Introduction

Cancer remains one of the leading causes of mortality worldwide. While cancer incidence rates vary across countries with differing income levels, it negatively impacts life expectancy in all regions1. Traditional cancer treatments primarily include surgical resection, chemotherapy, and radiotherapy. However, recent advances have led to the development of new treatments, such as immunotherapy and oncolytic viruses2,3,4. Despite these advancements, the heterogeneity among patients with different cancers remains a significant challenge. For instance, CAR-T therapy has shown promising results in hematological tumors but faces challenges in treating solid tumors5,6,7. Additionally, patients with the same tumor type can exhibit varying responses to the same treatment2,8. Although cancer types differ in their mechanisms of development, they share common characteristics, such as uncontrolled proliferation, increased invasiveness, and immune evasion9. Therefore, pan-cancer studies may uncover both commonalities and differences among cancers, potentially leading to the identification of novel therapeutic targets and biomarkers.

A disintegrin and metalloprotease (ADAM) family consists of transmembrane and secreted proteins characterized by metalloprotease domains10. These proteins are involved in cell adhesion and the hydrolytic processing of the ectodomains of various cell-surface receptors and signaling molecules10,11. They play a crucial role in regulating cellular behaviors, such as cell adhesion, migration, protein hydrolysis, and signaling12,13,14,15. ADAM12, a member of this family, has been shown to play a pro-carcinogenic role in various cancers. It promotes the migration and invasion of breast cancer cells16, the proliferation of hepatocellular carcinoma cells17, and the resistance of rectal cancer cells to radiotherapy18. Nevertheless, most studies on ADAM12 have focused on individual cancer types, failing to provide a comprehensive overview at the pan-cancer level.

In this study, multi-omics data were used to explore the potential role and prognostic value of ADAM12 in pan-cancer, with a particular focus on gliomas. Experimental validation revealed that ADAM12 expression is upregulated in glioma tissues and promotes glioma cells proliferation, migration, and invasion. These findings suggest that ADAM12 this study, multi-omics data were used to explore the potential role and prognostic value of ADAM12 in pan-cancer, with a particular focus on gliomas. Experimental validation revealed that ADAM12 could be a potential therapeutic target for glioma treatment.

Materials and methods

Acquisition of data sources

Harmonized and standardized pan-cancer expression data from TCGA, along with normal tissue expression data from GTEx, were downloaded from the UCSC database. Subsequently, a log2(x + 0.001) transformation was applied to the mRNA sequencing data for ADAM12 in 34 distinct tumor tissues and 31 normal tissues, along with associated clinical information. Cell line data and images from the Cancer Cell Line Encyclopedia (CCLE) and the Human Protein Atlas (HPA) datasets were utilized. The Gene Expression Profiling Interactive Analysis (GEPIA) website (http://gepia2.cancer-pku.cn/index.html) was used to assess the variance in ADAM12 mRNA expression across tumor and normal tissues, and to explore its relationship with overall survival and cancer stage. Additional clinical and RNA-seq data from the mRNAseq_325 dataset and GSE11016 were retrieved from the CGGA database (http://www.cgga.org.cn/) and the Gene Expression Omnibus (GEO) databases for further analysis.

Diagnostic and prognostic analysis

The prognostic significance of ADAM12 was analyzed via overall survival (OS) and disease-specific survival (DSS) using data from the TCGA database. The diagnostic potential of ADAM12 in GBMLGG was evaluated using ROC curve analysis with data from the TCGA database. The prognostic significance of ADAM12 was initially assessed using univariate Cox analysis, followed by multivariate Cox analysis to confirm its independent prognostic significance.

Relationship between ADAM12 levels and tumor heterogeneity

The correlation between ADAM12 levels and mutant-allele tumor heterogeneity (MATH), tumor mutational burden (TMB), and microsatellite instability (MSI) was studied across multiple cancer forms. Utilization of the TCGA Pan-Cancer dataset (PANCAN, N = 10535, G = 60499) from the UCSC database (https://xenabrowser.net/) was carried out. Additionally, the level 4 Simple Nucleotide Variation dataset for all TCGA samples, handled by the MuTect2 software (DOI: 10.1038/nature08822), was obtained from the GDC portal (https://portal.gdc.cancer.gov/). The TMB for each tumor was determined using the TMB function from the maftools R package (version 2.8.05). Gene expression data were integrated with MATH/TMB/MSI data, with each expression value transformed using log2 (x + 0.001).

Correlations between ADAM12 and the immune environment

The deconvo_quantiseq method from the R package IOBR (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8283787/), based on the study “QUANTISEQ: Molecular and Pharmacological Modulators of the Tumor Immune Contexture Revealed by Deconvolution of RNA-seq Data” (DOI: 10.1186/s13073-019-0638-6), was employed to reassess the infiltration scores for each tumor in every patient. These scores included assessments for B cells, M1 and M2 macrophages, monocytes, neutrophils, NK cells, CD4 + T and CD8 + T cells, Tregs, dendritic cells, and others based on gene expression data. Using the ESTIMATE algorithm, the infiltration levels of stromal and immune cells in tumor tissue were assessed through three scoring systems: stromal score, immune score, and estimate score.

