2.1 Data Acquisition and Download

In this study, 693 glioma patients were sourced from the mRNAseq_693 data set in the CGGA database [19], and 697 glioma patients were obtained from the TCGA database (Table 1) [20]. Additionally, the CGGA database offered sc-RNA data for glioma patients, encompassing a total of 6148 cells from 73 biopsy samples collected via surgery from 14 patients. All data utilized in this study are publicly accessible.

2.2 Functional Enrichment Analysis

The gene set associated with APOBEC3C was subjected to Pearson's correlation analysis (R > 0.5, p < 0.001) and subsequently analyzed using the DAVID database (https://david.ncifcrf.gov). Functional enrichment analysis was conducted based on Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, with official gene symbols used as identifiers and Homo sapiens designated as the species. In this study, all enrichment results are ranked in ascending order according to their p values.

2.3 Gene Set Variation Analysis (GSVA)

We retrieved the immune response-related gene set from AmiGO2 (http://amigo.geneontology.org/amigo). The relationship between APOBEC3C and the immune response was investigated using Pearson's correlation analysis, and a heatmap was generated using the Pheatmap package.

2.4 APOBEC3C and TME

To investigate the potential relationship between APOBEC3C and the glioma TME, we evaluated its correlation with immune cell populations using Pearson's correlation coefficient. Furthermore, the association between APOBEC3C and diverse immune-infiltrating cell types was assessed through the TIMER database.

2.5 Single-Cell Sequencing

To investigate the association between APOBEC3C and the TME in GBM, single-cell RNA-seq data were obtained from the CGGA database. The expression matrices from single-cell sequencing were processed using the Seurat package in R. Gene expression data underwent normalization, followed by principal component analysis (PCA) using the “RunPCA” function. For data visualization, the “UMAP” package was employed, while cell annotation was performed using the “SingleR” package. The similarity of cell clusters was determined based on their characteristic genes, with higher gene similarity indicating greater cluster similarity.

2.6 Reagents and Antibodies

Antibodies targeting N-cadherin, E-cadherin, Vimentin, p65, and phosphorylated p65 (Ser536) were sourced from Cell Signaling Technology (Danvers, Massachusetts, USA). The anti-GAPDH antibody was acquired from Protein Technology (Wuhan, China). Additionally, polyvinylidene fluoride (PVDF) membranes and chemiluminescent reagents were supplied by Millipore (Massachusetts, USA).

2.7 Cell Culture

The glioma cells were acquired from Wuhan Punosai Life Science and Technology Co Ltd. These cells were maintained in DMEM high-glucose medium enriched with 10% fetal bovine serum (FBS) and a dual-antibiotic solution. Cultivation was performed in a humidified incubator at 37°C with 5% CO₂. The complete medium was replaced every 2–3 days, adjusted according to the proliferation rate of the cells.

2.8 Cell-Scratch Assay

Following our previous method [21], well-grown glioma cells were inoculated into 6-well plates. When the cell growth density reached ~80%, each well was scratched horizontally with a 200 μL pipette tip. Subsequently, the cells were washed with PBS buffer. Finally, siRNA transfection was carried out according to the instructions of the transfection reagents provided by Neofect (Beijing) Biotech Co. Ltd., and the cells were continuously incubated in the incubator. Images were captured at 0, 12, and 24 h to observe the changes in the scratch width.

2.9 Transwell Assays

The migratory and invasive properties of cells were evaluated using the Transwell assay. For the migration assay, transfected cells were placed in the upper chamber and maintained in serum-free DMEM, while the lower chamber contained DMEM supplemented with 10% FBS. In the invasion assay, the upper chamber was pre-coated with Matrigel. Following a 24-h incubation period, the cells were fixed with 20% formaldehyde and stained with 0.5% crystal violet. Images were subsequently captured using an inverted microscope.

2.10 Western Blot Analysis

Following cell transfection, cellular proteins were extracted according to the manufacturer's protocol. Proteins were initially separated via electrophoresis and then transferred onto a PVDF membrane. The membrane was blocked with skimmed milk at room temperature for 2 h. After blocking, the membrane was incubated with the primary antibody overnight at 4°C. Subsequently, the membrane was treated with the secondary antibody for 2 h, and protein bands were visualized using a chemiluminescent detection system.

