Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Evidence that a polymorphism within the 3′UTR of glutathione peroxidase 4 is functional and is associated with susceptibility to colorectal cancer

Evidence that a polymorphism within the 3′UTR of glutathione peroxidase 4 is functional and is... Genes Nutr (2007) 2:225–232 DOI 10.1007/s12263-007-0052-3 RESEARCH PAPER Evidence that a polymorphism within the 3 UTR of glutathione peroxidase 4 is functional and is associated with susceptibility to colorectal cancer G. Bermano Æ V. Pagmantidis Æ N. Holloway Æ S. Kadri Æ N. A. G. Mowat Æ R. S. Shiel Æ J. R. Arthur Æ J. C. Mathers Æ A. K. Daly Æ J. Broom Æ J. E. Hesketh Received: 24 November 2006 / Accepted: 18 January 2007 / Published online: 13 October 2007 Springer-Verlag 2007 Abstract Low selenium (Se) status has been associated study was carried out in cohorts of patients with either with increased risk of colorectal cancer (CRC). Se is adenomatous polyps, colorectal adenocarcinomas and in present as the amino acid selenocysteine in selenoproteins, healthy controls. A higher proportion of individuals with such as the glutathione peroxidases. Se incorporation CC genotype at the GPx4 T/C 718 SNP was present in the requires specific RNA structures in the 3 untranslated cancer group, but not in the polyp group, compared with region (3 UTR) of the selenoprotein mRNAs. A single the control group (P \ 0.05). The present data demonstrate nucleotide polymorphism (SNP) occurs at nucleotide 718 the functionality of the GPx4 T/C 718 SNP and suggest (within the 3 UTR) in the glutathione peroxidase 4 gene. In that T genotype is associated with lower risk of CRC. the present study, Caco-2 cells were transfected with constructs in which type 1 iodothyronine deiodinase coding Keywords Colorectal cancer  GPx4  Reporter gene 0 0 region was linked to the GPx4 3 UTR with either C or T Selenium  SNP  3 Untranslated region variant at position 718. Higher reporter activity was observed in cells expressing the C variant compared to those expressing the T variant, under either Se-adequate or Introduction Se-deficient conditions. In addition, a disease association Selenium (Se) is a dietary micronutrient that is essential for human health [12, 34] and low Se status has been associ- ated with increased mortality [1]. Worldwide, mortality G. Bermano and V. Pagmantidis have contributed equally to this work. G. Bermano  N. Holloway  J. Broom J. C. Mathers School of Life Sciences, The Robert Gordon University, School of Clinical Medical Sciences and Human Nutrition Aberdeen AB25 1HG, UK Research Centre, Newcastle University, Newcastle upon Tyne NE2 4HH, UK V. Pagmantidis Institute for Cell and Molecular Biosciences, J. E. Hesketh (&) Newcastle University, Institute for Cell and Molecular Biosciences and Human Newcastle upon Tyne NE2 4HH, UK Nutrition Research Centre, Newcastle University, Newcastle upon Tyne NE2 4HH, UK S. Kadri  N. A. G. Mowat e-mail: [email protected] Aberdeen Royal Infirmary, Aberdeen AB25 2ZN, UK A. K. Daly R. S. Shiel School of Clinical and Laboratory Sciences, School of Agriculture and Rural Development, Newcastle University, Newcastle upon Tyne NE2 4HH, UK Newcastle University, Newcastle upon Tyne NE1 7RU, UK J. R. Arthur Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK 123 226 Genes Nutr (2007) 2:225–232 from cancers, including colorectal cancer (CRC), is lipoxygenase products implies that the SNP may affect inversely correlated with estimated Se intake [11]. Addi- inflammatory signalling, and, since inflammation in the tionally in case-control studies, low Se intake is associated colon is regarded as a frequent initiating event in carci- with increased risk of CRC [35] and, in a 10-year pro- nogenesis [26, 28], the second objective was to carry out a spective study, initial plasma Se was inversely correlated disease-association study focusing on the GPx4 T/C 718 with subsequent risk of CRC [9]. Moreover, a clinical trial SNP in patients with CRC and in controls. The data indi- in the United States indicated that a daily supplement of cate that the T/C 718 SNP has functional effects on Se 200 lg Se led to reduced mortality from CRC. Carcino- incorporation and that the C variant is more common in genesis in the colon is a multistage process and CRC patients diagnosed with CRC. develops via an adenoma–carcinoma sequence from microscopic dysplastic aberrant crypt foci within the epi- thelium to benign adenomatous polyps, which can then Materials and methods potentially, but not necessarily, undergo malignant change. The sequence of biochemical changes underlying this Plasmid construction sequence is not known but oxidant damage to DNA is a critical event in tumour formation [13, 20, 39]. Se is nec- Constructs containing the rat type 1 iodothyronine deio- essary for the function of *25 selenoproteins, including dinase (IDI) coding region linked to GPx4 3 UTR with the antioxidant enzymes cytosolic glutathione peroxidase either a T or C at position 718, were cloned to assess (GPx1), gastrointestinal specific glutathione peroxidase functionality of the T/C polymorphism in the human GPx4 (GPx2) and phospholipid hydroperoxide glutathione per- 3 UTR at position 718 [42]. The rat IDI coding region was oxidase (GPx4) [8, 41] which provide defence against amplified by PCR from IDI–IDI construct in pCDNA3 [4] oxidative damage. and the human GPx4 3 UTR (T variant) was amplified In selenoproteins, Se is present as the amino-acid sele- from pRc/CMV [45]. PCR primers used for the amplifi- nocysteine (Sec) and its incorporation occurs during cation reactions are shown in Table 1. Following ligation synthesis of the proteins. This incorporation requires the of the rat IDI coding region (BamHI/EcoRI) to the human recoding of a UGA codon that involves the formation of a GPx4 3 UTR (EcoRI/XbaI) (Fig. 1), the product was complex between an RNA stem-loop structure (SECIS) cloned into pCDNA3.1/Zeo (+) (Invitrogen, Paisley, UK) 0 0 located in the 3 untranslated region (3 UTR) of seleno- for expression in mammalian cells. PCR products were protein mRNAs and various trans-acting proteins [6]. Thus purified using QIAquick PCR purification kit (Qiagen; selenoprotein expression can depend on both dietary Se Crawley, UK) according to the manufacturers’ instructions. intake and genetic factors. Furthermore, variants in genes The construct in which the IDI coding region was linked to encoding any of the GPxs affecting antioxidant capacity, the C variant (position 718) of the GPx4 3 UTR was made may change susceptibility to genomic damage by oxidants by site-directed mutagenesis using QuickChange Site- and thus influence neoplastic change. A SNP has been identified in the GPx1 gene which results in a proline to Table 1 Primer sequences used for plasmid construction, mutagen- leucine substitution at codon 198 [18, 19] and this SNP is esis and genotyping associated with lung and bladder cancer risk [23, 24]. Two 0 0 SNPs within the Sep15 gene occur in a region of the 3 UTR Cloning Primers 5 –3 and, although the function of the encoded protein is not rIDI coding region known, there is an association of alleles at these two BamHI CGGGATCCATGGGGCTGTCCCAG positions with susceptibility to breast cancer [22]. Despite EcoRI CGGAATTCCTAGAACTGAGGCATGTGTG the report that Se supplementation reduces CRC mortality hGPx4 3 UTR [10], there is little information on whether SNPs in GPx EcoRI CGGAATTCCTCCACAAGTGTGTGGC genes affect susceptibility to CRC. XbaI GCTCTAGACTAATTTGTCTGTTTATTCCC A SNP (rs713041) has been found at position 718 of the Mutagenesis GPx4 gene mRNA in the 3 UTR region; the regulatory Sense CTGCCCACGCCCTCGGAGCCTTCCACC region required for incorporation of Se into selenoproteins. Antisense GGTGGAAGGCTCCGAGGGCGTGGGCAG The C allele is associated with increased levels of lym- Genotyping phocyte lipoxygenase products [42]. However, since the hGPx4 3 UTR GACCTGCCCCACTATTTCTA effects of the SNP on Se incorporation are not known, the GTCTGTTTATTCCCACAAGG first objective of this work was to use a reporter gene assay in transfected Caco-2 cells [4] to assess the functionality of X: Nucleotides added, X: restriction enzyme recognition sequence this SNP. The effect of the SNP on levels of lymphocyte added, h: human, r: rat 123 Genes Nutr (2007) 2:225–232 227 B E X 40.5 nM) (Se-adequate cells) were added. Two days after 65nt the transfer to serum-free medium, cells were harvested for T/C 718 SNP RNA extraction and for measurement of IDI activity. 