B3GNT3 is an oncogenic and prognostic biomarker in human tumors via pan-cancer analysis combined with experimental validation

Introduction

In recent years, the global cancer burden has been steadily increasing. According to projections by the International Agency for Research on Cancer (IARC), new cancer cases worldwide are expected to exceed 28.4 million by 2040, with the highest proportion occurring in Asia (1). In China, the most populous country, approximately 4.82 million new cancer cases and 2.57 million cancer-related deaths were reported in 2022, with lung, colorectal and thyroid cancers being the most commonly diagnosed (2). Notably, there has been an increasing trend in the younger population being affected by cancer. Among individuals aged 25–29 years, the annual growth rate of incidence of obesity-related cancers, such as colorectal and breast cancers, has reached 15%, which is substantially higher than the 1.55% increase observed in individuals aged 60 years and older. This significant difference highlights the profound impact of metabolic disorders and changes in lifestyle on younger cohorts (3). Advances in diagnostic and therapeutic technologies have contributed to a decline in mortality rates for certain types of cancer. Pan-cancer biomarkers based on liquid biopsy, such as circulating tumor DNA (ctDNA) and T-cell receptor repertoires, when combined with artificial intelligence algorithms, have significantly enhanced the sensitivity of early screening for malignancies like lung cancer (4). Despite advances in cancer treatment, digestive system cancers with poor prognosis, such as hepatocellular carcinoma and gastric cancer, still account for over 40% of cancer-related deaths in China. This highlights the urgent need for precision prevention and treatment strategies. Future research should further explore the clinical application of pan-cancer molecular biomarkers and integrate multi-omics data to optimize early screening systems, thereby addressing the rising burden of cancer.

Beta-1,3-N-acetylglucosaminyltransferase 3 (B3GNT3), is a member of the glycosyltransferase family. This gene is widely distributed in the Golgi apparatus and plasma membrane, where it primarily catalyzes the β-1,3 linkage between N-acetylglucosamine (GlcNAc) and galactose (Gal). It is implicated in the synthesis of polylactosamine chains and regulates protein glycosylation, thereby influencing critical cellular processes such as cell adhesion, signal transduction, and immune evasion (5,6). Recent studies demonstrated that B3GNT3 is aberrantly overexpressed in various malignant tumors and is significantly associated with tumor progression, metastasis and poor prognosis (7-14). In gynecological malignancies, such as cervical, ovarian and endometrial cancers, elevated B3GNT3 expression is correlated with reduced overall survival (OS) (7,8). Mechanistically, it promotes tumor cell proliferation via NF-κB pathway activation and reshapes the immunosuppressive tumor microenvironment (TME) by reducing CD8⁺ T cell infiltration and inducing M2-type macrophage polarization (7). Bioinformatics analysis and functional experiments at the cellular level demonstrated that B3GNT3 play a significant role in the development, migration and invasion of liver cancer, lung cancer, oesophageal squamous cell carcinoma, pancreatic adenocarcinoma (PAAD) and gynaecological tumours. It was also confirmed to be a biomarker for early diagnosis and prognostic evaluation (8-14). Studies by Wang and Zhuang et al. shown that B3GNT3 overexpression promotes tumor migration and chemoresistance by inhibiting CD8⁺ T cell infiltration and activating the RhoA/RAC1 pathway (8,15). Additionally, B3GNT3 is crucial for the epidermal growth factor-induced signaling of programmed cell death protein 1 (PD-1) and programmed death-ligand 1 (PD-L1) in triple-negative breast cancer (13,16). Down-regulation of B3GNT3 may enhance cytotoxic T cell-mediated antitumor effects. Notably, B3GNT3 expression levels are closely associated with the dynamic remodeling of the glycocalyx, and it may influence tumor cell interactions with the microenvironment by modifying mucin-type glycoproteins, such as podocalyxin (6). However, the regulatory networks, underlying mechanisms, and targeted therapeutic strategies of B3GNT3 in the context of pan-cancer research warrant further investigation.

In our study, we first utilized databases such as The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO) to conduct a pan-cancer analysis of the B3GNT3 gene. We comprehensively analyzed multiple factors, including gene expression, protein expression, survival status, genetic alterations, TME and associated cellular pathways, to investigate the potential molecular mechanisms of B3GNT3 in the oncogenic mechanisms or clinical outcomes of various cancers. Our findings provide comprehensive insights into the oncogenic role of B3GNT3 and highlight its potential as a prognostic biomarker in cancer. We present this article in accordance with the REMARK reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-748/rc).

Methods

Source of research data

Public data were obtained from TCGA program (https://portal.gdc.cancer.gov/) and Genotype-Tissue Expression (GTEx) (https://gtexportal.org/home/) website. The detailed analysis workflow is presented in Figure S1. In brief, analyses of B3GNT3 gene expression, its correlation with pathological stage, survival analysis and co-expressed gene correlation were performed using Gene Expression Profiling Interactive Analysis 2.0 (GEPIA2.0) (http://gepia2.cancerpku.cn/) (17). The correlation between B3GNT3 and cancer-associated fibroblast (CAF) infiltration was analyzed via Tumor Immune Response Estimation and Analysis 2.0 (TIMER2.0) (http://timer.cistrome.org/) (18). Protein expression levels of B3GNT3 in tumors were analyzed through University of Alabama Cancer Database (UALCAN) (http://ualcan.path.uab.edu/) (19) and the Human Protein Atlas (HPA) (https://www.proteinatlas.org/) (20). Protein-protein interaction (PPI) network analysis was conducted using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database (https://string-db.org/) (21). Variants in B3GNT3 were analyzed for frequency and type using the cBioPortal platform (https://www.cbioportal.org/) (22). The online Database for Annotation, Visualization and Integrated Discovery (DAVID) (https://davidbioinformatics.nih.gov/) was used for Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) pathway analysis (23,24).

Analysis of gene expression

Gene expression analysis was performed using the GEPIA2.0 database. First, the “Expression DIY_Profile” module was used to analyze the expression differences of B3GNT3 in 33 types of tumor tissues and their corresponding normal tissues. Then, the “Box Plots” module was used to analyze the expression differences of B3GNT3 in 33 types of cancers. To investigate the protein expression levels of B3GNT3 in tumor and normal control tissues, comparison box plots were generated using Clinical Proteomic Tumor Analysis Consortium (CPTAC) data from the UALCAN database. Additionally, data were downloaded from the HPA database to analyze the expression of B3GNT3 in different human tissues and organs. Finally, the “Survival Map” module of the GEPIA2.0 database was used to study the relationship between B3GNT3 expression and tumor pathological staging.

