Introduction

Cancer is the second most common cause of death and imposes a significant economic burden on society globally1. Targeted therapies and immunotherapies have produced remarkable clinical responses in many solid tumors and improved patients’ prognosis, which makes them among the mainstream and most effective strategies for cancer treatment2. However, due to the inadequacy of targeted therapies and the heterogeneity of tumours, the cure rate and prognosis of patients are still unsatisfactory3,4. Furthermore, the therapeutic efficacy of targeted drugs is limited by the emergence of drug resistance5. Therefore, new and unique targets for tumour treatment must be explored.

Researchers have often observed that certain key genes play important roles in one or several types of cancer. However, the potential value of these genes, especially in other types of tumours, has not been fully explored. With the continuous development and improvement of public databases such as The Cancer Genome Atlas (TCGA), a pan-cancer analysis can characterize genes of interest in multiple cancers, which allows a better evaluation of the role of genes in cancer and key molecular mechanisms and facilitates further discovery of new targets for targeted or immunotherapies.

Rho guanosine triphosphatases (Rho GTPases), which act as regulators of cell proliferation, cytoskeletal organization, and cell movement, are crucial for the spread of cancer6. The activation of Rho GTPases is controlled by two families of proteins: guanine nucleotide exchange factors (RhoGEFs) and GTPase-activating proteins (RhoGAPs)7. RhoGEFs can stimulate the exchange of GDP for GTP and thereby restore Rho GTPases to the active GTP-binding state8. Neuroepithelial cell transforming gene 1 (NET1) is an important member of the RhoGEF family and a key protein involved in the activation of Rho GTPases9. Moreover, research has shown that NET1 can regulate the malignant biological functions of tumour cells, such as growth, invasion, and metastasis, and is closely related to the progression of pancreatic cancer, gastric cancer and liver cancer10,11,12,13. Despite these findings, the comprehensive role and mechanistic function of NET1 in other types of cancer remain largely unexplored. A deeper understanding of the role of NET1 may provide new insights into the molecular mechanisms of cancer progression and metastasis. Additionally, it may help identify new therapeutic targets for a range of cancers. This study aims to fill this knowledge gap and provide a more comprehensive understanding of the role of NET1 in cancer biology.

Therefore, this study systematically analysed the expression level of NET1 in various cancer and adjacent tissues, as well as its relationship with the clinical prognosis, and explored the biological processes in which NET1 may be involved. Further in vivo and in vitro experiments, along with real-world data, revealed that NET1 may contribute to the malignant progression of triple-negative breast cancer (TNBC) through the AKT signalling pathway and mediate a poor prognosis.

Materials and methods

Study overall design

The overall design of this study can be seen in Fig. 1. The abbreviations for various types of cancer are listed in Supplementary Table 1.

Fig. 1
figure 1

Overall flowchart of the study design.

Data download and differential expression analysis

TCGA and Genotype-Tissue Expression (GTEx) data, DNA methylation data, RNA stemness scores (RNAs) based on mRNA expression, and DNA stemness scores (DNAss) based on DNA methylation patterns were obtained from the UCSC database (https://xenabrowser.net/). In addition, the gene expression levels extracted from each sample were standardized. The Sangerbox online tool was used to visualize differential gene expression (http://www.sangerbox.com/home.html)14. mRNA expression data for breast cancer cell lines were obtained from the Human Protein Atlas platform (https://www.proteinatlas.org/).

Prognostic and diagnostic analyses

According to TCGA data, the patients were divided into two groups based on the threshold equal to 50% of the NET1 expression level. The “survival” package of R software was used to plot the survival curves, and the log-rank test was used to evaluate whether the difference in overall survival (OS) between the two groups was significant. Furthermore, we used Kaplan‒Meier (KM) plotter (https://kmplot.com/analysis/) to further analyse possible categories of cancer with differences in survival, serving as a supplement to TCGA survival prognosis data. KM plotter is capable of assessing the correlation between the expression of all genes (mRNAs, miRNAs, proteins, and DNA) and survival in 35,000 + samples from 21 tumour types. We chose mRNA expression levels and divided them into high and low groups based on the median expression level of NET1. The log-rank test was used to calculate the differences in survival between groups.

The coxph function of the “survival” package in R software was then used to establish a univariate Cox proportional hazards regression model for the analysis of the relationships of NET1 expression with overall survival (OS), disease-specific survival (DSS), the disease-free interval (DFI), and the progression-free interval (PFI) of patients with each cancer type in TCGA.

An unpaired Student’s t test was used for pairwise significance analyses to evaluate the differences in expression in tumour samples from patients classified into groups based on different clinical features, and analysis of variance was used to assess the differences among multiple groups of samples.

The “pROC” R package in R software was used to calculate the sensitivity and specificity of the tumour diagnosis based on NET1 gene expression. The diagnostic value was quantified based on the area under the curve (AUC).

Analysis of gene alterations

By accessing the cBioPortal database (https://www.cbioportal.org/), we explored the types and frequencies of NET1 gene alterations in TCGA pan-cancer dataset. The GISTIC2.0 method was used to estimate the copy number variations (CNVs) in genes in TCGA database15. Using the GSCA database (http://bioinfo.life.hust.edu.cn/GSCA/#/), the correlation between NET1 expression and CNVs was analysed by performing Spearman’s correlation analysis. The UALCAN database (https://ualcan.path.uab.edu/) was used to explore the differences in the DNA methylation levels of the NET1 promoter between cancer and adjacent tissues16.

Analysis of tumour characteristics

The tumour mutation burden (TMB) and microsatellite instability (MSI) data from TCGA database were obtained from the Genomic Data Commons (GDC) (https://portal.gdc.cancer.gov/)17,18. The “ESTIMATE” package in R software was used to calculate the stromal score, immune score, and ESTIMATE score. The relationships of NET1 expression with the TMB, MSI, DNAss, RNAss, stromal score, immune score and ESTIMATE score were analysed using Spearman’s correlation analysis. The “ggplot2” package in R software was used to visualize the research results.

Gene correlation analysis

Spearman’s correlation analysis was used to evaluate the correlations between NET1 expression and RNA modifications, including genes related to N1-methyladenosine (m1A), 5-methylcytosine (m5C), and N6-methyladenosine (m6A). The correlations of NET1 expression with both oncogenes and immune-related genes (chemokines, receptors, major histocompatibility complex (MHC), immunoinhibitors, and immunostimulators) were also analysed. The “ggplot2” package in R software was used to visualize the research results.

Immune cell infiltration and single-cell distribution characteristics

The TIMER2.0 database comprehensively characterizes molecular and tumour immune interactions19. Using the TIMER2.0 database, we explored the correlation between NET1 expression and the level of immune cell infiltration in the tumor microenvironment of 33 types of cancer. The Tumour Immune Single-cell Hub (TISCH; http://tisch.comp-genomics.org/) is an single-cell RNA sequencing database specifically developed to study the single-cell landscape of the tumour microenvironment (TME)20. UMAP plots were generated to visualize the cell types and NET1 expression levels. We also obtained single-cell data from CancerSEA (http://biocc.hrbmu.edu.cn/CancerSEA/) to study the relationships of NET1 expression with 14 different cancer-related biological functions21.