461nt 221nt RNA extraction and Northern hybridization IDI coding region GPx4 3’UTR T/C 718 Total RNA was extracted using Trizol reagent (Invitrogen; Paisley, UK) and following the manufacturer’s instruc- Fig. 1 Schematic diagram of IDI-GPx4 T/C 718 reporter constructs. tions. Northern blotting was carried out as described The T/C 718 SNP within the GPx4 3 UTR is indicated by a dotted line at the base of the SECIS element. The size of the 3 UTR is previously [3, 32]. The rat IDI probe was a 2.1 kb fragment indicated, as well as the distance of the SECIS element from the UGA excised with BstXI/NotI from IDI–IDI construct in codon for Se incorporation [data obtained from the NCBI database pCDNA3 [4] whereas the glyceraldehyde-3-phosphate (Accession number NM_002085)] and the restriction sites used for dehydrogenase (GAPDH) probe was a 780 bp PstI–XbaI cloning. U UGA codon for selenocysteine incorporation, S SECIS element, B BamHI, E EcoRI, X XbaI fragment obtained from human foetal liver as previously described [3]. Quantification of the bound probe was car- Directed Mutagenesis Kit (Stratagene, Amsterdam, The ried out using the Canberra Packard Instantimager (Perkin Netherlands) and the primers are shown in Table 1. All Elmer; Boston, USA), and results were expressed per unit constructs were confirmed by sequencing. of hybridisation achieved with the GAPDH cDNA. DNA transfection Type 1 iodothyronine deiodinase (IDI) activity assay Endotoxin-free pCDNA3.1/Zeo (+) constructs were stably IDI enzyme activity was measured in cells that had been transfected into human colon adenocarcinoma cells (Caco- washed and scraped in ice cold PBS (pH 7.4). Cells were 2) using Lipofectamine 2000 reagent (1 mg/ml; Invitrogen, collected by centrifugation at 10,000g for 15 min at 4C. Paisley, UK) according to the manufacturer’s instructions. The cell pellets were then resuspended in 200 ll phosphate Briefly, 2.5 · 10 cells were plated out in a 60 mm dish, buffer (0.125 M KPO /1 mM EDTA pH 7.4 containing and transfected after 1 day when they were 90–95% con- 0.1% (v/v) Triton X-100), and sonicated (10% power; fluent. The ratio of DNA (lg) to Lipofectamine 2000 (ll) Status 200, Philip Harris Scientific; Hyde, UK) three times used was 1:3.5. The day after transfection, cells were split for 10 s. Sonicated samples were centrifuged at 2,400g for at a 1:5 ratio and grown in normal growth medium for 10 min at 4C, and the supernatant fluids were used for the further 24 h. A selective medium containing 200 lg/ml of assay; 5 ll of cell lysate were incubated with [ I] reverse the antibiotic zeocin (Invitrogen; Paisley, UK) was added T3 (L-3,3’,5’-[ I]-triodothyronine (rT3) (1 lM) and to the cells 48 h after the start of the transfection. Stable 1 mM dithiothreitol (DTT) in phosphate buffer (0.125 M transfected colonies were identified after about 1 week but K PO /1 mM EDTA pH 7.4), in a final volume of 300 llat 3 4 were allowed to grow for an extra 2–3 weeks in the 37C for 30 min. The reaction was terminated, by precip- selective medium before a mixed population of transfected itating rT3, with the addition of heat-inactivated horse cells was obtained. serum and 10% (v/v) trichloroacetic acid (TCA). The free I released by the activity of IDI in both the supernatant fluid and the precipitate was counted using a gamma Cell culture conditions counter (Cobra counter, Canberra Instruments/Perkin Elmer; Boston, USA) and the release of I was quantified Stock of stable transfected Caco-2 cells were grown in as described by Arthur and co-workers [2]. Protein content normal Dulbecco’s Modified Eagle’s Medium (with 4.5 g/l of the cell lysates was measured by the Bicinchoninic acid glucose and Glutamax) supplemented with 10% (v/v) (BCA) method [40]. foetal calf serum (Sigma; Poole, UK), 60 lg/ml gentamy- cin, 1% (v/v) non-essential amino acids and penicillin/ streptomycin (100 units/ml and 100 lg/ml, respectively), Study subjects until they reached 80% confluence. Cells were then trans- ferred to serum-free medium to which either insulin (5 lg/ Participants in the study were recruited through two centres: ml) and transferrin (5 lg/ml) (Se-deficient cells), or insu- the Gastro-Intestinal Unit at Aberdeen Royal Infirmary, and lin, transferrin and selenium as sodium selenite (7 ng/ml; Newcastle and North Tyneside Health district. The study 123 228 Genes Nutr (2007) 2:225–232 was approved by the Grampian Ethics Committee and by the restriction enzyme Sty1 to generate either two frag- the Newcastle and North Tyneside Health Authorities Joint ments of 159 and 62 bp (C allele) or three fragments of 97, Ethics Committee. Written informed consent was obtained 62 and 62 bp (T allele). Genotype frequencies in patient from each study participant. Individuals attending clinic for groups were compared by a chi-squared test and Fischer’s colonoscopy were invited to participate in the study and, if exact test and odds ratios calculated using binary logistic consent was given, blood samples were obtained at the time regression and corrected for age and gender. of colonoscopy. A total number of 546 participants were recruited and identified as having colorectal adenocarci- noma, adenomatous polyps or no evidence of tumour or Results polyp at the time of colonoscopy. Subsequently, diagnosis was confirmed histologically. Reporter gene constructs and cell culture Incorporation of Sec into selenoproteins depends on the Genotyping SECIS structure within the 3 UTR and reporter gene studies show that the 3 UTR sequences from the various DNA was extracted from whole blood; 35 ml cell lysis selenoprotein mRNAs differ in their ability to allow Sec buffer pH 8 (10 mM Tris–HCl, 320 mM sucrose, 5 mM incorporation [4, 6, 44]. To study the effect of the T/C MgCl , 1% (v/v) Triton X-100) was added to 5–10 ml variation at position 718 in the GPx4 3 UTR, reporter gene blood and centrifuged at 900g for 10 min at 4C. The pellet constructs were made (Fig. 1) in which the rat type 1 IDI was then resuspended in 2 ml of nuclear lysis buffer coding region was linked to either the T or C variant of the (400 mM Tris–HCl, 60 mM EDTA, 150 mM NaCl, 1% human GPx4 3 UTR, and stable cell populations expressing (w/v) SDS), 0.5 ml of 5 M sodium perchlorate added and the constructs were generated in Caco-2 cells by transfec- the sample mixed at room temperature for 15 min and then tion. IDI is not expressed at detectable levels in Caco-2 incubated at 65C for 30 min; 2.5 ml of chloroform were cells and therefore using IDI as a reporter linked to the then added, the sample mixed for 10 min and the aqueous GPx4 3 UTR variants rather than over-expressing the full phase separated by centrifugation at 250g for 10 min at GPx4 transcripts avoided any complications due to 4C. The aqueous phase containing the DNA was trans- endogenous transcripts. ferred to a fresh tube, two volumes of 100% (v/v) ethanol Levels of the IDI mRNA and its enzyme activity were were added and the sample gently rotated. The DNA pre- then measured to assess Se incorporation into active sele- cipitated to form a white pellet, which was spooled with a noprotein. Untransfected cells did not express IDI as microbiological loop. The DNA was air dried for 10 min judged by Northern hybridisation or enzyme activity (data and resuspended in 200 ll of 1 mM Tris–HCl pH 8, not shown). Cells expressing the chimaeric constructs in overnight at 60C. For a small number of samples, only a which the T or C GPx4 variants were linked to the IDI limited amount of blood (\1 ml) was available. For these coding region exhibited closely similar levels of IDI tran- samples, DNA was extracted using the QIAamp DNA scripts (Table 2). However, the cells expressing IDI linked Blood Mini kit (Qiagen; Crawley, UK), according to the to the C variant 3 UTR showed a considerably higher (3.8- manufacturers’ instructions. fold) IDI reporter activity when the cells were grown under One hundred nanogram of DNA was subjected to PCR Se-adequate conditions (Table 2). Furthermore, the ratio of to amplify the 221 bp fragment corresponding to the activity/mRNA, a measure of Sec incorporation during 3 UTR of the human GPx4 RNA (Acc. N. NCBI: translation that also accounts for any differences in level of NM_002085; nt 654–874). The PCR primers used are mRNA expression [4, 5], was 2.7-fold higher with the C shown in Table 1 and conditions were as follows: one variant than with the T in cells grown under Se-adequate cycle at 94C for 2 min, 53.5C for 30 s, 72C for 1 min, culture conditions. This suggests that the C variant is more followed by 30 cycles at 94C for 30 s, 53.5C for 30 s and effective in incorporating Se under Se-adequate conditions. 72C for 1 min and a final elongation step at 72C for When cells were grown under Se-deficient conditions [4], 7 min. PCR products were purified by adsorption to a the levels of IDI mRNA did not change but IDI activity fell polyethylene glycol (PEG) mix (26.2% (w/v) PEG 8000, by 69 and 78% in IDI-GPx4 T and in IDI-GPx4 C trans- 6.6 mM MgCl , 0.6 M Sodium Acetate pH 5.2), washed fected cells, respectively (P \ 0.05, Mann–Whitney U test, with 70% (v/v) ethanol, air-dried and resuspended in 18 ll Table 2). Under these conditions IDI activity was still H O. The integrity of the PCR product was confirmed by greater in the cells expressing the C variant but the dif- agarose gel electrophoresis and the T/C variation at posi- ference between T and C variants was less marked. Indeed, tion 718 determined by direct sequencing (MWG Biotech; ratio of IDI activity in Se-deficient conditions compared Ebersberg, Germany) or restriction fragment analysis using with that in Se-adequate conditions was significantly 123 Genes Nutr (2007) 2:225–232 229 Table 2 Type 1 iodothyronine deiodinase (IDI) reporter studies to assess the effect of T to C mutation at position 718 in GPx4 3 UTR Cell line IDI mRNA levels (mean ± SEM) n = 4 IDI enzyme activity (mean ± SEM) n = 4 Ratio activity/mRNA levels (mean) +Se –Se +Se –Se +Se –Se IDI-GPx4 T 1.01 ± 0.27 0.93 ± 0.16 46.88 ± 3.00 14.43 ± 0.82 46.4 15.5 IDI-GPx4 C 1.44 ± 0.24 1.32 ± 0.14 177.90 ± 19.00 39.80 ± 5.30 123.4 30.0 Caco-2 cells were stably transfected with constructs in which the IDI coding region was linked to either the T or C variant of the GPx4 3 UTR. Cells were grown under Se-adequate or Se-deficient conditions as described in the text. RNA was extracted for measurement of IDI mRNA levels by Northern hybridisation and protein extracts prepared for assay of IDI activity. IDI mRNA levels are expressed in arbitrary units and activity expressed as pmol iodine released/min/mg protein. Se deficiency did not induce any changes in the levels of IDI mRNA abundance for the two different 3 UTRs (T or C). IDI activity levels decreased significantly (P \ 0.05) in Se-deficient conditions. An account of the translational efficiency (selenocysteine read-through) of the different 3 UTRs used (ratio of activity/mRNA abundance) [4, 5] is shown in the last column. Statistical analysis was performed using Mann–Whitney U test (P \ 0.05) smaller in the cells expressing the C variant adenomatous polyps was similar to that of the control group. (mean ± SEM 0.22 ± 0.01) compared with cells express- In contrast, in the group with adenocarcinoma, TT individ- ing the T variant (0.31 ± 0.01). These data show that the uals represented a smaller proportion (17%), and those with change from T to C at position 718 in the GPx4 3 UTR CC a larger proportion (30%), of the cohort. Indeed, statis- altered the ability of the 3 UTR to drive Se incorporation in tical analysis using a chi square test showed that the a reporter protein and that, regardless of Se conditions, the frequency of the CC genotype was significantly higher in the C variant led to greater Se incorporation, although the cancer group compared with the control group (P \ 0.05), expression driven by the T variant is relatively less sensi- and higher in the cancer group compared with the group with tive to Se depletion. adenomatous polyps (P \ 0.05). A multivariable analysis indicated that the odds ratio for the protective effect of carriage of one or more T alleles at the GPx4 T/C 718 SNP, Clinical study independent of age and gender, was 0.60 (95% CI 0.37–0.96, P = 0.033) in the cancer group compared with controls and A total of 546 participants took part to the study: 252 with 0.54 (95% CI 0.30–0.96, P = 0.035) in the cancer group adenocarcinoma, 107 with adenomatous polyps and 187 compared with the polyp group. There was no statistically controls. The three groups were comparable with respect to significant difference between genotype frequencies in the mean age and the control and cancer groups were similar in control and polyp groups. In view of the reporter gene data gender distribution (Table 3). A 221 bp sequence of the indicating functionality of the SNP and a lack of haplotype GPx4 gene corresponding to the 3 UTR was amplified by and other functional SNP data on the GPx4 gene, the focus of PCR from DNA extracted from blood taken from all the the study was restricted to the GPx4 T/C 718 SNP rather than participants. Sequencing of the PCR products showed a T/C extending it to include other variants by using a tagging or variation at position 718, confirming the presence of the SNP haplotype approach. previously identified at this position in a group of healthy volunteers [42]. The frequency distribution of the T/C 718 SNP in the control group showed a high number of hetero- Discussion zygotes (60%) with both homozygotes being relatively common (20% CC, 20% TT) (Table 3). The distribution of The T/C allelic variants at position 718 investigated in the the three genotypes in the group of patients with present study occur in a region of the GPx4 gene that Table 3 Distribution of GPx4 T/C 718 genotype among patients with cancer and polyps, and controls GPx4 T/C 718 genotype Cancer (n = 252) Polyp (n = 107) Control (n = 187) CC n = 76 (30%)* n = 20 (19%) n = 38 (20%) CT n = 132 (53%) n = 60 (56%) n = 111 (60%) TT n = 44 (17%) n = 27 (25%) n = 38 (20%) Age (mean ± SD) 68 ± 10 64 ± 6 63 ± 13 Gender distribution (% of males:females) 55:45 78:22 51:49 Results are expressed as number of individuals of each specific genotype with the % frequency in parentheses. Groups were compared by a chi- squared test and *indicates a statistically significant difference from controls and polyp groups (both P \ 0.05) 123 230 Genes Nutr (2007) 2:225–232 Fig. 2 Prediction of the secondary structure of the region containing Watson–Crick quartet), as well as the same number of stems and the SECIS element in the GPx4 3 UTR. The C (a) and T (b) variants loops in the core of the SECIS but there were differences in the region of the polymorphism at position 718 (indicated by arrow) were close to the SECIS. Prediction patterns were obtained using the predicted to result into two patterns of the RNA folding. Both SECIS-search secondary structure prediction program [30] predicted structures contain the same consensus sequence (Non- 0 0 corresponds to the 3 UTR of the mRNA [42] and, notably, and that two SNPs within the 3 UTR of the 15 kDa sele- is close to the base of the selenocysteine insertion sequence noprotein influence the ability of the 3 UTR to support Se (SECIS) element (Fig. 2), a region of the mRNA required incorporation [22]. for incorporation of the amino acid Sec into GPx4. Higher In addition, the present data also provide evidence for a reporter activity was observed in cells transfected with the link between the SNP T/C 718 and disease risk. Since IDI-GPx4 C variant and grown under either Se-adequate or GPx4 has been implicated in the regulation of leukotriene Se-deficient conditions compared with the corresponding metabolism [8, 25] and individuals of CC and TT genotype activities observed with the T variant. However, the for the GPx4 T/C 718 SNP have been reported to have activity driven by T variant was significantly less sensitive different levels of lymphocyte 5-lipoxygenase products to Se depletion. This may reflect the much lower starting [42], it is possible that the SNP has functional conse- level for activity for the T variant in Se-adequate condi- quences in vivo and it may cause differences in tions but it suggests that, in vivo, the C variant leads to inflammatory responses. The lower frequency of TT GPx4 synthesis that is greater at optimal Se intake but that genotypes for the GPx4 T/C 718 SNP observed in patients is more sensitive to Se depletion. In contrast, there was no suffering from ulcerative colitis [33] and the evidence for a significant difference between the mRNA levels of IDI- link between longstanding ulcerative colitis and colon GPx4 T 718 and IDI-GPx4 C 718 constructs in Se-deficient cancer [36, 38] led us to investigate the frequency of this and Se-adequate conditions, suggesting that any effect of particular SNP, in patients diagnosed with CRC or ade- the 3 UTR on Se incorporation occurs at a translational, nomatous polyps and healthy controls. The frequency of rather than transcriptional level. The present data, there- the TT, TC and CC variations at this position was found to fore, indicate that the T and C variants of the GPx4 3 UTR be different in patients with CRC from those patients with differ in their ability to drive synthesis of a reporter adenomatous polyps or controls with no colo-rectal neo- selenoprotein. plasia, suggesting that the risk of CRC, but not the risk of These results are the first direct evidence that the SNP T/ having adenomatous polyps, is affected by genotype at the C 718 in the GPx4 3 UTR has functional consequences. T/C 718 SNP in the GPx4 gene. It is possible that the GPx4 They are supported by computational analysis that suggests T/C 718 SNP causes differences in inflammatory responses that there may be structural differences around the two core in the colon, responses that have been suggested to affect SECIS elements (Fig. 2) in the GPx4 T and GPx4 C progression of polyps to tumour formation [26, 28]. 0 0 3 UTRs. The ability of a small change in 3 UTR sequence However, the link between the C variant, possibly to affect Se incorporation is compatible with previous higher GPx4 activity and higher levels of lipoxygenase studies showing that the efficiency of SECIS elements to products and disease risk has not been defined. One pos- incorporate Se differs between selenoproteins in vitro [29] sibility is that alterations in GPx4 lead to changes in 5- and 123 Genes Nutr (2007) 2:225–232 231 15-lipoxygenase activity, cycloxygenase-2 activity and provides evidence that genotype at this SNP is associated activation of NF-jB pathways [7, 21, 37), reflecting the with altered risk of CRC. Our hypothesis is that a combi- role of GPx4 as a regulator of inflammatory responses. A nation of sub-optimal Se status and a particular genotype at second possible mechanism is based on the hierarchy of the GPx4 T/C 718 SNP increases susceptibility to CRC but selenoprotein synthesis in which selenoproteins essentially not to the risk of developing adenomatous polyps. Further compete for available Se [6] and, on such a model, work is required to define the link between GPx4 T/C 718 increased synthesis of GPx4 due to the C variant would be variants, selenoprotein synthesis and cell function. expected to cause a withdrawal of Se from synthesis of Acknowledgments We thank the World Cancer Research Fund for other selenoproteins such as GPx1 and so lead to an altered financial support (grants 2000/10, 2002/41), clinical staff for assis- pattern of overall selenoprotein synthesis. The observed tance with sampling and the patients for consenting to take part in this colitis in the combined GPx1/GPx2 knock-out mice where study. We greatly acknowledge Brian Burtle, Rebecca Lamb, Leanne Boulding and Rachel Ledward for help with genotyping and Fergus GPx1 has a minor role in counteracting inflammation [16, Nicol for help with IDI activity assay. JRA’s lab is funded by The 17) is compatible with this hypothesis. Furthermore, such Scottish Executive Environment and Rural Affairs Department (SE- indirect effects, through changes in the selenoprotein ERAD). Conflict of interest-None declared. hierarchy, not only could account for the link between the C variant and disease risk but are also consistent with the References observed beneficial effect of increased Se intake on CRC risk [27]. However, a third possibility is that high GPx4 1. Akbaraly NT, Arnaud J, Hininger-Favier I, Gourlet V, Roussel expression may lead directly to lower apoptosis [31]. AM, Berr C (2005) Selenium and mortality in the elderly: results Several lines of evidence suggest that low Se status from the EVA study. Clin Chem 51:2117–2123 increases risk of mortality from CRC. Firstly, colon cancer 2. Arthur JR, Nicol F, Beckett GJ (1990) Hepatic type 1 iodothyr- onine 5 -deiodinase. The role of selenium. Biochem J 272:537– mortality has been inversely correlated with estimated Se intake [12]. Secondly, low Se intake or low plasma Se is 3. Bermano G, Nicol F, Dyer JA, Sunde RA, Beckett GJ, Arthur JR, associated with increased risk of CRC [12, 35]. Thirdly, a Hesketh JE (1995) Tissue-specific regulation of selenoenzyme gene expression during selenium deficiency in rats. Biochem J recent meta-analysis indicates that higher Se status is 311:425–430 associated with lower risk of a recurrence of colonic 4. Bermano G, Arthur JR, Hesketh JE (1996) Role of the 3 tumours [27]. Finally, a clinical trial indicates that Se untranslated region in the regulation of cytosolic glutathione supplementation reduces CRC mortality rates [10]. In view peroxidase and phospholipid-hydroperoxide glutathione peroxi- dase gene expression by selenium supply. Biochem J 320:891– of these reported effects of low Se status on CRC risk [12, 27, 35, 43], it is possible that the impact of genotype at the 6. Berry MJ (2005) Insights into the hierarchy of selenium incor- 718 SNP is modulated by Se status. Plasma Se and eryth- poration. Nat Genet 37:1162–1163 rocyte GPx1 activity are commonly used measures of 5. Berry MJ, Banu L, Chen YY, Mandel SJ, Kieffer JD, Harney JW, functional Se status [14]. Interestingly, preliminary data Larsen PR (1991) Recognition of UGA as a selenocysteine codon in type I deiodinase requires sequences in the 3 untranslated indicate that in a sub-group of the cancer patients studied region. Nature 353:273–276 here both plasma Se levels and erythrocyte GPx1 activity 8. Brigelius-Flohe R (1999) Tissue-specific functions of individual were significantly lower in individuals with CC genotype glutathione peroxidases. Free Radic Biol Med 27:951–965 (mean ± SEM plasma Se = 0.80 ± 0.06 lmol/l, mean ± 7. Brigelius-Flohe R, Friedrichs B, Maurer S, Schultz M, streicher R (1997) Interleukin-1-induced nuclear factor kappa B activation is SEM GPx1 activity = 4.5 ± 0.5 U/ml) than in individuals inhibited by overexpression of phospholipid hydroperoxide glu- with CT or TT genotype (mean plasma Se = 0.99 ± 0.04 tathione peroxidase in a human endothelial cell line. Biochem J and 1.00 ± 0.05 lmol/l, respectively and mean GPx1 328:199–203 activity = 5.3 ± 0.2 and 6.1 ± 0.7 U/ml respectively). In 9. Clark LC, Hixson LJ, Combs GF, Reid ME, Turnbull BW, Sampliner RE (1993) Plasma selenium concentration predicts the contrast, in the polyp group, plasma Se concentration and prevalence of colorectal adenomatous polyps. Cancer Epidemiol erythrocyte GPx1 activity were not significantly affected Biomark Prev 2:41–46 by genotype. The observed differences in plasma Se and 10. Clark LC, Combs GF Jr, Turnbull BW, Slate EH, Chalker DK, Chow J, Davis LS, Glover RA, Graham GF, Gross EG, Krongrad erythrocyte GPx1 activity with genotype and clinical A, Lesher JL Jr, Park HK, Sanders BB Jr, Smith CL, Taylor JR diagnosis suggest that the combination of Se status and (1996) Effects of selenium supplementation for cancer prevention GPx4 T/C 718 genotype affects development of colorectal in patients with carcinoma of the skin. A randomized controlled cancer but not the risk of developing polyps. This is con- trial. Nutritional Prevention of Cancer Study Group. J Am Med Assoc 276:1957–1963 sistent with the previous suggestion that patients with 12. Combs GF (2005) Current evidence and research needs to support polyps and the lowest Se levels are at higher risk of a health claim for selenium and cancer prevention. J Nutr malignant transformation [15]. 135:343–347 In summary, the present study demonstrates the func- 11. Combs GF Jr, Gray WP (1998) Chemopreventive agents: sele- tionality of the SNP T/C 718 in the GPx4 gene and nium. Pharmacol Ther 79:179–192 123 232 Genes Nutr (2007) 2:225–232 13. Cummings OW (2000) Pathology of the adenoma–carcinoma 30. Kryukov GV, Kryukov VM, Gladyshev VN (1999) New mam- sequence: from aberrant crypt focus to invasive carcinoma. Se- malian selenocysteine-containing proteins identified with an min Gastrointest Dis 11:229–237 algorithm that searches for selenocysteine insertion sequence 14. Diplock AT (1993) Indexes of selenium status in human popu- elements. J Biol Chem 274:33888–33897 lations. Am J Clin Nutr 57:256S–258S 31. Nomura K, Imai H, Koumura T, Arai M, Nakagawa Y (1999) 15. Dworkin BM, Rosenthal WS, Mittelman A, Weiss L, Applebee- Mitochondrial phospholipid hydroperoxide glutathione peroxi- Brady L, Arlin Z (1988) Selenium status and the polyp-cancer dase suppresses apoptosis mediated by a mitochondrial death sequence: a colonoscopy controlled study. Am J Gastroenterol pathway. J Biol Chem 274:22294–29302 83:748–751 32. Pagmantidis V, Bermano G, Villette S, Broom J, Arthur J, 16. Esworthy RS, Aranda R, Martin MG, Doroshow JH, Binder SW, Hesketh JE (2005) Effects of Se-depletion on glutathione per- Chu FF (2001) Mice with combined disruption of Gpx1 and Gpx2 oxidase and SelW gene expression in the colon. FEBS Lett genes have colitis. Am J Physiol Gastrointest Liver Physiol 579:792–796 281:G848–55 33. Qatatsheh A, Seal CJ, Jowett SL, Welfare MR, Hesketh JE (2005) 17. Esworthy RS, Yang L, Frankel PH, Chu FF (2005) Epithelium- Patients with ulcerative colitis show an altered frequency distri- specific glutathione peroxidase, Gpx2, is involved in the pre- bution of a single nucleotide polymorphism in the gene encoding vention of intestinal inflammation in selenium-deficient mice. J the phospholipid hydroperoxide glutathione peroxidase. Pro- Nutr 135:740–745 ceedings of the Nutrition Society 64, OCA, 20A (Abstract) 18. Forsberg L, de Faire U, Morgenstern R (1998) To identify genetic 34. Rayman MP (2005) Selenium in cancer prevention: a review of polymorphisms in the ‘‘Expressed Sequence Tag’’ (EST) data- the evidence and mechanism of action. Proc Nutr Soc 64:527– base. Technical tips on-line http://www.elsevier.com_/locate_/tto 542 , T01440 35. Russo MW, Murray SC, Wurzelmann JI, Woosley JT, Sandler RS 19. Forsberg L, de Faire U, Marklund SL, Andersson PM, Stegmayr (1997) Plasma selenium levels and the risk of colorectal adeno- B, Morgenstern R (2000) Phenotype determination of a common mas. Nutr Cancer 28:125–129 Pro-Leu polymorphism in human glutathione peroxidase 1. Blood 36. Rutter M, Saunders B, Wilkinson K, Rumbles S, Schofield G, Cells Mol Dis 26:423–426 Kamm M, Williams C, Price A, Talbot I, Forbes A (2004) 20. Haklar G, Sayin-Ozveri E, Yuksel M, Aktan AO, Yalcin AS Severity of inflammation is a risk factor for colorectal neoplasia (2001) Different kinds of reactive oxygen and nitrogen species in ulcerative colitis. Gastroenterology 126:451–459 were detected in colon and breast tumors. Cancer Lett 165:219– 37. Schnurr K, Belkner J, Ursini F, Schewe T, Kuhn H (1996) The 224 selenoenzyme phospholipid hydroperoxide glutathione peroxi- 21. Heirman I, Ginneberge D, Brigelius-Flohe R, Hendrickx N, Ag- dase controls the activity of the 15-lipoxygenase with complex ostinis P, Brouckaert P, Rottiers P. And Grooten J (2006) substrates and preserves the specificity of the oxygenation prod- Blocking tumor cell eicosanoid synthesis by GPx4 impedes tumor ucts. J Biol Chem 271:4653–4658 growth and malignancy. Free Radic Biol Med 40:285–94 38. Seril DN, Liao J, Yang GY, Yang CS (2003) Oxidative stress and 23. Hu YJ, Diamond AM (2003) Role of glutathione peroxidase 1 in ulcerative colitis-associated carcinogenesis: studies in humans breast cancer: loss of heterozygosity and allelic differences in the and animal models. Carcinogenesis 24:353–362 response to selenium. Cancer Res 63:3347–3351 39. Shamberger RJ, Tytko SA, Willis CE (1976) Antioxidants and 22. Hu YJ, Korotkov KV, Mehta R, Hatfield DL, Rotimi CN, Luke cancer. Part VI. Selenium and age-adjusted human cancer mor- A, Prewitt TE, Cooper RS, Stock W, Vokes EE, Dolan ME, tality. Arch Environ Health 31:231–235 Gladyshev VN, Diamond AM (2001) Distribution and functional 40. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, consequences of nucleotide polymorphisms in the 3 -untranslated Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC region of the human Sep15 gene. Cancer Res 61:2307–2310 (1985) Measurement of protein using bicinchoninic acid. Ann 24. Ichimura Y, Habuchi T, Tsuchiya N, Wang L, Oyama C, Sato K, Biochem 150:76–85 Nishiyama H, Ogawa O, Kato T (2004) Increased risk of bladder 41. Stadtman TC (1996) Selenocysteine. Ann Rev Biochem 65:83– cancer associated with a glutathione peroxidase 1 codon 198 100 variant. J Urol 172:728–732 42. Villette S, Kyle JA, Brown KM, Pickard K, Milne JS, Nicol F, 25. Imai H, Nakagawa Y (2003) Biological significance of phos- Arthur JR, Hesketh JE (2002) A novel single nucleotide poly- pholipid hydroperoxide glutathione peroxidase (PHGPx, GPx4) morphism in the 3 untranslated region of human glutathione in mammalian cells. Free Radic Biol Med 34:145–169 peroxidase 4 influences lipoxygenase metabolism. Blood Cells 26. Itzkowitz SH, Yio X (2004) Inflammation and cancer IV. Colo- Mol Dis 29:174–178 rectal cancer in inflammatory bowel disease: the role of 43. Whanger PD (2004) Selenium and its relationship to cancer: an inflammation. Am J Physiol Gastrointest Liver Physiol 287:G7– update dagger. Br J Nutr 91:11–28 G17 44. Wingler K, Bo ¨ cher M, Flohe L, Kollmus H, Brigelius-Flohe R 27. Jacobs ET, Jiang R, Alberts DS, Greenberg ER, Gunter EW, (1999) mRNA stability and selenocysteine insertion sequence Karagas MR, Lanza E, Ratnasinghe L, Reid ME, Schatzkin A, efficiency rank gastrointestinal glutathione peroxidase high in the Smith-Warner SA, Wallace K, Martinez ME (2004) Selenium hierarchy of selenoproteins. Eur J Biochem 259:149–157 and colorectal adenoma: results of a pooled analysis. J Natl 45. Yagi K, Komura S, Kojima H, Sun Q, Nagata N, Ohishi N, Cancer Instit 96:1669–1675 Nishikimi M (1996) Expression of human phospholipid hydro- 28. Jacobson-Brown P, Neuman MG (2004) Colorectal polyposis and peroxide glutathione peroxidase gene for protection of host cells immune-based therapies. Canad J Gastroenterol 18:239–249 from lipid hydroperoxide-mediated injury. Biochem Biophys Res 29. Kollmus H, Flohe L, McCarthy JE (1996) Analysis of eukaryotic Commun 219:486–491 mRNA structures directing cotranslational incorporation of selenocysteine. Nucleic Acids Res 24:1195–1201 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Genes & Nutrition Springer Journals

Evidence that a polymorphism within the 3′UTR of glutathione peroxidase 4 is functional and is associated with susceptibility to colorectal cancer

Download PDF
8 pages

Loading next page...
