ProbioEase Flat-Roofed Building

Overview

Our project's dry lab component features the development of ProbiEase, a novel software platform designed to accelerate research in therapeutic probiotics. ProbiEase addresses a critical bottleneck in synthetic biology: the gap between vast, unstructured scientific literature and actionable, data-driven experimental design. By integrating a curated database, a natural language interface, and interactive visualizations, the platform serves as an end-to-end solution for candidate generation and wet lab candidate selection.

Function

ProbiEase is built on two core components: a comprehensive database and a multi-module user interface.

1. Curated Probiotic-Disease Database

（1）We systematically collected and organized data on over 300 probiotic strain-disease associations, covering both potential and clinically validated therapeutic effects.

（2）Crucially, every association is linked to its supporting scientific literature, ensuring all data is verifiable, credible, and adheres to scientific standards of evidence. This curated dataset forms the knowledge backbone of the entire platform.

2. The ProbiEase Platform

（1）Strain Card Display: Provides a concise, high-level overview of each probiotic strain in a "card" format. This feature allows for rapid browsing and comparison. Users can click through to access a detailed page with comprehensive data and linked publications.

（2）Natural Language Q&A (RAG-based): To make our database accessible to users without computational expertise, we implemented a Retrieval-Augmented Generation (RAG) system. Users can pose complex questions in natural language (e.g., "Which probiotics show potential for alleviating Parkinson's disease symptoms?"). The system retrieves relevant data from our curated database and leverages a Large Language Model (LLM) to synthesize a coherent, evidence-based answer. This significantly lowers the barrier to entry for complex data exploration.

（3）Knowledge Graph Visualization: We developed an interactive network graph to visualize the complex relationships between probiotics and diseases. Users can explore the data from three perspectives: a strain-centric view, a disease-centric view, or a global overview. This visual tool helps uncover non-obvious connections and patterns that are difficult to discern from text-based data alone.

Static--Dynamic Dual FBA Modeling

Background and Motivation

To further confirm the reliability, transparency, and safety of ProbioEase-recommended probiotic strains in human-like environments, we introduce Flux Balance Analysis (FBA) and Dynamic Flux Balance Analysis (dFBA) for dual-dimensional modeling. These approaches simulate metabolic behaviors and interactions, enabling a comprehensive evaluation of potential risks associated with probiotic candidates from the ProbiEase platform.

FBA

With the recommended strains obtained, we conducted FBA modelling to simulate their growth within the nasal cavity, thereby enabling preliminary screening.

Specifically, our contribution in this module can be concluded as follows:

1. Predict steady-state metabolic flux distributions of individual strains under simulated human conditions by constructing a stoichiometric matrix and optimizing an objective (e.g., biomass growth).
2. Identify key metabolites (e.g., lactate, short-chain fatty acids) produced by each strain and screen their standalone safety by cross-referencing toxicity resources, without requiring kinetic parameters.
3. Establish a single-strain metabolic "fingerprint" (nutrient needs, secretion profile, limiting steps) that sets intake bounds and boundary conditions for downstream dFBA scenarios.
4. Provide a baseline safety layer for individual metabolite outputs, forming a foundation for subsequent interactive risk evaluation.

dFBA

By integrating FBA with ordinary differential equations (ODEs), dFBA modelling's contribution can be drawn as follow:

1. Extend FBA into time-dependent settings by coupling with ODEs to simulate multi-strain interactions (nutrient competition, cross-feeding, succession) in human-like environments.
2. Detect process-level risks emergent from interactions, such as excessive organic acid accumulation leading to pH imbalance, toxin buildup, or inflammation triggers.
3. Use co-culture simulations (e.g., Lactobacillus and Bifidobacterium, adjustable) to reveal risks not seen in isolated strains and to locate critical exchange fluxes and pathways.
4. Support scenario testing of environmental and dosing conditions to anticipate thresholds where harmless single-strain outputs become community-level hazards.

Result

This static-dynamic dual-dimensional modeling---static FBA for individual metabolite safety and dFBA for interactive community risks---further refines ProbiEase-recommended strains. After confirming the safety of the strains, further medical therapy or wet experiment can be conducted。

More details about the dual-dimensional modeling can be found in the 'Model' [Link]

Drug delivery

Compared to traditional drug delivery strategies, such as oral drugs or delivery systems based on intestinal colonisation with probiotics, an innovative nasal drug delivery method was used in this study. This strategy has several significant advantages over traditional methods:

1. Nasal drug delivery can effectively avoid the first-pass effect, thus improving the bioavailability of the drug.
2. It is able to bypass the blood-brain barrier, which enhances the efficiency and targeting of drug delivery to the brain.
3. The engineered probiotic platform constructed in this study is able to continuously secrete therapeutic molecules to achieve stable and long-lasting drug release to cope with the side effects caused by drug fluctuations during Parkinson's disease treatment.

