= Solution
\b[<NMR> of <membrane proteins> became practical through coordinated improvements in sample preparation, <isotope labelling>, relaxation control, instrumentation and solid-state methods.] The obstacle is not simply that <cell membranes> are insoluble. Hydrophobic surfaces must remain shielded, while the <protein> must retain function and form a sufficiently homogeneous sample. In solution, the <protein> plus its detergent or lipid assembly tumbles slowly; rapid <transverse relaxation> then broadens resonances until assignments and structural restraints become difficult to obtain.
Recombinant expression and controlled <isotope labelling> allow selective observation of the <protein> against its environment. Uniform <nitrogen-15> and <carbon-13> labelling supports multidimensional assignments. <deuteration> reduces much proton–proton dipolar relaxation; reintroducing selected protons or labelled methyl groups retains sensitive reporters while limiting the relaxation network. <selective methyl labelling> is particularly useful for large complexes, but the sparse reporters do not automatically supply enough restraints for a unique complete structure. Exchangeable amide protons may also need back-exchange after growth in deuterated medium, and buried sites can exchange incompletely. Sample production, rather than the spectrometer alone, is therefore a central technical development.
The <membrane mimic> must balance biological fidelity and spectroscopic tractability. Detergent <micelles> shield hydrophobic surfaces and can produce relatively small assemblies; inappropriate detergents may alter packing, activity or oligomeric state. <bicelles> combine bilayer-forming lipids with components that stabilize their edges, and <lipid bilayer nanodiscs> use a scaffold to enclose a bilayer patch. Smaller, well-defined <nanodiscs> reduce the additional rotational burden. <Franz Hagn> and colleagues engineered optimized small <nanodiscs> and obtained high-resolution <NMR> information on the <membrane protein> <OmpX>, connecting scaffold design with a more tractable lipid-embedded sample.<Optimized phospholipid bilayer nanodiscs facilitate high-resolution structure determination of membrane proteins by Franz Hagn and others>[Small scaffold design and OmpX structure determination]{ref} No mimic is automatically native: compare <ligand binding>, transport or catalytic activity with an appropriate <cell membrane> preparation, and check concentration-dependent aggregation and oligomerization.
<transverse relaxation-optimized spectroscopy> selects a slowly relaxing component of a coupled spin system. For suitable nuclei and fields, dipolar and <chemical shift anisotropy> contributions partly cancel for one component while reinforcing others. Combined with <deuteration>, this preserves narrower lines in systems that would otherwise be inaccessible. The original <TROSY> experiments established the relaxation-interference principle rather than merely increasing signal by averaging longer.<Attenuated T2 relaxation by mutual cancellation of dipole-dipole coupling and chemical shift anisotropy indicates an avenue to NMR structures of very large biological macromolecules in solution by Konstantin Pervushin and others>[Theory and experimental demonstration of relaxation interference]{ref} Methyl-based variants exploit favourable methyl relaxation properties and selective labelling. Increasing magnetic field improves sensitivity and frequency dispersion, but also changes relaxation contributions; it is not correct to assume that every resonance becomes narrower at every higher field.
High-field magnets, stable radiofrequency electronics and cryogenic detection probes increase accessible sensitivity and resolution. Three-dimensional and higher-dimensional experiments spread overlapping signals and connect resonances through known covalent pathways. Once assignments are secure, <nuclear Overhauser effects> provide distance information, while <residual dipolar couplings> and <paramagnetic relaxation enhancement> can add orientational and longer-range constraints. <chemical shift> perturbations report changes in local environment on <ligand binding>, but do not by themselves determine a contact geometry. Each additional measurement restricts the structural model in a different way, which is particularly valuable when only a sparse set of membrane-protein resonances remains visible.
<solid-state nuclear magnetic resonance> provides a complementary route that does not require rapid overall tumbling. <proteins> can be studied in bilayers or other immobilized assemblies, using pulse sequences and sample orientation or <magic angle spinning> to control anisotropic interactions. The magic angle satisfies $3\cos^2\theta-1=0$, giving $\theta\simeq54.74^\circ$. Spinning averages appropriate orientation-dependent terms, while decoupling, <isotope> dilution or <deuteration> and fast spinning reduce broadening from dense spin networks. Proton detection at very high spinning rates improves sensitivity; <Loren B. Andreas> and colleagues demonstrated structures of fully protonated <proteins> using fast-spinning proton-detected experiments, an important general methodological advance rather than a demonstration that every <cell membrane> sample is thereby solved.<Structure of fully protonated proteins by proton-detected magic-angle spinning NMR by Loren B. Andreas and others>[Proton detection with approximately 100 kHz magic-angle spinning]{ref} <dynamic nuclear polarization> can additionally enhance signal through electron-to-nuclear polarization transfer, although radicals and cryogenic conditions can affect line widths and the conformational states observed.
The outcome is both structural and functional information. Spectra can distinguish conformations, map <ligand> responses and measure exchange even when no single static structure describes the system. Empty <micelles> or scaffold-only <nanodiscs> help identify environmental signals; unlabelled samples, independent assignments and distance-restraint validation guard against incorrect peaks and overfitting. Compare preparations and activity before and after long acquisitions, because aggregation or degradation can selectively remove signals. A missing peak may reflect exchange broadening or overlap, not an absent residue. \b[The enabling combination is a functional and homogeneous <cell membrane> environment with an <isotope> and pulse-sequence strategy matched to the size and dynamics of the assembly; solution and solid-state <NMR> cover complementary regimes.]
Back to article page
