Biochem. J. (1986) 238, 671-675 (Printed in Great Britain)

671

Riboflavin-binding protein Concentration and fractional saturation in chicken eggs as a function of dietary riboflavin

Har'old B. WHITE III,* John ARMSTRONG and Colin C. WHITEHEAD AFRC Poultry Research Centre, Roslin, Midlothian EH25 9PS, Scotland, U.K.

The concentration of riboflavin and riboflavin-binding protein were determined in the plasma, egg yolk and albumen from hens fed a riboflavin-deficient diet (1.2 mg/kg) supplemented with 0, 1, 2, 3, 10 and 40 mg of riboflavin/kg. We observed that the deposition of riboflavin in egg yolk and albumen is dependent on dietary riboflavin and reaches half-maximal values at about 2 mg of supplemental riboflavin/kg. The maximal amount of riboflavin deposited in the yolk is limited stoichiometrically by the amount of riboflavin-binding protein, whereas the maximum amount of riboflavin deposited in albumen is limited by other factors before saturation occurs. The amount of riboflavin-binding protein in yolk and albumen is independent of dietary riboflavin. If there is a specific oocyte receptor for riboflavin-binding protein, it cannot distinguish between the apo and holo forms of the protein. Riboflavin-binding protein is about six times more concentrated in yolk than in plasma. INTRODUCTION The hen's egg is perhaps the most complete single food (Williams et al., 1971). It contains all the nutrients for the 21-day development of a chick embryo. The yolk, which contains most of the nutrients, is derived from the plasma by endocytosis (Griffin et al., 1984). Our interest is in the deposition of riboflavin in the egg and the role of riboflavin-binding protein in this process. Petersen et al. (1947) showed that up to 30 % of ingested riboflavin can be transferred to the egg by a laying hen. They also showed that the riboflavin content of eggs is proportional to dietary riboflavin over the range 0-5 mg of riboflavin/kg of feed and more-or-less independent of dietary riboflavin above that range. This saturation phenomenon is attributable to a specific riboflavin-binding protein (Rhodes et al., 1959) to which all of the riboflavin in an egg is bound (Blum, 1967). Hens genetically unable to synthesize a functional riboflavin-binding protein deposit insignificant amounts of riboflavin in their eggs (Maw, 1954; Winter et al., 1967; Farrell et al., 1970). Thus the amount of riboflavin-binding protein appears to set an upper limit on the amount of riboflavin that can be deposited in an egg. On the basis of these observations and others, the presence of a specific receptor that recognizes the binding protein, but not the vitamin, was postulated (Muniyappa & Adiga, 1979; Miller et al., 198 la). There is indirect evidence both for and against this hypothesis. Experiments by Miller and others have shown that modification of protein-bound carbohydrate (Miller et al., 198 1a,b, 1982a) or phosphate (Miller et al., 1982b), or derivatization of amino acid side chains (Miller et al., 1982b; Hammer et al., 1971), can drastically reduce the uptake of the radiolabelled protein by the oocyte. Because not all of these modifications result in rapid clearance of riboflavin-binding protein from the plasma by the liver or other tissues, an oocyte receptor that distinguishes between the native and modified *

protein seems likely. Furthermore, specific receptors for vitellogenin (Yusko et al., 1981), immunoglobulins (Roth et al., 1976), and low-density lipoproteins (Krumins & Roth, 1981; Perry et al., 1984) have been demonstrated in the hen's oocyte plasma membrane. Despite these precedents for protein-specific receptors in this system, continuing efforts to demonstrate directly a receptor for riboflavin-binding protein have been unsuccessful (Benore-Parsons, 1986). Because the primary function of riboflavin-binding protein is to deposit riboflavin in the egg, it would seem likely that a specific receptor on the oocyte plasma membrane would be able to discriminate between apo and holo forms of the protein. However, Benore-Parsons et al. (1985) reported that deposition of the protein occurs in the absence of bound riboflavin. In order to verify this observation, we fed laying hens diets differing in their riboflavin content and monitored the amount of riboflavin-binding protein and its fractional saturation in the plasma, yolk and albumen. Both apo and holo forms of the protein are transferred to the egg. We conclude that, if a receptor exists, it does not distinguish between these two forms of the protein. EXPERIMENTAL Hens and diets The birds (ISA Brown) were obtained at 1 day old from a commercial hatchery. From 18 weeks of age they were housed in single-bird battery-cage units and had free access to the standard layer's diet and water. Individual egg production was recorded daily. The experimental period started when the birds were 26 weeks old. Diet compositions are given in Table 1. The standard layer's diet was found by analysis to contain 6.5 mg of riboflavin/kg, an amount that exceeded the minimum requirement of 2.2 mg/kg for egg-laying in hens

Permanent address and address for correspondence and reprint requests: Department of Chemistry, University of Delaware, Newark, DE 19716,

U.S.A.

