Kiyomi Kikuchi

Generation of metabolically functioning hepatocytes from human pluripotent stem cells by FOXA2 and HNF1α transduction

BACKGROUND & AIMS Hepatocyte-like cells differentiated from human embryonic stem cells (hESCs) and induced pluripotent stem cells (hiPSCs) can be utilized as a tool for screening for hepatotoxicity in the early phase of pharmaceutical development. We have recently reported that hepatic differentiation is promoted by sequential transduction of SOX17, HEX, and HNF4α into hESC- or hiPSC-derived cells, but further maturation of hepatocyte-like cells is required for widespread use of drug screening. METHODS To screen for hepatic differentiation-promoting factors, we tested the seven candidate genes related to liver development. RESULTS The combination of two transcription factors, FOXA2 and HNF1α, promoted efficient hepatic differentiation from hESCs and hiPSCs. The expression profile of hepatocyte-related genes (such as genes encoding cytochrome P450 enzymes, conjugating enzymes, hepatic transporters, and hepatic nuclear receptors) achieved with FOXA2 and HNF1α transduction was comparable to that obtained in primary human hepatocytes. The hepatocyte-like cells generated by FOXA2 and HNF1α transduction exerted various hepatocyte functions including albumin and urea secretion, and the uptake of indocyanine green and low density lipoprotein. Moreover, these cells had the capacity to metabolize all nine tested drugs and were successfully employed to evaluate drug-induced cytotoxicity. CONCLUSIONS Our method employing the transduction of FOXA2 and HNF1α represents a useful tool for the efficient generation of metabolically functional hepatocytes from hESCs and hiPSCs, and the screening of drug-induced cytotoxicity.

A new paradigm for drug-induced torsadogenic risk assessment using human iPS cell-derived cardiomyocytes

INTRODUCTION Human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) are anticipated to be a useful tool for conducting proarrhythmia risk assessments of drug candidates. However, a torsadogenic risk prediction paradigm using hiPSC-CMs has not yet been fully established. METHODS Extracellular field potentials (FPs) were recorded from hiPSC-CMs using the multi-electrode array (MEA) system. The effects on FPs were evaluated with 60 drugs, including 57 with various clinical torsadogenic risks. Actual drug concentrations in medium were measured using the equilibrium dialysis method with a Rapid Equilibrium Dialysis device. Relative torsade de pointes (TdP) scores were determined for each drug according to the degree of FP duration prolongation and early afterdepolarization occurrence. The margins were calculated from the free concentration in medium and free effective therapeutic plasma concentration. Each drugs results were plotted on a two-dimensional map of relative TdP risk scores versus margins. RESULTS Each drug was categorised as high, intermediate, or low risk based on its location within predefined areas of the two-dimensional map. We categorised 19 drugs as high risk; 18 as intermediate risk; and 17 as low risk. We examined the concordance between our categorisation of high and low risk drugs against the torsadogenic risk categorisation in CredibleMeds®. Our system demonstrated high concordance, as reflected in a sensitivity of 81%, specificity of 87%, and accuracy of 83%. DISCUSSION These results indicate that our torsadogenic risk assessment is reliable and has a potential to replace the hERG assay for torsadogenic risk prediction, however, this system needs to be improved for the accurate of prediction of clinical TdP risk. Here, we propose a novel drug induced torsadogenic risk categorising system using hiPSC-CMs and the MEA system.

1,25-dihydroxyvitamin D3 promotes vitamin K2 metabolism in human osteoblasts

Abstract. It has been reported that vitamin K2 (menaquinone-4) promoted 1,25-dihydroxyvitamin D3 (1,25(OH)2D3)-induced mineralization and enhanced γ-carboxyglutamic acid (Gla)-containing osteocalcin accumulation in cultured human osteoblasts. In the present study, we investigated whether menaquinone-4 (MK-4) was metabolized in human osteoblasts to act as a cofactor of γ-glutamyl carboxylase. Both conversions of MK-4 to MK-4 2,3-epoxide (epoxide) and epoxide to MK-4 were observed in cell extracts of cultured human osteoblasts. The effect of 1,25(OH)2D3 and warfarin on the vitamin K cycle to cultured osteoblasts were examined. With the addition of 1 nM 1,25(OH)2D3 or 25 μM warfarin in cultured osteoblasts, the yield of epoxide from MK-4 increased. However, the conversion of epoxide to MK-4 was strongly inhibited by the addition of warfarin (2.5–25 μM), whereas it was almost not inhibited by 1,25(OH)2D3 (0.1–10 nM). To clarify the mechanism for this phenomenon, a cell-free assay system was studied. Osteoblast microsomes were incubated with 10 μM epoxide in the presence or absence of warfarin and 1,25(OH)2D3. Epoxide reductase, one of the enzymes in the vitamin K cycle was strongly inhibited by warfarin (2.5–25 μM), whereas it was not affected by 1,25(OH)2D3 (0.1–1 nM). Moreover, there was no effect of pretreatment of osteoblasts with 1 nM 1,25(OH)2D3 on the activity of epoxide reductase. However, the activity of epoxidase, that is the γ-glutamyl carboxylase was induced by the pretreatment of osteoblasts with 1 nM 1,25(OH)2D3. In the present study, it was demonstrated that the vitamin K metabolic cycle functions in human osteoblasts as well as in the liver, the post-translational mechanism, by which 1,25(OH)2D3 caused mineralization in cooperation with vitamin K2 was clarified.

