Adsorption and Exchange of β-Lactoglobulin onto Spread

Spread monolayer isotherms of β-lactoglobulin, monopalmitin, monolaurin, and their mixtures were measured at the air-water interface. The isotherms e...
0 downloads 0 Views 138KB Size
1216

Langmuir 2000, 16, 1216-1225

Adsorption and Exchange of β-Lactoglobulin onto Spread Monoglyceride Monolayers at the Air-Water Interface Michel Cornec and Ganesan Narsimhan* Biochemical and Food Process Engineering, Department of Agricultural and Biological Engineering, Purdue University, West Lafayette, Indiana 47907-1146 Received March 19, 1999. In Final Form: September 2, 1999 Spread monolayer isotherms of β-lactoglobulin, monopalmitin, monolaurin, and their mixtures were measured at the air-water interface. The isotherms exhibited hysteresis behavior. A phase transition from a liquid expanded (LE) to liquid condensed (LC) was observed for monolayers containing monopalmitin. At highest surface pressures, isotherms of mixed β-lactoglobulin-monoglycerides coincided with those of pure monoglycerides, which indicated that the protein was squeezed out from the interface. Monopalmitin monolayer exhibited higher values of surface pressure as well as dilatational elasticity as compared to monolaurin and β-lactoglobulin. The dynamics of surface pressure (π) and of surface concentration (Γ) of 14C radiolabeled β-lactoglobulin adsorbed onto spread monolayers of monoglyceride (monopalmitin and monolaurin) at the air-water interface was measured. The adsorption of β-lactoglobulin onto loosely packed spread monoglyceride monolayers was enhanced at short times with a corresponding increase in the surface pressure, indicating synergism due to possible dissolution of the protein molecules into loosely or moderately packed monoglyceride layers. The rate of adsorption of [14C]-β-lactoglobulin was enhanced onto a spread monolayer of monopalmitin of 19 and 27 Å2/molecule at short times. However, the amounts of protein adsorbed after 10 h were lower with the values being 0.8 and 1.2 mg/m2, respectively. Spreading of a monopalmitin monolayer of close-packed area after 2.5 h onto the air-water interface with adsorbed β-lactoglobulin led to complete displacement of the protein from the interface, possibly because of the surface pressure and steric exclusion effects. On the other hand, the rate of adsorption of [14C]-β-lactoglobulin was lowered by the presence of a monolaurin layer at the interface, but the steady-state surface concentration of the protein was not changed. Spreading of monolaurin monolayer of close-packed area after 2.5 h onto the air-water interface with adsorbed β-lactoglobulin resulted in an initial displacement of the protein from the interface. However, this resulted in an initial jump in the surface pressure immediately followed by its relaxation to much smaller values possibly due to the rearrangement of the monolayer. This resulted in slow readsorption of β-lactoglobulin at longer times.

Introduction Whey proteins and surfactants, such as monoglycerides, are widely used in the food industry as emulsifiers and stabilizers in foams and emulsions. Both molecular species, because of their amphiphilicity, absorb at liquid interfaces. The resulting decrease in the interfacial tension reduces the amount of energy needed to disrupt the oil droplets and promotes emulsification.1,2 In addition, proteins and surfactants stabilize emulsions against coalescence by forming a protective coating around the droplets. Proteins form a condensed viscoelastic film of highly self-interacting molecules at the interface which resists local deformation.3,4 In contrast, surfactants form a fluid adsorbed layer in which adsorbed molecules can diffuse laterally toward regions of high surface tension conferring stability via the Marangoni effect.5 Individually, the viscoelastic and Marangoni mechanisms are very effective at stabilizing foams and emulsions but are mutually incompatible. Low molecular weight surfactants, because of their much lower molecular weight, pack more efficiently at the interface and thus reduce the interfacial tension to a greater extent * To whom all correspondence should be addressed. Telephone, (765)-494-1199; Fax, (765)-496-1115; e-mail, narsimha@ ecn.purdue.edu. (1) Halling, P. J. CRC Crit. Rev. Food Sci. Nutrition 1981, 15, 155. (2) Dickinson, E.; Stainsby, G. In Colloids in Food; Applied Science Publishers: London, 1982. (3) Clark, D. C.; Mackie, A. R.; Wilde, P. J.; Wilson, D. R. Faraday Discuss. 1994, 98, 253. (4) Clark, D. C. In Characterization of Food: Emerging Methods; Gaonkar, A. G., Ed.; Elsevier: Amsterdam, 1982; p 23. (5) Ewers, W. E.; Sutherland, K. L. Aust. J. Sci. Res. Ser. A 1952, 5, 697.

than proteins.6 On the other hand, intermolecular interactions between molecules of low molecular weight surfactant are much weaker and development of high mechanical strength is not possible.7 For these reasons, both classes of molecules are used in food systems. This may cause a problem as competition between the two mechanisms arises, leading to instability of the system. The mechanism of the destabilization is related to the displacement of the protein from the interface. Desorption of protein from the interface occurs by two extreme mechanisms:8 (i) the solubilization mechanism, where the protein is solubilized in the aqueous phase by formation of a complex between protein and a water-soluble surfactant, and (ii) the replacement mechanism, where displacement of adsorbed protein arises when interaction between surface and surfactant is stronger than the interaction between the surface and protein. A great amount of work has dealt with competitive adsorption between protein and surfactant at both air-water9-13 and (6) McClements, D. J. In Food Emulsions: Principles, Practice and Techniques; CRC Press: Boca Raton, FL, 1998. (7) Dickinson, E. In An Introduction to Food Colloids; Oxford University Press: Oxford, U.K., 1992. (8) Dickinson, E. J. Chem. Soc., Faraday Trans. 1998, 94, 1657. (9) Coke, M.; Wilde, P. J.; Russel, E. J.; Clark, D. C. J. Colloid Interface Sci. 1990, 138, 489. (10) Clark, D. C.; Wilde, P. J.; Wilson, D. R. Colloids Surf. 1991, 59, 209. (11) Wilde, P. J.; Clark, D. C. J. Colloid Interface Sci. 1993, 155, 48. (12) Sarker, D. K., Wilde, P. J.; Clark, D. C. J. Agric. Food Chem. 1995, 43, 295. (13) Sarker, D. K.; Wilde, P. J.; Clark, D. C. Colloids Surf., B 1995, 3, 349.

