6:00 pm - 8:30 pm
Bridging the Gap: Human Stem Cell Solutions for Better Predictivity
Add this event to your calendar
8:30 am - 9:30 am
Moscone Convention Center
Human Induced Pluripotent Stem Cell Technology in Predictive and Mechanism-based Drug Discovery and Toxicity Testing Using Photometric-based Assays
Add this event to your calendar
Characterization and Utilization of Induced Pluripotent Stem Cell-derived Cardiomyocytes in Identifying Novel Mechanisms of Cardiotoxicity
Dinesh Puppala, Pfizer Inc.
9:30 am - 12:30 pm
Abstract: 420 | Poster: 642
Cardiotoxicity is one of the leading causes of early and late stage drug attrition which has significant impact on patients' welfare and in particular for cancer patients. Current in vitro models insufficiently predict cardiotoxicity and there is a definite need for alternate physiologically relevant in vitro models. Induced pluripotent stem cells offer the promise for development of tissue specific in vitro models. In here, we describe the use of induced pluripotent stem cell-derived cardiomyocytes (iCell Cardiomyocytes) in identifying novel molecular mechanisms of cardiotoxicity. Towards this end, we characterized iCell Cardiomyocytes using gene expression analyses over a 42 day period post thaw in culture. Based on our initial analyses iCell Cardiomyocytes express relevant cardiac markers such as ion channels (SCN5A, KCNJ2, KCNQ1 and KCNH2), cardiac tissue markers(MYH6, MYLPF, MYBPC3, DES, TNNT2 and TNNI3), cardiac transcription factors (NKX2.5, GATA4 and GATA6) and lacked the expression of stem cell markers(FOXD3 , GBX2, NANOG, POU5F1, SOX2, and ZFP42). We then selected ten in vivo structural cardiotoxicants which were not flagged by our current in vitro assays and generated gene expression data in response to these compounds in both iCell Cardiomyocytes and rat heart derived cell line H9C2. To further evaluate the data, we used our internal Casual Reasoning Engine(CRE) platform, which analyzes experimental gene expression data, in the context of prior biological knowledge to generate testable upstream molecular hypotheses. Based on our initial analyses across all compounds, iCell iCell Cardiomyocytes exhibited increases in structural cardiotoxicity injury signals, inflammation signals and TGF beta signaling pathways whereas H9C2 cells exhibited increase in muscle injury signals, decreases in both TGF beta and inflammation signals. In conclusion, iCell Cardiomyocytes appear to represent a better in vitro cardiomyocyte model for understanding cardiotoxicity mechanisms and further validation will be performed to confirm these findings.
Predictive Assays for High-throughput Assessment of Cardiac Toxicity and Drug Safety
Evan F. Cromwell, Molecular Devices, Inc.
9:30 am - 12: 30 pm
Abstract: 416 | Poster: 638
A large percentage of new drugs fail in clinical studies due to cardiac toxicity. Therefore development of highly predicative in vitro assays suitable for safety and efficacy testing is extremely important for drug development. Human cardiomyocytes derived from stem cell sources can greatly accelerate the development of new chemical entities and improve drug safety by offering more clinically relevant cell-based models than those presently available. Human induced pluripotent stem cell (iPSC) derived cardiomyocytes are especially attractive because they express ion channels and demonstrate spontaneous mechanical and electrical activity similar to native cardiac cells. Once such cell type, the iCell® Cardiomyocytes, has been extensively characterized and tested for gene expression, ion channel activity, and purity and is well suited for a predictive assay. Here we demonstrate cell based assays for measuring the impact of pharmacological compounds on the rate of beating cardiomyocytes with different assay platforms. We developed methods for the ImageXpress® Micro and the FLIPR® Tetra systems that enable determination of beating rate from a series of time-lapse images. The systems employ calcium sensitive dyes to monitor changes in intensity of Ca2+ fluxes synchronous with cell beating. The calcium signal is used as a surrogate marker for action potential and the method allows monitoring of drug impact on the beat rate, amplitude, and other contraction parameters in 96 or 384 well formats. We have shown dose-dependent atypical patterns caused by hERG inhibitors and other ion channel blockers. Cardioactive compounds are used in clinical treatment of heart failure, arrhythmia or other cardiac diseases. We have demonstrated effects of several positive (epinephrine, etc.) and negative (a and b blockers) chronotropic agents on cardiac rates and determined IC50 values. These methods are well suited for safety testing and can be used to estimate efficacy and dosing of drug candidates prior to clinical studies.
