Integrated proteomics and transcriptomics analysis reveals key regulatory genes between ER-positive/PR-positive and ER-positive/PR-negative breast cancer

Abstract

Purpose: Hormone receptor-positive breast cancer is characterized by the expression of estrogen receptor (ER) or progesterone receptor (PR), it is generally associated with less aggressive clinical features and more favorable prognostic outcomes, primarily due to the effectiveness of endocrine therapy. However, the loss of PR expression has been correlated with endocrine resistance and poorer prognosis. To date, there is limited research elucidating the underlying mechanisms distinguishing ER-positive/PR-positive from ER-positive/PR-negative breast cancer. This study aims to investigate the molecular mechanisms associated with these two subtypes and to propose recommendations for precision therapy.

Patients and methods: Fresh tumor tissues from ER + /PR + patients (n = 5) and ER + /PR- patients (n = 5) were subjected to proteomic analysis to identify differentially expressed proteins. Transcriptomic data were obtained from the TCGA database, encompassing 937 breast cancer patients divided into three subgroups: ER + /PR + (n = 627), ER + /PR- (n = 112), and ER-/PR- (n = 198). Clinical characteristics and prognostic data were collected to analyze disease-specific survival (DSS) and overall survival (OS) across the three subtypes. Differential expression data for both transcripts and proteins were extracted, and Cox regression along with Least Absolute Shrinkage and Selection Operator (LASSO) regression were applied to identify key regulatory genes. A risk scoring formula was employed to classify patients into high-risk and low-risk groups. Kaplan-Meier curves, Gene Set Enrichment Analysis (GSEA), immune cell infiltration analysis, and OncoPredict drug sensitivity predictions were conducted to provide insights into the underlying mechanisms and clinical treatment strategies for this patient cohort. The accuracy of this model was further validated using external GEO datasets (GSE21653, GSE20685, and GSE42568). Additionally, we collected data from 97 hormone receptor-positive breast cancer patients who underwent neoadjuvant chemotherapy at our center between January 2021 and December 2023, assessing their response to chemotherapy using the Miller-Payne score.

Results: In the TCGA database, patients with ER + /PR- breast cancer exhibited poorer 5-year DSS and OS compared to those with ER + /PR + status (DSS: P = 0.038; OS: P = 0.052), which was similar to those with ER-/PR- status (DSS: P = 0.47; OS: P = 0.77). 186 differentially expressed proteins (110 up-regulated and 76 down-regulated) were identified based on proteomic analysis. After COX regression and Lasso regression, five key differential genes with prognostic and diagnostic value of ER + /PR + and ER + /PR- patients were finally included, that is HPN, FSCN1, FGD3, LRIG1, and TBC1D7. HPN, FSCN1 and FGD3 can be regarded as a tumor suppressor gene. And LRIG1, and TBC1D7 can be regarded as a risk-associated gene. Patients with high-risk scores had significantly lower survival probabilities compared to those with low-risk scores. Additionally, there were differences in functional pathway enrichment analysis (galactose_metabolism, glycolysis_gluconeogenesis, jak_stat_signaling_pathway, pentose_phosphate_pathway, et al.) and immune cell infiltration (CD8 T cell, Macrophages M1, et al.) between the high-risk and low-risk groups. Drug sensitivity analysis indicated that the low-risk patients may be more sensitive to endocrine drug like fulvestrant, while high-risk patients may be more sensitive to chemotherapy drugs like docetaxel, paclitaxel, and vinorelbine. Of the 97 patients underwent neoadjuvant chemotherapy in our center, the proportion of patients achieving Miller-Payne (MP) score 4 and 5 was higher in ER + /PR- patients (44%) compared to ER + /PR + patients (9.8%).

Conclusion: We confirmed that ER + /PR- breast cancer patients exhibited worse survival compared with ER + /PR + patients. Five key regulatory genes were identified and potential mechanisms and biological pathways were discovered, our prediction of drug sensitivity offers new insights for developing precise pharmacological treatment strategies for ER + /PR- breast cancer.

Keywords: Estrogen receptor; Progesterone receptor; Prognosis; Proteomic analysis.

Fig. 1

Flow chart of data collection, analysis and validation

Fig. 2

Survival analysis of patients in TCGA database and process for screening genes. A Survival analysis of the ER +/PR +, ER +/PR- and ER-/PR- groups. Kaplan–Meier survival curves were plotted and the log-rank test was performed to compare the DSS of these groups in TCGA database. B The OS of these groups. C The volcano map was drawn for 186 differential proteins, the red dots indicate significant upward adjustments, the blue dots indicate significant downward adjustments, and the gray dots indicate no significant differences. D Survival analyses for 37 genes using univariate Cox regression model. E Screening of prognostic model genes using LASSO regression. F Cross-validation of LASSO regression parameter selection

Fig. 3

Identifying target Genes and performing data analysis. A Heatmap showing the gene expression profiles of high-risk group and low-risk group. B PR-dependent ROC curves of risk score in train cohort. C PR-dependent ROC curves of risk score in the external validation, GSE21653. D PR-dependent ROC curves with HER2-negative data of risk score in train cohort. E Time-dependent ROC curves with HER2-negative data of risk score in train cohort on OS. F Risk scores of ER +/PR +, ER +/PR- and ER-/PR- groups. G The biological functions and pathways of dysregulated genes in both groups through KEGG pathway analysis were displayed. F The correlation heatmap between the risk model and the immune microenvironment were displayed

Fig. 4

Prognosis analysis of patients in different cohorts. A The probability of early cumulative mortality between high-risk patients and low-risk patients (left panel), Time-dependent ROC analysis at 1, 3, and 5 years (middle panel), Kaplan–Meier survival curves between high-risk patients and low-risk patients (right panel) were displayed in the TCGA cohort. B in the GSE 21653 cohort, C in the GSE 20685 cohort, D in the GSE 42568 cohort

Fig. 5

The estimated IC50s of clinical chemotherapeutic drugs and endocrine therapy drugs in high-risk and low-risk groups. A, B The estimated IC50s of Fulvestrant. C, D The estimated IC50s of Docetaxel. E, F The estimated IC50s of Paclitaxel. G The estimated IC50s of Vinorelbine. H The efficacy of neoadjuvant chemotherapy of the ER +/PR + and ER +/PR- groups was displayed.