Drug sensitivity analysis

Based on the Genomics Committee on Drug Sensitivity Scale (GDSC) and the Cancer Therapeutics Response Portal (CTRP) databases, the GSCA online website (http://bioinfo.life.hust.edu.cn/GSCA/#/drug) was used to explore potential drugs whose sensitivity may be correlated with NET1 gene expression. Using the “ggplot2” package in R software, the research results are presented in the form of a bubble chart showing the relationships between the half-maximal inhibitory concentrations of the drugs and NET1 expression.

Functional enrichment analysis

The signalling pathways through which NET1 might be involved in regulating cancer were explored through gene set enrichment analysis (GSEA) to explore the mechanism underlying the effect of NET1 expression on the prognosis of cancer patients22. Based on TCGA expression data, groups with high and low NET1 expression levels were established using a 30% threshold; the “limma” R package in R software was used to identify differentially expressed genes (DEGs). Genes with an adjusted P value < 0.05 were considered DEGs. The h.all.v7.2.symbols.gmt gene set was selected as a reference, and the normalized enrichment score (NES) and false discovery rate (FDR) were determined.

Furthermore, the DEGs obtained from second-generation sequencing before and after NET1 overexpression in MDA-MB-231 cells were evaluated. Using the DAVID online website (https://david.ncifcrf.gov/), gene function annotations were determined using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses. The results were visualized using the “ggplot2” package in R software.

GEO data analysis

NET1 expression data were retrieved from the Gene Expression Omnibus (GEO) dataset. After log2 transformation of the NET1 expression data, IBM SPSS Statistics V26.0 was used to calculate pattern of expression, including the number, mean, and standard deviation of each experimental and control group. The “survival” and “survminer” packages in R software were used to analyse the prognostic value of NET1 in terms of OS, distant metastasis-free survival (DMFS) and recurrence-free survival (RFS). A meta-analysis was performed to merge the same results from different datasets. The mean difference (MD) and 95% confidence interval (95% CI) were used to represent NET1 expression, and the hazard ratio (HR) and 95% CI were used to indicate OS, DMFS, and RFS. If the heterogeneity was low (I2 < 50%, P < 0.10), a fixed-effects model was used; otherwise, a random-effects model was used. The meta-analysis was implemented using the “meta” package in R software, and P < 0.05 was considered to indicate a significant difference.

Acquisition and cultivation of cell lines

The MDA-MB-231, SUM-159PT, and BT-549 cell lines were obtained from Procell Life Science & Technology Co., Ltd. (China) and cultured according to the manufacturer’s instructions.

Cell proliferation assay

MDA-MB-231 (1,000 cells/well), SUM-159PT (1,000 cells/well), and BT-549 (1,500 cells/well) cells with and without NET1 overexpression were seeded in 6-well plates. After 10 days of culture, the cells were fixed with 4% paraformaldehyde for 15 min and stained with 0.1% crystal violet for 30 min.

Cell proliferation levels were quantified using the CCK-8 method. Control and experimental cells were seeded into 96-well plates at a density of 1,000 cells (MDA-MB-231 and SUM-159PT) or 1,500 cells (BT-549) per well. After an incubation for 1, 2, 3, 4, or 5 days, cell viability was evaluated using the CCK-8 method (NCM, Cat. No. C6005, China), and the optical density (OD) was measured at 450 nm with a microplate reader (Thermo Scientific, USA). The results were visualized using GraphPad Prism 10.0.

Cell migration assay

Control and experimental cells were seeded into 6-well plates. When the cell confluence reached approximately 90%, a scratch test was performed, and images were captured at 0 h, 24 h, 36 h, 48 h, and 72 h after scratching. Before each imaging session, the cells were washed three times with PBS. The scratch area was calculated with ImageJ software. For each group, at least three fields were evaluated. The scratch area healing rate was calculated as follows: (scratch area at 0 h—scratch area at each time point)/scratch area at 0 h × 100%.

Cell apoptosis and cell cycle detection

The cells were digested with trypsin and collected, and the cell density was then adjusted to 1 × 106 cells/0.1 ml. Subsequently, cell apoptosis was detected by double staining according to the instructions of the Nest Kit (Annexin V-FITC/PI; Cat No. 210838, China), and the cell apoptosis rate was detected with a CYTECTM NL-3000 spectral flow cytometer (USA). Additionally, 1 × 106 cells/group were collected and fixed overnight with 75% ethanol, and the cell cycle distribution was then detected with a Solarbio DNA content detection kit (Cat No. CA1510, China). A CYTECTM NL-3000 spectral flow cytometer was used to detect the cell cycle distribution.

Immunohistochemical staining and semiquantitative analysis

Paraffin-embedded sections were dewaxed with xylene and a gradient of alcohol solutions. After the inactivation of endogenous peroxidases with 0.3% hydrogen peroxide and blocking with bovine serum albumin, the sections were incubated overnight with a NET1 antibody at 4 °C. After washing, the sections were incubated with a secondary antibody using a rabbit two-step detection kit (ZSGB-BIO, Cat No. PV-9001, China), and colour development was performed with DAB. The sections were subsequently washed, counterstained with haematoxylin, mounted, and photographed under a microscope.

The immunohistochemical results were independently evaluated by two pathologists. The final score was determined by the product of the percentage of positive cells and the intensity of the positively stained cells23. The percentage of positive cells was scored according to the infiltration ratio as follows: 1 point (1–10%), 2 points (10–40%), 3 points (40–75%), and 4 points (75–100%). The staining intensity was scored as follows: 0 points (negative), 1 point (weak), 2 points (moderate), and 3 points (strong). A score of 4 or less was considered to indicate low expression, whereas a score greater than 4 was considered to indicate high expression.

Reverse transcription–quantitative PCR (RT‒qPCR)

RNA was extracted from cells using the Steadypure Quick RNA Extraction Kit (AG, Cat No. AG21023, China) and then reverse transcribed into cDNA using the HiScript 1st-Strand cDNA Synthesis Kit (+ gDNA wiper) (Vazyme Company, China). RT‒qPCR was performed in a 20 µl reaction mixture using 2X Universal SYBR Green (ABclonal, Cat No. RM21203, China) and a real-time PCR detection system (TianLong, Model: Gentier 96R, China). The comparative ΔΔCt method was used to analyse the results; RQ = 2-ΔΔCt was calculated using the ΔΔCt values. The data are presented as the fold differences in gene expression relative to that of the reference sample (normalized to the housekeeping gene GAPDH), as shown in each figure legend. The sequences of the forwards and reverse primers for NET1 (transcript variant 1, the longest transcript) were 5’-GTCGCCTAGTCAAATACCCT-3' and 5’-ATGTAATACTGGCACTCGGAT-3', respectively. Similarly, the sequences of the forwards and reverse primers for GAPDH were 5’-TGACCTGCCGTCTAGAAAAACCT-3’ and 5’-GCTGTTGAAGTCAGAGGAGACCA-3’, respectively.

