Team:NUS Singapore/Description

NUS iGEM 2019


Description and Inspiration
Engineered organisms are being used to solve global problems today, from cleaning up our environment to diagnosing diseases. Yet, bottlenecks in engineered microbes lie in their limited functional lifespan and inherent stochasticity. Inspired by the phenomenon of hibernation in higher organisms, as well as the ability of our electrical devices to turn off when unused, we aim to overcome these limitations by engineering an ‘ON/OFF’ switch for the cells, giving the ability to control their productivity and extend their productive lifespan. Aided by modelling, we developed this switch using toxin-antitoxin modules which target global translational process and cellular metabolism to enable dormancy. Our technology allows the insertion of different input control modules to regulate these circuits in a plug-and-play manner. To ensure the safety and retainability of our circuits, a biocontainment module was designed to prevent the unwanted spread of our modules to other organisms. Finally, we successfully demonstrated regulated and lengthened productivity in the context of bioluminescence production and biosensing. Our technology thus pushes the boundaries of synthetic biology, bringing it closer to real world adoption.
Background & Proposal
Synthetic biology is a field that has been experiencing exponential growth over the past years. More and more effort, in both the private and the public sector, has been put into developing the industry and there are many different thriving areas in which synthetic biology is playing an increasingly important role.


Fig. 1: An overview of the field of synthetic biology. Illustration obtained from NUS Yong Loo Lin School of Medicine (Medicine, n.d.).

While synthetic biology can encompass many different things, from adding genetic circuits to cells to building entire cells from the bottom up (Ausländer, Ausländer, & Fussenegger, 2017; Church, Elowitz, Smolke, Voigt, & Weiss, 2014; Ganzinger & Schwille, 2019; Khalil & Collins, 2010), the majority of synthetic biology work involves engineering living cells to carry out specific functions. One major example of this would be biomanufacturing, whereby large bioreactors are filled with engineered yeast or E. coli that produce different compounds. Another example of this would be in the field of bioremediation, where E. coli is modified to use PET as a food source, with the aim of tackling the plastic pollution problem (Caruso, 2015).

The advantages of using living matter for such applications are clear. Living cells can divide, allowing them to achieve a large scale with relatively simple starting nutrients. However, as with all things living, one major limitation to their usage is simply that these machines are variable, and as a result cannot be as simply controlled as in silica devices (Bandiera, Furini, & Giordano, 2016). As such, this inherent limitation is something that all aspiring biotechnologists have to accept.

Yet, we believe otherwise. We believe that in successful engineering, one must be able to attain fine and precise control of the various activities of the device in question. As such, we believe we have developed a way to provide scientists and engineers a way to strictly control the cell’s growth - allowing them to be truly turned on and off. As a result, this significantly expands the longevity of the device as they would not be growing and using up nutrients unnecessarily - opening up the use of such devices for a larger market.

Stochasticity & Control
Engineered organisms range from simple bacterial E. coli cells to more complex eukaryotes such as human T-cells. However, they all share one thing in common - being living systems, they have unpredictability inbuilt.


Fig. 2: Image showing variability of gene expression in E. coli. Picture obtained here.

Stochasticity in living systems serve multiple purposes. The original proposal of stochastic gene expression being common in individual cells was elegantly described by Elowitz et al (Elowitz, Levine, Siggia, & Swain, 2002), whereby it was shown that cells have an inherent distribution of gene expression by the visualization of the production of a fluorescent protein marker. The field of stochastic biological studies has grown steadily, with multiple hypotheses as to why such stochasticity happens in the natural environment. Some groups have proposed that this stochasticity is what enables the adaptability of life to the multitude of ever-changing environments and stimuli, while some have argued that it is a hindrance - that life has to instead overcome this randomness in order to flourish (Honegger & de Bivort, 2018).

While this debate is ongoing in the natural world, we as engineers of biology have a different objective - stochasticity in engineered devices is a significant proble which has to be overcome or managed. Thus, the field of synthetic and systems biology has put in significant work to better understand and characterize our living systems, thereby possibly coming up with ways to reduce or eliminate this inherent randomness entirely.

We aim to support our fellow engineers in this endeavour. We seek to create ways to directly control the growth of the cell, using engineered genetic circuits which are not as susceptible to variability in their activity to do so. This will then allow indirect but more holistic control of the functionality of the engineered cell.

Longevity of Engineered Bacteria - A question of metabolic state
Scientists have long endeavored to create cells that can sustain their function for a lengthened period of time. In this section we elaborate on how metabolic activity is intrinsically connected to cellular lifespan, and propose a way to control it.

