Variability of bacterial behavior in the mammalian gut captured using a growth-linked single-cell synthetic gene oscillator

David T Riglar, David L Richmond, Laurent Potvin-Trottier, Andrew A Verdegaal, Alexander D Naydich, Somenath Bakshi, Emanuele Leoncini, Johan Paulsson, Pamela A Silver

Preprint posted on November 17, 2018

Using the synthetic oscillator - repressilator v2.0 to monitor bacteria behaviour in mammalian gut..

Selected by Meng Zhu


To keep life at a right pace, many biological systems display periodic fluctuations. Some well-known natural biological oscillators include the cell cycle, somite segmentation clock and circadian rhythms. Theoretically, a gene circuit composed of a negative feedback loop would be sufficient to generate periodic behaviors.  This is well demonstrated by the first synthetic oscillator – termed repressilator. Repressilator was built in E.coli cells, and is made up of only three proteins that antagonize each other in turn (Figure 1) – LacI inhibits the transcription of the second repressor TetR, which then inhibits CI. Finally, CI inhibits lacI expression to complete the circle. Each bacterium also carries a reporter plasmid which produces GFP upon the repression of tetR. In this way, the progress of the oscillation can be read out by GFP signal intensity.


Figure 1. Repressilator system design. Repressilator consists of three basic elements that repress each other in turn. A reporter plasmid is used to monitor repressilator dynamics. From Elowitz and Leibler., 2000 (see further reading).


Repressilator has been later modified by adding several repressor binding sites (and is therefore termed repressilator v2.0). These modifications significantly improved the regularity and stability of the waves and so massively increased its potential for biological engineering or biomedicine applications.  In this preprint, the authors proposed to use repressilator to monitor the growth rate of bacteria in the mammalian gut, and demonstrated the feasibility of this idea by using the mouse as a model. This work offers one of many possibilities that the repressilator system can be used for medical applications.


Key results

When a bacterium expressing repressilator v2.0 system is seeded on a plate, only the cells on the periphery will divide. In this manner, the oscillatory expression of fluorescent proteins during bacterial growth will be propagated on the plate radially as the colony expands. Importantly, the phase information can be read out by the radius of the rings. This property would allow one to read the initial phase of the seeding bacterium. In addition, the fact that roughly 14 generations are needed to complete an oscillation enables the growth rate to be theoretically calculated. Practically, the authors first developed a Repressilator-based Inference of Growth at Single-cell level (RINGS) system to analyze the fluorescence ring radius of each colony (Figure 2). By comparing the growth measurements generated by classical ways or by RINGS, the authors validated that the phase of the bacteria can indeed be a good readout for growth, and furthermore, they found that the oscillation periods of repressilator are insensitive to either strain or environmental variations, therefore allowing to apply this system to a wide range of conditions.

Figure 2. RINGs analysis has been used to measure the phase of the colony. Figure 1c from the preprint.

The authors decided to use this workflow to test the bacteria growth rate in the mouse gut. Considering the critical function of the mammalian gut microbiota’s composition in maintaining proper adult physiological status, this system would certainly have great application potential. Experimentally, the authors fed the mice with the repressilator v2.0 expressing bacteria and collected the bacteria back from the fecal samples with roughly every 6 hours over a day, and analyzed the bacteria phase by RINGS. As the first test of the system, the authors compared the growth rate of the bacteria from the mice treated or untreated with streptomycin, with two different bacteria strains, and different time durations. As good news, the repressilator v2.0 system works relatively stable:  control and streptomycin groups share similar bacteria growth rate and this behavior is conserved between two bacteria strains, and moreover, the bacteria seem to remain functional for over 16 days.

Next, the authors compared the bacterial growth in mice treated or untreated with Dextran Sulfate Sodium (DSS) which is used as an inflammation model. To increase the sensitivity of the system, the authors synchronized the bacteria before feeding. Interestingly, the authors found that, albeit displaying a similar growth rate to control at the beginning of the experiments (within 5hrs), the DSS treated bacteria population display stronger heterogeneity in the long term (>17hrs). This result suggests that the inflammation condition make the bacteria population become desynchronized.


Specific questions to the authors:

Overall, I think this is a very nice work showing a case of how an artificial oscillatory network can be used for medical treatments. Of course, the application potential of this system can be much broader and more possibilities could be tested. For example, as a developmental biologist in training, I am interested in seeing how this system can be used for developmental contexts. For example, can we adapt the design of this system to generate synthetic somites?


But coming back to the preprint I have three specific questions:

  1. What do the authors think are the possible explanations for the de-synchronization seen in the inflammation experiment? Could it be the differential niches that the bacteria encounter? Could this result somehow actually reflect the low robustness of this system? In fact, no in vitro experiments have been performed to test whether the DSS treatment would impair the synchronization of the oscillation in vitro.
  2. Do the authors think that this pipeline can be applied to monitor the other health issues related to the gut? (for example, parasite?)
  3. Are there any potential risks of using this system? Although there is no doubt that using repressilator v2.0 to measure the growth of the bacteria, especially in the way demonstrated by the authors, is truly feasible, feeding the bacteria to healthy individuals sounds quite scary. What potential risks do the authors think this proposed method can confront?


