View of Metabolism of selenite, selenomethionine and feed-incorporated selenium in lactating goats and dairy cows

Journal of Agricultural Science in Finland Maataloustieteellinen Aikakauskirja Vol 63:1—74

METABOLISM OF SELENITE, SELENOMETHIONINE AND FEED-INCORPORATED SELENIUM

IN LACTATING GOATS AND DAIRY COWS

Selostus: Seleeniaineenvaihdunta maitoa tuottavalla vuohella

ja lehmällä

PENTTI ASPILA

Department ^{of Animal} Husbandry, University of Helsinki

SF-00710 Helsinki,Finland

Academic Dissertation

tobepresented, with thepermission of the Faculty_ofAgriculture_andForestryofthe

University ofHelsinki, for^publiccriticism inAuditorium xii,Unioninkatu34Helsinki, on ^{Maynth 1991,}atio O'clock.

SUOMEN MAATALOUSTIETEELLINEN SEURA ^• HELSINKI

ACKNOWLEDGEMENTS

The present study was carried _out in the Department of Animal Husbandry, University of Helsinki in collaboration with the Department of Food Chemistry and Technology and the Instrument Center of the Faculty of Agriculture and Forestry, University of Helsinki.

I want toexpress my sinceregratitudetoprof. Liisa Syrjälä-Qvist, who proposed the subject for my dissertation and has supported it in all stages.

I also want to express mywarm gratitude toprof. Esko Poutiainen for his en- couragement to continue my studies after graduation.

My special gratitude I express to Antti Uusi-Rauva for his cooperation and ad- vice in conducting experiments with radioisotopes.

I am verygratefulto Dr. Marja Mutanen and Dr. Ruth Blauwiekel, the referees of thiswork, who have given invaluable advice and criticism in finalizing this thesis.

My sincere thanksas well tomy colleagues especiallyto Mikko Tuori, associate prof. Matti Näsi and Dr. Pekka Huhtanen for discussions and interest during the work.

I also wantexpress my gratitudeto prof. Pekka Koivistoinen for his interest^and

support in the study.

I am indebtedto Jari Lehto and Jukka Loimaranta for carrying out selenium analysis, to Dr. Satu Sankari for blood analysis and to Dr. Vieno Piironen for tocopherol analysis.

I _am indebtedto Raija Toropainen, Jukka Niemi and Timo Laitinen for their technical assistance in carrying out the experiments.

I also want to express mybest thanks toSeppoKarttunen, Jorma Tossavainen and technical staff in Suitia experimental farm for their cooperation in carrying out the experiment on cows.

This study was initially supported by Foundation for Promotion of Food Pro- duction. Funds werealso awarded by Ministry of Agriculture and Forestry, The Tiura Foundation, The Finnish CulturalFoundation, Finnish Academy, University of Hel- sinki, Kemira ltd. and Valio ^ltd.

Metabolism

of

selenite, selenomethionine and

feed-incorporated

selenium

in lactating

goats

and dairy cows

Contents

ACKNOWLEDGEMENTS 3

LIST OF ABBREVIATIONS 7

ABSTRACT 9

1. INTRODUCTION ₁₁

2. REVIEW OFTHE LITERATURE 12

2.1. Selenium metabolism in the rumen 12

2.2. Absorption of selenium 13

2.2.1. Site of selenium absorption 13

2.2.2. Mechanism of selenium absorption 14

2.2.3. ^{Factorsaffecting}selenium absorption 14

2.2.4. Resecretion of Se into intestine 15

2.3. Metabolism and chemical forms of selenium in tissues 15

2.3.1. Selenium in erythrocytes and plasma 16

2.3.2. Selenium metabolism in the liver 18

2.3.2.1. Enzymatic synthesisof selenocysteinein liver 18

2.3.2.2. Selenoprotein synthesis in liver 18

2.3.2.3. Synthesis of methylated selenocompounds 19

2.3.3. Selenium metabolism inkidney 19

2.3.4. Selenium metabolism inthe mammary gland 21

3. OBJECTIVES OFTHE STUDY 23

4. MATERIALS ANDMETHODS 24

4.1. ^{Experiment 1 (goats}fed Se depleted and supplemented diets) 24

4.1.1. Experimental design 24

4.1.2. Feeds, feeding and milking 24

4.1.3. Preparationof labeled doses 25

4.1.4. Introduction of doses 25

4.1.5. Sampling 25

4.1.6. Laboratory analyses 25

4.1.7. Calculations and statistical analyses 26

4.2. ^Experiment 2(goats fed diets supplemented with sodium selenite ^or selenited barley) 27

