Background
The human breast cancer susceptibility gene 1 product, BRCA1 is involved in important cellular processes, including DNA repair, and loss of BRCA1 can result in genomic instability. Loss of BRCA1 expression occurs in a subset of breast cancer cases, and inherited mutations of the
BRCA1 gene account for about 5% of all breast cancer cases [
1‐
8]. The regulation of
BRCA1 expression has been studied extensively, including investigations of alternative mechanisms for reduced expression of
BRCA1 in sporadic cases [
9,
10]. A set of transcription factors and/or co-factors has been shown to regulate
BRCA1 expression through a region in close proximity to the
BRCA1 transcription start site. This proximal promoter region of
BRCA1 is bidirectional and includes a 218 bp intergenic region between the
BRCA1 and
NBR2 genes [
11]. Within this region, it has been demonstrated that a short segment (-204 to -148) relative to
BRCA1 exon 1a start site may be responsible for
BRCA1 promoter activities [
12]. Another report confirmed these findings using a different strategy of deletion analysis and showed that a slightly shorter region (-202 to -166) was required for
BRCA1 promoter activation [
13,
14]. Further studies of this region (-202 to -166) revealed two sub-elements: 1) a RIBS element (-204 to -182) that binds to and is activated by a GABPα/β complex [
15] and 2) a CREB/ATF1 element (-174 to -167) that acts as a constitutive transcriptional activation element bound by CREB [
16,
17]. In addition, the E2F family of transcription factors binds to two regions, -41 to -31 and -21 to -11, and activates or represses
BRCA1 expression depending on the co-factors recruited [
18,
19], BRCA1 itself has been shown to be one of the co-factors [
20]. An element (-40 to -25) that overlaps one of the E2F binding sites can be bound and activated by 53BP1 in a sequence-specific manner and functions as a positive regulatory element [
21]. An ER-α transcription complex binds an AP1 element (+246 to +250) and activates
BRCA1 transcription upon estrogen stimulation [
22]. This ER-α dependent activation can be modulated by an aromatic hydrocarbon receptor complex that binds two consecutive xenobiotic-responsive elements located upstream of the ER-α binding region (+17 to +21 and +175 to +179) [
23]. A relatively long segment in a 5 kb region in
BRCA1 intron 2 that is highly conserved in multiple species contains a CNS-1 (Conserved Nucleotide Site-1) and CNS-2, which appear to act as repression and activation elements, respectively [
24].
c-Myc is a transcription factor involved in growth, proliferation, differentiation, and apoptosis of cells and regulates up to 15% of human genes [
25]. c-Myc regulates transcription through several mechanisms, and
cis-regulatory elements modulate specific subsets of c-Myc targets. One of the
cis-regulatory elements, E box, is common in c-Myc targeted genes [
26]. Serial analysis of gene expression performed after adenoviral expression of c-Myc in primary human umbilical vein endothelial cells has implicated
BRCA1 as one of the activated gene targets for c-Myc [
27]. However, it was not clear whether c-Myc could transcriptionally regulate
BRCA1 expression through a
cis-regulatory element, particularly in breast cancer cells. In this report, we show that depletion of c-Myc is correlated with a reduction in
BRCA1 mRNA and BRCA1 protein levels and decreased
BRCA1 promoter activities were observed in the cells following depletion of c-Myc. On the other hand, ectopic expression of c-Myc activated
BRCA1 promoter activities. DNA sequence analysis revealed two novel E boxes within the distal
BRCA1 promoter. A chromatin immunoprecipitation assay demonstrated that c-Myc binds to these E boxes
in vivo. Furthermore, we show that
BRCA1 promoter/reporters containing nucleotide substitutions in these E boxes abrogate their c-Myc dependent activation. These data suggest that c-Myc activates
BRCA1 expression through the E boxes in the distal
BRCA1 promoter region. Finally we observed that cells treated with c-Myc specific siRNAs had reduced DNA repair activity, a biological process associated with BRCA1.
