Cancer-associated Fibroblasts in Pan-Cancer Drive CXCR2+VNN2+ Neutrophils Reprogramming to Mediate Immunosuppression and Immunotherapy Resistance

Cancer-associated Fibroblasts in Pan-Cancer Drive CXCR2+VNN2+ Neutrophils Reprogramming to Mediate Immunosuppression and Immunotherapy Resistance | 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 Cancer-associated Fibroblasts in Pan-Cancer Drive CXCR2+VNN2+ Neutrophils Reprogramming to Mediate Immunosuppression and Immunotherapy Resistance Zhiyu Guo, Xujia Li, Lingli Huang, Yue Yan, Mengge Gao, Jinsheng Huang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6566788/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 Neutrophils are the most abundant granulocyte population and have important functions such as defense against pathogens. However, they show significant heterogeneity and play more complex roles in tumors. The theory of two-tiered differentiation of neutrophils is insufficient to summarize their phenotypic and functional heterogeneity. Therefore, specific regulatory mechanisms remain to be explored and neutrophil-based therapeutic regimens remain challenging. Here, we generated a single-cell atlas of neutrophils from 462 patients with 21 cancer types, revealing their heterogeneity, with CXCR2+ VNN2+ Neu as the main functional subpopulation exerting immunosuppressive effects. Spatial transcriptomic data from the pan-cancer elucidated that fibroblast regulated the phenotypic shift of CXCR2+ VNN2+ Neu in tumor tissues and enabled it to acquire immunosuppressive functions through receptor ligands, cytokines, and extracellular vesicles, which suggested that the tumor microenvironment component was a key reason for the heterogeneity of the prognostic association between neutrophils and pan-cancer patients. Subsequently, we constructed a gene regulatory network to demonstrate the specific regulatory mechanisms of this subpopulation and confirmed that the relevant transcription factors were closely associated with its immunosuppressive function. The pan-cancer immunotherapy cohort proved that the CXCR2+ VNN2+ Neu phenotypic shift was also an important cause of immunotherapy resistance in patients. We finally constructed a deep learning model named Deepsurv to accurately stratify pan-cancer patients based on the CXCR2+ VNN2+ Neu phenotypic shift gene regulatory network (CVN-GRN) and predict the prognosis of the patients, which achieved the desired results. Neutrophils Fibroblasts Immunosuppression Single-cell transcriptome Pan-Cancer Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Neutrophils, the most abundant granulocytes in circulation, play a crucial role in host defense against pathogens and inflammatory responses. Beyond their well-characterized antimicrobial functions, emerging evidence indicates their involvement in immune regulation, tissue repair, and tumor progression [ 1 , 2 ] . Although typically short-lived, neutrophils exhibit prolonged survival in tumor microenvironments, where they display significant phenotypic and functional heterogeneity [ 3 ] . Tumor-associated neutrophils (TANs) exhibit two primary phenotypes: the anti-tumoral N1 and pro-tumoral N2 subsets. Their polarization depends on cytokine milieu, with TGF-β driving the N2 phenotype and IFN-I promoting N1 differentiation [ 4 , 5 ] . While high TAN infiltration typically predicts poor prognosis, it correlates with improved outcomes in colon adenocarcinoma (COAD), indicating that the N1/N2 paradigm inadequately represents TAN heterogeneity [ 6 – 8 ] . Recent studies have identified neutrophil-intrinsic factors influencing TAN phenotypes, including developmental stage and metabolic reprogramming [ 9 – 11 ] . The tumor microenvironment (TME) - comprising stromal cells, vasculature, immune components, and extracellular matrix - critically regulates tumor progression [ 12 ] . However, the mechanisms by which the TME modulates TAN evolution, plasticity and functional heterogeneity remain poorly understood [ 13 ] . Current neutrophil-targeting therapies face significant limitations, including poor specificity and transient efficacy. These approaches may indiscriminately affect anti-tumor neutrophils and other immune cells, compromising immune homeostasis and hindering therapeutic development [ 14 ] . Precise targeting requires identification of TME components governing neutrophil behavior and discovery of specific immunosuppressive neutrophil biomarkers, which are critical for developing effective neutrophil-directed strategies and advancing personalized oncology. Result Construction of a single-cell atlas of pan-cancer neutrophils To establish a pan-cancer neutrophil atlas, we analyzed 299 tumor and 182 normal tissue samples from 462 patients across 21 cancer types ( Figure 1A and Table S1 ). After preprocessing, we identified 19,5542 neutrophils from 2,364,733 total cells using established marker genes ( Table S2-3 ) [15-24] . Neutrophil distribution varied significantly by cancer type, with highest abundance in GBC, CESC, and NSCLC tumors ( Figure 1B and Figure S1A ). Comparative analysis revealed tumor-specific neutrophil enrichment in BLCA, HNSCC, and COAD, while CESC and BRCA showed predominant normal tissue localization ( Figure 1C ). To characterize neutrophil phenotypes and functions, we identified 8 distinct subpopulations following Harmony batch correction ( Figure 1D ) [25] , with defining marker genes shown in Figure 1E ( Table S4 ). These subpopulations demonstrated differential distribution patterns, with most enriched in tumor versus normal tissues across cancer types ( Figure 1F and Figure S1B ). Functional analysis revealed tumor-enriched CXCR2+VNN2+ Neu subpopulations were associated with immunosuppression and fibroblast proliferation, while APOE+C1QB+ Neu and IGKC+HLA-DPB1+ Neu subpopulations showed immune activation and antigen presentation functions ( Figure 1G and Figure S1C ). These results demonstrate cancer-type specific neutrophil distribution patterns with distinct functional specializations. Evolution, Functional Transition, and Survival Correlation of Pan-Carncer Neutrophil Subpopulations Our analyses demonstrate that neutrophils undergo marked phenotypic and functional transitions in tumor tissues, modulating both immune processes and stromal components. To delineate their evolutionary trajectory, we applied five computational approaches (CytoTrace2, Slingshot, Monocle2, Vector, scTour), revealing SPRR1B+ S100A7+ Neu as the progenitor population and TIMP1+ EREG+ Neu/CXCR2+ VNN2+ Neu as terminal states ( Figure 2A and Figure S1D-E ) [26-29] . Functional characterization of neutrophil subpopulations revealed progressive increases in Neutrophil Aging Scores during differentiation, with accelerated aging following the APOE+ C1QB+ Neu and IGKC+ HLA-DPB1+ Neu stages. These mature subsets exhibited peak Antigen Presentation, Phagocytosis, and Azurophilic Granule Scores before subsequent decline - a phenomenon more pronounced in tumor versus normal tissues ( Figure 2B and Figure S2A ). Our integrated analyses identify APOE+ C1QB+ Neu and IGKC+ HLA-DPB1+ Neu as terminally differentiated neutrophils with antigen presenting capacity [30] , while TIMP1+ EREG+ Neu and CXCR2+ VNN2+ Neu represent senescent subsets that acquire immunosuppressive properties in tumors through accelerated aging. The senescent neutrophil subsets TIMP1+ EREG+ Neu and CXCR2+ VNN2+ Neu acquire enhanced immunosuppressive functions through tumor microenvironment-mediated senescence acceleration. These subsets, previously implicated in PAAD and NSCLC progression through TIMP1 and CXCR2 overexpression [31-33] , exhibited elevated expression of co-inhibitory molecules (particularly CD274) compared to other neutrophil populations. Correlation analyses revealed their strong association with T cell exhaustion ( Figure 2C-D and Figure S1F-G ). Pathway enrichment confirmed significant involvement in immunosuppressive mechanisms within tumor tissues ( Figure 2E-F ). The senescent neutrophil subsets exhibited distinct cancer-type distributions: CXCR2+ VNN2+ Neu predominated in digestive and breast cancers, while TIMP1+ EREG+ Neu was enriched in GBM, NSCLC, and RCC ( Figure 2G ). Survival analysis demonstrated that these subsets correlated with poor prognosis in most malignancies ( Figure 2H-I, Figure S3A-B, Figure S4A-B and Figure S5A ). Notably, in COAD and SKCM, these populations showed paradoxical associations with favorable outcomes ( Figure S3C, Figure S4C and Table S5 ). Collectively, these findings demonstrate that the phenotypic shift of TIMP1+ EREG+ Neu and CXCR2+ VNN2+ Neu subsets toward immunosuppressive states in tumor tissues correlates with adverse clinical outcomes, suggesting tumor microenvironmental factors critically regulate this functional transformation. CAFs interact with CXCR2+ VNN2+ Neu via surface receptor ligands, secreted chemokines and extracellular vesicles To identify tumor microenvironment (TME) factors driving neutrophil phenotypic transitions, we analyzed spatial transcriptomic data ( Table S6 ). CellTrek deconvolution [34] revealed significant CXCR2+ VNN2+ Neu-fibroblast interactions in LIHC, NSCLC, HNSCC, OV and PAAD, but not in COAD, SKCM and normal tissues. This spatial specificity may explain the favorable prognosis associated with neutrophils in COAD and SKCM, where CXCR2+ VNN2+ Neu likely maintains its original phenotype. These findings strongly implicate TME components in regulating both neutrophil phenotypic plasticity and clinical outcomes. Notably, TIMP1+ EREG+ Neu and CXCR2+ VNN2+ Neu demonstrated co-localization patterns, suggesting fibroblasts may critically mediate CXCR2+ VNN2+ Neu phenotypic shifts ( Figure 3A and Figure S6A-B ). To validate this hypothesis, we characterized cancer-associated fibroblast (CAFs) subpopulations across multiple tumor types. MMP11+ Fibro, RGS5+ Fibro, and CCL4+ Fibro were significantly enriched in tumor tissues, while CFD+ Fibro predominated in normal tissues, consistent with prior reports implicating these subsets in HCC, PAAD, and NSCLC progression ( Figure 3B ) [35-38] . Functional enrichment analysis revealed that MMP11+ Fibro, RGS5+ Fibro, CCL4+ Fibro, and RPL38+ Fibro subsets exhibited both immunosuppressive properties and neutrophil chemotactic activity in tumors ( Figure S7A ). Figure 3C-D demonstrate the distribution patterns and marker genes of CAFs subpopulations across various cancers ( Table S7 ). Further investigation of CAFs-CXCR2+ VNN2+ Neu interactions revealed that tumor-associated MMP11+ Fibro engages CXCR2+ VNN2+ Neu through ANXA1/COL1A1/FN1-CD44 binding and secretes CXCL3/CXCL6 chemokines, consistent with known roles of CD44+ cells in gastrointestinal cancers ( Figure 3E-G and Figure S7B-C ) [39, 40] . Additionally, RGS5+ Fibro, RPL38+ Fibro and CCL4+ Fibro exhibited tumor specific extracellular vesicles mediated communication with CXCR2+ VNN2+ Neu ( Figure 3H and Figure S7D ). Metabolic flux analysis demonstrated limited metabolite exchange between CAFs and CXCR2+ VNN2+ Neu in both tumor and normal microenvironments ( Figure 3I and Figure S7E-G ). Our integrated findings indicate that CAFs subpopulations primarily interact with CXCR2+ VNN2+ Neu through three mechanisms: (1) direct receptor-ligand binding, (2) chemokine signaling, and (3) extracellular vesicle transfer, potentially driving their immunosuppressive phenotypic conversion. CAFs in Pan-cancer TME Promote CXCR2+ VNN2+ Neu Phenotype Switching and Mediate its Immunosuppressive Function To determine whether CAF subpopulations enhance CXCR2+ VNN2+ Neu immunosuppression through these mechanisms, we examined MMP11+ Fibro mediated gene regulation. MMP11+ Fibro significantly modulated 38 target genes in CXCR2+ VNN2+ Neu via ligand-chemokine signaling ( Figure 4A ), including upregulated BHLHE40/CCL3 and downregulated DDIT4/EHD1 ( Figure 8A and Figure 9A ). Functional enrichment revealed upregulated genes were associated with immune suppression and fibroblast proliferation, while downregulated genes correlated with granulocyte activation and immune response pathways. MMP11+ Fibro bidirectionally regulates CXCR2+ VNN2+ Neu target genes to promote immunosuppression ( Figure 4B ). Extracellular vesicles mediated regulation by RGS5+ Fibro, RPL38+ Fibro, and CCL4+ Fibro through miR-24-3p, miR-145-3p upregulated RPS24, EEF1A1 and other genes expression in tumors ( Figure S10A ). These targets showed enrichment for cytoplasmic translation and immune suppression pathways, with weaker activity in normal tissues ( Figure 4C-D and Figure S10B ). SCENIC analysis [41] identified tumor-enriched transcription factors (TFAP2A, SPIB) regulating this network ( Figure 4E-F ), confirmed at protein level across multiple cancers (HSCLC, GBM, PCC, LIHC, PAAD, OSCC) ( Figure 4G ). These findings demonstrate that TME CAFs coordinately bi-directionally regulate CXCR2+ VNN2+ Neu through multiple mechanisms to drive immunosuppressive phenotypic conversion. Genes encoding phenotype switching-related transcription factors in CXCR2+ VNN2+ Neu promote their immunosuppressive function To investigate the transcriptional regulation of CXCR2+ VNN2+ Neu immunosuppression, we analyzed 10 CRISPR datasets from five studies, identifying Immune Resistant and Immune Sensitive gene sets in colorectal cancer and melanoma ( Table S8-9 ). The top 40 genes from each category are shown in Figure 5A-B . Analysis of transcription factor Z-scores in the CXCR2+ VNN2+ Neu regulatory network revealed HLTF, GTF2IRD1, BACH1, and ATF2 as key mediators of tumor immunosuppression ( Figure 5C ). Intersection analysis demonstrated that all identified CRISPR transcription factors except HLTF were expressed in CXCR2+VNN2+Neu ( Figure 5D ). Virtual knockdown of GTF2IRD1, BACH1, and ATF2 across 10 cancer types (including PAAD) significantly downregulated immunosuppression related pathways (GO:0001915, GO:0002698, GO:0002683), with pathway associated gene counts shown in Figure 5E and expression changes in