Protein dynamics from solution NMR

Solution nuclear magnetic resonance (NMR) spectroscopy is unique in its ability to elucidate the details of atomic-level structural and dynamical properties of biological macromolecules under native-like conditions. Recent advances in NMR techniques and protein sample preparation now allow comprehensive investigation of protein dynamics over timescales ranging 14 orders of magnitude at nearly every atomic site. Thus, solution NMR is poised to reveal aspects of the physico-chemical properties that govern the ensemble distribution of protein conformers and the dynamics of their interconversion. We review these advances as well as their recent application to the study of proteins.     

Comparison of backbone dynamics of monomeric and domain-swapped stefin A

Three-dimensional domain swapping has been observed in increasing number of proteins and has been implicated in the initial stages of protein aggregation, including that of the cystatins. Stefin A folds as a monomer under native conditions, while under some denaturing conditions domain-swapped dimer is formed. We have determined the backbone dynamics of the monomeric and domain-swapped dimeric forms of stefin A by (15)N relaxation using a model-free approach. The overall correlation times of the molecules were determined to be 4.6 +/- 0.1 ns and 9.2 +/- 0.2 ns for the monomer and the dimer, respectively. In the monomer, decreased order parameters indicate an increased mobility for the N-terminal trunk, the first and the second binding loops. At the opposite side of the molecule, the loop connecting the alpha-helix with strand B, the beginning of strand B and the loop connecting strands C and D show increased localized mobility. In the domain-swapped dimer, a distinctive feature of the structure is the concatenation of strands B and C into a single long beta-strand. The newly formed linker region between strands B and C, which substitutes for the first binding loop in the monomer, has order parameters typical for the remainder of the beta-strands. Thus, the interaction between subunits that occurs on domain-swapping has consequences for the dynamics of the protein at long-range from the site of conformational change, where an increased rigidity in the newly formed linker region is accompanied by an increased mobility of loops remote from that site.     

Solution conformation and dynamics of a fungal cell wall polysaccharide isolated from Microsporum gypseum

The conformational and dynamical features of a branched mannan isolated from a fungal cell wall have been analysed by homo and heteronuclear NMR methods, employing different magnetic fields. 1HNMR cross relaxation times have been obtained for this polysaccharide and have been interpreted qualitatively using different motional models. 13C NMR relaxation parameters (T1, T2, NOE) have also been measured and interpreted using different approximations based on the Lipari and Szabo model free approach. The analysis of the data indicate the existence of important flexibility for the different linkages of the polysaccharide. Motions in the range of 4–6 ns contribute to the relaxation of the macromolecule, although faster internal motions in the 500 ps and 100 ps timescales are also present. These time scales indicate that segmental motions as well as internal motions around the glycosidic linkages are the major sources of relaxation for this molecule at 318 K. Molecular dynamics simulations have also been performed. The obtained results also indicate that the polysaccharide possess a substantial amount of conformational freedom.     

Backbone dynamics and thermodynamics of Borrelia outer surface protein A

Nuclear spin relaxation experiments performed at 298K, 308K and 318K are used to characterize the intramolecular dynamics and thermodynamics of outer surface protein A (OspA), a key protein in the life-cycle of Borrelia burgdorferi, the causative agent of Lyme disease. It has recently been demonstrated that OspA specifically binds to the gut of the intermediate tick host (Ixodes scapularis), and that this interaction is mediated, at least in part, by residues in the C-terminal domain of OspA that are largely inaccessible to solvent in all X-ray structures of this protein. Our analysis of 15N relaxation parameters in OspA shows that the putative-binding region contains and is surrounded by flexible residues, which could facilitate accessibility to solvent and ligands. In addition, residues with similar activation energies are clustered in a manner that suggests locally collective motions. We have used molecular modeling to show that these collective motions are consistent with a hinge-bending mechanism that exposes residues implicated in binding. Characteristic temperatures describing the energy landscape of the OspA backbone are derived from the temperature dependence of the N-H bond vector order parameters, and a comparison is made between the N and C-terminal globular domains and the unusual single-layer beta-sheet connecting them. The average characteristic temperatures in the three regions indicate that, with an increase in temperature, a larger increase in accessible conformational states occurs for N-H bond vectors in the single-layer central beta-sheet than for bond vectors in the globular N and C-terminal domains. These conformational states are accessible without disruption of hydrogen bonds, providing a conformational entropic gain, upon increase in temperature, without a significant enthalpic penalty. This increase in heat capacity may help to explain the unexpected thermal stability of the unusual single-layer beta-sheet.     

