Experimental Measurement of Nonuniform Dioxygen Accessibility to Ribonuclease A Surface and Interior

Detecting O2 binding sites in protein cavities

Internal cavities are important elements in protein structure, dynamics, stability and function. Here we use NMR spectroscopy to investigate the binding of molecular oxygen (O₂) to cavities in a well-studied model for ligand binding, the L99A mutant of T4 lysozyme. On increasing the O₂ concentration to 8.9 mM, changes in ¹H, ¹⁵N, and ¹³C chemical shifts and signal broadening were observed specifically for backbone amide and side chain methyl groups located around the two hydrophobic cavities of the protein. O₂-induced longitudinal relaxation enhancements for amide and methyl protons could be adequately accounted for by paramagnetic dipolar relaxation. These data provide the first experimental demonstration that O₂ binds specifically to the hydrophobic, and not the hydrophilic cavities, in a protein. Molecular dynamics simulations visualized the rotational and translational motions of O₂ in the cavities, as well as the binding and egress of O₂, suggesting that the channel consisting of helices D, E, G, H, and J could be the potential gateway for ligand binding to the protein. Due to strong paramagnetic relaxation effects, O₂ gas-pressure NMR measurements can detect hydrophobic cavities when populated to as little as 1%, and thereby provide a general and highly sensitive method for detecting oxygen binding in proteins.

Cosolute paramagnetic relaxation enhancements detect transient conformations of human uracil DNA glycosylase (hUNG)

The human DNA repair enzyme uracil DNA glycosylase (hUNG) locates and excises rare uracil bases that arise in DNA from cytosine deamination or through dUTP incorporation by DNA polymerases. Previous NMR studies of hUNG have revealed millisecond time scale dynamic transitions in the enzyme-nonspecific DNA complex, but not the free enzyme, that were ascribed to a reversible clamping motion of the enzyme as it scans along short regions of duplex DNA in its search for uracil. Here we further probe the properties of the nonspecific DNA binding surface of {(2)H(12)C}{(15)N}-labeled hUNG using a neutral chelate of a paramagnetic Gd(3+) cosolute (Gd(HP-DO3A)). Overall, the measured paramagnetic relaxation enhancements (PREs) on R(2) of the backbone amide protons for free hUNG and its DNA complex were in good agreement with those calculated based on their relative exposure observed in the crystal structures of both enzyme forms. However, the calculated PREs systematically underestimated the experimental PREs by large amounts in discrete regions implicated in DNA recognition and catalysis: active site loops involved in DNA recognition (268-274, 246-250), the uracil binding pocket (143-148, 169-170), a transient extrahelical base binding site (214-216), and a remote hinge region (129-132) implicated in dynamic clamping. These reactive hot spots were not correlated with structural, hydrophobic, or solvent exchange properties that might be common to these regions, leaving the possibility that the effects arise from dynamic sampling of exposed conformations that are distinct from the static structures. Consistent with this suggestion, the above regions have been previously shown to be flexible based on relaxation dispersion measurements and course-grained normal-mode analysis. A model is suggested where the intrinsic dynamic properties of these regions allows sampling of transient conformations where the backbone amide groups have greater average exposure to the cosolute as compared to the static structures. We conclude that PREs derived from the paramagnetic cosolute reveal dynamic hot spots in hUNG and that these regions are highly correlated with substrate binding and recognition.

Dioxygen transmembrane distributions and partitioning thermodynamics in lipid bilayers and micelles

