Membrane Protein Structures in Lipid Bilayers; Small-Angle Neutron Scattering With Contrast-Matched Bicontinuous Cubic Phases

Abstract

This perspective describes advances in determining membrane protein structures in lipid bilayers using small-angle neutron scattering (SANS). Differentially labeled detergents with a homogeneous scattering length density facilitate contrast matching of detergent micelles; this has previously been used successfully to obtain the structures of membrane proteins. However, detergent micelles do not mimic the lipid bilayer environment of the cell membrane in vivo. Deuterated vesicles can be used to obtain the radius of gyration of membrane proteins, but protein-protein interference effects within the vesicles severely limits this method such that the protein structure cannot be modeled. We show herein that different membrane protein conformations can be distinguished within the lipid bilayer of the bicontinuous cubic phase using contrast-matching. Time-resolved studies performed using SANS illustrate the complex phase behavior in lyotropic liquid crystalline systems and emphasize the importance of this development. We believe that studying membrane protein structures and phase behavior in contrast-matched lipid bilayers will advance both biological and pharmaceutical applications of membrane-associated proteins, biosensors and food science.

Keywords: lipid cubic phase; membrane protein; peptide structure; self-assembly; small-angle scattering.

Figure 1

(A–C) SANS data showing H-MO and peptide encapsulation in 92% D-MO/8% H-MO (M-MO) at a range of peptide concentrations at 20°C in D₂O. (A,B) Gramicidin A (1.9 kDa) concentrations reflect 1.6, 3.1, and 5.3 mol% with respect to MO and at 40% v/v D₂O: diamond cubic ${\text{Q}}_{\text{II}}^{\text{D}}$ phase. (C) WALPS53 (5.9 kDa) at 0.2 and 0.4 mol% with respect to MO and at 35% v/v D₂O: gyroid cubic ${\text{Q}}_{\text{II}}^{\text{G}}$ phase. The SANS data was treated by first subtracting the MO cubic phase as background, then scaled to peptide concentration, and then a further incoherent background was subtracted. The calculated form factors were obtained using CRYSON. (D–G) were created using PyMOL and show cartoon representations of the secondary structures with side-chains shown as lines. They were obtained using PDB ID (D) 1MAG: helical dimer and (E) 1BDW: double helix. (F) WALPS53–1 dimer as generated using the iTasser server (Yang et al., 2015) and (G) WALPS53–2 (extended version of WALPS53–1).