Team:NDNF China/Proof Of Concept
Overview
There is no denying that synbio is a hugely potential field. But for engineered strains designed to function beyond laboratories, some obstacles are holding them back. NDNF_China hopes to help these strains work beyond the laboratory in a safe, stable and traceable way.
Here we present Hidro: a hydrogel system enclosing engineered bacterial strains. The outer layer of Hidro is a compact shell, offering both protection and containment, preventing the strains from escaping into the wild; the inner core of Hidro provides a supportive environment for them under harsh conditions, thus enabling their stable function; A genome-integrated Tracing and Control system offers tracking and specific killing of engineered strains in case of emergencies. We have experimentally demonstrated that Hidro can be implemented in diverse scenarios, such as heavy metal sensing, food-quality detection, drug secretion, etc.
The Hidro system has the great potential to promote synthetic biology applications beyond the laboratory.
Figure 1: The Design diagram of Hidro system.
The Design of Hidro
The production of the Hidro system is divided into two main sections.
The first section is the material part of Hidro: This shell material is extremely tough and resistant to fracture, yet retains permeability to small molecules. It ensures the biocontainment and stability of the Hidro system in which the engineered strains can carry out engineered functions;
The second section is the genetic design part of Hidro: the genetic designs including customized barcode and kill switch are introduced into the bacterial genome. It could provide a "Trace and Control" system to allow dynamical monitoring of the engineered microbes beyond the laboratory.
The material part of Hidro
Figure 2：The procedures to encapsulate cells in Hidro hydrogel system.
The Hidro system includes a two-layer structure: (1) an alginate-based hydrogel core and (2) a tough hydrogel shell.
To incorporate living cells into the inner core (Figure 2), liquid cultures of Escherichia coli were mixed with alginate in 50 μl droplets that were crosslinked with calcium ions to form spheres. The alginate core is preloaded with nutrients for growth support.
Then the cores were coated with the tough polyacrylamide-alginate hydrogel layer. The shell is an extremely stretchable and tough hydrogel, made by mixing two types of crosslinked polymers: ionically crosslinked alginate, and covalently crosslinked polyacrylamide. This shell material is extremely tough and resistant to fracture, yet retains permeability to small molecules (pore size with 5-50 nm). The design of this shell was inspired by the article published on Nature[1].
Table 1: A table of hydrogel shell raw materials and their function
To make a Hidro bead, we first mixed sodium alginate with liquid cultures of bacteria to form spheres. Subsequent soaking in calcium chloride solution formed the core. Sodium alginate was mixed with Acr-Bis solution, then preloaded into a mould, where the previously made core could be lodged in. After its solidification, it could be soaked into the solution of MES and crosslinkers.(for details, please go to Protocols and Engineering Success)
Optimize Hidro's materials to reduce toxicity
We plan for Hidro to be used in a variety of food and health-related applications. However, one key component, acrylamide, is known to be highly toxic, causing noticeable negative effects on animals during studies, with a LD50 (50% lethal dose) of 150~180mg/kg on rodents. We found that polyacrylamide, as its derivative, on the other hand, is nontoxic.（figure 3)
Figure 3: The chemical structure of acrylamide(left) and polyacrylamide (right).
Two groups of composition of ingredients were taken into consideration:
1. Originally: 30% acrylamide(Acr) It is composed of 30% acrylamide, 0.015% N-methylenebiscrylamide(Bis), 2% alginate, 0.046% ammonium persulfate and 0.05%N,N,N9,N9- tetramethylethylenediamine(TEMED). This was the composition we came over during initial researching of literature.
2. And another with 25% acrylamide, 5% N-methylenebiscrylamide, 2% alginate, 0.046% ammonium persulfate and 0.05% TEMED. At this ratio, N-methylenebiscrylamide and acrylamide is more inclined to form polyacrylamide which reduces the content of acrylamide monomer, thus reducing toxicity.
During our own experiment, we incubated Hidro at 37℃ for 24 hours. The result was in favour of our assumption that the toxicity of 25% acrylamide was considerably smaller, with 770% increase in survival rate compared to those with 30% acrylamide (Figure 4).
Figure 4: The toxicity of Hidro with 25% acrylamide was considerably smaller than those with 30% acrylamide.