Cellular respiration, mediated by the passive diffusion of oxygen across lipid membranes, is key to many basic cellular processes. In this work, we report the detailed distribution of oxygen across lipid bilayers and examine the thermodynamics of oxygen partitioning via NMR studies of lipids in a small unilamellar vesicle (SUV) morphology. Dissolved oxygen gives rise to paramagnetic chemical shift perturbations and relaxation rate enhancements, both of which report on local oxygen concentration. From SUVs containing the phospholipid sn-2-perdeuterio-1-myristelaidoyl, 2-myristoyl-sn-glycero-3-phosphocholine (MLMPC), an analogue of 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC), we deduced the complete trans-bilayer oxygen distribution by measuring (13)C paramagnetic chemical shifts perturbations for 18 different sites on MLMPC arising from oxygen at a partial pressure of 30 bar. The overall oxygen solubility at 45 °C spans a factor of 7 between the bulk water (23.7 mM) and the bilayer center (170 mM) and is lowest in the vicinity of the phosphocholine headgroup, suggesting that oxygen diffusion across the glycerol backbone should be the rate-limiting step in diffusion-mediated passive transport of oxygen across the lipid bilayer. Lowering of the temperature from 45 to 25 °C gave rise to a slight decrease of the oxygen solubility within the hydrocarbon interior of the membrane. An analysis of the temperature dependence of the oxygen solubility profile, as measured by (1)H paramagnetic relaxation rate enhancements, reveals that oxygen partitioning into the bilayer is entropically favored (ΔS° = 54 ± 3 J K(-1) mol(-1)) and must overcome an enthalpic barrier (ΔH° = 12.0 ± 0.9 kJ mol(-1)).

A solution NMR approach to the measurement of amphiphile immersion depth and orientation in membrane model systems

Oxygen and Ni(II) are ideal paramagnetic species for NMR studies of immersion depth since they establish prominent concentration gradients across the membrane-water interface of either bilayers or micelles. Corresponding gradients of paramagnetic shifts and relaxation rates are observed by NMR for membrane embedded amphiphiles. Specifically, upon dissolution of oxygen at a partial pressure of 20 bar or more, (13)C NMR spectra of membrane embedded amphiphiles reveal chemical shift perturbations which depend sensitively on average immersion depth in the membrane. Similarly, depth-dependent enhancements of spin-lattice relaxation rates can be detected by (1)H NMR. Generally, such paramagnetic effects depend both on steric or accessibility factors and on the local concentration of the paramagnet. The steric terms can be factored out by combining paramagnetic rates arising from O(2) and Ni, in the form of a ratio, R(1P)(O(2))/R(1P)(Ni). The natural logarithm of this ratio is shown to depend linearly on immersion depth in a micelle. The analysis is verified using molecular dynamics simulations of dodecylphosphocholine in a detergent micelle, while thorough consideration of the paramagnetic rate data also allows for the determination of the orientation of imipramine in the micelle. Thus, a complete picture of topology arises from this approach which is readily applicable to studies of drugs and amphiphiles in fast-tumbling bicelles, small unilamellar vesicles, and micelles.

Spin-diffusion couples proton relaxation rates for proteins in exchange with a membrane interface

Changes in nuclear spin-lattice relaxation rates that are induced by a freely diffusing paramagnetic relaxation agent are examined for a protein in solution and compared to the case where the protein binds to a membrane. In the solution case, the intramolecular cross-relaxation rates are modest and large differences are observed in the oxygen induced protein-proton relaxation rates. In the case where a dynamic equilibrium between solution and membrane-bound environments is established, the intramolecular (1)H cross-relaxation rates for the protein protons increase dramatically because of the slow reorientational motion in the membrane-bound environment. As a consequence, all protein protons relax with nearly the same spin-lattice relaxation rate constants when bound to the membrane, and site specific relaxation effects of the diffusing paramagnet are suppressed. Slowly reorienting sites or rotationally immobilized sites sampled by observable molecules in vivo will demonstrate similar relaxation leveling effects.

Oxygen accessibility to ribonuclease a: quantitative interpretation of nuclear spin relaxation induced by a freely diffusing paramagnet

The nuclear spin relaxation induced by a freely diffusing paramagnetic center provides a direct measure of intermolecular accessibility. A number of factors are involved in a quantitative interpretation of relaxation data including excluded volume effects, solvation differences, and the details of the electron spin relaxation in the paramagnetic center. In the case where the electron relaxation time is short compared with correlation times describing the electron-nuclear coupling, the nuclear spin relaxation rates may be related to the effective local concentration of the paramagnetic center at different locations about the solute of interest. The local concentrations may in turn be related to differences in the local free energies of interaction between the diffusing paramagnet and the cosolute. We demonstrate this approach for the case of ribonuclease A and deduce surface free energy differences for a large number of protein proton sites. We find that the oxygen accessibility is poorly represented by hard-sphere models such as computed solvent or steric accessibility. There is a distribution of local intermolecular interactions with a width of the order of RT that dominates the report of the intermolecular exploration of the protein by this simple solute.

