Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Nov 17:13:1072253.
doi: 10.3389/fphar.2022.1072253. eCollection 2022.

Serine and glycine metabolism-related gene expression signature stratifies immune profiles of brain gliomas, and predicts prognosis and responses to immunotherapy

Siliang Chen  1 , Shuxin Zhang  1   2 , Wentao Feng  1 , Junhong Li  3 , Yunbo Yuan  1 , Wenhao Li  1 , Zhihao Wang  1 , Yuan Yang  1 , Yanhui Liu  1
Affiliations

Serine and glycine metabolism-related gene expression signature stratifies immune profiles of brain gliomas, and predicts prognosis and responses to immunotherapy

Siliang Chen et al. Front Pharmacol. .

Abstract

Glioma is one of the most lethal cancers and causes more than 200,000 deaths every year. Immunotherapy was an inspiring therapy for multiple cancers but failed in glioma treatment. The importance of serine and glycine and their metabolism has been well-recognized in the physiology of immune cells and microenvironment in multiple cancers. However, their correlation with prognosis, immune cells, and immune microenvironment of glioma remains unclear. In this study, we investigated the relationships between the expression pattern of serine and glycine metabolism-related genes (SGMGs) and clinicopathological features, prognosis, and tumor microenvironment in glioma based on comprehensive analyses of multiple public datasets and our cohort. According to the expression of SGMGs, we conducted the consensus clustering analysis to stratify all patients into four clusters with remarkably distinctive clinicopathological features, prognosis, immune cell infiltration, and immune microenvironment. Subsequently, a serine and glycine metabolism-related genes signature (SGMRS) was constructed based on five critical SGMGs in glioma to stratify patients into SGMRS high- and low-risk groups and tested for its prognostic value. Higher SGMRS expressed genes associated with the synthesis of serine and glycine at higher levels and manifested poorer prognosis. Besides, we confirmed that SGMRS was an independent prognostic factor and constructed nomograms with satisfactory prognosis prediction performance based on SGMRS and other factors. Analyzing the relationship between SGMRS and immune landscape, we found that higher SGMRS correlated with 'hotter' immunological phenotype and more immune cell infiltration. Furthermore, the expression levels of multiple immunotherapy-related targets, including PD-1, PD-L1, and B7-H3, were positively correlated with SGMRS, which was validated by the better predicted response to immune checkpoint inhibitors. In conclusion, our study explored the relationships between the expression pattern of SGMGs and tumor features and created novel models to predict the prognosis of glioma patients. The correlation of SGMRS with immune cells and microenvironment in gliomas suggested an essential role of serine and glycine metabolism in reforming immune cells and microenvironment. Finally, the results of our study endorsed the potential application of SGMRS to guide the selection of immunotherapy for gliomas.

Keywords: glioma; glycine; immune checkpoint inhibitor; immune infiltration; metabolism; prognosis; serine; tumor microenvironment.

