Gpx (glutathione peroxidase)-1 enzyme activity and mRNA levels decrease dramatically in Se (selenium) deficiency, whereas other selenoproteins are less affected by Se deficiency. This hierarchy of Se regulation is not understood, but the position of the UGA selenocysteine codon is thought to play a major role in making selenoprotein mRNAs susceptible to nonsense-mediated decay. Thus in the present paper we studied the complete selenoproteome in the mouse to uncover additional selenoprotein mRNAs that are highly regulated by Se status. Mice were fed on Se-deficient, Se-marginal and Se-adequate diets (0, 0.05 and 0.2 μg of Se/g respectively) for 35 days, and selenoprotein mRNA levels in liver and kidney were determined using microarray analysis and quantitative real-time PCR analysis. Se-deficient mice had liver Se concentrations and liver Gpx1 and thioredoxin reductase activities that were 4, 3 and 3% respectively of the levels in Se-adequate mice, indicating that the mice were Se deficient. mRNAs for Selh (selenoprotein H) and Sepw1 (selenoprotein W) as well as Gpx1 were decreased by Se deficiency to <40% of Se-adequate levels. Five and two additional mRNAs were moderately down-regulated in Sedeficient liver and kidney respectively. Importantly, nine selenoprotein mRNAs in liver and fifteen selenoprotein mRNAs in the kidney were not significantly regulated by Se deficiency, clearly demonstrating that Se regulation of selenoprotein mRNAs is not a general phenomenon. The similarity of the response to Se deficiency suggests that there is one underlying mechanism responsible. Importantly, the position of the UGA codon did not predict susceptibility to Se regulation, clearly indicating that additional features are involved in causing selenoprotein mRNAs to be sensitive to Se status.

INTRODUCTION

Gpx (glutathione peroxidase)-1 was the first identified biochemical role for the essential nutrient Se (selenium) [1], and this activity is used to assess the risk of Se-dependent disease and to determine Se status. This discovery was followed rapidly by the identification of Se-dependent enzymes in bacteria [2] and then by the identification of Sec (selenocysteine) as the Se cofactor present and incorporated into the peptide backbone of these selenoproteins [3,4]. These discoveries were followed over the next 30 years by the identification of additional selenoproteins in higher animals as well as bacteria, including plasma Sepp1 (selenoprotein P) containing ten selenocysteines, all three mammalian deiodinases, all three mammalian Txnrds (thioredoxin reductases) and three additional Gpxs with unique tissue distribution or substrate utilization [5,6]. Characterization of this growing superfamily of selenoproteins revealed that Sec is encoded by a UGA codon [7], that a unique 3′-UTR (untranslated region) stem-loop structure called a SECIS (Sec insertion sequence) element is required in the mRNA to reinterpret the UGA codon for Sec rather than as a termination of translation codon [8,9], and that a series of unique and novel gene products are required to synthesize an activated form of Se and then use serine attached to a unique tRNA to synthesize Sec [10,11]. Se in the form of Sec is incorporated co-translationally via this cellular machinery into the growing polypeptide chain at the position specified by the UGA codon. Previously, a computational method based on these elements was developed to identify in-frame UGA-containing selenoproteins [12] in sequenced genomes. Kryukov et al. [12] thus elegantly found that the human selenoproteome consists of 25 selenoproteins and that the rodent selenoproteome consists of 24 selenoproteins. This is the first example where the complete proteome specific for a cofactor has been identified.

Gpx1 activity decreases dramatically under conditions of Se deficiency and increases during Se repletion, thus making Gpx1 a useful biomarker for Se status [13]. Other selenoproteins, such as Gpx4 (phospholipid hydroperoxide Gpx), only decrease in activity to approx. 40–50% of maximal activity in Se-deficient animals [14]. A more complete understanding of this regulation emerged when it was found that Gpx1 mRNA decreases dramatically under conditions of Se deficiency [15], whereas Gpx4 mRNA levels do not decrease [14,16]. Other selenoprotein levels, such as Txnrd, decrease to nearly the same extent as Gpx1, but Txnrd mRNA levels are only modestly decreased [17]. This hierarchy of Se regulation of the various selenoproteins and their mRNAs is not currently understood.

The degradation of mRNA levels in Se deficiency appears to be due to NMD (nonsense-mediated decay) [18,19]. The decrease in Gpx1 mRNA levels is not due to transcriptional regulation, but rather due to NMD of Gpx1 mRNA when Se is lacking. mRNA levels for several other well-characterized selenoproteins, however, are not susceptible to NMD. The underlying mechanism appears to require the UGA codon to be positioned >55 nt upstream of an intron splice junction [2022], but this condition is also true for Gpx4 mRNA, which is decreased in Se-deficient cultured cells, but not in intact Se-deficient animals, whereas Gpx1 mRNA is decreased by nearly an order of magnitude [14,23,24]. A number of hypotheses have been proposed to explain the hierarchy of the susceptibility of selenoprotein mRNAs to degradation [8,18,19,2528], but this issue is complicated because Se deficiency has a much more profound impact in animals than in cultured cells [18,23]. Thus we conducted the present study of the complete selenoproteome in mice fed on differing levels of dietary Se, with the hypothesis that we would uncover additional selenoprotein mRNAs that are highly regulated by Se status and so might help to better understand the underlying mechanism of Se regulation of mRNA stability.

To evaluate the complete selenoproteome for Se regulation of mRNA levels, selenoprotein mRNA levels in total RNA from liver and kidney of Se-deficient, Se-marginal and Se-adequate mice were determined using microarray analysis and qRT-PCR (quantitative real-time PCR) analysis. We found two selenoprotein mRNAs in addition to Gpx1 mRNA that were highly down-regulated in Se-deficient tissue, and six additional mRNAs that were moderately down-regulated. We also found, however, that the majority of selenoprotein mRNAs were not decreased significantly by Se status. Furthermore, the sensitivity of these selenoprotein mRNAs to Se regulation is not predicted by current models for the underlying mechanism.

MATERIALS AND METHODS

Reagents

Molecular biology reagents were purchased from Promega (Madison, WI, U.S.A.), Invitrogen (Carlsbad, CA, U.S.A.) or Sigma (St Louis, MO, U.S.A.). All other chemicals were of molecular biology or reagent grade.

Animal husbandry

Male mouse pups from Se-adequate dams in our wild-type colony (predominantly C57 Black background) were weaned 18 days after birth and housed individually in hanging-wire cages for 35 days. The basal Se-deficient diet was a rodent torula-yeast diet containing 0.005 μg of Se/g and supplemented with 100 mg/kg of all-rac-α-tocopherol acetate to ensure prevention of liver necrosis, and supplemented with 0.4% L-methionine to ensure adequate growth, as described previously [16]. Mice were fed on the basal diet supplemented with 0, 0.05 or 0.2 μg of Se/g as Na2SeO3 to provide Se-deficient, Se-marginal or Se-adequate diets [29], and had free access to feed and water. The care and treatment protocol was approved by the University of Wisconsin institutional animal care and use committee.

Tissue analyses

Mice were anaesthetized with ether, blood was drawn by cardiac puncture using heparinized syringes, and were then killed by exsanguination under anaesthesia. Livers were perfused in situ with ice-cold 0.15 M potassium chloride, and all tissues were removed and snap-frozen in liquid nitrogen for later enzyme and RNA analysis. Blood was centrifuged [1500 g for 15 min at 4°C, Eppendorf 5415R centrifuge, F-45-24-11 rotor (Brinkmann, Westbury, NY, U.S.A.)] to separate the plasma from red cells, and the red cells were reconstituted to the original volume using PBS [76 mM NaCl and 50 mM sodium phosphate (pH 7.4)]. Liver, kidney and diet Se concentrations were determined by neutron-activation analysis [30].

