Mol Biol Rep DOI 10.

1007/s11033-012-1715-8

Association between genetic variants in glutathione peroxidase 1 gene and risk of prostate cancer: a meta-analysis
Liu Liwei Zhang Wei Han Ruifa Liu Chunyu

Received: 8 February 2012 / Accepted: 6 June 2012 Springer Science+Business Media B.V. 2012

Abstract To examine the association between glutathione peroxidase 1 (GPx1) gene Pro198Leu polymorphism with the development and progression of prostate cancer. A comprehensive search was conducted to identify all case control studies of GPx1 polymorphisms and prostate cancer. Statistical analysis was performed with the software program Stata, version 11.0, and Review Manage, version 4.2. A total of 7 eligible studies relating the GPx1 polymorphism to the risk of prostate cancer were identied. The results indicated no signicant association between GPx1 polymorphisms and prostate cancer susceptibility in the dominant model (random effects OR 0.75, 95 % CI 0.481.18), recessive model (random effects OR 0.47, 95 % CI 0.221.01) and co-dominant genetic model (random effects OR 0.72, 95 % CI 0.431.21). For the analysis of GPx1 polymorphism and progression of prostate cancer, no signicant association were found in the dominant model (xed effects OR 1.20, 95 % CI 0.951.52), recessive model (xed effects OR 0.69, 95 % CI 0.481.00) and co-dominant genetic model (xed effects OR 0.95, 95 % CI 0.791.15). Eggers test showed that publication bias was not present in all the comparisons. Keywords Polymorphism Prostate cancer GPx1

Introduction Prostate cancer continues to be the most frequently diagnosed neoplasm and a major age-related malignancy. Evidence from
L. Liwei Z. Wei H. Ruifa L. Chunyu (&) Tianjin Institute of Urology, Tianjing Medical University, Tianjin 300211, Peoples Republic of China e-mail: lwluu@yahoo.com.cn

epidemiological, experimental and clinical studies suggests that prostate cancer cells are exposed to an increased oxidative stress [1]. Reactive oxygen species, most notably the hydroxyl radicals, generated endogenously by cellular metabolism are known to cause oxidative DNA damage that has been implicated in prostate carcinogenesis [2]. A number of antioxidant systems are involved in the scavenging of reactive oxygen species, including the glutathione peroxidase (GPx). Up to date, four different GPx izoenzymes have been identied. The most abundant one is GPx1. GPx1 is a selenium-dependent enzyme that is ubiquitously expressed and protects cells against oxidative damage by reducing hydrogen peroxide and wide range of organic peroxides with reduced glutathione [3]. The GPx1 gene is located at chromosome 3p21 and is composed of 2 exons [4]. A genetic polymorphism was dened which is associated with C to T substitution in the exon 2 of GPx1 gene encoding for either proline (Pro) or leucine (Leu) at codon 198(rs1050450) [5]. GPx1 polymorphisms have been investigated in several types of malignancies, such as colorectal cancer [6], brain tumors [7], breast cancer [8] and lung cancer [9]. Also, several studies have been performed to elucidate the association of GPx1 polymorphisms and prostate cancer [2, 5, 1014]. However the results remain inconsistent. Some studies [5, 12, 13] found polymorphism for the GPx1 Pro 198 Leu is associated with the occurrence of prostate cancer. Although, Cheng [10] and Choi [11] described that the polymorphism Pro 198 Leu in GPx1 does not inuence on prostate cancer risk. Also some studies found GPx1 Pro 198 Leu polymorphism associated with the progression of prostate cancer [5, 13]. Although some studies described no signicant difference was found in GPx1 polymorphism between aggressive and non-aggressive prostate cancer patients [2, 11, 12].

123

Mol Biol Rep

The individual studies might have been underpowered to detect the overall results. Given the amount of the accumulated data, a quantitative synthesis of the evidence has been deemed important. In this meta-analysis, we sought to obtain summary estimates for the strength of the postulated genetic association with prostate cancer susceptibility and progression.

