Eduardo Horjales

TOM: a FRODO subpackage for protein-ligand fitting with interactive energy minimization

We have implemented the docking subpackage TOM into FRODO1 for studying protein-ligand interactions with interactive energy-minimization procedures. When using TOM, models of protein-ligand complexes are first created in a graphics display unit, followed by interactive energy-minimization treating both ligand and parts of the receptor as flexible units. The potential energy function includes Coulomb and van der Waals interactions. At present, two versions of TOM are available, running on a DEC VAX computer, an Evans and Sutherland PS300 display unit, or a Silicon Graphics IRIS workstation.

Structure and catalytic mechanism of glucosamine 6-phosphate deaminase from Escherichia coli at 2.1 å resolution

BACKGROUND Glucosamine 6-phosphate deaminase from Escherichia coli is an allosteric hexameric enzyme which catalyzes the reversible conversion of D-glucosamine 6-phosphate into D-fructose 6-phosphate and ammonium ion and is activated by N-acetyl-D-glucosamine 6-phosphate. Mechanistically, it belongs to the group of aldoseketose isomerases, but its reaction also accomplishes a simultaneous amination/deamination. The determination of the structure of this protein provides fundamental knowledge for understanding its mode of action and the nature of allosteric conformational changes that regulate its function. RESULTS The crystal structure of glucosamine 6-phosphate deaminase with bound phosphate ions is presented at 2.1 A resolution together with the refined structures of the enzyme in complexes with its allosteric activator and with a competitive inhibitor. The protein fold can be described as a modified NAD-binding domain. CONCLUSIONS From the similarities between the three presented structures, it is concluded that these represent the enzymatically active R state conformer. A mechanism for the deaminase reaction is proposed. It comprises steps to open the pyranose ring of the substrate and a sequence of general base-catalyzed reactions to bring about isomerization and deamination, with Asp72 playing a key role as a proton exchanger.

Cross-linked crystals of chloroperoxidase

Chloroperoxidase from Caldariomyces fumago was crystallized. The crystals were modified with several cross-linkers, but only glurataldehyde was able to produce catalytically active and insoluble crystals. Unlike other immobilized chloroperoxidase preparations, these catalytic crystals are more thermostable than the unmodified soluble enzyme. The enhanced stability is probably due to the structure conservation in the crystalline matrix. In addition, non-cross-linked chloroperoxidase crystals retained more activity than the soluble enzyme after incubation in an organic solvent with low water content. Although the cross-linked crystals were catalytically active, they showed lower specific activity than the soluble enzyme. This low activity may be due to non-specific reactions between the cross-linker and essential residues for catalysis. Alternative cross-linking strategies are discussed.

The allosteric transition of glucosamine-6-phosphate deaminase: the structure of the T state at 2.3 Å resolution

BACKGROUND The allosteric hexameric enzyme glucosamine-6-phosphate deaminase from Escherichia coli catalyses the regulatory step of N-acetylglucosamine catabolism, which consists of the isomerisation and deamination of glucosamine 6-phosphate (GlcN6P) to form fructose 6-phosphate (Fru6P) and ammonia. The reversibility of the catalysis and its rapid-equilibrium random kinetic mechanism, among other properties, make this enzyme a good model for studying allosteric processes. RESULTS Here we present the structure of P6(3)22 crystals, obtained in sodium acetate, of GlcN6P deaminase in its ligand-free T state. These crystals are very sensitive to X-ray radiation and have a high (78%) solvent content. The activesite lid (residues 162-185) is highly disordered in the T conformer; this may contribute significantly to the free-energy change of the whole allosteric transition. Comparison of the structure with the crystallographic coordinates of the R conformer (Brookhaven Protein Data Bank entry 1 dea) allows us to describe the geometrical changes associated with the allosteric transition as the movement of two rigid entities within each monomer. The active site, located in a deep cleft between these two rigid entities, presents a more open geometry in the T conformer than in the R conformer. CONCLUSIONS The differences in active-site geometry are related to alterations in the substrate-binding properties associated with the allosteric transition. The rigid nature of the two mobile structural units of each monomer seems to be essential in order to explain the observed kinetics of the deaminase hexamer. The triggers for both the homotropic and heterotropic allosteric transitions are discussed and particular residues are assigned to these functions. A structural basis for an entropic term in the allosteric transition is an interesting new feature that emerges from this study.