19F NMR studies of solvent exposure and peptide binding to an SH3 domain

(19)F NMR was used to study topological features of the SH3 domain of Fyn tyrosine kinase for both the free protein and a complex formed with a binding peptide. Metafluorinated tyrosine was biosynthetically incorporated into each of 5 residues of the G48M mutant of the SH3 domain (i.e. residues 8, 10, 49 and 54 in addition to a single residue in the linker region to the C-terminal polyhistidine tag). Distinct (19)F NMR resonances were observed and subsequently assigned after separately introducing single phenylalanine mutations. (19)F NMR chemical shifts were dependent on protein concentration above 0.6 mM, suggestive of dimerization via the binding site in the vicinity of the tyrosine side chains. (19)F NMR spectra of Fyn SH3 were also obtained as a function of concentration of a small peptide (2-hydroxynicotinic-NH)-Arg-Ala-Leu-Pro-Pro-Leu-Pro-diaminopropionic acid -NH(2), known to interact with the canonical polyproline II (PPII) helix binding site of the SH3 domain. Based on the (19)F chemical shifts of Tyr8, Tyr49, and Tyr54, as a function of peptide concentration, an equilibrium dissociation constant of 18 +/- 4 microM was obtained. Analysis of the line widths suggested an average exchange rate, k(ex), associated with the peptide-protein two-site exchange, of 5200 +/- 600 s(-1) at a peptide concentration where 96% of the FynSH3 protein was assumed to be bound. The extent of solvent exposure of the fluorine labels was studied by a combination of solvent isotope shifts and paramagnetic effects from dissolved oxygen. Tyr54, Tyr49, Tyr10, and Tyr8, in addition to the Tyr on the C-terminal tag, appear to be fully exposed to the solvent at the metafluoro position in the absence of binding peptide. Tyr54 and, to some extent, Tyr10 become protected from the solvent in the peptide bound state, consistent with known structural data on SH3-domain peptide complexes. These results show the potential utility of (19)F-metafluorotyrosine to probe protein-protein interactions in conjunction with paramagnetic contrast agents.

Spin relaxation measurements of electrostatic bias in intermolecular exploration

We utilize the paramagnetic contribution to proton spin-lattice relaxation rate constants induced by freely diffusing charged paramagnetic centers to investigate the effect of charge on the intermolecular exploration of a protein by the small molecule. The proton NMR spectrum provided 255 resolved resonances that report how the explorer molecule local concentration varies with position on the surface. The measurements integrate over local dielectric constant variations, and, in principle, provide an experimental characterization of the surface free energy sampling biases introduced by the charge distribution on the protein. The experimental results for ribonuclease A obtained using positive, neutral, and negatively charged small nitroxide radicals are qualitatively similar to those expected from electrostatic calculations. However, while systematic electrostatic trends are apparent, the three different combinations of the data sets do not yield internally consistent values for the electrostatic contribution to the intermolecular free energy. We attribute this failure to the weakness of the electrostatic sampling bias for charged nitroxides in water and local variations in effective translational diffusion constant at the water-protein interface, which enters the nuclear spin relaxation equations for the nitroxide-proton dipolar coupling.

Dynamics of xenon binding inside the hydrophobic cavity of pseudo-wild-type bacteriophage T4 lysozyme explored through xenon-based NMR spectroscopy

Wild-type bacteriophage T4 lysozyme contains a hydrophobic cavity with binding properties that have been extensively studied by X-ray crystallography and NMR. In the present study, the monitoring of 1H chemical shift variations under xenon pressure enables the determination of the noble gas binding constant (K = 60.2 M(-1)). Although the interaction site is highly localized, dipolar cross-relaxation effects between laser-polarized xenon and nearby protons (SPINOE) are rather poor. This is explained by the high value of the xenon-proton dipolar correlation time (0.8 ns), much longer than the previously reported values for xenon in medium-size proteins. This indicates that xenon is highly localized within the protein cavity, as confirmed by the large chemical shift difference between free and bound xenon. The exploitation of the xenon line width variation vs xenon pressure and protein concentration allows the extraction of the exchange correlation time between free and bound xenon. Comparison to the exchange experienced by protein protons indicates that the exchange between the open and closed conformations of T4 lysozyme is not required for xenon binding.

