-mir-7-1-3p, POU2F2 regulated PMAIP1 through hsa-let-7a-2-3p, as well as TP53 affected ERN1 through bonding to hsa-mir-9-3p and hsa-miR-545-5p (Fig.6c). 3.6 Biomarkers affected PA development through multiple pathways According to the results of GSEA, TWIST1 was enriched in 42 pathways, ERN1 was proved to participate in 19 pathways, PMAIP1 was found in 18 pathways, and BCL2 was involved in 16 pathways, in which oxidative phosphorylation was a co-pathway for all biomarkers, as well as ERN1, PMAIP1, and BCL2 could all function in ribosome to affect PA development (Fig.7a-d, Table S8-11). 3.7 Biomarkers affected PA development through multiple immune cells To explore the immune microenvironment that helped biomarkers affect PA development, totally 20 immune cells were validated to exhibit remarkable differences of enrichment scores between PA and normal samples (p <0.05) and renamed as differential immune cells, in which central memory CD4 T cell, immature dendritic cell and gamma delta (γδ) T cell had the highest enrichment scores in two groups (Fig.8a-b). Moreover, four differential immune cells, including CD56bright natural killer (NK) cell, γδ T cell, T follicular helper cell and type 2 T help cell, all existed higher enrichment scores in PA samples, which suggested that they might play significant roles in PA development. In subsequent analyses, multiple correlations were proved to exist between differential immune cells, in which the strongest positive correlation was detected between activated dendritic cell and central memory CD8 T cell (cor = 0.92, p < 0.05), as well as immature dendritic cell was validated to have the strongest negative correlation with type 2 T helper cell (cor = -0.73, p < 0.05) (Fig.8c). All these findings provided novel references for the study of immune response of PA. In addition to this, this study also revealed the correlations between differential immune cells and biomarkers. Interestingly, except CD56bright NK cell, γδ T cell, T follicular helper cell and type 2 T help cell, all remaining cells were all had strong positive correlations with all biomarkers, in which activated dendritic cell was proved to most strongly correlated with TWIST1 (cor = 0.88, p < 0.05) (Fig.8d). Among these immune cells, regulatory T cell exhibited the highest expression level of BCL2, closely followed by memory CD4 T cell and memory 8 cell. The expression levels in these cells were notably above 5 nTPM (Fig.8e). Basophils exhibited the highest expression level of ERN1, reaching approximately 18 nTPM, which was significantly higher than other cell types (Fig.8f). Basophils had the highest expression level of PMAIP1, reaching approximately 160 nTPM. Eosinophils followed with a relatively high expression, around 100 nTPM (Fig.8g). However, TWIST1 was not expressed in immune cells. These findings underscore the intricate interplay between immune cells and biomarkers in PA, unveiling potential therapeutic targets and shedding light on the complex immunopathogenesis of PA. 3.8 Drug prediction for PA According to CTD database, totally 36 drugs were predicted for BCL2, PMAIP1 and TWIST1, in which 33 drugs were found for BCL2, six drugs for PMAIP1, and only valproic acid was predicted for TWIST1 (Table S12). Referring to the drug-biomarkers network, sodium arsenite was predicted to interact with BCL2, cyclosporine was predicted to interact with PMAIP1, as well as valproic acid could interact with TWIST1, which all provided significant insights for the action mechanisms of BCL2, PMAIP1 and TWIST1, as well as furnished important references for optimizing the treatment strategies of PA (Fig.9). 4. Discussion 4.1 Clinical and Biological Context of Pituitary Adenomas Pituitary adenomas (PAs) are common central nervous system tumors with distinct clinical management paradigms[35–37]. Although most PAs are benign, invasive subtypes frequently exhibit high recurrence rates due to incomplete resection[38,39], necessitating novel therapeutic targets. The integrated stress response (ISR), an evolutionarily conserved signaling pathway, is a key regulator of proteostasis and disease pathogenesis[10]. Dysregulation of the ISR has been implicated in diverse pathological processes, including neurodegenerative diseases, tumorigenesis, and metabolic disorders[10,13], positioning the ISR pathway as a potential therapeutic target. Based on transcriptomic datasets from the Gene Expression Omnibus (GEO), this study identified four potential biomarkers—BCL2, ERN1, PMAIP1, and TWIST1—using bioinformatic methods. Further analyses, including Gene Set Enrichment Analysis (GSEA) and immune infiltration assessment, explored their biological functions and underlying regulatory mechanisms, offering new insights for the treatment of pituitary adenomas. 4.2 Mechanistic Insights into ISR-Related Biomarkers 4.2.1 BCL2: Apoptosis Regulation and Immune Modulation Located at chromosome 18q21.33, BCL2(Apoptosis regulator Bcl-2) serves as a master regulator of apoptotic signaling through its mitochondrial and ER membrane localization[40]. In PA pathogenesis, BCL2 overexpression correlates with tumor invasiveness and therapeutic resistance[41]. Our bioinformatics analyses corroborate its central role in ISR networks, suggesting that ISR-mediated stabilization of BCL2 may create a pro-survival niche[42]. Notably, the observed high BCL2 expression in regulatory T cells (Tregs) implies dual functionality: direct tumor cell protection via apoptosis inhibition[40] and indirect immune suppression through Treg-mediated IL-10/TGF-β secretion. 4.2.2 ERN1: ER Stress Sensor and Metabolic Adaptor As the molecular transducer of the unfolded protein response (UPR), ERN1 (IRE1α) coordinates cellular adaptation to ER stress through XBP1 splicing[43–45]. In PAs, ERN1 activation promotes tumor cell survival under metabolic stress[46]. Our findings extend this understanding by revealing its significant enrichment in basophils, suggesting crosstalk between ER stress signaling and basophil-mediated inflammatory responses (e.g., IL-4/IL-13 release). This crosstalk may foster chronic inflammation and angiogenesis, creating a tumor-permissive microenvironment. 4.2.3 PMAIP1: Context-Dependent Apoptotic Regulator The BH3-only protein PMAIP1 (Phorbol-12-myristate-13-acetate-induced protein 1) demonstrates paradoxical roles in PA progression. While generally suppressed in gonadotroph tumors through TP53-dependent mechanisms[47], its marked upregulation in invasive subtypes[48] suggests stage-specific functionality. Our multi-omics analyses position PMAIP1 as an ISR-ATF4 transcriptional target[49,50], potentially explaining its stress-inducible expression patterns. The observed basophil enrichment further implicates PMAIP1 in inflammatory microenvironment remodeling. 4.2.4 TWIST1: EMT Orchestrator and Immune Modulator TWIST1's (Twist-related protein 1) functional dichotomy in solid tumors [51–55] finds particular relevance in PAs. While absent in immune cells, its strong correlation with activated dendritic cells (DCs) reveals a novel immune-regulatory dimension. Experimental evidence suggests TWIST1 may drive DC dysfunction through antigen presentation impairment[56], while simultaneously promoting EMT-mediated tumor invasion[57,58]. This dual mechanism positions TWIST1 as a master regulator connecting tumor cell plasticity with immune evasion. 4.3 Therapeutic Implications and Model Validation 4.3.1 Dual-Targeting Therapeutic Strategy The ISR biomarkers' dual functionality - direct tumor promotion and immune microenvironment remodeling - necessitates combinatorial therapeutic approaches[10]. Our proposed strategy integrates ISR inhibitors (e.g., ISRIB) with immune checkpoint modulators (e.g., anti-PD-1 antibodies) to concurrently target both tumor cell-autonomous pathways and immune evasion mechanisms. 4.3.2 Diagnostic Model Performance and Limitations While the BCL2/ERN1/PMAIP1/TWIST1-based nomogram demonstrates exceptional discriminative capacity (AUC=1), potential overfitting in the GSE26966 training cohort mandates external validation. Future model optimization should incorporate imaging parameters (e.g., Knosp grading) and serum biomarkers to enhance clinical applicability. 4.3.3 Valproic Acid: A Repurposing Candidate Using a screening approach based on the Connectivity Map Database (CTD), valproic acid (VPA) has been identified as a potential therapeutic candidate associated with the transcription factor TWIST1. Initially discovered serendipitously as an antiepileptic drug[59], VPA's core pharmacological actions include enhancing GABAergic neurotransmission[60], and it has been validated through decades of clinical use[61]. Notably, VPA exhibits complex context-dependent effects in oncology: on one hand, multiple studies indicate that VPA and its analogs may possess pro-cancer potential[62], with mechanisms involving activation of specific signaling pathways such as Wnt/β-catenin[63] or promotion of epithelial-mesenchymal transition (EMT)-related processes[64]; on the other hand, its property as a potent histone deacetylase (HDAC) inhibitor[65] forms a critical mechanistic basis for its antitumor activities[66]. This HDAC inhibitory activity is also considered central to its antiseizure effects[67], mood-stabilizing properties[68], and potential neuroprotective roles being explored for neurodegenerative diseases and multiple sclerosis[69]. However, its related mechanisms of action in pituitary adenomas warrant further investigation. 5. Conclusions This study systematically identified four integrated stress response (ISR)-related biomarkers—BCL2, ERN1, PMAIP1, and TWIST1—through integrative bioinformatics and multi-omics analysis. It elucidated their regulatory networks in the pathogenesis of pituitary adenomas (PA), their roles in metabolic reprogramming of PA via pathways such as oxidative phosphorylation and ribosomal functions, and their dynamic interaction mechanisms with immunological microenvironment components including activated dendritic cells and T follicular helper cells. Additionally, drug prediction screening identified 36 potential therapeutic agents such as valproic acid and cyclosporine, providing new insights for precision medicine in PA. The findings fill the knowledge gap in ISR-driven molecular mechanisms in PA, construct a theoretical framework for the crosstalk between stress responses and the tumor microenvironment, and lay a foundation for ISR-targeted personalized therapy. Abbreviations Legends Abbreviation Full Name ISR Integrated Stress Response PA Pituitary Adenoma ISR-RGs Integrated Stress Response-related Genes GEO Gene Expression Omnibus DEGs Differentially Expressed Genes GO Gene Ontology KEGG Kyoto Encyclopedia of Genes and Genomes PPI Protein-Protein Interaction STRING Search Tool for the Retrieval of Interacting Genes LASSO Least Absolute Shrinkage and Selection Operator TFs Transcription Factors GSEA Gene Set Enrichment Analysis ssGSEA Single-sample Gene Set Enrichment Analysis CTD Comparative Toxicogenomics Database HPA Human Protein Atlas HL test Hosmer-Lemeshow Test DCA Decision Curve Analysis EMT Epithelial-Mesenchymal Transition UPR Unfolded Protein Response VPA Valproic Acid HDAC Histone Deacetylase TF-miRNA-mRNA Transcription Factor-MicroRNA-Messenger RNA Regulatory Network lncRNA Long Non-coding RNA miRNA MicroRNA rms package R package for nomogram construction (Citation: PMID: 30686944) rmda package R package for decision curve plotting (Citation: PMID: 38855330) GSVA package R package for gene set variation analysis (Citation: PMID: 23285768) psych package R package for statistical analysis (Citation: PMID: 28321695) glmnet package R package for LASSO algorithm (Citation: PMID: 20809303) Boruta package R package for feature selection (Citation: PMID: 33799563) clusterProfiler package R package for enrichment analysis (Citation: PMID: 22475548) ComplexHeatmap package R package for heatmap plotting (Citation: PMID: 27098862) VennDiagram package R package for Venn diagram plotting (Citation: PMID: 21544600) RCircos package R package for chromosome localization plotting (Citation: PMID: 25335086) enrichplot package R package for visualization of enrichment results (Citation: PMID: 30511128) Cytoscape Software for network analysis (Version: v 3.8.2) JASPAR Transcription Factor Binding Site Database (https://jaspar.elixir.no/) TarBase miRNA Target Gene Database (v 9.0, https://dianalab.e-ce.uth.gr/tarbasev9) StarBase lncRNA-miRNA Interaction Database (https://rnasysu.com/encori/) Ensembl Genome Database (https://useast.ensembl.org/) Protein database Protein Sequence Database (https://ncbi.nlm.nih.gov/protein/) mRNALocater Subcellular Localization Prediction Database (http://bio-bigdata.cn/mRNALocater/) GeneMANIA Gene Interaction Network Prediction Database (https://genemania.org/) MSigDB Molecular Signatures Database (http://software.broadinstitute.org/gsea/msigdb) Knosp classification Classification system for evaluating PA invasiveness (based on internal carotid artery relationship) NK cell Natural Killer Cell Treg Regulatory T Cell DCs Dendritic Cells TP53 Tumor Protein 53 STAT3 Signal Transducer and Activator of Transcription 3 Declarations Acknowledgments and Funding We would like to express our thanks to the Neurosurgery of The First Affiliated Hospital of Guangdong Pharmaceutical University for the technical assistance. This work is supported by the Guangdong Provincial Engineering and Technology Research Center of Stem Cell Therapy for Pituitary Disease. Footnote Conflicts of Interest: The authors declare that they have no conflicts of interest. Limitations: Based on bioinformatics approaches, this study systematically identified key biomarkers associated