Identification of biomarkers regulated by rexinoids (LGD1069, LG100268 and Ro25-7386) in human breast cells using Affymetrix microarray

Retinoids possess anti-proliferative properties, which suggests that they possess chemopreventive and therapeutic potential against cancer. In the current study, genes modulated by rexinoids (retinoid X receptor (RXR)-pan agonists, LGD1069 and LG100268; and the RXRα agonist, Ro25-7386) were identified using an Affymetrix microarray in normal and malignant breast cells. It was observed that LGD1069, LG100268 and Ro25-7386 suppressed the growth of breast cells. Secondly, several rexinoid-regulated genes were identified, which are involved in cell death, cell growth/maintenance, signal transduction and response to stimulus. These genes may be associated with the growth-suppressive activity of rexinoids. Therefore, the identified genes may serve as biomarkers and novel molecular targets for the prevention and treatment of breast cancer.


Identification of biomarkers regulated by rexinoids (LGD1069, LG100268 and Ro25-7386) in human breast cells using Affymetrix microarray
mammary tumor development with reduced toxicity compared with RAR-selective retinoids (18)(19)(20). Rexinoids are additionally active in animals with tamoxifen-resistant breast cancer (17,21) and in ATRA-resistant breast cancer cells (22). Thus, rexinoids appear to be promising chemopreventive and therapeutic agents with improved efficiency as compared with RAR-selective ligands. Among the rexinoids, LGD1069 (Bexarotene) was confirmed as a safe and well-tolerated agent in clinical trials of cutaneous T-cell lymphoma, breast cancer and lung cancer (22,23). Thus, we focussed on rexinoids and their cognate receptor, RXR, in breast cells, and aimed to investigate their regulatory activity on the transcription of genes involved in growth suppression. In particular, the present study investigated the RXRα isoform, which has been suggested as a potential therapeutic target in breast cancer cells, due to the observation that overexpression of RXRα sensitized breast cancer cells lines to the antiproliferative effects of RXR-selective ligands (24). In addition, infection with adenoviral RXRα induced nucleoplasmic overexpression of RXRα and resulted in apoptosis with treatment with an RXR ligand in retinoid-resistant MDA-MB-231 cells (25). Thus, in the current study, the growth-suppressive activity of RXR pan agonists (LGD1069 and LG100268) and an RXRα specific ligand (Ro25-7386) were investigated in normal human mammary epithelial cells (HMECs) and four breast cancer cell lines (MCF-7, T47D, MDA-MB-231 and MDA-MB-435) using an MTS assay. Subsequently, the genes regulated by rexinoids that may be involved in their antiproliferative activity were investigated with an Affymetrix microarray.
Cells and culture materials. Human normal mammary epithelial cells (HMECs) were obtained from Lonza Group (San Diego, CA, USA). Cells between passages 10 and 11 were used for experiments and the cells were grown and maintained in mammary epithelial basal medium supplemented with 13 mg/ml bovine pituitary extract, 0.5% serum, 5 µg/ml insulin, 10 ng/ml human recombinant epidermal growth factor, 0.5 mg/ml hydrocortisone, 50 µg/ml gentamicin and 50 µg/ml amphotericin-β (all Clonetics, Lonza Group, San Diego, CA, USA). Cells were maintained in a humidified environment at 37˚C with 5% CO 2 in air.
Four different human breast cancer cell lines (MCF-7, T47D, MDA-MB-231 and MBA-MB-435) purchased from the American Type Culture Collection (Manassas, VA, USA) were grown and maintained in appropriate growth media; minimal essential medium for MCF-7 and RPMI 1640 for T47D, MDA-MB-231 and MBA-MB-435 (Invitrogen Life Technologies, Carlsbad, CA, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Welgene, Daegu, Korea). L-glutamine, penicillin, streptomycin and gentamicin (Life Technologies Korea, LLC, Seoul, Korea) were used at the usual concentrations. For all experiments, breast cancer cells were harvested by trypsinization (0.25% trypsin and 0.02% EDTA; Life Technologies Korea, LLC), seeded and grown in the appropriate media containing 10% FBS in a humidified 95% air 5% CO 2 atmosphere.
Cell growth rate measurements. The CellTiter 96 ® AQ ueous Non-Radioactive Cell Proliferation Assay (Promega Corporation, Madison, WI, USA) was used for the measurement of cell growth rate in breast cancer cells according to the manufacturer's instructions. The CellTiter 96 ® AQ ueous Assay is composed of solutions of a novel tetrazolium compound [3-(4,5-dimethylthiazol-2-yl)-5-(-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt; MTS] and an electron coupling reagent (phenazine methosulfate; PMS). Briefly, HMECs, MCF-7 and T47D (1,000 cells/well) were plated in 96-well plates. Following a 24 h resting period, LGD1069, LG100268 and Ro25-7386 were added into the growth media and cell culture continued for 8-12 days. Each measurement day (every 2 days), MTS (Promega Corporation) was added to the cells (20 µl combined MTS/PMS solution per 100 µl culture medium) and further incubation was conducted for 2 h. MTS is bioreduced by cells into a formazan product that is soluble in tissue culture medium. The absorbance of the formazan at 490 nm was measured directly using an ELISA plate reader (Gemini EM Microplate reader, Versa Max, Fluorescence readers; Molecular Devices, Sunnyvale, CA, USA). Each data point was performed in quadruplicate and the results were presented as the mean absorption (optical density).