The Effect of Divalent Cations on the Membrane Properties and Pharmacokinetics in Rat of the Lipid A Analogue E5531

To obtain information on the effects of Ca2+ on the membrane properties of the lipid A analogue E5531, we have determined the aggregate size, zeta potential, membrane fluidity, micropolarity and permeability of the E5531 membrane as a function of Ca2+ levels.

Quantitative determination of a potent lipopolysaccharide antagonist, E5564, in rat and dog plasma by high-performance liquid chromatography with fluorescence detection

The assay method was established for the quantification of a potent lipopolysaccharide (LPS) antagonist, E5564, in rat and dog plasma using HPLC. E5564 and the I.S. (an analogue of E5564) were extracted and derivatized with 9-Anthryldiazomethane (ADAM reagent) to be given fluorescence. LC-MS analysis indicated that single molecule of E5564 was coupled with two molecules of ADAM reagent at one on each of the phosphorus groups. After solid-phase extraction, ADAM derivatives of E5564 and the I.S. were separated on an ODS column using methanol/ethanol containing sodium acetate as a mobile phase at 1.2 ml/min (gradient elution), and detected by a fluorescence detector (excitation: 254 nm, emission: 415 nm). The intra-day and inter-day precision were less than 14.4%, and accuracy were within +/-13.0% in the concentration range of 30 to 20,000 ng/ml plasma in both species. E5564 was stable for at least 13 days in rat and dog plasma at -20 degrees C, and the processed sample was stable for up to 14 days at 4 degrees C. This validated method was successfully applied to the evaluation of the pharmacokinetics of E5564 in rats and dogs after single bolus intravenous doses.

Control of the dispersing process and pharmacokinetics in rats for lipid A analogue, E5531.

E5531 is a synthetic disaccharide analogue of lipid A which has a low toxicity but retains the ability to reduce production of tumour necrosis factor. This analogue has potential for use in the treatment of septic shock. An injectable formulation of E5531 would be useful, but dispersion in aqueous solution is a problem. In the present study the dispersing process for E5531 was evaluated using the pH‐jump method (pH 11·0·7·3). The size of the aggregates was decreased (reaching 20 nm) with increasing dispersing time in 0·003 M NaOH (pH 11·0). The membrane fluidity of the aggregates increased with increasing dispersing time. When prepared by the normal dilution method (pH 7·3·7·3), the size of the aggregates remained constant at 140 nm and the membrane fluidity was smaller than that of samples prepared by the pH‐jump method. This indicates that during dispersing at basic pH, the hydration proceeded in a normal manner, but then stopped, just after adjustment of the pH to 7·3. This suggests that the degree of hydration of the membrane is dependent on the dispersing time at pH 11·0. Using samples with different degrees of hydration and different membrane fluidity prepared by the pH‐jump method, the pharmacokinetics and stability of the aggregates were evaluated after intravenous injection into rats. The data thus obtained confirmed that the membrane fluidity was correlated with the pharmacokinetics and stability in rat plasma.

The Effect of the Membrane Fluidity on Pharmacokinetics for Lipid A Analog E5531

The effect of the dispersing procedure on the aggregate size, membrane fluidity and the pharmacokinetics were evaluated for the lipid A analog E5531. The size of the aggregates prepared by the pH-jump method (pH 11.0 → 7.3) was decreased, reaching 20 nm with increasing dispersing time in 0.003 N NaOH (pH 11.0). The membrane fluidity of the aggregates increased with increasing dispersing time. When prepared by the normal dilution method (pH 7.3 → 7.3), the size of the aggregates remained constant at 150 nm and the membrane fluidity was smaller compared to samples prepared by the pH-jump method. Using samples with different degrees of hydration and different membrane fluidities prepared by the pH-jump method, the pharmacokinetics after intravenous administration into rats were evaluated, and the data obtained confirmed that the membrane fluidity was correlated with the pharmacokinetics in rat. In addition, E5531 vials were stable for 24 months at room temperature when used within 24 hr after reconstitution.