Background
Metastasis is responsible for ~90% of deaths in patients with solid tumors [
1‐
4], including those originating in the breast [
5‐
7]. Metastasis has always been portrayed as the ultimate step of the progressing breast cancers. Recent evidence, however, indicates that about a third of women diagnosed with small asymptomatic breast tumors (~4 mm) already harbor disseminated BC cells in their bone marrow [
8]. Moreover, these micrometastases can remain dormant for years before reemerging as incurable secondary tumors and surprisingly insensitive to adjuvant chemotherapies that were originally effective against the primary tumor [
9,
10]. Adding to this problem is the fact that BC is a heterogeneous disease comprised of at least 5 genetically distinct subtypes, which together are the second leading cause of cancer deaths in women in the United States [
11‐
13]. Within BC subtypes, those classified as Triple-Negative BCs (TNBCs) exhibit dismal survival rates due to their highly aggressive and metastatic behavior, and to their propensity to rapidly recur [
14‐
20]. The TNBC subtype is characterized by lack of expression of hormone receptors (ER-α and PR) and HER2, harbor BRCA1-defects and/or deficiencies, and remain p53-positive [
14,
16‐
20]. Moreover, the absence of novel therapies capable of specifically targeting this very aggressive TNBC subtype reflects in part a lack of sufficient knowledge about TNBC development and progression [
2,
4,
9,
21].
microRNAs (miRs) are small noncoding RNAs, usually 20- to 22-nucleotides long, which regulate gene expression at the post-transcriptional level. To date, close to 1000 human miRs have been identified, which are thought to regulate more than 50% of human genes. miRs are now widely regarded as the most powerful regulators of gene expression in complex cellular processes including cancer cell invasion and metastasis [
1,
22‐
25]. In fact, several miRs, miR-15a, miR-16-1, and let-7 [
23,
26‐
30] function as tumor suppressors, and others, miR-155, miR-17-5p, and miR-21 [
23,
31,
32], possess oncogenic properties
Several recent reports have identified a major role of miR31 in cancer metastasis [
33‐
36] With regard to BC, we reported that miR-31 expression is lost in aggressive basal-type breast cancer cell lines compared to the non-invasive luminal counterparts. This observation was extended to human breast cancer tumors where we found an inverse correlation between miR-31 expression levels and advanced stages of BC [
24]. Also, in our previously published work, we reported a highly significant correlation between the expression levels of WAVE3 and advanced stages of BC [
37], supporting the function of WAVE3 as a metastasis promoter protein [
38‐
40]. Linking these observations, we found that miR-31 regulated WAVE3 expression and activity during the invasion-metastasis cascade [
24,
25]. However, the upstream mechanisms of transcriptional regulation of miR-31 are not well understood and are the focus of the present study.
A recent study has predicted miR-31 to be transcribed from within the first intron of a host gene, LOC554202, on human chromosome 9 [
41]. Our in silico analyses have confirmed these findings and suggest that LOC554202 is transcribed into a long non-coding RNA, (Lnc)RNA. We also identified a major CpG island upstream of the miR-31 locus, which also spans the first exon of LOC554202, suggesting an epigenetic regulation by methylation of both miR31 and its host gene. Here, we report that the expression pattern of miR-31 follows that of LOC554202 in the TNBC basal versus the luminal BC cell lines, supporting the hypothesis that miR-31 is under the transcriptional regulation as LOC554202. Next we show that loss of miR-31 expression in the aggressive TNBC cell lines is a direct consequence of hypermethylation of its associated promoter which also regulates LOC554202, the host gene for miR-31. Using both methylation-specific PCR (MSP) and bisulfite-modified DNA sequencing, we directly demonstrate that the miR-31 promoter is heavily methylated in basal TNBC compared to luminal BC cell lines. Our results not only identify a novel mechanism for miR-31 regulation but also clearly support the important role of promoter methylation in the suppression of miR-31 during tumor progression.
Discussion
Altered expression of microRNAs is frequently observed in human cancer, including ones originating in the breast [
44‐
46], but the mechanisms underlying their regulation are poorly understood. We and others have previously shown that miR-31, a BC metastasis suppressor gene, is a major contributor to BC progression and metastasis by regulating a cohort of a pro-metastatic target genes, including WAVE3 [
24], RhoA, Radexin [
36] and several integrin subunits [
47] that regulate key steps in the invasion-metastasis cascade. miR-31 expression levels are high in early stage BC tumors, allowing for its pro-invasive, pro-metastatic target genes to remain under tight control [
24,
36]. Expression levels of miR-31diminish as the tumors progress to more invasive stages [
24] and become undetectable in metastatic BC tumors [
36]. Loss of miR-31 expression is accompanied by concomitant increase in expression of its pro-invasive and pro-metastatic target genes, therefore, allowing for the tumor to become more invasive and ultimately metastasizes [
24,
36]. In BC mouse models, re-expression of miR-31 in the triple-negative MDA-MB-231 BC cells, which do not express endogenous miR-31, almost completely inhibits the metastatic potential of these cells without affecting the growth of the primary tumors while specifically inhibiting its pro-metastatic target genes ([
36] and Sossey-Alaoui, unpublished data). On the other hand, knockdown of miR-31 in the non invasive luminal MCF7 BC cells results in lifting the inhibitory effect imposed by miR-31 on its target genes and imparts aggressive and metastatic phenotype to these cells comparable to observed in MDA-MB-231 cells [
24,
36]. Together, these published studies clearly demonstrate the important role that miR-31 plays during the invasion-metastasis cascade of BC tumors.
