Structure and phospholipid transfer activity of human PLTP: analysis by molecular modeling and site-directed mutagenesis.

The plasma phospholipid transfer protein (PLTP) is an important regulator of high density lipoprotein (HDL) metabolism. We have here, based on sequence alignments of the plasma LPS-binding/lipid transfer protein family and the X-ray structure of the bactericidal/permeability increasing protein (BPI), modeled the structure of PLTP. The model predicts a two-domain architecture with conserved lipid-binding pockets consisting of apolar residues in each domain. By site-directed mutagenesis of selected amino acid residues and transient expression of the protein variants in HeLa cells, the pockets are shown to be essential for PLTP-mediated phospholipid transfer. A solid phase ligand binding assay was used to determine the HDL-binding ability of the mutants. The results suggest that the observed decreases in phospholipid transfer activity of the N-terminal pocket mutants cannot be attributed to altered HDL-binding, but the C-terminal lipid-binding pocket may be involved in the association of PLTP with HDL. Further, the essential structural role of a disulfide bridge between cysteine residues 146 and 185 is demonstrated. The structural model and the mutants characterized here provide powerful tools for the detailed analysis of the mechanisms of PLTP function.

A theoretical model for the Gla-TSR-EGF-1 region of the anticoagulant cofactor protein S: from biostructural pathology to species-specific cofactor activity.

Protein S (PS), which functions as a species-specific anticoagulant cofactor to activated protein C (APC), is a mosaic protein that interacts with the phospholipid membrane via its gamma-carboxyglutamate-rich (Gla) module. This module is followed by the thrombin-sensitive region (TSR), sensitive to thrombin cleavage, four epidermal growth factor (EGF)-like modules and a last region referred to as the sex hormone binding globulin (SHBG) domain. Of these, the TSR and the first EGF-like regions have been shown to be important for the species-specific interaction with APC. Difficulties in crystallising PS have so far hindered its study at the atomic level. Here, we report theoretical models for the Gla and EGF-1 modules of human PS constructed using prothrombin and factor X experimental structures. The TSR was built interactively. Analysis of the model linked with the large body of biochemical literature on PS and related proteins leads to suggestions that (i) the TSR stabilises the calcium-loaded Gla module through hydrophobic and ionic interactions and its conformation depends on the presence of the Gla module; (ii) the TSR does not form a calcium binding site but is protected from thrombin cleavage in the calcium-loaded form owing to short secondary structure elements and close contact with the Gla module; (iii) the PS missense mutations in this region are consistent with the structural data, except in one case which needs further investigation; and (iv) the two PS 'faces' involving regions of residues Arg49-Gln52-Lys97 (TSR-EGF-1) and Thr103-Pro106 (EGF-1) may be involved in species-specific interactions with APC as they are richer in nonconservative substitution when comparing human and bovine protein S. This preliminary model helps to plan future experiments and the resulting data will be used to further validate and optimise the present structure.

Structural analysis of an anti-estradiol antibody.

An anti-estradiol antibody with improved specificity is searched for by combining steroid analog binding studies, mutant antibodies obtained from a phage-display library and structural modeling. Three-dimensional models for the anti-estradiol antibody 57-2 were constructed by comparative model building. Estradiol and analogs were docked into the combining site and molecular dynamics simulation was used to further refine this area of the protein. Cross-reactivities measured against 36 steroid analogs were used to help in the docking process and to evaluate the models. The roles of a number of residues were assessed by characterization of cross-reactivity mutants obtained from a phage display library. The cross-reactivity data and the results observed for mutants are explained by the structural model, in which the estradiol D-ring inserts deeply into the binding site and interacts with the antibody through at least one specific hydrogen bond. The binding data strongly suggest that this hydrogen bond connects the estradiol 17-hydroxyl group with the side chain of Gln H35. As expected for the binding of a small aromatic molecule, the antibody binding site contains many aromatic residues, e.g. Trp H50, H95 and L96 and Tyr L32, L49 and Phe L91.

Cello-oligosaccharide hydrolysis by cellobiohydrolase II from Trichoderma reesei. Association and rate constants derived from an analysis of progress curves.

