MINRMS: an efficient algorithm for determining protein structure similarity using root-mean-squared-distance

MOTIVATION: Existing algorithms for automated protein structure alignment generate contradictory results and are difficult to interpret. An algorithm which can provide a context for interpreting the alignment and uses a simple method to characterize protein structure similarity is needed. RESULTS: We describe a heuristic for limiting the search space for structure alignment comparisons between two proteins, and an algorithm for finding minimal root-mean-squared-distance (RMSD) alignments as a function of the number of matching residue pairs within this limited search space. Our alignment algorithm uses coordinates of alpha-carbon atoms to represent each amino acid residue and requires a total computation time of O(m(3) n(2)), where m and n denote the lengths of the protein sequences. This makes our method fast enough for comparisons of moderate-size proteins (fewer than approximately 800 residues) on current workstation-class computers and therefore addresses the need for a systematic analysis of multiple plausible shape similarities between two proteins using a widely accepted comparison metric.     

Conformational dynamics of chymotrypsin inhibitor 2 by coarse-grained simulations

A coarse-grained dynamic Monte Carlo (MC) simulation method is used to investigate the conformational dynamics of chymotrypsin inhibitor 2 (CI2). Each residue is represented therein by two interaction sites, one at the alpha-carbon and the other on the amino acid side-chain. The energy and geometry parameters extracted from databank structures are used. The calculated rms fluctuations of alpha-carbon atoms are in good agreement with crystallographic temperature factors. The two regions of the protein that pack against each other to form the main hydrophobic core exhibit negatively correlated fluctuations. The conformational dynamics could efficiently be probed by the time-delayed orientational and conformational correlation functions of the virtual bonds: the active site loop, excluding the active site bond, the turn region, and the N-terminal of the alpha-helix are relatively more mobile regions of the structure. A correlation is observed between the hydrogen/deuterium (H/D) exchange behavior and the long-time orientational and conformational autocorrelation function values for CI2. A cooperativity in the rotations of the bonds near in sequence is observed at all time windows, whereas the cooperative rotations of the bonds far along the sequence appear at long time windows; these correlations contribute to the stability of the secondary structures and the tertiary structure, respectively.     

Reconstruction of 3D structures from protein contact maps

The prediction of the protein tertiary structure from solely its residue sequence (the so called Protein Folding Problem) is one of the most challenging problems in Structural Bioinformatics. We focus on the protein residue contact map. When this map is assigned it is possible to reconstruct the 3D structure of the protein backbone. The general problem of recovering a set of 3D coordinates consistent with some given contact map is known as a unit-disk-graph realization problem and it has been recently proven to be NP-Hard. In this paper we describe a heuristic method (COMAR) that is able to reconstruct with an unprecedented rate (3-15 seconds) a 3D model that exactly matches the target contact map of a protein. Working with a non-redundant set of 1760 proteins, we find that the scoring efficiency of finding a 3D model very close to the protein native structure depends on the threshold value adopted to compute the protein residue contact map. Contact maps whose threshold values range from 10 to 18 Ångstroms allow reconstructing 3D models that are very similar to the proteins native structure.     

A new method for the measurement of protein turnover.

A new technique for the determination of rate constants of protein degradation is described. By using the method, half-lives of total soluble protein of Lemna minor during growth on full culture medium and distilled water were measured. The method involves incubating Lemna on a growth medium containing 3H2O. After a short exposure (20 min) to 3H-labelled culture medium, 3H was found in soluble amino acids, especially aspartate, glutamate, glutamine and alanine. After transfer to a 3H-free medium for 30 min, 80% of the 3H originally present in soluble amino acids was lost. These results suggest that 3H enters and leaves amino acids at the alpha-carbon atom, a conclusion supported by the observed labelling of glutamates. The exchange of H and 3H on the alpha-carbon atom is catalysed by transaminases and the speed of this exchange ensures that when the 3H2O is removed, the 3H in free amino acids is rapidly lost, thereby eliminating problems connected with metabolic pools and recycling. After an exposure of 20 min to 3H-labelled medium all protein amino acids, except for arginine, were found to be radioactive. The loss of radioactivity from protein amino acids was used to measure protein degradation.     

Characterization by NMR of the heme-myoglobin adduct formed during the reductive metabolism of BrCCl3. Covalent bonding of the proximal histidine to the ring I vinyl group

The reductive debromination of BrCCl3 by ferrous deoxymyoglobin leads to the covalent bonding of the prosthetic heme to the protein. We have previously shown, by the use of peptide mapping and mass spectrometry, that histidine residue 93 is covalently bound to the heme moiety. In the present study the structure of the heme adduct was more completely determined by 1H and 13C NMR techniques. We have found that the ring I vinyl group of the prosthetic heme was altered by the addition of a histidine imidazole nitrogen to the alpha-carbon and a CCl2 moiety to the beta-carbon. The electronic absorption spectra of the oxidized and reduced states of the altered heme-protein indicated that the heme-iron exists in a bis-histidine-ligated form. Analysis of the crystal structure of native myoglobin suggested that for the altered heme-protein, histidine residues 97 and 64 are ligated to the heme-iron and that residue 97 has replaced the native proximal histidine residue 93. These movements, in effect a "histidine shuffle" at the active site, may be responsible for the enhanced reducing activity of the altered protein.     

