Association of Glutathione S -Transferase,EPHX,and p53 codon 72 Gene Polymorphisms with Adult Acute Myeloid Leukemia

Abstract

Polymorphisms in genes encoding detoxification enzymes have been suggested as susceptibility factors for many solid tumors. However, their association with hematological malignancies is controversial. A case–control study was done to determine the association between glutathione S-transferase M1 (GSTM1), GSTT1, GSTP1, EPHX1, and p53 codon 72 polymorphisms as risk factors in 120 adult acute myeloid leukemia (AML) cases and 202 healthy controls by polymerase chain reaction–restriction fragment length polymorphism techniques. Data were analyzed using χ ² and conditional logistic regression model. None of the polymorphisms studied alone was associated with increased risk for AML. However, the frequency of GSTT1 null genotype was higher among controls (28.7%) than AML cases (21.6%), which showed a protective effect of the null genotype (odds ratio = 0.58, 95% confidence interval: 0.33–1.05, p = 0.07). In a combined analysis, both EPHX1 (His113His) and GSTP1 (Ile/Val) genes imparted a fourfold risk for adult AML but did not reach statistical significance (odds ratio = 4.22, 95% confidence interval: 0.992–17.99, p = 0.05). These findings suggest that the etiology of adult AML cannot be explained by polymorphism at a single locus, perhaps because of complexity involved in the metabolism of diverse xenobiotic compounds, and therefore, multiple gene–gene interactions should be investigated to predict the risk of AML.

Introduction

A cute myeloid leukemia (AML) is a very heterogeneous disease with respect to clinical features and acquired genetic alterations. The etiology of AML appears to be multifactorial and the exposure to benzene, ionizing radiation, and cytotoxic therapy have been implicated as risk factors in the pathogenesis of AML. Every individual is considered to be exposed to benzene either from vehicle exhaust emission or from active and passive smoking (Groves et al., 1994; Infante-Rivard et al., 1999). Exposure to pesticides and alcohol are other environmental risk factors that are found to be inadequately and inconsistently associated with the development of AML. Carcinogens present in these environmental factors may cause DNA damage at the level of hematopoietic progenitors, which is an essential prerequisite for the development of AML. Genetic polymorphism in xenobiotic metabolizing enzymes seems to play an important role in determining individual susceptibility to these carcinogens (Taningher et al., 1999). Recent epidemiological studies have demonstrated that genetic susceptibility and environmental exposure together may play a significant role in the etiology of AML (Majumdar et al., 2008; Lee et al., 2009; Xiong et al., 2009).

Primary candidates for gene–environment interaction study in the carcinogenesis are enzymes involved in the activation and detoxification of xenobiotic compounds. Genetic polymorphisms have been identified in many of these enzymes and these may be responsible for variation in the enzymatic activity, thereby leading to interindividual and interethnic variation in the metabolism of carcinogens. Genetic polymorphisms of two such enzymes have been widely implicated in cancer susceptibility. Microsomal epoxide hydrolase (EPHX1), an important phase I biotransformation enzyme, is involved in the first-pass metabolism of highly reactive epoxide intermediates and oxygen radicals. EPHX1 plays a dual role in carcinogenesis depending on the exposure to type of environmental substrates. Besides providing protection against the toxicity of reactive epoxides intermediate, EPHX1 along with CYP enzymes plays a role in the metabolic activation of procarcinogens such as benzo(a)pyrene present in tobacco smoke, which leads to highly reactive carcinogenic diolepoxides (Miyata et al., 1999). Literature about the role of EPHX1 polymorphism in the development of AML is very scanty. One study reported a protective role of EPHX1 in leukemogenesis of childhood acute lymphoblastic leukemia (ALL), whereas another study considered it as a risk factor in adult AML with t(8:21) translocation (Lebailly et al., 2002; Silveira Vda et al., 2009).

Glutathione S-transferases (GSTs) are phase II biotransformation enzymes that convert the activated metabolites of procarcinogens of phase I reactions into nonreactive and water-soluble compounds. GSTM1 and GSTT1 both exhibit deletion polymorphism that results in inactivation of the enzyme (Seidegård et al., 1988). GSTP1, located on chromosome 11 (11q13), encodes the major enzyme involved in the inactivation of tobacco-related procarcinogens. The GSTP1 Ile105Val polymorphism is associated with reduced catalytic activity, which may result in an increased susceptibility to cancer (Ali-Osman et al., 1997). Individuals with the GSTM1 null genotype and GSTP1 Val/Val allelic variant have significantly higher levels of hydrophobic DNA adducts (Ryberg et al., 2002). These GST variants have been investigated as a risk factor for predisposition to many cancers, but there are very few studies reported on the association of GST polymorphism with AML risk in adults.

Genetic polymorphism in the critical cell-cycle checkpoint control genes may also predispose to cancer. The polymorphism related to changes in the function of the p53 protein is strongly associated with increased risk of developing certain tumors (Wang et al., 1999; Damin et al., 2006; Ignaszak-Szczepaniak et al., 2006). Benzopyrene diol epoxide, an intermediate product of polycyclic aromatic hydrocarbon present in the automobile exhaust and tobacco smoke, is able to inactivate the p53 oncogene by forming benzopyrene diol epoxide–DNA adducts, which are associated with decreased repair capability and reduced apoptotic potential (Dong et al., 2004).

Although these genes have been extensively investigated in solid tumors, little is known about the role of individual gene susceptibility and gene–gene interaction in the development of a heterogeneous disease such as AML. Functional polymorphisms in the genes encoding detoxification enzymes and in coordination with p53 activity can regulate the effect of DNA adduct, which may explain that the differences in leukemia risk are as a result of interplay of genetic susceptibility and exogenous exposure. Current data on the potential associations between AML risk and the genes encoding the enzymes metabolizing xenobiotic compounds are inconsistent. The aim of the present study was to examine the association between genetic polymorphisms in genes involved in the phase I and phase II metabolism of the xenobiotic compounds (EPHX1 and GSTM1, GSTT1, GSTP1) and p53 codon 72 polymorphism with the susceptibility of adult AML.

