Influence of the completeness of chemical shift assignments on NMR structures obtained with automated NOE assignment

Reliable automated NOE assignment and structure calculation on the basis of a largely complete, assigned input chemical shift list and a list of unassigned NOESY cross peaks has recently become feasible for routine NMR protein structure calculation and has been shown to yield results that are equivalent to those of the conventional, manual approach. However, these algorithms rely on the availability of a virtually complete list of the chemical shifts. This paper investigates the influence of incomplete chemical shift assignments on the reliability of NMR structures obtained with automated NOESY cross peak assignment. The program CYANA was used for combined automated NOESY assignment with the CANDID algorithm and structure calculations with torsion angle dynamics at various degrees of completeness of the chemical shift assignment which was simulated by random omission of entries in the experimental ¹H chemical shift lists that had been used for the earlier, conventional structure determinations of two proteins. Sets of structure calculations were performed choosing the omitted chemical shifts randomly among all assigned hydrogen atoms, or among aromatic hydrogen atoms. For comparison, automated NOESY assignment and structure calculations were performed with the complete experimental chemical shift but under random omission of NOESY cross peaks. When heteronuclear-resolved three-dimensional NOESY spectra are available the current CANDID algorithm yields in the absence of up to about 10% of the experimental ¹H chemical shifts reliable NOE assignments and three-dimensional structures that deviate by less than 2 Å from the reference structure obtained using all experimental chemical shift assignments. In contrast, the algorithm can accommodate the omission of up to 50% of the cross peaks in heteronuclear- resolved NOESY spectra without producing structures with a RMSD of more than 2 Å to the reference structure. When only homonuclear NOESY spectra are available, the algorithm is slightly more susceptible to missing data and can tolerate the absence of up to about 7% of the experimental ¹H chemical shifts or of up to 30% of the NOESY peaks.Abbreviations: BmPBP^A – Bombyx mori pheromone binding protein form A; CYANA – combined assignment and dynamics algorithm for NMR applications; NMR – nuclear magnetic resonance; NOE – nuclear Overhauser effect; NOESY – NOE spectroscopy; RMSD – root-mean-square deviation; WmKT – Williopsis mrakii killer toxin     

Protein NMR structure determination with automated NOE assignment using the new software CANDID and the torsion angle dynamics algorithm DYANA

Combined automated NOE assignment and structure determination module (CANDID) is a new software for efficient NMR structure determination of proteins by automated assignment of the NOESY spectra. CANDID uses an iterative approach with multiple cycles of NOE cross-peak assignment and protein structure calculation using the fast DYANA torsion angle dynamics algorithm, so that the result from each CANDID cycle consists of exhaustive, possibly ambiguous NOE cross-peak assignments in all available spectra and a three-dimensional protein structure represented by a bundle of conformers. The input for the first CANDID cycle consists of the amino acid sequence, the chemical shift list from the sequence-specific resonance assignment, and listings of the cross-peak positions and volumes in one or several two, three or four-dimensional NOESY spectra. The input for the second and subsequent CANDID cycles contains the three-dimensional protein structure from the previous cycle, in addition to the complete input used for the first cycle. CANDID includes two new elements that make it robust with respect to the presence of artifacts in the input data, i.e. network-anchoring and constraint-combination, which have a key role in de novo protein structure determinations for the successful generation of the correct polypeptide fold by the first CANDID cycle. Network-anchoring makes use of the fact that any network of correct NOE cross-peak assignments forms a self-consistent set; the initial, chemical shift-based assignments for each individual NOE cross-peak are therefore weighted by the extent to which they can be embedded into the network formed by all other NOE cross-peak assignments. Constraint-combination reduces the deleterious impact of artifact NOE upper distance constraints in the input for a protein structure calculation by combining the assignments for two or several peaks into a single upper limit distance constraint, which lowers the probability that the presence of an artifact peak will influence the outcome of the structure calculation. CANDID test calculations were performed with NMR data sets of four proteins for which high-quality structures had previously been solved by interactive protocols, and they yielded comparable results to these reference structure determinations with regard to both the residual constraint violations, and the precision and accuracy of the atomic coordinates. The CANDID approach has further been validated by de novo NMR structure determinations of four additional proteins. The experience gained in these calculations shows that once nearly complete sequence-specific resonance assignments are available, the automated CANDID approach results in greatly enhanced efficiency of the NOESY spectral analysis. The fact that the correct fold is obtained in cycle 1 of a de novo structure calculation is the single most important advance achieved with CANDID, when compared with previously proposed automated NOESY assignment methods that do not use network-anchoring and constraint-combination.     

