Medium- and short-chain dehydrogenase/reductase gene and protein families : Medium-chain and short-chain dehydrogenases/reductases in retinoid metabolism.

Retinoic acid (RA), the most active retinoid, is synthesized in two steps from retinol. The first step, oxidation of retinol to retinaldehyde, is catalyzed by cytosolic alcohol dehydrogenases (ADHs) of the medium-chain dehydrogenase/reductase (MDR) superfamily and microsomal retinol dehydrogenases (RDHs) of the short-chain dehydrogenase/reductase (SDR) superfamily. The second step, oxidation of retinaldehyde to RA, is catalyzed by several aldehyde dehydrogenases. ADH1 and ADH2 are the major MDR enzymes in liver retinol detoxification, while ADH3 (less active) and ADH4 (most active) participate in RA generation in tissues. Several NAD(+)- and NADP(+)-dependent SDRs are retinoid active. Their in vivo contribution has been demonstrated in the visual cycle (RDH5, RDH12), adult retinoid homeostasis (RDH1) and embryogenesis (RDH10). K(m) values for most retinoid-active ADHs and RDHs are close to 1 microM or lower, suggesting that they participate physiologically in retinol/retinaldehyde interconversion. Probably none of these enzymes uses retinoids bound to cellular retinol-binding protein, but only free retinoids. The large number of enzymes involved in the two directions of this step, also including aldo-keto reductases, suggests that retinaldehyde levels are strictly regulated.

Nucleotide binding and affinity labelling support the existence of the phosphate-binding subsite p2 in bovine pancreatic ribonuclease A.

When the reaction of bovine pancreatic ribonuclease A with 6-chloropurine riboside 5'-monophosphate was carried out in the presence of several natural mononucleotides, a decrease of 25-75% was found in the amount of the reaction product derivative II (the main product of the reaction which has the nucleotide label at the alpha-NH2 group of Lys-1). The efficiency of inhibition followed the order 3'-AMP greater than 5'CMP approximately equal to 5'AMP greater than 3'CMP. Previous studies indicate that this order reflects the extent of occupancy of p2, a phosphate-binding subsite adjacent to the catalytic centre. This finding suggests that derivative II is the result of affinity labelling and that the phosphate group of the halogenated nucleotide binds to p2 before the reaction takes place. The dissociation constants and stoichiometry of the interaction between native enzyme, derivative II and derivative E (homologous to derivative II, but labelled with a nucleoside instead of a nucleotide) with 3'AMP and 5'AMP at several pH values were also determined. Although in general one strong binding site was found, no strong binding occurs between 3'AMP and derivative II. It is concluded that the phosphate of the label occupies the same site p2, as the phosphate of 3'AMP. Finally, the pH dependence for the binding of 3'AMP and 5'AMP to RNAase A indicates that they bind to different protein groups. The results presented support the structure of the active site of ribonuclease A postulated previously (Parés, X., Llorens, R., Arús, C. and Cuchillo, C.M. (1980) Eur. J. Biochem. 105, 571-579).

1H-NMR studies on the binding subsites of bovine pancreatic ribonuclease A.

The titration curves of the C-2 histidine protons of an RNAase derivative (a covalent derivative obtained by reaction of bovine pancreatic RNAase A (EC 3.1.27.5) with 6-chloropurine 9-beta-D-ribofuranosyl 5'-monophosphate) were studied by means of 1H-NMR spectroscopy at 270 MHz. The interaction of natural (5'AMP, 5'GMP, 5'IMP) and halogenated purine mononucleotides (cl6RMP, br8AMP) with RNAase A was also monitored by using the same technique. The slight change observed in the pK values of the active centre histidine residues of the RNAase derivative, with respect to those in the native enzyme, can be considered as evidence that the phosphate of the label does not interact directly either with His-12 or 119 in the p1 site, but the p2 site as proposed previously (Parés, X., Llorens, R., Arús, C. and Cuchillo, C.M. (1980) Eur. J. Biochem. 105, 571--579). Lys-7 and/or Arg-10 are proposed as part of the p2 phosphate-binding subsite. The pK values of His-12 and 119 and the shift of an aromatic resonance of the native enzyme found on interaction with some purine nucleotides, can be interpreted by postulating that the interaction of 5'AMP, 5'GMP and 5'IMP takes place not only in the so-called purine-binding site B2R2p1 but also in the primary pyrimidine-binding site B1R1 and p0 of RNAase A.