Apart from directly lowering the toxicity, hydrogel composition with 25% acrylamide, while exhibiting less toxicity, was also quicker in gel formation. This particular composition also increases the rate of polyacrylamide network formation through enlarging the relative ratio of the crosslinking agent: Bis. This noticeably decreased the contact time of microbes with acrylamide and further benefitted their survival. This video compares the gel-formation speed of two groups of hydrogel material. The gelling speed of the group with 25% acrylamide is much quicker than 30% acrylamide.
Video 1: The gel formation time of Hidro with 25% acrylamide and 30% acrylamide.
Further, this material in whole was also indistinguishable from Hidro with 25% acrylamide in terms of the key functions of Hidro (More details can be found below in 3.3 Biosafety section), and thus was preferred over the 30% one in following experiments.
The genetic design of Hidro
To achieve our goal of dynamically monitoring and controlling engineered microbes beyond the laboratory, we incorporated a barcode system along with a toxic protein system to the microbial genome through gene editing. The genetic circuit design is shown in the following diagram (Figure 5B).
Figure 5：(A) The code table for translating information into DNA barcode; (B) The design scheme of Trace and Control System in Hidro;(C) The gel image of the result of genome integration of DNA barcode and kill switch into fimA site.
Tracing
In order to achieve Trace and Control in the Hidro system, we have incorporated a DNA barcode into engineered strains, which stores user-defined information to distinguish engineered bacteria from natural bacteria. This Barcode can effectively help researchers track possible escaped bacterial individuals and distinct engineered strains or natural ones in an out-of-lab environment.
To design the desired barcode, we used online tools to generate a DNA Barcode sequence to store "NDNFCHINAIGEM2021“ in the DNA sequence “CTCTACCTCGCTTCACGTCTGCTCACTCTGGTCCTAACAGCGATAGCGTCT” (Figure 5A). (Note that the barcode used here is only an example. Users can customize the required barcode information according to their needs.)
Since the barcode needs to be subsequently detected by CRISPR-Cas12a based nucleic acids, we added the PAM sequence (TTTA) required for Cas12a at the 5' end of the barcode. So the final sequence was "TTTACTCTACCTCGCTTCACGTCTGCTCACTCTGGTCCTAACAGCGATAGCGTCT" (Figure 5B).
Control
After achieving the tracking of the bacteria, we also set up a kill switch system. This system can kill the target bacteria efficiently, thus eliminating any potential risk of escape.
In the design, the main parts of the kill switch are 1. a user-customizable inducible promoter, we chose arabinose promoter as an example here; and 2. a toxin gene - Kid (Figure 5B).
The kill switch has been successfully integrated into the E. coli Top10 genome using the CRISPR-Cas9 system, thus maintaining the stability of the suicide genetic circuit (Figure 5C). Please see the 3.5 Traceability section below for the characterization of this system.
The Properties of Hidro
After completing the above design, we tested whether the performance of the Hidro system could meet the requirements. The result exceeded our expectations, and its performance was mainly reflected in the following three points.
1. Biosafety: Within 72 hours, no escape occurred. Combining the material and genetic design, the overall escape frequency of the Hidro system is lower than 10^-10, which has met the NIH requirements.
2. Stability: We have tested Hidro's protection of internal bacteria against a variety of harsh environments. The results showed that higher bacterial survival rates were maintained in Hidro than the control group in extreme environments.
3. Traceability: Using Barcode & CRISPR-Cas12a and Kill switch system, we can track and effectively control escaped organisms when an escape event occurs (though the possibility is extremely low).
Next, please allow us to describe the three points above in detail.
Biosafety
bicontainment efficiency is higher than the NIH standard
Theoretically, pores formed by the outer hybrid network are only 5~50 nanometers in size, over 20 times smaller than E. coli (which usually range from 1~2 micrometers), and would be sufficient to prevent their escape.
We thus carried out a test to determine whether Hidro could effectively prevent the leakage of engineered strains. Specifically, we encapsulated roughly 10^7 bacteria (10^9 per ml) in each bead. Beads lacking a tough shell allowed bacteria to escape into the surrounding media and to grow to high densities after overnight incubation, whereas there was no physical escape of bacteria from Hidro (both 25% acrylamide and 30% acrylamide) even after 72h of incubation (Figure 6).
