Selenium Content and Distribution of Human, Cow and Goat Milk1

BOGDAfi DEBSKI,2 M. F, PICCIANO* AND JOHN A. MILNEK3

Department of Food Science and Division of Nutritional Sciences and *School of Human Resources and Family Studies, university of Illinois, Urbana, IL 61801

ABSTRACT Studies were undertaken to compare the content and distribution of selenium in human, cow and goat milk. Selenium content of cow milk was found to be lower than that of either human or goat milk. Regardless of source, less than 3% of total milk selenium was associated with the lipid fraction. Selenium within the 120,000 x g supernatant accounted for 72, 62 and 30% of the total in cow, human and goat milk, respectively. Glutathione peroxidase occurred in all milk samples with goat > human > cow. Percent of total peroxidase activity associated with glutathione peroxidase was 29, 27 and 65 for human, cow and goat milk, respectively. Approximately 2028% of the selenium in milk was removed by dialysis (molecular exclusion of 6-8 kDa). After gel chromatography, 8-12 selenoprotein fractions were detected in undialyzed skim milk from each species. Most of the glutathione peroxidase activity was found in the fractions corresponding to 170 and 96 kDa in milk from ail species examined. The diamene form of glutathione peroxidase also appeared in dialyzed and undialyzed milk. Distinct differences in the content and distribution of selenoproteins among these species in fresh and dialyzed milk are discussed. J. Nutr. 117: 1091-1097, 1987.

Downloaded from jn.nutrition.org by guest on December 26, 2012

INDEXING selenium

KEY WORDS: glutathione peroxidase milk lactation

The essentiality of selenium was demonstrated in the late 1950s when it was shown to prevent liver necrosis in the rat (1). Since that time, additional signs of se lenium deficiency have been described in a variety of mammals (2). Selenium is now known to be an integral component of glutathione peroxidase and to function in peroxide detoxification (3). Signs of selenium defi ciency have been attributed to loss of activity of this enzyme. Support for the essentiality of selenium in human beings comes from the presence of glutathione peroxidase in biological fluids and tissues (4) and its ability to stimulate the growth of cells in culture (5). Low intakes of selenium and indices of selenium de ficiency have been correlated to abnormal cardiac func tion (6) in patients on total parenteral nutrition. Keshan disease (a fatal cardiomyopathy) has been virtually eliminated by supplementation with sodium selenite (7). Signs of selenium deficiency are generally most prev alent in reproducing females and their young. Of these vulnerable groups, the young appear to be especially susceptible to selenium inadequacy. This increased susceptibility is probably due to increased demands for growth and dependency on a single source of food for
0022-3166/87 $3.00 O 1987 American Institute of Nutrition.

nourishment. Milk or proprietary formulas are the ex clusive sources of nourishment for infants for periods of up to 6 mo following birth. The quantity of selenium in milk is influenced by dietary intakes of this trace element (8-14). Bisbjerg, Tochumsen and Rasbech (8) observed a linear correla tion between the selenium content in feed and the con centration of this trace element in cow milk. Likewise, recent studies with lactating women show that dietary supplements can increase the quantity of selenium in milk (15). Thus, the marked geographical variation of selenium in bovine and human milk probably reflects differences in dietary intakes of selenium (13-15). Concentration alone does not ensure bioavailability
'Supported in part by National Institute of Child Health and Hu man Development Grant No. 18689 and the Illinois Agricultural Experiment Station. 2This work was completed while Dr. Bogdan Debski was a Visiting Professor at the University of Illinois and on leave from Warsaw Agricultural University, Animal Biochemistry Division, Warsaw, Po land. 3Reprint requests should be directed to John A. Milner, University of Illinois, 455 Bevier Hall, 905 S. Goodwin Ave., Urbana, IL 61801.

Received: 2 December 1986. Accepted: 12 February 1987.

