Genetic association between CDKN2B-AS1 polymorphisms and the susceptibility of primary open-angle glaucoma (POAG): a meta-analysis from 21,775 subjects

Background Primary open-angle glaucoma (POAG) is affected by both genetics and environmental factors. CDKN2B-AS1 polymorphisms have been reported to be involved in the pathogenesis of POAG. However, the results of the genetic associations between the CDKN2B-AS1 polymorphisms and POAG risk were inconclusive. Aims This study aimed to evaluate the correlation of CDKN2B-AS1 polymorphisms and POAG susceptibility using a meta-analysis. Methods Meta-analysis was performed by searching PubMed, Web of science, the Cochrane database of system reviews, CNKI, and Embase databases. The relationship of CDKN2B-AS1 rs4977756, rs10120688, rs2157719, and rs7049105 polymorphisms and POAG risk was evaluated by the odds ratios (ORs) and 95% confidence intervals (CIs). Results Eleven studies with 8290 cases and 13,485 controls were included in the present meta-analysis. The alleles of rs4977756 and rs10120688 significantly increased the risk of POAG (rs4977756: OR = 1.20, 95%CI = 1.03–1.39, p = 0.02; rs10120688: OR = 1.36, 95%CI = 1.29–1.44, p < 0.00001). As for ethnicity, rs4977756 polymorphism significantly increased POAG risk in Caucasians (OR = 1.33, 95%CI = 1.12–1.57, p = 0.0009), but not in Asians. In addition, the rs2157719 allele was significantly associated with POAG risk in Asians (OR = 0.66, 95%CI = 0.55–0.80, p < 0.0001), but not in Caucasians (p > 0.05). Conclusions The CDKN2B-AS1 rs4977756 might increase the POAG risk in Caucasian population, and rs2157719 might decrease the POAG risk in Asian population, while rs10120688 might increase the risk of POAG.

CDKN2B-AS1 is an antisense RNA that may influence the nearby CDKN2A and CDKN2B genes via regulatory mechanisms [20] and was determinate to be a genetic susceptibility locus for several age-related complex diseases including POAG [21,22]. Recently, several GWA studies have identified a number of POAG-associated single-nucleotide polymorphisms (SNPs) such as rs4977756, rs10120688, rs2157719, and rs7049105 [23,24]. Burdon et al. revealed CDKN2B-AS1 rs4977756 can significantly increase the risk of POAG in a cohort of 590 cases and 3956 controls in Australia [23]. Ng et al. also illustrated the genetic association between the CDKN2B-AS1 rs4977756 and POAG in Australians [25]. It is suggested that the stronger genetic signals at the 9p21 locus among females may contribute to the observed sex bias for POAG. However, several studies on the association between CDKN2B-AS1 rs4977756 and POAG in Chinese [26], Japanese [27], Pakistan [28], and American Caucasian [29] populations have shown inconsistent results. For limited studies and relatively small sample size in the correlation of CDKN2B-AS1 rs10120688, rs2157719, and rs7049105 and POAG, inconclusive results were observed.
Meta-analysis is powerful to obtain a more precise conclusion that was inconclusive in previous individual study. Considering the inconclusive results and limited sample size in previous studies, we performed a meta-analysis to further evaluate the genetic associations between CDKN2B-AS1 rs4977756, rs10120688, rs2157719, and rs7049105 polymorphisms and the susceptibility of POAG.

Search strategy
The relevant literatures published as of April 1, 2021 were searched in PubMed, Chinese National Knowledge Infrastructure (CNKI), CQVIP, Chinese Biomedical Literature Database (CBM), Web of science, Embase, and Wanfang database. "cyclin-dependent kinase inhibitor 2B antisense RNA 1" or "CDKN2B-AS1" and "polymorphism" or "single nucleotide polymorphisms" or "SNPs" and "Primary Open-Angle Glaucoma" or "Open-Angle Glaucoma" or "POAG" or "OAG" were used as search terms. At the same time, the references included and related reviews were reviewed. Language is unlimited.

