ENGINEERING
Design
In the course of clinical treatment, the EEG signals of epilepsy patients during seizures often contain characteristic spike waves, slow waves, and spike-and-slow wave complexes. These features can be utilized to evaluate the patients' health conditions and seizure risks. However, in practical clinical settings, the judgment usually relies on the extensive experience of clinicians. This not only imposes a significant burden on doctors but also introduces uncertainties for patients. To provide a reliable adjunct for clinical treatment, we are dedicated to developing an EEG-based epilepsy monitoring and prediction model. The epilepsy dataset we utilized was sourced from OpenNEURO(https://openneuro.org/datasets/ds003555/versions/1.0.1), which contains electroencephalogram (EEG) recordings of high-frequency oscillations (HFO) from 30 pediatric epilepsy patients. The authors also shared code on GitHub detailing data processing workflows including signal filtering, baseline calibration, and event detection. By analyzing EEG data from these patients and applying the provided GitHub code for feature computation, we generated labeled time segments of epileptic seizures. However, the labels contained false positives mixed with true positives. To further optimize true positive prediction, we implemented additional features beyond those specified by the authors, including L1 norm calculation, Shannon entropy analysis, sample entropy metrics, and permutation entropy evaluation. Through automated identification of these anomalous feature values, we conducted predictive classification tasks to distinguish between true and false positives using deep learning models.
Build
In this section, we extracted time-domain and frequency-domain features from columns 6 to 14 of the TSV file for machine learning integration. Using column 15's true positive/false positive predictions as the reference target, we developed three models: a random forest model, a CNN model, and an LSTM-CNN hybrid model. After evaluating performance metrics including accuracy rates across these models, we ultimately selected the optimized CNN model as the final implementation.
Test
During the model's testing phase, we use the accuracy result of model training as the measurement standard. We divide the existing dataset into a training set and a validation set. We input the positive and negative sets of the training set into the model for learning. After completion, we input the validation set to check the training results. The epilepsy detection model was evaluated on the independent test set with an accuracy of 80.85% and a sensitivity of 82.3%. The prediction confidence of a single prediction was sometimes low, indicating the uncertainty of decision.
Learn
Analysis shows that data imbalance and the loss of time-domain information due to frequency-domain features lead to performance limitations. More data needs to be collected, time-domain features should be integrated, and the model architecture should be optimized, such as adding LSTM to enhance generalization ability. By combining machine learning methods, the accuracy of distinguishing the resting-state EEG of normal patients from that of epilepsy patients can reach over 90%, but the model for calculating the true positive/false positive labels during epileptic seizures needs to be further optimized based on specific criteria. In the next cycle of the DBTL engineering loop, we plan to divide the dataset more reasonably to enhance the model's sensitivity, accuracy and precision.
Design
Our design goal is to construct an efficient, safe, stable and highly compatible BHB synthesis pathway in engineered E.coli Nissle 1917 ( EcN ). We aim to design a self-sustaining metabolic cycle. The innovation lies in the use of the two-way reaction characteristics of PCT : the CoA group on BHB-CoA is transferred to acetic acid to form the final product BHB, and acetic acid is converted to acetyl-CoA, which is the starting substrate of the BHB synthesis pathway. This design realizes the regeneration of acetyl-CoA, bringing multiple theoretical advantages including reducing metabolic load, improving pathway sustainability, and solving the accumulation toxicity of potential metabolites.
The intestine is rich in short-chain fatty acids ( SCFAs ), of which acetic acid is one of the most important components. The design of the PCT pathway cleverly transforms this environmental feature into an advantage. Our engineered bacteria can directly use intestinal endogenous acetic acid as a ' raw material ' to drive the regeneration of acetyl-CoA, reducing the competition for host direct energy substances ( such as glucose ) and reducing the risk of potential interference with host metabolism. This pathway enables engineered bacteria to be more harmoniously integrated into the intestinal micro-ecosystem, and its metabolic activity complements the intestinal chemical environment, improving the stability of colonization and the persistence of therapeutic effects. Intestinal short-chain fatty acids ( SCFAs ), of which acetic acid is one of the most important components. The design of the PCT pathway cleverly transforms this environmental feature into an advantage. Our engineered bacteria can directly use intestinal endogenous acetic acid as a ' raw material ' to drive the regeneration of acetyl-CoA, reducing the competition for host direct energy substances ( such as glucose ) and reducing the risk of potential interference with host metabolism. This pathway enables engineered bacteria to be more harmoniously integrated into the intestinal micro-ecosystem, and its metabolic activity complements the intestinal chemical environment, improving the stability of colonization and the persistence of therapeutic effects.
