Binding of Lipophilic Nutrients to β-Lactoglobulin Prepared by Bioselective Adsorption1

Binding of Lipophilic Nutrients to b-Lactoglobulin Prepared by Bioselective Adsorption1 QIWU WANG, JONATHAN C. ALLEN, and HAROLD E. SWAISGOOD2 Southeast Dairy Foods Research Center, Department of Food Science, North Carolina State University, Raleigh, 27695-7624

ABSTRACT The binding of the lipophilic nutrients, retinal, vitamin D2, and retinyl palmitate by b-lactoglobulin was measured by analysis of changes in the fluorescence of the tryptophanyl residue of b-lactoglobulin or the retinyl moiety. The fluorescence intensity of the tryptophanyl residue was quenched by retinoid or vitamin D binding but was enhanced by palmitate binding. The analysis of competitive binding experiments with palmitate indicated that retinal and palmitate did not compete for the same site; however, vitamin D2, which binds with a stoichiometry of 2, appeared to displace palmitate at higher concentrations. Also, the retinoids and vitamin D2 were bound more tightly than was palmitate. The results are consistent with the model in which the retinoids and vitamin D2 bind in the calyx formed by the b-barrel; palmitate and a second molecule of vitamin D2 bind in a surface pocket near the dimer contact region. Retinyl palmitate, which has both moieties, appeared to bind at both sites. ( Key words: binding lipophilic nutrients, blactoglobulin, bioselective adsorption) INTRODUCTION b-Lactoglobulin, the major protein in bovine whey, is classified as an octinin member of the lipocalycin family (5, 6, 18). Thus, its structure is represented by an eight-stranded antiparallel b-barrel, with an added a-helix lying on the surface (14, 17). Because of its high affinity binding of retinoids (3, 4, 8, 26) and some fatty acids (10, 20, 22), b-LG has been proposed to function biologically in retinol transport ( 1 2 ) or in the enhancement of pregastric esterase

Received June 1, 1998. Accepted September 23, 1998. 1Paper Number FSR 98-14 of the Journal Series of the Department of Food Science, North Carolina State University, Raleigh 27695-7624. The use of trade names in this publication does not imply endorsement by the North Carolina Agricultural Research Service of the products named nor criticism of similar ones not mentioned. 2Corresponding author. 1999 J Dairy Sci 82:257–264

through the binding of inhibitory fatty acids (18). This protein has two hydrophobic pockets that potentially are capable of binding lipophilic molecules: one in the calyx formed by the b-barrel ( 1 7 ) and the other between the a-helix and the surface of the barrel (14). Most evidence indicates that retinoids bind in the calyx (2, 3, 16, 23, 24) and palmitate binds elsewhere (16, 22); the highest affinity site is most likely near the monomer contact surface in the dimer (10, 25). Thus, the characteristics of palmitate binding are strongly dependent on protein concentration (25), but binding of retinoids is independent of protein concentration (unpublished observation). Binding of retinol by b-LG quenches the fluorescence of Trp19 ( 3 ) , which lies at the bottom of the calyx (17); hence, fluorescence changes have been used to characterize the binding of retinoids (6, 11, 23) and vitamin D (24). However, interaction of bLG with palmitate increases the tryptophanyl fluorescence (10, 19); consequently, this enhancement of fluorescence has been used to characterize the binding of fatty acids. b-Lactoglobulin isolated by bioselective adsorption on N-retinyl-Celite ( 2 6 ) exhibits a high affinity and a larger capacity for binding retiniods than does b-LG in whey (23). These characteristics suggest that this protein ingredient could be an excellent carrier of lipophilic nutrients in low fat or nonfat foods. In this study, both the quenching and enhancement of fluorescence were used to characterize binding of palmitate, retinal, vitamin D2, and retinyl palmitate. The latter compound, which is commonly added as a vitamin supplement, contains both retinyl and palmitate moieties and thus may bind in several sites. MATERIALS AND METHODS Materials Palmitic acid (Sigma grade, approximately 99%), vitamin D2 (ergocalciferol, reference standard), alltrans-retinal, and all-trans-retinyl palmitate were purchased from Sigma Chemical Co. (St. Louis, MO). Solutions of the appropriate concentrations determined by weight of each were prepared in absolute

257

258

WANG ET AL.