Analysis of ADAM12 at the single-cell level

The CancerSEA database was utilized to analyze ADAM12 expression levels across various cell types in pan-cancer, with the findings displayed as heatmaps and scatterplots (https://biocc.hrbmu.edu.cn/CancerSEA/). The Seurat package in R software organized and analyzed single-cell RNA-seq data. Genes present in less than ten cells or with under 200 unique counts were omitted. The “vst” method of Seurat was applied to select the top 1,000 highly variable genes. Dimensionality reduction of the data was achieved using UMAP and t-SNE algorithms. Cell clusters were annotated using the SingleR algorithm. The UCSC Cell Browser (http://cells.ucsc.edu/) and TISCH database (http://tisch.comp-genomics.org/home/) were used for visualization.

Functional enrichment analysis

Using the UCSC Xena database, the disease datasets TCGA-GBMLGG and TARGET/GTEx were obtained, while related clinical information was sourced from the GDC TCGA. Tumor samples were filtered (Genes.group.txt) and then grouped based on the median value of ADAM12 expression. Differential analysis was performed using the limma package (version 3.54.2) with a significance threshold of p-value < 0.05 and |logFC|> 1. Heatmaps were produced with the R package pheatmap (v1.0.12), and volcano plots were generated using ggplot2 (v3.4.2). Important pathways involved in cancer were analyzed using the R package GSVA, with the method set to ‘ssGSEA’. The KEGG pathway database is copyrighted by Kanehisa laboratories19,20. All KEGG imageries conform to the citation guidelines: www.kegg.jp/kegg/kegg1.html. Spearman correlation was employed to analyze the relationship between genes and pathway scores.

Cell culture and transfection

Human glioblastoma multiforme (GBM) cell lines (U87MG and U251MG) were sourced from Procell Life Science & Technology Co., Ltd (Wuhan, China) and verified via short tandem repeat (STR) profiling. In a humidified incubator set at 37 °C with 5% CO2, these cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, USA) with 10% fetal bovine serum (FBS) (Gibco, USA). Three different small interfering RNA (siRNA) sequences targeting ADAM12 (siRNA 1-3) and negative control (siRNA-NC) were synthesized by Wuhan GeneCreate Biological Engineering Co., Ltd. Transfection of cells was performed using Lipofectamine™ 3000 transfection reagent (Invitrogen, Carlsbad, USA). The siRNA sequence demonstrating the highest efficiency in interfering with ADAM12 expression was selected for further analysis. The target sequences for ADAM12 siRNAs are as follows:

si-control: 5′-ATCTCGCTTGGGCGAGAGTAAGTA-3′

ADAM12 si1: 5′-GACTGCTGTTTACAAATAATT-3′

ADAM12 si2: 5′-CCTCAGACCTGCTCCACAATT-3′

ADAM12 si3: 5′-GCAGATAACCAAGGTTTAATT-3′

Tumor samples collection

Human glioma specimens were collected from the Department of Neurosurgery, Union Hospital of Tongji Medical College, Huazhong University of Science and Technology, between January 2023 and January 2024. The study cohort consisted of 12 glioma patients, classified into WHO grade II (n = 4), WHO grade III (n = 4), and WHO grade IV (n = 4). None of the patients had undergone chemotherapy or radiotherapy prior to sample collection.

Cell proliferation and colony formation assay

Transfected U87 and U251 cells were distributed into 96-well culture plates, each well containing 2000 cells. These cells were maintained at 37°℃, and their viability was monitored daily using the CCK-8 kit over a four-day period. Absorbance readings were taken at 450 nm according to the reagent supplier’s guidelines, with triplicate measurements for each experimental condition.

For the colony formation assay, 1,000 control or ADAM12 siRNA-transfected glioma cells were plated in six-well plates and cultured for two weeks, with the medium replenishment every four days. Following the incubation period, the cells were fixed with methanol and subsequently stained using Crystal Violet (Sigma-Aldrich). The stained colonies were photographed and quantified using ImageJ for further analysis.

Transwell and wound healing assays

Transwell assays were performed using U87 and U251 cells, both transfected and controlled. The cell suspension containing 20,000 cells in 200 µL was seeded into Transwell chambers. The the upper chambers were filled with the FBS-free media. The lower chambers were supplemented with a culture medium containing 10% FBS. Cells were incubated for 24 h, then fixed and stained with 0.05% crystal violet for 30 min before undergoing microscopic imaging.

For the wound healing assay, evenly transfected cells were plated in six-well plates and grown to confluence. A sterile 200 µL pipette tip was employed to create a scratch in the cell monolayer, marking the starting point (0 h). Cells were washed with PBS to remove debris and then incubated in FBS-free media for 24 h.The migration of cells into the wound area was documented at 0 and 24 h, and the migration rate was calculated.