Materials and Methods

Selection of patients

The study was conducted on samples of 120 patients (77 men, 43 women; age [mean ± standard deviation] at diagnosis was 35.6 ± 17.2 years, range 14–85, with a median age of 32 years) with a confirmed diagnosis of AML collected between 2005 and 2009 at the Department of Hematology, Safdarjung Hospital, New Delhi. A sample size of 120 adult AML was determined on the basis of available literature from an Indian study that reports GSTM1 null polymorphism as 52%. Taking this as the sample proportion (Majumdar et al., 2008), the minimum sample size at 5% level of significance with 6% of the absolute precision and 80% power gave a minimum sample size of 114 to be studied. The diagnosis of AML was made on routinely stained bone marrow aspiration/biopsies and peripheral blood smears and evaluated according to the French-American-British criteria. Peripheral blood samples/bone marrow aspirates obtained from these patients were collected in ethylenediaminetetraacetic acid/heparin for the study. Immunophenotyping studies were also performed at the time of diagnosis as previously described (Bhushan et al., 2010). Exclusion criteria for patients in this study were therapy-related AML cases and patients treated with other than the standard “3 + 7” induction chemotherapy. Details of the treatment protocol are mentioned in our previous publication (Chauhan et al., 2010). Clinical data, including age, sex, white blood cells count, hemoglobin level, platelets count, and presence of organomegaly at diagnosis, were collected prospectively. All in vitro procedures were performed according to protocols approved by the ethical committee of the institute and Safdarjung Hospital. Peripheral blood samples from 202 (126 men, 76 women; age [mean ± standard deviation] was 41.8 ± 17.3 years, range 12–82, with a median age of 45 years) healthy volunteers were used as control samples. Controls had no history of cancer and were not related to the patients. Controls were matched to the cases on the basis of age (±5 years) and sex. Three AML patients and six controls were below 15 years of age. Peripheral blood samples were stored at −70°C until processed for DNA extraction. All the patients and controls were from the North India. The participation of all the patients was voluntary and informed consent was obtained from each patient or accompanying family member.

Genotyping

Blood/bone marrow samples were collected into tubes containing ethylenediaminetetraacetic acid as an anticoagulant. DNA was extracted from 300 μL of peripheral blood/bone marrow samples using a HiPurA™ Blood Genomic Kit (Himedia, Mumbai, India) according to the manufacturer's protocol. A multiplex polymerase chain reaction was used to detect the null genotype of GSTT1 and GSTM1 genes using β-globin as an internal control according to Chen et al. (1996). The polymorphisms in GSTP1 (Ile105Val), p53 (Arg72Pro), and EPHX1 exon 3 (Tyr113His) and exon 4 (His139Arg) were detected using polymerase chain reaction–restriction fragment length polymorphism techniques as described in the literature (Smith and Harrison, 1997; Zhao et al., 2001). Interpretations of gels were done blinded to the case/control status of the sample.

Statistical analysis

The test for Hardy–Weinberg equilibrium (HWE) was determined by χ ² test/Fisher's exact test (two tailed). Associations between genetic variants and AML were assessed by odds ratios (ORs) and 95% confidence intervals (CIs) using a conditional logistic regression model that was adjusted for age and sex. In the regression model, the enter method was used to include the variables under consideration so as to assess their association with AML. ORs were calculated for each genotype compared with the homozygous group of the most frequent allele as well as for the dominant and recessive model for both univariate and multivariable models. Demographic variables associated with smoking, alcohol consumption, dietary habits, and occupational information were not available in all the patients. Therefore, data could not be analyzed on these variables. All analysis were performed using Stata 8.0 software. The characteristics of the study population are presented in Table 1. The main effect of each analyzed polymorphism was assessed for the risk of AML in the univariate and multivariable models (Table 2).

Table 1.

Characteristics of the Acute Myeloid Leukemia Cases and Control Subjects

Characteristics	AML cases n = 120 (%)	Control n = 202 (%)
Sex
Male	77 (64.2)	126
Female	43 (35.8)	76
Age (years)
<50	97 (80)	136
≥50	23 (20)	66
Hb level (g/dL)
≤10	104 (87)
≥10	16 (1)
WBC
≤30,000	60 (50)
≥30,000	60 (50)
Platelets (1 × 10⁹ L⁻¹)
≤50,000	83 (69)
≥50,000	37 (31)
PB blast (%)
≤50	22 (18)
≥50	98 (82)
FAB
M0/M1	17 (14)
M2	41 (31)
M3	7 (6)
M4	18 (15)
M5	20 (17)
Others	17 (14)
Lymphadenopathy	29 (24)
Hepatosplenomegaly	64 (53)
Auer rod present	54 (45)
Follow-up (n = 31)
CR	23 (74)

AML, acute myeloid leukemia; CR, complete remission; FAB, French-American-British; Hb, hemoglobin; PB, peripheral blood; WBCs, white blood cells.

Table 2.

Genotype Frequency and Odds Ratio of GSTM1, GSTT1, GSTP1, EPHX, and p53 codon 72 Polymorphism in Acute Myeloid Leukemia Cases and Controls

