Response of rat selenoprotein and fate of its selenium RAYMOND
F. BURK,
KRISTINA
P to selenium
E. HILL,
ROBERT
READ,
administration
AND
TERRI
BELLEW
Division of Gastroenterology, Department of Medicine, and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, Tennessee37232
F., KRISTINA E. HILL, ROBERT READ, AND Response of rat selenoprotein P to selenium administration and fate of its selenium. Am. J. Physiol. 261 (Endocrinol. Metab. 24): E26-E30, 1991.-Selenoprotein P is a glycoprotein that contains ~60% of the selenium in rat plasma. Physiological experiments were undertaken to gain insight into selenoprotein P function. Selenium-deficient rats were injected with doses of selenium ranging from 25 to 200 pg/kg, and the appearance of selenoprotein P was compared with the appearance of glutathione peroxidase activity in plasma and in liver. Selenoprotein P concentration increased to 35% of control by 6 h, whereas glutathione peroxidase activity increased minimally or not at all. Moreover, in rats given 100 and 200 PcLgselenium/kg, selenoprotein P reached 75% of its concentration in control rats at 24 h, whereas glutathione peroxidase activity reached only 6% of control. Cycloheximide pretreatment blocked the appearance of selenoprotein P in response to selenium injection. Male and female rats had similar concentrations of selenoprotein P. Partially purified selenoprotein P and plasma glutathione peroxidase labeled with 7”Se were administered intravenously to seleniumdeficient and control rats. 7”Se given as selenoprotein P disappeared more rapidly from plasma than did 7”Se given as glutathione peroxidase. Selenium deficiency did not significantly affect 7”Se disappearance from plasma. At 2 h, brain, but not other tissues, took up more 75Se in selenium-deficient rats than in control rats when 7”Se was given as selenoprotein P. This suggests that brain has a specific uptake mechanism for selenium given in the form of selenoprotein P. These results demonstrate that several physiological properties distinguish selenoprotein P from glutathione peroxidase. However, they do not clearly indicate its function. BURK, RAYMOND TERRI BELLEW.
plasma selenium; nium uptake
plasma
glutathione
peroxidase;
brain
sele-
P is a plasma selenoprotein that has recently been purified from rats by immunoaffinity chromatography (12). It is a glycoprotein that contains seven or more selenium atoms per molecule in the form of selenocysteine (8). Selenoprotein P is quantitatively important because it accounts for ~60% of plasma selenium in rats with normal selenium status (8). The function of selenoprotein P is not known. Suggestions have been advanced that it is involved in selenium transport (7) and that it is an oxidant defense protein (12). The present paper reports studies of selenoprotein P metabolism. They were carried out with the hope that they would provide insight into its function.
SELENOPROTEIN
E26
0193-1849/91
$1.50 Copyright
METHODS
Animals. Weanling Sprague-Dawley rats were fed a selenium-deficient diet or the same diet containing 0.5 mg selenium/kg as sodium selenate (2) ad libitum for 8 wk or longer. Table 1 indicates that the experimental diet produces selenium deficiency. Except where specified, rats were males. They were housed in a room with a 12:12-h light-dark cycle. Tap water was provided ad libitum. Rats were anesthetized with pentobarbital sodium (65 mg/kg ip) and exsanguinated by aortic puncture. After removal, using a syringe and needle, blood was treated with disodium EDTA (1 mg/ml) to prevent coagulation, and plasma was separated by centrifugation. For the dose-response and protein synthesis inhibition experiments, selenium was injected intraperitoneally in the form of sodium selenite dissolved in normal saline. Injection volume was 1 ml/kg in the dose-response experiment and 1 ml in the protein synthesis inhibition experiment. Cycloheximide (5 mg/kg) was administered intraperitoneally in saline (1 ml/kg) 30 min before selenium injection and 2 and 4 h after it to inhibit protein synthesis (11). Half-life and tissue distribution experiment. Selenoprotein P and plasma glutathione peroxidase labeled with 75Se were used for these studies. A stock solution of 75Selabeled sodium selenite was prepared. The specific activity was 0.67 mCi/mg, and it was divided into aliquots and frozen. Each day for 20 days an aliquot was thawed, and 1 ml (10 pg selenium) was injected intraperitoneally into each of six selenium-deficient adult male rats. After an overnight fast and 24 h after the last injection, plasma was collected from the rats. Thus the selenium-deficient rats were repleted with labeled selenium so 75Se could be used to quantitate selenium. Labeled plasma from the repleted animals was passed over an Affi-Gel Blue column with clean separation of glutathione peroxidase from selenoprotein P (12). The respective peaks were pooled and concentrated. The buffer was changed to phosphate-buffered saline [(in mM) 137 NaCl, 2.6 KCl, and 10 sodium phosphate, pH 7.41 by gel filtration on a Sephacryl S-200-HR column. The volumes were adjusted so that the two preparations had the same 75Se concentration. The total selenium concentrations of these two protein preparations should have been very similar, because the growing rats were repleted only with labeled selenite for the 20 days before death. A similar concentration of the original 75Se-so-
0 1991 the American
Physiological
Society
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (018.218.056.169) on August 13, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
SELENOPROTEIN
1. Effect of gender and selenium status on selenoprotein P concentration in rat plasma TABLE
Selenoprotein Selenium Control Glutathione Selenium Control
P, pg/ml deficient peroxidase, deficient
Female
Male
3.6t0.3 31.2,t3.0
33.9s.4
58t6 1,820+140
3421 2,360+450
E27
P METABOLISM
2.1kO.6
units/ml
Values are means & SD, n = 4 rats. Rats had been fed respective diets for 8 wk from time of weaning. A unit of glutathione peroxidase is 1 nmol NADPH oxidized/min.