The life and death of various living organisms are usually dictated via many common factors. One major factor is that of nutrient supply to the organism. In particular, for prokaryotes such as the ubiquitous E. coli, the growth of the cell is almost strictly limited in laboratory conditions by the presence of varying nutrients. While other factors such as temperature and UV radiation also cause cellular damage and death, these aspects are tightly controlled in the laboratory. As a result, one can clearly correlate cellular death with the lack of nutrients (Sezonov, Joseleau-Petit, & D’Ari, 2007).

Yet, in the living environment, cells are unable to control the composition of their surroundings. However, they can alter their cellular state to better suit these surroundings. Work to date has shown that E. coli has proven to be resilient to alterations in its external environment. It has been shown that E. coli can retain viability in a stagnant, non-refreshed culture for many years on end (Westphal et al., 2018) and recent research has delved deeper into the mechanisms in which E. coli has developed to withstand the many environmental stressors it is subjected to. One phenomenon, termed as the Growth Advantage in Stationary Phase (GASP) phenotype, has been subjected to much study in recent years (Finkel, 2006). Other groups have also suggested that these cells which are seemingly alive after a prolonged duration of nutrient stress also appear to be able to produce exogenous proteins upon induction (Gefen, Fridman, Ronin, & Balaban, 2014). Other areas of research have also suggested that E. coli can respond differently to different nutrient loads, with growth and survival differing depending on the substrate being nitrogen or carbon limiting (Iyer, Le, Park, & Kim, 2018).

The phenomenon of bacterial persistence, a state of dormancy with lowered metabolic activity, has also been widely characterized in recent years, due to it playing a role in the increasing trend of antibiotic-resistant bacteria (Amato et al., 2014). However, the ability of cells to enter a persistent state has not been widely studied, and has much more to be discovered.


Fig. 3: Bacterial life cycle, showing how a significant portion of bacteria die off after a few days and subsequently enter a poorly-studied long-term phase of persistency. Figure taken from Kram & Finkel (Kram & Finkel, 2015).

Yet this ability of cells to enter a persistent, metabolically inactive state was exactly what drew us to the causes of this phenomenon. We hypothesized that since nutrient load was effectively the root cause of cell death, ways to slow down unnecessary consumption of nutrients or conserving cellular resources by manipulating the metabolic state of the cells would be key to extending cellular lifespan. This thought process led us to the phenomenon of hibernation among animals, whereby long winters lacking nutrients are overcome by drastically lowering metabolic activity. Similarly, from the engineering point of view, a key aspect of control is the ability of a circuit to be turned off when unused, conserving energy. We realized that despite the relative simplicity of such a concept, there was no real way to turn cells 'off' when they are not needed.

This led us to our project - to use toxin-antitoxin and glucose inhibition systems to control cellular metabolism, thereby creating an 'ON/OFF' switch and a tunable knob, bringing synthetic biology one step closer to engineering.
Growth Control Systems
1) Toxin-antitoxin System
Toxin-antitoxin systems are naturally occurring protein pairs, which as the name suggests, have a toxic and a corresponding recuperative effect on various cell types. There exists multiple types of toxin-antitoxin systems (Harms, Brodersen, Mitarai, & Gerdes, 2018), with varying functions, from being a key component of plasmid retention in bacteria via the post-segregational killing mechanism (Gerdes, Jacobsen, & Franch, 2012) to possibly playing a role in the phenomenon of persistence in bacteria, a key problem facing clinicians around the world - although this particular link has been hotly disputed in recent times (Page & Peti, 2016; Ronneau & Helaine, 2019).

Toxin-antitoxin systems, irregardless of type, generally function by inhibiting the growth of the bacteria, causing either a bactericidal or bacteriostatic effect. The toxin causes this effect by inhibiting transcription or translation, or even NAD+. The antitoxin, on the other hand, neutralizes this effect by sequestering the toxin away. While the exact stoichiometric ratios of toxins and antitoxins differ between different systems, it is generally shown that having a larger amount of antitoxin than toxin neutralizes this effect, and vice versa (Hall, Gollan, & Helaine, 2017).

Having this understanding, we hypothesized that since the growth of the cell can be controlled by the amount of toxin and antitoxin present in the cell, we could then utilize this system to create our ‘ON/OFF’ switch. Currently, most people lower cellular activity by simply lowering the temperature. But we believed that this lowering of temperature is unsuitable for many of the use cases that we intend for engineered bacteria to fulfil as it requires a well-controlled environment and cannot be done in situ. Thus, we decided that we could repurpose the supposed negative properties of toxins and antitoxins to benefit us and provide us with a solution to our problem of shortened bacterial lifespan.

