Cell cycle-dependent regulation of the bi-directional overlapping promoter of human BRCA2/ZAR2 genes in breast cancer cells

Human breast cancer cells were obtained from ATCC (Manassas, VA) and cultured in ATCC-recommended media [21, 24]. We synchronized the cells by serum starvation and evaluated by FACS analysis, as described earlier [21]. For transfection and synchronization experiments, we transfected the cells with the plasmids (see below) and let them recover for 2 h in RPMI medium with 10% fetal bovine serum (FBS). This complete medium was then replaced with starvation medium (RPMI 1640, phenol red free, 0% FBS). After 36 h, the cells were stimulated to re-enter the proliferative cell cycle by replacing the starvation medium with medium containing 20% FBS. Cells were harvested at specific time points following serum stimulation and were processed. The progression of cells through the cell cycle during these experiments was monitored by flow cytometric analysis of replicate samples of propidium iodide-stained cells [18, 21]. Each transfection experiment was repeated at least twice with triplicate samples each time. We used different human cell types including human mammary epithelial cells (HMEC), human breast cancer cells like MCF7, MDA-MB-468, MDA-MB-231, BT549, as described [21], for further verification and confirmation experiments.

The human BRCA2 gene promoter (-187 to +310) was PCR amplified from the genomic DNA using primers P1 and P2 (Table 1), and cloned into pCR2.0-Topo (Invitrogen, Carlsbad, CA). The cloned insert (497 bp) was digested with Eco RI and cloned at the Eco RI site of the pRL-Null vector (Promega, Madison, WI). The resulting plasmids, pRL-FP (forward promoter) and pRL-RP (reverse promoter) (Fig. 1B), were transiently transfected into MCF7 cells using Lipofectamine 2000 transfection reagent (Invitrogen) along with the pGL3-control vector (Promega). Protein lysates were prepared from the cells, and dual-luciferase activities were measured as described previously [21, 24]. Renilla luciferase (Rluc) activity was normalized with respect to firefly luciferase (Fluc) activity and presented as a ratio (relative light units; RLU). We also made a dual reporter construct (Fig. 1B) in which the BRCA2/ZAR2 promoter sequence is flanked by two reporter gene ORFs, Rluc is transcribed by forward (BRCA2) activity and Fluc is transcribed by the reverse (ZAR2) activity. To clone the firefly luciferase (Fluc) gene opposite to the Rluc gene in pRL-FP, Fluc ORF was taken out from pGL3-basic plasmid with Hind III and Sal I digestions and the DNA fragment was gel purified. The Fluc ORF was then cloned at the Hind III/Xho I sites of pRL-FP to obtain dual reporter plasmid pRL-DR. The identities of the plasmid constructs were verified by restriction mapping and nucleotide sequencing.

The transcription start site of the ZAR2 gene was determined with the reagents from the GeneRacer Kit (Invitrogen). RNA was isolated from MCF7 cells using TRIzol reagent. The DNAse (RQ1, Promega)-treated RNA was then digested with calf intestinal phosphatase to remove 5'-phosphate from broken RNAs (if any). The 5'-caps of the intact mRNAs were removed by digestion with tobacco acid pyrophosphatase followed by ligation of a RNA oligonucleotide (44 nt) by T4 RNA ligase following suppliers protocols and using their reagents. These RNAs were used as template to make the cDNAs with GeneRacer oligo(dT) primer. The 5'-end of the ZAR2 mRNA was amplified using the GeneRacer 5'-primer and ZAR2 gene specific antisense primer (P3 and P4, Table 1) and Platinum Pfx DNA polymerase (Invitrogen). The PCR conditions were 1 cycle at 94°C for 2 min; 5 cycles at 94°C for 30 sec, 72°C for 2 min; 5 cycles at 94°C for 30 s, 70°C for 2 min; 25 cycles at 94°C for 30 sec, 60°C for 30 s, and 68°C for 2 min, and 1 cycle at 68°C for 10 min. The PCR product (814 bp: 770 bp from the ZAR2 mRNA + 44 bp from the RNA anchor) was gel purified, cloned and the nucleotide sequences were determined [21, 24].