dium selenite was also prepared. It was used only for the 24-h time point (Table 2). The labeled material was injected in a volume of -0.5 ml into the femoral vein while the rat was anesthetized. Syringes were weighed before and after injection, and standards were prepared by injection into counting tubes. This allowed a determination of the dose injected into each rat. Counting standards were included with each batch of samples to allow correction for radioactive decay. The dose of selenium injected was 6 ng for the 5 min experiment and 39-44 ng for the other time points. The label injected varied from 10,000 to 15,000 counts/ min (cpm) per rat. The experiment that yielded the 5min time point was done several months before the experiment that yielded all the other time points and utilized a separate preparation of labeled selenoproteins. Blood and tissue collections were carried out under pentobarbital anesthesia. Urine was collected in metabolic cages from the rats that were killed at 24 h. Four rats were studied for each dietary group at each time point.
substrate (6). Protein was determined using the Bradford method (1). Selenoprotein P was determined in plasma by a competitive radioimmunoassay (12) using monoclonal antibody 8Fll (8). 75Se was determined using a Packard model 5230 scintillation counter (Packard Instruments, Downers Grove, IL) with a counting efficiency of -50%. Statistics. Data were analyzed with an Apple Macintosh SE using Statview SE+ graphics (Abacus Concepts, Berkeley, CA). Selenium-deficient and control values were compared using an unpaired t test. Slopes of lines were compared according to Steel and Torrie (10) for statistical significance (P < 0.05). Fisher’s protected least-significant difference test was used to determine statistical differences (~0.05) between time points in Fig. 1. Chemicals. 7”Se as H&Se03 was purchased from Du Pont-New England Nuclear Products (Boston, MA). Affi-Gel Blue was purchased from Bio-Rad Laboratories (Richmond, CA). Other chemicals were American Chemical Society grade or higher. RESULTS
Dose response of selenoprotein P to selenium injection.
Selenium-deficient rats were administered doses of selenium ranging from 25 to 200 pg/kg. Selenoprotein P in plasma and glutathione peroxidase in plasma and liver were measured at 6, 12, and 24 h after dosing. Figure 1 shows the results. The doses of 100 and 200 pg/kg were equivalent for the glutathione peroxidases and selenoprotein P, suggesting that the organism could not utilize X00 pg of selenium/kg as a single injection. Figure 1A shows that selenoprotein P increased significantly by 6 Tissue uptake of selenium from different dosesof sele- h after all injections. Glutathione peroxidase activity did noprotein P. 75Se-labeled selenoprotein P was prepared not increase significantly in the liver until after 6 h (Fig. as described above, and the experiment was carried out 1B) or in the plasma until after 12 h (Fig. 1C). Selenoprotein P concentration reached 75% of control by 24 h, in a similar fashion to the half-life experiment but with a single time point of 2 h. The stock sodium selenite whereas glutathione peroxidase activity increased only solution used to replete the rats had a higher 75Se specific to 6% of control. This demonstrates a preferential and activity than the one used in the half-life experiment. more rapid utilization of selenium for synthesis of seleThis produced selenoprotein P with a higher specific noprotein P in comparison with glutathione peroxidase. Role of protein synthesis in selenium incorporation into activity and allowed the study of smaller amounts of the protein. The label injected varied from 13,000 to selenoprotein P. The role of protein synthesis in the 1,500,OOOcpm in the three groups. The volume injected appearance of selenoprotein P after selenium injection was -0.5 ml, and three rats were studied for each dietary was studied using the protein synthesis inhibitor cyclogroup at each dose. heximide. Selenium-deficient rats had a plasma selenoAssays. Glutathione peroxidase was measured as de- protein P concentration that was 4.6 t 0.7% (n = 4) of scribed before using 0.25 mM hydrogen peroxide as a control. Six hours after injection of 200 pg/kg selenium, TABLE
2. Effect of form given and selenium status on distribution of 75Se24 h after administration Selenoprotein
Plasma Liver Kidney Heart Muscle (leg) Urine (24 h)
P
Glutathione
Peroxidase
Selenite
Selenium deficient
Control
Selenium deficient
Control
Selenium deficient
Control
l.OOt0.14 1.08t0.13" 25320.41 0.38-eO.08 0.05t0.04 0.83kO.31”
1.21kO.38 0.67~0.11" 2.25-t-0.52 0.34-c-0.07 0.06kO.02 7.52-t-1.97”
3.09t0.45 0.86t0.23 2.81kO.38" 0.80t0.12 0.04-t-0.04 0.68+0.2gf
3.08kO.39 0.64t0.10 1.81+0.20b 0.72t0.10 0.06kO.07 4.14kO.54'
0.83t0.22 1.48t0.44 2.98k0.25 0.60t0.02' 0. 10+O.Old 7.42t2.37"
1.06t0.08 1.30to.13 2.86t0.28 0.36kO.01' 0.05+0.01d 25.8t3.F
Values are means & SD in %dose/ml or g except for urine, which Figs. 2 and 3. Control rats weighed 348 t 17 g and selenium-deficient 0.05) by unpaired t test.
shown in represents %excreted in 24 h; n = 4. Data are from experiment rats weighed 267 t 29 g. Pairs with same superscript are different (P c
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (018.218.056.169) on August 13, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
E28
SELENOPROTEIN
25
-80
-60 -40
-20
-0
-0
-
80.
o------o -
25pg -
5009 IOOpg 2OOpg
Se/kg - .