A complete algorithm to resolve ambiguity for intersubunit NOE assignment in structure determination of symmetric homo-oligomers

Assignment of nuclear Overhauser effect (NOE) data is a key bottleneck in structure determination by NMR. NOE assignment resolves the ambiguity as to which pair of protons generated the observed NOE peaks, and thus should be restrained in structure determination. In the case of intersubunit NOEs in symmetric homo-oligomers, the ambiguity includes both the identities of the protons within a subunit, and the identities of the subunits to which they belong. This paper develops an algorithm for simultaneous intersubunit NOE assignment and C(n) symmetric homo-oligomeric structure determinations, given the subunit structure. By using a configuration space framework, our algorithm guarantees completeness, in that it identifies structures representing, to within a user-defined similarity level, every structure consistent with the available data (ambiguous or not). However, while our approach is complete in considering all conformations and assignments, it avoids explicit enumeration of the exponential number of combinations of possible assignments. Our algorithm can draw two types of conclusions not possible under previous methods: (1) that different assignments for an NOE would lead to different structural classes, or (2) that it is not necessary to uniquely assign an NOE, since it would have little impact on structural precision. We demonstrate on two test proteins that our method reduces the average number of possible assignments per NOE by a factor of 2.6 for MinE and 4.2 for CCMP. It results in high structural precision, reducing the average variance in atomic positions by factors of 1.5 and 3.6, respectively.     

An automated method for consistent helix assignment using turn information

A new automated helix assignment method is presented that leads to a more consistent definition of the helix termini, especially of the helix C-terminus. The method assigns a helix to segments of protein chain where adjacent helical turn structures are observed, capped by specific distorted turn types (e.g., open helical turns without a hydrogen bond) or capping motifs (e.g., the Schellman motif). Helix termini are detected by observing the behavior of the NH group in N-termini and the CO group in C-termini; in each case, the respective group must be free to interact with hydrogen bonding partners outside of the putative helix for a helix terminus to be assigned. The presented assignment method and SHAFT-assigned helices are part of Secbase and are made available with Relibase+ 3.0 and the free web version of Relibase 3.0. The method can also be used for the helix assignments of additional protein structures.     

Ansig for Windows: An interactive computer program for semiautomatic assignment of protein NMR spectra

Assignment of NMR spectra is a prerequisite for structure determination of proteins using NMR. The time spent on the assignment is comparatively long compared to that spent on other parts in the protein structure determination process, but it can be shortened by using either interactive or fully automated computer programs. To benefit from the advantages of both types of program we have developed a version of the interactive assignment program ANSIG to include automatized, yet user-supervised, routines. The new program includes tools for (i) semiautomatic sequential assignment, (ii) plotting of distances from PDB structure files directly in NMR spectra and (iii) statistical analysis of distance restraint violations with the possibility to directly zoom to violated NOEs in NOESY spectra.     

MONTE: An automated Monte Carlo based approach to nuclear magnetic resonance assignment of proteins

A general-purpose Monte Carlo assignment program has been developed to aid in the assignment of NMR resonances from proteins. By virtue of its flexible data requirements the program is capable of obtaining assignments of both heavily deuterated and fully protonated proteins. A wide variety of source data, such as inter-residue scalar connectivity, inter-residue dipolar (NOE) connectivity, and residue specific information, can be utilized in the assignment process. The program can also use known assignments from one form of a protein to facilitate the assignment of another form of the protein. This attribute is useful for assigning protein-ligand complexes when the assignments of the unliganded protein are known. The program can be also be used as an interactive research tool to assist in the choice of additional experimental data to facilitate completion of assignments. The assignment of a deuterated 45 kDa homodimeric Glutathione-S-transferase illustrates the principal features of the program.     

