Immune dysregulation and prognostic signatures associated with Epstein–Barr virus in acute lymphoblastic leukemia: an integrated transcriptomic and single-cell analysis

Immune dysregulation and prognostic signatures associated with Epstein–Barr virus in acute lymphoblastic leukemia: an integrated transcriptomic and single-cell analysis | 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 Immune dysregulation and prognostic signatures associated with Epstein–Barr virus in acute lymphoblastic leukemia: an integrated transcriptomic and single-cell analysis Lili Pan, Chunping Wu, Dezheng Ning, Yongzhi Zheng This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7725271/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Epstein–Barr virus (EBV) is associated with lymphoid malignancies; however, its contribution to acute lymphoblastic leukemia (ALL) is unclear. Therefore, we investigated the associations between EBV-related transcriptional activity and immune remodeling in ALL across publicly available cohorts. Methods This study integrated transcriptomics, weighted gene co-expression network analysis, protein–protein interaction network analyses, and single-cell RNA sequencing (scRNA-seq). Results We identified 401 EBV-related differentially expressed genes and constructed a prognostic model comprising the following nine critical immune-related genes: interleukin (IL)-18, Toll-like receptor 1, perforin 1 [ PRF1 ], IL-6 receptor, C-C motif chemokine ligand 2 [ CCL2 ], TTK protein kinase [ TTK ], CD19 molecule [ CD19 ], cathepsin S, and C-C motif chemokine ligand 4 [ CCL4 ]. Our model robustly stratified patients into high- and low-risk groups. The high-risk group exhibited significantly poorer survival than did the low-risk group ( P = 0.015). External validation confirmed the predictive accuracy of the model (area under the curve values: 0.82, 0.68, and 0.68 for 1-, 3-, and 5-year survival, respectively). scRNA-seq further revealed distinct expression patterns of the nine prognostic genes across immune cell subsets: TTK was enriched in B and T cells; PRF1 was predominantly expressed in T and natural killer cells; and CCL2, CCL4 , and CD19 were highly expressed in monocytes and B cells. Conclusions Associations between these genes, immune composition/proliferation signals, and survival in ALL were highlighted. These findings are hypothesis-generating and may reflect EBV-related transcriptional activity as well as lineage and immune-infiltration states; causal roles require validation in clinically EBV-characterized cohorts. acute lymphoblastic leukemia Epstein–Barr virus immune microenvironment Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 BACKGROUND Acute lymphoblastic leukemia (ALL) is a malignancy of lymphoid progenitors and the most common pediatric cancer (~ 75% of childhood leukemias); moreover, it comprises 20–30% of adult acute leukemias [ 1 , 2 ]. Despite therapeutic advances, relapse remains a major challenge, underscoring the need for deeper biological insights. Notably, infectious exposures have been implicated in leukemogenesis via persistent inflammation, genomic instability, and aberrant signaling [ 3 ]. Epstein–Barr virus (EBV), a ubiquitous oncogenic herpesvirus, establishes lifelong latency in memory B cells and is linked to multiple cancers, primarily via latent programs that modulate host pathways and enable immune evasion [ 4 – 7 ]. Evidence implicates EBV in the development and progression of Burkitt’s lymphoma, Hodgkin’s lymphoma, nasopharyngeal carcinoma, and gastric cancer. EBV-related activity has been associated with tumor microenvironment remodeling, including M2 macrophage enrichment, increased MMP9, T-cell exhaustion, and treatment resistance [ 6 , 7 ]. These observations indicate that EBV-related immunologic states may accompany, but do not by themselves prove, causal viral oncogenesis in every context. In ALL, the contribution of EBV remains uncertain and context-dependent. Pediatric data report higher EBV seroprevalence and LMP1 positivity compared to those of controls, alongside expansion of regulatory T cells and elevated IL-10, suggesting immune modulation that could intersect with leukemic biology [ 8 – 10 ]. These findings indicate that EBV-related immune modulation could intersect with leukemic biology, potentially shaping immune evasion and disease behavior, while not establishing direct causality. Moreover, chronic inflammation is a key contributor to chemotherapy resistance in various cancers [ 11 ], which may explain the poor treatment outcomes observed in such patients [ 12 ]. Therefore, we integrated publicly available bulk and single-cell transcriptomes to evaluate the associations between EBV-related transcriptional features, immune context, and prognosis in ALL. The study workflow is shown in Fig. 1 . METHODS Methods Gene expression microarray datasets associated with EBV infection (GSE45918 and GSE85599) and ALL (GSE26713 and GSE67684) were obtained from the GEO database. Consistent Affymetrix Probe Set IDs within the same sample type were selected, and batch effects among different datasets were corrected using the “ComBat” function from the “sva” package in R. Transcriptomic data and clinical information for the training cohort were retrieved from the TARGET database phases TARGET-ALL-P1 and TARGET-ALL-P2, comprising 679 samples obtained from patients with leukemia. The validation cohort, which included 1479 samples from patients with leukemia, was obtained from the MP2PRT-ALL database. scRNA-seq data were sourced from GSE196214, which involved 21 patients newly diagnosed with B/T-cell acute lymphoblastic leukemia. Differential expression analysis Differential expression analyses comparing EBV-infected samples and controls, as well as ALL samples and controls, were conducted using the “limma” package in R. Limma uses linear models combined with empirical Bayesian methods, providing a robust statistical inference particularly suitable for datasets with limited sample sizes. Differentially expressed genes (DEGs) were identified based on a threshold of P 1. Count data from the TARGET and MP2PRT cohorts were subjected to DEG analysis using the “edgeR” package, which applies a negative binomial distribution model. Data were normalized using the trimmed mean M-values. Finally, dispersion estimates were calculated for statistical testing to detect DEGs. In addition, we conducted weighted gene co-expression network analysis (WGCNA) on datasets from patients infected with EBV and control groups to identify gene modules significantly associated with EBV infection characteristics. By examining the relationships between modules and features and using hierarchical clustering to merge modules with high similarity, we further revealed the interconnectivity of the gene modules. Protein–protein interaction (PPI) network analysis A PPI network was constructed using the Search Tool for the Retrieval of Interacting Genes/Proteins database (http://string-db.org). Key subnetworks were extracted using the Molecular Complex Detection (MCODE) plugin within the Cytoscape software (version 3.10.2), resulting in two major modules comprising 157 genes, which were subsequently utilized in downstream analyses. To further investigate the correlation between EBV-associated ALL-related genes (EARGs) and the immune landscape in ALL samples, we used the Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data (ESTIMATE) algorithm to calculate immune scores, stromal scores, ESTIMATE scores, and tumor purity. Tumor immune infiltration Tumor purity and stromal cell composition were assessed using estimation algorithms. According to the median immune score, the samples were divided into two groups, followed by differential and survival analyses. Single-sample gene set enrichment analysis for 29 immune-related signatures was performed using the “GSVA” package (method = “ssgsea”), with gene sets represented via “GSEAbase.” Graphical representations were rendered using the “ggplot2” and “pheatmap” packages in R. Prognostic model construction Univariate Cox proportional hazards regression analysis was used to identify candidate prognostic genes ( P < 0.05). Redundant variables were eliminated using the least absolute shrinkage and selection operator (LASSO) regression. LASSO was implemented with the “glmnet” package, and the optimal penalty parameter (λ) was selected by cross-validation. A prognostic model was then developed using multivariate Cox proportional hazards regression. Patients were stratified into high- and low-risk groups based on median risk scores. Kaplan–Meier survival analysis with log-rank tests was used to evaluate survival differences between the groups. Time-dependent receiver operating characteristic (ROC) curves were generated using the “timeROC” package to assess 1-, 3-, and 5-year performance. The prognostic model was externally validated in the MP2PRT-ALL validation cohort. scRNA-seq analysis The top 2000 highly variable genes were identified using the variance-stabilizing transformation method implemented in the FindVariableFeatures function in the Seurat R package. The data were scaled using the ScaleData function, whereas dimensionality reduction was performed via principal component analysis (PCA) using the RunPCA function. Batch effects among the patient samples were corrected using the harmony algorithm (R package “harmony”). Clustering was performed based on the first 30 significant principal components using FindNeighbors and FindClusters (resolution = 0.25). Dimensionality reduction was visualized using t-distributed stochastic neighbor embedding (RunTSNE). Cell types were identified using SingleR, followed by manual verification, and visualized using the “scCustomize” package. RESULTS EBV-related gene expression and module identification To identify DEGs regulated by EBV infection, we analyzed two gene expression datasets: GSE45918 (EBV-infected samples) and GSE85599 (normal controls). The PCA plot demonstrated clear separation between EBV-infected and control samples (Fig. 2A and B), consistent with EBV-associated transcriptional remodeling but not proof of causality. The volcano plot revealed 401 DEGs, including 230 upregulated and 171 downregulated genes (Online Resource 1: Supplementary Table S1). Subsequently, we performed a WGCNA using the EBV-related gene expression dataset (Fig. 2C). By evaluating the scale-free topology fit index and mean connectivity under various soft-thresholding powers, we determined the optimal soft-threshold (Fig. 2D and E) to confirm the stability and robustness of the network. With the merging threshold set at cutHeight = 0.25, similar gene modules were merged, resulting in the identification of six distinct co-expressed modules (Online Resource 2: Supplementary Figure S1 A and B). Based on biological plausibility and module eigengene patterns, we focused on four modules (cyan, tan, pink, and green), comprising 3,122 genes in total (Fig. 2F; Online Resource 1: Supplementary Table S2). Notably, the 401 DEGs were distributed across the selected modules rather than present in every module. Functional annotations of genes within the selected modules indicated enrichment for DNA repair, immune signaling, and cell-cycle regulation, processes broadly implicated in leukemogenesis . Accordingly, we describe these signals as associational rather than as EBV-specific causal effects. EBV-associated ALL genes and functional modules Differential expression analysis between patients with ALL and healthy controls (GEO datasets GSE26713 and GSE67684) identified 882 DEGs (Online Resource 1: Supplementary Table S3, Fig. 3A). We then took the union of the EBV-related DEGs and ALL DEGs to generate an EBV-associated ALL gene set (EARGs; n = 1,175) (Fig. 3B), highlighting genes that co-occur across EBV-related and ALL comparisons without implying direct viral regulation in ALL. To further investigate putative interactions among EARGs, we constructed a PPI network in STRING and extracted two dense subnetworks (Clusters 1 and 2) using the MCODE plugin in Cytoscape (Fig. 3C–D). Functional enrichment analyses of these modules (GO/KEGG) revealed distinct pathway patterns (Fig. 3E–F). Cluster 1 was enriched for terms related to EBV infection, NF-κB signaling, and JAK–STAT signaling, which are consistent with immune/inflammatory signaling described in EBV-associated contexts and are also broadly implicated in leukemic biology [13–15]. In contrast, Cluster 2 encompassed a broader set of genes enriched for cell cycle, DNA replication, and p53 signaling, alongside GO terms such as nuclear division, chromosome segregation, and mitotic cell-cycle transition. These findings indicate prominent proliferation/cell-cycle programs typical of leukemogenesis. Collectively, the EARG PPI highlights two thematic axes—(i) immune/inflammatory signaling and (ii) cell-cycle/proliferation—associated with EBV-related gene sets in our integrated analysis. These signals may partly reflect immune composition and proliferative states inherent to ALL and should not be over-interpreted as evidence of EBV-specific causal mechanisms without direct virological confirmation. Immune landscape in ALL The results of the ESTIMATE algorithm are shown in Fig. 4A. Tumor immune infiltration analysis was performed on the TARGET-ALL dataset using prescreened immune/stromal-related gene sets. Using bulk gene-expression profiles as proxies, we estimated stromal and immune components within the leukemic microenvironment and stratified samples into high and low immune-score groups by the cohort median. As expected from the ESTIMATE framework, the high-score group exhibited lower inferred tumor purity and higher stromal scores than did the low-score group (Fig. 4B–E; Online Resource 1: Supplementary Table S4). In Kaplan–Meier analysis, the high immune-score group exhibited worse overall survival than did the low-score group (Fig. 4F). These findings suggest that immune cells in the tumor microenvironment promote immune evasion in ALL and that high immune infiltration is associated with poor prognosis in ALL, consistent with findings in acute myeloid leukemia [16]. Prognostic model Differential expression analysis between the high and low immune score groups identified 2031 DEGs (Fig. 4G). The intersection of these DEGs from the TARGET-ALL dataset with the EARGs produced 49 candidate genes (Fig. 5A), which, combined with clinical information, were analyzed using univariate Cox regression analysis using the survival package. All 49 genes significantly affected the overall survival (OS) of patients with ALL. Subsequently, LASSO shrinkage and selection operator regression were performed using the glmnet package, resulting in the identification of 17 feature genes. These 17 genes were further analyzed using multivariate Cox regression, leading to the identification of nine genes that were significantly associated with prognosis (Fig. 5B), including interleukin 18 ( IL18 ), Toll-like receptor 1 ( TLR1 ), perforin 1 ( PRF1 ), interleukin 6 receptor ( IL6R ), C-C motif chemokine ligand 2 ( CCL2 ), TTK protein kinase ( TTK ), CD19 molecule ( CD19 ), cathepsin S ( CTSS ), and C-C motif chemokine ligand 4 ( CCL4 ). These genes corroborated emerging evidence suggesting the critical roles of immune-related genes such as IL18 in modulating the immune microenvironment and influencing the response to cancer therapy [17]. Subsequently, a prognostic risk score index using the immune-related gene pair index (IRGPI) was constructed based on the expression of these nine genes (Fig. 5C, Online Resource 1: Supplementary Table S5) using the following equation: Riskscore = 0.128 × IL18 - 0.286 × TLR1 - 0.112 × PRF1 - 0.282 × IL6R + 0.118 × CCL2 - 0.133 × TTK + 0.240 × CD19 + 0.416 × CTSS + 0.173 × CCL4 . Using this model, risk scores were calculated for the clinical samples. After generating ROC curves, the corresponding areas under the curves (AUCs) for 1-, 3-, and 5-year survival were 0.72, 0.84, and 0.85, respectively (Fig. 5D), indicating the good discriminatory power of the prognostic model in predicting patient outcomes, further validating its potential clinical utility. Based on the median risk score as a cutoff, patients with high-risk scores exhibited significantly worse OS than did those with low-risk scores ( P < 0.001; Fig. 5E). Stratification of patients into high- and low-risk groups based on this model underscored its prognostic value in guiding treatment decisions. The risk score distribution, survival status, and feature gene heatmaps for the high- and low-risk score groups (Fig. 5F and G) showed higher CCL2, CCL4 , and CD19 expression and lower TTK and PRF1 expression in the former than in the latter. Thus, elevated CCL2 and CCL4 expression may enhance the inflammatory microenvironment in ALL to promote immune evasion and disease progression. Additionally, the lower TTK and PRF1 expression may contribute to impaired immune surveillance and uncontrolled cell proliferation, which are critical for leukemia development. Validation using the external independent dataset MP2PRT-ALL confirmed the findings from the TARGET dataset, in which patients in the high-risk group showed significantly poorer OS compared with those in the low-risk group (Online Resource 2: Supplementary Figure S2A). These results strengthened the external applicability and robustness of our prognostic model, ensuring its generalizability across different patient populations. Time-dependent ROC curve analysis further supported the robustness of the prognostic model, with AUC values of 0.82, 0.68, and 0.68 for 1-, 3-, and 5-year survival, respectively (Online Resource 2: Supplementary Figure S2B and C). These values provide additional evidence for the utility of the model for predicting long-term survival outcomes in patients with ALL. Gene co-expression network and functional enrichment To explore the potential biological functions of the prognostic genes, we used the GeneMANIA database (http://www.genemania.org/) to identify functionally associated genes and construct a gene co-expression network. Overall, 89.9% of the connections in the network were attributed to co-expression relationships, highlighting the interconnected nature of the identified genes, suggesting that these genes are functionally related and that their collective activity may contribute to the molecular mechanisms underlying ALL progression. Functional enrichment analysis indicated that this network was strongly associated with immune cell activity, highlighting the immune-related nature of the model (Online Resource 2: Supplementary Figure S3). Single-cell profiling of gene expression in ALL cellular subtypes We further validated the prognostic model based on the nine identified genes by assessing its predictive performance and conducting a survival analysis. To explore the biological relevance of the prognostic genes in the cellular context of ALL, we applied scRNA-seq to identify the spatial expression patterns of these genes within the ALL immune landscape. We identified 13 distinct clusters representing a diverse array of cellular subtypes (Fig. 6A–C). Subsequently, we analyzed these clusters to gain insight into their immune cell-type composition and potential functional roles in the progression of ALL. Cell markers for each of the 13 clusters were identified using the FindAllMarkers function (logFC = 0.5, MinPct = 0.25), which allowed for precise characterization of the cellular subtypes within the tumor microenvironment (Online Resource 1: Supplementary Table S6). To explore the cellular distribution of the prognostic genes, we examined their expression profiles across the identified cell types (Fig. 6D), demonstrating that TTK is predominantly expressed in B and T cells, thereby suggesting its involvement in modulating immune cell proliferation and differentiation in ALL. PRF1 , a key immune effector gene, was highly expressed in T and NK cells, indicating its role in mediating cytotoxic responses in the tumor microenvironment. IL18, IL6R, CTSS , and TLR1 were evenly distributed across various cell types, suggesting their broader involvement in immune modulation and cell signaling within the ALL microenvironment. Moreover, CD19 and CCL4 were predominantly expressed in B cells, consistent with their established roles in B cell activation and migration, which are central to leukemia pathogenesis. Conversely, CCL2 exhibited higher expression in monocytes, highlighting its potential role in their recruitment and immune suppression within the tumor microenvironment (Fig. 6D). Collectively, these findings demonstrated the utility of scRNA-seq for providing a detailed map of gene expression at the single-cell level, thereby offering novel insights into the role of prognostic genes in cellular subtypes within the ALL microenvironment. The spatial distribution of these genes indicates that immune cell interactions are key to ALL progression and treatment resistance. DISCUSSION The integrated findings from WGCNA and PPI network analyses indicated that EBV infection perturbs gene expression modules closely associated with DNA repair, cell cycle regulation, and immune modulation. Enrichment of NF-κB and JAK–STAT pathways plus immune/stromal signatures aligns with inflammatory and immune-evasion states reported in EBV-associated contexts and leukemic biology; however, it does not alone establish EBV-specific causality [ 18 ]. These patterns may also co-vary with lineage composition and proliferative burden that our datasets cannot fully disentangle. Overall, the data support a hypothesis-generating model in which EBV-related activity may coincide with immune remodeling and cell-cycle programs in subsets of ALL, requiring validation in clinically EBV-characterized cohorts and functional studies. LMP1 can mimic constitutively active CD40, recruiting TRAF adaptors and activating NF-κB and JNK to support B-cell proliferation and survival; similarly, LMP1 has been reported to dampen p53-mediated apoptosis via A20 induction and upregulation of BCL-2 family members [ 19 – 21 ]. In addition, MYC can transactivate the EBV receptor CR2/CD21, potentially enhancing viral entry in MYC-dysregulated B cells [ 22 ]. These mechanisms are well described in EBV-positive B-cell contexts and help explain lymphoproliferation. Here, enrichment of NF-κB/JAK–STAT and cell-cycle programs aligns with pathways implicated in EBV-associated biology and leukemic proliferation. Nevertheless, the present data do not establish EBV-specific causality in ALL. Our analysis demonstrated that patients with high immune scores exhibited worse overall survival (OS) than did those with low scores (P = 0.015), corroborating previous findings that immune-microenvironment features correlate with ALL prognosis [ 23 ]. Because ESTIMATE-derived immune scores in hematologic malignancies largely reflect cell-mixture proportions and disease burden, our results are associational rather than proof that immune infiltration per se drives outcome differences. Several EBV-associated malignancies show adverse prognosis and chemoresistance [ 24 ]; however, extrapolation to ALL is uncertain without uniform, sample-level EBV ascertainment in our cohorts. If clinically EBV-characterized ALL studies confirm similar relationships, EBV status together with immune-infiltration profiles could serve as exploratory adjuncts for risk stratification, immunomonitoring, and therapy planning. At present, these markers remain hypothesis-generating, pending prospective validation and functional evidence. Beyond validating the prognostic signature, our scRNA-seq analysis provided cell-type context for the signature genes—clarifying their localization across immune and proliferative compartments. The reduced TTK and PRF1 expression observed in the high-risk group may reflect impaired proliferative regulation of lymphoid cells and compromised cytotoxic immune responses, respectively. TTK , a dual-specificity protein kinase essential for mitotic progression, has been implicated in maintaining chromosomal stability during lymphocyte proliferation; hence, its downregulation may lead to aberrant cell cycle control and genomic instability [ 25 ]. Meanwhile, decreased expression of PRF1 , a key effector molecule in T and NK cell-mediated cytotoxicity, suggests impaired antitumor immune defense, consistent with previous studies linking PRF1 deficiency to tumor immune evasion [ 26 ]. In contrast, elevated CCL2, CCL4 , and CD19 expression indicated an inflammatory yet immunosuppressive microenvironment favorable for leukemia progression. CCL2 recruits myeloid-derived suppressor cells and regulatory monocytes, both of which contribute to immune evasion and tumor support [ 27 ]. Similarly, high CCL4 expression can enhance leukemic cell migration, disrupt tumor microenvironment stability, and participate in carcinogenesis and tumor progression [ 28 , 29 ]. CD19—a canonical B-lineage marker—likely reflects lineage composition/disease burden in bulk profiles, capturing expansion of malignant B-cell populations rather than acting as a direct driver [ 30 ]. Collectively, these signals should be viewed as associational; whether CCL2/CCL4-centered axes directly mediate immune suppression in ALL—particularly in clinically EBV-characterized subsets—requires targeted validation. Future studies should explore the critical mechanisms by which EBV contributes to the tumorigenic process in ALL. The IRGPI model, constructed based on EARGs in this study, provides preliminary evidence linking EBV-related gene signatures to ALL prognosis. Nonetheless, current evidence on the role of EBV in promoting ALL through the regulation of DNA repair, immune evasion, and cell