Introduction
MicroRNAs (miRNAs) are a class of small non-coding RNA molecules 19-24 nucleotides in length that suppress gene expression post-transcriptionally by base-pairing with the 3
′-untranslated regions (3
′-UTRs) of target mRNA[
1]. Recent studies have shown that miRNAs are involved in multiple processes of cancer progression including cancer cell proliferation and metastasis[
2]. Large scale profiling approaches have revealed that miRNAs are globally down-regulated in breast cancer[
3]. The same study identified a set of miRNAs as being differentially expressed in breast tumors and showed that the miRNA profile could be used to distinguish between breast cancer and normal breast tissue[
3]. Studies have also revealed the correlation between down-regulation of certain miRNAs and clinicopathological features such as ER/PR positivity, tumor size, lymph node status, and metastasis status[
3,
4]. It was reported that the level of miR-31 was down-regulated[
5], while the level of miR-10b was up-regulated in metastatic breast tumors[
6]. With an experimental murine model, two reports have shown that tumors can be effectively suppressed by either silencing pro-cancer miRNA (miR-10b)[
7] or expressing anti-cancer miRNA (miR-26a)[
8]. Taken together, these findings clearly demonstrate that miRNAs play a critical role in breast cancer development and progression.
Metastasis is the main cause of breast cancer mortality and involves multiple complicated processes[
9,
10]. The ability of mammary tumor cells to invade and destroy neighboring tissues and organs, as well as migrate to other parts of the body, is crucial to the metastatic process[
9,
10]. It has been increasingly recognized that miRNAs regulate cell migration and invasion and play an important role in the invasiveness of breast cancer cells[
11,
12]; however, a systematic investigation of how miRNAs affect the invasive behavior of breast cancer cell lines has not been conducted. In this study, we performed an integrated analysis of miRNA and mRNA expression profiles in 12 breast cancer cell lines and identified a group of miRNA that are differentially expressed in invasive breast cancer cell lines when compared to less-invasive cell lines. We identified 35 functional target genes of three significantly down-regulated miRNAs in invasive cell lines, namely miR-200c, miR-205, and miR-375. Extensive validation studies were performed to confirm the functional interaction of the three miRNAs and their target genes. Finally, we characterized one of the target genes of miR-200c,
CFL2, and demonstrated that
CFL2 is overexpressed in invasive breast cancer cell lines and regulated by miR-200c. Tissue microarray analysis (TMA) further demonstrated that CFL2 expression in primary breast cancer tissue was positively correlated with tumor grades.
Materials and methods
Tissue culture and RNA isolation
Breast cancer cell lines BT474, MDA-MB-468, T47D, ZR-75-1, MCF7, SK-BR3, MDA-MB-231, HS578T, BT549, SUM159, and HeLa cell line were cultured in DMEM media supplemented with 10% fetal bovine serum (FBS). Immortalized breast epithelium cell lines MCF10A and MCF12A were cultured in DMEM/F12 supplemented with 5% horse serum, 20 ng/mL EGF, 10 μg/mL insulin, 100 ng/ml cholera toxin, and 500 ng/ml hydrocortisone (all from Sigma Aldrich, St Louis, MO). Total RNA was extracted using the QIAzol™ Lysis reagent (Qiagen, Valencia, CA). Small molecular weight RNA was extracted using the mirVana™ miRNA Isolation Kit (Invitrogen, Carlsbad, CA) per manufacturer’s protocol.
Transfection
miR-200c, miR-205, miR-375 mimic or scrambled negative control (Ambion, Austin, TX) at a concentration of 50 nM were incubated with Lipofectamine 2000 (Invitrogen) in culture medium before addition to cells according to the manufacturer's protocol. CFL2 siRNA and scramble control siRNA were purchased from Dharmacon (Lafayette CO.) and used at a concentration of 30 nM as described above.
