Interaction of cytochrome P4501A1 genotypes with other risk factors and susceptibility to lung cancer

Available online at www.sciencedirect.com

Mutation Research 639 (2008) 1–10

Interaction of cytochrome P4501A1 genotypes with other risk factors and susceptibility to lung cancer Parag P. Shah a , Arvind P. Singh a , Madhu Singh a , Neeraj Mathur a , Mohan C. Pant b , Bhartendu. N. Mishra c , Devendra Parmar a,∗ a

Developmental Toxicology Division, Industrial Toxicology Research Centre, P.O. Box 80, M.G. Marg, Lucknow 226001, India b Department of Radiotherapy, King George’s Medical University, Shahmina Road, Lucknow 226001, India c Department of Biotechnology, IET, Sitapur Road, Lucknow 226021, India Received 9 June 2007; received in revised form 9 October 2007; accepted 23 October 2007 Available online 30 October 2007

Abstract Lung cancer is the most common cause of death throughout the world with cigarette smoking being established as the major etiological factor in lung cancer. Since not much information is available regarding the polymorphism in drug metabolizing enzymes and lung cancer risk in the Indian population, the present case–control study attempted to investigate the association of polymorphisms in cytochrome P450 1A1 (CYP1A1) and glutathione-S-transferase M1 (GSTM1) with risk to squamous cell carcinoma of lung malignancy. Patients suffering from lung cancer (n = 200) and visiting OPD facility of Department of Radiotherapy, King George’s Medical University, Lucknow, were included in the study. Equal number (n = 200) of age and sex matched healthy individuals were also enrolled in the study. Our data revealed that the variant genotypes of CYP1A1*2A, CYP1A1*2C and CYP1A1*4 were found to be over represented in the lung cancer patients when compared to controls. CYP1A1*2A variant genotypes (combined heterozygous and mutant genotypes) revealed significant association towards the lung cancer risk (OR: 1.93, 95%CI: 1.28–2.89, p = 0.002). Likewise, GSTM1 null genotypes were found to be over represented in patients when compared to controls. Haplotype analysis revealed that CYP1A1 haplotype, C-G-C increased the lung cancer risk (OR: 3.90, 95%CI: 1.00–15.04, p = 0.025) in the patients. The lung cancer risk was increased several two-to fourfold in the patients carrying the genotype combinations of CYP1A1*2A and GSTM1 suggesting the role of gene–gene interaction in lung cancer. Cigarette smoking or tobacco chewing or alcohol consumption was also found to interact with CYP1A1 genotypes in increasing the risk to lung cancer further demonstrating the role of gene–environment interaction in development of lung cancer. © 2007 Elsevier B.V. All rights reserved. Keywords: Polymorphism; CYP1A1; Lung cancer; Interaction; Gene–gene; Gene–environment

1. Introduction Lung cancer is the most frequent and the lethal malignancy of all cancers worldwide. Tobacco smoking

∗ Corresponding author. Tel.: +91 522 2627586x261; fax: +91 522 2628227/2621547. E-mail address: parmar [email protected] (D. Parmar).

0027-5107/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.mrfmmm.2007.10.006

has been established as the most important etiological factor of lung cancer [1]. However, only a small number of smokers develop lung cancer while others do not depending on the extent of smoking and exposure to other carcinogens present in the environment [2]. The gene–environment interaction for lung cancer development is largely attributed to the action of drug metabolizing enzymes [3]. Individual differences in the bioactivation of procarcinogens and in detoxification of

2

P.P. Shah et al. / Mutation Research 639 (2008) 1–10

carcinogens, arising from alterations in drug metabolizing enzymes may explain varying susceptibilities to lung cancer [4,5]. Genetic polymorphisms has been reported for cytochrome P4501A1 (CYP1A1) and glutathione-Stransferase M1 (GSTM1) isoenzymes, that are involved in the bioactivation and detoxification of chemical carcinogens present in the tobacco smoke and may influence host susceptibility to lung carcinoma [6,7]. Two functionally important polymorphisms have been described in CYP1A1 gene, which may result in an increased activity. An isoleucine/valine substitution in exon 7 at the heme binding region (CYP1A1*2C) and a thymine/cytosine point mutation in the MspI restriction site (CYP1A1*2A) has been reported in CYP1A1 gene [8,9]. The third polymorphism in the CYP1A1, i.e. (BsaI) which results in the amino acid substitution of threonine to asparagines has also been reported, though the functional role of this polymorphism is yet to be understood. Similarly homozygous deletions have been reported for the GSTM1 gene having no GSTM1 enzyme activity. Lack of these enzymes may potentially increase cancer susceptibility because of decreased ability to detoxify carcinogens such as benzo(a)pyrene-7-8-diol epoxide, the activated form of benzo(a)pyrene [10,11]. Ethnic differences have been reported in the distribution of CYP1A1, 1B1 and GSTM1 genotypes [3,12]. While, both CYP1A1*2A and CYP1A1*2C has been reported in Asians [13], the frequency of CYP1A1*2C is rare in the Caucasians [14]. However, the null genotype of GSTM1 is widely distributed in both Caucasian and Asians, with 48–57% of Caucasians and 35–63% of Asians lacking GSTM1 genotype [15]. Though the polymorphism in these genes has been associated with lung cancer, consistency in results is lacking among the different populations. In Japanese and Chinese, both CYP1A1*2A and CYP1A1*2C polymorphism have been associated with increased lung cancer risk, especially in relation to tobacco smoking [16–18], though such association has not been consistently reported in Caucasians [19,20]. Likewise, meta-analysis of case–control studies suggested an association between GSTM1 null genotype and lung cancer with the risks seemed to be higher among Asians and for squamous cell and small cell carcinomas [22,23]. Further, combinations of CYP1A1 polymorphism with GSTM1 null genotype significantly increased the lung cancer risk [6,24]. As compared to the other Asian populations, not much information is available on the association of lung cancer with polymorphism in CYP1A1, CYP1B1 and GSTM1 isoenzymes in the Indian population. Sobti et al. [25,26]

