Hydrophobic core in domains of immunoglobulin-like fold.

This work analyzes proteins which contain an immunoglobulin fold, focusing on their hydrophobic core structure. The "fuzzy oil drop" model was used to measure the regularity of hydrophobicity distribution in globular domains belonging to proteins which exhibit the above-mentioned fold. Light-chain IgG domains are found to frequently contain regular hydrophobic cores, unlike the corresponding heavy-chain domains. Enzymes and DNA binding proteins present in the data-set are found to exhibit poor accordance with the hydrophobic core model.

Frog: a FRee Online druG 3D conformation generator.

In silico screening methods based on the 3D structures of the ligands or of the proteins have become an essential tool to facilitate the drug discovery process. To achieve such process, the 3D structures of the small chemical compounds have to be generated. In addition, for ligand-based screening computations or hierarchical structure-based screening projects involving a rigid-body docking step, it is necessary to generate multi-conformer 3D models for each input ligand to increase the efficiency of the search. However, most academic or commercial compound collections are delivered in 1D SMILES (simplified molecular input line entry system) format or in 2D SDF (structure data file), highlighting the need for free 1D/2D to 3D structure generators. Frog is an on-line service aimed at generating 3D conformations for drug-like compounds starting from their 1D or 2D descriptions. Given the atomic constitution of the molecules and connectivity information, Frog can identify the different unambiguous isomers corresponding to each compound, and generate single or multiple low-to-medium energy 3D conformations, using an assembly process that does not presently consider ring flexibility. Tests show that Frog is able to generate bioactive conformations close to those observed in crystallographic complexes. Frog can be accessed at http://bioserv.rpbs.jussieu.fr/Frog.html.

RPBS: a web resource for structural bioinformatics.

RPBS (Ressource Parisienne en Bioinformatique Structurale) is a resource dedicated primarily to structural bioinformatics. It is the result of a joint effort by several teams to set up an interface that offers original and powerful methods in the field. As an illustration, we focus here on three such methods uniquely available at RPBS: AUTOMAT for sequence databank scanning, YAKUSA for structure databank scanning and WLOOP for homology loop modelling. The RPBS server can be accessed at http://bioserv.rpbs.jussieu.fr/ and the specific services at http://bioserv.rpbs.jussieu.fr/SpecificServices.html.

Active immunization against murine TNFalpha peptides in mice: generation of endogenous antibodies cross-reacting with the native cytokine and in vivo protection.

New lines of treatment targeting cytokines have been successfully developed recently and are now widely used in therapy. They are based on passive administration of cytokine inhibitors either soluble receptors or mAbs and the major example is TNFalpha in rheumatoid arthritis (RA). Since a few years, our group has developed a novel alternative approach targeting cytokines by using active immunization against biologically inactive but immunogenic cytokine derivatives. In the present work, we present a new aspect of this research, based on immunization against specific cytokine peptides chosen by molecular modelling. We could elicit a significant humoral response against four TNFalpha peptides by active immunization, and show that the Abs generated cross-reacted with the native cytokine with good titers as determined by ELISA. Interestingly, during coimmunization experiments with couples of peptides, one showed a clear immunodominant effect over the other. Overall, we could not show the neutralization of TNFalpha biological activity in vitro by the immunized sera, but it seems that it is not a prerequisite to observe clinical efficacy. Indeed, using the LPS/galactosamine-induced shock, we could demonstrate that one of the four peptides tested conferred a clinical protection. These results validate the feasibility and efficacy of active immunization against cytokine peptides, and confirm that active immunization against cytokines could represent in the future an alternative to passive immunization in many diseases.

Distribution of tightened end fragments of globular proteins statistically matches that of topohydrophobic positions: towards an efficient punctuation of protein folding?

Using a set of 372 proteins representative of a variety of 56 distinct globular folds, a statistical correlation was observed between two recently revealed features of protein structures: tightened end fragments or 'closed loops', i.e. sequence fragments that are able in three-dimensional (3D) space to nearly close their ends (a current parameter of polymer physics), and 'topohydrophobic positions', i.e. positions always occupied in 3D space by strong hydrophobic amino acids for all members of a fold family. Indeed, in sequence space, the distribution of preferred lengths for tightened end fragments and that for topohydrophobic separation match. In addition to this statistically significant similarity, the extremities of these 'closed loops' may be preferentially occupied by topohydrophobic positions, as observed on a random sample of various folds. This observation may be of special interest for sequence comparison of distantly related proteins. It is also important for the ab initio prediction of protein folds, considering the remarkable topological properties of topohydrophobic positions and their paramount importance within folding nuclei. Consequently, topohydrophobic positions locking the 'closed loops' belong to the deep cores of protein domains and might have a key role in the folding process.

