Gsk3beta Rs334558 And Rs6438552 Polymorphism

Published Date: 02 Nov 2017

1.Introduction

Parkinson’s disease (PD; Online Mendelian Inheritance in Man, OMIM #168600) is the most common movement disorder in the elderly, characterized by bradykinesia, rigidity, tremor, and postural instability(Van Den Eeden et al., 2003). This neurodegenerative disorder is caused by selective midbrain dopamine cell depletion(Forno, 1996) ,which results in a deficiency of the dopamine in the basal ganglia circuit. The exact etiology of PD has not been fully elucidated, however, accumulated evidence suggested that multiple factors contributed to the pathogenesis of PD, especially environmental and genetic factors(Schapira and Jenner, 2011) .

The protein encoded by GSK3beta(located on chromosome 3q13.3) is a serine-threonine kinase and is widely expressed in human tissue, with particularly abundant levels in the brain(Embi et al., 1980). It is known to be involved in several important physiological processes, such as affecting the antioxidant cell defense (Rojo et al., 2008) , phosphorylating several nuclear and cytoplasmic proteins including tau (Jope and Johnson, 2004)and regulating the interactions between neural stem cell and endothelial cell (Li et al., 2011). Moreover, GSK3beta is also considered to participate in the pathogenesis of several neurodegenerative diseases, such as PD, Alzheimer’s disease (Kalinderi et al., 2011; Zhang et al., 2011).Additionally, Kwok’s research(Kwok et al., 2005) showed that two functional single nucleotide polymorphisms (rs334558 and rs6438552) altered transcription and splicing of GSK3beta and were associated with PD risk.

Various studies have investigated the relationship between GSK3beta polymorphism (rs334558 and rs6438552) and PD risk, but the results remain controversial(Kwok et al., 2005; Goris et al., 2007; Infante et al., 2010; Kalinderi et al., 2011; Wider et al., 2011; Zhao et al., 2012).Three studies about Caucasian population and Hong Kong Chinese population didn’t detect significant association between the two SNPs and PD risk without considering gene-gene interactions(Kwok et al., 2005; Goris et al., 2007; Wider et al., 2011). However, some studies about mainland Chinese and Greek population indicated that the GSK3beta rs334558 polymorphism could reduce the risk of PD. Furthermore, the data from a Spanish cohort showed that AA genotype of GSK3beta rs6438552 polymorphism increased PD susceptibility (Infante et al., 2010).Hence, we performed this meta-analysis to combine eligible studies and explore the association between two functional GSK3beta variants (rs334558 and rs6438552) and PD susceptibility.

2. Materials and methods

2.1.Identification of eligible studies

A systematic literature search of MEDLINE, Embase , Cochrane Library, China National Knowledge Infrastructure (CNKI) and the Chinese Biomedicine Database was conducted (the last search update was performed on January 11,2013) by using the following keywords "glycogen synthase kinase 3", "GKS 3" ,"GSK-3" ,"glycogen synthase kinase 3 beta", "glycogen synthase kinase 3B", "glycogen synthase kinase 3β" ,"GSK3beta", "GSK3B", "GSK3β","GSK-3beta" , "GSK-3B" ,"GSK-3β"and ‘‘Parkinson’s disease’’, ‘‘Parkinson disease’’, ‘‘Parkinsons disease’’, ‘‘PD’’ in combination with ‘‘variant’’, ‘‘mutation’’ ,"polymorphism,’’ , ‘‘genetics’’, "SNP" or "single nucleotide polymorphism". In addition, we performed a handsearch of the related reference articles to identify other relevant publications. We have only used data from the fullpublished paper, not from any unpublished data. Moreover，all researches were limited to English and Chinese language articles. In this meta-analysis, the following inclusion criteria were used for selecting the studies: (1) articles about the GSK3beta polymorphism and PD susceptibility, (2) case-control design, and (3) sufficient genotype data for estimating an odds ratio (OR) with a 95% confidence interval (CI). Articles that were not about PD research, contained duplicated previous research, did not include usable genotype data, or the study only involved a case population were excluded. No restrictions we placed on ethnicity or geographic area.