Western blot (WB) analysis

The cells were collected and lysed on ice in RIPA buffer (NCM, Cat No. Western blot3100, China). Protein concentrations were quantified using a BCA protein assay kit (NCM, Cat No. Western blot W6501, China). Forty micrograms of protein were separated via electrophoresis on a 10% SDS‒polyacrylamide gel and subsequently transferred to a polyvinylidene fluoride membrane. Nonspecific staining was blocked at room temperature for 1 h in 1 × Tween Tris-buffered saline containing 5% nonfat milk. The membrane was then incubated overnight at 4 °C with the following specific primary antibodies: NET1 (Proteintech, Cat No. 28180–1-AP, China), GAPDH (Proteintech, Cat No. HRP-60004, China), AKT (Proteintech, Cat No. 60203–2-Ig, China), phospho-Akt (Ser473) (Cell Signaling Technology, Cat No. 4060S, USA) and DYKDDDDK-tagged rabbit polyclonal antibodies (Proteintech, Cat No. 20543–1-AP, China). The membrane was incubated at room temperature for 1 h with a goat anti-rabbit IgG secondary antibody (Abiowell, Cat No. Aws00026, China) or a goat anti-mouse IgG secondary antibody (Abiowell, Cat No. AWS0001b, China). The bands were visualized in a dark room using SuperSensitive ECL Luminescent Fluid (Meilunbio, Cat No. MA0186-2, China), and the band density was analysed using SageCapture™ image analysis software. The semiquantitative analysis of the WB images was performed using ImageJ, and the results were visualized and statistically analysed using GraphPad Prism 10.0.

Endpoint of the follow-up study

Disease-free survival (DFS) was defined as the time from the patient’s diagnosis to the time of tumour recurrence, metastasis, or death from any cause.

Statistical analysis

We used Student’s t test to determine the statistical significance of the differences between two groups. Paired t tests were used to compare the expression levels of NET1 in tumour tissues with those in paired normal tissues. We evaluated the prognostic significance of NET1 expression through log-rank tests and Cox regression analyses. Spearman’s correlation analysis was used to assess the correlations between NET1 expression and various features. A P value less than 0.05 was considered to indicate statistical significance.

Results

Landscape of NET1 expression

By merging the data from TCGA and GTEx databases, we conducted a comparative analysis of NET1 expression levels in tumour and normal tissues. Compared with that in adjacent tissues, NET1 expression was significantly upregulated in LGG, BRCA, CESC, LUAD, ESCA, STES, COAD, PRAD, STAD, KIRC, LUSC, LIHC, WT, OV, PAAD, ALL, LAML, and PCPG tumours (Fig. 2A, all P < 0.05) and significantly downregulated in GBM, UCEC, KIRP, KIPAN, SKCM, THCA, TGCT, ACC, and KICH tumours (Fig. 2A, all P < 0.05). However, no significant change in NET1 expression was detected in HNSC, BLCA, READ, UCS, or CHOL tissues compared with adjacent tissues.

Fig. 2
figure 2

Differential analysis of NET1 gene expression across cancers. (A) Expression levels of the NET1 mRNA in pancancer tissues and corresponding normal tissues in TCGA and GTEx databases. (B) Differential expression of NET1 in tumour tissues and paired adjacent tissues in TCGA cohort. Abbreviations: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, and—not significant.

Next, we further analysed tumour and matched normal tissue samples from TCGA database. Compared with that in adjacent tissues, NET1 expression was significantly increased in BRCA, CHOL, STAD, LUAD, LIHC, LUSC, and ESCA tumour samples (Fig. 2B, all P < 0.05) and significantly downregulated in UCEC, KIRC, COAD, THCA, KIRP, KICH, and GBM tumour samples (Fig. 2B, all P < 0.05).

Prognostic and diagnostic value of NET1

We evaluated the relationship between NET1 expression and the OS of patients using various cancer tissues from TCGA database to investigate the effect of NET1 expression on the prognosis of patients with tumours (Fig. 3A-F). As shown in Fig. 3, high levels of NET1 were strongly correlated with shorter OS of patients with ACC (Fig. 3A), LIHC (Fig. 3B), LUSC (Fig. 3C) and PAAD (Fig. 3D) (all P < 0.05). In contrast, in patients with KIRC (Fig. 3E) and LGG (Fig. 3F), high levels of NET1 were strongly correlated with prolonged OS (all P < 0.05).

Fig. 3
figure 3

Analysis of the OS of patients with different types of cancer stratified by NET1 expression. (A-F) Relationships between NET1 expression levels and the prognosis of patients with various tumour types in TCGA database. (G-J) Relationships between NET1 expression levels and the prognosis of patients with various tumour types plotted via the KM Plotter website. Abbreviations: KM, Kaplan–Meier; OS, overall survival; HR, hazard ratio.

Furthermore, we constructed a KM plot to further analyse possible categories of cancer associated with differences in survival, serving as a supplement to TCGA survival prognosis data. As shown in Fig. 3, high levels of NET1 were strongly correlated with shorter OS of patients with BRCA (Fig. 3G), LIHC (Fig. 3H) and PAAD (Fig. 3I) (all P < 0.05). In contrast, high levels of NET1 were strongly correlated with prolonged OS of patients with KIRC (Fig. 3J) (P < 0.05).

We then conducted a univariate Cox regression analysis to evaluate the correlations of NET1 expression with several survival outcomes, including OS, DSS, DFI, and PFI, for patients with each cancer type in TCGA cohort. The results are presented in a forest plot. As shown in Fig. 4A, our study revealed that high NET1 expression was significantly associated with shorter OS of patients with LIHC, PAAD, ACC and BRCA (all P < 0.05). In contrast, high NET1 expression was significantly associated with longer OS of patients with LGG, KIRC, MESO, SKCM-M, and SKCM (all P < 0.05). Furthermore, high NET1 expression was associated with a shorter DFI in PAAD and LIHC patients (Fig. 4B, all P < 0.05). A significant correlation was found between high NET1 expression and low DSS of patients with LIHC, PAAD, ACC and LUSC, whereas high NET1 expression was associated with improved DSS of patients with LGG, KIRC, SKCM-M, and SKCM (Fig. 4C, all P < 0.05). In addition, as shown in Fig. 4D, high levels of NET1 were closely related to a shorter PFI for patients with LIHC, PAAD, ACC, HNSC and TGCT (all P < 0.05), whereas high NET1 expression was associated with prolonged DSS of LGG, MESO, KIRC, and GBM patients (all P < 0.05).

Fig. 4
figure 4

Univariate Cox regression analysis was performed to determine the prognostic value of NET1. (A) The prognostic value of NET1 expression for determining the OS, DFI (B), DSS (C), and PFI (D) of patients with various tumours. Abbreviations: OS, overall survival; DFI, disease-free interval; DSS, disease-specific survival; PFI, progression-free interval; CI, confidence interval.

We further assessed the diagnostic accuracy of NET1 for different cancer types using receiver operating characteristic (ROC) curves. As shown in Supplementary Fig. 1, NET1 could serve as a highly accurate diagnostic marker for KICH and SKCM (Supplementary Fig. 1A-B, all AUCs > 0.85). In GBM, ACC, and PUM1, NET1 exhibited moderate diagnostic performance (Supplementary Fig. 1C-E; all AUCs = 0.70–0.85). In KIRP, UCEC, and READ, the diagnostic value of NET1 was poor (Supplementary Fig. 1F-H; all AUCs = 0.50–0.70).