		Cases	Control	Conditional logistic regression
Genes	Genotype	n = 120 (%)	n = 202 (%)	Unadjusted OR (95% CI) Adjusted OR (95% CI)	p
GSTM1 ^a	Present	75 (62.5)	105 (52)	1.00
	Null	45 (37.5)	97 (48)	0.62 (0.38–1.0)	0.05
				0.70 (0.43–1.16)	0.17
GSTT1	Present	94 (78.3)	144 (71.3)	1.00
	Null	26 (21.6)	58 (28.7)	0.53 (0.30–0.93)	0.02
				0.58 (0.32–1.05)	0.07
GSTP1	Ile/Ile	59 (49.2)	103 (51)	1.00
	Ile/Val	52 (43.3)	84 (41.8)	1.21 (0.73–1.99)	0.45
				1.17 (0.70–1.96)	0.54
	Val/Val	9 (7.5)	15 (7.4)	1.39 (0.55–3.53)	0.47
				1.13 (0.51–3.39)	0.57
p53	Arg/Arg	32 (26.7)	47 (23.3)	1.00
	Arg/Pro	66 (55.0)	114 (56.4)	0.88 (0.50–1.55)	0.66
				0.71 (0.49–1.68)	0.71
	Pro/Pro	22 (18.3)	41 (20.3)	0.84 (0.41–1.71)	0.64
				0.75 (0.43–1.82)	0.75
EPHX1 exon 3	Tyr/Tyr	50 (41.7)	77 (38.1)	1.00
	Tyr/His	48 (40.0)	85 (42.1)	0.79 (0.47–1.34)	0.39
				0.75 (0.44–1.29)	0.30
	His/His	22 (18.3)	40 (19.8)	0.89 (0.46–1.71)	0.74
				0.78 (0.40–1.53)	0.48
EPHX1 exon 4	His/His	75 (62.5)	115 (56.9)	1.00
	His/Arg	38 (31.7)	78 (38.6)	0.65 (0.39–1.08)	0.10
				0.68 (0.40–1.15)	0.15
	Arg/Arg	7 (5.8)	9 (4.5)	1.10 (0.38–3.16)	0.85
				1.07 (0.36–3.11)	0.89

χ² = 3.37, df = 1, p = 0.08.

CI, confidence interval; OR, odds ratio.

Results

No association was detected between AML and GSTM1, GSTP1, and p53 codon 72 polymorphisms, which occurred at approximately equal frequencies in cases and control (p = 0.087, p = 0.94, and p = 0.76, respectively). However, the frequency of GSTT1 null genotype was higher among controls (28.7%) when compared with AML cases (21.6%) and imparted 42% less risk, which was marginally significant (OR = 0.58, 95% CI: 0.32–1.05, p = 0.07) (Table 2).

The test for HWE was performed and the allelic frequency for GSTP1, p53 codon 72, and EPHX1 Tyr113His and His139Arg polymorphisms in AML cases and controls were in HWE (0.59 vs. 0.70, 0.23 vs. 0.06, 0.09 vs. 0.06, and 0.54 vs. 0.87). The frequency of GSTP1 Ile/Val (43.3%) and Val/Val (7.5%) homozygote genotypes was marginally different in AML patients in comparison to controls (41.8% and 7.4%) but imparted insignificant risk for AML (OR = 1.17, CI: 0.70–1.96, p = 0.57). The frequency of three p53 genotypes Arg/Arg, Arg/Pro, and Pro/Pro found in the AML cases and controls were 26.7% versus 23.3%, 55% versus 56.4%, and 18.3% versus 20.3%, respectively. The distribution showed no significant differences between the AML cases and controls (χ ² = 0.53, p = 0.76). The distribution of combined genotypes of GST and p53 was also studied and no significant association was found between p53 and GSTM1, GSTT1, and GSTP1 in two way gene–gene interactions (χ ² = 2.1, p = 0.31; χ ² = 3.69, p = 0.15; χ ² = 0.67, p = 0.95; data not shown). Further, the OR of combined genotypes was also calculated with different combinations of GST and p53, but the estimated risk was not statistically significant (Table 3).

Table 3.

Combined Frequency of GST and p53 Genotype in Acute Myeloid Leukemia Cases and Controls

Locus	Genotype ^a	Cases	Control	OR (95% CI)	p-Value
GSTT1 GSTM1	Null × Null	12 (10)	40 (19.8)	0.587 (0.190–1.815)	0.356
GSTT1 GSTP1	Null × Ile/Val	8 (6.6)	24 (11.8)	0.663 (0.200–2.201)	0.503
	Null × Val/Val	3 (2)	3 (1.4)	2.24 (2.178–18.597)	0.452
GSTM1 GSTP1	Null × Ile/Val	22 (18.3)	38 (18.8)	1.56 (0.562–4.361)	0.395
	Null × Val/Val	1 (0.8)	8 (3.9)	0.192 (0.016–2.270)	0.191
GSTT1 p53	Null × Arg/Pro	14 (11.6)	31 (15.3)	1.064 (0.279–4.047)	0.927
	Null × Pro/Pro	5 (4.1)	14 (6.9)	0.626 (0.121–3.221)	0.575
GSTM1 p53	Null × Arg/Pro	28 (23.3)	56 (27.7)	1.653 (0.491–5.56)	0.416
	Null × Pro/Pro	9 (7.5)	19 (9.4)	2.026 (0.440–9.32)	0.364
GSTP1 p53	Ile/val × Arg/Pro	28 (23.3)	50 (24.7)	0.774 (0.234–2.56)	0.675
	Ile/val × Pro/Pro	11 (9.1)	18 (8.9)	1.136 (0.251–5.141)	0.868
	Val/Val × Arg/Pro	6 (5)	8 (3.9)	6.425 (0.469–87.89)	0.163
	Val/Val × Pro/Pro	2 (1.6)	2 (0.9)	8.167 (0.341–195.5)	0.195

The combination with lowest risk was taken as a reference.

