PGRN Is Associated With Late-Onset Alzheimer's Disease:A Case-Control Replication Study And Meta-Analysis Of Related Literature

ObjectiveThis study is mainly focused on the associations between the polymorphism of PGRNrs5848 and the genetic susceptibility to LOAD in Northern Chinease.SubjectsOur study consisted of 992 sporadic LOAD patients(?65 years;584 female)and 1358healthy control subjects(?65 years;748 female)matched for age and gender.All subjects in our study were unrelated northern Han Chinese residents originally from Shandong province.All patients wereclinically diagnosed as?probable AD?according to the criteria of the National Institute of Neurological and Communication Disorders and Stroke-Alzheimer Disease and Related Disorders Association(NINCDS-ADRDA).MethodGenomic DNA was isolated from peripheral blood leukocytes of AD patients and healthy controls using the Wizard Genomic DNA Purification Kit.The SNPs(rs5848 for GRN,rs429358 and rs7412 for APOE)were genotyped using the improved multiplex ligase detection reaction(imLDR)technique.To identify all articles that examined the association of GRN polymorphism with AD,we conducted a systematic literature search of PubMed,EMBASE,the Cochrane library,and BIOSIS previews up to July 2015 using a combination of the following terms?progranulin?or?PGRN?,?rs5848?,?polymorphism?and?Alzheimer's disease?or?AD?For the limitation of language,only English publications were included.The study quality was assessed by the Newcastle-Ottawa quality assessment scale(NOS)on the following aspects:(1)the selection,(2)the comparability,and(3)the exposure.Studies with a score equal to or higher than 7 were considered.to be high quality.Deviations of the characteristics for case and controls were examined by the?test.Differences in allele and genotype frequencies of the two groups were assessed using the Pearson?test.Multivariate logistic regression analyses,adjusting for gender,APOE?4 status(presence or absence ofAPOE?4 allele),and age were used to estimate odds ratio(OR)and the 95%confidence interval(CI).To assess genotypic and allelic associations with LOAD,three various genetic models were defined as 1(TT+CT)versus 0(CC)for dominant,1(TT)versus 0(CT+CC)for recessive,and 0(CC)versus 1(CT)versus 2(TT)for additive(C:major allele;T:minor allele).The significance of the SNP×APOEinteraction was also tested by logistic regression.The level of significance for all statistical tests was defined asP<0.05.All statistical analyses were performed by SPSS17.0 software.The meta-analyses were done with Stata 12.0 software.Firstly,we estimated the risk of rs5848 leading to LOAD by allele model T vs C and then evaluated the risks under two genetic models(dominant model:CT and TT vs CC,recessive model:TT vs CT and TT),respectively.Estimation of the statistical power was performed with STPLAN 4.5software.The random effects(I–V heterogeneity method)model was used to calculate the pooled effect estimates.The pooled OR results were examined by the Z test(P<0.05 was considered statistically significant).The heterogeneity between studies was measured with chi-square test(I~2<50%as lack of heterogeneity).ResultsNo significant differences were observed for age(P=0.187)and gender(P=0.067)between cases and controls.MMSE scores were significantly lower in patients with AD than in control subjects(LOAD:11.94±6.21,control:28.49±1.09,P<0.001).The APOE?4allele frequency was also significantly higher in patients with LOAD than in control subjects(P<0.001,OR=2.451,95%CI=1.995-3.011).In the univariate Analysis,Genotype distribution of rs5848 was in HWE in control subjects(P>0.05).Both the genotype distribution(P=0.009)and the minor allele frequency of rs5848(P=0.005)were associated with LOAD.The multivariate logistic regression analysis adjusting for age,gender,and APOE states found a significant association between rs5848 and LOAD in all the three models(dominant:P=0.038,OR=1.195,95%CI=1.010-1.413;recessive:P=0.025,OR=1.386,95%CI=1.042-1.844;additive:P=0.009,OR=1.187,95%CI=1.044-1.351).The interact-tion between these two loci significantly altered the risk for AD(P<0.05).When the subjects were divided by APOE?4 allele status,the discrepancy of genotype distribution and minor allele frequency of rs5848 between cases and controls were more remarkable in the APOE?4 carriers(genotype distribution:P=0.001,minor allele frequency:P=0.001).However,this diversity did not reach significantly in APOE?4 allele noncarriers.After application of the search strategy,four studies met the inclusion criteria.Combing our study,the pooled subjects included 3236 cases and 3405 controls from both Caucasians(three studies)and Asians(two tudies).The allele model revealed that T allele had association with a higher risk for AD(OR=1.20,95%CI=1.10-1.31).We identified that rs5848 was significantly associated with increased risk of AD both in dominant model(OR=1.199,95%CI=1.017–1.413)and recessive model(OR=1.473,95%CI=1.114-1.947).In the following subgroup analysis,the Asian group remains significant in all three models(allele model:OR=1.21,95%CI=1.07-1.34;dominant model:OR=1.20;95%CI=1.02-1.41;recessive model:OR=1.47;95%CI=1.11-1.95),while in caucasians,the association were only found in recessive model(OR=1.24,95%CI=1.00-1.53).Begg's test for the pooled analysis of different models showed that there was no significant publication bias in this meta-analysis(P>0.05).ConclusionOur study suggests that the polymorphism of PGRNrs5848 is significantly associated with the susceptibility to LOAD in northern Han Chinease.It indicated that PGRN gene mutation may paly a important role in the pathogenesis of AD.