PubMed Disclaimer

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Figures

FIGURE 1
FIGURE 1
Clustering of gliomas based on expression pattern of SGMGs. (A) tSNE map for SGMGs expression patterns of four consensus clusters. (B) Heatmap for expression of 21 SGMGs based on four clusters. (C) The expression levels of PHGDH, PSAT1, PSPH, and SHMT1 among four clusters. (D) K-M curves based on four consensus clusters in (D) TCGA, (E) CGGA325, (F) CGGA693, and (G) WCH cohorts.
FIGURE 2
FIGURE 2
Functional enrichment and clinicopathological characteristics of the four consensus clusters in TCGA cohort. (A) Top five pathways with the highest NES in the REACTOME gene set between cluster 1 and cluster 4. (B) Top five pathways with the highest NES in the KEGG gene set between cluster 1 and cluster 4. (C) Top 20 mutated genes of the four consensus clusters. (D) Heatmap for copy number variations of the four clusters. (E) The differences in (E) tumor grade, (F) histological diagnosis, (G) MGMT promoter status, and (H) TERT promoter status among four clusters. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
FIGURE 3
FIGURE 3
Different immunological landscapes of tumor microenvironment among four clusters. (A) Boxplots for infiltration fraction of four types of immune cells based on CIBERSORTx in TCGA cohort. (B) Differences in stromal, immune, and ESTIMATE scores among four clusters in TCGA cohort. (C) Difference in tumor purity among four clusters in TCGA cohort. (D) TIP score and related gene expression heatmap among four clusters in TCGA cohort. (E) Difference in TIP score among four clusters in TCGA cohort. (F) TIP score and related gene expression heatmap among four clusters in CGGA325 cohort. (G) Difference in TIP score among four clusters in CGGA325 cohort. (H) Differences in expression levels of CD274 and CD276 among four clusters in TCGA and CGGA 325 cohorts. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
FIGURE 4
FIGURE 4
Construction of SGMRS and its relationship with prognosis. (A) Average of coefficients of five critical SGMGs in the LASSO Cox regression at each lambda value. (B) The prognostic effect of each critical SGMG in glioma. (C) Optima cutoff value of SGMRS in all four cohorts. (D) K-M curves of the (D) TCGA, (E) CGGA325, (F) CGGA693, and (G) WCH cohorts based on SGMRS high- and low-risk groups. (H) ROC curves and matched AUC of 1-, 2-, 3-year survival rate in all four cohorts.
FIGURE 5
FIGURE 5
Clinicopathological features of SGMRS risk groups. (A) Expression level of five critical SGMGs and its relationship with clinicopathological features. (B) Gene mutations of five critical SGMGs and top eight frequently mutated genes in gliomas ordered by SGMRS risk groups. (C) Difference in tumor mutation burden between SGMRS high- and low-risk groups. (D) Copy number variation and its relationship with clinicopathological features ordered by SGMRS risk groups. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
FIGURE 6
FIGURE 6
Functional enrichment analyses between two SGMRS risk groups. (A) Pathways with high confidence and odds ratio in REACTOME gene sets. (B) Pathways with high confidence and odds ratio in KEGG gene sets. (C) Top 12 pathways in REACTOME gene set with the highest GSVA scores. (D) Top 12 pathways in KEGG gene set with the highest GSVA scores. (E) Top five pathways in REACTOME gene set with the highest normalized enrichment scores. (F) Top five pathways in KEGG gene set with the highest normalized enrichment scores.
FIGURE 7
FIGURE 7
Prognostic value of SGMRS and construction of SGMRS-based nomograms. (A) Univariate analysis of potential prognostic factors in glioma. (B) Multivariate analysis to identify independent prognostic factors in glioma. (C) Nomogram of 1-, 2-, and 3-year survival rate of glioma patients in TCGA cohort. (D) Calibration plots for the nomogram of TCGA cohort. (E) Nomogram of 1-, 2-, and 3-year survival rate of glioma patients in CGGA325 cohort. (F) Calibration plots for the nomogram of CGGA325 cohort.
FIGURE 8
FIGURE 8
Analyses on immune landscapes of tumor microenvironment between SGMRS high- and low-risk groups. (A) Boxplot for the estimated infiltration fraction of 22 types of immune cells in tumors. (B) Differences in the stromal, immune, and ESTIMATE scores between two risk groups in TCGA, CGGA325, and WCH cohorts. (C) Differences in tumor purity between two risk groups in TCGA, CGGA325, and WCH cohorts. (D) Analyses of correlations of SGMRS with the (D) stromal score, (E) immune score, (F) ESTIMATE score, and (G) tumor purity in TCGA, CGGA325, and WCH cohorts. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
FIGURE 9
FIGURE 9
Differences in expression of immunotherapy-related genes, immunological phenotype, and response to ICIs between two SGMRS risk groups. (A) Boxplot for the expression level of 33 immunotherapy-related genes in two risk groups in TCGA cohort. (B) Analyses of correlations between SGMRS and the expression levels of CD274, CD276, CD44, and PD-1 in TCGA cohort. (C) Analysis of TIP score and related gene expression levels ordered by SGMRS in TCGA cohort. (D) Difference in TIP score between two risk groups in TCGA cohort. (E) Analysis of correlation between SGMRS and TIP score in TCGA cohort. (F) Analysis of TIP score and related gene expression levels ordered by SGMRS in CGGA325 cohort. (G) Difference in TIP score between two risk groups in CGGA325 cohort. (H) Analysis of correlation between SGMRS and TIP score in CGGA325 cohort. (I) Difference in proportion of patients with high cyto-toxic T lymphocytes infiltration between two risk groups in TCGA cohort. (J) Difference in proportion of patients with predictive response to immune checkpoint inhibitors between two risk groups in TCGA cohort.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Anderson N. R., Minutolo N. G., Gill S., Klichinsky M. (2021). Macrophage-based approaches for cancer immunotherapy. Cancer Res. 81 (5), 1201–1208. 10.1158/0008-5472.Can-20-2990 - DOI - PubMed
    1. Aran D., Sirota M., Butte A. J. (2015). Systematic pan-cancer analysis of tumour purity. Nat. Commun. 6, 8971. 10.1038/ncomms9971 - DOI - PMC - PubMed
    1. Chaneton B., Hillmann P., Zheng L., Martin A. C. L., Maddocks O. D. K., Chokkathukalam A., et al. (2012). Serine is a natural ligand and allosteric activator of pyruvate kinase M2. Nature 491 (7424), 458–462. 10.1038/nature11540 - DOI - PMC - PubMed
    1. Chen Y. M., Shiu J. Y., Tzeng S. J., Shih L. S., Chen Y. J., Lui W. Y., et al. (1998). Characterization of glycine-N-methyltransferase-gene expression in human hepatocellular carcinoma. Int. J. Cancer 75 (5), 787–793. 10.1002/(sici)1097-0215(19980302)75:5<787::aid-ijc20>3.0.co;2-2 - DOI - PubMed
    1. Chinot O. L., Wick W., Mason W., Henriksson R., Saran F., Nishikawa R., et al. (2014). Bevacizumab plus radiotherapy-temozolomide for newly diagnosed glioblastoma. N. Engl. J. Med. 370 (8), 709–722. 10.1056/NEJMoa1308345 - DOI - PubMed