microRNA expression profiling
The GeneChip miRNA 1.0 array (Affymetrix, Santa Clara, CA) was used for the miRNA expression profiling in breast cancer cell lines. One μg of small RNA from each sample was labeled with biotin using the FlashTag Biotin RNA Labeling Kit (Genisphere, Hatfield, PA). Array hybridization, washing, and scanning of the slides were carried out according to Affymetrix's recommendations. Data was extracted from the images, quantile-normalized, summarized (median polish), and log
2-transformed with the miRNA QC software from Affymetrix. Partek Genomic Suites (Partek, St. Louis, MO) was used to analyze the array results, and TargetScan6.2 (
http://www.targetscan.org/) was used to predict miRNA-mRNA pairs. All microarray data has been submitted to NCBI Gene Expression Omnibus (
http://ncbi.nlm.nih.gov/geo/) under accession number GSE40059.
mRNA expression profiling
The GeneChip® Human Genome U133 Plus 2.0 Array (Affymetrix) was used for the mRNA expression profiling in 12 breast cancer cell lines. Biotinylated cRNA was synthesized from total RNA using the Affymetrix 3′ IVT Express Kit according to manufacturer’s protocols. The GeneChip® Human Gene 1.0 ST Array (Affymetrix) was used for the mRNA expression analysis in the miRNA mimic transfected MDA-MB-231 cells. The cRNA was synthesized using Ambion WT Expression Kit and labeled using Affymetix GeneChip WT Terminal Labeling Kit. Array hybridization, washing, and scanning of the slides were carried out according to Affymetrix's protocols. The gene expression data was analyzed using Partek Genomic Suites 6.5. The Ingenuity Pathway analysis (IPA) was used to identify functional groups and molecular networks from the microarray data sets generated in the miRNA mimic transfected MDA-MB-231 cells.
qRT-PCR analysis of miRNA expression
One μg of small RNA was used for reverse transcription with the RT2 miRNA First Strand Kit (SA Biosciences, Frederick, MD). Quantitative RT-PCR was carried out using a Light Cycler 480 II instrument (Roche, Indianapolis, IN). The PCR primers for U6, miR-200c, miR-205, miR-375, and miR-146a were purchased from SABiosciences. RT2 SYBR Green Master Mixes (SA Biosciences) were used in the real time PCR reaction according to the manufacturer’s suggested protocols. The relative gene expression was calculated using the equation 2-ΔCt, where ΔCt = Ct (miRNA) − Ct (U6).
qRT-PCR analysis of mRNA expression
Two μg of the total RNA was reverse-transcribed using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA). All PCR reactions were carried out as described above. The primer sequences used for RT-PCR can be found in Additional file
1: Table S1. Each sample was run in duplicate. Fold change in gene expression was calculated using ΔΔCt method.
Transwell migration and invasion assay
miRNA mimic or siRNA treated and control cells were starved in serum-free medium for 2 hours, detached, and then re-suspended in medium with 2.5% fetal bovine serum at a density of 4 × 105 cells/mL. For the migration assay, 500 μL of the cell suspension was added to the upper chamber of the transwell inserts (BD Biosciences, Sparks, MD). 750 μL of medium containing 10% fetal bovine serum was added into the bottom of a 24 well plate to act as a chemoattractant. After an 8-hour migration period, non-migratory cells in the upper chamber were removed with cotton swabs, and the cells on the lower surface of the inserts were fixed and stained using DIFF-QUICK (IMEB Inc, San Marcos, CA). The number of migratory cells was calculated by counting five different fields under a phase-contrast microscope in three independent inserts. Invasion assays were done in a similar manner as the migration assays described above, except that the inserts were pre-coated with Matrigel (BD Biosciences). The cells were allowed to invade for 24 hours before proceeding with fixation and staining.
Luciferase reporter assay
The 3
′-UTR of
CDH11,
CFL2,
SEC23A,
ZEB-1,
PTPRM, and
LDHB were generated by PCR using DNA isolated from HeLa cells. The PCR fragments were subcloned into the pmirGLO dual-luciferase reporter vector (Promega, Madison, WI, USA). The primers used for 3
′-UTR amplification can be found in Additional file
2: Table S2. The reporter gene constructs were cotransfected into HeLa cells containing a miR mimic control or miR-200c/205/375 mimic for 48 hours. The dual luciferase system (Promega) was used to measure luciferase activity per manufacturer’s protocol. Normalized firefly luciferase activity (firefly luciferase activity/Renilla luciferase activity) was used to compare each respective sample to the control. For each transfection, luciferase activity was averaged from three replicates.
