Crystal structure of the complex of diphtheria toxin with an extracellular fragment of its receptor.

We describe the crystal structure at 2.65 A resolution of diphtheria toxin (DT) complexed 1:1 with a fragment of its cell-surface receptor, the precursor of heparin-binding epidermal-growth-factor-like growth factor (HBEGF). HBEGF in the complex has the typical EGF-like fold and packs its principal beta hairpin against the face of a beta sheet in the receptor-binding domain of DT. The interface has a predominantly hydrophobic core, and polar interactions are formed at the periphery. The structure of the complex suggests that part of the membrane anchor of the receptor can interact with a hinge region of DT. The toxin molecule is thereby induced to form an open conformation conducive to membrane insertion. The structure provides a basis for altering the binding specificity of the toxin, and may also serve as a model for other EGF-receptor interactions.

The three-dimensional structure of Escherichia coli porphobilinogen deaminase at 1.76-A resolution.

Porphobilinogen deaminase (PBGD) catalyses the polymerization of four molecules of porphobilinogen to form the 1-hydroxymethylbilane, preuroporphyrinogen, a key intermediate in the biosynthesis of tetrapyrroles. The three-dimensional structure of wild-type PBGD from Escherichia coli has been determined by multiple isomorphous replacement and refined to a crystallographic R-factor of 0.188 at 1.76 A resolution. the polypeptide chain of PBGD is folded into three alpha/beta domains. Domains 1 and 2 have a similar overall topology, based on a five-stranded, mixed beta-sheet. These two domains, which are linked by two hinge segments but otherwise make few direct interactions, form an extensive active site cleft at their interface. Domain 3, an open-faced, anti-parallel sheet of three strands, interacts approximately equally with the other two domains. The dipyrromethane cofactor is covalently attached to a cysteine side-chain borne on a flexible loop of domain 3. The cofactor serves as a primer for the assembly of the tetrapyrrole product and is held within the active site cleft by hydrogen-bonds and salt-bridges that are formed between its acetate and propionate side-groups and the polypeptide chain. The structure of a variant of PBGD, in which the methionines have been replaced with selenomethionines, has also been determined. The cofactor, in the native and functional form of the enzyme, adopts a conformation in which the second pyrrole ring (C2) occupies an internal position in the active site cleft. On oxidation, however, this C2 ring of the cofactor adopts a more external position that may correspond approximately to the site of substrate binding and polypyrrole chain elongation. The side-chain of Asp84 hydrogen-bonds the hydrogen atoms of both cofactor pyrrole nitrogens and also potentially the hydrogen atom of the pyrrole nitrogen of the porphobilinogen molecule bound to the proposed substrate binding site. This group has a key catalytic role, possibly in stabilizing the positive charges that develop on the pyrrole nitrogens during the ring-coupling reactions. Possible mechanisms for the processive elongation of the polypyrrole chain involve: accommodation of the elongating chain within the active site cleft, coupled with shifts in the relative positions of domains 1 and 2 to carry the terminal ring into the appropriate position at the catalytic site; or sequential translocation of the elongating polypyrrole chain, attached to the cofactor on domain 3, through the active site cleft by the progressive movement of domain 3 with respect to domains 1 and 2. Other mechanisms are considered although the amino acid sequence comparisons between PBGDs from all species suggest they share the same three-dimensional structure and mechanism of activity.

Structural studies of the roles of residues 82 and 85 at the interactive face of cytochrome c.

A combination of structural, functional, and mutagenic experiments has been used to study the roles of the invariant Phe82 and highly conserved Leu85 residues in cytochrome c, especially with respect to the complexation interface with electron transfer partners and maintenance of the hydrophobic heme pocket. Structural analyses show that the F82Y, L85A, and F82Y/L85A mutant proteins all retain the characteristic cytochrome c fold, but that conformational alterations are introduced in the direct vicinity of the mutation sites. In particular, the additional hydroxyl group of Tyr82 is in direct spatial conflict with the side chain of Leu85 in the F82Y mutant protein, leading to rotation of the side chain of Tyr82 out toward the protein surface. This strain is relieved in the F82Y/L85A mutant protein where the phenyl ring of Tyr82 is accommodated in a conformation comparable to that of the phenylalanine normally present at this location. In addition, the available space vacated by the replacement of Leu85 with an alanine allows for the inclusion of two new internal water molecules, one of which is bound to Tyr82 and the other to Arg13. In contrast, in the L85A mutant protein, no internal water molecules are observed in this exclusively hydrophobic pocket, which is partially filled by shifts in nearby side chains. Overall, the conformational changes observed result from the optimization of side chain packing to reflect the spatial requirements of new side chains, the minimization of both vacant internal space and the solvent exposure of hydrophobic groups, and the attainment of maximal hydrogen bonding between available polar groups.(ABSTRACT TRUNCATED AT 250 WORDS)

The three-dimensional structures of mutants of porphobilinogen deaminase: toward an understanding of the structural basis of acute intermittent porphyria.