Methods
Cell culture
Breast cancer cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA). MCF-7 cells (ATCC: HTB-22) were cultured in DMEM (Mediatech, Inc., Manassas, VA) supplemented with 10% fetal bovine serum (Lonza, Inc, Allendale, NJ) and 1% antibiotic-antimycotic solution (Mediatech Inc.). MDA-MB-231 cells (ATCC: HTB-26) were cultured in RPMI1640 (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (Lonza, Inc.) and 1% antibiotic-antimycotic solution (Mediatech, Inc.).
Plasmid construction
A vector containing human c-Myc cDNA was obtained from ATCC (ATCC: 5233860) and used to construct a c-Myc-GFP expression vector designated pMYC-GFP. The full length c-Myc protein coding DNA fragment was inserted in the 5' end of GFP under the control of the cytomegalovirus (CMV) promoter. First the full-length coding region of c-Myc was amplified by PCR using primers (5'-TCCCGCGACG ATGCCCCTCA ACGTTAGCTTCA-3' and 5'-CACAAGAGTT CCGTAGCTGT TCAAGTTTGTG-3') and the c-Myc cDNA vector (ATCC 5233860) as a template. Then the PCR products were cloned into pcDNA3.1/CT-GFP-TOPO (Invitrogen) following the manufacturer's protocol, yielding pMYC-GFP. The c-Myc encoding sequence was confirmed by DNA sequence analysis to be identical to that of GenBank: BC000917.
A promoter/luciferase reporter construct containing the BRCA1 promoter region (-1714 to +42) in pGL4.10 (Promega, Madison, WI) was generated as follows. The required BRCA1 promoter region was amplified by PCR using the primer pair: 5'-CTA GGTACC TTGGGAGGGG GCTCGGGCAT GGC-3' and 5'-CAT AAGCTT CCAGGAAGTC TCAGCGAGCT CACG-3' (KpnI and HindIII sites were underlined, respectively) and human placental genomic DNA (Sigma-Aldrich, St. Louis, MO) was used as a template. The PCR products were digested with KpnI and HindIII, and the resulting fragments were inserted into pGL4.10 at identical sites. The resulting construct was named pCYL42. The inserted DNA sequence of the BRCA1 promoter was determined to be identical to that of GenBank: L78833 by DNA sequence analysis.
We used a QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA) to generate E box nucleotide substitutions following the manufacturer's protocol. To generate single E box nucleotide substitution constructs, we used pCYL42 as the original template and corresponding E box mX and E box mY primer pairs. The primers (sense strands shown) used are shown below with nucleotide substitutions underlined: E box mX: 5'-AGATTGGCTC TTACAAAATG TCCCTCAAAA CGAC-3'; and E box mY: 5'-GCGAGGGCTG CTAGAAAATT GTCACCTCGC ATTCT-3'. To generate the double E box mutant, the mutagenesis reactions were conducted using the primers E box mX and E box Y nucleotide substitution constructs as templates. All constructs were confirmed by DNA sequence analysis.
The pMyc-TA-Luc and pTA-Luc plasmids used as c-Myc activated promoter/luciferase controls were purchased (Clontech Laboratories, Inc., Mountain View, CA). The pMyc-TA-Luc vector is a derivative of pTA-Luc and contains six tandem copies of the E box consensus sequence (CACGTG) located upstream of the minimal TA promoter, followed by the firefly luciferase gene.
Small interfering RNA (siRNA)
All related reagents, including c-Myc siRNAs, control siRNA, and transfection reagents, were purchased from Qiagen (Valencia, CA). Transfections were performed according to the manufacturer's protocol.
Quantitative reverse transcriptase PCR
Total RNA was extracted from cultured cells using the RNeasy mini kit (Qiagen). Reverse transcription reactions were performed with the SuperScript III First-Strand Synthesis System (Invitrogen) using 1 μg of DNase-treated RNA and oligo (dT) primer. Real-time PCR was performed in a Light Cycler (Roche Diagnostics, Indianapolis, IN) using the LightCycler FastStart DNA Master PLUS SYBR Green I kit (Roche Diagnostics). Primers for
BRCA1 and glyceraldehyde-3-phosphate dehydrogenase (
GAPDH) were used as previously described [
9,
28]. Thermal cycling for amplification of
BRCA1 or
GAPDH was initiated by heating at 95°C for 10 min, followed by 40 cycles of denaturation at 94°C for 10 sec, annealing for 10 sec at 54°C and 57°C for
BRCA1 and
GAPDH, respectively, and elongation at 72°C for 15 sec. After completion of the PCR cycles, melting curve analyses and electrophoreses of the products on 2% agarose gels were performed to validate generation of each specific, expected PCR product. The fold change in
BRCA1 cDNA (target gene) relative to the
GAPDH control was determined by the 2
-ΔCt method [
29]. Experiments were conducted independently twice in triplicate.
