Langmuir 1990, 6, 376-378
376
-5.00-
-15.0. UA
.
-25.0.
-35.0-
,-.600 '
- . i o '-.-io0
.
.do0
'
,200
V vs SCE
Figure 7. "Synthesized" voltammogram for an array of five active sites. The solid curve with asterisks is the total current (eq 11). The steady-state currents at the five active sites are calculated from eq 8 with C* = 1.8 X lo* mol cm3 and I/sm cm /s, (B)7.4 pA and A,k? equal to (A) 3.7 pA and 1.05 X and 1.05 X cm3/s, (C) 14.7 MAand 1.05 X lo+ cm3/s, (D) 29.5 FA and 1.05 X lo-' cm3/s,and (E) 59.0 PA and 1.05 X lo-' cm3/s.
I
more distant from the electrode surface. Because of the number of adjustable parameters used in the model (10 in Figures 6 and 7), a more quantitative interpretation is precluded. The active sites are tentatively identified as (a) regions of collapsed monolayer and/or (b) sites where the redox molecules actually penetrate the monolayer. Such sites may exist at grain boundaries, atomic steps on the gold
surface, or packing discontinuities in the monolayer required by the tilt in the octadecyl chain.' The redox molecules are presumed to approach closer than the thickness of the ClsSH monolayer because tunneling currents across 20 A are expected to be well below the currents obtained at the largest overpotentials. Tunneling currents between metal films separated by a single monolayer4' or between an electrode and an adsorbed cobalt complex41yield a tunneling constant of 1.6 1\ per order of magnitude decrease in current. From the tunneling constant, we estimate that the kinetic current at a given potential should be reduced by 10-13 orders of magnitude. The latter figure assumes that tunneling occurs across the full 20 A of the monolayer, while the former figure assumes that electron transfer is adiabatic over reducing the tunneling distance accordthe first 5 ingly. For Ru(NH,):+at -0.6 V, the tunneling current A, considerably below is predicted to be from lo-'' to the observed 5 X loF5A in Figure 6. In summary, pinholes in a C,,SH monolayer on a gold electrode can be characterized in terms of average size and separation by a combination of electrochemicalmethods. The pinholes can be permanently passivated by electrochemical polymerization of phenol in sulfuric acid. The residual faradaic currents obtained on the ClsSH + PPO coated electrode are attributed to electron tunneling between the electrode and redox molecules close to (40%), practically identical isotherms were obtained. The adsorption data were analyzed by using the Langmuir equation, and it appeared that increase in modification led to decrease in area per adsorbed protein molecule, while the adsorption constant remained unchanged. Introduction
The ability to control physical adsorption of proteins has great importance in applications such as enzyme immoCasali Institute of Applied Chemistry, School of Applied Science and Technology. Department of Biological Chemistry, Life Sciences Institute.
*
0743-7463/90/2406-0376$02.50/0
bilization, immunodiagnostics, chromatography, and emulsion ~tabilization.'-~ (1) Unger, K. CRC Handbook of HPLC for separation of amino acids peptides and proteins; CRC: Boca Raton, 1979;Vol. 1. (2) Huang, A.;Txoo, Y. S.; Kennel, S. J.; Hung, L. Biochim. Biophys' Acta 'I6, 140* ( 3 ) Zittle, C. A. Adu. Enzymol. 1963,14, 319. (4) Pearce, K. N.; Kinsella, J. E. J. Agric. Food Chem. 1978,26, 716.
0 1990 American Chemical Society
Adsorption of Stearyl-Modified Ovalbumin
Langmuir, Vol. 6, No. 2, 1990 377 Table I. Effect of EstedProtein Ratio in Synthesis of Stearic-Modified Ovalbumin ester, ovalbumin, mol of ester/ % mg mg mol of ovalbumin modification 300
Figure 1. Synthesis of modified ovalbumin: R is stearyl group; P is the ovalbumin.
The adsorption of proteins on various surfaces was studied intensively by Lyklema et al.59' and recently reviewed.' From these studies, a model emerged in which the adsorption is influenced by conformation and charge of proteins and by the physicochemical properties of the adsorbent.' However, most of the adsorption studies were performed with native proteins which cannot offer the possibility of varying only the hydrophobicity of the native protein. Therefore, it was of interest to study how modification of hydrophobicity within the same molecule could affect its adsorption properties. Hydrophobicity of proteins may be modified either by covalent attachment of various groups' or by adsorption of surfactants such as sodium dodecyl sulfate onto the protein mole~ule.~ Such modifications of proteins were used in order to improve emulsification of various oildo*'' and to increase activity of enzymes such as chemotrypsin a t the oil-water interface." In this study, a model protein, ovalbumin, was chemically modified with stearic residues, and its adsorption on hydrophobic silica was investigated.
