ARTICLE pubs.acs.org/JAFC
Antiproliferative Activity of Pomiferin in Normal (MCF-10A) and Transformed (MCF-7) Breast Epithelial Cells Raymond Yang,†,§,4 Heather Hanwell,†,X Jing Zhang,§ Rong Tsao,§,|| and Kelly Anne Meckling*,†,|| †
Department of Human Health and Nutritional Sciences, University of Guelph, Guelph, Ontario, Canada N1G 2W1 Guelph Food Research Centre, Agriculture and Agri-Food Canada, 93 Stone Road West, Guelph, Ontario, Canada N1G 5C9
§
ABSTRACT: Pomiferin and osajin are prenylated isoflavones from Osage orange fruit that both have potent antioxidant activity in a variety of assays. Pomiferin, in particular, has strong activity against the superoxide anion in a photochemiluminescence (PCL) assay system. In vitro, pomiferin, but not osajin, demonstrated selective antiproliferative activity against the tumorigenic breast epithelial cell line MCF-7 (IC50 = 5.2 μM) with limited toxicity toward nontumorigenic breast epithelial cells (MCF-10A). The differential sensitivity of normal and tumorigenic cells to the antiproliferative action of pomiferin was examined further by using cDNA microarrays. With a stringent cutoff of p < 0.01, a total of 94 genes were significantly differentially expressed between MCF-7 and MCF-10A cells; 80 up-regulated and 14 down-regulated when cells were exposed to 5 μM pomiferin for 24 h. Fold changes by microarray analysis were confirmed using RT-qPCR, and the most significant changes were found with genes related to antioxidant enzymes. Genes involved in mitotic inhibition and apoptotic regulations were also found to be up-regulated. Pomiferin is therefore a good anticancer candidate agent that may be useful either alone or in combination with other therapeutic agents and, because of its selectivity toward tumor cells, likely to have fewer side effects that classic chemotherapy drugs. KEYWORDS: prenylated isoflavones, pomiferin, antiproliferation, antioxidant, MCF-7, MCF-10A, nutrigenomics, gene expression, microarray, RT-qPCR
’ INTRODUCTION Polyphenols are a general class of plant phytochemicals that have a wide number of biological activities and presumed roles in the prevention and treatment of several chronic diseases.1 Cardiovascular disease and cancer are two of the most prevalent diseases in the Western world, and their incidence has been associated with excessive oxidant stress.2 Polyphenols are strong antioxidants that complement and add to the body’s defense system, which includes antioxidant vitamins and enzymes. Superoxide anion, one of the most damaging reactive oxygen species (ROS) in the cell, is neutralized by endogenous superoxide dismutases to the less toxic metabolite hydrogen peroxide, which is further broken down by catalase.3 The largest group of polyphenols comprises the flavonoids, which are further subdivided depending on the structure of the flavan nucleus. Several structureactivity relationships have been established for the flavonoids using in vitro models. Among them, multiple hydroxyl groups in the various rings of the flavan nucleus have been found to confer substantial antioxidant, chelating, and prooxidant activities.4 Methoxylation of the hydroxyl groups on rings A and C introduces unfavorable steric effects and increases the lipophilicity and membrane partitioning potentially affecting in vitro and in vivo activities. Introducing a double bond or carbonyl function in the heterocycle, or polymerization of the nuclear structure, increases activity by affording a more stable flavonoid radical through conjugation and electron delocalization.4,5 Using a microsomal lipid peroxidation (LPO) and Fe2+ chelation model, large differences were found in the chelating capacity of the flavonoids. For example, a catechol moiety on ring B is required, and the 3-OH group in combination with a C2/C3 double bond on ring C increases the scavenging activity.6 Flavonoids with prenyl substituent(s), that is, the r 2011 American Chemical Society
prenylated flavonoids, have been found to possess strong antioxidant activities.7 Pomiferin is one of the major prenylated isoflavones in the fruits of Osage orange (Maclura pomifera (Raf.) Schneid) (Figure 1). Prenyl substitution on ring A and the 30 -OH group (catechol) on ring B were found to be critical for the antioxidant activities of pomiferin, which were overwhelmingly stronger than those of osajin, an analogue of pomiferin that lacks the 30 -OH, and those of genistein and daidzein, two isoflavones that lacked the prenyl groups.8 Strong antiproliferative activity of pomiferin has also been found in a couple of human cancer cell lines.9,10 Whereas antioxidant activity is undoubtedly one of the best studied and most obvious activities of the flavonoids, recent in vitro studies suggest several other potential mechanisms of action. These include regulation of cell signaling pathways involving both nitric oxide and lipids,11 activity as pro- or antiestrogens,12 regulation of calcium- and calmodulin-promoted phosphodiesterase activity,13 and modulation of tumor angiogenesis.14 One approach to identifying potential novel targets of bioactive phytochemicals including the flavonoids is the multidisciplinary field of microarray analysis. It enables simultaneous assessment of the transcription of tens of thousands of genes and of their relative expression between normal cells and diseased cells or before and after exposure to various treatments. To further study the potential health benefits of pomiferin, we evaluated the antioxidant activities of pomiferin and osajin from the Osage orange and compared them with those of genistein, daidzein, formononetin, and biochanin A using the Received: July 19, 2011 Revised: September 27, 2011 Accepted: November 16, 2011 Published: November 16, 2011 13328
dx.doi.org/10.1021/jf202898g | J. Agric. Food Chem. 2011, 59, 13328–13336
Journal of Agricultural and Food Chemistry
Figure 1. Chemical structure of pomiferin.