ethanol. Solutions of vitamin D2 and retinal were prepared fresh daily in the dark and under anaerobic conditions to prevent oxidation and isomerization. b-Lactoglobulin was obtained by bioselective adsorption on N-retinyl-Celite as previously described (26). This preparation was greater than 96% pure as determined by size-exclusion chromatography.

data (4, 7, 23). Data were plotted according to the linear relationship PTa = (1/n)[ RTa/(1 – a)] – Kapp d /n

[1]

where RT = total ligand concentration, PT = total protein concentration, and a = fraction of unoccupied ligand-binding sites on the protein given by

Fluorescence Spectroscopy Fluorescence spectra were recorded at 27°C with a System 3 Scanning Spectrofluorometer (Optical Technology Devices, Inc., Elmsford, NY) using the ratio mode. Fluorescence changes caused by binding of the ligands were measured upon titration of 2.5 ml of a 20 mM b-LG solution in 50 mM sodium phosphate, pH 7.0, with 3-ml increments of the ligand solution. To eliminate the effects of protein dilution and the possible fluorescence changes induced by ethanol, a blank b-LG solution was titrated with ethanol, and the fluorescence changes were subtracted from those of the protein-ligand solution at each point. In these experiments, the fluorescence intensity at 332 nm of the original protein solution was normalized to 1 (10). A typical emission spectrum for b-LG was obtained with a 20 mM solution of the protein in 50 mM sodium phosphate, pH 7.0, both in the presence and absence of 20 mM palmitate. In the presence of palmitate, fluorescence emission at 332 nm (excitation at 287 nm) was enhanced, but binding of vitamin D2 or retinal quenched the 332-nm fluorescence. The interaction of retinyl palmitate with b-LG (10 mM) was examined by recording the emission spectrum in the presence and absence of 10 mM retinyl palmitate from 200 to 400 nm with an excitation wavelength of 280 nm and a scan speed of 50 nm/min. Also, the emission spectrum of retinyl palmitate (10 mM) was recorded from 350 to 600 nm in the presence and absence of 10 mM b-LG at an excitation wavelength of 342 nm and a scan speed of 50 nm/min to characterize the fluorescence change resulting from binding of the retinyl moiety. Binding was determined by titrating 3 ml of a 10 mM solution of b-LG in 50 mM sodium phosphate, pH 7.0, with 3-ml increments of ligand solution. The ethanol added to the protein solution never exceeded 3% (vol/vol). Three blank controls were used: retinyl palmitate, palmitate, and ethanol at concentrations identical to those used in the protein titration. Both the number of binding sites per molecule of protein, n, and the apparent dissociation constant of the complex, Kapp d , were evaluated from the titration Journal of Dairy Science Vol. 82, No. 2, 1999

a = ( F – Fsat) / ( F o – Fsat) where F = fluorescence intensity of the solution, Fsat = fluorescence intensity of b-LG saturated with ligand, and Fo = initial fluorescence intensity. Competitive Binding of Palmitate, Retinal, and Vitamin D2 b-Lactoglobulin (0.2 mM in 50 mM sodium phosphate, pH 7.0) was incubated in ultrafiltration cones overnight at 23°C with 1 ) 0.2 mM palmitate, 2 ) 0.2 mM palmitate plus 0.2 mM all-trans-retinal, or 3 ) 0.2 mM palmitate plus 0.4 mM vitamin D2. After incubation, the solution was centrifuged at 7000 rpm for 10 min, and the palmitate concentration in the ultrafiltrate was determined colorimetrically ( 1 ) . Benzene (1.0 ml), 1.0 ml of cupric acetate-pyridine solution (5% aqueous solution of cupric acetate adjusted to pH 6.0 with pyridine), and 100 ml of the ultrafiltrate or standard palmitate solution were added to a test tube. The phases were vortexed for 30 s and then were separated by centrifugation for 5 min with a clinical centrifuge (International Equipment Co., Needham, MA). The concentration of palmitate in the organic phase was determined from a standard curve prepared by measuring the absorbance at 715 nm. Effect of pH on Binding and Fluorescence Measurements Although the fluorescence of indole is independent of pH over the approximate range of 3 to 11, the fluorescence of Trp in proteins may be affected by ionization of neighboring prototropic groups or by conformational changes (9, 13). Therefore, the fluorescence of b-LG as a function of pH was determined at 332 nm with excitation at 287 nm. The effect of pH on ligand binding was determined from similar measurements of solutions containing 19.2 mM b-LG and molar ratios of 1, 1, and 2 of palmitate, retinal, or vitamin D, respectively, in 50 mM phosphate adjusted to various pH with phosphoric acid.

