Novel zwitterionic reverse micelles for encapsulation of proteins in low-viscosity media

Large proteins remain inaccessible to structural NMR studies because of their unfavorable relaxation properties. Their solubilization in the aqueous core of reverse micelles, in a low-viscosity medium, represents a promising approach, provided that their native tertiary structure is maintained. However, the use of classical ionic surfactants may lead to protein unfolding, due to strong electrostatic interactions between the polar head groups and the protein charges. To design reverse micelles in which these interactions are weakened, a new zwitterionic surfactant molecule was synthesized and studied by high-resolution NMR spectroscopy, for which cytochrome C and 15N-labeled ubiquitin were used as guest candidates. At different ionization states, both proteins are encapsulated in the absence of salts or other additives, in a folded conformation similar to the native one.     

New reverse micelle surfactant systems optimized for high-resolution NMR spectroscopy of encapsulated proteins

Sodium bis(2-ethylhexyl)sulfosuccinate (AOT) is a surfactant commonly used to encapsulate water soluble proteins within the aqueous core of a reverse micelle. In the context of high-resolution NMR studies of encapsulated proteins the size of the resulting reverse micelle is critically important. We have designed and synthesized a short AOT analogue, 3,3-dimethyl-1-butylsulfosuccinate sodium salt and determined that it is able to form reverse micelles and to encapsulate the protein ubiquitin with high structural fidelity. AOT is often found to significantly destabilize encapsulated proteins, largely through charge-charge interactions between the anionic headgroup and the surface of the protein. Here we demonstrate, for the first time, that proportional mixtures of anionic and cationic surfactants can form reverse micelles that are also capable of protein encapsulation with high fidelity.     

Simulations of the confinement of ubiquitin in self-assembled reverse micelles

We describe the effects of confinement on the structure, hydration, and the internal dynamics of ubiquitin encapsulated in reverse micelles (RM). We performed molecular dynamics simulations of the encapsulation of ubiquitin into self-assembled protein/surfactant reverse micelles to study the positioning and interactions of the protein with the RM and found that ubiquitin binds to the RM interface at low salt concentrations. The same hydrophobic patch that is recognized by ubiquitin binding domains in vivo is found to make direct contact with the surfactant head groups, hydrophobic tails, and the iso-octane solvent. The fast backbone N-H relaxation dynamics show that the fluctuations of the protein encapsulated in the RM are reduced when compared to the protein in bulk. This reduction in fluctuations can be explained by the direct interactions of ubiquitin with the surfactant and by the reduced hydration environment within the RM. At high concentrations of excess salt, the protein does not bind strongly to the RM interface and the fast backbone dynamics are similar to that of the protein in bulk. Our simulations demonstrate that the confinement of protein can result in altered protein dynamics due to the interactions between the protein and the surfactant.     

Encapsulation and NMR on an aggregating peptide before fibrillogenesis

The early stages of peptide and protein aggregation include the formation of soluble oligomers, some of which may be cytotoxic. There is a paucity of structural information on these oligomers, however, because they are temporally unstable and tend to aggregate further into insoluble protofibrils and fibrils. To obtain structural information on soluble oligomers, we have developed a procedure for encapsulating a fibril-forming peptide, Peptide 1 (NH2-SDDYYYGFGSNKFGRPRDD-COOH), in 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine single bilayer vesicles (POPC SBVs). We also encapsulated a non-fibril forming peptide, Peptide 2 (NH2-EEWEE-COOH), in POPC SBVs. The nominal concentration of Peptide 1 in the resulting 40 nm diameter SBVs was 2.4 +/- 0.1 mM, well above the concentration at which Peptide 1 forms fibrils. We demonstrated that these peptides had indeed been encapsulated by measuring longitudinal relaxation times (T1) in the presence and absence of a paramagnetic substance, 1 mM Gd-EDTA, by NMR spectroscopy. When the peptides were free in solution, they showed the expected shortening of T1 times and broadening of NMR peaks. In contrast, peptide encapsulated in POPC SBVs were shielded from the effects of Gd-EDTA and showed preservation of T1 values and NMR line widths. To demonstrate that encapsulation inhibits fibril formation, we measured one-dimensional proton (1D-1H) NMR spectra of the peptides in solution, and of the encapsulated peptides immediately after encapsulation, and 4 days after encapsulation, because Peptide 1 forms fibrils within 1 day. A 2.8 mM solution of Peptide 1 shows the loss of NMR signal expected for a fibrillizing peptide. In contrast, the 1D-1H spectra of encapsulated Peptide 1 measured immediately after encapsulation and 4 days after encapsulation were essentially identical, with preservation of line width at 4 days, i.e., well within the time frame of most high-resolution NMR experiments. Encapsulation may provide a means to obtain high-resolution NMR data on unstable soluble oligomers of peptides implicated in amyloidoses such as Alzheimer's Disease and provide the first detailed structural information about these possibly cytotoxic species that have hitherto been inaccessible to analysis.     