Vol. 238

672

H. B. White III, J. Armstrong and C. C. Whitehead

Table 1. Composition of diets

Composition (g/kg) Standard

Riboflavindeficient

Barley 255 Wheat 193 300 Maize 242 Maize starch 348 Isolated soybean 170 protein Soybean meal 104 (extracted) Herring meal 48 Meat and bone meal 19 Grass meal 50 Limestone flour 68 80 CaHPO4,2H20 14 40 Salt 2 4 Vegetable oil 50 Mineral supplement* 2.5 2.5 Vitamin supplement At 2.5 Supplement Bt 5.5 * Supplied (mg/kg of diet): copper, 3.5; iodide, 0.4; iron, 80; magnesium, 300; manganese, 100; zinc, 50. t Supplied (per kg of diet): retinol, 2.0 mg; cholecalciferol, 20,g; a-tocopherol, 17 mg; riboflavin, 4 mg; nicotinic acid, 28 mg; pantothenic acid, 10 mg. t Supplied (per kg of diet): retinol, 3.0 mg; cholecalciferol, 50 jg; a-tocopherol, 10 mg; menadione, 1.2 mg; thiamin, 1.5 mg; folic acid, 0.4 mg; pyridoxine, 4 mg; nicotinic acid, 20 mg; panthothenic acid, 20 mg; cyanocobalamin, 10 jug; choline chloride, 0.5 g; DL-methionine, 3 g, L-tryptophan, 0.6 g.

(National Research Council, 1984). The riboflavindeficient diet was found to contain 1.2 mg of riboflavin/ kg. Analysis of supplemented diets confirmed that the appropriate amounts of riboflavin had been added to each. Experimental design A total of 12 hens, all with a high rate of egg-laying, were allocated to each of eight experimental groups. Hens were housed in individual battery cages. At the start of the experiment, six groups were transferred to the riboflavin-deficient diet supplemented with 0, 1, 2, 3, 10, or 40 mg of riboflavin/kg. The remaining two groups were maintained on the standard layer diet as controls. Plasma samples from each bird were taken at weekly intervals and pooled by groups. Individual egg production was monitored, and eggs laid on days 0, 7, 14, and 21 were collected for analysis of riboflavin and riboflavinbinding protein. If an egg was not available from a hen on a particular day, the egg laid by that bird on the previous day was taken, if available. Preparation of samples A 1 ml sample of blood was taken from a wing vein of each bird and pooled with blood samples from the other hens in the group in a chilled plastic tube containing 15 jug of a 15 mg/ml solution of heparin (Sigma Chemical Co., St. Louis, MO, U.S.A.). Red blood cells were removed by centrifugation (10 min at 1000 g) and the plasma was stored frozen at -20 'C.

