Systematic pan-cancer analysis reveals the prognostic and immunological roles of ectonucleoside triphosphate diphosphohydrolase 6

Abstract

BACKGROUND

Ectonucleoside triphosphate diphosphohydrolase 6 (ENTPD6), a member of the ENTPD family, has been implicated in certain cancers, yet a comprehensive analysis across multiple cancer types remains lacking.

AIM

To systematically evaluate ENTPD6’s expression, prognostic significance, and functions across multiple cancer types.

METHODS

In this study, we performed a pan-cancer analysis to investigate the correlation between ENTPD6 expression and various factors, including prognosis, genetic alterations, epigenetic modification, immune infiltration, immunotherapy responses, functional enrichment, and drug sensitivity. A tissue microarray of gastrointestinal tumors was used to validate differential ENTPD6 protein expression.

RESULTS

Pan-cancer analysis revealed that ENTPD6 expression was significantly elevated in many cancers. Immunohistochemistry staining analysis revealed that ENTPD6 expression was significantly higher in esophageal carcinoma, stomach adenocarcinoma, colon adenocarcinoma, rectal adenocarcinoma, and pancreatic adenocarcinoma compared to normal tissues. Furthermore, ENTPD6 expression was strongly associated with immune-infiltrating cells, particularly clusters of differentiation 8+ T cells and natural killer cells, and correlated with immune-related genomic features including tumor mutational burden and microsatellite instability. Pathway analysis indicated that ENTPD6 expression was primarily linked to purine and pyrimidine metabolism pathways. Drug sensitivity analysis revealed that high ENTPD6 expression was sensitive to RDEA119, selumetinib, and PD-0325901.

CONCLUSION

This pan-cancer study elucidates the pivotal role of ENTPD6 in tumor progression and establishes its potential as a therapeutic target for immunotherapeutic approaches in specific malignancies.

Key Words: Ectonucleoside triphosphate diphosphohydrolase 6; Pan-cancer analysis; Immunotherapy; Biomarker; Molecular docking

Core Tip: Ectonucleoside triphosphate diphosphohydrolase 6 (ENTPD6), a member of the ENTPD family, was highly expressed in kidney renal clear cell carcinoma, liver hepatocellular carcinoma, kidney renal papillary cell carcinoma, endocervical adenocarcinoma, and other malignancies, indicating it could be a potential biomarker. In this study, we performed a comprehensive analysis examining the correlation between ENTPD6 expression and various factors, including prognosis, genetic alterations, epigenetic modification, immune infiltration, immunotherapy responses, functional enrichment, and drug sensitivity. Our findings yield novel insights into the role of ENTPD6 in cancer biology and bolster global research on its potential as a therapeutic target.

INTRODUCTION

Cancer is a leading cause of death globally, accounting for 16.8% of all deaths and 22.8% of deaths from noncommunicable diseases worldwide[1]. Between 2005 and 2020, a significant rise in cancer-related mortality was observed in China[2]. Cancer treatment typically involves a combination of surgery, chemotherapy, radiotherapy, immunotherapy, and targeted therapies. However, the discovery of new genes that serve as biomarkers can provide opportunities for developing targeted therapies to enhance the efficacy of immune checkpoint inhibitors[3]. Identifying multiple biomarkers based on patients’ clinical-pathological characteristics, drug mechanisms, and safety profiles is essential for advancing personalized treatment strategies. These biomarkers can accelerate the development of novel anti-tumor therapies and enhance personalized treatment strategies.

Ectonucleoside triphosphate diphosphohydrolase 6 (ENTPD6), also known as interleukin six signal transducer-2 or cluster of differentiation (CD) 39 antigen-like 2 or nucleoside triphosphate diphosphohydrolase 6, belongs to the ENTPD family[4]. Previous studies characterize soluble ENTPD6 as a diphosphatase, with predominant cardiac expression where it modulates platelet activation and recruitment[5,6]. Its primary function might involve the conversion of nucleotides into nucleosides, thereby regulating nucleotide levels in cellular organelles[7]. Notably, extracellular nucleotide levels are closely linked to various tissue functions, including development, inflammation, secretion, blood flow, and immune reaction[8]. Interestingly, ENTPD6 has been identified as a novel therapeutic biomarker in testicular cancer[9], where it plays a role in the malignant progression. Concurrently, it is characterized as a metabolism-related gene in squamous lung cancer[10]. Preliminary evidence further suggested that the mutated ENTPD6 could serve as a neoantigen in hepatocellular carcinoma, suggesting its potential as an immunotherapeutic target[11].

Despite the critical role of ENTPD6 in tumors, a comprehensive bioinformatics analysis examining its expression, prognostic relevance, and functions across multiple cancer types has not yet been conducted. In this study, we perform a thorough analysis of ENTPD6 expression and its correlation with the prognostic value, functional pathways, DNA methylation, genomic instability, immune cell infiltration, immunotherapeutic response, and drug sensitivity in pan-cancer. Furthermore, the oncogenic role of ENTPD6 in six gastrointestinal tumors was validated using immunohistochemistry (IHC). This study systematically evaluates the biomarker potential of ENTPD6 for prognosis, diagnosis, and targeted therapy in specific malignancies. The findings will provide valuable insights for researchers worldwide, advancing the understanding of ENTPD6’s role in cancer mechanisms.

MATERIALS AND METHODS

Pan-cancer data collection

Data on ENTPD6 expression in normal and tumor tissues from The Cancer Genome Atlas (TGCA) and in normal tissues from the Genotype-Tissue Expression (GTEx) databases were retrieved from the University of California, Santa Cruz Xena platform (https://xenabrowser.net/datapages/). The pan-cancer clinical data, DNA methylation profiles, and sample annotations were also obtained from this platform.

