a. We imported the genome-scale metabolic network of Saccharomyces
cerevisiae from the BiGG Models database. We integrated six exogenous functional
modules—adsorption, degradation, therapy, and their combinations—each encoding
specific proteins. Using Python’s
COBRApy package, we evaluated how the expression of these foreign genes affects
the metabolic homeostasis of the chassis cell.
c. To ensure the robustness and practical reliability of our results, we
performed a sensitivity analysis on the uncertain biomass parameter—the energy
cost associated with protein secretion.
The results demonstrated that our model is highly robust and biologically meaningful.
d. Overall, this analysis defined the physiological safety
boundary of our genetic design and provided a quantitative
basis for the subsequent development of the project’s safety module.
a. To mitigate the metabolic toxicity described above, our
team actively explored the design of biosafety modules and proposed a
yeast–lactobacillus
symbiotic system as a novel Metabolic Safety Strategy. Through
cross-species metabolic complementation,
this system redistributes energy
demands and metabolite fluxes, thereby alleviating the physiological burden on the
engineered host at the systems level.
b. As conducting in vivo experiments in humans would
violate safety guidelines, we instead constructed a metabolic
coupling model of yeast and Lactobacillus using Python and the COBRApy
framework.
We further introduced an FVA-constrained model to analyze metabolic flux
distribution and maintenance energy demand (ATPM) under symbiotic conditions. The
results showed that bidirectional exchange of key metabolites—such as lactate, amino
acids, and riboflavin—significantly
reduced the yeast’s maintenance energy burden and mitigated the metabolic toxicity
caused by heterologous gene expression.
The mathematical formulation of our model is presented as follows.
c. Under multiple functional perturbations, the symbiotic system consistently
exhibited higher growth rates and greater metabolic stability than
monoculture models,
demonstrating a safety-regulatory mechanism of “metabolic complementation – energy
sharing – homeostasis enhancement.”
Members of our human practice group interviewed Professor Wang Yongming from Fudan
University. Through the interview, we learned about the risks of metabolic toxicity.
After further in-depth exploration, we understood the solutions to reduce metabolic
toxicity. The following are the guidance and research work we have obtained:
In the early days, genetic engineering experiments tended to be regarded as being
successful once a certain level of expression of an introduced gene was achieved.
However, this perspective, for the most part, ignored any physiological changes to
the host organism that might have occurred as a consequence of the introduction of
the foreign DNA, which is called “metabolic burden”. The "metabolic burden" refers
to a phenomenon in which the induced expression of foreign genes in recombinant host
cells or organisms typically consumes a large amount of host cell resources and
brings adverse effects to the host[4]. Due to metabolic burden, the
biochemistry and physiology of host cells may undergo significant changes. For
instance, this can impede the expression of endogenous essential genes within the
cells, leading to a decrease in cell growth rate or a reduction in the synthesis of
their own protein. On an industrial scale, these are reflected in low production
titers and the loss of newly acquired traits, especially during fermentation
runs[5]. The main mechanisms by which overexpression of foreign genes
causes metabolic burden are the stringent response due to amino acid/charged tRNA
starvation[6] and the increase in the number of misfolded proteins caused
by translation errors leads to an increase in the stress on intracellular chaperone
proteins and proteases, thereby activating heat shock and nutrient starvation
responses[7]. To address the adverse effects of metabolic burden, the
following solutions are proposed:
a. Use low, rather than high, copy number plasmid vectors to avoid the host cells
wasting resources in synthesizing unneeded antibiotic resistance marker gene
products.
b. Use strong but regulatable promoters to control. The promoter of the target gene
is in a closed state during the cell growth stage. And at the induction stage, the
promoter is in an open state.
For us, the key point of metabolic toxicity lies in ensuring the growth of the
target yeast while resolving the conflict between yeast growth and gene expression.
In terms of experiments and circuit design:
(1) Firstly, our Saccharomyces cerevisiae is inherently deficient in Ura and His.
Using nutritional defect screening not only ensures safety but also reduces the
competition for nutrients by non-target bacteria.
(2) Secondly, our symbiotic system mainly involves lactic acid bacteria providing
nutrients to the yeast, which can also reduce metabolic toxicity.
(3) In our therapeutic circuit, the yeast has enhanced expression of SOD1 and CTT1.
Although the signal peptide enables their secretion, the residual proteins remaining
in the cells can reduce the oxidative stress caused by toxicity. We will later study
the secretion efficiency of the signal peptide (the ratio of target proteins inside
and outside the cell to explore the balance between therapeutic efficiency and
attenuation effect)
(4) Finally, we selected a dynamic regulation system to control some exogenous genes.
By using the gal promoter to control the expression of downstream target genes,
while the chassis is Saccharomyces cerevisiae with gal80 knocked out. This can
inhibit the expression of exogenous genes in the early growth stage of Saccharomyces
cerevisiae (when it is in the growth phase), and concentrate on expressing the
target genes in the later growth stage[2].
(5) Furthermore, the relevant suggestions from the Advisor also pointed out issues
regarding the orthogonality of the pathways. We drew inspiration from the regulation
of the gal promoter. The CRISPR/dCas9 system only functions in a high-sugar
environment during feeding, and at this time, the dynamic regulatory system in the
fourth point inhibits the expression of the adsorption module. Thus, the functions
of the degradation module and the adsorption module become orthogonal, and the
expression of different modules initially has a temporal sequence.
(6) We constructed plasmids for some of the genes and integrated the remaining genes
into the genome using CRISPR/Cas9. This hybrid system is conducive to achieving the
optimal balance among stability, flexibility, and metabolic burden[3].
2. Product safety
2.1 Product form
(1) Strain selection
In the strain selection process, Safety & Security has always been our top priority.
