Introduction
Breast cancer is the most common cause of cancer-associated deaths amongst women in developed and developing countries [
1]. Metastasis, the spread of a tumor to distant organs, accounts for 90 % of breast cancer patients’ mortality [
2]. Important advances have been achieved to understand the complicated process of metastasis, and various molecules have shown promising anti-metastatic properties [
3‐
5]. However, detailed mechanisms remain to be defined.
The microRNAs (miRNAs) are small non-coding RNAs of ∼ 22 nt that regulate gene expression at the post-transcriptional level. These molecules add a new dimension for understanding cancer progression. An increasing paradigm has clearly shown that miRNAs are involved in breast cancer metastasis. For example, miR-10b promotes breast cancer cell invasion and metastasis by targeting syndecan-1 (SDC1) in MDA-MB-231 and MCF-7 cells [
6]. The invasion of MDA-MB-231 and BT-20 cells is diminished by over-expression of c-Met-targeting miR-335 [
7]. Furthermore, miR-135 and miR-203 reduce tumor growth and metastasis of MD-MB-231 cells to the bones by targeting the runt-related transcription factor 2 (Runx2) [
8].
MiR-212 and miR-132 are tandem miRNAs at the same location on chromosome 17 in humans, called miR-212/132 cluster, and they share the same seed sequence and the transcriptional regulatory elements. Extensive studies have revealed important roles of this miRNA cluster in the different body systems, which may suggest potential therapeutic strategy. For example, miR-212/132 cluster is involved in mammary gland development [
9,
10], neuronal differentiation and cognitive processes [
11‐
13], cardiac hypertrophy and cardiomyocyte autophagy [
14], autoimmune inflammation [
15], vasodilatory function and angiogenic responses [
16].
In breast cancer, miR-132 suppresses cell proliferation, invasion, migration and metastasis of different breast cancer cells through direct suppression of hematological and neurological expressed 1 (HN1) [
17]. Over-expression of this miRNA suppresses proliferation and colony formation of MDA-MB-231 and MCF-7 [
18]. Moreover, miR-132 causes expression changes of genes involved in metabolism, DNA damage and cell motility in immortalized fibroblasts co-cultured with epithelial columnar cell hyperplasia (CCH) cells [
19]. Although the role of miR-212 has been investigated in different cancer types [
20,
21], it has never been investigated alone or in a combination with miR-132 in breast cancer.
The aryl hydrocarbon receptor (Ahr) is an environmentally responsive transcription factor activated by structurally diverse agonists see [
22]. It is demonstrated that the Ahr-active omeprazole decreases invasion and metastasis in estrogen receptor (ER)-negative breast cancer cell lines by down-regulation of matrix metalloproteinase-9 (MMP-9) and C-X-C chemokine receptor 4 (CXCR4) [
23]. Activation of Ahr by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and 6-methyl-1,3,-trichlorodibenzofuran (MCDF) suppresses metastasis of ER-negative breast cancer cells to the lungs [
24,
25]. Zhang and colleagues [
25] suggest that both TCDD and MCDF induce miR-335 targeting the pro-metastatic mediator SRY-related HMG-box4 (SOX4). However, no more studies were performed to provide more miRNA-based mechanistic explanations.
Activation of Ahr by 3,3′-diindolylmethane (DIM) suppresses breast cancer through repression of CXCR4 and/or CXCL12, and thereby, lowering the invasive and metastatic potential of MDA-MB-231 and MCF-7 cell [
26]. In ER-negative breast cancer cell lines, DIM suppresses cell proliferation and motility of MDA-MB-231 by inhibition of phosphorylation of hepatocyte growth factor (HGF) and c-Met at the tyrosines residues [
27]. Furthermore, oral treatment of DIM inhibits metastasis of 4T1 cells accompanied by reduced levels of MMP, adhesion molecules, and pro-inflammatory cytokines [
28]. Importantly, the underlying mechanisms of the anti-cancer activities of DIM are not simply attributed to the Ahr since DIM is a relatively weak agonist. For example, the DIM inhibits carcinogen-induced mammary tumor growth in Sprague–Dawley rats and this is not concomitant with the hepatic CYP1A1-dependent activity [
29]. In addition, no studies have investigated the involvement of miRNAs in the anti-metastatic effect of DIM.