10.1021/la990326p CCC: $19.00 © 2000 American Chemical Society Published on Web 11/24/1999

Adsorption and Exchange of β-Lactoglobulin

oil-water14-34 interfaces. It has been shown that, at high enough surfactant to protein molar ratio, milk proteins are completely displaced by nonionic water-soluble surfactants.15,17 On the other hand, oil-soluble surfactants are less effective in displacing the protein from the interface.27,33,34 Depending on the surfactant to protein molar ratio, non-ionic lipophilic surfactants (such as monoglyceride) have been observed to have a positive or negative effect on protein surface coverage,24-26 whereas ionic surfactants, which form a strong interfacial complex with the adsorbed protein, are much less effective in displacing the protein.26 It is therefore important to gain information on the molecular interactions between proteins and surfactants present at the interface, to tailor the structure of these systems as well as their composition in order to increase emulsion stability. Despite a considerable amount of experimental investigations using various techniques,35-45 little is known in a quantitative sense on the molecular mechanisms of protein-lipid interactions at the interface. Electrostatic interactions between protein molecules and lipid monolayers have been shown to be important at the liquid-air interface.40,45-51 In addition, hydrophobic interactions have also been shown to play an important role,52,53 the nature of this interaction being dependent on (14) Courthaudon, J. L.; Dickinson, E.; Christie, W. W. J. Agric. Food Chem. 1991, 39, 1365. (15) Courthaudon, J. L.; Dickinson, E.; Dalgleish, D. G. J. Colloid Interface Sci. 1991, 145, 390. (16) Courthaudon, J. L.; Dickinson, E.; Matsumara, Y.; Clark, D. C. Colloids Surf. 1991, 56, 293. (17) Courthaudon, J. L.; Dickinson, E.; Matsumara, Y.; Williams, A. Food Structure 1991, 10, 109. (18) Dickinson, E.; Tanai, S. J. Agric. Food Chem. 1992, 40, 179. (19) Chen, J.; Dickinson, E.; Iveson, G. Food Structure 1993, 12, 135. (20) Chen, J.; Dickinson, E. J. Sci. Food Agric. 1993, 62, 283. (21) Chen, J.; Dickinson, E. Food Hydrocolloids 1995, 9, 35. (22) Chen, J.; Dickinson, E. Colloids Surf., A 1995, 100, 255. (23) Chen, J.; Dickinson, E. Colloids Surf., A 1995, 100, 267. (24) Chen, J.; Dickinson, E. Colloids Surf., A 1995, 101, 77. (25) Dickinson, E.; Iveson, G. Food Hydrocolloids 1993, 6, 533. (26) Dickinson, E.; Hong, S. T. J. Agric. Food Chem. 1994, 42, 1602. (27) Dickinson, E.; Owusu, R. K.; Tan, S.; Williams, A. J. Food Sci. 1993, 58, 295. (28) Fang, Y.; Dalgleish, D. G. Colloids Surf., B. 1993, 1, 357. (29) Mackie, A. R.; Wilde, P. J.; Wilson, D. R.; Clark, D. C. J. Chem. Soc., Faraday Trans. 1993, 89, 2755. (30) Mackie, A. R.; Nativel, S.; Wilson, D. R.; Ladha, S.; Clark, D. C. J. Sci. Food Agric. 1996, 70, 413. (31) Fang, Y.; Dalgleish, D. G. J. Agric. Food Chem. 1996, 44, 59. (32) Fang, Y.; Dalgleish, D. G. J. Am. Oil Chem. Soc. 1996, 73, 437. (33) Cornec, M.; Mackie, A. R.; Wilde, P. J.; Clark, D. C. Colloids Surf., A 1996, 114, 237. (34) Cornec, M.; Wilde, P. J.; Gunning, P. A.; Mackie, A. R.; Husband, F.; Parker, M. L.; Clark, D. C. J. Food Sci. 1998, 63, 39. (35) Bos, M.; Nylander, T.; Arnebrant, T.; Clark, D. C. In Food emulsifiers and their applications; Hasenhuettl, G. L., Hartel, W., Eds.; Chapman & Hall: New York, 1997; pp 95-146. (36) Nylander, T.; Erickson, B. In Food Emulsions; Friberg, S. E., Larsson, K., Eds.; Marcel Dekker Inc.: New York, 1997; pp 189-234. (37) Nylander, T. In Proteins at liquid interfaces; Mo¨bius, D., Miller, R., Eds.; Elsevier: Amsterdam, 1998; pp 385-431. (38) Li, J. B.; Zhao, J.; Wu, J.; Miller, R. Nahrung 1998, 42, 232. (39) Li, J. B.; Zhao, J.; Miller, R. Nahrung 1998, 42, 234. (40) Cornell, D. G.; Patterson, D. L. J. Agric. Food Chem. 1989, 37, 1455. (41) Kristensen, D.; Nylander, T.; Rasmussen, J. T.; Paulsson, M.; Carlsson, A. Int Dairy J.l 1997, 7, 87. (42) Rodriguez Nino, M. R.; Wilde, P. J.; Clark, D. C.; Rodriguez Patino, J. M. Langmuir 1998, 14, 2160. (43) Heckl, W. M.; Zaba, B. N.; Mo¨hwald, H. Biochim. Biophys. Acta 1987, 903, 166. (44) Mackie, A. R.; Gunning A. P.; Wilde, P. J.; Morris, V. J. J. Colloid Interface Sci. 1999, 210, 157. (45) Kozorac, Z.; Dhathathregan, A.; Mo¨bius, D. FEBS 1988, 229, 372. (46) Aynie´, S.; Le Meste, M.; Colas, B.; Lorient, D. J. Food Sci. 1992, 57, 883. (47) Quinn, P. J.; Dawson, R. M. C. Biochem J. 1969, 113, 791. (48) Quinn, P. J.; Dawson, R. M. C. Biochem J. 1969, 115, 65.