Human ZAR2 (hZAR2) gene ORF (NM_001136571; 1-966 bp) was PCR amplified from the cDNAs derived from MCF7 cells with Hind III and Bam HI site-containing primers (P5 and P6, Table 1). The reverse primer did not have the endogenous stop codon. The PCR product was cloned at the Hind III/Bam HI sites of p3xFLAG-CMV-14 vector (Sigma Chemical Co. St Louis, MO) to get the C-terminal FLAG tagged construct. For the expression of N-terminal FLAG-tagged ZAR2, ZAR2 ORF was amplified from the cDNAs derived from MCF7 cells with Hind III and Xba I site-containing primers (P7 and P8, Table 1). The reverse primer retained the endogenous stop codon. The PCR product was cloned at the Hind III/Xba I sites of p3xFLAG-CMV-10 vector (Sigma). The clones were sequence verified. The plasmid DNA was transfected in the MCF7 cells using Lipofectamine 2000 (Invitrogen) following supplier's protocol. After 36 h, cells were lyzed in TRIzol (Invitrogen) for RNA isolation or in Cell lytic reagent with protease inhibitors (Sigma) for Western blotting analysis [21, 24]. Stable transfectants were also selected in G418 (1000 μg/ml) containing growth medium.

Cells were transfected in 6-well plate with either the C-terminal FLAG-tagged construct (p3XFLAG-CMV14 or p3XFLAG-CMV14-ZAR2) or with the N-terminal FLAG tagged constructs (p3XFLAG-CMV10 or p3XFLAG-CMV10-ZAR2) using Lipofectamine 2000 (Invitrogen). After 16 h of transfection, cells were trypsinized and plated in 8-well chamber slides for 24 h in complete growth medium, washed with PBS, fixed with ice-cold methanol for 10 min and permeabilized in 50 mM NH₄Cl and 0.2% Triton X100 in PBS. After blocking with 5% goat serum in PBS, the cells were incubated with anti-FLAG M2 monoclonal antibody (Sigma) in the blocking buffer overnight at 4°C [24]. After 16 h, the slides were washed 5 times with PBS and treated with secondary antibody conjugated with the red fluorescent dye (Alexa Fluor R555-conjugated donkey anti-mouse IgG, Invitrogen) for 1 h at room temperature. When performed dual labeling, unsynchronized MCF7 cells expressing C-terminal FLAG-tagged ZAR2 protein were processed similarly but incubated with both anti-FLAG M2 mouse monoclonal antibody (Sigma) and cyclin A monoclonal rabbit antibody (Abcam, Cambridge, MA) in the blocking buffer overnight at 4°C. After 16 h, the slides were washed 5 times with PBS and treated with secondary antibody conjugated with the red fluorescent dye (Alexa Fluor R555-conjugated donkey anti-mouse IgG, Invitrogen for FLAG) and the green fluorescent dye (Alexa Fluor R488-conjugated donkey anti-rabbit IgG, Invitrogen for cyclin A) for 1 h at room temperature. The cells were subsequently washed with PBS four times and stained with DAPI (Sigma) or Topro (Invitrogen). Finally, each slide was examined by confocal fluorescence microscopy (Nikon TE2000-U-CI confocal microscope). Each representative image was examined and digitally recorded at the same cellular level and magnification [24].

Total RNA was isolated from cultured cells using TRIzol reagent (Invitrogen) according to the manufacturer's instructions. Isolated RNA was treated with DNase (RQ1, Promega) and then first strand cDNA was synthesized for real-Time RT-PCR (19). First strand cDNAs were synthesized using iScript cDNA synthesis kit (Biorad) with 5 μg of total RNA per reaction. For the real-time RT-PCR reaction cDNA (0.5 μl) was used per well in a total reaction volume of 25 μl. The iQSYBR green supermix (Biorad) containing the antibody-mediated hot start iTaq DNA polymerase was used for the PCR reaction. RT-PCR conditions were 1 cycle at 95°C for 3 min; 40 cycles at 95°C for 30 sec and 55°C for 1 min; 1 cycle at 95°C for 1 min; 1 cycle of 55°C for 1 min then 80 cycles for 10 sec each with 0.5°C increment after cycle two starting at 55°C, to collect the melt curve data and hold at 20°C [21, 24]. The end point RT-PCR was done using Taq PCR master mix (Qiagen). PCR conditions were 1 cycle at 94°C for 5 min; 40 cycles: 94°C for 1 min, 55°C for 1 min and 72°C for 1 min; 1 cycle at 72°C for 10 min and then hold at 4°C.