Western blot

Tissues or cells were lysed in appropriate lysis buffer, followed by electrophoresis on a 4–12% FuturePAGE™ Protein Precast Gel (ACE Biotechnology, China). Proteins were subsequently transferred to a polyvinylidene fluoride (PVDF) membrane. Membranes were incubated overnight at 4°℃ with primary antibodies:

GAPDH (60004-1-Ig, Proteintech Group) and ADAM12 (A7940, ABclonal Technology). Then these were incubated with secondary antibodies, and relative protein abundance was quantified using Image Lab analysis software.

Quantitative real-time PCR (qRT-PCR)

Total RNA was isolated from fresh tissues and cells using TRIzol Reagent (Invitrogen) following the manufacturer’s guidelines. mRNA reverse transcription and qPCR were conducted using the VoM-MLV One Step RT-qPCR Kit (SYBR) (AG11732, ACCURATE, China) in a single-tube reaction. GAPDH served as an internal control. The sequences of the primers were as follows:

ADAM12 forward: 5′-CAGGCACAAAGTGTGCAGAT-3′; reverse:5′-GCTTGTGCTTCCTCCAAAGC-3′.

GAPDH forward: 5′- GCACCGTCAAGGCTGAGAAC-3′; reverse: 5′-TGGTGAAGACGCCAGTGGA-3′.

Immunohistochemical analysis

Fresh samples were fixed in 4% paraformaldehyde and paraffin-embedded for 24 h. The paraffin blocks were sectioned into 5 µm thick slices. Sections were blocked overnight at 4°℃ and stained using an anti-ADAM12 antibody (A7940, ABclonal Technology). After washing, sections were treated with a secondary antibody, and images were captured using a microscope. Quantitative analysis was performed using ImageJ software.

Statistical analysis

A univariate Cox proportional hazards analysis identified hub genes significantly linked to patient survival (p < 0.05). Genes significantly associated with patient survival in univariate analysis were further examined using multivariate Cox regression analysis. All analyses were conducted using GraphPad Prism 8.0 (GraphPad Software, La Jolla, CA, USA). All experiments were conducted independently a minimum of three times. Data are presented as mean ± standard deviation (SD). Comparisons between groups were conducted using an independent sample t-test. A p-value below 0.05 was considered statistically significant.

Results

Upregulation of ADAM12 in multiple human cancers

We acquired expression data for ADAM12 across 33 different cancer typesand subsequently analyzed the differences in ADAM12 expression between normal and tumor tissues. The data indicated that ADAM12 was significantly upregulated in 25 tumor types, including pan-glioma (GBMLGG), breast invasive carcinoma (BRCA), lung adenocarcinoma (LUAD), esophageal carcinoma (ESCA), stomach and esophageal carcinoma (STES), kidney renal papillary cell carcinoma (KIRP), pan-kidney cohort (KIPAN), colon adenocarcinoma (COAD), colon and rectum adenocarcinoma (COADREAD), stomach adenocarcinoma (STAD), head and neck squamous cell carcinoma (HNSC), kidney renal clear cell carcinoma (KIRC), lung squamous cell carcinoma (LUSC), liver hepatocellular carcinoma (LIHC), Wilms’ tumor (WT), skin cutaneous melanoma (SKCM), bladder urothelial carcinoma (BLCA), thyroid carcinoma (THCA), rectum adenocarcinoma (READ), pancreatic adenocarcinoma (PAAD), uterine carcinosarcoma (UCS), acute myeloid leukemia (LAML), pheochromocytoma and paraganglioma (PCPG), adrenocortical carcinoma (ACC), and cholangiocarcinoma (CHOL). In contrast, ADAM12 expression was found to be downregulated in prostate adenocarcinoma (PRAD), ovarian serous cystadenocarcinoma (OV), acute lymphoblastic leukemia (ALL), and kidney chromophobe (KICH) (Fig. 1A). Figure 1B illustrates the differential expression levels of ADAM12 across various cancer cell lines, as cataloged in the Cancer Cell Line Encyclopedia (CCLE). It is clearly demonstrated that ADAM12 is significantly overexpressed in glioma cell lines (Fig. 1B). In line with these findings, data from the Human Protein Atlas (HPA) also confirmed elevated ADAM12 expression in brain and kidney cancers (Fig. 1C). The expression of ADAM12 across various glioma cell lines is depicted in Fig. 1D. Immunofluorescence imaging further demonstrated that ADAM12 predominantly localizes within the nucleoplasm and nuclear speckles. And the images were obtained from the HAP database (Fig. 1E). These findings consistently highlight the widespread upregulation of ADAM12 in cell lines derived from diverse tumor tissues, in alignment with data from the Cancer Genome Atlas database (TCGA).

Fig. 1
figure 1

Pan-cancer analysis of ADAM12 expression. (A) ADAM12 expression levels across various cancer types, including both tumor and normal tissues, based on data from TCGA and GTEx datasets. (B) ADAM12 expression in tumor cell lines from the CCLE dataset. (C) Expression of ADAM12 in diverse cancer cell lines as derived from the HPA dataset. (D) ADAM12 expression in brain cancer cell lines from the HPA dataset. (E) Cellular localization of ADAM12 as observed in the HPA database (*p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001).

