Structure determination of membrane proteins in their native phospholipid bilayer environment by rotationally aligned solid-state NMR spectroscopy

Abstract

One of the most important topics in experimental structural biology is determining the structures of membrane proteins. These structures represent one-third of all of the information expressed from a genome, distinguished by their locations within the phospholipid bilayer of cells, organelles, or enveloped viruses. Their highly hydrophobic nature and insolubility in aqueous media means that they require an amphipathic environment. They have unique functions in transport, catalysis, channel formation, and signaling. Researchers are particularly interested in G-protein coupled receptors (GPCRs) because they modulate many biological processes, and about half of the approximately 800 of these proteins within the human genome are or can be turned into drug receptors that affect a wide range of diseases. Because of experimental difficulties, researchers have studied membrane proteins using a wide variety of artificial media that mimic membranes, such as mixed organic solvents or detergents. More sophisticated mimics include bilayer discs (bicelles) and the lipid cubic phase (LCP), but both of these contain a very large detergent component, which can disrupt the stability and function of membrane proteins. To have confidence in the resulting structures and their biological functions and to avoid disrupting these delicate proteins, the structures of membrane proteins should be determined in their native environment of liquid crystalline phospholipid bilayers under physiological conditions. This Account describes a recently developed general method for determining the structures of unmodified membrane proteins in phospholipid bilayers by solid-state NMR spectroscopy. Because it relies on the natural, rapid rotational diffusion of these proteins about the bilayer normal, this method is referred to as rotationally aligned (RA) solid-state NMR. This technique elaborates on oriented sample (OS) solid-state NMR, its complementary predecessor. These methods exploit the power of solid-state NMR, which enables researchers to obtain well-resolved spectra from "immobile" membrane proteins in phospholipid bilayers, to separate and measure frequencies that reflect orientations with respect to the bilayer normal, and to make complementary distance measurements. The determination of the structures of several membrane proteins, most prominently the chemokine receptor CXCR1, a 350-residue GPCR, has demonstrated this approach.

Figure 1

Schematic three-dimensional and cross-sectional views of the “fluid mosaic model” of globular membrane proteins that are completely or partially embedded within a lipid matrix. (Modified from Singer and Nicholson)

Figure 2

¹³C solid-state NMR spectra of uniformly ¹³C/¹⁵N labeled MerFt in proteoliposomes. The majority of the resonance intensity shown in the expanded region, centered near 175 ppm, is from ¹³C′ backbone sites. The spectra in panels C. and G. were obtained at 25°C where the protein undergoes fast rotational diffusion. The spectra in panels D. and H. were obtained at 10°C) where the protein is immobile on the timescale of a static ¹³C′ chemical shift powder pattern (∼10⁵ Hz). The simulated spectra are for a single ¹³C′ group in a trans membrane helix aligned approximately parallel to the lipid bilayer normal. B. and F. are for a static protein. A. and E. are for a protein undergoing fast rotational diffusion about the bilayer normal. The experimental spectra in G. and H. were obtained from samples undergoing slow (5 kHz) magic angle spinning. The corresponding simulated spectra in E. and F. confirm that a family of sidebands is observable in the absence of rotational diffusion of the membrane protein.

Figure 3

Two-dimensional magic angle spinning spectra of uniformly ¹³C/¹⁵N labeled MerFt in DMPC proteoliposomes at 10°C. A. ¹³C/¹³C homonuclear spin-exchange correlation spectrum. B. ¹³C/¹⁵N heteronuclear correlation spectrum. The experiments were performed at 750 MHz under conditions of 11.11 kHz MAS.

Figure 4

Examples of spectroscopic data for residue L31 obtained from two- and three-dimensional NMR spectra under the same conditions as in Figure 3 except at 25oC. A. Two-dimensional ¹H-¹⁵N DC / ¹³C Shift SLF spectrum. B. Two-dimensional ¹H-¹⁵N DC / ¹³C Shift SLF spectral plane from a three-dimensional spectrum selected at the ¹⁵N isotropic chemical shift frequency of 118.6 ppm. C. ¹H-¹³C DC / ¹³C shift plane obtained at the same ¹⁵N isotropic chemical shift frequency of 118.6 ppm as Panel B. D. ¹H-¹³C DC / ¹H-¹⁵N DC correlation plane obtained at a ¹³C shift frequency of 54.6 ppm. All three spectral planes are associated with residue L31; the dashed line traces the correlation among the resonance frequencies. E, F, H, I, Simulated Pake-doublets for a static peptide bond. E. Rigid powder pattern for ¹H-¹⁵N dipolar coupling constant of 10.5 kHz. F. Motionally averaged powder pattern for ¹H-¹⁵N dipolar coupling constant of 9.4 kHz. G. Experimental one-dimensional ¹H-¹⁵N dipolar slice obtained from three-dimensional SLF experiment for ¹³C and ¹⁵N shifts at 54.8 ppm and 118.6 ppm respectively. H. Rigid powder pattern for ¹H-¹³C dipolar coupling constant of 22.68 kHz. I. Motionally averaged powder pattern for ¹H-¹³C dipolar coupling constant of 7.6 kHz. J. Experimental one-dimensional ¹H-¹³C dipolar slice obtained from three-dimensional SLF experiment for ¹³C and ¹⁵N shifts at 54.8 ppm and 118.6 ppm respectively.

Figure 5

Ribbon drawing of the three-dimensional structure of MerF. The two proximal mercury-binding sites are labeled with arrows. The scale bar corresponds to the 23 A thickness of the hydrocarbons in the lipid bilayers. (From Ref.)

Figure 6

The large conformational rearrangement of the N-terminal domain between the truncated (MerFt) construct (in magenta color) and the full-length (MerF) construct (in cyan color). The MerFt structure and MerF structure are aligned on the transmembrane region, and the N-terminal domains between the two structures show an orientation difference of close to 90°. (From Ref.)

Figure 7

Three-dimensional structure of CXCR1 in phospholipid bilayers under physiological conditions. Backbone representation of CXCR1 showing helices (TM1-TM7 and H8) in aqua, extracellular loops (ECL1-ECL3) in gray, and intracellular loops in blue (ICL1), green (ICL2) and red (ICL3). Note that ICL3, which is essential for activity, has a well-defined tertiary structure. Disulfide bonded Cys pairs (C30-C277; C110-C187) are shown as sticks.