Database for three dimensional structures of pepsin-like enzymes using the internal coordinate system.

A database for 3D structures of pepsin-like enzymes has been created on the basis of a novel approach using the Internal Coordinate System (ICS). It allows rapid comparison of multiple structures of pepsin-like enzymes without the need for preliminary calculations. Atomic displacements measured by this approach are very close to those estimated by the superposition procedures widely employed in comparing three-dimensional structures of proteins. Any new structure of pepsin-like enzyme converted to the ICS automatically becomes superimposed with all structures in a database. The ICS approach can be used for any class of enzymes and is especially efficient for families containing a large number of homologous structures.

Structure of Escherichia coli glutamate decarboxylase (GADalpha) in complex with glutarate at 2.05 angstroms resolution.

Glutamate decarboxylase (GAD) is a pyridoxal enzyme that catalyzes the conversion of L-glutamate into gamma-aminobutyric acid and carbon dioxide. The Escherichia coli enzyme exists as two isozymes, referred to as GADalpha and GADbeta. Crystals of the complex of the recombinant isozyme GADalpha with glutarate as a substrate analogue were grown in space group R3, with unit-cell parameters a = b = 117.1, c = 196.4 angstroms. The structure of the enzyme was solved by the molecular-replacement method and refined at 2.05 angstroms resolution to an R factor of 15.1% (R(free) = 19.9%). The asymmetric unit contains a dimer consisting of two subunits of the enzyme related by a noncrystallographic twofold axis which is perpendicular to and intersects a crystallographic threefold axis. The dimers are related by a crystallographic threefold axis to form a hexamer. The active site of each subunit is formed by residues of the large domains of both subunits of the dimer. The coenzyme pyridoxal phosphate (PLP) forms an aldimine bond with Lys276. The glutarate molecule bound in the active site of the enzyme adopts two conformations with equal occupancies. One of the two carboxy groups of the glutarate occupies the same position in both conformations and forms hydrogen bonds with the N atom of the main chain of Phe63 and the side chain of Thr62 of one subunit and the side chains of Asp86 and Asn83 of the adjacent subunit of the dimer. Apparently, it is in this position that the distal carboxy group of the substrate would be bound by the enzyme, thus providing recognition of glutamic acid by the enzyme.

Cysteines beta93 and beta112 as probes of conformational and functional events at the human hemoglobin subunit interfaces.

Three variants of tetrameric human hemoglobin, with changes at the alpha1beta2/alpha2beta1-interface, at the alpha1beta1/alpha2beta2-interface, and at both interfaces, have been constructed. At alpha1beta2/alpha2beta1-interface the beta93 cysteine was replaced by alanine (betaC93A), and at the alpha1beta1/alpha2beta2-interface the beta112 cysteine was replaced by glycine (betaC112G). The alpha1beta2 interface variant, betaC93A, and the alpha1beta1/alpha1beta2 double mutant, beta(C93A+C112G), were crystallized in the T-state, and the structures determined at 2. 0 and 1.8 A resolution, respectively. A comparison of the structures with that of natural hemoglobin A shows the absence of detectable changes in the tertiary folding of the protein or in the T-state quaternary assembly. At the beta112 site, the void left by the removal of the cysteine side chain is filled by a water molecule, and the functional characteristics of betaC112G are essentially those of human hemoglobin A. At the beta93 site, water molecules do not replace the cysteine side chain, and the alanine substitution increases the conformational freedom of beta146His, weakening the important interaction of this residue with beta94Asp. As a result, when Cl- is present in the solution, at a concentration 100 mM, the Bohr effect of the two mutants carrying the beta93Cys-->Ala substitution, betaC93A and beta(C93A+C112G), is significantly modified being practically absent below pH 7.4. Based on the crystallographic data, we attribute these effects to the competition between beta94Asp and Cl- in the salt link with beta146His in T-state hemoglobin. These results point to an interplay between the betaHis146-betaAsp94 salt bridge and the Cl- in solution regulated by the Cys present at position beta93, indicating yet another role of beta93 Cys in the regulation of hemoglobin function.

Crystallographic, molecular modeling, and biophysical characterization of the valine beta 67 (E11)-->threonine variant of hemoglobin.

