PARP-1 Val762Ala polymorphism is associated with reduced risk of non-Hodgkin lymphoma in Korean males

Background Poly(ADP-ribose) polymerase-1 (PARP-1) is a nuclear enzyme that plays a role in DNA repair, differentiation, proliferation, and cell death. The polymorphisms of PARP-1 have been associated with the risk of various carcinomas, including breast, lung, and prostate. We investigated whether PARP-1 polymorphisms are associated with the risk of non-Hodgkin lymphoma (NHL). Methods Subjects from a Korean population consisting of 573 NHL patients and 721 controls were genotyped for 5 PARP-1 polymorphisms (Asp81Asp, Ala284Ala, Lys352Lys, IVS13+118A>G, and Val762Ala) using High Resolution Melting polymerase chain reaction (PCR) and an automatic sequencer. Results None of the 5 polymorphisms were associated with overall risk for NHL. However, the Val762Ala polymorphism was associated with reduced risk for NHL in males [odds ratio (OR), 0.62; 95% confidence interval (CI), 0.41-0.93 for CC genotype and OR, 0.84; 95% CI, 0.60-1.16 for TC genotype] with a trend toward a gene dose effect (p for trend, 0.02). The Asp81Asp (p for trend, 0.04) and Lys352Lys (p for trend, 0.03) polymorphisms revealed the same trend. In an association study of PARP-1 haplotypes, the haplotype-ACAAC was associated with decreased risk of NHL in males (OR, 0.75; 95% CI, 0.59-0.94). Conclusion The present data suggest that Val762Ala, Asp81Asp, and Lys352Lys polymorphisms and the haplotype-ACAAC in PARP-1 are associated with reduced risk of NHL in Korean males.


Background
Non-Hodgkin lymphoma (NHL) is the most common hematologic malignancy worldwide. It represents 4% of cancers, and is the fifth commonly diagnosed cancer in the United States [1]. Its rates are over 10/100,000 in the United States, Australia, and Western Europe, while less than 5/100,000 in Southern and Eastern Asia [2].
PARP-1 is a nuclear enzyme that catalyzes the poly (ADP-ribosyl)ation of target proteins in response to DNA damage; it is involved in DNA repair, cell death, proliferation, genome integrity, and modulation of gene transcription [3,4]. PARP-1 is involved in base excision reaction, which repairs DNA damage induced by chemical alterations, reactive oxidative species, and ionizing radiation. In addition, PARP-1 activation mediates apoptosis through the induction of translocation of apoptosis-inducing factor from the mitochondria to the nucleus [5].
Although many studies have searched for the association between PARP-1 polymorphisms and the risk of malignancy, the results are inconsistent in different organs and in different ethnic groups. We performed a population-based, case-control study to identify the association between PARP-1 polymorphisms and NHL risk in Korean subjects.

Study population
This case-control study included NHL patients (n = 573) and cancer-free controls (n = 721). All cases and controls were unrelated Korean individuals; the cases were adults (median age at diagnosis ± S.D 55.0 ± 14.6 years, range 15.0-90.0 years), histologically diagnosed with NHL between 1997 and 2006 at Chonnam National University Hospital (Gwangju, Korea) or Chonnam National University Hwasun Hospital (Hwasun, Korea) [21]. Lymph nodes or extranodal tissues were obtained and histologically diagnosed according to the World Health Organization (WHO) classification [22]. A staging workup included physical examination, laboratory examination, computed tomography scan, bone marrow biopsy, and lumbar puncture. Stages were defined according to the Ann Arbor system [23]. Types of NHL included diffuse large B-cell lymphomas (n = 330), T-cell lymphomas (n = 122), and other lymphomas (n = 121). The controls were frequency matched to patients by age within 5 years, sex, and county of residence. All of the healthy controls have no history of cancer (427 males, 304 females; median age ± S.D. 56.0 ± 15.3 years, range 17.0-82.5 years) ( Table 1). All cases and controls provided informed consent for study participation. The study was approved by the Institutional Review Board of the Chonnam National University Hwasun Hospital in Hwasun, Korea.

Statistical analysis
Hardy-Weinberg equilibrium testing was performed using the chi-square goodness-of-fit test. The Pearson chi-square test was used to examine the genotype distribution differences between cases and controls. Adjusted odds ratios (ORs) were calculated using a logistic regression model that controlled for gender and age and included 95% confidence intervals (CIs). Statistical significance was defined at p values less than 0.05, and all statistical tests were 2-sided. Haplotypes and haplotype frequencies were calculated using the SNPAnalyzer-Pro version 1.8 software (Istech, Goyang, Korea) to analyze associations between haplotypes and NHL risk after stratification by age at diagnosis, sex, and NHL subtype. All calculations were performed using an SPSS software package, version 13.0 (SPSS, Chicago, IL). PARP-1 inter-SNP linkage disequilibrium was calculated using Haploview software version 4.1 http://www.broad.mit.edu/ mpg/haploview for the set of control samples (n = 721).