2.2.Data extraction

One investigator (Yongsheng Yuan) independently extracted all eligible publications using the selection criteria listed above , and the result was reviewed by the second investigator (Qing Tong). Any disagreement was resolved by discussion with our research team. We extracted the following information from each eligible publication: the first author’s name, year of publication, region, ethnics of the study population(Chinese or Caucasian ) and controls, genotyping method , the Hardy-Weinberg equilibrium (HWE) among the controls, number of cases and the available genotype and allele frequency information form the GSK3beta rs334558 and rs6438552 polymorphism

2.3. Statistical analysis

We conducted this meta-analysis to explore the relationship between the GSK3beta rs334558 and rs6438552 polymorphism and PD susceptibility. Allele frequencies were estimated using the allele counting method. The Chi-square test was used to assess Hardy-Weinberg equilibrium in the control groups of each study. All studies with control groups that were not in HWE (p<0.05) were excluded. We used the OR and 95% CI in all involved case-control studies to evaluate the strength of the association between the GSK3beta polymorphism and PD susceptibility. The following four genetic models of inheritance were examined : dominant model (GA+AA versus GG), recessive model(AA versus AG+GG),codominant model(AA versus GG and AG versus GG) and additive model(2AA + AG versus 2GG + GA) ; then subgroup analyses were performed based on ethnicity(Chinese or Caucasian).

The heterogeneity between studies was assessed by the chi-square test-based Q-statistic (Cochran, 1954) and if Q-statistic P < 0.05 , indicating that the between-study heterogeneity was significant, then the random-effect model (DerSimonian and Laird, 1986) was selected to analyze the combined ORs , otherwise the fixed-effect model was selected(Mantel and Haenszel, 1959) . We used I2 (I2 = 100% × (Q-df)/Q) statistic to quantify the between-study heterogeneity, with I2 <25%, 25-75% and >75% representing low, moderate or high degrees of heterogeneity, respectively (Higgins and Thompson, 2002; Higgins et al., 2003). Z test (P < 0.05 was considered statistically significant) was employed to evaluate the significance of the pooled OR. Additionally, sensitivity analyses were performed after sequential removal of each study to assess the influence of each study on the overall estimate and examine the possible biases of this meta-analysis. Cumulative meta-analyses were also conducted through an assortment of all included studies by case sample size. The Begg’s funnel plot and Egger’s linear regression test were chosen to assess the potential publication bias (P< 0.05 was considered a significant publication bias)(Egger et al., 1997).

All analyses were performed using statistical software STATA/SE (version 12.0, STATA Corporation, College Station, Texas).Two-sided P values less than 0.05 were considered statistically significant, unless otherwise stated.

3. Results

3.1.Eligible studies

The article by Kwok and colleague(Kwok et al., 2005) contained two separate case-control studies, which were separately included in the pooling analysis. Therefore, our meta-analysis consisted of a total of six articles including seven studies in English met the inclusion criteria. The characteristics of eligible studies are listed in Table 1. Infante’s(Infante et al., 2010) article only detected the association between rs6438552 and PD risk, and Zhao’s(Zhao et al., 2012) publication only detected the relationship between rs334558 and PD risk. Two of the seven studies were about Chinese population ,and the others were about Caucasian from Australia, United Kingdom, Greece, USA, Ireland, Spain and Norway. All studies were stratified into two groups(Chinese or Caucasian ) according to ethnicity. The detailed selection procedure of the included publications was shown in Fig.1.

3.2.Data Synthesis

The main results of meta-analysis for the association between GSK3beta polymorphism (rs334558 and rs6438552) with PD susceptibility was shown in Table 2 and 3. The ORs with 95% CIs for each genetic model were calculated both in random-effect model and fix-effect model , and if Pheterogeneity <0.05,the results from random-effect model were used, otherwise, the results from fixed-effect model were adopted.

3.2.1.GSK3beta promoter rs334558 polymorphism and PD susceptibility

As shown in Table 2, no association between GSK3beta promoter rs334558 polymorphism and PD was found in all genetic models based on all studies (dominant model-GA+AA versus GG: OR=0.97, 95% CI: 0.70,1.35; recessive model-AA versus AG+GG: OR = 1.02, 95% CI: 0.92, 1.14;codominant model-AA versus GG: OR = 0.99, 95% CI: 0.85, 1.16, and AG versus GG: OR = 0.98, 95% CI: 0.68, 1.40; additive model-2AA + AG versus 2GG + GA: OR = 0.98, 95% CI: 0.87, 1.10).

