Probing protein structure by solvent perturbation of nmr spectra. II. Determination of surface and buried residues in homologous proteins

The experimental assignment of most residues in a protein to the surface or interior is in principle possible without prior solution of a complete three-dimensional structure. The method described is based on nmr measurements that determine the amino acid composition of the surface of a protein [A. Petros, L. Mueller, and K. D. Kopple (1990) Biochemistry, Vol. 29, pp. 10041–10048; G. Esposito, A. M. Lesk, H. Molinari, A. Motta, N. Niccolai, and A. Pastore (1992) Journal of Molecular Biology, Vol. 224, pp. 659–670]. If these measurements are carried out on several homologous proteins of known sequence, it is possible to combine the results to determine, in most cases, which positions in the sequence contain exposed residues. © 1993 John Wiley & Sons, Inc.     

Probing protein dynamics by nuclear magnetic resonance

Proteins are dynamic molecules that often undergo conformational changes while performing their specific functions, such as target recognition, ligand binding and catalysis. NMR spectroscopy is uniquely suited to study protein dynamics, because site-specific information can be obtained for motions that span a broad range of time scales. The information obtained from NMR dynamics experiments has provided insights into specific structural changes or conformational energetics associated with molecular function. In the last decade, a number of new advancements in NMR methodologies have further extended our ability to characterize protein dynamics. Here, we present an overview of current NMR technology that is used to monitor the dynamic properties of proteins.     

Probing protein structure by solvent perturbation of NMR spectra: the surface accessibility of bovine pancreatic trypsin inhibitor.

In the absence of specific interactions, the relative attenuation of protein NMR signals due to added stable free radicals such as TEMPOL should reflect the solvent accessibility of the molecular surface. The quantitative correlation between observed attenuation and surface accessibility was investigated with a model system, i.e., the small protein bovine pancreatic trypsin inhibitor. A detailed discussion is presented on the reliability and limits of the approach, and guidelines are provided for data acquisition, treatment, and interpretation. The NMR-derived accessibilities are compared with those obtained from x-ray diffraction and molecular dynamics data. Although the time-averaged accessibilities from molecular dynamics are ideally suited to fit the NMR data, better agreement was observed between the paramagnetic attenuations of the fingerprint cross-peaks of homonuclear proton spectra and the total NH and H alpha accessibilities calculated from x-ray coordinates, than from time-averaged molecular dynamics simulations. In addition, the solvent perturbation response appears to be a promising approach for detecting the thermal conformational evolution of secondary structure elements in proteins.     

Nuclear magnetic resonance and protein structure

NMR provides a wealth of structural information about proteins in solution, but does not, by itself, permit an unambiguous determination of a unique structure. A rigorous interpretation of NMR data to obtain the entire family of structures compatible with a given data set requires extensive, systematic and unbiased sampling of the conformational space of the polypeptide chain. Methods of sampling based on the exclusion paradigm--i. e. those that generate structures, check constraints and accept or reject members of the family on that basis, avoid the problem of generating erroneous structures by converging on local minima, which is a common pitfall of methods based on the optimization paradigm. Their much higher computational cost can be reduced by solving the structure in stages, using abstract representations of partial structures, and guiding the computation by control heuristics. The heuristic refinement method developed at Stanford and encoded in the expert system PROTEAN yields more or less extensive families of structures, depending on the size of the NMR data set, and defines the "allowed volume" in which each atom (or other substructure) may lie, with all experimental constraints satisfied. The allowed volume is a measure of the uncertainty of our knowledge of the structure, to which both the limitations of the data and the uncertainty of position resulting from molecular motion may contribute. Prediction of the experimental NMR spectra by solving the generalized Bloch equations (or the Redfield density matrix) for the protein, using atomic coordinates that lie within the allowed atomic volume, provides the final test for the correctness of the proposed structure.(ABSTRACT TRUNCATED AT 250 WORDS)     

Nuclear magnetic resonance studies of amino acids and proteins. Rotational correlation times of proteins by deuterium nuclear magnetic resonance spectroscopy

