Differences in Endothelin B Receptor Isoforms Expression and Function in Breast Cancer Cells

Abstract

The endothelins and their receptors are best known for their regulation of the vascular system. Their widespread expression in epithelial cells and their overexpression in some tumors has prompted investigation into their ability to regulate cancer progression. In this study, we assessed the mRNA expression of the major endothelin B receptor gene (EDNRB) isoforms and found differences in both mRNA and protein expression in normal breast cells and breast cancer cell lines. Knocking down the EDNRB gene in breast cancer cells altered invasiveness toward endothelin 3 (ET3), and we observed EDNRB isoform-specific regulation of breast cancer cell invasion and cell signaling, as well as isoform- and subtype-specific differences in breast cancer patient survival. The results reported in this study emphasize the importance of the endothelin B receptor in breast cancer. To our knowledge, this study is the first to clarify the differential expression and roles of specific EDNRB isoforms in breast cancer.

Figure 1

EDNRB mRNA expression in breast cancer cell lines. (A) Schematic of ENDRB splice variants; open boxes are non-coding exon sequences and closed boxes indicate coding exon sequences. * indicates the location of the EDNRB-all primers; red arrows indicate the location of the EDNRB-532 primers; blue arrows indicate the location of the EDNRB-436 primers; green arrows represent the location of the EDNRB-532/409 primers. (B)-(F) Semi-quantitative RT-PCR results from EDNRB-all primers tested with pooled biological triplicates from HMEC, MDA-MB-231, MCF7, BT-549, and ZR-75 cell with (B) EDNRB-all primers (ANOVA p<0.0001); (C) EDNRB-442 primers (ANOVA p=0.0003); (D) EDNRB-532 primers (ANOVA p=0.028); (E) EDNRB-436 primers (ns); and (F) deduced EDNRB-409 (see methods; ns). (G) Semi-quantitative RT-PCR results from EDNRB isoform-specific primers with RNA extracted from uninvolved/normal breast tissue and grade III metastatic invasive ductal carcinoma (n=3); EDNRB-all p=0.015; EDNRB-532 p=0.024; right- The ratio of EDNRB-442/532 from RTPCR analysis shows differences in low-grade vs high-grade IDC. (H) TCGA dataset analysis of EDNRB isoform expression by breast cancer subtype and normal tissue. (2-way ANOVA p<0.0001; Tukey's comparison shows significance between normal and all subtypes for EDNRB-all and EDNRB-436/442 p<0.0001; normal vs all subtypes for EDNRB-532 p≤0.01).

Figure 2

EDNRB protein expression in breast cancer cell lines. (A) Representative Western blot showing EDNRB protein expression in HMEC, MCF-7, MDA-MB-231 and BT-549 cell lines; (B) Semi-quantitative analysis of EDNRB relative to beta-actin (ACTB) in 3 biological replicates (2-way ANOVA p<0.0001); (C) representative flow cytometry charts showing EDNRB cell-surface expression in HMECs (top), MCF7 (middle), and MDA-MB-231 (bottom) cell lines.

Figure 3

EDNRB isoforms differentially regulate breast cancer invasion. Transfecting siRNA specific to EDNRB alongside control siRNA into breast cancer cell lines results in (A) significant reduction in mRNA expression (p=0.027) and a significant increase in in vitro invasion toward ET3 (p=0.02) and (B) significant reduction in mRNA expression (p=0.018) but a non-significant increase in invasion toward ET3 (p=0.18) in MDA-MB-231 cells. Over-expressing mammalian plasmids encoding for (C) EDNRB-442 significantly increased mRNA expression (p=.0002) and significantly decreased in vitro invasion toward ET3 (p=0.008); however, expressing EDNRB isoforms EDNRB-436 (D) and EDNRB-532 (E) significantly increased expression (p=0.0002; p=0.0026) but did not alter invasion toward ET3 (p=0.39; p=0.25, respectively). All graphs are from 3 biological replicates.

Figure 4

EDNRB isoforms differentially regulate cellular signaling. Western blotting of lysates extracted from (A) MDA-MB-231 cells transfected with EDNRB isoforms and (B) MCF-7 cells transfected EDNRB-specific siRNA were probed with antibodies specific for phosphorylated AKT1 and phosphorylated AKT2 at S473 (pAKT1, pAKT2), pan AKT, phosphorylated ERK (pERK), pan ERK, and GAPDH. Bands were quantitated using ImageJ Software and normalized to GAPDH intensity. (A) MDA-MB-231 cells transfected with EDNRB-442 had significantly higher levels of pAKT1 (ANOVA p=0.013) but not pAKT2 (p=0.28) or pERK (p=0.42); (B) MCF-7 cells transfected with EDNRB-siRNA had significantly reduced pAKT1/AKT levels (p=0.097). (C) Analysis of breast cancer cell lines from Cancer Cell Line Encyclopedia data (CCLE) shows a non-significant positive association between EDNRB and active AKT (left); this trend was not observed in non-transformed breast cell lines (right). (D) TCGA breast cancer exon expression data was separated by subtype and analyzed by median EDNRB expression using two isoform-specific probes that recognize EDNRB-442 or EDNRB-532; pAKT levels at both S473 and T308 sites from the RPPA dataset were compared between EDNRB-high and low groups (n=60 for both groups) in basal and (left) luminal A (right) breast cancer subtypes, and (E) normal breast tissue.

Figure 5

EDNRB isoform and ligand expression may predict survival outcomes in basal breast cancers. Survival data from the TCGA dataset was sorted based on average EDNRB survival by exon-specific probe. Basal breast cancers high in EDNRB-532 (n=7) had significantly poorer survival outcomes than basal breast cancers low in EDNRB-532 (n=12) (p=0.03, Student's t-test). (B) Sorting basal breast cancers by combined EDNRB-all and EDN3 RNASeq data showed a significant difference in survival between cancers sorted by low EDNRB/high EDN3 (n=89) vs low EDNRB/low EDN3 expression (n=158) (p=0.013).