New Cell Lines
Our objective is to use induced pluripotent stem (iPS) cell technology to produce a cell-based test for long QT syndrome (LQTS), a major form of sudden cardiac death. Nearly 500,000 people in the US die of sudden cardiac death each year. LQTS can be triggered by drug exposure or stresses. Drug-induced LQTS is the single most common reason for drugs to be withdrawn from clinical trials, causing major setbacks to drug discovery efforts and exposing people to dangerous drugs. In most cases, the mechanism of drug-induced LQTS is unknown. However, there are genetic forms of LQTS that should allow us to make iPS cell–derived heart cells that have the key features of LQTS. Despite the critical need, current tests for drug-induced LQTS are far from perfect. As a result, potentially unsafe drugs enter clinical trials, endangering people and wasting millions of dollars in research funds. When drugs causing LQTS such as terfenadine (Seldane) enter the market, millions of people are put at serious risk. Unfortunately it is very difficult to know when a drug will cause LQTS, since most people who develop LQTS have no known genetic risk factors. The standard tests for LQTS use animal models or hamster cells that express human heart genes at high levels. Unfortunately, cardiac physiology in animal models (rabbits and dogs) differs from that in humans, and hamster cells lack many key features of human heart cells. Human embryonic stem cells (hESCs) can be differentiated into heart cells, but we do not know the culture conditions that would make the assay most similar to LQTS in a living person. These problems could be solved if we had a method to grow human heart cells from people with genetic LQTS mutations, so that we know the exact test conditions that would reflect the human disease. This test would be much more accurate than currently available tests and would help enable the development of safer human pharmaceuticals. Our long-term goal is to develop a panel of iPS cell lines that better represent the genetic diversity of the human population. Susceptibility to LQTS varies, and most people who have life-threatening LQTS have no known genetic risk factors. We will characterize iPS cells that have well-defined mutations that have clinically proven responses to drugs that cause LQTS. These iPS cell lines will be used to refine testing conditions. To validate the iPS cell–based test, the results will be directly compared to the responses in people. These studies will provide the foundation for an expanded panel of iPS cell lines from people with other genetic mutations and from people who have no genetically defined risk factor but still have potentially fatal drug-induced LQTS. This growing panel of iPS cell lines should allow for testing drugs for LQTS more effectively and accurately than any current test.
Statement of Benefit to California:
Heart disease is the leading killer of adults in the Western world. Nearly 500,000 people in the US die of sudden cardiac death each year. Our goal is to develop a cell-based test to screen for drugs that can cause sudden cardiac death. Drug-induced cardiac side effects are the most common reason for withdrawal of drugs from clinical trials, causing major setbacks to drug discovery efforts. Therefore our test we will improve the safety of pharmaceuticals. Our test will also reduce the change that a drug in development will fail during clinical trials, thereby decreasing the financial risk for pharmaceutical companies. The results of our studies will help develop new technology that is likely to contribute to the California biotechnology industry. Our studies will develop multiple lines of iPS cells with unique genetic characteristics. These cell lines could be valuable for biotechnology companies and researchers who are screening for drug compounds. We are working closely with California companies to develop new microscopes, assay devices, and analytical software that could be the basis for new product lines or new businesses. If therapies do come to fruition, we anticipate that California medical centers will be leading the way. The most important contribution of this study will be to improve the health of Californians. Heart disease is a major cause of mortality and morbidity, resulting in billions of dollars in health care costs and lost days at work. Our goal is to contribute research that would ultimately improve the quality of life and increase productivity for millions of people who suffer from heart disease.
