Background
Basal-like tumors are aggressive estrogen receptor-α (ERα) negative breast carcinomas that have been identified due to their peculiar gene expression profile [
1‐
3]. Such tumors display a stem cell-like gene expression profile, including the over-expression of cancer stem cells (CSCs) markers, such as CD133 [
1,
4] and CD44 [
5‐
9]. CD44 and CD133 are also over-expressed in multicellular spheroids (called mammospheres), derived from breast cancer tissues and cell lines [
10,
11]. Mammosphere-forming subpopulation of breast cancer cells are endowed with highly enhanced tumor-initiating capability and with resistance to cancer therapy, and are currently dubbed as breast CSCs [
12‐
14]. Similarly to basal-like tumors, breast CSCs lack ERα expression [
1‐
3,
15,
16]. Basal-like tumors also over-express the pro-inflammatory cytokine Interleukin-6 (IL-6), a potent growth factor for breast cancer cells that enhances mammospheres growth capacity and malignant features in a paracrine/autocrine fashion [
3,
5,
10].
Basal-like breast cancers carry inactivating mutations of the tumor suppressor
p53 in about 80% of cases [
1‐
3]. It has been reported that p53 represses the expression of IL-6 and CD44, via direct promoter binding [
17,
18]. p53 exerts various check-point activities, including the repression of gene transcription through the methylation of DNA promoters, a mechanism of epigenetic regulation catalyzed by DNA (cytosine-5)-methyltransferases at CpG dyads dinucleotides [
19‐
21]. Interestingly, basal-like cells and tissues exhibit a peculiar promoter methylation pattern and over-express genes involved in genomic DNA and histone methylation [
22‐
25].
We therefore hypothesized that IL-6, CD44, CD133 and ERα take part to the basal-like gene expression profile throughout the epigenetic modification of their promoter regions.
We show that p53 deficiency induces the loss of methylation at the IL-6 promoter. This phenomenon starts an autocrine IL-6 loop that favours the loss of methylation at IL-6, CD44 and CD133 promoter 1, as well as the gain of methylation at ERα promoter. In parallel, the expression of IL-6, CD44 and CD133 is enhanced, and that of ERα is blunted. Moreover, IL-6 induces the methylation of IL-6 distal promoter and of CD133 promoter region 2, which contain putative repressor binding sites.
We conclude that p53 deficiency induces an IL-6 dependent epigenetic reprogramming that drives breast carcinoma cells towards a basal-like/stem cell-like gene expression profile.
Materials and methods
Chemicals and reagents
αIL-6, a monoclonal antibody that blocks the IL-6 receptor/ligand interaction [
10], recombinant human IL-6, 4-hydroxytamoxifen (4OHT, Tamoxifen) and the demethylathing agent 5-aza-2'-deoxycytidine (5azadC) were purchased from Sigma (Sigma, St-Louis, MO, USA).
Cell cultures
MCF-7 cells (carrying wild type p53) were cultured in RPMI medium supplemented with fetal bovine serum (FBS 10%), 100 IU/mL penicillin, and 100 μg/mL streptomycin. MCF-7 cells stably transduced with pBabe retroviral vector encoding p53 dominant-negative mini-protein were cultured as previously described [
26,
27]. MCF-7 derived mammospheres were obtained as previously described [
4,
10,
27]. P53 deficient MDA-MB231 breast cancer cell line (carrying R280K mutation) [
5] were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with fetal bovine serum (FBS 10%), 100 IU/mL penicillin, and 100 μg/mL streptomycin.
RNA extraction and RT-PCR analysis
Total RNA was extracted from cultured cells using the RNA-extracting reagent TRIzol (Invitrogen) according to the manufacturer's instructions.
Reverse transcription reaction was performed in a 20 μl volume with 2 μg of total RNA using the M-MLV Reverse Transcriptase, following the manufacturer's protocol. Oligo-(dT) 12-18 primers (Invitrogen) were used for the first strand synthesis. PCR primers (Additional file
1 Table 1) and reagents were purchased from Invitrogen.
Transient RNA interference
Double-strand RNA oligonucleotides (siRNA) directed against p53, IL-6 and ERα mRNA (Stealth validated RNAi DuoPaks), and appropriate control scrambled siRNA, were purchased from Invitrogen. siRNAs were transfected to adherent MCF-7 cells (105 cells in a 3-cm2well) at a concentration of 1 μg/well using Lipofectamine 2000 (Invitrogen).