 
/lp/springer-journals/evidence-that-a-polymorphism-within-the-3-utr-of-glutathione-0h1hZpvvb1

References (52)

Publisher
Springer Journals
Copyright
Copyright © 2007 by Springer-Verlag
Subject
Biomedicine; Nutrition; Biomedicine general; Human Genetics
ISSN
1555-8932
eISSN
1865-3499
DOI
10.1007/s12263-007-0052-3
pmid
18850177
Publisher site
See Article on Publisher Site

Abstract

Genes Nutr (2007) 2:225–232 DOI 10.1007/s12263-007-0052-3 RESEARCH PAPER Evidence that a polymorphism within the 3 UTR of glutathione peroxidase 4 is functional and is associated with susceptibility to colorectal cancer G. Bermano Æ V. Pagmantidis Æ N. Holloway Æ S. Kadri Æ N. A. G. Mowat Æ R. S. Shiel Æ J. R. Arthur Æ J. C. Mathers Æ A. K. Daly Æ J. Broom Æ J. E. Hesketh Received: 24 November 2006 / Accepted: 18 January 2007 / Published online: 13 October 2007 Springer-Verlag 2007 Abstract Low selenium (Se) status has been associated study was carried out in cohorts of patients with either with increased risk of colorectal cancer (CRC). Se is adenomatous polyps, colorectal adenocarcinomas and in present as the amino acid selenocysteine in selenoproteins, healthy controls. A higher proportion of individuals with such as the glutathione peroxidases. Se incorporation CC genotype at the GPx4 T/C 718 SNP was present in the requires specific RNA structures in the 3 untranslated cancer group, but not in the polyp group, compared with region (3 UTR) of the selenoprotein mRNAs. A single the control group (P \ 0.05). The present data demonstrate nucleotide polymorphism (SNP) occurs at nucleotide 718 the functionality of the GPx4 T/C 718 SNP and suggest (within the 3 UTR) in the glutathione peroxidase 4 gene. In that T genotype is associated with lower risk of CRC. the present study, Caco-2 cells were transfected with constructs in which type 1 iodothyronine deiodinase coding Keywords Colorectal cancer  GPx4  Reporter gene 0 0 region was linked to the GPx4 3 UTR with either C or T Selenium  SNP  3 Untranslated region variant at position 718. Higher reporter activity was observed in cells expressing the C variant compared to those expressing the T variant, under either Se-adequate or Introduction Se-deficient conditions. In addition, a disease association Selenium (Se) is a dietary micronutrient that is essential for human health [12, 34] and low Se status has been associ- ated with increased mortality [1]. Worldwide, mortality G. Bermano and V. Pagmantidis have contributed equally to this work. G. Bermano  N. Holloway  J. Broom J. C. Mathers School of Life Sciences, The Robert Gordon University, School of Clinical Medical Sciences and Human Nutrition Aberdeen AB25 1HG, UK Research Centre, Newcastle University, Newcastle upon Tyne NE2 4HH, UK V. Pagmantidis Institute for Cell and Molecular Biosciences, J. E. Hesketh (&) Newcastle University, Institute for Cell and Molecular Biosciences and Human Newcastle upon Tyne NE2 4HH, UK Nutrition Research Centre, Newcastle University, Newcastle upon Tyne NE2 4HH, UK S. Kadri  N. A. G. Mowat e-mail: [email protected] Aberdeen Royal Infirmary, Aberdeen AB25 2ZN, UK A. K. Daly R. S. Shiel School of Clinical and Laboratory Sciences, School of Agriculture and Rural Development, Newcastle University, Newcastle upon Tyne NE2 4HH, UK Newcastle University, Newcastle upon Tyne NE1 7RU, UK J. R. Arthur Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK 123 226 Genes Nutr (2007) 2:225–232 from cancers, including colorectal cancer (CRC), is lipoxygenase products implies that the SNP may affect inversely correlated with estimated Se intake [11]. Addi- inflammatory signalling, and, since inflammation in the tionally in case-control studies, low Se intake is associated colon is regarded as a frequent initiating event in carci- with increased risk of CRC [35] and, in a 10-year pro- nogenesis [26, 28], the second objective was to carry out a spective study, initial plasma Se was inversely correlated disease-association study focusing on the GPx4 T/C 718 with subsequent risk of CRC [9]. Moreover, a clinical trial SNP in patients with CRC and in controls. The data indi- in the United States indicated that a daily supplement of cate that the T/C 718 SNP has functional effects on Se 200 lg Se led to reduced mortality from CRC. Carcino- incorporation and that the C variant is more common in genesis in the colon is a multistage process and CRC patients diagnosed with CRC. develops via an adenoma–carcinoma sequence from microscopic dysplastic aberrant crypt foci within the epi- thelium to benign adenomatous polyps, which can then Materials and methods potentially, but not necessarily, undergo malignant change. The sequence of biochemical changes underlying this Plasmid construction sequence is not known but oxidant damage to DNA is a critical event in tumour formation [13, 20, 39]. Se is nec- Constructs containing the rat type 1 iodothyronine deio- essary for the function of *25 selenoproteins, including dinase (IDI) coding region linked to GPx4 3 UTR with the antioxidant enzymes cytosolic glutathione peroxidase either a T or C at position 718, were cloned to assess (GPx1), gastrointestinal specific glutathione peroxidase functionality of the T/C polymorphism in the human GPx4 (GPx2) and phospholipid hydroperoxide glutathione per- 3 UTR at position 718 [42]. The rat IDI coding region was oxidase (GPx4) [8, 41] which provide defence against amplified by PCR from IDI–IDI construct in pCDNA3 [4] oxidative damage. and the human GPx4 3 UTR (T variant) was amplified In selenoproteins, Se is present as the amino-acid sele- from pRc/CMV [45]. PCR primers used for the amplifi- nocysteine (Sec) and its incorporation occurs during cation reactions are shown in Table 1. Following ligation synthesis of the proteins. This incorporation requires the of the rat IDI coding region (BamHI/EcoRI) to the human recoding of a UGA codon that involves the formation of a GPx4 3 UTR (EcoRI/XbaI) (Fig. 1), the product was complex between an RNA stem-loop structure (SECIS) cloned into pCDNA3.1/Zeo (+) (Invitrogen, Paisley, UK) 0 0 located in the 3 untranslated region (3 UTR) of seleno- for expression in mammalian cells. PCR products were protein mRNAs and various trans-acting proteins [6]. Thus purified using QIAquick PCR purification kit (Qiagen; selenoprotein expression can depend on both dietary Se Crawley, UK) according to the manufacturers’ instructions. intake and genetic factors. Furthermore, variants in genes The construct in which the IDI coding region was linked to encoding any of the GPxs affecting antioxidant capacity, the C variant (position 718) of the GPx4 3 UTR was made may change susceptibility to genomic damage by oxidants by site-directed mutagenesis using QuickChange Site- and thus influence neoplastic change. A SNP has been identified in the GPx1 gene which results in a proline to Table 1 Primer sequences used for plasmid construction, mutagen- leucine substitution at codon 198 [18, 19] and this SNP is esis and genotyping associated with lung and bladder cancer risk [23, 24]. Two 0 0 SNPs within the Sep15 gene occur in a region of the 3 UTR Cloning Primers 5 –3 and, although the function of the encoded protein is not rIDI coding region known, there is an association of alleles at these two BamHI CGGGATCCATGGGGCTGTCCCAG positions with susceptibility to breast cancer [22]. Despite EcoRI CGGAATTCCTAGAACTGAGGCATGTGTG the report that Se supplementation reduces CRC mortality hGPx4 3 UTR [10], there is little information on whether SNPs in GPx EcoRI CGGAATTCCTCCACAAGTGTGTGGC genes affect susceptibility to CRC. XbaI GCTCTAGACTAATTTGTCTGTTTATTCCC A SNP (rs713041) has been found at position 718 of the Mutagenesis GPx4 gene mRNA in the 3 UTR region; the regulatory Sense CTGCCCACGCCCTCGGAGCCTTCCACC region required for incorporation of Se into selenoproteins. Antisense GGTGGAAGGCTCCGAGGGCGTGGGCAG The C allele is associated with increased levels of lym- Genotyping phocyte lipoxygenase products [42]. However, since the hGPx4 3 UTR GACCTGCCCCACTATTTCTA effects of the SNP on Se incorporation are not known, the GTCTGTTTATTCCCACAAGG first objective of this work was to use a reporter gene assay in transfected Caco-2 cells [4] to assess the functionality of X: Nucleotides added, X: restriction enzyme recognition sequence this SNP. The effect of the SNP on levels of lymphocyte added, h: human, r: rat 123 Genes Nutr (2007) 2:225–232 227 B E X 40.5 nM) (Se-adequate cells) were added. Two days after 65nt the transfer to serum-free medium, cells were harvested for T/C 718 SNP RNA extraction and for measurement of IDI activity. 