MAGOS: multiple alignment and modelling server.

UNLABELLED: MAGOS is a web server allowing automated protein modelling coupled to the creation of a hierarchical and annotated multiple alignment of complete sequences. MAGOS is designed for an interactive approach of structural information within the framework of the evolutionary relevance of mined and predicted sequence information. AVAILABILITY: The web server is freely available at http://pig-pbil.ibcp.fr/magos.

A common mechanism for ATP hydrolysis in ABC transporter and helicase superfamilies.

ABC (ATP-binding cassette) transporters and helicases are large superfamilies of seemingly unrelated proteins, whose functions depend on the energy provided by ATP hydrolysis. Comparison of the 3D structures of their nucleotide-binding domains reveals that, besides two well-characterized ATP-binding signatures, the folds of their nucleotide-binding sites are similar. Furthermore, there are striking similarities in the positioning of residues thought to be important for ATP binding or hydrolysis. Interestingly, structures have recently been obtained for two ABC proteins that are not involved in transport activities, but that have a function related to DNA modification. These ABC proteins, which contain a nucleotide-binding site akin to those of typical ABC transporters, might constitute the missing link between the two superfamilies.

ANTHEPROT: an integrated protein sequence analysis software with client/server capabilities.

Programs devoted to the analysis of protein sequences exist either as stand-alone programs or as Web servers. However, stand-alone programs can hardly accommodate for the analysis that involves comparisons on databanks, which require regular updates. Moreover, Web servers cannot be as efficient as stand-alone programs when dealing with real-time graphic display. We describe here a stand-alone software program called ANTHEPROT, which is intended to perform protein sequence analysis with a high integration level and clients/server capabilities. It is an interactive program with a graphical user interface that allows handling of protein sequence and data in a very interactive and convenient manner. It provides many methods and tools, which are integrated into a graphical user interface. ANTHEPROT is available for Windows-based systems. It is able to connect to a Web server in order to perform large-scale sequence comparison on up-to-date databanks. ANTHEPROT is freely available to academic users and may be downloaded at http://pbil.ibcp.fr/ANTHEPROT.

Identification of related proteins with weak sequence identity using secondary structure information.

Molecular modeling of proteins is confronted with the problem of finding homologous proteins, especially when few identities remain after the process of molecular evolution. Using even the most recent methods based on sequence identity detection, structural relationships are still difficult to establish with high reliability. As protein structures are more conserved than sequences, we investigated the possibility of using protein secondary structure comparison (observed or predicted structures) to discriminate between related and unrelated proteins sequences in the range of 10%-30% sequence identity. Pairwise comparison of secondary structures have been measured using the structural overlap (Sov) parameter. In this article, we show that if the secondary structures likeness is >50%, most of the pairs are structurally related. Taking into account the secondary structures of proteins that have been detected by BLAST, FASTA, or SSEARCH in the noisy region (with high E: value), we show that distantly related protein sequences (even with <20% identity) can be still identified. This strategy can be used to identify three-dimensional templates in homology modeling by finding unexpected related proteins and to select proteins for experimental investigation in a structural genomic approach, as well as for genome annotation.

Involvement of electrostatic interactions in the mechanism of peptide folding induced by sodium dodecyl sulfate binding.

Sodium dodecyl sulfate (SDS) has consistently been shown to induce secondary structure, particularly alpha-helices, in polypeptides, and is commonly used to model membrane and other hydrophobic environments. However, the precise mechanism by which SDS induces these conformational changes remains unclear. To examine the role of electrostatic interactions in this mechanism, we have designed two hydrophilic, charged amphipathic alpha-helical peptides, one basic (QAPAYKKAAKKLAES) and the other acidic (QAPAYEEAAEELAKS), and their structures were studied by CD and NMR. The design of the peptides is based on the sequence of the segment of residues 56-70 of human platelet factor 4 [PF4(56-70), QAPLYKKIIKKLLES]. Both peptides were unstructured in water, and in the presence of neutral, zwitterionic, or cationic detergents. However, in SDS at neutral pH, the basic peptide folded into an alpha-helix. By contrast, the pH needed to be lowered to 1.8 before alpha-helix formation was observed for the acidic peptide. Strong, attractive electrostatic interactions, between the anionic groups of SDS and the cationic groups of the lysines, appeared to be necessary to initiate the folding of the basic peptide. NMR analysis showed that the basic peptide was fully embedded in SDS-peptide micelles, and that its three-dimensional alpha-helical structure could be superimposed on that of the native structure of PF4(56-70). These results enabled us to propose a working model of the basic peptide-SDS complex, and a mechanism for SDS-induced alpha-helical folding. This study demonstrates that, while the folding of peptides is mostly driven by hydrophobic effects, electrostatic interactions play a significant role in the formation and the stabilization of SDS-induced structure.

