Genetic variants in XPD gene and glioma susceptibility in Chinese children: A multicenter case–control study

Abstract Background Glioma is one of the central nervous system (CNS) tumors in children, accounting for 80% of malignant brain tumors. Nucleotide excision repair (NER) is a vital pathway during DNA damage repair progression. Xeroderma pigmentosum group D (XPD) or excision repair cross‐complementing group 2 (ERCC2) is a critical factor in the NER pathway, playing an indispensable role in the DNA repair process. Therefore, the genetic variants in XPD may be associated with carcinogenesis induced by defects in DNA repair. Methods We are the first to conduct a multi‐center case‐control study to investigate the correlation between XPD gene polymorphisms and pediatric glioma risk. We chose three single nucleotide polymorphisms and genotyped them using the TaqMan assay. Results Although there is no significant association of these genetic variations with glioma susceptibility, the stratified analysis revealed that in the subtype of astrocytic tumors, the rs13181 TG/GG genotype enhanced glioma risk than the TT genotype, and carriers with two to three genotypes also elevated the tumor risk than 0‐1 genotypes. Conclusion In conclusion, our findings provided an insight into the impact of XPD genetic variants on glioma risk.


| INTRODUCTION
Glioma is the most frequent tumor in children, originating from glial and precursor cells.Glioma accounts for above 51% of pediatric tumors, and the incidence of malignant glioma has increased for nearly two decades among children aged 0-14 years [1].Based on the World Health Organization (WHO) classification, gliomas were classified into benign astrocytomas (low-grade for grades I-II), anaplastic astrocytomas (high-grade for grade III), and glioblastoma (highgrade for grade IV) [2].The most common pediatric gliomas are astrocytomas, oligodendrogliomas, ependymomas, and brainstem gliomas.Among gliomas, pilocytic astrocytoma is considered nonmalignant, recognized by the WHO and clinical practice.Glioblastoma and other diffuse gliomas were categorized into malignancies, frequently occurring in adults [3].Most gliomas in children were pediatric low-grade gliomas (PLGG), which developed slowly, referred to as grades I-II by WHO.Meanwhile, some pediatric gliomas, namely high-grade gliomas (PHGG), were in rapid occurrence and development classified as grades III-IV [4].The current standard of care for glioma is surgery, radiation, and alkylating agents [5].However, the overall survival rate of glioma is unsatisfactory.Pediatric gliomas may develop into adult gliomas because of low-grade lesions and remaining recurrence for many years [6].
DNA repair pathways mainly have nonhomologous end joining (NHEJ), base excision repair (BER), and nucleotide excision repair (NER) [7].As one of the classical general impaired repair pathways, the NER pathway is highly conservative.Its function is the removal of bulky adducts caused by UV radiation and several chemical agents [8].In addition, several essential NER genes associated closely with xeroderma pigmentosum (XP) are XPA, XPB, XPC, XPD, XPE, XPF, and XPG, respectively.They together ensure the DNA repair procedure [9].XP group D (XPD/ERCC2), an 87 kDa protein, is a transcription factor IIH (TFIIH) component.XPD gene is located at chromosome 19q13.3and contains 22 exons and 21 introns, encoding 2.3 kb messenger RNA (mRNA) [10].XPD played an essential role in the NER pathway because it could encode a core ATP-dependent DNA helicase and utilize 5′-3′ polarity, unwinding the DNA duplex near impaired sites.However, this XPD helicase is optional for transcription [11].Additionally, XPD was found in mitochondria, and it could keep the mitochondrial genome from being impaired by oxidative DNA [12].Mutations in XPD principally were from single residue alterations, sometimes in neighboring residues [11].Not only were the mutants in the XPD gene associated with XP, trichothiodystrophy, and Cockayne's syndrome, but they affected the cancer susceptibility [13].
It is still unsure whether the numerous genetic variations in many genes are considered the cause or the effect of cancer.If the consideration is the reason, the genetic variants studies benefit enormously from the effect of single nucleotide polymorphisms (SNPs) on cancer occurrence and progression.Growing research demonstrated that genetic modifications in DNA repair genes like SNPs could influence protein expression during repair, containing encoding, transcription, and translation.Finally, these influences impaired DNA repair capacity and led to genetic instability, even carcinogenesis [14].Given the importance of the XPD gene in impacting DNA repair capacity, genetic variants in this core gene were likely to alter cancer risk.For instance, XPD rs1799787 is closely associated with lung cancer risk, and the combined effect with other NER genes and SNPs contributed to elevated lung cancer risk [15].Variant alleles in XPD also enhanced pancreatic cancer susceptibility [16].Recent genome-wide association studies have identified 13 new loci for glioma, such as polymorphisms in RAVER2 (rs12752552), MDM4 (rs4252707), and AKT3 (rs12076373) [17].However, the relationship between XPD gene polymorphisms and pediatric glioma risk has not been studied.Therefore, we conducted a case-control study to explore the impact of XPD SNPs on glioma susceptibility in Chinese children for the first time.