Results
Characteristics of the study population are summarized in Table 1. There were no differences in the frequency distribution for age (p = 0.33) or sex (p = 0.77) between cases and controls. The control genotype distributions were in agreement with the Hardy-Weinberg equilibrium (p = 0.61 for Asp81Asp, p = 0.31 for Ala284Ala, p = 0.83 for Lys352Lys, p = 0.38 for IVS13 +118 A>G, p = 0.84 for Val762Ala). Haploview version 4.1 determined that the PARP-1 locus contained 1 block of linkage disequilibrium ( Figure. 1). Two groups of polymorphisms were present in the block; one group consisted of Asp81Asp, Lys352Lys, and Val762Ala polymorphism, and the other included Ala284Ala and IVS13 +118A>G polymorphism. Both revealed high intragroup linkage disequilibrium (r 2 , ≥ 0.96), but low intergroup linkage disequilibrium (r 2 , 0.22-0.23). Table 2 summarizes the association study between the 5 polymorphisms and the overall risk for NHL. Analysis revealed no significant associations between polymorphisms and overall NHL risk. Neither the CC genotype of Val762Ala (OR, 0.85; 95%CI, 0.62-1.17) nor the TC genotype (OR, 0.93; 95%CI, 0.72-1.19) were associated with risk for NHL. The genotypes of the remaining 4 polymorphisms were also associated with risk for NHL.
As presented in Additional file 1: table S1, males with the TC genotype of Val762Ala demonstrated a slightly reduced NHL risk (OR, 0.84; 95% CI, 0.60-1.16) compared to those with the TT genotype. And the males with the CC genotype had a significantly reduced risk (OR, 0.62; 95% CI, 0.41-0.93) which revealed a trend toward a gene dose effect (p for trend, 0.02). However, there was no association between Val762Ala polymorphism and NHL risk in females. There was significant interaction seen between sex and the Val762Ala polymorphism (p for interaction, 0.048). Asp81Asp (p for trend, 0.04) and Lys352Lys (p for trend, 0.03) polymorphisms revealed similar trends, although interactions between sex and polymorphism were a borderline significance levels (p for interaction of Asp81Asp, 0.062; p for interaction of Lys352Lys, 0.067). However, PARP-1 polymorphisms did not show associations with the NHL subtypes (Additional file 1: table S2) or age (data not shown).
The distribution of PARP-1 haplotype frequencies and association of haplotypes with risk for NHL are summarized in Additional file 1: table S3. Three common haplotypes (ACAAC, GCGAT, and GTGGT) and several rare ones (with frequency < 5%) were inferred, and rare haplotypes were excluded in the association analyses. None of the haplotypes were associatied with overall risk for NHL. However the haplotype-ACAAC in males was associated with decreased risk for NHL (OR, 0.74; 95% CI, 0.59-0.94) when haplotype-GCGAT was used as a reference.

Discussion
We investigated the association between 5 PARP-1 polymorphisms and NHL risk in a Korean population. Three polymorphisms (Asp81Asp, Lys352Lys, and Val762Ala) were associated with decreased risk of NHL in males. A association testing of haplotype-ACAAC revealed a similar result. However, none of the PARP-1 polymorphisms or haplotypes were associated with overall NHL risk.
Our data appear to support the hypothesis that accumulated DNA damage may lead to enhanced apoptosis during cell division, which could result in protection from development of malignancy with intact apoptotic mechanisms. Concordant findings with the results of this study have been identified in case-control studies of squamous cell carcinoma of the head and neck [15], breast carcinoma [18], and adult glioma [16]. In animal experiments, overexpression of dominant negative PARP-1 prevented in vivo tumor formation by HeLa cells in nude mice due to increased apoptosis of tumor cells [25].
PARP-1 polymorphisms were associated with decreased risk of NHL in males in this study. The sexual dimorphism might be associated with the female sex hormone, 17-β-estradiol. PARP strongly interacts with estrogen receptor α and DNA in the presence of 17-β-estradiol, which leads to inhibition of PARP activation [26]. Sex-dependent response to PARP activation has also been reported in an endotoxin-induced inflammation and vascular change model, female mice were more resistant to endotoxin than male mice [26]. In a neonatal-stroke mouse model, disruption of the PARP-1 gene selectively protected male mice against brain injury [27]. Also in an adult-stroke mouse model, female neuronal nitric oxide synthase knockout (nNOS-/-) mice were more severely damaged after middle cerebral artery occlusion than wild-type females. In addition, male nNOS-/-littermates were protected [28]. In a case-control study, PARP-1 polymorphism has been associated with reduced risk of adult glioma in men [16]. Inconsistency of the association between PARP-1 polymorphism and malignancy risk might result from differences in genetic background, environmental factors, organ specificity, and study sample size. For example, the frequency of the Val762Ala genetic polymorphism differs between ethnic groups. In the present study of Koreans, the frequency of the C allele is 0.448 which is consistent with that observed in previous studies of Koreans (0.444) and Han Chinese (0.389) [12,29]. However, the frequency is 0.145 in Caucasian-Americans and 0.045 among African-Americans [8]. PARP-1 expression levels vary in carcinomas of different organs. It is high in lymphoma [30] and endometrial carcinoma [31], and low in breast carcinoma [32,33] and laryngeal carcinoma [34].
In this present study, haplotype-ACAAC was also associated with reduced risk of NHL in males, although there was no association between any haplotypes and overall NHL risk. It agrees with the results that the A allele of Asp81Asp, A allele of Lys352Lys, and C allele of Val762Ala are significantly associated with decreased risk of NHL.
As the sample size of this study is not sufficiently large and is restricted to Korean population, the present data should be validated in larger samples and in other ethnic groups. Although PARP-1 polymorphisms are associated with decreased risk of NHL in the present study, we could not exclude the possibility that other un-typed variants located in or near the PARP-1 locus might be associated with reduced risk of NHL. Additional surveys examining other variants around the locus are required to resolve this issue. A gene-gene interaction study examining genes involved in apoptosis or DNA repair might allow for a more comprehensive perspective.

Conclusion
In a polulation-based, case-control study to identify the association between polymorphisms in PARP-1 and NHL risk in Koreans, we found that Val762Ala, Asp81Asp, and Lys352Lys polymorphisms and the haplotype-ACAAC in PARP-1 were associated with decreased risk of NHL in males.