Abstract
Galectins (gal) are multifunctional proteins whose expression changes under different physiological or pathological conditions, including cancer. However, so far, most studies have focused on gal-1 and gal-3, and to a lesser extent to gal-7 and gal-9. We still know very little about other galectins, especially the recently discovered ones, such as gal-14, a prototype galectin highly expressed at the maternal-fetal interface. Here, using in silico and in vitro approaches, we report a correlation between lgals14 expression and ovarian cancer. We found that high expression of gal-14 mRNA in ovarian cancer cells is associated with a shorter survival. Consistent with this observation, we also found that lgals14 is preferentially expressed in high grade serous adenocarcinoma (HGSA) ovarian cancer. Our in vitro data with ovarian cancer cell lines confirmed that lgals14 is readily expressed in HGSA. Interestingly, de novo expression of gal-14 in HEK-293 cells increased apoptosis, both at the basal level and following exposure to low doses of etoposide. Thus, although the study of this galectin is still in its infancy, we were able to provide novel insights into the expression patterns of this galectin and its involvement in cancer.
Introduction
Galectins are small molecular weight soluble proteins with have high affinity for β-galactosides [1]. They were first recognized as lactose-specific lectins, or more specifically β-galactoside-binding lectins, a group of soluble proteins widely found in extracts of vertebrate tissues and whose activity could be inhibited by lactose, galactose or related carbohydrates, such as N-acetylgalactosamine [2–4]. To date, in mammals, a total of 19 different galectins have been identified. They are involved in a large number of cellular functions, most notably in the regulation of cell survival. In vertebrates, galectins are found in a variety of tissues and in almost every cell type. Gal-1, for example, is widely expressed by various cell types, such as vascular endothelial cells, epithelial cells and cells from both primary and secondary lymphoid organs [5–7]. Gal-3, another extensively documented galectin, is expressed in a variety of human tissues and cells, such as in normal human peripheral blood monocytes, activated macrophages, mast cells, T cells, certain epithelial cells and sensory neurons [8–10]. Gal-7 is largely expressed in normal epithelial cells and is often considered a marker of epithelial differentiation [11, 12]. The two tandem-repeat galectins, gal-8 and −9, are expressed in vascular and lymphatic endothelial cells, and in the thymus, respectively [13, 14]. The Charcot-Leyden Crystal protein, or gal-10, is abundantly expressed in the bone marrow, the birthplace of immune cells, and in several subpopulations of immune cells [5, 15, 16]. The expression pattern and role of of other galectins is, however, less well known. We do know, however, that gal-12 is predominantly expressed in adipocytes while gal-13, −14 and −16 are preferentially expressed in the placenta [16–19].
Gal-14, also known as Placental Protein 13-like or PPL 13, is one the less well known galectins. Gal-14 is a prototype galectin with significant homology to gal-13 [17]. Galectin-14 gene is identified as lgals14 and located within a cluster of galectin genes on the q13.2 region of chromosome 19 [16, 17]. Although it has been mostly known for its expression in the placenta, studies with animal models have shown that gal-14 is also expressed in eosinophils from sheep, being released from these cells in response to allergen stimulation, suggesting a role of gal-14 in the regulation of eosinophils activity during allergic responses [20]. Its expression and role in cancer, however, remains largely unknown. In the present work, we provide novel insights into the expression patterns of this galectin and explore its potential role in regulating cell survival in cancer.
Results
In silico analysis of galectin-14 expression in normal and cancer tissues
To obtain a global view of gal-14 expression in normal and cancer tissues, we first interrogated TCGA datasets. As expected, the tissue expressing the highest number of lgals14 transcripts was the placenta as compared with other normal tissues (Figure 1). In cancer tissues, lgals14 was expressed at a higher level in several types of cancer, including liver, breast, liver, uterine and ovarian cancer (Figure 2a). In contrast, levels of gals14 were relatively low in prostate, uveal, and pancreas cancer, indicating that lgals14 expression in cancer is cancer-type specific. Analysis of lgals14 expression in the NCI-60 panel of human cancer cell lines that is commonly used to assess the expression of specific genes in cancer confirmed that lgals14 expression is significantly higher in human ovarian cancer cell lines when compared to other cancer cell lines (Figure 2b).