MPSA: integrated system for multiple protein sequence analysis with client/server capabilities.

UNLABELLED: MPSA is a stand-alone software intended to protein sequence analysis with a high integration level and Web clients/server capabilities. It provides many methods and tools, which are integrated into an interactive graphical user interface. It is available for most Unix/Linux and non-Unix systems. MPSA is able to connect to a Web server (e.g. http://pbil.ibcp.fr/NPSA) in order to perform large-scale sequence comparison on up-to-date databanks. AVAILABILITY: Free to academic http://www.ibcp.fr/mpsa/ CONTACT: c.blanchet@ibcp.fr

Molecular model, calcium sensitivity, and disease specificity of a conformational thyroperoxidase B-cell epitope.

While studying the humoral mechanisms involved in thyroid autoimmunity, we located a B-cell autoepitope in the extracellular C-terminal region of human thyroperoxidase. Structural modeling showed that this region encompasses both a Sushi-like and an epidermal growth factor-like domain, the flexible arrangement of which was putatively stabilized by calcium. The recombinant peptide was found to contain the previously identified conformational thyroperoxidase autoepitope. The occurrence of a calcium-induced conformational change was confirmed using a recombinant peptide monoclonal antibody, the decrease of which in binding to calcium-saturated thyroperoxidase was reversed by a chelating agent. The disease specificity of recombinant peptide, which was more frequently recognized by Hashimoto's than by Graves' patients, adds to its potential value as a diagnostic and preventive tool in the context of B-cell autoimmunity.

Improved performance in protein secondary structure prediction by inhomogeneous score combination.

MOTIVATION: In many fields of pattern recognition, combination has proved efficient to increase the generalization performance of individual prediction methods. Numerous systems have been developed for protein secondary structure prediction, based on different principles. Finding better ensemble methods for this task may thus become crucial. Furthermore, efforts need to be made to help the biologist in the post-processing of the outputs. RESULTS: An ensemble method has been designed to post-process the outputs of discriminant models, in order to obtain an improvement in prediction accuracy while generating class posterior probability estimates. Experimental results establish that it can increase the recognition rate of protein secondary structure prediction methods that provide inhomogeneous scores, even though their individual prediction successes are largely different. This combination thus constitutes a help for the biologist, who can use it confidently on top of any set of prediction methods. Moreover, the resulting estimates can be used in various ways, for instance to determine which areas in the sequence are predicted with a given level of reliability. AVAILABILITY: The prediction is freely available over the Internet on the Network Protein Sequence Analysis (NPS@) WWW server at http://pbil.ibcp.fr/NPSA/npsa_server.ht ml. The source code of the combiner can be obtained on request for academic use.

The prolyl aminopeptidase from Lactobacillus delbrueckii subsp. bulgaricus belongs to the alpha/beta hydrolase fold family.

Prolyl aminopeptidase (PepIP) of Lactobacillus delbrueckii subsp. bulgaricus displays the Gly-x-Ser-x-Gly-Gly consensus motif surrounding the catalytic serine of the prolyl oligopeptidases family. Sequence comparison revealed that this motif and two other domains appear well conserved among bacterial PepIPs and members of the alpha/beta hydrolase fold family. Secondary structural predictions of PepIP were performed from amino acid sequence and corroborated by circular dichroism analysis. These predictions well matched the core structure of alpha/beta hydrolases organised in eight beta-sheets connected by alpha-helices. We obtained 26 mutants of PepIP by chemical or site-directed mutagenesis. Most substitutions associated with stable and inactive mutant proteins were mainly located in the three conserved boxes (including the catalytic serine motif). Taken together, our results strongly suggest that PepIP belongs to the alpha/beta hydrolase fold family and that Ser107, Asp246 and His273 constitute the catalytic triad of the enzyme.

Molecular features of the collagen V heparin binding site.