The genotype frequencies of both the homozygous EPHX1 His113His and Arg139Arg genotypes did not show significant difference between AML patients and controls (Table 2). The frequency of His/His allele was 0.38 in AML cases and 0.41 in controls for Tyr113His polymorphism and 0.22 and 0.24 for Arg/Arg in His139Arg polymorphism, respectively. The estimated adjusted OR showed 22% less chance of risk for AML for His113His genotype (OR = 0.78, CI: 0.40–1.53, p = 0.48) compared with the wild-type Tyr113Tyr genotype. These results suggest a protective effect of the His113His genotype of EPHX1 gene against AML, but there was no statistical significance. Similarly, with EPHX1 His139Arg polymorphism, the risk for the homozygous Arg139Arg genotype was not significant (OR = 1.07, CI: 0.36–3.11, p = 0.89) compared with the wild-type His139His genotype. We divided the combined EPHX1 Tyr113His and His139Arg genotypes into four groups based on in vitro enzyme activities according to classification of Benhamou et al. (1998) to estimate the AML risk, but we did not find any significant difference for very low activity genotype versus other genotype between AML patient and controls (data not shown). The EPHX1, GSTT1, GSTM1, and GSTP1 genotypes were analyzed together to investigate whether a combination of these genetic polymorphisms was associated with the development of AML (Table 4). The presence of combined EPHX1 His113His/GSTP1Ile/Val genotype resulted in an increased fourfold risk compared with reference, but it was not statistically significant (OR = 4.22, CI: 0.992–17.99, p = 0.051). All other combinations of EPHX1 Tyr113His-GST also did not show significant interactions. We also did not observe any significant interaction of EPHX1 and p53 codon 72 polymorphism for the risk of AML (Table 5).

Table 4.

Combined Frequency of GST (GSTT1, GSTM1, and GSTP1) and EPHX1 Genotype in Acute Myeloid Leukemia Cases and Controls

Locus	Genotype ^a	Cases	Control	OR (95% CI)	p-Value
GSTT1 EPHX1 exon 3	Null × Tyr/His	10 (8.3)	26 (12.8)	1.04 (0.316–3.418)	0.948
	Null × His/His	4 (3.3)	8 (3.9)	1.40 (0.280–7.067)	0.678
GSTT1 EPHX1 exon 4	Null × His/Arg	10 (8.3)	27 (13.3)	0.899 (0.294–2.74)	0.853
	Null × Arg/Arg	0	3 (1.4)	—
GSTM1 EPHX1 exon 3	Null × Tyr/His	17 (14.1)	47 (23.2)	0.609 (0.208–1.782)	0.366
	Null × His/His	7 (5.8)	15 (7.4)	1.518 (0.372–6.188)	0.560
GSTM1 EPHX1 exon 4	Null × His/Arg	18 (15)	40 (19.8)	1.28 (0.471–3.50)	0.624
	Null × Arg/Arg	1 (0.8)	4 (1.9)	0.267 (0.019–3.700)	0.325
GSTP1 EPHX1 exon 3	Ile/Val × Tyr/His	15 (12.5)	34 (19.3)	0.823 (0.270–2.507)	0.732
	Ile/Val × His/His	13 (10.8)	12 (5.9)	4.22 (0.992–17.99)	0.051
	Val/Val × Tyr/His	4 (3.3)	6 (2.9)	1.169 (0.140–9.766)	0.885
	Val/Val × His/His	2 (1.6)	4 (1.9)	2.147 (0.174–26.45)	0.551
GSTP1 EPHX1 exon 4	Ile/Val × His/Arg	16 (13.3)	29 (14.3)	1.09 (0.382–3.14)	0.862
	Ile/Val × Arg/Arg	1 (0.8)	3 (1.4)	0.371 (0.022–6.027)	0.486
	Val/Val × His/Arg	3 (2.5)	4 (1.9)	1.817 (0.240–13.74)	0.563
	Val/Val × Arg/Arg	1 (0.8)	2 (0.9)	0.364 (0.016–7.97)	0.521

The combination with lowest risk was taken as a reference.

Bold indicates a p value near to significance.

Table 5.

Combined Frequency of EPHX1 and p53 Genotype in Acute Myeloid Leukemia Cases and Controls

Locus	Genotype ^a	Cases	Control	OR (95% CI)	p-Value
EPHX1 exon 3 × exon 4	Tyr/His × His/Arg	14 (11.6)	33 (16.3)	0.986 (0.323–3.00)	0.981
	Tyr/His × Arg/Arg	4 (3.3)	3 (1.4)	2.28 (0.199–27.19)	0.513
	His/His × His/Arg	5 (4.1)	12 (5.9)	0.647 (0.147–2.831)	0.563
	His/His × Arg/Arg	1 (0.8)	2 (0.9)	0.629 (0.027–14.50)	0.773
EPHX1 exon 3 × p53	Tyr/His × Arg/Pro	29 (24.1)	47 (23.2)	2.21 (0.616–7.99)	0.223
	His/His × Arg/Pro	11 (9.1)	23 (11.3)	0.607 (0.127–2.89)	0.532
	Tyr/His × Pro/Pro	9 (7.5)	16 (7.9)	2.08 (0.420–10.30)	0.369
	His/His × Pro/Pro	2 (1.6)	9 (4.4)	1.68 (0.019–1.44)	0.104
EPHX1 exon 4 × p53	His/Arg × Arg/Pro	22 (18.3)	41 (20.2)	1.397 (0.409–4.765)	0.593
	Arg/Arg × Arg/Pro	2 (1.6)	5 (2.4)	0.383 (0.029–5.040)	0.466
	His/Arg × Pro/Pro	6 (5)	17 (8.4)	1.049 (0.222–4.950)	0.951
	Arg/Arg × Pro/Pro	2 (1.6)	2 (0.9)	1.763 (0.085–36.19)	0.713

The combination with lowest risk was taken as a reference.

Further, when the data were analyzed for dominant and recessive models for all the polymorphisms, GSTP1 was associated with AML risk in the dominant model. However, this was not statistical significant (OR = 1.19, CI: 0.72–1.95, p = 0.48) (Table 6).

Table 6.

