An automated classification of the structure of protein loops

Enhancement of protein modeling by human intervention in applying the automatic programs 3D-JIGSAW and 3D-PSSM

Fourteen models were constructed and analyzed for the comparative modeling section of Critical Assessment of Techniques for Protein Structure Prediction (CASP4). Sequence identity between each target and the best possible parent(s) ranged between 55 and 13%, and the root-mean-square deviation between model and target was from 0.8 to 17.9 A. In the fold recognition section, 10 of the 11 remote homologues were recognized. The modeling protocols are a combination of automated computer algorithms, 3D-JIGSAW (for comparative modeling) and 3D-PSSM (for fold recognition), with human intervention at certain critical stages. In particular, intervention is required to check superfamily assignment, best possible parents from which to model, sequence alignments to those parents and take-off regions for modeling variable regions. There now is a convergence of algorithms for comparative modeling and fold recognition, particularly in the region of remote homology.

Human procarboxypeptidase B: three-dimensional structure and implications for thrombin-activatable fibrinolysis inhibitor (TAFI)

Besides their classical role in alimentary protein degradation, zinc-dependant carboxypeptidases also participate in more selective regulatory processes like prohormone and neuropeptide processing or fibrinolysis inhibition in blood plasma. Human pancreatic procarboxypeptidase B (PCPB) is the prototype for those human exopeptidases that cleave off basic C-terminal residues and are secreted as inactive zymogens. One such protein is thrombin-activatable fibrinolysis inhibitor (TAFI), also known as plasma PCPB, which circulates in human plasma as a zymogen bound to plasminogen. The structure of human pancreatic PCPB displays a 95-residue pro-segment consisting of a globular region with an open-sandwich antiparallel-alpha antiparallel-beta topology and a C-terminal alpha-helix, which connects to the enzyme moiety. The latter is a 309-amino acid residue catalytic domain with alpha/beta hydrolase topology and a preformed active site, which is shielded by the globular domain of the pro-segment. The fold of the proenzyme is similar to previously reported procarboxypeptidase structures, also in that the most variable region is the connecting segment that links both globular moieties. However, the empty active site of human procarboxypeptidase B has two alternate conformations in one of the zinc-binding residues, which account for subtle differences in some of the key residues for substrate binding. The reported crystal structure, refined with data to 1.6A resolution, permits in the absence of an experimental structure, accurate homology modelling of TAFI, which may help to explain its properties.

Solution structure and backbone dynamics of an antigen-free heavy chain variable domain (VHH) from Llama

Camelids, (dromedaries, camels, and llamas) produce heavy-chains antibodies, with their antigen recognition sites composed of a single VH-like domain, referred to as VHH. The solution structure of one of these VHHs domains (VHH-H14), raised against the alpha subunit of the human chorionic gonadotropin hormone (hCG), has been determined by (15)N heteronuclear three-dimensional NMR spectroscopy. The framework is well resolved within the set of 20 best-calculated NMR structures and is close to that of classical VH domains from vertebrate antibodies, consisting of two antiparallel beta-sheets organized in a beta-barrel. Loops display a lower precision, especially the Complementarity Determining Regions (CDRs), involved in antigen recognition. Comparison of the three-dimensional VHH-H14 solution structure with its previously solved crystal structure (Spinelli et al., Nature Struct. Biol. 1996;3:752-757) reveals a high similarity to the framework, whereas significant conformational differences occur on CDRs, leading to the assumption that the antigen recognition site is a more mobile part. In order to deepen our insights into the dynamics of VHH-H14 in solution, (15)N relaxation was measured with longitudinal R1 and transverse R2 self-relaxation rates, and (15)N steady-state heteronuclear nuclear Overhauser enhancements (NOE), making it possible to probe picosecond-to-millisecond internal motions. Determination of dynamic parameters (S(2), tau(e), and Rex) through the Lipari-Szabo Model-free approach enables the identification of several regions with enhanced dynamics. Especially, the mobility measurements from NMR confirm that the antigen recognition site is the most mobile part of the VHH-H14 domain on picosecond-to-nanosecond fast time scales. Several residues belonging to the three CDRs are submitted to chemical exchange processes occurring on slow microsecond-to-millisecond time scales, suggesting that the formation of the VHH/antigen complex should be accompanied by structural changes.

Evaluating conformational free energies: the colony energy and its application to the problem of loop prediction

In this paper, we introduce a method to account for the shape of the potential energy curve in the evaluation of conformational free energies. The method is based on a procedure that generates a set of conformations, each with its own force-field energy, but adds a term to this energy that favors conformations that are close in structure (have a low rmsd) to other conformations. The sum of the force-field energy and rmsd-dependent term is defined here as the "colony energy" of a given conformation, because each conformation that is generated is viewed as representing a colony of points. The use of the colony energy tends to select conformations that are located in broad energy basins. The approach is applied to the ab initio prediction of the conformations of all of the loops in a dataset of 135 nonredundant proteins. By using an rmsd from a native criterion based on the superposition of loop stems, the average rmsd of 5-, 6-, 7-, and 8-residue long loops is 0.85, 0.92, 1.23, and 1.45 A, respectively. For 8-residue loops, 60 of 61 predictions have an rmsd of less than 3.0 A. The use of the colony energy is found to improve significantly the results obtained from the potential function alone. (The loop prediction program, "Loopy," can be downloaded at http://trantor.bioc.columbia.edu.)