A heparin binding region is known to be present within the triple helical part of the alpha1(V) chain. Here we show that a recombinant alpha1(V) fragment (Ile824 to Pro950), referred to as HepV, is sufficient for heparin binding at physiological ionic strength. Both native individual alpha1(V) chains and HepV are eluted at identical NaCl concentrations (0.35 M) from a heparin-Sepharose column, and this binding can be inhibited specifically by the addition of free heparin or heparan sulfate. In contrast, a shorter 23-residue synthetic peptide, containing the putative heparin binding site in HepV, fails to bind heparin. Interestingly, HepV promotes cell attachment, and HepV-mediated adhesion is inhibited specifically by heparin or heparan sulfate, indicating that this region might behave as an adhesive binding site. The same site is equally functional on triple helical molecules as shown by heparin-gold labeling. However, the affinities for heparin of each of the collagen V molecular forms tested are different and increase with the number of alpha1(V) chains incorporated in the molecules. Molecular modeling of a sequence encompassing the putative HepV binding sequence region shows that all of the basic residues cluster on one side of the helical face. A highly positively charged ring around the molecule is thus particularly evident for the alpha1(V) homotrimer. This could strengthen its interaction with the anionic heparin molecules. We propose that a single heparin binding site is involved in heparin-related glycosaminoglycans-collagen V interactions, but the different affinities observed likely modulate cell and matrix interactions between collagen V and heparan sulfate proteoglycans in tissues.

Three-dimensional structure of the DNA-binding domain of the fructose repressor from Escherichia coli by 1H and 15N NMR.

FruR is an Escherichia coli transcriptional regulator that belongs to the LacI DNA-binding protein family. By using 1H and 15N NMR spectroscopy, we have determined the three-dimensional solution structure of the FruR N-terminal DNA-binding domain consisting of 57 amino acid residues. A total of 809 NMR-derived distances and 54 dihedral angle constraints have been used for molecular modelling with the X-PLOR program. The resulting set of calculated structures presents an average root-mean-square deviation of 0.37 A at the main-chain level for the first 47 residues. This highly defined N-terminal part of the structure reveals a similar topology for the three alpha-helices when compared to the 3D structures of LacI and PurR counterparts. The most striking difference lies in the connection between helix II and helix III, in which three additional residues are present in FruR. This connecting segment is well structured and contains a type III turn. Apart from hydrophobic interactions of non-polar residues with the core of the domain, this connecting segment is stabilised by several hydrogen bonds and by the aromatic ring stacking between Tyr19 of helix II and Tyr28 of the turn. The region containing the putative "hinge helix" (helix IV), that has been described in PurR-DNA complex to make specific base contacts in the minor groove of DNA, is unfolded. Examination of hydrogen bonds highlights the importance of homologous residues that seem to be conserved for their ability to fulfill helix N and C-capping roles in the LacI repressor family.

Protein structure prediction. Implications for the biologist.

Recent improvements in the prediction of protein secondary structure are described, particularly those methods using the information contained into multiple alignments. In this respect, the prediction accuracy has been checked and methods that take into account multiple alignments are 70% correct for a three-state description of secondary structure. This quality is obtained by a 'leave-one out' procedure on a reference database of proteins sharing less than 25% identity. Biological applications such as 'protein domain design' and structural phylogeny are given. The biologist's point of view is also considered and joint predictions are encouraged in order to derive an amino acid based accuracy. All the tools described in this paper are available for biologists on the Web (http/www.ibcp.fr/predict.html).

Trimeric assembly and three-dimensional structure model of the FACIT collagen COL1-NC1 junction from CD and NMR analysis.

The 3D structure of the COL1-NC1 junction of FACIT type XIV collagen was investigated using GYCDPSSCAG and (GPP*)3GYCDPSSCAG synthetic peptides, circular dichroism, and NMR. At -20 degrees C and under air oxidation catalyzed by Cu2+, the peptide (GPP*)3GYCDPSSCAG is able to self-associated with high yield into a stable triple disulfide bonded trimer. The presence of a triple helical conformation was confirmed by circular dichroism. The analysis of the trimer by 2D NMR provided a set of distance constraints for the noncollagenous part. Molecular models for the 3D structure of COL1-NC1 junction were calculated, using the NMR distance constraints in combination with the 3D structural data recently established by X-ray crystallography [Bella, J., Eaton, M., Brodsky, B., & Berman, H. M. (1994) Science 266, 75-81] for a collagenous triple helix. From the eight theoretically possible arrangements for the three interchain disulfide bonds, only two close disulfide conformers are compatible with the experimental data. The main feature of the trimer structure is the asymmetry of the molecule due to the disulfide bond pattern that induces a particular folding of one chain. This chain forms a turn-like structure locked by two disulfide bonds with the two other chains. The turn-like folding is close to that observed for the cyclized oxidized monomeric peptide. This is the first report of the 3D structure model for a junction between a collagenous triple helical domain and a noncollagenous domain.

