Polymorphism of alcohol dehydrogenase genes in alcoholic and nonalcoholic individuals from Valencia (Spain)

Hereditas 126: 247-253 (1997)

Polymorphism of alcohol dehydrogenase genes in alcoholic and nonalcoholic individuals from Valencia (Spain) C. ESPINOS, F. SANCHEZ, C. RAMIREZ, F. JUAN and C. NAJERA Departumento de Genttica, Facultad de Ciencias Biologicas, Universidud de Vulenciu, Vulencia, Spain

C. Espinos and F. Sanchez are equal first authors Espinbs, C., Sanchez, F., Ramirez, C., Juan, F. and Nhjera, C. 1997. Polymorphism of alcohol dehydrogenase genes in alcoholic and non alcoholic individuals from Valencia (Spain). -Hereditas 126: 247-253. Lund, Sweden. ISSN 0018-0661. Received April 3, 1997. Accepted June 4, 1997 Polymorphisms in the two variable ADHloci (ADH2 and ADH3) were analyzed in two groups (alcoholicsand nonalcoholics) from a Spanish population. The frequencies were similar to those reported for other Causcasian groups. ADH2-I and ADH3-I genotypes were more frequent in alcoholics than in nonalcoholics, but the differences were not significant. Carmen Najera, Departamento de GenPtica, Facultad de Ciencias Bioldgicas, Universidad de Valencia, Dr. Moliner, 50, ES-46100 Burjassot, Valencia (Spain). E-mail: [email protected]

Alcoholism is one of the most important problems facing modern society, with about 19 YOof males and 4 YOof females in the developed countries manifesting alcohol abuse or dependence (CLONINGER 1987). The elucidation and modeling of the genetic and environmental interactions involved constitute major challenges. Taken together, adoption and twin studies have provided practically irrefutable evidence of a genetic factor predisposing to alcoholism (MURRAYet al. 1983a,b; SCHUCKIT1985; STABENAU and HESSELBROCK 1983). Studies of biological markers capable of identifying individuals at risk for developing alcoholism have suggested numerous candidate genes, including the alcohol metabolizing enzymes, alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH). Both ADH and ALDH exhibit genetic heterogeneity (SMITH1986). Variations at the ADHl gene locus are rare. Nevertheless, allelic variants at both the ADH2 and ADH3 loci have been described (DUESTERet al. 1986). A variant enzyme form produced at the polymorphic ADH2 locus is commonly known as “atypical ADH”. The atypical *enzyme contains a variant p2 subunit instead of the usual PI subunit, exhibits much higher catalytic activity, and migrates more cathodic in starch gel electrophoresis (BOSRONand LI 1986). A new enzyme form at the ADH2 locus has been reported in black Americans (BOSRONet al. 1983) and is called ADH Indianapolis or p3 (BOSRONand LI 1987). The ADH3 locus codifies two different subunits termed y1 and y2. In Caucasians, the PI allele is predominant over the p2 allele; p2 is observed in only 10 YOof these individuals (STAMATOYANNOPOULOS et al. 1975), while 85 % of the Japanese (HARADA et al. 1980) and 89 YO

of the Chinese (TENGet al. 1979) are carriers of the atypical phenotype. In contrast, the latter study observed no carriers of the atypical phenotype among 43 Indians. BOSRONet al. (1983) showed that the frequency of p3 was 0.16 in the black American population, though in neither Caucasians nor Asians was this allele present (AGARWALet al. 1981). The polymorphic variants of y ADH also occur in racially distinct gene patterns. Caucasians possess the y, and y2 alleles at gene frequencies of 0.6 and 0.4, respectively (SMITH1986), while in Orientals y1 predominates over the y2 allele with a gene frequency of 0.91 (TENG et al. 1979). Similar findings apply to black Americans and Brazilians. In contrast, y2 is more frequent in American Indians (BOSRONand LI 1981; REX et al. 1985). An association between the ADH3-1 phenotype and the atypical variant of the ADH2 locus has been proposed (EHAOet al. 1994). The 7c enzyme (ADH4 gene product) was considered by LI et al. (1977) to be particularly important in the metabolism of alcohol in alcoholics, but variant forms of this enzyme have not been identified. Regarding the second enzyme of alcohol catabolism, there is no evidence for genetic variations in the composition of ALDH in Caucasian populations, though 50 YOof Oriental subjects show isozyme deficiency (DYCK1993; WALLet al. 1993). This absence has been related to the alcohol-flush reaction which makes alcohol consumption more difficult. This may explain the low incidence of alcohol problems in Oriental countries (SAUNDERS and WILLIAMS 1983). The nucleotide triplets coding for the p,, p2 and p3 ADH alleles (ADH2-1, ADH2-2 and ADH2-3) and the nucleotides which code for the ADH y1 and y2 alleles (ADH3-I and ADH3-2) have been characterized (DUESTERet al. 1986; IKUTAet al. 1986), as have the nucleotides which give rise to the ALDH2-1 and ALDH2-2 alleles (IMPRAIM et al. 1982).

