Gliomas, the most common malignant tumours of the nervous system, have a poor prognosis, and conventional surgery and postoperative radiotherapy improve five-year survival, but still do not achieve satisfactory results[5]. Therefore, the search for new molecular targets is particularly important for the treatment of gliomas as well as for prognostic prediction.DIRAS3 is a tumour suppressor gene encoding a 26 kD GTPase, and DIRAS3 down-regulation occurs in ovarian, breast, lung, prostate, colon, brain, and thyroid cancers.DIRAS3 re-expression inhibits signal transduction through the PI3 kinase/AKT, JAK/STAT, and RAS/MAPK inhibitors, blocking signal transduction via PI3 kinase/AKT, JAK/STAT, and RAS/MAPK. MAPK to inhibit signal transduction, block malignant transformation, inhibit cancer cell growth and viability, and prevent angiogenesis.DIRAS3 is critical for autophagy and triggers this process through multiple mechanisms.Re-expression of DIRAS3 induces dormancy in nude mice/nude mice ovarian cancer xenograft models, inhibiting cancer cell growth and angiogenesisDIRAS3-mediated autophagy induction promotes the survival of dormant cancer cells in a nutrient-poor environment Survival of dormant cancer cellsDIRAS3 expression in dormant, drug-resistant autophagic cancer cells could serve as a biomarker and a target for new therapies to eliminate residual tumours after conventional therapies[29].
In gliomas, the role of DIRAS3 is not clear. We found that high DIRAS3 expression was associated with age, grade, IDH status, and 1p/19q deletion, but not gender.DIRAS3 expression was strongly correlated with tumour grade and increased with tumour grade.DIRAS3 expression was strongly correlated with OS (P < 0.001), progression-free interval (PFI) (P < 0.001), and disease-specific survival (DSS ) (P < 0.001).DIRAS3 expression was also strongly associated with race (P < 0.05), histological type (P < 0.01) and primary treatment outcome (P < 0.01)(Fig. 2–3). Univariate and multivariate Cox regression analyses showed that age, WHO class, primary treatment outcome, and DIRAS3 were independent risk factors for low OS. In addition, among the 3381 DIRAS3-related differentially expressed genes screened in the TCGA database with |logFC | >1.5, corrected P-value < 0.01, SAA1, which had the highest positive correlation with DIRAS3, is a sensitive acute-phase high-density lipoprotein mainly produced by the liver as a response to acute inflammation and tissue injury in humans, assessing the acute-phase human inflammatory response process, and SAA1 may be involved in the body's immune system to promote the repair of damaged tissues, as well as being used as a diagnostic or prognostic marker for many diseases [30], which is consistent with our results. The search for DIRAS3-binding proteins by STRING and the aforementioned intersection cross-tabulation analysis yielded three common members:MET, PLEKHG4B and MAP3K19.The MET receptor binds to its ligand, HGF, and induces MET dimerisation, tyrosine residue autophosphorylation, substrate docking, and activation of downstream signalling pathways[31].The aberrant MET/HGF axis pathway is involved in the proliferation, survival, invasion and metastasis of tumour cells [31].PLEKHG4B, a Rho-Guanine Nucleotide Exchange Factor (Rho-GEF), is localised to cell junctions and is involved in cell-cell junction maturation by decreasing myosin activity at the late stages of cell-cell junction formation[32] .The role of MAP3K19 is now less well studied, and has been reported to be associated with cancer progression associated with cancer progression [33]. Therefore, DIRAS3 may be involved in the regulation of SAA1, MET, PLEKHG4B and MAP3K19 to alter the tumour immune microenvironment and thereby promote glioma progression(Figure1E).
