Expression,purification, crystallization,and preliminary X-ray analysis of recombinant human saposin B

Saposin B (also known as cerebroside sulfate activator or CSAct) is a small non-enzymatic glycoprotein required for the breakdown of cerebroside sulfates (sulfatides) in lysosomes. Saposin B contains three intramolecular disulfide bridges, exists as a dimer and is remarkably heat, protease, and pH stable. We have expressed the protein in a thioredoxin reductase deficient strain of Escherichia coli and purified the protein by heat treatment, followed by ion-exchange, gel filtration, and hydrophobic interaction chromatographies. The protein is properly folded as judged by the observed disulfide bond topology, the hydrogen-deuterium exchange rate, and the level of stimulation of sulfatide hydrolysis by arylsulfatase A. Crystals of human saposin B were grown by vapor diffusion and diffract to a resolution of 2.2A. Despite obtaining only merohedrally twinned P3(1) native crystals, an untwined seleomethionine-substituted crystal belonging to space group P3(1)21 was also grown. The three-dimensional structure of saposin B protein will provide insights into how this 79 amino acid protein is able to solubilize relatively large membrane-bound lipid ligands.     

Monovalent cation binding to cubic insulin crystals.

Two localized monovalent cation binding sites have been identified in cubic insulin from 2.8 A-resolution difference electron density maps comparing crystals in which the Na+ ions have been replaced by Tl+. One cation is buried in a closed cavity between insulin dimers and is stabilized by interaction with protein carbonyl dipoles in two juxtaposed alternate positions related by the crystal dyad. The second cation binding site, which also involves ligation with carbonyl dipoles, is competitively occupied by one position of two alternate His B10 side chain conformations. The cation occupancy in both sites depends on the net charge on the protein which was varied by equilibrating crystals in the pH range 7-10. Detailed structures of the cation binding sites were inferred from the refined 2-A resolution map of the sodium-insulin crystal at pH 9. At pH 9, the localized monovalent cations account for less than one of the three to four positive counterion charges necessary to neutralize the negative charge on each protein molecule. The majority of the monovalent counterions are too mobile to show up in the electron density maps calculated using data only at resolution higher than 10 A. Monovalent cations of ionic radius less than 1.5 A are required for crystal stability. Replacing Na+ with Cs+, Mg++, Ca++ or La+++ disrupts the lattice order, but crystals at pH 9 with 0.1 M Li+, K+, NH4+, Rb+ or Tl+ diffract to at least 2.8 A resolution.     

Serotypes of verocytotoxigenic Escherichia coli isolated from cattle and buffalo calf diarrhoea

Abstract Parasporal crystals of the recently isolated Bacillus thuringiensis var. tenebrionis are toxic for coleopteran larvae. Unlike those of other strains they are soluble either in aqueous solutions of NaBr at neutral pH or in water after titration to pH values above pH 10.0. The dissolved crystal protein readily forms crystals after removal of the salt or neutralization. The crystal protein was not found to differ much in the amino acid composition from other crystal proteins. The parasporal crystals are composed of subunits of M _r 68 000 which are not linked by disulfide bridges.     

Carbohydrate-binding properties of human neo-CRP and its relationship to phosphorylcholine-binding site

Binding characteristics of two types of ligands for human neo-C-reactive protein (neo-CRP), which is a conformationally altered but physiologically relevant form of CRP, were studied fluorometrically by probing CRP immobilized on a polystyrene surface with europium-labeled ligands. Two Eu-ligands used were bovine serum albumin derivatives that contain on average 40 residues of ligand structures, one derivative containing phosphorylcholine (PC) and the other lactosyl residues. The PC-containing ligands required the presence of calcium for binding, whereas galactose-containing derivatives bound in the absence of calcium. The optimal pH for the PC-dependent binding was broad (pH 6-8), whereas the best binding pH for the galactose-dependent binding was around 6. The carbohydrate-mediated binding is rather nonspecific: the binding site prefers galactose configuration, but other hexoses can be accommodated. The two best monosaccharide inhibitors at this site were galactose-6-phosphate and galacturonic acid, suggesting the importance of having a negatively charged group at C-6 position of galactose. In fact, the phosphate-binding site is common to both PC and sugar phosphates, and the choline- and the sugar-binding sites are probably located on either side of the phosphate-binding site. Binding characteristics of Eu-labeled PC-BSA to neo-CRP are quite similar to that found for native CRP in solution phase [Lee et al. (2002) J. Biol. Chem., 277, 225-232], whereas binding of sugar phosphates by neo-CRP shows considerably less stringent requirements compared to native CRP. For instance, galactose-alpha1-phosphate was not inhibitory at all in the native CRP binding assay, whereas it was a good inhibitor in the neo-CRP assay.     

