ong-Ping Chen · Yu-XiangY Liao · Zhen-Jian Zhuo · Li uYan · Hui-Ran Lin · Lei Miao · Xia Li · Xiao-Kai Huang ·Jing-Ying Zhou · Jun Bian · Jing He

Glioma is an aggressive malignant primary tumor in the central nervous system [ 1, 2]. Although the pathogenesis of glioma is unclear, the only identif ied environmental risk factor is ionizing radiation [ 3]. However, other factors also may alter the glioma risk, among which gene polymorphisms may play prominent role in glioma susceptibility. Base excision repair (BER) is a DNA repair mechanism that maintains genome integrity and guarantees DNA base modif ication properly. Many signif icant and common proteins involved in the BER pathway are associated with cancer. They are poly (ADP-ribose) polymerase 1 (PARP1), human 8-oxoguanine DNA glycosylase (hOGG1), f lap endonuclease 1(FEN1), apurinic/apyrimidinic endonuclease 1 (APEX1 or APE1), DNA ligase III (LIG3), and X-ray repair cross-complementing group 1 (XRCC1). PARP1 can regulate proteins by poly(ADP-ribosyl)ation and can repair damaged DNA by recruiting and modifying DNA repair complexes.

OGG1

gene is composed of six introns and seven exons, encoding HOGG1, which can initiate the BER pathway by hydrolyzing the glycosidic bond to produce abasic sites. FEN1 is a critical enzyme of the RAD2 structure-specif ic nuclease family that displays a variety of endonuclease and exonuclease activities and provides relevant biological functions.APEX1 is a multifunctional protein that has an essential role in cell proliferation, apoptosis, and diff erentiation as a transcription co-activator. LIG3 can participate in DNA doublestrand broken repair due to strong binding activity, with the majority in the nucleus and the minority in mitochondria.XRCC1 interacts with LIG3 and PARP to form complexes,playing a substantial role in the BER pathway. DNA repair mechanisms have been expected to be promising and critical targets to treat glioma. The primary therapy for glioma is DNA alkylation but produces N7-methylguanine and N3-methyladenine, removal by BER [ 4]. Given the importance of the BER pathway in DNA repair and that DNA repair is closely associated with glioma, the genetic variants in BER genes may confer glioma risk. However, the relationship between BER gene polymorphisms and glioma risk has not been fully understood.We conducted a case-control study, including 314 cases and 380 controls free of glioma. The inclusion criteria of the cases were patients who had been clinically diagnosed with glioma. The controls were randomly chosen from children who visited the hospital simultaneously with the cases.All participants were matched by age and gender distribution, and they all signed informed consent. This study was approved by the institutional review board of the hospital. We selected 20 single nucleotide polymorphisms (SNPs)in six genes for analysis, containing

PARP1

,

hOGG1

,

FEN1

,

APEX1

,

LIG3

, and

XRCC1

. The frequency distribution of selected variables between glioma patients and cancer-free controls is detailed in Supplementary Table 1. The case and control were frequency matched by age (

P

= 0.461) and gender (

P

= 0.379) with no signif icant diff erence. SAS v10.0 was used for all statistical analyses. Statistical signif icance was

P

＜ 0.05.As described in Table 1, the

PARP1

rs2666428 T ＞ C conferred elevated glioma susceptibility [recessive model:adjusted odds ratio (AOR) = 2.79, 95% conf idence interval(CI) 1.19—6.59,

P

= 0.019]. Under the dominant model,some SNPs could enhance the risk of glioma, including

FEN1

rs4246215 T ＞ G (AOR = 1.88, 95% CI 1.31—2.69,

P

= 0.0006),

APEX1

rs1130409 T ＞ G (AOR = 1.65,95% CI 1.20—2.27,

P

= 0.002),

APEX1

rs1760944 T ＞ G(AOR = 1.60, 95% CI 1.16—2.20,

P

= 0.004),

XRCC1

rs25487 C ＞ T (AOR = 1.39, 95% CI 1.03—1.88,

P

= 0.033),and

XRCC1

rs3810378 G ＞ C (AOR = 1.40, 95% CI 1.04—1.90,

P

= 0.028), while the protective eff ect of

FEN1

rs174538 A ＞ G on glioma risk was detected (AOR = 0.64,95% CI 0.47—0.89,

P

= 0.007). No pronounced association with glioma risk was detected among the remaining SNPs.Some SNPs which were signif icantly related to glioma risk and their combined risk genotypes were stratif ied in the analysis.Regarding