All candidate chassis bacteria were rigorously compared: Escherichia coli K-12,
although having mature tools, lacks sufficient safety; Nissle 1917 has limited
public acceptance; some Bacteroides and Clostridium species have colonization
advantages but may have controversies regarding toxicity or pathogenicity;
Lactobacillus plantarum, as a food-grade probiotic, has good safety and tolerance,
but has low expression efficiency and is not suitable for engineering; Saccharomyces
cerevisiae, due to its GRAS certification, is widely used in food and medicine,
ensuring safety, but has limited intestinal colonization. Our initial research led
to the first conclusion: no single strain can fully meet the requirements. However,
E. coli Nissle 1917 and Saccharomyces cerevisiae perform well in terms of safety and
engineering tools, while Lactobacillus plantarum has unique anti-inflammatory and
adsorption advantages.
After the engineering cycle, to balance Safety & Security, we finally chose the
food-grade symbiotic system solution, consisting of Saccharomyces cerevisiae and
Lactobacillus DT88. The yeast is responsible for executing the complex synthetic
pathways and therapeutic modules, while DT88, as a food-grade probiotic, not only
enhances the safety and stability of the system but also has natural microplastic
adsorption and intestinal probiotic effects. The interdependence of the two reduces
the possibility of the single strain's uncontrollable spread in the environment,
which is more in line with biosafety requirements.
In the specific design, we additionally introduced multiple safety strategies,
including an inducible expression system to prevent the loop from going out of
control, an environment-dependent mechanism to ensure that the strain cannot survive
long-term outside the host, and an exogenous controllable removal module to ensure
that the engineered bacteria can be quickly removed when needed. And there is also
spore prevention for the yeast chassis bacteria. These measures ensure that the
engineered bacterial system complies with the "controllable, limited, and removable"
biosafety principles, and the details will be explained in other parts.
(2) Product preservation method
Traditional live probiotic and engineered microbial products are highly sensitive to
environmental conditions such as temperature, oxygen, light, and humidity. In
traditional liquid or refrigerated storage, the number of live bacteria decays
rapidly, and a costly cold chain system is required to maintain the efficacy and
safety of the products. The interruption of the cold chain not only affects the
quality of the products but may also accelerate the inactivation and deterioration
of microorganisms.
Lyophilization is a mature industrial solution to address these challenges. This
technology is carried out under low temperature and vacuum conditions, effectively
removing moisture through ice sublimation, avoiding the thermal damage and oxygen
poisoning caused by traditional hot drying. The product after freeze-drying is
transformed into a lightweight and stable solid powder, significantly extending the
shelf life. This powder form of the product can be directly transported and stored
at room temperature without the need for a cold chain, greatly improving the
accessibility and application scope of the product, especially for remote areas
lacking stable low-temperature storage facilities, which constitutes an important
application safety advantage.
Before freeze-drying, the strains are first microencapsulated by the
alginate-chitosan (Alginate-Chitosan, A-C) system. Microencapsulation plays a triple
role in preservation and delivery:
a. Physical protection: The A-C capsules act as the first physical barrier,
protecting the symbiotic strains inside from ice crystal formation and dehydration
stress during freeze-drying;
b. Enhanced mechanical stability: The chitosan-alginate complex forms through
electrostatic interactions, showing higher mechanical stability, lower drug leakage
rate, and less burst release compared to pure alginate capsules. This mechanical
integrity is crucial for subsequent freeze-drying and storage;
c. Targeted delivery: The A-C coating gives the product the ability to resist
degradation by gastric acid and bile salts, ensuring that the live bacteria can be
released in a controlled manner in the target sites, such as the small intestine or
colon, after passing through the stomach.
The preservation strategy shifts the safety paradigm of the product from "biological
activity" to "biological dormancy". The greatest safety risk of live microbial
products stems from their metabolic activity and reproductive ability. Through
freeze-drying, the product's state shifts from "potential biological risk" to
"physical and metabolic locked solid biological material". This forward-looking
safety design shifts the focus of safety from post-treatment clearance or inhibition
to pre-treatment activity locking, providing a more reliable guarantee for the
long-term safety of the product.
In addition, the freeze-dried powder form provides physical isolation, while
metabolic locking eliminates the risk of horizontal gene transfer of live bacteria
to the environment or foreign bacterial communities during storage and
transportation, maximizing the biological safety during the storage period.
2.2 The use of the product
When the product is used, ensuring that the yeast can function properly in the human
body is also an important aspect of safety. Our team mainly considers this issue
from two main directions: Firstly, the strain needs to survive normally in the human
body. Secondly, the target function of the strain must be guaranteed in the human
body.
(1) Strain life support
a. Symbiotic system support
One of the core advantages of this product lies in its unique dual-bacteria
symbiotic system. This system is composed of genetically engineered Saccharomyces
cerevisiae and probiotic Lactiplantibacillus plantarum DT88. This design is not a
simple stacking of strains, but rather the creation of a miniature ecosystem. The
primary purpose of this design is to provide key life support for the engineered
yeast in the complex intestinal environment, thereby directly enhancing the Safety
and Security of the product. We mainly investigated the advantages of this system in
terms of "metabolic mutualism" and "stress tolerance", and also analyzed the
potential risks and provided solutions.
a.1 Nutrient mutualism
In an environment rich in nitrogen sources in the intestinal tract, Saccharomyces
cerevisiae actively secretes various amino acids, providing essential growth factors
for amino acid-deficient Lactobacillus plantarum [8]. This targeted
nutrient supply significantly promotes the growth and survival of Lactobacillus
plantarum in a co-culture system, with its biomass being much higher than that in a
single culture condition[9]. In return, Lactobacillus plantarum can
decompose complex carbohydrates (such as lactose) that yeast cannot utilize and
convert them into monosaccharides like glucose, providing additional energy sources
for yeast. This two-way nutrient supply creates a stable "nutrient support complex"
in the resource-variable intestinal environment.