The hydrolysis of soluble cello-oligosaccharides, with a degree of polymerisation of 4-6, catalysed by cellobiohydrolase II from Trichoderma reesei was studied using 1H-NMR spectroscopy and HPLC. The experimental progress curves were analysed by fitting numerically integrated kinetic equations, which provided cleavage patterns and kinetic constants for each oligosaccharide. This analysis procedure accounts for product inhibition and avoids the initial slope approximation. No glucose was detected at the beginning of the reaction indicating that only the internal glycosidic linkages are attacked. For cellotetraose only the second glycosidic linkage was cleaved. For cellopentaose and cellohexaose the second and the third glycosidic linkage from the non-reducing end were cleaved with approximately equal probability. The degradation rates of these cello-oligosaccharides, 1-12 s-1 at 27 degrees C, are about 10-100 times faster than for the 4-methylumbelliferyl substituted analogs or for collotriose. No intermediate products larger than cellotriose were released. The degradation rate for cellotetraose were higher than its off-rate, which accounts for the processive degradation of cellohexaose. A high cellohexaose/enzyme ratio caused slow reversible inactivation of the enzyme.

The active site of Trichoderma reesei cellobiohydrolase II: the role of tyrosine 169.

Trichoderma reesei cellobiohydrolase II (CBHII) is an exoglucanase cleaving primarily cellobiose units from the non-reducing end of cellulose chains. The beta-1,4 glycosidic bond is cleaved by acid catalysis with an aspartic acid, D221, as the likely proton donor, and another aspartate, D175, probably ensuring its protonation and stabilizing charged reaction intermediates. The catalytic base has not yet been identified experimentally. The refined crystal structure of CBHII also shows a tyrosine residue, Y169, located close enough to the scissile bond to be involved in catalysis. The role of this residue has been studied by introducing a mutation Y169F, and analysing the kinetic and binding behavior of the mutated CBHII. The crystal structure of the mutated enzyme was determined to 2.0 A resolution showing no changes when compared with the structure of native CBHII. However, the association constants of the mutant enzyme for cellobiose and cellotriose are increased threefold and for 4-methylumbelliferyl cellobioside over 50-fold. The catalytic constants towards cellotriose and cellotetraose are four times lower for the mutant. These data suggest that Y169, on interacting with a glucose ring entering the second subsite in a narrow tunnel, helps to distort the glucose ring into a more reactive conformation. In addition, a change in the pH activity profile was observed. This indicates that Y169 may have a second role in the catalysis, namely to affect the protonation state of the active site carboxylates, D175 and D221.

Structural investigation of the alpha-1-antichymotrypsin: prostate-specific antigen complex by comparative model building.

Prostate-specific antigen (PSA), produced by prostate cells, provides an excellent serum marker for prostate cancer. It belongs to the human kallikrein family of enzymes, a second prostate-derived member of which is human glandular kallikrein-1 (hK2). Active PSA and hK2 are both 237-residue kallikrein-like proteases, based on sequence homology. An hK2 model structure based on the serine protease fold is presented and compared to PSA and six other serine proteases in order to analyze in depth the role of the surface-accessible loops surrounding the active site. The results show that PSA and hK2 share extensive structural similarity and that most amino acid replacements are centered on the loops surrounding the active site. Furthermore, the electrostatic potential surfaces are very similar for PSA and hK2. PSA interacts with at least two serine protease inhibitors (serpins): alpha-1-antichymotrypsin (ACT) and protein C inhibitor (PCI). Three-dimensional model structures of the uncleaved ACT molecule were developed based upon the recent X-ray structure of uncleaved antithrombin. The serpin was docked both to PSA and hK2. Amino acid replacements and electrostatic complementarities indicate that the overall orientation of the proteins in these complexes is reasonable. In order to investigate PSA's heparin interaction sites, electrostatic computations were carried out on PSA, hK2, protein C, ACT, and PCI. Two heparin binding sites are suggested on the PSA surface and could explain the enhanced complex formation between PSA and PCI, while inhibiting the formation of the ACT-PSA complex, PSA, hK2, and their preliminary complexes with ACT should facilitate the understanding and prediction of structural and functional properties for these important proteins also with respect to prostate diseases.

Structure of a cyclic peptide with a catalytic triad, determined by computer simulation and NMR spectroscopy.

We report the design of a cyclic, eight-residue peptide that possesses the catalytic triad residues of the serine proteases. A manually built model has been relaxed by 0.3 ns of molecular dynamics simulation at room temperature, during which no major changes occurred in the peptide. The molecule has been synthesised and purified. Two-dimensional NMR spectroscopy provided 35 distance and 7 torsion angle constraints, which were used to determine the three-dimensional structure. The experimental conformation agrees with the predicted one at the beta-turn, but deviates in the arrangement of the disulphide bridge that closes the backbone to a ring. A 1.2 ns simulation at 600 K provided extended sampling of conformation space. The disulphide bridge reoriented into the experimental arrangement, producing a minimum backbone rmsd from the experimental conformation of 0.8 A. At a later stage in the simulation, a transition at Ser3 produced more pronounced high-temperature behaviour. The peptide hydrolyses p-nitrophenyl acetate about nine times faster than free histidine.