461nt 221nt RNA extraction and Northern hybridization IDI coding region GPx4 3’UTR T/C 718 Total RNA was extracted using Trizol reagent (Invitrogen; Paisley, UK) and following the manufacturer’s instruc- Fig. 1 Schematic diagram of IDI-GPx4 T/C 718 reporter constructs. tions. Northern blotting was carried out as described The T/C 718 SNP within the GPx4 3 UTR is indicated by a dotted line at the base of the SECIS element. The size of the 3 UTR is previously [3, 32]. The rat IDI probe was a 2.1 kb fragment indicated, as well as the distance of the SECIS element from the UGA excised with BstXI/NotI from IDI–IDI construct in codon for Se incorporation [data obtained from the NCBI database pCDNA3 [4] whereas the glyceraldehyde-3-phosphate (Accession number NM_002085)] and the restriction sites used for dehydrogenase (GAPDH) probe was a 780 bp PstI–XbaI cloning. U UGA codon for selenocysteine incorporation, S SECIS element, B BamHI, E EcoRI, X XbaI fragment obtained from human foetal liver as previously described [3]. Quantification of the bound probe was car- Directed Mutagenesis Kit (Stratagene, Amsterdam, The ried out using the Canberra Packard Instantimager (Perkin Netherlands) and the primers are shown in Table 1. All Elmer; Boston, USA), and results were expressed per unit constructs were confirmed by sequencing. of hybridisation achieved with the GAPDH cDNA. DNA transfection Type 1 iodothyronine deiodinase (IDI) activity assay Endotoxin-free pCDNA3.1/Zeo (+) constructs were stably IDI enzyme activity was measured in cells that had been transfected into human colon adenocarcinoma cells (Caco- washed and scraped in ice cold PBS (pH 7.4). Cells were 2) using Lipofectamine 2000 reagent (1 mg/ml; Invitrogen, collected by centrifugation at 10,000g for 15 min at 4C. Paisley, UK) according to the manufacturer’s instructions. The cell pellets were then resuspended in 200 ll phosphate Briefly, 2.5 · 10 cells were plated out in a 60 mm dish, buffer (0.125 M KPO /1 mM EDTA pH 7.4 containing and transfected after 1 day when they were 90–95% con- 0.1% (v/v) Triton X-100), and sonicated (10% power; fluent. The ratio of DNA (lg) to Lipofectamine 2000 (ll) Status 200, Philip Harris Scientific; Hyde, UK) three times used was 1:3.5. The day after transfection, cells were split for 10 s. Sonicated samples were centrifuged at 2,400g for at a 1:5 ratio and grown in normal growth medium for 10 min at 4C, and the supernatant fluids were used for the further 24 h. A selective medium containing 200 lg/ml of assay; 5 ll of cell lysate were incubated with [ I] reverse the antibiotic zeocin (Invitrogen; Paisley, UK) was added T3 (L-3,3’,5’-[ I]-triodothyronine (rT3) (1 lM) and to the cells 48 h after the start of the transfection. Stable 1 mM dithiothreitol (DTT) in phosphate buffer (0.125 M transfected colonies were identified after about 1 week but K PO /1 mM EDTA pH 7.4), in a final volume of 300 llat 3 4 were allowed to grow for an extra 2–3 weeks in the 37C for 30 min. The reaction was terminated, by precip- selective medium before a mixed population of transfected itating rT3, with the addition of heat-inactivated horse cells was obtained. serum and 10% (v/v) trichloroacetic acid (TCA). The free I released by the activity of IDI in both the supernatant fluid and the precipitate was counted using a gamma Cell culture conditions counter (Cobra counter, Canberra Instruments/Perkin Elmer; Boston, USA) and the release of I was quantified Stock of stable transfected Caco-2 cells were grown in as described by Arthur and co-workers [2]. Protein content normal Dulbecco’s Modified Eagle’s Medium (with 4.5 g/l of the cell lysates was measured by the Bicinchoninic acid glucose and Glutamax) supplemented with 10% (v/v) (BCA) method [40]. foetal calf serum (Sigma; Poole, UK), 60 lg/ml gentamy- cin, 1% (v/v) non-essential amino acids and penicillin/ streptomycin (100 units/ml and 100 lg/ml, respectively), Study subjects until they reached 80% confluence. Cells were then trans- ferred to serum-free medium to which either insulin (5 lg/ Participants in the study were recruited through two centres: ml) and transferrin (5 lg/ml) (Se-deficient cells), or insu- the Gastro-Intestinal Unit at Aberdeen Royal Infirmary, and lin, transferrin and selenium as sodium selenite (7 ng/ml; Newcastle and North Tyneside Health district. The study 123 228 Genes Nutr (2007) 2:225–232 was approved by the Grampian Ethics Committee and by the restriction enzyme Sty1 to generate either two frag- the Newcastle and North Tyneside Health Authorities Joint ments of 159 and 62 bp (C allele) or three fragments of 97, Ethics Committee. Written informed consent was obtained 62 and 62 bp (T allele). Genotype frequencies in patient from each study participant. Individuals attending clinic for groups were compared by a chi-squared test and Fischer’s colonoscopy were invited to participate in the study and, if exact test and odds ratios calculated using binary logistic consent was given, blood samples were obtained at the time regression and corrected for age and gender. of colonoscopy. A total number of 546 participants were recruited and identified as having colorectal adenocarci- noma, adenomatous polyps or no evidence of tumour or Results polyp at the time of colonoscopy. Subsequently, diagnosis was confirmed histologically. Reporter gene constructs and cell culture Incorporation of Sec into selenoproteins depends on the Genotyping SECIS structure within the 3 UTR and reporter gene studies show that the 3 UTR sequences from the various DNA was extracted from whole blood; 35 ml cell lysis selenoprotein mRNAs differ in their ability to allow Sec buffer pH 8 (10 mM Tris–HCl, 320 mM sucrose, 5 mM incorporation [4, 6, 44]. To study the effect of the T/C MgCl , 1% (v/v) Triton X-100) was added to 5–10 ml variation at position 718 in the GPx4 3 UTR, reporter gene blood and centrifuged at 900g for 10 min at 4C. The pellet constructs were made (Fig. 1) in which the rat type 1 IDI was then resuspended in 2 ml of nuclear lysis buffer coding region was linked to either the T or C variant of the (400 mM Tris–HCl, 60 mM EDTA, 150 mM NaCl, 1% human GPx4 3 UTR, and stable cell populations expressing (w/v) SDS), 0.5 ml of 5 M sodium perchlorate added and the constructs were generated in Caco-2 cells by transfec- the sample mixed at room temperature for 15 min and then tion. IDI is not expressed at detectable levels in Caco-2 incubated at 65C for 30 min; 2.5 ml of chloroform were cells and therefore using IDI as a reporter linked to the then added, the sample mixed for 10 min and the aqueous GPx4 3 UTR variants rather than over-expressing the full phase separated by centrifugation at 250g for 10 min at GPx4 transcripts avoided any complications due to 4C. The aqueous phase containing the DNA was trans- endogenous transcripts. ferred to a fresh tube, two volumes of 100% (v/v) ethanol Levels of the IDI mRNA and its enzyme activity were were added and the sample gently rotated. The DNA pre- then measured to assess Se incorporation into active sele- cipitated to form a white pellet, which was spooled with a noprotein. Untransfected cells did not express IDI as microbiological loop. The DNA was air dried for 10 min judged by Northern hybridisation or enzyme activity (data and resuspended in 200 ll of 1 mM Tris–HCl pH 8, not shown). Cells expressing the chimaeric constructs in overnight at 60C. For a small number of samples, only a which the T or C GPx4 variants were linked to the IDI limited amount of blood (\1 ml) was available. For these coding region exhibited closely similar levels of IDI tran- samples, DNA was extracted using the QIAamp DNA scripts (Table 2). However, the cells expressing IDI linked Blood Mini kit (Qiagen; Crawley, UK), according to the to the C variant 3 UTR showed a considerably higher (3.8- manufacturers’ instructions. fold) IDI reporter activity when the cells were grown under One hundred nanogram of DNA was subjected to PCR Se-adequate conditions (Table 2). Furthermore, the ratio of to amplify the 221 bp fragment corresponding to the activity/mRNA, a measure of Sec incorporation during 3 UTR of the human GPx4 RNA (Acc. N. NCBI: translation that also accounts for any differences in level of NM_002085; nt 654–874). The PCR primers used are mRNA expression [4, 5], was 2.7-fold higher with the C shown in Table 1 and conditions were as follows: one variant than with the T in cells grown under Se-adequate cycle at 94C for 2 min, 53.5C for 30 s, 72C for 1 min, culture conditions. This suggests that the C variant is more followed by 30 cycles at 94C for 30 s, 53.5C for 30 s and effective in incorporating Se under Se-adequate conditions. 72C for 1 min and a final elongation step at 72C for When cells were grown under Se-deficient conditions [4], 7 min. PCR products were purified by adsorption to a the levels of IDI mRNA did not change but IDI activity fell polyethylene glycol (PEG) mix (26.2% (w/v) PEG 8000, by 69 and 78% in IDI-GPx4 T and in IDI-GPx4 C trans- 6.6 mM MgCl , 0.6 M Sodium Acetate pH 5.2), washed fected cells, respectively (P \ 0.05, Mann–Whitney U test, with 70% (v/v) ethanol, air-dried and resuspended in 18 ll Table 2). Under these conditions IDI activity was still H O. The integrity of the PCR product was confirmed by greater in the cells expressing the C variant but the dif- agarose gel electrophoresis and the T/C variation at posi- ference between T and C variants was less marked. Indeed, tion 718 determined by direct sequencing (MWG Biotech; ratio of IDI activity in Se-deficient conditions compared Ebersberg, Germany) or restriction fragment analysis using with that in Se-adequate conditions was significantly 123 Genes Nutr (2007) 2:225–232 229 Table 2 Type 1 iodothyronine deiodinase (IDI) reporter studies to assess the effect of T to C mutation at position 718 in GPx4 3 UTR Cell line IDI mRNA levels (mean ± SEM) n = 4 IDI enzyme activity (mean ± SEM) n = 4 Ratio activity/mRNA levels (mean) +Se –Se +Se –Se +Se –Se IDI-GPx4 T 1.01 ± 0.27 0.93 ± 0.16 46.88 ± 3.00 14.43 ± 0.82 46.4 15.5 IDI-GPx4 C 1.44 ± 0.24 1.32 ± 0.14 177.90 ± 19.00 39.80 ± 5.30 123.4 30.0 Caco-2 cells were stably transfected with constructs in which the IDI coding region was linked to either the T or C variant of the GPx4 3 UTR. Cells were grown under Se-adequate or Se-deficient conditions as described in the text. RNA was extracted for measurement of IDI mRNA levels by Northern hybridisation and protein extracts prepared for assay of IDI activity. IDI mRNA levels are expressed in arbitrary units and activity expressed as pmol iodine released/min/mg protein. Se deficiency did not induce any changes in the levels of IDI mRNA abundance for the two different 3 UTRs (T or C). IDI activity levels decreased significantly (P \ 0.05) in Se-deficient conditions. An account of the translational efficiency (selenocysteine read-through) of the different 3 UTRs used (ratio of activity/mRNA abundance) [4, 5] is shown in the last column. Statistical analysis was performed using Mann–Whitney U test (P \ 0.05) smaller in the cells expressing the C variant adenomatous polyps was similar to that of the control group. (mean ± SEM 0.22 ± 0.01) compared with cells express- In contrast, in the group with adenocarcinoma, TT individ- ing the T variant (0.31 ± 0.01). These data show that the uals