Mol Biol Rep DOI 10.1007/s11033-012-2009-x
Racial and tissue-specific cancer risk associated with PARP1 (ADPRT) Val762Ala polymorphism: a meta-analysis Noel Pabalan • Ofelia Francisco-Pabalan • Hamdi Jarjanazi • Hong Li • Lillian Sung • Hilmi Ozcelik
Received: 19 January 2012 / Accepted: 1 October 2012 Ó Springer Science+Business Media Dordrecht 2012
Abstract The Val762Ala polymorphism poly [ADPribose] polymerase 1 (PARP1) gene [ADPRT (adenosine diphosphate ribosyltransferase) gene] affects enzymatic activity, which modulates cancer susceptibility among human populations. Individual data on 13,745 cases and 16,947 controls from 28 published case–control studies were re-evaluated. Odds ratios (OR) were estimated for ethnic group, cancer type, smoking joint effects and studies confined to the Hardy–Weinberg equilibrium. We applied subgroup, sensitivity and outlier analyses as well as the Bonferroni correction for multiple testing. The results show strong evidence that the variant (C) allele confers significant increased risk in the Chinese (OR 1.20–1.44,
Electronic supplementary material The online version of this article (doi:10.1007/s11033-012-2009-x) contains supplementary material, which is available to authorized users. N. Pabalan Office of Research and Development, Angeles University Foundation, Angeles City, Philippines O. Francisco-Pabalan Canadian Grain Commission, 5 Water Street, Winnipeg, MB R3C 3G8, Canada H. Jarjanazi Ontario Ministry of the Environment, 125 Resources Road, Toronto, ON M9P 3V6, Canada H. Li H. Ozcelik (&) Fred A. Litwin Centre for Cancer Genetics, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 60 Murray St. Room L6-303, Box 29, Toronto, ON M5T 3L9, Canada e-mail:
[email protected] L. Sung Division of Hematology/Oncology, Hospital for Sick Children, Toronto, ON, Canada
P \ 0.0001–0.002), exacerbated by smoking (OR 1.66–2.53, P \ 0.0001) and joint interaction with XRCC1 Arg399Gln (OR 1.39, P \ 0.0001) as well as adjustment for tumor type (gastric carcinoma ORs 1.39–2.01, P \ 0.0001). These significant effects were unaltered following conservative correction for multiple tests. By contrast, this procedure erased the protective significance in Caucasians, but not in two American subgroups, (i) those in the brain tumor category (0.77–0.79, P \ 0.0001) and (ii) smokers in the dominant model (OR 0.86, P \ 0.0001). These differential findings between the two ethnicities maybe correlated with significantly (P \ 0.0001) greater allele frequency of the variant allele (C) among the Chinese compared to Caucasians. Our racial and tissue-specific summary estimates imply consideration of the Val762Ala polymorphism as candidate gene marker for screening cancer patients’ best suited for PARP inhibitor therapy. Keywords Cancer ADPRT PARP1 Polymorphism Val762Ala Meta-analysis
Introduction Poly [ADP-ribose] polymerase 1 (PARP1), also known as nicotinamide adenine dinucleotide (NAD?) ADP-ribosyltransferase 1 or poly[ADP-ribose] synthase 1 is a nuclear chromatin-associated enzyme encoded by the adenosine diphosphate-ribosyltransferase (ADPRT) gene on chromosome 1q41-42. PARP1 is involved in repair of singlestranded DNA breaks whereby poly ADP-ribosylation modifies nuclear proteins. Activation of PARP1 has been shown in a variety of diseases [1–3]. Although present in all eukaryotic cells, PARP1 is more highly expressed in nuclei of actively proliferating cells [4] and observed in
123
Mol Biol Rep
several cancer types such as hepatocellular, colorectal carcinomas and brain tumors [5–7]. Degree of PARP1 activation modulates the decision to die by apoptosis or necrosis. Medium levels of DNA damage optimize PARP1 activity. However, high levels deplete cellular NAD? and the resulting lack of adenosine triphosphate (ATP) renders the cell unable to execute cell death by the energy consuming apoptotic process, leading to cell death by necrosis [8, 9]. PARP1 inhibitors effectively disarm the ability of cancer cells to repair themselves and cause their death by preservation of cellular ATP (prevents NAD? depletion) necessary to accomplish apoptosis, and therefore avoiding necrotic death. A T to C nucleotide transition characterizes a commonly occurring PARP1 polymorphism (rs13641) leading to Val762Ala variation [valine (Val) (T) to alanine (Ala) (C) amino acid substitution at codon 762] in the encoded PARP protein. This polymorphism is located in the 6th helix of the C-terminal catalytic domain which binds NAD? leading to poly ADP-ribosylation [10]. This is in fact the domain that actually enabled development of potent PARP inhibitors [11]. The PARP1 762Ala (C) allele was shown to contribute significantly to low poly(ADPribosyl)ation activities in a dosage-dependent manner [12]. Furthermore, structural evidence points to a steric change in the catalytic domain of the Ala (C) allele leading to impairment of enzymatic activity [13] rendering individuals with this polymorphism to increased cancer risk [14]. To date, evidence accumulating from number of studies suggests genetic association of the Val762Ala with cancer development although discrepancies among the findings exist. The increasing number of reports and discrepancy of the findings prompted us to carry the meta-analysis of the Val762Ala to understand its role in cancer development
Materials and methods Publication search and data extraction We searched PubMed as of January 2011 resulting in identification of 29 eligible articles, which used a total 31 case–control populations to investigate the association of this polymorphism with cancer risk in various ethnic populations including Caucasians of the United States [US] [5, 6, 12, 15–24], Europe [25–27], French Canada [28] and Eurasia [29]; Asians of Chinese [30–39] and Korean origin [40] and one AfricanAmerican group [22]. To improve homogeneity, we focused on the two main large groups, primarily Caucasian American (13 articles) and Asian Chinese (10 articles) and secondarily, the Central and Western European populations (three articles). We excluded the rest that were not in these ethnic/geographic categories. Genotype frequencies in three types of brain tumors
123
