Biosynthesis of lipopolysaccharide in Escherichia coli. Cytoplasmic enzymes that attach 3-deoxy-D-manno-octulosonic acid to lipid A

Previous studies in our laboratory led to the elucidation of the covalent structure of a tetraacyldisaccharide 1,4'-bisphosphate precursor of lipid A (designated lipid IVA), that accumulates at 42 degrees C in temperature-sensitive mutants defective in 3-deoxy-D-manno-octulosonic acid (KDO) biosynthesis (Raetz, C. R. H., Purcell, S., Meyer, M. V., Qureshi, N., and Takayama, K. (1985) J. Biol. Chem. 260, 16080-16088). Using [4'-32P]lipid IVA as the probe, we now demonstrate the existence of cytoplasmic KDO-transferases in Escherichia coli capable of attaching 2 KDO residues, derived from CMP-KDO, to lipid IVA. A partial purification has been developed to obtain a cytoplasmic subfraction that adds these 2 KDO residues with a 90% yield. The product is shown to have the stoichiometry of (KDO)2-IVA by fast atom bombardment mass spectrometry and NMR spectroscopy. The partially purified enzyme can utilize alternative lipid-disaccharide cosubstrates bearing five or six fatty acyl chains, but it has an absolute requirement for a monophosphate residue at position 4' of the lipid acceptor. When reincubated with a crude cytoplasmic fraction, a nucleoside triphosphate and Mg2+, (KDO)2-IVA is rapidly metabolized to more polar substances, the identity of which is unknown. The KDO-transferase(s) described in the present study should be very useful for the semisynthetic preparation of complex lipopolysaccharide substructures and analogs.     

Inhibition of exogenous 3-deoxy-D-manno-octulosonate incorporation into lipid A precursor of toluene-treated Salmonella typhimurium cells.

Analogs of 3-deoxy-D-manno-octulosonate (KDO) were designed to inhibit CTP:CMP-KDO cytidylyltransferase (CMP-KDO synthetase). Since these analogs lacked whole-cell antibacterial activity, a permeabilized-cell method was developed to measure intracellular compound activity directly. The method employed a mutant of Salmonella typhimurium defective in KDO-8-phosphate synthetase (kdsA), which accumulated lipid A precursor at 42 degrees C. Cells permeabilized with 1% toluene were used to evaluate inhibitor effect on [3H]KDO incorporation into preformed lipid A precursor. KDO incorporation proceeded through the enzymes CMP-KDO synthetase and CMP-KDO:lipid A KDO transferase. Optimum KDO incorporation occurred between pH 8 and 9 and required CTP, prior lipid A precursor accumulation, and a functional kdsB gene product, CMP-KDO synthetase. The apparent Km for KDO in this coupled system at pH 7.6 was 1.38 mM. The reaction products isolated and characterized contained 1 and 2 KDO residues per lipid A precursor molecule. Several KDO analogs produced concentration-related reductions of KDO incorporation in toluenized cells with 50% inhibitory concentrations comparable to those obtained in purified CMP-KDO synthetase systems. Two compounds, 8-amino-2-deoxy-KDO (A-60478) and 8-aminomethyl-2-deoxy-KDO (A-60821), competitively inhibited KDO incorporation, displaying Kis of 4.2 microM for A-60478 and 2.5 microM for A-60821. These data indicated that the inactivity of the KDO analogs on intact bacteria was the result of poor permeation into cells rather than intracellular inactivation.     

Biosynthesis of lipid A in Escherichia coli. Acyl carrier protein-dependent incorporation of laurate and myristate