cycle dysregulation is primarily derived from in vitro studies, small-scale cohorts, and indirect observations, with a notable lack of large-scale clinical validation. Therefore, future investigations should integrate multicenter cohort studies, animal model experiments, and single-cell multi-omics approaches to systematically elucidate the specific molecular mechanisms underlying EBV infection in ALL progression and characterize the dynamic changes within the immune microenvironment during disease development. Our integrative analysis links EBV-related transcriptional features with immune context and prognosis in ALL across public cohorts and, with scRNA-seq localization, provides a cell-type–aware framework to interpret a parsimonious prognostic signature. Key limitations include mixed B-ALL/T-ALL cohorts, lack of uniform, sample-level EBV ascertainment, bulk-level composition confounding, and an observational design that precludes causal inference. Future work should use clinically EBV-characterized, lineage-stratified cohorts and composition-adjusted models to test signature generalizability and prospectively validate clinical utility. CONCLUSIONS EBV-related transcriptional activity is associated with immune/proliferation programs and risk in ALL. A 9-gene signature stratified prognosis; however, without uniform EBV status and with mixed lineages, the results remain hypothesis-generating. Prospective, EBV-characterized cohorts and functional studies are needed to establish clinical utility and causality. Declarations Declaration of AI and AI-assisted technologies We used ChatGPT (OpenAI) to assist in improving the clarity and grammar of the manuscript's English language during the writing and revision stages. The AI tool was not used for generating scientific content, designing the study, data analysis, or creating citations. All scientific interpretations, data analyses, and conclusions were solely developed by the authors. Acknowledgements The authors gratefully acknowledge the public availability of the GEO datasets and thank the researchers who generated and shared these valuable resources. Author contributions Conceptualization and Design of the Analysis: Pan L, Wu C; Data Collection: Pan L, Ning D; Provision of Data or Analysis Tools: Ning D; Data Analysis: Pan L, Ning D; Manuscript Writing: Pan L, Wu C, Ning D, Zheng Y. Funding This work was supported by the National Key Clinical Specialty Discipline Construction Program (2021-76); Fujian Provincial Clinical Research Center for Hematological Malignancies (2020Y2006); Mechanistic Study of CTHRC1-Mediated Regulation of Chemosensitivity in Acute T-Cell Lymphoblastic Leukemia Through the Hippo Signaling Pathway (2022QH1031). Data availability The datasets used in this study are publicly available in the Gene Expression Omnibus (GEO) database under the accession numbers GSE45918, GSE85599, GSE26713, GSE67684, and GSE13159. Additional transcriptomic and clinical data were obtained from the TARGET (TARGET-ALL-P1, TARGET-ALL-P2) and MP2PRT-ALL databases. All data supporting the conclusions of this article are available from the corresponding author upon reasonable request. Compliance with Ethical Standards Disclosure of potential conflicts of interest The authors declare that there is no conflict of interest regarding the publication of this paper. Research involving Human Participants and/or Animals This study did not involve any prospective enrollment of human participants or the collection of new human specimens. All analyses were performed on publicly available, de-identified transcriptomic datasets obtained from the Gene Expression Omnibus (GEO: GSE45918, GSE85599, GSE26713, GSE67684, GSE13159) and from publicly accessible consortia (TARGET-ALL-P1, TARGET-ALL-P2, MP2PRT-ALL, and GSE196214). Because these resources were generated under Institutional Review Board (IRB) approval and released with anonymized identifiers, no additional IRB review or written informed consent was required for the present secondary-data analysis, in accordance with the Declaration of Helsinki and the author guidelines of Cancer Research and Treatment. Informed consent Because these resources were generated under Institutional Review Board (IRB) approval and released with anonymized identifiers, no additional IRB review or written informed consent was required for the present secondary-data analysis, in accordance with the Declaration of Helsinki and the author guidelines of Cancer Research and Treatment. References Tran TH, Hunger SP (2022) The genomic landscape of pediatric acute lymphoblastic leukemia and precision medicine opportunities. Semin Cancer Biol 84:144–152. https://doi.org/10.1016/j.semcancer.2020.10.013 Terwilliger T, Abdul-Hay M (2017) Acute lymphoblastic leukemia: a comprehensive review and 2017 update. Blood Cancer J 7:e577. https://doi.org/10.1038/bcj.2017.53 Greaves M (2018) A causal mechanism for childhood acute lymphoblastic leukaemia. Nat Rev Cancer 18:471–484. https://doi.org/10.1038/s41568-018-0015-6 Damania B, Kenney SC, Raab-Traub N (2022) Epstein-Barr virus: biology and clinical disease. Cell 185:3652–3670. https://doi.org/10.1016/j.cell.2022.08.026 Okuno Y, Murata T, Sato Y, Muramatsu H, Ito Y, Watanabe T et al (2019) Defective Epstein-Barr virus in chronic active infection and haematological malignancy. Nat Microbiol 4:404–413. https://doi.org/10.1038/s41564-018-0334-0 Atri-Schuller A, Abushukair H, Cavalcante L, Hentzen S, Saeed A, Saeed A (2022) Tumor molecular and microenvironment characteristics in EBV-associated malignancies as potential therapeutic targets: focus on gastric cancer. Curr Issues Mol Biol 44:5756–5767. https://doi.org/10.3390/cimb44110390 Chen Y, Ouyang D, Wang Y, Pan Q, Zhao J, Chen H et al (2024) EBV promotes TCR-T-cell therapy resistance by inducing CD163 + M2 macrophage polarization and MMP9 secretion. J Immunother Cancer 12:e008375. https://doi.org/10.1136/jitc-2023-008375 Ahmed HG, Osman SI, Ashankyty IM (2012) Incidence of Epstein-Barr virus in pediatric leukemia in the Sudan. Clin Lymphoma Myeloma Leuk 12:127–131. https://doi.org/10.1016/j.clml.2011.11.006 Halabi MA, Jaccard A, Moulinas R, Bahri R, Al Mouhammad H, Mammari N et al (2016) Clonal deleted latent membrane protein 1 variants of Epstein-Barr virus are predominant in European extranodal NK/T lymphomas and disappear during successful treatment. Int J Cancer 139:793–802. https://doi.org/10.1002/ijc.30128 Ateyah ME, Hashem ME, Abdelsalam M (2017) Epstein-Barr virus and regulatory T cells in Egyptian paediatric patients with acute B lymphoblastic leukaemia. J Clin Pathol 70:120–125. https://doi.org/10.1136/jclinpath-2016-203803 Li Y, Wu Y, Cao S, Yu B, Zhang Q, Xia Z et al (2024) PD-1 inhibitors plus chemotherapy for refractory EBV-positive DLBCL: a retrospective analysis. Blood Res 59:36. https://doi.org/10.1007/s44313-024-00042-6 Deng W, Xu Y, Yuan X (2022) Clinical features and prognosis of acute lymphoblastic leukemia in children with Epstein-Barr virus infection. Transl Pediatr 11:642–650. https://doi.org/10.21037/tp-22-146 Silva JM, Alves CEC, Pontes GS (2024) Epstein-Barr virus: the mastermind of immune chaos. Front Immunol 15:1297994. https://doi.org/10.3389/fimmu.2024.1297994 Khan A, Zhang Y, Ma N, Shi J, Hou Y (2024) NF-κB role on tumor proliferation, migration, invasion and immune escape. Cancer Gene Ther 31:1599–1610. https://doi.org/10.1038/s41417-024-00811-6 Owen KL, Brockwell NK, Parker BS (2019) JAK-STAT Signaling: A double-edged sword of immune regulation and cancer progression. Cancers 11:2002. https://doi.org/10.3390/cancers11122002 Xu H, Chen D, Lu J, Wang L, Chen F, Zhong L (2025) LST1 expression correlates with immune infiltration and predicts poor prognosis in acute myeloid leukemia. Discov Oncol 16:305. https://doi.org/10.1007/s12672-025-02082-5 Guilbaud E, Galluzzi L (2025) Tumor immunology: a short IL18 cleavage product promotes cancer immunosurveillance. Signal Transduct Target Ther 10:129. https://doi.org/10.1038/s41392-025-02213-y Münz C (2025) Epstein-Barr virus pathogenesis and emerging control strategies. Nat Rev Microbiol 23:667–679. https://doi.org/10.1038/s41579-025-01181-y Šimičić P, Batović M, Stojanović Marković A, Židovec-Lepej S (2024) Deciphering the role of Epstein-Barr virus latent membrane protein 1 in immune modulation: A multifaced signalling perspective. Viruses 16:564. https://doi.org/10.3390/v16040564 Kieser A, Sterz KR (2015) The latent membrane protein 1 (LMP1). Curr Top Microbiol Immunol 391:119–149. https://doi.org/10.1007/978-3-319-22834-1_4 Chatterjee K, Das P, Chattopadhyay NR, Mal S, Choudhuri T (2019) The interplay between Epstein-Bar virus (EBV) with the p53 and its homologs during EBV associated malignancies. Heliyon 5:e02624. https://doi.org/10.1016/j.heliyon.2019.e02624 Molina E, García-Gutiérrez L, Junco V, Perez-Olivares M, De Yébenes VG, Blanco R et al (2023) MYC directly transactivates CR2/CD21, the receptor of the Epstein-Barr virus, enhancing the viral infection of Burkitt lymphoma cells. Oncogene 42:3358–3370. https://doi.org/10.1038/s41388-023-02846-9 Yu WL, Hua ZC (2022) Identification of immune and stromal cell infiltration-related gene signature for prognosis prediction in acute lymphoblastic leukemia. Aging 14:7470–7504. https://doi.org/10.18632/aging.204292 Soo Hoo WI, Higa K, McCormick AA (2023) Vaccination against Epstein-Barr latent membrane protein 1 protects against an Epstein-Barr virus-associated B cell model of lymphoma. Biology (Basel) 12:983. https://doi.org/10.3390/biology12070983 Xie Y, Wang A, Lin J, Wu L, Zhang H, Yang X et al (2017) Mps1/TTK: a novel target and biomarker for cancer. J Drug Target 25:112–118. https://doi.org/10.1080/1061186X.2016.1258568 Guan X, Guo H, Guo Y, Han Q, Li Z, Zhang C (2024) Perforin 1 in cancer: mechanisms, therapy, and outlook. Biomolecules 14:910. https://doi.org/10.3390/biom14080910 Pozzi S, Satchi-Fainaro R (2024) The role of CCL2/CCR2 axis in cancer and inflammation: the next frontier in nanomedicine. Adv Drug Deliv Rev 209:115318. https://doi.org/10.1016/j.addr.2024.115318 Mukaida N, Sasaki SI, Baba T (2020) CCL4 signaling in the tumor microenvironment. Adv Exp Med Biol 1231:23–32. https://doi.org/10.1007/978-3-030-36667-4_3 Korbecki J, Kojder K, Simińska D, Bohatyrewicz R, Gutowska I, Chlubek D, Baranowska-Bosiacka I (2020) CC chemokines in a tumor: a review of pro-cancer and anti-cancer Properties of the Ligands of Receptors CCR1, CCR2, CCR3, and CCR4. Int J Mol Sci 21:8412. https://doi.org/10.3390/ijms21218412 Scheuermann RH, Racila E (1995) CD19 antigen in leukemia and lymphoma diagnosis and immunotherapy. Leuk Lymphoma 18:385–397. https://doi.org/10.3109/10428199509059636 Additional Declarations No competing interests reported. Supplementary Files OnlineResource1SupplementaryTables.xlsx Online Resource 1 (.xlsx). Title of data: Supplementary material Description of data: Supplementary Tables S1 to S6 OnlineResource2SupplementaryFigures.docx Online Resource 2 (word) Title of data: Supplementary material Description of data: Supplementary Figures S1 to S3 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7725271","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":531794080,"identity":"417058a4-c5b1-4412-89b0-3bb8db9d61cc","order_by":0,"name":"Lili Pan","email":"","orcid":"","institution":"Fujian Institute of Hematology, Fujian Medical University Union Hospital","correspondingAuthor":false,"prefix":"","firstName":"Lili","middleName":"","lastName":"Pan","suffix":""},{"id":531794082,"identity":"b2cea982-0b37-4e55-ab0d-916b4df7c2ab","order_by":1,"name":"Chunping Wu","email":"","orcid":"","institution":"Fujian Institute of Hematology, Fujian Medical University Union Hospital","correspondingAuthor":false,"prefix":"","firstName":"Chunping","middleName":"","lastName":"Wu","suffix":""},{"id":531794084,"identity":"e7979d7d-c972-480b-bbfb-47c53de667f9","order_by":2,"name":"Dezheng Ning","email":"","orcid":"","institution":"Shanghai Jiao Tong University","correspondingAuthor":false,"prefix":"","firstName":"Dezheng","middleName":"","lastName":"Ning","suffix":""},{"id":531794087,"identity":"cc983287-ffec-4b9f-ae83-c1147ac9720a","order_by":3,"name":"Yongzhi Zheng","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8ElEQVRIiWNgGAWjYBACAxDBw2BTzyb/+OCDDwY2dsRqSUvgZ0hLNpxRkJZMrJbDCZINOWbSPB8OMTYQ0mIuffjZg7dtaXkGBw4YG9sYHGBmYD98dAM+LZZ9aeaGc9tsig0ONiQ+zjG4w8fAk5Z2A6/DzjCYSfO2pTFuOMxw2DjH4BkzgwSPGQEt7N+AWg4zbjjG2CZtYXCYsYGwFh6QLYcTZ/Yws0kzEKPFsoenTHLOuTRjfgk2ZsMeg7RkNkJ+Medh3ybxpsxGjk2C/+ODH39s7PjZDx/DqwUMGNmQOGw4laGAP8QpGwWjYBSMghEKACu1Sok2hRWdAAAAAElFTkSuQmCC","orcid":"","institution":"Fujian Institute of Hematology, Fujian Medical University Union Hospital","correspondingAuthor":true,"prefix":"","firstName":"Yongzhi","middleName":"","lastName":"Zheng","suffix":""}],"badges":[],"createdAt":"2025-09-27 01:53:21","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7725271/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7725271/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":94137827,"identity":"298336c1-7ad0-412b-bc1c-a89c2fd3c4fb","added_by":"auto","created_at":"2025-10-22 19:23:33","extension":"tif","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2132956,"visible":true,"origin":"","legend":"","description":"","filename":"Fig1.