Determinations of 15N chemical shift anisotropy magnitudes in a uniformly 15N,13C-labeled microcrystalline protein by three-dimensional magic-angle spinning nuclear magnetic resonance spectroscopy

Amide 15N chemical shift anisotropy (CSA) tensors provide quantitative insight into protein structure and dynamics. Experimental determinations of 15N CSA tensors in biologically relevant molecules have typically been performed by NMR relaxation studies in solution, goniometric analysis of single-crystal spectra, or slow magic-angle spinning (MAS) NMR experiments of microcrystalline samples. Here we present measurements of 15N CSA tensor magnitudes in a protein of known structure by three-dimensional MAS solid-state NMR. Isotropic 15N, 13C alpha, and 13C' chemical shifts in two dimensions resolve site-specific backbone amide recoupled CSA line shapes in the third dimension. Application of the experiments to the 56-residue beta1 immunoglobulin binding domain of protein G (GB1) enabled 91 independent determinations of 15N tensors at 51 of the 55 backbone amide sites, for which 15N-13C alpha and/or 15N-13C' cross-peaks were resolved in the two-dimensional experiment. For 37 15N signals, both intra- and interresidue correlations were resolved, enabling direct comparison of two experimental data sets to enhance measurement precision. Systematic variations between beta-sheet and alpha-helix residues are observed; the average value for the anisotropy parameter, delta (delta = delta(zz) - delta(iso)), for alpha-helical residues is 6 ppm greater than that for the beta-sheet residues. The results show a variation in delta of 15N amide backbone sites between -77 and -115 ppm, with an average value of -103.5 ppm. Some sites (e.g., G41) display smaller anisotropy due to backbone dynamics. In contrast, we observe an unusually large 15N tensor for K50, a residue that has an atypical, positive value for the backbone phi torsion angle. To our knowledge, this is the most complete experimental analysis of 15N CSA magnitude to date in a solid protein. The availability of previous high-resolution crystal and solution NMR structures, as well as detailed solid-state NMR studies, will enhance the value of these measurements as a benchmark for the development of ab initio calculations of amide 15N shielding tensor magnitudes.     

Eta(z)/kappa: a transverse relaxation optimized spectroscopy NMR experiment measuring longitudinal relaxation interference

NMR spin relaxation experiments provide a powerful tool for the measurement of global and local biomolecular rotational dynamics at subnanosecond time scales. Technical limitations restrict most spin relaxation studies to biomolecules weighing less than 10 kDa, considerably smaller than the average protein molecular weight of 30 kDa. In particular, experiments measuring eta(z), the longitudinal (1)H(N)-(15)N dipole-dipole (DD)/(15)N chemical shift anisotropy (CSA) cross-correlated relaxation rate, are among those least suitable for use with larger biosystems. This is unfortunate because these experiments yield valuable insight into the variability of the (15)N CSA tensor over the polypeptide backbone, and this knowledge is critical to the correct interpretation of most (15)N-NMR backbone relaxation experiments, including R(2) and R(1). In order to remedy this situation, we present a new (1)H(N)-(15)N transverse relaxation optimized spectroscopy experiment measuring eta(z) suitable for applications with larger proteins (up to at least 30 kDa). The presented experiment also yields kappa, the site-specific rate of longitudinal (1)H(N)-(1)H(') DD cross relaxation. We describe the eta(z)/kappa experiment's performance in protonated human ubiquitin at 30.0 degrees C and in protonated calcium-saturated calmodulin/peptide complex at 20.0 degrees C, and demonstrate preliminary experimental results for a deuterated E. coli DnaK ATPase domain construct at 34 degrees C.     