Full size image

Elevated expression of ADAM12 in glioma

Further investigation of ADAM12 expression was conducted by comparing data from TCGA and the Genotype-Tissue Expression (GTEx) project via the GEPIA2 platform (Fig. 2A). The analysis revealed that ADAM12 expression levels were significantly elevated in several cancer types compared to normal tissues, notably in BRCA, GBMLGG, BLCA, ESCA, KIRC, KIRP, OV, LUAD, LUSC and ACC. Furthermore, the expression of ADAM12 was found to correlate with disease stages/histological grades in GBMLGG, BRCA, BLCA, KICH, KIRC, OV, KIRP, ACC, LIHC, LUAD, and OV, as depicted in Fig. 2B–L. The aforementioned findings further delineate a positive correlation between ADAM12 expression and the histopathological grading of gliomas (Fig. 2B), with pronounced differential expression observed among glioma molecular subtypes (Fig. 2C). Notably, the mesenchymal (MES) subtype exhibited significantly elevated ADAM12 levels compared to the proneural (PN) and classical (CL) subtypes. This observation suggests the potential of ADAM12 as a prognostic biomarker for glioma pathogenesis and subtype classification.

Fig. 2
figure 2

The expression levels of ADAM12 in various tumors. (A) ADAM12 gene expression in Breast invasive carcinoma (BRCA), Glioblastoma-Lower Grade Glioma(GBMLGG), Bladder Cancer (BLCA), Esophageal Carcinoma (ESCA), Kidney Chromophobe (KICH),Kidney Clear Cell Carcinoma (KIRC), Kidney Papillary Cell Carcinoma (KIRP), Ovarian Cancer (OV), Lung Adenocarcinoma (LUAD), Lung Squamous Cell Carcinoma (LUSC), Adrenocortical Cancer (ACC) using GEPIA2. (B–L) The correlation between ADAM12 expression and different cancer stages/histological grades of tumors. (B) GBMLGG, (C) GBM, (D) BRCA, (E) KIRC, (F) KIRP, (G) KICH, (H) BLCA, (I) LIAD, (J) LIHC, (K) ACC, (L) OV.

Full size image

Evaluating ADAM12 for diagnostic and prognostic purposes in pan-cancer

Our research focused on evaluating the prognostic significance of ADAM12 expression in cancer patients. GEPIA2-based analyses revealed the survival map of gene ADAM12 has significant results in ACC, BLCA, GBMLGG, KIRP, LIHC, MESO, STAD and UVM (Fig. 3A). As shown in Fig. 3B, the correlations between ADAM12 expression and OS and DSS across 38 tumor types are displayed in forest plots. The data indicated that elevated ADAM12 expression was significantly linked to worse OS and DSS in GBMLGG. Specifically, patients with high ADAM12 expression had poorer OS compared to those with low expression, as evidenced by the TCGA-GBMLGG dataset (Fig. 3B). The ROC curve analysis suggested that ADAM12 could be a potential predictor of survival in GBMLGG, with areas under the curve of 0.852, 0.882, and 0.796 for 1-, 3-, and 5-year survival, respectively (Fig. 3C).

Fig. 3
figure 3

Survival analysis of ADAM12 across pan-cancer from TCGA database. (A) Survival map depicting ADAM12 gene expression across pan-cancer types. (B) Forest plots illustrating the correlations between OS/DSS and ADAM12 expression across 38 cancer types. (C) ADAM12 expression, survival time, and status in the TCGA-GBMLGG dataset. Kaplan–Meier survival estimates for patients at all grades, along with ROC curves and AUC values for ADAM12 at 1-, 3-, and 5- years in the TCGA-GBMLGG dataset. (D) Kaplan–Meier analyses of the association between ADAM12 expression and OS in GBMLGG, Pan-kidney cohort KIPAN (KICH + KIRC + KIRP), LIHC, BRCA, KIRC, UVM, OV, MESO, KIRP, ACC and LUAD. (E) P-values, risk coefficients (HR), and confidence intervals were analyzed through univariate and multivariate Cox regression. A nomogram predicting 1-, 2- and 3-years overall survival of glioma patients. Calibration curves for the overall survival nomogram model in the discovery group. The dashed diagonal line represents the ideal model, while the blue, red, and orange lines represent the observed nomogram at 1-, 2-, and 3- years, respectively.

Full size image

Additionally, ADAM12 effectively discriminated between high-risk and low-risk groups when the optimal cut-off value was applied (Fig. 3D and Supplementary Fig.S1A). The results of the Kaplan–Meier analysis showed a negative correlation between the expression of ADAM12 and the survival time in MESO, GBMLGG, UVM, KIPAN, LIHC, KIRC, KIRP, OV, LUAD and ACC. These findings suggest that higher ADAM12 expression is associated with a worse prognosis in cancer progression. To assess ADAM12’s prognostic value in GBMLGG, we applied univariate and multivariate Cox regression models. The results identified ADAM12 as an independent prognostic factor. Furthermore, in order to assist with patient prognosis evaluation in clinical practice, we designed a nomogram based on ADAM12 expression and tumor grade (Fig. 3E). The calibration curves demonstrated the accuracy of this model in predicting glioma patient prognosis at 1-, 2-, and 3- year (Fig. 3E).

