LINC01133 inhibits breast cancer invasion and metastasis by negatively regulating SOX4 expression through EZH2

Abstract Mounting evidence highlights long non‐coding RNAs (lncRNAs) as crucial regulators in multiple types of biological processes and contributing to tumourigenesis. LINC01133, located in chromosome 1q23.2, was a recently identified novel lncRNA with a length of 1154nt. It was involved in the development of colorectal cancer and non‐small cell lung cancer. However, its clinical relevance, biological functions and potential molecular mechanism in breast cancer are still unclear. In this study, we found that the LINC01133 expression was significantly down‐regulated in breast cancer samples and was associated with progression and poor prognosis of breast cancer. Further experiments demonstrated that overexpression of LINC01133 inhibited invasion and metastasis in breast cancer both in vitro and in vivo. Mechanistic investigations revealed that LINC01133 repressed SOX4 expression by recruiting EZH2 to SOX4 promoter. Moreover, rescue experiments further confirmed that LINC01133 functional acted as an anti‐oncogene, at least partly, via repressing SOX4 in breast cancer. Taken together, these findings imply that LINC01133 could serve as a novel prognostic biomarker and potential therapeutic target for breast cancer.


| INTRODUC TI ON
Breast cancer, originated from breast tissue, is the most commonly diagnosed malignant tumour and one of the leading causes of cancer-related death among women worldwide. 1,2 Although dramatic advancement has been made in the early diagnosis, and complex treatment such as surgical resection, radio-chemotherapy, endocrine therapy and immunotherapy, the prognosis of breast cancer patients is still poor due to high rate of lethal distant metastasis. 3 Therefore, it is urgent for us to identify novel diagnostic and therapeutic marker and unravel the potential mechanism behind the progression and metastasis of breast cancer to improve the prognosis.
With the rapid development of microarray technique and the whole-genome sequencing technology over the past decades, it is gradually wide accepted that only less than 2% of human genome is protein-coding genes, whereas more than 90% of human genome is transcribed into non-coding RNA without encoding proteins. [4][5][6] Long non-coding RNAs (lncRNAs), a newly discovered subgroup of non-coding RNAs, are transcripts longer than 200 nucleotides without apparent protein-coding ability. 7 LncRNAs were initially considered to be transcriptional noise because of the lack of the ability to code protein. However, increasing evidence has indicated that lncRNAs can act as a powerful transcriptional and post-transcriptional mediator to participate in a large number of physiological and pathological processes, including cell differentiation, cell proliferation, cell apoptosis, cell autophagy, cell invasion and chromosome inactivation. Therefore, ln-cRNAs are implicated in the initiation and development of human cancers. [8][9][10][11][12] LINC01133, located in chromosome 1q23.2, was a recently identified novel lncRNA with a length of 1154nt. It was reported that LINC01133 was involved in colorectal cancer and non-small cell lung cancer. [13][14][15] However, its clinical relevance, biological functions and potential mechanism in breast cancer are still unclear. In this study, we found the LINC01133 expression significantly reduced in breast cancer tissues compared with paired adjacent normal breast tissues.

| RNA extraction and qPCR assays
The total RNA was extracted from tissue samples or cultured cell lines with TRIzol reagent (Invitrogen), according to the manufacturer's protocol. Total RNA (1 μg) was reverse transcribed in a final volume of 20 μL under standard conditions for the PrimeScript RT reagent Kit (TaKaRa). qRT-PCR was performed using SYBR Premix Ex Taq (TaKaRa) to determine LINC01133 and targets expression levels, following the manufacturer's instructions.
Glyceraldehydes-3-phosphate dehydrogenase (GAPDH) was used as an internal reference gene to normalize RNA expression levels between different samples for an exact comparison of transcription levels, and then, the relative expression was calculated using the 2 −ΔΔCt method. qRT-PCR and data collection were performed on an ABI 7500 system (Applied Biosystems). The sequences of PCR primers are listed in Table S1.  Finally, five randomly selected fields were counted by using an inverted microscope.