Conditional Logistic Regression Analysis for Dominant and Recessive Models

	Recessive model		Dominant model
Locus	Unadjusted OR (95% CI) Adjusted OR (95% CI)	p	Unadjusted OR (95% CI) Adjusted OR (95% CI)	p
GSTP1 ^a	0.77 (0.31–1.91)	0.58	1.23 (0.76–1.99)	0.38
	0.81 (0.32–2.04)	0.66	1.19 (0.72–1.95)	0.48
p53 ^b	1.08 (0.5–1.96)	0.78	0.87 (0.50–1.49)	0.62
	1.04 (0.56–1.91)	0.89	0.89 (0.51–1.55)	0.69
EPHX1 exon 3^c	0.99 (0.54–1.80)	0.99	0.82 (0.51–1.33)	0.43
	1.10 (0.59–2.04)	0.74	0.76 (0.46–1.24)	0.28
EPHX1 exon	0.75 (0.26–2.12)	0.59	0.69 (0.43–1.13)	0.14
4^d	0.79 (0.28–2.26)	0.67	0.72 (0.44–1.19)	0.20

Adjusted for GSTM1, GSTT1, p53, EPHX3, and EPHX4.

Adjusted for GSTM1, GSTT1, GSTP1, EPHX3, and EPHX4.

Adjusted for GSTM1, GSTT1, p53, GSTP1, and EPHX4.

Adjusted for GSTM1, GSTT1, p53, GSTP1, and EPHX3.

Discussion

Environmental exposures and genetic susceptibility plays a significant role in the etiology of AML. DNA adduct–forming compounds present in the environmental xenobiotics and reactive oxygen species generated from the endogenous metabolites of these carcinogens are responsible for the mutation, which consequently leads to the progression of cancer. During last decade, numerous studies have investigated the impact of the genetic background of the host on solid tumors predisposition, but the findings are inconclusive in relation to hematological malignancies. We hypothesize that polymorphisms in genes encoding enzymes involved in the metabolism of carcinogens and DNA repair genes play a role in the development of AML. This is the first study to screen for the EPHX1 and p53 codon 72 polymorphism in Indian AML patients. The present study analyzed the genetic polymorphism of the biotransformation enzyme involved in phase I (EPHX1) and phase II (GST) metabolism of xenobiotic compounds along with p53 codon 72 polymorphism and correlated the data with the risk of AML.

The literature on the association of GST polymorphism with AML risk is conflicting. A recent metaanalysis of 15 published case–control studies in adult AML suggested a significant effect of GSTM1 polymorphism on predisposition to AML, but borderline significance was seen with GSTT1 polymorphism (Das et al., 2009). Aydin-Sayitoglu et al. (2006) also reported significant association of GSTM1 null genotype with adult AML, but no association was seen with GSTT1. In another study with a limited number of samples, both GSTM1 and GSTT1 null genotype were found to be significantly associated with the increased risk of adult AML (Arruda et al., 2001). However, several other studies have reported an absence of association of GSTT1 or GSTM1 null polymorphism with AML (Crump et al., 2000; Naoe et al., 2000).

In the Indian population, Joseph et al. (2004) studied pediatric ALL and reported an association of GSTM1 null genotype, whereas Bajpai et al. (2007) and Mondal et al. (2005) studied chronic myeloid leukemia (CML) patients and found no association of GSTM1 polymorphism. However, GSTT1 polymorphism was found to be significantly associated with risk of CML (Mondal et al., 2005; Bajpai et al., 2007), whereas Majumdar et al. (2008) observed significant association of AML with GSTM1 null genotype but not with GSTT1. The present study did not find evidence for association between GSTM1 and the GSTT1 null genotype, alone or in combination, with increased risk for the occurrence of AML. However, the GSTT1 null genotype was found to confer a protective effect to AML but did not reach statistical significance. Reports are available along these lines from various countries including India for various cancers such as GSTT1 null genotype in breast (García-Closas et al., 1999), bladder (Kim et al., 2002), oral cancer (Anantharaman et al., 2007), and gall bladder cancer (Pandey et al., 2006). Although GSTT1 is primarily involved in detoxification, certain compounds such as dihalomethanes and other chlorinated hydrocarbons are reported to be bioactivated by the GSTT1 pathway (Shimada et al., 1996). Further, individuals with GSTT1 null genotype were found to be more resistant to DNA damage caused by polycyclic aromatic hydrocarbon exposure (Garte et al., 2007). Possible explanations for the dual role (protective and risk) of GSTT1 could be, first, due to different substrate specificity and, second, due to possible interaction with environmental factors, which modify the risk associated with the gene. Jain et al. (2006) earlier reported that interaction of smoking and alcohol with GSTM1 and GSTP1 polymorphisms moderately enhanced the risk of esophageal cancer in an Indian population. Similarly, Anantharaman et al. (2007) found that GSTT1 null genotype was a protective factor for oral cancer among tobacco chewers.

In hematological malignancies, p53 mutation/polymorphism is less frequent compared with solid tumors. Epidemiological and functional studies of the p53 codon 72 polymorphism suggested that Arg and Pro alleles may confer increased susceptibilities to different types of solid cancers (Wang et al., 1999; Damin et al., 2006; Ignaszak-Szczepaniak et al., 2006). Our study did not find any significant association of p53 genotypes with AML risk. This is in concordance with recent studies reporting absence of association of p53 codon 72 polymorphism with AML risk (Ellis et al., 2008; Xiong et al., 2009). However, another study has reported an association of p53 variant with adult T-cell leukemia/lymphoma (Takeuchi et al., 2005). Individuals with GST null polymorphism cannot detoxify carcinogenic metabolites efficiently. Moreover, the risk of cancer increases in these individuals in the presence of p53 codon 72 Pro allele, because it has lower potential to induce apoptosis (Honma et al., 2008). However, in the present study, results suggest that polymorphism of GST and p53 gene alone or in combination did not impart risk for AML development.

