Association between NER Pathway Gene Polymorphisms and Wilms Tumor Risk.

Nucleotide excision repair (NER) is an essential mechanism of the body to defend against exogenous carcinogen-induced DNA damage. Defects in NER may impair DNA repair capacity and, therefore, increase genome instability and cancer susceptibility. To explore genetic predispositions to Wilms tumor, we conducted a case-control study totaling 145 neuroblastoma cases and 531 healthy controls. We systematically selected 19 potentially functional SNPs in six key genes within the NER pathway (ERCC1, XPA, XPC, XPD, XPF, and XPG). The odds ratio (OR) and 95% confidence interval (CI) were calculated to measure the strength of associations. We identified significant associations between two XPD SNPs and Wilms tumor risk. The XPD rs3810366 polymorphism significantly enhanced Wilms tumor risk (dominant model: adjusted OR = 2.12, 95% CI = 1.26-3.57). Likewise, XPD rs238406 conferred a significantly increased risk for the disease (dominant model: adjusted OR = 2.30, 95% CI = 1.40-3.80; recessive model: adjusted OR = 1.64, 95% CI = 1.11-2.44). Moreover, online expression quantitative trait locus (eQTL) analysis demonstrated that these two polymorphisms significantly affected XPD gene expression in transformed fibroblast cells. Our study provides evidence of the association between the two XPD polymorphisms and Wilms tumor risk. However, these findings warrant validation in larger studies.

Introduction

Wilms tumor (WT) is a complex childhood embryonal tumor of the kidney, affecting about one child per 10,000 worldwide under 15 years of age. WT is derived from the embryonal nephric mesenchyme. It is characterized by the copresence of the whole spectrum of nephrogenic differentiation, from primitive blastema to mature epithelial and stromal elements, which normally appear in the different developmental stages of the kidney. Compared with other cancers, treatment of WT has been quite successful. Several clinical trials have achieved overall survival rates of over 90%, carried out by the Children’s Oncology Group (COG), the International Society of Pediatric Oncology (SIOP), and others. The encouraging outcomes of patients in clinical trials largely benefit from personalized therapy founded on clinical (e.g., age, tumor size, volume, and response to chemotherapy) and genetic (e.g., loss of heterozygosity [LOH] at chromosomes 1p and 16q) risk factors. Despite the favorable prognosis of WT, it should be noted that approximately 25% survivors experience severe chronic disorders. Moreover, 25% WT patients have high-risk WT (i.e., poor histologic and molecular characteristics, tumors on both sides, and relapsed disease), and their survival rates are below 90%. Therefore, it is indispensable to refine treatment modalities to reduce sequelae and complications and to develop novel therapies for high-risk WT.

WT is a genetically heterogeneous and complex disease. Well established genetic risk factors include mutations in Wnt/β-catenin pathway-related Wilms tumor gene 1 (WT1), catenin beta 1 (CTNNB1), and Wilms Tumor gene on the X chromosome (WTX), which are involved in the etiology of approximately one-third of WT. Moreover, LOH of chromosome 16q, gain of chromosome 1q, and microRNA (miRNA)-processing gene mutations are also frequently observed in WT.3, 4 However, the alterations of additional genes that contribute to Wilms tumorigenesis still warrant intensive investigations.

The human genome is frequently subjected to damage resulting from both environmental agents (e.g., UV light and inhaled cigarette smoke) and endogenous weak mutagens (reactive oxygen species and metabolites like alkylating agents). To maintain genome integrity, several DNA repair mechanisms continuously inspect chromosomes to fix damaged nucleotide residues induced by the great variety of DNA-damaging agents. These mechanisms include base excision repair (BER), nucleotide excision repair (NER), and mismatch repair (MMR), which are responsible for distinct forms of DNA damage.5, 6 NER primarily eliminates bulky adducts arising from exposure to environmental agents. Many core proteins are involved in NER, among which хeroderma pigmentosum A (XPA) to XPG were identified from хeroderma pigmentosum. Moreover, excision repair cross-complementation group 1 (ERCC1), replication protein A (RPA), RAD23 homolog A (RAD23A), and RAD23 homolog B (RAD23B) also participate in NER. SNPs in NER genes have been linked to various cancer types, including lung, bladder, skin, breast, prostate, and head and neck cancers.8, 9, 10, 11, 12, 13, 14, 15, 16 Accumulating evidence has indicated that some SNPs in DNA repair genes or their regulatory elements can induce phenotypical alterations, affecting DNA repair capacity and promoting cancer initiation and development. In this study, we genotyped 19 potential functional NER pathway gene SNPs in 145 WT cases and 531 controls to intensively investigate their association with WT risk.