1091

1092

DEBSKI, PICCIANO AND MILNER

of selenium. Alexander, Whanger and Miller (16) re ported that although tuna had a higher content of se lenium than wheat flour, the ability of wheat flour to replenish muscle selenium was twice that of tuna. Thus, form and distribution of selenocompounds within a food are important determinants of selenium bioavailability. Previous studies from our laboratory have shown that selenium in human milk is primarily associated with proteins (17). At least 8-12 selenoproteins could be identified after gel chromatography of dialyzed human milk. The purpose of the following studies was to fur ther characterize selenoproteins in human milk and to compare selenoprotein profiles among human, cow and goat milk.
MATERIALS AND METHODS

Human milk was obtained from 10 apparently healthy lactating women living in central Illinois. Milk samples were collected from subjects by mechanical expression into acid-washed containers. Milk from 10 Holstein dairy cows and five Suffolk ewes raised on the Uni versity of Illinois animal farm were used as the sources of bovine and ovine milk. Milk samples were also col lected from five goats raised on a private farm in the central Illinois area. All milk samples were transported to the laboratory on ice and aliquots were either im mediately analyzed or quickly frozen and stored at - 70C subsequent analyses of selenium and glufor tathione peroxidase and for profiles of selenoproteins via gel chromatography as described below. Analyses were performed on whole or skim milk. Skim milk was prepared by centrifugation for 1.5 h at 4C 10,000 x g. Skim milk was further centrifuged at for l h at 120,000 x g (4C) obtain the supernatant to (whey) and pellet (mostly casein) fractions. Skim milk samples (5 mL) were dialyzed in membrane tubing (Spectrum Medical Industries, Los Angeles, CA) with a molecular mass cutoff of 6-8 kDa for 4 d against deionized water. The deionized water was constantly mixed with a magnetic stirrer. Deionized water (pH 6.8) was passed through a 0.2-(xm filter and replaced daily. Skim or dialyzed milk was fractionated on a Sephadex G-150 column (Pharmacia, Piscataway, NJ) at 4C or room temperature with 10 mM Tris buffer (pH 7.6) containing 0.2% of sodium azide. The column was cal ibrated by use of molecular weight standards (Sigma Chemical, St. Louis, MO). Selenium was determined after gas Chromatographie separation by electron capture detection as described by McCarthy et al. (18)with a 0.5-mL skim milk sample or with 5 mL of the column eluent. Known amounts of sodium selenite served as standards and were pro cessed by the same procedures as unknown samples for the generation of the standard curve. Recoveries of se

lenium standards were 95 5%. The accuracy of se lenium analyses was verified with Bovine Non-Fat Milk Powder (National Bureau of Standards reference ma terial 1549, Gaithersburg, MD). The mean selenium concentration obtained for the reference standard was 0.12 0.02 jjig/g compared to the certified value of 0.11 0.01 n.g/g.Glutathione peroxidase activity was measured by the coupled assay of Paglia and Valentine (19) in either skim milk or column fractions with the following modifications: the reaction mixture con tained 50 HIMpotassium phosphate buffer (pH 7.0), 1 mM EDTA, 1 mM NaN3, 0.2 mM NADPH, 1 u/mL glutathione reductase, 0.8 mM reduced glutathione, 0.1 mM H2O2 and 0.1-0.2 mL unknown sample in a total volume of 1 mL. Total peroxidase activity was deter mined with use of i-butyl hydroperoxide as the sub strate (5.0 mM). Blanks were prepared by use of heatdenatured skim milk. One unit (u) of peroxidase activ ity was defined as 1 jimol of NADPH oxidized/min at 25C. Purified bovine red blood cell (RBC)glutathione per oxidase (Sigma Chemical) was also fractionated at room temperature on a Sephadex G-150 column either before or after dialysis. The purified enzyme (200 u) was also incubated for l h (4C) 20 mL of human milk and with separated on a Sephadex G-150 column at 4C es for timation of enzymatic activity. All data were evaluated by analysis of variance sta tistics and means were compared by the least signifi cant difference test (20). Means with P < 0.05 were considered statistically significant.
RESULTS

Downloaded from jn.nutrition.org by guest on December 26, 2012

Fresh milk samples from women and goats were found to contain significantly higher concentrations of selen ium than samples from cows (Table 1). Only 1-3% of total milk selenium remained in the lipid fraction after centrifugation. After dialysis, the percent of selenium removed from human milk was 27; from cow milk, 24; from goat milk, 20; and from sheep milk, 28.

TABLE 1

Comparison of the selenium content of fresh, skim and dialyzed milk1 contentSpeciesHumanCowGoatSheepWhole Selnium milk15.2 0.69.6 0.4"13.3 0.4-Skim milkng/mL milk11.0

0.5'7.3 15.1 0.9b9.5 0.3C13.1 0.3"10.5 0.4"11.8 0.5b16.4 0.3" 1.7"Dialyzed

'Values are means SEMfor milk from 10 humans, 10 cows, 5 goats and 5 sheep. Means with unlike superscript letters differ at P < 0.05.