A divide and conquer approach to fast loop modeling

We describe a fast ab initio method for modeling local segments in protein structures. The algorithm is based on a divide and conquer approach and uses a database of precalculated look-up tables, which represent a large set of possible conformations for loop segments of variable length. The target loop is recursively decomposed until the resulting conformations are small enough to be compiled analytically. The algorithm, which is not restricted to any specific loop length, generates a ranked set of loop conformations in 20-180 s on a desktop PC. The prediction quality is evaluated in terms of global RMSD. Depending on loop length the top prediction varies between 1.06 A RMSD for three-residue loops and 3.72 A RMSD for eight-residue loops. Due to its speed the method may also be useful to generate alternative starting conformations for complex simulations.

Synthesis, 3-D structure, and pharmacology of a reticulated chimeric peptide derived from maurotoxin and Tsk scorpion toxins

Maurotoxin (MTX) is a 34-mer scorpion toxin cross-linked by four disulfide bridges that acts on both Ca(2+)-activated (SK) and voltage-gated (Kv) K(+) channels. A 38-mer chimera of MTX, Tsk-MTX, has been synthesized by the solid-phase method. It encompasses residues from 1 to 6 of Tsk at N-terminal, and residues from 3 to 34 of MTX at C-terminal. As established by enzyme cleavage, Tsk-MTX displays half-cystine pairings of the type C1-C5, C2-C6, C3-C7 and C4-C8 which, contrary to MTX, correspond to a disulfide bridge pattern common to known scorpion toxins. The 3-D structure of Tsk-MTX, solved by (1)H NMR, demonstrates that it adopts the alpha/beta scaffold of scorpion toxins. In vivo, Tsk-MTX is lethal by intracerebroventricular injection in mice (LD(50) value of 0.2 microg/mouse). In vitro, Tsk-MTX is as potent as MTX, or Tsk, to interact with apamin-sensitive SK channels of rat brain synaptosomes (IC(50) value of 2.5 nM). It also blocks voltage-gated K(+) channels expressed in Xenopus oocytes, but is inactive on rat Kv1.3 contrary to MTX.

A novel main-chain anion-binding site in proteins: the nest. A particular combination of phi,psi values in successive residues gives rise to anion-binding sites that occur commonly and are found often at functionally important regions

Main-chain conformations where one amino acid residue can be described as gamma(R) (or alpha(R)) and an adjacent one as gamma(L) (or alpha(L)) mostly result in the three main-chain NH groups (of the two residues and the one following) forming a depression that can accommodate an atom with a whole or partial negative charge. We propose the name nest for this feature. The negatively charged atom, when present, is also stabilized by hydrogen-bonding with the NH groups. In an average protein, 8 % of residues are involved in a nest. The anion, or partially negatively charged atom, that often occupies the nest may be a main-chain carbonyl oxygen atom as in the paperclip, also called the Schellman loop, and the oxyanion hole of serine proteases. It can be a phosphate group, as in the P-loop superfamily that binds ATP and GTP. Overlapping, compound, nests are observed often, as in the P-loop, which has five successive NH groups that bind the beta phosphate group of nucleotide triphosphate. The longest compound nests are found surrounding cysteine-bound [2Fe2S] and [4Fe4S] iron-sulfur centers, which are also anionic; nests may encourage binding of the more reduced forms. The nest is a novel feature in the sense of not having been described as a unique motif with anion-binding potential before, although some of the situations where it occurs are familiar.

Protein folding: looping from hydrophobic nuclei

Protein structure can be viewed as a compact linear array of nearly standard size closed loops of 25-30 amino acid residues (Berezovsky et al., FEBS Letters 2000; 466: 283-286) irrespective of details of secondary structure. The end-to-end contacts in the loops are likely to be hydrophobic, which is a testable hypothesis. This notion could be verified by direct comparison of the loop maps with Kyte and Doolittle hydropathicity plots. This analysis reveals that most of the ends of the loops are hydrophobic, indeed. The same conclusion is reached on the basis of positional autocorrelation analysis of protein sequences of 23 fully sequenced bacterial genomes. Hydrophobic residues valine, alanine, glycine, leucine, and isoleucine appear preferentially at the 25-30 residues distance one from another. These observations open a new perspective in the understanding of protein structure and folding: a consecutive looping of the polypeptide chain with the loops ending primarily at hydrophobic nuclei.

The architecture of parallel beta-helices and related folds

Three-dimensional structures have been determined of a large number of proteins characterized by a repetitive fold where each of the repeats (coils) supplies a strand to one or more parallel beta-sheets. Some of these proteins form superfamilies of proteins, which have probably arisen by divergent evolution from a common ancestor. The classical example is the family including four families of pectinases without obviously related primary sequences, the phage P22 tailspike endorhamnosidase, chrondroitinase B and possibly pertactin from Bordetella pertusis. These show extensive stacking of similar residues to give aliphatic, aromatic and polar stacks such as the asparagine ladder. This suggests that coils can be added or removed by duplication or deletion of the DNA corresponding to one or more coils and explains how homologous proteins can have different numbers of coils. This process can also account for the evolution of other families of proteins such as the beta-rolls, the leucine-rich repeat proteins, the hexapeptide repeat family, two separate families of beta-helical antifreeze proteins and the spiral folds. These families need not be related to each other but will share features such as relative untwisted beta-sheets, stacking of similar residues and turns between beta-strands of approximately 90 degrees often stabilized by hydrogen bonding along the direction of the parallel beta-helix.Repetitive folds present special problems in the comparison of structures but offer attractive targets for structure prediction. The stacking of similar residues on a flat parallel beta-sheet may account for the formation of amyloid with beta-strands at right-angles to the fibril axis from many unrelated peptides.