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/76303f0232cc2d092cbf37bf.tif"},{"id":94137826,"identity":"83157576-709c-4c9b-9037-b0c41b7b8ad8","added_by":"auto","created_at":"2025-10-22 19:23:33","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53996,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/5d793a3bf2dc49a59d6acd8b.docx"},{"id":94137856,"identity":"51b87698-1718-40aa-be09-d6c4e98329ea","added_by":"auto","created_at":"2025-10-22 19:23:35","extension":"tif","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2252132,"visible":true,"origin":"","legend":"","description":"","filename":"Fig2.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/b6616e6a0b8c1840874fd560.tif"},{"id":94137829,"identity":"845abaac-e1a0-460e-bf64-79f443d16092","added_by":"auto","created_at":"2025-10-22 19:23:33","extension":"tif","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1977420,"visible":true,"origin":"","legend":"","description":"","filename":"Fig3.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/24bef598fbf368d86c9af172.tif"},{"id":94140271,"identity":"214f820b-5998-46ad-8f21-946c38dedfb2","added_by":"auto","created_at":"2025-10-22 19:39:33","extension":"tif","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3806184,"visible":true,"origin":"","legend":"","description":"","filename":"Fig4.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/2c3427c53a23608f5ed7f340.tif"},{"id":94137818,"identity":"57a7e7cb-3efd-4fd4-a504-c67a2cd803cd","added_by":"auto","created_at":"2025-10-22 19:23:32","extension":"tif","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3085228,"visible":true,"origin":"","legend":"","description":"","filename":"Fig5.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/885b7e9ffa9b7fc55dfe4eee.tif"},{"id":94139051,"identity":"be2acd5b-d26d-4ec9-8bf4-ff2ac53b783c","added_by":"auto","created_at":"2025-10-22 19:31:33","extension":"tif","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3062136,"visible":true,"origin":"","legend":"","description":"","filename":"Fig6.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/4127b9c48277417a279706cb.tif"},{"id":94140270,"identity":"f2c57cdc-00c4-44d1-a427-eb5482ea5928","added_by":"auto","created_at":"2025-10-22 19:39:33","extension":"json","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6619,"visible":true,"origin":"","legend":"","description":"","filename":"5d81bd302ca54cc48d8e8741124d6a5a.json","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/0ba25933fdfa4b6b44fb1eb7.json"},{"id":94139055,"identity":"e7987a80-8669-4716-b8f6-26e2ab36bc40","added_by":"auto","created_at":"2025-10-22 19:31:34","extension":"xlsx","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3097439,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource1SupplementaryTables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/18c92e3d20fbe0f64783d667.xlsx"},{"id":94137857,"identity":"a4a54df6-1fac-4a7d-823a-8144e00f502b","added_by":"auto","created_at":"2025-10-22 19:23:35","extension":"docx","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2633598,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource2SupplementaryFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/8145737a50c1a64e5443e90b.docx"},{"id":94139097,"identity":"7b12f336-fe83-486e-bee0-91448b61d241","added_by":"auto","created_at":"2025-10-22 19:31:35","extension":"xml","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":95155,"visible":true,"origin":"","legend":"","description":"","filename":"5d81bd302ca54cc48d8e8741124d6a5a1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/26abfdd425a14affda0e837d.xml"},{"id":94137849,"identity":"72a904be-aea3-4b31-939d-9cd606567934","added_by":"auto","created_at":"2025-10-22 19:23:34","extension":"tif","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2132956,"visible":true,"origin":"","legend":"","description":"","filename":"Fig1.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/8dc3dfd58aa3b6f3fc90d5ef.tif"},{"id":94137837,"identity":"575b238d-c12c-41f7-91e3-4d313c7f1096","added_by":"auto","created_at":"2025-10-22 19:23:34","extension":"tif","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2252132,"visible":true,"origin":"","legend":"","description":"","filename":"Fig2.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/885546c3ddc39a9c76a242d4.tif"},{"id":94137822,"identity":"b61eda38-9b27-41f2-9997-10fe47420d47","added_by":"auto","created_at":"2025-10-22 19:23:32","extension":"tif","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1977420,"visible":true,"origin":"","legend":"","description":"","filename":"Fig3.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/34e6f8e120aa29c19d56b1bd.tif"},{"id":94139099,"identity":"b99c76c1-0e75-46d6-a928-75c29eb839b4","added_by":"auto","created_at":"2025-10-22 19:31:35","extension":"tif","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3806184,"visible":true,"origin":"","legend":"","description":"","filename":"Fig4.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/41d1a59ccfecb692b18fd335.tif"},{"id":94137865,"identity":"6b7d91cf-a257-4285-95ce-505691cc6f8a","added_by":"auto","created_at":"2025-10-22 19:23:35","extension":"tif","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3085228,"visible":true,"origin":"","legend":"","description":"","filename":"Fig5.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/f0af3220fe10b56e3c2d23a5.tif"},{"id":94137838,"identity":"fc877c77-01dd-4346-8c02-ac4a9a5054d1","added_by":"auto","created_at":"2025-10-22 19:23:34","extension":"tif","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3062136,"visible":true,"origin":"","legend":"","description":"","filename":"Fig6.tif","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/3d3bb157ba14a02dc6602a7a.tif"},{"id":94137825,"identity":"18fd98e4-0476-44dc-9fbf-35143a940c0f","added_by":"auto","created_at":"2025-10-22 19:23:33","extension":"xml","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":95408,"visible":true,"origin":"","legend":"","description":"","filename":"5d81bd302ca54cc48d8e8741124d6a5a1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/54910c08c6b268458ac8dfd6.xml"},{"id":94137869,"identity":"8d863617-4f24-454d-aca2-4bfb12dd0939","added_by":"auto","created_at":"2025-10-22 19:23:35","extension":"html","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":104878,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/11bb5050f02bc8e0935f27fd.html"},{"id":94137819,"identity":"febff695-8772-4750-8db5-86a8c15c7212","added_by":"auto","created_at":"2025-10-22 19:23:32","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":293003,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStudy workflow.\u003c/strong\u003e TARGET-ALL, Therapeutically Applicable Research to Generate Effective Treatments; MP2PRT-ALL, Molecular Profiling to Predict Response to Therapy in Acute Lymphoblastic Leukemia; ALL, acute lymphoblastic leukemia; EB, Epstein–Barr\u003c/p\u003e","description":"","filename":"OnlineFig1.png","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/9bcc5803a7222a5e4451191c.png"},{"id":94137841,"identity":"8aa13f4d-818d-4861-80a3-e0c59d1cbf86","added_by":"auto","created_at":"2025-10-22 19:23:34","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":296354,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWGCNA analysis of EBV-related gene expression\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003ePCA plot demonstrating a clear distinction between EBV-infected and control samples.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(B)\u003c/strong\u003eVolcano plot showing the distribution of DEGs between the two groups.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(C)\u003c/strong\u003eScale-free topology fit index (R²) plotted against various soft-thresholding powers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(D)\u003c/strong\u003eMean connectivity plotted against soft-thresholding powers, showing decreased connectivity with increasing power, indicating network sparsity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(E)\u003c/strong\u003eClustering of module eigengenes and merging of similar modules. Modules with high similarity were merged at cutHeight = 0.25, resulting in six distinct co-expression modules.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(F)\u003c/strong\u003eFour significant modules—cyan, tan, pink, and green-yellow—were selected based on their significant correlations with the trait of interest (P \u0026lt; 0.05).\u003c/p\u003e\n\u003cp\u003eWGCNA, weighted gene co-expression network analysis; EBV, Epstein–Barr virus; PCA, principal components analysis; DEGs, differentially-expressed genes\u003c/p\u003e","description":"","filename":"OnlineFig2.png","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/a66420352d2ff481daf9f707.png"},{"id":94137851,"identity":"a6c56ea1-c75a-4610-8edb-790b0a211277","added_by":"auto","created_at":"2025-10-22 19:23:35","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":288634,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEARG identification and functional characterization\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eA) \u003c/strong\u003eVolcano plot showing DEGs between pediatric patients with ALL and healthy controls (upper plot). The Venn diagram shows key genes related to EBV (lower plot). A total of 1,175 EARGs were identified.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eB\u003c/strong\u003e) Functional module detection using the MCODE plugin in Cytoscape. Two major gene clusters, clusters 1 and 2, were identified in the EARG PPI network.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eC\u003c/strong\u003e) GO and KEGG enrichment analyses of the genes in clusters 1 and 2.\u003c/p\u003e\n\u003cp\u003eEARGs, EBV-associated ALL-related genes; EBV, Epstein–Barr virus; ALL, acute lymphoblastic leukemia; DEGs, differentially-expressed genes; MCODE, Molecular Complex Detection; PPI, protein-protein interaction; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes\u003c/p\u003e","description":"","filename":"OnlineFig3.png","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/2c010bcf919c766624c37aa7.png"},{"id":94137817,"identity":"d8235016-a2a9-47d4-b80b-d1d8ff6a4f6a","added_by":"auto","created_at":"2025-10-22 19:23:32","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":281272,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eImmune landscape and differential gene expression in TARGET-ALL samples\u003c/strong\u003e\u003cbr\u003e\n(\u003cstrong\u003eA\u003c/strong\u003e) Tumor immune infiltration analysis on TARGET-ALL, using prescreened stromal/immune-related gene sets to predict immune and stromal cell infiltration. Samples were stratified into high and low immune score groups based on the median immune score.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eB\u003c/strong\u003e) Bar plots showing tumor purity (\u003cstrong\u003eB\u003c/strong\u003e), ESTIMATE scores (\u003cstrong\u003eC\u003c/strong\u003e), immune scores (\u003cstrong\u003eD\u003c/strong\u003e), and stromal scores (\u003cstrong\u003eE\u003c/strong\u003e) for groups with high and low immune scores.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eF\u003c/strong\u003e) Kaplan–Meier survival analysis indicating worse overall survival in the high immune score group.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eG\u003c/strong\u003e) Differential expression analysis between the high and low immune score groups showing 2031 differentially expressed genes.\u003c/p\u003e\n\u003cp\u003eTARGET-ALL, Therapeutically Applicable Research to Generate Effective Treatments; ESTIMATE, Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data\u003c/p\u003e","description":"","filename":"OnlineFig4.png","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/381270518e9074d94d0a07c2.png"},{"id":94137846,"identity":"a528f7d1-ccef-4c1a-a107-81df094b1e8f","added_by":"auto","created_at":"2025-10-22 19:23:34","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":278143,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConstruction and Validation of the Prognostic Model for ALL\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eA\u003c/strong\u003e) Intersection of the 2031 differentially expressed genes from the TARGET-ALL dataset with EARGs, resulting in 49 candidate genes.