Results

Characteristics of the Study Population

The current study was composed of 145 cases and 531 cancer-free controls (average age: 26.17 ± 21.48 months versus 29.73 ± 24.86 months). No statistically significant differences in age (p = 0.725) and gender (p = 0.956) were detected between the case and control groups. Patients were staged by following the National Wilms Tumor Study-5 (NWTS-5) criteria. Specifically, 4 (2.76%), 49 (33.79%), 50 (34.48%), and 33 (22.76%) individuals were determined to bear clinical stage I, II, III, and IV tumors, respectively. It should be noted that evaluation and staging failed in 9 cases (6.21%) because of inadequate information (Table S1).

Associations between NER Pathway Gene SNPs and WT Susceptibility

In total, 19 SNPs in the NER pathway genes were genotyped in 145 cases and 531 controls. Specifically, there were 3, 2, 5, 3, 1, and 5 SNPs in the ERCC1, XPA, XPC, XPD, XPF, and XPG genes, respectively (Table 1). Of them, two SNPs in the XPD gene were found to significantly modify WT risk (Figures 1 and 2). The XPD rs3810366 polymorphism was found to significantly increase the risk of developing WT (recessive model: adjusted odds ratio [OR] = 2.12, 95% confidence interval [CI] = 1.26–3.57), whereas XPD rs238406 was significantly associated with an increased risk of the disease (dominant model: adjusted OR = 2.30, 95% CI = 1.40–3.80; recessive model: adjusted OR = 1.64, 95% CI = 1.11–2.44).

Table 1

Association of Polymorphisms in Nucleotide Excision Repair Pathway Genes with Wilms Tumor Susceptibility

Gene	SNP	Allele		Case (n = 145)			Control (n = 531)			Adjusted OR (95% CI)a	pa	Adjusted OR (95% CI)b	pb	HWE
		A	B	AA	AB	BB	AA	AB	BB
ERCC1	rs2298881	C	A	62	63	20	213	231	87	0.89 (0.62–1.30)	0.557	0.85 (0.50–1.44)	0.544	0.072
ERCC1	rs3212986	C	A	54	72	19	236	231	64	1.35 (0.93–1.98)	0.118	1.09 (0.63–1.89)	0.758	0.519
ERCC1	rs11615	G	A	86	51	8	302	185	44	0.90 (0.62–1.31)	0.576	0.64 (0.29–1.39)	0.261	0.043
XPA	rs1800975	T	C	30	72	42	124	281	126	1.14 (0.73–1.79)	0.566	1.31 (0.87–1.98)	0.203	0.178
XPA	rs3176752	G	T	106	33	5	408	115	8	1.19 (0.78–1.82)	0.419	2.55 (0.81–8.00)	0.110	0.975
XPC	rs2228001	A	C	52	73	20	218	245	68	1.26 (0.86–1.84)	0.238	1.06 (0.62–1.82)	0.827	0.948
XPC	rs2228000	C	T	64	59	22	205	250	76	0.80 (0.55–1.16)	0.232	1.05 (0.63–1.76)	0.851	0.988
XPC	rs2607775	C	G	134	11	0	477	54	0	0.71 (0.36-1.40)	0.328	–	–	0.217
XPC	rs1870134	G	C	92	51	2	339	166	26	1.04 (0.71–1.52)	0.862	0.27 (0.06–1.17)	0.080	0.335
XPC	rs2229090	G	C	50	73	20	191	255	85	1.04 (0.70–1.53)	0.853	0.86 (0.51–1.46)	0.580	0.994
XPD	rs3810366	C	G	19	76	50	128	248	155	2.12 (1.26–3.57)c	0.005c	1.26 (0.85–1.87)	0.242	0.143
XPD	rs238406	G	T	21	73	51	149	250	132	2.30 (1.40–3.80)c	0.001c	1.64 (1.11–2.44)c	0.014c	0.186
XPD	rs13181	T	G	128	16	1	462	65	4	0.87 (0.50–1.54)	0.639	0.86 (0.10–7.80)	0.895	0.312
XPF	rs2276466	C	G	76	51	12	301	201	29	1.08 (0.74–1.57)	0.696	1.60 (0.79–3.24)	0.188	0.543
XPG	rs2094258	C	T	61	62	19	203	254	74	0.81 (0.55–1.17)	0.260	0.94 (0.55–1.62)	0.819	0.701
XPG	rs751402	C	T	50	72	20	208	241	82	1.20 (0.81–1.76)	0.366	0.93 (0.54–1.58)	0.776	0.380
XPG	rs2296147	T	C	96	41	5	343	170	18	0.87 (0.59–1.29)	0.492	1.06 (0.39–2.92)	0.910	0.583
XPG	rs1047768	T	C	76	58	8	307	198	26	1.19 (0.82–1.72)	0.371	1.20 (0.53–2.73)	0.659	0.409
XPG	rs873601	G	A	40	67	35	137	270	124	0.88 (0.58–1.33)	0.547	1.04 (0.67–1.61)	0.859	0.686