In recent studies, we found that TCDD and 6-formylindolo[3,2-b]carbazole (FICZ) induced the highly conserved miR-212/132 cluster in the murine cellular immune compartment [
15,
30]. Therefore, it was hypothesized here that the miR-212/132 cluster may be induced in human breast cancer cells by Ahr agonists, and may contribute to their anti-metastatic properties. Thus, the effects of TCDD and DIM on miR-212/132 expression and metastatic features in human breast cancer cells were investigated. The current results, for the first time, demonstrated that toxic and non-toxic Ahr agonists suppressed breast cancer metastasis through triggering the transcription of SOX4-targeting miR-212/132 cluster. It was further shown that miR-212/132 cluster is a metastasis suppressor in breast cancer cells.
Discussion
The miRNAs are small non-coding RNAs that regulate gene expression by complementary binding on the 3′UTR of the target mRNA. Accumulating evidences have demonstrated that miRNAs are involved in the progression or inhibition of different tumors including breast cancer [
33,
34]. In breast cancer, metastasis is a complicated process that contributes to a high mortality rate among other cancer patients [
2]. It has been indicated that miRNAs modulate metastasis of breast cancer cells to different organs of the body [
35,
36]. However, their interaction with the transcription factors that play prominent roles in breast cancer metastasis is rarely investigated.
The Ahr has been extensively studied in different tumors and cell lines of breast cancer, and now it is clear that Ahr plays critical roles in modulating tumor progression [
37,
38]. Current study showed that TCDD inhibited invasion of MDA-MB-231 cells
in vitro, whereas DIM showed inhibitory effects on both cell lines. These results were in a line with previous studies showed that TCDD inhibited cell invasion in ER-negative breast cancer cell lines in an Ahr-dependent fashion [
25,
39]. The inhibitory effects of DIM on cell invasion in ER-negative and ER-positive cell line were also reported [
40‐
42]. The
in vivo results showed that DIM, but not TCDD, decreased the primary tumor weight that formed at the T47D injection site. In contrast, previous studies showed that TCDD and TCDD-related compounds inhibited mammary tumor growth [
43,
44]. These discrepancies were likely attributed to the rodent model used, doses of Ahr agonists and the number of these doses.
In previous studies, we demonstrated that activation of Ahr by TCDD and FICZ induced the highly conserved miR-212/132 cluster in murine cellular immune compartment, and supporting results were obtained in Ahr
−/− mice [
15,
30,
45]. Therefore, it was predicted in the current study that Ahr agonists induce miR-212/132 in human breast cancer cell lines, and these miRNAs may contribute to the anti-tumor properties of these agonists. Activation of Ahr by TCDD and DIM suppressed migration of MDA-MB-231 and proliferation-based expansion of T47D cells, and knockdown of Ahr reversed these effects. Inhibition of miR-212/132 in Ahr-agonist treated cells mitigated the anti-invasive effects of the agonists, and transfection with mimics showed supporting results, suggesting that miR-212/132 cluster mediated, at least partially, the anti-invasive effects of TCDD and DIM. Also, these results indicated an involvement of other molecules, probably other miRNAs, in the anti-invasive effects of Ahr agonists.
MiR-212 and miR-132 are tandem miRNAs located in an intergenic region on chromosome 17 in humans, and they share the same seed sequence AACAGUCU. Analysis of the regulatory elements of the miR-212/132 gene using online software, i.e., Promo V3.0.2 [
31], revealed the presence of two XRE boxes (GCGTG) located within 1 kb relative to the transcription site. A direct regulatory role of Ahr on the miR-212/132 gene by association with XRE boxes was tested by ChIP assay, and confirmed by luciferase activity. The results identified a functional XRE box, located at 830 bp from the transcription site, on which Ahr was bound. Together, these results confirmed for the first time that Ahr directly regulated the transcription of miR-212/132 gene.
