Integrated pan-cancer and melanoma-specific analysis of angiopoietin-2: prognostic value, immune microenvironment modulation, and ceRNA network regulation

Abstract

Objective

This research seeks to comprehensively explore the expression patterns of Angiopoietin-2 (ANGPT2) in pan-cancer and examine its relationship with clinical outcomes, tumor immune microenvironment dynamics, and biological functions, with particular emphasis on skin cutaneous melanoma (SKCM).

Methods

Data from six databases, including UCSC Xena, TCGA, GTEx, TIMER2.0, GEPIA, and cBioPortal, were analyzed to assess ANGPT2 expression in pan-cancer. Immunohistochemical images were sourced from the HPA database. We explored the correlation between ANGPT2 expression and prognosis, diagnostic value, genetic alterations, and immune cell infiltration in cancers. Functional enrichment and gene set enrichment analyses were performed to uncover the biological roles of ANGPT2. Additionally, miRWalk, miRDB, Starbase 2.0 ENCORI, and Cytoscape were utilized to identify and construct the lncRNA-miRNA-ANGPT2 ceRNA regulatory network in SKCM.

Results

Our comprehensive pan-cancer analysis revealed that ANGPT2 undergoes genetic alterations in several tumor types, including mutations, amplifications, and deep deletions. ANGPT2 expression varies across cancers, with high levels detected in 21 tumor types. ANGPT2 demonstrated significant diagnostic and prognostic value in various cancers. Genes related to ANGPT2 in cancers with high expression were found to be involved in critical pathways such as tumor angiogenesis, growth factor signaling, extracellular matrix (ECM) structure, and cell surface receptor activation. In addition, ANGPT2 was found to modulate the immune microenvironment in multiple tumors, promoting immune evasion and contributing to tumor progression, particularly in SKCM. A lncRNA-miRNA-ANGPT2 ceRNA regulatory network was identified, which may play a role in SKCM progression.

Conclusions

ANGPT2 demonstrates potential as a biomarker for cancer diagnosis and prognosis, with evidence suggesting its involvement in immune microenvironment regulation and associated gene networks. Therapeutic approaches targeting ANGPT2, either as a stand-alone treatment or in combination with immunotherapies, could offer new and effective strategies for future cancer treatments.

Introduction

SKCM is one of the most aggressive types of skin cancer, characterized by its silent onset, high metastatic potential, and poor prognosis. In 2022, there were 331,722 new cases of SKCM and 58,667 related deaths worldwide [1]. Due to its high somatic mutation burden, melanoma exhibits significant immunogenicity [2]. In patients with advanced melanoma receiving immune checkpoint inhibitors targeting PD-1 and CTLA-4, around 70% attain a 3-year overall survival rate [3–5]. However, 30–50% of patients exhibit primary (de novo) resistance to these inhibitors, largely attributed to the immunosuppressive state of the tumor microenvironment (TME) [6, 7]. This underscores the pressing demand for new prognostic indicators and molecular targets to forecast patient outcomes and support personalized treatment approaches.

Vascular stability is essential in influencing the immunosuppressive TME. Irregular tumor blood vessels cause localized hypoxia, which triggers the hypoxia-inducible factor (HIF) pathway. This activation enhances tumor cell invasiveness and encourages the recruitment of immunosuppressive cells like myeloid-derived suppressor cells (MDSCs) [8, 9]. Hypoxia within the TME also suppresses the activity of anti-tumor effector cells, including CD8⁺ T cells and natural killer (NK) cells [10, 11]. Furthermore, the abnormal vasculature acts as a barrier, limiting the infiltration of immune effector cells and hindering the activation of anti-tumor immune responses, which exacerbates immune evasion [12, 13]. The effectiveness of immune checkpoint inhibitors, such as PD-1 and CTLA-4 blockers, relies on functional vasculature to facilitate sufficient immune cell infiltration [14]. Unfortunately, vascular instability often compromises these therapies and contributes to treatment resistance [15].

