Structural analysis of variant of Helicobacter pylori MotB in its activated form, engineered as chimera of MotB and leucine zipper

Rotation of the bacterial flagellum is powered by a proton influx through the peptidoglycan (PG)-tethered stator ring MotA/B. MotA and MotB form an inner-membrane complex that does not conduct protons and does not bind to PG until it is inserted into the flagellar motor. The opening of the proton channel involves association of the plug helices in the periplasmic region of the MotB dimer into a parallel coiled coil. Here, we have characterised the structure of a soluble variant of full-length Helicobacter pylori MotB in which the plug helix was engineered to be locked in a parallel coiled coil state, mimicking the open state of the stator. Fluorescence resonance energy transfer measurements, combined with PG-binding assays and fitting of the crystal structures of MotB fragments to the small angle X-ray scattering (SAXS) data revealed that the protein's C-terminal domain has a PG-binding-competent conformation. Molecular modelling against the SAXS data suggested that the linker in H. pylori MotB forms a subdomain between the plug and the C-terminal domain, that 'clamps' the coiled coil of the plug, thus stabilising the activated form of the protein. Based on these results, we present a pseudo-atomic model structure of full-length MotB in its activated form.

Design, purification and characterization of a soluble variant of the integral membrane protein MotB for structural studies

The bacterial flagellar motor is an intricate nanomachine powered by a transmembrane electrochemical gradient. Rotation is driven by the cumulative action of several peptidoglycan-anchored stator complexes on the rotor. In proton-motive force-driven motors, the stator complex is composed of a motility protein B (MotB) dimer surrounded by four copies of MotA, where both MotA and MotB are integral membrane proteins. The lack of full-length MotA and MotB structures hinders understanding of the mechanism of torque generation. Given the low levels of expression and low stability of detergent-solubilized MotB, a soluble chimaeric variant was engineered, where the two transmembrane helices of the MotB dimer were replaced by a leucine zipper. The biochemical and biophysical analysis of the resultant protein showed that it was properly folded, stable, behaved as a monodisperse dimer at low pH, had molecular dimensions close to those expected for native MotB and yielded reproducible crystals. The chimaeric protein is, therefore, a good candidate for structural studies. This 'solubilization by design' approach may be generally applicable to the production of soluble forms of other dimeric, trimeric and tetrameric single-span membrane proteins for functional and structural studies.

Two distinct regions of latency-associated peptide coordinate stability of the latent transforming growth factor-beta1 complex

Transforming growth factor-beta1 (TGF-beta1) is secreted as part of an inactive complex consisting of the mature dimer, the TGF-beta1 propeptide (latency-associated peptide (LAP)), and latent TGF-beta-binding proteins. Using in vitro mutagenesis, we identified the regions of LAP that govern the cooperative assembly and stability of the latent TGF-beta1 complex. Initially, hydrophobic LAP residues (Ile(53), Leu(54), Leu(57), and Leu(59)), which form a contiguous epitope on one surface of an amphipathic alpha-helix, interact with mature TGF-beta1 to form the small latent complex. TGF-beta1 binding is predicted to alter LAP conformation, exposing ionic residues (Arg(45), Arg(50), Lys(56), and Arg(58)) on the other side of the alpha-helix, which form the binding site for latent TGF-beta-binding proteins. The stability of the resultant large latent complex is dependent upon covalent dimerization of LAP, which is facilitated by key residues (Phe(198), Asp(199), Val(200), Leu(208), Phe(217), and Leu(219)) at the dimer interface. Significantly, genetic mutations in LAP (e.g. R218H) that cause the rare bone disorder Camurati-Engelmann disease disrupted dimerization and reduced the stability of the latent TGF-beta1 complex.

14-3-3:Shc scaffolds integrate phosphoserine and phosphotyrosine signaling to regulate phosphatidylinositol 3-kinase activation and cell survival

Integrated cascades of protein tyrosine and serine/threonine phosphorylation play essential roles in transducing signals in response to growth factors and cytokines. How adaptor or scaffold proteins assemble signaling complexes through both phosphotyrosine and phosphoserine/threonine residues to regulate specific signaling pathways and biological responses is unclear. We show in multiple cell types that endogenous 14-3-3zeta is phosphorylated on Tyr(179) in response to granulocyte macrophage colony-stimulating factor. Importantly, 14-3-3zeta can function as an intermolecular bridge that couples to phosphoserine residues and also directly binds the SH2 domain of Shc via Tyr(179). The assembly of these 14-3-3:Shc scaffolds is specifically required for the recruitment of a phosphatidylinositol 3-kinase signaling complex and the regulation of CTL-EN cell survival in response to cytokine. The biological significance of these findings was further demonstrated using primary bone marrow-derived mast cells from 14-3-3zeta(-/-) mice. We show that cytokine was able to promote Akt phosphorylation and viability of primary mast cells derived from 14-3-3zeta(-/-) mice when reconstituted with wild type 14-3-3zeta, but the Akt phosphorylation and survival response was reduced in cells reconstituted with the Y179F mutant. Together, these results show that 14-3-3:Shc scaffolds can act as multivalent signaling nodes for the integration of both phosphoserine/threonine and phosphotyrosine pathways to regulate specific cellular responses.