photochemiluminescence (PCL) method that is based on scavenging the superoxide anion.15 The antiproliferative activities of pomiferin on MCF-7 human breast cancer cells and the immortalized human breast epithelial cells (MCF-10A) were studied, and the molecular mechanism was investigated. cDNA microarray was used to identify the transcription changes after MCF-7 and MCF-10A cells were treated with pomiferin and to analyze the modulatory effect of pomiferin on the expression of antioxidant associated genes as well as genes involved in cell cycle control, apoptosis, and differentiation. The expression profile of a selected subset of the list was confirmed using reverse transcriptase quantitative PCR (RT-qPCR).
’ MATERIALS AND METHODS Chemicals and Media. All culture media were obtained from Invitrogen Inc. (Mississauga, ON, Canada); fetal bovine serum was from Gibco-BRL (Missisauga, ON, Canada); hydrocortisone, EGF, and cholera toxin were from Sigma-Aldrich Co (Mississauga, ON, Canada); and all HPLC grade solvents including ethyl acetate, hexanes, and dimethyl sulfoxide (DMSO) were from Caledon Laboratory Chemicals (Georgetown, ON, Canada). Genistein and daidzein were purchased from Sigma-Aldrich (St. Louis, MO), and formononetin and biochanin A from Indofine Chemicals (Somerville, NJ). Isolation and Purification of Pomiferin and Osajin from Osage Orange. Osage orange fruits (ca. 1 kg) collected locally in
Guelph, ON, Canada (2007), were sliced with a kitchen knife into 1 1 2 cm wedges. To minimize oxidation, fruit was added to an amber bottle, filled completely with ethyl acetate (1:4, w/v) sealed, and extracted for 48 h. The mixture was filtered through a Whatman no. 1 filter paper, and the remaining fruit was rinsed and soaked (repeating the above process) three more times. The combined extract was concentrated in vacuo (e40 °C) until all of the ethyl acetate was gone and the aqueous solution back extracted with 3 equal volumes of ethyl acetate. The combined ethyl acetate layer was dried over anhydrous sodium sulfate, evaporated, and reconstituted in ca. 500 mL hexane and ethyl acetate mixture (1:1, v/v). An aliquot of the sample solution (∼50 mL) was loaded onto a column (5.0 cm i.d. 68 cm) packed with silica gel (mesh size = 70230, Sigma-Aldrich Co.) in hexane and ethyl acetate (1:1, v/v). The same solvent mixture was used as mobile phase and delivered by a preparative pump (LabAlliance, State College, PA) isocratically at 9 mL/min. Fractions (50 mL) were collected manually and analyzed by HPLC as reported previously.8 Fractions with pure pomiferin or osajin were combined separately and dried in vacuo using a rotary evaporator (e40 °C). The identities of pomiferin and osajin were confirmed using mass spectrometry.8 The pure crystalline compounds were stored at 20 °C before use. Antioxidant Activity by the PCL Assay. The principle of PCL was based on photoinduced generation of the superoxide anion radical (O2•). A commercial PCL instrument, the Photochem system (Analytik-Jena, Berlin, Germany), was used in this study, in which luminol was used as a photosensitizer, which generates superoxide radicals, and a chemiluminogenic probe for free radicals.16 Complete reagent kits (ACW) were purchased from the manufacturer. For the
ARTICLE
