Geometrical Patterning and Constituent Cell Heterogeneity Facilitate Electrical Conduction Disturbances in a Human Induced Pluripotent Stem Cell-Based Platform: An In vitro Disease Model of Atrial Arrhythmias.

Ectopic foci from pulmonary veins (PVs) comprise the main trigger associated with the initiation of atrial fibrillation (AF). An abrupt anatomical narrow-to-wide transition, modeled as in vitro geometrical patterning with similar configuration in the present study, is located at the junction of PVs and the left atrium (LA). Complex cellular composition, i.e., constituent cell heterogeneity, is also observed in PVs and the PVs-LA junction. High frequency triggers accompanied with anatomical irregularity and constituent cell heterogeneity provoke impaired conduction, a prerequisite for AF genesis. However, few experiments investigating the effects of these factors on electrophysiological properties using human-based cardiomyocytes (CMs) with atrial properties have been reported. The aim of the current study was to estimate whether geometrical patterning and constituent cell heterogeneity under high frequency stimuli undergo conduction disturbance utilizing an in vitro two-dimensional (2D) monolayer preparation consisting of atrial-like CMs derived from human induced pluripotent stem cells (hiPSCs) and atrial fibroblasts (Fbs). We induced hiPSCs into atrial-like CMs using a directed cardiac differentiation protocol with the addition of all-trans retinoic acid (ATRA). The atrial-like hiPSC-derived CMs (hiPSC-CMs) and atrial Fbs were transferred in defined ratios (CMs/Fbs: 100%/0% or 70%/30%) on manually fabricated plates with or without geometrical patterning imitating the PVs-LA junction. High frequency field stimulation emulating repetitive ectopic foci originated in PVs were delivered, and the electrical propagation was assessed by optical mapping. We generated high purity CMs with or without the ATRA application. ATRA-treated hiPSC-CMs exhibited significantly higher atrial-specific properties by immunofluorescence staining, gene expression patterns, and optical action potential parameters than those of ATRA-untreated hiPSC-CMs. Electrical stimuli at a higher frequency preferentially induced impaired electrical conduction on atrial-like hiPSC-CMs monolayer preparations with an abrupt geometrical transition than on those with uniform geometry. Additionally, the application of human atrial Fbs to the geometrically patterned atrial-like hiPSC-CMs tended to further deteriorate the integrity of electrical conduction compared with those using the atrial-like hiPSC-CM alone preparations. Thus, geometrical narrow-to-wide patterning under high frequency stimuli preferentially jeopardized electrical conduction within in vitro atrial-like hiPSC-CM monolayers. Constituent cell heterogeneity represented by atrial Fbs also contributed to the further deterioration of conduction stability.

Increased susceptibility to oxidative stress-induced toxicological evaluation by genetically modified nrf2a-deficient zebrafish.

INTRODUCTION: Oxidative stress plays an important role in drug-induced toxicity. Oxidative stress-mediated toxicities can be detected using conventional animal models but their sensitivity is insufficient, and novel models to improve susceptibility to oxidative stress have been researched. In recent years, gene targeting methods in zebrafish have been developed, making it possible to generate homozygous null mutants. In this study, we established zebrafish deficient in the nuclear factor erythroid 2-related factor 2a (nrf2a), a key antioxidant-responsive gene, and its potential to detect oxidative stress-mediated toxicity was examined. METHODS: Nrf2a-deficient zebrafish were generated using the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated 9 technique. The loss of nrf2a function was confirmed by the tolerability to hydrogen peroxide and hydrogen peroxide-induced gene expression profiles being related to antioxidant response element (ARE)-dependent signaling. Subsequently, vulnerability of nrf2a-deficient zebrafish to acetaminophen (APAP)- or doxorubicin (DOX)-induced toxicity was investigated. RESULTS: Nrf2a-deficient zebrafish showed higher mortality than wild type accompanied by less induction of ARE-dependent genes with hydrogen peroxide treatment. Subsequently, this model showed increased severity and incidence of APAP-induced hepatotoxicity or DOX-induced cardiotoxicity than wild type. DISCUSSION: Our results demonstrated that anti-oxidative response might not fully function in this model, and resulted in higher sensitivity to drug-induced oxidative stress. Our data support the usefulness of nrf2a-deficient model as a tool for evaluation of oxidative stress-related toxicity in drug discovery research.

