CPMG sequences with enhanced sensitivity to chemical exchange.

Improved relaxation-compensated Carr-Purcell-Meiboom-Gill pulse sequences are reported for studying chemical exchange of backbone 15N nuclei. In contrast to the original methods [J. P. Loria, M. Rance, and A. G. Palmer, J. Am. Chem. Soc. 121, 2331-2332 (1999)], phenomenological relaxation rate constants obtained using the new sequences do not contain contributions from 1H-1H dipole-dipole interactions. Consequently, detection and quantification of chemical exchange processes are facilitated because the relaxation rate constant in the limit of fast pulsing can be obtained independently from conventional 15N spin relaxation measurements. The advantages of the experiments are demonstrated using basic pancreatic trypsin inhibitor.

The carboxyl terminus of phage HK022 Nun includes a novel zinc-binding motif and a tryptophan required for transcription termination.

The amino-terminal arginine-rich motif of the phage HK022 Nun protein binds phage lambda nascent mRNA transcripts while the carboxy-terminal domain binds RNA polymerase and arrests transcription. The role of specific residues in the carboxy-terminal domain in transcription termination were investigated by mutagenesis, in vitro and in vivo functional assays, and NMR spectroscopy. Coordination of zinc to three histidine residues in the carboxy-terminus inhibited RNA binding by the amino-terminal domain; however, only two of these histidines were required for transcription arrest. These results suggest that additional zinc-coordinating residues are supplied by RNA polymerase in the context of the Nun-RNA polymerase complex. Substitution of the penultimate carboxy-terminal tryptophan residue with alanine or leucine blocks transcription arrest, whereas a tyrosine substitution is innocuous. Wild-type Nun fails to arrest transcription on single-stranded templates. These results suggest that Nun inhibition of transcription elongation is due in part to interactions between the carboxy-terminal tryptophan of Nun and double-stranded DNA, possibly by intercalation. A model for the termination activity of Nun is developed on the basis of these data.

A TROSY CPMG sequence for characterizing chemical exchange in large proteins.

A new NMR spin relaxation experiment is described for measuring chemical exchange time constants from approximately 0.5 ms to 5 ms in 15N-labeled macromolecules. The pulse sequence is based on the Carr-Purcell-Meiboom-Gill technique [Carr and Purcell (1954) Phys. Rev., 94, 630-638; Meiboom and Gill (1958) Rev. Sci. Instrum., 29, 688-691; Loria et al. (1999) J. Am. Chem. Soc., 121, 2331-2332], but implements TROSY selection [Pervushin et al. (1997) Proc. Natl. Acad. Sci. USA, 94, 12366-12371] to permit measurement of exchange linebroadening contributions to the narrower component of the 1H-15N scalar-coupled doublet. This modification extends the size limitation imposed on relaxation measurements due to the fast decay of transverse magnetization in larger macromolecules. The new TROSY-CPMG experiment is demonstrated on a [U-98% 15N] labeled sample of basic pancreatic trypsin inhibitor and a [U-83% 2H, U-98% 15N] labeled sample of triosephosphate isomerase, a 54 kDa homodimeric protein.

Transverse-relaxation-optimized (TROSY) gradient-enhanced triple-resonance NMR spectroscopy.

Two modifications to sensitivity-enhanced gradient-selected TROSY-based triple-resonance NMR experiments are proposed that reduce the overall duration of the pulse sequences and minimize radiation damping effects on water-flipback solvent suppression. The modifications are illustrated for the HNCO-TROSY experiment, but are applicable to all triple-resonance experiments that detect proton magnetization after a reverse polarization transfer step from a (15)N spin. The methods are applied to yeast triosephosphate isomerase, a symmetric dimer with 248 amino acid residues per monomer.

Backbone dynamics of a short PU.1 ETS domain.