Chromium(III) complexes as intermolecular probes

Metal ion complexes provide flexible paramagnetic centers that may be used to define intermolecular contacts in a variety of solution phase environments because both the charge and electronic relaxation properties of the complex may be varied. For most complex ions, there are several proton equilibria that may change the effective charge on the complex as a function of pH which in turn affects the efficacy of application for defining the electrostatic surfaces of co-solute molecules. We report here spectrophotometric and nuclear spin relaxation studies on aqueous solutions of chromium(III) complexes of EDTA, DTPA, and bis-amides of both. The effective charges available from these paramagnetic centers range from -3 to +1 and we report the pH ranges over which the effective charge is defined with confidence for application in magnetic relaxation experiments.

Probing protein-peptide binding surfaces using charged stable free radicals and transverse paramagnetic relaxation enhancement (PRE)

Nitroxide species, which have an unpaired electron localized on a nitrogen atom, can be useful as NMR probes to identify areas of the surface of a protein involved in the formation of a complex. The proximity of an electron spin leads to higher NMR relaxation rates for protein nuclei. If a protein-ligand complex is formed the radical is excluded from certain sites on the protein surface, protecting them from relaxation effects. We show here that charged nitroxide species can be helpful for identifying regions of the surface of the 4F1(5)F1 module pair from human fibronectin involved in peptide binding.

An NMR study of the origin of dioxygen-induced spin-lattice relaxation enhancement and chemical shift perturbation

Due to its depth-dependent solubility, oxygen exerts paramagnetic effects which become progressively greater toward the hydrophobic interior of micelles, and lipid bilayer membranes. This paramagnetic gradient, which is manifested as contact shift perturbations (19F and 13C NMR) and spin-lattice relaxation enhancement (19F and 1H NMR), has been shown to be useful for precisely determining immersion depth, membrane protein secondary structure, and overall topology of membrane proteins. We have investigated the influence of oxygen on 19F and 13C NMR spectra and spin-lattice relaxation rates of a semiperfluorinated detergent, (8,8,8)-trifluoro (3,3,4,4,5,5,6,6,7,7)-difluoro octylmaltoside (TFOM) in a model membrane system, to determine the dominant paramagnetic spin-lattice relaxation and shift-perturbation mechanism. Based on the ratio of paramagnetic spin-lattice relaxation rates of 19F and directly bonded 13C nuclei, we conclude that the dominant relaxation mechanism must be dipolar. Furthermore, the temperature dependence of oxygen-induced chemical shift perturbations in 9F NMR spectra suggests a contact interaction is the dominant shift mechanism. The respective hyperfine coupling constants for 19F and 13C nuclei can then be estimated from the contact shifts <(deltav/v0)19F> and <(deltav/v0)13C>, allowing us to estimate the relative contribution of scalar and dipolar relaxation to 19F and 13C nuclei. We conclude that the contribution to spin-lattice relaxation from the oxygen induced paramagnetic scalar mechanism is negligible.

Mapping oxygen accessibility to ribonuclease a using high-resolution NMR relaxation spectroscopy

Paramagnetic contributions to nuclear magnetic spin-lattice relaxation rate constant induced by freely diffusing molecular oxygen measured at hundreds of different protein proton sites provide a direct means for characterizing the exploration of the protein by oxygen. This report focuses on regions of ribonuclease A where the rate constant enhancements are either quite large or quite small. We find that there are several regions of enhanced oxygen affinity for the protein both on the surface and in interior pockets where sufficient free volume permits. Oxygen has weak associative interactions with a number of surface crevices that are generally between secondary structural elements of the protein fold. Several regions near the surface have higher than expected accessibility to oxygen indicating that structural fluctuations in the protein provide intermolecular access. Oxygen penetrates part of the hydrophobic interior, but affinity does not correlate simply with hydrophobicity indices. Oxygen is excluded from regions of high interior packing density and a few surface sites where x-ray diffraction data have indicated the presence of specific hydration with high occupancy.