Association of ADAM12 expression levels with the clinicopathologic features of glioma

We analyzed the relationships between ADAM12 expression and factors such as WHO grade, pathological stage, IDH1 status, and clinical outcomes across three different datasets. Figure 4A illustrates a positive correlation between the expression of ADAM12 mRNA and tumor grade in all datasets. Elevated ADAM12 expression was linked to increased malignancy in glioma (Fig. 4B). Additionally, the IDH-wildtype group exhibited significantly higher ADAM12 expression levels compared to the IDH-mutant subgroup (Fig. 4C). Similarly, the 1p/19q non-codeletion group showed markedly higher ADAM12 expression than the 1p/19q codeletion group (Fig. 4D), determined by the Wilcoxon rank-sum test. These results suggest that ADAM12 is more highly expressed in the 1p19q-non-codeletion and IDH-wildtype groups.

Fig. 4
figure 4

Association between ADAM12 mRNA expression and clinicopathologic/prognostic data in TCGA-GBMLGG, CGGA, and GSE16011. The clinicopathologic features analyzed include WHO Grades (A), Histology (B), IHD_status (C), 1p/19q_status (D), and overall survival analysis of patients with high-and low-expression levels of ADAM12 and time-dependent ROC curves for OS (E). * p < 0.05, ** p < 0.01, *** p < 0.001.

Full size image

With the three datasets, we performed an integrated survival analysis to investigate the correlation between ADAM12 expression and glioma patient survival outcomes (Fig. 4E). In the TCGA Dataset, patients with high ADAM12 expression levels in primary glioma exhibited a significantly worse prognosis (p < 0.001). Time-dependent receiver operating characteristic (ROC) analysis revealed that the prognostic accuracy for GBMLGG at 1-, 3-, and 5- years, with AUC values of 0.856, 0.885, and 0.789, respectively. Similarly, in the CGGA_mRNAseq_325 dataset, the high-expression group had a significantly worse prognosis (p < 0.001). The high-risk group had poor overall survival (OS), with 1-, 3-, and 5-year AUC values of 0.717, 0.804, and 0.819, respectively. Additionally, in the GSE16011 dataset, the high-expression group also exhibited a significantly unfavorable prognosis in primary glioma (p < 0.001), with AUC values of 0.717, 0.804, and 0.819 for 1-, 3-, and 5-year survival, respectively.

Correlation between ADAM12 expression and MATH, TMB, MSI in various cancer types

We examined the relationship between ADAM12 expression and MATH status across 36 cancer types. Statistically significant negative correlations were found between ADAM12 expression and MATH in GBMLGG, GBM, and LGG (Fig. 5A,D). Additionally, ADAM12 expression was negatively correlated with MSI in 17 cancer types, with the most significant correlation observed in GBMLGG (Fig. 5B,E). Furthermore, ADAM12 expression was positively correlated with TMB in 8 cancer types, including GBMLGG, ACC, READ, LGG, CAODREAD, LUAD, COAD, and UVM (Fig. 5C,F). These findings underscore the significant association of ADAM12 expression with MATH, TMB, and MSI, particularly in glioma.

Fig. 5
figure 5

Correlation between ADAM12 expression and MATH, TMB, and MSI in various cancer types. Spearman correlation analysis of ADAM12 expression with MATH (A), MSI (B), and TMB (C) in the TCGA datasets. The horizontal axis represents the correlation coefficient, while the vertical axis represents different tumor types. The size of the dots reflects the magnitude of the correlation coefficient, and colors indicate the significance of the p-value. The deeper the blue, the smaller the p-value. (D) Correlation analysis between ADAM12 gene expression and MATH in TCGA-GBMLGG/LGG/GBM. (E) Correlation analysis between ADAM12 gene expression and MSI in TCGA-GBMLGG/LGG/GBM. (F) Correlation analysis between ADAM12 gene expression and TMB in TCGA-GBMLGG/LGG/GBM. The abscissa represents gene expression distribution, and the ordinate represents MATH/TMB/MSI score distribution. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Full size image

Immunological insights into ADAM12 in the glioma tumor microenvironment

To clarify the relationship between ADAM12 and immune cells in the tumor microenvironment in pan-cancer, we applied the quanTIseq algorithm to quantify immune cell populations. The results indicated a strong association between ADAM12 expression and various immune cell types in pan-cancer. Notably, higher ADAM12 expression was significantly correlated with increased levels of T cell regulators, CD4 + T cells, CD8 + T cells, neutrophils, B cells, and macrophages (Fig. 6A). And in our separate analysis of immune cell infiltration in GBMLGG, we noticed a positive correlation between the expression of ADAM12 and the infiltration of macrophages, myeloid dendritic cells, CD4 + T cells, and CD8 + T cells in glioma (Fig. 6B). As shown in Fig. 6C a wide range of immune cell populations, including macrophage M1 and M2 cells, CD4 + and CD8 + T cells, and myeloid dendritic cells, displayed a positive correlation with ADAM12 expression. In contrast, 18 immune cell types, such as B cells, monocytes, and natural killer cells, showed an inverse correlation with ADAM12 expression. Upon stratification of GBMLGG into high and low ADAM12 expression groups, a Tumor Immune Estimation Resource (TIMER) analysis indicated that the high expression cohort of ADAM12 had statistically significant increases in the levels of macrophage M1, macrophage M2, CD4 + T cells, CD8 + T cells, and myeloid dendritic cells, as depicted in Fig. 6D.