Comparative modeling of the three CP modules of the beta-chain of C4BP and evaluation of potential sites of interaction with protein S.

A computer model of the beta-chain of C4b-binding protein (C4BP) was constructed, using the backbone fold of the NMR structures of the sixteenth CP module of factor H (H16) and of a pair of modules consisting of the fifteenth and sixteenth CPs of factor H (H15-16). The characteristic hydrophobic core responsible for dictating the three-dimensional structure of the CP family is conserved in the amino acid sequence of C4BP beta 1, beta 2 and beta 3. The distribution of the electrostatic potential shows that the model is mainly covered by a negative contour. Interestingly, a positive area is observed in the C-terminal region of the first CP module, enclosing peptide 31-45, known to be a binding site for protein S. This observation suggests that electrostatic interactions can be of importance for the interaction of C4BP to protein S. A solvent-accessible hydrophobic patch, located nearby and involving the peptide 31-45, was also found in the model, further confirming that this area is involved in the interaction with protein S. The contribution of beta-chain residues 31-45 to the affinity for protein S was studied further by means of synthetic mutant peptides. The results suggest that both electrostatic and hydrophobic interactions are important for the binding to protein S.

Extracellular domain of type I receptor for transforming growth factor-beta: molecular modelling using protectin (CD59) as a template.

We have observed that the extracellular domain of T beta RI and protectin (CD59), an inhibitor of the membrane attack complex of complement, share structural features, a distinct spacing of ten cysteines and a C-terminal 'Cys-box'. Based on these common features and the recently determined NMR-structure of protectin, a three-dimensional model for the extracellular domain of T beta RI was constructed. After energy minimization and molecular dynamics simulation, a structure with four extending fingers (pes quattvordigitorum) and two clusters of charged residues was obtained. This model provides a view to the understanding of interactions between T beta RI, T beta RII and TGF beta during ligand recognition and signal transduction.

Effects of pH and high ionic strength on the adsorption and activity of native and mutated cellobiohydrolase I from Trichoderma reesei.

Cellobiohydrolase I (CBHI) is the major cellulase of Trichoderma reesei. The enzyme contains a discrete cellulose-binding domain (CBD), which increases its binding and activity on crystalline cellulose. We studied cellulase-cellulose interactions using site-directed mutagenesis on the basis of the three-dimensional structure of the CBD of CBHI. Three mutant proteins which have earlier been produced in Saccharomyces cerevisiae were expressed in the native host organism. The data presented here support the hypothesis that a conserved tyrosine (Y492) located on the flat and more hydrophilic surface of the CBD is essential for the functionality. The data also suggest that the more hydrophobic surface is not directly involved in the CBD function. The pH dependence of the adsorption revealed that electrostatic repulsion between the bound proteins may also control the adsorption. The binding of CBHI to cellulose was significantly affected by high ionic strength suggesting that the interaction with cellulose includes a hydrophobic effect. High ionic strength increased the activity of the isolated core and of mutant proteins on crystalline cellulose, indicating that once productively bound, the enzymes are capable of solubilizing cellulose even with a mutagenized or with no CBD.

Progress-curve analysis shows that glucose inhibits the cellotriose hydrolysis catalysed by cellobiohydrolase II from Trichoderma reesei.

NMR spectroscopy and HPLC were used to investigate the hydrolysis of cellotriose by cellobiohydrolase II from Trichoderma reesei. Substrate and product concentrations were followed as a function of time. Progress curves were calculated by forward numerical integration of the full kinetic equations and were fitted to the experimental data. Binding and rate constants were obtained from this fit, whereby no initial slope or Michaelis-Menten approximation was used. The progress curves from a single experiment sufficed to produce agreement with the Michaelis-Menten model (eight experiments). The absence of a kinetic isotope effect was proven. The progress-curve analysis showed that a simple degradation model cannot describe the experimental time-courses at substrate concentrations greater than 1 mM. A model containing competitive inhibition from cellobiose as well as non-competitive inhibition from glucose was developed. This four-parameter model accurately reproduces about 1000 experimental data points covering five orders of magnitude in oligosaccharide concentrations. Glucose binding to the enzyme/cellotriose complex retards, in a non-competitive fashion, cellotriose hydrolysis by at least a factor of 30. A structural model for the non-competitive inhibition is discussed. The NMR experiment also produced individual progress curves for the alpha and beta anomers. The beta anomer of cellotriose was degraded 2.5-times faster than the alpha anomer.

Structure of the Ca(2+)-free Gla domain sheds light on membrane binding of blood coagulation proteins.

