【作者】 费琦；

【导师】 邱贵兴；

【作者基本信息】 中国协和医科大学 ， 骨外科学， 2008， 博士

【摘要】 研究背景:先天性脊柱侧凸(congenital scoliosis,CS)定义为“由于胚胎期脊柱椎体发育异常而引起的脊柱侧凸”,它在脊柱的冠状面、矢状面和水平面上可引起侧凸、前凸、后凸和旋转畸形。CS是由于胚胎期中轴骨发育障碍引起的,而胚胎发育过程中一种重要的过渡性组织——“体节(somite)”决定中轴骨的形成,已证明体节发育受影响可导致脊椎畸形,包括椎体形成障碍(半椎体、楔形椎和蝴蝶椎)和椎体分节不良。椎体畸形可因发育相关基因突变所致,也可能是妊娠中环境因素或这两个因素共同作用所致,其确切病因还不清楚。近年来,随着分子生物学和遗传学水平的发展,国外已开始了CS候选基因的初步研究,CS的遗传病因学假说开始引起越来越多学者的重视。目前的研究支持CS是一种多基因遗传疾病,一个或多个基因导致不同的表型,任何一个基因均可能使患者处于易感状态,环境因素的影响使基因表型趋于复杂化。近20年通过对小鼠的胚胎学研究获得了大量关于脊椎发育的分子胚胎学信息,研究进展表明相对于物种进化来说不同物种间在核苷酸序列和氨基酸序列上有较高的保守性,因此我们可以系统回顾有关小鼠骨骼、尾部或神经肌肉突变表型的基因组数据以发现人类CS的可能候选基因,通过基因图的位置以同线性保守基因序列为基础进一步确立人类的可能候选基因。Giampietro等2003年报道利用人-小鼠同线性分析发现了小鼠的27个突变基因座及对应的人类CS的候选基因,还推荐了7个重要的候选基因——PAX1、DLL3、WNT3A、MYLK、LMX1A、FBN2和SIM2。从小鼠基因组数据库中我们还可发现TBX6、DVL2等基因也是人类CS的候选基因。目前对人类先天性脊柱侧凸(CS)的候选基因研究尚处于初步阶段,国内尚是空白。研究目的:■通过对7个候选基因(PAX1、SIM2、WNT3A、TBX6、DVL2、LMX1A、DLL3)上的SNPs的筛查,探索候选基因与人类CS之间的关联,依据研究结果提出可能的病因学假说■通过对候选基因SNPs基因型与CS临床表型的关联分析,探索可能潜在的致病基因,从遗传学上丰富CS临床表型的内容■从基因水平探索候选基因与CS及CS不同临床表型之间的关系研究方法:本研究采用医院为基础的病例-对照研究设计。■研究对象:根据入选和排除标准入选2005年10月——2007年9月期间北京协和医院骨科收治并已确诊的127例中国汉族先天性脊柱侧凸(CS)患者。127例非先天性脊柱侧凸对照为性别、年龄、民族与病例相匹配,年龄差别≤3岁的同期北京协和医院收住的感染(43%)、炎症性疾病(41%)、创伤(11%)、过敏性疾病及其他(5%)患者。■研究方法:(?)QIAamp DNA Blood Mini Kit试剂盒提取全血DNA;(?)根锯国际人类基因组单体型图计划(http://www.Hapmap.org)提供的基因型数据,应用Haploview 4.0软件选取各候选基因的标签SNPs,优先选取杂合度高于10%,位于外显子区域或有错义突变的SNPs和位于3′-、5′-调控区的SNPs作为遗传标记,共选取24个SNPs;(?)将病例组根据椎体畸形特点、畸形部位、椎体畸形受累程度、有无合并肋骨畸形和有无合并椎管内畸形进一步分为不同的临床表型;(?)在所有127例CS患者和127例对照中应用超高通量的SNP分型系统——SNPstream UHT Genotyping System对所选SNPs进行基因型鉴定;(?)拟合优度x~2检验(Goodness-of-fit Chi-square test)分析病例、对照组基因型频率的分布是否符合H-W平衡;(?)基于基因型/等位基因频率的关联分析;(?)在基因型/等位基因频率关联分析基础上,应用在线软件(SNPstats)进行非条件Logistic回归模型评估单位点基因型与CS发生风险的相关程度,并计算优势比(OR)及95%可信区间(CIs);(?)连锁不平衡和单倍体型分析:应用Haploview 4.0和UNPAHSED软件计算对照组人群中同一基因SNPs两两间连锁不平衡统计值D’和r~2,并评估所有单倍体型在病例组和对照组之间的分布差异。研究结果:■先天性脊柱侧凸(CS)病例组和对照组年龄、性别分布无统计学差异。■因DLL3基因的3个标签SNPs引物设计失败,故研究中放弃该基因;最终对筛查所得到的分布于6个基因(PAX1,SIM2、WNT3A、TBX6、DVL2、LMX1A)的20个SNPs数据进行统计分析,20个SNP位点均具有多态性。■PAX1基因筛选两个位点:SNP1(rs17861031)和SNP2(rs6047590),其基因型分布在病例组和对照组中均符合Hardy-Weinberg平衡;两位点间不存在连锁不平衡;两位点的基因型/等位基因与CS的发生风险之间不存在相关性。■SIM2基因共筛选3个位点:SNP3(rs2073601)、SNP4(2073417)和SNP5(2051397),其中SNP4基因型分布在病例组和对照组中不符合Hardy-Weinberg平衡,另两个位点均符合Hardy-Weinberg遗传平衡;三位点两两间不存在连锁不平衡;三个位点的基因型/等位基因与CS的发生风险之间不存在相关性。■WNT3A基因共筛选2个位点:SNP6(rs964941)和SNP7(rs752107),其基因型分布在病例组和对照组中均符合Hardy-Weinberg平衡:两位点间不存在连锁不平衡;两位点的基因型/等位基因与CS的发生风险之间不存在相关性。■TBX6基因共筛选2个位点:SNP8(rs2289292)和SNP9(rs3809624),其基因型分布在对照组中均符合Hardy-Weinberg平衡,但在病例组中不符合Hardy-Weinberg平衡;病例/对照组中等位基因分布频率分别为:SNP8C=108(0.43)/135(0.53),SNP8T=146(0.57)/119(0.47),SNP9C=144(0.57)/120(0.47),SNP9T=110(0.43)/134(0.53),单位点分析显示两位点等位基因在病例和对照组中的分布频率统计学上存在显著差异(P=0.017和P=0.033),进一步分析得知SNP8等位基因T、SNP9等位基因C与CS的易感性升高相关;病例/对照组中基因型分布频率统计学无显著差异;在非条件Logistic回归分析中,经年龄和性别校正后,SNP8位点基因型分布符合Dominant遗传模型(OR=0.56;95%CI=0.33-0.96,p=0.03,AIC=355.2)和Log—additive(OR=0.69;95%CI=0.50-0.97,p=0.029,AIC=354.9)遗传模型,后者最佳(AIC值较小);在多位点单倍体分析中,发现2种阳性单倍体:TBX6基因的SNP08T-SNP09C和SNP08C-SNP09T。■DVL2基因共筛选5个位点:SNP10(rs2074222)、SNP11(rs222837)、SNP12(rs222835)、SNP13(rs10671352)和SNP14(rs222836),其基因型分布在病例和对照组中均符合Hardy-Weinberg平衡;5个位点处于完全连锁不平衡状态(各位点两两间:D′=1;r~2＞0.5);5个位点的基因型/等位基因/单倍体型与CS的发生风险之间不存在相关性。