High-level expression of human liver monoamine oxidase B in Pichia pastoris

The high-level heterologous expression, purification, and characterization of the mitochondrial outer membrane enzyme human liver monoamine oxidase B (MAO B) using the methylotrophic yeast Pichia pastoris expression system are described. A 2-L culture of P. pastoris expresses approximately 1700 U of MAO B activity, with the recombinant enzyme associated tightly with the membrane fraction of the cell lysate. By a modification of the published procedure for purification of bovine liver MAO B [Salach, J. I. (1979) Arch. Biochem. Biophys. 192, 128-137], recombinant human liver MAO B is purified in a 34% yield ( approximately 200 mg from 2 L of cell culture). The isolated enzyme exhibits an M(r) of approximately 60, 000 on SDS-PAGE and 59,474 from electrospray mass spectrometry measurements, which is in good agreement with the mass predicted from the gene sequence and inclusion of the covalent FAD. One mole of covalent FAD per mole of MAO B is present in the purified enzyme and is bound by an 8alpha-S-cysteinyl(397) linkage, as identified by electrospray mass spectrometry of the isolated tryptic/chymotryptic flavin peptide. Recombinant human liver MAO B and bovine liver MAO B are shown to be acetylated at the seryl residues at their respective amino termini. The benzylamine oxidase activity of recombinant MAO B ranges from 3.0 to 3.4 U/mg and steady-state kinetic parameters for this enzyme preparation compare well with those published for the bovine liver enzyme: k(cat) = 600 min(-1), K(m)(benzylamine) = 0.50 mM, and K(m)(O(2)) = 0.33 mM. Kinetic isotope effect parameters using [alpha,alpha-(2)H(2)]benzylamine are also similar to those found for the bovine enzyme. Recombinant MAO B exhibits a (D)k(cat) = 4.7, a (D)[k(cat)/K(m)(benzylamine)] = 4.5, and a (D)[k(cat)/K(m)(O(2))] = 1.0. In contrast to bovine liver MAO B, no evidence was found for the presence of any anionic flavin radical either by UV-vis or by EPR spectroscopy in the resting form of the enzyme. These data demonstrate the successful heterologous expression of a functional, membrane-bound MAO B, which will permit a number of mutagenesis studies as structural and mechanistic probes not previously possible.     

Topological probes of monoamine oxidases A and B in rat liver mitochondria: inhibition by TEMPO-substituted pargyline analogues and inactivation by proteolysis

TEMPO-substituted pargyline analogues differentially inhibit recombinant human monoamine oxidase A (MAO A) and B (MAO B) in intact yeast mitochondria, suggesting these membrane-bound enzymes are located on differing faces of the mitochondrial outer membrane [Upadhyay, A., and Edmondson, D. E. (2009) Biochemistry 48, 3928]. This approach is extended to the recombinant rat enzymes and to rat liver mitochondria. The differential specificities exhibited for human MAO A and MAO B by the m- and p-amido TEMPO pargylines are not as absolute with the rat enzymes. Similar patterns of reactivity are observed for rat MAO A and B in mitochondrial outer membrane preparations expressed in Pichia pastoris or isolated from rat liver. In intact yeast mitochondria, recombinant rat MAO B is inhibited by the pargyline analogue whereas MAO A activity shows no inhibition. Intact rat liver mitochondria exhibit an inhibition pattern opposite to that observed in yeast where MAO A is inhibited and MAO B activity is unaffected. Protease inactivation studies show specificity in that MAO A is sensitive to trypsin whereas MAO B is sensitive to β-chymotrypsin. In intact mitochondrial preparations, MAO A is readily inactivated in rat liver but not in yeast upon trypsin treatment and MAO B is readily inactivated by β-chymotrypsin in yeast but not in rat liver. These data show MAO A is oriented on the cytosolic face and MAO B is situated on the surface facing the intermembrane space of the mitochondrial outer membrane in rat liver. The differential mitochondrial outer membrane topology of MAO A and MAO B is relevant to their inhibition by drugs designed to be cardioprotectants or neuroprotectants.     

cDNA cloning and sequencing of rat monoamine oxidase A: comparison with the human and bovine enzymes.

1. The nucleotide and deduced amino acid sequences of rat liver MAO A were determined, and sequence identities among MAO A and B from rat, human and bovine were compared. 2. MAO A from rat exhibited greater than 85% sequence identity with bovine and human MAO A, and 70% identity with rat MAO B. 3. Rat adrenal cDNAs were restriction mapped, partially sequenced and found to be identical to rat liver MAO A, suggesting that these two tissues express the same polypeptide.     