ANGPT2, a key regulator of angiogenesis and vascular instability, is typically upregulated in regions of unstable tumor vasculature, particularly in areas undergoing active vascular remodeling and in hypoxic tumor microenvironments. High ANGPT2 expression is frequently associated with poor prognosis and disease progression in various cancers, including gastric, colorectal, and lung cancer [16–19]. Additionally, ANGPT2 upregulation is linked to the recruitment of immunosuppressive cells, such as regulatory T cells and MDSCs. Recent studies have shown that blocking ANGPT2 enhances CD8 + T cell infiltration into the tumor core, thereby boosting the anti-tumor effects of these effector cells [20].

While angiogenesis is a hallmark of cancer, ANGPT2’s role in vascular instability and immune evasion makes it particularly relevant to melanoma. In melanoma, ANGPT2 upregulation correlates with hypoxia-driven immune suppression, including MDSC recruitment, and contributes to resistance against PD-1/PD-L1 inhibitors. Furthermore, elevated ANGPT2 levels are closely correlated with increased expression of immune checkpoint molecules like PD-L1. Research indicates that ANGPT2 inhibitors can reduce PD-L1 expression, thereby improving the effectiveness of PD-1 inhibitors in cancer therapy [21, 22]. These findings suggest that ANGPT2 may be a relevant target for enhancing immune checkpoint blockade efficacy in melanoma. However, the underlying molecular mechanisms and functional significance of ANGPT2 are not yet fully understood, and its role across multiple cancer types remains to be comprehensively explored.

This study aims to investigate the involvement of ANGPT2 across a pan-cancer landscape to determine its diagnostic and prognostic value, as well as its genetic alterations. Special focus is given to its interactions with the immune TME. Additionally, we examine the lncRNA-miRNA-ANGPT2 regulatory network in SKCM. By uncovering the regulatory mechanisms of ANGPT2, we seek to gain deeper insights into its potential as a therapeutic target across multiple cancers.

Materials and methods

Data sources

The expression data for ANGPT2 and clinical information for patients across various cancers were obtained from The Cancer Genome Atlas (TCGA) [23] and Genotype-Tissue Expression (GTEx) databases [24], accessed via the UCSC XENA platform (https://xenabrowser.net/datapages/). Immunohistochemistry(IHC) images of both normal and tumor tissues were retrieved from the Human Protein Atlas (HPA) (https://www.proteinatlas.org/) [25]. For identifying the top 100 genes most correlated with ANGPT2, we utilized the Gene Expression Profiling Interactive Analysis (GEPIA2) database using data from TCGA [26], while genetic alterations were analyzed through cBioPortal [27]. Immune cell infiltration analysis was conducted using TIMER2.0 [28], based on data from the TCGA cohort. As all datasets were sourced from publicly accessible repositories, no ethical approval was necessary. The study workflow is outlined in Fig. 1.

Fig. 1.

Workflow of the study. Flowchart depicting the overall study design and analysis process

Genetic alteration analysis

The cBioPortal platform [29] (http://www.cbioportal.org) was used to analyze genetic alterations in ANGPT2. The “Cancer Types Summary” module provided information on mutation types, copy number alterations (CNA), and mutation frequencies in different cancers.

ANGPT2 expression analysis in cancers

We compared ANGPT2 mRNA expression levels between normal and tumor tissues using TCGA_GTEx, TCGA, and paired TCGA samples. Processed using DESeq2 to compute normalized counts (log2(TPM + 1)) and remove batch effects between TCGA and GTEx using ComBat with linear modeling. Differential expression analysis was performed using the limma R package (significance threshold: padj < 0.05 and |log2FC| >1), and ggplot2 was used for data visualization. ANGPT2 protein levels in normal and tumor tissues were also investigated using the HPA database.