① Safety and Effectiveness Goals: Enhance effectiveness: This mutually beneficial
symbiotic relationship ensures that
our engineered bacterial system can maintain stable biomass and metabolic activity
even in the fierce competition of the native intestinal microbiota. This directly
guarantees the continuous and stable secretion of therapeutic proteins and ensures
the reliability of product functionality.
② Endogenous biological containment: More importantly, this "nutrient interlock"
mechanism builds a survival-dependent relationship. It firmly binds the two strains
together, significantly reducing the possibility of a single engineered strain
(especially yeast) surviving independently after accidental leakage into the
external environment. Once deprived of the nutritional support of Lactobacillus
plantarum, the survival ability of the engineered yeast will be significantly
limited. This is an important endogenous containment link in our biosafety design,
aiming to minimize environmental release risks from the source.
a.2 Stress tolerance
① The advantage in the intestinal environment - bile salt protection[10]:
Saccharomyces cerevisiae can secrete extracellular peptide substances known as
"reactivation factors". In the upper part of the intestine, high concentrations of
bile salts have a strong killing effect on most oral microorganisms. These
"reactivation factors" act like a "protective shield", significantly protecting the
coexisting lactic acid bacteria from the stress damage caused by bile salts,
ensuring that the entire symbiotic system can safely pass through the stomach and
the upper part of the small intestine in an alive form.
Safety and effectiveness goals: The purpose of this mechanism is to ensure that
the
engineered bacteria with an "effective dose" can reach the target action area in the
lower part of the intestine. This directly overcomes one of the biggest challenges
of oral live bacterial preparations - low bioavailability, thereby ensuring the
therapeutic effect of the product. Of course, we continue to optimize tolerance in
that direction.
② The advantage in the intestinal environment - ethanol detoxification[11]: In
the
anaerobic environment deep in the intestine, yeast undergoes alcohol fermentation,
producing ethanol that also inhibits its own growth. Research has found that
co-culturing with Lactobacillus plantarum can significantly improve the ethanol
tolerance of Saccharomyces cerevisiae.
Safety and effectiveness goals: The purpose of this mechanism is to extend the
functional lifespan of the engineered bacteria in the intestine. By alleviating the
self-toxicity of ethanol, the symbiotic system can maintain higher activity and
population density in the intestine for a longer time. This not only means a more
lasting therapeutic effect but also enhances the stability of the system, avoiding
the interruption of the therapeutic effect due to the premature collapse of the
population, ensuring the predictability and safety of the entire treatment cycle. It
also more clearly demonstrates the dependence of the two, reducing the potential
risk of the engineered yeast cells leaving the symbiotic system and causing harm to
the human body.
③ The advantage in the intestinal environment - inhibiting immunity[9]:
Anti-inflammation itself is a therapeutic effect, but this effect also reduces
the
risk of the symbiotic system being attacked by the immune system. Studies have
confirmed that the yeast-lactobacillus co-culture system can synergistically
significantly increase the production of IL-10 and inhibit TNF-α. This indicates
that this system can actively push the immune response towards a tolerant state.
Safety and effectiveness goals: This mechanism is an extension of the
thinking of
anti-inflammatory assistance, which involves the analysis of the immune
microenvironment around the symbiotic system. Enhancing its anti-inflammatory
effect
is also to reduce the stress of the immune response on the symbiotic system, and
the
continued survival of the strain can also continue to exert the
anti-inflammatory
effect, forming a positive feedback loop. Of course, this is based on the
symbiotic
system, and if the engineered yeast cells leave the symbiotic system in the
intestine, they will be significantly stressed.
a.3 Security considerations and risk control in symbiotic systems
Although symbiotic systems offer significant functional advantages, their complexity
also introduces risks that must be managed with caution. Our design has always taken
the following risk points as the core consideration and has adopted multi-level
mitigation strategies. See Attachment:
Security
Considerations and Risk Control forSymbiotic Systems.
b. Anti-gastric acid bile salts
Following the iGEM "Safe-by-Design" concept, we recognize that a responsible
engineering project requires a safety system that is actively constructed and
multi-layered with redundancy. The design of the alginate-chitosan microcapsules is
the first and most crucial physical defense line in our entire safety framework.

>
b.1 Ensure targeted delivery and biological containment
Although the symbiotic system we designed enhanced the survival ability through its
internal "stress tolerance" mechanism, to safely apply a living, genetically
engineered organism to the human body, the problem of protection during oral
delivery must be solved. The primary challenge faced by oral live bacterial
preparations is the stressful environment in the upper part of the digestive tract.
The strong acidity in the stomach (pH 1.5–3.5) and the bile salts in the upper part
of the small intestine not only cause the live bacteria to lose their viability, but
more importantly, we need a physical barrier to ensure that the engineered bacteria
do not interact with non-target environments before reaching the target action area,
thereby minimizing the risks of accidental exposure and environmental release to the
greatest extent.
To address this challenge, we designed and verified a dual-layer composite
microcapsule system. This system consists of a sodium alginate core and a chitosan
shell, using the electrostatic interaction between the two GRAS-level materials to
form a denser and more stable physical barrier that can resist stomach acid and bile
salts.
*PS:
Anti-acid mechanism: In the highly acidic environment of the stomach (pH
1.5–3.5), the negatively charged sodium alginate undergoes protonation to form
insoluble alginate, and its gel network contracts and becomes denser, thereby
physically encapsulating the internal bacteria firmly. At the same time, the
outer chitosan, due to the protonation of its amino groups, acquires positive
charges and forms a stable polyelectrolyte complex with the alginate, further
enhancing the structural integrity and acid resistance of the capsule,
effectively resisting the erosion of gastric acid[12].