(glioma, meningioma and acoustic neuroma) of one American paper [6] were treated separately. This resulted in a total of 15 studies for the Caucasian American population. Each of the three European [25–27] and 10 Asian articles [30–39] had a single population, two of which were in the Chinese language [32, 39]. Thus, the combination of Caucasian American, European and Asian Chinese (hence, Chinese) populations totaled 28 case–control studies (Table 1). Two investigators independently extracted data and reached consensus on all the items. The following information was obtained from each publication: first author’s name, published year, country of origin, dominant ancestry of the study populations, matching criteria, genotype data as well as number of cases and controls. Inclusion criteria were: (i) provision of genotypic data, and (ii) case–control design of the study. Departures of genotypic frequencies from the Hardy–Weinberg Equilibrium (HWE) in control subjects were determined with the v2 test and we found three such studies [33, 37, 39] (Table 1) indicating potential biases in the selection of controls. Assuming an odds ratio (OR) of 1.5 at a genotypic risk level of a = 0.05 (twosided), power was considered adequate at C80 %. Data were analyzed using Review Manager (RevMan 4.2, Oxford, England), SigmaStat 2.03 and SigmaPlot 9.01. Significance was set at a P value of \0.05 throughout except in heterogeneity estimation. Subgroup selection In the Chinese population, summary effects were obtained with and without the studies deviating from the HWE. Of the 15 cancer types, we analyzed four distinct tumor sites, stomach [32, 33, 37, 39], breast [22, 24, 25, 28, 36], brain [6, 19, 20, 29] and bladder [16, 23, 26, 35] with C4 studies each. Of the four sites, gastric carcinoma was specific to the Chinese and the brain tumor studies to the Americans. The brain tumor category was composed of five studies, three of which investigated glioma [6, 19, 29] and the remaining three focusing on glioblastoma [20], meningioma [6] and acoustic neuroma [6]. The remaining cancer types, with number of studies ranging from one to two, were combined under other cancers [5, 12, 15, 17, 18, 21, 27, 30, 31, 34, 38] (Table 1). We investigated the role of smoking in the Chinese and American populations from six studies [16, 17, 22, 31, 33, 38] as well as joint effects of PARP1 Val762Ala and XRCC1 Arg399Gln from five studies [31, 33, 36, 38, 39]. Statistical analysis We used crude OR to compare effects on the same baseline. Based on associations of the Val762Ala variant allele,
Mol Biol Rep Table 1 Characteristics of the 28 studies included in the meta-analysis First author [ref]
Years
Cancer type
N Cases
Controls
Ethnicity (country)
Controls in HWE
Maf in controls
Power (v2 = 0.05, OR = 1.5) 26.8
1
Cao [25]
2007
Breast
83
100
Caucasian (France)
Yes
0.14
2
Smith [22]a
2008
Breast
314
397
Caucasian (US)
Yes
0.17
75.3
3
Zhai [36]
2006
Breast
302
639
Asian (China)
Yes
0.43
81.6
4
Zhang [24]
2006
Breast
1,715
1,371
Caucasian (US)
Yes
0.17
99.9
5
Liu [19]a
2009
Glioma
267
236
Caucasian (US)
Yes
0.20
77.0
6
McKean [20]a
2009
Glioblastoma
987
1,935
Caucasian (US)
Yes
0.18
99.9
7 8
Rajaraman [6] Rajaraman [6]
2010 2010
Neuroma Meningioma
65 121
463 463
Caucasian (US) Caucasian (US)
Yes Yes
0.18 0.18
32.6 49.8
9
Rajaraman [6]
2010
Glioma
340
464
10
Figueroa [26]
2007
Bladder
1,138
1,131
11
Huang [16]
2007
Bladder
606
595
Caucasian (US)
Yes
0.16
93.4
12
Wang [35]
2010
Bladder
234
253
Asian (China)
Yes
0.44
59.5
13
Wu [23]
2006
Bladder
606
595
Caucasian (US)
Yes
0.16
93.4
14
Kang [32]
2010
Stomach
150
152
Asian (China)
Yes
0.26
41.0
15
Miao [33]a
2006
Stomach
500
1,000
Asian (China)
No
0.36
95.4
16
Zhang [39]a
2006
Stomach
236
708
Asian (China)
No
0.38
75.7
17
Zhang [37]
2009
Stomach
236
320
Asian (China)
No
0.27
64.3
18
Berndt [5]
2007
Colorectal
691
702
Caucasian (US)
Yes
0.17
96.2
19
Stern [34]
2007
Colorectal
307
1,173
Asian (Singapore)
Yes
0.43
87.6
20
Landi [27]
2006
Lung
292
307
21
Zhang [38]
2005
Lung
1,000
1,000
22 23
Shen [21] Gao [15]
2006 2010
NHL Prostate
535 453
24
Lockett [12]
2004
Prostate
25
Hao [31]
2004
26
Li [17]a
2007
27
Li [18]a
28
Chiang [30]
Caucasian (US)
Yes
0.18
80.0
Caucasian (Spain)
Yes
0.12
99.7
Caucasian (Europe)
Yes
0.18
68.5
Asian (China)
Yes
0.39
99.4
455 119
Caucasian (US) Caucasian (US)
Yes Yes
0.15 0.19
88.0 49.1
438
427
Caucasian (US)
Yes
0.14
83.6
Esophageal
414
479
Asian (China)
Yes
0.41
84.5
Head and neck
830
854
Caucasian (US)
Yes
0.16
98.4
2006
Skin
602
603
Caucasian (US)
Yes
0.17
93.4
2008
Thyroid
283
469
Asian (China)
Yes
0.41
75.6
a
Outlier studies; Power was calculated with the G* Power program (http://www.psycho.uniduesseldorf.de/aap/projects/gpower) as probability of detecting an association between Val762Ala and cancer assuming odds ratios (OR) of 1.5 (small effects size); maf minor allele frequency, NHL Non-Hodgkin Lymphoma
we estimated OR and 95 % confidence interval (CI) of cancer associated with low activity allele(s) compared with the reference high activity allele(s) (Ala vs. Val) and the homozygous model (Ala–Ala vs. Val–Val). To examine genotype specific risk, we considered the common genotype (Val–Val) as reference for calculating summary ORs for the Ala–Ala Ala-Val genotypes. We also evaluated the odds of Ala–Ala vs. Ala-Val ? Val–Val and Ala–Ala ? Ala-Val vs. Val–Val, assuming recessive and dominant effects of the variant (Ala) allele, respectively. Pooled ORs (summary estimates) were obtained using either the fixed (Mantel– Haenszel) or random (DerSimonian-Laird) effects models. The fixed-effects model was used in the absence of heterogeneity [41] while the random-effects model was used in its presence [42]. Assuming genuine diversity in the results of various studies, the random-effects model incorporates
between study variance. Sensitivity analysis involved omitting one study at a time and recalculating the pooled OR. We identified eight studies [12, 16, 21, 26, 33, 36, 39, 40] whose individual absence from the analysis altered heterogeneity of the summary effects. Heterogeneity between studies was estimated using the v2-based Q test [43]. Given the low power of this test [44], significance threshold was set at P = 0.10. Effect of heterogeneity was quantified with the I2 statistic which measures the degree of inconsistency among the studies [45]. The Galbraith plot [46] was used to identify potential outlier studies after which their influence on pooled effects and heterogeneity was graphically examined [47]. Table 1 and Supplementary Material, Fig. S2, show the seven studies which are outputs of outlier analysis, two Chinese [33, 39] (Fig. S2a) five Caucasian [17–20, 22] (Fig. S2b).