This is the first study investigating the association of EPHX1 polymorphism with AML risk in the Indian population. The enzyme activity of EPHX1 differs considerably in the population and the two common coding Tyr113His and His139Arg polymorphisms confer slow and fast enzyme activity, respectively (Maekawa et al., 2003). In the present study, the EPHX1 His113His genotypes posed a decreased but insignificant risk. Similarly, a metaanalysis on chronic obstructive pulmonary disease found a protective effect of EPHX1 Tyr113His polymorphism, whereas Huang et al. (2005) found it as a risk factor in colorector cancer, which increases in smokers (Brøgger et al., 2006). The protective role of EPHX Tyr113His polymorphism has been also observed in studies on lung cancer (To-Figueras et al., 2001; Gsur et al., 2003; Voho et al., 2006). For EPHX1 His139Arg polymorphism, a slight insignificant increased risk was observed for homozygous Arg139Arg genotype, compared with the homozygous wild-type His139His genotype in AML cases. In contrast to our result, for lung cancer, Gsur et al. (2003) reported no relationship of EPHX1 His139Arg polymorphism risk, whereas To-Figueras et al. (2001) found it protective. Similarly, the EPHX1 His139Arg polymorphism was found to be associated with a significantly decreased esophageal cancer risk in the Indian population (Jain et al., 2008). Because of the broad array of epoxide substrates, the activity of EPHX1 might be both procarcinogenic or anticarcinogenic and this may partly explain the divergent role of EPHX1 polymorphism and its association with increased or decreased cancer risk.

The effect of a possible interaction between EPHX1 and GSTs was investigated in a combined analysis. In the present study, a combination of both His113His genotype of EPHX1 gene and 105Ile/105Val of GSTP1 was overrepresented in AML cases. Individuals with slow EPHX1 (His113His) enzyme activity and GSTP1 105Ile/105Val genotype may possess a decreased ability to detoxify xenobiotic compounds, thereby resulting in an increased risk for AML development. Interaction between variants of EPHX1 and GSTP1 enzymes with increased risk of lung and laryngeal carcinoma has been earlier reported by To-Figueras et al. (2001, 2002).

Although the risk associated with environmental factors such as pesticide, benzene exposure, smoking, and alcohol is small, it is possible that the interaction of these risk factors with the genotypes influences the susceptibility toward the occurrence of AML. The present study, however, excluded the analysis of environmental factors because of incomplete information.

Conclusion

This study confirms previous data reporting an absence of association of GSTs and p53 alone or in combination with the risk of adult AML. On the other hand, GSTT1 deletion seems to play a protective role. This is the first study to provide evidence of a combined effect of GSTP1 and EPHX1 on the increased risk of AML, but there was no statistical significance. However, these findings are empirical observations of polymorphism with cancer and the influence of ethnicity, geography, and other environmental factors require further investigation. Further in vivo studies of a combined activity of both enzymes in the bioactivation of xenobiotic compounds may lead to better insights to explain the biology of AML. The present results suggest that further studies on multiple gene–gene interactions may provide a better insight into the complex etiology of AML.

Footnotes

Acknowledgment

This work was supported by a grant under the Council of Scientific and Industrial Research fellowship [e.No.09/630(0014)/2006-EMR-1].

Disclosure Statement

No competing financial interests exist.

References

Ali-Osman

, Akande

, Antoun

, Mao

J.X.

, Buolamwini

1997. Molecular cloning, characterization, and expression in Escherichia coli of full-length cDNAs of three human glutathione S-transferase Pi gene variants. Evidence for differential catalytic activity of the encoded proteins. J Biol Chem, 272:10004–10012.

Anantharaman

, Chaubal

P.M.

, Kannan

, Bhisey

R.A.

, Mahimkar

M.B.

2007. Susceptibility to oral cancer by genetic polymorphisms at CYP1A1, GSTM1 and GSTT1 loci among Indians: tobacco exposure as a risk modulator. Carcinogenesis, 28:1455–1462.

Arruda

V.R.

, Lima

C.S.

, Grignoli

C.R.

, de Mel

M.B.

, Lorand-Metze

, Alberto

F.L.

, Saad

S.T.

, Costa

F.F.

2001. Increased risk for acute myeloid leukaemia in individuals with glutathione S-transferase mu 1 (GSTM1) and theta 1 (GSTT1) gene defects. Eur J Haematol, 66:383–388.

Aydin-Sayitoglu

, Hatirnaz

, Erensoy

, Ozbek

2006. Role of CYP2D6, CYP1A1, CYP2E1, GSTT1, and GSTM1 genes in the susceptibility to acute leukemias. Am J Hematol, 81:162–170.

Bajpai

, Tripathi

A.K.

, Agrawal

2007. Increased frequencies of glutathione-S-transferase (GSTM1 and GSTT1) null genotypes in Indian patients with chronic myeloid leukemia. Leuk Res, 31:1359–1363.

Benhamou

, Reinikainen

, Bouchardy

, Dayer

, Hirvonen

1998. Association between lung cancer and microsomal epoxide hydrolase genotypes. Cancer Res, 58:5291–5293.

Bhushan

, Chauhan

P.S.

, Saluja

, Verma

, Mishra

A.K.

, Siddiqui

, Kapur

2010. Aberrant phenotypes in childhood and adult acute leukemia and its association with adverse prognostic factors and clinical outcome. Clin Exp Med, 10:33–40.

Brøgger

, Steen

V.M.

, Eiken

H.G.

, Gulsvik

, Bakke

2006. Genetic association between COPD and polymorphisms in TNF, ADRB2 and EPHX1. Eur Respir J, 27:682–688.

Chauhan

P.S.

, Bhushan

, Mishra

A.K.

, Singh

L.C.

, Saluja

, Verma

, Gupta

D.K.

, Mittal

, Chaudhry

, Kapur

2010. Mutation of FLT3 gene in acute myeloid leukemia with normal cytogenetics and its association with clinical and immunophenotypic features. Med Oncol, March 31[Epub ahead of print].

10.

Chen

C.L.

, Liu

, Relling

M.V.

1996. Simultaneous characterization of glutathione S-transferase M1 and T1 polymorphisms by polymerase chain reaction in American whites and blacks. Pharmacogenetics, 6:187–191.

11.

Crump

, Chen

, Appelbaum

F.R.

, Kopecky

K.J.

, Schwartz

S.M.

, Willman

C.L.

, Slovak

M.L.

, Weiss

N.S.