isoflavones in this study, the assay mixture contained 1 mL of reagent 1 (sample solvent), 1.5 mL of reagent 2 (reaction buffer), 25 μL of diluted reagent 3 (luminol), and 10 μL of reagent 4 (ascorbic acid) for the calibration curve or 10 μL of sample solution for the antioxidant activity. The evaluation of the activity was based on the lag phase in seconds.15 All isoflavone samples were dissolved in 80% methanol, and when necessary, samples were diluted so that the PCL curves fell within the linear range of the standard, ascorbic acid (0.050.3 mM). All results were expressed as ascorbic acid equivalents (AAE, mM). All samples were prepared and run in triplicate. Cell Culture. The MCF-7 human breast epithelial cancer cell line and the spontaneously immortalized human breast epithelial cell line (MCF-10A) were obtained from American Tissue Type Culture Collection (ATCC, Bethesda, MD). MCF-7 cells were cultured in α-modified Eagle’s medium (α-MEM) supplemented with 10% fetal bovine serum (FBS), 1 mM sodium pyruvate, 100 U/mL penicillin, 100 ng/mL streptomycin, and 10 μg/mL insulin. MCF-10A cells were cultured in 50:50 DMEM/HAMS’s F12 nutrient mixture supplemented with 5% horse serum, 10 μg/mL insulin, 0.5 μg/mL hydrocortisone, 20 ng/mL EGF, 100 ng/mL cholera toxin, 100 U/mL penicillin, and 100 ng/mL streptomycin. All cells were grown in a humidified atmosphere at 37 °C, with 5% CO2 in air. The division and morphology of the cells were monitored periodically. Cells were passaged by trypsinization and maintained at subconfluent densities during all experimental procedures. Proliferation Assays. For cell proliferation assays, 5 103 cells (in 200 μL growth media) were loaded in each well of a 96-well roundbottom plate, followed by the addition of various concentrations of pomiferin. All wells had the same final concentration of solvent vehicle (DMSO). After 24 h of incubation at 37 °C, proliferation was estimated using the sulforhodamine B (SRB) dye-binding assay as previously described.17 Results are expressed on a relative proliferation index scale (mean ( SD). All experiments were run at least four times. The IC50 values were calculated using an approximated sigmoidal curve. RNA Extraction. MCF-7 and MCF-10A cells grown in 175 cm2 tissue culture flasks were used for RNA preparation. After reaching 70% confluence, the cultures were treated with pomiferin (5 μM) or vehicle for 24 h. Cells were then released by brief exposure to trypsin/EDTA, and the RNeasy Mini Kit (Qiagen, Mississauga, ON, Canada) was used for RNA isolation according to the manufacturer’s instructions. Purified RNA was stored at 80 °C. Aliquots of the RNA were thawed on ice. The integrity of the total RNA was initially evaluated by running samples on 1.5% agarose gels containing ethidium bromide (10 μL of 1.5% ethidium bromide in 30 mL of gel) with 1 TBE (Tris/Borate/EDTA) running buffer and visualization under UV light. The concentration of RNA was measured using a NanoDrop ND-1000 spectrophotometer (Wilmington, DE) at 280 and 260 nm, and the quality was further confirmed with the RNA 6000 Pico LabChip Kit on an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA) according to the manufacturer’s instructions. Preparation of Microarray Slides. For our study, the human 19K Combo Array kits containing more than 19000 human clones were purchased from the University Health Network (Toronto, ON, Canada). A SuperScript Plus Indirect cDNA Labeling System from Invitrogen Inc. (Mississauga, ON, Canada) was used to prepare the hybridization slides. In brief, 20 μg of total RNA from control and treated cells was used for the first-strand cDNA synthesis. Sixteen microliters of total RNA and 2 μL of anchored oligo(dT)20 primer (2.5 μg/ μL) were incubated at 70 °C for 5 min in an 18 μL volume and chilled on ice. Six microliters of first-strand synthesis buffer was then added (1.5 μL of 0.1 M DTT, 1.5 μL of 10 mM dNTP mixture, 1 μL of RNaseOUT, and 2 μL of SuperScript III RT) to each tube, and the tube was incubated at 46 °C for 3 h. RNA was degraded with NaOH and neutralized with HCl. Next, the PureLink PCR Purification System 13329
dx.doi.org/10.1021/jf202898g |J. Agric. Food Chem. 2011, 59, 13328–13336
Journal of Agricultural and Food Chemistry
ARTICLE
Table 1. Primersa of MCF-7 Genes Used for RT-qPCR gene symbol
