Control strategies for multitasking cellular computers

Cells can host engineered networks of regulatory proteins, the so-called genetic circuits, that are able to sense inputs and generate outputs according to predefined rules. These cellular computers are often single-purposed, so they will exclusively perform the function encoded in a specific genetic circuit. The implementation of new computing functions requires to re-design and re-engineer the cells - a time-consuming refactoring process.

We designed a system in which more than one genetic circuit coexist, and control strategies select which one is functional (i.e. which task runs) at a given time. Since circuits are typically engineered in plasmids, multitasking here refers to the presence of different plasmids that encode for different functions, and the ability to switch between them on demand. Key to this approach is the control of plasmid copy number across a bacterial population. To this end, our model builds on: (1) horizontal gene transfer (HGT) to facilitates the spread of plasmids through non-parental lineages, and (2) inhibition of plasmid replication to decrease copy numbers after a mother cell splits in two daughters. Computational simulations highlight the role of HGT to give rise to circuit clustering across a population, thus avoiding plasmid loss and allowing for reversible computations. Multitasking, a common feature of digital computers, allows more efficient use of hardware. In synthetic biology, cells (hardware) would host different circuits (software) simultaneously which can be executed at will according to external control.