Genotoxic effects of alcohol in human peripheral lymphocytes modulated by ADH1B and ALDH2 gene polymorphisms

Mutation Research 615 (2007) 134–142

Genotoxic effects of alcohol in human peripheral lymphocytes modulated by ADH1B and ALDH2 gene polymorphisms Hitoshi Ishikawa a,c,∗ , Takashi Ishikawa a , Hidetaka Yamamoto b , Akira Fukao c , Kazuhito Yokoyama a a

c

Department of Public Health and Occupational Medicine, Mie University Graduate School of Medicine, Edobashi 2-174, Tsu 514-8507, Japan b Department of Microbiology, Suzuka University of Medical Science, Suzuka 510-0293, Japan Department of Public Health, Yamagata University Graduate School of Medicine, Iida-Nishi 2-2-2, Yamagata 990-9585, Japan Received 29 August 2006; received in revised form 14 November 2006; accepted 17 November 2006

Abstract Ethanol is almost totally broken down by oxidative metabolism in vivo. Ethanol per se is considered to be neither carcinogenic, mutagenic nor genotoxic. However, during the metabolic conversion of ethanol to acetaldehyde and acetate, the organism is exposed to both ethanol and acetaldehyde and therefore ethanol is suspected to be co-carcinogenic. The genetic polymorphisms of alcohol dehydrogenase-2 (ADH1B) and acetaldehyde dehydrogenase-2 (ALDH2) influence the metabolism of alcohol. The ADH1B*1/*1 genotype encodes the low-activity form of ADH1B, and ALDH2*1/*2 and ALDH2*2/*2 genotype encode inactive ALDH2. The aim of this study was to test the hypothesis that polymorphisms of the ADH1B and ALDH2 genes are significantly associated with genotoxicity induced by alcohol drinking, measured using the cytokinesis-block micronucleus (CBMN) assay, an established biomarker of genome instability, in peripheral blood lymphocytes of 286 healthy Japanese men. There was a significant trend for the mean micronuclei (MN) frequency in habitual or moderate drinkers without a smoking habit to increase as the numbers of the *1 allele in ADH1B increased (P = 0.039 or P = 0.029) and the *2 allele in ALDH2 increased (P = 0.019 or P = 0.037). A logistic regression analysis showed that the number of subjects with MN frequency levels more than median value of MN (3.0) was significantly higher in the subjects with the ADH1B*1 allele as adjusted estimates (OR 2.08, 95% C.I. 1.24–3.48), when the OR for the subjects with the ADH1B*2/*2 genotype was defined as 1.00. The number of subjects with MN frequency levels more than median value of MN was also significantly higher in the subjects with the ALDH2*2 allele as adjusted estimates (OR 1.79, 95% C.I. 1.04–3.11), when the OR for the subjects with the ALDH2*1/*1 genotype was defined as 1.00. The results of this study have identified important novel associations between ADH1B/ALDH2 polymorphisms and genotoxicity in alcohol drinkers. © 2006 Elsevier B.V. All rights reserved. Keywords: Acetaldehyde; Ethanol; Micronuclei; ADH1B; ALDH2

1. Introduction ∗

Corresponding author. Tel.: +81 23 628 5259; fax: +81 23 628 5261. E-mail address: [email protected] (H. Ishikawa). 0027-5107/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.mrfmmm.2006.11.026

The International Agency for Research on Cancer concluded that epidemiological data were conclusive enough to classify alcoholic beverages as group I

H. Ishikawa et al. / Mutation Research 615 (2007) 134–142

carcinogens in humans [1]. Epidemiological data have also identified chronic alcohol consumption as a significant risk factor for alimentary tract cancer [2]. Ethanol is almost totally broken down by oxidative metabolism in vivo. Numerous experiments in prokaryotic and eukaryotic cell cultures, as well as animal models, have shown that acetaldehyde, a reactive and toxic metabolite of ethanol, could induce genotoxic and mutagenic effects [3–7]. On the other hand, ethanol per se is considered to be neither carcinogenic, mutagenic nor genotoxic [8,9]. However, during metabolic conversion of ethanol to acetaldehyde an organism is exposed to both substances and therefore ethanol is suspected to be co-carcinogenic [10]. Individuals who accumulate acetaldehyde due to polymorphic differences in the genes encoding the enzymes responsible for acetaldehyde generation and detoxification have been thought to show an ethanolassociated carcinogenesis [2]. In humans, more than 90% of ingested alcohol is eliminated via metabolic degradation mainly in the liver. Ethanol is first metabolized into acetaldehyde through several enzymatic and nonenzymatic mechanisms, the main enzymatic pathways being alcohol dehydrogenase (ADH), cytochrome P4502E1 (CYP2E1) and catalase, although aldehyde dehydrogenase (ALDH) metabolizes the bulk of acetaldehyde in the liver [11]. The efficiency in converting ethanol to acetaldehyde, and subsequent conversion to acetate, is largely determined by the ADH and ALDH gene families, with large potential interindividual differences in acetaldehyde exposure due to the presence of some well-studied common genetic variants with a functional role [12]. Two of the ADH genes, ADH1B and ADH1C, are polymorphic resulting in between-person variation in ethanol metabolism [13]. In the genes encoding enzymes responsible for alcohol metabolism, polymorphic variants that influence Vmax , and Km have been well described for the ADH1B gene [14]. The ADH1B has a polymorphism (Arg47His), and its ADH1B*2 allele, which is only common in East Asians and exists in more than 90% of the populations, but in fewer than 20% of Caucasians and Africans [15]. The ADH1B*2 allele encodes a super-reactive subunit of ADH1B and has about 100-times the oxidation capability of the ADH1B*1 allele [16], and the super-reactive ADH1B*2 homodimer has about 40-times higher Vmax than the less-active ADH1B*1/*1 form of ADH1B [14]. The individuals with ADH1B*2 allele show a higher elimination rate for blood ethanol compared with those that have the ADH1B*1/*1 genotype [17]. Acetaldehyde is metabolized into acetate by ALDH. Although there are multiple forms of ALDH in the liver [18], the