Reversible membrane binding of gamma-carboxyglutamic acid (Gla)-containing coagulation factors requires Ca(2+)-binding to 10-12 Gla residues. Here we describe the solution structure of the Ca(2+)-free Gla-EGF domain pair of factor x which reveals a striking difference between the Ca(2+)-free and Ca(2+)-loaded forms. In the Ca(2+)-free form Gla residues are exposed to solvent and Phe 4, Leu 5 and Val 8 form a hydrophobic cluster in the interior of the domain. In the Ca(2+)-loaded form Gla residues ligate Ca2+ in the core of the domain pushing the side-chains of the three hydrophobic residues into the solvent. We propose that the Ca(2+)-induced exposure of hydrophobic side chains is crucial for membrane binding of Gla-containing coagulation proteins.

Molecular dynamics simulation of fungal cellulose-binding domains: differences in molecular rigidity but a preserved cellulose binding surface.

A total of 23 fungal cellulose-binding domain (CBD) sequences were aligned. Structural models of the cellulose-binding domain of an exoglucanase (CBHII) and of three endoglucanases (EGI, EGII and EGV) from Trichoderma reesei cellulases were homology modelled based on the NMR structure of the fungal cellobiohydrolase CBHI, from the same organism. The completed models and the known structure of the CBHI cellulose-binding domain were refined by molecular dynamics simulations in water. All four models were found to be very similar to the structure of the CBHI cellulose-binding domain and sequence comparison indicated that in general the three-dimensional structures of fungal cellulose-binding domains are very similar. In all the CBDs studied, two disulphide bridges apparently stabilize the polypeptide fold. From the models, and additional disulphide bridge was predicted in EGI and CBHII, and in eight further CBDs from other organisms. Three highly conserved aromatic residues on the hydrophilic side of the wedge make this surface flat. This surface is expected to make contact with the substrate. Three invariant amino acids, Gln7, Asn29 and Gln34, on this flat face are in suitable positions for hydrogen bonding with the cellulose surface. Analysis of the differences in the protein surface properties indicated that the endoglucanases tend to be more hydrophilic than the exoglucanases. The largest structural variation was found around positions 12-16. The fungal CBD sequences are discussed in relation to variations in function and pH dependence. Comparison of the modelled structures with experimental binding data for the CBHI and EGI allowed the formulation of a qualitative relationship to cellulose affinity.(ABSTRACT TRUNCATED AT 250 WORDS)

Atomic force microscope images of lipid layers spread from vesicle suspensions.

The layer formation of unilamellar vesicles of L-alpha-dimyristoyl phosphatidylcholine (DMPC) spread onto the air/liquid interface has been investigated. The layers were transferred to clean glass slides and onto slides made hydrophobic with multilayers of Cd arachidate. Aged vesicle suspensions aggregate during storage and are transferred as large domains as imaged with atomic force microscopy (AFM). Freshly prepared vesicles fuse and can be transferred as monolayers to hydrophobic supports. Furthermore, AFM images reveal the importance of positioning the solid support parallel to the moving barrier in order to obtain more uniform deposition of Cd arachidate.

Introduction of lysine residues on the light chain constant domain improves the labelling properties of a recombinant Fab fragment.

Europium chelates provide a non-radioactive alternative for sensitive labelling of antibodies for diagnostic immunoassays. Lysine residues at antibody surfaces are ready targets for labelling by an isothiocyanate derivative of the europium chelate (Eu3+). Here the labelling efficiency of a recombinant anti-human alpha-fetoprotein (hAFP) Fab fragment has been improved by increasing its lysine content by protein engineering. Molecular modelling was used to identify three light chain constant domain surface arginine residues, R154, R187 and R210, which were mutated to lysine residues. The mutations did not influence the affinity of the lysine-enriched Fab fragment and its labelling efficiency was found to be approximately 40% higher than that of the wild-type Fab fragment. With low degree of labelling, the affinities of the two Fab fragments were identical and comparable with that of the original monoclonal anti-hAFP IgG. With a higher degree of labelling the affinities of both Fab fragments decreased more than that of the intact IgG since more lysine residues are available for labelling in the additional heavy chain constant domains of the larger molecule. Electrostatic adsorption and covalent immobilization of the Fab fragments were characterized by BIAcore and the lysine-enriched Fab fragment was found to be more efficiently immobilized to an activated carboxymethyl surface.

A novel, small endoglucanase gene, egl5, from Trichoderma reesei isolated by expression in yeast.

