Glyceollins trigger anti-proliferative effects through estradiol-dependent and independent pathways in breast cancer cells
- 643 Downloads
Estrogen receptors (ER) α and β are found in both women and men in many tissues, where they have different functions, including having roles in cell proliferation and differentiation of the reproductive tract. In addition to estradiol (E2), a natural hormone, numerous compounds are able to bind ERs and modulate their activities. Among these compounds, phytoestrogens such as isoflavones, which are found in plants, are promising therapeutics for several pathologies. Glyceollins are second metabolites of isoflavones that are mainly produced in soybean in response to an elicitor. They have potentially therapeutic actions in breast cancer by reducing the proliferation of cancer cells. However, the molecular mechanisms driving these effects remain elusive.
First, to determine the proliferative or anti-proliferative effects of glyceollins, in vivo and in vitro approaches were used. The length of epithelial duct in mammary gland as well as uterotrophy after treatment by E2 and glyceollins and their effect on proliferation of different breast cell line were assessed. Secondly, the ability of glyceollin to activate ER was assessed by luciferase assay. Finally, to unravel molecular mechanisms involved by glyceollins, transcriptomic analysis was performed on MCF-7 breast cancer cells.
In this study, we show that synthetic versions of glyceollin I and II exert anti-proliferative effects in vivo in mouse mammary glands and in vitro in different ER-positive and ER-negative breast cell lines. Using transcriptomic analysis, we produce for the first time an integrated view of gene regulation in response to glyceollins and reveal that these phytochemicals act through at least two major pathways. One pathway involving FOXM1 and ERα is directly linked to proliferation. The other involves the HIF family and reveals that stress is a potential factor in the anti-proliferative effects of glyceollins due to its role in increasing the expression of REDD1, an mTORC1 inhibitor.
Overall, our study clearly shows that glyceollins exert anti-proliferative effects by reducing the expression of genes encoding cell cycle and mitosis-associated factors and biomarkers overexpressed in cancers and by increasing the expression of growth arrest-related genes. These results reinforce the therapeutic potential of glyceollins for breast cancer.
KeywordsGlyceollins Estrogen receptors Cell proliferation Transcriptomic Gene expression Breast cancer
Estrogen responsive element
Linear model for microarray data
Selective estrogen receptor modulator
Breast cancer is the most prevalent cancer in women worldwide and has an incidence of 89.7 per 100,000 women in Western Europe. The WHO estimates that breast cancer was responsible for over 500,000 deaths in 2011 . Among the different types of breast cancer, the most common is estrogen receptor (ER)-positive cancer, which represents approximately 80% of diagnosed cases of cancer. The ER belongs to the nuclear receptor superfamily and is divided into two subtypes, ERα and ERβ. The ER acts in cells by directly binding to DNA on a responsive element called the estrogen-responsive element (ERE) or by interacting with other transcription factors, such as stimulating protein 1 (Sp1) or activator protein 1 (AP1), which are already bound to responsive elements in promoter regions. The ER also modulates signaling pathways such as the MAPK and PI3K/AKT pathways . Thus, ERα, the major isoform in breast tissue, plays an essential role in normal mammary gland development and function as well as in breast cancer initiation and growth. ERs are bound by the natural hormone estradiol (E2), which has a pleiotropic effect and is responsible for the proliferation and survival of breast epithelial cells. Therefore, E2 plays an important role in breast cancer growth. However, E2 is also essential for maintaining cell differentiation, which consequently limits metastatic potential. Hence, ERα is a good prognostic marker and a prime target for therapy. Endocrine therapies such as tamoxifen or fulvestrant are effective for ER-positive cancer, but frequent relapses are observed . Currently, the focus is on the discovery of new compounds with selective ER-modulator (SERM) activities. Phytoestrogens appear to be promising candidates and have been well studied . Phytoestrogens could be ERα agonist or antagonist depending of cellular type and phenotype studied such as cell proliferation and differentiation . Among this family, pterocarpans, which are second metabolites of isoflavones and the best-known members are glyceollins, have been studied since the 2000s.
