Introduction
We have previously demonstrated that BEX2, a member of
B rain
E xpressed
X-linked gene family, is differentially expressed in breast tumors and BEX2 expression predicts the response to tamoxifen therapy [
1]. Although BEX2 shows a relatively higher expression in 15% of breast cancers, this gene is expressed in the majority of breast tumors and breast cancer cell lines [
1,
2]. The BEX genes were originally found to have a developmental function and a role in the neurological diseases such as accumulation in retinal ganglion cells after optic nerve stroke [
3,
4]. However, recent studies strongly suggest their involvement in cancer biology. For example BEX1 is overexpressed in neuroendrocrine tumors and is down-regulated in glioblastoma cells compared to normal tissue [
5,
6]. BEX3 is shown to be expressed in teratocarcinoma cells, is associated with the mitochondria, and is required for cell cycle entry in these cancer cells [
7]. In addition to our data in breast cancer, BEX2 is found to be differentially expressed in acute myeloid leukemia with a higher expression observed in MLL subtype [
8]. It has been reported that BEX2 is a binding partner of LMO2, a T-cell oncogene with recurrent chromosomal translocations in T-cell acute leukemias [
9], and enhances the transcriptional activity of LMO2-NSCL2 complex [
10]. Furthermore, in AML and glioblastomas BEX2 expression is regulated by epigenetic mechanisms such as promoter methylation [
6,
8]. However, we have not found any correlation between BEX2 expression and promoter methylation in breast tumors or any evidence for gene amplification to explain the differential expression of BEX2 in breast cancer [
1]. These suggest that disturbances in transcriptional regulation may be a mechanism for the observed pattern of BEX2 expression in breast cancer.
Moreover, we have demonstrated that BEX2 has a significant role in promoting cell survival and growth in breast cancer cells [
1,
2]. BEX2 down-regulation induces mitochondrial apoptosis and sensitizes breast cancer cells to pro-apoptotic agents and conversely, BEX2 overexpression protects these cells against mitochondrial apoptosis [
1,
2]. In addition, we have shown that this effect of BEX2 is mediated through the modulation of Bcl-2 protein family, including the regulation of Bcl-2 and BAD phosphorylation [
2]. Furthermore, our data suggest that BEX2 expression is required for the normal cell cycle progression during G1 in breast cancer cells through the regulation of cyclin D1 [
2]. Importantly, we have shown that BEX2 down-regulation results in a higher activity of Protein Phosphatase 2A (PP2A), [
2]. The modulation of PP2A, which is known to regulate several key proteins involved in mitochondrial apoptosis and G1 cell cycle [
11,
12], provides a possible mechanism to explain the BEX2-mediated cellular effects.
In this study we investigate the mechanism of transcriptional regulation of BEX2 and demonstrate that the BEX2 gene is a target of c-Jun and p65/RelA transcription factors. Furthermore, we show that BEX2 is necessary for the phosphorylation of c-Jun/JNK and p65 in breast cancer cells. This study suggests that BEX2 has a functional interplay with c-Jun/JNK and p65, which has significant implications for the biology of breast cancer.
Discussion
We have previously demonstrated that BEX2 has a significant role in promoting cell survival and growth in breast cancer cells [
1,
2]. In this respect, BEX2 expression protects breast cancer cells against mitochondrial apoptosis and is necessary for the normal transition of these cells through G1 cell cycle [
2]. In addition, it has recently been shown that down-regulation of BEX1 and BEX2 sensitize LNT-229 glioma cells to the chimeric tumor suppressor-1 (CST-1), a dominant-positive variant of p53, and up-regulation of BEX1 protects these cells to CST-1-induced cell death [
22]. These findings further support a pro-survival function for BEX1 and BEX2 using a glioma model. Moreover, BEX2 is differentially expressed in breast tumors and is associated with a characteristic gene-expression signature in this disease [
1]. Therefore, understanding the transcriptional regulation of BEX2 is a critical step to advance our knowledge about the function of this gene in the biology of breast cancer.
