The epithelial adhesin 1 (Epa1p) from the human-pathogenic yeast Candida glabrata: structural and functional study of the carbohydrate-binding domain.

The yeast Candida glabrata represents the second major cause of clinical candidiasis cases in the world. The ability of this opportunistic pathogen to adhere to human epithelial and endothelial cells relies on the Epa adhesins, a large set of cell-wall proteins whose N-terminal domains are endowed with a calcium-dependent lectin activity. This feature allows the yeast cells to adhere to host cells by establishing multiple interactions with the glycans expressed on their cell membrane. The ligand-binding domain of the Epa1p adhesin, which is one of the best characterized in the Epa family, was expressed in Escherichia coli, purified and crystallized in complex with lactose. Sequence identity with the domain of another yeast adhesin, the Flo5p flocculin from Saccharomyces cerevisiae, was exploited for molecular replacement and the structure of the domain was solved at a resolution of 1.65 Å. The protein is a member of the PA14 superfamily. It has a ß-sandwich core and a DcisD calcium-binding motif, which is also present in the binding site of Flo5p. However, Epa1p differs from this homologue by the lack of a Flo5-like subdomain and by a significantly decreased accessibility of the solvent to the binding site, in which a calcium ion still plays an active role in the interactions with carbohydrates. This structural insight, together with fluorescence-assay data, confirms and explains the higher specificity of Epa1p adhesin for glycan molecules compared with the S. cerevisiae flocculins.

A camelid antibody fragment inhibits the formation of amyloid fibrils by human lysozyme.

Amyloid diseases are characterized by an aberrant assembly of a specific protein or protein fragment into fibrils and plaques that are deposited in various organs and tissues, often with serious pathological consequences. Non-neuropathic systemic amyloidosis is associated with single point mutations in the gene coding for human lysozyme. Here we report that a single-domain fragment of a camelid antibody raised against wild-type human lysozyme inhibits the in vitro aggregation of its amyloidogenic variant, D67H. Our structural studies reveal that the epitope includes neither the site of mutation nor most residues in the region of the protein structure that is destabilized by the mutation. Instead, the binding of the antibody fragment achieves its effect by restoring the structural cooperativity characteristic of the wild-type protein. This appears to occur at least in part through the transmission of long-range conformational effects to the interface between the two structural domains of the protein. Thus, reducing the ability of an amyloidogenic protein to form partly unfolded species can be an effective method of preventing its aggregation, suggesting approaches to the rational design of therapeutic agents directed against protein deposition diseases.

Antigen binding and solubility effects upon the veneering of a camel VHH in framework-2 to mimic a VH.

Heavy chain only antibodies of camelids bind their antigens with a single domain, the VHH, which acquired adaptations relative to classical VHs to function in the absence of a VL partner. Additional CDR loop conformations, outside the canonical loop structures of VHs, broaden the repertoire of the antigen-binding site. The combined effects of part of the CDR3 that folds over the "former" VL binding site and framework-2 mutations to more hydrophilic amino acids, enhance the solubility of VHH domains and prevent VL pairing. cAbAn33, a VHH domain specific for the carbohydrate moiety of the variant surface glycoprotein of trypanosomes, has a short CDR3 loop that does not cover the former VL binding site as well as a VH-specific Trp47 instead of the VHH-specific Gly47. Resurfacing its framework-2 region (mutations Tyr37Val, Glu44Gly and Arg45Leu) to mimic that of a human VH restores the VL binding capacity. In solution, the humanised VHH behaves as a soluble, monomeric entity, albeit with reduced thermodynamic stability and affinity for its antigen. Comparison of the crystal structures of cAbAn33 and its humanised derivative reveals steric hindrance exerted by VHH-specific residues Tyr37 and Arg45 that prevent the VL domain pairing, whereas Glu44 and Arg45 are key elements to avoid insolubility of the domain.

Crystallization of ornithine acetyltransferase from yeast by counter-diffusion and preliminary X-ray study.

A study is presented on the crystallization of ornithine acetyltransferase from yeast, which catalyzes the fifth step in microbial arginine synthesis. The use of the counter-diffusion technique removes the disorder present in one dimension in crystals grown by either the batch or hanging-drop techniques. This makes the difference between useless crystals and crystals that allow successful determination of the structure of the protein. The crystals belong to space group P4, with unit-cell parameters a = b = 66.98, c = 427.09 A, and a data set was collected to 2.76 A.