| Study subjects
This case-control study included 314 glioma cases and 380 healthy controls, which were collected from Guangzhou Women and Children's Medical Center (171 cases and 228 controls), Xiangya Hospital (85 cases and 132 controls), and the Second Affiliated Hospital and Yuying Children's Hospital of Wenzhou Medical University (58 cases and 20 controls).The cases were confirmed as patients with glioma.The controls group was healthy participants recruited randomly from hospital visitors during the same period as the cases.They were matched based on age and gender distribution.All the participants signed informed consent to use their samples before the study.Our research was approved by the institutional review board of Guangzhou Women and Children's Medical Center.
The rs238406 also has the potential function to influence exonic splicing enhancer or exonic splicing silencer.Additionally, rs3810366 may influence the activity of transcription factor binding sites.DNA was extracted from blood samples with the QIAamp DNA Blood Mini kit (QIAGEN).Selected SNPs were genotyped using TaqMan real-time PCR (Applied Biosystems).The details for genotyping are the following conditions: preread stage at 60℃ for 30 s, holding stage at 95℃ 10 min, repeated 45 cycles each of denaturation at 95℃ for 15 s, annealing, and extension at 60℃ for 1 min.Then, we selected the standard run mode and added the reaction volume (5 µl for each well in a 384-well reaction plate) into the instrument.Finally, we loaded the reaction plate, then started the run, genotyping 10% of the samples randomly and blindly, exhibiting a 100% concordance rate.

| Statistical analysis
Using the goodness-of-kit χ 2 test among control subjects to determine whether genotypes complied with Hardy-Weinberg equilibrium (HWE), we employed the χ 2 test or t-test to compare the differences in clinical variables between the cases and controls.Multivariate logistic regression analysis was performed to estimate the association between the SNPs and glioma risk by the age-and gender-adjusted odds ratios (ORs) and 95% confidence intervals (CIs).We further conducted the stratification analysis for cases in age, gender, subtypes, and clinical-stage subgroups.All the statistical analyses were completed using the SAS v10.0 (SAS Institute Inc.), and a two-sided p < 0.05 was considered significant.To compare the XPD expression between glioma groups from WHO grades II-IV, we utilized one-way ANOVA by online analysis from the Chinese Glioma Genome Atlas (CGGA).

| Characteristics of study participants
Detailed population characteristics between 314 cases and 380 controls are shown in Supporting Information: Table S1.Cases and controls were well matched by age (p = 0.461) and gender (p = 0.379), with no statistically significant difference.Among the patients, the astrocytic tumors (214) accounted for 68.15%, the ependymoma (61) for 19.43%, the neuronal and mixed neuronal-glial tumors (25) for 7.96%, the embryonal tumors (12) for 3.82%, and 2 for 0.64% could not be classified.Cases were staged by WHO classification, including 151 (48.09%) for grade I, 73 (23.23%) for grade II, 36 (11.46%) for grade III, and 53 (16.88%) for grade IV.
Three SNPs in the XPD gene were genotyped successfully in 313 cases and 380 controls, as presented in Table 1.Regrettably, we could not find significant associations between selected XPD gene polymorphisms and glioma risk in single-locus and combined effect analyses.

| Stratification analysis
We conducted the stratified analysis by age, gender, subtypes, and clinical-stage subgroups.The XPD rs13181 heterozygote of thymine and guanine/homozygous guanine (TG/GG) genotypes was observed with enhanced glioma risk in astrocytic tumors when compared to homozygous thymine genotypes (adjusted OR = 1.60, 95% CI = 1.02-2.54,p = 0.043).Furthermore, rs3810366 GG, rs238406 homozygous guanine/heterozygote of guanine and thymine (GG/ GT), and rs13181 TG/GG genotypes were referred to as risk genotypes.In comparison to 0-1 genotypes, carriers with two to three genotypes were associated with increased glioma risk in the subtype of astrocytic tumors (adjusted OR = 1.69, 95% CI = 1.09-2.62,p = 0.019) (Table 2).We gained the published functional relevance from GTEx (http://www.gtexportal.org/home/)and evaluated the mRNA levels varied with XPD genotypes.As denoted in Figure 1a, the rs13181 G genotype dipped XPD mRNA levels conspicuously in cell-cultured fibroblasts and nervetibial than the rs13181 T genotype.Simultaneously, the rs13181 G genotype altered mRNA levels of genes in the vicinity, involving KLC3, CD3EAP, and MARK4.