135

mitochondrial enzyme encoded by the ALDH2 has a very low Km for acetaldehyde and is responsible for oxidation of acetaldehyde. The ALDH2 has a polymorphism (Glu487Lys) and its alleles encoding the active and inactive subunits have been termed as ALDH2*1 and ALDH2*2, respectively [19]. This polymorphism is prevalent in more than 50% of East Asians but has not been found in Caucasians or Africans [20]. The individuals with at least one ALDH2*2 allele show markedly higher blood acetaldehyde concentrations compared with those that have the ALDH2*1/*1 genotype [21]. Therefore, one may predict that, if exposure to acetaldehyde is the primary alcohol-related carcinogen for alimentary tract cancer, the effect of alcohol would be greater among fast metabolizing individuals, i.e. those possessing ADH1B*2/*2 and ALDH2*1/*2 genotypes. Many studies, mainly investigating East Asians, have reported that the ADH1B*1/1 and ALDH2*1/2 genotypes were more strongly associated with the development of esophageal cancer than their counterpart genotypes [22–27]. Similarly, it has been also suggested that a significantly higher increased risk of colorectal cancer [28] was observed in subjects with both the inactive ADH1B*1 allele and ALDH2*1/2 genotype. These findings were contrary to the above-mentioned original hypothesis that fast metabolism of alcohol would lead to an increased peak of acetaldehyde exposure and therefore greater risk. This would also argue against an important role of acetaldehyde in carcinogenesis and mutagenesis in alcohol metabolism. In this respect, one possible explanation would be that the less active form of ADH1B*1 and ALDH2*1/*2 might cause more accumulation of ethanol and acetaldehyde compared to those with active ADH1B*2 and ALDH2*1/*1, respectively. Subsequently, the levels of genotoxicity induced by ethanol and acetaldehyde would be elevated in specific tissues and lead to carcinogenesis. The primary aim of this study was to test the hypotheses that ethanol has a genotoxicity as well as acetaldehyde and polymorphisms of ADH1B and ALDH2 genes are significantly associated with genotoxicity induced by alcohol drinking, measured using the cytokinesis-block micronucleus (CBMN) assay in human peripheral blood lymphocytes. 2. Materials and methods 2.1. Study populations Some parts of this study consisted of a population enrolled in a previous study [29] for assessment of the effects of several genetic polymorphisms on MN frequencies. Studies were

136

H. Ishikawa et al. / Mutation Research 615 (2007) 134–142

carried out on 286 healthy Japanese men. The participation of each subject was voluntary and the subjects could withdraw at any time during the study (according to the Helsinki II declaration). Written informed consent was obtained from each subject according to the ethical guidelines of the Japanese Government. This study was approved by the Ethics Committee at Mie University School of Medicine. 2.2. Questionnaires Each participant completed a structured questionnaire concerning his alcohol drinking and smoking habits. Regarding smoking, the subjects were divided into current smokers, former smokers and never smokers. Former-smokers were defined as subjects who had quit smoking 1 year or more before completing the questionnaire. Regarding smoking status, subjects, who were never or current smokers, were stratified into five subgroups: never, 1–10 cigarettes/day, 11–20 cigarettes/day, 21–30 cigarettes/day and ≥31 cigarettes/day. With regarding to the frequency of alcohol drinking, the intensity of drinking was evaluated as never drinking, non-habitual drinking (≤3 times per week) and habitual drinking (>3 times per week). Furthermore, regarding alcohol consumption, subjects were categorized into three levels, never, moderate and heavy drinker. It has been widely realized that the volume of alcohol intake is directly proportional to alcohol-induced liver damage [30]. In Japan, 3 “GOU” units of Japanese Sake once is accepted as a generally concept of heavy drinking in Japan [31]. This is equivalent to 60 g of ethanol. Therefore, we defined intake of a mean of 60 g or more of alcohol once as the criterion for “heavy drinker” in this study. Heavy drinkers were defined as those who drank a mean of 60 g or more of alcohol each time they consumed alcohol, while moderate drinkers were defined as those who drank a mean of less than 60 g of alcohol each time they consumed alcohol. 2.3. Sample collection All subjects contributed a single blood donation during the period February 2000 to March 2005. Blood was collected by venipuncture from fasted subjects, and aliquoted in heparinized tubes, used for genotoxicity analysis, and in tubes devoid of anticoagulant, used for DNA extraction and amplification. Tubes were transferred to the laboratory within a few hours and immediately used for lymphocyte cultures or stored −80 ◦ C until analysis. For each individual, all analyses were carried out using aliquots of the same blood sample. 2.4. Analysis of lymphatic MN Some parts of data regarding MN were obtained within the framework of our previous study on the effects of lifestyle factors and genetic polymorphisms on spontaneous MN frequency [29]. MN were analyzed using the cytokinesis-block method [32] with Giemsa staining. MN were scored among 1000 lymphocyte binucleated cells according to published criteria [33].

2.5. ADH1B and ALDH2 genotyping The ADH1B single nucleotide polymorphism (SNP) examined is a G to A transition in exon 3 which translates hystidine instead of arginine at the residue 47 (Arg47Hys). The ALDH2 SNP examined is a G to A transition in exon 12 which translates lysine instead of glutamic acid at the residue 487 (Glu487Lys). PCR-RFLP methods were performed on lymphocyte DNA samples from all participants, without knowledge of MN frequency status, to determine ADH1B [34] and ALDH2 [35]. 2.6. Statistical analysis The significance of a difference in mean MN frequencies between two groups was estimated by the Student t-test, and among more than two groups by the one-way ANOVA test followed by Fisher PLSD multiple comparison test. The logistic regression analysis was carried out to calculate ORs adjusted for different possible covariates (age, smoking and alcohol drinking habits) using a dichotomous variable for MN frequency levels (above/equal to or below 3.0 MN frequency level), because the median value of MN was 3.0 among all subjects. The Chi-square test was used to examine the distributions of subjects with the polymorphic genotypes between the younger (≤43) and older (>43) age groups. This test was also used to compare the allele proportions and to determine whether the ADH1B and ALDH2 allelic frequencies conformed to the Hardy–Weinberg equilibrium. Values are presented as the mean ± S.E. unless otherwise stated. The level of significance was P < 0.05. All statistical analyses were carried out using the SPSS 13.0 software package.