Glyceollins are phytoalexins produced mainly by soybeans after elicitation by different types of stressors, such as UV, low temperatures or microorganisms. The glyceollin family includes three compounds: glyceollin I (GI), glyceollin II (GII) and glyceollin III (GIII). In plants, glyceollins are involved in host defense against pathogens such as fungi [6, 7] or nematodes . Glyceollins are promising therapeutic compounds for numerous human pathologies, including breast cancer . Interactions between glyceollins and ERα or ERβ were described for the first time in 2000 . Glyceollins act as antiestrogenic compounds that directly interact with both ER isoforms [11, 12], and they have the capacity to suppress tumorigenesis of breast and ovarian cancer . Glyceollin I is the most potent antiestrogenic molecule, and docking experiments have shown that glyceollin I can interact with the ER in a similar manner to tamoxifen to exert antagonist activity . Thus, chemically synthesized glyceollin I was generated and assessed. It was shown that the natural enantiomer exerted anti-proliferative activities against numerous cell lines, including ER-positive breast cancer cells , and that the compound is a potent inhibitor of ER activation .
However, the precise antiestrogenic mechanisms associated with glyceollins in ER-positive breast cancer remain elusive. In this work, we synthetized natural enantiomers of glyceollin I and II to determine their impact in vivo on the growth of galactophore ducts in mouse mammary glands as well as in vitro in different breast cell lines. The ability of glyceollins to bind and activate ERs as well as their effects on the expression of endogenous E2-dependent genes were assessed. Glyceollins showed surprising effects on gene expression, which led us to perform transcriptomic analysis of the ER-positive breast cell line MCF-7 to better elucidate the mechanisms underlying the actions of these compounds. We found that glyceollins exert their effects through both ER-dependent and ER-independent pathways involving different transcription factors.
Description of randomized mouse groups and their treatments
E2 10 μg/kg
Glyceollin I Low (GI L)
GI 50 mg/kg
Glyceollin II Low (GII L)
GII 50 mg/kg
Estradiol + Glyceollin I Low (E2 + GI L)
E2 10 μg/kg + GI 50 mg/kg
Estradiol + Glyceollin II Low (E2 + GII L)
E2 10 μg/kg + GII 50 mg/kg
Glyceollin I High (GI H)
GI 100 mg/kg
Glyceollin II High (GI H)
GII 100 mg/kg
Estradiol + Glyceollin I High (E2 + GI H)
E2 10 μg/kg + GI 100 mg/kg
Estradiol + Glyceollin II High (E2 + GII H)
E2 10 μg/kg + GII 100 mg/kg
The uterus was carefully dissected at the level of the vaginal fornix, trimmed of fascia and fat, gently blotted on moistened filter paper and weighed.
Mammary whole-mount preparations and immunostaining
The first inguinal mammary fat pads were removed and stained as described by Tian et al. . Briefly, mammary fat pads were spread as flat as possible on a glass surface and fixed with 4% paraformaldehyde. For assessment of epithelial duct length, the mammary glands were stained overnight in carmine alum (0.2% carmine, 0.5% aluminum potassium sulfate), dehydrated in an ethanol gradient and clarified overnight in xylene. Tissues were then photographed under a SteREO Discovery V8 microscope (Zeiss, original magnification, ×1). For Ki-67 and Epcam immunostaining, mammary fat pads were embedded in Tissue Tek mounting medium (Sakura) and sliced with a cryostat. The slices were then incubated at room temperature with a rabbit anti-Ki-67 antibody (Abcam) and a rat anti-Epcam (sc53532, Santa Cruz) for 1 h in PBS supplemented with 0.3% Triton ×100 and 0.5% milk. A dye-conjugated secondary antibody was then incubated with the sections at room temperature for 1 h in PBS supplemented with 0.3% Triton ×100 and 0.5% milk. Images were obtained with an Imager.Z1 ApoTome AxioCam (Zeiss) microscope and processed with Axio Vision Software. The percentage of Ki-67-positive cells was determined by counting the total numbers of ductal epithelial cells and Ki-67-positive cells using ImageJ software.
