How cell shape and behavior reveal future cell identity
During development, cells gradually acquire their final identities through a series of coordinated changes. Traditionally, these cell fate decisions have been studied by analyzing gene expression profiles using advanced sequencing technologies, which provide molecular snapshots of cells at discrete time points. However, a new study published in Molecular Systems Biology by Associate Professor Jakub Sedzinski’s group at reNEW Copenhagen, and led by PhD student Mari Tolonen, explores a different question: can the physical behavior of cells reveal what they are going to become?
“Having observed how dynamic developing cells are under the microscope, I wanted to design a single-cell study of cell fate by measuring each cell’s shape, movement, and spatial context, rather than relying solely on gene expression,”
To address this question, the team used the frog embryo mucociliary epithelium, an easily accessible developing tissue, as a model system. They established a live-imaging and computational framework capable of tracking thousands of individual cells over time as they differentiated within the tissue. From these long-term recordings, the researchers extracted a large data set of morphodynamic features, including information about cell shape, movement, and spatial position. They then used supervised machine-learning approaches to determine whether these features could predict the future identities of individual cells. To achieve this, the models were trained using cells whose final fates were known through backtracking from the end of the recordings.
“Most studies of cell fate focus on molecular signatures. We wanted to ask whether the dynamic physical behaviour of cells – how they change shape, move, and interact with their neighbours – also contain information about their future identity,”
The analyses revealed that cell shape and behaviour indeed contain predictive information about future cell identity. However, this information does not appear as an early deterministic signal. Instead, predictive power gradually increases as cells progress through differentiation, suggesting that cell fate commitment is a continuous process that emerges over time rather than a fixed decision made at an early stage.
The study further shows that predictive information is distributed across multiple aspects of cell behaviour and tissue organisation. Cell morphology, movement patterns, and local spatial context all contribute to distinguishing future cell identities. Together, these findings highlight that development cannot be fully understood from molecular information alone and that the physical dynamics of cells provide an important complementary layer of information.
“As we are now able to capture the dynamic behaviours of cells within their native tissue environment at the single-cell level, this work provides a foundation for integrating physical and molecular descriptions of development. Combining morphodynamic analyses with gene expression data may ultimately provide a more complete understanding of how cell fate decisions unfold in living tissues,” Mari Tolonen concludes.
The significance of the study was further recognised by the journal Molecular Systems Biology, which selected an illustration inspired by the work as the cover image for the July 2026 issue.
Read the article here:
Single-cell morphodynamics predict cell fate decisions during mucociliary epithelial differentiation
Cover image illustration: Anna Wiemer @wmr_illustrations.