3. Results 3.1. Effects of demographic characteristics on MN frequency The median age of the subjects was 43.0 years (range: 19–65). As shown in Table 1, the subjects of >43 years of age had a higher mean MN frequency than those of ≤43 years of age (P < 0.001). The MN frequency was positively correlated with age (r = 0.29, P < 0.0001). No significant differences in the mean MN frequencies were observed among smoking and alcohol drinking subgroups. 3.2. Effects of ADH2 and ALDH2 genotypes on MN frequency Table 2 summarized the distributions of the ADH1B and ALDH2 genotypes and their mean MN frequencies. The genotype and allele frequencies for the polymorphisms analyzed were calculated by direct counting and found to be in Hardy–Weinberg equilibrium. There was a

ADH1B 1/1 1/2 2/2 ALDH2 1/1 1/2 2/2 a

n (%)

Mean ± S.E.

25 (8.7) 80 (28.0) 181 (63.3) P for HWa 0.082

4.16 ± 0.44 3.61 ± 0.26 3.30 ± 0.18 P for trend 0.041

151 (52.8) 118 (41.3) 17(5.9) P for HWa 0.524

3.24 ± 0.18 3.70 ± 0.21 3.88 ± 0.80 P for trend 0.140

P values for deviation from Hardy–Weinberg equilibrium.

3.47 ± 0.32 (55) 4.10 ± 0.55 (20) (0) 0.299 3.41 ± 0.31 (46) 4.83 ± 0.45 (18) 4.00 (1) 0.019 3.45 ± 0.22 (101) 4.45 ± 0.36 (38) 4.00 (1) 0.019

5.00 ± 0.93 (8) 4.46 ± 0.55 (13) 3.41 ± 0.30 (44) 0.039 4.25 ± 0.60 (16) 3.97 ± 0.35 (30) 3.55 ± 0.24 (94) 0.156 2.00 (1) 4.19 ± 0.75 (16) 3.45 ± 0.52 (29) 0.390

3.54 ± 0.81 (13) 3.57 ± 0.50 (30) 5.33 ± 2.33 (3) 0.670 2.36 ± 0.35 (28) 3.17 ± 0.47 (24) (0) 0.163 2.73 ± 0.36 (41) 3.39 ± 0.34 (54) 5.33 ± 2.33 (3) 0.088 Non-smokers include never- and former-smokers.

Genotypes

a

Table 2 Subjects and mean MN frequency by ADH1B and ALDH2 genotypes

3.00 ± 0.85 (7) 3.10 ± 0.62 (10) 4.00 ± 1.27 (9) 0.738

The mean MN frequencies in relation to the ADH1B and ALDH2 genotypes regarding to drinking and smoking status are shown in Tables 3 and 4. The mean MN

4.00 ± 1.00 (2) 3.31 ± 0.42 (16) 2.50 ± 0.87 (4) 0.248

3.3. Effects of alcohol drinking and ADH2 and ALDH2 genotypes on MN frequency

3.22 ± 0.68 (9) 3.23 ± 0.34 (26) 3.54 ± 0.92 (13) 0.704

significant trend for the mean MN frequency to increase as the numbers of the *1 allele in ADH1B increased (P = 0.041).

ALDH2 1/1 1/2 2/2 P for trends

P < 0.001 by the Student t-test. Excluding former smokers.

4.80 ± 0.97 (5) 2.61 ± 0.53 (18) 2.45 ± 0.34 (29) 0.154

0.21 0.56 0.33 0.37 0.48

4.33 ± 0.92 (6) 3.35 ± 0.46 (34) 2.95 ± 0.31 (58) 0.151

± ± ± ± ±

4.00 (1) 3.58 ± 0.60 (12) 3.15 ± 0.94 (13) 0.220

3.11 2.60 4.11 3.44 2.95

3.00 ± 1.00 (2) 3.25 ± 1.03 (4) 3.25 ± 0.41 (16) 0.882

a b

84 15 70 41 21

3.33 ± 0.67 (3) 3.50 ± 0.50 (16) 3.21 ± 0.47 (29) 0.420

Smoking exposureb Never 1–10 cigarettes/day 11–20 cigarettes/day 21–30 cigarettes/day More than 30 cigarettes/day

ADH2 1/1 1/2 2/2 P for trends

3.11 ± 0.21 3.67 ± 0.30 3.61 ± 0.21

Non-smokersa Mean ± S.E. (n)

84 55 147

Overall Mean ± S.E. (n)

Smoking status Never Former Current

Smokers Mean ± S.E. (n)

3.31 ± 0.33 3.47 ± 0.17 3.64 ± 0.30

Mean ± S.E. (n)

48 202 36

Overall Mean ± S.E. (n)

Alcohol consumption Never Moderate Heavy

Smokers Mean ± S.E. (n)

3.31 ± 0.33 3.17 ± 0.25 3.72 ± 0.19

Mean ± S.E. (n)

48 98 140

Overall Mean ± S.E. (n)

Drinking frequency Never Non-habitual drinkers Habitual drinkers

Habitual drinkers

2.97 ± 0.16 4.04 ± 0.22a

Non-smokersa

147 139

Non-habitual drinkers

Age Younger subjects (≤43) Elderly subjects (>43)

Non-smokersa

MN frequency (mean ± S.E.)

