Type I diabetes is classified as a chronic disease because affected individuals can manage the condition by repeated insulin injection. However, a large number of patients eventually suffer from serious, life-threatening complications of their diabetic state, including blindness, kidney failure, amputation of limbs, heart disease, and stroke. Our goal is finding a cure for type I (and a severe form of type II) diabetes.
Recent transplantation of islets cells isolated from human cadavers into individuals with type 1 diabetes has shown that these patients became free from insulin-dependency for a significant period of time, raising the possibility that someday the disease may be cured by transplanting healthy, exogenous islet cells. However, the cadaver approach is hampered by severe limitations associated with the sourcing of sufficient tissue for the large pool of patients in need. We will never come close to obtaining enough islet cells to meet the demands of the over 1.5 million type 1 diabetics in the US using the present isolation methods. What is now required is an essentially unlimited supply of physiologically competent human islet cells.
Naturally, the current focus of diabetes research is directed towards generating replacement cells that can mimic the function of normal pancreatic b-cells. An approach attempted by many scientists has been to convert stem cells directly into b-cells in vitro. Despite efforts to date, the protocols currently available for driving the differentiation of ES cells into a pancreatic lineage are inefficient. In order to explore an alternative approach, we attempt to leverage the close developmental relatedness of liver and pancreas in designing a two-stage process to transform ES cells into b-cells. Since differentiating ES cells into liver cells appears to be somewhat simpler, human ES cells will be steered into the liver lineage first, and then to pancreas via directed expression of particular genes specific to pancreatic development.
This approach has several advantages. First, available methods for differentiating ES cells into liver are simpler and more efficient that those for pancreatic differentiation. Second, we have identified molecular factors involved whose activities show promise in assisting an in vitro conversion of liver to pancreas. Lastly, recently published pieces of evidence support the notion that a method including generation of b-cells from liver holds promise. If we are successful in our two-stage approach beginning with ES cells, we may even be able to identify liver cells within the body capable of being changed into insulin-producing cells. The significance of being able to use a diabetic’s own liver tissue to generate b-cells for transplantation is that this would not only eliminate the risk of transplant rejection, but also alleviate concerns over potential tumor formation from use of ES cells.
Statement of Benefit to California:
Diabetes is the leading cause of the blindness, kidney failure, non-traumatic amputations in adults, and contributes significantly to rates of heart disease and stroke. Because the disease affects millions of people in so many different ways due to diabetic complications, diabetes is increasingly being referred to as the epidemic of the New Millennium. The disease (type I and type II) affects over 17 million people nationwide, with approximately 2 million of these in California, based on a 1992 report. Approximately 95% of the overall diabetic population suffers from type II diabetes with the remaining 5% having type I. The total number of diabetics in California is expected to double by the year 2020.
The adverse effects of diabetes disproportionately burden the elderly, women, and people of color, stemming from poor economic circumstances, genetic background, or both. Numbers In all groups are expected to rise in the future. The occurrence of juvenile diabetes (type I) is also increasing, with type I diabetes expected to strike 1 out of every 200 children by 2010. Therefore, for many of Californians, the question is not whether they will develop the disease, but when its symptoms will begin to have major negative impacts on their lives.
The diabetes-related costs to the nation in terms of medical expenses, disability payments, lost work, and premature mortality is currently more than $100 billion, with California’s share of this cost at approximately $12 billion. There are over 300,000 diabetes-related hospitalizations in California each year, at annual cost of $3.4 billion (data from Diabetes Prevention and Control Program). Clearly, diabetes has a major negative impact on California’s economy. A cure for diabetes would allow physicians to radically improve the quality of patients’ lives while significantly relieving stress upon the economy by simultaneously reducing medical costs and improving productivity at the work place. The overall positive effects from a cure for diabetes upon society are immeasurable.
SYNOPSIS: The PI proposes that hESC will be a) differentiated into liver cells and b) then induced to express pancreatic proteins. The goal is to produce grafts for Type I diabetics. The first Specific Aim questions whether the protocols for differentiation of mESCs into liver cells can be used effectively with hESCs. hESCs will be transfected with a construct containing a reporter gene (tdTomato) under the albumin promoter. Differentiation to liver cells will be induced with a published protocol. Following successful conversion, selected pancreatic transcription factors encoded in retroviral constructs will be used to affect pancreatic “transdifferentiation” of the target mES and hES cells. The efficiency of the ESC to liver and liver to pancreas steps will be optimized in both species and the gene expression profiles of the resulting cells will be determined using Affymetrix chips. “Pancreatic” cells may be implanted under the kidney capsule of STZ-treated mice.
