【作者】 唐劲松；

【导师】 陈晓岗； 刘春宇；

【作者基本信息】 中南大学 ， 精神病与精神卫生学， 2008， 博士

【摘要】 第一部分早发精神分裂症神经网络连接障碍的磁共振研究目的:采用血氧依赖水平的功能磁共振(BOLD—fMRI)磁共振技术,探讨早发精神分裂症静息状态下脑默认网络功能连接情况;采用弥散张量成像(DTI)磁共振技术,探讨早发精神分裂症脑白质完整性情况。方法:采用第四版美国精神障碍诊断与统计手册(DSM-Ⅳ)诊断标准和CCMD-3诊断标准的早发性精神分裂症患者和年龄、性别、教育程度匹配的对照组进行静息状态下fMRI检查和DTI检查。共41例患者和42例正常人完成了MRI扫描,排除头动导致的数据质量问题,进入统计分析的fMRI部分早发精神分裂症病人26例(13女13男),健康对照28例(15女13男);DTI部分早发精神分裂症病人33例(16女17男),健康对照组33例(16女17男)分别对fMRI和DTI数据进行分析。采用功能连接分析方法,提取静息状态下默认网络,在病例组和对照组中分别计算默认网络各脑区两两间的功能连接,比较各连接在两组之间的差异。以默认网络为感兴趣区,比较两组间全脑与默认网络间的功能连接的差异。从DTI数据中提取全脑所有体素(voxel)的FA(fractional anisotropy)值,采用基于体素的分析方法(Voxel-based analysis,VBA)对早发精神分裂症组和对照组进行分析。结果:1.fMRI研究通过种子点网络分析提取了静息状态下默认网络,默认网络共包含13个脑区:楔前叶/后扣带回、双侧额上回、尾状核、腹侧中内侧前额叶、腹侧中内侧前额叶、双侧外侧项叶、双侧颞下回、左侧海马旁回、小脑扁桃体和右侧下半月小叶。早发精神分裂症组在默认网络存在5条异常连接。其中3条连接表现为连接增强:腹侧中内侧前额叶-右侧颞下回(p=0.0078)、腹侧中内侧前额叶-左侧外侧顶叶(p=0.0091)、腹侧中内侧前额叶-背侧中内侧前额叶(p=0.0163)。2条连接表现为连接减弱:右侧外侧顶叶-小脑扁桃体(p=0.0223)、左侧额上回-右侧下半月小叶(p=0.0294)。默认网络与其他脑区的功能连接分析发现,早发精神分裂症组默认网络与双侧额下回框部功能连接增强。2.DTI早发精神分裂症患者的右侧尾状核、右侧前扣带回、右侧额中回、左侧额上回、左侧顶下小叶、左侧下丘脑/海马区的白质FA值较正常对照组低;未见患者组白质FA值显著高于对照组的区域。结论:1.本研究首次报道早发精神分裂症组病人存在默认网络功能的异常。在网络内部,默认网络的连接异常主要为连接增强(与小脑部分的默认网络连接异常表现为连接减弱),网络连接的异常的部位主要涉及中内侧前额叶;网络作为整体与其他脑区直接也存在连接异常,主要表现在与双侧额下回框部的功能连接增强。早发精神分裂症的默认网络连接障碍可能是导致精神分裂症内向性思维、妄想等思维障碍的病理基础之一。2.早发精神分裂症存在广泛的脑白质完整性异常,异常脑区主要涉及前额叶和皮层下脑区神经网络相关脑区,以及顶下小叶。说明早发精神分裂症病人存在广泛的解剖连接障碍。第二部分重性精神疾病基因组DNA拷贝数变异分析目的:了解重性精神疾病基因组DNA拷贝数变异情况,探讨拷贝数变异位点与重性精神疾病的关系方法:研究样本为美国Stanley医学中心所提供的重性精神疾病和对照组的脑组织标本,样本包括50个精神分裂症病人,49个双相情感障碍病人,15个重性抑郁和49个对照,常规方法提取DNA。采用Affymetrix 5.0 SNP基因芯片对所有样本进行全基因组DNA拷贝数变异检测。采用Partek软件包对杂交信号进行处理。首先利用对照组样本生成拷贝数变异的参考基线,将所有样本的基因芯片杂交信号根据拷贝数变异参考基线进行标准化,然后采用隐性马尔可夫模型(HMM)检测基因组中的拷贝数变异位点,并规定每个拷贝数变异位点必须包括3个以上连续探针。同时对LOH和等位基因信号比率进行检测,作为拷贝数变异位点的补充信息。结果:在163个样本中共检测到2524个CNVs(1246位点),其中常染色体2502个(1226位点),X染色体22个(20位点),人均CNVs15.5个。2524个CNVs中,1148个为杂合拷贝缺失,562个为纯合拷贝缺失,814个为拷贝增加。CNVs长度介于89bp-29Mb之间,长度中位数为31K,在9号染色体处发现3个超长CNVs,长度接近30MBps,其中2个位于精神分裂症病人,一个位于双相情感障碍病人。所有CNVs中906个为孤本(singleton)CNVs,1618个为非孤本(non-singleton)CNVs。与已知CNVs数据库比较,156个CNVs为新CNVs。对频率＞5%的43个常见CNVs与疾病进行关联分析,发现CNV-304与精神分裂症存在临界相关(simulation P=0.002,Bonferroni P=0.086)。2524个CNV中有1087个位点对已知基因产生了影响,其中35个基因为病例组所特有,这些病例组特有的基因主要与细胞生长和分化、免疫反应、癌症和生殖疾病相关。结论:在精神分裂症病人和双相情感障碍病人的9号染色体区域检测到3个超长CNVs;通过关联分析,我们发现CNV-304与精神分裂症存在临界关联;通过对受CNVs影响基因进行分析,发现35个基因仅在病例组中存在,这些基因可能为重性精神疾病的重要候选基因。此小样本分析中的结果暂不能结论CNV与重性精神疾病的明确相关,得到的候选区域和基因有待大样本研究证实。更多还原

【Abstract】 PART-ⅠMRI study on brain neuronetwork connections in early onset schizophreniaObject The aim of this study was to explore the brain default network function connection and white matter integrity in first-episode early onset schizophrenia(EOS),by blood oxygenation level dependent functional magnetic resonance imaging(BOLD-fMRI) and diffusion tensor imaging(DTI) techniques.Methods First-episode adolescences with EOS(early onset schizophrenia) according to the diagnosis criteria of CCMD-Ⅲand DSM-Ⅳ) and age and gender-matched healthy controls were tested in a resting-state fMRI