Analysis of survival

Survival analysis was performed using the “Survival Analysis” module of the GEPIA2.0 database to evaluate the relationship between B3GNT3 mRNA expression and OS and disease-free survival (DFS) in TCGA cancer patients. High- and low-expression groups were determined using the upper and lower 50% expression cutoffs, respectively. The log-rank test was used for hypothesis testing, and survival curves were generated via the “Survival Analysis” module of GEPIA2.0.

Analysis of genetic variation

Genetic variation analysis of B3GNT3 was performed using the “TCGA Pan-cancer Atlas Studies” dataset of cBioPortal platform. This analysis included the distribution of mutation types across domains in the “Mutations” module. Additionally, we randomly selected the “Comparison” module of the lung squamous cell carcinoma (LUSC) dataset to investigate the correlation between B3GNT3 mutation status and TCGA cancer patients’ OS, DFS, disease-specific survival (DSS) and progression-free survival (PFS). Kaplan-Meier survival curves were generated, and log-rank test P values were calculated.

Analysis of immune microenvironment

The immune microenvironment analysis of B3GNT3 was performed using the “Immune” module of the TIMER2.0 online database. Four algorithms-Estimating the Proportions of Immune and Cancer cells (EPIC), Microenvironment Cell Populations-counter (MCP_COUNTER), Tumor Immune Dysfunction and Exclusion (TIDE) and Cell Type Enrichment Analysis (XCELL) were applied to evaluate the correlation between B3GNT3 and CAFs across different types of cancer. Spearman rank correlation tests, adjusted for purity, were used to obtain P values and partial correlation (cor) values. The data were visualized via heatmaps and scatter plots. To ensure reliability, results consistent across all four algorithms were selected to assess B3GNT3 infiltration in tumors. This approach comprehensively reflects the potential association between B3GNT3 and CAFs, providing a robust foundation for future research.

Analysis of function and mechanism

We investigated potential B3GNT3-binding proteins via the STRING database, setting parameters for network edge measurements (evidence), active interaction sources (experiments), minimum required interaction scores (low confidence) and a maximum of 50 displayed participants. Using the “Similar Genes Detection” module of GEPIA2.0, we analyzed the top 100 genes correlated with B3GNT3 expression across all TCGA tumor and normal tissue datasets. Visualization included partial Spearman rank correlation coefficients (cor) and P values adjusted for tumor purity. Additionally, the TIMER2.0 “Gene_Corr” module was used for pairwise Pearson correlation analysis between B3GNT3 and the five most similar genes selected, with scatter plots generated using log₂ transformed transcripts per million (TPM) values and annotated with P values and correlation coefficients (R). Finally, pathway enrichment analysis was performed in the DAVID online database to explore the functions and related pathways of these genes.

Analysis of human tumor tissue and main reagents

In accordance with the recommendations of the Anatomic Pathology Committee (AP) (25) and the Anatomic Pathology Patient Interest Association (APPIA) (26), a minimum of three histological staining should be performed to ensure the repeatability of validation. Therefore, in our study, we randomly selected 15 real-world cases each of PAAD, colon adenocarcinoma (COAD), lung adenocarcinoma (LUAD), uterine corpus endometrial cancer (UCEC), and ovarian cancer (OV) from The Eighth Affiliated Hospital of Southern Medical University (The First People’s Hospital of Shunde, Foshan). All cases were confirmed by two senior pathologists through slide review. For each case, tumor tissue and adjacent normal pancreatic tissue (about one centimeter from the tumor) were collected and stained via immunohistochemistry on a Roche Ventana instrument (Medical Systems, Tucson, USA), following a strict standard operating procedure (SOP). The immunohistochemical reagents were purchased from Proteintech Group, Inc. (Catalog No. 18098-1-AP) and used at a dilution of 1:100. The reagent was validated for laboratory performance prior to use and used and stored according to the manufacturer’s instructions.

Immunohistochemical (IHC) staining

Automated Ventana IHC assay was carried out in this study according to the manufacturer’s instructions. Briefly, tissue slides were dewaxed using 1X EZ prep solution (supplied by Ventana Medical Systems, Tucson, USA) through heat and vortex mixing. Heated paraffin wax floated out from tissue sections through the aqueous solution and was efficiently removed by vortex mixing. For antigen retrieval, Cell Conditioning one (CC1, supplied by Ventana Medical Systems, Tucson, USA) was used in the presence of heat. The tissue was then incubated with the pre-diluted primary antibody, along with the UltraView DAB (3,3'-Diaminobenzidine) IHC detection kit, on the Benchmark UltraView stainer. Each case was also stained with a matched rabbit/mouse monoclonal negative control IgG antibody (27).

Staining was evaluated using a semiquantitative scoring method, where scores for staining intensity (one to four points) and the percentage of positive cells were multiplied to obtain a final score. Staining intensity was scored as one point (negative), two points (weak, light yellow), three points (positive, brownish yellow) or four points (strong, brown).

Statistical analysis

Statistical analysis was performed using GraphPad Prism 8.0.1 software. Comparisons between groups were conducted using the two-sample independent t-test. Differences were considered statistically significant when P<0.05.

Ethics statement

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by ethics board of The Eighth Affiliated Hospital of Southern Medical University (The First People’s Hospital of Shunde, Foshan) (No. KYLS20250401) and informed consent was obtained from all individual participants.

Results

B3GNT3 was expressed inconsistently in tumor tissues of different cancers and was associated with tumor progression

To investigate the mRNA expression levels of B3GNT3 (transcript ID: NM_014256) across all TCGA tumor types, we first conducted a pan-cancer analysis using the “Expression DIY” module of the GEPIA2.0 database (Figure 1A). Subsequent single-pair analyses for individual cancer types revealed that B3GNT3 was upregulated in tumor tissues compared to adjacent normal tissues in 15 cancer types: bladder urothelial carcinoma (BLCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), COAD, kidney renal papillary cell carcinoma (KIRP), liver hepatocellular carcinoma (LIHC), LUAD, LUSC, OV, PAAD (Figure 1B), rectum adenocarcinoma (READ), stomach adenocarcinoma (STAD), thyroid carcinoma (THCA), UCEC and uterine carcinosarcoma (UCS) (Figure 1C). In contrast, B3GNT3 was downregulated in tumor tissues in skin cutaneous melanoma (SKCM) (Figure 1D), sarcoma (SARC) and kidney chromophobe (KICH) (Figure 1E). Overall, B3GNT3 exhibited upregulated expression in the majority of tumor types, suggesting its potential oncogenic role in these cancers.

