A live-cell image-based machine learning strategy for reducing variability in iPSC differentiation systems

The differentiation of induced pluripotent stem cells (iPSCs) into diverse functional cell types provides a promising solution to support drug discovery, disease modeling, and regenerative medicine. However, functional cell differentiation is currently limited by the substantial line-to-line and batch-to-batch variability, which severely impede the progress of scientific research and the manufacturing of cell products. For instance, iPSC-to-cardiomyocyte (CM) differentiation is vulnerable with great sensitivity to inappropriate doses of CHIR99021 (CHIR) that are applied in the initial stage of mesoderm differentiation. Here, by harnessing live-cell bright-field imaging and machine learning (ML), we realize real-time cell recognition in the entire differentiation process, e.g., CMs, cardiac progenitor cells (CPCs), iPSC clones, and even misdifferentiated cells. This enables non-invasive prediction of differentiation efficiency, purification of ML-recognized CMs and CPCs for reducing cell contamination, early assessment of the CHIR dose for correcting the misdifferentiation trajectory, and evaluation of initial iPSC colonies for controlling the start point of differentiation, all of which provide a more invulnerable differentiation method with resistance to variability. Moreover, with the established ML models as a readout for the chemical screen, we identify a CDK8 inhibitor to further improve the cell resistance to the overdose of CHIR. Together, this study indicates that artificial intelligence is able to guide and iteratively optimize iPSC differentiation to achieve consistently high efficiency across cell lines and batches, providing a better understanding and rational modulation of the differentiation process for functional cell manufacturing in biomedical applications. With the assistance of RNA sequencing, we transcriptomically characterize the image-recognized CPCs (IR-CPCs), the cells treated with different CHIR doses at stage I, iPSC-CM treated with or without BI-1347, and cells treated with or without BI-1347 at stage I. 1. A total of 12 samples of IR-CPC, non-CPC, CM, and iPSC (including three biological repetitions) were collected and analyzed together, in which IR-CPCs were purified through DACT-1 photoactivation method; non-CPC were the cells under inappropriate CHIR conditions on day 6. 2. A total of 10 samples of cells under different CHIR doses (iPSC; CHIR 2 μM 48h, 6 μM 24h, 6 μM 36h, 10 μM 24h, 8 μM 36h, 6 μM 48h, 12 μM 24h, 12 μM 36h, and 10 μM 48h) were collected before withdrawing CHIR. 3. A total of 9 samples of iPSC-CM with BI-1347, iPSC-CM without BI-1347, and iPSC (including three biological repetitions) were collected and analyzed together. iPSC-CM with BI-1347 (induced by CHIR 12 uM 48h and BI-1347 0.5 uM 48h) and iPSC-CM without BI-1347 (induced by CHIR 8 uM 48h) were collected after 12 days of cardiac differentiation. 4. A total of 14 samples were collected and analyzed for exploring the mechanism of BI-1347 (including two biological repetitions), including two iPSC samples and 6 sample pairs from 2 groups, in which CM was induced in the presence or absence of BI-1347 (0.5 μM, 48h) under the overdose CHIR (16 μM, 48h). Guided by this experimental design, cells were collected at three time points (24h, 48h, and 72h).