The SOX4 transcription factor belongs to the SOX (SRY-related HMG-box) family that is involved in embryonic development and cell fate. It is demonstrated that an inhibition of SOX4 is associated with a decreased invasion and metastasis of breast cancer cells [
46]. It is also shown that TCDD and MCDF down-regulate SOX4 in MDA-MB-231 and BT474 by inducing miR-335 [
25]. These effects are abrogated by Ahr knockdown and mitigated by miR-335 antisense in Ahr agonists-treated cells. These results are in complement with those obtained in the current study. Both TCDD and DIM reduced the luciferase activity of the 3′UTR-SOX4 construct that contains the predicted binding sites for the miR-212/132 cluster, and mutation in these sites restored the luciferase activity. Also, co-transfection of the 3′UTR-SOX4-luc and miR-212/132 mimics or miRNA mimics alone showed supporting results. Taken together, the results identified SOX4 as a new target gene of the miR-212/132 cluster in human breast cancer cells. Notably, the effects of Ahr agonists on SOX4 were repealed by Ahr knockdown and partially reversed by miR-212/132 antisense in Ahr agonist-treated cells, suggesting an involvement of other molecules, i.e., miRNAs, in the regulatory role of Ahr agonists on SOX4.
The involvement of Ahr-miR-212/132-SOX4 module in the anti-metastatic properties of TCDD and DIM on MDA-MB-231 and T47D were investigated in the pulmonary nodules using a demonstrated model of spontaneous metastasis. Previous studies showed that oral treatment of TCDD and DIM in breast cancer cells-injected mice resulted in a reduction in the pulmonary tumor nodules [
24,
28]. These results were in line with those obtained in the current study using MDA-MB-231 and T47D cells. Consistent with the
in vitro results, TCDD induced a reciprocal correlation between miR-212/132 and SOX4
in vivo, which may explain, at least partially, the anti-metastatic properties. The DIM treatment induced miR-212/132
in vivo, to a lesser extent compared with TCDD, and mean value of SOX4 did not reach the significance level (
P < 0.05). These differences may be attributed to the agonists’ chemical structure. TCDD contains four chlorine residues that give this agonist stability and longer half-life, which attribute to strong and sustained Ahr activation see [
22].
Previous studies have suggested different mechanistic explanations for the
in vitro anti-invasive and the
in vivo anti-metastatic properties of Ahr agonists. For example, TCDD and MCDF induce SOX4-targeting miR-335 in MDA-MB-231 and BT474 cells
in vitro [
25]. It has been also suggested that TCDD disrupt the CXCR4/CXCL12 axis in an
in vitro chemotaxis assay [
47]. Oral administration of DIM suppresses 4T1 metastasis to the lungs by inhibition of two MMPs, adhesion molecules, and pro-inflammatory cytokines [
28]. Furthermore, different Ahr agonists show anti-estrogen effects by inhibitory cross talk between Ahr and ER in ER-positive breast cancer cells [
48,
49]. In the current study, over-expression of miR-212/132 showed anti-invasive properties, and inhibition of Ahr agonist-induced miRNA cluster abrogated the agonists’ anti-invasive properties in MDA-MB-231 and T47D. These results suggest a new mechanism through which miR-212/132 mediate the anti-metastatic properties of TCDD and DIM by targeting the pro-metastatic factor SOX4.
It has been shown that the constitutively active Ahr enhances growth and motility of breast cancer cells by different mechanisms such as transcription of breast cancer gene 1 (BRCA1) oncogene and CYP1B1, and activation of epiregulin and Wnt signaling pathway see [
50]. Therefore, inhibition of certain endogenous Ahr agonists has suppressive effects on breast cancer progression. This does not reflect contradiction with the inhibitory properties of Ahr agonists, as agonists may force Ahr to do different functions. For example, while endogenous ligands enhance cancer by transcription of BRCA1 oncogene, activation of Ahr by exogenous agonists suppress BRCA1 gene expression [
51].