Figure 1 Expression profile of B3GNT3 in different tumors. (A) Expression of B3GNT3 in 33 types of tumor tissues. Red font is upregulation, green font is downregulation, black font is no significant difference. (B-E) Expression of B3GNT3 in PAAD, UCS, SKCM, KICH tumor tissues and corresponding normal tissues. *, P<0.05. The red box indicates the tumor group, and the blue box indicates the normal group. (F-L) Relationship between expression of B3GNT3 and pathological stage in BLCA, ESCA, KIRP, LUAD, PAAD, THCA and UCS. ACC, adrenocortical carcinoma; B3GNT3, beta-1,3-N-acetylglucosaminyltransferase 3; BLCA, bladder urothelial carcinoma; BRCA, breast invasive carcinoma; CESC, cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL, cholangiocarcinoma; COAD, colon adenocarcinoma; DLBC, lymphoid neoplasm diffuse large B-cell lymphoma; ESCA, esophageal carcinoma; GBM, glioblastoma multiforme; HNSC, head and neck squamous cell carcinoma; KICH, kidney chromophobe; KIRC, kidney renal clear cell carcinoma; KIRP, kidney renal papillary cell carcinoma; LAML, acute myeloid leukemia; LGG, brain lower grade glioma; LIHC, liver hepatocellular carcinoma; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; MESO, mesothelioma; N, normal; OV, ovarian serous cystadenocarcinoma; PAAD, pancreatic adenocarcinoma; PCPG, pheochromocytoma and paraganglioma; PRAD, prostate adenocarcinoma; READ, rectum adenocarcinoma; SARC, sarcoma; SKCM, skin cutaneous melanoma; STAD, stomach adenocarcinoma; T, tumor; TGCT, Testicular Germ Cell Tumors; THCA, thyroid carcinoma; THYM, thymoma; TPM, transcripts per million; UCEC, uterine corpus endometrial carcinoma; UCS, uterine carcinosarcoma; UVM, uveal melanoma.

We further explored the relationship between B3GNT3 expression and pathological staging using the “Expression Analysis” module of GEPIA2.0. Results showed that B3GNT3 expression was associated with pathological staging in several cancers, including BLCA (Figure 1F), esophageal carcinoma (ESCA) (Figure 1G), KIRP (Figure 1H), LUAD (Figure 1I), PAAD (Figure 1J), THCA (Figure 1K) and UCS (Figure 1L). This suggests that the high expression of B3GNT3 may be closely associated with carcinogenesis and progression, which is in line with the findings of Kong et al. (9-13) in their studies on ESCA, LUAD, and PAAD.

B3GNT3 protein expression level was different in different types of cancer

We analyzed B3GNT3 protein levels using the CPTAC database. Results showed that B3GNT3 was significantly overexpressed in COAD (Figure 2A), PAAD (Figure 2B), and UCEC (Figure 2C), but underexpressed in head and neck squamous cell carcinoma (HNSC) (Figure 2D), LUAD (Figure 2E), LUSC (Figure 2F) and OV (Figure 2G). Additionally, immunohistochemical staining of tumor tissues via the HPA database revealed moderate cytoplasmic/membranous staining in more than 75% of pancreatic adenocarcinoma cells (Figure 2H), compared to weak staining in less than 25% of normal pancreatic ductal epithelial cells (Figure 2I). In hepatocellular carcinoma, B3GNT3 was completely negative in tumor cells (Figure 2J), while corresponding liver tissues showed moderate cytoplasmic/membranous staining in more than 75% of cells (Figure 2K). These protein expression differences across tumor types suggest that B3GNT3 may have distinct roles and mechanisms in various cancers, indicating complex regulatory mechanisms in vivo.

Figure 2 Protein analysis of B3GNT3 in 33 types of tumors. (A-G) Expression of B3GNT3 in COAD, PAAD, UCEC, HNSC, LUAD, LUSC and OV tumor tissues and corresponding normal tissues. ***, P<0.001. (H-K) IHC staining of B3GNT3 (HPA ID Number: HPA024298) in PAAD and LIHC tumor tissues and corresponding normal tissues from the Human Protein Atlas database (https://www.proteinatlas.org/). Image credit goes to the HPA. The links to the staining of B3GNT3 protein in tumor tissue and corresponding control tissue are as follows: PAAD_Tumor (https://www.proteinatlas.org/ENSG00000179913-B3GNT3/cancer/pancreatic+cancer#img), PAAD_Normal (https://www.proteinatlas.org/ENSG00000179913-B3GNT3/tissue/pancreas#img), LIHC_Tumor (https://www.proteinatlas.org/ENSG00000179913-B3GNT3/cancer/liver+cancer#img), LIHC_Normal (https://www.proteinatlas.org/ENSG00000179913-B3GNT3/tissue/liver#img). Scale bar, 100 μm. B3GNT3, beta-1,3-N-acetylglucosaminyltransferase 3; COAD, colon adenocarcinoma; HPA, Human Protein Atlas; HNSC, head and neck squamous cell carcinoma; IHC, immunohistochemical; LIHC, liver hepatocellular carcinoma; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; NOS, not otherwise specified; OV, ovarian serous cystadenocarcinoma; PAAD, pancreatic adenocarcinoma; UCEC, uterine corpus endometrial carcinoma.

B3GNT3 impacted the survival of patients with different types of cancer

Using the “Survival Analysis” module of GEPIA2.0 database, we evaluated the prognostic significance of B3GNT3 across 33 cancer types (Figure 3A). Results showed that high B3GNT3 expression was positively associated with OS in kidney renal clear cell carcinoma (KIRC) (Figure 3B) and READ (Figure 3C) patients, but negatively associated in mesothelioma (MESO) (Figure 3D), OV (Figure 3E) and PAAD (Figure 3F) patients.

