General Features of Adsorption from Single-Protein Solutions
Solution Chemistry Effects on Adsorption. The net charge of a protein in solution is dependent on the difference between pH of the solution and the isoelectric point (pI) of the protein. If the pH of the solution is greater than the pI, the net charge of the protein would be negative, whereas if the pH is less than the pI, the net charge of the protein would be positive. It is generally accepted that maximum adsorption occurs at the isoelectric point. As the out-of-balance charge of a protein increases, it will be in a more extended form than when the net charge is zero (11).
Norde and Lyklema (12) suggested that the degree to which pH affects the adsorption of a protein is determined by its conformational stability. They found that plateau values of adsorbed mass were independent of pH for structurally stable proteins, whereas those of less stable proteins varied considerably, apparently because less stable proteins were able to change structure with solution conditions. The effect of pH on protein adsorption and desorption can depend on solution history as well (13). Kondo and Higashitani (14) studied the adsorption of model proteins with wide variation in molecular properties. They explained the pH dependence of adsorbed mass in terms of lateral interactions. In particular, they suggested that lateral interactions between large protein molecules are stronger than those between small molecules. Large proteins would thus be expected to show maximum adsorption around their pIs, whereas the effect of pH on smaller proteins would be less pronounced.
The degree to which ionic strength affects protein adsorption is a function of the role electrostatic plays in the adsorption driving force. At low ionic strength, protein surface charge fully contributes to the total electrostatic interaction (11). At high ionic strength, the surface charges of proteins are shielded, reducing electrostatic interactions between proteins, whether attractive or repulsive (13). Luey et al. (15) showed that ionic strength effects on adsorbed mass are very much related to solid surface properties. They observed that increased ionic strength reduced the electrostatic repulsion between negatively charged j-lactoglobulin molecules and the hydrophilic, negatively charged surface they studied, increasing adsorbed mass. By contrast, increased ionic strength resulted in little change in adsorbed mass at hydrophobic surfaces.
Surface-Induced Conformational Changes. It is well accepted that a given protein can exist in multiple adsorbed conformational states on a surface (16-20). These states can be distinguished by differences in occupied area, binding strength, propensity to undergo exchange events with other proteins, and catalytic activity or function. All these features of adsorbed protein are interrelated and can be time-dependent. For example, decreases in surfactant-mediated elution of proteins from an adsorbed layer (an indirect measure of binding strength) are observed as protein-surface contact time increases (21). This time-dependence is illustrated in Figure 1. As conformational change proceeds, the likelihood of desorption decreases.
It has been observed that the extent of conformational change experienced by adsorbed fibrinogen increases with contact surface hydrophobicity (22). This is consistent with findings of Elwing et al. (23), who used ellipsometry to make inferences regarding conformational changes experienced by complement factor III, a plasma protein, on hy-drophilic and hydrophobic silica surfaces. The results of tl t t
Figure 1. Surface-induced conformational changes undergone by adsorbed protein, resulting in multiple noncovalent bonds with the surface and coverage of greater interfacial area per molecule.
Figure 1. Surface-induced conformational changes undergone by adsorbed protein, resulting in multiple noncovalent bonds with the surface and coverage of greater interfacial area per molecule.
Elwing et al. also showed that greater values of adsorbed mass were found on hydrophobic as opposed to hydrophilic surfaces. Protein molecules are assumed, in general, to change conformation to a greater extent on hydrophobic surfaces. This is due to the effect of hydrophobic interactions between the solid surface and hydrophobic regions in the protein molecule. In fact, surface-induced unfolding is often characterized as entropically driven, because the hy-drophobic protein interior associates with hydrophobic regions of the surface during unfolding. These interactions can give the molecule an extended structure, covering a relatively large area of the surface. If the repulsive force normally acting between native protein molecules is decreased for such structurally altered molecules, one would expect to measure a greater adsorbed mass on hydrophobic as opposed to hydrophilic surfaces. On the other hand, adsorption of positively charged protein to hydrophilic (negatively charged) silica can result in greater conformational change than adsorption of the same protein to hydrophobic silica, even with a greater extent of adsorption being observed at the hydrophobic silica surface (24). It is thus important to recognize that multiple factors affect the extents of protein adsorption and conformational change.