123
Mol Biol Rep
Publication bias was statistically evaluated with Egger’s regression asymmetry test which detects whether the intercept deviates significantly from zero in a regression of the standardized effect estimates against their precision [48]. The Begg-Mazumdar diagnostic (nonparametric s correlation coefficient) was also applied to evaluate whether the magnitude of observed association was related to the variance of each study [49]. The Bonferroni method corrected for multiple testing [corrected P = 0.001 (0.05/34)]
Results Allelic difference PARP1 polymorphism data were obtained for two major populations, Asian Chinese (3,662 cases/6,193 controls) and Caucasians (10,083 cases/10,754 controls). Interestingly, allelic frequency of Ala (C) allele was found to be significantly higher in Chinese controls (=0.39 ± 0.05) when compared to Caucasians (=0.17 ± 0.02) [Mann– Whitney U test: P \ 0.0001] (Supplementary Material, Fig. S1). In the following analyses, we have estimated the allelic and genotype specific cancer risk associated with Ala (C) compared to the reference Val (T) in both Asian and Caucasians populations. Associations with cancer Of the 10 Chinese studies, half [31, 33, 34, 36, 38] had C80 % statistical power to demonstrate an association (Table 1). Controls in nine studies were matched with cases, all by age. Of the nine, five were matched by gender [31, 33, 35, 37, 38] and one using the region criterion [37]. Combination of these three matching criteria was found in five studies [31, 33, 35, 37, 38]. Although three studies [31, 33, 38] that investigated smoking (1,164 cases/1,246 controls) were composed of three different tumor types (esophageal, gastric and lung), they had were statistically strong and the controls were matched with cases by age and gender. On the other hand, of the age-matched Chinese studies [32, 33, 37, 39] that investigated gastric carcinoma (1,122 cases/2,180 controls), only one [33] had adequate power. Without evidence of publication bias (Table 2), analysis of the Chinese studies showed increased risk (ORs 1.20–1.44) associated with cancer, significant (P \ 0.0001–0.001) under all genetic models (Fig. 1a). Obtained mostly under heterogeneous conditions (I2 = 47–54 %), these heterogeneity disappeared (I2 = 0–13 %) when re-analyzed with studies that agreed with the HWE (N = 7:2,690 cases/4,165 controls). Of the four significant of the pooled ORs (ORs 1.13–1.27) in HWE (Fig. 1b), only the allele effect (OR 1.13,
123
P = 0.0007) survived the Bonferroni correction for multiple testing. Omission of the outliers [33, 39] retained Chinese significance (P \ 0.0001–0.0002) but impacted on heterogeneity (Fig. 1c). Sensitivity analysis identified two studies [33, 36] that improved heterogeneity without affecting significance. In Chinese smokers, the Ala allele (C) and Ala–Ala (CC) genotypes were significantly (all at P \ 0.0001) associated with increased risk of cancer (ORs 1.66–2.53). Significance and homogeneity (I2 = 0–14 %) of the pooled ORs in this subgroup (Fig. 1a) were unchanged by the following: (i) confined to studies in HWE (ORs 1.64–2.28) in all comparisons (Fig. 1b), and (ii) when subjected to sensitivity analysis. These nonchanges underpin robustness of the smoker summary effects. The gastric cancer studies may not account for the overall Chinese effect of susceptibility because omitting them from the analysis did not materially affect the pooled ORs which remained significant in all genetic models (ORs 1.12–1.27, P values: 0.002–0.02). Under all genetic models, the Chinese were susceptible to gastric cancer (ORs 1.39–2.01, P \ 0.0001). Highest impact was observed in the homozygous model (OR 2.01, 95 % CI 1.60–2.52, P \ 0.0001). These summary effects were obtained under highly homogeneous (I2 = 0–5 %) conditions (Fig. 3). Significant P values (all \0.001) from the outlier and sensitivity analyses, smoking subgroup as well gastric cancer were unaffected by multiple testing corrections. Of the 18 Caucasian studies, over half (61.1 %) were adequately powered (Table 1). Controls in 16 studies were matched by age, sex, ethnicity and residence, 12 [5, 6, 16– 18, 20, 23, 26, 27] of these used one or more combinations of these matching criteria. The Caucasian population is composed of US-based Americans (15 studies: (8,570 cases/9,679 controls) and Europeans (three studies: 1,513 cases/1,538 controls). The three (868 cases, 624 controls)
Table 2 Evaluation of publication bias Egger’s regression
Begg and Mazumdar correlation
Intercept
P value
Kendall’s s
Allele
-0.34
0.86
0.11
0.66
Dominant
-0.40
0.80
0.11
0.66
Recessive
-0.95
0.62
0.02
0.93
Homozygous
-1.18
0.55
0.07
0.79
Allele
-1.31
0.19
-0.18
0.31
Dominant
-0.99
0.35
0.00
1.00
P value
Chinese
Caucasian
Only significant effects from the two populations were tested. Significance was set at P \ 0.05
Mol Biol Rep
Fig. 1 Results of the meta-analysis for PARP1 Val762Ala polymorphism and cancer risk in the Chinese population
smoking studies [16, 17, 22] had the limitation of being composed of three different tumor types (bladder, head and neck, breast). However, two [16, 17] of the three studies had [90 % statistical power to demonstrate an association and all the controls were matched to cases by age, two by gender and ethnicity [16, 17]. All five brain tumor (1,780 cases/3,561 controls) studies [6, 19, 20] were US-based; two [6, 20] had statistical power of [80 % and all were matched by age, three by gender and residence [6], one by ethnicity [20] and four [6, 20] with a combination of the aforementioned criteria. Figure 2a shows that in the respective allele and dominant models, the significant (P = 0.02) Caucasian protective (ORs 0.91 and 0.89) effects were unchanged (P \ 0.0001 and =0.002) when the analysis was confined to Americans only (ORs 0.86–0.89). These pooled ORs were primarily obtained under heterogeneous conditions (I2 = 40–74 %) and without evidence of publication bias (Table 2) Fig. 2b shows that Caucasian re-analysis without the outlier studies [17–20, 22] resulted in loss of