Figure 3 Survival analysis of B3GNT3 in 33 types of tumors. (A) Analysis of B3GNT3 expression and overall survival in 33 types of tumors. (B-F) Analysis of B3GNT3 expression and overall survival in KIRC, READ, MESO, OV and PAAD. (G) Analysis of B3GNT3 expression and disease-free survival in 33 types of tumors. (H-K) Analysis of B3GNT3 expression and disease-free survival in CESC, CHOL, PAAD and KIRC. ACC, adrenocortical carcinoma; B3GNT3, beta-1,3-N-acetylglucosaminyltransferase 3; BLCA, bladder urothelial carcinoma; BRCA, breast invasive carcinoma; CESC, cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL, cholangiocarcinoma; COAD, colon adenocarcinoma; DLBC, lymphoid neoplasm diffuse large B-cell lymphoma; ESCA, esophageal carcinoma; GBM, glioblastoma multiforme; HNSC, head and neck squamous cell carcinoma; HR, hazard ratio; KICH, kidney chromophobe; KIRC, kidney renal clear cell carcinoma; KIRP, kidney renal papillary cell carcinoma; LAML, acute myeloid leukemia; LGG, brain lower grade glioma; LIHC, liver hepatocellular carcinoma; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; MESO, mesothelioma; OV, ovarian serous cystadenocarcinoma; PAAD, pancreatic adenocarcinoma; PCPG, pheochromocytoma and paraganglioma; PRAD, prostate adenocarcinoma; READ, rectum adenocarcinoma; SARC, sarcoma; SKCM, skin cutaneous melanoma; STAD, stomach adenocarcinoma; TGCT, testicular germ cell tumors; THCA, thyroid carcinoma; THYM, thymoma; UCEC, uterine corpus endometrial carcinoma; UCS, uterine carcinosarcoma; UVM, uveal melanoma.

DFS analysis (Figure 3G) revealed that B3GNT3 upregulation was significantly linked to poor prognosis in CESC (Figure 3H), CHOL (Figure 3I) and PAAD (Figure 3J) patients. In contrast, in KIRC (Figure 3K), B3GNT3 downregulation was associated with poor prognosis. These findings indicate that B3GNT3 expression correlates differentially with patient prognosis across cancer types. However, its impacts on both OS and DFS were consistent in KIRC (positive correlation) and PAAD (negative correlation).

B3GNT3 alteration analysis in different types of cancer

Using the cBioPortal database, we analyzed B3GNT3 alterations in TCGA cancers. Results showed that amplification was the primary alteration type, followed by mutation (Figure 4A). Specifically, UCS, CHOL, MESO, ESCA, uveal melanoma (UVM) and adrenocortical carcinoma (ACC) exhibited only amplification, while PAAD and BLCA showed only mutation. No B3GNT3 alterations were observed in thymoma (THYM), testicular germ cell tumors (TGCT), Pheochromocytoma and Paraganglioma (PCPG), KIRP, KIRC, KICH, lymphoid neoplasm diffuse large B-cell lymphoma (DLBC) and acute myeloid leukemia (AML). Most of other cancers showed coexistence of amplification and mutation, such as in LUSC, where 2.05% (10/487) of B3GNT3 alterations were missense mutations and 0.62% (3/487) were amplifications. We further explored the B3GNT3 mutation profile (Figure 4B). Survival analysis in LUSC patients revealed that B3GNT3 mutations were associated with shorter DFS (P=0.02) (Figure 4C-4F). These findings suggest that B3GNT3 mutations may contribute to tumorigenesis.

Figure 4 Genetic variation analysis of B3GNT3 in tumors. (A) Mutation types of B3GNT3 in 24 types of tumors. (B) Mutations in various domains of B3GNT3. (C-F) Correlation between B3GNT3 mutation status and OS, DFS, DSS and PFS of LUSC, the altered group includes all of mutation types. ACC, adrenocortical carcinoma; AML, acute myeloid leukemia; B3GNT3, beta-1,3-N-acetylglucosaminyltransferase 3; BLCA, bladder urothelial carcinoma; BRCA, breast invasive carcinoma; CESC, cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL, cholangiocarcinoma; CNV, copy number variation; COAD, colon adenocarcinoma; DFS, disease-free survival; DLBC, lymphoid neoplasm diffuse large B-cell lymphoma; DSS, disease-specific survival; ESCA, esophageal carcinoma; GBM, glioblastoma multiforme; HNSC, head and neck squamous cell carcinoma; KICH, kidney chromophobe; KIRC, kidney renal clear cell carcinoma; KIRP, kidney renal papillary cell carcinoma; LGG, brain lower grade glioma; LIHC, liver hepatocellular carcinoma; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; MESO, mesothelioma; OV, ovarian serous cystadenocarcinoma; OS, overall survival; PAAD, pancreatic adenocarcinoma; PCPG, pheochromocytoma and paraganglioma; PFS, progression-free survival; PRAD, prostate adenocarcinoma; SARC, sarcoma; SKCM, skin cutaneous melanoma; STAD, stomach adenocarcinoma; SV, structural variation; TGCT, testicular germ cell tumors; THCA, thyroid carcinoma; THYM, thymoma; UCEC, uterine corpus endometrial carcinoma; UCS, uterine carcinosarcoma; UVM, uveal melanoma.

Correlation analysis of B3GNT3 expression and CAFs in different types of cancer

CAFs, a key component of the TME, are highly heterogeneous and activated fibroblasts. They promote tumor progression, migration, inflammation, and drug resistance by secreting chemokines and cytokines (28-30), and modulate the functions of various tumor-infiltrating immune cells (31,32). Using the “Immune” module of TIMER2.0 database, we analyzed the potential correlation between B3GNT3 expression and CAFs. As shown in Figure 5A, consistent results were observed across four algorithms in ten cancer types: breast cancer-luminal A (BRCA-LumA) (Figure 5B), LIHC (Figure 5C), MESO (Figure 5D), SKCM-metastasis (Figure 5E), TGCT (Figure 5F), ESCA (Figure 5G), HNSC (Figure 5H), HNSC-human papillomavirus- (HNSC-HPV-) (Figure 5I), STAD (Figure 5J) and UCEC (Figure 5K). Darker colors indicate higher correlations. Notably, B3GNT3 expression was positively correlated with CAFs in BRCA-LumA (Figure 5B, Rho =0.149, P<0.001), LIHC (Figure 5C, Rho =0.423, P<0.001), MESO (Figure 5D, Rho =0.426, P<0.001), SKCM-metastasis (Figure 5E, Rho =0.133, P=0.01), and TGCT (Figure 5F, Rho =0.691, P<0.001). These findings suggest that B3GNT3 may drive tumor progression by regulating CAFs.