PARP1

(Supplementary Table 2), rs2666428 CC genotype increased the glioma risk significantly in the patients ＜ 60 months of age (AOR = 10.92, 95% CI 1.35—88.40,

P

= 0.025), females (AOR = 8.81, 95% CI 1.07—72.72,

P

= 0.043), children with astrocytic tumors(AOR = 2.70, 95% CI 1.06—6.88,

P

= 0.038) and embryonal tumors (AOR = 14.08, 95% CI 2.34—84.87,

P

= 0.004),in subgroups of stage I (AOR = 3.32, 95% CI 1.24—8.89,

P

= 0.017) and I + II (AOR = 2.75, 95% CI 1.10—6.86,

P

= 0.031). We considered rs1136410 AG/GG, rs2666428 CC and rs8679 AG/GG as risk genotypes to perform combining risk genotypes analysis. It was obvious that when compared to 0—1 risk genotypes, those with 2—3 risk genotypes were more dominant on increased glioma risk in females(AOR = 4.32, 95% CI 1.17—15.87,

P

= 0.028), ependymoma (AOR = 2.95, 95% CI 1.13—7.72,

P

= 0.027), stage III (AOR = 3.04, 95% CI 1.02—8.99,

P

= 0.045) and III + IV(AOR = 2.29, 95% CI 1.02—5.14,

P

= 0.045). Compared with the AA genotype, the

FEN1

rs174538 AG/GG genotype (Supplementary Table 3) decreased glioma risk among patients in subgroups of age ＜ 60 months (AOR = 0.62,95% CI 0.39—0.99,

P

= 0.044), males (AOR = 0.56, 95%CI 0.36—0.85,

P

= 0.007), astrocytic tumors (AOR = 0.58,95% CI 0.40—0.82,

P

= 0.003), stage I (AOR = 0.48, 95%CI 0.32—0.71,

P

= 0.0002) and I + II (AOR = 0.58, 95%CI 0.41—0.82,

P

= 0.002). However, the

FEN1

rs4246215 TG/GG genotype (Supplementary Table 3) could increase the glioma risk across the age ＜ 60 (AOR = 2.20, 95%CI 1.28—3.76,

P

= 0.004) and ≥ 60 (AOR = 1.68, 95%CI 1.04—2.71,

P

= 0.036), females (AOR = 2.22, 95%CI 1.29—3.82,

P

= 0.004) and males (AOR = 1.65, 95%CI 1.02—2.65,

P

= 0.041), subtypes of astrocytic tumors(AOR = 1.66, 95% CI 1.11—2.48,

P

= 0.013) and ependymoma (AOR = 2.99, 95% CI 1.37—6.54,

P

= 0.006). In the clinical stage subgroup, all but stage I showed evidences of increased glioma risk (stage II, AOR = 2.03,95% CI 1.07—3.86,

P

= 0.030; stage III, AOR = 3.56, 95%CI 1.23—10.36,

P

= 0.020; stage IV, AOR = 2.41, 95% CI 1.09—5.33,

P

= 0.030; stage I + II, AOR = 1.62, 95% CI 1.10—2.40,

P

= 0.015; stage III + IV, AOR = 2.83, 95% CI 1.48—5.40,

P

= 0.002).

