【作者】 姚琳；

【导师】 惠宁；

【作者基本信息】 第二军医大学 ， 妇产科学（专业学位）， 2013， 博士

【摘要】 研究背景神经管畸形（neural tube defects， NTDs）是最常见、最严重的出生缺陷之一。NTDs的发生是由于囊胚期神经管闭合异常所导致的神经管开放。在世界范围内，其发生率仅次于先天性心脏病，而我国的发生率为全世界最高。除致死性NTDs外，轻型闭合型NTDs不仅将给患儿带来伴随终生的先天残疾及严重的心理缺陷，也会给患儿家庭及社会带来沉重的经济负担。随着孕期叶酸（folic acid，FA）补充的推广，NTDs的流行趋势明显下降，但大规模的研究证实使用FA补充剂后仍有30-50%的NTDs无法预防。另外过量补充FA还可能引起包括新生儿胰岛素抵抗或肥胖在内的一系列非治疗作用，因此FA补充是否应当成为强制饮食添加剂，何种剂量为适宜补充剂量，在国际上引起了广泛的争议。除叶酸缺乏外，一线抗癫痫药物丙戊酸（Valproic acid，VPA）的应用也是NTDs的诱因之一。对小鼠模型的研究提示，VPA可能通过FA代谢途径导致NTDs，并且给予FA补充可能在一定程度上起到保护作用，从而避免NTDs发生。反对使用VPA的学者强调其致畸性对胎儿的严重影响，但是对于部分常见癫痫类型VPA仍然是无法替代的最佳选择。因此很难在妊娠前期或妊娠期更换其他药物，这使VPA在妊娠期的使用陷入进退两难的境地。神经管闭合这一过程始于胚胎第20天，且是贯续发生的发育过程，这使以往的观察研究遇到了很大的困难。同时由于研究对象的特殊性，特别是人类研究存在的伦理问题，使我们无法直接对人类胚胎进行在体实验。此外由于胚胎发育的不可逆性，我们亦无法获知神经管畸形的发生的真实过程，只能通过回顾性分析获得间接的证据。这就使动物模型中获得的信息无法在人类中获得验证。因此如何建立一个人类的NTDs体外研究模型，成为亟待解决的问题。本研究探索利用胚胎干细胞（ES cell）和体细胞来源的诱导多能性干细胞（iPScell）,结合体外神经系定向诱导分化技术，建立人类干细胞来源神经管研究模型。尝试通过不同组别的药物作用实验，了解FA及VPA对神经管形成及神经系发育的作用。同时在此基础上进行药物浓度筛选的初步应用研究，为在人类系统中验证动物模型中所获理论创造条件。第一部分人类多能干细胞来源神经管体外模型建立方法：对人类多能干细胞系H1；H9；iPS4进行培养传代。同时进行神经系分化：去除bFGF，悬浮培养细胞，诱导EB形成；利用神经上皮干细胞（NSC）培养基对EB进行贴壁培养，诱导玫瑰花环样神经管结构（RS）形成,利用NSC标记物进行免疫荧光染色鉴定；利用定向诱导分化技术进行进一步分化：缺省分化大脑皮质大椎体细胞；腹侧尾侧化脊髓前脚运动神经元；神经胶质细胞系定向诱导分化；神经脊系细胞定向诱导分化；对分化获得细胞进行特异性标记物免疫荧光染色鉴定。结果：人类多能干细胞系体外培养传代后仍表达多能转录因子Oct4, Nanog；H1；H9；iPS4均能形成EB；三系均能在诱导条件下形成RS结构，且Pax6，Nestin阳性；定向诱导分化：获得前脑皮质大椎体细胞，Pax6阳性；获得人γ-氨基丁酸能神经元，GABA阳性；获得多巴胺能神经元，TH阳性；获得脊髓前脚运动神经元,ChAT阳性；获得星形胶质细胞，GFAP，S100阳性；获得神经脊细胞，P75，Sox10，SMA阳性。结论：利用三个人类多能干细胞系，通过定向诱导分化，我们首次提出建立人类干细胞来源神经管样结构模型。此模型可以在发育时间，细胞构成，分化潜能上模拟在体神经管的发育过程。它的建立为后续药物浓度筛选提供了条件，为NTDs发生及预防机制研究奠定了基础。第二部分叶酸缺乏及补充对神经管结构形成及神经发育的影响方法：利用第一部分中建立的人类多能干细胞体外神经管分化模型，选择iPS4作为研究对象，使用无FA的PRIM1640培养基作为空白对照组，根据FA梯度分组：0.02uM组，0.2uM组，2uM组，20uM组，80uM组，160uM组。在无FA的PRIM1640培养基中加入相应浓度FA，对iPS4细胞进行分化。分化过程中比较不同FA分组的：EB形成情况；EB中多能标记物Oct4，Nanog和Sox2的mRNA及蛋白质表达水平；RS形成情况；RS中NSCs标记物Pax6及Nestin的mRNA及蛋白质表达水平；远期神经分化率差别；神经胶质细胞分化率差别；神经脊细胞分化率差别。结果：FA缺乏组（0-2uM）：人类诱导多能干细胞分化EB形成率降低（P＜0.05），多能因子表达量下降；RS形成明显减少（P＜0.05）,结构紊乱；NSCs标记物Pax6，Nestin的mRNA，蛋白质表达下调；远期神经元，神经脊分化率呈剂量依赖性降低。FA补充组（20-160uM）：EB形成率升高，多能因子表达量与iPS细胞相似，RS形成数量增多（P＜0.05），神经系标记物表达量上调，远期神经元及神经脊系分化率上升。但在FA补充组中，组间差异均无统计意义。FA缺乏组及FA补充组间神经胶质细胞分化率无显著差别。结论：在人类胚胎发育过程中，FA缺乏或缺失将影响早期神经发育。FA缺乏或缺失将抑制NSCs分化，引起RS形成障碍，进而影响远期神经元及神经脊系的分化，最终导致一系列先天缺陷及综合征。对于FA代谢正常围孕期女性，补充FA使其维持在正常或稍高水平（20uM-80uM，即常规富含FA饮食或400ug/day补充剂量）将有利于神经分化及神经管的形成，而无需给予更高剂量FA（160uM或600ug/day），带来不必要的非治疗作用。第三部分VPA及叶酸对神经系发育的影响及其相互作用方法：利用第一部分中建立的人类多能干细胞体外神经管分化模型，选择iPS4作为研究对象，使用DMEM/F12+N2培养基作为空白对照组，实验分组：VPA组（1mM VPA），VPA+FA组（1mM VPA+160uM FA）。对不同处理组别的iPS细胞进行神经分化，分化过程中比较不同分组的： RS形成情况；RS中NSCs标记物的mRNA及蛋白质表达水平；NSCs殖及凋亡情况；NSCs细胞周期；FA受体Forl在mRNA及蛋白质水平的表达情况；及远期神经分化率差别。结果：与对照组相比，VPA处理组：RS形成明显减少（P＜0.01）,结构紊乱；NSCs标记物Pax6，Nestin的mRNA，蛋白质表达下调（P＜0.001）；NSCs增殖减少(41.01±3.53%VS.80.09±4.1%，P＜0.01）；凋亡显著增多（21.38%VS.3.71%，P＜0.01），凋亡早期细胞亦增高（29.64%VS.9.76%，P＜0.05）；处于G2-S期细胞比例下降（34.75%VS.63.67%，P＜0.05），G0-G1期细胞增多（65.25%vs.36.33%，P＜0.05）；FA受体Forl mRNA及蛋白质表达量下调；远期神经元分化率降低。VPA+FA组：RS形成较VPA单独处理组增多,但较对照组仍减少（P＜0.05），但RS结构较对照无明显差别；NSCs标记物Pax6，Nestin mRNA，蛋白质表达较VPA组明显升高（P＜0.01），与对照组差异无统计学意义；NSCs增殖较VPA单独处理组高，但仍低于对照（58.0±4.67%VS.80.09±4.1%，P＜0.05）；与VPA单药组相比，凋亡稍降低，但差别不具有统计学意义，与对照组相比，细胞死亡多（15.46%VS.3.71%，P＜0.01），但凋亡早期细胞比例无明显差别（9.76%VS.9.86%，P＞0.05）；与VPA单独处理组相比，处于G2-S期细胞比例增多(43.35%VS.34.75%)，G0/G1期的细胞比例减少（56.65%VS.65.25%），但与对照组相比，G0-G1期细胞仍高（56.651%VS.36.34%，P＜0.05）；FA受体Forl的mRNA及蛋白质表达量下调；与对照相比，远期神经元分化率仍降低。结论：结果证实治疗剂量的VPA抑制人类多能干细胞的体外神经分化。其通过抑制细胞增殖，促进细胞凋亡，阻碍FA代谢，破坏正常神经管样结构的形成，导致NTDs的发生，同时影响远期神经分化。补充FA实验组证明FA在其他致畸诱因的存在下，可以在一定程度上保护早期神经胚发育，弥补细胞增殖，抑制细胞凋亡，促进FA代谢，但与对照相比仍存在一定异常。这为解决临床治疗存在的矛盾提供了新的研究平台，在药物筛选中有着一定的应用前景。创新点与小结本课题首次将人类多能干细胞体外神经管分化模型应用于出生缺陷—NTDs的研究。在建立人类干细胞来源神经管体外模型的基础上，进行了FA浓度筛选及VPA处理等药物筛选的初步应用研究。首次在人类分化发育系统中验证了动物模型中所获得的理论；为后续制备患者个体化神经管分化模型，研究NTDs发病机制，致病基因筛选，优化预防配方奠定了基础；在提高人口质量，最终避免NTDs发生的道路上，迈出了新的一步。更多还原

