To explore the association of SIGLEC7 with cancer, we employed GEPIA (Gene Expression Profiling Interactive Analysis) to compare SIGLEC7 expression levels across various tumors and their respective normal tissues. Our findings revealed a consistent elevation in SIGLEC7 expression across multiple tumor types, notably in LGG and GBM tissues, indicating a potential oncogenic role of SIGLEC7 in gliomas ( Supplementary Figure 1 ). Based on the TCGA and CGGA databases, we analyzed the relationship between the expression level of SIGLEC7 and clinicopathological features of glioma patients, including histological diagnosis, WHO grade, gender, age, IDH mutation status, 1p/19q co-deletion status, MGMT promoter methylation status, and pathological subtypes, and described the patients’ relevant survival outcomes ( Figures 1A, B ). Contrasting these different groupings in the two datasets revealed that these results were not randomly distributed; notably, SIGLEC7 expression was significantly upregulated in higher-grade gliomas ( Figures 1C, D ). In clinical practice, IDH mutation status, 1p/19q co-deletion status, and MGMT promoter methylation status are of significant importance in predicting prognosis, treatment response, and clinical management of gliomas (15). In samples without 1p/19q co-deletion ( Figures 1E, F ), without MGMT promoter methylation ( Figures 1G, H ), and in IDH wild-type glioma patients ( Figures 1I, J ), we also found that patients had higher SIGLEC7 levels compared to the control group. Although the expression differences in these groupings in the CGGA database were not as pronounced as in the TCGA database, the trends were consistent between the two. Furthermore, we analyzed the survival outcomes of patients with different levels of SIGLEC7 expression ( Figures 1A, B ). The data showed that with increasing levels of SIGLEC7 expression, patient overall survival decreased, which was also observed in the CGGA database. In summary, this reveals that high expression of SIGLEC7 is associated with higher malignancy and poorer prognosis in glioma patients.

Glioma transcriptional subtypes, analyzed based on transcriptomic data, provide more accurate molecular classifications and guide prognosis assessment and personalized treatment for gliomas, particularly the mesenchymal subtype, which suggests a malignant prognosis and has been widely recognized worldwide (16). Hence, we explored SIGLEC7 distribution across various subtypes utilizing the TCGA and CGGA databases. Our analysis revealed a notable upregulation of SIGLEC7 in the mesenchymal subtype compared to other subtypes ( Figures 2A, B ). To validate the specificity of our results, we utilized receiver operating characteristic (ROC) curves for evaluation ( Figures 2C, D ). As anticipated, the expression of SIGLEC7 in the mesenchymal subtype reached a significant area under the curve (AUC) value of 94.6% (P <0.0001) in the TCGA database, and 87.1% (P <0.0001) in the CGGA database. Therefore, we have reason to believe that the high expression of SIGLEC7 in the mesenchymal subtype of glioma could serve as one of the indicators for diagnosing the mesenchymal subtype of glioma. Additionally, these results align with our previous findings regarding the association of SIGLEC7 with the malignancy of glioma, further validating that high expression of SIGLEC7 predicts a poorer prognosis for patients.

To explore the association between SIGLEC7 expression and biological functions in glioma, we conducted Pearson correlation analysis using TCGA and CGGA databases, identifying genes most positively correlated with SIGLEC7 expression (Pearson R > 0.5, p < 0.0001). In TCGA, 1464 genes were selected, and in CGGA, 687 genes were selected ( Supplementary Tables 1 and 2 ). Subsequently, we performed gene ontology and KEGG pathway analyses based on these gene sets. In TCGA, biological processes ranked by significance from high to low were inflammation response, immune response, innate immune response, signal transduction, and positive regulation of tumor necrosis factor production ( Figure 3A ). The most correlated cellular component with SIGLEC7 was extracellular exosome ( Figure 3B ), and the associated molecular function was protein binding ( Figure 3C ). Similar enrichment results were obtained in the CGGA database ( Figures 3E-G ). KEGG analysis in both TCGA and CGGA databases showed close associations between SIGLEC7 and immune system function as well as tumor progression ( Figures 3D, H ). Further analysis by intersecting positively correlated genes from both databases consistently indicated that SIGLEC7 primarily participates in the immune and inflammatory processes of glioma ( Supplementary Figure 2 ). Based on these findings, we speculate that SIGLEC7 predominantly engages in the immune response processes in glioma, consequently influencing its malignant progression.

