The CCAAT/Enhancer-Binding Protein Beta-2 Isoform (CEBPβ-2) Upregulates Galectin-7 Expression in Human Breast Cancer Cells

Galectin-7 is considered a gene under the control of p53. However, elevated expression of galectin-7 has been reported in several forms of cancer harboring an inactive p53 pathway. This is especially true for breast cancer where galectin-7 expression is readily expressed in a high proportion in basal-like breast cancer tissues, conferring cancer cells with increased resistance to cell death and metastatic properties. These observations suggest that other transcription factors are capable of inducing galectin-7 expression. In the present work, we have examined the role of CCAAT/enhancer-binding protein beta (C/EBPβ) in inducing expression of galectin-7. C/EBP proteins have been shown to contribute to breast cancer by upregulating pro-metastatic genes. We paid particular attention to C/EBPβ-2 (also known as LAP2), the most transcriptionally active of the C/EBPβ isoforms. Our results showed that ectopic expression of C/EBPβ-2 in human breast cancer cells was sufficient to induce expression of galectin-7 at both the mRNA and protein levels. In silico analysis further revealed the presence of an established CEBP element in the galectin-7 promoter. Mutation of this binding site abolished the transcriptional activity of the galectin-7 promoter. Chromatin immunoprecipitation analysis confirmed that C/EBPβ-2 binds to the endogenous galectin-7 promoter. Analysis of galectin-7 protein expression in normal epithelia and in breast carcinoma by immunohistochemistry further showed the expression pattern of C/EBPβ closely micmicked that of galectin-7, most notably in mammary myoepithelial cells and basal-like breast cancer where galectin-7 is preferentially expressed. Taken together, our findings suggest that C/EBPβ is an important mediator of galectin-7 gene activation in breast cancer cells and highlight the different transcriptional mechanisms controlling galectin-7 in cancer cells.


Introduction
Galectins constitute a family of lectins defined by shared consensus amino acid sequences and affinity for b-galactosecontaining oligosaccharides [1]. In mammals, the distribution of galectins is tissue-specific and their expression is developmentally regulated [1,2]. They play an important role in several physiological processes, including embryonic development, wound healing, apoptosis, intercellular adhesion, cell migration, and immune response. They are also involved in a number of pathological conditions, including infectious diseases and cancer. Most of our knowledge on galectins has been obtained while studying galectin-1 and galectin-3. In contrast, galectin-7 remains an unfamiliar member of the galectin family. This galectin was initially described as a marker of differentiation of stratified epithelia by Magnaldo and colleagues [3]. Functionally, galectin-7 has been shown to be associated with UVB-induced apoptosis in epidermis since sunburn/apoptotic keratinocytes express abnormally high levels of galectin-7 [4]. In fact, galectin-7 has been considered an apoptosis regulator in many cell systems since its discovery by the group of Bert Vogelstein as one of the 14 transcripts out of 7,202 induced in colorectal cancer cells by the expression of p53, whose major function is to control apoptosis [5].
In sharp contrast to the intuitively expected negative roles played by galectin-7 in tumor development, a study by Lu et al. reported that galectin-7 was overexpressed in tumors induced in an experimental model of chemically induced mammary carcinomas [6]. This provided the first indication that galectin-7 expression could be modulated in breast cancer. We and others have since shown that galectin-7 is also expressed in breast cancer tissues in humans. For example, genomic profiling data reported by Perou et al., who provided a molecular portrait of human breast tumors from 42 individuals, has revealed that galectin-7 was highly expressed in estrogen receptor (ER)-negative tumors [7]. Subsequent microarray studies examining normal breast luminal and myoepithelial cells by Jones and his colleagues identified galectin-7 as a myoepithelial-specific gene [8]. We have since confirmed the specific expression pattern of galectin-7 in breast cancer tissues using tissue microarrays (TMAs) constructed from samples obtained from normal individuals and in patients with breast carcinomas [9]. Using preclinical mouse models, we have further shown that high levels of galectin-7 expression in breast cancer cells increase their ability to metastasize to lungs and bones [9].
While a clear picture of the function of galectin-7 in breast cancer is emerging, very little is known about the molecular mechanisms regulating galectin-7 expression in human breast cancer cells. Among the transcription factors that could possibly regulate galectin-7 is CCAAT/enhancer binding protein beta (C/ EBPb). Members of the C/EBPb family have been shown to contribute to tumor progression by controlling the expression of genes involved in invasion, cellular proliferation, survival and apoptosis [10][11][12][13]. In mammary tissues, this factor has been shown to be critical for normal growth and differentiation of the mammary gland [14][15][16]. It also contributes to malignant conversion of the human breast [12]. Of the three C/EBPb isoforms, a particular attention has been paid to C/EBPb-2 because overexpression of this isoform induces epithelial-mesenchymal transition [17]. A significant increase in C/EBPb is observed in estrogen and progesterone-receptor-negative breast cancer as compared to tumors positive for these receptors. Increased C/EBPb levels also correlate with metastatic breast cancer and a high tumor grade [12]. In the present work, we have examined whether C/EBPb regulates expression of galectin-7 in breast cancer cells.