Materials and methods Identication and eligibility of relevant articles We considered that all studies examined the association of the GPx1 polymorphism with prostate cancer. The sources included MEDLINE and EMBASE. (latest search update performed in January 2012). The search strategy included the combination of prostate cancer, glutathione peroxidase 1, GPx1, and polymorphism. References of the retrieved articles were also screened. Nonfamilial case control studies were eligible if they determined the distribution for this polymorphism in unrelated patients with prostate cancer and in a concurrent control group of healthy subjects using molecular methods for genotyping. Of the studies with the same or overlapping data by the same investigators, we selected the most recent ones with the most subjects. Data extraction Two investigators (Liu Liwei and Liu Chunyu) independently extracted data. The following information was sought from each report: authors, year of publication, number of genotyped cases and controls, crude odds ratio (ORs), and method of genotyping. The frequencies of the alleles were calculated, for the cases and controls from the corresponding genotype distributions. Meta-analysis The distribution of the genotypes in the control group was tested to determine whether it was in HardyWeinberg equilibrium (P C 0.05). The strength of the association between the prostate cancer risk and GPx1 Pro 198 Leu polymorphism were estimated by the ORs and the 95 % condence intervals (CIs). We estimated the ORs and 95 % CIs of Pro/Pro vs Pro/Leu ? Leu/Leu, Pro/Pro ? Pro/Leu vs Leu/Leu, and the Pro allele frequency, which assumed recessive, dominant, and co-dominant genetic models of the variant. For the association of GPx1 Pro 198 Leu polymorphism and progression of prostate cancer, We also estimated the ORs and the 95 % condence intervals (CIs) of recessive, dominant, and co-dominant genetic models.

Heterogeneity between studies was tested using the Cochrane Q test. If P \ 0.05, heterogeneity was considered statistically signicant. The data were combined using both xed effects (MantelHaenszel) and random effects models (DerSimonian and Laird). Random effects were more appropriate when heterogeneity was present. Eggers test was used to provide a diagnosis of publication bias (Linear regression analysis). All the calculations were performed with computer program Stata, version 9.0, and Review Manage, version 4.2 (Oxford, England, UK). All P values are 2-sides.

Results Eligible studies We identied 7 eligible studies [2, 5, 1014]. The publishing year of the studies was 20072012. Appropriate molecular methods for genotyping were stated in all studies (Table 1), three studies [10, 13, 14] of which was Sequenom, two studies [5, 11] was polymerase chain reaction restriction fragment length polymorphism, and two studies [2, 12] was real-time PCR. The genotype distribution of GPx1 Pro 198 Leu polymorphism in the control group of all eligible studies were in HardyWeinberg equilibrium. Meta-analysis databases For the association of GPx1 Pro 198 Leu polymorphism and risk of prostate cancer, heterogeneity was observed in both recessive model (P \ 0.05), dominant model (P \ 0.05) and co-dominant model comparisons. The results indicated no signicant association between GPx1 polymorphisms and prostate cancer susceptibility in the dominant model (random effects OR 0.75, 95 % CI 0.481.18), recessive model (random effects OR 0.47, 95 % CI 0.221.01) and co-dominant genetic model (random effects OR 0.72, 95 % CI 0.431.21) (Fig. 1). Eggers test showed that publication bias was not present in these 3 comparisons. (P = 0.859, P = 0.949, and P = 0.950). For the analysis of GPx1 Pro 198 Leu polymorphism and progression of prostate cancer, no heterogeneity was observed in both recessive model (P [ 0.05), dominant model (P [ 0.05) and co-dominant model (P [ 0.05) comparisons. No signicant association between GPx1 polymorphisms and progression of prostate cancer were found in both dominant model (xed effects OR 1.20, 95 % CI 0.951.52), recessive model (xed effects OR 0.69, 95 % CI 0.481.00) and co-dominant genetic model (xed effects OR 0.95, 95 % CI 0.791.15) (Fig. 2). Eggers test showed that publication bias was not present in these 3 comparisons (P = 0.849, P = 0.109, and P = 0.358).