Tyramine and monoamine oxidase inhibitors as modulators of the mitochondrial membrane permeability transition

Incubation of rat liver mitochondria with 100-500 mM tyramine, a substrate for monoamine oxidases A and B (MAOs), in the presence of 30 mM Ca2+ induces matrix swelling, accompanied by collapse of membrane potential, efflux of endogenous Mg2+ and accumulated Ca2+ and oxidation of endogenous pyridine nucleotides. These effects are completely abolished in the presence of cyclosporin A, ADP, dithioerythritol and N-ethylmaleimide, thus confirming the induction of the mitochondrial membrane permeability transition (MPT). The observed partial protective effect exerted by catalase indicates the involvement of both MAO-derived hydrogen peroxide and aldehyde. Higher concentrations of tyramine (1-2 mM) are less effective or even completely ineffective. At these high concentrations tyramine has an inhibitory effect when the MPT is induced by 100 mM Ca2+. The MAO inhibitors clorgyline (50 mM) and pargyline (500 mM) completely protect against MPT induction by 100 mM tyramine but also inhibit the phenomenon, although with different efficacy, when it is induced by 100 mM Ca2+ in the absence of tyramine. Taken together, our data suggest that tyramine, clorgyline and pargyline act as modulators of the MPT either through a direct inducing/protective effect or by controlling hydrogen peroxide and aldehyde generation.     

The in vitro insertion of monoamine oxidase B into mitochondrial outer membranes

Bovine monoamine oxidase (MAO) B has been synthesized in vitro using a reticulocyte lysate translation system directed by bovine liver poly(A)+ RNA. The newly synthesized enzyme apparently lacks a cleavable N-terminal extension, but MAO B is readily incorporated into mitochondria or isolated mitochondrial outer membranes prepared from rat liver. ATP is not required for the binding of the newly synthesized enzyme to the outer membranes, but is necessary for the insertion of MAO B into these membrane vesicles. The ATP is not required to generate a mitochondrial membrane potential as assembly occurs under conditions that preclude either the formation or the maintenance of the potential. MAO B will bind to but not become incorporated into outer membrane vesicles which have been treated with trypsin, suggesting that the insertion of MAO B also depends on protein factors present on the outer membranes.     

Human brain monoamine oxidase: solubilization and kinetics of inhibition by octylglucoside

Complete solubilization of both the A and B forms of human brain monoamine oxidase (MAO) occurred when crude mitochondria were incubated in the presence of 50 mM octylglucoside (OG). Upon removal of this nonionic detergent by dialysis, approximately 100% of the starting activity was present in the dialysate. The effects of solubilization were examined by comparison of several properties of the membrane-bound and OG-treated oxidases. The percentage inhibition of phenylethylamine (PEA) and the 5-hydroxytryptamine (5-HT) deamination by deprenyl and clorgyline were identical. The Km values obtained for the deamination of PEA, a B-selective substrate, 5-HT, an A-selective substrate, and tyramine (TYR), a nonselective substrate, were also comparable. OG was found to inhibit type A (I50 = 8.1 mM) and B (I50 = 4.7 mM) MAO activities at concentrations at least 10-fold below those used to solubilize the oxidases. Kinetic studies revealed that OG was an apparent competitive inhibitor of PEA deamination whereas OG produced a mixed-type pattern of inhibition when 5-HT was the variable substrate. Inhibition of TYR deamination by either the A or B form of MAO produced a mixed pattern of inhibition. The findings herein suggest that solubilization of the A and B forms of MAO by OG does not significantly alter the substrate and inhibitor specificity of the oxidases following removal of detergent. However, in the presence of concentrations of OG 50 times less than the critical micellar concentration of this detergent, marked inhibition of deamination by both forms of human brain MAO is observed. Accordingly, the usefulness of OG is limited to situations where the detergent is completely removed before quantitation of MAO activity.     

The effect of isatin (tribulin) on metabolism of indoles in the rat brain and pineal: In vitro and in vivo studies

Isatin (Tribulin) produced a dose-dependent inhibition of both MAO A and MAO B in broken cell preparations from rat brain and pineal. However, isatin administered in vivo (80–160 mg/kg) to the intact animal significantly increased brain, but not pineal, serotonin and did not affect 5HIAA or other indoles in either brain or pineal. Further, in vivo administration did not produce detectable MAO inhibition in either tissue. In pineal organ culture, addition of isatin up to 1mM had no influence on the concentrations of pineal indoles or the activities of monoamine oxidase or serotonin N-acetyltransferase. However, the diazepam augmentation of beta adrenergic induction of serotonin N-acetyltransferase activity was blocked by isatin. The results of these studies call into question the proposed role of isatin as an endogenous monoamine oxidase inhibitor but support a possible role as a benzodiazepine receptor blocker.     