248

Hereditas 126 (1997)

C. Esuinds et a[.

Table 1. Primers (Smith, personal commun.) and PCR conditions used to amplify ADH loci Name

Exon

Sequence

Annealing T“

PB3.1 PB3.2 PCRB2.1 PCRB2.2 G P1 GP2

9 9 3 3 8 8

S’-GGACTCTCACAACAAGCATG-3’ 5’-TGGCAAAGGTGACACAGTAG-3’ 5’-CTCTTTATTCTGTAGATGGT-3’ 5'-GTGTGA ATCCTGT ACCTGGT-3' 5’-GAATCTGTCCCCAAACTTGT-3’ 5’-CTCTTTCCAGAGCGAAGCAG-3’

60°C 60°C 58°C 58°C 50°C 50°C

otide probes directed to the point mutations that determine the coding differences between the A D H l , ADHZ, and ADH3 alleles. 1. Identification of ADH2-3 (Arg-369-Cys). An aliquot of the amplified mixture was digested with Bgl I1 and size fractioned over 0.8 ‘YO agarose gels, followed by transfer to nylon membranes (SOUTHERN 1975). Prehybridization was performed for at least 3 h in 5 x SSC/1 ‘YOblocking agent/O.l ‘YON-lauMATERIALS AND METHODS roylsarcosine/0.02 % SDS (prehybridization buffer). The hybridization buffer was constituted by 10 ng of Patients and controls. -Blood specimens were oblabeled probe with digoxigenine in 6 ml of prehytained from 71 alcoholic patients and 71 healthy bridization buffer. Hybridization with the probes controls from the Valencian Community (Spain). B3N and B3M (Table 2) was carried out at 45°C Blood samples from alcoholics were collected with overnight. The membranes were washed twice for 15 the help of different alcoholic associations including min at 62°C and detected using an antibody conjualcohol-dependent individuals trying to stop drinkgate (antidigoxigenin alkaline phosphatase conjugate) ing. Blood samples from nonalcoholic subjects were by chemiluminescent reaction with Lumigen PPD and extracted in the Hematology Department of Valencia exposure to X-rays. University General Hospital (Spain). The characteris2. Identification of ADH2-2 (Arg-47-His). The amtics of both samples were identical with regard to age, plified DNA was digested with Mae 111, because the sex, and ethnic origin. change creates a Mae 111 site (Fig. 2). The digested DNA extraction and ampl$cation. -Total genomic DNA was size fractioned on 20 YOacrylamide gel and DNA was extracted from peripheral blood (KUNKEL stained with silver. For better identification, SSCP et al. 1982). PCR amplifications were performed in a analysis was also carried out (ORITAet al. 1989). total volume of 50 p1 containing 200 ng of human Thus, after amplification, samples were denatured, genomic DNA, 50 mM Tris-HC1 pH 8.3, 1.5 mM electrophoresed on 8 ‘YO polyacrylamide gels (5 YO MgCl,, 0.01 % gelatin, 200 mM dNTPs, 1 pM of glycerol and 1 x TBE) at 700 V for approximately 4 each primer, and 1 unit of Taq DNA polymerase h at room temperature, and stained with silver. (F-501-L Dinazyme). The MgCl, was increased to 3 3. Identification of ADH3-1 and ADH3-2. After mM for PCR reactions containing PCRB2 primers. amplification, 1 pl of amplified DNA was denatured The primers used for identification of the different (95”C, 1 min) and DOT blotting was made. The alleles and the PCR conditions are reflected in Table DNA was fixed by ultraviolet exposure. The prehy1. Fig. 1 shows the gene structure around exons 3 and bridization was performed for at least 3 h in prehy9 of the ADH2 gene and the putative exon 8 of the bridization buffer, and hybridization with the probes ADH3 gene, along with the location and sequence of G1 and G2 (Table 2) was performed at 37°C the primers and probes used. When possible, amplifiovernight. The membranes were washed twice for 15 cation primers were chosen from regions in which the min at 46°C and detected by the chemiluminescent ADH2 and ADH3 genes were different, in order to method. decrease interference between them. Statistical analysis. -A contingency x2 was used for Determination of genotype. -Genotypes were detertesting differences between alcoholic and nonalcomined by hybridization to allele-specific oligonucleholic samples in both genes. Since in the Caucasian populations there are polymorphisms at the two ADH loci but not at the A L D H locus, and since the allelic forms of ADH2 and ADH3 differ in their efficacy for utilizing ethanol as a substrate, we decided to investigate whether certain alleles are more common/rare in individuals suffering from alcoholism.

