The potential mechanism of stromal cells in predicting the prognosis of glioma

The potential mechanism of glioma stromal cells in predicting the prognosis of glioma.
Glioma is the most common primary malignant tumor of the central nervous system and has a poor prognosis
.

In recent years, immunotherapy has been extensively studied in human malignant tumors.
However, the "cold immunity" characteristics of glioma have made only a few patients benefit from immunotherapy
.

The current research fever focuses on the understanding of the complex interaction between glioma and the immune system in order to exert the effect of glioma immunotherapy
.

An article published in Frontiers Oncology in April 2021 "Glioma-Associated Stromal Cells Stimulate Glioma Malignancy by Regulating the Tumor Immune Microenvironment" (IF: 6+) on glioma-associated stromal cells (GASC) as the prognosis of glioma The underlying mechanism of markers is discussed
.

Abbreviations: GASC, glioma-associated stromal cell TCGA, The Cancer Genome Atlas CGGA, Chinese Glioma Genome Atlas HGG, high-grade glioma; High-grade glioma ssGSEA , Single-sample Gene Set Enrichment Analysis algorithm PCA, Principle Component Analysis DEGs, differentially expressed genes; CNV, copy number variations; ICPs, immune checkpoints.
Immunity Checkpoint TIDE, The Tumor Immune Dysfunction and Exclusion Research Background Glioma-associated stromal cells are an important stromal cell in the newly discovered glioma microenvironment, and have potential value for predicting prognosis and treatment prospects
.

However, the underlying mechanism of GASCs in the glioma microenvironment is still largely unknown
.

There is evidence that GASCs promote angiogenesis, invasion and tumor growth
.

The author’s study aimed to explore the mechanism of action of GASCs in gliomas, especially in high-grade gliomas (HGG)
.

Research methods This study used glioma datasets from TCGA and CGGA
.

Downloaded the gene sequencing data and clinical information data of 702 cases of TCGA and 693 cases of CGGA glioma samples, and performed methylation analysis
.

Use a single-sample gene set enrichment analysis algorithm to calculate the GASCs enrichment score of each sample, calculate the stem cell score, mesenchymal-epithelial-macrophage transition (EMT) score, tumorigenic factor score and angiogenesis activity score, and The corresponding genome
.

Next, the SsGSEA algorithm was used to distinguish between high and low patients in the GASC group
.

Then Morpheus was used to identify genes that were significantly differentially expressed between the high-GASC group and the low-GASC group, the xCELL and CIBERSORT algorithms were used to analyze the composition of stromal cells and immune cells, and TIDE was used to predict the clinical efficacy of immune checkpoint inhibitors for each glioma sample Response
.

Finally, a risk score and a nomogram model were constructed to predict the prognosis of glioma
.

Research results (1) In order to analyze the potential mechanism of GASCs in the glioma microenvironment, the author analyzed the mRNA sequencing data of 702 samples from TCGA and 693 samples from CGGA, and then used the ssGSEA algorithm to calculate the GASC score of each sample
.

According to the median of GASC score, the samples of TCGA and CGGA were divided into high GASC group and low GASC group (Figure 1).
Figures 1A and B show the information of high GASC group and low GASC group
.

In addition, 393 HGG samples from TCGA and 504 high-grade glioma (HGG) samples from CGGA were also separately classified and divided into high-GASC and low-GASC groups
.

Kaplan-Meier survival curve results showed that in all glioma groups and HGG, the higher the GASC score, the worse the OS (p<0.
0001; Figure 1C-F)
.

The results of principal component analysis showed that there were significant differences in the expression profiles of GASC markers between the high GASC group and the low GASC group (Figure 1G-J)
.

Figure 1 distinguishes high-GASC and low-GASC groups
.

(AB) represents the heat map of GASC markers in all glioma populations; (CF) represents Kaplan-Meier of samples from all glioma populations and high-grade glioma populations in the high-GASC and low-GASC groups Survival (OS) curve; (GJ) represents the principal component analysis results of all glioma groups and the high- and low-GASC groups of high-grade gliomas (A, C, G, I represent TCGA, B, D, H, J On behalf of CGGA)
.

(2) The relationship between GASCs and stromal cells The xCELL algorithm is used to evaluate the relationship between GASCs and other stromal cells.
The results indicate the content of endothelial cells, lymphatic endothelial cells and microvascular endothelial cells in the high GASC group in all glioma populations in the TCGA data Both are high, and the coefficient matrix shows that the GASC score is positively correlated with the above cell level
.

Univariate Cox regression results also indicate that mesenchymal stem cells are a protective factor for glioma (Figure 2)
.

Figure 2 The relationship between GASCs and stromal cells in the glioma population
.

(A, D) The bar graphs show the difference in xCELL scores between the high GASC group and the low GASC group; (B, E) Univariate Cox regression analysis forest plot of stromal cells; (C, F) the correlation between GASC scores and stromal cells
.

(A, B, C are TCGA, D, E, F are CGGA)
.

Remarks: p>0.
05, *P≤0.
05, **p≤0.
01, ***p≤0.
001, ****p≤0.
0001
.

(3) The immune landscape of the high-GASC group and the low-GASC group used the CIBERSORT algorithm to calculate the relative abundance of 22 immune cells, and at the same time analyzed the correlation between GASCs and 14 ICPs
.