represented a smaller proportion (17%), and those with change from T to C at position 718 in the GPx4 3 UTR CC a larger proportion (30%), of the cohort. Indeed, statis- altered the ability of the 3 UTR to drive Se incorporation in tical analysis using a chi square test showed that the a reporter protein and that, regardless of Se conditions, the frequency of the CC genotype was significantly higher in the C variant led to greater Se incorporation, although the cancer group compared with the control group (P \ 0.05), expression driven by the T variant is relatively less sensi- and higher in the cancer group compared with the group with tive to Se depletion. adenomatous polyps (P \ 0.05). A multivariable analysis indicated that the odds ratio for the protective effect of carriage of one or more T alleles at the GPx4 T/C 718 SNP, Clinical study independent of age and gender, was 0.60 (95% CI 0.37–0.96, P = 0.033) in the cancer group compared with controls and A total of 546 participants took part to the study: 252 with 0.54 (95% CI 0.30–0.96, P = 0.035) in the cancer group adenocarcinoma, 107 with adenomatous polyps and 187 compared with the polyp group. There was no statistically controls. The three groups were comparable with respect to significant difference between genotype frequencies in the mean age and the control and cancer groups were similar in control and polyp groups. In view of the reporter gene data gender distribution (Table 3). A 221 bp sequence of the indicating functionality of the SNP and a lack of haplotype GPx4 gene corresponding to the 3 UTR was amplified by and other functional SNP data on the GPx4 gene, the focus of PCR from DNA extracted from blood taken from all the the study was restricted to the GPx4 T/C 718 SNP rather than participants. Sequencing of the PCR products showed a T/C extending it to include other variants by using a tagging or variation at position 718, confirming the presence of the SNP haplotype approach. previously identified at this position in a group of healthy volunteers [42]. The frequency distribution of the T/C 718 SNP in the control group showed a high number of hetero- Discussion zygotes (60%) with both homozygotes being relatively common (20% CC, 20% TT) (Table 3). The distribution of The T/C allelic variants at position 718 investigated in the the three genotypes in the group of patients with present study occur in a region of the GPx4 gene that Table 3 Distribution of GPx4 T/C 718 genotype among patients with cancer and polyps, and controls GPx4 T/C 718 genotype Cancer (n = 252) Polyp (n = 107) Control (n = 187) CC n = 76 (30%)* n = 20 (19%) n = 38 (20%) CT n = 132 (53%) n = 60 (56%) n = 111 (60%) TT n = 44 (17%) n = 27 (25%) n = 38 (20%) Age (mean ± SD) 68 ± 10 64 ± 6 63 ± 13 Gender distribution (% of males:females) 55:45 78:22 51:49 Results are expressed as number of individuals of each specific genotype with the % frequency in parentheses. Groups were compared by a chi- squared test and *indicates a statistically significant difference from controls and polyp groups (both P \ 0.05) 123 230 Genes Nutr (2007) 2:225–232 Fig. 2 Prediction of the secondary structure of the region containing Watson–Crick quartet), as well as the same number of stems and the SECIS element in the GPx4 3 UTR. The C (a) and T (b) variants loops in the core of the SECIS but there were differences in the region of the polymorphism at position 718 (indicated by arrow) were close to the SECIS. Prediction patterns were obtained using the predicted to result into two patterns of the RNA folding. Both SECIS-search secondary structure prediction program [30] predicted structures contain the same consensus sequence (Non- 0 0 corresponds to the 3 UTR of the mRNA [42] and, notably, and that two SNPs within the 3 UTR of the 15 kDa sele- is close to the base of the selenocysteine insertion sequence noprotein influence the ability of the 3 UTR to support Se (SECIS) element (Fig. 2), a region of the mRNA required incorporation [22]. for incorporation of the amino acid Sec into GPx4. Higher In addition, the present data also provide evidence for a reporter activity was observed in cells transfected with the link between the SNP T/C 718 and disease risk. Since IDI-GPx4 C variant and grown under either Se-adequate or GPx4 has been implicated in the regulation of leukotriene Se-deficient conditions compared with the corresponding metabolism [8, 25] and individuals of CC and TT genotype activities observed with the T variant. However, the for the GPx4 T/C 718 SNP have been reported to have activity driven by T variant was significantly less sensitive different levels of lymphocyte 5-lipoxygenase products to Se depletion. This may reflect the much lower starting [42], it is possible that the SNP has functional conse- level for activity for the T variant in Se-adequate condi- quences in vivo and it may cause differences in tions but it suggests that, in vivo, the C variant leads to inflammatory responses. The lower frequency of TT GPx4 synthesis that is greater at optimal Se intake but that genotypes for the GPx4 T/C 718 SNP observed in patients is more sensitive to Se depletion. In contrast, there was no suffering from ulcerative colitis [33] and the evidence for a significant difference between the mRNA levels of IDI- link between longstanding ulcerative colitis and colon GPx4 T 718 and IDI-GPx4 C 718 constructs in Se-deficient cancer [36, 38] led us to investigate the frequency of this and Se-adequate conditions, suggesting that any effect of particular SNP, in patients diagnosed with CRC or ade- the 3 UTR on Se incorporation occurs at a translational, nomatous polyps and healthy controls. The frequency of rather than transcriptional level. The present data, there- the TT, TC and CC variations at this position was found to fore, indicate that the T and C variants of the GPx4 3 UTR be different in patients with CRC from those patients with differ in their ability to drive synthesis of a reporter adenomatous polyps or controls with no colo-rectal neo- selenoprotein. plasia, suggesting that the risk of CRC, but not the risk of These results are the first direct evidence that the SNP T/ having adenomatous polyps, is affected by genotype at the C 718 in the GPx4 3 UTR has functional consequences. T/C 718 SNP in the GPx4 gene. It is possible that the GPx4 They are supported by computational analysis that suggests T/C 718 SNP causes differences in inflammatory responses that there may be structural differences around the two core in the colon, responses that have been suggested to affect SECIS elements (Fig. 2) in the GPx4 T and GPx4 C progression of polyps to tumour formation [26, 28]. 0 0 3 UTRs. The ability of a small change in 3 UTR sequence However, the link between the C variant, possibly to affect Se incorporation is compatible with previous higher GPx4 activity and higher levels of lipoxygenase studies showing that the efficiency of SECIS elements to products and disease risk has not been defined. One pos- incorporate Se differs between selenoproteins in vitro [29] sibility is that alterations in GPx4 lead to changes in 5- and 123 Genes Nutr (2007) 2:225–232 231 15-lipoxygenase activity, cycloxygenase-2 activity and provides evidence that genotype at this SNP is associated activation of NF-jB pathways [7, 21, 37), reflecting the with altered risk of CRC. Our hypothesis is that a combi- role of GPx4 as a regulator of inflammatory responses. A nation of sub-optimal Se status and a particular genotype at second possible mechanism is based on the hierarchy of the GPx4 T/C 718 SNP increases susceptibility to CRC but selenoprotein synthesis in which selenoproteins essentially not to the risk of developing adenomatous polyps. Further compete for available Se [6] and, on such a model, work is required to define the link between GPx4 T/C 718 increased synthesis of GPx4 due to the C variant would be variants, selenoprotein synthesis and cell function. expected to cause a withdrawal of Se from synthesis of Acknowledgments We thank the World Cancer Research Fund for other selenoproteins such as GPx1 and so lead to an altered financial support (grants 2000/10, 2002/41), clinical staff for assis- pattern of overall selenoprotein synthesis. The observed tance with sampling and the patients for consenting to take part in this colitis in the combined GPx1/GPx2 knock-out mice where study. We greatly acknowledge Brian Burtle, Rebecca Lamb, Leanne Boulding and Rachel Ledward for help with genotyping and Fergus GPx1 has a minor role in counteracting inflammation [16, Nicol for help with IDI activity assay. JRA’s lab is funded by The 17) is compatible with this hypothesis. Furthermore, such Scottish Executive Environment and Rural Affairs Department (SE- indirect effects, through changes in the selenoprotein ERAD). Conflict of interest-None declared. hierarchy, not only could account for the link between the C variant and disease risk but are also consistent with the References observed beneficial effect of increased Se intake on CRC risk [27]. However, a third possibility is that high GPx4 1. Akbaraly NT, Arnaud J, Hininger-Favier I, Gourlet V, Roussel expression may lead directly to lower apoptosis [31]. AM, Berr C (2005) Selenium and mortality in the elderly: results Several lines of evidence suggest that low Se status from the EVA study. Clin Chem 51:2117–2123 increases risk of mortality from CRC. Firstly, colon cancer 2. Arthur JR, Nicol F, Beckett GJ (1990) Hepatic type 1 iodothyr- onine 5 -deiodinase. The role of selenium. Biochem J 272:537– mortality has been inversely correlated with estimated Se intake [12]. Secondly, low Se intake or low plasma Se is 3. Bermano G, Nicol F, Dyer JA, Sunde RA, Beckett GJ, Arthur JR, associated with increased risk of CRC [12, 35]. Thirdly, a Hesketh JE (1995) Tissue-specific regulation of selenoenzyme gene expression during selenium deficiency in rats. Biochem J recent meta-analysis indicates that higher