significance (allele: 0.94, P = 0.18; dominant: 0.97, P = 0.56). Outlier analysis of the Americans only resulted in losses of both heterogeneity (I2 = 33 %) and of significance (P = 0.21) in the allele and dominant models, respectively (Fig. 2b). Sensitivity analysis elicited loss of heterogeneity with removal of three studies [12, 21, 26] in this subgroup. Reduced risks in the dominant (OR 0.82, 95 % CI 0.55–1.21, P = 0.31) and allelic (OR 0.54, 95 % CI 0.33–0.86, P = 0.01) models among American smokers, obtained under heterogeneous conditions (I2 = 60 % and 74 %) (Fig. 2a) were altered with sensitivity analysis resulting in gain in significance (OR 0.66, 95 % CI 0.46–0.93, P = 0.02) and loss of heterogeneity (I2 = 0 %), respectively with removal of Huang et al. [16]. The significant protection of Americans from developing brain tumors (OR 0.64–0.77, P = 0.04–0.0001) (Fig. 3) was unchanged with outlier analysis (data not shown). The brain tumor effects may account for the significant protection in Caucasians because omitting these cancer type studies and redoing the analysis resulted in loss
123
Mol Biol Rep
of significance of the pooled ORs in the allele contrast (OR 0.95, P = 0.27) and dominant model (OR 0.94, P = 0.31). Significant effects, mainly allelic, from the Caucasian, American and smoker analyses as well as the recessive and homozygous comparisons from the brain tumor analyses were consistent in failing the Bonferroni correction. The breast (N = 4: 2,414 cases/2,507 controls) and bladder (N = 4: 2,584 cases/2,574 controls) cancer studies suggested null associations (ORs 0.86–1.04). Heterogeneity of the breast cancer studies was zero (I2 = 0 %) while that of bladder cancer was high (I2 = 44–94 %) (Fig. 3) Heterogeneity of the null effects (OR 1.03) in other cancer types (11 studies: 5,845 cases/6,588 controls) was also high (I2 = 54–64 %) (data not shown)
protein 1) Arg399Gln. Homogeneous (I2 = 44 %) joint effects indicate increased risk in the allele contrast (OR 1.39). This significant (P \ 0.00001) effect survived application of the multiple correction test but not the marginally significant (P = 0.05) recessive OR (1.92) obtained heterogeneously (I2 = 78 %). Sensitivity analysis elicited the following effects: (i) non-significant effects in the dominant (OR 1.07, P = 0.20) and homozygous (OR 2.10, P = 0.12) models were materially altered to significance (OR 1.15, P = 0.04 and 2.89, P = 0.0001, respectively) when recalculated without Zhai et al. [36]. (ii) The homogeneous (Pheterogeneity: 0.15) allele effect disappeared (Pheterogeneity: 0.07) when reanalyzed without Zhang et al. [39].
Gene–gene interaction
Discussion
Figure 1a summarizes the interaction between PARP1 Val762Ala and XRCC1 (X-ray repair cross-complementing
Compared to major allele frequency Val (T) (61 % in Asians, 83 % in Caucasians), the minor allele frequency
(a)
(b)
Fig. 2 Results of the meta-analysis for PARP1 Val762Ala polymorphism and cancer risk in the Caucasian population
123
Mol Biol Rep Fig. 3 Results of the metaanalysis for PARP1 Val762Ala polymorphism and cancer risk in four tumor types
Ala (C) (39 % in Asians, 17 % in Caucasians) conferred strong susceptibility to Chinese (up to 1.3-fold) and protection to Caucasians (up to 9–17 %) accounted by the brain related cancer studies. These findings suggest that different ethnicities might engender different roles of the PARP1 Val762Ala polymorphism because of differing genetic background. Population-specific differences of the PARP1 effects have been documented [28, 36, 50]. Chance effects are less likely to impact our findings for the following reasons: (i) the two populations have adequate number of studies and large sample sizes, (ii) half of Chinese and more than half of Caucasian studies have adequate statistical power, (iii) statistical homogeneity of the two populations particularly the Chinese in HWE and
Americans in all but one genetic model, and (iv) we applied the conservative correction for multiple tests. These reasons render selection bias unlikely and reduce the risk of confounding. Given the ethnic and geographical homogeneity that mark the two populations, the individual studies are unlikely to show population admixture. Removal of outlier studies found in the ethnic categories has provided, for the most part, a more homogeneous but less significant associations with cancer. Caucasians, brain tumors and PARP1 Summary effects in the allele and dominant models of the brain tumor category were shown to be robust for two
123
Mol Biol Rep
reasons: (i) they were unchanged by outlier and sensitivity analyses, and (ii) the significant P values in these genetic models survived the Bonferroni correction. Given these outcomes, it could then be confidently stated that the variants of PARP1 have shown 21–23 % reduced risk of brain tumors in Americans [20]. This protection may correlate with elevated population frequencies of the, Val (T) that probably translate to increased poly(ADP-ribosyl)ation activity and PAR accumulation. Study-specific protective effects of the variant, Ala (C) ranged from 20 % in glioblastoma [20] to 36 % in meningioma [6] to 50 % in neuroma [6] and significant (OR 0.71, 95 % CI 0.52–0.97) in one study [19] for glioma. The protective effects from brain tumors seen in our meta-analysis probably reflect the synthetic lethal interactions between PARP inhibitors and PTEN (phosphatase and tensin homolog) loss of function [51, 52]. Cell death in gliomas has been targeted therapeutically. Cell damage through PARP1 