Background
The tumor suppressor protein BRCA2 is implicated in the regulated growth and proliferation of human breast [
1‐
4], prostate [
5,
6], ovarian [
7,
8], esophageal [
9], and pancreatic [
10,
11] cells. About 25% of autosomal dominant familial breast cancers are proposed to be caused by germline mutations in
BRCA2 gene [
12,
13]. The mutations of
BRCA2 gene predispose the cells towards neoplastic development. BRCA2 protein is over-expressed in most of the sporadic breast cancer cells [
1‐
4]. The consequence of this over-expression of BRCA2 is not clearly understood. The notion could be that unique cellular mechanisms are triggered in the breast cancer cells to stimulate
BRCA2 gene expression as a temporary measure to regulate the growth of the breast cancer cells. One potential mechanism of BRCA2 involvement in breast cancer progression may be through deregulation of the
BRCA2 gene expression.
In humans,
BRCA2 is a 3418-amino acid protein localized in the nucleus [
14,
15]. Loss of BRCA2 function has been shown to lead to centrosome amplification, chromosomal rearrangement, aneuploidy, and reduced efficiency of homologous recombination-mediated double-strand break repair. BRCA2 is known to directly bind to RAD51, BCCIP, PALB2, and BRAF35 proteins that are involved in meiotic/mitotic recombination, DNA double-strand break (DSB) repair, and chromosome segregation [
1‐
4,
15].
BRCA2 gene expression is stringently regulated during the cell cycle. BRCA2 expression is proportional to the rate of cell proliferation [
14,
16]. While BRCA2 expression is involved in cell cycle checkpoints and DNA repair, the mechanisms of cell cycle-dependent regulation of BRCA2 gene expression remains elusive. Analysis of the minimal promoter sequence of
BRCA2 recognized several conserved binding sites for transcription factors such as E-box, E2F, and Ets recognition motifs [
15,
17‐
20]. USF1 and USF2 bind to the E-box [
18,
19] and Elf1, an Ets family protein, binds to the Ets recognition motifs [
18] to activate
BRCA2 gene expression [
18,
19]. Another transcription factor, NF-κB, has also been shown to bind to the promoter and induce
BRCA2 gene expression [
19]. The tumor suppressor protein TP53 represses the
BRCA2 promoter by blocking the binding of USF [
20]. Recently, poly-(ADP-ribose) polymerase-1 was reported to negatively regulate
BRCA2 gene promoter by binding to it [
15].
We have reported previously that the transcriptional repressor protein SLUG negatively regulates
BRCA2 gene expression in SLUG-positive breast cancer cells by binding to an E2-box flanked by two Alu sequences in the -701 to -921-bp region [
17,
21]. Deletion of this sequence resulted in a 5-7-fold activation of the
BRCA2 promoter. But the mechanism of cell cycle dependent regulation of BRCA2 gene expression in SLUG-negative cells remains unclear.
Here, we provide experimental evidence for the bi-directional activity of human
BRCA2 gene promoter. We have shown here that the reverse activity of this promoter indeed transcribes a protein (ZAR1-like, we named ZAR2) that has significant similarity (36% identity), particularly at the C-terminal amino acid sequence of the C4-type zinc finger containing homeodomain protein, zygote arrest 1 (ZAR1) [
22,
23]. The similarity between ZAR1 and ZAR2 may indicate that these proteins belong to a unique family of transcriptional regulators. The chromosomal context of
BRCA2 and
ZAR2 genes are highly conserved among vertebrates studied.
BRCA2 and
ZAR2 gene expressions are reciprocally related during the cell cycle in human breast cells. Our studies suggest that negative regulation of BRCA2 gene expression by the ZAR2 at the G0/G1 phase of human breast cell growth may provide an additional mechanism of cell cycle-dependent regulation of its expression in both SLUG-positive and SLUG-negative cells.
Discussion
BRCA2 levels go up in many aggressively growing breast cancer cells [
17‐
21]. It appears that the level of BRCA2 protein in the cell must commensurate with the need of the cells to avoid detrimental consequences in the cellular physiology. No BRCA2 in the dividing breast cells will predispose them to non-homologous end joining mode of DNA double strand break repair thus to potential oncogenesis [
34]. Understanding the mechanisms of this stringent mechanism of
BRCA2 gene expression regulation is critical to evaluate etiology of human breast cancer.
Human genome is riddled with bi-directional promoters [
23,
35,
36]. In this study we characterized the bidirectional promoter that expresses
BRCA2 and
ZAR2 genes. Human BRCA1 gene and about 11% of the total other human gene promoters have bi-directional activities [
36]. While assessing the activities of human
BRCA2 gene promoter (Fig.