High expression of galectin-14 is associated with decreased survival of OC patients
Because epithelial ovarian cancer (EOC) is an aggressive type of cancer with limited treatment options, we focused our investigations on EOC, specially high grade serous ovarian adenocarcinoma (HGSA), the most aggressive subtype of EOC. Data mining using the Expression Atlas database, an open source that allows to study expression of specific genes across tissues or cell lines showed that lgals14 was expressed in many EOC cells lines (Figure 3). Interestingly, EOC cell lines expressing the highest level of lgals14 were derived from patients with HGSA. This finding included OVCAR-3, a common cell model to study EOC. An overall survival curve calculated by Kaplan-Meier method showed that high levels of lgals14 were associated with a poorer survival (p = 0.016) (Figure 4). The expression of lgals14 was also significantly higher in EOC tissues as compared to control (normal) tissues (p = 0.03) (Figure 5a). Furthermore, comparison between lgals14 expression in EOC cases that are sensitive and resistant to treatment was very close to the threshold of significance (p = 0.1) (Figure 5b). Whether lgals14 expression is driven by genetic alterations in EOC cell lines is a real possibility as lgals14 had the highest percentage of genetic amplification in EOC when compared to other members of the galectin family, most notably lgals1 and lgals3 (Figure 6). Genetic alterations (mostly amplifications) of lgals14 is indeed frequently observed in serous ovarian carcinoma and correlates with a poorer survival of patients with ovarian cancer (Supplementary figure S1 and S2).
Validation of in silico data
We next carried out PCR and Western Blot analysis to validate the expression of gal-14 at the mRNA and protein levels. We also took the opportunity to look at other less well known galectin genes, including lgals10, lgals12, lgals13, and lgals16. We first confirmed that OVCAR-3 cells readily express high levels of lgals14 (Figure 7a). No expression was found in HL-60 cell line, a common human leukemia cell model. Lgals10 and lgals16 were also expressed in OVCAR-3 cells, albeit at lower levels than lgals14. Expression of lgals14 in EOC cell lines was also observed in TOV-1369 (TR), another HGSA cell line [21]. The TOV-1369(TR) cell line is a pre-chemotherapy cell line derived from a solid tumor (TOV) located in the fallopian tube (Supplementary Table 1). After ovarian cancer diagnosis, this patient received a treatment of paclitaxel and carboplatin for 5 months, two cycles of paclitaxel alone, 11 cycles of doxorubicin and a final treatment with Topotecan [21]. The OV-1369(R2) cell line is a metastatic cell line that was derived from the same patient at the end of the treatments [21]. Low but detectable expression of lgals14 was also found in OV4453, another pre-chemotherapy HGSA cell line derived from ascites (OV) collected from a 70-year-old patient [22]. Both patients had a prior history of breast cancer. No detectable expression was observed in A2780, an ovarian endometroid adenocarcinoma cell line derived from an untreated patient, and SK-OV-3, a cell line derived from the ascites of a 64 year-old patient with an ovarian serous cystadenocarcinoma [23]. As expected, constitutive expression of lgals14 was observed in placenta tissues. It was also found in the AML-14-3D10, an eosinophilic cell line derived from a patient with an acute myelogenous leukemia (AML) expressed lgals14, consistent with previous reports showing that lgals14 is expressed in eosinophils, but not in other lymphoid or myeloid cells, including lymphocytes and macrophages [20, 24].
Galectin-14 protein levels
We next examined the expression of gal-14 at the protein level in EOC cell lines. First, we compared two commercially available anti-gal-14 antibodies using placenta and recombinant human gal-14. Our results showed that while both antibodies were capable of detecting gal-14 proteins, one of the antibodies showed low but detectable cross-reactivity with human gal-7 and gal-1 (Figure 7c, d). This antibody is a mouse monoclonal antibody (MAB5744, from R&D Systems) that was developed following immunization with human recombinant gal-14. The other antibody (ab150427, from Abcam) is a rabbit polyclonal antibody generated from an unknown (proprietary) peptide sequence. We thus used the Abcam antibody for the remaining of our experiments. We could not, however, detect gal-14 in protein extracts from ovarian cancer cell lines, nor with the AML-14.3D10 cells (Figure 7e). Our results obtained following transfection of a vector encoding human gal-14 in HEK-293 embryonic kidney cells, which do not normally express lgals14 (Figure 8a), showed that we can detect gal-14 protein form in cellular extracts (Figure 8b). This suggests that while EOC and AML-14.3D10 cells express detectable levels of lgas14, they do not express detectable levels of gal-14 proteins.
