Alexandre Kabla – Herding Cells
Animals, including us, are most often studied as individuals. However, looking at them down a microscope, they can also be seen as societies of cells, surpassing the complexity of beehives or ant colonies, and far more harmonious than human societies.
This high level of cooperation between cells is surprising considering that nearly all of them are doomed to die without offspring, with the sole function of helping a handful of them to pass their common genetic material on to the next generation. Cancer occasionally gives them back the freedom to proliferate and run away, if only this insurrection was not so short-lived. Unicellular organisms might have reasons to laugh at the unfortunate evolutionary pathway taken by their distant cousins.
The theory of evolution explains well why such a level of cooperation is possible and indeed likely to appear at all. But a lot of mystery remains about how organisms develop from a single fertilised egg, cell division after cell division. Molecular biology and genetics have provided remarkable insights into the sub-cellular aspects of development. Yet we do not understand how the whole process can work so reliably, how thousands of billions of cells end up most of the time at the right location, performing the right function. Looking at the molecular or genetic level of development only provides limited information, as Sir David Smithers wrote in The Lancet in 1962: ‘Cancer is no more a disease of cells than a traffic jam is a disease of cars. A lifetime of study of the internal combustion engine would not help anyone understand our traffic problems.’ To progress with both cancer and developmental biology, a higher level of description is required, involving cell behaviours and interactions. This is where physics can help.
The science of self-organisation stems from the principle that order and complexity can emerge from simple forms of interactions combined with a bit of noise. A remarkable example is the occurrence of phase transitions, whereby a small decrease in temperature can cause atoms of a material to arrange in a crystalline manner over very large distances. This process is remarkably well captured by principles of statistical physics. The same type of approach can be used to explain structures ranging from soap films and proteins to physical interactions within groups of animals. Over the last 20 years, the minimal rules needed for collective behaviours in animal populations have been narrowed down to three simple interactions: short-range repulsion (animals do not like to move on top of each other), long-range attraction (nor do they like to be isolated from the group) and alignment (animals can use their sensory inputs, such as vision, to adjust their direction of migration according to that of their neighbours). This replicates the dynamics of starling flocks, sardine schools and bison herds. A distinctive feature of these migratory patterns is that they are leaderless, free from any internal hierarchy.
In our research, we explore the use of these principles to understand the behaviour of cells within organisms, and in particular during embryo development where cells are not yet highly specialised and remain highly motile. Although cells do not have eyes and brains to align their trajectories, it is now well established that temporal persistence of cell orientation and cellular adhesion can lead to flocking-like behaviours and coordinated cell movements. During embryo development, mechanics and geometry play critical roles. One has to understand how different tissues constrain the motion of cells, and how these constraints in turn influence the result and robustness of developmental processes.
Developing such research requires multi-disciplinary teamwork, involving collaborations across different research groups interested in the same question but bringing different perspectives. We develop instruments and models in our group. These are to manipulate embryos, to image their cells in vivo and analyse their behaviours, but also to study these questions in more indirect and abstract ways, using in vitro and in silico experimentation. For instance, one can culture a particular cell type and study how it behaves confined to specific geometries, or under a known mechanical constraint. This enables us to isolate particular aspects of environmental interactions that would be impossible in vivo.
Computer models are tremendously helpful in identifying the simplest principles that account for complex behaviours in populations of living cells. Elementary rules can be implemented in idealised systems to test if they are sufficient or necessary to account for complex behaviours. Although reductionist approaches rarely capture the details of living processes, they nevertheless strongly contribute to their understanding and efficiently guide experimental design and analysis. Because cell migration plays a fundamental role in both normal development and cancer cell metastasis, the scientific community is currently exploring the analogy between the two processes in the hope of developing novel therapies.
Simulated segregation patterns emerging from a mixture of two cell populations of different motile properties (J. R. Soc. Interface, 9:3268–78, 2012). Green/blue cells are motile, yellow/red cells are non-motile.