cleavage observed in a study indicates that PARP inhibitors have a potential role in glioma treatment [53]. Understanding the exact mode of cell death and specific signaling mechanism involved on gliomas is significant to inhibit unwanted forms of cell death [54] Chinese susceptibility to gastric cancer In the Chinese, susceptibility (1.5 and 2-fold) to gastric carcinoma has been associated with Ala (C) allele and Ala/ Ala (CC) genotype compared with the 762Val (T) allele and 762Val/Val (TT) genotypes, respectively [37]. The possibility of inflating Type I error [55] because of HWEdeviating studies maybe limiting. However, this is offset with the strength of ethnic and statistical homogeneity (I2 = 0 %). Incidence data point to China as having high rates of gastric cancer [56] with an age-adjusted mortality rate exceeding 70 deaths per 100,000 males. High starch intake, salted and spicy foods may have abrasive qualities that predispose to degeneration of the gastric mucosa [32]. Given the likely dietary and broader environmental effects, where, compared to the white American population, the Chinese have been shown to have a relative risk of 1.87 (95 % CI 1.56–2.21) developing gastric cancer [57]. Another factor that may account for elevated gastric cancer risk is interaction of Helicobacter pylori infection and Val762Ala [58]. A study-specific effect of a Chinese hospital-based case–control study [37] on interaction of H. pylori infection and Val762Ala for risk of gastric cancer reported a significant 1.9-fold association (P = 0.01) of the variant genotype (Ala/Ala). This may agree with the finding of high H. pylori-sero ? rate ([70 %) among the Chinese [59] but appears to conflict with a population-based cohort study which failed to find a significant association (OR 0.79, 95 % CI 0.50–1.24) but not when evaluated by
123
smoking which indicated significant increased risk (OR 1.59, 95 % CI 1.03–2.45) [60]. Another study [61] found that an H. pylori factor activates PARP1. Interactions with environment and with other genes Tobacco smoke contains various carcinogens that can cause DNA damage [62]. If the Ala (C) variant diminishes base-repair excision capacity, smokers carrying the 762Ala/Ala (CC) homozygous variant genotype will be at higher risk for cancer compared to those carrying the common genotype. Our study appears to support this hypothetical statement where we found increased risk among Chinese smokers. In a bladder cancer study, susceptibility in smokers were found to be 1.95-fold for individuals with the Val/Val (TT) genotype and 2.2-fold for those with the Val/Val (TT) ? Ala/Ala (CC) genotypes [16]. An even higher significant risk of 4.3-fold for lung cancer associated the variant Val762Ala polymorphism among smokers with the Ala/Ala (CC) genotype [38]. This significant finding among Chinese smokers agrees with our meta-analysis results in all genetic models (1.7–2.5-fold). However, in the gene–gene interaction analysis, Zhang et al.’s [38] significant 5.9-fold finding agrees with our results only in the recessive (1.92-fold) and allelic models (1.39-fold) which were based on five and four studies, respectively. While the recessive model results were obtained under heterogeneous conditions (Pheterogeneity = 0.001) with broad confidence intervals indicating low precision, the allele contrast findings were homogeneous (Pheterogeneity = 0.15) with narrow confidence intervals indicating high precision. The apparent discrepancy between Chinese and American smokers may initially be characterized in the following manner. Increased risk among Chinese smokers were composed of studies whose heterogeneity was virtually absent (I2 = 0–14 %), but substantial (I2 = 74 %) among American smokers with significant decreased risk effects in the allele contrast. In fact, the contrasting narrow and broad confidence interval effects in all genetic models indicate that precision was high among Chinese smokers but low among American smokers, respectively. Furthermore, two studies of the American smokers effects admit to inherent and uncontrolled biases as well as low penetrance of the common SNPs selected that likely render results by chance [16, 17]. It maybe then that the single gene approach is limiting to studying the global functioning of the BER pathway [16]. XRCC1 and PARP1 play a joint role in repairing oxidative damage and single-strand breaks during base excision repair. XRCC1 encodes a scaffolding protein that is pivotal in base excision repair by recruiting PARP1, DNA ligase III and DNA polymerase-b at the site of DNA
Mol Biol Rep
damage [33]. The 1.4 and 1.9-fold respective joint allelic and recessive effects must be interpreted with caution because of deviation from the HWE of the controls in all but one [31] of the component studies. Nevertheless, the combined effects between XRCC1 and PARP1 may be explained by a reduction of interaction between Val762Ala and Arg399Gln. The mechanism of this attenuated interaction involves lowering autopoly(ADP-ribosyl)ation of PARP1 in the central automodification domain [33], which impairs base excision repair capacity [63]. Given the preferential interaction of XRCC1 with autoribosylated PARP1 [63], a reduced interaction between these two proteins is not unexpected. Individuals homozygous for both variant genotypes were found to have a 6-fold (lung and stomach cancers) to 8-fold (esophageal carcinoma) increased risk compared with those having other genotypes [31, 33, 38] Clinical implications Given greater susceptibility of low enzymatic activity Ala/ Ala–CC individuals to cancer development, and that the
Table 3 Summary of PARP1 effects across genotypes differentiating associations between groups PARP1 genotypes CC Ala/ Ala
CT Ala/Val
TT Val/ Val
Lower
Medium
Higher
