Mouse model to assess putative beta cell progenitor cells

我们建立了灵敏和强大的体内试验系统，以评估假定的干细胞分化为能够生理调节胰岛素分泌的细胞的能力。利用我们的小鼠实验性自身免疫性糖尿病（EAD）模型系统，我们设计了一种体内试验来测试干细胞是否能在持续的自身免疫性反应中有效地产生和维持β细胞。该模型的具体特征包括：(i)正常宿主代谢维持较长时间，随后（ii）胰岛功能逐渐下降，需要新的β细胞，（iii）胰岛内炎症可能为调节干细胞向胰腺迁移并分化为β细胞提供关键线索，（iv）新开发的干细胞衍生的β细胞可抵抗自身抗原特异性T细胞介导的攻击。使用该系统，我们测试了一些报道的造血干细胞（HSC）的粗制剂，这些造血干细胞含有能够转分化为胰腺细胞的细胞。我们发现HSC在体内没有形成β细胞的功能相关能力。使用该模型，我们彻底评估并通过实验排除了几个关键问题，例如体内不能产生产生胰岛素的β细胞是否可能是由于新发育的β细胞无法承受炎症环境的毒性，这种炎症环境通常与免疫介导的胰岛破坏有关。通过证实供体来源的胰岛（干细胞基因型）比宿主类型的胰岛在混合胰岛移植排斥反应中选择性存活，我们获得的数据有力地支持了来自骨髓供体来源的新生成的β细胞可以免受免疫介导攻击的说法。作为该模型的阳性对照，我们已经证明，从小鼠胚胎胰腺中分离的单细胞，然后注射到成年小鼠胰腺中，将分化成成熟的胰岛。更具体地说，使用芝加哥大学Manami Hari博士提供的小鼠，其中绿色荧光蛋白的β细胞表达由小鼠胰岛素启动子驱动，我们已经证明，可以将e14.5天胰腺芽细胞的单细胞悬液注射到成年同基因小鼠的胰腺中，并且来自胚胎小鼠胰腺的固定祖细胞将在成年小鼠胰腺中分化为成熟的胰岛。此外，与胚胎供体分化的胰岛通过其绿色明显区分。在相关工作中，Pechhold博士与合作者使用细胞标记技术证明，胰岛β细胞不能有效地去分化为快速增殖的细胞（即所谓的上皮细胞向间充质细胞转化），然后在体外重新分化为产生胰岛素的细胞。

We established sensitive and robust in vivo test systems to assess putative stem cells for their ability to differentiate into cells capable of physiologically regulated insulin secretion. Using our mouse experimental autoimmune diabetes (EAD) model system, we have designed an in vivo assay to test if stem cells can efficiently generate and maintain beta cells during an ongoing autoimmune response. Specific features of the model include: (i) normal host metabolism is maintained for an extended time followed by (ii) gradually declining pancreatic islet function creating a need for new beta cells, (iii) intra-islet inflammation which may provide critical cues for regulated stem cell migration to the pancreas and differentiation into beta cells, and (iv) newly developed, stem cell-derived beta cells are resistant to autoantigen-specific T cell mediated attack. Using this system, we tested crude preparations of hematopoetic stem cells (HSC) reported by some to contain cells capable of transdifferentiating into pancreatic beta cells. We found that the HSC had no functionally relevant capacity to form beta cells in vivo. Using the model, we thoroughly evaluated and experimentally ruled out several critical questions such as whether the failure to generate insulin producing beta cells in vivo might have been due to the inability of newly developed beta cells to withstand the toxicity of an inflammatory environment typically associated with immune-mediated pancreatic islet destruction. Data we've generated strongly support the claim that newly generated beta cells from the bone marrow donor origin would be protected from immune mediated attack, by confirming the selective survival of donor-derived pancreatic islets (stem-cell genotype) over host-type islets during rejection of mixed islet transplants. As a positive control for this model, we've shown that single cells isolated from embryonic mouse pancreata, then injected into the adult mouse pancreas, will differentiate into mature islets. More specifically, using mice provided by Dr. Manami Hari (University of Chicago) in which beta cell expression of green fluorescent protein is driven by the mouse insulin promoter, we've shown that a single cell suspension of day e14.5 pancreatic bud cells can be injected into the pancreas of adult syngeneic mice and that committed progenitor cells from the embryonic mouse pancreas will differentiate in the adult mouse pancreas into mature appearing islets. Moreover, the islets differentiating from the embryonic donors are clearly differentiated by their green color. In related work, Dr. Pechhold demonstrated with collaborators using cell labeling techniques that islet beta cells do not efficiently de-differentiate into rapidly proliferating cells (so called epithelial to mesenchymal transition) that can then be re-differentiated into insulin producing cells in vitro. We have also demonstrated and reported that a gene-therapy approach described by others as efficiently promoting transdifferentiation, i.e. converting liver cells into insulin producing cells, was in fact very inefficient. Our published studies suggested that mouse liver Pdx-1 expression driven by adenoviral vectors generated less hepatic insulin production than 16 islets, an amount clearly insufficient to maintain glucose homeostasis. Last, in view of our promising model system for identifying putative islet progenitor cells, we were joined in 2009 by a professor from Brazil who elected to spend his sabbatical year with us further develop and characterize the system, and to identify the pancreatic islet progenitor cell. A manuscript describing these findings is being prepared.