Introduction
Tumor progression is a highly complex process that involves both intrinsic genetic changes and extrinsic signals mediated via specialized tumor microenvironments [
1]. Tumor cells can actively recruit a variety of cell types, including mesenchymal stem cells (MSCs), into the tumor microenvironment [
2‐
4]. MSCs are pluripotent progenitor cells that contribute to the maintenance and regeneration of a variety of connective tissues, including bone, adipose, cartilage, and muscle [
5,
6]. Accumulating evidence suggests that the cross-talk between MSCs and breast cancer cells may impact tumorigenesis, growth, and metastasis [
7‐
9]. Previously, we co-cultured adipose-derived (AD)-MSCs with the breast cancer cells MCF-7 in vitro and found that the co-cultured MCF-7 cells displayed increased migratory and invasive capacities and underwent epithelial-mesenchymal transition (EMT) changes in a time-dependent manner [
10]. Further investigation of the molecular mechanism underlying the MSC-tumor interplay may facilitate the development of new interventions targeting tumor growth and metastasis.
Long noncoding RNAs (lncRNAs) are defined as RNA transcripts longer than 200 nucleotides that lack apparent protein coding potential. Genome-wide analyses based on the basis of GENCODE annotations have identified approximately 14,000 lncRNA genes in the human genome [
11]. Numerous studies have revealed that lncRNAs play important roles in multiple biological processes, including development, differentiation, and carcinogenesis. For example, HOTAIR induces breast cancer metastasis by operating as a tether that links EZH2 (PRC2) and LSD1, thereby coordinating their epigenetic regulatory functions [
12,
13]. lncRNA-ATB mediates TGF-β-induced EMT and colonization by competitively binding the miR-200 family and IL-11 mRNA, thus promoting the invasion-metastasis cascade in hepatocellular carcinoma [
14]. MALAT1 is a conserved non-coding transcript whose expression has been associated with metastasis and reduced survival in patients with multiple tumor types [
15]. More importantly, MALAT1 is not required for normal development, which makes it a potential therapeutic target against cancer progression [
16]. To date, only a few lncRNAs have been well studied for the cause and progression of cancers.
In this study, we aimed to identify the potential lncRNA deregulations associated with breast cancer malignancy instigated by MSC-MCF-7 co-culture. We profiled the expression changes of lncRNAs in MCF-7 cells during EMT induced by co-culture with hAD-MSCs and observed an intergenic lncRNA KB-1732A1.1 (transcript uc003ykw.2, termed LincK) that was significantly elevated. Linck partly overlapped with lncRNA GASL1, which was found to inhibit cell cycle in some cancer cells [
17]. Here, we reported for the first time the role of LincK in controlling EMT and proliferation in breast cancer.
Materials and methods
Cell lines
The cell lines MCF-7, 293 T, MDA-MB-453, and MDA-MB-231 were directly obtained from the Cell Resource Center, Peking Union Medical College (which is the headquarters of the National Infrastructure of Cell Line Resource, NSTI).
Isolation and expansion of adipose-derived hAD-MSCs
Human adipose tissue was obtained from patients undergoing tumescent liposuction according to the procedures approved by the Ethics Committee at the Chinese Academy of Medical Sciences and Peking Union Medical College. Isolation of hAD-MSCs was followed by procedures as previously reported [
5].
Cell-cell co-culture and enzyme-linked immunosorbent assay
A Transwell® chamber with a 0.4-μm pore size permeable membrane (Corning) was used to co-culture the MCF-7 cells, MDA-MB-453, or MDA-MB-231 with the hAD-MSCs at a ratio of 1:1. The cells were continuously passaged when needed.
To detect the secretion of interleukin-6 (IL-6), TGF-β, and VEGF, hAD-MSCs with or without co-culture with MCF-7 for 7 days were plated in 6-well plates at a cell density of 2 × 105/ml. After incubation in serum-free medium for 24 h, the supernatant was collected for analysis using enzyme-linked immunosorbent assay (ELISA) kits (NeoBioscience, China) according to the manufacturer’s instructions.
Microarray and data analysis
Total RNA was isolated using Trizol from MCF-7 cells cultured alone (0 day) and co-cultured with hAD-MSCs for the indicated times (2, 4, 6, 8, and 10 days). Sample processing and hybridization were conducted as described in a previous report [
18,
19].
