Activities of human alcohol dehydrogenases in the metabolic pathways of ethanol and serotonin.

Alcohols and aldehydes in the metabolic pathways of ethanol and serotonin are substrates for alcohol dehydrogenases (ADH) of class I and II. In addition to the reversible alcohol oxidation/aldehyde reduction, these enzymes catalyse aldehyde oxidation. Class-I gammagamma ADH catalyses the dismutation of both acetaldehyde and 5-hydroxyindole-3-acetaldehyde (5-HIAL) into their corresponding alcohols and carboxylic acids. The turnover of acetaldehyde dismutation is high (kcat = 180 min-1) but saturation is reached first at high concentrations (Km = 30 mm) while dismutation of 5-HIAL is saturated at lower concentrations and is thereby more efficient (Km = 150 microm; kcat = 40 min-1). In a system where NAD+ is regenerated, the oxidation of 5-hydroxytryptophol to 5-hydroxyindole-3-acetic acid proceeds with concentration levels of the intermediary 5-HIAL expected for a two-step oxidation. Butanal and 5-HIAL oxidation is also observed for class-I ADH in the presence of NADH. The class-II enzyme is less efficient in aldehyde oxidation, and the ethanol-oxidation activity of this enzyme is competitively inhibited by acetate (Ki = 12 mm) and 5-hydroxyindole-3-acetic acid (Ki = 2 mm). Reduction of 5-HIAL is efficiently catalysed by class-I gammagamma ADH (kcat = 400 min-1; Km = 33 microm) in the presence of NADH. This indicates that the increased 5-hydroxytryptophol/5-hydroxyindole-3-acetic acid ratio observed after ethanol intake may be due to the increased NADH/NAD+ ratio on the class-I ADH.

A new HPLC-based method for the quantitative analysis of inner stratum corneum lipids with special reference to the free fatty acid fraction.

The inner stratum corneum is likely to represent the location of the intact skin barrier, unperturbed by degradation processes. In our studies of the physical skin barrier a new high-performance liquid chromatography (HPLC)-based method was developed for the quantitative analysis of lipids of the inner stratum corneum. All main lipid classes were separated and quantitated by HPLC/light scattering detection (LSD) and the free fatty acid fraction was further analysed by gas-liquid chromatography (GLC). Mass spectrometry (MS) was used for peak identification and flame ionization detection (FID) for quantitation. Special attention was paid to the free fatty acid fraction since unsaturated free fatty acids may exert a key function in the regulation of the skin barrier properties by shifting the physical equilibrium of the multilamellar lipid bilayer system towards a noncrystalline state. Our results indicated that the endogenous free fatty acid fraction of the stratum corneum barrier lipids in essence exclusively consisted of saturated long-chain free fatty acids. This fraction was characterized as a very stable population (low interindividual peak variation) dominated by saturated lignoceric acid (C24:0, 39 molar%) and hexacosanoic acid (C26:0, 23 molar%). In addition, trace amounts of very long-chain (C32-C36) saturated and monounsaturated free fatty acids were detected in human forearm inner stratum corneum. Our analysis method gives highly accurate and precise quantitative information on the relative composition of all major lipid species present in the skin barrier. Such data will eventually permit skin barrier model systems to be created which will allow a more detailed analysis of the physical nature of the human skin barrier.

Coupling of ethanol metabolism to lipid biosynthesis: labelling of the glycerol moieties of sn-glycerol-3-phosphate, a phosphatidic acid and a phosphatidylcholine in liver of rats given [1,1-2H2]ethanol.

The mechanism behind ethanol-induced fatty liver was investigated by administration of [1,1-2H2]ethanol to rats and analysis of intermediates in lipid biosynthesis. Phosphatidic acid and phosphatidylcholine were isolated by chromatography on a lipophilic anion exchanger and molecular species were isolated by high-performance liquid chromatography in a non-aqueous system. The glycerol moieties of palmitoyl-linoleoylphosphatidic acid, the corresponding phosphatidylcholine and free sn-glycerol-3-phosphate were analysed by GC/MS of methyl ester t-butyldimethylsilyl derivatives. The deuterium labelling in the glycerol moiety of the phosphatidic acid was 2-3-times higher than in free sn-glycerol-3-phosphate, indicating that a specific pool of sn-glycerol-3-phosphate was used for the synthesis of phosphatidic acid in liver. The results indicate that NADH formed during ethanol oxidation is used in the formation of a pool of sn-glycerol-3-phosphate that gives rise to triacylglycerol and possibly fatty liver.