Experimental Section Materials. Stearic acid (BDH),N-hydroxysuccinimide, and o-phthalaldehyde analytical reagents were obtained from Sigma. Grade V ovalbumin (Sigma) and Brij 35 (Atlas Europol) were used without further purification. Hydrophobic silica, Lichrosorb RP-8, with a particle size of 7 pm and surface area of 0.25 m2/ mg was supplied by Merck. The average pore size of Lichrosorb RP-8 is 148 A.' Methods. 1. N-Hydroxysuccinimide Active Ester Synthesis. Stearic ester was synthesized according to Lapidot et 2. Modified Ovalbumin Synthesis. The modification was achieved by reacting the native albumin with the active ester as shown in Figure 1. Ovalbumin (1OO-3OOmg) was dissolved in 100 mL of 50 mM sodium carbonate-bicarbonate buffer (pH 10). Active ester (10300) mg was dissolved in 3 mL of 1,4-dioxaneand was dropwise added to the ovalbumin solution with stirring. The mixture was incubated for 72 h at 37 "C and then filtered through a 0.2-pm filter. The filtrate was dialyzed against 100 volumes of sodium carbonate-bicarbonate buffer and then 5 times against distilled water. The modified protein was isolated by lyophilization. 3. Determination of Degree of Modification. The average degree of modification was determined as follows: the total concentration of ovalbumin (C,) was determined according to the Lowry method,14which measures the concentration of tyrosine (5) Norde, W.; Lyklema, J. J. Colloid Interface Sci. 1979, 71, 350. (6) Lyklema, J. Colloid Surf. 1984, IO, 33. (7) Proteins at Interfaces; ACS Symposium Series 343; Brash, L. J., Horbett, Eds.; American Chemical Society: Washington, D.C., 1987. (8) Ansari, A. A.; Kidwai, S. A.; Salahuldin, A. J.Biol. Chem. 1975, 28,1925. (9) Kato, A.; Nakai, S. Biochim. Biophys. Acta 1980, 624, 13. (10) Franzen, K. L.; Kinssela, J. E. J. Agric. Food Chem. 1976,24, 788. (11) Pearce, K. N.; Kinsella, J. E. J. Agric. Food Chem. 1978, 26, 716. (12) Torchilin, V. P.; Amelyantro, V. G.; Alibanov, A. L.; Mikharilov, A. I.; Goldonski, V. 1.; Smirnov, V. N. Biochim. Biophys. Acta 1980, 602, 511. (13) Lapidot, J.; Rappoport, S.; Wolman, J. J. Lipid Res. 1967, 8, 142. (14) Peterson, G. L. Methods Enrymol. 1983,91,95.
100 50 20
200 300 300 300
179 39 20
8
57 33 24 13.5
residues, and then measured by a fluorescent method14 (C,), which measures the concentration of amine groups. The average degree of modification was calculated by % modification = [(C, - Cf)/C,]lOO
(1) Ovalbumin contains 20 lysine residues/molecule.16 4. Adsorption of Ovalbumin. Adsorption isotherms were obtained with various modified proteins (0.2-2 mg protein/mL of sodium carbonate-bicarbonate buffer) and a constant weight of hydrophobic silica (2 mg). The suspensions were mixed overnight (25 "C) in glass vials by rotating end to end on a vertically rotating table, avoiding foam formation. The suspensions were filtered (0.2 pm Millipore). Within the concentration range studied, the adsorption was constant after 8 h. The protein concentration in the filtrate was determined according to L0wy,l4 and the adsorbed amount was calculated by eq 2
S = (C,-Cz)V/M (2) where S is the adsorbed amount (mg of protein/mg of silica), C, the initial concentration (mg/mL), C, the final concentration (mg/mL), V the solution volume (mL), and M the silica weight (mg).