Evaluation of an algorithm for the automated sequential assignment of protein backbone resonances: A demonstration of the connectivity tracing assignment tools (CONTRAST) software package

Summary The peptide sequential assignment algorithm presented here was implemented as a macro within the CONnectivity TRacing ASsignment Tools (CONTRAST) computer software package. The algorithm provides a semi- or fully automated global means of sequentially assigning the NMR backbone resonances of proteins. The program's performance is demonstrated here by its analysis of realistic computer-generated data for III^Glc, a 168-residue signal-transducing protein of Escherichia coli [Pelton et al. (1991) Biochemistry, 30, 10043–10057]. Missing experimental data (19 resonances) were generated so that a complete assignment set could be tested. The algorithm produces sequential assignments from appropriate peak lists of nD NMR data. It quantifies the ambiguity of each assignment and provides ranked alternatives. A best first approach, in which high-scoring local assignments are made before and in preference to lower scoring assignments, is shown to be superior (in terms of the current set of CONTRAST scoring routines) to approaches such as simulated annealing that seek to maximize the combined scores of the individual assignments. The robustness of the algorithm was tested by evaluating the effects of imposed frequency imprecision (scatter), added false signals (noise), missing peaks (incomplete data), and variation in userdefined tolerances on the performance of the algorithm.     

Automated error-tolerant macromolecular structure determination from multidimensional nuclear Overhauser enhancement spectra and chemical shift assignments: improved robustness and performance of the PASD algorithm

We report substantial improvements to the previously introduced automated NOE assignment and structure determination protocol known as PASD (Kuszewski et al. (2004) J Am Chem Soc 26:6258-6273). The improved protocol includes extensive analysis of input spectral data to create a low-resolution contact map of residues expected to be close in space. This map is used to obtain reasonable initial guesses of NOE assignment likelihoods which are refined during subsequent structure calculations. Information in the contact map about which residues are predicted to not be close in space is applied via conservative repulsive distance restraints which are used in early phases of the structure calculations. In comparison with the previous protocol, the new protocol requires significantly less computation time. We show results of running the new PASD protocol on six proteins and demonstrate that useful assignment and structural information is extracted on proteins of more than 220 residues. We show that useful assignment information can be obtained even in the case in which a unique structure cannot be determined.     

Variability in automated assignment of NOESY spectra and three-dimensional structure determination: A test case on three small disulfide-bonded proteins

Three independent runs of automatic assignment and structure calculations were performed on three small proteins, calcicludine from the venom of the green mamba Dendroaspis angusticeps, -conotoxin PVIIA from the purple cone Conus purpurascens and HsTX1, a short scorpion toxin from the venom of Heterometrus spinnifer. At the end of all the runs, the number of cross peaks which remained unassigned (0.6%, 1.4% and 2% for calcicludine, -conotoxin and HsTX1, respectively), as well as the number of constraints which were rejected as producing systematic violations (2.7%, 1.0%, and 1.4% for calcicludine, -conotoxin and HsTX1, respectively) were low. The conformation of the initial model used in the procedure (linear model or constructed by homology) has no influence on the final structures. Mainly two parameters control the procedure: the chemical shift tolerance and the cut-off distance. Independent runs of structure calculations, using the same parameters, yield structures for which the rmsd between averaged structures and the rmsd around each averaged structure were of the same order of magnitude. A different cut-off distance and a different chemical shift tolerance yield rmsd values on final average structures which did not differ more than 0.5 Å compared to the rmsd obtained around the averaged structure for each calculation. These results show that the procedure is robust when applied to such a small disulfide-bonded protein.     

Extensive distance geometry calculations with different NOE calibrations: New criteria for structure selection applied to Sandostatin and BPTI

Summary To generate structures efficiently, a version of the distance geometry program DIANA for a parallel computer was developed, new objective criteria for the selection of NMR solution structures are presented, and the influence of using different calibrations of NOE intensities on the final structures are described. The methods are applied to the structure determination of Sandostatin, a disulfide-bridge octapeptide, and to model calculations of BPTI. On an Alliant FX2800 computer using 10 processors in parallel, the calculations were done 9.2 times faster than with a single processor. Up to 7000 Sandostatin structures were calculated with distance and angular constraints. The procedure for selecting acceptable structures is based on the maximum values of pairwise RMSDs between structures. Suitable target function cut-offs are defined independent of the number of starting structures. The method allowed for an objective comparison of three groups of Sandostatin structures that were calculated from different sets of upper distance constraints which were derived from the same NOE intensity data using three empirical calibration curves. The number of converged structures and the target function values differed significantly among the three groups, but the structures were qualitatively and quantitatively very similar. The conformation is well determined in the cyclic region Cys²–Cys⁷ and adopts a -turn centered at d-Trp⁴–Lys⁵. The criteria for structure selection were further tested with BPTI. Results obtained from sets of structures calculated with and without using the REDAC strategy are consistent and suggest that the structure selection method is objective and generally applicable.     