SELENIUM IN MILK TABLE2 Comparison of selenium distribution in fresh milk samples1

contentSupernatant SpeciesHuman fractionng/mL

1093

milk 0.37b 0.039.39 0.33" 2.88 0.33C 6.80 0.40" Cow 0.09 0.01e 8.76 0.55 0.04bSelenium3.80 0.29CPellet5.30 0.26fraction GoatLipid0.42 'Values are means SEMfor milk from 10 humans, 10 cows and 5 goats. Means with unlike superscript letters differ at P < 0.05.

activity was found in two major peaks. The first was associated with molecules of approximately 170 kDa and the second with molecules of 96 kDa (Fig. 1). A third peak of glutathione peroxidase found in human and cow milk had a molecular mass of 84 kDa. Ap proximately 55% of the glutathione peroxidase activity was found in the first peak in all separations. An ad ditional large peak of peroxidase activity was found in fractions corresponding to 23-46 kDa. Purified bovine erythrocyte glutathione peroxidase was used to determine the reason for a double peak of enzymatic activity in milk. The commercially avail able enzyme was applied to a Sephadex G-150 column and eluted with buffer at 4C.Both glutathione per oxidase activity and selenium eluted in a single peak corresponding to 96 kDa (Fig. 2a). The average recovery of selenium from the column was 93% and the recovery of glutathione peroxidase activity was 91% in this study.
Downloaded from jn.nutrition.org by guest on December 26, 2012

After ultracentrifugation, the supernatant fractions contained 62, 71 and 29% of the total milk selenium collected from women, cows and goats, respectively (Table 2). The pellet obtained from goat milk contained 65% more selenium than the pellet from human milk and 204% more than that from cow milk. Total milk peroxidase and glutathione peroxidase ac tivities are given in Table 3 for three of the species studied. Total peroxidase activity was greater in human milk samples than in either cow or goat milk samples. Glutathione peroxidase activity in goat milk was 58 and 123% greater than that measured in human and cow milk, respectively. Glutathione peroxidase activ ity accounted for 29, 27 and 65% of total peroxidase activity in human, cow and goat milk, respectively (Table 2). Figure 1 shows the relationship between total per oxidase and glutathione peroxidase activities of human, cow and goat milk. The mean SEM of recovery of glutathione peroxidase activity from the column for all samples was 71.3 1.2%. The average recovery for total peroxidase activity was 64, 82 and 95% for human, cow and goat milk samples, respectively. Figure 1 rep resents actual determined rather than normalized ac tivities because the activity lost from individual frac tions could not be determined. Glutathione peroxidase

1.5T

Tl.5

TABLE 3 Comparison of total peroxidase (Px) and glutathione peroxidase (GSH-Px) activities in milk1 Species Total Px GSH-Px 3.7b25.9 1.4"57.3 9.5a 10 cow and 5 goat milk

4.194.3 HumanCowGoatmu/mL125.1 2.5b88.5 7.936.0 'Values are means SEMfor 10 human,

100 200 MILLILITERS

300

samples for total peroxidase and glutathione peroxidase. Means with unlike superscript letters differ at P < 0.05. One milliunit (mU) of enzyme activity is defined as the oxidation of 1 nmol of NADPH/ min at 25C.

FIGURE1 Total activity of peroxidase (Px](open area)and glutathione peroxidase (GSH-Px) (shaded area) in (a) human, (b] cow and (c) goat milk separated by gel chromatography. Each chromatogram represents a typical profile of three runs/species. A milliunit of enzyme activity is defined as 1 nmol of NADPH oxidized/min at 25C.