Aldehyde dismutase activity of human liver alcohol dehydrogenase.

Human alcohol dehydrogenases of class I and class II but not class III catalyse NAD+-dependent aldehyde oxidation in addition to the NADH-dependent aldehyde reduction. The two reactions are coupled, i.e. the enzymes display dismutase activity. Dismutase activity of recombinantly expressed human class I isozymes beta1beta1 and gamma2gamma2, class II and class III alcohol dehydrogenases was assayed with butanal as substrate by gas chromatographic-mass spectrometric quantitations of butanol and butyric acid. The class I gamma2gamma2 isozyme showed a pronounced dismutase activity with a high kcat, 1300 min(-1), and a moderate Km, 1.2 mM. The class I beta1beta1 isozyme and the class II alcohol dehydrogenase showed moderate catalytic efficiencies for dismutase activity with lower kcat values, 60-75 min(-1). 4-Methylpyrazole, a potent class I ADH inhibitor, inhibited the class I dismutation completely, but cyanamide, an inhibitor of mitochondrial aldehyde dehydrogenase, did not affect the dismutation. The dismutase reaction might be important for metabolism of aldehydes during inhibition or lack of mitochondrial aldehyde dehydrogenase activity.

Metabolic profiles of steroids in urine of alcoholics after withdrawal.

The metabolic profiles of steroids in urine were analyzed in 13 male alcoholics during long-term abstinence, in most cases exceeding 3 months. The ratios of 5 beta- to 5 alpha-reduced steroid metabolites (etiocholanolone/androsterone and tetrahydrocortisol/allotetrahydrocortisol) were initially elevated but decreased slowly following withdrawal. The half-life of this normalization exceeded 3 weeks. The change was most marked in patients with signs of liver injury, and may reflect a relative decrease of the activity of hepatic 5 alpha-reductase. The ratio between cortisol metabolites carrying a 11 beta-hydroxy and an 11-oxo group was elevated in the patients and showed no tendency to normalize. This might reflect a decrease in the peripheral inactivation of cortisol.

Inhibition of aldehyde dehydrogenases by methylene blue.

The effect of the redox dye methylene blue on the stability of NADH and on the activity of the enzyme aldehyde dehydrogenase (ALDH; EC 1.2.1.3) was examined. NADH was measured by HPLC with fluorometric or spectrophotometric detection. The ALDH activity assays were carried out by following the formation of 3,4-dihydroxyphenylacetic acid (DOPAC) from 3,4-dihydroxyphenylacetaldehyde (DOPAL) using HPLC and electrochemical detection. Incubation of NADH solutions in the presence of methylene blue resulted in a time-dependent direct oxidation of NADH. Methylene blue inhibited the human erythrocyte and leukocyte ALDHs and the rat liver mitochondrial low-Km ALDH in a concentration-dependent manner. The inactivation was reversible by dilution, and kinetic analysis indicated that methylene blue inhibits the rat liver mitochondrial low-Km and human erythrocyte ALDHs competitively with respect to DOPAL, while no effect of the NAD+ concentration was apparent. For the rat liver low-Km ALDH, a Ki of 8.4 +/- 2.8 microM (mean +/- SD; N = 5) was calculated. The inhibition of ALDH and the resulting decrease in the redox effect on the NAD system bound to alcohol dehydrogenase (EC 1.1.1.1) could explain the protective effect of methylene blue against metabolic redox effects of ethanol.

Dehydrogenase-dependent metabolism of alcohols in gastric mucosa of deer mice lacking hepatic alcohol dehydrogenase.

Deer mice (Peromyscus maniculatus) lacking hepatic alcohol dehydrogenase (ADH) have been used as a model for studies of ethanol elimination catalysed by non-ADH systems like catalase and cytochrome P450. However, in an in vivo study on these animals (ADH- deer mice), we detected reversibility in the oxidation of [2H]ethanol, indicating that a major part of the ethanol elimination was due to a dehydrogenase (Norsten et al., J Biol Chem 264: 5593-5597, 1989). In the present investigation, we found significant ethanol oxidizing activity in the gastric mucosa of the deer mice. Reversibility was demonstrated by the use of [2H]acetaldehyde and gas chromatography-mass spectrometry of the products. The kinetic 2H isotope effect of the gastric system was about 3.0 and the system was comparatively insensitive to inhibition by 4-methylpyrazole. The behavior of the deer mice gastric ADH in isoelectric focusing and its higher activity with longer alcohols as substrates indicated similarity with the previously described human class IV enzymes. Our data are in agreement with results obtained in vivo and indicate that ethanol is oxidized extrahepatically in ADH- deer mice. This has to be taken into account when deer mice are used to study non-ADH-dependent ethanol oxidation in vivo.