Differential expression and subcellular localization profiling

Data processing leveraged the stringr, tidyr, stringi, and AnnotationDbi packages, while result visualization employed ggplot2 and RColorBrewer. Between-group comparisons were performed using Student’s t-test when data were normally distributed and the Wilcoxon rank-sum test otherwise. For three or more groups, one-way ANOVA was applied under normality assumptions, whereas the Kruskal-Wallis test was used for non-normally distributed data. The Human Protein Atlas dataset (https://www.proteinatlas.org/) was utilized for displaying the subcellular localization of ENTPD6[12].

IHC staining

A tissue microarray of gastrointestinal tumors (HDgS-C140PT-01) was purchased from Shanghai Outdo Biotech Company (Shanghai, China). The tissue microarray included six kinds of gastrointestinal tumors and adjacent normal tissues, with 10 samples for each type, including esophageal carcinoma (ESCA), stomach adenocarcinoma (STAD), colon adenocarcinoma (COAD), rectal adenocarcinoma (READ), hepatocellular carcinoma, and pancreatic adenocarcinoma (PAAD). It also included six types of normal tissues, each with 2-5 samples, including the esophagus, stomach, colon, rectum, liver, and pancreas. A polyclonal antibody (dilution 1:100) was purchased from Proteintech Group (IL, United States). Two experienced pathologists assessed ENTPD6 expression levels by evaluating the cytoplasmic staining intensity and the percentage of tumor cells. The pathologists scored the staining intensity as 0 (negative), 1 (weak), 2 (medium), and 3 (strong). The percentage of tumor cells with staining was categorized as 0 (negative), 1 (< 25%), 2 (25%-50%), 3 (50%-75%), and 4 (> 75%). The positive staining score, calculated by multiplying staining intensity and area, was classified as follows: Scores of 1-3 indicated weak expression, 4-6 indicated moderate expression, and 7-12 indicated strong expression.

Prognostic analysis

To assess the prognosis of ENTPD6 across various cancer types, we constructed Cox proportional hazards survival regression models to analyze the relationship between ENTPD6 expression and multiple survival outcomes, including overall survival (OS), progression-free interval (PFI), disease-free interval, and disease-specific survival (DSS). These analyses were performed using the survival and survminer packages. Meanwhile, patients were divided into high and low groups based on median ENTPD6 expression. Kaplan-Meier curves were then applied to compare survival outcomes.

Genetic alterations and genomic instability analyses

The cBioPortal platform (https://www.cbioportal.org/) was utilized to analyze ENTPD6 genetic alterations, including genomic mutations, structural variants, amplifications, deep deletions, and multiple alterations[13]. The ‘Mutations’ module was employed to detect and visualize ENTPD6’s mutation sites and types. Additionally, the relationship between ENTPD6 expression and copy-number alterations was investigated.

Epigenetic modification analysis

Building on the Shiny Methylation Analysis Resource Tool platform (http://www.bioinfo-zs.com/smartapp/)[14], we systematically examined DNA methylation patterns of ENTPD6 across pan-cancer datasets. We further assessed the differential methylation status of ENTPD6 probes in the upstream and gene body regions. Additionally, we assessed associations between ENTPD6 expression and RNA epigenetic modifications, including N1-methyladenosine, 5-methylcytosine, N7-methylguanosine, and N6-methyladenosine, were also performed. Genes associated with the modifications of N6-methyladenosine, 5-methylcytosine, N1-methyladenosine, and N7-methylguanosine were collected from previous articles[15,16].

Correlation between ENTPD6 expression and tumor microenvironment

Immune cell infiltration profiles were obtained from the Tumor Immune Estimation Resource 2.0 platform (http://timer.cistrome.org/)[17]. Immune scores were calculated using the estimate package. Spearman correlation was employed to evaluate the correlation between ENTPD6 expression and ImmuneScore, Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data score, as well as StromalScore. Cell-type identification by estimating relative subsets of RNA transcripts was employed to quantify the relationship between ENTPD6 expression and immune-cell infiltration. To explore relationships between ENTPD6 expression and lymphocyte populations, major histocompatibility complex genes, immunoinhibitory genes, immunostimulatory genes, chemokines, and chemokine receptors, we interrogated the tumor-immune system interaction database (TISIDB) (http://cis.hku.hk/TISIDB/index.php)[18].

Correlation between ENTPD6 expression and immunotherapy

The association between ENTPD6 expression and mismatch repair (MMR) genes (mutL homolog 1, PMS2, mutS homolog 6, mutS homolog 6)[19], as well as 51 immunological checkpoint genes (ICGs), has been studied[20]. TGCAplot package was used to explore the relationship between ENTPD6 expression and tumor mutational burden (TMB)/microsatellite instability (MSI)[21].

Related-gene network construction and functional enrichment analyses

The protein-protein interaction network of ENTPD6 was established via the STRING database (https://cn.string-db.org/)[22]. Then, Cytoscape software was applied to visualize the results. The ENTPD6-interacting genes were collected from the Gene Expression Profiling Interaction Analysis 2 tool (http://gepia2.cancer-pku.cn/#index) and GeneMANIA platform (http://www.genemania.org)[23,24]. Subsequent analyses, including the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis and Gene Ontology (GO) enrichment analysis[25,26], were performed using ‘clusterProfiler’ and ‘org.Hs.eg.db’ packages. The GO analysis encompassed three main categories: Biological processes, cellular components, and molecular functions.