The concept that adsorbed proteins can exist in multiple states on a surface plays a role in interpretation of most if not all experiments in protein adsorption. Biophysicists rather routinely gain information relevant to protein structure in solution with circular dichroism (CD). It would be attractive to use CD in structural study of adsorbed protein as well. A recent innovation has made CD more applicable to study of structural changes during adsorption, and that is the use of colloidal silica particles, or nanoparticles (1620). In these tests, particles have ranged from less than 10 to about 30 nm and are small enough not to interfere with the CD spectra. Individual molecules are allowed to adsorb to nanoparticles, resulting in a stable suspension of adsorbed protein. In this way, structural changes on adsorption have been unambiguously measured.
Work by Billsten et al. (19) and Tian et al. (20) have provided the most direct illustration of the effect of stability on structural rearrangements at a solid surface. Using site-directed mutants of bacteriophage T4 lysozyme, these investigators showed that both the rate and extent of secondary structure loss upon adsorption to colloidal silica were clearly related to protein thermal stability. With the same mutants, Froberg et al. (25) used the interferometric surface force technique to study structural characteristics of adsorbed layers of T4 lysozyme. The results demonstrated that less-stable mutants lose their tertiary structure upon adsorption, whereas more stable mutants retain their globular shape.
Steady-State Adsorption Behavior. A great deal is known about how various conditions affect the steady-state adsorbed mass of protein. Numerous protein adsorption isotherms have been constructed and compared on the basis of temperature, pH, ionic strength, conformational stability of the protein in solution, and solid surface charge and hydrophobicity. The effects of protein conformational stability and solid surface properties are perhaps best revealed with reference to effects of pH and ionic strength.
In general, the effect of pH and ionic strength on protein adsorption is dependent on which type of interactions predominate (e.g., electrostatic, hydrophobic, or van der Waals interactions). At a negatively charged surface, if electrostatic interactions predominate, adsorbed mass should be greater at pH values below the isoelectric point relative to pH values above it. Below the isoelectric point, the protein and surface would be of opposite charge, whereas both the protein and surface would be negatively charged at pH values greater than the isoelectric point. As ionic strength increases, the electrostatic interaction would be reduced because of shielding of the protein by counterions; consequently, increasing the ionic strength should decrease adsorbed mass at pH values less than the isoelectric point and increase the adsorbed mass at greater values of pH. The relationship between adsorbed mass and changes in pH and ionic strength becomes inextricably linked to protein conformational stability. In general, pH and ionic strength conditions that lead to a less stable conformation for the protein in solution will lead to an increased adsorbed mass, assuming that the protein molecule would be more stable on the solid surface (15).
Another observation of importance is that protein adsorption is often an apparently irreversible process, at least in the sense that is often irreversible to dilution or buffer elution. The adsorbed mass remains constant or decreases very little when the solution in contact with the solid surface is depleted of protein. This irreversibility is more pronounced as protein-surface contact time increases. However, although spontaneous desorption is not generally observed, adsorbed protein can undergo exchange reactions with similar or dissimilar protein molecules adsorbing from solution (26). Such exchange reactions are shown schematically in Figure 2. Adsorbed protein exchange rates are likely state-dependent, being slower for more conformationally altered protein.
Some workers have reported that protein adsorbs onto a solid surface in more than one layer. Arnebrant et al. (27) studied adsorption of /-lactoglobulin and ovalbumin on hy-drophilic and hydrophobic chromium surfaces using ellip-sometry and electrical potential measurements. On hydro-philic surfaces, their results showed that a highly hydrated layer is obtained that can be partially removed by rinsing. They suggested that the protein adopts a bilayer formation
on the surface, with the layer in direct contact with the surface being unfolded and attached by strong polar bonds. Rinsing showed that the outer protein layer is loosely attached, which would imply that molecules in the outer layer have a structure closer to that of their native state. This adsorption behavior was described in terms of surface-induced conformational changes and charge interactions between the protein and surface. In particular, there are always polar amino acid side chains that can interact strongly with a surface, even if both the protein and surface are negatively charged. Such binding might be expected to result in unfolding of the protein. A consequence of this might be exposure of hydrophobic regions into aqueous solution; adsorption of a second protein layer would therefore reduce the interfacial free energy. In the case of protein in contact with a hydrophobic metal surface, values of adsorbed mass were found to be consistent with formation of a monolayer. Arnebrant and Nylander (28) reported possible bilayer formation upon adsorption of oligomeric units of insulin as well.
