As outlined in Proposition 71, very many people in California and beyond suffer from devastating diseases and injuries that are currently incurable, including cancer, diabetes, heart disease, Alzheimer’s, Parkinson’s, spinal cord injuries, blindness and so on. It is has been shown that normal, healthy tissues are renewed and regenerated constantly from special cells, called tissue-specific stem cells, that reside in all tissues. Many of the diseases and conditions listed above are caused or are accompanied by the failure of normal tissue regeneration. It is believed that these conditions could be vastly improved, if not cured, if medical science could figure out a way to restore normal tissue regeneration. A very promising approach is to generate tissue-specific stem cells in the laboratory and transplant them into patients to replace failing tissue. Currently the best source for tissue-specific stem cells (in the following called “progenitors”) is to derive such cells by culturing human embryonic stem cells (hESC) in the laboratory under conditions that resemble those in the tissue of interest. Transplantation of hESC-derived progenitors has shown good responses in animal models of human disease such as diabetes and Parkinson’s. However, before such cells can be used in patients, many issues still have to be addressed. One fundamental issue is the safety of such therapies, particularly the very important concern that transplanted progenitors in the long-term could cause cancer in the recipient. This concern is based on studies of hESC and some progenitors that have been shown to cause cancer in animals. However there are also studies in which progenitors have survived and functioned in animals without causing cancer. It is not known why some progenitors lead to cancer whereas others do not. We have designed a study using state-of-the-art technology to identify for the first time in a systematic fashion what factors might cause progenitors to form cancer. We will study the fate of many different, well-characterized progenitors in mice that allow the growth of human cells. We will follow the fate of the injected cells over the life time of the mouse, approximately 1 year. Once we establish which of the different progenitors cause cancer and which do not, we will analyze the genes and other properties of these cells to determine which factors individually, or as a group, correlate with the development of cancers. The results of this study will be of fundamental importance for all physicians and scientists currently working to cure various diseases and conditions with stem cell therapies and for all patients and their loved ones who hope to benefit one day from such therapies.
Statement of Benefit to California:
California, like much of the United States, is facing a staggering challenge to provide treatments for a population that is living longer than ever before. Increasingly physicians are treating chronic, debilitating, and therefore expensive diseases associated with aging. This is made all the worse by the demographic wave caused by the entry of the Baby Boomers into retirement. As a result, the percentage of the elderly in California is expected to grow from 14 percent in 1990 to 22 percent in 2030. (Source: California Department of Finance, Population Projections 1993). Many of the chronic devastating diseases of an aging population are the degenerative diseases. Generally speaking, degenerative diseases are those diseases caused by the loss or dysfunction of cells. Examples include osteoarthritis (loss of cartilage cells that protect the ends of the bones), Parkinson’s disease (the loss of dopaminergic neurons), osteoporosis (dysfunction of osteoblasts), macular degeneration (dysfunction of retinal pigment cells) and so on. More significantly, the loss or dysfunction of cells in the heart (or the vessels that supply the heart with blood) results in heart disease, the leading cause of death in California. In 2001 (the most recent year data is available) heart disease caused 68,234 deaths (29% of all the deaths in the state). Stroke is also a vascular disease and the third leading cause of death in California. Regenerative medicine represents the effort of cell biologists to provide a new approach to the problem of degenerative disease. Human embryonic stem cells (hESC) have the potential to become all of the cells in the human body, and their unique properties give researchers the hope that these primitive cells can result in new therapies. However, before hESC-derived cells can be used in the clinic, safety concerns must be addressed. The major concern regarding the transplantation of hESC-derived cell types is their potential to cause cancer. This study will illuminate the factors that contribute to the tumorigenicity of hESC-derived cells and will define parameters for their safe application in the clinic. Thus this study addresses a fundamental issue of all stem cell therapies rather than focuses on a disease specific application and therefore provides a universal resource for progress in regenerative medicine.
