Association of MPO Expression with the Immune Microenvironment in Breast Cancer: Insights from Bioinformatics and Single-Cell Analyses

Summary

This article presents a reproducible bioinformatics and single-cell workflow for exploring associations between myeloperoxidase (MPO) expression and immune/myeloid features in breast cancer. Because the analyses are based on public datasets and in silico methods, the findings are interpreted as exploratory and hypothesis-generating.

Abstract

Breast cancer remains a major cause of cancer-related mortality, and exploratory computational workflows can help prioritize immune-associated markers for further investigation. Here, we used the cancer genome atlas breast invasive carcinoma (TCGA-BRCA) bulk transcriptomic data and the public single-cell dataset GSE161529 to examine associations between myeloperoxidase (MPO) expression, clinical outcomes, immune infiltration, methylation, upstream-regulator annotations, single-cell expression patterns, virtual knockdown sensitivity outputs, drug–gene interaction retrieval, and absorption, distribution, metabolism, excretion, and toxicity (ADMET) annotation. MPO expression was lower in breast cancer tissues than in adjacent non-tumor tissues. Higher MPO expression was associated with a longer progression-free interval, whereas its associations with overall survival and disease-specific survival were not statistically significant. Receiver operating characteristic (ROC) analysis suggested tumor–normal separation within the analyzed public dataset, but this should not be interpreted as clinical diagnostic validation. Immune deconvolution and enrichment analyses indicated that MPO expression mainly tracked with immune- and myeloid-related transcriptional features, rather than establishing tumor-intrinsic regulation of the immune microenvironment. At single-cell resolution, the MPO signal was sparse, with only 85 MPO-positive cells detected before k-nearest neighbor (KNN)-based neighborhood expansion. Detectable MPO signal and MPO-associated scores were interpreted cautiously because they may be influenced by sparse expression, cell-type annotation uncertainty, dropout, doublets, or ambient RNA. In silico virtual knockdown suggested candidate immune- and inflammatory-related transcriptional changes, but these results were considered exploratory and require validation. Drug-gene interaction database (DGIdb)-based drug-gene retrieval and ADMET annotation were used only as preliminary chemical annotations and were not interpreted as therapeutic evidence. Overall, this study provides a reproducible in silico workflow for generating hypotheses about MPO-associated immune/myeloid features in breast cancer, which require external cohort validation and experimental confirmation.

Introduction

Breast cancer is a highly heterogeneous immune-related malignancy¹. Its disease progression, risk of recurrence and metastasis, and treatment response are closely associated with the composition and functional status of the tumor immune microenvironment (TIME)². Despite ongoing optimization of comprehensive treatment strategies, some patients still experience progression or recurrence, underscoring the urgent need to identify molecular biomarkers that characterize TIME status and support risk stratification, while elucidating their underlying mechanisms.

Myeloperoxidase (MPO) is a heme-contain....

Protocol

Acquisition from the TCGA database

RNA sequencing data and clinical information for the TCGA breast invasive carcinoma (TCGA-BRCA) cohort were obtained from the genomic data commons portal¹⁴. STAR workflow RNA-seq data in transcripts per million (TPM) format were extracted together with matched clinical annotations. RNA-seq samples lacking corresponding clinical information were excluded. For expression-based analyses, TPM values were transformed as log2(TPM + 1). MPO expression was extracted using the gene symbol MPO and Ensembl gene ID ENSG00000005381.8. For analyses requiring MPO-high and MPO-low gr....

Results

MPO expression patterns and exploratory survival associations in breast cancer

To describe MPO expression patterns across cancer datasets, we analyzed MPO RNA-seq data from the TCGA pan-cancer dataset and observed lower MPO expression in tumor tissues from bladder urothelial carcinoma (BLCA), breast invasive carcinoma (BRCA), glioblastoma multiforme (GBM), head and neck squamous cell carcinoma (HNSC), kidney chromophobe (KICH), liver hepatocellular carcinoma (.......

Discussion

This study presents an exploratory public-dataset and in silico workflow for examining associations between MPO expression and immune/myeloid features in breast cancer. The TCGA-BRCA analyses showed that MPO expression was lower in tumor tissues than in adjacent non-tumor tissues and that higher MPO expression was associated with a longer progression-free interval. However, overall survival and disease-specific survival were not statistically significant. Therefore, MPO should not be interpreted as a robust or establishe.......

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
CellChat	R package/Open source	https://github.com/sqjin/CellChat	Cell-cell communication analysis
ChIP-Atlas	Public database	https://chip-atlas.org/	TF target screening; 2021 update
clusterProfiler	Bioconductor	https://bioconductor.org/packages/clusterProfiler/	GO/KEGG enrichment analysis; v4.4.4
Cytoscape	Cytoscape Consortium	https://cytoscape.org/	Network visualization and topology analysis
DGIdb	Washington University/Public database	https://www.dgidb.org/	Drug-gene interaction retrieval
GDC/TCGA-BRCA	National Cancer Institute	https://portal.gdc.cancer.gov/	Bulk transcriptomic and clinical data source
Gene Expression Omnibus: GSE161529	NCBI	https://www.ncbi.nlm.nih.gov/geo/	Single-cell dataset source
GSEA/MSigDB	Broad Institute	https://www.gsea-msigdb.org/gsea/msigdb	Gene set enrichment analysis and gene-set reference; Version 3.0
GSVA	Bioconductor	https://bioconductor.org/packages/GSVA/	Gene set variation/ssGSEA-related scoring; Version 1.46.0
GTRD	Public database	http://gtrd.biouml.org/	TF target screening; 2021
KnockTF	Public database	http://www.licpathway.net/KnockTF/index.html	TF perturbation resource; Version 2.0
R	R Foundation for Statistical Computing	https://www.r-project.org/	Statistical computing environment
scTenifoldKnk	R package/Open source	https://github.com/cailab-tamu/scTenifoldKnk	Virtual knockdown analysis
Seurat	R package/Open source	https://satijalab.org/seurat/	Single-cell preprocessing and clustering
STRING	ELIXIR/Public database	https://string-db.org/	Protein-protein interaction analysis; v11
SwissADME	SIB Swiss Institute of Bioinformatics	http://www.swissadme.ch/	Drug-likeness assessment; 2017 release/web tool
TIMER	Public web resource	https://timer.cistrome.org/	Immune infiltration analysis; TIMER2.0
UCSC Xena or linked TCGA portal	UCSC	https://xenabrowser.net/	Exploratory data access/validation