Improved protein loop prediction from sequence alone

The SLoop database of supersecondary fragments, first described by Donate et al. (Protein Sci., 1996, 5, 2600-2616), contains protein loops, classified according to structural similarity. The database has recently been updated and currently contains over 10 000 loops up to 20 residues in length, which cluster into over 560 well populated classes. The database can be found at http://www-cryst.bioc.cam.ac.uk/~sloop. In this paper, we identify conserved structural features such as main chain conformation and hydrogen bonding. Using the original approach of Rufino and co-workers (1997), the correct structural class is predicted with the highest SLoop score for 35% of loops. This rises to 65% by considering the three highest scoring class predictions and to 75% in the top five scoring class predictions. Inclusion of residues from the neighbouring secondary structures and use of substitution tables derived using a reduced definition of secondary structure increase these prediction accuracies to 58, 78 and 85%, respectively. This suggests that capping residues can stabilize the loop conformation as well as that of the secondary structure. Further increases are achieved if only well-populated classes are considered in the prediction. These results correspond to an average loop root mean square deviation of between 0.4 and 2.6 A for loops up to five residues in length.

Loop fold nature of globular proteins

Protein chains make numerous returns in globules, thus forming loops, closed by tight residue-to-residue contacts-closed loops. Previous statistical analysis of the sizes and locations of the closed loops in all major protein folds revealed that the loops have an almost standard contour length of 25-30 amino acid residues and follow one after another along the chain. In this work the closed loops of the major folds are presented in three dimensions. A special image filtering procedure is introduced that allows one to visualize the standard size closed loops for the first time. The loop positions along the sequences are verified by detection of loop-end clusters.

Optimization of solvation models for predicting the structure of surface loops in proteins

A novel procedure for optimizing the atomic solvation parameters (ASPs) sigma(i) developed recently for cyclic peptides is extended to surface loops in proteins. The loop is free to move, whereas the protein template is held fixed in its X-ray structure. The energy is E(tot) = E(FF)(epsilon = nr) + summation operator sigma(i)A(i), where E(FF)(epsilon = nr) is the force-field energy of the loop-loop and loop-template interactions, epsilon = nr is a distance-dependent dielectric constant, and n is an additional parameter to be optimized. A(i) is the solvent-accessible surface area of atom i. The optimal sigma(i) and n are those for which the loop structure with the global minimum of E(tot)(n, sigma(i)) becomes the experimental X-ray structure. Thus, the ASPs depend on the force field and are optimized in the protein environment, unlike commonly used ASPs such as those of Wesson and Eisenberg (Protein Sci 1992;1:227-235). The latter are based on the free energy of transfer of small molecules from the gas phase to water and have been traditionally combined with various force fields without further calibration. We found that for loops the all-atom AMBER force field performed better than OPLS and CHARMM22. Two sets of ASPs [based on AMBER (n = 2)], optimized independently for loops 64-71 and 89-97 of ribonuclease A, were similar and thus enabled the definition of a best-fit set. All these ASPs were negative (hydrophilic), including those for carbon. Very good (i.e., small) root-mean-square-deviation values from the X-ray loop structure were obtained with the three sets of ASPs, suggesting that the best-fit set would be transferable to loops in other proteins as well. The structure of loop 13-24 is relatively stretched and was insensitive to the effect of the ASPs.

Review: protein design--where we were, where we are, where we're going

Protein design has become a powerful approach for understanding the relationship between amino acid sequence and 3-dimensional structure. In the past 5 years, there have been many breakthroughs in the development of computational methods that allow the selection of novel sequences given the structure of a protein backbone. Successful design of protein scaffolds has now paved the way for new endeavors to design function. The ability to design sequences compatible with a fold may also be useful in structural and functional genomics by expanding the range of proteins used for fold recognition and for the identification of functionally important domains from multiple sequence alignments.

Van der Waals locks: loop-n-lock structure of globular proteins

In a globular protein the polypeptide chain returns to itself many times, making numerous chain-to-chain contacts. The stability of these contacts is maintained primarily by van der Waals interactions. In this work we isolated and analysed van der Waals contacts that stabilise spatial structures of nine major folds. We suggest a specific way to identify the tightest contacts of prime importance for the stability of a given crystallized protein and introduce the notion of the van der Waals lock. The loops closed by the van der Waals interactions provide a basically novel view of protein globule organization: the loop-n-lock structure. This opens a new perspective in understanding protein folding as well: the consecutive looping of the polypeptide chain and the locking of the loop ends by tight van der Waals interactions.