The crystal structure of the mutant deoxyhemoglobin in which the beta-globin Val67(E11) has been replaced with threonine [Fronticelli et al. (1993) Biochemistry 32, 1235-1242] has been determined at 2.2 A resolution. Prior to the crystal structure determination, molecular modeling indicated that the Thr67(E11) side chain hydroxyl group in the distal beta-heme pocket forms a hydrogen bond with the backbone carbonyl of His63(E7) and is within hydrogen-bonding distance of the N delta of His63(E7). The mutant crystal structure indicates only small changes in conformation in the vicinity of the E11 mutation confirming the molecular modeling predictions. Comparison of the structures of the mutant beta-subunits and recombinant porcine myoglobin with the identical mutation [Cameron et al. (1993) Biochemistry 32, 13061-13070] indicates similar conformations of residues in the distal heme pocket, but there is no water molecule associated with either of the threonines of the beta-subunits. The introduction of threonine into the distal heme pocket, despite having only small perturbations in the local structure, has a marked affect on the interaction with ligands. In the oxy derivative there is a 2-fold decrease in O2 affinity [Fronticelli et al. (1993) Biochemistry 32, 1235-1242], and the rate of autoxidation is increased by 2 orders of magnitude. In the CO derivative the IR spectrum shows modifications with respect to that of normal human hemoglobin, suggesting the presence of multiple CO conformers. In the nitrosyl derivative an interaction with the O gamma atom of Thr67(E11) is probably responsible for the 10-fold increase in the rate of NO release from the beta-subunits. In the aquomet derivative there is a 6-fold decrease in the rate of hemin dissociation suggesting an interaction of the Fe-coordinated water with the O gamma of Thr67(E11).

Heparin binding by fibronectin module III-13 involves six discontinuous basic residues brought together to form a cationic cradle.

The thirteenth type III domain of fibronectin binds heparin almost as well as fibronectin itself and contains a so-called heparin-binding consensus sequence, Arg6-Arg7-Ala8-Arg9 (residues 1697-1700 in plasma fibronectin). Barkalow and Schwarzbauer (Barkalow, F.J., and Schwarzbauer, J.E. (1991) J. Biol. Chem. 266, 7812-7818) showed that mutation of Arg6-Arg7 in domain III-13 of recombinant truncated fibronectins abolished their ability to bind heparin-Sepharose. However, synthetic peptides containing this sequence have negligible affinity for heparin (Ingham, K.C., Brew, S.A., Migliorini, M. M., and Busby, T.F. (1993) Biochemistry 32, 12548-12553). We generated a three-dimensional model of fibronectin type III-13 based on the structure of a homologous domain from tenascin. The model places Arg23, Lys25, and Arg54 parallel to and in close proximity to the Arg6-Arg7-Ala8-Arg9 motif, suggesting that these residues may also contribute to the heparin-binding site. Domain III-13 and six single-site mutants containing Ser in place of each of the above-mentioned basic residues were expressed in Escherichia coli. All of the purified mutant domains melted reversibly with a Tm near that of the wild type indicating that they were correctly folded. When fluorescein-labeled heparin was titrated at physiological ionic strength, the wild type domain increased the anisotropy in a hyperbolic fashion with a Kd of 5-7 microM, close to that of the natural domain obtained by proteolysis of fibronectin. The R54S mutant bound 3-fold weaker and the remaining mutants bound at least 10-fold weaker than wild type. The results point out that the Arg6-Arg7-Ala8-Arg9 consensus sequence by itself has little affinity for heparin under physiological conditions, even when presented in the context of a folded domain. Thus, the heparin-binding site in fibronectin is more complex than previously realized. It is formed by a cluster of 6 positively charged residues that are remote in the sequence but brought together on one side of domain III-13 to form a "cationic cradle" into which the anionic heparin molecule could fit.

Chloride ion independence of the Bohr effect in a mutant human hemoglobin beta (V1M+H2deleted).