MCF7 cells stably transfected to over express C-terminal FLAG-tagged ZAR2 were used for this study. ChIP assays were done with the EZ MagnaChIP kit reagents and protocols (Upstate-Millipore). Briefly, cells were treated with formaldehyde (1% final concentration) for 10 min at 37°C. Cross-linking was terminated with addition of glycine (0.125 M final concentration). Cells were washed twice with ice-cold PBS containing protease inhibitor cocktail (Sigma). The chromatin pellets were sonicated in SDS lysis buffer [1% SDS, 10 mM EDTA, 50 mM Tris (pH 8.1)] to an average DNA size of 500 bp with a Fisher model 50 Sonic Dismembranator using an optimized sonication condition. The sonicated extract was centrifuged for 10 min at maximum speed and diluted with ChIP dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl (pH 8.1), 167 mM NaCl). The diluted ChIP lysates were pre-cleared with Magna beads for 30 min at 4°C. Immunoprecipitations were performed at 4°C overnight with either FLAG (Sigma) or normal mouse IgG (Santa Cruz Biotechnology). After 1 h incubation with 20 μl Protein A-Magna beads suspension, the conjugates were collected by magnetic separator. Immunocomplexes were washed twice sequentially in low salt immune complex wash buffer [0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl (pH 8.1), 150 mM NaCl], high salt immune complex wash buffer [0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl (pH 8.1), 500 mM NaCl], LiCl immune complex wash buffer [0.25 M LiCl, 1% IGEPAL-CA-630, 1% deoxycholic acid, 1 mM EDTA, 10 mM Tris (pH 8.1)], and TE buffer [10 mM Tris-HCl, 1 mM EDTA (pH 8.0)]. Elution of the immunocaptured chromatin complexes were performed with ChIP elution buffer provided in the kit. The DNA-protein cross-linking was reversed by incubating at 62°C for 2 h with Proteinase K. DNA fragments were obtained using Qiagen DNA purification column. DNA samples and standards were analyzed using real-time PCR system (BioRad) and iQSYBR Green PCR Master Mix (BioRad). Primers used to amplify BRCA2/ZAR2 promoter were as described [21] (Table 1). The following cycling parameters were used: 95°C for 3 min, 40 cycles of 95°C for 15 s, 55°C for 30 s, and 72°C for 1 min. Dissociation curve analyses were performed to confirm specificity of the amplification products. All samples were run in triplicate and all data were normalized with control IgG and 1% input DNA amplification [21, 24].

ZAR2 mRNA was knocked down in MCF7 cells using 50 pmol/ml of ZAR2 specific stealth siRNA#1 (P15 and P16, Table 1) and stealth siRNA#2 (P17 and P18, Table 1). Corresponding control stealth siRNAs used were (#1: P19/P20 and #2: P21/P22, Table 1). Stealth siRNAs were custom designed and synthesized by Invitrogen. Transfection of the cells with siRNAs was done by lipofection following Invitrogen-provided protocol. After 68 h cells were lysed in TRIzol for RNA isolation. First strand cDNAs were synthesized using iSCRIPT cDNA synthesis kit (BioRad). Evaluation of the levels of ZAR2 and BRCA2 mRNAs was done by real-time RT-PCR using iQSYBR green Supermix (BioRad). To evaluate the effect of ZAR2 mRNA knockdown on the activities of the forward (BRCA2) and the reverse (ZAR2) promoters, MCF7 cells were transfected with 50 pmol/ml of the ZAR2 or control siRNA for 68 h. The cells were then transfected with the BRCA2 forward or reverse promoter containing reporter constructs (Fig. 1). Dual luciferase assay was performed after 24 h as described previously [21] following the supplier's protocol (Promega).