Fig. 6
figure 6

Pan-cancer analysis of the relationship between ADAM12 expression and immune cell infiltration. (A) Immune cell infiltration across pan-cancer types analyzed using QUANTISEQ algorithms. *p < 0.05, **p < 0.01, ***p < 0.001. (B) The heat map in the schematic represents the correlation analysis of the immune score in TCGA-GBMLGG. (C) Spearman correlation analysis of ADAM12 expression with immune infltrating cells in glioma. (D) Distribution of QUANTISEQ immune scores in high and low ADAM12 expression groups of glioma. The horizontal axis represents immune cell types, while the vertical axis indicates the distribution of immune scores across different groups. G1: high group of ADAM12 in glioma. G2: low group of ADAM12 in glioma.

Full size image

To investigate the impact of the risk score on immune infiltration and response to immunotherapy, we computed the ESTIMATE score, ImmuneScore, and StromalScore, which reflect immune and stromal cell infiltration levels. Our findings revealed a significant in GBM, GBMLGG and LGG (Supplementary Fig.S1B). Additionally, immune checkpoint analysis demonstrated a positive correlation between ADAM12 levels and immune checkpoint levels in several cancers, including GBMLGG, KIPAN, SARC, KIRP, ACC, KIRC, LUSC and BLAC, suggesting that ADAM12 may serve as a potential marker to guide immunotherapy in these cancers (Supplementary Fig.S2A). We also investigated the relationship between ADAM12 expression and chemokines, as well as chemokine receptors, using the TISIDB database. Our results revealed a positive correlation between ADAM12 and most chemokines (Supplementary Fig.S2B,C). In the case of GBMLGG, the top 6 Chemokines were GBM-CCL7, GBM-CCL13, GBM-CXCL5, GBM-CXCL3, GBM-CXCL6, and HLA-DPA1. The correlations between ADAM12 expression and 18 chemokine receptors are shown in Supplementary Fig.S2D,E. These results suggest a strong positive correlation between ADAM12 expression and several chemokine receptors in GBMLGG; the top 6 chemokine receptors were GBM-CCR1, GBM-CCR2, GBM-CCR5, LGG-CCR2, LGG-CCR4, and LGG-CXCR6.

Analysis of the potential function of ADAM12 in single-cell datasets

Further analysis of ADAM12 expression in relation to seven canonical cancer-related functional states, using single-cell sequencing data from CancerSEA, demonstrated a positive correlation between ADAM12 and key processes such as angiogenesis, differentiation, and metastasis across various tumor types (Fig. 7A). Notably, in gliomas, ADAM12 expression was strongly associated with differentiation, angiogenesis, proliferation, and inflammation (Fig. 7B). Additional single-cell sequencing analysis of ADAM12 expression across various normal tissues, including the Brain, Kidney, Lung, and Liver, revealed that ADAM12 expression was predominantly confined to oligodendrocyte progenitor cells within brain tissue, with minimal expression in other normal tissues (Fig. 7C,D and Supplementary Fig.S3).

Fig. 7
figure 7

Single-cell RNA sequencing analysis of the functional state of ADAM12 in Pan-cancer and its expression in normal tissues. (A) Single-cell analysis illustrating the functional state of ADAM12 in Pan-cancer. (B) CancerSEA analysis demonstrated that ADAM12 positively correlated with differentiation, angiogenesis, and proliferation in glioma. Single-cell sequencing analysis showed the expression of ADAM12 in normal tissues from Tabula Muris DataBase, including the Brain (C) and Kidney (D).

Full size image

Further examination of ADAM12 expression across tumor and stromal cells across various cancers, including GBM, LIHC, OV and UVM, revealed a pronounced co-expression in both cancerous and stromal cells, particularly in malignant cells, mast cells, T cells, B cells, fibroblasts, and endothelial cells (Fig. 8). These findings suggest that ADAM12 expression is preferentially increased in malignant glioma cells, potentially contributing to tumor proliferation.

Fig. 8
figure 8

Single-cell RNA sequencing analysis of the expression of ADAM12 in tumors. (A) ADAM12 was mainly expressed in stromal cells and glioma cells in the UCSC Cell Browser. (B) ADAM12 was mainly expressed in Malignant cells in Glioma_GSE148842 in TISCH DataBase. (C)ADAM12 was mainly expressed in Mast cells in LIHC_GSE140228 in the TISCH Database. (D)ADAM12 was mainly expressed in Fibroblast cells in OV_GSE147082 in the TISCH Database. (E) ADAM12 was mainly expressed in Malignant cells and Endothelial cells in UVM_GSE138433 in the TISCH Database.