Usefulness of cardiotoxicity assessment using calcium transient in human induced pluripotent stem cell-derived cardiomyocytes.

Monitoring dramatic changes in intracellular calcium ion levels during cardiac contraction and relaxation, known as calcium transient, in human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) would be an attractive strategy for assessing compounds on cardiac contractility. In addition, as arrhythmogenic compounds are known to induce characteristic waveform changes in hiPSC-CMs, it is expected that calcium transient would allow evaluation of not only compound-induced effects on cardiac contractility, but also compound arrhythmogenic potential. Using a combination of calcium transient in hiPSC-CMs and a fast kinetic fluorescence imaging detection system, we examined in this study changes in calcium transient waveforms induced by a series of 17 compounds that include positive/negative inotropic agents as well as cardiac ion channel activators/inhibitors. We found that all positive inotropic compounds induced an increase in peak frequency and/or peak amplitude. The effects of a negative inotropic compound could clearly be detected in the presence of a ß-adrenergic receptor agonist. Furthermore, most arrhythmogenic compounds raised the ratio of peak decay time to peak rise time (D/R ratio) in calcium transient waveforms. Compound concentrations at which these parameters exceeded cutoff values correlated well with systemic exposure levels at which arrhythmias were reported to be evoked. In conclusion, we believe that peak analysis of calcium transient and determination of D/R ratio are reliable methods for assessing compounds' cardiac contractility and arrhythmogenic potential, respectively. Using these approaches would allow selection of compounds with low cardiotoxic potential at the early stage of drug discovery.

Troglitazone Inhibits Bile Acid Amidation: A Possible Risk Factor for Liver Injury.

Troglitazone and pioglitazone were developed as thiazolidinedione-type antidiabetes drugs, but only troglitazone was withdrawn from the markets due to severe liver injury. As both troglitazone and its sulfate metabolite are strong inhibitors of the bile salt export pump (BSEP), troglitazone-induced bile acid (BA) retention is thought to be one of the underlying mechanisms of liver injury. However, pioglitazone is also a strong BSEP inhibitor, indicating other mechanisms may also be involved in troglitazone-induced BA retention. Although retention of hydrophobic BAs (eg, chenodeoxycholic acid [CDCA]: a nonamidated BA) is known to cause hepatocyte injury, little is known about the hepatic conversion of nonamidated, hydrophobic BA species into less toxic hydrophilic BAs (eg, glycochenodeoxycholic acid: amidated BA) as a mechanism of drug-induced liver injury. In this study, we, therefore, investigated the effects of troglitazone and pioglitazone on BA amidation and the role of amidated BAs in troglitazone-associated BA-mediated hepatotoxicity. We also evaluated the intracellular BA composition of human hepatocytes treated with nonamidated BA species (CDCA or deoxycholic acid [DCA]) in the presence of troglitazone or pioglitazone. Amidation of CDCA and DCA was significantly inhibited by troglitazone (IC50: 5 and 3 µmol/l, respectively), but not pioglitazone. Moreover, treatment with troglitazone led to the retention of CDCA and DCA and decrease of glycine-amidation in hepatocytes. From these results, we suggest that troglitazone-induced liver injury might be caused by the accumulation of nonamidated BAs in hepatocytes due to inhibition of BA amidation.

Hepatotoxicants induce cytokine imbalance in response to innate immune system.