We studied the minimal promoter of human BRCA2 gene as established by several groups [15, 17‐21] (Fig. 1A). We cloned the 497 bp promoter DNA sequence (-187 to +310) of human BRCA2 gene in front of Renilla luciferase (Rluc) gene in pRL-Null plasmid (Promega) to obtain clones with the insert either in the forward or in the reverse orientation with respect to the luciferase gene (Fig. 1B). We also developed a dual reporter construct in which the firefly luciferase (Fluc) gene ORF is transcribed by the reverse activity of the cloned promoter and the forward activity of the promoter will transcribe Rluc gene ORF (Fig 1B) from a single plasmid DNA. The single reporter construct with the forward orientation of the promoter yielded the BRCA2 gene promoter activity whereas the construct with the reverse orientation of the promoter revealed the activity of ZAR2 gene. We found that both the forward and the reverse orientations of the promoter have significant promoter activities (Table 2). Surprisingly, the activity of the reverse promoter was significantly higher in comparison to that of the forward promoter in different human breast cancer cells. These were unsynchronized cells with 90-95% confluency and the difference between the two activities varied depending upon the cell line tested (Table 2). When these cells were transiently transfected with the reporter constructs and the promoter activities were assayed at 40-60% confluency, the forward activity was higher than the reverse activity (data not shown), suggesting cell cycle dependency in the regulation of these promoters.

Regulation of BRCA2 gene expression during cell cycle progression is well-documented [14, 18, 19, 21, 25]. To evaluate whether the forward and the reverse promoter activities of BRCA2 gene promoter is regulated during the cell cycle, we synchronized different human breast cancer cell lines that were transiently transfected with either of the single-reporter plasmid constructs (Fig. 1B) and pGL3-Control plasmid (as a transfection control; Promega). Our data suggest that ZAR2 gene promoter activity is 2-7 folds higher than the BRCA2 gene promoter activity in the G0/G1 phase cells whereas it is significantly less than the BRCA2 gene promoter activity in the S/G2 phase cells (Fig. 1C). Similar results were obtained with MCF10A and HMEC cells. The ratio of the Rluc/Fluc from the cells transiently transfected with the dual-reporter construct (Fig. 1B) increased when the cells were shifted to S/G2 phase from G0/G1 phase (Table 3). These data are in agreement with the suggestion that these two promoters are reciprocally regulated during the cell cycle.

Human ZAR2 gene is designated in the NCBI database as ZAR1L (#LOC646799). This gene is recently reported by in silico analysis in human as well as in bovine cells [22, 23]. It is reported to have four exons; exon 1 starts with the translation start codon. The 5'-UTR sequence for ZAR2 mRNA has not yet been reported. Since we hypothesized that ZAR2 gene is transcribed from the BRCA2 gene promoter, the transcription start site of ZAR2 gene should also be located within the 497 bp promoter. To determine the 5'-UTR as well as the transcription start site of the ZAR2 gene, we employed GeneRacer technology (Invitrogen). We amplified the 5'-RACE product (814 bp: 770 bp from mRNA + 44 bp from the Gene Racer RNA anchor) (Fig. 2A) and determined the nucleotide sequence of the cloned insert. Alignment of the nucleotide sequence of the 5'-RACE product with genomic DNA sequence revealed that the transcription start site (TSS) of ZAR2 is located within the exon 1 of BRCA2 gene while the exon 1-intron 1 junction of ZAR2 is located within the upstream sequence of the BRCA2 gene (Fig. 2B). As a control for the GeneRacer technique, we verified the transcription start site and the first splice donor site of human BRCA2 gene in parallel. The locations of the transcription start sites and the first splice donor site at the reverse strand of the BRCA2 gene promoter are shown in Fig. 2B. Relative maps of the BRCA2 and ZAR2 transcription start sites (TSS) in the bi-directional promoter tested are shown in Fig. 2C. These data suggest that BRCA2 and ZAR2 transcripts have 111 nt complementary and antiparallel overlaps with each other at their 5'-ends (see below).