Selenium derivatization of nucleic acids for crystallography

The high-resolution structure of the DNA (5′-GTGTACA-C-3′) with the selenium derivatization at the 2′-position of T2 was determined via MAD and SAD phasing. The selenium-derivatized structure (1.28 Å resolution) with the 2′-Se modification in the minor groove is isomorphorous to the native structure (2.0 Å). To directly compare with the conventional bromine derivatization, we incorporated bromine into the 5-postion of T4, determined the bromine-derivatized DNA structure at 1.5 Å resolution, and found that the local backbone torsion angles and solvent hydration patterns were altered in the structure with the Br incorporation in the major groove. Furthermore, while the native and Br-derivatized DNAs needed over a week to form reasonable-size crystals, we observed that the Se-derivatized DNAs grew crystals overnight with high-diffraction quality, suggesting that the Se derivatization facilitated the crystal formation. In addition, the Se-derivatized DNA sequences crystallized under a broader range of buffer conditions, and generally had a faster crystal growth rate. Our experimental results indicate that the selenium derivatization of DNAs may facilitate the determination of nucleic acid X-ray crystal structures in phasing and high-quality crystal growth. In addition, our results suggest that the Se derivatization can be an alternative to the conventional Br derivatization.     

Comparison of Disulfide Contents and Solubility at Alkaline pH of Insecticidal and Noninsecticidal Bacillus thuringiensis Protein Crystals

We compared two insecticidal and eight noninsecticidal soil isolates of Bacillus thuringiensis with regard to the solubility of their proteinaceous crystals at alkaline pH values. The protein disulfide contents of the insecticidal and noninsecticidal crystals were equivalent. However, six of the noninsecticidal crystals were soluble only at pH values of ≥12. This lack of solubility contributed to their lack of toxicity. One crystal type which was soluble only at pH ≥12 (strain SHP 1-12) did exhibit significant toxicity to tobacco hornworm larvae when the crystals were presolubilized. In contrast, freshly prepared crystals from the highly insecticidal strain HD-1 were solubilized at pH 9.5 to 10.5, but when these crystals were denatured, by either 8 M urea or autoclave temperatures, they became nontoxic and were soluble only at pH values of ≥12. These changes in toxicity and solubility occurred even though the denatured HD-1 crystals were morphologically indistinguishable from native crystals. Our data are consistent with the view that insecticidal crystals contain distorted, destabilized disulfide bonds which allow them to be solubilized at pH values (9.5 to 10.5) characteristic of lepidopteran and dipteran larval midguts.     

Structure of porphobilinogen synthase from Pseudomonas aeruginosa in complex with 5-fluorolevulinic acid suggests a double Schiff base mechanism