We show that measurement of the spin-lattice (T1) and spin-spin (T2) relaxation times (or line widths) of irrotationally bound 2H nuclei in macromolecules undergoing isotropic rotational motion outside of the extreme narrowing limit (i.e., for the case omega 02 tau R2 much greater than 1) permits determination of both the rotational correlation time (tau R) of the macromolecule and the electric quadrupole coupling constant (e2qQ/h) of the 2H label. The technique has the advantage over 13C nuclear magnetic resonance (NMR) that no assumptions about bond lengths (which appear to the sixth power in 13C relaxation studies) or relaxation mechanisms need to be made, since relaxation will always be quadrupolar, even for aromatic residues at high field. Asymmetry parameter (eta) uncertainties are shown to cause negligible effects on tau R determinations, and in any case it is shown that both e2qQ/h and eta may readily be determined in separate solid-state experiments. By way of example, we report 2H NMR results on aqueous lysozyme (EC 3.2.1.17) at 5.2 and 8.5 T (corresponding to 2H-resonance frequencies of 34 and 55 MHz). Interpretation of the results in terms of the isotropic rigid-rotor model yields e2qQ/h values of approximately equal to 170 or approximately equal to 190 kHz, respectively, for the imidazolium and free-base forms of [epsilon 1-2H] His-15 lysozyme in solution, in excellent agreement with e2qQ/h values of approximately 167 and approximately 190 kHz obtained for the free amino acids in the solid state. In principle, the method may in suitable cases permit comparison between the dynamic structures of proteins in solution and in the crystalline solid state.     

Homonuclear three-dimensional NOE-NOE nuclear magnetic resonance/spectra for structure determination of proteins in solution.

The solution structures of two proteins (CMTI-I, a trypsin inhibitor from Cucurbita maxima, and hisactophilin, an actin binding protein of 118 amino acids) have been determined based on the NOE data derived solely from the homonuclear 3D NOE-NOE magnetic resonance spectroscopy. Two different approaches for extraction of the structural information from the 3D NOE-NOE experiment were tested. One approach was based on the transformation of the 3D intensities into distance constraints. In the second, and more robust approach, the 3D NOE intensities were used directly in structure calculations, without the need to transform them into distance constraints. A new 2D potential function representing the 3D NOE-NOE intensity was developed and used in the simulated annealing protocol. For CMTI-I, a comparison between structures determined with the 3D NOE-NOE method and various 2D NOE approaches was carried out. The 3D data set allowed better definition of the structures than was previously possible with the 2D NOE procedures that used the isolated two-spin approximation to derive distance information.     

Nuclear magnetic resonance transverse relaxation in muscle water.

The origin of the nonexponentiality of proton spin echoes of skeletal muscle has been carefully examined. It is shown that the slowly decaying part of the proton spin echoes is not due to extracellular water. First, for muscle from mice with in vivo deuteration, the deuteron spin echoes were also nonexponential, but the slowly decaying part had a larger weighing factor. Second, for glycerinated muscle in which cell membranes were disrupted, the proton spin echoes were similar to those in intact muscle. Third, the nonexponentiality of the proton spin echoes in intact muscle increased when postmortem rigor set in. Finally, when the lifetimes of extracellular water and intracellular water were taken into account in the exchange, it was found that the two types of water would not give two resolvable exponentials with the observed decay constants. It is suggested that the unusually short T2's and the nonexponential character of the spin echoes of proton and deuteron in muscle water are mainly due to hydrogen exchange between water and functional groups in the protein filaments. These groups have large dipolar or quadrupolar splittings, and undergo hydrogen exchange with water at intermediate rates. The exchange processes and their effects on the spin echoes are pH-dependent. The dependence of transverse relaxation of pH was observed in glycerinated rabbit psoas muscle fibers.     

