Functional and physical interaction between the selenium-binding ...

Carcinogenesis vol.31 no.8 pp.1360–1366, 2010 doi:10.1093/carcin/bgq114 Advance Access publication June 7, 2010

Functional and physical interaction between the selenium-binding protein 1 (SBP1) and the glutathione peroxidase 1 selenoprotein Wenfeng Fang1,2,3, Marci L.Goldberg1, Nicole M.Pohl1, Xiuli Bi1, Chang Tong1, Bin Xiong3, Timothy J.Koh4, Alan M.Diamond1,5 and Wancai Yang1,5,
1 Department of Pathology, University of Illinois at Chicago, 840 South Wood Street, CSN 130, Chicago, IL 60612, USA, 2The Open Laboratory for Overseas Scientists and 3Department of Oncology, Zhongnan Hospital, Wuhan University, Wuhan 430071, China and 4Department of Kinesiology and Nutrition and 5Cancer Center, University of Illinois at Chicago, Chicago, IL 60612, USA
To whom correspondence should be addressed. Tel: þ1 312 355 4154; Fax: þ1 312 996 7586; Email: [email protected]

Selenium-binding protein (SBP) 1 is present in reduced levels in several cancer types as compared with normal tissues, and lower levels are associated with poor clinical prognosis. Another selenium-containing protein, glutathione peroxidase 1 (GPX1), has been associated with cancer risk and development. The interaction between these representatives of different classes of selenoproteins was investigated. Increasing SBP1 levels in either human colorectal or breast cancer cells by transfection of an expression construct resulted in the reduction of GPX1 enzyme activity. Increased expression of GPX1 in the same cell types resulted in the transcriptional and translational repression of SBP1, as evidenced by the reduction of SBP1 messenger RNA and protein and the inhibition of transcription measured using an SBP1 reporter construct. The opposing effects of SBP1 and GPX1 on each other were also observed when GPX1 was increased by supplementing the media of these tissue culture cells with selenium, and the effect of selenium on SBP1 was shown to be GPX1 dependent. Decreasing or increasing GPX1 levels in colonic epithelial cells of mice fed a selenium-deficient, -adequate or -supplemented diet resulted in the opposing effect on SBP1 levels. These data are explained in part by the demonstration that SBP1 and GPX1 form a physical association, as determined by coimmunoprecipitation and fluorescence resonance energy transfer assay. The results presented establish an interaction between two distinct selenium-containing proteins that may enhance the understanding of the mechanisms by which selenium and selenoproteins affect carcinogenesis in humans.

Introduction Mammalian selenium-containing proteins fall into three distinct categories (1,2). In one, selenium is erroneously substituted for sulfur in sulfur-containing amino acids due to the similarity in structure between these two elements. The second class consists of proteins in which selenium is a constituent of the amino acid selenocysteine, which is inserted cotranslationally into selenoproteins in response to a UGA codon in the corresponding messenger RNA (mRNA) (3). The third class is composed of selenium-binding proteins (SBPs), which bind selenium by a mechanism that has yet to be clarified. Given the wealth of data indicating that low non-toxic levels of selenium can prevent cancer in animal models, and human epidemiology establishing an inverse association between dietary selenium intake and the risk of several cancer types (4–10), there is considerable interest in investigating the role of selenium-containing proteins in cancer etiology, although the Selenium and Vitamin E Cancer Prevention Trial Abbreviations: cDNA, complementary DNA; GFP, green fluorescent protein; GPX1, glutathione peroxidase 1; HA, hemagglutinin; mRNA, messenger RNA; PCR, polymerase chain reaction; SBP, selenium-binding protein 1.

