Evidence for a low molecular weight selenium-containing protein in muscle, now called selenoprotein W, was first reported in 1969 (Pedersen et al. 1969). The presence of this protein was noted in subsequent work (Black et al. 1978, Pedersen et al. 1972), but its purification occurred only recently (Vendeland et al. 1993). Interest in this protein arose from its possible involvement in the etiology of the sheep nutritional myopathy, white muscle disease. Failure to incorporate selenium into this protein in selenium-deficient lambs is associated with the myopathy (Pedersen et al. 1972). Selenium was demonstrated to be present as selenocysteine in a partially pure preparation of the protein (Beilstein et al. 1981), which is consistent with the form present in other selenoproteins (Behne et al. 1990, Berry et al. 1991, Burk and Hill 1993, Hill et al. 1991).

Synthesis of several other selenocysteine-containing proteins is regulated by dietary selenium (Behne et al. 1990, Burk and Hill 1993, Sunde 1990). Previous studies of the relationship of selenoprotein W to dietary selenium in sheep (Black et al. 1978) were based on the observation of the incorporation of radiolabeled ⁷⁵Se into a low molecular weight fraction of muscle cytosol. The availability of antibodies which recognized the selenoprotein in a number of species (Yeh et al. 1995) made it feasible to determine levels of the protein precisely during dietary manipulations. The specificity of the antibody for selenoprotein W has been demonstrated (Yeh et al. 1995). In the current study, selenoprotein W was measured in muscle biopsies from sheep that were either supplemented with or depleted of selenium. After 10.5 wk of consuming these diets, the sheep were killed and the selenoprotein concentration was examined in other tissues.

MATERIALS AND METHODS

Biopsies of the biceps femoris muscle were taken from each sheep initially and at 3.5, 7.0 and 10.5 wk after beginning the experimental diets. For surgery, sheep were first anesthetized with Biotal (Boehringer Ingelheim Animal Health, St. Louis, MO) and maintained under anesthesia with Halothane (Halocarbon Laboratory, River Edge, NJ). A lateral incision was made to access the biceps femoris. Fascia and connective tissue were then cleaned to remove a 1-g portion of the caudal biceps femoris. This muscle was immediately frozen with dry ice and kept at 70^o C until analysis. The remaining biceps femoris was sutured and the wound closed. Animals were sprayed with Furox topical antibiotic (Solvay Animal Health, Mendota Heights, MN) and injected intramuscularly with Liquimycin LA-200 antibiotic (Pfizer Animal Health, New York, NY) to prevent infection. Sheep were monitored for 2 wk following surgery to ensure complete recovery. Blood was taken from the jugular vein at the time of each biopsy. After a sample was taken for selenium analysis, the blood was centrifuged at 1400 × g to separate the plasma from erythrocytes. The plasma was frozen at 70⁰ C until analysis. Selenium concentration and glutathione peroxidase (GPX, EC 1.11.1.9) activity were determined on the plasma samples. At the end of the experiment, sheep were killed by exsanguination after being stunned with a stun-gun, and samples of two muscles (biceps femoris and semitendinosus), heart, tongue, brain, lung, spleen, kidney and liver were quickly removed and frozen on dry ice. Tissue samples were subsequently stored at 70^o C until analysis.

Tissues were homogenized in 6 volumes of buffer [20 mmol/L Tris (pH 7.5), 0.25 mol/L sucrose, 1 mmol/L EGTA, 5 mmol/L EDTA, 1 mmol/L phenylmethylsulfonyl fluoride, 50 mmol/L 2-mercaptoethanol and 0.025 g/L leupeptin], and the homogenates were centrifuged at 17,000 × g for 10 min at 4^oC to obtain cytosolic extracts. GPX activity was determined in extracts by the method of Paglia and Valentine (1967), and protein concentrations were measured by the dye-binding assay (Bradford 1976) using bovine serum albumin as a standard. After digestion with nitric and perchloric acids, selenium concentration of whole blood, tissue samples and extracts was determined by a semi-automated fluorometric assay (Brown and Watkinson 1977) with an Alpchem II system (Alpchem, Milwaukie, OR).