To further validate our hypothesis, we conducted Gene Set Variation Analysis (GSVA) to compute the enrichment scores of gene sets in both TCGA and CGGA datasets. Specifically, we focused on the enrichment scores of three major categories: regulation of immune response, regulation of immune effector process, and regulation of responses to tumor cells ( Figures 3I-N ). Interestingly, we found that in the TCGA database, although the expression of SIGLEC7 was positively correlated with these enrichment scores, the correlation with inhibitory functions was often higher than that with promotive functions, particularly in the regulation of responses to tumor cells, where the correlation with negative modulation was more significant compared to positive regulation. Similar results were obtained in the CGGA database. Additionally, tumor cells often evade immunity by upregulating immune checkpoint genes (17). We investigated the relationship between SIGLEC7 and some well-known immune checkpoint genes using the TCGA and CGGA databases, including TIM-3, HVEM, CD200R1, CD47, TIGIT, CTLA4, PD-1, and PD-L2 ( Figures 4A, B ). The data revealed a significant positive correlation between these genes and SIGLEC7. Based on these results, we hypothesize that SIGLEC7 may collaborate with other immune checkpoints to collectively negatively regulate the organism’s responsiveness to tumors.

To investigate the impact of SIGLEC7 on the inflammatory response in glioma, we conducted expression analyses of several immunoproteins of significance, including hematopoietic cell kinase (HCK), immunoglobulin G (IgG), interferon, lymphocyte-specific kinase (LCK), major histocompatibility complex (MHC), and signal transducer and activator of transcription 1/2 (STAT1/2) (18). According to the results ( Figures 4C, D ), SIGLEC7 exhibited a positive correlation with various inflammatory proteins associated with glioma, while showing a negative correlation with B cell IgG secretion. Additionally, to observe the impact of SIGLEC7 on immune cell infiltration in glioma, we imported data from two databases into the TIMER2.0 database for analysis and stratified patients into high and low expression groups based on the median expression of SIGLEC7. The results indicated a significant positive correlation between high SIGLEC7 expression groups and infiltration of M2-type macrophages in both databases ( Figures 5A, B , Supplementary Figure 3 ). M2-type macrophages have been widely demonstrated to play crucial roles in suppressing immune responses and promoting tumor proliferation (19). Based on these findings, we conclude that upregulation of SIGLEC7 is significantly correlated with infiltration of M2-type macrophages, while also inhibiting certain functions of B cells. Therefore, we hypothesize that SIGLEC7 may exert immunosuppressive effects in glioma.

Considering the significant positive correlation between SIGLEC7 and infiltration of M2 macrophages, we conducted single-cell sequencing analysis of different glioma patient types in the GSE131928, CGGA, and GSE89567 databases to explore their substantive connection. Initially, we performed simple clustering of glioma cells and conducted positional analysis using several typical markers. Among them, CD68, TMEM119, and CD163 are typical markers for macrophages, microglia, and M2 macrophages, respectively (20–22). During the analysis, we compared their distribution with that of SIGLEC7 and found a noticeable consistency between the distribution of SIGLEC7 and CD163, with this result being consistently observed across the three databases ( Figures 6A-C ). Additionally, using the Single Cell Portal database (GSE131928), we also observed predominant expression of SIGLEC7 in macrophages ( Supplementary Figure 4 ). Based on this analysis and in conjunction with the preceding discussion, we hypothesize that the high expression of SIGLEC7 may lead to a predisposition of macrophages towards polarization into M2 macrophages, thereby altering the immune microenvironment. This alteration, through the immunosuppressive action of M2 macrophages, promotes the malignant progression of gliomas.