Inclusion and exclusion criteria
The following criteria were used for the literature inclusion: (1) papers should concern CDKN2BAS polymorphism and glaucoma risk, (2) case-control and/or cohort designed studies, (3) contained SNP genotype data both in case and control groups, (4) adequate data for the calculation of odds ratios (ORs) and 95% confidence intervals (CIs), (5) the genotype distribution in control groups was in Hardy-Weinberg equilibrium (HWE). In addition, studies were excluded when they were (1) studies that contained overlapping data with other literatures; (2) data came from case-reports, reviews, or abstracts; (3) not case-control and/or cohort designed studies; (4) genotype frequencies were unavailable; and (5) the control group did not confirm to HWE.

Data extraction
Liu SS and Chen SW independently extracted all data from each eligible study. Disputes was resolved by discussion. The following information from each study were extracted: the name of first author, publication, ethnicity, mean age, gender, type of glaucoma, mean intraocular pressure (IOP) (mm Hg), mean vertical cup-to-disc ratio (VCDR), and the number of cases and controls.

Statistical analysis
Data was processed with RevMan 5 (Oxford, UK) and STATA12.0. The association between CDKN2B-AS1 polymorphisms and POAG risk was evaluated by pooled OR) and 95%CI. The significance of the pooled OR was assessed by the Z test. I 2 was used to evaluate the heterogeneity between studies. If I 2 was less than 50% (p > 0.05), the combined effect value OR and its 95%CI were calculated using the fixed-effect model; otherwise, the random effects model was used. Subgroup analyses were conducted based on ethnicity. Egger's test and Begg's test were used to evaluate and analyze the publication bias of each study. Metaregression analysis was used to analyze the potential sources of inter-study heterogeneity. p value < 0.05 was considered to be significant difference.

The characters of eligible studies
After an exhaustive search, a total of 2154 articles were retrieved from electric databases. According to the inclusion and exclusion criteria, we excluded 507 duplicated studies, 382 not original studies, and 1244 not case-control designed studies (Fig. 1). Finally, 11 studies with 8290 cases and 13,485 controls were enrolled in the present meta-analysis [15,20,23,[25][26][27][28][29][39][40][41]. Among these articles, 7 studies refer to rs4977756, 3 studies refer to rs10120688 and rs2157719 respectively, and 2 studies refer to rs7049105. Furthermore, 4 studies were conducted on Asian populations, and 7 were on Caucasian populations. The basic information and genotype distribution of the included literatures are shown in Table 1.

Sensitive analysis and publish bias
There was no deviation from Hardy-Weinberg equilibrium or OR 3.0 in the 11 studies on CDKN2B-AS1 rs4977756, rs10120688, rs2157719, and rs7049105 polymorphisms in the present meta-analysis. Thus, the association between CDKN2B-AS1 rs4977756, rs10120688, rs2157719, and rs7049105 and POAG was stable and reliable. As shown in Table 3 and Figs. 3 and 4, no evidence of publication bias was found by using Begg' s test and Egger' s test.

