Bacteria are responsive to the environment within the mammalian gut. Scientists have long desired to harness bacteria to detect disease or therapeutically influence the gut environment, but to date, synthetic genetic circuits in bacteria have proven susceptible to mutation and unpredictable function when they colonize the gut for an extended period of time.

New work, led by Pamela Silver of The Wyss Institute of Biologically Inspired Engineering at Harvard (USA), appears to have overcome this problem. The research, which has been called “incredibly exciting” by other scientists, involved the design of a bacterial trigger circuit that detects and responds to tetrathionate—a transient product of reactive oxygen species that is produced during intestinal inflammation. While tetrathionate may even be a general marker of inflammation, until now, scientists were unable to detect it non-invasively.

Silver’s team worked with the mouse commensal Escherichia coli strain NGF-1, inserting a synthetic genetic circuit with two elements: a trigger element adopted from S. typhimurium that specifically recognizes the biomarker (tetrathionate); and a memory element, resembling an on/off switch, from a bacterial virus. The memory element was switched on (expressing Cro protein and beta-galactosidase) when the tetrathionate recognition gene was expressed, turning the bacteria from colourless to blue on solid media. The synthetic gene circuits of the strain were stable over many generations of bacteria; the researchers called this new engineered strain E. coli PAS638.

After in vitro testing, the researchers used a mouse model of S. typhimurium-induced colitis and found E. coli PAS638 could detect tetrathionate in vivo. In these mice, inflammation was quantified via the lipocalin-2 (LCN-2) biomarker in combination with blinded scoring of the histopathology of cecum and colon sections. Infected mice whose fecal samples contained memory-on state colonies of E. coli PAS638 showed higher cumulative LCN-2 values at four days post-infection than those with memory-off colonies. Thus, E. coli PAS638’s sensing of tetrathionate corresponded with a more acute inflammatory response. The researchers confirmed in additional experiments that tetrathionate was detectable in the same manner in mice that were genetically predisposed to inflammation. The engineered strain of bacteria retained its ability to detect tetrathionate in mice for at least 200 days, or more than 1600 generations of bacteria.

This research marks the first time researchers have engineered stable bacterial strains that maintain function for a long period of time (six months) in the mammalian gut. The next challenge for the research group will be to investigate the applicability of this engineered bacteria to humans, creating what they call “living diagnostics”.

“Our approach is to use the bacteria’s sensing ability to monitor the environment in unhealthy tissue or organs. By adding gene circuits that retain memory, we envision giving humans probiotics that record disease progression by a simple and non-invasive fecal test,” says first author David Riglar of Harvard, in a press release.

Donald Ingber, Harvard professor and Founding Director of the Wyss Institute, adds: “If successful in humans, their technology would offer a much less expensive and more specific way to monitor gut function at home than sophisticated imaging instruments used today.”




Riglar DT, Giessen TW, Baym M, et al. Engineered bacteria can function in the mammalian gut long-term as live diagnostics of inflammation. Nature Biotechnology. 2017. doi: 10.1038/nbt.3879