Mitochondria as the primary site of acetaldehyde metabolism in beef and pig liver slices

Oxidative-stress-related proteome changes in Helicobacter pylori-infected human gastric mucosa

Helicobacter pylori infection leads to gastroduodenal inflammation, peptic ulceration and gastric carcinoma. Proteomic analysis of the human gastric mucosa from the patients with erosive gastritis, peptic ulcer or gastric cancer, which were either infected or not with H. pylori, was used to determine the differentially expressed proteins by H. pylori in the human gastric mucosa in order to investigate the pathogenic mechanism of H. pylori -induced gastric diseases. Prior to the experiment, the expression of the main 18 proteins were identified in the gastric mucosa and used for a proteome map of the human gastric mucosa. Using two-dimensional electrophoresis of the protein isolated from the H. pylori -infected tissues, Coomassie Brilliant Blue staining and computerized analysis of the stained gel, the expression of eight proteins were altered in the H. pylori -infected tissues compared with the non-infected tissues. MS analysis (matrix-assisted laser desorption/ionization-time of flight MS) of the tryptic fragment and a data search allowed the the identification of the four increased proteins (78 kDa glucose-regulated protein precursor, endoplasmin precursor, aldehyde dehydrogenase 2 and L-lactate dehydrogenase B chain) and the four decreased proteins (intracellular chloride channel protein 1, glutathione S-transferase, heat-shock protein 60 and cytokeratin 8) caused by H. pylori infection in the gastric mucosa. These proteins are related to cell proliferation, carcinogenesis, cytoskeletal function and cellular defence mechanism. The common feature is that these proteins are related to oxidative-stress-mediated cell damage. In conclusion, the established gastric mucosal proteome map might be useful for detecting the disease-related protein changes. The H. pylori -induced alterations in protein expression demonstrate the involvement of oxidative stress in the pathogenesis of H. pylori -induced gastric diseases, including inflammation, ulceration and carcinogenesis.

Increase in class 2 aldehyde dehydrogenase expression by arachidonic acid in rat hepatoma cells

Aldehyde dehydrogenase (ALDH) is a family of several isoenzymes important in cell defence against both exogenous and endogenous aldehydes. Compared with normal hepatocytes, in rat hepatoma cells the following changes in the expression of ALDH occur: cytosolic class 3 ALDH expression appears and mitochondrial class 2 ALDH decreases. In parallel with these changes, a decrease in the polyunsaturated fatty acid content in membrane phospholipids occurs. In the present study we demonstrated that restoring the levels of arachidonic acid in 7777 and JM2 rat hepatoma cell lines to those seen in hepatocytes decreases hepatoma cell growth, and increases class 2 ALDH activity. This latter effect appears to be due to an increased gene transcription of class 2 ALDH. To account for this increase, we examined whether peroxisome-proliferator-activated receptors (PPARs) or lipid peroxidation were involved. We demonstrated a stimulation of PPAR expression, which is different in the two hepatoma cell lines: in the 7777 cell line, there was an increase in PPAR alpha expression, whereas PPAR gamma expression increased in JM2 cells. We also found increased lipid peroxidation, but this increase became evident at a later stage when class 2 ALDH expression had already increased. In conclusion, arachidonic acid added to the culture medium of hepatoma cell lines is able to partially restore the normal phenotype of class 2 ALDH, in addition to a decrease in cell growth.

Making an Oriental equivalent of the yeast cytosolic aldehyde dehydrogenase as well as making one with positive cooperativity in coenzyme binding by mutations of glutamate 492 and arginine 480

