Meaghan E. Ward

Proton-Detected Solid-State NMR Reveals Intramembrane Polar Networks in a Seven-Helical Transmembrane Protein Proteorhodopsin

We used high-resolution proton-detected multidimensional NMR to study the solvent-exposed parts of a seven-helical integral membrane proton pump, proteorhodopsin (PR). PR samples were prepared by growing the apoprotein on fully deuterated medium and reintroducing protons to solvent-accessible sites through exchange with protonated buffer. This preparation leads to NMR spectra with proton resolution down to ca. 0.2 ppm at fast spinning (28 kHz) in a protein back-exchanged at a level of 40%. Novel three-dimensional proton-detected chemical shift correlation spectroscopy allowed for the identification and resonance assignment of the solvent-exposed parts of the protein. Most of the observed residues are located at the membrane interface, but there are notable exceptions, particularly in helix G, where most of the residues are susceptible to H/D exchange. This helix contains Schiff base-forming Lys231, and many conserved polar residues in the extracellular half, such as Asn220, Tyr223, Asn224, Asp227, and Asn230. We proposed earlier that high mobility of the F-G loop may transiently expose a hydrophilic cavity in the extracellular half of the protein, similar to the one found in xanthorhodopsin. Solvent accessibility of helix G is in line with this hypothesis, implying that such a cavity may be a part of the proton-conducting pathway lined by this helix.

Conformational dynamics of a seven transmembrane helical protein Anabaena Sensory Rhodopsin probed by solid-state NMR.

The ability to detect and characterize molecular motions represents one of the unique strengths of nuclear magnetic resonance (NMR) spectroscopy. In this study, we report solid-state NMR site-specific measurements of the dipolar order parameters and (15)N rotating frame spin-lattice (R1ρ) relaxation rates in a seven transmembrane helical protein Anabaena Sensory Rhodopsin reconstituted in lipids. The magnitudes of the observed order parameters indicate that both the well-defined transmembrane regions and the less structured intramembrane loops undergo restricted submicrosecond time scale motions. In contrast, the R1ρ rates, which were measured under fast magic angle spinning conditions, vary by an order of magnitude between the TM and exposed regions and suggest the presence of intermediate time scale motions. Using a simple model, which assumes a single exponential autocorrelation function, we estimated the time scales of dominant stochastic motions to be on the order of low tens of nanoseconds for most residues within the TM helices and tens to hundreds of nanoseconds for the extracellular B-C and F-G loops. These relatively slow time scales could be attributed to collective anisotropic motions. We used the 3D Gaussian axial fluctuations model to estimate amplitudes, directions, and time scales of overall motions for helices and the extracellular B-C and F-G loops. Within this model, the TM helices A,B,C,D,E,F undergo rigid body motions on a time scale of tens of nanoseconds, while the time scale for the seventh helix G approaches 100 ns. Similar time scales of roughly 100-200 ns are estimated for the B-C and F-G loops.

Cysteine-Specific Labeling of Proteins with a Nitroxide Biradical for Dynamic Nuclear Polarization NMR

Dynamic nuclear polarization (DNP) enhances the signal in solid-state NMR of proteins by transferring polarization from electronic spins to the nuclear spins of interest. Typically, both the protein and an exogenous source of electronic spins, such as a biradical, are either codissolved or suspended and then frozen in a glycerol/water glassy matrix to achieve a homogeneous distribution. While the use of such a matrix protects the protein upon freezing, it also reduces the available sample volume (by ca. a factor of 4 in our experiments) and causes proportional NMR signal loss. Here we demonstrate an alternative approach that does not rely on dispersing the DNP agent in a glassy matrix. We synthesize a new biradical, ToSMTSL, which is based on the known DNP agent TOTAPOL, but also contains a thiol-specific methanethiosulfonate group to allow for incorporating this biradical into a protein in a site-directed manner. ToSMTSL was characterized by EPR and tested for DNP of a heptahelical transmembrane protein, Anabaena sensory rhodopsin (ASR), by covalent modification of solvent-exposed cysteine residues in two (15)N-labeled ASR mutants. DNP enhancements were measured at 400 MHz/263 GHz NMR/EPR frequencies for a series of samples prepared in deuterated and protonated buffers and with varied biradical/protein ratios. While the maximum DNP enhancement of 15 obtained in these samples is comparable to that observed for an ASR sample cosuspended with ~17 mM TOTAPOL in a glycerol-d8/D2O/H2O matrix, the achievable sensitivity would be 4-fold greater due to the gain in the filling factor. We anticipate that the DNP enhancements could be further improved by optimizing the biradical structure. The use of covalently attached biradicals would broaden the applicability of DNP NMR to structural studies of proteins.