Cell lines and reagents
Breast cancer cell lines were a generous gift from Dr. P. Siegel (the Goodman Cancer Centre, McGill University, Montreal, Qc) [18]. Immortalized human keratinocytes (HaCaT) were provided by Dr. T. Magnaldo (Université de Nice) [4]. All cell lines were maintained in Dulbecco's modified Eagle's medium. Culture media were supplemented with 10% [v/v] fetal bovine serum, 2 mM L-glutamine, 10 mM HEPES buffer, 1 mM non-essential amino acids and 1 mM Sodium Pyruvate. All cell culture products were purchased from Life Technologies (Burlington, ON, Canada).

Transient transfection and luciferase assay
Vectors encoding human C/EBPa or b (SC303472 and SC319561; Origene, Burlington, ON) and vectors encoding C/ EBPb-2 and C/EBPb-3 (No. 15738 and No. 15737; Addgene, Cambridge, MA) were obtained commercially. The cDNA encoding the human galectin-7 (provided by Dr. T. Magnaldo) was cloned in the sra eukaryotic expression vector (kind gift of Dr. François Denis) using SpeI and BamHI restriction sites. Cells were transfected using the Lipofectamine 2000 reagent (Life Technologies) according to the manufacturer's protocol. For reporter assays, NF-kB or a C/EBP luciferase reporter vectors (Cat. No. 219078 and 240112 respectively; Stratagene, Santa Clara, CA) were used. The (empty) pCMV5 vector was used as a control. Transfection efficiency was measured using the pCMV/b-gal reporter vector (Promega, Madison, WI). Luciferase activity was measured using the Luciferase Assay System protocol (Promega) and a luminometer (Lumat LB 9507, Berthold). b-galactosidase activity was detected using a colorimetric enzyme assay using the Luminescent b-Galactosidase Detection Kit II according to the manufacturer's instructions (Clontech Laboratories, Mountain View, CA). Luciferase expression levels were normalized to the levels of b-galactosidase expression.

Confocal microscopy
Cells were cultured onto glass coverslips to semi-confluency. After 24 h of transfection with vector encoding C/EBPb-2, cells were washed with cold PBS, fixed with 3% (v/v) paraformaldehyde in PBS and permeabilized with 0.1% Triton X-100 in PBS and blocked with 1% (v/v) BSA in PBS (PBA) for 30 min. Cells were first incubated overnight at 4uC with a goat anti-human galectin-7 polyclonal antibody (1:100; R&D Systems, Minneapolis, MN) with PBA. After several washes, cells were incubated with an Alexa Fluor 488-conjugated donkey anti-goat IgG (1:500; Life Technologies) for 1 h at room temperature. Samples were mounted using the ProLong gold antifade reagent together with 49-6-diamino-2-phenylindole (DAPI) (Life Technologies) on glass slides and visualized using a Zeiss LSM780 laser scanning microscope (Carl Zeiss Microimaging, Thornwood, NY).

RNA Isolation and RT-PCR
Total cellular RNA was isolated from cells using the TRIzol reagent (Life Technologies) according to the manufacturer's instructions. First-strand cDNA was prepared from 2 mg of cellular RNA in a total reaction volume of 20 mL using the reverse transcriptase Omniscript (QIAGEN, Mississauga, ON, Canada). After reverse transcription, human galectin-7 (gene ID 3963, sense primer: 59-ACC AAC CCG GTC CCA G -39 and antisense primer: 59-GCG GGC TAA CGC TTT ATT TGC -39) and GAPDH (gene ID 2597, sense primer: 59-CGG AGT CAA CGG ATT TGG TCG TAT-39 and antisense primer: 59-CAG AAG TGG TGG TAC CTC TTC CGA -39) cDNAs were amplified using the following conditions: 94uC for 3 min, followed by 35 cycles of the following: 94uC for 1 min, 60uC for 1 min, and 72uC for 1 min, followed by a final extension step at 72uC for 10 min. PCR was performed in a thermal cycler (MJ Research, Watertown, MA). The amplified products were analyzed by electrophoresis using 1.5% (w/v) agarose gels and SYBR Safe DNA gel (Invitrogen) staining and UV illumination.