The reciprocal relationship between BRCA2 and ZAR2 genes should be of significant biological importance because the bi-directional arrangements of these genes are highly conserved in several vertebrate species (Fig. 3A). Database analysis (NCBI GenBank) showed that although several other surrounding genes have changed in their relative positional arrangements, BRCA2/ZAR2 pair remains intact from bird to human and perhaps in others (Fig. 3A). Like BRCA2, ZAR2 is present in many vertebrates (Fig. 3B). Unlike BRCA2, which is found in many organisms from protozoa to human [26‐29], we could not find any significant ortholog or paralog of ZAR2 mRNA or protein in non-vertebrates or microbes by database search.

GeneRacer technology revealed a new exon and a new intron at the 5'-flank of the ZAR2 gene (Fig. 4A). The new exon 1 of ZAR2 is 245 bp, of which 111 bp and 134 bp are complementary and antiparallel to BRCA2 exon 1 and upstream sequences, respectively. With respect to BRCA2 gene transcription start site (TSS), the location of ZAR2 exon 1 is -134 to +111 (Fig. 2). Our study also revealed 431 bp additional sequences added before the translation start site of ZAR2 gene (Fig. 4B). Based upon these analyses, ZAR2 now has 5 exons and 4 introns (Fig. 4A) instead of 4 exons and 3 introns, as entered in the gene database. Our study revealed that the translation start site (ATG codon) is located within exon 2, instead of exon 1.

The nucleotide sequence of the ZAR2 cDNA is shown in Fig. 4B. The 676 nt 5'-untranslated region (UTR) of the ZAR2 mRNA has several upstream AUG codons (uAUGs) and out of frame upstream open reading frames (uORFs). Thus, translation of ZAR2 mRNA may potentially be regulated by these uAUGs and uORFs [30, 31].

The 966 nt coding sequence of the ZAR2 mRNA codes for a 321 amino acid protein (Fig. 5A). We found two unique C4 type zinc fingers at the carboxyl terminus of this protein (Fig. 5A). These fingers are Cys-X2-Cys-X23-Cys-X2-Cys and Cys-X2-Cys-X24-Cys-X2-Cys where X is any amino acid. Such zinc fingers are found in nuclear receptors and GATA family of transcription factors among others [32]. The uniqueness of the ZAR2 zinc fingers is that they are comparatively longer (23-24 amino acids) than those of other proteins. The ZAR2 protein is largely similar with the ZAR1 protein, particularly at the C-terminus and the zinc finger Cys residues are highly conserved (Fig. 5B). The zinc fingers are also highly conserved among the ZAR2 orthologs from different animals (Fig. 6). These analyses suggest that ZAR2 is a potential transcription factor, perhaps binding to the promoters of its target genes through its C-terminal zinc finger domains. Alternatively, ZAR2 could be a non-DNA binding transcriptional regulator like FOG-1 [33] and perhaps regulates the function of BRCA2 gene promoter by restraining some regulatory protein.

Subcellular location of a protein often reflects upon its biological function. To understand subcellular location of ZAR2 in human breast cells, we expressed N-terminal and C-terminal FLAG-tagged ZAR2 in these cells (Fig. 7A and 7B). Expression of both of these tagged proteins was necessary because we did not know whether tagging will compromise the subcellular localization of this protein. This was particularly important since the putative nuclear localization signals of ZAR2 are located at the C-terminal region of this protein (Fig. 5A). RT-PCR analysis with ZAR2 specific primer (P9, Table 1) and FLAG-tag specific primer (P23, Table 1) showed significant expression of the recombinant ZAR2 transcripts in the cells transiently transfected with either of the constructs (Fig. 7A). Similar data were obtained with Western blot analysis for the FLAG-tagged ZAR2 protein with anti-FLAG monoclonal antibody (Fig. 7B). Although ZAR2 is a zinc finger protein with attributes of a nuclear protein and has putative strong nuclear localization signals (Fig. 5A), our immunofluorescence microscopy analysis data with FLAG antibody showed predominant cytosolic localization of the ZAR2 protein in the transiently transfected human breast cells (Fig. 7C). This is true for both N-terminal as well as C-terminal FLAG-tagged ZAR2 proteins (Fig. 7C). Few cells in the population had significant levels of FLAG-tagged ZAR2 inside the nucleus (Fig. 7C). In an unsynchronized population of MCF7 cells transfected with FLAG-tagged ZAR2 construct also showed major presence of FLAG-ZAR2 (red) in the cytosol while the S-phase marker cyclin A (green) is predominantly located in the nucleus of the cell (Fig. 7D).