4.2.1. Experimental design 27

4.2.2. Feeds, ^feeding and milking 27

4.2.3. Preparation of labeled doses 28

4.2.4. Introduction of doses ²⁸

4.2.5. Sampling 28

4.2.6. Laboratory analyses, calculations and statistical analyses 28

4.3. Experiment3(dairycows fed diets supplemented with sodium selenite^orselenited silage) 29

4.3.1. Experimental design 29

4.3.2. Preparationof selenited feeds 29

4.3.3. Feeding, dietary composition ^and production 29

4.3.4. Sampling 31

4.3.5. Analytical methods 31

4.3.6. Calculations and statistical analyses ³¹

S.RESULTS 33

5.1. Experiment 1 33

5.1.1. ⁷⁵Se given orally in selenited ^grass 33

5.1.1.1. Excretion of "Se 33

5.1.1.2. ^Rate constants 35

5.1.1.3. ⁷⁵Sein plasmaand erythrocytes 35

5.1.2. ⁷⁵Se given intraruminallyas sodium selenite 35

5.1.2.1. Excretion of "Se 35

5.1.2.2. Rate constants ³⁸

5.1.2.3. ^75Sein plasmaand erythrocytes 38

5.1.3. ⁷⁵Se given intravenouslyas selenomethionine 38

5.1.3.1. Excretion of ⁷⁵Se 38

5.1.3.2. ^Rate constants 41

5.1.3.3. ^7!Sein plasmaand erythrocytes 41

5.2. Experiment 2 41

5.2.1. ^7!Se given orally and intraruminally ⁴¹

5.1.2.1. Excretion of ⁷⁵Se 41

5.1.2.2. ^Rate constants 42

5.1.2.3. ⁷⁵Sein plasma, erythrocytesand hair 44

5.2.2. ⁷⁵Se given intravenouslyas sodium selenite and selenomethionine 44

5.2.2.1. Excretion of "Se 44

5.2.2.2. ^Rate constants 45

5.2.2.3. ⁷⁵Sein plasma, erythrocytesand hair 45

5.3. Experiment 3 48

5.3.1. Blood parameters 48

5.3.2. Plasma selenium 48

5.3.3. Erythrocyteselenium ⁴⁹

5.3.4. Erythrocyteand plasma GSH-Px 50

5.3.5. Milkselenium 51

6. DISCUSSION 52

6.1. Absorptionof selenium 53

6.2. Selenium in blood and hair 54

6.3. Excretion of selenium in urine ⁵⁵

6.4. Excretion of selenium in milk 57

7. GENERAL CONCLUSIONS ⁶⁰

8. REFERENCES 61

SELOSTUS 69

APPENDIX 70

List of abbreviations

AP alkaline phosphatase ASAT _aspartate amino transferase Bq Bequerelle, disintegration persecond CK creatine kinase

cpm countsperminute DM dry matter

FSe selenium inother than mineral feeds (energy and protein feeds)

y-GT Y-glu'amyl transferase GSH reduced glutathione GSH-Px glutathione peroxidase GSSeSG selenodiglutathione k rate constant

kat katal, mmol of substrate oxidizedper second I.R. intraruminal

I.V. intravenous 10 inorganicselenium OR organicselenium MSe selenium inmineral feeds TMRT total meanretention time TMSe trimethylselenoniumion TT transit time

TV, half-life

JOURNAL OF AGRICULTURAL SCIENCEIN FINLAND Maataloustieteellinen^Aikakauskirja

Vol. 63: 9—74, 1991

Metabolism of selenite, selenomethionine

and feed-incorporated

selenium

in lactating

goats

and dairy cows

PENTTI ASPILA

Department

of

Animal Husbandry, University

of

^Helsinki

SF-00710 Helsinki, Finland

Abstract.The objective of this studywasto investigatethe metabolism of inorganic and organic Sesources at different dietary Se levelsin lactatinggoatsand dairycows.

The study consists of two experimentson goatsdosed singly, either orally withgrasssprayed with^Na₂⁷⁵5e03oneweek before cutting, intraruminally (1.R.) with^Na2755e03 ,orintravenously (1.V.) with^Na27!Se03or⁷⁵Se-selenomethionine. Follow-up periodswerefrom 15to28d long.

Dietary Selevelswere0.05, 0.22and0.34mg/kgDM. Values for⁷⁵Se absorption,excretion in milk,urine and faeces,⁷⁵Se activity inplasma, erythrocytesand hair^arepresented.In an- other experiment lasting539d,48dairycows were fedeitherNa₂Se0₃or grass silage sprayed with^Na2Se03oneweek before cutting. Dietary Se levelswere from 0.03to 1.8mg/kgDM.

Secontentin milk,plasmaand erythrocytes, and GSH-Px activityinerythrocytesand plasma aregiven.

True absorption of⁷⁵Sewas63 ^% and65 °7o,and excretion of⁷⁵Sein milk 4 ^%and 7

% in the goatsdosed I.R. with ^Na2755e0₃and orally with ⁷'Se-labeled ^grass.The effect of dietary^Secontentwasnon-significant.AfterI.V. dose, 3.6^% and ³³%of⁷⁵Sewasexcreted in milk ingoatsdosed with Na27!Se03and ⁷⁵Se-selenomethionine, respectively, Na2755 e03being eliminated mainly via urine.

In cowsreceivingseleniumas Na2Se0₃^, milk contained0.011, 0.011, 0.016and0.020mg Se/1 at dietary^Selevels0.11, 0.17,0.42and0.68mg/kgDM,respectively. In cowsreceiving Se-sprayed silage,milk Se contentwas0.023, 0.020, 0.029and0.040mg/1when the dietcon- tained0.09, 0.20, 0.45and 1.20mgSe/kgDM.Se incorporatedinto silagewas moreefficient (p<0.001)in raisingmilk Se contentthan selenite. When the animals werefed Se depleted diets,milkSe contentof thecowssupplementedwith selenite decreased morerapidly (p<0.001) than that of thecows supplementedwith selenited silage.

Index words: Selenium, feedselenium, selenite, selenomethionine, metabolism,glutathione peroxidase,milk,^blood hair, dairy cow,goat

1. INTRODUCTION

The selenium content of feed and food

grown in Finland has been well below theac- cepted dietary requirements (Sippola 1979, Varo ^&Koivistoinen 1981). There has also been evidence linking dietary seleniumcontent tosome diseases (e.g. cardiovasculardiseases, Salonen etal. 1982). These assumptions led in the early 1980'sto the need to investigate possibilities _toincrease dietary selenium intake

in the Finnish population.

Even though since 1969 commercial mineral mixtures had been fortified with inorganic selenium this hadnot had any distinct impact on selenium_content in milk or meat. There wasalso evidence from previous studies that organic forms of selenium would be trans- ferred into milk and retained in tissuesmore efficiently than inorganic Se (e.g. Jacobsenet al. 1965, Conrad and Moxon 1979, Maus et al. 1980). Thus fortifying selenium in fer- tilizers was anticipated to be the best means of supplying the Finnish population and also domestic animals with adequate selenium. In- vestigations of plant uptake of various sele- nocompoundswere succesfully carried^outby Dr. Toivo Yläranta. Based on his results, it wasconcluded that sodium selenate would be the most suitable form of selenium to be addedtofertilizers given prevailing conditions

in Finland (Yläranta

1984

^a).

This research programmewas established toinvestigate appropriate feed seleniumcon- tentfor achieving suitable seleniumcontent in milk and meatproducts. Prof. Liisa Syrjälä-

Qvist

^wasthe chairman of the group and Mr.