Cell culture and reagents
MCF-7 cells were maintained in DMEM, 4.5 g/L glucose supplemented with non-essential amino acids (NEAA) (Invitrogen) and 10% fetal bovine serum (FBS) (Biowest). T47D cells were maintained in RPMI 1640 supplemented with NEAA, sodium pyruvate (Invitrogen) and 10% FBS (Biowest). HC-11 cells were maintained in RPMI 1640 supplemented with 2 mM L-glutamine, 5 μg/mL insulin (Invitrogen), 0.01 μg/mL epidermal growth factor (EGF) (Abcys) and 10% FBS (Biowest). MCF10-A cells were maintained in DMEM/F12 supplemented with 0.5 μg/mL hydrocortisone, 10 μg/mL insulin (Invitrogen), 20 ng/mL EGF (Abcys), 100 ng/mL cholera toxin (Sigma) and 5% horse serum (Invitrogen) All cell lines were cultured with penicillin/streptomycin (Invitrogen) at 37 °C under 5% CO2. For steroid treatments, cells were cultured for at least 24 h in steroids and serum-free DMEM without phenol red and with 2.5% or 5% charcoal/dextran-stripping FBS (Biowest) for MCF-7 and T47D cells, respectively. E2 was purchased from Sigma. Natural enantiomers (6aS and 11aS asymmetric carbon configurations) of glyceollin I and glyceollin II were chemically synthesized by HPC Pharma adapting the synthesis method described by Khupse et al.  and Luniwal et al. . The purity was determined at 98% and 99% for glyceollin I and glyceollin II, respectively.
Cells (7500 cells/well for HC-11, 20,000 cells/well for MCF-7 and MCF10-A, and 40,000 cells/well for T47D) were plated in 24-well plates and then deprived of steroids and serum for 72 h. The cells were treated with different doses of glyceollin I or II with or without 10−9 M E2 for 6 days with renewal of the treatment mixture on day 3. After treatment, the cells were trypsinized, and the cell number was determined using a TC10 Automated Cell Counter (Bio-Rad).
MCF-7 cells (30,000 cells/well) were plated in 24-well plates. After serum and steroid deprivation, the cells were transfected overnight with 100 ng of an ERE-TK-luciferase vector, which encodes luciferase under the control of one ERE, and with 20 ng of a CMV-β galactosidase vector, which served as a control of transfection efficiency control. JetPEI was used as a transfection reagent (Polyplus transfection). Next, the cells were treated with 10−9 M E2 and/or with different doses of glyceollin I or II. ICI182.780 (Tocris) was used as ER-inhibitor. The cells were lysed in Passive Lysis Buffer (Promega), and luciferase activity was determined using a commercial luciferase assay system (Promega).
RNA extraction and real-time PCR
Gene names and primer sequences used in real-time PCR experiments
Gene name and symbol
Progesterone receptor (PgR)
Growth regulation in breast cancer 1 (GREB1)
Trefoil Factor 1(TFFI/pS2)
Forkhead box M1 (FOXM1)
Estrogen receptor 1 (ESR1/ERα)
Estrogen receptor 2 (ESR2/ERβ)
FBJ murine osteosarcoma viral oncogene homolog (FOS)
Peroxisome proliferator-activated receptor gamma (PPARG)
Hypoxia inducible factor 1, alpha subunit (HIF1A)
Endothelial PAS domain protein 1 (EPAS1/HIF2α)
Vascular endothelial growth factor A (VEGFA)
DNA-damage-inducible transcript 4 (DDIT4/REDD1)
Nuclear Receptor Subfamily 2 Group F member 1 (NR2F1/COUP-TFI)
Chemokine (C-X-C motif) receptor 4 (CXCR4)
Atypical chemokine receptor 3 (ACKR3/CXCR7)
Chemokine (C-X-C motif) ligand 12 (CXCL12)
Glyceraldehyde-3-phosphate dehydrogenase (GAPDH)
TATA box binding protein (TBP)
MCF-7 cells (250,000 cells/well) were plated in 6-well plates. After 30 h of serum and steroid deprivation, the cells were treated for 24 h with solvent as a control, with 10−9 M E2, with 10−5 M glyceollin I or II, or with an E2 and glyceollin I or II co-treatment. Total RNA was prepared as described above. RNA quantity and purity were determined using a Nanodrop (Thermo Fisher). Only RNA samples with 260/280 and 260/230 ratios >1.8 were selected. RNA quality was analyzed with a bioanalyzer (Agilent), and RNA samples with a RIN > 8.5 and an 18S/28S ratio > 1.7 were selected for spotting on a SurePrint G3 Human Gene Expression v2 8x60K Microarray (Agilent Technologies). Total RNA was reverse-transcribed and labeled according to the manufacturer’s instructions. All samples were prepared and spotted in quadruplicate. Sample hybridization, microarray scanning and results extraction were performed by the GeT-Biopuces Platform in Toulouse, France.
