Figure 1.
BST2 overexpression is associated with high histological grade of primary breast cancer.
A. Low grade invasive primary breast tumor showing weak or no BST2 expressing cancer cells. Blue – hematoxylin counterstain. B. High grade invasive primary breast tumor displaying strong BST2 localization in cancer cells (brown). C. Homogeneous BST2 immunostaining in cells of coexisting ductal carcinoma in situ (DCIS), and invasive tumor. Bar –50 µm.
Table 1.
BST2 immunolocalization in primary invasive breast cancer.
Figure 2.
Differential BST2 expression in primary breast cancer of varying histological grade is maintained in tumor-derived cell lines.
A. QPCR based BST2 transcript levels in 17 breast cancer cell lines normalized to expression in non-malignant breast epithelium. B. Western blot analysis of BST2 protein (25–35 kd) in breast cancer cells. Tubulin used as a loading control. C. Microscopic images of BST2 immunostaining (green) in fixed, permeabilized breast cancer cells. Nuclei counterstained with propidium iodide (red). Bar –50 µm. D. BST2 immunolocalized at the cell membrane in live, unfixed breast cancer cells. Bar –25 µm.
Figure 3.
Regulation of BST2 expression by TGFβ in breast cancer cells.
A. Reduction in baseline BST2 expression detected in 5 grade 1 & 2 cell lines after 24 hr TGFβ (4 ng/ml) treatment. Each QPCR reaction was carried out in triplicate, and values normalized to ACTB expression and to vehicle-only controls. Asterisks represent statistically significant differences (p<0.01) between untreated and TGFβ-treated samples. B. TGFβ-induced shift in AP2 and STAT3 transcript levels under the same treatment conditions employed in panel A. BST2 reduction in grade 1 (CCdl22) & 2 (CCdl66) cell lines is accompanied by a decline in STAT3 levels, while AP2 transcripts increase significantly. No significant changes were observed in grade 3 (CCdl54) cells. Each assay was done in triplicate. Asterisks indicate a significant difference (p<0.01) between untreated and TGFβ-treated samples. C. Inhibition of the TGFβ pathway with 24 hr SB-431542 treatment increased BST2 expression in grade 1 cell lines (CCdl22, CCdl68). Data acquired in triplicate shows significant differences (p<0.01) between test and vehicle-only cell samples, indicated by asterisks.
Figure 4.
Differential AP2 binding to the BST2 promoter in primary breast cancer cell lines of varying histologic grade.
A. Schematic of the BST2 promoter region spanning 239 bp upstream of the transcriptional start site, including 111 bp of exon 1. Numbering is relative to the translation start site, highlighted in green (+1). Potential cis-regulatory elements shown are either on the plus (+) or minus (–) DNA strands. Regulatory binding sequences: AP2 (blue), STAT1 (red) and STAT3 (purple). B. Top panel - Chromatin immunoprecipitation (ChIP) performed with anti AP2, or control IgG in 8 breast cancer cell lines. Significant AP2 recruitment to the BST2 promoter observed only in grade 1 (CCdl22, CCdl68, CCdl67) & grade 2 (CCdl66, CCdl61) cell lines. Bottom panel - DNA from ChIP samples in top panel analyzed by QPCR. Primers encompassing putative AP2 binding sites from −221 to +6 were used. Melt curves were analyzed to ensure amplification of a single product. Each reaction was performed in triplicate. The plot represents relative binding efficiency determined by 2-ΔΔCT, where ΔCT is the difference between input CT and immunoprecipitated CT; ΔΔCT is the difference between AP2-immunoprecipitated ΔCT and IgG-immunoprecipitated ΔCT. Asterisks represent statistical significance (p<0.01). C. Inhibition of AP2 binding to the BST2 promoter in TGFβ-responsive primary breast cancer cell lines. Prior to processing for ChIP, cells were treated with vehicle, 4 ng/ml TGFβ, or 20 uM TGFβ inhibitor - SB-431542. Note striking reduction in AP2 binding in the presence of SB-431542 in 3 independent test cell lines. D. Hypothetical representation of BST2 transcriptional regulation by the TGFβ axis. Intact TGFβ regulation mediated by AP2 binding to the BST2 promoter enables maintenance of low baseline expression in grade 1 & 2 breast cancer cells.
