Summary Deep learning is a significant step forward for developing autonomous tasks. One of its branches, computer vision, allows image recognition with high accuracy thanks to the use of convolutional neural networks (CNNs). Our goal was to train a CNN with transmitted light microscopy images to distinguish pluripotent stem cells from early differentiating cells. We induced differentiation of mouse embryonic stem cells to epiblast-like cells and took images at several time points from the initial stimulus. We found that the networks can be trained to recognize undifferentiated cells from differentiating cells with an accuracy higher than 99%. Successful prediction started just 20 min after the onset of differentiation. Furthermore, CNNs displayed great performance in several similar pluripotent stem cell (PSC) settings, including mesoderm differentiation in human induced PSCs. Accurate cellular morphology recognition in a simple microscopic set up may have a significant impact on how cell assays are performed in the near future.

Mesenchymal stem/stromal cells (MSCs) obtained from pluripotent stem cells (PSCs) constitute an interesting alternative to classical MSCs in regenerative medicine. Among their many mechanisms of action, MSC extracellular vesicles (EVs) are a potential suitable substitute for MSCs in future cell-free-based therapeutic approaches. Unlike cells, EVs do not elicit acute immune rejection, and they can be produced in large quantities and stored until ready to use. Although the therapeutic potential of MSC EVs has already been proven, a thorough characterization of MSC EVs is lacking. In this work, we used a label-free liquid chromatography tandem mass spectrometry proteomic approach to identify the most abundant proteins in EVs that are secreted from MSCs derived from PSCs (PD-MSCs) and from their parental induced PSCs (iPSCs). Next, we compared both datasets and found that while iPSC EVs enclose proteins that modulate RNA and microRNA stability and protein sorting, PD-MSC EVs are rich in proteins that organize extracellular matrix, regulate locomotion, and influence cell–substrate adhesion. Moreover, compared to their respective cells, iPSCs and iPSC EVs share a greater proportion of proteins, while the PD-MSC proteome appears to be more specific. Correlation and principal component analysis consistently aggregate iPSCs and iPSC EVs but segregate PD-MSC and their EVs. Altogether, these findings suggest that during differentiation, compared with their parental iPSC EVs, PD-MSC EVs acquire a more specific set of proteins; arguably, this difference might confer their therapeutic properties.