RT-PCR analysis showed the expression of BRCA2 and ZAR2 mRNAs in unsynchronized dividing human breast cells (Fig. 8A). To evaluate whether the expressions of BRCA2 and ZAR2 genes at G0/G1 and S/G2 phases follow the pattern of their promoter activities, we measured the mRNA levels by real-time RT-PCR. In all the cells tested BRCA2 and ZAR2 expressions are inversely related. ZAR2 mRNA levels are higher at the G0/G1 phase with significant lower levels of BRCA2 mRNA (Fig. 8B). On the other hand, at the S/G2 phase the levels of BRCA2 mRNA are significantly higher than those of ZAR2 mRNAs (Fig. 8B).

We synchronized C-terminal FLAG-tagged ZAR2-expressing MCF7 cells and evaluated the subcellular location of this protein at G0/G1 and S/G2 phase by immunofluorescence confocal microscopy using FLAG antibody. Our data suggest that ZAR2 protein is predominantly concentrated in the nucleus of these cells at the G0/G1 phase whereas it is mainly present in the cytosol at the S/G2 phase cells (Fig. 8C). Evaluation of ZAR2 protein distribution in the subcellular fractions by Western blotting analysis also revealed similar localization pattern (data not shown). These data suggest that not only the expression of ZAR2 gene is strictly regulated cell cycle-dependently but its subcellular localization is also controlled in a growth stage-dependent manner. ZAR2 and BRCA2 gene expressions are thus inversely related.

In order to understand the role of cell cycle-dependent regulation of ZAR2 gene expression on BRCA2 levels, we explored whether the C4-type zinc finger protein ZAR2 can bind to the BRCA2 promoter DNA. We used MCF7 cells over expressing C-terminal FLAG-tagged ZAR2 and employed quantitative ChIP techniques using commercially available FLAG antibody for this purpose. We found that ZAR2 binds tightly with BRCA2/ZAR2 bi-directional promoter preferentially at the G0/G1 phase of MCF7 cells (Fig. 9A and 9B). The nature (direct or indirect) and exact site(s) of binding of the ZAR2 protein to this promoter are yet to be determined. The outcome of over expression of ZAR2 and its binding to the BRCA2 promoter is not yet known. We found that over expression and DNA-binding of ZAR2 is associated with decrease in the level of BRCA2 mRNA (Fig. 10A) as well as inhibition of both BRCA2 and ZAR2 promoter activities (Fig. 10B), particularly at the S/G2 phase of MCF7 cells. Correlation between ZAR2 binding to the promoter and the repression of BRCA2 gene expression needs to be validated by rigorous mutational analyses.

To verify further the potential role of ZAR2 in the regulation of BRCA2 gene expression we knocked down the expression of ZAR2 by two different siRNAs designed from its ORF. When the mRNA levels of ZAR2 were knocked down to 50-55% at the G0/G1 phase of MCF7 cells, the levels of BRCA2 mRNA went up significantly (3.5-4 folds) (Fig. 11A). Interestingly, the activities of both of the ZAR2 and BRCA2 promoters were increased in the ZAR2 knocked down cells (Fig. 11B). Increased expressions of ZAR2 mRNA may explain why we could not down regulate the ZAR2 mRNA levels more than 50% with the siRNAs.