Characteristics PARP1 enzymatic activity DNA damage repair capacity
Higher
Medium
Lower
Apoptotic ability
Lower
Medium
Higher
Necrotic ability
Higher
Medium
Lower
Asians
Higher
Medium
Lower
Caucasians
Lower
Medium
Higher
Frequency by ethnicity/race
role of PARP1 inhibitor is to inhibit its activity, treatment decision may require genotype state of this polymorphism in patients [13]. For example, individuals homozygous for the common genotype, Val/Val (TT) may require elevated doses of inhibitors since this allele is associated with high enzymatic activity. Therapeutic application of PARP1 inhibitors necessitates identifying patients most likely to benefit from this anti-cancer strategy [64] which may be addressed from the perspective of differential treatment according to distribution of the Val762Ala polymorphism and or relevant low PARP1 activity in patients [13]. Chinese patients have a high frequency (45 %) of the low enzymatic activity Ala (C) allele [13] and our meta-analysis shows increased risk exacerbated by behaviour (smoking) and tumor type (gastric carcinoma). By contrast, Caucasian patients with 20 % allele distribution [13], showed in our study that this ethnic group had reduced risk, particularly from brain tumors. Given this differential enzymatic activity between the two ethnic/racial groups, treating a patient with a PARP1 inhibitor may consider the status of this polymorphism as a better guide for dosage dependent treatment. This polymorphism, however, may more likely be relevant in gastric carcinoma among Asians given its higher incidence rates (0.01 %) compared to Caucasians (0.008 %). By contrast, incidence rates for brain tumors are orders of magnitude lower where Caucasians have 0.007 % against 0.004 % in Asians (http:// seer.cancer.gov/statfacts/html). Because we were limited only to two ethnic subgroups as well as number of studies in other cancer types, future investigations of this polymorphism warrant close attention to design, methodological features and expansion into other less studied population groups as well as tumor types. Well designed epidemiological studies based on sample sizes commensurate with the detection of small genotypic risks as well as multicentric case–control studies would help illuminate the complex landscape of base excision repair and cancer risk
Risk for cancer type Risk for gastric carcinoma (Chinese)
Higher
Medium
Lower
Lower
Medium
Higher
Chinese
Higher
Medium
Lower
American
Lower
Medium
Higher
Higher
Medium
Lower
PARP1 Inhibitor
Lower
Medium
Higher
3-Aminobenzamide
Same
Same
Same
Risk for brain tumors (American) Risk for smokers
Gene–gene interaction with XRCC1 PARP1 treatment
PARP1 inhibition treatment decisions may need to consider these genotypic differences across ethnic subgroups as influenced by environmental and joint-effects variables
Conclusion In conclusion, results of this meta-analysis demonstrate that the variant allele (Ala) of the PARP1 Val762Ala polymorphism appears to elicit both racial and tissue-specific genetic effects in carcinogenesis. Our findings underscore susceptibility of the Chinese population and protection in Americans. Gene–environment interaction indicates a multiplicative joint effect on susceptibility of the Chinese where smokers exhibited more pronounced associations in all genotypic and allelic comparisons. Table 3 summarizes our observations that could be useful in assessing susceptibility of different populations to
123
Mol Biol Rep
cancer and contribute to better predictions of risk. To the best of our knowledge, there is no published study exploring the Val762Ala polymorphism of PARP1 in patients due for inhibition therapy. PARP1 inhibition currently in human clinical trials may well be served to consider racial and tissuespecific-based treatment decisions. Acknowledgments This work was supported by the Canadian Breast Cancer Foundation (CBCF) grant of Hilmi Ozcelik and by the Philippine Council for Health Research and Development grant-in-aid of Noel Pabalan. We thank Stephen Pisle, NCI Post-Baccalaureate Fellow of the National Institute of Health, USA. Competing interests peting interests
The authors declare that they have no com-
References 1. Masutani M, Nakagama H, Sugimura T (2005) Poly(ADP-ribosyl)ation in relation to cancer and autoimmune disease. Cell Mol Life Sci 62:769–783 2. Nikitin AG, Chudakova DA, Strokov IA, Bursa TR, Chistiakov DA, Nosikov VV (2008) Leu54Phe and Val762Ala polymorphisms in the poly(ADP-ribose)polymerase-1 gene are associated with diabetic polyneuropathy in russian type 1 diabetic patients. Diabetes Res Clin Pract 79:446–452 3. von Lukowicz T, Hassa PO, Lohmann C, Boren J, Braunersreuther V, Mach F, Odermatt B, Gersbach M, Camici GG, Stahli BE, Tanner FC, Hottiger MO, Luscher TF, Matter CM (2008) PARP1 is required for adhesion molecule expression in atherogenesis. Cardiovasc Res 78:158–166 4. Virag L, Szabo C (2002) The therapeutic potential of poly(ADPribose) polymerase inhibitors. Pharmacol Rev 54:375–429 5. Berndt SI, Huang WY, Fallin MD, Helzlsouer KJ, Platz EA, Weissfeld JL, Church TR, Welch R, Chanock SJ, Hayes RB (2007) Genetic variation in base excision repair genes and the prevalence of advanced colorectal adenoma. Cancer Res 67: 1395–1404 6. Rajaraman P, Hutchinson A, Wichner S, Black PM, Fine HA, Loeffler JS, Selker RG, Shapiro WR, Rothman N, Linet MS, Inskip PD (2010) DNA repair gene polymorphisms and risk of adult meningioma, glioma, and acoustic neuroma. Neuro Oncol 12:37–48 7. Shiobara M, Miyazaki M, Ito H, Togawa A, Nakajima N, Nomura F, Morinaga N, Noda M (2001) Enhanced polyadenosine diphosphate-ribosylation in cirrhotic liver and carcinoma tissues in patients with hepatocellular carcinoma. J Gastroenterol Hepatol 16:338–344 8. Eguchi Y, Shimizu S, Tsujimoto Y (1997) Intracellular ATP levels determine cell death fate by apoptosis or necrosis. Cancer Res 57:1835–1840 9. Leist M, Single B, Castoldi AF, Kuhnle S, Nicotera P (1997) Intracellular adenosine triphosphate (ATP) concentration: a switch in the decision between apoptosis and necrosis. J Exp Med 185:1481–1486 10. Cottet F, Blanche H, Verasdonck P, Le Gall I, Schachter F, Burkle A, Muiras ML (2000) New polymorphisms in the human poly(ADP-ribose) polymerase-1 coding sequence: lack of association with longevity or with increased cellular poly(ADPribosyl)ation capacity. J Mol Med 78:431–440