Se status is 311:425–430 associated with lower risk of a recurrence of colonic 4. Bermano G, Arthur JR, Hesketh JE (1996) Role of the 3 tumours [27]. Finally, a clinical trial indicates that Se untranslated region in the regulation of cytosolic glutathione supplementation reduces CRC mortality rates [10]. In view peroxidase and phospholipid-hydroperoxide glutathione peroxi- dase gene expression by selenium supply. Biochem J 320:891– of these reported effects of low Se status on CRC risk [12, 27, 35, 43], it is possible that the impact of genotype at the 6. Berry MJ (2005) Insights into the hierarchy of selenium incor- 718 SNP is modulated by Se status. Plasma Se and eryth- poration. Nat Genet 37:1162–1163 rocyte GPx1 activity are commonly used measures of 5. Berry MJ, Banu L, Chen YY, Mandel SJ, Kieffer JD, Harney JW, functional Se status [14]. Interestingly, preliminary data Larsen PR (1991) Recognition of UGA as a selenocysteine codon in type I deiodinase requires sequences in the 3 untranslated indicate that in a sub-group of the cancer patients studied region. Nature 353:273–276 here both plasma Se levels and erythrocyte GPx1 activity 8. Brigelius-Flohe R (1999) Tissue-specific functions of individual were significantly lower in individuals with CC genotype glutathione peroxidases. Free Radic Biol Med 27:951–965 (mean ± SEM plasma Se = 0.80 ± 0.06 lmol/l, mean ± 7. Brigelius-Flohe R, Friedrichs B, Maurer S, Schultz M, streicher R (1997) Interleukin-1-induced nuclear factor kappa B activation is SEM GPx1 activity = 4.5 ± 0.5 U/ml) than in individuals inhibited by overexpression of phospholipid hydroperoxide glu- with CT or TT genotype (mean plasma Se = 0.99 ± 0.04 tathione peroxidase in a human endothelial cell line. Biochem J and 1.00 ± 0.05 lmol/l, respectively and mean GPx1 328:199–203 activity = 5.3 ± 0.2 and 6.1 ± 0.7 U/ml respectively). In 9. Clark LC, Hixson LJ, Combs GF, Reid ME, Turnbull BW, Sampliner RE (1993) Plasma selenium concentration predicts the contrast, in the polyp group, plasma Se concentration and prevalence of colorectal adenomatous polyps. Cancer Epidemiol erythrocyte GPx1 activity were not significantly affected Biomark Prev 2:41–46 by genotype. The observed differences in plasma Se and 10. Clark LC, Combs GF Jr, Turnbull BW, Slate EH, Chalker DK, Chow J, Davis LS, Glover RA, Graham GF, Gross EG, Krongrad erythrocyte GPx1 activity with genotype and clinical A, Lesher JL Jr, Park HK, Sanders BB Jr, Smith CL, Taylor JR diagnosis suggest that the combination of Se status and (1996) Effects of selenium supplementation for cancer prevention GPx4 T/C 718 genotype affects development of colorectal in patients with carcinoma of the skin. A randomized controlled cancer but not the risk of developing polyps. This is con- trial. Nutritional Prevention of Cancer Study Group. J Am Med Assoc 276:1957–1963 sistent with the previous suggestion that patients with 12. Combs GF (2005) Current evidence and research needs to support polyps and the lowest Se levels are at higher risk of a health claim for selenium and cancer prevention. J Nutr malignant transformation [15]. 135:343–347 In summary, the present study demonstrates the func- 11. Combs GF Jr, Gray WP (1998) Chemopreventive agents: sele- tionality of the SNP T/C 718 in the GPx4 gene and nium. Pharmacol Ther 79:179–192 123 232 Genes Nutr (2007) 2:225–232 13. Cummings OW (2000) Pathology of the adenoma–carcinoma 30. Kryukov GV, Kryukov VM, Gladyshev VN (1999) New mam- sequence: from aberrant crypt focus to invasive carcinoma. Se- malian selenocysteine-containing proteins identified with an min Gastrointest Dis 11:229–237 algorithm that searches for selenocysteine insertion sequence 14. Diplock AT (1993) Indexes of selenium status in human popu- elements. J Biol Chem 274:33888–33897 lations. Am J Clin Nutr 57:256S–258S 31. Nomura K, Imai H, Koumura T, Arai M, Nakagawa Y (1999) 15. Dworkin BM, Rosenthal WS, Mittelman A, Weiss L, Applebee- Mitochondrial phospholipid hydroperoxide glutathione peroxi- Brady L, Arlin Z (1988) Selenium status and the polyp-cancer dase suppresses apoptosis mediated by a mitochondrial death sequence: a colonoscopy controlled study. Am J Gastroenterol pathway. J Biol Chem 274:22294–29302 83:748–751 32. Pagmantidis V, Bermano G, Villette S, Broom J, Arthur J, 16. Esworthy RS, Aranda R, Martin MG, Doroshow JH, Binder SW, Hesketh JE (2005) Effects of Se-depletion on glutathione per- Chu FF (2001) Mice with combined disruption of Gpx1 and Gpx2 oxidase and SelW gene expression in the colon. FEBS Lett genes have colitis. Am J Physiol Gastrointest Liver Physiol 579:792–796 281:G848–55 33. Qatatsheh A, Seal CJ, Jowett SL, Welfare MR, Hesketh JE (2005) 17. Esworthy RS, Yang L, Frankel PH, Chu FF (2005) Epithelium- Patients with ulcerative colitis show an altered frequency distri- specific glutathione peroxidase, Gpx2, is involved in the pre- bution of a single nucleotide polymorphism in the gene encoding vention of intestinal inflammation in selenium-deficient mice. J the phospholipid hydroperoxide glutathione peroxidase. Pro- Nutr 135:740–745 ceedings of the Nutrition Society 64, OCA, 20A (Abstract) 18. Forsberg L, de Faire U, Morgenstern R (1998) To identify genetic 34. Rayman MP (2005) Selenium in cancer prevention: a review of polymorphisms in the ‘‘Expressed Sequence Tag’’ (EST) data- the evidence and mechanism of action. Proc Nutr Soc 64:527– base. Technical tips on-line http://www.elsevier.com_/locate_/tto 542 , T01440 35. Russo MW, Murray SC, Wurzelmann JI, Woosley JT, Sandler RS 19. Forsberg L, de Faire U, Marklund SL, Andersson PM, Stegmayr (1997) Plasma selenium levels and the risk of colorectal adeno- B, Morgenstern R (2000) Phenotype determination of a common mas. Nutr Cancer 28:125–129 Pro-Leu polymorphism in human glutathione peroxidase 1. Blood 36. Rutter M, Saunders B, Wilkinson K, Rumbles S, Schofield G, Cells Mol Dis 26:423–426 Kamm M, Williams C, Price A, Talbot I, Forbes A (2004) 20. Haklar G, Sayin-Ozveri E, Yuksel M, Aktan AO, Yalcin AS Severity of inflammation is a risk factor for colorectal neoplasia (2001) Different kinds of reactive oxygen and nitrogen species in ulcerative colitis. Gastroenterology 126:451–459 were detected in colon and breast tumors. Cancer Lett 165:219– 37. Schnurr K, Belkner J, Ursini F, Schewe T, Kuhn H (1996) The 224 selenoenzyme phospholipid hydroperoxide glutathione peroxi- 21. Heirman I, Ginneberge D, Brigelius-Flohe R, Hendrickx N, Ag- dase controls the activity of the 15-lipoxygenase with complex ostinis P, Brouckaert P, Rottiers P. And Grooten J (2006) substrates and preserves the specificity of the oxygenation prod- Blocking tumor cell eicosanoid synthesis by GPx4 impedes tumor ucts. J Biol Chem 271:4653–4658 growth and malignancy. Free Radic Biol Med 40:285–94 38. Seril DN, Liao J, Yang GY, Yang CS (2003) Oxidative stress and 23. Hu YJ, Diamond AM (2003) Role of glutathione peroxidase 1 in ulcerative colitis-associated carcinogenesis: studies in humans breast cancer: loss of heterozygosity and allelic differences in the and animal models. Carcinogenesis 24:353–362 response to selenium. Cancer Res 63:3347–3351 39. Shamberger RJ, Tytko SA, Willis CE (1976) Antioxidants and 22. Hu YJ, Korotkov KV, Mehta R, Hatfield DL, Rotimi CN, Luke cancer. Part VI. Selenium and age-adjusted human cancer mor- A, Prewitt TE, Cooper RS, Stock W, Vokes EE, Dolan ME, tality. Arch Environ Health 31:231–235 Gladyshev VN, Diamond AM (2001) Distribution and functional 40. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, consequences of nucleotide polymorphisms in the 3 -untranslated Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC region of the human Sep15 gene. Cancer Res 61:2307–2310 (1985) Measurement of protein using bicinchoninic acid. Ann 24. Ichimura Y, Habuchi T, Tsuchiya N, Wang L, Oyama C, Sato K, Biochem 150:76–85 Nishiyama H, Ogawa O, Kato T (2004) Increased risk of bladder 41. Stadtman TC (1996) Selenocysteine. Ann Rev Biochem 65:83– cancer associated with a glutathione peroxidase 1 codon 198 100 variant. J Urol 172:728–732 42. Villette S, Kyle JA, Brown KM, Pickard K, Milne JS, Nicol F, 25. Imai H, Nakagawa Y (2003) Biological significance of phos- Arthur JR, Hesketh JE (2002) A novel single nucleotide poly- pholipid hydroperoxide glutathione peroxidase (PHGPx, GPx4) morphism in the 3 untranslated region of human glutathione in mammalian cells. Free Radic Biol Med 34:145–169 peroxidase 4 influences lipoxygenase metabolism. Blood Cells 26. Itzkowitz SH, Yio X (2004) Inflammation and cancer IV. Colo- Mol Dis 29:174–178 rectal cancer in inflammatory bowel disease: the role of 43. Whanger PD (2004) Selenium and its relationship to cancer: an inflammation. Am J Physiol Gastrointest Liver Physiol 287:G7– update dagger. Br J Nutr 91:11–28 G17 44. Wingler K, Bo ¨ cher M, Flohe L, Kollmus H, Brigelius-Flohe R 27. Jacobs ET, Jiang R, Alberts DS, Greenberg ER, Gunter EW, (1999) mRNA stability and selenocysteine insertion sequence Karagas MR, Lanza E, Ratnasinghe L, Reid ME, Schatzkin A, efficiency rank gastrointestinal glutathione peroxidase high in the Smith-Warner SA, Wallace K, Martinez ME (2004) Selenium hierarchy of selenoproteins. Eur J Biochem 259:149–157 and colorectal adenoma: results of a pooled analysis. J Natl 45. Yagi K, Komura S, Kojima H, Sun Q, Nagata N, Ohishi N, Cancer Instit 96:1669–1675 Nishikimi M (1996) Expression of human phospholipid hydro- 28. Jacobson-Brown P, Neuman MG (2004) Colorectal polyposis and peroxide glutathione peroxidase gene for protection of host cells immune-based therapies. Canad J Gastroenterol 18:239–249 from lipid hydroperoxide-mediated injury. Biochem Biophys Res 29. Kollmus H, Flohe L, McCarthy JE (1996) Analysis of eukaryotic Commun 219:486–491 mRNA structures directing cotranslational incorporation of selenocysteine. Nucleic Acids Res 24:1195–1201

Journal

Genes & Nutrition – Springer Journals

Published: Oct 13, 2007

There are no references for this article.