In previous studies we described enzyme(s) from Escherichia coli that transfer two 3-deoxy-D-manno-octulosonate (KDO) residues from two CMP-KDO molecules to a tetraacyldisaccharide-1,4'-bis-phosphate precursor of lipid A, termed lipid IVA (Brozek, K. A., Hosaka, K., Robertson, A. D., and Raetz, C. R. H. (1989) J. Biol. Chem. 264, 6956-6966). The product, designated (KDO)2-IVA, can be prepared in milligram quantities and/or radiolabeled with 32P at position 4' of the IVA moiety. We now demonstrate the presence of enzymes in E. coli extracts that transfer laurate and/or myristate residues from lauroyl or myristoyl-acyl carrier protein (ACP) to (KDO)2-IVA. Thioesters of coenzyme A are not substrates. The cytosolic fraction catalyzes rapid acylation with lauroyl-ACP, but not with myristoyl, R-3-hydroxymyristoyl, palmitoyl, or palmitoleoyl-ACP. The membrane fraction transfers both laurate and myristate to (KDO)2-IVA. Evidence for the enzymatic acylation of (KDO)2-IVA is provided by (a) conversion of [4'-32P](KDO)2-IVA to more rapidly migrating products in the presence of the appropriate acyl-ACP, (b) incorporation of [1-14C]laurate or [1-14C]myristate into these metabolites in the presence of (KDO)2-IVA, (c) fast atom bombardment-mass spectrometry, and (d) 1H NMR spectroscopy. At protein concentrations less than 0.5 mg/ml, the acylation of (KDO)2-IVA by the cytoplasmic fraction is absolutely dependent upon the addition of exogenous acyl-ACP. These acyltransferases cannot utilize lipid IVA as a substrate, demonstrating that they possess novel KDO recognition domains. The unusual substrate specificity of these enzymes provides compelling evidence for their involvement in lipid A biosynthesis. Depending on the conditions it is possible to acylate (KDO)2-IVA with 1 or 2 lauroyl residues, with 1 or 2 myristoyl residues, or with 1 of each.     

Chemical and biological properties of lipopolysaccharide, lipid A and degraded polysaccharide from Wolinella recta ATCC 33238

Lipopolysaccharide (LPS) was isolated and purified from Wolinella recta ATCC 33238 by the phenol-water procedure and RNAase treatment. The sugar components of the LPS were rhamnose, mannose, glucose, heptose, 2-keto-3-deoxyoctonate (KDO) (3-deoxy-D-manno-octulosonate) and glucosamine. The degraded polysaccharide prepared from LPS by mild acid hydrolysis was fractionated by Sephadex G-50 gel chromatography into three fractions: (1) a high-molecular-mass fraction, eluting just behind the void volume, consisting of a long chain of rhamnose (22 mols per 3 mols of heptose residue) with attached core oligosaccharide; (2) a core oligosaccharide containing heptose, glucose and KDO, substituted with a short side chain of rhamnose; (3) a low-molecular-mass fraction containing KDO and phosphate. The main fatty acids of the lipid A were C12:0, C14:0, 3-OH-C14:0 and 3-OH-C16:0. The biological activities of the LPS were similar to those of Salmonella typhimurium LPS in activation of the clotting enzyme of Limulus amoebocytes, the Schwartzman reaction and mitogenicity for murine lymphocytes, although all the biological activities of lipid A were lower than those of intact LPS.     

Lipid A mutants of Salmonella typhimurium. Purification and characterization of a lipid A precursor produced by a mutant in 3-deoxy-D-mannooctulosonate-8-phosphate synthetase.

We describe here the isolation, purification, and structural characterization of a lipid A precursor synthesized under nonpermissive conditions by a mutant of Salmonella typhimurium conditionally defective in the synthesis of the 3-deoxy-D-mannoctulosonate (2-keto-3-deoxyoctonate, KDO) region of the lipopolysaccharide. The precursor was isolated free from lipopolysaccharide, murein, and phospholipids by extraction of delipidated cells with 90% phenol/CHCL3/petroleum ether. The molecule was recovered from the phenol phase after precipitation of lipopolysaccharide with H2O and subsequently purified by DEAE-cellulose chromatography. Structural analyses showed that the lipid A precursor is a phosphorylated glucosamine disaccharide containing one ester and two amide-linked residues of beta-hydroxymyristate. In contrast to lipid A, the precursor disaccharide lacks ester-linked 12:0 and 14:0 fatty acids as well as KDO. The molecule contains 2 phosphate residues both of which were identified as phosphomonoesters by 31P NMR spectroscopy. One of the phosphomonoesters is located in position 1 of the reducing terminal glucosamine residue; the location of the other phosphomonoester was not determined. The structure of the precursor provides strong support for the conclusion that KDO incorporation occurs at an early stage in lipid A biosynthesis prior to the incorporation of ester-linked saturated fatty acids.     