In summary, the Hidro system successfully achieved our aims of escape prevention. We encapsulated roughly 10^7 bacteria per bead, which means that the escape frequency is lower than 10^-7. Also, since there was no difference between Hidro with 25% acrylamide and 30% acrylamide in terms of biocontainment, and the 25% acrylamide composition showed less toxicity, Hidro with 25% acrylamide was used in all subsequent experiments.
Figure 6：Hidro has a highly efficient biocontainment ability even after 72hours. (A) Hidro with 25% acrylamide; (B) Hidro with 30% acrylamide.
Stability
Hidro maintains high bacterial survival rate in harsh environments
Hydrogels are desirable materials for encapsulating living cells as they provide an aqueous environment that can be infused with nutrients, allowing for cell growth and sensing, while also providing protection against environmental hazards.
To prove that Hidro can provide a stable environment for internal microorganisms in a variety of adverse environments, we put Hidro under diverse harsh conditions to test whether engineered strains can still survive. The results show that Hidro has the ability to maintain high bacterial survival rates, which will facilitate better application of the Hidro system in beyond lab environments.
Extreme low pH
Low pH can seriously damage the survival of microorganisms. Many industrial or natural environments have a low pH environment, such as industrial wastewater. If microorganisms cannot tolerate the low pH environment, then the expected biological functions such as wastewater contamination detection are difficult to achieve.
For testing acidic resistance, the Hidro in the control group were incubated at pH 7 for 1 hour, and the Hidro in the experimental group were incubated at pH 1.5 for 1 hour. The microorganisms in LB medium were treated in the same way as the control group.
After an hour, we homogenized the hydrogel core after the shell was removed and then performed a ten-fold gradient dilution along with the treated liquid culture on the LB agar plate.
Figure 7：Hidro can keep strains stable even in low pH environment.
As you can see from the picture shown above (Figure 7), the survival rate of microorganisms in Hidro was 90% at pH 1.5 compared to 0.03% in LB medium (where, without protection of Hidro, the microorganisms were directly exposed to the acidic environment). We concluded that, the Hidro system was able to provide a stable environment for internal microorganisms in acidic conditions.
Extreme high pH
Similar to low pH environments, extreme high pH environments can also seriously damage the survival of microorganisms.
For testing alkaline resistance, Hidro and bacteria in liquid medium were incubated in pH 7 and pH 10 respectively for 1 hour. We then performed a ten-fold gradient dilution after homogenizing the hydrogel core in the same way as the above.
Figure 8：Hidro can keep strains stable even in high pH environment.
As is shown in the picture shown above, it is noticeable that for the Hidro system, the survival rate of microorganisms was 53.3% at pH 10 compared to 0.21% in LB medium (Figure 8). It means the Hidro system could provide a stable environment for internal microorganisms in alkaline environments as well.
Extreme Heat
Temperature, especially extreme high temperatures, is another main obstacle outside the laboratory that affects the function of engineered modified strains. We likewise tested the performance of Hidro under extreme high temperature conditions. Hidro and LB medium containing bacteria were placed in 37°C and 90°C respectively for 2 hours. We subsequently performed a ten-fold gradient dilution on LB plate using the method previously mentioned.
It is noticeable that for the Hidro system, the survival rate of bacteria was 3.75% at 90°C compared to 0% in LB medium (Figure 9). So, the Hidro system is able to provide a supportive environment for internal microorganisms even under extreme heat conditions.
Figure 9：Hidro can support strains survive even in high temperature.
Traceability
Hidro can dynamically trace and control engineered bacteria
Tracing
Customizable Barcode and CRISPR-Cas12a offers an efficient tracing system
In the Tracing section above, we have described that a customized DNA barcode 「NDNFChinaiGEM2021」 has been integrated into the genome of E. coli. To enable efficient tracking of this Barcode, we utilized the highly efficient CRISPR-Cas12a nucleic acid detection system. The steps to achieve this detection have been shown in Figure A.