c

221 bp

0 2 (ADH?)

GP2

Fig. 1. Structure of the ADH2 and ADH3 genes in the amplified regions. Double lines (= = ) represent exons, and discontinued lines (----)represent introns. Thick arrows ( 4 )denote products of PCR and double lined arrows (*) denote amplification primers. Vertical arrow (1) mark restriction site, and mutations are indicated with asterisks (*).

250

C. EsDinos et al.

Hereditas 126 (1997)

Table 2. Probes (Smith, personal commun.) and hybridization temperatures used to identijy ADH alleles Exon

Name

Hybridization T”

Sequence ~

S’-GCAGTATCCGTACCGTCCTG-3’ 5’-CAGGACGGTACAGATACTGC-3’ 5’-ATAACAAATATTTTACCTTT-3’

B3N B3M G1 G2

5'-AAAGGTA A AACATTTGTTAT-3'

RESULTS Identification of ADH2-3

The sequence of ADH2-1 starting 15-nt upstream of exon 9 (DUESTERet al. 1986) is identical to that of ADH2-3, except for the single substitution that changes the Arg 369 in ADH2-I to Cys in ADH2-3 (BURNELL et al. 1987). After amplification of blood cell DNA, a 366-pb fragment could be clearly observed by ethidium bromide staining. The amplified DNA was digested by Bgl I1 to identify ADH2 from the other class I loci. Hybridization with B3M (mutant typing oligo) was not observed; however, all the samples hybridized with B3N (normal typing oligo). Thus, there was no p, allele in the analyzed samples. Identification of ADH2-2

Amplified DNA shows a 169-bp fragment with a Mae 111 restriction site only for the ADH2-2 phenotype. Fig. 3 shows the SSCP analysis used to detect the ADH2 polymorphism. Table 3 shows the different ADH2 phenotypes for alcoholic and nonalcoholic subjects. The contingencyx2 shows that differences in allele frequencies between nonalcoholic and alcoholic subjects were not significant.

45°C 45°C 37°C 37°C

only y,, only yz, or heterozygotes for both alleles, as can be observed in Fig. 4. Table 4 shows the different ADH3 phenotypes for alcoholic and nonalcoholic subjects. The differences in the frequencies between alcoholic and nonalcoholic subjects were non significant, (P < 0.2). DISCUSSION Alcoholism appears to involve a complex phenotype sequentially modified by interaction among multiple gene loci and many environmental factors. It cannot be diagnosed in persons unexposed to the triggering environmental agent ethanol. It is important to identify the risk alleles at the relevant loci in order to prevent the disease. With regard to polymorphism of the alcohol metabolizing enzymes, GOEDDE et al. (1992) determined the distribution of human liver alcohol dehydrogenase (ADH2) and aldehyde dehydrogenase (ALDH2) genotypes in 21 different populations comprising Mongoloids, Caucasoids, and Negroids by hybridization of the amplified genomic DNA with allele-specific oligonucleotide probes. Whereas the frequency

Identification ofADH3-1 and ADH3-2

Amplification of DNA produced a 124-bp segment. The y, probe differs from both ADHl and ADH2 genes by 3-nt, and the y2 probe by 2-nt; consequently, neither hybridizes under the conditions used. There is a clear discrimination between samples containing Mae I l l

Mae Ill

rn1.1~1...................... JI ..................... JI _____________ 169bp 67 bp

69 bp

Mae ___________MaeJI_______Mae &____________----_---_ JI -----------Ill

ADHZ2

34 bP

34bp

111

35bp A

Ill

67 bp

1@bP

34 bP

4

Fig. 2. ldentification of ADH: by means of a new Mae 111 site. Vertical arrows (I) mark restriction sites.

Fig. 3. Polymorphism detected in the ADH2 locus by SSCP’s analysis. The arrow indicates p2 allele. Lanes 2, 3 and 6 show ADH; homozygotes, and lanes I , 4 and 5 heterozygotes.

