Searching for signatures of cold adaptations in modern and archaic humans: hints from the brown adipose tissue genes

Adaptation to low temperatures has been reasonably developed in the human species during the colonization of the Eurasian landmass subsequent to Out of Africa migrations of anatomically modern humans. In addition to morphological and cultural changes, also metabolic ones are supposed to have favored human isolation from cold and body heat production and this can be hypothesized also for most Neandertal and at least for some Denisovan populations, which lived in geographical areas that strongly experienced the last glacial period. Modulation of non-shivering thermogenesis, for which adipocytes belonging to the brown adipose tissue are the most specialized cells, might have driven these metabolic adaptations. To perform an exploratory analysis aimed at looking into this hypothesis, variation at 28 genes involved in such functional pathway was investigated in modern populations from different climate zones, as well as in Neandertal and Denisovan genomes. Patterns of variation at the LEPR gene, strongly related to increased heat dissipation by mitochondria, appeared to have been shaped by positive selection in modern East Asians, but not in Europeans. Moreover, a single potentially cold-adapted LEPR allele, different from the supposed adaptive one identified in Homo sapiens, was found also in Neandertal and Denisovan genomes. These findings suggest that independent mechanisms for cold adaptations might have been developed in different non-African human groups, as well as that the evolution of possible enhanced thermal efficiency in Neandertals and in some Denisovan populations has plausibly entailed significant changes also in other functional pathways than in the examined one.

High-throughput SNP discovery in the rabbit (Oryctolagus cuniculus) genome by next-generation semiconductor-based sequencing

The European rabbit (Oryctolagus cuniculus) is a domesticated species with one of the broadest ranges of economic and scientific applications and fields of investigation. Rabbit genome information and assembly are available (oryCun2.0), but so far few studies have investigated its variability, and massive discovery of polymorphisms has not been published yet for this species. Here, we sequenced two reduced representation libraries (RRLs) to identify single nucleotide polymorphisms (SNPs) in the rabbit genome. Genomic DNA of 10 rabbits belonging to different breeds was pooled and digested with two restriction enzymes (HaeIII and RsaI) to create two RRLs which were sequenced using the Ion Torrent Personal Genome Machine. The two RRLs produced 2 917 879 and 4 046 871 reads, for a total of 280.51 Mb (248.49 Mb with quality >20) and 417.28 Mb (360.89 Mb with quality >20) respectively of sequenced DNA. About 90% and 91% respectively of the obtained reads were mapped on the rabbit genome, covering a total of 15.82% of the oryCun2.0 genome version. The mapping and ad hoc filtering procedures allowed to reliably call 62 491 SNPs. SNPs in a few genomic regions were validated by Sanger sequencing. The Variant Effect Predictor Web tool was used to map SNPs on the current version of the rabbit genome. The obtained results will be useful for many applied and basic research programs for this species and will contribute to the development of cost-effective solutions for high-throughput SNP genotyping in the rabbit.

The prediction of protein-protein interacting sites in genome-wide protein interaction networks: the test case of the human cell cycle

In this paper we aim at investigating possible correlations between the number of putative interaction patches of a given protein, as inferred by an algorithm that we have developed, and its degree (number of edges of the protein node in a protein interaction network). We focus on the human cell cycle that, as compared with other biological processes, comprises the largest number of proteins whose structure is known at atomic resolution both as monomers and as interacting complexes. For predicting interaction patches we specifically develop a HM-SVM based method reaching 71% overall accuracy with a correlation coefficient value equal to 0.43 on a non redundant set of protein complexes. To test the biological meaning of our predictions, we also explore whether interacting patches contain energetically important residues and/or disease related mutations and find that predicted patches are endowed with both features. Based on this, we propose that mapping the protein with all the predicted interaction patches bridges the molecule to the interactome at the cell level. To test our hypothesis we downloaded interaction data from interaction data bases and find that the number of predicted interaction patches significantly correlates (Pearson correlation value >0.3) with the number of the known interactions (edges) per protein in the human interactome, as contained in MINT and IntAct. We also show that the correlation increases (Pearson correlation value >0.5) when the subcellular co-localization and the co-expression levels of the interacting partners are taken into account.