SIGNIFICANCE AND INNOVATION: The idea of transforming liver cells into insulin-producing cells is certainly not new. However, using hESC as a starting point offers interesting academic and potentially practical possibilities. Clearly, however, the obstacles to clinical application remain daunting, including the histocompatibility requirements, in addition to those of a purely technical nature. These requirements are addressed in part with the suggestion that the understanding to be derived from these experiments will enable the transformation of autologous adult liver stem cells into insulin secreting "pancreatic" cells. The proposed approach is certainly unorthodox. It entails a very interesting and rather innovative idea to differentiate the hESC into hepatic cells, and then convert them to pancreatic cells by the overexpression of specific differentiation factors. The chance that Aim I will work may not be high, but the pay off would certainly be high.
STRENGTHS: The PI has extensive experience in liver cell differentiation to pancreatic cells using Xenopus laevis as a model, and has worked on mHSC differentiation to hepatocytes. In addition, the PI's lab, with help from Drs. Donovan and Lock, should be technically qualified to do the work with S9 cells. There seems to be preliminary work done with mESCs and with Xenopus to suggest that hESCs may be transformed into cells expressing albumin and that these cells might acquire the ability of producing insulin by constitutively overexpressing pancreatic transcription factors. The PI has a JDRF grant to do the same work with mES cells and has already performed some critical experiments. There is a chance that transcription factors converting hepatic cells to pancreatic cells may be identified through Specific Aim 2. Although mESCs and hESCs are quite different, the insight gained through Xenopus and ideas refined in mESCs may facilitate the work, and the PI seems to be aware of the potential hurdles.
WEAKNESSES: This is a very high risk proposal that requires many elements to work well. The description of the work to be done is unnecessarily complicated, and it is not clear whether the objectives are to be accomplished in mESC, or whether insulin production or other pancreatic phenotypes has been already obtained. Specific Aim 2 is written as if it were to be done with mESC, although in the discussion of overlap with the JDRF grant, the PI states that no work with mESCs will be done in this CIRM SEED Grant.
There are a number of significant concerns about the experimental design. First, there are significant differences between hESCs and mESCs. It may take time to differentiate hESCs into liver cells, and even then the percentage of differentiated cells could be very low. Second, Pdx alone is known to be insufficient in the human system. Third, experimental steps are sometimes omitted. For example, in selecting "liver" cells for G418 resistance, the PI does not describe introducing the corresponding gene. Moreover, G418 selection is probably too slow for selecting differentiated cells. Finally, why use lentivirus? Why not use nucleoporation?
DISCUSSION: This proposal aims to study the creation of beta cells from hepatocytes. The approach is to go from liver cell differentiation to pancreatic protein expression. Differentiation efficiencies will be optimized in both species by studying gene expression profiles as determined by phenotypic markers and chips. Cells with pancreatic markers will be transplanted into kidney capsules to look for glucose stasis. This is not a new approach - mouse work suggests that the same protocols in human cells might work to produce insulin, which is a logical approach. The applicant has done substantial work with Xenopus and mESCs, and the lab is technically qualified to perform the proposed studies; however, there was substantial concern that the work plan is unnecessarily complicated, and while the idea is very interesting, the approach needs much refocusing. These elements make the proposal very high risk. Another major concern is that it is not clear how much work will be done in mESCs versus hESCs. There was also discussion over why the applicant is differentiating into hepatocytes first as a necessary intermediate for generating pancreatic beta cells. The protocols for generating hepatocytes from mESCs are well described in mice, and perhaps the efficiency of differentiating hepatocytes into beta cells is higher in the mouse system than going directly from mESCs into beta cells. While there may have been success with trans-differentiating hepatocytes in some scenarios to pancreatic cells, this may be a red herring given the recent success in directly differentiating hESCs to pancreatic cells. Also, the use of retroviral transduction to generate beta cells is considered a dead end strategy to get beta cells for therapy. Overall, reviewers suggested that this proposal has some very interesting features and should be refocused for resubmission.