Before determining the final scheme, we systematically evaluated other major BHB synthesis pathways reported in the literature and weighed them based on our design goals :
PhaZ pathway ( polymerization-depolymerization pathway ) :
Mechanism : Polyhydroxybutyrate ( P ( 3HB ) ) particles were synthesized by PhaA, PhaB, and PhaC, and then BHB was produced by PhaZ depolymerization. Disadvantages : This approach is too lengthy and needs to coordinate the two processes of aggregation and disaggregation that may conflict in time and space ; due to its long cycle of more than 100 hours and low productivity ( ~ 0.07 g / L / h ), it is completely unable to meet the requirements of rapid response and dynamic regulation required for in vivo treatment.
TesB pathway ( thioesterase pathway ) :
Mechanism : BHB-CoA was synthesized by PhaA and PhaB, and then hydrolyzed by thioesterase ( TesB ) to produce BHB.
Disadvantages : Although simpler than the PhaZ pathway, the TesB-catalyzed hydrolysis reaction is a one-way, irreversible consumption process. The production of each BHB molecule permanently consumes one acetyl-CoA and one CoA, which will impose an increasing burden on the central metabolism and lack sustainability, which may affect the long-term survival and function of the engineered bacteria.
Ptb-Buk pathway ( phosphotransferase pathway ) :
Mechanism : BHB-CoA is converted to BHB by phosphotransbutyrase ( Ptb ) and butyrate kinase ( Buk ).
Disadvantages : Similar to the TesB pathway, it is also a CoA consumption pathway, and requires ATP consumption, and does not have an advantage in energy economics.
After comparison, the PCT approach has become the best choice to meet all our core design standards due to its outstanding theoretical advantages of regeneration cycle, environmental compatibility, and high efficiency and sustainability.
Build
We assigned specific functions and optimization considerations to each core enzyme element : PhaA ( β-ketothiolase, derived from Ralstonia eutropha ) catalyzes the Claisen condensation reaction of two molecules of acetyl-CoA to produce acetyl-CoA. This is the rate-limiting initial step of the entire pathway. In order to ensure high catalytic efficiency, we selected the PhaA isoenzyme, which has been characterized in the bacteria and has high affinity for acetyl-CoA. At the same time, codon optimization was performed on its coding sequence to adapt to the translation system of E.coli and avoid low expression efficiency and translation errors caused by rare codons.
PhaB ( acetylacetyl-CoA reductase, derived from Ralstonia eutropha ) uses NADPH as a coenzyme to stereoselectively reduce acetylacetyl-CoA to ( R ) -BHB-CoA. This is a key step in determining the chiral purity of the product. The PhaB variant with high ( R ) -stereoselectivity was strictly selected to ensure the optical purity of the product more than 99 % from the source. At the same time, we focused on its coenzyme circulation efficiency with NADPH / NADP + to ensure that it can be coordinated with the reducing power supply of host cells.
Test
Before starting the wet experiment, we conducted a systematic comparative analysis of the public performance data of multiple channels, which constitutes a crucial ' test ' link in our decision-making process. We have established an assessment framework that focuses on titer , productivity and pathway sustainability.
Existing studies have shown that although the PhaZ pathway can obtain a higher final yield ( such as 7.3 g / L ), it requires a two-step fermentation process of more than 100 hours, resulting in a generally low volumetric productivity ( about 0.07 g / L / h ). For in vivo treatment that requires rapid response and dynamic regulation, a lengthy production cycle is unacceptable. Therefore, despite its high absolute yield, low productivity makes it unsuitable for our application scenarios. Liu et al. ( 2007 ) reported that the TesB pathway can achieve 4.0 g / L production within 24-48 hours, and the productivity is 0.17 g / L / h. The TesB pathway is more direct, but its hydrolysis consumes coenzyme A ( CoA ) and cannot regenerate acetyl-CoA. This means that the synthesis of each BHB molecule is at the expense of permanent consumption of central metabolites, which may limit the sustainability of long-term production.