Egg yolk and albumen were separated. Albumen adhering to yolks was removed by gently rolling yolks on paper towels. The yolk and albumen from each egg was weighed. The separately pooled albumen and yolk samples from each group were homogenized with a motor-driven propeller. Albumen samples were stored frozen at -20 C° without dilution. Yolk samples were diluted (25% w/w) and homogenized with 25 mmTris/HCl, pH 7.5, before being frozen. Radioligand binding assay for riboflavin-binding protein The assay for riboflavin-binding protein (Lotter et al., 1982; White, 1986) is based on the protein's ability to bind riboflavin specifically with high affinity (KD 2 nM). D-[2-14C]Riboflavin (52.4 mCi/mmol; Amersham International) was incubated with diluted samples at room temperature for 10 min by which time added radioactive riboflavin had come to isotope equilibrium with endogenous unlabelled riboflavin. Bound and free riboflavin were separated by DEAEcellulose chromatography and the bound radioactivity was determined by liquid-scintillation counting. The concentration of riboflavin-binding sites and the fractional saturation by endogenous riboflavin was determined graphically. Although a fractional saturation of less than zero is impossible, values less than zero can be obtained by graphical analysis on samples that contain little or no endogenous riboflavin. This method assumes that binding is reversible, that isotope equilibrium is attained during the incubation period, that free and bound riboflavin are completely separated and that there is a total recovery of bound riboflavin. In order to correct for any losses, a sample of homogeneous apo-(riboflavin-binding protein) was used as a standard in each set of analyses. It was observed that the concentration of the binding protein based on the specific radioactivity of the bound [14C]riboflavin underestimated the actual protein concentration by 40 %. Fluorescence assay for riboflavin Because virtually all the flavin in egg and plasma from laying hens is riboflavin bound non-covalently to riboflavin-binding protein, there was no need to acid-hydrolyse these samples before analysis. The following procedure was adapted from that described by Koziol (1971). Plasma, albumen and yolk samples were thawed and diluted 10-, 20- and 40-fold (w/v) respectively with 25 mM-Tris/HCl, pH 7.5. A 2 ml portion of diluted sample was mixed with an equal volume of 20% (w/v) trichloroacetic acid. After 10 min, the precipitated protein was removed by centrifugation. To 3.0 ml of the supernatant were added 0.75 ml of 4 M-KH2PO4 and the fluorescence was measured at 530 nm with excitation at 450 nm in a Perkin-Elmer 3000 fluorimeter. All transfers and incubations were conducted in the dark or in subdued light. Riboflavin was recrystallized twice from acetic acid and its concentration in stock solutions was determined spectrophotometrically at 260 nm, 375 nm and 450 nm (Yagi, 1962). Internal and external standards were used to determine riboflavin concentrations in the samples. Light-scattering caused by turbidity was a problem in albumen samples. The light-scattering contribution to the measurements on these samples was determined by re-analysing the samples after photolysing the riboflavin 1986

Deposition of riboflavin in eggs

673

Table 2. Egg production and riboflavin content of plama, yolk and albumen from hens fed different amounts of riboflavin

Treated

Supplemental riboflavin

(mg/kg)...

0

1

2

3

10

40

Control

Egg production (% hen day)§ Pretreatment week 90.0 + 6.9* Week 3 75, 85t 32 81 74 76 80 83 Riboflavin content of: Plasma (sg/ml) Pretreatment 0.88 + 0.04* Day 7 0.18 0.38 0.57 0.70 0.80 0.87 Day 14 0.04 0.17 0.37 0.42 0.68 0.53 Day 21 0.04 0.18 0.31 0.44 0.70 0.89 0.87, 0.75t Yolk (jug/g) Pretreatment 4.94+0.31* Day 7 2.38 3.38 1.61 3.77 4.61 4.56 Day 14 0.85 1.64 2.66 3.74 4.72 4.64 Day 21 0.77 1.18 2.28 2.90 4.74 4.51 5.08,. 5.62t Albumen (,ug/g)$ Pretreatment 3.90+0.16 2.47 Day 7 0.58 1.54 2.75 3.47 3.47 Day 14 0.49 0.96 2.05 3.37 4.07 4.42 Day 21 0.58 0.68 1.59 2.55 3.98 4.22 4.05, 3.96 * Mean + S.D. for the six groups immediately before the treatment period. t Values for two groups of hens maintained on the control diet throughout the experimental period. t All values are corrected for light-scattering. § '% Hen * day' is a convenient unit used in poultry research to express the rate of egg laying for a group of hens. Its value is given by the equation: % hen -day = a/bc x 100, where a is the number of eggs laid by b hens in c days.

in sunlight for 1 h. Destruction of riboflavin is complete under these conditions. Riboflavin in the standard and riboflavin-deficient diets was determined as described above, but after acid hydrolysis (Koziol, 1971). There was a significant amount of fluorescent material in the riboflavin-deficient feed samples that was not destroyed by photolysis. The portion that was destroyed by sunlight (- 40%) was considered to be riboflavin. Because other fluorescent compounds may also be destroyed, the riboflavin estimate for the riboflavin-deficient sample is an upper limit. RESULTS Riboflavin content of plasma, yolk and albumen The concentration of riboflavin in plasma, yolk and albumen from laying hens displays saturation behaviour in relation to the amount of riboflavin in the diet (Table 2). Half-maximal deposition in yolk and albumen occurs on diets containing a total of about 3 mg of riboflavin/kg (the riboflavin-deficient diet supplemented with 2 mg of riboflavin/kg). There is little, if any, additional deposition of riboflavin when the dietary riboflavin is increased from 10 to 40 mg/kg. The transfer of hens from the control diet to diets containing 3 mg of supplemental riboflavin/kg or less caused a marked decrease in the riboflavin content of plasma, yolk and albumen. It would appear from the kinetics of these changes that internal sources of riboflavin turn over and equilibrate with dietary riboflavin within 2 weeks. Vol. 238

