hOGG1 rs1052133 Polymorphism and Prostate Cancer Risk: A Chinese Case-Control Study and Meta-Analysis

Background

The role of DNA damage in tumor formation and development is well established [1]. As a result of oxidative stress, inflammation, or environmental carcinogens, DNA damage can accumulate in the prostate, possibly increasing the risk for prostate cancer (PCa) [2–4]. The base excision repair (BER) pathway is the most often-used approach for removing minor damage from DNA. Despite its structural independence, it is also crucial in cellular defense against many types of DNA lesions [5,6].

The human 8-oxoguanine glycosylase 1 (hOGG1) gene, located on chromosome 3p26, plays an essential role in DNA damage repair by initiating the BER pathway [7–9]. Rs1052133 is a commonly occurring polymorphism in hOGG1, and also referred to as Ser326Cys polymorphism. Cysteine can be substituted by serine at 326 amino acids of hOGG1 protein. Notably, there is an altered susceptibility for diverse cancer types related to its genotypes [10–12].

In recent years, the hOGG1 rs1052133 polymorphism has been often investigated in relation to PCa across diverse ethnic populations, but the results were inconsistent [13–24]. Here, we assessed whether genotypes of the hOGG1 rs1052133 polymorphism were related to an elevated PCa risk in a Chinese Han population. Additionally, we performed a comprehensive updated meta-analysis of published investigations to determine the exact relationship between them.

Material and Methods

Case-Control Study

SELECTION OF ELIGIBLE PCA CASES AND HEALTHY CONTROLS:

We enrolled 160 PCa patients and 243 healthy controls from the First Affiliated Hospital of Anhui Medical University between January 2016 and December 2020. A biopsy or postoperative pathology of transurethral resection prostate (TURP) was used to diagnose PCa patients within 1 year of enrollment. In the present study, patients pathologically confirmed to have prostate adenocarcinoma were enrolled for subsequent analysis. Prior to invasive manipulation, PSA levels were assessed, excluding those patients who had received endocrine therapy. Gleason grades were determined by biopsy or radical prostatectomy. Based on the radical prostatectomy specimens or the results of computed tomography, magnetic resonance imaging, or bone scans, the tumor stage was determined. All the controls were cancer-free individuals who underwent regular physical examinations at the hospital. Study controls with serum PSA levels >4 ng/ml were excluded to rule out prostate cancer. The investigators excluded patients with incomplete medical records. The Research Ethics Committee at the First Affiliated Hospital of Anhui Medical University approved the study. Furthermore, each participant signed an informed consent form in duplicate.

GENOTYPING: Commercially available DNA extraction kits (Cat. No. 51106; Qiagen, Inc., Valencia, CA) were used to extract DNA from each participant’s blood sample. The site sequence of the hOGG1 rs1052133 polymorphism was obtained from the NCBI db SNP database. For each DNA sample, the hOGG1 rs1052133 polymorphism was detected by using the SNaPshot SNP assay by Shanghai Tianhao company (Shanghai Tianhao Industrial Co., Ltd., Shanghai, China) [25,26]. We used the Multiplex SNaPshot technology (Applied Biosystems, Foster City, CA, USA) to determine the genotype of hOGG1 gene. Primer3 online software (v.0.4.0) (http://frodo.wi.mit.edu/primer3/) was used to design the primers for PCR and the SNaPshot extension reactions based on the sequences provided in dbSNP (http://www.ncbi.nlm.nih.gov/SNP) (rs1052133F: CCAGGTGGCCCTAAAGGACTCT; rs1052133R: GTGGGGATGGGGAGAGAGAAGT; primer extension: TGGCTCCTGAGCATGGCGG). The products were sequenced by ABI3130XL Sequencer (Applied Biosystems, Foster City, CA, USA), and GeneMapper 4.0 (Applied Biosystems, Foster City, CA, USA) was used to analyze the data. The type of nucleotide presented in the SNP locus was used to determine the genotype of each sample, which was analyzed by 1 or 2 distinct colored peaks on the graph.

STATISTICAL ANALYSIS FOR REAL-WORLD COHORT:

Differences between the 2 groups were determined using the 2-tailed unpaired t test. We presented the data as the mean±standard deviation (mean±SD), and P value less than 0.05 was deemed as statistically significant. The different distribution of genotypes of rs1052133 in the subgroups of PSA, Gleason score, and TNM stage were evaluated by chi-square test. Based on simple counting, genotypic and allelic frequencies were determined. We compared Hardy-Weinberg equilibrium (HWE) test results, genotype frequencies (CC, G/C, and GG), and distributions between cases and controls. Odds ratios (ORs) and 95% confidence intervals (95% CIs) were generated for each comparison to reveal its risk to prostate cancer. Procedures of statistical analyses were performed using SPSS version 17.0 (SPSS, Inc., Chicago, United States).

