Overview. A longstanding goal for synthetic biology has been the engineering of emergent behaviors at increasing length scales through genetic programming of molecular-level interactions. Considerable effort has focused on the bottom-up design of emergent pattern formation in commonly studied microbes, primarily Escherichia coli. Feats in the engineering of pattern formation have mainly utilized microcolonies, requiring specialized imaging equipment for detection, or swimming motility in unstable liquid environments. Yet a remarkably diverse and untapped array of bacterial species naturally form complex, macroscopic patterns on solid surfaces via a rapid and coordinated motility mechanism known as swarming. Motivated by the scale and robustness of swarming, as well as the power of visual patterns as vesicles of information, the proposed research strove to engineer bacterial swarming to record information in macroscopic patterns. Demonstrating this concept, we developed a platform for engineering the swarming, bullseye-forming bacterium Proteus mirabilis to record exposure to prescribed signals within changes to its ring pattern, analogous to the recording of environmental information within the rings of a tree trunk. Rounding out our macroscale bacterial recorder, we composed an extensive computational toolkit to accurately decode the bacterium’s spatial record of inputs. By notably engineering a copper-sensing strain and using widely available technology within our methodology, we have shown a proof-of-concept of how our system can serve as a low-cost, visual recorder of environmental pollutants in low-resources settings that might lack access to specialized equipment for image acquisition and analysis. The development of work under this award aimed and proved to be impactful and interdisciplinary, arising from and contributing to fields that span synthetic biology, computer science, education, and the arts. 

Intellectual Merit. We specifically engineered an array of P. mirabilis strains with novel single- and dual-input genetic circuits that, in response to various chemical inputs such as copper, tune the expression levels of swarming-related genes, each of which distinctly and gradually modify the bacterium’s bullseye pattern in quantifiable ways. The modular and diverse collection of genetic circuits, built from biological parts of both standard and novel use, extends the engineering toolkit to P. mirabilis, which has historically remained outside the scope of synthetic biology research. Our repertoire of custom computational methods, ranging from statistical  to state-of-the-art deep learning-based classification and segmentation methods, enables the decoding of spatially encoded inputs under various conditions and restraints. We have thus far shared our macroscale bacterial recording platform with various communities via two peer-reviewed scientific publications (Nature Chemical Biology, 2023; IEEE International Symposium on Biomedical Imaging, 2022), their preceding preprints available on bioRxiv, and publicly available Github repositories (github.com/daninolab) housing our computational methods and raw image data. This dissemination represents a significant advancement, as no publicly accessible datasets and state-of-the-art approaches previously existed for analysis of macroscopic P. mirabilis colony patterns. Upon accessing our image dataset, researchers can inform and augment their studies on the macroscopic spatial consequences of genetically and environmentally regulated microbial behaviors. Overall, our methodology creates a framework for future studies on bacterial motility mechanisms, clinically and environmentally relevant biofilms, and more broadly, engineering coordinated cellular behaviors using diverse model systems.

Broader Impact. Throughout the lifetime of this award, we recruited and mentored several trainees, the majority young females at various stages in their research careers. The core team included two PhD students, one technician, two Master’s students, and four undergraduate students, several of whom gained their first wet-lab experiences with us and all of whom uniquely contributed to the development and dissemination of the proposed work. Under the direct mentorship of Dr. Tal Danino (PI) and deep learning collaborators Dr. Andrew Laine and Dr. Jia Guo, all trainees gained ample experience in synthetic biology methodology, computational analysis, manuscript preparation, and research presentations. The two primary PhD-level research assistants, one of whom successfully defended her dissertation within the past year, delivered a total of thirteen oral and poster presentations on the awarded work at local and international conferences and symposia, as well as three guest lectures for Columbia undergraduate and graduate courses. This list of presentations expands with those given by the undergraduate students at various summer research programs. Outside of research and academia, we have spread awareness of engineering the microbiological world to a broader audience by creating and sharing bacteria artworks such as through the organization of an art exhibition at the BMES conference (2023). This award has accelerated our efforts toward realizing the potential of bacterial patterns as both a biotechnological tool and medium for educational outreach, in turn advancing synthetic biology and public awareness of the complexity of the microbial world.

Last Modified: 04/16/2024
Modified by: Tal Danino