REPRODUCTIONRESEARCH

Polymorphisms in double-strand breaks repair are associated with impaired fertility in Chinese population

Guixiang Ji, Lifeng Yan1,, Wei Liu1, Cong Huang1, Aihua Gu1 and Xinru Wang1 Nanjing Institute of Environmental Sciences/Key Laboratory of Pesticide Environmental Assessment and Pollution Control, Ministry of Environmental Protection, Nanjing 210042, People’s Republic of China and 1State Key Laboratory of Reproductive Medicine, School of Public Health, Institute of Toxicology, Nanjing Medical University, Nanjing 210029, People’s Republic of China Correspondence should be addressed to A Gu; Email: [email protected]; X Wang; Email: [email protected]

G Ji and L Yan contributed equally to the study and they should be regarded as joint first authors

Abstract

The DNA double-strand breaks (DSBs) repair pathway plays a critical role in repairing double-strand breaks, and genetic variants in DSBs repair pathway genes are potential risk factors for various diseases. To test the hypothesis that polymorphisms in DSBs genes are associated with susceptibility to male infertility, we examined 11 single nucleotide polymorphisms in eight key DSBs genes (XRCC3, XRCC2, BRCA2, RAG1, XRCC5, LIG4, XRCC4 and ATM) in 580 infertility cases and 580 controls from a Chinese population-based case–control study (NJMU Infertility Study). Genotypes were determined using the OpenArray platform, and sperm DNA fragmentation was detected using the TUNEL assay. The adjusted odds ratio (OR) and 95% CI were estimated using logistic regression. The results indicate that LIG4 rs1805388 (Ex2C54COT, Thr9Ile) T allele could increase the susceptibility to male infertility (adjusted ORZ2.78; 95% CI, 1.77–4.36 for TT genotype; and adjusted ORZ1.58; 95% CI, 1.77–4.36 for TC genotype respectively). In addition, the homozygous variant genotype GG of RAG1 rs2227973 (AOG, K820R) was associated with a significantly increased risk of male infertility (adjusted OR, 1.44; 95% CI, 1.01–2.04). Moreover, linear regression analysis revealed that carriers of LIG4 rs1805388 or RAG1 rs2227973 variants had a significantly higher level of sperm DNA fragmentation and that T allele carriers of LIG4 rs1805388 also had a lower level of sperm concentration when compared with common homozygous genotype carriers. This study demonstrates, for the first time, to our knowledge, that functional variants of RAG1 rs2227973 and LIG4 rs1805388 are associated with susceptibility to male infertility. Reproduction (2013) 145 463–470

Introduction packaging, apoptosis, oxidative damage, replication errors and genetically programmed processes such as Male infertility is a common reproductive disorder and a recombination in meiosis, and variable (diversity) join- major cause of conception failure in humans that is ing (V(D)J) recombination (Gellert et al. 1999, Dresser responsible for 40–60% of infertility cases (Esteves et al. 2011, Ji et al. 2012). Though the aetiologies of w50% of 2000, Khanna & Jackson 2001, Agarwal & Said 2003, cases are still poorly understood (De Kretser & Baker Zha et al. 2007, Lieber 2010). 1999), sperm DNA damage is thought to be a cause of In mammalian cells, DSBs are mostly repaired by the male infertility and has been an area of intense research DSBs repair system. The repair of DSBs is a complex throughout the past decade (Sakkas et al. 1999). It has process. The first response is sensing a DSB. This process been demonstrated that DNA damage in human is mediated by the MRN (MRE11A (MRE11), RAD50 and spermatozoa is associated with poor semen quality, NBN (NBS1)) complex and ATM kinase (Lee & low fertilisation rates, impaired preimplantation Paull 2005). Once DSBs are recognised, cells can repair development, increased abortion and an elevated a break lesion in one of the two ways: non-homologous incidence of disease in the offspring (e.g. childhood end-joining (NHEJ) and cancer) (Ji et al. 1997, Borini et al. 2006, Cohen-Bacrie (HR). HR is of high fidelity and error free, whereas et al. 2009). Among the many types of DNA damage, NHEJ is potentially error prone (Khanna & Jackson DNA double-stranded breaks (DSBs) are the most severe 2001). In the NHEJ pathway, the XRCC5/XRCC6 protein (Khanna & Jackson 2001) and are a result of ionising dimer initially binds to broken ends of DNA and radiation, chemotherapeutic drugs, defective chromatin activates DNA-PKcs (Abe et al. 2008). Ligation is carried

q 2013 Society for Reproduction and Fertility DOI: 10.1530/REP-12-0370 ISSN 1470–1626 (paper) 1741–7899 (online) Online version via www.reproduction-online.org Downloaded from Bioscientifica.com at 09/24/2021 08:34:38PM via free access 464 G Ji, L Yan and others