NMR investigation of the dynamics of tryptophan side-chains in hemoglobins

NMR relaxation measurements of 15N spin-lattice relaxation rate (R(1)), spin-spin relaxation rate (R(2)), and heteronuclear nuclear Overhauser effect (NOE) have been carried out at 11.7T and 14.1T as a function of temperature for the side-chains of the tryptophan residues of 15N-labeled and/or (2H,15N)-labeled recombinant human normal adult hemoglobin (Hb A) and three recombinant mutant hemoglobins, rHb Kempsey (betaD99N), rHb (alphaY42D/betaD99N), and rHb (alphaV96W), in the carbonmonoxy and the deoxy forms as well as in the presence and in the absence of an allosteric effector, inositol hexaphosphate (IHP). There are three Trp residues (alpha14, beta15, and beta37) in Hb A for each alphabeta dimer. These Trp residues are located in important regions of the Hb molecule, i.e. alpha14Trp and beta15Trp are located in the alpha(1)beta(1) subunit interface and beta37Trp is located in the alpha(1)beta(2) subunit interface. The relaxation experiments show that amino acid substitutions in the alpha(1)beta(2) subunit interface can alter the dynamics of beta37Trp. The transverse relaxation rate (R(2)) for beta37Trp can serve as a marker for the dynamics of the alpha(1)beta(2) subunit interface. The relaxation parameters of deoxy-rHb Kemspey (betaD99N), which is a naturally occurring abnormal human hemoglobin with high oxygen affinity and very low cooperativity, are quite different from those of deoxy-Hb A, even in the presence of IHP. The relaxation parameters for rHb (alphaY42D/betaD99N), which is a compensatory mutant of rHb Kempsey, are more similar to those of Hb A. In addition, TROSY-CPMG experiments have been used to investigate conformational exchange in the Trp residues of Hb A and the three mutant rHbs. Experimental results indicate that the side-chain of beta37Trp is involved in a relatively slow conformational exchange on the micro- to millisecond time-scale under certain experimental conditions. The present results provide new dynamic insights into the structure-function relationship in hemoglobin.     

Extended Impact of Pin1 Catalytic Loop Phosphorylation Revealed by S71E Phosphomimetic

Pin1 is a two-domain human protein that catalyzes the cis–trans isomerization of phospho-Ser/Thr–Pro (pS/T–P) motifs in numerous cell-cycle regulatory proteins. These pS/T–P motifs bind to Pin1's peptidyl-prolyl isomerase (PPIase) domain in a catalytic pocket, between an extended catalytic loop and the PPIase domain core. Previous studies showed that post-translational phosphorylation of S71 in the catalytic loop decreases substrate binding affinity and isomerase activity. To define the origins for these effects, we investigated a phosphomimetic Pin1 mutant, S71E-Pin1, using solution NMR. We find that S71E perturbs not only its host loop but also the nearby PPIase core. The perturbations identify a local network of hydrogen bonds and salt bridges that is more extended than previously thought, and includes interactions between the catalytic loop and the α2/α3 turn in the PPIase core. Explicit-solvent molecular dynamics simulations and phylogenetic analysis suggest that these interactions act as conserved “latches” between the loop and PPIase core that enhance binding of phosphorylated substrates, as they are absent in PPIases lacking pS/T–P specificity. Our results suggest that S71 is a hub residue within an electrostatic network primed for phosphorylation, and may illustrate a common mechanism of phosphorylation-mediated allostery.     

Structural and dynamics characterization of norovirus protease

Norovirus protease is an essential enzyme for proteolytic maturation of norovirus nonstructural proteins and has been implicated as a potential target for antiviral drug development. Although X‐ray structural studies of the protease give us wealth of structural information including interactions of the protease with its substrate and dimeric overall structure, the role of protein dynamics in the substrate recognition and the biological relevance of the protease dimer remain unclear. Here we determined the solution NMR structure of the 3C‐like protease from Norwalk virus (NV 3CLpro), a prototype strain of norovirus, and analyzed its backbone dynamics and hydrodynamic behavior in solution. ¹⁵N spin relaxation and analytical ultracentrifugation analyses demonstrate that NV 3CLpro is predominantly a monomer in solution. Solution structure of NV 3CLpro shows significant structural variation in C‐terminal domain compared with crystal structures and among lower energy structure ensembles. Also, ¹⁵N spin relaxation and Carr–Purcell–Meiboom–Gill (CPMG)‐based relaxation dispersion analyses reveal the dynamic properties of residues in the C‐terminal domain over a wide range of timescales. In particular, the long loop spanning residues T123–G133 show fast motion (ps‐ns), and the residues in the bII–cII region forming the large hydrophobic pocket (S2 site) undergo conformational exchanges on slower timescales (μs–ms), suggesting their important role in substrate recognition.     