showed lack of efficacy of selenium (11) and other studies have indicated that higher selenium intake may be associated with increased risk of diabetes (11–13). The human SBP gene (SBP1, SELENBP1 or hSP56) (14) is located on chromosome 1 at q21–22 and is the homologue of the mouse SP56 gene that was originally reported as a 56 kDa mouse protein that stably bound 75selenium (15,16). The human complementary DNA (cDNA) contains a 472 amino acid-encoding open reading frame (14) and protein resides both in the nucleus and in the cytoplasm (17). It is expressed in a variety of tissue types, including the heart, lung, kidney and tissues of the digestive tract. The function of SBP1 is unknown although it may be involved in intra-golgi transport (18). SBP1 was recently reported to interact with the von Hippel–Lindau protein (pVHL)-interacting deubiquitinating enzyme 1 (VDU1) and might play a role in ubiquitination/deubiquitination-mediated protein degradation pathways in a selenium-dependent manner (19). We and others have recently reported that the levels of SBP1 are significantly decreased in various epithelial cancers, including those of the prostate (20), stomach (21), ovary (22), lung (17) and colon (23,24). More recently, we found that SBP1 becomes silenced by methylation in colon cancers and can influence sensitivity to oxidative stress, cell migration and tumorigenesis (25). Decreased expression of SBP1 in lung and colorectal cancer is also associated with poor prognosis (17,23,24). Among the 25 human selenocysteine-containing proteins, there is considerable evidence that the cytosolic form of glutathione peroxidase 1 (GPX1) is associated with cancer risk. GPX1 is a ubiquitously expressed enzyme that detoxifies hydroperoxides using reducing equivalents from glutathione. Human studies have demonstrated that a functional GPX1 gene polymorphism at codon 198 resulting in either a proline (Pro) or a leucine (Leu) at that position is associated with increased risk for several types of cancer, including those of the head and neck, breast, colon, lung and prostate (26,27). In addition, allelic loss at the GPX1 locus is a common event in cancer development (28,29). This, as well as in vitro data indicating that reduced GPX1 expression can sensitize mammalian cells to DNA damage (30) and that overexpression can have the opposite effect (31), support a role for GPX1 levels in determining cancer risk. Given the cumulative data indicating possible roles of both SBP1 and GPX1 in cancer development and/or outcome, the interaction of these two seleniumassociated proteins was investigated in several model systems in the present study. Materials and methods Cell lines and cell culture The human colon cancer cell line HCT116 and breast cancer cell line MCF-7 were obtained from the American Type Culture Collection (Manassas, VA). Derivative MCF-7 cell lines obtained by transfection of a GPX1 expression construct (designed MCF-7-GPX1) and a control vector-only (pLNCX)transfected cell line MCF-7 (designed MCF-7-vector) were previously generated (28). HCT116 cells were cultured in McCoy’s 5A medium with 10% fetal bovine serum and antibiotics (10 000 U/ml penicillin and 10 lg/ml streptomycin). The selenium concentration of the fetal bovine serum was determined to be 253 nM by the Analytical Services laboratory at South Dakota State University. MCF-7 cells were cultured in modified Eagle’s medium supplemented with 10% fetal bovine serum and antibiotics (10 000 U/ml penicillin and 10 lg/ml streptomycin). Cells were cultured at 37°C in a humidified atmosphere containing 5% CO2. Sodium selenite (Sigma, St Louis, MO) was dissolved in phosphate-buffered saline and stored at 20°C until used. Plasmids construction and establishment of SBP1 stable transfection cell line A human SBP1 expression plasmid (pIRES2-SBP1) was generated using normal colon mucosa DNA as template for polymerase chain reaction (PCR) amplification and subcloning into the pIRES2 vector (Clontech Laboratory,

Ó The Author 2010. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]

1360

GPX1 and SBP1 interaction and regulation

Mountain View, CA) using the following primers: forward 5#-ATCTCGAGCTATGGCTACGAAATGTGGGAAT-3# and reverse 5#-ATGGATCCTCAAATCCAGATGTCAGAGCTAC-3#. The PCR product was digested with XhoI and HindIII and the digestion products were ligated into the vector. Another SBP1 expression plasmid was generated by inserting human SBP1 cDNA into pcDNA3-hemagglutinin (HA) vector (Clontech Laboratory) using the forward primer 5#-ATCTCGAGATGGCTACGAAATGTGGGAATT-3# and reverse primer 5#- ATTCTAGATCAAATCCAGATGTCAGAGCTAC-3# with XhoI and XbaI digestion. Similarly, a human GPX1 cDNA was amplified using the forward primer 5#-TCAAGCTTGTGCTGCTCGGCTAGCGGCG-3# and reverse primer 5#-ATGGATCCCTGCTACACCCGGCACTTTAT-3#, respectively, and inserted into pECFP-C1 at the HindIII and BamH1 sites. A mutant GPX1 plasmid (pECFP-GPX1-Cys) was constructed by changing the selenocysteine (UGA) into a cysteine (UGC) using two-step PCR and ligation. Two sets of primers were used for the PCR amplification: set 1 forward primer 5#-TCAAGCTTGTGCTGCTCGGCTAGCGGCG-3# and GPX1-Cys reverse primer 5#-. CGCTGCAGCTCGTTCATCTGGGTGTAGTCCCGGACCGTGGTGCCGCA-3#; set 2 GPX1-Cys forward primer 5#-AGCTGCAGCGGCGCCTCGGA3# and reverse primer 5#-ATGGATCCCTGCTACACCCGGCACTTTAT-3#. These two products were digested with PtsI and ligated and then were inserted into the pECFP-C1 vector at the HindIII and BamH1 sites. A human SBP1 promoter (2572 to þ1) reporter construct (designed pGL4-SBP1) was generated by PCR amplification of the human SBP1 promoter and insertion into the pGL4 luciferase reporter at the XhoI and HindIII sites using a forward primer 5#-ATCTCGAGCCCATACATACCAAGCAA-3# and reverse primer 5#-TCAAGCTTCGGGTTTGCTGTGCTGGTGTC-3#. The sequences of all generated recombinant constructs were confirmed by sequencing. Transfection of plasmids into the human colon cancer cells HCT116 was accomplished using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). Clones were selected using 500 lg/ml of G418. Quantitative real-time PCR To quantify expression, total RNA was isolated and cDNA was synthesized using TaqMan MultiScribe Reverse Transcriptase (Applied Biosystems, Foster City, CA) as described recently (24). Quantitative real-time PCR analysis was performed using an ABI Prism 7900-HT Sequence Detection System (96-well, Applied Biosystems) using the forward primer 5#-CCAAGCTCATCACCTGGTCT-3# and reverse primer 5#-TCGATGTCAATGGTCTGGAA3# for GPX1. Primers used for quantification of mRNA levels for SBP1 and b-actin were previously reported (24). Measurement of GPX activity GPX activity was measured using a standard coupled spectrophotometric method as described by Samuels et al. (32). Briefly, cells were washed with ice-cold phosphate-buffered saline twice, scraped and transferred to a 1.5 ml tube. Cells were resuspended in sodium phosphate buffer (0.1 M, pH7.0) and lysed by sonication. Determination of GPX1 activity was calculated based on the consumption of reduced nicotinamide adenine dinucleotide phosphate in a reaction containing reduced glutathione, glutathione reductase, cell lysate and hydrogen peroxide. Units of GPX1 activity are expressed as nanomoles of reduced nicotinamide adenine dinucleotide phosphate oxidized per minute per milligram protein. Assays were conducted in triplicate by using three independently generated lysates from separate cultures. Immunoblotting Cells were harvested and suspended in RIPA lysis buffer (Upstate Biotechnology, Lake Placid, NY) containing a protease inhibitor cocktail (Sigma–Aldrich Corporation, St Louis, MO). After incubation on ice for 15 min, cell lysates were centrifuged at 12 000 r.p.m. for 15 min at 4°C. The protein content of the supernatant was determined using the Bio-Rad Protein Assay Reagent (Bio-Rad Laboratories, Hercules, CA). Proteins were separated by 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes (Bio-Rad Laboratories). Blots were probed with antibodies specific for SBP1, GPX1 (Medical and Biological Laboratories Co., Nagoya, Japan) and b-actin (Sigma–Aldrich Corporation). Signals were detected by enhanced chemiluminescence Plus reagents (Amersham Pharmacia, Piscataway, NJ). The signals were quantified by densitometry using Quantity One software (Bio-Rad) and normalized to b-actin. SBP1 promoter activity Transfected cells were plated in 24-well plates and cotransfected with either the pGL4-SBP1 or the pGL4 control luciferase reporter plasmid, along with a Renilla luciferase plasmid used for transfection efficiency normalization using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. At 48 h after transfection, the cells were harvested and the firefly and Renilla luciferase activities were measured using a dual luciferase kit