The seeds of jack fruit (Artocarpus integrifolia) contain two tetrameric lectins, jacalin and artocarpin. Jacalin was the first lectin found to exhibit the beta-prism I fold, which is characteristic of the Moraceae plant lectin family. Jacalin contains two polypeptide chains produced by a post-translational proteolysis which has been shown to be crucial for generating its specificity for galactose. Artocarpin is a single chain protein with considerable sequence similarity with jacalin. It, however, exhibits many properties different from those of jacalin. In particular, it is specific to mannose. The structures of two crystal forms, form I and form II, of the native lectin have been determined at 2.4 and 2.5 A resolution, respectively. The structure of the lectin complexed with methyl-alpha-mannose, has also been determined at 2.9 A resolution. The structure is similar to jacalin, although differences exist in details. The crystal structures and detailed modelling studies indicate that the following differences between the carbohydrate binding sites of artocarpin and jacalin are responsible for the difference in the specificities of the two lectins. Firstly, artocarpin does not contain, unlike jacalin, an N terminus generated by post-translational proteolysis. Secondly, there is no aromatic residue in the binding site of artocarpin whereas there are four in that of jacalin. A comparison with similar lectins of known structures or sequences, suggests that, in general, stacking interactions with aromatic residues are important for the binding of galactose while such interactions are usually absent in the carbohydrate binding sites of mannose-specific lectins with the beta-prism I fold.     

Crystallization and preliminary X-ray investigation reveals that tumor necrosis factor is a compact trimer furnished with 3-fold symmetry

Crystals of recombinant human tumor necrosis factor produced by Escherichia coli have been obtained under different conditions. Crystals suitable for X-ray studies are produced by a vapor diffusion technique using sodium phosphate as both precipitant and buffer at pH 6.5. The crystals belong to the cubic space group, P2(1)3 with unit cell dimensions a = b = c = 95.7 A (1 A = 0.1 nm). Preliminary photography reveals that the crystals are moderately stable to X-rays and diffract to at least 3 A resolution. The diffraction data for native crystals have been collected on a diffractometer at 3 A resolution. Another crystal form, which appeared in a solution containing sodium phosphate at pH 8.0, has the trigonal space group P3 with unit cell dimensions a = b = 63.8 A and c = 54.4 A, and produces measurable reflections to a resolution of 3 A. Hexagonal crystals also have been obtained by the use of polyethylene glycol as precipitant in the range pH 7.6 to 8.0; however, the crystals are fragile and unstable to X-rays. Conservation of 3-fold symmetry in the different crystal forms obtained could reflect the ability of tumor necrosis factor molecules to form trimers in solution and probably the nature of binding of the molecules to cellular receptors.     

Properties of the galactose binding protein of Salmonella typhimurium and Escherichia coli.

The galactose binding protein implicated in transport and in chemotaxis has been purified to homogeneity from the shock fluids of Salmonella typhimurium and Escherichia coli. Both proteins are monomers of molecular weight 33 000 and exhibit cross-reactivity with antibody. The Salmonella galactose receptor showed binding of 1 mol of [14C]galactose or 1 mol of [14C]glucose at saturation. The dissociation constants were 0.38 and 0.17 muM, respectively. In light of the previously published report that the E. coli protein contains two binding sites with two different affinities, the binding characteristics of this protein were reexamined. Using highly purified radiolabeled substrate and homogeneous protein, a single binding site and single binding affinity were seen galactose (KD = 0.48 muM) or for glucose (KD = 0.21 muM). The competition between glucose and galactose for the same site is intriguing in view of the competition between ribose and galactose at the receptor level.     

Preliminary crystallographic analysis of the ATP-hydrolysing domain of the Escherichia coli DNA gyrase B protein.

The 43 kDa N-terminal ATPase domain of the Escherichia coli DNA gyrase B protein has been purified from an over-expressing strain. This protein has been crystallized in two crystal forms, both in the presence of the non-hydrolysable ATP analogue 5'-adenylyl-beta,gamma-imidodiphosphate. The first crystal form is monoclinic P2(1), with cell dimensions a = 76 A, b = 88 A, c = 82 A, beta = 105.5 degrees, and diffracts to at least 2.7 A resolution using synchrotron radiation. Crystal density measurements suggest that there are two molecules in the asymmetric unit (Vm = 3.08 A3/Da). The second crystal form is orthorhombic C222(1), with cell dimensions a = 89.2 A, b = 143.1 A and c = 79.8 A. The crystals diffract to beyond 3 A and are stable for at least 100 hours when exposed to X-rays from a rotating anode source. The asymmetric unit of this crystal form appears to contain one molecule (Vm = 2.96 A3/Da). Data have already been collected to 5 A resolution from native crystals of this second form, and to 6 A resolution from three heavy-atom derivatives. Electron density maps calculated using phases obtained from these derivatives show features consistent with secondary structural elements, and have allowed the molecular boundary to be determined. Higher resolution native and derivative data are being collected.     