Effects of toxins Pi2 and Pi3 on human T lymphocyte Kv1.3 channels: the role of Glu7 and Lys24.

Abstract.Pandinus imperator scorpion toxins Pi2 and Pi3 differ only by a single amino acid residue (neutral Pro7 in Pi2 vs. acidic Glu7 in Pi3). The binding kinetics of these toxins to human Kv1.3 showed that the decreased on rate (kON= 2.18 × 108m−1sec−1 for Pi2 and 1.28 × 107m−1sec−1 for Pi3) was almost entirely responsible for the increased dissociation constant (Kd) of Pi3 (Kd= 795 pm) as compared to Pi2 (Kd= 44 pm). The ionic strength dependence of the association rates was exactly the same for the two toxins indicating that through-space electrostatic interactions can not account for the different on rates. Results were further analyzed on the basis of the three-dimensional structural models of the toxins. A 3D structure of Pi3 was generated from the NMR spectroscopy coordinates of Pi2 by computer modeling. The Pi3 model resulted in a salt bridge between Glu7 and Lys24 in Pi3. Based on this finding our interpretation of the reduced on rate of Pi3 is that the intramolecular salt bridge reduces the local positive electrostatic potential around Lys24 resulting in decreased short-range electrostatic interactions during the binding step. To support our finding, we constructed a 3D model of the Ser-10-Asp Charybdotoxin mutant displaying distinctly reduced affinity for Shaker channels. The mutant Charybdotoxin structure also displayed a salt bridge between residues Asp10 and Lys27 equivalent to the one between Glu7 and Lys24 in Pi3.

A Single Mutation at the Sheet Switch Region Results in Conformational Changes Favoring λ6 Light-Chain Fibrillogenesis

Systemic amyloid light-chain (LC) amyloidosis is a disease process characterized by the pathological deposition of monoclonal LCs in tissue. All LC subtypes are capable of fibril formation although lambda chains, particularly those belonging to the lambda6 type, are overrepresented. Here, we report the thermodynamic and in vitro fibrillogenic properties of several mutants of the lambda6 protein 6aJL2 in which Pro7 and/or His8 was substituted by Ser or Pro. The H8P and H8S mutants were almost as stable as the wild-type protein and were poorly fibrillogenic. In contrast, the P7S mutation decreased the thermodynamic stability of 6aJL2 and greatly enhanced its capacity to form amyloid-like fibrils in vitro. The crystal structure of the P7S mutant showed that the substitution induced both local and long-distance effects, such as the rearrangement of the V(L) (variable region of the light chain)-V(L) interface. This mutant crystallized in two orthorhombic polymorphs, P2(1)2(1)2(1) and C222(1). In the latter, a monomer that was not arranged in the typical Bence-Jones dimer was observed for the first time. Crystal-packing analysis of the C222(1) lattice showed the establishment of intermolecular beta-beta interactions that involved the N-terminus and beta-strand B and that these could be relevant in the mechanism of LC fibril formation. Our results strongly suggest that Pro7 is a key residue in the conformation of the N-terminal sheet switch motif and, through long-distance interactions, is also critically involved in the contacts that stabilized the V(L) interface in lambda6 LCs.

Structure-Function Relationships in Fungal Large-Subunit Catalases

Neurospora crassa has two large-subunit catalases, CAT-1 and CAT-3. CAT-1 is associated with non-growing cells and accumulates particularly in asexual spores; CAT-3 is associated with growing cells and is induced under different stress conditions. It is our interest to elucidate the structure-function relationships in large-subunit catalases. Here we have determined the CAT-3 crystal structure and compared it with the previously determined CAT-1 structure. Similar to CAT-1, CAT-3 hydrogen peroxide (H(2)O(2)) saturation kinetics exhibited two components, consistent with the existence of two active sites: one saturated in the millimolar range and the other in the molar range. In the CAT-1 structure, we found three interesting features related to its unusual kinetics: (a) a constriction in the channel that conveys H(2)O(2) to the active site; (b) a covalent bond between the tyrosine, which forms the fifth coordination bound to the iron of the heme, and a vicinal cysteine; (c) oxidation of the pyrrole ring III to form a cis-hydroxyl group in C5 and a cis-gamma-spirolactone in C6. The site of heme oxidation marks the starts of the central channel that communicates to the central cavity and the shortest way products can exit the active site. CAT-3 has a similar constriction in its major channel, which could function as a gating system regulated by the H(2)O(2) concentration before the gate. CAT-3 functional tyrosine is not covalently bonded, but has instead the electron relay mechanism described for the human catalase to divert electrons from it. Pyrrole ring III in CAT-3 is not oxidized as it is in other large-subunit catalases whose structure has been determined. Different in CAT-3 from these enzymes is an occupied central cavity. Results presented here indicate that CAT-3 and CAT-1 enzymes represent a functional group of catalases with distinctive structural characteristics that determine similar kinetics.