Comparison of Two Cell Model Systems, Primary Human Hepatocytes and Human iPS Cell-derived Hepatocytes to Determine the Hepatotoxicity of Three Candidate Drugs Developed for Rheumatoid Arthritis
Sei Kameoka, F. Hoffmann-La Roche
1:00 pm - 4: 30 pm
Abstract: 519 | Poster: 237
In vitro hepatocyte culture serves as a very useful model system to detect potential hepatotoxicity without sacrificing animals. However, current in vitro model systems have limitations. Primary human hepatocytes (PHH) are known to exhibit donor-specific phenotypic variability and availability is too limited for high throughput assay. Immortalized hepatoma cell lines, such as HepG2 and HepaRG, offer more reproducible and homogenous culture systems, but interpreting the toxicant sensitivity in cells of tumor origin poses a considerable problem. Using human induced pluripotent stem cell (hiPSC)-derived hepatocytes is an important new tool which offers unlimited supply of euploid cells from single donors. Here, we tested the hepatotoxicity of three internal candidate drugs developed for the treatment of rheumatoid arthritis. In vivo, two of these compounds induced canine liver toxicity, while one compound showed no toxicity. This toxicological profile of three compounds is recapitulated in vitro, both in cultures of PHH and the hiPSC-derived hepatocytes. IC50 values of the three compounds were determined by 24 hour ATP assay (>50, 20, 14µM by PHH; >50, 24, 13µM by iPSC-derived HC) to be nearly identical. This pilot study shows that despite some difference in metabolism (cytochrome P450 and Phase II enzyme levels) between PHH and hiPSC-derived hepatocytes, stem cell-derived hepatocytes may provide an valuable model system for preclinical drug safety study and disease modeling.
HTS for Small Molecules Capable of Modulating the Cellular Phenotypes of Human iPS Cell-derived Hepatocytes
Brian Leonard, F. Hoffmann-La Roche
1:00 pm - 4: 30 pm
Abstract: 503 | Poster: 221
The liver is the major drug metabolizing organ and a frequent site of drug induced toxicity. In drug discovery, robust in vitro methods are needed for modeling liver function. Current methods employing primary human hepatocyte cultures and cell lines have well-documented shortcomings, namely donor to donor variability and functional instability. Although pluripotent stem cell derived tissues hold promise to address these problems, thus far most reports on human induced pluripotent stem cell (hiPSC)-derived hepatocytes indicate a higher similarity to fetal tissues than adult in many aspects including metabolic activity, which could make their extrapolation to the adult situation difficult. To address these shortcomings, human iPSC-derived hepatocytes were exposed to a library of small molecules to identify compounds that resemble functionality of an adult liver. High-throughput, microfluidic qRT-PCR was used to examine the expression of 26 genes that span a spectrum of hepatocyte functions that were either low or exhibited an immature phenotype in hiPSC-derived hepatocytes when compared to adult primary human hepatocytes. During the primary screen, multiple compounds significantly increased maturation-associated genes. These hits were confirmed in a secondary screen, and the ability for a sustained response was quantified. These data suggest it may be possible to modulate the maturation of stem cell-derived tissues, potentially generating even more relevant models for drug discovery and safety platforms.
Identification of Biomarkers of Cardiotoxicity Using Metabolomics of Human Pluripotent Stem Cell-derived Cardiomyocytes
Robert E. Burrier, Stemina Biomarker Discovery, Inc.
