NOE Constraints Research Articles

NMR Data-Driven Docking of HDM2-Inhibitor Complexes.

We present an automated NMR-guided docking workflow that can be used to generate models of protein-ligand complexes based on data from NOE NMR experiments. The first step is to generate a number of intermolecular distance constraints from experimental NOE data. Then, the ligand is docked on an ensemble of receptor structures to account for protein flexibility, and multiple poses are generated. Finally, we use the NOE-based constraints to filter and score docking poses based on the percentage of NOE constraints that are consistent with protein-ligand interatomic distances. This workflow was successfully used during a lead optimization project to generate models of synthetic protein-protein interaction (PPI) inhibitors bound to the HDM2 protein.

Guiding automated NMR structure determination using a global optimization metric, the NMR DP score.

ASDP is an automated NMR NOE assignment program. It uses a distinct bottom-up topology-constrained network anchoring approach for NOE interpretation, with 2D, 3D and/or 4D NOESY peak lists and resonance assignments as input, and generates unambiguous NOE constraints for iterative structure calculations. ASDP is designed to function interactively with various structure determination programs that use distance restraints to generate molecular models. In the CASD-NMR project, ASDP was tested and further developed using blinded NMR data, including resonance assignments, either raw or manually-curated (refined) NOESY peak list data, and in some cases (15)N-(1)H residual dipolar coupling data. In these blinded tests, in which the reference structure was not available until after structures were generated, the fully-automated ASDP program performed very well on all targets using both the raw and refined NOESY peak list data. Improvements of ASDP relative to its predecessor program for automated NOESY peak assignments, AutoStructure, were driven by challenges provided by these CASD-NMR data. These algorithmic improvements include (1) using a global metric of structural accuracy, the discriminating power score, for guiding model selection during the iterative NOE interpretation process, and (2) identifying incorrect NOESY cross peak assignments caused by errors in the NMR resonance assignment list. These improvements provide a more robust automated NOESY analysis program, ASDP, with the unique capability of being utilized with alternative structure generation and refinement programs including CYANA, CNS, and/or Rosetta.

Utilization of paramagnetic relaxation enhancements for high-resolution NMR structure determination of a soluble loop-rich protein with sparse NOE distance restraints.

NMR structure determination of soluble proteins depends in large part on distance restraints derived from NOE. In this study, we examined the impact of paramagnetic relaxation enhancement (PRE)-derived distance restraints on protein structure determination. A high-resolution structure of the loop-rich soluble protein Sin1 could not be determined by conventional NOE-based procedures due to an insufficient number of NOE restraints. By using the 867 PRE-derived distance restraints obtained from the NOE-based structure determination procedure, a high-resolution structure of Sin1 could be successfully determined. The convergence and accuracy of the determined structure were improved by increasing the number of PRE-derived distance restraints. This study demonstrates that PRE-derived distance restraints are useful in the determination of a high-resolution structure of a soluble protein when the number of NOE constraints is insufficient.

E. Coli F1Fo ATP Synthase Subunit C: Solution NMR Structure by CS-Rosetta

The ATP synthase is the main source of ATP in cells, whose function is to inter-convert chemical energy and transmembrane electrochemical potential difference. The enzyme contains two domains: a soluble F1 domain where ATP is synthesized/hydrolyzed and a transmembrane FO domain, which functions as an ion pump. Multiple (9-15) copies of subunit c form an oligomeric ring in the FO domain. The ion translocation activity takes place at the interface of c-ring and subunit a where an essential acidic residue in middle of the outer helix in subunit c undergoes a protonation/deprotonation process, which couples with the relative rotation of c-ring to subunit a. Although high-resolution X-ray structures of the c-rings from several organisms have been solved, the mechanism of ion translocation still needs to be clarified at atomic resolution.We aim to solve the structures of E. coli subunit c at both protonation states of the essential Asp61 and elucidate any conformational changes at the active site during the ion translocation process. Due to the difficulty of obtaining unambiguous long-range NOE constraints in membrane proteins, traditional solution NMR methods were unsuccessful in this case. We employed CS-Rosetta, which utilizes minimal NMR constraints, in our study.Our study of the subunit c in the protonated state shows that the choice of proper scoring weights is essential in the application of CS-Rosetta to membrane proteins. Using limited NMR restraints this method converges on a structure for E. coli subunit c that is similar to the X-ray structures of subunit c from other organisms. Further improving the quality of the structure may be achieved by optimizing the weights in the calculation. We are also using this approach on deprotonated subunit c and hoping to model the c-ring from our results.

