Increased SIX‐1 expression promotes breast cancer metastasis by regulating lncATB‐miR‐200s‐ZEB1 axis

Abstract Patients with advanced breast cancer (BC) showed a higher incidence of regional and distant metastases. Sine oculis homeobox homolog 1 (SIX‐1) has been confirmed to be a key tumorigenic and metastatic regulator in BC progression. Yet, molecular mechanisms behind SIX‐1‐induced BC metastases remain largely unknown. Here we found that SIX‐1 was frequently up‐regulated in BC and correlated with poor outcomes when tested in human BC tissue microarray. Then, we manipulated the expression of SIX‐1 by via shRNA‐mediated knockdown and lentivirus‐mediated overexpression. Transwell assay in vitro and lung metastases model of nude mice in vivo showed that SIX‐1 promoted BC cell invasion and migration in vitro, and facilitated metastases in vivo. Mechanistically, SIX‐1 could promote the transcription of lncATB, which exerts critical pro‐metastatic role in BC by directly binding to the miR‐200 family, especially for miR‐200c, to induce EMT and promote metastases. In conclusion, SIX‐1 exerts its pro‐metastatic role in BC through lncATB/miR‐200s axis of EMT signalling pathway and could act as an important diagnostic marker as well as a significant therapeutic target for clinically advanced BC.

confirmed that SIX-1 served as a metastatic regulator that associated with transforming growth factor β (TGF-β)-induced EMT in BC. 12,13 However, the exact mechanism of SIX-1 promoting the invasion and metastasis of BC is still largely unclear.
Emerging evidences suggested that dysregulation of microRNAs (miRNAs) plays a key role in tumorigenic and metastatic by controlling downstream genes. [14][15][16] Among which, miR-200 family has been recognized to be critical in the process of cancer metastasis. 17 The miR-200 family (miR-200s) possess five members organized as two clusters, miRs-200b/a/429 and miRs-200c/141. 18 Previous studies demonstrated that miR-200 family members could inhibit EMT by directly targeting zinc-finger E-box-binding homeobox factor 1 (ZEB1) and zinc-finger E-box-binding homeobox factor 2 (ZEB2). In turn, ZEB1 and ZEB2 could suppress the miR-200s promoter activity by binding to the E-box elements, making a feedback loop with miR-200s to promote EMT process. [19][20][21] Recent reports proved that several transcription factors and long non-coding RNAs (lncRNAs) could control miR-200s expression in metastases. [22][23][24] However, the specific transcription factors or lncRNAs that could control miR-200s expression and regulating EMT to promote metastasis are not well studied.
In this study, we found that SIX-1 was frequently up-regulated in BC and correlated with poor outcomes. In addition, SIX-1 promoted BC cell invasion and migration in vitro, and facilitated metastases in vivo. Mechanistically, SIX-1 could indirectly promote the transcription of lncATB, which exerts critical pro-metastatic role in BC by directly binding to the miR-200 family, especially for miR-200c, to induce EMT and promote metastases. Furthermore, SIX-1 may act as an important diagnostic marker as well as a significant therapeutic target for clinically advanced BC.

| Human tissues collection and treatment
Tissue microarray containing 45 pairs of BC tissues and paired nontumour breast tissues were purchased from Outdo Biotech Co. Ltd.
A total of 183 of BC tissues and 20 non-tumour breast tissues were randomly obtained with informed consent from surgery patients of BC (First Affiliated Hospital of Xi'an Jiaotong University, Xi'an, Shaanxi, China). Patients enrolled for this study were not received any chemotherapy, radiotherapy or targeted therapy. The tissues collected were snap-frozen in liquid nitrogen immediately for further RNA and protein extraction. All the human studies were performed according to ethical consent granted from the Clinical Research Ethics Committee of the Xi'an Jiaotong University. and were cultured under normal conditions at 37°C with 5% CO 2 . BC cells were cultured in DMEM medium (Gibco) with 10% foetal bovine serum (Gibco, USA) and 1% antibiotics (Gibco). Cells were tested for mycoplasma contamination using mycoplasma detection PCR and were found to be negative for mycoplasma contamination . For cell   transfection, miR-200a, miR-200b, miR-200c, miR-141 and miR-429 mimics, ZEB1 siRNA and their corresponding negative controls were synthesized by RiboBio. MCF-7 or MDA-MB-231 BC cells were seeded into 6-well plates (1 × 10 6 cells/well) and transfected with miR-200a, miR-200b, miR-200c, miR-141 and miR-429 mimics (50 nmol/L), siRNA (100 nmol/L) or the corresponding negative control. SIX-1 overexpression and knockdown lentiviral vectors and their corresponding control vectors were constructed by GenePharma. All the transfection procedures were conducted following the manufacturer's protocol.

