Document 5LV1bNN4p8rJ0V9g4BYyg2VYD

Arch Toxicol (19W^ 71: 481-488 Ml I \ lit11 K \( I I \ \ I I < i \ INAC I IVA I ION Spnnger-Verlag 1997 Chung-Ying Huang - Kuo-Liang Huang Tsun-Jen Cheng Jung-Der Wang Ling-Ling Hsieh The GST T1 and CYP2E1 genotypes are possible factors causing vinyl chloride induced abnormal liver function '7 f-O/- 9 Received 19 November 1996 / Accepted: 17 March 1997 Abstract Vinyl chloride monomer (VCM) is hepatotoxic as well as carcinogenic in humans. There are re ports that exposure to VCM seems to induce abnormal liver function, liver fibrosis, cirrhosis, portal hyperten sion, and angiosarcoma of the liver. In vivo, VCM is metabolized by cytochrome P450 2E1 (CYP2E1) to form the electrophilic metabolites, chloroethylene oxide (CEO) and chloroacetaldehyde (CAA), which may ei ther cause cell damage or be further metabolized and detoxified by glutathione S-transferases (GSTs). This study investigated whether or not the genotypes CYP2E1, glutathione S-transferase 0 (GST Tl) and p. (GST Ml) correlated with abnormal liver function found in vinyl chloride exposed workers. For this study, 251 workers from five polyvinyl chloride plants were enrolled. The workers were classified into two exposure groups (high and low) and the degree of exposure was determined based on their job titles and airborne VCM concentration. The activity of serum alanine amino transferase (ALT) was used as the parameter of liver function. The genotypes CYP2E1, GST Tl and GST Ml were determined by polymerase chain reaction and re striction fragment length polymorphism on peripheral white blood cell DNA. Other potential risk factors were also ascertained and the confounding effect was adjusted accordingly. Stratified analyses were used to explore the correlation between the alteration of liver function and the genotypes CYP2E1, GST Tl and GST Ml among the workers exposed to different levels of VCM. The following results were obtained (1) at low VCM expo sure, the odds ratio (OR) of positive GST Tl on ab- C -Y Huung K-.-L. Huang T.-J. Cheng J -D. Wang Institute of Occupational Medicine and Industrial Hygiene, College of Public Health. National Taiwan University, Taipei. Taiwan L.-L. Hsieh m) Department of Public Health. Chang Gung College of Medicine and Technology, 259 Wen-Hwa 1 Road. Kwei San. Tao-Yuan 5332. Taiwan normal ALT was 3.8 (95% Cl 1.2-14.5) but the CYP2E1 genotype was not associated with abnormal ALT. (2) At high VCM exposure, a c2c2 CYP2E1 genotype was as sociated with increased OR on abnormal ALT (OR 5.4, 95% Cl 0.7-35.1) and positive GST Tl was significantly associated with decreased OR on abnormal ALT (OR 0.3, 95% Cl 0.1-0.9). (3) Multiple linear and logistic regression also showed strong interactions of the VCM exposure to CYP2E1 as well as to the GST Tl genotype. These observations suggest that the two genotypes, CYP2E1 and GST Tl, may play important roles in the biotransformation of VCM, the effect of which leads to liver damage. Key words Vinyl chloride Glutathione S-transferases Cytochrome P-450 2E1 Liver function Introduction Vinyl chloride monomer (VCM) has been associated with hepatotoxic effects in humans including liver fi brosis, cirrhosis, hepatomegaly, splenomegaly, portal hypertension, and angiosarcoma of the liver (Creech and Johnson 1974; Lilis et al. 1975; Smith and Crossley 1976; Simonato et al. 1991; Wang et al. 1991). Exposure to VCM may also cause abnormalities m one or more liver function tests, such as aspartate aminotransferase (AST), alanine aminotransferase (ALT) and gamma glutamyl transpeptidase (GGT; Makk et al. 1974; Lilis et al. 1975; Marsteller and Lelbach 1975; Ho et al. 1991; Du et al. 1995). The liver is the primary target