F-actin staining
Cells were fixed in 3.7% formaldehyde solution and extracted with a solution of 0.1% Triton X-100 in PBS for 5 minutes. The cells were then washed three times with PBS and stained using a rhodamine phalloidin (Invitrogen) solution for 20 minutes at room temperature. The cells were washed three times with PBS and mounted in a Mounting Medium for Fluorescence (Vector Laboratories, Inc. Burlingame, CA).
Immunohistochemistry
CFL2 levels in breast tumors and normal breast tissues were evaluated by IHC using anti-CFL2 polyclonal antibody (1:250 dilution, sc-32160, Santa Cruz Biotechnology, Santa Cruz, CA) on commercial tissue arrays (Shanghai Outdo Biotech Co., Shanghai) as previously described[
13]. The array contained 5 normal breast tissues and 211 breast tumor specimens. Staining intensity of each sample was given a modified histochemical score (MH-score) that considers both the intensity and the percentage of cells stained at each intensity[
14]. The intensity of each grade is the average of MH-score of all samples in that grade. Clinicopathological data of the 211 tumors used in TMA is provided in Additional file
3: Table S3.
Statistical analysis
Each experiment was repeated at least in triplicate. Numerical data are presented as mean ± s.d. Student’s t-test was used to analyze the differences between two samples; differences were considered statistically significant at p < 0.05. One-way ANOVA was performed in SPSS 17.0 (SPSS Inc. Chicago, IL) to analyze the association of CFL2 and tumor grades.
Discussion
The complications arising from metastasis are the major causes of death from cancer. Mounting evidence suggests that miRNAs may promote or suppress tumor metastasis[
12], thus offering a new perspective on the metastatic process. However, there has not been a systematic evaluation of the role of each miRNA, or a combination of coordinately expressed miRNAs, in the metastatic process of breast cancer cells. In this study, we attempt to perform a systematic evaluation of the role of miRNAs in the invasiveness of breast cancer cells using a group of 12 well-characterized breast cancer cell lines. Previous gene expression studies have shown that these cell lines resemble the primary tumor cells and can be grouped as basal or luminal-like breast cancer subtypes[
17]. These cell lines are highly variable in their migration and invasion capability; BT474, MDA-MB-468, T47D, ZR-75-1, MCF7, SK-BR3 represent less-invasive breast cancer cell lines, while MDA-MB-231, HS578T, BT549, SUM159 are more invasive. MCF10A and MCF12A are two immortalized breast epithelial cell lines that are non-tumorigenic, but were classified as basal-like cell lines based on gene expression profiles. Using highly stringent criteria, we identified a group of 11 miRNAs that were most differentially expressed between the invasive and less-invasive cell lines, including several miRNAs that have been linked to breast cancer previously, such as miR-200c, miR-203, miR-205, miR-375, miR-141, and miR-146a[
15,
16,
18]. Introduction of exogenous miR-200c, miR-205, and miR-375 mimics in the invasive breast cancer cell line MDA-MB-231 inhibited cell migration and invasion suggesting that these miRNAs may play important roles in maintaining the invasiveness of this cell line.
Using gene expression array analysis, we also identified candidate genes affected by the exogenous miRNA mimics in MDA-MB-231 cells. IPA analysis revealed that genes affected by miR-200c, miR-205, and miR-375 are involved in cellular movement and cell-to-cell signaling. Most significantly, our integrated analysis suggested that miR-200c might play a pivotal role in regulating cell migration and invasion, as 27 of the 35 differentially expressed genes were miR-200c targets. This is consistent with the migration and invasion experimental results showing that the miR-200c mimic is more potent than miR-205 and miR-375 mimics in inhibiting the migration and invasion of MDA-MB-231 cells. CDH1 (E-cadherin) was found to be at the core of the molecular network regulated by miR-200c. CDH1 acts as a hub connected by several neighborhood genes that play important roles in cell migration and invasion (Additional file
6: Figure S2A).