HicA-HicB
The HicA-HicB toxin-antitoxin system is one of the multiple canonical toxin-antitoxin systems that are found endogenously in many different species of bacteria, including E. coli. They are considered to be a type II system, as HicA and HicB are both proteins produced from gene sequences and exert their effect in that form. The HicA toxin has been experimentally tested to bind to specific sequences in mRNA, degrading them and resulting in a lowered rate of translation of proteins. This global reduction in translation thereby results in an overall lowering of growth (Jørgensen, Pandey, Jaskolska, & Gerdes, 2009; Turnbull & Gerdes, 2017; Winter et al., 2018).

The HicB antitoxin neutralizes the effect of the HicA toxin by binding to it, resulting in the HicA toxin being unable to bind to and degrade mRNA. In normal cellular conditions, the HicB antitoxin is present in larger amounts compared to the HicA toxin, thus allowing for growth to occur. In situations of stress, the HicB antitoxin is speculated to degrade faster than the HicA toxin due to its less stable structure as well as via lon-protease mediated degradation, resulting in the arrest of growth. Subsequent HicB expression has been shown to induce a recovery in growth, indicating that the HicA-HicB system is bacteriostatic in nature. This therefore encouraged us to use this system for our experiments as it implied that states of dormancy and growth can be controlled by the levels of HicA and HicB respectively.

Fig. 4: Illustration of the mechanism of action of HicA-HicB in a normal and stressed state.

We hypothesize that HicA inhibition of growth by translation inhibition results in a change in the metabolic profile of the cell, resulting in the conservation or accumulation of resources that can then be used up upon turning ‘on’ the cell. This hypothesis was inspired by research into the action of chloramphenicol on bacteria, which has a similar effect of translation inhibition. It was noted that upon inhibition of translation by chloramphenicol, an accumulation of ATP was observed (Lobritz et al., 2015). It has also been previously shown that cells with an enhanced level of ATP production resulted in increased protein production, thus drawing a link between ATP levels and protein expression (Hara & Kondo, 2015; Kim, Kwon, Lee, & Kim, 2012). We thus postulated that this accumulation of ATP resulted in an energy store that the cell could subsequently use upon ‘waking’ with the neutralization of the toxin with antitoxin expression.

RES-Xre
The RES-Xre system is another toxin-antitoxin system that is found in various organisms. Of note is that this system is not endogenous to E. coli, making it unique compared to the many different commonly studied toxin-antitoxin systems around. This system is hypothesized to be a type II TA system, with the antitoxin gene found in front of the toxin gene as well as an overlap in the open reading frames of both genes. It was found to be an essential TA loci for the growth of Sinorhizobium meliloti, a nitrogen-fixing bacterial symbiont, and has also been speculated to play a role in the pathogenicity of Photorhabdus luminescens, a γ‐proteobacterium endemic in insects and nematodes. The RES toxin has been found to exert its toxic effect via a mechanism unique to toxin-antitoxin systems, as it acts as an NADase enzyme, reducing intracellular levels of NAD+ in the cell and thereby arresting growth. The toxin-antitoxin complex formed with RES-Xre is also in an unusual 1:2 ratio, and it is further postulated that the antitoxin is similarly involved in its own gene regulation due to the presence of Helix-Turn-Helix DNA binding domains in the protein (Skjerning, Senissar, Winther, Gerdes, & Brodersen, 2019).

Fig. 5: Illustration of the mechanism of action of RES-Xre in a normal and stressed state.

Since the mechanism for the effect of RES-Xre is different from that of HicA-HicB, we postulated that the reason for resource conservation would likely differ from that of HicA-HicB. One possible reason could be due to the sequestering of NAD+ by RES. This might lead to lowered metabolic activity as NAD+ plays an important role in metabolic processes (Smith, 1992) thus possibly resulting in growth inhibition. Subsequently, upon neutralization of RES by Xre, NAD+ would likely be upregulated – resulting in the usage of the accumulated resources (e.g. glucose, ATP, etc.) and thus protein production and growth.

2) Glucose Uptake System
While we decided that the toxin and antitoxin systems were crucial to the core functioning of the system, we did not simply focus on one system to control cellular growth. Research has also shown that regulating the glucose intake of cells could also limit cellular growth (Rice & Vanderpool, 2011).