In recent years, attention has been paid to innate immune systems as mechanisms to initiate or promote drug-induced liver injury (DILI). Kupffer cells are hepatic resident macrophages and might be involved in the pathogenesis of DILI by release of pro- and anti-inflammatory mediators such as cytokines, chemokines, reactive oxygen species, and/or nitric oxides. The purpose of this study was to investigate alterations in mediator levels induced by hepatotoxic compounds in isolated Kupffer cells and discuss the relation between balance of each cytokine or chemokine and potential of innate immune-mediated DILI. Primary cultured rat Kupffer cells were treated with hepatotoxic (acetaminophen, troglitazone, trovafloxacin) or non-hepatotoxic (pioglitazone, levofloxacin) compounds with or without lipopolysaccharide (LPS). After 24 hr treatment, cell supernatants were collected and various levels of mediators released by Kupffer cells were examined. Although hepatotoxicants had no effect on the LPS-induced tumor necrosis factor-alpha (TNF-α) secretion, they enhanced the release of pro-inflammatory cytokine interleukin-1 beta (IL-1ß) and suppressed the anti-inflammatory cytokines interleukin-6 (IL-6) and interleukin-10 (IL-10) induced by LPS. These cytokine shifts were not associated with switching the phenotypes of M1 and M2 macrophages in Kupffer cells. In conclusion, the present study suggested that the levels of some specific cytokines are affected by DILI-related drugs with LPS stimulation, and imbalance between pro- and anti-inflammatory cytokines, induced by the up-regulation of IL-1ß and the down-regulation of IL-6 or IL-10, plays a key role in innate immune-mediated DILI.

Establishment of a novel experimental protocol for drug-induced seizure liability screening based on a locomotor activity assay in zebrafish.

As drug-induced seizures have severe impact on drug development, evaluating seizure induction potential of candidate drugs at the early stages of drug discovery is important. A novel assay system using zebrafish has attracted interest as a high throughput toxicological in vivo assay system, and we tried to establish an experimental method for drug-induced seizure liability on the basis of locomotor activity in zebrafish. We monitored locomotor activity at high-speed movement (> 20 mm/sec) for 60 min immediately after exposure, and assessed seizure liability potential in some drugs using locomotor activity. However this experimental procedure was not sufficient for predicting seizures because the potential of several drugs with demonstrated seizure potential in mammals was not detected. We, therefore, added other parameters for locomotor activity such as extending exposure time or conducting flashlight stimulation (10 Hz) which is a known seizure induction stimulus, and these additional parameters improved seizure potential detection in some drugs. The validation study using the improved methodology was used to assess 52 commercially available drugs, and the prediction rate was approximately 70%. The experimental protocol established in this present study is considered useful for seizure potential screening during early stages of drug discovery.

Improvement of the evaluation method for teratogenicity using zebrafish embryos.

The zebrafish has been considered as a suitable animal model for drug discovery, especially for evaluation of the teratogenicity, due to their small size, rapid development, transparency, and developmental similarities to mammalian development. These features of zebrafish make it possible to maintain them in culture plates, evaluate the teratogenicity in short term, conduct morphological assessment of each organ without any autopsy operation. The purpose of the present study was to improve an evaluation method for the teratogenicity of test compounds with high throughput ability and prediction rateusing zebrafish embryos. In this study, we established a modified evaluation method as using non-dechorionated embryos and observation a limited number of parameters without grading. Zebrafish embryos were exposed to test compounds from 5-6 to 144 hr post-fertilization, (hpf) corresponding to the organogenesis period. Morphological changes or functional abnormalities induced by test compound treatments were assessed and scored at 11 endpoints, and the potential of teratogenicity was judged based on the score. As a validation assay of the system, the potentials of 59 known teratogenic or non-teratogenic test compounds were evaluated using the present standard zebrafish assay, and the teratogenicity was correctly predicted in 90% (53/59) of all compounds with low false negative and false positive rates. These results indicated that the evaluation method using zebrafish for the teratogenicity we have improved was a valuable tool for early stage screening in drug discovery.

2-({6-[(3R)-3-amino-3-methylpiperidine-1-yl]-1,3-dimethyl-2,4-dioxo-1,2,3,4-tetrahydro-5H-pyrrolo[3,2-d]pyrimidine-5-yl}methyl)-4-fluorobenzonitrile (DSR-12727): a potent, orally active dipeptidyl peptidase IV inhibitor without mechanism-based inactivation of CYP3A.

We report on the identification of 2-({6-[(3R)-3-amino-3-methylpiperidine-1-yl]-1,3-dimethyl-2,4-dioxo-1,2,3,4-tetrahydro-5H-pyrrolo[3,2-d]pyrimidine-5-yl}methyl)-4-fluorobenzonitrile (DSR-12727) (7a) as a potent and orally active DPP-4 inhibitor without mechanism-based inactivation of CYP3A. Compound 7a showed good DPP-4 inhibitory activity (IC(50)=1.1 nM), excellent selectivity against related peptidases and other off-targets, good pharmacokinetic and pharmacodynamic profile, great in vivo efficacy in Zucker-fatty rat, and no safety concerns both in vitro and in vivo.

