LOADING
Our hardware design aims to address two critical challenges: experimental characterization of genetic circuits and industrial-scale production and application. To gain deeper insights into the development and practical application of oral bacterial therapeutics, we constructed an oral hydrogel microsphere delivery system for in vivo bacterial delivery and an in vitro turbidostat model for genetic circuit characterization. Together, these systems provide crucial insights for developing ultra-long-acting GLP-1-based therapies for diabetes and weight management.
To ensure the intact delivery of engineered E. coli to the intestinal tract for drug synthesis, we continuously optimized the hydrogel delivery system for gastric acid and digestive enzyme resistance, controlled intestinal release and long-term colonization, providing the necessary conditions for EcN to function.
The in vitro turbidostat, meanwhile, aims to simulate the dynamic equilibrium of bacterial density within the gut, reflecting induction efficiency in a near-physiological context. It addresses the issue of unsynchronized bacterial growth and genetic circuit expression in standard batch cultures, simulating colonization kinetics and providing deeper understanding of genetic component response times to facilitate future in vivo experiments.
These two systems form a complete "Delivery-Characterization" technological pipeline for engineered probiotics, laying the foundation for the practical production of oral formulations and the precise regulation of genetic circuits.
Figure 1. Our integrated hardware module including delivery and characterization
The hydrogel microsphere delivery system is the core hardware for the oral delivery of engineered EcN. Further optimized with enteric capsule coating and thiolation modification, it can achieve the dual functions of "Protected Delivery + Enhanced Colonization".
Oral probiotics must survive the harsh gastric environment with low pH and high digestive enzyme concentration before reaching the intestines. Conventionally unprotected probiotics are easily inactivated by gastric acid, resulting in very low intestinal survival rates. Through our preliminary Engineering iterations, we found that plain hydrogel alone was insufficient for fully protecting E. coli in the stomach. To solve this, we designed and optimized a "hydrogel microsphere + enteric capsule" composite system. The enteric capsule remains intact in Simulated Gastric Fluid (SGF, ~pH 2.0) and achieves delayed, site-specific release in Simulated Intestinal Fluid (SIF, ~pH 6.8), preserving bacterial viability. This prevents direct contact between the hydrogel-encapsulated EcN and gastric acid, while the hydrogel microspheres act as a secondary barrier, slowing EcN release to prevent rapid probiotic washout and facilitate sustained colonization[1].
Figure 2. Schematic diagram of EcN encapsulation based on hydrogel microspheres and gelatine enteric capsules
Even if probiotics successfully reach the intestines, they are easily cleared by peristalsis, hindering long-term colonization. To enhance the adhesion of the delivery system to both the intestinal mucosa and the EcN itself, we modified one of the hydrogel materials, Carboxymethyl Cellulose (CMC), via thiolation (-SH) to form CMC-SH. This effectively strengthens adhesion and promotes long-term colonization. Thiol groups can form stable disulfide bonds with cysteine and thiol-containing proteins on the intestinal mucosa and EcN surface, significantly improving hydrogel adhesiveness. By using the hydrogel as an intermediary, enhanced adhesion to both the gut and the bacteria slows the washout of EcN during digestion and peristalsis, prolonging the residence time of the engineered bacteria in the gut and ensuring sustained GLP-1 synthesis.[2]
Figure 3. Schematic diagram of improved EcN adhesion and colonization via thiol-modification of hydrogels
(1) Preparation of enteric capsule-coated hydrogel microspheres and controlled release experiments.
(2) Adhesion experiments of thiolated hydrogel (CMC-SH).
Enteric capsules maintained structural integrity after 1 hour in SGF, whereas uncoated hydrogel microspheres were severely damaged, leading to a significant drop in EcN survival, with the survival rate being nearly zero. Upon transfer to 37°C SIF, the enteric capsules dissolved completely within 10 minutes, releasing the hydrogel microspheres. EcN release from capsule-protected microspheres was slower. After 7 hours in SIF, the release was only 54.2% of that from the unprotected group, and the release process hadn't reached a plateau, indicating sustained release. The cumulative release of EcN from hydrogels after 7 hours in SIF exceeded the initial loading, suggesting that EcN can proliferate within the hydrogels to some extent.
After rinsing, bacteria in uncoated and CMC-coated areas were easily washed away, whereas the CMC-SH coated area and the bacteria adhering to its surface remained largely unaffected, showing no significant difference from the unrinsed control. This demonstrates that thiol modification significantly enhances the hydrogel's ability to adhere to EcN, effectively retaining even bacteria not fully embedded within the hydrogel matrix.