(Promega, Madison, WI). The firefly luciferase data for each sample was normalized based on transfection efficiency using Renilla luciferase activity. Each experiment was repeated at least three times. Mouse diets and tissue preparation Fifteen C57Bl/6 mice were purchased from Harlan Laboratories (Indianapolis, IN) and provided a Torula yeast-based diet containing 0, 0.1 or 0.4 p.p.m. selenium in the form of sodium selenite for 10 weeks, based on the feeding scheme described by Novoselov et al. (33). The 0.1 p.p.m. diet would approximate the 55 mg/day intake in humans that is the Recommended Daily Allowance and 0.4 p.p.m would approximate the 200 mg/day commonly used in human chemoprevention trials with selenium. Mouse colonic and duodenal epithelial cells were isolated by incubating with 15 mM ethylenediaminetetraacetic acid buffer for 30 min at 37°C, and cell pellets were collected by centrifuging at 4°C as described recently (34). Cell lysates were used for immunoblotting analysis using Anti-SBP1 (Medical and Biological Laboratories Co.) and anti-GPX1 (Abcam, Cambridge, MA) antibodies. Coimmunoprecipitation analysis HCT116 cells were cotransfected with the SBP1 expression plasmid pcDNA3HA-SBP1 and GPX1 expression plasmid pECFP-GPX1, GPX1 mutation plasmid pECFP-GPX1-Cys or empty vector pECFP. Twenty-four hours posttransfection, the cells were harvested and lysed in RIPA lysis buffer (20 mM Tris–HCL, pH 8.0, 150 mM NaCl, 1% Triton X-100 and protease inhibitor cocktail) and then the cell lysates were immunoprecipitated with antigreen fluorescent protein (GFP) antibody (recognition of pECFP-GPX1), the latter precipitate was probed with anti-HA antibody (for SBP1). The immunocomplex was visualized using enhance chemiluminescence reagent plus (Amersham Biosciences, Piscataway, NJ).

Results SBP1 negatively regulates GPX1 activity SBP1 and GPX1 represent distinct classes of proteins with the selenium moiety of GPX1 present as the amino acid selenocysteine. In order to determine whether the levels of SBP1 could influence GPX activity, we used a human colon cancer cell line (HCT116) that has undetectable levels of SBP1 but significant levels of GPX1 protein and enzyme activity (Figure 1). Increasing SBP1 levels in these cells by transfection of an expression construct resulted in an 50% reduction in GPX1 activity as compared with cells transfected with parental pIRES2 vector alone (Figure 1A). The reduction in GPX1 activity observed in SBP1-transfected cells was considerably more than the amount seen when cells were transfected with the pIRES2 vector alone. Changes in GPX enzyme activity as a consequence of the transfection process of the magnitude shown have been observed previously (Yang, et al, unpublished results), indicating the necessity for this particular control. As seen in Figure 1, the observed reduction in GPX enzyme activity was posttranslational, not being associated with lowering either GPX1 protein (Figure 1B) or mRNA levels (Figure 1C). Although colon-derived cells typically express both the GPX1 and the gastrointestinal-GPX, GPX2, HCT116 cells contain relatively little GPX2 (Yang, et al, unpublished results). To confirm that SBP1 overexpression could specifically inhibit GPX1 activity and extend these observations to a different cell type, SBP1 expression was increased in MCF-7 human breast carcinoma cells by transfection of the SBP1 expression construct. MCF-7 cells do not express detectable levels of GPX1 mRNA and display low or undetectable total GPX1 enzyme activity (28). Using a previously generated derivative MCF-7 cell line engineered such that all of its detectable GPX activity is exclusively due to GPX1 (28) as the recipient of the SBP1 expression construct, it was shown that overexpression of SBP1 caused a dramatic reduction, 90%, of GPX activity in these cells (Figure 1D). GPX1 negatively regulates SBP1 The results shown in Figure 1 establish that SBP1 could negatively regulate GPX1 activity. To determine whether GPX1 could affect SBP1 abundance, parental and GPX1-expressing MCF-7 cells were examined for SBP1 expression and protein levels (Figure 2). Increased expression of GPX1 resulted in a significant decline in SBP1 protein levels as compared with either the parental MCF-7 or