reported that CYP1A1*2A and CYP1A1*2C polymorphism is not significantly associated with lung cancer risk, though the risk was found to increase in heavy smokers. Sreeja et al. [27] reported significant association of CYP1A1*2A polymorphism with lung cancer risk which however decreased with cigarette smoking. Sobti et al. [25] also reported comparatively higher frequency of variant alleles of CYP1A1*2C polymorphism in their study. The present case–control study was therefore initiated to provide convincing evidence for the association of CYP1A1 polymorphism with squamous cell carcinoma of lung malignancy. To further identify the involvement of gene–environment interactions in lung cancer, the interaction of CYP1A1 genotypes and susceptibility to lung cancer was also studied in cigarette smokers, tobacco chewers and alcohol drinkers. As the lung cancer risk is increased in patients carrying combinations of phase I and phase II variant genotypes, attempts were also made to study the lung cancer risk in patients carrying combination of genotypes of phase I (CYP1A1) and phase II (GSTM1) enzymes. 2. Materials and methods A case–control study was initiated to investigate the association of functionally important polymorphisms in CYP1A1 (CYP1A1*2A, CYP1A1*2C and CYP1A1*4) genes, involved in the metabolic activation of carcinogens & GSTM1, which detoxifies reactive intermediates, with squamous cell carcinoma of lung malignancy. Patients suffering from lung cancer (n = 200) and visiting OPD facility of Department of Radiotherapy, King George’s Medical University, Lucknow, India, were included in the study. All cases were diagnosed with squamous cell carcinoma of lung by cytological, imaging and histopathological examinations. The average age of the patients was 56 ± 9 years. The control group consisted of 200 healthy men (average age 43 ± 12 years) belonging from same geographical location (Northern India) and the socio-economic condition with the same ethnicity. The volunteers (both controls and patients) were selected at random during the same period. All were informed about the study and their consent was taken prior to the study. The controls and patients were asked to fill up the detailed questionnaire regarding their family history, medical history, life style habits, etc. The questionnaire included among other details such as frequency of smoking/tobacco chewing/alcohol intake per day. Individuals having regular smoking habits and smoking index (S.I.) (cigarettes/day × 365 days) of 730 or more were classified as smokers [6]. Like-

P.P. Shah et al. / Mutation Research 639 (2008) 1–10

wise, the smokeless tobacco dose was estimated as ‘chewing year’ (i.e. CY = frequency of tobacco chewedkept/day × duration of year). Those who had CY of 365 or more were considered as tobacco chewers [32]. Similarly, cumulative exposure of alcohol drinking was derived by multiplying the total yearly consumption of alcohol (in liters/year) by the duration of habitual alcohol drinking (in years). Those who had cumulative exposure of alcohol beverage of about 90 L for a year were considered as regular alcohol users in our study [31]. 2.1. DNA isolation and determination of CYP1A1 and GSTM1 genotypes About 1ml of blood was drawn into citrate containing tubes from all patients as well as controls. DNA was isolated from whole blood with the QIAamp DNA mini kit (Qiagen, CA) following the manufacturer’s protocol. PCR-RFLP assay was used for identifying the functionally important polymorphisms in CYP1A1 (CYP1A1*2A, CYP1A1*2C and CYP1A1*4) gene. The method of Kawajiri et al. [9] was followed for genotyping the CYP1A1*2A polymorphism (MspI) of CYP1A1, while for detecting CYP1A1*2C polymorphisms (Ile/Val), the methodology described by Oyama et al. [28] was used. Similarly the method of Cascorbi et al. [29] was followed for genotyping the CYP1A1*4 polymorphism (BsaI). GSTM1 genotypes were determined by the method of Zhong et al. [30]. 2.2. Statistical analysis We determined whether genotype or allele frequencies of CYP1A1, CYP1B1 among cases and controls were in Hardy Weinberg Equilibrium (HWE) using standard X2 statistics. The haplotype analyses (haplotype frequency estimation and pair wise linkage disequilibrium between the SNPs) were carried out using Haploview (http://www.broad.mit. edu/mpg/haploview/). The association between individual variable demographic characteristics and environmental factors, or genetic polymorphisms or haplotypes and risk of lung cancer was estimated by conditional logistic regression. Using multiple logistic regression models, we determined the relationship of CYP1A1 and GSTM1 polymorphisms with lung risk after adjusting for other covariates. Those covariates include age, cigarette smoking, tobacco chewing and alcohol drinking in the study of lung cancer. Interaction between genotypes or between genotypes and environmental factors were also estimated by conditional logistic regression. The presence of gene–environment interac-

3

Table 1 Frequency distribution of demographic variables and putative risk factors of lung cancer Characteristics

Controls n (%)

Patients n (%)

Subjects Age (mean ± S.D.) Non-smoker Tobacco chewers Alcohol users Smoker Non-tobacco chewer Smokers Alcohol users Tobacco chewer Non-alcohol users Smokers Tobacco chewers Alcohol users

200 43 ± 12 138 (69) 23 (17) 13 (9) 62 (31) 148 (74) 33 (22) 17 (11) 52 (26) 171 (85) 46 (27) 40 (23) 29 (15)

200 56 ± 9 80 (40) 27 (34) 11 (14) 120 (60) 138 (69) 85 (62) 31 (22) 62 (31) 154 (77) 85 (55) 47 (31) 46 (23)

tion was tested by calculating second-order interaction terms between each genetic polymorphism and the environmental factor smoking in unconditional multiple logistic regression analyses [33]. All statistical analysis was performed with the SPSS software package (version 11.0 for windows; SPSS Chicago, IL). 3. Results The main characteristics of the study populations are summarized in Table 1. The two populations were very similar in numbers of male subjects and ethnic origin. The mean ages for controls and patients were 43 ± 12 and 56 ± 9, respectively. Amongst the cases, 60% were found to be smokers whereas only 31% in the controls have history of smoking. However, amongst the nonsmokers in the control group, 17% and 9% were tobacco chewers and alcohol users, respectively while 34% of the non-smoking cases were tobacco chewers and 14% were alcohol users (Table 1). Likewise, 31% of the cases and 26% of the controls were found to be tobacco chewers, i.e. have the habit of chewing tobacco on daily basis. Amongst the nontobacco chewers in control group, 22% were smokers and 11% were alcohol users while 62% on non-tobacco chewing cases were smokers and 22% were alcohol users (Table 1). About 23% of the cases were reported to be alcohol drinkers based on the daily consumption of alcohol, as compared to 15% found amongst the controls. Similarly, amongst the non-alcohol users in control group, 23% individuals were tobacco chewers and 27% individuals were smokers. Amongst the nonalcohol users in cases, 31% were tobacco chewers and 55% were smokers (Table 1).

