The monomer structure is the same in both the complexed and uncomplexed crystal forms. The dimers differ in the relative positions of the two monomers at the dimer interface. Of the 55 residues that are different in CE from those in C. rugosa lipase 1, 23 are located in the active site and at the dimer interface. The altered substrate specificity is a direct consequence of these substitutions.

Conformational models of the three characterized classes of mammalian liver alcohol dehydrogenase were constructed using computer graphics based on the known three-dimensional structure of the E subunit of the horse enzyme (class 1) and the primary structures of the three human enzyme classes. This correlates the substratebinding pockets of the class I subunits (a, and y in the human enzyme) with those of the class I1 and I11 subunits (z and x, respectively) for three enzymes that differ in substrate specificity, inhibition pattern and many other properties. The substrate-binding sites exhibit pronounced differences in both shape and properties. Comparing human class I subunits with those of class 11 and I11 subunits there are no less than 8 and 10 replacements, respectively, out of 11 residues in the substrate pocket, while in the human class I isozyme variants, only 1 -3 of these 11 positions differ. A single residue, Va1294, is conserved throughout. The liver alcohol dehydrogenases, with different substrate-specificity pockets, resemble the patterns of other enzyme families such as the pancreatic serine proteases.The inner part of the substrate cleft in the class I1 and I11 enzymes is smaller than in the horse class I enzyme, because both Ser48 and Phe93 are replaced by larger residues, Thr and Tyr, respectively. In class 11, the residues in the substrate pocket are larger in about half of the positions. It is rich in aromatic residues, four Phe and one Tyr, making the substrate site distinctly smaller than in the class I subunits. In class Ill, the inner part of the substrate cleft is narrow but the outer part considerably wider and more polar than in the class I and I1 enzymes. In addition, Ser (or Thr) and Tyr in class I1 and I11 instead of His51 may influence proton abstraction/donation at the active site.Mammalian zinc-containing alcohol dehydrogenases constitute an enzyme family of multiple forms. Subunit types a, ,/I and y [l] in dimeric combinations constitute the isozymes of the human class I enzyme [2] and are all homologous to the E subunit of the horse EE isozyme [3], the only alcohol dehydrogenase crystallographically analyzed [4]. The class I1 and I11 enzymes differ considerably in primary structure [5 -81, constituting essentially separate and distinct enzymes [9] with different evolutionary rates [S].Previous model-building studies have shown large similarities between the isozymes within class I and have explained the consequences of the replacements that occur [lo]. These residue exchanges are few but have effects on substrate [lo] and coenzyme [113 binding. The inter-class differences are large and affect charge, enzyme activity, inhibition pattern, and other properties utilized for detection and purification [2]. Ethanol at 5 mM saturates the traditional class I enzymes, while at fhis concentration class I1 contributes less than 15% to the total ethanol oxidation of the liver [12]. Class I11 of human liver alcohol dehydrogenase, with x subunits, is almost inactive towards ethanol and ev...

The primary structure of class III alcohol dehydrogenase (dimeric with chi subunits) from human liver has been determined by peptide analyses. The protein chain is a clearly distinct type of subunit distantly related to those of both human class I and class II alcohol dehydrogenases (with alpha, beta, gamma, and pi subunits, respectively). Disregarding a few gaps, residue differences in the chi protein chain with respect to beta 1 and pi occur at 139 and 140 positions, respectively. Compared to class I, the 373-residue chi structure has an extra residue, Cys after position 60, and two missing ones, the first two residues relative to class I, although the N-terminus is acetylated like that for those enzymes. The chi subunit contains two more tryptophan residues than the class I subunits, accounting for the increased absorbance at 280 nm. There are also four additional acidic and two fewer basic side chains than in the class I beta structure, compatible with the markedly different electrophoretic mobility of the class III enzyme. Residue differences between class III and the other classes occur with nearly equal frequency in the coenzyme-binding and catalytic domains. The similarity in the number of exchanges relative to that of the enzymes of the other two classes supports conclusions that the three classes of alcohol dehydrogenase reflect stages in the development of separate enzymes with distinct functional roles. In spite of the many exchanges, the residues critical to basic functional properties are either completely unchanged--all zinc ligands and space-restricted Gly residues--or partly unchanged--residues at the coenzyme-binding pocket.(ABSTRACT TRUNCATED AT 250 WORDS)