Figure 5F and Figure S11A-C . These findings establish CXCR2+ VNN2+ Neu phenotype associated transcription factors as critical regulators of its immunosuppressive function. CAFs promote CXCR2+ VNN2+ Neu phenotypic shift and immunosuppressive function leading to immunotherapy resistance To assess CAFs mediated CXCR2+ VNN2+ Neu modulation in immunotherapy response, we analyzed seven immunotherapy cohorts (BCC, COAD, LIHC, NSCLC, RCC, SCC, SKCM; Table S10 ). Following data integration, batch correction, and quality control, we identified and clustered neutrophils into eight distinct subpopulations based on marker genes ( Figure 6A-B and Table S11 ). The C1_Neu subset showed significant enrichment in non-responders across cancer types ( Figure 6C and Figure S12A-B ), revealing a potential association with treatment resistance. AddModuleScore analysis revealed significantly higher CXCR2+ VNN2+ Neu scores in C1_Neu and C0_Neu compared to other subpopulations, confirming their phenotypic similarity ( Figure 6D ). Functional enrichment demonstrated distinct pathway activation patterns: non-responders showed C1_Neu enrichment for ATP biosynthesis and immune suppression, while responders exhibited membrane biogenesis and leukocyte cytotoxicity pathways. Parallel findings in C0_Neu suggest CXCR2+ VNN2+ Neu undergoes comparable phenotypic switching in treatment resistant patients ( Figure 6E and Figure S12C ). Building on these observations, we investigated fibroblast involvement by clustering them into eight distinct subpopulations ( Figure 6F and Figure S12D ). Analysis revealed C1_Fibro enrichment in non-responders ( Figure 6G and Figure S12E-F ), with cellular communication studies demonstrating robust C1_Fibro interactions with C0_Neu/C1_Neu through ANXA1/COL1A1 ligand-receptor pairs and CXCL9/CXCL2 chemokines in non-responders ( Figure 6H and Figure 6J ). These interactions were attenuated in responders ( Figure S12G ), while extracellular vesicles mediated communication primarily occurred between neutrophils ( Figure 6I and Figure S12H ). Metabolite exchange analysis showed no differential fibroblasts-neutrophils interactions between response groups ( Figure S13A-D ). These interactions mediated the upregulation of BHLHE40 and CCL4 in CXCR2+VNN2+Neu, enhancing its immunosuppressive functions including inhibition of T cell cytotoxicity and immune system processes ( Figure 6K-L ). Collectively, our findings demonstrate that pan-cancer CAFs drive CXCR2+ VNN2+ Neu phenotypic conversion and immunosuppression primarily through ligand-chemokine signaling, ultimately contributing to patient immunotherapy resistance. Deepsurv deep learning model accurately predicts prognosis of pan-cancer patients based on CVN-GRN To investigate the prognostic value of the CXCR2+ VNN2+ Neu gene regulatory network (CVN-GRN), we employed 10 machine learning and 2 deep learning models for survival prediction. Using five-fold cross-validation, we randomly partitioned 8,278 TCGA patients across 22 cancer types (including BLCA and BRCA) into training (4/5) and validation (1/5) sets. Model performance was evaluated using Harrell's C-index, Begg's C-index, Uno's C-index, GH C-index, and time-dependent AUC values ( Figure 7A ). The deepsurv model demonstrated superior predictive performance, with five-fold cross-validation results detailed in Table S12 . External validation across seven independent cohorts confirmed its robust generalizability ( Figure 7B and Table S13 ). Survival analysis revealed significantly worse outcomes for high CVN-GRN score patients in both training and test sets ( Figure 7C-D ). The model architecture, comprising an input layer, three hidden layers, and an output layer optimized through backpropagation, is shown in Figure 7E . Consistent results were observed across all external validation cohorts ( Figure 7F and Figure S13E ), establishing deepsurv as an effective tool for CVN-GRN-based prognostic stratification and potential treatment guidance in pan-cancer patients. Materials and Methods Data collection Single-cell sequencing data This study utilized single-cell RNA sequencing data obtained from public repositories: Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/), EMBL-EBI (https://www.ebi.ac.uk/), National Genomics Data Center (NGDC, https://ngdc.cncb.ac.cn/omix/), and China National GeneBank DataBase (CNGBdb, https://db.cngb.org/). We analyzed datasets comprising 462 patients across 21 malignancies: basal cell carcinoma (BCC), bladder urothelial carcinoma (BLCA), breast invasive carcinoma (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), gallbladder cancer (GBC), glioblastoma multiforme (GBM), head and neck squamous cell carcinoma (HNSCC), liver hepatocellular carcinoma (LIHC), non-small-cell lung carcinoma (NSCLC), oral squamous cell carcinoma (OSCC), ovarian serous cystadenocarcinoma (OV), pancreatic adenocarcinoma (PAAD), prostate adenocarcinoma (PRAD), renal cell carcinoma (RCC), squamous cell carcinoma (SCC), skin cutaneous melanoma (SKCM), stomach adenocarcinoma (STAD), and thyroid carcinoma (THCA). Complete dataset details are provided in Table S1 . Spatial transcriptome data Spatial transcriptomic data were obtained from seven cancer types across multiple repositories: HRA000437 (LIHC/normal liver) and GSE203612 (PAAD/OV) from GEO; E-MTAB-13530 (NSCLC/normal lung) from EMBL-EBI; and COAD data from the Cancer Diversity Asia portal ((http://www.cancerdiversity.asia/scCRLM/)). Bulk transcriptome data Bulk transcriptomic and clinical data were sourced from TCGA (8,278 patients across 22 cancers) and six validation cohorts: CGGA (301/325/693 for GBM), METABRIC (BRCA), PRAD-SU-2019 (PRAD), GSE13507 (BLCA), and E-MTAB-6134 (PAAD), totaling 3,718 additional patients. Proteomics data Proteomic data were sourced from the Clinical Proteomic Tumor Analysis Consortium (CPTAC, https://proteomics.cancer.gov/programs/cptac) database, which includes data for six cancer types: NSCLC, GBM, RCC, LIHC, PAAD, and OSCC. Immunotherapy cohorts Immunotherapy cohorts were obtained from multiple sources: GSE123813 (BCC/SCC), GSE205506 (COAD), GSE207422 (NSCLC), PRJNA705464 (RCC), and GSE120575 (SKCM) from GEO, along with LIHC data from Mendeley Data. CRISPR dataset This study analyzed CRISPR/Cas9 screening data from five established studies (Freeman [42] , Kearney [43] , Manguso [44] , Pan [45] , Pate [46] ), which were reorganized into 10 datasets focusing primarily on COAD and SKCM models. Pan-cancer single cell sequencing data processing and integration Single-cell data processing was conducted using Seurat (v4.3.0.1) with the following workflow: mitochondrial gene filtering (>25% threshold), PCA-based dimensionality reduction, and Harmony (v0.1.1) batch correction. The integrated dataset was generated through UMAP projection and subsequent merging of harmonized Seurat objects for downstream analyses. Trajectory analysis Pan-cancer neutrophils trajectory analysis was performed using five complementary algorithms: CytoTRACE2 (v1.0.0) for potency scoring, Slingshot (v2.8.0) for trajectory reconstruction, Monocle2 (v2.28.0), Vector (gmodels v2.18.1.1), and scTour for independent validation. Computational consistency across all methods confirmed trajectory reliability. Spatial transcriptome data analysis Spatial transcriptomic data processing involved: (1) data import (Read10X/Read10X_h5 for expression matrices, Read10X_Image for spatial coordinates); (2) Seurat-based normalization and clustering; (3) cellular deconvolution using CellTrek (v0.0.94) with paired scRNA-seq as reference; and (4) spatial co-localization analysis via scoloc network construction and visualization. Cell-Communication analysis Cellular communication analysis was performed using CellChat (v2.1.2) for ligand-receptor and chemokine interactions, nichenetr (v2.0.1) for ligand-target gene regulation, miRTalk (v1.0) for extracellular vesicle interactions, and mebocost for metabolite-mediated signaling. Overexpressed interactions were identified and aggregated through CellChat's analytical pipeline. Gene regulatory network analysis Gene regulatory networks were constructed using SCENIC through three computational steps: (1) GRNBoost-based inference of TF-target gene co-expression networks, (2) RcisTarget motif analysis to identify enriched regulatory modules, and (3) AUCell scoring of regulator activity at single-cell resolution. CRISPR data analysis We analyzed 10 previously mentioned CRISPR/Cas9 datasets focusing on COAD and SKCM to evaluate transcription factor-immune response associations. Using logFC of sgRNA reads between CTL-treated and control conditions, we calculated Z-scores for 21,304 genes, where lower values indicated immune sensitivity and higher values corresponded to immune resistance. scTenifoldKnk analysis We employed scTenifoldKnk (v1.0.1) [47] for virtual TF knockout analysis targeting three immune-related pathways (GO:0001915, GO:0002698, GO:0002683) in neutrophils from five cancer types (PAAD, NSCLC, RCC, HNSCC, OV). Single-cell regulatory networks were reconstructed to quantify expression perturbations following virtual knockouts, revealing TF-mediated immune modulation. Construction of the CVN-GRN score We evaluated 12 prognostic models (10 machine learning including Lasso-Cox/MTLR/CoxBoost and 2 deep learning algorithms) on TCGA pan-cancer data (22 tumor types) using five validation metrics (Harrell's/Begg's/Uno's/GH C-indices, time-AUC). Following 50-fold cross-validation, the optimal model (Deepsurv) was used to develop the CVN-GRN score, which effectively stratified patients into prognostic subgroups based on CXCR2+VNN2+Neu transcriptional networks. Statistical analysis R v4.3.1 was applied to conduct all statistical analyses in this study. The Wilcox test was implemented to compare the GSVA scores of marker genes between two different subgroups. The log-rank test was utilized to assess the significance of observed differences in overall survival (OS). Statistical significance was determined by a two tailed p-value less than 0.05, unless explicitly specified otherwise. Disscusion Emerging evidence establishes neutrophils as critical modulators of tumor progression within the tumor microenvironment [ 48 , 49 ] . While the classical N1/N2 paradigm posits antitumor and protumor neutrophil subsets respectively [ 50 , 51 ] , recent studies reveal this dichotomy inadequately captures their functional plasticity [ 52 , 53 ] . The mechanisms governing neutrophil phenotypic switching and its clinical implications remain poorly understood. Our pan-cancer analysis of single-cell data from 462 patients across 21 cancer types identified eight neutrophil subpopulations, six of which (excluding HIST1H4C + STMN1 + Neu and KRT7 + ALDH1A3 + Neu) showed tumor-specific enrichment. Functional characterization revealed distinct activation patterns: TIMP1 + EREG + Neu and CXCR2 + VNN2 + Neu exhibited enhanced cytokine signaling, while APOE + C1QB + Neu and IGKC + HLA-DPB1 + Neu demonstrated immune-activating properties, suggesting antitumor potential. Notably, CXCR2 + VNN2 + Neu acquired immunosuppressive functions including immune system inhibition and lymphocyte activation suppression in tumors. Trajectory analysis using five independent algorithms (CytoTRACE, Slingshot, Monocle2, Vector, scTour) consistently identified SPRR1B + S100A7 + Neu as the progenitor population and TIMP1 + EREG + Neu/CXCR2 + VNN2 + Neu as terminal differentiation states. Comparative functional analysis along the neutrophil differentiation trajectory revealed peak antigen presentation and phagocytic activity in APOE + C1QB + Neu and IGKC + HLA-DPB1 + Neu subsets, followed by progressive decline to minimal levels in CXCR2 + VNN2 + Neu, consistent with their established antitumor roles. Notably, this functional attenuation was markedly accelerated in tumor microenvironments, suggesting TME-mediated promotion of CXCR2 + VNN2 + Neu phenotypic conversion. Subsequent validation demonstrated elevated co-inhibitory molecule expression, strong T-cell exhaustion correlation, and immunosuppressive pathway enrichment in TIMP1 + EREG + Neu and CXCR2 + VNN2 + Neu subsets, with pan-cancer prognostic analysis confirming their association with poor clinical outcomes. Spatial neighborhood analysis revealed fibroblast-neutrophil co-localization across multiple cancer types, except in COAD, SKCM, and normal tissues, potentially explaining the favorable prognosis associated with neutrophils in these malignancies [ 6 ] . Our findings identify CAFs as key regulators of neutrophil plasticity. Tumor-associated MMP11 + Fibro engages CXCR2 + VNN2 + Neu through ANXA1 receptor binding and CCL3 cytokine signaling, while RGS5 + Fibro, CCL4 + Fibro, and RPL38 + Fibro communicates via hsa-miR-24-3p-containing extracellular vesicles. Functional genomic analyses demonstrated these interactions bidirectionally modulate CXCR2 + VNN2 + Neu gene expression, driving phenotypic conversion toward an immunosuppressive state. To elucidate CAFs mediated regulation of CXCR2 + VNN2 + Neu, we constructed its gene regulatory network and identified key transcription factors, subsequently validated by pan-cancer proteomic data showing their tumor-specific overexpression. Analysis of five CRISPR studies revealed these transcription factors mediate cancer cell immune resistance. Virtual knockdown experiments confirmed their functional importance, demonstrating significant downregulation of immunosuppressive pathways (GO:0001915, GO:0002698, GO:0002683) in CXCR2 + VNN2 + Neu, establishing a direct link between these transcriptional regulators and neutrophil immunosuppressive polarization. Analysis of pan-cancer immunotherapy cohorts revealed distinct CXCR2 + VNN2 + Neu phenotypic shifts in non-responders, mediated by CAF subpopulations through ANXA1 receptor ligands and CCL3 chemokine signaling that drive immunosuppressive gene expression programs. Prognostic modeling using 10 machine and 2 deep learning approaches identified Deepsurv as optimal for CVN-GRN-based stratification, with validation across seven independent cohorts confirming its clinical utility for outcome prediction and therapeutic guidance. Our study has several limitations. First, the CRISPR analysis was restricted to COAD and SKCM, leaving the association between CXCR2 + VNN2 + Neu transcriptional regulators and immunosuppression unvalidated in other cancer types. Second, the findings require further experimental validation through mechanistic studies. Conclusion Our study establishes a comprehensive pan-cancer neutrophils atlas that delineates phenotypic and functional heterogeneity, uncovering key mechanisms of neutrophil plasticity. We demonstrate that specific fibroblasts subpopulations critically regulate neutrophils phenotypic conversion, which directly contributes to immunotherapy resistance. Furthermore, we developed the CVN-GRN scoring system based on CXCR2 + VNN2 + Neu transcriptional networks and validated its clinical utility through deep learning-based prognostic stratification, providing a valuable framework for therapeutic decision-making in pan-cancer patients. Declarations Authors’ Contributions Conceptualization, Zhiyu Guo and Jinsheng Huang; Investigation, Zhiyu Guo, Xujia Li and Lingli Huang; Software, Zhiyu Guo; Formal Analysis, Zhiyu Guo; Writing—original draft, Xujia Li and Lingli Huang; Writing—review & editing, Mengge Gao and Jinsheng Huang; Supervision, Jinsheng Huang; Visualization, Zhiyu Guo and Xujia Li; Funding acquisition, Jinsheng Huang and Yue Yan. Availability of data and materials All data generated in this study are included in this published article and its supplementary information ( Table S1 ). Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. Ethical approval and consent to participate The data utilized in this study were sourced from publicly accessible databases and were managed under approved ethical exemptions. Consent for publication The authors agreed to publication in the journal. Competing interests The authors declare that they have no competing interests. Acknowledgments Not applicable. References Zhang F, Xia Y, Su J, Quan F, Zhou H, Li Q, et al. Neutrophil diversity and function in health and disease . Signal transduction and targeted therapy 2024; 9(1):343. Huang X, Nepovimova E, Adam V, Sivak L, Heger Z, Valko M, et al. Neutrophils in Cancer immunotherapy: friends or foes? Molecular cancer 2024; 23(1):107. He W, Yan L, Hu D, Hao J, Liou YC, Luo G. Neutrophil heterogeneity and plasticity: unveiling the multifaceted roles in health and disease. MedComm 2025; 6(2):e70063. Fridlender ZG, Sun J, Kim S, Kapoor V, Cheng G, Ling L, et al. Polarization of tumor-associated neutrophil phenotype by TGF-beta: "N1" versus "N2" TAN . Cancer cell 2009; 16(3):183-194. Andzinski L, Kasnitz N, Stahnke S, Wu CF, Gereke M, von Köckritz-Blickwede M, et al. Type I IFNs induce anti-tumor polarization of tumor associated neutrophils in mice and human . International journal of cancer 2016; 138(8):1982-1993. Zhao JJ, Pan K, Wang W, Chen JG, Wu YH, Lv L, et al. The prognostic value of tumor-infiltrating neutrophils in gastric adenocarcinoma after resection. PloS one 2012; 7(3):e33655. Sconocchia G, Zlobec I, Lugli A, Calabrese D, Iezzi G, Karamitopoulou E, et al. Tumor infiltration by FcγRIII (CD16)+ myeloid cells is associated with improved survival in patients with colorectal carcinoma . International journal of cancer 2011; 128(11):2663-2672. Takakura K, Ito Z, Suka M, Kanai T, Matsumoto Y, Odahara S, et al. Comprehensive assessment of the prognosis of pancreatic cancer: peripheral blood neutrophil-lymphocyte ratio and immunohistochemical analyses of the tumour site. Scandinavian journal of gastroenterology 2016; 51(5):610-617. Bodac A, Meylan E. Neutrophil metabolism in the cancer context . Seminars in immunology 2021; 57:101583. Singhal S, Bhojnagarwala PS, O'Brien S, Moon EK, Garfall AL, Rao AS, et al. Origin and Role of a Subset of Tumor-Associated Neutrophils with Antigen-Presenting Cell Features in Early-Stage Human Lung Cancer. Cancer cell 2016; 30(1):120-135. Perez C, Botta C, Zabaleta A, Puig N, Cedena MT, Goicoechea I, et al. Immunogenomic identification and characterization of granulocytic myeloid-derived suppressor cells in multiple myeloma. Blood 2020; 136(2):199-209. Jin MZ, Jin WL. The updated landscape of tumor microenvironment and drug repurposing . Signal transduction and targeted therapy 2020; 5(1):166. Yu X, Li C, Wang Z, Xu Y, Shao S, Shao F, et al. Neutrophils in cancer: dual roles through intercellular interactions . Oncogene 2024; 43(16):1163-1177. Zhang Y, Guoqiang L, Sun M, Lu X. Targeting and exploitation of tumor-associated neutrophils to enhance immunotherapy and drug delivery for cancer treatment. Cancer biology & medicine 2020; 17(1):32-43. Salcher S, Sturm G, Horvath L, Untergasser G, Kuempers C, Fotakis G, et al. High-resolution single-cell atlas reveals diversity and plasticity of tissue-resident neutrophils in non-small cell lung cancer. Cancer cell 2022; 40(12):1503-1520.e1508. Zhao Y, Liu Z, Liu G, Zhang Y, Liu S, Gan D, et al. Neutrophils resist ferroptosis and promote breast cancer metastasis through aconitate decarboxylase 1 . Cell metabolism 2023; 35(10):1688-1703.e1610. Meng Y, Ye F, Nie P, Zhao Q, An L, Wang W, et al. Immunosuppressive CD10(+)ALPL(+) neutrophils promote resistance to anti-PD-1 therapy in HCC by mediating irreversible exhaustion of T cells. Journal of hepatology 2023; 79(6):1435-1449. Xie Y, Zhou T, Li X, Zhao K, Bai W, Hou X, et al. Targeting ESE3/EHF With Nifurtimox Inhibits CXCR2(+) Neutrophil Infiltration and Overcomes Pancreatic Cancer Resistance to Chemotherapy and Immunotherapy. Gastroenterology 2024; 167(2):281-297. Chen Z, Soni N, Pinero G, Giotti B, Eddins DJ, Lindblad KE, et al. Monocyte depletion enhances neutrophil influx and proneural to mesenchymal transition in glioblastoma . Nature communications 2023; 14(1):1839. Hu S, Lu H, Xie W, Wang D, Shan Z, Xing X, et al. TDO2+ myofibroblasts mediate immune suppression in malignant transformation of squamous cell carcinoma . The Journal of clinical investigation 2022; 132(19). Jin X, Wang Q, Luo F, Pan J, Lu T, Zhao Y, et al. Single-cell transcriptomic analysis of tumor heterogeneity and intercellular networks in human urothelial carcinoma. Chinese medical journal 2023; 136(6):690-706. Ma R, Zhou X, Zhai X, Wang C, Hu R, Chen Y, et al. Single-cell RNA sequencing reveals immune cell dysfunction in the peripheral blood of patients with highly aggressive gastric cancer. Cell proliferation 2024; 57(5):e13591. Hu J, Chen Z, Bao L, Zhou L, Hou Y, Liu L, et al. Single-Cell Transcriptome Analysis Reveals Intratumoral Heterogeneity in ccRCC, which Results in Different Clinical Outcomes . Molecular therapy : the journal of the American Society of Gene Therapy 2020; 28(7):1658-1672. Wang H, Xiong A, Chen X, Guo J, Tang Z, Wu C, et al. CXCR1(+) neutrophil infiltration orchestrates response to third-generation EGFR-TKI in EGFR mutant non-small-cell lung cancer. Signal transduction and targeted therapy 2024; 9(1):342. Korsunsky I, Millard N, Fan J, Slowikowski K, Zhang F, Wei K, et al. Fast, sensitive and accurate integration of single-cell data with Harmony . Nature methods 2019; 16(12):1289-1296. Kang M, Armenteros JJA, Gulati GS, Gleyzer R, Avagyan S, Brown EL, et al. Mapping single-cell developmental potential in health and disease with interpretable deep learning . bioRxiv : the preprint server for biology 2024. Street K, Risso D, Fletcher RB, Das D, Ngai J, Yosef N, et al. Slingshot: cell lineage and pseudotime inference for single-cell transcriptomics . BMC genomics 2018; 19(1):477. Qiu X, Mao Q, Tang Y, Wang L, Chawla R, Pliner HA, et al. Reversed graph embedding resolves complex single-cell trajectories . Nature methods 2017; 14(10):979-982. Li Q. scTour: a deep learning architecture for robust inference and accurate prediction of cellular dynamics. Genome biology 2023; 24(1):149. Wu Y, Ma J, Yang X, Nan F, Zhang T, Ji S, et al. Neutrophil profiling illuminates anti-tumor antigen-presenting potency . Cell 2024; 187(6):1422-1439.e1424. Schoeps B, Eckfeld C, Prokopchuk O, Böttcher J, Häußler D, Steiger K, et al. TIMP1 Triggers Neutrophil Extracellular Trap Formation in Pancreatic Cancer . Cancer research 2021; 81(13):3568-3579. Steele CW, Karim SA, Leach JDG, Bailey P, Upstill-Goddard R, Rishi L, et al. CXCR2 Inhibition Profoundly Suppresses Metastases and Augments Immunotherapy in Pancreatic Ductal Adenocarcinoma . Cancer cell 2016; 29(6):832-845. Cheng Y, Mo F, Li Q, Han X, Shi H, Chen S, et al. Targeting CXCR2 inhibits the progression of lung cancer and promotes therapeutic effect of cisplatin. Molecular cancer 2021; 20(1):62. Wei R, He S, Bai S, Sei E, Hu M, Thompson A, et al. Spatial charting of single-cell transcriptomes in tissues . Nature biotechnology 2022; 40(8):1190-1199. Liu Y, Dong G, Yu J, Liang P. Integration of single-cell and spatial transcriptomics reveals fibroblast subtypes in hepatocellular carcinoma: spatial distribution, differentiation trajectories, and therapeutic potential. Journal of translational medicine 2025; 23(1):198. Wang Z, Guo X, Li X, Wang J, Zhang N, Amin B, et al. Cancer-associated fibroblast-derived MMP11 promotes tumor progression in pancreatic cancer . Cancer science 2025; 116(3):643-655. Forsthuber A, Aschenbrenner B, Korosec A, Jacob T, Annusver K, Krajic N, et al. Cancer-associated fibroblast subtypes modulate the tumor-immune microenvironment and are associated with skin cancer malignancy. Nature communications 2024; 15(1):9678. Koppensteiner L, Mathieson L, Neilson L, O'Connor RA, Akram AR. IFNγ and TNFα drive an inflammatory secretion profile in cancer-associated fibroblasts from human non-small cell lung cancer. FEBS letters 2025; 599(5):713-723. Garay J, Piazuelo MB, Majumdar S, Li L, Trillo-Tinoco J, Del Valle L, et al. The homing receptor CD44 is involved in the progression of precancerous gastric lesions in patients infected with Helicobacter pylori and in development of mucous metaplasia in mice. Cancer letters 2016; 371(1):90-98. Tang F, Zhu Y, Shen J, Yuan B, He X, Tian Y, et al. CD44(+) cells enhance pro-tumor stroma in the spatial landscape of colorectal cancer leading edge. British journal of cancer 2025. Aibar S, González-Blas CB, Moerman T, Huynh-Thu VA, Imrichova H, Hulselmans G, et al. SCENIC: single-cell regulatory network inference and clustering . Nature methods 2017; 14(11):1083-1086. Freeman AJ, Vervoort SJ, Ramsbottom KM, Kelly MJ, Michie J, Pijpers L, et al. Natural Killer Cells Suppress T Cell-Associated Tumor Immune Evasion . Cell reports 2019; 28(11):2784-2794.e2785. Kearney CJ, Vervoort SJ, Hogg SJ, Ramsbottom KM, Freeman AJ, Lalaoui N, et al. Tumor immune evasion arises through loss of TNF sensitivity . Science immunology 2018; 3(23). Manguso RT, Pope HW, Zimmer MD, Brown FD, Yates KB, Miller BC, et al. In vivo CRISPR screening identifies Ptpn2 as a cancer immunotherapy target . Nature 2017; 547(7664):413-418. Pan D, Kobayashi A, Jiang P, Ferrari de Andrade L, Tay RE, Luoma AM, et al. A major chromatin regulator determines resistance of tumor cells to T cell-mediated killing . Science (New York, NY) 2018; 359(6377):770-775. Patel SJ, Sanjana NE, Kishton RJ, Eidizadeh A, Vodnala SK, Cam M, et al. Identification of essential genes for cancer immunotherapy . Nature 2017; 548(7669):537-542. Osorio D, Zhong Y, Li G, Xu Q, Yang Y, Tian Y, et al. scTenifoldKnk: An efficient virtual knockout tool for gene function predictions via single-cell gene regulatory network perturbation. Patterns (New York, NY) 2022; 3(3):100434. Xue R, Zhang Q, Cao Q, Kong R, Xiang X, Liu H, et al. Liver tumour immune microenvironment subtypes and neutrophil heterogeneity . Nature 2022; 612(7938):141-147. Shaul ME, Fridlender ZG. Tumour-associated neutrophils in patients with cancer . Nature reviews Clinical oncology 2019; 16(10):601-620. Zhang X, Shi H, Yuan X, Jiang P, Qian H, Xu W. Tumor-derived exosomes induce N2 polarization of neutrophils to promote gastric cancer cell migration. Molecular cancer 2018; 17(1):146. Zhang J, Gu J, Wang X, Ji C, Yu D, Wang M, et al. Engineering and Targeting Neutrophils for Cancer Therapy . Advanced materials (Deerfield Beach, Fla) 2024; 36(19):e2310318. Ng LG, Ostuni R, Hidalgo A. Heterogeneity of neutrophils . Nature reviews Immunology 2019; 19(4):255-265. Xiong S, Dong L, Cheng L. Neutrophils in cancer carcinogenesis and metastasis . Journal of hematology & oncology 2021; 14(1):173. Supplementary Files SupplementaryFigure.docx SupplementaryTable.xlsx 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-6566788","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":456671150,"identity":"c6dd13b5-a5db-452f-9e81-0112b3f2d328","order_by":0,"name":"Zhiyu Guo","email":"","orcid":"","institution":"Sun Yat-sen University Cancer Center","correspondingAuthor":false,"prefix":"","firstName":"Zhiyu","middleName":"","lastName":"Guo","suffix":""},{"id":456671151,"identity":"11fa7434-9fb6-4939-8ada-f1789060f528","order_by":1,"name":"Xujia Li","email":"","orcid":"","institution":"Sun Yat-sen University Cancer Center","correspondingAuthor":false,"prefix":"","firstName":"Xujia","middleName":"","lastName":"Li","suffix":""},{"id":456671152,"identity":"4e4ac56e-060c-462f-89bd-0263b73678a2","order_by":2,"name":"Lingli Huang","email":"","orcid":"","institution":"Zhuzhou Hospital Affiliated to Central South University: Zhuzhou Central Hospital","correspondingAuthor":false,"prefix":"","firstName":"Lingli","middleName":"","lastName":"Huang","suffix":""},{"id":456671153,"identity":"c820d86b-5c92-4bad-a338-9f8f1e31c3be","order_by":3,"name":"Yue Yan","email":"","orcid":"","institution":"Sun Yat-sen University Cancer Center","correspondingAuthor":false,"prefix":"","firstName":"Yue","middleName":"","lastName":"Yan","suffix":""},{"id":456671154,"identity":"31d5b533-d800-40bc-9a1f-a912c88f8268","order_by":4,"name":"Mengge Gao","email":"","orcid":"","institution":"Huadu District People's Hospital of Guangzhou","correspondingAuthor":false,"prefix":"","firstName":"Mengge","middleName":"","lastName":"Gao","suffix":""},{"id":456671155,"identity":"5b47fc34-2c90-45ba-a297-aac4e38844ea","order_by":5,"name":"Jinsheng Huang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+UlEQVRIiWNgGAWjYBACAzDJI8HAIMF8gGQtbAmkaAEBCR4DPOqQgLlE8rOHX2Qs8uRn93z+zPPnsDx/A/Oxj18Y7PJwabGckWZuLMMjUWxw5+w2ad62w4YzDrAlz5ZhSC7G6bAbCWbSEjwSiRskcrcx8zYcTjBg4DFmlmA4kNiAU0v6N7CW+TNyHoMcRoyWHDPJD0AtDTdyGKR52CBaGD/g03LmTZk0A8hhN9LMJOe2pRvOOMyWzMxgkIxby/H0bZI/e+qADkt+/OHNH2t5/vbmw4w/KuxwagEBZt4eCIOJB8wFIkJxxPjjB4zBgMYYBaNgFIyCUQAEAAVwU4tJ6irZAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-5859-4994","institution":"Sun Yat-sen University Cancer Center","correspondingAuthor":true,"prefix":"","firstName":"Jinsheng","middleName":"","lastName":"Huang","suffix":""}],"badges":[],"createdAt":"2025-04-30 16:51:37","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6566788/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6566788/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":84335882,"identity":"2dcfe5f0-9aa3-4f62-94b2-1041f9bbff13","added_by":"auto","created_at":"2025-06-10 17:18:49","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":6969003,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConstruction of a single-cell atlas of pan-cancer neutrophils.