The resonance assignments, secondary structure and backbone dynamics of the ETS domain of the transcription factor PU.1 have been determined for the free protein in solution by NMR spectroscopy. The secondary structure for the free ETS domain is similar to that observed in the crystal structure of the PU.1 protein complexed with DNA, except that helix alpha2 and recognition helix alpha3 are shorter for the free protein in solution. Backbone dynamics of the protein have been examined using amide hydrogen-deuterium exchange and (15)N laboratory-frame spin relaxation measurements. A significant probability of local unfolding of helix alpha2, which precedes the loop-helix-loop DNA recognition domain, is inferred from the very fast hydrogen-deuterium exchange for amide protons in this helix. The (15)N relaxation measurements indicate that the protein is partially oligomerized at a concentration of 2.5 mM, but monomeric at a concentration of 0.3 mM. The (15)N relaxation data for the low concentration sample were interpreted, using the model-free formalism, to provide insight into protein dynamics on picosecond-nanosecond and microsecond-millisecond time scales. High flexibility of the protein backbone is observed for the residues in the loop between alpha2 and alpha3. This loop is variable in length and in structure within the class of winged helix proteins and is partially responsible for binding to DNA. The dynamic properties observed for alpha2, alpha3 and the intervening loop may indicate a correlation between protein plasticity in particular structural elements and recognition of specific DNA sequences.

Temperature dependence of intramolecular dynamics of the basic leucine zipper of GCN4: implications for the entropy of association with DNA.

The basic leucine zipper domain of the yeast transcription factor GCN4 consists of a C-terminal leucine zipper and an N-terminal basic DNA-binding region that achieves a stable structure only after association with DNA. Backbone dynamics of a peptide encompassing the basic and leucine zipper bZip domain (residues 226-281) are described using NMR spectroscopy. The 15N longitudinal relaxation rates, 15N transverse relaxation rates, and {1H}-15N nuclear Overhauser effects were measured for the backbone amide nitrogen atoms at 290 K, 300 K, and 310 K. The relaxation data were interpreted using reduced spectral density mapping to determine values of the spectral density function, J(omega), at the frequencies 0, omegaN, and 0.87omegaH to characterize overall and intramolecular motions on picosecond-nanosecond timescales. To account for the temperature dependence of overall rotational diffusion, the J(0) values were normalized using Stoke's Law. At 310 K, the 13Calpha and 13CO chemical shifts in conjunction with the spectral density values indicate that the leucine zipper sequence forms a highly ordered alpha-helix, while the basic region populates an ensemble of highly dynamic transient structures with substantial helical character. The normalized values of J(0) and the values of J(0.87omegaH) for residues in the leucine zipper dimerization domain are independent of temperature. In contrast, residues in the basic region exhibit pronounced increases in the normalized J(0) and decreases in J(0.87omegaH) as temperature is decreased. A strong correlation exists between the temperature dependence of 13CO chemical shifts and of J(0.87omegaH). These results suggest that, for the basic region, lowering the temperature increases the population of transient helical conformations, and concomitantly reduces the amplitude or timescale of conformational fluctuations on picosecond-nanosecond timescales. Changes in the conformational dynamics of the peptide backbone of the basic region that accompany DNA binding contribute to the overall thermodynamics of complex formation. The change in backbone conformational entropy derived from NMR spin-relaxation data agrees well with the result calculated from calorimetric measurements. Restriction of the conformational space accessible to the basic region may significantly reduce the entropic cost associated with formation of the basic region helices consequent to DNA binding.

CO_H(N)CACB experiments for assigning backbone resonances in 13C/15N-labeled proteins.

A triple resonance NMR experiment, denoted CO_H(N)CACB, correlates 1HN and 13CO spins with the 13C alpha and 13C beta spins of adjacent amino acids. The pulse sequence in an 'out-and-back' design that starts with 1HN magnetization and transfers coherence via the 15N spin simultaneously to the 13CO and 13C alpha spins, followed by transfer to the 13C beta spin. Two versions of the sequence are presented: one in which the 13CO spins are frequency labeled during an incremented t1 evolution period prior to transfer of magnetization from the 13C alpha to the 13C beta resonances, and one in which the 13CO spins are frequency labeled in a constant-time manner during the coherence transfer to and from the 13C beta resonances. Because 13CO and 15N chemical shifts are largely uncorrelated, the technique will be especially useful when degeneracy in the 1Hn-15N chemical shifts hinders resonance assignment. The CO_H(N)CACB experiment is demonstrated using uniformly 13C/15N-labeled ubiquitin.