Table 1 Distribution of selected characteristics between cases and Genetic variants of genes involved in the DSBs repair fertile controls. pathway have been extensively studied in various Cases Controls diseases (Han et al. 2004, Auranen et al. 2005, Figueroa (nZ580); (nZ580); et al. 2007, Liu et al. 2007, Tseng et al. 2009). However, Variables n (%) n (%) P a whether genetic variants in the DSBs repair pathway Age genes affect susceptibility to male infertility remains R20 and !30 326 (56.3) 299 (51.6) 0.067 largely unknown. Thus, the aim hereby was to R30 and !40 246 (42.4) 278 (47.9) R40 and !50 8 (1.4) 3 (0.5) investigate the relationship between 11 functional Age (meanGS.D.) 28.2G3.3 28.1G3.2 0.600 polymorphisms in eight key genes (XRCC3, XRCC2, Smoking status BRCA2, RAG1, XRCC5, LIG4, XRCC4 and ATM) of the Never 272 (46.9) 306 (52.8) 0.046 DSBs repair pathway and the risk of male infertility. In Ever 308 (53.1) 274 (47.2) b Pack-years (meanGS.D.) 4.7G4.4 4.1G4.3 0.019 addition, we also evaluated their influence on DNA Drinking status damage detected with TUNEL assay. Never 500 (86.2) 504 (86.8) 0.731 Ever 80 (13.8) 76 (13.1) BMI (kg/m2) !25 423 413 0.246 Results 25–29.9 149 153 R30 8 14 2 Characteristics of the study population BMI (meanGS.D.) (kg/m ) 23.3G0.1 23.4G0.1 0.825 Abstinence time (days) The frequency distributions of selected characteristics of 3–5 304 (52.4) 321 (55.3) 0.272 the case patients and control subjects are presented in O5 276 (47.6) 259 (44.6) Semen parameters (meanGS.D.) Table 1. Briefly, cases and controls were well matched Concentration (!106/ml) 73.6G49.37 115.2G76.58 !0.001 for age, drinking status, BMI and abstinence time Motility (%) 37.9G13.49 65.3G12.52 !0.001 (PO0.05). However, there were more smokers in the Volume (ml) 2.78G1.44 2.80G1.37 0.809 Z Sperm DNA 20.98G15.30 ND case group compared with the control group (P 0.046). fragmentation (%) Among smokers, cases also reported significantly greater cigarette consumption than controls, as assessed by the ND, not detected. Z aP values were derived from the c2 test for categorical variables (age, mean number of pack-years (P 0.019). Moreover, there BMI, smoking and drinking status) and t-test for continuous variables was a significant difference between cases and controls (age, pack-years and BMI). bAmong ever smokers. with respect to semen concentration and sperm motility (P!0.001), but not semen volume (PZ0.809). out by the LIG4–XRCC4 complex to complete the repair (Grawunder et al. 1997). This pathway is also required Individual single nucleotide polymorphism, –gene for the completion of RAG-initiated lymphocyte-specific interaction and susceptibility to male infertility V(D)J recombination (Lieber 2010). In HR pathway, Primary information on the 11 functional single RAD51 mediates the crucial step of searching for a nucleotide polymorphisms (SNPs) found in the Chinese homologous duplex template. RAD51 then interacts population in the HapMap database is shown in Table 2. with other important repair , including BRCA1, All tested genotypes were in Hardy–Weinberg equili- BRCA2, XRCC2 and XRCC3 at various steps of HR (van brium in control groups (PO0.05). The main effect Gent et al. 2001, San Filippo et al. 2008). models for all the genotypes are presented in Table 3.

Table 2 Primary information for 11 genotyped SNPs in DNA DSBs repair pathway genes. Polymorphisms

Base (amino acid) P value for Symbol Gene name ID no. change MAFa HWE test XRCC3 X-ray repair cross-complementing group 3 rs861539 COT (T241M) 0.058 0.923 rs1799794 GOA(50-UTR) 0.478 0.156 rs1799796 AOT (IVS6-14) 0.289 0.317 XRCC2 X-ray repair cross-complementing group 2 rs3218385 T/G (50-UTR) 0.155 0.706 BRCA2 Breast cancer 2 rs1799943 GOA(50-UTR) 0.279 0.410 rs15869 AOC(30-UTR) 0.256 0.135 RAG1 Recombination activating gene 1 rs2227973 AOG (K820R) 0.465 0.362 XRCC5 X-ray repair cross-complementing group 5 rs1051685 AOG(30-UTR) 0.070 0.868 LIG4 Ligase 4 rs1805388 COT (T9I) 0.261 0.756 XRCC4 X-ray repair cross-complementing group 4 rs1805377 AOG (splice site) 0.291 0.208 ATM Ataxia-telangiectasia mutated rs189037 GOA(50-UTR) 0.389 0.142 MAF, minor allele frequency; HWE, Hardy–Weinberg equilibrium. aMinor allele frequency in the Chinese (CHB, Han-Chinese in Beijing, China) population, as reported in dbSNP database.

Reproduction (2013) 145 463–470 www.reproduction-online.org

Downloaded from Bioscientifica.com at 09/24/2021 08:34:38PM via free access DSBs repair gene SNPs and male infertility 465

Table 3 Associations between SNPs in DNA DSBs repair pathway genes and risk of male infertility. Homozygotes (for common allele) Heterozygotes Homozygotes (for rarer allele)

a a b SNP Cases Controls Cases Controls OR (95% CI) Cases Controls OR (95% CI) Ptrend XRCC3: rs861539 511 498 62 75 0.79 (0.55–1.12) 4 3 1.27 (0.28–5.69) 0.356 XRCC3: rs1799794 155 165 292 271 1.09 (0.83–1.44) 128 141 0.92 (0.66–1.27) 0.899 XRCC3: rs1799796 268 262 254 245 0.95 (0.75–1.22) 56 69 0.75 (0.50–1.10) 0.419 XRCC2: rs3218385 420 432 141 134 1.07 (0.81–1.40) 12 9 1.35 (0.56–3.24) 0.411 BRCA2: rs1799943 254 268 271 256 1.07 (0.84–1.36) 59 52 1.14 (0.76–1.72) 0.280 BRCA2: rs15869 307 330 216 203 1.10 (0.86–1.41) 52 43 1.25 (0.81–1.93) 0.149 RAG1: rs2227973 175 206 293 285 1.17 (0.90–1.52) 106 84 1.44 (1.01–2.04) 0.025 XRCC5: rs1051685 478 487 93 84 1.11 (0.80–1.53) 5 4 1.25 (0.33–4.69) 0.420 LIG4: rs1805388 249 330 257 213 1.58 (1.23–2.01) 68 32 2.78 (1.77–4.36) !0.001 XRCC4: rs1805377 273 289 248 229 1.12 (0.88–1.43) 64 58 1.14 (0.77–1.69) 0.264 ATM: rs189037 206 193 288 303 0.84 (0.65–1.08) 82 79 0.92 (0.63–1.32) 0.665 Data in boldface represent P!0.05. aOdds ratio was age, cigarette smoking, alcohol use and BMI. bFalse discovery rate (FDR) corrected P value.