123

Mol Biol Rep Table 1 Characteristics of studies included in meta-analysis References Casea Controla Aggressive cancera,b 8/11/4 10/22/17 Non aggressive cancera,c 3/6/1 22/40/23 Genotyping method HardyWeinberg equilibrium in controls

Erdem et al. [2] Kucukgergin et al. [5] Cheng et al. [10]

11/17/5 32/62/40

40/41/10 78/61/20

Real-time PCR Yes PCRRFLP Sequenom Yes Yes

49/53 ProPro/ 371/342 ProPro/ (ProLeu (ProLeu ? LeuLeu) ? LeuLeu) 169/137/50 57/47/19 264/181/42 704/578/109 33/7/8 171/177/41 16/6/2

Abe et al. [11] Arsova-Saranovska 54/17/11 et al. [12] Steinbrecher et al. [13] Choi et al. [14]
a b c

PCRRFLP Real-time PCR Yes Yes Yes

123/108/16 249/213/38

52/38 ProPro/ 68/85 ProPro/ Sequenom (ProLeu ? LeuLeu) (ProLeu ? LeuLeu) Sequenom

PCRRFLP polymerase chain reaction restriction fragment length polymorphism Common homozygote ProPro/heterozygote ProLeu/rare homozygote Leu.Leu Aggressive prostate cancer: Gleason score 710 Non aggressive cancer: Gleason score 26

Fig. 1 Forest plots of odds ratios with 95 % condence intervals for GPx1 Pro 198 Leu polymorphisms and risk of prostate cancer. Top Pro/Pro vs (Pro/Leu ? Leu/Leu) and middle (Pro/Pro ? Pro/Leu) vs Leu/Leu and Pro allele

123

Mol Biol Rep

Fig. 2 Forest plots of odds ratios with 95 % condence intervals for GPx1 Pro 198 Leu polymorphisms and progression of prostate cancer. Top Pro/Pro vs (Pro/Leu ? Leu/Leu) and middle (Pro/Pro ? Pro/Leu) vs Leu/Leu and Pro allele

Comment Meta analysis is a useful strategy for elucidating genetic factors in cancer. Because of the nature of cancer, the effect of one genetic component on the development of the disease can be easily masked by other genetic and environmental factors. A meta analysis is a powerful strategy because it potentially investigates a large number of individuals and can estimate the effect of a genetic factor on the risk of the disease [15]. The present study included data from 7 casecontrol association studies that had investigated the relationship between the GPx1 polymorphism and prostate cancer. The overall results indicated no signicant association between the GPx1 Pro 198 Leu polymorphism and prostate cancer risk. Eggers test showed that publication bias was not present in all the comparisons. Furthermore, in our study we found no association between the GPx1 Pro 198 Leu polymorphism and progression of prostate cancer. GPX is part of the defence system that neutralizes hydrogen peroxide. There are variable reports on the

activity of this enzyme in prostate cancer. Jung et al. [16] found no differences in the antioxidant enzymatic activities of prostatic epithelial cell cultures between benign and malign tissue. In other studies, malignant epithelial cells in prostatic adenocarcinoma have been found to express lower levels of antioxidant enzymes than do benign prostatic epithelium [2, 17]. Also, the results of association of GPx1 Pro 198 Leu polymorphism and GPx activity are inconsistent. Erdem et al. [2] reported there were no relationships between GPx1 polymorphism and erythrocyte GPx activity. On the contrary, Kucukgergin et al. [5] reported that the Leu/Leu genotype was correlated with lower GPx activity among both controls and PCa patients. The explanations for these inconsistent results may be that different dietary and lifestyle factors may inuence GPX enzyme activity. There were some limitations of this meta-analysis. First, because of the lack of sufcient studies, our analysis did not evaluate the relationship of GPx1 Pro 198 Leu polymorphism and prostate cancer in different ethnicity descent in stratication analyses. Second, control subjects consisted

123

Mol Biol Rep

of men who did not have a history of prostate cancer have potential risks of developing prostate cancer, which might lead to misclassication. The results might have been masked in the studies in which the cancer-free control group contained cases of occult prostate cancer. Therefore, larger and well-designed studies are needed to further evaluate the association between GPx1 polymorphism and prostate cancer susceptibility.
Acknowledgments This study was supported by grant from the Natural Science Foundation of Tianjin, China (project number 11JCYBJC13900).

