Introduction
Breast cancer is a family of diseases that involve unregulated breast epithelial cell growth and division, which is caused by many different carcinogenic factors. The exact cause of breast cancer is unclear. Many risk factors may increase the chance of having breast cancer, such as endocrine disorders, genetic mutations and declines in immune function. However, unregulated mammary epithelial cell proliferation and apoptosis, which are caused by an accumulation of gene mutations and by dysregulated gene expression, is the essential reason for breast cancer. As numerous genes are predicted to be regulated by microRNA (miRNA), mammary tumorigenesis and metastasis is likely to be regulated by several tissue-specific miRNAs.
The
miR-183/-96/-182 cluster is a highly conserved polycistronic miRNA cluster which was first identified by Dr Xu in sensory organs [
1]. Members of this cluster are located within a 5-kb region on human chromosome 7q32.2 and are transcribed in the same direction from telomere to centromere. Previous studies showed that the
miR-183/-96/-182 cluster is abnormally expressed in a variety of tumors and is directly involved in human cancers. But the role of this miRNA cluster in tumors is still unclear. It may function as an oncogene or tumor suppressor gene, depending on the type, location and stage of the cancer. We summarize its reported functions in cancers and its target genes in Table
1.
Table 1
Role of miR-183/-96/-182 in cancer based on recent publications within the last five years
miR-96 | Oncogene | Hepatocellular carcinoma | Increases proliferation and colony formation |
FOXO1, FOXO3a
| |
miR-96 | Oncogene | Prostate cancer | Inhibits zinc uptake |
ZIP1, ZIP3, ZIP7, ZIP9, ZnT1, ZnT7
| |
miR-182 |
miR-183 |
miR-96 | Oncogene | Medullo-blastoma | Inhibits apoptosis, destroys DNA repair, promotes cell migration |
See reference
| |
miR-182 |
miR-183 |
miR-96 | Oncogene | Breast cancer | Induces proliferation |
FOXO3a
| |
miR-96 | Oncogene | Breast cancer | Increases cell number |
FOXO1
| |
miR-182 |
miR-182 | Oncogene | Glioma | Promotes glioma cell aggression |
CYLD
| |
miR-182 | Oncogene | Melanoma | Promotes cell migration and survival |
FOXO3
| |
miR-183 | Oncogene | Synovial sarcoma | Promotes tumor cell migration |
EGR1
| |
PTEN
|
miR-183 | Oncogene | Hepatocellular carcinoma | Iinhibits TGF-beta1-induced apoptosis |
PDCD4
| |
miR-96 | Tumor suppressor | Pancreatic cancer | Decreases cell invasion, migration and tumor growth |
KRAS
| |
miR-183 | Tumor suppressor | Breast cancer | Inhibits migration |
Ezrin
| |
miR-183 | Tumor suppressor | Osteosarcoma | Inhibits migration and invasion |
Ezrin
| |
miR-182 | Tumor suppressor | Lung cancer | Inhibits cancer cell proliferation |
RGS17
| |
The
miR-183/-96/-182 cluster has not yet been extensively studied in breast cancer. Forkhead box O (
FOXO) proteins, which are a family of tumor suppressor transcription factors involved in cell growth, proliferation, differentiation, and longevity, are the main targets for this cluster in breast cancer. Both
FOXO1 and
FOXO3a are regulated by
miR-96 and
miR-182 [
5],[
6]. It seems that this miRNA cluster functions as onco-microRNA in breast cancer. However, in 2010,
Lowery et al. reported that
miR-183 inhibits cell migration in breast cancer by repressing
Ezrin, which plays a key role in cell-surface structure adhesion, migration, and organization [
12]. These conflicting results may be ascribed to two reasons. One possibility is that these three miRNAs are transcribed or processed in different way and they function separately and differently; the other possibility is that this cluster plays different roles in different breast cancer types. In fact, the level of
miR-183 was lower in estrogen receptor (
ER)-positive breast tumors compared to
ER-negative tumors, and higher in human epidermal growth factor receptor-2 (
HER2)/neu-receptor-positive tumors compared to
HER2/neu-receptor-negative tumors [
12], suggesting the roles of
miR-183 in different breast cancer cells are different.