Establishment of a nomogram

Samples were integrated from TGCA and GTEx databases. For model development, 70% were randomly partitioned into the training cohort, reserving the remaining 30% as an independent testing cohort. Receiver operating characteristic (ROC) curves were generated to comprehensively assess ENTPD6’s diagnostic accuracy across multiple cancer types. A dynamic nomogram was constructed using the rms package. Decision-curve analysis (DCA) was performed with the ggDCA package. ROC curves were generated using the ‘pROC’ package. The bootstrap resampling method was used to validate the model, with a sampling frequency of 1000. Theoretically, if the area under the curve (AUC) value is closer to 1, the area below the curve is more extensive, indicating a higher prediction model accuracy and vice versa[27]. A standard curve is a straight line with a slope of 1 through a coordinate axis dot, indicating the nomogram is perfectly calibrated[28].

Drug sensitivity analysis

The Gene Set Cancer Analysis (GSCA) platform (https://guolab.wchscu.cn/GSCA/#/drug) was employed to systematically evaluate ENTPD6’s sensitivity to a panel of drugs[29]. The three-dimensional structure of ENTPD6 protein was obtained from the AlphaFold Protein Structure Database (https://www.alphafold.ebi.ac.uk/entry/AF-O75354)[30,31]. The three-dimensional structures of selumetinib, SB590885, RDEA119, trametinib, dabrafenib, and PD-0325901 were downloaded from the PubChem platform (https://pubchem.ncbi.nlm.nih.gov/)[32]. The molecular docking experiments were performed using AutoDockTool (version 1.5.7), with visualization conducted in PyMOL (version 3.0.3).

Statistical analysis

All statistical analyses were performed using R software (version 4.3.3). The correlations between ENTPD6 and other genes were evaluated using Spearman’s correlation method. All statistical comparisons were rigorously controlled for multiple-testing bias using the false discovery rate. P < 0.05 was considered statistically significant.

RESULTS

ENTPD6 expression and its subcellular localization

Our study began with an analysis of ENTPD6 expression across various cancers, revealing significantly elevated expression in bladder urothelial carcinoma (BLCA), breast invasive carcinoma (BRCA), cervical squamous cell carcinoma and cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), COAD, ESCA, head and neck squamous cell carcinoma (HNSC), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), PAAD, prostate adenocarcinoma (PRAD), READ, STAD and uterine corpus endometrial carcinoma (UCEC), whereas it was reduced in thyroid carcinoma (THCA) (Figure 1A). To compensate for the lack of normal samples, we integrated the data from the GTEx database. These results showed that ENTPD6 expression was significantly higher in lymphoid neoplasm diffuse large B-cell lymphoma (DLBC), BRCA, PAAD, STAD, skin cutaneous melanoma (SKCM), READ, COAD, LIHC, ESCA, BLCA, acute myeloid leukemia (LAML), CHOL, HNSC, thymoma (THYM) (Figure 1B). In contrast, it was lower in ovarian serous cystadenocarcinoma (OV), brain lower-grade glioma (LGG), LUSC, THCA, adrenocortical carcinoma (ACC), uterine carcinosarcoma, testicular germ cell tumors (TGCT), glioblastoma multiforme, and kidney renal clear cell carcinoma (KIRC) (Figure 1B). ENTPD6 was predominantly localized to the Golgi apparatus (Figure 1C). Additionally, ENTPD6 expression levels were found to be associated with tumor stages in HNSC and uveal melanoma (UVM) (Figure 1D and E). In BLCA, HNSC, KIRC, LIHC, and STAD, ENTPD6 expression was significantly correlated with different tumor grades (Figure 1F-J). These findings suggest that ENTPD6 may be implicated in tumor aggressiveness and metastasis, providing valuable insights for further exploring its functional mechanisms across various cancers.

Figure 1 Ectonucleoside triphosphate diphosphohydrolase 6 expressions in pan-cancer. A: Ectonucleoside triphosphate diphosphohydrolase 6 (ENTPD6) expression levels both in the tumor and normal tissues from The Cancer Genome Atlas dataset; B: ENTPD6 expression levels in the paired tumor/normal samples of the datasets from The Cancer Genome Atlas and Genotype-Tissue Expression databases; C: Subcellular localization of ENTPD6 in SK-MEL-30 cell line from The Human Protein Atlas database, scale bar: 20 μm. Red, green, and blue represent microtubules, the target protein, and the nucleus, respectively; D and E: The association between ENTPD6 expression and pathological stages in head and neck squamous cell carcinoma and uveal melanoma; F-J: ENTPD6 expression correlated with tumor grades in bladder urothelial carcinoma, head and neck squamous cell carcinoma, kidney renal clear cell carcinoma, liver hepatocellular carcinoma, and stomach adenocarcinoma. ^aP < 0.05; ^bP < 0.01; ^cP < 0.001. ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6; TPM: Transcripts per million; TCGA: The Cancer Genome Atlas; GTEx: Genotype-Tissue Expression; HNSC: Head and neck squamous cell carcinoma; UVM: Uveal melanoma; BLCA: Bladder urothelial carcinoma; KIRC: Kidney renal clear cell carcinoma; LIHC: Liver hepatocellular carcinoma; STAD: Stomach adenocarcinoma.

IHC staining validation of ENTPD6 expression

Our bioinformatics analysis showed that ENTPD6 was overexpressed in most gastrointestinal tumors compared to normal tissues. Therefore, we conducted in vitro experimental validation on six common gastrointestinal tumors. IHC analysis showed that ENTPD6 was overexpressed in PAAD, STAD, ESCA, READ, and COAD compared to adjacent normal tissues (Figure 2A-F). ROC analysis demonstrated that the AUC value was 0.839, indicating high diagnostic accuracy for these cancers (Figure 2G).