Figure 4. Comparison of adhesion performance of CMC and CMC-SH hydrogel coatings on LB agar plates for EcN based on the running water rinsing experiment
(1) Incomplete Release due to Double-Layer Protection: High mass transfer resistance might prevent complete bacterial release. Missing the optimal release window could lead to bacterial clearance, or incomplete hydrogel disintegration might result in colonization levels too low to compete with native flora, hindering effective chemical induction and long-term drug synthesis.
(2) Storage of Capsule-Encapsulated Formulations: As the internal hydrogel is hydrous, and given the biosafety circuits involved, storage outside 37°C conditions can lead to spontaneous bacterial death and decreased viability over time within the hydrogel.
(3) Manufacturing Reproducibility and Industrial Scale-up: To ensure semi-quantifiable, customizable treatment with live biotherapeutics, consistent batch quality control and shelf-life validation are essential. Unexpected bacterial death during storage, which is hard to quantify, would require potency titration before administration.
(4) Biosafety Considerations: Engineered E. coli used in oral formulations must be strictly contained to prevent environmental release outside the lab. Safeguards against leakage and horizontal gene transfer are necessary to mitigate environmental risks.
(1) Optimization of Dual-Layer Carriers: Future designs could explore double-layer hydrogel structures with optimized surface area for contact with GI fluids. Fine-tuning the cross-linking density can validate formulations that balance gastric protection with timely release.[3] Personalized customization based on individual digestive/GI states may be needed.
(2) Improved Storage of Oral Formulations: Incorporating protective agents like glycerol and developing lyophilized powder formulations could reduce metabolic activity, extend shelf life, and ensure recovery viability, preventing irreversible damage.
(3) Industrial Production of Hydrogels: Leveraging established batch production equipment, such as microfluidic bacterial injection and fluidized bed granulation facilities, can ensure minimal batch-to-batch variation. Regular random sampling for viability and potency testing, along with master cell bank maintenance, can provide genetically stable bacterial sources.
(4) Enhanced Biosafety: Biosafety requires ensuring viability before administration and implementing multiple safety layers against environmental escape including biocontainment circuits. Strict adherence to sterile protocols in labs and production facilities is crucial to confine engineered bacteria to their intended therapeutic purpose and minimize side effects.
GlucoXpert's "Enteric Capsule + Hydrogel" design overcomes oral delivery barriers and promotes long-term colonization. Compared to complex designs like double-layer hydrogels, this system offers advantages in simplicity, reliability and public acceptance, facilitating subsequent industrial production and clinical translation. Simultaneously, it provides a stable delivery platform for engineered EcN, creating conditions for optimizing carrier structure such as porosity and size and establishing quantitative models linking carrier properties to probiotic function.
A turbidostat is a microbial cultivation device that maintains a constant microbial concentration by monitoring culture turbidity and dynamically adjusting the nutrient feed rate, providing suitable conditions for simulating in vivo microenvironments.[4] Our goal was to build a simple in vitro turbidostat to simulate the state of engineered GLP-1-producing bacteria after successful colonization and reaching dynamic equilibrium in the gut, thereby providing more realistic data for optimizing inducer addition and timing.
We performed periodic sampling for validation and characterization once the turbidostat reached dynamic stability. After achieving a dynamic equilibrium in bacterial density, we injected a specific volume of inducer at designated time points to simulate the effect of consuming green tea or other induction conditions. Subsequent analysis involved interval sampling and detection of collected samples. This approach helps us precisely monitor and assess GLP-1 secretion and genetic component response times post-induction.
· DF-101S Heated Magnetic Stirrer
· Rubber stopper with ports
· Kamoer Peristaltic Pumps (Models: DKCP-S10, NKCP-C-B08B)
· 100 mL Glass Three-Neck Round-Bottom Flask
· Glass inlet/outlet elbows and connectors
· 1 m Silicone tubing for peristaltic pumps
· 1000 mL Media feed bottle
· Tubing flow regulator
· Beakers
· Sterile sealing film
· Collection tube
· Magnetic stir bar
· Sterile syringes
Figure 5. Schematic diagram of the in vitro turbidostat
The setup is shown in Figure 5. The core is a three-neck flask containing the engineered bacterial culture, stirred by a magnetic stirrer. The flask is immersed in a water bath for temperature control at 37 °C. Fresh LB medium is continuously supplied via a peristaltic pump (left), with its flow rate dynamically adjusted based on real-time OD600 readings (using a flow regulator and pump speed: increase flow if OD600 > 1.4; decrease if < 1.4). The insertion depth of the right outlet tube is fixed to maintain a constant culture volume of ~50 mL. Effluent is continuously removed by another peristaltic pump (right) into a collection tube for monitoring OD600 and Flou/OD600. The top neck is sealed with a breathable sterile sealing film to maintain oxygen levels and prevent contamination. All components, including media and tubing, were sterilized before use.
