Peter F. Knowles

Crystal structure of a quinoenzyme: copper amine oxidase of Escherichia coli at 2 A resolution.

BACKGROUND Copper amine oxidases are a ubiquitous and novel group of quinoenzymes that catalyze the oxidative deamination of primary amines to the corresponding aldehydes, with concomitant reduction of molecular oxygen to hydrogen peroxide. The enzymes are dimers of identical 70-90 kDa subunits, each of which contains a single copper ion and a covalently bound cofactor formed by the post-translational modification of a tyrosine side chain to 2,4,5-trihydroxyphenylalanine quinone (TPQ). RESULTS The crystal structure of amine oxidase from Escherichia coli has been determined in both an active and an inactive form. The only structural differences are in the active site, where differences in copper coordination geometry and in the position and interactions of the redox cofactor, TPQ, are observed. Each subunit of the mushroom-shaped dimer comprises four domains: a 440 amino acid C-terminal beta sandwich domain, which contains the active site and provides the dimer interface, and three smaller peripheral alpha/beta domains (D1-D3), each of about 100 amino acids. D2 and D3 show remarkable structural and sequence similarity to each other and are conserved throughout the quinoenzyme family. In contrast, D1 is absent from some amine oxidases. The active sites are well buried from solvent and lie some 35 A apart, connected by a pair of beta hairpin arms. CONCLUSIONS The crystal structure of E. coli copper amine oxidase reveals a number of unexpected features and provides a basis for investigating the intriguing similarities and differences in catalytic mechanism of members of this enzyme family. In addition to the three conserved histidines that bind the copper, our studies identify a number of other conserved residues close to the active site, including a candidate for the catalytic base and a fourth conserved histidine which is involved in an interesting intersubunit interaction.

Cooperativity of the phase transition in single- and multibilayer lipid vesicles

The effect of membrane morphology on the cooperativity of the ordered-fluid, lipid phase transition has been investigated by comparing the transition widths in extended, multibilayer dispersons of dimyristoyl phosphatidyl-choline, and also of dipalmitoyl phosphatidylcholine, with those in the small, single-bilayer vesicles obtained by sonication. The electron spin resonance spectra of three different spin-labelled probes, 2,2,6,6-tetramethylpiperdine-N-oxyl, phosphatidylcholine and stearic acid, and also 90 degrees light scattering and optical turbidity measurements were used as indicators of the phase transition. In all cases the transition was broader in the single-bilayer vesicles than in the multibilayer dispersions, corresponding to a decreased cooperativity on going to the small vesicles. Comparison of the light scattering properties of centrifuged and uncentrifuged, sonicated vesicles suggests that these are particularly sensitive to the presence of intermediate-size particles, and thus the spin label measurements are likely to give a more reliable measure of the degree of cooperativity of the small, single-bilayer vesicles. Application of the Zimm and Bragg theory ((1959) J. Chem. Phys. 31, 526-535) of cooperative transitions to the two-dimensional bilayer system shows that the size of the cooperative unit, 1/square root sigma, is a measure of the mean number of molecules per perimeter molecule, in a given region of ordered or fluid lipid at the centre of the transition. From this result it is found that it is the vesicle size which limits the cooperativity of the transition in the small, single-bilayer vesicles. The implications for the effect of membrane structure and morphology on the cooperativity of phase transitions in biological membranes, and for the possibility of achieving lateral communication in the plane of the membrane, are discussed.

ESR spin-label studies of lipid-protein interactions in membranes.

Lipid spin labels have been used to study lipid-protein interactions in bovine and frog rod outer segment disc membranes, in (Na+, K+)-ATPase membranes from shark rectal gland, and in yeast cytochrome oxidase-dimyristoyl phosphatidylcholine complexes. These systems all display a two component ESR spectrum from 14-doxyl lipid spin-labels. One component corresponds to the normal fluid bilayer lipids. The second component has a greater degree of motional restriction and arises from lipids interacting with the protein. For the phosphatidylcholine spin label there are effectively 55 +/- 5 lipids/200,000-dalton cytochrome oxidase, 58 +/- 4 mol lipid/265,000 dalton (Na+, K+)-ATPase, and 24 +/- 3 and 22 +/- 2 mol lipid/37,000 dalton rhodopsin for the bovine and frog preparations, respectively. These values correlate roughly with the intramembrane protein perimeter and scale with the square root of the molecular weight of the protein. For cytochrome oxidase the motionally restricted component bears a fixed stoichiometry to the protein at high lipid:protein ratios, and is reduced at low lipid:protein ratios to an extent which can be quantitatively accounted for by random protein-protein contacts. Experiments with spin labels of different headgroups indicate a marked selectivity of cytochrome oxidase and the (Na+, K+)-ATPase for stearic acid and for cardiolipin, relative to phosphatidylcholine. The motionally restricted component from the cardiolipin spin label is 80% greater than from the phosphatidylcholine spin label for cytochrome oxidase (at lipid:protein = 90.1), and 160% greater for the (Na+, K+)-ATPase. The corresponding increases for the stearic acid label are 20% for cytochrome oxidase and 40% for (Na+, K+)-ATPase. The effective association constant for cardiolipin is approximately 4.5 times greater than for phosphatidylcholine, and that for stearic acid is 1.5 times greater, in both systems. Almost no specificity is found in the interaction of spin-labeled lipids (including cardiolipin) with rhodopsin in the rod outer segment disc membrane. The linewidths of the fluid spin-label component in bovine rod outer segment membranes are consistently higher than those in bilayers of the extracted membrane lipids and provide valuable information on the rate of exchange between the two lipid components, which is suggested to be in the range of 10(6)-10(7) s-1.