Assessment of the Effects of Increased Relaxation Dispersion Data on the Extraction of 3-site Exchange Parameters Characterizing the Unfolding of an SH3 Domain

Recently a suite of six CPMG relaxation dispersion experiments has been described for quantifying millisecond time-scale exchange processes in proteins. The methodology has been applied to study the folding reaction of a G48M Fyn SH3 domain mutant that exchanges between the native state, and low populated unfolded and intermediate states. A complex non-linear global optimization protocol allows extraction of the kinetics and thermodynamics of the 3-site exchange process from the experimental data, as well as reconstruction of the amide group chemical shifts of the excited states. We show here, through a series of Monte-Carlo simulations on various synthetic data sets, that the 3-site exchange parameters extracted for this system on the basis of ¹⁵N single-quantum (SQ) dispersion profiles exclusively, recorded at a single temperature, are significantly in error. While a temperature dependent ¹⁵N study improves the robustness of extracted parameters, as does a combined analysis of ¹⁵N and ¹H SQ data sets measured at a single temperature, the best agreement is observed in cases where the full complement of six dispersion profiles per residue is analyzed.     

Inherent backbone dynamics fine-tune the functional plasticity of Tudor domains

1. Download : Download high-res image (162KB)
2. Download : Download full-size image

     

Substrate induced structural and dynamics changes in human phosphomevalonate kinase and implications for mechanism

Phosphomevalonate kinase (PMK) catalyzes an essential step in the mevalonate pathway, which is the only pathway for synthesis of isoprenoids and steroids in humans. PMK catalyzes transfer of the gamma-phosphate of ATP to mevalonate 5-phosphate (M5P) to form mevalonate 5-diphosphate. Bringing these phosphate groups in proximity to react is especially challenging, given the high negative charge density on the four phosphate groups in the active site. As such, conformational and dynamics changes needed to form the Michaelis complex are of mechanistic interest. Herein, we report the characterization of substrate induced changes (Mg-ADP, M5P, and the ternary complex) in PMK using NMR-based dynamics and chemical shift perturbation measurements. Mg-ADP and M5P K(d)'s were 6-60 microM in all complexes, consistent with there being little binding synergy. Binding of M5P causes the PMK structure to compress (tau(c) = 13.5 nsec), whereas subsequent binding of Mg-ADP opens the structure up (tau(c) = 15.6 nsec). The overall complex seems to stay very rigid on the psec-nsec timescale with an average NMR order parameter of S(2) approximately 0.88. Data are consistent with addition of M5P causing movement around a hinge region to permit domain closure, which would bring the M5P domain close to ATP to permit catalysis. Dynamics data identify potential hinge residues as H55 and R93, based on their low order parameters and their location in extended regions that connect the M5P and ATP domains in the PMK homology model. Likewise, D163 may be a hinge residue for the lid region that is homologous to the adenylate kinase lid, covering the "Walker-A" catalytic loop. Binding of ATP or ADP appears to cause similar conformational changes; however, these observations do not indicate an obvious role for gamma-phosphate binding interactions. Indeed, the role of gamma-phosphate interactions may be more subtle than suggested by ATP/ADP comparisons, because the conservative O to NH substitution in the beta-gamma bridge of ATP causes a dramatic decrease in affinity and induces few chemical shift perturbations. In terms of positioning of catalytic residues, binding of M5P induces a rigidification of Gly21 (adjacent to the catalytically important Lys22), although exchange broadening in the ternary complex suggests some motion on a slower timescale does still occur. Finally, the first nine residues of the N-terminus are highly disordered, suggesting that they may be part of a cleavable signal or regulatory peptide sequence.     

Comparison of the dynamics of the membrane-bound form of fd coat protein in micelles and in bilayers by solution and solid-state nitrogen-15 nuclear magnetic resonance spectroscopy

Solid-state and solution 15N nuclear magnetic resonance experiments on uniformly and specifically 15N labeled coat protein in phospholipid bilayers and in detergent micelles are used to describe the dynamics of the membrane-bound form of the protein. The residues in the N- and C-terminal portions of the coat protein in both phospholipid bilayers and in detergent micelles are mobile, while those in the hydrophobic midsection are immobile. There is evidence for a gradient of mobility in the C-terminal region of the coat protein in micelles; at 25 degrees C only the last two residues are mobile on the 10(9)-Hz timescale, while the last six to eight residues appear to be mobile on slower timescales and highly mobile at higher temperatures. Since all of the C-terminal residues are immobile in the virus particles, the mobility of these residues in the membrane-bound form of the protein may be important for the formation of protein-DNA interactions in the assembly process.     