Further optimization of a hybrid united-atom and coarse-grained force field for folding simulations: Improved backbone hydration and interactions between charged side chains.

PACE, a hybrid force field which couples united-atom protein models with coarse-grained (CG) solvent, has been further optimized, aiming to improve itse ciency for folding simulations. Backbone hydration parameters have been re-optimized based on hydration free energies of polyalanyl peptides through atomistic simulations. Also, atomistic partial charges from all-atom force fields were combined with PACE in order to provide a more realistic description of interactions between charged groups. Using replica exchange molecular dynamics (REMD), ab initio folding using the new PACE has been achieved for seven small proteins (16 - 23 residues) with different structural motifs. Experimental data about folded states, such as their stability at room temperature, melting point and NMR NOE constraints, were also well reproduced. Moreover, a systematic comparison of folding kinetics at room temperature has been made with experiments, through standard MD simulations, showing that the new PACE may speed up the actual folding kinetics 5-10 times. Together with the computational speedup benefited from coarse-graining, the force field provides opportunities to study folding mechanisms. In particular, we used the new PACE to fold a 73-residue protein, 3D, in multiple 10 - 30 μs simulations, to its native states (C(α) RMSD ~ 0.34 nm). Our results suggest the potential applicability of the new PACE for the study of folding and dynamics of proteins.

Structurally Diverse Cyclisation Linkers Impose Different Backbone Conformations in Bicyclic Peptides

Combinatorial libraries of structurally diverse peptide macrocycles offer a rich source for the development of high-affinity ligands to targets of interest. In this work we have developed linkers for the generation of genetically encoded bicyclic peptides and tested whether the peptides cyclised by them have significant variations in their backbone conformations. Two new cyclisation reagents, each containing three thiol-reactive groups, efficiently and selectively cyclised linear peptides containing three cysteine moieties. When the mesitylene linker of the bicyclic peptide PK15, a potent inhibitor of plasma kallikrein (K(i)=2 nM), was replaced by the new linkers, its inhibitory activity dropped by a factor of more than 1000, suggesting that the linkers impose different conformations on the peptide. Indeed, structural analysis by solution-state NMR revealed different NOE constraints in the three bicyclic peptides, indicating that these relatively small linkers at the centres of bicyclic peptide structures significantly influence the conformations of the peptides. These results demonstrate the prominent structural role of linkers in peptide macrocycles and suggest that application of different cyclisation linkers in a combinatorial fashion could be an attractive means to generate topologically diverse macrocycle libraries.

Experimentally assessing molecular dynamics sampling of the protein native state conformational distribution

The acute sensitivity to conformation exhibited by amide hydrogen exchange reactivity provides a valuable test for the physical accuracy of model ensembles developed to represent the Boltzmann distribution of the protein native state. A number of molecular dynamics studies of ubiquitin have predicted a well-populated transition in the tight turn immediately preceding the primary site of proteasome-directed polyubiquitylation Lys 48. Amide exchange reactivity analysis demonstrates that this transition is 103-fold rarer than these predictions. More strikingly, for the most populated novel conformational basin predicted from a recent 1ms MD simulation of bovine pancreatic trypsin inhibitor (at 13% of total), experimental hydrogen exchange data indicates a population below 10−6. The most sophisticated efforts to directly incorporate experimental constraints into the derivation of model protein ensembles have been applied to ubiquitin, as illustrated by three recently deposited studies (PDB codes 2NR2, 2K39 and 2KOX2K392KOX). Utilizing the extensive set of experimental NOE constraints, each of these three ensembles yields a modestly more accurate prediction of the exchange rates for the highly exposed amides than does a standard unconstrained molecular simulation. However, for the less frequently exposed amide hydrogens, the 2NR2 ensemble offers no improvement in rate predictions as compared to the unconstrained MD ensemble. The other two NMR-constrained ensembles performed markedly worse, either underestimating (2KOX) or overestimating (2K39) the extent of conformational diversity.