2.11 Statistical Analysis

All data analyses and graphical visualizations were conducted using R software (version 4.3.1) and GraphPad. The t-test was employed to evaluate the significance of differences between two sample means, while analysis of variance (ANOVA) was utilized to assess the significance of differences among more sample means. The log-rank test was applied to evaluate the significance of prognostic indicators. Multivariate analysis was conducted using Cox regression models to identify independent prognostic factors in glioma patients. In all statistical tests, a p < 0.05 was considered statistically significant.

3.1 Differential Expression of APOBEC3C in Pan-Cancer

The TCGA database presents the APOBEC3C protein expression levels in 20 different types of cancer, suggesting that the expression of APOBEC3C in tumor samples is significantly higher than that in normal samples (Figure 1A). The GEPIA database (http://gepia.cancer-pku.cn/) further highlights the difference in APOBEC3C expression between normal tissues and tumor patients (Figure 1B,C).

3.2 APOBEC3C Is Enriched in Glioma Patients With Different Clinical Features and Molecular Differences

We observed that patients with varying levels of APOBEC3C expression exhibited distinct clinical and molecular pathological features, such as WHO grade, 1p/19q co-deletion status, MGMT promoter methylation status, and IDH mutation status. In both the CGGA and TCGA databases, the distribution of APOBEC3C expression levels displayed an asymmetric pattern (Figure 2A,B). Analysis of the CGGA database revealed that APOBEC3C expression was positively correlated with glioma WHO grade (Figure 2C), elevated in samples lacking 1p/19q codeletion (Figure 2D), and higher in IDH wild-type patients (Figure 2F). These results were corroborated in the TCGA database (Figure 2G,H,J). Additionally, APOBEC3C expression was significantly increased in samples without MGMT promoter methylation in the TCGA database (Figure 2I), with a similar trend observed in the CGGA database (Figure 2E). Collectively, these findings suggest that gliomas with more aggressive phenotypes are characterized by elevated APOBEC3C expression levels.

3.3 An APOBEC3C Biomarker May Help Diagnose Glioma Mesenchymal Subtypes

We explored the levels of APOBEC3C expression among different histological subtypes. In both the CGGA and TCGA databases, APOBEC3C was significantly overrepresented in mesenchymal cells (p < 0.05) (Figure 3A,C). To evaluate the specificity of APOBEC3C expression in glioma mesenchymal subtypes, we utilized receiver operating characteristic (ROC) curves. The analysis showed an AUC of 84.9% in the CGGA database (Figure 3B) and an AUC of 89.1% in the TCGA database (Figure 3D). APOBEC3C is preferentially enriched in the mesenchymal subtype with a poor prognosis, indicating its potential as a mesenchymal biomarker for gliomas. Since the mesenchymal subtype of GBM is more aggressive than other subtypes, these findings further confirm that APOBEC3C is positively correlated with the malignancy level of gliomas.

3.4 Glioma Immune and Inflammatory Responses Are Regulated by APOBEC3C

To clarify the biological functions of APOBEC3C, we carried out Pearson's correlation analysis to identify genes most closely related to APOBEC3C. Subsequently, these gene sets were subjected to GO and KEGG analyses. The biological processes most closely related to APOBEC3C include the innate immune response, positive regulation of NF-κB signaling, and inflammatory response in CGGA (Figure 4A). Moreover, APOBEC3C is located in exosomes and membranes (Figure 4B). Its molecular functions mainly involve protein binding, cytokine receptor activity, and signaling receptor activity, among others (Figure 4C). The most relevant signaling pathways are the lysosome and tumor necrosis factor signaling pathways (Figure 4G). These findings were reproduced in TCGA (Figure 4D–F,H). So, these data imply that APOBEC3C may play a significant role in glioma immune and inflammatory responses, making it an important immune target.

