The human HA1800 and, 1321N1 cell lines were acquired from the European Collection of Cell Culture. They were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (04-001-1Acs, BI) and 1% streptomycin and penicillin (15140122, Gbico), and maintained at 37°C in a humidified atmosphere composed of 95% air and 5% CO2, with medium changes every other day.

RNA was extracted from the cells as described previously (10). The extracted RNA was then reverse-transcribed to cDNA using the RT Kit (HY-K0511A, MedChemExpress). Subsequently, the SuperReal PreMix Plus was used for PCR analysis and run in the Real-Time PCR System. We used Primer Premier 5.0 program to design the TBC1D1 primer. TBC1D1: Forward: 5’-CCTGCGCTACATCACACCC-3’ and reverse: 5’-CATGCGGTCTGGAACACTC-3’ and amplicon size was 175 bp. GAPDH: 5’-TGTGGGCATCAATGGATTTGG-3’ (forward) and 5’-ACACCATGTATTCCGGGTCAAT-3’ (reverse) and amplicon size was 116 bp (Sangon). The resulting data were normalized to GAPDH expression via the CFX Manager software (version 3.0), facilitating the evaluation of TBC1D1 mRNA expression levels.

Protein extraction was performed using RIPA lysis buffer (p0013B, Beyotime), followed by separation through 12% SDS-PAGE electrophoresis and transfer onto PVDF membranes. Thereafter, the membranes underwent blocking with 5% nonfat milk, succeeded by treatment utilizing either the TBC1D1 antibody (22124-1-AP, Proteintech) or the β-actin antibody (66009-1-Ig, Proteintech), and subsequently, on the following day, exposure to an HRP-linked anti-rabbit IgG antibody (BL003A, Biosharp). The visualization of the protein bands was attained using ECL reagents (34577, Thermo Fisher), and Image J software was employed to calculate the optical densities to determine TBC1D1 protein expression levels.

To acquire bone marrow cells, femurs were harvested from 8- to 10-week-old C57BL/6 mice (purchased from Ensiweier) and red blood cells were lysed using erythrocytes lysate (BL503A, Biosharp). Cell culture involved the use of complete DMEM medium (HyClone) with 20% L929 cell conditioned medium at 37°C and 5% CO2. The medium was refreshed on day 4 to obtain mature M0 macrophages by day 7, while M2 macrophages were induced by IL4 protein (1 ng/mL, HY-P70653, MedChemExpress), M1 macrophages by lipopolysaccharide stimulation, and MDSC differentiation was initiated by IL6 protein (5 ng/mL, HY-P7063, MedChemExpress). Single cell suspensions were obtained from spleens collected from wild-type C57BL/6 following erythrocyte lysis. T cell activation was achieved using anti-CD3 antibody (2.5 mg/mL, 145-2C11, BioLegend) and anti-CD28 antibody (3 mg/mL, 102102, BioLegend). Lastly, fibroblasts were isolated from tumors harvested from wild-type C57BL/6 mice, and the expression levels of TBC1D1 were confirmed in each of these cell types.

To explore the potential association between TBC1D1 expression and overall survival (OS), we retrieved data from several databases: CGGA (http://www.cgga.org.cn), TCGA (https://www.cancer.gov/ccg/research/genome-sequencing/tcga), GTEx (https://www.genome.gov/Funded-Programs-Projects/Genotype-Tissue-Expression-Project), and GEO (https://www.ncbi.nlm.nih.gov/geo). These databases provided access to expression matrices of the TBC1D1 gene as well as clinical information across normal (N = 1152) and tumor (N = 523) tissues. Prior to analysis, we standardized the data using R (version 4.3.0). Subsequently, we utilized the limma package to identify genes with differential expression linked to TBC1D1 expression, applying a threshold of |log2 FC| ≥ 1 and an adjusted P-value (FDR) of < 0.05.

The Human Protein Atlas (HPA) database, established in Sweden in, 2003, endeavors to comprehensively map all human proteins within cells, tissues, and organs by employing a variety of omics technologies including antibody-based imaging, mass spectrometry-based proteomics, systems biology, and transcriptomics (11, 12). In this study, we leveraged the HPA database to validate the intracellular localization of the TBC1D1 protein, assess its mRNA expression in somatic cells, tissues, and immune cells, and compare its protein expression in normal and tumor tissues.

LASSO analysis was conducted as previously described (10). In this study, we utilized LASSO analysis to build a prognostic model and identify genes associated with prognosis.

In this study, cancer patients were divided into two groups based on TBC1D1 expression. Subsequently, we employed CIBERSORT to assess the levels of immune cell infiltration in the tumor tissues.

Tumor Immune Single-cell (TISCH) database is an available resource facilitates the exploration of the tumor microenvironment (TME) via single-cell RNA sequencing (scRNA-seq) data (13, 14). We analyzed the distribution and expression of TBC1D1 in diverse cell types within the glioma microenvironment using the TISCH.