4

P.P. Shah et al. / Mutation Research 639 (2008) 1–10

Table 2 Distribution of CYP1A1 genotypes (CYP1A1*2A, CYP1A1*2C and CYP1A1*4) and GSTM1 among lung cancer patients and controls Genotypes

Controls, n = 200 (%)

Cases, n = 200 (%)

Crude OR, (95%CI), p-value

Adjusted ORa (95%CI), p-value

CYP1A1*2A m1/m1 m1/m2 or m2/m2

137 (69) 63 (31)

106 (53) 94 (47)

1.0 (Ref.) 1.93 (1.28–2.89) 0.002b

1.0 (Ref.) 1.34 (0.79–2.25) 0.26

CYP1A1*2C Ile/Ile Ile/Val or Val/Val

156 (78) 44 (22)

133 (67) 67 (33)

1.0 (Ref.) 1.78 (1.14–2.78) 0.01b

1.0 (Ref.) 1.41 (0.80–2.47) 0.22

CYP1A1*4 Thr/Thr Thr/Asn or Asn/Asn

187 (93) 13 (6)

159 (80) 41 (20)

1.0 (Ref.) 3.70 (1.92–7.16) 0.00b

1.0 (Ref.) 3.57 (1.65–7.69) 0.001b

GSTM1 GSTM1 (+) GSTM1 (−)

152 (76) 48 (24)

111 (56) 89 (44)

1.0 (Ref.) 2.53 (1.65–3.89) 0.00b

1.0 (Ref.) 2.26 (1.3–3.80) 0.00b

Ref.: reference category. a Odds ratio adjusted in multivariate logistic regression models including age, smoking status, daily consumption of alcohol, tobacco chewing and CYP1A1 genotypes. b Level of significance (α) was considered to be <0.025 as per Bonferroni adjustment for n = 2, OR: Odds ratio, 95%CI: 95% confidence interval.

Table 2 summarizes the genotype frequencies for three polymorphic variants of CYP1A1 (CYP1A1*2A, CYP1A1*2C and CYP1A1*4) and GSTM1 genes. The genotype frequencies among controls were found to be in Hardy–Weinberg equilibrium. The prevalence of variant genotypes (combined heterozygous and homozygous mutant) of CYP1A1*2A was found to be increased in patients (47%) when compared to controls (31%). The crude odds ratio (OR) for variant genotypes was found to be 1.93 (95%CI: 1.28–2.89), which was found to be statistically significant (Table 2). After adjusting for age, cigarette smoking, tobacco chewing and alcohol consumption using multivariate logistic analysis, our data revealed that the risk was slightly reduced in the patients carrying variant genotypes, which was not found to be statistically significant. As observed with CYP1A1*2A the proportion of the individuals with variant genotypes of CYP1A1*2C was found to be higher in the patients (33%) than the controls (22%). While significant increase in the crude OR was observed for variant genotypes, the decrease in the relative risk was found to

persist in the patients with variant genotypes even when the data was adjusted for age, cigarette smoking, tobacco chewing and alcohol consumption habits (Table 2). As observed with CYP1A1*2A and CYP1A1*2C, the frequency of variant genotypes of CYP1A1*4 was also found to be higher in the patients (20%) than in the controls (6%). As evident from OR, the risk was also found to be higher in patients carrying valiant genotypes which was found to be statistically significant (Table 2). The frequency of the null genotype of GSTM1 was found to be higher in patients when compared to controls (Table 2). The risk (crude and adjusted OR) for GSTM1 null genotype was found to be relatively higher in the patients when compared to respective controls, which was further found to be statistically significant (Table 2). Haplotype approach revealed that five possible haplotypes were observed both in the patients and the controls. The haplotype, C-A-T was considered to be the reference carrying wild type alleles. The frequency of the other four haplotypes (C-A-C, C-G-T, A-A-T, C-G-C)

Table 3 Distribution of CYP1A1 haplotypes among lung cancer patients and controls CYP1A1 haplotypes (C2453A, 2455G, T3927C)

Control frequency % (n)

Case frequency % (n)

OR (95%CI)

p-Value

C-A-T C-A-C C-G-T A-A-T C-G-C

0.74 (148) 0.14 (28) 0.05 (10) 0.04 (8) 0.03 (6)

0.57 (114) 0.20 (40) 0.07 (14) 0.07 (14) 0.09 (18)

1.0 (Ref.) 1.85 (0.86–3.98) 1.82 (0.54–6.02) 2.27 (0.63–8.13) 3.90 (1.00–15.0)

0.110 0.322 0.197 0.025*

Ref.: reference category. * Level of significance (α) was considered to be <0.025 as per Bonferroni adjustment for n = 2, OR: Odds ratio, 95%CI: 95% confidence interval.

