Co-expression of key gene modules and pathways of human breast cancer cell lines

Abstract

Breast cancer (BC) is the most common leading cause of cancer-related death in women worldwide. Gene expression profiling analysis for human BCs has been studied previously. However, co-expression analysis for BC cell lines is still devoid to date. The aim of the study was to identify key pathways and hub genes that may serve as a biomarker for BC and uncover potential molecular mechanism using weighted correlation network analysis. We analyzed microarray data of BC cell lines (GSE 48213) listed in the Gene Expression Omnibus database. Gene co-expression networks were used to construct and explore the biological function in hub modules using the weighted correlation network analysis algorithm method. Meanwhile, Gene ontology and KEGG pathway analysis were performed using Cytoscape plug-in ClueGo. The network of the key module was also constructed using Cytoscape. A total of 5000 genes were selected, 28 modules of co-expressed genes were identified from the gene co-expression network, one of which was found to be significantly associated with a subtype of BC lines. Functional enrichment analysis revealed that the brown module was mainly involved in the pathway of the autophagy, spliceosome, and mitophagy, the black module was mainly enriched in the pathway of colorectal cancer and pancreatic cancer, and genes in midnightblue module played critical roles in ribosome and regulation of lipolysis in adipocytes pathway. Three hub genes CBR3, SF3B6, and RHPN1 may play an important role in the development and malignancy of the disease. The findings of the present study could improve our understanding of the molecular pathogenesis of breast cancer.

Keywords: GEO; WGCNA; breast cancer cell lines.

Figure 1. Sample hierarchical clustering plot and trait heatmap

Basal represent white, claudin-low represents blue, luminal represents red, and non-malignant represents green.

Figure 2. Analysis of network topology for various soft-thresholding powers

Figure 3. Clustering dendrograms of genes with dissimilarity based on the topological overlap, together with assigned module color

Figure 4. Module-trait associations

Each row represents a module eigengene, each column to a trait. Each cell contains the corresponding correlation and P-value.

Figure 5. A scatterplot of gene significance versus MM

Plot (A) is for brown module, plot (B) is for the black module, plot (C) is for dark red module, and plot (D) is for midnight blue module.

Figure 6. Heat map view of TOM of co-expressed genes in different modules in the top 1500 genes

Figure 7. Analysis of connectivity of eigengenes in a different module

(A) Cluster analysis of eigengenes and (B) the heatmap of connectivity of eigengenes.

Figure 8. GO terms shows as an interaction network using Cytoscape plug-in ClueGO

Plot (A) is for brown module, plot (B) is for the black module, plot (C) is for dark red module, and plot (D) is for midnight blue module.

Figure 9. KEGG pathways displayed as an interaction network using Cytoscape plug-in ClueGO

Plot (A) is for the brown module, plot (B) is for the black module, and plot (C) is for dark red module.

Figure 10. The visualization of four key modules

Black nodes, black module; dark red nodes, dark red module; brown nodes, brown module; and midnight blue nodes, midnight blue module.

Figure 11. Three significant modules identified from the whole network using Cytoscape plug-in MCODE with a score of >6.0

Module I: black module, MCODE score = 34.979; module II: midnight blue module, MCODE score = 8.889; module III: brown module, MCODE score = 7.579. The hub genes in the modules were bold with red.

Figure 12. The prognostic value of mRNA level of hub genes in BC patients (OS in Kaplan–Meier plotter)

(A) CBR3 (205379_at); (B) SF3B6 (223416_at); (C) RHPN1 (228405_at).

Figure 13. The mRNA expression level of three key genes between BRCA and adjunct noncancerous breast tissues based on TCGA data

(A) CBR3; (B) SF3B6; (C) RHPN1.

Figure 14. The ROC curves analysis for diagnostic power prediction by the three key genes

Figure 15. The two mRNA (CBR3 and RHPN1) were high expressed in kidney cancer cell lines from CCLE analysis