A method is presented for the isolation of genes encoding hydrolytic enzymes without any knowledge of the corresponding proteins. cDNA made from the organism of interest is cloned into a yeast vector to construct an expression library in the yeast Saccharomyces cerevisiae. Colonies producing hydrolytic enzymes are screened by activity plate assays. In this work, we constructed a yeast expression library from the filamentous fungus Trichoderma reesei and isolated a new beta-1,4-endoglucanase gene on plates containing beta-glucan. This gene, egl5, codes for a previously unknown small protein of 242 amino acids. Despite its small size, the protein contains two conservative domains found in Trichoderma cellulases, namely the cellulose-binding domain (CBD) and the linker region that connects the CBD to the catalytic core domain. Molecular modelling of the EGV CBD revealed some interesting structural differences compared to the CBD of the major cellulase CBHI from T. reesei. The catalytic core of EGV is unusually small for a cellulase and represents a new family of cellulases (Family K) and of glycosyl hydrolases (Family 45) together with the endoglucanase B of Pseudomonas fluorescens and the endoglucanase V of Humicola insolens on the basis of hydrophobic cluster analysis.

How an epidermal growth factor (EGF)-like domain binds calcium. High resolution NMR structure of the calcium form of the NH2-terminal EGF-like domain in coagulation factor X.

Domains homologous to the epidermal growth factor (EGF) are important building blocks for extracellular proteins. Proteins containing these domains have been shown to function in such diverse biological processes as blood coagulation, complement activation, and the developmental determination of embryonic cell fates. Many of these proteins require calcium for their biological function. In the case of coagulation factors IX and X and anticoagulants proteins C and S, calcium has been found to bind to the EGF-like domains. We have now determined the three-dimensional structure of the calcium-bound form of the NH2-terminal EGF-like domain in coagulation factor X by two-dimensional NMR and simulated folding. Ligands to the calcium ion are the two backbone carbonyls in Gly-47 and Gly-64, as well as the side chains in Gln-49, erythro-beta-hydroxyaspartic acid (Hya) 63, and possibly Asp-46. The conserved Asp-48 is not a ligand in our present structures. The remaining ligands are assumed to be solvent molecules or, in the intact protein, ligands from neighboring domains. Other proteins interacting in a calcium-dependent manner may also contribute ligands. A comparison with the calcium-free form shows that calcium binding induces strictly local structural changes in the domain. Residues corresponding to the side chain ligands in factor X are conserved in many other proteins, such as the integral membrane protein TAN-1 of human lymphocytes and its developmentally important homolog, Notch, in Drosophila. Calcium binding to EGF-like domains may be crucial for numerous protein-protein interactions involving EGF-like domains in coagulation factors, plasma proteins, and membrane proteins. Therefore, there is reason to believe that this novel calcium site plays an important role in the biochemistry of extracellular proteins.

Three-dimensional structure of the apo form of the N-terminal EGF-like module of blood coagulation factor X as determined by NMR spectroscopy and simulated folding.

The three-dimensional structure of a 42-residue fragment containing the N-terminal EGF-like module of blood coagulation factor X was determined by means of 2D NMR spectroscopy and computer simulation. The spectroscopic data consisted of 370 NOE distances and 27 dihedral angle constraints. These were used to generate peptide conformations by molecular dynamics simulation. The simulations used a novel functional form for the constraint potentials and were performed with two time steps to ensure rapid execution. Apart from preliminary runs to aid assignment of NOEs, 60 runs resulted in 13 accepted structures, which have two antiparallel beta sheets, no alpha helices, and five tight turns. There is no hydrophobic cluster. The root mean square deviation for the backbone of the 13 conformations is 0.65 +/- 0.11 A against their mean conformation. About half of the side chains have well-defined structure. The overall conformation is similar to that of murine EGF.

Molecular dynamics simulation of the renin inhibitor H142 in water.

H142 is a synthetic decapeptide designed to inhibit renin, an enzyme acting in the regulation of blood pressure. The inhibiting effect of H142 is caused by a reduction of a -Leu-Val-peptide bond (i. e. C(= O)-NH----CH2-NH). The conformational and dynamical properties of H142 and its unreduced counterpart (H142n) was modelled by means of molecular dynamics simulations. Water was either included explicitly in the simulations or as a dielectric continuum. When water molecules surround the peptides, they remain in a more or less extended conformation through the simulation. If water is replaced by a dielectric continuum, the peptides undergo a conformational change from an extended to a folded state. It is not clear whether this difference is a consequence of a too short simulation time for the water simulations, a force-field artifact promoting extended conformations, or if the extended conformation represents the true conformational state of the peptide. A number of dynamic properties were evaluated as well, such as overall rotation, translational diffusion, side-chain dynamics and hydrogen bonding.