Increased shedding of angiotensin-converting enzyme by a mutation identified in the stalk region.

Angiotensin-converting enzyme (ACE), an enzyme that plays a major role in vasoactive peptide metabolism, is a type 1 ectoprotein, which is released from the plasma membrane by a proteolytic cleavage occurring in the stalk sequence adjacent to the membrane anchor. In this study, we have discovered the molecular mechanism underlying the marked increase of plasma ACE levels observed in three unrelated individuals. We have identified a Pro(1199) --> Leu mutation in the juxtamembrane stalk region. In vitro analysis revealed that the shedding of [Leu(1199)]ACE was enhanced compared with wild-type ACE. The solubilization process of [Leu(1199)]ACE was stimulated by phorbol esters and inhibited by compound 3, an inhibitor of ACE-secretase. The results of Western blot analysis were consistent with a cleavage at the major described site (Arg(1203)/Ser(1204)). Two-dimensional structural analysis of ACE showed that the mutated residue was critical for the positioning of a specific loop containing the cleavage site. We therefore propose that a local conformational modification caused by the Pro(1199) --> Leu mutation leads to more accessibility at the stalk region for ACE secretase and is responsible for the enhancement of the cleavage-secretion process. Our results show that different molecular mechanisms are responsible for the common genetic variation of plasma ACE and for its more rare familial elevation.

Molecular characterization of the major membrane skeletal protein in the ciliate Tetrahymena pyriformis suggests n-plication of an early evolutionary intermediate filament protein subdomain.

Epiplasmin C is the major protein component of the membrane skeleton in the ciliate Tetrahymena pyriformis. Cloning and analysis of the gene encoding epiplasmin C showed this protein to be a previously unrecognized protein. In particular, epiplasmin C was shown to lack the canonical features of already known epiplasmic proteins in ciliates and flagellates. By means of hydrophobic cluster analysis (HCA), it has been shown that epiplasmin C is constituted of a repeat of 25 domains of 40 residues each. These domains are related and can be grouped in two families called types I and types II. Connections between types I and types II present rules that can be evidenced in the sequence itself, thus enforcing the validity of the splitting of the domains. Using these repeated domains as queries, significant structural similarities were demonstrated with an extra six heptads shared by nuclear lamins and invertebrate cytoplasmic intermediate filament proteins and deleted in the cytoplasmic intermediate filament protein lineage at the protostome-deuterostome branching in the eukaryotic phylogenetic tree.

Voronoï tessellation reveals the condensed matter character of folded proteins.

The packing geometry of amino acids in folded proteins is analyzed via a modified Voronoï tessellation method which distinguishes bulk and surface. From a statistical analysis of the Voronoï cells over 40 representative proteins, it appears that the packings are in average similar to random packings of hard spheres encountered in condensed matter physics, with a quite strong fivefold local symmetry. Moreover, the statistics permits one to establish a classification of amino acids in terms of increasing propensity to be buried in agreement with what is known from chemical considerations.

Beta-sheet modeling by helical surfaces.

We present a topological description of a beta-sheet in terms of a piece of helical surface. It requires only two easy-to-handle parameters: the twist, i.e. the turn of the helical surface per residue, and the coiling, which is a curvature along the strands or in the direction perpendicular to the strands of the sheet. This method applies fairly well to three- and four-strand sheets, forming a too limited structure to be able to build a barrel. From an analysis of beta-sheets derived from a structural database, we show that this picture can even be reduced to the use of one main value, the twist angle. The dependence of beta-sheet twisting on the number of strands in a sheet, and also on the length and direction of strands, has been demonstrated. The applications of such a description may include the rapid modeling of 3D structures.

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.

Sequence analysis identifies a ras-associating (RA)-like domain in the N-termini of band 4.1/JEF domains and in the Grb7/10/14 adapter family.

RA (RalGEF/AF6 or Ras-associating) domains are found in a wide variety of proteins, several of which are known to be Ras-GTP effectors. The three dimensional structure of the RA domain has been experimentally determined in Ral-guanine nucleotide exchange factor (Ral-GEF) and found to be similar to that of the Ras-binding domain of c-Raf1, in spite of a very low level of sequence identity. Using various approaches of sequence analysis, including automatic procedures such as BLAST2, profilescan, and hidden Markov models (HMM), as well as the bidimensional hydrophobic cluster analysis (HCA), here we found that a region with a similar structure is also present at the N-terminus of the band 4.1/JEF domain of KIAA0316 (a human cDNA open reading frame) and H09G03.2 (a related protein sequence predicted from C. elegans genome cloning), as well as in a particular class of adapter proteins including Grb7, Grb10, Grb14, MIG-10, and PRP48. Although the structural conservation of this motif does not necessarily imply a conservation of its ability to bind small GTPases of the Ras superfamily, several proteins with a band 4.1/JEF domain and adapters of the Grb7 group have close functional relationships with such small GTPases. Thus, our finding raises the intriguing possibility of a direct interaction between members of these two groups of proteins and Ras-like GTP-binding proteins.