Tissue extracts were electrophoresed on 7.5-15% gradient SDS-polyacrylamide gels as described by Laemmli (1970), and proteins were transferred onto nitrocellulose membranes (0.2 µm: BA-S83, Schleicher & Schuell, Keene, NH) according to the method of Towbin et al. (1979). Pure rat selenoprotein W, ranging from 5 to 20 ng, was included in each gel to use as standards. Selenoprotein W contents in tissue extracts were then determined in Western blot analysis as described by Yeh et al. (1995), using a rabbit polyclonal antibody against the peptide sequence corresponding to amino acid residues 13 to 31 of rat selenoprotein W. After hybridization with horseradish peroxidase-conjugated goat anti-rabbit IgG (Bio-Rad, Richmond, CA), blots were incubated with ECL chemiluminescent reagent (Amersham Life Sciences, Arlington Heights, IL) and exposed to Kodak X-OMAT film (Eastman Kodak, Rochester, NY). Developed films were scanned with a Personal Densitometer SI (Molecular Dynamics, Sunnyvale, CA) and analyzed by the ImageQuanT program (Molecular Dynamics).

Data were examined for equal variance and normal distribution prior to statistical analysis. Where necessary to achieve homogeneity, the data were subjected to logarithmic transformations. Statistical analyses were performed by ANOVA with Fisher's least-significant difference (LSD) method for comparison within treatment groups and the Student-Newman-Keuls t test for comparison between treatment groups (Steel and Torrie 1980). Correlation coefficients were also calculated on some of the data. A significance level of at least 5% was adopted for all means to be considered statistically different.

RESULTS

Figure 2 shows the Western blot for muscle biopsy samples at various times from the two groups of sheep. Based on the intensity of the bands, the levels of selenoprotein W increased over time in supplemented sheep whereas levels decreased in deficient sheep.

The selenoprotein W in the muscle (Fig. 3) increased with time of selenium supplementation. The concentration of this selenoprotein gradually dropped from about 23.9 to 15.1 ng/mg protein at the end of the study in sheep fed the deficient diet (P < 0.05). In contrast, there was a significant increase (P < 0.05) of this selenoprotein in the muscle of supplemented sheep. The selenoprotein W concentrations in muscles of supplemented sheep were significantly higher (P < 0.05) at 7.0 wk compared with the initial values, and the values at 10.5 wk were significantly higher (P < 0.05) compared with the initial and 3.5-wk values.

The selenium concentration in muscles of deficient sheep decreased gradually (P < 0.05), whereas that in muscles of supplemented sheep increased markedly (P < 0.05) and reached a plateau after 3.5 wk of supplementation (Fig. 4). The selenium concentration in the muscle cytosols followed the same patterns as the selenoprotein W content (data not shown).

The GPX activity of the muscle followed similar trends but did not duplicate the results with selenoprotein W (Fig. 5). As expected, GPX activity dropped (P < 0.05) after 3.5 wk in the deficient sheep, whereas GPX activity of supplemented sheep increased dramatically (P < 0.05) and reached a plateau after 3.5 wk of supplementation. The patterns of muscle GPX activity and muscle selenium concentration in supplemented sheep were similar.

The selenoprotein W level of the various tissues is shown in Figure 6. The selenoprotein W content was significantly lower (P < 0.05) in all tissues of the controls, except brain, than the supplemented sheep. The selenoprotein W level was highest in skeletal muscle and lowest in liver.

The selenium concentration in the whole tissues followed a different pattern than that of selenoprotein W (Fig. 7). The selenium concentration was highest in the liver and lowest in skeletal muscles (combined biceps femoris and semitendinosus) in decreasing order. The selenium concentration was significantly higher (P < 0.01) in all tissues from the supplemented sheep than in those from sheep fed the deficient diet. The selenium in the cytosols from these tissues followed the same patterns (data not shown).

The GPX activity in the various sheep tissues did not follow the pattern of either the selenium concentration or the selenoprotein W levels (Fig. 8). This activity was lower in all tissues from the deficient sheep than in tissues from the supplemented group (P < 0.05). GPX activity was highest in the spleen and lowest in the skeletal muscles (combined biceps and semitendinosus). Thus, it is evident that the distributions of selenium, selenoprotein W and GPX among tissues do not follow similar patterns. There were no significant correlations (P > 0.05) among the relative tissue distributions of selenium, selenoprotein W level and GPX activity.

DISCUSSION

These results indicate that there are no correlations among the relative tissue distributions of selenium, GPX and selenoprotein W. Selenoprotein W was highest in muscle and heart (Fig. 6), selenium was highest in liver (Fig. 7), whereas GPX was highest in the spleen of selenium-supplemented sheep (Fig. 8). There was a significant correlation of muscle selenoprotein W concentration with the selenium concentrations (r = 0.64, P < 0.002), with muscle cytosolic selenium (r = 0.59, P < 0.005), and with muscle GPX activity (r = 0.62, P < 0.003). As expected, when all tissues were considered, there was no correlation of tissue selenoprotein W concentration with tissue selenium, tissue cystosol selenium or with tissue GPX activity. White muscle disease is a selenium deficiency myopathy in which lesions occur in the muscle and heart (Schubert et al. 1961). Because selenoprotein W is highest in these organs in sheep given selenium, it is tempting to speculate that it is involved in the prevention of this disorder in these organs.