Structure and function of ribonuclease A binding subsites.

Ribonuclease A binds nucleic acids through multiple electrostatic interactions between the phosphates of the polynucleotide and the positive groups (side chains of lysines and arginines) of the protein subsites. The bases only play a significant role in the binding at the active site. The active centre p1R1B1 sites determine the specificity of the catalytic cleavage. The phosphate-binding subsites p2 (Lys-7 and Arg-10), p1 (Lys-41, His-12 and His-119) and p0 (Lys-66) are essential for an effective catalysis and are conserved in all mammalian pancreatic ribonucleases. Additional phosphate-binding subsites confer further catalytic efficiency, probably by avoiding non-productive binding. The minimum chain size for optimum catalysis is probably longer than six or seven nucleotides. The full occupancy of binding sites by the long chain polynucleotides would explain the preference of the enzyme for these substrates. The multiplicity of binding subsites is responsible for the helix-destabilizing activity of ribonuclease A. Its capacity for destroying the secondary structure of single-stranded nucleic acids may be of importance for the complete hydrolysis of RNA in the digestive tract. A large variety of proteins, with very different structures and functions, interact with nucleic acids. An analysis of their binding properties shows that there is no general model for protein-nucleic acid interaction. However, the vast amount of work on the ribonuclease A binding subsites should serve as a model for the study of the binding properties of many other proteins that recognize nucleic acids.

Molecular modelling of human gastric alcohol dehydrogenase (class IV) and substrate docking: differences towards the classical liver enzyme (class I).

A three-dimensional model of the human class IV alcohol dehydrogenase has been calculated based upon the X-ray structure of the class I enzyme. As judged from the model, the substrate-binding site is wider than in class I, compatible with the differences in substrate specificities and the large difference in Km value for ethanol. Substrate docking performed for the class I structure and the class IV model show all-trans-retinol and 11-cis-retinol to bind better to the class IV enzyme. The calculations also indicate that 16-hydroxyhexadecanoic acid binds in a different manner for the two enzyme classes. A simulation of coenzyme-binding indicates that the adenine ring of the coenzyme might be differently bound in class IV than in class I, decreasing the interactions with Asp-223 which is compatible with the higher k(cat) values for class IV.

Class IV mammalian alcohol dehydrogenase. Structural data of the rat stomach enzyme reveal a new class well separated from those already characterized.

The stomach form of alcohol dehydrogenase has been structurally evaluated by peptide analysis covering six separate regions of the rat enzyme. Overall, this new structure differs widely (32-40% residue differences) from the structures of three classes of alcohol dehydrogenase characterized before from the same species. Consequently, this novel enzyme constitutes a true fourth class of mammalian alcohol dehydrogenase. In particular, differences are extensive also towards class II, although enzymatic and physicochemical properties initially suggested overall similarities with class II. The new structure establishes the presence of one further alcohol dehydrogenase mammalian gene, extends the enzyme family derived from repeated gene duplications, and confirms tissue-specific expressions.

Cephalopod alcohol dehydrogenase: purification and enzymatic characterization.

Octopus, squid and cuttle-fish organs were examined for alcohol dehydrogenase activity. Only one form was detectable, with properties typical of mammalian class III alcohol dehydrogenase. The corresponding protein was purified from octopus and enzymatically characterized. Ion-exchange and affinity chromatography produced a pure protein in excellent yield (73%) after 1600-fold purification. Enzymatic parameters with several substrates were similar to those for the human class III alcohol dehydrogenase, demonstrating a largely conserved function of the enzyme through wide lines of divergence covering vertebrates, cephalopods and bacteria. The results establish the universal occurrence of class III alcohol dehydrogenase and its strictly conserved functional properties in separate living forms. The absence of other alcohol dehydrogenases in cephalopods is compatible with the emergence of the ethanol-active class I type at a later stage, in lineages leading to vertebrates.