DNA methylation assay
Methylation specific PCR was performed as previously described [
28]. DNA was extracted with phenol/chloroform (Sigma) and Proteinase K (Invitrogen, Carlsbad, CA, USA) and was bisulphite-modified with EZ-Methylation Gold-Kit (Zymo Research Corporation Orange, CA U.S.A) according to the manufacturer's instructions. Bisulphite modified DNA was amplified with primers designed using design Methyl Express
® Software v1.0 (Applied Biosystems Foster City, CA USA) and Beacon Designer 3.0 (Premier Biosoft International, Palo Alto CA USA; Additional file
2 Table 2). PCR primers and reagents were purchased from Invitrogen. PCR protocols were performed as follows: pre-denaturation step at 95°C for 2 min, 35 to 40 cycles of denaturation at 95°C for 30 sec, annealing at the appropriate temperature for 30 sec, extension at 72°C for 1 min; final extension at 72°C for 7 min. Sequence of genomic DNA promoters are:
IL-6 [GenBank: M18403],
CD133 p1 and p2 promoters [GenBank: ay275524],
CD44 [GenBank: M59040],
ERα [GenBank: X03635] (Additional file
3 Figure S1). Search for transcriptional factor binding sites was performed by TESS: Transcription Element Search Software on the WWW, Jonathan Schug and G. Christian Overton, Technical Report CBIL-TR-1997-1001-v0.0 Computational Biology and Informatics Laboratory, School of Medicine University of Pennsylvania, 1997 URL:
http://www.cbil.upenn.edu/tess.
Fluorometry
Amplified fragments were resolved onto a 1.8% agarose gel with ethidium bromide.
Gels were imaged with FluorSMultiImager (Bio-Rad, Hercules, CA) using UV excitation and a barrier filter of 520 nm. Emission of amplified bands was analysed with QuantityOne 4.6.6 software (Bio-Rad) using the same reading frame. Given values are ratios between the Unmethylated "U" and Methylated "M" band emissions of the same fragment. Experimental values were normalized with the U/M ratio of control fragments to which a value of 1 was assigned.
Luciferase Assay
DNA transfection of MCF-7 cells was performed with Lipofectamine 2000 (Invitrogen). One day before transfection, the cells were seeded at a density of 1.5 × 10
5 cells/well on 6-well plates and transfected with 1 μg of luciferase reporters driven by either p53 responsive elements (Stratagene, La Jolla, CA USA), or the -2,161 to -41bp IL6 promoter fragment (kindly provided by W. L. Farrar, NCI-Frederick Cancer Research and Development Center, USA) [
29], or the -1192bp to +10b CD133 promoter 1 fragment (Kindly provided by K. Tabu, Department of Stem Cell Regulation, Medical Research Institute, Tokyo Medical and Dental University, Japan) [
30]. IL-6 promoter reporter activity was also tested when co-transfected with 1 μg of IRF-1 or IRF-2 encoding pCAG vectors (kindly provided by T. Taniguchi, Department of Immunology, Graduate School of Medicine and Faculty of Medicine, University of Tokyo, Japan).
[
31]. Firefly Luciferase was normalized by co-trasfecting 10 ng of Thymidine Kinase Renilla Luciferase reporter (Promega Corporation, Madison Wisconsin USA). All luciferase assays were performed in triplicates following manufacturer's instructions (Promega).
Luciferase assay on in vitro methylated Luciferase reporter
In vitro plasmid DNA methylation was performed as previously described [
30]. Briefly, 4 μg of -1192bp to +10b CD133 promoter 1 fragment [
30] were incubated with 5 units of SssI (CpG) methylase (Zymo Research Corporation) for 4 h, per 1 μg of plasmid DNA in presence (methylated) or absence (unmethylated) of 0.64 mM S-adenosylmethionine. After phenol purification, equal amounts of methylated and unmethylated reporter constructs were assessed by digestion with the methylation-sensitive restriction enzyme HpaII (Promega) and were then assessed in a luciferase assay as above described.