scan and diffusion tensor imaging scan.Functional connectivity analysis was used to reconstruct the default network in patients and healthy volunteers.Differences were examined between two groups.Functional connectivity analysis also was evaluated between the default network and other regions of the brains.Fractional anisotropy(FA) was measured in EOS and healthy volunteers with an automated voxel-based method of analysis.Results Forty-one first-episode patients and forty-two controls finished MRI scan.After excluding data from subjects with head movement,26 EOS(13 females and 13 males) and 28 age and gender-matched healthy volunteers were remained for fMRI analysis;35 EOS(17females and 18 males) and 33 age and gender-matched healthy volunteers were analyzed in DTI analysis.1.Default network analysis: the default network included the posterior cingulate/precuneus,left and right Superior Frontal gyrus,Caudate,the dorsal and ventral medial prefrontal cortex(MPFC),left and right angular gyrus,left and right Inferior Temporal Gyrus,left Parahippocampa Gyrus,right Inferior Semi-Lunar Lobule,and Cerebellar Tonsil.By comparing the functional connections within the default network between groups,we found five significantly abnormal connections comparing the patients group with controls.Three increased connections in patients are the connections between:ventral MPFC and right Inferior Temporal Gyrus (p=0.0078),ventral MPFC and left angular gyrus(p=0.0091),ventral MPFC and dorsal MPFC(p=0.0163).Two decreased connections in patients are the connections between:right angular gyrus and right Inferior Semi-Lunar Lobule(p=0.0223),left Superior Frontal gyrus and Cerebellar Tonsil(p=0.0294)。Comparing the connections of the default network and the other regions of the whole brain in two groups,the bilateral orbital middle frontal gyrus showed increased positive correlation with the default network in patients.2.DTI analysis: Comparing with controls,EOS patients exhibited significantly decreased fractional anisotropy(FA) values in right caudate,right anterior cingulate, right middle frontal gyrus,left superior frontal gyrus,left inferior Parietal lobe and left thalamus/hippocampus.There are no regions showing significantly increased FA in EOS patients.Conclusions 1.To my knowledge,this is the first study reporting that functional disintegration of the default network in first-episode EOS.Within the default network,the major abnormal connections were correlated to MPFC,and showed increased positive connections in EOS patients.The bilateral orbital inferior frontal gyrus showed increased positive correlation with the default network in patients. The abnormalities could be the source of the abnormal of introspectively oriented mental actives.2.EOS showed decreased fractional anisotropy in prefrontal and sub cortical regions suggested that widespread structural dysconnectivity,including the subcortical region,is