Phenylethanolamine is a specific substrate for type B monoamine oxidase

The characteristics of phenylethanolamine as both a competitive inhibitor and as a substrate for monoamine oxidase (MAO) were studied using rat brain and liver homogenates. Although phenylethanolamine, even at high concentrations (1 mM), produced minimal inhibition of MAO when serotonin (a substrate for type A MAO) was used as the substrate, it was a potent competitive inhibitor (K_i=11 μM) of the deamination of phenylethylamine (a substrate for type B MAO). When phenylethanolamine was used as a substrate, deprenyl, a selective inhibitor of type B MAO, was found to produce a single sigmoid inhibition curve at low concentrations of the inhibitor (pI₅₀=7.5). These results indicate that phenylethanolamine is a specific substrate for type B MAO. Identification of the products formed under the assay conditions show that phenylethanolamine is converted to both mandelic acid and phenylethylene glycol by liver homogenates but only to the latter, neutral metabolite by brain homogenates.     

Inhibition of Type A Monoamine Oxidase by Methylquinolines and Structurally Related Compounds

Abstract: A series of methylquinolines (MQ) were found to inhibit markedly type A monoamine oxidase (MAO) in human brain synaptosomal mitochondria. 4-MQ and 6-MQ inhibited type A MAO (MAO-A) competitively and 7- and 8-MQ inhibited MAO-A noncompetitively. Among these four isomers of MQ, 6-MQ was the most potent inhibitor; the K _i value toward MAO-A was 23.4 ± 1.8 μ M , which was smaller than the K _m value toward kynuramine, ± amine substrate, 46.2 ± 2.8 μ M . On the other hand, MQ were very weak inhibitors of type B MAO (MAO-B) and 8-MQ did not inhibit MAO-B in brain synaptosomal mitochondria. The inhibition of MAO-A proved to be reversible; by dialysis the inhibition of MQ was completely reversible. The affinity of these isomers of MQ toward MAO-A or -B was confirmed further with human liver mitochondria as sources of MAO-A and -B and with human placental mitochondria and rat pheochromocytoma PC12h cell line as sources of MAO-A. The relationship of the chemical structure of structurally related quinoline and isoquinoline derivatives to inhibition of the activity of type A or B MAO was examined.     

Isoelectric analysis of 3H-pargyline-labelled monoamine oxidase in rat and carp

1. After selective binding of [3H]pargyline to either monoamine oxidase (MAO) A or MAO B in the rat liver, MAO B alone in the rat brain and MAO in carp brain and liver, molecular weight and isoelectric points (pI) of these MAO were determined by sodium dodecyl sulphate (SDS)-polyacrylamide gel electrophoresis and isoelectric focusing and results obtained were compared. 2. For all tissues tested, SDS-polyacrylamide gel electrophoresis of [3H]pargyline-bound samples revealed a labelled protein band of an apparent mol. wt of 60,000 da. 3. Estimation of radioactivity of [3H]pargyline bound after isoelectric focusing revealed a single protein band with acidic pI values of about 5.5 for rat brain and liver MAO B. 4. Moreover, the pI values of about 7.5 were obtained for carp brain and liver MAO. This basic value was also found for MAO A in the rat liver MAO A.     

Side-chain metabolism of propranolol: involvement of monoamine oxidase and mitochondrial aldehyde dehydrogenase in the metabolism of N-desisopropylpropranolol to naphthoxylactic acid in rat liver