RNA target preparation/Affymetrix microarray analysis.
Total RNA was extracted from different breast cells treated with rexinoids using the guanidinium isothiocynate method (TRIzol reagent; Invitrogen Life Technologies) followed by purification using an RNeasy column (Qiagen, Valencia, CA, USA). RNA quality was assessed using the 2100 Bioanalyzer Instrument (Agilent Technologies, Inc., Palo Alto, CA, USA). A total of 10 µg total RNA was processed for use on the microarray using the Affymetrix GeneChip One-Cycle Target Labeling kit (Affymetrix, Inc., Santa Clara, CA, USA) according to the manufacturer's instructions. The resultant biotinylated cRNA was fragmented and then hybridized to the Affymetrix U133 Plus 2.0 GeneChip. The arrays were washed, stained and scanned using the Affymetrix 450 Fluidics Station and GeneChip Scanner 3000 7G (Affymetrix, Inc.) according to the manufacturer's recommendations. Expression values were generated using Microarray Suite software, version 5.0 (Affymetrix, Inc.).
Statistical analysis of microarray data. Background subtraction and normalization using the robust multi-array average algorithm method was performed using GeneSpring GX 11.5 software (Agilent Technologies) for gene expression. Fold change values for genes were calculated as the ratio of the signal values of the experimental group compared with the control group. Alterations in gene expression >2-fold were considered to be statistically significant. Genes of interest were selected by referring to the PathArt program which shows intersection of genes in several signaling pathways.
Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) analysis. Cells were cultured to 80-90% confluence. Total RNA was prepared using the Qiagen RNeasy Mini kit (Qiagen). The RT reaction was performed using 1 µg total RNA which was reverse-transcribed into cDNA using a random hexamer primer (GeneAmp RNA PCR Core kit; Applied Biosystems Life Technologies, Foster City, CA, USA), according to the manufacturer's instructions. cDNA of the 7 selected genes and an internal reference gene (GAPDH) was produced from each sample and was quantified using a fluorescence-based real-time detection method (iCycler; Bio-Rad Laboratories, Inc., Hercules, CA, USA). RT-qPCR analysis was performed using the standard methods recommended by the RT-qPCR kit supplier (SYBR ® Green Dye-Based Gene Expression Detection; Applied Biosystems Life Technologies). Primer sequences used for detection of RXRα-regulated genes are shown in Table I (Cosmo Genetech, Seoul, Korea). For the endogenous control, human GAPDH labeled with VIC TM dye provided by Applied Biosystems Life Technologies was used. The amplification conditions were as follows: 30 sec at 95˚C and 3 min at 95˚C, and 30 sec at 95˚C and 60 sec at 65˚C for 40 cycles, followed by a final extension for 20 min at 72˚C. The ratio between the values obtained provided the relative gene expression levels.
Western blot analysis. Whole cell extracts were prepared using 1X sodium dodecyl sulfate (SDS) laemmlli lysis buffer (125 mM Tris-HCl, pH 6.8; 1% SDS; 2% β-mercaptoethanol). Total cell lysates with equal quantities of protein (30 µg) were subjected to 10% SDS-PAGE and subsequently electrotransferred onto a nitrocellulose membrane (Bio-Rad Laboratories, Inc.). Membranes were blocked with 5% skimmed milk in PBST (phosphate-buffered saline containing 0.1% Tween 20) for 1 h at room temperature, then incubated overnight with primary antibodies in PBST containing 2.5% bovine serum albumin (1:1,000 dilution). Subsequent to washing with PBST, the blot was further incubated for 1 h at room temperature with peroxidase conjugated anti-rabbit or anti-mouse antibodies (Pierce Technology Corporation, Holmdel, NJ, USA) in PBST and then visualized using the enhanced chemiluminescence system (GE Healthcare Life Sciences, Chalfont, UK). Protein expression was normalized using β-actin expression.
Statistical analysis. All experiments were performed in triplicate. Statistical analyses were performed using Microsoft Excel 2007 (Microsoft Corporation, Albuquerque, NM, USA). The data for the MTS assay and RT-qPCR are expressed as the mean ± standard deviation. Student's t-test was used for single variable comparisons, and P<0.05 was considered to indicate a statistically significant difference.