Terlipressin infusion induces ischemia of breast skin in a cirrothic patient with hepatorenal syndrome

We report a case of a 65-year-old woman with hepatitis C virus-related decompensated cirrhosis with hepatorenal syndrome, treated by high dose of terlipressin. Few hours after the highest dose was started, the patient complained burning pain in breasts, followed by the development of extensive bilateral cyanosis of breast's skin. When terlipressin was immediately stopped, pain and skin cyanosis rapidly disappeared. The peculiarity of our case is that cyanosis did not develop in common peripheral sites (e.g. fingers, toes, etc.) but in an atypical area, as skin of the breasts. Probably, this particular behaviour could be explained by the anatomical position of her large size breasts, that resulting as an extremely sloping and stretching region thus filling the maximum effect of gravity.

Binding mode of the tyrosine kinase inhibitor bosutinib (SKI-606) to Abl kinase

7049

Background: Bosutinib (SKI-606) is a 4-anilino-3-quinolinecarbonitrile Src/Abl kinase inhibitor. The crystal structure of Abl kinase domain in complex with bosutinib has not yet been determined and the mode of binding of this inhibitor is therefore unknown. Methods: In the present study, we used a molecular docking approach to investigate the following: a) Which Abl conformation(s) bosutinib is more likely to bind? b) Which Abl residues are likely to be involved in bosutinib binding? c) Which imatinib-resistant Abl mutants bosutinib is likely to be active against? A three-dimensional model of bosutinib was generated by Chemsketch ( http://www.acdlabs.com ). Modelling of the human Abl kinase in its inactive (closed) conformation was performed with the software Modeller, v7.7 ( http://salilab.org/modeller ) adopting the highly related Mus musculus Abl homologue as a template structure (PDB entry: 1OPJ , 0.175nm resolution). Active (open) Abl conformation is available as PDB entry 1OPL . Flexible docking of bosutinib to Abl was performed with Autodock v3.0 ( http://www.scripps.edu/mb/olson ). Results: Our binding hypothesis predicts that a) similarly to IM, the interaction between bosutinib and Abl seems to be more stable when the latter is in its inactive (closed) conformation; b) 21 Abl residues (including Y253, T315, F317 and F359 among others) surround bosutinib and engage H-bonds or Van der Waals interactions; c) bosutinib is likely to retain the ability to bind and inhibit the IM-resistant Y253H, M351T, F359V, F317L and E255V Abl mutants. In contrast, bosutinib binding to E255K and T315I mutants is predicted to be less efficient. Detailed models will be presented. Conclusions: Bosutinib is a promising second-line treatment for Ph+ leukemia patients harbouring several IM-resistant Abl mutations. The T315I remains the biggest hurdle for pharmacological therapy of Ph+ leukemias. Supported by European LeukemiaNet, AIRC, AIL, COFIN and FIRB projects, and by Fondazione del Monte di Bologna e Ravenna.

No significant financial relationships to disclose.

Molecular mechanism of thioredoxin regulation in photosynthetic A2B2-glyceraldehyde-3-phosphate dehydrogenase

Chloroplast glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is a light-regulated, NAD(P)H-dependent enzyme involved in plant photosynthetic carbon reduction. Unlike lower photosynthetic organisms, which only contain A(4)-GAPDH, the major GAPDH isoform of land plants is made up of A and B subunits, the latter containing a C-terminal extension (CTE) with fundamental regulatory functions. Light-activation of AB-GAPDH depends on the redox state of a pair of cysteines of the CTE, which can form a disulfide bond under control of thioredoxin f, leading to specific inhibition of the NADPH-dependent activity. The tridimensional structure of A(2)B(2)-GAPDH from spinach chloroplasts, crystallized in the oxidized state, shows that each disulfide-containing CTE is docked into a deep cleft between a pair of A and B subunits. The structure of the CTE was derived from crystallographic data and computational modeling and confirmed by site-specific mutagenesis. Structural analysis of oxidized A(2)B(2)-GAPDH and chimeric mutant [A+CTE](4)-GAPDH revealed that Arg-77, which is essential for coenzyme specificity and high NADPH-dependent activity, fails to interact with NADP in these kinetically inhibited GAPDH tetramers and is attracted instead by negative residues of oxidized CTE. Other subtle changes in catalytic domains and overall conformation of the tetramers were noticed in oxidized A(2)B(2)-GAPDH and [A+CTE](4)-GAPDH, compared with fully active A(4)-GAPDH. The CTE is envisioned as a redox-sensitive regulatory domain that can force AB-GAPDH into a kinetically inhibited conformation under oxidizing conditions, which also occur during dark inactivation of the enzyme in vivo.