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eB\u003c/strong\u003e) LASSO regression to select the most significant prognostic genes. Left: relationship between the log-transformed penalty parameter (Log[λ]) and partial likelihood deviance, with the optimal λ value (dashed line) selected by cross-validation to minimize deviance and prevent overfitting. Right: changes in the coefficients of the 17 selected prognostic genes with varying λ, highlighting the most influential genes contributing to the model.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eC\u003c/strong\u003e) Forest plot showing the HRs of the nine selected prognostic genes based on multivariate Cox regression analysis. Genes with HRs \u0026gt;1 and \u0026lt;1 are associated with increased risk and protective effects on overall survival, respectively.\u003c/p\u003e\n\u003cp\u003eROC curves and corresponding AUC values for 1-, 3-, and 5-year survival using the prognostic model, with AUC values of 0.72, 0.84, and 0.85, respectively.\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eE\u003c/strong\u003e) Kaplan–Meier survival curves showing significantly worse overall survival among patients with high-risk scores than those with low-risk scores (P \u0026lt; 0.001, log-rank test).\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eF–G\u003c/strong\u003e) Feature gene heatmaps, risk score distribution, and survival status of the high- and low-risk score groups.\u003c/p\u003e\n\u003cp\u003eALL, acute lymphoblastic leukemia; DEGs, differentially expressed genes; TARGET-ALL, Therapeutically Applicable Research to Generate Effective Treatments; EARGs, EBV-associated ALL-related genes; EBV, Epstein–Barr virus; LASSO, least absolute shrinkage and selection operator; HR, hazard ratio; ROC, receiver operating characteristic; AUC, area under the curve\u003c/p\u003e","description":"","filename":"OnlineFig5.png","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/62a991e105cbf76fb7113c24.png"},{"id":94137843,"identity":"8abf8ff1-7d59-451a-ba07-ddf69869cff9","added_by":"auto","created_at":"2025-10-22 19:23:34","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":427885,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSingle-cell RNA sequencing analysis of prognostic gene expression in ALL\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003eOriginal single-cell dataset from ALL samples.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(B)\u003c/strong\u003et-SNE dimensionality reduction of the single-cell RNA sequencing data, resulting in 13 distinct clusters.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(C)\u003c/strong\u003eIdentification of five distinct cell types (T cells, B cells, NK cells, MEPs, and monocytes) within the 13 clusters.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(D)\u003c/strong\u003eHeatmap showing the distributions of the nine prognostic genes across the 13 cell clusters.\u003c/p\u003e\n\u003cp\u003eALL, acute lymphoblastic leukemia; t-SNE, t-distributed stochastic neighbor embedding; NK, natural killer; MEP, megakaryocyte-erythroid progenitor\u003c/p\u003e","description":"","filename":"OnlineFig6.png","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/de8698f6170abc7b81e9de56.png"},{"id":96251892,"identity":"464a39fa-577a-42f8-b82d-6d206b0d013f","added_by":"auto","created_at":"2025-11-19 07:40:09","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3539818,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/f8a36868-eae0-4898-a881-174b4cf1141f.pdf"},{"id":94139054,"identity":"1dc5d737-a783-4408-9a70-bef15beebbf3","added_by":"auto","created_at":"2025-10-22 19:31:34","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3097439,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOnline Resource 1\u003c/strong\u003e (.xlsx).\u003c/p\u003e\n\u003cp\u003eTitle of data: Supplementary material\u003c/p\u003e\n\u003cp\u003eDescription of data: Supplementary Tables S1 to S6\u003c/p\u003e","description":"","filename":"OnlineResource1SupplementaryTables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/858df6071a76191f64562bca.xlsx"},{"id":94137861,"identity":"b198d94c-8b8c-441e-8d1f-d2ae206eaae3","added_by":"auto","created_at":"2025-10-22 19:23:35","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2633598,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOnline Resource 2\u003c/strong\u003e (word)\u003c/p\u003e\n\u003cp\u003eTitle of data: Supplementary material\u003c/p\u003e\n\u003cp\u003eDescription of data: Supplementary Figures S1 to S3\u003c/p\u003e","description":"","filename":"OnlineResource2SupplementaryFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-7725271/v1/a800c7fd017cb4d6f4e0964e.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Immune dysregulation and prognostic signatures associated with Epstein–Barr virus in acute lymphoblastic leukemia: an integrated transcriptomic and single-cell analysis","fulltext":[{"header":"BACKGROUND","content":"\u003cp\u003eAcute lymphoblastic leukemia (ALL) is a malignancy of lymphoid progenitors and the most common pediatric cancer (~\u0026thinsp;75% of childhood leukemias); moreover, it comprises 20\u0026ndash;30% of adult acute leukemias [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Despite therapeutic advances, relapse remains a major challenge, underscoring the need for deeper biological insights. Notably, infectious exposures have been implicated in leukemogenesis via persistent inflammation, genomic instability, and aberrant signaling [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eEpstein\u0026ndash;Barr virus (EBV), a ubiquitous oncogenic herpesvirus, establishes lifelong latency in memory B cells and is linked to multiple cancers, primarily via latent programs that modulate host pathways and enable immune evasion [\u003cspan additionalcitationids=\"CR5 CR6\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Evidence implicates EBV in the development and progression of Burkitt\u0026rsquo;s lymphoma, Hodgkin\u0026rsquo;s lymphoma, nasopharyngeal carcinoma, and gastric cancer. EBV-related activity has been associated with tumor microenvironment remodeling, including M2 macrophage enrichment, increased MMP9, T-cell exhaustion, and treatment resistance [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. These observations indicate that EBV-related immunologic states may accompany, but do not by themselves prove, causal viral oncogenesis in every context.\u003c/p\u003e\u003cp\u003eIn ALL, the contribution of EBV remains uncertain and context-dependent. Pediatric data report higher EBV seroprevalence and LMP1 positivity compared to those of controls, alongside expansion of regulatory T cells and elevated IL-10, suggesting immune modulation that could intersect with leukemic biology [\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. These findings indicate that EBV-related immune modulation could intersect with leukemic biology, potentially shaping immune evasion and disease behavior, while not establishing direct causality. Moreover, chronic inflammation is a key contributor to chemotherapy resistance in various cancers [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], which may explain the poor treatment outcomes observed in such patients [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eTherefore, we integrated publicly available bulk and single-cell transcriptomes to evaluate the associations between EBV-related transcriptional features, immune context, and prognosis in ALL. The study workflow is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"METHODS","content":"\u003ch2\u003eMethods\u003c/h2\u003e\n\u003cp\u003eGene expression microarray datasets associated with EBV infection (GSE45918 and GSE85599) and ALL (GSE26713 and GSE67684) were obtained from the GEO database. Consistent Affymetrix Probe Set IDs within the same sample type were selected, and batch effects among different datasets were corrected using the \u0026ldquo;ComBat\u0026rdquo; function from the \u0026ldquo;sva\u0026rdquo; package in R. Transcriptomic data and clinical information for the training cohort were retrieved from the TARGET database phases TARGET-ALL-P1 and TARGET-ALL-P2, comprising 679 samples obtained from patients with leukemia. The validation cohort, which included 1479 samples from patients with leukemia, was obtained from the MP2PRT-ALL database. scRNA-seq data were sourced from GSE196214, which involved 21 patients newly diagnosed with B/T-cell acute lymphoblastic leukemia.\u003c/p\u003e\n\u003ch2\u003eDifferential expression analysis\u003c/h2\u003e\n\u003cp\u003eDifferential expression analyses comparing EBV-infected samples and controls, as well as ALL samples and controls, were conducted using the \u0026ldquo;limma\u0026rdquo; package in R. Limma uses linear models combined with empirical Bayesian methods, providing a robust statistical inference particularly suitable for datasets with limited sample sizes. Differentially expressed genes (DEGs) were identified based on a threshold of \u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u0026lt; 0.05, |log2 fold change| \u0026gt;1. Count data from the TARGET and MP2PRT cohorts were subjected to DEG analysis using the \u0026ldquo;edgeR\u0026rdquo; package, which applies a negative binomial distribution model. Data were normalized using the trimmed mean M-values. Finally, dispersion estimates were calculated for statistical testing to detect DEGs. In addition, we conducted weighted gene co-expression network analysis (WGCNA) on datasets from patients infected with EBV and control groups to identify gene modules significantly associated with EBV infection characteristics. By examining the relationships between modules and features and using hierarchical clustering to merge modules with high similarity, we further revealed the interconnectivity of the gene modules.\u003c/p\u003e\n\u003ch2\u003eProtein\u0026ndash;protein interaction (PPI) network analysis\u003c/h2\u003e\n\u003cp\u003eA PPI network was constructed using the Search Tool for the Retrieval of Interacting Genes/Proteins database (http://string-db.org). Key subnetworks were extracted using the Molecular Complex Detection (MCODE) plugin within the Cytoscape software (version 3.10.2), resulting in two major modules comprising 157 genes, which were subsequently utilized in downstream analyses.\u003c/p\u003e\n\u003cp\u003eTo further investigate the correlation between EBV-associated ALL-related genes (EARGs) and the immune landscape in ALL samples, we used the Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data (ESTIMATE) algorithm to calculate immune scores, stromal scores, ESTIMATE scores, and tumor purity.\u003c/p\u003e\n\u003ch2\u003eTumor immune infiltration\u003c/h2\u003e\n\u003cp\u003eTumor purity and stromal cell composition were assessed using estimation algorithms. According to the median immune score, the samples were divided into two groups, followed by differential and survival analyses. Single-sample gene set enrichment analysis for 29 immune-related signatures was performed using the \u0026ldquo;GSVA\u0026rdquo; package (method = \u0026ldquo;ssgsea\u0026rdquo;), with gene sets represented via \u0026ldquo;GSEAbase.\u0026rdquo; Graphical representations were rendered using the \u0026ldquo;ggplot2\u0026rdquo; and \u0026ldquo;pheatmap\u0026rdquo; packages in R.\u003c/p\u003e\n\u003ch2\u003ePrognostic model construction\u003c/h2\u003e\n\u003cp\u003eUnivariate Cox proportional hazards regression analysis was used to identify candidate prognostic genes (\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u0026lt; 0.05). Redundant variables were eliminated using the least absolute shrinkage and selection operator (LASSO) regression. LASSO was implemented with the \u0026ldquo;glmnet\u0026rdquo; package, and the optimal penalty parameter (\u0026lambda;) was selected by cross-validation. A prognostic model was then developed using multivariate Cox proportional hazards regression. Patients were stratified into high- and low-risk groups based on median risk scores. Kaplan\u0026ndash;Meier survival analysis with log-rank tests was used to evaluate survival differences between the groups. Time-dependent receiver operating characteristic (ROC) curves were generated using the \u0026ldquo;timeROC\u0026rdquo; package to assess 1-, 3-, and 5-year performance. The prognostic model was externally validated in the MP2PRT-ALL validation cohort.\u003c/p\u003e\n\u003ch2\u003escRNA-seq analysis\u003c/h2\u003e\n\u003cp\u003eThe top 2000 highly variable genes were identified using the variance-stabilizing transformation method implemented in the FindVariableFeatures function in the Seurat R package. The data were scaled using the ScaleData function, whereas dimensionality reduction was performed via principal component analysis (PCA) using the RunPCA function. Batch effects among the patient samples were corrected using the harmony algorithm (R package \u0026ldquo;harmony\u0026rdquo;). Clustering was performed based on the first 30 significant principal components using FindNeighbors and FindClusters (resolution = 0.25). Dimensionality reduction was visualized using t-distributed stochastic neighbor embedding (RunTSNE). Cell types were identified using SingleR, followed by manual verification, and visualized using the \u0026ldquo;scCustomize\u0026rdquo; package.