Stereospecific binding of diastereomeric peptides to salmon sperm deoxyribonucleic acid. Further evidence for partial intercalation

A series of diastereomeric dipeptide amides, containing an N-terminal L-lysyl residue and a C-terminal L- or D-amino acid with a derivatized aromatic ring on the side chain, was synthesized to determine the dependence of (1) the chirality of the N-terminal amino acid alpha-carbon and (2) the length of the N-terminal amino acid side chain for intercalation of the aromatic ring. The nature of the complex between the peptide and DNA (i.e., electrostatic, intercalative, or a combination of these) was determined by UV and CD studies, viscometric titrations, and 1H NMR studies. The results of these studies reveal distinct differences in the binding site of the aromatic rings of the various peptides. In particular, the results suggest that the alpha- and epsilon-amino groups of the lysyl residue bind electrostatically to adjacent phosphates on the DNA backbone in a stereospecific manner. As a result of this stereospecificity, the aromatic rings of the peptides with the L-L designation point toward the DNA helix, while those of the peptides of the L-D designation point away from the helix. This is completely consistent with previously reported work [Gabbay, E.J., Adawadkar, P. D., & Wilson, W. D. (1976) Biochemistry 15, 146; Gabbay, E. J., Adawadkar, P. D., Kapicak, L., Pearce, S., & Wilson, W. D. (1976) Biochemistry 15, 152]. The results also indicate a great dependence on the length of the side chain for intercalation of the aromatic ring. Specifically, if the side chain is long enough, and flexible enough, the aromatic ring can fully or partially intercalate, regardless of the chirality of the N-terminal amino acid alpha-carbon. However, if the side chain is too short, only partial intercalation is observed for peptides of the L-D designation, and no intercalation is observed for peptides of the L-D designation.     

ProDCoNN: Protein design using a convolutional neural network

Designing protein sequences that fold to a given three-dimensional (3D) structure has long been a challenging problem in computational structural biology with significant theoretical and practical implications. In this study, we first formulated this problem as predicting the residue type given the 3D structural environment around the C _α atom of a residue, which is repeated for each residue of a protein. We designed a nine-layer 3D deep convolutional neural network (CNN) that takes as input a gridded box with the atomic coordinates and types around a residue. Several CNN layers were designed to capture structure information at different scales, such as bond lengths, bond angles, torsion angles, and secondary structures. Trained on a very large number of protein structures, the method, called ProDCoNN (protein design with CNN), achieved state-of-the-art performance when tested on large numbers of test proteins and benchmark datasets.     

Improving the antigen affinity of an antibody Fv-fragment by protein design.

The affinity of an antibody for its ligand 2-phenyloxazolone was improved by protein design. For the design two-dimensional nuclear magnetic resonance spectroscopy, protein engineering and molecular modelling were used in an interactive scheme. Initially the binding site was localized with the help of transferred nuclear Overhauser enhancement signals from two, site specifically assigned tyrosine side-chains in the complementarity-determining regions of the antibody to the ligand 4-glycyl-2-phenyloxazolone. On their basis the hapten was placed into a model of the Fv-fragment built according to the principles of canonical antibody structures. From the model, unfavourable contacts between hapten and an aspartyl side-chain in complementarity-determining region 3 of the heavy chain were predicted. Substitution of the aspartyl residue by alanine resulted in a threefold increase in affinity of the antibody Fv-fragment for two hapten derivatives when compared with the wild-type. Nuclear magnetic resonance analysis of the improved Fv-fragment revealed an interaction between the alpha-carbon proton of alanyl residue with the ligand, which was not seen for the aspartyl residue. This interaction is not entirely in accordance with the model, which predicts an interaction between the side-chain of this residue and the hapten. However, it shows that by combined use of nuclear magnetic resonance analysis and molecular modelling, a residue that is critical for antigen binding was identified, whose mutation allowed the design of an improved antibody combining site.     