Definition of a consensus DNA-binding site for the Escherichia coli pleiotropic regulatory protein, FruR.

The FruR regulator of Escherichia coli controls the initiation of transcription of several operons encoding a variety of proteins involved in carbon and energy metabolism. The sequence determinants of the FruR-binding site were analysed by using 6x His-tagged FruR and a series of double-stranded randomized oligonucleotides. FruR consensus binding sites were selected and characterized by several consecutive rounds of the polymerase chain reaction-assisted binding-site selection method (BSS) using nitrocellulose-immobilized DNA-binding protein. FruR was demonstrated to require, for binding, an 8 bp left half-site motif and a 3 bp conserved right half-site with the following sequence: 5'-GNNGAATC/GNT-3'. In this sequence, the left half-site AATC/ consensus tetranucleotide is a typical motif of the DNA-binding site of the regulators of the GalR-Lacl family. On the other hand, the high degree of degeneracy found in the right half-site of this palindrome-like structure indicated that FruR, which is a tetramer in solution, interacts asymmetrically with the two half-sites of its operator. However, potentially FruR-target sites showing a high degree of symmetry were detected in 13 genes/operons. Among these, we have focused our interest on the pfkA gene, encoding phosphofructo-kinase-1, which is negatively regulated by FruR.

Cloning and characterization of the Bacillus subtilis prkA gene encoding a novel serine protein kinase.

We have cloned and sequenced a 3574-bp Bacillus subtilis (Bs) DNA fragment located between the nrdA and citB genes at about 169 degrees on the chromosome. An Escherichia coli strain, LBG1605, carrying a mutated ptsH gene (encoding HPr (His-containing protein) of the bacterial phosphotransferase system (PTS)) and complemented for PTS activity with the ptsH of Staphylococcus carnosus, exhibited reduced mannitol fermentation activity when transformed with a plasmid bearing this 3574-bp Bs fragment. This fragment contained an incomplete and two complete open reading frames (ORFs). The product of the first complete ORF, a protein composed of 235 amino acids (aa) (25038 Da), was found to be responsible for the observed reduced mannitol fermentation. The 3' part of this 705-bp second ORF and the 428-bp incomplete first ORF encode aa sequences exhibiting almost 40% sequence identify. However, the function of these two proteins remains unknown. The third ORF, the 1893-bp prkA gene, encodes a protein (PrkA) of 72889 Da. PrkA possesses the A-motif of nucleotide-binding proteins and exhibits distant homology to eukaryotic protein kinases. Several of the essential aa in the loops known to form the active site of cyclic adenosine 3',5'-monophosphate (cAMP)-dependent protein kinase appeared to be conserved in PrkA. After expression of prkA and purification of PrkA, we could demonstrate that PrkA can indeed phosphorylate a Bs 60-kDa protein at a Ser residue.

SOPMA: significant improvements in protein secondary structure prediction by consensus prediction from multiple alignments.

Recently a new method called the self-optimized prediction method (SOPM) has been described to improve the success rate in the prediction of the secondary structure of proteins. In this paper we report improvements brought about by predicting all the sequences of a set of aligned proteins belonging to the same family. This improved SOPM method (SOPMA) correctly predicts 69.5% of amino acids for a three-state description of the secondary structure (alpha-helix, beta-sheet and coil) in a whole database containing 126 chains of non-homologous (less than 25% identity) proteins. Joint prediction with SOPMA and a neural networks method (PHD) correctly predicts 82.2% of residues for 74% of co-predicted amino acids. Predictions are available by Email to deleage@ibcp.fr or on a Web page (http:@www.ibcp.fr/predict.html).

ANTHEPROT 2.0: a three-dimensional module fully coupled with protein sequence analysis methods.