This study introduces this drug delivery strategy to the iGEM community, aiming to provide an innovative delivery route for research teams working on the treatment of neurological diseases and seeking to overcome the limitations of the blood-brain barrier, with a view to expanding research ideas and therapeutic strategies in this field.

Adhension Module

We enhanced the colonisation ability of Lactobacillus plantarum WCFS1 in the olfactory epithelium by overexpressing its OppA protein gene, and constructed a three-bacteria copolymerisation system based on the principle of antigen-antibody specific binding. In addition, AutoDock and HDOCK simulation, we predicted the binding ability between the constructed Lp_0018 protein and acetylheparin sulfate on the surface of olfactory epithelial cells, systematically evaluated the advantage of Lp_0018 over other OppA proteins and surface adhesion proteins in terms of the binding performance, and examined the effect of the structure of heparin on the interaction.

Our contributions are reflected in the following aspects:

1. Stable adhesion between Gram-positive and negative bacteria was successfully achieved, and a three-bacteria copolymerisation system with controlled colonisation was constructed.
2. It provides key technological support for the development of drug-delivery probiotic platforms and offers new strategies that can be applied to the design of multimicrobial systems.
3. This study preliminarily theorized the specific colonization capability of Lactobacillus plantarum WCFS1 in the olfactory epithelium by performing molecular docking between the oppA protein and various ubiquitous endogenous ligand molecules, thereby laying a theoretical foundation for subsequent advancement toward practical clinical applications.

Quorum Sensing Module

In our research, we have achieved two-way communication between Gram-negative and Gram-positive bacteria

This year, the TJUSX team has engineered a cross-species bi-directional communication Part Collection, which provides a universal solution for the exchange of signals and the regulation of interactions between different species in the synthetic bacterial community.

Our contributions are reflected in the following aspects:

1. The first bi-directional communication framework between Lactobacillus and E.coli.
2. Validation and standardisation of the core device Part uploads.
3. Functional extension: communication signals drive downstream functional gene expression.

This provides an important library of tools for future multi-species colony engineering, which can be used for: rational colony design, symbiotic system regulation, metabolic collaboration optimisation. We hope that this part collection will become a "standard communication interface" for the design of future multispecies systems.

Therapeutic Module

In this study, we constructed engineered Escherichia coli Nissle 1917 strains capable of synthesising levodopa and glutathione, respectively, and systematically evaluated the yield of both types of engineered strains and their growth status in the production of target therapeutic molecules. In addition, we conducted literature research on the therapeutic effects of levodopa and glutathione.

Our work provides theoretical and experimental data support for subsequent iGEM team studies on the application of such therapeutic molecules.

Safety Module

In this project, the engineered probiotic was designed to colonise the olfactory epithelium, so we focused on how to prevent the strain from escaping to the external environment or other non-target body sites. Based on the differences in environmental characteristics between the olfactory epithelium and other environments, we designed a dual safety switch system: a temperature-sensitive switch to inhibit the survival of the strain in the external environment at low temperatures, and an anoxic switch to inhibit the proliferation of the bacterium in areas of low oxygen (e.g., the intestinal tract), thus avoiding ectopic colonisation. We experimentally verified the reliability and effectiveness of the above two types of switches under different environmental conditions.

Our contributions are reflected in the following aspects:

1. The safety strategy adopted in this study provides a practical reference for other teams to design controlled colonisation and biocontainment systems in synthetic biology systems, especially live bacterial therapeutics.
2. The validated temperature-sensitive and hypoxic switches not only complement the range of available safety switches, but also save time and resources for preliminary validation in subsequent studies.

Fluorescent Characterization of Microbiota

In our wet experiment section, we also explored fluorescence characterisation of microbial communities. Our contributions to this piece are categorised as follows:

1. Fluorescent labelling tool selection: three fluorescent proteins were selected from the iGEM 2025 Distribution to validate the utility of this standardised component library and provide a selection reference for similar studies.
2. Gram-positive bacteria labelling challenge: The signal intensity of the fluorescent protein expressed in Lactobacillus plantarum was found to be significantly lower than that of the negative bacteria, presumably because the cell wall of Gram-positive bacteria is thicker, and the fluorescence is not significant, and this result provides an important reference for the selection of fluorescent labelling strategies against Gram-positive bacteria.
3. Validation of strain coexistence: The stable coexistence of the two strains was confirmed by multi-technical means, and a characterisation strategy was established to provide methodological support for the construction and assessment of synthetic microbial communities.
4. Measurement of gene expression regulatory elements: The promoter strength of Lactobacillus plantarum was quantitatively determined, and the relevant data have been shared to provide reliable data support for subsequent gene expression regulation in this bacterium.