| Functional annotation of XPD
We further acquired outright data for functional annotation of XPD in the mRNAseq_325 data set (www.cgga.org.cn)from the Chinese Glioma Genome Atlas (CGGA).Through one-way ANOVA, the outcome illuminated that these was a statistical difference in the XPD gene expression between WHO grades II-IV (Figure 2a).The survival analysis was also implemented to scrutinize the correlation between XPD expression and survival probability, which indicated that strengthened XPD expression threatened glioma patients (Figure 2b).These results might conduce to discover the XPD implication in pediatric glioma.

| DISCUSSION
The existing problems, such as complicated treatment and poor prognosis in glioma, have disturbed children's health.Glioma has discrepancies between children and adults at the molecular level.Additionally, different from adult glioma, most pediatric glioma belongs to WHO grade I. Thus, although the molecular pathogenesis research in adult glioma has grown tremendously, these genetic findings may not explain the etiology of pediatric glioma.In addition, how the XPD polymorphisms impact the underlying molecular mechanism of glioma tumorigenesis in children still has not been entirely investigated.To systematically discover the strength of associations between XPD gene SNPs and glioma risk in Chinese children for the first time, we performed a multicenter case-control study with 314 cases and 380 controls.Moreover, we also evaluated the glioma risk in carriers with SNPs risk genotypes.Various environmental factors have been reported to influence glioma, among which the effect of ionizing radiation was confirmed [18].Genetic factors also occupied an essential role, and many molecular diagnostic markers in glioma have been discovered, such as IDH, codeletion of chromosomal arms 1p and 19q (1p/19q codeletion), H3F3A, nuclear alpha-thalassemia/mental retardation X-linked syndrome (ATRX) gene, O 6methylguanine-DNA methyltransferase (MGMT), and telomerase reverse transcriptase (TERT) [19].Many DNA repair mechanisms are categorized as direct and indirect repair types [20].The NER pathway belongs to one of the excision repair systems in the indirect repair mechanisms, which completes repair progress by excising and removing about 24-32nt impaired DNA fragments after DNA replication.The most frequent DNA damage, processed by NER, is bulky adduct from DNA mutations due to mutagenic agent exposure [21].The dysfunctional NER pathway would influence cancer risk.In a study corresponding to colorectal cancer in the Chinese population, some NER genetic variants (XPA rs10817938 and XPC rs2607775) were illustrated to alter disease risk.Moreover, this study revealed that genetic biomarkers in the NER pathway were associated with predicting colorectal cancer risk and clinical outcome [22].Recent research elaborated that the disequilibrium of the NER pathway increased the risk of hematological malignancies among patients with XP group C. In these patients, bulky purine DNA lesions could not be repaired entirely, leading to multiple mutations in transcription and replication, consequently the higher risk of internal cancers [23].There are two subpathways in the NER pathway: global genome NER (GG-NER) and transcription-coupled NER (TC-NER).The GG-NER is used for whole-genome repair.First, it mainly relies on the XPC-RAD23B complex and UV-DDB (UV-damaged DNA-binding activity) to complete damage identification.Then, it unwinds and stabilizes impaired DNA through XPG, XPB, and XPD.Subsequently, it cleaved the damaged DNA fragment mainly with the help of XPG, ERCC1-XPF, and finally used various DNA polymerases to fill the gaps [24].Besides, when its defects, it can result in cancer predisposition.The TC-NER is initiated by the suppressed RNA polymerase, which is indispensable for damage recognition, occurring at the damaged transcriptional strand in active genes.It works through combination with transcription mechanisms, playing a role in numerous syndromes [8,25].
DNA repair pathway maintains genome stability by protecting the genome from damage by endogenous and exogenous hazardous agents, such as environmental and genetic factors and their interactions.The variants in DNA repair genes may alter cancer risk [7].Especially, XPD mediated various biological functions to stimulate disease growth and development [26].There are many SNPs in the XPD gene impacting genic exons and introns.Some nonsynonymous SNPs may affect gene encoding ability and protein expression levels [27].Some XPD polymorphisms have been suggested to modulate DNA repair capacity and promote tumorigenesis [28].XPD Asp312Asn (rs1799793) was related to prostate cancer pathology [29].XPD Lys751Gln (rs13181) polymorphism could increase the susceptibility of such cancers [30].Of note, genetic variations that arise spontaneously, referred to as SNPs, showed potential for predicting the glioma risk [18].Moreover, several XPD gene polymorphisms may influence the DNA repair ability in glioma [27].Our chosen SNPs are rs3810366, rs238406, and rs13181, all in chromosome 19.The rs3810366 polymorphism is at the position of 45370684 with the G/A/C/T