Enzyme assays

To prepare tissue supernatant, frozen tissues were homogenized in nine volumes of sucrose buffer [20 mM Tris/HCl (pH 7.4), 0.25 M sucrose, 1 mM EDTA and 0.1% peroxide-free Triton X-100] and centrifuged [10000 g for 15 min at 4°C, model J2-21M, JA-21 rotor (Beckman Instruments, Palo Alto, CA, U.S.A.)]. Gpx1 activity in liver, kidney and red blood cells (designated as Gpx1) and Gpx3 activity in plasma [designated as Gpx3 (plasma Gpx)] were measured by the coupled assay procedure [31] using 120 μM H2O2. Gpx4 activity was measured by the coupled assay procedure [14] using 78 μM phosphatidylcholine hydroperoxide, the specific substrate. For both assays, 1 unit is the amount of enzyme that will oxidize 1 μmole of GSH per min under these conditions. Txnrd was measured using the gold-inhibition assay with DTNB [5,5′-dithiobis-(2-nitrobenzoic acid)] used as a substrate [32]. The protein concentration of each sample was determined by the method of Lowry et al. [33].

RNA isolation and processing

Total RNA from mouse liver and kidney (50–100 mg of tissue, n=3 per diet group) was isolated using the guanidinium isothiocyanate method using TRIzol reagent (Invitrogen), following the manufacturer's protocol. mRNA was purified using an oligo(dT)-linked Oligotex resin (Qiagen, Valencia, CA, U.S.A.), and then used to synthesize double-stranded cDNA [34]. The cDNA was purified using phenol/chloroform extraction [35], and was used for in vitro transcription of biotin-labelled cRNA using the Enzo BioArray HighYield RNATranscript Labelling Kit (Enzo Life Sciences, Farmingdale, NY, U.S.A.). Finally, purified cRNA was fragmented at 94°C for 35 min in a buffer containing 200 mM Tris/acetate (pH 8.0), 500 mM potassium acetate and 150 mM magnesium acetate. The integrity and size of total RNA, cRNA and fragmented cRNA for each sample was assessed by formaldehyde–agarose-gel electrophoresis. RNA or DNA obtained at each step was quantified using an ND-1000 UV-Vis Spectrophotometer (NanoDrop Technologies, Wilmington, DE, U.S.A.).

Microarray analysis

Each fragmented cRNA sample (10 μg) was hybridized to one Affymetrix Mouse Genome 430 2.0 array following the manufacturer's protocols (Affymetrix, Santa Clara, CA, U.S.A.). After hybridization, the arrays were washed and stained with streptavidin phycoerythrin (Invitrogen) in a GC450 Fluidics Station (Affymetrix) and scanned using a GC3000 7G scanner (Affymetrix). Hybridization, washing and scanning of the arrays were performed by the Gene Expression Center at the University of Wisconsin-Madison. Expression data were generated from raw fluorescence values using Microarray Suite 5.0 (MAS 5.0) (Affymetrix) within the GCOS (GeneChip Operating Software version 1.4) (Affymetrix). The GCOS-generated detection calls and scaled expression values for the probe sets, targeting each of the 24 mouse selenoprotein genes, were obtained from each single array analysis. If multiple probe sets were found to target the same gene, the probe set giving the highest expression in Se-adequate samples was used for analysis. In addition, the expression of eight non-selenoprotein genes was determined, including three housekeeping genes [Gapdh (glyceraldehyde-3-phosphate dehydrogenase), Actb (β-actin) and Rps14 (ribosomal protein S14)], and including five orthologues of selenoprotein genes that encode cysteine-containing rather than Sec-containing proteins [Mrps14 (mitochondrial Rps14), Gpx7, Msr (methionine sulfoxide reductase), a Msrb2 and Msrb3]. Selenoprotein mRNA levels were normalized to the expression of all of the probe sets within each array by the GCOS software, the expression ratios for each gene from each of the nine arrays per tissue were calculated relative to the average Se-adequate expression (set to 100) and visualized using Treeview software [36].

qRT-PCR

The relative mRNA abundance was also determined by qRT-PCR. RNA (1 μg) was reverse transcribed to cDNA using the RETROscript kit (AM1710, Ambion, Austin, TX, U.S.A.), following the manufacturer's instructions. Gene-specific primers were designed to span a splice junction and amplify an approx. 150 base-pair segment (see Table 1). The final real time reactions (25 μl final volume) contained 10 ng of reverse-transcribed RNA, 0.2 mM gene-specific forward and reverse primers and 1× SybrGreen PCR Master Mix (#4309155, Applied Biosystems, Foster City, CA, U.S.A.). Reactions were performed in an ABI Prism 7000 Sequence Detection System (Applied Biosystems), with initial stages of 50°C for 2 min and 95°C for 10 min, followed by 50 cycles consisting of 95°C for 15 s and 60°C for 2 min. A dissociation curve was run for each plate to confirm the production of a single product. The amplification efficiency for each gene was determined using the DART program [37]. The relative abundance of each mRNA was calculated using the method of Pfaffl [38], accounting for gene-specific efficiencies, normalized to Actb levels, and set relative to the mean of the Se-adequate (0.2 μg of Se/g of diet) animals.

Table 1
qRT-PCR primers for mouse selenoproteins used in the present study

Selo, selenoprotein O; Sels, selenoprotein S; Selt, selenoprotein T.

