Skip to main content
Figure 1 | Breast Cancer Research

Figure 1

From: Bone marrow stromal antigen 2 expressed in cancer cells promotes mammary tumor growth and metastasis

Figure 1

BST-2 mRNA is prevalent in highly aggressive tumors and associates with patients’ poor survival. (A) RNA-seq data (n = 100) of paired tumor versus normal breast tissues from The Cancer Genome Atlas (TCGA) breast-invasive carcinoma (BRCA) data portal presented as scatter plot and heat map show that BST-2 is significantly elevated in tumor tissues compared to matched normal breast tissues. (B) Levels of BST-2 in tumor tissues of patients bearing different subtypes of invasive breast carcinomas show that BST-2 is upregulated in different breast tumors subtypes with the exception of the basal subtype. (C) BST-2 expression in tumors from Uppsala (Sweden) breast cancer patients obtained from GSE4922 was segregated into three BST-2 expression levels (relative units): low = 6.0 to 7.5, intermediate = 7.5 to 9.0, and high = 9.0 to 11.0. (D) Tumor size in patients with low, intermediate, or high levels of BST-2 is shown. (E) BST-2 GeneChip Robust Multiarray Averaging (GC-RMA) signal scores from healthy, ductal carcinoma in situ (DCIS), and invasive ductal carcinoma (IDC) tumor-bearing patients obtained from GSE21422. (F) BST-2 levels from normal, primary tumors (tumors), and metastatic tumors (metastatic) of patients bearing invasive breast cancer (TCGA). (G) Kaplan-Meier survival analysis using TCGA (BRCA) primary tumor samples segregated into high and low BST-2 levels show a significant link between low BST-2 and patient survival. The median overall survival (OS) time and the area under the curve (AUC) for each group are shown. (H) Mammary epithelial and stromal cells obtained from normal and invasive breast cancer patients (GSE10797) show elevated BST-2 expression in cancerous epithelial cells but not in cancerous stromal cells. In all panels, numbers correspond to P values. The relative units for BST-2 RNA levels acquired from TCGA and Gene Expression Omnibus (GEO) datasets are SEM-normalized and centralized log2(x + 1). Error bars represent standard deviations and significance was taken at P <0.01**. ns = not significant.

Back to article page