We employed the GSEA computational method to assess the statistical significance of a preselected gene set (15–18). Following correlation analysis, we generated an initial list of gene categories. These categories were then segmented into various groups for each analysis, involving, 1000 permutations of gene sets to identify any disparities among them. The results of this analysis aided in identifying the critical genetic functions and signaling pathways linked to TBC1D1.

STRING database is dedicated to predicting protein-protein interactions (PPI) and employs computational predictions, cross-organism knowledge transfer, and curated data from other databases to identify both direct and indirect associations (19, 20). In this study, we utilized the STRING database to investigate protein-protein interactions.

Tumor immune dysfunction and exclusion (TIDE) tool is designed to predict transcriptomic biomarkers for immunotherapy response by analyzing the gene expression profile of a tumor before treatment (21–23). In this study, we used the TIDE to explore the association between TBC1D1 expression and T cell function in multiple human cancers, with a specific focus on glioma.

We analyzed the RNA-sequencing data using R (version 4.3.0). To compare the outcomes between the experimental and control groups, we employed a two-tailed Student’s t-test and visualized data by using GraphPad Prism (version 8.4.0). Furthermore, for multiple comparisons, we performed a one-way ANOVA (P < 0.05 = “*”, P < 0.01 = “**”, P < 0.001 = “***”, P < 0.0001 = “****”, P > 0.05 = ns).

Survival data analyses from CGGA and TCGA databases for glioma cancer patients revealed, 2309 and, 7398 differentially expressed genes, respectively. The intersection of these gene sets identified, 1770 genes linked to overall survival (OS) ( Figure 1A ). Among non-responders with drug therapy resistance, 4736 genes showed significant upregulation ( Figure 1B ). Further analysis of the upregulated genes for enrichment in signaling pathways using GSEA revealed a significant enrichment of the PI3K/AKT/mTOR pathway in the tumor tissue of non-responders ( Figure 1C ). The validation of core genes associated with patient prognosis and therapy resistance involved analyzing the common genes from three datasets: OS-related genes, up-regulated genes from non-responders, and those from the PI3K/AKT/mTOR signaling pathway. This analysis revealed fourteen core genes, which were further examined using LASSO regression, indicating that seven critical genes, including ITGA5, TBC1D1, GNG12, ITGA2, GNG5, LAMC1, and OSMR, were significant ( Figures 1D–F ). A random forest tree analysis confirmed the importance of ITGA5 and TBC1D1 as the most critical genes ( Figure 1G ). Interestingly, while ITGA5 showed no impact on patient survival, high expression of TBC1D1 was linked to an unfavorable prognosis ( Figure 1H ). Patients with high levels of TBC1D1 in their tumors exhibited poor therapy efficacy ( Figure 1I ). Overall, TBC1D1 emerged as a prognostic marker positively associated with therapy resistance.

Transcriptome sequencing data from the TCGA and GTEx databases were analyzed to evaluate TBC1D1 expression levels. The findings indicated a notable upregulation of TBC1D1 expression in glioma tissues compared to normal tissues ( Figure 2A ). This pattern was consistent across various human tumors, including ACC, CHOL, DLBC, HNSC, KICH, LAML, LIHC, MESO, OV, PCPG, TGCT, UCS, and UVM ( Supplementary Figure S1A ). Paired sample analysis also confirmed elevated TBC1D1 expression in tumor tissues ( Supplementary Figure S1B ). A recent genomewide mutational analysis of gliomas uncovered somatic mutations in the isocitrate dehydrogenase 1 (IDH1) gene in a subset of these tumors. Similarly, glioma patients with high-TBC1D1 tumors exhibited a higher frequency of mutations in IDH1 ( Figure 2B ). Immunohistochemical analysis validated higher TBC1D1 expression levels in tumor tissues, and a substantial increase in TBC1D1 expression was observed in the glioma cell line, 1321N1 compared to the normal cell line HA1800 ( Figures 2C–G ). In addition, the study identified the concurrent expression of 10 genes alongside TBC1D1 in gliomas, implying their potential contribution to tumor development ( Figure 2H ).

Logistic regression analysis was employed to investigate the relationship between TBC1D1 and various clinicopathological variables using a relevant clinical dataset. The findings revealed a notable elevation in TBC1D1 levels in glioma patients over 40 years old compared to those under 40 years old ( Figure 3A ). Additionally, a significant correlation between TBC1D1 expression and tumor grade was observed, with an increase in expression as the tumor grade advanced from 2 to 3 ( Figure 3C ). Remarkably, varying levels of TBC1D1 were observed in different histological types, with higher expression in astrocytomas compared to oligoastrocytomas or oligodendrogliomas ( Figure 3D ). In contrast, gender showed less significance as a clinical variable ( Figure 3B ). In conclusion, the results suggest a notable correlation between TBC1D1 expression and key clinical variables such as patient age, tumor grade, and histological type.

