| The subject of this thesis is characterization of protein-surfactant interactions, with particular emphasis on integral membrane proteins. Integral membrane proteins are an important, poorly understood class of proteins involved in a wide range of biological processes. Their hydrophobic character renders them insoluble and makes their characterization difficult. It also necessitates use of surfactants to isolate them for further study. By developing a physical basis for many observed phenomena of membrane protein solubilization, structure, activity and crystallization, insight was gained into how surfactant influences each of these processes. This insight leads to several suggestions on how to improve methods for membrane protein characterization, the utility of which are illustrated in this thesis.; Initially, the nature of protein-surfactant interactions were examined during solubilization of the human adenosine A3 receptor, human adenosine A2a receptor, E. coli diacylglycerol kinase and H. halobium bacteriorhodopsin using a wide range of surfactants. A striking correlation between activity or recovery of each was found with the hydrophile-lipophile balance (HLB) number of the surfactants, in contrast to other common surfactant properties such as critical micelle concentration (CMC), with an unique optimal HLB range for each. HLB is an empirical number that describes the hydrophobicity of a surfactant based on its structure, and has been used for many years to guide formulation of stable emulsions. Theories of non-ionic surfactant emulsions relate HLB to the geometry a surfactant aggregate assumes, notably the molecular packing parameter (P). Our results indicate a similar optimum with packing parameter, suggesting optimal activity for membrane proteins are a consequence of appropriate surfactant packing in a PDC.; Based on this idea of surfactant packing in a PDC, we developed a correlation between the HLB of surfactants used in crystallization of several membrane proteins and their lipid-exposed transmembrane surface area (TMSA). Surprisingly, this correlation, while strong, implies spherical packing with increasing surface area. Therefore, additional factors beyond surfactant packing are necessary to explain optimal surfactants for membrane protein crystallization. In order to determine TMSA for membrane proteins of unknown structure, we developed a correlation with TMSA based on molecular weight. The fact that this correlation is linear in MW suggests, in contrast to soluble proteins, that membrane proteins are significantly non-spherical. The correlations were successfully used to predict optimal surfactants for the human adenosine A 2a receptor without a need for extensive screening, which points to the validity of our results. (Abstract shortened by UMI.)... |