Prognostic and diagnostic value of ANGPT2

RNA sequencing data from TIMER2.0 and TCGA were used to evaluate the relationship between ANGPT2 expression and cancer prognosis and diagnosis. Survival outcomes were assessed using the log-rank test, with Kaplan-Meier curves generated via the survival and survminer R packages. Hazard ratios (HR) with 95% confidence intervals (CI) and P-values were calculated, and forest plots created using forestplot. Diagnostic value was assessed using ROC curves and AUC values, generated with the pROC package. Statistical significance was set at P < 0.05.

Functional enrichment analysis of ANGPT2-related genes in cancers with high ANGPT2 expression

The top 100 genes correlated with ANGPT2 in cancers with high ANGPT2 expression were identified using GEPIA2. Gene Ontology (GO) analysis was performed to explore biological processes (BP), cellular components (CC), and molecular functions (MF), using the clusterProfiler and org.Hs.eg.db R packages. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was also conducted to assess potential functions of ANGPT2.

ANGPT2 expression and immune infiltration analysis

The TIMER2.0 platform (http://timer.comp-genomics.org/) was employed to analyze the correlation between ANGPT2 expression and various immune cell types across cancers in the TCGA dataset. Immune scores for the TCGA-SKCM dataset were generated using the estimate R package, which assesses tumor purity, immune cell infiltration, and stromal components. The CIBERSORT R package was used to explore the relationship between ANGPT2 expression and immune cell infiltration in SKCM, with results visualized using ggplot2.

lncRNA-miRNA-ANGPT2 ceRNA network construction

To predict miRNAs targeting ANGPT2, we used StarBase2.0 ENCORI [30] (https://starbase.sysu.edu.cn/), miRWalk [31] (http://mirwalk.umm.uni-heidelberg.de/), and miRDB [32] (http://mirdb.org/miRDB/). Only miRNAs present in all three databases were considered target miRNAs. The Long non-coding RNAs (lncRNAs)-miRNA interactome was analyzed using StarBase2.0 ENCORI, and correlations between miRNAs and ANGPT2 were assessed. The filtering criteria included “mammals, humans, hg19, strict stringency (≥ 4) of CLIP data, with or without degradome data”. The miRNA-lncRNA interactome and lncRNA-miRNA-ANGPT2 ceRNA network were visualized using R (with the ggalluvial and ggplot2 packages) and Cytoscape, respectively.

Statistical analysis

All statistical analyses were performed using R software (version 4.4.1). Differences between groups were analyzed using the Wilcoxon rank-sum test for continuous variables and χ² tests for categorical variables. Associations were evaluated using Pearson correlation analysis. Multiple testing correction was performed using the Benjamini-Hochberg procedure (α = 0.05). Data are presented as mean ± standard deviation (SD), with significance levels reported as *P < 0.05, **P < 0.01, and ***P < 0.001. Missing data were imputed using the knnImpute method (for missing rate < 10%) or excluded from survival analyses.

Results

Genetic alteration analysis of ANGPT2

We conducted a pan-cancer genetic alteration analysis of ANGPT2 using the cBioPortal platform. Among 10,967 patient samples from 32 cancer types, 4% exhibited ANGPT2 genetic alterations (Fig. 2A). The majority of these mutations were concentrated in the Fibrinogen_C domain of the ANGPT2 protein, with the R368C mutation—where arginine (R) is substituted by cysteine (C)—identified as a prominent hotspot (Fig. 2B). Mutation frequencies varied across cancer types, ranging from 0 to 8%. Colorectal, endometrial, and ovarian cancers exhibited higher mutation frequencies, with gene amplifications (red) and deep deletions (blue) being the most prevalent alterations. In contrast, melanoma, pancreatic cancer, and prostate cancer displayed lower mutation frequencies, though a notable number of gene mutations (green) were still detected (Fig. 2C). High-frequency mutations, such as 8% in colorectal cancer, may indicate a driver role for ANGPT2 in tumorigenesis, while lower frequencies, such as those seen in melanoma (< 1%), may represent passenger mutations or reflect limitations in detection sensitivity. Further analysis revealed that ANGPT2 mRNA expression levels were elevated in cancers such as melanoma, colorectal cancer, and lung adenocarcinoma, often in conjunction with gene amplifications or other genetic alterations (Fig. 2D). These findings highlight the diverse genetic alterations of ANGPT2 across various cancer types and suggest its potential involvement in tumor progression.