\u003c/p\u003e"},{"header":"RESULTS","content":"\u003ch2\u003eEBV-related gene expression and module identification\u003c/h2\u003e\n\u003cp\u003eTo identify DEGs regulated by EBV infection, we analyzed two gene expression datasets: GSE45918 (EBV-infected samples) and GSE85599 (normal controls). The PCA plot demonstrated clear separation between EBV-infected and control samples (Fig. 2A and B), consistent with EBV-associated transcriptional remodeling but not proof of causality. The volcano plot revealed 401 DEGs, including 230 upregulated and 171 downregulated genes (Online Resource 1: Supplementary Table S1). Subsequently, we performed a WGCNA using the EBV-related gene expression dataset (Fig. 2C). By evaluating the scale-free topology fit index and mean connectivity under various soft-thresholding powers, we determined the optimal soft-threshold (Fig. 2D and E) to confirm the stability and robustness of the network. With the merging threshold set at cutHeight = 0.25, similar gene modules were merged, resulting in the identification of six distinct co-expressed modules (Online Resource 2: Supplementary Figure S1 A and B). Based on biological plausibility and module eigengene patterns, we focused on four modules (cyan, tan, pink, and green), comprising 3,122 genes in total (Fig. 2F; Online Resource 1: Supplementary Table S2). Notably, the 401 DEGs were distributed across the selected modules rather than present in every module. Functional annotations of genes within the selected modules indicated enrichment for DNA repair, immune signaling, and cell-cycle regulation, processes broadly implicated in leukemogenesis . Accordingly, we describe these signals as associational rather than as EBV-specific causal effects.\u003c/p\u003e\n\u003ch2\u003eEBV-associated ALL genes and functional modules\u003c/h2\u003e\n\u003cp\u003eDifferential expression analysis between patients with ALL and healthy controls (GEO datasets GSE26713 and GSE67684) identified 882 DEGs (Online Resource 1: Supplementary Table S3, Fig. 3A). We then took the union of the EBV-related DEGs and ALL DEGs to generate an EBV-associated ALL gene set (EARGs; n = 1,175) (Fig. 3B), highlighting genes that co-occur across EBV-related and ALL comparisons without implying direct viral regulation in ALL.\u003c/p\u003e\n\u003cp\u003eTo further investigate putative interactions among EARGs, we constructed a PPI network in STRING and extracted two dense subnetworks (Clusters 1 and 2) using the MCODE plugin in Cytoscape (Fig. 3C\u0026ndash;D). Functional enrichment analyses of these modules (GO/KEGG) revealed distinct pathway patterns (Fig. 3E\u0026ndash;F). Cluster 1 was enriched for terms related to EBV infection, NF-\u0026kappa;B signaling, and JAK\u0026ndash;STAT signaling, which are consistent with immune/inflammatory signaling described in EBV-associated contexts and are also broadly implicated in leukemic biology [13\u0026ndash;15]. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn contrast, Cluster 2 encompassed a broader set of genes enriched for cell cycle, DNA replication, and p53 signaling, alongside GO terms such as nuclear division, chromosome segregation, and mitotic cell-cycle transition. These findings indicate prominent proliferation/cell-cycle programs typical of leukemogenesis.\u003c/p\u003e\n\u003cp\u003eCollectively, the EARG PPI highlights two thematic axes\u0026mdash;(i) immune/inflammatory signaling and (ii) cell-cycle/proliferation\u0026mdash;associated with EBV-related gene sets in our integrated analysis. These signals may partly reflect immune composition and proliferative states inherent to ALL and should not be over-interpreted as evidence of EBV-specific causal mechanisms without direct virological confirmation.\u003c/p\u003e\n\u003ch2\u003eImmune landscape in ALL\u003c/h2\u003e\n\u003cp\u003eThe results of the ESTIMATE algorithm are shown in Fig. 4A. Tumor immune infiltration analysis was performed on the TARGET-ALL dataset using prescreened immune/stromal-related gene sets. Using bulk gene-expression profiles as proxies, we estimated stromal and immune components within the leukemic microenvironment and stratified samples into high and low immune-score groups by the cohort median. As expected from the ESTIMATE framework, the high-score group exhibited lower inferred tumor purity and higher stromal scores than did the low-score group (Fig. 4B\u0026ndash;E; Online Resource 1: Supplementary Table S4). In Kaplan\u0026ndash;Meier analysis, the high immune-score group exhibited worse overall survival than did the low-score group (Fig. 4F).\u003c/p\u003e\n\u003cp\u003eThese findings suggest that immune cells in the tumor microenvironment promote immune evasion in ALL and that high immune infiltration is associated with poor prognosis in ALL, consistent with findings in acute myeloid leukemia [16].\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003ePrognostic model\u003c/h2\u003e\n\u003cp\u003eDifferential expression analysis between the high and low immune score groups identified 2031 DEGs (Fig. 4G). The intersection of these DEGs from the TARGET-ALL dataset with the EARGs produced 49 candidate genes (Fig. 5A), which, combined with clinical information, were analyzed using univariate Cox regression analysis using the survival package. All 49 genes significantly affected the overall survival (OS) of patients with ALL. Subsequently, LASSO shrinkage and selection operator regression were performed using the glmnet package, resulting in the identification of 17 feature genes. These 17 genes were further analyzed using multivariate Cox regression, leading to the identification of nine genes that were significantly associated with prognosis (Fig. 5B), including interleukin 18 (\u003cem\u003eIL18\u003c/em\u003e), Toll-like receptor 1 (\u003cem\u003eTLR1\u003c/em\u003e), perforin 1 (\u003cem\u003ePRF1\u003c/em\u003e), interleukin 6 receptor (\u003cem\u003eIL6R\u003c/em\u003e), C-C motif chemokine ligand 2 (\u003cem\u003eCCL2\u003c/em\u003e), TTK protein kinase (\u003cem\u003eTTK\u003c/em\u003e), CD19 molecule (\u003cem\u003eCD19\u003c/em\u003e), cathepsin S (\u003cem\u003eCTSS\u003c/em\u003e), and C-C motif chemokine ligand 4 (\u003cem\u003eCCL4\u003c/em\u003e). These genes corroborated emerging evidence suggesting the critical roles of immune-related genes such as \u003cem\u003eIL18\u003c/em\u003e in modulating the immune microenvironment and influencing the response to cancer therapy [17]. Subsequently, a prognostic risk score index using the immune-related gene pair index (IRGPI) was constructed based on the expression of these nine genes (Fig. 5C, Online Resource 1: Supplementary Table S5) using the following equation: Riskscore = 0.128 \u0026times; \u003cem\u003eIL18\u003c/em\u003e - 0.286 \u0026times; \u003cem\u003eTLR1\u003c/em\u003e - 0.112 \u0026times; \u003cem\u003ePRF1\u003c/em\u003e - 0.282 \u0026times; \u003cem\u003eIL6R\u003c/em\u003e + 0.118 \u0026times; \u003cem\u003eCCL2\u003c/em\u003e - 0.133 \u0026times; \u003cem\u003eTTK\u003c/em\u003e + 0.240 \u0026times; \u003cem\u003eCD19\u003c/em\u003e + 0.416 \u0026times; \u003cem\u003eCTSS\u003c/em\u003e + 0.173 \u0026times; \u003cem\u003eCCL4\u003c/em\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eUsing this model, risk scores were calculated for the clinical samples. After generating ROC curves, the corresponding areas under the curves (AUCs) for 1-, 3-, and 5-year survival were 0.72, 0.84, and 0.85, respectively (Fig. 5D), indicating the good discriminatory power of the prognostic model in predicting patient outcomes, further validating its potential clinical utility.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBased on the median risk score as a cutoff, patients with high-risk scores exhibited significantly worse OS than did those with low-risk scores (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001; Fig. 5E). Stratification of patients into high- and low-risk groups based on this model underscored its prognostic value in guiding treatment decisions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe risk score distribution, survival status, and feature gene heatmaps for the high- and low-risk score groups (Fig. 5F and G) showed higher \u003cem\u003eCCL2, CCL4\u003c/em\u003e, and \u003cem\u003eCD19\u003c/em\u003e expression and lower \u003cem\u003eTTK\u003c/em\u003e and \u003cem\u003ePRF1\u003c/em\u003e expression in the former than in the latter. Thus, elevated \u003cem\u003eCCL2\u003c/em\u003e and \u003cem\u003eCCL4\u003c/em\u003e expression may enhance the inflammatory microenvironment in ALL to promote immune evasion and disease progression. Additionally, the lower \u003cem\u003eTTK\u003c/em\u003e and\u003cem\u003e\u0026nbsp;PRF1\u003c/em\u003e expression may contribute to impaired immune surveillance and uncontrolled cell proliferation, which are critical for leukemia development.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eValidation using the external independent dataset MP2PRT-ALL confirmed the findings from the TARGET dataset, in which patients in the high-risk group showed significantly poorer OS compared with those in the low-risk group (Online Resource 2: Supplementary Figure S2A). These results strengthened the external applicability and robustness of our prognostic model, ensuring its generalizability across different patient populations.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTime-dependent ROC curve analysis further supported the robustness of the prognostic model, with AUC values of 0.82, 0.68, and 0.68 for 1-, 3-, and 5-year survival, respectively (Online Resource 2: Supplementary Figure S2B and C). These values provide additional evidence for the utility of the model for predicting long-term survival outcomes in patients with ALL.\u003c/p\u003e\n\u003ch2\u003eGene co-expression network and functional enrichment\u003c/h2\u003e\n\u003cp\u003eTo explore the potential biological functions of the prognostic genes, we used the GeneMANIA database (http://www.genemania.org/) to identify functionally associated genes and construct a gene co-expression network. Overall, 89.9% of the connections in the network were attributed to co-expression relationships, highlighting the interconnected nature of the identified genes, suggesting that these genes are functionally related and that their collective activity may contribute to the molecular mechanisms underlying ALL progression. Functional enrichment analysis indicated that this network was strongly associated with immune cell activity, highlighting the immune-related nature of the model (Online Resource 2: Supplementary Figure S3).\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eSingle-cell profiling of gene expression in ALL cellular subtypes\u003c/h2\u003e\n\u003cp\u003eWe further validated the prognostic model based on the nine identified genes by assessing its predictive performance and conducting a survival analysis.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo explore the biological relevance of the prognostic genes in the cellular context of ALL, we applied scRNA-seq to identify the spatial expression patterns of these genes within the ALL immune landscape. We identified 13 distinct clusters representing a diverse array of cellular subtypes (Fig. 6A\u0026ndash;C).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSubsequently, we analyzed these clusters to gain insight into their immune cell-type composition and potential functional roles in the progression of ALL. Cell markers for each of the 13 clusters were identified using the FindAllMarkers function (logFC = 0.5, MinPct = 0.25), which allowed for precise characterization of the cellular subtypes within the tumor microenvironment (Online Resource 1: Supplementary Table S6).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo explore the cellular distribution of the prognostic genes, we examined their expression profiles across the identified cell types (Fig. 6D), demonstrating that \u003cem\u003eTTK\u003c/em\u003e is predominantly expressed in B and T cells, thereby suggesting its involvement in modulating immune cell proliferation and differentiation in ALL. \u003cem\u003ePRF1\u003c/em\u003e, a key immune effector gene, was highly expressed in T and NK cells, indicating its role in mediating cytotoxic responses in the tumor microenvironment. \u003cem\u003eIL18, IL6R, CTSS\u003c/em\u003e, and \u003cem\u003eTLR1\u003c/em\u003e were evenly distributed across various cell types, suggesting their broader involvement in immune modulation and cell signaling within the ALL microenvironment. Moreover, \u003cem\u003eCD19\u003c/em\u003e and \u003cem\u003eCCL4\u003c/em\u003e were predominantly expressed in B cells, consistent with their established roles in B cell activation and migration, which are central to leukemia pathogenesis. Conversely, \u003cem\u003eCCL2\u003c/em\u003e exhibited higher expression in monocytes, highlighting its potential role in their recruitment and immune suppression within the tumor microenvironment (Fig. 6D).