To confirm our previous speculations, we conducted relevant organizational and cellular experiments. Utilizing immunohistochemistry staining, we investigated the expression of SIGLEC7 in glioma tissues, with results consistent with RNA sequencing data, indicating an enriched expression of SIGLEC7 in higher-grade glioma tissues ( Figure 7A ). THP-1 cells are commonly used to simulate in vivo macrophages and tumor microenvironments (23). We successfully silenced the expression of SIGLEC7 and compared it with untreated macrophages. After polarization towards M2-type macrophages, the results showed reduced CD163 expression in the experimental group, seemingly validating our viewpoint on SIGLEC7 promoting M2-type macrophage polarization ( Figures 7B-D ). Subsequently, we co-cultured SIGLEC7-silenced polarized macrophages with glioma cells. Compared to the control group, the invasive ability of glioma cells in the experimental group decreased ( Figures 7E, F ).

Based on our research, we can confidently confirm a significant correlation between the expression of SIGLEC7 and poor prognosis in gliomas. Therefore, we further analyzed the prognostic predictive value of SIGLEC7 in glioma patients. We conducted analyses using Kaplan-Meier survival curves and Cox regression analysis on the TCGA and CGGA databases, yielding consistent results across both databases. The results revealed ( Figures 8A, B ) a clear distinction in survival curves when patients were grouped based on SIGLEC7 expression levels using the median cutoff. Patients with high SIGLEC7 expression exhibited markedly shorter overall survival, a trend corroborated across different grades of glioma patients ( Supplementary Figure 5 ). Additionally, in Cox regression analysis, we included SIGLEC7 expression levels, WHO grade, age at diagnosis, IDH status, 1p/19q co-deletion status, and MGMT promoter methylation status as factors in both univariate and multivariate survival analyses, further confirming the significant prognostic impact of SIGLEC7 expression levels in glioma patients ( Figures 8C, D ). Finally, utilizing the TCGA database, we constructed a survival prediction model for glioma patients. This model incorporated SIGLEC7 expression levels, WHO grade, IDH mutation status, 1p/19q status, and patient age, demonstrating the probabilities of patient survival at 1, 2, 3, and 5 years under various conditions ( Figure 9A ). At the predictive level, the model’s Harrell’s C-index exceeded that of other factors within the model ( Figure 9B ), and further validation of the model’s predictive accuracy was achieved through calibration curves using data from both databases ( Figure 9C ).

This study collected all selected cases of WHO grade II to IV gliomas from The Cancer Genome Atlas (TCGA) database (702 cases) (http://cancergenome.nih.gov/) and the Chinese Glioma Genome Atlas (CGGA) database (693 cases) (http://www.cgga.org.cn). All clinical information of patients was presented in Table 1 . In the analysis process, cases with incomplete information were handled by treating them as missing data. Single-cell data were obtained from the CGGA database as well as the Gene Expression Omnibus (GEO) databases GSE89567 and GSE131928 (https://www.ncbi.nlm.nih.gov/geo/). Samples for immunohistochemistry were obtained from Sanbo Brain Hospital Capital Medical University with approval from the hospital’s board/ethics committee.

Pearson correlation coefficient was utilized to select the gene list most closely related to SIGLEC7. Subsequently, this list was uploaded to https://david.ncifcrf.gov/ for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, with results visualized using heatmaps.

Biological process gene sets were downloaded from the GSEA website (https://www.gsea-msigdb.org/gsea). The enrichment scores for each biological process per patient were calculated using the default parameters of the R package GSVA. Interested biological processes were visualized using correlation scatter plots.

The patient mRNA expression matrices were uploaded to the online analysis platform TIMER2.0 (http://timer.cistrome.org) for analysis. Utilizing the CIBERSORT algorithm, the abundance of 22 immune cell types infiltrating the tumors was obtained. Subsequently, these results were visualized using box plots and bar graphs.

Three glioma patient databases were subjected to cell clustering analysis using the R package Seurat. Initially, poor-quality cells and genes were filtered out, followed by batch effect removal and data normalization. Dimensionality reduction of cells was achieved using Uniform Manifold Approximation and Projection (UMAP), with cluster results visualized in scatter plots.