NMR studies of structure, hydrogen exchange, and main-chain dynamics in a disrupted-core mutant of thioredoxin.

Core-packing mutants of proteins often approach molten globule states, and hence may have attributes of folding intermediates. We have studied a core-packing mutant of thioredoxin, L78K, in which a leucine residue is substituted by lysine, using 15N heteronuclear two- and three-dimensional NMR. Chemical shift differences between the mutant and wild-type main-chain resonances reveal that structural changes caused by the mutation are localized within 12 A of the altered side chain. The majority of resonances are unchanged, as are many 1H-1H NOEs indicative of the main-chain fold, suggesting that the structure of L78K is largely similar to wild type. Hydrogen exchange studies reveal that residues comprising the central beta-sheet of both mutant and wild-type proteins constitute a local unfolding unit, but with the unfolding/folding equilibrium approximately 12 times larger in L78K. The dynamics of main-chain NH bonds in L78K were studied by 15N spin relaxation and compared with a previous study of wild type. Order parameters for angular motion of NH bonds in the mutant are on average lower than in wild type, suggesting greater spatial freedom on a rapid time scale, but may also be related to different rotational correlation times in the two proteins. There is also evidence of greater conformational exchange in the mutant. Differences between mutant and wild type in hydrogen exchange and main-chain dynamics are not confined to the vicinity of the mutation. We infer that mispacking of the protein core in one location affects local dynamics and stability throughout.     

Internal dynamics of d(CGCAAATTTGCG)2: a comparison of NMR relaxation measurements with a molecular dynamics simulation

We report the analysis of a 250 ps molecular dynamics simulation of the dodecamer d(CGCAAATTT-GCG)₂ immersed in a rectangular box of 3469 water molecules with 22 Na⁺ counterions. The internal dynamics of the molecule were investigated by studying the relevant autocorrelation functions related to the ¹³C-NMR relaxation parameters of the C1′-H1′ bonds of the sugar rings. The calculated effective correlation times τ _e (∼13 ps) and the order parameter S² (∼0.82) of the Lipari and Szabo formalism (Lipari and Szabo 1982a, b) are in satisfactory agreement with those determined previously by NMR (Gaudin et al. 1995, 1996). ¹H-¹H NOE buildups have also been measured experimentally and agree with those computed from the simulation. These results validate the simulation, and a more detailed analysis of the internal dynamics of the dodecamer was undertaken. Analysis of the distributions and of the autocorrelation functions of the glycosidic angle flucuations χ shows that the rotational motion of the sugar rings about their glycosidic bond conforms to a restricted diffusion mechanism. The amplitude of the motions and the diffusion constant are 20° and 17.10⁹ rad²s^–1 respectively. These values are in good agreement with ¹³C NMR data. Furthermore the simulation allows us to rule out another model also consistent with the experiment, consisting of a two-state jump between a syn and an anti conformation. Received: 19 November 1996 / Accepted: 17 March 1997     

Native state fluctuations in a peroxiredoxin active site match motions needed for catalysis

1. Download : Download high-res image (171KB)
2. Download : Download full-size image

     

Propagation of experimental uncertainties using the Lipari-Szabo model-free analysis of protein dynamics

In this paper we make use of the graphical procedure previously described [Jin, D. et al. (1997) J. Am. Chem. Soc., 119, 6923–6924] to analyze NMR relaxation data using the Lipari-Szabo model-free formalism. The graphical approach is advantageous in that it allows the direct visualization of the experimental uncertainties in the motional parameter space. Some general rules describing the relationship between the precision of the relaxation measurements and the precision of the model-free parameters and how this relationship changes with the overall tumbling time (m) are summarized. The effect of the precision in the relaxation measurements on the detection of internal motions not close to the extreme narrowing limit is analyzed. We also show that multiple timescale internal motions may be obscured by experimental uncertainty, and that the collection of relaxation data at very high field strength can improve the ability to detect such deviations from the simple Lipari-Szabo model.     