Crystal structures of Escherichia coli KDO8P synthase complexes reveal the source of catalytic irreversibility

The enzyme 3-deoxy-D-manno-2-octulosonate-8-phosphate synthase (KDO8PS) catalyses the condensation of arabinose 5-phosphate (A5P) and phosphoenol pyruvate (PEP) to obtain 3-deoxy-D-manno-2-octulosonate-8-phosphate (KDO8P). We have elucidated initial modes of ligand binding in KDO8PS binary complexes by X-ray crystallography. Structures of the apo-enzyme and of binary complexes with the substrate PEP, the product KDO8P and the catalytically inactive 1-deoxy analog of arabinose 5-phosphate (1dA5P) were obtained. The KDO8PS active site resembles an irregular funnel with positive electrostatic potential situated at the bottom of the PEP-binding sub-site, which is the primary attractive force towards negatively charged phosphate moieties of all ligands. The structures of the ligand-free apo-KDO8PS and the binary complex with the product KDO8P visualize for the first time the role of His202 as an active-site gate. Examination of the crystal structures of KDO8PS with the KDO8P or 1dA5P shows these ligands bound to the enzyme in the PEP-binding sub-site, and not as expected to the A5P sub-site. Taken together, the structures presented here strengthen earlier evidence that this enzyme functions predominantly through positional catalysis, map out the roles of active-site residues and provide evidence that explains the total lack of catalytic reversibility.     

A single point mutation in 3-deoxy-D-manno-octulosonate-8-phosphate synthase is responsible for temperature sensitivity in a mutant strain of Salmonella typhimurium

Salmonella typhimurium mutants conditionally deficient in 3-deoxy-d-manno-octulosonate-8-phosphate (KDO8P) synthase activity play a central role in our understanding of lipopolysaccharide function in enteric bacteria. The detailed characterization of KDO8P synthase from such a mutant, however, has not been previously reported. To address this issue KDO8P synthase from S. typhimurium AG701 and from a related temperature-sensitive strain (S. typhimurium AG701i50) have been overexpressed in Escherichia coli and purified to homogeneity. The enzyme from the temperature-sensitive strain has a single proline to serine substitution at position 145, leading to an increase in K(m) for both substrates, d-arabinose 5-phosphate and phosphoenolpyruvate. Analytical gel filtration and native polyacrylamide gel electrophoresis indicate that this enzyme also has an altered oligomeric state. These observations are rationalized through an examination of the structure of E. coli KDO8P synthase, which has 93% sequence identity to the enzyme from S. typhimurium.     

A metal bridge between two enzyme families. 3-deoxy-D-manno-octulosonate-8-phosphate synthase from Aquifex aeolicus requires a divalent metal for activity

The enzymes 3-deoxy-d-manno-octulosonic acid-8-phosphate synthase (KDO8PS) and 3-deoxy-d-arabino-heptulosonic acid-7-phosphate synthase (DAHPS) catalyze analogous condensation reactions between phosphoenolpyruvate and d-arabinose 5-phosphate or d-erythrose 4-phosphate, respectively. While several similarities exist between the two enzymatic reactions, classic studies on the Escherichia coli enzymes have established that DAHPS is a metalloenzyme, whereas KDO8PS has no metal requirement. Here, we demonstrate that KDO8PS from Aquifex aeolicus, representing only the second member of the KDO8PS family to be characterized in detail, is a metalloenzyme. The recombinant KDO8PS, as isolated, displays an absorption band at 505 nm and contains approximately 0.4 and 0.2-0.3 eq of zinc and iron, respectively, per enzyme subunit. EDTA inactivates the enzyme in a time- and concentration-dependent manner and eliminates the absorption at 505 nm. The addition of Cu(2+) to KDO8PS produces an intense absorption at 375 nm, while neither Co(2+) nor Ni(2+) produce such an effect. The EDTA-treated enzyme is reactivated by a wide range of divalent metal ions including Ca(2+), Cd(2+), Co(2+), Cu(2+), Fe(2+), Mg(2+), Mn(2+), Ni(2+), and Zn(2+) and is reversibly inhibited by higher concentrations (>1 mm) of certain metals. Analysis of several metal forms of the enzyme by plasma mass spectrometry suggests that the enzyme preferentially binds one, two, or four metal ions per tetramer. These observations strongly suggest that A. aeolicus KDO8PS is a metalloenzyme in vivo and point to a previously unrecognized relationship between the KDO8PS and DAHPS families.     