Se/kg Se/kg Se/
T
P METABOLISM
necessary for selenium to effect an increase in selenoprotein P concentration. Selenoprotein P in female rats. The concentration of selenoprotein P in female rat plasma was compared with that in male rat plasma. Rats had been fasted overnight to avoid lipemic plasma. Overnight fasting does not affect selenoprotein P levels (Read and Burk, unpublished observations). Table I shows that male and female selenoprotein P values were comparable in animals fed the control diet and reflected the selenium status of the animals fed the selenium-deficient diet. Plasma glutathione peroxidase activities were similar. Thus plasma concentration of selenoprotein P is not affected by gender Half-life in blood and tissue distribution of selenium administered as selenoprotein P and as glutathione perof oxidase. Partially purified 75Se-labeled preparations selenoprotein P and plasma glutathione peroxidase were administered intravenously to selenium-deficient and control rats. The amounts of the preparations were adjusted so that the same quantity of 75Se was given in each form. Figure 2 shows that 75Se administered as selenoprotein P disappeared from blood faster (initial half-life 3-4 h) than 75Se administered as glutathione peroxidase (initial half-life -12 h). The selenium status of the recipient animal did not significantly affect the rate of 75Se disappearance. The similar disappearance rates in selenium-deficient and control animals suggests
-6
60,
n
hours
after
injection
FIG. 1. Time course of selenoprotein appearance after injection of selenium-deficient rats with different doses of selenium. Results in A and C were obtained from assays of plasma. Results in B were obtained from assay of 105,000 g supernatant of liver (6). Values are means & SD, n = 4. Values at 6,12, and 24 h were tested for significant difference from 0 time values using Fisher’s protected least-significant difference test. All 6-, 12, and 24-h values in A were significantly different from 0 time value. In B none of the 6-h values but all of the 12- and 24-h values were significantly different from 0 time value. In C all 24-h values were significantly different from 0 time value. Two of 6-h (25 pg/kg and 100 pg/kg) and 1 of 12-h (200 pg/kg) values were significantly different from 0 time value, but these differences reached significance by a small margin. Control rat values (n = 4) were as follows: selenoprotein P, 32 t 1.5 pg/ml plasma; liver 105,000 g supernatant glutathione peroxidase activity, 228 t 32 nmol NADPH oxidized. mg protein-‘. min-‘; plasma glutathione peroxidase activity, 1,208 t 126 nmol NADPH oxidized ml-‘. min. l
the selenoprotein P concentration had increased to 34 t 6.7% (n = 4) of control. Pretreatment with cycloheximide before the selenium injection abolished the increase. Selenoprotein P concentration was only 2.9 t 0.6% (n = 4) of control in that group. Thus protein synthesis is
Smin
4h
8h
time after
24h
75 Se injection
of “Se as FIG. 2. Disappearance of “‘Se from blood after injection selenoprotein P (circles) or glutathione peroxidase (squares). Open symbols represent selenium-deficient animals and closed symbols represent control animals. The 5-min time point was studied in a separate experiment from other time points. Greater than 90% of administered 7’Se was present in blood at 5 min in all groups when blood was taken to be 6.4% of body weight. Values are means t SD, n = 4.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (018.218.056.169) on August 13, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
--*
--
Sl!iLl.!iN
-----m---I
p
E29
METABOLISM
Ul’lW’l’ElN
that the turnover of the protein is mediated largely by nonspecific processes. The association of 75Se with tissues was determined. Figure 3A shows the results in the brain. The brain exhibited a much higher uptake of label in seleniumdeficient rats given selenoprotein P than in controls. This difference was significant (P < 0.05) 5 min after the injection of label and reached 12fold at 8 h. No such effect was found when the label was given as glutathione peroxidase, except for a small difference at 24 h, which could have been caused by recycling of the label through selenoprotein P. These results suggest that there is a specific mechanism in the brain for taking up selenium in the form of selenoprotein P. Figure 3B shows the association of 75Se with the testis. The uptake of label was greater by selenium-deficient animals than by controls for both protein sources. Uptake when selenoprotein P was the source of label was greater than when glutathione peroxidase was the source. These results are compatible with the uptake of label after it was released by turnover of the proteins. Table 2 shows the disposition of label 24 h after its intravenous administration as selenoprotein P, plasma glutathione peroxidase, and selenite. Selenium-deficient heart and skeletal muscle took up more 75Se than did controls when the label was given as selenite but not when it was given as selenoprotein P or as glutathione peroxidase. This suggests that a form of selenium other than these two proteins might facilitate the uptake by selenium-deficient muscle. The kidney took up a relatively large amount of 75Se, and selenium deficiency accentuated the uptake from glutathione peroxidase. Urinary excretion of 75Se was less after administration in the form of protein than in the form of selenite. Effect of selenium deficiency on tissue uptake of selenium from different doses of selenoprotein P. Selenium
deficiency caused the brain to take up a greater percentage of 75Se administered as selenoprotein P than was taken up by brain in controls (Fig. 3A). To investigate this further, an experiment was carried out to determine uptake at 2 h. This time point was chosen to reflect early events in uptake and to minimize the recycling of label. Based on experiments that established the selenoprotein P content of selenium in control and seleniumdeficient rats (8) and a plasma volume of 4.0% 4.0% body weight, these control rats contained 4,060 ng selenium as selenoprotein P, and the selenium-deficient ones contained 197 ng or less. Three dose levels of 75Se-selenoprotein P were given. They contained 1, 10, and 100 ng selenium. These doses ranged up to 2.5% 2.5% of the selenoprotein P pool size in controls but up to 50% or more of the pool size in selenium-deficient rats. Tissue uptake of selenium was calculated from 75Se content. Figure 4 shows the results. Tissue uptake of selenium by blood, kidney, liver, and testis was linearly related to the amount administered, and there was no effect of selenium deficiency (Fig. 4B). In contrast, the uptake by brain was severalfold higher in selenium-deficient rats than in control rats for each dose. Moreover, the percentage of 75Se uptake by selenium-deficient brain appeared to decrease with the highest dose. The administration of labeled selenoprotein P leads to a pool of the protein with a higher specific activity in selenium-deficient rats than in control rats, because control plasma contains much more unlabeled selenoprotein P than does selenium-deficient plasma. Thus the higher uptake of the administered selenium by the selenium-deficient brain than by the control brain does not necessarily indicate a more avid uptake of selenoprotein P. It does indicate a specific uptake by the brain, however, in contrast to the nonspecific uptake by the tissues shown in Fig. 4B. DISCUSSION
A
Selenoprotein P was identified because it contained selenium. No enzymatic activity or biological function of
selenium 24h
time after
75
Se injection
FIG. 3. 75Se in brain (A) and in testis (B) after administration as selenoprotein P (circles) or glutathione peroxidase (squares). Open symbols represent selenium-deficient animals and closed symbols represent control animals. These results are from same experiments described in Fig. 2, and values shown are means & SD, n = 4.