forward primer
reverse primer
size of product
SOD (Mn)
GGAAGCCATCAAACGTGACT
CTGATTTGGACAAGCAGCAA
HSPA1A
GCCGAGAAGGACGAGTTTGA
TCCGCTGATGATGGGGTTAC
FTL
GCGTCTCCTGAAGATGCAAA
AGGAAGTGAGTCTCCAGGAAGT
213
GPX3
TTGATGGGGAGGAGTACATCC
AGACCGAATGGTGCAAGCTC
139
ULBP2
GTGGTGGACATACTTACAGAGC
CTGCCCATCGAAACTGAACTG
150
CANX
GCATCATGCCATCTCTGCTA
GATACCCGTTTTGGGGTTTT
271
TOP2A
AGTCATTCCACGAATAACCA
TTCACACCATCTTCTTGAG
108
ID2 MCM7
TGCAGCACCTCATCGACTACA GGAAATATCCCTCGTAGTATCAC
TCTGGTGATGCAGGCTGACA CTGAGAGTAAACCCTGTACC
81 144
HMGB1
CGGGAGGAGCATAAGAAGAAGCACC
CAATGGACAGGCCAGGATGTTCTCC
273
H2AFJ
GGCAAAGTGACCATCGCTCA
GTCAGGGTCATTTGCTCTTC
101
TXNRD1
CTTTTTCATTCCTGCTACTCTACC
CTCTCTCCTTTTCCCTTTTCC
200
PSMA5
GCCCAGCAGCATTGAGAAAAT
CTTGGGTCACACTCTCCACTG
164
BCAP31
GATGCCGTGCGCGAAATTC
AAGCCAGCAATGTAGAGATTCC
131
18s rRNA
AATTGACGGAAGGGCACCAC
CGGACATCTAAGGGCATCACAG
305
202 69
a
The primers of SODs were from Suzuki et al. (2002); HSPA1A from Ishida et al. (2002); FTL from Feng et al. (2005); GPX3, PSMA5, and BCAP31 from the Primer Bank provided by The Massachusetts General Hospital (http://pga.mgh.harvard.edu/primerbank/index.html); CANX from York et al. (2005); TOP2A and MCM7 from Steinau et al. (2007); ID2 from Husson et al. (2002); HMGB1 from Yoshihara et al. (2006); H2AFJ from Yao et al. (2006); TXNRD1 from Reichard et al. (2007); and 18s rRNA from Hammamieh et al. (2007).
(Invitrogen Inc.) with spin columns was used to purify the cDNA. The cDNA was then coupled with fluorescent dye with SuperScript Plus Indirect cDNA Labeling System, and the labeled cDNA was purified with the PCR purification kit from Qiagen Inc. (Mississauga, ON, Canada). Hybridization of Cy3 and Cy5 labeled cDNA probes to the arrays was carried out at 42 °C overnight (18 h). Finally, washes were carried out at high stringency by soaking the slides in a series of solutions containing sodium chloride, sodium citrate (SSC), and sodium dodecyl sulfate (SDS), followed by drying before scanning. All cDNA concentrations and the dye incorporation efficiency were determined with the NanoDrop ND-1000 spectrophotometer. The experiments were run for three independent samples with dye swap assays. Thus, for each cell line, a total of six slides were prepared. Scanning and Data Processing. Hybridized and washed arrays were scanned on an Axon 4000B dual laser scanner (555/647 nm wavelengths). The voltage across the photomultiplier tubes was adjusted until the intensity ratio of the red and green acquisition histograms was between 0.9 and 1.1. The fluorescence intensities for each spot on the array were determined using the GenePix Pro program (version 3.0) from Axon Instruments. Array-specific data normalization was then done using the locally weighted regression and smoothing scatter plots (LOWESS) procedure. All fluorescent signals above 100 units were included in the analysis. The fold changes of the samples between the treatment and control were calculated. The P values of 0.05 and 0.01 were used to identify significant gene expression differences. The data obtained from the GenePix were further analyzed by loading into GeneSpring (version 7.0) software (Agilent Technologies) and the Database for Annotation, Visualization and Integrated Discovery (DAVID) (http://david.abcc.ncifcrf.gov/) for functional annotation. RT-qPCR. Total RNA from control and pomiferin-treated cells was used for carrying out one-step RT-qPCR in 96-well plates using the ABI prism 7000 system (Applied Biosystems Ltd., ON, Canada). The primers for RT-qPCR were chosen from the literature, and their specificity was verified in BLAST domain (http://www.ncbi.nlm.nih. gov/BLAST). These primers were synthesized with the ABI DNA synthesizer 3900 (Applied Biosystems Ltd.), and their sequences are listed in Table 1. A parallel amplification reaction using 18S rRNA was carried out as an internal reference. The Power SYBR Green RT-PCR Reagents Kit containing the Power SYBR Green RT-PCR Master Mix