Acetylated N-terminal structures of class III alcohol dehydrogenases. Differences among the three enzyme classes.

The protein chains of mammalian alcohol dehydrogenases typically lack free alpha-amino groups. The blocked N-terminal regions of the class III type of the rat (ADH-2), human (chi chi) and horse enzymes were isolated by digestions with proteases, and characterized by mass-spectrometry supplemented with chemical analysis of the peptides and their redigestion fragments. Results were confirmed by synthesis of the corresponding peptides, followed by chromatographic comparisons of the native and synthetic products. The N-terminal regions of the three class III alcohol dehydrogenase subunits are homologous but differ from the class I and II enzymes in both the exact start position and the amino acid sequence, which suggests that different N-terminal structures are typical for each of the three classes.

Class III alcohol dehydrogenase from Saccharomyces cerevisiae: structural and enzymatic features differ toward the human/mammalian forms in a manner consistent with functional needs in formaldehyde detoxication.

Alcohol dehydrogenase class III (glutathione-dependent formaldehyde dehydrogenase) from Saccharomyces cerevisiae was purified and analyzed structurally and enzymatically. The corresponding gene was also analyzed after cloning from a yeast genome library by screening with a probe prepared through PCR amplification. As with class III alcohol dehydrogenase from other sources, the yeast protein was obtained in two active forms, deduced to reflect different adducts/modifications. Protein analysis established N-terminal and C-terminal positions, showing different and specific patterns in protein start positions between the human/mammalian, yeast, and prokaryotic forms. Km values with formaldehyde differ consistently, being about 10-fold higher in the yeast than the human/mammalian enzymes, but compensated for by similar changes in kcat values. This is compatible with the different functional needs, emphasizing low formaldehyde concentration in the animal cells but efficient formaldehyde elimination in the microorganisms. This supports a general role of the enzyme in formaldehyde detoxication rather than in long-chain alcohol turnover.

Class IV alcohol dehydrogenase (the gastric enzyme). Structural analysis of human sigma sigma-ADH reveals class IV to be variable and confirms the presence of a fifth mammalian alcohol dehydrogenase class.

Human gastric alcohol dehydrogenase (sigma sigma-ADH) was submitted to peptide analysis at picomole scale. A total of 72 positions were determined in the protein chain, providing information on three aspects of alcohol dehydrogenase structures in general. First, the data establish the presence of a unique class of the enzyme, now confirmed as class IV, expressed in gastric tissue and separate from another novel class, now termed class V. Second, the class IV gastric enzyme has active site relationships compatible with an ethanol-active, zinc-containing alcohol dehydrogenase. Third, this enzyme class is of the variable type, like that for the 'variable', classical liver alcohol dehydrogenase of class I, and in contrast to that for the 'constant' class III enzyme. Known human alcohol dehydrogenase structures now prove the presence of at least seven human genes for the enzyme and nine for the whole protein family.

Retinoids, omega-hydroxyfatty acids and cytotoxic aldehydes as physiological substrates, and H2-receptor antagonists as pharmacological inhibitors, of human class IV alcohol dehydrogenase.

Kinetic constants of human class IV alcohol dehydrogenase (sigmasigma-ADH) support a role of the enzyme in retinoid metabolism, fatty acid omega-oxidation, and elimination of cytotoxic aldehydes produced by lipid peroxidation. Class IV is the human ADH form most efficient in the reduction of 4-hydroxynonenal (k(cat)/Km: 39,500 mM(-1) min(-1)). Class IV shows high activity with all-trans-retinol and 9-cis-retinol, while 13-cis-retinol is not a substrate but an inhibitor. Both all-trans-retinoic and 13-cis-retinoic acids are potent competitive inhibitors of retinol oxidation (Ki: 3-10 microM) which can be a basis for the regulation of the retinoic acid generation and of the pharmacological actions of the 13-cis-isomer. The inhibition of class IV retinol oxidation by ethanol (Ki: 6-10 mM) may be the origin of toxic and teratogenic effects of ethanol. H2-receptor antagonists are poor inhibitors of human and rat classes I and IV (Ki > 0.3 mM) suggesting a small interference in ethanol metabolism at the pharmacological doses of these common drugs.