GeneForward primerReverse primer
ActB 5′-AGCCATGTACGTAGCCATCC-3′ 5′-CTCTCAGCTGTGGTGGTGAA-3′ 
Dio1 5′-GGAACCATAGGCATTGGAAA-3′ 5′-AGTGCCAGAGAGCCAGATTC-3′ 
Gapdh 5′-GTGTTCCTACCCCCAATGTG-3′ 5′-AGGAGACAACCTGGTCCTCA-3′ 
Gpx1 5′-GGTTCGAGCCCAATTTTACA-3′ 5′-CCCACCAGGAACTTCTCAAA-3′ 
Gpx3 5′-GATGTGAACGGGGAGAAAGA-3′ 5′-CCCACCAGGAACTTCTCAAA-3′ 
Gpx4 5′-CTCCATGCACGAATTCTCAG-3′ 5′-ACGTCAGTTTTGCCTCATTG-3′ 
Rps14 5′-TGTCTGCCACATCTTTGCAT-3′ 5′-GAGGACTCATCTCGGTCAGC-3′ 
Selh 5′-GCGAGATTTGAACTTTGCATC-3′ 5′-TTGTCCACCGTCTCCATAGG-3′ 
Selk 5′-GCTGGTGGATGAGGAAGGTA-3′ 5′-CTCATTCATCTGTGGGGACA-3′ 
Selm 5′-TTTGTCACCGAGGACATTCA-3′ 5′-TGTACCAGCGCATTGATCTC-3′ 
Selo 5′-TGCACAGAAAGCCATTGAAG-3′ 5′-GGAAGATTGCTCCTCAGTGC-3′ 
Sels 5′-GCCTTACGCACACTTTCACA-3′ 5′-GTGGCCTAATGGCAATGTCT-3′ 
Selt 5′-TGTGGCAACAGAAAGGGATT-3′ 5′-CAGGTGGCATCAACATCAAG-3′ 
Sep15 5′-CTCACCAGTGAAACGCTTTG-3′ 5′-TCAAAGAGCACACAGCAAGG-3′ 
Sepp1 5′-GCAATTGCTTGACAGTGTGC-3′ 5′-TTCATGGGCTGATTTTGTCA-3′ 
Sepw1 5′-CCCAAGTACCTCCAGCTCAA-3′ 5′-GCCATCACCTCTCTTCTTGG-3′ 
Sepx1 5′-ACAGTTGTTGCCCCATTAGC-3′ 5′-GGAGTGGGTCTCAGCTTCAG-3′ 
Txnrd1 5′-TTCGACCTGATCATCATTGG-3′ 5′-CCACATTCACACACGTTCCT-3′ 
Txnrd2 5′-CCCTAGAGTGTGCTGGCTTC-3′ 5′-AAGCATGATCCTCCCAAGTG-3′ 
Txnrd3 5′-CCTTTCCCAGTTGCTAGTGC-3′ 5′-GTGCTACACTCTGGGCAACA-3′ 
GeneForward primerReverse primer
ActB 5′-AGCCATGTACGTAGCCATCC-3′ 5′-CTCTCAGCTGTGGTGGTGAA-3′ 
Dio1 5′-GGAACCATAGGCATTGGAAA-3′ 5′-AGTGCCAGAGAGCCAGATTC-3′ 
Gapdh 5′-GTGTTCCTACCCCCAATGTG-3′ 5′-AGGAGACAACCTGGTCCTCA-3′ 
Gpx1 5′-GGTTCGAGCCCAATTTTACA-3′ 5′-CCCACCAGGAACTTCTCAAA-3′ 
Gpx3 5′-GATGTGAACGGGGAGAAAGA-3′ 5′-CCCACCAGGAACTTCTCAAA-3′ 
Gpx4 5′-CTCCATGCACGAATTCTCAG-3′ 5′-ACGTCAGTTTTGCCTCATTG-3′ 
Rps14 5′-TGTCTGCCACATCTTTGCAT-3′ 5′-GAGGACTCATCTCGGTCAGC-3′ 
Selh 5′-GCGAGATTTGAACTTTGCATC-3′ 5′-TTGTCCACCGTCTCCATAGG-3′ 
Selk 5′-GCTGGTGGATGAGGAAGGTA-3′ 5′-CTCATTCATCTGTGGGGACA-3′ 
Selm 5′-TTTGTCACCGAGGACATTCA-3′ 5′-TGTACCAGCGCATTGATCTC-3′ 
Selo 5′-TGCACAGAAAGCCATTGAAG-3′ 5′-GGAAGATTGCTCCTCAGTGC-3′ 
Sels 5′-GCCTTACGCACACTTTCACA-3′ 5′-GTGGCCTAATGGCAATGTCT-3′ 
Selt 5′-TGTGGCAACAGAAAGGGATT-3′ 5′-CAGGTGGCATCAACATCAAG-3′ 
Sep15 5′-CTCACCAGTGAAACGCTTTG-3′ 5′-TCAAAGAGCACACAGCAAGG-3′ 
Sepp1 5′-GCAATTGCTTGACAGTGTGC-3′ 5′-TTCATGGGCTGATTTTGTCA-3′ 
Sepw1 5′-CCCAAGTACCTCCAGCTCAA-3′ 5′-GCCATCACCTCTCTTCTTGG-3′ 
Sepx1 5′-ACAGTTGTTGCCCCATTAGC-3′ 5′-GGAGTGGGTCTCAGCTTCAG-3′ 
Txnrd1 5′-TTCGACCTGATCATCATTGG-3′ 5′-CCACATTCACACACGTTCCT-3′ 
Txnrd2 5′-CCCTAGAGTGTGCTGGCTTC-3′ 5′-AAGCATGATCCTCCCAAGTG-3′ 
Txnrd3 5′-CCTTTCCCAGTTGCTAGTGC-3′ 5′-GTGCTACACTCTGGGCAACA-3′ 

Statistics

Data are presented as means±S.E.M. (n=3/diet group). Results were analysed by ANOVA using a fixed model testing the main effect of diet. F-tests and least-squares means were calculated using the mixed procedures of SAS (SAS, Cary, NC, U.S.A.). When the main effect of diet was significant (P<0.05), the means were separated by the least-significant-difference test.

RESULTS

Physiological markers

The weanling mice weighed 8.4 g at the start of the study, and after 35 days the final masses were 28.1±1.2, 27.5±0.9 and 28.7±2.3 (P=0.82) for mice fed on 0, 0.05 and 0.2 μg of Se/g of diet respectively. The average gain was 0.57 g/day. There was no effect of dietary Se level on body mass throughout the study (see Supplementary Figure S1 at http://www.bioscirep.org/bsr/029/bsr0290329add.htm). Similarly, there was no effect of dietary Se on liver and kidney masses, expressed as total organ mass or as a percentage of body mass (results not shown). There were no overt signs of Se deficiency in these mice at any point in the study.

Se analysis indicated that tissue Se concentrations in mice fed on the Se-deficient diet decreased to 0.5 and 2.3 nmol/g of tissue in liver and kidney respectively (Figure 1A). Liver Se was 4% of the level in mice fed on an Se-adequate diet containing 0.2 μg of Se/g, clearly indicating that the mice were Se deficient. Kidney Se in Se-deficient mice was only 17% of Se-adequate levels. Liver and kidney Se levels in mice fed on a diet containing 0.05 μg of Se/g were 49% and 64% of the levels in Se-adequate mice respectively, showing that these Se-marginal mice were intermediate in Se status.

Effect of Se status on tissue Se concentration and on tissue selenoenzyme activity

Figure 1
Effect of Se status on tissue Se concentration and on tissue selenoenzyme activity

At weaning, mice were fed on diets containing 0, 0.05 or 0.2 μg of Se/g of diet to provide Se-deficient, Se-marginal or Se-adequate diets, and killed 35 days later. Values are means±S.E.M. (n=3) and, for a given analysis, values not sharing a common letter (uppercase or lowercase) are significantly different (P<0.05). (A) Tissue Se concentration. Kidney, ●; liver, ○. (B) Gpx3 (●) and red blood cell Gpx1 (○) activities. (C) Tissue Gpx1 activities. Kidney, ●; liver, ○. (D) Liver Gpx4 (○) and Txnrd (●) activities. EU, enzyme unit.

Figure 1
Effect of Se status on tissue Se concentration and on tissue selenoenzyme activity

At weaning, mice were fed on diets containing 0, 0.05 or 0.2 μg of Se/g of diet to provide Se-deficient, Se-marginal or Se-adequate diets, and killed 35 days later. Values are means±S.E.M. (n=3) and, for a given analysis, values not sharing a common letter (uppercase or lowercase) are significantly different (P<0.05). (A) Tissue Se concentration. Kidney, ●; liver, ○. (B) Gpx3 (●) and red blood cell Gpx1 (○) activities. (C) Tissue Gpx1 activities. Kidney, ●; liver, ○. (D) Liver Gpx4 (○) and Txnrd (●) activities. EU, enzyme unit.

Biochemical markers

Gpx3 activity was decreased in Se-deficient mice to 13% of Se-adequate levels (Figure 1B), with Gpx3 activity in Se-marginal mice 26% of the level in Se-adequate mice. Gpx1 activity in Se-deficient red blood cells (Figure 1B) only decreased to 37% of Se-adequate levels, as expected due to the long half-life of the red blood cell in rodents [39]; Se-marginal red blood cell Gpx1 activity was 57% of Se-adequate levels.

In Se-deficient liver, liver Gpx1 activity was 3.4% of Se-adequate levels, and Se-marginal liver only had 40% of the Se-adequate Gpx1 activity (Figure 1C). The Se-deficient kidney appeared to be less Se deficient, with kidney Gpx1 activity at 17% of Se-adequate levels. In Se-deficient liver, Txnrd activity was 2.8% of Se-adequate levels, whereas Se-marginal liver had 68% of Se-adequate Txnrd activity (Figure 1D). Kidney Txnrd and Gpx4 activities were not determined due to lack of sufficient material. Thus, based on these biochemical markers, and mice fed on the Se-deficient diet were clearly Se deficient, mice fed on the Se-marginal diet had less than half of the level of Se-adequate enzyme activity for liver and kidney Gpx1, for liver Txnrd and for Gpx3, indicating that these mice were marginal in Se status.

