Prognostic role of EGR1 in breast cancer: a systematic review

Abstract

EGR1 (early growth response 1) is dysregulated in many cancers and exhibits both tumor suppressor and promoter activities, making it an appealing target for cancer therapy. Here, we used a systematic multi-omics analysis to review the expression of EGR1 and its role in regulating clinical outcomes in breast cancer (BC). EGR1 expression, its promoter methylation, and protein expression pattern were assessed using various publicly available tools. COSMIC-based somatic mutations and cBioPortal-based copy number alterations were analyzed, and the prognostic roles of EGR1 in BC were determined using Prognoscan and Kaplan-Meier Plotter. We also used bc-GenEx- Miner to investigate the EGR1 co-expression profile. EGR1 was more often downregulated in BC tissues than in normal breast tissue, and its knockdown was positively correlated with poor survival. Low EGR1 expression levels were also associated with increased risk of ER+, PR+, and HER2- BCs. High positive correlations were observed among EGR1, DUSP1, FOS, FOSB, CYR61, and JUN mRNA expression in BC tissue. This systematic review suggested that EGR1 expression may serve as a prognostic marker for BC patients and that clinicopathological parameters influence its prognostic utility. In addition to EGR1, DUSP1, FOS, FOSB, CYR61, and JUN can jointly be considered prognostic indicators for BC. [BMB Reports 2021; 54(10): 497-504].

Fig. 1

Analysis of EGR1 protein structure, post-translational modification, and expression in breast cancer (BC). (A) Schematic diagram of EGR1 protein structure and post-translational modifications. (B) mRNA expression of EGR1 in normal and BC tissue (IDBC, invasive ductal breast carcinoma) was derived from Oncomine database. (C) mRNA expression of EGR1 in breast normal and cancer tissues was derived from UALCAN web using TCGA database. (D) Protein expression of EGR1 in breast normal and cancer tissues by immunohistochemistry (IHC) was derived from Human Protein Atlas web. The intensity of EGR1 expression was quantified by ImageJ and plotted by GraphPad Prism 7 software (right panel). (E-G) mRNA expression of EGR1 in BC clinicopathological subtypes was analyzed using the BRCA TCGA datasets through UCSC Xena web. Box plots showing the EGR1 mRNA expression in BC subcategories including PAM50 subtypes (E), clinical subtypes (F), and stages (G). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig. 2

Methylation status and genetic alterations of EGR1 in BC. (A) Methylation level of the EGR1 gene promoter in BC (TCGA Wanderer web tool). Median methylation level of the EGR1 gene promoter in BC. The box plot comparing specific CpG sites of EGR1 promoter methylation in normal (blue plot) and cancer tissue (red plot) was derived from the TCGA database (Methylation 450K) through the TCGA Wanderer web tool. The P values were obtained after an unpaired t-test using GraphPad Prism 7 software. (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). (B) EGR1 mutation in human BC. Table showed the percentage of the mutation type of EGR1 in BC according to COSMIC database. (C) Alteration frequency of EGR1 mutation in BC was analyzed by using cBioPortal web. (D) Alteration frequency of EGR1 copy number in BC was analyzed by using cBioPortal web.

Fig. 3

Relationship between EGR1 mRNA expression and clinical outcomes in BC patients (PrognoScan and Kaplan Meier plotter Database). (A-E) The survival curve comparing the patient with high (red) and low (blue) expression of EGR1 (probe: 201693_s_at) was plotted from PrognoScan database in BC patients. (F-M) The survival curve comparing the patient with high (red) and low (blue) expression of EGR1 (probe: 201693_s_at) was plotted from Kaplan Meier plotter in BC patients. The threshold of cox P-value < 0.05. Meta-Analysis of Studies of BC studies with EGR1 mRNA expression. Forest plots of GEO datasets evaluating association of EGR1 mRNA expression with OS (N), RFS (O), DMFS (P), and PPS (Q) in BC. Hazard ratio (HR) with 95% confidential interval (CI) and p-value were labeled in the right column of each forest plot.

Fig. 4

EGR1 mRNA expression is correlated to DUSP1, FOS, FOSB, CYR61, and JUN mRNA expression in BC. (A) Top 20 genes positively correlated with EGR1 mRNA expression based on 2, 136 BC samples in Curtis Breast (PMID: 22522925). Analysis was performed using Oncomine database. (B) The heat map of EGR1, DUSP1, FOSB, CYR61, and JUN mRNA expression across PAM50 BC subtypes in TCGA database. Data was analyzed using UCSC Xena (http://xena.ucsc.edu/). (C, D) Regression analysis showed that EGR1, DUSP1, and FOS had positively high correlation coefficients. Data was analyzed using UCSC Xena (http://xena.ucsc.edu/). (E) Data mining in bc-GenExMiner 4.0 confirmed the positive correlation between EGR1, DUSP1, FOSB, and JUN mRNA expression across DNA microarray data. (F, G) Regression analysis confirmed that EGR1, DUSP1, and FOS had positively high correlation coefficients across DNA microarray data. Data was analyzed using bc-GenExMiner 4.0 web. (H, I) The survival curve comparing the patient with high (red) and low (blue) expression of EGR1, DUSP1, FOS, FOSB, CYR61, and JUN was plotted from Kaplan Meier plotter in BC patients. The threshold of cox P-value < 0.05.