3D accordion spectroscopy for measuring 15N and 13CO relaxation rates in poorly resolved NMR spectra.

An experimental approach for the measurement of nuclear magnetic spin relaxation rate constants that combines triple-resonance techniques and accordion spectroscopy is described. Pulse sequences are discussed for the measurement of backbone 15N and 13CO R1 relaxation rate constants. The three-dimensional HNCO triple-resonance technique is employed to gain improved spectral resolution over conventional two-dimensional methods by frequency labeling both the 15N and 13CO spins. Accordion spectroscopy is used to reduce the dimensionality of the relaxation experiment. The "negative-time accordion" approach (A. M. Mandel and A. G. Palmer (1994), J. Magn. Reson. A 110, 62-72) is used for extracting rate constants from the t1 interferograms. The experiments are demonstrated using a 13C/15N isotopically enriched sample of the third fibronectin type III domain of human tenascin.

Backbone dynamics of the EGF-like domain of heregulin-alpha.

The backbone dynamics of the 63 residue epidermal growth factor (EGF)-like domain of heregulin-alpha (HRG-alpha) have been characterized by measurement of longitudinal relaxation rate constants (R1), transverse relaxation rate constants (R2), and steady-state ¿1H¿-15N nuclear Overhauser effects for the 15N nuclear spins using proton-detected heteronuclear NMR spectroscopy. Analysis of the R2/R1 ratios in conjunction with the known structure of the HRG-alpha EGF-like domain yields a rotational correlation time of approximately 8.4 ns, suggesting that the protein aggregates under the solution conditions used (3.8 mM protein, 50 mM sodium acetate, pH 4.5, 20 degreesC), and that it tumbles with an axially symmetric diffusion tensor (D parallel/D perpendicular=1.4). Sedimentation equilibrium experiments confirm that the EGF-like domain of HRG-alpha undergoes weak self-association under these conditions and are consistent with a simple monomer-dimer equilibrium with a dimer-dissociation constant Kd=1.6(+/-0.4) mM. The relaxation data were analyzed using a reduced spectral density mapping approach to avoid systematic effects of aggregation on the usual model-free formalism. The reduced spectral densities show that residues near the N terminus (residues 3 to 5 and 7 to 12), in the Omega-loop between beta-strands 2 and 3 (residues 24 to 31), and in particular the C-terminal 13 residues (residues 51 to 63), have significant mobility on a picosecond/nanosecond time-scale. In addition, conformational exchange on a microsecond time-scale was identified for residues 44 to 46 on the basis of observed differences in R2 at 11.7 and 14.1 T. The mobility identified near the N terminus and in the vicinity of residues 44 to 46 may be important in allowing the interactions of heregulin with multiple receptors.

Pervasive conformational fluctuations on microsecond time scales in a fibronectin type III domain.

A novel off-resonance rotating-frame 15N NMR spin relaxation experiment is used to characterize conformational fluctuations with correlation times between 32 and 175 microseconds in the third fibronectin type III domain of human tenascin-C. Conformational fluctuations of contiguous regions of the beta-sandwich structure of the type III domain may represent collective motions, such as transient twisting or breathing of the beta-sheets. Flexibility of the loop containing the Arg-Gly-Asp (RGD) tripeptide may affect the accessibility of this motif in protein-protein interactions.

Conformation of the trypanocidal pharmaceutical suramin in its free and bound forms: transferred nuclear overhauser studies.