The developed nomogram revealed a robust association between TBC1D1 and various clinical variables ( Figure 4A ). Specifically, the nomogram demonstrated exceptional predictive accuracy at three distinct time points, with C-indexes and AUC values of 0.817, 0.901, 0.897, and 0.803, respectively. Moreover, the calibration curves for 1-year, 3-year, and 5-year predictions displayed a high level of concordance between the predicted and actual outcomes ( Figures 4B–D ). These findings further validate the dependable performance of the TBC1D1-based nomogram in accurately predicting the prognosis of glioma patients.

Our GSEA analysis revealed strong evidence of a significant correlation between elevated expression of TBC1D1 and pathways related to type I diabetes mellitus, complement and coagulation cascades, allograft rejection, autoimmune thyroid disease, and graft versus host disease signaling ( Figure 5A ). Additionally, gene ontology (GO) analysis highlighted TBC1D1’s primary associations with pattern specification processes in biological functions and the extracellular matrix containing collagen in cellular components. Notably, TBC1D1’s molecular function was significantly linked to DNA-binding transcriptional activator activity and specificity for RNA polymerase II ( Figure 5B ). These findings emphasize the critical role of TBC1D1 in regulating these essential signaling pathways and gene functions in tumor progression.

In our investigation of the relationship between TBC1D1 and the TME, scRNA-seq data from glioma obtained from the GEO database were analyzed. The results revealed high expression of TBC1D1 in immune cells, particularly in macrophage populations ( Figures 6A–C ). Subsequent analysis of macrophage subpopulations showed predominant expression of TBC1D1 in the M2 subpopulation ( Supplementary Figures S2A–I ). Expanding our inquiry, we induced the differentiation of bone marrow-derived monocytes into different subtypes of macrophages and examined the protein and mRNA levels of TBC1D1 in cancer-associated fibroblasts and splenic T cells. Our findings indicated that TBC1D1 was expressed in M2-like macrophages, akin to tumor cells ( Figures 6D–F ). Stratifying glioma patients into two groups based on their TBC1D1 levels, it was observed that patients with elevated TBC1D1 levels exhibited high macrophage infiltration within the TME, while a noticeable negative correlation was observed between TBC1D1 and T cell infiltration levels ( Figures 6G, H ). Importantly, this trend was also observed in other types of gliomas, including glioblastoma multiforme and brain low-grade glioma ( Supplementary Figures S3A, B ). Additionally, TBC1D1 displayed a positive correlation with immunosuppressive macrophage markers such as CD163 and ARG1 ( Figures 6I, J ), suggesting that TBC1D1-positive macrophages could function as an immunosuppressive cell population, facilitating the establishment of an immunosuppressive TME and promoting tumor immune escape, ultimately accelerating tumorigenesis and progression.

An analysis of treatment outcomes in glioma patients from the TCGA database revealed a strong association between high TBC1D1 expression and non-responsive patients ( Figure 7A ). We further investigated the impact of TBC1D1 on treatment tolerance, focusing on its effect on the function of cytotoxic T lymphocytes (CTL), which play a crucial role in eliminating tumor cells. Our findings clearly indicated that high TBC1D1 expression impairs favorable prognosis in the context of high CTL function in tumor tissues, leading to poor survival outcomes among glioma patients. Conversely, low TBC1D1 expression was positively correlated with enhanced CTL function and improved patient survival ( Figure 7B ). This trend was observed in breast, colon, endometrial, melanoma, myeloma, and ovarian cancers ( Supplementary Figures S4A–F ). Our mechanistic study revealed that TBC1D1-mediated inhibition of CTL function could be attributed to the upregulation of various immune checkpoint molecules, including ARG1, CD68, PDCD1, CD274, TGFB1, and CTLA4, collectively impeding the anti-tumor immune response ( Figure 7C ). In animal models receiving immunotherapy in the TISMO database, high TBC1D1 expression significantly hindered the effectiveness of immune checkpoint blockade (ICB) treatment, resulting in an increase in non-responders ( Supplementary Figure S5A ). Analysis of the TIGER database of patients receiving anti-PD-1 immunotherapy for melanoma revealed that TBC1D1 impedes the effectiveness of the treatment, leading to suboptimal patient survival rates ( Supplementary Figure S5B ). Additionally, patients with high TBC1D1 expression exhibited increased infiltration of M2 tumor-associated macrophages (TAMs), showing significant elevation in markers including CD68 and ARG1 ( Figure 7C ). The M2 TAMs signature is a significant risk factor in patients with melanoma treated with immunotherapy ( Supplementary Figure S6A ), and patients with tumors exhibiting high M2 TAMs signature expression had a poor survival probability ( Supplementary Figure S6B ). Therefore, we posited that TBC1D1-positive macrophages could inhibit the effectiveness of immunotherapy in glioma patients, leading to immunotherapy tolerance and poor prognosis. Furthermore, we demonstrated the prognostic value of TBC1D1 using Kaplan-Meier curves ( Figure 7D ) and ROC curves and AUC values at different time points ( Figures 7E, F ). Overall, the study results indicated that macrophages with elevated TBC1D1 levels may play a crucial role in immunotherapy resistance and can act as a compelling biomarker when predicting the chances of survival for patients with gliomas.