Never drinkers

No. of subjects

Table 3 Mean MN frequencies for ADH1B and ALDH2 genotypes according to drinking frequency and smoking status

Age and habits

Smokers Mean ± S.E. (n)

Table 1 Subjects and mean MN frequency by age, drinking and smoking

137 3.50 ± 0.73 (8) 3.59 ± 0.45 (17) 3.68 ± 0.37 (50) 0.920

H. Ishikawa et al. / Mutation Research 615 (2007) 134–142

3.50 ± 0.62 (10) 5.00 (1) (0) 0.422

3.55 ± 0.42 (20) 4.00 ± 0.71 (5) (0) 0.580

frequency increased significantly in the genotype order: *1/*1, *1/*2 and *2/*2 genotypes in ADH1B (P = 0.039 or P = 0.029) and *1/*1, *1/*2 and *2/*2 genotypes in ALDH2 (P = 0.019 or P = 0.037) in habitual and moderate drinkers without a smoking habit. We also examined the effects on mean MN frequency of combined genotypes for ADH1B and ALDH2 and drinking and smoking status. Our analysis showed that subjects with at least one ADH1B*1 allele and at least ALDH2*2 revealed the highest mean MN frequency in the habitual and moderate drinkers (Tables 5 and 6). In contrast, subjects with the significantly lowest mean MN frequency in the habitual and moderate drinkers without smoking habit had a combined genotype of ALDH2*1/*1 with ADH1B*2/*2 (P < 0.05). In the nonhabitual drinkers without smoking habit, subjects with both ADH1B*2/*2 and ALDH2*1/*1 genotypes showed the significantly lowest mean MN frequency (P < 0.05). To investigate the possible modulation of MN frequency by age, the distributions of the ADH1B and ALDH2 genotypes were compared between the younger and elderly subjects and the ADH1B and ALDH2 genes polymorphisms were equally distributed in both groups (data not shown).

3.46 ± 0.39 (48) 3.76 ± 0.40 (45) 5.33 ± 2.33 (3) 0.426 2.94 ± 0.26 (64) 3.85 ± 0.36 (41) 4.00 (1) 0.037

3.4. Logistic regression analysis for the relationship between MN frequency and age, smoking and alcohol drinking habits, and ADH2 and ALDH2 genotypes

Non-smokers include never- and former-smokers. a

4.00 ± 1.00 (2) 3.31 ± 0.42 (16) 2.50 ± 0.87 (4) 0.248 3.22 ± 0.68 (9) 3.23 ± 0.34 (26) 3.54 ± 0.92 (13) 0.704 ALDH2 1/1 1/2 2/2 P for trends

3.00 ± 0.85 (7) 3.10 ± 0.62 (10) 4.00 ± 1.27 (9) 0.738

3.16 ± 0.22 (112) 3.80 ± 0.27 (86) 5.00 ± 1.68 (4) 0.037

3.53 ± 0.34 (30) 4.17 ± 0.60 (6) (0) 0.426

2.50 ± 1.50 (2) 4.33 ± 0.88 (3) 3.67 ± 0.88 (6) 0.723 4.00 ± 0.71 (5) 3.93 ± 0.51 (27) 3.52 ± 0.35 (64) 0.197 5.36 ± 0.66 (11) 3.29 ± 0.45 (28) 2.97 ± 0.24 (67) 0.029 3.00 ± 1.00 (2) 3.25 ± 1.03 (4) 3.25 ± 0.41 (16) 0.882 3.33 ± 0.67 (3) 3.50 ± 0.50 (16) 3.21 ± 0.47 (29) 0.420 ADH2 1/1 1/2 2/2 P for trends

Smokers Mean ± S.E. (n) Mean ± S.E. (n)

Overall Mean ± S.E. (n)

4.00 (1) 3.58 ± 0.60 (12) 3.15 ± 0.94 (13) 0.220

4.94 ± 0.52 (16) 3.60 ± 0.34 (55) 3.24 ± 0.21 (131) 0.013

2.50 ± 0.85 (6) 3.89 ± 0.48 (9) 3.86 ± 0.40 (21) 0.204

Non-smokersa Mean ± S.E. (n) Overall Mean ± S.E. (n) Smokers Mean ± S.E. (n) Mean ± S.E. (n)

Overall Mean ± S.E. (n)

Heavy drinkers

Non-smokersa Moderate drinkers

Non-smokersa Never drinkers

Table 4 Mean MN frequencies for ADH1B and ALDH2 genotypes according to alcohol consumption and smoking status

2.50 ± 1.19 (4) 3.67 ± 0.62 (6) 3.93 ± 0.45 (15) 0.199

H. Ishikawa et al. / Mutation Research 615 (2007) 134–142 Smokers Mean ± S.E. (n)

138

The number of subjects with MN frequency levels more than median value of MN (3.0) was significantly higher in the subjects with the ADH1B*1 allele as adjusted estimates (OR 2.08, 95% C.I. 1.24–3.48), when the OR for the subjects with homozygotes of the ADH1B*2/*2 genotype was defined as 1.00 (Table 7). Furthermore, the number of subjects with MN frequency levels more than median value of MN was significantly higher in the subjects with the ALDH2*2 allele as adjusted estimates (OR 1.79, 95% C.I. 1.04–3.11), when the OR for the subjects with homozygotes of the ALDH2*1/*1 genotype was defined as 1.00. Adjusted OR was significantly higher both in elderly subjects (OR 2.14, 95% C.I. 1.29–3.54) and in habitual alcohol drinkers (OR 2.30, 95% C.I. 1.31–4.04) (Table 7). 4. Discussion To our knowledge, the current study is the first to report a significant effect of the ADH1B and ALDH2 polymorphisms on the genotoxicity for alcohol drinking in humans. We have shown that the ADH1B*1 or

H. Ishikawa et al. / Mutation Research 615 (2007) 134–142

139

Table 5 Combined effects of ADH1B and ALDH2 genotypes on mean MN frequencies stratified by drinking frequency and smoking status Never drinkers ADH1B 1/1 or 1/2 n Overall ALDH2 1/1 0 1/2 or 2/2 3 P valuesa Non-smokersd ALDH2 1/1 0 1/2 or 2/2 2 P valuesa Smokers ALDH2 1/1 0 1/2 or 2/2 1 P valuesa

Non-habitual drinkers ADH1B 2/2

Mean ± S.E. n

ADH1B 1/1 or 1/2

Habitual drinkers ADH1B 2/2

ADH1B 1/1 or 1/2

ADH1B 2/2

Mean ± S.E. n

Mean ± S.E. n

Mean ± S.E. n

Mean ± S.E.

n

Mean ± S.E.