of VCM because it is believed that VCM is activated by cytochrome P450 2E1 (CYP2E1), an enzyme mainly existing in the liver (Guengerich et al. 1991; Koop 1992). CYP2E1 catalyses VCM oxidation into an epoxide intermediate, chloro ethylene oxide (CEO), which is spontaneously rear ranged into 2-chloroacetaldehyde (CAA). The electro philic metabolites, CEO and CAA. considered to be the most important intermediates in the VCM carcinogenic ASI 000014021 process react with DNA bases to form adducts. The DNA-adducts are mutagenic in the bacterial system as well as in mammalian cells or covalently bind to proteins to impair cell function (Bolt 1986; Brandt-Rauf et al. 1995) . These two metabolites may be further processed by glutathione S-transferases (GSTs; Bolt 1986; Easter and Von Burg 1994). The activities of these enzymes, GSTs and CYP2E1, are known to vary genetically among individuals. Thus, inter-individual differences in susceptibility to VCM exposure may be attributable to the differences in the activities of CYP2E1 and GSTs. Metabolism of VCM may proceed via different path ways at different concentrations (Easter and Von Burg 1994). The pathway described above mainly proceeds at high concentraion. At low concentration, VCM appears to be oxidized by alcohol dehydrogenase to form chloroethanol, CAA and chloroacetic acid. The latter two products may further conjugate with glutathione (GSH; Easter and Von Burg 1994). Therefore, the sus ceptibility of an individual to the induction of liver damage may depend on the exposure concentration of VCM. In other words, individuals with the same genetic trait exposed to different levels of VCM may exhibit different health effects. The genetic variation in CYP2E1 is supported by recent identification of the restriction fragment length polymorphisms (RFLPs) of the GYP2E1 5'-flanking region detected by Pstl and Rsal digestion. The two restriction sites, PsrI and Rsal, are tightly linked and such genotype may affect the transcriptional regulation of CYP2E1 (Hayashi et al. 1991; Watanabe et al. 1994). The relationship has been reported between CYP2E1 genetic polymorphisms and susceptibility to lung cancer (Perrson et al. 1993), alcoholic liver disease (Maezawa et al. 1994; Tsutsumi et al. 1994; Chao et al. 1995), he patocellular carcinoma (Yu et al. 1995) and nasopha ryngeal carcinoma (Hildesheim et al. 1995) respectively. Concerning GSTs, the mammalian GST supergene family, composed of a, n, p, and newly discovered 9, catalyses the conjugation of GSH with electrophilic substrates. Some individuals carry a large deletion (null genotype) in the glutathione S-transferase 9 (GST Tl) and glutathione S-transferase |i (GST Ml) gene, res pectively, An individual without an intact GST Ml may have a higher risk of developing cancer or cytogenetic damage (Wiencke et al. 1990; Seidegard et al. 1986, 1990). The role of the newly discovered GST Tl remains controversial. An in vitro study shows that GST Tl deficiency may increase chromosome damage by ethyl ene oxide (Hallier et al. 1993) or by diepoxybutane (Norppa et al. 1995). On the other hand, there are some classes of agents (e.g. halomethanes) that are activated by conjugating with GSH to form mutagenic interme diates (Thier et al. 1993). Thus, different substrates metabolized by GST Tl may have different conse quences (Hallier et al. 1993; Norppa et al. 1995; Thier et al. 1993; Nelson et al. 1995). Several studies performed in the 1970s showed a higher prevalence of abnormal liver function in VCM exposed workers (Lilis et al. 1975; Makk et al. 1974; Marsteller and Lelbach 1975). However, other studies indicate that the liver function tests used are not sensi tive enough to detect