FET: Medium: Programming multi-cellular systems with spatially-defined computation

Many biological systems, most prominently networks of neurons, control the flow of information by physically organizing component connectivity and spatial organization. However, although recent technological advancements make it possible to comprehensively map the layout of biological networks (e.g., the connectome), researchers are still far from understanding how spatial architecture enables and relates to information processing. Investigators of this project propose to take an engineering approach to understanding how complex information processing can be achieved with a very limited set of molecular and cellular components simply by controlling the spatial organization of these components. The presented approach will result in novel insights into the biophysics of computation. Rather than asking how information processing occurs in a specific, existing biological system, investigators ask what the minimal requirements are for complex computation and information processing in a synthetic multicellular, spatially organized system. Using this principle, they argue that almost arbitrarily complex computation can be achieved with a very small set of engineered cells. Importantly, and unlike prior work, investigators do not need to engineer new molecular components or cell types to increase circuit complexity but can expand the circuit footprint in space to accommodate additional building blocks. Investigators will develop technological innovations to scale up cellular computing and memory systems using a minimal set of molecular parts. They will achieve this goal by overcoming limitations in multiple domains, including orthogonal circuit components, cell-to-cell communication and cell storage capacities. A key innovation is the use of spatial organization to enable systematic reuse of molecular signaling and computation units. The result will be systems that enable on-demand programming of arbitrary circuit architectures into multicellular yeast biofilms. To accomplish these goals, the biofilms will be composed of two different yeast strains, each containing a different circuit component. The two types of components are for logic integration and signal propagation. In this system, a dual rail logic encoding will be used in which the logical '0' will be represented by the presence of the chemical signaling molecule alpha-factor, and the logical '1' will be encoded by the presence of the small molecule auxin. Logic integration will be carried out by ``gate" cells, which will contain an intracellular transcriptional NOR logic gate. For signal propagation, a single strain of ``wire" cells will be used to transmit the Boolean outputs (alpha-factor or auxin) via active signaling along spatially-defined paths that connect regions of gate cells, and gate cell outputs to downstream gate cells and/or circuit output cells. Circuit input cells will have the potential to be programmed with any arbitrary set of biosensors (e.g., electrogenetic redox-based sensors), while the output cells can be programmed with molecular memory systems that can record and permanently store transient circuit outputs directly in cellular DNA (e.g., using CRISPR-based molecular recorders). This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.