SgrS
In E. coli, the glucose intake of the cell is regulated by a particular protein known as PtsG, which functions as a glucose transporter (Negrete, Ng, & Shiloach, 2010). Other groups have shown that E. coli regulates the activity and expression of this protein via two unique mechanisms. Firstly, the production of a small RNA known as SgrS directly interacts with the ptsG mRNA and causes its degradation, resulting in a lowered concentration of ptsG proteins. Furthermore, it was also discovered that the SgrS RNA encodes a protein known as SgrT, which also interacts with the ptsG protein, inhibiting its function (Wadler & Vanderpool, 2007).

With this understanding, we sought to add a tunable aspect to our ‘ON/OFF’ switch design. Since the growth inhibition by SgrS is, as characterized by other teams, a linear trend rather than a stepwise one from the toxin-antitoxin system, we decided to combine them together to provide our fellow synthetic biologists with a broader toolbox to control cellular growth and expression.


Fig. 6: Illustration of the mechanism of action of SgrS.

Modularity
We conceived of our ‘ON/OFF’ switch module as one that is, as our terminology implies, modular - that it could fit into potentially any use case as so desired by the user. As such, we theorized that various inducer systems could control our module, and therefore allow the user a much finer level of control.

We decided to showcase the modularity of our system by demonstrating our control using chemical inducers as well as blue-light inducers. In chemical induction, the standards of lactose and arabinose were used to control the toxin and antitoxin induction, while due to the conflating effect of arabinose and glucose, the Tet operon was used to control the SgrS system instead. We also utilized the blue-light EL222 system, which is able to control gene expression via the light-activated dimerization of the EL222 DNA-binding protein (Jayaraman et al., 2016), to regulate the expression of the antitoxin. The main advantage of the blue-light system is that it is reversible, allowing us to turn on and off expression of the antitoxin and by extension cellular growth. This opens up the flexibility of our system for other users.


Fig. 7: Illustration of the control of the growth switch using blue light.

In order to provide an even tighter level of control over the metabolic state of the cell, one potential future direction would be to incorporate a bistable toggle switch into our system. This would be a chemically inducible one, as detailed by Collins et al. in their seminal paper whereby two chemical inducers control the activity of a particular protein (Gardner, Cantor, & Collins, 2000). The toggle switch allows for a clean transition between the expression of two opposing proteins, which in this case would be our toxin and antitoxin respectively. We believe that this system would serve to reduce any potential leakiness in our circuitry, as well as reduce the amount of reagents needed to keep the cell in their dormant state.


Fig. 8: Illustration of a hypothetical toggle switch design incorporated with the growth switch.

Plasmid Retention & Biocontainment
In the process of understanding how engineered bacteria can last in their environment, we realized that the current method for ensuring that plasmids are retained by cells, which is the utilization of antibiotics as a selection pressure, was unsuitable for most of the use cases as antibiotics can degrade over time, with most antibiotics being ineffective if stored at room temperature for a prolonged period of time. Furthermore, in order to induce antibiotic resistance, a resistance gene had to be transformed into the cell. With the increasing pandemic of drug-resistant bacteria sweeping the world, as well as the propensity for horizontal gene transfer between bacterial species, we wanted to avoid the unnecessary use of antibiotics as a selection pressure. Given that we intend to ensure that engineered bacteria can retain their function for a prolonged period of time, we had to seek ways other than using antibiotics to allow cells to continuously retain the plasmids within them.

We thus conceptualized a method, also using toxin-antitoxin systems, to achieve this effect of plasmid retention. Drawing inspiration from Fedorec et al.’s paper on plasmid retention (Fedorec et al., 2019), whereby it was shown that certain toxin-antitoxin pairs were able to significantly increase the duration in which plasmids were retained in cells compared to the wild type, we decided to use another set of toxin-antitoxin pairs, separate from the controllable module regulating the ‘ON/OFF’ switch as mentioned above. In this system, the toxin-antitoxin pairs themselves act as a selection pressure, as detailed in the phenomenon of Post-Segregational Killing. Thus, our plasmid with the desired module would be selectively retained in the cell as well.

Every single application for engineered bacteria, as long as it has the intention of being released into the external environment, would necessarily have to have appropriate safety measures built into it. Our intended system is no different. In fact, because we are essentially aiming to induce a state of so-called persistency within our engineered bacteria, further precautions have to be taken into account to prevent the unwanted dissemination of this 'persistency' module into the environment.

