Hydrophobic interactions of proteins

• Main Author: Louis, Charles F.
• Published: University of Oxford 1968
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.731991

This work is concerned with an investigation of the hydrophobic interactions that are involved in the equilibria of proteins with non-polar surfaces. The strength of hydrogen bonds, as well as salt linkages, will increase in a non-aqueous environment so they have also been considered when studying these equilibria. The object of this work was to study not only the adsorption of proteins on hydrocarbon surfaces but also the interaction of hydrocarbon molecules with protein molecules in aqueous solution. The work was mostly confined to a study of extracellular proteins as it was found that intracellular proteins tend to unfold irreversibly at hydrocarbon/ water interfaces. Emulsions of hydrocarbons in water were used for studying the adsorption of proteins onto non-polar/aqueous interfaces since this provides the largest surface area for a given volume of hydrocarbon. When a solution of protein was equilibrated with decane an emulsion formed and it was found that a proportion of the protein was lost from solution; this reached a maximum with a given volume of decane and then remained constant. It was discovered that In the presence of alcohols the amount of protein lost from solution reached a maximum with a given volume of alcohol-in-decane solution and then decreased as the volume was further increased. This was thought to be governed by two equilibria; namely equilibrium of protein between aqueous solution and the interface, and equilibrium of alcohol between decane or aqueous solution and the interface. The alcohol and protein probably form a mixed film at the interface. The volume of decane solution at this maximum is called the "volume at the dip". In order to investigate this phenomenon in greater detail, the adsorption of various proteins on 0.25% octanol-in-decane (referred to as decane B') emulsions was investigated over a range of protein concentrations and pH values. It was observed with bovine serum albumin and lysozyme that, if the volume of decane B' at the dip was plotted against the initial concentration of protein, then it increased linearly until, at a certain concentration of protein, it became constant. This was connected with the fact that the percentage loss of protein from solution decreased once the volume at the dip became constant. Experiments with insulin and ribonuclease showed no levelling off in the volume of decane B' at the dip as the concentration of protein was increased. The explanation of these results is not clear. The effect of pH on the adsorption of proteins on decane B' emulsions was found to be related to the conformation or state of aggregation of the protein molecules. The N-F transition of bovine serum albumin, which occurs at pH 4, coincided with a change in the volume of decane B' at the dip; the F form requiring the volume of decane B' at the dip as compared to the N form. A further change in bovine serum albumin over the range pH 2 to pH 7 was indicated by the fact that the percentage loss of protein from solution at the dip rose gradually from 20% at pH 2 to 65% at pH 7. With lysozyme the percentage loss of protein from solution at the dip increased sharply over the range pH 5.5 to pH 6.0 where dimerization of the protein is known to occur. Similarly with insulin, the percentage loss of protein from solution is greater at pH 7 than at pH 2 owing to the increase in molecular weight from 12,000 to 36,000. Ribonuclease, which does not undergo any change in conformation or state of aggregation with pH, showed no change in its surface properties between pH 2 and pH 8. This relationship between properties of proteins in solution and in the presence of decane B' emulsions provides a sensitive method for comparing the solution properties of different proteins. An investigation of the difference spectra of proteins in equilibrium with emulsions showed that they undergo slight conformational changes but it appears that the extent of the change is related to the volume of decane B'. This suggests that the conformation of the protein must be altered a definite amount every time it comes into contact with the surface of an emulsion. It was found that the addition of an excess of alcohol displaced protein from an emulsion stabilized by protein; the difference spectrum of the protein solution so obtained was almost identical to that of the solution which had been originally in equilibrium with the emulsion. It was also found that lysozyme could be removed from emulsions by re-equilibrating the emulsion with fresh buffer at pH 2. Besides studying the adsorption of proteins on emulsions, the combination of decane with protein in aqueous solution was also investigated. When lysozyme, bovine serum albumin or insulin were equilibrated with a given volume of pure decane, then the number of moles of decane bound per mole of protein (^&minus;<sub style='position: relative; left: -0.5em;'>v</sub>) increased with decreasing concentration of protein; similar results were obtained when the binding of decanol to bovine serum albumin was investigated. This binding phenomenon is probably controlled by the ratio of the concentration of protein in solution to the volume of decane present i.e. the interfacial area of the deeane present. When decane B' was substituted for pure decane, then the binding again increased with decreasing concentration of protein but below a given concentration of protein it decreased dramatically. This result was shown to be linked to the fact that the amount of binding is affected by the amount of emulsion present, i.e. the maximum value of ^&minus;<sub style='position: relative; left: -0.5em;'>v</sub> for a given concentration of protein occurs when this concentration is equilibrated with the volume of decane B' at the dip corresponding to this concentration. The effect of pH on the binding of decane and decanol to bovine serum albumin is closely linked to the N-F transition and ^&minus;<sub style='position: relative; left: -0.5em;'>v</sub> was shown to increase between pH 3 and pH 7. The fact that lysozyme and insulin do not bind decane at 2&deg;C, but do at 22&deg;C, gives weight to Kauzmann's original precept, that the strength of hydrophobic interactions should increase with increasing temperature. When the percentage of octanol in decane was varied and different volumes of these solutions were added to a 2 x 10^-5M solution of lysozyme at pH 2.0 both the volume at the dip and the percentage loss of protein from solution at the dip remained constant. This effect was only observed when the percentage of octanol in decane was 0.2% or greater. However the sharpness of the dip increased with increasing percentage of octanol in the decane. When this experiment was repeated at pH 7.0, using bovine serum albumin, the results were very different since both the volume of decane solution at the dip, as well as the percentage loss of protein from solution at the dip, varied with the percentage of octanol in decane. The difference between these two sets of results probably lies in the fact that, with a 2 x 10^-5M solution of lysozyme at pH 2.0, the volume of decane B' at the dip was still increasing with increasing concentration of protein while, with a 1.1 x 10^-5M solution of bovine serum albumin at pH 7.0, the volume remained constant with increasing concentration of protein. The results obtained with bovine serum albumin provided further insight into the nature of the N-F transition since the amount of protein lost from solution and the volume of decane solution at the dip are identical for a 1.1 x 10^-5M solution of bovine serum albumin using a 0.8% octanol-in-decane solution at pH 7.0 or a 0.25% octanol-in-decane solution at pH 2.0. Thus the N form of bovine serum albumin requires approximately 3 times as much octanol at the interface to interact with the protein compared to the F form under similar conditions. This result is in good agreement with that described previously where the N form of bovine serum albumin required 3 times the volume of decane B' at the dip compared to the F form. These results indicate that a definite stoichiometry between octanol and protein molecules must exist at the interface. The effect of different alcohols on the relation between the volume of decane solution at the dip and the initial concentration of protein shows that these plots are different from those obtained using decane B'. The effect of secondary alcohols on emulsions stabilized with lysozyme was also investigated but their effect appeared to be similar to those of the homologous primary alcohols.