Nuclear magnetic resonance studies of amino acids and proteins. Deuterium nuclear magnetic resonance relaxation of deuteriomethyl-labeled amino acids in crystals and in Halobacterium halobium and Escherichia coli cell membranes

We have obtained deuterium (2H) Fourier transform nuclear magnetic resonance (NMR) spectra of zwitterionic L-[beta-2H3]alanine, DL-[gamma-2H6]valine, DL-[beta, gamma-2H4]threonine, L-[delta-2H3]leucine, and L-[alpha, beta, gamma, gamma', delta-2H10]isoleucine in the crystalline solid state and have determined the deuteriomethyl group spin-lattice relaxation rates as a function of temperature. The results yield the Arrhenius activation energies (delta E) for methyl rotation, and through use of a suitable mathematical model, rotational correlation times, tau c. For alanine, valine, threonine, leucine, and isoleucine at 37 degrees C, tau c and delta E values are 780, 100, 40, 38, and 18 ps and 22, 14.0, 17.6, 15.5, and 8.6 kJ, respectively. For L-[beta-2H3]alanine in the zwitterionic lattice, a spin-lattice relaxation time (T1) minimum of 2.1 +/- 0.3 ms is observed (at 0 degree C), in excellent agreement with the 1.92-ms prediction of the mathematical model. Similar tau c and delta E measurements are reported for bacteriorhodopsin in the purple membrane of Halobacterium halobium R1 and for Escherichia coli cell membranes. Overall, our results demonstrate a great similarity between the dynamics in amino acid crystals and in membrane proteins. However, threonine exhibits a nonlinear Arrhenius behavior in bacteriorhodopsin, and in the valine-, leucine-, and isoleucine-labeled membrane samples at higher temperatures (approximately greater than 37 degrees C), there is evidence of an additional slow side-chain motion. The lipid phase state in E. coli does not appear to influence, on the average, the dynamics of the valine side chains. These results indicate that the sensitivity of the deuterium NMR technique is now adequate to study in moderate detail the dynamics of most types of amino acids in a membrane protein and that adequate sensitivity, in some instances, should be available for the study of individual amino acids in suitably labeled membrane proteins.     

Nuclear magnetic resonance of lipid A--the influence of solvents on spin relaxation and spectral quality

Nuclear magnetic resonance (NMR) spectroscopy of lipid A is limited by rapid transversal relaxation and subsequent line broadening caused by the tendency of these glycolipids to form aggregates in all solvents. To examine the influence of solvents on NMR spectra, hexa-acyl lipid A from Escherichia coli F515 was investigated. Line widths at half height, longitudinal relaxation times, and transversal relaxation times were measured in different solvents, lipid A concentrations, and temperatures. Chloroform-d, dioxane-d(8), and pyridine-d(5) each mixed with 25% methanol-d(4) as well as sole DMSO-d(6) and 0.1M triethylamine-d(15) (TEA-d(15)) in D(2)O caused good spectral resolutions and allowed structure analysis. ROESY and HMBC spectra gave an insight into the influence of transversal relaxation times on spectral quality in two-dimensional spectra. Solvent depending differences of interglycosidic NOEs indicated dissimilarities of the conformations in the interglycosidic linkage and allowed conclusions about the lipid A solution state.     

Sequential resonance assignments in protein 1H nuclear magnetic resonance spectra. Basic pancreatic trypsin inhibitor

The assignment of the ¹H nuclear magnetic resonance spectrum of the basic pancreatic trypsin inhibitor with the use of two-dimensional ¹H nuclear magnetic resonance techniques at 500 MHz is described. The assignments are based entirely on the known amino acid sequence and the nuclear magnetic resonance data. Individual resonance assignments were obtained for all backbone and C^β protons, with the exception of those of Arg1, Pro2, Pro13 and the amide proton of Gly37. The side-chain resonance assignments are complete, with the exception of Pro2 and Pro13, the N^δ protons of Asn44 and the peripheral protons of the lysine residues and all but two of the arginine residues.     

Probing the electronic structure of transition metal ion centres in proteins by coherent Raman-detected electron paramagnetic resonance spectroscopy

The simultaneous excitation of a paramagnetic sample with optical (laser) and microwave radiation can cause an amplitude or phase modulation of the transmitted light at the microwave frequency. The detection of this modulation indicates the presence of coupled optical and electron paramagnetic resonance (EPR) transitions in the sample. Here we report the first application of this technique to a biomolecule: the blue copper centre of Pseudomonas aeruginosa azurin. Using optical excitation at 686 nm, in the thiol to copper(II) charge transfer band, we measure a coherent Raman-detected EPR spectrum of a frozen aqueous solution. Its lineshape is characteristic of the magnetic circular dichroism along each principal g-value axis. This information allows electronic and structural models of transition metal ion centres in proteins to be tested.     