HWE, Hardy-Weinberg equilibrium.

Adjusted for age and gender for the dominant model.

Adjusted for age and gender for the recessive model.

The results if the 95% CI excluded 1 or p < 0.05.

Figure 1

Forest Plot for the Association between NER Gene Polymorphisms and Wilms Tumor Susceptibility under the Dominant Model (AB/BB versus AA)

For each SNP, the estimates of odds ratio and its 95% confidence interval are plotted with a box and a horizontal line.

Figure 2

Forest Plot for the Association between NER Gene Polymorphisms and Wilms Tumor Susceptibility under the Recessive Model (BB versus AA/AB)

For each SNP, the estimates of odds ratio and its 95% confidence interval are plotted with a box and a horizontal line.

Stratified Analysis

Participants were further stratified by age, gender, and clinical stage (Table 2). Stratified analyses were performed for the two significant XPD polymorphisms (rs3810366 and rs238406) and the combined risk genotypes of the XPD gene. Intriguingly, the XPD rs3810366 polymorphism was shown to associate with WT risk in older participants (>18 months of age), males, and those with clinical stage I/II disease. Moreover, the association with the rs238406 polymorphism remained significant among all strata, except for females. A borderline significant association was found in females. We next evaluated the combined effects of XPD polymorphisms. Among the participants carrying 1–3 risk genotypes, males and older children (>18) were at significantly elevated risk of WT compared with non-carriers. Children with 1–3 risk genotypes were more likely to develop stage I/II disease.