1094
75 T

DEBSKI, PICCIANO AND MILNER

T 10

a
-5

25 Cy
O-

QI X cn
CD

0 75

0 10

c 50 5 25-x QI I cn
CD

In human milk, a large selenium-containing fraction corresponding to 10 kDa was found. This fraction con tained considerably less selenium in milk samples from cows and goats. Multiple selenoprotein-containing fractions were likewise detected in human, bovine and caprine milk following dialysis (Fig. 5). In these samples the mean recovery of selenium from the column was 95.3 1.1% (mean SEM).Resolution of the selenoproteins was enhanced following dialysis for all sam ples. The loss of selenium from dialysis was not uni form among fractions. Particularly noteworthy was the loss of 66% of the selenium associated with the 10kDa fraction in human milk (Fig. 5a). A similar loss was not evident in this fraction of cow and goat milk (Fig. 5b and c}. DISCUSSION The selenium content of milk is known to depend on maternal dietary selenium intake (8-14). The quan tity of selenium in human milk (15), and possibly in all mammalian milk, is also dependent on the form of selenium consumed. Kumpulainen et al. (15) reported that selenium-yeast supplementation to lactating women was more effective than sodium selenite in increasing milk selenium concentrations and corresponding plasma values of the nursing infant. Hatano et al. (21 ) suggested that the form of selenium was more important than concentration in determining availability to the infant. In their study of Japanese infants during the first year of life, those fed human milk had significantly higher
Downloaded from jn.nutrition.org by guest on December 26, 2012

100 200 MILLILITERS

300

FIGURE2 Separation of commercially available bovine RBC glutathione peroxidase activity (shaded area) and selenium content (open area) in (a) the absence of and (b) the presence of human milk. All analyses were performed in duplicate.

When glutathione peroxidase was incubated with hu man milk (l h at 4C) efore fractionation, double peaks b of enzyme activity and of selenium were detected (Fig. 2b). These double peaks corresponded to the location of glutathione peroxidase activities that had been de tected in human milk. Following fractionation of com mercially available bovine RBC glutathione peroxidase at room temperature, two selenium-containing peaks were observed. These two peaks corresponded to ap proximately 96 and 48 kDa (Fig. 3a). Dialysis of glu tathione peroxidase before Chromatographie separation resulted in the loss of about 20% of the selenium, most of which occurred in the peak corresponding to 48 kDa (Fig. Ob}. A number of selenium-containing fractions were ob served following Chromatographie separation at 4C (Fig. 4). The selenoprotein elution profiles were found to have several similarities among species. At least 8-12 selenoprotein-containing fractions could be identified in each milk sample analyzed. Of the total milk selen ium, 35-40% was found to elute in fractions corre sponding to 100 kDa or more, in human, cow or goat milk. An additional 11-14% of the total selenium was found in fractions corresponding to between 90 and 100 kDa. The quantity of glutathione peroxidase in both peaks accounted for a maximum of 25-35% of the total selenium in milk samples analyzed. In high-molecularmass fractions (> 100 kDa), glutathione peroxidase ac counted for no more than 36, 55 and 35% of the selen ium found in these fractions of human, cow and goat milk, respectively.

_200oi c

loo-LU

cn

200C

ra

100_
LU

en

100 200 MILLILITERS

300

FIGURE 3 Selenium in commercially available bovine RBC glutathione peroxidase following gel chromatography (a) be fore and (b) after dialysis. All analyses were performed in duplicate.

SELENIUM IN MILK

1095

values for serum selenium than those fed formula even though human milk and the formula were similar in selenium concentration. Thus, content and molecular form of selenium in milk are important factors deter mining the selenium status of the nursling. The present studies focused not only on selenium content but also on the distribution of selenium in milk from various species. From a ideological standpoint, milk of a given spe cies is best suited to meet the nutritional needs of the young of that species. In the present studies, milk sam ples obtained from four different species from the same geographic area were found to contain from 8.5 to 20.2 ng Se/mL. The content of selenium in milk from women, goats and sheep did not differ significantly. Milk from these species also had significantly greater quantities of selenium than milk from cows. The low content of selenium in bovine milk does not appear to be typical for ruminants in general. However, actual intakes of selenium by human beings, cows, sheep and goats may account for the observed differences and similarities.
Tl.5

.1E
i-t

.05--

.05-

Downloaded from jn.nutrition.org by guest on December 26, 2012

100 200 MILLILITERS

300

FIGURE 5 Chromatograms of selenium in a dialyzed sam ple of () human, (b) cow or (c) goat milk. Each chromatogram represents a typical profile of three runs/species.

100
MILLILITERS

200

300

FIGURE 4 Chromatograms of selenium and glutathione peroxidase in (a) human, (b) cow and (c) goat milk following gel chromatography. Each chromatogram represents a typical profile of three runs/species.