Further reading:

  • Elowitz, M. B. & Leibler, S. A synthetic oscillatory network of transcriptional regulators. Nature 403, 335-338 (2000).
  • Potvin-Trottier, L., Lord, N. D., Vinnicombe, G. & Paulsson, J. Synchronous long-term oscillations in a synthetic gene circuit. Nature 538, 514-517 (2016).


Tags: human health, repressilator, synthetic oscillator

Posted on: 25th January 2019 , updated on: 26th January 2019

Read preprint (No Ratings Yet)

  • Have your say

    Your email address will not be published. Required fields are marked *

    This site uses Akismet to reduce spam. Learn how your comment data is processed.

    Sign up to customise the site to your preferences and to receive alerts

    Register here

    Also in the systems biology category:

    Lineage tracing on transcriptional landscapes links state to fate during differentiation

    Caleb Weinreb, Alejo E Rodriguez-Fraticelli, Fernando D Camargo, et al.

    Selected by Yen-Chung Chen


    Short-range interactions govern cellular dynamics in microbial multi-genotype systems

    Alma Dal Co, Simon van Vliet, Daniel Johannes Kiviet, et al.


    Rapid microbial interaction network inference in microfluidic droplets

    Ryan H Hsu, Ryan L Clark, Jin Wei Tan, et al.

    Selected by Connor Rosen

    High-throughput functional analysis of lncRNA core promoters elucidates rules governing tissue-specificity

    Kaia Mattioli, Pieter-Jan Volders, Chiara Gerhardinger, et al.

    Selected by Clarice Hong

    Variability of bacterial behavior in the mammalian gut captured using a growth-linked single-cell synthetic gene oscillator

    David T Riglar, David L Richmond, Laurent Potvin-Trottier, et al.

    Selected by Meng Zhu

    Charting a tissue from single-cell transcriptomes

    Mor Nitzan, Nikos Karaiskos, Nir Friedman, et al.

    Selected by Irepan Salvador-Martinez

    Large-scale analyses of human microbiomes reveal thousands of small, novel genes and their predicted functions

    Hila Sberro, Nicholas Greenfield, Georgios Pavlopoulos, et al.

    Selected by Ganesh Kadamur

    Symmetry breaking in the embryonic skin triggers a directional and sequential front of competence during plumage patterning

    Richard Bailleul, Carole Desmarquet-Trin Dinh, Magdalena Hidalgo, et al.

    Selected by Alexa Sadier

    RNase L reprograms translation by widespread mRNA turnover escaped by antiviral mRNAs

    James M Burke, Stephanie L Moon, Evan T Lester, et al.

    Selected by Connor Rosen

    Acquired interbacterial defense systems protect against interspecies antagonism in the human gut microbiome

    Benjamin D. Ross, Adrian J. Verster, Matthew C. Radey, et al.

    Selected by Connor Rosen

    DNA microscopy: Optics-free spatio-genetic imaging by a stand-alone chemical reaction

    Joshua A. Weinstein, Aviv Regev, Feng Zhang

    Selected by Theo Sanderson


    The Toll pathway inhibits tissue growth and regulates cell fitness in an infection-dependent manner

    Federico Germani, Daniel Hain, Denise Sternlicht, et al.

    Selected by Rohan Khadilkar

    LCM-seq reveals unique transcriptional adaption mechanisms of resistant neurons in spinal muscular atrophy

    Susanne Nichterwitz, Helena Storvall, Jik Nijssen, et al.


    Axon-seq decodes the motor axon transcriptome and its modulation in response to ALS

    Jik Nijssen, Julio Cesar Aguila Benitez, Rein Hoogstraaten, et al.

    Selected by Yen-Chung Chen

    Memory sequencing reveals heritable single cell gene expression programs associated with distinct cellular behaviors

    Sydney M Shaffer, Benjamin L Emert, Ann E. Sizemore, et al.

    Selected by Leighton Daigh


    Conserved phosphorylation hotspots in eukaryotic protein domain families

    Marta J Strumillo, Michaela Oplova, Cristina Vieitez, et al.

    Selected by Gautam Dey

    LADL: Light-activated dynamic looping for endogenous gene expression control

    Mayuri Rege, Ji Hun Kim, Jacqueline Valeri, et al.

    Selected by Ivan Candido-Ferreira

    A minimal "push-pull" bistability model explains oscillations between quiescent and proliferative cell states.

    Sandeep Krishna, Sunil Laxman

    Selected by Lauren Neves