Polymorphism of alcohol dehydrogenase genes

Hereditas 126 (1997)

25 1

Table 3. Frequencies of the dijfeerent ADH2 alleles and genotypes in alcoholic and nonalcoholic individuals n

Nonalcoholics Alcoholics

71 71

Genotypes

Alleles

PI P I

PI P 2

58

12

62

9

P2P2

ADH2-1

ADH2-2

Contingency

1 0

0.90 0.94

0.10 0.06

1.56 ns

of the ADH2-1 allele was found to be relatively high among Caucasoids, Mexican Mestizos, Brazilian Indians, Swedish Lapps, Papua New Guineans, and Negroids, the frequency of the ADH2-2 gene was considerably higher in Mongoloids and Australian Aborigenes. The atypical ALDH2 gene (ALDH2-2) was found to be extremely rare in Caucasoids, Negroids, Papua New Guineans, Australian Aborigens, and Aurocanians (Southern Chile). In contrast, this mutant gene was found to be common among Mongoloids (50 ‘YO). Differences in drinking behavior due to uncomfortable or embarrasing acute flushing reactions to ethanol have been attributed to inherited polymorphic differences in metabolism of ethanol and acetaldehyde (GOEDDE et al. 1979; HARADAet a1 1980; SHIBUYAand YOSHIDA1988a,b); thus, individuals who have inactive low Km aldehyde dehydrogenase (ALDH2) are much less likely to develop alcoholism than those with active ALDH2 (SUZUKYet al. 1992; YOSHIDA1992). Those Japanese Americans who exhibit the fast flushing response tend to drink less than those who do not flush (NAKAWATASE et al. 1993). On the other hand, SLUTSKE et al. (1995) concluded that alcohol-related flushing is not a protective factor against alcoholism in Caucasians and may in fact constitute a risk factor. The enzymatically inactive mitochondria1 ALDH isozyme, combined with a rapid ethanol metabolism catalyzed by the superactive atypical ADH, might be primarily responsible for the intense adverse reaction commonly observed in Orientals after drinking a small dose of alcohol. Thus, polymorphism of the genes involved in alcohol metabolism seems to be an important predisposing or protective factor for alcoholism in Oriental popula-

Fig. 4. Different ADH3 phenotypes for alcoholic and nonalcoholic individuals identified by GI and G2 probes. Lanes 1 and 5 show ADH: homozygotes, lane 2 ADH: homozygote and lanes 3 and 4 heterozygotes.

x2

tions. In fact, CHAOet al. (1994) found that Chinese alcoholics without liver disease had significantly lower frequencies of the ADH2-2 and ADH3-1 alleles, compared with nonalcoholic individuals. The data strongly suggest that genetic variation of both ADH and ALDH may influence drinking behavior and the risk of alcoholism developing through acetaldehyde formation. The ADH atypical allele in our population was less frequent among alcoholics than in nonalcoholics, though the difference was not significant. A study of another Spanish population (VIDALet al. 1993), analyzing the presence of atypical liver alcohol dehydrogenase isozyme in samples of liver tissue from 222 alcoholic and nonalcoholic subjects, showed that atypical ADH was present in 14.9% of alcoholics and in 17.40/0 of nonalcoholics (P = ns). Our data confirm these nonsignificant differences, indicating that the presence of this isozyme does not play a role in the development of alcohol liver disease in the population analyzed. With respect to the ADH3 genotype, non-significant differences between nonalcoholics and alcoholics were found in our population (P < 0.20). On the other hand, the alcoholic group does not fit Hardy-Weinberg equilibrium due to a deficiency in the observed proportion of heterozygotes (Wahlund’s effect). Among the possible reasons of this deficiency, we think that the most likely one is positive assortative mating, since it is evident that social and economic barriers exist that prevent or limit marriages. Other authors have also studied these polymorphisms in other Caucasian populations in order to compare alcoholic and nonalcoholic genotypes. GILDERet al. (1993) used DNA analysis to type 82 Caucasian patients receiving treatment for alcohol-related problems and 84 controls for variants in the alcohol dehydrogenase (ADH2 and ADH3) and mitochondrial aldehyde dehydrogenase (ALDH2) gene loci; they found no association between individual or combined gene frequencies and the presence of alcohol-related problems. SMITHet al. (1972) found that in Caucasians with the atypical form of ADH, the ADH3-1 allele occurs with a higher frequency than expected. However, in our population there was no linkage disequilibrium