| Western blot analysis and antibodies
Proteins were separated by 10% SDS-polyacrylamide gel electrophoresis, then transferred to Hybond membranes. We blocked membrances for 1.5 hours at room temperature using 5% fat-free.
Then, primary antibodies purchased from Abcam, including SOX4 (ab80261, 1:1000) and β-actin (ab8227, 1:5000) were incubated overnight at 4°C. And then, the membrane was washed three times with TBST and the secondary antibodies were added for 1 hour at room temperature. Finally, protein was visualized using an enhanced chemiluminescence system and visualized after X-ray film exposure.

| Subcellular fractionation
The PARIS Kit (Life Technologies) was used to conduct the separation of the cytosolic and nuclear fractions in accordance with the manufacturer's protocol.

| Fluorescence in situ hybridization (FISH)
RNA FISH was performed according to the manufacturer's instruc-

| RIP
The Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore) was used to conduct RNA immunoprecipitation (RIP) assays in accordance with the manufacturer's protocol. The EZH2 antibody for this experiment was purchased from Abcam.

| RNA-protein Pull-down assay, Mass spectrometry
We performed this assay adopting Magnetic RNA-Protein Pull-Down Kit (Pierce) according to the manual. Firstly, TranscriptAid T7 High Yield Transcription Kit (ThermoFisher Scientific) was used to yield full length of LINC01133 and labelled using Pierce RNA3′ End Desthiobiotinylation Kit (ThermoFisher Scientific). And then, RNA-bound beads were added to the cell protein lysate for immunoprecipitation and the beads were washed five times. Finally, the proteins were identified by mass spectrometry (MS) and confirmed by western blotting.

| ChIP assays
The EZ-CHIP KIT (Millipore) was used to conduct the Chromatin immunoprecipitation (ChIP) assays in accordance with the manufacturer's protocol. Firstly, cross-linked chromatin was broken to 200-1000 bp fragments using enzymatic digestion. Then, we immunoprecipitate the chromatin adopting anti-EZH2, anti-H3K27me3 and anti-RNA Pol II antibodies. Besides, the normal mouse immunoglobulin G (IgG) was adopted as negative control. Finally, qRT-PCR was performed.

| Mouse tail-vein assay
To assess the effect of sh-LINC01133 on breast cancer metastasis, 1 × 10 6 MDA-MB-231 cells infected with sh-LINC01133 or sh-NC were injected into the tail vein of female nude mice (ten in each group) to establish a lung metastatic model of breast cancer. Four weeks later, the mice were sacrificed and lung metastasis was quantified by counting the number of tumour foci. The micrometastases in the lung were fixed in formalin and sectioned for H&E staining.

| Immunohistochemistry
The tissues were sectioned and subsequently deparaffinized, rehydrated and antigen retrieval. Following blocking with goat serum, SOX4 antibody (1:200; Abcam) was incubated overnight at 4°C and then the tissues were incubated with biotin-labelled secondary antibodies. The sections were stained and visualized by using 3,3′diaminobenzidine (Maixin Biotech) solution. Images were taken with microscope analysis. All sections were scored blindly by at least two pathologists, and scores were obtained.

| Statistical analysis
All statistical analysis was performed with SPSS 21.0 (SPSS). All the data were shown in terms of mean ± SD (standard deviation

| LINC01133 expression is down-regulated in human breast tissues and correlated with lymph node metastasis and advanced TNM stage
The expression of LINC01133 was examined in 74 human breast cancer tissues and adjacent normal tissues using qRT-PCR with normalization to GAPDH. The results showed that the expression of LINC01133 was significantly down-regulated in breast cancer tissues than that in the corresponding non-cancerous tissues ( Figure 1A).

| Downregulation of LINC01133 predicts a poor prognosis and could be considered as an independent biological predictor for overall survival of breast cancer patients
To investigate the value of LINC01133 expression in the prognosis of breast cancer patients, Kaplan-Meier survival analysis and logrank tests were adopted. The results demonstrated that the downregulation of LINC01133 predicted a poor prognosis in breast cancer patients ( Figure 1D). Simultaneously, univariate and multivariate Cox regression analyses further revealed that downregulation of LINC01133, together with lymph node metastasis and TNM stage can be regarded as an independent prognostic indicator for breast cancer patients (Table 2).