Gpx4 activity in Se-deficient liver, in contrast to Gpx1, Gpx3 and Txnrd activities, was only 34% of Se-adequate levels (Figure 1D). Se-marginal liver had 79% of Se-adequate Gpx4 activity.

Microarray studies

Studies with the Affymetrix Mouse Genome 430 2.0 array were then conducted to determine the Se regulation of all 24 selenoprotein mRNAs in the complete selenoproteome in the mouse. Microarray analysis identified 17 liver selenoprotein mRNAs as ‘present’ (Figure 2), including nine that were significantly down-regulated in Se-deficient liver [Gpx1, Selh (selenoprotein H), Sepw1 (selenoprotein W), Txnrd1, Txnrd2, Selk (selenoprotein K), Dio1 (deiodinase, iodothyronine, type I), Selt (selenoprotein T) and Sep15 (15 kDa selenoprotein)]. In kidney, 20 selenoprotein mRNAs were identified as ‘present’, including eight that were significantly down-regulated in Se-deficient kidney [Gpx1, Selh, Sepw1, Selm (selenoprotein M), Txnrd1, Dio1, Sep15 and Sels (selenoprotein S)]. In addition, eight non-selenoprotein mRNAs (Gapdh, Actb, Rps14, Mrps14, Gpx7, Msra, Msrb2 and Msrb3) were also evaluated by microarray analysis and found not to be significantly affected by Se status.

Visualization of effect of Se status on selenoprotein mRNA expression as assessed by microarray analysis

Figure 2
Visualization of effect of Se status on selenoprotein mRNA expression as assessed by microarray analysis

Expression ratios for each gene from each of the nine arrays per tissue were calculated relative to average Se-adequate expression and set to 100, as described in the Materials and methods section. Expression levels were visualized using Treeview software [36], as shown in the colour scale. (A) Liver and (B) kidney mRNA from mice fed on diets containing 0, 0.05 or 0.2 μg of Se/g for 35 days were subjected to microarray analysis for all 24 selenoproteins in the mouse genome, and for the expression of eight non-selenoprotein genes, including three housekeeping genes (Gapdh, Actb and Rps14), and including five orthologues of selenoprotein genes that encode cysteine-containing rather than Sec-containing proteins (Mrps14, Gpx7, Msra, Msrb2 and Msrb3). Shown are the seventeen liver selenoprotein mRNAs and twenty kidney selenoprotein mRNAs described as ‘present’ by the GCOS software. Seli, selenoprotein I; Selo, selenoprotein O; Sels, selenoprotein S; Selt, selenoprotein T; Sephs2, selenophosphate synthetase 2.

Figure 2
Visualization of effect of Se status on selenoprotein mRNA expression as assessed by microarray analysis

Expression ratios for each gene from each of the nine arrays per tissue were calculated relative to average Se-adequate expression and set to 100, as described in the Materials and methods section. Expression levels were visualized using Treeview software [36], as shown in the colour scale. (A) Liver and (B) kidney mRNA from mice fed on diets containing 0, 0.05 or 0.2 μg of Se/g for 35 days were subjected to microarray analysis for all 24 selenoproteins in the mouse genome, and for the expression of eight non-selenoprotein genes, including three housekeeping genes (Gapdh, Actb and Rps14), and including five orthologues of selenoprotein genes that encode cysteine-containing rather than Sec-containing proteins (Mrps14, Gpx7, Msra, Msrb2 and Msrb3). Shown are the seventeen liver selenoprotein mRNAs and twenty kidney selenoprotein mRNAs described as ‘present’ by the GCOS software. Seli, selenoprotein I; Selo, selenoprotein O; Sels, selenoprotein S; Selt, selenoprotein T; Sephs2, selenophosphate synthetase 2.

Normalized fluorescence of the selenoprotein mRNAs in the microarray analysis suggested that Sepp1, Gpx1, Gpx4 and Sepx1 (also known as Msrb1) (highest to lowest) were the selenoprotein transcripts with the apparent highest abundance in Se-adequate liver (see Supplementary Figure S2 at http://www.bioscirep.org/bsr/029/bsr0290329add.htm). These estimates match previous analyses that found that Sepp1, Gpx1 and Gpx4 transcript levels were comparable to Gapdh levels in Se-adequate liver [16,23]. In Se-adequate kidney, Gpx3, Sepp1, Gpx1 and Gpx4 (highest to lowest) were the selenoprotein transcripts with the apparent highest abundance (see Supplementary Figure S3 at http://www.bioscirep.org/bsr/029/bsr0290329add.htm).

qRT-PCR studies

To confirm that the levels of these mRNAs are regulated by Se status, qRT-PCR analyses were conducted for all selenoprotein mRNAs identified by microarray analysis as being Se regulated in either tissue, plus Sepp1, Gpx4 and Sepx1 in both tissues, and Gpx3, Selo (selenoprotein O) and Txnrd3 in kidney. The resulting expression profiles, determined by both microarray analysis and qRT-PCR, are shown in Figures 3 and 4. In general, there was very good agreement between microarray analysis and qRT-PCR analysis for the regulation of mRNA levels by Se status. Using qRT-PCR, transcript levels of three housekeeping genes, Gapdh, Actb and Rps14, were also found not to be significantly affected by Se status (results not shown).

Effect of Se status on liver selenoprotein mRNA expression

Figure 3
Effect of Se status on liver selenoprotein mRNA expression

Liver mRNA from mice fed on diets containing 0, 0.05 or 0.2 μg of Se/g for 35 days was subjected to microarray analysis (○) and to qRT-PCR analysis (●) as described in the Materials and methods section. For microarray analysis, relative expression (%) was calculated relative to the mean of the Se-adequate (0.2 μg of Se/g) abundance (set to 100) after the GCOS software had normalized selenoprotein mRNA levels to the expression of all probe sets within each array. For qRT-PCR analysis, relative mRNA abundance was calculated using the method of Pfaffl [38] to account for gene-specific efficiencies, normalized to Actb levels in each sample, and set relative to the mean of the Se-adequate (0.2 μg of Se/g of diet) abundance. Values are means±S.E.M. (n=3) and, for a given analysis, values not sharing a common letter are significantly different (P<0.05). Shown are mRNA expression levels for (A) Gpx1, (B) Selh, (C) Sepw1, (D) Txnrd1, (E) Txnrd2, (F) Selk, (G) Gpx4 and (H) Sepx1.

Figure 3
Effect of Se status on liver selenoprotein mRNA expression

Liver mRNA from mice fed on diets containing 0, 0.05 or 0.2 μg of Se/g for 35 days was subjected to microarray analysis (○) and to qRT-PCR analysis (●) as described in the Materials and methods section. For microarray analysis, relative expression (%) was calculated relative to the mean of the Se-adequate (0.2 μg of Se/g) abundance (set to 100) after the GCOS software had normalized selenoprotein mRNA levels to the expression of all probe sets within each array. For qRT-PCR analysis, relative mRNA abundance was calculated using the method of Pfaffl [38] to account for gene-specific efficiencies, normalized to Actb levels in each sample, and set relative to the mean of the Se-adequate (0.2 μg of Se/g of diet) abundance. Values are means±S.E.M. (n=3) and, for a given analysis, values not sharing a common letter are significantly different (P<0.05). Shown are mRNA expression levels for (A) Gpx1, (B) Selh, (C) Sepw1, (D) Txnrd1, (E) Txnrd2, (F) Selk, (G) Gpx4 and (H) Sepx1.