Lachnospiraceae bacterium ND2006 Cas12a (LbCas12a), which belongs to the class 2 type V-A CRISPR-Cas system, performed collateral cleavage on non-targeted ssDNAs upon the formation of the Cas12a/crRNA/target DNA ternary complex[3]. Similar to the SHERLOCK based on Cas13, Cas12a is of high sensitivity and specificity and is very convenient in the detection of target DNA barcodes on the bacteria genome. If a target DNA exists in the reaction system, the Cas12a/crRNA binary complex forms a ternary complex with the target DNA, which will then trans-cleave non-targeted ssDNA reporter in the system, illuminating the fluorescence which can be easily read by the plate reader or test strip.
It is reported that the minimum detectable concentration for Cas12a-crRNA was approximately 0.1 nM; However, When combined with PCR, the detectable concentration could be as low as 10 aM. Considering that tracking assays for the Hidro system need to be done outside the lab, we need a DNA amplification method that can be used on site without the need for complex equipment. Recombinase polymerase amplification (RPA) is a good choice.
RPA is a single tube, isothermal alternative to PCR. Because it is isothermal, RPA can use much simpler equipment than PCR. Operating at room temperature means RPA reactions can in theory be run quickly simply by holding a tube by hand. This makes RPA an excellent candidate for developing low-cost, rapid, point-of-care molecular tests.
Figure 10：(A) Steps in CRISPR-Cas12a-based nucleic acid detection system to track escaped strains from Hidro system; (B) A band is shown on a test strip when CRISPR-Cas12a-crRNA complex detects the barcode on the E. coli genome; (C) The output of CRISPR-Cas12a can also be read by fluorescence using plate reader.
Once the Hidro system is performing a task or finishing a task, we can use the above system to perform the detection assay to know whether there are any escape events happening. As shown in Figure 10, CRISPR-Cas12a could efficiently detect the samples mimicking strain escape from Hidro. Positive results include a color band shown on the test strips (Figure 10B) as well as an increase in fluorescence signal (Figure 10C).
Theoretically, the detectable concentration could be as low as 10 aM[4], which greatly increases the sensitivity to detect very low-possibility escaped individuals throughout the Hidro system and allows us to more easily and dynamically monitor the development of escape events.
Control
Kill Switch to eliminate any potential risk of escape
After tracking the escaped bacteria, we can activate the kill switch in the engineered strain, thus eliminating any potential risk of escape.
The kill switch contains a user-customizable inducible promoter and a toxin gene. In this project, we chose L-arabinose-induced promoter and Kid toxin as an example. Kid is a toxin protein of post-segregational killing system and can work as an endonuclease enzyme to cleave mRNA. The kill switch will eventually be integrated into the E. coli Top10 genome using the CRISPR-Cas9 system, thus maintaining the stability of the suicide genetic circuit.
Figure 11：(A) A user-customizable kill switch could eliminate any potential risk of escape; (B) The viability of E. coli with the Kid kill switch with or without L-arabinose inducer.
Kid kill switch has an escape frequency lower 10^-3, according to the data shown in Figure 11B. Combined with the previous Hidro system's intrinsic biocontainment efficiency of more than 10^7, the whole Hidro system's escape frequency is lower than 10^-10, which has fully met the NIH requirements.
Application
Since Hidro system only offers protection and support for engineered bacteria inside, it has an astonishing wide range of applicable environments, including water, soil, food and human body. Below are several genetic systems we designed for our project as examples of how our system works in diverse scenarios.
Basic induction function
The nanoporous structures of the hydrogel shell and alginate core should allow rapid diffusion of small molecules and ions. To determine whether encapsulated bacteria in Hidro can respond to these external signal inputs, we encapsulated Hidro with the bacteria containing a genetic circuit that expresses chromoprotein (eforRed) in response to L-arabinose induction (Figure 12A).
Figure 12: (A) The schematic of a L-arabinose–controlled genetic switch to produce chromoprotein eforRed; (B) eforRed protein production in Hidro.
We then incubated the beads at 37 °C in the absence of L-arabinose or in the presence of 10mM L-arabinose. We found that encapsulated cells exposed to L-arabinose exhibited a pink colour compared with encapsulated cells not exposed to L-arabinose (Figure 12B). Thus, gene expression in Hidro can be exogenously controlled by chemical inducers.