Two mammalian glucosamine-6-phosphate deaminases: a structural and genetic study

Glucosamine‐6‐phosphate deaminase (EC 3.5.99.6) is an allosteric enzyme that catalyzes the reversible conversion of D‐glucosamine‐6‐phosphate into D‐fructose‐6‐phosphate and ammonium. Here we describe the existence of two mammalian glucosamine‐6‐phosphate deaminase enzymes. We present the crystallographic structure of one of them, the long human glucosamine‐6‐phosphate deaminase, at 1.75 Å resolution. Crystals belong to the space group P212121 and present a whole hexamer in the asymmetric unit. The active‐site lid (residues 162–182) presented significant structural differences among monomers. Interestingly the region with the largest differences, when compared with the Escherichia coli homologue, was found to be close to the active site. These structural differences can be related to the kinetic and allosteric properties of both mammalian enzymes.

Structural flexibility, an essential component of the allosteric activation in Escherichia coli glucosamine-6-phosphate deaminase.

A new crystallographic structure of the free active-site R conformer of the allosteric enzyme glucosamine-6-phosphate deaminase from Escherichia coli, coupled with previously reported structures of the T and R conformers, generates a detailed description of the heterotropic allosteric transition in which structural flexibility plays a central role. The T conformers external zone [Horjales et al. (1999), Structure, 7, 527-536] presents higher B values than in the R conformers. The ligand-free enzyme (T conformer) undergoes an allosteric transition to the free active-site R conformer upon binding of the allosteric activator. This structure shows three alternate conformations of the mobile section of the active-site lid (residues 163-182), in comparison to the high B values for the unique conformation of the T conformer. One of these alternate R conformations corresponds to the active-site lid found when the substrate is bound. The disorder associated with the three alternate conformations can be related to the biological regulation of the K(m) of the enzyme for the reaction, which is metabolically required to maintain adequate concentrations of the activator, which holds the enzyme in its R state. Seven alternate conformations for the active-site lid and three for the C-terminus were refined for the T structure using isotropic B factors. Some of these conformers approach that of the R conformer in geometry. Furthermore, the direction of the atomic vibrations obtained with anisotropic B refinement supports the hypothesis of an oscillating rather than a tense T state. The concerted character of the allosteric transition is also analysed in view of the apparent dynamics of the conformers.

Antibody BCF2 against scorpion toxin cn2 from Centruroides noxius hoffmann: Primary structure and three-dimensional model as free fv fragment and complexed with its antigen†

The antibody BCF2 generated against the mammal‐specific toxin Cn2 of the scorpion Centruroides noxius Hoffmann neutralizes the effect of both the toxin and the venom. We cloned and sequenced the genes coding for the Fv fragment of BCF2. A three‐dimensional (3D) model of the Fv fragment was generated using a knowledge‐based approach. Furthermore, a 3D model of the complex Cn2–BCF2 was built using the nuclear magnetic resonance (NMR) structure of Cn2 and experimental results on a putative epitope region around the N and C termini. The initial complex conformations were submitted to a new refinement procedure of rigid‐body energy minimization combined with flexible‐side‐chain molecular dynamics. The final complex, selected after an extensive evaluation, uses the loop 7–11 as the central part of the epitope. The generated complex allows the following conclusions: 1) the neutralizing capacity of BCF2 toward the venom of C. noxius might rather be caused by the high venom concentration and toxicity of Cn2 than by a broad specificity, 2) the region involved in the binding of Cn2 to the Na+ channel, should overlap with the employed epitope region, and 3) contact residues SerL91, AsnL92, LeuH50, AspH56, TyrH95, and TyrH98 of BCF2 are candidates for mutations to broaden its specificity. Proteins 1999;37:130–143.