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

The differentiation of pluripotent stem cells (PSCs) 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 variabilities, which severely impede the progress of scientific research and the manufacturing of cell products. For instance, PSC-to-cardiomyocyte (CM) differentiation is vulnerable 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), PSC 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 PSC 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 that can further improve the cell resistance to the overdose of CHIR. Together, this study indicates that artificial intelligence is able to guide and iteratively optimize PSC 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.

Single-Molecule Peptide Identification Using Fluorescence Blinking Fingerprints

The ability to identify peptides with single-molecule sensitivity would lead to next-generation proteomics methods for basic research and clinical applications. Existing single-molecule peptide sequencing methods can read some amino acid sequences, but they are limited in their ability to distinguish between similar amino acids or post-translational modifications. Here, we demonstrate that the fluorescence intermittency of a peptide labeled with a spontaneously blinking fluorophore contains information about the structure of the peptide. Using a deep learning algorithm, this single-molecule blinking pattern can be used to identify the peptide. This method can distinguish between peptides with different sequences, peptides with the same sequence but different phosphorylation patterns, and even peptides that differ only by the presence of epimerized residues. This study builds the foundation for a targeted proteomics method with single-molecule sensitivity.

Dual-Activatable Cell Tracker for Controlled and Prolonged Single-Cell Labeling

Cell trackers are fluorescent chemical tools that facilitate imaging and tracking cells within live organisms. Despite their versatility, these dyes lack specificity, tend to leak outside of the cell, and stain neighboring cells. Here, we report a dual-activatable cell tracker for increased spatial and temporal staining control, especially for single-cell tracking. This probe overcomes the typical problems of current cell trackers: off-target staining, high background signal, and leakage from the intracellular medium. Staining with this dye is not cytotoxic, and it can be used in sensitive primary cells. Moreover, this dye is resistant to harsh fixation and permeabilization conditions and allows for multiwavelength studies with confocal microscopy and fluorescence-activated cell sorting. Using this cell tracker, we performed in vivo homing experiments in mice with primary splenocytes and tracked a single cell in a heterogeneous, multicellular culture environment for over 20 h. These experiments, in addition to comparative proliferation studies with other cell trackers, demonstrated that the signal from this dye is retained in cells for over 72 h after photoactivation. We envision that this type of probes will facilitate the analysis of single-cell behavior and migration in cell culture and in vivo experiments.