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCollectively, these findings demonstrated the utility of scRNA-seq for providing a detailed map of gene expression at the single-cell level, thereby offering novel insights into the role of prognostic genes in cellular subtypes within the ALL microenvironment. The spatial distribution of these genes indicates that immune cell interactions are key to ALL progression and treatment resistance.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThe integrated findings from WGCNA and PPI network analyses indicated that EBV infection perturbs gene expression modules closely associated with DNA repair, cell cycle regulation, and immune modulation. Enrichment of NF-κB and JAK\u0026ndash;STAT pathways plus immune/stromal signatures aligns with inflammatory and immune-evasion states reported in EBV-associated contexts and leukemic biology; however, it does not alone establish EBV-specific causality [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. These patterns may also co-vary with lineage composition and proliferative burden that our datasets cannot fully disentangle. Overall, the data support a hypothesis-generating model in which EBV-related activity may coincide with immune remodeling and cell-cycle programs in subsets of ALL, requiring validation in clinically EBV-characterized cohorts and functional studies.\u003c/p\u003e\u003cp\u003eLMP1 can mimic constitutively active CD40, recruiting TRAF adaptors and activating NF-κB and JNK to support B-cell proliferation and survival; similarly, LMP1 has been reported to dampen p53-mediated apoptosis via A20 induction and upregulation of BCL-2 family members [\u003cspan additionalcitationids=\"CR20\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. In addition, MYC can transactivate the EBV receptor CR2/CD21, potentially enhancing viral entry in MYC-dysregulated B cells [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. These mechanisms are well described in EBV-positive B-cell contexts and help explain lymphoproliferation. Here, enrichment of NF-κB/JAK\u0026ndash;STAT and cell-cycle programs aligns with pathways implicated in EBV-associated biology and leukemic proliferation. Nevertheless, the present data do not establish EBV-specific causality in ALL.\u003c/p\u003e\u003cp\u003eOur analysis demonstrated that patients with high immune scores exhibited worse overall survival (OS) than did those with low scores (P\u0026thinsp;=\u0026thinsp;0.015), corroborating previous findings that immune-microenvironment features correlate with ALL prognosis [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Because ESTIMATE-derived immune scores in hematologic malignancies largely reflect cell-mixture proportions and disease burden, our results are associational rather than proof that immune infiltration per se drives outcome differences. Several EBV-associated malignancies show adverse prognosis and chemoresistance [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]; however, extrapolation to ALL is uncertain without uniform, sample-level EBV ascertainment in our cohorts. If clinically EBV-characterized ALL studies confirm similar relationships, EBV status together with immune-infiltration profiles could serve as exploratory adjuncts for risk stratification, immunomonitoring, and therapy planning. At present, these markers remain hypothesis-generating, pending prospective validation and functional evidence.\u003c/p\u003e\u003cp\u003eBeyond validating the prognostic signature, our scRNA-seq analysis provided cell-type context for the signature genes\u0026mdash;clarifying their localization across immune and proliferative compartments. The reduced \u003cem\u003eTTK\u003c/em\u003e and \u003cem\u003ePRF1\u003c/em\u003e expression observed in the high-risk group may reflect impaired proliferative regulation of lymphoid cells and compromised cytotoxic immune responses, respectively. \u003cem\u003eTTK\u003c/em\u003e, a dual-specificity protein kinase essential for mitotic progression, has been implicated in maintaining chromosomal stability during lymphocyte proliferation; hence, its downregulation may lead to aberrant cell cycle control and genomic instability [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Meanwhile, decreased expression of \u003cem\u003ePRF1\u003c/em\u003e, a key effector molecule in T and NK cell-mediated cytotoxicity, suggests impaired antitumor immune defense, consistent with previous studies linking \u003cem\u003ePRF1\u003c/em\u003e deficiency to tumor immune evasion [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eIn contrast, elevated \u003cem\u003eCCL2, CCL4\u003c/em\u003e, and \u003cem\u003eCD19\u003c/em\u003e expression indicated an inflammatory yet immunosuppressive microenvironment favorable for leukemia progression. \u003cem\u003eCCL2\u003c/em\u003e recruits myeloid-derived suppressor cells and regulatory monocytes, both of which contribute to immune evasion and tumor support [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Similarly, high \u003cem\u003eCCL4\u003c/em\u003e expression can enhance leukemic cell migration, disrupt tumor microenvironment stability, and participate in carcinogenesis and tumor progression [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eCD19\u0026mdash;a canonical B-lineage marker\u0026mdash;likely reflects lineage composition/disease burden in bulk profiles, capturing expansion of malignant B-cell populations rather than acting as a direct driver [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Collectively, these signals should be viewed as associational; whether CCL2/CCL4-centered axes directly mediate immune suppression in ALL\u0026mdash;particularly in clinically EBV-characterized subsets\u0026mdash;requires targeted validation.\u003c/p\u003e\u003cp\u003eFuture studies should explore the critical mechanisms by which EBV contributes to the tumorigenic process in ALL. The IRGPI model, constructed based on EARGs in this study, provides preliminary evidence linking EBV-related gene signatures to ALL prognosis. Nonetheless, current evidence on the role of EBV in promoting ALL through the regulation of DNA repair, immune evasion, and cell cycle dysregulation is primarily derived from \u003cem\u003ein vitro\u003c/em\u003e studies, small-scale cohorts, and indirect observations, with a notable lack of large-scale clinical validation. Therefore, future investigations should integrate multicenter cohort studies, animal model experiments, and single-cell multi-omics approaches to systematically elucidate the specific molecular mechanisms underlying EBV infection in ALL progression and characterize the dynamic changes within the immune microenvironment during disease development.\u003c/p\u003e\u003cp\u003eOur integrative analysis links EBV-related transcriptional features with immune context and prognosis in ALL across public cohorts and, with scRNA-seq localization, provides a cell-type\u0026ndash;aware framework to interpret a parsimonious prognostic signature. Key limitations include mixed B-ALL/T-ALL cohorts, lack of uniform, sample-level EBV ascertainment, bulk-level composition confounding, and an observational design that precludes causal inference. Future work should use clinically EBV-characterized, lineage-stratified cohorts and composition-adjusted models to test signature generalizability and prospectively validate clinical utility.\u003c/p\u003e"},{"header":"CONCLUSIONS","content":"\u003cp\u003eEBV-related transcriptional activity is associated with immune/proliferation programs and risk in ALL. A 9-gene signature stratified prognosis; however, without uniform EBV status and with mixed lineages, the results remain hypothesis-generating. Prospective, EBV-characterized cohorts and functional studies are needed to establish clinical utility and causality.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eDeclaration of AI and AI-assisted technologies\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe used ChatGPT (OpenAI) to assist in improving the clarity and grammar of the manuscript\u0026apos;s English language during the writing and revision stages. The AI tool was not used for generating scientific content, designing the study, data analysis, or creating citations. All scientific interpretations, data analyses, and conclusions were solely developed by the authors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors gratefully acknowledge the public availability of the GEO datasets and thank the researchers who generated and shared these valuable resources.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization and Design of the Analysis: Pan L, Wu C; Data Collection: Pan L, Ning D; Provision of Data or Analysis Tools: Ning D; Data Analysis: Pan L, Ning D; Manuscript Writing: Pan L, Wu C, Ning D, Zheng Y.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Key Clinical Specialty Discipline Construction Program (2021-76); Fujian Provincial Clinical Research Center for Hematological Malignancies (2020Y2006); Mechanistic Study of CTHRC1-Mediated Regulation of Chemosensitivity in Acute T-Cell Lymphoblastic Leukemia Through the Hippo Signaling Pathway (2022QH1031).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used in this study are publicly available in the Gene Expression Omnibus (GEO) database under the accession numbers GSE45918, GSE85599, GSE26713, GSE67684, and GSE13159. Additional transcriptomic and clinical data were obtained from the TARGET (TARGET-ALL-P1, TARGET-ALL-P2) and MP2PRT-ALL databases. All data supporting the conclusions of this article are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompliance with Ethical Standards\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDisclosure of potential conflicts of interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that there is no conflict of interest regarding the publication of this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResearch involving Human Participants and/or Animals\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study did not involve any prospective enrollment of human participants or the collection of new human specimens. All analyses were performed on publicly available, de-identified transcriptomic datasets obtained from the Gene Expression Omnibus (GEO: GSE45918, GSE85599, GSE26713, GSE67684, GSE13159) and from publicly accessible consortia (TARGET-ALL-P1, TARGET-ALL-P2, MP2PRT-ALL, and GSE196214). Because these resources were generated under Institutional Review Board (IRB) approval and released with anonymized identifiers, no additional IRB review or written informed consent was required for the present secondary-data analysis, in accordance with the Declaration of Helsinki and the author guidelines of Cancer Research and Treatment.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInformed consent\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBecause these resources were generated under Institutional Review Board (IRB) approval and released with anonymized identifiers, no additional IRB review or written informed consent was required for the present secondary-data analysis, in accordance with the Declaration of Helsinki and the author guidelines of Cancer Research and Treatment.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eTran TH, Hunger SP (2022) The genomic landscape of pediatric acute lymphoblastic leukemia and precision medicine opportunities. Semin Cancer Biol 84:144\u0026ndash;152. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.semcancer.2020.10.013\u003c/span\u003e\u003cspan address=\"10.1016/j.semcancer.2020.10.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTerwilliger T, Abdul-Hay M (2017) Acute lymphoblastic leukemia: a comprehensive review and 2017 update. Blood Cancer J 7:e577. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/bcj.2017.53\u003c/span\u003e\u003cspan address=\"10.1038/bcj.2017.53\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGreaves M (2018) A causal mechanism for childhood acute lymphoblastic leukaemia. Nat Rev Cancer 18:471\u0026ndash;484. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41568-018-0015-6\u003c/span\u003e\u003cspan address=\"10.1038/s41568-018-0015-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDamania B, Kenney SC, Raab-Traub N (2022) Epstein-Barr virus: biology and clinical disease. Cell 185:3652\u0026ndash;3670. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cell.2022.08.026\u003c/span\u003e\u003cspan address=\"10.1016/j.cell.2022.08.026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOkuno Y, Murata T, Sato Y, Muramatsu H, Ito Y, Watanabe T et al (2019) Defective Epstein-Barr virus in chronic active infection and haematological malignancy. Nat Microbiol 4:404\u0026ndash;413. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41564-018-0334-0\u003c/span\u003e\u003cspan address=\"10.1038/s41564-018-0334-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAtri-Schuller A, Abushukair H, Cavalcante L, Hentzen S, Saeed A, Saeed A (2022) Tumor molecular and microenvironment characteristics in EBV-associated malignancies as potential therapeutic targets: focus on gastric cancer. Curr Issues Mol Biol 44:5756\u0026ndash;5767. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/cimb44110390\u003c/span\u003e\u003cspan address=\"10.3390/cimb44110390\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen Y, Ouyang D, Wang Y, Pan Q, Zhao J, Chen H et al (2024) EBV promotes TCR-T-cell therapy resistance by inducing CD163\u0026thinsp;+\u0026thinsp;M2 macrophage polarization and MMP9 secretion. J Immunother Cancer 12:e008375. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/jitc-2023-008375\u003c/span\u003e\u003cspan address=\"10.1136/jitc-2023-008375\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAhmed HG, Osman SI, Ashankyty IM (2012) Incidence of Epstein-Barr virus in pediatric leukemia in the Sudan. Clin Lymphoma Myeloma Leuk 12:127\u0026ndash;131. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.clml.2011.11.006\u003c/span\u003e\u003cspan address=\"10.1016/j.clml.2011.11.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHalabi MA, Jaccard A, Moulinas R, Bahri R, Al Mouhammad H, Mammari N et al (2016) Clonal deleted latent membrane protein 1 variants of Epstein-Barr virus are predominant in European extranodal NK/T lymphomas and disappear during successful treatment. Int J Cancer 139:793\u0026ndash;802. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ijc.30128\u003c/span\u003e\u003cspan address=\"10.1002/ijc.30128\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAteyah ME, Hashem ME, Abdelsalam M (2017) Epstein-Barr virus and regulatory T cells in Egyptian paediatric patients with acute B lymphoblastic leukaemia. J Clin Pathol 70:120\u0026ndash;125. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/jclinpath-2016-203803\u003c/span\u003e\u003cspan address=\"10.1136/jclinpath-2016-203803\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi Y, Wu Y, Cao S, Yu B, Zhang Q, Xia Z et al (2024) PD-1 inhibitors plus chemotherapy for refractory EBV-positive DLBCL: a retrospective analysis. Blood Res 59:36. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s44313-024-00042-6\u003c/span\u003e\u003cspan address=\"10.1007/s44313-024-00042-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDeng W, Xu Y, Yuan X (2022) Clinical features and prognosis of acute lymphoblastic leukemia in children with Epstein-Barr virus infection. Transl Pediatr 11:642\u0026ndash;650. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21037/tp-22-146\u003c/span\u003e\u003cspan address=\"10.21037/tp-22-146\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSilva JM, Alves CEC, Pontes GS (2024) Epstein-Barr virus: the mastermind of immune chaos. Front Immunol 15:1297994. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fimmu.2024.1297994\u003c/span\u003e\u003cspan address=\"10.3389/fimmu.2024.1297994\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKhan A, Zhang Y, Ma N, Shi J, Hou Y (2024) NF-κB role on tumor proliferation, migration, invasion and immune escape. Cancer Gene Ther 31:1599\u0026ndash;1610. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41417-024-00811-6\u003c/span\u003e\u003cspan address=\"10.1038/s41417-024-00811-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOwen KL, Brockwell NK, Parker BS (2019) JAK-STAT Signaling: A double-edged sword of immune regulation and cancer progression. Cancers 11:2002. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/cancers11122002\u003c/span\u003e\u003cspan address=\"10.3390/cancers11122002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXu H, Chen D, Lu J, Wang L, Chen F, Zhong L (2025) LST1 expression correlates with immune infiltration and predicts poor prognosis in acute myeloid leukemia. Discov Oncol 16:305. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s12672-025-02082-5\u003c/span\u003e\u003cspan address=\"10.1007/s12672-025-02082-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGuilbaud E, Galluzzi L (2025) Tumor immunology: a short IL18 cleavage product promotes cancer immunosurveillance. Signal Transduct Target Ther 10:129. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41392-025-02213-y\u003c/span\u003e\u003cspan address=\"10.1038/s41392-025-02213-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eM\u0026uuml;nz C (2025) Epstein-Barr virus pathogenesis and emerging control strategies. Nat Rev Microbiol 23:667\u0026ndash;679. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41579-025-01181-y\u003c/span\u003e\u003cspan address=\"10.1038/s41579-025-01181-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eŠimičić P, Batović M, Stojanović Marković A, Židovec-Lepej S (2024) Deciphering the role of Epstein-Barr virus latent membrane protein 1 in immune modulation: A multifaced signalling perspective. Viruses 16:564. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/v16040564\u003c/span\u003e\u003cspan address=\"10.3390/v16040564\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKieser A, Sterz KR (2015) The latent membrane protein 1 (LMP1). Curr Top Microbiol Immunol 391:119\u0026ndash;149. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-319-22834-1_4\u003c/span\u003e\u003cspan address=\"10.1007/978-3-319-22834-1_4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChatterjee K, Das P, Chattopadhyay NR, Mal S, Choudhuri T (2019) The interplay between Epstein-Bar virus (EBV) with the p53 and its homologs during EBV associated malignancies. Heliyon 5:e02624. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.heliyon.2019.e02624\u003c/span\u003e\u003cspan address=\"10.1016/j.heliyon.2019.e02624\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMolina E, Garc\u0026iacute;a-Guti\u0026eacute;rrez L, Junco V, Perez-Olivares M, De Y\u0026eacute;benes VG, Blanco R et al (2023) MYC directly transactivates CR2/CD21, the receptor of the Epstein-Barr virus, enhancing the viral infection of Burkitt lymphoma cells. Oncogene 42:3358\u0026ndash;3370. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41388-023-02846-9\u003c/span\u003e\u003cspan address=\"10.1038/s41388-023-02846-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYu WL, Hua ZC (2022) Identification of immune and stromal cell infiltration-related gene signature for prognosis prediction in acute lymphoblastic leukemia. Aging 14:7470\u0026ndash;7504. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.18632/aging.204292\u003c/span\u003e\u003cspan address=\"10.18632/aging.204292\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSoo Hoo WI, Higa K, McCormick AA (2023) Vaccination against Epstein-Barr latent membrane protein 1 protects against an Epstein-Barr virus-associated B cell model of lymphoma. Biology (Basel) 12:983. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/biology12070983\u003c/span\u003e\u003cspan address=\"10.3390/biology12070983\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXie Y, Wang A, Lin J, Wu L, Zhang H, Yang X et al (2017) Mps1/TTK: a novel target and biomarker for cancer. J Drug Target 25:112\u0026ndash;118. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/1061186X.2016.1258568\u003c/span\u003e\u003cspan address=\"10.1080/1061186X.2016.1258568\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGuan X, Guo H, Guo Y, Han Q, Li Z, Zhang C (2024) Perforin 1 in cancer: mechanisms, therapy, and outlook. Biomolecules 14:910. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/biom14080910\u003c/span\u003e\u003cspan address=\"10.3390/biom14080910\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePozzi S, Satchi-Fainaro R (2024) The role of CCL2/CCR2 axis in cancer and inflammation: the next frontier in nanomedicine. Adv Drug Deliv Rev 209:115318. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.addr.2024.115318\u003c/span\u003e\u003cspan address=\"10.1016/j.addr.2024.115318\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMukaida N, Sasaki SI, Baba T (2020) CCL4 signaling in the tumor microenvironment. Adv Exp Med Biol 1231:23\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-030-36667-4_3\u003c/span\u003e\u003cspan address=\"10.1007/978-3-030-36667-4_3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKorbecki J, Kojder K, Simińska D, Bohatyrewicz R, Gutowska I, Chlubek D, Baranowska-Bosiacka I (2020) CC chemokines in a tumor: a review of pro-cancer and anti-cancer Properties of the Ligands of Receptors CCR1, CCR2, CCR3, and CCR4. Int J Mol Sci 21:8412. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ijms21218412\u003c/span\u003e\u003cspan address=\"10.3390/ijms21218412\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eScheuermann RH, Racila E (1995) CD19 antigen in leukemia and lymphoma diagnosis and immunotherapy. Leuk Lymphoma 18:385\u0026ndash;397. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3109/10428199509059636\u003c/span\u003e\u003cspan address=\"10.3109/10428199509059636\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"acute lymphoblastic leukemia, Epstein–Barr virus, immune microenvironment","lastPublishedDoi":"10.21203/rs.3.rs-7725271/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7725271/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eEpstein\u0026ndash;Barr virus (EBV) is associated with lymphoid malignancies; however, its contribution to acute lymphoblastic leukemia (ALL) is unclear. Therefore, we investigated the associations between EBV-related transcriptional activity and immune remodeling in ALL across publicly available cohorts.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eThis study integrated transcriptomics, weighted gene co-expression network analysis, protein\u0026ndash;protein interaction network analyses, and single-cell RNA sequencing (scRNA-seq).\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eWe identified 401 EBV-related differentially expressed genes and constructed a prognostic model comprising the following nine critical immune-related genes: interleukin (IL)-18, Toll-like receptor 1, perforin 1 [\u003cem\u003ePRF1\u003c/em\u003e], IL-6 receptor, C-C motif chemokine ligand 2 [\u003cem\u003eCCL2\u003c/em\u003e], TTK protein kinase [\u003cem\u003eTTK\u003c/em\u003e], CD19 molecule [\u003cem\u003eCD19\u003c/em\u003e], cathepsin S, and C-C motif chemokine ligand 4 [\u003cem\u003eCCL4\u003c/em\u003e]. Our model robustly stratified patients into high- and low-risk groups. The high-risk group exhibited significantly poorer survival than did the low-risk group (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.015). External validation confirmed the predictive accuracy of the model (area under the curve values: 0.82, 0.68, and 0.68 for 1-, 3-, and 5-year survival, respectively). scRNA-seq further revealed distinct expression patterns of the nine prognostic genes across immune cell subsets: \u003cem\u003eTTK\u003c/em\u003e was enriched in B and T cells; \u003cem\u003ePRF1\u003c/em\u003e was predominantly expressed in T and natural killer cells; and \u003cem\u003eCCL2, CCL4\u003c/em\u003e, and \u003cem\u003eCD19\u003c/em\u003e were highly expressed in monocytes and B cells.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e\u003cp\u003eAssociations between these genes, immune composition/proliferation signals, and survival in ALL were highlighted. These findings are hypothesis-generating and may reflect EBV-related transcriptional activity as well as lineage and immune-infiltration states; causal roles require validation in clinically EBV-characterized cohorts.\u003c/p\u003e","manuscriptTitle":"Immune dysregulation and prognostic signatures associated with Epstein–Barr virus in acute lymphoblastic leukemia: an integrated transcriptomic and single-cell analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-22 19:23:21","doi":"10.21203/rs.3.rs-7725271/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"37933c69-ef39-41db-9a67-4feccb52acb3","owner":[],"postedDate":"October 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-17T16:39:11+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-22 19:23:21","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7725271","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7725271","identity":"rs-7725271","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}