We examined the metabolism of N-desisopropylpropranolol (NDP), which is generated from propranolol (PL) by side-chain N-desisopropylation, to naphthoxylactic acid (NLA) in rat liver. S(-)-NDP (S-NDP) and R(+)-NDP (R-NDP) were enantioselectively metabolized to NLA in isolated rat hepatocytes and in an enzyme reaction system of rat liver mitochondria with cofactor NAD+. Furthermore, the clearance profiles of NDP enantiomers were examined in an enzyme reaction system of rat liver mitochondria without NAD+. The amounts of S-NDP remaining in the incubation medium were similar to those of R-NDP, suggesting that monoamine oxidase (MAO) catalyzes the deamination of NDP to the aldehyde intermediate, but fails to deaminate enantioselectively S-NDP or R-NDP. Cyanamide, a potent inhibitor of aldehyde dehydrogenase (ALDH), markedly decreased the formation of NLA from racemic NDP in the enzyme reaction system of rat liver mitochondria with NAD+. When rat liver cytosol and microsomes were added to this enzyme reaction system, no significant alterations were observed in the amount of NLA generated from racemic NDP. We concluded that MAO deaminates NDP to an aldehyde intermediate, and that mitochondrial ALDH subsequently catalyzes the enantioselective metabolism of the aldehyde intermediate to NLA in rat liver.     

Evidence for the generation of transaminase inhibitor(s) during ethanol metabolism by rat liver homogenates: a potential mechanism for alcohol toxicity

Since ethanol consumption decreases hepatic aminotransferase activities in vivo, mechanisms of ethanol-mediated transaminase inhibition were explored in vitro using mitochondria-depleted rat liver homogenates. When homogenates were incubated at 37 degrees with 50 mM ethanol for 1 hr, alanine aminotransferase decreased by 20%, while aspartate aminotransferase was unchanged. After 2 hr, aspartate aminotransferase decreased by 20% and by 3 hr, alanine and aspartate aminotransferases were decreased by 31 and 23%, respectively. Levels of acetaldehyde generated during ethanol oxidation were 525 +/- 47 microM at 1 hr, 855 +/- 14 microM at 2 hr, and 1293 +/- 140 microM at 3 hr. Although inhibition of alcohol oxidation with methylpyrazole or cyanide markedly decreased ethanol-mediated transaminase inhibition, neither incubation with acetate nor generation of reducing equivalents by oxidation of lactate, malate, xylitol, or sorbitol altered the activity of either enzyme. However, semicarbazide, an aldehyde scavenger, prevented inhibition of both aminotransferases by ethanol. Moreover, incubation with 5 mM acetaldehyde for 1 hr inhibited alanine and aspartate aminotransferases by 36 and 26%, respectively. Cyanamide, an aldehyde dehydrogenase inhibitor, had little effect on ethanol-mediated transaminase inhibition. Thus, metabolism of ethanol by rat liver homogenates produces transaminase inhibition similar to that described in vivo and this effect requires acetaldehyde generation but not acetaldehyde oxidation. Since addition of pyridoxal 5'-phosphate to assay mixes did not reverse ethanol effects, aminotransferase inhibition does not result from displacement of vitamin B6 coenzymes.     

Oxidation of β-Phenylethylamine by Both Types of Monoamine Oxidase: Examination of Enzymes in Brain and Liver Mitochondria of Eight Species

Abstract: β-Phenylethylamine (PEA) was characterized as a substrate for type A and type B monoamine oxidase (MAO) in brain and liver mitochondria of eight species at different substrate concentrations. In all species, at 10.0 μM, PEA was almost specific for type B MAO. At 1000 μM, however, the amine was common for both types of MAO in rat brain and liver, human brain and liver, mouse brain, guinea pig brain and liver, and bovine brain, while it was specific for type B MAO in mouse liver, rabbit brain and liver, bovine liver, pig brain and liver, and chicken brain and liver. From the present study, when PEA is used as a type B substrate, it is recommended that the substrate concentration should be sufficiently low to avoid the effects of species and tissue differences.     

Inhibition of pancreatic islet monoamine oxidase by adrenergic antagonists and ethanol.

Although the alpha-adrenergic antagonist phentolamine potentiates glucose-stimulated insulin secretion of intact animals, it either does not alter, or it inhibits in vitro insulin secretion. This may be because in the higher concentration used in in vitro studies, phentolamine exerts a second pharmacological effect that counterbalances its primary effect of blocking monoamine action. We recently demonstrated that pancreatic islets contain substantial amounts of monoamine oxidase (MAO), and that MAO inhibitors such as iproniazid and tranylcypromine can alter insulin secretion. In the present study, we determined if other drugs that affect insulin secretion, alter the MAO activity of homogenates of rabbit pancreatic islets (collagenase technique) or liver. Phentolamine, phenoxybenzamine and propranolol (10 muM and 100 muM) inhibit islet and hepatic MAO. Haloperidol (10muM) inhibits hepatic but not islet MAO, while haloperidol (10muM) does not inhibit MAO in either tissue. Ethanol (270 to 2.7mM) inhibits islet MAO. Hepatic MAO is inhibited by high (270 to 180mM) but not by low (27 to 2.7mM) concentrations of ethanol. Collagenase digestion does not increase the sensitivity of islet and liver MAO to inhibition by phentolamine or ethanol. In the absence of added monoamines, phentolamine and phenoxybenzamine do not alter basal or glucose-stimulated insulin secretion from rabbit pancreas. Preincubation of rabbit pancreas with the serotonin precursor 5-hydroxytryptophan (5-HTP) increases the beta cell serotonin content and inhibits glucose-stimulated insulin secretion. Alpha adrenergic antagonists not only fail to block, but actually potentiate the serotonin inhibition of insulin secretion. We conclude that inhibition of islet MAO may cause an increase in islet monoamine content and these monoamines may alter in vitro insulin secretion. One mechanism through which adrenergic antagonists and ethanol modify in vitro insulin secretion may be by inhibiting pancreatic islet MAO.     