Discussion
In the present study, we have assessed the genetic association of CDKN2B-AS1 rs4977756, rs10120688, rs2157719, and rs7049105 and the risk of POAG and found that the CDKN2B-AS1 rs4977756 and rs10120688 can significantly increase the risk of POAG. However, the CDKN2B-AS1 rs4977756 was associated with POAG risk only in Caucasian, but not in Asian. In addition, the rs2157719 decreases the POAG risk only in Asian. No association between rs7049105 and the risk of POAG was detected.
GWAS data have identified significant associations between several genes in chromosome 9p21 and multiple common diseases [30][31][32]. CDKN2B-AS1 is a non-coding gene (also known as ANRIL) with an unknown function, located on chromosome 9p21.3 [33]. Multiple tissue-specific splice variants of this gene have been reported. This gene is likely to play a role in regulating expression of genes through epigenetic mechanisms. CDKN2B-AS1 polymorphisms were illustrated to be associated with coronary artery disease (CAD) [34] and type 2 diabetes mellitus risk initially [35]. Subsequently, the polymorphisms in CDKN2B-AS1 were identified to be significantly associated with the development of POAG [29]. However, little is known about the biological meanings underlying this locus. Song et al. have shown that 33 enhancers were identified in 9p21, some of which being within CDKN2B-AS1 [36]. Thus, it is suggested that the variants in this gene would probably affect the expression level of the downstream genes CDKN2A and CDKN2B, which is response to the elevated IOP in glaucoma. Altered gene expression may also affect cell cycle regulation and lead to a tendency toward apoptosis of retinal ganglion cells. The cells of POAG patients carrying the risk alleles of CDKN2B-AS1 gene may be more sensitive to IOP. Moreover, the variants in this gene might also affect the distant genes that correlated with the complex pathogenesis of glaucoma. The rs4977756 is mapped 59 kb telomeric to CDKN2B within a 122-kb region of LD at 9p21.3, which encompasses the CDKN2A-CDKN2B tumor suppressor genes and was shown to be associated with the risk of glioma [37,38]. In this meta-analysis, a total of 5144 cases and 11,285 controls from 7 publications were included. And the CDKN2B-AS1 rs4977756 polymorphism was significantly associated with increased POAG risk. To our knowledge, this is the first time a significant genetic association between CDKN2B-AS1 rs4977756 polymorphism and the susceptibility of POAG was detected using a meta-analysis. As for ethnicity, rs4977756 polymorphism was associated with increased risk of POAG in Caucasians, but not in Asians. Interestingly, GWAS data has demonstrated that the rs4977756 polymorphism was a risk factor for POAG in Japanese, which may suggest that the CDKN2B-AS1 locus in previous study using a Japanese population seemed to be shared with the Caucasian subjects, but not with the Chinese and other Asian populations. Thus, the CDKN2B-AS1 rs4977756 for POAG is still ethnicity related.
The current meta-analysis had various limitations. Firstly, the ethnicity included in the present meta-analysis was relatively limited. Subgroup analysis was only divided into Caucasian and Asian. Considering the significant influence of African ancestry in the pathogenesis of POAG, studies on the association between CDKN2B-AS1 polymorphisms and POAG risk in other ethnicities such as African and Latin American are necessary in the future. Secondly, age and gender were shown to be involved in the pathogenesis of POAG. However, we could not perform further stratification analyses based on age and gender due to limited data. Thirdly, we failed to assess the potential influence of genetic and environmental factors, as well as the gene-gene and gene-environment interactions on POAG for lack of relevant data. Fourthly, CDKN2B-AS1 rs4977756, rs10120688, rs2157719, and rs7049105 polymorphisms would not be enough to explain the associations between CDKN2B-AS1 gene and POAG risk. Fifthly, the dominant and recessive models of these polymorphisms were not available for lack of relevant data.

Conclusions
The CDKN2B-AS1 rs4977756 might increase the POAG risk in Caucasian population. rs10120688 might be associated with an increased risk of POAG. rs2157719 might decrease the risk of POAG in Asian population. To confirm these results, a larger number of subjects with different ethnicity are necessary to explore the role of CDKN2B-AS1 rs4977756 and rs10120688 in POAG.
Author contribution Data curation, methodology, software, validation: Liu SS and Chen SW. Project administration, Funding acquisition: Niu TT. Writing-original draft: Niu TT. Writing-review and editing: Liu SS and Chen SW.
Funding This work was supported by the National Natural Science Foundation of China (Grant 81271001).

Availability of data and material
The data used to support the founding of this study are included within the article.
Code availability Not applicable.

Declarations
Ethics approval Not applicable.

Competing interests The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.