The available data in different cancers suggest that BEX2 expression can be regulated by a variety of mechanisms. Le Mercier et al. have recently reported that galectin 1, a key player in astroglioma and oligodendroglioma cell migration, has a regulatory effect on BEX2 expression in oligodendroglioma cells [
23]. These authors have demonstrated that down-regulation of galactin 1 in oligodendroglioma cells results in a marked reduction of BEX2 expression [
23]. Furthermore, decreasing BEX2 expression in these cells impairs neoangiogenesis and cell migration [
23]. It is also notable that galactin 1 is up-regulated in breast cancer and has a possible role in tumor-stroma interaction in this disease [
24]. Furthermore, in MLL wild-type AML and glioblastoma BEX2 expression is regulated by epigenetic silencing such as promoter methylation [
6,
8]. However, in MLL mutant AML cells there is a constitutive expression of BEX2 accompanied by promoter hypomethylation [
8]. It is notable that in contrast to these cancer types, we have not found any correlation between BEX2 expression and promoter methylation in breast tumors [
1]. Importantly, as opposed to the down-regulation of BEX2 expression observed in gliobalstoma there is a relative overexpression of this gene in breast tumors, which suggests a difference in the transcriptional regulation of BEX2 between these cancers [
1,
6]. Interestingly, BEX2 has a higher expression in low grade oligodendroglioma compared to glioblastoma and there are differences in the biological function of this gene between these tumor types [
23], which suggest a variation in the transcriptional regulation and function of BEX2 in different brain malignancies.
In order to investigate the transcriptional regulation of BEX2, we first examined the factors involved in the regulation of BEX2 expression in breast cancer cells. We confirmed our previous observation that ceramide treatment has a striking effect on the induction of BEX2 expression and showed that this effect can be almost completely reversed using IкBα phosphorylation inhibitor BAY11 or the overexpression of IκBα-DN (Figure
1A). These findings suggested that transcription factors known to be activated by ceramide signaling and NF-κB activation are potentially involved in the transcriptional regulation of BEX2. Transcription factors c-Jun/AP-1 and AP-2 are known to be activated by the ceramide signaling pathway [
13,
14,
25]. Coordinated induction of ceramide and c-Jun/JNK has an important role in stress-induced apoptosis[
13,
25]. In addition, ceramide induction of intercellular adhesion molecule-1 (ICAM-1) expression requires the activation of AP-2 through a cytochrome c-dependent mitochondrial pathway [
14]. Furthermore, ceramide activates transcription factor NF-κB including both p65/RelA and p50/NF-κB1components of this protein complex [
17,
25]. Moreover, the bioinformatics analysis of BEX2 promoter identified several candidate binding sites for c-Jun/AP-1, NF-κB/p65, and AP-2 transcription factors on BEX2 promoter including six binding sites for c-Jun/AP-1 (Figure
1B). Importantly, we observed a significant induction of BEX2 promoter by 11-fold for c-Jun and by 2.7 to 5-fold for the other transcription factors (Figure
1C), providing strong experimental support for the bioinformatics analysis. In addition to showing a strong effect in the functional transcriptional assay, we also proved that c-Jun and p65/RelA are physically present at the BEX2 promoter with a panel of ChIP assays (Figure
1D and
1E). Moreover, there was a 2-fold increase in the observed enrichment by c-Jun antibody following ceramide treatment of MCF-7 cells (Figure
1F). A similar pattern of increase in enrichment following ceramide treatment has been reported with another c-Jun target gene Beclin1, which is also inducible by ceramide [
26]. These findings demonstrate that BEX2 is a target gene of c-Jun and p65/RelA. Moreover, c-Jun has a clear role in the ceramide-mediated induction of BEX2 expression.
We have also demonstrated that the transcriptional regulation of BEX2 by c-Jun and p65/RelA translated through to BEX2 protein expression and we were able to show that there is a strong correlation between BEX2 and c-Jun expression levels in primary breast tumors. Moreover, we have previously demonstrated that p65-nuclear staining by IF is approximately 2-fold higher in primary breast tumor samples with a relative overexpression of BEX2 [
2]. Overall, these findings demonstrate that BEX2 expression has a positive correlation with the expression of c-Jun and activation of p65 (nuclear) in primary breast tumors. These data using actual breast cancer tissue support our
in vitro findings regarding the transcriptional regulation of BEX2 by c-Jun and p65/RelA. Furthermore, our findings suggest that the relative overexpression of BEX2 in a subset of breast tumors can be explained by a higher expression/activation of c-Jun and p65 transcription factors in this subset.