Pentti Aspila was requested _to work _as researcher. The programmewasimplemented in 1982 and its objective was to develop recommendations for raising the selenium content in cow milk and in beef to an ap- propriatelevel. ^Thechemical form ^{of Se}em- ployed in this study was sodium selenite, ^a reduced product of sodium selenate,because it is the best documented seleno compound and in foliar applications it has been similar to sodium selenate (Yläranta

1984

b). In this doctoral thesis the results from two experi- ments onlactatinggoats and the results from one experiment _on lactating dairy _{cows are} presented.

Since 1984 Finnish fertilizers have been sup- plemented with sodiumselenate, which has in- creased ^theselenium contentof feeds from the level of0.02 mg/kg DM to the level of 0.2 mg/kg DM. As_aresult the Secontentof ani- mal products has also increased and at the

present, the intake of Se from the average Finnish diet meets standard recommenda- tions. Meat products contribute 40 °/o and

milk products 20^% of the total selenium in- take (Ekholm et al. 1991).

2. REVIEW OF THE LITERATURE 2.1. Selenium metabolism in the rumen

Investigations on the chemical _nature of seleno-compounds of plant origin have dem- onstrated the presence of several selenocom- pounds, e.g. selenomethionine, ^selenocys- teine, selenite, selenocysteic acid, ^selenocys- tathione, Se-methylselenomethionine, Se- methylselenocysteine and others. (Peterson and Butler 1962, Jenkins and ^Hidiroglou 1967,^Shrift 1969,^Olson etal. 1970, ^Nigam and McConnell 1976, ^Burk 1976, ^Ulrey 1981, Gissel-Nielsen 1987). Of these the main selenium compounds ingested by rumi- nants are isologs of methionine and cystine.

In soya proteins selenium is probably in the form of selenocysteine (Mason and Weaver 1988). About 70 ^% of selenium is protein- bound in lucerne (Peterson and ^Spedding

1963, Jones and Godwin 1963) and about 60^% in brome grass (Jenkins and Hidiro-

glou 1967). Gissel-Nielsen(1976)suggested that normally 80 °/o of selenite-Se given to plants is incorporated into either protein or free amino acids. Fortifying sulphur and nitrogen might havesomeinfluenceonthe dis- tribution and chemical form of selenium in plants (Gissel-Nielsen 1982).

In therumen,plant proteins undergoexten- sive bacterial proteolysis resulting in the liber- ation ofaconsiderable proportion of the ami- no acids and reutilization for microbial pro- tein synthesis. Compared tosulphur, selenium has been shown to be incorporated in vitro morerapidly into bacterial protein; the incor- poration of selenite is faster than that for selenate _orselenomethionine(Paulson etal.

1968). A smallamount of inorganic selenium could also be incorporated into seleno-amino acids by rumen microbes. The compounds identifiedaremainlyselenomethionine, ^sele- nomethionineselenoxide, selenocysteineand compounds resembling taurine and homocys- teine (Hidiroglou et al. 1968, Hudman and Glenn 1984). Selenite and selenate, however, were not incorporated into selenomethionine even though under the _same conditions sul- phate-sulphur wasutilized in methionine syn- thesis with subsequent incorporation into microbial proteins (Paulson etal. 1968). At low sulphur intake (Pope et al. 1979)^or at high selenium intake (Rosenfeld 1962) incor- poration of inorganicselenium into microbial protein tends to increase. Incorporation of selenite is inhibited by sulphite and nitrite (Hudman and Glenn 1984).

Some selenomethioninecanbe metabolized by therumenbacteria toform selenocysteine in asuch way that both compoundsareincor- porated into bacterial protein (Whanger et al. 1967,^{Hidiroglou et}al. 1974). This con- version, ^{however, has not} been found in all studies (Paulson etal. 1968). Thereason for this might be the instability of selenocysteine (Butler and Peterson 1967). Incorporation of seleno-amino acids into_rumen bacteria is arapid process resulting in about half of the dosed selenomethionine being associated with rumen bacteria within the first 6 hrs. There- after the proportion of selenium associated withrumen bacteria is slowly decreased with acoinciding increase in proportion of ingesta

+protozoa associated Se. It is assumed that only a minor proportion of this selenium is bound to the _protozoa. The proportion of protein-bound selenium is foundtobe about 12

80^%in ingesta⁺protozoal fraction, but only about ^65% in bacterial protein (Hidiroglou et al. 1974). The average turnover rate for protein-bound selenium has been only 70 ^% of that observed for sulphur (Hidiroglou ^et al. ^1968).

Supplementing seleniumto sheep has been showntoincrease the number ofrumenbac- teria and also tohave some effecton rumen bacterial composition (Hidiroglou et al.

1968). This might result in improved microbial protein synthesis. The effect of sulphur on selenium metabolism may arise from the fact that increasing sulphur in the diet has in- creased the population of Desulphovibrio bac- teria in the rumen (Shrift 1973) and thus more selenium is enzymatically reduced to H₂Se (Popeetal. 1979). H₂Se is an unstable compound, which might be reduced further to elemental seleniumor to highly insoluble metal selenides (Hidiroglou and Jenkins 1973). Some of H2Se _can also be eliminated from the rumen via eructation(Lopez etal.

1968). Handerek and Godwin (1970) found one percent of dosed selenium in expired air.

However, Paulson et al. (1968) could not find any volatile selenium^evenunder thesame conditions in whichsomevolatile sulphurwas formed. High selenium intake(Lopez etal.

1968) and also insome caseshigh dietary pro- tein level (Ganther etal. 1966) may increase expired selenium.

Part of the selenium giventoanimals in or- ganic form or as sodium seleniteor selenate is metabolizedto elementalselenium, which is utilized to some extentby rumenmicrobes (Handerek and Godwin 1970, ^{Hudman and} Glenn 1984),or further to highly insoluble selenides which canbe utilized neither by mi- crobes nor ^by the host (Pope et al. 1979).

2.2. ^Absorption of selenium 2.2.1. Site

of

Se absorption

There isno netabsorption of selenite from the rumen (Wright and Bell 1966) although a small amount of selenomethioninemay be

absorbed (Hidiroglou and Jenkins 1973).