Overall, two SNPs exhibited significant associations with were negatively associated with sperm concentration the risk of male infertility. Carriers with RAG1 rs2227973 (bZK0.164, PZ0.006). homozygous variant genotype (GG) exhibited a signi- ficantly increased risk of male infertility (adjusted odds ratio (OR), 1.44; 95% CI, 1.01–2.04) when compared Discussion with the common homozygous genotype (AA). More- Accumulating evidence demonstrates that the DSBs over, LIG4 rs1805388 variant genotypes (TT or CT) were repair pathway plays a critical role in repairing DSBs, associated with significantly increased risks of male and polymorphisms in DSBs repair pathway genes are infertility compared with the common homozygous CC potential risk factors for various diseases. Considering genotype (adjusted OR, 1.58; 95% CI, 1.23–2.01 for CT; the fact that inadequacy or defects in sperm DSBs repair and adjusted OR, 2.78; 95% CI, 1.77–4.36 for TT can lead to large-scale loss of genetic information and respectively). Trend c2 test showed that the male affect the germ line for generations (Lewis & Aitken infertility risk was significantly increased in a dose- 2005), we hypothesised that polymorphisms in DSBs ! dependent manner (Ptrend 0.001). repair pathway genes were involved in susceptibility to We further evaluated the combined effects of the two male infertility. high-risk genotypes on male infertility risk by summing In this study, we assessed the effects of 11 functional the unfavourable genotypes of RAG1 rs2227973 and SNPs in DSBs repair genes on male infertility and found LIG4 rs1805388. As shown in Table 4, 46.3% of the cases that two nonsynonymous SNPs of RAG1 rs2227973 and and 32.5% of the controls had variant genotypes at both LIG4 rs1805388 were associated with significantly loci (RAG1 KR/RR and LIG4 TI/II), and carriers of both loci increased male infertility risks, even after adjustment had an increased risk of male infertility (adjusted OR, for multiple testing. In parallel, strong associations were 1.63; 95% CI, 1.20–2.21). However, no significant detected between these two significant SNPs and sperm DNA fragmentation or sperm concentration. These additive interaction between the variant genotypes of findings support our previous hypothesis, and, to the rs2227973 and rs1805388 was detected for the develop- best of our knowledge, this is the first systematic ment of male infertility (P Z0.130). interaction pathway-based study on the role of polymorphisms of DSBs repair pathway genes on male infertility risk. Association of sperm concentration and sperm DNA The LIG4 protein plays a key role in NHEJ with regard fragmentation with gene polymorphisms to ligating and filling the gap in the broken DNA strands

Considering the important role of DSBs repair pathway Table 4 Interaction of RAG1 R820K and LIG4 T9I for male infertility genes in maintaining DNA integrity, we further eval- risk. uated the associations of sperm concentration and sperm RAG1 LIG4 Cases, Controls, OR DNA fragmentation with all the tested gene polymorph- K820R T9I n (%) n (%) (95% CI)a isms (Table 5). Despite the high variation in DNA KK TT 116 (20.21) 148 (25.74) 1.00 fragmentation among the patients (Fig. 1), we observed KK TI/II 59 (10.28) 58 (10.09) 1.17 (0.75–1.80) that the RAG1 rs2227973 and the LIG4 rs1805388 KR/RR TT 133 (23.17) 182 (31.65) 0.84 (0.60–1.16) KR/RR TI/II 266 (46.34) 187 (32.52) 1.63 (1.20–2.21) variants were positively associated with higher levels of b Pinteraction 0.130 sperm DNA fragmentation (all bO0andP!0.05). aOdds ratio was age, cigarette smoking, alcohol use and BMI. bP for Moreover, we found that LIG4 rs1805388 variants additive interaction. www.reproduction-online.org Reproduction (2013) 145 463–470

Downloaded from Bioscientifica.com at 09/24/2021 08:34:38PM via free access 466 G Ji, L Yan and others

Table 5 Effects of SNPs in DNA DSBs repair pathway genes on sperm concentration and sperm DNA fragmentation. Sperm concentration Sperm DNA fragmentation

SNP ba 95% CI Pb ba 95% CI Pb XRCC3: rs861539 K0.085 K0.252, 0.082 0.318 K0.082 K0.200, 0.036 0.173 XRCC3: rs1799794 K0.011 K0.201, 0.179 0.909 0.049 K0.084, 0.182 0.470 XRCC3: rs1799796 0.065 K0.052, 0.183 0.277 K0.032 K0.152, 0.089 0.605 XRCC2: rs3218385 K0.006 K0.182, 0.169 0.943 0.013 K0.110, 0.137 0.831 BRCA2: rs1799943 K0.004 K0.174, 0.166 0.964 0.036 K0.084, 0.156 0.554 BRCA2: rs15869 0.103 K0.027, 0.232 0.119 K0.002 K0.122, 0.118 0.973 RAG1: rs2227973 K0.149 K0.332, 0.035 0.113 0.172 0.002, 0.342 0.047 XRCC5: rs1051685 K0.012 K0.184, 0.181 0.984 K0.026 K0.154, 0.102 0.691 LIG4: rs1805388 K0.164 K0.282, K0.047 0.006 0.216 0.046, 0.386 0.013 XRCC4: rs1805377 K0.112 K0.281, 0.056 0.191 K0.010 K0.177, 0.158 0.909 ATM: rs189037 0.150 K0.019, 0.320 0.082 K0.013 K0.132, 0.106 0.829 b, regression coefficient. Data in boldface represent P!0.05. aAll analyses were done using linear regression models, adjusted for age, smoking status, drinking status, abstinence time and BMI. bFalse discovery rate (FDR) corrected P value. with XRCC4 to complete repair (Karran 2000). Northern Regarding RAG1, it initiates V(D)J recombination with blots have indicated that LIG4 is present at higher levels RAG2 by introducing DSBs at recombination signal in testis than in other tissues (Wei et al. 1995). In sequences (McBlane et al. 1995). It then aids in the mammalian cells, the LIG4 protein exists as a pre- DNA ligation processing along with the additional adenylate complex (Robins & Lindahl 1996), with a components of the NHEJ pathway (Kale et al. 2001). The N-terminal conserved ligase domain and two C-terminal RAG1 K820R polymorphism results in a nonsynonymous BRCT domains (Girard et al. 2004). The LIG4 rs1805388 amino acid change from lysine to arginine at C-terminal (T9I) polymorphism results in a nonsynonymous amino acid change from threonine to isoleucine in the N-terminal region of the LIG4 protein that is essential 300 300 /ml) /ml) 6 for its activity (Grawunder et al. 1998). It has been 6 demonstrated that this substitution in the N-terminal region of LIG4 mildly but reproducibly reduces adenyla- 200 200 tion and ligation activities (two- to threefold) and increases the hydrophobic nature of this region of the 100 100 protein (O’Driscoll et al. 2001, Girard et al. 2004). So, it is conceivable that the LIG4 T9I alteration may alter the Sperm (× 10 concentration interactions of LIG4 with XRCC4, or with other 0 0 components of the NHEJ pathway. Consistent with the GG GA AA CC CT TT known functionality of the polymorphism, we found that (n = 175)(n = 293)(n = 106) (n = 249)(n = 257) (n = 68) the LIG4 9I variants were associated with significantly RAG1: rs2227973 LIG4: rs1805388 decreased DNA repair capacity and increased risk of 100.00 100.00 male infertility. Some previous studies have found evidence that LIG4 80.00 80.00 polymorphism is a risk factor for certain diseases. The LIG4 T9I polymorphism was initially identified in an LIG 60.00 60.00