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e Workflow of the study. \u003cstrong\u003e(B)\u003c/strong\u003eUMAP graph showing the distribution and percentage of neutrophils in different 21 cancer types, different colors in the left graph indicate different cancer types, and the darker color in the right graph represents the higher number of neutrophils. \u003cstrong\u003e(C)\u003c/strong\u003e The relative abundance of neutrophils in tumor and normal tissues of different cancer types, the red area represents the higher abundance of neutrophils in tumor tissues and the green area represents the higher abundance of neutrophils in normal tissues. \u003cstrong\u003e(D) \u003c/strong\u003eUMAP plot showing the dimensionality reduction clustering of 8 subpopulations of neutrophils in pan-cancer species. \u003cstrong\u003e(E)\u003c/strong\u003eHeat map demonstrating marker genes of neutrophils subpopulations. \u003cstrong\u003e(F)\u003c/strong\u003eDifferential distribution of neutrophils subpopulations in tumor tissues (orange) and normal tissues (blue). \u003cstrong\u003e(G) \u003c/strong\u003eBar graph demonstrating the functional enrichment analysis of neutrophils subpopulations in tumor tissues.\u003c/p\u003e","description":"","filename":"figure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6566788/v1/b467c47875c93efeacc79d04.jpg"},{"id":84335123,"identity":"e415c4b5-8f39-412a-b47e-a6c926162439","added_by":"auto","created_at":"2025-06-10 17:02:49","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":10000952,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEvolution, Functional Conversion, and Survival Correlation of Neutrophil Subpopulations in Pan-Cancer Species\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e CytoTRACE (blue), Slingshot (green), Monocle2 (yellow), Vector (purple) and scTour (red) 5 algorithms demonstrating neutrophils evolutionary trajectories. \u003cstrong\u003e(B)\u003c/strong\u003e Changes in antigen presentation function (red), phagocytosis (blue) and chemotaxis (orange) during neutrophils evolution in tumor tissue (left) and normal tissue (right). \u003cstrong\u003e(C) \u003c/strong\u003eMountain range diagram demonstrating differences in the expression of neutrophils subpopulations CD274, CEACAM1 co-repressor molecules in tumor tissues. \u003cstrong\u003e(D) \u003c/strong\u003eBar graph of the correlation between CXCR2+ VNN2+ Neu and CD8+ effector T-cell depletion in different cancer types, with darker colors representing stronger correlation. \u003cstrong\u003e(E-F)\u003c/strong\u003e Lollipop plots of CXCR2+ VNN2+ Neu \u003cstrong\u003e(E)\u003c/strong\u003e, TIMP1+ EREG+ Neu \u003cstrong\u003e(F)\u003c/strong\u003e functional enrichment in tumor tissue (red) and normal tissue (blue). \u003cstrong\u003e(G)\u003c/strong\u003e CXCR2+ VNN2+ Neu (orange), TIMP1+ EREG+ Neu (blue) distribution preference, red represents high Ro/e index and blue represents low Ro/e index. \u003cstrong\u003e(H)\u003c/strong\u003e TCGA data on the association between the two subgroups of different cancer types CXCR2+ VNN2+ Neu (E), TIMP1+ EREG+ Neu, and patient prognosis, with red representing a poor prognosis (p \u0026lt; 0.05), blue representing a good prognosis (p \u0026lt; 0.05), and gray representing no distribution of cancer types, with the white area in the middle indicating no significant correlation (p \u0026gt; 0.05). \u003cstrong\u003e(I)\u003c/strong\u003eKaplan-Meier survival curves demonstrating the prognostic difference between patients in the CXCR2+ VNN2+ Neu high (red) and low (blue) groups of PAAD, LIHC, GBM, LGG, LUSC and HNSCC species.\u003c/p\u003e","description":"","filename":"figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6566788/v1/6d0c73cee0b974b3a9d768c8.jpg"},{"id":84335884,"identity":"c1c1cdb4-3acf-4608-bf03-d34162519bee","added_by":"auto","created_at":"2025-06-10 17:18:49","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":16265457,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCAFs interact with CXCR2+ VNN2+ Neu via surface receptor ligands, secreted chemokines and extracellular vesicles\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e Spatial transcriptome demonstrating the spatial distribution of neutrophils subpopulations in LIHC, LIVER (normal liver tissue), NSCLC, LUNG (normal lung tissue), HNSCC, OV, COAD, and SKCM, as well as neighboring cell subpopulations. \u003cstrong\u003e(B)\u003c/strong\u003e UMAP plots of the downscaled clustering of different fibroblasts subpopulations in tumor tissues (right) and normal tissues (left) as well as the percentage of each subpopulation (bottom). \u003cstrong\u003e(C)\u003c/strong\u003e Bar graph demonstrating the differences in the distribution of the 10 fibroblasts subpopulations in different cancer types. \u003cstrong\u003e(D) \u003c/strong\u003eBubble map of marker genes for neutrophils subpopulations, with the depth of color representing the level of high or low expression. \u003cstrong\u003e(E)\u003c/strong\u003eCellChat cell communication analysis demonstrating the communication relationship between fibroblasts subpopulations and neutrophils subpopulations in tumor tissues. \u003cstrong\u003e(F-G)\u003c/strong\u003e Crosstalk association mediated by different fibroblasts subpopulations and CXCR2+ VNN2+ Neu receptor ligand \u003cstrong\u003e(F)\u003c/strong\u003e and cytokine \u003cstrong\u003e(G)\u003c/strong\u003e in tumor tissues, darker color indicates stronger crosstalk. \u003cstrong\u003e(H)\u003c/strong\u003e Sankey diagram demonstrating extracellular vesicle-mediated interactions between fibroblasts subpopulations and centriolar granulocyte subpopulations in tumor tissue, with signal senders on the left and signal receivers on the right. \u003cstrong\u003e(I)\u003c/strong\u003e Metabolite-mediated communication between fibroblasts subpopulations and neutrophils subpopulations in tumor tissue, with signal sender, metabolite, sensor, and signal receiver from left to right.\u003c/p\u003e","description":"","filename":"figure3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6566788/v1/103f281cde9736bfb6295eb4.jpg"},{"id":84335114,"identity":"e215c73c-73a0-4f6f-b61a-615c4fa582ff","added_by":"auto","created_at":"2025-06-10 17:02:49","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":4501678,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCAFs in pan-cancer TME promote CXCR2+ VNN2+ Neu phenotype switching and mediate its immunosuppressive function\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e Heatmap demonstrating the regulation of CXCR2+ VNN2+ Neu target genes by MMP11+ Fibro in tumor tissues, with darker red representing stronger regulation. \u003cstrong\u003e(B)\u003c/strong\u003e Histogram of GO enrichment analysis of up-regulated genes (purple) and down-regulated genes (orange) in tumor tissue CXCR2+ VNN2+ Neu. \u003cstrong\u003e(C)\u003c/strong\u003e Regulation of target genes in CXCR2+ VNN2+ Neu by RGS5+ Fibro (green), RPL38+ Fibro (yellow), CCL4+ Fibro (red), and MMP11+ Fibro (blue) via extracellular vesicles. \u003cstrong\u003e(D)\u003c/strong\u003eHistogram of GO enrichment analysis of extracellular vesicle-regulated target genes. \u003cstrong\u003e(E)\u003c/strong\u003e Gene regulatory network of CXCR2+ VNN2+ Neu by fibroblasts subpopulations in tumor tissues, with genes represented in red, associated transcription factors in blue, and extracellular vesicle-associated gene regulatory networks in green boxes. \u003cstrong\u003e(F)\u003c/strong\u003e SCENIC analysis of the rank of CXCR2+ VNN2+ Neu phenotype regulation-related transcription factors (red) in tumor tissues (left) and normal tissues (right). \u003cstrong\u003e(G)\u003c/strong\u003e Cloud rain plot demonstrating the difference in protein expression scores of transcription factors in tumor tissues (red) and normal tissues (green) in six cancer types, NSCLC, GBM, RCC, LIHC, PAAD and OSCC.\u003c/p\u003e","description":"","filename":"figure4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6566788/v1/072803cb4181329239798b00.jpg"},{"id":84336336,"identity":"ddf9fb83-41c2-4cb6-a264-1fc2712a6a93","added_by":"auto","created_at":"2025-06-10 17:26:49","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":6085755,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGenes encoding phenotype switching-related transcription factors in CXCR2+ VNN2+ Neu promote their immunosuppressive function\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A-B)\u003c/strong\u003eHeatmap of immunosuppressive genes \u003cstrong\u003e(A)\u003c/strong\u003e and immunosensitive genes \u003cstrong\u003e(B)\u003c/strong\u003ein the 10 CRISPR datasets, with immunosuppressive genes (red) on the left, the darker the color the lower the Z-score, and immunosensitive genes (blue) on the right, the darker the color the higher the Z-score. \u003cstrong\u003e(C)\u003c/strong\u003e Heatmap demonstrating the Z-score of genes encoding CXCR2+ VNN2+ Neu phenotype switch-related transcription factors in tumor tissues, with red indicating lower Z-score and blue indicating higher Z-score. \u003cstrong\u003e(D)\u003c/strong\u003e Wayne diagram showing the intersection between transcription factor-encoding genes and CXCR2+ VNN2+ Neu genes in the CRISPR dataset, which contains a total of 38 genes. \u003cstrong\u003e(E)\u003c/strong\u003eHistogram of the number of genes related to three immunosuppressive pathways, GO:0001915 (red), GO:0002698 (blue), and GO:0002683 (yellow). \u003cstrong\u003e(F)\u003c/strong\u003eDifferences in the expression of three immunosuppressive pathway-related genes by knockdown of GTF2IRD1, BACH1, and ATF2 in PAAD, NSCLC, and RCC, respectively.\u003c/p\u003e","description":"","filename":"figure5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6566788/v1/0e17c762c7fc18ac1bef7ab8.jpg"},{"id":84335886,"identity":"e9d7b793-5d1d-4ef6-86ff-39b2727c3269","added_by":"auto","created_at":"2025-06-10 17:18:49","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":6021088,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCAFs promote CXCR2+ VNN2+ Neu phenotypic shift and immunosuppressive function leading to immunotherapy resistance\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003eUMAP plot of neutrophils dimensionality reduction clustering in 7 immunotherapy cohorts. \u003cstrong\u003e(B)\u003c/strong\u003e Bubble plots showing 8 neutrophils subpopulations marker genes. \u003cstrong\u003e(C)\u003c/strong\u003e Percentage of neutrophils subpopulations in immunotherapy-responding and immunotherapy-non responding patients. \u003cstrong\u003e(D)\u003c/strong\u003eImmunotherapy cohort neutrophils subpopulation CXCR2+ VNN2+ Neu scores are shown, with orange representing high scores and purple representing low scores. \u003cstrong\u003e(E)\u003c/strong\u003e Lollipop plot showing C1_Neu functional enrichment analysis in immunotherapy non-responding patients (red) and immunotherapy responding patients (blue). \u003cstrong\u003e(F)\u003c/strong\u003e UMAP plot of downscaled clustering of fibroblasts from 7 immunotherapy cohorts. \u003cstrong\u003e(G)\u003c/strong\u003e Bar graph showing the percentage of 8 fibroblasts subpopulations in immunotherapy-responding and immunotherapy-non responding patients. \u003cstrong\u003e(H)\u003c/strong\u003e Bubble plot of ligand pair-mediated cellular communication between fibroblasts subpopulations and C0_Neu and C1_Neu in immunotherapy-non responding patients, with red indicating stronger communication and blue indicating weaker cellular communication. \u003cstrong\u003e(I)\u003c/strong\u003eSankey diagram demonstrating extracellular vesicle-mediated crosstalk between fibroblasts and neutrophils in immunotherapy-non responding patients. \u003cstrong\u003e(J)\u003c/strong\u003eHeat map demonstrating cytokine-mediated crosstalk between fibroblasts subpopulations and neutrophils subpopulations in immunotherapy-non responding patients. \u003cstrong\u003e(K)\u003c/strong\u003e Heatmap of the regulatory relationship between fibroblasts subpopulations on C0_Neu and C1_Neu target genes in immunotherapy-non responding patients. \u003cstrong\u003e(L)\u003c/strong\u003e Bar graph demonstrating functional enrichment analysis of C0_Neu and C1_Neu target genes in immunotherapy non-responding patients.