LIPOPHILIC NUTRIENT BINDING TO β-LACTOGLOBULIN

259

TABLE 1. Apparent dissociation constants ( Kapp d ) for palmitate bound to b-LG and binding stoichiometry ( n ) in the presence and absence of retinal or vitamin D2.1 Condition2

Kapp d

b-LG b-LG + Retinal b-LG + Vitamin D2

(× X 4.9 1.4 0.9

107

n M) SD 4.3 1.3 0.4

(mol/mol of protein) X SD 1.17 0.05 0.81 0.08 0.72 0.05

1All fluorescence measurements were at 332 nm and 27°C with excitation at 287 nm. 2Each solution contained 19.3 mM b-LG in 50 mM phosphate buffer at pH 7.0. Retinal was added at a concentration of 19.3 mM, and vitamin D2 was added at a concentration of 38.6 mM.

Figure 1. Fluorescence emission spectrum of b-LG in the absence ( ◊) and in the presence ( ♦) of palmitate. Measurements were at 27°C with 19.3 mM protein in 50 mM sodium phosphate buffer, pH 7.0, using an excitation wavelength of 287 nm, an excitation and emission band width of 10 nm, and a scan speed of 50 nm/min. Palmitate was added at a concentration of 19.3 mM. Inset: plot of the fluorescence data at 332 nm according to Equation [1] (see Materials and Methods). RT = Total ligand concentration, and a = the fraction of unoccupied ligand-binding sites on the protein.

bound. The data that were obtained in the presence of varying concentrations of ascorbic acid are included as a control because binding of this ligand with a resulting fluorescence change does not occur (21). Tests for Competitive Binding between Palmitate, Retinal, and Vitamin D2

Comparative Fluorescence Changes upon Binding of Palmitate, Retinal, and Vitamin D2

Solutions of b-LG were titrated with one ligand in the presence of saturating concentrations of a second ligand. Results from analyses of plots of data according to Equation [1] for titration of the protein with palmitate in the presence of retinal or vitamin D2 are listed in Table 1. Calculations based on the reported equilibrium dissociation constants for retinal and

The fluorescence emission spectrum of b-LG resulting from excitation at 287 nm in the presence or absence of palmitate is shown in Figure 1. Enhancement of the relative intensity around 332 and 295 nm occurred upon palmitate binding. These results for fluorescence enhancement and the spectrum obtained are similar to those previously reported (5, 22) but are clearly different from that observed upon binding of retinal, which quenches tryptophanyl fluorescence at 332 nm at both pH 5.1 and 7.0 (23). Furthermore, the binding of retinoids does not change the fluorescence below 300 nm (23). The apparent dissociation constant and moles bound per mole of protein were calculated from a plot of the data for enhancement of fluorescence at 332 nm upon palmitate binding according to Equation [1] (Figure 1, inset). The values obtained are listed in Table 1. A comparison of the tryptophanyl fluorescence enhancement resulting from palmitate binding is compared in Figure 2 to the quenching observed when retinal or vitamin D2 was

Figure 2. Relative fluorescence emission intensity at 332 nm for 19.3 mM b-LG solutions as a function of the protein and ligand molar ratios for palmitate ( ⁄) , ascorbate ( o) , vitamin D2 ( ÿ) , and retinal ( ◊) . For vitamin D2, the ratio scale is 0 to 2.0.

RESULTS

Journal of Dairy Science Vol. 82, No. 2, 1999

260

WANG ET AL.

vitamin D (23, 24) indicate that the protein should be saturated with these ligands at the concentrations used. The values for the parameters listed for palmitate binding in Table 1 are similar, regardless of the presence or absence of retinal even though the titrations were performed in the same concentration range. Values for these same parameters in the presence of vitamin D2 were slightly lower. The relative fluorescence intensities of the tryptophanyl residues in the presence of saturating palmitate as a function of retinal and vitamin D2 concentrations are shown in Figure 3. The dependence of the quenching of the intensity on retinal concentration is similar in the presence or absence of palmitate except for a positive displacement when palmitate was