To further understand the function of the three miRNAs, it was necessary to identify their true targets in breast cancer cells. The computational prediction of miRNA targets still faces significant challenges. The most widely used tools (miRanda, TargetScan, PicTar, PITA, and RNAhybrid) are characterized by a significant proportion of false-positive interactions because post-transcriptional regulation is context-dependent. On the basis of increasing experimental evidence supporting the hypothesis that miRNAs can act through target degradation, it has been proposed that target predictions could be integrated with miRNA and gene expression profiles to select functional miRNA-mRNA relationships[
19‐
24]. In this study, we performed an integrated analysis by combining multiple data sets: 1) differentially expressed miRNA between invasive and less invasive cell lines; 2) differentially expressed genes between the two groups of cell lines; 3) miRNA target genes predicted by TargetScan; 4) genes significantly down-regulated upon forced expression of a miRNA. The integrated analysis yielded 35 genes that satisfied the selection criteria. Out of the 27 differentially expressed target genes identified for miR-200c, 17 have been shown to influence the migration, invasion, or metastasis of cancer cells both
in vitro and
in vivo (Table
1). Of these genes,
LOX, which encodes Lysyl oxidase, is essential for hypoxia-induced metastasis[
25] and it was reported that breast cancer cell metastasis can be attenuated by lysyl oxidase inhibitors[
26]. EMP1 may represent a novel immunohistochemical marker helpful in distinguishing between invasive ductal and lobular carcinomas[
27]. Fascin (
FSCN1) is a key regulator of breast cancer invasion[
28]. CDH11 may play a role in recruiting Trio to the plasma membrane where Trio activates Rac, leading to cell migration[
29].
SEC23A, which mediates the secretion of metastasis-suppressive proteins including
IGFBP4, has been shown to be a direct target of miR-200c, and the subsequent down-regulation has been correlated to an increase in metastatic colonization[
30]. Interestingly,
SEC23A was identified as a target of both miR-200c and miR-375 (Table
1), suggesting crosstalk can occur between these two miRNAs.
PRKCA overexpression is strongly associated with a more invasive and metastatic phenotype in breast cancer[
31]. Furthermore, it has been shown that
PRKCA expression is increased in the invasive breast cancer cell lines MDA-MB-231 and HS578T, and overexpression can lead to a significant increase in both the migration and invasion ability of the cell lines[
32]. Finally, SLC7A11, which is functional subunit of the cystine/ glutamate transporter, plays an important role in breast tumor metastasis and maybe a potential target for cancer therapy[
33].
Among the target genes identified,
CFL2 was most significantly down-regulated by miR-200c. Hurteau et al. first indicated that CFL2 as a potential target gene of miR-200c[
30], and later re-reported by Gregory et al[
34]. Korpal et al showed recently that knockdown of
Cfl2 in a mouse mammary tumor cell lines 4TO7 significantly decreased cell migration[
35]. In this study, we further revealed that
CFL2 plays a crucial role in regulating actin turnover and is intimately linked to the cell migration and invasion ability. An increased level of CFL2 would allow for a much higher F-actin turnover rate, thus allowing the cell to become much more mobile[
36]. Therefore by slowing the F-actin turnover rate, miR-200c helps to maintain the anchoring filaments that tend to hold the cells in place, leading to the decreased migration and invasion ability. CFL2 expression was also noted to increase significantly along with the grade of the tumor. This increase in CFL2 production likely assists the tumor in spreading both locally and metastatically at a much greater rate.
In summary, we have performed the first systematic screening of miRNA-mRNA target pairs that are differentially expressed between invasive and less-invasive breast cancer cells. We identified a group of negatively correlated miRNA-mRNA target pairs via integrated analysis of miRNA and mRNA expression profiles. The subsequent confirmation studies demonstrated that this integrated approach is very effective. Our results further emphasize the important role of miR-200c and its target genes in maintaining the invasiveness of breast cancer cells. Further analysis of the candidate miRNAs and their target genes identified in this study may ultimately lead to the identification of novel prognostic biomarkers and therapeutic targets.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
DL, LH and HS designed the study. DL, JMW, NH, JL, and LP performed the experiments. DL, NH, LH and SH participated in the data analysis and interpretation. SH contributed vital reagents and analytical tools. DL, JMW and HS co-wrote the manuscript. All authors read and approved the final manuscript.