Suramin is a lead compound for treatment of cancer, HIV, and trypanosomiasis. The conformations of suramin in its free form and bound to phosphoglycerate kinases from T. brucei and S. cerevisae, have been studied in aqueous solutions using nuclear Overhauser effect (NOE) and transferred NOE NMR spectroscopies. The NOE data of the free drug can be accommodated by a model in which many of the single bonds of suramin are unrestricted at room temperature, consistent with molecular mechanics calculations. The angle between the naphthalene ring and the adjoining amide is essentially locked by a strong amide-sulfonate hydrogen bond into one preferred conformation. Another degree of freedom near the termini of the molecule has a rather pronounced preference, and a third exhibits a nearly perpendicular arrangement between the amide and adjacent aromatic ring. The other two degrees of freedom have weaker preferences. Molecular mechanics calculations using AMBER force field and charges on amides and sulfonates obtained from semiempirical or ab initio calculations reproduced the extent of nonplanarity but not the detailed preferences. 13C spin-lattice relaxation, proton NOE, and light-scattering measurements for free suramin indicate that the correlation time of the molecule is approximately 3 ns at 5 mM concentration, suggesting that suramin is multimeric. Lowering the temperature to 5 degrees C causes a dramatic broadening of all of the resonances in the NMR spectra of 5 mM suramin. This broadening probably is associated with further aggregation into micelles. Suramin is monomeric at 0.5 mM and room temperature, and the NOE cross-relaxation rate constants are close to the cancellation condition for a 500 MHz proton frequency; this concentration is typical of blood serum concentrations when the drug is utilized in humans. Changes in the conformational preferences for terminal degrees of freedom are observed in the bound states of suramin based upon the transferred NOE data. The data for the bound state cannot be accommodated by a symmetric conformer. Analysis of the transferred NOESY buildup curves indicates complex kinetics of binding, probably involving an electrostatically bound encounter complex. Despite the weak binding constant, the buildup curves cannot be treated as population-weighted averages of the free and bound cross-relaxation rates, and therefore complete relaxation-exchange matrix analysis has been performed to simulate the data sets.

Probing molecular motion by NMR.

Recently developed solution NMR methods for measuring 2H, 13C, and 15N spin relaxation, coupled with biosynthetic isotopic enrichment, permit the characterization of backbone and sidechain dynamical properties of proteins on picosecond/nanosecond and microsecond/millisecond timescales. Theoretical interpretations of the relaxation data provide insights into the biophysical and functional properties of proteins.

Accuracy and precision of NMR relaxation experiments and MD simulations for characterizing protein dynamics.

Model-free parameters obtained from nuclear magnetic resonance (NMR) relaxation experiments and molecular dynamics (MD) simulations commonly are used to describe the intramolecular dynamical properties of proteins. To assess the relative accuracy and precision of experimental and simulated model-free parameters, three independent data sets derived from backbone 15N NMR relaxation experiments and two independent data sets derived from MD simulations of Escherichia-coli ribonuclease HI are compared. The widths of the distributions of the differences between the order parameters for pairs of NMR data sets are congruent with the uncertainties derived from statistical analyses of individual data sets; thus, current protocols for analyzing NMR data encapsulate random uncertainties appropriately. Large differences in order parameters for certain residues are attributed to systematic differences between samples for intralaboratory comparisons and unknown, possibly magnetic field-dependent, experimental effects for interlaboratory comparisons. The widths of distributions of the differences between the order parameters for two NMR sets are similar to widths of distributions for an NMR and an MD set or for two MD sets. The linear correlations between the order parameters for an MD set and an NMR set are within the range of correlations observed between pairs of NMR sets. These comparisons suggest that the NMR and MD generalized order parameters for the backbone amide N-H bond vectors are of comparable accuracy for residues exhibiting motions on a fast time scale (< 100 ps). Large discrepancies between NMR and MD order parameters for certain residues are attributed to the occurrence of "rare" motional events over the simulation trajectories, the disruption of an element of secondary structure in one of the simulations, and lack of consensus among the experimental data sets. Consequently, (easily detectable) severe distortions of local protein structure and infrequent motional events in MD simulations appear to be the most serious artifacts affecting the accuracy and precision, respectively, of MD order parameters relative to NMR values. In addition, MD order parameters for motions on a fast (< 100 ps) timescale are more precisely determined than their NMR counterparts, thereby permitting more detailed dynamic characterization of biologically important residues by MD simulation than is sometimes possible by experimental methods. Proteins 28:481-493, 1997.