Full size image

Gene set enrichment analysis for ADAM12 in glioma

To explore the molecular mechanisms underlying ADAM12’s role in glioma, we categorized glioma samples into ADAM12-high and ADAM12-low groups. The differentially expressed genes (DEGs) between these groups are presented in Fig. 9A and Fig. 9B. In order to better grasp the functional consequences of these DEGs, we executed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses. GO analysis indicated that, in terms of biological processes (BP), ADAM12-associated genes are primarily involved in glial cell differentiation, cell growth, the G2/M phase transition of the cell cycle, and oligodendrocyte differentiation, among others. In terms of molecular function (MF), these genes are implicated in activities such as growth factor binding, immune receptor binding, collagen binding, and contributing to the structural composition of the extracellular matrix (ECM), among others (Fig. 9C). KEGG pathway enrichment analysis further highlighted the association between ADAM12 and several pathways in glioma, including ECM-receptor interactions (hsa04512), the cell cycle (hsa04110), the PI3K-Akt signaling pathway (hsa04151), and pathways involved in small cell lung cancer (hsa05222) and bladder cancer (hsa05219), among others (Fig. 9D). Additionally, Gene Set Enrichment Analysis (GSEA) of the DEGs revealed that the pathways related to the cell cycle, ECM-receptor interactions, the JAK-STAT signaling pathway, cancer-related pathways, the P53 signaling pathway, DNA replication, primary immunodeficiency, and small cell lung cancer were significantly enriched (Fig. 9E). We also examined the protein–protein interaction (PPI) network of ADAM12, identifying several key interactions with EGF, EGFR, SH3D19, ADAM9, and others (Supplementary Fig.S4B). These findings highlight ADAM12’s essential role in regulating critical biological processes, emphasizing its central involvement in cellular activities crucial for cancer progression.

Fig. 9
figure 9

Enrichment function for ADAM12 in glioma. (A) Heatmaps displaying the expression of the top 50 genes positively and negatively correlated with ADAM12. (B) Volcano plot of DEGs between high and low ADAM12 expressions. (C) GO analysis of the DEGs based on the ADAM12 expression in glioma. (D) KEGG analysis of the DEGs based on the ADAM12 expression in glioma. (E) The GSEA analysis results of the DEGs based on the ADAM12 expression in glioma.

Full size image

ADAM12 promotes malignant progression of glioma cells

To validate these observations, we collected six pairs of glioma tissues and adjacent normal tissues. Western blot analyses demonstrated a significant increase in ADAM12 expression in tumor tissues relative to normal tissues (Fig. 10A). Immunohistochemistry further revealed a positive correlation between ADAM12 expression levels and the pathological grades of glioma tissues (Supplementary Fig.S4A). To investigate the oncogenic potential of ADAM12 in glioma more deeply, we established an ADAM12 knockdown model using small interfering RNA (siRNA) in U87 and U251 glioma cell lines. Both quantitative PCR (qPCR) and Western blot assays confirmed effective knockdown of ADAM12 in these cell lines (Fig. 10B and Fig. 10C). Cell viability assays (CCK-8) demonstrated that ADAM12 knockdown significantly decreased the viability of glioma cells (Fig. 10D). Additionally, wound healing assays revealed that ADAM12 silencing markedly impaired the migratory capacity of glioma cells (Fig. 10E,F). Colony formation assays further indicated that ADAM12 knockdown substantially inhibited the proliferative capacity of glioma cells (Fig. 10G). Additionally, transwell invasion assays showed a significant reduction in the invasive ability of glioma cells upon ADAM12 silencing (Fig. 10H,I). Overall, these results suggest that ADAM12 plays a critical role in promoting the malignant biological behaviors of glioma cells.

Fig. 10
figure 10

ADAM12 promotes the malignant progression of glioma cells. (A) Protein levels of ADAM12 in six pairs of glioma tissues (T) and adjacent non-tumor tissues (N).Full-length original Western blots are presented in Supplementary Fig S5A. (B, C) Validation of siRNA interference efficiency in knockdown of ADAM12 expression in U87 and U251 cells by RT-qPCR and Weston blotting.Full-length original Western blots are presented in Supplementary Fig S5B. (D) CCK8 assay to detect the proliferative capacity of transfected U87 and U251. (E, F) Wound healing assay for negative control and ADAM12 knockout groups. (G) Colony formation assay for negative control and ADAM12 knockout groups. (H, I) Transwell assay images of migration and invasion in the negative control and ADAM12 knockout groups. *p < 0.05; **p < 0.01; ***p < 0.001.