Quantitative analysis of protein backbone dynamics in microcrystalline ubiquitin by solid-state NMR spectroscopy

Characterization of protein dynamics by solid-state NMR spectroscopy requires robust and accurate measurement protocols, which are not yet fully developed. In this study, we investigate the backbone dynamics of microcrystalline ubiquitin using different approaches. A rotational-echo double-resonance type (REDOR-type) methodology allows one to accurately measure (1)H-(15)N order parameters in highly deuterated samples. We show that the systematic errors in the REDOR experiment are as low as 1% or even less, giving access to accurate data for the amplitudes of backbone mobility. Combining such dipolar-coupling-derived order parameters with autocorrelated and cross-correlated (15)N relaxation rates, we are able to quantitate amplitudes and correlation times of backbone dynamics on picosecond and nanosecond time scales in a residue-resolved manner. While the mobility on picosecond time scales appears to have rather uniform amplitude throughout the protein, we unambiguously identify and quantitate nanosecond mobility with order parameters S(2) as low as 0.8 in some regions of the protein, where nanosecond dynamics has also been revealed in solution state. The methodology used here, a combination of accurate dipolar-coupling measurements and different relaxation parameters, yields details about dynamics on different time scales and can be applied to solid protein samples such as amyloid fibrils or membrane proteins.     

Temperature dependence of anisotropic protein backbone dynamics

The measurement of (15)N NMR spin relaxation, which reports the (15)N-(1)H vector reorientational dynamics, is a widely used experimental method to assess the motion of the protein backbone. Here, we investigate whether the (15)N-(1)H vector motions are representative of the overall backbone motions, by analyzing the temperature dependence of the (15)N-(1)H and (13)CO-(13)C(alpha) reorientational dynamics for the small proteins binase and ubiquitin. The latter dynamics were measured using NMR cross-correlated relaxation experiments. The data show that, on average, the (15)N-(1)H order parameters decrease only by 2.5% between 5 and 30 degrees C. In contrast, the (13)CO-(13)C(alpha) order parameters decrease by 10% over the same temperature trajectory. This strongly indicates that there are polypeptide-backbone motions activated at room temperature that are not sensed by the (15)N-(1)H vector. Our findings are at variance with the common crank-shaft model for protein backbone dynamics, which predicts the opposite behavior. This study suggests that investigation of the (15)N relaxation alone would lead to underestimation of the dynamics of the protein backbone and the entropy contained therein.     

De novo determination of bond orientations and order parameters from residual dipolar couplings with high accuracy

We report the de novo determination of 15N-1H bond orientations and motional order parameters for the protein ubiquitin with high accuracy based solely on NMR residual dipolar coupling measurements made in six distinct alignment media. The resulting bond orientations are in agreement with RDC-refined orientations of either solid or solution state coordinates to within approximately 2 degrees , which is also the estimated precision of the resulting orientations. The squared generalized order parameters, which reflect amplitudes of motion spanning the picosecond to millisecond time scales, exhibit a correlation with picosecond time scale order parameters derived from conventional NMR 15N spin relaxation methods. Provided that RDC measurements can be obtained using many different alignment media, this approach (called direct interpretation of dipolar couplings) may significantly impact the attainable accuracy and the molecular weight range accessible to NMR structure determination in the solution state, as well as provide a route for the study of motions occurring on the nanosecond to microsecond time scales, which have been traditionally difficult to study at atomic resolution.     

Impact of 15N R2/R1 relaxation restraints on molecular size, shape, and bond vector orientation for NMR protein structure determination with sparse distance restraints

(15)N R(2)/R(1) relaxation data contain information on molecular shape and size as well as on bond vector orientations relative to the diffusion tensor. Since the diffusion tensor can be directly calculated from the molecular coordinates, direct inclusion of (15)N R(2)/R(1) restraints in NMR structure calculations without any a priori assumptions is possible. Here we show that (15)N R(2)/R(1) restraints are particularly valuable when only sparse distance restraints are available. Using three examples of proteins of varying size, namely, GB3 (56 residues), ubiquitin (76 residues), and the N-terminal domain of enzyme I (EIN, 249 residues), we show that incorporation of (15)N R(2)/R(1) restraints results in large and significant increases in coordinate accuracy that can make the difference between being able or unable to determine an approximate global fold. For GB3 and ubiquitin, good coordinate accuracy was obtained using only backbone hydrogen-bond restraints supplemented by (15)N R(2)/R(1) relaxation restraints. For EIN, the global fold could be determined using sparse nuclear Overhauser enhancement (NOE) distance restraints involving only NH and methyl groups in conjunction with (15)N R(2)/R(1) restraints. These results are of practical significance in the study of larger and more complex systems, where the increasing spectral complexity and number of chemical shift degeneracies reduce the number of unambiguous NOE assignments that can be readily obtained, resulting in progressively reduced NOE coverage as the size of the protein increases.     