Langmuir, Vol. 16, No. 3, 2000 1217

the aliphatic chain length,53 size of the headgroup,53 physical state of the lipid phase,54 and packing of the lipid54,55 and protein structure.56 In a recent study, Cho et al.57 investigated the exchange of radiolabeled bovine serum albumin (BSA) with spread phospholipid monolayers at the air-water interface. Adsorption of BSA was found to be enhanced at short times, indicating synergism due to possible dissolution of BSA into losely or moderately packed monolayers. BSA, however, was expelled from the interface by a close-packed lecithin monolayer. In summary, most studies on mixed lipid-protein monolayers are focused on interactions between proteins and polar phospholipids. Less attention has been given to nonpoplar monoglycerides,42,58,59 even though they are used abundantly as emulsifiers in the food industry.60,61 In addition, in most of previous studies of mixed proteinlipid monolayers, adsorption of protein into the lipid monolayer has been monitored by measuring the change in the surface pressure at constant surface area (or the changes in surface area at constant surface pressure). Only a few studies have attempted to probe the changes in surface concentration of the protein.46,47,54,57 In the present study, the interaction between radiolabeled β-lactoglobulin and monoglyceride of different chain lengths at air-water interface has been investigated. Spread monolayer isotherms of the mixture of different compositions have been measured. The exchange of [14C]-βlactoglobulin with spread monolayer of monolaurin (C12: 0) as well as monopalmitin (C16:0) at the air-water interface has been measured by using a radiotracer method. Experimental Section Materials and Apparatus. β-Lactoglobulin (Prod No. L0130, lot number 114H7055), 1 monolauroyl-rac-glycerol (C12:0), and 1-monopalmitoyl-rac-glycerol (C16:0) were purchased from Sigma Chemical Inc. (St. Louis, MO). Reagent grade n-hexane and ethanol were purchased from Aldrich Chemical. Monoglyceride solutions were made in hexane-ethanol (v/v ) 9:1) and were used within 2 days. Isotopes of [14C]formaldehyde (37.3%) and [14C]sodium stearate were purchased from Sigma Chemical Inc. Sodium cyanoborohydride (NaCNBH3) was purchased from Aldrich Chemical Inc. All the experiments were carried out at pH 7.4 using 10 mM commercial sodium phosphate buffer containing 0.9% NaCl. Thoughout the experiments, ultrapure deionized water was used. A Langmuir minitrough (with dimensions of 330 × 75 × 6.5 mm) from KSV (Helsinki, Finland) was used for both surface pressure and surface concentration measurements. A gas proportional detector (Ludlum model 120, with a 2 × 2 in. Mylar window) with a digital scaler/counter (Ludlum model 520) was used for detecting radioactivity, in counts per minute (CPM), from the adsorbed monolayer at the air-water interface. Radioactivity was measured under P-10 gas (10% methane in argon) flowing through the detector chamber. (49) Cornell, D. G.; Patterson, D. L.; Hoban, N. J. Colloid Interface Sci. 1990, 140, 428. (50) Brewkink, E.; Demel, R. A.; De Korte-Kool, G.; De Kruijff, B. Biochemistry 1992, 31, 1119. (51) Grimard, R.; Tancrede, P.; Gicquaud, C. Biochem. Biophys. Res. Commun. 1993, 190, 1017. (52) Bos, M. A.; Nylander, T. Langmuir 1996, 12, 2791. (53) Du, Y.; An, J.; Tang, J.; Jiang, L. Colloids Surf., B 1996, 7, 129. (54) Ibdah, J. A.; Phillips, M. C. Biochemistry 1988, 27, 7155. (55) Fidelio, G. D.; Maggio, B.; Cumar, F. A. Biochim. Biophys. Acta 1986, 862, 49. (56) Hanssens, I.; Van Cauwelaert, F. H. Biochim. Biophys. Res. Commun. 1978, 84, 1088. (57) Cho, D.; Narsimhan, G.; Franses, E. I. Langmuir 1997, 13, 4710. (58) Rodriguez Nino, M. R.; Rodriguez Patino, J. M. J. Am. Oil Chem. Soc. 1998, 75, 1233. (59) Rahman, A.; Sherman, P. Colloid Polym. Sci. 1982, 260, 1035. (60) Boyle, E.; German, J. B. CRC Food Sci. Nutr. 1996, 36, 785. (61) Henry, C. Cereal Foods World 1995, 40, 734.