Chromatin immunoprecipitation (ChIP)
ChIP assays were performed using the EZ-Chromatin-Immunoprecipitation Assay Kit (Millipore, Billerica, MA). For this purpose, MCF-7 cells were grown overnight in 100-mm dishes to ,60-70% confluency and then transfected with either the empty pCMV5 control vector or vectors encoding C/EBPb-2 or C/ EBPb-3 using Lipofectamine 2000 reagent. Cells were fixed in 1% (v/v) paraformaldehyde. Nuclei were isolated, sonicated, and precleaned with protein G agarose/salmon sperm DNA. The precleaned chromatin solution was either set aside as input DNA or incubated with anti-NF-kB/p50 (Cat. No. 06-886; Millipore) or anti-C/EBPb (C-19, sc-150; Santa-Cruz Biotechnology) on a rotation platform at 4uC overnight. Input DNA and mouse IgGpulled DNA were used as controls for all the experiments. After reversal of the cross-linking, DNA was purified from the immune complex and amplified using PCR primers specific for the galectin-7 promoter region encompassing +1 to 2200 bp: sense: 59-CCT GGG TGA TGG GGG GAT CAGG -39 and antisense: 59-CCA TGT CCG TGA GTG CTC CAG GG-39. The samples were incubated for 3 min at 94uC, followed by 40 cycles as follows: 1 min at 94uC, 1 min at 62uC, and 1 min at 72uC, with a final extension at 72uC for 10 min.

Scratch wound healing assay
Confluent monolayers were obtained by seeding 3610 5 cells onto 6-well glass bottom culture plates (MatTek Corporation, Ashland, MA). Cells were transfected with vectors encoding specific cDNA using the Lipofectamine 2000 reagent (Life Technologies) according to the manufacturer's protocol. After 24 h, a scratch with a pipet tip was made in the cell monolayer, followed by washing with PBS to remove cell debris. The plates were moved to an incubator and migration was visualized with a Carl Zeiss LSM780 confocal microscope (Carl Zeiss). Images were captured every 10 min for 2 h. For each cell type, the movement of 30 or 60 different cells was measured. Cell movement was analyzed using the Image J plugins manual tracking and chemotaxis tool.

Statistical analysis
Statistical significance was carried out using an unpaired Student's t-test. Results were considered statistically significant at P#0.05.

C/EBPb-2 induces galectin-7 expression in breast cancer cell lines
We first set out to determine if the C/EBPb-2 isoform could induce galectin-7 gene expression in human breast cancer cell lines. Using semi-quantitative RT-PCR, we found that galectin-7 mRNA increased in all cell lines tested following transfection of a vector encoding the C/EBPb-2 isoform ( Figure 1A). This expression by C/EBPb-2 was independent of the p53 status of the cells. It was observed in cells expressing a wild-type form of p53, such as MCF-7, or in cells harboring an inactive p53 pathway, such as the p53 null MDA-MB-453 or MDA-MB-231, which express the CEBPb-2 Regulates Galectin-7 PLOS ONE | www.plosone.org p53 R280K mutant protein. It was also found in HaCaT cells, which express the transcriptionally inactive TP53 H179Y/R282W alleles. In all cases, de novo expression of galectin-7 was not induced following transfection of a vector encoding C/EBPb-3 (also known as LIP). This increased expression of galectin-7 by C/EBPb-2 was also dose-dependent ( Figure 1B). C/EBPb-3, which lacks the Nterminal transactivation domains but represses the transcription activity of other C/EBPs by competing for C/EBP consensus binding sites or by forming inactive heterodimers with other C/ EBPs [12,19], did not repress the constitutive expression of galectin-7 in MDA-MB-468 or HaCaT cells, nor did it repress C/EBPb-2-induced galectin-7 ( Figure 1C). The ability of C/EBPb-2 to induce galectin-7 was confirmed at the protein level by confocal microscopy in both MDA-MB-231 and MCF-7 cells (Figure 2). Taken together, these results indicate that increased expression of C/EBPb-2 is sufficient to induce galectin-7 in breast cancer cell lines and possibly other types of cells.

