A unicellular walker embodies a finite state machine [article]

Benjamin T Larson, Jack Garbus, Jordan B Pollack, Wallace F Marshall
2021 bioRxiv   pre-print
Cells are complex biochemical systems whose behavior emerges from interactions among myriad molecular components. The idea that cells execute computational processes is often invoked as a general framework for understanding cellular complexity. However, the manner in which cells might embody computational processes in a way that the powerful theories of computation, such as finite state machine models, could be productively applied, remains to be seen. Here we demonstrate finite state
more » ... ke processing embodied in cells, using the walking behavior of Euplotes eurystomus, a ciliate that walks across surfaces using fourteen motile appendages called cirri. We found that cellular walking entails a discrete set of gait states. Transitions between these states are highly regulated, with distinct breaking of detailed balance and only a small subset of possible transitions actually observed. The set of observed transitions decomposes into a small group of high-probability unbalanced transitions forming a cycle and a large group of low-probability balanced transitions, thus revealing stereotypy in sequential patterns of state transitions. Taken together these findings implicate a machine-like process. Cirri are connected by microtubule bundles, and we find an association between the involvement of cirri in different state transitions and the pattern of attachment to the microtubule bundle system, suggesting a mechanical basis for the regularity of state transitions. We propose a model where the actively controlled, unbalanced transitions establish strain in certain cirri, the release of which from the substrate causes the cell to advance forward along a linear trajectory. This demonstration of a finite state machine embodied in a living cell opens up new links between theoretical computer science and cell biology and may provide a general framework for understanding and predicting cell behavior at a super-molecular level.
doi:10.1101/2021.02.26.433123 fatcat:x6g6xiikmjhunevtbofqfednli