3.5 APOBEC3C Is Positively Correlated With Glioma-Related Immune Pathways

In the context of immunogenic cancer cell death, lymphocytes are activated and chemokines and cytokines are released, which leads to apoptosis or necrosis [22]. Therefore, we explored the impact of APOBEC3C activation on immune responses. GSVA was carried out on the public databases to determine the immune process enrichment scores. A heatmap of the correlation analysis between the enrichment scores and APOBEC3C expression showed a positive correlation between most of the immune responses and APOBEC3C expression (Figure 5A). The results from the TCGA database confirmed these findings (Figure 5B). These data indicate a potential link between APOBEC3C and glioma immune responses.

3.6 APOBEC3C Is Associated With Immune Checkpoint Inhibitors and Inflammatory Activity

We found that APOBEC3C is involved in immune responses and closely related to inflammatory activity in the glioma TME. Using the CGGA and TCGA databases, we explored the correlation between APOBEC3C and inhibitory immune checkpoints. We randomly selected common ones such as PDCD1LG2, HAVCR2, CD274, TNFRSF14, CD200R1, CD47, PDCD1, and CTLA4. These checkpoints were strongly positively correlated with APOBEC3C (Figure 6A), indicating that APOBEC3C may suppress immune responses against gliomas.

To further investigate APOBEC3C-related inflammatory activities, we identified 7 gene clusters consisting of 104 genes that represent various inflammatory and immune responses [23]. Correlation matrix plots were used to show the correlation between APOBEC3C and the seven gene clusters in both databases. The results indicated that APOBEC3C expression was correlated with most of the inflammatory responses, with the exception of the immunoglobulin G (IgG) metagenome, which was negatively correlated with APOBEC3C (Figure 6B,C).

3.7 The Relationship Between APOBEC3C and Immune Cell Infiltration in Gliomas

To clarify the mechanism by which high APOBEC3C expression is associated with a poor prognosis in gliomas, the TIMER was used to reveal the association between APOBEC3C and the infiltration levels of six immune cell subtypes. In low-grade gliomas (LGG), APOBEC3C expression was positively correlated with B cells (Cor = 0.596, p = 2.98E−47), CD8⁺ T cells (Cor = 0.312, p = 2.85E−12), CD4⁺ T cells (Cor = 0.659, p = 1.22E−60), macrophages (Cor = 0.682, p = 5.32E−66), dendritic cells (Cor = 0.769, p = 3.14E−94), and neutrophils (Cor = 0.67, p = 2.75E−63) (Figure 7A). Notably, patients with high levels of immune cell infiltration had a lower cumulative survival rate (Figure 7B). However, in GBM, immune cells were less associated with APOBEC3C (Figure 7C,D). Remarkably, there was no significant correlation between APOBEC3C expression and CD8⁺ T-cell infiltration levels in both LGG and GBM, and a clear negative correlation was observed in GBM patients (Figure 7A,C). High expression of APOBEC3C may lead to an increase in macrophages and dendritic cells while decreasing CD8⁺ T cells, thus contributing to the poor cumulative survival rate of glioma patients (Figure 7B). Tumor-infiltrating lymphocytes, such as tumor-associated macrophages (TAMs) and tumor-infiltrating neutrophils, can affect the efficacy of chemotherapy and immunotherapy [24]. APOBEC3C may inhibit the antitumor immune process mediated by T cells by activating macrophages and dendritic cells.

3.8 Relationship Between APOBEC3C and Markers of Immune Infiltration in Gliomas

Infiltrating immune cells in the TME influence the biological behavior of tumors and patient outcomes. The TIMER was used to explore the correlation between APOBEC3C and multiple immune cell markers. Gliomas with APOBEC3C expression presented significant correlations with various immune markers in different immune cells and T-cell subsets (Table 2). In gliomas, APOBEC3C was positively correlated with markers of monocytes (Figure 8A,D), TAMs (Figure 8B,E), and M2 macrophages (Figure 8C,F). APOBEC3C is involved in regulating macrophage polarization in gliomas. Moreover, APOBEC3C has been associated with several markers of regulatory T cells (Tregs) and T-cell exhaustion, such as TIM-3. Depletion of Tregs and T cells is crucial for tumor immune evasion. Therefore, APOBEC3C may play an immunomodulatory role in gliomas. Significantly, T-cell immunoglobulin mucin 3 (TIM-3), a key gene regulating T-cell exhaustion, was positively correlated with APOBEC3C expression. This shows that high APOBEC3C expression is related to TIM-3-mediated T-cell exhaustion. Furthermore, it was confirmed that APOBEC3C is specifically associated with infiltrating immune cells in gliomas, suggesting that APOBEC3C plays a vital role in the immune evasion mechanism of gliomas.