Recently, attention has focused on the target genes of these miRNAs; however, little is known about the regulation mechanism of the miRNA cluster itself. Most miRNA genes are transcribed by RNA polymerase II [
15], which means miRNA biogenesis is controlled elaborately through various regulatory pathways just as protein-coding mRNAs. Chromatin structure analysis, genomic and RNA sequence analysis and RNA polymerase II chromatin immuneprecipitation assays have been applied to predict the transcription start site (TSS) and promoter region of miRNAs [
16]-[
19], but few results have been confirmed by experiments. The Ozsolak [
16], Wang [
18], and Chien [
19] laboratories predicted that the TSS of
miR-183/-96/-182 was 5068 bp, 5200 bp and 5207 bp upstream of the
miR-183 precursor, respectively. However, the promoter region of
miR-183/-96/-182 and the transcription regulators remain unknown.
Here, we investigated the function of the miR-183/-96/-182 cluster in breast cancer. We found that the miR-183/-96/-182 cluster was highly expressed in most breast cancers. These three miRNAs were transcribed in the same pri-miRNA and this miRNA cluster was regulated by HSF2 and ZEB1. We also demonstrated that the miR-183/-96/-182 cluster functioned as an onco-miRNA in breast cancer. Overexpression of the miR-183/-96/-182 cluster increased the cell proliferation rate and promoted cell migration while inhibition of the miR-183/-96/-182 cluster decreased cell growth rate, and even induced cell death. MiR-183 targeted RAB21 directly in breast cancer and accumulated nucleus number aberration cells. Our results suggested that the miR-183/-96/-182 cluster plays an important role in tumorigenesis and in the migration of breast cancer cells.
Methods
Clinical cancer samples and cell lines
All cancer samples were obtained from the Affiliated Tumor Hospital of XiangYa Medical School of Central South University, and stored at -80°C until analyzed. All experiments were conducted in accordance with the Declaration of Helsinki and were approved by the Xiangya Hospital Medical Ethics Committee in Central South University.
Breast cancer cell lines MCF-7,MDA-MB-231,SK-BR-3,T47D, ZR-75-1, MCF-10A and human embryonic kidney cell HEK-293 were used in the study. MCF-7 and MDA-MB-231 were obtained from NeuronBiotech (Shanghai, China). SK-BR-3, T47D, ZR-75-1 and MCF-10A were obtained from Dingguo, Co. (Beijing, China). HEK-293 was obtained from Xiangya experiment center (Changsha, China). All the cells were cultured in complete DMEM high glucose medium (Hyclone, Logan, UT, USA) supplemented with 10% FBS (Hyclone) and 1% penicillin and streptomycin sulfate (Solarbia, Co., Beijing, China). Cells were incubated at 37°C with 5% CO2 and medium was changed every 2 or 3 days.
Virion and cell line constructions
To establish the miRNA overexpression cell lines, partial
mir-96,
mir-182 and
mir-183 pri-microRNA sequences flanked by EcoRI and AgeI restriction sites were inserted into the CMV promoter of lentivirus infectious virions pLKD-CMV-G&PR-U6-shRNA (Hpcoo3) (Additional file
1: Figure S1A). MCF-7 or T47D cells were infected with these viruses and selected under the pressure of 1 μg/ml puromycin (Invitrogen, San Diego, CA, USA). The green fluorescent protein (GFP) signal of the infected cells was detected under microscope (Additional file
1: Figure S1B), and the expression of the
miR-183/-96/-182 cluster in each cell line was measured by reverse transcription (RT)-PCR (Additional file
1: Figure S1C).