Folding of the human protein FKBP. Lattice Monte-Carlo simulations.

Monte-Carlo simulations of folding of the human protein FKBP are presented. The protein is confined in a simple cubic lattice and only nearest-neighbour interactions are considered. The evolution of protein structure, energy and diameter is followed over time. Starting from different extended conformations, compact globular forms with a hydrophobic core are reached above a critical temperature Tc, while below Tc the protein 'freezes' into high-energy, non-compact states. In the temperature range of folding, all the recorded intermediate states belong to two structural groups, where the process spends most of its time, separated by relatively fast transitions. During folding, the protein is successively composed of three and two compact fragments, whose separation occurs at loop positions. From comparisons performed on a domain of the family sharing 24% identity with FKBP, it appears that the number of fragments, and therefore their location, are sequence dependent.

The D152H mutation found in growth hormone insensitivity syndrome impairs expression and function of human growth hormone receptor but is silent in rat receptor.

In two patients with growth hormone (GH) insensitivity syndrome (Laron syndrome), in whom the GH receptor is able to bind the hormone, the D152H mutation was identified, and lack of dimerization was proposed to explain GH resistance in these patients. To examine further the consequences of the substitution of conserved aspartate 152 on the function of the GH receptor (GHR), we reproduced the mutation in vitro on the full length GH receptor cDNA from man and rat. Effects of the mutation on expression and activity of the GHR were analyzed in 293 cells transfected with wild-type and mutant GHR cDNAs. Mutant human receptor protein was expressed at a lower level than wild-type receptor and its activity was reduced: GH-dependent signal transducer and activator of transcription 5 (Stat5)-mediated transactivation of a reporter gene was lower in 293 cells transfected with mutant GHR cDNA than in transfected cells expressing a comparable level of wild-type GHR. The membrane-bound form of the mutant and of the wild-type human GHR were able to homodimerize, as suggested by the size of the complexes detected in cross-linking experiments with 125I-human (h) GH, and also by the activity in the functional test. With the soluble GHR resulting from proteolysis of the wild-type membrane form, no dimeric complexes could be detected. However, when a soluble receptor lacking the transmembrane and cytoplasmic domains of the receptor was expressed, wild-type and not mutant GH binding protein (GHBP) was able to form dimers in the presence of hGH. The amino acid substitution has no effect on either expression or function of the rat receptor. Structural modeling of D152H soluble human and rat GHR (GHBP) supports the species-specific functional consequences of the mutation. Evaluation of the functional importance of the mutation strongly suggests that impairment in expression and activity of the mutant receptor, rather than complete lack of dimerization, explains the GH resistance of the patients.

Deciphering protein sequence information through hydrophobic cluster analysis (HCA): current status and perspectives.

Ten years after the idea of hydrophobic cluster analysis (HCA) was conceived and first published, theoretical and practical experience has shown this unconventional method of protein sequence analysis to be particularly efficient and sensitive, especially with families of sequences sharing low levels of sequence identity. This extreme sensitivity has made it possible to predict the functions of genes whose sequence similarities are hardly if at all detectable by current one-dimensional (1D) methods alone, and offers a new way to explore the enormous amount of data generated by genome sequencing. HCA also provides original tools to understand fundamental features of protein stability and folding. Since the last review of HCA published in 1990 [1], significant improvements have been made and several new facets have been addressed. Here we wish to update and summarize this information.

Molecular modeling of immunoglobulin light chains implicates hydrophobic residues in non-amyloid light chain deposition disease.

Light chain deposition disease is a severe complication of certain immunoproliferative disorders, due to the secretion of a monoclonal light chain which precipitates close to basement membranes of several tissues. A kappa isotype restriction and an unusual frequency of a variable region subgroup (VkappaIV) suggest that precise structural features govern the propensity of pathogenic light chains to precipitate in extracellular spaces. We studied primary structures of light chains from six patients with light chain deposition disease in comparison with light chains from other pathological conditions. Sequence alignment revealed the presence of certain amino acids only in light chain deposition disease, in particular non-polar replacing hydrophilic residues. To determine the role of these residues, structures of the variable domain from four kappa chains belonging to VkappaI and VkappaIV subgroups responsible for deposition disease were modeled using known immunoglobulins as templates. The most evident structural features shared by all pathogenic light chains were hydrophobic residues exposed to the solvent in complementarity determining regions 1 or 3. In contrast to immunoglobulin light chain-related amyloidosis, where deposition of organized material might be due to electrostatic interactions between light chain dimers, hydrophobic interactions could enhance amorphous precipitation in non-amyloid light chain deposition disease.