Figure 6. Photograph of the successfully assembled turbidostat
We used E. coli harboring the AND-gate plasmid for testing. Inducers including protocatechuic acid (PCA) and sodium cholate were added via sterile syringe into the stabilized turbidostat system.
EcN with an initial OD600 of 0.02 was inoculated into a three-neck flask with 50ml LB medium for a 12h pre-culture to reach the exponential phase. Post-pre-culture, samples were taken periodically from the collection tube to measure OD600. The inflow/outflow pump rates were adjusted to maintain the OD600 fluctuating around 1.4 for approximately 5 hours, indicating system stability at the setpoint. Once stabilized, PCA and sodium cholate were injected into the flask (defining this as time zero) to final concentrations of 400μM and 25μM, respectively to simulate the gastrointestinal environment and validate AND-gate functionality. Samples were taken every 0.5 hours to measure OD600 and fluorescence intensity, calculating the Fluo/OD600 ratio (normalized against negative controls to eliminate density effects) and analyzing the fluorescence trend over time post-induction.
The system maintained a stable turbidostatic state for nearly 5 hours, with OD600 fluctuating around 1.4 without a significant rising or falling trend. Following inducer addition, the Fluo/OD600 ratio of the AND-gate engineered bacteria increased, peaked around 3 hours, and then declined. This indicates that the AND-gate sensing bacteria can rapidly respond to PCA induction in the "simulated gut" and synthesize the downstream product GLP-1 around 3 hours. This hardware system can be preliminarily applied for validating and characterizing wet lab results.
We successfully constructed a turbidostat to simulate the dynamic equilibrium OD600 level of engineered bacteria post-colonization in vivo. Experimentally, we determined the time-dependent accumulation of GFP post-induction. Our data demonstrate that the engineered bacteria can achieve regulated, rapid and controlled GLP-1 secretion in a simulated in vivo setting. This implies the feasibility, controllability and potential effectiveness of using colonized E. coli for GLP-1 secretion in humans' body. In the future, we will continue exploring conditions that more accurately mimic human colonization, such as incorporating interference from native gut microbiota, to further refine the simulation and measurement of engineered bacterial colonization and GLP-1 secretion.
The "Turbidostat-based Simulated In Vivo Colonization Validation" platform established by GlucoXpert provides a reproducible in vitro characterization tool for engineered probiotics. Subsequent studies incorporating native gut microbiota interference and co-culture induction can provide critical data references for gut microbiome modulation and the clinical translation of Live Biotherapeutic Products (LBPs).
This project addresses both delivery and characterization needs through dedicated hardware designs, functionally validated with wet lab results. While currently at a preliminary in vitro stage, with detailed genetic circuit characterization and oral formulation safety/efficacy requiring further in vivo validation, these hardware designs significantly enhance our understanding of oral biotherapeutic development and efficacy. They offer a novel perspective for the industrial development of such formulations.
Aiming to develop a truly long-acting GLP-1-based "live medicine", our hardware design focuses on industrial feasibility. Preliminary verification suggests strong potential for scalable production. Standardized fermentation of engineered bacteria combined with established enteric capsule delivery technology offers a significant cost reduction compared to complex biosynthetic GLP-1 processes. This paves the way for a novel, patient-compliant treatment regimen.
[1]Toshifumi Udo,Zijin Qin,Yang Jiao,Rakesh K. Singh & Fanbin Kong.(2025).A comprehensive review of mathematical modeling in probiotic microencapsulation.Food Engineering Reviews,17(2),1-21.
[2]Zhang, H., Liu, Z., Fang, H., Chang, S., Ren, G., Cheng, X., Pan, Y., Hu, R., Liu, H., & Wu, J. (2023). Construction of probiotic double-layered multinucleated microcapsules based on sulfhydryl-modified carboxymethyl cellulose sodium for increased intestinal adhesion of probiotics and therapy for intestinal inflammation induced by Escherichia coli O157:H7. ACS Applied Materials & Interfaces, 15(15), 18569–18589. h
[3]Aleš Ručigaj & Tilen Kopač.(2025).Mathematical modeling of swelling and shrinking dynamics in pH-sensitive hydrogels composed of alginate, anionic cellulose, and chitosan.Polymer,328,128452-128452.
[4]Bryson V, Szybalski W. "Microbial selection". Science, 1952, 116(115)