early VCM-induced liver damage (Williams et al, 1976; Sugita et al. 1986). Although AST, ALT and GGT have been proposed as indicators of liver damage, their specificity and sensitivity remain un known, Ho et al, (1991) report that ALT is the earliest and most sensitive indicator of VCM-induced liver dysfunction followed by GGT. It is believed that ALT is more specific than AST as an indicator of liver damage since ALT is found primarily in the liver, whereas AST is also present in many other tissues including heart, skeletal muscle, kidney and brain (Podolsky and Isselbacher 1994). Because ALT is more sensitive and specific than AST or GGT, and more suitable for the screening of occupational toxic exposure (Redlich and Brodkin 1994; Widmann 1989), ALT was used as an indicator of liver damage in this study. There may be interactions between the susceptible factors and chemical exposure. Hence, the magnitude and direction of associations between exposures and diseases may differ according to the extent of genetic susceptibilities (Vine and Mcfarland 1990). On the other hand, the effect of genetic susceptibilities may depend on the exposure concentrations when the metabolism of the chemical proceeds via different pathways at different concentrations. In this study, we attempted to determine whether or not there were interactions between the ge notypes of CYP2E1, GST Tl, and GST Ml and the VCM exposure in inducing liver damage. Materials and methods - Subjects Two hundred and eighty volunteers were recruited from five polyvinyl chloride (PVC) plants. Female workers (n = 13) were excluded from this study because of insufficient number. Those who were newly employed or changed workplace within 6 months (n = 13) were also excluded, on consideration that the time for the induction of liver damage from VCM exposure might not be suf ficient (Marsteller and Lelbach 1975). Some data were missing from three workers, who were likewise excluded. Thus, a total of 251 male workers was employed for the final analysis. All workers re ceived complete physical examinations and structured question naires including demographic data, smoking and drinking habits, and detailed occupational history. Questionnaires were adminis tered by well-trained interviewers. Subjects were divided into two exposure groups, high and low, according to the nature of their current job. The high-exposure group was defined as those individuals involved in unloading PVC or VCM, adding catalyst, tank cleaning, working in a control room, tank stripping, and being a foreman; the low-exposure group were those individuals involved in the drying operation, laboratory testing, managing industrial hygiene, machine maintenance, PVC storage, gatehouse keeping, administration and car driving. Ex posure assessment performed 1 year before this study showed that the 8 h nme-weighted average (TWA) of VCM in the low-exposure group was always <1 ppm, while the TWA of VCM in the highexposure group ranged from 1 to 5 ppm; occasionally concentra tions up to 40 ppm were found in certain jobs including unloading of PVC. adding of catalyst and tank cleaning. ASI 000014022 484 Samples Ten millilitres of venous blood were drawn into heparinized tubes (Vacutamer) and stored at 4 C. The whole blood was separated into plasma, buffy coat, and red blood cells by centrifugation within 18 h of obtaining the blood, then stored in a -70 C freezer. Genomic DNA was extracted and purified from buffy coat as previously described (Hsieh et al. 1996). r- <M Sr- CMO^-tO S Measurement of ALT ALT was analysed with a Hitachi 7050 autoanalyser. Enzyme im munoassay (Austria-II, Abbott Lab., Chicago, II., USA) was used to determine the presence or absence