The direct study of Matsumoto et al. ( 2013 ) under the same strain background and comparable conditions provided us with key evidence. The data showed that the yield of the PCT pathway was as high as 5.2 g / L and the productivity reached 0.22 g / L / h after acetic acid supplementation, which was significantly higher than that of the PhaZ pathway ( 0.07 g / L / h ) and the TesB pathway ( 0.17 g / L / h ), which proved its advantage in speed.
In terms of mechanism verification, the literature confirmed by 13C isotope labeling experiments that exogenous acetic acid was directly integrated into BHB molecules, providing solid experimental evidence for the ' acetyl-CoA regeneration ' mechanism, which is far better than pure theoretical speculation. In terms of product quality, the purity of ( R ) -BHB enantiomers reported in the literature was as high as 99.2 %, which met the treatment requirements. In terms of environmental adaptability, the yield increases with the increase of acetic acid concentration, which is perfectly consistent with the acetic acid-rich human intestinal environment, indicating that our engineered bacteria can effectively utilize endogenous acetic acid in the body to achieve more stable and sustainable production.
Learn
Through systematic comparison and analysis of these key literature data, we have obtained clear and reliable conclusions :
The PCT pathway performs best in key performance indicators ( productivity ) and can meet the needs of in vivo treatment for rapid onset.
Its unique ' acetyl-CoA regeneration ' mechanism has been strongly supported by experimental data, laying a metabolic foundation for long-term and sustainable production.
This pathway has a high degree of internal compatibility with the target colonization environment ( gut ), which is a ' born for the gut ' pathway.
Therefore, this in-depth test based on literature data convinces us that choosing the PhaA-PhaB-PCT pathway is a data-driven and rational optimal decision. It not only surpasses the alternative in performance, but also its inherent characteristics are highly consistent with our final application goals, laying a solid foundation for subsequent gene circuit design and security control. This method minimizes the risk of R & D and embodies the core spirit of rational engineering design of synthetic biology.
Design
Centering on the core objective of achieving BHB synthesis in Escherichia coli Nissle 1917, based on the previously screened BHB synthesis pathway, it was determined that four key target gene fragments, namely phaA, phaB, pcT, and rplO, need to be introduced. After collaborative discussion with the PI on the plasmid construction strategy, the initial design was to clone phaA, phaB, and pcT genes respectively using the pET-21a vector (constructing three recombinant plasmids), while cloning the rplO gene using the pRSFDuet-1 vector (constructing one recombinant plasmid). Ultimately, all genes would be introduced into the host bacteria through "co-transformation of four plasmids". The core consideration of this design is the "single gene - single vector" modular construction, which has the advantages of intuitive operation and clear steps, and initially meets the basic requirements for pathway assembly.
Build
Strictly follow the design plan to carry out the experiment:
①Firstly, specifically amplify the four target gene fragments of phaA, phaB, pcT, and rplO through PCR technology to ensure the length and sequence accuracy of the target fragments;
②Use seamless cloning technology to connect each gene to the corresponding vector (phaA/phaB/pcT to pET-21a, rplO to pRSFDuet-1), completing the construction of four recombinant plasmids;
③To achieve efficient amplification of the plasmids, first transfer the recombinant plasmids into the competent cells of Escherichia coli DH5α (DH5α is a plasmid replication-optimized strain that can quickly obtain high-copy plasmids), and plan to extract the plasmids to obtain high-concentration products, and then transfer them to the final host Escherichia coli Nissle 1917 to complete the initial construction of the engineered bacteria.
Test
Centering on "verifying the feasibility of the design plan", two major test indicators were set:
①Whether the concentration of the plasmid after amplification meets the subsequent transformation requirements;
②The host adaptability of the co-transformation of four plasmids.
During the experimental verification, key problems were discovered:
①The plasmid extraction experiment was repeated four times, and the concentration of the recombinant plasmid obtained each time was extremely low, far from reaching the concentration threshold required for the transformation of engineered bacteria, directly leading to the inability to proceed with the subsequent transformation experiments;
②Combined with the discussion of the PI team and further analysis: when four plasmids are simultaneously introduced into a single host bacterium, it will significantly increase the burden of exogenous gene expression on the bacteria (including promoter competition, consumption of plasmid replication resources, etc.), and both theoretically and practically, it has been confirmed that there are inherent defects in the success rate of this plan.