| RNA extraction and quantitative real-time RT-PCR
Tissues and cells were collected, and Trizol (Invitrogen) was used for total RNA isolation according to the manufacturer's instructions. For messenger RNA (mRNA) expression assay, complementary DNA (cDNA) was obtained from 1.0 μg of total RNA using the PrimeScriptTM RT reagent Kit (Takara Biotechnology, Dalian, China).
Real-time PCR was performed in triplicate with the diluted cDNAs by using iTaq Universal SYBR green PCR system (Takara Biotechnology).
Quantitative data of mRNA expression were calculated by using β-actin as an internal control. For microRNA (miRNA) expression assay, real-time PCR was performed in triplicate, and quantitative data of miRNA expression were calculated using U6 as an internal control.

| Wound healing assay
For wound healing assay, cells (5 × 10 5 ) were seeded into 6-well plates until it reaches a confluence of 90%-95%. Then, the straight wounds were generated by using a 200-μL plastic sterile tips, and the wound closure was recorded and calculated by using a microscope. Cell migration was analysed using NIH ImageJ software.

| Transwell migration and invasion assays
In vitro functional studies of cell migration and invasion were evaluated by transwell migration and invasion assay. Cell migration and invasion capacity were also determined by transwell migration assay by using an 8μm pore size transwell chambers (Corning). Briefly, for migration assay, cells (5 × 10 4 ) were seeded in the upper chamber covered with serum-free medium, while medium supplemented with 20% FBS was added into the lower chamber. After 24 hours of incubation, cell was fixed and stained with 0.1% crystal violet, while cells still reserved on the upper membrane were removed. Cell invasion assay was performed in the same manner but with matrigel covered in the upper chamber previously. The images were taken with inverted microscope (CX41, Olympus) and analysed using NIH ImageJ software.

| In vivo experiments
Male nude mice (4-week-old, n = 10/group) were purchased from Beijing HFK Biotechnology Co, Ltd. Animal study was approved by the Ethics Committee of the Xi'an Jiaotong University. In vivo F I G U R E 1 SIX-1 was up-regulated in BC. A, Representative images of immunohistochemical (IHC) staining of SIX-1 staining in 45 BC tissues and 45 adjacent normal tissues. B, Chi-square analysis of the SIX-1 level in 45 BC tissues and 45 adjacent normal tissues. **represents Kruskal-Wallis test P < .01. C, Kaplan-Meier curve depicting the overall survival (OS) of 1402 BC patients, the relapse-free survival(RFS) of 1754 BC patients and the distant metastasis-free survival (DMFS) of 1746 GC patients. The data were analysed by a Kaplan-Meier plotter (http://kmplot.com/analy sis/). The programme was allowed to automatically determine the best cut-off between expression groups, and samples were censored at the 'all' threshold metastases were evaluated by injecting cells (3 × 10 6 ) through tail vein of the nude mice. After 8 weeks of injection, mice were killed and the lung were removed and fixed. Number and size of metastatic tumour in the lung were calculated by pathologists after haematoxylin and eosin (HE) staining.

| Immunohistochemical (IHC) staining
Immunohistochemical analysis for SIX-1 expression of BC was performed according to standard protocols. Briefly, BC tissue microarray was deparaffinized, rehydrated, washed, the antigen was retrieved, and the endogenous peroxidase activity was blocked.
Then, the slide was incubated with anti-SIX-1 primary antibody overnight and then with biotinylated secondary antibody for 1 hour. Subsequently, the slide was stained with the diaminobenzidine (DAB) and haematoxylin. The histological scores were evaluated and defined by experienced pathologists in blind as previously described. 26

| Immunofluorescence (IF) staining
Cells were cultured and fixed, and then incubated with primary antibodies, followed by incubation with according secondary antibody.