Table 1 Association of polymorphisms in base excision repair pathway genes with glioma susceptibility in Chinese children

Bold letters indicating signif icant diff erences. poly (ADP) ribose polymerase 1, human 8-oxo-guanine DNA glycosylase, f lap endonuclease 1, apurinic/apyrimidinic endonuclease 1, DNA ligase III, X-ray repair cross-complementing group 1,adjusted odds ratio, conf idence interval, Hardy—Weinberg equilibrium, wild-type allele mutant allele, homozygous wild-type genotype, heterozygous genotype, homozygous mutant genotype. a Adjusted for age and sex for dominant model (MM/WM vs. WW); b adjusted for age and sex for recessive model (MM vs. WW/WM). “/” not available

Gene Polymorphism Allele Cases Controls AOR (95% CI) a P a AOR (95% CI) b P b HWE W M WW WM MM WW WM MM PARP1 rs1136410 A G 98 150 66 126 182 72 1.11 (0.80—1.53) 0.526 1.17 (0.81—1.71) 0.402 0.663 PARP1 rs2666428 T C 210 87 17 246 126 8 0.90 (0.66—1.24) 0.519 2.79 (1.19–6.59) 0.019 0.075 PARP1 rs8679 A G 276 38 0 339 41 0 1.16 (0.72—1.86) 0.540 / / 0.266 hOGG1 rs1052133 G C 110 153 54 122 191 67 0.88 (0.64—1.21) 0.417 0.90 (0.60—1.34) 0.587 0.602 hOGG1 rs159153 T C 254 55 5 297 80 3 0.83 (0.57—1.21) 0.337 1.88 (0.44—7.99) 0.395 0.340 hOGG1 rs293795 A G 284 28 2 339 39 2 0.87 (0.53—1.43) 0.579 1.08 (0.15—7.81) 0.940 0.451 FEN1 rs174538 A G 120 130 64 110 190 80 0.64 (0.47–0.89) 0.007 0.94 (0.65—1.37) 0.757 0.903 FEN1 rs4246215 T G 59 200 55 116 185 79 1.88 (1.31–2.69) 0.0006 0.80 (0.54—1.17) 0.242 0.740 APEX1 rs1130409 T G 96 158 60 157 164 59 1.65 (1.20–2.27) 0.002 1.30 (0.87—1.93) 0.198 0.142 APEX1 rs1760944 T G 92 150 72 149 163 68 1.60 (1.16–2.20) 0.004 1.38 (0.95—2.01) 0.088 0.048 APEX1 rs3136817 T C 262 49 3 313 65 2 0.94 (0.63—1.40) 0.751 1.92 (0.32—11.68) 0.477 0.482 LIG3 rs1052536 C T 146 137 31 176 176 28 1.00 (0.74—1.35) 0.990 1.40 (0.82—2.40) 0.217 0.073 LIG3 rs3744356 C T 305 9 0 369 11 0 0.91 (0.37—2.24) 0.835 / / 0.775 LIG3 rs4796030 A C 88 158 68 116 193 71 1.15 (0.83—1.60) 0.404 1.21(0.83—1.75) 0.324 0.556 XRCC1 rs1799782 G A 173 117 24 188 157 35 0.81 (0.60—1.09) 0.159 0.83 (0.48—1.43) 0.500 0.788 XRCC1 rs25487 C T 149 140 25 212 136 32 1.39 (1.03–1.88) 0.033 0.94 (0.54—1.62) 0.820 0.133 XRCC1 rs25489 C T 256 53 5 303 63 14 0.91 (0.62—1.33) 0.628 0.41 (0.14—1.15) 0.090 3.2 × 10 —5 XRCC1 rs2682585 G A 244 68 2 292 79 9 0.94 (0.66—1.35) 0.743 0.26 (0.05—1.20) 0.083 0.195 XRCC1 rs3810378 G C 148 138 28 211 136 33 1.40 (1.04–1.90) 0.028 1.02 (0.60—1.73) 0.938 0.106 XRCC1 rs915927 T C 247 65 2 295 78 7 0.94 (0.65—1.35) 0.724 0.34 (0.07—1.65) 0.181 0.490

Regarding

APEX1

genotypes (Supplementary Table 4),the rs1130409 TG/GG and rs1760944 TG/GG genotypes were associated with increased glioma risk in children ≥ 60 (rs1130409 TG/GG, AOR = 2.15, 95% CI 1.40—3.29,

P

= 0.0005; rs1760944 TG/GG, AOR = 1.69,95% CI 1.11—2.59,

P

= 0.016), males (rs1130409 TG/GG,AOR = 1.75, 95% CI 1.13—2.71,

P

= 0.012; rs1760944 TG/GG, AOR = 1.63, 95% CI 1.05—2.52,

P

= 0.028), subtypes of astrocytic tumors (rs1130409 TG/GG, AOR = 2.07, 95% CI 1.43—3.01,