Results and Discussion A series of ovalbumin at various degrees of modification was prepared by changing the ratio of active ester to native ovalbumin. As shown in Table I, increasing this ratio led to increased degrees of modification. It was found that the solubility of the modified protein decreased with the increase in extent of modification. Thus, at -40% modification, the protein solution was saturated at 2 mg/mL. This limited solubility set the limit of protein concentrations in the adsorption experiments. The adsorption isotherms of the native and modified ovalbumins (Figure 2) demonstrate that an increasing degree of modification resulted, in general, in higher adsorption. It is worth noting that even a low degree of modification, such as 13.5%,which is equivalent to about two stearyl groups per protein molecule, has led to a significant increase in adsorption. Increasing the modification from 33% to 57% (-6 to -11 stearyl groups) led to identical adsorption isotherms. It should be noted that at pH 10 both the native and the modified proteins are negatively charged. Calculation based on amino acid composition shows that the difference in charge between the mostly modified and the native protein is smaller than 2 % . Therefore, although electrostatic effects may play a role during the adsorption process, they cannot account for the difference in adsorption properties between the native and the modified ovalbumins. Many investigators found that protein adsorption onto various solid surfaces follows the Langmuir isotherm and calculated the maximum adsorption and adsorption energy by the Langmuir equation for various proteins.1e-18 The (15) Lewis, J. J. Biol. Chem. 1950, 186, 23. (16) Morrssev, B. W.: Stromberp, R. R. J. Colloid Interface Sci. 1974, 46, 152. (17) Kamyshnyi, A. L. Russ. J.Phys. Chem. 1981,55(3), 319. (18) Barford, R. A,; Sliwinski, B. J. J. Dispers. Sci. Technol. 1985, 6, 1.
378 Langmuir, Vol. 6, No. 2, 1990
Magdassi et al. 1
1
1.4
-
41
'i /
-.
' . .
o
IO
20
30
40
50
LO
% Modification
Figure 4. , S
(calculated by the Langmuir equations)vs degree of modification.
0.2
0.0
0.4
0.6
0.6
1.0
1.2
1.4
1.6
C2(mglmI)
Figure 2. Adsorption isotherms of native and modified oval, native ovalbumin; X, 13%; A,24%; 0,33% + 57% bumins: . modification. J
-
2.0
. . i
E e0
1.5-
v yl
1.0-
U N
0.5
-
00
02
04
0 6
0 8
1 0
1 2
C2(mg/ml)
Figure 3. Adsorption isotherms of native and modified ovalbumin presented in Figure 2.
adsorption data for the modified proteins in the present study were analyzed by using the linear form of the Lang muir equation, as shown in Figure 3. From these plots, the maximum adsorption and the adsorption energy were calculated. The adsorption isotherm of the native ovalbumin was analyzed by a nonlinear curve-fitting procedure, due to dispersity of the data which resulted from the low sensitivity of concentration measurements at low adsorption. As presented in Figure 4, the maximum adsorption, S,,, increases with the increase in degree (19) Chriatofferson, M. R.; Christofferson, J.; Ibsen, P.; Ipsen, H. Colloids Surf. 1986, 18, 1.
of modification until it reaches a constant value at modifications above 33%. Assuming a monomolecular layer, these values of, ,S reflect the following areas per molecule: 4600,1500,670, 610,and 580 A2 for native ovalbumin, 13%, 2 4 % , 3 3 % , and 57?& modifications, respectively. From these results it appears that either a change in the orientation of the adsorbed protein occurs or a significant decrease in the dimensions of the modified proteins occurs. As reported by Ansari et al.,' modification of ovalbumin with acetyl groups leads to an increase in the length of the molecule (which is an ellipsoid, 30 X 100A): and a slight decrease in its width. Therefore, it seems that the above adsorption results could be explained by a change from a sideon orientation to an end-on orientation of the adsorbed proteins. It should be emphasized that these calculations are based on the validity of the Langmuir equation and on the assumption that all the area of Licrosorb RP-8 is available for protein adsorption. Support for this assumption is found in the similar area per ovalbumin molecule which was reported for adsorbed ovalbumin on calcium hydroxyapatite." The adsorption constant, K, was also calculated and is in the range 0.64-0.96mg/mL, independent of degree of modification. The adsorption energy calculated from these constants is -9 kcal/mol, which is within the range reported for protein adsorption." It is rather surprising, since it could be expected that increasing hydrophobicity of the protein would lead to an increase in entropy upon adsorption. Since adsorption of proteins is mostly entropically driven?6 it could be expected that the energy of adsorption would increase with the increase of degree of modification. Registry No. N-Hydroxysuccinimide stearic ester, 1446432-5.