■LMX1A基因共筛选6个位点:SNP15(rs1819768)、SNP16(rs12023709)、SNP17(rs16841013)、SNP18(rs4656435)和SNP19(rs4657412)和SNP20(rs4657411),其基因型分布在病例和对照组中均符合Hardy-Weinberg平衡;在基于基因型的“病例-对照”关联分析中,筛到阳性位点SNP15和SNP16,病例/对照组中基因型分布频率分别为:SNP15AA=7(0.16)/17(0.13)、SNP15AC=62(0.49)/45(0.35)、SNP15CC=58(0.46)/65(0.51),SNP16AA=58(0.46)/65(0.51)、SNP16AG=62(0.49)/45(0.35)、SNP16GG=7(0.16)/17(0.13),单位点分析显示两位点基因型在病例和对照组中的分布频率统计学上存在显著差异(P=0.026和P=0.026),进一步分析得知SNP15基因型A/C、SNP16基因型A/G与CS的易感性升高相关;在非条件Logistic回归分析中,经年龄和性别校正后,SNP15和SNP16两位点基因型分布均符合Ressessive(OR=0.38;95%CI=0.15-0.94,p=0.029,AIC=354.9)或Overdominant(OR=1.73;95%CI=1.05-2.8,p=0.032,AIC=355.1)遗传模型,但前者最佳(AIC值较小);SNP15、SNP16、SNP17和SNP20处于连锁不平衡状态,SNP18和SNP19也处于完全连锁不平衡状态,但单倍体型与CS的发生风险之间不存在相关性。■在各SNP位点基因型与临床表型Ⅰ(根据CS畸形特点进一步分层)的关联分析中,我们发现LMX1A基因的SNP15位点基因型AC型、SNP16位点基因型AG型、SNP17位点基因型CT型与有椎体形成障碍CS的易感性升高有关;TBX6基因的SNP8位点基因型TT型、SNP9位点基因型CC型与有分节不良CS的易感性升高有关。■在各SNP位点基因型与CS其他临床表型(性别、畸形部位、椎体受累程度、有无合并肋骨畸形和椎管内异常)的关联分析中,没有发现阳性位点。结论:■在中国汉族人群中TBX6和LMX1A基因遗传变异与先天性脊柱侧凸(CS)发生相关,有可能是决定CS个体遗传易感性的重要因素。■不同基因的遗传变异可能与先天性脊柱侧凸(CS)不同临床表型的发生相关,LMX1A基因可能与有椎体形成障碍CS的易感性升高有关;TBX6基因可能与有分节不良CS的易感性升高有关。■在中国汉族人群中PAX1、SIM2、WNT3A和DVL2基因可能不是引起先天性脊柱侧凸(CS)及其不同临床表型的主要因素,有待于进一步深入研究。更多还原

【Abstract】 BackgroundCongenital scoliosis(CS) is defined as a lateral curvature of the spine due to a developmental abnormality,which arise from defects in the development of the axial skeleton.During embryogenesis,the axial skeleton is formed by a process called somitogcnesis,which produces transient segments of tissue known as somites. Disruptions in somitogenesis have been shown to result in vertebral malformations, including uneven segments(hemivertebrae and wedge vertebrae),fused segments (block vertebrae),and problems in midline fusion(butterfly vertebrac).Vertebral defects can arise from the disruption of genes involved in development. environmental insults during gestation,or a combination of these two factors.As the development of genetics and molecular biology in recent years,several preliminary studies about candidate gene contributed to vertebral malformation in human were reported and the genetic etiology hypothesis of CS has caused more and more interests.More evidences suggest that CS may be multigenetic disease.The interaction of environmental factors and the genes that play a role in regulation of somite segmentation,is thought to be disrupted in congenital vertebral deformities such as congenital scoliosis.A multigenic mechanism could be responsible for the type and spacial identity associated with a particular type of congenital scoliosis.Within the past two decades a great deal of information has been learned about the molecular embryology of spine development through the study of mouse and chick embryos.Human and mouse genes share high degree of homology.due the relatively recent evolutionary divergence of these mammals.By studying what happens to the axial skeleton when these genetic instructions are disrupted in the mouse,we have learned a great deal about the most important