To further explore the role of DIRAS3 in gliomas, in Fig. 7 we analysed by GO and KEGG enrichment, the DIRAS3 co-expressed genes appeared to be mainly involved in humoral immune response, complement activation, classical pathway, phagocytosis, recognition, circulating immunoglobulin-mediated humoral immune response, B-cell-mediated immunity, immunoglobulin complexes, immunoglobulin complexes, circulation, plasma membrane outer, blood particles, T-cell receptor complex, antigen binding, immunoglobulin receptor binding, receptor ligand activity, signalling receptor activator activity and cytokine activity.KEGG analysis showed that DIRAS3 co-expressed genes were enriched in cytokine-cytokine receptor interactions, systemic lupus erythematosus, viral proteins interacting with cytokines and cytokine receptors, S. aureus staphylococcal infection and haematopoietic cell lines. Tumour cells can functionally shape their microenvironment by secreting various cytokines, chemokines and other factors. This leads to reprogramming of the surrounding cells so that they play a decisive role in tumour survival and progression. Immune cells are important components of the tumour stroma and participate in this process [34]. GSEA was next applied to identify DIRAS3-related signalling pathways in the high and low DIRAS3 expression groups, as shown in Fig. 8 and Supplementary Table S4, 14 signalling pathways were highly enriched in the high DIRAS3 expression phenotype, antigen processing and presentation, cytokine-cytokine receptor interactions, JAK-STAT signalling pathway, natural killer cell-mediated cytotoxicity, NOD-like receptor signalling pathway, TOLL-like receptor signalling pathway, anti-inflammatory response to Leishmania parasiticum infection, base excision repair, FCERI-mediated activation of NF-kB, neutrophil degranulation, interleukin signalling and IL18 signalling pathway. This further suggests that DIRAS may promote glioma progression by altering the tumour immune microenvironment.
Interestingly, we also found a strong correlation between the degree of immune cell infiltration and DIRAS3 expression by GSEA analysis of individual samples (Fig. 9A).DIRAS3 expression was positively correlated with immune activation-associated cells (e.g., macrophages, eosinophils, neutrophils, activated dendritic cells, T-cells, and immature dendritic cells) (Fig. 9B-G). There is growing evidence that innate immune cells (macrophages, neutrophils, dendritic cells, innate lymphocytes, myeloid-derived suppressor cells and natural killer cells) as well as adaptive immune cells (T and B cells) are involved in tumour progression when the tumour microenvironment (TME) is present [6]. This suggests that DIRAS3 may promote immune cell infiltration abundance and thus glioma progression. Furthermore, we found that DIRAS3 was significantly correlated with the expression levels of genetic markers in monocytes and macrophages (including M1 macrophages, M2 macrophages and tumour-associated macrophages) (Table 1), and that in healthy tissues, macrophages usually express a mixed M1/M2 phenotype; thus, "M1" and "M2" polarisation is considered to be the extreme endpoint of a continuum of activation states, the precise extent of which depends on the precise mixing of local signals in a given microenvironment; in summary, TAMs enhance tumour growth and progression[35].
Based on these findings, we speculated whether DIRAS3 could be used as a diagnostic biomarker for glioma prognosis. In the time-dependent ROC curves evaluating the role of DIRAS3 expression in distinguishing glioma from normal tissues, the AUC values predicting 1-, 2-, 3-, 4-, and 5-year survival were 0.841, 0.851, 0.841, 0.805, and 0.803, respectively (Fig. 5A-F). These results suggest that DIRAS3 expression is a valid biomarker for the diagnosis of glioma. As shown in Fig. 5G-H and Supplementary Table S2, univariate and multivariate Cox regression analyses showed that age, WHO classification, primary treatment outcome and DIRAS3 were independent risk factors for poor OS. In addition, Fig. 5J showed that DIRAS3 accurately predicted 1-, 3-, and 5-year survival. In conclusion, our findings suggest that high expression of DIRAS3 is a risk factor for poor prognosis in glioma patients.
In conclusion, elevated expression of DIRAS3 in gliomas may enhance glioma development by regulating the levels of proteins such as SAA1, MET, PLEKHG4B, and MAP3K19 to increase the infiltration of immune cells and alter the tumour immune microenvironment in gliomas, and may serve as a diagnostic marker positively correlating with poor prognosis in glioma patients.