P

= 0.0001; rs1760944 TG/GG, AOR = 1.70, 95%CI 1.18—2.46,

P

= 0.005), with tumors in stage I (rs1130409 TG/GG, AOR = 2.07, 95% CI 1.36—3.16,

P

= 0.0007;rs1760944 TG/GG, AOR = 1.81, 95% CI 1.19—2.75,

P

= 0.005) and I + II (rs1130409 TG/GG, AOR = 1.80, 95%CI 1.26—2.57,

P

= 0.001; rs1760944 TG/GG, AOR = 1.73,95% CI 1.21—2.48,

P

= 0.003). In addition, we also found rs1760944 TG/GG genotype related with enhanced glioma risk in the subtype of neuronal and mixed (AOR = 4.84,95% CI 1.42—16.53,

P

= 0.012). Then, rs1130409 TG/GG,rs1760944 TG/GG, and rs3136817 CC genotypes were further analyzed as risk genotypes for the combined eff ect.Compared to 0—1 risk genotypes, carriers with 2—3 risk genotypes were likely to increase glioma susceptibility at the age of ≥ 60 months, whole gender subgroup, patient with astrocytic tumors and neuronal and mixed, and stage I and I + II.Regarding

XRCC1

genotypes (Supplementary Table 5), the rs25487 CT/TT and rs3810378 GC/CC genotypes were associated with increased glioma risk in tumor grade III (rs25487 CT/TT, AOR = 2.43, 95% CI 1.17—5.03,

P

= 0.017; rs3810378 GC/CC, AOR = 2.40,95% CI 1.16—4.98,

P

= 0.018), IV (rs25487 CT/TT,AOR = 2.59, 95% CI 1.39—4.83,

P

= 0.003; rs3810378 GC/CC, AOR = 2.46, 95% CI 1.33—4.55,

P

= 0.004), and III + IV (rs25487 CT/TT, AOR = 2.61, 95% CI 1.60—4.26,

P

= 0.0001; rs3810378 GC/CC, AOR = 2.46, 95% CI 1.52—4.00,

P

= 0.0003). Besides, rs3810378 GC/CC genotype could develop the glioma risk in patients ≥ 60(AOR = 1.51, 95% CI 1.01—2.27,

P

= 0.044). Then we continued a combined analysis of rs1799782 GG/GA, rs25487 CT/TT, rs25489 CC/CT, rs2682585 GG/GA, rs3810378 GC/CC, and rs915927 TT/TC as risk genotypes. We observed the carriers with f ive to six risk genotypes were more likely to develop the glioma risk when compared to zero to four risk genotypes in some subgroups of children ≥ 60, females, subtypes of astrocytic tumors, grade III, grade IV and grade III + IV.We assessed the biological eff ects of seven signif icant gene SNPs from Genotype-Tissue Expression database.The rs2666428 T allele had signif icantly lower

PARP1

mRNA levels than the rs2666428 C allele in the tibial nerve, cerebellum, and cerebellar hemisphere. However,this eff ect was reversed in cultured f ibroblasts (Supplementary Fig. 1). We also observed that samples with

FEN1

rs174538 A and rs4246215 T genotypes could aff ect some distant genes’ mRNA expression levels [fatty acid desaturase (

FADS

)

1

,

FADS2

,

FADS3

, transmembrane protein

258,

and myelin regulatory factor] in some tissues (Supplementary Figs. 2, 3). The

APEX1

rs1130409 G genotype was associated with decreased O-sialoglycoprotein endopeptidase mRNA levels in cultured f ibroblasts. Those with the rs1760944 G allele were correlated with slightly decreased

APEX1

mRNA levels than the rs1760944 T allele in cultured f ibroblasts (Supplementary Fig. 4). It also showed that the rs25487 C allele had signif icantly lower

XRCC1

mRNA levels than the rs25487 T allele,while the rs3810378 C allele increased

XRCC1

mRNA levels than the rs3810378 G allele in the tibial nerve (Supplementary Fig. 5).A prior study has shown that SNPs of some genes involved in the BER pathway were considered to inf luence DNA repair capacity and modify cancer risk [ 5]. Liu et al.reviewed much evidence corresponding to BER polymorphisms in sporadic colorectal carcinogenesis [ 6]. Our group has shown that some gene SNPs in BER can change Wilms tumor risk [ 7, 8].

