Cheap and safe human pluripotent stem cell culture

How a screening of small molecules and analyses of their effects led to the development of a growth–factor-free culture system for human pluripotent stem cells by controlling specific signaling pathways.

Like Comment

Research into generating and expanding induced pluripotent stem cells (iPSCs) has been the focus of considerable efforts in regenerative medicine in investigations of the causes of intractable diseases. However, the media used for cultivating human pluripotent stem cells (hPSCs)—especially xeno-free medium—is costly, constituting a large part of funds for research and development. It is therefore important to develop cheaper media so that regenerative medicine can eventually benefit large numbers of patients.


In our paper published in Nature Biomedical Engineering, we report the AKIT medium for hPSC culture, which is about one-tenth cheaper than conventional media used for this application due to the replacement of protein components with small molecules.


My supervisor, Kouichi, investigates the mechanisms of hPSC self-renewal and differentiation, and previously found that canonical Wnt signaling, specifically Wnt3a, in combination with ID-8, a chemical inhibitor of dual-specificity tyrosine phosphorylation-regulated kinase (DYRK), can support hPSC self-renewal and replace FGF and TGF-ß in the culture medium.1 However, recombinant Wnt protein is expensive.  He therefore went on to hypothesize that the combination of ID-8 and a small molecule activator of canonical Wnt signaling could maintain the pluripotency and support self-renewal of hPSCs. He tried several GSK3 inhibitors, which activate Wnt signaling, but both CHIR99021 (a well known and highly specific GSK3 inhibitor) and BIO (which was reported as supporting hPSC self-renewal) were unsuccessful in combination with ID-8. The results were disappointing, and the project was suspended.


Years after these events, my predecessor Shin-ya restarted the project and performed a chemical screening based on the same hypothesis. The correctness of the hypothesis was evidenced by the combination of ID-8 and 1-azakenpaullone (AKI medium). However, although our AKI medium could maintain hPSC pluripotency, the cell expansion rate was very low. Because our objective was aimed at the development of a practical medium for large-scale stem cell culture, we had to improve the expansion rate. Hosein, another predecessor of mine, screened for potential targets of DYRK and Wnt signaling pathways to find additional factors that may be involved in proliferation. He found that calcinuerin/NFAT signaling is activated by the AKI ‘slow cell growth medium’ but not E8 or other ‘rapid cell growth medium’. Finally, we found that the addition of tacrolimus, a calcinuerin/NFAT inhibitor, accelerated cell expansion.


The combination of AKI and tacrolimus (AKIT) could maintain the pluripotency of hPSCs and expand them in a similar way to any other culture media. But we also wanted to know the proliferation and differentiation potential of hPSCs cultured in AKIT. hPSCs in the AKIT culture system formed flat, tightly packed colonies that were similar to those seen in the conventional on-feeder culture system. However, our hPSC colonies were flatter and had a lower cell density when grown using the E8 feeder-free system.


At this point I joined this research project and engaged in the characterization of hPSCs in the AKIT culture system. I conducted comprehensive RNA-seq gene expression analyses comparing the embryonic stem cell line KhES-1 propagated in AKIT, E8 and conventional on-feeder cultures with knockout serum replacement (KSR) and bFGF medium. The global gene expression data indicated that the cells cultured in the AKIT system were different from those cultured in either E8 or KSR/bFGF media. In particular, the expression levels of factors in the tricarboxylic acid cycle and oxidative phosphorylation pathways were elevated in the AKIT culture conditions. Moreover, the cells in the AKIT culture system formed undifferentiated colonies in the presence of the glycolysis inhibitor 2-deoxy-D-glucose. This phenotype was very similar to that of cells in the naive state. Therefore, we expected that oxygen consumption rate in cells cultured in the AKIT system would be higher than in the E8 feeder-free system. However, unlike what is observed in naive cells, the oxygen consumption rate for the cells cultured in AKIT was not significantly different from the cells in E8. If we resolve this discrepancy between activation of the tricarboxylic acid cycles and absence of increase in the oxygen consumption rate, the expansion rate may well be higher. In fact, we are currently approaching a solution to this problem, and believe we can provide the most scalable cost-effective hPSC culture system in the near future.


Our paper: Shin-ya et al. Chemically defined and growth-factor-free culture system for the expansion and derivation of human pluripotent stem cells. Nat. Biomed. Eng. doi:10.1038/s41551-018-0200-7 (2018).

Reference: 1) Hasegawa, K. et al. Wnt signaling orchestration with a small molecule DYRK inhibitor provides long-term xeno-free human pluripotent cell expansion. Stem Cells Transl. Med. 1, 18–28 (2012).


Ikeda Tatsuhiko

Researcher, Kyoto University