Results
Anti-proliferative activity of rexinoids. In Fig. 1, the structures of LGD1069 and LG100268 are presented. The anti-proliferative effects of rexinoids in normal and malignant breast cells were investigated. It was observed that LGD1069 and LG100268 significantly suppressed cell growth in HMECs at 10 µM; whereas co-treatment with LGD1069 and LG100268 reduced cell growth at 1 and 10 µM suggesting that these two rexinoids possess the capacity to prevent mammary cell growth (Fig. 2). By contrast, LGD1069 weakly (10 µM, P<0.05) inhibited cell growth in MCF-7 cells while the compound strongly and significantly suppressed cell growth in a dose-dependent manner in T47D cells (0.1 µM, P<0.01; 1 and 10 µM, P<0.001) (Fig. 3). Notably, LGD1069 induced mild inhibition (P<0.05) of cell growth in MDA-MB-231 cells at 10 µM while rexinoids did not affect cell growth in MDA-MB-435 cells (Fig. 3). This result indicates that LGD1069 is able to inhibit the growth of ER-negative breast cancer with therapeutic potency.
In addition, Ro25-7386, the RXRα agonist significantly suppressed cell growth in a dose-dependent manner in HMECs. Ro25-7386 strongly reduced T47D cell growth at 1 µM and induced suppression of cell growth in MCF-7 cells at day 8 at 1 µM (Fig. 4). These results suggest that RXRα is Table I. Forward and reverse primers for amplification of targeted genes with reverse transcription-quantitative polymerase chain reaction.

Target gene
Forward primer Reverse primer important in the suppression of growth induced by rexinoids in breast cells.
Expression of RXRα in breast cells. The RXRα level in normal and malignant breast cells was next determined. It was observed that all breast cell lines express RXRα but with different intensities. MCF-7 and T47D expressed higher levels of RXRα (Fig. 5). Notably, the ER-negative breast cancer cell lines, MDA-MB-231 and MDA-MB-435, also expressed RXRα.
Identification of target genes regulated by rexinoids in normal and malignant breast cells by Affymetrix microarray. Finally, the genes regulated by rexinoids in normal (HMECs) and  The relative cell growth rate was measured by an MTS assay after 10 days. The growth rate of the vehicle-treated cells was set to 100% and the relative reduction in cell viability resulting from the treatment with rexinoids was expressed as a percentage of the control. Data are presented as the mean of three independent experiments (error bars denote the standard deviation; * P<0.05, ** P<0.01 and *** P<0.001 vs. control). HMECs, human mammary epithelial cells; DMSO, dimethyl sulfoxide. . The relative cell growth rate was measured by an MTS assay after 10 days. The growth rate of the vehicle-treated cells was set to 100%, and the relative reduction in cell viability resulting from the treatment with rexinoids was expressed as a percentage of the control. Data are presented as the mean of three independent experiments (error bars denote the standard deviation; * P<0.05, ** P<0.01 and *** P<0.001 vs. control). DMSO, dimethyl sulfoxide. In HMECs, 638 genes upregulated and 347 genes downregulated by Ro25-7386 with alterations in fold induction >2-fold were identified. A total of 22 genes were strongly upregulated (>10-fold) and 5 genes were strongly downregulated (>4-fold) in expression levels by Ro25-7386 (Table IIA and B). Among them, several genes were notable, including integrin β4, E-cadherin (CDH1), C-terminal binding protein 1 (CtBP1), integrin α6, paxillin (PAX), BAX, forkhead box O3A (FOXO3A) and signal transducer and activator of transcription 3 (STAT3) (upregulated genes), and collagen type VI α3 and cell division cycle 42 (CDC42) (downregulated genes).
In T47D cells, 16 upregulated genes and 3 downregulated genes modulated by LGD1069 were observed (Table IV), whereas 3 upregulated genes and 5 downregulated genes    were identified to be modulated by LG100268 (Table V) with alterations in fold induction >2-fold. According to the data, several notable genes induced by LGD1069 and LG100268 in T47D cells were identified, including cytochrome P450, dehydrogenase/reductase member 3, metallothionein, neuro-oncological ventral antigen 1 and regulator of G-protein signaling 1 (for LGD1069), and chemokine, glutamate receptor, colon carcinoma-related protein and insulin-like growth factor binding protein 7 (for LG100268). In addition, 3 upregulated genes and 5 downregulated genes by Ro25-7386 were identified with alterations in fold induction >2-fold in T47D cells. Among them, chemokine (upregulated genes), and glutamate receptor, ionotropic kainite 2, colon carcinoma-related protein, insulin-like growth factor binding protein 7 and growth differentiation factor 8 were identified (Table VI). In MDA-MB-231 cells, a total of 335 upregulated genes and 320 downregulated genes modulated by LGD1069 were observed (Table VII); whereas 118 upregulated genes and 432 downregulated genes were modulated by LGD100268 (Table VIII) with alterations in fold induction >2-fold. According to the data, several notable genes were identified, including several types of hypothetical protein, zinc finger homeobox 1b, recombination activating gene 2 and tumor protein D52 (for LGD1069), and zinc finger protein 21, Mdm2, and gonadotropin-releasing hormone 1 (for LG100268).