2000. Glutathione S-transferase theta 1 gene deletion and risk of acute myeloid leukemia. Cancer Epidemiol Biomarkers Prev, 9:457–460.

12.

Damin

A.P.

, Frazzon

A.P.

, Damin

D.C.

, Roehe

, Hermes

, Zettler

, Alexandre

C.O.

2006. Evidence for an association of TP53 codon 72 polymorphism with breast cancer risk. Cancer Detect Prev, 30:523–529.

13.

Das

, Shaik

A.P.

, Bammidi

V.K.

2009. Meta-analysis study of glutathione-S-transferases (GSTM1, GSTP1, and GSTT1) gene polymorphisms and risk of acute myeloid leukemia. Leuk Lymphoma, 50:1345–1351.

14.

Dong

, Bonala

R.R.

, Suzuki

, Johnson

, Grollman

A.P.

, Shibutani

2004. Mutagenic potential of benzo[a]pyrene derived DNA adducts positioned in codon 273 of the human P53 gene. Biochemistry, 43:15922–15928.

15.

Ellis

N.A.

, Huo

, Yildiz

, Worrillow

L.J.

, Banerjee

, Le Beau

M.M.

, Larson

R.A.

, Allan

J.M.

, Onel

2008. MDM2 SNP309 and TP53 Arg72Pro interact to alter therapy-related acute myeloid leukemia susceptibility. Blood, 112:741–749.

16.

García-Closas

, Kelsey

K.T.

, Hankinson

S.E.

, Spiegelman

, Springer

, Willett

W.C.

, Speizer

F.E.

, Hunter

D.J.

1999. Glutathione S-transferase mu and theta polymorphisms and breast cancer susceptibility. J Natl Cancer Inst, 91:1960–1964.

17.

Garte

, Taioli

, Popov

, Kalina

, Sram

, Farmer

2007. Role of GSTT1 deletion in DNA oxidative damage by exposure to polycyclic aromatic hydrocarbons in humans. Int J Cancer, 120:2499–2503.

18.

Groves

F.D.

, Linet

M.S.

, Devesa

S.S.

1994. Epidemiology of human leukemia. Curr Opin Hematol, 1:321–326.

19.

Gsur

, Zidek

, Schnattinger

, Feik

, Haidinger

, Hollaus

, Mohn-Staudner

, Armbruster

, Madersbacher

, Schatzl

, Trieb

, Vutuc

, Micksche

2003. Association of microsomal epoxide hydrolase polymorphisms and lung cancer risk. Br J Cancer, 89:702–706.

20.

Honma

H.N.

, De Capitani

E.M.

, Perroud

M.W.

Jr. , Barbeiro

A.S.

, Toro

I.F.

, Costa

D.B.

, Lima

C.S.

, Zambon

2008. Influence of p53 codon 72 exon 4, GSTM1, GSTT1 and GSTP1*B polymorphisms in lung cancer risk in a Brazilian population. Lung Cancer, 61:152–162.

21.

Huang

W.Y.

, Chatterjee

, Chanock

, Dean

, Yeager

, Schoen

R.E.

, Hou

L.F.

, Berndt

S.I.

, Yadavalli

, Johnson

C.C.

, Hayes

R.B.

2005. Microsomal epoxide hydrolase polymorphisms and risk for advanced colorectal adenoma. Cancer Epidemiol Biomarkers Prev, 14:152–157.

22.

Ignaszak-Szczepaniak

, Horst-Sikorska

, Sawicka

, Kaczmarek

, Slomski

2006. The TP53 codon 72 polymorphism and predisposition to adrenocortical cancer in Polish patients. Oncol Rep, 16:65–71.

23.

Infante-Rivard

, Labuda

, Krajinovic

, Sinnett

1999. Risk of childhood leukemia associated with exposure to pesticides and with gene polymorphisms. Epidemiology, 10:481–487.

24.

Jain

, Kumar

, Rastogi

, Lal

, Ghoshal

U.C.

, Tiwari

, Pant

M.C.

, Baiq

M.Q.

, Mittal

2006. GSTT1, GSTM1 and GSTP1 genetic polymorphisms and interaction with tobacco, alcohol and occupational exposure in esophageal cancer patients from North India. Cancer Lett, 242:60–67.

25.

Jain

, Tilak

A.R.

, Upadhyay

, Kumar

, Mittal

2008. Microsomal epoxide hydrolase (EPHX1), slow (exon 3, 113His) and fast (exon 4, 139Arg) alleles confer susceptibility to squamous cell esophageal cancer. Toxicol Appl Pharmacol, 230:247–251.

26.

Joseph

, Kusumakumary

, Chacko

, Abraham

, Radhakrishna Pillai

2004. Genetic polymorphism of CYP1A1, CYP2D6, GSTM1 and GSTT1 and susceptibility to acute lymphoblastic leukaemia in Indian children. Pediatr Blood Cancer, 43:560–567.

27.

Kim

W.J.

, Kim

C.H.

, Lee

M.S.

, Oh

B.R.

, Lee

H.M.

, Katoh

2002. GSTT1-null genotype is a protective factor against bladder cancer. Urology, 60:913–918.

28.

Lebailly

, Willett

E.V.

, Moorman

A.V.

, Roman

, Cartwright

, Morgan

G.J.

, Wild

C.P.

2002. Genetic polymorphisms in microsomal epoxide hydrolase and susceptibility to adult acute myeloid leukaemia with defined cytogenetic abnormalities. Br J Haematol, 116:587–594.

29.

Lee

K.M.

, Ward

M.H.

, Han

, Ahn

H.S.

, Kang

H.J.

, Choi

H.S.

, Shin

H.Y.

, Koo

H.H.

, Seo

J.J.

, Choi

J.E.

, Ahn

Y.O.

, Kang

2009. Paternal smoking, genetic polymorphisms in CYP1A1 and childhood leukemia risk. Leuk Res, 33:250–258.

30.