Interproton distance bounds from 2D NOE intensities: Effect of experimental noise and peak integration errors

Summary The effect of experimental and integration errors on the calculations in interproton distances from NOE intensities is examined. It is shown that NOE intensity errors can have a large impact on the distances determined. When multiple spin (spin diffusion) effects are significant, the calculated distances are often underestimated, even when using a complete relaxation matrix analysis. In this case, the bias of distances to smaller values is due to the random errors in the NOE intensities. We show here that accurate upper and lower bounds of the distances can be obtained if the intensity errors are properly accounted for in the complete relaxation matrix calculations, specifically the MARDIGRAS algorithm. The basic MARDIGRAS algorithm has been previously described [Borgias, B.A. and James, T.L. (1990) J. Magn. Reson., 87, 475–487]. It has been shown to provide reasonably good interproton distance bounds, but experimental errors can compromise the quality of the resulting restraints, especially for weak cross peaks. In a new approach introduced here, termed RANDMARDI (random error MARDIGRAS), errors due to random noise and integration errors are mimicked by the addition of random numbers from within a specified range to each input intensity. Interproton distances are then calculated for the modified intensity set using MARDIGRAS. The distribution of distances that define the upper and lower distance bounds is obtained by using N randomly modified intensity sets. RANDMARDI has been used in the solution structure determination of the interstrand cross-link (XL) formed between 4-hydroxymethyl-4,5,8-trimethylpsoralen (HMT) and the DNA oligomer d(5-GCGTACGC-3)₂ [Spielmann, H.P. et al. (1995) Biochemistry, 34, 12937–12953]. RANDMARDI generates accurate distance bounds from the experimental NOESY cross-peak intensities for the fixed (known) interproton distances in XL. This provides an independent internal check for the ability of RANDMARDI to accurately fit the experimental data. The XL structure determined using RANDMARDI-generated restrains is in good agreement with other biophysical data that indicate that there is no bend introduced into the DNA by the cross-link. In contrast, isolated spin-pair approximation calculations give distance restraints that, when applied in a restrained molecular dynamics protocol, produce a bent structure.Abbreviations NOE nuclear Overhauser effect - SD standard deviation - HMT 4-hydroxymethyl-4,5,8-trimethylpsoralen - XL psoralen-DNA interstrand cross-link     

The program ASNO for computer-supported collection of NOE upper distance constraints as input for protein structure determination

Summary A new program, ASNO (ASsign NOes), for computer-supported NOE cross-peak assignments is described. ASNO is used for structure refinement in several rounds of NOESY cross-peak assignments and 3D structure calculations, where the preliminary structures are used as a reference to resolve ambiguities in NOE assignments which are otherwise based on the chemical shifts available from the sequence-specific resonance assignments. The practical use of ASNO for proteins is illustrated with the structure determination of Dendrotoxin K from Dendroaspis polylepis polylepis.Abbreviations Toxin K dendrotoxin K (or trypsin inhibitor homologue K) from the venom of the black mamba Dendroaspis polylepis polylepis - NOE nuclear Overhauser effect - NOESY NOE spectroscopy - REDAC use of redundant dihedral angle constraints - RMSD root-mean-square deviation To whom correspondence should be addressed.     

Domain assignment for protein structures using a consensus approach: characterization and analysis.

A consensus approach for the assignment of structural domains in proteins is presented. The approach combines a number of previously published algorithms, and takes advantage of the elevated accuracy obtained when assignments from the individual algorithms are in agreement. The consensus approach is tested on a data set of 55 protein chains, for which domain assignments from four automated methods were known, and for which crystallographers assignments had been reported in the literature. Accuracy was found to increase in this test from 72% using individual algorithms to 100% when all four methods were in agreement. However a consensus prediction using all four methods was only possible for 52% of the dataset. The consensus approach [using three publicly available domain assignment algorithms (PUU, DETECTIVE, DOMAK)] was then used to make domain assignments for a data set of 787 protein chains from the Protein Data Bank. Analysis of the assignments showed 55.7% of assignments could be made automatically, and of these, 13.5% were multi-domain proteins. Of the remaining 44.3% that could not be assigned by the consensus procedure 90.4% had their domain boundaries assigned correctly by at least one of the algorithms. Once identified, these domains were analyzed for trends in their size and secondary structure class. In addition, the discontinuity of each domain along the protein chain was considered.     