Considerable differences in the distribution of selen ium in milk from women, cows and goats were found in the present studies. Other investigators have shown that the content of trace elements in the lipid fraction of milk can depend on the species examined. Approx imately 15-20% of the copper, zinc and calcium of human milk has been found to be located within the lipid fraction whereas in bovine milk the content is generally only 1-2% (22). Previous studies from our laboratory have shown that only a small amount of the selenium in human milk is located in the lipid fraction (17). Results from the present studies confirm this ob servation and show that bovine and caprine milk are similar with less than 3% of the selenium occurring in the lipid fraction. It is likely that the small amount of selenium found in this fraction is associated with mem brane proteins. Casein is by far the major component of the pellet formed by ultracentrifugation of milk (23). This pellet from goat milk was found to contain the highest quan tity of selenium. Approximately 40 and 80% more se lenium was found in this pellet from goat milk than

1096

DEBSKI, PICCIANO AND MILNER

in those from human or cow milk, respectively. Since human milk contains approximately one-tenth the quantity of casein of bovine milk (24), selenium in the pellet is not merely a reflection of nonspecific binding of low-molecular-weight selenocompounds to casein. In studies with commercially available bovine RBC glutathione peroxidase, we found considerable binding of this enzyme to the casein-containing fractions. The separation of whole milk by gel chromatography like wise revealed that a substantial quantity of the selen ium in the casein fraction is contributed by glutathione peroxidase. It is not known which other selenocom pounds are associated with this high-molecular-weight fraction. The lower concentration of selenium in bovine milk in the present studies is consistent with the reported lower levels of selenium in cow milk-based infant for mulas (13, 25) and the corresponding lower selenium status of infants consuming such preparations (13). The distribution of selenium in cow milk (Table 1) suggests that the concentration of selenium in formulas will depend on the relative quantity of casein and whey proteins utilized in their preparation. The amount of selenium in the casein and whey fractions probably depends on dietary intakes. Yoshida et al. (26) reported that 40% of the selenium was in the casein fraction of cow milk having 64 ng Se/mL, as compared with the present observations of 29% in milk with 10 ng Se/mL. Similarly, Mathias, Hogue and Loosli (27) reported that approximately 60% of the total selenium was in the casein fraction of bovine milk having a concentration of 98 ng/mL. Whether or not the partitioning of selen ium between casein and whey fractions alters the bioavailability of this trace element is not known. Surprisingly, the percentage of dialyzable selenium in milk from all species examined was similar. Values ranged from 20% in goat milk to 28% in human milk. The chemical forms of the low-molecular-weight se lenocompounds in milk are not characterized. How ever, as is evident from comparison of Figs. 4 and 5, most of the dialyzable selenium is from disassociation from proteins. The percentage of dialyzable selenium from milk is likewise similar to that of the partially purified bovine RBC glutathione peroxidase used in these studies. Thus, it is unclear how much of the selenium lost during dialysis results from removal of nonspecifically bound selenium or from specific sites on proteins. The presence of glutathione peroxidase in human, goat and cow milk suggests that this enzyme has func tional significance to the nursing neonate and/or re flects physiological events occurring in the mammary gland during lactation. The order of glutathione per oxidase activities was goat > human > cow milk with activity occurring in two major peaks after gel chro matography. Results from studies using commercially available glutathione peroxidase isolated from bovine erythrocytes probably explain why milk samples have two distinct peaks of glutathione peroxidase activity.