Expression of gal-14 in HEK-293 cells
Our experiments with HEK-293 cells provided us the opportunity to explore, for the first time, the functional role of gal-14. Given the propensity of galectins to modulate apoptosis, we first compared PARP-1 cleavage in a (empty vector) control and cells transfected with a vector encoding human gal-14. Our results showed that de novo expression of gal-14 significantly (p < 0.05) increased the level of apoptosis in HEK-293 cells (Fig. 8c). It also increased apoptosis induced by low doses of etoposide, a common cancer chemotherapeutic agent (Fig. 8d). Not surprisingly, de novo expression of gal-14 in HEK-293 cells significantly reduced their in vitro growth (Supplementary figure S4).
Discussion
In this study, using in silico and in vitro approaches, we have shed some light on the expression profile of gal-14 and started to explore its potential role in cancer. During our investigation, one type of cancer caught our attention: ovarian cancer. Our findings showed that lgals14 expression is elevated in ovarian cancer cell lines derived from patients with HGSA, the most aggressive subtype of ovarian cancer. In addition, we observed that lgals14 harbors more genomic alterations in patients with ovarian cancer than some of its famous family relatives, such as lgals1 and lgals3. We also found that high expression of lgals14 is associated with a shorter survival of patients with EOC, consistent with its preferential expression in HGSA. Additionally, we found that gal-14 expression in ovarian cancer cell lines is detectable at the mRNA level but not at the protein level. Our findings associated with HEK-293 transfectants suggest that this inability to detect gal-14 at the protein level is most likely due to the fact that these cells express very low levels of gal-14 proteins despite the fact that lgals14 is readily detectable. This would be consistent with the our data showing that high expression of gal-14 increases basal levels of apoptosis. These results thus provide the first indications that gal-14 has a role of in cell death processes. Taken together, although the study of this protein is still in its infancy, we were able to provide novel insights into the expression patterns of this galectin and its involvement in cancer.
Our in silico study of lgals14 expression in normal tissues agrees with previous studies showing that gal-14 is preferentially expressed in the placenta [16, 25]. To this day, the only data available about the expression of galectin-14 in other human normal tissues was published more than 15 years ago by researchers who studied gal-14 mRNA levels in multiple tissues by Northern Blot [17]. Yang’s team could not detect lgals14 in other tissues. Our in silico study suggest, however, that gal-14 is possibly expressed at low levels in some tissues, most notably in adrenal and salivary glands, and possibly in the kidney and colon. Whether these low levels reflect expression in specific subpopulations within these tissues is a real possibility. Further studies using cell lines and tissue microarrays will be needed to address these issues.
Very few studies have examined the role of gal-14 in cancer. Using a gal-14-specific aptamer, Cho and colleagues have shown that gal-14 was expressed, albeit at low levels, in hepatocellular carcinoma cell lines, including SNU761, Huh-7, and SNU3058 [26]. Interestingly, the result of our in silico analysis of 53,622 DNA microarray samples from 214 cancer studies shows an elevated mRNA expression of lgals14 in several cancers, including liver cancer. Yet, one interesting finding was that increased lgals14 expression and mutation frequency of lgals14 were clearly evident in tumors of the female reproductive tract, including breast, uterine and ovarian cancer, including the most malignant subtypes of ovarian cancer [27]. The involvement of galectins in female tract neoplasia has already been documented for some prototype galectins, such as gal-1 and −7. For example, increased expression of galectin-1 is found in uterine, breast, cervical, and ovarian cancers [28–32]. Gal-7 is also expressed at abnormally high levels in breast, ovarian and cervical cancer [12, 33–35]. When it comes the proteins homologous to gal-14, −10 and −13, the data is, again, very limited. The majority of studies have focused on the role of gal-10 in eosinophil-specific processes, and the functions of gal-13 in the maternal-fetal interface and in pregnancy-related events [36–39]. Therefore, our results are the first to give some insights into the expression of gal-14 in cancer, more specifically, in cancers of the female reproductive tract. In fact, our findings suggest that there might be an association between gal-14 and progression of EOC. The fact that lgals14 is preferentially expressed in HGSA is particularly interesting given the limited treatment options for this subtype of cancer. Understanding the molecular mechanisms controling expression of gal-14 will also be needed to determine which pathways controls its expression in specific tumor subtypes. It is interesting to note that the promoter of lgals14 contains consensus sites for both Nf-kB, Sp-1, and c/EBP, all of which have previoulsy been shown to control expression of gal-7 in cancer cells [33, 40].