Figure 2 Ectonucleoside triphosphate diphosphohydrolase 6 expressions in gastrointestinal tumors. A-E: Immunohistochemistry staining results of ectonucleoside triphosphate diphosphohydrolase 6 (ENTPD6) expression in gastrointestinal tumors and adjacent normal tissues, including esophageal carcinoma, colon adenocarcinoma, stomach adenocarcinoma, pancreatic adenocarcinoma, rectal adenocarcinoma, scale bar: 200 μm; F: Differential expression plot of ENTPD6; G: The receiver operating characteristic curve and area under the curve value of ENTPD6 in six tumors. ^aP < 0.05; ^cP < 0.001. ESCA: Esophageal carcinoma; COAD: Colon adenocarcinoma; STAD: Stomach adenocarcinoma; PAAD: Pancreatic adenocarcinoma; READ: Rectal adenocarcinoma; ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6; AUC: Area under the curve.

Relationship between ENTPD6 expression and prognosis

Univariate Cox regression analysis demonstrated that elevated ENTPD6 expression was significantly associated with poorer OS and DSS in KIRC, but acted as a protective factor in UVM (Figure 3A and B). Although false discovery rate-adjusted analyses showed non-significant associations between ENTPD6 expression and disease-free interval/PFI across 33 cancers, nominally significant raw P-values (P < 0.05) in specific malignancies warrant further investigation (Figure 3C and D). The Kaplan-Meier curves revealed that high ENTPD6 expression was associated with worse OS in LIHC, KIRC, and kidney renal papillary cell carcinoma (KIRP), whereas it correlated with improved survival outcomes in THYM and READ (Figure 3E-I). For PFI, high ENTPD6 expression predicted poor outcomes in CESC, DLBC, HNCS, and LUSC, but showed a positive result in STAD (Figure 3J-N). For DSS, high ENTPD6 expression was associated with worse outcomes in LUSC, KIRC, and KIRP (Figure 3O-Q).

Figure 3 Prognosis and diagnostic value analysis of ectonucleoside triphosphate diphosphohydrolase 6 in various cancers. A-D: The univariate Cox regression analysis of overall survival, disease-specific survival, disease-free interval, and progression-free interval; E-I: The results of Kaplan-Meier curves analysis of overall survival in liver hepatocellular carcinoma, kidney renal clear cell carcinoma, kidney renal papillary cell carcinoma, thymoma, and rectal adenocarcinoma; J-N: The results of the Kaplan-Meier curves analysis of progression-free interval in cervical squamous cell carcinoma and endocervical adenocarcinoma, diffuse large B-cell lymphoma, head and neck squamous cell carcinoma, lung squamous cell carcinoma, and stomach adenocarcinoma; O-Q: The results of the Kaplan-Meier curves analysis of disease-specific survival in lung squamous cell carcinoma, kidney renal papillary cell carcinoma, and kidney renal clear cell carcinoma. ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6; OS: Overall survival; DSS: Disease-specific survival; DFI: Disease-free interval; PFI: Progression-free interval; LIHC: Liver hepatocellular carcinoma; KIRC: Kidney renal clear cell carcinoma; KIRP: Kidney renal papillary cell carcinoma; THYM: Thymoma; READ: Rectal adenocarcinoma; CESC: Cervical squamous cell carcinoma and endocervical adenocarcinoma; DLBC: Diffuse large B-cell lymphoma; HNSC: Head and neck squamous cell carcinoma; LUSC: Lung squamous cell carcinoma; STAD: Stomach adenocarcinoma.

Genetic alterations and genomic instability

Genetic alterations and chromosomal instability are contributors to gene mutation and potential cancer progression. ENTPD6 exhibited marked genetic diversity and frequent mutations, with mutation rates of 4.35%, 2.49%, 1.59%, and 1.23% in UCEC, SKCM, STAD, and LUSC, respectively (Figure 4A). Furthermore, KIRP and LAML exhibited only a single mutation type, with frequencies of 0.71% and 0.5%, respectively (Figure 4A). Uterine carcinosarcoma had the highest incidence of “amplification” type of copy number amplifications, with a frequency of 3.51%. ENTPD6 expression exhibited deep deletions in colorectal adenocarcinoma, STAD, LUAD, LUSC, OV, LGG, BRCA, and PRAD (Figure 4A). In various cancers, 92 variants of uncertain significance were identified in ENTPD6 expression, primarily missense mutations (Figure 4B). We characterized the relationship between copy number alterations (CNAs) and ENTPD6 expression. ENTPD6 expression levels demonstrated significant variation across CNA categories, peaking in amplified tumors (Figure 4C), and exhibited a strong positive correlation with CNAs (Figure 4D). These results suggest an important association between ENTPD6 expression and genomic instability.

Figure 4 The correlation between ectonucleoside triphosphate diphosphohydrolase 6 expression and genetic alterations, as well as genomic instability. A: Pan-cancer analyses of genomic changes in ectonucleoside triphosphate diphosphohydrolase 6 (ENTPD6) expression were conducted, including analyses of mutations, amplifications, and deep deletions; B: Genetic alterations of ENTPD6, including missense, frameshift deletion, and splice site mutations; C: ENTPD6 expression in different copy number alteration categories; D: The correlation between copy number alteration and ENTPD6 expression. CAN: Copy number alteration; VUS: Variant of uncertain significance; ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6; RSEM: RNA-Seq by Expectation Maximization; GSTIC: Genomic Identification of Significant Targets in Cancer.