1361

W.Fang et al.

Fig. 2. GPX1 negatively regulated SBP1 expression. (A) SBP1 protein was inhibited by GPX1. MCF-7 parental, MCF-7-vector and MCF-7-GPX1 stable transfected cells were lysed and subjected to immunoblotting for SBP1 and GPX1. SBP1 and GPX1 signals were quantified and normalized to b-actin, a loading control. Relative intensities were indicated under lanes. (B) SBP1 mRNA was significantly reduced in MCF-7-GPX1 cells compared with the MCF-7-vector cells (
P , 0.01) (n 5 3). (C) GPX1 inhibited SBP1 promoter activity. MCF-7-vector and MCF-7-GPX1 cells were cotransfected with Renilla and the pGL4 control or pGL4-SBP1 promoter. Twenty-four hours after transfection, cells were harvested and luciferase activity determined (
P , 0.01). (n 5 3). Each bar represents the mean ± SD of samples obtained in triplicate. Fig. 1. SBP1 inhibited GPX1 activity in HCT116 cells and MCF-7 cells but did not affect GPX1 protein and mRNA levels. (A) Endogenous GPX1 activity was inhibited by SBP1 in HCT116 cells. HCT116 parental (mock), pIRES2 empty vector control (pIRES2) and SBP1 overexpressing HCT116 cells (SBP1) were lysed and GPX1 activity determined (
P , 0.001) (n 5 3, indicating the experiments were repeated three times). (B) GPX1 protein was not affected by SBP1 overexpression. The HCT116 parental, pIRES2 empty vector control (pIRES2) and SBP1 overexpressing HCT116 cells (SBP1) were lysed and immunoblotted to determine GPX1 and SBP1 protein levels. GPX1 signals were quantified by densitometry and normalized to b-actin, a loading control. (C) GPX1 mRNA levels were not affected by SBP1 overexpression. Quantitative real-time PCR was used to quantify GPX1 mRNA in HCT116 parental, pIRES2 control and SBP1 overexpressing HCT116 cells. The data were normalized to HCT116 parental cells (n 5 3). (D) GPX1 activity derived from an expression construct was inhibited by SBP1 in MCF-7-GPX1 cells. MCF-7-GPX1 cells were treated with transfection reagents without plasmid (mock), pIRES2 control vector or pIRES2-SBP1 plasmid. Forty-eight hours after transfection, cells were harvested and GPX1 activity was determined (
P , 0.001) (n 5 3).

1362

the MCF-7-vector cells (Figure 2A). This reduction in SBP1 protein levels was associated with 50% reduction in SBP1 mRNA levels as determined by Quantitative real-time PCR (Figure 2B). To determine whether the observed reduction in SBP1 RNA by GPX1 was due to either an effect on transcription or RNA stability, a reporter construct was generated using human SBP1 promoter sequences and luciferase as the reporter protein. As seen in Figure 2C, overexpression of GPX1 inhibited transcription from the SBP1 promoter driving the reporter gene. Contrasting effects of selenium supplementation on SBP1 and GPX1 The data presented above indicated that increasing GPX1 by transfection resulted in reduction of SBP1 and increasing SBP1 by transfection resulted in the reduction of GPX1 activity. An alternative approach to enhance GPX1 is by the addition of sodium selenite. GPX1 activity in cell lines is generally inducible by low physiologically relevant doses of selenium (27). The media of HCT116-SBP1 and MCF-7-GPX1 cells were each supplemented with 25, 50, 100 or