Escherichia coli YrbI is 3-deoxy-D-manno-octulosonate 8-phosphate phosphatase

3-Deoxy-d-manno-octulosonate 8-phosphate (KDO 8-P) phosphatase, which catalyzes the hydrolysis of KDO 8-P to KDO and inorganic phosphate, is the last enzyme in the KDO biosynthetic pathway for which the gene has not been identified. Wild-type KDO 8-P phosphatase was purified from Escherichia coli B, and the N-terminal amino acid sequence matched a hypothetical protein encoded by the E. coli open reading frame, yrbI. The yrbI gene, which encodes for a protein of 188 amino acids, was cloned, and the gene product was overexpressed in E. coli. The recombinant enzyme is a tetramer and requires a divalent metal cofactor for activity. Optimal enzymatic activity is observed at pH 5.5. The enzyme is highly specific for KDO 8-P with an apparent K(m) of 75 microm and a k(cat) of 175 s(-1) in the presence of 1 mm Mg(2+). Amino acid sequence analysis indicates that KDO 8-P phosphatase is a member of the haloacid dehalogenase hydrolase superfamily.     

A bacteriophage-associated glycanase cleaving beta-pyranosidic linkages of 3-deoxy-D-manno-2-octulosonic acid (KDO)

A bacteriophage growing on Escherichia coli K13, K20, and K23 strains carries a glycanase that catalyzes the hydrolytic cleavage of the beta-ketopyranosidic linkages of 3-deoxy-D-manno-2-octulosonic acid (KDO) in the respective capsular polysaccharides. The main cleavage product of the K23 polysaccharide has been identified by 1H- and 13C-n.m.r. spectroscopy as beta beta Ribfl----7 beta KDOp2----3-beta Ribfl----7KDO. Cleavage of polysaccharides containing alpha-pyranosidic, or 5-substituted beta-pyranosidic KDO is not catalyzed by the enzyme.     

Purification and characterization of cytidine 5''-triphosphate:cytidine 5''-monophosphate-3-deoxy-D-manno-octulosonate cytidylyltransferase

Cytidine 5'-triphosphate:cytidine 5'-monophosphate-3-deoxy-D-manno-octulosonate cytidylyltransferase (CMP-KDO synthetase) was purified 2,300-fold from frozen Escherichia coli B cells. The enzyme catalyzed the formation of CMP-KDO, a very labile product, from CTP and KDO. No other sugar tested could replace KDO as an alternate substrate. Uridine 5'-triphosphate at pH 9.5 and deoxycytidine 5'-triphosphate at pH 8.0 and 9.5 could be used as alternate substrates in place of CTP. CMP-KDO synthetase required Mg2+ at a concentration of 10.0 mM for optimal activity. The pH optimum was determined to be between 9.6 and 9.3 in tris(hydroxymethyl)aminomethane-acetate or sodium-glycine buffer. This enzyme had an isoelectric point between pH 4.15 and 4.4 and appeared to be a single polypeptide chain with a molecular weight of 36,000 to 40,000. The apparent Km values for CTP and KDO in the presence of 10.0 mM Mg2+ were determined to be 2.0 X 10(-4) and 2.9 X 10(-4) M, respectively, at pH 9.5. Uridine 5'-triphosphate and deoxycytidine 5'-triphosphate had apparent Km values of 8.8 X 10(-4) and 3.4 X 10(-4) M. respectively, at pH 9.5.     

A novel protein kinase target for the lipid second messenger phosphatidic acid.