Base dynamics in a UUCG tetraloop RNA hairpin characterized by 15N spin relaxation: correlations with structure and stability.

Intramolecular dynamics of guanine and uracil bases in a 14-nt RNA hairpin including the extraordinarily stable UUCG tetraloop were studied by 15N spin relaxation experiments that are sensitive to structural fluctuations occurring on a time scale of picoseconds to nanoseconds. The relaxation data were interpreted in the framework of the anisotropic model-free formalism, using assumed values for the chemical shift anisotropies of the 15N spins. The rotational diffusion tensor was determined to be symmetric with an axial ratio of 1.34 +/- 0.12, in agreement with estimates based on the ratio of the principal moments of the inertia tensor. The model-free results indicate that the bases of the G x U pair in the tetraloop are at least as rigid as the interior base pairs in the stem, whereas the 5'-terminal guanine is more flexible. The observed range of order parameters corresponds to base fluctuations of 19-22 degrees about the chi torsion angle. The results reveal dynamical consequences of the unusual structural features in the UUCG tetraloop and offer insights into the configurational entropy of hairpin formation.

Solution conformations of a peptide containing the cytoplasmic domain sequence of the beta amyloid precursor protein.

The cytoplasmic domain of the beta amyloid precursor protein (betaAPP) may play a role in cellular events that lead to the secretion of the Abeta peptide, the major constituent of amyloid plaques found in the brains of individuals affected by Alzheimer's disease, by interacting with cellular factors involved in betaAPP function or processing. In order to elucidate the structural basis of cytoplasmic domain activity, the conformations adopted in solution by a peptide containing the 47-residue C-terminal sequence of betaAPP have been investigated by NMR and CD spectroscopy. The peptide does not have a stable tertiary structure, but local regions of the polypeptide chain populate defined conformations. In particular, the amino acid sequences TPEE and NPTY form type I reverse turns. These structured regions correspond to sequences within the cytoplasmic domain implicated in the biological activity of betaAPP.

Backbone dynamics of homologous fibronectin type III cell adhesion domains from fibronectin and tenascin.

BACKGROUND: Fibronectin type III domains are found as autonomously-folded domains in a large variety of multidomain proteins, including extracellular matrix proteins. A subset of these domains employ an Arg-Gly-Asp (RGD) tripeptide motif to mediate contact with cell-surface receptors (integrins). This motif mediates protein-protein interactions in a diverse range of biological processes, such as in tissue development, would healing and metastasis. The molecular basis for affinity and specificity of cell adhesion via type III domains has not been clearly established. The tenth type III domain from fibronectin (FNfn10) and the third type III domain from tenascin-C (TNfn3) have 27% sequence identity and share the same overall protein fold, but present the RGD motifs in different structural contexts. The dynamical properties of the RGD motifs may affect the specificity and affinity of the FNfn10 and TNfn3 domains. Structure-dynamics correlations for these structurally homologous proteins may reveal common molecular features which are important to the dynamical properties of proteins. RESULTS: The intramolecular dynamics of the protein backbones of FNfn10 and TNfn3 have been studied by 15N nuclear spin relaxation. The FG loop in FNfn10, which contains the RGD motif, exhibits extensive flexibility on picosecond to nanosecond timescales, but motions on microsecond to millisecond timescales are not observed. The equivalent region in TNfn3 is as rigid as regular elements of secondary structure. The CC' loop also is more flexible on picosecond-nanosecond timescales in FNfn10 than in TNfn3. Conformational exchange, reflecting flexibility on microsecond-millisecond timescales, is observed in beta strands A and B of both FNfn10 and TNfn3. CONCLUSIONS: Comparison of the structures of the FNfn10 and TNfn3 reveals several features related to their different dynamical properties. The larger amplitude motions of loops in FNfn10 are consistent with the hypothesis that flexibility of these regions facilitates induced-fit recognition of fibronectin by multiple receptors. Similarly, the more rigid loops of TNfn3 may reflect greater specificity for particular integrins. The correlations observed between structural features and dynamical properties of the homologous type III domains indicate the influence of hydrogen bonding and hydrophobic packing on dynamical fluctuations in proteins.