Prediction of a 12-residue loop in bovine pancreatic trypsin inhibitor: effects of buried water

The x-ray conformations of 5-, 7-, 9-, and 12-residue loops in bovine pancreatic trypsin inhibitor (BPTI) were predicted by the use of multiple independent Monte Carlo simulating annealing (MCSA) runs starting from random conformations. Four buried water molecules interacted with a 12-residue loop that started at residue 8 and ended at residue 19, and that included the binding region. The final conformation at the end of an MCSA run was characterized. Solvation free energy based on the solvent accessible surface area was included in the energy function at low simulated annealing temperatures. Conformational states were interactively separated by a recently developed algorithm. Computed loops were characterized in terms of total energy, and backbone and side chain root mean square deviations (RMSDs) between computed native loop conformations and the x-ray conformation. The 12-residue loop was computed with and without buried water [called WL12(8-19) and L12(8-19), respectively]. The backbone was reliably and reproducibly computed to within 1.1 A in L12(8-19) and 0.9 A in WL12(8-19). L12(8-19) required significantly more MCSA runs to achieve the same level of reproducibility as WL12(8-19). Based on the size of the cluster of low energy native loop conformations, and the computational effort, WL12(8-19) had greater entropy. In calculations of 7-, 9-, and 12-residue loops without buried water, the effects of buried water became obvious in the 12-residue loop calculation, which interacted with all four buried water molecules. Nearly all conformations of the native loop conformer had a hydrogen bond between the Lys 15 side chain and the backbone of Gly 12, Pro 13, and Cys 14, which may have implications in the rate of exchange of buried water with bulk solvent and in protein folding. The present version of MCSA program was more efficient than earlier versions.

CODA: a combined algorithm for predicting the structurally variable regions of protein models

CODA, an algorithm for predicting the variable regions in proteins, combines FREAD a knowledge based approach, and PETRA, which constructs the region ab initio. FREAD selects from a database of protein structure fragments with environmentally constrained substitution tables and other rule-based filters. FREAD was parameterized and tested on over 3000 loops. The average root mean square deviation ranged from 0.78 A for three residue loops to 3.5 A for eight residue loops on a nonhomologous test set. CODA clusters the predictions from the two independent programs and makes a consensus prediction that must pass a set of rule-based filters. CODA was parameterized and tested on two unrelated separate sets of structures that were nonhomologous to one another and those found in the FREAD database. The average root mean square deviation in the test set ranged from 0.76 A for three residue loops to 3.09 A for eight residue loops. CODA shows a general improvement in loop prediction over PETRA and FREAD individually. The improvement is far more marked for lengths six and upward, probably as the predictive power of PETRA becomes more important. CODA was further tested on several model structures to determine its applicability to the modeling situation. A web server of CODA is available at http://www-cryst.bioc.cam.ac.uk/~charlotte/Coda/search_coda.html.

Comparative protein structure modeling of genes and genomes

Comparative modeling predicts the three-dimensional structure of a given protein sequence (target) based primarily on its alignment to one or more proteins of known structure (templates). The prediction process consists of fold assignment, target-template alignment, model building, and model evaluation. The number of protein sequences that can be modeled and the accuracy of the predictions are increasing steadily because of the growth in the number of known protein structures and because of the improvements in the modeling software. Further advances are necessary in recognizing weak sequence-structure similarities, aligning sequences with structures, modeling of rigid body shifts, distortions, loops and side chains, as well as detecting errors in a model. Despite these problems, it is currently possible to model with useful accuracy significant parts of approximately one third of all known protein sequences. The use of individual comparative models in biology is already rewarding and increasingly widespread. A major new challenge for comparative modeling is the integration of it with the torrents of data from genome sequencing projects as well as from functional and structural genomics. In particular, there is a need to develop an automated, rapid, robust, sensitive, and accurate comparative modeling pipeline applicable to whole genomes. Such large-scale modeling is likely to encourage new kinds of applications for the many resulting models, based on their large number and completeness at the level of the family, organism, or functional network.

Solution structure of the module X2 1 of unknown function of the cellulosomal scaffolding protein CipC of Clostridium cellulolyticum

Multidimensional, homo- and heteronuclear magnetic resonance spectroscopy combined with dynamical annealing has been used to determine the structure of a 94 residue module (X2 1) of the scaffolding protein CipC from the anaerobic bacterium Clostridium cellulolyticum. An experimental data set comprising 1647 nuclear Overhauser effect-derived restraints, 105 hydrogen bond restraints and 66 phi torsion angle restraints was used to calculate 20 converging final solutions. The calculated structures have an average rmsd about the mean structure of 0.55(+/-0.11) A for backbone atoms and 1.40(+/-0.11) A for all heavy atoms when fitted over the secondary structural elements. The X2 1 module has an immunoglobulin-like fold with two beta-sheets packed against each other. One sheet contains three strands, the second contains four strands. An additional strand is intercalated between the beta-sandwich, as well as two turns of a 3(.10) helix. X2 1 has a surprising conformational stability and may act as a conformational linker and solubility enhancer within the scaffolding protein. The fold of X2 1 is very similar to that of telokin, titin Ig domain, hemolin D2 domain, twitchin immunoglobulin domain and the first four domains of the IgSF portion of transmembrane cell adhesion molecule. As a consequence, the X2 1 module is the first prokaryotic member assigned to the I set of the immunoglobulin superfamily even though no sequence similarity with any member of this superfamily could be detected.