K-Fold: a tool for the prediction of the protein folding kinetic order and rate

K-Fold is a tool for the automatic prediction of the protein folding kinetic order and rate. The tool is based on a support vector machine (SVM) that was trained on a data set of 63 proteins, whose 3D structure and folding mechanism are known from experiments already described in the literature. The method predicts whether a protein of known atomic structure folds according to a two-state or a multi-state kinetics and correctly classifies 81% of the folding mechanisms when tested over the training set of the 63 proteins. It also predicts as a further option the logarithm of the folding rate. To the best of our knowledge, the tool discriminates for the first time whether a protein is characterized by a two state or a multiple state kinetics, during the folding process, and concomitantly estimates also the value of the constant rate of the process. When used to predict the logarithm of the folding rate, K-Fold scores with a correlation value to the experimental data of 0.74 (with a SE of 1.2).

AVAILABILITY

http://gpcr.biocomp.unibo.it/cgi/predictors/K-Fold/K-Fold.cgi.

SUPPLEMENTARY INFORMATION

http://gpcr.biocomp.unibo.it/~emidio/K-Fold/K-Fold_help.html.

Predicting the insurgence of human genetic diseases associated to single point protein mutations with support vector machines and evolutionary information

MOTIVATION

Human single nucleotide polymorphisms (SNPs) are the most frequent type of genetic variation in human population. One of the most important goals of SNP projects is to understand which human genotype variations are related to Mendelian and complex diseases. Great interest is focused on non-synonymous coding SNPs (nsSNPs) that are responsible of protein single point mutation. nsSNPs can be neutral or disease associated. It is known that the mutation of only one residue in a protein sequence can be related to a number of pathological conditions of dramatic social impact such as Alzheimer's, Parkinson's and Creutzfeldt-Jakob's diseases. The quality and completeness of presently available SNPs databases allows the application of machine learning techniques to predict the insurgence of human diseases due to single point protein mutation starting from the protein sequence.

RESULTS

In this paper, we develop a method based on support vector machines (SVMs) that starting from the protein sequence information can predict whether a new phenotype derived from a nsSNP can be related to a genetic disease in humans. Using a dataset of 21 185 single point mutations, 61% of which are disease-related, out of 3587 proteins, we show that our predictor can reach more than 74% accuracy in the specific task of predicting whether a single point mutation can be disease related or not. Our method, although based on less information, outperforms other web-available predictors implementing different approaches.

AVAILABILITY

A beta version of the web tool is available at http://gpcr.biocomp.unibo.it/cgi/predictors/PhD-SNP/PhD-SNP.cgi

A novel RGDS-analog inhibits angiogenesis in vitro and in vivo

In this study the anti-angiogenic action of a novel non-peptide RGDS-analog named RAM was tested in vitro and in vivo. RAM inhibited FGF-2-induced chemotaxis by 80% in an adhesion-independent way. Further, it induced HUVEC-apoptosis in collagen-seeded HUVEC, indicating that such pro-apoptotic effect was adhesion-independent. In vivo studies revealed that RAM inhibited FGF-2 induced angiogenesis by 60% in the mouse Matrigel-assay and in the chicken-egg chorion-allantoic membrane assay. Finally, RAM was markedly more stable in serum as compared to the template RGDS and after 24 h incubation in 100% serum was significantly more active than RGDS. Taken together these results show that RAM exerts anti-chemotactic and pro-apoptotic effects, by an unexpected adhesion-independent mechanism, as we have recently shown for the template RGDS molecule [Blood 103 (2004) 4180], and has in vivo relevant anti-angiogenic properties, with marked stability in serum; therefore, RAM represents a novel promising anti-angiogenic molecule.

Dynamics of the minimally frustrated helices determine the hierarchical folding of small helical proteins

In this paper we aim at determining the key residues of small helical proteins in order to build up reduced models of the folding dynamics. We start by arguing that the folding process can be dissected into concurrent fast and slow dynamics. The fast events are the quasiautonomous coil-to-helix transitions occurring in the minimally frustrated initiation sites of folding in the early stages of the process. The slow processes consist in the docking of the fluctuating helices formed in these critical sites. We show that a neural network devised to predict native secondary structures from sequence can be used to estimate the probabilities of formation of these helical traits as they are embedded in the protein. The resulting probabilities are shown to correlate well with the experimental helicities measured in the same isolated peptides. The relevance of this finding to the hierarchical character of folding is confirmed within the framework of a diffusion-collision-like mechanism. We demonstrate that thermodynamic and topological features of these critical helices allow accurate estimation of the folding times of five proteins that have been kinetically studied. This suggests that these critical helices determine the fundamental events of the whole folding process. A remarkable feature of our model is that not all of the native helices are eligible as critical helices, whereas the whole set of the native helices has been used so far in other reconstructions of the folding mechanism. This stresses that the minimally frustrated helices of these helical proteins comprise the minimal set of determinants of the folding process.