Analysis by qRT-PCR showed that Gpx1 mRNA levels in Se-deficient liver were decreased to 26% of Se-adequate levels (Figure 3A). Unlike liver Se and liver Gpx1 activity, however, liver Gpx1 mRNA levels in Se-marginal liver were not significantly different from Se-adequate Gpx1 mRNA levels. In kidney, a similar pattern was observed, with Se-deficient and Semarginal Gpx1 mRNA levels decreased to 27% and 79% of Se-adequate levels respectively (Figure 4A). Gpx4 mRNA levels were not significantly affected by the level of dietary Se in either liver or kidney (Figures 3G and 4F). Thus the present study with mice shows differential regulation of mRNA levels as well as selenoenzyme activities.

Effect of Se status on kidney selenoprotein mRNA expression

Figure 4
Effect of Se status on kidney selenoprotein mRNA expression

Kidney mRNA levels in mice fed on diets containing 0, 0.05 or 0.2 μg of Se/g. Values were calculated as described in the legend to Figure 3. Shown are mRNA expression levels for (A) Gpx1, (B) Selh, (C) Sepw1, (D) Gpx3, (E) Selm and (F) Gpx4.

Figure 4
Effect of Se status on kidney selenoprotein mRNA expression

Kidney mRNA levels in mice fed on diets containing 0, 0.05 or 0.2 μg of Se/g. Values were calculated as described in the legend to Figure 3. Shown are mRNA expression levels for (A) Gpx1, (B) Selh, (C) Sepw1, (D) Gpx3, (E) Selm and (F) Gpx4.

In the present study, Sepw1 and Selh mRNA levels in Sedeficient liver were decreased to 26 and 33% of levels in Se-adequate liver respectively, and thus the transcript levels of these genes, along with Gpx1, are highly regulated by Se status. In addition, mRNAs for Txnrd1 and Txnrd2 were moderately and significantly decreased in Se-deficient liver to 45–50% of Se-adequate levels, and mRNAs for Sepp1, Selk and Selm were significantly decreased to 60–70% of Se-adequate levels. In all, eight selenoprotein mRNAs were significantly decreased by Se deficiency in mouse liver to <70% of Se-adequate levels. The mRNA levels in Se-marginal liver were not significantly different from Se-adequate levels for most Se-regulated mRNAs. For the Se-regulated genes, the response curves appeared to be hyperbolic, with the largest increase in mRNA level occurring in the transition from Se-deficient to Se-marginal states; mRNA levels in Se-marginal liver were not significantly different from Se-adequate levels for most Se-regulated mRNAs. Lastly, nine of seventeen selenoprotein mRNAs present in liver, including Gpx4 and Sepx1 mRNAs (Figures 3G and 3H), were not significantly regulated by Se status.

In kidney, Gpx1, Sepw1 and Selh mRNAs were also highly regulated by Se status, decreasing in Se-deficient kidney to 27, 32 and 37% of Se-adequate levels respectively (Figures 4A–4C) as assessed by qRT-PCR. These values are again very similar to the values determined by microarray analysis. In Se-deficient kidney, Selm was 64% of Se-adequate levels by qRT-PCR (Figure 4E), and 47% of Se-adequate levels by microarray analysis. Kidney Se-deficient Gpx3 mRNA level was 64% of Se-adequate levels by qRT-PCR (Figure 4D), but not significantly different by microarray analysis. Thus five selenoprotein mRNAs were significantly decreased by Se deficiency in mouse kidney. Again the response curves for the regulated kidney selenoprotein mRNAs appeared to be hyperbolic, as stated for liver. Lastly, fifteen of the twenty selenoprotein mRNAs present in kidney (Figure 4), including Gpx4 mRNA (Figure 4F), were not significantly regulated by Se status.

DISCUSSION

The impact of Se deficiency in the present study was very similar to previous reports studying pups from Se-adequate dams [39,40]. There was no significant effect on growth or the apparent health of these mice supplemented with normal levels of vitamin E and the sulfur amino acids. Feeding on an Se-deficient diet for 35 days, however, dramatically decreased liver Se, and liver Gpx1 and Txnrd activity to 4, 3 and 3% of Se-adequate levels respectively. These decreases are very similar to levels reported previously [40]. This clearly shows that these young rapidly growing mice were biochemically Se deficient without overt alterations in their overall physiology.

Liver and kidney mRNA levels for Gpx1 mRNA were also decreased to 26 and 27% of Se-adequate levels respectively. These decreases are not as large as those observed in rats [23], most probably due to the differences between these two rodent models in growth and thus in the dilution of initial Se stores. Just as in rats, Gpx4 activity was only moderately affected by Se deficiency, and Gpx4 mRNA was unchanged by Se deficiency in these mice, illustrating that these mRNAs for Gpx1 and Gpx4 are differentially regulated by Se status in mice. Thus this mouse model allowed us to detect alterations in mRNA levels for the complete selenoproteome due to changes in Se status, including the six newly discovered but uncharacterized rodent selenoproteins [12].

In both liver and kidney, Sepw1 and Selh mRNA levels assessed by microarray analysis were highly regulated by Se status, as was Gpx1 mRNA, with all three mRNAs decreasing in Se-deficient mice to <40% of Se-adequate levels. This regulation was confirmed by qRT-PCR analysis for Gpx1, Sepw1 and Selh mRNAs. The present study is the first report in which Selh mRNA is regulated directly by Se status. Sepw1 mRNA has been reported to be decreased by Se deficiency in mouse muscle [41], but other researchers were not able to detect Sepw1 mRNA in liver [42]. In addition, Txnrd1, Txnrd2, Sepp1, Selk and Selm mRNAs were moderately down-regulated in Se-deficient liver, and Selm and Gpx3 mRNAs were moderately down-regulated in Sedeficient kidney. Importantly, nine selenoprotein mRNAs in liver and fifteen selenoprotein mRNAs in kidney, including Gpx4 mRNA in both tissues, were not significantly regulated by Se deficiency, clearly demonstrating that Se regulation of selenoprotein mRNAs is not a general phenomenon for all selenoproteins.

The mechanism underlying the Se regulation of the selenoprotein mRNA level is not understood. For Gpx1 and Gpx4 mRNA, at least, it is clear that this regulation is not due to transcriptional regulation or mRNA processing and export. Previous studies with a recombinant chimaera of Gpx1 indicate that the decrease in Gpx1 mRNA in Se deficiency is due to NMD [18,19]. Positioning of an in-frame UGA codon sufficiently upstream of a splice junction in β-globin along with a SECIS element in the 3′-UTR will confer Se regulation on to β-globin mRNA [18]. Susceptibility to NMD appears to require a UGA codon placed at least 55 nt upstream of the exon/intron splice junction [2022]. This explains the high regulation of Gpx1 mRNA (UGA codon at 105 nt upstream), but it does not explain why Gpx4 mRNA (UGA codon at 105 nt upstream) is not susceptible to Se regulation in vivo [14,23]. This is further complicated by a previous study which shows that transiently expressed Gpx4 mRNA or endogenous Gpx4 mRNA in cultured cells is susceptible to NMD [24], indicating that additional factors are likely to be involved.