Mutations in the human gene for the enzyme porphobilinogen deaminase give rise to an inherited disease of heme biosynthesis, acute intermittent porphyria. Knowledge of the 3-dimensional structure of human porphobilinogen deaminase, based on the structure of the bacterial enzyme, allows correlation of structure with gene organization and leads to an understanding of the relationship between mutations in the gene, structural and functional changes of the enzyme, and the symptoms of the disease. Most mutations occur in exons 10 and 12, often changing amino acids in the active site. Several of these are shown to be involved in binding the primer or substrate; none modifies Asp 84, which is essential for catalytic activity.

Identification of five novel mutations in the porphobilinogen deaminase gene.

We have studied the porphobilinogen deaminase gene transcripts from seven unrelated patients from the West of Scotland, all suffering from acute intermittent porphyria. This was achieved by reverse transcription and PCR amplification of mRNA followed by asymmetric amplification and direct sequencing. Five novel and two previously described mutations were identified and found to be single base substitutions. Of the five novel mutations, three were missense (R116Q, T2691, G274R) and two were nonsense (Q204 Stop, W283 Stop). Using Escherichia coli PBGD as a model, it is possible to predict and explain the deleterious effects that these mutations might have on the function and structure of the enzyme.

Structural studies on porphobilinogen deaminase.

The X-ray crystallographic analysis of porphobilinogen deaminase (hydroxymethylbilane synthase, EC 4.3.1.8) shows the polypeptide chain folded into three domains, (1) N-terminal, (2) central and (3) C-terminal, of approximately equal size. Domains 1 and 2 have a similar overall topology, a modified doubly wound parallel beta-sheet. Domain 3 is an open-faced three-stranded antiparallel beta-sheet, with one face covered by three alpha-helices. The active site is located between domains 1 and 2. The dipyrromethane cofactor linked to cysteine 242 protrudes from domain 3 into the mouth of the cleft. Flexible segments between domains 1 and 2 are thought to have a role in a hinge mechanism, facilitating conformational changes. The cleft is lined with positively charged, highly conserved, arginine residues which form ion pairs with the acidic side chains of the cofactor. Aspartic acid 84 has been identified as a critical catalytic residue both by its proximity to the cofactor pyrrole ring nitrogen and by structural and kinetic studies of the Asp-84-->Glu mutant protein. The active site arginine residues have been altered by site-directed mutagenesis to histidine residues. The mutant proteins have been studied crystallographically in order to reconcile the functional changes in the polymerization reaction with structural changes in the enzyme.

Detection of a high mutation frequency in exon 12 of the porphobilinogen deaminase gene in patients with acute intermittent porphyria.

Direct cDNA sequencing was performed on asymmetrically amplified transcripts from the porphobilinogen deaminase (PBG-D) gene of thirteen unrelated individuals with acute intermittent porphyria. Four different mutations and a polymorphic site were detected in exon 12 of the gene, four being the result of single base substitutions and one being caused by dinucleotide deletion. All of these mutations are located in domain 3 of the PBG-D molecule, with the single base substitutions affecting the hydrophobic interfaces between domains 1 and 3. The dinucleotide deletion results in a frame-shift producing a premature stop codon.

Purification, characterization, crystallisation and X-ray analysis of selenomethionine-labelled hydroxymethylbilane synthase from Escherichia coli.