1A) in both orientations, reverse orientation serving as a negative control, we made three significant observations: (i) The human
BRCA2 gene promoter is active in both the forward and the reverse orientations; (ii) The
BRCA2 gene promoter is more active in the reverse orientation than in the forward orientation when the cells are in the non-dividing stage (G0/G1), and (iii) when the cells are in the dividing state (S/G
2), the forward activity of the promoter is higher than the reverse activity (see below). The reverse activity was insignificant when we did not include the exon 1 and part of the intron 1 sequence of the
BRCA2 gene (26). We have repeated this experiment with different human cell types including human mammary epithelial cells (HMEC), human breast cancer cells like MDA-MB-468, MDA-MB-231, BT549, immortalized human breast cells like MCF10A, MCF10AT, human liver cells HepG2 and human monocytes U937. In all these cells the promoter behaved similarly. Thus, we believe that this cell cycle dependent differential bi-directional promoter activity of the
BRCA2 gene is an intrinsic property of
BRCA2 and
ZAR2 genes. Recently, a ZAR2 paralog, Xzar2, has been cloned from the African clawed frog
Xenopus laevis [
36]. Xzar2 was shown to be involved in epidermal fate determination mainly through signaling pathways distinct from that of BMP-Smad during early embryogenesis [
37].
As mentioned above,
BRCA2 gene expression is tightly regulated in human breast cells [
14,
15,
17‐
21]. The
BRCA2 mRNA and protein are only significantly expressed in the S/G2 phase cells and they are undetectable in the G0/G1 phase cells [
18‐
21]. Over expression of BRCA2 protein was shown to be lethal for the survival of human pancreatic cancer cell line Capan-1 [
38].
Several mechanisms are known to be operative in breast cancer cells to regulate BRCA2 gene expression [
15,
18‐
21]. We reported previously that cell cycle stage-dependent regulation of
BRCA2 gene expression in SLUG-positive breast cells occurs through a distal E2-box/Alu repeat containing silencer element located upstream of the
BRCA2 gene transcription start site [
17,
21]. The zinc-finger transcriptional repressor, SLUG, binds to the uniquely located E2-box sequence in the silencer element in the non-dividing cells and blocks the expression of
BRCA2 gene by chromatin remodeling [
21]. We recently found that peroxiredoxin 5 competes with SLUG for the binding to the
BRCA2 gene silencer in the dividing cells and thus de-silences the expression of
BRCA2 gene in the dividing human breast cells (Misra, S. and Chaudhuri, G., unpublished). Transcription factors other than SLUG that have been reported to regulate human
BRCA2 gene expression include USF1 and 2 [
18‐
20], P53 [
20], NFkB [
19], ElF1 [
18], and PARP1 [
15]. A recent report indicated the presence of a SNP (G to A) at the -26 position of human
BRCA2 gene [
39]. This SNP is in the exon 1 of ZAR2 gene (Fig.
2B). Whether TP53 also regulates
ZAR2 gene expression and whether this SNP affects its promoter activity is yet to be determined.
The bi-directional promoter of BRCA2/ZAR2 gene produces two partially overlapping transcripts. Whether these RNAs hybridize with each other and form double-stranded (ds) RNA and whether this ds-RNA has any role in regulating the activities of the promoter is yet to be determined. One of the potential roles of the ds-RNA could be siRNA-mediated transcriptional gene silencing through DNA methylation [
40,
41].
The biological function of ZAR2 protein is not known. It has two putative C4-type zinc fingers and potentially could be a transcription factor. We found BRCA2 and ZAR2 gene expressions have inverse relationships during the cell cycle. It is possible that ZAR2 protein somehow inhibits BRCA2 gene expression. Although ZAR2 protein has two putative NLS sequences, in the dividing stage of the human breast cells ZAR2 is trapped predominantly in the cytoplasm. Thus, ZAR2 in the dividing breast cells may not have any significant effect on the BRCA2 gene expression. At the non-dividing (G0/G1) phase ZAR2 protein predominantly accumulates in the cell nucleus, binds to the BRCA2/ZAR2 gene promoter and consequently, both ZAR2 and BRCA2 gene expressions are inhibited. While it is tempting to speculate that ZAR2 represents a mechanism of cell cycle dependent regulation of BRCA2 gene expression, direct involvement of ZAR2 in BRCA2 gene transcription is yet to be determined.
As ZAR2 over expression decreased the levels of BRCA2 in the cells, this gene, if disregulated, and over expressed in the cells, it may promote the growth of the tumor. On the other hand, ZAR2 may be needed to suppress BRCA2 expression in the quiescent cells. Expression of BRCA2 in these cells could be detrimental for the cell growth and survival [
37]. We made an interesting observation while knocking down ZAR2 mRNA levels in different breast cancer cells. Out of four cell lines tested (MCF7, MDA-MB-231, MDA-MB-468 and BT549), only BT549 died at the G0/G1 phase in the ZAR2 knocked down cells. We found that ZAR2 knockdown in the quiescent cells leads to the elevation of the levels of BRCA2 which should be detrimental to the cells [
38]. But the ability to suppress the growth of the cells by BRCA2, the cell may need to have high MAGE-D1 level [
42]. Our explanation for the essentiality of ZAR2 in the BT549 cells is that only these cells among the four cells tested have high levels of MAGE-D1 [
42]. ZAR2 protein thus may have multiple balancing roles in the biology of BRCA2 and perhaps other molecules in the cells.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SM, SS, AA, SVK, MKT, and MKM participated in the acquisition of data. SM and GC were involved with the study concept and design. SM and GC contributed to the statistical analyses. SM, MKM, and GC participated in manuscript preparation. All authors participated in the interpretation of results and critical revision of the manuscript for important intellectual content. All authors read and approved the final manuscript.