| Statistical analysis
All statistical analyses were performed by using the GraphPad Prism version 5.0 (GraphPad Software, CA) and SPSS 20.0 software (SPSS Inc). Student's t test, Pearson chi-square test, Pearson correlation analysis, Kaplan-Meier analysis and log-rank test were used for comparisons as indicated. A P value < .05 was considered statistically significant.

| SIX-1 was frequently up-regulated in BC, and correlated with poorly outcomes
To clarify whether progression of BC correlated with dysregulation of SIX-1, we first detected the expression of SIX-1 on a tissue chip by immunohistochemistry (IHC), which contains 45 pairs of BC tissues and paired non-tumour breast tissues. As determined by IHC staining, SIX-1 expression was significantly up-regulated in BC tissues compared with paired non-tumour breast tissues ( Figure 1A,B). Importantly, up-regulation of SIX-1 showed strong association with lymph node metastasis and AJCC stage in BC patients (Table 1). These data suggested that high expression of SIX-1 in BC may contribute to BC progression. To further draw attention to whether increased expression of SIX-1 affects survival probability of BC patient, we screened a public database (http://kmplot.com/analy sis/index. php?p=servi ce&cance r=breast) and confirmed that the overall survival (OS), recurrence-free survival (RFS) and distant metastasis-free survival (DMFS) of BC patients were longer in the low-SIX-1-expression group than in the high-SIX-1-expression group ( Figure 1C), suggesting that SIX-1 plays a pivotal role in BC development.

| SIX-1 promoted invasion and migration of BC cells in vitro and facilitated metastases in vivo
To verify the pro-metastatic role of SIX-1 in BC, we overexpressed  in MCF-7 cell line. In MDA-MB-231 cell line, SIX-1 knockdown led to 11.2% (Sh-1) and 13.2% (Sh-2) area being healed, compared to 60.0% of empty vector set (P value < .05) ( Figure

| SIX-1 promoted BC progression by regulating EMT
Previous studies showed that SIX-1 usually exerts its pro-metastatic ability by regulating EMT. 13,28 Morphological assay showed that Representative images of the experiments are shown. *represents Student's t test *P < .05 and **P < .01. G, Indicated cells were injected into nude mice (n = 10 for each group) via the tail vein and animals were sacrificed at 8 wk after the injections. Representative HE staining of lung tissue samples is shown. H, The number of lung metastatic foci observed in each group. *represents Kruskal-Wallis test P < .05, **P < .01 and ***P < .001 ZEB1 ( Figure 3D,F). These data taken together confirmed that SIX-1 contributes to BC metastases at least partly by inducing EMT.

| SIX-1 induced EMT through controlling TGF-β and ZEB1 expression
To explain in-depth the mechanisms of SIX-1-induced EMT in BC metastases, we performed PCR array assay in the SIX-1 Sh-1 group and NC group in MDA-MB-231 cells or SIX-1 OE group and vector group in MCF-7 cells. In the SIX-1 down-regulation group, 13 genes were up-regulated more than twice, such as OCLN, CAV2, KRT19 and DSC2, and 8 genes were down-regulated, such as TGFB1, ZEB1, FN1 and CDH2 ( Figure 4A). Meanwhile, in the SIX-1 up-regulation group, the 8 genes were up-regulated more than twice, such as FN1, VIM, TGFB1 and ZEB1, and 17 genes were down-regulated more than twice, such as VSP13A, DESI1, CAV2 and TCF3 ( Figure 4B). It was known that TGF-β1 and ZEB1 can regulate the EMT process of various cancer cells. 29,30 The changes of these two genes in BC cells were consistent with SIX-1, so we selected TGFB1 and ZEB1 for further study. PCR analysis also verified that both TGF-β1 and ZEB1