1218

Langmuir, Vol. 16, No. 3, 2000

Methods. Radiolabeling of Proteins. β-Lactoglobulin (30 mg) was dissolved in 0.05 M phosphate buffer (pH 7) and mixed with 0.1 M sodium cyanoborohydride and [14C]formaldehyde (102 µCi) and allowed to react for 2 h at room temperature.62 After the reaction, the mixture was dialyzed against a 0.05 M phosphate buffer for 30 h at 4 °C for complete removal of unreacted species. β-Lactoglobulin was found to have 1 amide group labeled per molecule on the average (2.52 µCi/mg of protein) as analyzed with a scintillation counter (Tri-carb 4000, from Packard instruments). Protein concentration was determined by using the BCA assay.63 It is to be noted that the degree of modification due to radiolabeling is small and therefore does not significantly affect the surface properties of the protein. Moreover, comparison of the spread monolayer isotherm of native and radiolabeled bovine serum albumin indicated no significant differences in the surface activity between the two because of radiolabeling.57 In addition, it was shown that radiolabeling did not affect the surface activity of different proteins by comparing the adsorption data at air-water interfaces using radiotracer and ellipsometry.64 Surface Pressure-Molecular Area (π-A) Isotherms. The Langmuir trough was first filled with phosphate buffer. Then, the surface was cleaned by sweeping it with the Teflon barrier, and any surface-active contaminants were removed by suction (aspiration) of the interface. The monoglyceride was spread over the clean-air-water interface by applying the solution dropwise from a Hamilton syringe. For preparing spread monolayers of protein, the Trurnit’s monolayer spreading method65 was used. Aliquots of 50 mL of a 0.0247 wt % protein solution were dripped from the top of a glass rod (5-mm diameter and 5 cm long) positioned across the air-water interface. The solution was spread uniformly on the interface. As detailed previously, there was negligible loss of protein to the bulk due to desorption.66 Consequently, it is reasonable to assume that all the proteins spread are adsorbed on the surface. For mixed monolayers, first a β-lactoglobulin monolayer was formed with Trurnit’s method and allowed to rest for 10 min. Then, various aliquots of monoglyceride solution in hexaneethanol were spread at several spots on the surface. Another 10 min was allowed for the monolayers to equilibrate. Surface areas were then reduced from 247.5 to 30 cm2 with the minitrough Teflon barrier at 4 cm2/min, with the surface pressure continuously recorded. Immediately after the end of the compression stage, the area was expanded at the same rate. Adsorption from Solution. The adsorption of [14C]-β-lactoglobulin to spread lipid monolayers was studied as described below. The trough was filled with the buffer solution without any protein (surface tension ) γ0) and the surface was carefully aspirated to remove surface impurities before the surface pressure was adjusted to zero (π ) γ0 - γ). Then, the protein solution was gently poured into the trough. After that a lipid monolayer was spread using a Hamilton syringe and the surface pressure and surface concentration (via radioactivity measurements) were monitored up to 10 h. In another set of experiments, monoglyceride was spread on top of adsorbed protein monolayers prepared by allowing the β-lactoglobulin to adsorb for 150 min. Calibration of Ludlum Gas Proportional Detector. To convert CPMs to surface concentrations, the Ludlum gas proportional detector was calibrated with radioactive samples of known surface or bulk concentrations. The bulk radioactivity calibration procedure of Hunter et al.62 was employed for background correction. [14C]H3COONa was employed for the calibration of the bulk radioactivity. Different amounts (25, 45, 80, and 120 µL) of 50 µCi/mL of sodium acetate solutions in ethanol were added to 175 mL of phosphate buffer to give solutions of bulk radioactivities of 7.14 × 103, 1.285 × 104, 2.285 × 104, and 3.4 × 104 µCi/m3 respectively. (62) Hunter, J. R.; Kilpatrick, P. K.; Carbonell, R. G. J. Colloid Interface Sci. 1991, 142, 429. (63) Smith, P. K.; Krohn, R. H.; Hermanson, G. T.; Mallia, A. K.; Gartner, F. H.; Provenzano, M. D.; Fujimoto, E. K.; Goeke, N. M.; Olson, B. J.; Klenk, D. C. Anal. Biochem. 1985, 150, 76. (64) Graham, D. E.; Phillips, M. C J. Colloid Interface Sci. 1979, 70, 403. (65) Trurnit, H. J. J. Colloid Sci. 1960, 15, 1. (66) Cho, D.; Franses, E. I.; Narsimhan, G. Colloids Surf., A 1996, 117, 45.

Cornec and Narsimhan

Figure 1. π-A h (area per molecule) isotherm of (a) β-lactoglobulin and (b) monoglycerides: curve 1, monopalmitin (C16: 0); curve 2, monolaurin (C12:0). [14C]-β-Lactoglobulin, rather than [14C]stearic acid, was used for calibration of the surface radioactivity, since by using [14C]stearic acid one tends to underestimate the surface concentration because of its much smaller molecular size compared to β-lactoglobulin.57,67 A total of 175 mL of 0.01 M phosphate buffer (pH 7, containing 0.9% NaCl) was poured into the Langmuir minitrough and a spread monolayer of [14C] protein was formed by using Trurnit’s method. The detector was placed at a distance of 3 mm above the air-water interface and the steady-state CPM was measured. The area of the air-water interface was compressed in stages to provide several surface radioactivities, which were used, for the detector calibration.

Results Surface Isotherms. π-A isotherms of β-lactoglobulin, monolaurin (C12:0) and monopalmitin (C16:0) are shown in Figure 1. The β-lactoglobulin monolayers were gaslike at area greater than 4,000 Å2/molecule, beyond which an apparent phase transition was noted to liquidlike with much lower monolayer compressibility (Figure 1a). Another pronounced transition, probably to a more condensed monolayer phase, seems to occur for A h lower than 1400 Å2/molecule (π ≈ 22 mN/m). Upon compression, molecules of β-lactoglobulin are forced to adopt a closepack form with only solvated polar amino acids submerged in the subphase. At higher surface pressure (>22 mN/m), whole protein molecules or segments of the protein are probably squeezed out into the subphase. However, the fact that the compressed monolayer was able to recover to its initial state upon decompression and that no shift (67) Xu, S.; Damodaran, S. J. Colloid Interface Sci. 1993, 157, 485.

Adsorption and Exchange of β-Lactoglobulin

in the molecular area did occur upon subsequent compression ruled out massive loss of polypeptide molecule to the subphase. However, the observed hysteresis suggests that the recovery is slow (Figure 1a). Monolaurin monolayer exhibited gaslike behavior for A between 55 and 90 Å2/molecule, liquidlike for smaller areas without any monolayer collapse up to 11 Å2/molecule (Figure 1b). Hysteresis and a shift of successive full compression isotherms toward lower molecular area were observed. In addition, π was lower after subsequent compression. The isotherm of the monopalmitin monolayer showed a gradual increase in the surface pressure until a plateau occurred at molecular area between 40 and 60 Å2/molecule (Figure 1b). The plateau reveals a phase transition between a liquid expanded (LE) state and a liquid condensed (LC) state. Gehlert et al.68 using Brewster Angle Microscopy (BAM) observed in the region corresponding to the beginning of the plateau some condensed phase domains surrounded by a homogeneous fluid phase of low density. As the molecular area is decreased, the domains grew in area at the expense of the fluid phase. Upon further compression, the domains started to contact and deform each others, filling the gaps in the condensed phase, which was accompanied by an increase in surface pressure. Above a surface pressure of 25 mN/m, the domains were compressed to close-packing without visible gap. At a surface pressure of 55 mN/m, corresponding to an area of 25 Å2/molecule, a monolayer collapse was observed. Hysteresis was observed for both monoglyceride monolayers with A h being smaller at a certain π during the expansion than the compression cycle. Mixed monolayer isotherms are shown in Figure 2a for weight ratios monopalmitin to β-lactoglobulin (R C16/β-lg) from 3 to 0.33. A LE-LC transition is visible on all isotherms at about 7-8 mN/m. The shift of the transition to higher areas when the proportion of β-lactoglobulin is increased point out the presence of the two constituents at the interface. Nevertheless, the conservation of the transition LE-LC is indicative of the great influence of the monoglyceride properties during compression.69,70 At higher surface pressures, the mixed isotherms exhibited another transition, which was also present in the isotherm of the pure β-lactoglobulin. Monopalmitin-rich monolayers exhibited smoother phase transitions than β-lactoglobulinrich monolayer. At the lowest RC16/β-lg ratio, the transition was reduced to a kink, which may be due to the expulsion of the protein molecules from the mixed film leading to the reduction of the film compressibility to a level similar to that of pure monoglyceride. As the film was further compressed, the isotherms of the mixed monolayer asymptotically approached that of the monoglyceride monolayer, and at high enough surface pressure the isotherm coincided, thereby indicating that the β-lactoglobulin molecule was squeezed out from the monopalmitin monolayer at high surface pressures. However, the collapse surface pressure was observed to be lower for mixed monolayers than for pure monoglycerides. The compression curve of the mixed layers formed with β-lactoglobulin and monolaurin were found to look similar to that of the β-lactoglobulin monolayer (Figure 2b). Unlike that of a monopalmitin-β-lactoglobulin mixture of different compositions, the isotherm did not exhibit a LELC phase transition. In fact, such a phase transition was (68) Gehlert, U.; Weidemann, G.; Vollhardt D. J. Colloid Interface Sci. 1995, 174, 392. (69) Boury, F.; Ivanova, I.; Panaiotov, I.; Proust, J. E. Langmuir 1995, 11, 599. (70) Boury, F.; Ivanova, I.; Panaiotov, I.; Proust, J. E. Langmuir 1995, 11, 2131.