1:00 pm - 4: 30 pm
Abstract: 615 | Poster: 360
Cardiotoxicity is one of the most prevalent adverse effects of drugs that afflicts patients and, in the pharmaceutical industry, is a leading cause of drug candidate attrition and the withdrawal of FDA-approved drugs from the market. We have developed an in vitro assay that predicts a drug's propensity to induce cardiotoxicity. To accomplish this goal, several types of cardiac cells, including primary cardiomyocytes, human embryonic stem cell-derived cardiac precursors, and human induced pluripotent stem (hiPS) cell-derived cardiomyocytes, were evaluated. As a result, the hiPS cell-derived cardiomyocytes were selected for further study based on their purity, availability, reproducibility, and ability to function similarly to their in vivo counterparts. Preliminary studies with these cardiomyocytes indicated that a cell density of 50K cells per well in a 96 well plate provides an optimal level of secreted features to be evaluated using mass spectrometry. hiPS cell-derived cardiomyocytes were then exposed to a training set consisting of compounds from different chemical classes that are of known inducers (15 compounds) and noninducers (10 compounds) for cardiotoxicity. Doses for the compounds used for metabolomic analysis were based on known human cardiac exposures and the effects on cell viability measured in culture. The cardiomyocytes were exposed to compounds for 72 hours and the spent culture media was analyzed following treatment using LC-ESI-QTOF mass spectrometry. Univariate and multivariate statistical analysis was used identify a metabolic signature of features that altered in response to cardiotoxicants. These features were then used to test the potential of predictive models capable of classifying cardiotoxic compounds from noncardiotoxins. Our current activities are aimed to understand the small molecules that gave rise to the observed LC/MS features and map them with in metabolic pathways.
Assessment of hERG Channel Function Using Stem Cell-derived Cardiomyocytes
Xiaoyu Zhang, ACEA Biosciences, Inc.
9:00 am - 12: 30 pm
Abstract: 1157 | Poster: 422
The vast majority of drugs withdrawn from the market due to association with TdP arrhythmia interfere with the Ikr repolarization current mediated through the hERG. Consequently, the ICH S7B guidelines recommend that all new chemical entities should be subjected to hERG repolarization assay, typically using cell lines that recombinantly express hERG protein. However, in the last decade it has become evident that not all hERG channel inhibitors result in TdP and not all compounds that induce QT prolongation and TdP necessarily inhibit hERG. In order to better understand and assess the different kinds of drug liabilities associated with hERG channel inhibition and modulation we have used a panel of drugs and compounds which (i) directly bind and inhibit hERG channel function (overt inhibitors); (ii) compounds which inhibit hERG as well as other channels and therefore compensate for the Ikr block (covert inhibitors) and (iii) compounds which interfere with the trafficking of hERG channel protein to the plasma membrane (trafficking inhibitors). We have assessed the activity of these compounds using human induced pluripotent stems cell-derived cardiomyocytes (iCells) together with a system that can measure the beating activity of the spontaneously beating cardiomyocytes. Our data clearly shows that overt hERG channel inhibitors disrupts the periodicity of beating of iCell cardiomyocytes leading to plateau oscillations and a signature waveform that is typical of these class of compounds. Covert hERG channel inhibitors at physiologic concentrations do not appear to affect cardiac function and therefore appear to be safe. hERG trafficking inhibitors display a time-dependent effect on the periodicity of beating that manifests several hours after compound dosing. In summary, the results clearly show that dynamic monitoring of iCell cardiomyocyte beating can be used in a predictive way to assess various types of hERG channel modulators and provide additional information to electrophysiological methods.
Beat Rate Analysis Improves the Prediction of in vivo ECG Alterations by the iPS Cell-derived Human Cardiomyocytes-based Proarrhythmic Model
Liang Guo, F. Hoffmann-La Roche
9:00 am - 12: 30 pm
Abstract: 1168 | Poster: 433
Direct detection of arrhythmic events by the high-speed impedance measurement in human induced pluripotent stem-cell (hiPSC) derived cardiomyocytes has been demonstrated recently as a novel and robust model to identify a drug's proarrhythmic risk in humans. While the prediction of clinical arrhythemogenicity by this model achieved a high degree of accuracy for the set of reference drugs tested, integration of this model into early screening strategy is challenging, since alterations in the ECG waveform, particularly changes in QT interval, not the overt arrhythmia are used as the standard surrogate biomarkers of proarrhythmic risk. Retrospective analysis on the data set of 24 reference drugs revealed a close correlation of changes in the spontaneously beating rate (BR) in this in vitro cardiomyocytes model to abnormal ECGs observed in the clinic. Indeed, a change of 20% or greater in BR at <= 30 µM (BR20) could correctly pinpointed 17 drugs out of 19 that presented abnormal ECGs in the clinic. In the extended validation study that included 27 internal compounds, the analysis of arrhythmic beats failed to identify 5 compounds out of 17 that caused abnormal ECGs in the non-rodent conscious animals, whereas using 30 µM as a cut-off of BR20 could correctly flag all 17 compounds. These results suggest the beat rate change is a more sensitive predictor and more translational biomarker of potential ECG alternations. Hence, including the beat rate analysis adds additional predictive value to this cardiomyocyte arrhythmogenesis model in assessing the likelihood of compounds to cause ECG alterations in the down-stream in vivo studies.