123
11. Simonin F, Hofferer L, Panzeter PL, Muller S, de Murcia G, Althaus FR (1993) The carboxyl-terminal domain of human poly(ADP-ribose) polymerase. Overproduction in escherichia coli, large scale purification, and characterization. J Biol Chem 268:13454–13461 12. Lockett KL, Hall MC, Xu J, Zheng SL, Berwick M, Chuang SC, Clark PE, Cramer SD, Lohman K, Hu JJ (2004) The ADPRT V762A genetic variant contributes to prostate cancer susceptibility and deficient enzyme function. Cancer Res 64:6344–6348 13. Wang XG, Wang ZQ, Tong WM, Shen Y (2007) PARP1 Val762Ala polymorphism reduces enzymatic activity. Biochem Biophys Res Commun 354:122–126 14. Zaremba T, Ketzer P, Cole M, Coulthard S, Plummer ER, Curtin NJ (2009) Poly(ADP-ribose) polymerase-1 polymorphisms, expression and activity in selected human tumour cell lines. Br J Cancer 101:256–262 15. Gao R, Price DK, Dahut WL, Steinberg SM, Reed E, Figg WD (2010) Genetic polymorphisms in xrcc1 associated with radiation therapy in prostate cancer. Cancer Biol Ther 10:13–18 16. Huang M, Dinney CP, Lin X, Lin J, Grossman HB, Wu X (2007) High-order interactions among genetic variants in DNA base excision repair pathway genes and smoking in bladder cancer susceptibility. Cancer Epidemiol Biomarkers Prev 16:84–91 17. Li C, Hu Z, Lu J, Liu Z, Wang LE, El-Naggar AK, Sturgis EM, Spitz MR, Wei Q (2007) Genetic polymorphisms in DNA baseexcision repair genes ADPRT, XRCC1, and APE1 and the risk of squamous cell carcinoma of the head and neck. Cancer 110: 867–875 18. Li C, Liu Z, Wang LE, Strom SS, Lee JE, Gershenwald JE, Ross MI, Mansfield PF, Cormier JN, Prieto VG, Duvic M, Grimm EA, Wei Q (2006) Genetic variants of the ADPRT, XRCC1 and APE1 genes and risk of cutaneous melanoma. Carcinogenesis 27:1894–1901 19. Liu Y, Scheurer ME, El-Zein R, Cao Y, Do KA, Gilbert M, Aldape KD, Wei Q, Etzel C, Bondy ML (2009) Association and interactions between DNA repair gene polymorphisms and adult glioma. Cancer Epidemiol Biomarkers Prev 18:204–214 20. McKean-Cowdin R, Barnholtz-Sloan J, Inskip PD, Ruder AM, Butler M, Rajaraman P, Razavi P, Patoka J, Wiencke JK, Bondy ML, Wrensch M (2009) Associations between polymorphisms in DNA repair genes and glioblastoma. Cancer Epidemiol Biomarkers Prev 18(4):1118–1126 21. Shen M, Zheng T, Lan Q, Zhang Y, Zahm SH, Wang SS, Holford TR, Leaderer B, Yeager M, Welch R, Kang D, Boyle P, Zhang B, Zou K, Zhu Y, Chanock S, Rothman N (2006) Polymorphisms in DNA repair genes and risk of non-hodgkin lymphoma among women in connecticut. Hum Genet 119:659–668 22. Smith TR, Levine EA, Freimanis RI, Akman SA, Allen GO, Hoang KN, Liu-Mares W, Hu JJ (2008) Polygenic model of DNA-repair genetic polymorphisms in human breast cancer risk. Carcinogenesis 29(11):2132–2138 23. Wu X, Gu J, Grossman HB, Amos CI, Etzel C, Huang M, Zhang Q, Millikan RE, Lerner S, Dinney CP, Spitz MR (2006) Bladder cancer predisposition: a multigenic approach to DNA-repair and cell-cyclecontrol genes. Am J Hum Genet 78:464–479 24. Zhang Y, Newcomb PA, Egan KM, Titus-Ernstoff L, Chanock S, Welch R, Brinton LA, Lissowska J, Bardin-Mikolajczak A, Peplonska B, Szeszenia-Dabrowska N, Zatonski W, Garcia-Closas M (2006) Genetic polymorphisms in base-excision repair pathway genes and risk of breast cancer. Cancer Epidemiol Biomarkers Prev 15:353–358 25. Cao WH, Wang X, Frappart L, Rigal D, Wang ZQ, Shen Y, Tong WM (2007) Analysis of genetic variants of the poly(ADP-ribose) polymerase-1 gene in breast cancer in French patients. Mutat Res 632:20–28
Mol Biol Rep 26. Figueroa JD, Malats N, Real FX, Silverman D, Kogevinas M, Chanock S, Welch R, Dosemeci M, Tardon A, Serra C, Carrato A, Garcia-Closas R, Castano-Vinyals G, Rothman N, GarciaClosas M (2007) Genetic variation in the base excision repair pathway and bladder cancer risk. Hum Genet 121:233–242 27. Landi S, Gemignani F, Canzian F, Gaborieau V, Barale R, Landi D, Szeszenia-Dabrowska N, Zaridze D, Lissowska J, Rudnai P, Fabianova E, Mates D, Foretova L, Janout V, Bencko V, GioiaPatricola L, Hall J, Boffetta P, Hung RJ, Brennan P (2006) DNA repair and cell cycle control genes and the risk of young-onset lung cancer. Cancer Res 66:11062–11069 28. Durocher F, Labrie Y, Ouellette G, Simard J (2007) Genetic sequence variations and ADPRT haplotype analysis in French Canadian families with high risk of breast cancer. J Hum Genet 52:963–977 29. Yosunkaya E, Kucukyuruk B, Onaran I, Gurel CB, Uzan M, Kanigur-Sultuybek G (2010) Glioma risk associates with polymorphisms of DNA repair genes, XRCC1 and PARP1. Br J Neurosurg 24:561–565 30. Chiang FY, Wu CW, Hsiao PJ, Kuo WR, Lee KW, Lin JC, Liao YC, Juo SH (2008) Association between polymorphisms in DNA base excision repair genes XRCC1, APE1, and ADPRT and differentiated thyroid carcinoma. Clin Cancer Res 14:5919–5924 31. Hao B, Wang H, Zhou K, Li Y, Chen X, Zhou G, Zhu Y, Miao X, Tan W, Wei Q, Lin D, He F (2004) Identification of genetic variants in base excision repair pathway and their associations with risk of esophageal squamous cell carcinoma. Cancer Res 64:4378–4384 32. Kang S-LL, Yu-Min; He; Wen-Ting; Liu Tao; Li Xun; Zhou, Wen-Ce; Yi, Jian-Fen; Zeng, Xiang-Ting (2010) Association between parp-1 polymorphisms and susceptibility to gastric cancer. World J Gastroenterol 18:1434–1441 