Langmuir, Vol. 16, No. 3, 2000 1219

Figure 2. π-A h isotherms of mixed monolayers for varied mass ratio of monoglyceride to β-lactoglobulin: curve 1, RMG/βlg ) 3; curve 2, RMG/βlg ) 2; curve 3, RMG/βlg ) 1; curve 4, RMG/βlg ) 0.5; and curve 5, RMG/βlg ) 0.33, where MG is (a) monopalmitin and (b) monolaurin. The average molecular area of the mixture, A h ) A/(nL + nβlg), where A is the total area and nL and nβlg are the total number of molecules of 1-monopalmitoyl-rac-glycerol and β-lactoglobulin, respectively.

not observed even for pure monolaurin. A phase transition, however, was observed at higher surface pressures possibly due to expulsion of some of the adsorbed protein segments. The specific area at which this transition occurred depended on the composition and the condensed phase was observed for a wider range of molecular areas for β-lactoglobulin-rich monolayers (low RC12/βlg). Whereas monolaurin-rich monolayers exhibited smoother phase transitions, β-lactoglobulin-rich monolayers exhibited sharper transitions with smaller slopes. The surface pressure reached at maximum compression was closer to that of the monolaurin monolayer than that of the protein monolayer. This further indicated that the monoglyceride seems to be the main contributor to the surface pressure. This was more pronounced at the close-packed limit, at which the β-lactoglobulin molecules might have been expelled from the surface layer. For an insoluble monolayer, the dilatational elastic modulus may be derived from the π-A isotherm:

|E| )

dπ d ln A h

The dilatational elasticity modulus for β-lactoglobulin, monolaurin, and monopalmitin monolayers at different surface pressures is shown in Figure 3. The values of |E| are higher for the monopalmitin than for β-lactoglobulin and the monolaurin, that is the interactions between the

1220

Langmuir, Vol. 16, No. 3, 2000

Cornec and Narsimhan

Figure 3. Elasticity modulus versus surface pressure of β-lactoglobulin (0), monopalmitin (\) and monolaurin (2).

monopalmitin molecules are stronger than that for the monolaurin or β-lactoglobulin molecules at the interface. For β-lactoglobulin, |E| exhibited a maximum at π ) 18 mN/m. Upon further compression (higher π), structural changes and/or desorption of polypeptides resulted in a decrease in |E|. For monopalmitin films, the sudden drop in |E| observed at lower surface pressure (π ) 8 mN/m) was associated with the LE to LC transition. Then, |E| increased rapidly until a plateau was reached at saturation. At highest π (π > 50 mN/m), the film collapsed resulting in a sharp decrease in the modulus. In case of the monolaurin film, there was no drop in |E| at lower surface pressures. However, a plateau in |E| was observed at π ) 12 mN/m. Also, a decrease in the modulus was observed at higher surface pressures (π > 25 mN/m), indicating a possible unstability of the film. Variations of |E| with π for mixed β-lactoglobulinmonopalmitin and β-lactoglobulin-monolaurin monolayers are shown in parts a and b of Figure 4, respectively. For β-lactoglobulin-monolaurin mixed films, |E| did not vary significantly with monolayer composition for π up to 27 mN/m (Figure 4b). However, at higher π, |E| was observed to increase up to a value close to the elasticity developed by a monolayer of monolaurin. In case of a mixed β-lactoglobulin-monopalmitin film, |E| exhibited two minima, one at low π (π ≈ 8 mN/m) corresponding to the LE to LC transition and one at higher π (π ≈ 27 mN/m) (Figure 4a). At higher π (π > 30 mN/m), i.e., when the protein molecule was squeezed out of the surface, |E| was observed to increased rapidly up to a maximum, this maximum value increasing as the proportion of the monoglyceride in the monolayer increased. At surface pressures greater than 45 mN/m, a sudden drop in |E| was observed due to monolayer collapse. Absorption and Exchange for Protein-Lipid Surfactant Mixtures. Interactions of proteins and spread lipid monolayers were investigated by spreading a lipid monolayer on an adsorbing protein solution. First, control experiments were performed in order to validate the methodology used in this study. Spreading of solvent only (hexane-ethanol mixture without any lipid) on an absorbing protein solution resulted in a small overshoot in π but no significant change in Γ was noted. It was concluded that the spreading solvent did not significantly affect the protein adsorption. β-Lactoglobulin-Monopalmitin Mixtures. Spreading a small amount (12.35 µg or 110 Å2/molecule) of mono-

Figure 4. Elasticity modulus versus surface pressure of mixed monolayer for varied mass ratio of (a) monopalmitin to β-lactoglobulin and (b) monolaurin to β-lactoglobulin: (9) RMG/βlg ) 3, (4) RMG/βlg ) 2, (O) RMG/βlg ) 1, (2) RMG/βlg ) 0.5, and (0) RMG/βlg ) 0.33.