Self-Assembly of a Designed Amyloid Peptide Containing the Functional Thienylalanine Unit

The self-assembly of a peptide based on a sequence from the amyloid beta peptide but incorporating the non-natural amino acid beta-2-thienylalanine (2-Thi) has been investigated in aqueous and methanol solutions. The peptide AAKLVFF was used as a design motif, replacing the phenylalanine residues (F) with 2-Thi units to yield (2-Thi)(2-Thi)VLKAA. The 2-Thi residues are expected to confer interesting electronic properties due to charge delocalization and pi-stacking. The peptide is shown to form beta-sheet-rich amyloid fibrils with a twisted morphology, in both water and methanol solutions at sufficiently high concentration. The formation of a self-assembling hydrogel is observed at high concentration. Detailed molecular modeling using molecular dynamics methods was performed using NOE constraints provided by 2D-NMR experiments. The conformational and charge properties of 2-Thi were modeled using quantum mechanical methods, and found to be similar to those previously reported for the beta-3-thienylalanine analogue. The molecular dynamics simulations reveal well-defined folded structures (turn-like) in dilute aqueous solution, driven by self-assembly of the hydrophobic aromatic units, with charged lysine groups exposed to water.

Sequential nearest-neighbor effects on computed 13Cα chemical shifts

To evaluate sequential nearest-neighbor effects on quantum-chemical calculations of 13Cα chemical shifts, we selected the structure of the nucleic acid binding (NAB) protein from the SARS coronavirus determined by NMR in solution (PDB id 2K87). NAB is a 116-residue α/β protein, which contains 9 prolines and has 50% of its residues located in loops and turns. Overall, the results presented here show that sizeable nearest-neighbor effects are seen only for residues preceding proline, where Pro introduces an overestimation, on average, of 1.73 ppm in the computed 13Cα chemical shifts. A new ensemble of 20 conformers representing the NMR structure of the NAB, which was calculated with an input containing backbone torsion angle constraints derived from the theoretical 13Cα chemical shifts as supplementary data to the NOE distance constraints, exhibits very similar topology and comparable agreement with the NOE constraints as the published NMR structure. However, the two structures differ in the patterns of differences between observed and computed 13Cα chemical shifts, Δca,i, for the individual residues along the sequence. This indicates that the Δca,i -values for the NAB protein are primarily a consequence of the limited sampling by the bundles of 20 conformers used, as in common practice, to represent the two NMR structures, rather than of local flaws in the structures.

Application of SAIL phenylalanine and tyrosine with alternative isotope-labeling patterns for protein structure determination