A high quality nuclear magnetic resonance solution structure of peptide deformylase from Escherichia coli: Application of an automated assignment strategy using GARANT

The NMR structure of the peptide deformylase (PDF) (1–150) from Escherichia coli, which is an essential enzyme that removes the formyl group from nascent polypeptides and represents a potential target for drug discovery, was determined using 15N/13C doubly labeled protein. Nearly completely automated assignment routines were employed to assign three-dimensional triple resonance, 15N-resolved and 13C-resolved NOESY spectra using the program GARANT. This assignment strategy, demonstrated on a 17 kDa protein, is a significant advance in the automation of NMR data assignment and structure determination that will accelerate future work. A total of 2302 conformational constraints were collected as input for the distance geometry program DYANA. After restrained energy minimization with the program X-PLOR the 20 best conformers characterize a high quality structure with an average of 0.43 Å for the root-mean-square deviation calculated from the backbone atoms N, C and C, and 0.81 Å for all heavy atoms of the individual conformers relative to the mean coordinates for residues 1 to 150. The globular fold of PDF contains two -helices comprising residues 25–40, 125–138, six -strands 57–60, 70–77, 85–88, 98–101, 105–111, 117–123 and one 310 helix comprising residues 49–51. The C-terminal helix contains the HEXXH motif positioning a zinc ligand in a similar fashion to other metalloproteases, with the third ligand being cysteine and the fourth presumably a water. The three-dimensional structure of PDF affords insight into the substrate recognition and specificity for N-formylated over N-acetylated substrates and is compared to other PDF structures.     

Potential of microsatellites for individual assignment: the North Atlantic redfish (genus Sebastes) species complex as a case study

We used the four redfish taxa (genus Sebastes) from the North Atlantic to evaluate the potential of multilocus genotype information obtained from microsatellites in assigning individuals at two different levels of group divergence. We first tested the hypothesis that microsatellites can diagnostically discriminate individual redfish from different groups. Second, we compared two different methods to quantify the effect of number of loci and likelihood stringency levels on the power of microsatellites for redfish group membership. The potential of microsatellites to discriminate individuals from different taxa was illustrated by a shared allele distance tree in which four major clusters corresponding to each taxa were defined. Concomitant with this strong discrimination, microsatellites also proved to be powerful in reclassifying specimens to the taxon of origin, using either an empirical or simulated method of estimating assignment success. By testing for the effect of both the number of loci and the level of stringency on the assignment success, we found that 95% of all specimens were still correctly reclassified with only four loci at the most commonly used criterion of log0. In contrast, the results obtained at the population level within taxa highlighted several problems of assignment that may occur at low levels of divergence. Namely, a drastic decrease of success with increasing stringency illustrated the lack of power of our set of loci. Strong discrepancy was observed between results obtained from the empirical and simulated methods. Finally, the highest assignment success was obtained when reducing the number of loci used, an observation previously reported in studies of human populations.     

Assigning large proteins in the solid state: a MAS NMR resonance assignment strategy using selectively and extensively <Superscript>13</Superscript>C-labelled proteins

In recent years, solid-state magic-angle spinning nuclear magnetic resonance spectroscopy (MAS NMR) has been growing into an important technique to study the structure of membrane proteins, amyloid fibrils and other protein preparations which do not form crystals or are insoluble. Currently, a key bottleneck is the assignment process due to the absence of the resolving power of proton chemical shifts. Particularly for large proteins (approximately >150 residues) it is difficult to obtain a full set of resonance assignments. In order to address this problem, we present an assignment method based upon samples prepared using [1,3-¹³C]- and [2-¹³C]-glycerol as the sole carbon source in the bacterial growth medium (so-called selectively and extensively labelled protein). Such samples give rise to higher quality spectra than uniformly [¹³C]-labelled protein samples, and have previously been used to obtain long-range restraints for use in structure calculations. Our method exploits the characteristic cross-peak patterns observed for the different amino acid types in ¹³C-¹³C correlation and 3D NCACX and NCOCX spectra. An in-depth analysis of the patterns and how they can be used to aid assignment is presented, using spectra of the chicken α-spectrin SH3 domain (62 residues), αB-crystallin (175 residues) and outer membrane protein G (OmpG, 281 residues) as examples. Using this procedure, over 90% of the Cα, Cβ, C′ and N resonances in the core domain of αB-crystallin and around 73% in the flanking domains could be assigned (excluding 24 residues at the extreme termini of the protein). Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.