GPX1 and SBP1 interaction and regulation

tained by western blotting (data not shown). One possible explanation for these results is that SBP1 and GPX1 are competing for the available cellular selenium pool. A selenium titration study was performed on HCT116-SBP1 cells using an expanded range of selenium supplementation, from 0 to 5 lM (Figure 4B). GPX1 levels increased and SBP1 declined with supplementation, reaching a maximum at 250 nM and then the trends reversed with increasing supplementation with GPX1 declining and SBP1 increasing at the higher selenium concentrations. Similar results were obtained with MCF-7-GPX1 transfectants (data not shown). Effects of selenium status on SBP1 and GPX1 in mouse intestinal epithelial cells The inverse association between SBP1 and GPX1 observed in tissue culture cells was examined in intestinal epithelial cells obtained from C57Bl/6 mice fed a selenium-deficient (0 p.p.m.), -adequate (0.1 p.p.m.) or -enriched (0.4 p.p.m.) diet. The levels of selenium in the diet were selected as the lowest will result in the a dramatic reduction in GPX1 in most tissues, the mid range corresponds to the amount of selenium recommended for human intake and the highest dose is one used for chemoprevention studies and would correspond to the level used in human chemoprevention trials (35). Animals were fed the respective diets for 10 weeks and colon and duodenal epithelial cells were isolated for western blotting analysis using anti-SBP1 and antiGPX1 antibodies. As shown in Figure 5, there were very low levels of GPX1 in cells derived from the animals maintained on the seleniumdeficient diet, much higher levels in those fed the ‘adequate’ diet and GPX1 protein was the highest in cells obtained from mice on selenium-supplemented diet. Consistent with the cell culture studies, there was an inverse association between GPX1 levels and SBP1 levels in intestinal epithelial cells derived from both colon and duodenum. Fig. 3. Contrasting effects of sodium selenite on SBP1 and GPX1. (A) Sodium selenite increased GPX1 expression and reduced SBP1 expression in the MCF-7-GPX1 cells with addition of 0, 25, 100 or 250 nM sodium selenite to the basal media, which contained at least 30 nM Se from the 10% fetal bovine serum for 48 h. Cells were harvested and immunoblotted to detect SBP1 and GPX1. SBP1 and GPX1 signals were quantified by densitometry and normalized to b-actin, a loading control, with the relative intensities indicated under the lanes. Similar experiments were performed in a HCT116SBP1 stably transfected cell line. (B) SBP1 was not affected by sodium selenite in the MCF-7 cells that did not express GPX1. (C) All experiments were repeated at least three times and representative blots shown.

250 nM of selenium in the form of sodium selenite for 48 h and levels of GPX1 and SBP1 were determined by western blotting (Figure 3A and B). As is typically observed, there was a dose–response with GPX enzyme activity increasing with selenium supplementation (data not shown). Consistent with what was observed in the transfection studies, there was an inverse association between the increase of GPX1 that occurred in a dose-dependent manner and the reduction of SBP1. Since MCF-7 is null for GPX1 (28), use of these cells allows one to determine whether the reduction in SBP1 that occurred with increasing selenium concentration was GPX1 dependent. MCF-7 cells transfected only with an empty vector were incubated with increasing selenium concentration under the same conditions described above for the GPX1 expressing lines and GPX1 and SBP1 were quantified. As expected, no GPX1 signal was detectable in the western blot (Figure 3C). In contrast to the results obtained with GPX1-expressing cells, the MCF-7 cells did not exhibit a selenium-dependent decline in SBP1, indicating that the selenium-mediated increment of GPX1 was responsible for the corresponding decline of SBP1. The relationship between GPX1 and SBP1 was further examined by quantifying GPX1 enzyme activity in HCT116 transfected with either SBP1 or vector and incubating the cells with increasing amounts of selenium (Figure 4A). It is apparent that baseline GPX1 activity was lower in the SBP1-expressing cells and the seleniummediated increase in activity was attenuated at the tested doses. The obtained enzyme activity corresponded to the levels of protein ob-

Direct interaction between SBP1 and GPX1 The ability of SBP1 and GPX1 to influence each others abundance and/or activity raised the possibility that there were direct interactions between these two proteins. To investigate this possibility, a plasmid that expressed an HA-tagged SBP1 (pcDNA3-HA-SBP) was cotransfected with an expression construct containing a GFP-GPX1 fusion protein expression construct (pECFP-GPX1), mutant GPX1 plasmid (pECFP-GPX1-CYS) or the empty vector (pECFP) into HCT116 cells. Extracts from transfectants were analyzed by western blotting (Figure 6A). Using antibodies directed against GFP, it is apparent that the vector-only transfectant expressed GFP, whereas the cells expressing the GFP-GPX1 fusion protein revealed a larger protein whose molecular weight was consistent with that expected from the fusion (49 KD). As seen with the earlier studies, cells expressing the GFP– GPX1 fusion also expressed less SBP1 as indicated by the reduced signal seen with anti-HA antibodies. In order to determine whether SBP1 and GPX1 physically bound each other, extracts from the three transfectants were immunoprecipitated using anti-GFP antibodies and probed with anti-HA to probe for the SBP1-HA-tagged protein in the precipitates. As seen in Figure 6B, HA-tagged SBP1 was only present in immunoprecipitates from cells that were also expressing the GFP– GPX1 fusion proteins, including a derivative GPX1 in which the selenocysteine coding UGA triplet was changed to the codon of cysteine (GPX1-Cys), indicating the binding of SBP1 to GPX1. This also excluded the possibility that the interaction occurs by the binding of SBP1 to the selenium moiety of GPX1. The interaction between SBP1 and GPX1 was further supported by fluorescence resonance energy transfer assay (supplementary Figure S1 is available at Carcinogenesis Online). Discussion The goal of the work presented in this manuscript was to investigate whether two proteins representing different classes of seleniumcontaining molecules functionally interacted. The first of these, SBP1, binds selenium and low levels of SBP1 are associated with

1363

W.Fang et al.