Heavy metal detection
Heavy metal pollution in water bodies is a human-caused environmental issue that many governments and institutes tried to deal with. For example, mercury ion like Hg2+ is a well-known and widespread environmental contaminant that can adversely affect human health. To demonstrate that encapsulated bacteria can function in a real-world setting, we used an Hidro to detect the presence of metal ions in water samples.
Figure 13：(A) The schematic of a Hg2+ sensing circuits in Hidro; (B) Hg2+ sensing circuits output.
We characterized the fluorescence output induced by Hg2+ in Hidro harboring the genetic design shown in the Figure 13A. In the Hg2+ sensing circuits, fluorescent protein GFP can only be expressed when Hg2+ combines with the sensory protein, merR.
Hidro beads with bacteria harboring the Hg2+ sensing circuits were incubated in liquid added Hg2+. As you can see from Figure b, exposure to 10μM Hg2+ resulted in the expression of GFP in E. coli chacterized by plate reader after breaking beads and homogenization, indicating successful detection of Hg2+ ions. Thus, this result highlighted the potential of Hidro to detect toxic levels of heavy metals in natural environment water bodies.
(All experiments involving heavy metals follow the appropriate laboratory safety measures, please see more information in Safety)
Food quality detection
Food is closely related to our health, with spoilage of fish causing particularly serious food health problems and significant economic losses and chronic diseases each year. Timely detection of spoilage can circumvent these problems in advance. Detection of Quorum-sensing (QS) molecules secreted by spoilage-associated bacteria like Pseudomonas aeruginosa in food is a feasible approach.
Collaborating with team GCGS_China, we had designed a food quality detection circuits that reponds to the C4-HSL, a QS molecule secreted from Pseudomonas aeruginosa. In the circuits shown in Figure 14A, when C4-HSL combines with the sensory protein, lasR, green fluorescent protein can be expressed in the sensing strain encapsulated in Hidro.
Figure 14：(A) The schematic of a C4-HSL sensing circuits in Hidro; (B) The image of Hidro taken by a fluorescence microscope under blue light.
As you can see from Figure 14B, exposure to 1 μM C4-HSL resulted in the emergence of a cell population expressing high levels of GFP in Hidro, indicating successful detection of C4-HSL. These results were confirmed visually under fluorescence microscope emitting blue light. The results demonstrated the feasibility of Hidro system in food detection.
Caffeine
controlled genetic switch for drug secreting
After successfully verifying the function of Hidro on sensing signals, we hope to use Hidro to secrete some important signaling molecules, thus realizing the complete function of Hidro as a platform for engineering biological design.
Health and medical care are important application directions of synthetic biology. Hidro's safe, stable, traceable specificity is particularly well suited in health-related application scenarios, as we wrote in the Proposed Implementation, where we can use Hidro to secrete specific drug-related molecules in the intestinal environment or as "tea bags" to secrete specific health molecules in food, but at the same time will NOT affect the original composition of the intestinal flora or the quality of the food.
Design a caffeine–controlled genetic switch
To realize the application of Hidro in health, the first thing to consider is how to effectively regulate the way Hidro functions in the human environment like oral or gut.
Recent advances in synthetic biology have required the design of application-specific control systems that are functionalized to perform user-defined, precisely controlled regulation process[5]. Initially, some common inducers like IPTG, tetracycline were used for the control of gene expression, but these wildly used inducers raised issues such as antibiotic resistance and side effects, especially in longterm applications. Traceless inducers, such as light or temperature, have recently been developed, but ambient light and ambient temperature make them less orthogonal than would be desirable.
The ideal inducer would be inexpensive, would have no side effects, and would be present in only a specific set of known sources. It has been proposed that ideal trigger molecules would be natural, nontoxic, highly soluble, inexpensive, and perhaps even origin from daily life[6].
Caffeine is a strong candidate. The caffeine is non-toxic, cheap to produce, and present in specific beverages, such as coffee and tea. Every day, more than two billion cups of coffee are being consumed worldwide, making coffee one of the most popular beverages after water.
Here, we have developed a sensitive engineered genetic system in response to dietary intake of coffee or other caffeine-containing beverages and characterised them in Hidro. This beverage-derived caffeine–controlled gene circuit expands the synthetic biology toolbox available for constructing safe and clinically relevant cell functions and has the potential to substantially advance Hidro application in health like bacterial therapies.