To disrupt the activity of the
miR-183/-96/-182 cluster, we generated
miR-183/-96/-182 cluster sponge lentivirus virion. Basically, 10 copies each of complementary sequences to
miR-183,
miR-96 and
miR-182, each with mismatches at positions 9 to 12 for improved stability [
20],[
21], were introduced into the pLOV-CMV-eGFP-EF1a-PuroR lentivirus infective virion (Additional file
2: Figure S2). A moderate multiplicity of infection (MOI) of 1 was used for transduction. The infection efficiency and cell morphology were monitored under microscope every day. After 3 days of transduction, cells were collected for cell cycle analysis and RNAs were collected for real-time PCR.
To research the function of transcription factors, the coding sequences of HSF2 and ZEB1 flanked by XhoI and KpnI restriction sites were inserted into vector GV219. The plasmids were transfected into MCF-7 cells and the cells were selected with a culture medium containing 600 μg/ml G418-Geneticin (GenView, Galveston, TX, USA) for 2 months.
LNA-based Northern Blotting
Total RNAs were extracted from cancer samples with the mirVanaTM miR isolation kit and 10 μg of total RNA was used for each assay. All procedures followed manufacturer's instructions for the miRCURY LNA™ microRNA detection probes (Exiqon, Woburn, MA, USA). After fractionation by electrophoresis on a denaturing 12% polyacrylamide gel containing 8 M urea, RNAs were transferred to Nytran N membrane (Amersham Biosciences, Piscataway, NJ, USA) and fixed by UV crosslinking. Blots were prehybridized for 1 h at 45°C in PerfectHyb™ Plus Hybridization Buffer (Sigma, St Louis, MO, USA) and hybridized overnight at 45°C in hybridization buffer containing 0.1 nM probe, then washed twice for 30 minutes at 65°C in 0.1SSC/0.1% SDS. As the probes were 5'-DIG labeled, we detected the signal by PhototopeR-Star Kit (New England BioLabs Inc, Ipswich, MA, USA), and the densities were quantified by the Image J program. Because the
miR-183, miR-96 and
miR-182 sequences are similar, we tested the probe specificities before doing the experiments (Additional file
3: Figure S3). Mimic oligonucleotides were designed based on miRNA sequences registered in the miRBase Sequence Database (see Additional file
4: Table S1).
RT-PCR and real-time PCR
For mRNA RT-PCR and real-time PCR, total RNAs were extracted from cancer samples or cultured cells with Trigol (Dingguo, Co.) reagent. Primer sets were designed within the exon junction areas listed in Additional file
4: Table S2. For miRNA real-time PCR, miRNAs were extracted from cells using a mirVana miRNA isolation kit (Ambion, Austin, TX, USA). All primers, including the YRBIO™ miRNA qPCR Detection primer sets and U6 snRNA PCR primer set were purchased from Yingrun Biotechnology (Changsha, China).
In brief, mRNA and miRNA were reverse-transcribed with an M-MLV First Strand kit (Invitrogen). Then 50 ng cDNA was mixed with All-in-one™ qPCR Mix (Genecopoeia, Rockville, MD, USA) and the target gene primer set (final concentration: 1 μM for each primer) to produce a 20-μl reaction mixture. All real-time PCR experiments were carried out with an ABI Step One Plus Real-time PCR System (Applied Biosystems, Carlsbad, CA, USA). All real-time PCR reactions were done in triplicates, and the average ΔCT (Δ cycle threshold) for the triplicates was used in subsequent analysis.
Plasmid, miR-Down™ antagomir and transfection
Large-scale plasmids were extracted by PureYield™ Plasmid Midiprep System (Promega, Madison, WI, USA), and small-scale plasmids were extracted by Mini DNA purification kit (Dingguo). Chemically modified antisense oligonucleotides (miR-Down™ antagomir, GenePharm Co. Ltd, Shanghai, China) were used to inhibit miR-96, miR-182 and miR-183 expression. A scrambled oligonucleotide was used as control. Plasmid and miR-Down™ antagomir transfections were conducted with Lipofectamine™ 2000 reagent (Invitrogen).