Fig. 4. (A) SBP1 attenuated sodium selenite-induced increase of GPX1 activity in HCT116 cells. HCT116-pIRES2 control and HCT116-SBP1 stable cells were treated with addition of 0, 25, 100 or 250 nM sodium selenite to the basal media, which contained at least 30 nM Se from the 10% fetal bovine serum. Forty-eight hours after treatment, the cells were lysed and GPX1 activity was determined. (B) GPX1 levels increased and SBP1 declined with supplementation of sodium selenite, reaching a maximum at 250 nM and then the trends reversed with increasing supplementation with GPX1 declining and SBP1 increasing at the higher selenium concentrations. SBP1 and GPX1 signals were quantified by densitometry and normalized to b-actin, a loading control, with the relative intensities indicated under the lanes. All experiments were repeated at least three times and representative blots shown.

cancer and poor clinical outcome (17,23,24). The second, GPX1, contains selenium as the amino acid selenocysteine (2), and increasing or decreasing GPX1 levels has an inverse effect of susceptibility of cells to DNA damage (31,36–39). Human data have indicated that functional polymorphisms in GPX1 are associated with increased cancer risk and allelic loss of GPX1 frequently occurs during cancer development (26,28,29). SBP1 and/or GPX1 are therefore potential mediators of the chemopreventive effects of selenium identified by human epidemiology and animal supplementation studies. Increasing the levels of SBP1 in either colon-derived HCT116 cells or MCF-7 breast carcinoma cells resulted in a consequential decline in GPX1 enzyme activity. How SBP1 affected GPX1 activity remains to be determined, but neither GPX1 RNA nor protein levels were changed (Figure 1). One possible mechanism to explain these results would involve a direct physical interaction between the two proteins resulting in reduced GPX activity. Support for this notion can be seen in the coprecipitation data presented in Figure 6 showing that SBP1 and GPX1 do form a complex. How these proteins interact remains to be determined, but it is unlikely that this interaction occurs by the binding of SBP1 to the selenium atom at the GPX1 active site as the selenium is probably to be inaccessible to a molecule the size of SBP1 and SBP1 is able to bind mutant GPX1 (Figure 6B). As seen in Figure 2, overexpression of GPX1 resulted in the reduction of SBP1 in the same cell lines used to investigate the effects of SBP1 overexpression. This effect is due, at least in part, to the inhibition of SBP1 transcription, as determined by the GPX1-mediated reduction in both SBP1 mRNA (Figure 2B) and expression from an

1364

SBP1 promoter-driven reporter gene (Figure 2C). Little is known about the regulation of SBP1 transcription, but it is possible that SBP1 expression is dependent on reactive oxygen-responsive transcriptional elements, and the reduction in peroxides that is expected to be achieved with increased GPX1 activity results in the attenuation of transcription. A search for an antioxidant response element (40) in the DNA region upstream of the SBP1 transcription start site revealed two sequences (GTGACCCTGGC and TGACACCAGC) that are very similar to the consensus antioxidant response element recognition motif (40), located just 3# of the transcription start site, although the functionality of this sequence as an antioxidant response element remains to be established. Also consistent with the studies using ectopic expression of either SBP1 or GPX1, selenium titration experiments indicated an inverse association in the expression of these proteins. GPX1 protein levels rose with selenium supplementation of HCT116- and MCF-7-derived cell lines with a corresponding decline in SBP1. This effect was dependent on the presence of GPX1 and not some other consequence or function of selenium. This was established by showing the seleniumassociated decline in SBP1 did not occur in GPX1 null MCF-7 cells (Figure 3C). The relationship between SBP1 and GPX1 was also apparent in mice fed a selenium-deficient, -adequate or -supplemented diet. Selenium deficiency resulted in a dramatic reduction of GPX1 in colonic and duodenal epithelial cells, whereas selenium supplementation resulted in elevated GPX1 as compared with those animals fed a selenium-adequate diet. For each level of dietary intake, there was a corresponding decline in SBP1 levels,

GPX1 and SBP1 interaction and regulation

Fig. 5. Dietary selenium status affected GPX1 and SBP1 expression in mouse colonic (A) and duodenal epithelial cells (B). Mice were fed with seleniumadequate diet (0.1 p.p.m.), selenium-deficient diet (0 p.p.m., Se-deficiency) or selenium-enriched diet (0.4 p.p.m. Se-supplemental) for 10 weeks. Intestinal epithelial cells were isolated for western blotting analysis for GPX1 and SBP1. b-actin was used as loading control. Each lane represented one individual mouse and three mice from each group were studied.