In conclusion, the test results clearly show that the initial design plan did not meet the expectations in both "plasmid yield" and "host compatibility", and targeted optimization is needed.
Learn
In response to the issue of "low concentration and high host stress due to excessive plasmid quantity" identified during the Test phase, a scheme optimization was carried out:
①Iteration of vector selection: Switch to dual-promoter vectors pET-Duet-1 and pRSFDuet-1 (both support independent expression of two genes on the same vector), and clone phaB and pcT in tandem to pET-Duet-1, and phaA and rplO in tandem to pRSFDuet-1, constructing only two recombinant plasmids;
②Optimization logic: Through the integration of "multiple genes - single vector", not only the number of plasmid constructions is reduced (from 4 to 2), simplifying the operation complexity and reducing the risk of extraction failure, but also the exogenous expression pressure on the host bacteria is significantly alleviated, fundamentally solving the compatibility issue of co-transformation.
Ultimately, through experimental verification, the optimized dual-plasmid co-transformation scheme successfully achieved the construction of engineered bacteria, and the BHB synthesis pathway could be normally initiated, confirming the feasibility of this iterative design and completing the closed-loop optimization of the DBTL cycle.
Design
We envision using hydrogels to encapsulate engineered bacteria, which can physically adhere directly to the intestinal tract without the need to introduce additional elements into the bacteria for expressing adhesive substances. This design can physically isolate the engineered bacteria from the intestinal flora to avoid horizontal exchange of genetic material and ensure the stable presence of the bacteria in the digestive tract.
Through literature review, we have designed a hydrogel mainly composed of sodium alginate (SA) and polydopamine (PDA), following a three-layer structure concept. It consists of an outer pH-responsive layer, a middle adhesive layer, and an inner nutrient layer. The outermost layer is a hydrogel made of alginate, which features responsiveness to the weakly alkaline pH of the intestinal tract and easily swells and ruptures in the intestine to expose the middle layer. The middle layer is a coating composed of polydopamine, sodium alginate, and polyacrylamide, which has adhesiveness to adhere to the intestinal tract for colonization. Additionally, the pore size of this middle layer does not allow macromolecules such as DNA to pass through, thereby preventing gene leakage. The innermost layer is a mixture of alginate and E. coli culture medium, which provides nutrients for the growth of E. coli.
Biuld
1.Preparation of Hydrogel Alginate Core
Mix fresh bacterial culture (10⁸–10⁹ cells/mL in LB medium) with a 5 w.t.% sodium alginate (SA) solution at a 1:1 volume ratio to achieve a final SA concentration of 2.5 w.t.%. Load this bacteria-alginate premix into a syringe, then drop it into a pre-sterilized 5 w.t.% CaCl₂ solution. Allow solidification for 15 minutes to form bead-like droplets with a controllable diameter of approximately 3 mm.
2.Preparation of Hydrogel Adhesive Middle Layer
Add a certain amount of dopamine hydrochloride (DA) to a weakly alkaline 2 wt.% sodium alginate (SA) solution, and stir at room temperature for 24 hours to obtain a PDA-SA solution.
Under an ice-water bath (approximately 4℃), sequentially add a certain amount of acrylamide (AM) and cross-linking agent to the above solution. After thorough stirring, add initiator and catalyst to the mixed solution. Mix well, transfer the mixture into a glass petri dish, and immerse the alginate cores into the pre-gel solution to form a middle-layer coating. Immerse the hydrogels in MES buffer (0.1 M MES, 0.5 M NaCl, pH 6.0) containing cross-linking agent and catalyst to form covalent bonds between the alginate and polyacrylamide networks, and maintain this state for 3 hours.
3.Preparation of Hydrogel pH-Responsive Outer Layer
Place the hydrogel beads with the middle layer prepared into a -20 °C refrigerator for freezing for 5 hours. Then immerse them into a solution containing 3% calcium ions and 1% SA. The calcium ions on the surface of the beads react with SA to form a calcium alginate gel film. Store the final hydrogels in a humid environment.