Bacillus thuringiensis crystal protein: effect of chemical modification of the cysteine and lysine residues.

The 16 cysteine residues of reduced protoxin from Bacillus thuringiensis subsp. kurstaki HD-73 can be quantitatively reacted with: (a) iodoacetic acid, to give carboxymethyl protoxin; (b) iodoacetamide, giving carbaminomethyl protoxin and (c) N-(beta-iodoethyl)trifluoroacetamide to give aminoethyl protoxin. The carboxymethyl derivative was found to be significantly more soluble at neutral pH values where both the native protoxin and the carbaminomethyl derivative exhibit low solubilities. At the alkaline pH values (pH 9.5-10.5) normally used to solubilize the crystal protein, the native protein was slightly more soluble than either the carboxymethyl or the carbaminomethyl derivatives. The aminoethyl derivative had an extremely low solubility at all pH values. Succinic anhydride reacted with only 35% of the lysine residues in both the carboxymethyl and the carbaminomethyl protoxin derivatives. Nonetheless, these succinylated protoxins exhibited significantly increased solubilities at neutral pH values. All the derivatives were found to retain full insecticidal activity toward spruce budworm (Choristeneura fufimerana) larvae. It is concluded that all the cysteine residues and modified lysine residues are on the surface of the protein and that derivatization does not alter the conformation of the solubilized protoxin.     

The redox state of cysteines 201 and 317 of the erythrocyte anion exchanger is critical for ankyrin binding

Previous studies have demonstrated that modification of erythrocyte membrane cysteine residues via disulfide cross-briding or direct derivatization with thiol reagents promotes massive morphological, rheological, and structural changes in the cell. To determine whether disruption of the band 3-ankyrin interaction, the major membrane-cytoskeletal linkage, might contribute to the above lesions, we quantitatively measured the band 3-ankyrin interaction following modification of Cys-201 and/or Cys-317 of the cytoplasmic domain of band 3. It was observed that irreversible alkylating agents (e.g. N-ethylmaleimide or iodoacetamide and its derivatives), reversible derivatizing compounds (.e.g. p-chloromercuribenzenesulfonate or glutathione), and native disulfide bond formation all blocked the ankyrin interaction. Comparison of the extent of sulfhydryl modification with the degree of inhibition of ankyrin binding further confirmed that cysteine modification was directly responsible for the inhibition. However, analysis of the site of sulfhydryl derivatization revealed that inhibition of ankyrin binding could be initiated in some cases with derivatization of Cys-201, while in other cases obstruction of Cys-317 appeared to be essential. This apparent discrepancy was resolved by demonstrating that Cys-201 of one strand of the cytoplasmic domain of band 3 dimer could disulfide bond with Cys-317 of the opposite strand, thus demonstrating that all four cysteines of the band 3 dimer are clustered at the interface between subunits. We argue that derivatization or disulfide cross-linking of these cysteines can block ankyrin binding by both conformational and steric mechanisms.     

Engineered interdomain disulfide in the periplasmic receptor for sulfate transport reduces flexibility. Site-directed mutagenesis and ligand-binding studies

Using recombinant DNA techniques, an Escherichia coli periplasmic sulfate receptor or sulfate-binding protein involved in active transport has been overexpressed and characterized. This protein is essentially identical in size, sequence, antigenicity, and ligand affinity and specificity to the sulfate receptor from Salmonella typhimurium whose crystal structure has been refined at 2 A resolution. The dehydrated sulfate is bound in the deep cleft between the two lobes of the bilobate protein. Using the structure of the S. typhimurium as a guide, three site-directed mutants (Ser129Cys, Gly46Cys, and Ser129Cys/Gly46Cys) have been made. In the Cys129/Cys46 mutant the disulfide has been successfully introduced across the opening of the ligand-binding site cleft of the E. coli sulfate-binding protein. The dissociation of sulfate from the double mutant protein is very slow under oxidizing conditions and increases more than 200-fold when reducing agent is added. This effect is attributed to a loss of interdomain structural flexibility in the presence of the disulfide, and underscores the importance of protein conformational change in binding protein function.     