P.P. Shah et al. / Mutation Research 639 (2008) 1–10

5

Table 4 Genotype combinations of CYP1A1 (CYP1A1*2A) and GSTM1 in cases and controls CYP1A1*2A (MspI)

GSTM1

Controls, n = 200 (%)

Cases, n = 200 (%)

OR (95%CI), p-value

ORa (95%CI), p-value

m1/m1

Positive Null

103 (52) 34 (17)

60 (30) 1.0 (Ref.) 46 (23)

1.0 (Ref.) 2.32 (1.34–4.00), 0.002b

2.85 (1.53–5.32), 0.001b

m1/m2

Positive Null

44 (22) 12 (6)

41 (20) 32 (16)

1.59 (0.94–2.72), 0.082 4.58 (2.19–9.55), 0.00b

1.46 (0.77–2.75), 0.247 2.71 (1.11–6.60), 0.02b

m2/m2

Positive Null

5 (2) 2 (1)

10 (5), 3.43 (1.12–10.51), 0.02b 11 (6)

2.81 (0.68–11.50), 0.150 9.44 (2.02–44.03), 0.01b

3.28 (0.65–16.50), 0.150

Ref.: reference category. a Odds ratio adjusted in multivariate logistic regression models including age, smoking status, daily consumption of alcohol and tobacco chewing. b Level of significance (α) was considered to be <0.025 as per Bonferroni adjustment for n = 2, OR: Odds ratio, 95%CI: 95% confidence interval.

was found to be increased in the patients resulting in the increase in the OR in the patients (Table 3). The patients carrying haplotype (C-G-C) exhibited 3.90-fold increase in risk as reflected by increase in OR. The haplotypes, C-A-C, C-G-T and A-A-T were also associated with increase in risk as reflected by increased OR, though the increase was not found to be statistically significant (Table 3). The distribution of genotype combinations for CYP1A1*2A and GSTM1 is summarized in Table 4. The risk genotype combinations, i.e. heterozygous genotype of CYP1A1*2A and null genotype of GSTM1 resulted in 4.5-fold increase in the lung cancer risk (OR: 4.58, 95%CI: 2.19–9.55) which was further found to be statistically significant. The risk was found to be reduced after adjusting for age, smoking, tobacco chewing and alcohol consumption using multivariate logistic analysis (OR: 2.71, 95%CI: 1.11–6.60), which was further found to be statistically significant. Almost tenfold increase in the risk (OR: 9.44, 95%CI: 2.02–44.03) was observed in the patients carrying genotype combination of homozygous mutant of CYP1A1*2A and null genotype of GSTM1 which was found to be statistically significant. Interestingly, after adjusting the data by multivariate logistic analysis the risk was found to be reduced (OR: 3.28, 95%CI: 0.65-16.50) which was found to be statistically nonsignificant. The effect of the interaction of the risk modifiers, i.e. cigarette smoking, tobacco chewing and alcohol consumption and the genotypes distribution of CYP1A1*2A and CYP1A1*2C in controls and patients is summarized in Table 5. The distribution of variant genotypes of CYP1A1*2A was found to be increased in smokers amongst the patients resulting in an increase in lung cancer risk. Cigarette smoking in patients with either heterozygous or homozygous mutant genotype of CYP1A1*2A

resulted in almost sixfold increase in the lung cancer risk (OR: 6.13, 95%CI: 3.04–12.39) which was also found to be statistically significant. Similarly almost fourfold increase in the risk was associated with cigarette smoking in the patients carrying variant genotypes of CYP1A1*2C (OR: 4.35, 95%CI: 1.96–9.62) when compared with non-smokers (Table 5). Tobacco chewing also significantly increased the risk in the patients when compared to the non-tobacco chewers (Table 5). The risk associated with tobacco chewing increased several fold in the patients with variant genotypes of CYP1A1*2A (OR: 4.15, 95%CI: 1.85–9.29) or CYP1A1*2C (OR: 2.47, 95%CI: 1.04–5.85). This increase in risk associated with tobacco chewing was further found to be statistically significant for CYP1A1*2A but not for CYP1A1*2C. As observed with tobacco chewing and smoking, daily alcohol use was also significantly found to increase the lung cancer risk in the patients (Table 5). Alcohol use was found to be associated with several fold increase in the risk in the patients with variant genotypes for CYP1A1*2A (OR: 5.36, 95%CI: 1.89–15.17) or CYP1A1*2C (OR: 2.86, 95%CI: 1.05-7.77) when compared to the patients who are not regular alcohol users (Table 5). 4. Discussion Our data indicating prevalence of variant genotypes of CYP1A1*2A in the patients is consistent with the earlier reports indicating significant association of lung cancer with MspI polymorphism in Asians [13]. In contrast, the risk of lung cancer with CYP1A1*2A polymorphism was earlier not clearly established in the Caucasians or other European populations [19,35] as the Caucasian populations suffer from an extremely low prevalence of the mutant genotype of CYP1A1*2A polymorphic site [10,35,36]. Taoli et al. [37], however, in a pooled analy-

6

Table 5 Interaction of CYP1A1 genotypes and smoking, tobacco chewing and alcohol consumption on risk of squamous cell carcinoma of lung Genotypes

Non-smoker

Smoker Cases (n = 80) (%)

OR

95%CI

p-Value

CYP1A1*2A Wild typea Variantb

89 (64) 49 (36)

63 (79) 17 (21)

1.0 (Ref.) 0.49

0.25–0.92

0.03

CYP1A1*2C Wild type Variant

103 (75) 35 (25)

64 (80) 16 (20)

1.0 (Ref.) 0.73

0.37–1.43

0.367

Genotypes

Non-tobacco

Controls (n = 62) (%)

Cases (n = 120) (%)

OR

95%CI

p-Value

48 (77) 14 (23)

43 (36) 77 (64)

1.0 (Ref.) 6.13

3.04–12.39

0.00c

53 (85) 9 (15)

69 (58) 51 (42)

1.0 (Ref.) 4.35

1.96–9.62

0.0001c

Tobacco

Controls (n = 148) (%)

Cases (n = 138) (%)

OR

95%CI

p-Value

Controls (n = 52) (%)

Cases (n = 62) (%)

OR

95%CI

p-Value

98 (66) 50 (34)

80 (58) 58 (42)

1.0 (Ref.) 1.42

0.87–2.29

0.150

39 (75) 13 (25)

26 (42) 36 (58)

1.0 (Ref.) 4.15

1.85–9.29

0.0003c

CYP1A1*2C Wild type 114 (77) Variant 34 (23)

94 (68) 44 (32)

1.0 (Ref.) 1.56

0.92–2.65

0.09

42 (81) 10 (19)

39 (63) 23 (37)

1.0 (Ref.) 2.47

1.04–5.85

0.03

CYP1A1*2A Wild type Variant

Genotypes

Non-alcohol user Controls (n = 171) (%)

Alcohol user Cases (n = 154) (%)

OR

95%CI

p-Value

CYP1A1*2A Wild type 115 (67) Variant 56 (33)

89 (58) 65 (42)

1.0 (Ref.) 1.49

0.95–2.35

0.078

CYP1A1*2C Wild type 135 (79) Variant 36 (21)

111 (72) 43 (28)

1.0 (Ref.) 1.45

0.87–2.41

0.149

OR: odds ratio; CI: confidence interval; Ref.: reference category. a Wild type genotype. b Heterozygous and homozygous mutant genotype. c Level of significance (α) was considered to be <0.025 as per Bonferroni adjustment for n = 2.