already present in their first episode of illness. PARTⅡDNA copy number variations in major psychosisObject The aim of this study was to identify copy number variations(CNVs) in schizophrenia,bipolar disorder and major depression,and to test correlations between CNVs and these diseases.Methods Postmortem brain tissue from individuals with DSM-IV diagnosed SZ(n = 50),BD(n = 49),major depression(15) and matched controls(n = 49),provided by the Stanley Medical Research Institute Brain-Array and Consortium Collections.DNA was extracted from each specimen and purified.The Affymetrix 5.0 SNP array was hybridized to the genomic DNA sample of each subject.Assays were carried out according to the manufacturer’s protocol.Partek software was used to obtain estimates of copy number variations(CNVs).First,copy number baseline was created from pools of all control samples,and the intensity of each sample’s data was normalized to the copy number baseline.Then, CNVs were detected using a Hidden Markov Model,with a region defined as consistent variation in at least three adjacent probes.This finds the most likely state at each genomic locus by assigning a hidden state at each locus based on the observed data and the neighboring states. Secondary analyses of the genomic DNA hybridization data were performed,on allele signal ratio and loss of heterozygosity(LOH),which can give support information on CNVs.Results 2502 CNVs(from 1226 loci) were detected from autosomes. 22 CNVs(from 20 loci) were detected from X chromosome.The average number of CNVs detected per individual was 15.5.Of 2524 CNVs,1148 CNVs were heterozygous deletions,562 CNVs were homozygous deletions,and 814 CNVs were duplications.The median size of CNVs was 31kb,and the mean size was 189 kb.Three large CNVs on chromosome 9 were detected in schizophrenia and bipolar diseases. We compared the CNVs identified in our analysis to those present in the Database of Genomic Variants(DGV).156 CNVs(96 loci) in our data were novel CNVs.Of 2524 CNVs,906 CNVs are singleton CNVs,and 1618 CNVs were non-singleton CNVs.In tests for association for 43 common CNVs(frequency＞5%),CNV-304 has marginal significant P value(p=0.086) after Bonferroni correction.We further investigated genes that were deleted or duplicated exclusively in patients(ie,not observed in controls in our study,and not present in the Database of Genomic Variants).35 genes in CNVs were recorded exclusively in patients with major psychosis:6 in bipolar disorder,8 in major depression and 21 in schizophrenia.Conclusions 1.We found three very big CNVs in schizophrenia and bipolar diseases which suggested that these big CNVs may contribute genetic risk to these diseases.2.CNVs-304 is has marginal association with schizophrenia(Simulation P=0.002,Bonferroni P=0.086).3.There are 35 genes exclusively in major psychosis,which suggested these genes are valuable candidate genes for further study.更多还原