Modeling of loops in protein structures

Comparative protein structure prediction is limited mostly by the errors in alignment and loop modeling. We describe here a new automated modeling technique that significantly improves the accuracy of loop predictions in protein structures. The positions of all nonhydrogen atoms of the loop are optimized in a fixed environment with respect to a pseudo energy function. The energy is a sum of many spatial restraints that include the bond length, bond angle, and improper dihedral angle terms from the CHARMM-22 force field, statistical preferences for the main-chain and side-chain dihedral angles, and statistical preferences for nonbonded atomic contacts that depend on the two atom types, their distance through space, and separation in sequence. The energy function is optimized with the method of conjugate gradients combined with molecular dynamics and simulated annealing. Typically, the predicted loop conformation corresponds to the lowest energy conformation among 500 independent optimizations. Predictions were made for 40 loops of known structure at each length from 1 to 14 residues. The accuracy of loop predictions is evaluated as a function of thoroughness of conformational sampling, loop length, and structural properties of native loops. When accuracy is measured by local superposition of the model on the native loop, 100, 90, and 30% of 4-, 8-, and 12-residue loop predictions, respectively, had <2 A RMSD error for the mainchain N, C(alpha), C, and O atoms; the average accuracies were 0.59 +/- 0.05, 1.16 +/- 0.10, and 2.61 +/- 0.16 A, respectively. To simulate real comparative modeling problems, the method was also evaluated by predicting loops of known structure in only approximately correct environments with errors typical of comparative modeling without misalignment. When the RMSD distortion of the main-chain stem atoms is 2.5 A, the average loop prediction error increased by 180, 25, and 3% for 4-, 8-, and 12-residue loops, respectively. The accuracy of the lowest energy prediction for a given loop can be estimated from the structural variability among a number of low energy predictions. The relative value of the present method is gauged by (1) comparing it with one of the most successful previously described methods, and (2) describing its accuracy in recent blind predictions of protein structure. Finally, it is shown that the average accuracy of prediction is limited primarily by the accuracy of the energy function rather than by the extent of conformational sampling.

Role of amino acid residues at turns in the conformational stability and folding of human lysozyme

To clarify the role of amino acid residues at turns in the conformational stability and folding of a globular protein, six mutant human lysozymes deleted or substituted at turn structures were investigated by calorimetry, GuHCl denaturation experiments, and X-ray crystal analysis. The thermodynamic properties of the mutant and wild-type human lysozymes were compared and discussed on the basis of their three-dimensional structures. For the deletion mutants, Delta47-48 and Delta101, the deleted residues are in turns on the surface and are absent in human alpha-lactalbumin, which is homologous to human lysozyme in amino acid sequence and tertiary structure. The stability of both mutants would be expected to increase due to a decrease in conformational entropy in the denatured state; however, both proteins were destabilized. The destabilizations were mainly caused by the disappearance of intramolecular hydrogen bonds. Each part deleted was recovered by the turn region like the alpha-lactalbumin structure, but there were differences in the main-chain conformation of the turn between each deletion mutant and alpha-lactalbumin even if the loop length was the same. For the point mutants, R50G, Q58G, H78G, and G37Q, the main-chain conformations of these substitution residues located in turns adopt a left-handed helical region in the wild-type structure. It is thought that the left-handed non-Gly residue has unfavorable conformational energy compared to the left-handed Gly residue. Q58G was stabilized, but the others had little effect on the stability. The structural analysis revealed that the turns could rearrange the main-chain conformation to accommodate the left-handed non-Gly residues. The present results indicate that turn structures are able to change their main-chain conformations, depending upon the side-chain features of amino acid residues on the turns. Furthermore, stopped-flow GuHCl denaturation experiments on the six mutants were performed. The effects of mutations on unfolding-refolding kinetics were significantly different among the mutant proteins. The deletion/substitutions in turns located in the alpha-domain of human lysozyme affected the refolding rate, indicating the contribution of turn structures to the folding of a globular protein.

Refinement of modelled structures by knowledge-based energy profiles and secondary structure prediction: application to the human procarboxypeptidase A2

Knowledge-based energy profiles combined with secondary structure prediction have been applied to molecular modelling refinement. To check the procedure, three different models of human procarboxypeptidase A2 (hPCPA2) have been built using the 3D structures of procarboxypeptidase A1 (pPCPA1) and bovine procarboxypeptidase A (bPCPA) as templates. The results of the refinement can be tested against the X-ray structure of hPCPA2 which has been recently determined. Regions miss-modelled in the activation segment of hPCPA2 were detected by means of pseudo-energies using Prosa II and modified afterwards according to the secondary structure prediction. Moreover, models obtained by automated methods as COMPOSER, MODELLER and distance restraints have also been compared, where it was found possible to find out the best model by means of pseudo-energies. Two general conclusions can be elicited from this work: (1) on a given set of putative models it is possible to distinguish among them the one closest to the crystallographic structure, and (2) within a given structure it is possible to find by means of pseudo-energies those regions that have been defectively modelled.

Solution structure of BmKTX, a K+ blocker toxin from the Chinese scorpion Buthus Martensi

BmKTX is a toxin recently purified from the venom of Buthus Martensi, which belongs to the kaliotoxin family. We have determined its solution structure by use of conventional two-dimensional NMR techniques followed by distance-geometry and energy minimization. The calculated structure is composed of a short alpha-helix (residues 14 to 20) connected by a tight turn to a two-stranded antiparallel beta-sheet (sequences 25-27 and 32-34). The beta-turn connecting these strands belongs to type I. The N-terminal segment (sequence 1 to 8) runs parallel to the beta-sheet although it cannot be considered as a third strand. Comparison of the conformation of BmKTX and toxins of the kaliotoxin family clearly demonstrates that they are highly related. Therefore, analysis of the residues belonging to the interacting surface of those toxins allows us to propose a functional map of BmKTX slightly different from the one of KTX and AgTX2, which may explain the variations in affinities of these toxins towards the Kv1.3 channels.