Epigenetic modification

In the mammalian genome, DNA methylation is an epigenetic mechanism that adds a methyl group to the C5 position of the cytosine, facilitated by DNA methyltransferase[33]. Accumulating evidence establishes dysregulated DNA methylation as a critical driver in cancer pathogenesis, orchestrating tumor initiation, persistence, and metastatic progression across diverse malignancies[34,35]. Based on the Shiny Methylation Analysis Resource Tool platform, we analyzed ENTPD6 DNA methylation levels in both tumor and normal tissues. The results indicated that ENTPD6 DNA methylation levels in tumor tissues were significantly lower than those in normal tissues, including BLCA, BRCA, CESC, CHOL, COAD, HNSC, KIRP, LIHC, LUAD, LUSC, PAAD, READ, THCA, and UCEC (Figure 5A). In contrast, the opposite was observed in KIRC (Figure 5A). Hypomethylation of some ENTPD6 probes was observed in BLCA, CHOL, LIHC, KIRP, LUAD, THCA, and UCEC, suggesting potential oncogene activation (Figure 5B). The opposite trend was observed in HNSC and KIRC (Figure 5B). In addition, we explored the correlation between ENTPD6 expression and RNA modification genes. These results showed a positive correlation between ENTPD6 expression and epigenetic modifications in pan-cancer, such as BLCA, DLBC, HNSC, kidney chromophobe, LAML, LIHC, LUAD, OV, and pheochromocytoma and paraganglioma (PCPG) (Figure 5C-F). These findings demonstrate that ENTPD6 may regulate tumorigenesis and development through DNA methylation or RNA modification.

Figure 5 Epigenetic modification analyses. A: The DNA methylation levels of ectonucleoside triphosphate diphosphohydrolase 6 (ENTPD6) were obtained from the Shiny Methylation Analysis Resource Tool database; B: Differential methylation of DNA methylation probes in upstream (> 2 kb) and gene body regions of ENTPD6; C-F: Correlation analysis between ENTPD6 expression and the genes related to m6A, m1A, m5C, and m7G modifications. ^aP < 0.05; ^bP < 0.01; ^cP < 0.001. ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6.

The correlation between ENTPD6 expression and tumor microenvironment

The tumor microenvironment (TME) is a complex ecosystem that includes diverse immune cells such as CD8+ T cells, regulatory T cells, B cells, macrophages, monocytes, and natural killer (NK) cells[36]. We evaluated ENTPD6’s role in TME remodeling using stromal, immune, and Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data scores. Pan-cancer analysis revealed significantly negative correlations between ENTPD6 expression and three TME metrics across multiple malignancies, while conversely demonstrating a strong positive correlation in DLBC (Figure 6A). Analysis of immune cell infiltration revealed distinct correlation patterns between ENTPD6 expression and immune subsets across multiple cancers. In BRCA, ENTPD6 expression positively correlated with M2 macrophages, while negatively correlated with M1 macrophages (Figure 6B). A significant negative correlation with M1 macrophages and M2 macrophages was observed in CESC. Notably, ENTPD6 expression showed a consistent negative correlation with CD8+ T cell infiltration in PAAD, LUSC, HNSC, and STAD (Figure 6B). In T-regulatory cells, ENTPD6 expression exhibited a significant positive correlation in KIRC, ESCA, and LIHC, whereas it was markedly negatively correlated in HNSC and LGG (Figure 6B). Regarding NK cell compartments, ENTPD6 expression positively correlated with resting NK cells in CESC, STAD, and OV, while demonstrating a concurrent negative correlation with activated NK cells in these same cancer types (Figure 6B). Finally, we applied the tumor-immune system interaction database to investigate the relationship between ENTPD6 and immune-related genes. These findings revealed significant correlations between ENTPD6 expression levels and lymphocyte populations in ACC, COAD, LGG, PAAD, PRAD, and UVM (Figure 6C). ENTPD6 expression levels were significantly associated with immunostimulatory factors (ectonucleoside triphosphate diphosphohydrolase 1, poliovirus receptor, tumor necrosis factor receptor superfamily member 14, and tumor necrosis factor receptor superfamily member 25), and immunoinhibitory genes (programmed death-ligand 1, programmed cell death 1 ligand 2, kinase insert domain receptor, poliovirus receptor-related 2, type I transforming growth factor-beta receptor gene, and galectin 9) (Figure 6D and E). Moreover, ENTPD6 co-expression with major histocompatibility complex molecules obtained a positive result in some cancers, including ACC, mesothelioma, OV, TGCT, and THCA (Figure 6F). ENTPD6 expression levels were negatively correlated with some chemokines [C-C motif chemokine ligand 2 (CCL 2), CCL22, CCL14, C-X-C motif chemokine ligand 12, and C-X-C motif chemokine ligand 9] and chemokine receptors [C-C motif chemokine receptor 1 (CCR1), CCR2, CCR4, CCR5, C-X3-C motif chemokine receptor 1, and C-X3-C motif chemokine receptor 4] in various cancers (Figure 6G and H).

Figure 6 The correlation between ectonucleoside triphosphate diphosphohydrolase 6 expression and tumor microenvironment. A: The relationship between ectonucleoside triphosphate diphosphohydrolase 6 (ENTPD6) expression levels and Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data scores, immune scores, and stromal scores; B: The correlation between ENTPD6 expression and immune cells; C-H: The relationships between ENTPD6 expression and immune-related genes in the tumor-immune system interaction database, including lymphocyte, immunostimulator, immunoinhibitor, major histocompatibility complex molecule, chemokine, and receptor. ^aP < 0.05; ^bP < 0.01; ^cP < 0.001. ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6; MHC: Major histocompatibility complex molecule.