META-ANALYSIS:

The meta-analysis was performed following the PRISMA Statement [27]. Figure 1 shows the whole process of the study.

PUBLICATION SEARCH:

Our electronic literature search utilized PubMed, Web of Science, and EMBASE to find all eligible studies that evaluated the risk associated between the hOGG1 rs1052133 polymorphism and the risk of PCa up to April 2022. In this search, the following MeSH terms and keywords were employed: (“PCa” OR “prostate cancer” OR “prostate tumor” OR “prostate adenocarcinoma” OR “prostate neoplasm”) AND (“hOGG1” OR “human 8-oxoguanine DNA glycosylase 1”) AND (“gene” OR “polymorphism” OR “variant” OR “allele” OR “mutation”) AND (“rs1052133” OR “Ser326Cys” OR “1245C>G” OR “C8069G”). The entire search process was conducted in English. If the same population was presented in multiple publications, only the publication with the largest sample size was enrolled.

INCLUSION AND EXCLUSION CRITERIA:

Eligible studies were recruited to the meta-analysis when the following criteria were reached: (1) a case-control study; (2) a cohort study; (3) evaluation of the presence of the rs1052133 polymorphism and the risk of PCa; (3) adequate data were available; and (4) genotype distributions following Hardy-Weinberg equilibrium (HWE).

Accordingly, we excluded studies with the following characteristics: (1) cross-sectional cohort studies published in other languages, non-original studies, dissertations, and thesis studies; (2) insufficient data or studies lacking genotype distribution data; and (3) study was not relevant to hOGG1 rs1052133 or the PCa polymorphism. Author(s), publication date, country, ethnicity, examined genes (SNPs), sample number, genotyping method used, genotypic frequencies, and allelic frequencies were collected. Two reviewers (HJX and MZ) independently evaluated the qualities of enrolled studies in our meta-analysis. The qualities of the included studies were assessed by the Newcastle-Ottawa Scale (NOS) score method [28] from the different aspects, including selection, comparability, exposure/outcome (Supplementary Table 1).

META-ANALYSIS:

Genotype models (the heterozygote, homozygote, dominant, and recessive genetic models) were employed to determine the correlation of the hOGG1 rs1052133 polymorphism with PCa risk by using the crude OR with related 95% CI.

We determined the methods to calculate the OR from the χ²-based Q-test and checked the heterogeneity of the current study. If P>0.1 indicated that the Q-tests were not heterogeneous, then the ORs were obtained by using the Mantel-Haenszel method (fixed effects). As an alternative, the DerSimonian and Laird model was used to check the significance of a pooled OR. We performed a stratification analysis by ethnicity (White, Asian, and mixed populations). Moreover, a visual assessment of publication bias was performed by inspecting Begg’s funnel plot for asymmetry, as well as the Egger’s linear regression test. There is a possibility of publication bias when P≤0.05 and the scheme is considered to be asymmetric. We used STATA 12.0 software (Stata Corp., College Station, TX) to conduct the statistical analyses. We considered two-sided P values <0.05 to be statistically significant.

Results

Case-Control Study

DEMOGRAPHIC FEATURES: The average age in the PCa group was 70.09±7.14 years compared to 71.30±8.80 years in the negative control group (t test, P=0.148). In the PCa group, 46 (28.75%) subjects showed a PSA level <10 ng/ml, 50 (31.25%) subjects showed a PSA level of 10–20 ng/ml, and 64 (40.00%) subjects showed a PSA level >20 ng/ml. The number of PCa cases whose Gleason score ≤7 was 99 (61.88%), and the number with a Gleason score >7 was 61 (38.12%). Moreover, there were 119 (74.38%) patients and 41 (25.62%) patients with stage <T3 and ≥T3, respectively. We summarized these features in Table 1.

ASSOCIATION OF HOGG1 GENOTYPES WITH PCA RISK: The prevalence of the hOGG1 rs1052133 polymorphism in healthy controls followed HWE (P=0.380). As indicated, hOGG1 polymorphism was not correlated with an increased risk of PCa in the 4 genetic models in our case-control study (Table 2). We further analyzed the correlation between subgroups (PSA level, Gleason score, and TNM stage) in PCa cases to validate the correlation with the predictive risk of PCa (Table 3). However, none of these differences were statistically significant, which was possibly due to the limited sample size.

ELIGIBLE STUDIES: Twelve eligible studies were found by using the selected keywords (Table 4) [13–24]; however, 1 repeated study was removed [19], and 2 studies that failed to meet the HWE balance were excluded [20,21]. Finally, 10 studies were included in the meta-analysis, including the data that were generated from the present case-control study.