Fig. 2.

Genetic alteration analysis of ANGPT2. A Genetic alterations of ANGPT2 across various cancers, showing a 4% alteration rate in pan-cancer samples. B Specific alteration sites within the ANGPT2 protein structure. C Frequency and types of ANGPT2 genetic alterations observed across different cancer types. D Correlation between ANGPT2 mRNA expression and genetic alterations across various cancers

Expression of ANGPT2 in pan-cancer

Following the genetic analysis, we systematically evaluated ANGPT2 expression across 33 tumor types using TCGA_GTEx and TCGA datasets. TCGA data showed significant differences in ANGPT2 expression between tumor tissues and paired normal tissues in cancers such as CHOL, COAD, ESCA, HNSC, KIRC, LIHC, LUAD, READ, SKCM, STAD, and THCA (Fig. 3A and C). In addition, integrated TCGA and GTEx data indicated that ANGPT2 was significantly upregulated in cancers like ACC, CHOL, COAD, DLBC, ESCA, GBM, HNSC, KICH, KIRC, LAML, LGG, LIHC, OV, PAAD, PCPG, READ, SKCM, STAD, TGCT, THYM, and UCS (P < 0.05), whereas it showed a downregulation trend in BRCA, CESC, PRAD, and THCA (P < 0.05) (Fig. 3B). Notably, ANGPT2 was significantly upregulated in primary SKCM compared to normal skin tissues, with even higher expression observed in metastatic SKCM (Fig. 3A and B). IHC images from the HPA database confirmed elevated ANGPT2 protein expression in cancers such as KICH, LGG, LIHC, PAAD, and SKCM, which was consistent with the mRNA expression data (Fig. 3D).

Fig. 3.

Expression of ANGPT2 in pan-cancer. A ANGPT2 mRNA expression levels in tumor versus normal tissues from the TIMER2.0 database. B ANGPT2 mRNA expression comparison between tumor and normal tissues from the combined TCGA and GTEx datasets. C Expression levels of ANGPT2 in tumors compared to paired adjacent normal tissues from the TCGA dataset. D ANGPT2 protein expression in tumor and normal tissues, based on data from the Human Protein Atlas (HPA). (ns: not significant, P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001)

Correlation of ANGPT2 with pan-cancer prognosis and diagnosis

We evaluated the prognostic significance of ANGPT2 using forest plots and Cox regression analysis. High ANGPT2 expression was associated with poorer overall survival in cancers such as BRCA-LumA, CESC, LGG, MESO, SKCM-Primary, and UVM (P < 0.05, HR > 1). Conversely, ANGPT2 expression appeared to serve as a protective factor in BRCA-Basal, HNSC-HPV-, and KIRC (P < 0.05, HR < 1) (Fig. 4A). Kaplan-Meier survival analysis indicated that higher ANGPT2 expression was correlated with reduced disease-specific survival (DSS) in cancers such as ESCA, LIHC, LGG, and SKCM (Fig. 4B). ROC curve analysis demonstrated that ANGPT2 had strong diagnostic value in ACC, PAAD, and STAD, with AUC values approaching or exceeding 0.9, though its predictive power was weaker in LGG, KICH, and UCS (Fig. 4C). These findings highlight ANGPT2’s significant prognostic and diagnostic potential across various cancers, including SKCM.

Fig. 4.