with the integrated stress response (ISR) in pituitary adenomas and established a robust diagnostic prediction model, providing novel insights and valuable references for research on the molecular mechanisms and clinical diagnosis of pituitary adenomas. However, several limitations exist. Firstly, the study relies on transcriptomic data from public databases, lacking experimental validation to confirm the specific biological functions of BCL2 , ERN1 , PMAIP1 , and TWIST1 ; secondly, the sample size is limited and the hormone-secreting subtypes of pituitary adenomas were not differentiated, which may compromise the generalizability of the results. Future studies should validate the diagnostic value of these biomarkers in larger-scale, multicenter cohorts, further elucidate their specific mechanisms of action in the ISR pathway and tumor microenvironment through functional experiments, and explore their potential as therapeutic targets. Data Availability Statement: The transcriptomic datasets used in this study were all derived from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/), specifically the datasets with accession numbers GSE26966 (platform GPL570) and GSE63357 (platform GPL570). The 529 integrated stress response (ISR)-related genes employed in the study were obtained from a published article (DOI: 10.1016/j.celrep.2024.114236), and detailed information is available in Supplementary Table S1. For additional supplementary data supporting the conclusions of this study, please contact the corresponding author through established institutional communication channels. Ethical Statement: The information used in our research does not involve any direct interaction with human subjects, nor does it allow for the identification of individual participants. According to the ethical guidelines and regulations, when research is based solely on publicly accessible, anonymized data, ethical approval and informed consent from participants are not required. Therefore, this study is exempt from the ethical review process, and ethical approval is not applicable. References Melmed S. Pituitary-Tumor Endocrinopathies. N Engl J Med 2020;382:937–50. https://doi.org/10.1056/NEJMra1810772. Kasuki L, Raverot G. Definition and diagnosis of aggressive pituitary tumors. Rev Endocr Metab Disord 2020;21:203–8. https://doi.org/10.1007/s11154-019-09531-x. Hashmi FA, Shamim MS. Pituitary Adenoma: A review of existing classification systems based on anatomic extension and invasion. J Pak Med Assoc 2020;70:368–70. Fang Y, Pei Z, Chen H, Wang R, Feng M, Wei L, et al. Diagnostic value of Knosp grade and modified Knosp grade for cavernous sinus invasion in pituitary adenomas: a systematic review and meta-analysis. Pituitary 2021;24:457–64. https://doi.org/10.1007/s11102-020-01122-3. Khan DZ, Hanrahan JG, Baldeweg SE, Dorward NL, Stoyanov D, Marcus HJ. Current and Future Advances in Surgical Therapy for Pituitary Adenoma. Endocr Rev 2023;44:947–59. https://doi.org/10.1210/endrev/bnad014. Dai C, Liang S, Sun B, Kang J. The Progress of Immunotherapy in Refractory Pituitary Adenomas and Pituitary Carcinomas. Front Endocrinol (Lausanne) 2020;11:608422. https://doi.org/10.3389/fendo.2020.608422. Wang AT, Mullan RJ, Lane MA, Hazem A, Prasad C, Gathaiya NW, et al. Treatment of hyperprolactinemia: a systematic review and meta-analysis. Syst Rev 2012;1:33. https://doi.org/10.1186/2046-4053-1-33. Zamanipoor Najafabadi AH, Zandbergen IM, De Vries F, Broersen LHA, Van Den Akker-van Marle ME, Pereira AM, et al. Surgery as a Viable Alternative First-Line Treatment for Prolactinoma Patients. A Systematic Review and Meta-Analysis. The Journal of Clinical Endocrinology & Metabolism 2020;105:e32–41. https://doi.org/10.1210/clinem/dgz144. Ježková J, Hána V, Kosák M, Kršek M, Liščák R, Vymazal J, et al. Role of gamma knife radiosurgery in the treatment of prolactinomas. Pituitary 2019;22:411–21. https://doi.org/10.1007/s11102-019-00971-x. Costa-Mattioli M, Walter P. The integrated stress response: From mechanism to disease. Science 2020;368:eaat5314. https://doi.org/10.1126/science.aat5314. Tian X, Zhang S, Zhou L, Seyhan AA, Hernandez Borrero L, Zhang Y, et al. Targeting the Integrated Stress Response in Cancer Therapy. Front Pharmacol 2021;12:747837. https://doi.org/10.3389/fphar.2021.747837. Lu H, Koju N, Sheng R. Mammalian integrated stress responses in stressed organelles and their functions. Acta Pharmacol Sin 2024;45:1095–114. https://doi.org/10.1038/s41401-023-01225-0. Pakos‐Zebrucka K, Koryga I, Mnich K, Ljujic M, Samali A, Gorman AM. The integrated stress response. EMBO Reports 2016;17:1374–95. https://doi.org/10.15252/embr.201642195. Vanselow S, Neumann-Arnold L, Wojciech-Moock F, Seufert W. Stepwise assembly of the eukaryotic translation initiation factor 2 complex. J Biol Chem 2022;298:101583. https://doi.org/10.1016/j.jbc.2022.101583. Mellado W, Willis DE. Stressing out translation. Science 2021;373:1089–90. https://doi.org/10.1126/science.abk3261. Ghaddar N, Wang S, Woodvine B, Krishnamoorthy J, van Hoef V, Darini C, et al. The integrated stress response is tumorigenic and constitutes a therapeutic liability in KRAS-driven lung cancer. Nat Commun 2021;12:4651. https://doi.org/10.1038/s41467-021-24661-0. Licari E, Sánchez-Del-Campo L, Falletta P. The two faces of the Integrated Stress Response in cancer progression and therapeutic strategies. Int J Biochem Cell Biol 2021;139:106059. https://doi.org/10.1016/j.biocel.2021.106059. Xue YH, Ge YQ. Construction of lncRNA regulatory networks reveal the key lncRNAs associated with Pituitary adenomas progression. Math Biosci Eng 2020;17:2138–49. https://doi.org/10.3934/mbe.2020113. Peng H, Deng Y, Wang L, Cheng Y, Xu Y, Liao J, et al. Identification of Potential Biomarkers with Diagnostic Value in Pituitary Adenomas Using Prediction Analysis for Microarrays Method. J Mol Neurosci 2019;69:399–410. https://doi.org/10.1007/s12031-019-01369-x. Lior C, Barki D, Halperin C, Iacobuzio-Donahue CA, Kelsen D, Shouval RS-. Mapping the tumor stress network reveals dynamic shifts in the stromal oxidative stress response. Cell Rep 2024;43:114236. https://doi.org/10.1016/j.celrep.2024.114236. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43:e47. https://doi.org/10.1093/nar/gkv007. Gustavsson EK, Zhang D, Reynolds RH, Garcia-Ruiz S, Ryten M. ggtranscript: an R package for the visualization and interpretation of transcript isoforms using ggplot2. Bioinformatics 2022;38:3844–6. https://doi.org/10.1093/bioinformatics/btac409. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 2016;32:2847–9. https://doi.org/10.1093/bioinformatics/btw313. Zheng Y, Gao W, Zhang Q, Cheng X, Liu Y, Qi Z, et al. Ferroptosis and Autophagy-Related Genes in the Pathogenesis of Ischemic Cardiomyopathy. Front Cardiovasc Med 2022;9:906753. https://doi.org/10.3389/fcvm.2022.906753. Yu G, Wang L-G, Han Y, He Q-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 2012;16:284–7. https://doi.org/10.1089/omi.2011.0118. Liu P, Xu H, Shi Y, Deng L, Chen X. Potential Molecular Mechanisms of Plantain in the Treatment of Gout and Hyperuricemia Based on Network Pharmacology. Evid Based Complement Alternat Med 2020;2020:3023127. https://doi.org/10.1155/2020/3023127. Chen H, Boutros PC. VennDiagram: a package for the generation of highly-customizable Venn and Euler diagrams in R. BMC Bioinformatics 2011;12:35. https://doi.org/10.1186/1471-2105-12-35. Friedman J, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw 2010;33:1–22. Younis H, Anwar MW, Khan MUG, Sikandar A, Bajwa UI. A New Sequential Forward Feature Selection (SFFS) Algorithm for Mining Best Topological and Biological Features to Predict Protein Complexes from Protein-Protein Interaction Networks (PPINs). Interdiscip Sci 2021;13:371–88. https://doi.org/10.1007/s12539-021-00433-8. An J, Lai J, Sajjanhar A, Batra J, Wang C, Nelson CC. J-Circos: an interactive Circos plotter. Bioinformatics 2015;31:1463–5. https://doi.org/10.1093/bioinformatics/btu842. Correction to Lancet Psych 2022; 9: 938. Lancet Psychiatry 2023;10:e10. https://doi.org/10.1016/S2215-0366(22)00411-4. Wang L, Wang D, Yang L, Zeng X, Zhang Q, Liu G, et al. Cuproptosis related genes associated with Jab1 shapes tumor microenvironment and pharmacological profile in nasopharyngeal carcinoma. Front Immunol 2022;13:989286. https://doi.org/10.3389/fimmu.2022.989286. Sparks R, Lau WW, Liu C, Han KL, Vrindten KL, Sun G, et al. Influenza vaccination reveals sex dimorphic imprints of prior mild COVID-19. Nature 2023;614:752–61. https://doi.org/10.1038/s41586-022-05670-5. Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics 2013;14:7. https://doi.org/10.1186/1471-2105-14-7. Han C, Lin S, Lu X, Xue L, Wu ZB. Tumor-Associated Macrophages: New Horizons for Pituitary Adenoma Researches. Front Endocrinol 2021;12:785050. https://doi.org/10.3389/fendo.2021.785050. Daly AF, Beckers A. The Epidemiology of Pituitary Adenomas. Endocrinol Metab Clin North Am 2020;49:347–55. https://doi.org/10.1016/j.ecl.2020.04.002. Oruçkaptan HH, Senmevsim Ö, Özcan OE, Özgen T. Pituitary adenomas: results of 684 surgically treated patients and review of the literature. Surgical Neurology 2000;53:211–9. https://doi.org/10.1016/S0090-3019(00)00171-3. Scheithauer BW, Kovacs KT, Laws ER, Randall RV. Pathology of invasive pituitary tumors with special reference to functional classification. Journal of Neurosurgery 1986;65:733–44. https://doi.org/10.3171/jns.1986.65.6.0733. Chen Y, Wang CD, Su ZP, Chen YX, Cai L, Zhuge QC, et al. Natural History of Postoperative Nonfunctioning Pituitary Adenomas: A Systematic Review and Meta-Analysis. Neuroendocrinology 2012;96:333–42. https://doi.org/10.1159/000339823. Radha G, Raghavan SC. BCL2: A promising cancer therapeutic target. Biochimica et Biophysica Acta (BBA) - Reviews on Cancer 2017;1868:309–14. https://doi.org/10.1016/j.bbcan.2017.06.004. Shi C, Ye Z, Han J, Ye X, Lu W, Ji C, et al. BRD4 as a therapeutic target for nonfunctioning and growth hormone pituitary adenoma. Neuro-Oncology 2020;22:1114–25. https://doi.org/10.1093/neuonc/noaa084. Asuzu DT, Alvarez R, Fletcher PA, Mandal D, Johnson K, Wu W, et al. Pituitary adenomas evade apoptosis via noxa deregulation in Cushing’s disease. Cell Reports 2022;40:111223. https://doi.org/10.1016/j.celrep.2022.111223. Walter P, Ron D. The Unfolded Protein Response: From Stress Pathway to Homeostatic Regulation. Science 2011;334:1081–6. https://doi.org/10.1126/science.1209038. Iurlaro R, Muñoz‐Pinedo C. Cell death induced by endoplasmic reticulum stress. The FEBS Journal 2016;283:2640–52. https://doi.org/10.1111/febs.13598. Ron D, Walter P. Signal integration in the endoplasmic reticulum unfolded protein response. Nat Rev Mol Cell Biol 2007;8:519–29. https://doi.org/10.1038/nrm2199. Wang C, Bai M, Wang X, Tan C, Zhang D, Chang L, et al. Estrogen receptor antagonist fulvestrant inhibits proliferation and promotes apoptosis of prolactinoma cells by regulating the IRE1/XBP1 signaling pathway. Mol Med Report 2018. https://doi.org/10.3892/mmr.2018.9379. Michaelis KA, Knox AJ, Xu M, Kiseljak-Vassiliades K, Edwards MG, Geraci M, et al. Identification of Growth Arrest and DNA-Damage-Inducible Gene β (GADD45β) as a Novel Tumor Suppressor in Pituitary Gonadotrope Tumors. Endocrinology 2011;152:3603–13. https://doi.org/10.1210/en.2011-0109. Yuan L, Li P, Li J, Peng J, Zhouwen J, Ma S, et al. Identification and gene expression profiling of human gonadotrophic pituitary adenoma stem cells. Acta Neuropathol Commun 2023;11:24. https://doi.org/10.1186/s40478-023-01517-w. Wang Q, Mora-Jensen H, Weniger MA, Perez-Galan P, Wolford C, Hai T, et al. ERAD inhibitors integrate ER stress with an epigenetic mechanism to activate BH3-only protein NOXA in cancer cells. Proc Natl Acad Sci USA 2009;106:2200–5. https://doi.org/10.1073/pnas.0807611106. Guikema JE, Amiot M, Eldering E. Exploiting the pro-apoptotic function of NOXA as a therapeutic modality in cancer. Expert Opinion on Therapeutic Targets 2017;21:767–79. https://doi.org/10.1080/14728222.2017.1349754. Khan MdA, Chen H, Zhang D, Fu J. Twist: a molecular target in cancer therapeutics. Tumor Biol 2013;34:2497–506. https://doi.org/10.1007/s13277-013-1002-x. Elias MC, Tozer KR, Silber JR, Mikheeva S, Deng M, Morrison RS, et al. TWIST is Expressed in Human Gliomas, Promotes Invasion. Neoplasia 2005;7:824–37. https://doi.org/10.1593/neo.04352. Mikheeva SA, Mikheev AM, Petit A, Beyer R, Oxford RG, Khorasani L, et al. TWIST1 promotes invasion through mesenchymal change in human glioblastoma. Mol Cancer 2010;9:194. https://doi.org/10.1186/1476-4598-9-194. Cosset E, Hamdan G, Jeanpierre S, Voeltzel T, Sagorny K, Hayette S, et al. Deregulation of TWIST-1 in the CD34+ compartment represents a novel prognostic factor in chronic myeloid leukemia. Blood 2011;117:1673–6. https://doi.org/10.1182/blood-2009-11-254680. Ansieau S, Morel A-P, Hinkal G, Bastid J, Puisieux A. TWISTing an embryonic transcription factor into an oncoprotein. Oncogene 2010;29:3173–84. https://doi.org/10.1038/onc.2010.92. Luo Y, Chen J, Liu M, Chen S, Su X, Su J, et al. Twist1 promotes dendritic cell-mediated antitumor immunity. Experimental Cell Research 2020;392:112003. https://doi.org/10.1016/j.yexcr.2020.112003. Jia W, Zhu J, Martin TA, Jiang A, Sanders AJ, Jiang WG. Epithelial-mesenchymal Transition (EMT) Markers in Human Pituitary Adenomas Indicate a Clinical Course. Anticancer Res 2015;35:2635–43. Zhang F, Zhang Q, Zhu J, Yao B, Ma C, Qiao N, et al. Integrated proteogenomic characterization across major histological types of pituitary neuroendocrine tumors. Cell Res 2022;32:1047–67. https://doi.org/10.1038/s41422-022-00736-5. López-Muñoz F, Baumeister AA, Hawkins MF, Alamo C. The role of serendipity in the discovery of the