Ethanol metabolism in isolated hepatocytes. Effects of methylene blue, cyanamide and penicillamine on the redox state of the bound coenzyme and on the substrate exchange at alcohol dehydrogenase.

Ethanol metabolism in hepatocytes increases the NADH/NAD+ ratio. The mechanism was investigated by measurements of the redox state of the coenzyme bound to alcohol dehydrogenase and of ethanol-acetaldehyde exchange and concomitant hydrogen transfer between ethanol molecules. Isolated hepatocytes from fed rats were incubated with cyclohexanone and cyclohexanol or with [1,1-2H2]-and [2,2,2-2H3]ethanol, followed by gas chromatographic determination of the redox state and isotope analysis of the ethanol by gas chromatography-mass spectrometry, respectively. Cyanamide and methylene blue decreased the redox shift caused by ethanol and increased the rates of acetaldehyde reduction during the exchange. Both drugs increased the extent of hydrogen transfer between ethanol molecules during oxidoreduction. Penicillamine had no significant effect on the ethanol-induced change in redox state of the bound coenzyme although it decreased the rate of acetaldehyde reduction. The results indicate that methylene blue inhibits aldehyde dehydrogenase and that accumulation of acetaldehyde decreases the redox effects of ethanol. The redox effect appears to result primarily from rapid elimination of acetaldehyde and equilibration with the NAD system on the alcohol dehydrogenase, but is not enhanced by further decreases in acetaldehyde concentration. Thus, penicillamine could probably be used to decrease the concentration of acetaldehyde without increasing the redox effects.

Decreased content of arachidonoyl species of phosphatidylinositol phosphates in pancreas of rats fed on an ethanol-containing diet.

Phosphatidylinositol (PtdIns), phosphatidylinositol 4-phosphate (PtdIns4P) and phosphatidylinositol 4,5-diphosphate [PtdIns(4,5)P2] were isolated from the pancreas of rats fed an ethanol-containing liquid diet for 24 days and from the corresponding pair-fed controls. The isolation involved chromatography on a lipophilic anion exchanger in the phosphate form. The species composition was determined by fast-atom bombardment mass spectrometry. The compositions of PtdIns4P and PtdIns(4,5)P2 were similar to that of PtdIns, with the stearoyl/arachidonoyl species constituting about 32% of the total, compared with 38% in PtdIns. PtdIns(4,5)P2 contained a larger fraction of fully saturated species than PtdIns. The PtdIns species having an arachidonoyl group were about half as abundant in the ethanol-treated as in the control rats. The differences between ethanol-fed and control rats were qualitatively similar for PtdIns, PtdIns4P and PtdIns(4,5)P2, but were less marked for PtdIns4P and PtdIns(4,5)P2. The results indicate that the species compositions of PtdIns4P and PtdIns(4,5)P2 reflect that of PtdIns, and that the changes of PtdIns4P and PtdIns(4,5)P2 in ethanol-treated rats are secondary to changes in PtdIns.

Transfer of deuterium from [1,1-2H2]ethanol to steroids and organic acids in the rat testis.

Rats were given [1,1-2H2]ethanol in a single dose, and the 2H content was determined in testicular steroids and in organic acids of low molecular mass in the testis, liver and blood. The acids were quantified by capillary gas chromatography/mass spectrometry of t-butyldimethylsilyl derivatives with [2H4]lactate as internal standard. In addition to lactate, pyruvate, 3-hydroxybutyrate and acids of the tricarboxylic acid cycle, the testis was shown to contain 2-hydroxybutyrate, 2-hydroxy-2-methylbutyrate, 2-hydroxyisohexanoate and glycerate. No 2H was found in pregnenolone, 5-androstene-3 beta,17 beta-diol or testosterone, whereas the abundance of monodeuterated molecules of 5 alpha-androstane-3 alpha,17 beta-diol and its 3 beta-isomer were 7.6% and 11.2% respectively. The abundance of monodeuterated lactate was 7.0% in the testis and 5.3% in the blood. The other acids were less labelled but 3-hydroxybutyrate had a higher 2H content in the testis (3.1%) than in the liver. These results support the contention that ethanol is oxidized in an alcohol dehydrogenase-catalysed reaction in testis in vivo and that the acute inhibition of the testosterone production is due at least partly to a redox effect. The labelling and increased concentration of 3-hydroxybutyrate in the testis indicate that a change in the mitochondrial redox state might be involved.