3.33 ± 0.67

9 36

3.22 ± 0.68 3.33 ± 0.40 0.992

6 0

4.33 ± 0.92

35 57

2.46 ± 0.37 3.49 ± 0.35 0.076

12 4

3.67 ± 0.70 6.00 ± 0.71

89 35

3.42 ± 0.24 4.26 ± 0.37 0.052

3.00 ± 1.00

2 18

4.00 ± 1.00 3.17 ± 0.41 0.791

5 0

4.80 ± 0.97

23 24

1.83 ± 0.27b 3.17 ± 0.47 0.004

6 2

4.50 ± 1.12 6.50 ± 1.50

40 17

3.25 ± 0.32c 4.59 ± 0.44 0.028

7 18

3.00 ± 0.85 3.50 ± 0.71 0.903

1 0

2.00

4.00

12 33

3.67 ± 0.87 3.73 ± 0.49 0.841

6 2

2.83 ± 0.79 5.50 ± 0.50

49 18

3.55 ± 0.34 3.94 ± 0.60 0.526

a P values were calculated by one-way ANOVA among four subgroups for the combinatorial genotypes ADH1B/ALDH2 in overall, non-smoking and smoking groups. b P < 0.05 vs. other three subgroups. c P < 0.05 vs. ALDH2 1/2 or 2/2 subgroups. d Non-smokers include never- and former-smokers.

Table 6 Combined effects of ADH1B and ALDH2 genotypes on mean MN frequencies stratified by alcohol consumption and smoking status Never drinkers ADH1B 1/1 or 1/2 n Overall ALDH2 1/1 0 1/2 or 2/2 3 P valuesa Non-smokersc ALDH2 1/1 0 1/2 or 2/2 2 P valuesa Smokers ALDH2 1/1 0 1/2 or 2/2 1 P valuesa

Moderate drinkers ADH1B 2/2

ADH1B 1/1 or 1/2

Heavy drinkers ADH1B 2/2

ADH1B 1/1 or 1/2

ADH1B 2/2

Mean ± S.E.

n

Mean ± S.E. n

Mean ± S.E.

n

Mean ± S.E. n

Mean ± S.E.

n

Mean ± S.E.

3.33 ± 0.67

9 36

3.22 ± 0.68 3.33 ± 0.40 0.992

13 3

4.69 ± 0.59 6.00 ± 1.00

99 87

2.96 ± 0.23b 3.78 ± 0.27 0.007

5 1

1.80 ± 0.58 6.00

25 5

3.88 ± 0.36 3.80 ± 0.58 0.053

3.00 ± 1.00

2 18

4.00 ± 1.00 3.17 ± 0.41 0.791

9 2

5.11 ± 0.75 6.50 ± 1.50

55 40

2.58 ± 0.25b 3.73 ± 0.35 0.0003

2 0

4.69 ± 0.59

8 1

3.75 ± 0.70 5.00 0.588

4.00

7 18

3.00 ± 0.85 3.50 ± 0.71 0.903

4 1

3.75 ± 0.85 5.00

44 47

3.43 ± 0.42 3.83 ± 0.41 0.868

3 1

1.33 ± 0.33 6.00

17 4

3.94 ± 0.42 3.50 ± 0.65 0.055

a P values were calculated by one-way ANOVA among four subgroups for the combinatorial genotypes ADH1B/ALDH2 in overall, non-smoking and smoking groups. b P < 0.05 vs. other three subgroups. c Non-smokers include never- and former-smokers.

140

H. Ishikawa et al. / Mutation Research 615 (2007) 134–142

Table 7 Multivariate logistic regression analysis of risk factors for MNa Variables

Adjusted OR

95% C.I.

P value

Age (>43)b Smoking (no)c Alcohol drinking (yes)d ADH1B*1e ALDH2*2f

2.14 0.89 2.30 2.08 1.79

1.29–3.54 0.54–1.46 1.31–4.04 1.24–3.48 1.04–3.11

0.003 0.646 0.004 0.006 0.037

a MN per 1000 binucleated cells: ≥3.0 vs. <3.0 (3.0: median level of all subjects). b Age: vs. ≤43 (43: median level of all subjects). c Smoking: vs. yes. d Alcohol drinking habit: vs. no. e ADH1B: vs. *2/*2. f ALDH2: vs. *1/*1.