We here report on the purification and characterization of glucose-1-phosphate thymidylyltransferase, the first of four enzymes commited to biosynthesis of dTDP-L-rhamnose from Salmonella enterica strain LT2. The purification was greatly facilitated by the cloning of the IjhA gene encoding this enzyme. Pure enzyme was obtained by 109-fold enrichment in three chromatography steps.The glucose-1 -phosphate thymidylyltransferase catalyzes a reversible bimolecular group transfer reaction and kinetic measurements indicate that it acts by a 'ping-pong' mechanism. The K , values for dTTP and a-D-glucose 1-phosphate in the forward reaction are 0.020 mM and 0.11 mM, respectively. In the reverse reaction the K , values for dTDP-D-glucose and diphosphate are 0.083 mM and 0.15 mM, respectively. The enzyme also accepts UTP and UDP-D-glucose and a-D-glucosamine 1-phosphate is accepted equally as well as a-D-glucose 1-phosphate.The NH,-terminal sequence of glucose-1 -phosphate thymidylyltransferase agrees with the sequence predicted from the nucleotide sequence of the 0lf6.l gene of the rj& gene cluster. The SDS/ PAGE estimated subunit mass of 31 kDa agrees well with that calculated from the amino acid composition deduced from the nucleotide sequence of the 0rf6.l gene (32453 Da).Saccharides are important surface structures in eukaryotic and prokaryotic cells and involved in cell recognition phenomena. They occur as free saccharides and as glycoconjugates linked to proteins and lipids. One of the problems in studies of their activities is the difficulty in obtaining them in pure form in sufficient quantities. In recent years advances have been made in the chemical synthesis of saccharides [l]. However, saccharide synthesis is still a relatively complicated procedure, in particular when larger saccharides or large quantities are desirable. In the last few years we have been investigating the feasibility of in vitro enzymatic synthesis of the Salmonella enterica 0-antigen-specific oligosaccharide and the nucleotide sugar precursors using enzymes overexpressed by cloned genes [Z].The 0-antigen part of S. enterica sero-group B strains is a repeating tetramer oligosaccharide composed of abequose, mannose, rhamnose and galactose. This saccharide is assembled from appropriate nucleoside &phosphate monosaccharides. The enzymes which participate in the biosynthesis of S. enterica 0-antigen polysaccharide are encoded by genes Abbreviations. Buffer A, 50mM Tris/HCl pH7.6, 10mM MgCl, and 1 mM EDTA; buffer B, 20 mM Tris/HC pH 8.0, 1 mM MgC1, and 22% glycerol; buffer C, 50mM sodium phosphate pH 7.0, 1 M ammonium sulphate and 22% glycerol.Enzymes. Glucose-I-phosphate thymidylyltransferase (EC 2.7.7.24), inorganic pyrophoshatase (EC 3.6.1.1).which are located mostly in the rj8 gene cluster [3]. The rj& gene cluster has been cloned [4-61, and Escherichia coli K12 strains harboring plasmids containing different parts of the rj8 gene cluster of S. enten'ca LT2 was the source of the enzymes used for in vitro enzymatic synthesis of dTDP-Lrhamnose, ...

The 500-residue amino acid sequence of the subunit of mitochondrial human liver aldehyde dehydrogenase is reported. It is the first structure determined for this enzyme type from any species, and is based on peptides froin treatments with trypsin, CNBr, staphylococcal Glu-specific protease, and hydroxylamine. The chain is not blocked (in contrast to that of the acetylated cytosolic enzmye form), but shows N-terminal processing heterogeneity over the first seven positions. Otherwise, no evidence for subunit microheterogeneities was obtained. The structure displays 68% positional identity with that of the corresponding cytosolic enzyme, and comparisons allow functional interpretations for several segments.A region with segments suggested to participate in coenzyme binding is the most highly conserved long segment of the entire structure (positions 194 -274). Cys-302, identified in the cytosolic enzyme in relation to the disulfiram reaction, is also present in thc mitochondrial enzyme. A new model of the active site appears possible and involves a hydrophobic cleft. Near-total lack ofconservation of the N-terminal segments may reflect a role of the N-tcrminal region in signaling the transport of the mitochondrial protein chains. Non-conservaiion of interior regions may reflect the differences between the two enzyme forms in subunit interactions, explaining the lack of heterotetrameric molecules. The presence of some internal repeat structures is also noted as well as apparently general features of differences between cytosolic and mitochondrial enzymes.