extending the results obtained in cell culture to animals. It is particularly noteworthy that supra nutritional amounts of selenium in the diet that have been associated with cancer prevention in animals studies, including those investigating colon cancer prevention (41), resulted in changes in the levels of the two investigated seleniumcontaining proteins. Collectively, the work presented herein establishes an interaction among selenium availability, the amount of a protein that binds selenium—SBP1 and another protein in which selenium is present as the amino acid selenocysteine. These data raise questions as to whether higher levels of GPX1 in tumors promote carcinogenesis by suppressing SBP1 and whether SBP1 has biological functions beyond its interaction with GPX1. Furthermore, there may be biological consequences of the regulatory interaction between GPX1 and SBP1 that reflect on the potential health consequences of selenium dietary intake. An inverse association between dietary selenium intake and cancer risk has been reported for several organ systems (42). Encouraged by the results that were initially presented for the Nutritional Prevention of Cancer selenium supplementation trial (43,44), the large cancer prevention study selenium and vitamin E cancer prevention trial that investigated the efficacy of selenium, both in concert with vitamin E and alone, in the prevention of prostate cancer was initiated (45). This study was terminated early, in part due to the lack of efficacy of selenium (11), although the analysis of the cohort following stratification by selenium status and genetic polymorphisms has not yet been reported. Data from selenium and vitamin E cancer prevention trial, as well as an additional recent studies, have indicated that higher selenium intake may also be associated with increased risk of diabetes (11–13). Whether any of the possible benefits or risks of selenium intake are due to consequential effects on GPX1 and SBP1 remains to be determined. Supplementary material Supplementary FigureS1 can be found at http://carcin.oxfordjournals .org/ Funding National Cancer Institute (CA112081 to W.Y. and CA127943 to A.M.D.); American Institute for Cancer Research (05A121 to W.Y.); State Scholarship Fund of China (2007-3020) to F.W. Acknowledgements We would like to thank Dr Frank J.M.van Kuppeveld (Radboud University Nijmegen Medical Centre, Netherlands) for providing the pECFP and pEYFPECFP plasmids and thank Dr Dong Hu (University of Illinois at Chicago, IL) for technical assistance. Conflict of Interest Statement: None declared.

References Fig. 6. SBP1 interacts with GPX1 as determined by immunoprecipitation (IP). HCT116 cells were cotransfected with SBP1 expression plasmid pcDNA-HA-SBP1 and GPX1 expression plasmid pECFP-GPX1 or empty vector pECFP or Sec-Cys GPX1 plasmid (pECFP-GPX1-Cys). (A) Wholecell lysates were analyzed by immunoblotting (IB) to confirm transfection efficiency and downregulation of SBP1 by GPX1. (B) The cell lysates were incubated with anti-GFP antibody to immunoprecipitate pECFP-GPX1 and pECFP-GPX1-Cys and then probed with anti-HA antibody to detect the HAtagged SBP1. Non-specific peptide (NSP) in the same blot showed the equal immunoprecipitation of pECFP, pECFP-GPX1 and pECFP-GPX1-Cys. Lane 1, cell lysate from cells cotransfected with HA-SBP1 and pECFP; lane 2, cell lysate from cells cotransfected with HA-SBP1 and pECFP-GPX1 and lane 3, cell lysate from cells cotransfected with HA-SBP1 and pECFP-GPX1-Cys. Equal amount of protein were added to the gel and all experiments were repeated at least three times with representative blots shown.

1. Behne,D. et al. (2001) Mammalian selenium-containing proteins. Annu. Rev. Nutr., 21, 453–473. 2. Kryukov,G.V. et al. (2003) Characterization of mammalian selenoproteomes. Science, 300, 1439–1443. 3. Hatfield,D.L. et al. (2002) How selenium has altered our understanding of the genetic code. Mol. Cell. Biol., 22, 3565–3576. 4. Wang,L. et al. (2009) Methyl-selenium compounds inhibit prostate carcinogenesis in the transgenic adenocarcinoma of mouse prostate model with survival benefit. Cancer Prev. Res. (Phila Pa), 2, 484–495. 5. Clark,L.C. et al. (1996) Effects of selenium supplementation for cancer prevention in patients with carcinoma of the skin. A randomized controlled trial. Nutritional Prevention of Cancer Study Group. JAMA, 276, 1957–1963. 6. Facompre,N. et al. (2009) Potential stages for prostate cancer prevention with selenium: implications for cancer survivors. Cancer Res., 69, 2699–2703. 7. Greenwald,P. et al. (2007) Clinical trials of vitamin and mineral supplements for cancer prevention. Am. J. Clin. Nutr., 85, 314S–317S.

1365

W.Fang et al.