252

C. Espinos et al.

Hereditas 126 (1 997)

Table 4. Frequencies of the different ADH3 alleles and genotypes in alcoholic and nonalcoholic individuals

n

Nonalcoholics Alcoholics

71 69

Genotypes

Alleles

YI YI

YlY2

Y2Y2

ADH3-1

ADH3-2

Contingency

29 40

28 17

14 12

0.60 0.70

0.40 0.30

4.55(p < 0.2)

between both alleles. One possible explanation is that linkage disequilibrium exists between the ADH2 a n d ADH3 alleles because the ADH2-2 allele was introduced into the European population relatively recently by a founder with the ADH3-1 genotype, a n d that the equilibrium not yet reached in the population a t that time has now been attained.

ACKNOWLEDGMENTS To the Ministerio de Educacion y Ciencia and the Ministerio de Asuntos Sociales (Spain) for financial support.

REFERENCES Agarwal DP, Meier-Tackman D, Harada S and Goedde HW, (1981). A search for the Indianapolis variant of human alcohol dehydrogenase in liver samples from Northern Germany and Japan. Hum. Genet. 50: 170171. Bosron WF and Li TK, (1981). Genetic determinants of alcohol and aldehyde dehydrogenases and alcohol metabolism. Sem. Liver Dis. 1: 179-188. Bosron WF and Li TK, (1986). Genetic polymorphism of human liver alcohol and aldehyde dehydrogenase, and their relationship to alcohol metabolism and alcoholism. Hepatology 6: 502-510. Bosron WF and Li TK, (1987). Catalytic properties of human liver alcohol dehydrogenase isoenzymes. Enzyme 37: 19-28. Bosron WF, Magnes LJ and Li TK, (1983). Human liver alcohol dehydrogenase: ADH Indianapolis results from genetic polymorphism at the ADH2 gene locus. Biochem. Genet. 21: 735-744. Burnell C, Carr LG, Dwulet FE, Edenberg HJ, Li TK and Bosron WF, (1987). The human p3 alcohol dehydrogenase subunit differs from PI by a Cys for Arg,,, substitution which decreases NAD(H) binding. Biochem. Biophys. Res. Commun. 146: 1227-1233. Chao YC, Liou SR, Chung YY, Tang HS, Hsu CT, Li TK and Yin SJ, (1994). Polymorphism of alcohol and aldehyde dehydrogenase genes and alcoholic cirrhosis in Chinese patients. Hepatology 19: 360-366. Cloninger CR, (1987). Recent advances in family studies of alcoholism, In: Genetics and Alcoholism (eds HE Goedde and DP Agarwal.) A.R. Liss, N.Y., p. 47-60. Duester G, Smith M, Bilanchone V and Hatfield GW, (1986). Molecular analysis of the human class I alcohol dehydrogenase gene family and nucleotide sequence of the gene encoding the subunit. J. Biol. Chem. 261: 2027-2033. Dyck LE, (1993). Absence of the atypical mitochondria1 aldehyde dehydrogenase (ALDH2) isozyme in Saskatchewan Cree Indians. Hum, Hered. 43: 116- 120.