bound. This result is consistent with the simultaneous binding of both molecules, the positive displacement being caused by the enhancement of intensity upon palmitate binding. Conversely, the dependence of the relative intensity on vitamin D2 concentration in the presence or absence of palmitate followed a somewhat different pattern. At low vitamin D2 concentrations, the intensity in the presence of palmitate was enhanced over that in its absence; however, at high concentrations of vitamin D2, the intensity was quenched even more than in solutions without palmitate. This result suggests that some of the palmitate was displaced by high concentrations of vitamin D2. Analysis of these data using Equation [1] gives the values listed in Table 2 for the apparent dissociation constant of the complex and the number of binding sites per molecule of protein. Values for these parameters for retinal binding are similar in the presence or absence of palmitate. However, in the presence of palmitate, the apparent dissociation constant for vitamin D2 is much larger, and the molar ratio of vitamin D2 bound is less. Palmitate binding to b-LG in the presence or absence of retinal or vitamin D2 was determined directly by measuring its concentration in ultrafiltrates of equilibrated solutions. The results given in Table 3 show that the presence of retinal did not significantly affect the amount of palmitate bound; however, when vitamin D2 was present, the amount of palmitate bound was decreased. Effect of pH on the Fluorescence Changes Data plotted in Figure 4 indicate that, in the presence of palmitate, the tryptophanyl fluorescence was enhanced throughout the range of pH from 2.79 to 7.12, suggesting that palmitate is bound in this entire pH range. The intensity enhancement appeared to reach a maximum near pH 5.0. Quenching of the fluorescence in the presence of retinal or vitamin D2 also was observed throughout this pH range. The quenching of fluorescence in the presence of vitamin D2 appeared to be a combination of the changes caused by palmitate and retinal binding. In fact, the sum of the effects of the latter two ligands falls on the same curve as that formed by vitamin D2 (Figure 4). Fluorescence Changes that Occur upon Binding of Retinyl Palmitate

Figure 3. Relative fluorescence intensities at 332 nm for 19.3 mM solutions of b-LG in sodium phosphate buffer, pH 7.0, without ( ◊) and with 19.3 mM palmitate added ( ♦) , as a function of all-transretinal concentration (upper) or vitamin D2 concentration (lower). Journal of Dairy Science Vol. 82, No. 2, 1999

The results indicate that retinal and palmitate both bind independently at different sites on b-LG; therefore, retinyl palmitate, the commonly used vita-

LIPOPHILIC NUTRIENT BINDING TO β-LACTOGLOBULIN

261

TABLE 2. Apparent dissociation constants ( Kapp d ) and the binding stoichiometry ( n ) for retinal and vitamin D2 bound to b-LG in the presence and absence of palmitate.1 Condition2

Ligand

Kapp d (×

Retinal Retinal3 Vitamin D2 Vitamin D24

b-LG + palmitate b-LG b-LG + palmitate b-LG

X 7.6 2.0 5.6 0.5

n 108

M) SD 2.3 1.9 3.5 0.3

(mol/mol of protein) X SD 1.02 0.07 1.00 0.16 1.67 0.17 2.00 0.16

1All

fluorescence measurements were at 332 nm and 27°C with excitation at 287 nm. solutions contained 19.3 mM b-LG in 50 mM phosphate buffer at pH 7.0. Palmitate was added at a concentration of 19.3 mM. 3Data were taken from Wang et al. (23). 4Data were taken from Wang et al. (24). 2All

min additive, potentially could bind at both sites. The fluorescence emission spectrum of b-LG in the presence of retinyl palmitate with excitation at 280 nm is shown in Figure 5. In comparison with the spectrum for b-LG alone, the spectrum with retinyl palmitate did not show much change above 300 nm. However, the apparent relative intensity at 284 nm, most likely resulting from resonance Raman scattering, was substantially increased. These results are similar to those obtained with palmitate (Figure 1). Nevertheless, the intensity enhancement appeared to be quantitatively related to ligand binding. Therefore, the data for emission at 284 nm from a titration with retinyl palmitate were analyzed according to Equation [1] (Figure 5, inset), and the binding parameters obtained are listed in Table 4. The binding of retinyl palmitate to b-LG can also be determined by measuring the fluorescence changes of the retinyl moiety that occurs when it binds in the calyx of the b-barrel (11). The emission spectrum for

the retinyl moiety obtained upon excitation at 342 nm is given in Figure 6. A broad maximum in the difference spectrum was observed between 440 and 480 nm. A plot of the data according to Equation [1] for the relative intensity change at 450 nm with excitation at 342 nm obtained by titration of b-LG with retinyl palmitate is shown in the inset of Figure 6. The values calculated for the binding parameters are listed in Table 4. DISCUSSION b-Lactoglobulin has two potential sites for binding lipophilic nutrients: one in the calyx formed by the b-