PARP1

,

hOGG1

,

FEN1

,

APEX1

,

LIG3

,and

XRCC1

are critical genes in BER, and 20 SNPs in these genes have been focused on in the previous study [ 9]. In addition, the released data from SNPinfo demonstrate the potential functions of these 20 SNPs. As described in Supplementary Fig. 6, some SNPs located within transcription factor binding sites are rs159153, rs174538, rs1760944,rs2682585, rs3136817, rs3810378, and rs4246215. Besides,rs1052133, rs1130409, rs1136410, rs174538, rs1760944,rs1799782, rs25487, rs25489, rs2666428, rs3744356,rs4796030, and rs915927 aff ect splicing exonic splicing enhancer or exonic splicing silencer. The rs1136410 and rs4796030 have the potential function to inf luence splicing abolish domain. Additionally, the function of rs1052133,rs1052536, rs2666428, rs293795, rs4246215, rs4796030,and rs8679 can be involved in impacting miRNA. Moreover, the non-synonymous coding SNP is susceptible to rs1052133, rs1130409, rs1136410, rs1799782, rs25487,rs25489, and rs3744356. The studies regarding the relationship of these genes with glioma susceptibility have been highly documented, some of which conducted a single gene investigation, multiple genes analysis, and even expanded to the whole DNA repair pathway. However, they did not systematically analyze these genes together in the BER pathway. Owing to the joint-gene investigations in our precedent study revealing optimistic signif icance to cancer risk and the various potential functions of these gene SNPs, it is promising to focus on glioma risk [ 7— 9]. Furthermore,whether these genes’ polymorphisms in the BER pathway impact glioma susceptibility among Chinese children is still ambiguous. Therefore, we conducted a case—control study to investigate this association.Our study implied that the

FEN1

rs4246215 T ＞ G,

APEX1

rs1130409 T ＞ G, rs1760944 T ＞ G, and

XRCC1

rs25487 C ＞ T, rs3810378 G ＞ C could confer increased glioma risk under the dominant model, but the same eff ect for

PARP1

rs2666428 T ＞ C under the recessive model.Only

FEN1

rs174538 A ＞ G was associated with reduced glioma risk. In the stratif ied analysis, the chosen SNPs in

PARP1

,

FEN1

,

APEX1

, and

XRCC1

have signif icant associations with glioma risk in certain subgroups, as depicted in the results. The expression quantitative trait loci analysis demonstrated that these signif icant SNPs could impact local or distant gene transcript levels. In particular, the rs174538 A allele and rs4246215 T allele could regulate many other genes’ mRNA expression levels, which might be the underlying mechanism by which glioma risk was altered.In summary, this is a case—control study to investigate the BER pathway common genes polymorphisms on glioma risk among Chinese children. Our result demonstrated that

PARP1

,

FEN1

,

APEX1

, and

XRCC1

SNPs are signif icantly associated with glioma susceptibility. However, intensive genetic studies need larger sample sizes and a Bonferronicorrected method to validate this relationship and discover potential molecular mechanisms.

Supplementary Information The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s12519- 022- 00562-0.

Author contributions

CYP, LYX, and ZZJ contributed equally to this work. CYP and LYX designed and performed the study and wrote the manuscript. ZZJ, BJ, and HJ coordinated the study over the entire time.YL, LHR, ML, HXK, and ZJY collected the samples and information,and participated in analyzing the data. LX participated in analyzing data. All the authors approved the f inal version of the manuscript.

Funding

This study was supported by grants from the National Natural Science Foundation of China (81802346) and the Discipline Promotion Project of the First Affi liated Hospital of Air Force Medical University(XJZT18MJ49).

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethical approval

This study was approved by the Institutional Review Board of Guangzhou Women and Children’s Medical Center (No.2016021650).

Conflict of interest

The authors declare that there is no conf lict of interest.