Luciferase reporter assays
For promoter analysis, promoter region sequences or their mutants flanked by XhoI and KpnI restriction sites were inserted into the upstream region of luciferase reporter gene in pGL3-Basic vector (Promega). MCF-7 cells were transfected with 200 ng reporter construct and 1 μg GV219 vector with or without transcription factor sequence. Also, 40 ng of pRL-CMV-Renilla plasmid was transfected as an internal control.
For target analysis, 33 bp of RAB21 3'-UTRs including the seed sequence were flanked by XbaI and FseI restriction sites and inserted between the Luciferase coding sequence and SV40 polyadenylation element in pGL3-Promoter vector (Promega). HEK-293 cells were transfected with 200 ng reporter construct and 1 μg Hpcoo3 vector with or without partial pri-microRNA sequence of miR-183/-96/-182 cluster. Also, 40 ng of pRL-CMV-Renilla plasmid was transfected as an internal control.
The luciferase reporter assays (Promega) were performed 48h after transfection, and luciferase activity was determined with a GloMax 20/20 Luminometer (Promega). Relative luciferase activities were calculated as ratios of firefly to renilla luciferase activities.
Assays: 3-(4, 5-dimethyl-2-thiazolyl)-2, 5-diphenyl-2H-tetrazolium bromide (MTT)
Cells were seeded on 96-well plates (5 × 103 cells per well) and incubated for 24 h in 0.2 ml medium. After reaction with 20 μl 5 mg/ml sterile MTT (Sigma) for 4 h at 37°C, culture media was removed and 150 μl of dimethyl sulphoxide (DMSO) was added. The absorbance was measured with the ELISA reader (BioTek, Vermont, VT, USA) at 490 nm and 540 nm and the reactions were performed in triplicates.
Cell wound-healing assays
Cells were seeded on 6-well plates (5 × 105 cells per well) and incubated for 24 h. Adherent cell monolayers were scratched with a 10-μl pipette tip and cultured in 2 ml DMEM high-glucose medium without FBS or antibiotics. Cell migration was monitored under microscopy later.
The culture dish was covered by 2 ml bottom gel (0.5% basic agar in RPMI medium 1640 (Invitrogen) supplemented with 10% FBS and 1% penicillin/streptomycin) and 1.5 ml top gel (0.7% agar in RPMI-1640 medium supplemented with 10% FBS and 1% penicillin/streptomycin) mixed with 10,000 cells. Cells were incubated for 16 days and the colonies were stained with 0.5ml 0.005% crystal violet overnight followed by washing with PBS (Hyclone) three times. The pictures of cell colonies were taken by a digital camera.
Cell cycle analysis
Cells were digested with 0.05% trypsin (Thermo Scientific, Logan, UT, USA) for 2 minutes to dissociate them from the plates. After fixation in 70% pre-chilled (−20°C) ethanol in PBS at 4°C overnight, cells were treated with 10 μg/ml of RNase (Auragene, Co., Shenzhen, China) in PBS at 37°C for 2 h and stained with 50 μg/ml of propidium iodide (PI) (Sigma) for 5 minutes. Flow cytometry was conducted on a BD FACSCalibur flow cytometer (BD Biosciences, Franklin, IN, USA) and data were analyzed by ModFit LT software.
Western blotting
Total proteins were lysed in RIPA buffer (150 mM NaCl, 0.1% SDS, 0.5% sodium deoxycholate, 1% NP-40 and 50 mM Tris-HCl, pH 7.6) with a proteinase inhibitor cocktail (Roche, Mannheim, Germany). After separation by 15% polyacrylamide gels and transfer to 0.45 μM membrane (Millipore, Billerica, MA, USA), proteins were detected by anti-RAB21 (Abcam, HongKong, China) and anti-β-tubulin (Sigma) antibodies.