8.

9.

10.

11.

References
1. Khandrika L, Kumar B, Koul S et al (2009) Oxidative stress in prostate cancer. Cancer Lett 282:125136 2. Erdem O, Eken A, Akay C et al (2012) Association of GPX1 polymorphism, GPX activity and prostate cancer risk. Hum Exp Toxicol 31(1):2431 3. Arthur JR (2000) The glutathione peroxidases. Cell Moll Life Sci 57:18251835 4. Bhabak KP, Mugesh G (2010) Functional mimics of glutathione peroxidase: bioinspired synthetic antioxidants. Acc Chem Res 43(11):14081419 5. Kucukgergin C, Gokpinar M, Sanli O et al (2011) Association between genetic variants in glutathione peroxidase 1 (GPx1) gene, GPx activity and the risk of prostate cancer. Minerva Urol Nefrol 63(3):183190 6. Hansen RD, Krath BN, Frederiksen K et al (2009) GPX1 Pro(198)Leu polymorphism, erythrocyte GPX activity, interaction with alcohol consumption and smoking, and risk of colorectal cancer. Mutat Res 664(12):1319 7. Bhatti P, Stewart PA, Hutchinson A et al (2009) Lead exposure, polymorphisms in genes related to oxidative stress, and risk of

12.

13.

14.

15.

16.

17.

adult brain tumors. Cancer Epidemiol Biomarkers Prev 18(6): 18411848 Udler M, Maia AT, Cebrian A et al (2007) Common germline genetic variation in antioxidant defense genes and survival after diagnosis of breast cancer. J Clin Oncol 25(21):30153023 Raaschou-Nielsen O, Srensen M, Hansen RD et al (2007) GPX1 Pro198Leu polymorphism, interactions with smoking and alcohol consumption, and risk for lung cancer. Cancer Lett 247(2): 293300 Cheng TY, Barnett MJ, Kristal AR et al (2011) Genetic variation in myeloperoxidase modies the association of serum a-tocopherol with aggressive prostate cancer among current smokers. J Nutr 141(9):17311737 Abe M, Xie W, Regan MM et al (2011) Single-nucleotide polymorphisms within the antioxidant defence system and associations with aggressive prostate cancer. BJU Int 107(1):126134 Arsova-Saranovska Z, Matevska N, Eken A et al (2009) Glutathione peroxidase 1 (GPX1) genetic polymorphism, erythrocyte GPX activity, and prostate cancer risk. Int Urol Nephrol 41(1): 6370 Steinbrecher A, Meplan C, Hesketh J (2010) Effects of selenium status and polymorphisms in selenoprotein genes on prostate cancer risk in a prospective study of European men. Cancer Epidemiol Biomarkers Prev 19(11):29582968 Choi JY, Neuhouser ML, Barnett M et al (2007) Polymorphisms in oxidative stress-related genes are not associated with prostate cancer risk in heavy smokers. Cancer Epidemiol Biomarkers Prev 16(6):11151120 Liwei L, Chunyu L, Jie L et al (2011) Association between broblast growth factor receptor-4 gene polymorphism and risk of prostate cancer: a meta-analysis. Urol Int 87(2):159164 Jung K, Seidel B, Rudolph B et al (1997) Antioxidant enzymes in malignant prostate cell lines and in primary cultured prostatic cells. Free Radic Biol Med 23(1):127133 Baker AM, Oberley LW, Cohen MB (1997) Expression of antioxidant enzymes in human prostatic adenocarcinoma. Prostate 32(4):229233

123