| lncRNA ATB was up-regulated by TGF-β1 and was associated with the miR-200 family
We next evaluated how SIX-1 controlled miR-200 family in BC metastases. Previous study showed that the lncRNA-activated by TGF-β (lncRNA ATB), which was up-regulated by TGF-β1, could promote EMT by competitively binding the miR-200 family. 30 Our data showed that TGF-β1 level increased significantly as SIX-1 expression level up-regulated in BC cells. Therefore, we hypothesis that lncRNA ATB may be able to function as a ceRNA for miR-200 family in SIX-1-induced EMT of BC. To verify our hypothesis, we found that multiple miR-200 family binding sites existed on LncATB ( Figure 6A). For further confirmation, we constructed luciferase reporters containing the 3' 1000nt of lncRNA ATB, which contains wild-type (pmirGLO-lncRNA ATB wt), or the binding sites of miR-  Figure 6F), suggesting that lncRNA ATB is required for SIX-1-mediated miR-200s expression. In addition, transfection with wt-lncRNA ATB could also lead to change of ZEB1 expression but mutant lncRNA ATB not ( Figure 6G), indicating that ZEB1 could also be regulated by lncRNA ATB. Importantly, Western blot assay further confirmed that transfection with wt-lncRNA ATB but not mutant lncRNA ATB in MDA-MB-231-SIX-1-KD cells could reverse the EMT process ( Figure 6H). Transwell assay also showed that the invasion and migration ability of MDA-MB-231-SIX-1-KD cells was reversed when transfected with wt-lncRNA ATB but not mutant lncRNA ATB ( Figure 6I). Altogether, our result confirmed that lncRNA ATB may be able to function as a ceRNA for miR-200 in SIX-1-induced EMT of BC.

| D ISCUSS I ON
By using EMT-related PCR array assay in MCF-7-SIX-1-OE and MDA-MB-231-SIX-1-KD cells, we further found and confirmed that besides TGF-β1, ZEB1 are also responsible for SIX-1-induced EMT in BC metastases ( Figure 3). ZEB1 is an important transcription factor activated during EMT, and ZEB1 activation was associated with poor overall survival in advanced BC. 34 Figure 5I). Besides, miR-200 family showed a global influence on ZEB1-mediated EMT of BC cells ( Figure 5H). In addition, transfected miR-200 family mimics in MCF-7-SIX-1-OE cells could reverse SIX-1-mediated EMT ( Figure 5G). Thus, our data provided strong evidences suggesting that miR-200 family is necessary for SIX-1-induced ZEB1 signalling in BC metastases. However, no explanations are currently available on either how SIX-1 controlled miR-200 family in BC metastases or how TGF-β1 involved in these complex processes. Therefore, we further investigated and payed attention to the lncRNA-activated by TGF-β (lncRNA ATB), which was up-regulated by TGF-β1, has been reported could promote EMT by competitively binding the miR-200 family. 30 Luciferase reporter assay showed a strong interaction between lncRNA ATB and miR-200 family, as evidenced by the activity of the reporter constructs was most significantly reduced when miR-200 family mimic constructs were cotransfected with the wild-type IncATB 3'-UTR reporter into MCF-7 cells ( Figure 6B). Moreover, our results also confirmed that SIX-1 could regulate the expression level of lncRNA ATB via TGF-β1 and that lncRNA ATB is required for SIX-1-mediated miR-200s expression, suggesting that SIX-1/ln-cRNA ATB/miR-200s axis is one of the possible mechanisms that control BC metastases. Meanwhile, we also found that transfection with wt-lncRNA ATB could also lead to change of ZEB1 expression but not mutant lncRNA ATB ( Figure 6G), indicating that ZEB1 could also be regulated by lncRNA ATB. Taken together, we confirmed that lncRNA ATB may be able to function as a ceRNA for miR-200s in SIX-1 induced EMT of BC. We finally emphasized the clinical significance of SIX-1 in BC metastases by assessing the correlation between the SIX-1 expression level and the clinicopathological characteristics, and showed that SIX-1 was a promising diagnosis marker to predict poorly clinical outcomes of advanced BC.
In conclusion, SIX-1 exerts its pro-metastatic role in BC through lncATB/miR-200s axis of EMT signalling pathway and could act as an important diagnostic marker as well as a significant therapeutic target for clinically advanced BC (Figure 8).

ACK N OWLED G EM ENTS
We thank the members of Department of Pathology for histological analysis and technical support of the specimens.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest in this study.

AUTH O R S' CO NTR I B UTI O N S
BW and KW designed the study; LZ, SJ and SY collected the data; SJ, SY, SP and PX were involved in contribution of new reagents or analytical tools; LZ, HC and XL analysed the data; LZ, SJ, XL, KW and BW prepared the manuscript. Male nude mice (4-week-old, n = 10/group) were purchased from

E TH I C S A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
Beijing HFK Biotechnology Co, Ltd. Animal study was approved by the Ethics Committee of the Xi'an Jiaotong University.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.