A C/EBPb consensus site in human galectin-7 promoter
In silico computational analysis of the human galectin-7 promoter region (using the TFSEARCH program) revealed eight potential C/EBP binding sites within the proximal 1.5 kb 59 flanking region CEBPb-2 Regulates Galectin-7 PLOS ONE | www.plosone.org of the galectin-7 gene. Using progressive deletion reporter constructs of the 59 flanking region of the galectin-7 promoter, we found that deletion of the distal C/EBP binding sites did not modulate the transcriptional activity of the promoter as compared to the transcriptional activity of a reporter construct containing the two promoter-proximal C/EBP binding sites located at positions 2105/293 bp and 2147/2132 bp ( Figure 3A). Mutational analysis showed that mutation (D-103-98) of the C/EBPb binding site located at 2105/293 bp had minimal effect on the transcriptional activity of galectin-7 promoter compared to the wild-type promoter. On the other hand, disruption of C/EBPb binding site located at position 2147/2132 bp (D-145-140) resulted in a 2.5-fold decrease in the promoter activity ( Figure 3B). A 3.5-fold decrease was observed when both C/ EBPb binding sites were disrupted (D-145-140/-103-98). These findings provide additional evidence that C/EBP plays an important role in controlling galectin-7 promoter activity.

Expression of galectin-7 and C\EBPb in epithelial tissues
To further examine whether we could find an association between C/EBPb and galectin-7 expression profiles in epithelial tissues, we looked at the Human Protein Atlas database, which contains high-resolution images showing the spatial distribution of proteins in normal and cancer tissues. We paid a particular attention to epithelial tissues known to express galectin-7 constitutively, including skin, esophagus, oral mucosa and cervical tissues. As expected, immunohistochemistry staining of normal epithelial tissues showed a predominantly nuclear pattern of C/ EBPb protein expression while galectin-7 was mostly found in the cytosolic and nuclear compartments ( Figure 4). In all cases, the distribution of galectin-7 expression co-located with that of C/ EBPb. Identical findings were observed in normal mammary tissues where galectin-7 and C/EBPb were both specifically expressed in myoepithelial cells. Using the ONCOMINE public cancer microarray database, we also found that C/EBPb was expressed at significantly higher levels in oestrogen receptor (ER)negative and triple-negative (TN) breast cancer tissues ( Figure 5), a pattern identical to that of galectin-7 in human breast cancer tissues [9].