The mechanisms for upstream regulation of miR-31 leading to its loss during the invasion-metastasis cascade has been heretofore unknown. In this study, we report the contribution of epigenetic modifications as a novel mechanism by which miR-31 is regulated in breast cancer. First we showed that miR-31 is transcribed from the intronic sequence of LOC554202, a newly identified lncRNA. While both miR-31 and its host gene LOC554202 are expressed abundantly in the non-invasive BC cell lines of luminal subtype, their expression is lost in more aggressive TNBC cell lines of basal subtype, clearly suggesting that the transcription regulation of miR-31 might be under the control of its host gene LOC554202. Second, we identified a strong CpG island in the LOC554202-associated promoter, prompting us to hypothesize that both miR-31 in its host gene might be regulated by promoter methylation. Indeed, we were able to enhance expression of both miR-31 and LOC554202 in the TNBC cell lines after treatment with either the methylase inhibitor 5Aza2Cd alone or in combination with the acetylase inhibitor TSA. To further confirm the contribution of promoter hypermethylation to the loss of miR-31 in the TNBC cell lines we performed both methylation specific PCR and sequencing of bisulfite-modified DNA from both luminal (MCF7) and basal TNBC (MDA-MB-231 and BT549) cell lines. The combined number of CpG dinucleotides surveyed by these two assays allowed coverage of at least one third of total length of the CpG island. We found that while the LOC554201-associated promoter was significantly hypermethylated the in basal TNBCs, it was significantly hypomethylated in the luminal counterpart, further confirming that miR-31 expression is regulated in the TNBCs at least in part by promoter methylation. It is well established that hypermethylation of CpG islands associated with specific genes increases during the growth and progression of the primary tumor, providing a mechanism to inactivate tumor suppressor genes, DNA repair genes, cell cycle regulators and transcription factors. Based on our RT-PCR results, treatment of the TNBC cells with 5Aza2Cd or in combination with TSA enhanced expression of both miR-31 and its host gene LOC554202 to levels similar to those found in luminal BC subtypes. The restoration of miR-31 expression by these maneuvers was also very significant but did not reach the levels observed in luminal BC (Figure
3). One possible explanation for this difference is that, in addition to promoter methylation that regulates both miR-31 and LOC554202, other mechanisms may selectively regulate miR-31.
It is worth noting that our study appears to be the first to report that LOC554202 might belong to the lncRNA family. Our preliminary in silico analyses show that the LOC554202 locus spans more that 100 kilobases (kb) of genomic sequence and that its RNA is transcribed from 4 exons resulting in a spliced transcript of ~2.2 kb (Figure
1), which does not contain an open-reading frame that could be translated into a functional protein, and therefore can be classified a lncRNA. It is possible that this new lncRNA may have a function in chromatin remodeling and epigenetic regulation of gene expression similar to that of the well known XIST and HOTAIR lncRNAs [
48,
49]. More experimental analyses are however required to investigate the exact function of LOC554202 in breast cancer invasion and metastasis.
Conclusion
The present study, although conducted in established cell lines, clearly identifies promoter hypermethylation as a novel mechanism by which miR-31 is silenced during the invasion-metastasis cascade of BC. Future studies using biological specimens with associated clinico-pathological parameters and disease outcome, are required to further confirm these findings and to assess whether miR-31 promoter methylation can be used a prognostic marker for BC progression and survival outcome.
Methods
Cell lines and their treatment
Human non-malignant breast epithelial cell line MCF10A and the human breast cancer cell lines were obtained from American Type Culture Collection (Rockville, MD). Cells were cultured at 37°C with 5% CO
2 in their specified basic culture medium supplemented with 4.5 g/L glucose, 10% fetal bovine serum (Invitrogen), 2 mmol/L glutamine and antibiotics. The demethylating agent 5Aza2dC (Sigma, MO, USA) was freshly prepared in double-distilled H
2O and filter sterilized. Cells (5-10 × 10
5) were seeded in a 100 mm tissue culture dish in culture medium at 37°C, 10% CO
2. The next day, cells were treated with 0.5 μ
M of 5aza2dC. The culture medium containing the demethylating agent was replaced every day for 7 days. For the 5Aza2dC-Trichostatin A (TSA, Sigma, MO, USA) dual treatment, TSA (0.3 μ
M final concentration) was added to the culture at day 5 for a 48-h treatment period. At the end of the treatment period, total RNA was prepared using TRIzol (Invitrogen, CA, USA), according to the manufacturer's instructions. The BAC clones were obtained from the Rowell Park Cancer Institute BAC Library and BAC DNA was isolated using the the QIAprep Spin Miniprep kit (Qiagen Sciences, MD, USA). The total genomic DNA was prepared using proteinase K digestion as previously described [
50].