Anti-bile salt mechanism: After entering the small intestine, the high
concentration of bile salts acts as a biological surfactant and can damage the
microbial cell membranes. The dense composite microcapsules provide an effective
physical barrier to prevent the penetration of bile salts, protecting the
internal bacteria and ensuring their survival and reaching the lower part of the
intestinal tract[13].
Through an in vitro experiment simulating the gastric environment, we rigorously
verified the protective ability and biological containment effectiveness of the
microcapsules. The experimental results clearly demonstrated that this composite
microcapsule could effectively resist the erosion of simulated gastric fluid and
successfully contain the live bacteria within the capsule. Its pH-responsive design
ensured that the microcapsule would disintegrate only when it reached the specific
pH environment of the intestine, meaning that the GMM could only be released at the
predetermined application site. This "targeted release" mechanism significantly
reduces the possibility of its colonization in non-target areas or being maliciously
extracted and misused, reflecting our careful consideration of the Security aspect
of the project.
b.2 A safety commitment based on absolute harmlessness
Our design starts from the first principle of material selection: absolute safety.
To minimize risks from the very beginning, we systematically investigated various
biocompatible materials and strictly limited ourselves to the GRAS (Generally
Recognized as Safe) level.
We learned that both are natural polysaccharides, having decades of safe application
history in the food and medical fields. They are non-toxic and non-irritating to the
human body, and their biodegradable products are also safe. This choice ensures that
our physical containment system itself does not introduce any new chemical or
biological safety risks, fully meeting the basic requirements for project safety set
by iGEM.
*PS:
Sodium Alginate: This is a polysaccharide extracted from natural brown algae and
is recognized as a GRAS substance by the US Food and Drug Administration (FDA).
It is widely used in the food industry as a thickening agent and stabilizer, and
is also commonly used in the medical field for wound dressings and drug
delivery. It has good biocompatibility, cannot be digested or absorbed by the
human body, and will eventually be excreted through feces, with no toxic side
effects[14].
Chitosan: It is a natural polysaccharide obtained by deacetylation of chitin from
the shells of crustaceans such as shrimp and crab. Chitosan also has excellent
biocompatibility, biodegradability, and extremely low toxicity. It has been
widely studied and applied in food, cosmetics, and drug delivery
systems[15].
At this point, we have completed the fabrication of the "physical armor" for the
strain. Our ultimate goal is to combine this external physical containment strategy
with our internal genetically engineered containment strategy to form a
dual-protective system in order to address more complex security challenges and
maximize the safety of users, communities, and the environment.
(2) Strain safety effect
Engineered bacteria can survive stably and function in the intestinal tract, and
also ensure safety and harmlessness during actual action, bringing positive effects
to the host.
a. Specificity and harmlessness
The degradation enzymes secreted by the enzyme degradation module, such as PETase
and MHETase, have clear substrate specificity and only act on the microplastic
structure, without degrading the host's proteins or polysaccharide components.
The products after microplastic degradation are small molecule monomers or
oligomers, without toxicity, and have no adverse effects on the host or the
intestinal microbiota.
The adsorption module uses the HFBI surface to display protein to bind to
microplastic particles, and the target object is limited to hydrophobic surfaces,
without conflicting with the attachment sites of intestinal tissues or normal
microbiota.
b. Reduce potential damage caused by microplastics
The synergistic effect of the enzyme degradation and adsorption modules can
accelerate the clearance of microplastics, shorten their residence time in the
intestine, avoid long-term physical stimulation and chronic inflammation, and at the
same time reduce the risk of oxidative stress and barrier disruption caused by them,
and reduce the burden on the intestinal environment.
c. Promote intestinal health and protection
The SOD1, TFF3, and CTT1 secreted by the treatment module, respectively, undertake
the functions of eliminating free radicals, repairing the mucosal barrier, and
decomposing peroxides. These molecules can work in synergy with the host's natural
defense factors to alleviate oxidative stress and inflammation, not only being
harmless but also protecting intestinal homeostasis and reducing further harm to the
bloodstream.
d. Stable colonization and immune friendliness
Lactobacillus has acid resistance, bile salt resistance, and colonization ability.
While ensuring the safe delivery of yeast to the intestine, it can form protective
biofilms locally, and it can symbiotically coexist with yeast and metabolize
complementarily, enabling the engineered microbiota to exist stably and express
normally in the intestine. The formed system can jointly adsorb microplastics,
reduce their contact with the intestinal mucosa, and exert anti-inflammatory and
immune regulatory effects through the secretion of lactic acid, amino acids, etc.,
which is beneficial and harmless to the human body.
(3) Disposal of products
a. Safety Route Review
To further ensure safety, we began to explore biological containment strategies. We
believe that the safest and most effective method is through a mode:
signal-controlled sterilization. It involves sensors and effectors, just like the
reflexes in the human body, simple yet highly effective. Our team conducted a
partial systematic review of sensors and effectors based on the requirements of our
project. Although they come from different platforms, they can still provide ideas
for our biological containment strategies.
a.1 Sensor
Based on our project application scenarios, we classify the input signals into two
different types: 1) Input signals that are regulated by the inherent conditions of
the body - "passive" input signals, 2) Input signals that require external
conditions to be applied for regulation - "active" input signals.
a.1.1 "Passive" input signal
① The temperature control system of the engineered bacteria utilizes the inherent
temperature difference between the internal and external body environment. This
enables the triggering of cell death at lower external temperatures. Advantages: It
utilizes the natural temperature difference and does not require manual
intervention. Disadvantages: It needs to consider the possibility of bacterial
escape during the transportation process in the presence of low temperatures and hot
weather. For our project, this issue can be addressed. Our product is preserved
through freeze-drying technology, ensuring that it will not be abnormally activated
before entering the body. Regarding the hot weather issue, we believe that the
environment where the excretions are located will not exceed the body temperature.