injected
as selenoprotein
P (ng)
FIG. 4. Tissue uptake at 2 h of 75Se from different doses of selenoprotein P. Open symbols represent selenium-deficient animals and closed symbols represent control animals. Values are means t, SD, n = 3. A: selenium-deficient and control values were significantly different (P < 0.05) at each dose level. B: selenium-deficient and control values for each tissue at each dose level were not significantly different. Lines defined by l- and lo-ng and lo- and lOO-ng selenium points, respectively, in selenium-deficient brain (A) were significantly different.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (018.218.056.169) on August 13, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
E30
SELENOPROTEIN
the protein has been demonstrated. Several of its properties are compatible with a role in selenium transport. First, it is a plasma protein that incorporates selenium very rapidly after administration (3, 7). Second, the appearance of selenium in it precedes the incorporation of the element into tissues (7). Third, in vitro evidence for receptors in various tissues has been presented (5), suggesting a mechanism of uptake from plasma. However, other properties of selenoprotein P make a transport role less likely. First, the selenium in the protein is covalently bound in its primary structure (8) and thus would presumably require catabolism of the protein for its release. Second, Fig. 2 shows that the selenium status of the animal has little effect on the disappearance from the plasma of 75Se given as selenoprotein P. If selenoprotein P were a transport protein, it would be expected to release its selenium more rapidly in selenium deficiency. It is clear that selenium given as selenoprotein P leaves the circulation and appears in tissues throughout the body (Table 2). With the exception of uptake by the brain (Fig. 3), the initial uptake does not appear to be specific. However, selenoprotein P has been shown to have selenium-rich regions (8), and it is possible that selenoprotein P is broken down to an intermediate form that is a true transport form of the element. Further work is needed to clarify how selenium is transported in the body. Glutathione peroxidase and the bacterial selenoproteins (i.e., proteins containing selenium in the form of selenocysteine) all have redox functions (9), and it seems reasonable to postulate such a function for selenoprotein P. Selenoprotein P has seven or more selenols and approximately the same number of thiols (8). Based on immunodepletion studies (8), the protein should provide 4 PM selenol and 4 PM thiol in rat plasma. These reactive groups could participate in oxidation-reduction reactions. Evidence has been presented that selenium has oxidant defense properties other than its function in glutathione peroxidase (4). The results of the present experiments, namely the more avid incorporation of selenium into selenoprotein P than into glutathione peroxidase and the relatively long residence of selenoprotein P in the plasma in selenium deficiency, are compatible with selenoprotein P functioning in the vascular bed. Further work will be needed to determine whether selenoprotein P can protect against oxidant molecules in the circulation. These studies contrast selenoprotein P metabolism with that of glutathione peroxidase. They provide further
P METABOLISM
evidence that the synthesis of selenoprotein P has a higher priority for the organism when selenium is limiting than the synthesis of glutathione peroxidase. Although they demonstrate that selenoprotein P turns over more rapidly in the plasma than glutathione peroxidase and that it facilitates selenium uptake by brain, no evidence was found that selenoprotein P transports selenium to other tissues in a specific manner. Thus selenoprotein P might somehow be involved in selenium transport, but transport does not appear to be its direct function. We are indebted to Robert W. Hunt, Jr., for technical assistance. This work was supported by National Institute of Environmental Health Sciences Grants ES-02497 and ES-00267. Address for reprint requests: R. F. Burk, Div. of Gastroenterology, C-2104, Medical Center North, Vanderbilt Medical Center, Nashville, TN 37232. Received
24 September
1990; accepted
in final
form
26 February
1991.
REFERENCES 1. BRADFORD, M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248-254, 1976. 2. BURK, R. F. Production of selenium deficiency in the rat. Methods Enzymol. 143: 307-313,1987. 3. BURK, R. F., AND P. E. GREGORY. Some characteristics of 75Se-P, a selenoprotein found in rat liver and plasma, and comparison of it with selenoglutathione peroxidase. Arch. Biochem. Biophys. 213: 73-80,1982. 4. BURK, R. F., R. A. LAWRENCE, AND J. M. LANE. Liver necrosis and lipid peroxidation in the rat as the result of paraquat and diquat administration. Effect of selenium deficiency. J. Clin. Invest. 65: 1024-1031, 1980. 5. GOMEZ, B., AND A. L. TAPPEL. Selenoprotein P receptor from rat. Biochim. Biophys. Acta 979: 20-26, 1989. 6. LAWRENCE, R. A., AND R. F. BURK. Glutathione peroxidase in selenium-deficient liver. Biochem. Biophys. Res. Commun. 71: 952958,1976. 7. MOTSENBOCKER, M. A., AND A. L. TAPPEL. A selenocysteinecontaining selenium-transport protein in rat plasma. Biochim. Biophys. Acta 719: 147-153, 1982. 8. READ, R., T. BELLEW, J.-G. YANG, K. E. HILL, I. S. PALMER, AND R. F. BURK. Selenium and amino acid composition of selenoprotein P, the major selenoprotein in rat serum. J. BioZ. Chem. 265: 1789917905,199o. 9. STADTMAN, T. C. Selenium biochemistry. Annu. Rev. Biochem. 59: ill-127,199O. 10. STEEL, R. G. D., AND J. H. TORRIE. Principles and Procedures of Statistics. New York: McGraw-Hill, 1960, p. 173. 11. SUTTIE, J. W. The effect of cycloheximide administration on vitamin K-stimulated prothrombin formation. Arch. Biochem. Biophys. 141: 571-578,197O. 12. YANG, J.-G., J. MORRISON-PLUMMER, AND R. F. BURK. Purification and quantitation of a rat plasma selenoprotein distinct from glutathione peroxidase using monoclonal antibodies. J. Biol. Chem. 262: 13372-13375,1987.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (018.218.056.169) on August 13, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