\u003c/p\u003e","description":"","filename":"figure6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6566788/v1/1981b63088284190fe94536e.jpg"},{"id":84335147,"identity":"afe51cd6-32e0-4908-a4f5-cc97a2d6ed14","added_by":"auto","created_at":"2025-06-10 17:02:50","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":5162436,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDeepsurv deep learning model accurately predicts the prognosis of pan-cancer patients based on CVN-GRN\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003eHeatmap demonstrating the efficacy of 10 machine learning algorithms as well as 2 deep learning algorithms for predicting the prognosis of patients in the validation set of pan-cancer, with the algorithms on the vertical axis and evaluation metrics on the horizontal axis, where purple color represents good prediction performance and green color represents weak prediction performance. \u003cstrong\u003e(B)\u003c/strong\u003eHeatmap of Deepsurv model's effectiveness in predicting the prognosis of 7 test set cohorts (including GBM, PRAD, BRAC, PAAD, BLCA), vertical axis is the algorithms, horizontal axis is the evaluation indexes, purple color represents good prediction performance, and green color represents weaker prediction performance. \u003cstrong\u003e(C-D) \u003c/strong\u003eKaplan-Meier survival curves demonstrating the difference in prognosis between the Deepsurv model for the high scoring group (red) and the low scoring group (blue) in the training and validation sets. \u003cstrong\u003e(E)\u003c/strong\u003eDeepsurv model structure demonstration, green represents the input layer, blue represents the hidden layers red represents the output layer. \u003cstrong\u003e(F)\u003c/strong\u003e 6 external cohorts containing GBM, PAAD, BRCA, and BLCA constitute the test sets, with prognostic Kaplan-Meier survival curves for patients in the high scoring group (red) and patients in the low scoring group (blue).\u003c/p\u003e","description":"","filename":"figure7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6566788/v1/5b5e51bc34ebe3ce0283358e.jpg"},{"id":84336473,"identity":"38acb1e2-100a-4f8a-b85e-bf808a563bc4","added_by":"auto","created_at":"2025-06-10 17:35:14","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":26346004,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6566788/v1/0c66e8b5-925c-487b-8575-ab1faaf9d552.pdf"},{"id":84335768,"identity":"8dbf91e2-8a9c-4e59-99c4-a29bf79aa48a","added_by":"auto","created_at":"2025-06-10 17:10:49","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2872686,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigure.docx","url":"https://assets-eu.researchsquare.com/files/rs-6566788/v1/3367dbab06891ecd5243a7ff.docx"},{"id":84335765,"identity":"23290644-e052-4acb-8d88-c3baea3faad0","added_by":"auto","created_at":"2025-06-10 17:10:49","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2941114,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6566788/v1/34a4d32b4681881113ccf36e.xlsx"}],"financialInterests":"","formattedTitle":"Cancer-associated Fibroblasts in Pan-Cancer Drive CXCR2+VNN2+ Neutrophils Reprogramming to Mediate Immunosuppression and Immunotherapy Resistance","fulltext":[{"header":"Introduction","content":"\u003cp\u003eNeutrophils, the most abundant granulocytes in circulation, play a crucial role in host defense against pathogens and inflammatory responses. Beyond their well-characterized antimicrobial functions, emerging evidence indicates their involvement in immune regulation, tissue repair, and tumor progression\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/sup\u003e. Although typically short-lived, neutrophils exhibit prolonged survival in tumor microenvironments, where they display significant phenotypic and functional heterogeneity\u003csup\u003e[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTumor-associated neutrophils (TANs) exhibit two primary phenotypes: the anti-tumoral N1 and pro-tumoral N2 subsets. Their polarization depends on cytokine milieu, with TGF-β driving the N2 phenotype and IFN-I promoting N1 differentiation\u003csup\u003e[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]\u003c/sup\u003e. While high TAN infiltration typically predicts poor prognosis, it correlates with improved outcomes in colon adenocarcinoma (COAD), indicating that the N1/N2 paradigm inadequately represents TAN heterogeneity\u003csup\u003e[\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]\u003c/sup\u003e. Recent studies have identified neutrophil-intrinsic factors influencing TAN phenotypes, including developmental stage and metabolic reprogramming\u003csup\u003e[\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]\u003c/sup\u003e. The tumor microenvironment (TME) - comprising stromal cells, vasculature, immune components, and extracellular matrix - critically regulates tumor progression\u003csup\u003e[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]\u003c/sup\u003e. However, the mechanisms by which the TME modulates TAN evolution, plasticity and functional heterogeneity remain poorly understood\u003csup\u003e[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eCurrent neutrophil-targeting therapies face significant limitations, including poor specificity and transient efficacy. These approaches may indiscriminately affect anti-tumor neutrophils and other immune cells, compromising immune homeostasis and hindering therapeutic development\u003csup\u003e[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]\u003c/sup\u003e. Precise targeting requires identification of TME components governing neutrophil behavior and discovery of specific immunosuppressive neutrophil biomarkers, which are critical for developing effective neutrophil-directed strategies and advancing personalized oncology.\u003c/p\u003e"},{"header":"Result","content":"\u003cp\u003e\u003cstrong\u003eConstruction of a single-cell atlas of pan-cancer neutrophils\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo establish a pan-cancer neutrophil atlas, we analyzed 299 tumor and 182 normal tissue samples from 462 patients across 21 cancer types (\u003cstrong\u003eFigure 1A\u003c/strong\u003e and \u003cstrong\u003eTable S1\u003c/strong\u003e). After preprocessing, we identified 19,5542 neutrophils from 2,364,733 total cells using established marker genes (\u003cstrong\u003eTable S2-3\u003c/strong\u003e)\u003csup\u003e[15-24]\u003c/sup\u003e. Neutrophil distribution varied significantly by cancer type, with highest abundance in GBC, CESC, and NSCLC tumors (\u003cstrong\u003eFigure 1B\u003c/strong\u003e and \u003cstrong\u003eFigure S1A\u003c/strong\u003e). Comparative analysis revealed tumor-specific neutrophil enrichment in BLCA, HNSCC, and COAD, while CESC and BRCA showed predominant normal tissue localization (\u003cstrong\u003eFigure 1C\u003c/strong\u003e). To characterize neutrophil phenotypes and functions, we identified 8 distinct subpopulations following Harmony batch correction (\u003cstrong\u003eFigure 1D\u003c/strong\u003e)\u003csup\u003e[25]\u003c/sup\u003e, with defining marker genes shown in \u003cstrong\u003eFigure 1E\u0026nbsp;\u003c/strong\u003e(\u003cstrong\u003eTable S4\u003c/strong\u003e). These subpopulations demonstrated differential distribution patterns, with most enriched in tumor versus normal tissues across cancer types (\u003cstrong\u003eFigure 1F\u003c/strong\u003e and \u003cstrong\u003eFigure S1B\u003c/strong\u003e). Functional analysis revealed tumor-enriched CXCR2+VNN2+ Neu subpopulations were associated with immunosuppression and fibroblast proliferation, while APOE+C1QB+ Neu and IGKC+HLA-DPB1+ Neu subpopulations showed immune activation and antigen presentation functions (\u003cstrong\u003eFigure 1G\u003c/strong\u003e and \u003cstrong\u003eFigure S1C\u003c/strong\u003e). These results demonstrate cancer-type specific neutrophil distribution patterns with distinct functional specializations.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEvolution, Functional Transition, and Survival Correlation of Pan-Carncer Neutrophil Subpopulations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur analyses demonstrate that neutrophils undergo marked phenotypic and functional transitions in tumor tissues, modulating both immune processes and stromal components. To delineate their evolutionary trajectory, we applied five computational approaches (CytoTrace2, Slingshot, Monocle2, Vector, scTour), revealing SPRR1B+ S100A7+ Neu as the progenitor population and TIMP1+ EREG+ Neu/CXCR2+ VNN2+ Neu as terminal states (\u003cstrong\u003eFigure 2A\u003c/strong\u003e and \u003cstrong\u003eFigure S1D-E\u003c/strong\u003e)\u003csup\u003e[26-29]\u003c/sup\u003e. Functional characterization of neutrophil subpopulations revealed progressive increases in Neutrophil Aging Scores during differentiation, with accelerated aging following the APOE+ C1QB+ Neu and IGKC+ HLA-DPB1+ Neu stages. These mature subsets exhibited peak Antigen Presentation, Phagocytosis, and Azurophilic Granule Scores before subsequent decline - a phenomenon more pronounced in tumor versus normal tissues (\u003cstrong\u003eFigure 2B\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Figure S2A\u003c/strong\u003e). Our integrated analyses identify APOE+ C1QB+ Neu and IGKC+ HLA-DPB1+ Neu as terminally differentiated neutrophils with antigen presenting capacity\u003csup\u003e[30]\u003c/sup\u003e, while TIMP1+ EREG+ Neu and CXCR2+ VNN2+ Neu represent senescent subsets that acquire immunosuppressive properties in tumors through accelerated aging. The senescent neutrophil subsets TIMP1+ EREG+ Neu and CXCR2+ VNN2+ Neu acquire enhanced immunosuppressive functions through tumor microenvironment-mediated senescence acceleration. These subsets, previously implicated in PAAD and NSCLC progression through TIMP1 and CXCR2 overexpression\u003csup\u003e[31-33]\u003c/sup\u003e, exhibited elevated expression of co-inhibitory molecules (particularly CD274) compared to other neutrophil populations. Correlation analyses revealed their strong association with T cell exhaustion (\u003cstrong\u003eFigure 2C-D\u003c/strong\u003e and \u003cstrong\u003eFigure S1F-G\u003c/strong\u003e). Pathway enrichment confirmed significant involvement in immunosuppressive mechanisms within tumor tissues (\u003cstrong\u003eFigure 2E-F\u003c/strong\u003e). The senescent neutrophil subsets exhibited distinct cancer-type distributions: CXCR2+ VNN2+ Neu predominated in digestive and breast cancers, while TIMP1+ EREG+ Neu was enriched in GBM, NSCLC, and RCC (\u003cstrong\u003eFigure 2G\u003c/strong\u003e). Survival analysis demonstrated that these subsets correlated with poor prognosis in most malignancies (\u003cstrong\u003eFigure 2H-I, Figure S3A-B, Figure S4A-B\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Figure S5A\u003c/strong\u003e). Notably, in COAD and SKCM, these populations showed paradoxical associations with favorable outcomes (\u003cstrong\u003eFigure S3C, Figure S4C\u003c/strong\u003e and \u003cstrong\u003eTable S5\u003c/strong\u003e). Collectively, these findings demonstrate that the phenotypic shift of TIMP1+ EREG+ Neu and CXCR2+ VNN2+ Neu subsets toward immunosuppressive states in tumor tissues correlates with adverse clinical outcomes, suggesting tumor microenvironmental factors critically regulate this functional transformation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCAFs interact with CXCR2+ VNN2+ Neu via surface receptor ligands, secreted chemokines and extracellular vesicles\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify tumor microenvironment (TME) factors driving neutrophil phenotypic transitions, we analyzed spatial transcriptomic data (\u003cstrong\u003eTable S6\u003c/strong\u003e). CellTrek deconvolution\u003csup\u003e[34]\u003c/sup\u003e revealed significant CXCR2+ VNN2+ Neu-fibroblast interactions in LIHC, NSCLC, HNSCC, OV and PAAD, but not in COAD, SKCM and normal tissues. This spatial specificity may explain the favorable prognosis associated with neutrophils in COAD and SKCM, where CXCR2+ VNN2+ Neu likely maintains its original phenotype. These findings strongly implicate TME components in regulating both neutrophil phenotypic plasticity and clinical outcomes. Notably, TIMP1+ EREG+ Neu and CXCR2+ VNN2+ Neu demonstrated co-localization patterns, suggesting fibroblasts may critically mediate CXCR2+ VNN2+ Neu phenotypic shifts (\u003cstrong\u003eFigure 3A\u0026nbsp;\u003c/strong\u003eand \u003cstrong\u003eFigure S6A-B\u003c/strong\u003e). To validate this hypothesis, we characterized cancer-associated fibroblast (CAFs) subpopulations across multiple tumor types. MMP11+ Fibro, RGS5+ Fibro, and CCL4+ Fibro were significantly enriched in tumor tissues, while CFD+ Fibro predominated in normal tissues, consistent with prior reports implicating these subsets in HCC, PAAD, and NSCLC progression (\u003cstrong\u003eFigure 3B\u003c/strong\u003e)\u003csup\u003e[35-38]\u003c/sup\u003e. Functional enrichment analysis revealed that MMP11+ Fibro, RGS5+ Fibro, CCL4+ Fibro, and RPL38+ Fibro subsets exhibited both immunosuppressive properties and neutrophil chemotactic activity in tumors (\u003cstrong\u003eFigure S7A\u003c/strong\u003e). \u003cstrong\u003eFigure 3C-D\u003c/strong\u003e demonstrate the distribution patterns and marker genes of CAFs subpopulations across various cancers (\u003cstrong\u003eTable S7\u003c/strong\u003e). Further investigation of CAFs-CXCR2+ VNN2+ Neu interactions revealed that tumor-associated MMP11+ Fibro engages CXCR2+ VNN2+ Neu through ANXA1/COL1A1/FN1-CD44 binding and secretes CXCL3/CXCL6 chemokines, consistent with known roles of CD44+ cells in gastrointestinal cancers (\u003cstrong\u003eFigure 3E-G\u0026nbsp;\u003c/strong\u003eand \u003cstrong\u003eFigure S7B-C\u003c/strong\u003e)\u003csup\u003e[39, 40]\u003c/sup\u003e. Additionally, RGS5+ Fibro, RPL38+ Fibro and CCL4+ Fibro exhibited tumor specific extracellular vesicles mediated communication with CXCR2+ VNN2+ Neu (\u003cstrong\u003eFigure 3H\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Figure S7D\u003c/strong\u003e). Metabolic flux analysis demonstrated limited metabolite exchange between CAFs and CXCR2+ VNN2+ Neu in both tumor and normal microenvironments (\u003cstrong\u003eFigure 3I\u003c/strong\u003e and\u003cstrong\u003e\u0026nbsp;Figure S7E-G\u003c/strong\u003e). Our integrated findings indicate that CAFs subpopulations primarily interact with CXCR2+ VNN2+ Neu through three mechanisms: (1) direct receptor-ligand binding, (2) chemokine signaling, and (3) extracellular vesicle transfer, potentially driving their immunosuppressive phenotypic conversion.