Hydroxymethylbilane synthase (HMBS) catalyses the conversion of porphobilinogen into hydroxymethylbilane, a linear tetrapyrrolic intermediate in the biosynthesis of haems, chlorophylls, vitamin B12 and related macrocycles. In the course of an investigation of the crystal structure of this enzyme, we intended to follow a new strategy to obtain the X-ray phase information, i.e. the collection of multiwavelength anomalous diffraction data from a crystal of a seleno-L-methionine (SeMet)-labelled variant of the protein. We have expressed and purified HMBS from Escherichia coli (34268 Da) in which all (six) methionine (Met) residues are replaced by SeMet. Complete replacement, as shown by amino acid composition analysis and by electrospray mass spectrometry, was achieved by growing the Met-requiring mutant E. coli PO1562 carrying the plasmid pPA410 in a medium containing 50 mg/l SeMet as the sole source of Met. [SeMet]HMBS exhibits full enzyme activity, as reflected by unchanged steady-state kinetic parameters relative to native enzyme. Rhombohedral crystals of [SeMet]HMBS could be grown at the pH optimum (7.4) of the enzyme (solutions containing 30 mg/ml protein, 0.4 mM EDTA, 20 mM dithiothreitol, 3 M NaCl and 15 mM Bristris-propane buffer were equilibrated by vapour diffusion at 20 degrees C against reservoirs of saturated NaCl). However, being very thin plates, these crystals were not suitable for X-ray analysis. Alternatively, rectangular crystals were obtained at pH 5.3 using conditions based on those reported for wild-type HMBS [sitting drops of 50 microliters containing 6-7 mg/ml protein, 0.3 mM EDTA, 15 mM dithiothreitol, 10% (mass/vol.) poly(ethylene glycol) 6000 and 0.01% NaN3 in 0.1 M sodium acetate were equilibrated by vapour diffusion at 20 degrees C against a reservoir of 10-20 mg solid dithiothreitol]. X-ray diffraction data of the crystals were complete to 93.8% at 0.21 nm resolution and showed that [SeMet]HMBS and native HMBS crystallise isomorphously. A difference Fourier map using FSeMet-Fnative and phases derived from the native structure, which has recently been determined independently by multiple isomorphous replacement, showed positive difference peaks centered at or close to where the sulphur atoms of the Met side chains appear in the native structure. In addition, paired positive/negative peaks in the difference map near the cofactor of HMBS indicate conformational differences in the active site, probably due to differences in the state of oxidation of the cofactor in the two crystalline samples.

Structure of porphobilinogen deaminase reveals a flexible multidomain polymerase with a single catalytic site.

The three-domain structure of porphobilinogen deaminase, a key enzyme in the biosynthetic pathway of tetrapyrroles, has been defined by X-ray analysis at 1.9 A resolution. Two of the domains structurally resemble the transferrins and periplasmic binding proteins. The dipyrromethane cofactor is covalently linked to domain 3 but is bound by extensive salt-bridges and hydrogen-bonds within the cleft between domains 1 and 2, at a position corresponding to the binding sites for small-molecule ligands in the analogous proteins. The X-ray structure and results from site-directed mutagenesis provide evidence for a single catalytic site. Interdomain flexibility may aid elongation of the polypyrrole product in the active-site cleft of the enzyme.

High-resolution refinement of yeast iso-1-cytochrome c and comparisons with other eukaryotic cytochromes c.

The structure of yeast iso-1-cytochrome c has been refined against X-ray diffraction data to a nominal resolution of 1.23 A. The atomic model contains 893 protein atoms, as well as 116 water molecules and one sulfate anion. Also included in the refinement are 886 hydrogen atoms belonging to the protein molecule. The crystallographic R-factor is 0.192 for the 12,513 reflections with F greater than or equal to 3 sigma (F) in the resolution range 6.0 to 1.23 A. Co-ordinate accuracy is estimated to be better than 0.18 A. The iso-1-cytochrome c molecule has the typical cytochrome c fold, with the polypeptide chain organized into a series of alpha-helices and reverse turns that serve to envelop the heme prosthetic group in a hydrophobic pocket. Inspection of the conformations of helices in the molecule shows that the local environments of the helices, in particular the presence of intrahelical threonine residues, cause distortions from ideal alpha-helical geometry. Analysis of the internal mobility of iso-1-cytochrome c, based on refined crystallographic temperature factors, shows that the most rigid parts of the molecule are those that are closely associated with the heme group. The degree of saturation of hydrogen-bonding potential is high, with 90% of all polar atoms found to participate in hydrogen bonding. The geometry of intramolecular hydrogen bonds is typical of that observed in other high-resolution protein structures. The 116 water molecules present in the model represent about 41% of those expected to be present in the asymmetric unit. The majority of the water molecules are organized into a small number of hydrogen-bonding networks that are anchored to the protein surface. Comparison of the structure of yeast iso-1-cytochrome c with those of tuna and rice cytochromes c shows that these three molecules have very high structural similarity, with the atomic packing in the heme crevice region being particularly highly conserved. Large conformational differences that are observed between these cytochromes c can be explained by amino acid substitutions. Additional subtle differences in the positioning of the side-chains of several highly conserved residues are also observed and occur due to unique features in the local environments of each cytochrome c molecule.(ABSTRACT TRUNCATED AT 400 WORDS)

High-resolution three-dimensional structure of horse heart cytochrome c.