Of course, this reminds us to pay attention to the temperature threshold for
activation. In addition, the method of activation is also diverse. For example, the
NMU-China 2024 and Athens 2022 teams have all applied this sensing method using
different technologies in the safety module[16][17].
② Besides temperature, we also referred to other lines that can sense the
differences in internal signals. We applied the following ideas to our degradation
and adsorption modules, which are in line with our previous engineering cycle goals.
— The deep part of the human intestinal tract is a typical hypoxic/microoxic
environment, while the external environment (air, water) is oxygen-rich. This huge
oxygen concentration gradient provides an excellent opportunity for designing safety
switches. The PTSH-Taiwan 2023 team utilized an oxygen-sensitive promoter in their
project to control the expression of toxins, ensuring that the engineered bacteria
could be effectively eliminated when they were in the oxygen-rich environment
outside the intestine.
— SZU-China 2021 developed a glucose starvation system involving the use of glucose
starvation receptors. When glucose is absent, these receptors will induce the
expression of downstream genes. This system uses the inherent difference in glucose
concentration between the intestinal tract and the external environment to induce
bacterial death in the external environment.
③ The phosphate receptor is a receptor proposed by SZU-China 2021. It utilizes the
difference in phosphate concentration between the inside and outside of blood
vessels to prevent engineered bacteria from entering the circulation through minor
vascular ruptures in the intestinal tract during inflammation. We believe that this
system ingeniously solves the problem of preventing bacteria from entering the
bloodstream, which is a problem that other teams have rarely considered. Regarding
the issue of blood entry, we currently have a preliminary blocking effect through
symbiosis to enhance the intestinal mucosal barrier[18][19].
a.1.2 "Active" input signal
These two types of signals can fully exert the stabilizing effect of the artificial
control circuit and can effectively control the intensity of the input signal for
testing, ultimately achieving the optimal result. In terms of nature, they include
physical signals: light, electricity, temperature, etc., and signal perception.
Here, taking the light receptor as an example, SZPT-CHINA 2021 uses a simple
promoter to sense blue light. This regulatory system is simple and stable. In shady
areas, when the light intensity is insufficient, it may be necessary to artificially
apply blue light to induce death. Chemical signals are mostly regarded as "drug"
signals. Our NMU-China team has discussed the use of "drugs" in the safety module in
previous editions of IGEN. This inspired us to compare the advantages and
disadvantages of different "drug" signals in intestinal applications to obtain the
best option.
a.2 Effector
a.2.1 Toxin-antitoxin,(TA) system
Toxin-antitoxin (TA) system is one of the most fundamental and classic tools for
constructing "self-destruct switches" in living organisms. Such systems typically
consist of a stable toxin that can disrupt key physiological processes of the cell,
and an unstable antitoxin that can neutralize this toxin. By placing the expression
of one or both components under the control of specific environmental promoters, the
team can program the cells to survive only in the pre-defined ecological niche. The
following reviews how previous iGEM teams have creatively applied various TA
systems.
— CcdB/CcdA system: This system utilizes the CcdB toxin, which can lethally inhibit
the DNA gyrase (an enzyme crucial for DNA replication). Its effect is neutralized by
the CcdA antitoxin protein. For instance, the GEMS_Taiwan 2022 team employed this
system[20].
— Hok/Sok System: This is an I-type TA system. In this system, the toxin (Hok) is a
protein that damages the cell membrane, while the antitoxin (Sok) is an unstable
antisense RNA that binds to the mRNA of Hok, preventing its translation and
promoting its degradation. For example, the Wageningen_UR 2021 team has used this
system[21].
— MazF/MazE system: This is a thoroughly studied type II TA system. Among them,
MazF is a stable mRNA endonuclease that can cut cellular mRNA, while MazE is an
unstable protein antitoxin that will be rapidly degraded by cellular proteases. For
example, the Fudan 2020 team used this system[22].
— VapD/VapX system: This is a relatively novel type II TA system. In this system,
VapD is a ribonuclease toxin, and VapX is its protein antitoxin. The
Vilnius-Lithuania 2021 team was the first to characterize and apply it in
iGEM[23].
— Holin/Antiholin System: Holin is a toxin protein that can "make holes" in the
cell membrane. These holes allow other molecules (such as endolysins) to enter and
degrade the cell wall. Antiholin is an antitoxin that can bind to Holin and inhibit
its hole-making activity. For example, the Cornell 2020 team used this
system[24].
a.2.2 Hijacking Apoptosis
This strategy no longer relies on exogenous toxins, but instead precisely controls
cell fate by activating the endogenous, highly conserved programmed cell death
(apoptosis) pathway within eukaryotic cells.
— Bax is a core pro-apoptotic protein in the apoptotic pathway of mammalian cells.
When expressed in yeast, Bax can target the outer mitochondrial membrane, form
channels, disrupt mitochondrial function, and ultimately induce apoptotic-like death
in the cells. The GreatBay-SCIE 2023 team envisioned in their project a
self-destruct switch using Bax, for clearing the engineered yeast after completing
its task. Their design concept is to induce the expression of Bax after the
engineered bacteria leave the target environment, thereby triggering the endogenous
death program[25].
a.2.3 Targeted gene system
These effectors carry out "execution" by permanently modifying the genetic material
of cells. Once triggered, there is almost no possibility of escape, representing the
cutting edge of current biosecurity design[26].