Microarray data analysis and gene filtration
Data analysis was performed using the AMEN suite of tools . Briefly, probes showing a signal higher than a given background cutoff (median of the normalized dataset, cutoff 5.48) and at least a 2-fold change in at least one pairwise comparison were selected. To define a set of 1852 transcripts displaying significant statistical changes across comparisons, the LIMMA (linear models for microarray data) package was used (F-value adjusted with the false discovery rate method, p ≤ 0.05) . The resulting probes were then partitioned into eight expression patterns (termed P1-P8) using the k-means algorithm.
Functional data mining
The enrichment analysis module implemented in AMEN  was employed to identify human diseases, biological processes, molecular pathways and subcellular components significantly over-represented in each expression pattern by calculating Fisher’s exact probability using the Gaussian hypergeometric function (FDR-adjusted p-value ≤0.01, number of probes in a given group associated with a given annotation term ≥5).
Regulatory network analysis
Protein-gene regulation data were downloaded from the Transcription Factor Encyclopedia database . A network representation showing all known protein-gene interactions between transcripts differentially expressed in the current project was drawn using AMEN software.
Mann-Whitney tests were performed using the BiostaTGV website (http://marne.u707.jussieu.fr/biostatgv/), and significant p-values were adjusted with Bonferroni correction.
Glyceollins have anti-proliferative properties
We next tested the in vitro effects of glyceollin I and II on the growth of ER-positive breast cell lines (MCF-7, T47D and HC-11) and an ER-negative nonmalignant breast cell line (MCF10-A). The cells were treated with vehicle and different doses of glyceollin I and II either alone or in combination with E2 for six days. As the proliferation of ER-positive cells is controlled by estrogens, the MCF-7 and T47D cell lines were also treated with 10−9 M E2, which respectively led to 5- and 3.5-fold increases in cell number (Fig. 1c and d). Interestingly, when cells were treated with E2 in combination with glyceollin I or II, a significant anti-proliferative effect was observed for a doses of 10−6 M glyceollin I and 5 × 10−6 M glyceollin II compared to the E2-treated cells. Thus, glyceollin I showed stronger anti-proliferative effects than glyceollin II. However, globally, glyceollin treatment alone did not augment cell number, except for glyceollin I at a dose of 10−6 M, which showed a very low proliferative effect in MCF-7 cells. It should also be noted that both glyceollins showed a weak but significant anti-proliferative effect at a dose of 10−5 M compared to the vehicle-treated control cells. E2 did not have any effect on the proliferation of the mouse mammary gland cell line HC-11 (data not shown). However, the glyceollins showed a strong anti-proliferative effect that was statistically significant starting from 5 × 10−6 M (Fig. 1e). Unlike the other ER-positive cell lines used above, glyceollin I and II showed the same dose effect. Because HC-11 cells require EGF and insulin for growth, it is possible that the glyceollins affected the signaling pathways involved by these two growth factors. To assess the precise role of the ER in this phenotype, ER-negative MCF10-A breast cancer cells were treated with different doses of glyceollin I and II for six days (Fig. 1f). A significant decrease in proliferation was observed with both glyceollins at 10−5 M compared to vehicle-treated control cells. However, the decrease was less evident in these cells compared to the MCF-7 or T47D cell lines. Altogether, these data show that the anti-proliferative effects of glyceollins are primarily produced by the ER but also occur through other ER-independent pathways. To establish whether the glyceollins are cytostatic or cytotoxic, cell cycle and apoptosis were assessed in MCF-7 cells after treatment with 10−5 M glyceollins I or II either alone or in combination with E2 (Additional files 1: Figure S1b and c, and 2 respectively). Cell cycle was analyzed by flow cytometry and showed that E2 induced MCF-7 cells to enter the cell cycle by significantly increasing the percentage of cells in S and G2/M phases compared to control cells. In contrast, glyceollins I and II only weakly induced cells to enter S phase and blocked their passage to G2/M phase. In combination with E2, glyceollins I and II reduced cell entry into S phase compared to treatment with E2 alone, but the combination clearly blocked cells from entering G2/M phase. Apoptosis was analyzed by TUNEL assay. As described in numerous previous studies, E2 reduced the percentage of apoptotic cells. In contrast, 10−5 M glyceollin, whether alone or in combination with E2, did not significantly increase the percentage of apoptotic cells. Thus, glyceollins appear to be more cytostatic than cytotoxic.
