Lipid Metabolic-Related Signature CYP19A1 is a Potential Biomarker for Prognosis and Immune Cell Infiltration in Gastric Cancer

Figures

• 

  Figure 1. Identification of lipid metabolism-related genes from TCGA-STAD. (A) Heatmap of differentially expressed lipid metabolism-related genes between GC tumor (n=375) and normal tissues (n=32). (B) Volcano plot of the differentially expressed lipid metabolism-related genes. (C) Scatter plots of mRNA expression of differentially expressed lipid metabolism-related genes. (D) The top 30 significant terms of GO function enrichment. (E) GO circle plot of biological process (BP) differentially enriched. (F) KEGG circle plot of the logFC of the specified gene. The higher Z-score value indicated, the higher expression of the enriched pathway.

• 

  Figure 2. Construction of the risk model in the training cohort. (A) The forest plots of univariate Cox analysis results of the 10 lipid metabolism-related genes significantly correlated with overall survival of GC patients. (B) Heatmap illustrating the 4 metabolic genes signature expression pattern. (C) Risk score distribution and survival status of GC patients based on the 4 metabolic genes signature. (D) Survival curves of patients in the low- and high-risk groups based on the 4 lipid metabolism-related genes. (E) Time-dependent ROC analysis for the risk model at 1, 3, 5 years. (F and G) Forest plot of univariate and multivariate Cox regression analysis in the TCGA cohort. (H) ROC analysis of OS for the four-signature and the clinicopathologic parameters.

• 

  Figure 3. Verification of lipid metabolism-related gene signature in an independent cohort. (A) Kaplan–Meier curves of GC patients in the low - and high -risk groups based on the 4 lipid metabolism-related genes in an GSE84437 cohort. (B) Heatmap of the 4 lipid metabolism-related genes expression pattern in GSE84437 cohort. (C) Risk score distribution and survival status of GC patients based on the 4 lipid metabolism-related genes. (D) ROC curves to evaluate 1, 3, and 5-year prediction efficiency of the 4 lipid metabolism-related signature in GSE84437 cohort.

• 

  Figure 4. Expression level and prognostic value of CYP19A1 in gastric cancer. (A) The expression level of CYP19A1 in STAD cases was higher than that in the adjacent normal tissues. (B) CYP19A1 mRNA expression were higher in gastric cancer tissues compared with normal tissues in TCGA cohort. (C) CYP19A1 mRNA expression were significantly higher in gastric cancer tissues compared with normal tissues in GSE54129 cohort. (D) Representative IHC image show CYP19A1 protein level were higher in GC tumour tissues relative to adjacent normal tissues among 24 patients by immunohistochemistry, scale bar=100 μm. (E) Different distribution of CYP19A1 was statistically analyzed with Chi-square test. (F) The mRNA expression level of CYP19A1 in different tumor stages based on TCGA-STAD. (G) The alteration frequency of CYP19A1 were determined by cBioportal database. *P<0.05, **P<0.01, ***P<0.001.

• 

  Figure 5. Enrichment plots by GSEA. (A) GSEA results showed that higher CYP19A1 expression was associated with activation of glycerophospholipid and glycerolipid-related metabolism pathways. (B) Correlation between CYP19A1 expression level and lipid metabolism-related genes FASN (right) and SCD1(left).

• 

  Figure 6. Comprehensive analysis of immune characteristics. (A) The CYP19A1 expression significantly correlated with infiltration levels of immune cells in GC. The upper triangle in each tile indicates coefficients calculated by Pearson’s correlation test, and the lower triangle indicates log10 transformed P-value. *P < 0.05. (B) CYP19A1 CNV affects the infiltrating levels of B cells, CD8+T cells, CD4+T cells, Macrophage, neutrophils, and dendritic cells in GC. (C) The change ratio of 22 immune cell subtypes in the high (blue) and low (red) CYP19A1 expression groups in GC samples. P < 0.05 was considered significant. Analysis of CYP19A1 mRNA in different ESTIMATE score (D), stromal score (E) and immune score (F) of GC tissues from TCGA database. *P<0.05, **P<0.01, ***P<0.001.