The role of 2',3'-cyclic phosphodiesters in the bovine pancreatic ribonuclease A catalysed cleavage of RNA: intermediates or products?

There is a considerable degree of ambiguity in the literature regarding the role of the 2',3'-cyclic phosphodiesters formed during the reaction of RNA cleavage catalysed by ribonuclease. Usually the reaction is considered to take place in two steps: in the first step there is a transphosphorylation of the RNA 3',5'-phosphodiester bond broken yielding a 2',3'-cyclic phosphodiester which in the second step is hydrolysed to a 3'-nucleotide. Although in many occasions, either explicitly or implicitly, the reaction is treated as taking place sequentially, this is not the case as it has been shown that the 2',3'-phosphodiesters are actually released to the medium as true products of the reaction and that no hydrolysis of these cyclic compounds takes place until all the susceptible 3',5'-phosphodiester bonds have been cyclised. Comparison of the hydrolysis and alcoholysis of the 2',3'-phosphodiesters catalysed by RNase A indicates that the hydrolysis reaction has to be considered formally as a special case of the transphosphorylation back reaction in which the R group of the R-OH substrate is just H. It is thus concluded that the 2',3'-cyclic phosphodiesters formed in the ribonuclease A reaction are true products of the transphosphorylation reaction and not intermediates as usually considered.

Fast atom bombardment mass spectrometry and chemical analysis in determinations of acyl-blocked protein structures.

Peptide generation and fast atom bombardment mass spectrometry in combination with conventional chemical analysis was used to identify the blocking group and establish the N-terminal structure of six different proteins at the nanomole level. In this manner, the first terminal structures of three non-mammalian alcohol dehydrogenases were determined, demonstrating the presence of N-terminal acetylation in these piscine, amphibian, and avian enzymes. Similarly, two different yeast glucose-6-phosphate dehydrogenases and a minor variant of a human alcohol dehydrogenase were found to be acetylated. The exact end location of C-terminal structures was also established. Together, the analyses permit the definition of terminal regions and blocking groups, thus facilitating the delineation of remaining structures.

Alcohol dehydrogenase of human and rat blood vessels. Role in ethanol metabolism.

Alcohol dehydrogenase (ADH) activity has been detected in all arteries and veins examined from humans and rat. In distinct human autopsy vessels, activity values range from 0.9 +/- 0.2 to 9.9 +/- 7.7 mU/mg. Distribution of the activity in human aorta was: intima (23.5%), media (74%) and adventia (2.5%). In most of the samples the beta1 beta1 isozyme of class I ADH was the only form responsible for the ADH activity. Class IV ADH (sigma sigma-ADH) was present in three of the 28 individuals examined. The rat blood vessels showed class IV, but not class I, ADH localized in endothelium and media. The physiological role of vascular ADH is probably related to retinoid metabolism and elimination of lipid peroxidation aldehydes. A contribution to human ethanol metabolism is supported by the significant amount of low-Km activity and the extension of the vascular system.

Immunocytochemical and biochemical demonstration of formaldhyde dehydrogenase (class III alcohol dehydrogenase) in the nucleus.