Absorption ofselenomethionine from theru- menis^arapid process, but accountsfor only about2 ^% of the total absorption of seleno- methionine. The absorption is similartothat of methionine(Venkov 1969). Cysteine is ab- sorbed from the _rumen to amuch lesser ex- tent than methionine (Lazarov and Ivanov

1970), but there_{are no} studies of the ruminal absorption of selenocysteine. Some selenium is, ^however, recycled back into therumen, mostly via saliva (Dejneka et al. 1979), resulting in almost _zeronet absorption from the rumen (Hidiroglou and Jenkins 1974).

In the omasum some selenomethionine is absorbed, but somework suggeststhat more selenium is secreted ^{back to the omasum} leading to the zero or even negative net ab- sorption of selenium from the omasum (Hidiroglou and Jenkins 1973). However, Langlands etal. (1986) found a slight ten- dency for decreased outflow of Se from the omasumwhen comparedtooutflow from the rumen.

Selenite may be absorbed in the abomasum to some extent in sheep (Wright and Bell 1966), but not in the rat (Whanger et al.

1976)orin swine (Wright and Bell 1966). In contrast toselenite,^some selenomethionineis absorbed in mice (Hanson and Jacobsen 1966),rats (Whanger etal. 1976) and sheep (Langlands etal. 1986). Absorption isgreater from the pyloric and fundic portions of the abomasum than from the mucosal epithelial

segment (Hanson and Jacobsen 1966).

The main site for selenium absorption varies from species tospecies, but is always located ina small intestine. In rats (Whanger et al.

1976)and in chicks ^(Pesti and Combs 1976,

Humalojaand ^Mykkänen 1986) selenite and selenomethionine are absorbed from the duodenum more efficiently than from the jejunumor ileum,but in sheep absorption is most efficient the from midjejunum (Hidi-

roglou and Jenkins 1974).Selenate-Se is ab-

sorbed inrats mostefficiently from the ileum followed in descending order by the proximal

jejunum and large intestine (caecum and co- lon) (Wolframm et al. 1985). In swine and sheep net absorption is _greatest in the distal 4/5 of the small intestine (Wright and Bell 1966), probably dueto extensive secretion of Se into the duodenum.

2.2.2. Mechanism

of

Se absorption

Most investigations on the absorption mechanism of Seare carriedoutwith labora-

tory animals and results in ruminants are limited. Ruminant nutritionists therefor_are compelled toextrapolatefrom data obtained from laboratory animals. Selenomethionine is transferred actively acrossthe intestinal wall in the golden hamster (McConnelland Cho 1967), the chick (Humaloja and ^Mykkänen 1986) and the rat (Thomson and Sterwart 1973). Transport is dependent upon the ATP- dependant sodium pump (McConnell and Cho 1967). The mechanism is probably the same as that for methionine. Also selenate transport may be active in rat ileum (Arduser etal. 1985,^{Wolframm et}al. 1985).

Absorption of selenate is morerapid than that of selenite(Turner etal. 1990). Selenate ab- sorption probably involvesa carrier-mediated mechanism, which is relatively unspecific for selenate (Wolframm et al. 1985), is depen- dant _{on a} Na⁺ gradient across the intestinal membrane (Arduser etal 1985), and ^mayde- rive energy from Na⁺K⁺-ATPase (Turner ^et al. 1990).

In contrast to selenomethionine, simple diffusion probably dominates in the absorp- tion of selenite (McConnell and Cho 1965, Wolframm et al. 1985) and selenocysteine (McConnell and Cho 1965). ^Humalojaand

Mykkänen (1986) have proposed, however, thatacarriersystem transportsselenite in the chick intestine.Furthermore, there is evidence that Se-dicysteine could be formed extracel- lularly from selenite and cysteine in pig (Wolframm et al. 1988) and sheep jejunum (Wolframmetal. 1987). Se-dicysteine would then be transportedacrossthe intestinal brush border by the active Na⁺-dependant trans-

port system existing for neutral amino acids (Prestonetal. 1974,^Wolframmetal. 1989).

L-cysteine stimulates absorption of Se from selenite probably by generation of selenodi- cysteine and cysteine selenopersulfide. In- tracellular formation of Se-dicysteine may contribute tothe stimulatory effect of L-cys- teine on the uptake of Se from selenite by maintaining the concentration gradient for passive uptake of selenite (Wurmli et al.

1989).

Reduced glutathione (GSH) and y-gluta- myltransferase (y-GT) may playsomerole in the absorption of selenite in therat intestine (Anundi etal. 1984). A largeamount of^GSH is excreted in the bile in rats(Siesetal. 1979, Eberle etal. 1981) probably resulting in the formation of oxidized selenodiglutathione (GSSeSG). This wouldact as a substrate for y-GT and result in the formation of amino acids and dipeptides, for which active trans- port mechanisms exist in intestinal cells. Se- lenium boundtothese products might thus be- come available for active transport (Anundi et al. 1984).

2.2.3. Factors affecting Se absorption The efficiency of selenium absorption ap- parently is not regulated by dietary selenium level (Lopez et al. 1968, ^{Cary et al. 1973,} Kiker and Burk 1974), although ^Humaloja and ^Mykkänen(1986) reported thatat atoxic Se dietary level selenium absorption tendedto increase. Supplementing methionineto ham- sters (McConnel and Cho 1967)and sulphur to sheep (Pope etal. 1979)or rats (Arduser etal. 1985) tendedtodecrease therateof sele- nium absorption although the results with sul- phur are not consistent (e.g. Paulson etal.

1966, White and Somers 1977). Some ca- tions, especially bivalent cations like silver (Rahim etal. 1986)and lead(Mykkänen and

Humaloja 1984,^Neatheryetal. 1987), may decrease selenium absorption when ad- ministered together with Se. However,neither Cu (Rahim et al 1986, ^{Koenig et} al. 1989), Fe, Cd, Mo, Mn (Rahim et al. 1986)nor Co 14

(van ^Ryssenetal. 1987) affected selenite ab- sorption. Absorption of selenium may be decreased on low or high calcium intake (Alfaro^etal. 1987), with the optimal dietary calcium level being 8 g/kg DM for selenium absorption in the dairy cow (Harrison and Conrad

1984

^b).

