Background

Human induced pluripotent stem cell-derived Cardiomyocytes (hiPSC-CMs) are expected to become a cell source for applications in regenerative medicine of heart diseases and drug discovery. Heart diseases remain the leading cause of death worldwide (WHO, 2018). In non-human primates, transplanted hiPSC-CMs improve cardiac contraction function and are sufficient to regenerate the infarcted heart (Shiba et al., 2016). Previous studies, represented by the CiPA Project (Blinova et al., 2017) and the JiCSA Project (Yamazaki et al., 2018), have shown the utility of hiPSC-CMs for prediction of the risk of cardiotoxicity, like Torsades de Pointes. It was also shown that maturity of the generated hiPSC-CMs alters the response to drugs (da Rocha et al., 2017). Thus, the development of a high-throughput and robust protocol for generating high-quality hiPSC-CMs is urgent for the realization of regenerative medicine and drug discovery.

hiPSCs could be directly differentiated into cardiomyocytes in either 2D adhesion culture or 3D cell aggregates/spheroid culture methods. In particular, the spheroid culture method is easier to scale-up and better at promoting maturation of the generated hiPSC-CMs, compared with the 2D culture method (Pal et al., 2013). In the cardiac differentiation process, regulation of canonical and non-canonical Wnt at the right time strongly promotes pluripotent stem cell differentiation into cardiomyocytes, not other mesodermal cell types (Loh et al., 2016). In addition, expression levels of the Wnt signal-related genes Wnt5A and Wnt11 are affected by spheroid size (Hwang et al., 2009, Chen et al., 2015). Therefore, we sought to develop a highly efficient and robust cardiac differentiation protocol by achieving high-throughput spheroid formation with simultaneous active spheroid size control.

In this study, we used the uniquely shaped microfabric dishes, “EZSPHERE,” designed for high-throughput production of uniform spheroids on one culture dish, with its high-density micro-wells (Sato et al., 2016). Initially, a basic spheroid culture protocol, optimized for cardiac differentiation (Yang et al., 2008, Bauwens et al., 2014, Kempf et al., 2014), was combined with the EZSPHERE dishes. For spheroid formation from the hiPSCs, Type 900 EZSPHERE dishes with a diameter of 100 mm were used. Each of them had approximately 14,000 individual micro-wells with regular size (400-500 μm in diameter and approximately 100 μm in depth). On calculation, approximately 14,000 spheroids could be formed in each EZSPHERE dish, and their initial seeding cell number per spheroid ranged 200-220. As per standard protocol, activin A, BMP-4, and FGF2 were added to the spheroids in the EZSPHERE to induce differentiation into Primitive Streak on day 1. On day 4, the spheroids were collected, washed, and transferred onto a flat bottom low cell adhesion dish, followed by treatment with Wnt inhibitor. On day 6, the size uniformity of the spheroids was lost, but 62% of the cell population had differentiated into PDGFR-α/KDR double-positive cardiac mesoderm-like cells (Kattman et al., 2011; Birket et al., 2015). To control spheroid size once more, the differentiated spheroids were collected and dissociated into single cells. Then, the single cells obtained were re-aggregated by re-seeding onto the dishes of Type 903 EZSPHERE (with a diameter of 35 mm), which have approximately 1,000 individual micro-wells with larger size (approximately 800 μm diameter and 400 μm depth) in each dish. The re-aggregated spheroids increased the number of cardiac Troroponin T (CTNT) positive cells faster than that of the non-reaggregated spheroids. In addition, the cardiac maturity (estimated by the gene expression level of MYL2 per MYL7) was much higher in the re-aggregated spheroids compared with that of the spheroids that did not undergo the re-aggregation process. Performing the re-aggregation process also raised the expression levels of the non-canonical Wnt signals Wnt5A and Wnt11. In mice, Wnt5A and Wnt11 are essential for the generation of second heart field progenitor cells (Cohen et al., 2012). Wnt11 is also known for its role in generating myocardial electrical gradient patterns through the regulation of non-canonical Wnt/Ca²⁺ signaling during zebrafish heart development (Panakova et al., 2010). To reveal the detail of its molecular mechanism, further study is required. However, these data strongly suggest that the cardiomyocyte differentiation process performed on EZSPHERE, including a re-aggregation step for the induced cardiac mesoderm/progenitor, is effective in producing cardiomyocytes with high purity and maturation levels.