Exploring the conformational diversity of loops on conserved frameworks

Loops are structurally variable regions, but the secondary structural elements bracing loops are often conserved. Motifs with similar secondary structures exist in the same and different protein families. In this study, we made an all-PDB-based analysis and produced 495 motif families accessible from the Internet. Every motif family contains some variable loops spanning a common framework (a pair of secondary structures). The diversity of loops and the convergence of frameworks were examined. In addition, we also identified 119 loops with conformational changes in different PDB files. These materials can give some directions for functional loop design and flexible docking.

Skewed distribution of protein secondary structure contents over the conformational triangle

A conformational triangle method is presented to analyze the secondary structure contents of 1028 structurally known proteins in the non-redundant data set of the recent 25% PDB_SELECT. The secondary structure contents of each protein are mapped on to a point in the triangle. It was found that the distribution of the 1028 points is strongly skewed in the triangle and about 42% of the whole area is empty, which is called the forbidden area. The detailed border between the allowable and forbidden areas was calculated. The possible explanation of the skewed distribution is discussed. The distributions of the mapping points for enzymes and non-enzymes in this non-redundant data set are compared. It was found that a necessary rather than a sufficient condition for an enzyme molecule is that its coil content must be >/=0.223. It is hoped that the skewed distribution observed here could be used to test the secondary structure and threading predictions.

New efficient statistical sequence-dependent structure prediction of short to medium-sized protein loops based on an exhaustive loop classification

A bank of 13,563 loops from three to eight amino acid residues long, representing motifs between two consecutive regular secondary structures, has been derived from protein structures presenting less than 95 % sequence identity. Statistical analyses of occurrences of conformations and residues revealed length-dependent over-representations of particular amino acids (glycine, proline, asparagine, serine, and aspartate) and conformations (alphaL, epsilon, betaPregions of the Ramachandran plot). A position-dependent distribution of these occurrences was observed for N and C-terminal residues, which are correlated to the nature of the flanking regions. Loops of the same length were clustered into statistically meaningful families on the basis of their backbone structures when placed in a common reference frame, independent of the flanks. These clusters present significantly different distributions of sequence, conformations, and endpoint residue Calphadistances. On the basis of the sequence-structure correlation of this clustering, an automatic loop modeling algorithm was developed. Based on the knowledge of its sequence and of its flank backbone structures each query loop is assigned to a family and target loop supports are selected in this family. The support backbones of these target loops are then adjusted on flanking structures by partial exploration of the conformational space. Loop closure is performed by energy minimization for each support and the final model is chosen among connected supports based upon energy criteria. The quality of the prediction is evaluated by the root-mean-square deviation (rmsd) between the final model and the native loops when the whole bank is re-attributed on itself with a Jackknife test. This average rmsd ranges from 1.1 A for three-residue loops to 3.8 A for eight-residue loops. A few poorly predicted loops are inescapable, considering the high level of diversity in loops and the lack of environment data. To overcome such modeling problems, a statistical reliability score was assigned for each prediction. This score is correlated to the quality of the prediction, in terms of rmsd, and thus improves the selection accuracy of the model. The algorithm efficiency was compared to CASP3 target loop predictions. Moreover, when tested on a test loop bank, this algorithm was shown to be robust when the loops are not precisely delimited, therefore proving to be a useful tool in practice for protein modeling.

Protein loops on structurally similar scaffolds: database and conformational analysis

A general problem in comparative modeling and protein design is the conformational evaluation of loops with a certain sequence in specific environmental protein frameworks. Loops of different sequences and structures on similar scaffolds are common in the Protein Data Bank (PDB). In order to explore both structural and sequential diversity of them, a data base of loops connecting similar secondary structure fragments is constructed by searching the data base of families of structurally similar proteins and PDB. A total of 84 loop families having 2-13 residues are found among the well-determined structures of resolution better than 2.5 A. Eight alpha-alpha, 20 alpha-beta, 19 beta-alpha, and 37 beta-beta families are identified. Every family contains more than 5 loop motifs. In each family, no loops share same sequence and all the frameworks are well superimposed. Forty-three new loop classes are distinguished in the data base. The structural variability of loops in homologous proteins are examined and shown in 44 families. Motif families are characterized with geometric parameters and sequence patterns. The conformations of loops in each family are clustered into subfamilies using average linkage cluster analysis method. Information such as geometric properties, sequence profile, sequential and structural variability in loop, structural alignment parameters, sequence similarities, and clustering results are provided. Correlations between the conformation of loops and loop sequence, motif sequence, and global sequence of PDB chain are examined in order to find how loop structures depend on their sequences and how they are affected by the local and global environment. Strong correlations (R > 0.75) are only found in 24 families. The best R value is 0.98. The data base is available through the Internet.