F. BURK,
KRISTINA
P to selenium
E. HILL,
ROBERT
READ,
administration
AND
TERRI
BELLEW
Division of Gastroenterology, Department of Medicine, and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, Tennessee37232
F., KRISTINA E. HILL, ROBERT READ, AND Response of rat selenoprotein P to selenium administration and fate of its selenium. Am. J. Physiol. 261 (Endocrinol. Metab. 24): E26-E30, 1991.-Selenoprotein P is a glycoprotein that contains ~60% of the selenium in rat plasma. Physiological experiments were undertaken to gain insight into selenoprotein P function. Selenium-deficient rats were injected with doses of selenium ranging from 25 to 200 pg/kg, and the appearance of selenoprotein P was compared with the appearance of glutathione peroxidase activity in plasma and in liver. Selenoprotein P concentration increased to 35% of control by 6 h, whereas glutathione peroxidase activity increased minimally or not at all. Moreover, in rats given 100 and 200 PcLgselenium/kg, selenoprotein P reached 75% of its concentration in control rats at 24 h, whereas glutathione peroxidase activity reached only 6% of control. Cycloheximide pretreatment blocked the appearance of selenoprotein P in response to selenium injection. Male and female rats had similar concentrations of selenoprotein P. Partially purified selenoprotein P and plasma glutathione peroxidase labeled with 7”Se were administered intravenously to seleniumdeficient and control rats. 7”Se given as selenoprotein P disappeared more rapidly from plasma than did 7”Se given as glutathione peroxidase. Selenium deficiency did not significantly affect 7”Se disappearance from plasma. At 2 h, brain, but not other tissues, took up more 75Se in selenium-deficient rats than in control rats when 7”Se was given as selenoprotein P. This suggests that brain has a specific uptake mechanism for selenium given in the form of selenoprotein P. These results demonstrate that several physiological properties distinguish selenoprotein P from glutathione peroxidase. However, they do not clearly indicate its function. BURK, RAYMOND TERRI BELLEW.
plasma selenium; nium uptake
plasma
glutathione
peroxidase;
brain
sele-
P is a plasma selenoprotein that has recently been purified from rats by immunoaffinity chromatography (12). It is a glycoprotein that contains seven or more selenium atoms per molecule in the form of selenocysteine (8). Selenoprotein P is quantitatively important because it accounts for ~60% of plasma selenium in rats with normal selenium status (8). The function of selenoprotein P is not known. Suggestions have been advanced that it is involved in selenium transport (7) and that it is an oxidant defense protein (12). The present paper reports studies of selenoprotein P metabolism. They were carried out with the hope that they would provide insight into its function.
SELENOPROTEIN
E26
0193-1849/91
$1.50 Copyright
METHODS
Animals. Weanling Sprague-Dawley rats were fed a selenium-deficient diet or the same diet containing 0.5 mg selenium/kg as sodium selenate (2) ad libitum for 8 wk or longer. Table 1 indicates that the experimental diet produces selenium deficiency. Except where specified, rats were males. They were housed in a room with a 12:12-h light-dark cycle. Tap water was provided ad libitum. Rats were anesthetized with pentobarbital sodium (65 mg/kg ip) and exsanguinated by aortic puncture. After removal, using a syringe and needle, blood was treated with disodium EDTA (1 mg/ml) to prevent coagulation, and plasma was separated by centrifugation. For the dose-response and protein synthesis inhibition experiments, selenium was injected intraperitoneally in the form of sodium selenite dissolved in normal saline. Injection volume was 1 ml/kg in the dose-response experiment and 1 ml in the protein synthesis inhibition experiment. Cycloheximide (5 mg/kg) was administered intraperitoneally in saline (1 ml/kg) 30 min before selenium injection and 2 and 4 h after it to inhibit protein synthesis (11). Half-life and tissue distribution experiment. Selenoprotein P and plasma glutathione peroxidase labeled with 75Se were used for these studies. A stock solution of 75Selabeled sodium selenite was prepared. The specific activity was 0.67 mCi/mg, and it was divided into aliquots and frozen. Each day for 20 days an aliquot was thawed, and 1 ml (10 pg selenium) was injected intraperitoneally into each of six selenium-deficient adult male rats. After an overnight fast and 24 h after the last injection, plasma was collected from the rats. Thus the selenium-deficient rats were repleted with labeled selenium so 75Se could be used to quantitate selenium. Labeled plasma from the repleted animals was passed over an Affi-Gel Blue column with clean separation of glutathione peroxidase from selenoprotein P (12). The respective peaks were pooled and concentrated. The buffer was changed to phosphate-buffered saline [(in mM) 137 NaCl, 2.6 KCl, and 10 sodium phosphate, pH 7.41 by gel filtration on a Sephacryl S-200-HR column. The volumes were adjusted so that the two preparations had the same 75Se concentration. The total selenium concentrations of these two protein preparations should have been very similar, because the growing rats were repleted only with labeled selenite for the 20 days before death. A similar concentration of the original 75Se-so-
0 1991 the American
Physiological
Society
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (018.218.056.169) on August 13, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
SELENOPROTEIN
1. Effect of gender and selenium status on selenoprotein P concentration in rat plasma TABLE
Selenoprotein Selenium Control Glutathione Selenium Control
P, pg/ml deficient peroxidase, deficient
Female
Male
3.6t0.3 31.2,t3.0
33.9s.4
58t6 1,820+140
3421 2,360+450
E27
P METABOLISM
2.1kO.6
units/ml
Values are means & SD, n = 4 rats. Rats had been fed respective diets for 8 wk from time of weaning. A unit of glutathione peroxidase is 1 nmol NADPH oxidized/min.