Discussion
Galectin-7 has generally been considered a gene under the control of p53. There is an increasing number of reports, however, showing that galectin-7 is overly expressed in cancer cells, most notably in as esophageal, lung and buccal squamous cell carcinomas, thyroid carcinomas, bladder cancer, lymphoma, and breast cancer [7,[20][21][22][23][24][25]. This is somewhat paradoxical since p53 is often inhibited by mutations within its DNA binding domain that lead to the expression of a transcriptionally inactive p53. Here, we have provided a possible explanation for this paradox. More specifically, we found that increased expression of C/EBPb-2 is CEBPb-2 Regulates Galectin-7 PLOS ONE | www.plosone.org sufficient to upregulate galectin-7 expression at both mRNA and protein levels in breast cancer cells. Such a role for C/EBPb was specific to C/EBPb-2 since C/EBPb-3, another isoform that often acts as a dominant negative factor against C/EBPb-2, could not induce galectin-7 nor did it repress the constitutive expression of galectin-7 or the C/EBPb-2-induced galectin-7 expression. The role of C/EBPb-2 as an important transcription factor for galectin-7 is further supported by our mutational analysis showing that mutation in the C/EBP binding motif located at position 2 147/2132 bp of the galectin-7 promoter region strongly inhibits its transcriptional activity. It is also supported by data found in public databases, which showed that C/EBPb and galectin-7 have an almost identical distribution pattern in both normal and cancer tissues. Taken together, these results identify a novel regulatory pathway that regulates galectin-7 expression in human breast cancer cells.
Close examination of a number of public databases supports the view that C/EBP is a positive regulator of galectin-7 in breast cancer cells. For example, a significant increase in C/EBPb mRNA has been observed in estrogen and progesterone-receptornegative breast cancer versus those tumors positive for these receptors [26][27][28]. An increase of C/EBPb mRNA also correlates with metastatic breast cancer and high tumor grade, making it an interesting biomarker for subsets of tumors with a poor prognosis [28][29][30]. Moreover, our preliminary results have shown that transfection of C/EBPb-2 in MCF-7 cells increases their motility ( Figure S1). Similar results were obtained when we increased galectin-7 expression ( Figure S2), consistent with our previous findings that high levels of galectin-7 increase metastasis of breast cancer cells to the bone and the lung [9]. In fact, our results may have implications in other types of cancer where gal-7 is expressed, most notably in transformed keratinocytes. Both gal-7 and C/ EBPb-2 have been shown to be involved in the differentiation of keratinocytes [31]. Whether suppression of C/EBPb-2 will necessarily reduces gal-7 is currently unknown but likely since previous studies showing that retinoic acid, which suppresses the expression of C/EBPb-2 target genes in keratinocytes [32][33], also inhibits gal-7 expression [34]. Interestingly, C/EBPb-2 has been shown to be expressed at high levels in most mammary tumors and is the most frequent isoform found in human breast cancer cell lines [35]. Its expression in MCF10A cells has also been shown to induce in vitro epithelial to mesenchymal transition associated with increased invasive properties [17]. However, while C/EBPb-2 is generally considered the most transcriptionally active of all three C/EBPb isoforms, we cannot completely rule out the implication of other isoforms in regulating gal-7 expression. The C/EBPb-3 isoform, for instance, is expressed at high levels in some breast carcinoma, most notably in basal-like breast cancer [36][37]. Future studies with isoform-specific antibodies will thus be needed to correlate the expression of gal-7 with specific C/EBPb isoforms in breast cancer tissues. It is important to note, however, that other transcription factors can also upregulate gal-7 and may thus compensate for suppression of C/EBPb-2. This is particularly true for breast cancer cells which often express a mutant form of p53, which is capable of inducing gal-7 [38]. Future investigations will thus be needed to determine whether direct suppression of galectin-7 expression or via specific targeting C/EBPb are valuable alternatives to inhibit breast cancer progression.
Our in silico analysis of the human galectin-7 promoter has revealed some other interesting features, most notably within the C/EBP binding site located at position 2147/2132 bp. This site contains an overlapping consensus binding motif for NF-kB. Overlapping binding motifs for transcription factors are frequently found genome-wide in both eukaryotic and prokaryotic cisregulatory regions of a gene promoter. Such overlap often results in competitive binding of transcription factors to the overlapping site [20]. Our preliminary investigations on the relevance of the overlap in the consensus binding motifs for C/EBP and NF-kB at position 2147/2132 bp have shown that a mutation (D-147-145) in the NF-kB recognition sequence of the galectin-7 reporter construct that leaves intact the C/EBP recognition site increased rather than reduced the activity of the promoter ( Figure S3). This would suggest that endogenous NF-kB units, such as transcriptionally inactive p50 complexes bound to this site, hinder binding C/EBP for the same site. This possibility is supported by our results showing that caffeic acid phenethyl ester (CAPE), which prevents NF-kB binding to DNA, increased C/EBPb-2-induced galectin-7 expression ( Figure S3). The ability of C/EBPb to displace NF-kB on the galectin-7 promoter was confirmed by ChIP analysis which showed that while endogenous p50 is constitutively bound to the galectin-7 promoter, transfection of C/EBPb-2 displaced the p50 homodimers, leading to a strong activation of galectin-7 expression. However, in presence of transcriptionally active NF-kB complexes, containing c-Rel for example, NF-kB could possibly exert a positive influence of galectin-7 expression, as we recently showed [38]. In other words, galectin-7 expression in cancer cells with an inactive p53 pathway possibly involves C/EBP and/or NF-kB. These observations also consistent with a model where expression of galectin-7 gene is repressed by transcriptionally inactive p50:p50 homodimers that bind to this overlapping C/ EBPb/NF-kB site on galectin-7 promoter. During malignant transformation, increased expression of C/EBPb-2 or a transcriptionally active NF-kB complex displaces the p50 homodimers, leading to a strong activation of galectin-7 expression. Future investigations will be needed to determine how specific signaling pathways can dictate an NF-kB or C/EBPb response.  Figure S3 Binding of C/EBPb and NF-kB on galectin-7 promoter. (A) Transcriptional activity of the galectin-7 promoter following a mutation (D-147-145) that disrupts the NF-kB binding site located at 2148 bp of a luciferase vector encoding a promoter construct of the human galectin-7 gene. Transfection efficiency was normalized by co-transfection with a b-galactosidase reporter vector. (B) Effect of CAPE on C/EBPb-2-induced galectin-7 expression in MCF-7 cells. Expression was measured by semiquantitative RT-PCR. GAPDH was used as loading control. (C) ChIP analyses showing displacement of the p50 complexes by C/ EBPb-2. Comparative binding of p50 complexes and C/EBPb on the endogenous galectin-7 promoter in control cells MCF-7 and cells transfected with an expression vector encoding C/EBPb-2. Binding to the promoter was measured using anti-p50 and anti-C/ EBPb using genomic DNA. An isotypic antibody (IgG) was used as a negative control. Total DNA extract was used as a positive control (input). (TIFF)