Figure 15：(A) The design scheme of a caffeine–controlled genetic switch; (B) The response curve of caffeine–controlled genetic switch. Samples prepared in triplicate, data represent the mean ± 1 s.d.
We used two of these domains: the single-domain VHH camelid antibody; (referred to as aCaffVHH) that homodimerizes in the presence of caffeine. In two separated acVHH domains, each was fused with a 10-residue linker, into the contiguous M86 intein. The intein was already inserted between ECF20(1–101) and ECF20(102–193)[7]. The resulting constructs were bipartite proteins, with each part driven by a constitutively-expressed promoter J23110. So in the presence of caffeine, they could homodimerize and reconstitute a complete ECF20 with the promotion of M86 intein and activate downstream promoter, pECF20.
In order to test the performance of caffeine–controlled genetic switch, we first chose RFP to be our reporter gene. The engineered strain containing this switch is first incubated with different concentrations of caffeine molecules in a liquid medium. Plate Reader test results show that fluorescence intensity increased with the increase of caffeine concentration, which reveals the successful design of the caffeine sensing circuit (Figure 15B).
After encapsulating engineered bacteria into Hidro beads and incubating with and without caffeine molecules respectively (Figure 16A), we sliced the beads at thickness of roughly 0.5 mm. The sliced samples are then imaged with a fluorescence microscope under green light. It was confirmed visually that Hidro exposed to caffeine exhibited strong red fluorescence (Figure 16B).
Figure 16：(A) The schematic of caffeine–controlled genetic switch encapsulated in the Hidro system; (B) The image of Hidro taken by a fluorescence microscope under green light.
Apply caffeine–controlled genetic switch to secrete L-dopa
The Hidro system has potential for safe drug delivery in food and the human gut (See more information in NDNF Proposed Implementation). So we then designed a caffeine–controlled L-Dopa production system based on the switch designed here (Figure 17A).
Derived from the biosynthesis of l-tyrosine, L-Dopa is a naturally occurring amino acid which acts as a precursor to a number of neurochemicals such as adrenaline and dopamine (Figure 17B). L-Dopa is an amino acid which is made and used as part of normal biological functions in humans.
We replace the output gene with L-Dopa metabolic pathway HpaB and HpaC, each with a specifically designed RBS. After 24 hours of incubation in tube, we removed the Hidro beads, centrifuged the liquid and took out the supernatant for measurement. Then we applied measurement protocol on levodopa concentration by the method we have mentioned before. From the result shown in Figure c, we can see that Hidro group showed L-dopa production compared to those in group group. It means that Hidro can realize the sensing and secreting process (Figure 17C). It further proves that Hidro has a great potential to be applied in Health.
Figure 17：(A) The schematic of caffeine–controlled genetic switch encapsulated in the Hidro system to produce L-dopa; (B) The metabolic pathway from L-tyrosine to L-Dopa. (C) L- Dopa concentration measurement in the caffeine–controlled L-Dopa production Hidro system.
References
1. Tang, T.-C. et al. Hydrogel-based biocontainment of bacteria for continuous sensing and computation. Nat Chem Biol 1–8 (2021) doi:10.1038/s41589-021-00779-6.
2. Bodzoń-Kułakowska, A. et al. Proteomic Profiling and Analytical Chemistry. 25–77 (2013) doi:10.1016/b978-0-444-59378-8.00003-7.
3. Li, S.-Y. et al. CRISPR-Cas12a-assisted nucleic acid detection. Cell Discov 4, 20 (2018).
4. Gootenberg, J. S. et al. Multiplexed and portable nucleic acid detection platform with Cas13, Cas12a, and Csm6. Science 360, eaaq0179 (2018).
5. Bojar, D., Scheller, L., Hamri, G. C.-E., Xie, M. & Fussenegger, M. Caffeine-inducible gene switches controlling experimental diabetes. Nat Commun 9, 2318 (2018).
6. Yin, J. et al. A green tea–triggered genetic control system for treating diabetes in mice and monkeys. Sci Transl Med 11, eaav8826 (2019).
7. Ho, T. Y. H. et al. A systematic approach to inserting split inteins for Boolean logic gate engineering and basal activity reduction. Nat Commun 12, 2200 (2021).