The association between ENTPD6 expression and immunotherapy

To deeply study the role of ENTPD6 expression in immunotherapy response, we analyzed its relationship with immune checkpoint gene expression across various cancers. Results demonstrated that ENTPD6 expression exhibited positive correlations with the majority of immune checkpoint genes in DLBC, LIHC, OV, PCPG, and LAML, whereas it showed negative correlations with most immune checkpoint genes in PAAD, LGG, COAD, CESC, and SKCM (Figure 7A). Emerging evidence positions MMR deficiency, TMB, and MSI as pivotal immunotherapy biomarkers[37]. Therefore, we analyzed the links between ENTPD6 expression and TMB, MSI, and MMR status in human cancers to see whether ENTPD6 could be an immunotherapy response biomarker. An apparent positive correlation was observed between ENTPD6 expression and the expression levels of MMR genes, including THYM, THCA, PRAD, LUAD, LUSC, and COAD (Figure 7B). ENTPD6 expression positively correlated with MSI in LUSC, LUAD, KIRC, CESC, BLCA, UCEC, STAD, and SKCM, but inversely correlated with MSI in READ and COAD (Figure 7C). ENTPD6 expression was positively associated with TMB in patients with SKCM, LUAD, and KIRC, while showing negative correlations with TMB in COAD and KIRP (Figure 7D). Collectively, our pan-cancer analysis establishes ENTPD6 expression as a composite biomarker of tumor immunogenicity, significantly correlated with TMB, MSI, and MMR statuses, positioning it as a promising candidate for stratifying immunotherapy responders in specific cancers.

Figure 7 Correlation between ectonucleoside triphosphate diphosphohydrolase 6 expression and immunotherapy. A: The association between ectonucleoside triphosphate diphosphohydrolase 6 expression and 51 immune checkpoint genes; B-D: Mismatch repair genes, microsatellite instability, as well as tumor mutational burden. ^aP < 0.05; ^bP < 0.01; ^cP < 0.001. ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6; MMR: Mismatch repair; MSI: Microsatellite instability; TMB: Tumor mutational burden.

Related-gene network and functional enrichment analyses

We employed a multi-platform bioinformatics strategy to delineate the ENTPD6 interaction landscape, beginning with functional association network construction using GeneMANIA (Figure 8A). Subsequent Gene Expression Profiling Interactive Analysis tool identified 400 genes significantly associated with ENTPD6 expression. The top 76 ENTPD6-interacting proteins were extracted from the STRING database and visualized as a protein-protein interaction network in Cytoscape (Figure 8B). Intersectional analysis across three datasets identified two common interactors: Glycogen phosphorylase B and ENTPD5 (Figure 8C). Subsequent pan-cancer analysis revealed significant positive co-expression between ENTPD6 and both glycogen phosphorylase B and ENTPD5 expression (Figure 8D). To explore the functional enrichment of ENTPD6, we performed GO function and KEGG pathway enrichment analyses. KEGG pathway analysis showed that ENTPD6-related genes were associated with purine metabolism, nucleotide metabolism, pyrimidine metabolism, drug metabolism-other enzymes, nicotinate and nicotinamide metabolism, biosynthesis of cofactors, biosynthesis of nucleotide sugars, amino sugar and nucleotide sugar metabolism, renin secretion, and morphine addiction (Figure 8E). GO enrichment analysis revealed that the biological processes of ENTPD6-interacting genes were linked to the nucleotide biosynthetic process, nucleotide catabolic process, carbohydrate derivative catabolic process, organophosphate catabolic process, and glycoprotein metabolic process (Figure 8F). The cellular components of ENTPD6-related genes were enrichment in transport vesicle, Golgi apparatus subcompartment, trans-Golgi network membrane, trans-Golgi network, etc. (Figure 8F). The results of molecular functions analysis indicated that ENTPD6-related genes were involved in phosphoric ester hydrolase activity, guanyl nucleotide binding, cyclic-nucleotide phosphodiesterase activity, magnesium ion binding, calmodulin binding, etc. (Figure 8F).

Figure 8 Functional enrichment analysis. A: The ectonucleoside triphosphate diphosphohydrolase 6 (ENTPD6) interactome was generated with GeneMANIA; B: 76 putative ENTPD6-binding partners were retrieved from STRING; C: A Venn diagram highlights the genes shared by GeneMANIA, STRING, and Gene Expression Profiling Interactive Analysis; D: A pan-cancer heatmap illustrates the correlation between ENTPD6 and ENTPD5/glycogen phosphorylase B expression; E and F: Kyoto Encyclopedia of Genes and Genomes pathway and Gene Ontology enrichment analyses delineate the functional landscape of ENTPD6-associated genes. ^aP < 0.05; ^bP < 0.01; ^cP < 0.001. ENTPD5: Ectonucleoside triphosphate diphosphohydrolase 5; PYGB: Glycogen phosphorylase B; KEGG: Kyoto Encyclopedia of Genes; GO: Gene Ontology.

The establishment of a dynamic nomogram

ROC curves demonstrated ENTPD6’s high diagnostic accuracy in various cancers, including OV, THYM, TGCT, THCA, ACC, PAAD, and SKCM, with the AUC values above 0.8 (Figure 9A). Based on the results of the ROC analysis and the tumor sample sizes, THCA with a larger sample size and a higher AUC value was selected to construct a clinical diagnostic model. Table 1 summarizes the baseline characteristics of the training and testing sets. A logistic regression model predicting the likelihood of tumor (vs normal) identified both ENTPD6 and gender as significant predictors (Table 2). Higher ENTPD6 expression was strongly associated with reduced odds of tumor occurrence (Figure 9B). Males had significantly lower tumor odds compared to females (Figure 9B). The model showed excellent fit, with substantial deviance reduction from the null model, confirming its explanatory power beyond baseline. We then validated the model using calibration plots for predictive model calibration, ROC for prediction efficiency evaluation, and DCA for demonstrating the clinical benefits of each model. The calibration curves demonstrated that the training and testing sets exhibited a close fit (Figure 9C and D). The model demonstrated high predictive efficacy, with AUC values of 0.911 and 0.934 in the training and testing sets, respectively (Figure 9E and F). The DCA showed that the model provided a net benefit over the treat-all and treat-none strategies within a high-risk threshold range of 0.7 to 0.9 (Figure 9G and H). All the findings suggest that the dynamic nomogram could serve as a valuable tool for clinicians to make beneficial clinical decisions.