SYNOPSIS: The primary goal of this proposal is to test the in vivo tumorigenic propensity of a panel of cell lines derived from hES cells by the applicant's collaborators (Advanced Cell Technology) using a complex set of differentiation conditions. The lines were clonally derived and characterized by microarray. The applicant will modify these lines to express luciferase using viral vectors and then monitor their maintenance in transplanted immunodeficient mice using bioluminescence. Additional studies of a subset of the lines will use PET to track the cells. In addition, the applicant will test the minimal dose of contaminating hES cells that can be tolerated without tumor development in mixed hES/progenitor grafts. Finally, by analyzing gene expression data from each line, the applicant will attempt to define a "signature" of tumorigenic hES derived cells. IMPACT AND SIGNIFICANCE: A major barrier to translation of hES or hES-derived cell therapies is our limited understanding of the long-term potential for tumors arising from transplanted cells; not all hESC-derived progenitors develop teratomas, and it is not known which factors are involved in tumorigenicity. In this regard, the applicant is attacking a very important problem. Since ES cells have the potential to generate tumors when injected into adult tissues, there is a concern for the use of these cells and their derivatives in regenerative medicine. Enthusiasm for hESC-based therapies has been tempered by the fact that hESC form teratomas when injected into immune deficient mice. This proposal will use an immunodeficient mouse model to test the tumor-forming potential of a panel of hES derived cells, and may identify those with reduced tumorigenicity. In addition, the applicant will determine the minimal level of hES contamination that leads to tumor formation. The applicants propose to transplant a broad array of progenitors of different lineages into SCID mice and follow them long-term with bioluminescence imaging in Phase 1, while in Phase 2 they will investigate the dose of the progenitors that appears to be tumorigenic using both bioluminescence and PET imaging techniques. The use of state-of-the-art imaging modalities in this research is a significant strength, as the applicants will be able to monitor the formation (or not) of teratomas, starting with only a few hundred cells. Combining PET imaging by a reporter gene strategy with FHBG will enable the investigators to look at hESC transplant in deep-seated tissues. It is reasonable to assume, given the current state of the art, that hES-derived progenitors are more likely to be used clinically than undifferentiated stem cells (unless the control issues were to be better understood). By investigating a large number of progenitors with diverse phenotypes, and correlating these with various markers, much could be learned regarding improving the safety of hESC transplants. The mouse model is appropriate for the total body scans since it is the only animal model that can be imaged this way with current technologies. The multidisciplinary nature of the team is also laudable. However, while these findings may have relevance to the use of hES in human patients, exactly how well these findings in immunocompromised animals will translate to humans is unclear. Also, the likelihood that any of the ACT progenitor lines to be studied will in the future be introduced into humans is impossible to determine right now, as information about these lines is still rudimentary. QUALITY OF THE RESEARCH PLAN: This well-written and clear proposal describes research to investigate the tumorigenic potential of hESC-derived progenitor cells. Aim 1 is to transplant ~ 50 of these lines subcutaneously, after lentiviral infection/labeling with FLuc, and follow the cells using total body bioluminescence imaging every two weeks (for about 60 wk) and pathologic exam. Aim 2 is dependent on results of Aim 1: the most tumorigenic cells will be implanted either under kidney capsule or orthotopically, and again followed with bioluminescence and also PET imaging. Cells for PET imaging will have to be infected with HSV1-tk, so that an F18-tracer will be used to target them. If lines are tumorigenic, the studies will be repeated with lower numbers of cells to determine minimal tumorigenic cell number. Pathologic exam of selected organs is also an end-point of analysis. Aim 3 is to mix hESC with progenitors in various ratios to determine if hESC (and at what number) alter the tumorigenicity of transplanted progenitor cell populations. The hESC will be labeled to be distinguishable from the progenitor cells. While the research plan is relatively straightforward, the rationale for these studies is not clearly defined. The research plan could be considerably improved as it pertains to the progenitor lines chosen for study, which stems in large part from the fact that the panel of hES-derived cells proposed for study is very poorly defined. The assignment of these cell lines to different "lineages" appears to be based solely on gene expression