palmitin on a nearly clean interface was observed to slightly enhance the adsorption of β-lactoglobulin. The steady surface concentration reached after 10 h of adsorption was 1.82 mg/m2 compared to 1.63 mg/m2 when no monoglyceride was at the interface (Figure 5). In addition, the presence of the monoglyceride at the interface was observed to increase the initial rate of adsorption. dΓ/dt, as estimated from the initial slope of the curve Γ-t1/2 (insert Figure 5), was 1.22 × 10-9 m2 s-1 or about 4 times faster than that when β-lactoglobulin only was present. When the amount of monopalmitin spread was increased to 49.4 µg (27 Å2/molecule), π jumped to 14.5 mN/m within seconds. Then π increased gradually until it reached a plateau at 22 mN/m (Figure 6a). The surface concentration of [14C]-β-lactoglobulin also increased rapidly to 1.2 mg/ m2. The initial rate of adsorption was found to be 21 times faster (D ) 6.81 × 10-9 m2 s-1) than that for the protein alone. However, the steady-state surface concentration was lowered to 1.2 mg/m2. The enhancement of the β-lactoglobulin adsorption was not due to any interaction between the solvent and the protein, since spreading solvent only did not significantly affect the β-lactoglobulin adsorption. So, it was inferred that the monopalmitin enhances the adsorption of β-lactoglobulin at short times. The higher the amount of monopalmitin spread was, the faster the rate of adsorption and the lower the amount of protein adsorbed at the interface (Figure 6b). If a close pack monolayer of monopalmitin (113 µg or 12.01 Å2/ molecule) was present at the surface, the adsorption of β-lactoglobulin was totally impeded (Figure 6c).

Adsorption and Exchange of β-Lactoglobulin

Langmuir, Vol. 16, No. 3, 2000 1221

Figure 5. Comparison of dynamic of adsorption densities of [14C]-β-lactoglobulin (10-4 wt %) from (O) protein solution or (9) from protein solution in the presence of a spread monolayer of monoplalmitin (12.35 µg or 110 Å2/molecule). In insert is shown the initial rate of adsorption. The lines indicated the regression used to calculate the diffusion coefficient. Table 1. Effect of the Amount of Spread Monolaurin and Monopalmitin on the Initial Rate of Adsorption of β-Lactoglobulin at the Air-Water Interface amt of monoglyceride area per molecule spread, µg Å/molecule

3.16 × 10-10

0 monopalmitin

monolaurin

diffusion coefficient, m2/s

12.35 49.05 74.1

110 27 18.31

1.22 × 10-9 6.81 × 10-9 8.01 × 10-9

12.35 37.05 49.05 147

91 30.4 22.97 7.66

1.91 × 10-10 1.46 × 10-10 4.0 × 10-11 3.02 × 10-13

To test how an adsorbed monolayer of β-lactoglobulin would respond to spreading of a monopalmitin monolayer, the spreading was delayed by 150 min, when the surface monolayer was nearly at steady state. β-Lactoglobulin was enhanced from ≈1.66 to 2.15 mg/m2 by spreading of monopalmitin (24.7 µg or 54.96 Å2/molecule) to a value higher than the steady-state surface concentration of β-lactoglobulin, when no monoglyceride was present at the interface (Figure 7a). On the other hand, π surged to 21 mN/m and relaxed to a final value of 19 mN/m. Using 74.1 µg of monopalmitin rather than 24.7 decreased the monoglyceride molecule area from 54.96 to 18.31 Å2/ molecule and increased π from 9 to 30 mN/m (Figure 7b). As observed earlier, β-lactoglobulin adsorption was first enhanced but subsequently some protein molecules were apparently desorbed and Γ remained at about 1.4 mg/m2 at longer times. Spreading a close-packed monolayer of

monopalmitin resulted in a complete desorption of the β-lactoglobulin molecules from the interface (Figure 7c). β-Lactoglobulin-Monolaurin Mixtures. Spreading a monolaurin monolayer, instead of a monopalmitin, led to different results (Figures 8 and 9 and Table 1). Spreading a small amount of monolaurin (12.35 µg or 91.25 Å2/ molecule) did not change the steady-state surface concentration of protein adsorbed (see Figures 5 and 8a), but the initial rate of adsorption was lowered (Table 1). Increasing the amount of monolaurin spread (Figure 8b) resulted in a slower rate of protein adsorption (Table 1) but did not change the amount of protein adsorbed significantly. When a close-packed monolaurin was spread initially, however, protein did not adsorb at small times (Figure 8c). After an induction period, slow adsorption of protein was observed. Spreading the monoglyceride monolayer induced an overshoot in the surface pressure at very small times followed by a decrease at larger times. Except for close-packed monolayer, π was found to increase again. For example, when 37.05 µg of monolaurin were spread, π surged to 6 mN/m, following by a rapid decrease to 4.5 mN/m. Then, π increased again due to the protein adsorption. Spreading a close-packed monolayer (147 µg or 7.66 Å2/molecule) induced a jump in the surface pressure to 39 mN/m, immediately followed by a sudden drop and π relaxed to 16 mN/m (Figure 8c). The response of an adsorbed β-lactoglobulin monolayer to the spreading of a monolaurin was also tested by spreading different amounts of monoglyceride on an adsorbing protein solution. Results are shown in Figure 9. As previously observed for monopalmitin-β-lactoglobulin mixtures, the effect of the spread monolaurin

1222

Langmuir, Vol. 16, No. 3, 2000

Cornec and Narsimhan

Figure 6. Time-dependent changes of π (s) and Γ (b) when a solution of [14C]-β-lactoglobulin (cb ) 10-4 wt %) was poured in the trough, followed by immediate spreading of a monopalmitin (a) 49.4 µg (27.67 Å2/molecule), (b) 74.1 µg (18.31 Å2/ molecule), and (c) 113 µg (12.01 Å2/molecule).