A natural grouping of motifs with an aspartate or asparagine residue forming two hydrogen bonds to residues ahead in sequence: their occurrence at alpha-helical N termini and in other situations

Examination of the ways side-chain carboxylate and amide groups in high-resolution protein crystal structures form hydrogen bonds with main-chain atoms reveals that the most common category is a two-hydrogen-bond four to five residue motif with an aspartate or asparagine (Asx) at the first residue, for which we propose the name Asx-motif. Similar motifs with glutamate or glutamine residues at that position are rare. Asx-motifs occur typically as (1) a common feature of the N termini of alpha-helices called the Asx N-cap motif; (2) an independent motif, usually a beta-turn with an appropriately hydrogen-bonded Asx as the first residue; and (3) a motif incorporated in a beta-bulge loop. Asx-motifs are common, there being just under two-and-a-half in an average-sized protein subunit; of these, about 55 % are Asx N-cap motifs. Because they occur often in many situations, it seems that these motifs have an inherent propensity to form on their own rather than just being a feature stabilised at the end of a helix. Asx-motifs also occur in functionally interesting situations in aspartyl proteases, citrate synthase, EF hands, haemoglobins, lipocalins, glutathione reductase and the alpha/beta hydrolases.

Solution structure of potassium channel-inhibiting scorpion toxin Lq2

Lq2 is a unique scorpion toxin. Acting from the extracellular side, Lq2 blocks the ion conduction pore in not only the voltage- and Ca2+ -activated channels, but also the inward-rectifier K+ channels. This finding argues that the three-dimensional structures of the pores in these K+ channels are similar. However, the amino acid sequences that form the external part of the pore are minimally conserved among the various classes of K+ channels. Because Lq2 can bind to all the three classes of K+ channels, we can use Lq2 as a structural probe to examine how the non-conserved pore-forming sequences are arranged in space to form similar pore structures. In the present study, we determined the three-dimensional structure of Lq2 using nuclear magnetic resonance (NMR) techniques. Lq2 consists of an alpha-helix (residues S10 to L20) and a beta-sheet, connected by an alphabeta3 loop (residues N22 to N24). The beta-sheet has two well-defined anti-parallel strands (residues G26 to M29 and residues K32 to C35), which are connected by a type I' beta-turn centered between residues N30 and K31. The N-terminal segment (residues Z1 to T8) appears to form a quasi-third strand of the beta-sheet.

Solution structure of two new toxins from the venom of the Chinese scorpion Buthus martensi Karsch blockers of potassium channels

The solution structure of BmTX2 purified from the venom of the Chinese Buthid Buthus martensi has been determined by 2D NMR spectroscopy techniques which led to the description of its 3D conformation. The structure consists of a triple-stranded beta-sheet connected to a helical structure. This helix encompasses 10 residues, from 11 to 20, begins with a turn of 310 helix, and ends with an alpha helix. The three strands of beta sheet comprise residues 2-6, with a bulge covering residues 4 and 5, 26-29, and 32-35, with a type I' beta turn centered on residues 30-31. We also characterized the solution structure of BmTX1. The two toxins which are potent blockers of both large-conductance calcium-activated potassium channels (BKCa channels) and voltage-gated potassium channels (Kv1. 3) are highly superimposable and possess the same structural characteristics. Analysis of these structures allows us to hypothesize that, besides the main surface of interaction described by the functional map of charybdotoxin, one can expect that the binding of scorpion toxins on BKCa channels may involve residues on the edge of this surface.

Structure and evolution of parallel beta-helix proteins

Three bacterial pectate lyases, a pectin lyase from Aspergillus niger, the structures of rhamnogalacturonase A from Aspergillus aculeatus, RGase A, and the P22-phage tailspike protein, TSP, display the right-handed parallel beta-helix architecture first seen in pectate lyase. The lyases have 7 complete coils while RGase A and TSP have 11 and 12, respectively. Each coil contains three beta-strands and three turn regions named PB1, T1, PB2, T2, PB3, and T3 in their order of occurrence. The lyases have homologous sequences but RGase A and TSP do not show obvious sequence homology either to the lyases or to each other. However, the structural similarities between all these molecules are so extensive that divergence from a common ancestor is much more probable than convergence to the same fold. The region PB2-T2-PB3 is the best conserved region in the lyases and shows the clearest structural similarity. Not only is the pleating and the direction of the hydrogen bonding in the sheets conserved, but so is the unusual alphaL-conformation turn between the two sheets. However, the overall shape, the position of long loops, a conserved alpha-helix that covers the amino-terminal end of the parallel beta-helix and stacks of residues in alphaR-conformation at the start of PB1 all suggest a common ancestor. The functional similarity, that the enzymes all bind alpha-galactose containing polymers at an equivalent site involving PB1 and its two flanking turn regions, further supports divergent evolution. We suggest that the stacking of the coils and the unusual near perpendicular junction of PB2 and PB3 make the parallel beta-helix fold especially likely to maintain similar main chain conformations during divergent evolution even after all vestige of similarity in primary structure has vanished.

Prediction of local structure in proteins using a library of sequence-structure motifs

We describe a new method for local protein structure prediction based on a library of short sequence pattern that correlate strongly with protein three-dimensional structural elements. The library was generated using an automated method for finding correlations between protein sequence and local structure, and contains most previously described local sequence-structure correlations as well as new relationships, including a diverging type-II beta-turn, a frayed helix, and a proline-terminated helix. The query sequence is scanned for segments 7 to 19 residues in length that strongly match one of the 82 patterns in the library. Matching segments are assigned the three-dimensional structure characteristic of the corresponding sequence pattern, and backbone torsion angles for the entire query sequence are then predicted by piecing together mutually compatible segment predictions. In predictions of local structure in a test set of 55 proteins, about 50% of all residues, and 76% of residues covered by high-confidence predictions, were found in eight-residue segments within 1.4 A of their true structures. The predictions are complementary to traditional secondary structure predictions because they are considerably more specific in turn regions, and may contribute to ab initio tertiary structure prediction and fold recognition.