A common biosynthetic pathway governs the dimerization and secretion of inhibin and related transforming growth factor beta (TGFbeta) ligands

The assembly and secretion of transforming growth factor beta superfamily ligands is dependent upon non-covalent interactions between their pro- and mature domains. Despite the importance of this interaction, little is known regarding the underlying regulatory mechanisms. In this study, the binding interface between the pro- and mature domains of the inhibin alpha-subunit was characterized using in vitro mutagenesis. Three hydrophobic residues near the N terminus of the prodomain (Leu(30), Phe(37), Leu(41)) were identified that, when mutated to alanine, disrupted heterodimer assembly and secretion. It is postulated that these residues mediate dimerization by interacting non-covalently with hydrophobic residues (Phe(271), Ile(280), Pro(283), Leu(338), and Val(340)) on the outer convex surface of the mature alpha-subunit. Homology modeling indicated that these mature residues are located at the interface between two beta-sheets of the alpha-subunit and that their side chains form a hydrophobic packing core. Mutation of these residues likely disturbs the conformation of this region, thereby disrupting non-covalent interactions with the prodomain. A similar hydrophobic interface was identified spanning the pro- and mature domains of the inhibin beta(A)-subunit. Mutation of key residues, including Ile(62), Leu(66), Phe(329), and Pro(341), across this interface was disruptive for the production of both inhibin A and activin A. In addition, mutation of Ile(62) and Leu(66) in the beta(A)-propeptide reduced its ability to bind, or inhibit the activity of, activin A. Conservation of the identified hydrophobic motifs in the pro- and mature domains of other transforming growth factor beta superfamily ligands suggests that we have identified a common biosynthetic pathway governing dimer assembly.

T cell receptor recognition of a 'super-bulged' major histocompatibility complex class I-bound peptide

Unusually long major histocompatibility complex (MHC) class I-restricted epitopes are important in immunity, but their 'bulged' conformation represents a potential obstacle to alphabeta T cell receptor (TCR)-MHC class I docking. To elucidate how such recognition is achieved while still preserving MHC restriction, we have determined here the structure of a TCR in complex with HLA-B(*)3508 presenting a peptide 13 amino acids in length. This complex was atypical of TCR-peptide-MHC class I interactions, being dominated at the interface by peptide-mediated interactions. The TCR assumed two distinct orientations, swiveling on top of the centrally bulged, rigid peptide such that only limited contacts were made with MHC class I. Although the TCR-peptide recognition resembled an antibody-antigen interaction, the TCR-MHC class I contacts defined a minimal 'generic footprint' of MHC-restriction. Thus our findings simultaneously demonstrate the considerable adaptability of the TCR and the 'shape' of MHC restriction.

The crystal structures of glutathione S-transferases isozymes 1-3 and 1-4 from Anopheles dirus species B

Glutathione S-transferases (GSTs) are dimeric proteins that play an important role in cellular detoxification. Four GSTs from the mosquito Anopheles dirus species B (Ad), an important malaria vector in South East Asia, are produced by alternate splicing of a single transcription product and were previously shown to have detoxifying activity towards pesticides such as DDT. We have determined the crystal structures for two of these alternatively spliced proteins, AdGST1-3 (complexed with glutathione) and AdGST1-4 (apo form), at 1.75 and 2.45 A resolution, respectively. These GST isozymes show differences from the related GST from the Australian sheep blowfly Lucilia cuprina; in particular, the presence of a C-terminal helix forming part of the active site. This helix causes the active site of the Anopheles GSTs to be enclosed. The glutathione-binding helix alpha2 and flanking residues are disordered in the AdGST1-4 (apo) structure, yet ordered in the AdGST1-3 (GSH-bound) structure, suggesting that insect GSTs operate with an induced fit mechanism similar to that found in the plant phi- and human pi-class GSTs. Despite the high overall sequence identities, the active site residues of AdGST1-4 and AdGST1-3 have different conformations.

Structural biology and its applications to the health sciences

Part of the decipherment of genomic information lies in understanding the structure and function of the protein products of these genes. Protein structure is of further importance because of the molecular basis of many diseases. Structural biology is the field of research focusing on the experimental determination of the structure of biological molecules. We review the field of structural biology and its application to medical research and drug discovery, and describe the structural results recently obtained in our laboratory for the detoxifying enzyme glutathione S-transferase from the Asian mosquito Anopheles dirus species B, an important malaria vector. These enzymes have detoxifying activity toward pesticides and thus contribute to pesticide resistance in insects.