Amide proton relaxation measurements employing a highly deuterated protein

Proton NMR longitudinal and transverse relaxation rates of unlabelled proteins are generally dominated by the many 1H-1H dipolar interactions so that spin diffusion, rather than molecular or internal motions, governs longitudinal relaxation. Here, relaxation measurements of backbone amide proton (1H(N)) magnetisations have been carried out employing the 99% 2H, 98% 15N labelled, small 2F2 protein domain in 10%/90% H(2)O/D(2)O solution. Under these conditions, the longitudinal relaxation rates exhibit time constants, T(1)*=1/R(1)* if described by a mono-exponential, within the range of 3.0 to 18.7s-a wide range which indicates that the phenomenon of spin diffusion has been greatly reduced. The majority of 1H(N) nuclei in this sample (pH 4.0 and 5 degrees C) exhibit chemical exchange with solvent that couples their longitudinal relaxation to that of the solvent. For the subset of 1H(N) nuclei not undergoing detectable solvent chemical exchange, the R(1)* rates correlate well with their individual 1H(N,O)/2H(N,O) structural environments. The correlation for corresponding transverse relaxation rates, R(2)* was found to be less good. Longitudinal relaxation measurements in 1%/99% H(2)O/D(2)O solution identify a further subset of 1H(N) nuclei which exhibit essentially indistinguishable R(1)* rates in both 1% and 10% H(2)O, implying that averaging of rates from spin diffusion processes and different 2F2 isotopomer populations are negligible for these 1H(N) sites. In addition to a high sensitivity to structural parameters, model calculations predict 1H(N) relaxation rates to exhibit pronounced sensitivity to internal dynamics.

The distribution of lipid attached spin probes in bilayers: application to membrane protein topology

The distribution of the lipid-attached doxyl electron paramagnetic resonance (EPR) spin label in 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine membranes has been studied by (1)H and (13)C magic angle spinning nuclear magnetic resonance relaxation measurements. The doxyl spin label was covalently attached to the 5th, 10th, and 16th carbons of the sn-2 stearic acid chain of a 1-palmitoyl-2-stearoyl-(5/10/16-doxyl)-sn-glycero-3-phosphocholine analog. Due to the unpaired electron of the spin label, (1)H and (13)C lipid relaxation rates are enhanced by paramagnetic relaxation. For all lipid segments the influence of paramagnetic relaxation is observed even at low probe concentrations. Paramagnetic relaxation rates provide a measure for the interaction strength between lipid segments and the doxyl group. Plotted along the membrane director a transverse distribution profile of the EPR probe is obtained. The chain-attached spin labels are broadly distributed in the membrane with a maximum at the approximate chain position of the probe. Both (1)H and (13)C relaxation measurements show these broad distributions of the doxyl group in the membrane indicating that (1)H spin diffusion does not influence the relaxation measurements. The broad distributions of the EPR label result from the high degree of mobility and structural heterogeneity in liquid-crystalline membranes. Knowing the distribution profiles of the EPR probes, their influence on relaxation behavior of membrane inserted peptide and protein segments can be studied by (13)C magic angle spinning nuclear magnetic resonance. As an example, the location of Ala residues positioned at three sites of the transmembrane WALP-16 peptide was investigated. All three doxyl-labeled phospholipid analogs induce paramagnetic relaxation of the respective Ala site. However, for well ordered secondary structures the strongest relaxation enhancement is observed for that doxyl group in the closest proximity to the respective Ala. Thus, this approach allows study of membrane insertion of protein segments with respect to the high molecular mobility in liquid-crystalline membranes.

The effects of dissolved oxygen upon amide proton relaxation and chemical shift in a perdeuterated protein

The effects of dissolved molecular oxygen upon amide proton ((1)H(N)) longitudinal and transverse relaxation rates and chemical shifts were studied for a small protein domain, the second type 2 module of fibronectin ((2)F2)-isotopically enriched to 99% (2)H, 98% (15)N. Longitudinal relaxation rate enhancements, R(O(2))((1)H(N)), of individual backbone (1)H(N) nuclei varied up to 14 fold between a degassed and oxygenated (1 bar) solution, indicating that the oxygen distribution within the protein is inhomogeneous. On average, smaller relaxation rate enhancements were observed for (1)H(N) nuclei associated with the core of the protein compared to (1)H(N) nuclei closer to the surface, suggesting restricted oxygen accessibility to some regions. In agreement with an O(2)-(1)H(N) hyperfine interaction in the extreme narrowing limit, the (1)H(N) transverse relaxation rates showed no significant change, up to an oxygen pressure of 9.5 bar (the maximum pressure used in this study). For most (1)H(N) resonances, small deltadelta(O(2))((1)H(N)) hyperfine chemical shifts could be detected between oxygen pressures of 1 bar and 9.5 bar.