As such, we decided to look into ways in which we could contain the spread of our plasmid to the external environment. The literature on this topic was abundant, with various methods, such as the engineering of auxotrophic systems, the addition of genetic circuits to trigger upon target signals and transgene compartmentalization being used and proposed (Lee, Chan, Slomovic, & Collins, 2018). However, many of these methods involve the addition of yet another layer of genetic circuitry and complexity, which we felt would create an undesirable burden on the cell. In particular, since our system is very much focused on the stringent control of growth and thereby activity, the various biocontainment methods proposed above would significantly impact our claim.

However, we similarly realized that our desire to use a plasmid retention system could also serve as a form of biocontainment if the toxins and their respective antitoxins were kept on two separate plasmids. This would be able to encompass both of our desired outcomes of retention and biocontainment without having to design or add a new layer of circuitry. As a result, we came up with the following design.


Fig. 9: Concept of the biocontainment and retention system.

From this design, it can be clearly seen that all the components for most of our system is neatly contained within two modules - the retention-cum-biocontainment module, and the primary 'ON/OFF' switch module. The biocontainment aspect comes into play as both plasmids must be present within the cell in order for it to survive. Therefore, the chances of successful horizontal gene transfer is significantly lowered. Furthermore, this system, by doubling up as a way to retain the plasmids, allows us to remove the need for antibiotic resistance - thereby further rendering it less dangerous when released into the environment.

For the toxin-antitoxin systems used in this retention and biocontainment module, we utilized the Txe-Axe and Hok-Sok toxin-antitoxin systems. Both are purported to be bactericidal in nature. In the paper published by Fedorec et. al., it was also found that both systems were able to increase the retention rate of the plasmid in which the system was inserted into (Fedorec et al., 2019). With this knowledge, we posited that a swapping of the toxins and corresponding antitoxins on different plasmids would then likely have an even more significant retention and killing effect, thereby achieving our goal. We did not use the same toxin-antitoxin systems as our sleep-wake module for this containment and retention system as we did not want to have any possible interference between the two, which would make it difficult to characterize their individual effects.

Our biocontainment system notably only works to prevent the spread of our genetic circuit to the external environment. It does not tackle the other aspect of biocontainment, which is to prevent the unwanted spread of the cell containing our circuitry itself. Future work could involve the further addition of kill switches developed by others to enable this aspect of biocontainment and ensure a more robust and safe cell (Chan et al., 2016; Stirling et al., 2017).
Applications
We believe that our project constitutes a fundamental, foundational development that will have a big impact on synthetic biology as a whole. This control system has been designed from the very beginning to have the traits of modularity as well as tunability, which provides users with greater flexibility in using our technology.

To demonstrate the immediate benefits our system can bring, we decided to look into two common and straightforward applications of engineered bacteria:

1) Bioluminescence Production

2) Whole-Cell Biosensors

2) Bioluminescence
With the increasing threat of global climate change, there is increasing effort to search for sustainable solutions to power our world. One large consumer of power is towards lighting, which is considered an essential part of civilization. Yet, many areas of the world are still unlighted. How can we reconcile this desire to improve the economic and environmental infrastructure of such areas, yet also be cognizant of the possible damaging effects this change might bring to our climate? The answer is in alternative methods of powering our infrastructure.

One alternative method recently being proposed by many people around the world to address the issue of lighting hard-to-access and rural areas is via the use of bioluminescence. Bioluminescence is a natural and eco-friendly lighting source, with naturally glowing bacteria derived from places such as oceans and fish providing the initial inspiration and development of larger scale bioluminescent efforts (Fleiss & Sarkisyan, 2019). However, one major limitation of the various lighting devices incorporating such bacteria is that of time - the light usually lasts for only a few hours, which prevents any long-term usage of such devices.

We thus envision that our technology can solve this problem of longevity by providing the ability for the bioluminescent device to truly turn ‘on’ and ‘off’ - much like an actual torch. This would then allow the bacteria to emulate a battery - and inch us ever closer to using bioluminescent bacteria as a sustainable and viable light source.


Fig. 11: Cultures of bioluminescent bacteria.

1) Whole-cell Biosensors
One promising use in engineered bacteria is to act as responsive biosensors. Whole cell-based biosensors are a particularly attractive proposition in many use cases due to their ability to adapt to the living environment and respond accordingly. However, being a live cell, these biosensors require a constant supply of nutrients in order to become viable. While cell-free biosensors are slowly being developed, whole-cell based biosensors are still desirable as they can potentially be reusable and longer lasting depending on the environmental conditions, and are easier to engineer due to the innate ability of many living cells to sense and respond that is difficult to reconstitute outside of a cell (Tsien, 1998).