Phalloidin and 4',6-diamidino-2-phenylindole (DAPI) staining
For imaging of fixed cells, cells were seeded on acid-washed, glass coverslips coated with 5 μg/ml of collagen. Cells were then fixed with 3.7% paraformaldehyde in PBS permeabilized with 0.2% Triton X-100 in PBS for 15 minutes. Then we co-stained the cells with fluorescein isothiocyanate (FITC)-conjugated phalloidin (Beyotime, Shangai, China) to detect the F-actin, and with DAPI (Invitrogen) to detect the nuclear. Coverslips were mounted with Microscopy Aquatex® mounting medium (Merck, Darmstadt, Germany), and then detected under the Leica Tcs-sp5-II confocal microscope (Leica, Wetzlar, Germany).
Statistical analysis
Data were expressed as means ± SD, and the statistical software SPSS 11.5 (IBM, Armonk, NY, USA) was used for analysis of variance (ANOVA) and analysis using Student's t-test. Statistical probability (P) in tables, figures, and figure legends are expressed as follows: *P <0.05, **P <0.01, *** P <0.001.
Discussion
The
MiR-183/-96/-182 cluster is a conserved polycistronic miRNA cluster that is highly expressed in several tumor types. Although it is well known that the expression level of this miRNA cluster is increased in breast cancer, its biological roles and the regulatory mechanisms governing
MiR-183/-96/-182 expression in breast cancer are still unclear. Here, we report that
miR-96,
miR-182 and
miR-183 expression levels are significantly higher in breast cancer compared to the NAT, and the transcription pattern of
miR-183/-96/-182 is irregular in breast cancer as the correlation between
miR-182 and
miR-183 expression dropped dramatically in tumor samples. The expression of
miR-183/-96/-182 is not upregulated in a specific breast cancer subtype. It is overexpressed in all kinds of breast cancer - ductal or lobular, luminal or basal, early-stage or late-stage - but there are some differences in their expression patterns. For example,
miR-96 and
miR-183 were lower in lobular carcinoma than in ductal carcinoma and other types of carcinoma. The levels of
miR-96 and
miR-183 were also lower in ER+ and PR+ cancers than in ER− and PR− cancers, but
miR-182 was almost the same, even a little higher in ER+ cancers. Among the four different subtypes of breast cancer,
miR-96 and
miR-183 levels were higher in HER2-enriched breast cancers than other types;
miR-182 was lower but
miR-183 was higher in basal-like breast cancers than other types of breast cancer. We also compared the miRNA expression levels in different breast cancer cell lines based on their molecular markers. We found that
miR-96 is only upregulated in SK-BR-3 and BT-20 cells, whereas
miR-182 and
miR-183 are upregulated in most of the breast cancer cell lines tested except for MDA-MB-231. Basically, the cell line data closely match the clinical analysis.
MiR-96 and
miR-183 levels are higher in HER2-enriched cell line SK-BR-3.
MiR-96 is lower in ER+ and PR+ cancers than in ER- and PR- cancers.