Figure 9 The development of a nomogram. A: Receiver operating characteristic curve analysis revealed tumor types with area under the curve values > 0.6, based on the combined The Cancer Genome Atlas and Genotype-Tissue Expression databases; B: Risk factors of gender and ectonucleoside triphosphate diphosphohydrolase 6 expression were used to construct a nomogram in thyroid carcinoma; C and D: The calibration curves of the training and testing sets; E and F: The receiver operating characteristic analysis of the training and testing sets; G and H: The decision-curve analysis results of the training and testing sets. ^bP < 0.01, ^cP < 0.001. ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6; AUC: Area under the curve; CI: Confidence interval; ROC: Receiver operating characteristic.

Table 1 The baseline characteristics of the training and validation sets, n (%).

Variable	Testing (n = 254)	Training (n = 596)	P value
ENTPD6	5.26 (0.66)	5.23 (0.90)	0.705
Gender			0.481
Female	150 (59.1)	369 (61.9)
Male	104 (40.9)	227 (38.1)

ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6.

Table 2 The logistic regression results for the training set, mean ± SD/n (%).

Dependent (type)1	Normal (n = 237)	Tumor (n = 359)	OR (univariable)	OR (multivariable)
ENTPD6	5.8 ± 1.2	4.9 ± 0.4	0.05 (0.03-0.07)b	0.06 (0.03-0.09)b
Gender
Female
Male	135 (57)	92 (25.6)	0.26 (0.18-0.37)b	0.47 (0.30-0.74)a

^aP < 0.01.

^bP < 0.001.

¹Tumor vs normal.

ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6; OR: Odds ratio.

Drug sensitivity and molecular docking simulations

To identify potential drugs whose efficacy was tightly linked to ENTPD6 expression, we performed a drug-sensitivity analysis. Gene Set Cancer Analysis results revealed that ENTPD6 expression was negatively correlated with the response to SB590885, PD-0325901, RDEA119, dabrafenib, trametinib, and selumetinib, indicating that high ENTPD6 expression was associated with increased sensitivity to these drugs (Figure 10A). In contrast, a positive correlation was observed between ENTPD6 expression and BX-912, BX-795, FK866, and camptothecin, suggesting that high ENTPD6 expression could potentially lead to these drugs’ resistance (Figure 10A). Considering the tight correlation between drug sensitivity and ENTPD6 expression, we applied molecular docking techniques to simulate docking between the negative correlation drugs and the ENTPD6 protein. Interestingly, the binding energies between ENTPD6 and several drugs were calculated as follows: SB590885 at -7.44 kcal/mol, selumetinib at -6.41 kcal/mol, RDEA119 at -6.35 kcal/mol, PD-0325901 at -6.19 kcal/mol, dabrafenib at -8.27 kcal/mol, and trametinib at -7.64 kcal/mol, respectively (Figure 10B-G).

Figure 10  Drug sensitivity analysis. A: Drug-sensitivity profiling of ectonucleoside triphosphate diphosphohydrolase 6 (ENTPD6) was conducted online using the Gene Set Cancer Analysis platform; B-G: Molecular docking results of ENTPD6 protein with SB590885, selumetinib, RDEA119, PD-0325901, dabrafenib, and trametinib. Silvery white represents ENTPD6 protein, blue represents amino acid residue, yellow represents hydrogen atoms, and magenta represents drugs. GDSC: Genomics of Drug Sensitivity in Cancer; ENTPD6: Ectonucleoside triphosphate diphosphohydrolase 6; FDR: False discovery rate.

DISCUSSION

This study comprehensively analyzed the molecular characteristics, predictive potential, genomic associations, immunotherapy relevance, and drug sensitivity profiles of ENTPD6 across pan-cancer. We observed significant ENTPD6 upregulation in multiple cancers, including ESCA, COAD, READ, PAAD, and STAD. Given the limited literature exploring ENTPD6’s relationship with most cancer types, our findings highlight a significant gap warranting further investigation.

The influence of genomic alterations on tumor progression has been well-documented in previous studies[38]. Genomic instability, often linked to low gene methylation, can increase tumor heterogeneity and drive cancer evolution[39]. In this study, we identified the associations between ENTPD6 and mutational profiles in specific cancers, coupled with hypomethylation, suggesting its potential role in tumorigenesis and progression. Furthermore, ENTPD6 expression correlated with RNA modification-related genes, implying potential regulation by post-transcriptional mechanisms. However, further experimental verification is required to confirm whether post-translational modifications regulate ENTPD6 expression.

M1 macrophages mediate anti-tumor immunity through the secretion of a variety of cytokines, such as tumor necrosis factor-alpha, interleukin (IL)-6, IL-12, and IL-23, whereas M2 macrophages enhance the proliferation and migration of tumor cells[40]. Tumor-infiltrating CD8+ T cells mediate potent antitumor effects in multiple malignancies through a coordinated cascade of immunologic processes, including antigen presentation, T cell priming, and functional execution[41]. NK cells mediate critical innate immunosurveillance against malignancies through their capacity to intrinsically recognize abnormal cells and execute rapid cytolytic elimination[42]. We found that ENTPD6 expression was positively associated with M2 macrophage infiltration in BRCA, negatively correlated with CD8+ T-cell abundance in PAAD, LUSC, HNSC, and STAD, and positively linked to activated NK-cell infiltration in PAAD. Previous study showed that ENTPD6 exhibited efficient hydrolysis of uridine diphosphate (UDP) and guanosine diphosphate (GDP), while demonstrating reduced catalytic activity toward cytidine triphosphate, deoxythymidine triphosphate, cytidine diphosphate, adenosine diphosphate, and guanosine triphosphate in bacterial models[43]. Extracellular UDP exploits immunosuppressive tumor-associated macrophages via P2Y6 receptor activation, thereby affecting cytotoxic T cell infiltration, proliferation, and effector function[44]. Ding et al[45] reported that GDP-mannose combined with anti-programmed cell death 1 antibody significantly increased CD8+ T-cell tumor infiltration compared to anti-programmed cell death 1 monotherapy in murine models. Based on these findings, we speculate that ENTPD6 exquisitely modulates extracellular purine and pyrimidine balance by hydrolyzing GDP and UDP, thereby shaping the TME.

