【学位级别】医学博士

【论文完成日期】2004-5-18

【论文中文题名】NET与DAT基因多态性与哌甲酯治疗ADHD疗效的关联

【论文外文题名】Associatin of the norepinephrine transporter gene (NET) and dopamine transporter gene (DAT) polymorphisms with methylphenidate response in ADHD

【论文著者】姓名：杨莉 yang li

【论文著者】学号：B10199765

【论文著者】系别：精神卫生研究所

【论文著者】专业：精神病与精神卫生学

【论文著者】研究方向：注意缺陷多动障碍

【指导教师】姓名：王玉凤 wang yu feng

【指导教师】学校：北京大学医学部

【指导教师】系别：精神卫生研究所

【指导教师】专业：精神病与精神卫生学

【中文文摘】目的 注意缺陷多动障碍(ADHD)是儿童期最常见的行为障碍，中国学龄儿童患病率为4.31%~5.83%，国外报告为3%~6%。哌甲酯可减轻ADHD患者的注意和多动冲动症状，但约30%患者治疗无效。已知遗传多态性是影响治疗反应的主要因素之一，本研究目的是运用药物基因组学理论和方法 (1) 研究去甲肾上腺素和多巴胺转运体基因多态性与哌甲酯疗效的关联；(2)分析遗传与非遗传因素对疗效的影响；(3) 寻找哌甲酯治疗反应的预测指标。 方法 对门诊符合DSM-IV ADHD诊断标准且适宜哌甲酯治疗的汉族患者，采用国际标准滴定程序滴定哌甲酯剂量，以父母和教师填写的Conners儿童行为量表以及ADHD评定量表为工具评定疗效。在家长知情同意后留取患儿静脉血，用蛋白沉淀法提取DNA。选择去甲肾上腺素转运体(NET)基因1148-13 C/A、G1287A、1389+9 G/A以及多巴胺转运体(DAT)基因C-839T、1398-21 G/A和40bp VNTR为候选基因多态性，通过聚合酶链反应(PCR)扩增包含候选位点的DNA片段，再经限制性酶切或直接电泳检测基因型。采用关联分析的方法对表型和基因型资料进行分析，并用logistic回归分析影响疗效的因素，建立预测模型。 结果 1．单个基因多态性与哌甲酯疗效的关联分析； 1) NET G1287A和1389+9 G/A多态性与哌甲酯疗效存在关联，特别是与多动冲动的疗效，治疗无效者两个多态性位点A/A基因型频率均高于有效者，G/G和G/A基因型频率均低于有效者(p < 0.01)；在男性、低年龄组、治疗前症状为中度和日服药2次或3次者，上述关联均存在。多态性1148-13 C/A与疗效无关联。 2) DAT基因多态性中，1398-21 G/A与男孩多动冲动的疗效存在关联，无效者G/A基因型频率高于有效者，A/A基因型频率低于有效者(p < 0.05)。40bp VNTR和C-839T多态性与疗效的关联无显著性(p > 0.05)，但男孩Conners多动指数无明显下降者有40bp VNTR 10/10基因型频率高的趋势(p = 0.088)，高年龄、治疗前中度症状和日服药2次或3次的患者注意缺陷无效者有10/10基因型频率低的趋势(p值分别为0.093，0.068，0.094)，日服药2次或3次、Conners多动指数无明显下降者与C-839T T/T基因型频率高，T/C基因型频率低有关联趋势(p = 0.058)。 3) 在极端敏感和抵抗表型中，NET G1287A与疗效的关联有显著性(p < 0.05)，药物抵抗者A/A基因型频率高于敏感者，G/G基因型频率低于敏感者；NET 1389+9 G/A与疗效的关联有显著性趋势(p = 0.095)。 2．单倍型分析 对上述分析发现与疗效存在稳定关联的NET G1287A和1389+9 G/A多态性进行单倍型分析，估计单倍型1287G/1389+9G和1287A/1389+9A的频率共计96.85%，单倍型1287G/1389+9A和1287A/1389+9G的频率仅3.15%，两位点连锁不平衡模型较不连锁模型拟合优度的改善有统计学意义(p < 0.001)，D = 0.21，D’ = 0.97。 3．多因素logistic回归分析 在NET 1389+9 G/A多态性、DAT 1398-21 G/A多态性、性别、年龄、以及治疗前严重程度诸因素中，NET 1389+9 G/A多态性对疗效的影响有统计学意义(p < 0.01)。以该多态性位点基因型为自变量，建立疗效预测的logistic回归模型为 logit (P) = -1.645 + 1.981 X1 结论 1． NET G1287A和1389+9 G/A多态性与哌甲酯疗效有关，尤其是多动冲动的疗效。DAT1 1398-21 G/A多态性与男孩多动冲动的疗效有关。NET 1148-13 C/A多态性与哌甲酯疗效无关。 2． NET G1287A和1389+9 G/A多态性很可能存在连锁不平衡，二者与疗效的关联可能都与1389+9 G/A多态性的功能有关，或者可用其附近的其它功能多态性解释。 3． 在遗传和非遗传因素中，NET 1389+9 G/A多态性进入预测模型，A/A基因型患者预示治疗可能无效。