Table 2

Stratification Analysis of XPD Gene Variant Genotypes with Wilms Tumor Risk

Variables	rs3810366 (Case/Control)		AOR (95% CI)	pa	rs238406 (Case/Control)		AOR (95% CI)	pa	Risk Genotype (Case/Control)		AOR (95% CI)	pa
	CC	CG/GG			GG	GT/TT			0	1–3
Age (Months)
≤18	10/56	56/177	1.74 (0.83–3.64)	0.142	10/65	56/168	2.15 (1.03–4.46)b	0.041b	9/51	57/182	1.75 (0.81–3.79)	0.152
>18	9/72	70/226	2.48 (1.18–5.21)b	0.017b	11/84	68/214	2.43 (1.23–4.83)b	0.011b	9/65	70/233	2.18 (1.03–4.61)b	0.041b
Gender
Females	10/59	54/174	1.85 (0.89–3.87)	0.102	11/67	53/166	1.95 (0.96–3.96)	0.066	10/53	54/180	1.60 (0.76–3.35)	0.216
Males	9/69	72/229	2.38 (1.13–5.02)b	0.022b	10/82	71/216	2.69 (1.32–5.47)b	0.006b	8/63	73/235	2.42 (1.11–5.29)b	0.027b
Clinical Stages
I/II	6/128	47/403	2.56 (1.06–6.15)b	0.036b	7/149	46/382	2.58 (1.13–5.86)b	0.024b	5/116	48/415	2.72 (1.05–7.02)b	0.039b
III/IV	12/128	71/403	1.87 (0.98–3.56)	0.057	13/149	70/382	2.09 (1.12–3.90)b	0.020b	12/116	71/415	1.65 (0.86–3.14)	0.130

AOR, adjusted odds ratio.

Obtained in logistic regression models with adjustment for age and gender, omitting the corresponding stratification factor.

The results if the 95% CI excluded 1 or p < 0.05.

Haplotype Analysis

The effects of the haplotypes of the XPD gene were also explored. We found that the GTG haplotype was significantly associated with an increased WT risk compared with the CCG reference haplotype in the order of rs3810366, rs238406, and rs13181 (Table 3).

Table 3

The Frequency of Inferred Haplotypes of the XPD Gene and Wilms Tumor Risk

Haplotypesa	Cases (n = 290)	Controls (n = 1,062)	OR (95% CI)	p	AOR (95% CI)b	pb
CGG	10 (3.45)	48 (4.52)	1.00	–	1.00	–
CGT	100 (34.48)	456 (42.94)	1.05 (0.52–2.15)	0.888	1.08 (0.53–2.21)	0.658
CTG	0 (0.00)	0 (0.00)	–	–	–	–
CTT	4 (1.38)	0 (0.00)	–	–	–	–
GGG	2 (0.69)	20 (1.88)	0.48 (0.10–2.39)	0.370	0.47 (0.09–2.34)	0.355
GGT	3 (1.03)	24 (2.26)	0.60 (0.15–2.39)	0.468	0.63 (0.16–2.49)	0.505
GTG	6 (2.07)	5 (0.47)	5.76 (1.47–22.63)c	0.012c	6.10 (1.54–24.09)c	0.010c
GTT	165 (56.90)	509 (47.93)	1.56 (0.77–3.15)	0.218	1.59 (0.79–3.21)	0.198

The haplotype order was rs3810366, rs238406, and rs13181.

Obtained in logistic regression models with adjustment for age and gender.

The results if the 95% CI excluded 1 or p < 0.05.

Expression Quantitative Trait Loci

We further explored biological effects of the two significant SNPs in the XPD gene expression by investigating a public database, GTEx portal. We observed that genotypes of both SNPs were significantly associated with XPD gene expression in transformed fibroblasts cells (Figure 3).

Figure 3

Analysis of the rs3810366 and rs238406 Polymorphisms in the XPD Gene in Transformed Fibroblast Cells

Shown is the eQTL Analysis for the (A) rs3810366 and (B) rs238406 polymorphisms in the XPD gene in transformed fibroblast cells (GTEx portal).

Discussion

Several lines of evidence implicate genetic variants in WT: ethnic differences in WT susceptibility are more prominent than geographic differences, WT occurs in both sporadic and familial forms, and several syndromes, harboring mutations in WT1 or epigenetic defects at 11p15.3, had a greatly increased risk of WT. A more comprehensive understanding of tumor biology would help us to make progress in prevention, therapy, and prognosis of this disease. A genome-wide association study (GWAS) has been performed in 757 cases and 1,879 controls, attempting to determine common WT-predisposing variants. Ten significant SNPs were further validated in two separate study populations from the United Kingdom (769 cases and 2,814 controls) and the United States (719 cases and 1,037 controls). Two loci showed significant association with WT susceptibility: 2p24 (rs3755132 and rs807624) and 11q14 (rs790356).