Correlation of ANGPT2 with Pan-Cancer Prognosis and Diagnosis and Functional Enrichment Analysis of Related Genes. A Forest plot showing the association between ANGPT2 expression and overall survival in pan-cancer analysis (P < 0.05). B Kaplan-Meier survival curves displaying the relationship between ANGPT2 expression and disease-specific survival (DSS) in cancers such as ACC, ESCA, KICH, LGG, LIHC, PAAD, SKCM, STAD, and UCS. C ROC curve analysis demonstrating the diagnostic value of ANGPT2 in cancers like ACC, ESCA, KICH, LGG, LIHC, PAAD, SKCM, STAD, and UCS. D Gene Ontology (GO) enrichment analysis of biological processes (BP), cellular components (CC), and molecular functions (MF) of the top 100 ANGPT2-related genes in tumors with high ANGPT2 expression. E KEGG pathway analysis of the top 100 ANGPT2-related genes in tumors with high ANGPT2 expression, showing involvement in key cancer-related pathways

Functional enrichment analysis of ANGPT2-related genes

We identified the top 100 genes most correlated with ANGPT2 expression in cancers with high ANGPT2 expression using the GEPIA2 database. GO analysis revealed that these genes are involved in key biological processes such as “regulation of vasculature development,” “angiogenesis,” “endothelial cell proliferation,” and “vascular endothelial growth factor receptor signaling.” These findings suggest that ANGPT2-related genes are significantly enriched in pathways associated with vascular regulation and tumor angiogenesis. The cellular component analysis highlighted their role in structures such as the “collagen-containing extracellular matrix,” “plasma membrane,” and “cell-cell junctions,” suggesting involvement in cell communication and ECM interactions. Molecular function analysis revealed enrichment in “growth factor binding,” “transmembrane receptor protein kinase activity,” and “integrin binding,” reinforcing ANGPT2’s role in cell signaling and tumor growth (Fig. 4D). KEGG pathway analysis indicated enrichment in cancer-related signaling pathways, including PI3K-Akt, focal adhesion, and Ras, suggesting ANGPT2’s involvement in tumor cell proliferation, ECM interactions, angiogenesis, and signal transduction (Fig. 4F). These results provide strong evidence of the involvement of ANGPT2 in tumor progression and its potential as a therapeutic target.

ANGPT2 expression and tumor immune microenvironment in pan-cancer

The tumor immune microenvironment plays a critical role in tumorigenesis and progression. We analyzed the relationship between ANGPT2 expression and immune cell infiltration using the TIMER2.0 database. In cancers such as SKCM, ANGPT2 expression was positively correlated with the infiltration of cancer-associated fibroblasts and endothelial cells, suggesting that high ANGPT2 expression may be linked to fibroblast enrichment and angiogenesis regulation within the tumor microenvironment (Fig. 5A). We further employed the ESTIMATE scoring system to assess tumor purity, immune cell infiltration, and stromal content in the tumor microenvironment. The results indicated a positive correlation between ANGPT2 expression and both the ESTIMATE score and StromalScore in SKCM (Figs. 5B-C), suggesting that high ANGPT2 expression may be associated with increased immune and stromal components in the tumor microenvironment. Moreover, we used the CIBERSORT algorithm to assess immune cell infiltration in the high and low ANGPT2 expression groups in SKCM. The results showed that T cell CD4 memory resting cells and M0 macrophages were significantly enriched in the high ANGPT2 expression group, whereas follicular helper T cells, regulatory T cells (Tregs), and CD8 + T cells were decreased (Fig. 5D). These results suggest that ANGPT2 may contribute to immune evasion and tumor progression in SKCM by modulating the immune microenvironment.

Fig. 5.