Western blot
Western blotting was carried out as previously described [
30]. Mouse anti-BRCA1 antibody (AB-4, Oncogene, San Diego, CA), mouse anti-c-Myc antibody (sc-40, Santa Cruz Biotechnology, Inc., Santa Cruz, CA), and goat anti-actin antibody (sc-1616, Santa Cruz Biotechnology, Inc.) were used for western blotting. An ECL western blotting detection reagent kit (GE Healthcare, Buckinghamshire, UK) was used for detection.
Transient transfection and luciferase assay
One day prior to transfection, 6 × 105 cells were seeded into one well of a six-well tissue-culture plate (BD Biosciences, San Jose, CA). After 24 hours, cotransfections were conducted. The transfection mixture contained a promoter/luciferase reporter construct (980 ng), a CMV/renilla luciferase vector, pGL4.75 (20 ng, Promega) as a transfection efficiency control, and an expression construct for c-Myc (pMYC-GFP, 1000 ng or other indicated amounts) or additional control vector pcDNA3.1/CT-GFP (Invitrogen) to give consistent amounts of total vector of 2000 ng. Then 6 μl of FuGENE HD (Roche Diagnostics) was added to the mixture, and cotransfections were done according to the manufacturer's instruction. Forty-eight hours after transfection, cell lysates were prepared using passive lysis buffer (Promega). Luciferase activities in the lysates were measured using a dual luciferase assay system (Promega) with a 1450 microbeta trilux jet scintillation and luminescence counter (PerkinElmer Life and Analytical Sciences, Downers Grove, IL). Experiments were performed in triplicate.
In reporter assays with depletion of c-Myc in MCF-7 and MDA-MB-231 cells, 24 hours after siRNA treatment, cotransfections were done using the mixture of pCYL42 (1000 ng) and pGL4.75 (20 ng) combined with 2 μl of FuGENE HD as described above. Luciferase activity was determined 48 hours after the transfection. Experiments were performed in triplicate.
Chromatin immunoprecipitation
Chromatin immunoprecipitation (ChIP) assays were performed following the manufacturer's protocol (Upstate Biotechnology, Lake Placid, NY) using 2 × 106 MCF-7 or MDA-MB-231 cells. An anti-c-Myc antibody (sc-764, Santa Cruz Biotechnology, Inc.) or an anti-CTCF antibody (sc-15914, Santa Cruz Biotechnology, Inc) was used to precipitate DNA-protein complexes, and the normal isotype-matched IgG from the same species was used as negative control. The PCR primers used for detecting DNA fragments containing E boxes are: E box X: ChIPX-FW: 5'-AATGCAAAGA CCGTCCGCTG CCA-3'and ChIPX-RV: 5'-TCCACCCCTC AGCCCCAGTG TTT-3'; E box Y: ChIPY-FW: 5'-TGAAGGGCTC CTCCAGCACG-3' and ChIPY-RV: 5'-TGAGGGACCG AGTGGGCGAA-3', and non E box: ChIPNE-FW: 5'-CGAGAGACGC TTGGCTCTTT CTGT-3' and ChIPNE-RV: 5'-GCCCAGTTAT CTGAGAAACC CCAC-3'. The amplified PCR product size was 233 bps for the detection of E box X, 146 bps for E box Y, and 214 bps for non E box, respectively. The band densities of the targeted PCR products were quantified with the software QUANTITY ONE version 4.5.1 (Bio-Rad, Hercules, CA).