Large conformational changes of the epsilon subunit in the bacterial F1F0 ATP synthase provide a ratchet action to regulate this rotary motor enzyme

The F(1)F(0) ATP synthase is the smallest motor enzyme known. Previous studies had established that the central stalk, made of the gamma and epsilon subunits in the F(1) part and c subunit ring in the F(0) part, rotates relative to a stator composed of alpha(3)beta(3)deltaab(2) during ATP hydrolysis and synthesis. How this rotation is regulated has been less clear. Here, we show that the epsilon subunit plays a key role by acting as a switch of this motor. Two different arrangements of the epsilon subunit have been visualized recently. The first has been observed in beef heart mitochondrial F(1)-ATPase where the C-terminal portion is arranged as a two-alpha-helix hairpin structure that extends away from the alpha(3)beta(3) region, and toward the position of the c subunit ring in the intact F(1)F(0). The second arrangement was observed in a structure determination of a complex of the gamma and epsilon subunits of the Escherichia coli F(1)-ATPase. In this, the two C-terminal helices are apart and extend along the gamma to interact with the alpha and beta subunits in the intact complex. We have been able to trap these two arrangements by cross-linking after introducing appropriate Cys residues in E. coli F(1)F(0), confirming that both conformations of the epsilon subunit exist in the enzyme complex. With the C-terminal domain of epsilon toward the F(0), ATP hydrolysis is activated, but the enzyme is fully coupled in both ATP hydrolysis and synthesis. With the C-terminal domain toward the F(1) part, ATP hydrolysis is inhibited and yet the enzyme is fully functional in ATP synthesis; i.e., it works in one direction only. These results help explain the inhibitory action of the epsilon subunit in the F(1)F(0) complex and argue for a ratchet function of this subunit.

Crystallization of two glutathione S-transferases from an unusual gene family

Two glutathione S-transferase isozymes from the mosquito Anopheles dirus (AdGST1-3 and AdGST1-4) from an alternately spliced gene family have been expressed, purified and crystallized. The isozymes share an N-terminal domain derived from a single exon and C-terminal domains from unique exons. Despite the high level of sequence identity (64% overall), the two isozymes crystallize in different space groups, the 1-3 isozyme in P3(1)21 or P3(2)21 (unit-cell parameters a = 49.9, c = 271.8 A at 100 K) and the 1-4 isozyme in P4(1) or P4(3) (unit-cell parameters a = 87.8, c = 166.1 at 100 K). Determination of these structures will advance our understanding of how these enzymes inactivate pesticides and the structural consequences of alternate splicing.

Structure of the RTP-DNA complex and the mechanism of polar replication fork arrest

The coordinated termination of DNA replication is an important step in the life cycle of bacteria with circular chromosomes, but has only been defined at a molecular level in two systems to date. Here we report the structure of an engineered replication terminator protein (RTP) of Bacillus subtilis in complex with a 21 base pair DNA by X-ray crystallography at 2.5 A resolution. We also use NMR spectroscopic titration techniques. This work reveals a novel DNA interaction involving a dimeric 'winged helix' domain protein that differs from predictions. While the two recognition helices of RTP are in close contact with the B-form DNA major grooves, the 'wings' and N-termini of RTP do not form intimate contacts with the DNA. This structure provides insight into the molecular basis of polar replication fork arrest based on a model of cooperative binding and differential binding affinities of RTP to the two adjacent binding sites in the complete terminator.

Crystal structure of auracyanin, a "blue" copper protein from the green thermophilic photosynthetic bacterium Chloroflexus aurantiacus

Auracyanin B, one of two similar blue copper proteins produced by the thermophilic green non-sulfur photosynthetic bacterium Chloroflexus aurantiacus, crystallizes in space group P6(4)22 (a=b=115.7 A, c=54.6 A). The structure was solved using multiple wavelength anomalous dispersion data recorded about the CuK absorption edge, and was refined at 1.55 A resolution. The molecular model comprises 139 amino acid residues, one Cu, 247 H(2)O molecules, one Cl(-) and two SO(4)(2-). The final residual and estimated standard uncertainties are R=0.198, ESU=0.076 A for atomic coordinates and ESU=0.05 A for Cu---ligand bond lengths, respectively. The auracyanin B molecule has a standard cupredoxin fold. With the exception of an additional N-terminal strand, the molecule is very similar to that of the bacterial cupredoxin, azurin. As in other cupredoxins, one of the Cu ligands lies on strand 4 of the polypeptide, and the other three lie along a large loop between strands 7 and 8. The Cu site geometry is discussed with reference to the amino acid spacing between the latter three ligands. The crystallographically characterized Cu-binding domain of auracyanin B is probably tethered to the periplasmic side of the cytoplasmic membrane by an N-terminal tail that exhibits significant sequence identity with known tethers in several other membrane-associated electron-transfer proteins.