Executive Summary This is an interesting proposal to develop new human induced pluripotent stem (iPS) cell lines derived from patients that present with long QT syndrome (LQTS). The goal is not to use these cells for regenerative medicine, but rather to use them to develop research reagents, specifically cellular assays that accurately reflect susceptibility to a form of drug-induced arrhythmia. The plan is to derive iPS cells from patients with LQTS, and demonstrate pluripotency. These iPS cells will be differentiated to cardiomyocytes, and the electrophysiologic profiles will be compared with normal iPS-derived cardiomyocytes. Finally, the principal investigator (PI) plans to adapt culture conditions of iPS cell-derived cardiomyocytes for high-throughput preclinical screening of drugs. LQTS is a serious and potentially deadly condition that reflects an abnormal cardiac ventricular repolarization interval. It is a major cause of arrhythmia and sudden cardiac death, often in response to drug exposure or physiologic stresses. Genetic forms of LQTS result from mutations in ten genes, including a specific ion channel, which is the focus of this study. Drug-induced LQTS often also occurs in people with no detectable genetic mutation. The PI seeks to develop a panel of iPS cell lines constituting a range of LQTS susceptibility and to develop in vitro assays to better predict drug-induced responses. This is an important problem and significance is potentially very high, since drug-induced LQTS is a major reason that drugs are withdrawn from clinical trials. There are several major strengths in this proposal. First, the team assembled to do the work appears to have the expertise and commitment to the project and notably includes a pioneer in the development of the iPS technology. Second, the PI will generate the iPS cells from patients with known mutations in a specific ion channel, and can compare the electrophysiologic findings with responses of the patients themselves, and possibly with cells from unaffected relatives. Finally, one underlying assumption in the proposal is well-founded: there is an imperfect relationship between electrophysiologic findings obtained by screening in conventionally-used in vitro assays, and animal models may carry similar liabilities. A leading hypothesis to explain this disconnect is failure of heterologous systems to express all cardiac proteins, some of which may interact with ion channels to produce a human phenotype. Thus, the incorporation of a mutant channel into an authentic human cardiomyocyte background could represent an important step to developing better models. Reviewers identified some weaknesses with the proposal. The proposal lacked key preliminary data attesting to the ability to generate cardiomyocytes from these human fibroblasts. Work by this team in the mouse is reassuring, but one reviewer noted that mouse electrophysiology is different from humans, and myocytes derived from larger mammals display considerable heterogeneity within the same chamber. It was not clear to reviewers how the homogeneity and maturity of the derived cells will be validated. One reviewer felt this could also impact the third aim to develop high throughput assays, in which reproducibility and homogeneity of cell types, rather than efficiency, is key. A final caveat from one reviewer is that the PI should carefully consider how to choose which cardiomyocytes to study. If the ultimate goal of the screen is to understand the underlying mechanism or to determine how a specific patient will respond to a drug, then cells with the mutations are appropriate. However, if the goal is to predict which drugs will cause abnormal effects in a population, the appropriate cells to study are those from normal individuals. In summary, this is a proposal to establish a cell-based test for LQTS that can be used to identify drugs that cause LQTS and its attendant arrhythmias. The strengths are the investigative team, and the possibility that new reagents that would assist in understanding mechanisms underlying these arrhythmias would be developed. The proposal was felt to be responsive to the RFA and novel in seeking to develop a unique research toolkit, rather than regenerative protocols. Programmatic Discussion: A motion was made to recommend that this application be moved to Tier 1 – Recommended for Funding. The panel highlighted the strength of the research team and the utility of the proposed screening system as reasons to recommend this application. Technical difficulties with the proposal were acknowledged, but the panel’s sentiment was that this was the team that could accomplish the work. The motion to move this application to Tier 1 carried. Reviewer Synopsis This is an interesting proposal to develop new human iPS cell lines derived from patients that present with long QT syndrome. The goal is not to use these cells for regenerative medicine, but rather to use them to develop research reagents, specifically cellular assays that accurately reflect susceptibility to drug-induced arrhythmia. Reviewer One Comments Significance: This is a proposal to use the technology of generating induced pluripotent stem (iPS) cells from patient-specific fibroblasts to create a new cell based test for