DNA repair assay
We used a fluorescence-based DR-GFP reporter system with modifications [
31,
32]. The efficiency of homologous recombination was assessed using a restriction endonuclease I-SceI expression plasmid pCβASce and pDR-GFP, an I-SceI repair reporter plasmid composed of two differentially mutated GFP genes, one of which contained a unique I-SceI restriction site. A double strand break of DR-GFP plasmids created by I-SceI digestion was repaired by gene conversion to produce a functional GFP. The cells expressing GFP, representing homologous recombination activity, were detected with flow cytometry. Briefly, MDA-MB-231 cells were seeded in a well of a six well plate. On the next day, the cells were treated with equal amounts of siRNA against c-Myc or control siRNA with Hiperfect as instructed by the manufacturer (Qiagen). 48 hours later, pDR-GFP plasmid, along with either pCβASce (expressing endonuclease I-SceI to create double strand breaks (DSBs)) or pCAGGS (empty vector) were co-transfected into cells by Fugene HD (Roche) as described above. After incubation for 48 hours, cells were trypsinized, harvested, and finally suspended in PBS. Consequently, the cells were analyzed at the flow cytometry facility of the University of Chicago with a Becton Dickinson FCAScan (BD Biosciences). The channel FL-1 (green) and FL-2 (orange) were recorded and used to calculate the frequency of GFP positive cells.
Statistical analysis
For quantitative RT-PCR, after logarithmic transformation,
BRCA1 mRNA level was analyzed using ANOVA. Then the regression coefficients were transformed back to obtain the fold change or geometric mean difference between study groups [
33].
Means and standard deviations (SDs) were calculated for relative luciferase activities, densitometry ratios, and homologous recombinant efficiencies. Two-tailed Student's t-tests were used to compare only 2 groups. ANOVA were implemented for comparisons with multiple groups. After the overall analysis, Tukey tests were used to make the pairwise comparison.
Discussion
In the present report, we used a targeted identification strategy for characterizing regulatory DNA elements to show transcriptional activation of
BRCA1 by c-Myc through c-Myc's interaction with two E boxes in the distal
BRCA1 promoter region. Genome-wide analysis has shown that c-Myc could bind to the regulatory regions of up to 15% of all human genes, and these data also indicated that c-Myc-DNA interactions were not sufficient for promoter regulation in some cases [
35,
36]. Indeed, it has been shown that although c-Myc binds to the
CCL5 gene promoter in both MCF-7 and MDA-MB-231 cells, c-Myc-dependent regulation of
CCL5 was evident in MCF-7 but not in MDA-MB-231 cells. Consistently, gene expression profiling analysis has shown that the spectrum of genes regulated by c-Myc in these cell lines is intrinsic to each cell line [
37]. While the mechanism of how c-Myc functions in such a cell context-dependent manner is not clear, possibilities include, but are not limited to, post-translational modifications of c-Myc. For example, phosphorylation and acethylation of c-Myc could affect its protein-DNA and protein-protein interactions, which would contribute to its selective binding of target gene promoters. Recently Benassi
et al. showed that phosphorylation of S62 of c-Myc could increase its binding ability to the promoter region of the γ-glutamyl cysteine synthetase gene [
38]. However, the conclusive connections between c-Myc-DNA binding abilities and c-Myc transcriptional activities in a given condition are still obscure, as in the case of the
CCL5 gene. In this study, we identified two E boxes within the
BRCA1 promoter region and provided evidence that these were necessary for c-Myc-dependent promoter activation in MCF-7 and MDA-MB-231 cells (Figure
1). Although the co-factors recruited by c-Myc to these sites were not elucidated in our present experiments, we speculate that it is highly likely that a c-Myc-containing transcription complex would play an important role in
BRCA1 expression.
Our data showed that
BRCA1 activation by c-Myc only increased promoter activities by 1.8 fold under the conditions used (Figure
2B), and disrupting c-Myc-responsive elements (two E boxes) by different combinations correlated with loss of promoter activity ranging from 30-40% (Figure
5A and
5C). This observation is consistent with a previous report that
BRCA1 mRNA expression was detectable in
c-Myc knockout
c-Myc
-/-
Rat1A cells, and expression of c-Myc by serum stimulation in the parent
c-Myc
+/+
Rat1A cells slightly increased
BRCA1 mRNA expression [
27].