IV syndrome patient characterised by microcephaly, 40.00 40.00 characteristic facial features, growth retardation, developmental delay and immunodeficiency (Roddam 20.00 20.00 Sperm (%) DNA fragmentation et al. 2002, Ben-Omran et al. 2005), thus suggesting a Sperm (%) DNA fragmentation Sperm (× 10 concentration deleterious effect of the polymorphisms upon LIG4 0.00 0.00 function. Recently, Yin et al. (2012) reported that LIG4 GG GA AA CC CT TT (n = 175)(n = 293)(n = 106) T9I polymorphism had a significant effect on the risk of (n = 249) (n = 257) (n = 68) RAG1: rs2227973 developing severe radiation pneumonitis among non- LIG4: rs1805388 small-cell lung cancer (NSCLC) patients. Moreover, a significant association between LIG4 T9I polymorphism Figure 1 Box-and-whisker plots for sperm DNA fragmentation in study and NSCLC susceptibility was found in a Chinese case– subjects, divided into three groups according to genotype. The boxes represent the 25th and 75th percentiles; whiskers are lines extending control study with an overall adjusted OR of 1.64 (95% from each end of the box covering the extent of the data on 1.5!inter- CI 1.03–2.62) (Tseng et al. 2009). These findings indicate quartile range. The median value is denoted by the line that bisects the that diseases may share some fraction of genetic factors. boxes. Circles represent the outlier values.

Reproduction (2013) 145 463–470 www.reproduction-online.org

Downloaded from Bioscientifica.com at 09/24/2021 08:34:38PM via free access DSBs repair gene SNPs and male infertility 467 domain, a domain that binds double-stranded DNA in the promoter of the ATM gene, affects the expression cooperatively and nonspecifically and mediates coding- of ATM mRNA by differentially binding to activator end contacts (Arbuckle et al. 2001, Mo et al. 2001). protein 2a (AP-2a; Chen et al. 2010), which then affect Consistent with the functionality of the RAG1 K820R the process of sensing DNA damage. In this case–control polymorphism, we found that the RAG1 K820R variant study, we did not find any association between these was associated with significantly decreased DNA recom- SNPs and risk of male infertility, possibly due to bining capacity and increased risk of male infertility. insufficient statistical power. Thus, large-scale studies To date, only a few studies have demonstrated that are needed to confirm our results in the future. RAG1 K820R polymorphism is a risk factor for a Several strengths of our study should be acknowl- particular disease. One study reported that homozygotes edged. i) A pathway-based approach was conducted to for the RAG1 820 R missense substitution had a 2.7-fold estimate the effects of selected gene polymorphisms on increased risk of non-Hodgkin lymphoma and presented male infertility. ii) All tested SNPs were in Hardy– data suggesting a gene dosage effect (Hill et al. 2006). Weinberg equilibrium in controls. iii) All cases and Another study also found that carriers of variant RAG1 controls were racially homogeneous (all Han-Chinese) 820R allele exhibited a significantly increased risk for and well matched with regard to age and drinking status, bladder cancer (Wu et al. 2006). These findings raise the which minimises potential confounding bias. iv) The possibility that decreased RAG1 activity plays a role in genotyping method for the study was standardised, and human diseases and indicates that our findings are quality control samples indicated high reproducibility of biologically plausible. Thus, studies exploring the effects the genotyping results. However, there are some of the RAG1 K820R polymorphism on susceptibility to limitations of our study. i) We were unable to explore additional conditions and diseases, including male the specific biological mechanism of how RAG1 infertility, should be conducted in the future. rs2227973 and LIG4 rs1805388 polymorphisms affect In addtion to the two SNPs discussed earlier, we male infertility. ii) The moderate sample size might not examined another nine SNPs in six DSB repair genes, be sufficient for us to adequately detect a small effect including XRCC3 (rs861539, rs1799794 and from very low-penetrance SNPs and evaluate gene–gene rs1799796), XRCC2 rs3218385, BRCA2 (rs1799943 interactions between the two high-risk SNPs. iii) We and rs15869), XRCC5 rs1051685, XRCC4 rs1805377 used the TUNEL assay to detect sperm DNA fragmenta- and ATM rs189037. XRCC3 belongs to the RecA/Rad51- tion. However, the TUNEL assay measures both single- related protein family, which is involved in HR repair and double-strand DNA fragmentation by detecting a pathway. XRCC3 rs861539 SNP changes the amino acid definitive end point (free 3-OH terminus). So, using this from a neutral hydrophilic residue with a hydroxyl group method might influence interpretation of the results. to a hydrophobic one with a methyl sulphur group, Among other commonly used tests to measure sperm which may result in a substantial change in protein DNA damage, such as sperm chromatin structure assay, structure and function (Mittal et al. 2012). The XRCC3 acridine orange test, alkaline and neutral comet assay, rs1799794 SNP, which is located in the 50-UTR region, only the neutral comet assay is specific for identification causes the loss of an upstream open reading frame of double-stranded DNA breaks (Olive et al. 1991). (Fachal et al. 2012) and then affects . Interestingly, it has been reported that the output of XRCC3 rs1799796 is non-coding. XRCC2 is an essential TUNEL and neutral comet assays are correlated to some part of the HR repair pathway and a functional candidate degree (Caglar et al. 2007). So, we believe that the for involvement in tumour progression (Mohindra et al. potential misclassification and bias, if any, is unlikely to 2002, Thacker & Zdzienicka 2004). The rs3218385 SNP be substantial. in XRCC2 is located in the 50-UTR, which might affect In conclusion, of the 11 potential functional poly- gene expression. The BRCA2 gene functions as a tumour- morphisms investigated here, we provide the first suppressor gene and plays a pivotal role in the control of evidence that the LIG4 rs1805388 and RAG1 the HR repair pathway (Venkitaraman 2002). BRCA2 rs2227973 polymorphisms contribute to the risk of rs1799943 SNP might modulate BRCA2 transcript levels male infertility. These findings may be helpful in as indicated by in silico analysis (Rajasekaran et al. improving our understanding of the aetiology of male 2008). The BRCA2 rs15869 SNP is located in the 30-UTR, infertility. Future large-scale studies with ethnically which might affect mRNA stability and translation. diverse populations and functional evaluations are XRCC5 rs1051685, located in an exonic splice enhancer needed to validate our findings. sequence, results in exon skipping, errors in alternative splicing patterns and affects mRNA stability and translation (Cibeira et al. 2011). XRCC4 rs1805377 Materials and Methods may have functional significance as the nucleotide change from A to G potentially abolishes an acceptor Subjects and sample collection splice site at exon 8 (Dore et al. 2006) and subsequently The study was approved by the Ethics Review Board of Nanjing affects mRNA stability. The ATM rs189037 SNP, located Medical University. All studies involving human subjects were www.reproduction-online.org Reproduction (2013) 145 463–470