Yeast has at least three partially characterized aldehyde dehydrogenases. Previous studies by gene disrupted in our laboratory revealed that the Saccharomyces cerevisiae cytosol ALDH1 played an important role in ethanol metabolism as did the class 2 mitochondrial enzyme. To date, few mutagenesis studies have been performed with the yeast enzymes. An important human variant of ALDH is one found in Asian People. In it, the glutamate at position 487 is replaced by a lysine. This glutamate interacts with an arginine (475) that is located in the subunit that makes up the dimer pair in the tetrameric enzyme. Sequence alignment shows that these two residues are located at positions 492 and 480, respectively, in the yeast class 1 enzyme which shares just 45% sequence identity with the human enzymes. Mutating glutamate 492 to lysine produced an enzyme with altered kinetic properties when compared to the wild-type glutamate-enzyme. The K(m) for NADP of E492K increased to nearly 3600 microM compare to 40 microM for wild-type enzyme. The specific activity decreased more than 10-fold with respect to the recombinant wild-type yeast enzyme. Moreover, substituting a glutamine for a glutamate was not detrimental in that the E492Q had wild-type-like K(m) for NADP and V(max). These properties were similar to the changes found with the human class 2 E487K mutant form. Further, mutating arginine 480 to glutamine produced an enzyme that exhibited positive cooperativity in NADP binding. The K(m) for NADP increased 11-fold with a Hill coefficient of 1.6. The NADP-dependent activity of R480Q mutant was 60% of wild-type enzyme. Again, these results are very similar to what we recently showed to occur with the human enzyme [Biochemistry 39 (2000) 5295-5302]. These findings show that the even though the glutamate and arginine residues are not conserved, similar changes occur in both the human and the yeast enzyme when either is mutated.

Betaine aldehyde, betaine, and choline levels in rat livers during ethanol metabolism

Betaine aldehyde levels were determined in rat livers following 4 weeks of ethanol feeding, employing the Lieber-De Carli liquid diet. The results showed that the levels of betaine aldehyde are unaffected by alcohol feeding to rats. These levels in both experimental and control animals were found to be quite low, 5.5 nmol/g liver. Betaine aldehyde levels have not been determined previously in mammalian liver because of methodological difficulties. This investigation employed fast atom bombardment-mass spectroscopy to determine the levels of betaine aldehyde, betaine, and choline. The decrease in betaine levels following ethanol administration confirmed the results of other investigators. Choline levels determined during this investigation were lower than previously reported. The reason for starting this investigation was the fact that the enzyme that catalyzes betaine aldehyde dehydrogenation to betaine, which is distributed in both mitochondria and the cytoplasm, was found to also metabolize acetaldehyde with K(m) and V(max) values lower than those for betaine aldehyde. Thus, it appeared likely that the metabolism of acetaldehyde during ethanol metabolism might inhibit betaine aldehyde conversion to betaine and thereby result in decreased betaine levels (Barak et al., Alcohol 13: 395-398, 1996). The fact that betaine aldehyde levels in alcohol-fed animals were similar to those in controls demonstrates that competition between acetaldehyde and betaine aldehyde for the same enzyme does not occur. This complete lack of competition suggests that betaine aldehyde dehydrogenase in the mitochondrial matrix may totally metabolize betaine aldehyde to betaine without any involvement of cytoplasmic betaine aldehyde dehydrogenase.

Structure of mitochondrial aldehyde dehydrogenase: the genetic component of ethanol aversion

BACKGROUND

The single genetic factor most strongly correlated with reduced alcohol consumption and incidence of alcoholism is a naturally occurring variant of mitochondrial aldehyde dehydrogenase (ALDH2). This variant contains a glutamate to lysine substitution at position 487 (E487K). The E487K variant of ALDH2 is found in approximately 50% of the Asian population, and is associated with a phenotypic loss of ALDH2 activity in both heterozygotes and homozygotes. ALDH2-deficient individuals exhibit an averse response to ethanol consumption, which is probably caused by elevated levels of blood acetaldehyde. The structure of ALDH2 is important for the elucidation of its catalytic mechanism, to gain a clear understanding of the contribution of ALDH2 to the genetic component of alcoholism and for the development of specific ALDH2 inhibitors as potential drugs for use in the treatment of alcoholism.

RESULTS

The X-ray structure of bovine ALDH2 has been solved to 2.65 A in its free form and to 2.75 A in a complex with NAD+. The enzyme structure contains three domains; two dinucleotide-binding domains and a small three-stranded beta-sheet domain, which is involved in subunit interactions in this tetrameric enzyme. The E487K mutation occurs in this small oligomerization domain and is located at a key interface between subunits immediately below the active site of another monomer. The active site of ALDH2 is divided into two halves by the nicotinamide ring of NAD+. Adjacent to the A-side (Pro-R) of the nicotinamide ring is a cluster of three cysteines (Cys301, Cys302 and Cys303) and adjacent to the B-side (Pro-S) are Thr244, Glu268, Glu476 and an ordered water molecule bound to Thr244 and Glu476.

