Development of a human heart-on-a-chip model using induced pluripotent stem cells, fibroblasts and endothelial cells

Abstract

Abstract Background The use of animal models in cardiovascular research is associated with two serious, intrinsic problems: inaccuracy in the extrapolation of data obtained from animals such as rodents due to different cardiac physiology and animal ethics. Purpose To develop an artificial human heart model for cardiovascular research using organ-on-a-chip technology and human cells Methods Organ chips made of silicone (polydimethylsiloxane) that have two microfluidic channels, a top channel and a bottom channel, separated by a 50-μm thick membrane with 7-μm pores hexagonally packed at 40-μm intervals (Figure 1), were used in this study. We seeded 10,000 human umbilical vein endothelial cells on the membrane surface in the bottom channel to mimic the vasculature. Next, we seeded a mixture of 100,000 human-induced pluripotent stem cells (hiPSCs) and 50,000 human gingival fibroblasts on the membrane surface in the top channel (Figure 1). Human gingival fibroblasts facilitate cardiac differentiation of hiPSCs [1]. We performed the cardiac differentiation of hiPSCs using a previously described protocol [1]. Culture medium was perfused at a constant rate of 60 μl/h to maintain the culture. Results We observed spontaneous contraction of hiPSC-derived cardiomyocytes 20–26 days after the start of the differentiation protocol. Simultaneously, we conducted live intracellular calcium imaging using a fluorogenic calcium-sensitive dye, Cal-520 AM (5 μM in the culture medium). hiPSC-derived cardiomyocytes exhibited a periodic, coordinated pattern of calcium influx synchronised with their contraction under fluorescence microscopy (Figure 2A). Moreover, we found that the β-adrenergic agonist noradrenaline elevated the heart rate of hiPSC-derived cardiomyocytes on the organ chips in a dose-dependent manner (Figures 2A, B). After observing the contraction and intracellular calcium influx of hiPSC-derived cardiomyocytes, we performed immunocytochemistry. Confocal microscopy indicated that the fluorescent signal obtained from anti-cardiac troponin T antibody staining in the top channel exhibited a typical striated pattern with 1.56±0.12-μm interval that reflected sarcomere structure (Figure 2C, yellow). Moreover, the fluorescent signal obtained from anti-CD31 antibody staining in the bottom channel exhibited a typical pattern at the boundary between cells, which is expected at the cell–cell junction of endothelial cells. Conclusion We developed a human heart-on-a-chip model that was confirmed by the functional response to noradrenaline and the histological evidence of sarcomere structure and vasculature, with a capability of live imaging. We expect that this model would be useful for examining the physiological function and for the pharmacological analysis of not only the normal heart but also the heart that reflects specific patient's pathophysiology using patient-derived hiPSCs. Funding Acknowledgement Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Japan Society for the Promotion of Science Figure 1Figure 2