【作者】 寻广磊；

【导师】 赵靖平；

【作者基本信息】 中南大学 ， 精神病与精神卫生学， 2010， 博士

【摘要】 目的：(1)研究Reelin信号通路相关基因的多态性与汉族儿童孤独症的关联；(2)研究Reelin信号通路相关基因的交互作用与汉族儿童孤独症易感性的关系；(3)研究Reelin信号通路相关基因的多态性与汉族儿童孤独症临床表型的关系。方法：(1)采用Illumina CNV 370-Duo芯片对455例符合DSM-IV诊断标准的儿童孤独症患者及97例正常对照进行全基因组SNP分型,从中选取Reelin信号通路6个基因(RELN, DAB1、VLDLR、LRP8、FYN, CDK5)所有SNP的数据,与中南大学医学遗传学国家重点实验室遗传资源库的另外358例正常对照的相应数据合并,对数据进行筛选,然后对每个基因均采用Haploview4.1软件进行单位点的病例-对照关联分析,以及基于连锁不平衡分析基础上的单体型病例-对照关联分析。置换检验被用于对关联分析的结果进行校正(置换次数=1000)。(2)采用“标记SNP+连锁结构域”策略选择基因间交互作用研究的SNP位点,6个基因共计入选48个位点。多因子降维分析(MDR)方法被用于探索孤独症患者Reelin信号通路相关基因间的交互作用,选择具有最大的交叉验证一致性、最大检验准确度且置换检验结果具有显著性的组合作为基因间交互作用的最佳模型。(3)剔除15例临床资料不完整的患者,将440例孤独症患者按照“退化”临床表型分为退化组(180例)和非退化组(260例),按照“不遵从行为”临床表型分为不遵从组(177例)和遵从组(263例),χ2检验被用于比较上述48个位点基因型频率和等位基因频率的组间差异。结果：(1)经置换检验后,单位点关联分析显示RELN的99个SNPs与孤独症均无显著关联(P均大于0.05)；单体型关联分析发现AAGTATCATGGG单体型频率和CTTAGCCG单体型频率在对照组显著高于孤独症组(χ2分别为8.658,10.000；校正后的P值分别为0.015,0.002)；(2)经置换检验后,单位点关联分析显示DAB1的154个SNPs与孤独症均无显著关联(P均大于0.05),单体型关联分析发现GGTTTGGCATTTC单体型频率在孤独症组显著高于对照组(χ2=8.461,校正后的P=0.019)；(3)经置换检验后,LRP8、VLDLR、FYN和CDK5所有SNPs的等位基因频率及单体型频率在孤独症组和对照组无显著性差异(P均大于0.05)；(4) MDR分析显示三位点基因型组合(rs1454627-rs6943822-rs722187)是孤独症患者Reelin信号通路相关基因间交互作用的最佳模型(交叉验证一致性9/10,检验准确度0.822,P<0.001)；(5)RELN的rs10487160位点的基因型分布和等位基因分布在退化型孤独症亚组和非退化型孤独症亚组间均存在显著性差异(χ2分别为10.570和10.190,P值分别为0.005和0.001),退化亚组GG基因型频率与G等位基因频率显著高于非退化亚组；rs12666897位点的基因型分布和等位基因分布在两个亚组间亦存在显著性差异(χ2分别为6.019和4.874,P值分别为0.049和0.027),退化亚组AC基因型频率及C等位基因频率显著高于非退化亚组；(6)RELN的rs17157128位点基因型分布和等位基因分布在不遵从组和遵从组之间有显著性差异(χ2分别为9.165和4.472,P值分别为0.010和0.034),不遵从亚组AA基因型频率和A等位基因频率显著高于遵从亚组；(7)DAB1的rs3131735位点的基因型分布在不遵从亚组和遵从亚组间存在显著性差异(χ2=8.430,P=0.015),AG基因型频率在不遵从组的显著高于遵从组；(8) LRP8的rs1288520位点基因型分布和等位基因分布在退化组和非退化组之间有显著性差异(χ2分别为7.790和7.095,P值分别为0.020和0.008),GG基因型频率和G等位基因频率在退化亚组显著高于非退化亚组。(9)VLDLR的rs1454627位点基因型分布和等位基因分布在退化组和非退化组之间有显著性差异(χ2分别为7.143和4.127,P值分别为0.028和0.042),TT基因型频率和T等位基因频率在退化亚组显著高于非退化亚组；结论：(1)RELN可能不是中国汉族儿童孤独症的易感基因,其AAGTATCATGGG单体型和CTTAGCCG单体型反而可能是孤独症的保护因素；(2)DAB1上的GGTTTGGCATTTC单体型与孤独症存在显著的正性关联,DAB1可能是中国汉族儿童孤独症的易感基因之一；(3)LRP8、VLDLR、FYN和CDK5的单核甘酸多态性均与中国汉族儿童孤独症无关；(4)DAB1-RELN-VLDLR基因间交互作用与中国汉族儿童孤独症的发病机制有关；(5)RELN的rs10487160位点以及rs12666897位点多态性与孤独症患者的“退化”临床表型有关；(6)RELN的rs17157128位点多态性与孤独症患者的“不遵从行为”临床表型有关；(7) DAB1的rs3131735位点多态性与孤独症患者的“不遵从行为”临床表型有关；(8) LPR8的rs1288520位点多态性与孤独症患者的“退化”临床表型有关；(9)VLDLR的rs1454627位点多态性与孤独症患者的“退化”临床表型有关。更多还原