Figure 5 Analysis of B3GNT3 in immune microenvironment. (A) Correlation between B3GNT3 and CAFs calculated by four algorithms. (B-K) The tumors were correlated with expression level of B3GNT3 in BRCA-LumA, LIHC, MESO, SKCM-metastasis, TGCT, ESCA, HNSC, HNSC-HPV-, STAD, and UCEC. The ordinate is the expression of B3GNT3 transformed by log₂ (TPM). ACC, adrenocortical carcinoma; B3GNT3, beta-1,3-N-acetylglucosaminyltransferase 3; BLCA, bladder urothelial carcinoma; BRCA, breast invasive carcinoma; CAF, cancer-associated fibroblast; CESC, cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL, cholangiocarcinoma; COAD, colon adenocarcinoma; DLBC, lymphoid neoplasm diffuse large B-cell lymphoma; EPIC, estimating the proportions of immune and cancer cells; ESCA, esophageal carcinoma; GBM, glioblastoma multiforme; HNSC, head and neck squamous cell carcinoma; HPV, human papillomavirus; KICH, kidney chromophobe; KIRC, kidney renal clear cell carcinoma; KIRP, kidney renal papillary cell carcinoma; LAML, acute myeloid leukemia; LGG, brain lower grade glioma; LIHC, liver hepatocellular carcinoma; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; MCP_COUNTER, microenvironment cell populations-counter; MESO, mesothelioma; OV, ovarian serous cystadenocarcinoma; PAAD, pancreatic adenocarcinoma; PCPG, pheochromocytoma and paraganglioma; PRAD, prostate adenocarcinoma; READ, rectum adenocarcinoma; SARC, sarcoma; SKCM, skin cutaneous melanoma; STAD, stomach adenocarcinoma; TGCT, testicular germ cell tumors; THCA, thyroid carcinoma; THYM, thymoma; TIDE, tumor immune dysfunction and exclusion; TPM, transcripts per million; UCEC, uterine corpus endometrial carcinoma; UCS, uterine carcinosarcoma; UVM, uveal melanoma; XCELL, cell type enrichment analysis.

The potential molecular mechanism of B3GNT3 in tumorigenesis was tight junction (TJ)

To explore the potential molecular mechanisms of B3GNT3 in tumorigenesis, we conducted a PPI network analysis via the STRING online tool, identifying 50 proteins interacting with B3GNT3 and their interaction network (Figure 6A). We also integrated TCGA tumor expression data into the GEPIA2.0 database to obtain the top 100 genes correlated with B3GNT3 expression. We analyzed the correlation between B3GNT3 and the expression levels of the top five genes using TIMER2.0, visualizing the results in Figure 6B. Further analysis in GEPIA2.0 revealed positive correlations between B3GNT3 and these five genes, with correlation coefficients of plastin 1 (PLS1) (R=0.68) (Figure 6C), mitotic spindle positioning (MISP) (R=0.68) (Figure 6D), G protein-coupled receptor 35 (GPR35) (R=0.67) (Figure 6E), epidermal growth factor receptor pathway substrate 8 like 3 (EPS8L3) (R=0.66) (Figure 6F) and fucosyltransferase 3 (FUT3) (R=0.65) (Figure 6G), all with P<0.001. Subsequent pathway enrichment analyses showed that KEGG analysis highlighted “Tight Junction” (Figure 6H) as the most significant mechanism related to B3GNT3 in tumorigenesis. GO analysis indicated these genes are involved in metabolic pathways such as “protein O-linked glycosylation” (Figure 6I), “bicellular tight junction” (Figure 6J) and “actin filament binding” (Figure 6K). These findings reveal the potential molecular mechanisms of the B3GNT3 gene in tumor development.

Figure 6 Analysis of molecular mechanism of B3GNT3 in tumorigenesis. (A) Analysis of protein-protein interaction network of B3GNT3. (B) Correlation between B3GNT3 and the first five genes in different types of tumors. (C-G) Analysis of the correlation between the expression of B3GNT3 and PLS1, MISP, GPR35, EPS8L3 and FUT3 in different types of tumors. (H-K) Pathway enrichment analysis of the first 100 genes related to B3GNT3. ACC, adrenocortical carcinoma; B3GNT3, beta-1,3-N-acetylglucosaminyltransferase 3; BLCA, bladder urothelial carcinoma; BP, biological process; BRCA, breast invasive carcinoma; CC, cellular component; CESC, cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL, cholangiocarcinoma; COAD, colon adenocarcinoma; DLBC, lymphoid neoplasm diffuse large B-cell lymphoma; EPS8L3, epidermal growth factor receptor pathway substrate 8 like 3; ESCA, esophageal carcinoma; FUT3, fucosyltransferase 3; GBM, glioblastoma multiforme; GO, Gene Ontology; GPR35, G protein-coupled receptor 35; HNSC, head and neck squamous cell carcinoma; KEGG, Kyoto Encyclopedia of Genes and Genomes; KICH, kidney chromophobe; KIRC, kidney renal clear cell carcinoma; KIRP, kidney renal papillary cell carcinoma; LGG, brain lower grade glioma; LIHC, liver hepatocellular carcinoma; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; MESO, mesothelioma; MF, molecular function; MISP, mitotic spindle positioning; OV, ovarian serous cystadenocarcinoma; PAAD, pancreatic adenocarcinoma; PCPG, pheochromocytoma and paraganglioma; PLS1, plastin 1; PRAD, prostate adenocarcinoma; READ, rectum adenocarcinoma; SARC, sarcoma; SKCM, skin cutaneous melanoma; STAD, stomach adenocarcinoma; TGCT, testicular germ cell tumors; THCA, thyroid carcinoma; THYM, thymoma; TPM, transcripts per million; UCEC, uterine corpus endometrial carcinoma; UCS, uterine carcinosarcoma; UVM, uveal melanoma.

B3GNT3 was highly expressed in real-world patients with PAAD, COAD, LUAD, UCEC, and OV

In our prior analysis, we observed that in patients with PAAD, B3GNT3 exhibited consistent associations across gene expression, pathological staging, protein expression, and prognostic survival. To verify whether B3GNT3 is significantly overexpressed in real-world PAAD patient cases, we selected tumor tissues and corresponding control tissues from 15 PAAD patients for immunohistochemical staining. The results showed that normal pancreatic epithelial cells were almost unstained, whereas in PAAD lesions, more than 75% of tumor cells exhibited at least moderate cell membrane/cytoplasmic staining intensity (Figure 7A). The staining scores showed a significant difference, with P<0.001 (Figure 7B), which is consistent with the study by Kong et al. (13). In COAD, normal intestinal epithelium exhibited weakly positive staining in approximately 35% of cells, with a few cells showing moderate positive staining. In contrast, in COAD tumor tissues, over 75% of cells displayed diffuse and uniform positive staining of at least moderate intensity (Figure 7C). The staining score results revealed a significant difference, with P<0.001 (Figure 7D). Similarly, in immunohistochemical experiments of tumor tissues from LUAD, UCEC, and OV, B3GNT3 was found to be significantly overexpressed (Figure S2). The results of real-world immunohistochemical validation experiments suggest that B3GNT3 may be an important factor in the formation of tumor. A thorough investigation of its molecular mechanisms of action could play a crucial role in tumor diagnosis and treatment, and B3GNT3 holds promise as a novel tumor biomarker.