Figure 7. Time-dependent changes of π (s) and Γ (b) for a [14C]-β-lactoglobulin (cb )10-4 wt %) solution. At t ) 150 min, (a) 24.7 µg (54.96 Å2/molecule), (b) 74.1 µg (18.31 Å2/molecule), and (c) 113 µg (12.01 Å2/molecule) of monopalmitin was spread at the surface.

monolayer on the adsorption of β-lactoglobulin depended on the amount of monolaurin spread. Protein adsorption was also enhanced by spreading 24.7 µg (45.62 Å2/ molecule) of monolaurin at the surface. However, when this amount was increased up to 49.05, 74.1, and 113 µg (22.97, 15.2, and 9.97 Å2/molecule, respectively), the protein was, first, partially displaced from the surface and readsorbed at the interface almost immediately (see Figure 9b). β-Lactoglobulin was completely displaced from the interface when a close-packed monolayer (147 mg or 7.66 Å2/molecule) of monolaurin was spread. In contrast to the β-lactoglobulin-monopalmitin mixture, the protein was able to slowly readsorb at the surface at longer times (Figure 9c). Also, spreading the monolaurin resulted in a sudden increase in π to a value depending on the amount of monolaurin spread. Then π relaxed rapidly to a value close to 18 mN/m.

remains underneath the condensed monoglyceride film either through hydrophobic interaction between protein and lipid or by local anchoring through the monoglyceride layer.71 The interaction between the molecules at the interface can be characterized by the excess free energy of mixing ∆Gexc defined as:

Discussion At high surface pressures, isotherms of mixed monolayers composed of β-lactoglobulin-monolaurin or β-lactoglobulin-monopalmitin reached the molecular area of the fully condensed monolayer of pure monoglycerides. This indicated that β-lactoglobulin molecules are expelled from the mixed monolayer upon compression. However, upon expansion and further compression, the typical shape of the mixed monolayer was retained, whatever its composition. This suggests that the protein reenters the mixed monolayer and supports the idea that the protein

∆Gexe )

∫0π Aexc dπ

where Aexe is the excess area of the mixed monolayer, i.e., the amount that the measured mean area per molecule differs from the ideal mean area per molecule.72

Aexc ) A - (x1A1 + x2A2) where A is the mean molecular area in the mixed monolayer, A1 and A2 are the molecular areas in the pure monolayers, and x1 and x2 are their molecular fractions. ∆Gexc can be evaluated directly from the π-A isotherms of the pure and mixed monolayers. A positive value for ∆Gexc indicates repulsion between the two different species, whereas a negative value indicates attraction. Figure 10 shows the excess energy ∆Gexc of mixing for β-lactoglobulin-monopalmitin (Figure 10a) and for β-lactoglobulinmonolaurin (Figure 10b) mixed films at different surface (71) Li, J. B.; Kra¨gel, J.; Makievski, A. V.; Fainermann, V. B.; Miller, R.; Mo¨hwald, H. Colloids Surf., A 1998, 142, 355. (72) Garofalakis, G.; Murray, B. S. Colloids Surf., B 1999, 12, 231.

Adsorption and Exchange of β-Lactoglobulin

Figure 8. Time-dependent changes of π (s) and Γ (b) when a solution of [14C]-β-lactoglobulin (cb ) 10-4 wt %) was poured in the trough, followed by immediate spreading of a monolaurin: (a) 12.35 µg (91.25 Å2/molecule), (b) 37.05 µg (30.41 Å2/ molecule), (c) 147 µg (7.66 Å2/molecule).

pressures. ∆Gexc was found to be negative in the whole range of molar ratio studied for all surface pressures. This seems to indicate the existence of attractive molecular interaction at the interface, and thus, there is no phase separation in the mixed film. However, as pointed out by Garofalakis and Murray,72 this thermodynamic treatment can be applied to most binary mixtures of low molecular weight surfactant. In the case of protein, the interpretation of the excess area is more difficult since the protein conformation may change in the presence of other compounds, leading to a possible change in the number of segments at the interface. In particular, β-lactoglobulin interacts with numerous lipid molecules,52 and depending on the nature of the lipid, this interaction leads to more or less conformational stability.41 In addition, a very narrow range of molar ratios has been investigated in this study. Interaction between monoglycerides and β-lactoglobulin was evidenced by spreading various amounts of monoglycerides on an adsorbing solution of protein. Spreading a loosely packed monolayer of monopalmitin was found to enhance the adsorption of β-lactoglobulin soon after the surfactant was spread. Increasing the amount of monopalmitin at the interface resulted in an increase in the initial rate of adsorption, but the steady-state surface concentration was lowered. If a close-packed monolayer was spread at the surface, the protein was excluded or desorbed from the surface. Similar results were obtained by Cho et al.57 with another system composed of BSA and

Langmuir, Vol. 16, No. 3, 2000 1223

Figure 9. Time-dependent changes of π (s) and Γ (b) for a [14C]-β-lactoglobulin (cb )10-4 wt %) solution. At t ) 150 min, (a) 24.70 µg (45.62 Å2/molecule), (b) 73 µg (60 Å2/molecule), and (c) 147 µg (7 Å2/molecule) of monolaurin was spread at the surface.

egg-yolk lecithin. As pointed out by Cho et al.,57 the initial enhancement in the protein adsorption was due to the formation of a complex between the hydrophobic patches of the protein and the hydrophobic tails of the lipid surfactant. The adsorption energy Ead of a protein molecule can be written as

Ead ) Esp + Eel + Ehy where Esp is the work that needs to be done by an adsorbing protein molecule to anchor itself at the interface and acts as the surface pressure energy barrier due to the steric interactions of the molecules already present at the interface, Eel is the electrostatic energy due to the formation of a electrical double layer after the adsorption of protein molecules at the surface, and Ehy is the hydrophobic energy due to the exposure of the hydrophobic patches of the protein toward air.73,74 Eel and Esp being positive will oppose protein adsorption, whereas negative Ehy will promote adsorption. The larger the surface concentration of monoglyceride, the more favorable is the interaction energy Ehy. Protein may form a complex with the monoglyceride molecule if Ehy is greater than the sum Eel + Esp. When the monoglyceride monolayer is closepacked, the contribution from Esp may be dominant and the protein may be excluded or expelled from the interface. (73) Cho, D.; Narsimhan, G.; Franses, E. I. J. Colloid Interface Sci. 1997, 191, 312. (74) Narsimhan, G.; Uraizee, F. Biotechnol. Prog. 1992, 8, 187.