of the serum hepatitis B virus surface antigen (HBs Ag) and anti-hepatitis C virus antibody (anl'HCV Ab). All these tests were-performed in the central laboratory of the National Taiwan University Hospital; the cut-off value of 31 IU/1 adopted here for the abnormality of the ALT was that used by the central laboratory. Genotyping of the GST T1 and GST Ml gene A simple assay based on polymerase chain reaction (PCR) tech nology was available to determine the presence or absence of GST T1 (Pemble et al. 1994) and GST Ml (Comstock et al. 1990). Briefly, PCR was performed in a 25 pi mixture containing the buffer supplied by Promega (Madison, Wis., USA), 250 ng of ge nomic DNA. Taq DNA polymerase (1 U). four bases (dNTP), and 200 ng of each of the primers for GST T1 (5'-TTCCTTACTGGTCCTCACATCTC-3' and 5'-CACCGGATCATGGCCAGCA-3'), GST Ml (5'-TGCCCTACTTGATTGATGGG-3' and 5'-CTGGATTGTAGCAGATCATGC-3') and the internal con trol, |3-globin (5'-CACAACTGTGTTCACTAGC-3' and 5'-CAACTTCATCCACGTTCACC-3'). The running conditions were: 30 cycles at 95 C for 30 s, 61 C for 15 s, and 72 C for 1 min. The PCR product was electrophoresed in 6% polyacrylamide gel, stained with ethidium bromide and photographed under UV light. Individuals without an intact GST T1 or GST Ml gene showed no amplification of the 480 bp GST T1 fragment or the 273 bp GST Ml fragment and a positive internal control, the 100 bp (3-globin fragment (Fig. 1). 5 r- CM CO VlfKOf'- 120 bp Fig. 2 Representauve results of PCR-RFLP analysis of the CYP2E1 gene from PVC workers. Genomic DNAs were amplified using PCR, and the products digested with Pstl, electrophoresed in polyacryl amide gels, and stained with ethidium bromide. The 410 bp DNA fragment corresponds to the cl allele which does not have a /VI site, while the paired fragments of 290 and 120 bp. correspond to the c2 allele containing the Pstl site (RFLP Restriction fragment length polymorphism) Determination of PstI fragment polymorphism of the CYP2E1 gene To determine the CYP2E1 fragment polymorphism, genomic DNA was amplified in a 20 pi mixture containing 500 ng of genomic DNA, and 160 ng of each primer (5'-CAGTCGAGTCTACATTGTC-3' and 5'-TTCATTCTGTCTTCTAACTG-3') flanking the polymorphic PstI site, for 30 cycles (Hayashi et al. 1991) using the running conditions: 95 C for 30 s, 57 C for 15 s. and 72 C for 1 min. The PCR product was digested overnight with PstI (7.5 units) at 37 C, electrophoresed in 6% polyacrylamide gels, stained with ethidium bromide and photographed under UV light. The PCR product of CYP2E1 digested with restriction enzyme Pstl showed three genotypes of CYP2E1: type clcl was not digested by PstI and showed only one band of 410 bp; heterozygous type clc2 showed three bands of 410, 290 and 120 bp; and type c2c2 showed 2 bands of 290 and 120 bp by polyacrylamide gel electrophoresis (F`g. 2). Statistical analysis Fig. 1 Representative results of polymerase chain reaction (PCR) of the GST Ml and GST T1 gene from PVC workers. Genomic DNAs were amplified using PCR, and the products electrophoresed in polyacrylamide gels, and stained with ethidium bromide. The 480 and 273 bp DNA fragments correspond to GST T1 and GST MI, respectively, while the 100 bp DNA fragment corresponds to the pglobtn gene as internal control (PVC Polyvinyl chloride, GST glutathione S-transferase) Statistical Analysis System (SAS) version 6.10 and EGRET (Epi demiologic Graphic, Estimation, and Testing package) were used for the statistical analysis. x*-test was used to compare the differ ence of basic characteristics between the high- and low-VCM ex posure groups. To assess the effect of potential risk factors on liver function, HBs Ag, anti-HCV Ab, habitual alcohol drinking (ha bitual dnnkmg of at least once a week), age, employment duration, body mass index (BM1), VCM exposure and the genotypes GST Tl, GST Ml, and CYP2E1 were included in the univariate ana lyses. Stratified analyses were used to explore the correlation be tween the alteration of liver function and the genotypes, CYP2E1, GST Tl and GST Ml, among the workers exposed to different levels of VCM. Subsequently, a multiple logistic regression model was used to estimate the adjusted odds ratio (OR) of the above risk factors and the effect on abnormal liver function of interactions between the genotypes GST Tl, GST Ml, and CYP2E1 and al cohol drinking (i.e. habitual alcohol dnnkmg x GST Tl. GST Ml or CYP2E1) or VCM exposure, (i e. VCM exposure x GST Tl, GST M1 or CYP2E1). The stepwise and backward procedures were used to select the best model. Similar procedures were also used to select the best multiple linear regression model. AS! 000014023 Rest The in T have than P< BMI HBs feren num HBs the ir HBs was t distn the t frequ was - Table worke (HBs. anti-hf Age <40 >40 ; Emplo' <15 215 ; Bods n <25 >25 ' Dnnkir No Yes HBs At Nega Posit Anti-H' Nega Posit; Virus in Nega; Posit! GST Tl Null Positl GST M Null Posnr CYP2E! clcl clc2 c2c2 x; = ' ' Dnnkir h Virus i! Results Table 2 Odds ratio (OR, with 95% confidence interval) for ab normal ALT in PVC workers (ALT Alanine aminotransferase) The basic characteristics of 251 male workers are shown in Table 1. The high-VCM exposure group seemed to have a higher prevalence of habitual alcohol drinking than the low-VCM exposure group (x2 = 7.426, P < 0.01); otherwise all the other characteristics of age, BMI, employment duration, and prevalence of positive HBs Ag and anti-HCV Ab, were not statistically dif ferent between these two exposure groups. Because the number of positive anti-HCV Ab was small, the data on HBs Ag and anti-HCV Ab were combined and used as the indicator of hepatitis virus infection. When either the HBs Ag or anti-HCV Ab test was positive, the person was scored as positive for virus infection. The genotypic distribution of GST Tl, GST Ml and CYP2E1 among the two VCM exposure groups was similar. The total frequency of the null-genotype of GST Tl and GST Ml was 47.4 and 60.6%, respectively. The frequency of the Table 1 Basic characteristics of the polyvinyl chlonde (PVC) workers stratified by vinyl chloride monomer (VCM) exposure (HBsAg Serum hepatitis B virus surface antigen, anti-HCV Ab anti-hepatitis C virus antibody, GST glutathione S-transferase) VCM exposure Low (n = 111) n (%) High n (%) a" It| Total (n = 25!) n (%) Age <40 years >40 years Employment duration <15 years >15 years Body mass index <25 kg/irr >25 kg/m* Drinking3 No Yes HBs Ag Negative Positive Anti-HCV Ab Negative Positive Virus infection6 Negative Positive GST Tl genotype Null Positive GST Ml genotype Null Positive CYP2E1 genotype clcl clc2 c2c2 43 (38.7) 68 (61.3) 70 (50) 70 (50) 113 (45.0) 138 (55.0) 46 (41.4) 65 (58.6) 72 (51.4) 118 (47.0) 68 (48 6) 133 (53.0) 90 (81 1) 103 (73 6) 193 (76.9) 21 (18.9) 37 (26.4) 58 (23.1) 104 (93.7) 115 (82.1) 219 (87.3) 7 (6.3)* 25 (17.9)* 32 (12.7) 91 (82.0) 110 (78.6) 201 (80.1) 20 (18.0) 30 (21.4) 50 (19.9) 109 (98.2) 136 (97,1) 245 (97.6) 2(1.8) 4 (2.9) 6 (2.4) 89 (80.2) 108 (77.1) 197 (78.5) 22 (19.8) 32 (22.9) 54 (21.5) 58 (52.3) 53 (47.7) 61 (43.6) 119 (47.4) 79 (56.4) 132 (52.6) 70 (63 1) 41 (36.9) 82 (58.6) 152 (60 6) 58 (41.4) 99 (39.4) 61 (55.0) 44 (39.6) 6(5 4) 78 (55.7) 55 (39.3) 7 (5.0) 139 (55.4) 99 (39.4) 13(5.2) *X; = 7 426. P < 0 01 'Drinking, Habitual alcohol drinking > once a week 6 Virus infection positive, HBs Ag positive or anti-HCV Ab positive Normal n (%) Abnormal OR(95% Cl) n (%) VCM