EHD2 is a Predictive Biomarker of Chemotherapy Efficacy in Triple Negative Breast Carcinoma

Abstract

EHD2 is a mechanotransducing ATPase localized in caveolae invaginations at the plasma membrane. EHD2 has recently been associated with several human cancers, however the significance of EHD2 transcript levels in cancer prognosis remains debated. Breast cancer is the most commonly occurring cancer in women and prognosis is variable depending on the subtypes. Triple negative breast cancer (TNBC) often has a poor therapeutic response. The aim of this study was to assess the prognostic significance of EHD2 transcripts and protein expression levels in breast carcinomas. We found that low EHD2 levels were associated with enhanced proliferation, migration and invasion of TNBC cells. EHD2 expression was significantly reduced in TNBC tissues and the loss of EHD2 led to higher expression of the pro-tumoral cytokine IL-8. In apparent contradiction with in vitro data, multivariate analysis of two independent cohorts of breast cancer patients revealed that low EHD2 was in fact associated with good prognosis in the highly proliferative TNBC subtype. Accordingly, TNBC low EHD2 expressers were found to benefit the most from chemotherapy when compared to all subtypes of breast cancers. Our study validates EHD2 expression level as an independent prognostic factor of metastasis-free survival and as a new predictive marker of chemotherapy efficacy in TNBC patients.

Figure 1

Low EHD2 expression is associated with TNBC patients. (a) Representative EHD2 immunohistochemistry staining on sections of human breast tumor (TMA). Scale bar = 100 μm. (b) Quantification of EHD2 H-score (immunohistochemistry intensity multiplied by percentage of positively stained cells) in luminal, HER2+ and TNBC tissues. (c) Quantification of EHD2 H-score (immunohistochemistry intensity multiplied by percentage of positively stained cells) in stroma cells of luminal, HER2+ and TNBC tissues. (d) EHD2 protein levels and standard clinicopathological parameters in a series of 377 breast cancers (TMA). Underexpressed EHD2 corresponds to EHD2 H-score <50 whereas overexpressed EHD2 correspond to EHD2 H-score ≥50. ns = non-significant; *P < 0.05; **P < 0.01; ***P < 0.001; (b) Bonferroni’s multiple comparison test; mean ± s.e.m.

Figure 2

EHD2 expression in breast epithelial and cancer cell lines. (a) EHD2 mRNA levels in 22 breast cancer cell lines representing various human TNBC (blue) and normal breast tissue cell lines (white). *Indicates cell lines selected for experiments. (b) Immunoblot analysis for EHD2 expression in 16 breast cancer cell lines representing human TNBC and normal breast tissue cell lines. (c) The correlation between EHD2 mRNA and EHD2 protein in various TNBC and normal breast tissue cell lines. Pearson’s correlation coefficients for TNBC and normal breast cell lines were scored; *P < 0.05; (c) two tailed t-test; mean ± s.e.m.

Figure 3

EHD2 downregulation is associated with breast cancer cell aggressiveness. (a–c) Representative transmitted light images (left) and quantification (right) of cell migration using a wound healing assay in Hs578T cells EHD2 depleted (siEHD2) or not (CTRL) (a), in MDA-MB-231 cells overexpressing EHD2 or not (CTRL) (b), and in MDA-MB-436 cells overexpressing EHD2 or not (CTRL) (c). Scale bar =10 μm. Cell migration into the wound site was assessed after 16 h. (d–f) Quantification of invasion using Transwell chamber inserts in Hs578T cells EHD2 depleted (siEHD2) or not (CTRL) in the absence (negative control) or presence of serum (d), in MDA-MB-231 cells overexpressing EHD2 or not (CTRL) (e), and in MDA-MB-436 cells overexpressing EHD2 or not (CTRL) (f). (g) Measurement of cell proliferation of the Hs578T cells EHD2 depleted (siEHD2) or not (CTRL), and in MDA-MB-436 cells overexpressing EHD2 or not (CTRL). (h,i) Quantification of IL-8 mRNA levels in Hs578T cells EHD2 depleted (siEHD2) or not (CTRL) (h), and in MDA-MB-231 cells overexpressing EHD2 (i); n ≥ 3 independent experiments; ns = non-significant; *P < 0.05; **P < 0.01; (a–c,e–i) two tailed t-test; (d) Bonferroni’s multiple comparison test; mean ± s.e.m.

Figure 4

Low EHD2 mRNA expression is associated with good prognosis in TNBC and is a new independent prognostic factor of MFS. Analysis of MFS in (a) 101 or (b) 228 patients with TNBC was determined as the interval between initial diagnosis and detection of the first metastasis. Survival distributions were plotted using the Kaplan–Meier method, and the significance of the difference was ascertained with the log-rank test using optimal cutoffs (a: 0.18, b: 0.15). Patients with high EHD2 mRNA expression have a significantly poorer prognosis compared with patients with reduced EHD2 mRNA expression (a: P = 0.0066, b: P = 0.017). (c) Multivariate COX analysis of MFS for EHD2 mRNA expression in the series of (c) 101 or (d) 228 patients with TNBC.

Figure 5

EHD2 expression levels predict chemotherapy efficacy in TNBC. (a) Analysis of MFS in 236 patients treated with chemotherapy was determined by the interval between initial diagnosis and detection of the first metastasis. Survival distributions were plotted using the Kaplan–Meier method, and the significance of the difference was ascertained with the log-rank test using optimal cutoffs (0.385). (b) MFS of 61 patients with TNBC treated with chemotherapy was determined by the interval between initial diagnosis and detection of the first metastasis. Survival distributions were plotted using the Kaplan–Meier method, and the significance of the difference was ascertained with the log-rank test using optimal cutoffs (0.3955). Patients with high EHD2 mRNA expression had a significantly poorer prognosis compared to patients with reduced EHD2 mRNA expression (P = 0.027). (c) Multivariate COX analysis of MFS for EHD2 mRNA expression in the series of 61 patients with TNBC treated with chemotherapy.