Structure of the gamma-epsilon complex of ATP synthase

ATP synthases (F(1)F(o)-ATPases) use energy released by the movement of protons down a transmembrane electrochemical gradient to drive the synthesis of ATP, the universal biological energy currency. Proton flow through F(o) drives rotation of a ring of c-subunits and a complex of the gamma and epsilon-subunits, causing cyclical conformational changes in F(1) that are required for catalysis. The crystal structure of a large portion of F(1) has been resolved. However, the structure of the central portion of the enzyme, through which conformational changes in F(o) are communicated to F(1), has until now remained elusive. Here we report the crystal structure of a complex of the epsilon-subunit and the central domain of the gamma-subunit refined at 2.1 A resolution. The structure reveals how rotation of these subunits causes large conformational changes in F(1), and thereby provides new insights into energy coupling between F(o) and F(1).

Macromolecular crystallography as a tool for investigating drug, enzyme and receptor interactions

1. Protein crystallography is an essential tool for the discovery and investigation of pharmacological interactions at the molecular level. It allows investigators to directly visualize the three-dimensional structures of proteins, including enzymes, receptors and hormones. 2. Increasingly, knowledge of these interactions is being used in the drug-discovery process. This is popularly called structure-based drug design. The desired drug could be an enzyme inhibitor or an agonist that mimics endogenous transmitters or hormones. 3. Once the 3-D structure of a pharmacologically relevant target is known, computational processes can be used to search databases of compounds to identify ones that may interact strongly with the target. Lead compounds can be improved using the 3-D structure of the complex of the lead compound and its biological target. 4. The present review describes the processes involved in the determination of a structure by means of protein crystallography and the use of structures in the drug-discovery process. A number of successful examples of structure-based drug design are described. The limitations of the techniques are discussed.

Structure and function of the bacterial mechanosensitive channel of large conductance

Mechanosensation in bacteria involves transducing membrane stress into an electrochemical response. In Escherichia coli and other bacteria, this function is carried out by a number of proteins including MscL, the mechanosensitive channel of large conductance. MscL is the best characterized of all mechanosensitive channels. It has been the subject of numerous structural and functional investigations. The explosion in experimental data on MscL recently culminated in the solution of the three-dimensional structure of the MscL homologue from Mycobacterium tuberculosis. In this review, much of these data are united and interpreted in terms of the newly published M. tuberculosis MscL crystal structure.

Rubredoxin from Clostridium pasteurianum. Structures of G10A, G43A and G10VG43A mutant proteins. Mutation of conserved glycine 10 to valine causes the 9-10 peptide link to invert

The four cysteine ligands which coordinate the Fe atom in the electron-transfer protein rubredoxin lie on loops of the polypeptide which form approximate local twofold symmetry. The cysteine ligands in the protein from Clostridium pasteurianum lie at positions 6, 9, 39 and 42. Two glycine residues adjacent to the cysteine ligands at positions 10 and 43 are conserved in all rubredoxins, consistent with the proposal that a beta-carbon substituent at these positions would eclipse adjacent peptide carbonyl groups [Adman et al. (1975). Proc. Natl Acad. Sci. USA, 72, 4854-4858]. X-ray crystal structures of the three mutant proteins G10A, G43A and G10VG43A are reported. The crystal structures of the single-site mutations are isomorphous with the native protein, space group R3; unit-cell parameters are a = 64.3, c = 32.9 A for G10A and a = 64.4, c = 32.8 A for G43A. The crystals of the double mutant, G10VG43A, were in space group P43212, unit-cell parameters a = 61.9, c = 80.5 A, with two molecules per asymmetric unit. The observed structural perturbations support the hypothesis that mutation of the conserved glycine residues would introduce strain into the polypeptide. In particular, in the G10VG43A protein substitution of valine at Gly10 causes the 9-10 peptide link to invert, relieving steric interaction between Cys9 O and Val10 Cbeta. This dramatic change in conformation is accompanied by the loss of the 10N-HcO6 hydrogen bond, part of the chelate loop Thr5-Tyr11. The new conformation allows retention of the 11N-HcS9 hydrogen bond, but converts it from a type II to a type I hydrogen bond. This occurs at the cost of a less tightly packed structure. The structural insights allow rationalization of 1H NMR data reported previously for the 113CdII-substituted proteins and of the negative shifts observed in the FeIII/FeII mid-point potentials upon mutation.

The structure of plastocyanin from the cyanobacterium Phormidium laminosum

The crystal structure of the 'blue' copper protein plastocyanin from the cyanobacterium Phormidium laminosum has been solved and refined using 2.8 A X--ray data. P. laminosum plastocyanin crystallizes in space group P43212 with unit-cell dimensions a = 86.57, c = 91.47 A and with three protein molecules per asymmetric unit. The final residual R is 19.9%. The structure was solved using molecular replacement with a search model based on the crystal structure of a close homologue, Anabaena variabilis plastocyanin (66% sequence identity). The molecule of P. laminosum plastocyanin has 105 amino-acid residues. The single Cu atom is coordinated by the same residues - two histidines, a cysteine and a methionine - as in other plastocyanins. In the crystal structure, the three molecules of the asymmetric unit are related by a non-crystallographic threefold axis. A Zn atom lies between each pair of neighbouring molecules in this ensemble, being coordinated by a surface histidine residue of one molecule and by two aspartates of the other.