We also studied the correlations between NET1 expression and various clinicopathological features. According to the results shown in Supplementary Fig. 2A, NET1 expression was significantly positively correlated with the size of THCA and PRAD tumours (all P < 0.05) and significantly negatively correlated with the tumour size in KIPAN and KIRC (all P < 0.05). As shown in Supplementary Fig. 2B, in THCA and CHOL patients, an increase in NET1 expression was significantly positively correlated with lymph node metastasis (all P < 0.05). In addition, an increase in NET1 expression was significantly positively correlated with the tumour stage in THCA, PAAD, ACC, and BLCA patients and significantly negatively correlated with the tumour stage in the KIPAN and KIRC patients (Supplementary Fig. 2C; all P < 0.05). Supplementary Fig. 2D shows that an increase in NET1 expression was significantly positively correlated with the histological grade of PAAD but significantly negatively correlated with the histological grades of CESC, STES, KIPAN, KIRC and LGG (all P < 0.05).

NET1 gene alterations and correlations of NET1 expression with tumour characteristics

Studying the types of specific gene alterations is crucial for understanding the occurrence and development of cancer. For this purpose, we analysed gene alterations using the online resources cBioPortal and GSCALite. As shown in Supplementary Fig. 3A, the main type of genetic alteration detected in NET1 was “amplification”, most notably in BLCA (5.68%), OV (5.29%), ESCA (2.25%), BRCA (1.85%), and SARC (1.1%). Moreover, “mutations” were found mainly in UCS (2.2%), UCEC (0.88%), SKCM (0.78%), and COAD (0.54%), and “deep deletions” were observed mainly in DLBC (2.2%), SARC (0.88%), TGCT (0.78%), and MESO (0.54%). Across cancers, the mutation frequencies of “structural variations” and “multiple changes” in NET1 were generally less than 0.5%.

In addition, CNVs are important distortions that alter gene expression during the growth of tumours24. In most types of tumours, a relatively large proportion of single-copy deletions and low-level amplifications were observed in the NET1 gene, and homozygous deletions and high-level amplifications were found in a small proportion of samples (Supplementary Fig. 3B). Spearman’s correlation analysis of the LUSC, STAD, BRCA, LGG, HNSC, SARC, OV, BLCA, SKCM, ESCA, COAD, CESC, LIHC, LUAD, ACC, GBM, UCS, KIRC, PRAD, PAAD, READ, and PCPG cohorts revealed that NET1 mRNA expression was significantly positively correlated with the CNVs of the NET1 gene (all P < 0.05; Supplementary Fig. 3C and Supplementary Table 2).

The TMB and MSI are predictors of immunotherapy response25. Spearman’s correlation analysis revealed that NET1 expression was significantly positively correlated with the TMB in LUAD, PAAD, STAD, KIRP, BRCA, PRAD, ESCA, ACC, and THYM (all P < 0.05) and negatively correlated with the TMB in LGG and UCS (all P < 0.05; Supplementary Fig. 3D). In addition, in TGCT, STAD, READ, GBM, ESCA, and COAD, the expression level of NET1 was significantly positively correlated with MSI (Supplementary Fig. 3E). In contrast, in SKCM and CESC (Supplementary Fig. 3E), NET1 expression was significantly negatively correlated with MSI (all P < 0.05).

The stemness index of cancer has been shown to be closely related to drug resistance in tumours and the inherent immunosuppressive characteristics of tumours26,27. Tumour stemness can be measured by RNAss and DNAss. The expression of NET1 in THYM (cor = 0.43), CHOL (cor = 0.40), PCPG (cor = 0.25), THCA (cor = 0.20), and STES (cor = 0.15) was positively correlated with its DNAss value, and the differences were significant (all P < 0.05; Supplementary Fig. 3F); in contrast, the expression of NET1 in TGCT (cor = -0.73), LGG (cor = -0.50), MESO (cor = -0.46), UCS (cor = -0.31), KICH (cor = -0.26), HNSC (cor = -0.11), and KIPAN (cor = -0.10) was negatively correlated with its DNAss value, and the differences were significant (all P < 0.05; Supplementary Fig. 3F). Among the RNAss values, NET1 expression was significantly correlated with the RNAss of most cancer types (all P < 0.05; Supplementary Fig. 3G). The three tumours with the strongest positive correlations were LAML (cor = 0.32), and LGG (cor = 0.25). In contrast, NET1 expression in THYM (cor = -0.76), TGCT (cor = -0.62), THCA (cor = -0.54), UCS (cor = -0.45), KIPAN (cor = -0.44), CHOL (cor = -0.33), and MESO (cor = -0.32) was negatively correlated with the RNAss values (Supplementary Fig. 3G).

NET1 and epigenetic modifications

Dysregulation of epigenetic modifications has been proven to be a common feature of most human cancers28. Our analysis revealed strong positive correlations between NET1 expression and the expression of genes related to m1A, m5C, and m6A in almost all tumour types (Supplementary Fig. 4A). Moreover, dysregulation of DNA methylation is closely associated with the onset of various diseases, including cancers29. We also obtained information on different methylation sites in different cancers from TCGA database. The methylation levels of multiple sites were significantly negatively correlated with the NET1 mRNA expression levels in BLCA, BC, CESC, DLBC, HNSC, KIRC, KIRP, LAML, LIHC, LUAD, LUSC, MESO, SARC, SKCM, THCA, and THYM (Supplementary Fig. 4B). In addition, abnormal DNA methylation in the gene promoter region can lead to changes in chromatin structure and DNA stability, culminating in the disruption of gene expression within the organism30. Therefore, we further analysed the differences in the DNA methylation levels of the NET1 promoter between tumour and normal tissues using the UALCAN database. The results showed that the methylation levels of the NET1 gene promoter in SARC, KIRC, HNSC, and CESC tissues were significantly higher than those in corresponding normal tissues, whereas the opposite findings were obtained for COAD, ESCA, KIRP, BLCA, LIHC, TGCT, THCA, and BRCA (all P < 0.05; Supplementary Fig. 5A).

Correlation of NET1 expression with the tumour immune landscape

Investigating gene expression within tumours and its association with immune cell infiltration is important for understanding the regulatory relationship within the TME. The pan-cancer analysis revealed that the NET1 expression level was negatively correlated with the infiltration of various immune cells in multiple cancers (Fig. 5A), which typically included natural killer T cells, CD4 + TH1 cells, memory B cells and M1 macrophages (Fig. 5D-5G), whereas the infiltrating immune cells that were positively correlated with each other were mainly common lymphoid progenitors and mast cells (Fig. 5B‒5C). We studied the distribution of NET1 expression in the TME of various cancers, including ALL, AML, BRCA, CHOL, CRC, ESCA, glioma, KIRC, LIHC, LSCC, MB, MM, NSCLC, OV, PAAD, PRAD, STAD, SS, and UVM, to investigate its expression at the single-cell level in the TME (Fig. 6A-S). The results showed that NET1 was highly expressed mainly in the malignant cells of these cancers but was rarely expressed in immune effector cells, such as CD8 + cells.

Fig. 5
figure 5

Correlation analysis of NET1 expression with the infiltration of multiple immune cell types across cancers. (A) Heatmap showing the correlations between NET1 expression and infiltration of multiple immune cell types. (B) Scatter plots showing the results from the correlation analysis of NET1 expression with the infiltration of common lymphoid progenitor cells, mast cells (C), CD4 + TH1 cells (D), natural killer T cells (E), memory B cells (F) and M1 macrophages (G) across cancers. Abbreviations: Neg, negative correlation; No, not significant; Pos, positive correlation.