Western blot
Protein concentration was determined by Protein Assay reagent (Bio-Rad, Richmond, CA, USA). Sixty μg of proteins were separated by SDS-PAGE and transferred to a nitrocellulose filter that was subsequently incubated with TBS buffer containing 5% dried nonfat-milk for 2 hour at room temperature (RT). Filters were probed with mouse monoclonal antibodies to p53 (DO-1 Santa Cruz Biotechnology, Santa Cruz, CA, USA), human STAT-3 and phosphorylated STAT-3 (Cell Signaling Technology, Danvers, MA USA) and to β-Actin (Santa Cruz). Bound antibodies were detected with peroxidase-labelled goat antibody to mouse or rabbit IgG and visualized by enhanced chemiluminescence reagents (Amersham Pharmacia Biotech, Freiburg, Germany).
IL-6 ELISA Assay
Quantitative detection of human IL-6 in cell culture supernatants was determined by Human IL-6 ELISA kit (Immunological sciences, Roma, Italy). The assay was performed in duplicates, following manufacturer's instructions. The plate was read by Thermo Labsystems Multiskan Ascent Photometric plate reader for 96 and 384 well plates (American Instrument Exchange, Inc., Haverhill, MA, USA).
Immunofluorescence
Adherent cells (seeded at a density of 5 × 104) and mammospheres were fixed with 4% paraformaldehyde for 10 minutes, permeabilized with Triton X-100 0.2% for 30 minutes and incubated with anti-CD44 mouse monoclonal antibody (1:250, Cell Signaling) in PBS-1%BSA for 1 h at 37°C. After PBS washing, cells and mammospheres were incubated with anti-mouse fluorescein-conjugated antibody (1:250, Santa Cruz) in PBS-1% BSA for 45 min at 37°C in dark room and with DAPI solution (1:1000, 4',6-diamidino-2-phenylindole, KPL, Gaithersburg, MD USA) for 15 min, and mounted in anti-fade Pro long reagent mounting medium (Molecular Probes Inc, Eugene, Oregon, USA). Images were captured using a Leica DMI 6000B inverted microscope (Leica Microsystems GmbH, Wetzlar, Germany).
Statistical analysis
Data were analyzed by t-Student test (SPSS, Chicago, IL, USA). Data were considered significant when p < 0.05.
Discussion
In this investigation, we show that the abrogation of p53 function, a distinctive feature of basal-like breast carcinomas, is functionally associated with the loss of methylation at the
IL-6 proximal promoter, a crucial region for
IL-6 gene expression [
29,
31]. In p53 deficient breast cancer cells, the loss of methylation at
IL-6 proximal promoter is maintained by an autocrine loop which further takes to an
IL-6 dependent loss of methylation at
CD133 and
CD44 promoters. These epigenetic modifications also occur when recombinant IL-6 is exogenously administered to the cells. In addition, IL-6 administration elicits the gain of methylation at
ERα promoter, whose epigenetic regulation is of primary importance in breast cancer biology [
1‐
3,
33,
34]. Paralleling these phenomena, we observed an up-regulation of CD44 and CD133 mRNA coupled with a down-regulation of ER
α mRNA. Current literature indicates that the above pattern of gene expression is proper of basal-like tumors and CSCs [
2,
3,
15,
16,
27].
Previously, IL-6 itself was found to be over-expressed in basal-like tumors and CSCs, and to enhance mammosphere forming capacity [
10,
35]. Accordingly, we report in the present work that MCF-7 derived mammospheres show an over-expression of
IL-6,
CD133 and
CD44 genes, as a possible consequence of the loss of promoter methylation. Considering the close relationship existing between mammospheres and CSCs [
4,
10‐
14], these findings support the very recent observation [
35] that IL-6 driven epigenetic changes are associated with CSCs features in breast cancer cells. In line with this reasoning, both CD133 and CD44 have been previously reported to be regulated by promoter methylation [
30,
31,
36]. Moreover, multiple epigenetic promoter modifications have been found to control IL-6 gene expression [
32,
37]. In fact, basal-like cancer show complex changes in the genomic DNA methylation pattern [
22‐
25]. Here, we report that at least two genomic regions located in the promoters of IL-6 and
CD133 (
IL-6dist,
CD133p2 see Additional file
3 Figure S1 E and F) genes gain methylation in response to IL-6. In both regions, binding sites for putative repressors (IRF-1/2 and ERα) are likely to be present. We therefore we speculate that
IL-6 and
CD133 gene transcription can be enhanced by a combination of loss and gain of methylation at genomic regions with opposite functional roles.