Side-chain metabolism of propranolol: involvement of monoamine oxidase and aldehyde reductase in the metabolism of N-desisopropylpropranolol to propranolol glycol in rat liver

The further metabolism of N-desisopropylpropranolol (NDP), a side-chain metabolite of propranolol (PL), was investigated in isolated rat hepatocytes. Propranolol glycol (PGL) was generated from NDP as a major metabolite. Naphtetrazole (NTE), a potent inhibitor of monoamine oxidase (MAO), significantly retarded the disappearance of NDP from the incubation medium, suggesting the involvement of MAO in the deamination of NDP to an aldehyde intermediate. In a reaction mixture of rat liver mitochondria and cytosol with NADPH, phenobarbital, a specific inhibitor of aldehyde reductase, and 4-nitrobenzaldehyde (4-NBA), a substrate inhibitor of aldehyde reductase, decreased the formation of PGL from NDP. 4-NBA was a competitive inhibitor of the enzyme responsible for the PGL formation. The optimal pH for the formation of PGL from NDP in the reaction mixture was approximately 8.0. Based on these results, we propose the possibility that, in the rat liver, MAO catalyzes the oxidative deamination of NDP to an aldehyde intermediate and the formed aldehyde intermediate is subsequently reduced to PGL by aldehyde reductase. Furthermore, the enantioselective metabolism of NDP to PGL was examined. In isolated rat hepatocytes, the amount of PGL formed from S-NDP [S(-)-form of NDP] was larger than that of PGL formed from R-NDP [R(+)-form of NDP].     

In vitro inhibition of alkaline phosphatase activities from intestine, bone, liver, and kidney by phenobarbital.

A kinetic study of the inhibition of several alkaline phosphatase (AP isoenzyme activities by phenobarbital was carried out using p-nitrophenylphosphate (10 mM) as a substrate at pH 9.8 in a 300-mM Hepes buffer. AP from bovine kidney, calf intestine, bovine liver, and rat bone was used. Over a phenobarbital concentration range of 20-400 mM, all these isoenzymes were inhibited in an uncompetitive manner with a Ki of 200 mM for intestinal AP, and in a linear mixed-type manner for all the other isoenzymes tested. The Ki values were 10, 40 and 55 mM for kidney, bone and liver AP, respectively. The use of 15 mM carbonate-bicarbonate or 400 mM diethanolamine buffer did not modify the degree of inhibition of intestinal AP activity. Dixon plots of the reciprocal of reaction velocity versus inhibitor concentration either at different substrate concentration or at different DEA concentration indicate uncompetitive inhibition for the intestinal enzyme. This in vitro inhibitory effect of phenobarbital is in contrast to its in vivo stimulating action on AP. However, in the whole animal, the effects of phenobarbital administration probably represent the sum of multiple effects.     

A new type of mitochondrial monamine oxidase distinct from type-A and type-B

The characteristics of mitochondrial monoamine oxidase (MAO) in carp liver were studied with MAO inhibitors and substrates. This enzyme was thermolabile, but was stabilized in the presence of bovine serum albumin. With clorgyline and deprenyl, single-sigmoidal curves for inhibition of the activity towards tyramine or 5-hydroxytryptamine were obtained; the sensitivities to the two inhibitors were identical. The activity towards β-phenylethylamine was not completely inhibited by clorgyline or deprenyl, but the remaining activity was inhibited by semicarbazide and the inhibition curves by either clorgyline or deprenyl and semicarbazide were also identical to the curves with the other two substrates. These results suggest that carp liver mitochondria contain “classical” MAO and a clorgyline- and deprenyl-resistant amine oxidase and that the classical MAO does not seem to be MAO-A or MAO-B, which are present in mitochondria of most mammalian tissues.     