Egg production by the riboflavin-deficient- group was drastically decreased, because several hens ceased laying and those that continued to lay produced fewer eggs (Table 2). Analyses from this group were made on eight eggs at day 14 and six eggs at day 21, whereas a minimum of ten, but usually 11 or 12 eggs, were represented in the samples from the other groups. Because plasma samples were taken from all birds in a group regardless of their laying status and pooled, the riboflavin-deficient plasma samples come from a heterogeneous population and are not directly comparable with corresponding plasma samples from the other groups. Riboflavin-binding-protein content of plasma, yolk and albumen The concentration of riboflavin-binding protein in plasma, yolk and albumen is independent of dietary riboflavin (Table 3). The values for samples taken on successive weeks were so similar that they were averaged by groups. The concentration of riboflavin-binding protein in yolk is about six times greater than that in the plasma from which it is derived (6.08 + 0.48 for all but the riboflavin-deficient group). This value in turn is similar to the factor by which riboflavin is concentrated in yolk relative to plasma (6.75 + 1.35). In other words, the concentration ratio is virtually the same for the vitamin and its binding protein and independent of dietary

riboflavin.

Thepercentagesaturationofriboflavin-bindingprotein,

determined by the graphical method of Lotter et al. (1982) and tabulated in Table 3 or by comparing the concentrations of riboflavin in Table 2 with the

H. B. White III, J. Armstrong and C. C. Whitehead

674

Table 3. Concentration and saturation of riboflavin-binding proteins in the plasma, yolk and albumen from hens fed different amounts of riboflavin Treated

Supplemental riboflavin

(mg/kg) ...

0

1

2

3

10

40

Control

[Riboflavin-binding protein]* in:

Plasma (mg/l) Pretreatment (n = 6) 64+7 64+14 59+19 Treatment 91, 92t 67+12 85+27 65+9 -: Yolk (mg/kg) Pretreatment (n = 6) 568+ 119 390+35 407+24 504+ 118 427+43 Treatment 314+41 405+ 15 501,515t Albumen (mg/kg) 905+75 Pretreatment (n = 6) 757+33 820+148 685+19 811,800t 728+53 786+112 782+13 Treatment Percentage saturation* in: Plasma 109+28 _ Pretreatment 90+12 104+50 48+30 65+ 10 97,103t 23+ 15 Treatment Yolk 78+40 _ _ _ Pretreatment 118+40 11 + 19 64+48 85+ 15 108+25 113,103t 31 +6 Treatment Albumen 42+12 Pretreatment 21+26 30+12 21+17 17+11 40,37t -16+16§ -9+17§ Treatment * Values are means + S.D. for samples taken on day 0 for the pretreatment group and on days 7, 14 and 21 for the treatment groups; 1 mg of riboflavin-binding protein binds 12.5 ,g of riboflavin. t Values at day 21 for two groups of hens maintained on the control diet throughout the experimental period. t Values declined through the experimental period as more hens ceased to lay. § Saturation values were determined graphically. Negative values are 'physically impossible', but here provide an indication of the error in the method. -

-

riboflavin-binding capacity derived from the data in Table 3, show that riboflavin-binding protein is saturated with riboflavin only when dietary riboflavin is high (> 10 mg/kg). There is little evidence for significant amounts of unbound riboflavin in any of the samples. DISCUSSION Riboflavin deposition in eggs Our results for the deposition of riboflavin in the yolk and albumen of chicken eggs in response to dietary riboflavin agree with those of Stamberg et al. (1946) and Petersen et al. (1947). At the time of their work, it was not known that the transfer of riboflavin to the egg was dependent upon riboflavin-binding protein. The data presented here demonstrate that the amount of riboflavin in an egg is limited by the amount of riboflavin-binding protein and that, even at high riboflavin intake, little, if any, unbound riboflavin appears in the egg. Whereas the deposition of riboflavin in yolk involves the direct transfer of the vitamin-protein complex from plasma, the deposition of riboflavin in albumen is not understood. Chickens are atypical birds in that more than half of the riboflavin in their eggs is in the albumen (Feeney & Allison, 1969). As a consequence, their albumin has a definite yellow colour compared with the colourless albumen of most other birds. The fact that all bird eggs examined have riboflavin-binding protein in their albumen (Feeney & Allison, 1969) that is synthesized in the oviduct (Mandeles & Ducay, 1962), implies that the oviduct in the chicken has the capacity to accumulate riboflavin from the blood. It has been