8. Hawk,E.T. et al. (2005) Colorectal cancer prevention. J. Clin. Oncol., 23, 378–391. 9. Ip,C. (1998) Lessons from basic research in selenium and cancer prevention. J. Nutr., 128, 1845–1854. 10. Peters,U. et al. (2008) Selenium and the prevention of prostate and colorectal cancer. Mol. Nutr. Food Res., 52, 1261–1272. 11. Lippman,S.M. et al. (2009) Effect of selenium and vitamin E on risk of prostate cancer and other cancers: the Selenium and Vitamin E Cancer Prevention Trial (SELECT). JAMA, 301, 39–51. 12. Stranges,S. et al. (2007) Effects of long-term selenium supplementation on the incidence of type 2 diabetes: a randomized trial. Ann. Intern. Med., 147, 217–223. 13. Stranges,S. et al. Higher selenium status is associated with adverse blood lipid profile in British adults. J. Nutr., 140, 81–87. 14. Chang,P.W. et al. (1997) Isolation, characterization, and chromosomal mapping of a novel cDNA clone encoding human selenium binding protein. J. Cell. Biochem., 64, 217–224. 15. Bansal,M.P. et al. (1989) Evidence for two selenium-binding proteins distinct from glutathione peroxidase in mouse liver. Carcinogenesis, 10, 541–546. 16. Bansal,M.P. et al. (1990) DNA sequencing of a mouse liver protein that binds selenium: implications for selenium’s mechanism of action in cancer prevention. Carcinogenesis, 11, 2071–2073. 17. Chen,G. et al. (2004) Reduced selenium-binding protein 1 expression is associated with poor outcome in lung adenocarcinomas. J. Pathol., 202, 321–329. 18. Porat,A. et al. (2000) A 56-kDa selenium-binding protein participates in intra-Golgi protein transport. J. Biol. Chem., 275, 14457–14465. 19. Jeong,J.Y. et al. (2009) Human selenium binding protein-1 (hSP56) interacts with VDU1 in a selenium-dependent manner. Biochem. Biophys. Res. Commun., 379, 583–588. 20. Yang,M. et al. (1998) Differential expression and androgen regulation of the human selenium-binding protein gene hSP56 in prostate cancer cells. Cancer Res., 58, 3150–3153. 21. He,Q.Y. et al. (2004) Diverse proteomic alterations in gastric adenocarcinoma. Proteomics, 4, 3276–3287. 22. Huang,K.C. et al. (2006) Selenium binding protein 1 in ovarian cancer. Int. J. Cancer, 118, 2433–2440. 23. Kim,H. et al. (2006) Suppression of human selenium-binding protein 1 is a late event in colorectal carcinogenesis and is associated with poor survival. Proteomics, 6, 3466–3476. 24. Li,T. et al. (2008) Expression of selenium-binding protein 1 characterizes intestinal cell maturation and predicts survival for patients with colorectal cancer. Mol. Nutr. Food Res., 52, 1289–1299. 25. Pohl,N.M. et al. (2009) Transcriptional regulation and biological functions of selenium-binding protein 1 in colorectal cancer in vitro and in nude mouse xenografts. PLoS ONE, 4, e7774. 26. Zhuo,P. et al. (2009) Molecular mechanisms by which selenoproteins affect cancer risk and progression. Biochim. Biophys. Acta, 1790, 1546–1554. 27. Lei,X.G. et al. (2007) Metabolic regulation and function of glutathione peroxidase-1. Annu. Rev. Nutr., 27, 41–61.

1366

28. Hu,Y.J. et al. (2003) Role of glutathione peroxidase 1 in breast cancer: loss of heterozygosity and allelic differences in the response to selenium. Cancer Res., 63, 3347–3351. 29. Hu,Y. et al. (2005) Allelic loss of the gene for the GPX1 selenium-containing protein is a common event in cancer. J. Nutr., 135, 3021S–3024S. 30. Baliga,M.S. et al. (2008) Selenoprotein deficiency enhances radiationinduced micronuclei formation. Mol. Nutr. Food Res., 52, 1300–1304. 31. Baliga,M.S. et al. (2007) Selenium and GPx-1 overexpression protect mammalian cells against UV-induced DNA damage. Biol. Trace Elem. Res., 115, 227–242. 32. Samuels,B.L. et al. (1991) Increased glutathione peroxidase activity in a human sarcoma cell line with inherent doxorubicin resistance. Cancer Res., 51, 521–527. 33. Novoselov,S.V. et al. (2005) Selenoprotein deficiency and high levels of selenium compounds can effectively inhibit hepatocarcinogenesis in transgenic mice. Oncogene, 24, 8003–8011. 34. Hu,D. et al. (2009) GSK3beta is involved in JNK2-mediated beta-catenin inhibition. PLoS One, 4, e6640. 35. Bellinger,F.P. et al. (2009) Regulation and function of selenoproteins in human disease. Biochem. J., 422, 11–22. 36. Fu,Y. et al. (2001) Opposite roles of selenium-dependent glutathione peroxidase-1 in superoxide generator diquat- and peroxynitrite-induced apoptosis and signaling. J. Biol. Chem., 276, 43004–43009. 37. Gouaze,V. et al. (2002) Glutathione peroxidase-1 protects from CD95induced apoptosis. J. Biol. Chem., 277, 42867–42874. 38. Nasr,M.A. et al. (2004) GPx-1 modulates Akt and P70S6K phosphorylation and Gadd45 levels in MCF-7 cells. Free Radic. Biol. Med., 37, 187–195. 39. Brigelius-Flohe,R. et al. (2009) Glutathione peroxidases in different stages of carcinogenesis. Biochim. Biophys. Acta, 1790, 1555–1568. 40. Rushmore,T.H. et al. (1991) The antioxidant responsive element. Activation by oxidative stress and identification of the DNA consensus sequence required for functional activity. J. Biol. Chem., 266, 11632– 11639. 41. Davis,C.D. et al. (2002) Selenium-enriched broccoli decreases intestinal tumorigenesis in multiple intestinal neoplasia mice. J. Nutr., 132, 307–309. 42. Gromadzinska,J. et al. (2008) Selenium and cancer: biomarkers of selenium status and molecular action of selenium supplements. Eur. J. Nutr., 47 (suppl. 2), 29–50. 43. Clark,L.C. et al. (1996) Effects of selenium supplementation for cancer prevention in patients with carcinoma of the skin. A randomized controlled trial. JAMA, 276, 1957–1963. 44. Duffield-Lillico,A.J. et al. (2003) Selenium supplementation, baseline plasma selenium status and incidence of prostate cancer: an analysis of the complete treatment period of the Nutritional Prevention of Cancer Trial. BJU Int., 91, 608–612. 45. Lippman,S.M. et al. (2005) Designing the Selenium and Vitamin E Cancer Prevention Trial (SELECT). J. Natl Cancer Inst., 97, 94–102. Received September 23, 2009; revised May 25, 2010; accepted May 29, 2010