The deficient and 3 mg selenium/kg diets were compared to determine whether there was a difference in selenoprotein W levels in sheep fed such widely differing amounts. Now that it has been established that selenoprotein W levels increase with supernutritional levels, a level routinely used in carcinogenic studies (Ip and Lisk, 1994), nutritional levels will now be tested. We have observed that in rats selenoprotein W levels in muscle reach a plateau only after 1 mg selenium/kg diet is fed. It will be of interest to determine whether similar patterns are obtained with sheep.

There were no differences (P > 0.05) in the brain selenoprotein W levels between the sheep fed the deficient diet and those given selenium (Fig. 6), but there was a 53% difference in the total brain selenium concentration (Fig. 7). The GPX activity was 30% lower (P < 0.05) in this organ of the deficient sheep than in that of supplemented sheep (Fig. 8). The regulation of selenium and selenoprotein synthesis in brain appears to be different from those in other tissues. Selenium status will have an influence on selenoprotein levels. For example, Davidson and Kennedy (1993) indicated that there is greater synthesis of selenoproteins in deficient than in supplemented sheep. Because our study showed that selenoprotein W content in brain was not affected by selenium deficiency or supplementation, this organ may regulate its selenoprotein levels to compensate for deficiency or excess of selenium. The respective decrease of GPX activity and selenium concentration by 30 and 53% in brain of deficient sheep indicates a preferential retention of selenoprotein W. Therefore, the regulations of selenoprotein W and GPX are different in brain and further studies are required to investigate the importance of these observations.

Neither the tissue distribution nor its characteristics offer any clues to the metabolic functions of selenoprotein W. It contains one gram atom of selenium per mole protein (Vendeland et al. 1993), and reduced glutathione is bound to two species of this protein (Beilstein et al. 1996). Because all of the selenoenzymes which have been identified thus far are involved in redox reactions (Burk and Hill 1993, Sunde 1990), it is tempting to postulate a role for selenoprotein W as an antioxidant, especially because glutathione is one of its binding moieties.

Selenoprotein W is one of two mammalian selenoproteins that have been well characterized without an identified function. The other, selenoprotein P, has been studied to the greatest extent. The majority of the selenium in plasma of rats and humans (Deagen et al. 1993) is associated with this protein. This selenoprotein contains the highest concentration of selenium of any known protein with 10 selenocysteine residues per mole protein (Hill et al. 1991). Similar to selenoprotein W in tissues, the level of selenoprotein P in plasma is affected by the selenium status of the animal (Yang et al. 1989).

Two families of GPX and deiodinases are the only mammalian selenoenzymes presently known. The cellular GPX was recognized over 20 y ago (Rotruck et al. 1973), and the plasma GPX was clearly shown to be a separate gene product from the cellular protein (Takahashi et al. 1990). A third GPX was identified which reduces fatty acid hydroperoxides esterified to phospholipids and which is assumed to play an important role in reducing peroxides in membranes (Schuckelt et al. 1991). A fourth GPX has been found predominantly in the gastrointestinal mucosa (Chu et al. 1993). Type I was the first deiodinase shown to be a selenoenzyme (Behne et al. 1990, Berry et al. 1991). Types II (Davey et al. 1995) and III (Croteau et al. 1995) have also recently been shown to be selenoenzymes. These deiodinases provide a metabolic link between selenium and iodine. The relationship of selenoprotein W to selenoenzymes is not known, but the present results with selenoprotein W and GPX activity indicate that they differ in tissue distribution.

Up to 13 selenium-containing proteins have been reported in rat tissues (Behne et al. 1988) and up to 9 have been reported in ovine tissues (Davidson and Kennedy 1993). Up to four selenium-containing proteins were found in ovine heart and muscle. Thus far, only GPX and selenoprotein W have been identified in these tissues, which suggests that there are at least two more to be identified. Presumably the deiodinases could be one of them. More selenium-containing proteins were present in other tissues such as the liver and pituitary, providing sufficient reason to believe that the identification of additional selenoproteins is forthcoming.