In this study, we explore the association of NER pathway gene SNPs and WT risk in 145 neuroblastoma cases and 531 healthy controls. With the same study population, we validated that SNPs in several genes were associated with WT, including the BARD1, TP53, LIN28B, LOM1, and HACE1 genes. Overall, 19 potentially functional SNPs in 6 key NER pathway genes (ERCC1, XPA, XPC, XPD, XPF, and XPG) were genotyped. The association of DNA repair gene SNPs with cancer susceptibility has been widely investigated worldwide. A number of SNPs within the NER pathway have been found to associate with the risk of various types of cancer in Chinese populations, including laryngeal cancer, pancreatic cancer, breast cancer, prostate cancer, gastric cancer, colorectal cancer, and hepatocellular cancer. Among six key NER genes, association with significantly increased WT risk was identified for two SNPs (rs3810366 and rs238406) in the XPD gene. The XPD gene encodes a 760-amino acid polypeptide of 87 kDa that participates in transcription-coupled NER. Defects in this gene have been known to be related to three different conditions: the cancer-prone syndrome xeroderma pigmentosum complementation group D, photosensitive trichothiodystrophy, and Cockayne syndrome. XPD rs238406 (R156R) was found to be marginally significantly associated with epidermal growth factor receptor (EGFR)-mutant lung adenocarcinoma. Romanowicz et al. and Michalska et al. identify this SNP as a risk factor for ovarian cancer and endometrial cancer, respectively, in Poland. In a Chinese population, rs238406 significantly increased esophageal squamous cell carcinoma susceptibility. However, studies of XPD rs3810366 (at promoter −114) are relatively few. Several studies showed no association between rs3810366 and cancer susceptibility in Taiwan.9, 12, 29, 30

Numerous SNPs have been proven to be functional. Most recently, an uncommon missense polymorphism, rs149418249 (c.C1520T, p.P507L) located in the TPP1 (alias ACD, gene ID 65057), a component of the shelterin complex, was reported to confer colorectal cancer susceptibility. This variant can cause telomere dysfunction by disrupting TPP1 interaction with TIN2. Moreover, an exome-wide analysis identified a variant (rs138478634) in CYP26B1 associated with esophageal squamous cell carcinoma. This SNP can modify the enzymatic activity of CYP26B1. Functional analysis was not performed for the significant SNPs in the current study. However, online expression quantitative trait loci (eQTL) analysis indicated that genotypes of both SNPs significantly correlated with XPD gene expression levels. An in vitro model indicated that XPD rs13181 (Lys751Gln) variants reduce the capacity of an XPD-deficient cell line (UV5) in repairing DNA damage induced by benzo[a]pyrene (B[a]P), an exogenous carcinogen. Moreover, previous studies have demonstrated that decreased expression of NER genes (e.g., ERCC1, XPB, XPG, and CSB) in peripheral blood lymphocytes is associated with an increased risk of squamous cell carcinoma of the head and neck.8, 34

Several limitations should be noted in this study. First, we performed this association study in a relatively small number of samples. It cannot be ruled out that chance may account for one or more of the associations in the present study. Therefore, replication of our findings is encouraged. Second, WT carcinogenesis resulted from complicated gene-environment interactions. If possible, environmental factors, including paternal exposure, should be included in the risk estimation. Third, although 19 SNPs were genotyped in this study, more SNPs within the NER pathway should be investigated. Finally, the information from the online eQTL analysis is only suggestive. Tissue samples should be used to confirm the results in the future.

In summary, among 19 genotyped SNPs, the XPD rs3810366 and rs238406 polymorphisms were shown to significantly enhance WT susceptibility. Online eQTL analysis suggested that these two SNPs might influence XPD gene expression. However, replication studies and functional analyses of SNPs are needed to validate our findings.