Full size image

Discussion

Tumors, as malignant diseases driven by uncontrolled cell proliferation, share some common underlying pathogenic mechanisms, despite their varying clinical manifestations across different cancer types21,22. In recent years, large-scale pan-cancer studies have revealed crucial information about the mechanisms that propel cancer progression23,24,25. ADAM12 has been reported to contribute to cancer progression in several tumor types, including breast16,26, liver17, rectal18, and gastric27 cancers. However, the role of ADAM12 across multiple cancer types, particularly its similarities and differences in function, has yet to be fully elucidated. In this study, we integrated pan-cancer data from TCGA to generate a comprehensive map of ADAM12 expression patterns and their potential functions across various cancers. Additionally, our experimental work further confirmed the pro-cancer role of ADAM12 in the malignant progression of glioma, contributing to a better understanding of its significance in tumor biology.

The aberrant expression pattern of ADAM12 has been observed in various tumors18,27,28. To further investigate the expression trends of ADAM12 across different tumors, we analyzed its expression in various tumor tissues and cell lines. Our results demonstrated that ADAM12 expression is dysregulated across multiple cancers, suggesting its potential role in tumorigenesis and progression. Further analysis identified a correlation between ADAM12 expression and the grading of various cancers, including LGG/GBM, BRCA, KIRC and LUAD. Gliomas, the most common primary malignant brain tumors in adults, are classified into grades I-IV according to WHO grade. Patients with glioblastoma (WHO IV) face an very poor prognosis, with a median survival of about 15 months29,30,31,32. Notably, we found that brain tumor cell lines exhibited the highest ADAM12 expression among all tumor cell lines, and ADAM12 expression increased with glioma grade. In GBM, ADAM12 is highly correlated with the mesenchymal subtype. These findings suggest that ADAM12 may be associated with tumor malignancy, particularly in gliomas.

The advent of tumor immunotherapy has significantly improved the prognosis for many patients. However, the response to treatment varies considerably among patients, even those with the same tumor type2. This variability limits the broad applicability of immunotherapy. Identifying biomarkers that predict a patient’s response to specific treatments and overall prognosis has become a major focus in oncology research33,34. Luo et al. showed that LTBR, FXYD6 and MAL2 DNA methylation were the biomarkers for the diagnosis and prognosis of glioma35,36,37. In this study, elevated ADAM12 expression correlated with poor prognosis across multiple cancers, such as GBM/LGG, KIPAN, LIHC, MESO, and others. In the TCGA pan-glioma cohort, both univariate and multivariate Cox analyses confirmed that ADAM12 is an independent prognostic factor. By integrating other clinicopathologic features, we constructed a nomogram to predict the prognosis of LGG patients. Calibration curves confirmed the accuracy of this nomogram. These results demonstrate that ADAM12 is a robust prognostic marker, particularly in glioma patients. Additionally, we examined the relationship between ADAM12 and other features of glioma across multiple datasets. Our findings showed that ADAM12 is associated with more malignant features, such as the pathological type of GBM, IDH wild-type status, and 1p19q non-co-deletion, indicating that ADAM12 characterizes more malignant glioma subtypes. Previous studies have reported that MATH, MSI, and TMB can serve as indicators of patient prognosis and response to immunotherapy38,39,40,41. Our study revealed a high correlation between ADAM12 expression and MATH, MSI, and TMB, suggesting the potential of ADAM12 in predicting patients’ responses to immunotherapy. Additionally, our findings indicate that ADAM12 is linked to the infiltration of immunosuppressive cells, including as regulatory T cells and M2-type macrophages, in various cancers. ADAM12 was also linked to several of immunosuppressive molecules and chemokines. The results indicate that ADAM12 may play a role in shaping the tumor immunosuppressive microenvironment.

Recent advancements in single-cell sequencing technologies have greatly expanded our understanding of the tumor microenvironment42,43,44. Driven by these innovations, we examined the role of ADAM12 within the tumor microenvironment using single-cell sequencing data. Our results revealed that ADAM12 was minimally detected in the cellular composition of normal brain, kidney, lung, and liver tissues but showed elevated expression in the microenvironments of glioma, LIHC, OC, and UVM tumor tissues. Gene enrichment analysis further indicated that ADAM12 is linked to the activation of several pro-oncogenic pathways in glioma, indicating its potential role in promoting tumorigenesis. To confirm the elevated expression of ADAM12 in tumor tissues, we collected glioma and adjacent non-tumor tissue specimens. Western blot analysis confirmed a high level of ADAM12 expression in glioma tissues. Additionally, ADAM12 knockdown cell lines were constructed, and the results demonstrated that ADAM12 enhances the proliferation, migration, and invasion of glioma cells.These findings highlight the significant involvement of ADAM12 in the malignant progression of glioma, positioning it as a potential therapeutic target.

In conclusion, our study provides an extensive overview of ADAM12 across various cancers, emphasizing its expression patterns and prognostic value in a wide range of tumors. Our results suggest that ADAM12 holds potential as a biomarker for assessing patient prognosis and tumor malignancy, especially in gliomas. However, several limitations remain. Further investigation is needed to explore the underlying mechanisms through which ADAM12 facilitates malignant tumor progression in glioma. Additionally, its role in the immune microenvironment warrants validation through additional experimental studies. We foresee that this research will provide the groundwork for future studies aimed at unraveling the specific mechanisms by which ADAM12 influences tumor biology.