The mitotic checkpoint regulator RAE1 induces aggressive breast cancer cell phenotypes by mediating epithelial-mesenchymal transition

The gene RAE1 encodes ribonucleic acid export 1 (RAE1), which is involved in mRNA export and is known to serve as a mitotic checkpoint regulator. In addition, RAE1 haplo-insufficiency leads to chromosome missegregation and early aging-associated phenotypes. In humans, a positive correlation has been found between RAE1 copy number abnormalities and gene amplification in breast cancer cells. However, the precise functional role of RAE1 in breast cancer remains to be determined. An in silico analysis of data retrieved from GENT and cBio-Portal identified RAE1 upregulation in breast cancer tissues relative to normal breast cells. Functional studies of various cell lines showed that RAE1 induced invasive and migratory abilities by regulating epithelial-mesenchymal transition signals. A tissue microarray was constructed to demonstrate the interrelationship between clinicopathological features and RAE1 expression. Immunohistochemistry revealed a positive correlation between RAE1 expression and a high histologic grade. Furthermore, RAE1 overexpression was associated with considerably poorer disease-free survival and distant metastasis-free survival, especially in patients with oestrogen receptor-positive tumours. In summary, RAE1 may be a prognostic marker and therapeutic intervention target in malignant breast cancers.

Scientific RepoRts | 7:42256 | DOI: 10.1038/srep42256 duration among patients with breast cancer 15 . However, the exact roles of RAE1 and related abnormalities in breast cancer remain unclear.
In this study, which aimed to explore the relationship between RAE1 expression and breast cancer progression, we performed functional studies of breast cancer cell lines and analysed the relationships of RAE1 expression with clinicopathological features and prognosis in patients with breast cancer. Through RAE1 overexpression and knockdown studies, we revealed that RAE1 enhanced aggressive breast cancer cell phenotypes by inducing epithelial-mesenchymal transition (EMT) signals. A combined tissue microarray (TMA) and survival analysis revealed the prognostic significance of RAE1 and a positive correlation between RAE1 expression and histologic grade in invasive ductal carcinomas.

RAE1 abnormalities in breast cancer.
To investigate the relationship between RAE1 expression and breast cancer, we analysed in silico data from the Gene Expression across Normal and Tumour tissue database (GENT; http://medical-genome.kribb.re.kr/GENT). A comparison of 271 normal breast tissues with 2,658 breast cancer tissues clearly demonstrated significant upregulation of RAE1 in the latter (Fig. 1a). An analysis of retrieved data from cBio-Portal (http://www.cbioportal.org) specified the classes of these abnormalities. Among 825 evaluated patients with breast cancer, 16% (n = 129) harboured abnormalities in RAE1, including amplification, mRNA up/downregulation, and missense mutations. Of those abnormalities, mRNA upregulation was the most common (120/129, 79%; Fig. 1b). These data demonstrate that RAE1 upregulation might associate with breast cancer development.
The relationship between RAE1 and EMT. After observing that RAE1 induced migratory and invasive abilities in breast cancer cells in our in vitro system, we tested whether the EMT mediated this phenomenon. To  evaluate changes in EMT-related proteins, we performed western blotting for major epithelial (E-cadherin and β -catenin) and mesenchymal markers (vimentin and N-cadherin) in RAE1-overexpressing cells. Notably, epithelial markers were downregulated, whereas mesenchymal markers were upregulated (Fig. 4a). In contrast, the protein levels of E-cadherin and β -catenin were upregulated in RAE1-knockdown cells (Fig. S4). Interestingly, RAE1 overexpression in MCF7 cells induced a morphological change from a normal fibroblast spindle cell shape to a cobblestone-like shape (Fig. 4b). Similar morphological changes were also detected in RAE1-overexpressing T47D cells (Fig. S5a). The opposite morphological pattern was observed in RAE1-knockdown MDA-MB-231 cells (Fig. S5b). Taken together, our in vitro data suggest that RAE1 induces EMT and thus promotes the migration and invasion of breast cancer cells.