Biosensors can sense a range of materials, from heavy metals to pathogenic biomarkers (Zhang, Jensen, & Keasling, 2015). However, engineering these sensing systems usually impose a heavy metabolic burden on the cell. This therefore results in an increased resource load, which is undesirable when the target to be sensed is not present. As a result, it is desirable for such cells to have the ability to reduce their metabolic activity in the absence of their target, and subsequently 'activate' only in the presence of the target substrate. This is where we believe our technology could play a substantial role by giving the ability to do precisely as stated above.

Most biosensors work by producing proteins within the cell that are then activated by an external biomarker, and subsequently triggers the expression of a reporter - which is essentially an inducible system. In theory, any inducible system would be suitable for our module. To demonstrate a simple example of a system that is inducible by specific external environmental cues, we designed a system inducible by the accumulation of acyl homoserine lactones (AHLs) produced by Pseudomonas aeruginosa, a pathogenic bacteria known to cause respiratory and gastrointestinal infections. This system has been previously demonstrated in Saeidi et al., whereby a genetic circuit was constructed to sense AHLs and release a killing substrate, pyocin (Saeidi et al., 2011). We envision that such engineered cells could be improved significantly if the cell would be dormant until it senses AHL protein, upon which it would then ‘wake’ up and express the desired proteins - be it RFPs for fluorescence detection, or pyocins for killing. Therefore, we designed a system utilizing the LasR protein and the PlasI promoter, which would control the expression of the antitoxin in our system, while the toxin is under the expression of a constitutive promoter. This would then allow the cell to be in its ‘off’ state for a majority of the time. Upon the expression of AHLs by P. aeruginosa, the LasR protein would then form a complex and activate the PlasI promoter, transcribing antitoxins and ‘waking’ up the cell. The cell could then produce RFP or pyocins as so desired by the user. The system is illustrated in the figure below.


Fig. 10: Incorporation of the growth switch with an AHL quorum sensing system.

E.co Grow Software and Modelling
The study of growth is something that is common to almost all biologists in some way or another. With our project so heavily centered on the manipulation of growth, we would obtain a significant amount of data pertaining to how various factors affect the growth rate of bacteria, and how perturbations in our system would lead to changes in growth. Thus, we decided to utilize all the data to generate a model that could help others predict and conduct better growth experiments.

The model, as further detailed on the modelling page, utilizes most of the growth data generated from our experiments to determine and construct the equations behind the growth curves. This would therefore allow us as well as others to predict how various factors such as differing inducers and their concentrations would affect the growth of the engineered bacteria with our switches, giving them more insight into their own experiments. We endeavor to create a software tool that can assist fellow scientists in planning their own experiments with growth, and helping them gain more insights into their work without having to go through the same process we did of laboriously constructing multiple growth curves. We hope that this tool will prove useful to scientists studying how various inducers and proteins would affect the growth rate of the engineered bacteria, and allow them to account for such factors. We aim to make this a collaborative effort, and welcome other users to send us their growth curve data to make our models more effective over time.

References
Amato, S. M., Fazen, C. H., Henry, T. C., Mok, W. W. K., Orman, M. A., Sandvik, E. L., … Brynildsen, M. P. (2014). The role of metabolism in bacterial persistence. Frontiers in Microbiology. https://doi.org/10.3389/fmicb.2014.00070

Ausländer, S., Ausländer, D., & Fussenegger, M. (2017). Synthetic Biology—The Synthesis of Biology. Angewandte Chemie - International Edition. https://doi.org/10.1002/anie.201609229 Bandiera, L., Furini, S., & Giordano, E. (2016). Phenotypic variability in synthetic biology applications: Dealing with noise in microbial gene expression. Frontiers in Microbiology. https://doi.org/10.3389/fmicb.2016.00479

Caruso, G. (2015). Plastic Degrading Microorganisms as a Tool for Bioremediation of Plastic Contamination in Aquatic Environments. Journal of Pollution Effects & Control. https://doi.org/10.4172/2375-4397.1000e112

Chan, C. T. Y., Lee, J. W., Cameron, D. E., Bashor, C. J., & Collins, J. J. (2016). “Deadman” and “Passcode” microbial kill switches for bacterial containment. Nature Chemical Biology. https://doi.org/10.1038/nchembio.1979

Church, G. M., Elowitz, M. B., Smolke, C. D., Voigt, C. A., & Weiss, R. (2014). Realizing the potential of synthetic biology. Nature Reviews Molecular Cell Biology. https://doi.org/10.1038/nrm3767