CONCLUSIONS

Although there is a recognizable Rossmann-type fold, the coenzyme-binding region of ALDH2 binds NAD+ in a manner not seen in other NAD+-binding enzymes. The positions of the residues near the nicotinamide ring of NAD+ suggest a chemical mechanism whereby Glu268 functions as a general base through a bound water molecule. The sidechain amide nitrogen of Asn169 and the peptide nitrogen of Cys302 are in position to stabilize the oxyanion present in the tetrahedral transition state prior to hydride transfer. The functional importance of residue Glu487 now appears to be due to indirect interactions of this residue with the substrate-binding site via Arg264 and Arg475.

Allosteric inhibition of human liver aldehyde dehydrogenase by the isoflavone prunetin

Isoflavonoid derivatives including prunetin (4',5-dihydroxy-7-methoxyisoflavone) were shown to be potent inhibitors of human aldehyde dehydrogenases (Keung W-M and Vallee BL, Proc Natl Acad Sci USA 90: 1247-1251, 1993). The inhibition reaction was reinvestigated using recombinantly expressed human aldehyde dehydrogenases. The kinetic analyses showed that prunetin inhibits competitively against both NAD and propionaldehyde with the mitochondrial and cytoplasmic enzymes. The Ki value for the mitochondrial enzyme was much lower than for the cytoplasmicenzyme. A mixed pattern of inhibition was obtaiend with the mitochondrial enzyme in the presence of Mg2+. Only one mole of prunetin binds per mole of tetrameric mitochondrial enzyme, which remains unaltered in the presence of Mg2+. Prunetin did not displace NADH from the enzyme-NADH complex. Propionaldehyde did not reverse the loss of fluorescence obtained due to enzyme-prunetin complex formation, indicating that prunetin may not be interacting at the substrate site. The esterase activity of the mitochondrial enzyme was also inhibited by prunetin in a competitive manner. The replacement of lysine 192 by glutamine resulted in a mutant with a 20% kcat and a 100-fold increase in the Km for NAI) compared with the native enzyme. However, the Ki value of prunetin against NAD was similar to that observed with the native enzyme. Prunetin, even at a very high concentration, was not an inhibitor of alcohol and malate dehydrogenase. It was concluded that prunetin may act as an allosteric inhibitor of aldehyde dehydrogenase.

Heterotetramers of human liver mitochondrial (class 2) aldehyde dehydrogenase expressed in Escherichia coli. A model to study the heterotetramers expected to be found in Oriental people

About 50% of the Oriental population have less liver mitochondrial aldehyde dehydrogenase (ALDH2) activity than do other people. It was found that they possessed an enzyme with a lysine at position 487 (E487K) instead of glutamate (Glu487). We previously found that the Km for NAD of recombinant human and rat E487K enzymes increased more than 150-fold (Farrés, J., Wang X., Takahashi, K., Cunningham, S. J. , Wang, T.T., and Weiner, H (1994) J. Biol. Chem. 269, 13854-13860). Many aldehyde dehydrogenase-deficient people were found to be heterozygous when genotyped for ALDH2. In this study liver tissue from heterozygous people was analyzed and found to possess mRNAs for both the glutamate and the lysine subunits. Western blot analysis showed that the glutamate subunit was present. The cDNAs for Glu487 and E487K were coexpressed on one plasmid in Escherichia coli, and the enzyme forms were separated from each other by isoelectric focusing to show that heterotetramers were formed. Only one Km value for NAD could be measured with the purified heterotetrameric enzyme that possessed just 16-18% activity of the glutamate homotetrameric enzyme. The E487K homotetramers had 8% specific activity of the Glu487 enzyme. There was no pre-steady state burst of NADH formation with the heterotetramer, a property found with the glutamate enzyme. Similar results were found for the coexpressed rat liver enzyme, except that a higher specific activity, 48%, was obtained. Thus, we conclude that presence of the lysine subunit altered the activity of the glutamate subunit in the heterotetramer to make it function more like an E487K enzyme.