Sodium-23 and potassium-39 nuclear magnetic resonance relaxation in eye lens. Examples of quadrupole ion magnetic relaxation in a crowded protein environment.

Single and multiple quantum nuclear magnetic resonance (NMR) spectroscopic techniques were used to investigate the motional dynamics of sodium and potassium ions in concentrated protein solution, represented in this study by cortical and nuclear bovine lens tissue homogenates. Both ions displayed homogeneous biexponential magnetic relaxation behavior. Furthermore, the NMR relaxation behavior of these ions in lens homogenates was consistent either with a model that assumed the occurrence of two predominant ionic populations, "free" and "bound," in fast exchange with each other or with a model that assumed an asymmetric Gaussian distribution of correlation times. Regardless of the model employed, both ions were found to occur in a predominantly "free" or "unbound" rapidly reorienting state. The fraction of "bound" 23Na+, assuming a discrete two-site model, was approximately 0.006 and 0.017 for cortical and nuclear homogenates, respectively. Corresponding values for 39K+ were 0.003 and 0.007, respectively. Estimated values for the fraction of "bound" 23Na+ or 39K+ obtained from the distribution model (tau C greater than omega L-1) were less than or equal to 0.05 for all cases examined. The correlation times of the "bound" ions, derived using either a two-site or distribution model, yielded values that were at least one order of magnitude smaller than the reorientational motion of the constituent lens proteins. This observation implies that the apparent correlation time for ion binding is dominated by processes other than protein reorientational motion, most likely fast exchange between "free" and "bound" environments. The results of NMR visibility studies were consistent with the above findings, in agreement with other studies performed by non-NMR methods. These studies, in combination with those presented in the literature, suggest that the most likely role for sodium and potassium ions in the lens appears to be the regulation of cell volume by affecting the intralenticular water chemical potential.     

High-resolution phosphorus nuclear magnetic resonance spectra of yeast phenylalanine transfer ribonucleic acid. Melting curves and relaxation effects.

In a continuation of our studies on structural effects on the 31P chemical shifts of nucleic acids, we present 31P NMR spectra of yeast phenylalanine tRNA in the presence and absence of Mg2+. Superconducting field (146 MHz) and 32-MHz 31P NMR spectra reveal approximately 15 nonhelical diester signals spread over approximately 7 ppm besides the downfield terminal 3'-phosphate monoester. In the presence of 10 mM Mg2+, most scattered and main cluster signals do not shift between 22--66 degrees C, thus supporting our earlier hypothesis that 31P chemical shifts are sensitive to phosphate ester torsional and bond angles. At 70 degrees C, all of the signals merge into a single random coil conformation signal. Similar effects are observed in the absence of Mg2+ except that the transition melting temperature is approximately 20 degrees C lower. Measured spin-lattice and spin-spin relaxation times reveal another lower temperature transition besides the thermal denaturation process. A number of the scattered peaks are shifted (0.2--1.7 ppm) and broadened between 22 and 66 degrees C in the presence of Mg2+ as a result of this conformational transition between two intact tertiary structures. The loss of the scattered peaks in the absence of Mg2+ occurs in the temperature range expected for melting of a tertiary structure. An attempt to simulate the 31P spectra of tRNA Phe based upon the X-ray crystallographically determined phosphate ester torsional agles supports the suggestion that the large shifts in the scattered peaks are due to bond angle distortions in the tertiary structure.     

Nuclear magnetic resonance study of spin relaxation and magnetic field gradients in maple leaves.

1H Nuclear magnetic resonance techniques were used to measure the distributions of spin-spin relaxation times, T2, and of magnetic field gradients in both the chloroplast and nonchloroplast water compartments of maple leaves (Acer platanoides). Results showed that encounters between water molecules and membranes inside chloroplasts provide an inefficient relaxation mechanism; i.e., chloroplast membranes interact weakly with water molecules. Gradient measurements indirectly measured the sizes of chloroplasts by showing that water in the chloroplasts is confined to small compartments a few microns in diameter. A comparison between measured gradients and gradients calculated for a model leaf indicated that chloroplasts are somewhat more likely to occupy positions along cell walls adjacent to air spaces, but also they may be found in the interiors of cells.     