Vitamin A (Combs 1976) and ascorbic acid (Combs and Pesti 1976, Combs and Scott 1974) in chicks and ascorbic acid in rats (Rahim 1985) increase selenium absorption, suggesting that factors which inhibit the oxi- dation of dietary seleniumpromoteits absorp- tion. In contrast to this ^Mykkänen and Mutanen (1983) reported that moderate doses of ascorbic acid caninhibit the intestinal absorption of selenite in chicks when ad- ministered together intraduodenally. In their later study in non-fasted chicks (Mykkänen and Mutanen 1986) concluded that orally ad- ministered ascorbic acid had no effect _on selenium absorption, but there might besome differences between the Se ^sources.Lowered absorption duetoascorbic acid might be due toreduction of seleniteto elemental Se (Hill 1979). Ascorbic acid status of chicks has no effect on selenium absorption (Mykkänen and Mutanen 1983). The ionophores narasin and monensin have enhanced selenite absorp- tion in steers (Costa etal. 1985).

Inratselenium absorption is enhanced with increasing age (Raghib et al. 1986), but the opposite has been found insheep (Grace and Watkinson 1988).

2.2.4. Resecretion

of

Se into intestine A significant ^amountof absorbed selenium is excreted back to the intestine in rat (Im-

bach and ^Sternberg 1967, ^Gregus and Klaassen 1986), sheep (Hidiroglou and Jenkins 1974,^Langlandsetal. 1986) and bo- vine (Symonds et al.

1981 a and

^{b). Two to}

three times the amount of ingested Se enters the proximal portion of the small intestine in sheep (Langlands et al. 1986). Secretion is highest in the first fifth of the small intestine

(Wright and Bell 1966)or in the midjeju-

num (Hidiroglou and Jenkins 1974) in sheep and in the first fifth of the small intestine in swine (Wright and Bell 1966). Resecreted Se entersthe gastrointestinaltractmainly via sali- vaand bile. In sheep saliva Se concentration is reported tobe0.3 ng/1 and in bile is nearly three times higher (Langlands et al. 1986).

There _{was no} correlation between Se _concen- tration in blood and that of saliva or bile.

Dejneka et al. (1979) reported about 2 ^%, butLanglandsetal. (1986) asmuch as28 ^% ofa Se dosetobe excreted via bile in sheep.

The corresponding figures for theratare4^% (Levander and Baumann 1966)_to about6 ^% (Imbach and ^Sternberg 1967) and for cattle about2 °?o with selenite and about 1 ^% with selenate (Symonds_etal.

1981

b). Biliary _excre- tion doesnot increase proportionally with dos- agesuggesting that the hepatobiliarytransport

of selenium is saturable (Gregus and Klaas- sen 1986). The total secretion of selenite into the intestine in cattle is about 9 ^% of the to- tal dosage of selenite and about 14—17^% of selenate (Symonds et al.

1981

b), suggesting that digestive juice plays a major role. Inrats 11 ^%of selenitewassecreted in digestive juice (Imbach and ^Sternberg 1967). The major selenocompound secreted in bile is probably selenotaurocholic acid (Rosenfeld 1962). In cats 1.4 °/o of intravenously injected 75Se-se- lenomethionine was excreted in pancreatic

juice, mainly in protein-bound form (Hanson and Blau 1963).

2.3. Metabolism and chemical forms of selenium in tissues

Although the chemistry of seleniumresem- bles that of sulphur in severalrespects, these elements have important biochemical differ- encesand are thereforenot completely inter- changeable in animals. The quadrivalent Se in selenite tendstoundergoreduction, while the quadrivalent S in sulfite tends _to undergo oxidation. Thus Se compounds tend to be metabolized in animals to more reduced states, while S compounds tendto be oxidized (Combs and Combs 1984). Although the

analogous oxyacids of Se and S have com- parable strengths, the hydride H2Se is a much stronger acid than H2S. This difference is reflected in the dissociationconstantsof the selenohydryl group of selenocystine (pKa 5.24) and the sulphydryl group of cystine (pKa 8.25) (Huber and Criddle 1967). Whereas thiols such as cystine are mainly protonated atphysiological pH, the selenohydryl groups of selenols such as selenocystine are largely dissociated. This behavior appears tobe im-

portant in the catalytic role of selenium in selenoenzymes.

2.3.1. Selenium in erythrocytes and plasma Studies concerning selenium metabolism in erythrocytesare abundant and have beencar- ried out in rats (Burk 1973, Butler et al.

1985), mice (Sandholm, 1973

a,

Sandholm 1974), chicks (Jenkins et al. 1969), humans (Burketal. 1967,^{Lee et}al. 1969, Burk 1974, Mas et al. 1988), sheep (Wright and Bell 1966) and cattle (Sandholm

1973

^{b, Jenkins}

and Hidiroglou 1988). Glutathione peroxi- dase (GSH-Px) is the best documented seleno- compound in animals. The molecular weight of this enzyme in cattle erythrocytes has been reported to be 84 000 to 88 000 U (Oh etal.

1974, Jenkins and ^Hidiroglou 1988). The role of Se in GSH-Pxwasfirst demonstrated by Rotruck et al. (1973) and Flohe et al.

(1973) and the structure was described by Ladenstein (1979).

Plasma proteins do not bind selenium_as selenite unless erythrocytes are present, sug-

gesting that a transformation of seleniteoc- curs within the erythrocytes (Mas et al.

1988). Exchange of selenite between plasma and the erythrocyte isarapid process. With- in oneminute50 80 ^% of selenite selenium was accumulated inside erythrocytes in hu- mans (Lee et al. 1969), in mice (Sandholm

1973

^{a) and in rats} (Gasiewicz and Smith 1978). The uptake of selenite bybovine,chick and ovine erythrocytes is also rapid, but not asrapid_asthat of humanorrat erythrocytes (Jenkins 1968, Jenkins and^Hidiroglou 1972,

Sandholm 1973^{b, McMurray} and Davidson 1979). In a cow atlow plasma seleniumcon- centration(<0.2 mg/1) selenite^{wastaken up} in one to two minutes by erythrocytes and thereafter the ejection exceeded the uptake. At higher selenium concentrations (>2 mg/1) the bulk of selenium remained associated with erythrocytes. At averyhigh seleniumconcen- tration (200 mg/1) the uptake of Se by the red blood cells wasrestricted (Sandholm

1973

^b).