Controls (n = 29) (%)

Cases (n = 46) (%)

OR

95%CI

p-Value

22 (76) 7 (24)

17 (37) 29 (63)

1.0 (Ref.) 5.36

1.89–15.17

0.001c

21 (72) 8 (28)

22 (48) 24 (52)

1.0 (Ref.) 2.86

1.05–7.77

0.03

P.P. Shah et al. / Mutation Research 639 (2008) 1–10

Controls (n = 138) (%)

P.P. Shah et al. / Mutation Research 639 (2008) 1–10

sis of lung cancer in subjects under 45 years, in a Greek population, reported a significant association between the homozygous variant of the CYP1A1*2A polymorphism and lung cancer, though the association was found to be confined to never smokers. Pooled analysis of 22 case–control studies [34] and report of Dialyna et al. [38] however showed clear association between homozygous genotype of CYP1A1*2A and lung cancer in the Caucasian population. In contrast to the significant association of CYP1A1*2C polymorphism of CYP1A1 with lung cancer in Asians [17,39,40], no significant association of lung cancer was observed with mutant genotype of CYP1A1*2C in our study. As against relatively higher frequency of Val/Val genotype in Japanese and Chinese populations [17,41], the Val/Val genotype has been reported to be rare in Indians [42]. No mutant homozygous genotype of CYP1A1*2C was also observed in our study. Low frequency of the mutant genotype of CYP1A1*2C has also been reported in the Caucasians [14,21], though individuals with CYP1A1*2C polymorphism were at increased risk of lung cancer, particularly of SCC. This association however appeared to be stronger in Caucasians than the Asians but was limited to ‘never smokers’ [43]. As observed in Caucasians as well as Asians, the frequency of the mutant genotype of CYP1A1*4 was found to be very low in our case–control study. Singh et al. [44] also reported very low frequency of this homozygous mutant genotype in his study using controls of the same geographical location. Although Schwarz et al. [45] reported that CYP1A1*4 variants exhibit the greatest efficiency amongst the different SNPs of CYP1A1, no association of CYP1A*4 polymorphism with lung cancer risk has been reported so far. Haplotype analysis have further shown that unlike in the Caucasians, 3801T > C and 2455 A > G polymorphisms of CYP1A1 did not exhibited LD in our study [46]. Our data, using combined genotype analysis with a haplotypes approach revealed a stronger association of CYP1A1 haplotype with risk to lung cancer than individual SNPs in CYP1A1. Our data have provided evidence that though the frequency of the CYP1A1 haplotypes C-A-C, C-G-T, A-A-T and C-G-C is increased in the patients, haplotype C-G-C was associated with approximately fourfold increase risk for lung cancer. Though no such association of CYP1A1 haplotypes has been reported with lung cancer in either Caucasians or the Asians, studies have shown that C-A-C haplotypes is associated with higher enzyme activity [46]. Although null genotypes of GSTM1 were found to be at increased risk, further increase in the risk in the

7

patients carrying combination of variant genotypes of CYP1A1*2A and null genotype of GSTM1 has provided evidence for gene–gene interaction and risk to lung cancer. Studies have earlier reported that GSTM1 (−) became more markedly expressed in the patients with CYP1A1*2C genotype [40,47]. A similar effect, though less strong has been shown for non-smoker patients in pooled analysis while patients with CYP1A1 Val variants and GSTM1 null genotype exhibited much more increase in the risk for lung cancer in the Caucasians [14]. Likewise, though the results obtained from Caucasians are inconsistent, a population based study including Caucasians, Japanese and Hawaiians showed increased OR for lung cancer was associated with the CYP1A1*2A, especially when combined with the GSTM1 deletions genotype [18]. Significant increase in the risk amongst the smokers in the patients who carry CYP1A1 variant alleles have further shown the potential modifying effect of CYP1A1 genotypes on the relationship between tobacco and lung cancer. It is well established that cigarette smoking is associated with increased risk of lung cancer [48]. An increase in the lung cancer risk with Val allele of CYP1A1*2C has been reported in light smokers (<30 pack years), but not in heavy smokers [17]. It has been hypothesized that at lower level of tobacco smoke exposure, the relatively minor variability in carcinogen metabolism due to deficient GSTM1 and Val to Ile substitution may be more important than at higher exposure levels, where numerous effects of smoking predominate in the heavy smokers. Though our study did not differentiated between light and heavy smokers, increased risk in the smokers have further provided support to the earlier studies that tobacco smoking is an important modifier of genetic susceptibility to lung cancer. Similarly two- to fourfold increase in the risk in tobacco chewers carrying variant genotypes of CYP1A1 could be of significance especially for our country, as tobacco chewers constitute a significant proportion of the tobacco users. Though the carcinogenic potential of tobacco products for chewing is suggested in animal models, it has been shown that tobacco chewing does not seem to have an effect on lung cancer risk in our population [49,50]. However, a case–control study from Pakistan reported an elevated OR of lung cancer for heavy tobacco chewers [51]. Likewise, several fold (fiveto sixfold) increase in the risk in the alcohol users with the variant genotypes of CYP1A1*2A and CYP1A1*2C have demonstrated the interaction of alcohol with lung cancer risk. Epidemiological studies have also reported a positive relationship between alcohol consumption and lung cancer. As smoking is correlated with alcohol con-