Downloaded from Bioscientifica.com at 09/24/2021 08:34:38PM via free access 468 G Ji, L Yan and others conducted under full compliance with government policies Diego, CA, USA) according to the manufacturer’s protocol. and the Helsinki Declaration. A total of 1657 infertile patients, Briefly, sperm were washed and resuspended in 2% paraf- diagnosed with unexplained male factor infertility, were drawn ormaldehyde for 30 min at room temperature. After being from the Centre of Clinical Reproductive Medicine (NJMU rinsed with PBS, samples were resuspended in permeabilisa- infertility centre) between April 2005 and March 2009. All tion solution (0.2% Triton X-100, 0.1% sodium citrate) for patients had undergone complete historical and physical 10 min on ice. Fifty millilitres of TUNEL reagent were then examinations. Those with a history of orchitis, cryptorchidism, added to the sample. For each batch, samples that were not obstruction, congenital bilateral absence of vas deferens, treated with the Tdt enzyme were used as negative controls, cytogenetic abnormalities and Y microdeletions and samples treated with DNase I were included as positive were excluded from this study (Wu et al. 2007, Lu et al. 2009). controls. After incubation for 1 h at 37 8C, samples were A total of 580 subjects ranging from 24 to 42 years old were analysed immediately by flow cytometry (FACSCalibur; BD chosen for this study. The control group consisted of 580 fertile Biosciences Pharmingen) (Sergerie et al. 2000). men ranging from 25 to 40 years old who had fathered at least one child without assisted reproductive technologies. All participants were ethnic Han-Chinese and completed an Statistical analysis informed consent form. Participants completed a detailed The statistical analyses were performed with the Statistical questionnaire that included information pertaining to their age, Analysis System (version 9.1.3, SAS Institute, Cary, NC, USA). cigarette smoking frequency, alcohol and tea consumption, Differences in selected demographic variables as well as vitamin regimen and abstinence time. Each subject then smoking and alcohol status between the cases and the controls donated 5 ml blood that was used to extract genomic DNA. were evaluated by the c2 test. The Student’s t-test was used to In addition, the semen samples were collected after at least evaluate continuous variables, including age and pack-years of 3 days of sexual abstinence. The liquefied samples were cigarette smoking. Hardy–Weinberg equilibrium was tested by divided into two parts, one for the routine semen analysis and a goodness-of-fit c2 test. The risk of male infertility was the other for the assessment of sperm DNA fragmentation. The estimated as OR and 95% CI using unconditional multivariate semen was analysed for sperm concentration, motility and logistic regression. When appropriate, the ORs were adjusted morphology, and the analysis was performed according to the for age, cigarette smoking and alcohol use. Linear regression WHO criteria (Lu et al. 2010). models were used to estimate the association of the natural logarithm (ln) transformation sperm concentration and In-transformed sperm DNA fragmentation values for each SNP selection and genotyping SNP independently. The SNPs were selected based on the following criteria: i) they The potential gene–gene interactions were evaluated by had a high frequency of the rare allele (to allow the highest logistic regression analyses and tested by comparing the statistical power to detect associations) and ii) they were of changes in deviance (K2 log likelihood) between models of potential functional significance, i.e. those located at 50-flaking main effects with or without the interaction term. The false regions, UTRs, coding regions or 30-UTRs, according to the discovery rate method was applied to the P values for trends to databases of the International HapMap Project and NCBI reduce the potential for spurious findings due to multiple dbSNPs. Finally, we identified 11 potential functional poly- testing. All P values were two-sided, with P!0.05 considered morphisms from eight crucial genes involved in DSBs repair to be statistically significant. pathway (Table 2). Genomic DNA was isolated from leukocyte pellets of the venous blood by phenol–chloroform extraction with proteinase Declaration of interest K digestion. Genotyping was performed using the OpenArray The authors declare that there is no conflict of interest that platform (Applied Biosystems), which employs a chip-based could be perceived as prejudicing the impartiality of the Taq-Man genotyping technology. Genotyping was conducted research reported. according to the manufacturer’s protocol, and genotypes were identified by the OpenArray SNP Genotyping Analysis Soft- ware version 1.0.3. For quality control, the genotyping was Funding performed without knowledge of the subjects’ case/control status, and a random 5% of cases and controls were genotyped This work was supported by the Key Project of National Natural twice by different individuals; the reproducibility rate was Science Foundation of China (Grant No. 30930079); National 100%. To confirm the genotyping results, PCR-amplified DNA Natural Science Foundation of China (Grant No. 81202243 samples (nZ2 for each genotype) were examined by DNA and Grant No. 81172694); and the Natural Science Foundation sequencing. These results were also consistent. of Jiangsu Province (Grant No. BK2012087). A project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). DNA fragmentation analysis The TUNEL assay has been shown to be a feasible and sensitive Acknowledgements way to detect DNA fragmentation (Muratori et al. 2008). We used the APO-DIRECT kit (BD Biosciences Pharmingen, San The authors thank Yongyue Wei for statistical analysis.