ALDH2*2 allele had an effect on elevated MN frequency in habitual and moderate drinkers. Subjects with the significantly lowest mean MN frequency in the habitual and moderate drinkers without smoking habit had a combined genotype of ALDH2*1/*1 with ADH1B*2/*2. However, this relationship disappeared in the habitual and heavy drinkers with a smoking habit. Heavy alcohol intake and smoking are highly correlated. In a large Japanese cohort study, smoking was shown to be strongly associated with heavy alcohol intake [36]. Alcohol acts as a solvent for tobacco carcinogens. Furthermore, it has been shown that smoking changes the oral bacterial flora rapidly from Gram-negative to Gram-positive bacteria, which leads to acetaldehyde concentrations 50–60% higher compared to those observed without smoking [37]. Further clarification on a synergistic effect of alcohol drinking and smoking on acetaldehyde concentrations is needed. Although the combined effects of ADH1B and ALDH2 polymorphisms with alcohol drinking behavior on genotoxicity has so far not been proposed by other researchers, it may be reasonable to explain this as follows: the persons that carry the ADH1B*1 and ALDH2*2 alleles might metabolize ethanol more slowly into acetaldehyde, and also metabolize acetaldehyde more slowly into acetate; consequently, they have a tendency to retain both ethanol and acetaldehyde. In ADH1 and ALDH2 null mice by gene targeting knockout, Adh1−/− mice showed a severe defect in clearing ethanol and its damage [38], whereas Aldh2−/− mice having a lower clearance in acetaldehyde showed its toxicity [39,40]. In vitro experiments have also shown that ethanol as well as acetaldehyde caused DNA strand breaks in colonic mucosa cells [41] and in human derived liver cells [42]. So, there exists a possibility that ethanol- or/and acetaldehyde-induced DNA damages

may produce higher MN frequency in alcohol drinkers carrying ADH1B*1 and ALDH2*2 alleles, as indicated in the present study. Furthermore, multivariate analysis in the present study indicates that ethanol may have a more severe genotoxicity than acetaldehyde, because the subjects with ADH1B*1 allele had a higher OR for high MN frequency than those with the ALDH2*2 allele. Mutations in genes that encode for important mitotic proteins can shape the process defined as genome instability, creating an environment where cancer can develop [43]. The genome instability induced by ethanol- and acetaldehyde-mediated pathways in the present study might be able to explain ADH1B/ALDH2 polymorphic effects on alcohol-induced carcinogenesis. Recently, Matsuo et al. [28] have shown that a combination of ADH1B*1 and ALDH2*2 carrier revealed the highest risk of colorectal cancer in moderate consumers of alcohol. The genotypes of the subjects who showed the highest risk of colorectal cancer indicated by Matsuo et al. [28] is compatible with genotypes combinations possessing the highest MN frequency in the present study. On the other hand, subjects with the ADH1B*2/*2 and ALDH2*1/*1 genotype can promptly metabolize ethanol into acetaldehyde, and subsequently eliminates acetaldehyde, leading to lower MN frequencies. Thereby the subjects with aforementioned genotypes are protected against genotoxicity derived from ethanol and acetaldehyde. Other epidemiological studies have also shown that the combined genotypes of ADH1B*1/*1 with ALDH2*1/*2 revealed the highest risk of oropharyngolaryngeal [27] and esophageal [22] cancers among drinkers. As well as the above-mentioned colorectal cancer [28], it is reasonable to consider that drinkers possessing the combined genotypes of ADH1B*1/*1 with ALDH2*1/*2 might be exposed to increased ethanol and acetaldehyde, which might linger in blood and tissue of the body. In fact, it has been indicated that elimination rates for blood ethanol among ADH1B*1/*1 genotype carriers [17] and for blood acetaldehyde among ALDH2*1/*2 genotype carriers [21,44] are lower compared to their counterpart genotypes. Another possibility is that less active forms of ADH may increase alcohol metabolism through non-ADH pathways such as the CYP2E1, catalase and non-oxidative pathways. These pathways may be linked to accelerated DNA damage through enhanced formation of free radicals and fatty acid ethyl esters [45,46]. In fact, our previous study showed the CYP2E1 polymorphism as well as ALDH2 in a habitual drinker results in the increased genotoxicity [29].

H. Ishikawa et al. / Mutation Research 615 (2007) 134–142

Our study has several potential limitations. Significant linkage disequilibrium has been detected between the ADH1B and ADH1C gene polymorphisms in Japanese [22] and the effect of ADH1C polymorphism remains unclear. The sample size in the present study, which was insufficient for performing subgroups analyses among combined genotypes of ADH1B and ALDH2 and smoking and alcohol drinking habits, is limited and our findings need to be validated in larger populations. The MN assay is a genotoxicity assay that provides simultaneous information on a variety of chromosomal damage endpoints that reflect not only chromosomal breakage (MN) but also chromosome rearrangements indicated by nucleoplasmic bridges (NPBs) and gene amplification indicated by nuclear buds (NBUDs) [47]. EI-Zein et al. [48] provided the evidence that MN, NPBs and NBUDs are associated with lung cancer risk in smokers and the association is stronger for NPBs and NBUDs. Bonassi et al. [49] also provided the evidence that MN frequency in lymphocytes is a predictive biomarker of cancer risk within a population of healthy subjects. Therefore, the MN assay including NPBs and NBUDs should be used in the planning and validation of cancer surveillance and prevention programs in future. In conclusion, the results of this study have identified important novel associations between ADH1B/ALDH2 polymorphisms and micronuclei, an established biomarker of genome instability, via metabolisms of ethanol and acetaldehyde in alcohol drinkers. These results have also suggested that carcinogenic, mutagenic and genotoxic effects of ethanol per se on humans needs to be further tested allowing for ADH1B/ALDH2 polymorphisms.

[5] [6]

[7]

[8] [9] [10]

[11] [12]

[13]

[14]

[15]

[16]

[17]

Acknowledgements The study was supported by a Grant-in Aid for Scientific Research from the Ministry of Education, Science, Sports, Culture and Technology of Japan. The authors are grateful to Mr. Y. Miyatsu and Mr. K. Kurihara for counting micronuclei.