Our results warrant further investigations on the role of gal-14 in EOC, most notably in its potential role in controlling cell death and the apparent discrepancy between the observations that higher lgals14 expression is associated with lower survival of patients with EOC and the pro-apoptotic role of gal-14 observed in HEK-293 cells. One possibility is that gal-14’s pro-apoptotic function is cell type-specific. Another possibility is that the pro-apoptotic function of gal-14 in HEK-293 cells was due to the abnormally high levels of gal-14 following gene transfer, while EOC cells barely express the gal-14 protein, as we found in OEC cell lines, which keeps them protected from gal-14-induced apoptosis. This would explain why OV-1369(R2) cells, which were established following chemotherapy does not express gal-14 while TOV-1369(TR) (established prior to pre-chemotherapy) does. This would also explain why we were not able to detect gal-14 proteins in OVCAR-3 cells despite the presence of lgals14. This result was somewhat unexpected, given that transcription levels of a given gene often dictates the presence/absence of the corresponding protein. Yet, there are examples of galectins showing this type of molecular behavior. For instance, Satelli and colleagues showed that gal-1 and gal-3 proteins could not be detected in HEK-293 cells even though these cells expressed gal-1 and gal-3 mRNA [41]. Similar observations were also reported with gal-8 [41]. These findings support the existence of some type of translational repression mechanisms in ovarian cancer cells that keep gal-14 from being translated or specific proteolytic activity that effectively eliminates gal-14 as soon as it is synthesized.
It is becoming clear that members of the galectin family play an important role in cancer. Research over the past years has identified key roles for these lectins and their ligands in cancer immuno-editing and metastasis. We still need, however, to better understand the contribution of less well known galectins in cancer. This is particularly important for the development of anti-galectin drugs and for limiting their potential off-target effects. Overall, this preliminary study allowed us not only to provide novel insights into the expression patterns of this galectin in normal and cancer cells, but also to explore its potential role in cancer. It certainly sets the ground for future investigations of gal-14 in cancer, most notably in ovarian cancer. Although the study of gal-14 is still in its infancy, on a long-term basis, our results may contribute to the development of new and valuable anti-cancer strategies.
Material and Methods
In silico analysis
In silico analysis of galectin-14 expression was first conducted using datasets from the Gene Expression Omnibus (GEO) repository of the National Center for Biotechnology Information (NCBI) [42]. We examined microarray datasets obtained from profiling human normal tissues (GDS1096) [43], cancer cell lines (GDS4296) [44] and ovarian cancer cells (GDS3592 and GDS4950) [45, 46]. Additionally, other analysis were conducted using the public RNAseq datasets of large-scale cancer genomics projects obtained from the cBio Cancer Genomics Portal and the gene expression data across biological conditions from Expression Atlas database [47–49]. The putative transcription factor binding sites (TFBS) of galectin-14 gene sequence were identified using ALGGEN PROMO, a virtual laboratory for the study of TFBS in DNA sequences [50]. Analysis of public RNAseq datasets from the Cancer Genome Atlas (TCGA) were also conducted to correlate the expression of galectin-14 mRNA and the overall survival of patients with ovarian cancer.
Cell lines, tissues, and reagents
The A2780 and OVCAR-3 ovarian cancer cell lines were kindly provided by Dr. Eric Asselin (Université du Québec à Trois-Rivières, QC, Canada), and the AML-14.3D10 eosinophil cell line was provided by Dr. Denis Girard (INRS-Institut Armand-Frappier, Laval, QC, Canada). These cells were maintained in RPMI 1640 medium supplemented with 10% [v/v] fetal bovine serum (FBS), 2 mM L-glutamine, 10 mmol/L HEPES buffer, and 1 mmol/L sodium pyruvate. The SK-OV-3 ovarian cancer cell line was grown in McCoy’s 5A medium supplemented with 10% [v/v] FBS, 2mM L-glutamine and 10 mM HEPES buffer at 37°C in a humidified atmosphere containing 5% CO2. The HEK-293 cell line was provided by Dr. Stéphane Lefrançois (INRS-Institut Armand-Frappier, Laval, QC, Canada) and it was maintained in Dulbecco’s modified Eagle complete medium (DMEM) supplemented with 10% [v/v] FBS. The high grade serous (HGS) ovarian carcinoma cell lines were a generous gift from Dr. Anne-Marie Mes-Masson (CRCHUM-Université de Montréal, Montreal, QC, Canada) and they were established from the ascites of a patient (OV-4453) and from a HGS ovarian tumor of a patient before (TOV-1369TR) and after (OV-1369 (R2)) chemotherapy treatment [21, 22]. These cell lines were maintained in ovarian surface epithelium (OSE) medium (Wisent, QC, Canada) supplemented with 10% [v/v] fetal bovine serum. All cell culture products were purchased from Life Technologies (Burlington, ON, Canada). The placenta tissue was provided by Dr. Cathy Vaillancourt (INRS-Institut Armand-Frappier, Laval, QC, Canada).