Immune checkpoint inhibitors modulate T cell activity and antitumor immune responses through key signaling pathways, thereby reinstating immune-mediated cancer cell elimination[46]. Our study revealed a positive correlation between ENTPD6 expression and multiple immunological checkpoint genes in DLBC, LIHC, OV, PCPG, and LAML, suggesting that higher ENTPD6 expression may predict better immunotherapy response. In colorectal cancers, MMR deficiency or MSI-high status is associated with favorable responses to immunotherapy[37]. MSI-high is often linked to TMB-high, which correlates with the production of more neo-antigens and better outcomes after immune checkpoint inhibitor therapy[47]. However, MMR deficiency was recognized as the leading cause of higher MSI in specific cancers[48]. We further observed that ENTPD6 expression positively correlated with MMR genes in most cancer types. Given that mutations in MMR genes impair DNA repair and promote tumorigenesis[49], the strong associations between ENTPD6 expression and TMB/MSI in certain cancers indicate that ENTPD6 could serve as a valuable biomarker for predicting immunotherapy response. These findings support its potential utility in developing personalized treatment strategies.

KEGG enrichment analysis indicated ENTPD6 involvement in purine metabolism, nucleotide metabolism, and pyrimidine metabolism. A multivariable Mendelian randomization analysis has demonstrated that genetically elevated ENTPD6 expression levels were associated with an increased risk of visceral obesity[50]. Biochemical characterization studies further elucidated ENTPD6’s enzymatic properties, but its negligible hydrolytic capacity for adenosine triphosphate[43]. These findings were corroborated by subsequent research, which demonstrated that soluble ENTPD6 had a limited ability to hydrolyze nucleoside triphosphates and a complete absence of nucleoside monophosphates function[5]. Instead, it showed a clear preference for pyrimidine and purine nucleoside diphosphates, particularly those with oxygen at special positions[5]. This conserved enzymatic specificity across experimental models validates our pathway enrichment results and confirms ENTPD6’s critical role in nucleotide metabolism regulation. The enhanced nucleotide metabolism pathway is vital for cancer cell behaviors, such as proliferation, chemotherapy resistance, immune evasion, and metastasis[51]. Therefore, we speculate that ENTPD6 may influence tumor occurrence and development through its effects on nucleotide and purine metabolism.

GO enrichment analysis revealed that ENTPD6-coexpressed genes were significantly enriched in transport vesicle, Golgi apparatus sub-compartment, and trans-Golgi network, consistent with the results of the subcellular localization analysis. Notably, related-gene network analysis showed ENTPD5 was closely linked to ENTPD6. ENTPD5, an endoplasmic reticulum enzyme upregulated by active AKT signaling, catalyzes the hydrolysis of UDP to uridine monophosphate to maintain protein N-glycosylation and folding fidelity[52]. High ENTPD5 expression is consistently associated with gain-of-function p53 mutations across a broad spectrum of tumor entities[53]. The mutp53-ENTPD5 axis drives N-glycoproteins integrin-α5/integrin-β1 expression and enhances tumor-cell motility via the calnexin/calreticulin cycle, thus promoting cancer metastasis[54]. Structurally, both ENTPD5 and ENTPD6 lack the C-terminal transmembrane domain[55], existing as soluble monomeric enzymes[43]. Unlike cell surface-localized nucleoside triphosphate diphosphohydrolases (1, 2, 3, 8), they exhibit intracellular localization with regulated secretion[8]. This shared architecture underlies their functional synergy, corroborating our interaction data. Clinically, ENTPD5 overexpression has been implicated in the progression of some malignancies, including TGCT, prostate cancer, and lung cancer, suggesting parallel oncogenic roles for ENTPD6[56-58].

To identify therapeutic agents targeting ENTPD6, we performed drug-sensitivity profiling. SB590885, dabrafenib, and trametinib demonstrated high sensitivity against ENTPD6 and exhibited significantly low docking binding energies. Given the ENTPD5/6 homology, these agents warrant further evaluation for dual-targeting therapeutic strategies. Although this study systematically revealed the relationship between ENTPD6 expression and pan-cancer, it has some limitations. Firstly, we encountered some contradictory findings in specific cancers, indicating that further investigations using cell and animal experiments are needed to clarify the biological functions and molecular mechanisms. Secondly, the relatively limited sample sizes for some cancer types may affect the precision of these results.

CONCLUSION

Our study provides a comprehensive analysis of ENTPD6 expression patterns across various cancers, highlighting its potential as a novel biomarker and therapeutic target. This is the first investigation to integrate ENTPD6 expression with patient prognosis, DNA methylation, genomic alterations, KEGG pathway and GO enrichment analyses, epigenetic modifications, genomic instability, immunotherapy response, IHC staining, drug sensitivity, and immune cell infiltration in the context. Our findings demonstrate that ENTPD6 acts as an oncogene in specific cancers. Based on the existing literature and our research outcomes, future studies should focus on validating ENTPD6-targeted therapies in preclinical models and examining ENTPD6’s potential synergy with current checkpoint inhibitors.