Fig. 6
figure 6

Single-cell sequencing analysis of NET1 expression. The expression levels of NET1 in the TME of ALL (A), AML (B), BRCA (C), CHOL (D), CRC (E), ESCA (F), glioma (G), KIRC (H), LIHC (I), LSCC (J), MB (K), MM (L), NSCLC (M), OV (N), PAAD (O), PRAD (P), STAD (Q), and SS (R) patients. Distribution of NET1 expression in several BRCA datasets (S).

Correlation of NET1 expression with immune indicators

We further examined the relationship between NET1 expression and the expression of immunoregulatory genes. The heatmap shows that NET1 was coexpressed with most immune-related genes in all cancers. Specifically, in THYM, KICH, KIPAN, THCA, KIRC, OV, BRCA, HNSC, PCPG, and THCA, NET1 expression generally exhibited a positive correlation with the expression of five immune-related gene sets: chemokines, receptors, MHC, immunoinhibitors, and immunostimulators. Conversely, in LGG and TGCT, NET1 expression generally displayed a negative correlation with the expression of the abovementioned immune-related gene sets (Fig. 7A). Among the immune activation-related genes, RAETIE, TNFSF18, TNFSF4, NT5E, PVR, ULBP1, and CD276 were positively correlated with NET1 expression in most tumours (Fig. 7A).

Fig. 7
figure 7

Correlation analysis of NET1 expression with immune-related genes and tumour scores in various cancer types. (A) Heatmap showing the correlations between NET1 expression and the expression of chemokine-, receptor-, MHC-, immunoinhibitor-, and immunostimulator-related genes. (B) Correlations of NET1 expression with the ESTIMATE score, immune score (C) and stromal score (D) across cancers. Abbreviations: MHC, major histocompatibility complex; *P < 0.05.

We then analysed the relationship between the expression of NET1 and the immune score of the TME. NET1 expression was negatively correlated with the ESTIMATE scores of COAD, LUAD, LUSC, UCEC, LGG, GBM, BRCA, CESC, SARC, LAML, PAAD, THYM, STAD, UVM and ESCA, and was positively correlated with the ESTIMATE scores of THCA, KICH, OV, and SKCM (Fig. 7B). The expression of NET1 was negatively correlated with the immune scores of BLCA, PRAD, LUSC, LGG, UCEC, LUAD, COAD, BRCA, READ, STAD, PAAD, CESC, LAML, MESO, SARC, UVM, ESCA, LIHC, TGCT, and THYM, and positively correlated with the immune scores of THCA, KICH, and OV (Fig. 7C). NET1 expression was negatively correlated with the stromal scores of CESC, LGG, BRCA, LIHC, PAAD, STAD, UVM, and ESCA, and positively correlated with the stromal scores of TGCT, KICH, THCA, OV, THYM, and SKCM (Fig. 7D).

Role of NET1 in predicting the immunotherapy response and drug efficacy

We evaluated the predictive function of NET1 expression in the immunotherapy response using the TISMO database. As shown in Fig. 8A, significant differences in NET1 expression were observed before and after immune checkpoint blockade (ICB) treatment and between responders and nonresponders (all P < 0.05). Additionally, significant differences in NET1 expression were detected before and after cytokine treatment in the six cell lines (Fig. 8B). We also evaluated NET1 as a biomarker based on the TIDE score. As a biomarker, NET1 had a better ability to predict survival in the 12 immunotherapy cohorts (Fig. 8C). The GDSC dataset was subsequently used to analyse the half-maximal inhibitory concentrations of drugs based on NET1 expression. TW 37 (a BCL-2 inhibitor), 52–7-oxozeaenol (a selective TAKI inhibitor), and pazopanib (a tyrosine kinase inhibitor) were identified as the three drugs most positively correlated with NET1 expression, and afatinib (an EGFR inhibitor), lapatinib (a tyrosine kinase inhibitor), and gefitinib (a tyrosine kinase inhibitor) were the three drugs most negatively correlated with NET1 expression (Fig. 8D and Supplementary Table 3). Based on the correlation between NET1 expression and the half-maximal inhibitory concentrations of drugs in the CTRP database, BRD-K99006945 (a BAF inhibitor), vemurafenib (an inhibitor of the mutated form of BRAF serine–threonine kinase), and ML210 (a glutathione peroxidase 4 inhibitor) were the three drugs most strongly correlated with NET1 expression, and austocystin D (a tumour inhibitor), afatinib (a tyrosine kinase inhibitor) and erlotinib (a tyrosine kinase inhibitor) were the three drugs most negatively correlated with NET1 expression (Fig. 8E and Supplementary Table 4).

Fig. 8
figure 8

Correlations of NET1 expression with the immunotherapy response and drug sensitivity. (A) Differences in NET1 expression between the immune treatment group and the control group. (B) Differences in NET1 expression between the cytokine treatment group and the control group. (C) Ability of NET1 to serve as a biomarker compared with other markers in different immune therapy cohorts. (D) The GDSC and CTRP (E) databases were used to predict drugs that are influenced by NET1 expression levels. Abbreviations: Neg, negative correlation; Pos, positive correlation; *P < 0.05, **P < 0.01, and ***P < 0.001.

Pathways and functions related to NET1 expression

We conducted a GSEA of differentially expressed genes in the high and low NET1 expression groups using data from 33 tumours obtained from TCGA database to explore the biological pathways through which NET1 may affect tumour onset and development. As shown in Fig. 9A, we observed the enrichment of immune-related pathways, including the TNF/NF-κB pathway, IFN-α response, IFN-γ response, inflammatory response, IL-6/JAK/STAT3 pathway, and IL-2/STAT5 pathway, especially in tumours such as BLCA, BRCA, COAD, GBM, and PRAD. In addition, we observed positive correlations between NET1 expression and the MYC-targeted signalling pathway, the MTORC1 signalling pathway, and other cancer-related pathways. A correlation analysis with oncogenes also revealed that NET1 expression was strongly correlated with the expression of well-known oncogenes, such as CDC6, EIF3A, FOXM1, HMGA1, NME1, and UHRF1 (Supplementary Fig. 4A-B), and these correlations were observed in various cancers. We further investigated the relationship between NET1 expression and the functional status of different cancers by analysing single-cell sequencing data from 14 cancers obtained from the CancerSEA database. In most cancers, NET1 expression was positively correlated with angiogenesis, differentiation, metastasis, quiescence, and stemness and was negatively correlated with apoptosis, the cell cycle, DNA damage, and DNA repair (Supplementary Fig. 7A).

Fig. 9
figure 9

GSEA of differentially expressed genes in the high and low NET1 expression groups across cancers. Abbreviations: NES, normalized enrichment score; FDR, false discovery rate.