[18]

[19] [20]

References [21] [1] IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, no. 44, Alcohol Drinking, IARC, Lyon, 1988. [2] G. P¨oscl, H.K. Seitz, Alcohol and cancer, Alcohol Alcohol. 39 (2004) 155–165. [3] N.P. Singh, A. Khan, Acetaldehyde: genotoxicity and cytotoxicity in human lymphocytes, Mut. Res. 337 (1995) 9–17. [4] G. Obe, R. Jonas, S. Schmidt, Metabolism of ethanol in vitro produces a compound which induces sister-chromatid exchanges in

[22]

141

human peripheral lymphocytes in vitro: acetaldehyde not ethanol is mutagenic, Mut. Res. 174 (1986) 47–51. V.L. Dellarco, A mutagenicity assessment of acetaldehyde, Mut. Res. 195 (1988) 1–20. A. Helander, K. Lindahl-Kiessling, Increased frequency of acetaldehyde-induced sister-chromatid exchanges in human lymphocytes treated with an aldehyde dehydrogenase inhibitor, Mut. Res. 264 (1991) 103–107. V.J. Feron, H.P. Til, F. de Vrijer, R.A. Woutersen, F.R. Cassee, P.J. van Bladeren, Aldehydes: occurrence, carcinogenic potential, mechanism of action and risk assessment, Mut. Res. 259 (1991) 363–385. A.S. Ketcham, H. Wexler, N. Mantel, Affects of alcohol in mouse neoplasia, Cancer Res. 23 (1963) 667–670. B.J. Phillips, P. Jenkinson, Is ethanol genotoxic? A review of the published data, Mutagenesis 16 (2001) 91–101. L.M. Anderson, V.L. Souliotis, S.K. Chhabra, T.J. Moskal, S.D. Harbaugh, S.A. Kyrtopoulos, N-nitrosodimethylamine-derived O(6)-methylguanine in DNA of monkey gastrointestinal and urogenital organs and enhancement by ethanol, Int. J. Cancer 66 (1996) 130–134. V.A. Ramchandani, W.F. Bosron, T.K. Li, Research advances in ethanol metabolism, Pathol. Biol. 49 (2001) 676–682. J.B. Whitfield, ADH and ALDH genotypes in relation to alcohol metabolic rate and sensitivity, Alcohol Alcohol. (suppl. 2) (1994) 59–65. P. Brennan, S. Lewis, M. Hashibe, D.A. Bell, P. Boffetta, C. Bouchardy, et al., Pooled analysis of alcohol dehydrogenase genotypes and head and neck cancer: a HuGE review, Am. J. Epidemiol. 159 (2004) 1–16. W.F. Bosron, T.-K. Li, Genetic polymorphism of human liver alcohol and aldehyde dehydrogenases, and their relationship to alcohol metabolism and alcoholism, Hepatology 6 (1986) 502–510. A. Yoshida, L.C. Hsu, M. Yasunami, Genetic of human alcoholmetabolizing enzymes, Prog. Nucl. Acid Res. Mol. Biol. 40 (1991) 255–287. W.F. Bosron, D.W. Crabb, T.K. Li, Relationship kinetics of liver alcohol dehydrogenase and alcohol metabolism, Pharmacol. Biochem. Behav. 18 (Suppl. 1) (1983) 223–227. Y.D. Neumark, Y. Friedlander, R. Durst, E. Leitersdorf, D. Jaffe, V.A. Ramchandani, et al., Alcohol dehydrogenase polymorphisms influence alcohol-elimination rates in a male Jewish population, Alcohol. Clin. Exp. Res. 28 (2004) 10–14. S. Harada, D.P. Agarwal, H.W. Goedde, Aldehyde dehydrogenase and alcohol metabolism in alcoholics, Alcohol 2 (1985) 391– 392. S. Harada, Polymorphism of aldehyde dehydrogenase and its application to alcoholism, Electrophoresis 10 (1983) 652–655. H.W. Goeddle, D.P. Agarwal, G. Fritze, D. Meier-Tackmann, S. Singh, G. Beckmann, et al., Distribution of ADH2 and ALDH2 genotypes in different populations, Hum. Genet. 88 (1992) 344–346. K. Yamamoto, Y. Ueno, Y. Mizoi, Y. Tatsuno, Genetic polymorphism of alcohol and aldehyde dehydrogenase and the effects on alcohol metabolism, Jpn. J. Alcohol Drug Depend. 28 (1993) 13–25. A. Yokoyama, H. Kato, T. Yokoyama, T. Tsujinaka, M. Muto, T. Omori, et al., Genetic polymorphisms of alcohol and aldehyde dehydrogenases and glutathione S-transferase M1 and drinking, smoking, and diet in Japanese men with esophageal squamous cell carcinoma, Carcinogenesis 23 (2002) 1851–1859.