analysis, and the sensitivity of this analysis is unclear. Microarray analysis to characterize the cells is hard to use as a baseline input. Functional data regarding the differentiation potential of these cells is largely lacking, and no basis for the selection of the 50 or so to be studied is provided. Were the microarrays done at just one time point? What about rates of proliferation? Furthermore, the list of oncogenes presented is hardly exhaustive. Are other things missed by simply looking at message? Are these oncogenes that were stable with passaging? Given that the environment into which the cells are transplanted after Aim 1 will also influence the cells, the chances of picking up meaningful data are really diluted by the experimental design. There are so many different cell types, so many expression patterns, so many places for transplant. It is understandable to want to perform a large-scale screen with the wonderful resource of the clonally-derived lines. But it is easy to argue that a focus on one progenitor type, chosen because it expresses a particular repertoire, and/or manipulated to express various potential tumorigenic factors, then transplanted in one place, would yield more information. Also, without knowing the potential for therapeutic utility of these lines, it is difficult to gauge the importance of determining their tumorigenicity. In addition, that comparison of these lines will yield a gene "signature" for cells that give rise to tumors is far from certain, particularly since tumor forming potential probably won't be a simple "yes/no" characteristic of these cells. It more likely will be a "gradient", with some cells generating tumors at low cell doses and others only at high doses. This is a problem for the scheme described in Aim 1 for classifying individual lines as tumorigenic or non-tumorigenic, as only one cell dose is proposed for this aim. Alternatively, none of these lines may form tumors, which would prohibit the studies proposed in Aim 2. Specific Aim 1 will determine the long term fate of hESC-derived progenitor cells transplanted into immune deficient mice. While the applicants acknowledge that the site of implantation may influence survival and tumorigenicity, the progenitor cells will be implanted subcutaneously (sc), since this location has allowed the development of teratomas in previous studies. It is also easy to survey many cell lines by injecting the progenitor cells sc. After sc injection of cells, the SCID mice will be imaged every other week to follow the fate of the cells and whether an increase in cell number occurs. If an increase in cell number is observed, the mouse will be euthanized and a necropsy performed. Tissues will be stained for markers of the original progenitor line as well as for stem cell markers. The applicants clearly outlined their expected results and potential problems. However, a question arises regarding only injecting cells sc. How will it be known whether injecting the progenitor cells into a different site might cause a teratoma? If no tumors form, is it only due to the location the cells were transplanted? Specific Aim 2 in part answers the question from Specific Aim 1, in that they plan to inject tumorigenic lines under the kidney capsule of SCID mice, and monitor tumor development by bioluminescence. Additionally, if tumor formation does occur, then lower doses of that cell line will be implanted to determine the tumorigenic dose. The implantation of cells in these deeper-seated locations in the mouse will then be monitored by PET imaging. The progenitors will be transfected with HSV1-tk, which then allows for PET imaging using F-18-FHBG. These studies will be performed at UCSF, and this group is well-suited to do these studies, since they have experience with SPECT imaging using I-123-FIAU. The proposal does not discuss the limit of detection by PET, although preliminary data presented with I-123-FIAU are impressive. Still, there are some problems with the experimental approaches chosen for imaging. The use of bioluminescence, while attractive in providing real time analysis, is problematic in its relative lack of sensitivity and resolution. How, for instance, will stable engraftment of the cells as differentiated progeny be distinguished in these animals from tumor maintenance? What will be the criteria for tumorigenicity? Also, it is unclear how one would translate the doses of hES that give tumors in mice into the doses that would be tumorigenic in humans. The imaging is in some ways a very crude measure of tumorigenicity, such that the real information is in the necropsy data. This very labor-intensive pathologic work would seem to require a really good mouse pathologist. How benign/malignant are tumors? How will they be graded? The animals that don’t develop tumors based on the imaging should be examined very carefully as well as the baseline, and this important control is not really handled in the research plan. Will all organs be sectioned and examined by H&E, and then by more specialized studies? Specific Aim 3 will