clinical effects of psychotropic drugs: beyond of the myth. Actas Esp Psiquiatr 2012;40:34–42. Meunier H, Carraz G, Neunier Y, Eymard P, Aimard M. [Pharmacodynamic properties of N-dipropylacetic acid]. Therapie 1963;18:435–8. Perucca E. Pharmacological and Therapeutic Properties of Valproate: A Summary After 35 Years of Clinical Experience. CNS Drugs 2002;16:695–714. https://doi.org/10.2165/00023210-200216100-00004. Blaheta R, Nau H, Michaelis M, Cinatl.Jr J. Valproate and Valproate-Analogues: Potent Tools to Fight Against Cancer. CMC 2002;9:1417–33. https://doi.org/10.2174/0929867023369763. Riva G, Cilibrasi C, Bazzoni R, Cadamuro M, Negroni C, Butta V, et al. Valproic Acid Inhibits Proliferation and Reduces Invasiveness in Glioma Stem Cells Through Wnt/β Catenin Signalling Activation. Genes 2018;9:522. https://doi.org/10.3390/genes9110522. Yang L, Chang Y, Cao P. CCR7 preservation via histone deacetylase inhibition promotes epithelial-mesenchymal transition of hepatocellular carcinoma cells. Experimental Cell Research 2018;371:231–7. https://doi.org/10.1016/j.yexcr.2018.08.015. Phiel CJ, Zhang F, Huang EY, Guenther MG, Lazar MA, Klein PS. Histone Deacetylase Is a Direct Target of Valproic Acid, a Potent Anticonvulsant, Mood Stabilizer, and Teratogen. Journal of Biological Chemistry 2001;276:36734–41. https://doi.org/10.1074/jbc.M101287200. Chateauvieux S, Morceau F, Dicato M, Diederich M. Molecular and Therapeutic Potential and Toxicity of Valproic Acid. Journal of Biomedicine and Biotechnology 2010;2010:1–18. https://doi.org/10.1155/2010/479364. Calabresi P, Galletti F, Rossi C, Sarchielli P, Cupini LM. Antiepileptic drugs in migraine: from clinical aspects to cellular mechanisms. Trends in Pharmacological Sciences 2007;28:188–95. https://doi.org/10.1016/j.tips.2007.02.005. Tariot PN, Loy R, Ryan JM, Porsteinsson A, Ismail S. Mood stabilizers in Alzheimer’s disease: symptomatic and neuroprotective rationales. Advanced Drug Delivery Reviews 2002;54:1567–77. https://doi.org/10.1016/S0169-409X(02)00153-9. Nielsen NM, Svanström H, Stenager E, Magyari M, Koch‐Henriksen N, Pasternak B, et al. The use of valproic acid and multiple sclerosis. Pharmacoepidemiology and Drug 2015;24:262–8. https://doi.org/10.1002/pds.3692. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 26 Dec, 2025 Reviews received at journal 18 Dec, 2025 Reviews received at journal 17 Dec, 2025 Reviews received at journal 11 Dec, 2025 Reviewers agreed at journal 08 Dec, 2025 Reviewers agreed at journal 05 Dec, 2025 Reviewers agreed at journal 03 Dec, 2025 Reviewers invited by journal 03 Dec, 2025 Editor invited by journal 01 Dec, 2025 Editor assigned by journal 25 Nov, 2025 Submission checks completed at journal 22 Nov, 2025 First submitted to journal 22 Nov, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8124800","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":555423397,"identity":"967b49dd-1bdc-48a3-8933-91bb589dc058","order_by":0,"name":"Yuan Sui","email":"","orcid":"","institution":"First Affiliated Hospital of Guangdong Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Yuan","middleName":"","lastName":"Sui","suffix":""},{"id":555423398,"identity":"04175523-f2fa-44b5-8758-3ddb27284fe6","order_by":1,"name":"Zhiqian Yang","email":"","orcid":"","institution":"Guangdong Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Zhiqian","middleName":"","lastName":"Yang","suffix":""},{"id":555423399,"identity":"e7bd0361-84c0-4062-abcd-ac8f1c738355","order_by":2,"name":"Wei Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2ElEQVRIiWNgGAWjYBACNvbmgw8+GEjY8TMzH3yQUFFDWAsfz7FkwxkVNsmS7W3JBg/OHCOsRU4iR02Y40wa44YzZ8wkH7YwE+EwnjNszIxth5kNbiSYVSQ2sDHwt3cnEPBL77HHhW2H+SRvJKTdSNwhwyBx5uwGAracSzeeCbSF70bCsRuJZ9gYDCRyCWiRyDGT5m07zNhwI7GtILGNmUgtPEDvTzhzmI2BOC1IgcwskXDmGA9Bv8i3w6OS/+PHHxU1cvztvfi1YAAe0pSPglEwCkbBKMAKAOrvTnegj52BAAAAAElFTkSuQmCC","orcid":"","institution":"First Affiliated Hospital of Guangdong Pharmaceutical University","correspondingAuthor":true,"prefix":"","firstName":"Wei","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2025-11-16 02:53:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8124800/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8124800/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97669926,"identity":"40c33f88-40b2-41fe-879d-6ee0631e5bed","added_by":"auto","created_at":"2025-12-08 09:29:24","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1864287,"visible":true,"origin":"","legend":"","description":"","filename":"BiomarkersassociatedwithintegratedstressresponseinPA2025.11.docx","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/515126cdab73106ce57f436c.docx"},{"id":97521288,"identity":"e2970e73-d5f7-4088-9ef5-f09c4555d399","added_by":"auto","created_at":"2025-12-05 11:17:23","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5971,"visible":true,"origin":"","legend":"","description":"","filename":"a7cdfc86d1fe4870b26a77d947449110.json","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/9f1448367a0a20ae4ecc17e9.json"},{"id":97521291,"identity":"06f98340-61c0-4a72-8463-c56a831132a6","added_by":"auto","created_at":"2025-12-05 11:17:23","extension":"eps","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":618455,"visible":true,"origin":"","legend":"","description":"","filename":"drawingimage1.eps","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/3a6ab737ca8f230d2b79d941.eps"},{"id":97671616,"identity":"4061c321-eeca-4052-9753-5447a13ac4b3","added_by":"auto","created_at":"2025-12-08 09:32:48","extension":"jpeg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":544247,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/40b5db1f5f02b8f8cf4cd704.jpeg"},{"id":97670993,"identity":"92b09fd8-1b90-47cd-bdfd-4314fec205bd","added_by":"auto","created_at":"2025-12-08 09:31:41","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":210616,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/569dfc47e763a7f750d8c44a.png"},{"id":97521298,"identity":"8b74456f-a00e-4bc2-a89f-ce84aa1f192b","added_by":"auto","created_at":"2025-12-05 11:17:23","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":146675,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/d61f08ca4f29fcb37d75fd11.jpeg"},{"id":97670481,"identity":"c0b02fa0-f270-4945-b047-50eb36e28bac","added_by":"auto","created_at":"2025-12-08 09:30:47","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":150216,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/eede873655bd534e03f54cb3.jpeg"},{"id":97521299,"identity":"1273b052-2479-4134-bc5f-a852aa6dc9de","added_by":"auto","created_at":"2025-12-05 11:17:23","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":141461,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/985dbdcab8cad0923e823c82.jpeg"},{"id":97521307,"identity":"d9ed85e6-369e-42cb-b8e4-e101c37d3cd7","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":152600,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/3c592054ff958cedc122b330.jpeg"},{"id":97521317,"identity":"c52d4341-ef6e-4784-8f98-68965a3b7ca8","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"jpeg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":137491,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/d193a8362019d0b92f824f44.jpeg"},{"id":97521300,"identity":"e3e673fc-8579-4369-9b9f-cc0d86186962","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"jpeg","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":297509,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/8294cb3e5226ca6f52b30b43.jpeg"},{"id":97672140,"identity":"fada30f8-d7c2-4495-b41f-e0d36ac85867","added_by":"auto","created_at":"2025-12-08 09:34:23","extension":"jpeg","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":134101,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/e4a929840976e067c678d8d1.jpeg"},{"id":97521311,"identity":"9be051f8-ccd2-4fa3-b970-aa400caf80fc","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"jpeg","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":371164,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/500a8205c9776f2a7ab743a7.jpeg"},{"id":97521313,"identity":"9601e904-0aab-4040-b05e-0c30895caf9d","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":109276,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/1088bb67b98ee943bae6d546.png"},{"id":97521308,"identity":"11e2be9d-cc23-49ea-afeb-f47b73e942e1","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":39416,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/e9ffbd0ca8ab09446c5dd12e.png"},{"id":97670774,"identity":"e77a20ac-3fd4-4827-a31f-a9254ea64a05","added_by":"auto","created_at":"2025-12-08 09:31:19","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":82329,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/5c1e631af6ca8ba3a2c4fcd7.png"},{"id":97671198,"identity":"0e03f49f-3899-4ac9-915c-ab99cfce25d4","added_by":"auto","created_at":"2025-12-08 09:32:07","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":137328,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/923c45f80006cf21ee6339a6.png"},{"id":97671038,"identity":"b5a8cf33-93fe-4193-8974-908922955dee","added_by":"auto","created_at":"2025-12-08 09:31:47","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":74375,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/b918891fc30e61c116d8ca37.png"},{"id":97521304,"identity":"d35a1413-aa28-4b71-a336-ad3421edc2fe","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":34579,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/1c3e9d793d8a835eb5991664.png"},{"id":97521305,"identity":"4f172c02-75eb-4e33-80ab-5d6dadb5c72b","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":81204,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/00b76fcb5c6e100047eaf548.png"},{"id":97521315,"identity":"ecfd4bbd-069b-4a4d-9dbd-30463243b515","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":74694,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/f5d85097673f0a3218eb6bb7.png"},{"id":97669830,"identity":"b715c129-5174-4bbc-8f02-b85d0e613cc2","added_by":"auto","created_at":"2025-12-08 09:29:04","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":80541,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/2aaed7b39b3aa204b9b69840.png"},{"id":97521314,"identity":"48a48bf5-86d8-4b77-88f7-8173ad083ed4","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":74372,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/af4af07e81246c1406c6a368.png"},{"id":97521286,"identity":"431cac04-9c46-476e-896b-2a5eb3bf6f82","added_by":"auto","created_at":"2025-12-05 11:17:23","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":146675,"visible":true,"origin":"","legend":"\u003cp\u003eScreening and functional enrichment analysis of differentially expressed genes in Integrated Stress Response-Related Genes(ISR-RGs) and Pituitary Adenomas:(a) Volcano plots of DEGs (b) Heatmap of DEGs (c) Veen diagram displaying candidate genes between DEGs and ISR-RGs (d) GO enrichment analysis of cadidate genes (e) KEGG enrichment analysis of cadidate genes\u003c/p\u003e","description":"","filename":"1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/e1a47a3ea44c07122fce9af1.jpeg"},{"id":97671662,"identity":"3ced5d6e-4680-446e-ac7f-909171cefa0e","added_by":"auto","created_at":"2025-12-08 09:32:53","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":150216,"visible":true,"origin":"","legend":"\u003cp\u003eProtein-protein interaction (PPI) network construction and hub gene screening of candidate genes: (a)PPI network of 91 candidate genes (b) Venn diagram for hub gene screening\u003c/p\u003e","description":"","filename":"2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/ede307dab833d1490b8035ce.jpeg"},{"id":97521289,"identity":"74b267d1-ff70-4201-bfc1-17bc66ea8d17","added_by":"auto","created_at":"2025-12-05 11:17:23","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":141461,"visible":true,"origin":"","legend":"\u003cp\u003eScreening and validation of biomarkers from 18 hub genes:(a) LASSO logistic coefficient penalty plot of 18 hub genes;(b) LASSO model summary coefficient distribution plot of 18 hub genes;(c) Brouta analysis of 18 hub genes;(d) Venn diagram of intersected genes from LASSO and Brouta analyses; (e) Box-plot of expression levels of 7 potential candidate biomarkers in the training set (GSE26966); (f) \u0026nbsp;Box-plot of expression levels of 7 potential candidate biomarkers in the validation set (GSE63357)\u003c/p\u003e","description":"","filename":"3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/da5c8df22285ac9cd08fdb0e.jpeg"},{"id":97669840,"identity":"081af084-0de7-478b-aac6-0ad91e64863a","added_by":"auto","created_at":"2025-12-08 09:29:05","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":137692,"visible":true,"origin":"","legend":"\u003cp\u003eEstablishment and evaluation of the PA diagnostic nomogram based on 4 biomarkers:(a) Nomogram for predicting PA diagnosis probability;(b) Calibration curve of the PA diagnostic model;(c) Decision curve analysis (DCA) of the PA diagnostic model\u003c/p\u003e","description":"","filename":"4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/80da9d508306904a16c421e8.jpeg"},{"id":97669898,"identity":"e6ad63fd-b85c-4b64-8c9c-03a6c711b86c","added_by":"auto","created_at":"2025-12-08 09:29:21","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":137491,"visible":true,"origin":"","legend":"\u003cp\u003eChromosomal localization and sub-cellular localization analysis of 4 biomarkers:(a) Chromosomal localization map of 4 biomarkers;(b) Sub-cellular localization analysis map of 4 biomarkers\u003c/p\u003e","description":"","filename":"5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/1339eccd90ffe3b90979cdbb.jpeg"},{"id":97521294,"identity":"3c328d65-171f-4c8b-81e7-a47d88151f31","added_by":"auto","created_at":"2025-12-05 