exposure Low 92 (43,2) High 121 (56.8) GST Tl genotype Null 101 (47.4) Positive 112 (52.6) GST Ml genotype Null 129 (60.6) Positive 84 (39.4) CYP2E1 genotype clcl/clc2 204 (95.8) c2c2 9 (4.2) Virus infection Negative 176 (82.6) Positive 37 (17 4) Drinking* Negative 187 (87.8) Positive 26 (12.2) Body mass index <25 kg/m 172 (80.8) >25 kg/m"1 41 (19.2) Employment duration <15 years 100 (46.9) >5 years 113 (53.1) 19 (50.0) 19 (50.0) 1 0.8 (0.4, 1.62) 18 (47.4) 20 (52.6) 1 1.0 (0.5, 2.1) 23 (60.5) -- 1 15 (39.5)' 1.0 (0.5, 2.1) ^* 34 (89.5)' I 4(10.5) 2.7 (0.6, 10.2)* 21 (55.3) 17 (44.7) 1 3.8 (1.7, 8.5)** 32 (84.2) 6 (15.8) 1 1.4 (0.4, 3.7) 21 (55.3) 17 (44.7) 1 3.4 (1.5, 7,4)** 18 (47,4) 20 (52.6) 1 1.0 (0.5, 2.1) *P = 0.12; ** P < 0.01; other P-value >0.1 * Drinking, Habitual alcohol drinking > once a week clcl, clc2 and c2c2 genotypes of CYP2E1 was 55,4, 39.4 and 5.2%, respectively. The results of univariate analysis are shown in Table 2. Virus infection [OR (odds ratio) 3.8, P < 0.01, 95% Cl 1.7-8.5], and BMI (OR 3.4, P < 0.01, 95% Cl 1.5-7.4) were significantly associated with abnormal ALT. Individuals with CYP2E1 c2c2 genotype had a higher frequency of abnormal ALT but without statis tical significance (OR 2.7, P - 0.12, 95% Cl 0.6-10.2). The factors VCM exposure, the genotypes of GST Tl and GST Ml, habitual alcohol drinking, and employ ment duration, were not associated with abnormal ALT in univariate analyses. Because the metabolism of VCM may differ at dif ferent concentrations, the prevalence of abnormal ALT associated with the genotypes of these metabolic en zymes may be affected by the levels of VCM exposure. In the low-VCM exposure group, the frequency of the GST Tl positive genotype was significantly higher in workers with abnormal ALT (OR 3.8, 95% Cl 1.2-14.5) but the CYP2E1 genotype was not associated with ab normal ALT. In the high-VCM exposure group, the frequency of the GST Tl positive genotype was signifi cantly lower in workers with abnormal ALT (OR 0.3, 95% Cl 0.1-0.9) and the c2c2 genotype of CYP2E1 was associated with increased OR on abnormal ALT (OR 5.4, 95% Cl 0.7-35.1). No association was found be tween the GST Ml genotype and abnormal ALT whether the VCM exposure was low or high. A signifi cant interaction between the GST Tl genotype and ASl 000014024 486 Table 3 OR (with 95% Cl) of abnormal ALT in relation to the GST Tl, GST Ml and CYP2E1 genotypes VCM exposure Low High Normal n Abnormal n OR (95% Cl) Normal n Abnormal n GST Tl genotype Null Positive GST Ml genotype Null Positive CYP2E1 genotype clcl/clc2 c2c2 53 5 1 39 14 3.8 (1.2, 14.5)** Maentel-Haenszel test for homogeneity, P < 0 001 48 73 58 12 1 34 7 1.0 (0.3, 3 1) Maentel-Haenszel test for homogeneity, P = 0.96 71 50 87 18 1 5 1 1.0 (0.0, 9.4) Maentel-Haenszel test for homogeneity, P - 0.19 117 4 13 6 11 8 16 3 *P = 0.053; **0.01 < P < 0.05; other /'-value > 0,1 OR (95% Cl) 1 0,3 (0.1, 0.9)** 1 ' 1.0(0 3,3.1) 1 5.4 (0.7, 35.1)* VCM exposure was indicated based on the analysis using a multiplicative model (P < 0.001, Maentel-Haenszel test for homogeneity). Although the OR for the c2c2 genotype was different in the low or high VCM exposure groups, a statistical test for the interaction between the CYP2E1 genotype and VCM exposure in dicated non-significance (P = 0.19, Maentel-Haenszel test for homogeneity; Table 3) The data of Table 4 give the best fitted multiple lo gistic and linear regression models. In the multiple lo gistic regression model, virus infection (OR 6.4, P < 0.01, 95% Cl 2.7-15.3), and BMI of >25 kg/m2 (OR 5.7, P < 0.01, 95% Cl 2.4-13.7) were significantly associated with increased OR on abnormal ALT. In teractions between VCM exposure and the genotypes of