【Abstract】 IntroductionNeural tube defects (neural tube defects, NTDs) are the most common and severe birthdefects. It is due to the abnormal in neural tube closure at blastocyst stage. The incidenceof NTDs is second only to congenital heart disease worldwide. Moreover, the rate ofChina is the highest in the world. Except for fatal NTDs, closed NTDs will not only causelifelong congenital disabilities and severe mental defects of infected children, but also puta heavy financial burden on their families and society.With the folic acid (FA) supplements during pregnancy, the prevalence of NTDs isremarkably decreased. However, large-scale studies showed that there is still30-50%ofNTDs which can not be prevented by FA supplements. Excessive complement FA alsomay cause a series of non-treatment effects including neonatal insulin resistance orobesity. These initiated an extensive controversy in the field including the role of the FAsupplement in NTDs prevention, whether it should become mandatory food additives andwhat dose is appropriate in pregnancy.In addition to folic acid deficiency, the application of the first-line antiepileptic drugValproic acid (VPA) is one of other participating factors. A mouse model study showedthat the VPA might lead to NTDs through the FA pathway and FA supplement may playa protective role in NTDs prevention caused by VPA. Scholars against VPA emphasizeits teratogenic effects on fetus, but there is no other better choice besides VPA for somecommon type of epilepsy. And this makes it difficult to change VPA to others duringpre-pregnancy or pregnancy.Neural tube closure occurs from the20th day after fertilization, and is a consistentlydeveloping process. This makes the previous studies encounter a lot of difficulties inobservation. As a special study object, especially considering the ethical issues of humanresearch, direct experiments on human embryos in vivo is impossible. In addition, due tothe irreversibility of development, we can not gain the information from the real processof NTDs in human. The only evidence can be obtained is through retrospective analysisindirectly. This makes the information form animal models hard to be verified in human.Therefore, building a human NTDs vitro model is very important for future study. Our research explores the use of embryonic stem cells (ES) and induced pluripotent stemcells (iPS) from somatic cells. Combined with in vitro induced differentiation of neurallinage, we setup human pluripotent stem cell-derived neural tube models. To investigatethe role of the FA and VPA during neural tube formation and nervous systemdevelopment, we applied different dug groups in our neural tube model developmentsystem. Meanwhile based on these studies, we want to initiate some application researchon drug screening, also to validate theory from animal models systems in human.Section one: Human pluripotent stem cell-derived in vitro neural tube modelMethods: Human pluripotent stem cell lines, H1, H9and iPS4were cultured andpassaged. Induced neural linage differentiation: Removed bFGF in suspension cultureand induced EB formation; EB were adherently cultured in NSC medium to induceRosette like structure (RS) formation; Immunofluorescence staining analysis of NSCmarkers; Further directly induced the NSCs ifferentiate into different cell types: defaultdifferentiation to large pyramidal cells of the cerebral cortex; ventral and caudalizeddifferentiation to motor neuron of the spinal cord; induced differentiation to glial cells;induced differentiation to neural crest cells; Immunofluorescence staining analysis of celltype-specific markers.Results: Human pluripotent stem cell lines cultured and passaged in vitro still expressedOct4, and Nanog; they are able to form EB; all three Lines can form the RS structureunder inducing conditions. And the cells in RS were Pax6, Nestin positive; Furtherinduced differentiation: large pyramidal cells of the cerebral cortex is Pax6positive;γ-aminobutyric acid neurons is GABA positive; Dopaminergic neurons is TH positive;Motor neurons of the spinal cord is ChAT positive; Astrocytes is GFAP and S100positive; Neural crest cells is P75, Sox10and SMA positive.Conclusion: Through induced differentiation of three human pluripotent stem cell lines,we first proposed the establishment of a human stem cell-derived neural tube-likestructure model in vitro. This model simulates the development of the neural tube in vivo,in the aspects of developmental time, cellularity and differentiation potential. It providesa novel potential in vitro system for drug screening and mechanism research of NTDs.Section two: the influence of Folate deficiency and supplement on neural tube formation and neurodevelopmentMethods: Based on the human pluripotent stem cell-derived RS model in section one, weselect iPS4as the object of study. Using the PRIM1640/woFA medium as a blank controlgroup, we grouped the experiment according to the FA concentration:0.02uM,0.2uM,2uM,20uM,80uM,160uM. iPS4cells were differentiated in PRIM1640/woFA mediumadding corresponding concentration of FA. Compare the following aspects among groups:EB formation; the mRNA and protein expression levels of pluripotent markers as Oct4and Nanog; RS formation; the mRNA and protein expression levels of NSC markers inRS as Pax6and Nestin; The differentiation rate in neurons, glial cells and neural crestcells.Results: Lack of FA groups (0-2μM): EB formation rate is lower (P <0.05), and theexpression levels of pluripotent factors decreased; RS formation was significantly lessthan high dosage groups (P <0.05); The mRNA and protein expression levels of Pax6,and Nestin were down regulated; Differentiation rates of neuron and neural crest reducedin a dose-dependent way. FA supplemented groups (20-160uM): EB formation ratesincreased, pluripotent factors expression levels were similar with iPS cells; RS formationincreased (P <0.05), the expression of markers of NSCs unregulated and thedifferentiation rates of neuron and neural crest increased. There is no statisticalsignificance among FA supplemented groups. The glial cell differentiation had nosignificant difference among all groups.Conclusion: During human embryonic development, the absence or lack of FA willaffect early neurodevelopment. FA deficiency inhibited the NSC differentiation. Theresulting RS formation inhibition will affect further differentiation of neuron and neuralcrest lineages. This then leads to