drug-induced long QT syndrome. The plan would be to acquire skin fibroblasts from individuals with Type 2 congenital long QT syndrome and family controls, and to generate from these fibroblasts pluripotent stem cells which would then be differentiated into cardiac myocytes and embryoid bodies. Since Type 2 long QT syndrome involves loss of function mutations in KCNH2 (HERG) and since the mechanism for drug-induced long QT syndrome is almost always block of the potassium channel encoded by this gene, the assumption is that cardiomyocytes derived from mutation carriers will be more sensitive to drug block than those not carrying such mutations. Eventually, this approach would then be expanded to include stem cells from patients with congenital long QT syndrome in whom mutations in HERG or other candidate genes have not been identified. In this case, studying the electrophysiology would hopefully inform the ion current(s) that might be the culprit, and therefore, direct genetic screening toward specific candidate pathways. There are 2 major strengths in this proposal. First, the team assembled to do the work appears to have the expertise and commitment to the project and notably includes Dr. Yamanaka, a pioneer in the development of the iPS technology. Second, one underlying assumption in the proposal is well-founded: there is an imperfect relationship between electrophysiologic findings obtained by screening wild-type or mutant ion channels in heterologous expression systems and in vitro phenotypes, and conventionally-used animal models, such as whole rabbit or tissues derived from dog, may carry similar liabilities. A leading hypothesis to explain this disconnect is failure of heterologous systems to express all cardiac proteins, some of which may interact with ion channels to produce a human phenotype. Thus, the incorporation of a mutant channel into authentic human cardiomyocyte background could represent an important step to developing better models. Indeed, even the development of cardiomyocytes carrying “wild type” channels, or channels with common polymorphisms whose electrophysiologic effects have been controversial, might represent an important advance in the field. Feasibility: Weaknesses are identified. One weakness is the choice of patients. If the goal is to understand the mechanisms underlying susceptibility to drug-induced long QT syndrome, then surely the right fibroblasts from which to obtain cardiomyocytes are those in patients with drug-induced long QT syndrome. Occasionally, patients with “subclinical” congenital long QT syndrome (sometimes involving KCNH2, sometimes one of the other disease genes implicated in this disorder) have declared themselves after developing an episode of arrhythmia upon exposure to a QT prolonging drug. This seems no more common with KCNH2 mutation carriers than with carriers of mutations in other genes, notably KCNQ1, which underlies potassium currents (IKs). Thus, one could justify, perhaps even more easily than a KCNH2 model, incorporating mutations in other ion channel genes here. Again, as outlined above, the current approach to assessing drug liability is to screen for block of wild type KCNH2 channels expressed in heterologous systems, and so a major advance that this research may provide is the generation of authentic human cardiomyocytes expressing the wild type channel. A related weakness is the real lack of any phenotypic information. The fibroblasts will be obtained from patients followed by the clinical genetic arrhythmia group at UCSF, which is internationally recognized for its expertise in this area, and this is reassuring. However, which individuals in an affected kindred would be chosen (the ones with longest QT intervals? The ones who are mutation carriers but with normal QT intervals?), and whether individuals exposed to drugs would be chosen as well, is not addressed. Another weakness is the lack of key preliminary data attesting to the ability to generate cardiomyocytes from these fibroblasts. The fact that this has been accomplished in mice is reassuring, but mouse electrophysiology is known to be quite different from that in humans, and indeed myocytes derived from larger mammals display considerable heterogeneity within the same chamber. Thus, it is entirely possible that the cardiomyocytes to be derived will be largely “epicardial” yet the primary myocytes in which screening would be desirable are from the mid-myocardial region, or some other combination. Simply deriving “a” cardiomyocyte is simplistic, since there are many subcellular types and how the investigators will go about choosing these for drug testing is not clear. I have no comments on the investigative team, which is strong, or on the budgets. Summary This is a proposal to generate human cardiomyocytes with the goal of identifying individuals, or drugs, at risk for drug-induced long QT syndrome and its attendant arrhythmias. The strengths are the investigative team, and the possibility that new reagents that would assist in understanding mechanisms underlying these arrhythmias would be developed. The weaknesses are primarily a lack of preliminary data attesting the feasibility of the study (hard to generate that without the funding), and a failure to think