In the Chinese population, the AG genotype, GA+AA genotype and 2AA + AG genotype reduced the PD risk compared with GG genotype, GG genotype and 2GG + GA genotype, respectively(OR=0.60, 95%CI:0.48,0.74; OR=0.63,95%CI:0.51,0.78; OR=0.82,95%CI:0.71,0.94, respectively). However, no significant association was found in all genetic models in the Caucasian population. As shown in Table 2, the ORs and 95% CIs from the fixed-effect model and random-effect model for each genetic model didn’t bring about different outcomes, particularly in the Chinese population, the fixed-effect model and random-effect model almost created the same results. Hence, we used the forest plot(Fig.2) in the dominant model based on the random-effect model to describe the pooled ORs in the overall population ,Chinese population and Caucasian population.

3.2.2.GSK3beta intronic rs6438552 polymorphism and PD susceptibility

As shown in Table 3, we also had not detected any association between GSK3beta intronic rs6438552 polymorphism and PD for all genetic models based on the studies (dominant model-GA+AA versus GG: OR = 1.00, 95% CI: 0.87,1.15; recessive model-AA versus AG+GG: OR = 1.03, 95% CI: 0.92,1.14;codominant model-AA versus GG: OR = 1.03, 95% CI: 0.89,1.21, and AG versus GG: OR = 0.99, 95% CI: 0.86,1.15; additive model-2AA + AG versus 2GG + GA: OR =1.00, 95% CI: 0.88,1.13).

Similar to GSK3beta promoter rs334558 polymorphism , in the Chinese population, the AG genotype, GA+AA genotype and 2AA + AG genotype for rs6438552 reduced the PD risk compared with GG genotype, GG genotype and 2GG + GA genotype, respectively(OR=0.62, 95%CI:0.39,0.97; OR=0.57,95%CI:0.37,0.87; OR=0.66,95%CI=0.49,0.88 ,respectively). Moreover, AA genotype also reduced the PD risk compared with GG genotype(OR=0.45,95%CI:0.24,0.84). However, no significant association was found in all genetic models in the Caucasian population. Here, the forest plot(Fig.3) in the dominant model based on the fixed-effect model was selected to perform a vivid description of the pooled ORs in the overall population ,Chinese population and Caucasian population.

3.3.Sensitivity analysis

Due to the significant between-study heterogeneity for GSK3beta rs334558 polymorphism, we conducted a sensitivity analysis to evaluate the effects of each study on the combined ORs by sequential removal of each eligible study. A random-effect model was adopted since significant heterogeneity. The results of the sensitivity analysis showed that none of the studies could affect the significance of the combined ORs in the dominant genetic model. Moreover, the analysis indicated that the study contributing the most to heterogeneity was conducted by Kalinderi et al(Kalinderi et al., 2011).(Fig.4) The overall heterogeneity was moderately decreased by exclusion of that study: OR = 0.97 ,95% CI: 0.70,1.35; Pheterogeneity = 0.000; I2 =82.2 %) and OR=0.84,95% CI: 0.66, 1.08; Pheterogeneity =0.016; I2 = 67.1%) before and after removal, respectively. Furthermore, in the Caucasian population, the exclusion of that study effectively reduced the heterogeneity(before removal: OR = 1.19 ,95% CI: 0.87,1.65; Pheterogeneity = 0.019; I2 =69.9%;after removal: OR =1.02,95% CI: 0.85,1.21; Pheterogeneity =0.953 ; I2 =0.0%), which showed that Kalinderi’s study greatly contributed to the between-study heterogeneity in the Caucasian subgroup .A sensitivity analysis was also done for GSK3beta rs6438552 polymorphism in a fixed-effect model, and the results for the dominant genetic model indicated that the overall estimate of this meta-analysis was stable and the study conducted in China HK by Kwok et al (Kwok et al., 2005)may be the origin of the heterogeneity(data not shown).(Fig.4)

3.4.Test of heterogeneity

There was significant heterogeneity in the heterozygote (AG versus GG), additive model(2AA + AG versus 2GG + GA),and dominant model (GA+AA versus GG)comparisons for GSK3beta promoter rs334558 polymorphism (Table2). Significant heterogeneity also existed in the additive genetic model(2AA + AG versus 2GG + GA) of the GSK3beta intronic rs6438552 polymorphism (Table 3).However, the heterogeneity decreased to varying extent via the analysis of the ethnic subgroups in both variants(Table2,3), which indicated that ethnicity was an important factor for the overall between-study heterogeneity.