The 1.94 A resolution three-dimensional structure of oxidized horse heart cytochrome c has been elucidated and refined to a final R-factor of 0.17. This has allowed for a detailed assessment of the structural features of this protein, including the presence of secondary structure, hydrogen-bonding patterns and heme geometry. A comprehensive analysis of the structural differences between horse heart cytochrome c and those other eukaryotic cytochromes c for which high-resolution structures are available (yeast iso-1, tuna, rice) has also been completed. Significant conformational differences between these proteins occur in three regions and primarily involve residues 22 to 27, 41 to 43 and 56 to 57. The first of these variable regions is part of a surface beta-loop, whilst the latter two are located together adjacent to the heme group. This study also demonstrates that, in horse cytochrome c, the side-chain of Phe82 is positioned in a co-planar fashion next to the heme in a conformation comparable to that found in other cytochromes c. The positioning of this residue does not therefore appear to be oxidation-state-dependent. In total, five water molecules occupy conserved positions in the structures of horse heart, yeast iso-1, tuna and rice cytochromes c. Three of these are on the surface of the protein, serving to stabilize local polypeptide chain conformations. The remaining two are internally located. One of these mediates a charged interaction between the invariant residue Arg38 and a nearby heme propionate. The other is more centrally buried near the heme iron atom and is hydrogen bonded to the conserved residues Asn52, Tyr67 and Thr78. It is shown that this latter water molecule shifts in a consistent manner upon change in oxidation state if cytochrome c structures from various sources are compared. The conservation of this structural feature and its close proximity to the heme iron atom strongly implicate this internal water molecule as having a functional role in the mechanism of action of cytochrome c.

A polypeptide chain-refolding event occurs in the Gly82 variant of yeast iso-1-cytochrome c.

The replacement of Phe82 in yeast iso-1-cytochrome c by a glycine residue substantially alters both the tertiary structure and electron transfer properties of this protein. The largest structural change involves a polypeptide chain refolding of residues 79 through 85. Refolding places glycines 82, 83 and 84 immediately adjacent to the plane of the heme group in a spatial positioning comparable to that of the phenyl ring of Phe82 in the wild-type protein. Despite this perturbation in structure, solvent accessibility computations show that heme solvent exposure has not increased in the Gly82 variant protein. However, refolding does result in the introduction of a number of polar groups into the hydrophobic heme pocket. This appears to be responsible for the decreased reduction potential of the heme in this protein. The present study, along with that of the Ser82 variant protein (Louie et al., 1988b), clearly establishes the link between dielectric constant within the heme crevice and reduction potential. The further anomalously low electron transfer activity of the Gly82 variant protein would appear to arise from two factors. First, the polypeptide chain medium now adjacent to the heme is unable to facilitate electron transfer in a manner similar to that of the aromatic side-chain of Phe82. Second, polypeptide chain refolding significantly alters the surface contour of the Gly82 protein rendering it less suitable to interact with the corresponding complementary surfaces of redox partners. Our data support the conclusion that Phe82 plays a number of roles in the electron transfer process mediated by yeast iso-1-cytochrome c. These include the maintenance of the heme environment, provision of an optimal medium along the path of electron transfer and formation of interactions at the contact interface in complexes with redox partners.

A moderate form of hemophilia B is caused by a novel mutation in the protease domain of factor IXVancouver.

A genomic phage library was constructed using lymphocyte DNA from a patient with cross-reacting material-positive, moderately severe hemophilia B. The library was screened by using a full-length factor IX cDNA as a hybridization probe. DNA sequence analysis of the factor IX exons and intron/exon junctions revealed a single point mutation at nucleotide 31,311 of the gene. This mutation occurs in the protease domain of factor IXa and changes the codon for isoleucine 397 (ATA) to a threonine codon (ACA). The resulting abnormal protein has been named factor IXVancouver. Factor IXVancouver was isolated from the patient's plasma by barium citrate adsorption, affinity chromatography on a Ca2+-dependent antibody bound to agarose, and anion-exchange chromatography. On gel electrophoresis, the purified protein exhibited a normal molecular weight and a normal pattern of activation cleavages with bovine factor XIa. Kinetic studies on the purified protein indicated that the Km of factor IXaVancouver for human factor X was 3.4 times higher than that of normal factor IXa. The kcat of factor IXaVancouver was 12.5% of the kcat of normal factor IXa. Structural models of the protease domain of human factor IXa and of factor IXaVancouver were constructed, based on the homology of factor IXa with related serine proteases of known structure. The factor IXaVancouver model suggests that hydrogen bonding between the side chain hydroxyl group of threonine 397 and the carbonyl oxygen of tryptophan 385 reduces the ability of factor IXaVancouver to bind factor X in a configuration favoring catalysis.