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCAFs in Pan-cancer TME Promote CXCR2+ VNN2+ Neu Phenotype Switching and Mediate its Immunosuppressive Function\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo determine whether CAF subpopulations enhance CXCR2+ VNN2+ Neu immunosuppression through these mechanisms, we examined MMP11+ Fibro mediated gene regulation. MMP11+ Fibro significantly modulated 38 target genes in CXCR2+ VNN2+ Neu via ligand-chemokine signaling (\u003cstrong\u003eFigure 4A\u003c/strong\u003e), including upregulated BHLHE40/CCL3 and downregulated DDIT4/EHD1 (\u003cstrong\u003eFigure 8A\u003c/strong\u003e and \u003cstrong\u003eFigure 9A\u003c/strong\u003e). Functional enrichment revealed upregulated genes were associated with immune suppression and fibroblast proliferation, while downregulated genes correlated with granulocyte activation and immune response pathways. MMP11+ Fibro bidirectionally regulates CXCR2+ VNN2+ Neu target genes to promote immunosuppression (\u003cstrong\u003eFigure 4B\u003c/strong\u003e). Extracellular vesicles mediated regulation by RGS5+ Fibro, RPL38+ Fibro, and CCL4+ Fibro through miR-24-3p, miR-145-3p upregulated RPS24, EEF1A1 and other genes expression in tumors (\u003cstrong\u003eFigure S10A\u003c/strong\u003e). These targets showed enrichment for cytoplasmic translation and immune suppression pathways, with weaker activity in normal tissues (\u003cstrong\u003eFigure 4C-D\u003c/strong\u003e and \u003cstrong\u003eFigure S10B\u003c/strong\u003e). SCENIC analysis\u003csup\u003e[41]\u003c/sup\u003e identified tumor-enriched transcription factors (TFAP2A, SPIB) regulating this network (\u003cstrong\u003eFigure 4E-F\u003c/strong\u003e), confirmed at protein level across multiple cancers (HSCLC, GBM, PCC, LIHC, PAAD, OSCC) (\u003cstrong\u003eFigure 4G\u003c/strong\u003e). These findings demonstrate that TME CAFs coordinately bi-directionally regulate CXCR2+ VNN2+ Neu through multiple mechanisms to drive immunosuppressive phenotypic conversion.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGenes encoding phenotype switching-related transcription factors in CXCR2+ VNN2+ Neu promote their immunosuppressive function\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo investigate the transcriptional regulation of CXCR2+ VNN2+ Neu immunosuppression, we analyzed 10 CRISPR datasets from five studies, identifying Immune Resistant and Immune Sensitive gene sets in colorectal cancer and melanoma (\u003cstrong\u003eTable S8-9\u003c/strong\u003e). The top 40 genes from each category are shown in \u003cstrong\u003eFigure 5A-B\u003c/strong\u003e. Analysis of transcription factor Z-scores in the CXCR2+ VNN2+ Neu regulatory network revealed HLTF, GTF2IRD1, BACH1, and ATF2 as key mediators of tumor immunosuppression (\u003cstrong\u003eFigure 5C\u003c/strong\u003e). Intersection analysis demonstrated that all identified CRISPR transcription factors except HLTF were expressed in CXCR2+VNN2+Neu (\u003cstrong\u003eFigure 5D\u003c/strong\u003e). Virtual knockdown of GTF2IRD1, BACH1, and ATF2 across 10 cancer types (including PAAD) significantly downregulated immunosuppression related pathways (GO:0001915, GO:0002698, GO:0002683), with pathway associated gene counts shown in \u003cstrong\u003eFigure 5E\u003c/strong\u003e and expression changes in \u003cstrong\u003eFigure 5F\u003c/strong\u003e and \u003cstrong\u003eFigure S11A-C\u003c/strong\u003e. These findings establish CXCR2+ VNN2+ Neu phenotype associated transcription factors as critical regulators of its immunosuppressive function.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCAFs promote CXCR2+ VNN2+ Neu phenotypic shift and immunosuppressive function leading to immunotherapy resistance\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo assess CAFs mediated CXCR2+ VNN2+ Neu modulation in immunotherapy response, we analyzed seven immunotherapy cohorts (BCC, COAD, LIHC, NSCLC, RCC, SCC, SKCM; \u003cstrong\u003eTable S10\u003c/strong\u003e). Following data integration, batch correction, and quality control, we identified and clustered neutrophils into eight distinct subpopulations based on marker genes (\u003cstrong\u003eFigure 6A-B\u003c/strong\u003e and \u003cstrong\u003eTable S11\u003c/strong\u003e). The C1_Neu subset showed significant enrichment in non-responders across cancer types (\u003cstrong\u003eFigure 6C\u003c/strong\u003e and \u003cstrong\u003eFigure S12A-B\u003c/strong\u003e), revealing a potential association with treatment resistance. AddModuleScore\u0026nbsp;analysis revealed significantly higher CXCR2+ VNN2+ Neu scores in C1_Neu and C0_Neu compared to other subpopulations, confirming their phenotypic similarity (\u003cstrong\u003eFigure 6D\u003c/strong\u003e). Functional enrichment demonstrated distinct pathway activation patterns: non-responders showed C1_Neu enrichment for ATP biosynthesis and immune suppression, while responders exhibited membrane biogenesis and leukocyte cytotoxicity pathways. Parallel findings in C0_Neu suggest CXCR2+ VNN2+ Neu undergoes comparable phenotypic switching in treatment resistant patients (\u003cstrong\u003eFigure 6E\u003c/strong\u003e and \u003cstrong\u003eFigure S12C\u003c/strong\u003e). Building on these observations, we investigated fibroblast involvement by clustering them into eight distinct subpopulations (\u003cstrong\u003eFigure 6F\u0026nbsp;\u003c/strong\u003eand \u003cstrong\u003eFigure S12D\u003c/strong\u003e). Analysis revealed C1_Fibro enrichment in non-responders (\u003cstrong\u003eFigure 6G\u0026nbsp;\u003c/strong\u003eand\u003cstrong\u003e\u0026nbsp;Figure S12E-F\u003c/strong\u003e), with cellular communication studies demonstrating robust C1_Fibro interactions with C0_Neu/C1_Neu through ANXA1/COL1A1 ligand-receptor pairs and CXCL9/CXCL2 chemokines in non-responders (\u003cstrong\u003eFigure 6H\u003c/strong\u003e and \u003cstrong\u003eFigure 6J\u003c/strong\u003e). These interactions were attenuated in responders (\u003cstrong\u003eFigure S12G\u003c/strong\u003e), while extracellular vesicles mediated communication primarily occurred between neutrophils (\u003cstrong\u003eFigure 6I\u003c/strong\u003e and \u003cstrong\u003eFigure S12H\u003c/strong\u003e). Metabolite exchange analysis showed no differential fibroblasts-neutrophils interactions between response groups (\u003cstrong\u003eFigure S13A-D\u003c/strong\u003e). These interactions mediated the upregulation of BHLHE40 and CCL4 in CXCR2+VNN2+Neu, enhancing its immunosuppressive functions including inhibition of T cell cytotoxicity and immune system processes (\u003cstrong\u003eFigure 6K-L\u003c/strong\u003e). Collectively, our findings demonstrate that pan-cancer CAFs drive CXCR2+ VNN2+ Neu phenotypic conversion and immunosuppression primarily through ligand-chemokine signaling, ultimately contributing to patient immunotherapy resistance.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeepsurv deep learning model accurately predicts prognosis of pan-cancer patients based on CVN-GRN\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo investigate the prognostic value of the CXCR2+ VNN2+ Neu gene regulatory network (CVN-GRN), we employed 10 machine learning and 2 deep learning models for survival prediction. Using five-fold cross-validation, we randomly partitioned 8,278 TCGA patients across 22 cancer types (including BLCA and BRCA) into training (4/5) and validation (1/5) sets. Model performance was evaluated using Harrell\u0026apos;s C-index, Begg\u0026apos;s C-index, Uno\u0026apos;s C-index, GH C-index, and time-dependent AUC values (\u003cstrong\u003eFigure 7A\u003c/strong\u003e). The deepsurv model demonstrated superior predictive performance, with five-fold cross-validation results detailed in \u003cstrong\u003eTable S12\u003c/strong\u003e. External validation across seven independent cohorts confirmed its robust generalizability (\u003cstrong\u003eFigure 7B\u003c/strong\u003e and \u003cstrong\u003eTable S13\u003c/strong\u003e). Survival analysis revealed significantly worse outcomes for high CVN-GRN score patients in both training and test sets (\u003cstrong\u003eFigure 7C-D\u003c/strong\u003e). The model architecture, comprising an input layer, three hidden layers, and an output layer optimized through backpropagation, is shown in \u003cstrong\u003eFigure 7E\u003c/strong\u003e. Consistent results were observed across all external validation cohorts (\u003cstrong\u003eFigure 7F\u003c/strong\u003e and\u003cstrong\u003e\u0026nbsp;Figure S13E\u003c/strong\u003e), establishing deepsurv as an effective tool for CVN-GRN-based prognostic stratification and potential treatment guidance in pan-cancer patients.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003e\u003cstrong\u003eData collection\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSingle-cell sequencing data\u0026nbsp;\u003c/strong\u003eThis study utilized single-cell RNA sequencing data obtained from public repositories: Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/), EMBL-EBI (https://www.ebi.ac.uk/), National Genomics Data Center (NGDC, https://ngdc.cncb.ac.cn/omix/), and China National GeneBank DataBase (CNGBdb, https://db.cngb.org/). We analyzed datasets comprising 462 patients across 21 malignancies: basal cell carcinoma (BCC), bladder urothelial carcinoma (BLCA), breast invasive carcinoma (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), gallbladder cancer (GBC), glioblastoma multiforme (GBM), head and neck squamous cell carcinoma (HNSCC), liver hepatocellular carcinoma (LIHC), non-small-cell lung carcinoma (NSCLC), oral squamous cell carcinoma (OSCC), ovarian serous cystadenocarcinoma (OV), pancreatic adenocarcinoma (PAAD), prostate adenocarcinoma (PRAD), renal cell carcinoma (RCC), squamous cell carcinoma (SCC), skin cutaneous melanoma (SKCM), stomach adenocarcinoma (STAD), and thyroid carcinoma (THCA). Complete dataset details are provided in \u003cstrong\u003eTable S1\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSpatial transcriptome data\u0026nbsp;\u003c/strong\u003eSpatial transcriptomic data were obtained from seven cancer types across multiple repositories: HRA000437 (LIHC/normal liver) and GSE203612 (PAAD/OV) from GEO; E-MTAB-13530 (NSCLC/normal lung) from EMBL-EBI; and COAD data from the Cancer Diversity Asia portal ((http://www.cancerdiversity.asia/scCRLM/)).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBulk transcriptome data\u0026nbsp;\u003c/strong\u003eBulk transcriptomic and clinical data were sourced from TCGA (8,278 patients across 22 cancers) and six validation cohorts: CGGA (301/325/693 for GBM), METABRIC (BRCA), PRAD-SU-2019 (PRAD), GSE13507 (BLCA), and E-MTAB-6134 (PAAD), totaling 3,718 additional patients.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eProteomics data\u0026nbsp;\u003c/strong\u003eProteomic data were sourced from the Clinical Proteomic Tumor Analysis Consortium (CPTAC, https://proteomics.cancer.gov/programs/cptac) database, which includes data for six cancer types: NSCLC, GBM, RCC, LIHC, PAAD, and OSCC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImmunotherapy cohorts\u0026nbsp;\u003c/strong\u003eImmunotherapy cohorts were obtained from multiple sources: GSE123813 (BCC/SCC), GSE205506 (COAD), GSE207422 (NSCLC), PRJNA705464 (RCC), and GSE120575 (SKCM) from GEO, along with LIHC data from Mendeley Data.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCRISPR dataset\u0026nbsp;\u003c/strong\u003eThis study analyzed CRISPR/Cas9 screening data from five established studies (Freeman\u003csup\u003e[42]\u003c/sup\u003e, Kearney\u003csup\u003e[43]\u003c/sup\u003e, Manguso\u003csup\u003e[44]\u003c/sup\u003e, Pan\u003csup\u003e[45]\u003c/sup\u003e, Pate\u003csup\u003e[46]\u003c/sup\u003e), which were reorganized into 10 datasets focusing primarily on COAD and SKCM models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePan-cancer single cell sequencing data processing and integration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSingle-cell data processing was conducted using Seurat (v4.3.0.1) with the following workflow: mitochondrial gene filtering (\u0026gt;25% threshold), PCA-based dimensionality reduction, and Harmony (v0.1.1) batch correction. The integrated dataset was generated through UMAP projection and subsequent merging of harmonized Seurat objects for downstream analyses.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTrajectory analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePan-cancer neutrophils trajectory analysis was performed using five complementary algorithms: CytoTRACE2 (v1.0.0) for potency scoring, Slingshot (v2.8.0) for trajectory reconstruction, Monocle2 (v2.28.0), Vector (gmodels v2.18.1.1), and scTour for independent validation. Computational consistency across all methods confirmed trajectory reliability.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSpatial transcriptome data analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSpatial transcriptomic data processing involved: (1) data import (Read10X/Read10X_h5 for expression matrices, Read10X_Image for spatial coordinates); (2) Seurat-based normalization and clustering; (3) cellular deconvolution using CellTrek (v0.0.94) with paired scRNA-seq as reference; and (4) spatial co-localization analysis via scoloc network construction and visualization.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCell-Communication analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCellular communication analysis was performed using CellChat (v2.1.2) for ligand-receptor and chemokine interactions, nichenetr (v2.0.1) for ligand-target gene regulation, miRTalk (v1.0) for extracellular vesicle interactions, and mebocost for metabolite-mediated signaling. Overexpressed interactions were identified and aggregated through CellChat\u0026apos;s analytical pipeline.