Correlation of RAE1 protein expression with clinicopathological features and prognosis in patients with breast cancer. As RAE1 induces EMT and invasion, we examined the potential correlations of clinicopathological features with RAE1 expression in patients with breast cancer. We used breast cancer tissues from 98 patients to construct a TMA, which was subjected to immunohistochemistry (IHC) to evaluate RAE1 protein expression (Fig. 5). Breast cancer tissues with an Allred score of ≥ 3 are typically considered positive. In our samples, although 25.5% (25/98) of breast cancer tissues were defined as RAE1-positive (Table 1), we were unable to find any correlations with various clinicopathological features. However, using an Allred score of 5 as a cut-off value, we observed a positive correlation between RAE1 and a high histologic grade in 16.3% of tissues (16/98) with strong signals (Allred score ≥ 5; Table 2). However, the prognostic impact of RAE1 overexpression could not be determined because of the low number of positive cases in our dataset. Therefore, we examined the effect of RAE1 expression on the prognosis of breast cancer patients using a Kaplan-Meier online tool (http://kmplot.com). Among all patients, higher RAE1 expression was associated with worse distant metastasis-free survival (DMFS; p = 0.00036; Fig. 6a) and overall survival (OS; p = 0.018; Fig. 6b).  Supplementary Fig. 3. (b) A morphological change from a cobblestone shape to an elongated spindle shape was observed in MCF7 cells after inducing RAE1 overexpression. Scale bar = 200 μ m. Among ER-positive patients, high RAE1 expression associated strongly with a poor DMFS (p = 6.3e-05; Fig. 6c) and OS (p = 0.0024; Fig. 6d).

Discussion
This is the first study to investigate in depth the functional roles and clinical significance of RAE1 in breast cancer using both in vitro system and patient samples, together with an in silico analysis. Our integration of in silico resources confirmed the overexpression of RAE1 in patients with breast cancer, in accordance with previous reports that demonstrated RAE1 copy number variation (CNV) and gene expression abnormalities 14,15 . In addition, our confirmation of RAE1-positive patients among a larger population of patients with breast cancer and the relationship of this parameter with clinicopathological data via TMA analysis provides worthwhile information. Our important findings, namely the correlations of RAE1 overexpression with a high histologic grade and poor prognosis and identification of a functional role of RAE1 in cancer cell migration and invasion, strongly support the notion that RAE1 contributes to breast cancer progression.
Our in vitro experiment demonstrated that the expression of EMT markers, such as E-cadherin, vimentin, and N-cadherin, was modulated by RAE1 expression and, moreover, that the phenotypes of various cell lines shifted between epithelial and mesenchymal states depending on RAE1 expression levels. Because EMT enhances the metastatic potential of breast cancer cells, a critical determinant of the prognosis of a patient with breast cancer 16 , our results strengthen the relationship between RAE1 activity and breast cancer aggressiveness. The histologic grades of invasive breast cancers are assigned based on the structures of cancer cells. A high histologic grade indicates a poorly differentiated tumour with insufficient tubule formation. Given the role of E-cadherin in the maintenance of duct formation and epithelial integrity [17][18][19] , our data provide a plausible explanation for the in vivo function of RAE1; specifically, strong RAE1 expression may facilitate an EMT-like switch and act as an unfavourable metastatic factor, leading to an invasive ductal histology and high histological grade.