142

H. Ishikawa et al. / Mutation Research 615 (2007) 134–142

[23] H. Hori, T. Kawano, M. Endo, Y. Yuasa, Genetic polymorphisms of tobacco- and alcohol-related metabolizing enzymes and human esophageal squamous cell carcinoma susceptibility, J. Clin. Gastroenterol. 25 (1997) 568–575. [24] Y.C. Chao, L.S. Wang, T.Y. Hsieh, C.W. Chu, F.Y. Chang, H.C. Chu, Chinese alcoholic patients with esophageal cancer are genetically different from alcoholics with acute pancreatitis and liver cirrhosis, Am. J. Gastroenterol. 95 (2000) 2958–2964. [25] K. Matsuo, N. Hamajima, M. Shinoda, S. Hatooka, M. Inoue, T. Takezaki, et al., Gene-environment interaction between an aldehyde dehydrogenase-2 (ALDH2) polymorphism and alcohol consumption for the risk of esophageal cancer, Carcinogenesis 22 (2001) 913–916. [26] C.-F. Wu, D.-C. Wu, H.-K. Hsu, E.-L. Kao, J.-M. Lee, C.-C. Lin, et al., Relationship between genetic polymorphisms of alcohol and aldehyde dehydrogenases and esophageal squamous cell carcinoma risk in males, World J. Gastroenterol. 11 (2005) 5103– 5108. [27] A. Yokoyama, T. Muramatsu, T. Omori, T. Yokoyama, S. Matsushita, S. Higuchi, et al., Alcohol and aldehyde dehydrogenase gene polymorphisms and oropharyngolaryngeal, esophageal and stomach cancers in Japanese alcoholics, Carcinogenesis 22 (2001) 433–439. [28] K. Matsuo, K. Wakai, K. Hirose, H. Ito, T. Saito, T. Suzuki, et al., A gene-gene interaction between ALDH2 Glu487Lys and ADH2 His47Arg polymorphisms regarding the risk of colorectal cancer in Japan, Carcinogenesis 27 (2005) 1018–1023. [29] H. Ishikawa, Y. Miyatsu, K. Kurihara, K. Yokoyama, Geneenvironmental interactions between alcohol-drinking behavior and ALDH2 and CYP2E1 polymorphisms and their impact on micronuclei frequency in human lymphocytes, Mut. Res. 594 (2006) 1–9. [30] W.K. Lelbach, Cirrhosis in the alcoholics and its relation to the volume of alcohol abuse, Ann. NY Acad. Sci. 252 (1979) 85– 105. [31] A. Takada, Epidemiology: the relationship of alcoholic liver damage with the volume of alcohol intale, in: A. Takada, C. Hirayama (Eds.), Liver and Alcohol, first ed., Chu gai Igaku Sha, Tokyo, 1986. [32] M. Fenech, A.A. Morley, Measurement of micronuclei in human lymphocytes, Mut. Res. 148 (1985) 29–36. [33] M. Fenech, The cytokinesis-block micronucleus technique and its application to genotoxicity studies in human populations, Environ. Health Perspect. 101 (suppl. 3) (1993) 101–107. [34] K. Suzuki, A. Uchida, Y. Mizoi, T. Fukunaga, A study on ADH2 and ALDH2 genotyping by PCR-RFLP and SSCP analyses with description of allele and genotype frequencies in Japanese, Finn, and Lapp populations, Alcohol Alcohol. 29 (1994) 21–27. [35] S. Harada, S. Zhag, New strategy for detection of ALDH2 mutant, Alcohol Alcohol. Suppl. 28 (1993) 11–13.

[36] M. Inoue, S. Tsugane, Impact of alcohol drinking on total cancer risk: data from a large-scale population-based cohort study in Japan, Br. J. Cancer 92 (2005) 182–187. [37] M.P. Salaspuro, Acetaldehyde, microbes, and cancer of the digestive tract, Crit. Rev. Clin. Lab. Sci. 40 (2003) 183–208. [38] L. Deltour, M.H. Foglio, G. Duester, Metabolic deficiencies in alcohol dehydrogenase adh1, adh3, and adh4 null mutant mice, J. Biol. Chem. 274 (1999) 16796–16801. [39] T. Isse, T. Oyama, K. Kitagawa, K. Matsuno, A. Matsumoto, A. Yoshida, et al., Diminished alcohol preference in transgenic mice lacking aldehyde dehydrogenase activity, Pharmacogenetics 12 (2002) 621–626. [40] T. Isse, T. Oyama, K. Matsuno, M. Ogawa, R. Narai-Suzuki, T. Yamaguchi, et al., Paired acute inhalation test reveals that acetaldehyde toxicity is higher in aldehyde dehydrogenase 2knockout mice than wild-type mice, J. Toxicol. Sci. 30 (2005) 329–337. [41] J. Blasiak, A. Trzeciak, E. Malecka-Panas, J. Drzewoski, M. Wojew´odzka, In vitro genotoxicity of ethanol and acetaldehyde in human lymphocytes and the gastrointestinal tract mucosa cells, Toxicol. In Vitro 14 (2000) 287–295. [42] B.J. Majer, V. Mersch-Sundermann, F. Darroudi, B. Laky, K. de Wit, S. Knasm¨uller, Genotoxic effects of dietary and lifestyle related carcinogens in human derived hepatoma (HepG2, Hep3B) cells, Mut. Res. 551 (2004) 153–166. [43] C. Leugauer, K.W. Kinzler, B. Vogelstein, Genetic instabilies in human cancers, Nature 396 (1998) 643–649. [44] E. Yoshihara, K. Ameno, K. Nakamura, M. Ameno, S. Itoh, I. Ijiri, et al., The effects of the ALDH2*1/2, CYP2E1 C1/C2 and C/D genotypes on blood ethanol elimination, Drug. Chem. Toxicol. 23 (2000) 371–379. [45] C.S. Lieber, Cytochrome P-4502E1: its physiological and pathological role, Physiol. Rev. 77 (1997) 517–544. [46] C.S. Lieber, Ethanol metabolism, cirrhosis and alcoholism, Clin. Chim. Acta 257 (1997) 59–84. [47] M. Fenech, Cytokinesis-block micronucleus assay evolves into a “cytome” assay of chromosomal instability, mitotic dysfunction and cell death, Mutat. Res. 600 (2006) 58–66. [48] R.A. EI-Zein, M.B. Schabath, C.J. Etzel, M.S. Lopez, J.D. Franklin, M.R. Spitz, Cytokinesis-blocked micronucleus assay as a novel biomarker for lung cancer, Cancer Res. 66 (2006) 6449–6456. [49] S. Bonassi, A. Znaor, M. Ceppi, C. Lando, W.P. Chang, N. Holland, M. Kirsch-Volders, E. Zeiger, S. Ban, R. Barale, M.P. Bigatti, C. Bolognesi, A. Cebulska-Wasilewska, E. Fabianova, A. Fucic, L. Hagmar, G. Joksic, A. Martelli, L. Migliore, E. Mirkova, M.R. Scarfi, A. Zijno, H. Norppa, M. Fenech, An increased micronucleus frequency in peripheral blood lymphocytes predicts the risk of cancer in humans, Carcinogenesis (2006), Advance Access published on September 14, 2006 doi:10.1093/carcin/bgl177.