8

P.P. Shah et al. / Mutation Research 639 (2008) 1–10

sumption, the major concern in the examination of an association between alcohol consumption and lung cancer has been the failure to control for confounding by smoking. In a pooled analysis of cohort studies, it was found that alcohol consumption was associated with increased risk to lung cancer in male never smokers [52]. Our data demonstrating increased risk in the tobacco chewers and alcohol consumers have suggested that the possibility of interaction of CYP1A1 genotypes with alcohol or tobacco chewing in increasing the lung cancer risk. Though tobacco smoking or chewing or alcohol use was found to increase the risk, the present study, particularly after stratification with respect to risk factors such as tobacco chewing or alcohol drinking, have its limitations of a small sample size. Due to the small sample size, the effect of estimates may have low precision and could have occurred by chance or could have been biased. Further studies with a much larger sample size are needed to support the present findings. In conclusion, the results of the present study have provided convincing evidence that CYP1A1 polymorphism is an important modifying factor in determining susceptibility to lung cancer. Significant increase in the risk in the individuals carrying variant genotypes of CYP1A1 and GSTM1 have further provided evidence that gene–gene interaction may play an important role in the development of lung cancer. Likewise significant interactions of CYP1A1 genotypes with tobacco, both in the form of tobacco smoking or tobacco chewing and alcohol have demonstrated the important of gene–environment interactions in modifying the susceptibility to lung cancer. Acknowledgements The authors are grateful to the Director, Industrial Toxicology Research Centre, Lucknow for his keen interest and support in carrying out the study. The financial support of ICMR, N. Delhi, in carrying out the above studies is gratefully acknowledged. The technical assistance of Mr. B.S. Pandey and Mr. Rajesh Misra and computer help of Mr. Mohd Aslam is also gratefully acknowledged. ITRC Communication Number: 2598. References [1] D. Gupta, P. Boffatta, V. Gaborieau, S.K. Jindal, Risk factors of lung cancer in Chandigarh, India, J. Med. Res. 113 (2001) 142–150. [2] S.S. Hecht, Tobacco smoke carcinogens and lung cancer, J. Natl. Cancer Inst. 14 (1999) 1194–1210. [3] K. Inoue, T. Asao, T. Shimada, Ethnic-related differences in the frequency distribution of genetic polymorphisms in the CYP1A1

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

and CYP1B1 genes in Japanese and Caucasian populations, Xenobiotica 30 (2000) 285–295. T. Shimada, C.L. Hayes, H. Yamazaki, S. Amin, S.S. Hecht, F.P. Guengerich, T.R. Sutter, Activation of chemically diverse procarcinogens by human cytochrome P-450 1B1, Cancer Res. 56 (1996) 2979–2984. R. Barouki, Y. Morel, Repression of cytochrome P450 1A1 gene expression by oxidative stress: mechanisms and biological implications, Biochem. Pharmacol. 61 (2001) 511–516. L. Quinones, D. Lucas, J. Godoy, D. Caceres, F. Berthou, N. Varela, K. Lee, C. Acevedo, L. Martinez, A.M. Aguilera, L. Gil, CYP1A1, CYP2E1 and GSTM1 genetic polymorphisms. The effect of single and combined genotypes on lung cancer susceptibility in Chilean people, Cancer Lett. 174 (2001) 35–44. J. Schneider, U. Bernges, M. Philipp, H.J. Woitowitz, CYP1A1 and CYP1B1 polymorphism and lung cancer risk in relation to tobacco smoking, Cancer Genomics Proteomics (2004) 189–198. S. Hayashi, J. Watanabe, K. Nakachi, K. Kawajiri, Genetic linkage of lung cancer-associated MspI polymorphisms with amino acid replacement in the heme binding region of the human cytochrome P450IA1 gene, J. Biochem. 110 (1991) 407–411 (Tokyo). K. Kawajiri, H. Eguchi, K. Nakachi, T. Sekiya, M. Yamamoto, Association of CYP1A1 germ line polymorphisms with mutations of the p53 gene in lung cancer, Cancer Res. 56 (1996) 72–76. A. d’Errico, N. Malats, P. Vineis, P. Boffetta, Review of studies of selected metabolic polymorphisms and cancer, IARC Sci. Publ. 148 (1999) 323–393. R.C. Strange, A.A. Fryer, The glutathione-S-transferases: influence of polymorphism on cancer susceptibility, in: P. Vineis, N. Malats, M. Lang, A. d’Errico, N. Caporaso, J. Cuzick, P. Boffetta (Eds.), Metabolic Polymorphisms and Susceptibility to Cancer, vol. 148, IARC Scientific Publications, Lyon, France, 1999, pp. 231–249. S. Antilla, I. Kalina, Le. Marchand, S.J. London, A. Rannug, M. Romkes, J. Salagovic, B. Schoket, L. Gaspari, E. Taioli, R.J. Hung, P. Boffetta, J. Brockmoller, D. Butkiewicz, I. Cascorbi, M.L. Clapper, S. Garte, A. Haugen, A. Hirvonen, CYP1A1 and GSTM1 genetic polymorphisms and lung cancer risk in Caucasians non-smokers: a pooled analysis, Carcinogenesis 24 (2003) 875–882. N. Song, W. Tan, D. Xing, D. Lin, CYP1A1 polymorphism and risk of lung cancer in relation to tobacco smoking: a case–control study in China, Carcinogenesis 22 (2002) 11–16. R.J. Hung, P. Boffetta, J. Brockmoller, D. Butkiewicz, I. Cascorbi, M.L. Clapper, S. Garte, A. Haugen, A. Hirvonen, S. Anttila, I. Kalina, Le Marchand, S.J. London, A. Rannug, M. Romkes, J. Salagovic, B. Schoket, L. Gaspari, E. Taioli, CYP1A1 and GSTM1 genetic polymorphisms and lung cancer risk in Caucasian non-smokers: a pooled analysis, Carcinogenesis 24 (2003) 875–882. S. Garte, L. Gaspari, A.K. Alexandrie, C. Ambrosone, H. Autrup, Metabolic gene polymorphism frequencies in control populations, Cancer Epidemiol. Biomark. Prev. 10 (2001) 1239–1248. Y.S. Hong, J.H. Chang, O.J. Kwon, Y.A. Ham, J.H. Choi, Polymorphism of the CYP1A1 and glutathione-S-transferase gene in Korean lung cancer patients, Exp. Mol. Med. 30 (1998) 192– 198. H. Sugimura, K. Wakai, K. Genka, K. Nagura, H. Igarashi, K. Nagayama, A. Ohkawa, S. Baba, B.J. Morris, S. Tsugane, Y. Ohno, C. Gao, Z. Li, T. Takezaki, K. Tajima, T. Iwamasa, Association of Ile462Val (exon 7) polymorphism of cytochrome P450 IA1 with lung cancer in the Asian population: further evidence from