ANGPT2 expression and immune infiltration analysis. A Correlation between ANGPT2 expression and immune cell infiltration across pan-cancer samples. B Association between ANGPT2 mRNA expression and ESTIMATE scores in SKCM, indicating the tumor microenvironment composition. C Relationship between ANGPT2 mRNA expression and stromal scores in SKCM. D Comparison of immune cell infiltration between high and low ANGPT2 expression groups in SKCM

lncRNA-miRNA-ANGPT2 network construction in SKCM

To further elucidate the complex regulatory mechanisms of ANGPT2 in SKCM progression, we constructed a competitive endogenous RNA (ceRNA) regulatory network related to ANGPT2. Using the miRWalk, miRDB, and Starbase2.0 ENCORI databases, we identified miRNAs targeting ANGPT2 in SKCM. From these sources, we identified 1866, 166, and 32 ANGPT2-related miRNAs, respectively, and ultimately filtered seven common target miRNAs, including hsa-miR-145-5p, hsa-miR-205-5p, hsa-miR-514b-5p, hsa-miR-3121-3p, hsa-miR-3163, hsa-miR-4436a, and hsa-miR-5000-3p (Fig. 6A). Typically, mRNA expression is inversely correlated with the expression of its targeting miRNAs [33]. We found that ANGPT2 expression was significantly negatively correlated with hsa-miR-205-5p, hsa-miR-514b-5p, hsa-miR-3121-3p, hsa-miR-3163, and hsa-miR-5000-3p (Fig. 6B). To explore the target lncRNAs of these five miRNAs, we used Starbase2.0 ENCORI to identify the corresponding lncRNAs (Fig. 6C). lncRNAs can function as ceRNAs by competing with miRNAs, thereby regulating mRNA expression [34]. We identified lncRNAs targeted by these five miRNAs, which may participate in ANGPT2 regulation through a negative regulatory mechanism. Ultimately, we constructed a lncRNAs-miRNAs-ANGPT2 ceRNA regulatory network for SKCM (Fig. 6D).

Fig. 6.

lncRNA-miRNA-ANGPT2 network construction in SKCM. A Venn diagram showing predicted ANGPT2-targeting miRNAs from miRDB, miRWalk, and Starbase 2.0 ENCORI databases. B Correlation analysis between ANGPT2 expression and miRNAs, including hsa-miR-205-5p, hsa-miR-514b-5p, hsa-miR-3121-3p, hsa-miR-3163, and hsa-miR-5000-3p. C Relationships between the identified miRNAs and their target lncRNAs. D A ceRNA regulatory network involving lncRNAs, miRNAs, and ANGPT2 was constructed for SKCM

Discussion

Early detection and timely treatment are vital for improving cancer prognosis. Recently, multi-omics pan-cancer analyses have increasingly been used to identify gene mutations, RNA alterations, and key pathways involved in cancer progression, providing valuable insights into potential tumor biomarkers. In this study, we performed a comprehensive pan-cancer and melanoma-specific analysis to evaluate the prognostic value of ANGPT2, its role in modulating the immune microenvironment, and its regulation within the ceRNA network.

To date, no large-scale bioinformatics analysis has explored ANGPT2’s function across multiple databases and cancer types. Gene mutations caused by sequence alterations are critical drivers of cancer development and progression [35]. Our analysis revealed that ANGPT2 exhibits a high frequency of mutations across various tumor types. Notably, these mutations are concentrated in the Fibrinogen_C domain of the ANGPT2 protein, with the R368C mutation being particularly prominent. The substitution of arginine with cysteine at this site may alter protein function, affecting ANGPT2’s interactions with other molecules or signaling pathways, thereby influencing tumor angiogenesis and progression. Furthermore, ANGPT2 is associated with vascular abnormalities, including capillary malformations and Sturge-Weber syndrome, which involves the GNAQ p.R183Q mutation [36, 37]. ANGPT2 genetic alterations are particularly common in colorectal, endometrial, and ovarian cancers, with gene amplification and deletions being the primary forms of variation. Amplifications, in particular, are often linked to overexpression, which can promote tumor cell survival by enhancing angiogenesis or altering vascular structure. Thus, ANGPT2 holds promise as a potential diagnostic biomarker for detecting mutations and conducting pan-cancer analyses.