A global taxonomy of loops in globular proteins.

A bank of loops from three to eight amino acid residues long has been constituted. On the basis of statistical analysis of occurrences of conformations and residue, loops could be divided into two parts: the side residues directly bonded to the secondary structure flanking element, and the inner part. The conformations of the side residues are correlated to the nature of their neighboring flanks, while the inner residues adopt conformations uncorrelated from one residue to the next; thus they are unrelated to the flanks. Two zones in the Ramachandran plot are important: alpha L and beta P. In particular, the high occurrence of alpha L, mainly occupied by glycine residues, is necessary to induce flexibility and thus allow loops to comply with the geometrical constraints of the flanks. An algorithm of clustering has been used to aggregate loops of the same length within families of similar 3D structures. At each position in each cluster, sequence and conformational signatures have been deduced if the occurrence of a residue (or a conformation) is higher than an equiprobable distribution over all clusters. The result is that some positions favor particular amino acids and conformations, which are typical of a cluster although not unique. This is an indication of a relation between structure and sequence in loops. A taxonomy is proposed that classifies the various clusters. It relies on two terms: the mean distance between the first and last C alpha in one cluster and, perpendicular to this line, the distance to the center of gravity of the cluster. It is noteworthy that the differently populated clusters represented in such 2D plots can be separated. Thus, although the conformations of loops in globular proteins could cover a continuum, it has been possible to cluster them into a limited number of well populated families and superfamilies. This basic feature of protein architecture could be further exploited to better predict their geometry.

SUBIM: a program for analysing the Kabat database and determining the variability subgroup of a new immunoglobulin sequence.

Although various programs are available to extract all the information included in protein sequence databases, none is dedicated to immunoglobulins. For this purpose, we designed a program, SUBIM, which is adapted to the Kabat database specialized in immunoglobulin sequences. Besides all the possibilities of any database searching program, SUBIM analyses new sequences of variable regions and determines the variability subgroup they belong to. It also numbers the new sequence according to the system established by Kabat and co-workers for an easier comparison with the other immunoglobulins, thus realizing an automatic alignment with other members of a given type of immunoglobulin chain. This program is largely machine independent and requires very little memory, and should help biochemists concerned with new immunoglobulin sequences.

Automat and BLAST: comparison of two protein sequence similarity search programs.

Since the early 1980s, protein/DNA sequence similarity search has become of major importance to biologists, and the need for fast and efficient tools grows with the size of databanks. Two programs use the strategy of finite state deterministic automatons to accomplish these searches. One of these two is BLAST, which is now widely used, and the other Automat, which has just been published. The differences and similarities in their basic principles, their use and their performances are analysed in this paper in order to allow optimal use of these important softwares.

Structural comparisons lead to the definition of a new superfamily of NAD(P)(H)-accepting oxidoreductases: the single-domain reductases/epimerases/dehydrogenases (the 'RED' family).

Using both primary- and tertiary-structure comparisons, we have established new structural similarities shared by reductases, epimerases and dehydrogenases not previously known to be related. Despite the low sequence identity (down to 10%), short consensus segments are identified. We show that the sequence, the active site and the supersecondary structure are well conserved in these proteins. New homologues (the protochlorophyllide reductases) are detected, and we define a new superfamily composed of single-domain dinucleotide-binding enzymes. Rules for the cofactor-binding specificity are deduced from our sequence alignment. The involvement of some amino acids in catalysis is discussed. Comparison with two-domain dehydrogenases allows us to distinguish two general mechanisms of divergent evolution.

A computational and experimental study of the bending induced at a double-triple helix junction.

We have studied the conformation of a 17 base-pair homopyrimidine.homopurine triple helix formed on a fragment of duplex DNA derived from Simian Virus SV40. Gel retardation assays indicate that an 80 base-pair fragment has an altered conformation when the triple helix is formed, which is most likely to result from an induced bend in the DNA. Investigation of the detailed conformation of the double helix-triple helix junctions has been performed by means of molecular modelling. Bending on the 5' and 3' sides of the third strand oligonucleotide are not located at equivalent positions with respect to the junctions, which is explained in terms of base stacking. The junction effects on DNA structure, induced by the requirement for cytosine protonation in the Hoogsteen-bonded strand to form CGC+ base triplets, are also discussed.