ANTHEPROT is a fully interactive graphics program devoted to the analysis of the sequences and structures of proteins. This program, originally developed to facilitate the protein sequence analysis coupled with multiple alignments and predicted secondary structures of proteins, now comprises a powerful 3D module to display and handle macromolecular structures. All the methods that were previously integrated into ANTHEPROT are now directly coupled with a 3D window that provides the user all the classic features of a molecular modeling package. Indeed, it allows real-time rotation and translation of 3D structures with many kinds of models in depth-cueing mode (space filling, backbone, wire models, main chain, and ribbons), selections (atom type, residue type, segments, and chain), color-coding systems (amino acid properties, predicted or observed secondary structures, temperature B factor, and subunits), geometric calculations (Ramachandran plot, distances, and angles), and fitting molecules. Stereo views are possible as well as HPGL standard files. A module specifically devoted to the determination of 3D structures using nuclear magnetic resonance is also available. This major release of our program for IBM rs6000 workstations is available by anonymous ftp to ibcp.fr for academic institutions.

Dimerization kinetics of HIV-1 and HIV-2 reverse transcriptase: a two step process.

The dimerization processes of the human immunodeficiency virus (HIV) types 1 and 2 reverse transcriptase (RTs) from their subunits have been investigated using a number of complementary approaches (fluorescence spectroscopy, size exclusion-HPLC and polymerase activity assay). The formation of the native heterodimeric form of HIV-1 and HIV-2 RT occurs in a two step process. The first step is a concentration-dependent association of the two subunits (p66 and p51) to give a heterodimeric intermediate, which slowly isomerizes to the "mature" heterodimeric form of the enzyme. For both RTs, the first step behaves as a second order reaction with similar association rate constants (in the range of 2 x 10(4) to 4 x 10(4) M-1 s-1). This initial dimerization results in a 25% quenching of the intrinsic fluorescence and a 30% decrease in the accessibility of the tryptophan hydrophobic cluster to solvent as revealed by iodide quenching experiments and by monitoring the binding of 1-anilino-8-naphthalenesulphonate. The formation of the intermediate-RT form appears to involve hydrophobic regions of the subunits containing tryptophan residues. This intermediate form is devoid of polymerase activity, but is able to bind primer/template with high affinity. The final stage of the mature RT-heterodimer formation occurs in a slow first order reaction, which is 12-fold faster for HIV-2 (1.2 h-1) than HIV-1 RT (0.1 h-1). At micromolar concentrations, this slow isomerization constitutes the rate limiting step of the RT maturation and the structural change involved appears to be partly associated with the catalytic site, as shown using fluorescent labelled primer/template. On the basis of both the presently available X-ray structure of the HIV-1 RT and the predicted structure of HIV-2 RT, the thumb subdomain of the p51 subunit seems to be involved in this maturation step, which is probably the interaction of this domain with the RNAse H domain of the large subunit. The placement of the fingers subdomain of p51 in the palm subdomain of the p66 subunit may also be associated with formation of mature heterodimeric RTs.

Purification and partial amino acid sequence of fuctinin, an endogenous inhibitor of fucosyltransferase activities.

A powerful endogenous protein inhibitor of fucosyltransferase activities, called fuctinin, was purified to homogeneity from rat small-intestinal mucosa. The purification scheme involved DEAE-cellulose ion-exchange chromatography, ammonium sulfate fractionation, hexyl-agarose hydrophobic chromatography and size-exclusion HPLC. Active native fuctinin has an isoelectric point of 4.55 and apparent molecular mass approximately 66 kDa, whereas a single protein band with a molecular mass of approximately 24 kDa was obtained by denaturing polyacrylamide gel electrophoresis, suggesting that fuctinin is an oligomeric protein. Two-dimensional polyacrylamide gel electrophoresis displayed eight spots in this single band. Comparisons of the N-terminal amino acid sequences of each spot support the idea of the existence of three related polypeptides and suggest a proteolytic N-terminal cleavage despite the use of an efficient protease inhibitor throughout the purification. In spite of the presence of an N-glycosylation site, fuctinin is not glycosylated. One of the three polypeptides, peptide 3, possesses two consensus sequences for phosphorylation and a consensus sequence for myristoylation. The sequences of functinin-related peptides, especially peptide 3, exhibit high similarity to the N-terminal domain of the Set protein and a putative human leukocyte antigen-associated protein. The possible implications of these results are discussed.