Structure and mechanism of a proline-specific aminopeptidase from Escherichia coli

The structure of the proline-specific aminopeptidase (EC 3.4.11.9) from Escherichia coli has been solved and refined for crystals of the native enzyme at a 2.0-A resolution, for a dipeptide-inhibited complex at 2.3-A resolution, and for a low-pH inactive form at 2.7-A resolution. The protein crystallizes as a tetramer, more correctly a dimer of dimers, at both high and low pH, consistent with observations from analytical ultracentrifuge studies that show that the protein is a tetramer under physiological conditions. The monomer folds into two domains. The active site, in the larger C-terminal domain, contains a dinuclear manganese center in which a bridging water molecule or hydroxide ion appears poised to act as the nucleophile in the attack on the scissile peptide bond of Xaa-Pro. The metal-binding residues are located in a single subunit, but the residues surrounding the active site are contributed by three subunits. The fold of the protein resembles that of creatine amidinohydrolase (creatinase, not a metalloenzyme). The C-terminal catalytic domain is also similar to the single-domain enzyme methionine aminopeptidase that has a dinuclear cobalt center.

Crystal structures of the copper-containing amine oxidase from Arthrobacter globiformis in the holo and apo forms: implications for the biogenesis of topaquinone

The crystal structures of the copper enzyme phenylethylamine oxidase from the Gram-positive bacterium Arthrobacter globiformis (AGAO) have been determined and refined for three forms of the enzyme: the holoenzyme in its active form (at 2.2 A resolution), the holoenzyme in an inactive form (at 2.8 A resolution), and the apoenzyme (at 2.2 A resolution). The holoenzyme has a topaquinone (TPQ) cofactor formed from the apoenzyme by the post-translational modification of a tyrosine residue in the presence of Cu2+. Significant differences between the three forms of AGAO are limited to the active site. The polypeptide fold is closely similar to those of the amine oxidases from Escherichia coli [Parsons, M. R., et al. (1995) Structure 3, 1171-1184] and pea seedlings [Kumar, V., et al. (1996) Structure 4, 943-955]. In the active form of holo-AGAO, the active-site Cu atom is coordinated by three His residues and two water molecules in an approximately square-pyramidal arrangement. In the inactive form, the Cu atom is coordinated by the same three His residues and by the phenolic oxygen of the TPQ, the geometry being quasi-trigonal-pyramidal. There is evidence of disorder in the crystals of both forms of holo-AGAO. As a result, only the position of the aromatic group of the TPQ cofactor, but not its orientation about the Cbeta-Cgamma bond, is determined unequivocally. In apo-AGAO, electron density consistent with an unmodified Tyr occurs at a position close to that of the TPQ in the inactive holo-AGAO. This observation has implications for the biogenesis of TPQ. Two features which have not been described previously in amine oxidase structures are a channel from the molecular surface to the active site and a solvent-filled cavity at the major interface between the two subunits of the dimer.

Crystallization, structural determination and analysis of a novel parasite vaccine candidate: Fasciola hepatica glutathione S-transferase

Glutathione S-transferases (GSTs) represent the major class of detoxifying enzymes from parasitic helminths. As a result, they are candidates for chemotherapeutic and vaccine design. Indeed, GSTs from Fasciola hepatica have been found to be effective for vaccinating sheep and cattle against fasciolosis. This helminth contains at least seven GST isoforms, of which four have been cloned. The cloned isoforms (Fh51, Fh47, Fh7 and Fh1) all belong to the mu class of GSTs, share greater than 71% sequence identity, yet display distinct substrate specificities. Crystals of Fh47 were obtained using the hanging drop vapour diffusion technique. The crystals belong to space group I4122, with one monomer in the asymmetric unit, which corresponds to a very high solvent content of approximately 75%. The physiological dimer is generated via a crystallographic 2-fold rotation. The three-dimensional structure of Fh47 was solved by molecular replacement using the Schistosoma japonicum glutathione S-transferase (Sj26) crystal structure as a search model. The structure adopts the canonical GST fold comprising two domains: an N-terminal glutathione-binding domain, consisting of a four-stranded beta-sheet and three helices whilst the C-terminal domain is entirely alpha-helical. The presence of Phe19 in Fh47 results in a 6 degrees interdomain rotation in comparison to Sj26, where the equivalent residue is a leucine. Homology models of Fh51, Fh7 and Fh1, based on the Fh47 crystal structure, reveal critical differences in the residues lining the xenobiotic binding site, particularly at residue positions 9, 106 and 204. In addition, differences amongst the isoforms in the non-substrate binding site were noted, which may explain the observed differential binding of large ligands. The major immunogenic epitopes of Fh47 were surprisingly found not to reside on the most solvent-exposed regions of the molecule.