-

-

-

-

assumed that this occurs by the uptake and lysosomal destruction of plasma riboflavin-binding protein (Miller et al., 1982a). This releases riboflavin to be bound by the newly synthesized binding protein. The distinctive post-translational glycosylations of plasma and albumenderived riboflavin-binding protein (Miller et al., 1982a; Hamazume et al., 1984; Norioka et al., 1985) seem to preclude a direct transfer from the plasma to the albumen. Our observation that riboflavin-binding protein in chicken albumen remains less than half-saturated implies that it is the capacity of the plasma-to-oviduct transfer process rather than the amount of binding protein that limits the amount of riboflavin deposited in albumen. Deposition of riboflavin-binding protein in the egg The results of Benore-Parsons et al. (1985), which suggest that apo-(riboflavin-binding protein) is deposited in yolk, have been confirmed. This indicates, that, if there is a receptor for this protein on the oocyte plasma membrane, it is unable to exclude the apoprotein. Furthermore, the fact that the concentration ratio of riboflavin between yolk and plasma is indistinguishable from that of riboflavin-binding protein suggests there is no discrimination between apo and holo forms during yolk deposition. This is consistent with the physicochemical behaviour of phospho groups on riboflavinbinding protein. Miller et al. (1982b) showed that removal of phospho groupsfrom riboflavin-bindingproteindrastically reduced the deposition of the 125I-labelled protein in oocytes. It was presumed that a receptor on the oocyte plasma 1986

Deposition of riboflavin in eggs

membrane was recognizing these phospho groups and that the conformation of the highly phosphorylated region of the sequence (Fenselau et al., 1985) would change upon binding riboflavin. However, the 31P-n.m.r. spectra of apo- and holo-proteins were indistinguishable, despite the presence of well-dispersed resonances (Miller et al., 1984). Our observations imply that the conformational differences between apo- and holo-protein observed with c.d.. and o.r.d. by Zak et al. (1972) also are not recognized by a receptor, if such exists. Although these results are contrary to our expectations, they may relate to a more generalized and less discriminating process associated with the endocytosis of yolk proteins. The results observed for riboflavin and riboflavinbinding protein here are in contrast with those for biotin and its two binding proteins in laying hens (White & Whitehead, 1985). There only the saturated proteins are found in plasma and yolk, and the production of the proteins is in some way controlled by the availability of biotin. Here the production and deposition of riboflavinbinding protein is unaffected by the availability of riboflavin. Dynamics of riboflavin-binding protein in plasma There have been a number of reports on the half-life of labelled riboflavin-binding protein in laying hens and the effects of various modifications (Hammer et al., 1971; Miller et al., 198 1a,b, 1982a,b). The observed values of 2 h or less are in contrast with the 6-10 h for immature birds without functioning ovaries (Murthy & Adiga, 1977). In the studies with laying hens, only 10-12% of the injected labelled protein was deposited in yolk. Thus the observed half-lives may reflect atypical clearance of damaged protein rather than typical metabolism of the native protein. If all riboflavin-binding protein in the plasma were destined for the yolk with a half-life of 2 h, the total amount of the circulating protein should be about 1 mg. This is considerably less than the value of 4 mg that we calculate in an estimated blood volume of 105 ml (Newell & Shaffner, 1950). It seems apparent under normal physiological conditions that either a relatively small proportion of the circulating riboflavin-binding protein is destined for the yolk or the plasma half-lives are considerably longer than estimated with labelled riboflavin-binding protein. We thank Ms. Christine Murnin for her expert technical assistance. H. B. W. was supported by National Institutes of Health grants AM27873 and AM34445, and the AFRCUnderwood Fund during a sabbatical leave at the Poultry Research Centre.