BRCA1 expression activated by c-Myc could be the final outcome of the interplay within multi-component transcriptional networks containing factors important for breast tumorigenesis. It has been shown that HIF-1α transcriptional machinery activated by hypoxia signaling pathway abrogates c-Myc activation of
BRCA1 expression in colon cancer cells [
39]. c-Myc activates the expression of
p53 through an E box within the
p53 promoter [
40]. On the other hand, p53 represses
c-Myc transcription by binding to the
c-Myc promoter and recruiting the general repressor mSin3a; this repression is required for cell cycle arrest and differentiation but not apoptosis [
41]. In addition, p53 up-regulates
ER-α by increasing gene transcription [
42]. c-Myc is up-regulated by activated ER-α and plays a critical role in enhancing estrogen-induced breast cell proliferation [
43,
44]. Although not directly mediated by the ER-α pathway, estrogen stimulation up-regulates
p53 expression [
45]. Thus, further studies to elucidate transcription factor occupancy on the
BRCA1 promoter, including proximal and distal regions, in breast cancer cells with distinct molecular signatures would be critical to understanding the complex regulation of
BRCA1 expression during tumor progression.
Accumulating evidence has demonstrated that BRCA1 is a major component of the DNA damage repair complex required for normal cellular processes. Depletion of BRCA1 has been associated with decreased DNA damage repair and increased chromosomal instability [
46]. In order to maintain the basal level of DNA damage repair, BRCA1 may be required for c-Myc associated cell proliferation. In this report we show that c-Myc can activate
BRCA1 expression in breast cancer cells, and depletion of c-Myc reduced BRCA1-dependent DNA repair. However, the complicated interaction between c-Myc and BRCA1 may be cell context dependent. In breast cancer patients,
c-Myc amplification was found more often in patients with BRCA1 deficiency than in patients with normal BRCA1 [
47], suggesting that c-Myc amplification could be linked to decreased BRCA1 but the mechanism is poorly understood. These data suggested that c-Myc overexpression and BRCA1 loss seemed highly correlated in a large portion of basal-like breast cancers. This could indicate that loss of BRCA1 with c-Myc overexpression might lead to the development of basal-like breast cancer. Interestingly, we have reported that reduction of BRCA1 expression could be due to
BRCA1 promoter methylation [
9]. c-Myc has been shown to recruit DNA methyl transferases such as DNMT3a to repress p21Cip1 gene expression [
48]. It raises the questions of c-Myc's involvement in the DNA methylation complex in breast cancer cells; particularly if
BRCA1 is a target. It is worth noting that aberrant expression of DNA methyl transferase or its alternative splicing forms have been detected in cancerous cells, and could affect the distribution of DNA methylation [
49].
Although our data support a model in which
BRCA1 expression is regulated by transcription factors, such as c-Myc, it has been shown that other mechanisms could also affect
BRCA1 expression. Epigenetic regulation such as
BRCA1 promoter methylation has been shown to account for BRCA1 loss in 10-31% of sporadic breast cancer tumors [
9]. The accessibility of transcription factors to the methylated
BRCA1 promoter is reduced, and consequently the expression of
BRCA1 mRNA and BRCA1 protein is decreased [
33]. On the other hand, nucleosome occupancy has been identified as an alternative means to regulate
BRCA1 expression [
50]. Taken together, chromatin structure, methylation status of the gene promoter and dynamic transcription factor recruitment may work together to affect BRCA1 expression in breast cancer cells. Newly identified regulation by means of microRNA could also play a role in
BRCA1 expression at the translational level. Aberrant expression of a set of microRNAs have been found in breast tumor samples, and some of these microRNAs could target
BRCA1 mRNA [
51], thus possibly regulating
BRCA1 expression.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
YC and JX designed and conducted the experiments, and wrote the manuscript. SB and CC performed experiments, and DH participated in the experiments and conducted the statistical analysis. OIO conceived the study, participated in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.