Usefulness of field potential as a marker of embryonic stem cell-derived cardiomyocytes, and endpoint analysis of embryonic stem cell test.

The embryonic stem cell test (EST) is a validated method and a useful screening tool for drug discovery. EST requires microscopic observation of beating cells to be considered cardiomyocytes as an endpoint assay. However, this procedure is time-consuming and limits the throughput performance. Instead of microscopic observation, we previously established a novel assay method based on cardiac field potential as an endpoint. However, cardiac specificity of this field potential is not yet clarified, because beating cells have not been rigorously evaluated as skeletal or cardiomyocyte. Here, we investigated the relationships between field potential, beating, and cardiac troponin T (cTnT) expression, selected as a cardiomyocyte-specific marker, and evaluated suitability of the field potential as a marker for cardiomyocyte in vehicle or 5-fluorouracil treated embryo bodies. Embryoid bodies of mouse embryonic stem cells (D3) were differentiated in a chamber with multi-electrode array for 5 days, and field potential and beating were measured at the end of differentiation. In addition, these chambers were immunohistochemically stained with anti-cTnT antibody, and the correlation between field potential, beating, and cTnT expression was examined. These results indicated the area of field potential or beating mainly coincided with that of cTnT expression. 5-fluorouracil treatment decreased not only the number of field potential detecting electrodes and beating area, but also cTnT expression, and the area of these parameters was also nearly identical. These results indicate that field potential can be used as a suitable cardiac differentiation marker, and can be a promising parameter of EST.

Improvement of the embryonic stem cell test endpoint analysis by use of field potential detection.

The embryonic stem cell test (EST) is a validated in vitro method to assess the embryotoxic potential of compounds and is a promising tool for drug screening. EST requires microscopic observation of beating cardiomyocytes differentiated from embryonic stem cells as a toxicological endpoint. However, this process is time-consuming and lacks throughput performance. To improve the analysis, we introduced an electrophysiological method with a microelectrode array system for the evaluation of differentiated cardiomyocytes. Embryotoxic (valproic acid, verapamil, and 5-fluorouracil) and non-embryotoxic (penicillin G, d-camphor, and isoniazid) compounds were assessed with the system. Mouse embryonic stem cells were differentiated into cardiomyocytes and treated with each compound during the differentiation process. The embryotoxicity of each compound was then assessed by measuring the field potentials of differentiated cardiomyocytes using the microelectrode array system, as well as by microscopic evaluation. All the embryotoxic compounds dose-dependently inhibited the field potential formation and the myocardial beating of differentiated cells, while the non-embryotoxic compounds did not affect either endpoint. The detection capabilities of the two assay methods were similar. These results indicated that the field potential measurements can be used as an alternative endpoint of EST. Moreover, the field potential can be measured automatically, introducing a high throughput performance compared to the conventional microscopic observation. We therefore concluded that the endpoint analysis with the microelectrode array system improves the original EST and can be useful for the assessment of the embryotoxic potential of compounds.

Comparison of potential risks of lactic acidosis induction by biguanides in rats.

Lactic acidosis has been considered to be a side effect of some biguanides, after phenformin was withdrawn from the market because of its association with lactic acidosis. The potential of lactic acidosis induced by biguanides at human therapeutic exposure levels, however, has not been examined. Then, we compared the risk of lactic acid at doses providing exposure levels comparable to human therapeutic doses. Metformin and phenformin were orally administered to rats for up to 28 days, and plasma drug concentrations and blood lactic acid levels were examined. Metformin did not elevate lactic acid levels at the dose corresponding to higher systemic drug exposure than human therapeutic level, even for repeated doses. In contrast, phenformin elevated lactic acid levels at the dose corresponding to lower exposure than human therapeutic level, and sustained high levels were observed up to 24h post-dose; furthermore, these changes were enhanced by repeated doses. Direct comparison at each rat equivalent dose clearly indicated that lactic acid levels of phenformin were higher than those of metformin. These non-clinical findings suggest that metformin dose not increase lactic acid levels like phenformin does, and therefore may not increase the risk for lactic acidosis at human therapeutic exposure level.