P.P. Shah et al. / Mutation Research 639 (2008) 1–10

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

a case–control study in Okinawa, Cancer Epidemiol. Biomark. Prev. 7 (1998) 413–417. N. Song, W. Tan, D. Xing, D. Lin, CYP 1A1 polymorphism and risk of lung cancer in relation to tobacco smoking: a case–control study in China, Carcinogenesis 22 (2001) 11–16. P.G. Shields, N.E. Caporaso, R.T. Falk, H. Sugimura, G.E. Trivers, B.F. Trump, R.N. Hoover, A. Weston, C.C. Harris, Lung cancer, race, and a CYP1A1 genetic polymorphism, Cancer Epidemiol. Biomark. Prev. 2 (1993) 481–485. C. Bouchardy, H. Wikman, S. Benhamou, A. Hirvonen, P. Dayer, K.H. Pursiainen, CYP1A1 genetic polymorphisms, tobacco smoking and lung cancer risk in a French Caucasian population, Biomarkers 2 (1997) 131–134. A.S. Wenzlaff, M.L. Cote, C.H. Bock, S.J. Land, S.K. Santer, D.R. Schwartz, A.G. Schwartz, CYP1A1 and CYP1B1 polymorphisms and risk of lung cancer among never smokers: a population-based study, Carcinogenesis 26 (2005) 2207–2212. S. Benhamou, W.J. Lee, A.K. Alexandrie, P. Boffeta, C. Bouchardy, D. Butkiewicz, Meta and pooled analyses of the effects of glutathione-S-transferases M1 polymorphism and smoking on lung cancer risk, Carcinogenesis 23 (2002) 1343–1350. P. Vineis, M. Malats, M. Lang, A. d’Errico, N. Caporaso, J. Cuzick, P. Boffetta, Metabolic polymorphisms and susceptibility to cancer eds, IARC Sci. Publ. 148 (1999) 1–510. X.R. Yang, S. Wacholder, Z. Xu, M. Dean, V. Clark, B. Gold, L.M. Brown, B.J. Stone, J.F. Fraumeni, N.E. Caporaso, CYP1A1 and GSTM1 polymorphisms in relation to lung cancer risk in Chinese women, Cancer Lett. 214 (2004) 197–204. R.C. Sobti, S. Sharma, A. Joshi, S.K. Jindal, A. Janmeja, CYP1A1 and CYP2D6 polymorphism and risk of lung cancer in a North Indian population, Biomarkers 8 (2003) 415–428. R.C. Sobti, S. Sharma, A. Joshi, S.K. Jindal, A. Janmeja, Genetic polymorphism of the CYP1A1, CYP2E1, GSTM1 and GSTT1 genes and lung cancer susceptibility in a North Indian population, Mol. Cell Biochem. 266 (2004) 1–9. L. Sreeja, V. Syamala, S. Hariharan, J. Madhavan, S.C. Devan, R. Ankathil, Possible risk modification by CYP1A1, GSTM1 and GSTT1 gene polymorphisms in lung cancer susceptibility in a South Indian population, J. Hum. Genet. 50 (2005) 618–627. T. Oyama, T. Mitsudomi, T. Kawamoto, A. Ogami, T. Osaki, Kodama, Y. Yasumoto, Detection of CYP1A1 gene polymorphism using designed RFLP and distributions of CYP1A1 genotypes in Japanese, Int. Arch. Occup. Environ. Health 67 (1995) 253–256. I. Cascorbi, J. Brockmoller, I. Roots, A C4887A polymorphism in exon 7 of human CYP1A1: population frequency, mutation linkages, and impact on lung cancer susceptibility, Cancer Res. 56 (1996) 4965–4969. S. Zhong, A. Howie, B. Ketterer, J. Taylor, J.D. Hayes, G. Beckette, C.G. Wathen, C.R. Wolf, N.K. Spurr, Glutathione-Stransferase mu locus: use of genotyping and phenotyping assays to assess association with lung cancer susceptibility, Carcinogenesis 12 (1991) 1533–1537. H.C. Hung, J. Chuang, Y.C. Chien, H.D. Chern, C.P. Chiang, Y.S. Kuo, A. Chen, C.J. Hildesheim, Genetic polymorphisms of CYP2E1, GSTM1, and GSTT1; Environmental factors and risk of oral cancer, Cancer Epidemiol. Biomark. Prev. 6 (1997) 901–905. N. Sikdar, S.K. Mahmud, R.R. Paul, B. Roy, Polymorphism in CYP1A1 and CYP2E1 genes and susceptibility to leukoplakia in Indian tobacco users, Cancer Lett. 195 (2003) 33–42.