A neural network approach to evaluate fold recognition results

Fold recognition techniques assist the exploration of protein structures, and web-based servers are part of the standard set of tools used in the analysis of biochemical problems. Despite their success, current methods are only able to predict the correct fold in a relatively small number of cases. We propose an approach that improves the selection of correct folds from among the results of two methods implemented as web servers (SAMT99 and 3DPSSM). Our approach is based on the training of a system of neural networks with models generated by the servers and a set of associated characteristics such as the quality of the sequence-structure alignment, distribution of sequence features (sequence-conserved positions and apolar residues), and compactness of the resulting models. Our results show that it is possible to detect adequate folds to model 80% of the sequences with a high level of confidence. The improvements achieved by taking into account sequence characteristics open the door to future improvements by directly including such factors in the step of model generation. This approach has been implemented as an automatic system LIBELLULA, available as a public web server at http://www.pdg.cnb.uam.es/servers/libellula.html.

Characterisation of immunodominant protein encoded by the F1L gene of orf virus strains isolated in Italy

We analysed the molecular properties of the immunodominant protein of different orf virus strains isolated in Italy. The F1L encoding genes and the deduced amino acid sequences of all strains were determined and compared, and they showed several mutations. Structural analysis was carried out in order to assess the influence of amino acid variations on protein structure demonstrating a conservation of the secondary structure. Western blot analysis and immunogold electron microscopy showed that all orf virus strains were antigenically identical. The results of our study confirmed the immunogenicity of the F1L protein; furthermore, our data suggest a possible involvement of the protein in the virus cycle.

Progress in predicting inter-residue contacts of proteins with neural networks and correlated mutations

This article presents recent progress in predicting inter-residue contacts of proteins with a neural network-based method. Improvement over the results obtained at the previous CASP3 competition is attained by using as input to the network a complex code, which includes evolutionary information, sequence conservation, correlated mutations, and predicted secondary structures. The predictor was trained and cross-validated on a data set comprising the contact maps of 173 non-homologous proteins as computed from their well-resolved three-dimensional structures. The method could assign protein contacts with an average accuracy of 0.21 and with an improvement over a random predictor of a factor greater than 6, which is higher than that previously obtained with methods only based either on neural networks or on correlated mutations. Although far from being ideal, these scores are the highest reported so far for predicting protein contact maps. On 29 targets automatically predicted by the server (CORNET) the average accuracy is 0.14. The predictor is poorly performing on all alpha proteins, not represented in the training set. On all beta and mixed proteins (22 targets) the average accuracy is 0.16. This set comprises proteins of different complexity and different chain length, suggesting that the predictor is capable of generalization over a broad number of features.

Protein structure prediction and biomolecular recognition: from protein sequence to peptidomimetic design with the human beta3 integrin

Computational tools can bridge the gap between sequence and protein 3D structure based on the notion that information is to be retrieved from the databases and that knowledge-based methods can help in approaching a solution of the protein-folding problem. To this aim our group has implemented neural network-based predictors capable of performing with some success in different tasks, including predictions of the secondary structure of globular and membrane proteins, the topology of membrane proteins and porins and stable alpha-helical segments suited for protein design. Moreover we have developed methods for predicting contact maps in proteins and the probability of finding a cysteine in a disulfide bridge, tools which can contribute to the goal of predicting the 3D structure starting from the sequence (the so called ab initio prediction). All our predictors take advantage of evolution information derived from the structural alignments of homologous (evolutionary related) proteins and taken from the sequence and structure databases. When it is necessary to build models for proteins of unknown spatial structure, which have very little homology with other proteins of known structure, non-standard techniques need to be developed and the tools for protein structure predictions may help in protein modeling. The results of a recent simulation performed in our lab highlights the role of high performing computing technology and of tools of computational biology in protein modeling and peptidomimetic design.