and TagMan Reverse Transcription Reagents (Applied Biosystems Ltd.) was used following the manufacturer’s protocol. In brief, the reaction mixture (total volume = 50 μL) contained 25 μL of master mix, 2 μL of template, 0.25 μL of reverse transcription reagents, 1.0 μL of RNase inhibitor, 1.5 μL of forward and reverse primers, respectively, and 18.75 μL of RNase-free water. A no-template control (NTC) and a noamplification control (NAC) were run in parallel to assess the overall specificity of the reaction. The real-time cycler conditions were as follows: reverse transcription for 30 min at 48 °C followed by PCR initial activation step at 95 °C for 10 min, 40 cycles each of melting at 95 °C for 15 s, and annealing/extension at 60 °C for 1 min. Data were collected using an ABI Prism 7000 SDS analytical thermal cycler (Applied Biosystems Ltd.). A dissociation curve was generated for each sample to evaluate the specificity of the PCR product. The products were evaluated by using 1.5% agarose gels containing ethidium bromide with 1 TBE running buffer to verify the product sizes. The experiments were run for three independent biological samples. The 2ΔΔCt method was used for calculation of fold changes,18,19 and 18S rRNA was used as an endogenous reference.
’ RESULTS Antioxidant Activities of Prenylated and Nonprenylated Isoflavones. The PCL assay showed that among the two
prenylated isoflavones pomiferin and osajin and the four nonprenylated isoflavones genistein, daidzein, formononetin, and biochanin A, pomiferin alone was effective in scavenging the superoxide anion (Figure 2). The antioxidant activity of pomiferin was 22.1 mmol AAE/mmol pomiferin; that is, pomiferin was >22-fold stronger as an antioxidant compared to ascorbic acid. This is similar to the results of previous studies using other in vitro model systems.8 Antiproliferative Activity of Prenylated Isoflavones. A variety of pure flavonoids, crude Osage extracts, and HPLCpurified pomiferin and osajin were used in vitro to examine antiproliferative activity against nontumorigenic (MCF-10A) and tumorigenic (MCF-7) cells. Using the SRB dye binding assay, pomiferin exhibited selective antiproliferative activity 13330
dx.doi.org/10.1021/jf202898g |J. Agric. Food Chem. 2011, 59, 13328–13336
Journal of Agricultural and Food Chemistry
ARTICLE
Table 2. Number of Genes Regulated by Pomiferin in MCF-7 and MCF-10A Cell Lines (P < 0.05 and P < 0.01) P < 0.05 upregulated MCF-7
Figure 2. Antioxidant activities of isoflavones and L-ascorbic acid measured by the PCL assay. By using one-way ANOVA and followed by Fisher’s least significant difference (LSD) test, pomiferin showed a much greater activity than L-ascorbic acid (P < 0.001). Other isoflavones did not show any activity compared to the control.
Figure 3. Cell growth inhibition by osajin and pomiferin in MCF-7 and MCF-10A cells following 24 h of incubation with graded concentrations of the purified compounds. Sulforhodamine dye binding in 96-well plates was carried out in subconfluent monolayers as described under Materials and Methods. The degree of growth inhibition is expressed as a percentage of control cultures in the presence of vehicle (DMSO) only.
against the MCF-7, but not the MCF-10A, cells (Figure 3). The IC50 value was 5.2 ( 0.9 μM for pomiferin in MCF-7 cells and >20 μM in MCF-10A cells. In contrast, osajin, which is structurally very similar to pomiferin, showed little toxicity against either cell line in this same assay (Figure 3). Microarray Analysis. MCF-7 and MCF-10A cells were treated with 5 μM pomiferin for 24 h, and RNA was isolated for microarray analysis. After genes with low expression levels (