Inhibition of gene expression by triple helix formation in hepatoma cells

The aim of this study was to selectively inhibit human mitochondrial aldehyde dehydrogenase (ALDH2) gene expression by triple helix assembly. Eight 21-mer oligodeoxyribonucleotides were designed to bind to two purine-rich sequences in the 5'-flanking region of the human ALDH2 gene. Gel mobility shift assays showed that triplex formation is sequence-specific for the target duplex and the third strand oligonucleotide. In the presence of Mg2+, but absence of K+, triplex-forming oligonucleotides bind to their target sites with apparent dissociation constants (Kd) in the 10(-7) to 10(-9) M range. Potassium cation virtually suppressed the triplex formation of G-C-rich duplex DNA with natural oligonucleotides, but did not prevent triplex formation with phosphorothioate-modified oligonucleotides. Phosphorothioate-modified oligonucleotides were delivered into human hepatoma Hep G2 cells by cationic liposomes. The reduction in ALDH2 mRNA levels in the cells was determined by the competitive reverse transcription-polymerase chain reaction. One of the phosphorothioate-modified oligonucleotides designed to forma an antiparallel triplex with a target in the 5'-flanking region of human ALDH2 gene (-105 to -125 from the translation initiation codon ATG) reduced by 80-90% the ALDH2 mRNA levels without affecting albumin mRNA levels. Data suggest that triple-helix formation may provide a means to selectively inhibit hepatic ALDH2 gene expression for therapeutic use.

The aldehyde dehydrogenase ALDH2*2 allele exhibits dominance over ALDH2*1 in transduced HeLa cells

Individuals heterozygous or homozygous for the variant aldehyde dehydrogenase (ALDH2) allele (ALDH2*2), which encodes a protein differing only at residue 487 from the normal protein, have decreased ALDH2 activity in liver extracts and experience cutaneous flushing when they drink alcohol. The mechanisms by which this allele exerts its dominant effect is unknown. To study this effect, the human ALDH2*1 cDNA was cloned and the ALDH2*2 allele was generated by site-directed mutagenesis. These cDNAs were transduced using retroviral vectors into HeLa and CV1 cells, which do not express ALDH2. The normal allele directed synthesis of immunoreactive ALDH2 protein (ALDH2E) with the expected isoelectric point. Extracts of these cells contained increased aldehyde dehydrogenase activity with low Km for the aldehyde substrate. The ALDH2*2 allele directed synthesis of mRNA and immunoreactive protein (ALDH2K), but the protein lacked enzymatic activity. When ALDH2*1-expressing cells were transduced with ALDH2*2 vectors, both mRNAs were expressed and immunoreactive proteins with isoelectric points ranging between those of ALDH2E and ALDH2K were present, indicating that the subunits formed heteromers. ALDH2 activity in these cells was reduced below that of the parental ALDH2*1-expressing cells. Thus, the ALDH2*2 allele is sufficient to cause ALDH2 deficiency in vitro.

Daidzin suppresses ethanol consumption by Syrian golden hamsters without blocking acetaldehyde metabolism

Daidzin is a potent, selective, and reversible inhibitor of human mitochondrial aldehyde dehydrogenase (ALDH) that suppresses free-choice ethanol intake by Syrian golden hamsters. Other ALDH inhibitors, such as disulfiram (Antabuse) and calcium citrate carbimide (Temposil), have also been shown to suppress ethanol intake of laboratory animals and are thought to act by inhibiting the metabolism of acetaldehyde produced from ingested ethanol. To determine whether or not daidzin inhibits acetaldehyde metabolism in vivo, plasma acetaldehyde in daidzin-treated hamsters was measured after the administration of a test dose of ethanol. Daidzin treatment (150 mg/kg per day i.p. for 6 days) significantly suppresses (> 70%) hamster ethanol intake but does not affect overall acetaldehyde metabolism. In contrast, after administration of the same ethanol dose, plasma acetaldehyde concentration in disulfiram-treated hamsters reaches 0.9 mM, 70 times higher than that of the control. In vitro, daidzin suppresses hamster liver mitochondria-catalyzed acetaldehyde oxidation very potently with an IC50 value of 0.4 microM, which is substantially lower than the daidzin concentration (70 microM) found in the liver mitochondria of daidzin-treated hamsters. These results indicate that (i) the action of daidzin differs from that proposed for the classic, broad-acting ALDH inhibitors (e.g., disulfiram), and (ii) the daidzin-sensitive mitochondrial ALDH is not the one and only enzyme that is essential for acetaldehyde metabolism in golden hamsters.