Alcohol dehydrogenase (ADH), the major enzyme catalyzing the biological oxidation of ethanol in mammals, includes four classes with very different capacities for ethanol oxidation. Class III ADH is present in all the tissues and is well conserved throughout evolution. This enzyme has a low activity with ethanol, is specific for the glutathione-dependent oxidation of formaldehyde, and is therefore a formaldehyde dehydrogenase (FALDH). Until now there have been few and conflicting studies concerning its intracellular distribution, which is important for the understanding of its role in cell function. In the present work we used biochemical and immunocytochemical methods to assess the distribution of FALDH in rat hepatocytes and astroglial cells. With the glutathione-dependent formaldehyde dehydrogenase assay, we found the highest activity in the cytosol of hepatocytes and brain cells (12 and 2.6 mU/mg protein, respectively), but nuclei also exhibited significant activity (1.16 and 2.1 mU/mg protein, respectively). The immunocytochemical results showed the presence of FALDH binding sites in both the cytoplasm and the nucleus of the different cell types studied. Whereas no specific gold particle labeling was seen associated with any cytoplasmic component, in the nucleus the particles were found mainly over condensed chromatin and interchromatin regions. Finally, the gold particle density over both the nucleus and cytoplasm was greater in differentiated than in proliferating astrocytes in primary culture. In contrast, class I ADH, primarily responsible for ethanol metabolism, was found only in the cytoplasm of hepatocytes. We propose that one of the functions of FALDH is to protect cell structures, including DNA, from the toxic effects of endogenous formaldehyde, which is an intermediate in many metabolic process.

Alcohol dehydrogenase isoenzymes in rat development. Effect of maternal ethanol consumption.

The alcohol dehydrogenase (ADH) isoenzymes (alcohol:NAD oxidoreductase, EC 1.1.1.1) of classes I, III and IV were investigated by activity and starch gel electrophoresis analyses during rat ontogeny. Class I was studied in the liver, class III in the brain and class IV in the stomach and eyes. Classes I and IV exhibited very low activity during the fetal period, reaching 12% and 3%, respectively, of the adult value at birth. Class III was relatively more active in the fetus, with 38% of the adult activity at birth. In the three cases, activity increased after birth and adult values were found around day 20 (classes I and III), day 39 (stomach class IV) and after day 91 (eye class IV). The very low activity of the isoenzymes responsible for ethanol oxidation, i.e. liver class I and stomach class IV, in the fetus demonstrates that metabolism of ethanol during gestation is essentially performed by the maternal tissues. Development of ADH isoenzymes were also studied in the offspring of rats exposed to an alcoholic liquid diet. Activities of liver class I and stomach class IV were severely reduced: they were only 30% and 50%, respectively, of the control values. In contrast, eye class IV activity did not change and brain class III showed a 30% increase. Moreover, the concentration of liver soluble protein exhibited a 1.3-1.5-fold increase with respect to control animals. The effects on activities and liver protein were more pronounced in the adult than in the perinatal period, and they seem irreversible since normal values were not recovered after 6 weeks of feeding with a non-alcoholic diet. The low activities of the alcohol-oxidizing isoenzymes indicate tht maternal ethanol consumption results in an impaired ethanol metabolism of the offspring.

Recommended nomenclature for the vertebrate alcohol dehydrogenase gene family.

The alcohol dehydrogenase (ADH) gene family encodes enzymes that metabolize a wide variety of substrates, including ethanol, retinol, other aliphatic alcohols, hydroxysteroids, and lipid peroxidation products. Studies on 19 vertebrate animals have identified ADH orthologs across several species, and this has now led to questions of how best to name ADH proteins and genes. Seven distinct classes of vertebrate ADH encoded by non-orthologous genes have been defined based upon sequence homology as well as unique catalytic properties or gene expression patterns. Each class of vertebrate ADH shares <70% sequence identity with other classes of ADH in the same species. Classes may be further divided into multiple closely related isoenzymes sharing >80% sequence identity such as the case for class I ADH where humans have three class I ADH genes, horses have two, and mice have only one. Presented here is a nomenclature that uses the widely accepted vertebrate ADH class system as its basis. It follows the guidelines of human and mouse gene nomenclature committees, which recommend coordinating names across species boundaries and eliminating Roman numerals and Greek symbols. We recommend that enzyme subunits be referred to by the symbol "ADH" (alcohol dehydrogenase) followed by an Arabic number denoting the class; i.e. ADH1 for class I ADH. For genes we recommend the italicized root symbol "ADH" for human and "Adh" for mouse, followed by the appropriate Arabic number for the class; i.e. ADH1 or Adh1 for class I ADH genes. For organisms where multiple species-specific isoenzymes exist within a class, we recommend adding a capital letter after the Arabic number; i.e. ADH1A, ADH1B, and ADH1C for human alpha, beta, and gamma class I ADHs, respectively. This nomenclature will accommodate newly discovered members of the vertebrate ADH family, and will facilitate functional and evolutionary studies.