dium selenite was also prepared. It was used only for the 24-h time point (Table 2). The labeled material was injected in a volume of -0.5 ml into the femoral vein while the rat was anesthetized. Syringes were weighed before and after injection, and standards were prepared by injection into counting tubes. This allowed a determination of the dose injected into each rat. Counting standards were included with each batch of samples to allow correction for radioactive decay. The dose of selenium injected was 6 ng for the 5 min experiment and 39-44 ng for the other time points. The label injected varied from 10,000 to 15,000 counts/ min (cpm) per rat. The experiment that yielded the 5min time point was done several months before the experiment that yielded all the other time points and utilized a separate preparation of labeled selenoproteins. Blood and tissue collections were carried out under pentobarbital anesthesia. Urine was collected in metabolic cages from the rats that were killed at 24 h. Four rats were studied for each dietary group at each time point.
substrate (6). Protein was determined using the Bradford method (1). Selenoprotein P was determined in plasma by a competitive radioimmunoassay (12) using monoclonal antibody 8Fll (8). 75Se was determined using a Packard model 5230 scintillation counter (Packard Instruments, Downers Grove, IL) with a counting efficiency of -50%. Statistics. Data were analyzed with an Apple Macintosh SE using Statview SE+ graphics (Abacus Concepts, Berkeley, CA). Selenium-deficient and control values were compared using an unpaired t test. Slopes of lines were compared according to Steel and Torrie (10) for statistical significance (P < 0.05). Fisher’s protected least-significant difference test was used to determine statistical differences (~0.05) between time points in Fig. 1. Chemicals. 7”Se as H&Se03 was purchased from Du Pont-New England Nuclear Products (Boston, MA). Affi-Gel Blue was purchased from Bio-Rad Laboratories (Richmond, CA). Other chemicals were American Chemical Society grade or higher. RESULTS
Dose response of selenoprotein P to selenium injection.
Selenium-deficient rats were administered doses of selenium ranging from 25 to 200 pg/kg. Selenoprotein P in plasma and glutathione peroxidase in plasma and liver were measured at 6, 12, and 24 h after dosing. Figure 1 shows the results. The doses of 100 and 200 pg/kg were equivalent for the glutathione peroxidases and selenoprotein P, suggesting that the organism could not utilize X00 pg of selenium/kg as a single injection. Figure 1A shows that selenoprotein P increased significantly by 6 Tissue uptake of selenium from different dosesof sele- h after all injections. Glutathione peroxidase activity did noprotein P. 75Se-labeled selenoprotein P was prepared not increase significantly in the liver until after 6 h (Fig. as described above, and the experiment was carried out 1B) or in the plasma until after 12 h (Fig. 1C). Selenoprotein P concentration reached 75% of control by 24 h, in a similar fashion to the half-life experiment but with a single time point of 2 h. The stock sodium selenite whereas glutathione peroxidase activity increased only solution used to replete the rats had a higher 75Se specific to 6% of control. This demonstrates a preferential and activity than the one used in the half-life experiment. more rapid utilization of selenium for synthesis of seleThis produced selenoprotein P with a higher specific noprotein P in comparison with glutathione peroxidase. Role of protein synthesis in selenium incorporation into activity and allowed the study of smaller amounts of the protein. The label injected varied from 13,000 to selenoprotein P. The role of protein synthesis in the 1,500,OOOcpm in the three groups. The volume injected appearance of selenoprotein P after selenium injection was -0.5 ml, and three rats were studied for each dietary was studied using the protein synthesis inhibitor cyclogroup at each dose. heximide. Selenium-deficient rats had a plasma selenoAssays. Glutathione peroxidase was measured as de- protein P concentration that was 4.6 t 0.7% (n = 4) of scribed before using 0.25 mM hydrogen peroxide as a control. Six hours after injection of 200 pg/kg selenium, TABLE
2. Effect of form given and selenium status on distribution of 75Se24 h after administration Selenoprotein
Plasma Liver Kidney Heart Muscle (leg) Urine (24 h)
P
Glutathione
Peroxidase
Selenite
Selenium deficient
Control
Selenium deficient
Control
Selenium deficient
Control
l.OOt0.14 1.08t0.13" 25320.41 0.38-eO.08 0.05t0.04 0.83kO.31”
1.21kO.38 0.67~0.11" 2.25-t-0.52 0.34-c-0.07 0.06kO.02 7.52-t-1.97”
3.09t0.45 0.86t0.23 2.81kO.38" 0.80t0.12 0.04-t-0.04 0.68+0.2gf
3.08kO.39 0.64t0.10 1.81+0.20b 0.72t0.10 0.06kO.07 4.14kO.54'
0.83t0.22 1.48t0.44 2.98k0.25 0.60t0.02' 0. 10+O.Old 7.42t2.37"
1.06t0.08 1.30to.13 2.86t0.28 0.36kO.01' 0.05+0.01d 25.8t3.F
Values are means & SD in %dose/ml or g except for urine, which Figs. 2 and 3. Control rats weighed 348 t 17 g and selenium-deficient 0.05) by unpaired t test.
shown in represents %excreted in 24 h; n = 4. Data are from experiment rats weighed 267 t 29 g. Pairs with same superscript are different (P c
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (018.218.056.169) on August 13, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
E28
SELENOPROTEIN
25
-80
-60 -40
-20
-0
-0
-
80.
o------o -
25pg -
5009 IOOpg 2OOpg
Se/kg - .