Activation of phospholipase D occurs in response to a wide variety of hormones, growth factors, and other extracellular signals. The initial product of phospholipase D, phosphatidic acid (PA), is thought to serve a signaling function, but the intracellular targets for this lipid second messenger are not clearly identified. The production of PA in human neutrophils is closely correlated with the activation of NADPH oxidase, the enzyme responsible for the respiratory burst. We have developed a cell-free system, in which the activation of NADPH oxidase is induced by the addition of PA. Characterization of this system revealed that a multi-functional cytosolic protein kinase was a target for PA, and that two NADPH oxidase components were substrates for the enzyme. Partial purification of the PA-activated protein kinase separated the enzyme from known protein kinase targets of PA. The partially purified enzyme was selectively activated by PA, compared to other phospholipids, and phosphorylated the oxidase component p47-phox on both serine and tyrosine residues. PA-activated protein kinase activity was present in a variety of hematopoietic cells and cell lines and in rat brain, suggesting it has widespread distribution. We conclude that this protein kinase may be a novel target for the second messenger function of PA.     

Measurement of 1-aspartamido-beta-N-acetylglucosamine amidohydrolase activity in human tissues.

The activity of 1-aspartamido-beta-N-acetylglucosamine amidohydrolase (aspartylglucosylaminase, EC 3.5.1.26) was measured in normal and diseased human liver, brain and kidney. Organs from patients with aspartylglucosaminuria show very little activity. Crude homogenates of human organs show a reaction catalysed by a complex enzyme system. With homogenate, the formation of product was linear with time up to about 6 h. Reaction times longer than 6-7h resulted in a decrease in the total concentration of product. This phenomenon was not found with the partially purified enzyme fraction. Linearity of the enzyme activity with different protein concentrations was found, independent of the incubation time. Longer incubation of the crude homogenate resulted in the utilization of the product, N-acetylglucosamine. This phenomenon was not observed with the partially purified enzyme fraction. This amidase from human organs differs from that obtained from other sources and apparently represents a rather complex enzyme system.     

Human MD-2 discrimination of meningococcal lipid A structures and activation of TLR4

MD-2, a eukaryotic accessory protein, is an essential component for the molecular pattern recognition of bacterial endotoxins. MD-2 interacts with lipid A of endotoxins [lipopolysaccharide (LPS) or lipooligosaccharide (LOS)] to activate human toll-like receptor (TLR) 4. The structure of lipid A influences the subsequent activation of human TLR4 and the immune response, but the basis for the discrimination of lipid A structures is unclear. A recombinant human MD-2 (rMD-2) protein was produced in the Pichia pastoris yeast expression system. Human embryonic kidney (HEK293) cells were transfected with human TLR4 and were stimulated with highly purified LOS (0.56 pmol) from Neisseria meningitidis or LPS from other structurally defined bacterial endotoxins in the presence or absence of human rMD-2. Human rMD-2 restored, in a dose-dependent manner, interleukin (IL-8) responsiveness to LOS or LPS in TLR4-transfected HEK293 cells. The interaction of endotoxin with human rMD-2 was then assessed by enzyme-linked immunosorbent assays. Wild-type meningococcal LOS (Wt m LOS) bound human rMD-2, and binding was inhibited by an anti-MD-2 antibody to MD-2 dose-dependently (P < 0.005). Wt m LOS or meningococcal KDO(2)-lipid A had the highest binding affinity for human rMD-2; unglycosylated meningococcal lipid A produced by meningococci with defects in the 3-deoxy-d-manno-2-octulosonic acid (KDO) biosynthesis pathway did not appear to bind human rMD-2 (P < 0.005). The affinity of meningococcal LOS with a penta-acylated lipid A for human rMD-2 was significantly less than that for hexa-acylated LOS (P < 0.05). The hierarchy in the binding affinity of different lipid A structures for human rMD-2 was directly correlated with differences in TLR4 pathway activation and cytokine production by human macrophages.     