Reproduction (2013) 145 463–470 www.reproduction-online.org

Downloaded from Bioscientifica.com at 09/24/2021 08:34:38PM via free access DSBs repair gene SNPs and male infertility 469

References van Gent DC, Hoeijmakers JH & Kanaar R 2001 Chromosomal stability and the DNA double-stranded break connection. Nature Reviews. Genetics 2 Abe T, Ishiai M, Hosono Y, Yoshimura A, Tada S, Adachi N, Koyama H, 196–206. (doi:10.1038/35056049) Takata M, Takeda S, Enomoto T et al. 2008 KU70/80, DNA-PKcs, and Girard PM, Kysela B, Harer CJ, Doherty AJ & Jeggo PA 2004 Analysis of Artemis are essential for the rapid induction of apoptosis after massive DNA ligase IV mutations found in LIG4 syndrome patients: the impact of DSB formation. Cellular Signalling 20 1978–1985. (doi:10.1016/j. two linked polymorphisms. Human Molecular Genetics 13 2369–2376. cellsig.2008.07.006) (doi:10.1093/hmg/ddh274) Agarwal A & Said TM 2003 Role of sperm chromatin abnormalities and Grawunder U, Wilm M, Wu X, Kulesza P, Wilson TE, Mann M & Lieber MR DNA damage in male infertility. Human Reproduction Update 9 1997 Activity of DNA ligase IV stimulated by complex formation with 331–345. (doi:10.1093/humupd/dmg027) XRCC4 protein in mammalian cells. Nature 388 492–495. (doi:10.1038/ Arbuckle JL, Fauss LA, Simpson R, Ptaszek LM & Rodgers KK 2001 41358) Identification of two topologically independent domains in RAG1 and Grawunder U, Zimmer D & Leiber MR 1998 DNA ligase IV binds to XRCC4 their role in macromolecular interactions relevant to V(D)J recombina- via a motif located between rather than within its BRCT domains. Current tion. Journal of Biological Chemistry 276 37093–37101. (doi:10.1074/ Biology 8 873–876. (doi:10.1016/S0960-9822(07)00349-1) jbc.M105988200) Han J, Colditz GA, Samson LD & Hunter DJ 2004 Polymorphisms in DNA Auranen A, Song H, Waterfall C, Dicioccio RA, Kuschel B, Kjaer SK, double-strand break repair genes and skin cancer risk. Cancer Research Hogdall E, Hogdall C, Stratton J, Whittemore AS et al. 2005 64 3009–3013. (doi:10.1158/0008-5472.CAN-04-0246) Polymorphisms in DNA repair genes and epithelial ovarian cancer risk. Hill DA, Wang SS, Cerhan JR, Davis S, Cozen W, Severson RK, Hartge P, International Journal of Cancer 117 611–618. (doi:10.1002/ijc.21047) Wacholder S, Yeager M, Chanock SJ et al. 2006 Risk of non-Hodgkin Ben-Omran TI, Cerosaletti K, Concannon P, Weitzman S & Nezarati MM lymphoma (NHL) in relation to germline variation in DNA repair and 2005 A patient with mutations in DNA ligase IV: clinical features and related genes. Blood 108 3161–3167. (doi:10.1182/blood-2005-01- overlap with Nijmegen breakage syndrome. American Journal of 026690) Medical Genetics. Part A 137A 283–287. (doi:10.1002/ajmg.a.30869) Ji BT, Shu XO, Linet MS, Zheng W, Wacholder S, Gao YT, Ying DM & Jin F Borini A, Tarozzi N, Bizzaro D, Bonu MA, Fava L, Flamigni C & Coticchio G 1997 Paternal cigarette smoking and the risk of childhood cancer among 2006 Sperm DNA fragmentation: paternal effect on early post- offspring of nonsmoking mothers. Journal of the National Cancer Institute implantation embryo development in ART. Human Reproduction 21 89 238–244. (doi:10.1093/jnci/89.3.238) 2876–2881. (doi:10.1093/humrep/del251) Ji G, Gu A, Wang Y, Huang C, Hu F, Zhou Y, Song L & Wang X 2012 Genetic Caglar GS, Koster F, Schopper B, Asimakopoulos B, Nehls B, Nikolettos N, variants in antioxidant genes are associated with sperm DNA damage Diedrich K & Al-Hasani S 2007 Semen DNA fragmentation index, and risk of male infertility in a Chinese population. Free Radical Biology evaluated with both TUNEL and Comet assay, and the ICSI outcome. & Medicine 52 775–780. (doi:10.1016/j.freeradbiomed.2011.11.032) In Vivo 21 1075–1080. Kale SB, Landree MA & Roth DB 2001 Conditional RAG-1 mutants block Chen T, Dong B, Lu Z, Tian B, Zhang J, Zhou J, Wu H, Zhang Y, Wu J, Lin P the hairpin formation step of V(D)J recombination. Molecular and et al. 2010 A functional single nucleotide polymorphism in promoter of Cellular Biology 21 459–466. (doi:10.1128/MCB.21.2.459-466.2001) ATM is associated with longevity. Mechanisms of Ageing and Karran P 2000 DNA double strand break repair in mammalian cells. Current Development 131 636–640. (doi:10.1016/j.mad.2010.08.009) Opinion in Genetics & Development 10 144–150. (doi:10.1016/S0959- Cibeira