Se/kg Se/kg Se/
T
P METABOLISM
necessary for selenium to effect an increase in selenoprotein P concentration. Selenoprotein P in female rats. The concentration of selenoprotein P in female rat plasma was compared with that in male rat plasma. Rats had been fasted overnight to avoid lipemic plasma. Overnight fasting does not affect selenoprotein P levels (Read and Burk, unpublished observations). Table I shows that male and female selenoprotein P values were comparable in animals fed the control diet and reflected the selenium status of the animals fed the selenium-deficient diet. Plasma glutathione peroxidase activities were similar. Thus plasma concentration of selenoprotein P is not affected by gender Half-life in blood and tissue distribution of selenium administered as selenoprotein P and as glutathione perof oxidase. Partially purified 75Se-labeled preparations selenoprotein P and plasma glutathione peroxidase were administered intravenously to selenium-deficient and control rats. The amounts of the preparations were adjusted so that the same quantity of 75Se was given in each form. Figure 2 shows that 75Se administered as selenoprotein P disappeared from blood faster (initial half-life 3-4 h) than 75Se administered as glutathione peroxidase (initial half-life -12 h). The selenium status of the recipient animal did not significantly affect the rate of 75Se disappearance. The similar disappearance rates in selenium-deficient and control animals suggests
-6
60,
n
hours
after
injection
FIG. 1. Time course of selenoprotein appearance after injection of selenium-deficient rats with different doses of selenium. Results in A and C were obtained from assays of plasma. Results in B were obtained from assay of 105,000 g supernatant of liver (6). Values are means & SD, n = 4. Values at 6,12, and 24 h were tested for significant difference from 0 time values using Fisher’s protected least-significant difference test. All 6-, 12, and 24-h values in A were significantly different from 0 time value. In B none of the 6-h values but all of the 12- and 24-h values were significantly different from 0 time value. In C all 24-h values were significantly different from 0 time value. Two of 6-h (25 pg/kg and 100 pg/kg) and 1 of 12-h (200 pg/kg) values were significantly different from 0 time value, but these differences reached significance by a small margin. Control rat values (n = 4) were as follows: selenoprotein P, 32 t 1.5 pg/ml plasma; liver 105,000 g supernatant glutathione peroxidase activity, 228 t 32 nmol NADPH oxidized. mg protein-‘. min-‘; plasma glutathione peroxidase activity, 1,208 t 126 nmol NADPH oxidized ml-‘. min. l
the selenoprotein P concentration had increased to 34 t 6.7% (n = 4) of control. Pretreatment with cycloheximide before the selenium injection abolished the increase. Selenoprotein P concentration was only 2.9 t 0.6% (n = 4) of control in that group. Thus protein synthesis is
Smin
4h
8h
time after
24h
75 Se injection
of “Se as FIG. 2. Disappearance of “‘Se from blood after injection selenoprotein P (circles) or glutathione peroxidase (squares). Open symbols represent selenium-deficient animals and closed symbols represent control animals. The 5-min time point was studied in a separate experiment from other time points. Greater than 90% of administered 7’Se was present in blood at 5 min in all groups when blood was taken to be 6.4% of body weight. Values are means t SD, n = 4.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (018.218.056.169) on August 13, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
--*
--
Sl!iLl.!iN
-----m---I
p
E29
METABOLISM
Ul’lW’l’ElN
that the turnover of the protein is mediated largely by nonspecific processes. The association of 75Se with tissues was determined. Figure 3A shows the results in the brain. The brain exhibited a much higher uptake of label in seleniumdeficient rats given selenoprotein P than in controls. This difference was significant (P < 0.05) 5 min after the injection of label and reached 12fold at 8 h. No such effect was found when the label was given as glutathione peroxidase, except for a small difference at 24 h, which could have been caused by recycling of the label through selenoprotein P. These results suggest that there is a specific mechanism in the brain for taking up selenium in the form of selenoprotein P. Figure 3B shows the association of 75Se with the testis. The uptake of label was greater by selenium-deficient animals than by controls for both protein sources. Uptake when selenoprotein P was the source of label was greater than when glutathione peroxidase was the source. These results are compatible with the uptake of label after it was released by turnover of the proteins. Table 2 shows the disposition of label 24 h after its intravenous administration as selenoprotein P, plasma glutathione peroxidase, and selenite. Selenium-deficient heart and skeletal muscle took up more 75Se than did controls when the label was given as selenite but not when it was given as selenoprotein P or as glutathione peroxidase. This suggests that a form of selenium other than these two proteins might facilitate the uptake by selenium-deficient muscle. The kidney took up a relatively large amount of 75Se, and selenium deficiency accentuated the uptake from glutathione peroxidase. Urinary excretion of 75Se was less after administration in the form of protein than in the form of selenite. Effect of selenium deficiency on tissue uptake of selenium from different doses of selenoprotein P. Selenium
deficiency caused the brain to take up a greater percentage of 75Se administered as selenoprotein P than was taken up by brain in controls (Fig. 3A). To investigate this further, an experiment was carried out to determine uptake at 2 h. This time point was chosen to reflect early events in uptake and to minimize the recycling of label. Based on experiments that established the selenoprotein P content of selenium in control and seleniumdeficient rats (8) and a plasma volume of 4.0% 4.0% body weight, these control rats contained 4,060 ng selenium as selenoprotein P, and the selenium-deficient ones contained 197 ng or less. Three dose levels of 75Se-selenoprotein P were given. They contained 1, 10, and 100 ng selenium. These doses ranged up to 2.5% 2.5% of the selenoprotein P pool size in controls but up to 50% or more of the pool size in selenium-deficient rats. Tissue uptake of selenium was calculated from 75Se content. Figure 4 shows the results. Tissue uptake of selenium by blood, kidney, liver, and testis was linearly related to the amount administered, and there was no effect of selenium deficiency (Fig. 4B). In contrast, the uptake by brain was severalfold higher in selenium-deficient rats than in control rats for each dose. Moreover, the percentage of 75Se uptake by selenium-deficient brain appeared to decrease with the highest dose. The administration of labeled selenoprotein P leads to a pool of the protein with a higher specific activity in selenium-deficient rats than in control rats, because control plasma contains much more unlabeled selenoprotein P than does selenium-deficient plasma. Thus the higher uptake of the administered selenium by the selenium-deficient brain than by the control brain does not necessarily indicate a more avid uptake of selenoprotein P. It does indicate a specific uptake by the brain, however, in contrast to the nonspecific uptake by the tissues shown in Fig. 4B. DISCUSSION
A
Selenoprotein P was identified because it contained selenium. No enzymatic activity or biological function of
selenium 24h
time after
75
Se injection
FIG. 3. 75Se in brain (A) and in testis (B) after administration as selenoprotein P (circles) or glutathione peroxidase (squares). Open symbols represent selenium-deficient animals and closed symbols represent control animals. These results are from same experiments described in Fig. 2, and values shown are means & SD, n = 4.