Test
Conduct experiments for safety evaluation, outer-layer pH responsiveness verification, and middle-layer adhesion performance verification
By determining the growth curve of E. coli in the hydrogel beads, we verified that the designed hydrogel does not affect the normal growth of engineered bacteria, indicating that the material is relatively safe for bacteria. Then, we used passaged colon cancer cells to verify the adhesion performance of the hydrogels coated with the middle layer. Finally, by measuring the swelling ratio of the prepared hydrogel beads in PBS buffers with different pH values, we verified their responsiveness to the weakly alkaline pH of the intestinal tract.
1.Safety Evaluation – Determination of E. coli Growth Curve
Place hydrogel samples coated with the middle layer and samples with only the inner core into centrifuge tubes containing 1 mL of liquid LB medium, respectively. Homogenize using 3.2 mm stainless steel beads at a frequency of 60 Hz for 20 minutes on a tissue grinder. To quantify cell density, measure the homogenized samples with a spectrophotometer and adjust their concentrations appropriately. Using this as a reference, adjust the absorbance values of each experimental group, record the ratios, and add each sample to a 96-well plate in proportion. Measure the absorbance values of each sample continuously for 24 hours using a microplate reader, and plot the growth curves. The groups are divided as follows:
- Group A: LB medium
- Group B: LB medium + 5% SA (1:1 mixture)
- Group C: LB medium + 5% SA + CaCl₂ (cross-linking agent)
- Group D: LB medium + 5% SA + PDA solution (system: 5 mL LB + 4.5 mL SA + 0.5 mL PDA)
- Group E: LB medium + 5% SA + PDA solution + initiator (system: 5 mL LB + 4.5 mL SA + 0.5 mL PDA + 10 μL initiator)
- Group F: LB medium + E. coli (EcN)
- Group G: LB medium + 5% SA + E. coli (EcN)
- Group H: LB medium + 5% SA + CaCl₂ (cross-linking agent) + E. coli (EcN)
- Group I: LB medium + 5% SA + PDA + E. coli (EcN)
- Group J: LB medium + 5% SA + PDA solution + initiator + E. coli (EcN)
- Group K: LB medium + inner core
- Group L: LB medium + hydrogel beads after coating with middle layer
Add 200 μL of each above system to each well of the 96-well plate, with 5 wells per group. After excluding the maximum and minimum values, calculate the arithmetic mean. Plot the growth curve with time as the abscissa and OD₆₀₀ as the ordinate.
The result validates our hypothesis that the inner-layer hydrogel core can effectively provide nutrients to support the growth of E. coli!
However, when the middle layer was used as a control, issues arose with the experimental data. We have analyzed and discussed these issues, and further details can be found in the "Learn" section.
2. Verification of the Adhesion Performance of the Middle Layer
Take out 4 bottles of passaged Caco-2 cells from the incubator and divide them equally into two groups. Inside a biosafety cabinet, aspirate and discard the medium. Place 100 hydrogel cores and 100 hydrogel beads coated with the middle layer into the culture bottles respectively, ensuring the beads are evenly dispersed. Invert the culture bottles containing the beads on a shaker, and take photos to record the position of each bead in the culture bottle. Shake the shaker at a certain rotational speed, take photos every 5 minutes to observe and record the changes in the position of the beads in the culture bottle, and continue this process for 50 minutes. Repeat the above experiment with the shaker rotating at different speeds, and record the changes in the position of the beads under different rotational speeds.
The results showed that over time, no hydrogel beads detached in the group coated with the middle layer material, whereas the uncoated group exhibited varying degrees of detachment at each time point. This indicates that the middle PSP hydrogel layer possesses strong adhesive properties.
3. Verification of the pH Responsiveness of the Outer Layer
Place the prepared hydrogels at room temperature for 3 hours, then weigh them to obtain the initial weight (W₀). Immerse the samples separately into beakers containing phosphate-buffered saline (PBS) with pH values of 1.5, 7.4, and 11.0. Every 10 minutes, take out the samples, quickly blot the moisture on the sample surface with filter paper, and weigh them, recording the weight as Wₛ. This measurement process is conducted in a constant temperature and humidity chamber, with the temperature set at 37°C and the humidity at 60%. The swelling ratio is calculated using the formula: Swelling Ratio = (Wₛ - W₀) / W₀ × 100%
The data shows in the near-neutral intestinal environment, the outer network becomes loose with increased permeability, allowing the diffusion and release of the encapsulated substances.