Expression level and prognostic value of NET1 in breast cancer

We first merged and analysed 13 TNBC datasets from the GEO database using a meta-analysis approach and found that the expression of NET1 in tumour tissues was significantly higher than that in adjacent tissues (P = 0.002, 95% CI = 0.86 [0.31–1.40]) (Fig. 10A). We further combined TNBC datasets from GEO to analyse the difference in prognosis between patients with different NET1-related OS rates and found that high expression of NET1 was significantly associated with shorter OS (P = 0.030, HR: 1.57, 95% CI: 1.04–2.38) (Fig. 10B). We also analysed the differences in the prognostic value of NET1 based on metastasis-free survival (MFS) and RFS. The results revealed that high expression of NET1 was significantly associated with shorter RFS (P = 0.007, HR:1.97, 95% CI: 1.20–3.23). However, high expression of NET1 had no significant effect on MFS (P = 0.847, HR: 0.98, 95% CI: 0.76–1.25) (Fig. 10B). Figure 10C shows immunohistochemical staining for NET1 expression in different subtypes of breast cancer. Using real-world clinical samples, we analysed 236 breast cancer patients, including 80 TNBC patients. The median follow-up time was 41 (3–83) months. We analysed the relationship between NET1 expression and DFS in patients with breast cancer. The results revealed that high NET1 expression was significantly associated with shorter DFS of both breast cancer patients and TNBC patients (P < 0.001; P = 0.025, respectively) (Fig. 10D-E). Through univariate and multivariate analyses, we found that high expression of NET1 (P = 0.001, HR: 3.46 95% CI: 1.66–7.23) and a higher T stage (P = 0.045, HR: 5.18, 95% CI: 1.04–25.87) were independent prognostic factors for DFS in patients with breast cancer (Fig. 10F), and high expression of NET1 (P = 0.006, HR: 8.69, 95% CI: 1.88–40.17) and a higher N stage (P = 0.002, HR: 89.54, 95% CI: 5.08–1579.54) were independent prognostic factors for DFS in patients with TNBC (Fig. 10G).

Fig. 10
figure 10

The expression level of NET1 in breast cancer samples and its potential value as a prognostic marker. (A) The expression levels of NET1 in tumours and adjacent tissues were evaluated across 13 TNBC GEO datasets through a meta-analysis. (B) The relationships of NET1 expression with OS, MFS, and RFS were evaluated after TNBC data were merged from GEO through a meta-analysis. (C) Immunohistochemistry revealed the expression levels of NET1 in different subtypes of breast cancer. (D) Survival curves for the DFS of breast cancer patients with different NET1 expression levels in real-world clinical cohorts. (E) Survival curves for the DFS of TNBC patients with different NET1 expression levels in real-world clinical cohorts. (F) Univariate and multivariate analyses of factors affecting the DFS of patients with breast cancer. (G) Univariate and multivariate analyses of factors affecting the DFS of TNBC patients. Abbreviations: OS, overall survival; MFS, metastasis-free survival; RFS, recurrence-free survival; DFS, disease-free survival; HR, hazard ratio; 95% CI, 95% confidence interval; T, tumour; N, lymph node.

We further analysed the expression levels in TNBC cell lines using a website and selected TNBC cell lines with relatively low expression to construct stable NET1-overexpressing cell lines (Supplementary Fig. 8A). RT‒qPCR detection of the NET1 mRNA level in TNBC cell lines revealed that the mRNA expression level of NET1 was significantly increased in NET1-overexpressing TNBC cell lines (Supplementary Fig. 8B). In addition, the western blot analysis revealed that the NET1 protein expression level in stable TNBC cell lines was also significantly increased (Supplementary Fig. 8C-8D).

Study of NET1 function in TNBC cells

We conducted in vitro experiments to study the biological function of NET1 in TNBC cells. Through colony formation assays, we found that NET1 overexpression significantly increased the colony formation ability of TNBC cells (all P < 0.05; Fig. 11A). In addition, the results of the CCK-8 assay revealed that overexpression of NET1 increased the proliferation of TNBC cells (all P < 0.05; Fig. 11B). Furthermore, by conducting subcutaneous tumour formation experiments in nude mice using TNBC cells, we found that NET1 overexpression significantly increased the tumour volume compared with that in the control group (Fig. 11C-11D, all P < 0.05). However, no significant difference in mouse body weight was observed between the experimental group and the treatment group (Fig. 11C-D, all P > 0.05).

Fig. 11
figure 11

Study of NET1 function in TNBC cells. (A) Colony formation and quantitative analyses were conducted to evaluate the changes in the colony formation ability of TNBC cells with and without NET1 overexpression. (B) The CCK-8 detection method was used to evaluate the effect of NET1 overexpression on the proliferation of TNBC cells. (C-D) Subcutaneous tumour formation assays in nude mice were conducted using TNBC cells with and without NET1 overexpression to detect the changes in the tumour volume and mouse weight. (E) A wound healing assay was performed to evaluate the changes in the cell migration rate of TNBC cells with and without NET1 overexpression. (F) Transwell assays were conducted to evaluate the changes in the migration of TNBC cells with and without NET1 overexpression. (G) Changes in the cell cycle distribution of cells with and without NET1 overexpression were detected. (H) Changes in the apoptosis of cells with and without NET1 overexpression were detected. Abbreviations: ns, not significant, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

We subsequently conducted wound healing and transwell assays to evaluate the impact of NET1 overexpression on the migration of TNBC cells. The results revealed that the upregulation of NET1 expression significantly increased the migration of TNBC cells compared with that of control cells (Fig. 11E-F).

We further detected changes in the cell cycle before and after NET1 overexpression, and the results revealed a significantly higher percentage of cells in S phase in the NET1-overexpressing group of SUM-159PT cells (P < 0.05), but no significant difference in the percentage of cells in G2 phase was observed between the groups (P > 0.05; Fig. 11G). Although increasing trends were observed for the percentages of BT-549 cells in both the S and G2 phases before and after the overexpression of NET1, the differences were not significant (P > 0.05; Fig. 11G).

We also detected changes in the apoptosis of cells with and without NET1 overexpression, and the results obtained from both SUM-159PT cells and BT-549 cells revealed that the proportion of apoptotic cells in the NET1-overexpressing group was significantly decreased (all P < 0.05) (Fig. 11H).

Study of the mechanism of NET1 in TNBC cells

A sequencing analysis of MDA-MB-231 cells with and without NET1 overexpression and subsequent enrichment analysis of biological process (BP) terms indicated significant enrichment of genes related to cell apoptosis, cell proliferation, and cell migration (Fig. 12A). In addition, a KEGG analysis revealed that the PI3K-AKT signalling pathway was significantly activated (Fig. 12B). A WB analysis further revealed no significant difference in the AKT protein levels in TNBC cells with and without NET1 overexpression, but the level of P-AKT (Ser473) was significantly increased after NET1 overexpression (Fig. 12C-D). In addition, we clarified the upstream and downstream relationships between NET1 and P-AKT by detecting changes in NET1 and P-AKT levels after treatment with the inhibitor MK2206. MK2206 is an inhibitor of P-AKT (Ser473). In SUM-159PT cells overexpressing NET1, the WB analysis revealed that the P-AKT (Ser473) level was significantly decreased at 12 h and 24 h after drug administration, but no significant difference in NET1 expression was observed (Fig. 12E-F). Furthermore, through a CCK-8 assay, we verified whether the administration of MK2206 could reverse the effect of NET1 overexpression on cell proliferation. The results showed that after treatment with MK2206, the proliferation of NET1-overexpressing cells was significantly inhibited compared with that of control cells (Fig. 12G-I).