For the siRNA-mediated knockdown of LincK in MCF-7 cells, RNA was extracted 48 h after si-NC (Negative Control), si-LincK1, and si-LincK2 transfection. Three biological replicates were done. A fold change ≥ 2 in its expression level and a p value ≤ 0.05 were chosen as the cut-off criteria. All microarray data were uploaded to Gene Expression Omnibus (accession number, GSE109007 and GSE109008).
siRNA, plasmids, and virus infection
siRNAs used to knockdown target lncRNAs or mRNAs were designed by the online tool (BLOCK-iT™ RNAi Designer) and synthesized by GenePharma (Suzhou, China). For overexpression, the full length of LincK was inserted into lentivirus expression vector PCDH-CMV-MCS-EF1-puro (termed as LV-Control thereafter). For knockdown, lentivirus shRNA expression vectors targeting the same sequences as siRNAs were constructed and packaged by GenePharma.
In vitro migration and invasion assay
For MCF-7 and MDA-MB-453 (with LincK overexpression, knockdown, or their corresponding controls), tumor cells were co-cultured with hAD-MSCs for 2 weeks before being subjected to the migration and invasion assay. Then, tumor cells were resuspended in 200-μl serum-free medium at a density of 1 × 106/ml and seeded into the upper chamber of 24-well Transwell chambers (8-μm pore, costar) coated without (migration) and with (invasion) Matrigel (BD Biosciences). The lower chambers were filled with 600 μl of medium containing 20% FBS. After 24 h (migration) or 36 h (invasion), cells on the lower surface of the inserts were stained with 0.1% Crystal Violet.
For MDA-MB-231 cells, 5 × 104 cells were added into the top chamber and permitted to migrate for 8 h; 2 × 105 cells were seeded into the top chamber coated with Matrigel and permitted to invade for 24 h. Three randomly selected fields per filter were counted.
For MCF-7 and MDA-MB-231 cell lines, 2000 cells were suspended in 5 ml complete medium and seeded in a 60-mm dish. For MDA-MB-453 cell line, 2 × 104 cells were cultured in a 6-mm dish. After 2 weeks, colonies were stained by 0.1% crystal violet. Photos of colonies were taken by Cannon EOS 600D and number of colonies were analyzed by ImageJ software.
RNA extraction and quantitative reverse transcription-polymerase chain reaction
Total RNA was extracted using the Trizol reagent (Invitrogen), and quantitative reverse transcription-polymerase chain reaction (qRT-PCR) analysis of mRNA and miRNAs was performed as we previously described [
5]. All the primer sequences are listed in Additional file
1: Table S1.
Western blot
Western blotting was performed as we previously described [
5]. Antibodies against the following proteins were obtained as indicated: ZEB1, E-cadherin, N-cadherin, ZO-1, Vimentin, and GAPDH (Proteintech, China); PBK, p38 MAPK, and phosphorylated p38 MAPK (Cell Signaling Technology®).
Cell proliferation assay
Cells were plated in 96-well plates (2000 cells/well). Cell proliferation was determined every 24 h for 5 days according to the manufacturer’s instructions. Briefly, 10 μl of MTS (#G3582, Promega) was added to each well. After incubation at 37 °C for 1 h, the absorbance at 490 nm was detected.
BrdU proliferation assay
Cell proliferation was monitored using the BrdU-ELISA kit (#11647229001, Roche) according to the manufacturer’s instructions. Briefly, 1 × 104 cells were plated in 96-well plates for 48 h and then labeled with BrdU for 2 h. After incubation with BrdU antibody-peroxidase (POD), photometric detection was performed at 370 nm wavelength.
Northern blot
Northern blots were performed using the DIG Northern Starter Kit (#12039672910, Roche) as we described previously [
20]. Digoxigenin (DIG)-labeled LNA probes were designed using online software (Stellaris probe designer) and synthesized by Exiqon.
5′ and 3′ rapid amplification of cDNA ends
The transcriptional initiation and termination sites of LincK were detected using the FirstChoice RLM-RACE Kit (#AM1700, Ambion) according to the manufacturer’s instructions. The primer sequences are listed in Additional file
1: Table S1.
Subcellular fractionation
The separation of the nuclear and cytosolic fractions was performed using the NE-PER Nuclear and Cytoplasmic Extraction Reagents (#78833, Thermo Scientific) according to the manufacturer’s instructions. RNA was extracted, and qRT-PCR was performed to assess the relative proportion in the nuclear and cytoplasmic fractions.
Dual luciferase reporter assay
Wild types of full-length LincK and 3′UTR of ZEB1 and PBK were obtained by PCR or RT-PCR and cloned into the luciferase reporter vector psiCHECK2 (Promega). Mutants were prepared by deleting of 16 base-pair binding sequences of miR-200b. The cells were harvested 24 h after transfection, and Renilla and firefly luciferase activity were analyzed using the Dual-Luciferase® Reporter Assay System (#E1910, Promega).