Figure 7 Validation of case through immunohistochemical staining. (A) Representative images of PAAD are shown (H&E, 40×; IHC, 40×; IHC, 100×). (B) The staining intensity is quantified by immunohistochemical score, results show that B3GNT3 is significantly upregulated in PAAD tumor tissues compared to normal tissues (P<0.001). (C) Representative images of COAD are shown (H&E, 40×; IHC, 40×; IHC, 100×). (D) The staining intensity is quantified by immunohistochemical score, results show that B3GNT3 is significantly upregulated in COAD tumor tissues compared to normal tissues (P<0.001). B3GNT3, beta-1,3-N-acetylglucosaminyltransferase 3; COAD, colon adenocarcinoma; H&E, hematoxylin and eosin; IHC, immunohistochemical; PAAD, pancreatic adenocarcinoma.

Discussion

B3GNT3, a key glycosyltransferase family member, catalyzes the formation of β-1,3-glycosidic bonds between N-acetylglucosamine (GlcNAc) and galactose (Gal), participating in various glycoconjugate biosyntheses (5,6). Recently, its role in tumorigenesis has gained research attention. Pan-cancer analyses, integrating multi-omics data and experimental validation, show that B3GNT3 is significantly dysregulated in multiple malignancies (8-14), including gastric, thyroid, and endometrial cancers. As a key glycosylation regulator, B3GNT3 exhibits oncogenic properties in pan-cancer analysis, with its expression levels closely related to patient prognosis, the TME, and therapy resistance.

Using a pan-cancer analysis approach, we explored B3GNT3 expression in 33 types of tumor tissues relative to their corresponding normal tissues. B3GNT3 was upregulated in 15 and downregulated in three types of cancer, suggesting an oncogenic role across cancer types. Previous studies show B3GNT3 promotes esophageal cancer growth, invasion and migration (9) and may serve as a prognostic biomarker in cervical (33) and non-small cell lung cancers (12). Zhou et al. demonstrated that B3GNT3 drives gastric cancer progression by upregulating EphA2 and activating PI3K/AKT signaling, highlighting its potential clinical relevance in diagnosis and therapy (34). Research by Zhou et al. indicated that ETV4 can bind to the B3GNT3 promoter, activate its transcription, and thereby enhance liver cancer cell proliferation, migration, and invasion. In addition, B3GNT3 may act as an intermediary in the activation of TGF-β signaling by ETV4 (14). Given the high mortality from digestive system cancers in China, further research into the role of B3GNT3 in tumorigenesis and its molecular mechanisms is crucial for developing effective therapeutic targets.

Pan-cancer analysis in this study revealed the impact of B3GNT3 on patient survival across multiple cancer types. In cervical and ovarian cancer patients, high B3GNT3 expression was associated with a 30–40% reduction in 5-year OS compared to controls, with a hazard ratio (HR) of 1.89 [95% confidence interval (CI): 1.32–2.71] in endometrial cancer (7). In LUAD, patients with high B3GNT3 expression showed significantly shortened recurrence-free survival (RFS) and OS and multivariate Cox regression analysis confirmed it as an independent prognostic factor (HR =1.19, 95% CI: 1.08–1.32) (35). Mechanistically, B3GNT3 may mediate immune evasion by regulating PD-L1 expression. Experiments demonstrated that silencing B3GNT3 significantly increased tumor cell apoptosis and suppressed proliferation, indicating its potential as a therapeutic target, which is consistent with the findings of Li et al. (16). In our study, upregulation of B3GNT3 was negatively correlated with both OS and DFS in PAAD patients. Combined with existing research findings (7,10,12,35,36) and our analysis of the oncogenic role of B3GNT3 and protein expression in PAAD, COAD, LUAD, UCEC and OV, we conclude that B3GNT3 upregulation may be a potential biomarker for poor prognosis in PAAD, COAD, LUAD, UCEC and OV. However, the clinical application value of B3GNT3 still requires further exploration based on a larger number of clinical samples and a deeper understanding of its underlying mechanisms. This includes the potential for non-invasive and repeatable detection methods, such as blood testing.

CAFs, a key stromal component of the TME, have been shown to regulate tumor progression through multifaceted mechanisms (37-39). CAFs secrete cytokines like transforming growth factor (TGF)-β and interleukin-6 (IL-6), remodeling the extracellular matrix (ECM) and activating downstream pathways to promote tumor cell proliferation, invasion, and metastasis (37,40). They also contribute to tumor drug resistance, angiogenesis, and inflammation (41). Using spatial transcriptomics, Croft et al. constructed a spatial interaction map between CAFs and tumor cells, offering new directions for microenvironment-specific therapies. Elucidating CAF heterogeneity and dynamic regulatory networks is crucial for breaking through cancer therapy bottlenecks (42). Our analysis identified a consistent correlation between CAFs abundance and B3GNT3 expression across four distinct algorithms in ten cancer types: BRCA-LumA, SKCM-metastasis, LIHC, MESO, TGCT, ESCA, HNSC, HNSC-HPV-, STAD, and UCEC. These findings suggest that B3GNT3 may play a role in modulating CAF activation and formation.

Our analysis reveals that B3GNT3 has a dual regulatory effect on the tumor immune microenvironment. On one hand, its high expression is associated with increased infiltration of immunosuppressive cells, such as M2 macrophages and regulatory T cells. In cervical cancer, B3GNT3 expression positively correlates with CD163⁺ macrophage density (R=0.67, P=0.003) (7). On the other hand, B3GNT3 may influence T cell recruitment by regulating chemokine receptors like CCR4 and CXCR3. Experiments showed that inhibiting B3GNT5 (a family member) reduced SSEA-1 expression on T cells and weakened antitumor immunity (43), indicating a conserved immunomodulatory role of the B3GNT family. Notably, single-cell transcriptomic analysis found spatial colocalization between the collagen-CD44 signaling axis in CAFs and B3GNT3 expression, which may promote an immune-exclusion phenotype by remodeling the ECM (44,45). However, the specific immunomodulatory mechanisms of B3GNT3 in different types of cancer require further study.

Finally, we analyzed the coexpression levels of the top five genes (PLS1, MISP, GPR35, EPS8L3, FUT3) similar to B3GNT3 via PPI networks and the GEPIA2.0 and TIMER2.0 databases. KEGG pathway enrichment analysis suggested that B3GNT3 may contribute to tumorigenesis through TJs and impact clinical outcomes. TJs, crucial for maintaining cell polarity and barrier function in epithelial and endothelial cells, also participate in tumor development by regulating signaling transduction, cell proliferation, and migration (46-49). GO pathway enrichment analysis revealed significant enrichment in “protein O-linked glycosylation”, “bicellular tight junction” and “actin filament binding” indicating potential roles in critical mechanisms such as tumor occurrence, metastasis, and microenvironment remodeling (50,51). These molecular interaction networks offer new insights for the development of targeted cancer therapies.