— CRISPR-based genome degradation: The CRISPR/Cas9 system is like a programmable
"genetic scissors". By expressing Cas9 enzyme and one or more guide RNAs (gRNAs),
multiple essential genes in the genome can be precisely cut. The large number of DNA
double-strand breaks (DSBs) produced in a short period of time will far exceed the
cell's repair capacity, leading to genome collapse and irreversible cell death.
AQA_Unesp in 2017 designed a more sophisticated light-controlled switch. They split
the Cas9 protein into two inactive fragments and fused each with a photoreceptor
protein (such as CRY2 and CIB1). Only under blue light irradiation can the two
fragments reassemble into an active Cas9 enzyme, thereby cutting the targeted
multiple bacterial essential genes (such as dnaN, rpoC, etc.), achieving precise and
temporal control of the self-destruct program.
— Site-specific recombinase: The Cre/LoxP system is a representative example. The
Cre recombinase can recognize two LoxP sites and irreversibly excise or reverse the
DNA fragment between them. By placing the essential gene or its promoter between the
two LoxP sites, the function of the gene can be permanently disrupted by inducing
the expression of Cre. The PTSH-Taiwan 2023 team designed an extremely sophisticated
double-layer logical switch for closely related yeast. This design decouples the two
steps of "arming" and "firing" of the system in terms of time and space. In the
intestine (a safe environment with high hydrogen sulfide), both Cre recombinase and
the antitoxin are expressed simultaneously. The Cre enzyme permanently removes the
terminator upstream of the toxin gene, completing the "arming", but the cells
survive due to the presence of the antitoxin. When the cells leave the intestine
(low hydrogen sulfide, high oxygen), the antitoxin stops expressing, and at the same
time, an oxygen-sensitive promoter activates the expression of the toxin, completing
the "firing", killing the cells. This strategy of using irreversible genetic
modification for "state memory" greatly enhances the robustness of the
system[27].
a.2.4 Engineered Dependency
Unlike the active introduction of toxins, this strategy achieves passive containment
by modifying the metabolic network of cells to make them dependent on a specific
environment or symbiotic partner for survival.
— Conditional essential genes: The NEU-CHINA 2022 team did not use CRISPR as a
direct lethal weapon, but rather regarded it as a tool for constructing a safe
pathway. They utilized the CRISPR technology to replace the natural promoter of a
yeast essential gene - the heat shock transcription factor HSF1 - with a copper
ion-inducible promoter CUP1. Due to the much higher concentration of copper ions in
the intestinal environment compared to the outside, the survival of this engineered
yeast is tightly bound to the intestinal environment. Once it leaves, it will die
because it cannot express the essential gene[28].
— Reciprocal nutritional defect: The Calgary 2020 team successfully constructed a
"metabolic lock" system consisting of two yeast strains. Strain A was modified to be
a leucine nutritional defect strain, but it can over-secrete tryptophan; Strain B,
on the contrary, is a tryptophan nutritional defect strain, but it can over-secrete
leucine. These two strains can only "feed" each other and grow together when
co-cultured. Once they escape into the environment and are diluted and separated,
the individual strain will "starve" due to the inability to obtain the necessary
amino acids. The evolutionary stability of this strategy is higher because cells
need to escape through more difficult recovery mutations rather than simple gene
inactivation mutations[29].
b. The design of the stop switch in our project
After reviewing the above sensors and effectors, we gradually began to build the
prototype of the safety circuit.
b.1 First-generation design: A temperature-sensitive switch for in vitro biological
inhibition
The core objective of the first-generation design was to achieve in vitro biological
containment, aiming to prevent engineered bacteria from entering the environment and
causing potential ecological risks due to unexpected survival or horizontal gene
transfer (HGT). To this end, we constructed a safety switch that could quickly
initiate programmed cell death when the engineered bacteria completed their tasks
within the host and left the host. According to our reviewed safety pathways, we
believe that temperature is a relatively ideal sensing signal.
This design utilizes the stable physical temperature difference between the host's
intestinal tract (approximately 37°C) and the external environment (lower than 30°C)
as a signal. Driven by this temperature gradient, an "inside-off, outside-on" logic
gate is achieved, which we call the Cold-Triggered Kill Switch. Its core is a
reverse logic circuit: the heat shock promoter PHSP26 drives the expression of the
repressor protein TetR to turn off the killing genes at 37°C; when the temperature
drops to the external level, the expression of TetR is downregulated, thereby
releasing the inhibition on the downstream killing module.
Inspired by the feedback from the effectors, we concurrently developed two killing
effector modules, both targeting cellular genetic material:
Scheme 1 (Metabolic Blockade):
Through the Cre/LoxP system, the key gene ADE2 of the purine synthesis pathway is
knocked out at low temperatures, thereby blocking DNA replication.
Scheme 2 (Genome Degradation):
Express the NucA nucleases of Pseudomonas aeruginosa directly at low temperatures to
efficiently degrade genomic DNA.
Experimental and bioinformatics analyses have verified the feasibility of this
design. However, although this design can effectively control leakage in vitro, it
has a fundamental flaw: it lacks an in-body emergency clearance mechanism that is
actively controlled by the user. Based on user practice and expert consultation
feedback, this aspect is crucial for building patients' trust and sense of
security.
b.2 Second-generation design: Drug-induced toxin-antitoxin switch
To address the lack of in-body control in the first-generation design, after review,
we concluded that the drug signal has advantages for in-body control. In terms of
effectors, the second-generation design introduces a drug-induced toxin-antitoxin
(Toxin-Antitoxin) system, giving users the ability to actively remove engineered
bacteria from the body when necessary. This system is based on the K1 killing toxin
system of Saccharomyces cerevisiae. Among them, the unprocessed K1 toxin precursor
protein contains an immune region inside, which can protect the host from harm and
play the role of "antitoxin". The mature αβ heterodimer that is processed and
secreted outside the cell is the "toxin", which can lyse the yeast cell wall and
cause death.