Adsorption Kinetics. In considering the kinetics of any interfacial process, the question of transport versus reaction control must be addressed. Protein adsorption at an interface is dependent not only on the intrinsic kinetic rate, which is a function of protein, solution, and surface properties, but also by the rate of protein transport from the bulk solution through the concentration boundary layer near the interface.
Proteins are macromolecules, and they can possess domains that differ chemically and physically. Diffusion coefficients may vary widely among proteins, depending on their concentration and the electrostatic condition of the solution (29). The initial adsorption rate of protein molecules at an interface can be transport-limited at either low or high concentration. The diffusion limitation exists as long as there is a significant concentration gradient near the solid surface. With careful design of an experimental system to minimize the transport-limited period, however, an intrinsic adsorption kinetic rate can be estimated. Still, relatively little is known about the nature of the adsorbed layer, and predictive models to describe any aspect of adsorption as a function of protein and interfacial properties are lacking. Protein adsorption is characterized by the likely presence of a time dependence in the development of bonds with the surface, a time dependence in the lateral mobility and exchangeability of the protein molecules, and time-dependent conformational changes; it is thus very difficult to describe mathematically.
Many experimental observations have indicated that a major portion of the final adsorbed amount had been adsorbed within the first few minutes of contact. Soderquist and Walton (30) proposed the existence of three distinct processes contributing to protein adsorption kinetics on polymeric surfaces. First, rapid and reversible adsorption of the proteins occurs during a short period of time. At up to 50 to 60% of surface coverage there is a random arrangement of adsorbed molecules, but then some form of surface transition occurs that is probably in the direction of surface ordering, thereby allowing further protein adsorption. Second, molecules on the surface undergo structural transitions as a function of time that occur in the direction of optimizing the protein-surface interaction. Third, as time increases the probability of desorption decreases, and the adsorption becomes irreversible to dilution.
Molecular Structure and Interfacial Behavior. Study of molecular influences on protein adsorption has received much attention because of its relevance to better understanding of adsorption competition in multiprotein mixtures. Important contributions to current understanding of molecular influences on protein adsorption have evolved from several comparative studies of protein interfacial behavior, in which similar or otherwise very well-characterized proteins (31-34), genetic variants (35,36), or site-directed mutants (19,20,24,25,37-39) of a single protein had been selected for study. A number of factors are known to affect protein adsorption, and these studies have stressed the importance of protein charge, hydrophobicity, and structural stability in interfacial behavior.
Shirahama et al. (33) studied hen lysozyme, ribonucle-ase A, and a-lactalbumin adsorption to hydrophilic and hy-drophobic, polystyrene-coated silica (both negatively charged surfaces). At hydrophilic silica, they found that adsorbed mass increased with increasing charge contrast between the surface and protein. At the hydrophobic surface they found electrostatic interaction to have a lesser effect, in that the adsorbed mass was not clearly related to charge contrast between the surface and protein. Arai and Norde (31) described adsorption from single-protein solutions of hen lysozyme, ribonuclease A, myoglobin, and a-lactalbumin to synthetic materials varying in surface charge density and hydrophobicity. They concluded that at a given surface, adsorption of a globular protein is related to its structural stability. That is, proteins of high stability behave like hard particles at a surface, with the interactions governed by surface hydrophobicity and electrostatics, whereas adsorption of proteins of low stability (soft proteins) may be influenced by structural rearrangement, allowing adsorption to occur even under conditions of electrostatic repulsion.