genes in spinal development.To date,some abnormalities of genes involved in mouse somitogenesis have been found to cause human spinal deformities.Synteny conservation was used as the basis to identify potential human candidate genes by map position.Giampietro therefore undertook a systematic review of mouse mutations with skeletal,tail.or neuromuscular phenotypes and identified 27 potential human candidate genes sites for CS in 2003.Seven of these genes including PAXI、DLL3、WNT3A、MYLK、LMXIA、FBN2 and SIM2 are briefly discussed.Through latest mouse genome database,TBX6 and DVL2 genes were also considered as candidate genes contributed to CS.At present,the etiology of congenital scoliosis is still unknown,no molecular analyses and few clinical genetic studies have been reported.ObjectsTo identify the relationship between candidate genes(PAXI、SIM2、WNT3A、TBX6、DVL2、LMXIA、DLL3)associated with somite developmental pathway and CS with SNPs in Chinese Han population.■To identify the relationship between the genotypes of SNPs and the clinical phenotypes of CS.■To explore possible etiologic hypothesis contributed to the development of CS and the different clinical phenotypes of CS from gene level.Methods■A hospital-based case-control design was applied in this study.■A total of 127 patients(55 boys,72 girls,mean age 12.90 y/o)diagnosed with clinically confirmed congenital scoliosis admitted in Peking Union Medical College(PUMC) Hospital were enrolled in this study according the inclusion and exclusion criterias between October 2005 and September 2007.The diagnosis of scoliosis-free control subjects(55 boys,72 girls, mean age 13.27y/o) consisted of infection(43%)、inflammatory disease (41%)、trauma(11%)、hypersensitivity disease and others(5%) at the same hospital during the same study period.All the control subjects were frequency-matched to the cases on age(±3 years),gender,and nation (Chinese Han).■Genomic DNA was extracted from peripheral blood leukocytes of each subject who had signed informed consent,using QIAamp DNA Blood Mini Kit.■Based on genotype data from the International HapMap project (http://www.Hapmap.org),the tagging single nucleotide polymorphisms (tSNPs) initially were selected using Haploview 4.0 software.The MAF(minor allele frequency) of all selected SNPs were above 10%.The missense or nonsense mutation SNPs in exons,or the SNPs located in promoters,5’UTR or 3’UTR were prefered.■Hardy-Weinberg equilibrium both in control and in case groups were analyzed through Goodness-of fit Chi-square test.■Case group were classified into different clinical phenotypes according to vertebral defect type、the location of deformity、the extent of developmental disruption、combined rib malformations and neural canal deformity.■Genotying of all selected SNPs was done by SNPstream technology (Beckman Couher SNPstream) which is the combination of the fidelity polymerase mediate reaction,single base primer extension and microarray methods.As a high throughput genotyping system,SNPstream technology has been widely used in biochemistry and genetic research.