Oxidized and synchrotron cleaved structures of the disulfide redox center in the N-terminal domain of Salmonella typhimurium AhpF

The flavoprotein component (AhpF) of Salmonella typhimurium alkyl hydroperoxide reductase contains an N-terminal domain (NTD) with two contiguous thioredoxin folds but only one redox-active disulfide (within the sequence -Cys129-His-Asn-Cys132-). This active site is responsible for mediating the transfer of electrons from the thioredoxin reductase-like segment of AhpF to AhpC, the peroxiredoxin component of the two-protein peroxidase system. The previously reported crystal structure of AhpF possessed a reduced NTD active site, although fully oxidized protein was used for crystallization. To further investigate this active site, we crystallized an isolated recombinant NTD (rNTD); using diffraction data sets collected first at our in-house X-ray source and subsequently at a synchrotron, we showed that the active site disulfide bond (Cys129-Cys132) is oxidized in the native crystals but becomes reduced during synchrotron data collection. The NTD disulfide bond is apparently particularly sensitive to radiation cleavage compared with other protein disulfides. The two data sets provide the first view of an oxidized (disulfide) form of NTD and show that the changes in conformation upon reduction of the disulfide are localized and small. Furthermore, we report the apparent pKa of the active site thiol to be approximately 5.1, a relatively low pKa given its redox potential (approximately 265 mV) compared with most members of the thioredoxin family.     

Purification, crystallization, and preliminary x-ray diffraction studies of the flavoenzyme mercuric ion reductase from Bacillus sp. strain RC607

The flavoenzyme mercuric ion reductase from Bacillus sp. strain RC607 was purified by dye-ligand affinity chromatography. The protein was crystallized from solutions of high ionic strength, and one of the two crystal forms obtained has proven suitable for x-ray diffraction studies. Preliminary analysis showed that these crystals belong to the tetragonal space group 1422. The unit cell dimensions are a = b = 180.7 A; c = 127.9 A. The diffraction pattern extends to better than 3 A resolution. Crystal density measurements are consistent with one enzyme dimer of 2 x 69,000 Da comprising the asymmetric unit. Trypsin treatment of the native enzyme resulted in the removal of 157 amino acids at the N terminus. After purification, the remaining fragment (amino acids 158-631), which is still fully active in vitro, could be crystallized under the same conditions as native enzyme. Twinning problems, however, did not allow complete analysis of these crystals.     

Structural basis of calcium and galactose recognition by the lectin PA-IL of Pseudomonas aeruginosa

The structure of the tetrameric Pseudomonas aeruginosa lectin I (PA-IL) in complex with galactose and calcium was determined at 1.6 A resolution, and the native protein was solved at 2.4 A resolution. Each monomer adopts a beta-sandwich fold with ligand binding site at the apex. All galactose hydroxyl groups, except O1, are involved in a hydrogen bond network with the protein and O3 and O4 also participate in the co-ordination of the calcium ion. The stereochemistry of calcium galactose binding is reminiscent of that observed in some animal C-type lectins. The structure of the complex provides a framework for future design of anti-bacterial compounds.     

Amino sugars in the glycoprotein toxin from Bacillus thuringiensis subsp. israelensis.