Oxidoreduction of butanol in deermice (Peromyscus maniculatus) lacking hepatic cytosolic alcohol dehydrogenase.

In view of conflicting information in the literature regarding enzyme systems responsible for alcohol oxidation in deermice previously reported to lack hepatic alcohol dehydrogenase (ADH) activity, the reversibility of butanol oxidation was studied in vivo and in liver-perfusion systems. Mixtures of [1,1-2H2]ethanol and butanol were given intraperitoneally to deermice lacking (ADH-) or possessing (ADH+) ADH activity, followed by analysis of alcohols in blood by GC/MS. 2H exchange between the two alcohols was seen in all experiments. In ADH- deermice, the 2H excess of butanol increased steadily and reached 18 +/- 5% after 2.5 h. In ADH+ deermice, butanol was rapidly eliminated and the 2H excess was about 7% after 0.5 h. In similar experiments with rats, the 2H excess was about 40% for 2 h. Perfusions of livers from ADH- deermice with mixtures of unlabelled and 1-[2H]butanol showed significant but slow intermolecular hydrogen transfer at C1, indicating oxidoreduction catalyzed by a dehydrogenase. Slow reduction of butanal was observed in mitochondria from ADH- deermice. ADH activity with a pH optimum of 10 and Km for ethanol of 6 mM was detected in the inner mitochondrial membranes from rats and deermice. However, low rates of oxidation observed in experiments carried out with perfused livers and in vitro suggest that this enzyme system does not contribute significantly to alcohol oxidation in vivo. Thus, perfused liver from ADH- deermice appears to be a useful system for studies of ADH-independent oxidation of alcohols. The 2H exchange between the alcohols seen in vivo indicates that both ethanol and butanol are substrates for a common extrahepatic dehydrogenase in ADH- deermice.

Acetaldehyde as a substrate for ethanol-inducible cytochrome P450 (CYP2E1).

Liver microsomes from starved and acetone-treated rats catalyzed NADPH-supported metabolism of acetaldehyde at a rate 8-fold higher than corresponding control microsomes; the Vmax was about 6 nmol/mg microsomal protein/min and the apparent Km 30 microM. The reaction was efficiently inhibited by anti-CYP2E1 IgG, but not by control IgG. Reconstituted membranes containing rat CYP2E1 and cytochrome b5 metabolized acetaldehyde with a Vmax of 20 nmol/nmol/min and an apparent Km of 30 microM, whereas CYP2B4 containing vesicles or vesicles without b5 were ineffective. Gas chromatographic/mass spectrometric analysis of products formed from [2H4]-acetaldehyde with CYP2E1-containing reconstituted membrane vesicles revealed the formation of acetate as the only detectable product, although other water soluble products were also formed as evidenced from incubations with [1,2-14C]acetaldehyde. The results indicate that CYP2E1 is an aldehyde oxidase and thus metabolizes both ethanol and its primary oxidation product. This might have implications in vivo for acetaldehyde metabolism in liver and brain.

Stable isotope studies on steroid metabolism and kinetics: sulfates of 3 alpha-hydroxy-5 alpha-pregnane derivatives in human pregnancy.