In summary, our pan-cancer analysis of B3GNT3 reveals statistically significant correlations between its expression and clinical outcomes, pathological stages, and immune cell infiltration across multiple cancers. We have also preliminarily explored the potential oncogenic mechanisms of B3GNT3, providing a foundation for understanding its clinical significance in tumorigenesis and progression. Future research should focus on elucidating the specific roles of B3GNT3 in the TME, particularly its interactions with CAFs and immune cells, and investigate the potential therapeutic benefits of combining glycosylation inhibitors targeting B3GNT3 with immune checkpoint inhibitors.

Conclusions

Pan-cancer analysis of B3GNT3 reveals complex regulatory mechanisms across different types of cancer, emerging as a novel biomarker for cancer diagnosis and prognosis, and offers new avenues for targeted therapeutic development.

Acknowledgments

We would like to thank the authors of previous studies and the staff members of the cBioPortal, TCGA and GTEx for providing the available data.

Footnote

Reporting Checklist: The authors have completed the REMARK reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-748/rc

Data Sharing Statement: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-748/dss

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-748/prf

Funding: This study was supported by funding from the National Natural Scientific Foundation of China (No. 82303994) and the Scientific Research Start Plan of The Eighth Affiliated Hospital of Southern Medical University (Nos. SRSP2021006, SRSP2022003, SRSP2024008, SRSP2023025).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-748/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by ethics board of The Eighth Affiliated Hospital of Southern Medical University (The First People’s Hospital of Shunde, Foshan) (No. KYLS20250401) and informed consent was obtained from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
2. Han B, Zheng R, Zeng H, et al. Cancer incidence and mortality in China, 2022. J Natl Cancer Cent 2024;4:47-53. [Crossref] [PubMed]
3. Liu C, Yuan YC, Guo MN, et al. Rising incidence of obesity-related cancers among younger adults in China: A population-based analysis (2007-2021). Med 2024;5:1402-1412.e2. [Crossref] [PubMed]
4. Zhou K, Huo J, Gao C, et al. Applying T-classifier, binary classifiers, upon high-throughput TCR sequencing output to identify cytomegalovirus exposure history. Sci Rep 2023;13:5024. [Crossref] [PubMed]
5. Xie A, Wang J, Liu Y, et al. Impacts of β-1, 3-N-acetylglucosaminyltransferases (B3GNTs) in human diseases. Mol Biol Rep 2024;51:476. [Crossref] [PubMed]
6. Shi SM, Suh RJ, Shon DJ, et al. Glycocalyx dysregulation impairs blood-brain barrier in ageing and disease. Nature 2025;639:985-94. [Crossref] [PubMed]
7. Xu J, Guo Z, Yuan S, et al. Upregulation of B3GNT3 is associated with immune infiltration and activation of NF-κB pathway in gynecologic cancers. J Reprod Immunol 2022;152:103658. [Crossref] [PubMed]
8. Wang JS, Ruan F, Guo LZ, et al. B3GNT3 acts as a carcinogenic factor in endometrial cancer via facilitating cell growth, invasion and migration through regulating RhoA/RAC1 pathway-associated markers. Genes Genomics 2021;43:447-57. [Crossref] [PubMed]
9. Lu J, Lei T, Yu H, et al. The role of B3GNT3 as an oncogene in the growth, invasion and migration of esophageal cancer cells. Acta Cir Bras 2023;38:e380923. [Crossref] [PubMed]
10. Yang H, Chen Y, Huang X, et al. Bioinformatics Analysis Reveals a Novel Prognostic Model for Esophageal Squamous Cell Carcinoma. Int J Med Sci 2024;21:1213-26. [Crossref] [PubMed]
11. Sun Y, Liu T, Xian L, et al. B3GNT3, a Direct Target of miR-149-5p, Promotes Lung Cancer Development and Indicates Poor Prognosis of Lung Cancer. Cancer Manag Res 2020;12:2381-91. [Crossref] [PubMed]
12. Gao L, Zhang H, Zhang B, et al. B3GNT3 overexpression is associated with unfavourable survival in non-small cell lung cancer. J Clin Pathol 2018;71:642-7. [Crossref] [PubMed]
13. Kong K, Zhao Y, Xia L, et al. B3GNT3: A prognostic biomarker associated with immune cell infiltration in pancreatic adenocarcinoma. Oncol Lett 2021;21:159. [Crossref] [PubMed]
14. Zhou Z, Wu B, Chen J, et al. ETV4 facilitates proliferation, migration, and invasion of liver cancer by mediating TGF-β signal transduction through activation of B3GNT3. Genes Genomics 2023;45:1433-43. [Crossref] [PubMed]
15. Zhuang H, Zhou Z, Zhang Z, et al. B3GNT3 overexpression promotes tumor progression and inhibits infiltration of CD8(+) T cells in pancreatic cancer. Aging (Albany NY) 2020;13:2310-29. [Crossref] [PubMed]
16. Li CW, Lim SO, Chung EM, et al. Eradication of Triple-Negative Breast Cancer Cells by Targeting Glycosylated PD-L1. Cancer Cell 2018;33:187-201.e10. [Crossref] [PubMed]
17. Tang Z, Kang B, Li C, et al. GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis. Nucleic Acids Res 2019;47:W556-60. [Crossref] [PubMed]
18. Li T, Fu J, Zeng Z, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res 2020;48:W509-14. [Crossref] [PubMed]
19. Chandrashekar DS, Karthikeyan SK, Korla PK, et al. UALCAN: An update to the integrated cancer data analysis platform. Neoplasia 2022;25:18-27. [Crossref] [PubMed]
20. Thul PJ, Lindskog C. The human protein atlas: A spatial map of the human proteome. Protein Sci 2018;27:233-44. [Crossref] [PubMed]
21. Szklarczyk D, Gable AL, Nastou KC, et al. The STRING database in 2021: customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets. Nucleic Acids Res 2021;49:D605-12. [Crossref] [PubMed]
22. Cerami E, Gao J, Dogrusoz U, et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov 2012;2:401-4. [Crossref] [PubMed]