■All the data of 20 SNPs with polymorphism are analyzed by the association analysis based on a single SNP,the association analysis between phenotypes and SNPs.■Odds radios(ORs) and 95%confidence intervals(CIs) were computed by the unconditional logistic regression to estimate the relative risk for the single locus genotypes using online software—SNPstats.■And pairwise linkage disequilibrium values D’ and r~2 were calculated in the control population using Haploview 4.0 software.Haplotype frequencies were estimated and difference in haplotype distributions between cases and controls were assessed using UNPAHSED software.Results■There were no significant difference of average age and sex proportion between case group and control group(p＞0.05).■We failured in designing premier of three tSNPs of DLL3 and discasded the gene in this study.Eventually 20 common SNPs of PAXI.SIM2,WNT3A, TBX6,DVL2,and LMXIA genes were chosen to be tagging SNPs and all SNPs have polymorphism.■SNP1(rs17861031) and SNP2(rs6047590) of PAXI gene were genotyped and both polymorphisms were in Hardy-Weinberg equilibrium both in control and in ease groups.Both polymorphisms were not in linkage disequilibrium.No association is obseiwed between SNP1 and SNP2 genotypes/allele polymorphisms and risk of CS.■Three SNPs including SNP3(rs2073601),SNP4(rs 2073417) and SNP5 (rs 2051307)of SIM2 gene were genotyped and both polymorphisms of SNP3 and SNP5 were in Hardy-Weinberg equilibrium both in control and in case groups.All polymorphisms were not in linkage disequilibrium.No association is observed between SNP3.SNP4 and SNP5 genotypes/allele polymorphisms and risk of CS.■Two SNPs including SNP6(rs964941) and SNP7(rs752107) of WNT3A gent were genotyped and both polymorphisms were in Hardy-Weinberg equilibrium both in control and in case groups.Both polymorphisms were not in linkage disequilibrium.No association is observed between SNP6 and SNP7 genotypes/allele polymorphisms and risk of CS.■Two SNPs including SNP8(rs2289292) and SNP9(rs3809624) of TBX6 gene were genotyped.Both polymorphisms were in Hardy-Weinberg equilibrium both in control group,but not in case group.The minor allele frequencies in cases/controls were,respectively,as follows:SNP8C=108 (0.43)/135(0.53).SNP8T=146(0.57) /119(0.47),SNP9C=144 0.57)/120(0.47).SNP9T=110(0.43) /134(0.53).The single locus analysis revealed the allele frequency distributions of SNP8 and SNP9 were statistically significantly different between case patients and control subjects(P=0.017 and P=0.033).In the unconditional logistic regression analysis,after adjustment for age and gender,SNP8 showed significant difference in both dominant(OR=0.56:95%CI=0.33-0.96) and log-additive (OR=0.69;95%CI=0.50-0.97) models,and the P-values were 0.03 and 0.029,respectively.The log-additive model was accepted as the best inheritance model because of the smaller AIC(Akaike information) value (354.9).Both polymorphisms were in full linkage disequilibrium.In the haplotype analysis,we got 2 positive haplotypes:SNP08T-SNP09C and SNP08C-SNP09T■Five SNPs including SNP10(rs2074222),SNP11(rs222837)、SNP12 (rs222835)、SNP13(rs10671352) and SNP14(rs222836) of DVL2 gene were selected to be tagging SNPs.All polymorphisms were in Hardy-Weinberg equilibrium both in control and in case groups.All polymorphisms were in strong linkage disequilibrium(D’=1;r~2＞0.5).No association is observed between SNP10.SNP11,SNP12.SNP13 and SNP14 genotypes/haplotype polymorphisms and risk of CS.