Studies on the pargyline-binding site of different types of monoamine oxidase

[3H]Pargyline has been covalently linked to active sites of both type A and type B monoamine oxidase (MAO) obtained from various tissues. Rat heart and human placenta were chosen to represent predominantly type A MAO, pig and bovine livers to represent type B MAO, and rat liver and brain to represent mixed type A and type B MAO's. The [3H]pargyline-MAO adducts were isolated and hydrolyzed by proteolytic enzymes, and the labelled peptides (pargyline-binding sites) separated and compared by paper chromatography and by paper electrophoresis at various pH values. Only one common pargyline peptide was obtained from all the different MAO's. The alternative A and B sites were assessed after preincubation of rat liver MAO with the selective inhibitors deprenyl (to block the B site) and clorgyline (to block the A site). Following proteolysis of the [3H]pargyline of both type A and type B MAO from this pretreated rat liver, MAO has been purified by a series of chromatographic and electrophoretic procedures. Micro-Edman degradation, followed by dansylation, revealed the amino acid sequence to be Ser-Gly-Gly-Cys(X)-Tyr. It is concluded that the primary structures immediately surrounding the pargyline-binding sites are identical for both type A and type B MAO in these tissues.     

Deamination of Aliphatic Amines by Monoamine Oxidase A and B Studied Using a Bioluminescence Technique

Deamination of n-octylamine and n-decylamine has been studied in various tissues using a new bioluminescence technique. Selectivity of n-octylamine and n-decylamine as substrates for monoamine oxidase (MAO) A or B has been determined using both clorgyline and (-)-deprenyl inhibition curves and kinetic parameters. Homogenates of rat brain, liver and heart containing predominantly MAO-A or -B were prepared by preincubation for 60 min with (-)-deprenyl or clorgyline (30 nM), respectively. Human placenta (MAO-A) and platelet (MAO-B) were used as reference tissues containing only one MAO form. In tissues (rat liver, brain) containing both MAO forms in equal proportion, inhibition curve studies showed a preference of both substrates for the B form of the enzyme; however, where MAO-A was the major form (rat heart, human placenta), clorgyline was the more effective inhibitor. In the beef brain cortex n-octylamine showed marked preference for MAO-B, whereas n-decylamine was selective toward-MAO-A. Kinetic studies in general supported the picture of greater selectivity of the aliphatic amine substrates for deamination by MAO-B, as reflected by lower Km values for this enzyme type. However, n-octylamine was more selective for MAO-B than n-decylamine in both kinetic and inhibition curve studies. The deamination of these aliphatic amine substrates cannot be explained only by reference to the binary classification of MAO into types A and B.     

Involvement of monoamine oxidase and diamine oxidase in the metabolism of the cell differentiating agent hexamethylene bisacetamide (HMBA)

We have previously demonstrated a number of metabolites of hexamethylene bisacetamide (HMBA) in the urine of patients treated with HMBA. These include N-acetyl-1,6-diaminohexane (NADAH), 6-acetamidohexanoic acid (6AcHA), 1,6-diaminohexane (DAH) and 6-aminohexanoic acid (6AmHA). Because these compounds have potential roles in the dose-limiting metabolic acidosis and neurotoxicity associated with HMBA therapy, and are similar in structure to known substrates of monoamine oxidase (MAO) and diamine oxidase (DAO), we investigated the activities of these enzymes in the metabolic interconversion of HMBA metabolites. NADAH (5 mM) was incubated with MAO and aldehyde dehydrogenase. 6AcHA production was verified by gas chromatography-mass spectrometry and quantified by gas chromatography. 6AcHA production was linear for up to 4 hr. Complete inhibition of MAO activity was observed with 2 mM tranyl-cypromine or pargyline. Mouse liver microsomes, which do not contain MAO, did not convert NADAH to 6AcHA and, in control experiments, did not degrade 6AcHA. The HMBA metabolite, DAH, was a substrate for DAO, producing 3,4,5,6-tetrahydro-2H-azepine. Participation of DAO in the metabolism of HMBA implies potential interaction of HMBA and metabolites with polyamine metabolism and may represent a mechanism for HMBA's effects on cellular growth and differentiation. Metabolism of NADAH, also a differentiator, by MAO implies that concurrent use of HMBA and an MAO inhibitor may be clinically useful.