Horsley et al. (35) compared isotherms constructed for hen and human lysozymes at silica derivatized to exhibit negatively charged, positively charged, or hydrophobicsur-faces. Differences in adsorptive behavior observed between the two lysozyme variants were largely explained with reference to the fact that human lysozyme contains one less disulfide bond and is less thermally stable than hen lyso-zyme. Xu and Damodaran (36) compared adsorption kinetic data measured for native and denatured hen, human, and bacteriophage T4 lysozymes at the air-water interface. Their results showed substantial differences in adsorption dynamics among the three variants, as influenced by their structural state and the physical and chemical nature of the protein and surface.
Kato and Yutani (37) evaluated the interfacial behavior of six site-directed mutants of tryptophan synthase a-subunits, produced by amino acid substitution in the protein's interior, using surface tension, foaming, and emulsifying property measurements. The stability of these subunits, as measured by their free energy of denaturation in water, varied from about 5 to 17 kcal/mol. They were able to attribute differences in interfacial behavior to protein stability with good success. In particular, they observed that less stable mutants were most surface active, that is, they more rapidly adsorbed or more readily unfolded at the hydrophobic interfaces studied in that work.
Multicomponent Systems. Shirahama et al. (33) used hen lysozyme, ribonuclease A, and a-lactalbumin to study sequential and competitive adsorption at hydrophilic and hydrophobic, polystyrene-coated silica. At hydrophilic silica, they found that once an adsorbed layer of a given protein was formed, almost total displacement of that protein would occur upon introduction of a second protein to solution if the second protein had a more favorable capacity for electrostatic attraction with the surface (otherwise, sequential adsorption was not observed). In addition, when adsorbed from a mixture, the protein capable of more favorable electrostatic attraction with the surface preferentially adsorbed, essentially to the exclusion of the other protein. At the hydrophobic surface, they found that once an adsorbed layer of a given protein was formed, only partial displacement of that protein would occur upon introduction of a second protein to solution, even if the second protein had a more favorable capacity for electrostatic attraction with the surface. Moreover, the eventual make-up of a film adsorbed from a mixture was not related to charge contrast between the protein and surface. Other experimental observations (40,41) indicated that adsorbed protein molecules undergo exchange with protein from solution more readily on hydrophilic than on hydrophobic regions of a surface. Arai and Norde (32) studied the sequential and competitive adsorption behavior exhibited by hen lysozyme, ribonuclease A, myoglobin, and a-lactalbumin and concluded that whether introduced in sequence or in a mixture, adsorption of a globular protein is related to its structural stability. In particular, interfacial behavior of proteins of high stability is governed by surface hydrophobicity and electrostatics, whereas that of proteins of low stability are more influenced by structural rearrangement.
The elution of adsorbed protein by surfactant has been used to provide a measure of protein binding strength
(21,24,42-51). The essential steps of this type of experiment are illustrated in Figure 3. Adsorption is allowed to occur, followed by rinsing with protein-free buffer. A surfactant solution is then introduced, after which adsorbed protein is displaced or solubilized (45). This is followed by rinsing and comparison of the amounts of protein present before surfactant addition and after the final rinse. Elution by dissimilar protein has been used as a measure as binding strength as well. Slack and Horbett (52) evaluated the strength of attachment of fibrinogen to solid surfaces by measuring its time-dependent elution in plasma and modeled fibrinogen adsorption, with reference to its rate of conversion from a weakly bound (exchangeable) to a tightly bound (nonexchangeable) state. Wahlgren and Arnebrant (45,46) used in situ ellipsometry to continuously monitor the different effects of cationic and anionic surfactants on the elution of /-lactoglobulin and lysozyme from hydro-philic and hydrophobic surfaces as well as adsorption from protein and surfactant mixtures. The elution studies allowed postulation of four mechanisms for surfactant-mediated elution of adsorbed protein. With the aim of relating elutability to protein molecular properties, Wahl-gren et al. (47) studied removal of well-characterized proteins from silica surfaces using dodecyltrimethylammon-ium bromide. Some general trends regarding molecular property effects on elutability emerged from that work, but clear correlations between molecular properties and elut-ability remained difficult to quantify. By contrast, similar tests conducted with synthetic mutants of bacteriophage T4 lysozyme showed a clear correlation between protein stability and elutability (24). In particular, less stable proteins are more resistant to elution, presumably because they are more able to alter their conformation at the surfaces.