RNA immunoprecipitation assay
The Ago2-LincK RNA immunoprecipitation (RIP) assay was performed with the EZ-Magna RIP Kit (Millipore). The AGO2 antibody was purchased from Millipore (#03-110, Merck). The Flag-MS2bp-MS2bs-based RIP assay was conducted according to previous reports [
21]. Briefly, 293 T cells were co-transfected with pcDNA3-Flag-MS2bp and LincK-MS2bs or mutant-MS2bs, and the cell lysates were harvested with IP lysis buffer (#87787, Pierce) after 48 h. Anti-FLAG® M2 Magnetic Beads (Sigma) was used to immunoprecipitate Flag-MS2bp fusion protein. The co-precipitated RNAs associated with Flag-MS2bp were extracted by Trizol reagent (Invitrogen) and detected by qRT-PCR.
RNA pull down
Biotin-labeled RNAs were transcribed in vitro with the Biotin RNA Labeling Mix (Roche) and T7 RNA Pol II (Promega). Whole-cell extract, prepared from 1 × 107 cells in 1 ml of RIP buffer (150 mM KCl, 25 mM Tris (pH 7.4), 5 mM EDTA, 0.5 mM DTT, and 0.5% NP-40) containing RNase and protease inhibitors, was mixed with 3 μg of biotinylated RNA and incubated at RT for 1.5 h, then followed by the addition of 50 μl washed Streptavidin Dynabeads to incubate for another 1.5 h. The beads were washed for 5 × 5 min with RIP buffer containing 0.5% sodium deoxycholate. RNA associated with the beads were extracted by Trizol reagent (Invitrogen) and detected by qRT-PCR.
Clinical specimens, tissue microarrays, and RNAscope®
Breast tissue specimens of normal, benign, and invasive breast cancers were obtained with informed consent from patients in Peking Union Medical College Hospital (Beijing, China) during Oct 2016 to Feb 2017. All specimens were collected using the protocols approved by the Ethics Committee of Peking Union Medical College Hospital. TMA were prepared by Zhongshan Golden Bridge Biotechnology CO., LTD (Beijing, China). The RNAscope® probe targeting LincK was designed and synthesized by Advanced Cell Diagnostics, and the expression of LincK was detected using the RNAscope 2.0 HD detection kit (Brown) according to the manufacturer’s instructions [
22].
Xenograft assay in nude mice
For the subcutaneous model, MCF-7 cells with LincK knockdown, LincK overexpression, and the corresponding controls (stably transduced with lentivirus vectors) were mixed with hAD-MSCs at the ratio of 5:1. After washing for three times with PBS, these cells were resuspended in PBS supplemented with 10% Matrigel at a concentration of 6 × 106/100 μL. Female Balb/c nude mice (5–6 weeks old) were then injected with 100 μl of cell suspension subcutaneously. To promote the tumorigenesis, the mice were injected intraperitoneally with 1 mM estradiol cypionate (MCE, # HY-B1100, 100 μl/mouse) on the day of tumor incubation. The tumor sizes were measured every week for 5 weeks, and the volumes (in cubic millimeters) were calculated according to the equation: width2 × length × 0.5.
For metastasis assay (colonization), 1 × 106 MDA-MB-231 cells (shCtrl or shLincK2 in 100 μl of PBS suspension) were injected intravenously into immunodeficiency mice (n = 6). Lungs were subsequently collected at week 7, fixed in 10% formalin, and embedded in paraffin. Serial pathological sections from the embedded tissues were stained by standard HE procedure to detect metastasis.
Discussions
In the current study, we screened for functional lncRNAs in the MSCs-induced EMT by using microarray assay and identified a functional lncRNA (LincK). LincK can be upregulated by co-culture-induced EMT and by other EMT activators such as IL-6 and TGF-β. LincK promotes proliferation and EMT in breast cancer cells in vitro as well as tumor growth in immunodeficient mice. More importantly, LincK was frequently elevated in breast cancer compared with normal breast tissue in clinical samples.
It has been shown that some lncRNAs, dominantly localized in the cytoplasm, can act as a ‘sponge’ via the competitive binding of common miRNA and attenuating the suppression of miRNAs on protein-coding RNAs [
36,
37]. In this study, we reported that LincK functions in a similar way. LincK contained two putative miR-200 binding sites and was observed to compete with their endogenous targets ZEB1 and PBK. By modulating the level of LincK, we observed that the expression of ZEB1 and PBK highly correlated with the expression of LincK. The miR-200 family has been found to be deregulated in multiple types of cancers and play a pivotal role in tumor initiation, maintenance and malignant metastasis [
35,
38,
39].