Hyperbranched‐upon‐dendritic macromolecules as unimolecular hosts for controlled release

Novel amphiphilic hyperbranched‐upon‐dendritic polymers with a dendritic polyester core, a linear poly(ε‐caprolactone) (PCL) inner shell, and a hyperbranched polyglycerol outer shell have been prepared. The structures of the hyperbranched‐upon‐dendritic polymers were characterized by using NMR spectra. The critical aggregating concentrations (CACs) of those amphiphilic hyperbranched‐upon‐dendritic polymers were measured by using pyrene as the polarity probe. To study the encapsulation performances of those hyperbranched‐upon‐dendritic polymers as unimolecular hosts, inter‐molecular encapsulation was carefully prevented by controlling the host concentrations below their CACs and by washing with good organic solvents. The study on encapsulation of two model guest molecules, pyrene and indomethacin, was performed. The amounts of encapsulated molecules were dependent mainly on the size of inner linear shells. About three pyrene molecules or five indomethacin molecules were encapsulated in hyperbranched‐upon‐dendritic polymers with average PCL repeating units of two but different hyperbranched polyglycerol outer shells, whereas about five pyrene molecules or about 12 indomethacin molecules were encapsulated in those with PCL repeating units of nine. The encapsulated molecules could be released in a controlled manner. Thus, the hyperbranched‐upon‐dendritic polymers could be used as unimolecular nanocarriers with controllable molecular encapsulation dosage for controlled release. © 2010 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 48: 4013–4019, 2010     

Quantitative measurement of differential 15N-H(alpha/beta)T2 relaxation rates in a perdeuterated protein by MAS solid-state NMR spectroscopy

Dynamic parameters become more and more accessible in the study of uniformly isotopically enriched proteins by MAS solid-state NMR. We demonstrate that T(2)-related relaxation properties can quantitatively be determined in a sample of a perdeuterated microcrystalline protein by the measurement of (15)N,(1)H dipole, (15)N CSA cross-correlated relaxation rates. We find that the measured cross-correlated relaxation rates are independent of the MAS rotation frequency, and therefore reflect local dynamic fluctuations of the protein structure.     

An exchange-free measure of 15N transverse relaxation: an NMR spectroscopy application to the study of a folding intermediate with pervasive chemical exchange

A series of experiments are presented that provide an exchange-free measure of dipole-dipole (15)N transverse relaxation, R(dd), that can then be substituted for (15)N R(1rho) or R(2) rates in the study of internal protein dynamics. The method is predicated on the measurement of a series of relaxation rates involving (1)H-(15)N longitudinal order, anti-phase (1)H and (15)N single-quantum coherences, and (1)H-(15)N multiple quantum coherences; the relaxation rates of all coherences are measured under conditions of spin-locking. Results from detailed simulations and experiments on a number of protein systems establish that R(dd) values are independent of exchange and systematic errors from dipolar interactions with proximal protons are calculated to be less than 1-2%, on average, for applications to perdeuterated proteins. Simulations further indicate that the methodology is rather insensitive to the exact level of deuteration so long as proteins are reasonably highly deuterated (>50%). The utility of the methodology is demonstrated with applications involving protein L, ubiquitin, and a stabilized folding intermediate of apocytochrome b(562) that shows large contributions to (15)N R(1rho) relaxation from chemical exchange.     

Accurate measurement of longitudinal cross-relaxation rates in nuclear magnetic resonance

The accuracy of the determination of longitudinal cross-relaxation rates in NMR can be improved by combining symmetrical reconversion with suitable operator swapping methods that lead to the averaging of differences in autorelaxation rates and eliminate the effects of cross relaxation with the environment. The principles are first discussed for an isolated two-spin system comprising a pair of 15N and 1HN nuclei subjected to chemical shift anisotropy and dipole-dipole relaxation, and then extended to include further protons. The gains in accuracy are demonstrated experimentally for the protein ubiquitin.     

High-resolution NMR studies of encapsulated proteins in liquid ethane

Many of the difficulties presented by large, aggregation-prone, and membrane proteins to modern solution NMR spectroscopy can be alleviated by actively seeking to increase the effective rate of molecular reorientation. An emerging approach involves encapsulating the protein of interest within the protective shell of a reverse micelle and dissolving the resulting particle in a low viscosity fluid, such as the short chain alkanes. Here we present the encapsulation of proteins with high structural fidelity within reverse micelles dissolved in liquid ethane. The addition of appropriate cosurfactants can significantly reduce the pressure required for successful encapsulation. At these reduced pressures, the viscosity of the ethane solution is low enough to provide sufficiently rapid molecular reorientation to significantly lengthen the spin-spin NMR relaxation times of the encapsulated protein.     