【外文文摘】Objectives Attention-deficit hyperactivity disorder (ADHD) is the most common psychiatric disorder of childhood, affecting 4.31-5.83% of school age children in China and 3-6% in the world. Although it has been widely demonstrated that methylphenidate attenuates inattentive and hyperactive-impulsive symptoms of ADHD, about 30% of patients show poor response. As the major factors affecting medication response are genetic polymorphisms, the objectives of this study were to (1) analyze the association between polymorphisms of norepinephrine and dopamine transporter gene and methylphenidate response; (2) analyze the effect of genetic and other factors on the drug response; and (3) find the predicting marker for the response of methylphenidate treatment, using the principles and methods of pharmacogenomics. Method Subjects were drawn from the clinic samples of the Sixth Hospital, Peking University. Appropriate Han children with DSM-IV ADHD were titrated with methylphenidate. The medication responses were rated using the Conners Rating Scale and the ADHD Rating Scale completed by parents and teachers. Blood samples were obtained with parents’ informed content. DNA was extracted with protein-depositing method. The NET 1148-13 C/A, G1287A, 1389+9 G/A, DAT C-839T, 1398-21 G/A and 40bp VNTR were selected as the candidate gene polymorphisms. After the specific gene fragment was amplified by PCR, the genotyping was performed by direct agarose gel electrophoresis or followed the digestions by restriction enzyme. Then genotypes were tested by χ2 to assess the significance of any association with phenotypes. The factors affecting treatment response were determined by logistic regression, and predicting model was constituted. Results 1．Associations between specific gene polymorphism and methylphenidate response 1) NET G1287A and 1389+9 G/A were found to be significantly associated with methylphenidate response, especially the response of hyperactivity and impulsion. At these two loci, patients with poor response had more A/A genotypes but less G/G and G/A genotypes than those with good response (p < 0.01). The associations were still significant for boys, young children, patients with moderate symptoms before treatment, and patients who took the drug twice or three times per day. The association of 1148-13 C/A polymorphism and response was not significant. 2) For DAT gene polymorphisms, 1398-21 G/A was associated with response in hyperactive and impulsive symptoms of boys, with more G/A genotypes but less A/A genotypes in poor responders than in good responders(p < 0.05). The 40bp VNTR and the C-839T polymorphisms were not significant (p > 0.05). Tendencies were found that frequency of VNTR 10/10 genotype was higher in boys with little improvement for the hyperactive index of Conners Rating Scale compared to those with much improvement(p = 0.088), but among older children, patients with moderate symptoms and patients who took the drug twice or three times per day, the frequency of 10/10 genotype was lower in nonresponders for inattentive symptoms compared to good responders, (p = 0.093, 0.068, 0.094 separately). And in patients who took the drug twice or three times per day, children with little improvement compared with much improvement for the Conners’ Index showed more tendency of high frequency of C-839T T/T genotype but low T/C genotype (p = 0.058). 3) Re-evaluation of specific gene polymorphism in extremely sensitive and resistant phenotypes implied the significant association of NET G1287A and medication response (p < 0.05), the frequency of A/A genotype was higher, but that of G/G genotype was lower in the extremely sensitive patients than in the extremely resistant patients. The NET 1389+9 G/A presented tendency of association (p = 0.095). 2．Haplotype Analysis The haplotypes were estimated for NET G1287A and 1389+9 G/A which showed significant association with methylphenidate response. The haplotype frequency of 1287G/1389+9G and 1287A/1389+9A were 96.85%, and that of 1287G/1389+9A and 1287A/1389+9G were 3.15%. The improvement of the simulation with the linkage disequilibrium model was statistically significant (p < 0.001). D = 0.21, D’ = 0.97. 3．Logistic regression of multiple factors Among several potential factors including NET 1389+9 G/A polymorphism, sex, age and severity of symptoms before treatment, the effect of the NET 1389+9 G/A on the drug response was statistically significant (p < 0.01). The logistic regression model including it was constituted as follow: logit (P) = -1.645 + 1.981 X1 Conclusions 1． NET G1287A and 1389+9 G/A were associated with methylphenidate response, especially the response of hyperactivity and impulsion. DAT 1398-21 G/A was associated with response in hyperactive and impulsive symptoms of boys. NET 1148-13 C/A was unrelated with the drug response. 2． NET G1287A and 1389+9 G/A are probably in linkage disequilibrium. The association of these two loci and the methylphenidate response may be due to the function of 1389+9 G/A, or it can be interpreted by other functional polymorphism nearby. 3． Among several potential genetic and non-genetic factors, NET 1389+9 G/A entered the predicting model for methylphenidate response, and the homozygosity of A allele characterized poor response to methylphenidate therapy.