x2

Gilder FJ, Hodgkinson S and Murray RM, (1993). ADH and ALDH genotype profiles in Caucasians with alcoholrelated problems and controls. Addiction 88: 383-388. Goedde HW, Harada S and Agarwal DP, (1979). Racial differences in alcohol sensitivity: A new hypothesis. Hum. Genet. 51: 331-334. Goedde HW, Agarwal DP, Fritze G, Meier-Tackmann D, Singh S, Beckmann G, Bhatia K, Chen LZ and Fang B, (1992). Distribution of ADH2 and ALDH2 genotypes in different populations. Hum. Genet. 88: 344-346. Harada S, Misawa S, Agarwal DP and Goedde HW, (1980). Liver alcohol dehydrogenase and aldehyde dehydrogenase in the Japanese: Isozyme variation and its possible role in alcohol intoxication. Am. J. Hum. Genet. 32: 8-15. Ikuta T, Szeto S and Yoshida A, (1986). Three human alcohol dehydrogenase subunits: cDNA structure and molecular and evolutionary divergence. Proc. Natl. Acad. Sci. USA 83: 634-638. Impraim C, Wang G and Yoshida A, (1982). A structural mutation in a major human aldehyde dehydrogenase gene results in loss of enzyme activity. Am. J. Hum. Genet. 34: 837-841. Kunkel LM, Tantravah U, Eisenhard M and Latt SA, (1982). Regional localization on the human X of DNA sequences cloned from flow sorted chromosomes. Nucleic Acids Res. 10: 1557-1571. Li TK, Bosron WF, Dafeldecker WP, Lange LG and Vallee BL, (1977). Isolation of n alcohol dehydrogenase of human liver: is it a determinant of alcoholism? Proc. Natl. Acad. Sci. USA 77: 5784-5788. Murray RM, Clifford CA and Gurling HMD, (1983a). Twin and adoption studies: How good is the evidence for a genetic role?, In: Recent Developments in Alcoholism, vol. 1 (ed. M Galanter) Plenum Press, N. Y., p. 25-48. Murray RM, Clifford C, Curling HMD, Topham A, Clow A and Bernadt M, (1983b). Current genetic and biological approaches to alcoholism. Psychiatr. Dev. 1: 179192. Nakawatase TV, Yamamoto J and Sasao T, (1993). The association between fast-flushing response and alcohol use among Japanese Americans. J. Stud. Alcohol. 54: 48-53. Orita M, Suzuki Y, Sekiya T and Hayashi K, (1989). Rapid and sensitive detection of point mutations and DNA polymorphisms using the polymerase chain reaction. Genomics 5: 874-879. Rex DK, Bosron WF, Smialek JE and Li TK, (1985). Alcohol and aldehyde dehydrogenase isozymes in North American Indians. Alcohol Clin. Exp. Res. 9: 147-152. Saunders JB and Williams R, (1983). The genetics of alcoholism: Is there an inherited susceptibility to alcohol-related problems? Alcohol Alcoholism 18: 189-217. Schuckit MA, (1985). Genetics and the risk for alcoholism. JAMA 254: 2614-2617.

Hereditas 126 (1997)

Polymorphism of alcohol dehydrogenase genes

253

Shibuya A and Yoshida A, (1988a). Frequency of the atypical aldehyde dehydrogenase-2 gene (ALDH2-2) in Japanese and Caucasians. Am. J. Hum. Genet. 43: 741743. Shibuya A and Yoshida A, (1988b). Genotypes of alcohol metabolizing enzymes in Japanese with alcoholic liver diseases: a strong association of the usual Caucasian type aldehyde dehydrogenase (ALDH) with the disease. Am. J. Hum. Genet. 43: 744-748. Slutske WS, Heath AC, Madden PAF, Bucholz KK, Dinwiddie SH, Dunne MP, Statham DS, Whitfield JB and Martin NG, (1995). Is alcohol-related flushing a protective factor for alcoholism in Caucasians? Alcohol Clin. Exp. Res. 19: 582-592. Smith M, (1986). Genetics of human alcohol and aldehyde dehydrogenases. Adv. Hum. Genet. 15: 249-290. Smith M, Hopkinson DA and Harris H, (1972). Alcohol dehydrogenase isozymes in stomach and liver: Evidence for activity of the ADH3 locus. Ann. Hum. Genet. 35: 243-253. Southern E, (1975). Detecting of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98: 503-517.

Stabenau JR and Hesselbrock VM, (1983). Family pedigree of alcoholic and control patients. Int. J. Addict. 18: 351-363. Stamatoyannopoulos G, Chen SH and Fukui M, (1975). Liver alcohol dehydrogenase in Japanese: High frequency of atypical form and its role in alcohol sensitivity. Am. J. Hum. Genet. 27: 789-796. Suzuki K, Muramatsu T, Yamada K and Kono H, (1992). ALDH phenotype in eating disorders with and without alcoholism. Japan. J. Alcohol Drug Depend. 27: 629633. Teng YS, Jehan S and Lie-Ijo LE, (1979). Human alcohol dehydrogenase ADH2 and ADH3 polymorphisms in ethnic Chinese and Indians of West Malaysia. Hum. Genet 53: 87-90. Vidal F, PCrez J, Panisello J, Toda R, Gutierrez C and Richart C, (1993). Atypical liver alcohol dehydrogenase in the Spanish population: Its relation with the development of alcoholic liver disease. Alcohol Clin. Exp. Res. 17: 782-785. Wall TL, Gallen CC and Ehlers CL, (1993). Effects of alcohol on the EEG in Asian men with genetic variations of ALDH2. Biol. Psychiatry 34: 91-99. Yoshida A, (1992). Molecular genetics of human aldehyde dehydrogenase. Pharmacogenetics 2: 139- 147.