The metabolism and production rates of 3 alpha-hydroxy-5 alpha-pregnan-20-one sulfate and the 3-sulfate and 3,20-disulfate of 5 alpha-pregnane-3 alpha,20 alpha-diol in pregnant women were studied. The steroid sulfates were labeled with deuterium in the 3 beta,11,11- or 3 beta,11,11,20 beta-positions and were injected intravenously. The deuterium content of steroids in the monosulfate and disulfate fraction of plasma collected at different times after the injection was determined by capillary column gas chromatography/mass spectrometry. The injected steroid sulfates underwent oxidoreduction at C-20 and 16 alpha-hydroxylation. In addition, the 3-sulfate of 5 alpha-pregnane-3 alpha,20 alpha-diol became hydroxylated at C-21. The pregnanediol and pregnanetriol monosulfates were also converted to disulfates. No evidence was obtained for a metabolic sequence involving hydrolysis, oxidoreduction, and resulfation at the C-3 position. Production rates and rates of metabolic transformations were determined using different one- and two-pool models. The production rate of the pregnanolone/pregnanediol monosulfate couple was 0.08 to 0.5 mmol/24 h, the variability probably depending both on individual factors and stage of pregnancy. The half-life time for oxidation and reduction at C-20 was 0.1 to 0.4 hours, reduction being the faster process. The half-life time for the turnover of the steroid skeleton was 1.3 to 3.3 hours. The injected steroid monosulfates were 16 alpha-hydroxylated at a rate of 1 to 8 mumol/24 h. A significant fraction of these 16 alpha-hydroxylated steroid sulfates, 0.5 to 25 mumol/24 h, was formed from other, probably unconjugated, precursors. The 16 alpha-hydroxylated steroid monosulfates underwent rapid oxidoreduction at C-20. The 3-sulfate of 5 alpha-pregnane-3 alpha,20 alpha-diol was hydroxylated at C-21. The production rate of 5 alpha-pregnane-3 alpha,20 alpha,21-triol 3-sulfate was 8 to 36 mumol/24 h in four women and 180 mumol/24 h in one woman, and this steroid was not formed from other precursors to a significant extent. 5 alpha-Pregnane-3 alpha,20 alpha-diol disulfate was a metabolic end product accounting for a major part of the elimination of the steroids injected. Its half-life time was 1.4 to 2.8 hours. The results show that the formation of sulfated steroids with a 3 alpha-hydroxy-5 alpha configuration may account for 50% of the metabolism of progesterone in late pregnancy.

Analysis of aldehydic lipid peroxidation products in rat liver and hepatocytes by gas chromatography and mass spectrometry of the oxime-tert-butyldimethylsilyl derivatives.

A method for analysis of aliphatic aldehydes in biological samples is described. Cyclohexanone is added as internal standard and the samples are treated with hydroxylamine and perchloric acid. The oximes are extracted and converted to the oxime-tert-butyldimethylsilyl derivatives, which are quantitated by capillary gas chromatography and identified by mass spectrometry. The characteristic M-57 fragment ions in the mass spectra enabled a rapid identification of the derivatives of the aldehydes, alkanals, alk-2-enals, alka-2,4-dienals, and 4-hydroxyalk-2-enals, which in addition gave rise to characteristic double peaks in the gas chromatographic analysis. The method was applied to analysis of autoxidized arachidonic acid, ADP-Fe3(+)-treated rat hepatocytes, and rat liver given a single dose of ethanol, 5 g/kg. The amounts of hexanal and 4-hydroxynon-2-enal were not increased 6 h after the administration of ethanol.

Compartmentation of acetyl CoA studied by analysis of tricarboxylic acid cycle acids and 3-hydroxybutyrate in bile of rats given [2,2,2-2H3]ethanol.

Acetate, 3-hydroxybutyrate, pyruvate, lactate, citrate, 2-oxoglutarate, succinate, fumarate and malate were analysed in rat bile by gas chromatography and gas chromatography/mass spectrometry of their O-melthyloxime-t-butyldimethylsilyl derivatives. The concentration of acetate increased to about 1.8 mmol/l after administration of [2,2,2-2H3]ethanol. Acetate was formed from ethanol to an extent of about 82% and retained all of the 2H at C-2, whereas 15% of the 2H had been lost in the tricarboxylic acid cycle intermediates and 24% in 3-hydroxybutyrate. Thus the exchange of 2H for 1H takes place after formation of acetyl CoA. For citrate and 3-hydroxybutyrate, 41% and 11% respectively was formed from [2,2,2-2H3]ethanol. These results indicate that different pools of acetyl CoA are used for the synthesis of ketone bodies and citrate, with the latter being derived from ethanol to a much larger extent. Smaller fractions of 2-oxoglutarate (16%) and succinate (5%) were derived from [2,2,2--2H3]ethanol, indicating significant contributions from amino acids.

Concentration and turnover of estradiol in the rat uterus in vivo.