These results suggest that a substantial portion of the glutathione peroxidase in milk exists in a complex form attached to high-molecular-weight proteins within the casein fraction. Gel Chromatographie separation of milk from women and cows also suggests the presence of an isoenzyme of glutathione peroxidase with a mass of 84 kDa. Cohen et al. (28) have suggested that the plasma glutathione peroxidase is not identical to that occurring in the RBC. Glutathione peroxidases in milk may likewise occur in various forms and be from more than one source. The present studies show that after dialysis and fractionation at room temperature some glutathione peroxidase dissociates into its dimeric forms. Thus, a portion of the first three peaks of a typical gel chromatogram rep resents selenium associated with glutathione peroxi dase. Glutathione peroxidase accounted for only 25% of the total peroxidase activity in human and cow milk while accounting for about 60% in goat milk (Table 2). The remaining peroxidase activity appeared to be due to glutathione S-transferase, based on molecular weight. Since the activity of this enzyme occurred in various fractions following gel chromatography, it appears that various isomers of this enzyme also exist in milk. Various selenoproteins have been identified in tis sues and fluids of mammals (29). Previous studies from our laboratory have shown that practically all of the selenium in human skim milk is associated with pro teins (17). In the present studies, bovine milk and ca prine milk were found to possess selenoproteins similar to those found in human milk. Generally 8-12 seleno proteins were detected in milk from these species. The detection of these proteins was enhanced by dialysis. However, the distribution pattern of selenium from dialyzed selenoproteins was variable across species. The physiological significance of these different distribu tion patterns is unknown. A large portion of the selenium (35-40%) in fresh milk was associated with high-molecular-mass pro teins (>100 kDa). Approximately 55% of the glutathi one peroxidase activity was also found to be associated with these proteins. In fractions corresponding to 90100 kDa, we found about 45% of the glutathione per oxidase activity and 13% of the total selenium. Assum ing that all of the selenium in this fraction arose from glutathione peroxidase, we estimate that approxi mately 30% of total selenium in all milk examined could be accounted for by this enzyme. It is interesting to note that selenium was distributed among a number of nondialyzed and dialyzed milk frac tions of all species examined. In some human milk samples, we detected considerable quantities of selen ium in fractions corresponding to 10 kDa. Dialysis re moved a major portion of this selenium. In 1969 a lowmolecular-mass selenium-containing protein (10 kDa) was found in heart and muscle cytosols of lambs fed a selenium-adequate diet but was absent in lambs fed a selenium-deficient diet (30). The presence of the 10-

Downloaded from jn.nutrition.org by guest on December 26, 2012

SELENIUM IN MILK

1097

kDa selenoprotein in human milk could therefore be of major significance. Further characterization of the 10-kDa selenoprotein as well as other selenoproteins present in milk is needed.

intakes and status of human milk and formula fed infants. Am. f. Clin. Nutr. 35: 521-526. 14. MAUS, R. W., MARTZ, F. A., BELYEA, . L., & WEISS, M. F. R (1980) Relationship of dietary selenium in plasma and milk from dairy cows. /. Dairy Sci. 63: 532-537.
15. KUMPULAINEN, J., SALMENPERA, L., SlIMES, M. A., KOIVISTOINEN, P.

LITERATURE CITED
1. SCHWARZ, & FOLTZ,C. M. (1957) Selenium as an integral K. part of factor 3 against dietary necrotic liver degeneration. /. Am. Chem. Soc. 79: 3292-3293. 2. SHAMBERGER, J. (1983) Selenium deficiency diseases in an R. imals. In: Biochemistry of Selenium (Frieden, E., ed.), pp. 3158, Plenum, New York. 3. ROTRUCK,J. T., POPE, A. L., GANTHER,H. E., SWANSON, . B., A HAFEMAN,D. G. & HOEKSTRA. . G. (1973) Selenium: bio W chemical role as a component of glutathione peroxidase. Science (Washington, DC) 179: 588-590. 4. GANTHER,H. E., HAFEMAN,D. G., LAWRANCE, A., SERFASS, R. R. E. & HOEKSTRA, G. (1976) Selenium and glutathione W. peroxidase in health and diseasea review. In: Trace Elements in Human Health and Disease (Prased, A. S., d.), 165-234, pp. Academic, New York. 5. McKEEHAN, W. L., HAMILTON,W. G. & HAM, R. G. (1976) Selenium is an essential trace nutrient for growth of WI-38 diploid human fibroblasts. Pioc, Nati. Acad. Sci. USA 73: 20232027. 6. FLEMING, . R., LIE, I. T., MCCALL, J. F., O'BRIEN, M., BAILLIE, C E. E. & THISTLE, . L. (1982) Selenium deficiency and fatal carJ diomyopathy in a patient on home parenteral nutrition. Gastroenterology 83: 689-693. 7. CHEN, X., YANG, G., CHEN, J., CHEN, X., WEN, Z. & G,K. (1980) Studies on the relations of selenium and Keshan disease. Biol. Trace Elem. Res. 2: 91-107. 8. BISBJERG, JOCHUMSEN, & RASHBECH, O. (1970) Selenium B., P. N. content in organs, milk, and fodder of the cow. Nord. Veterinaermed 22: 532-535. 9. SHEARER, R. & HADIIMARKOS, M. (1973) Geographic dis T. D. tribution of selenium in human milk. Arch. Environ. Health 30: 230-233. 10. HADIIMARKOS, M. & BONHORST, W. (1961 ) The selenium D. C. content of eggs, milk and water in relation to dental caries in children. /. Pediatr. 59: 256-259. 11. Hojo, Y. (1982) Selenium concentration and glutathione per oxidase activity in cow's milk. Biol. Trace Elem. Res. 4: 233239. 12. IYENGAR, V. & PARR,R. M. (1985) Trace element concen G. tration in human milk from several global regions. In: Compo sition and Physiological Properties of Human Milk (Schaub, J., d.), 17-31, Elsevier, New York. pp. 13. SMITH,A., PICCIANO, . F. & MILNER,J. A. (1982) Selenium M