Improved prediction of the number of residue contacts in proteins by recurrent neural networks

Knowing the number of residue contacts in a protein is crucial for deriving constraints useful in modeling protein folding, protein structure, and/or scoring remote homology searches. Here we use an ensemble of bi-directional recurrent neural network architectures and evolutionary information to improve the state-of-the-art in contact prediction using a large corpus of curated data. The ensemble is used to discriminate between two different states of residue contacts, characterized by a contact number higher or lower than the average value of the residue distribution. The ensemble achieves performances ranging from 70.1% to 73.1% depending on the radius adopted to discriminate contacts (6Ato 12A). These performances represent gains of 15% to 20% over the base line statistical predictors always assigning an aminoacid to the most numerous state, 3% to 7% better than any previous method. Combination of different radius predictors further improves the performance. SERVER: http://promoter.ics.uci.edu/BRNN-PRED/.

Prediction of contact maps with neural networks and correlated mutations

Contact maps of proteins are predicted with neural network-based methods, using as input codings of increasing complexity including evolutionary information, sequence conservation, correlated mutations and predicted secondary structures. Neural networks are trained on a data set comprising the contact maps of 173 non-homologous proteins as computed from their well resolved three-dimensional structures. Proteins are selected from the Protein Data Bank database provided that they align with at least 15 similar sequences in the corresponding families. The predictors are trained to learn the association rules between the covalent structure of each protein and its contact map with a standard back propagation algorithm and tested on the same protein set with a cross-validation procedure. Our results indicate that the method can assign protein contacts with an average accuracy of 0.21 and with an improvement over a random predictor of a factor >6, which is higher than that previously obtained with methods only based either on neural networks or on correlated mutations. Furthermore, filtering the network outputs with a procedure based on the residue coordination numbers, the accuracy of predictions increases up to 0.25 for all the proteins, with an 8-fold deviation from a random predictor. These scores are the highest reported so far for predicting protein contact maps.

Prediction of disulfide connectivity in proteins

MOTIVATION

A major problem in protein structure prediction is the correct location of disulfide bridges in cysteine-rich proteins. In protein-folding prediction, the location of disulfide bridges can strongly reduce the search in the conformational space. Therefore the correct prediction of the disulfide connectivity starting from the protein residue sequence may also help in predicting its 3D structure.

RESULTS

In this paper we equate the problem of predicting the disulfide connectivity in proteins to a problem of finding the graph matching with the maximum weight. The graph vertices are the residues of cysteine-forming disulfide bridges, and the weight edges are contact potentials. In order to solve this problem we develop and test different residue contact potentials. The best performing one, based on the Edmonds-Gabow algorithm and Monte-Carlo simulated annealing reaches an accuracy significantly higher than that obtained with a general mean force contact potential. Significantly, in the case of proteins with four disulfide bonds in the structure, the accuracy is 17 times higher than that of a random predictor. The method presented here can be used to locate putative disulfide bridges in protein-folding.

AVAILABILITY

The program is available upon request from the authors.

CONTACT

Casadio@alma.unibo.it; Piero@biocomp.unibo.it.

Model of interaction of the IL-1 receptor accessory protein IL-1RAcP with the IL-1beta/IL-1R(I) complex

A preliminary model has been calculated for the activating interaction of the interleukin 1 receptor (IL-1R) accessory protein IL-1RAcP with the ligand/receptor complex IL-1beta/IL-1R(I). First, IL-1RAcP was modeled on the crystal structure of IL-1R(I) bound to IL-1beta. Then, the IL-1RAcP model was docked using specific programs to the crystal structure of the IL-1beta/IL-1R(I) complex. Two types of models were predicted, with comparable probability. Experimental data obtained with the use of IL-1beta peptides and antibodies, and with mutated IL-1beta proteins, support the BACK model, in which IL-1RAcP establishes contacts with the back of IL-1R(I) wrapped around IL-1beta.

Prediction of the transmembrane regions of beta-barrel membrane proteins with a neural network-based predictor

A method based on neural networks is trained and tested on a nonredundant set of beta-barrel membrane proteins known at atomic resolution with a jackknife procedure. The method predicts the topography of transmembrane beta strands with residue accuracy as high as 78% when evolutionary information is used as input to the network. Of the transmembrane beta-strands included in the training set, 93% are correctly assigned. The predictor includes an algorithm of model optimization, based on dynamic programming, that correctly models eight out of the 11 proteins present in the training/testing set. In addition, protein topology is assigned on the basis of the location of the longest loops in the models. We propose this as a general method to fill the gap of the prediction of beta-barrel membrane proteins.