Glyceollins interact with the ER
Glyceollins modulate E2-related gene expression
Genome-wide analysis of glyceollin’s effects
Genes linked to proliferation and growth arrest were differentially expressed
Estrogens are involved in multiple physiological processes and act on various tissues. In particular, they participate in the development and maintain the function of reproductive organs such as the gonads or the mammary gland through their binding to ERα and ERβ. ERα is the major isoform in the mammary gland and has a role in epithelial duct proliferation and differentiation . E2 also promotes cell survival, and due to its proliferative effect, this hormone has been linked to breast cancer . Many natural and synthetic chemicals in the environment have been reported to exhibit hormonal activity, particularly estrogenic potency . This is the case for the well-known compound bisphenol A, which has well-documented effects on breast cancer . Phytoestrogens, which are estrogenic compounds from plants, are also found in food, particularly in soy, and have been reported to decrease the risk of breast cancer at high doses [32, 33]. Among the phytoestrogens, glyceollins emerged as promising compounds in the 2000s.
A previous study from Burow et al.  reported that glyceollins bind the ER and act as antiestrogenic compounds by inhibiting cell proliferation. However, the detailed molecular mechanisms driving the anti-proliferative actions of these phytochemicals remain elusive and appear to be more complex than those based only on ER interactions . In the present work, we utilized synthetic glyceollins I and II to better delineate the modes of action exhibited by these compounds. We showed that glyceollin I and glyceollin II inhibit the trophic action of E2 in vivo in mouse mammary glands, but not in uteri. Our data differ from those obtained by Salvo et al., who reported that natural glyceollins antagonize the trophic effects of E2 in uteri . One explanation for these different observations regarding uterotrophy is that the referenced study used nude mice that were treated daily with the natural glyceollins for twenty days, whereas our current study treated mice for only three days. Nevertheless, our data clearly show that mammary epithelium growth is not influenced by glyceollins when they are administered alone; however, when administered together with E2, they are capable of inhibiting the stimulatory effect of E2 on ductal epithelium growth. It should be noted that since proliferation is partially blocked by glyceollins during ductal elongation, further in vivo experiments, such as TUNEL assays, would be required to determine other mechanisms of action by which glyceollins may block E2-mediated ductal elongation. However, to the best of our knowledge, this is the first in vivo study showing the anti-proliferative effect of glyceollins on epithelial ductal extension. In this way, glyceollins appear to be selective estrogen receptor modulators (SERM). In accordance with our results, a previous study investigated the effects of various SERM such as raloxifen, on mammary gland development . The authors showed that raloxifen alone induced a slight ductal tree invasion in the fat pad, whereas, in combination with conjugated estrogens, raloxifen had a clear antagonistic effect . Next, we determined the in vitro effects of glyceollins on cell proliferation. Interestingly, we found that glyceollins I and II exert anti-proliferative effects in both ER-positive and ER-negative breast cancer cells in accordance with the study of Rhodes et al. . This suggests that glyceollins do not act as conventional antiestrogens, such as tamoxifen, but rather act through both ER-dependent and ER-independent pathways, although the ER-dependent pathway seems to be predominant. Burow and collaborators reported that glyceollins could antagonize the E2-mediated stimulation of an ERE-luciferase reporter plasmid [11, 12, 15]. Although this was not observed in our study, the use of different ERE sequences and cell lines and differences in the duration of the treatment could account for this discrepancy. For example, in our experiments, an ERE-luciferase reporter plasmid was used that contains only a single ERE sequence upstream of the luciferase gene, whereas Burow and collaborators used a luciferase reporter with two ERE motifs in addition to pre-treating cells with glyceollins before E2 was added. Based on these observations, we suggest that glyceollins may prevent cooperative effects between ER dimers on ERE sequences, which could explain the decreased luciferase activity. Nevertheless, glyceollin I and II inhibited the expression of the endogenous PgR gene induced by E2 by over 50% in MCF-7 cells.
