Abstract
Small angle X-ray scattering (SAXS) experiments are performed on two non-ionic surfactants, the dodecyl β-maltoside (DDβM) and the propyl(bi)cyclohexyl α-malto-side (PCCαM), a maltoside derivative containing a rigid bicyclohexyl group as hydrophobic chain, in order to compare the influence of both hydrophobic moiety structure and anomeric form on micelle form factors and intermicellar interactions relevant for membrane protein crystallization. Density and refractive index measurements were performed in order to determine volumetric and optical properties of surfactants, essential for determination of micelle molar masses by both SAXS and SEC-MALLS. SAXS form factors were analyzed by Guinier approximation and inverse Fourier transformation, to obtain the radius of gyration (R G) and the pair distribution function (P(r)) of each surfactant. Form factor model fitting was also performed to describe the shape and the assembly of both surfactant micelles. Finally, second virial coefficients were measured at different percentages of polyethylene glycol 3350, in order to correlate surfactant intermicellar interactions and RC-LH1-PufX phase diagram. It is thus found that while size, shape, and dimensions of micelles are slightly similar for both surfactants, their molar mass and aggregation number differ significantly. PCCαM are more densely packed than DDβM, which reflects (1) an increase in van der Waals contacts between PCCαM hydrophobic chains in the micelle bulk and (2) a supplementary intermicellar attraction compared to DDβM. Finally addition of PEG, which induces a depletion attraction, decreases the solubility of the RC-LH1-PufX complex in PCCαM. ■ INTRODUCTION Membrane proteins (MPs) play a fundamental role in biology as they are at the heart of all communications and transfers between the inside of living cells and their immediate environment. They represent roughly 30% of the proteome of E. coli or humans and 60% of current drug targets. Despite their importance in many cellular processes, knowledge of their structure and details of their molecular mechanisms remain sketchy at bestPMs represent about 2% of known structures in the Protein Data Bank. This lack of structural information is related to a number of difficulties: production of sufficient quantities, purification while conserving the structure/function/ activity, and 3D crystallization for atomic resolution by crystallography. Regardless of how they are produced, MPs are usually purified using standard biochemical techniques, but require the use of detergents, surfactants which solubilize membrane lipids and make possible handling of membrane proteins in a hydrophobic environment. While increasing successes are obtained using in meso crystallization methods, i.e., by reconstitution in lipidic environment after solubilization and purification in detergents, growing high diffracting MP crystals directly in surfactant micellesin surfoa long-established method, remains a challenge in structural biology. Despite the use of high-throughput robotic platforms in combination with commercial or lab-made crystallization screens, crystallization by these trial-and-error methods is not frequently successful. Moving beyond empirical approaches requires a better understanding of the physics of these complex solutions containing both surfactant-solubilized MPs and surfactant micelles, whether it be for the supersaturation or the crystal growth step. In the 1990s, fundamental approaches have proven highly successful for the crystallization of soluble proteins. 1 Based on the study of interaction potentials between macromolecules in solution and second virial coefficient (A 2) measurements, 2,3 knowledge of crystallization mechanisms has progressed greatly thanks to a better understanding and control of physicochemical parameters that govern solution properties during supersaturation and crystal growth. 4 It has thus been shown that proteins crystallize in a regime where protein− protein interactions are attractive (or at least not strongly