Structure-based computational study of the catalytic and inhibition mechanisms of urease

The viability of different mechanisms of catalysis and inhibition of the nickel-containing enzyme urease was explored using the available high-resolution structures of the enzyme isolated from Bacillus pasteurii in the native form and inhibited with several substrates. The structures and charge distribution of urea, its catalytic transition state, and three enzyme inhibitors were calculated using ab initio and density functional theory methods. The DOCK program suite was employed to determine families of structures of urease complexes characterized by docking energy scores indicative of their relative stability according to steric and electrostatic criteria. Adjustment of the parameters used by DOCK, in order to account for the presence of the metal ion in the active site, resulted in the calculation of best energy structures for the nickel-bound inhibitors beta-mercaptoethanol, acetohydroxamic acid, and diamidophosphoric acid. These calculated structures are in good agreement with the experimentally determined structures, and provide hints on the reactivity and mobility of the inhibitors in the active site. The same docking protocol was applied to the substrate urea and its catalytic transition state, in order to shed light onto the possible catalytic steps occurring at the binuclear nickel active site. These calculations suggest that the most viable pathway for urea hydrolysis involve a nucleophilic attack by the bridging, and not the terminal, nickel-bound hydroxide onto a urea molecule, with active site residues playing important roles in orienting and activating the substrate, and stabilizing the catalytic transition state.

Prediction of the number of residue contacts in proteins

Knowing the number of residue contacts in a protein is crucial for deriving constraints useful in modeling protein folding and/or scoring remote homology search. Here we focus on the prediction of residue contacts and show that this figure can be predicted with a neural network based method. The accuracy of the prediction is 12 percentage points higher than that of a simple statistical method. The neural network is used to discriminate between two different states of residue contacts, characterized by a contact number higher or lower than the average value of the residue distribution. When evolutionary information is taken into account, our method correctly predicts 69% of the residue states in the data base and it adds to the prediction of residue solvent accessibility. The predictor is available at htpp://www.biocomp.unibo.it

RCNPRED: prediction of the residue co-ordination numbers in proteins

The RCNPRED server implements a neural network-based method to predict the co-ordination numbers of residues starting from the protein sequence. Using evolutionary information as input, RCNPRED predicts the residue states of the proteins in the database with 69% accuracy and scores 12 percentage points higher than a simple statistical method. Moreover the server implements a neural network to predict the relative solvent accessibility of each residue. A protein sequence can be directly submitted to RCNPRED: residue co-ordination numbers and solvent accessibility for each chain are returned via e-mail.

AVAILABILITY

Freely available to non-commercial users at http://prion.biocomp.unibo.it/rcnpred.html.

Predictions of protein segments with the same aminoacid sequence and different secondary structure: a benchmark for predictive methods

The most stringent test for predictive methods of protein secondary structure is whether identical short sequences that are known to be present with different conformations in different proteins known at atomic resolution can be correctly discriminated. In this study, we show that the prediction efficiency of this type of segments in unrelated proteins reaches an average accuracy per residue ranging from about 72 to 75% (depending on the alignment method used to generate the input sequence profile) only when methods of the third generation are used. A comparison of different methods based on segment statistics (2nd generation methods) and/or including also evolutionary information (3rd generation methods) indicate that the discrimination of the different conformations of identical segments is dependent on the method used for the prediction. Accuracy is similar when methods similarly performing on the secondary structure prediction are tested. When evolutionary information is taken into account as compared to single sequence input, the number of correctly discriminated pairs is increased twofold. The results also highlight the predictive capability of neural networks for identical segments whose conformation differs in different proteins.

Tryptophanyl fluorescence lifetime distribution of hyperthermophilic beta-glycosidase from molecular dynamics simulation: a comparison with the experimental data