【Abstract】 Objective:(1) To explore the association of reelin-signaling-pathway-related-genes polymorphisms with autism in Chinese Han population.(2) To explore the relationship between the gene-gene interaction of reelin-signaling-pathway-related-genes and the susceptibility of autism in Chinese Han population.(3) To explore the relationship between reelin-signaling-pathway-related-genes polymorphisms and the clinical phenotype of autism in Chinese Han population.Methods:(1) Genome-wide SNP genotyping was performed using Illumina CNV 370-Duo chip in 455 autistic children fulfilled with DSM-IV-TR criteria for autistic disorder and in 97 healthy controls. SNP data of the six genes in reelin signaling pathway were combined with the corresponding data of additional 358 controls, which was selected randomly from the genetic resource of State Key Laboratory of Medical Genetic of China. Single marker case-control association analysis and haplotype case-control association analysis based on linkage disequilibrium were conducted using Haploview 4.1 software after the data was screened. The significance of results was corrected by permutation test (number of permutation=1000).(2) 48 SNPs in reeelin-signaling-pathway-related-genes were involved in the analysis of gene-gene interaction using strategy of TagSNP & Linkage Block. Multifactor dimensionality reduction (MDR) was used to detect gene-gene interaction in autistic individuals. Model with the highest CV consistency, the highest testing accuracy as well as the significant permutation test results was considered as the best interaction-model.(3) The present study excluded 15 autistic individuals with incomplete clinical data.440 patients were divided into regression subgroup (180) and non-regression subgroup (260) according to the presence or absence of regression, alternatively divided into challenging-behavior subgroup (177) and non-challenging-behavior subgroup (263) according to the presence or absence of challenging behaviors. Chi square test was used to compare the inter-group differences of genotypic and allelic distribution to detect the association between polymorphisms of 48 SNPs mentioned above and autistic clinical phenotypes.Results:(1) Single marker case-control analysis did not find significant association between all 99 SNPs of RELN with autism after the P value was corrected by permutation test (P>0.05). Haplotype case-control analysis revealed that frequency of AAGTATCATGGG and CTTAGCCG were significantly higher in control group than in autism group after the P value was corrected by permutation test (χ2=8.658,10.000 respectively; P=0.015,0.002 respectively).(2) Single marker case-control analysis did not find significant association between all 154 SNPs of DAB1 with autism after the P value was corrected by permutation test (P>0.05). Haplotype case-control analysis revealed that frequency of GGTTTGGCATTTC was significantly higher in autism group than in control group after the P value was corrected by permutation test (χ2=8.461,P=0.019).(3) There were no significant differences in allelic frequency and haplotypic frequency between autism group and control group in all SNPs of LRP8, VLDLR, FYN and CDK5 after the P value was corrected by permutation test (P>0.05).(4) MDR analysis revealed that the three-loci genotypic combination (rs1454627-rs6943822-rs722187) was the best gene-gene interaction model of reeelin-signaling-pathway-related-genes in the etiology of autism (CV consistency=9/10, testing accuracy=0.822, P<0.001).(5) Significant differences were found in genotypic and allelic distribution of rs 10487160 within RELN between the regression subgroup and non-regression subgroup (χ2=10.570,10.190 respectively; P=0.005,0.001 respectively), genotypic frequency of GG and allelic frequency of G in regression subgroup were significantly higher than in non-regression subgroup. Significant differences were found in genotypic and allelic distribution of rs12666897 within RELN between the two subgroups (χ2=6.019,4.874 respectively; P=0.049,0.027 respectively), genotypic frequency of AC and allelic frequency of C in regression subgroup were significantly higher than in non-regression subgroup.(6) Significant differences were found in genotypic and allelic distribution of rs17157128 within RELN between the challenging subgroup and non-challenging subgroup (χ2=9.165,4.472 respectively; P=0.010,0.034 respectively). Genotypic frequency of AA and allelic frequency of A in challenging subgroup were significantly higher than in non-challenging subgroup.(7) Significant differences were found in genotypic distribution of rs3131735 within DAB1 between the challenging subgroup and non-challenging subgroup (χ2=8.430, P=0.015). Genotypic frequency of AG in challenging subgroup were significantly higher than in non-challenging subgroup.(8) Significant differences were found in genotypic and allelic distribution of rs1288520 within LRP8 between the regression subgroup and non-regression subgroup (χ2=7.790,7.095 respectively; P=0.020,0.008 respectively). Genotypic frequency of GG and allelic frequency of G in regression subgroup were significantly higher than in non-regression subgroup. (9) Significant differences were found in genotypic and allelic distribution of rs 1454627 within VLDLR between the regression subgroup and non-regression subgroup (χ2=7.143,4.127 respectively; P=0.028,0.042 respectively). Genotypic frequency of TT and allelic frequency of T in regression subgroup were significantly higher than in non-regression subgroup.Conclusion:(1) RELN is not likely to be susceptibility gene of autism. On the contrary, the haplotypes of AAGTATCATGGG and CTTAGCCG may be protective factors for childhood autism in Chinese Han population.(2) There is a positive association between the DAB1 haplotype of GGTTTGGCATTTC and autism, which suggests that DAB1 may be one of the susceptibility genes of autism in Chinese Han population.(3) There are no association between polymorphisms of LRP8, VLDLR, FYN, CDK5 and the susceptibility of autism in Chinese Han population.(4) Gene-gene interaction of DAB1-RELN-VLDLR may be involved in the etiology of autism in Chinese Han population.(5) Polymorphisms of rs10487160 and rs12666897 within RELN are associated with regression phenotype in autistic individuals.(6) Polymorphisms of rs17157128 within RELN is associated with challenging behavior phenotype in autistic individuals.(7) Polymorphisms of rs3131735 within DAB1 is associated with challenging behavior phenotype in autistic individuals.(8) Polymorphisms of rs1288520 within LRP8 is associated with regression phenotype in autistic individuals.(9) Polymorphisms of rs1454627 within VLDLR is associated with regression phenotype in autistic individuals.更多还原