Backbone dynamics measurements on leukemia inhibitory factor, a rigid four-helical bundle cytokine

The backbone dynamics of the four-helical bundle cytokine leukemia inhibitory factor (LIF) have been investigated using 15N NMR relaxation and amide proton exchange measurements on a murine-human chimera, MH35-LIF. For rapid backbone motions (on a time scale of 10 ps to 100 ns), as probed by 15N relaxation measurements, the dynamics parameters were calculated using the model-free formalism incorporating the model selection approach. The principal components of the inertia tensor of MH35-LIF, as calculated from its NMR structure, were 1:0.98:0.38. The global rotational motion of the molecule was, therefore, assumed to be axially symmetric in the analysis of its relaxation data. This yielded a diffusion anisotropy D(parallel)/D(perpendicular) of 1.31 and an effective correlation time (4D(perpendicular) + 2D(parallel))(-1) of 8.9 ns. The average values of the order parameters (S2) for the four helices, the long interhelical loops, and the N-terminus were 0.91, 0.84, and 0.65, respectively, indicating that LIF is fairly rigid in solution, except at the N-terminus. The S2 values for the long interhelical loops of MH35-LIF were higher than those of their counterparts in short-chain members of the four-helical bundle cytokine family. Residues involved in LIF receptor binding showed no consistent pattern of backbone mobilities, with S2 values ranging from 0.71 to 0.95, but residues contributing to receptor binding site III had relatively lower S2 values, implying higher amplitude motions than for the backbone of sites I and II. In the relatively slow motion regime, backbone amide exchange measurements showed that a number of amides from the helical bundle exchanged extremely slowly, persisting for several months in 2H2O at 37 degrees C. Evidence for local unfolding was considered, and correlations among various structure-related parameters and the backbone amide exchange rates were examined. Both sets of data concur in showing that LIF is one of the most rigid four-helical bundle cytokines.     

Fast structural dynamics in reduced and oxidized cytochrome c

The sub‐nanosecond structural dynamics of reduced and oxidized cytochrome c were characterized. Dynamic properties of the protein backbone measured by amide ¹⁵N relaxation and side chains measured by the deuterium relaxation of methyl groups change little upon change in the redox state. These results imply that the solvent reorganization energy associated with electron transfer is small, consistent with previous theoretical analyses. The relative rigidity of both redox states also implies that dynamic relief of destructive electron transfer pathway interference is not operational in free cytochrome c.     

Side-chain dynamics of the SAP SH2 domain correlate with a binding hot spot and a region with conformational plasticity

X-linked lymphoproliferative disease is caused by mutations in the protein SAP, which consists almost entirely of a single SH2 domain. SAP interacts with the Tyr281 site of the T<-->B cell signaling protein SLAM via its SH2 domain. Interestingly, binding is not dependent on phosphorylation but does involve interactions with residues N-terminal to the Tyr. We have used 15N and 2H NMR relaxation experiments to investigate the motional properties of the SAP SH2 domain backbone amides and side-chain methyl groups in the free protein and complexes with phosphorylated and non-phosphorylated peptides derived from the Tyr281 site of SLAM. The most mobile methyl groups are in side-chains with large RMSD values between the three crystal structures of SAP, suggesting that fast time-scale dynamics in side-chains is associated with conformational plasticity. The backbone amides of two residues which interact with the C-terminal part of the peptides experience fast time-scale motions in the free SH2 domain that are quenched upon binding of either the phosphorylated or non-phosphorylated peptide. Of most importance, the mobility of methyl groups in and around the binding site for residues in the N-terminus of the peptide is significantly restricted in the complexes, underscoring the dominance of this interaction with SAP and demonstrating a correlation between changes in rapid side-chain motion upon binding with local binding energy.     

Backbone dynamics of the human CC-chemokine eotaxin

Eotaxin is a CC chemokine with potent chemoattractant activity towards eosinophils. ¹⁵N NMR relaxation data have been used to characterize the backbone dynamics of recombinant human eotaxin. ¹⁵N longitudinal (R₁) and transverse (R₂) auto relaxation rates, heteronuclear ¹H-¹⁵N steady-state NOEs, and transverse cross-relaxation rates (_xy) were obtained at 30 °C for all resolved backbone secondary amide groups using ¹ H-detected two-dimensional NMR experiments. Ratios of transverse auto and cross relaxation rates were used to identify NH groups influenced by slow conformational rearrangement. Relaxation data were fit to the extended model free dynamics formalism, yielding parameters describing axially symmetric molecular rotational diffusion and the internal dynamics of each NH group. The molecular rotational correlation time (_m) is 5.09±0.02 ns, indicating that eotaxin exists predominantly as a monomer under the conditions of the NMR study. The ratio of diffusion rates about unique and perpendicular axes (D/D) is 0.81±0.02. Residues with large amplitudes of subnanosecond motion are clustered in the N-terminal region (residues 1–19), the C-terminus (residues 68–73) and the loop connecting the first two -strands (residues 30–37). N-terminal flexibility appears to be conserved throughout the chemokine family and may have implications for the mechanism of chemokine receptor activation. Residues exhibiting significant dynamics on the microsecond–millisecond time scale are located close to the two conserved disulfide bonds, suggesting that these motions may be coupled to disulfide bond isomerization.