Role of phenylalanine-82 in yeast iso-1-cytochrome c and remote conformational changes induced by a serine residue at this position.

A three-dimensional structural analysis of the reduced form of the Ser-82 mutant protein of yeast iso-1-cytochrome c has been completed to 2.8-A resolution. Replacement of Phe-82 with a serine residue results in conformational changes both near and remote from the mutation site. Those groups undergoing positional shifts near Ser-82 include Arg-13, Gly-83 and -84, and the CBB methyl of the heme group. Remote shifts are centered about the propionate of pyrrole ring A and principally involve Asn-52, Trp-59, and an internally buried water molecule, WAT-166. Placement of a serine side chain at position 82 also leads to the formation of a large solvent channel which substantially increases the solvent accessibility of the heme group. This would appear to account for the much lower reduction potential observed for this protein. The detrimental effect of Ser-82 on both the steady-state activity and the rate of electron transfer in complexation with cytochrome c peroxidase can also be interpreted in terms of the modified character of the region about the mutation site. The remote conformational changes observed appear to represent the equivalent of the initial conformational changes occurring as yeast iso-1-cytochrome c is converted to the fully oxidized state during an electron-transfer event. These results agree well with the proposal [Moore, G. R. (1983) FEBS Lett. 161, 171-175] that the trigger for conformational changes between oxidation states resides in the nature of the interactions between the heme iron atom and the pyrrole ring A propionate group.

Yeast iso-1-cytochrome c. A 2.8 A resolution three-dimensional structure determination.

A molecular replacement approach, augmented with the results of predictive modeling procedures, solvent accessibility studies, packing analyses and translational coefficient searches, has been used to elucidate the 2.8 A (1 A = 0.1 nm) resolution structure of yeast iso-1-cytochrome c. An examination of the polypeptide chain folding of this protein shows it to have unique conformations in three regions, upon comparison with the structures of other eukaryotic cytochromes c. These include: residues -5 to +1 at the N-terminal end of the polypeptide chain, which are in an extended conformation and project in large part off the surface of the protein; residues 19 to 26, which form a surface beta-loop on the His18 ligand side of the central heme group; and, the C-terminal end of the helical segment composed of residues 49 to 56, which serves to form a part of the heme pocket. Structural studies also show that the highly reactive sulfhydryl group of Cys102 is buried within a hydrophobic region in the monomer form of yeast iso-1-cytochrome c. Dimerization of yeast iso-1-cytochrome c through disulfide bond formation between two such residues would require a substantial conformational change in the C-terminal helix of this protein. Another unique structural feature, the trimethylated side-chain of Lys72, is located on the surface of yeast iso-1-cytochrome c near the solvent-exposed edge of the bound heme prosthetic group. On the basis of the results of these and other structural studies, an analysis of the spatial conservation of structural features in the heme pocket of eukaryotic cytochromes c has been conducted. It was found that the residues involved could be divided into three general classes. The current structural analyses and additional modeling studies have also been used to explain the altered functional properties observed for mutant yeast iso-1-cytochrome c proteins.

Characterization of human blood coagulation factor XII cDNA. Prediction of the primary structure of factor XII and the tertiary structure of beta-factor XIIa.

A human liver cDNA library was screened by colony hybridization with two mixtures of synthetic oligodeoxyribonucleotides as probes. These oligonucleotides encoded regions of beta-factor XIIa as predicted from the amino acid sequence. Four positive clones were isolated that contained DNA coding for most of factor XII mRNA. DNA sequence analysis of these overlapping clones showed that they contained DNA coding for part of an amino-terminal extension, the complete amino acid sequence of plasma factor XII, a TGA stop codon, a 3' untranslated region of 150 nucleotides, and a poly(A)+ tail. The cDNA sequence predicts that plasma factor XII consists of 596 amino acid residues. Within the predicted amino acid sequence of factor XII, we have identified three peptide bonds that are cleaved by kallikrein during the formation of beta-factor XIIa. Comparison of the structure of factor XII with other proteins revealed extensive sequence identity with regions of tissue-type plasminogen activator (the epidermal growth factor-like region and the kringle region) and fibronectin (type I and type II homologies). As the type II region of fibronectin contains a collagen-binding site, the homologous region in factor XII may be responsible for the binding of factor XII to collagen. The carboxyl-terminal region of factor XII shares considerable amino acid sequence homology with other serine proteases including trypsin and many clotting factors. A preliminary structural model of beta-factor XIIa is proposed based on the known high resolution x-ray diffraction structures of trypsin, chymotrypsin, and elastase.