A tyrosine-specific protein kinase from Ehrlich ascites tumor cells

A protein tyrosine kinase that phosphorylates both alpha and beta subunits of inactivated (Na+,K+)-ATPase from dog kidney was purified about 500-fold from Ehrlich ascites tumor cell membranes. The enzyme required divalent cations Mn2+, Mg2+, or Fe2+ but was inhibited by Cu2+ or Zn2+. The purified enzyme phosphorylated the beta subunit about five times faster than the alpha subunit of the (Na+,K+)-ATPase. The random polymer poly(Glu80Tyr20) was an excellent substrate while casein was only marginally phosphorylated. In contrast, the purified transforming gene product of Rous sarcoma virus phosphorylated all three substrates and the (Na+,K+)-ATPase was preferentially phosphorylated on the alpha subunit. The transforming gene product of Fujinami sarcoma visue and EGF receptor kinase from A431 cells phosphorylated (Na+,K+)-ATPase poorly whereas casein was an excellent substrate. The molecular weight of the partially purified protein tyrosine kinase from Ehrlich ascites tumor cells determined by gel filtration was about 60,000. One of two major phosphorylated phosphopeptides resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis had an Mr of 60 kDa, thus suggesting that it might be the autophosphorylated protein tyrosine kinase. A phosphatase that hydrolyzes phosphorylated histones or poly(Glu80Tyr20) was partially purified from the same membrane.     

Characterization of a soluble glycolipid galactosyltransferase which occurs in bovine milk.

Bovine milk was found to contain, in soluble form, an enzyme which transfers galactose from UDPgalactose to glucosylceramide. This enzyme was partially purified by the same procedure used to isolate the galactosyltransferase of lactose synthetase. The partially purified enzyme required detergents for activity, had a pH optimum of 7.2--7.3 and required Mn2+. The apparent Km calculated for glucosylceramide was 1.33 . 10(-4) M. With glucosylceramide as acceptor the product of the reaction was identified as lactosylceramide by autoradiography on thin-layer chromatograms. Lactosylceramide was also an effective acceptor for the transferase reaction but neutral glycosphingolipids or gangliosides with terminal galactose of N-acetylgalactosamine residues were ineffective or poorly effective as acceptors. Addition of alpha-lactalbumin inhibited the transferase reaction.     

Guanylate cyclase in E. coli. III. Purification and possible physiological role of GTPase]

A phosphohydrolase with a preferential activity for GTP has been isolated and partially purified from E. coli extracts. The enzyme purification has been achieved through precipitation by ammonium sulfate and chromatography on DEAE-cellulose, DEAE-Sephadex, Ultragel and a second DEAE-cellulose column. The phosphohydrolase activity is poly (C) dependent. The chromatographic analysis on PEI-cellulose has shown that the main product of GTP hydrolysis is GDP. The possibility that the enzyme partially purified in this work has an important role in the control of GTP availability as substrate for guanylate cyclase into the cells has been discussed.     

A reciprocal single mutation affects the metal requirement of 3-deoxy-D-manno-2-octulosonate-8-phosphate (KDO8P) synthases from Aquifex pyrophilus and Escherichia coli

The enzyme 3-deoxy-d-manno-2-octulosonate-8-phosphate (KDO8P) synthase is metal-dependent in one class of organisms and metal-independent in another. We have used a rapid transient kinetic approach combined with site-directed mutagenesis to characterize the role of the metal ion as well as to explore the catalytic mechanisms of the two classes of enzymes. In the metal-dependent Aquifex pyrophilus KDO8P synthase, Cys11 was replaced by Asn (ApC11N), and in the metal-independent Escherichia coli KDO8P synthase a reciprocal mutation, Asn26 to Cys, was prepared (EcN26C). The ApC11N mutant retained about 10% of the wild-type maximal activity in the absence of metal ions. Addition of divalent metal ions did not affect the catalytic activity of the mutant enzyme and its catalytic efficiency (kcat/Km) was reduced by only approximately 12-fold, implying that the ApC11N KDO8P synthase mutant has become a bone fide metal-independent enzyme. The isolated EcN26C mutant had similar metal content and spectral properties as the metal-dependent wild-type A. pyrophilus KDO8P synthase. EDTA-treated EcN26C retained about 6% of the wild-type activity, and the addition of Mn2+ or Cd2+ stimulated its activity to approximately 30% of the wild-type maximal activity. This suggests that EcN26C KDO8P synthase mutant has properties similar to that of metal-dependent KDO8P synthases. The combined data indicate that the metal ion is not directly involved in the chemistry of the KDO8P synthase catalyzed reaction, but has an important structural role in metal-dependent enzymes in maintaining the correct orientation of the substrates and/or reaction intermediate(s) in the enzyme active site.     