23. Huang da W. Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009;4:44-57. [Crossref] [PubMed]
24. Huang da W. Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res 2009;37:1-13. [Crossref] [PubMed]
25. Verbeke H, Van Hecke D, Bauraing C, et al. Belgian Recommendations for Analytical Verification and Validation of Immunohistochemical Tests in Laboratories of Anatomic Pathology. Appl Immunohistochem Mol Morphol 2024;32:1-16. [Crossref] [PubMed]
26. Lott RL, Riccelli PV, Sheppard EA, et al. Immunohistochemical Validation of Rare Tissues and Antigens With Low Frequency of Occurrence: Recommendations From The Anatomic Pathology Patient Interest Association (APPIA). Appl Immunohistochem Mol Morphol 2021;29:327-34. [Crossref] [PubMed]
27. Goldsmith JD, Troxell ML, Roy-Chowdhuri S, et al. Principles of Analytic Validation of Immunohistochemical Assays: Guideline Update. Arch Pathol Lab Med 2024;148:e111-53. [Crossref] [PubMed]
28. Zhang H, Yue X, Chen Z, et al. Define cancer-associated fibroblasts (CAFs) in the tumor microenvironment: new opportunities in cancer immunotherapy and advances in clinical trials. Mol Cancer 2023;22:159. [Crossref] [PubMed]
29. Jia H, Chen X, Zhang L, et al. Cancer associated fibroblasts in cancer development and therapy. J Hematol Oncol 2025;18:36. [Crossref] [PubMed]
30. Zhang H, Lu X, Lu B, et al. Measuring the composition of the tumor microenvironment with transcriptome analysis: past, present and future. Future Oncol 2024;20:1207-20. [Crossref] [PubMed]
31. Xu Y, Li W, Lin S, et al. Fibroblast diversity and plasticity in the tumor microenvironment: roles in immunity and relevant therapies. Cell Commun Signal 2023;21:234. [Crossref] [PubMed]
32. Raaijmakers KTPM, Adema GJ, Bussink J, et al. Cancer-associated fibroblasts, tumor and radiotherapy: interactions in the tumor micro-environment. J Exp Clin Cancer Res 2024;43:323. [Crossref] [PubMed]
33. Zhang W, Hou T, Niu C, et al. B3GNT3 Expression Is a Novel Marker Correlated with Pelvic Lymph Node Metastasis and Poor Clinical Outcome in Early-Stage Cervical Cancer. PLoS One 2015;10:e0144360. [Crossref] [PubMed]
34. Zhou H, Zhao J, Yang X, et al. Study on the Expression of β-1,3-N-acetylglucosaminyltransferase 3 in Gastric Cancer and the Mechanism Promoting Gastric Cancer Progression Based on the Extraction Method of Nanomagnetic Beads. J Biomed Nanotechnol 2022;18:677-92. [Crossref] [PubMed]
35. Wu Y, Luo J, Li H, et al. B3GNT3 as a prognostic biomarker and correlation with immune cell infiltration in lung adenocarcinoma. Ann Transl Med 2022;10:295. [Crossref] [PubMed]
36. Liang JX, Chen Q, Gao W, et al. A novel glycosylation-related gene signature predicts survival in patients with lung adenocarcinoma. BMC Bioinformatics 2022;23:562. [Crossref] [PubMed]
37. Sahai E, Astsaturov I, Cukierman E, et al. A framework for advancing our understanding of cancer-associated fibroblasts. Nat Rev Cancer 2020;20:174-86. [Crossref] [PubMed]
38. Liu Y, Sinjab A, Min J, et al. Conserved spatial subtypes and cellular neighborhoods of cancer-associated fibroblasts revealed by single-cell spatial multi-omics. Cancer Cell 2025;43:905-924.e6. [Crossref] [PubMed]
39. Zhang F, Ma Y, Li D, et al. Cancer associated fibroblasts and metabolic reprogramming: unraveling the intricate crosstalk in tumor evolution. J Hematol Oncol 2024;17:80. [Crossref] [PubMed]
40. Thuya WL, Cao Y, Ho PC, et al. Insights into IL-6/JAK/STAT3 signaling in the tumor microenvironment: Implications for cancer therapy. Cytokine Growth Factor Rev 2025;S1359-6101(25)00003-6.
41. Nurmik M, Ullmann P, Rodriguez F, et al. In search of definitions: Cancer-associated fibroblasts and their markers. Int J Cancer 2020;146:895-905. [Crossref] [PubMed]
42. Croft W, Pearce H, Margielewska-Davies S, et al. Spatial determination and prognostic impact of the fibroblast transcriptome in pancreatic ductal adenocarcinoma. Elife 2023;12:e86125. [Crossref] [PubMed]
43. Miao Z, Cao Q, Liao R, et al. Elevated transcription and glycosylation of B3GNT5 promotes breast cancer aggressiveness. J Exp Clin Cancer Res 2022;41:169. [Crossref] [PubMed]
44. Yang Y, Sun H, Yu H, et al. Tumor-associated-fibrosis and active collagen-CD44 axis characterize a poor-prognosis subtype of gastric cancer and contribute to tumor immunosuppression. J Transl Med 2025;23:123. [Crossref] [PubMed]
45. Hu S, Xiao Q, Gao R, et al. Identification of BGN positive fibroblasts as a driving factor for colorectal cancer and development of its related prognostic model combined with machine learning. BMC Cancer 2024;24:516. [Crossref] [PubMed]
46. Kumari L, Yadav R, Kumar Y, et al. Role of tight junction proteins in shaping the immune milieu of malignancies. Expert Rev Clin Immunol 2024;20:1305-21. [Crossref] [PubMed]
47. Vonniessen B, Tabariès S, Siegel PM. Antibody-mediated targeting of Claudins in cancer. Front Oncol 2024;14:1320766. [Crossref] [PubMed]
48. Nehme Z, Roehlen N, Dhawan P, et al. Tight Junction Protein Signaling and Cancer Biology. Cells 2023;12:243. [Crossref] [PubMed]
49. Konno T, Kohno T, Kikuchi S, et al. The interplay between the epithelial permeability barrier, cell migration and mitochondrial metabolism of growth factors and their inhibitors in a human endometrial carcinoma cell line. Tissue Barriers 2024;12:2304443. [Crossref] [PubMed]
50. Magalhães A, Duarte HO, Reis CA. The role of O-glycosylation in human disease. Mol Aspects Med 2021;79:100964. [Crossref] [PubMed]
51. Li JP, Liu YJ, Li Y, et al. Spatiotemporal heterogeneity of LMOD1 expression summarizes two modes of cell communication in colorectal cancer. J Transl Med 2024;22:549. [Crossref] [PubMed]