The extensive collection of NOE constraint data involving the aromatic ring signals is essential for accurate protein structure determination, although it is often hampered in practice by the pervasive signal overlapping and tight spin couplings for aromatic rings. We have prepared various types of stereo-array isotope labeled phenylalanines (epsilon- and zeta-SAIL Phe) and tyrosine (epsilon-SAIL Tyr) to overcome these problems (Torizawa et al. 2005), and proven that these SAIL amino acids provide dramatic spectral simplification and sensitivity enhancement for the aromatic ring NMR signals. In addition to these SAIL aromatic amino acids, we recently synthesized delta-SAIL Phe and delta-SAIL Tyr, which allow us to observe and assign delta-(13)C/(1)H signals very efficiently. Each of the various types of SAIL Phe and SAIL Tyr yields well-resolved resonances for the delta-, epsilon- or zeta-(13)C/(1)H signals, respectively, which can readily be assigned by simple and robust pulse sequences. Since the delta-, epsilon-, and zeta-proton signals of Phe/Tyr residues give rise to complementary NOE constraints, the concomitant use of various types of SAIL-Phe and SAIL-Tyr would generate more accurate protein structures, as compared to those obtained by using conventional uniformly (13)C, (15)N-double labeled proteins. We illustrated this with the case of an 18.2 kDa protein, Escherichia coli peptidyl-prolyl cis-trans isomerase b (EPPIb), and concluded that the combined use of zeta-SAIL Phe and epsilon-SAIL Tyr would be practically the best choice for protein structural determinations.

NOEnet–Use of NOE networks for NMR resonance assignment of proteins with known 3D structure

Motivation: A prerequisite for any protein study by NMR is the assignment of the resonances from the 15N−1H HSQC spectrum to their corresponding atoms of the protein backbone. Usually, this assignment is obtained by analyzing triple resonance NMR experiments. An alternative assignment strategy exploits the information given by an already available 3D structure of the same or a homologous protein. Up to now, the algorithms that have been developed around the structure-based assignment strategy have the important drawbacks that they cannot guarantee a high assignment accuracy near to 100%.Results: We propose here a new program, called NOEnet, implementing an efficient complete search algorithm that ensures the correctness of the assignment results. NOEnet exploits the network character of unambiguous NOE constraints to realize an exhaustive search of all matching possibilities of the NOE network onto the structural one. NOEnet has been successfully tested on EIN, a large protein of 28 kDa, using only NOE data. The complete search of NOEnet finds all possible assignments compatible with experimental data that can be defined as an assignment ensemble. We show that multiple assignment possibilities of large NOE networks are restricted to a small spatial assignment range (SAR), so that assignment ensembles, obtained from accessible experimental data, are precise enough to be used for functional proteins studies, like protein–ligand interaction or protein dynamics studies. We believe that NOEnet can become a major tool for the structure-based backbone resonance assignment strategy in NMR.Availability: The NOEnet program will be available under: http://www.icsn.cnrs-gif.fr/download/nmrContact: carine@icsn.cnrs-gif.fr; eric.guittet@icsn.cnrs-gif.frSupplementary Information: Supplementary data are available at Bioinformatics online.

Amino Acids Important for DNA Recognition by the Response Regulator OmpR

Response regulators undergo regulated phosphorylation and dephosphorylation at conserved aspartic acid residues in bacterial signal transduction systems. OmpR is a winged helix-turnhelix DNA-binding protein that functions as a global regulator in bacteria and is also important in pathogenesis. A detailed mechanistic picture of how OmpR binds to DNA and activates transcription is lacking. We used NMR spectroscopy to solve the solution structure of the C-terminal domain of OmpR (OmpR(C)) and to analyze the chemical shift changes that occur upon DNA binding. There is little overlap in the interaction surface with residues of PhoB that were reportedly involved in protein/protein interactions in its head-to-tail dimer. Multiple factors account for the lack of overlap. One is that the spacing between the OmpR half-sites is shorter than observed with PhoB, requiring the arrangement of the two OmpR molecules to be different from that of the PhoB dimer on DNA. A second is the demonstration herein that OmpR can bind to its high affinity site as a monomer. As a result, OmpR(C) appears to be capable of adopting alternative orientations depending on the precise base composition of the binding site, which also contributes to the lack of overlap. In the presence of DNA, chemical shift changes occur in OmpR in the recognition alpha-helix 3, the loop between beta-strand 4 and alpha-helix 1, and the loop between beta-strands 5 and 6. DNA contact residues are Val(203) (T), Arg(207) (G), and Arg(209) (phosphate backbone). Our results suggest that OmpR binds to DNA as a monomer and then forms a symmetric or asymmetric dimer, depending on the binding site. We propose that during activation OmpR binds to DNA and undergoes a conformational change that promotes phosphorylation of the N-terminal receiver domain, the receiver domains dimerize, and then the second monomer binds to DNA. The flexible linker of OmpR enables the second monomer to bind in multiple orientations (head-to-tail and head-to-head), depending on the specific DNA contacts.