Materials and Methods

Study Population

The 145 cases in the study were individuals newly diagnosed with neuroblastoma from the Guangzhou Women and Children’s Medical Center.19, 20, 21, 22, 23, 35 All tumors were histopathologically confirmed. Blood samples were collected at the time of diagnosis. Specimens were annotated with information including age at diagnosis, sex, and disease stage based on the NWTS-5 criteria. The 531 cancer-free controls were selected from children visiting the same hospital for a regular physical examination during the same period of time.36, 37, 38, 39 Frequency matching was performed for cases and controls on age and sex. Exclusion criteria included other types of tumors, secondary or recurrent tumors, and previous chemotherapy or radiotherapy. Participants were limited to the ethnic Chinese Han population. Prior to sample collection, all participants or their guardians were required to sign written informed consent compliant with the Declaration of Helsinki. The study was authorized by the Institutional Review Board of Guangzhou Women and Children’s Medical Center.

SNP Selection and Genotyping

SNPs were retrieved from the dbSNP database (https://www.ncbi.nlm.nih.gov/projects/SNP). We finally included 19 potentially functional SNPs in the core genes in the NER pathway that fit the following selection criteria: location of candidate SNPs were limited to the 5′ UTR, upstream promoter region, coding region, as well as the 3′ UTR of genes; a minor allele frequency of less than 5% in Chinese Han populations; and lack of linkage disequilibrium (R2 < 0.8) between each SNP pair. Potential functions of SNPs were predicted using SNPinfo (https://snpinfo.niehs.nih.gov/snpinfo/snpfunc.html); candidate SNPs should be able to modify the function of transcription factor binding sites or microRNA binding sites. Eventually, the following polymorphisms were included: ERCC1 (rs2298881 C > A, rs11615 G > A, rs3212986 C > A), XPA (rs1800975 G > A, rs3176752 C > A), XPC (rs2228001 A > C, rs2228000 C > T, rs2607775 C > G, rs1870134 G > C and rs2229090 G > C), XPD (rs3810366 C > G, rs238406 G > T, rs13181 T > G), XPF rs2276466 C > G, and XPG (rs2094258 C > T, rs751402 C > T, rs2296147 T > C, rs1047768 T > C, and rs873601 G > A).

DNA was collected from peripheral blood samples of participants with the QIAamp DNA Blood Mini Kit (QIAGEN, Valencia, CA). We carried out genotyping on the platform of the TaqMan real-time PCR method on a 7900 Sequence Detection System (Applied Biosystems, Foster City, CA). The technique details were described previously.40, 41, 42 Quality control was strictly executed; four duplicate positive controls and four negative controls (omitting DNA template) were loaded, along with samples in each of the 384-well plates. Furthermore, 10% of the tested samples were randomly picked and genotyped for a second time.43, 44, 45, 46 We observed an overall genotype concordance rate of 100% for each SNP in the subsequent genotyping.

Genotype-Phenotype Association

eQTL are regions of the genome containing DNA sequence variants that influence the expression level of one or more genes. We further explored the effects of the two significant SNPs on XPD gene expression by investigating a public database, GTEx portal (https://www.gtexportal.org/). The data in transformed fibroblasts were described previously.

Statistics

The Hardy-Weinberg equilibrium (HWE) was determined by using the goodness-of-fit χ2 test in control subjects. Student’s t test was used to compare the differences in age between cases and controls. χ2 tests were used to evaluate differences in the categorical variables between cases and controls, including sex and distributions of allele frequencies. Multivariate logistic regression analysis was performed. ORs and 95% CIs were computed to determine the strength of the association between SNPs and WT risk. We further carried out multivariate analysis using an unconditional logistic regression model to calculate ORs, with adjustment for age and sex. All statistical analyses were accomplished using version 9.1 SAS software (SAS Institute, Cary, NC). Two-sided p < 0.05 was adopted as a criterion of significance.

Author Contributions

J.Z., W.F., J.H., and G.-C.L. designed and performed the study and wrote the manuscript. W.F., W.J., and H.X. collected the samples and information. J.Z. and J.H. participated in data analysis. W.F., J.H., and G.-C.L. coordinated the entire study. All authors reviewed the final manuscript.

Conflicts of Interest

The authors declare no conflict of interest.