A mutant human hemoglobin, beta (V1M+H2 delta), has been constructed. Analysis of the oxygen binding curves obtained at pH 8.3, where the Bohr effect is inoperative, indicates that this mutation results in an additional stabilization of the T-state conformation by 0.9 kcal/mol. The crystal structure of deoxy-beta (V1M+H2 delta) has been determined to 2.2-A resolution and compared with the deoxy structure of human hemoglobin at the same resolution. In human hemoglobin, a sulfate anion is anchored to the beta-chains by a complex network of H-bonds and electrostatic interactions with the amino terminus and Lys beta 82. In the mutant hemoglobin, the shortening of the amino-terminal region of the A helix by 1 residue results in the formation of an intrachain electrostatic interaction between the amino-terminal amino and Asp beta 79. This eliminates the sulfate binding site, and the sulfate is replaced by two water molecules. At variance with human hemoglobin, the alkaline Bohr effect for beta (V1M+H2 delta) is not sensitive to the presence of Cl-. This indicates that the sulfate binding site in human hemoglobin also serves as a Cl- binding site, and that the amino-terminal Val beta 1 is essential for oxygen-linked Cl- binding to hemoglobin as well as the Cl(-)-dependent Bohr effect. Analysis of the oxygen binding curves indicates that the oxygen-linked Cl- ions are released upon binding of the first oxygen molecule.

Crystallization and preliminary X-ray diffraction data of two heparin-binding fragments of human fibronectin.

Two different heparin-binding fragments of human fibronectin have been crystallized in forms which are suitable for crystal structure analyses. The 30 kDa hep-2A fragment, consisting of type III domains 12-14, was crystallized from solutions containing ammonium sulfate or polyethylene glycol 6000. The crystals grown in ammonium sulfate solutions were orthorhombic with space group I222 or I2(1)2(1)2(1) with a = 68.1 A, b = 88.6 A, and c = 144.9 A. The crystals grown in polyethylene glycol solutions are hexagonal with space group P6(1)22 or P6(5)22 with a = b = 66.7 A and c = 245.7 A. The 40 kDa hep-2B fragment, consisting of type III domains 12-15, was also crystallized from solutions containing ammonium sulfate with the addition of glycerol. Glycerol proved an effective agent for reducing the number of crystals in the crystallization experiments, and thus, increasing the size of the crystals in these experiments. This crystal form is nearly isomorphous to the orthorhombic form of the hep-2A fragment with space group I222 or I2(1)2(1)2(1) and a = 67.5 A, b = 87.0 A, and c = 144.3 A. All crystal forms diffract to at least 3.5 A resolution and contain a single molecule in the asymmetric unit.

Isolation, crystallization and preliminary X-ray diffraction data of human progastricsin.

Human progastricsin, a zymogen of one of the gastric aspartic proteinases, was isolated and crystallized. The crystals belong to the tetragonal space group P4(2)2(1)2, and have unit cell dimensions a = b = 105.5 +/- 0.1 A, c = 70.6 A. The native crystals of progastricsin diffract X-rays at least to 2.5 A and are suitable for a high-resolution X-ray analysis.

[Various aspects of structural studies of aspartate proteinases]. / Nekotorye aspekty strukturnykh issledovanii aspartatnykh proteinaz.

The paper is a brief account of aspartic proteinases' structural studies developed in V.A. Engelhardt Institute of Molecular Biology during the last 3 years. The work on porcine pepsin has been finalized after the refinement of the monoclinic crystal form at 1.8 A resolution performed in collaboration with the group of protein structure and function studies of the University of Alberta in Canada. An important structural property of chymosin which explains the enzyme specificity has been found. Protein engineering work on chymosin is being developed. The structural template for aspartic proteinases has been elucidated and on the basis of this template the model of HIV-1 protease molecule has been built. Some approaches to the design of HIV-1 protease inhibitors were elucidated.

Possible role of some groups in the structure and function of HIV-1 protease as revealed by molecular modeling studies.

Retroviral proteases belong to the class of aspartic proteases. A molecular model of HIV-1 protease has been built on the basis of the consensus template specific for the domains of these enzymes. The template region comprises more than a half of the HIV-1 protease monomer structure, it includes the active site, formed at the junction of the two monomers, binding pockets of the enzyme, and some other molecular segments. These regions can be more conveniently described than other parts of the structure. Some properties of the HIV-1 protease molecule are discussed, as well as of probable inhibitors. The properties of the model structure are in good agreement with the recent results of crystallographic studies of Rous sarcoma virus protease.