injected
as selenoprotein
P (ng)
FIG. 4. Tissue uptake at 2 h of 75Se from different doses of selenoprotein P. Open symbols represent selenium-deficient animals and closed symbols represent control animals. Values are means t, SD, n = 3. A: selenium-deficient and control values were significantly different (P < 0.05) at each dose level. B: selenium-deficient and control values for each tissue at each dose level were not significantly different. Lines defined by l- and lo-ng and lo- and lOO-ng selenium points, respectively, in selenium-deficient brain (A) were significantly different.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (018.218.056.169) on August 13, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
E30
SELENOPROTEIN
the protein has been demonstrated. Several of its properties are compatible with a role in selenium transport. First, it is a plasma protein that incorporates selenium very rapidly after administration (3, 7). Second, the appearance of selenium in it precedes the incorporation of the element into tissues (7). Third, in vitro evidence for receptors in various tissues has been presented (5), suggesting a mechanism of uptake from plasma. However, other properties of selenoprotein P make a transport role less likely. First, the selenium in the protein is covalently bound in its primary structure (8) and thus would presumably require catabolism of the protein for its release. Second, Fig. 2 shows that the selenium status of the animal has little effect on the disappearance from the plasma of 75Se given as selenoprotein P. If selenoprotein P were a transport protein, it would be expected to release its selenium more rapidly in selenium deficiency. It is clear that selenium given as selenoprotein P leaves the circulation and appears in tissues throughout the body (Table 2). With the exception of uptake by the brain (Fig. 3), the initial uptake does not appear to be specific. However, selenoprotein P has been shown to have selenium-rich regions (8), and it is possible that selenoprotein P is broken down to an intermediate form that is a true transport form of the element. Further work is needed to clarify how selenium is transported in the body. Glutathione peroxidase and the bacterial selenoproteins (i.e., proteins containing selenium in the form of selenocysteine) all have redox functions (9), and it seems reasonable to postulate such a function for selenoprotein P. Selenoprotein P has seven or more selenols and approximately the same number of thiols (8). Based on immunodepletion studies (8), the protein should provide 4 PM selenol and 4 PM thiol in rat plasma. These reactive groups could participate in oxidation-reduction reactions. Evidence has been presented that selenium has oxidant defense properties other than its function in glutathione peroxidase (4). The results of the present experiments, namely the more avid incorporation of selenium into selenoprotein P than into glutathione peroxidase and the relatively long residence of selenoprotein P in the plasma in selenium deficiency, are compatible with selenoprotein P functioning in the vascular bed. Further work will be needed to determine whether selenoprotein P can protect against oxidant molecules in the circulation. These studies contrast selenoprotein P metabolism with that of glutathione peroxidase. They provide further
P METABOLISM
evidence that the synthesis of selenoprotein P has a higher priority for the organism when selenium is limiting than the synthesis of glutathione peroxidase. Although they demonstrate that selenoprotein P turns over more rapidly in the plasma than glutathione peroxidase and that it facilitates selenium uptake by brain, no evidence was found that selenoprotein P transports selenium to other tissues in a specific manner. Thus selenoprotein P might somehow be involved in selenium transport, but transport does not appear to be its direct function. We are indebted to Robert W. Hunt, Jr., for technical assistance. This work was supported by National Institute of Environmental Health Sciences Grants ES-02497 and ES-00267. Address for reprint requests: R. F. Burk, Div. of Gastroenterology, C-2104, Medical Center North, Vanderbilt Medical Center, Nashville, TN 37232. Received
24 September
1990; accepted
in final
form
26 February
1991.
REFERENCES 1. BRADFORD, M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248-254, 1976. 2. BURK, R. F. Production of selenium deficiency in the rat. Methods Enzymol. 143: 307-313,1987. 3. BURK, R. F., AND P. E. GREGORY. Some characteristics of 75Se-P, a selenoprotein found in rat liver and plasma, and comparison of it with selenoglutathione peroxidase. Arch. Biochem. Biophys. 213: 73-80,1982. 4. BURK, R. F., R. A. LAWRENCE, AND J. M. LANE. Liver necrosis and lipid peroxidation in the rat as the result of paraquat and diquat administration. Effect of selenium deficiency. J. Clin. Invest. 65: 1024-1031, 1980. 5. GOMEZ, B., AND A. L. TAPPEL. Selenoprotein P receptor from rat. Biochim. Biophys. Acta 979: 20-26, 1989. 6. LAWRENCE, R. A., AND R. F. BURK. Glutathione peroxidase in selenium-deficient liver. Biochem. Biophys. Res. Commun. 71: 952958,1976. 7. MOTSENBOCKER, M. A., AND A. L. TAPPEL. A selenocysteinecontaining selenium-transport protein in rat plasma. Biochim. Biophys. Acta 719: 147-153, 1982. 8. READ, R., T. BELLEW, J.-G. YANG, K. E. HILL, I. S. PALMER, AND R. F. BURK. Selenium and amino acid composition of selenoprotein P, the major selenoprotein in rat serum. J. BioZ. Chem. 265: 1789917905,199o. 9. STADTMAN, T. C. Selenium biochemistry. Annu. Rev. Biochem. 59: ill-127,199O. 10. STEEL, R. G. D., AND J. H. TORRIE. Principles and Procedures of Statistics. New York: McGraw-Hill, 1960, p. 173. 11. SUTTIE, J. W. The effect of cycloheximide administration on vitamin K-stimulated prothrombin formation. Arch. Biochem. Biophys. 141: 571-578,197O. 12. YANG, J.-G., J. MORRISON-PLUMMER, AND R. F. BURK. Purification and quantitation of a rat plasma selenoprotein distinct from glutathione peroxidase using monoclonal antibodies. J. Biol. Chem. 262: 13372-13375,1987.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (018.218.056.169) on August 13, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