Learn
1. How to enhance the adhesiveness of hydrogels has always been a challenging problem for us. After reviewing numerous literatures, we finally modified the middle-layer hydrogel material by adding dopamine hydrochloride, which was grafted onto sodium alginate through Michael addition reaction and Schiff base reaction, greatly enhancing the adhesiveness of the material.
2. When coating the middle layer of the hydrogel, in the initial experiments, the prepared hydrogel was too brittle and fragile to meet our expected results. After literature research and reflection, we found that it was because the hydrogel core was placed in the PDA-SA solution for too long, leading to excessive cross-linking and reduced mechanical properties. In the second preparation, we reduced the time of placing the hydrogel core in the PDA-SA solution, and obtained hydrogels with good mechanical properties.
3. When passaging Caco-2 cells, cell clumping is prone to occur. The reason may be that after adding serum-containing complete medium to terminate digestion, sufficient pipetting is not performed immediately, and the remaining trypsin will continue to act, causing cells to aggregate more easily due to damage. We found that pipetting the cell clumps repeatedly with appropriate force, especially along the bottom and side walls of the culture flask, can alleviate this problem and enable the cells to grow better.
4. Taking the hydrogel encapsulated with the middle layer as the research object, we analyzed the data and found anomalies. We hypothesized that the self-aggregation of PDA (polydopamine) caused its OD600 value to increase over time. Therefore, we took PDA as the research object and further determined the growth curve. Using the difference method, we obtained the OD600 values contributed by Escherichia coli (E. coli) in different media.
The results showed that E. coli could grow stably at a certain growth rate both in the middle-layer material and the hydrogel beads encapsulated with the middle layer. This indicates that neither the middle-layer material nor the encapsulation method has a significant impact on the growth of E. coli. The detailed iteration process is shown in the "Notebook" section.
This process reflects our cycle from one "Learn" phase to the next "Design" phase, and from one cycle to the subsequent cycle, embodying our "Engineering Success".
Design
We designed a simple and reliable in vitro experiment to quantitatively evaluate the gene leakage level of our hydrogel encapsulation system in a simulated intestinal basal environment.
1.Environmental simulation : We chose to prepare simulated intestinal fluid ( SIF ), and its key parameters included pH 6.8 and two key components : bile salt ( a surfactant that can destroy membrane structure ) and trypsin ( a protease mixture ). This creates a basic chemical stress environment.
2.Static conditions : The experiment was carried out under anaerobic, 37 ° C static conditions to focus on evaluating the intrinsic stability of the material / strain and excluding complex variables such as fluid dynamics.
3.Key separation steps : Use a 0.22 μm filter to filter the soaking solution to ensure that we detect ' leaky ' genetic material ( free DNA, subcellular particles ), not complete bacteria.
4.Sensitive detection : qPCR absolute quantification of specific genes ( such as resistance genes ) on our engineering plasmids was used to provide highly sensitive and quantifiable leakage data.
Build
Materials preparation : Synthesis and purification of our engineered strains. Preparation of hydrogel microspheres encapsulating engineering bacteria. At the same time, a blank hydrogel without bacteria was prepared as a control.
Reagent preparation : The simulated intestinal fluid ( pH 6.8 ) containing 0.5 % bile salt and 1.0 % trypsin was prepared and sterilized according to the Chinese Pharmacopoeia. Preparation of qPCR reagents : specific primers / probes for the target gene, purified plasmid DNA for the standard curve.
Establish a culture system : Hydrogel microspheres were incubated with simulated intestinal fluid in a sterile centrifuge tube. The culture tube was placed in an anaerobic tank filled with anaerobic mixture to simulate the intestinal hypoxia environment.
Test
Experiment execution and data acquisition :
1.Exposure and sampling : Parallel samples were taken from the anaerobic tank at the set time points ( days 1,3, and 7 ).
2.Sample treatment : After violent vortex, the soaking solution was taken and passed through a 0.22 μm filter. The collected filtrate represents the ' leakable component '.
3.DNA extraction and qPCR : Total DNA was extracted from the filtrate, and the qPCR program was run to generate a standard curve using a plasmid standard with a known concentration.