RNA isolation andRT-PCR analysis
Total cellular and tissue RNA were isolated from cells using the TRIzol reagent (Life Technologies) according to the manufacturer’s instructions. First-strand cDNA was prepared from 2 µg of cellular RNA in a total reaction volume of 20 µL using the reverse transcriptase Omniscript (QIAGEN, Mississauga, ON, Canada). After reverse transcription, the cDNAs were amplified by PCR with genes specific primers (see Table 1) and using the following conditions: 94°C for 3 minutes, followed by 30 to 35 cycles of the following: 94°C for 60 sec, 62°C (GAPDH) and 65°C (lgals14) for 60 sec, and 72°C for 60 sec, followed by a final extension step at 72°C for 10 minutes. 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, SYBR Safe (Life Technologies) DNA gel staining and UV illumination.
Western blot analysis
For whole cell extracts, cells were homogenized and resuspended in radioimmunoprecipitation assay (RIPA) lysis buffer (Thermo Fisher Scientific, Rockford, IL, USA) containing a cocktail of protease inhibitors (Roche, Laval, QC, Canada). For placenta extracts, the tissue was dissected, cut and homogenized in RIPA buffer with protease inhibitor, followed by centrifugation at 13,000 × g for 15 minutes at 4°C to pellet cell debris and collect lysate. Equal amounts of whole-cells and tissue extract (20-60 µg) were separated on 15% SDS-polyacrylamide gel and transferred onto nitrocellulose membranes (Bio-Rad Laboratories, Mississauga, ON, Canada). The membranes were first blocked with 5% milk [w/v] in TBS/0.05% Tween 20 [v/v] for 60 min at room temperature and subsequently blotted overnight at 4º C with primary antibodies (see Table 2). Secondary antibodies consisted of horseradish peroxidase-conjugated donkey anti-rabbit (GE Healthcare, Buckinghamshire, England), donkey anti-goat (R&D Systems) or sheep anti-mouse (GE Healthcare) IgG. Detection of immunoblots was performed by the enhanced chemiluminescence method using ECL detection reagent (GE Healthcare).
Transfection and apoptosis assays
The vector encoding human Gal-14 gene (HG11371-CH) was obtained from Sino Biological (North Wales, PA, USA). The plasmid encoding human Gal-7 gene was obtained through the cloning of galectin-7 cDNA into sr⍺ eukaryotic expression vector using SpeI and BamHI restriction enzymes as previously described (Campion et al. 2014). The controls were generated using the empty PCMV3 and Sr⍺ vectors. For transient transfection, the HEK-293 cells were plated at equal densities 24h prior to transfection and then transfected with the indicated vectors using the Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer’s protocol. After 24h, transfected cells were treated with increasing concentrations of etoposide (0–200 µM) for 24h. Cells were lysed with 50 µL of RIPA buffer containing protease inhibitor at 4°C for 30 min and centrifuged at 13,000 × g for 15 min at 4°C. Equal amounts of protein were then separated and processed by Western Blotting following to the protocol previously described. Nitrocellulose blotting membranes were then probed with a rabbit anti-Poly-(ADP-ribose) polymerase (PARP)-1 polyclonal antibody (1:1000; Epitomics, Burlingame, CA, USA) and a mouse anti-β-actin antibody (1:10000; Sigma-Aldrich, St. Louis, MO, USA), followed by incubation with secondary antibodies and chemiluminescence detection.
Statistics
Statistical significance of the experiments was evaluated using the unpaired Student’s t-test. Results were considered statistically significant at P≤ 0.05.
Financial support
This study was supported by the National Science and Engineering Research Council of Canada (Grant No. 298215-2012). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Acknowledgments
The authors wish to thank Drs. E. Asselin and Denis Girard for providing cell lines. They also thank Ms. Marlène Fortier for her technical help.
Footnotes
Abbreviations: galectin-14 (GAL-14), ovarian cancer (OC)