To further explore the molecular mechanisms underlying how glyceollins exert their anti-proliferative effects, we performed transcriptomic analysis of MCF-7 breast cancer cells exposed to glyceollins and created a gene regulatory network of differentially expressed genes. This integrative genomic approach was followed by the quantification of several key genes, which allowed us to identify, for the first time, two major pathways involving the ER and FOXM1 factors and the other including the hypoxia inducible factor (HIF) family (HIF1α and EPAS1/HIF2α). The first hub highlighted in our gene regulatory network is represented by the forkhead transcription factor FOXM1 and the ER. FOXM1 is a well-known key regulator of the cell cycle and is involved in G1/S and G2/M transition . Thus, the downregulation of this gene could explain the effects of glyceollins on cellular proliferation. FOXM1 gene expression involves ERα , and in return, expression of ERα involves FOXM1 . By targeting ERα, glyceollins could affect this auto-regulatory loop. The mechanisms by which glyceollins act through ERβ are not fully defined. Competition binding assays showed that glyceollins are able to bind both ERα and ERβ with, however, a greater sensitivity of glyceollins for ERα vs. ERβ . In addition, since MCF-7 cells express mainly ERα (the ratio ERα/ERβ is 8/1) , the effects of glyceollins are likely mediated by ERα signaling. Nevertheless, glyceollins may affect ERα/FOXM1 regulatory loop by another pathway that may potentially involve ERβ. Indeed, ERβ represses FOXM1 expression by displacing ERα from the FOXM1 promoter . Our data showed that glyceollins markedly decreased ERα expression but do not affect ERβ expression in MCF-7 cells. A change in the equilibrium of the ERα / ERβ ratio could then contribute to the antiestrogenic activity exerted by the glyceollins and could reinforce the possibility of an involvement of the ERβ. In addition to the auto-regulatory loop that exists between ERα and FOXM1, decreased expression and activity of these two factors could explain the downregulation of GREB1, at least with glyceollin II. Indeed, FOXM1 and ERα co-bind DNA in breast cancer cells and modulate the expression of specific genes . In the referenced work, the authors showed that FOXM1 knockout affected GREB1 expression. Therefore, one could easily hypothesize that glyceollins inhibit E2-related gene expression via this pathway. Moreover, overexpression of FOXM1 is a hallmark of many cancers and a sign of poor prognosis. In ER-positive breast cancer, overexpression of FOXM1 is associated with endocrine resistance and invasiveness because it favors the expansion of stem-like cancer cells . A recent study showed that the FOXM1 cistrome is a powerful index to predict breast cancer outcomes . Thus, it will be very interesting to test the plasticity of the binding interaction that exists between FOXM1 and ERα in response to glyceollin treatment.