Fig. 12
figure 12

Mechanistic study of NET1 in TNBC cells. (A) BP and (B) KEGG enrichment analyses of DEGs in NET1-overexpressing cells compared with non-NET1-overexpressing MDA-MB-231 cells. (E) Changes in the protein levels of AKT and P-AKT (Ser473) in TNBC cell lines with and without NET1 overexpression. (D) Semiquantitative statistical analysis of the changes in AKT and P-AKT (Ser473) protein levels in TNBC cell lines with and without NET1 overexpression. (E) Changes in the protein levels of NET1 and P-AKT (Ser473) after treatment with a P-AKT inhibitor (MK2206) in SUM-159PT cells overexpressing NET1. (F) Semiquantitative statistical analysis of the changes in the protein levels of NET1 and P-AKT (Ser473) after treatment with a P-AKT inhibitor (MK2206). (G-I) CCK-8 assay of changes in the proliferation of TNBC cells with and without NET1 overexpression after treatment with a P-AKT inhibitor (MK2206). Abbreviations: BP, biological process; KEGG, Kyoto Encyclopedia of Genes and Genomes; ns, not significant, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Discussion

Multidimensional data mining and analysis are crucial for exploring the heterogeneity and complexity of tumours and for identifying prognostic biomarkers. NET1, a vital member of the guanine nucleotide exchange factor family, can mediate malignant biological behaviours, such as proliferation and invasion, in cancer cells9. NET1 has been identified as a prognostic biomarker for gastric cancer, liver cancer, and pancreatic cancer, but its prognostic and biological significance in most other types of cancer remain unclear10,11,12. We conducted a comprehensive pan-cancer study of NET1 and thus revealed a potential role for NET1 as a biomarker. Our results suggest that the expression of NET1 in 18 cancers is significantly higher than that in adjacent tissues. High levels of NET1 expression in BRCA, LIHC, LUSC, and PAAD are associated with a poor prognosis. Conversely, high NET1 expression in KIRC, LGG, and MESO is correlated with a more favourable prognosis. Using the GEO database and real-world breast cancer clinical samples, we further verified that high expression of NET1 is associated with poor prognosis for TNBC patients.

We performed cellular experiments to further explore the molecular mechanisms underlying the poor prognosis of TNBC patients and found that high NET1 expression could promote cell proliferation and invasion, inhibit cell apoptosis, and accelerate cell division. In addition, the BP enrichment analysis of DEGs in MDA-MB-231 cells with and without the overexpression of NET1 indicated that NET1 is related to cell proliferation, migration, apoptosis, and the cell cycle. A subsequent KEGG enrichment analysis of the DEGs revealed that genes related to the AKT signalling pathway were significantly enriched. The AKT signalling pathway can regulate various biological processes, including cell growth, survival, proliferation, metabolism, and migration31.

AKT is also an essential mediator of the PI3K/Akt signalling pathway, is often activated in human cancers, and is a primary target of drug intervention32,33. This kinase promotes tumour cell proliferation and regulates apoptosis by activating downstream signalling pathways through phosphorylation34. Our study revealed that NET1 can activate the AKT signalling pathway in TNBC cells. The overexpression of NET1 significantly increased the level of P-AKT (Ser473). Moreover, after the addition of MK2206 (an effective P-AKT inhibitor), the level of P-AKT (Ser473) was significantly reduced, and the overexpression of NET1 no longer promoted tumour cell proliferation. These findings suggest that P-AKT (Ser473) plays a crucial role in tumour proliferation. In addition, recent research has shown that NET1 knockout can significantly inhibit the phosphorylation of AKT in HCC cells and the expression of genes downstream of AKT. Through the AKT signalling pathway, NET1 plays a carcinogenic role in the progression and metastasis of HCC. Therefore, NET1 could be used as a potential therapeutic target and prognostic marker for HCC11,35. We also detected changes in the expression of NET1 after treatment with MK2206 to further clarify the upstream and downstream relationships between NET1 and P-AKT (Ser473). We found that the treatment of cells with MK2206 for 12 or 24 h significantly decreased the P-AKT level but did not significantly alter the expression level of NET1. This finding suggested that NET1 may be upstream of the P-AKT molecule. Therefore, the NET1/P-AKT axis may be a key pathway regulating the proliferation of TNBC cells.

In summary, high expression of NET1 in TNBC induces proliferation, which suggests that NET1 is an important oncogene. Moreover, NET1 expression is strongly correlated with the expression of various oncogenes, including well-known oncogenes such as CDC6, EIF3A, FOXM1, HMGA1, NME1, and UHRF1, in multiple cancers36. The single-cell functional analysis also revealed a positive correlation between NET1 expression and most BPs, such as angiogenesis, differentiation, metastasis, dormancy, and stem cell properties, in tumours. GSEA of NET1 revealed that NET1 is associated with the activation of oncogenic signalling pathways, such as the MYC-targeted signalling pathway and the MTORC1 signalling pathway, in most tumours. Therefore, NET1 plays a strong cancer-promoting role in various cancers, and its activation of related oncogenes and signalling pathways to mediate malignant biological behaviours is a possible explanation for its carcinogenic effects.

Gene alterations play crucial roles in the occurrence, development, and prognosis of tumours37. The occurrence and progression of tumours always involve gene alterations38. The main types of NET1 alterations in tumours are “amplification” and “mutations”, which are mainly found in BLCA, OV, ESCA, BRCA, and SARC and in UCS, UCEC, SKCM, and COAD, respectively. We also noted that NET1 expression was significantly positively correlated with CNVs in multiple tumours, such as BRCA, CESC, HNSC, LUAD, LUSC, OV, and STES. CNVs are major genetic structural variations in the human cancer genome30. CNVs associated with a loss or gain of DNA sequences are associated with the development and progression of cancer39. Therefore, genetic changes in NET1 could also be a possible cause of its abnormal expression.

Epigenetic regulation occurs more frequently and significantly than somatic mutations40. Epigenetic changes that regulate gene expression are considered essential features of cancer, and the elucidation of epigenetic regulatory networks is useful for understanding the occurrence of cancer41. Interestingly, our study revealed that NET1 expression was positively correlated with the expression of m1A-, m5C-, and m6A-related genes in almost all cancers. These findings suggest that NET1 is closely related to epigenetic regulation and that epigenetic changes in genomic regulatory regions often participate in tumour occurrence and can drive the occurrence and progression of cancer by altering gene expression. Furthermore, a classic hallmark of the human cancer genome is abnormal DNA methylation, and high methylation in the promoter region often leads to the epigenetic silencing of tumour suppressor genes and thus the promotion of the occurrence and progression of cancer42. We further analysed the difference in the DNA methylation levels of the NET1 promoter between tumour and normal tissues via the UALCAN database. The results showed that the methylation levels of the NET1 gene promoter in SARC, KIRC, HNSC, and CESC tissues were significantly higher than those in corresponding normal tissues. Therefore, NET1 may also mediate the occurrence and development of cancer by affecting the expression levels of oncogenes and tumour suppressor genes through changes in promoter methylation levels.