Rotational diffusion anisotropy of proteins from simultaneous analysis of 15N and 13C alpha nuclear spin relaxation.

Current methods of determining the rotational diffusion tensors of proteins in solution by NMR spectroscopy exclusively utilize relaxation rate constants for backbone amide 15N spins. However, the distributions of orientations of N-H bond vectors are not isotropic in many proteins, and correlations between bond vector orientations reduce the accuracy and precision of rotational diffusion tensors extracted from 15N spin relaxation data. The inclusion of both 13C alpha and 15N spin relaxation rate constants increases the robustness of the diffusion tensor analysis because the orientations of the C alpha-H alpha bond vectors differ from the orientations of the N-H bond vectors. Theoretical and experimental results for calbindin D9k, granulocyte colony stimulating factor, and ubiquitin, three proteins with different distributions of N-H and C alpha-H alpha bond vectors, are used to illustrate the advantages of the simultaneous utilization of 13C alpha and 15N relaxation data.

(H)N(COCA)NH and HN(COCA)NH experiments for 1H-15N backbone assignments in 13C/15N-labeled proteins.

Triple resonance HN(COCA)NH pulse sequences for correlating 1H(i), 15N(i), 1H(i-1), and 15N(i-1) spins that utilize overlapping coherence transfer periods provide increased sensitivity relative to pulse sequences that utilize sequential coherence transfer periods. Although the overlapping sequence elements reduce the overall duration of the pulse sequences, the principal benefit derives from a reduction in the number of 180 degrees pulses. Two versions of the technique are presented: a 3D (H)N(COCA)NH experiment that correlates 15N(i), 1H(i-1), and 15N(i-1) spins, and a 3D HN(COCA)NH experiment that correlates 1H(i), 15N(i), 1H(i-1), and 15N(i-1) spins by simultaneously encoding the 1H(i) and 15N(i) chemical shifts during the t1 evolution period. The methods are demonstrated on a 13C/15N-enriched sample of the protein ubiquitin and are easily adapted for application to 2H/13C/15N-enriched proteins.

Dynamics of ribonuclease H: temperature dependence of motions on multiple time scales.

The temperature dependence of the backbone motions in Escherichia coli ribonuclease HI was studied on multiple time scales by 15N nuclear magnetic spin relaxation. Laboratory frame relaxation data at 285, 300, and 310 K were analyzed using the model-free and reduced spectral density approaches. The temperature dependence of the order parameters was used to define a characteristic temperature for the motions of the backbone N-H bond vectors on picosecond to nanosecond time scales. The characteristic temperatures for secondary structure elements, loops, and the C-terminus are approximately 1000, approximately 300, and approximately 170 K, respectively. The observed variation in the characteristic temperature indicates that the energy landscape, and thus the configurational heat capacity, is markedly structure dependent in the folded protein. The effective correlation times for internal motions do not show significant temperature dependence. Conformational exchange was observed for a large number of residues forming a contiguous region of the protein that includes the coiled coil formed by helices alpha A and alpha D. Exchange broadening in the CPMG experiments decreased with increased temperature, directly demonstrating that the microscopic exchange rate is faster than the pulse repetition rate of 1.2 ms. The temperature dependence of the exchange contributions to the transverse relaxation rate constant shows approximately Arrhenius behavior over the studied temperature range with apparent activation enthalpies of approximately 20-50 kJ/mol. Numerical calculations suggest that these values underestimate the activation barriers by at most a factor of 2. The present results obtained at 300 K are compared to those reported previously [Mandel, A. M., Akke, M., & Palmer, A. G., III (1995) J. Mol. Biol. 246, 144-163] to establish the reproducibility of the experimental techniques.