The second hub highlighted in our gene regulatory network is centered on the HIF family. The HIF family is composed of three O2-regulated members (HIF1α, EPAS1/HIF2α and HIF3α) that become stabilized under hypoxic conditions. To accomplish this, they heterodimerize with the constitutively expressed HIF1β (also known as ARNT) to regulate genes necessary for adaptation to low-oxygen conditions . It was surprising that the glyceollins in this study induced HIF1α expression under normoxic conditions because they have been previously described as inhibitors of this factor at both the synthesis and stability levels under hypoxic conditions . Under normoxic conditions, HIF1α is controlled by numerous stimuli, including reactive oxygen species (ROS) . Recently, it was shown that glyceollin at a concentration of approximately 18 μM induced ROS generation in a hepatic cell line . In our experiments utilizing 10 μM glyceollin, it is possible that moderate ROS production was induced that consequently induced HIF family activity. We identified DDIT4 (also known as REDD1) and DDIT4 L (also known as REDD2), both inhibitors of mTORC1 , in the HIF family community. REDD1 and REDD2 are stress-responsive genes induced by different stimuli, such as DNA damage or hypoxia. mTORC1 is a member of the PI3K/AKT signaling pathway and acts downstream of AKT. It participates in protein synthesis by promoting the phosphorylation of p70S6K. Therefore, glyceollins alter the phosphorylation of p70S6K in ER-positive breast cancer . Thus, inhibition of mTORC1 could be a factor involved in the anti-proliferative effects produced by glyceollins due to perturbations in the PI3K/AKT/mTOR pathway. PI3K mutations are frequently observed in ER-positive breast cancer. Many inhibitors of PI3K pathway are under clinical trials or approved as therapeutics such as everolimus which is a mTORC1 inhibitor . Thus, this observation reinforces the therapeutic potential of glyceollin in ER-positive breast cancer. Moreover, another study reported that loss of the REDD1 gene leads to an increase in HIF1 level and consequently an increase in tumorigenicity. The authors also showed that REDD1 localizes to mitochondria to regulate ROS production . Overall, REDD1 appears to act as a tumor suppressor that works through different levels, reinforcing the therapeutic potential of glyceollins.
Glyceollins are studied in part for their ability to inhibit E2-related gene expression . Thus, we were surprised to note that, unlike the other genes that were assessed, AREG expression was induced by 10−5 M glyceollins, and this effect was increased by E2 co-treatment. AREG is regulated by numerous transcription factors, including the ER . Recently, a role for EPAS1/HIF2α in the induction of AREG expression was described in MCF-7 cells . Thus, glyceollins might induce AREG expression through ERs and EPAS1/HIF2α, which would explain the synergistic effect. Moreover, high expression of EPAS1/HIF2α, AREG and WISP2 is linked to improved survival in breast cancer ; glyceollin treatment does not affect EPAS1/HIF2α expression in the absence of E2 treatment and even partially restores expression in E2-treated cells. AREG and WISP2 (Fig. 3 and Additional file 4: Table S1) are overexpressed in glyceollin-treated cells.
Finally, our transcriptomic analysis identified differential effects on the expression of the orphan receptor COUP-TFI and the chemokine CXCL12, as well as its receptors CXCR4 and CXCR7, following glyceollin treatment. Considering the important role of COUP-TFI in CXCL12 expression and the importance of the CXCL12 signaling axis in tumor growth and metastasis, the effects produced by glyceollins seem very important. Indeed, the chemokine CXCL12 plays critical roles in cell migration, angiogenesis, proliferation, and survival in many types of cancer, including breast cancer, by interacting with the transmembrane receptors CXCR4 and CXCR7 [51, 52]. CXCR4 is often overexpressed in metastatic tumors and promotes the migration of invasive cells to tissues where local CXCL12 secretion is increased, such as bone, liver, brain and lung [53, 54]. We recently reported that E2 controls the activity of the CXCL12/CXCR4/CXCR7 signaling axis in breast tumor cells and influences the proliferation and migration of breast cancer cells [27, 28]. Furthermore, COUP-TFI and the CXCL12 signaling axis are dysregulated in breast tumor biopsies compared to normal epithelium. Indeed, primarily in ER-positive invasive ductal cancer, we observed significant upregulation of COUP-TFI and CXCR4 and downregulation of CXCR7 and CXCL12, and the levels of these changes showed correlations with tumor grade . Downregulation of CXCL12 in cancer cells is frequently associated with promoter methylation, which encourages cells to migrate toward a CXCL12 gradient and establish metastases . Interestingly, glyceollins repress the expression of CXCR4 and do not affect CXCR7. However, they exert antiestrogenic activity in E2-mediated induction of CXCL12 and they maintain the expression of this gene; thus, they might limit the metastatic potential of tumor cells. In accordance with our observation, a previous study showed that glyceollins could reverse the epithelial to mesenchymal transition of letrozole resistant cells and thus decreased their invasion and migration . It would therefore be interesting to test the ability of glyceollins to limit the loss of expression of key genes in cancer and, in particular, the activation of enzymes involved in epigenetic modifications, as was previously described for genistein .