MT, de Larrea CF, Navarro A, Diaz T, Fuster D, Tovar N, Rosinol L, 437X(00)00069-1) Monzo M & Blade J 2011 Impact on response and survival of DNA repair Khanna KK & Jackson SP 2001 DNA double-strand breaks: signaling, repair single nucleotide polymorphisms in relapsed or refractory multiple and the cancer connection. Nature Genetics 27 247–254. (doi:10.1038/ myeloma patients treated with thalidomide. Leukemia Research 35 85798) 1178–1183. (doi:10.1016/j.leukres.2011.02.009) Lee JH & Paull TT 2005 ATM activation by DNA double-strand breaks Cohen-Bacrie P, Belloc S, Menezo YJ, Clement P, Hamidi J & Benkhalifa M 2009 Correlation between DNA damage and sperm parameters: a through the Mre11-Rad50-Nbs1 complex. Science 308 551–554. prospective study of 1,633 patients. Fertility and Sterility 91 1801–1805. (doi:10.1126/science.1108297) (doi:10.1016/j.fertnstert.2008.01.086) Lewis SE & Aitken RJ 2005 DNA damage to spermatozoa has impacts De Kretser DM & Baker HW 1999 Infertility in men: recent advances and on fertilization and pregnancy. Cell and Tissue Research 322 33–41. continuing controversies. Journal of Clinical Endocrinology and (doi:10.1007/s00441-005-1097-5) Metabolism 84 3443–3450. (doi:10.1210/jc.84.10.3443) Lieber MR 2010 The mechanism of double-strand DNA break repair by the Dore AS, Furnham N, Davies OR, Sibanda BL, Chirgadze DY, Jackson SP, nonhomologous DNA end-joining pathway. Annual Review of Biochem- Pellegrini L & Blundell TL 2006 Structure of an Xrcc4-DNA ligase IV istry 79 181–211. (doi:10.1146/annurev.biochem.052308.093131) yeast ortholog complex reveals a novel BRCT interaction mode. DNA Liu Y, Zhang H, Zhou K, Chen L, Xu Z, Zhong Y, Liu H, Li R, Shugart YY, Repair 5 362–368. (doi:10.1016/j.dnarep.2005.11.004) Wei Q et al. 2007 Tagging SNPs in non-homologous end-joining Dresser ME 2000 Meiotic chromosome behavior in Saccharomyces pathway genes and risk of glioma. Carcinogenesis 28 1906–1913. cerevisiae and (mostly) mammals. Mutation Research 451 107–127. (doi:10.1093/carcin/bgm073) (doi:10.1016/S0027-5107(00)00043-9) Lu C, Zhang J, Li Y, Xia Y, Zhang F, Wu B, Wu W, Ji G, Gu A, Wang S et al. Esteves SC, Miyaoka R & Agarwal A 2011 An update on the clinical 2009 The b2/b3 subdeletion shows higher risk of spermatogenic failure assessment of the infertile male. Clinics 66 691–700. (doi:10.1590/ and higher frequency of complete AZFc deletion than the gr/gr S1807-59322011000400026) subdeletion in a Chinese population. Human Molecular Genetics 18 Fachal L, Gomez-Caamano A, Peleteiro P, Carballo A, Calvo-Crespo P, 1122–1130. (doi:10.1093/hmg/ddn427) Sanchez-Garcia M, Lobato-Busto R, Carracedo A & Vega A 2012 Lu JC, Huang YF & Lu NQ 2010 WHO Laboratory Manual for the Association of a XRCC3 polymorphism and rectum mean dose with the Examination and Processing of Human Semen: its applicability to risk of acute radio-induced gastrointestinal toxicity in prostate cancer andrology laboratories in China. Zhonghua Nan Ke Xue 16 867–871. patients. Radiotherapy and Oncology 105 321–328. (doi:10.1016/j. McBlane JF, van Gent DC, Ramsden DA, Romeo C, Cuomo CA, Gellert M & radonc.2012.09.013) Oettinger MA 1995 Cleavage at a V(D)J recombination signal requires Figueroa JD, Malats N, Rothman N, Real FX, Silverman D, Kogevinas M, only RAG1 and RAG2 proteins and occurs in two steps. Cell 83 387–395. Chanock S, Yeager M, Welch R, Dosemeci M et al. 2007 Evaluation of (doi:10.1016/0092-8674(95)90116-7) genetic variation in the double-strand break repair pathway and bladder Mittal RD, Gangwar R, Mandal RK, Srivastava P & Ahirwar DK 2012 Gene cancer risk. Carcinogenesis 28 1788–1793. (doi:10.1093/carcin/ variants of XRCC4 and XRCC3 and their association with risk for bgm132) urothelial bladder cancer. Molecular Biology Reports 39 1667–1675. Gellert M, Hesse JE, Hiom K, Melek M, Modesti M, Paull TT, Ramsden DA (doi:10.1007/s11033-011-0906-z) & van Gent DC 1999 V(D)J recombination: links to transposition Mo X, Bailin T & Sadofsky MJ 2001 A C-terminal region of RAG1 contacts and double-strand break repair. Cold Spring Harbor Symposia on the coding DNA during V(D)J recombination. Molecular and Cellular Quantitative Biology 64 161–167. (doi:10.1101/sqb.1999.64.161) Biology 21 2038–2047. (doi:10.1128/MCB.21.6.2038-2047.2001) www.reproduction-online.org Reproduction (2013) 145 463–470