In silico study on the effect of F19T mutation on amyloid-b peptide (10-35)

The wild type (WT) amyloid-beta (10-35) peptide, Abeta (10-35), and its F19T mutant have been studied by molecular dynamics simulations at 340 K in explicit water solvent each for over 3.4 ms. The WT peptide has a strong preference to form an E22-K28 loop (44% of total conformations) and a reasonable stability for a strand-loop-strand (SLS, L17-M35) (9%). The F19T mutant has a significantly lower population of E22-K28 loop (14%) and SLS structure (1.7%), but has a high population of Q15-D23 loop (48%). A specific interaction pattern among D23, V24, E22 and K28 was found to stabilize the E22-K28 loop in WT. Our results are in agreement with several experimental observations including: (1) the NOE constraints for the Abeta are reproduced; (2) the regions (15-23) and (22-28) can form loops; (3) the WT peptide is more structured than the F19T mutant. The current results also support our early proposal that the SLS structure might be important intermediate for monomer deposition to fibril, which explains the experimental fact that F19T mutant resists deposition to fibril.

Assembling NMR structures for the intracellular loops of the human thromboxane A 2 receptor: Implication of the G protein-coupling pocket

It has been reported that the multiple intracellular loops (iLPs) of the thromboxane A 2 receptor (TP) are involved in the receptor G protein coupling. In this study, a high-resolution 2D NMR technique was used to determine the 3D structures of the first, second, and third iLPs of the TP using synthetic peptides constrained into the loop structures. 2D 1H NMR spectra, TOCSY and NOESY were obtained for the two peptides from proton NMR experiments. The NMR data was processed and assigned through the Felix 2000 program. Standard methods were used to acquire sequence-specific assignments. Structure calculations were processed through DGII and NMR refinement programs within the Insight II program. We were able to calculate and use the NOE constraints to obtain the superimposed structure of 10 structures for each iLP peptide. The NMR-determined structures of the iLP peptides were used to refine a homology model of the TP. A 3D G-protein-binding cavity, formed by the three intracellular loops, was predicted by the docking of the C-terminal domain of the Gαq. Based on the structural model and the previous mutagenesis studies, the residues, R130, R60, C223, F138, L360, V361, E358 and Y359, which are important for interaction with the G protein, were further highlighted. These results reveal the possibly important molecular mechanisms in TP signaling and provide structural information to characterize other prostanoid receptor signalings.

NMR assessment of the global shape of a non‐labelled DNA dodecamer containing a tandem of GT mismatches

We have carried out a solution study of two non-labelled self-complementary DNA dodecamers d(GACTGTACAGTC)2 and d(GACTGTGCAGTC)2 by NMR, the second sequence composed of two G-T mismatches. Structures were determined using distances extracted from NOE effects alone or using both NOE and RDC constraints, measured in three different liquid crystalline media. We ensured that our data on the influence of the mesogen on the DNA structures, and the way in which the RDCs were incorporated as constraints in the protocol refinement, were consistent. We also tested the influence of different sets of RDCs and the best means of optimizing the calculation of D(a) and R. Resolution and accuracy of the ten best energy final structures were compared. The addition of a small set of RDC constraints significantly improves the final determined structures. We took advantage of the specificity of the RDC, i.e. it contains orientational information, and explored the global shape of the DNA duplexes; it was found that the duplexes do not have a large curvature. For the G-T base pair, we observed, in this particular sequence (tandem of G-T mismatches), a new pattern of base pairing, which involved the formation of a bifurcated hydrogen bond.