Hydrophobic surface patches on LolA of Pseudomonas aeruginosa are essential for lipoprotein binding.

Many lipoproteins reside in the outer membrane (OM) of Gram-negative bacteria, and their biogenesis is dependent on the Lol (localization of lipoproteins) system. The periplasmic chaperone LolA accepts OM-destined lipoproteins that are released from the inner membrane by the LolCDE complex and transfers them to the OM receptor LolB. The exact nature of the LolA-lipoprotein complex is still unknown. The crystal structure of Escherichia coli LolA features an open beta-barrel covered by alpha helices that together constitute a hydrophobic cavity, which would allow the binding of one acyl chain. However, OM lipoproteins contain three acyl chains, and the stoichiometry of the LolA-lipoprotein complex is 1:1. Here we present the crystal structure of Pseudomonas aeruginosa LolA that projects clear hydrophobic surface patches. Since these patches are large enough to accommodate acyl chains, their role in lipoprotein binding was investigated. Several LolA mutant proteins were created, and their functionality was assessed by studying their capacity to release lipoproteins produced in sphaeroplasts. Interruption of the largest hydrophobic patch completely destroyed the lipoprotein-releasing capacity of LolA, while interruption of smaller patches apparently reduced efficiency. Thus, the results show a new lipoprotein transport model that places (some of) the acyl chains on the hydrophobic surface patches.

Molecular basis for the preferential cleft recognition by dromedary heavy-chain antibodies.

Clefts on protein surfaces are avoided by antigen-combining sites of conventional antibodies, in contrast to heavy-chain antibodies (HCAbs) of camelids that seem to be attracted by enzymes' substrate pockets. The explanation for this pronounced preference of HCAbs was investigated. Eight single domain antigen-binding fragments of HCAbs (VHH) with nanomolar affinities for lysozyme were isolated from three immunized dromedaries. Six of eight VHHs compete with small lysozyme inhibitors. This ratio of active site binders is also found within the VHH pool derived from polyclonal HCAbs purified from the serum of the immunized dromedary. The crystal structures of six VHHs in complex with lysozyme and their interaction surfaces were compared to those of conventional antibodies with the same antigen. The interface sizes of VHH and conventional antibodies to lysozyme are very similar as well as the number and chemical nature of the contacts. The main difference comes from the compact prolate shape of VHH that presents a large convex paratope, predominantly formed by the H3 loop and interacting, although with different structures, into the concave lysozyme substrate-binding pocket. Therefore, a single domain antigen-combining site has a clear structural advantage over a conventional dimeric format for targeting clefts on antigenic surfaces.

A transparent two-dimensional in situ beam-position and profile monitor for synchrotron X-ray beamlines.

A compact, inexpensive and easy-to-construct two-dimensional in situ beam-position and profile monitor for synchrotron X-ray beamlines is presented. The device is based on the collection of spatially resolved scattered radiation from a polyimide foil. The X-ray beam passes through a foil placed in the path of the beam, which absorbs no more than 3% of the beam at 12 keV. The scattered radiation is collected at an angle of 90 degrees through a collimator located below the foil onto a CCD sensor. The device was tested on bending-magnet beamline BM26 at the ESRF synchrotron radiation source and has a positional sensitivity better than 10 microm with a large working range of 25 mm x 25 mm. Although the device is optimized for use in the range 10-12 keV, it can easily be modified for use with higher-energy beams by using a suitably chosen scattering foil.

Strong in vivo maturation compensates for structurally restricted H3 loops in antibody repertoires.

A central paradigm in immunology states that successful generation of high affinity antibodies necessitates an immense primary repertoire of antigen-combining sites. Much of the diversity of this repertoire is provided by varying one antigen binding loop, created by inserting randomly a D (diversity) gene out of a small pool between the V and J genes. It is therefore assumed that any particular D-encoded region surrounded by different V and J regions adopts a different conformation. We have solved the structure of two lysozyme-specific variable domains of heavy-chain antibodies isolated from two strictly unrelated dromedaries. These antibodies recombined identical D gene sequences to different V and J precursors with significant variance in their V(D)J junctions. Despite these large differences, the D-encoded loop segments adopt remarkably identical architectures, thus directing the antibodies toward identical epitopes. Furthermore, a striking convergent maturation process occurred in the V region, adapting both binders for their sub-nanomolar affinity association with lysozyme. Hence, on a structural level, humoral immunity may rely more on well developed maturation and selection systems than on the acquisition of large primary repertoires.