In vivo effects of disulfiram and cyanamide on canine liver aldehyde dehydrogenase isoenzymes as detected by high-performance (pressure) liquid chromatography

Methods for analysis of aldehyde dehydrogenase isoenzymes using high-performance (pressure) liquid chromatography (HPLC) were used to determine in vivo effects of disulfiram and cyanamide on canine liver aldehyde dehydrogenase (ALDH) isoenzymes. Liver ALDH isoenzymes from control and disulfiram- or cyanamide-treated dogs were separated by ion-exchange HPLC, and enzyme activity was detected using a postcolumn reactor. Two major peaks of ALDH activity (peaks I and II) were detected. Varying the composition of the reaction column reagents resulted in alterations in the elution profiles consistent with the kinetic properties of individual isoenzymes (i.e., ALDH IB in peak I and ALDH IIB in peak II), including estimates of the Km for acetaldehyde and the effects of magnesium ions on ALDH activity. Disulfiram treatment decreased both peaks depending on disulfiram dose and length of treatment, with peak I being more sensitive to inactivation than peak II. Reagents containing MgCl2 (1 mM) decreased peak I and increased peak II compared with EDTA (1 mM) for samples from both control and disulfiram-treated animals. These data are consistent with the assignment of the disulfiram-sensitive isoenzyme (ALDH IB) to peak I and the isoenzyme stimulated by magnesium ions (ALDH IIB) to peak II. In vivo cyanamide treatment produced similar decreases in both peaks to a maximum decrease of approximately 30% of control depending on cyanamide dose. Peak I, however, was more sensitive than peak II to in vitro inactivation by cyanamide, which suggests that cytosolic ALDH in the dog (in contrast to other mammals) is more sensitive to inactivation than mitochondrial ALDH.

Cloning and expression of the full-length cDNAS encoding human liver class 1 and class 2 aldehyde dehydrogenase

The amino acid sequences of both human class 1 and 2 aldehyde dehydrogenase (ALDH) and the sequences of the genes coding for them are known. Based on this sequence data, we designed primers and isolated the full-length cDNAs encoding both isozymes from a human liver mRNA pool. cDNAs were subcloned in the plasmid pT7-7 and expressed in Escherichia coli with a yield of approximately 3 mg ALDH protein/liter of cell culture, although only one-third of the enzyme was soluble. The soluble recombinantly expressed ALDHs were purified to homogeneity using a hydroxyacetophenone-Sepharose affinity column. The mitochondrial isozyme had a subunit molecular weight of 55 kDa, an isoelectric point of 4.9, and a specific activity of 1.10 units/mg, which were in good agreement with that from the native enzyme. The expressed cytosolic isozyme had the same subunit molecular weight (55 kDa) and pI (5.4) as that reported for the native enzyme and had a specific activity of 0.26 units/mg. The expressed mitochondrial isozyme could be recognized by antibodies raised against rat mitochondrial ALDH, whereas the cytosolic isozyme could be recognized by antibody raised against horse cytosolic ALDH.

Crystallization and preliminary X-ray investigation of bovine liver mitochondrial aldehyde dehydrogenase

Aldehyde dehydrogenase from bovine liver mitochondria has been crystallized using the sitting drop method of vapor diffusion at 22 degrees C. The crystals formed from solutions containing, 40 mM-sodium citrate, 1 mM-NAD+ and 21% to 24% polyethylene glycol 3400 (pH 5.3 to 5.5). X-ray diffraction data collected from these crystals indicate that the crystals belong to the orthorhombic space group P2(1)2(1)2(1) with cell dimensions of a = 153.7 A, b = 159.37 A and c = 101.45 A. The crystals diffract to at least 2.9 A and a tetramer may comprise the asymmetric unit.

Polymorphism of the rat liver mitochondrial aldehyde dehydrogenase cDNA

In humans, a deficiency in mitochondrial aldehyde dehydrogenase (Class 2 ALDH) activity due to a single base-pair exchange in its structural gene serves as a deterrent to excessive alcohol consumption. Differences in Class 2 ALDH isozyme patterns on isoelectric focusing gels have been observed in the selectively bred, alcohol-preferring (P) and alcohol-nonpreferring (NP) lines of rats. To determine whether the differences are the result of sequence variation in the structural gene, we sequenced the cDNAs for Class 2 ALDH from P and NP rats. A synonymous exchange was seen in the codon for amino acid 473 in both lines, when compared with published sequences. Additionally, when the cDNA from P rats was used as reference, a substitution (G for A) was identified in the cDNA of NP rats which changes amino acid 67 from Gln (CAG codon; ALDH2Q allele) to Arg (CGG codon; ALDH2R allele). The Arg for Gln substitution makes the enzyme more basic and could account for the different electrophoretic mobilities. To determine whether the polymorphism was associated with drinking behavior, we genotyped the ALDH2 locus by amplifying rat genomic DNA encompassing the nucleotide exchange followed by probing with allele-specific oligonucleotides. There are highly significant differences in the frequencies of the two alleles in the P and NP rat lines. The frequency of the ALDH2R allele is 63% in the NP line and only 18% in the P line, whereas the frequency of the ALDH2Q allele is 82% in the P line and 37% in the NP line.