a series of birth defects and syndrome. For thepreconceptional women with normal FA metabolism, keeping the FA concentrationaround normal or slightly higher levels (20uM-80uM in serum, by taking FA rich diet or400ug/day supplement dose) will be conducive to neuronal differentiation and theformation of the neural tube. There is no need to have an even higher dose of FA (160uMor600ug/day) which will bring unnecessary non-therapeutic side effects.Section three: Influence of VPA and folic acid on the development of the nervoussystem and their interactionsMethods: Select iPS4as the object of study in Section three, using DMEM/F12+N2 medium as a blank control group. And the experimental groups were divided into: VPAtreatment group (1mM of VPA) and VPA+FA group (1uM VPA+160mM FA).Neuronal induced differentiation was preformed in different groups. Compare the resultsin: RS formation; the mRNA and protein expression levels of NSC markers; Analysis ofNSC proliferation and apoptosis; Cell cycle Analysis of NSCs; The mRNA and proteinexpression levels of the FA receptor Forl expression; Long-term neuronal differentiationrate.Results: VPA group: RS formation reduced significantly (P <0.01); Pax6and Nestinexpression lowered (P<0.001) in mRNA and protein; Proliferation of NSCs deseased(41.01±3.53%, P <0.01); Apoptosis increased significantly (21.18%vs.3.57%, P <0.01);The proportion of cells in G2-S phase declined (35.69%VS.63.37%, P <0.05) whileG0-G1phase increased (64.31%vs.36.63%, P <0.05); Forl expression lowered;Long-term neuron differentiation rate reduced. VPA+FA group: Compared with VPAgroup, RS formation increased; Pax6and Nestin expression are significantly higher (P<0.01); NSC proliferation and cells in G2-S (43.69%VS.35.69%) increased(58.0±4.67%, P <0.05); Compared with the control group, cell death (15.71%vs.3.57%, P<0.01) increased; percentage of G0-G1phase is still high (56.31%vs36.63%, P <0.05);Forl expression lowered; Long-term neuron differentiation rate still reduced.Conclusion: The results confirm that therapeutic dose of VPA will inhibit in vitro neuraldifferentiation of human pluripotent stem cells. It works through the inhibition of cellproliferation; activation of apoptosis; interference of FA metabolism and destruction ofnormal neural tube-like structure formation. In the presence of other teratogens, FAsupplementation plays some role in protecting early neural embryo development, incompensation for cell proliferation, inhibition of apoptosis. But there is still a certainabnormality compared with the control. This provides a new research platform to solvethe clinical contradiction and could be useful in drug screening.Summary and innovationFor the first time, our subject applied in vitro human pluripotent stem cells derived neuraltube model in research of human birth defects as NTDs. Based on the establishment ofthe neural tube model in vitro of the human source of stem cells, we preliminary appliedthe FA and VPA treatment screening. We firstly verified the theories from animal modelsin human differentiation and development system. This also laid the foundation for generation of patients’ individualized differentiation of neural tube model, study of NTDspathogenesis, screening of virulence gene and for optimizing prevention recipe. And ithas taken a new step on the road to improve the population quality, and to prevent NTDsultimately.更多还原