META-ANALYSIS: The detailed associations of the hOGG1 rs1052133 polymorphism with PCa susceptibility in all of the genetic models are presented in Table 5. The results showed that rs1052133 polymorphism was significantly related to an elevated PCa risk in 2 genetic models, including heterozygote (OR: 1.172, 95% CI: 1.033–1.330, P=0.014) and dominant models (OR: 1.227, 95% CI: 1.003–1.500, P=0.047). In the stratification analysis by ethnicity, we observed a significant association of the rs1052133 polymorphism with susceptibility to PCa risk in Asian population in 3 genetic models, including allele (OR: 1.217, 95% CI: 1.052–1.409, p=0.008), heterozygote (OR: 1.593, 95% CI: 1.160–2.187, P=0.004), and dominant models (OR: 1.299, 95% CI: 1.030–1.638, P=0.027), and in White populations in all of the genetic models, but it should be noted that there were 2 studies with small sample sizes that were included (Table 5).

Moreover, the findings were stable and robust, which was supported by the sensitivity analysis after removing any of the individual studies (Figure 2A). Both the plot of Begg’s funnel (Figure 2B) and test of Egger’s regression analysis (P>|t|=0.478) suggested that no publication bias existed.

Discussion

In developed and developing countries alike, PCa is a malignancy that is strongly influenced by genetic factors. According to Yamane et al [29], the hOGG1 gene is an integral component of the DNA repair pathway. hOGG1-Cys326 (also termed rs1052133) was much less effective at preventing mutations in human cells than hOGG1-Ser326. Studies have suggested that the hOGG1 rs1052133 polymorphism may be important in determining PCa susceptibility. There is disagreement over the relationship between the rs1052133 polymorphism in hOGG1 and PCa susceptibility.

The first evidence for the increased prostate cancer risk of men with the CC genotype (Ser326) was presented by Xu et al [13] in 2002, which compared this genotype with homozygous GG men. Similar findings have been confirmed in the Korean population [22] and White population [14]. Moreover, Dhillon et al [17] found that only G allele of the rs1052133 polymorphism was related to an elevated risk of PCa. In 2 studies that were conducted in the Chinese Han population, one study demonstrated that the hOGG1 rs1052133 polymorphism was correlated with an enhanced malignant potential of PCa [23], whereas the other study reported that this polymorphism was more associated with the risk of low-grade prostate cancer [24]. Lavender et al. [16] and Mittal et al. [18] proposed contrasting opinions that they failed to observe any significant association between the HOGG1 genotypes and PCa risk. These discordant and conflicting results may be due to limited sample sizes and different genetic backgrounds.

Several meta-analysis studies have made efforts to test the relationship of the hOGG1 rs1052133 polymorphism with PCa risk. Notably, our results are slightly different from those published studies via the improvement of some flaws [30–32]. In 2012, Zhu et al [30] performed a meta-analysis study, and they found that the rs1052133 polymorphism was related to an elevated risk of PCa in both Whites and Asians. However, in the study by Agalliu et al. [33], they focused on the rs3218997 polymorphism and PCa risk, which was incorrectly recruited and synthesized in the study by Zhu et al. [30]. In 2015, Chen et al [32] performed an updated meta-analysis comprising 11 studies. However, in the pooled analysis, their results failed to show any correlation between the rs1052133 polymorphism and PCa risk. Significant associations were only found in the Asian population. Two studies with duplicated samples were both enrolled, which would cause a potential bias [18,19]. Furthermore, for those studies that failed, the HWE balance should be removed from the pooled analysis. Thus, we aimed to perform a case-control study and comprehensive updated meta-analysis to reveal the associations. In our case-control study, we failed to identify any positive results between the genotypes and PCa risk, which is a result consistent with most previous studies. Notably, based on the newly generated data from our case-control study, we performed a more rigorous and updated meta-analysis to determine the association of the rs1052133 polymorphism and PCa risk. Finally, we recruited 10 case-control studies, including 2218 PCa cases and 2946 controls. As a result, we obtained a positive correlation between the hOGG1 rs1052133 polymorphism and an elevated risk of PCa. Our meta-analysis identified a statistically significant correlation between the rs1052133 and PCa risk.

However, there were several shortcomings to be addressed in this study. First, this was a case-control study based in a hospital; thus, a selection bias should be considered. Additionally, our case-control study only analyzed the Chinese Han population with small sample size; therefore, more research is needed to confirm these findings. Finally, the biological function of this polymorphism during the progression of PCa has not been investigated.

Conclusions

Our study indicates that the rs1052133 polymorphism in hOGG1 is related to an elevated risk of PCa. However, more studies are necessary to examine their precise effects and the genuine associations between them in different countries and ethnicities.