3.5.Cumulative meta-analysis

Cumulative meta-analyses of both variants were also conducted through an assortment of all included studies by case sample size (Fig.S1-S2).The results from the analyses showed that GSK3beta rs6438552 and rs334558 polymorphism both tended to be unrelated with PD risk in the overall population as the studies with larger sample size were added. Moreover, the 95%CIs of both variants became narrower and more stable with increasing sample size.

3.6.Evaluation of publication bias

The Begg’s funnel plot and Egger’s linear regression test were selected to evaluate the publication bias of included studies. It had shown that there were no evidence of publication bias in all comparisons of these two variants(data not shown). For instance, the shape of the Begg’s funnel plots seemed symmetrical in the homozygote comparison (AA versus GG ) for rs334558 and rs6438552 (Fig.5), which indicated that there was no significant publication bias.

4.Disccusion

Some basic and postmortem researches have revealed that GSK3beta was involved in the pathogenesis of PD(Chen et al., 2004; Kozikowski et al., 2006; Duka et al., 2009; Nagao and Hayashi, 2009; Wills et al., 2010; Haggerty et al., 2011; Dun et al., 2012; L'Episcopo et al., 2012; Yao et al., 2012).GSK3beta was activated in 6-hydroxydopamin(6-OHDA)-lesioned rats, and the applications with GSK3beta inhibitors could exert a protective effect on the lesioned dopaminergic neurons(Chen et al., 2004; Dun et al., 2012),which was similar with the (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine)MPTP-treated PD models(Kozikowski et al., 2006; Duka et al., 2009; L'Episcopo et al., 2012). Yao’s study (Yao et al., 2012)demonstrated that bis(3)-cognitin attenuated the toxic effects of 1-methyl-4-phenylpyridiniumion(MPP+)by inhibiting the activity of GSK3beta. Moreover, elevated tauopathy and GSK-3beta were detected in a transgenic model of PD (Haggerty et al., 2011) and postmortem studies(Nagao and Hayashi, 2009; Wills et al., 2010). However, the results from a series of case-control studies about GSK3beta polymorphism and PD susceptibility remained inconsistent and inconclusive(Kwok et al., 2005; Goris et al., 2007; Infante et al., 2010; Kalinderi et al., 2011; Wider et al., 2011; Zhao et al., 2012).

To the best of our knowledge, we performed the meta-analysis of GSK3beta polymorphism and PD susceptibility for the first time. The results demonstrated that GSK3beta polymorphism existed a protective effect on Chinese population not on Caucasian population and overall population.

Several limitations should be considered in the meta-analysis. Firstly, misclassifications on genotypes and disease status ,differences in control characteristics and various genotyping methods may have an impact on the results. Secondly, our literature searching was limited to English and Chinese language, so language bias might be considered. Thirdly, gene-gene interactions were not taken into account in this meta-analysis ,which may affect the results. Fourthly, there were only two studies concerning rs334558 and one study concerning rs6438552 in the subgroup analysis of Chinese population, which indicated that the statistical power was low, hence the results of Chinese subgroup should be cautiously interpreted. Finally, we combined unadjusted estimates because of the absence of available information. A more accurate analysis would be conducted based on adjusted estimates corrected for age at onset, sex and other factors if possible.

Despite some limitations ,our meta-analysis effectively increased the statistical power by pooling results from different studies and didn’t show significant publication bias. In summary, this meta-analysis found null association between two GSK3beta SNPs(rs334558 and rs6438552) and PD risk in the overall and Caucasian population, however, the two variants both reduced the risk of PD in the Chinese subgroup. More well-designed case-control studies with larger sample size are needed to affirm this association in Chinese.

Acknowledgments

Thanks to the authors of the primary studies.

Supplementary data

Fig.S1.Cumulative meta-analysis of associations between the GSK3beta promoter rs334558 polymorphism and PD risk in the dominant model based on the random- effect model.

Fig.S2. Cumulative meta-analysis of associations between the GSK3beta intronic rs6438552 polymorphism and PD risk in the dominant model based on the fixed- effect model.