Solution structure of candoxin, a novel three-finger toxin from the venom of Bungarus candidus

Candoxin, a novel toxin purified from the venom of the Malayan krait (Bungarus candidus), is a 66- residue polypeptide containing five disulfide bridges, and is a reversible antagonist of postjunctional nicotinic acetylcholine receptors of the neuromuscular junction. A family of structures were calculated using a combination of distance geometry and simulated annealing with nOe, hydrogen bond and dihedral angle constraints. After refinement 19 structures, which satisfy the experimental constraints, were obtained. A comparison of each of the final structures with the average structure, gives an RMSD of 1.20 A for backbone atoms. Candoxin has an overall conformation similar to other three-finger toxins with a two-stranded and a three-stranded β-sheets. In spite of the low sequence homology, the tertiary fold of candoxin closely resembles those of erabutoxin b and cobrotoxin, with the exception of the disulfide bridge in the Loop I. Though candoxin lags the most conserved Tyr at the origin of the central loop it posses the most popular 3-dimensional array of residues (W31, R35 and K49), which is reported to be critical for curaremimetic neurotoxicity. The presence of the kink at the tip of the first loop (Loop I) caused by the C6-C11 disulphide bridge, shortens the Loop and hence may isolate this portion of the molecule and prevent its interference with the rest of the molecule, and facilitate specific interaction with receptor/acceptor protein. These observations suggest that candoxin may bind to acetylcholine receptor with lower affinity compared to other neurotoxins and hence could account for its relatively lower toxicity.

Comparison of Protocols for Calculation of Peptide Structures from Experimental NMR Data.

In a comparison of structure calculation protocols we clearly demonstrate the need for generating independent starting structures, which is for peptides most efficiently achieved by distance geometry (DG) methods. Our test set consisted of 20 peptides with 7-9 amino acid residues additionally constrained by backbone cyclization and/or the presence of a disulfide bridge. Small peptides usually adopt defined conformational properties only upon introduction of additional constraints, such as cyclization. Therefore, we believe the results of our comparison to be applicable to a large and important class of molecules. The problems associated with the use of restrained molecular dynamics (MD) for conformational searching in the context of structure calculation consist in energy barriers that derive mainly but not exclusively from the experimental NOE constraints. A valid alternative to the DG approach, although for peptides computationally less efficient, is MD simulated annealing starting from random structures as commonly performed in the protein structure calculation from NMR data. As a consequence of our study it must be expected that a considerable fraction of published peptide structures are artificially well-defined or even wrong. Given the relevance of peptide studies for both drug development and protein folding we regard it highly important that structure calculations of peptides are performed with more consideration.

Solution structure of a common substrate mimetic of cyclooxygenase-downstream synthases bound to an engineered thromboxane A 2 synthase using a high-resolution NMR technique

Understanding the docking mechanism of the common substrate, prostaglandin H 2 (PGH 2), into the active sites of different cyclooxygenase(COX)-downstream synthases is a key step toward uncovering the molecular basis of the isomerization of PGH 2 to different prostanoids. A high-resolution NMR spectroscopy was used to determine the conformational changes and solution 3D structure of U44069, a PGH 2 analogue, bound to one of the COX-downstream synthases—an engineered thromboxane A 2 synthase (TXAS). The dynamic binding was clearly observed by 1D NMR titration. The detailed conformational change and 3D structure of U44069 bound to the TXAS were demonstrated by 2D 1H NMR experiments using transferred NOEs. Through the assignments for the 2D 1H NMR spectra, TOCSY, DQF-COSY, NOESY, and the structural calculations based on the NOE constraints, they demonstrated that the widely open conformation with a triangle shape of the free U44069 changed to a compact structure with an oval shape when bound to the TXAS. The putative substrate-binding pocket of the TXAS model fits the conformation of the TXAS-bound U44069 appropriately, but could not fit the free form of U44069. It was the first to provide structural information for the dynamic docking of the PGH 2 mimic of the TXAS in solution, and to imply that PGH 2 undergoes conformational changes when bound to different COX-downstream synthases, which may play important roles in the isomerization of PGH 2 to different prostanoids. The NMR technique can be used as a powerful tool to determine the conformations of PGH 2 bound to other COX-downstream synthases.

