ARTICLE IN PRESS Journal of Nutritional Biochemistry xx (2006) xxx–xxx Estrogen receptor a as a target for indole-3-carbinol Thomas T.Y. Wanga,4, Matthew J. Milnerb, John A. Milnerc, Young S. Kimc aPhytonutrients Laboratory, Beltsville Human Nutrition Research Center, Agricultural Research Service, U.S. Department of Agriculture, Beltsville, MD 20705, USA bDepartment of Biology, The Pennsylvania State University, University Park, PA, USA cNutritional Science Research Group, Division of Cancer Prevention, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Received 8 June 2005; received in revised form 14 October 2005; accepted 22 October 2005 Abstract A wealth of preclinical evidence supports the antitumorigenic properties of indole-3-carbinol (I3C), which is a major bioactive food component in cruciferous vegetables. However, the underlying molecular mechanism(s) accounting for these effects remain unresolved. In the present study, estrogen receptor alpha (ER-a) was identified as a potential molecular target for I3C. Treating MCF-7 cells with 100 AM I3C reduced ER-a mRNA expression by approximately 60% compared to controls. This reduction in ER-a transcript levels was confirmed using real-time polymerase chain reaction. The I3C dimer, 3,3V-diindolylmethane (DIM), was considerably more effective in depressing ER-a mRNA in MCF-7 cells than the monomeric unit. The suppressive effects of 5 AM DIM on ER-a mRNA was comparable to that caused by 100 AM I3C. DIM is known to accumulate in the nucleus and is a preferred ligand for aryl hydrocarbon receptor (AhR) to I3C. The addition of other AhR ligands, a-naphthoflavone (a-NF, 10 AM) and luteolin (10 AM), to the culture media resulted in a similar suppression in ER-a mRNA levels to that caused by 5 AM DIM. Thus, it is likely that the binding of ligands to AhR inhibits nuclear ER-a transcript. The results from these experiments suggest that the antitumorigenic effects of I3C in MCF-7 human breast cancer cells may arise from its ability to reduce ER-a expression through the binding of its metabolite, DIM, to the nuclear AhR. D 2006 Published by Elsevier Inc. Keywords: Indole-3-carbinol; 3,3V-Diindolylmethane; Estrogen receptor a 1. Introduction the critical role of ER-a in estrogen-dependent promotion of mammary tumor growth. Thus, it is likely that ER-a Breast cancer influenced by genetic and/or environmen- plays a significant role in breast cancer prevention and/or tal factors is one of the most prevalent cancers in the United therapy [6]. States. Fluctuation in estrogen exposure is considered one of Indole-3-carbinol (I3C) is a phytochemical that has been the important risk factors in the development of cancer at documented in numerous epidemiological and preclinical this site [1]. Many of estrogen’s biological activities are studies to possess mammary cancer preventive properties triggered by binding of estrogen to its receptors including [7,8]. Glucobrassicin, which is found in a variety of cruci- ER- [2]. ER- is a ligand-activated transcription factor and a a ferous vegetables including broccoli, Brussel’s sprout, a member of the nuclear receptor superfamily [3]. ER- is a cabbage, cauliflower and kale, is the primary source of involved not only with the proliferative but also with the I3C in the human diet. Once I3C reaches the acidity of the antiapoptotic effects of estrogen [4].UsinganER- a stomach, it can be converted to a number of derivatives [9]. knockout mouse model, Bocchinfuso et al. [5] demonstrated One of the most prevalent derivatives found in humans is 3,3V-diindolylmethane (DIM) [10]. Mounting preclinical evidence with cell cultures and animal models suggest that Abbreviations: ER-a, Estrogen receptor alpha; GAPDH, Glyceralde- both I3C and DIM have promising effects on breast cancer hyde-3-phosphate dehydrogenase; I3C, Indole-3-carbinol; DIM, 3,3V- prevention [11–13]. However, it remains unclear how these Diindolylmethane; RT-PCR, Reverse transcriptase-mediated polymerase chain reaction. interact with specific targets among the pathways involved 4 Corresponding author. Tel.: +1 301 504 8459. with estrogen. The current studies were designed to examine E-mail address: [email protected] (T.T.Y. Wang). ER-a as a key target for dietary indoles. 