Kinetic effects of a single-amino acid mutation in a highly variable loop (residues 114-120) of class IV ADH.

Class IV alcohol dehydrogenase shows a deletion at position 117 with respect to class I enzymes, which typically have a Gly residue. In class I structures, Gly117 is part of a loop (residues 114-120) that is highly variable within the alcohol dehydrogenase family. A mutant human class IV enzyme was engineered in which a Gly residue was inserted at position 117 (G117ins). Its kinetic properties, regarding ethanol and primary aliphatic alcohols, secondary alcohols and pH profiles, were determined and compared with the results obtained in previous studies in which the size of the 114-120 loop was modified. For the enzymes considered, a smaller loop was associated with a lower catalytic efficiency towards short-chain alcohols (ethanol and propanol) and secondary alcohols, as well as with a higher K(m) for ethanol at pH 7.5 than at pH 10.0. The effect can be rationalized in terms of a more open, solvent-accessible active site in class IV alcohol dehydrogenase, which disfavors productive binding of ethanol and short-chain alcohols, specially at physiological pH.

Characterization and functional role of Saccharomyces cerevisiae 2,3-butanediol dehydrogenase.

Using a conserved sequence motif, a new gene (YAL060W) of the MDR family has been identified in Saccharomyces cerevisiae. The expressed protein was a stereoespecific (2R,3R)-2,3-butanediol dehydrogenase (BDH). The best substrates were (2R,3R)-2,3-butanediol for the oxidation and (3R/3S)-acetoin and 1-hydroxy-2-propanone for the reduction reactions. The enzyme is extremely specific for NAD(H) as cofactor, probably because the presence of Glu223 in the cofactor binding site, instead of the highly conserved Asp223. BDH is inhibited competitively by 4-methylpyrazole with a K(i) of 34 microM. Yeast could grow on 2,3-butanediol or acetoin as a sole energy and carbon sources, and a 3.6-fold increase in BDH activity was observed when cells were grown in 2,3-butanediol, suggesting a role of the enzyme in 2,3-butanediol metabolism. However, the disruption of the YAL060W gene was not lethal for the yeast under laboratory conditions, and the disrupted strain could also grow in 2,3-butanediol and acetoin. This suggests that other enzymes, in addition to BDH, can also metabolize 2,3-butanediol in yeast.

Properties of rat retina alcohol dehydrogenase.

The rat eye fraction, including retina, pigment epithelium and choroid, contains an alcohol dehydrogenase (ADH) isoenzyme that is not present in rat liver. Starch gel electrophoresis of retina ADH shows an anodic band that can be visualized by activity staining, using either ethanol or pentanol as substrates. Ethanol is a poor substrate (Km: 336 mM, at pH 10.0) for the purified retina ADH which prefers long chain, 2-unsaturated and aromatic alcohols. The enzyme has a pH optimum of 10.0 for ethanol oxidation and it is inhibited by 4-methylpyrazole (KI: 10 microM). Electrophoretic and kinetic properties clearly differentiate the retina ADH from the hepatic cathodic ADH isoenzymes and from an anodic chi-ADH-like form that we have also detected in rat liver. At the pH and ethanol concentrations found "in vivo," retina ADH can oxidize ethanol to an appreciable extent. The subsequent production of acetaldehyde and redox change may be responsible for visual disorders during alcohol intoxication.