Strategy for the study of paramagnetic proteins with slow electronic relaxation rates by nmr spectroscopy: application to oxidized human [2Fe-2S] ferredoxin

NMR studies of paramagnetic proteins are hampered by the rapid relaxation of nuclei near the paramagnetic center, which prevents the application of conventional methods to investigations of the most interesting regions of such molecules. This problem is particularly acute in systems with slow electronic relaxation rates. We present a strategy that can be used with a protein with slow electronic relaxation to identify and assign resonances from nuclei near the paramagnetic center. Oxidized human [2Fe-2S] ferredoxin (adrenodoxin) was used to test the approach. The strategy involves six steps: (1) NMR signals from (1)H, (13)C, and (15)N nuclei unaffected or minimally affected by paramagnetic effects are assigned by standard multinuclear two- and three-dimensional (2D and 3D) spectroscopic methods with protein samples labeled uniformly with (13)C and (15)N. (2) The very broad, hyperfine-shifted signals from carbons in the residues that ligate the metal center are classified by amino acid and atom type by selective (13)C labeling and one-dimensional (1D) (13)C NMR spectroscopy. (3) Spin systems involving carbons near the paramagnetic center that are broadened but not hyperfine-shifted are elucidated by (13)C[(13)C] constant time correlation spectroscopy (CT-COSY). (4) Signals from amide nitrogens affected by the paramagnetic center are assigned to amino acid type by selective (15)N labeling and 1D (15)N NMR spectroscopy. (5) Sequence-specific assignments of these carbon and nitrogen signals are determined by 1D (13)C[(15)N] difference decoupling experiments. (6) Signals from (1)H nuclei in these spin systems are assigned by paramagnetic-optimized 2D and 3D (1)H[(13)C] experiments. For oxidized human ferredoxin, this strategy led to assignments (to amino acid and atom type) for 88% of the carbons in the [2Fe-2S] cluster-binding loops (residues 43-58 and 89-94). These included complete carbon spin-system assignments for eight of the 22 residues and partial assignments for each of the others. Sequence-specific assignments were determined for the backbone (15)N signals from nine of the 22 residues and ambiguous assignments for five of the others.     

Dynamics of lysine side-chain amino groups in a protein studied by heteronuclear 1H−15N NMR spectroscopy

Despite their importance in macromolecular interactions and functions, the dynamics of lysine side-chain amino groups in proteins are not well understood. In this study, we have developed the methodology for the investigations of the dynamics of lysine NH3(+) groups by NMR spectroscopy and computation. By using 1H?15N heteronuclear correlation experiments optimized for 15NH3(+) moieties, we have analyzed the dynamic behavior of individual lysine NH3(+) groups in human ubiquitin at 2 °C and pH 5. We modified the theoretical framework developed previously for CH3 groups and used it to analyze 15N relaxation data for the NH3(+) groups. For six lysine NH3(+) groups out of seven in ubiquitin, we have determined model-free order parameters, correlation times for bond rotation, and reorientation of the symmetry axis occurring on a pico- to nanosecond time scale. From CPMG relaxation dispersion experiment for lysine NH3(+) groups, slower dynamics occurring on a millisecond time scale have also been detected for Lys27. The NH3(+) groups of Lys48, which plays a key role as the linkage site in ubiquitination for proteasomal degradation, was found to be highly mobile with the lowest order parameter among the six NH3(+) groups analyzed by NMR. We compared the experimental order parameters for the lysine NH3(+) groups with those from a 1 μs molecular dynamics simulation in explicit solvent and found good agreement between the two. Furthermore, both the computer simulation and the experimental correlation times for the bond rotations of NH3(+) groups suggest that their hydrogen bonding is highly dynamic with a subnanosecond lifetime. This study demonstrates the utility of combining NMR experiment and simulation for an in-depth characterization of the dynamics of these functionally most important side-chains of ubiquitin.     

Off-resonance TROSY-selected R 1rho experiment with improved sensitivity for medium- and high-molecular-weight proteins

NMR spin relaxation techniques that utilize relaxation interference phenomena (TROSY) enable chemical exchange processes to be characterized in high-molecular-weight proteins. A TROSY-selected (TS) approach for measuring off-resonance R1rho relaxation in the spin-locked rotating reference frame is developed using three principles: (i) deuteration of nonexchangeable 1H sites to minimize remote dipole-dipole interactions, (ii) selective excitation of the slowly relaxing 15N doublet component to obtain optimal initial conditions, and (iii) selective inversion of one of the 15N doublet components to suppress cross-relaxation during the spin-lock period. The method is validated using [90%-15N, 70%-2H] ubiquitin at 280 K. The TROSY-selected R1rho experiment enables characterization of backbone dynamics on the microsecond time scale in large proteins.