1224

Langmuir, Vol. 16, No. 3, 2000

Figure 10. Excess free energy for (a) monopalmitin-βlactoglobulin and (b) monolaurin-β-lactoglobulin mixed monolayers as a function of varying film composition and surface pressure: ([) π ) 5; (0) π ) 10; (2) π ) 15; (O) π ) 20; (9) π ) 25 mN/m.

Based on the average surface hydrophobicity and molecular dimensions of β-lactoglobulin, the estimated value of Ehy is of the order of 60 kcal/mol. On the other hand, the estimated values of Eel were found to be negligible at any protein surface concentration (1.4 kcal/mol at the highest value of Γ ≈ 2.18 mg/m2). Ehy was found to be higher than the sum Eel + Esp for all surface pressures lower than 11 mN/m. Consequently, at all π < 11 mN/m, β-lactoglobulin adsorption is enhanced. However, hydrophobic interaction with monopalmitin may lead to larger Ehy, thus explaining why protein adsorption was observed to be enhanced at higher π. On the other hand, for a close-packed monopalmitin monolayer (π ) 47.5 mN/m), Esp was estimated to be of the order of 275 kcal/mol. Therefore, Ehy < Eel + Esp, so that the protein is expelled from the interface. Spreading a monolayer of monolaurin instead of monopamitin led to different results. The initial rate of adsorption was observed to decrease, whereas the steadystate surface concentration was unchanged. When a closepacked monolaurin monolayer was spread onto an adsorbed β-lactoglobulin solution, the protein was initially desorbed but was observed to slowly readsorb at longer times. Unlike the β-lactoglobulin-monopalmitin mixture, spreading of monolaurin monolayer was accompanied by a large overshoot in π, which was observed to relax immediately to a much lower value comparable to a monolayer of β-lactoglobulin. Interestingly, during this period of surface pressure relaxation, protein was expelled from the interface. Such a behavior seems to indicate that the spread monolayer of monolaurin rearranges during this relaxation period. For insoluble monolayer, at surface

Cornec and Narsimhan

Figure 11. Illustration of possible mechanism for proteinmonoglyceride interactions at the air-water interface: (a) Initially, when no monoglyceride monolayer, only a few protein molecules adsorb. (b) Once monoglyceride spreads on the surface, the attractive interactions between surfactant molecules and subphase protein molecules enhance protein adsorption. (c) Spread monoglyceride molecules rearrange leading to partial displacement of protein molecules. (d) If the spread monoglyceride monolayer is close-packed, protein molecules are expelled from the interface. (e) If π is higher than a critical surface pressure, the monoglyceride monolayer collapses, resulting in the formation of lenses, between which protein is allowed to readsorb.

pressure higher than the equilibrium surface pressure, πe, relaxation phenomenon is due to the transformation of homogeneous monolayer phase into a heterogeneous monolayer collapse phase system.75 Whenever a characteristic surface pressure is exceeded, collapse may occur, either by a macroscopic film fracture or by a process of nucleation and growth of bulk surfactant fragments.75 This may result in the formation of lenses at the interface, making room available for the protein to adsorb. Exclusion of protein during the relaxation period is believed to be due to the surface pressure energy barrier. For example, when π ) 37 and 47 mN/m, Esp ) 196 and 266 kcal, respectively. Then, Ehy , Eel + Esp, and the protein is expelled from the interface. After the relaxation of the surface pressure, however, the surface pressure energy barrier decreases to 57 kcal for a surface pressure of 10 mN/m. As a result, Ehy > Eel + Esp. Consequently, the protein tends to readsorb after the rearrangement of the monolayer. The above mechanism of interaction of protein with the spread monoglyceride monolayer is depicted schematically in Figure 11. It was recently reported that the stability of monoglyceride monolayer depends on factors such as the hydrocarbon chain length, the presence of a double liaison, and the pH of the subphase. Maximum stability was observed for monopalmitin monolayer whereas monolaurin monolayer exhibited the lowest stability.75 This is consistent with the results observed in this study. (75) Carrera Sa´nchez, C.; Rodrı´guez Nin˜o, M. R.; Rodrı´guez Patino, J. M. Colloids Surf., B 1999, 12, 175.

Adsorption and Exchange of β-Lactoglobulin

Conclusion π-A isotherms of spread monolayers of β-lactoglobulin, monopalmitin, monolaurin, and their mixtures were determined with a Langmuir minitrough. The isotherms of monolaurin-β-lactoglobulin mixtures were closer to that of protein alone for weight ratios of monolaurin to β-lactoglobulin of 0.33-3.0. All mixed monopalmitin-βlactoglobulin isotherms exhibited a phase transition between a liquid-expanded state to a liquid-condensed state, which indicated that the monoglyceride had a strong influence on the properties of the mixed monolayer during compression. Monopalmitin monolayer exhibited much higher surface elasticity than β-lactoglobulin and monolaurin. In both systems studied, the isotherms of the mixed monolayers approached asymptotically that of the monoglyceride monolayer, thus indicating that the protein was squeezed out from the monolayer at very small areas of the monolayer. The dynamics of adsorption of [14C]-βlactoglobulin for different spread amounts of monoglyceride (monopalmitin or monolaurin) and time delays after the initiation of protein adsorption were investigated

Langmuir, Vol. 16, No. 3, 2000 1225

through the measurements of surface pressure and protein surface concentration by using a radiotracer technique. Spreading a monoglyceride onto an adsorbing protein solution was found to enhance the rate of adsorption of the protein. For both systems, spreading a low amount of lipid at the interface resulted in an enhancement of the protein adsorption at short times leading to a protein surface concentration higher than the steady-state value in the absence of monoglyceride. Spreading high amount of monopalmitin decreased the amount of protein adsorbed, and spreading a close-packed lipid monolayer caused eventually desorption of protein from the interface, maybe because of the surface pressure and steric exclusion effects produced by the spread monopalmitin monolayer. β-Lactoglobulin was also expelled by high surface concentrations of monolaurin. However, instability in the monolaurin film caused the monolayer to rearrange, leading to a dramatic decrease in the surface pressure. This allowed β-lactoglobulin to readsorb. LA990326P