One-dimensional electronic conductivity in discotic liquid crystals

Abstract The discotic mesogen 2,3,6,7,10,11-hexa-hexyloxytriphenylene (HAT6) forms a columnar hexagonal phase which is an electrical insulator. Doping with 1 mol% of the Lewis acid AlCl 3 converts it to a p-type semiconductor with the preferred direction of conduction being along the axes of the columns. The electrical conductivity is envisaged to arise from the migration of positive holes created in the π-electron band of the triphenylene stack. The behaviour is established by electrical conductivity measurements, which show anisotropy, and ESR lineshapes which are consistent with Dysons theory of resonance absorption by conduction electrons.

Crystal structure of the precursor of galactose oxidase: An unusual self-processing enzyme

Galactose oxidase (EC 1.1.3.9) is a monomeric enzyme that contains a single copper ion and catalyses the stereospecific oxidation of primary alcohols to their corresponding aldehydes. The protein contains an unusual covalent thioether bond between a tyrosine, which acts as a radical center during the two-electron reaction, and a cysteine. The enzyme is produced in a precursor form lacking the thioether bond and also possessing an additional 17-aa pro-sequence at the N terminus. Previous work has shown that the aerobic addition of Cu2+ to the precursor is sufficient to generate fully processed mature enzyme. The structure of the precursor protein has been determined to 1.4 Å, revealing the location of the pro-sequence and identifying structural differences between the precursor and the mature protein. Structural alignment of the precursor and mature forms of galactose oxidase shows that five regions of main chain and some key residues of the active site differ significantly between the two forms. The precursor structure provides a starting point for modeling the chemistry of thioether bond formation and pro-sequence cleavage.

Protein-immobilized lipid in dimyristoylphosphatidylcholine-substituted cytochrome oxidase: Evidence for both boundary and trapped-bilayer lipid

Abstract Cytochrome oxidase-dimyristoyl phosphatidylcholine complexes have been prepared at defined lipid:protein ratios to study the effects of protein packing density on the lipid fluidity. All the complexes reveal a two-component ESR spectrum from an incorporated phosphatidylcholine spin label, corresponding to both an immobilized lipid boundary layer and fluid bilayer regions. Difference spectra, obtained by subtracting the same immobilized spectrum from the spectra of the various complexes, demonstrate a strong perturbation of the lipid bilayer fluidity which is quite distinct from the immobilized boundary layer formation.

THE DESIGN AND SYNTHESIS OF SIMPLE MOLECULAR TETHERS FOR BINDING BIOMEMBRANES TO A GOLD SURFACE

Abstract Molecular tethers have been synthesised for fixing biomembranes to a gold surface. These are comprised of a thiol at one end to bind to the gold, a polyethylenoxy chain of defined length (two, six or twelve ethylenoxy units) in the middle and a cholesteryl residue at the other end to insert into the biomembrane.

Lipid-substituted cytochrome oxidase: No absolute requirement of cardiolipin for activity

Abstract The endogenous lipid of yeast cytochrome oxidase has been replaced by dimyristoyl phosphatidylcholine. Thin layer chromatography of the total lipid extract from the substituted enzyme revealed phosphatidylcholine only and no cardiolipin. Gas-liquid chromatography showed that >99% of the lipid chains derived from the substituted lipid, and that cardiolipin must be

Binding of water to "types I and II" Cu2+ in proteins.

Water proton spin-lattice relaxation times have been measured at 30MHz between 280 – 333 K in aqueous solutions of proteins containing Type I Cu2+ ions (azurin and umecyanin) and Type II Cu2+ ions (benzylamine oxidase and superoxide dismutase). These measurements show that Type II Cu2+ is accessible to exchangeable water molecules but Type I is not. This behaviour is consistent with the EPR and optical properties of these ions and their likely biochemical functions.

Structure and mechanism of galactose oxidase: catalytic role of tyrosine 495

Abstract The catalytic mechanism of the copper-containing enzyme galactose oxidase involves a protein radical on Tyr272, one of the equatorial copper ligands. The first step in this mechanism has been proposed to be the abstraction of a proton from the alcohol substrate by Tyr495, the axial copper ligand that is weakly co-ordinated to copper. In this study we have generated and studied the properties of a Y495F variant to test this proposal. X-ray crystallography reveals essentially no change from wild-type other than loss of the tyrosyl hydroxyl group. Visible spectroscopy indicates a significant change in the oxidised Y495F compared to wild-type with loss of a broad 810-nm peak, supporting the suggestion that this feature is due to inter-ligand charge transfer via the copper. The presence of a peak at 420 nm indicates that the Y495F variant remains capable of radical formation, a fact supported by EPR measurements. Thus the significantly reduced catalytic efficiency (1100-fold lower kcat / Km) observed for this variant is not due to an inability to generate the Tyr272 radical. By studying azide-induced pH changes, it is clear that the reduced catalytic efficiency is due mainly to the inability of Y495F to accept protons. This provides definitive evidence for the key role of Tyr495 in the initial proton abstraction step of the galactose oxidase catalytic mechanism.