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGene regulatory network analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGene regulatory networks were constructed using SCENIC through three computational steps: (1) GRNBoost-based inference of TF-target gene co-expression networks, (2) RcisTarget motif analysis to identify enriched regulatory modules, and (3) AUCell scoring of regulator activity at single-cell resolution.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCRISPR data analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe analyzed 10 previously mentioned CRISPR/Cas9 datasets focusing on COAD and SKCM to evaluate transcription factor-immune response associations. Using logFC of sgRNA reads between CTL-treated and control conditions, we calculated Z-scores for 21,304 genes, where lower values indicated immune sensitivity and higher values corresponded to immune resistance.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003escTenifoldKnk analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe employed scTenifoldKnk (v1.0.1)\u003csup\u003e[47]\u003c/sup\u003e for virtual TF knockout analysis targeting three immune-related pathways (GO:0001915, GO:0002698, GO:0002683) in neutrophils from five cancer types (PAAD, NSCLC, RCC, HNSCC, OV). Single-cell regulatory networks were reconstructed to quantify expression perturbations following virtual knockouts, revealing TF-mediated immune modulation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConstruction of the CVN-GRN score\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe evaluated 12 prognostic models (10 machine learning including Lasso-Cox/MTLR/CoxBoost and 2 deep learning algorithms) on TCGA pan-cancer data (22 tumor types) using five validation metrics (Harrell\u0026apos;s/Begg\u0026apos;s/Uno\u0026apos;s/GH C-indices, time-AUC). Following 50-fold cross-validation, the optimal model (Deepsurv) was used to develop the CVN-GRN score, which effectively stratified patients into prognostic subgroups based on CXCR2+VNN2+Neu transcriptional networks.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eR v4.3.1 was applied to conduct all statistical analyses in this study. The Wilcox test was implemented to compare the GSVA scores of marker genes between two different subgroups. The log-rank test was utilized to assess the significance of observed differences in overall survival (OS). Statistical significance was determined by a two tailed p-value less than 0.05, unless explicitly specified otherwise.\u003c/p\u003e"},{"header":"Disscusion","content":"\u003cp\u003eEmerging evidence establishes neutrophils as critical modulators of tumor progression within the tumor microenvironment\u003csup\u003e[\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]\u003c/sup\u003e. While the classical N1/N2 paradigm posits antitumor and protumor neutrophil subsets respectively\u003csup\u003e[\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]\u003c/sup\u003e, recent studies reveal this dichotomy inadequately captures their functional plasticity\u003csup\u003e[\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]\u003c/sup\u003e. The mechanisms governing neutrophil phenotypic switching and its clinical implications remain poorly understood.\u003c/p\u003e \u003cp\u003eOur pan-cancer analysis of single-cell data from 462 patients across 21 cancer types identified eight neutrophil subpopulations, six of which (excluding HIST1H4C\u0026thinsp;+\u0026thinsp;STMN1\u0026thinsp;+\u0026thinsp;Neu and KRT7\u0026thinsp;+\u0026thinsp;ALDH1A3\u0026thinsp;+\u0026thinsp;Neu) showed tumor-specific enrichment. Functional characterization revealed distinct activation patterns: TIMP1\u0026thinsp;+\u0026thinsp;EREG\u0026thinsp;+\u0026thinsp;Neu and CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu exhibited enhanced cytokine signaling, while APOE\u0026thinsp;+\u0026thinsp;C1QB\u0026thinsp;+\u0026thinsp;Neu and IGKC\u0026thinsp;+\u0026thinsp;HLA-DPB1\u0026thinsp;+\u0026thinsp;Neu demonstrated immune-activating properties, suggesting antitumor potential. Notably, CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu acquired immunosuppressive functions including immune system inhibition and lymphocyte activation suppression in tumors. Trajectory analysis using five independent algorithms (CytoTRACE, Slingshot, Monocle2, Vector, scTour) consistently identified SPRR1B\u0026thinsp;+\u0026thinsp;S100A7\u0026thinsp;+\u0026thinsp;Neu as the progenitor population and TIMP1\u0026thinsp;+\u0026thinsp;EREG\u0026thinsp;+\u0026thinsp;Neu/CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu as terminal differentiation states. Comparative functional analysis along the neutrophil differentiation trajectory revealed peak antigen presentation and phagocytic activity in APOE\u0026thinsp;+\u0026thinsp;C1QB\u0026thinsp;+\u0026thinsp;Neu and IGKC\u0026thinsp;+\u0026thinsp;HLA-DPB1\u0026thinsp;+\u0026thinsp;Neu subsets, followed by progressive decline to minimal levels in CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu, consistent with their established antitumor roles. Notably, this functional attenuation was markedly accelerated in tumor microenvironments, suggesting TME-mediated promotion of CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu phenotypic conversion. Subsequent validation demonstrated elevated co-inhibitory molecule expression, strong T-cell exhaustion correlation, and immunosuppressive pathway enrichment in TIMP1\u0026thinsp;+\u0026thinsp;EREG\u0026thinsp;+\u0026thinsp;Neu and CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu subsets, with pan-cancer prognostic analysis confirming their association with poor clinical outcomes.\u003c/p\u003e \u003cp\u003eSpatial neighborhood analysis revealed fibroblast-neutrophil co-localization across multiple cancer types, except in COAD, SKCM, and normal tissues, potentially explaining the favorable prognosis associated with neutrophils in these malignancies\u003csup\u003e[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]\u003c/sup\u003e. Our findings identify CAFs as key regulators of neutrophil plasticity. Tumor-associated MMP11\u0026thinsp;+\u0026thinsp;Fibro engages CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu through ANXA1 receptor binding and CCL3 cytokine signaling, while RGS5\u0026thinsp;+\u0026thinsp;Fibro, CCL4\u0026thinsp;+\u0026thinsp;Fibro, and RPL38\u0026thinsp;+\u0026thinsp;Fibro communicates via hsa-miR-24-3p-containing extracellular vesicles. Functional genomic analyses demonstrated these interactions bidirectionally modulate CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu gene expression, driving phenotypic conversion toward an immunosuppressive state. To elucidate CAFs mediated regulation of CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu, we constructed its gene regulatory network and identified key transcription factors, subsequently validated by pan-cancer proteomic data showing their tumor-specific overexpression. Analysis of five CRISPR studies revealed these transcription factors mediate cancer cell immune resistance. Virtual knockdown experiments confirmed their functional importance, demonstrating significant downregulation of immunosuppressive pathways (GO:0001915, GO:0002698, GO:0002683) in CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu, establishing a direct link between these transcriptional regulators and neutrophil immunosuppressive polarization.\u003c/p\u003e \u003cp\u003eAnalysis of pan-cancer immunotherapy cohorts revealed distinct CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu phenotypic shifts in non-responders, mediated by CAF subpopulations through ANXA1 receptor ligands and CCL3 chemokine signaling that drive immunosuppressive gene expression programs. Prognostic modeling using 10 machine and 2 deep learning approaches identified Deepsurv as optimal for CVN-GRN-based stratification, with validation across seven independent cohorts confirming its clinical utility for outcome prediction and therapeutic guidance.\u003c/p\u003e \u003cp\u003eOur study has several limitations. First, the CRISPR analysis was restricted to COAD and SKCM, leaving the association between CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu transcriptional regulators and immunosuppression unvalidated in other cancer types. Second, the findings require further experimental validation through mechanistic studies.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eOur study establishes a comprehensive pan-cancer neutrophils atlas that delineates phenotypic and functional heterogeneity, uncovering key mechanisms of neutrophil plasticity. We demonstrate that specific fibroblasts subpopulations critically regulate neutrophils phenotypic conversion, which directly contributes to immunotherapy resistance. Furthermore, we developed the CVN-GRN scoring system based on CXCR2\u0026thinsp;+\u0026thinsp;VNN2\u0026thinsp;+\u0026thinsp;Neu transcriptional networks and validated its clinical utility through deep learning-based prognostic stratification, providing a valuable framework for therapeutic decision-making in pan-cancer patients.\u003c/p\u003e "},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; Contributions\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization, Zhiyu Guo and Jinsheng Huang; Investigation, Zhiyu Guo, Xujia Li and Lingli Huang; Software, Zhiyu Guo; Formal Analysis, Zhiyu Guo; Writing\u0026mdash;original draft, Xujia Li and Lingli Huang; Writing\u0026mdash;review \u0026amp; editing, Mengge Gao and Jinsheng Huang; Supervision, Jinsheng Huang; Visualization, Zhiyu Guo and Xujia Li; Funding acquisition, Jinsheng Huang and Yue Yan.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data generated in this study are included in this published article and its supplementary information (\u003cstrong\u003eTable S1\u003c/strong\u003e). Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data utilized in this study were sourced from publicly accessible databases and were managed under approved ethical exemptions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors agreed to publication in the journal.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eZhang F, Xia Y, Su J, Quan F, Zhou H, Li Q, et al. \u003cstrong\u003eNeutrophil diversity and function in health and disease\u003c/strong\u003e. \u003cem\u003eSignal transduction and targeted therapy \u003c/em\u003e2024; 9(1):343.\u003c/li\u003e\n\u003cli\u003eHuang X, Nepovimova E, Adam V, Sivak L, Heger Z, Valko M, et al. \u003cstrong\u003eNeutrophils in Cancer immunotherapy: friends or foes?\u003c/strong\u003e \u003cem\u003eMolecular cancer \u003c/em\u003e2024; 23(1):107.\u003c/li\u003e\n\u003cli\u003eHe W, Yan L, Hu D, Hao J, Liou YC, Luo G. Neutrophil heterogeneity and plasticity: unveiling the multifaceted roles in health and disease. \u003cem\u003eMedComm \u003c/em\u003e2025; 6(2):e70063.\u003c/li\u003e\n\u003cli\u003eFridlender ZG, Sun J, Kim S, Kapoor V, Cheng G, Ling L, et al. \u003cstrong\u003ePolarization of tumor-associated neutrophil phenotype by TGF-beta: \u0026quot;N1\u0026quot; versus \u0026quot;N2\u0026quot; TAN\u003c/strong\u003e. \u003cem\u003eCancer cell \u003c/em\u003e2009; 16(3):183-194.\u003c/li\u003e\n\u003cli\u003eAndzinski L, Kasnitz N, Stahnke S, Wu CF, Gereke M, von K\u0026ouml;ckritz-Blickwede M, et al. \u003cstrong\u003eType I IFNs induce anti-tumor polarization of tumor associated neutrophils in mice and human\u003c/strong\u003e. \u003cem\u003eInternational journal of cancer \u003c/em\u003e2016; 138(8):1982-1993.\u003c/li\u003e\n\u003cli\u003eZhao JJ, Pan K, Wang W, Chen JG, Wu YH, Lv L, et al. The prognostic value of tumor-infiltrating neutrophils in gastric adenocarcinoma after resection. \u003cem\u003ePloS one \u003c/em\u003e2012; 7(3):e33655.\u003c/li\u003e\n\u003cli\u003eSconocchia G, Zlobec I, Lugli A, Calabrese D, Iezzi G, Karamitopoulou E, et al. \u003cstrong\u003eTumor infiltration by Fc\u0026gamma;RIII (CD16)+ myeloid cells is associated with improved survival in patients with colorectal carcinoma\u003c/strong\u003e. \u003cem\u003eInternational journal of cancer \u003c/em\u003e2011; 128(11):2663-2672.\u003c/li\u003e\n\u003cli\u003eTakakura K, Ito Z, Suka M, Kanai T, Matsumoto Y, Odahara S, et al. Comprehensive assessment of the prognosis of pancreatic cancer: peripheral blood neutrophil-lymphocyte ratio and immunohistochemical analyses of the tumour site. \u003cem\u003eScandinavian journal of gastroenterology \u003c/em\u003e2016; 51(5):610-617.\u003c/li\u003e\n\u003cli\u003eBodac A, Meylan E. \u003cstrong\u003eNeutrophil metabolism in the cancer context\u003c/strong\u003e. \u003cem\u003eSeminars in immunology \u003c/em\u003e2021; 57:101583.\u003c/li\u003e\n\u003cli\u003eSinghal S, Bhojnagarwala PS, O\u0026apos;Brien S, Moon EK, Garfall AL, Rao AS, et al. Origin and Role of a Subset of Tumor-Associated Neutrophils with Antigen-Presenting Cell Features in Early-Stage Human Lung Cancer. \u003cem\u003eCancer cell \u003c/em\u003e2016; 30(1):120-135.\u003c/li\u003e\n\u003cli\u003ePerez C, Botta C, Zabaleta A, Puig N, Cedena MT, Goicoechea I, et al. Immunogenomic identification and characterization of granulocytic myeloid-derived suppressor cells in multiple myeloma. \u003cem\u003eBlood \u003c/em\u003e2020; 136(2):199-209.\u003c/li\u003e\n\u003cli\u003eJin MZ, Jin WL. \u003cstrong\u003eThe updated landscape of tumor microenvironment and drug repurposing\u003c/strong\u003e. \u003cem\u003eSignal transduction and targeted therapy \u003c/em\u003e2020; 5(1):166.\u003c/li\u003e\n\u003cli\u003eYu X, Li C, Wang Z, Xu Y, Shao S, Shao F, et al. \u003cstrong\u003eNeutrophils in cancer: dual roles through intercellular interactions\u003c/strong\u003e. \u003cem\u003eOncogene \u003c/em\u003e2024; 43(16):1163-1177.