Genomic Science Program
U.S. Department of Energy | Office of Science | Biological and Environmental Research Program

2024 Abstracts

Mapping Toxin-Antitoxin Systems for Microbial Community Biocontainment


Dante Ricci* ([email protected]), Jonathan Bethke, Jeff Kimbrel, Yongqin Jiao


Lawrence Livermore National Laboratory


The Lawrence Livermore National Laboratory Secure Biosystems Design Science Focus Area (SFA) aims to develop robust biosecurity tools at the sequence, cellular, and population levels to safeguard the deployment of genetically engineered bacteria for environmental applications. This project aims to exploit the competition between toxin-antitoxin (TA) systems to model and ultimately regulate horizontal gene transfer (HGT) within microbial communities for biocontainment.


Bacteria rapidly disseminate genetic information through HGT, a fundamental driver of microbial evolution. While there is enormous potential in the development of engineered microbial products that are compatible with native microbiota in target environments (e.g, gut or rhizosphere microbiomes), the challenge lies in controlling HGT to and from deployable engineered bacteria. This raises important questions regarding the maintenance of genetic stability over time, and the ecological containment of genetically modified organisms as well as the recombinant or synthetic constructs they harbor. Rather than attempting to suppress natural HGT in situ, the team’s goal is to examine the forces and barriers that shape HGT networks in microbial populations, and to leverage the principles uncovered to develop genetic tools that promote genetic stability of deployable engineered microbes.

The ubiquitous and mobile nature of TA systems in prokaryotes makes them versatile effectors of biocontainment mediated through HGT network interactions. Researchers systematically identified and mapped 40,000 TA systems onto the global bacterial plasmidome, discovering how TA systems are organized through HGT communities, rather than traditional taxonomic classifications. Machine learning models trained on the most common 10% of TA systems alone were able to assign plasmids to HGT communities with 95% accuracy, suggesting each HGT community has its own unique and predictable TA signature. The results of this study imply that HGT networks are constrained, at least in part, by the compatibility between TA systems and provide a coherent explanation for the otherwise erratic distribution across microbial genomes. Understanding and leveraging the dynamics of this innate competition between TA systems could form the basis of a TA-based mechanism for custom community-level invasion or biocontainment. Initial models will inform experiments to test the potential and limitations of TA-based design for controlling the horizontal spread of engineered plasmids outward and natural plasmids inward, in both simple and complex microbial consortia.

Funding Information

This work is supported by the U.S. DOE, Office of Science, BER program, Lawrence Livermore National Laboratory (LLNL) BioSecure Science Focus Area within the Secure Biosystems Design program. Work at LLNL is performed under the auspices of the U.S. DOE at LLNL under Contract DE-AC52-07NA27344.