Expression of transmembrane protein 26 (TMEM26) in breast cancer and its association with drug response

Abstract

We have previously shown that stromal cells desensitize breast cancer cells to the anti-estrogen fulvestrant and, along with it, downregulate the expression of TMEM26 (transmembrane protein 26). In an effort to study the function and regulation of TMEM26 in breast cancer cells, we found that breast cancer cells express non-glycosylated and N-glycosylated isoforms of the TMEM26 protein and demonstrate that N-glycosylation is important for its retention at the plasma membrane. Fulvestrant induced significant changes in expression and in the N-glycosylation status of TMEM26. In primary breast cancer, TMEM26 protein expression was higher in ERα (estrogen receptor α)/PR (progesterone receptor)-positive cancers. These data suggest that ERα is a major regulator of TMEM26. Significant changes in TMEM26 expression and N-glycosylation were also found, when MCF-7 and T47D cells acquired fulvestrant resistance. Furthermore, patients who received aromatase inhibitor treatment tend to have a higher risk of recurrence when tumoral TMEM26 protein expression is low. In addition, TMEM26 negatively regulates the expression of integrin β1, an important factor involved in endocrine resistance. Data obtained by spheroid formation assays confirmed that TMEM26 and integrin β1 can have opposite effects in breast cancer cells. These data are consistent with the hypothesis that, in ERα-positive breast cancer, TMEM26 may function as a tumor suppressor by impeding the acquisition of endocrine resistance. In contrast, in ERα-negative breast cancer, particularly triple-negative cancer, high TMEM26 expression was found to be associated with a higher risk of recurrence. This implies that TMEM26 has different functions in ERα-positive and -negative breast cancer.

Keywords: anti-estrogen resistance; estrogen receptor; integrin beta1; sonic hedgehog; triple-negative breast cancer.

Figure 1. TMEM26 RNA and protein are expressed in ERα-positive and -negative breast cancer cell lines

A. B. ERα-positive (pos.) and -negative (neg.) breast cancer cell lines were examined for TMEM26 RNA expression by Q-RT-PCR (A) and for TMEM26 protein expression by Western blot analysis after proteins had been fractionated (PM = plasma membrane fraction, CE = cytosolic fraction and NE = nuclear fraction) (B). (A) Statistical analyses of Q-PCR data were performed by student's t-test (*p < 0.05). Each bar represents the mean value ± S.D. of at least three independent experiments. (B) TMEM26 protein expression patterns were compared with the expression of various proteins and phospho-proteins (ITGB1 = integrin β1, IGF1R = insulin-like growth factor receptor 1, Her2 = human epidermal receptor 2, AKT, P-AKT = phospho-AKT, ERα = estrogen receptor a, ERK1/2 and P-ERK1/2 = phospho-ERK1/2). The blots were stained with Fast green to check for equal protein loading. C. The specificities of the interactions between the anti-TMEM26 antibody and the three major proteins p40^TMEM26, p44^TMEM26 and p53^TMEM26 were analyzed by the preincubating the anti-TMEM26 antibody with TMEM26 antigen in a molar ratio of ~1:50. For control reasons, the effect of the TMEM26 antigen on the interaction of the anti-Elf-1 antibody with the Elf-1 protein was also studied. D. Following transfection of MCF-7 cells with either the integrin β1-specific siRNA siIB1 or the control siRNA siL, the plasma membrane fraction was analyzed for TMEM26 and integrin β1 expression by Western blot analysis. E. MCF-7 cells were treated with insulin or mock for three days and analyzed for TMEM26 protein expression by Western blot analysis after protein fractionation. (C-E) To check for equal protein loading, proteins remaining in the gel after blotting were stained with Coomassie Blue. F. Immunocytochemical analyses of adherent MCF-7, T47D, BT20 and MDA-MB-231 cells for the expression of TMEM26.

Figure 2. p44^TMEM26 and p53^TMEM26 are N-glycosylated TMEM26 proteins

A. Proteins isolated from MCF-7 cells were either incubated with peptid-N-glycosidase F (PNG) alone or in combination with O-glycosidase (O-gly) and sialidase (sial) or mock-treated at 37°C o/n and analyzed for TMEM26, integrin β1 (ITGB1) or Elf-1 protein expression patterns by the Western blot technique. For comparison reasons, along with these samples, proteins isolated from insulin-treated MCF-7 cells were also analyzed for TMEM26 expression. B. Following treatment of MCF-7 cells with tunicamycin (Tun, 5μg/ml) for o/n, 2 or 5 days, TMEM26 protein expression pattern was determined in the plasma membrane (PM) (left panel), cytosolic (CE) (right panel) and nuclear fraction (NE) (right panel) by Western blot analysis. A, B. To check for equal protein loading, proteins remaining in the gel after blotting were stained with Coomassie Blue (Coom.). Exp. = exposure.

