Early embryonic development is often described as a carefully choreographed process of growth and differentiation, cells dividing, migrating, and specializing in precise sequence. But a new study suggests that hidden beneath this orderly surface lies a more strategic logic: individual cells make decisions about whether to kill their neighbors based on the embryo’s size relative to a viability threshold.
The finding, published June 13 in Nature Communications, emerges from a collaboration between the University of British Columbia, the University of Toronto, and the University of Pittsburgh. The team combined evolutionary game theory with human stem-cell-derived embryo models to reveal competitive cell behaviors that traditional imaging methods have missed.
The game framework
The researchers framed early embryonic development as a collective risk game, a class of evolutionary game theory models used to study public goods dilemmas. Each cell is treated as a player with two strategic options: cooperate (refrain from killing, which benefits the collective) or defect (kill neighboring cells, which is costly to the killer but may benefit the embryo as a whole).
The model predicts when competitive cell killing becomes a rational strategy. The key variable is embryo size: when the developing structure exceeds a viable target size, the probability of survival drops unless some cells are eliminated. At that point, individual cells that kill their neighbors are, paradoxically, acting in the collective interest, reducing the embryo to a size that can sustain healthy development.
The experimental system
To test the model, the team used heX-embryoids, human induced pluripotent stem cell (hiPSC)-derived embryo models developed by co-authors Joshua Hislop and Mo Ebrahimkhani (originally published in Nature, 2023). These structures recapitulate the bilaminar disc-like interface of the epiblast and hypoblast found in early human embryos, including amniotic and yolk sac cavities.
The high-throughput nature of the system allowed the team to generate hundreds of embryoids and track how variability in size correlated with cell behavior and developmental outcomes.
When they experimentally inhibited cell death, the embryoids grew larger, but failed more often. The structures that normally would have been trimmed by competitive killing became oversized and less viable, directly linking the killing behavior to embryo-level success.
Hidden behaviors
The game-theoretic approach revealed something invisible to standard microscopy. Traditional imaging captures static snapshots, which cells are where, which markers they express, which ones die. It can show that cell death occurs, but not why or under what strategic conditions.
The game model, by contrast, predicts the decision-making logic: cells that kill are not malfunctioning. They are responding to information about the collective state, the embryo’s size, and adjusting their behavior accordingly. The authors describe this as “responsive behavior” that had been hidden because it is dynamic, conditional, and collective: it only becomes visible when you frame individual cell actions in the context of the embryo-level outcome.
The mechanism by which individual cells sense the embryo’s overall size remains under investigation. The finding establishes that cells do respond to collective state, but the sensory pathways involved are not yet identified.
Caveats and context
The heX-embryoids are stem-cell-derived models, not actual human embryos. They capture key features of peri-implantation development (approximately days 8 to 14 post-fertilization) but may miss in vivo complexity, the maternal interface, nutrient gradients, and mechanical forces present in a developing uterus.
The computational predictions of the evolutionary game theory model also require further experimental validation across more conditions and species. The current study provides strong correlational and perturbational evidence linking cell killing to embryo size control, but the formal relationship between the game-theoretic predictions and the biological mechanism will need refinement.
The authors have declared competing interests: Hislop and Ebrahimkhani have filed a patent for the heX-embryoid technology and have formed a startup company to commercialize it.
What it means
The study adds to a growing picture of development not as a deterministic program but as a strategic system in which cells make conditional decisions based on local and global information. The “game” framing, familiar from economics and evolutionary biology, provides a formal language for describing these decisions that static imaging cannot.
For developmental biology, the approach offers a new tool: instead of asking “what happens during development,” researchers can ask “what would a rational cell do, given its circumstances?”, and then test whether real cells behave that way. For regenerative medicine, where controlling the size and composition of stem-cell-derived structures is a practical challenge, understanding the strategic logic of cell competition could help improve the yield and quality of engineered tissues.
Source: Abou Chakra, M., Hislop, J., Egilmez, I., Alkalai, R., Bashth, O., Maheden, K., Gundagathi, A., Ebrahimkhani, M.R., & Shakiba, N. (2026). The embryo game uncovers hidden cell behaviors. Nature Communications. DOI: 10.1038/s41467-026-74074-0
Disclosure: Co-authors Hislop and Ebrahimkhani have filed a patent for the heX-embryoid technology and are associated with a startup company commercializing it.