As far as protein expression, IL-6 secretion was found to substantially parallel the changes observed at mRNA level. Further, we observed that, similarly to what it has been recently reported [
38], CD44 protein was detectable in a low percentage of MCF-7 cells and that CD44 expressing cells became more frequent after the administration of IL-6 for at least 72 hours (Additional file
8 Figure 6S A). Moreover, long term (96 hours) exposure of MCF-7 cells to IL-6 elicited a substantial increase in CD44 expressing cells and in the generation of CD44-expressing mammospheres (Additional file
8 Figure 6S B). Of interest here is the finding that, despite the strict regulation of CD133 mRNA by promoter methylation [
30,
39], CD133 protein expression was not detected in our experimental models. This was not a completely unexpected finding, because recent published observations on colon CSCs show that CD133 mRNA expression is present in both CD133 positive and CD133 negative colon cancer cells, and that CD133 protein undergoes epitope masking during differentiation [
40]. It could be therefore hypothesized that CD133 negative cells expressing CD133 mRNA represent a "primed" population, ready to translate the protein under appropriate environmental conditions or that CD133 mRNA itself exerts regulatory functions. Our preliminary data suggest that CD133 mRNA undergoes cytoplasmic stabilization to be fully expressed in breast cancer cells (D'Uva et al., ms in preparation).
Overall, our data suggest that a remodelling of gene expression toward a basal/stem cell like phenotype may entail a complex reshaping of promoter methylation profile, where a loss of and gain of methylation at different promoter regions occurs (Additional file
9 Figure S7). The above changes are facilitated by the presence of functional p53 impairment. Interestingly, loss of p53 function was recently associated to the shift of cell division from an asymmetric to a symmetric pattern in breast cancer stem cells [
41]. Such a phenomenon was proposed as a main mechanism fuelling tumor growth [
41]. It could be therefore interesting to investigate whether genes shaping cell division patterns are part of the epigenetic modifications occurring in basal-like tumors [
22‐
25].
This study contributes to recent literature supporting the notion that epigenetic modifications driven by IL-6 are of relevance to determine the gene expression profile of cancer cells [
42,
43], and we can conclude that IL-6 blockage holds promises as a potential therapeutic strategy to combat breast cancer.
Acknowledgements
This work has been supported by Fondazione Banca del Monte di Bologna e Ravenna to P. Chieco, and by University of Bologna RFO funds-ex 60%, Cornelia Pallotti and Roberto Pallotti Foundation and PRIN 2008 "Clinical, diagnostic and therapeutics implications of studies of breast cancer stem cell" to M. Bonafè. We also thank Fondazione Cassa di Risparmio in Bologna for supporting the Center of Applied Biomedical Research. We thank W. L. Farrar (Cytokine Molecular Section, Laboratory of Molecular Immunoregulation, NCI-Frederick Cancer Research and Development Center) for IL-6 Promoter-Luc, T. Taniguchi (Department of Immunology, Graduate School of Medicine and Faculty of Medicine, University of Tokyo) for pCAG-IRF1 and pCAG-IRF2, K. Tabu (Department of Stem Cell Regulation, Medical Research Institute, Tokyo Medical and Dental University) for CD133promoters-Luc.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
LDA carried out experimental assays on methylation, manuscript drafting and experimental design. PS carried out RT-PCR analysis and experimental design. GS carried out data on mammospheres, RT-PCR analysis and data analysis. VM participated in cell cultures, viral infections and plasmid amplification. GDU participated in RT-PCR analysis, viral infections and Luciferase assay. PC helped to manuscript drafting and data analysis. MB participated in manuscript drafting, experimental conception and data interpretation. All the Authors read and approved the final version of the manuscript.