MiR-182 is higher in luminal breast cancer than basal breast cancer. MDA-MB-231 is the only exception. It is an ER- and PR- cancers, but its expression of
miR-183/-96/-182 is low. Because MDA-MB-231 is a basal B/claudin-low breast cancer cell line, which lacks common epithelial cell features and most closely resembles the mammary epithelial stem cell [
26], we think its regulation of
miR-183/-96/-182 is different to other breast cancer cell lines. Our data were similar to those reported by Riaz and colleagues. Based on their work, 51 human breast cancer cell lines were divided into two groups: the first major group included 33 cell lines, which was a luminal-like group; the second minor group included 18 cell lines, which was a basal-like group. Seventeen miRNAs, which included
miR-182, showed significantly higher expression in the major cluster compared with the other miRNAs. They also found that the expression of
miR-183/-96/-182 is low in MDA-MB-231 cells [
23]. Although the
miR-183/-96/-182 cluster is transcribed in the same pri-miRNA, the expression profile of each miRNA varies between different cell lines, which indicate that their subsequent processing or stability are regulated in different ways. An interesting phenomenon is that from the 102 patient samples of TCGA dataset,
miR-182 only increases 4.2 (± 1.1)-fold in tumor samples, but
miR-96 and
miR-183 increase 8.4 (± 1.1)- and 7.5 (± 1.1)-fold in tumor samples. The correlation between the expressions of
miR-182 and
miR-183 dropped dramatically in tumor samples. This phenomenon was also confirmed in HSF2 and ZEB1 overexpression cell lines, as the expressions of
miR-96 and
miR-183 were increased significantly, but not
miR-182. We think it is because the transcription of
miR-183/-96/-182 is very fast in cancer; some pri-miRNA is not complete and the transcription stalls before
miR-182.
We also identified two transcriptional factors that regulate the transcription of the
miR-183/-96/-182 cluster,
ZEB1 and
HSF2. ZEB1, which is a zinc finger transcription factor, is involved in the epithelial-mesenchymal transition and promotes metastasis in cancer [
27],[
28]. Although most work has concentrated on the capacity of
ZEB1 to repress gene expression, several groups demonstrated that
ZEB1 can also activate transcription of downstream targets [
28],[
29].
HSF2 binds heat shock promoter elements (HSE) and activates transcription. Although there is little evidence on the involvement of
HSF2 in tumorigenesis, it can play a role indirectly by modulating
HSF1 [
30]. Previous studies also report that
HSF2 regulates the proto-oncogene c-fos and may be involved in tumorigenesis [
31]. Our findings show that
ZEB1 and
HSF2 activate the transcription of the
miR-183/-96/-182 cluster, which gives us new insights into how
ZEB1 and
HSF2 enhance tumorigenesis.
The biological role of the
miR-183/-96/-182 cluster in breast cancer is complicated. In our experience, this cluster functions more like an oncogene in breast cancer as it increases cancer cell proliferation and migration. Most previous and recent publications support this conclusion, especially for
miR-182, which has been confirmed by many groups to induce breast cancer metastasis [
6],[
32]-[
34].
Mir-96 is also proposed to be an onco-miRNA in breast cancer [
5],[
6], but the role of
miR-183 is more complex. It represses the expression of
EGR1 and functions as an oncogene in breast cancer [
35], but it also targets the
Ezrin gene and inhibits cell migration in T47D cells [
12]. Our results support a pro-oncogenic role for
miR-183 in breast cancer, because upregulated expression of
miR-183 by lentivirus in MCF-7 cells induces cell proliferation and migration. The effects of knockdown of
miR-183/-96/-182 cluster are more complicated, and depend on the knockdown efficiency and specificity. We did not observe obvious changes after inhibition of
miR-183, but we found a significant decrease in cell growth rates and S phase cell percentages in
miR-96 and
miR-182-inhibited cells. Two reasons can explain these results. First, the knockdown efficiency of
miR-183 antagomir is lower than
miR-96 and
miR-182 antagomir. Second,
miR-96 and
miR-182 target
FOXO1, but
miR-183 does not [
6].
MiR-96 and
miR-182 might compensate partial functions of
miR-183, but
miR-183 cannot replace the function of
miR-96 and
miR-182 on inhibition of
FOXO1.
Long-term inhibition of three miRNAs by sponge elements induced cell death and apoptosis in T47D cells, but we did not detect apoptosis with a single antagomir transfection. Inhibition of two or three of the cluster members at one time induced apoptosis, though some of them were not statistically significant (Additional file
8: Figure S5). These data indicate that these three miRNAs are redundant; they may be complimentary to each other. Knockdown of
miR-183 had little effect on its own, but it had collaborative effects with the other two miRNAs.
