Measuring membrane protein bond orientations in nanodiscs via residual dipolar couplings

Abstract

Membrane proteins are involved in numerous vital biological processes. To understand membrane protein functionality, accurate structural information is required. Usually, structure determination and dynamics of membrane proteins are studied in micelles using either solution state NMR or X-ray crystallography. Even though invaluable information has been obtained by this approach, micelles are known to be far from ideal mimics of biological membranes often causing the loss or decrease of membrane protein activity. Recently, nanodiscs, which are composed of a lipid bilayer surrounded by apolipoproteins, have been introduced as a more physiological alternative than micelles for NMR investigations on membrane proteins. Here, we show that membrane protein bond orientations in nanodiscs can be obtained by measuring residual dipolar couplings (RDCs) with the outer membrane protein OmpX embedded in nanodiscs using Pf1 phage as an alignment medium. The presented collection of membrane protein RDCs in nanodiscs represents an important step toward more comprehensive structural and dynamical NMR-based investigations of membrane proteins in a natural bilayer environment.

Keywords: COCAINE; NMR; Pf1 phage; TROSY; alignment medium; membrane protein; nanodisc; residual dipolar coupling.

Figure 1

Schematic drawing of OmpX-loaded nanodiscs aligned with Pf1 phages. Two Pf1 phages are indicated as yellow cylinders aligned along the labeled static magnetic field of the magnet of the NMR spectrometer. A presumed model of the nanodisc composed of the cyan and pink colored apolipoprotein ApoA-I, the lipid bilayer colored in gray, and OmpX in green are shown. The x/y/z coordinate system is for each nanodisc indicated highlighting a preferred orientation of the nanodisc and concomitantly the embedded membrane protein OmpX.

Figure 2

RDCs measured in the NMR spectra of deuterated ¹⁵N,¹³C-labeled OmpX reconstituted in nanodiscs (at a temperature of 318 K in 10 mM Tris–HCl, 100 mM NaCl, 1 mM EDTA, pH 7.4). [¹⁵N,¹H]-TROSY (solid lines) and [¹⁵N,¹H]-COCAINE (dashed lines) spectra were recorded under (A) isotropic conditions and (B) aligned conditions, which was established by adding 10 mg/mL Pf1 phages into the isotropic sample. In (A) 0.5*¹J(HN,N) scalar couplings and (B) half of the sum of the scalar coupling and the RDCs of the backbone amide bonds are labeled, respectively. To alleviate severe line broadening due to the size of the system, the experiments were set up in such a manner that the peak position difference between the two spectra give rise to only half of the size of the couplings, that is, scalar and dipolar couplings using [¹⁵N,¹H]-COCAINE.

Figure 3

Correlation plots between the observed backbone amide bond RDCs and the calculated RDCs from the structures of OmpX determined (A) without (PDB accession code: 2M06) and (B) with RDCs (PDB accession code: 2MNH). (C) Comparison of the 3D structure of OmpX calculated without (blue) and with (green) RDCs. The two structures are represented each by a bundle that fulfills best the experimental input restraints. By introducing the observed backbone amide bond RDCs into the structure determination protocol, the orientations of the backbone amide bonds are significantly improved (compare A with B), as evidenced quantitatively with the Pearson correlation coefficient (r) and the quality factor Q (Q = rms[observed RDC − calculated RDC]/rms[observed RDC]) listed in A and B.