Materials and methods
Cell culture
The human breast cancer cell lines MDA-MB-231 and T47D were obtained from American Type Culture Collection (ATCC; Manassas, VA). The cells were maintained in a complete medium contining Dulbecco’s modified Eagle’s medium (DMEM)/Ham’s F12 nutrient mixture (F-12) at 1:1 (Sigma-Aldrich, St Louis, MO) supplemented with 10 % FBS and 1× of antibiotic antimycotic solution (Gibco, Rockville, MD). Cells were incubated in a humidified atmosphere with 5 % CO2 at 37 °C.
Wound healing and invasion assays
Cells were seeded in 6-wells plate in a complete medium for 24 h to attach. The cells were then treated with DMSO, TCDD (Accustandard, New Haven, CT) or DIM (Sigma-aldrich) for 24 h in 3 % charcoal-stripped FBS (Sigma-Aldrich) DMEM/F-12. When cells reached 60–80 % confluent, a scratch was made at the axis of the well using pipette tip, washed, and then treated again with DMSO or Ahr agonists for another 24 h. The T47D medium was supplemented with 10 nmol/L β-estradiol (E2; Santa Cruz Biotechnology, Santa Cruz, CA). A 24-wells matrigel coated Boyden chamber with 8.0 μm PET membrane (Corning, New York, NY) was used for invasion assay. The cells (2 × 104) were suspended in 200 μl serum-free medium containing DMSO or Ahr agonists and placed in the trans-well. The lower chamber contained DMSO or Ahr agonists in 750 μl medium supplement with 10 % FBS as an attractant. After 24 h, cells were fixed by formaldehyde and permeabilized with methanol, and then stained with Giemsa. Migrated cells with spread-out shape were counted in 4 different microscopic fields.
Cell proliferation and adhesion assays
4 × 104 MDA-MB-231 or T47D cells were seeded in 96-well and incubated in 3 % charcoal-stripped FBS DMEM/F-12 containing different concentration of Ahr agonists for 48 h. Cell proliferation was quantified by MTT assay using Cell-Counting Kit-8 (CCK-8; Dojindo, Baltimor, MD) following manufacturer’s instructions. For adhesion assay, 4 × 105 cells were seeded in 6-wells plate and incubated overnight to adhere, then washed to remove non-adherent cells. Cells were then re-incubated in charcoal-stripped medium containing DMSO or Ahr agonists and incubated for 48 h. A single wash was performed before examination.
Real-time PCR and western blot assays
Extracted RNA was reverse transcribed in a thermal cycler using RT enzyme. The real-time PCR was carried out in ViiA 7 system using TaqMan® gene expression assays. The cycling conditions were 50 °C for 2 min and 95 °C for 10 min, then 40 cycles of 95 °C for 15 s and 60 °C for 1 min. The comparative ΔΔCt method was applied to calculate fold change. For endogenous controls, GAPDH was used for Ahr, CYP1A1 and SOX4, and RNU6B was used for miR-212/132. System, reagents and kits for real-time PCR were all from Applied Biosystems, Grand Island, NY. For western blot, cell were lysed by RIPA lysis buffer system (Santa Cruz Biotechnology), lysate was then fractionated using SDS-PAGE system (Bio-Rad, Richmond, CA). The Ahr, CYP1A1, SOX4 and β-actin were detected using their rabbit polyclonal antibodies (Santa Cruz Biotechnology). The intensities of the protein bands were quantified via software [
http://imagej.nih.gov/ij/download.html] ImageJ v.1.48 [
52].
ChIP assay
Analysis of the regulatory elements of the miR-212/132 gene was performed using transcription factor prediction software [
http://alggen.lsi.upc.es/] Promo V3.0.2 [
31]. The ChIP assay was performed using ChiP-IT enzymatic kit (Active Motif, Carlsbad, CA) following manufacturer’s instructions. Briefly, MDA-MB-231 and T47D cells were treated with DMSO, TCDD or DIM for 24 h. After that Ahr was immunopreciptated using specific antibodies (Santa Cruz Biotechnology). Attached DNA was prepared using proteinase K and further purified using phenol/chloroform procedure. The PCR was done using the following primer sets: XRE-1: forward 5′-CTCCTTCTGCTCCGCGTC-3′, and reverse 5′-TCCGCGGTGCTGATCAAC-3′; XRE-2: forward 5′-AGAGCACTACACCCAGCAG-3′, and reverse 5′-CAGGTGTGAGACTTCCCCAG-3′. The positive control CYP1A1 primers were used as previously described [
53]: forward 5′-TCAGGGCTGGGGTCGCAGCGCTTCT-3′, and reverse 5′-GCTACAGCCTACCAGGACTCGGCAG-3′.