The present study was conducted in part to survey the full selenoproteome for additional selenoprotein mRNAs which are susceptible to Se regulation, presumably by NMD, and to relate this susceptibility to the position of its cognate UGA codon relative to the downstream splice junction (Table 2). Highly regulated Gpx1 mRNA has its UGA codon located at 105 nt from the splice junction. Two additional highly regulated mRNAs were identified; Selh fits the >55 nt rule, with its UGA codon located 136 nt upstream of the splice junction, but the UGA codon for Sepw1 only lies 15 nt upstream. For the moderately regulated mRNAs identified in the present study, the UGA codon positions for Txnrd1 and Txnrd2 fit the >55 nt rule, but the UGA codon positions for Selk, Selm and Gpx3 and the first UGA codon for Sepp1 are 5, 21, 22 and 26 nt upstream of the splice junction respectively. Several of the selenoprotein mRNAs found to have low Se regulation in the present study, including Gpx4 and Dio1, have UGA codons positioned much greater than 55 nt upstream of the splice junction. The present study thus provides a number of new examples showing that the >55 nt rule does not explain the susceptibility to degradation.

Table 2
Effect of UGA codon position on Se regulation

The susceptibility to Se regulation of selenoprotein mRNA levels in Se-deficient mouse tissue (liver and kidney), as assessed by qRT-PCR in Se-deficient versus Se-adequate mouse samples (high=20–40% of Se-adequate; moderate=40–70% of Se-adequate; low=>70% of Se-adequate). The exon location of the UGA codon in each selenoprotein gene [data from National Center for Biotechnology Information (NCBI, http://www.ncbi.nlm.nih.gov/)], and the distance in nt from the UGA (A) codon and the 3′-exon/intron splice junction are indicated. A, absent as deterimed by Affymetrix GCOS software analysis of microarray data; nd, not determined by qRT-PCR analysis, present but not Se regulated in microarray analysis; Dio2, deiodinase, iodothyronine, type II; Seli, selenoprotein I; Selo, selenoprotein O; Sels, selenoprotein S; Selt, selenoprotein T; Sephs2, selenophosphate synthetase 2.

Se geneUGA exon (nthof n exons)Distance(nt)Liver SeregulationKidney Seregulation
Gpx1 1st of 2 105 High High 
Selh 2nd of 4 136 High High 
Sepw1 2nd of 6 15 High High 
Selk 4th of 5 Moderate Low 
Selm 2nd of 5 21 Moderate Moderate 
Txnrd1 15th of 15 1685 Moderate Low 
Txnrd2 17th of 18 83 Moderate Low 
Sepp1 2nd of 5 26 Moderate Low 
Gpx3 2nd of 5 22 Moderate 
Dio1 2nd of 4 103 Low Low 
Selt 2nd of 6 101 Low Low 
Gpx4 3rd of 7 105 Low Low 
Sepx1 3rd of 5 34 Low Low 
Sep15 3rd of 5 28 Low Low 
H47/Sels 6th of 6 559 Low Low 
Selo 9th of 9 363 nd Low 
Txnrd3 16th of 16 738 Low 
Sephs2 1st of 1 1852 nd nd 
Seli 10th of 10 5464 nd nd 
Dio2 2nd of 2 5271 nd 
Se geneUGA exon (nthof n exons)Distance(nt)Liver SeregulationKidney Seregulation
Gpx1 1st of 2 105 High High 
Selh 2nd of 4 136 High High 
Sepw1 2nd of 6 15 High High 
Selk 4th of 5 Moderate Low 
Selm 2nd of 5 21 Moderate Moderate 
Txnrd1 15th of 15 1685 Moderate Low 
Txnrd2 17th of 18 83 Moderate Low 
Sepp1 2nd of 5 26 Moderate Low 
Gpx3 2nd of 5 22 Moderate 
Dio1 2nd of 4 103 Low Low 
Selt 2nd of 6 101 Low Low 
Gpx4 3rd of 7 105 Low Low 
Sepx1 3rd of 5 34 Low Low 
Sep15 3rd of 5 28 Low Low 
H47/Sels 6th of 6 559 Low Low 
Selo 9th of 9 363 nd Low 
Txnrd3 16th of 16 738 Low 
Sephs2 1st of 1 1852 nd nd 
Seli 10th of 10 5464 nd nd 
Dio2 2nd of 2 5271 nd 

There are several alternative hypotheses that may be involved in Se regulation of selenoprotein mRNA level. Some studies using chimaeric constructs expressed in cultured cells suggest that regions in both the coding region and in the 3′-UTR are involved in making selenoprotein mRNAs susceptible to decay in Se-deficient cells [25]. The extent of Se regulation in cultured cells, however, is much smaller than in the intact animal [18], making these studies difficult to interpret. NMD in mammals appears to be a consequence of conditions in the pioneering round of translation [22], and so competition for a limited supply of Se for translation may switch the destination of an mRNA from translation to NMD [23]. The UGA codon position itself relative to the start codon or to the SECIS element, or the local UGA context, can all affect translational efficiency and thus might affect mRNA sensitivity to NMD [43]. Early studies with chimaeric mRNA in cultured cells or oocytes found that SECIS elements differ in their ability to facilitate Sec incorporation in translation [8], and thus varied translational efficiency due to differences in SECIS elements might confer differential sensitivity to NMD [8,44]. Differential affinities of SECIS elements for isoforms of the Sec-tRNA also influence translational efficiency and are accompanied by changes in mRNA levels, and so may influence mRNA stability [26]; these shifts in relative levels of Sec-tRNA isoforms, however, are modest in Se-deficient mice [45], making it unlikely that this alone regulates selenoprotein mRNA stability in intact animals. Similarly, differing affinities of SBPs (SECIS-binding proteins), such as SBP2, for the SECIS element have also been proposed to explain the hierarchy of sensitivity to NMD [27,28]; these studies in cultured cells may explain the differential effects of translation of selenoprotein mRNAs, but studies with chimaeric constructs have shown that SECIS elements from selenoprotein mRNAs not susceptible to NMD are as sufficient to confer NMD sensitivity as those from NMD-susceptible mRNAs, thus indicating that Se regulation of NMD involves more than the SECIS element and SBPs [18]. Lastly, the presence of a putative second SECIS stem-loop immediately downstream of the UGA codon in the coding region in some selenoprotein mRNAs is thought to influence translational efficiency [46] but only Sepn1 (selenoprotein N1) mRNA, which we did not find to be present in mouse liver or kidney, has been studied. The two new mRNAs shown in the present study to be highly regulated in vivo may prove to be useful to better understand the mechanism underlying this regulation.

One overall striking feature of the Se-response curves in the present study (Figures 3 and 4) is that a marginal level of dietary Se (0.05 μg of Se/g of diet) raises selenoprotein mRNA levels to Se-adequate levels; the dramatic impact on selenoprotein mRNA stability of these Se-regulated mRNAs occurs at or <0.05 μg of Se/g of diet. In rats fed on intermediate levels of dietary Se to better define Se regulation, the response curves for Gpx1, Selh and Sepw1 mRNAs are hyperbolic, all overlap and reach plateau responses at or near 0.05 μg of Se/g of diet [47]. This strongly suggests that there is one underlying mechanism involved in Se regulation of selenoprotein mRNA levels, but further studies will be required to unravel this mechanism.

In summary, microarray and qRT-PCR analysis of mRNA levels for the complete selenoprotein proteome found two additional selenoprotein mRNAs, in addition to Gpx1, that were highly regulated by Se status. Several additional selenoprotein mRNAs were moderately regulated by Se status, but the majority of selenoprotein mRNAs were not down-regulated by Se deficiency in mouse liver and kidney. This clearly indicates that Se regulation of mRNA level is not a general phenomenon of selenoprotein mRNAs. The similarity of the Se response curves for those mRNAs sensitive to Se status, however, suggests that there is one underlying mechanism at work. Lastly, the position of the Sec codon, UGA, in the Se-regulated mRNAs does not predict at all the susceptibility to Se regulation, clearly indicating that additional features are involved in making selenoprotein mRNA levels sensitive to Se status. For these selenoproteins with Se-regulated mRNA levels, it will also be important to determine if the protein levels of the selenoproteins are also regulated by Se status.