The concentrations and turnover of estradiol isolated from cytosolic and nuclear fractions of uteri from ovariectomized rats given estradiol, either in single injections or in continuous infusion, were analyzed by gas chromatography-mass spectrometry. The analytical method was validated for different organs and lower limits of analysis were established. After infusion of 20 ng x h-1 for 18-22 h, mean estradiol levels were 2.0-2.4 fmol x mg-1 uterine wet weight in the nuclear fraction, and 1.2-1.5 fmol x mg-1 in the cytosolic fraction. The concentrations were about five times higher after a single injection of one microgram estradiol but the distribution between nuclear and cytosolic fractions was almost the same. The concentrations of estradiol in nuclei from liver and spleen were 50-200 times lower than those in uterus. Taken together with previous knowledge, the results indicate that the distributions of estradiol and its receptor are not the same and that hormone response cannot be predicted from the concentration of receptors alone. The exchange of estradiol molecules in the uterus was followed after a change of the infusion from unlabelled to [11,12,12-2H3]-labelled estradiol, or vice versa. The uterine uptake of estradiol was calculated to be about 0.7 fmol x h-1 x mg-1 uterine wet weight. The half-life time was calculated to be at least 4 h for estradiol molecules isolated from the nuclear fraction and 3 h (significantly shorter) for those isolated from the cytosolic fraction. The results indicate an uptake of 40-90% of all estradiol passing through the uterus in proestrus with only about 10% of available receptors becoming occupied. When the infusion was changed from estradiol to ethynylestradiol, estradiol disappeared from the uterus at the same rate as in the experiments above. Ethynylestradiol was taken up at a rate of about 0.3-0.4 fmol x h-1 x mg-1 tissue. The percentage of total steroid found in the nuclear fraction was higher for ethynylestradiol, about 70%, than for estradiol, about 60%, indicative of a more stable association of receptor to nuclear binding sites when ethynylestradiol is the ligand.

Dehydrogenase-dependent ethanol metabolism in deer mice (Peromyscus maniculatus) lacking cytosolic alcohol dehydrogenase. Reversibility and isotope effects in vivo and in subcellular fractions.

Elimination of [2H]ethanol in vivo as studied by gas chromatography/mass spectrometry occurred at about half the rate in deer mice reported to lack alcohol dehydrogenase (ADH-) compared with ADH+ deer mice and exhibited kinetic isotope effects on Vmax and Km (D(V/K] of 2.2 +/- 0.1 and 3.2 +/- 0.8 in the two strains, respectively. To an equal extent in both strains, ethanol elimination was accompanied by an ethanol-acetaldehyde exchange with an intermolecular transfer of hydrogen atoms, indicating the occurrence of dehydrogenase activity. This exchange was also observed in perfused deer mouse livers. Based on calculations it was estimated that at least 50% of ethanol elimination in ADH- deer mice was caused by the action of dehydrogenase systems. NADPH-supported cytochrome P-450-dependent ethanol oxidation in liver microsomes from ADH+ and ADH- deer mice was not stereoselective and occurred with a D(V/K) of 3.6. The D(V/K) value of catalase-dependent oxidation was 1.8, whereas a kinetic isotope effect of cytosolic ADH in the ADH+ strain was 3.2. Mitochondria from both ADH+ and ADH- deer mice catalyzed NAD+-dependent ethanol oxidation and NADH-dependent acetaldehyde reduction. The kinetic isotope effects of NAD+-dependent ethanol oxidation in the mitochondrial fraction from ADH+ and ADH- deer mice were 2.0 +/- 0.1 and 2.3 +/- 0.3, respectively. The results indicate only a minor contribution by cytochrome P-450 to ethanol elimination, whereas the isotope effects are consistent with ethanol oxidation by the catalase-H2O2 system in ADH- deer mice in addition to the dehydrogenase systems.

Mechanism and regulation of ethanol elimination in humans: intermolecular hydrogen transfer and oxidoreduction in vivo.

Ethanol metabolism was studied in four healthy volunteers by intravenous infusion of a mixture of [1,1-2H2]ethanol (1.0 mmol/kg) and [2,2,2-2H3]ethanol (1.0 mmol/kg) followed by blood sampling at 10-min intervals. The concentrations of ethanols labeled with 1, 2, 3, and 4 deuterium atoms were determined by gas chromatography/mass spectrometry of the 3,5-dinitrobenzoates. During the first 30 min mono- and tetradeuteriated molecules appeared rapidly, which indicates that a fraction of the ethanol was formed from acetaldehyde by exchange. This fraction was calculated to be 38-58% and the hydrogen incorporated during the reduction was mainly (63-82%) derived from C-1 of ethanol, indicating slow exchange of enzyme-bound NADH. After 30 min the elimination followed first-order kinetics with t1/2 of 18-31 min and with a small primary isotope effect (1.05-1.11). This indicates almost complete removal of ethanol from blood passing through the liver when the concentration is low (below 1 mM). The results indicate that as long as hepatic blood flow is not limiting, the rate of alcohol dehydrogenase-catalyzed elimination of a small dose of ethanol in vivo is limited by the dissociation of NADH from the enzyme and by the rates of oxidation of acetaldehyde and reoxidation of NADH.