TABLE 3. Direct measurement of the competitive binding of palmitate, retinal, and vitamin D2 to b-LG. Ligand

Palmitate bound1

Palmitate2 Palmitate + retinal2 Palmitate + vitamin D22

( % of total added) X SD 86 3 82 2 55 8

Normalized to palmitate (Relative % ) 100 97 65

1Calculated from the decrease in palmitate concentration of an ultrafiltrate measured colorimetrically. 2The palmitate concentration in each solution was 0.2 mM. Retinal was added at a concentration of 0.2 mM, and vitamin D2 was added at a concentration of 0.4 mM. The concentration of b-LG was 0.2 mM.

Figure 4. Changes (normalized to pH 7.12) in the relative fluorescence intensities at 332 nm for 19.3 mM b-LG solutions in 50 mM sodium phosphate as a function of pH and resulting from the addition of 19.3 mM palmitate ( ◊) , 38.6 mM vitamin D2 ( ⁄) , or 19.3 mM all-trans-retinal ( ÿ) . The symbol ( o) represents the sum of the changes induced by palmitate and retinal. Journal of Dairy Science Vol. 82, No. 2, 1999

262

WANG ET AL. TABLE 4. Apparent dissociation constants ( Kapp d ) and binding stoichiometry ( n ) for binding of retinyl palmitate to b-LG. Fluorescence wavelength Ligand

Excitation

Emission

(nm) Retinol1 Retinyl palmitate Retinyl palmitate 1Data

287 342 280

332 450 284

Concentration of b-LG

n

Kapp d

X 0.99 0.80 1.15

( × 107 M) X SD 0.36 0.11 1.24 1.21 0.39 0.23

( mM) 1 10 10

SD 0.02 0.21 0.16

were taken from Wang et al. (23).

barrel and the other in an external hydrophobic pocket between the a-helix and the b-barrel (14, 17). The sites for binding of palmitate and retinoids on bLG have been the subject of several investigations, but the results have not been clear. It has been suggested that retinol binds in the external pocket (14, 20); however, most of the experimental evidence points to the calyx as the binding site (2, 3, 15, 16). Conflicting conclusions also have been drawn regarding the binding site for palmitate. Some investigators have suggested that these two ligands bind in the same site (5, 20), but others (6, 10, 15, 16, 25) have

reported evidence indicating different binding sites for retinol versus those for other lipophilic ligands. A comparison of the affinities for these two ligands indicates that retinoids are bound more tightly; the dissociation constant is the order of 10–8 M (3, 6, 11, 23), and the dissociation constant for palmitate is in the order of 10–6 to 10–7 M (Table 1 ) (10, 16, 21, 25). Furthermore, competition between palmitate and retinal for binding to b-LG did not occur. Titration of the protein with palmitate when retinal was already bound yielded similar values for the dissociation constant and stoichiometry of palmitate binding as was

Figure 5. Fluorescence emission spectrum of 10-mM solutions of b-LG in 50 mM sodium phosphate buffer, pH 7.0, measured in the absence ( ♦) and in the presence of 10 mM retinyl palmitate ( o) . The spectrum for retinyl palmitate is also shown ( ◊) . Measurements were at 27°C using an excitation wavelength of 280 nm, an excitation and emission bandwidth 10 nm, and a scan speed of 50 nm/min. Inset: plot of the fluorescence data at 284 nm according to Equation [1] (see Materials and Methods). RT = Total ligand concentration, and a = the fraction of unoccupied ligand-binding sites on the protein.

Figure 6. Fluorescence emission spectrum of a 10-mM solution of retinyl palmitate in 50 mM sodium phosphate buffer, pH 7.0, in the absence ( ◊) and in the presence of 10 mM b-LG ( ♦) . Measurements were at 27°C using an excitation wavelength of 342 nm, an excitation and emission bandwidth of 10 nm, and a scan speed of 50 nm/ min. Inset: plot of the fluorescence data at 450 nm according to Equation [1] (see Materials and Methods). RT = Total ligand concentration, and a = the fraction of unoccupied ligand-binding sites on the protein.