Advanced solid-state NMR techniques for characterization of membrane protein structure and dynamics: Application to Anabaena Sensory Rhodopsin

Studies of the structure, dynamics, and function of membrane proteins (MPs) have long been considered one of the main applications of solid-state NMR (SSNMR). Advances in instrumentation, and the plethora of new SSNMR methodologies developed over the past decade have resulted in a number of high-resolution structures and structural models of both bitopic and polytopic α-helical MPs. The necessity to retain lipids in the sample, the high proportion of one type of secondary structure, differential dynamics, and the possibility of local disorder in the loop regions all create challenges for structure determination. In this Perspective article we describe our recent efforts directed at determining the structure and functional dynamics of Anabaena Sensory Rhodopsin, a heptahelical transmembrane (7TM) protein. We review some of the established and emerging methods which can be utilized for SSNMR-based structure determination, with a particular focus on those used for ASR, a bacterial protein which shares its 7TM architecture with G-protein coupled receptors.

Proton detection for signal enhancement in solid-state NMR experiments on mobile species in membrane proteins.

Direct proton detection is becoming an increasingly popular method for enhancing sensitivity in solid-state nuclear magnetic resonance spectroscopy. Generally, these experiments require extensive deuteration of the protein, fast magic angle spinning (MAS), or a combination of both. Here, we implement direct proton detection to selectively observe the mobile entities in fully-protonated membrane proteins at moderate MAS frequencies. We demonstrate this method on two proteins that exhibit different motional regimes. Myelin basic protein is an intrinsically-disordered, peripherally membrane-associated protein that is highly flexible, whereas Anabaena sensory rhodopsin is composed of seven rigid transmembrane α-helices connected by mobile loop regions. In both cases, we observe narrow proton linewidths and, on average, a 10× increase in sensitivity in 2D insensitive nuclear enhancement of polarization transfer-based HSQC experiments when proton detection is compared to carbon detection. We further show that our proton-detected experiments can be easily extended to three dimensions and used to build complete amino acid systems, including sidechain protons, and obtain inter-residue correlations. Additionally, we detect signals which do not correspond to amino acids, but rather to lipids and/or carbohydrates which interact strongly with membrane proteins.

Oligomeric Structure of Anabaena Sensory Rhodopsin in a Lipid Bilayer Environment by Combining Solid-State NMR and Long-range DEER Constraints

Oligomerization of membrane proteins is common in nature. Here, we combine spin-labeling double electron-electron resonance (DEER) and solid-state NMR (ssNMR) spectroscopy to refine the structure of an oligomeric integral membrane protein, Anabaena sensory rhodopsin (ASR), reconstituted in a lipid environment. An essential feature of such a combined approach is that it provides structural distance restraints spanning a range of ca 3-60Å while using the same sample preparation (i.e., mutations, paramagnetic labeling, and reconstitution in lipid bilayers) for both ssNMR and DEER. Direct modeling of the multispin effects on DEER signal allowed for the determination of the oligomeric order and for obtaining long-range DEER distance restraints between the ASR trimer subunits that were used to refine the ssNMR structure of ASR. The improved structure of the ASR trimer revealed a more compact packing of helices and side chains at the intermonomer interface, compared to the structure determined using the ssNMR data alone. The extent of the refinement is significant when compared with typical helix movements observed for the active states of homologous proteins. Our combined approach of using complementary DEER and NMR measurements for the determination of oligomeric structures would be widely applicable to membrane proteins where paramagnetic tags can be introduced. Such a method could be used to study the effects of the lipid membrane composition on protein oligomerization and to observe structural changes in protein oligomers upon drug, substrate, and co-factor binding.