The carbohydrate content of purified Bacillus thuriniensis subsp. israelensis crystal toxin was determined by six biochemical tests, column chromatography on an amino acid analyzer, and the binding of 11 fluorescent lectins. The crystals contained approximately 1.0% neutral sugars and 1.7% amino sugars. The amino sugars consisted of 70% glucosamine and 30% galactosamine. No N-acetylneuraminic acid (sialic acid) was detected. The presence of amino sugars was confirmed by the strong binding of fluorescent wheat germ agglutinin and the weak binding of fluorescent soybean agglutinin. These lectins recognize N-acetyl-D-glucosamine and N-acetyl-D-galactosamine, respectively. The lectin-binding sites appeared evenly distributed among the protein subunits of the crystal. The sugars were covalently attached to the crystal toxin because wheat germ agglutinin still bound alkali-solubilized toxin which had been boiled in sodium dodecyl sulfate, separate by polyacrylamide gel electrophoresis, and transferred to nitrocellulose membranes. This study demonstrates the covalent attachment of amino sugars and indicates that the B. thuringiensis subsp. israelensis protein toxins should be viewed as glycoprotein toxins. The crystals used in the present study were purified on sodium bromide density gradients. Studies employing crystals purified on Renografin density gradients can give artificially high values for the anthrone test for neutral sugars.     

Crystal structures of apo- and holo-bovine alpha-lactalbumin at 2. 2-A resolution reveal an effect of calcium on inter-lobe interactions

High affinity binding of Ca(2+) to alpha-lactalbumin (LA) stabilizes the native structure and is required for the efficient generation of native protein with correct disulfide bonds from the reduced denatured state. A progressive increase in affinity of LA conformers for Ca(2+) as they develop increasingly native structures can account for the tendency of the apo form to assume a molten globule state and the large acceleration of folding by Ca(2+). To investigate the effect of calcium on structure of bovine LA, x-ray structures have been determined for crystals of the apo and holo forms at 2.2-A resolution. In both crystal forms, which were grown at high ionic strength, the protein is in a similar global native conformation consisting of alpha-helical and beta-subdomains separated by a cleft. Even though alternative cations and Ca(2+) liganding solvent molecules are absent, removal of Ca(2+) has only minor effects on the structure of the metal-binding site and a structural change was observed in the cleft on the opposite face of the molecule adjoining Tyr(103) of the helical lobe and Gln(54) of the beta-lobe. Changes include increased separation of the lobes, loss of a buried solvent molecule near the Ca(2+)-binding site, and the replacement of inter- and intra-lobe H-bonds of Tyr(103) by interactions with new immobilized water molecules. The more open cleft structure in the apo protein appears to be an effect of calcium binding transmitted via a change in orientation of helix H3 relative to the beta-lobe to the inter-lobe interface. Calcium is well known to promote the folding of LA. The results from the comparison of apo and holo structures of LA provide high resolution structural evidence that the acceleration of folding by Ca(2+) is mediated by an effect on interactions between the two subdomains.     

A structural basis for the interaction of urea with lysozyme.

The effect of urea on the crystal structure of hen egg-white lysozyme has been investigated using X-ray crystallography. High resolution structures have been determined from crystals grown in the presence of 0, 0.7, 2, 3, 4, and 5 M urea and from crystals soaked in 9 M urea. All the forms are essentially isomorphous with the native type II crystals, and the derived structures exhibit excellent geometry and RMS differences from ideality in bond distances and angles. Comparison of the urea complex structures with the native enzyme (type II form, at 1.5 A resolution) indicates that the effect of urea is minimal over the concentration range studied. The mean difference in backbone conformation between the native enzyme and its urea complexes varies from 0.18 to 0.49 A. Conformational changes are limited to flexible surface loops (Thr 69-Asn 74, Ser 100-Asn 103), the active site loop (Asn 59-Cys 80), and the C-terminus (Cys 127-Leu 129). Urea molecules are bound to distinct sites on the surface of the protein. One molecule is bound to the active site cleft's C subsite, at all concentrations, in a fashion analogous to that of the N-acetyl substituent of substrate and inhibitor sugars normally bound to this site. Occupation of this subsite by urea alone does not appear to induce the conformational changes associated with inhibitor binding.