Confirmation of the alterations of modulation of RXRα target genes of HMECs by RT-qPCR and western blot analysis.
The induction of a total of 7 genes by rexinoid (mRNA levels) was confirmed by RT-qPCR assays. These 7 genes are as follows: Integrin β4, integrin α6, CDH1, PAX, BAX, FOXO3A and    STAT3; and upregulation of these genes by Ro25-7386 was confirmed as demonstrated in Fig. 6. The alterations in fold induction of protein levels of certain genes were confirmed by western blot analysis; thus upregulation of BAX, CDH1,    Fig. 7. Thorough investigation of the notable genes-CDH1, FOXO3A, BAX (HMEC-Ro25-7386), insulin-like growth factor binding protein 7 and growth differentiation factor 8 (T47D-Ro25-7386) and cathepsin S, TGFβ2, basigin, MCL-1 and BCL2L1 (MCF-7-Ro25-7386), may aid in the clarification of how RXRα agonists function to inhibit breast cell growth. Such notable genes are implicated in breast cancer management and are important for the treatment of breast cancer.    The current study may aid in the elucidation of novel preventive/therapeutic targets for breast cancer, and may contribute to the development of novel molecules, which may be able to inhibit breast cancer development.

Discussion
In order to investigate the molecular mechanism by which retinoids suppress breast cancer development, the current study focused upon RXR-specific ligands (rexinoids). These have been reported to suppress breast cancer development with minimal toxicity compared with RAR-specific ligands (21), and it was the RXRα isoform that was specifically focused upon in the present study that serves an important role in tumor suppression. The human RXRα gene spans over 40 kilobases in size and consists of a minimum of 10 exons separated by introns ranging in size from 700 base pairs (intron 3) to >7.8 kb (intron 4) (26). It was observed that all of the cell lines examined expressed RXRα. Notably, ER-negative breast cancer cells, which do not respond to retinoid treatment, such as MDA-MB-231 and MDA-MB-435 also expressed RXRα. This suggests that RXRα is non-functional, losing DNA binding activity or failing to recruit essential co-activators required for the activation of the gene in ER-negative cells. Different and inappropriate sub-localization of the receptor may also explain the unresponsiveness of the cells to retinoid treatment.
LGD1069, LG100268 and Ro25-7386 were observed to suppress the growth of breast cells, including the normal HMECs and ER-positive breast cancer cells (MCF-7 and T47D).
LGD1069 was observed to induce a mild inhibition of MDA-MB-231 cell growth at a dose of 10 µM.
LG100268 did not affect the cell growth as compared with LGD1069 in all four breast cancer cell lines suggesting its weaker activity. This result indicates that LGD1069 may possess the ability to inhibit the growth of ER-negative breast cancer.
The genes of interest were selected by referring to the PathArt program, which demonstrated the association between genes of several signaling pathways (data not shown). The alterations in gene expression were then analyzed using the Affymetrix microarray (human genome U133A 2.0) to determine which genes are associated with the inhibition of cell growth induced by the rexinoids. Among them, several genes were identified that are involved in cell death, cell growth/maintenance, signal transduction and response to stimulus, including E-cadherin, CtBP1, integrin β4, integrin α6, PAX, BAX, FOXO3A, STAT3, collagen type VI α3 and CDC42. It was additionally confirmed that Ro25-7386 upregulates the mRNA expression levels of FOXO3A, E-cadherin, BAX, PAX, STAT3, integrin α6 and integrin β4. In addition, Ro25-7386 was observed to increase the levels of BAX, E-cadherin and integrin α6 but reduce the level of CDC42. These results suggest that RXRa may have a role in the prevention and treatment of breast cancer development.
Further investigation regarding the functions of selected genes may aid in the elucidation of novel preventive/therapeutic targets for breast cancer, and may additionally contribute to the development of novel molecules, which may inhibit breast cancer progression.