Figure 3. TMEM26 expression is altered in fulvestrant-treated and -resistant ERα-positive breast cancer cell lines

TMEM26 RNA and protein expression in fulvestrant (fulv)-treated MCF-7 cells and in fulvestrant-resistant breast cancer cell lines MCF-7/FulvR, T47D/182^R-1 and T47D/182^R-2 (grown in the absence of fulvestrant) were compared with TMEM26 expression in the corresponding parental cell line by Q-RT-PCR A. or Western blot analysis B, C.. (A) Statistical analyses of Q-PCR data were performed by student's t-test (* p < 0.05). Each bar represents the mean value ± S.D. of at least three independent experiments. (B, C) The expression status of a number of other proteins and phospho-proteins were also analyzed (E-cadherin, ITGB1 = integrin β1, SHH = sonic hedgehog, AKT, P-AKT = phospho-AKT, ERα = estrogen receptor α, ERK1/2 and P-ERK1/2 = phospho-ERK1/2). B, C. To check for equal protein loading, either the proteins that remained in the gel after protein transfer were stained by Coomassie Blue B. or proteins transferred to the membranes were stained by Fast Green C. (PM = plasma membrane fraction, CE = cytosolic fraction and NE = nuclear fraction).

Figure 4. Knock-down of TMEM26 leads to changes in TMEM26 protein expression and to an increase in the integrin β1 level

A. The effect of the TMEM26-specific siRNA siTM on the TMEM26 RNA expression in MCF-7 cells as measured by Q-RT-PCR, siL = control siRNA. B, C. Western blot analyses of the expression of certain proteins and phospho-proteins (TMEM26, ITGB1 = integrin β1, IGF1R = insulin-like growth factor receptor 1, E-cadherin, AKT, P-AKT = phospho-AKT, ERα = estrogen receptor α, ERK1/2 and P-ERK1/2 = phospho-ERK1/2, Elf-1 = Ets-like factor-1) in the plasma membrane (PM), cytosolic (CE) and nuclear fractions (NE) as prepared from siTM- or siL-transfected MCF-7 (B) or MCF-7/FulvR cells (C).

Figure 5. Knock-down of TMEM26 results in a delay of spheroid formation by breast cancer cells A-D, H

MCF-7 (A), MCF-7/FulvR (B), T47D/182^R-1 (C), T47D/182^R-2 (D) and SKBR3 cells (H) were analyzed for their abilities to aggregate in 3D cultures in the presence of the TMEM26-specific siRNA siTM, the integrin β1-specific siRNA siIB1 or the control siRNA siL and in the presence or absence of fulvestrant. As a measure for the size of the spheroid the area occupied by the spheroid was determined as described in Materials and methods. Each bar represents the mean value ± S.D. of at least three independent experiments. E-G. Colony assays were performed for MCF-7/FulvR (E), T47D/182^R-1 (F) and T47D/182^R-2 cells (G) to assess whether siTM affects cell growth. For each condition, the sizes of at least 50 single colonies were determined. Statistical analyses were performed by student's t-test (A-D, H) or Wilcoxon matched pair test (E-G), * p < 0.05. ITGB1 = integrin β1, IGF1R = insulin-like growth factor 1 receptor.

Figure 6. Breast cancer samples show strong differences in TMEM26-specific immunoreactivity

TMEM26 specific immunoreactivity of tumor cells in breast cancer samples were determined by immunohistochemistry. A. Negative immunostaining (staining intensity = 0), invasive carcinoma of no special type, B. weak cytosolic immunostaining (staining intensity = 1) in an invasive carcinoma of no special type, C. strong immunostaining (staining intensity = 3), predominantly cytosolic, in an invasive carcinoma of no special type, D. strong immunostaining (staining intensity = 3), predominantly cytosolic, in a ductal carcinoma in situ; E. strong immunostaining (staining intensity = 3), exclusively membranous, in an invasive carcinoma of no special type, F. strong immunostaining (staining intensity = 3), predominantly cytosolic, in an invasive lobular carcinoma and lobular carcinoma in situ. Bar = 50 μm.

Figure 7. While TMEM26-specific immunoreactivity is not associated with risk of recurrence in ERα-positive cancer, high TMEM26 expression correlates with high risk of recurrence in ERα-negative tumors

Forest plot of risk of recurrence stratified by subgroups. The diamonds represent the point estimates of the hazard ratio (HR). The vertical bars show the estimated 95% confidence intervals. The size of the diamond is proportional to the precision of the estimate. ER = estrogen receptor α, PR = progesterone receptor, Her2 = human epidermal receptor 2, TAM, AI, chemo = treatment with tamoxifen, aromatase inhibitor or chemotherapeutics, respectively.