In conclusion, glyceollins I and II did not show any effect on mouse uterotrophy, whereas they did exert antiproliferative effects on mammary gland epithelial duct growth. This antagonistic activity was confirmed in different ER-positive and ER-negative breast cell lines. Our mechanistic studies revealed that glyceollins are more cytostatic than cytotoxic. Moreover, they have some similarity to SERMs which have partial agonist and antagonist properties depending on E2-target genes. For the first time, a genome-wide microarray was performed on an ER-positive breast cell line to identify pathways involved in the anti-proliferative effects of glyceollins. We identified two major pathways, centered on FOXM1/ERα and HIF1α/HIF2α, which could explain the activity of glyceollins on ER-positive and ER-negative cell lines. These results confirm and reinforce the therapeutic potential of glyceollins for managing breast cancer.
We kindly thank Catherine Martin, Adrien Alusse and Charly Jehanno for their technical assistance with mouse experiments and Rémy Le Guevel for his assistance with TUNEL analysis. We also thank the University of Rennes 1, Inserm and CNRS for supporting our research programs.
Availability and data materials
All data generated or analyzed during this study are included in this published article [and its additional information files].
This work was funded by FUI mVolio, Région Bretagne and Rennes Métropole.
SL conducted, analyzed and interpreted all cellular experiments and was a major contributor in writing the manuscript. FC analyzed the transcriptomic data. FF and FPe conducted the animal experiments. FF, CSa and ML participated to the in vivo data analyses. NP, CSu and TE produced and characterized the glyceollins. FPa supervised and coordinated the study, participated to the conception and design of data and was a major contributor in writing the manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate
All animal procedures were performed according to the guidelines for animal models in research defined by the Ethics Committee and approved by the Ministry of France (reference project number, 2015061812074056_V2). All experiments were conducted by FF and FPe, who are qualified in laboratory animal care and use procedures.
Consent for publication
The authors declare that they have no conflicts of interest.
- 1.WHO, Globocan 2012. Fact Sheets by Cancer. Available from: http://globocan.iarc.fr/Pages/fact_sheets_cancer.aspx
- 5.Lecomte S, Lelong M, Bourgine G, Efstathiou T, Saligaut C, Pakdel F. Assessment of the potential activity of major dietary compounds as selective estrogen receptor modulators in two distinct cell models for proliferation and differentiation. Toxicol Appl Pharmacol. 2017;325:61–70.PubMedGoogle Scholar
- 18.Luniwal A, Khupse R, Reese M, Liu J, El-Dakdouki M, Malik N, et al. Multigram synthesis of Glyceollin I. Org Process Res Dev. 2011;15:1149–62.Google Scholar
- 20.Smyth GK. Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat. Appl. Genet. Mol. Biol. 2004;3:1–25.Google Scholar
- 26.Bergamaschi A, Madak-Erdogan Z, Kim YJ, Choi Y-L, Lu H, Katzenellenbogen BS. The forkhead transcription factor FOXM1 promotes endocrine resistance and invasiveness in estrogen receptor-positive breast cancer by expansion of stem-like cancer cells. Breast Cancer Res. 2014;16:436.PubMedPubMedCentralGoogle Scholar
- 29.Couse JF, Korach KS. Reproductive phenotypes in the estrogen receptor-alpha knockout mouse. Ann Endocrinol. 1999;60:143–8.Google Scholar
- 33.Caan BJ, Natarajan L, Parker B, Gold EB, Thomson C, Newman V, et al. Soy food consumption and breast cancer prognosis. Cancer Epidemiol Biomark Prev. 2011;20:854–8.Google Scholar
- 47.Dey N, De P, Leyland-Jones B. PI3K-AKT-mTOR inhibitors in breast cancers: From tumor cell signaling to clinical trials. Pharmacol. Ther. 2017;175:91–106. [Epub ahead of print].Google Scholar
- 53.Kang H, Watkins G, Parr C, Douglas-Jones A, Mansel RE, Jiang WG. Stromal cell derived factor-1: its influence on invasiveness and migration of breast cancer cells in vitro, and its association with prognosis and survival in human breast cancer. Breast Cancer Res. 2005;7:R402–10.PubMedPubMedCentralGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.