LincK contains an Alu element in 3′ end. Recent studies have shown that some Alu element-containing lncRNAs, which were largely cytoplasmic and polyadenylated, might be involved in Staufen1 (STAU1)-mediated mRNA decay (SMD) [
40,
41]. During SMD, lncRNA could regulate the degradation of the mRNA target by forming dsRNA structures between the imperfect pairing of their Alu elements. We could not exclude the possibility that the Alu-containing lncRNA LincK plays a crucial role in SMD.
In this study, we used three breast cancer cell lines to investigate the function of LincK in vitro. For enhancing proliferation, LincK displayed a similar role in the three cell lines. However, for promoting EMT, LincK regulated the morphology of three cell lines differently. In MCF-7 cells, autonomous EMT was exhibited after ectopic expression of LincK by lentivirus infection and antibiotic selection for 2 weeks (Fig.
2d). In MDA-MB-453 cells, no obvious morphological changes were observed by simply regulating the expression of LincK. After co-culture with MSCs (Fig.
2g) or treatment with TGF-β (data not shown), the difference in EMT features between LincK overexpression and the controls became apparent. In mesenchymal-like MDA-MB-231 cells, depletion of LincK caused changes in expression of EMT markers while not in the shapes of cells. However, the migration and invasion abilities were clearly correlated with the expression of LincK in the three cell lines. These data demonstrated that LincK consistently promoted EMT progress in BCC, although it triggered morphological changes differently upon cell context.
Using the subcutaneous model, we showed that silencing of LincK dramatically blocked tumor growth and that the ectopic expression of LincK promoted tumor growth. This was consistent with the observation that LincK played a positive role in proliferation in vitro. Although these cells underwent EMT accompanied by increased migration and invasion capacity in vitro, we did not detect lung metastasis in subcutaneous model. There might have been at least three reasons responsible for the discrepancy. First, MCF-7 cells showed relatively low metastasis in vivo [
42]. Second, the occurrence of metastasis in mouse models using human breast cancer cells was low especially when cells were implanted subcutaneously [
43‐
45]. Another possible reason was that LincK was not sufficient enough to trigger metastasis in vivo. Using lung metastasis model by intravenous injection of MDA-MB-231 cells, we found knockdown of LincK significantly impaired the lung metastasis of MDA-MB-231 cells. Other models, such as orthotopic incubation or ventricle injection, are useful for further investigating the function of LincK on metastasis in vivo.
In this study, we used RNAscope® technology to detect the expression of LincK in clinical specimens. Despite limited cases and short-term observation, we proved that LincK was differently expressed among normal and breast cancer tissues. The results obtained from the open database TCGA additionally confirmed our findings. It should be noted that we did not observe a significant relation between LincK expression and metastasis in our clinical samples. All data suggested that it is important to conduct more detailed and long-term investigations on relationships between LincK expression and clinical features including histological subtype, treatment response, and survival.
Recently, Gasri-Plotnitsky et al. reported that lncRNA GASL1, transcribed from the same chromosome position as LincK, functions as tumor suppressor by inhibiting cell proliferation in osteosarcoma and lung carcinoma cells [
17]. In this study, we proved that LincK was upregulated in breast cancer patients and promoted growth and metastasis in breast cancer cells. GASL1 is 1536 base long consisting of four exons, and LincK is 1210 base long consisting of two exons (Fig.
1f). Although partly overlap between GASL1 and LincK, they confer different even opposite effects depending on different tissues or diseases.
Recently, increasing studies have demonstrated that MSCs within the tumor stroma may promote breast cancer progression by secreting cytokines such as CCL5, IL-6, or TGF-β [
7‐
9]. In addition to paracrine cytokines, MSCs can trigger a select set of miRNAs in BCC and promote breast cancer metastasis. However, currently, the lncRNAs underlying how MSCs contribute to tumor pathogenesis remain incompletely understood. Our data demonstrate that LincK plays a role in tumor-hADMSC interaction. Co-culture of tumors and hAD-MSCs provide an easy and useful platform for mimicking tumor microenvironments and can identify important members involved in tumor progress. Given the specific expression pattern and significant role of LincK in tumor growth and metastasis, we propose that LincK is a promising diagnostic biomarker or therapeutic target for human breast cancer. Recently, great advancements have been made in design and modification of lncRNAs inhibitors such as siRNAs, antisense oligonucleotides, and small molecular, which laid the foundation for lncRNA-based cancer therapy. However, there are still some obstacles that need to be overcome before their clinical application, including specific delivery to the target tissue or cell, stability, and toxicity control.
Acknowledgements
We thank the research members at R. Chen’s Lab: Dr. Guifeng Wei, Huaxia Luo, Haiyan Yue, Qinghua Liu, Dan Wang, Yu Sun, and Zhen Fan for their help, collaboration, and scientific support in the study.