Abbreviations

     
  • Actb

    β-actin

  •  
  • Dio1

    deiodinase, iodothyronine, type I

  •  
  • Gapdh

    glyceraldehyde-3-phosphate dehydrogenase

  •  
  • GCOS

    GeneChip Operating Software version 1.4

  •  
  • Gpx

    glutathione peroxidase

  •  
  • Gpx3

    plasma Gpx

  •  
  • Gpx4

    phospholipid hydroperoxide Gpx

  •  
  • Msr

    methionine sulfoxide reductase

  •  
  • NMD

    nonsense-mediated decay

  •  
  • qRT-PCR

    quantitative real-time PCR

  •  
  • Rps14

    ribosomal protein S14

  •  
  • Mrps14

    mitochondrial Rps14

  •  
  • Sec

    selenocysteine

  •  
  • SECIS

    Sec insertion sequence

  •  
  • SBP

    SECIS-binding protein

  •  
  • Selh

    selenoprotein H

  •  
  • Selk

    selenoprotein K

  •  
  • Selm

    selenoprotein M

  •  
  • Sep15

    15 kDa selenoprotein

  •  
  • Sepp1

    selenoprotein P

  •  
  • Sepw1

    selenoprotein W

  •  
  • Sepx1

    Msrb1

  •  
  • Txnrd

    thioredoxin reductase

  •  
  • UTR

    untranslated region.

FUNDING

This work was supported in part by the National Institutes of Health [grant numbers DK74184, T32-DK07665 (training grant supporting K. M. B.)], and by the University of Wisconsin Agricultural Experiment Station [grant number WIS04909].

References

References
Rotruck
 
J. T.
Pope
 
A. L.
Ganther
 
H. E.
Swanson
 
A. B.
Hafeman
 
D. G.
Hoekstra
 
W. G.
 
Selenium: biochemical role as a component of glutathione peroxidase
Science
1973
, vol. 
179
 (pg. 
588
-
590
)
Turner
 
D. C.
Stadtman
 
T. C.
 
Purification of protein components of the clostridial glycine reductase system and characterization of protein A as a selenoprotein
Arch. Biochem. Biophys.
1973
, vol. 
154
 (pg. 
366
-
381
)
Forstrom
 
J. W.
Zakowski
 
J. J.
Tappel
 
A. L.
 
Identification of the catalytic site of rat liver GSH-Px as selenocysteine
Biochemistry
1978
, vol. 
17
 (pg. 
2639
-
2644
)
Jones
 
J. B.
Dilworth
 
G. L.
Stadtman
 
T. C.
 
Occurrence of selenocysteine in the selenium-dependent formate dehydrogenase of Methanococcus vannielii
Arch. Biochem. Biophys.
1979
, vol. 
195
 (pg. 
255
-
260
)
Hatfield
 
D. L.
 
Selenium: Its Molecular Biology and Role in Human Health
2001
Norwood, MA.
Kluwer Academic Publishers
Sunde
 
R. A.
 
Bowman
 
B. A.
Russell
 
R. M.
 
Selenium
Present Knowledge in Nutrition
2001
Washington, D.C.
ILSI Press
(pg. 
352
-
365
)
Chambers
 
I.
Frampton
 
J.
Goldfarb
 
P. S.
Affara
 
N.
McBain
 
W.
Harrison
 
P. R.
 
The structure of the mouse glutathione peroxidase gene: the selenocysteine in the active site is encoded by the ‘termination’ codon
TGA. EMBO J.
1986
, vol. 
5
 (pg. 
1221
-
1227
)
Berry
 
M. J.
Banu
 
L.
Harney
 
J. W.
Larsen
 
P. R.
 
Functional characterization of the eukaryotic SECIS elements which direct selenocysteine insertion at UGA codons
EMBO J.
1993
, vol. 
12
 (pg. 
3315
-
3322
)
Walczak
 
R.
Westhof
 
E.
Carbon
 
P.
Krol
 
A.
 
A novel RNA structural motif in the selenocysteine insertion element of eukaryotic selenoprotein mRNAs
RNA 2
1996
(pg. 
367
-
379
)
Sunde
 
R. A.
Evenson
 
J. K.
 
Serine incorporation into the selenocysteine moiety of glutathione peroxidase
J. Biol. Chem.
1987
, vol. 
262
 (pg. 
933
-
937
)
Hatfield
 
D. L.
Berry
 
M. J.
Gladyshev
 
V. N.
 
Selenium: Its Molecular Biology and Role in Human Health
2006
New York, NY
Springer
Kryukov
 
G. V.
Castellano
 
S.
Novoselov
 
S. V.
Lobanov
 
A. V.
Zehtab
 
O.
Guigo
 
R.
Gladyshev
 
V. N.
 
Characterization of mammalian selenoproteins
Science
2003
, vol. 
300
 (pg. 
1439
-
1443
)
Sunde
 
R. A.
 
Hatfield
 
D. L.
Berry
 
M. J.
Gladyshev
 
V. N.
 
Regulation of glutathione peroxidase-1 expression
Selenium: Its Molecular Biology and Role in Human Health
2006
New York, NY
Springer
(pg. 
149
-
160
)
Lei
 
X. G.
Evenson
 
J. K.
Thompson
 
K. M.
Sunde
 
R. A.
 
Glutathione peroxidase and phospholipid hydroperoxide glutathione peroxidase are differentially regulated in rats by dietary selenium
J. Nutr.
1995
, vol. 
125
 (pg. 
1438
-
1446
)
Saedi
 
M. S.
Smith
 
C. G.
Frampton
 
J.
Chambers
 
I.
Harrison
 
P. R.
Sunde
 
R. A.
 
Effect of selenium status on mRNA levels for glutathione peroxidase in rat liver
Biochem. Biophys. Res. Commun.
1988
, vol. 
153
 (pg. 
855
-
861
)
Sunde
 
R. A.
Evenson
 
J. K.
Thompson
 
K. M.
Sachdev
 
S. W.
 
Dietary selenium requirements based on glutathione peroxidase-1 activity and mRNA levels and other selenium parameters are not increased by pregnancy and lactation in rats
J. Nutr.
2005
, vol. 
135
 (pg. 
2144
-
2150
)
Hadley
 
K. B.
Sunde
 
R. A.
 
Selenium regulation of thioredoxin reductase activity and mRNA levels in rat liver
J. Nutr. Biochem.
2001
, vol. 
12
 (pg. 
693
-
702
)
Weiss
 
S. L.
Sunde
 
R. A.
 
Cis-acting elements are required for selenium regulation of glutathione peroxidase-1 mRNA levels
RNA 4
1998
(pg. 
816
-
827
)
Moriarty
 
P. M.
Reddy
 
C. C.
Maquat
 
L. E.
 
Selenium deficiency reduces the abundance of mRNA for Se-dependent glutathione peroxidase 1 by a UGA-dependent mechanism likely to be nonsense codon-mediated decay of cytoplasmic mRNA
Mol. Cell. Biol.
1998
, vol. 
18
 (pg. 
2932
-
2939
)
Low
 
S. C.
Berry
 
M. J.
 
Knowing when not to stop: selenocysteine incorporation in eukaryotes
Trends Biochem. Sci.
1996
, vol. 
21
 (pg. 
203
-
208
)
Nagy
 
E.
Maquat
 
L. E.
 
A rule for termination-codon position within intron-containing genes: when nonsense affects RNA abundance
Trends Biochem. Sci.
1998
, vol. 
23
 (pg. 
198
-
199
)
Maquat
 
L. E.
 
Nonsense-mediated mRNA decay in mammals
J. Cell Sci.
2005
, vol. 
118
 (pg. 
1773
-
1776
)
Weiss Sachdev
 
S.
Sunde
 
R. A.
 
Selenium regulation of transcript abundance and relative translational efficiency of glutathione peroxidase 1 and 4 in rat liver
Biochem. J.
2001
, vol. 
357
 (pg. 
851
-
858
)
Sun
 
X.
Li
 
X.
Moriarty
 
P. M.
Henics
 
T.
LaDuca
 
J. P.
Maquat
 
L. E.
 
Nonsense-mediated decay of mRNA for the selenoprotein phospholipid hydroperoxide glutathione peroxidase is detectable in cultured cells but masked or inhibited in rat tissues
Mol. Biol. Cell
2001
, vol. 
12
 (pg. 
1009
-
1017
)
Muller
 
C.
Wingler
 
K.
Brigelius-Flohé
 
R.
 
