Genetic Analysis Of Morphogenetic Mechanism During Mouse Development

脊椎动物胚胎的发育过程至少部分是通过分泌生长因子等分子来调节的。骨形态发生蛋白(BMPs)是转化生长因子-β超家族成员之一，在小鼠发育过程中的功能是我们关注的焦点。为了揭示BMP信号转导的功能，我们用传统的基因打靶技术获得了BMP IA受体(BMPR1A或Alk3)和激活素IA受体(Alk2)缺失的突变小鼠。每个受体的缺失导致严重的胚胎死亡，突变的胚胎在胚胎7.5天(BMPR1A)或8.5天(Alk2)死亡。对于BMPR1A突变胚胎，我们发现1)没有形成中胚层，2)原肠形成前的细胞周期延长。对于Alk2突变的胚胎，我们发现1)中胚层已形成，但尚未完全分化，2)胚外区域(未来的胎盘)中的Alk2信号对原肠形成至关重要，3)Alk2突变的细胞不能贡献心脏或眼睛。这些结果表明，胚胎发育早期的BMP信号对细胞生长、原肠形成和心脏等特定器官的形成具有重要意义。 为了在发育的后期阶段对这些基因产物进行功能分析，我们引入了组织特异性基因打靶。利用这项技术，我们以骨骼特异的方式突变了BMPR1A(专用于分化的成骨细胞)。骨特异性BMPR1A基因缺陷小鼠存活，表明我们可以通过组织特异性基因打靶技术避免BMPR1A基因中断造成的胚胎死亡。突变小鼠比正常小鼠个头小，骨骼中的骨基质沉积较少。这些结果首次证明BMP信号是体内正常骨形成所必需的。有趣的是，衰老的突变小鼠表现出破骨细胞活性降低，导致骨基质增加。这表明BMP信号在分化的成骨细胞中的主要作用是年龄依赖性的。为了直接解决成骨细胞中BMP功能随年龄变化的问题，我们采用了他莫昔芬诱导的CRE系统。利用I型胶原启动子(Col1-Creer)在成骨细胞中特异性表达了一种Cre重组酶，该酶的活性可被他莫昔芬诱导。当给新生仔鼠服用他莫昔芬时，突变小鼠的骨骼密度(BMD)显示出比对照小鼠低。当给断奶小鼠在出生后1.5个月至5个月期间服用他莫昔芬时，突变小鼠的骨密度高于对照组。这些结果与我们之前的发现很好地一致，并强烈表明成骨细胞中BMP信号的年龄依赖性功能。在另一种方法中，我们以神经峰特异的方式突变了BMPR1A。胚胎在妊娠中期死亡，表现为神经脊细胞来源的组织如背根神经节和头面部的异常。这些胚胎还表现出心脏神经脊细胞迁移的流出道(OFT)的异常。初步结果显示，E11.5突变者的背主动脉出现反向血流，提示OFT患者的心垫功能异常。这是第一个证据表明OFT垫早在E11.5就起作用了，BMP信号是心脏神经脊细胞形成功能性OFT垫的关键因素之一。我们还在原肠胚阶段以外胚性特有的方式突变了BMPR1A。胚胎原肠发育后表现为神经组织过度生长。在突变胚胎中，体节等近轴中胚层扩张，但没有心脏发育的迹象。这些结果表明，BMP信号在神经组织发育的不同阶段以及中胚层模式中发挥着关键作用。 为了阐明ALK2信号在发育后期的功能，我们建立了嵌合分析来确定ALK2信号是否对心脏形态发生、眼睛发育以及沿着体轴的左右同一性的建立是必要的。为了建立Alk2的嵌合分析，我们分离了Alk2(-/-)的纯合子突变ES细胞，并将它们注射到野生型囊胚的腔内。ES细胞用β-半乳糖苷酶标记，因此，嵌合胚胎中的突变细胞很容易与野生型细胞区分开来。在嵌合胚胎中，由Alk2突变引起的异常被部分挽救，它们在子宫中可以比纯合子零突变多存活几天。在E9.0-E10.0处恢复的嵌合胚胎表现出多种表型，这取决于突变的ES细胞对Alk2的贡献程度。突变ES细胞贡献率较低(低于80%)的嵌合胚胎发育没有明显的表型。高贡献率的嵌合胚胎表现出生长迟缓，并在E11.0左右死亡。在正常胚胎中，心管在E9.0时向右环状，之后形成四腔结构。嵌合胚胎在旋转和心脏形态发生方面存在缺陷；心脏管环的形成是随机的。Nodeal、Pitx2和Lefty2在侧板中胚层有表达，但在正常胚胎中仅在左侧表达。有趣的是，在高度贡献的嵌合胚胎中，这些基因在侧中胚层的两侧都有表达。这些结果表明，嵌合胚胎未能在节点处建立左右不对称，从而导致左异构体。为了获得更深入的了解，我们在Alk2小鼠和Mox2-Cre小鼠身上建立了外胚细胞特异性的Alk2干扰。初步结果显示，结节内纤毛的形成受到损害，导致向左的结节流丧失。

Developmental process of vertebrate embryos is regulated, at least in part, by secreting molecules such as growth factors. We are focusing on the function of Bone Morphogenetic Proteins (BMPs) that are the members of TGF-beta superfamily during mouse development. To reveal the function of BMP signaling, we have generated a mutant mouse that is deficient for BMP type IA receptor (Bmpr1a or Alk3) and activin type IA receptor (Alk2) by conventional gene targeting technologies. Nullizygosity of each receptor caused severe embryonic lethality and mutant embryos die at embryonic day 7.5 (Bmpr1a) or 8.5 (Alk2). For Bmpr1a mutant embryos, we found that 1) no mesoderm was formed, and 2) cell cycles prior to gastrulation was prolonged. For the Alk2 mutant embryos, we found that 1) mesoderm was formed but not fully differentiated, 2) Alk2 signaling in the extraembryonic region (future placenta) was critical for gastrulation, and 3) Alk2 mutant cells were not capable to contribute heart or eye. These results suggest that BMP signaling at the early stage of embryogenesis is important for cell growth, gastrulation and formation of particular organs such as heart. For functional analysis of these gene products in a later stage of development, we introduced a tissue-specific gene targeting. Using this technology, we mutated Bmpr1a in a bone-specific manner (specific for differentiated osteoblasts). The bone-specific Bmpr1a deficient mice were viable indicating we can avoid embryonic lethality of Bmpr1a disruption with the tissue-specific gene targeting technology. Mutant mice were smaller than normal littermate and show less deposition of bone matrix in their bones. These results are the first evidences that BMP signaling is required for normal bone formation in vivo. Interestingly, the aged mutant mice showed reduced osteoclast activity that led to increased bone matrix. This suggests that primary function of BMP signaling in the differentiated