Our analysis of 33 cancer datasets from TCGA and GTEx revealed that ANGPT2 is significantly overexpressed in 21 cancer types, including SKCM, STAD, and ESCA, while it is downregulated in BRCA, CESC, PRAD, and THCA. ANGPT2 overexpression is closely associated with poor prognosis in several cancers, especially melanoma. Our data showed that ANGPT2 is highly expressed in primary melanoma and further elevated in metastatic melanoma, suggesting that it plays a key role in melanoma progression and metastasis. A detailed analysis of SKCM patient samples confirmed a correlation between ANGPT2 expression, clinical stage, and lymph node metastasis [38]. Additionally, previous studies demonstrated that malignant melanoma cells secrete high levels of ANGPT2. Building on this, we developed ANGPT2-targeted nanoparticles carrying small interfering RNA (siRNA) and applied them in a melanoma mouse model. This approach significantly reduced tumor angiogenesis, normalized vasculature, and promoted tumor cell apoptosis, leading to reduced tumor size [39, 40].

ANGPT2 antagonizes ANGPT1 by inhibiting Tie-2 receptor activation, leading to unstable vasculature [41]. ANGPT2 not only promotes angiogenesis but also influences tumor proliferation, invasion, and metastasis through extracellular matrix (ECM) remodeling and regulation of key signaling pathways. The ECM, a component of the stromal and epithelial vascular matrix, consists of structural and adhesive proteins, glycoproteins, and secreted factors that interact with cells to transmit extracellular signals [42]. ECM remodeling creates a permissive environment for tumor growth, enabling high proliferation, invasion, and metastasis [43–45]. KEGG pathway analysis revealed that ANGPT2-related genes are strongly associated with cancer progression, particularly through the PI3K-Akt and focal adhesion pathways. The PI3K-Akt pathway is commonly dysregulated in cancer and promotes cell proliferation and survival [46, 47]. The focal adhesion pathway regulates cell-ECM interactions, promoting tumor migration and invasion, further highlighting ANGPT2’s role in metastasis [48].

The immune TME is crucial in tumor development and progression [49]. Our analysis using the TIMER2.0 database revealed that high ANGPT2 expression is significantly correlated with the infiltration of cancer-associated fibroblasts and endothelial cells across multiple cancer types, particularly melanoma. High ANGPT2 expression was associated with higher tumor purity and stromal scores, along with increased infiltration of CD4⁺ memory T cells and M0 macrophages. In contrast, follicular helper T cells, regulatory T cells (Tregs), and CD8⁺ T cells were significantly reduced. This suggests that ANGPT2 promotes vascular instability, fibroblast infiltration, suppression of cytotoxic immune cells, and recruitment of immunosuppressive cells, creating an immunosuppressive TME.

This immunosuppressive environment fosters immune evasion, a key factor in resistance to immune checkpoint inhibitors. Moreover, higher ANGPT2 levels are strongly linked to increased expression of immune checkpoint molecules such as PD-L1. Studies suggest that inhibiting ANGPT2 can reduce PD-L1 expression, thus enhancing the efficacy of PD-1 inhibitors in cancer treatment [21, 22]. For immunotherapy-resistant tumors, targeted therapies are urgently needed. In advanced ovarian and colorectal cancers, combining ANGPT2 inhibition with immune checkpoint blockade has shown potential to limit tumor growth by inhibiting angiogenesis while enhancing immunotherapy by improving the TME [50]. A phase I clinical trial combining CTLA-4 and ANGPT2 blockade in melanoma demonstrated safety and efficacy [51], further supporting ANGPT2 as a promising biomarker for predicting response to immunotherapy across multiple cancers, including SKCM.