3′UTRs of glutathione peroxidases differentially affect selenium-dependent mRNA stability and selenocysteine incorporation efficiency
Biol. Chem.
2003
, vol. 
384
 (pg. 
11
-
18
)
Carlson
 
B. A.
Xu
 
X. M.
Gladyshev
 
V. N.
Hatfield
 
D. L.
 
Selective rescue of selenoprotein expression in mice lacking a highly specialized methyl group in selenocysteine tRNA
J. Biol. Chem.
2005
, vol. 
280
 (pg. 
5542
-
5548
)
Squires
 
J. E.
Stoytchev
 
I.
Forry
 
E. P.
Berry
 
M. J.
 
SBP2 binding affinity is a major determinant in differential selenoprotein mRNA translation and sensitivity to nonsense-mediated decay
Mol. Cell Biol.
2007
, vol. 
27
 (pg. 
7848
-
7855
)
Bubenik
 
J. L.
Driscoll
 
D. M.
 
Altered RNA binding activity underlies abnormal thyroid hormone metabolism linked to a mutation in selenocysteine insertion sequence-binding protein 2
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
34653
-
34662
)
National Research Council
Nutrient Requirements of Laboratory Animals
1995
Washington, DC
National Academy Press
McKown
 
D. M.
Morris
 
J. S.
 
Rapid measurement of selenium in biological samples using instrumental neutron activation analysis
J. Radioanal. Nucl. Chem.
1978
, vol. 
43
 (pg. 
411
-
420
)
Lawrence
 
R. A.
Sunde
 
R. A.
Schwartz
 
G. L.
Hoekstra
 
W. G.
 
Glutathione peroxidase activity in rat lens and other tissues in relation to dietary selenium intake
Exp. Eye Res.
1974
, vol. 
18
 (pg. 
563
-
569
)
Hill
 
K. E.
McCollum
 
G. W.
Burk
 
R. F.
 
Determination of thioredoxin reductase activity in rat liver supernatant
Anal. Biochem.
1997
, vol. 
253
 (pg. 
123
-
125
)
Lowry
 
O. H.
Rosebrough
 
N. J.
Farr
 
A. L.
Randall
 
R. J.
 
Protein measurement with the Folin phenol reagent
J. Biol. Chem.
1951
, vol. 
193
 (pg. 
265
-
275
)
Gubler
 
U.
Hoffman
 
B. J.
 
A simple and very efficient method for generating cDNA libraries
Gene
1983
, vol. 
25
 (pg. 
263
-
269
)
Ausubel
 
F. M.
Brent
 
R.
Kingston
 
R. E.
Moore
 
D. D.
Seidman
 
J. G.
Smith
 
J. A.
Struhl
 
K.
 
Current Protocols in Molecular Biology
1989
New York, NY
Wiley and Sons
Eisen
 
M. B.
Spellman
 
P. T.
Brown
 
P. O.
Botstein
 
D.
 
Cluster analysis and display of genome-wide expression patterns
Proc. Natl. Acad. Sci. U.S.A.
1998
, vol. 
95
 (pg. 
14863
-
14868
)
Peirson
 
S. N.
Butler
 
J. N.
Foster
 
R. G.
 
Experimental validation of novel and conventional approaches to quantitative real-time PCR data analysis
Nucleic Acids Res.
2003
, vol. 
31
 pg. 
e73
 
Pfaffl
 
M. W.
 
A new mathematical model for relative quantification in real-time RT-PCR
Nucleic Acids Res.
2001
, vol. 
29
 pg. 
e45
 
Weiss
 
S. L.
Evenson
 
J. K.
Thompson
 
K. M.
Sunde
 
R. A.
 
The selenium requirement for glutathione peroxidase mRNA level is half of the selenium requirement for glutathione peroxidase activity in female rats
J. Nutr.
1996
, vol. 
126
 (pg. 
2260
-
2267
)
Zhang
 
L.
Maiorino
 
M.
Roveri
 
A.
Ursini
 
F.
 
Phospholipid hydroperoxide glutathione peroxidase: specific activity in tissues of rats of different age and comparison with other glutathione peroxidases
Biochim. Biophys. Acta
1989
, vol. 
1006
 (pg. 
140
-
143
)
Hooven
 
L. A.
Butler
 
J.
Ream
 
L. W.
Whanger
 
P. D.
 
Microarray analysis of selenium-depleted and selenium- supplemented mice
Biol. Trace Elem. Res.
2006
, vol. 
109
 (pg. 
173
-
179
)
Vendeland
 
S. C.
Beilstein
 
M. A.
Yeh
 
J. Y.
Ream
 
W.
Whanger
 
P. D.
 
Rat skeletal muscle selenoprotein W: cDNA clone and mRNA modulation by dietary selenium
Proc. Natl. Acad. Sci. U.S.A.
1995
, vol. 
92
 (pg. 
8749
-
8753
)
Wen
 
W.
Weiss
 
S. L.
Sunde
 
R. A.
 
UGA codon position affects the efficiency of selenocysteine incorporation into glutathione peroxidase-1
J. Biol. Chem.
1998
, vol. 
273
 (pg. 
28533
-
28541
)
Wingler
 
K.
Bocher
 
M.
Flohé
 
L.
Kollmus
 
H.
Brigelius-Flohé
 
R.
 
mRNA stability and selenocysteine insertion sequence efficiency rank gastrointestinal glutathione peroxidase high in the hierarchy of selenoproteins
Eur. J. Biochem.
1999
, vol. 
259
 (pg. 
149
-
157
)
Chittum
 
H. S.
Hill
 
K. E.
Carlson
 
B. A.
Lee
 
B. J.
Burk
 
R. F.
Hatfield
 
D. L.
 
Replenishment of selenium deficient rats with selenium results in redistribution of the selenocysteine tRNA population in a tissue specific manner
Biochim. Biophys. Acta
1997
, vol. 
1359
 (pg. 
25
-
34
)
Howard
 
M. T.
Aggarwal
 
G.
Anderson
 
C. B.
Khatri
 
S.
Flanigan
 
K. M.
Atkins
 
J. F.
 
Recoding elements located adjacent to a subset of eukaryal selenocysteine-specifying UGA codons
EMBO J.
2005
, vol. 
24
 (pg. 
1596
-
1607
)
Barnes
 
K. M.
Evenson
 
J. K.
Raines
 
A. M.
Sunde
 
R. A.
 
Transcript analysis of the selenoproteome indicates that dietary selenium requirements based on selenium-regulated selenoprotein mRNAs are uniformly less than requirements based on glutathione peroxidase activity in rats
J. Nutr.
2009
, vol. 
139
 (pg. 
199
-
206
)

Author notes

2

Present address: Division of Animal and Veterinary Scienes, West Virginia University, G014 Agricultural Sciences Building, Morgantown, WV 26506, U.S.A.

Supplementary data