determine the tumorigenicity of mixed populations of hESC and hESC-derived progenitors. hESC cells will be transfected to express renilla luciferase (RLuc), while the progenitor cells express firefly luciferase (FLuc). This is a very innovative and interesting study to determine whether the presence of hESC can alter the tumorigenic properties of non-tumorigenic progenitors. STRENGHTS: The strengths of this proposal are that it is clear and well-written, and the applicants have picked an important problem. The multidisciplinary team of investigators and collaborators is also a strength. The collaboration itself with Advanced Cell Technologies (ACT) to obtain the over 50 progenitor lines is important, since the use of novel hES derived cell populations as a potential substrate of clonally-derived lines has merit for clean data, though the real identity and promise of these remains unclear. Finally, complementing the imaging by bioluminescence, which is likely the best way to image dividing cells long-term, with PET imaging using a reporter-gene strategy is a strength. WEAKNESSES: The lack of preliminary data characterizing the cell lines to be studied, and the lack of a rationale for the selection of these lines are the main weaknesses of this proposal. There also is a lack of preliminary data with these cells in the in vivo model. The characterization of the cell lines is superficial in terms of the data that would be needed to sort out WHY some lines are tumorigenic and others not. The question of whether non-tumorigenic cells lines injected sc would be tumorigenic if they were injected in another site also is not fully addressed. In addition, there is insufficient biologic analysis of the mice, and too much emphasis is placed on just the imaging. There are concerns with the sensitivity of the detection strategy and criteria for designating tumorigenicity. In fact, while much of the success of this endeavor is based on the imaging, the PET is something that no one in the group really has experience with. The imaging is directed by a very junior investigator, without PI experience. He finished residency in 2004. Though he managed a very impressive amount of research as a medical student and resident, the C.V. has case reports and conference presentations, with only one senior author paper (presumably in 2006, though there is no date in the citation). Presumably the animals will have to be moved to the site of the PET and the imaging will be supervised by this investigator on site, and it is unclear from the C.V. if he has any animal experience. Other issues of note include the relatively low recent productivity of PI, and inaccuracies in SCRO Form The experiments are diffuse in design such that meaningful data will be difficult to extract, and there are several specific issues with the feasibility of the work. First, the mice will be anesthetized every two weeks for the scans, and in Aim 2, there are a lot of survival surgeries before the scans. No account is made for loss of animals (10 per cell line). Second, the numbers of different doses of transplanted cells that will be used to calculate minimal dose is not specified (half the original dose, then 1/10th, etc?). It is hard to calculate the numbers of animals and experiments that will be needed. Third, it is not until the timeline that we learn that the microarray data will be supplemented by protein expression data, but the nature of the protein quantitation is not presented. Fourth, the timeline for Aim 1 is unrealistic. It will take some time to develop the tagged clonal lines (say 3-4 months) then they are going to be characterized (but how?). The mice are then followed for their lifespan (60 wks). This is all slated for Year 1. DISCUSSION: The tumorigenic capacity of stem cells is a problem worthy of analysis. The main concern with this application is the poor characterization of the ACT cell lines (panelists later clarified that the ACT lines are not ESC lines at all, but rather progenitor lines derived from hESC.) The focus of the proposal is to evaluate the tumorogenicity of these lines where 50 to 100 cells will be injected in a single dose into mice and analyzed by imaging and transcriptional profiling of the tumorigenic vs. non-tumorigenic lines. The question is why do we care about whether these cells are tumorigenic if we don't know whether cells will ever be used in patients? How would the numerical results in mice be translated to humans? In the first pass the PI will use a single dose of cells, in a plus/minus screen for tumor formation. Doing a single dose for all lines may be problematic because the applicants may end up with a gradient of tumors. In addition, while the focus on imaging is interesting, the collaborating PI for the PET imaging work is relatively junior having just finished a residency in 2004. However, the collaborating institution (UCSF) does have a good PET center, but reviewers note that the PI does not list senior people at the imaging facility as collaborators on the grant. The collaborators also lack animal experience.