11:17:23","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":173081,"visible":true,"origin":"","legend":"\u003cp\u003eRegulatory networks of PA biomarkers:(a) GeneMANIA-based gene-gene interaction (GGI) network of biomarkers;(b) lncRNA-miRNA-biomarker regulatory network;(c) TF-miRNA-mRNA regulatory network of biomarkers\u003c/p\u003e","description":"","filename":"6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/3d22456595212d5ec6a55258.jpeg"},{"id":97670578,"identity":"5ba53e22-bd64-4446-9992-2e87b6fbb103","added_by":"auto","created_at":"2025-12-08 09:30:58","extension":"jpeg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":218218,"visible":true,"origin":"","legend":"\u003cp\u003eGSEA pathway enrichment analysis of PA biomarkers involved in PA development:(a) \u0026nbsp;GSEA pathway enrichment results for TWIST1;(b) \u0026nbsp;GSEA pathway enrichment results for ERN1;(c) GSEA pathway enrichment results for PMAIP1;(d) GSEA pathway enrichment results for BCL2\u003c/p\u003e","description":"","filename":"7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/18bfd9db6dbb8d99e89fb148.jpeg"},{"id":97521302,"identity":"ee9bbd1e-e48c-49cb-95f7-f1e43790f897","added_by":"auto","created_at":"2025-12-05 11:17:24","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":371164,"visible":true,"origin":"","legend":"\u003cp\u003eAnalysis of differential immune cells and their correlation with core biomarkers in PA:(a) Heatmap of immune infiltrating cell enrichment scores between PA and control groups;(b) Differential analysis plot of immune cells between PA and control samples;(c) Correlation heatmap between differential immune cells;(d) Correlation analysis between differential immune cells and core biomarkers;(e)Expression plot of BCL2 in 18 immune cell types;(f) Expression plot of ERN1 in 18 immune cell types;(g) Expression plot of PMAIP1 in 18 immune cell types\u003c/p\u003e","description":"","filename":"8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/080eabb8a983955231393f9f.jpeg"},{"id":97521296,"identity":"fca1bcd1-75fd-4f6a-b78c-dd5247df8c0c","added_by":"auto","created_at":"2025-12-05 11:17:23","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":200029,"visible":true,"origin":"","legend":"\u003cp\u003eSmall molecule compound-core biomarkers interaction network for PA drug prediction\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/e53f44b33814e72e35ac33a3.png"},{"id":97677825,"identity":"00604547-5d8b-4899-890e-74663d4c87fd","added_by":"auto","created_at":"2025-12-08 09:54:43","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2955497,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8124800/v1/bc704d6e-f0d4-40d4-a7c2-4fb10f23d176.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Identification of biomarkers associated with integrated stress response in pituitary adenomas based on bioinformatics","fulltext":[{"header":"Highlight box","content":"\u003cp\u003e\u003cstrong\u003eKey findings\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eThis study identified four biomarkers for pituitary adenomas (PA), namely BCL2, ERN1, PMAIP1, and TWIST1.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eWhat is known and what is new?\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eIt is known that the integrated stress response (ISR) is involved in tumor development. However, the role of ISR - related genes in PA remains unclear.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eThis study, for the first time, utilized bioinformatics approaches to clarify the above - mentioned four biomarkers, revealing their regulatory networks, signaling pathways, and relationships with immune cells.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eWhat is the implication, and what should change now?\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eThe study provides important references for PA treatment and research. In the future, experimental verification of the functions of these biomarkers and the drug prediction results should be conducted to promote the clinical translation of precise PA classification and targeted therapy.\u003c/li\u003e\n\u003c/ul\u003e"},{"header":"1. Introduction ","content":"\u003cp\u003ePA is a common neuroendocrine tumor, a benign epithelial neoplasm originating from the anterior pituitary lobe. The incidence of PA ranks third among intracranial tumors, just behind glioma and meningioma[1]. The incidence of PA has been reported to account for 10-15% of all primary tumors in the central nervous system[1]. Approximately 10% of PAs exhibit active behavior, and among them, 0.2% develop metastases and are thus classified as pituitary carcinomas\u0026nbsp;[1,2]. Pituitary adenomas include various hormone-producing subtypes such as Prolactinomas, ACTH-secreting adenomas associated with Cushing\u0026apos;s syndrome, and neoplasms generating excessive growth hormone. This category also comprises adenomas that abnormally release TSH or gonadotrophin, along with rare mixed-type variants demonstrating combined hormonal secretion patterns. Each subtype exhibits distinct pathophysiological features requiring tailored diagnostic and therapeutic approaches[2].\u003c/p\u003e\n\u003cp\u003ePituitary adenomas (PAs) exhibit diverse classification criteria. Based on size, they are categorized as microadenomas (\u0026lt;1 cm), macroadenomas (1-4 cm), and giant adenomas (\u0026gt;4 cm). Classification based on radiological, biological characteristics, and invasiveness distinguishes functional from non-functional subtypes, as well as invasive from non-invasive subtypes. Invasiveness is commonly assessed using the Knosp classification system, which evaluates the tumor\u0026apos;s relationship to the internal carotid artery[3,4]. Due to the critical physiological functions of the pituitary gland and its proximity to neurovascular structures, PAs can lead to significant morbidity and mortality[5]. Therapeutic approaches include surgery (e.g., transsphenoidal surgery), radiotherapy, chemotherapy, immunotherapy, and molecularly targeted therapies[6]. However, current treatments have significant limitations: pharmacological therapies may cause adverse effects such as gastrointestinal reactions, while surgery and radiotherapy carry risks including hypopituitarism[7\u0026ndash;9]. Consequently, elucidating the molecular mechanisms underlying PAs and developing targeted therapies are of paramount importance.\u003c/p\u003e\n\u003cp\u003eIntegrated Stress Response (ISR) is an important stress-support pathway. By regulating the rate of protein synthesis, it is increasingly regarded as a determinant in tumorigenesis[10]. Four stress-sensing kinases \u0026mdash; EIF2AK1, EIF2AK2, EIF2AK3, and EIF2AK4 (Eukaryotic Translation Initiation Factor-2\u0026alpha; Kinases 1-4) \u0026mdash; mediate cellular responses to diverse stressors. These kinases converge on a shared molecular mechanism: phosphorylation of a conserved serine residue within the Eukaryotic Translation Initiation Factor 2 (eIF2) complex[11]. A key functional attribute of the ISR pathway is its ability to dynamically modulate intracellular levels of the Ternary Complex. The Ternary Complex is structurally defined by the heterotrimeric organization of eIF2, comprising three stoichiometrically equivalent subunits: \u0026alpha;, \u0026beta;, and \u0026gamma;. The functional state of the eIF2 heterotrimer is modulated by phosphorylation of its \u0026alpha; subunit. Within this complex, the \u0026gamma; subunit harbors a catalytic domain essential for GTPase-activating protein (GAP) activity and mediates the recruitment of the Ternary Complex to mRNA during translation initiation[12,13].\u003c/p\u003e\n\u003cp\u003eDysregulation of the eIF2\u0026gamma; subunit is linked to developmental and functional abnormalities in the hypothalamic-pituitary axis, hypopituitarism, and metabolic disorders such as pancreatic insufficiency and disrupted glucose homeostasis[14]. Studies have revealed the two-sided nature of the ISR, exhibiting both cytoprotective functions and apoptosis-inducing capacities. Targeted modulation of ISR signaling elements has emerged as a viable therapeutic approach in oncology treatment strategies[15,16]. Despite widespread activation of the Integrated Stress Response across various malignancies, the mechanistic role of ISR in organ-specific tumorigenesis and malignant evolution continues to pose significant research challenges[17].\u003c/p\u003e\n\u003cp\u003eThis study, based on public databases and ISR-related literature data, aims to identify ISR-related biomarkers in PA through bioinformatics methods such as the combination of the Least absolute shrinkage and selection operator (LASSO) algorithm and the Boruta algorithm. Employing immunological profiling, molecular interaction network architecture, and therapeutic compound screening, this investigation aims to elucidate groundbreaking perspectives for deciphering ISR-mediated pathways in PA pathogenesis and advancing patient-specific immunomodulatory interventions.\u003c/p\u003e"},{"header":"2 Materials and methods ","content":"\u003cp\u003e\u003cstrong\u003e2.1 \u003c/strong\u003e\u003cstrong\u003eData acquisition\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn this study, GSE26966 (GPL570 platform; containing 14 pituitary adenoma (PA) tissue samples and 9 normal pituitary tissue samples) and GSE63357 (GPL570 platform; containing 20 PA tissue samples and 5 normal pituitary tissue samples) were both acquired from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/) database, among which GSE26966 was the training set and GSE63357 was the testing set[18,19]. Additionally, totally 529 integrated stress response related genes (ISR-RGs) were found from public paper[20] (Table S1).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.2 Identification of candidate genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore the mechanism of PA development, the candidate genes were identified. The study firstly found the differentially expressed genes (DEGs) between PA and normal samples in GSE26966 using limma package [21](v 3.54.0) (p.adj \u0026lt; 0.05 and |log\u003csub\u003e2\u003c/sub\u003e fold change (FC)| \u0026gt; 1), and the results, including top five up-regulated and down-regulated DEGs in PA samples, were labeled in volcano diagram plotted by ggplot2 package[22] (v 3.4.1). Additionally, top ten up-regulated and down-regulated DEGs in PA samples were displayed in heatmap plotted by ComplexHeatmap package[23] (v 2.14.0). Subsequently, the DEGs that belonged to ISR-RGs were further picked out to participate in following analyses as candidate genes using ggvenn package[24] (v 0.1.9), whose results were also plotted as a Venn diagram.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.3 Functional analyses and construction of protein-protein interaction (PPI) network of candidate genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo further study the functions and pathways that candidate genes participated in, Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses (p \u0026lt; 0.05) were carried out using the clusterProfiler package[25] (v 4.2.2), in which top five significant terms in GO, comprising biological process (BP), cellular component (CC) and molecular function (MF), as well as KEGG were separately displayed using ggplot2 package (v 3.4.1). Moreover, all candidate genes were input into the Search Tool for the Retrieval of Interacting Genes (STRING, http://string-db.org) to establish the PPI network and analyze the protein interactions of candidate genes (confidence score \u0026ge; 0.4), whose results were plotted using Cytoscape software[26] (v 3.8.2). To further lock the crucial DEGs that played significant roles in PA development, the CytoHubba within Cytoscape (v 3.8.2) was firstly used to compute the scores for all candidate genes using maximal neighborhood component (MNC), maximal clique centrality (MCC), edge-betweenness path centrality (EPC), respectively. Then the hub genes were obtained through overlapping the top 20 genes with the highest score in each algorithm, and the results were displayed in a Venn diagram plotted by VennDiagram package[27] (v 1.7.1).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.4 Identification of biomarkers\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe crucial biomarkers played significant roles in PA development, and to lock crucial biomarkers, least absolute shrinkage and selection operator (LASSO) algorithm and Boruta analysis were separately conducted for all hub genes to perform this task. Firstly, LASSO characteristic genes were found from hub genes through 3-fold cross validation using glmnet package[28] (v 4.1.4), when the lambda approached the minimum. Secondly, Boruta characteristic genes were also obtained from all hub genes through Boruta analysis using Boruta package[29] (v 8.0.0) (pValue = 0.01, maxRuns = 50). The intersection genes of above two groups were then collected as potential candidate biomarkers using ggvenn package (v 0.1.9). Subsequently, expression validation for potential candidate biomarkers was successively carried out in GSE26966 and GSE63357 to find the biomarkers that existing consistent expression trends in two dataset using Wilcoxon test (p \u0026lt; 0.05).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.5 Chromosome localization analysis and subcellular localization of biomarkers\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo deeply explore the functions of biomarkers, the distributions of biomarkers in chromosomes were identified using Ensembl database (https://useast.ensembl.org/Homo_sapiens/Info/Index), and the results were plotted using RCircos package[30] (v 1.2.2). Additionally, the amino acid sequences of biomarkers were found from the Protein database (https://ncbi.nlm.nih.gov/protein/), and then the localizations of biomarkers in intracellular were predicted using mRNALocater database (http://bio-bigdata.cn/mRNALocater/result/), whose results were displayed using ggplot2 package (v 3.4.1).