| LINC01133 inhibited migration and invasion of breast cancer cells
qRT-PCR analysis was conducted to determine the expression of

| LINC01133 suppresses the metastasis of breast cancer cells in vivo
In order to investigate the role of LINC01133 on breast cancer metastasis in vivo, MDA-MB-231 cells transferred sh-LINC01133 or sh-Nc were injected into the tail vein to establish a lung metastatic mice model ( Figure 4A

| LINC01133 inhibits breast cancer cell metastasis by downregulating SOX4 expression
To identify the potential downstream targets of LINC01133, qRT-PCR was performed to examine the expression of the core transcript factors and cell adhesion molecules that have been confirmed to be essential in the migration and invasion of cancer cell (including SOX4, E2F7, E2F8, OCT4, Snail, FN1, Twist1, slug and β-catenin).
Among these analysed genes, SOX4 mRNA level was found to be the most significantly altered after LINC01133 knockdown ( Figure 5A).
Moreover, western blot analysis further showed that the protein level of SOX4 was markedly increased in response to LINC01133 knockdown MDA-MB-231 and MCF-7 cells ( Figure 5B).
Next, we examined the RNA level of SOX4 and LINC01133 in 40 clinical breast cancer samples. The data of qRT-PCR revealed that the expression of LINC01133 was negatively associated with SOX4 ( Figure 5C). Additionally, the expression of SOX4 was increased in LINC01133 knockdown lung metastatic tumours ( Figure 5D,E), while the expression of LINC01133 was decreased in LINC01133 knockdown lung metastatic tumours ( Figure 5F). These data indicated that the expression of SOX4 was negatively regulated by LINC01133 in breast cancer cells.
It is well-acknowledged fact that SOX4 contributes to the invasion and metastasis in various types of human cancer, including breast cancer. [16][17][18] Based on these observations, rescue assays were performed to validate whether SOX4 was involved in the anti-metastatic function of LINC01133.
After transfection with sh-LINC01133, MDA-MB-231 and MCF-7 cells were cotransfected with si-SOX4 ( Figure 5F). Our results showed that SOX4 knockdown compromised, at least partially, the effects of LINC01133 on breast cancer invasion ( Figure 5G) and cell viability ( Figure 5H,I). Taken together, our results indicated that LINC01133 inhibits breast cancer progression at least in part through suppressing SOX4.

| LINC01133 suppresses SOX4 expression by recruiting polycomb repressive complex 2 to SOX4 promoter
Firstly, we first determined the subcellular localization of LINC01133 to explore the potential mechanisms by which LINC011331 mediate SOX4 expression. The data of both subcellular fractionation ( Figure 6A) and FISH ( Figure 6B)   Therefore, we hypothesized that LINC01133 can suppress the expression of SOX4 in this manner. In order to test this hypothesis, the antibody against EZH2 which was widely acknowledged as an important subunit of the PRC2 complex was adopted to perform RIP assay. Compared with IgG control, a significant enrichment of LINC01133 with EZH2 was found ( Figure 6C). Furthermore, we performed RNA pull-down analysis to explore the interaction between EZH2 and LINC01133 ( Figure 6D,E). And then, to further confirm whether LINC01133 inhibited SOX4 transcription by recruiting  LINC01133 may serve as an important potential target for the diagnosis and treatment of breast cancer. However, further studies are still needed to clarify other potential target of LINC01133 and molecular regulation mechanism in breast cancer.

ACK N OWLED G EM ENTS
This research was supported in part by the National Nature Science

CO N FLI C T O F I NTE R E S T
None.

AUTH O R CO NTR I B UTI O N S
ZWS and CYD designed the study. ZWS, XZ, YL and SJL collated the data, carried out data analyses and produced the initial draft of the manuscript. YZW and CYD contributed to drafting and polishing the manuscript. All authors have read and approved the final submitted manuscript.