0955-2863/$ – see front matter D 2006 Published by Elsevier Inc. doi:10.1016/j.jnutbio.2005.10.012 ARTICLE IN PRESS 2 T.T.Y. Wang et al. / Journal of Nutritional Biochemistry xx (2006) xxx–xxx 2. Materials and methods into a 50-ml Falcon screw cap tube (Becton Dickinson, Franklin Lakes, NJ), and chloroform (200 l/ml Trizol) was 2.1. Reagents and cell lines A added. The tubes were then vortexed for 20 s and Microarray analyses were performed using the 3DNA centrifuged at 48C, 12,000 g for 15 min. The clear top  Submicro EX Expression Array Detection kits that were layer containing the RNA was recovered, and an equal purchased from Genisphere (Hatfield, PA). The cDNA volume of 70% ethanol was added. Samples were than microarray slides (Hs-UniGEM2) containing 10,368 genes mixed and loaded on to Qiagen RNeasy Midi Kit. were obtained from the Advanced Technology Center, Purification continued from this point following Qiagen’s National Cancer Institute (Gaithersburg, MD). Human protocols. For other experiments, total RNA was isolated as mammary cancer cell line MCF-7 was obtained from described previously [14]. American Type Culture Collection. Media and reagents for 2.5. Probe synthesis and detection of differential gene cell culture and Trizol were from Invitrogen (Carlsbad, CA). expression using cDNA microarray I3C (purity, 29%) was purchased from Sigma (St. Louis, MO) and used without additional purification. DIM was Differential expression of genes between control and purchased from LKT laboratories (St. Paul, MN). Qiagen treated cells was assessed using cDNA microarray RNeasy Midi Kit was purchased from Qiagen (Valencia, techniques. The Genisphere 3DNA detection methods CA). Alpha-naphthoflavone (a-NF) and luteolin were (3DNA Submicro EX Expression Array Detection kit, purchased from Sigma (St. Louis, MO). Genisphere, Hatfield, PA) were used for labeling and detection of differential expression. Briefly, total RNA 2.2. Cell culture from control and treated cells was reverse transcribed MCF-7 cells were cultured as described previously [4]. using RT primers tagged with either Cy5- (control) or The experiments were conducted in media containing 5% Cy3- (treated) specific 3DNA capture sequence. After charcoal-dextran-treated serum to minimize background hybridization of the probes to cDNA microarray slides, steroid levels. For microarray experiments, 1 108 cells/ the synthesized tagged cDNAs were then fluorescent la-  flask were plated into T-175 flasks. Twenty-four hours after beled by Cy3-3DNA or Cy5-3DNA based on the comple- plating, cells were treated with either 100 AM I3C or mentary capture sequence of 3DNA capture reagents. vehicle (dimethylsulfoxide). For reverse transcriptase-me- After stringency washes, the microarray slides were then diated polymerase chain reaction (RT-PCR) experiments, scanned using ScanArray 4000 microarray scanner (Per- 1 106 to 2 106 cells were plated on 6-well plates. Twenty- kin-Elmer, Boston, MA). The intensity of Cy3- and Cy5-   four hours after plating, cells were treated with either labeled spots was quantified using QuantArray software vehicle or test compounds. In all experiments, media con- (Perkin-Elmer). The results are expressed as a relative taining either test compounds or vehicle were replenished value to the control. daily. In experiments using the aryl hydrocarbon receptor 2.6. Determination of estrogen receptor alpha mRNA levels (AhR) antagonists, cells were treated with vehicle, 100 AM I3C or 10 AM DIM for 4 h prior to the addition of a-NF Semiquantitative RT-PCR was used to validate changes (10 AM) or luteolin (10 AM) and allowed to grow for a total in estrogen receptor alpha (ER-a) mRNA levels. RT was of 72 h. performed using 2 Ag of total RNA as previously described [4]. The TaqMan real-time PCR method was then used to 2.3. Cell number determination quantify ER-a message levels. TaqMan reactions were For experiments examining time-dependent effects of carried out using the TaqMan universal PCR master mix I3C or DIM on MCF-7 cells, cells were treated in the (Applied Biosystems, Branchburg, NJ) in a total volume of presence or absence of I3C (100 AM) or DIM (25 AM) for 25 Al on an ABI-PRISM 7000 Sequence Detector (Applied 0–72 h. Cells (1 105 cells/well) were plated in 24-well Biosystems). Oligonucleotide primers and TaqMan probe  plates, and treatments began 24 h after plating. To assess for ER-a mRNA were made according to published se- effects of I3C or DIM on estradiol, cells were treated with quences [15]. Primers and a probe for glyceraldehyde-3- 10 or without 10À M 17h-estradiol in the presence or phosphate dehydrogenase (GAPDH) reference gene were absence of I3C (100 AM) or DIM (5 AM). We used lower purchased from ABI-PE Applied Biosystems. For the de- DIM concentrations to avoid complications from cell death. termination of ER-a mRNA, 200 nM probe and 300 nM Cell number was assessed using a previously published of each primer were used. For GAPDH, 100 nM probe and method [14]. 40 nM of each primer were used. The amplifications were performed in triplicate for each sample, and the 2.4. RNA isolation PCR optimal conditions were 508C for 2 min, 958C for Total RNA for microarray studies was isolated using a 10 min followed by 40 cycles of 958C for 15 s and 608C combination of Trizol and Qiagen RNeasy Midi Kit for 1 min. The Ct method, as described in the manufac- methods. Briefly, cells were lysed directly in a T-175 flask turer’s protocol, was used to generate relative expres- with 5 ml Trizol (Invitrogen).