Crystal and solution structures of 7-amino-actinomycin D complexes with d(TTAGBrUT), d(TTAGTT) and d(TTTAGTTT)

The formation of the complex of 7-amino-actinomycin D with potentially single-stranded DNA has been studied by X-ray crystallography in the solid state, by NMR in solution and by molecular modelling. The crystal structures of the complex with 5'-TTAG[Br(5)U]T-3' provide interesting examples of MAD phasing in which the dispersive component of the MAD signal was almost certainly enhanced by radiation damage. The trigonal and orthorhombic crystal modifications both contain antibiotic molecules and DNA strands in the form of a 2:4 complex: in the orthorhombic form there is one such complex in the asymmetric unit, while in the trigonal structure there are four. In both structures the phenoxazone ring of the first drug intercalates between a BrU-G (analogous to T-G) wobble pair and a G-T pair where the T is part of a symmetry-related molecule. The chromophore of the second actinomycin intercalates between the BrU-G and G-BrU wobble pairs of the partially paired third and fourth strands. The base stacking also involves (A*T)*T triplets and Watson-Crick A-T pairs and leads to similar complex three-dimensional networks in both structures, with looping-out of unpaired bases. Although the available NOE constraints of a solution containing the antibiotic and d(TTTAGTTT) strands in the ratio 1:1 are insufficient to determine the structure of the complex from the NMR data alone, they are consistent with the intercalation geometry observed in the crystal structure. Molecular-dynamics (MD) trajectories starting from the 1:2 complexes observed in the crystal showed that although the thymines flanking the d(AGT) core are rather flexible and the G-T pairing is not permanently preserved, both strands remain bound to the actinomycin by strong interactions between it and the guanines between which it is sandwiched. Similar strong binding (hemi-intercalation) of the actinomycin to a single guanine was observed in the MD trajectories of a 1:1 complex. The dominant interaction is between the antibiotic and guanine, but the complexes are stabilized further by promiscuous base-pairing.

The fumarate sensor DcuS: progress in rapid protein fold elucidation by combining protein structure prediction methods with NMR spectroscopy

We illustrate how moderate resolution protein structures can be rapidly obtained by interlinking computational prediction methodologies with un- or partially assigned NMR data. To facilitate the application of our recently described method of ranking and subsequent refining alternative structural models using unassigned NMR data [Proc. Natl. Acad. Sci. USA 100 (2003) 15404] for such “structural genomics”-type experiments it is combined with protein models from several prediction techniques, enhanced to utilize partial assignments, and applied on a protein with an unknown structure and fold. From the original NMR spectra obtained for the 140 residue fumarate sensor DcuS, 1100 1H, 13C, and 15N chemical shift signals, 3000 1H– 1H NOESY cross peak intensities, and 209 backbone residual dipolar couplings were extracted and used to rank models produced by de novo structure prediction and comparative modeling methods. The ranking proceeds in two steps: first, an optimal assignment of the NMR peaks to atoms is found for each model independently, and second, the models are ranked based on the consistency between the NMR data and the model assuming these optimal assignments. The low-resolution model selected using this ranking procedure had the correct overall fold and a global backbone RMSD of 6.0Å, and was subsequently refined to 3.7 Å RMSD. With the incorporation of a small number of NOE and residual dipolar coupling constraints available very early in the traditional spectral assignment process, a model with an RMSD of 2.8 Å could rapidly be built. The ability to generate moderate resolution models within days of NMR data collection should facilitate large scale NMR structure determination efforts.