Predictivity of Cardiomyocyte Multielectrode Array (MEA) Assays for in vivo Cardiovascular Outcomes versus Standard Safety Assays: Three Examples
Hong Shi, Bristol-Myers Squibb
9:00 am - 12: 30 pm
Abstract: 1149 | Poster: 414
Using MEAs and cardiomyocytes derived from human induced pluripotent stem cells (hCMs) or rat fetus (rCMs), we evaluated the effects of three internal compounds from different programs whose in vivo cardiovascular outcomes were not predicted by standard in vitro safety assays. hCMs were purchased from CDI, and rCMs were isolated from E18 rats. Spontaneous field potentials (FPs) were recorded on MEAs following an equilibration period. FP beating rate (BR), waveform morphology and conduction time (CT) were analyzed with custom software written in Matlab. The first compound prolonged QRS interval in safety pharmacology (SP) telemetry and EP studies (12% at 0.6 μM free), but was a weak Na channel inhibitor in NaV1.5 patch-clamp (12% inhibition at 30 μM) and rabbit Purkinje fiber (Vmax 30% inhibition at 30 μM) assays. MEA assays with hCMs were more predictive of in vivo findings, with FP conduction delayed by 42, 58 and 79% at 0.3, 1 and 3 μM. A development compound in a second program reduced heart rate in humans at lower exposures than predicted based on SP telemetry studies. Mechanistic studies showed the compound was a direct inhibitor of sinoatrial node automaticity. The compound inhibited BR in hCM MEA assays by 27, 72 and 86% at 3, 10 and 30 μM, respectively, while inhibiting BR in rCMs by only 38% at 30 μM. The MEA assays were more predictive for heart rate slowing, and hCMs were more senstive than rCMs. In a final example, a development candidate prolonged QT interval in monkey SP studies at 2 μM free drug and even lower exposure in humans, despite having weak hERG potency in patch-clamp (10% at 10 μM). However, in MEA assays with hCMs, the compound prolonged FP duration at concentrations similar to in vivo findings. hCMs were more sensitive than rCMs. In summary, in the three cases presented, MEA assays using hCMs or rCMs had excellent predictivity for in vivo cardiovascular outcomes, and were more predictive than standard assays. hCMs were more sensitive than rCMs depending on the mechanism.
Bioenergetics Characterization of Human-induced Pluripotent Stem (iPS) Cell-derived Cardiomyocytes Grown in Different Carbon Sources
Payal Rana, Pfizer, Inc.
9:00 am - 12: 30 pm
Abstract: 1824 | Poster: 110
iCell® Cardiomyocytes (Cellular Dynamics International, Madison, WI) are highly purified human cardiomyocytes derived from induced pluripotent stem (iPS) cells through differentiation and purification protocols. Adult heart prefers fatty acids for energy utilization, however very little is known about the bioenergetics of iCells and their preferred substrate for energy utilization. Cells grown in high glucose are mostly glycolytic because they tend to use glycolysis for energy production despite of having fully functioning mitochondria. Here, we investigated the bioenergetics of iCells grown in different substrate media (high-glucose, low-glucose, galactose, and fatty acids (oleic acid and palmitate)) and their preferred substrate for energy utilization using the XF96 flux analyzer. We investigated effect of mitochondrial modulators (Rotenone, Antimycin, and Oligomycin,) on iCells growing in the different substrate media.. We observed that iCells grown in galactose and fatty acids media were more sensitive to mitochondrial modulators compared to cells grown in high/low glucose media. iCells grown in galactose and fatty acids media displayed higher mitochondrial reserve capacity and maximum respiratory capacity compared to cells cultured in high/low glucose media. Furthermore, complete inhibition of oxidative phosphorylation (OXPHOS) reduced ATP levels in cells growing in galactose and fatty acids suggesting that these cells have limited ability to increase glycolysis to preserve energy levels under these conditions. In contrast, inhibition of OXPHOS did not reduce ATP levels in cells grown in high/low glucose suggesting an increase in glycolytic activity. In summary, cultured iCell cardiomyoctyes can use a diverse set of substrates, including high and low-glucose, galactose and fatty acids. However, only iCells grown in galactose and fatty acids display an aerobic (heart like) phenotype.