Target gene prediction
Potential binding sites of miR-212/132 cluster (HGNC:31589/HGNC:31516) on the 3′UTR of SOX4 (HGNC:11200) were queried by the miRNA target prediction software microRNA.org [
http://www.microrna.org] miRaNda v.21 [
32].
Cell transfection and luciferase reporter assays
The fragment of miR-212/132 promoter that contains XRE-2 was cloned in the Xba1 restriction site of the basic PGL-3 vector (Promega, Madison, WI) using the following primers: forward 5′- AGATCGCCGTGTAATTCTAGAAGAGCACTACACCCAGCAG-3′, and reverse 5′- GCCGGCCGCCCCGACTCTAGACAGGTG TGAGACTTCCCCAG-3′. The siAhr and siNS control (Ambion, Austin, TX) were co-transfected with the miR-212/132 promoter reporter at the final concentration of 75 nmol/L. The 3′UTR-SOX4-luc construct (SwitchGear Genomics, Menlo Park, CA) was co-transfected with miR-212/132 mimics or antisense (Ambion) at 250 nmol/L as previously described [
45]. The binding specificity of miR-212/132 cluster on the SOX4 3′UTR was examined by mutation in the sequence on which miR-212/132 seed sequence bind (from GACTGTT to GAGACGG). The MDA-MB-231 and T47D cells were transfected using 4D-nucleofector device and cell-specific transfection kits (Lonza, Walkersville, MD). Cells were incubated for 6 h to recover, and then the medium was changed. After 24 h, the cells were lysed and luciferase activity was measured using luciferase reagents (Promega).
Orthotopic model of spontaneous metastasis
Female BALB/c athymic nude mice, 6–8 weeks old were purchased from King Faisal Specialist Hospital and Research Center, Riyadh, KSA. The orthotopic model of tumor growth and spontaneous metastasis was induced as previously described using ER-negative and ER-positive breast cancer cells [
54]. The MDA-MB-231 and T47D (2.5×10
5 cells) were suspended in 25 μl of serum-free DMEM. The cells were injected directly into the fat pad of the second left nipple through small incision. The mice receiving T47D cells were implanted with a 90-day release E2 pellets (Innovative Research of America, Sarasota, FL). Non-transplanted (Neg) mice were used as negative controls. The TCDD at 25 μg/kg/day, DIM at 50 mg/kg/day or corn oil were given orally for 10 consecutive days starting from the breast cancer cell injection day. Visible surface tumor nodules and internal nodules were excised under dissection microscope directly after mice euthanization. Lung samples from all mice were fixed in formaldehyde, sectioned to 5 μm thickness and stained with conventional hematoxylin and eosin (H&E) for metastases counting. All animals were maintained under specific pathogen-free conditions and had free access to sterilized feed and water. Animal experiments were performed in accordance with the protocols approved by Animal Care Committee of King Faisal University.
Statistics
Shown are mean ± SD of the results obtained from three independent experiments studied in triplicates; n = 9 in all experiments except when indicated. The significance was analyzed by analysis of variance (ANOVA) test for pooled data from all replicates. P < 0.05 was considered significant.
Acknowledgment
Special thanks for Sherr David for the scientific reading of the paper (School of Medicine, Boston University, USA), and for Al-hussein Khaled and Ghebeh Hazem (King Faial Specilist Hospital and Research Center, KSA), and Mohafez Omar and Shehata Tamer (Clinical Pharmacy, King Faisal University, KSA) for excellent technical support. Deep gratitude goes for Alzayer Mohammad (Veterinary Medicine, King Faisal University) for support and professional comments in histological studies. The work was financially supported by the Deanship of Scientific Research, King Faisal University (Grants 15008 and 160030).
Competing interest
The author declares that he has no competing interests.