Journal of Dairy Science Vol. 82, No. 2, 1999

LIPOPHILIC NUTRIENT BINDING TO β-LACTOGLOBULIN

obtained without retinal present. Binding parameters for retinal also were similar with and without a saturating concentration of palmitate. Moreover, direct measurement for competitive binding indicated that the presence of saturating concentrations of retinal did not affect the amount of palmitate that was bound. It is informative to compare these results for retinal and palmitate with those for vitamin D2 and retinyl palmitate. Previously, we (23, 24) have shown that the stoichiometry for binding retinal is 1 and that for binding vitamin D2 is 2. The apparent dissociation constant for vitamin D2 was an order of magnitude less, and the stoichiometry was less, in the presence of saturating palmitate when the evaluation was performed in the same concentration range. The dependence of the relative fluorescence intensity on vitamin D2 concentration also indicated that the vitamin displaced palmitate at high concentrations. Furthermore, direct measurement of palmitate concentrations showed that significantly less palmitate was bound when a high concentration of vitamin D2 was added. Hence, vitamin D2 binds not only to the calyx site but also to the palmitate-binding site. Because retinyl palmitate contains both moieties, it potentially could occupy both the retinal- and palmitate-binding sites. The analysis of the tryptophanyl fluorescence spectrum at 332 nm is complicated by the compensatory quenching of intensity because of the retinyl moiety binding in the calyx and the enhancement of intensity caused by palmitate binding in the external pocket. However, the apparent intensity at 284 nm, probably caused by resonance Raman scattering, is greatly enhanced upon binding either palmitate or retinyl palmitate, a change that does not occur when retinoids are bound (23). Palmitate binding has been shown ( 2 5 ) to depend on protein concentration and binds more tightly to the dimer. Thus, the binding of palmitate or retinyl palmitate may be coupled with association of the protein. The enhancement of intensity at 284 nm is clearly quantitatively associated with binding of palmitate and, therefore, can be used to evaluate the binding parameters even though it does not represent true fluorescence. Analysis of the change associated with the binding of the palmitate moiety yields a stoichiometry of 1, and results of analysis of changes in fluorescence of the retinyl moiety also gives a stoichiometry near 1. Comparison of the relative affinities suggests that the retinyl moiety is bound less tightly than most retinoids but that the palmitate moiety is bound more tightly than palmitate alone. The added bulkiness of the palmitate moiety may affect the fit of the retinyl group in the calyx, and the