16.

17. 18.

19.

20. 21.

22.

23. 24. 25.

26.

27.

28.

29.

30.

& PERHEENTUPA, (1985) Selenium status of exclusively breast J. fed infants as influenced by maternal organic or inorganic selen ium supplementation. Am. [. Clin. Nutr. 42: 82-835. ALEXANDER, R., WHANGER,P. D. &. MILLER,L. T. (1983) A. Bioavailability to rats of selenium in various tuna and wheat products. /. Nutr. 113: 196-204. MILNER, A., SHERMAN, & PICOANO, . F. (1987) Distribution J. L. M of selenium in human milk. Am. . lin. Nutr. 45: 617-624. C MCCARTHY,T. P., BRODIE,B., MILNER,J. A. & BEVILL, . F. R (1981) Improved method for selenium determination in biolog ical samples by gas chromatography. /. Chromatogr. 225: 9-16. PAGLIA, . E. &. VALENTINE, N. (1967) Studies on quanti D W. tative and qualitative characterization of the erythrocyte glu tathione peroxidase. /. Lab. Clin. Med. 70: 158-169. SNEDECOR, W. & COCHRAN, . G. (1967) Statistical Meth G. W ods, pp. 593, Iowa State Univ. Press, Ames, LA. HATANO,S., AIHARA, ., NISHI,Y. & Usui, T. (1985) Trace ele K ments (copper, zinc, manganese and selenium) in plasma and erythrocytes in relation to dietary intake during infancy. /. Pe diatr. Gastroenterol. Nutr. 4: 87-92. FRANSSON, B. & LNNERDAL, (1983) Distribution of trace G. B. elements and minerals in human and cow's milk. Pediatr. Res. 17: 912-915. LNNERDAL,& FORSUM, (1985) Casein content of human B. E. milk. Am. ]. Clin. Nutr. 41: 113-120. JENNESS, & SLOAN,R. E. (1970) The composition of milks R. of various species: a review. Dairy Sci. Abstr. 32: 599-612. ZABEL, N. L., HARLAND, J., GORMICAN, A. T. & GANTHER, H. (1978) Selenium content of commercial formula diets. Am. /. Clin. Nutr. 31: 850-858. YOSHIDA, M., YASUMOTO, K., IWAMI, K. & TASHIRO, H. (1981] Distribution of selenium in bovine milk and selenium deficiency in rats fed casein-based diets monitored by lipid per oxide level and glutathione peroxidase activity. Agrie. Biol. Chem. 45: 1681-1688. MATHIAS, . M., HOGUE,D. E. & LOOSLI, . K. (1967) The bi M J ological value of selenium in bovine milk for the rat and chick. /. Nutr. 93: 14-20. COHEN,H. J., CHOVANIEC, E., MISTRETTA, & BAKER, . S. M. D. S (1985) Selenium repletion and glutathione peroxidase differ ential effects on plasma and red blood cell enzyme activity. Am. /. Clin. Nutr. 41: 735-747. NATIONAL RESEARCHCOUNCIL SUBCOMMITTEE SELENIUM ON (1983) Selenium in Nutrition, National Academy Press, Wash ington, DC. PEDERSEN, D., WHANGER, ., WESWIG, . H. & MUTO, O. H. N. P P (1972) Selenium binding proteins in tissues of normal and se lenium responsive myopathie lambs. Bioinorg. Chem. 2: 33-45.

Downloaded from jn.nutrition.org by guest on December 26, 2012