We identified
RAB21 as a target gene of
miR-183 in both mRNA and protein levels, and also confirmed that overexpression of
miR-183 induced accumulation of bi- and multinuclear cells.
RAB21 is involved in the targeted trafficking of integrins via its association with integrin alpha tails. As a consequence,
RAB21 regulates cell adhesion and migration [
36]. In mitotic cells, integrin trafficking regulated by
RAB21 is necessary for cytokinesis and cytokinesis failure will induce aneuploidy and oncogenic transformation [
25],[
37]. This information may answer the question why
miR-183 has dual effects in breast cancer. In some cases, repression of
RAB21 results in decreased cell mobility, but in other cases, repression of
RAB21 may lead to cytokinesis failure and aneuploidy. The 3'-UTR of
RAB21 matches the seed sequence of
miR-183, but not
miR-96 nor
miR-182. So, only
miR-183 can inhibit the expression of
RAB21. As the phenotype is similar no matter which of the three miRNAs is overexpressed in MCF-7 and T47D cells,
RAB21 down regulation itself is not enough to explain the phenotype. Some other mechanisms are also involved in the regulation of cell proliferation and migration. For example, inhibition of
FOXO1 by
miR-96 and
miR-182 will increase cell proliferation.
We identified two regulators (
ZEB1 and
HSF2) and one target gene (
RAB21) for the
miR-183/-96/-182 cluster in breast cancer cell lines. How do they work in clinical samples? We looked for correlation between
miR-183/-96/-182 cluster miRNAs and their target/regulators by analysis of 508 clinical samples from TCGA data (Additional file
9). Because the correlations between miRNAs and their targets/regulators are not simply negative or positive correlations, we did not find any direct correlations between these miRNAs and the expressions of
HSF2,
ZEB1 and
RAB21 based on the TCGA data analysis. But there were some interesting correlations between them in different subtypes.
MiR-96 and
miR-183 weree lower in ER+ and PR+ breast cancers than ER- and PR- breast cancers; in the meantime, their regulator,
HSF2 level was lower and their target,
RAB21 level, was higher in ER+ and PR+ breast cancers than ER- and PR- breast cancers (Additional file
4: Table S4). Subtype analysis also confirmed our findings.
HSF2 level was high in basal breast cancers, which are
miR-183-enriched breast cancers; and
RAB21 level was low in HER2 and basal breast cancers, which are
miR-96- and/or
miR-183-enriched breast cancers (Additional file
4: Table S5).
MiR-182 was not strongly correlated with the levels of
HSF2 because its transcription is not controlled by
HSF2 (Figure
3D)
. There is still a complicated phenomenon that requires explanation, which is that the
ZEB1 level was negatively correlated with
miR-96 and
miR-183 (Additional file
4: Table S4, S5)
. In MCF-7 cells,
ZEB1 upregulates the expressions of
miR-96 and
miR-183 (Figure
3)
, and Graham
et al. also report that
ZEB1 is more expressed in ER/PR- breast cell lines than ER/PR+ breast cell lines [
38]. However, in clinical samples,
ZEB1 was enriched in ER/PR+ samples. Considering
ZEB1 is a transcription factor that can either activate or repress its target genes, we think it functions differently in breast cancer cell lines and breast cancer patients. In patients,
ZEB1 may repress the transcription of
miR-183/-96/-182 cluster. This conclusion needs further work for confirmation, but nevertheless,
ZEB1 plays an important role in the regulation of
miR-183/-96/-182 cluster.
Authors' contributions
PL carried out the molecular and cellular biology studies and drafted the manuscript. CS participated in the construction of plasmids and viruses. LLH participated in cellular biology studies. HZ collected cancer samples. LHH conducted real-time PCR and FACS. ZC participated in the design of the study and performed the statistical analysis. QZ conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.