Protein structure prediction

Genome sequencing projects continue to provide a flood of new protein sequences, and prediction methods remain an important means of adding structural information. Recently, there have been advances in secondary structure prediction, which feed, in turn, into improved fold recognition algorithms. Finally, there have been technical improvements in comparative modelling, and studies of the expected accuracy of three-dimensional structural models built by this method.

Automated classification of antibody complementarity determining region 3 of the heavy chain (H3) loops into canonical forms and its application to protein structure prediction

A computer-based algorithm was used to cluster the loops forming the complementarity determining region (CDR) 3 of the heavy chain (H3) into canonical classes. Previous analyses of the three-dimensional structures of CDR loops (also known as the hypervariable regions) within antibody immunoglobulin variable domains have shown that for five of the six CDRs there are only a few main-chain conformations (known as canonical forms) that show clear relationships between sequence and structure. However, the larger variation in length and conformation of loops within H3 has limited the classification of these loops into canonical forms. The clustering procedure presented here is based on aligning the Ramachandran-coded main-chain conformation of the residues using a dynamic algorithm that allows the insertion of gaps to obtain an optimum alignment. A total of 41 H3 loops out of 62 non-identical loops, extracted from the Brookhaven Protein Data Bank, have been automatically grouped into 22 clusters. Inspection of the clusters for consensus sequences or intra-loop interactions or invariant conformation led to the proposal of 13 canonical forms representing 31 loops. These canonical forms include a consideration of the geometry of both the take-off region adjacent to the bracing beta-strands and the remaining loop apex. Subsequently a new set of 15 H3 loops not included in the initial analysis was considered. The clustering procedure was repeated and nine of these 15 loops could be assigned to original clusters, including seven to canonical forms. A sequence profile was generated for each canonical form from the original set of loops and matched against the sequences of the new H3 loops. For five out of the seven new H3 loops that were in a canonical form, the correct form was identified at first rank by this predictive scheme.

Potato carboxypeptidase inhibitor, a T-knot protein, is an epidermal growth factor antagonist that inhibits tumor cell growth

Epidermal growth factor (EGF) and its receptor (EGFR) are involved in many aspects of the development of carcinomas, including tumor cell growth, vascularization, invasiveness, and metastasis. Because EGFR has been found to be overexpressed in many tumors of epithelial origin, it is a potential target for antitumor therapy. Here we report that potato carboxypeptidase inhibitor (PCI), a 39-amino acid protease inhibitor with three disulfide bridges, is an antagonist of human EGF. It competed with EGF for binding to EGFR and inhibited EGFR activation and cell proliferation induced by this growth factor. PCI suppressed the growth of several human pancreatic adenocarcinoma cell lines, both in vitro and in nude mice. PCI has a special disulfide scaffold called a T-knot that is also present in several growth factors including EGF and transforming growth factor alpha. PCI shows structural similarities with these factors, a fact that can explain the antagonistic effect of the former. This is the first reported example of an antagonistic analogue of human EGF.

Structure, function, and temperature sensitivity of directed, random mutants at proline 76 and glycine 77 in omega-loop D of yeast iso-1-cytochrome c

Residues 75-78 form a tight turn within Omega-loop D in Saccharomyces cerevisiae iso-1-cytochrome c. Directed, random mutagenesis of invariant residues proline 76 and glycine 77 in this turn were analyzed for the in vivo functionality and level of protein within the cell. All proteins, except Pro76Val, also exhibit a significant decrease in intracellular cytochrome c levels, ranging from 15% to 80% of wild type. Furthermore, all isolated mutant strains, except the one expressing Pro76Val, exhibit a significant decrease in growth on lactate medium, suggesting that the variant cytochromes are much less functional than wild type. This requirement for protein function is clearly the cause for the strict invariance of these residues in eukaryotic cytochromes c. Seven proteins with mutations just at Pro76 were purified and studied by circular dichroism spectroscopy. All proteins with mutations at Pro76 exhibit melting temperatures about 7 degreesC less than that of the wild-type protein, suggesting that mutation of Pro76 affects the entropy of the denatured state. It is proposed that the functional significance of Pro76 and Gly77 is the requirement for a type II (betagammaL) beta-turn in this loop, the conformation of which requires a glycine at the third position, and that a change occurs in this turn conformation upon a change in the redox state of the protein.

Helix capping

Helix-capping motifs are specific patterns of hydrogen bonding and hydrophobic interactions found at or near the ends of helices in both proteins and peptides. In an alpha-helix, the first four >N-H groups and last four >C=O groups necessarily lack intrahelical hydrogen bonds. Instead, such groups are often capped by alternative hydrogen bond partners. This review enlarges our earlier hypothesis (Presta LG, Rose GD. 1988. Helix signals in proteins. Science 240:1632-1641) to include hydrophobic capping. A hydrophobic interaction that straddles the helix terminus is always associated with hydrogen-bonded capping. From a global survey among proteins of known structure, seven distinct capping motifs are identified-three at the helix N-terminus and four at the C-terminus. The consensus sequence patterns of these seven motifs, together with results from simple molecular modeling, are used to formulate useful rules of thumb for helix termination. Finally, we examine the role of helix capping as a bridge linking the conformation of secondary structure to supersecondary structure.