Further research is needed to investigate how ANGPT2 expression is regulated in SKCM immunotherapy. lncRNAs regulate mRNA expression by competitively binding to miRNAs. In recent years, lncRNAs and miRNAs have gained attention for their roles in gene regulation in cancer [52, 53]. Studies have shown that a ceRNA network involving ANGPT1, ANGPT2, VEGF-A, and specific miRNAs (e.g., hsa-miR-190a-3p, hsa-miR-374c-5p, hsa-miR-452-5p, hsa-miR-889-3p), along with lncRNAs (e.g., AFAP1-AS1, KCNQ1OT1, and MALAT1), plays a critical role in colorectal cancer progression [54]. Additionally, downregulation of lncRNA GAS5 has been shown to inhibit ANGPT2 via miR-17-3p, reduce apoptosis, and promote ECM remodeling [55]. In this study, we constructed a ceRNA network based on hsa-miR-205-5p, hsa-miR-514b-5p, hsa-miR-3121-3p, hsa-miR-3163, and hsa-miR-5000-3p, along with lncRNAs such as SNHG14, NEAT1, and 21 other lncRNAs, to explore the underlying mechanisms of SKCM progression. This network operates through the SNHG14/NEAT1 lncRNAs, which act as ceRNA hubs by binding miRNAs that target ANGPT2. This regulation enhances angiogenesis via the PI3K-Akt/VEGF pathways and supports immunosuppression by promoting M0 macrophage enrichment and suppressing CD8 + T cell activity, thereby directly contributing to the progression of SKCM. In summary, our study underscores the complexity and diversity of ANGPT2’s role in tumorigenesis across different cancer types. Further investigation into ANGPT2’s regulatory mechanisms will not only enhance our understanding of tumor biology but also present valuable opportunities for developing targeted cancer therapies.

Limitations

Although we conducted a thorough data analysis, there are several limitations to consider. The relatively small sample size and lack of experimental validation may impact the robustness of our conclusions. While we applied reliable statistical methods for correlation and survival analysis, the use of public datasets introduces inherent limitations. For instance, the absence of confidence intervals in the immune infiltration analysis—due to algorithmic constraints—may reduce the accuracy of some findings. Furthermore, statistical methods such as the Benjamini-Hochberg correction, though commonly used, rely on assumptions that may not always apply to every dataset. Future research should aim to increase the sample size and include both in vivo and in vitro experiments to provide a more comprehensive understanding of ANGPT2’s regulatory role in SKCM.

Conclusions

In conclusion, our extensive study has demonstrated that ANGPT2 expression undergoes significant changes across a range of cancers, including skin cutaneous melanoma (SKCM), and is strongly linked to cancer development and patient survival outcomes. Moreover, ANGPT2 expression was found to be closely associated with the infiltration of cancer-associated fibroblasts (CAFs), CD8⁺ T cells, and regulatory T cells (Tregs), indicating its potential involvement in shaping the immune microenvironment. The lncRNA-miRNA-ANGPT2 ceRNA network also emerged as a critical regulator of SKCM progression. Although these results underscore the potential of ANGPT2 as a valuable prognostic marker and target for immunotherapy-based treatments, it is crucial to note that our study relies primarily on correlational data. Therefore, further experimental research, particularly focused on establishing causality and exploring the underlying mechanisms, is required to validate these conclusions and deepen our understanding of ANGPT2’s role in cancer progression.

Author contributions

XN, XW and BC analyzed the data, and wrote the manuscript; XS and BW conceived the project and revised the manuscript; while HX, LC and QL edited the manuscript; RK, TH and HY researched the data. All authors read and approved the final manuscript.

Funding

This study was funded by the National Natural Science Foundation of China(No.82472576); Natural Science Fundation of Fujian Province (No.2021J01244); Natural Science Fundation of Fujian Province (No. 2024J08167).

Data availability

Data Availability StatementThe study incorporated data from publicly accessible websites, as detailed in the “Methods” section. Additional information is available upon contacting the corresponding authors at biaowang@fjmu.edu.cn.

Declarations

Competing interests

The authors declare no competing interests.