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.6 Construction and verification of line nomogram\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore whether biomarkers can predict the prevalence of PA, in all samples of GSE26966, the rms package (PMID: 30686944) (v 6.8-1) was used to construct a nomogram. Each biomarker was scored separately, with one score corresponding to each biomarker. The scores of all factors were summed up to obtain the total point, and then the incidence of PA was inferred according to the total point. To evaluate the predictive performance of the nomogram model, the calibration curve generated by the rms package (v 6.8-1) was used to intuitively show the relationship between the predicted probability values and the true probability values. The Hosmer-Lemeshow test (HL test) served as the model fitting index. Its principle lies in judging the dispersion between the predicted values and the true values. If the p-value is greater than 0.05, it indicates that the HL test is passed. The rmda package (PMID: 38855330) (v 1.6) was employed to draw the decision curve for the nomogram model. \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.7 Construction of \u003c/strong\u003e\u003cstrong\u003egene-gene-interaction (GGI) and \u003c/strong\u003e\u003cstrong\u003emolecular regulatory networks\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore the interactions that could help biomarkers complete their functions, the GGI network was predicted using GeneMANIA database (https://genemania.org/). Furthermore, the lncRNA-miRNA-biomarker network for biomarkers was constructed using the miRNAs predicted from the TarBase (v 9.0) database (https://dianalab.e-ce.uth.gr/tarbasev9) and long noncoding RNAs (lncRNAs) targeted to miRNAs predicted from the StarBase database (https://rnasysu.com/encori/) (clipExpNum\u0026gt;30). Additionally, transcription factors (TFs) that were targeting biomarkers were predicted from the JASPAR (https://jaspar.elixir.no/) to further perfect the TF-miRNA-mRNA network. The above networks were all plotted by Cytoscape (v 3.8.2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.8 Gene set enrichment analysis (GSEA)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMoreover, to further study the functions that biomarkers participated in to affect the development of PA, GSEA was performed. Firstly, the Spearman correlations between biomarkers and the remaining genes in GSE26966 were computed using psych package[31] (v 2.1.6), and then the remaining genes were ranked based on the correlations. Secondly, GSEA was performed based on the \u0026ldquo;c2.cp.kegg.v2023.1.Hs.symbols.gmt\u0026rdquo; from the Molecular Signatures Database (MSigDB, http://software.broadinstitute.org/gsea/msigdb) using the clusterProfiler package (v 4.2.2) (p.adj \u0026lt; 0.05, |normalized enrichment score| \u0026gt; 1), and the results were then visualized in ridge diagrams using enrichplot package[32] (v 1.18.3).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.9 Immune infiltration analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo investigate the functional dynamics of biological indicators within immunological niches, the study implemented a methodological framework employing single-sample Gene Set Enrichment Analysis (ssGSEA) to systematically quantify infiltration indices across 28 distinct immune cell populations[33] using GSVA package[34] (v 1.42.0). Then the enrichment score differences of each immune cell between PA and normal samples were identified to find differential immune cells using Wilcoxon test (p \u0026lt; 0.05), in which the results were displayed in box diagrams plotted by ggplot2 package (v 3.4.1). Additionally, Spearman correlation analysis (|correlation coefficient (cor)| \u0026gt; 0.3, p \u0026lt; 0.05) revealed the correlations between differential immune cells, as well as between differential immune cells and biomarkers using the cor function within psych package (v 2.1.6).\u003c/p\u003e\n\u003cp\u003eDifferent immune cells perform distinct functions. The expressions of biomarkers vary among different immune cells, which is likely to have an impact on the functions of immune cells. Therefore, the Human Protein Atlas (HPA) database (https://www.proteinatlas.org/) is utilized to analyze the expression profiles of biomarkers in eighteen different types of immune cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.10 Drug prediction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo provide more references for the treatment of PA, the Comparative Toxicogenomics Database (CTD; http://ctdbase.org/) was used to predict the potential drugs that interacted with biomarkers (interaction count \u0026gt; 5), whose results were shown in a drug-biomarker network plotted by Cytoscape (v 3.8.2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.11 \u003c/strong\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eR language (v 4.2.3) was leveraged in this study for bioinformatics analyses. The Wilcoxon test was used to perform statistical analysis, and a p-value less than 0.05 was considered significant.\u003c/p\u003e"},{"header":"3 Results","content":"\u003cp\u003e\u003cstrong\u003e3.1 Identification of 18 hub genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore the roles that ISR-RGs played in PA, totally 2,654 DEGs, comprising 1,058 up-regulated and 1,596 down-regulated genes in PA samples, were firstly identified from the GSE26966 (Fig.1a-b, Table S2). Subsequently, a sum of 91 ISR-RGs were further proved to exhibit significant differences in PA and normal samples (p.adj \u0026lt; 0.05), and then were renamed as candidate genes for further analyses (Fig.1c, Table S3). In the following analyses, multiple pathways were found that candidate genes might participate in, comprising 1,121 BP terms, 63 CC terms, 82 MF terms, as well as 92 terms from KEGG (Fig.1d-e, Table S4-5), which suggested all candidate genes might affect PA development through multiple response ways to oxygen, as well as indicated they might function in membrane structures and cause multiple complications, such as colorectal cancer and hepatitis B.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMoreover, PPI network indicated that there were 243 interactions between proteins of 75 candidate genes (Fig.2a, Table S6), in which BCL2 was validated to interact with BTG2 and BRCA1, and TWIST1 had interactions with ERN1 and EPHA2. Subsequently, through overlapping the top 20 genes with the highest scores of three algorithm, totally 18 genes were identified as hub genes to further screen biomarkers for PA (Fig.2b, Table S7).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.2 BCL2, ERN1, PMAIP1 and TWIST1 as biomarkers for PA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the following analyses, LASSO algorithm firstly locked seven hub genes, including BCL2, ATF3, ATF4, ERN1, PMAIP1, BRCA1 and TWIST1, when log(lambda.min) equaled -5.4173 (Fig.3a-b). Meanwhile, a total of 17 hub genes except MYC from 18 hub genes were assessed as confirmed through Boruta analysis (Fig.3c). Furthermore, the intersection genes of above two groups, including BCL2, ATF3, ATF4, ERN1, PMAIP1, BRCA1 and TWIST1, were renamed as potential candidate biomarkers (Fig.3d). According to the expression validation, BCL2, ERN1, PMAIP1 and TWIST1 were all further validated to have the significant lower expression in PA samples in GSE26966 and GSE63357 (p \u0026lt; 0.05), and then selected as biomarkers to explore the mechanism of PA development (Fig.3e-f).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.3 Establishment and assessment of a nomogram for PA diagnosis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on the biomarkers (BCL2, ERN1, PMAIP1, TWIST1), a diagnostic model for PA was established. The higher the total score, the greater the probability of having PA (Fig.4a). The slope of the calibration curve is close to 1, and with a p-value of 0.668, it indicates that the HL test is passed. In other words, there is no significant difference between the predicted values and the true values, suggesting that the model has a good degree of fitting and prediction accuracy. Decision Curve Analysis (DCA) is a method for evaluating clinical prediction models, diagnostic tests, and molecular markers (Fig.4b). The DCA results indicated that patients can benefit from the diagnostic model developed using the four biomarkers with a threshold probability ranging from 0 to 1 (Fig.4c). Overall, it shows that our diagnostic model has a strong ability to distinguish between PA patients and healthy controls.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.4\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eChromosome localization and sub-cellular localization of biomarkers\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAccording to chromosome localization, BCL2 and PMAIP1 were both validated to locate in chromosome 18, ERN1 was detected in chromosome 17, and TWIST1 was found in chromosome 7 (Fig.5a), which was helpful to further study the function of biomarkers and the genetic patterns of related diseases. Additionally, BCL2, ERN1, and TWIST1 were proved to be able to affect the development of PA through functioning in cytoplasm and nucleus, and PMAIP1 mainly worked in extracellular region (Fig.5b).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.5 Regulatory networks of biomarkers\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on GeneMANIA database, GGI network was established for all biomarkers, which suggested that PMAIP1 and BCL2 both could affect the development of PA through multiple pathways, such as regulating the membrane permeability and functioning in mitochondrial membrane, as well as ERN1 could interact with BBC3 and BID to participate in such pathways (Fig.6a). Additionally, the lncRNA-miRNA-biomarker network uncovered that AC004687.1 could regulate PMAIP1 and ERN1 through hsa-miR-142-5p, as well as BCL2 and TWIST1 could be regulated by AC021078.1 through hsa-miR-106a-5p (Fig.6b). Furthermore, more regulatory pathways of biomarkers were revealed by the TF-miRNA-mRNA network, in which STAT3 could regulate BCL2 through hsa-mir-7-1-3p, POU2F2 regulated PMAIP1 through hsa-let-7a-2-3p, as well as TP53 affected ERN1 through bonding to hsa-mir-9-3p and hsa-miR-545-5p (Fig.6c).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.6 Biomarkers affected PA development through multiple pathways\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAccording to the results of GSEA, TWIST1 was enriched in 42 pathways, ERN1 was proved to participate in 19 pathways, PMAIP1 was found in 18 pathways, and BCL2 was involved in 16 pathways, in which oxidative phosphorylation was a co-pathway for all biomarkers, as well as ERN1, PMAIP1, and BCL2 could all function in ribosome to affect PA development (Fig.7a-d, Table S8-11).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.7 Biomarkers affected PA development through multiple immune cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore the immune microenvironment that helped biomarkers affect PA development, totally 20 immune cells were validated to exhibit remarkable differences of enrichment scores between PA and normal samples (p \u0026lt;0.05) and renamed as differential immune cells, in which central memory CD4 T cell, immature dendritic cell and gamma delta (\u0026gamma;\u0026delta;) T cell had the highest enrichment scores in two groups (Fig.8a-b). Moreover, four differential immune cells, including CD56bright natural killer (NK) cell,\u0026nbsp;\u0026gamma;\u0026delta;\u0026nbsp;T cell, T follicular helper cell and type 2 T help cell, all existed higher enrichment scores in PA samples, which suggested that they might play significant roles in PA development. In subsequent analyses, multiple correlations were proved to exist between differential immune cells, in which the strongest positive correlation was detected between activated dendritic cell and central memory CD8 T cell (cor = 0.92, p \u0026lt; 0.05), as well as immature dendritic cell was validated to have the strongest negative correlation with type 2 T helper cell (cor = -0.73, p \u0026lt; 0.05) (Fig.8c). All these findings provided novel references for the study of immune response of PA. In addition to this, this study also revealed the correlations between differential immune cells and biomarkers. Interestingly, except CD56bright NK cell,\u0026nbsp;\u0026gamma;\u0026delta;\u0026nbsp;T cell, T follicular helper cell and type 2 T help cell, all remaining cells were all had strong positive correlations with all biomarkers, in which activated dendritic cell was proved to most strongly correlated with TWIST1 (cor = 0.88, p \u0026lt; 0.05) (Fig.8d).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAmong these immune cells, regulatory T cell exhibited the highest expression level of BCL2, closely followed by memory CD4 T cell and memory 8 cell. The expression levels in these cells were notably above 5 nTPM (Fig.8e). Basophils exhibited the highest expression level of ERN1, reaching approximately 18 nTPM, which was significantly higher than other cell types (Fig.8f). Basophils had the highest expression level of PMAIP1, reaching approximately 160 nTPM. Eosinophils followed with a relatively high expression, around 100 nTPM (Fig.8g). However, TWIST1 was not expressed in immune cells. These findings underscore the intricate interplay between immune cells and biomarkers in PA, unveiling potential therapeutic targets and shedding light on the complex immunopathogenesis of PA.