3.9 APOBEC3C Is Enriched in T Cell in GBM

APOBEC3C is associated with various immune-infiltrating cell types in the glioma TME. We utilized single-cell scRNA-seq to analyze APOBEC3C expression levels in different cell clusters in GBM. Using GBM single-cell sequencing data, the R software “UMAP” package segregated the GBM single-cell sequencing data into 15 cell clusters and identified the differentially expressed genes of each cell cluster (Figure 9A). Based on the expression of characteristic markers, the 0th, 6th, and 8th cell clusters were characterized as macrophages, the 1st cell cluster as oligodendrocytes (Figure 9B), and the 9th and 13th cell clusters as T-cells (Figure 9D). The remaining cell clusters were classified as tumor cells (Figure 9C). It was observed that APOBEC3C exhibited high expression in the 9th and 13th cell clusters designated as lymphocytes (Figure 9E). Subsequently, lymphocytes were isolated for further analysis, revealing that they could be further divided into 6 cell clusters (Figure 9F), with APOBEC3C exhibiting consistency with CD8⁺ T cell markers (Figure 9G,I), but not CD4⁺ T cell markers (Figure 9H). These findings confirm the previous hypothesis that APOBEC3C is mediated by CD8⁺ T cells in the TME of GBM patients, and further influences the development of tumor cells and worsens prognosis. Additionally, we discovered that cytotoxic genes (GZMA, GZMB, PRF1, and GNLY) were highly expressed in two of these cell clusters (Figure 9J).

3.10 The APOBEC3C Gene Is an Independent Prognostic Factor in the Survival of Patients With Gliomas

Based on the CGGA and TCGA databases, we conducted Kaplan–Meier (K–M) and Cox proportional hazards model analyses to assess the prognostic significance of APOBEC3C in gliomas. Patients with high APOBEC3C expression exhibited significantly shorter OS compared to those with low expression in the CGGA database (Figure 10A). The TCGA database further confirmed the prognostic value of APOBEC3C (Figure 10D), consistently indicating a poor prognosis for patients with elevated APOBEC3C expression. To mitigate the impact of tumor grade, we reanalyzed the prognostic outcomes for patients with gliomas of varying grades (Figure 10B,C,E,F). Notably, APOBEC3C emerged as an independent prognostic factor in the Cox proportional hazards model analysis (Tables 3 and 4).

3.11 APOBEC3C Inhibits the Malignant Progression of Glioma via the NF-κB Signaling Pathway

To validate the effect of APOBEC3C on the malignant progression of gliomas, we analyzed the correlation between APOBEC3C and the EMT process as well as its representative genes. It was found that APOBEC3C had a significant correlation with the EMT process (Figure 11A). Furthermore, transwell experiments demonstrated that knocking down APOBEC3C using siRNA technology under in-vitro conditions could significantly inhibit the migration and invasion abilities of glioma cells (Figure 11F). Similarly, cell-scratch experiments showed that knocking down APOBEC3C could inhibit the horizontal migration ability of glioma cells U87-MG (Figure 11H). Further western blot experiments indicated that knocking down APOBEC3C could regulate the expression levels of EMT-related proteins (Figure 11B). Finally, to verify the results of GO enrichment analysis, we investigated the effect of APOBEC3C on the common NF-κB signaling pathway. Western blot results showed that knocking down APOBEC3C could significantly inhibit the activation of the NF-κB signaling pathway, thus exerting antitumor effects (Figure 11D). Not only that, but we also obtained similar results when using the U251 glioma cell line (Figure 11C,E,G,I).