4.Data analysis : The Ct value of qPCR was substituted into the standard curve to calculate the absolute copy number / mL of the target gene in each sample. The data of the experimental group were statistically compared with the negative control group.
Learn
The above experimental results will determine our next optimization and improvement of the hydrogel.
If the level of gene leakage is extremely low : This indicates that our hydrogel system has excellent closure under basic conditions. Based on this positive result, we will advance to more complex tests ( such as horizontal gene transfer assessment ) to challenge the safety of the system in more extreme environments.
If a significant leak is detected : We must analyze the leak pattern to determine the root cause. In the case of early sudden leakage, it indicates that the physical structure of the hydrogel is defective, and it is necessary to return to the Design stage to optimize the material formulation. If a leak that continues to rise over time is detected, it points to the cracking of the internal engineering bacteria. It is necessary to return to the Build stage to transform the strain and enhance its stability.
Regardless of the results, this experiment provides us with a clear direction for improvement, ensuring that each iteration of the project is based on experimental evidence.
[1] Tang T C, Tham E, Liu X, et al. Hydrogel-based biocontainment of bacteria for continuous sensing and computation[J]. Nature chemical biology, 2021, 17(6): 724-731.
[2] Su Zekun.Design, preparation and evaluation of adhesive hydrogel coated anti-migration intestinal scaffold [ D ].Dalian University of Technology, 2024. DOI : 10.26991 / d.cnki.gdllu.2024.000837.
[3] Zhao Q, Chen X, Ji W, Zhang J, Li X, Liu Z, Wang W, Liu H, Wang Y, Nan B, Li X, Wang Y, Liu J. From lab to table: Recent advances in the application of sodium alginate-based hydrogel beads in the food industry. Food Res Int. 2025 Oct;217:116843. doi: 10.1016/j.foodres.2025.116843. Epub 2025 Jun 9. PMID: 40597545.
[4] Chen Tiantian, Wang Congtian, Liu Wanxiang. et al. Preparation and drug release of pH-responsive cellulose nanofibrils / sodium alginate based hydrogels [ J ]. Chinese Journal of Papermaking, 2022,37 ( 01 ) : 16-21. DOI : CNKI : SUN : ZGZB.0.0.2022-01-003.
[5] Xu Bin, Duan Cuimi, Hao Tong, et al. Preparation of chitosan hydrogels and their adhesion and support to C2C12 cells [ J ]. Biomedical Engineering Research, 2008, ( 02 ) : 121-124. DOI : 10.19529 / j.cnki.1672-6278.2008.02.014.
[6] Shiraki M, Endo T, Saito T. Fermentative production of (R)-(−)-3-hydroxybutyrate using 3-hydroxybutyrate dehydrogenase null mutant of Ralstonia eutropha and recombinant Escherichia coli. J Biosci Bioeng. 2006;102:529–34.
[7] De Roo G, Kellerhals MB, Ren Q, Witholt B, Kessler B (2002) Production of chiral R-3-hydroxyalkanoic acids and R-3-hydroxyalkanoic acid methylesters via hydrolytic degradation of polyhydroxyalkanoate synthesized by Pseudomonas. Biotechnol Bioeng 77:717–722
[8] Liu Q, Ouyang S-P, Chung A, Wu Q, Chen GQ. Microbial production of R-3-hydroxybutyric acid by recombinant E. coli harboring genes of phbA, phbB, and tesB. Appl Microbiol Biotechnol. 2007;76:811–8.
[9] Matsumoto, K.i., Okei, T., Honma..., I. 2013. Efficient (R)-3-hydroxybutyrate production using acetyl CoA-regenerating pathway catalyzed by coenzyme A transferase. Applied Microbiology & Biotechnology, 97(1), 439-439.
[10] LIAO Junqi, HUANG Haining, CHEN Yinguang. CO2 promotes the conjunctive transfer of multiresistance genes by facilitating cellular contact and plasmid transfer [J]. Environment International, 2019, 129:333-342.
[11] ZHANG Ye,GU A Z,CEN Tianyu,et al. Subinhibitory concentrations of heavy metals facilitate the horizontal transfer of plasmid-mediated antibiotic resistance genes in water environment[J]. Environmental Pollution,2018,237:74-82.