Figure 5.
Functional consequences of endogenous BST2 overexpression in high grade primary breast cancer cells.
A. Top panel – Reduction of BST2 expression after transient knockdown by BST2 siRNA transfection of grade 3 tumor cells (CCdl54). Values normalized to ACTB expression. Bottom panel – BST2 immunolocalization (green) on day-6 post knockdown, showing sustained reduction in BST2 siRNA transfected cells plated in Matrigel; nuclei counterstained with PI (red). Bar - 50 µm. B. BST2 knockdown enhances drug-mediated apoptotic cell death in 6-day old tumor cultures plated in Matrigel, and treated with 5 µM tamoxifen, or 5 µM staurosporine for 24 hrs. Top panel – immunolocalization of anti cleaved caspase 3 (green); nuclei counterstained with PI (red). Bar - 50 µm. Bottom panel – Data plotted as percent apoptotic cells in 2 independent BST2 overexpressing cell lines (CCdl54, CCdl672). Total number of PI-stained nuclei (>100), and proportion of cleaved caspase 3 positive tumor cells was averaged for 3 optical fields scanned with a 20× objective. Differences in apoptotic cell yield between NS siRNA vs. BST2 siRNA-transfected cultures after treatment with each drug were significant (p<0.01). C. Growth reduction induced by BST2 knockdown measured by BrdU incorporation in primary tumor cells plated in Matrigel. Top panel - BrdU immunolocalization (yellow) in tumor nuclei (CCdl54) counterstained with PI (red) 7-day post transfection with BST2 siRNA. Bar - 25 µm. Middle panel - Data plots represent the fraction of BrdU positive proliferating cells in control (>100) and BST2 knockdown cultures averaged from 5 optical fields scanned with a 20× objective. Asterisk denotes a significant (p<0.01) difference. Bottom panel – Data plots demonstrate an increase in the number of small (4-cell) and medium (8- to 12-cell) colonies accompanied by a decrease in the number of large-sized colonies (>25-cell) in BST2 siRNA-treated Matrigel cultures. D. Stratification of hormone-treated ER+ breast cancer (n = 66) based on BST2 transcript levels. The Kaplan Meier plot suggests a trend whereby a relatively poor clinical outcome is conferred upon cases with moderate (red) or high (green) BST2 expression compared to those with low gene expression (blue).
Figure 6.
Resveratrol attenuates endogenous BST2 expression and disrupts downstream functional effects in high grade breast cancer cells.
A. Tumor cells (CCdl54) treated with resveratrol for 24 hrs displayed a significant decline in BST2 expression denoted by asterisks (p<0.01). QPCR analysis performed in triplicate, and normalized to ACTB expression. B. Enhancement of apoptotic cell death induced by 24 hr tamoxifen treatment of 6-day resveratrol (20 µM) supplemented tumor cultures plated in Matrigel. Top panel – immunolocalization of cleaved caspase 3 (green) in apoptotic cells; nuclei counterstained with PI (red). Bar - 25 µm. Bottom panel – Immunostaining data plotted as apoptotic index represents proportion of PI-stained nuclei (>100) vs. immunopositive cells in multiple microscopic fields scanned with a 20× objective, and averaged. Asterisk denotes significant (p<0.01) differences in the apoptotic index in the presence of tamoxifen alone vs. resveratrol pretreatment followed by tamoxifen. C. Resveratrol-mediated suppression of the cell cycle in high grade tumor cells (CCdl54, CCdl672) is similar to the effects of BST2 silencing. A statistically significant reduction in S-phase resulting from both approaches for BST2 attenuation is accompanied by an increase in the G1-fraction in independent cell lines. The data derived from BrdU labeling of 2 independent cell lines represents an average of 2 separate FACS runs on each cell line, with each treatment performed in triplicate sets. D. Induction of AP2 expression in high grade tumor cells (CCdl675, CCdl54) treated with increasing concentrations of resveratrol. QPCR analysis performed in triplicate, and normalized to ACTB expression. Asterisks indicate significant differences (p<0.01) between control and treated samples.