33. Miao X, Zhang X, Zhang L, Guo Y, Hao B, Tan W, He F, Lin D (2006) Adenosine diphosphate ribosyl transferase and x-ray repair cross-complementing 1 polymorphisms in gastric cardia cancer. Gastroenterology 131:420–427 34. Stern MC, Conti DV, Siegmund KD, Corral R, Yuan JM, Koh WP, Yu MC (2007) DNA repair single-nucleotide polymorphisms in colorectal cancer and their role as modifiers of the effect of cigarette smoking and alcohol in the Singapore Chinese health study. Cancer Epidemiol Biomarkers Prev 16:2363–2372 35. Wang M, Qin C, Zhu J, Yuan L, Fu G, Zhang Z, Yin C (2010) Genetic variants of xrcc1, ape1, and adprt genes and risk of bladder cancer. DNA Cell Biol 29(6):303–311 36. Zhai X, Liu J, Hu Z, Wang S, Qing J, Wang X, Jin G, Gao J, Wang X, Shen H (2006) Polymorphisms of ADPRT Val762Ala and XRCC1 Arg399Glu and risk of breast cancer in Chinese women: a case control analysis. Oncol Rep 15:247–252 37. Zhang Q, Li Y, Li X, Zhou W, Shi B, Chen H, Yuan W (2009) Parp-1 val762ala polymorphism, caga(?) h. Pylori infection and risk for gastric cancer in han chinese population. Mol Biol Rep 36(6):1461–1467 38. Zhang X, Miao X, Liang G, Hao B, Wang Y, Tan W, Li Y, Guo Y, He F, Wei Q, Lin D (2005) Polymorphisms in DNA base excision repair genes ADPRT and XRCC1 and risk of lung cancer. Cancer Res 65:722–726 39. Zhang Z, Miao XP, Tan W, Guo YL, Zhang XM, Lin DX (2006) Correlation of genetic polymorphisms in DNA repair genes ADPRT and XRCC1 to risk of gastric cancer. Ai Zheng 25:7–10 40. Jin XM, Kim HN, Lee IK, Park KS, Kim HJ, Choi JS, Juhng SW, Choi C (2010) PARP-1 Val762Ala polymorphism is associated with reduced risk of non-hodgkin lymphoma in Korean males. BMC Med Genet 11:38 41. Mantel N, Haenszel W (1959) Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 22:719–748
42. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials. Control Clin Trials 7:177–188 43. Lau J, Ioannidis JP, Schmid CH (1997) Quantitative synthesis in systematic reviews. Ann Intern Med 127:820–826 44. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Measuring inconsistency in meta-analyses. BMJ 327:557–560 45. Higgins JP, Thompson SG (2002) Quantifying heterogeneity in a meta-analysis. Stat Med 21:1539–1558 46. Galbraith RF (1988) A note on graphical presentation of estimated odds ratios from several clinical trials. Stat Med 7:889–894 47. Pabalan N, Bapat B, Sung L, Jarjanazi H, Francisco-Pabalan O, Ozcelik H (2008) Cyclin D1 Pro241Pro (CCND1-G870A) polymorphism is associated with increased cancer risk in human populations: a meta-analysis. Cancer Epidemiol Biomarkers Prev 17:2773–2781 48. Egger M, Davey Smith G, Schneider M, Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ (Clinical research ed) 315:629–634 49. Begg CB, Mazumdar M (1994) Operating characteristics of a rank correlation test for publication bias. Biometrics 50:1088–1101 50. Gao R, Price DK, Sissung T, Reed E, Figg WD (2008) Ethnic disparities in Americans of European descent versus Americans of African descent related to polymorphic ERCC1, ERCC2, XRCC1, and PARP1. Mol Cancer Ther 7:1246–1250 51. Dedes KJ, Wetterskog D, Mendes-Pereira AM, Natrajan R, Lambros MB, Geyer FC, Vatcheva R, Savage K, Mackay A, Lord CJ, Ashworth A, Reis-Filho JS (2010) PTEN deficiency in endometrioid endometrial adenocarcinomas predicts sensitivity to PARP inhibitors. Sci Transl Med 2:53ra75 52. Mendes-Pereira AM, Martin SA, Brough R, McCarthy A, Taylor JR, Kim JS, Waldman T, Lord CJ, Ashworth A (2009) Synthetic lethal targeting of PTEN mutant cells with PARP inhibitors. EMBO Mol Med 1:315–322 53. Bhaskara VK, Panigrahi M, Challa S, Babu PP (2005) Comparative status of activated ERK1/2 and PARP cleavage in human gliomas. Neuropathology 25:48–53 54. Bhaskara VK, Challa S, Panigrahi M, Babu PP (2009) Differential PARP cleavage: an indication for existence of multiple forms of cell death in human gliomas. Neurol India 57:264–268 55. Schaid DJ, Jacobsen SJ (1999) Biased tests of association: comparisons of allele frequencies when departing from HardyWeinberg proportions. Am J Epidemiol 149:706–711 56. Xiuti S, Ren M, Shangyou Z (2002) Analyses of gastric cancer mortality in china from 1990 to 1992. Chin J Oncol 24:4–8 57. Roder DM (2002) The epidemiology of gastric cancer. Gastric Cancer 5(Suppl 1):5–11 58. Figueiredo C, Quint W, Nouhan N, van den Munckhof H, Herbrink P, Scherpenisse J et al (2001) Assessment of helicobacter pylori vacA and cagA genotypes and host serological response. J Clin Microbiol 39:1339–1344 59. Zhang L, Blot WJ, You WC, Chang YS, Kneller RW, Jin ML, Li JY, Zhao L, Liu WD, Zhang JS, Ma JL, Samloff IM, Correa P, Blaser MJ, Xu GW, Fraumeni JF Jr (1996) Helicobacter pylori antibodies in relation to precancerous gastric lesions in a high-risk Chinese population. Cancer Epidemiol Biomarkers Prev 5:627–630 60. Li WQ, Zhang L, Ma JL, Zhang Y, Li JY, Pan KF, You WC (2009) Association between genetic polymorphisms of DNA base excision repair genes and evolution of precancerous gastric lesions in a Chinese population. Carcinogenesis 30:500–505 61. Nossa CW, Jain P, Tamilselvam B, Gupta VR, Chen LF, Schreiber V, Desnoyers S, Blanke SR (2009) Activation of the abundant nuclear factor poly(ADP-ribose) polymerase-1 by helicobacter pylori. Proc Natl Acad Sci USA 106:19998–20003 62. Hecht SS (1999) Tobacco smoke carcinogens and lung cancer. J Natl Cancer Inst 91:1194–1210
123
Mol Biol Rep 63. Masson M, Niedergang C, Schreiber V, Muller S, Menissier-de Murcia J, de Murcia G (1998) XRCC1 is specifically associated with poly(ADP-ribose) polymerase and negatively regulates its activity following DNA damage. Mol Cell Biol 18:3563–3571
123
64. Moeller BJ, Arap W, Pasqualini R (2010) Targeting synthetic lethality in DNA damage repair pathways as an anti-cancer strategy. Curr Drug Targets 1(10):1336–1340