\u003c/li\u003e\n\u003cli\u003eZhang Y, Guoqiang L, Sun M, Lu X. Targeting and exploitation of tumor-associated neutrophils to enhance immunotherapy and drug delivery for cancer treatment. \u003cem\u003eCancer biology \u0026amp; medicine \u003c/em\u003e2020; 17(1):32-43.\u003c/li\u003e\n\u003cli\u003eSalcher S, Sturm G, Horvath L, Untergasser G, Kuempers C, Fotakis G, et al. High-resolution single-cell atlas reveals diversity and plasticity of tissue-resident neutrophils in non-small cell lung cancer. \u003cem\u003eCancer cell \u003c/em\u003e2022; 40(12):1503-1520.e1508.\u003c/li\u003e\n\u003cli\u003eZhao Y, Liu Z, Liu G, Zhang Y, Liu S, Gan D, et al. \u003cstrong\u003eNeutrophils resist ferroptosis and promote breast cancer metastasis through aconitate decarboxylase 1\u003c/strong\u003e. \u003cem\u003eCell metabolism \u003c/em\u003e2023; 35(10):1688-1703.e1610.\u003c/li\u003e\n\u003cli\u003eMeng Y, Ye F, Nie P, Zhao Q, An L, Wang W, et al. Immunosuppressive CD10(+)ALPL(+) neutrophils promote resistance to anti-PD-1 therapy in HCC by mediating irreversible exhaustion of T cells. \u003cem\u003eJournal of hepatology \u003c/em\u003e2023; 79(6):1435-1449.\u003c/li\u003e\n\u003cli\u003eXie Y, Zhou T, Li X, Zhao K, Bai W, Hou X, et al. Targeting ESE3/EHF With Nifurtimox Inhibits CXCR2(+) Neutrophil Infiltration and Overcomes Pancreatic Cancer Resistance to Chemotherapy and Immunotherapy. \u003cem\u003eGastroenterology \u003c/em\u003e2024; 167(2):281-297.\u003c/li\u003e\n\u003cli\u003eChen Z, Soni N, Pinero G, Giotti B, Eddins DJ, Lindblad KE, et al. \u003cstrong\u003eMonocyte depletion enhances neutrophil influx and proneural to mesenchymal transition in glioblastoma\u003c/strong\u003e. \u003cem\u003eNature communications \u003c/em\u003e2023; 14(1):1839.\u003c/li\u003e\n\u003cli\u003eHu S, Lu H, Xie W, Wang D, Shan Z, Xing X, et al. \u003cstrong\u003eTDO2+ myofibroblasts mediate immune suppression in malignant transformation of squamous cell carcinoma\u003c/strong\u003e. \u003cem\u003eThe Journal of clinical investigation \u003c/em\u003e2022; 132(19).\u003c/li\u003e\n\u003cli\u003eJin X, Wang Q, Luo F, Pan J, Lu T, Zhao Y, et al. Single-cell transcriptomic analysis of tumor heterogeneity and intercellular networks in human urothelial carcinoma. \u003cem\u003eChinese medical journal \u003c/em\u003e2023; 136(6):690-706.\u003c/li\u003e\n\u003cli\u003eMa R, Zhou X, Zhai X, Wang C, Hu R, Chen Y, et al. Single-cell RNA sequencing reveals immune cell dysfunction in the peripheral blood of patients with highly aggressive gastric cancer. \u003cem\u003eCell proliferation \u003c/em\u003e2024; 57(5):e13591.\u003c/li\u003e\n\u003cli\u003eHu J, Chen Z, Bao L, Zhou L, Hou Y, Liu L, et al. \u003cstrong\u003eSingle-Cell Transcriptome Analysis Reveals Intratumoral Heterogeneity in ccRCC, which Results in Different Clinical Outcomes\u003c/strong\u003e. \u003cem\u003eMolecular therapy : the journal of the American Society of Gene Therapy \u003c/em\u003e2020; 28(7):1658-1672.\u003c/li\u003e\n\u003cli\u003eWang H, Xiong A, Chen X, Guo J, Tang Z, Wu C, et al. CXCR1(+) neutrophil infiltration orchestrates response to third-generation EGFR-TKI in EGFR mutant non-small-cell lung cancer. \u003cem\u003eSignal transduction and targeted therapy \u003c/em\u003e2024; 9(1):342.\u003c/li\u003e\n\u003cli\u003eKorsunsky I, Millard N, Fan J, Slowikowski K, Zhang F, Wei K, et al. \u003cstrong\u003eFast, sensitive and accurate integration of single-cell data with Harmony\u003c/strong\u003e. \u003cem\u003eNature methods \u003c/em\u003e2019; 16(12):1289-1296.\u003c/li\u003e\n\u003cli\u003eKang M, Armenteros JJA, Gulati GS, Gleyzer R, Avagyan S, Brown EL, et al. \u003cstrong\u003eMapping single-cell developmental potential in health and disease with interpretable deep learning\u003c/strong\u003e. \u003cem\u003ebioRxiv : the preprint server for biology \u003c/em\u003e2024.\u003c/li\u003e\n\u003cli\u003eStreet K, Risso D, Fletcher RB, Das D, Ngai J, Yosef N, et al. \u003cstrong\u003eSlingshot: cell lineage and pseudotime inference for single-cell transcriptomics\u003c/strong\u003e. \u003cem\u003eBMC genomics \u003c/em\u003e2018; 19(1):477.\u003c/li\u003e\n\u003cli\u003eQiu X, Mao Q, Tang Y, Wang L, Chawla R, Pliner HA, et al. \u003cstrong\u003eReversed graph embedding resolves complex single-cell trajectories\u003c/strong\u003e. \u003cem\u003eNature methods \u003c/em\u003e2017; 14(10):979-982.\u003c/li\u003e\n\u003cli\u003eLi Q. scTour: a deep learning architecture for robust inference and accurate prediction of cellular dynamics. \u003cem\u003eGenome biology \u003c/em\u003e2023; 24(1):149.\u003c/li\u003e\n\u003cli\u003eWu Y, Ma J, Yang X, Nan F, Zhang T, Ji S, et al. \u003cstrong\u003eNeutrophil profiling illuminates anti-tumor antigen-presenting potency\u003c/strong\u003e. \u003cem\u003eCell \u003c/em\u003e2024; 187(6):1422-1439.e1424.\u003c/li\u003e\n\u003cli\u003eSchoeps B, Eckfeld C, Prokopchuk O, B\u0026ouml;ttcher J, H\u0026auml;u\u0026szlig;ler D, Steiger K, et al. \u003cstrong\u003eTIMP1 Triggers Neutrophil Extracellular Trap Formation in Pancreatic Cancer\u003c/strong\u003e. \u003cem\u003eCancer research \u003c/em\u003e2021; 81(13):3568-3579.\u003c/li\u003e\n\u003cli\u003eSteele CW, Karim SA, Leach JDG, Bailey P, Upstill-Goddard R, Rishi L, et al. \u003cstrong\u003eCXCR2 Inhibition Profoundly Suppresses Metastases and Augments Immunotherapy in Pancreatic Ductal Adenocarcinoma\u003c/strong\u003e. \u003cem\u003eCancer cell \u003c/em\u003e2016; 29(6):832-845.\u003c/li\u003e\n\u003cli\u003eCheng Y, Mo F, Li Q, Han X, Shi H, Chen S, et al. Targeting CXCR2 inhibits the progression of lung cancer and promotes therapeutic effect of cisplatin. \u003cem\u003eMolecular cancer \u003c/em\u003e2021; 20(1):62.\u003c/li\u003e\n\u003cli\u003eWei R, He S, Bai S, Sei E, Hu M, Thompson A, et al. \u003cstrong\u003eSpatial charting of single-cell transcriptomes in tissues\u003c/strong\u003e. \u003cem\u003eNature biotechnology \u003c/em\u003e2022; 40(8):1190-1199.\u003c/li\u003e\n\u003cli\u003eLiu Y, Dong G, Yu J, Liang P. Integration of single-cell and spatial transcriptomics reveals fibroblast subtypes in hepatocellular carcinoma: spatial distribution, differentiation trajectories, and therapeutic potential. \u003cem\u003eJournal of translational medicine \u003c/em\u003e2025; 23(1):198.\u003c/li\u003e\n\u003cli\u003eWang Z, Guo X, Li X, Wang J, Zhang N, Amin B, et al. \u003cstrong\u003eCancer-associated fibroblast-derived MMP11 promotes tumor progression in pancreatic cancer\u003c/strong\u003e. \u003cem\u003eCancer science \u003c/em\u003e2025; 116(3):643-655.\u003c/li\u003e\n\u003cli\u003eForsthuber A, Aschenbrenner B, Korosec A, Jacob T, Annusver K, Krajic N, et al. Cancer-associated fibroblast subtypes modulate the tumor-immune microenvironment and are associated with skin cancer malignancy. \u003cem\u003eNature communications \u003c/em\u003e2024; 15(1):9678.\u003c/li\u003e\n\u003cli\u003eKoppensteiner L, Mathieson L, Neilson L, O\u0026apos;Connor RA, Akram AR. IFN\u0026gamma; and TNF\u0026alpha; drive an inflammatory secretion profile in cancer-associated fibroblasts from human non-small cell lung cancer. \u003cem\u003eFEBS letters \u003c/em\u003e2025; 599(5):713-723.\u003c/li\u003e\n\u003cli\u003eGaray J, Piazuelo MB, Majumdar S, Li L, Trillo-Tinoco J, Del Valle L, et al. The homing receptor CD44 is involved in the progression of precancerous gastric lesions in patients infected with Helicobacter pylori and in development of mucous metaplasia in mice. \u003cem\u003eCancer letters \u003c/em\u003e2016; 371(1):90-98.\u003c/li\u003e\n\u003cli\u003eTang F, Zhu Y, Shen J, Yuan B, He X, Tian Y, et al. CD44(+) cells enhance pro-tumor stroma in the spatial landscape of colorectal cancer leading edge. \u003cem\u003eBritish journal of cancer \u003c/em\u003e2025.\u003c/li\u003e\n\u003cli\u003eAibar S, Gonz\u0026aacute;lez-Blas CB, Moerman T, Huynh-Thu VA, Imrichova H, Hulselmans G, et al. \u003cstrong\u003eSCENIC: single-cell regulatory network inference and clustering\u003c/strong\u003e. \u003cem\u003eNature methods \u003c/em\u003e2017; 14(11):1083-1086.\u003c/li\u003e\n\u003cli\u003eFreeman AJ, Vervoort SJ, Ramsbottom KM, Kelly MJ, Michie J, Pijpers L, et al. \u003cstrong\u003eNatural Killer Cells Suppress T Cell-Associated Tumor Immune Evasion\u003c/strong\u003e. \u003cem\u003eCell reports \u003c/em\u003e2019; 28(11):2784-2794.e2785.\u003c/li\u003e\n\u003cli\u003eKearney CJ, Vervoort SJ, Hogg SJ, Ramsbottom KM, Freeman AJ, Lalaoui N, et al. \u003cstrong\u003eTumor immune evasion arises through loss of TNF sensitivity\u003c/strong\u003e. \u003cem\u003eScience immunology \u003c/em\u003e2018; 3(23).\u003c/li\u003e\n\u003cli\u003eManguso RT, Pope HW, Zimmer MD, Brown FD, Yates KB, Miller BC, et al. \u003cstrong\u003eIn vivo CRISPR screening identifies Ptpn2 as a cancer immunotherapy target\u003c/strong\u003e. \u003cem\u003eNature \u003c/em\u003e2017; 547(7664):413-418.\u003c/li\u003e\n\u003cli\u003ePan D, Kobayashi A, Jiang P, Ferrari de Andrade L, Tay RE, Luoma AM, et al. \u003cstrong\u003eA major chromatin regulator determines resistance of tumor cells to T cell-mediated killing\u003c/strong\u003e. \u003cem\u003eScience (New York, NY) \u003c/em\u003e2018; 359(6377):770-775.\u003c/li\u003e\n\u003cli\u003ePatel SJ, Sanjana NE, Kishton RJ, Eidizadeh A, Vodnala SK, Cam M, et al. \u003cstrong\u003eIdentification of essential genes for cancer immunotherapy\u003c/strong\u003e. \u003cem\u003eNature \u003c/em\u003e2017; 548(7669):537-542.\u003c/li\u003e\n\u003cli\u003eOsorio D, Zhong Y, Li G, Xu Q, Yang Y, Tian Y, et al. scTenifoldKnk: An efficient virtual knockout tool for gene function predictions via single-cell gene regulatory network perturbation. \u003cem\u003ePatterns (New York, NY) \u003c/em\u003e2022; 3(3):100434.\u003c/li\u003e\n\u003cli\u003eXue R, Zhang Q, Cao Q, Kong R, Xiang X, Liu H, et al. \u003cstrong\u003eLiver tumour immune microenvironment subtypes and neutrophil heterogeneity\u003c/strong\u003e. \u003cem\u003eNature \u003c/em\u003e2022; 612(7938):141-147.\u003c/li\u003e\n\u003cli\u003eShaul ME, Fridlender ZG. \u003cstrong\u003eTumour-associated neutrophils in patients with cancer\u003c/strong\u003e. \u003cem\u003eNature reviews Clinical oncology \u003c/em\u003e2019; 16(10):601-620.\u003c/li\u003e\n\u003cli\u003eZhang X, Shi H, Yuan X, Jiang P, Qian H, Xu W. Tumor-derived exosomes induce N2 polarization of neutrophils to promote gastric cancer cell migration. \u003cem\u003eMolecular cancer \u003c/em\u003e2018; 17(1):146.\u003c/li\u003e\n\u003cli\u003eZhang J, Gu J, Wang X, Ji C, Yu D, Wang M, et al. \u003cstrong\u003eEngineering and Targeting Neutrophils for Cancer Therapy\u003c/strong\u003e. \u003cem\u003eAdvanced materials (Deerfield Beach, Fla) \u003c/em\u003e2024; 36(19):e2310318.\u003c/li\u003e\n\u003cli\u003eNg LG, Ostuni R, Hidalgo A. \u003cstrong\u003eHeterogeneity of neutrophils\u003c/strong\u003e. \u003cem\u003eNature reviews Immunology \u003c/em\u003e2019; 19(4):255-265.\u003c/li\u003e\n\u003cli\u003eXiong S, Dong L, Cheng L. \u003cstrong\u003eNeutrophils in cancer carcinogenesis and metastasis\u003c/strong\u003e. \u003cem\u003eJournal of hematology \u0026amp; oncology \u003c/em\u003e2021; 14(1):173.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"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":"Neutrophils, Fibroblasts, Immunosuppression, Single-cell transcriptome, Pan-Cancer","lastPublishedDoi":"10.21203/rs.3.rs-6566788/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6566788/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eNeutrophils are the most abundant granulocyte population and have important functions such as defense against pathogens. However, they show significant heterogeneity and play more complex roles in tumors. The theory of two-tiered differentiation of neutrophils is insufficient to summarize their phenotypic and functional heterogeneity. Therefore, specific regulatory mechanisms remain to be explored and neutrophil-based therapeutic regimens remain challenging. Here, we generated a single-cell atlas of neutrophils from 462 patients with 21 cancer types, revealing their heterogeneity, with CXCR2+ VNN2+ Neu as the main functional subpopulation exerting immunosuppressive effects. Spatial transcriptomic data from the pan-cancer elucidated that fibroblast regulated the phenotypic shift of CXCR2+ VNN2+ Neu in tumor tissues and enabled it to acquire immunosuppressive functions through receptor ligands, cytokines, and extracellular vesicles, which suggested that the tumor microenvironment component was a key reason for the heterogeneity of the prognostic association between neutrophils and pan-cancer patients. Subsequently, we constructed a gene regulatory network to demonstrate the specific regulatory mechanisms of this subpopulation and confirmed that the relevant transcription factors were closely associated with its immunosuppressive function. The pan-cancer immunotherapy cohort proved that the CXCR2+ VNN2+ Neu phenotypic shift was also an important cause of immunotherapy resistance in patients. We finally constructed a deep learning model named Deepsurv to accurately stratify pan-cancer patients based on the CXCR2+ VNN2+ Neu phenotypic shift gene regulatory network (CVN-GRN) and predict the prognosis of the patients, which achieved the desired results.\u003c/p\u003e","manuscriptTitle":"Cancer-associated Fibroblasts in Pan-Cancer Drive CXCR2+VNN2+ Neutrophils Reprogramming to Mediate Immunosuppression and Immunotherapy Resistance","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-10 17:02:44","doi":"10.21203/rs.3.rs-6566788/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":"7a94e25f-a8b3-49fd-a906-031950eb21a1","owner":[],"postedDate":"June 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-06-10T17:02:44+00:00","versionOfRecord":[],"versionCreatedAt":"2025-06-10 17:02:44","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6566788","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6566788","identity":"rs-6566788","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}