On the steric course of the adenosylcobalamin-dependent 2-methyleneglutarate mutase reaction in Clostridium barkeri

The enzymatically active enantiomer of 3-methylitaconate in Clostridium barkeri has (R)-configuration. This was checked by fermentation of the racemate and reisolation of the (S)-enantiomer. In addition (R)-3-methylitaconate was synthesized by enzymatic isomerisation of 2,3-dimethylmaleate which was protonated at the Si-face. 2-Methylene[2-2H1]glutarate was synthesized via (R)-3-methyl[3-2H1]itaconate by brief incubation of 2,3-dimethylmaleate with a cell-free extract of Clostridium barkeri in 2H2O. The predominantly monodeuterated compound was oxidized to (S)-[2-2H1]succinate as analysed by circular dichroism. The results demonstrate that 2-methyleneglutarate mutase catalyses the reversible migration of an acryloyl residue from the alpha-carbon to the beta-carbon of propionate with inversion of configuration at the alpha-carbon.     

Gaussian neighborhood: A new measure of accessibility for residues of protein molecules

We introduce a new method for assessing the extent of residue exposure in proteins. For each atom of every residue a Gaussian-weighted atomic surroundings value (the G-neighborhood) is calculated. A normalized sum of G-neighborhood values over all the atoms of a residue is complementary to conventional surface accessibility characteristics. The G-0neighborhood value of a residue is a sensitive indicator of its location, strongly dependent on the 3D structure of a the protein. Correlations between secondary structures and patterns of G-neighborhood values for six different protein molecules are discussed. Comparison of the distribution of hydrophobic and charged residues in the 3D structure for the alcohol-soluble protein crambin and that of five water-soluble proteins (cytochrome c, flavodoxin, myoglobin, rhodanese, and Bence–Jones protein) shows striking differences in their G-neighborhood patterns. Contacts between the prosthetic group and the peptide portion of a protein as well as protein interdomain contacts and monomer–monomer contacts are characterized. © 1993 Wiley-Liss, Inc.     

Three-dimensional cluster analysis identifies interfaces and functional residue clusters in proteins

Three-dimensional cluster analysis offers a method for the prediction of functional residue clusters in proteins. This method requires a representative structure and a multiple sequence alignment as input data. Individual residues are represented in terms of regional alignments that reflect both their structural environment and their evolutionary variation, as defined by the alignment of homologous sequences. From the overall (global) and the residue-specific (regional) alignments, we calculate the global and regional similarity matrices, containing scores for all pairwise sequence comparisons in the respective alignments. Comparing the matrices yields two scores for each residue. The regional conservation score (C(R)(x)) defines the conservation of each residue x and its neighbors in 3D space relative to the protein as a whole. The similarity deviation score (S(x)) detects residue clusters with sequence similarities that deviate from the similarities suggested by the full-length sequences. We evaluated 3D cluster analysis on a set of 35 families of proteins with available cocrystal structures, showing small ligand interfaces, nucleic acid interfaces and two types of protein-protein interfaces (transient and stable). We present two examples in detail: fructose-1,6-bisphosphate aldolase and the mitogen-activated protein kinase ERK2. We found that the regional conservation score (C(R)(x)) identifies functional residue clusters better than a scoring scheme that does not take 3D information into account. C(R)(x) is particularly useful for the prediction of poorly conserved, transient protein-protein interfaces. Many of the proteins studied contained residue clusters with elevated similarity deviation scores. These residue clusters correlate with specificity-conferring regions: 3D cluster analysis therefore represents an easily applied method for the prediction of functionally relevant spatial clusters of residues in proteins.     

The protein sequence responsible for lipoprotein membrane localization in Escherichia coli exhibits remarkable specificity

Structural information defining an N-terminal sequence required for the membrane sorting of bacterial lipoproteins has been previously garnered through the study of a hybrid outer membrane (OM) lipo-beta-lactamase (LL) (Ghrayeb and Inouye (1984) J. Biol. Chem. 259, 463-467). Introduction of an aspartate as the second residue of mature LL (D2 mutant) causes an inner membrane (IM) localization of this protein (Yamaguchi, K., Yu, F., and Inouye, M. (1988) Cell 53, 423-432). Introduction of as aspartate at the third residue of mature LL (D3) causes a weaker IM sorting signal and when present as the fourth residue (D4), normal OM sorting occurs. A positively charged residue at the second position (K2) has no effect on OM localization. Remarkably, glutamate substitution at either the second (E2) or third (E3) position does not interfere with OM sorting. Sorting of the mutant D2 LL can be partially suppressed by introduction of a positively charged histidine (D2H3) or lysine (D2K3) at residue 3 of the mature protein. These results indicate that both the negative charge of the aspartate residue and some structural feature not present in a glutamate residue are required for sorting to the IM. The suppression of IM localization of the D2H3 LL double mutant can be eliminated by growing Escherichia coli at pH 8.4 to reduce the histidine partial positive charge. This result supports the essentiality of a negative charge in IM localization and indicates that the committed step in lipoprotein sorting is made in a cellular compartment, the periplasm, at equilibrium with the external pH.