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.8 Drug prediction for PA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAccording to CTD database, totally 36 drugs were predicted for BCL2, PMAIP1 and TWIST1, in which 33 drugs were found for BCL2, six drugs for PMAIP1, and only valproic acid was predicted for TWIST1 (Table S12). Referring to the drug-biomarkers network, sodium arsenite was predicted to interact with BCL2, cyclosporine was predicted to interact with PMAIP1, as well as valproic acid could interact with TWIST1, which all provided significant insights for the action mechanisms of BCL2, PMAIP1 and TWIST1, as well as furnished important references for optimizing the treatment strategies of PA (Fig.9).\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003e\u003cstrong\u003e4.1 Clinical and Biological Context of Pituitary Adenomas\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePituitary adenomas (PAs) are common central nervous system tumors with distinct clinical management paradigms[35\u0026ndash;37]. Although most PAs are benign, invasive subtypes frequently exhibit high recurrence rates due to incomplete resection[38,39], necessitating novel therapeutic targets. The integrated stress response (ISR), an evolutionarily conserved signaling pathway, is a key regulator of proteostasis and disease pathogenesis[10]. Dysregulation of the ISR has been implicated in diverse pathological processes, including neurodegenerative diseases, tumorigenesis, and metabolic disorders[10,13], positioning the ISR pathway as a potential therapeutic target.\u003c/p\u003e\n\u003cp\u003eBased on transcriptomic datasets from the Gene Expression Omnibus (GEO), this study identified four potential biomarkers\u0026mdash;BCL2, ERN1, PMAIP1, and TWIST1\u0026mdash;using bioinformatic methods. Further analyses, including Gene Set Enrichment Analysis (GSEA) and immune infiltration assessment, explored their biological functions and underlying regulatory mechanisms, offering new insights for the treatment of pituitary adenomas.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2 Mechanistic Insights into ISR-Related Biomarkers\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2.1 BCL2: Apoptosis Regulation and Immune Modulation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLocated at chromosome 18q21.33, BCL2(Apoptosis regulator Bcl-2) serves as a master regulator of apoptotic signaling through its mitochondrial and ER membrane localization[40]. In PA pathogenesis, BCL2 overexpression correlates with tumor invasiveness and therapeutic resistance[41]. Our bioinformatics analyses corroborate its central role in ISR networks, suggesting that ISR-mediated stabilization of BCL2 may create a pro-survival niche[42]. Notably, the observed high BCL2 expression in regulatory T cells (Tregs) implies dual functionality: direct tumor cell protection via apoptosis inhibition[40]\u0026nbsp;and indirect immune suppression through Treg-mediated IL-10/TGF-\u0026beta; secretion.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2.2 ERN1: ER Stress Sensor and Metabolic Adaptor\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAs the molecular transducer of the unfolded protein response (UPR), ERN1 (IRE1\u0026alpha;) coordinates cellular adaptation to ER stress through XBP1 splicing[43\u0026ndash;45]. In PAs, ERN1 activation promotes tumor cell survival under metabolic stress[46]. Our findings extend this understanding by revealing its significant enrichment in basophils, suggesting crosstalk between ER stress signaling and basophil-mediated inflammatory responses (e.g., IL-4/IL-13 release). This crosstalk may foster chronic inflammation and angiogenesis, creating a tumor-permissive microenvironment.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2.3 PMAIP1: Context-Dependent Apoptotic Regulator\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe BH3-only protein PMAIP1 (Phorbol-12-myristate-13-acetate-induced protein 1) demonstrates paradoxical roles in PA progression. While generally suppressed in gonadotroph tumors through TP53-dependent mechanisms[47], its marked upregulation in invasive subtypes[48]\u0026nbsp;suggests stage-specific functionality. Our multi-omics analyses position PMAIP1 as an ISR-ATF4 transcriptional target[49,50], potentially explaining its stress-inducible expression patterns. The observed basophil enrichment further implicates PMAIP1 in inflammatory microenvironment remodeling.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2.4 TWIST1: EMT Orchestrator and Immune Modulator\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTWIST1\u0026apos;s (Twist-related protein 1) functional dichotomy in solid tumors\u0026nbsp;[51\u0026ndash;55]\u0026nbsp;finds particular relevance in PAs. While absent in immune cells, its strong correlation with activated dendritic cells (DCs) reveals a novel immune-regulatory dimension. Experimental evidence suggests TWIST1 may drive DC dysfunction through antigen presentation impairment[56], while simultaneously promoting EMT-mediated tumor invasion[57,58]. This dual mechanism positions TWIST1 as a master regulator connecting tumor cell plasticity with immune evasion.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.3 Therapeutic Implications and Model Validation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.3.1 Dual-Targeting Therapeutic Strategy\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe ISR biomarkers\u0026apos; dual functionality - direct tumor promotion and immune microenvironment remodeling - necessitates combinatorial therapeutic approaches[10]. Our proposed strategy integrates ISR inhibitors (e.g., ISRIB) with immune checkpoint modulators (e.g., anti-PD-1 antibodies) to concurrently target both tumor cell-autonomous pathways and immune evasion mechanisms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.3.2 Diagnostic Model Performance and Limitations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWhile the BCL2/ERN1/PMAIP1/TWIST1-based nomogram demonstrates exceptional discriminative capacity (AUC=1), potential overfitting in the GSE26966 training cohort mandates external validation. Future model optimization should incorporate imaging parameters (e.g., Knosp grading) and serum biomarkers to enhance clinical applicability.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.3.3 Valproic Acid: A Repurposing Candidate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUsing a screening approach based on the Connectivity Map Database (CTD), valproic acid (VPA) has been identified as a potential therapeutic candidate associated with the transcription factor TWIST1. Initially discovered serendipitously as an antiepileptic drug[59], VPA\u0026apos;s core pharmacological actions include enhancing GABAergic neurotransmission[60], and it has been validated through decades of clinical use[61]. Notably, VPA exhibits complex context-dependent effects in oncology: on one hand, multiple studies indicate that VPA and its analogs may possess pro-cancer potential[62], with mechanisms involving activation of specific signaling pathways such as Wnt/\u0026beta;-catenin[63] or promotion of epithelial-mesenchymal transition (EMT)-related processes[64]; on the other hand, its property as a potent histone deacetylase (HDAC) inhibitor[65] forms a critical mechanistic basis for its antitumor activities[66]. This HDAC inhibitory activity is also considered central to its antiseizure effects[67], mood-stabilizing properties[68], and potential neuroprotective roles being explored for neurodegenerative diseases and multiple sclerosis[69]. However, its related mechanisms of action in pituitary adenomas warrant further investigation.\u003c/p\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003eThis study systematically identified four integrated stress response (ISR)-related biomarkers\u0026mdash;BCL2, ERN1, PMAIP1, and TWIST1\u0026mdash;through integrative bioinformatics and multi-omics analysis. It elucidated their regulatory networks in the pathogenesis of pituitary adenomas (PA), their roles in metabolic reprogramming of PA via pathways such as oxidative phosphorylation and ribosomal functions, and their dynamic interaction mechanisms with immunological microenvironment components including activated dendritic cells and T follicular helper cells. Additionally, drug prediction screening identified 36 potential therapeutic agents such as valproic acid and cyclosporine, providing new insights for precision medicine in PA. The findings fill the knowledge gap in ISR-driven molecular mechanisms in PA, construct a theoretical framework for the crosstalk between stress responses and the tumor microenvironment, and lay a foundation for ISR-targeted personalized therapy.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e\u003cstrong\u003eLegends\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eAbbreviation\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eFull Name\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eISR\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eIntegrated Stress Response\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003ePA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003ePituitary Adenoma\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eISR-RGs\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eIntegrated Stress Response-related Genes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eGEO\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eGene Expression Omnibus\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eDEGs\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eDifferentially Expressed Genes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eGO\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eGene Ontology\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eKEGG\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eKyoto Encyclopedia of Genes and Genomes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003ePPI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eProtein-Protein Interaction\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eSTRING\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eSearch Tool for the Retrieval of Interacting Genes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eLASSO\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eLeast Absolute Shrinkage and Selection Operator\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eTFs\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eTranscription Factors\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eGSEA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eGene Set Enrichment Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003essGSEA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eSingle-sample Gene Set Enrichment Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eCTD\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eComparative Toxicogenomics Database\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eHPA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eHuman Protein Atlas\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eHL test\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eHosmer-Lemeshow Test\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eDCA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eDecision Curve Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eEMT\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eEpithelial-Mesenchymal Transition\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eUPR\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eUnfolded Protein Response\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eVPA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eValproic Acid\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eHDAC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eHistone Deacetylase\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eTF-miRNA-mRNA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eTranscription Factor-MicroRNA-Messenger RNA Regulatory Network\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003elncRNA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eLong Non-coding RNA\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003emiRNA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eMicroRNA\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003erms package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for nomogram construction (Citation: PMID: 30686944)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003ermda package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for decision curve plotting (Citation: PMID: 38855330)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eGSVA package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for gene set variation analysis (Citation: PMID: 23285768)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003epsych package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for statistical analysis (Citation: PMID: 28321695)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eglmnet package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for LASSO algorithm (Citation: PMID: 20809303)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eBoruta package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for feature selection (Citation: PMID: 33799563)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eclusterProfiler package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for enrichment analysis (Citation: PMID: 22475548)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eComplexHeatmap package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for heatmap plotting (Citation: PMID: 27098862)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eVennDiagram package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for Venn diagram plotting (Citation: PMID: 21544600)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eRCircos package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for chromosome localization plotting (Citation: PMID: 25335086)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eenrichplot package\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR package for visualization of enrichment results (Citation: PMID: 30511128)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eCytoscape\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eSoftware for network analysis (Version: v 3.8.