Homology model for the human GSTT2 Theta class glutathione transferase

A tertiary model of the human GSTT2 Theta class glutathione transferase is presented based on the recently solved crystal structure of a related thetalike isoenzyme from Lucilia cuprina. Although the N-terminal domains are quite homologous, the C-terminal domains share less than about 20% identity. The model is used to consolidate the role of Ser 11 in the active site of the enzyme as well as to identify other residues and mechanisms of likely catalytic importance. The T2 subfamily of theta class enzymes have been shown to inactivate reactive sulfate esters arising from arylmethanols. A possible reaction pathway involving the conjugation of glutathione with one such sulfate ester, 1-menaphthyl-sulfate, is described. It is also proposed that the C-terminal region of the enzyme plays an important role in allowing substrate access to the active site.

Crystal structure of a eukaryotic (pea seedling) copper-containing amine oxidase at 2.2 A resolution

BACKGROUND

Copper-containing amine oxidases catalyze the oxidative deamination of primary amines to aldehydes, in a reaction that requires free radicals. These enzymes are important in many biological processes, including cell differentiation and growth, would healing, detoxification and signalling. The catalytic reaction requires a redox cofactor, topa quinone (TPQ), which is derived by post-translational modification of an invariant tyrosine residue. Both the biogenesis of the TPQ cofactor and the reaction catalyzed by the enzyme require the presence of a copper atom at the active site. The crystal structure of a prokaryotic copper amine oxidase from E. coli (ECAO) has recently been reported.

RESULTS

The first structure of a eukaryotic (pea seedling) amine oxidase (PSAO) has been solved and refined at 2.2 A resolution. The crystallographic phases were derived from a single phosphotungstic acid derivative. The positions of the tungsten atoms in the W12 clusters were obtained by molecular replacement using E. coli amine oxidase as a search model. The methodology avoided bias from the search model, and provides an essentially independent view of a eukaryotic amine oxidase. The PSAO molecule is a homodimer; each subunit has three domains. The active site of each subunit lies near an edge of the beta-sandwich of the largest domain, but is not accessible from the solvent. The essential active-site copper atom is coordinated by three histidine side chains and two water molecules in an approximately square-pyramidal arrangement. All the atoms of the TPQ cofactor are unambiguously defined, the shortest distance to the copper atom being approximately 6 A.

CONCLUSIONS

There is considerable structural homology between PSAO and ECAO. A combination of evidence from both structures indicates that the TPQ side chain is sufficiently flexible to permit the aromatic grouf to rotate about the Cbeta-Cgamma bond, and to move between bonding and non-bonding positions with respect to the Cu atom. Conformational flexibility is also required at the surface of the molecule to allow the substrates access to the active site, which is inaccessible to solvent, as expected for an enzyme that uses radical chemistry.

A structurally derived consensus pattern for theta class glutathione transferases

We have recently determined the first crystal structure of a theta class glutathione transferase. We have aligned the amino acid sequences of members of the family using the crystal structure as a guide. The alignment has revealed a consensus pattern of residues that first identifies a protein as belonging to the glutathione transferase superfamily, and second is able to distinguish theta class members from other classes of glutathione transferases. The consensus residues unique to the theta class are found to cluster mostly on the hydrophilic surface and flanking loops of helix 2, a region found to be structurally diverse amongst crystal structures of the different glutathione transferase classes. When the consensus pattern was scanned against sequence databases, a number of matches were made with proteins not formally identified as glutathione transferases. Some of these matches indicated that several stress-related proteins belong to the theta class GST family.

Crystallization and preliminary X-ray analysis of a bacterial glutathione transferase

Crystals of a bacterial glutathione S-transferase from Proteus mirabilis have been grown from polyethylene glycol by the hanging-drop vapour-diffusion method. Successful crystallization required the presence of the substrate glutathione. The crystals belong to the tetragonal space group P4 with cell dimensions a = b = 90.9 and c = 117.3 A. They contain between three and six monomers in the asymmetric unit and diffract to beyond 2.3 A resolution.

Evidence for an essential serine residue in the active site of the Theta class glutathione transferases

A consistent feature of the Alpha-, Mu- and Pi-class glutathione transferases (GSTs) is the presence near the N-terminus of a tyrosine residue that contributes to the activation of glutathione. While this residue appears to be conserved in many Theta-class GSTs, its absence in some suggested that the Theta-class GSTs may have a significantly different structure or catalytic mechanism. The elucidation of the crystal structure of the Theta-class GST from the Australian sheep blowfly, Lucilia cuprina, has indicated that a serine residue rather than a tyrosine residue can form a hydrogen bond with the glutathionyl sulphur atom. The present studies show that mutation of Ser-9 to alanine substantially inactivates the L. cuprina GST, confirming its importance in the reaction mechanism. As this serine is conserved in all Theta-class enzymes reported so far, it seems that an active-site serine is a significant factor that distinguishes the Theta-class GSTs from members of the Alpha-, Mu- and Pi-class isoenzymes.