Glioma tissue sections, with a thickness of four micrometers, were first immersed in Tris-EDTA buffer at pH 9.0. Antigen retrieval was carried out by heating the sections in a water bath at 100°C for 15 minutes. Endogenous peroxidases were then blocked using a 3% hydrogen peroxide solution. Following this, goat serum was applied for 30 minutes to prevent nonspecific binding. The sections were then treated with the appropriate concentration of SIGLEC7 antibody and co-incubated overnight. Finally, secondary antibody staining, conjugated with HRP, was performed, and the results were observed using an optical microscope.

The THP-1 cells were cultured at 37°C in a 5% CO2 atmosphere in RPMI1640 medium supplemented with 10% fetal bovine serum, 1% penicillin-streptomycin, and 0.1 mM β-mercaptoethanol. To induce differentiation into M0 macrophages, cells were treated with 100 ng/mL phorbol myristate acetate (PMA) for 48 hours. Subsequently, polarization into M2 macrophages was achieved by co-incubation with 20 ng/ml IL-4 and 20 ng/ml IL-13 in the presence of PMA. Concurrently, siRNA-NC and siRNA-SIGLEC7 were added to the control and experimental groups, followed by a 48-hour co-incubation.

After polarization and siRNA transfection, total protein was extracted from THP-1 cells. The proteins were quantified using the BCA method and then denatured by boiling with 5x loading buffer. Following this, protein samples were subjected to electrophoresis for separation and transferred onto membranes. These membranes were then blocked with 5% skim milk at room temperature for 2 hours. Subsequently, they were incubated overnight at 4°C with primary antibodies against SIGLEC7 (immunoway, YT5281, 1:1000), CD163 (Abcam, ab156769, 1:1000), and GAPDH (immunoway, YM3029, 1:5000). After washing with TBST buffer, the membranes were incubated with appropriate secondary antibodies, and chemiluminescence was detected using the Bio-Rad imaging system.

Cells were fixed with 4% paraformaldehyde for 20 minutes after multiple washes with PBS, permeabilized with 1% Triton X-100 for 10 minutes, and then blocked with goat serum for 20 minutes. Primary antibody solution containing SIGLEC7 (immunoway, YT5281, 1:200) and CD163 (Abcam, ab156769, 1:200) was added and incubated overnight at 4°C. Subsequently, cells were incubated with secondary antibodies (Goat anti-Mouse IgG (H+L) Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 488, A11001, 1:200; Goat anti-Mouse IgG (H+L) Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 594, A32740, 1:200) for 1 hour at room temperature. After three PBS washes, cells were mounted with Prolong™ Diamond Antifade Mountant containing DAPI (Invitrogen, P36962), and images were captured using confocal microscopy.

Transwell invasion assays were performed using 8μm polycarbonate membrane inserts pre-coated with Matrigel (Corning) in a 24-well plate. LN229 cells (1×10^5 cells/100μl) were seeded in the upper chamber, while M2 cells (1×10^5 cells/500μl) were added to the lower chamber containing 2% FBS medium. After 10 hours, non-invading cells were removed, and the invaded cells were fixed with 4% PFA and stained with 1% crystal violet. Images were captured using a microscope after cleaning the inserts.

This study primarily utilized R software (version 4.2.3, Windows platform) and IBM SPSS Statistics (version 27.0.1, Windows platform) for statistical analysis and chart generation. Various R packages were employed including ggplot2, pheatmap, circlize, GSVA, ggpubr, survminer, survival, corrgram, dplyr, Seurat, patchwork, and pROC for data visualization and analysis. Additionally, IBM SPSS Statistics was utilized for both univariate and multivariate Cox regression analysis. Pearson correlation analysis was conducted to assess the correlation between two sets of data. The significance of differences between two groups was evaluated using Student’s t-test, while one-way analysis of variance (ANOVA) was employed to assess differences among more than two groups. A p-value < 0.05 was considered statistically significant.