Kinetic and affinity predictions of a protein-protein interaction using multivariate experimental design.

We measured the influence of 14 mutations and 5 environmental variables (buffer perturbation) on the association and dissociation rate of a camel single domain antibody (cAb-Lys3) interacting with hen egg white lysozyme using a surface plasmon resonance-based biosensor. Based on this data set, we constructed quantitative predictive models for both kinetic (k(a) and k(d)) constants as for the affinity constant (K(d)). Mutations, after parameterization by quantitative descriptors, and buffers were selected using multivariate experimental design. These models were able to predict the corresponding parameters of four new variants of cAb-Lys3. Moreover, the models provide insights to the important chemical aspects of the interacting residues, which are difficult to deduce from the crystal structure. Our approach provides useful physicochemical information of protein-protein interactions in general. The information obtained from this kind of analysis complements and goes beyond that of conventional methods like alanine scanning and substitution by closely related amino acids. The mathematical modeling may contribute to a rational approach in the optimization of bio-molecules of biotechnological interest.

Cloning, preliminary characterization and crystallization of nucleoside hydrolases from Caenorhabditis elegans and Campylobacter jejuni.

The nucleoside hydrolases (NHs) are a family of nucleoside-modifying enzymes. They play an important role in the purine-salvage pathway of many pathogenic organisms which are unable to synthesize purines de novo. Although well characterized in protozoan parasites, their precise function and mechanism remain unclear in other species. For the first time, NHs from Caenorhabditis elegans and Campylobacter jejuni, which are representatives of mesozoa and bacteria, respectively, have been cloned and purified. Steady-state kinetics indicate a different substrate-specificity profile to previously described hydrolases. Native diffraction data sets were collected from crystals of NH from each organism. The hexagonal crystals (space group P6(2)22 or P6(4)22) of NH from C. elegans diffracted to a resolution of 2.8 A, while the data set from the orthorhombic crystals (space group I222 or I2(1)2(1)2(1)) of NH from C. jejuni could be processed to 1.7 A resolution. The unit-cell parameters were a = b = 102.23, c = 117.27 A in the former case and a = 101.13, b = 100.13, c = 81.37 A in the latter.

Enzyme-substrate interactions in the purine-specific nucleoside hydrolase from Trypanosoma vivax.

Nucleoside hydrolases are key enzymes in the purine salvage pathway of Trypanosomatidae and are considered as targets for drug design. We previously reported the first x-ray structure of an inosine-adenosine-guanosine preferring nucleoside hydrolase (IAG-NH) from Trypanosoma vivax (). Here we report the 2.0-A crystal structure of the slow D10A mutant in complex with the inhibitor 3-deaza-adenosine and the 1.6-A crystal structure of the same enzyme in complex with a genuine substrate inosine. The enzyme-substrate complex shows the substrate bound to the enzyme in a different conformation from 3-deaza-adenosine and provides a snapshot along the reaction coordinate of the enzyme-catalyzed reaction. The chemical groups on the substrate important for binding and catalysis are mapped. The 2'-OH, 3'-OH, and 5'-OH contribute 4.6, 7.5, and 5.4 kcal/mol to k(cat)/K(m), respectively. Specific interactions with the exocyclic groups on the purine ring are not required for catalysis. Site-directed mutagenesis indicates that the purine specificity of the IAG-NHs is imposed by a parallel aromatic stacking interaction involving Trp(83) and Trp(260). The pH profiles of k(cat) and k(cat)/K(m) indicate the existence of one or more proton donors, possibly involved in leaving group activation. However, mutagenesis of the active site residues around the nucleoside base and an alanine scan of a flexible loop near the active site fail to identify this general acid. The parallel aromatic stacking seems to provide the most likely alternative mechanism for leaving group activation.