In general, the high GASC group has a lower adaptive immunity level than the low GASC group, but the level of M2 macrophages is significantly higher in the high GASC group and is positively correlated with the GASC score, and is a risk factor for glioma
.

The expression levels of most ICP in the high GASC group were statistically different.
Cox regression results showed that the expression levels of PD-L2, TIM3, CD80, CD86, CD155 and CIITA were risk factors for glioma in all glioma populations and HGG populations.
And there is a strong positive correlation within the ICP, and the GASC score is also positively correlated with them (Figure 3) Figure 3 (1).
The relative abundance of 22 immune cells
.

(A is derived from TCAG, B is derived from CGGA) Figure 3 (2).
The correlation between GASC score and immune cell and ICP expression
.

(A).
Data correlation diagram of all glioma population; (B) Data correlation diagram of high-grade glioma population
.

Figure 3 (3)
.
The relationship between GASCs and ICP expression .

(ACEG) histogram illustrates the difference in expression of ICP in all glioma populations with high and low GASC groups and high glioma populations; (BDFH) all glioma populations (B stands for TCGA, D stands for CGGA) and high grades Forest plot of univariate Cox regression analysis of ICP expression in glioma population (F stands for TCGA, H stands for CGGA)
.

Remarks: p>0.
05, *P≤0.
05, **p≤0.
01, ***p≤0.
001, ****p≤0.
0001
.

(4) GASC score and prognostic survival analysis showed that the gene copy number variation of DR3 and CIITA in the high GASC group was significantly increased, which reduced the OS of patients with glioma
.

However, elevated levels of THY1 and CD80 methylation have better survival
.

Figure 4.
Methylation analysis of GASC markers and ICP
.

(A) The box plot shows the difference in the methylation levels of THY1, CD9, CD14, CD44, ITGAM and ACTA1 in the high GASC group, low GASC group and normal group; (B).
THY1, CD9, CD14, CD44, ITGAM and Kaplan-Meier overall survival curve of samples from ACTA1 high and low methylation group; (C).
Box plot shows the difference of Galectin-9, CD80, CD155 and LAG3 methylation levels in high and low GASC group and normal group; (D).
Kaplan-Meier survival curve of Galectin-9, CD80, CD155 and LAG3 hypermethylation group and hypomethylation group samples
.

(5) Prediction of immunotherapy response The TIDE tool is used to predict the possibility of immune response of each sample.
The results show that among all glioma patients, the low GASC group is more likely to respond to immunotherapy than the high GASC group.
In all glioma populations, the response group had lower levels of "T cell CD8" and "macrophage M0", while "mast cell activation" was higher, and the opposite was true in the HGG population
.

(Figure 5) Figure 5.
Potential immunotherapy response prediction between high and low GASC groups
.

(A) Potential immunotherapy responses of all glioma populations and high-grade glioma populations (A, C are TCGA, B, D are CGGA); (E) GASC and predict the correlation of immunotherapy response; (F, I) The bar graph illustrates the difference in immune cell scores between responders and non-responders in all glioma populations and high-grade glioma populations
.

(6) Construction of risk scoring system Finally, a nomogram model for predicting the prognosis of glioma was constructed, including risk score and clinicopathological characteristics
.

Univariate and multivariate Cox regression analysis showed that GASC risk score is an independent predictor of the prognosis of glioma
.

Figure 6.
Construction of risk scoring system and establishment and verification of nomogram survival model
.

(AB) Cox regression analysis of the training data set; (CD) Kaplan-Meier overall survival curve of the high-risk and low-risk groups; (E) Univariate and multivariate Cox regression analysis of the prognostic model; (F) 1 year prediction , 3-year or 5-year survival rate nomogram; (GH) nomogram prediction 1, 3, 5-year survival rate time-dependent ROC curve; (IJ) IDH mutation prediction 1, 3, 5-year survival rate versus time Changing ROC curve
.

(C, G, I are training data sets, D, H, J are validation data sets) Full text summary The author first observed GASC and M2 macrophages and immune checkpoints (PD-1, PD-L1, PD-L2, TIM3, Galectin-9, CTLA-4, CD80, CD86, CD155 and CIITA) levels were positively correlated, and were positively correlated with the GASC scores of all glioma groups and HGG groups
.

GASCs stimulate the malignant transformation of gliomas through M2 macrophages, and are related to the level of immune checkpoints in the glioma microenvironment
.

The high GASC group has higher copy number variation of DR3 and CIITA, and decreased THY1 methylation, which can be used as a prognostic indicator and treatment target for glioma
.

However, the underlying mechanism of GASCs in the glioma microenvironment, especially in HGG, and the risk score constructed by GASCs as an independent predictor of the prognosis of gliomas need to be confirmed by further studies
.

(Full text link: https://doi.
org/10.
3389/fonc.
2021.
672928) It is not easy for the editor to write articles.
For friendly copy, please contact the talking editor butler at 15510012760 (same number on WeChat) for related questions about Shengxin: 18501230653 (same on WeChat) No.
) Welcome to pay attention to Shengxin human transcriptome | methylation | resequencing | single cell | m6A | multiomics cytoscape | limma | WGCNA | Experimental gene | 2019-nCoV | Enrichment analysis | Joint analysis | Microenvironmental plague hunting | Idea summary | Scholars | Scientific research | Withdrawal | PhD reading | Work