osteoblasts is age dependent. To directly address the age-dependent alteration of BMP function in osteoblasts, we employed a tamoxifen inducible Cre system. A Cre recombinase of which activity can be induced by tamoxifen treatment was specifically expressed in osteoblasts using type I collagen promoter (Col1-CreER). When tamoxifen was administrated to newborn pups, bones in the mutant mice showed reduced bone mineral density (BMD) compared to the control littermate. When tamoxifen was administrated to the weaned mice between 1.5 month and 5 months after birth, bones in the mutant mice showed higher BMD than controls. These results are in good agreement with our previous finding and strongly suggest an age-dependent function of BMP signaling in osteoblasts. In another approach, we mutated Bmpr1a in a neural crest specific manner. Embryos die at the mid-gestation stage showing abnormalities in neural crest cell-derived tissues such as dorsal root ganglion and craniofacial region. These embryos also showed abnormalities in outflow tracts (OFT) where cardiac neural crest cells migrated. Preliminary results revealed a reverse flow in the dorsal aorta in E11.5 mutants, suggesting functional abnormalities in the cardiac cushion in OFT. This is the first evidence that the OFT cushion is functional as early as E11.5, and BMP signaling is one of the critical players for cardiac neural crest cells to form functional OFT cushion. We also mutated Bmpr1a in an epiblast-specific manner at the stage of gastrulation. Embryos showed an overgrowth of neural tissues after gastrulation. Paraxial mesoderm such as somite is expanded in the mutant embryos, but no sign of the heart development. These results indicate that BMP signaling plays a critical role in various stages of neural tissue development as well as mesodermal patterning. In order to elucidate the function of ALK2 signaling during the later stages of development, we set up chimeric analyses to determine if ALK2 signaling is essential for heart morphogenesis, eye development, and establishment of left-right identity along the body axis. To set up chimeric analysis for Alk2, we isolated homozygous mutant ES cells for Alk2 (-/-) and injected them into the cavities of wild type blastocysts. The ES cells are marked by beta-galactosidase, therefore, mutant cells can easily be distinguished from wild type cells in the chimeric embryos. In the chimeric embryos, abnormalities caused by Alk2 mutation are partially rescued and they can survive several more days in utero than a homozygous null mutant. Chimeric embryos recovered at E9.0-E10.0 showed a variety of phenotypes depending on the degree of contribution of Alk2 from the mutant ES cells. Chimeric embryos with low contribution of the mutant ES cells (less than 80%) developed without showing overt phenotype. Chimeric embryos with high contribution showed growth retardation and die around E11.0. In normal embryos, a heart tube exhibits rightward looping at E9.0, then forms a four-chamber structure later. Chimeric embryos showed defects in turning and heart morphogenesis; looping of the heart tube was randomized. Nodal, Pitx2, and lefty2 are expressed in lateral plate mesoderm, but their expression is limited to the left side in normal embryos. Interestingly, these genes were expressed on both sides of the lateral mesoderm in the highly contributed chimeric embryos. These results indicate that the chimeric embryos fail to establish left-right asymmetry at the node leading to left isomerism. To get further insights, we have set up an epiblast-specific disruption of Alk2 using floxed Alk2 mice and Mox2-Cre mice. Preliminary results revealed that cilia formation in the node was compromised leading to the loss of the leftward nodal flow.