Some of selenite maystayunaltered inside the erythrocytes andextract asselenite. This hap- pens whenevera relatively small ^{number of} erythrocytes is present (McMurray and Davidson 1979).

The uptake of selenium by erythrocytes is dependant upon the availability of reduced glutathione (GSH) (Jenkins and ^Hidiroglou 1972, Sandholm

1973

^b, Gasiewicz and Smith 1978) or sulphydryl groups (Porter et al.

1979) and the amount of selenium in blood (Sandholm

1973

^{b). However,} increased sele- nite concentration tends to deplete erythro- cyteGSH (Gasiewicz and Smith 1978). Hae- moglobin is reported to be the binding site for newly-reduced Se and thus haemoglobin would play a role in the erythrocyte uptake of selenite (Mas et al. 1988). Jenkins and

Hidiroglou(1988) reported that in calves fed seleniumatadietary level of 5 ppm 35^to40^% of erythrocyte Se was in GSH-Px, 50 ^% in haemoglobin and 5 ^% in a selenite plus sele- nopolypeptide fraction72 hrs ^postdosing. In pigs Se has been reported to be distributed equally between GSH-Px and haemoglobin (Xia et al. 1985). In rhesus monkeys fed selenite68 ^%of erythrocyte Sewasassociated with GSH-Px while in animals fed seleno- methionine only 34 ^% of erythrocyte Sewas in GSH-Px. In contrast, more Se was asso- ciated with haemoglobin in animals fed selenomethionine than selenite (Butler ^etal.

1990). In haemoglobin Se is retained in the globin portion (Beilstein and ^Whanger

1986^a).

Subsequentto selenite's appearance in the erythrocyte it reacts with GSH toform sele- nodiglutathione (GSSeSG) (Ganther 1968,

Sandholm and ^Sipponen 1973). GSSeSG is further reduced by ^Nadph and glutathione reductase_to form selenopersulphide (GSSeSH) and further to H₂Se (Gasiewicz and Smith 1978). H2Se _{or a} similar reduced product of GSSeSG is the final product of seleniteme- tabolism in the rat erythrocyte. ^McMurray and Davidson (1979) called this productcom- pound X. However, release of selenium from rat erythrocytes _occurs through a different mechanism when compared tothat observed for diglutathione (GSSG) (Gasiewicz and Smith 1978). Selenite metabolism in animals leading to the formation of methylated sele- nides is believed to occurvia H2Se (Ganther and Hsieh 1974, Hsieh and Ganther 1977).

H₂Se is an unstable compound, which may undergo oxidation to elemental selenium (Ganther 1971) and further, combine with tissue macromolecules such as albumin (Hsieh and Ganther 1975) or other seleno- compounds.However, when_receptor sites_on plasma proteinsexist,compound X is excreted from the erythrocyteto be boundto this ac- ceptor protein. If plasma proteins are satu- rated compound X is retained by the erythro- cyteand becomes attachedtoproteins within the erythrocyte. These processes arereversi- ble. Binding of selenium within the erythro-

cyte is an alternative process and formation of protein-bound selenium in plasma is usually favored (McMurray and Davidson 1979). In therat85 °/o (Gasiewicz and Smith 1978) and in the sheep 90 ^%(McMurray and Davidson 1979) of the selenium released into plasma is found tobe protein-bound. The majority of this selenium is incorporated into selenocys- tines in polypeptides (Cummins and Martin

1967, Jenkins 1968). Incorporation of seleni- um into_serumproteins neednotbe enzymatic.

In addition^toproteins, biologically significant sulphydryl reducing agents such as CoASH, cysteine and GSH have also been reported to form selenotrisulphides upon reaction with selenous acids (Ganther 1968, Sandholm and ^Sipponen 1973). The reaction ofselenite with the thiolgroups of these compounds is an important pathway through which inor-

ganic selenium is initially incorporated into living systems (Kice 1981).

In contrast to selenite and selenomethio- nine, selenate is probably taken into the erythrocyte only by diffusion (Jenkins and

Hidiroglou 1972). Even though selenite as such does not bindto plasma proteins, H₂Se and GSSeSG arereadily incorporated toplas- ma proteins even in the absence of erythro-

cytes (Gasiewicz and Smith 1978). Albumin (Sandholm 1973

a,

Mas et al 1988)and an- other protein having a molecular weight of greater than200 000 U (Mas etal. 1988)are the initial acceptors for the selenium com- pound released from erythrocytes. Increasing the selenium dose has generally increased the total selenium binding to albumin (Jenkins and Hidiroglou 1988), but decreased the binding to the a-globulins in rats and dogs (Hirooka and Galombos 1966). Jenkinsetal.

(1969) postulated that albumin binds selenium only when supraphysiological levels of selenite are employed. This was confirmed by Her- man and McConnell (1974). Also McMur-

ray and Davidson (1979) found that some proteins other than albumin play _an impor- tantrole in the binding of selenite metabolized by erythrocytes. P-lipoprotein is suggestedto bean important site for selenium attachment (Sandholm 1975). Onlyasmallpercentage of the total plasma selenium is identified as selenium-dependant GSH-Px (Behne and Wolters 1979). In erythrocytes, however, GSH-Pxrepresentsmostof the totalselenium, being 75 ^% in sheep (Oh et al. 1974) and 100 °7o inrat (Behne and Wolters 1979). The proportion of erythrocyte Se associated with GSH-Px is dependant on the Se source, or- ganic forms of selenium being incorporated more efficiently into other proteins than into GSH-Px (Beilstein and Whanger 1986

a,

1986

^{b, Butler etal. 1990).}

Plasma proteins are important carriers for the incorporation of selenium into lympho- cytes. Uptake of protein-bound selenium by

lymphocytes isthree times _ashigh _as uptake of selenite. There areneither energynorpro- tein synthesis requirements for the uptake of

17

seleno proteins orseleniteby lymphocytes, but there is need for sulphydryl groups for uptake of both protein bound seleniumand ^selenite (Porteretal. 1979). Protein synthesis is also notneeded in leucocytes for uptake of selenite (Cavalieri ^et al. 1967).