through some of the important operational issues in this experiment, such as whether all cardiomyocytes are the same (they are not), how to choose which cardiomyocytes to study, and whether, in fact, engineering cardiomyocytes with KCNH2 mutations is the most appropriate way to go about identifying culprit drugs or even patients susceptible to this arrhythmia. Reviewer Two Comments Significance: LQTS is a serious and potentially deadly condition that reflects an abnormal cardiac ventricular repolarization interval. It is a major cause of arrhythmia and sudden cardiac death. There are so far 10 disease genes that have been mapped, but the majority of LQTS is caused by mutations in ion channels, e.g. KCNQ1 or KCNH2 (HERG). For reasons that are generally not well understood, LQTS is often triggered by drug interactions or stress. Most importantly, the current assays used to test new drugs for potential to trigger LQTS are either in animal models or channel-expressing non-cardiac cells that are likely not optimized in the context of the clinical manifestation of the syndrome. Improperly tested drugs are therefore dangerous, and improper tests might invalidate potentially good drugs. Therefore, the PI seeks to develop a panel of iPS cell lines constituting a range of LQTS susceptibility and to develop in vitro assays to better predict drug-induced responses. This is an important problem and significance is potentially very high. Feasibility: Because in a majority of cases the KCNH2 channel is affected in LQTS, directly or indirectly, the PI has chosen to focus on this as a valuable predictor. He will start with fibroblasts derived from patients with known KCNH2 mutations that are sensitive to known drugs. Their behavior can thereby be compared directly with the clinical responsiveness of the patients themselves. This is a real strength of the proposal. Donors from the UCSF clinic are apparently already identified, and it may be possible to obtain samples from unaffected family members, although feasibility for this was not clearly described. The first Aim is to generate iPS lines from 10 defined LQTS patients and 10 controls. This will be carried out in the Gladstone Stem Cell Core (directed by the PI) with the consult of Dr. Yamanaka. Pluripotency will be evaluated by embryoid body (EB) formation and teratoma assays. In Aim 2, the PI will attempt to optimize differentiation of iPS lines to cardiac fate, using EBs, induction with END2 cells, or via the Laflamme et al. protocol from the Murry laboratory. There are some serious caveats here, since it is not clear how efficiently cardiac cells can be generated in vitro from iPS cells, nor is the maturity and homogeneity of such derivatives validated. Will embryonic-like cardiac cells suffice for the model, or will it require much more mature cardiomyocytes? To the advantage of the PI, efficiency may not be so important, since whole cell recordings or patch-clamping can be done in single cells, assuming they can be appropriately identified. The final aim is to develop high throughput assays for the development of clinical screens. Again, since efficiency and cell numbers might be a major challenge, the goal is instead to miniaturize the process, using micro-fluidics to immobilize dissociated cardiac myocytes for single cell recordings, with cells exposed to various drugs. A key here will be reproducibility and homogeneity rather than efficiency. Preliminary data that he can compare WT and mutant ES-derived cardiomyocytes is provided from a different study by the PI using the mouse ES system. A big advantage in the human system will be the ability to use cells with known mutations and known drug-susceptibility as a positive control. Responsiveness to RFA: The proposal is responsive by proposing to generate a unique panel of pluripotent human iPS lines, carrying defined genetic alterations in a clinically relevant gene. Given the expertise at the Gladstone it seems that the lines should be generated. Whether an appropriate cardiac differentiation assay will be developed is more risky, although the PI has expertise and has also recruited Dr. Aalto-Setala, a Finnish colleague with expertise in hES-derived cardiac cells. Other colleagues with expertise in micro-fluidics can assist in the assay development. The proposal is rather novel in seeking to develop a unique research toolkit, rather than regenerative protocols, and this is a strength regarding feasibility. Reviewer Three Comments Feasibility: - determine if iPS cells from LQTS syndrome are pluripotent; this is feasible - determine if iPS derived heart cells from LQTS have a physiological defect is also possible, however it is not clear to me how the investigators will distinguish between fetal or embryonic and adult cells - there is the issue of which specific cardiac myocytes will really be generated, since there is some heterogeneity within the cardiac muscle tissue - high throughput assay for drugs: it is not clear to me how this will be conducted or how realistic this aim is - ES cells on nanowires: this is an interesting idea however high risk and not clear how this would affect the overall project - differentiation into cardiac myocytes is likely to be more difficult than it was initially proposed