Elowitz, M. B., Levine, A. J., Siggia, E. D., & Swain, P. S. (2002). Stochastic Gene Expression in a Single Cell. Science, 297(5584), 1183 LP – 1186. Retrieved from http://science.sciencemag.org/content/297/5584/1183.abstract

Fedorec, A. J. H., Ozdemir, T., Doshi, A., Ho, Y. K., Rosa, L., Rutter, J., … Barnes, C. P. (2019). Two New Plasmid Post-segregational Killing Mechanisms for the Implementation of Synthetic Gene Networks in Escherichia coli. Food Science and Human Wellness. https://doi.org/10.1016/j.isci.2019.03.019

Finkel, S. E. (2006). Long-term survival during stationary phase: Evolution and the GASP phenotype. Nature Reviews Microbiology. https://doi.org/10.1038/nrmicro1340

Fleiss, A., & Sarkisyan, K. S. (2019). A brief review of bioluminescent systems (2019). Current Genetics, 65(4), 877–882. https://doi.org/10.1007/s00294-019-00951-5

Ganzinger, K. A., & Schwille, P. (2019). More from less – Bottom-up reconstitution of cell biology. Journal of Cell Science. https://doi.org/10.1242/jcs.227488

Gardner, T. S., Cantor, C. R., & Collins, J. J. (2000). Construction of a genetic toggle switch in Escherichia coli. Nature. https://doi.org/10.1038/35002131

Gefen, O., Fridman, O., Ronin, I., & Balaban, N. Q. (2014). Direct observation of single stationary-phase bacteria reveals a surprisingly long period of constant protein production activity. Proceedings of the National Academy of Sciences of the United States of America. https://doi.org/10.1073/pnas.1314114111

Gerdes, K., Jacobsen, J. S., & Franch, T. (2012). Plasmid Stabilization by Post-Segregational Killing. In Genetic Engineering. https://doi.org/10.1007/978-1-4615-5925-2_3

Hall, A. M., Gollan, B., & Helaine, S. (2017). Toxin–antitoxin systems: reversible toxicity. Current Opinion in Microbiology. https://doi.org/10.1016/j.mib.2017.02.003

Hara, K. Y., & Kondo, A. (2015). ATP regulation in bioproduction. Microbial Cell Factories. https://doi.org/10.1186/s12934-015-0390-6

Harms, A., Brodersen, D. E., Mitarai, N., & Gerdes, K. (2018). Toxins, Targets, and Triggers: An Overview of Toxin-Antitoxin Biology. Molecular Cell. https://doi.org/10.1016/j.molcel.2018.01.003

Honegger, K., & de Bivort, B. (2018). Stochasticity, individuality and behavior. Current Biology. https://doi.org/10.1016/j.cub.2017.11.058

Iyer, S., Le, D., Park, B. R., & Kim, M. (2018). Distinct mechanisms coordinate transcription and translation under carbon and nitrogen starvation in Escherichia coli. Nature Microbiology. https://doi.org/10.1038/s41564-018-0161-3

Jayaraman, P., Devarajan, K., Chua, T. K., Zhang, H., Gunawan, E., & Poh, C. L. (2016). Blue light-mediated transcriptional activation and repression of gene expression in bacteria. Nucleic Acids Research. https://doi.org/10.1093/nar/gkw548

Jørgensen, M. G., Pandey, D. P., Jaskolska, M., & Gerdes, K. (2009). HicA of Escherichia coli defines a novel family of translation-independent mRNA interferases in bacteria and archaea. Journal of Bacteriology. https://doi.org/10.1128/JB.01013-08

Khalil, A. S., & Collins, J. J. (2010). Synthetic biology: Applications come of age. Nature Reviews Genetics. https://doi.org/10.1038/nrg2775

Kim, H. J., Kwon, Y. D., Lee, S. Y., & Kim, P. (2012). An engineered Escherichia coli having a high intracellular level of ATP and enhanced recombinant protein production. Applied Microbiology and Biotechnology. https://doi.org/10.1007/s00253-011-3779-0

Kram, K. E., & Finkel, S. E. (2015). Rich medium composition affects Escherichia coli survival, glycation, and mutation frequency during long-term batch culture. Applied and Environmental Microbiology. https://doi.org/10.1128/AEM.00722-15

Lee, J. W., Chan, C. T. Y., Slomovic, S., & Collins, J. J. (2018). Next-generation biocontainment systems for engineered organisms. Nature Chemical Biology. https://doi.org/10.1038/s41589-018-0056-x