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eJASPAR\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eTranscription Factor Binding Site Database (https://jaspar.elixir.no/)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eTarBase\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003emiRNA Target Gene Database (v 9.0, https://dianalab.e-ce.uth.gr/tarbasev9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eStarBase\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003elncRNA-miRNA Interaction Database (https://rnasysu.com/encori/)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eEnsembl\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eGenome Database (https://useast.ensembl.org/)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eProtein database\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eProtein Sequence Database (https://ncbi.nlm.nih.gov/protein/)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003emRNALocater\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eSubcellular Localization Prediction Database (http://bio-bigdata.cn/mRNALocater/)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eGeneMANIA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eGene Interaction Network Prediction Database (https://genemania.org/)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eMSigDB\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eMolecular Signatures Database (http://software.broadinstitute.org/gsea/msigdb)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eKnosp classification\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eClassification system for evaluating PA invasiveness (based on internal carotid artery relationship)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eNK cell\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eNatural Killer Cell\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eTreg\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eRegulatory T Cell\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eDCs\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eDendritic Cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eTP53\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eTumor Protein 53\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eSTAT3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eSignal Transducer and Activator of Transcription 3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments and Funding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to express our thanks to the Neurosurgery of The First Affiliated Hospital of Guangdong Pharmaceutical University for the technical assistance. This work is supported by the Guangdong Provincial Engineering and Technology Research Center of Stem Cell Therapy for Pituitary Disease.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFootnote\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConflicts of Interest: The authors declare that they have no conflicts of interest.\u003c/p\u003e\n\u003cp\u003eLimitations:\u0026nbsp;Based on bioinformatics approaches, this study systematically identified key biomarkers associated with the integrated stress response (ISR) in pituitary adenomas and established a robust diagnostic prediction model, providing novel insights and valuable references for research on the molecular mechanisms and clinical diagnosis of pituitary adenomas. However, several limitations exist. Firstly, the study relies on transcriptomic data from public databases, lacking experimental validation to confirm the specific biological functions of \u003cem\u003eBCL2\u003c/em\u003e, \u003cem\u003eERN1\u003c/em\u003e, \u003cem\u003ePMAIP1\u003c/em\u003e, and \u003cem\u003eTWIST1\u003c/em\u003e; secondly, the sample size is limited and the hormone-secreting subtypes of pituitary adenomas were not differentiated, which may compromise the generalizability of the results. Future studies should validate the diagnostic value of these biomarkers in larger-scale, multicenter cohorts, further elucidate their specific mechanisms of action in the ISR pathway and tumor microenvironment through functional experiments, and explore their potential as therapeutic targets.\u003c/p\u003e\n\u003cp\u003eData Availability Statement: The transcriptomic datasets used in this study were all derived from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/), specifically the datasets with accession numbers GSE26966 (platform GPL570) and GSE63357 (platform GPL570). The 529 integrated stress response (ISR)-related genes employed in the study were obtained from a published article (DOI: 10.1016/j.celrep.2024.114236), and detailed information is available in Supplementary Table S1. For additional supplementary data supporting the conclusions of this study, please contact the corresponding author through established institutional communication channels.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eEthical Statement: The information used in our research does not involve any direct interaction with human subjects, nor does it allow for the identification of individual participants. According to the ethical guidelines and regulations, when research is based solely on publicly accessible, anonymized data, ethical approval and informed consent from participants are not required. Therefore, this study is exempt from the ethical review process, and ethical approval is not applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eMelmed S. Pituitary-Tumor Endocrinopathies. N Engl J Med 2020;382:937\u0026ndash;50. https://doi.org/10.1056/NEJMra1810772.\u003c/li\u003e\n\u003cli\u003eKasuki L, Raverot G. Definition and diagnosis of aggressive pituitary tumors. Rev Endocr Metab Disord 2020;21:203\u0026ndash;8. https://doi.org/10.1007/s11154-019-09531-x.\u003c/li\u003e\n\u003cli\u003eHashmi FA, Shamim MS. Pituitary Adenoma: A review of existing classification systems based on anatomic extension and invasion. J Pak Med Assoc 2020;70:368\u0026ndash;70.\u003c/li\u003e\n\u003cli\u003eFang Y, Pei Z, Chen H, Wang R, Feng M, Wei L, et al. Diagnostic value of Knosp grade and modified Knosp grade for cavernous sinus invasion in pituitary adenomas: a systematic review and meta-analysis. Pituitary 2021;24:457\u0026ndash;64. https://doi.org/10.1007/s11102-020-01122-3.\u003c/li\u003e\n\u003cli\u003eKhan DZ, Hanrahan JG, Baldeweg SE, Dorward NL, Stoyanov D, Marcus HJ. Current and Future Advances in Surgical Therapy for Pituitary Adenoma. Endocr Rev 2023;44:947\u0026ndash;59. https://doi.org/10.1210/endrev/bnad014.\u003c/li\u003e\n\u003cli\u003eDai C, Liang S, Sun B, Kang J. The Progress of Immunotherapy in Refractory Pituitary Adenomas and Pituitary Carcinomas. Front Endocrinol (Lausanne) 2020;11:608422. https://doi.org/10.3389/fendo.2020.608422.\u003c/li\u003e\n\u003cli\u003eWang AT, Mullan RJ, Lane MA, Hazem A, Prasad C, Gathaiya NW, et al. Treatment of hyperprolactinemia: a systematic review and meta-analysis. Syst Rev 2012;1:33. https://doi.org/10.1186/2046-4053-1-33.\u003c/li\u003e\n\u003cli\u003eZamanipoor Najafabadi AH, Zandbergen IM, De Vries F, Broersen LHA, Van Den Akker-van Marle ME, Pereira AM, et al. Surgery as a Viable Alternative First-Line Treatment for Prolactinoma Patients. A Systematic Review and Meta-Analysis. The Journal of Clinical Endocrinology \u0026amp; Metabolism 2020;105:e32\u0026ndash;41. https://doi.org/10.1210/clinem/dgz144.\u003c/li\u003e\n\u003cli\u003eJežkov\u0026aacute; J, H\u0026aacute;na V, Kos\u0026aacute;k M, Kr\u0026scaron;ek M, Li\u0026scaron;č\u0026aacute;k R, Vymazal J, et al. Role of gamma knife radiosurgery in the treatment of prolactinomas. Pituitary 2019;22:411\u0026ndash;21. https://doi.org/10.1007/s11102-019-00971-x.\u003c/li\u003e\n\u003cli\u003eCosta-Mattioli M, Walter P. The integrated stress response: From mechanism to disease. Science 2020;368:eaat5314. https://doi.org/10.1126/science.aat5314.\u003c/li\u003e\n\u003cli\u003eTian X, Zhang S, Zhou L, Seyhan AA, Hernandez Borrero L, Zhang Y, et al. Targeting the Integrated Stress Response in Cancer Therapy. Front Pharmacol 2021;12:747837. https://doi.org/10.3389/fphar.2021.747837.\u003c/li\u003e\n\u003cli\u003eLu H, Koju N, Sheng R. Mammalian integrated stress responses in stressed organelles and their functions. Acta Pharmacol Sin 2024;45:1095\u0026ndash;114. https://doi.org/10.1038/s41401-023-01225-0.\u003c/li\u003e\n\u003cli\u003ePakos‐Zebrucka K, Koryga I, Mnich K, Ljujic M, Samali A, Gorman AM. The integrated stress response. EMBO Reports 2016;17:1374\u0026ndash;95. https://doi.org/10.15252/embr.201642195.\u003c/li\u003e\n\u003cli\u003eVanselow S, Neumann-Arnold L, Wojciech-Moock F, Seufert W. Stepwise assembly of the eukaryotic translation initiation factor 2 complex. J Biol Chem 2022;298:101583. https://doi.org/10.1016/j.jbc.2022.101583.\u003c/li\u003e\n\u003cli\u003eMellado W, Willis DE. Stressing out translation. Science 2021;373:1089\u0026ndash;90. https://doi.org/10.1126/science.abk3261.\u003c/li\u003e\n\u003cli\u003eGhaddar N, Wang S, Woodvine B, Krishnamoorthy J, van Hoef V, Darini C, et al. The integrated stress response is tumorigenic and constitutes a therapeutic liability in KRAS-driven lung cancer. Nat Commun 2021;12:4651. https://doi.org/10.1038/s41467-021-24661-0.\u003c/li\u003e\n\u003cli\u003eLicari E, S\u0026aacute;nchez-Del-Campo L, Falletta P. The two faces of the Integrated Stress Response in cancer progression and therapeutic strategies. Int J Biochem Cell Biol 2021;139:106059. https://doi.org/10.1016/j.biocel.2021.106059.\u003c/li\u003e\n\u003cli\u003eXue YH, Ge YQ. Construction of lncRNA regulatory networks reveal the key lncRNAs associated with Pituitary adenomas progression. Math Biosci Eng 2020;17:2138\u0026ndash;49. https://doi.org/10.3934/mbe.2020113.\u003c/li\u003e\n\u003cli\u003ePeng H, Deng Y, Wang L, Cheng Y, Xu Y, Liao J, et al. Identification of Potential Biomarkers with Diagnostic Value in Pituitary Adenomas Using Prediction Analysis for Microarrays Method. J Mol Neurosci 2019;69:399\u0026ndash;410. https://doi.org/10.1007/s12031-019-01369-x.\u003c/li\u003e\n\u003cli\u003eLior C, Barki D, Halperin C, Iacobuzio-Donahue CA, Kelsen D, Shouval RS-. Mapping the tumor stress network reveals dynamic shifts in the stromal oxidative stress response. Cell Rep 2024;43:114236. https://doi.org/10.1016/j.celrep.2024.114236.\u003c/li\u003e\n\u003cli\u003eRitchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43:e47. https://doi.org/10.1093/nar/gkv007.\u003c/li\u003e\n\u003cli\u003eGustavsson EK, Zhang D, Reynolds RH, Garcia-Ruiz S, Ryten M. ggtranscript: an R package for the visualization and interpretation of transcript isoforms using ggplot2. Bioinformatics 2022;38:3844\u0026ndash;6. https://doi.org/10.1093/bioinformatics/btac409.\u003c/li\u003e\n\u003cli\u003eGu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 2016;32:2847\u0026ndash;9. https://doi.org/10.1093/bioinformatics/btw313.\u003c/li\u003e\n\u003cli\u003eZheng Y, Gao W, Zhang Q, Cheng X, Liu Y, Qi Z, et al. Ferroptosis and Autophagy-Related Genes in the Pathogenesis of Ischemic Cardiomyopathy. Front Cardiovasc Med 2022;9:906753. https://doi.org/10.3389/fcvm.2022.906753.\u003c/li\u003e\n\u003cli\u003eYu G, Wang L-G, Han Y, He Q-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 2012;16:284\u0026ndash;7. https://doi.org/10.1089/omi.2011.0118.\u003c/li\u003e\n\u003cli\u003eLiu P, Xu H, Shi Y, Deng L, Chen X. Potential Molecular Mechanisms of Plantain in the Treatment of Gout and Hyperuricemia Based on Network Pharmacology. Evid Based Complement Alternat Med 2020;2020:3023127. https://doi.org/10.1155/2020/3023127.\u003c/li\u003e\n\u003cli\u003eChen H, Boutros PC. VennDiagram: a package for the generation of highly-customizable Venn and Euler diagrams in R. BMC Bioinformatics 2011;12:35. https://doi.org/10.1186/1471-2105-12-35.\u003c/li\u003e\n\u003cli\u003eFriedman J, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw 2010;33:1\u0026ndash;22.\u003c/li\u003e\n\u003cli\u003eYounis H, Anwar MW, Khan MUG, Sikandar A, Bajwa UI. A New Sequential Forward Feature Selection (SFFS) Algorithm for Mining Best Topological and Biological Features to Predict Protein Complexes from Protein-Protein Interaction Networks (PPINs). Interdiscip Sci 2021;13:371\u0026ndash;88. https://doi.org/10.1007/s12539-021-00433-8.\u003c/li\u003e\n\u003cli\u003eAn J, Lai J, Sajjanhar A, Batra J, Wang C, Nelson CC. J-Circos: an interactive Circos plotter. Bioinformatics 2015;31:1463\u0026ndash;5. https://doi.org/10.1093/bioinformatics/btu842.\u003c/li\u003e\n\u003cli\u003eCorrection to Lancet Psych 2022; 9: 938. Lancet Psychiatry 2023;10:e10. https://doi.org/10.1016/S2215-0366(22)00411-4.\u003c/li\u003e\n\u003cli\u003eWang L, Wang D, Yang L, Zeng X, Zhang Q, Liu G, et al. Cuproptosis related genes associated with Jab1 shapes tumor microenvironment and pharmacological profile in nasopharyngeal carcinoma. Front Immunol 2022;13:989286. https://doi.org/10.3389/fimmu.2022.989286.