The TME includes various cell types (endothelial cells, fibroblasts, immune cells, etc.), which are nourished by the vascular network surrounding tumour cells and include extracellular components such as cytokines, growth factors, hormones, and the extracellular matrix43. The immune cells in the TME and adaptive immune cells interact with tumour cells through direct contact or chemotactic factor and cytokine signal transduction, and these interactions affect the tumour behaviour and the response to treatment44. Many researchers have shown that immune cells can both enhance and hinder treatment effects and that their activation status and location in the TME may vary45. Interestingly, we found that the expression level of NET1 was negatively correlated with the abundance of cells widely considered antitumour immune cells, including CD8 + T cells, NK cells, and M1 macrophages. We also noted that the expression level of NET1 was positively correlated with the abundance of common lymphoid progenitor and mast cells in the TME. All lymphocytes are thought to develop from common lymphoid progenitors46. Unfortunately, the specific biological functions of common lymphoid progenitors in the TME have not yet been determined. In tumours, mast cells interact with infiltrating immune cells, tumour cells, and the extracellular matrix through direct cell-to-cell interactions or the release of a series of mediators that can reshape the TME. Mast cells participate in angiogenesis by releasing classical proangiogenic factors such as VEGF, FGF-2, PDGF, and IL-6 and nonclassical proangiogenic factors such as trypsin and chymotrypsin, which results in the induction of neovascularization. Mast cells also support tumour invasion by releasing large amounts of matrix metalloproteinases47. In addition, NF-κB and AP-1 activities in tumour cells are enhanced by the remodelling of the inflammatory microenvironment by mast cells. Activated mast cells also exacerbate tumour immune suppression by releasing adenosine and increasing the number of regulatory T cells, which enhances the inhibitory effects of T cells and natural killer cells on tumours48. In addition, single-cell sequencing revealed that NET1 is highly expressed mainly on malignant cells but is rarely expressed on immune effector cells, such as CD8 + cells. Interestingly, NET1 expression was significantly correlated with the expression of immune-related genes, such as those encoding chemokines, receptors, MHC, immunoinhibitors, and immunostimulators, in almost all cancers. These findings indirectly indicate that NET1 may affect immunoregulatory genes, especially chemokines, and their receptor-related pathways and thereby affect the level of immune cells infiltrating the TME.

The immune score reflects the proportion of infiltrating immune cells in tumour tissues. Our results showed that NET1 expression was negatively correlated with the immune scores of BLCA, PRAD, LUSC, LGG, UCEC, LUAD, COAD, BRCA, READ, STAD, PAAD, CESC, LAML, MESO, SARC, UVM, ESCA, LIHC, TGCT, and THYM. This finding suggested that these types of tumours may have lower immunogenicity. The TMB is defined as the total number of base mutations per million cells in a tumour49. MSI is caused by abnormalities in DNA mismatch repair, which lead to genetic replication obstacles and tumour development and thus affect the tumour prognosis50. Currently, the TMB is widely believed to stimulate the production of tumour-specific and highly immunogenic antibodies and is a new target for predicting the efficacy of immunotherapy in tumours51. The MSI level can also indirectly reflect the efficacy of immunotherapy. Patients with a high TMB and MSI exhibit a better response to immunotherapy52. Our results suggest that the expression of NET1 in LUAD, PAAD, STAD, KIRP, BRCA, PRAD, ESCA, ACC, and THYM is significantly positively correlated with the TMB. Moreover, in TGCT, STAD, READ, GBM, ESCA, and COAD, the expression level of NET1 was significantly positively correlated with MSI. This finding may suggest that NET1 may induce a better immune response in some cancers. Our GSEA also revealed a close association between NET1 expression and immune-related pathways in various types of tumours, including the TNFA/NF-κB signalling pathway, IFN-α response, IFN-γ response, inflammatory response, IL-6/JAK/STAT3 pathway, and IL-2/STAT5 pathway. These findings further highlight the potential mechanism through which NET1 regulates immunogenicity.

We further analysed the effects of immunotherapy and cytokine therapy on NET1 expression. Significant differences in the expression of NET1 were detected in 6 patients before and after ICB therapy and between the responding and nonresponding groups. In addition, significant differences in NET1 expression were found among the six cell lines before and after cytokine treatment, indicating that NET1 plays a key role in the process of immunotherapy. Furthermore, we also examined the correlation between NET1 mRNA expression and the sensitivity to anticancer drugs. An analysis of the effects of various concentrations of drugs on NET1 expression using the GDSC and CTRP databases revealed that NET1 expression was negatively correlated with the concentrations of various tyrosine kinase inhibitors. These drugs can further guide the selection of clinical drugs. Interestingly, we also calculated the stemness indices of the tumours. The stemness index is an indicator of the similarity between tumours and stem cells. This index can measure tumour stemness through the RNAss and DNAss values. The cancer stemness index has been proven to be closely related to drug resistance in cancer and the intrinsic immunosuppressive characteristics of tumours26,53. The expression of NET1 in THYM, CHOL, PCPG, THCA, and STES cells was positively correlated with the DNAss values. In addition, the NET1 expression level was positively correlated with the RNAss value in LAML and LGG. These results suggest that high NET1 expression may also affect the sensitivity of patients with cancer to drug treatment.

This study has several limitations. First, we only conducted follow-up experimental verification in breast cancer. However, the cancer we chose may not be the one with the most significant differences in survival for verification because many studies on NET1 have focused on digestive system-related tumours, especially liver cancer, gastric cancer, and pancreatic cancer. At present, the incidence of breast cancer is increasing annually, and TNBC is the subtype of breast cancer with the worst prognosis. Due to the lack of effective treatment targets, its treatments are limited. Therefore, an urgent need is to find new treatment targets. In addition, our preliminary data from the GEO data meta-analysis revealed that TNBC patients with high NET1 expression have a poor prognosis, and we also verified the prognostic value of NET1 in a real-world cohort. Therefore, we chose TNBC for subsequent basic experimental verification. Second, the differences in the prognostic value of NET1 expression in different tumours have not been further verified, and the main causes of these differences should be further explored in subsequent research. Third, our exploration of the regulatory mechanism of NET1 is not deep enough and focuses only on the indirect mechanism. The inherent mechanism by which NET1 regulates malignant biological behaviour still needs to be further explored.

In summary, the NET1 gene can be altered, which is a warning signal for the onset of cancer. Moreover, NET1 may play roles in various processes, including the regulation of epigenetic modification, immune regulation, and immunotherapy. Second, NET1 is strongly associated with oncogenes and oncogenic signalling pathways in various tumours. Therefore, NET1 can serve as a prognostic marker for multiple types of cancer. We further showed that TNBC patients with high NET1 expression have poorer prognoses. NET1 can mediate the malignant proliferation of cells by regulating the AKT signalling pathway, but the specific internal regulatory mechanism of NET1 needs to be further clarified. In addition, the research value of NET1 in other cancers warrants further exploration.

Conclusions

NET1 can serve as a potential prognostic marker for LIHC, LUSC, PAAD, and BRCA tumours. Real-world data further suggest that NET1 can also serve as a prognostic indicator for TNBC. High expression of NET1 may contribute to the malignant proliferation of TNBC cells, potentially through the AKT signalling pathway. Moreover, NET1 may contribute to the formation of an immunosuppressive microenvironment that can promote tumour progression. Therefore, targeting NET1 may represents a promising approach for inhibiting tumour progression.