A protein structure from nuclear magnetic resonance data. lac repressor headpiece

A procedure is described to determine the three-dimensional structure of biomolecules from nuclear magnetic resonance data. This procedure combines model building with a restrained molecular dynamics algorithm, in which distance information from nuclear Overhauser effects is incorporated in the form of pseudo potentials. The method has been applied to the N-terminal DNA-binding domain or headpiece (amino acid residues 1 to 51) of the lac repressor from Escherichia coli, for which no crystal structure is available. The relative orientation of the three helices of the headpiece is similar to that of the three homologous helices found in the cI repressor of bacteriophage lambda.     

Structure of the dysprosium-glycocholate complex in submicellar aqueous solution: paramagnetic mapping by proton nuclear magnetic resonance spectroscopy. An approximation for the intrinsic "bound" relaxation rates in the case of nondilute paramagnetic systems

A paramagnetic NMR study of the structure of the calcium-glycocholate complex in submicellar solution, utilizing dysprosium as an isomorphous lanthanide replacement of calcium, is presented. The dysprosium-induced relaxation rate (1/T1) enhancements of certain glycocholate protons have been used to estimate internuclear distances between these protons and the metal ion. An approximation to calculate the intrinsic relaxation rate (1/T1) enhancements for a nondilute paramagnetic solution is given in the Appendix. From these data, and analysis based on conformation averaging and minimum energy conformations, a molecular model of the dysprosium-glycocholate complex in submicellar aqueous solution has been constructed. In this model the metal ion has a unidentate, first-sphere interaction with the proximal oxygen atom of the glycine carboxyl. The metal ion has second-sphere interactions with the peptide bond carbonyl oxygen (3.6 A) and the distal carboxyl oxygen (4.4 A). The metal ion to hydroxyl oxygen distances (8.4-12.4 A) are not compatible with any metal ion to hydroxyl coordination. The side chain appears to exist in one predominant conformation. All six oxygen atoms of glycocholate, the peptide bond carbonyl, the carboxyl group, and the hydroxyl groups are on the alpha face of the bile salt molecule. On the basis of these features we conclude that in the submicellar state the solution structure of the dysprosium-glycocholate complex displays a metal ion enhanced segregation of polar versus nonpolar groups to the two separate faces of the molecule, which may result in a facilitated hydrophobic interaction of different complex units.     

Interstitial sodium nuclear magnetic resonance relaxation times in perfused hearts.

Separation of intracellular and extracellular sodium nuclear magnetic resonance (NMR) signals would enable nondestructive monitoring of intracellular sodium. It has been proposed that differences between the relaxation times of intracellular and extracellular sodium be used either directly or indirectly to separate the signal from each compartment. However, whereas intracellular sodium relaxation times have been characterized for some systems, these times were unknown for interstitial sodium. In this study, the interstitial sodium NMR relaxation times have been measured in perfused frog and rat hearts under control conditions. This was achieved by eliminating the NMR signal from the extracardiac (perfusate) sodium, and then quantifying the remaining cardiac signal. The intracellular signal was measured to be 8% (frog) or 22% (rat) of the cardiac signal and its subtraction was found to have a negligible effect on the cardiac relaxation times. Therefore this cardiac signal is considered to provide a good estimate of interstitial relaxation behavior. For perfused frog (rat) hearts under control conditions, this signal was found to have a T1 of 31.6 +/- 3.0 ms (27.3 +/- 1.6 ms) and a biexponential T2 of 1.9 +/- 1.0 ms (2.1 +/- 0.3 ms) and 25.2 +/- 1.3 ms (26.3 +/- 3.2 ms). Due to the methods used to separate cardiac signal from perfusate signal, it is possible that this characterized only a part of the signal from the interstitium. The short T2 component attributable to the interstitial signal indicates that separation of the NMR signals from each compartment on the basis of relaxation times alone may be difficult.