A molecular dynamics simulation approach has been utilized to understand the unusual fluorescence emission decay observed for beta-glycosidase from the hyperthermophilic bacterium Solfolobus sulfotaricus (Sbeta gly), a tetrameric enzyme containing 17 tryptophanyl residues for each subunit. The tryptophanyl emission decay of Sbeta gly results from a bimodal distribution of fluorescence lifetimes with a short-lived component centered at 2.5 ns and a long-lived one at 7.4 ns (Bismuto E, Nucci R, Rossi M, Irace G, 1999, Proteins 27:71-79). From the examination of the trajectories of the side chains capable of causing intramolecular quenching for each tryptophan microenvironment and using a modified Stern-Volmer model for the emission quenching processes, we calculated the fluorescence lifetime for each tryptophanyl residue of Sbeta gly at two different temperatures, i.e., 300 and 365 K. The highest temperature was chosen because in this condition Sbeta gly evidences a maximum in its catalytic activity and is stable for a very long time. The calculated lifetime distributions overlap those experimentally determined. Moreover, the majority of trytptophanyl residues having longer lifetimes correspond to those originally identified by inspection of the crystallographic structure. The tryptophanyl lifetimes appear to be a complex function of several variables, such as microenvironment viscosity, solvent accessibility, the chemical structure of quencher side chains, and side-chain dynamics. The lifetime calculation by MD simulation can be used to validate a predicted structure by comparing the theoretical data with the experimental fluorescence decay results.

The effect of tryptophanyl substitution on folding and structure of myoglobin

Mammalian myoglobins contain two tryptophanyl residues at the invariant positions 7 (A-5) and 14 (A-12) in the N-terminal region (A helix) of the protein molecule. The crucial role of tryptophanyl residues has been investigated by site-directed mutagenesis and molecular dynamics simulation. The apomyoglobin mutants with a double W-->F substitution were found to be not correctly folded and therefore not expressed as holoprotein. The introduction of a tyrosyl residue at position 7, that is, W7YW14F, resulted in the expression of a correctly folded myoglobin. Not correctly folded apomyoglobins were found with the following mutants: W7FW14Y, W7EW14F, W7FW14E, W7KW14F, W7FW14K. Moreover, in all these cases, very low levels of expression were observed. The acid-induced denaturation curves of wild-type and folded mutant W7YW14F, obtained following the fluorescence variation of the extrinsic fluorophore 1-anilino-8-naphthalenesulfonate, revealed that the stability of the native state of mutant apoprotein is decreased, thus indicating that the replacement W-->Y in position 7 is able to restore a correct folding but not the same stability. Molecular dynamics simulation indicated that both tryptophans are involved in forming favorable, specific tertiary interactions in the native apomyoglobin structure. The lack of some of these interactions caused by tryptophanyl replacement affects the overall protein structure and may provide an explanation for the observed stability decrease. In the case of the double W-->F substitution, the simulated structure shows conclusively the domain formed by helices A, G and H to be not correctly folded. This effect is attenuated if at least one of the two residues is conserved or a tyrosyl residue replaces W7.

Ligand-induced conformational changes in tissue transglutaminase: Monte Carlo analysis of small-angle scattering data

Small-angle neutron and x-ray scattering experiments have been performed on type 2 tissular transglutaminase to characterize the conformational changes that bring about Ca(2+) activation and guanosine triphosphate (GTP) inhibition. The native and a proteolyzed form of the enzyme, in the presence and in the absence of the two effectors, were considered. To describe the shape of transglutaminase in the different conformations, a Monte Carlo method for calculating small-angle neutron scattering profiles was developed by taking into account the computer-designed structure of the native transglutaminase, the results of the Guinier analysis, and the essential role played by the solvent-exposed peptide loop for the conformational changes of the protein after activation. Although the range of the neutron scattering data is rather limited, by using the Monte Carlo analysis, and because the structure of the native protein is available, the distribution of the protein conformations after ligand interaction was obtained. Calcium activation promotes a rotation of the C-terminal with respect to the N-terminal domain around the solvent-exposed peptide loop that connects the two regions. The psi angle between the longest axes of the two pairs of domains is found to be above 50 degrees, larger than the psi value of 35 degrees calculated for the native transglutaminase. On the other hand, the addition of GTP makes possible conformations characterized by psi angles lower than 34 degrees. These results are in good agreement with the proposed enzyme activity regulation: in the presence of GTP, the catalytic site is shielded by the more compact protein structure, while the conformational changes induced by Ca(2+) make the active site accessible to the substrate.

A data base of minimally frustrated alpha helical segments extracted from proteins according to an entropy criterion

A data base of minimally frustrated alpha helical segments is defined by filtering a set comprising 822 non redundant proteins, which contain 4783 alpha helical structures. The data base definition is performed using a neural network-based alpha helix predictor, whose outputs are rated according to an entropy criterion. A comparison with the presently available experimental results indicates that a subset of the data base contains the initiation sites of protein folding experimentally detected and also protein fragments which fold into stable isolated alpha helices. This suggests the usage of the data base (and/or of the predictor) to highlight patterns which govern the stability of alpha helices in proteins and the helical behavior of isolated protein fragments.