Semi-quantitative and Real-time quantitative-PCR
Total RNA was extracted from cancer cell lines using TRIzol reagent (Invitrogen, Carlsbad, CA), following to the manufacturer's instructions. cDNA was generated and used as a template for semi-quantitative RT-PCR performed as previously described [
25,
37,
51,
52]. Expression levels of the precursor (pri-miR) and the mature forms of microRNA miR-31 were quantified by real-time quantitative RT-PCR using human TaqMan MicroRNA Assays Kits (Applied Biosystems, Carlsbad, CA). We used GAPDH to normalize the expression levels of LOC554202 transcripts. In addition, we found that both miR-16 and RNU6B were expressed at similar levels in all cell lines analyzed, when normalized to GAPDH (Additional File
1). Furthermore, treatment of BC cells with either 5Aza2dC, Trichostatin A or both, did not affect their expression levels when compared to the untreated cells (Additional File
1), and therefore, were used for normalization of miR-31 expression levels across the breast cancer cell lines and between treatments. The reverse transcription reaction was carried out with TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems, Carlsbad, CA) following the manufacturer's instructions. Quantitative PCR was performed on the BioRad (Hercules, CA) MyiQ2 iCycler PCR system where the reaction mixtures were incubated at 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. The cycle threshold (Ct) values were calculated with SDS 1.4 software (Bio-Rad). The expression levels of miR-31 were normalized using the 2
-ΔΔCt method [
53] relative to miR-16 or RNU6B. The ΔCt was calculated by subtracting the Ct values of miR-16 from the Ct values of miR-31. The ΔΔCt was then calculated by subtracting ΔCt of each breast cancer cell line from MCF10 cells. Fold change in the gene expression was calculated according to the equation 2
-ΔΔCt.
Preparation of bisulfite-modified DNA for methylation analysis
Genomic DNA (1-2 μg) was denatured in 0.3 M NaOH for 30 min at 42°C, and then the unmethylated cytosine residues were sulphonated by incubation in 3.12 M sodium bisulfite (pH 5.0; Sigma)/5 mM hydroquinone (Sigma, MO, USA) at 55°C for 16 h. The sulphonated DNA was recovered using the QIAquick Gel Extraction system (Qiagen, MA, USA), according to the manufacturer's recommendations. The conversion reaction was completed by desulphonating in 0.3 M NaOH for 5 min at room temperature. The DNA was ethanol precipitated and resuspended in double-distilled water.
CpG island prediction and primer design for methylation analysis
Sequencing of bisulfite-modified DNA
In total, 20-50 ng of bisulfite-treated DNA was used as template in each PCR reaction under the following conditions: 95°C for 5 min, followed by 40 cycles of 15 s of denaturation at 95°C, 20 sec at 55°C and 25 sec of extension at 72°C. The PCR reaction was terminated with an additional 7 min of extension and cooled to 4°C. The PCR products were resolved on a 2% agarose gel, stained with ethidium bromide, and the 250-bp bands were excised and gel-purified using the QIAquick Gel Extraction system (Qiagen, MA, USA). The purified PCR products were cloned into the pCR2·1-TOPO vector (Invitrogen, CA, USA), and at least 15 clones were sequenced from each cell line. The methylation status at each CpG site was analyzed using the MethTools software (
http://genome.imb-jena.de/methtools/). The overall methylation status in each cell line was calculated as a ratio of the number of unmethylated to methylated CpGs and plotted as a percentage of total number of CpGs analyzed.
Methylation specific PCR
Methylation specific PCR (MSP) was performed on bisufite-converted DNA using the MSP primer pairs described in Additional file
2. Each DNA sample was PCR-amplified using either the methylated or the unmethylated primer pairs. The PCR products were next resolved by agarose electrophoresis, stained with ethidium bromide and a picture recorded. The intensities of the PCR products between the methylated and unmethylated primer-pairs were compared by densitometry.
Oligonucleotide primer sequences
Sequences of the oligonucleotide primers used for genomic PCR, RT-PCR from IDT (San Diego, CA) and are listed Additional file
2.
Statistical analyses
The data are presented as the means ± standard errors of at least three independent experiments. The results were tested for significance using an unpaired Student's t test and p values of < 0.05 were considered statistically significant.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
KA carried out most of the experiments including the quantitative real-time RT-PCR for miR-31 and LOC445202 as well as the methylation specific PCR and assisted in drafting the manuscript. BM performed the bisufite sequencing experiments and was responsible for the tissue culture of cell lines and their treatments as well as DNA and RNA isolation. EFP discussed the design of the experiments, the results and assisted in writing. KSA supervised the overall design of the experiments and wrote the manuscript.