Crystal structure of a theta-class glutathione transferase

Glutathione S-transferases (GSTs) are a family of enzymes involved in the cellular detoxification of xenotoxins. Cytosolic GSTs have been grouped into four evolutionary classes for which there are representative crystal structures of three of them. Here we report the first crystal structure of a theta-class GST. So far, all available GST crystal structures suggest that a strictly conserved tyrosine near the N-terminus plays a critical role in the reaction mechanism and such a role has been convincingly demonstrated by site-directed mutagenesis. Surprisingly, the equivalent residue in the theta-class structure is not in the active site, but its role appears to have been replaced by either a nearby serine or by another tyrosine residue located in the C-terminal domain of the enzyme.

Insights into autoregulation from the crystal structure of twitchin kinase

Many protein kinases are self-regulated by an intrasteric mechanism where part of the enzyme's structure directly inhibits the active site. This inhibitory structure is called a pseudosubstrate and specific regulators are required to remove it from the active site to allow substrates access. Removal of the pseudosubstrate sequence from members of the myosin light-chain kinase subfamily, including twitchin kinase, activates them but it is not known whether the pseudosubstrate sequence binds to the active site. Native twitchin is a 753K protein (6,839 residues) located in muscle A-bands of the nematode Caenorhabditis elegans and because of its size has not been easy to study. We have determined the crystal structure, refined to 2.8 A resolution, of a recombinant fragment (residues 5,890 to 6,262) of twitchin kinase that contains the catalytic core and a 60 residue carboxy-terminal tail. The C-terminal tail extends through the active site, wedged between the small and large lobes of the structure and making extensive contacts with the catalytic core which accounts for autoinhibition and provides direct support for the intrasteric mechanism of protein kinase regulation.

Crystallization and preliminary X-ray diffraction studies of a glutathione S-transferase from the Australian sheep blowfly, Lucilia cuprina

Crystals of a glutathione S-transferase from the Australian sheep blowfly Lucilia cuprina have been grown from ammonium sulphate by the hanging drop vapour diffusion method. Successful crystallization required the presence of the inhibitor S-hexylglutathione. The crystals belong to the tetragonal space group P4(1)22 (or P4(3)22) with cell dimensions of a = b = 88.1 A and c = 66.9 A. They contain one monomer in the asymmetric unit and diffract beyond 2.8 A resolution.

Crystallization and preliminary X-ray analysis of the auto-inhibited twitchin kinase

An auto-inhibited fragment of twitchin kinase (residues 5890 to 6262) has been crystallized by vapor diffusion techniques using polyethylene glycol 4000 as the precipitant at pH 7.25 to 7.5 at 4 degrees C. We have found that MgSO4 and glycerol were essential for large crystal growth. The crystals belong to the orthorhombic space group P2(1)2(1)2, with unit cell dimensions of a = 144.1 A, b = 168.3 A and c = 60.6 A. They are suitable for X-ray analysis and diffract to a resolution of at least 2.8 A.

Vaccination of sheep against Fasciola hepatica with glutathione S-transferase. Identification and mapping of antibody epitopes on a three-dimensional model of the antigen

The glutathione S-transferases (FhGST) of the liver fluke Fasciola hepatica have been identified as novel vaccine candidates that protect sheep against a fluke infection. With the use of overlapping peptides covering the predicted amino acid sequences of four FhGST cDNAs, we have defined the linear epitopes recognized by polyclonal antibody from sheep vaccinated with FhGST. Dominant and minor epitopes were found to be present on all four of the sequences although some epitopes were shown to be specific to particular FhGST. A high percentage of the FhGST peptides were found to be antigenic although considerable variability in response to the peptides was observed among the animals. This analysis was extended to the IgG1 and IgG2 response at the peptide level. Based on the recently solved crystal structure of the rat mu-class GST 3-3, a three-dimensional model of one of the FhGST sequences was generated that allowed the predicted spatial localization of defined epitopes. Most epitopes were localized on regions of high flexibility and accessibility. A comparison of epitopes on FhGST with the B cell epitopes on Sm28, a 28-kDa GST from Schistosoma mansoni, has found few similarities. There was no correlation between an antibody response to linear peptide epitopes and the level of protection induced in sheep by vaccination with FhGST.