Maekawa

, Itoda

, Hanioka

, Saito

, Murayama

, Nakajima

, Soyama

, Ishida

, Ozawa

, Ando

, Sawada

2003. Non-synonymous single nucleotide alterations in the microsomal epoxide hydrolase gene and their functional effects. Xenobiotica, 33:277–287.

31.

Majumdar

, Mondal

B.C.

, Ghosh

, Dey

, Mukhopadhyay

, Chandra

, Dasgupta

U.B.

2008. Association of cytochrome P450, glutathione S-transferase and N-acetyl transferase 2 gene polymorphisms with incidence of acute myeloid leukemia. Eur J Cancer Prev, 17:125–132.

32.

Miyata

, Kudo

, Lee

Y.H.

, Yang

T.J.

, Gelboin

H.V.

, Fernandez-Salguero

, Kimura

, Gonzalez

F.J.

1999. Targeted disruption of the microsomal epoxide hydrolase gene. Microsomal epoxide hydrolase is required for the carcinogenic activity of 7,12-dimethylbenz[a]anthracene. J Biol Chem, 274:23963–23968.

33.

Mondal

B.C.

, Paria

, Majumdar

, Chandra

, Mukhopadhyay

, Chaudhuri

, Dasgupta

U.B.

2005. Glutathione S-transferase M1 and T1 null genotype frequency in chronic myeloid leukaemia. Eur J Cancer Prev, 14:281–284.

34.

Naoe

, Takeyama

, Yokozawa

, Kiyoi

, Seto

, Uike

, Ino

, Utsunomiya

, Maruta

, Jin-nai

, Kamada

, Kubota

, Nakamura

, Shimazaki

, Horiike

, Kodera

, Saito

, Ueda

, Wiemels

, Ohno

2000. Analysis of genetic polymorphism in NQO1, GST-M1, GST-T1, and CYP3A4 in 469 Japanese patients with therapy-related leukemia/myelodysplastic syndrome and de novo acute myeloid leukemia. Clin Cancer Res, 6:4091–4095.

35.

Pandey

S.N.

, Jain

, Nigam

, Choudhuri

, Mittal

2006. Genetic polymorphisms in GSTM1, GSTT1, GSTP1, GSTM3 and the susceptibility to gallbladder cancer in North India. Biomarkers, 11:250–261.

36.

Ryberg

, Skaug

, Hewer

, Phillips

D.H.

, Harries

L.W.

, Wolf

C.R.

, Ogreid

, Ulvik

, Vu

, Haugen

2002. Genotypes of glutathione transferase M1 and P1 and their significance for lung DNA adduct levels and cancer risk. Carcinogenesis, 18:1285–1289.

37.

Seidegård

, Vorachek

W.R.

, Pero

R.W.

, Pearson

W.R.

1988. Hereditary differences in the expression of the human glutathione transferase active on trans-stilbene oxide are due to a gene deletion. Proc Natl Acad Sci U S A, 85:7293–7297.

38.

Shimada

, Yamazaki

, Oda

, Hiratsuka

, Watabe

, Guengerich

F.P.

1996. Activation and inactivation of carcinogenic dihaloalkanes and other compounds by glutathione S-transferase 5–5 in Salmonella typhimurium tester strain NM5004. Chem Res Toxicol, 9:333–340.

39.

Silveira Vda

, Canalle

, Scrideli

C.A.

, Queiroz

R.G.

, Tone

L.G.

2009. Role of the CYP2D6, EPHX1, MPO, and NQO1 genes in the susceptibility to acute lymphoblastic leukemia in Brazilian children. Environ Mol Mutagen, 51:48–56.

40.

Smith

C.A.

, Harrison

D.J.

1997. Association between polymorphism in gene for microsomal epoxide hydrolase and susceptibility to emphysema. Lancet, 350:630–633.

41.

Takeuchi

, Matsushita

, Tsukasaki

, Takeuchi

, Tomonaga

, Komatsu

, Ikezoe

, Uehara

, Koeffler

H.P.

2005. P53 codon 72 polymorphism is associated with disease progression in adult T-cell leukaemia/lymphoma. Br J Haematol, 131:552–553.

42.

Taningher

, Malacarne

, Izzotti

, Ugolini

, Parodi

1999. Drug metabolism polymorphisms as modulators of cancer susceptibility. Mutat Res, 436:227–261.

43.

To-Figueras

, Gené

, Gómez-Catalán

, Piqué

, Borrego

, Caballero

, Cruellas

, Raya

, Dicenta

, Corbella

2002. Microsomal epoxide hydrolase and glutathione S-transferase polymorphisms in relation to laryngeal carcinoma risk. Cancer Lett, 187:95–101.

44.

To-Figueras

, Gené

, Gómez-Catalán

, Piqué

, Borrego

, Corbella

2001. Lung cancer susceptibility in relation to combined polymorphisms of microsomal epoxide hydrolase and glutathione S-transferase P1. Cancer Lett, 173:155–162.

45.

Voho

, Metsola

, Anttila

, Impivaara

, Järvisalo

, Vainio

, Husgafvel-Pursiainen

, Hirvonen

2006. EPHX1 gene polymorphisms and individual susceptibility to lung cancer. Cancer Lett, 237:102–108.

46.

Wang

Y.C.

, Lee

H.S.

, Chen

S.K.

, Chang

Y.Y.

, Chen

C.Y.

1999. Prognostic significance of p53 codon 72 polymorphism in lung carcinomas. Eur J Cancer, 35:226–230.

47.

Xiong

, Wang

, Liu

, Zhao

, Tian

, Wang

2009. Risk of MDM2 SNP309 alone or in combination with the p53 codon 72 polymorphism in acute myeloid leukemia. Leuk Res, 33:1454–1458.

48.

Zhao

, Lewis

, Gustafson

D.R.

, Wen

W.Q.

, Cerhan

J.R.

, Zheng

2001. No apparent association of GSTP1 A(313)G polymorphism with breast cancer risk among postmenopausal Iowa women. Cancer Epidemiol Biomarkers Prev, 10:1301–1302.