Specificity and mutational analysis of the metal-dependent 3-deoxy-d-manno-octulosonate 8-phosphate synthase from Acidithiobacillus ferrooxidans

3-Deoxy-d-manno-octulosonate 8-phosphate synthase (KDO8PS) catalyzes the reaction between phosphoenol pyruvate and d-arabinose 5-phosphate to generate KDO8P. This reaction is part of the biosynthetic pathway to 3-deoxy-d-manno-octulosonate, a component of the lipopolysaccharide of the Gram-negative bacterial cell wall. Two distinct groups of KDO8PSs exist, differing by the absolute requirement of a divalent metal ion. In this study Acidithiobacillus ferrooxidans KDO8PS has been expressed and purified and shown to require a divalent metal ion, with Mn²⁺, Co²⁺ and Cd²⁺ (in decreasing order) being able to restore activity to metal-free enzyme. Cd²⁺ significantly enhanced the stability of the enzyme, raising the T_m by 14 °C. d-Glucose 6-phosphate and d-erythrose 4-phosphate were not substrates for A. ferrooxidans KDO8PS, whereas 2-deoxy-d-ribose 5-phosphate was a poor substrate and there was negligible activity with d-ribose 5-phosphate. The ²⁴³AspGlyPro²⁴⁵ motif is absolutely conserved in the metal-independent group of synthases, but the Gly and Pro sites are variable in the metal-dependent enzymes. Substitution of the putative metal-binding Asp243 to Ala in A. ferrooxidans KDO8PS gave inactive enzyme, whereas substitutions Asp243Glu or Pro245Ala produced active enzymes with altered metal-dependency profiles. Prior studies indicated that exchange of a metal-binding Cys for Asn converts metal-dependent KDO8P synthase into a metal-independent form. Unexpectedly, this mutation in A. ferrooxidans KDO8P synthase (Cys21Asn) gave inactive enzyme. This finding, together with modest activity towards 2-deoxy-d-ribose 5-phosphate suggests similarities between the A. ferrooxidans KDO8PS and the related metal-dependent 3-deoxy-d-arabino-heptulosonate phosphate synthase, and highlights the importance of the AspGlyPro loop in positioning the substrate for effective catalysis in all KDO8P synthases.     

Identification and partial purification of a thiol endothelin-converting enzyme from porcine aortic endothelial cells.

Endothelin is a potent peptide vasoconstrictor. The final step in the processing of endothelin has been postulated to be the cleavage of the Trp21-Val22 peptide bond in proendothelin by a putative endothelin-converting enzyme. A soluble extract of primary porcine aortic endothelial cells was found to contain an enzyme activity that converted proendothelin-1 (proET-1) to an endothelin-1 (ET-1)-like peptide as determined by the rabbit aortic ring contraction assay. This enzyme was partially purified by DE52 ion-exchange chromatography. Incubation of proET-1 with the partially purified enzyme generated a product which had a retention time on HPLC identical to that of authentic ET-1. Further analysis of the product showed that it caused contraction of rabbit aortic rings, had a molecular weight identical to ET-1 as measured by fast atom bombardment mass spectrometry, and competed for [125I]ET-1 binding in an RIA using specific antibodies which recognize the carboxy terminal tryptophan of ET-1. The enzyme activity could be inhibited by thiol protease inhibitors such as Z-phe-pheCHN2 and p-hydroxymercuribenzoate, but not by serine- or metalloprotease inhibitors. The optimal pH for the enzymatic activity was between 7.0 and 7.5, and no activity was detected at pH 4.0. These results demonstrate that this thiol protease is a potential endothelin-converting enzyme.