\u003c/li\u003e\n\u003cli\u003eSparks R, Lau WW, Liu C, Han KL, Vrindten KL, Sun G, et al. Influenza vaccination reveals sex dimorphic imprints of prior mild COVID-19. Nature 2023;614:752\u0026ndash;61. https://doi.org/10.1038/s41586-022-05670-5.\u003c/li\u003e\n\u003cli\u003eH\u0026auml;nzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics 2013;14:7. https://doi.org/10.1186/1471-2105-14-7.\u003c/li\u003e\n\u003cli\u003eHan C, Lin S, Lu X, Xue L, Wu ZB. Tumor-Associated Macrophages: New Horizons for Pituitary Adenoma Researches. Front Endocrinol 2021;12:785050. https://doi.org/10.3389/fendo.2021.785050.\u003c/li\u003e\n\u003cli\u003eDaly AF, Beckers A. The Epidemiology of Pituitary Adenomas. Endocrinol Metab Clin North Am 2020;49:347\u0026ndash;55. https://doi.org/10.1016/j.ecl.2020.04.002.\u003c/li\u003e\n\u003cli\u003eOru\u0026ccedil;kaptan HH, Senmevsim \u0026Ouml;, \u0026Ouml;zcan OE, \u0026Ouml;zgen T. Pituitary adenomas: results of 684 surgically treated patients and review of the literature. Surgical Neurology 2000;53:211\u0026ndash;9. https://doi.org/10.1016/S0090-3019(00)00171-3.\u003c/li\u003e\n\u003cli\u003eScheithauer BW, Kovacs KT, Laws ER, Randall RV. Pathology of invasive pituitary tumors with special reference to functional classification. Journal of Neurosurgery 1986;65:733\u0026ndash;44. https://doi.org/10.3171/jns.1986.65.6.0733.\u003c/li\u003e\n\u003cli\u003eChen Y, Wang CD, Su ZP, Chen YX, Cai L, Zhuge QC, et al. Natural History of Postoperative Nonfunctioning Pituitary Adenomas: A Systematic Review and Meta-Analysis. Neuroendocrinology 2012;96:333\u0026ndash;42. https://doi.org/10.1159/000339823.\u003c/li\u003e\n\u003cli\u003eRadha G, Raghavan SC. BCL2: A promising cancer therapeutic target. Biochimica et Biophysica Acta (BBA) - Reviews on Cancer 2017;1868:309\u0026ndash;14. https://doi.org/10.1016/j.bbcan.2017.06.004.\u003c/li\u003e\n\u003cli\u003eShi C, Ye Z, Han J, Ye X, Lu W, Ji C, et al. BRD4 as a therapeutic target for nonfunctioning and growth hormone pituitary adenoma. Neuro-Oncology 2020;22:1114\u0026ndash;25. https://doi.org/10.1093/neuonc/noaa084.\u003c/li\u003e\n\u003cli\u003eAsuzu DT, Alvarez R, Fletcher PA, Mandal D, Johnson K, Wu W, et al. Pituitary adenomas evade apoptosis via noxa deregulation in Cushing\u0026rsquo;s disease. Cell Reports 2022;40:111223. https://doi.org/10.1016/j.celrep.2022.111223.\u003c/li\u003e\n\u003cli\u003eWalter P, Ron D. The Unfolded Protein Response: From Stress Pathway to Homeostatic Regulation. Science 2011;334:1081\u0026ndash;6. https://doi.org/10.1126/science.1209038.\u003c/li\u003e\n\u003cli\u003eIurlaro R, Mu\u0026ntilde;oz‐Pinedo C. Cell death induced by endoplasmic reticulum stress. The FEBS Journal 2016;283:2640\u0026ndash;52. https://doi.org/10.1111/febs.13598.\u003c/li\u003e\n\u003cli\u003eRon D, Walter P. Signal integration in the endoplasmic reticulum unfolded protein response. Nat Rev Mol Cell Biol 2007;8:519\u0026ndash;29. https://doi.org/10.1038/nrm2199.\u003c/li\u003e\n\u003cli\u003eWang C, Bai M, Wang X, Tan C, Zhang D, Chang L, et al. Estrogen receptor antagonist fulvestrant inhibits proliferation and promotes apoptosis of prolactinoma cells by regulating the IRE1/XBP1 signaling pathway. Mol Med Report 2018. https://doi.org/10.3892/mmr.2018.9379.\u003c/li\u003e\n\u003cli\u003eMichaelis KA, Knox AJ, Xu M, Kiseljak-Vassiliades K, Edwards MG, Geraci M, et al. Identification of Growth Arrest and DNA-Damage-Inducible Gene \u0026beta; (GADD45\u0026beta;) as a Novel Tumor Suppressor in Pituitary Gonadotrope Tumors. Endocrinology 2011;152:3603\u0026ndash;13. https://doi.org/10.1210/en.2011-0109.\u003c/li\u003e\n\u003cli\u003eYuan L, Li P, Li J, Peng J, Zhouwen J, Ma S, et al. Identification and gene expression profiling of human gonadotrophic pituitary adenoma stem cells. Acta Neuropathol Commun 2023;11:24. https://doi.org/10.1186/s40478-023-01517-w.\u003c/li\u003e\n\u003cli\u003eWang Q, Mora-Jensen H, Weniger MA, Perez-Galan P, Wolford C, Hai T, et al. ERAD inhibitors integrate ER stress with an epigenetic mechanism to activate BH3-only protein NOXA in cancer cells. Proc Natl Acad Sci USA 2009;106:2200\u0026ndash;5. https://doi.org/10.1073/pnas.0807611106.\u003c/li\u003e\n\u003cli\u003eGuikema JE, Amiot M, Eldering E. Exploiting the pro-apoptotic function of NOXA as a therapeutic modality in cancer. Expert Opinion on Therapeutic Targets 2017;21:767\u0026ndash;79. https://doi.org/10.1080/14728222.2017.1349754.\u003c/li\u003e\n\u003cli\u003eKhan MdA, Chen H, Zhang D, Fu J. Twist: a molecular target in cancer therapeutics. Tumor Biol 2013;34:2497\u0026ndash;506. https://doi.org/10.1007/s13277-013-1002-x.\u003c/li\u003e\n\u003cli\u003eElias MC, Tozer KR, Silber JR, Mikheeva S, Deng M, Morrison RS, et al. TWIST is Expressed in Human Gliomas, Promotes Invasion. Neoplasia 2005;7:824\u0026ndash;37. https://doi.org/10.1593/neo.04352.\u003c/li\u003e\n\u003cli\u003eMikheeva SA, Mikheev AM, Petit A, Beyer R, Oxford RG, Khorasani L, et al. TWIST1 promotes invasion through mesenchymal change in human glioblastoma. Mol Cancer 2010;9:194. https://doi.org/10.1186/1476-4598-9-194.\u003c/li\u003e\n\u003cli\u003eCosset E, Hamdan G, Jeanpierre S, Voeltzel T, Sagorny K, Hayette S, et al. Deregulation of TWIST-1 in the CD34+ compartment represents a novel prognostic factor in chronic myeloid leukemia. Blood 2011;117:1673\u0026ndash;6. https://doi.org/10.1182/blood-2009-11-254680.\u003c/li\u003e\n\u003cli\u003eAnsieau S, Morel A-P, Hinkal G, Bastid J, Puisieux A. TWISTing an embryonic transcription factor into an oncoprotein. Oncogene 2010;29:3173\u0026ndash;84. https://doi.org/10.1038/onc.2010.92.\u003c/li\u003e\n\u003cli\u003eLuo Y, Chen J, Liu M, Chen S, Su X, Su J, et al. Twist1 promotes dendritic cell-mediated antitumor immunity. Experimental Cell Research 2020;392:112003. https://doi.org/10.1016/j.yexcr.2020.112003.\u003c/li\u003e\n\u003cli\u003eJia W, Zhu J, Martin TA, Jiang A, Sanders AJ, Jiang WG. Epithelial-mesenchymal Transition (EMT) Markers in Human Pituitary Adenomas Indicate a Clinical Course. Anticancer Res 2015;35:2635\u0026ndash;43.\u003c/li\u003e\n\u003cli\u003eZhang F, Zhang Q, Zhu J, Yao B, Ma C, Qiao N, et al. Integrated proteogenomic characterization across major histological types of pituitary neuroendocrine tumors. Cell Res 2022;32:1047\u0026ndash;67. https://doi.org/10.1038/s41422-022-00736-5.\u003c/li\u003e\n\u003cli\u003eL\u0026oacute;pez-Mu\u0026ntilde;oz F, Baumeister AA, Hawkins MF, Alamo C. The role of serendipity in the discovery of the clinical effects of psychotropic drugs: beyond of the myth. Actas Esp Psiquiatr 2012;40:34\u0026ndash;42.\u003c/li\u003e\n\u003cli\u003eMeunier H, Carraz G, Neunier Y, Eymard P, Aimard M. [Pharmacodynamic properties of N-dipropylacetic acid]. Therapie 1963;18:435\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003ePerucca E. Pharmacological and Therapeutic Properties of Valproate: A Summary After 35 Years of Clinical Experience. CNS Drugs 2002;16:695\u0026ndash;714. https://doi.org/10.2165/00023210-200216100-00004.\u003c/li\u003e\n\u003cli\u003eBlaheta R, Nau H, Michaelis M, Cinatl.Jr J. Valproate and Valproate-Analogues: Potent Tools to Fight Against Cancer. CMC 2002;9:1417\u0026ndash;33. https://doi.org/10.2174/0929867023369763.\u003c/li\u003e\n\u003cli\u003eRiva G, Cilibrasi C, Bazzoni R, Cadamuro M, Negroni C, Butta V, et al. Valproic Acid Inhibits Proliferation and Reduces Invasiveness in Glioma Stem Cells Through Wnt/\u0026beta; Catenin Signalling Activation. Genes 2018;9:522. https://doi.org/10.3390/genes9110522.\u003c/li\u003e\n\u003cli\u003eYang L, Chang Y, Cao P. CCR7 preservation via histone deacetylase inhibition promotes epithelial-mesenchymal transition of hepatocellular carcinoma cells. Experimental Cell Research 2018;371:231\u0026ndash;7. https://doi.org/10.1016/j.yexcr.2018.08.015.\u003c/li\u003e\n\u003cli\u003ePhiel CJ, Zhang F, Huang EY, Guenther MG, Lazar MA, Klein PS. Histone Deacetylase Is a Direct Target of Valproic Acid, a Potent Anticonvulsant, Mood Stabilizer, and Teratogen. Journal of Biological Chemistry 2001;276:36734\u0026ndash;41. https://doi.org/10.1074/jbc.M101287200.\u003c/li\u003e\n\u003cli\u003eChateauvieux S, Morceau F, Dicato M, Diederich M. Molecular and Therapeutic Potential and Toxicity of Valproic Acid. Journal of Biomedicine and Biotechnology 2010;2010:1\u0026ndash;18. https://doi.org/10.1155/2010/479364.\u003c/li\u003e\n\u003cli\u003eCalabresi P, Galletti F, Rossi C, Sarchielli P, Cupini LM. Antiepileptic drugs in migraine: from clinical aspects to cellular mechanisms. Trends in Pharmacological Sciences 2007;28:188\u0026ndash;95. https://doi.org/10.1016/j.tips.2007.02.005.\u003c/li\u003e\n\u003cli\u003eTariot PN, Loy R, Ryan JM, Porsteinsson A, Ismail S. Mood stabilizers in Alzheimer\u0026rsquo;s disease: symptomatic and neuroprotective rationales. Advanced Drug Delivery Reviews 2002;54:1567\u0026ndash;77. https://doi.org/10.1016/S0169-409X(02)00153-9.\u003c/li\u003e\n\u003cli\u003eNielsen NM, Svanstr\u0026ouml;m H, Stenager E, Magyari M, Koch‐Henriksen N, Pasternak B, et al. The use of valproic acid and multiple sclerosis. Pharmacoepidemiology and Drug 2015;24:262\u0026ndash;8. https://doi.org/10.1002/pds.3692.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"discover-oncology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"dion","sideBox":"Learn more about [Discover Oncology](https://www.springer.com/12672)","snPcode":"","submissionUrl":"","title":"Discover Oncology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Pituitary adenoma, Integrated stress response, Regulatory network, Biomarkers","lastPublishedDoi":"10.21203/rs.3.rs-8124800/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8124800/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eBackground: Integrated stress response (ISR) participates in the development of tumor, however, the roles of ISR-related genes (ISR-RGs) in pituitary adenoma (PA) are unclear. The goal of this study was to identify the biomarkers related to ISR-RGs in PA and explore their action mechanism.\u003c/p\u003e\n\u003cp\u003eMethods: In the present study, the public datasets from Gene Expression Omnibus database (GEO) database and ISR-RGs from literature were analyzed to obtain biomarkers for PA using differential expression analysis, machine learning algorithm, Boruta analysis and expression validation.\u003c/p\u003e\n\u003cp\u003eResults: The results demonstrated that BCL2, ERN1, PMAIP1 and TWIST1, which all possessed significantly lower expression in PA samples in training and testing sets, were selected as biomarkers for PA. Subsequently, the nomogram showed good performance in predicting PA risk based on these biomarkers. BCL2 and PMAIP1 were both validated to locate in Chromosome 18, and BCL2 mainly functioned in cytoplasm and PMAIP1 worked in extracellular region. Additionally, the regulatory networks revealed that biomarkers might work through interacting with BBC3, AC004687, hsa-miR-142-5p, TP53 and STAT3, which provided significant insights into mechanism of biomarkers. Strong correlations were found to exist between activated dendritic cells and TWIST1, as well as between PMAIP1 and T follicular helper cells, which contributed to a more comprehensive understanding of the mechanism. Finally, we totally obtained 36 potential drugs interacting with biomarkers, including sodium arsenite, cyclosporine and valproic acid.\u003c/p\u003e\n\u003cp\u003eConclusions: This study identified BCL2, ERN1, PMAIP1 and TWIST1 as biomarkers for PA and revealed their action mechanism, which provided important references for treatment and study of PA.\u003c/p\u003e","manuscriptTitle":"Identification of biomarkers associated with integrated stress response in pituitary adenomas based on bioinformatics","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-05 11:17:18","doi":"10.21203/rs.3.rs-8124800/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-12-26T05:15:35+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-18T08:01:16+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-17T15:57:46+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-11T05:02:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"286368289131481590196653396253504042214","date":"2025-12-08T07:07:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"65147200491513936331164212620908079666","date":"2025-12-05T12:43:19+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"340215833113731837110652500647114554823","date":"2025-12-03T12:32:42+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-12-03T12:12:14+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-12-01T07:23:11+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-11-25T10:56:36+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-11-22T14:23:57+00:00","index":"","fulltext":""},{"type":"submitted","content":"Discover Oncology","date":"2025-11-22T14:21:12+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"discover-oncology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"dion","sideBox":"Learn more about [Discover Oncology](https://www.springer.com/12672)","snPcode":"","submissionUrl":"","title":"Discover Oncology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"3405274a-0cac-42be-8889-cd9ce52efd40","owner":[],"postedDate":"December 5th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-02-16T17:24:46+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-05 11:17:18","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8124800","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8124800","identity":"rs-8124800","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}