Allow adsorption to occur, then rinse:
Allow adsorption to occur, then rinse:
Introduce a surfactant:
Introduce a surfactant:
Figure 3. Experimental approach to evaluating adsorbed protein-binding strength using surfactant-mediated elution.
Modeling the Process. A number of mathematical models of protein adsorption at air-water and solid-water interfaces have been constructed (38,53-57). The problem is generally approached as an issue of molecular diffusion through a potential gradient, a reaction-diffusion problem involving interactions between diffusive-convective protein transport from the bulk solution and competitive adsorption and exchange kinetics on the surface, or as a ki-netically controlled phenomenon, involving adsorption, unfolding, and exchange. Such models and the kinetic simulations they allow provide a framework with which the complexity of protein adsorption can be better understood and quantified. In addition, they can be used to provide direction for further experiments, particularly involving surface modification.
Lundstrom (58) presented an equilibrium model of protein adsorption on solid surfaces. The model described the fractional surface coverage of adsorbed molecules as a function of equilibrium concentration and allowed for reversible adsorption and conformation change. Later, Lund-strom and Elwing (26) described a model that allowed for bulk-surface exchange reactions among proteins in single-component and binary mixtures. The work featured manipulation of the equations describing the fractional surface coverage of protein in specific states and simulations of total surface coverage as a function of equilibrium concentration and as a function of time. Although no experimental data were presented, the shapes of the curves were in qualitative agreement with experimental observations. Currently, there is no adequate method to directly monitor changes in fractional surface coverages of protein in different adsorbed states. A less-complex model that can be statistically compared with the available data would be useful, because it would enable individual rate constants to be related to surface, solution, and protein properties.
Past work with synthetic mutants of bacteriophage T4 lysozyme have involved in situ ellipsometry and surfactant-mediated elution (24,51), ring tensiometry (38), the interferometric surface force technique (25), CD (19,20), and spectrophotometric assays of bound enzyme activity (39). These studies have shown that structural alterations definitely occur upon adsorption, with the extent and rate of structural change being related to thermal stability. In addition, mutants exhibited resistances to surfactant-mediated elution that were proportional to thermal stability and consequently related to extent of structural change. Finally, concerning a number of T4 ly-sozyme variants, results could be explained by modeling adsorption as occurring such that molecules adopt one of only two states, each varying in binding strength and occupied area, with differences in behavior among the molecules attributable to the relative populations in each state.
The simplest adsorption mechanism consistent with the fact that adsorbed proteins can exist in multiple states would include two adsorbed states. Figure 4 shows such a mechanism. Rate constants k1 and k2 govern adsorption into states 1 and 2, respectively. Although the mechanism is drawn to depict molecules adsorbing directly into states 1 and 2 from solution, a more accurate and detailed mechanism might include a multistep path to state 2. However,
for modeling purposes, the actual path to state 2 is not consequential; we need only account for the different rates of generation of two functionally dissimilar forms of adsorbed protein. If adsorption of practically relevant proteins can be adequately described in this way, extension to the case of competitive protein adsorption would be straightforward.
Figure 5 shows a mechanism for competitive adsorption (between two proteins, A and B) based on Figure 4. In each case, all associated rate constants can be determined a priori. The protein-specific k1 C and k2 C of Figure 4 can be obtained from single-component kinetic data, and the various exchange constants can be determined through sequential adsorption experiments (59). Figure 5 can be easily redrawn to depict competitive adsorption of three proteins, A, B, and C or more. As long as sequential adsorption data are available for each pair permutation of A, B, and C, for example, an a priori estimate of all rate constants can be made, and the adsorption competition can be simulated. Such comparisons would provide a basis for design of further experiments to better resolve molecular ef-
Was this article helpful?
Discover How To Become Your Own Brew Master, With Brew Your Own Beer. It takes more than a recipe to make a great beer. Just using the right ingredients doesn't mean your beer will taste like it was meant to. Most of the time it’s the way a beer is made and served that makes it either an exceptional beer or one that gets dumped into the nearest flower pot.