The re-programming of post-natal somatic cells to induced pluripotent stem cells (iPSCs) via ectopic expression of stem cell specifying transcription factors has many exciting potential applications for improving human health. iPSCs were initially developed in the murine model, and shown to have the potential to contribute to all tissues via blastocyst complementation assays. Just a few years later, human iPS cells were created using a similar panel of transcription factors, and demonstrated to form teratomas in immunodeficient mice and share functional and gene expression characteristics with human embryonic stem cells. However, there are numerous hurdles to moving iPSC forward into clinical regenerative medicine applications. First and most important are safety concerns. Both murine and human iPSCs were initially derived by introducing the required transcription factor genes into target cells using integrating vectors, associated with risks due to ongoing or reactivated ectopic expression of the transcription factors, or insertional activation of genomic proto-oncogenes. Novel non-integrating vectors or protein transfer systems have begun to surmount this problem. However, much more serious concerns relate to the consequences of administering primitive pluripotent cells that may have the potential to form tumors, if differentiation is incomplete or inefficient. Second, there are significant challenges to the efficient differentiation of iPSCs into functional adult tissues. Protocols for differentiation of iPSCs towards even well-characterized hematopoietic stem cells are inefficient, inconsistent and result in aberrant or embryonic hematopoiesis. Design of methods for direct delivery or facilitation of homing of iPSCs or their progeny to appropriate locations in the body will also be a major challenge. While murine models are invaluable tools, it will be critical to develop more relevant models for clinical development of iPSCs. Murine and human embryonic stem cells behave quite differently in culture, require different cytokines and handling, and may be derived from different stage of embryonic development. Generation of murine iPSCs appears to be at least an order of magnitude more efficient that generation of human iPSCs. Telomeres in inbred laboratory mice are significantly longer than human telomeres, and may impact on the relative ease of immortalization of murine versus human cells and thus oncogenicity. Human iPSCs can be implanted in immunodeficient mouse strains and form teratomas, but the next steps in development, requiring functional differentiation and appropriate delivery or homing, may be impossible to model in xenografts. Scale-up of laboratory procedures developed in mice to human therapies would also be very difficult to develop solely using murine-murine or human-murine xenograft models. The rhesus macaque non-human primate (NHP) model could be a valuable resource to clear hurdles preventing clinical development. The close physiologic and genomic relationship between humans and NHPs results in cross-reactivity for most cytokines, antibodies and other reagents. Teratoma formation and other safety issues could be directly assessed utilizing autologous rhesus iPSCs. Differentiation, homing and other parameters critical for efficacy can be modeled. Tissue damage models such as pancreatic beta cell or hematopoietic stem cell ablation are well established in macaques. Rhesus embryonic stem cells were isolated prior to human ESCs, and their properties are well-characterized. Thus far, a single report has been published regarding derivation of macaque iPSCs13, and a second reporting marmoset iPSCs.14 The authors of these studies have not pursued development of a rhesus iPSC or marmoset in vivo model, and do not have access to a NHP transplantation or intensive medical support facility. Development of rhesus iPSCs at the NIH would take advantage of our unique expertise in NHP transplantation and in the development of novel cell and gene therapies in this valuable model. Our plans also focuses on the development of a suicide gene strategy to increase safety of utilization of iPSCs for tissue regeneration. There is a significant risk that residual pluripotent cells remaining following direct differentiation of iPSCs could form tumors in vivo. If integrating vectors are used to generate iPSCs, tumors could result from vector-related insertional mutagenesis or re-activation of reprogramming factors. Even if differentiation is complete and successful, iPSC-derived progeny might localize or proliferate inappropriately. In all these scenarios, the ability to ablate iPSCs in vivo would be desirable. Several promising suicide gene strategies have been developed over the past decade, allowing efficient killing of cells carrying the suicide gene vector via administration of a non-toxic drug. Several clinical trials utilizing allogeneic T cells carrying the herpes thymidine kinase (tk) suicide gene have been performed. Ganciclovir, a pro-drug only toxic to cells expressing herpes tk, was shown to ablate alloreactive T cells and successfully treat graft-versus-host disease. To avoid the immunogenicity of herpes tk, another promising suicide system utilizes human caspase 9 fused to the FK506 binding domain, allowing inducible dimerizer and caspase 9 activation following administration of an oral small molecule dimerizer. These suicide gene strategies hold great promise for iPSC safety, but need further clinical development in a relevant model such as the rhesus macaque. Thsi project began this year, and we have already published a paper taking advantage of our expertise in vector integration site retrieval and analysis to demonstrate that in human iPSCs, vector integration sites do not appear to play a role in promoting successful reprogramming of iPSCs. This is reassuring for at least preclinical and model development, allowing continued use of lentiviral vectors for reprogramming, given their much greater efficiency compared to non-integrating vectors. We have carried out a very intensive series of iterative experiments to finally derive rhesus iPSCs in the past two months. The conditions utilized for murine and human ESCs and iPSCs were not successful using rhesus cells, and we have developed new conditons, based on the optimal conditions for growing rhesus ESCs. Our rhesus iPSCs express alkaline phosphatase, shut off the reprogramming vectors and morphologically resemble rhesus ESCs. Lineage differentiation and teratoma studies will commence in the next year. We have developed an in vivo bone organoid model in the rhesus, in collaboration with Pam Robey's group in NIDCR. We are currently testing the model with rhesus MSCs, and plan to move into rhesus iPSCs in the next fiscal year.
Showing the most recent 10 out of 21 publications