9

[33] K. De Ruyck, M. Szaumkessel, I. De Rudder, A. Dehoorne, A. Vral, K. Claes, A. Velghe, J.V. Meerbeeck, H. Thierens, Polymorphism in base-excision repair and nucleotide-excision repair genes in relation to lung cancer risk, Mutat. Res. 631 (2007) 101–110. [34] P. Vineis, F. Veglia, S. Benhamou, D. Butkiewicz, I. Cascorbi, M.L. Clapper, V. Dolzan, A. Haugen, A. Hirvonen, M. IngelmanSundberg, M. Kihara, C. Kiyohara, P. Kremers, Le Marchand, S. Ohshima, R. Pastorelli, A. Rannug, M. Romkes, B. Schoket, P. Shields, R.C. Strange, I. Stucker, H. Sugimura, S. Garte, L. Gaspari, E. Taioli, CYP1A1 T3801 C polymorphism and lung cancer: a pooled analysis of 2451 cases and 3358 controls, Int. J. Cancer 104 (2003) 650–657. [35] A. Hirvonen, K. Husgafvel-Pursiainen, A. Karjalainen, S. Anttila, H. Vainio, Point-mutational MspI and Ile-Val polymorphisms closely linked in the CYP1A1 gene: lack of association with susceptibility to lung cancer in a Finnish study population, Cancer Epidemiol. Biomark. Prev. 1 (1992) 485–489. [36] N. Drakoulis, I. Cascorbi, J. Brockmoller, C.R. Gross, I. Roots, Polymorphisms in the human CYP1A1 gene as susceptibility factors for lung cancer: exon-7 mutation (4889 A to G), and a T to C mutation in the 3 -flanking region, Clin. Investig. 72 (1994) 240–248. [37] E. Taioli, L. Gaspari, S. Benhamou, P. Boffetta, J. Brockmoller, D. Butkiewicz, I. Carcorbi, M.L. Clapper, V. Dolzan, A. Haugen, A. Hirvonen, K. Husgafvel-Pursisinen, I. Kalina, P. Kremers, Le Marchand, S. London, A. Rannug, M. Romkes, B. Schoket, J. Seidegard, R.C. Strange, I. Stucker, J. To-Figueras, S. Garte, Polymorphisms in CYP1A1, GSTM1, GSTT1 and lung cancer below the age of 45 years, Int. J. Epidemiol. 32 (2003) 60–63. [38] I.A. Dialyna, S. Miyakis, N. Georgatou, Da. Spandidos, Genetic polymorphisms of CYP1A1, GSTM1 and GSTT1 genes and lung cancer risk, Oncol. Rep. 10 (2003) 1829–1835. [39] K. Nakachi, K. Imai, S. Hayashi, K. Kawajiri, Polymorphisms of the CYP1A1 and glutathione-S-transferase genes associated with susceptibility to lung cancer in relation to cigarette dose in a Japanese population, Cancer Res. 53 (1993) 2994–2999. [40] M. Kihara, M. Kihara, K. Noda, Risk of smoking for squamous and small cell carcinomas of the lung modulated by combinations of CYP1A1 and GSTM1 gene polymorphisms in a Japanese population, Carcinogenesis 16 (1995) 2331–2336. [41] E.J. Lee, B. Zhao, S.M. Moochhala, S.S. Ngoi, Frequency of mutant CYPIA1, NAT2 and GSTM1 alleles in a normal Chinese population, Pharmacogenetics 4 (1994) 355–358. [42] B. Zhao, E.J. Lee, J.Y. Wong, P.N. Yeoh, N.H. Gong, Frequency of mutant CYP1A1, NAT2 and GSTM1 alleles in normal Indians and Malays, Pharmacogenetics 5 (1995) 275–280. [43] Le Marchand, C. Guo, S. Benhamou, C. Bouchardy, I. Cascorbi, M.L. Clapper, S. Garte, A. Haugen, M. Ingelman-Sundberg, M. Kihara, A. Rannug, D. Ryberg, I. Stucker, H. Sugimura, E. Taioli, Pooled analysis of the CYP1A1 exon 7 polymorphism and lung cancer (United States), Cancer Causes Control 14 (2003) 339–346. [44] V. Singh, N. Rastogi, A. Sinha, A. Kumar, N. Mathur, M.P. Singh, A study on the association of cytochrome-P450 1A1 polymorphism and breast cancer risk in north Indian women, Breast Cancer Res. Treat. 101 (2007) 73–81. [45] D. Schwarz, P. Kisselev, I. Cascorbi, W.H. Schunck, I. Roots, Differential metabolism of benzo[a]pyrene and benzo[a]pyrene7,8-dihydrodiol by human CYP1A1 variants, Carcinogenesis 22 (2001) 453–459.

10

P.P. Shah et al. / Mutation Research 639 (2008) 1–10

[46] B.L. Chang, S.L. Zheng, S.D. Isaacs, A. Turner, G.A. Hawkins, K.E. Wiley, E.R. Bleecker, P.C. Walsh, D.A. Meyers, W.B. Isaacs, J. Xu, Polymorphism in the CYP1A1 gene is associated with prostate cancer risk, Int. J. Cancer 106 (2003) 375–378. [47] S. Chen, K. Xue, L. Xu, G. Ma, J. Wu, Polymorphisms of the CYP1A1 and GSTM1 genes in relation to individual susceptibility to lung carcinoma in Chinese population, Mutat. Res. 458 (2001) 41–47. [48] IARC, Monograph on the evaluation of the carcinogenic risk of chemicals to humans-tobacco smoking, IARC Sci. Publ. 38 (1986) 1–3. [49] R.P. Dikshit, S. Kanhere, Tobacco habits and risk of lung, oropharyngeal and oral cavity cancer: a population-based case–control study in Bhopal, India, Int. J. Epidemiol. 29 (2000) 609–614.

[50] V. Gajalakshmi, R.J. Hung, A. Mathew, C. Varghese, P. Brennan, P. Boffetta, Tobacco smoking and chewing, alcohol drinking and lung cancer risk among men in southern India, Int. J. Cancer 107 (2003) 441–447. [51] Y. Bhurgri, E. Decullier, A. Bhurgri, S. Nassar, A. Usman, P. Brennan, P. Boffetta, A case–control study of lung cancer in Karachi, Pakistan, Int. J. Cancer 98 (2002) 952–955. [52] J.L. Freudenheim, J. Ritz, S.A. Smith-Warner, D. Albanes, E.V. Bandera, P.A. van den Brandt, G. Colditz, D. Feskanich, R.A. Goldbohm, L. Harnack, A.B. Miller, E. Rimm, T.E. Rohan, T.A. Sellers, J. Virtamo, W.C. Willett, D.J. Hunter, Alcohol consumption and risk of lung cancer: a pooled analysis of cohort studies, Am. J. Clin. Nutr. 82 (2005) 657–667.