2.3.2. Selenium metabolism in the liver The liver playsanessential roleas the sele- nium processing center of the body. In the liver selenium is converted ^{to compounds} which are available for other organs and through which selenium is excreted from the body (Behne and Höfer-Bosse 1984).

In several studies the plasma concentration of intravenously given seleniumhas decreased sharply reaching the minimum 15to 40 min- utesafter the injection (Sandholm

1973

^b,Al-

len and Miller 1981

a,

^Symonds et al.

1981

b). During the first 30 minutes bovine liver takes up 40 °/o of the selenium removed from the plasma (Symonds et al.

1981 a, 1981

b). After reaching the minimumlevel, ^the plasma selenium beginsto increase duetothe release of selenium from the liver backtoplas-

ma, a- and Y-globulins bind the selenium metabolitesreleased from the liver (Symonds et al.

1981

b). Motsenbocker and ^Tappel (1982

a, ₁₉₈₂

c) found inrat and monkey liver and in plasmaaselenocysteine-containing pro- tein, ^{which had}a short half-life. They called

it selenoprotein P. Other studies have ^con- firmed the hypothesis that this protein isserv- ingas aselenium transportation protein from the liverto other tissues (Motsenbocker and

Tappel 1984, ^{Beilstein et}al. 1984, Gometz and ^Tappel 1989). A similar selenium trans- portprotein has been identified in swine plas- ma (XIA et al. 1985). At a high intake of selenite Se deposition occurs mainly in the cytosol fraction of^theliver, differing from muscle,^{kidney and testeswhere}an excess of Se is deposited in the nuclear fraction (Dea-

genand ^Whanger 1985).

2.3.2.1 Enzymatic synthesis of selenocysteine in liver

Synthesis ofselenocysteine playsa veryim-

portant role in selenium metabolism. More than80 °?o of theselenium in selenite fed rats is found tobe in the form of selenocysteine (Hawkes etal. 1983, Butler etal. 1985). The number of proteins containing selenocysteine has been estimated to be 9 _to 19 in rats (Hawkesetal. 1983), GSH-Px being themost important compound containing selenocys- teine (Deagen and ^Whanger 1985).

Se dosed as selenomethionine is first de- posited in liver asselenomethionine, ^{but then} is slowly convertedto selenocysteine (Butler etal. 1985) probably through thesamepath- way as cysteine from methionine: selenome- thionine ^—* Se-adenosylselenomethionine

—* Se-adenosylhomocysteine ^—» seleno- homocysteine ^—» selenocystathione ^—*

selenocysteine. These reactions arecatalyzed by cystathione 0-synthetase and cystathione Y-lyase (Esaki etal. 1981,^{Soda et}al. 1981).

Selenocysteine is synthesized also from selenite (Hawkes and ^Tappel 1983). Conversion ^of seleniteto selenocysteine _to be used for the GSH-Px synthesisoccurs faster thanconver- sion to selenomethionine (Butler et al.

1985). However, the mechanism of seleno- cysteine synthesis from selenite stillremains unclear. One possible pathway is the reverse of the selenocysteine lyase reaction, which would contribute selenocysteine formation from alanine and H2Se (Esaki et al. 1985).

However, no selenocysteine has been found

to be synthesized from H2Se and serine. In rat liver selenocysteine, but not cysteine, is catabolized rapidly by selenocysteine lyase into alanine and H2Se (Esaki et al. 1985).

Selenocysteine is synthesized and catabolized continuously and thus the selenium retention in various tissues is in constant flux.

2.3.2.2. Selenoprotein synthesis in liver GSH-Px and selenoprotein P are synthe- sized in rat liver (Burk and ^Gregory 1982, Motsenbocker and ^Tappel 1982^{c). In}

ruminants, liver GSH-Px synthesis does^not play a major role in selenium metabolism.

Only 10 °7o ofhepatic seleniumwas incorpo-

rated into GSH-Px inruminants whilst 75 ^% of hepatic selenium inrat liver was found in the form of GSH-Px (Sunde et al. 1978).

There exist two possible mechanisms for in- corporation of selenium into GSH-Px. Sunde and Hoekstra (1980 and 1981) proposed post-translational incorporation, in whichor- ganic selenium has tobe converted into the inorganic form before its attachment to an amino acid residue of the GSH-Px polypep- tide chain. However, this theorywascriticized by Hawkes et al. (1982), who introduced translational incorporation involving seleno- cysteine attachingtothe polypeptide chain of GSH-Px during the protein synthesis. They also found selenium-specific selenocysteyl- tßNA in rat liver. Selenocysteyl-tRNA _was later confirmedtobe synthesized from selenite in ratliver (Hawkes and^Tappel 1983) and in- corporating its selenocysteine moiety to the polypeptide chain of GSH-Px during protein synthesis.

SelenoproteinP is synthesized both in vitro and in vivo in therathepatocyte. The mecha-

nism by which selenocysteine attaches into selenoprotein P is probably thesame aspro- posed for GSH-Px (Motsenbocker and ^Tap-

pel

1982

c). Hepatic synthesis of selenopro- tein P isarapid process and the highest con- centration in plasma has been reached three hours after injection of selenium. Selenopro- tein P is also removed rapidly from plasma to other tissues(Motsenbocker and ^Tappel

1982

c). Selenium deficient_rats are shown _to retaintwice ^asmuch selenium in selenoprotein P when compared toseleniumsufficient^rats (Motsenbocker and^Tappel

1982

c). Obvious- ly the liver prioritizes the synthesis of selenium transportation protein above that of other selenoproteins(e.g. GSH-Px).This enables_an animal tomaintaina sufficient level of avail- able selenium for tissues where selenium is most needed (Burk and ^Gregory 1982). In this way the liver redistributes selenium be- tween different tissues and Se compounds in selenium deficient animals (Motsenbocker and Tappel 1982^c).

