In recent years, increasing evidences have shown that metabolic reprogramming is one of the major hallmarks of malignant tumor(42, 43). To meet the excessive energy consumption and biosynthetic demands associated with rapid proliferation, cancer cell generally exhibit metabolic disorders, involving glucose, amino acid, and lipid metabolism et al(44–46). As a membrane constituent of mammalian cells, cholesterol plays an essential role in maintaining the integrity and fluidity of the membrane, which is also indispensable for the survival and proliferation of cancer cells(8, 9). Metabolites of cholesterol, including bile acids, oxysterols, and steroid hormones can innate and promote a variety of human cancers(13, 15). Besides, several studies have revealed that cholesterol accumulated in TME also participate in promoting cancer progression and suppressing immune responses. However, few studies focus on the significance of cholesterol metabolism in LGGs.
In the present study, LGG patients were divided into two clusters based on the expression level of cholesterol metabolism-related genes. Then, we found that the patients in two clusters have different clinicopathological characteristics and clinical outcomes. These results suggested that cholesterol metabolism may play a key role in the malignant progression of glioma and affect the prognosis of patients. Thus, we constructed a cholesterol metabolism-related risk signature to evaluate the functional roles and prognostic value of cholesterol metabolism in LGGs. By investigating the distribution of risk score stratified patients, we found that the risk score was positively correlated with malignant clinicopathologic characteristics. Besides, it is worth mentioning that 7p11.2 contains EGFR, while 10q23.3 contains the PTEN locus. Chr 7 amplification accompanied Chr 10 loss, which eventually leads to EGFR amplification and PTEN loss, was generally considered to be a representative genomic alteration in GBM(47, 48). Therefore, the genomic alteration pattern of LGG samples in high-risk group was similar to GBM, suggesting that cholesterol metabolism might play a key role in the malignant progression of LGGs.
Bioinformatics analysis showed that the risk signature was tightly association with immune and inflammatory response, indicating a tight interaction between cholesterol metabolism and TIME in LGGs. In CIBERSORT analysis, the risk signature was positively correlated with the infiltration of resting memory CD4 + T cells and macrophages in M2 phase, which were considered to promote immunosuppression and malignant progression of tumors(49, 50). Relevant studies pointed out that cancer cells could promote cholesterol efflux in TAMs and polarize TAMs towards an M2-like phenotype, which ultimately accelerates the malignant progression of cancers(51, 52). Moreover, some research has shown that high levels of cholesterol activate the Toll-like receptor (TLR) present on macrophages and generate a chronic inflammation with in the TME that promotes cancer progression(53). In this study, the IHC analysis also suggested that cholesterol metabolism was positively correlated to the infiltration and M2 polarization of macrophages. Extracellular cholesterol accumulation in the TME had been shown to induce T cell functional exhaustion by promoting several well-known immune checkpoints, such as PD-1, TIM-3, 2B4, and LAG-3(19). In addition, 22-hydroxycholesterol (22HC), a metabolite of cholesterol, could inhibit T cell-mediated anti-tumor immunity by activating liver X receptors (LXRs) signaling pathway(54). More importantly, cholesterol metabolism could facilitate CD4 + and CD8 + T-cell differentiation and acquisition of effector function(55, 56). In the present study, the GSVA analysis also showed that enhanced cholesterol metabolism played an inhibitory role in T cell-mediated antitumor immune response in gliomas. Beyond TAMs and T cells, 22HC, 24-hydroxycholesterol (24HC) and 27-hydroxycholesterol (27HC) could also recruit neutrophils, which are emerging as an important immunosuppressive population in the TME(57–59). In brief, cholesterol metabolic status had significant impact on TIME and this finding provided an important reference for further improve the efficacy of immunotherapy of LGGs.
To investigate the prognostic value of this risk signature, we performed K-M survival and COX regression analyses in both TCGA and CGGA cohorts. Even if we divided patients into different groups according to molecular pathological characteristics, the risk signature still showed high prognostic value in each group. For IDH mutant patients with 1p/19q codeletion in CGGA cohort, the K-M survival analysis showed no statistical difference, which was mainly due to the small number of samples and short follow-up time. In brief, the results suggested that the risk signature was an independent prognostic factor for LGG patients. In this risk signature, each single gene was key mediator in the cholesterol metabolic signaling pathways and could serve as a prognostic factor of LGG patients. However, by constructing a risk signature, we could estimate the cholesterol metabolic status and predict prognosis of LGG patients more precisely. To improve the clinical application value of the risk signature, we also developed a nomogram model to predict 1-, 3- and 5-year survival of LGG patients. Compared with other LGG prognostic models proposed by previous studies(60, 61), the cholesterol metabolism-related signature showed similar prognostic value while accurately assessing cholesterol metabolism of gliomas. Therefore, the cholesterol metabolism-related signature was a reliable and robust prognostic indicator.
Because of the vital functions of cholesterol metabolism in cancer imitation and progression, impeding active cholesterol metabolism became a feasible treatment strategy and caused extensive concern(16). For example, HMG-CoA reductase (HMGCR) inhibitor, which was the most wildly used anti-cholesterol drug, has been proved to decrease mortality and prolong survival of patients with colorectal or prostate cancer(62–64). In addition, inhibitors of Acyl Coenzyme A: Cholesterol Acyltransferases 1 (ACAT1), which also known as sterol O-acyltransferase 1 (SOAT1), could suppress tumor growth by downregulating cholesterol esterification (CE) in leukemia and triple-negative breast cancer(65, 66). In the present study, as the most important contributor of the risk signature (Regression coefficients = 0.2773), the biological function of SOAT1 were also investigated in glioma cell lines. The results showed that silencing SOAT1 expression suppressed the proliferation and migration capacity of glioma cells, which may provide novel target for the treatment of gliomas. Interestingly, both agonists and inverse agonists of LXR have shown significant efficacy for the treatment of multiple human cancers, mainly by inhibiting cell proliferation and regulating anti-tumor immune responses(67–69). Based on findings of our study, we hypothesized that the eight genes in the cholesterol metabolism-related signature might also serve as promising therapeutic targets of LGGs. However, these conjectures need more experimental verification.
In brief, through comprehensively analyzing clinical and transcriptomic profiling data in TCGA and CGGA cohorts, we identified a cholesterol metabolism-related signature to evaluate the cholesterol metabolic status and predict the prognosis of patients with LGGs. Moreover, we found that cholesterol was involved in macrophage- and T cell-mediated antitumor immunity. Finally, the risk signature was revealed to be an independent prognostic indicator of LGG patients. Although, more in-depth exploration and prospective cohort studies were needed to estimate the significances of cholesterol metabolism in LGGs. These findings shed light on functional roles of cholesterol metabolism in LGGs and provided promising targets to enhance the anti-tumor therapies.