263

presence of the retinyl group may allow additional interaction with hydrophobic residues in the external pocket. These data suggest that retinyl palmitate binds in both sites, and, hence, b-LG is capable of binding 2 mol of retinyl palmitate/mol of protein. Thus, b-LG should be an excellent carrier of this nutrient additive. REFERENCES 1 Chen, J.-P., and H. Pai. 1991. Hydrolysis of milk fat with lipase in reversed micelles. J. Food Sci. 56:234–237. 2 Chen, S. X., C. C. Hardin, and H. E. Swaisgood. 1993. Purification and characterization of b-structural domains of blactoglobulin liberated by limited proteolysis. J. Protein Chem. 12:613–625. 3 Cho, Y., C. A. Batt, and L. Sawyer. 1994. Probing the retinolbinding site of bovine b-lactoglobulin. J. Biol. Chem. 269: 11102–11107. 4 Cogan, U., M. Kopelman, S. Mokady, and M. Shinitzky. 1976. Binding affinities of retinol and related compounds to retinol binding proteins. Eur. J. Biochem. 65:71–78. 5 Creamer, L. K. 1995. Effect of sodium dodecyl sulfate and palmitic acid on the equilibrium unfolding of bovine blactoglobulin. Biochemistry 34:7170–7176. 6 Dufour, E., C. Genot, and T. Haertle´. 1994. b-Lactoglobulin binding properties during its folding changes studied by fluorescence spectroscopy. Biochim. Biophys. Acta 1205:105–112. 7 Dufour, E., and T. Haertle´. 1991. Binding of retinoids and bcarotene to b-lactoglobulin. Influence of protein modifications. Biochim. Biophys. Acta 1079:316–320. 8 Dufour, E., M. C. Marden, and T. Haertle´. 1990. b-Lactoglobulin binds retinol and protoporphyrin IX at two different binding sites. FEBS Lett. 277:223–226. 9 Eftink, M. R. 1991. Fluorescence quenching reactions. Pages 1–9 in Biophysical and Biochemical Aspects of Fluorescence Spectroscopy. T. G. Dewey, ed. Plenum Press, New York, NY. 10 Frapin, D., E. Dufour, and T. Haertle´. 1993. Probing the fatty acid binding site of b-lactoglobulins. J. Protein Chem. 12: 443–449. 11 Fugate, R. D., and P. Song. 1980. Spectroscopic characterization of b-lactoglobulin-retinol complex. Biochim. Biophys. Acta 625: 28–42. 12 Godovac-Zimmermann, J. A. 1988. The structural motif of blactoglobulin and retinol-binding protein: a basic framework for binding and transport of small hydrophobic molecules? Trends Biochem. Sci. 13:64–66. 13 Lehrer, S. S. 1976. Perturbation of intrinsic protein fluorescence. Pages 515–539 in Biochemical Fluorescence: Concepts. R. F. Chen and H. Edelhoch, ed. Marcel Dekker, Inc., New York, NY. 14 Monaco, H. L., G. Zanotti, P. Spadon, M. Bolognesi, L. Sawyer, and E. Eliopoulos. 1987. Crystal structure of the trigonal form of bovine beta-lactoglobulin and of its complex with retinol at ˚ resolution. J. Mol. Biol. 197:695–706. 2.5 A 15 Narayan, M., and L. J. Berliner. 1997. Fatty acids and retinoids bind independently and simultaneously to b-lactoglobulin. Biochemistry 36:1906–1911. 16 Narayan, M., and L. J. Berliner. 1998. Mapping fatty acid binding to beta-lactoglobulin: ligand binding is restricted by modification of Cys 121. Protein Sci. 7:150–157. 17 Papiz, M. Z., L. Sawyer, E. E. Eliopoulos, A.C.T. North, J.B.C. Findlay, R. Sivaprasadarao, T. A. Jones, M. E. Newcomer, and P. J. Kraulis. 1986. The structure of b-lactoglobulin and its similarity to plasma retinol-binding protein. Nature (Lond.) 324:383–385. 18 Pe´rez, M. D., and M. Calvo. 1995. Interaction of b-lactoglobulin with retinol and fatty acids and its role as a possible biological function for this protein: a review. J. Dairy Sci. 78:978–988. Journal of Dairy Science Vol. 82, No. 2, 1999

264

WANG ET AL.

19 Pe´rez, M. D., C. Diaz de Villegas, L. Sanchez, P. Aranda, J. M. Ena, and M. Calvo. 1989. Interaction of fatty acids with blactoglobulin and albumin from ruminant milk. J. Biochem. 106:1094–1097. 20 Puyol, P., M. D. Pe´rez, J. M. Ena, and M. Calvo. 1991. Interaction of bovine b-lactoglobulin and other bovine and human whey proteins with retinol and fatty acids. Agric. Biol. Chem. 10: 2515–2520. 21 Puyol, P., M. D. Pe´rez, L. Mata, and M. Calvo. 1994. Study on interaction between b-lactoglobulin and other bovine whey proteins with ascorbic acid. Milchwissenschaft 49:25–27. 22 Spector, A. A., and J. E. Fletcher. 1970. Binding of long-chain fatty acids to b-lactoglobulin. Lipids 5:403–411.

Journal of Dairy Science Vol. 82, No. 2, 1999

23 Wang, Q., J. C. Allen, and H. E. Swaisgood. 1997. Binding of retinoids to b-lactoglobulin isolated by bioselective adsorption. J. Dairy Sci. 80:1047–1053. 24 Wang, Q., J. C. Allen, and H. E. Swaisgood. 1997. Binding of vitamin D and cholesterol to b-lactoglobulin. J. Dairy Sci. 80: 1054–1059. 25 Wang, Q., J. C. Allen, and H. E. Swaisgood. 1998. Protein concentration dependence of palmitate binding to blactoglobulin. J. Dairy Sci. 81:76–81. 26 Wang, Q., and H. E. Swaisgood. 1993. Characteristics of blactoglobulin binding to the all-trans-retinal moiety covalently immobilized on Celite. J. Dairy Sci. 76:1895–1901.