Lobritz, M. A., Belenky, P., Porter, C. B. M., Gutierrez, A., Yang, J. H., Schwarz, E. G., … Collins, J. J. (2015). Antibiotic efficacy is linked to bacterial cellular respiration. Proceedings of the National Academy of Sciences of the United States of America. https://doi.org/10.1073/pnas.1509743112

Medicine, Y. L. L. S. of. (n.d.). Synthetic Biology - NUS Medical Sciences Cluster. Retrieved from http://nusmedicine.nus.edu.sg/medical-sciences-cluster/research/synthetic-biology/

Negrete, A., Ng, W. I., & Shiloach, J. (2010). Glucose uptake regulation in E. coli by the small RNA SgrS: Comparative analysis of E. coli K-12 (JM109 and MG1655) and E. coli B (BL21). Microbial Cell Factories. https://doi.org/10.1186/1475-2859-9-75

Page, R., & Peti, W. (2016). Toxin-antitoxin systems in bacterial growth arrest and persistence. Nature Chemical Biology. https://doi.org/10.1038/nchembio.2044

Rice, J. B., & Vanderpool, C. K. (2011). The small RNA SgrS controls sugar-phosphate accumulation by regulating multiple PTS genes. Nucleic Acids Research. https://doi.org/10.1093/nar/gkq1219

Ronneau, S., & Helaine, S. (2019). Clarifying the Link between Toxin–Antitoxin Modules and Bacterial Persistence. Journal of Molecular Biology. https://doi.org/10.1016/j.jmb.2019.03.019

Saeidi, N., Wong, C. K., Lo, T. M., Nguyen, H. X., Ling, H., Leong, S. S. J., … Chang, M. W. (2011). Engineering microbes to sense and eradicate Pseudomonas aeruginosa, a human pathogen. Molecular Systems Biology. https://doi.org/10.1038/msb.2011.55

Sezonov, G., Joseleau-Petit, D., & D’Ari, R. (2007). Escherichia coli physiology in Luria-Bertani broth. Journal of Bacteriology. https://doi.org/10.1128/JB.01368-07

Skjerning, R. B., Senissar, M., Winther, K. S., Gerdes, K., & Brodersen, D. E. (2019). The RES domain toxins of RES-Xre toxin-antitoxin modules induce cell stasis by degrading NAD +. Molecular Microbiology. https://doi.org/10.1111/mmi.14150

Smith, C. A. (1992). Physiology of the Bacterial Cell. A Molecular Approach. Biochemical Education. https://doi.org/10.1016/0307-4412(92)90139-d

Stirling, F., Bitzan, L., O’Keefe, S., Redfield, E., Oliver, J. W. K., Way, J., & Silver, P. A. (2017). Rational Design of Evolutionarily Stable Microbial Kill Switches. Molecular Cell. https://doi.org/10.1016/j.molcel.2017.10.033

Tsien, R. Y. (1998). THE GREEN FLUORESCENT PROTEIN. Annual Review of Biochemistry. https://doi.org/10.1146/annurev.biochem.67.1.509

Turnbull, K. J., & Gerdes, K. (2017). HicA toxin of Escherichia coli derepresses hicAB transcription to selectively produce HicB antitoxin. Molecular Microbiology. https://doi.org/10.1111/mmi.13662

Wadler, C. S., & Vanderpool, C. K. (2007). A dual function for a bacterial small RNA: SgrS performs base pairing-dependent regulation and encodes a functional polypeptide. Proceedings of the National Academy of Sciences of the United States of America. https://doi.org/10.1073/pnas.0708102104

Westphal, L. L., Lau, J., Negro, Z., Moreno, I. J., Ismail Mohammed, W., Lee, H., … Kram, K. E. (2018). Adaptation of Escherichia coli to long-term batch culture in various rich media. Research in Microbiology. https://doi.org/10.1016/j.resmic.2018.01.003

Winter, A. J., Williams, C., Isupov, M. N., Crocker, H., Gromova, M., Marsh, P., … Crump, M. P. (2018). The molecular basis of protein toxin HicA– dependent binding of the protein antitoxin HicB to DNA. Journal of Biological Chemistry. https://doi.org/10.1074/jbc.RA118.005173

Zhang, J., Jensen, M. K., & Keasling, J. D. (2015). Development of biosensors and their application in metabolic engineering. Current Opinion in Chemical Biology. https://doi.org/10.1016/j.cbpa.2015.05.013