2.3.2.3. Synthesis of methylated selenocompounds

Selenomethionine, selenocysteine and H₂Se are toxic for animals and are converted to harmless dimethyl selenide and trimethyl- selenonium ions (Martin 1973).Methylated selenocompounds are mainly produced in the liver, but the kidney also is abletosynthesize dimethylselenide (Ganther and Hsieh 1974, Hsieh and Ganther 1977), adirect precursor of trimethylselenonium ion (TMSe). Toxicity of TMSe is only one-tenth that of selenome- thionine and selenite (Obermeyer etal. 1971).

The biological role of TMSe is negligible and it is excreted readily in urine (Tsay et al.

1970, Foster et al.

1986

b). Methylation of selenomethionine probably arises from the synthesis of selenocysteine, which is further catalyzed to H₂Se and alanine via selenocys- teine lyase (Esaki et al. 1981, 1982). H2Se then is metabolized ^{further to TMSe.}

Selenocysteine probably is not first catabo- lizedto H₂Se when synthesizing TMSe (Fos-

ter and Ganther 1984). Precursors for di- methylselenide are H₂Se or other interme- diate products from the reduction of selenite.

In this reaction the hydrogen of the precur- sors is substituted enzymatically by methyl groups (Ganther and Hsieh 1974). There exist at least ^two differenttypes of methyl- transferases, one existing in hepatic micro- somesand the other in the cytoplasm(Hsieh and Ganther 1977). If the methylation ^ca- pacity isexceeded ^{the excess}of dimethylsele- nide is excreted via the lungs. TMSe isnotan inert compound and may undergo demethyla- tion followed by remethylation (Foster etal.

1986^a).

2.3.3 Selenium metabolism in kidney More than 99 ^% of kidney selenium is found in thecortexand onlyasmall fraction in the medulla(Lopez et al. 1968). Most of the cellular selenium ofgoat kidney is found in the nuclear fraction (Allen and Miller

1981

^{b). Inrats} about half of the kidney sele- nium is distributed in the cell^organelles with the highest concentration found in the lyso-

somes (Motsenbocker and ^Tappel

1982

^b).

Only_asmall fraction of this lysosomal Sewas identified chemically. Xia et al. (1985) re- portedmostof kidney Setobe associated with GSH-Px in pigs. Motsenbocker and ^Tappel (1982b) speculated that selenium in plasma selenoprotein P is madeavailable for the tis- sues by catabolizing selenoprotein P in the lysosomes of the cell. Thus high lysosomal selenium ^{content might}represent intermediate products of selenoprotein P catabolism.

In rat kidney, specific activity of⁷⁵Sewas highest one hour after injection of 75Se-la- beled seleniteor selenomethionine (Millar et al. 1973). In the kidney selenium is rapidly in- corporated into proteins and within a few hours after injection of selenite, selenate or selenomethioninemost of selenium has been found in the protein fraction (Millar 1972, Millar etal. 1973).

At least four different selenoproteins have been discovered in the kidney. Only 10 °/o of the selenium in therat kidney is in the form of GSH-Px (Behne and Wolters 1983), with most of GSH-Px in the mitochondrial frac- tion (Motsenbocker and ^Tappel

1982

^{b). In}

addition to GSH-Px, selenoprotein-Pl and two smaller selenoproteins have been found in the kidney (Motsenbocker and Tappel

1982

^{b). Thesetwo}smaller proteins may have aspecificfunction in the kidney, because^these selenoproteinswerefoundneither ^{in the} liver nortestis. It seemslikely that there exists some specific function for selenium in the kidney, because in the selenium-deficient sheep ^(Lo-

pez et al. 1968) and ^rat (Burk et al. 1973, Behne and Höfer-Bosse 1984)a high propor- tion of selenium is retained in kidney. Boz-

kurt and Smith (1981) speculated that selenoproteins might have someeffect against the toxicity of cadmium. It is more likely, however, that cadmium_reacts with selenite, but not with selenoproteins (Motsenbocker and Tappel

1982

^b).

After absorption organic Se is incorporated immediately into tissue proteins (Vokal- Borek 1980) and is ^notfiltered by nephrons.

Free selenomethionine could be filtered in the

glomerulus, but is readily reabsorbed in the proximal nephron _asis methionine(Robinson etal. 1985).

Excretion of selenium into urine with in- creasing dietary selenium intake is due to elevated production of organic selenium metabolites, like TMSe. In addition^toTMSe, Kiker and Burk (1974) were able to demon- stratethree other urinary selenium-containing products, which they calledU-2,U-3 and U-4.

These productswerenotformed in vitro from selenite. The methodology used by Kiker and Burk (1974) to identify these compounds was, however, later criticized by^Nahapetian etal. (1983), even though they also foundat low Se intake that non-TMSe contributesmost of the Se excreted in urine. TMSe was the primary selenocompound in rat urine when selenate, selenomethionine, selenocysteine, Se-methylselenocysteine or wheat selenium were fed (Palmer etal. 1970,^Nahapetianet al. 1983). U-2 alsowas detected after all these selenium sources were fed. Some volatile selenocompounds have been detected in urine (Palmer etal. 1970), especially when organ- ic selenium has been involved. Kiker and Burk (1974) demonstrated in vitro synthesis ofsome urinary compounds, which they re- ferred to as inorganic compounds. ^Nahape-

tian etal. (1983) postulated, however, ^that little_{or no} selenite is excreted in urine in the rat. The first selenocompounds occurring in urine after selenite injection to rats are inor- ganic, but later the proportion of inorganic metabolites of selenium in urine plateaus at the level of 20 °7o of total urinary Se in sele- nium sufficientratsand at 50 °7o in selenium deficient ones.

The proportion of TMSe in urine rises dra- matically with increasing selenium dose. The change was from 10^% to69^% (Nahapetian et al. 1983)or from 2 ^% to 70 ^% (Morris and Levander 1986) when the Se dose was increased by a factor of 100. Thus very little TMSe is excreted with selenium deficientdiets, but it is the primary excretion metabolite when the intake exceeds the requirement (Kiker and Burk 1974, ^Nahapetian et al. 1983).