Downloaded from Bioscientifica.com at 09/24/2021 08:34:38PM via free access 470 G Ji, L Yan and others

Mohindra A, Hays LE, Phillips EN, Preston BD, Helleday T & Meuth M 2002 Thacker J & Zdzienicka MZ 2004 The XRCC genes: expanding roles in DNA Defects in homologous recombination repair in mismatch-repair- double-strand break repair. DNA Repair 3 1081–1090. (doi:10.1016/j. deficient tumour cell lines. Human Molecular Genetics 11 2189–2200. dnarep.2004.04.012) (doi:10.1093/hmg/11.18.2189) Tseng RC, Hsieh FJ, Shih CM, Hsu HS, Chen CY & Wang YC 2009 Lung Muratori M, Forti G & Baldi E 2008 Comparing flow cytometry and cancer susceptibility and prognosis associated with polymorphisms in fluorescence microscopy for analyzing human sperm DNA fragmenta- the nonhomologous end-joining pathway genes: a multiple genotype– tion by TUNEL labeling. Cytometry. Part A 73 785–787. (doi:10.1002/ phenotype study. Cancer 115 2939–2948. (doi:10.1002/cncr.24327) cyto.a.20615) Venkitaraman AR 2002 Cancer susceptibility and the functions of BRCA1 O’Driscoll M, Cerosaletti KM, Girard PM, Dai Y, Stumm M, Kysela B, and BRCA2. Cell 108 171–182. (doi:10.1016/S0092-8674(02)00615-3) Hirsch B, Gennery A, Palmer SE, Seidel J et al. 2001 DNA ligase IV Wei YF, Robins P, Carter K, Caldecott K, Pappin DJ, Yu GL, Wang RP, mutations identified in patients exhibiting developmental delay and Shell BK, Nash RA, Schar P et al. 1995 Molecular cloning and expression immunodeficiency. Molecular Cell 8 1175–1185. (doi:10.1016/S1097- of human cDNAs encoding a novel DNA ligase IV and DNA ligase III, an 2765(01)00408-7) enzyme active in DNA repair and recombination. Molecular and Olive PL, Wlodek D & Banath JP 1991 DNA double-strand breaks Cellular Biology 15 3206–3216. measured in individual cells subjected to gel electrophoresis. Cancer Wu X, Gu J, Grossman HB, Amos CI, Etzel C, Huang M, Zhang Q, Research 51 4671–4676. Millikan RE, Lerner S, Dinney CP et al. 2006 Bladder cancer Rajasekaran R, Doss GP, Sudandiradoss C, Ramanathan K, Rituraj P & predisposition: a multigenic approach to DNA-repair and cell-cycle- Sethumadhavan R 2008 Computational and structural investigation of control genes. American Journal of Human Genetics 78 464–479. deleterious functional SNPs in breast cancer BRCA2 gene. Sheng (doi:10.1086/500848) Wu Gong Cheng Xue Bao 24 851–856. (doi:10.1016/S1872-2075 Wu B, Lu NX, Xia YK, Gu AH, Lu CC, Wang W, Song L, Wang SL, Shen HB & (08)60042-4) Wang XR 2007 A frequent Y chromosome b2/b3 subdeletion shows Robins P & Lindahl T 1996 DNA ligase IV from HeLa cell nuclei. Journal of Biological Chemistry 271 24257–24261. (doi:10.1074/jbc.271.39. strong association with male infertility in Han-Chinese population. 24257) Human Reproduction 22 1107–1113. (doi:10.1093/humrep/del499) Roddam PL, Rollinson S, O’Driscoll M, Jeggo PA, Jack A & Morgan GJ 2002 Yin M, Liao Z, Liu Z, Wang LE, O’Reilly M, Gomez D, Li M, Komaki R & Genetic variants of NHEJ DNA ligase IV can affect the risk of developing Wei Q 2012 Genetic variants of the nonhomologous end joining gene multiple myeloma, a tumour characterised by aberrant class switch LIG4 and severe radiation pneumonitis in nonsmall cell lung cancer recombination. Journal of Medical Genetics 39 900–905. (doi:10.1136/ patients treated with definitive radiotherapy. Cancer 118 528–535. jmg.39.12.900) (doi:10.1002/cncr.26214) Sakkas D, Mariethoz E, Manicardi G, Bizzaro D, Bianchi PG & Bianchi U Zha S, Alt FW, Cheng HL, Brush JW & Li G 2007 Defective DNA repair and 1999 Origin of DNA damage in ejaculated human spermatozoa. increased genomic instability in Cernunnos-XLF-deficient murine ES Reviews of Reproduction 4 31–37. (doi:10.1530/ror.0.0040031) cells. PNAS 104 4518–4523. (doi:10.1073/pnas.0611734104) San Filippo J, Sung P & Klein H 2008 Mechanism of eukaryotic homo- logous recombination. Annual Review of Biochemistry 77 229–257. (doi:10.1146/annurev.biochem.77.061306.125255) Received 26 September 2012 Sergerie M, Ouhilal S, Bissonnette F, Brodeur J & Bleau G 2000 Lack of First decision 12 November 2012 association between smoking and DNA fragmentation in the spermato- zoa of normal men. Human Reproduction 15 1314–1321. (doi:10.1093/ Revised manuscript received 25 December 2012 humrep/15.6.1314) Accepted 5 March 2013

Reproduction (2013) 145 463–470 www.reproduction-online.org

Downloaded from Bioscientifica.com at 09/24/2021 08:34:38PM via free access