■Six SNPs including SNP15(rs1819768),SNP16(rs12023709),SNP17 rs16841013),SNP18(rs4656435),SNP19(rs4657412) and SNP20 rs4657411) of LMXIA gene were selected to be tagging SNPs.All polymorphisms were in Hardy-Weinberg equilibrium both in control and in case groups.In the association analysis between the genotypes of SNPs and CS,we got two positive SNPs:SNP15 and SNP16.The minor genotype frequencies in cases/controls were,respectively,as follows:SNP15AA=7 (0.16)/17(0.13).SNP15AC=62(0.49) /45(0.35),SNP15CC= 58(0.46)/65(0.51),SNP16AA=58(0.46)/65(0.51).SNP16AG =62(0.49)/45(0.35),SNP16GG=7(0.16)/17(0.13).The single locus analysis revealed the genotype distributions of SNP15 and SNP16 were statistically significantly different between case patients and control subjects(P=0.026 and P=0.026).In the unconditional logistic regression analysis,after adjustment for age and gender,SNP15 and SNP16 both showed significant difference in both Ressessive model(OR=0.38; 95%CI=0.15-0.94) and Overdominant model(OR=1.73;95%Cl=1.05-2.8), and the P-values were 0.029 and 0.032.respectively.The Resscssive model was accepted as the best inheritance model because of the smaller AIC (Akaike information) value(354.9).SNP15,SNP16;SNP17 and SNP20 were in strong linkage disequilibrium.Both SNP18 and SNP19 were also in strong linkage disequilibrium.But no association is observed between all haplotypes polymorphisms and risk of CS.■In the association analysis between the genotyhpes and the phenotype l(according to vertebral defect type) of CS,we got 5 positive SNPs.SNP15, SNP16 and SNP17 polymorphisms of LMXIA gene may related to CS with failure of formation.SNP8 and SNP9 polymorphisms of TBX6 gene may related to CS with a fallure of segmentation.■In the association analysis between the genotyhpes and other phenotypes of CS,we didn’t have got any positive SNPs.Conclusion■Genetie variants of TBX6 and LMXIA genes are associated with CS and may play an important role in mediating susceptibility to developing CS in a Chinese Han population.■Genetic variants of different genes may play an important role in mediating susceptibility to different clinical phenotypes of CS in a Chinese Han population.It was suggested that genetic variants of LMXIA gene may related to CS with failure of formation,and genetic variants of TBX6 gone may related to CS with a failure of segmentation.■It was suggested genetic variants of PAX1,SIM2,WNT3A and DVL2 genes may not be associated with the susceptibility to CS and different clinical phenotypes of CS in a Chinese Han population.Further studies are warranted before definitive conclusion can be drawn.更多还原