In addition to the impact of RAE1 gene overexpression, dysregulation of RAE1 can lead to chromosome missegregation 3 , chromosomal instability, and multipolar spindles, with consequent aneuploidy 6,20-22 . As chromosome number alterations, a consequence of chromosome missegregation, are a known hallmark of cancer cells [7][8][9]23 , RAE1 deficiency or loss could contribute to cancer development or progression. However, the low incidence of RAE1 mRNA downregulation or missense mutation events in human breast cancers (Fig. 1b) suggests that RAE1 deficiency might not be a major driving force in this type of cancer. Instead, RAE1 gene amplification and consequent RAE1 overexpression appear to be important risk factors in breast cancers. Recently, studies that  have identified and evaluated the functional consequences of CNV loci have received attention in the field of cancer biology 15,[24][25][26] . Similar to RAE1, amplification and overexpression of BUB3, a mitotic checkpoint regulator that shares extensive sequence homology with RAE1 3 , were found to correlate with the luminal A subtype of breast cancer 26 . Chromosome loci 8p11-12, 11q13-14, and 20q13 in which RAE1 is located, have been identified as amplification hotspots that correlated with poor survival in a group of patients with luminal A subtype breast cancer 15 . Furthermore, the functional significances of several genes, such as LSM1, TACC1, ADAM9, IKBKB, POLB, and FGFR1 at 8p11 and CCND1 and FGF3 at 11q13, have been identified 15 . In particular, ZNF217, which is located on 20q13 near the locus that encodes RAE1, was previously reported as a novel breast cancer gene along with EGFR1, ERBB2, and PPMID 26 . Among the related factors, FGFR1, IKBKB, and ERBB2 have been targeted for therapeutic drug development (PD173074 against FGFR1, PS-1145 against IKBKB, and trastuzumab against ERBB2) 26 .
In conclusion, our results, which demonstrate the functional role of RAE1 in cancer cell migration and invasion via EMT involvement, imply an important role for RAE1 during specific stages of cancer progression and metastasis. Given the findings regarding the relationship between RAE1 overexpression and clinicopathological significance in breast cancer, RAE1 might represent a new therapeutic target and prognostic marker.
Matrigel invasion and migration assays. The Matrigel ™ (BD, Franklin Lakes, NJ, USA) invasion assay was performed as previously described 27 . The migration assay was performed using the same protocol in the absence of Matrigel ™ . For these experiments, 5 × 10 4 cells were placed in each chamber. The same number of cells was used regardless of cell line. After a cell line-dependent 6-72-hour incubation, invading cells were stained with fluorochrome 4′ ,6-diamidino-2-phenylindole (DAPI) and observed via fluorescent microscopy. Acquired images were analysed using ImageJ software (National Institutes of Health, Bethesda, MD, USA).
Tissue microarray and immunohistochemistry (TMA-IHC). From among patients who underwent surgical resection of breast cancer between January 2006 and December 2010 at the National Health Insurance Service Ilsan Hospital, Gyeonggi-do, Korea, 103 patients with well-preserved paraffin-embedded tissue blocks were selected because sufficient tumour for TMA construction was available in a single tissue block. Data concerning the primary tumour histopathology and patient characteristics were retrospectively obtained by reviewing medical records; five patients who received preoperative chemotherapy were subsequently excluded. The Ethics Committee for the Clinical Research of the Institutional Review Board of the National Health Insurance Service Ilsan Hospital, Gyeonggi-do, Korea, approved this study protocol (2016-07-016). All tissue microarray experiments were performed in accordance with the guidelines and regulations set by the ethics committee. Furthermore, all patient samples were collected after obtaining informed consent. TMA construction and IHC analysis were performed as previously described 27 . For IHC, a primary antibody against RAE1 (Abcam) was used. Immunostaining signals were scored using the Allred scoring system.
In silico analysis. The Gene Expression across Normal and Tumour Tissue (GENT) web-accessible database (http://medical-genome.kribb.re.kr/GENT) was used to evaluate RAE1 expression patterns in breast cancer tissues. The web-accessible database cBioPortal (http://www.cbioportal.org) was also used to evaluate RAE1 abnormalities in breast cancer tissues. A Kaplan-Meier plotter (http://kmplot.com) was used for the survival analysis.
Statistical analysis. Clinicopathological variables were compared using the chi-square test or Fisher's exact test. All statistical tests were two-sided, and p-values of < 0.05 were considered statistically significant. SPSS for Windows, version 23.0 (SPSS Inc., Chicago, IL, USA) was used for the statistical analyses.