Determination of membrane immersion depth with O(2): a high-pressure (19)F NMR study

Oxygen is known to partition with an increasing concentration gradient toward the hydrophobic membrane interior. At partial pressures (P(O2)) of 100 Atm or more, this concentration gradient is sufficient to induce paramagnetic effects that depend sensitively on membrane immersion depth. This effect is demonstrated for the fluorine nucleus by depth-dependent paramagnetic shifts and spin-lattice relaxation rates, using a fluorinated detergent, CF3(CF(2))(5)C(2)H(4)-O-maltose (TFOM), reconstituted into a lipid bilayer model membrane system. To interpret the spin-lattice relaxation rates (R) in terms of a precise immersion depth, two specifically fluorinated cholesterol species (6-fluorocholesterol and 25-fluorocholesterol), whose membrane immersion depths were independently estimated, were studied by (19)F NMR. The paramagnetic relaxation rates, R, of the cholesterol species were then used to parameterize a Gaussian profile that directly relates R to immersion depth z. This same Gaussian curve could then be used to determine the membrane immersion depth of all six fluorinated chain positions of TFOM and of two adjacent residues of specifically fluorinated analogs of the antibacterial peptide indolicidin. The potential of this method for determination of immersion depth and topology of membrane proteins is discussed.

Molecular oxygen spin-lattice relaxation in solutions measured by proton magnetic relaxation dispersion

Proton spin-lattice relaxation rate constants have been measured as a function of magnetic field strength for water, water-glycerol solution, cyclohexane, methanol, benzene, acetone, acetonitrile, and dimethyl sulfoxide. The magnetic relaxation dispersion is well approximated by a Lorentzian shape. The origin of the relaxation dispersion is identified with the paramagnetic contribution from molecular oxygen. In the small molecule cases studied here, the effective correlation time for the electron-nuclear coupling may include contributions from both translational diffusion and the electron T(1). The electron T(1) for molecular oxygen dissolved in several solvents was found to be approximately 7.5 ps and nearly independent of solvent or viscosity.

Using O2 to probe membrane immersion depth by 19F NMR

A fluorinated detergent, CF(3)(CF(2))(5)C(2)H(4)-O-maltose, was reconstituted into a lipid bilayer model membrane system to demonstrate the feasibility of determining solvent accessibility and membrane immersion depth of each fluorinated group by (19)F NMR. Apolar oxygen, which is known to partition with an increasing concentration gradient toward the hydrophobic membrane interior, exhibits a range of paramagnetic relaxation effects on (19)F nuclei, depending on its depth in the membrane. This effect, which is predominately associated with spin-lattice relaxation rates (R(1)) and chemical shifts, can be amplified greatly with minimal line broadening by increasing the partial pressure of O(2) at least 100-fold (i.e., P(O(2)) greater than 20 bar). The differences of longitudinal relaxation rates at 20 bar of oxygen pressure to those under ambient pressure (R(1)(20bar) - R(1)(0)) are largest for those fluorine groups expected to be most deeply buried in the membrane bilayer. This result contrasts with the reverse trend, which is observed on addition of a membrane surface-associated paramagnetic species, 4-(N,N-dimethyl-N-hexadecyl) ammonium-2,2,6,6-tetramethylpiperidine-1-oxyl iodide (CAT-16) at ambient pressures. Thus, differential relaxation rates may be observed in (19)F-labeled membrane-associated molecules resulting from the addition of apolar oxygen under high pressure. The results demonstrate that the degree of solvent accessibility and membrane immersion depth of specific fluorinated species in membrane-associated macromolecules can be probed by (19)F NMR.