Vaccination of sheep against Fasciola hepatica with glutathione S-transferase. Identification and mapping of antibody epitopes on a three-dimensional model of the antigen

The glutathione S-transferases (FhGST) of the liver fluke Fasciola hepatica have been identified as novel vaccine candidates that protect sheep against a fluke infection. With the use of overlapping peptides covering the predicted amino acid sequences of four FhGST cDNAs, we have defined the linear epitopes recognized by polyclonal antibody from sheep vaccinated with FhGST. Dominant and minor epitopes were found to be present on all four of the sequences although some epitopes were shown to be specific to particular FhGST. A high percentage of the FhGST peptides were found to be antigenic although considerable variability in response to the peptides was observed among the animals. This analysis was extended to the IgG1 and IgG2 response at the peptide level. Based on the recently solved crystal structure of the rat mu-class GST 3-3, a three-dimensional model of one of the FhGST sequences was generated that allowed the predicted spatial localization of defined epitopes. Most epitopes were localized on regions of high flexibility and accessibility. A comparison of epitopes on FhGST with the B cell epitopes on Sm28, a 28-kDa GST from Schistosoma mansoni, has found few similarities. There was no correlation between an antibody response to linear peptide epitopes and the level of protection induced in sheep by vaccination with FhGST.

Conformational states of human placental glutathione transferase as probed by limited proteolysis

Limited proteolysis experiments have been carried out on human placental glutathione transferase in its different forms. The reduced enzyme, as well as the oxidized form and that inactivated with cystamine were all sensitive to 10% (w/w) trypsin, under nondenaturing conditions. The proteolytic cleavage was accompanied by a concomitant loss of enzymatic activity. On the contrary, the presence of glutathione or glutathione conjugates strongly protected the reduced enzyme against inactivation and from the proteolytic attack. Sodium dodecyl sulphate-polyacrylamide gel electrophoresis and peptide sequence analysis showed that only the peptide bond between Lys 44 and Ala 45 was cleaved. Since Lys 44 has been demonstrated to be involved in the glutathione binding, it is suggested that the region surrounding this amino acid residue (alpha B helix) could be more exposed to the solvent, in the absence of glutathione. Crystallographic data also indicated that this region is flexible, supporting the idea that it may be involved in the observed conformational change upon glutathione binding.

X-ray diffraction studies of fibrils formed from peptide fragments of transthyretin

Two synthetic peptide fragments of the plasma protein transthyretin (TTR), previously shown to form fibrillar structures in vitro, have been examined using electron microscopy and X-ray diffraction. The fibrils displayed all characteristics of cross beta-sheet conformation with antiparallel strand spacing of 4.7 A and intersheet spacings of 8-10 A as well as reflections indicating further lateral repeating units. A third peptide containing a substitution equivalent to a mutation in TTR known to increase the propensity of TTR to form amyloid was also examined. It also formed fibrils and showed similar cross beta-sheet structure, but with closer intersheet packing than its native equivalent.

High-performance liquid chromatography of amino acids, peptides and proteins. CXXII. Application of experimentally derived retention coefficients to the prediction of peptide retention times: studies with myohemerythrin

Amino acid retention coefficients were derived from the experimental retention data of 118 overlapping peptide heptamers related to the primary amino acid sequence of myohemerythrin. Individual retention coefficient values for each amino acid were derived by a multiple linear regression matrix approach. Retention data were derived for five different experimental conditions including different organic modifiers (acetonitrile, methanol or 2-propanol), different mobile phase additives (trifluoroacetic acid or potassium phosphate) and different silica-based stationary phase ligands (octadecyl or phenyl groups). A high degree of correlation was observed between these experimentally derived amino acid coefficients (EXP) and the coefficients (LIT) which we recently reported derived from the retention data of over 2000 peptides [M. C. J. Wilce et al., J. Chromatogr., 536 (1991) 165 and 548 (1991) 105]. These results demonstrated that the LIT and EXP coefficients can be used for the prediction of the retention of any peptide set. The effect of peptide length was also further investigated and the correlation results demonstrated the importance of peptide flexibility on the final value of the amino acid coefficient.

Duchenne muscular dystrophy and dystrophin: sequence homology observations

Duchenne muscular dystrophy (DMD) is a genetically transmitted disease characterized by progressive muscle weakness and usually leads to death. DMD results from the absence, deficiency or dysfunction of the protein dystrophin. Analysis of protein data bases, including homology alignments and domain recognition patterns, have located highly significant correlations between dystrophin and other calcium regulating proteins. In particular, a major portion of the dystrophin sequence has been found to contain repeating units of approximately 100 amino acid residues. These repeating units were found to exhibit significant homology to troponin I. Troponin I has been found to bind to the calcium binding proteins calmodulin and troponin C. The regions of highest homology were characterized by patterns of high localization of charged amino acids and thus could represent a possible calmodulin or troponin C surface accessible binding site. Since subcellular localization studies have indicated that dystrophin is associated with the triadic junction, these findings imply that dystrophin could be involved in controlling intracellular calcium homeostasis.