GST Tl and CYP2E1 could be seen. The effects of the GST Tl or CYP2E1 genotypes on inducing liver damage differed with different levels of VCM exposure; there fore, the odds ratio should be discussed separately on each category. Compared to individuals with low VCM exposure and null GST Tl, the OR for individuals with combined GST Tl positive and low VCM exposure on abnormal ALT was 7,1 times the null genotype at low VCM (OR 7.1, 95% Cl 2.1-24.1, P < 0.01), while the Table 4 Multiple linear and logistic regression model for ALT in PVC workers, n - 251 {BMI Body mass index) Virus infection Positive vs negative Drinking* Yes vs No BMI (kg/m3) >25 vs <25 VCM exposure High vs low GST Tl genotype Positive vs null Interactions between VCM exposure and GST Tl genotype6 CYP2E1 genotype c2c2 vs clcl/clc2 Interactions between VCM exposure and CYP2E1 genotype' Logistic regression OR (95% Cl) 6.4 (2.7. 15.3)* 5.7 (2.4, 13.7)* 2.7 (0,8. 9.0) 7 1 (2.1, 24.6)3 0.03 (0.004. 0,2)* 0.5 (0.04, 5.7) 17.2 (0.8, 350.0)** Linear regression 13 (95% Cl) 15 8 (11.2, 20.4)* 10.6 (4.8, 16.3)* 7.2 (2.7, 11.7)* 1.2 (-4.3, 6.8) 4.9 (-0 8. 10.6)* -10.7 (-18.3. -3.1)* -5.9 (-18.5, 6.7) 18.2 (1 1. 35.3)*** P = 0.09; **/' - 0.06; ***0.01 < P < 0,05; XP < 0.01: other P-value > 0.1 ' Drinking, Habitual alcohol drinking > once a week h Interactions between VCM exposure and GST Tl genotype were identified in the logistic regression (see the text) 1 Interactions between VCM exposure and CYP2E1 genotype were identified in the logistic regression (see the text) ASI 000014025 O hi 2. O w, (C 17 gr in sit Di In c2 Pr Cl ge to no fre rej Cl res be Stlwi ini str eff VC sot ex; ex; see the no wit no vie GS sh< GS tirr exf GS 7.1 low due VC tro ma et . me cer. OR for individuals with combined GST T1 positive and high VCM exposure differed by only 0.6 times (OR of 2.7 x 7.1 x 0.03 =e 0.6, 95% Cl 0.1-2.2, P > 0.1). The OR of the CYP2E1 c2c2 genotype on abnormal ALT was highly significant in the high VCM exposure group (OR of 2.7 x 0.5 x 17.2 ~ 23.3, P < 0.01, 95% Cl 3.1176.1), but was insignificant in the low VCM exposure group (OR 0.5, P > 0.1, 95% Cl 0.0-5.7). Similar interactions were also observed with the linear regres sion model (Table 4). Discussion In this study population, the frequency of the CYP2E1 c2 allele of 24.9%, is comparable to that shown in the previous study for Taiwanese (Stephens et al. 1994; Chao et al. 1995). The frequency of the GST Ml null genotype in our present study group (60.6%) is similar to that in the control group of a hepatocellular carci noma study in Taiwan (63.3%; Yu et al. 1995). The frequency of GST T1 in Taiwan, however, has not been reported. The frequency of the GST T1 null genotype in China is 64.4% (Nelson et al. 1995) differing from our result of 47.4%, Virus infection and BMI of >25 kg/m2 are shown to be the major risk factors for abnormal ALT in this study. Habitual alcohol drinking is weakly associated with abnormal ALT because alcohol may selectively inhibit ALT activity (Redlich and Brodkin 1994). In stratified analyses, the positive GST T1 shows a different effect on inducing abnormal ALT at different levels of VCM exposure, and CYP2E1 c2c2 is shown to be as sociated with abnormal ALT only in the high-VCM exposure group. Similar interactions between the VCM exposure and the genotypes of GST T1 and CYP2E1 are seen when analysed with the multiple logistic as well as the linear regression model. In the present study, the VCM exposure and the ge notypes of GST T1 and CYP2E1 are shown to interact with each other to affect ALT. Since these variables do not operate independently, the discussion of the indi vidual effect of VCM exposure and the genotypes of GST T1 and