The system's switch is controlled by exogenous arabinose: in the normal state, the
system remains in a safe condition. When the user orally ingests arabinose, the
inducible promoter pAra is activated, driving the efficient expression of the toxin
module. Eventually, a large amount of mature toxin proteins is produced and
secreted, whose killing effect exceeds the immune protection ability of the organism
itself, thereby achieving precise and targeted elimination of the engineered
bacteria in the body.
Literature research and preliminary experimental verification have confirmed the
feasibility of this scheme. However, this system still has potential areas for
optimization: Firstly, the arabinose promoter has the risk of background leakage
expression; Secondly, the individual differences in the host's absorption and
metabolism rate of arabinose may affect the stability of the induction efficiency.
These issues will be the focus of our subsequent work.
b.3 Further prospects and outlooks
Based on the practical experiences of the two generations of designs, we identified
the shortcomings of the safety circuits through comparison. Therefore, we have
planned the future research direction, aiming to build a more rigorous, intelligent,
and secure biological containment system.
b.3.1 The rigor and robustness of the safety switch: Regarding the leakage issue of
the arabinose system, future work will focus on constructing more precise regulatory
circuits, such as introducing toxin-antitoxin neutralization modules or designing
logical "AND gates" that require dual signal inputs, to fundamentally prevent
unintended toxin expression.
b.3.2 The uniformity and predictability of clinical applications: To overcome the
impact of individual metabolic differences, subsequent research needs to explore
standardized dosing schemes or screen and validate alternative inducers with clearer
metabolic pathways and smaller individual differences.
b.3.3 The intelligence and autonomous regulation of the system: To enhance the
precision of intervention, future work can integrate the quorum-sensing module into
the safety system. This will enable the engineered bacterial population to
autonomously activate the growth inhibition program when reaching a specific density
threshold, achieving dynamic balance and avoiding excessive proliferation.
b.3.4 Multiple redundant biological containment strategies: To address extreme
situations (such as the engineered bacteria breaking through the intestinal
barrier), the final safety design should include multiple, orthogonal containment
mechanisms. For example, on the basis of the existing killing switch, additional
nutritional defect designs can be integrated to ensure that the engineered bacteria
cannot get out of control in any environment.
c. Spore prevention
Although Saccharomyces cerevisiae belongs to BSL-1 microorganisms, its ability to
form spores may lead to Release Beyond Containment, which is explicitly listed as a
risk point in the iGEM safety policy[30]. Spores have extremely strong
environmental resistance, and once formed, they are difficult to effectively
control[31].
In our project, the engineered yeast will be excreted from the host after completing
its in vivo therapeutic task. If these yeasts encounter specific nutritional stress
(such as nitrogen source deficiency) in the external environment, they may initiate
the spore formation process[32]. This will bring two unacceptable high-level risks:
① Environmental release and ecological risk: Spores have high resistance to drying,
ultraviolet rays, and common disinfectants and can survive for a long time in the
natural environment. If the spores carrying the engineered circuit spread in the
environment, it may lead to gene drift, causing unpredictable and long-lasting
impacts on the local microbial ecosystem.
② Genetic information diffusion risk: Spores are dormant carriers of genetic
material. Their diffusion is equivalent to the "out-of-control spread" of our
engineered gene circuit in the environment, which increases the risk of horizontal
gene transfer (HGT) and may lead to the unintended use or abuse of our synthetic
biology tools.
To eliminate this risk, we used genetic engineering techniques to delete the core
regulatory gene IME1 that controls spore formation, thereby depriving the yeast of
the ability to form spores. IME1 is the "master switch" for spore formation,
functioning as a central processor that can integrate various signals (primarily
nutritional status) from both inside and outside the cell and make the final
decision on whether to initiate spore formation. Its absence will prevent diploid
yeast from forming spores under any induction conditions, without affecting mitosis
and normal metabolism. We chose to delete IME1 for the following reasons:
① Absolute core position: As the highest regulator of the spore formation signaling
pathway, deleting IME1 can block the entire process from the very top, ensuring that
the spore formation program cannot be initiated regardless of changes in the
external environment[33].
② High specificity, no off-target effects: The function of IME1 is highly specific,
and its main role is to initiate meiosis. Studies have confirmed that deleting this
gene will not have a significant negative impact on the yeast's normal mitosis
(i.e., normal growth and reproduction) and other core metabolic
functions[34]. This ensures that our engineered bacteria, when exerting
therapeutic effects, will not have their biological properties and stability
affected.
③ Permanent and stable strategy: The gene knockout achieved through genome editing
is permanent and genetically stable. This means that once the construction is
successful, all descendant strains will inherit the "no spore" characteristic,
without the need to worry about functional restoration or escape, and there will be
no new safety issues in the future.
We have designed a set of schemes using CRISPR/Cas9-mediated homologous
recombination to replace the coding region of IME1 with a fragment carrying His and
Ura, and ensure the successful knockout through resistance screening and PCR
verification.
Literature studies have proved that this strategy can eliminate the ability to form
spores while maintaining the biological stability and application value of the
strain. In subsequent experiments, we will place the modified yeast and the
wild-type yeast in standard spore formation media (such as potassium acetate medium)
and observe them under a microscope for several days. The wild-type should be able
to see typical tetrad spores, while the modified strain will completely fail to form
any spore structure.
In conclusion, this measure effectively avoids the risk of spore spread, ensuring
that our chassis strain meets the strict biological safety requirements of iGEM.
3. Part and Material Safety
3.1 Overview
(1) Our Basic Part
We designed 20 basic parts in total this year, integral to our whole cycle.