It has been shown that a number of c-Jun and p65/RelA target genes are involved in mediating the cellular functions of these proteins [
27‐
29]. For example NF-κB induction of Bcl-2 is functionally linked to its pro-survival activity [
28,
29]. In addition, HMG-I/Y is involved in c-Jun mediated anchorage-independent growth and the activation of c-Jun/JNK pathway can mediate Beclin 1 expression, which plays a key role in autophagic cell death in cancer cells [
26,
27]. We were able to detect a similar feedback loop in the BEX2 system. There was a significant induction of p65 nuclear localization following BEX2 overexpression, which was inhibited using IкBα phosphorylation inhibitor BAY11 and BEX2-KD reversed a ceramide-mediated increase in p65 DNA binding. It is notable that the inhibitory effect of BAY11 on p65 activation was not overcome by BEX2 overexpression. This is likely due to the fact that IкBα phosphorylation is a necessary step in p65/NF-κB activation [
30]. Moreover, our findings explain a possible mechanism underlying the observed effect of BEX2 expression on p65 activation, as there was a modest but reproducible reduction in p65 and IкBα phosphorylation following BEX2-KD. Overall, these findings indicate that BEX2 expression is required for the adequate activation and phosphorylation of p65 in an IкBα-dependent fashion. In addition, we observed similar functional effects of BEX2 expression in the regulation of c-Jun with striking reductions in c-Jun phosphorylation following BEX2-KD. This can be explained by our finding of marked reduction in JNK kinase activity following BEX2-KD. Since JNK is a key regulator of c-Jun phosphorylation, a reduction in JNK activity is a likely cause of the observed decrease in c-Jun phosphorylation level following BEX2-KD. Importantly, our data suggest that BEX2 regulates the phosphorylation of c-Jun and p65 at Ser63 and Ser468 sites, respectively. In turn, these phosphorylation sites are required for the effect of c-Jun and p65 in the transcriptional activation and binding to BEX2 promoter region. Taken together, these data show that the BEX2 pathway shares this feedback feature with some of the other c-Jun and p65/RelA target genes.
The functional data presented in this study suggest that BEX2 has a regulatory feedback loop with c-Jun and p65 signaling in breast cancer cells. Moreover, these findings are supported by a strong correlation between BEX2 and c-Jun expression patterns as well as a higher level of p65 activation associated with BEX2 overexpression in breast tumor samples [
2]. Considering the importance of c-Jun and p65/NF-κB pathways in breast tumor development and progression [
31,
32], this feedback mechanism has significant biological implications in breast cancer.
To gain a deeper understanding of the effects of BEX2 expression in c-Jun-mediated cellular functions we investigated the effect of BEX2 on cyclin D1 which is a known c-Jun target of obvious importance in breast cancer. To do this, we generated two stable c-Jun(+) cell lines. These had higher expression of cyclin D1than control lines, and their cyclin D1 levels were markedly reduced by BEX2 knock-down. Cyclin D1 is a c-Jun target gene and is involved in c-Jun-mediated G1 progression [
21,
33]. In addition, we noted a significant decrease in the baseline cell growth and c-Jun-mediated induction of cell proliferation following BEX2-KD (p < 0.01, Figure
5C). These findings suggest that BEX2 expression is necessary for c-Jun-mediated induction of cyclin D1 and cell proliferation in breast cancer cells. Moreover, we have previously reported that BEX2 down-regulation in breast cancer cells leads to a G1 arrest and a significant reduction of cyclin D1 expression [
2]. Considering the data presented here, the observed effects of BEX2 expression on G1 cell cycle and cyclin D1 can be a consequence of BEX2 regulation of c-Jun.
In this study, we demonstrate that BEX2 expression is required for the adequate phosphorylation of p65, IκBα, and c-Jun as well as JNK kinase activity. Importantly, these proteins are known to be directly regulated by PP2A [
12,
34‐
36]. Furthermore, we have recently shown that BEX2 regulates PP2A expression and activity in breast cancer cells [
2]. Moreover, here we found a significant increase in PP2A phosphatase activity following BEX2 down-regulation in c-Jun(+) stable lines (Figure
5D). Overall, these findings provide a possible mechanism for the functional effects of BEX2 expression on p65, IκBα, and c-Jun/JNK through the regulation of PP2A activity.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
AN conceived the study, performed the data analysis, and drafted the manuscript. AN and JL carried out the experiments. LHD contributed with scientific discussion and manuscript preparation. All authors read and approved the final manuscript.