Neural networks predict protein folding and structure: artificial intelligence faces biomolecular complexity

In the genomic era DNA sequencing is increasing our knowledge of the molecular structure of genetic codes from bacteria to man at a hyperbolic rate. Billions of nucleotides and millions of aminoacids are already filling the electronic files of the data bases presently available, which contain a tremendous amount of information on the most biologically relevant macromolecules, such as DNA, RNA and proteins. The most urgent problem originates from the need to single out the relevant information amidst a wealth of general features. Intelligent tools are therefore needed to optimise the search. Data mining for sequence analysis in biotechnology has been substantially aided by the development of new powerful methods borrowed from the machine learning approach. In this paper we discuss the application of artificial feedforward neural networks to deal with some fundamental problems tied with the folding process and the structure-function relationship in proteins.

Role of evolutionary information in predicting the disulfide-bonding state of cysteine in proteins

A neural network-based predictor is trained to distinguish the bonding states of cysteine in proteins starting from the residue chain. Training is performed by using 2,452 cysteine-containing segments extracted from 641 nonhomologous proteins of well-resolved three-dimensional structure. After a cross-validation procedure, efficiency of the prediction scores were as high as 72% when the predictor is trained by using protein single sequences. The addition of evolutionary information in the form of multiple sequence alignment and a jury of neural networks increases the prediction efficiency up to 81%. Assessment of the goodness of the prediction with a reliability index indicates that more than 60% of the predictions have an accuracy level greater than 90%. A comparison with a statistical method previously described and tested on the same database shows that the neural network-based predictor is performing with the highest efficiency. Proteins 1999;36:340-346.

Quantum mechanical analysis of oxygenated and deoxygenated states of hemocyanin: theoretical clues for a plausible allosteric model of oxygen binding

In this work with ab initio computations, we describe relevant interactions between protein active sites and ligands, using as a test case arthropod hemocyanins. A computational analysis of models corresponding to the oxygenated and deoxygenated forms of the hemocyanin active site is performed using the Density Functional Theory approach. We characterize the electron density distribution of the binding site with and without bound oxygen in relation to the geometry, which stems out of the crystals of three hemocyanin proteins, namely the oxygenated form from the horseshoe crab Limulus polyphemus, and the deoxygenated forms, respectively, from the same source and from another arthropod, the spiny lobster Panulirus interruplus. Comparison of the three available crystals indicate structural differences at the oxygen binding site, which cannot be explained only by the presence and absence of the oxygen ligand, since the geometry of the ligand site of the deoxygenated Panulirus hemocyanin is rather similar to that of the oxygenated Limulus protein. This finding was interpreted in the frame of a mechanism of allosteric regulation for oxygen binding. However, the cooperative mechanism, which is experimentally well documented, is only partially supported by crystallographic data, since no oxygenated crystal of Panulirus hemocyanin is presently available. We address the following question: is the local ligand geometry responsible for the difference of the dicopper distance observed in the two deoxygenated forms of hemocyanin or is it necessary to advocate the allosteric regulation of the active site conformations in order to reconcile the different crystal forms? We find that the difference of the dicopper distance between the two deoxygenated hemocyanins is not due to the small differences of ligand geometry found in the crystals and conclude that it must be therefore stabilized by the whole protein tertiary structure.

The structural basis for the regulation of tissue transglutaminase by calcium ions

The role of calcium ions in the regulation of tissue transglutaminase is investigated by experimental approaches and computer modeling. A three-dimensional model of the transglutaminase is computed by homology building on crystallized human factor XIII and is used to interpret structural and functional results. The molecule is a prolate ellipsoid (6.2 x 4.2 x 11 nm) and comprises four domains, assembled pairwise into N-terminal and C-terminal regions. The active site is hidden in a cleft between these regions and is inaccessible to macromolecular substrates in the calcium-free form. Protein dynamics simulation indicates that these regions move apart upon addition of calcium ions, revealing the active site for catalysis. The protein dimensions are consistent with results obtained with small-angle neutron and X-ray scattering. The gyration radius of the protein (3 nm) increases in the presence of calcium ions (3.9 nm), but it is virtually unaffected in the presence of GTP, suggesting that only calcium ions can promote major structural changes in the native protein. Proteolysis of an exposed loop connecting the N-terminal and C-terminal regions is linearly correlated with enzyme inactivation and prevents the calcium-induced conformational changes.