substitution.The rs238406 alleles variants T/G are at 45365051 in exon 6.The rs13181 T/A/G is located in 45351661 in exon 23, contributing to an amino acid transformation from Lys to Gln.Of them, XPD rs13181 is the most common variant.In addition, the rs13181 plays a prominent role in the work of XPD helicase via influencing the C-terminal domain [31].These SNPs may modify cancer risk and be treated as potential markers of carcinogenesis [32].In a correlation study from Poland, the DNA repair gene polymorphisms have significant associations with breast cancer risk in Polish women.Specifically, the XPD rs13181 allele strongly promoted the risk of breast cancer [33].However, our previous study could not find evidence that XPD genetic variations are associated with hepatoblastoma susceptibility in a single locus analysis [34].Worthy of note, XPD rs3810366 and rs238406 were identified to alter neuroblastoma risk.XPD rs3810366 could improve the susceptibility of neuroblastoma, while the protective effect of XPD rs238406 on neuroblastoma tumorigenesis was detected [35].Furthermore, Zhang et al. revealed that the essential prognostic factors for the lung cancer risk might be XPD rs13181 T > G and rs1799793 C > T. They found that both SNPs could enhance death risk, and the role of XPD rs13181 was more prominent in age, sex, and smoking [36].Maral Adel Fahmideh et al. conducted a systematic review and meta-analysis.They observed that the XPD rs13181 might be more susceptible to glioma risk, but no evidence of a significant relationship with glioma susceptibility has been accessed for another polymorphism of XPD, rs1799793 [37].However, no significant SNPs with glioma susceptibility were identified among the results in any statistical analysis but stratified analysis in our study.The stratified analysis demonstrated XPD rs13181 TG/GG and individuals with two to three risk genotypes promoted disease risk in astrocytic tumors.The interpretations of negative results may include inadequate sample size, population heterogeneity, and low-penetrance of a single variant.In a previous study, DNA repair gene XPD was considered a low-penetrance gene for glioma susceptibility [37].Existing research has elucidated the effect of XPD rs13181 on the risk of glioma [37], but has not yet probed into the correlation of XPD rs3810366 with glioma risk, so our negative result regarding rs3810366 will provide an insight into the potential role of this SNP in glioma etiology.Another negative outcome of our research was that the impact of rs238406 on glioma risk was unsuccessfully assessed in one comprehensive analysis and two meta-analyses [38][39][40].
Several limitations can be noted.First, an individual's cancer susceptibility may be an accumulative effect of various risk factors.Our research only concentrated on genes, without considerations like environmental agents or gene-environment interactions.Second, we only selected three SNPs in the XPD gene for the association research, and more potentially functional polymorphisms should be taken into further exploration.Third, a limited sample size might weaken the statistical power, and thereby the strength of correlation could not be objectively reflected.Fourth, the subjects were recruited from the Chinese Han children, so the ethnic discrepancy should be accounted for by applying results to other ethnicities.Finally, we did not conduct relevant biological experiments to validate the related protein and RNA expression levels, which should be further studied.
In sum, this study investigated the potential effect of XPD polymorphisms on the glioma risk in children.Our present evidence failed to observe any association between single SNP and glioma risk.However, we detect that the XPD rs13181 TG/GG and relevant variant genotypes might elevate glioma susceptibility in a specific subgroup.Larger scale studies are warranted to illuminate the underlying molecular mechanism of how XPD genetic variants affect glioma risk.
T A B L E 1 Association of XPD gene polymorphisms with glioma susceptibility in Chinese children ) b p b rs3810366 G > C (HWE = 0.603)

F I G U R E 1
Functional effect of XPD gene rs13181 polymorphism from GTEx portal.(a) Demonstrates the messenger RNA (mRNA) expression of XPD rs13181 genotypes in cell-cultured fibroblasts and nerve-tibial.(b) Explicates different mRNA expressions of its vicinal genes KLC3, CD3EAP, and MARK4 in different tissues.

F
I G U R E 2 The association between XPD expression and glioma development from the Chinese Glioma Genome Atlas database.(a) Manifests the relevance of XPD gene expression on glioma (classified by World Health Organization grade).(b) Kaplan-Meier method was used for survival analysis, revealing the distinguished patient survival by XPD gene expression.
Abbreviationsb Adjusted for age and sex.c Stratification analysis of risk genotypes with glioma susceptibility T A B L E 2Note: The statistically significant values are bold with p < 0.05.Abbreviations: CI, confidence interval; GG, homozygous guanine; OR, odds ratio; TG, heterozygote of thymine and guanine; TT, homozygous thymine.a Adjusted for age and sex, omitting the corresponding stratify factor.