Fig 1: GSEA functional enrichment and immune infiltration analysis. (A) The four hallmark pathways were all enriched in the training, validation 1 (GSE53622) and validation 2 (TCGA-ESCC) groups by GSEA functional enrichment analysis. (B) The risk score was positively correlated with Treg cell infiltration, and negatively correlated with plasma cells in the training, validation 1 cohort(GSE53622) and validation 2 cohorts (TCGA-ESCC). (C) IHC validation of three-gene metabolic-related signatures (INPP5E,CD38 and POLR3G) and FOXP3(Treg) in 20 ESCC tissues.
Fig 2: The relationship between the metabolism-related gene signature and clinicopathologic parameters. (A?C) Relationships between the expression of three genes (INPP5E, CD38 and POLR3G) and the tumor characteristics, including comparison with normal tissues, tumor grade (poorly, moderately and well), N stage (N0, N1, N2 and N3) and TNM stage (TNM1, TNM2 and TNM3) were analyzed. (D) The differential expression between tumor and normal tissues was verified by IHC analysis. (E) The ROC curve of each parameter with AUC scores in the training cohort, validation 1 cohort (GSE53622), validation 2 cohort (TCGA-ESCC) and independent validation cohort. *p < 0.05; **p < 0.01; ***p < 0.001.
Fig 3: The establishment of a metabolic-gene prognosis signature in the training group. (A) Kaplan-Meier survival analysis of overall survival, the distribution of patients’ risk scores, and survival status for the high- and low-risk groups. (B) The heatmap of three gene signatures. (C) The independent prognostic factor of the three gene signature by multivariate Cox regression analysis. (D) Kaplan-Meier survival analysis for INPP5E, CD38 and POLR3G expression.
Fig 4: Validation of the independent cohort by RT-qPCR and IHC analysis. (A) Kaplan-Meier survival analysis of overall survival, the distribution of patients’ risk scores, and survival status for the high- and low-risk group in an independent validation cohort with 49 ESCC patients. (B) Heatmap of three gene signatures in an independent validation cohort with 49 ESCC patients. (C) The independent prognostic factor of the three gene signature by multivariate Cox regression analysis in an independent validation cohort with 49 ESCC patients. (D) HE staining and INPP5E, CD38 and POLR3G expression were analyzed by IHC analysis in an external validation cohort of 95 ESCC patients. (E) Kaplan-Meier survival analysis of overall survival for the risk score, INPP5E, CD38 and POLR3G in the IHC validation cohort.
Fig 5: (A) Relative expression levels of miR-26a-5p and POLR3G in lung cancer cell lines. Expression levels of miR-26a-5p are relatively low in H23 and H460 cells revealed by miRNA expression assay (triplicate). (B) Expression levels of POLR3G are relatively high in H23 and H460 cells revealed by qPCR (triplicate). (C) After miR-26a-5p transfection, relative expression levels of POLR3G were down-regulated in H23 cells, revealed by qPCR (triplicate). (D) After miR-26a-5p transfection, relative expression levels of POLR3G were down-regulated in H460 cells, revealed by qPCR (triplicate). (E) After miR-26a-5p transfection, POLR3G protein level checked by western blots were decreased in H23 and H460 cells (singlicate). (F) The schematic diagram of miR-26a-5p binding sites in the 3'-untranslated regions (UTR) of POLR3G. (G) Renilla activity was significantly down-regulated after miR-26a-5p transfection (triplicate). The presented data depict the means ± SD of three independent experiments (except western blots in Figure1E). *P < 0.05, **P < 0.005. qPCR: quantitative polymerase chain reaction; miR-NC: microRNA negative control; SD: standard deviation.
Supplier Page from Abcam for Anti-POLR3G antibody