REFERENCES Benore-Parsons, M. (1986) Ph.D. Thesis, University of Delaware Received 6 January 1986/9 May 1986; accepted 21 May 1986

Vol. 238

675 Benore-Parsons, M., Yonno, L., Mulholland, L., Saylor, W. W. & White, H. B., III (1985) Fed. Proc. Fed. Am. Soc. Exp. Biol. 44, 1548 Blum, J.-C. (1967) Le Metabolisme de la Riboflavine chez la Poule Ponduse, p. 161, F. Hoffmann-LaRoche et cie, Paris Farrell, H. M., Buss, E. G. & Clagett, C. 0. (1970) Int. J. Biochem. 1, 168-172 Feeney, R. E. & Allison, R. G. (1969) Evolutionary Biochemistry of Proteins p. 290, Wiley-Interscience, New York Fenselau, C., Heller, D. N., Miller, M. S. & White, H. B., III (1985) Anal. Biochem. 150, 309-314 Griffin, H. D., Perry, M. M. & Gilbert, A. B. (1984) in Physiology and Biochemistry of the Domestic Fowl (Bell, D. J. & Freeman, B. M., eds.), vol. 5, pp. 345-380, Academic Press, London Hamazume, Y., Mega, T. & Ikenaka, T. (1984) J. Biochem. (Tokyo) 95, 1633-1644 Hammer, C., McDonald, K., Saylor, E. M., Buss, E. G. & Clagett, C. 0. (1971) Poultry Sci. 50, 938-944 Koziol, J. (1971) Methods Enzymol. 18B, 253-285 Krumins, S. A. & Roth, T. F. (1981) Biochem. J. 196, 481-488 Lotter, S. E., Miller, M. S., Bruch, R. C. & White, H. B., III (1982) Anal. Biochem. 125, 110-117 Mandeles, S. & Ducay, E. (1962) J. Biol. Chem. 237, 3196-3199 Maw, A. J. G. (1954) Poultry Sci. 33, 216-217 Miller, M. S., Buss, E. G. & Clagett, C. 0. (1981a) Biochim. Biophys. Acta 677, 225-233 Miller, M. S., Buss, E. G. & Clagett, C. 0. (198 ib) Comp. Biochem. Physiol. 69B, 681-686 Miller, M. S., Bruch, R. C. & White, H. B., III (1982a) Biochim. Biophys. Acta 715, 126-136 Miller, M. S., Benore-Parsons, M. & White, H. B., III (1982b) J. Biol. Chem. 257, 6818-6824 Miller, M. S., Mas, M. T. & White, H. B., III (1984) Biochemistry 23, 569-576 Muniyappa, K. & Adiga, P. R. (1979) Biochem. J. 177, 887-894 Murthy, U. S. & Aiga, P. R. (1977) Biochem. J. 166, 647-650 National Research Council (1984) Nutrient Requirements of Poultry, 8th edn., National Academy Press, Washington, DC Newell, G. W. & Shaffner, C. S. (1950) Poultry Sci. 29, 78-87 Norioka, N., Okada, T., Hamazume, Y., Mega, T. & Ikenaka, T. (1985) J. Biochem. (Tokyo) 97, 19-28 Perry, M. M., Griffin, H. D. & Gilbert, A. B. (1984) Exp. Cell Res. 151, 433-446 Petersen, C. F., Lampman, C. E. & Stamberg, 0. E. (1947) Poulty Sci. 26, 180-186 Rhodes, M. B., Bennett, N. & Feeney, R. E. (1959) J. Biol. Chem. 234, 2054-2060 Roth, T. F., Cutting, J. A. & Atlas, S. B. (1976) J. Supramol. Struct. 4, 527-548 Stamberg, 0. E., Petersen, C. F. & Lampman, C. E. (1946) Poultry Sci. 25, 327-329 White, H. B., III (1986) Methods Enzymol. 122, 221-226 White, H. B. III & Whitehead, C. C. (1985) Proc. Int. Congr. Nutr. 13th (abstr.) p. 53 Williams, R. J., Heffley, J. D. & Bode, C. W. (1971) Proc. Natl. Acad. Sci. U.S.A. 68, 2361-2364 Winter, W. P., Buss, E. G., Clagett, C. 0. & Boucher, R. V. (1967) Comp. Biochem. Physiol. 22, 897-906 Yagi, K. (1962) in Methods of Biochemical Analysis (Glick, D., ed), vol. 10, pp. 319-356, Interscience, New York Yusko, S. C., Roth, T. F. & Smith, T. (1981) Biochem. J. 200, 43-50 Zak, Z., Ostrowski, W., Steczko, J., Weber, M., Gizler, M. & Morawiecki, A. (1972) Acta Biochim. Pol. 19, 307-323