CYP2E1 on the induction of liver damage should be avoided. The OR for individuals with positive GST T1 atfftigh VCM exposure on abnormal ALT is 0.6 times the individuals with null GST TI at low VCM exposure. However, the OR for individuals with positive GST Tl at low VCM exposure on abnormal ALT was 7.1 times higher than individuals with null GST Tl at low VCM exposure. Thus, the effect of GST Tl on in ducing liver damage differed with different levels of VCM exposure. The role of newly discovered GST TI remains con troversial. Different substrates metabolized by GST Tl may have different effects (Hallier et al. 1993; Norppa et al 1995; Thier et al. 1993; Nelson et al. 1995). The metabolic pathway for VCM may depend on its con centration. At a low concentration, VCM may be pre dominantly metabolized by alcohol dehydrogenase to form 2-chloroethanol, CAA and 2-chloroacetic acid. At a high concentration, it may be oxidized to form CEO by cytochrome P-450s and/or spontaneously rearranged to form CAA (Easter and Von Burg 1994). The above alkylating metabolites including CAA and CEO may be further metabolized by the GSTs (Bolt 1986; Easter and Von Burg 1994). Thus, individuals with positive GST Tl exposed to different levels of VCM may have different effects on liver function due to different phase I metab olites formed. Furthermore, VCM may be directly me tabolized by GSTs to conjugate with GSH (Hefner et al. 1975) and these GSH-conjugates may have mutagenic activity, such as ethylene dibromide conjugated with GSH to form alkylating intermediates (Thier et al. 1993). The GST Tl positive genotype is significantly associated with increased OR on abnormal ALT only at low VCM exposure as observed in this study. Therefore, VCM may be mainly metabolized directly by GST Tl at low concentration to form more toxic intermediates. The transcriptional activity of the CYP2E1 c2c2 ge notype has been shown to be about 10 times the activity of the clcl genotype (Hayashi et al. 1991); however, the level of the final protein product has not been verified. VCM is primarily metabolized by CYP2E1 at high concentration, in partial explanation of why the signifi cant association of the CYP2E1 c2c2 genotype with in creased OR on abnormal ALT only at high VCM exposure. The above observation, therefore, indicates that GST Tl and CYP2E1 may play important roles in the bio transformation of VCM; hence, at low concentration, VCM is metabolized by GST Tl or alcohol dehydro genase and at high concentration, by CYP 2E1. Ac cordingly, the effect of these metabolic enzymes on the induction of abnormal ALT is different at different levels of VCM exposure. The association between abnormal ALT and the polymorphism of alcohol dehydrogenase among these PVC workers should be investigated fur ther to elucidate the roles of metabolic traits on the toxic effects of VCM. Acknowledgements We thank Dr Chung-Li Du for technical as sistance and Dr J.T. Ou for helpful discussions. This study was supported by NSC grant NSC86-2314-B002-332. Reference* Bolt HM (1986) Metabolic activation of vinyl chloride formation of nucleic acid adducts and relevance to carcinogenesis. IARC Sci Publ 70- 261-268 Brandt-Rauf PW. Marion ML DeVivo I (1995) Mutant p21 pro tein as a biomarker of chemical carcinogenesis in humans. In: Mendelsohn ML, Peetcrs JP, Normandy MJ (eds) Bio markers and occupational health - progress and perspectives, Joseph Henry, Washington DC. pp 163-173 Chao YC. Young TH, Chang WK, Tang HS, Hsu CT (1995) An investigation of whether polymorphisms of cytochrome P450 2E1 are genetic markers of susceptibility to alcoholic end-stage organ damage in a Chinese population. Hepatology 22: 1409-- 1414 AS I 000014026