(2) Our Composite Part
This year, we designed a total of 13 composite components, each assembled from
different Basic Parts, forming the core of the circuit system.
3.2 Parts
(1) Toxic protein
a. K1 Killer Toxin and NucA
The mature K1 toxin is an ionophoric protein toxin that kills sensitive cells by
forming pores in the plasma membrane. The K1 killer preprotoxin is a precursor
protein that also functions as an antitoxin, protecting the host cell from the
mature K1 toxin.
The nucA gene from Serratia marcescens encodes a non-specific endonuclease that
degrades both DNA and RNA, often used as a suicide gene in biocontainment
circuits.
These components form the core effectors of our “self-destruct switch,” designed to
enhance the project's biological containment capabilities. Their mechanism of action
exhibits no known toxicity to mammalian cells. Furthermore, their expression within
our circuit is tightly regulated by conditionally activated promoters (such as the
temperature-sensitive promoter pHSP26 or the arabinose-inducible pAra). This ensures
system safety under non-induced conditions and, due to their non-novel, low-risk
nature as toxins, poses minimal risk of malicious misuse.
(2) Functional protein
a. Fast-PETase and MHETase:
These two enzymes originate from non-pathogenic microorganisms and serve as core
functional proteins for degrading microplastics. They exhibit high substrate
specificity, acting exclusively on the chemical bonds of plastic polymers such as
PET or MHET, without degrading biological macromolecules like proteins or
polysaccharides within the gut. Consequently, they perform their degradation
function while remaining completely safe for both users and the gut microbiota. As
environmental remediation tools, they lack the potential for weaponization or
malicious disruption, presenting low dual-use risks.
b. SOD1,CTT1 and TFF3:
These proteins are key therapeutic components in our project. SOD1, CTT1 (catalase),
and TFF3 (trifolium factor) function as antioxidants and intestinal mucosal repair
agents, respectively, benefiting the host. Derived from safe species or homologous
to endogenous human proteins, they exhibit excellent biocompatibility with no known
toxicity or immunogenicity risks. These therapeutically functional proteins carry no
dual-use risks and cannot be repurposed for malicious purposes.
(3) Genetic Engineering Tools
a. dCas9:
dCas9 is an RNA-guided DNA-binding protein that lacks nuclease activity and serves
as a modular platform for targeting specific DNA sequences.Although dCas9 loses its
ability to cleave DNA, but fully retains its function of binding to specific DNA
sequences under the guidance of single-guide RNA (sgRNA). This characteristic makes
it a modular DNA-binding platform.In this project, the platform targets BAR1 to
enable MXI1 to inhibit gene transcription. Unlike traditional CRISPR/Cas9, dCas9 is
solely used for gene expression regulation (CRISPRi) without cleaving DNA, thus
eliminating the risk of gene drive.
b. Cre Recombinase
Cre is a site-specific recombinase targeting lox sequences, capable of recognizing
multiple variants (loxP, lox66, lox71, lox2272, etc.) and cleaving or inverting the
intervening DNA sequences depending on their orientation. Literature and reviews
indicate that Cre exhibits comparable recognition efficiency for single-mutant sites
(e.g., lox66 or lox71) to that of wild-type loxP. However, lox72—generated by
recombination between lox66 and lox71—exhibits significantly reduced affinity for
Cre, thereby preventing further recombination[35].
Additionally, the Cre recombinase sequence we used is BBa_K1680007, a yeast
codon-optimized recombinase.
Based on the above, the function of Cre recombinase is strictly confined to
performing a single, irreversible excision of a gene fragment (such as knockout of
ADE2). It serves as a controlled “genetic scalpel” rather than a self-propagating
element.
To sum up, the powerful potential of gene editing technology, along with an
explanation of the project's mitigation measures, such as not publicly releasing key
gRNA sequences, conducting bioinformatics screening on all designed sequences, and
implementing strict physical and digital asset access controls within the
laboratory.
3.3 Materials
(1) Physical containment Materials
a. Sodium Alginate and Chitosan
Sodium alginate and chitosan form the core of our physical containment strategy.
Both are U.S. FDA-certified “Generally Recognized as Safe” (GRAS) substances, widely
used in the food and pharmaceutical industries, non-toxic, and biodegradable. In
this project, they are utilized to construct microcapsules, providing an effective
physical barrier for engineered bacteria. Far from posing a risk, this constitutes a
proactive measure to enhance project safety. It effectively prevents the release of
bacterial strains into non-target areas, thereby elevating the overall safety and
security standards of the project. For specific information, please refer to
“Product Safety -
Anti-Gastric Acid and Bile Salt Module.”
b. Chemical Inducers
Arabinose is a naturally occurring sugar and a safe food ingredient. It is scarcely
absorbed by the human body and exhibits extremely low utilization within the body.
In our project, it is suitable for use as an “active” input signal to trigger a
user-controlled “self-destruct switch.” Its safety lies in providing users with a
reliable, harmless means to actively clear engineered bacteria from their bodies.
This “drug”-induced mechanism grants users or physicians ultimate control over the
engineered bacteria's lifecycle. This serves as a positive security feature,
ensuring the biosafety and controllability of the induction process within
engineered Saccharomyces cerevisiae[36].
(2) Target Material
a. PET
Microplastic powder is the target material for processing in our project, not a
project byproduct. We have identified key risks associated with handling
microplastics (particularly dry powder) during experiments, such as inhalation
hazards and electrostatic dispersion. To ensure personnel safety, all relevant
operations are conducted within chemical fume hoods or biosafety cabinets. Mandatory
use of personal protective equipment (PPE), including N95 respirators, safety
goggles, and gloves, is enforced to eliminate exposure risks. For details, please
refer to the Microplastics
Laboratory Safety section.