Molecular cloning of the mitochondrial aldehyde dehydrogenase gene of Saccharomyces cerevisiae by genetic complementation

Mutants of Saccharomyces cerevisiae deficient in mitochondrial aldehyde dehydrogenase (ALDH) activity were isolated by chemical mutagenesis with ethyl methanesulfonate. The mutants were selected by their inability to grow on ethanol as the sole carbon source. The ALDH mutants were distinguished from alcohol dehydrogenase mutants by an aldehyde indicator plate test and by immunoscreening. The ALDH gene was isolated from a yeast genomic DNA library on a 5.7-kb insert of a recombinant DNA plasmid by functional complementation of the aldh mutation in S. cerevisiae. An open reading frame which specifies 533 codons was found within the 2.0-kb BamHI-BstEII fragment in the 5.7-kb genomic insert which can encode a protein with a molecular weight of 58,630. The N-terminal portion of the protein contains many positively charged residues which may serve as a signal sequence that targets the protein to the mitochondria. The amino acid sequence of the proposed mature yeast enzyme shows 30% identity to each of the known ALDH sequences from eukaryotes or prokaryotes. The amino acid residues corresponding to mammalian cysteine 302 and glutamates 268 and 487, implicated to be involved at the active site, were conserved. S. cerevisiae ALDH was found to be localized in the mitochondria as a tetrameric enzyme. Thus, that organelle is responsible for acetaldehyde oxidation, as was found in mammalian liver.

Sequence of the precursor of bovine liver mitochondrial aldehyde dehydrogenase as determined from its cDNA, its gene, and its functionality

The partial sequence coding for bovine liver mitochondrial aldehyde dehydrogenase was previously reported (Farres, J., Guan, K.L., and Weiner, H., 1989, Eur. J. Biochem. 180, 67-74); cDNA coding for the N-terminal region has now been obtained by primer extension. The deduced 520 amino acids contained 499 residues corresponding to the mature mitochondrial aldehyde dehydrogenase which was one residue shorter than other mammalian mitochondrial aldehyde dehydrogenases. The N-terminal signal peptide had only 21 amino acids but showed high sequence identity to the signal sequences of human and rat enzymes, which have 17 and 19 amino acids, respectively, and had general characteristic features of typical mitochondrial signal peptides. Although the direct protein sequence of the signal peptide was not available, a mRNA coding for a 57-kDa precursor was proven to exist. The bovine signal peptide was able to direct the import of bovine aldehyde dehydrogenase precursor into isolated liver mitochondria. The complete nuclear gene coding for bovine mitochondrial aldehyde dehydrogenase was found to span at least 25 kb. A restriction map of this gene was constructed. Two potential transcription start sites were identified by primer extension and S1 nuclease protection. Several SpI transcription factor recognition sequences were found in the 5'-region of the gene. TATA box and CAAT box like sequences were found to be exist 18 and 64 bp upstream from the major transcription start site, respectively.

Purification and characterization of beef and pig liver aldehyde dehydrogenases

Beef liver cytosolic, mitochondrial, and pig liver mitochondrial aldehyde dehydrogenases (ALDH) had been purified to homogeneity. The two mitochondrial enzymes as with other mammalian mitochondrial enzymes had properties very similar to that of the corresponding human enzyme. These include immunological as well as basic kinetic properties such as low Km for aldehyde, activation by Mg2+ ions, and lack of inhibition by disulfiram. A major difference between these two enzymes and the human mitochondrial enzyme was that they contained an N-terminal-blocked amino acid. Cytosolic ALDHs from human and horse liver have been shown to possess an N-acetyl serine as the N-terminal residue; beef cytosolic ALDH was also found to be blocked. Tissue preparations and subcellular fractions from beef or pig liver could be used to study acetaldehyde oxidation. This is the subject of the accompanying paper (Cao Q-N, Tu G-C, Weiner H, Alcohol Clin Exp Res 12:xxx-xxx, 1988).