Home | TJUSX - iGEM 2025

What are the main challenges?

Engineered probiotics holds broad prospects for treating human diseases. The field faces several key challenges:

Lacks a unified and standardized development workflow.
Efforts remain fragmented , experience-dependent , and difficult to reproduce .

Final consequence: Significantly slows the translation of innovative therapies from lab to clinic.

Our story unfolds in three parts:

Platform Design

PD Treatment

Wet-Lab Execution

Building the ProbiEase

Architecture and Development Pipeline

The Closed-Loop Paradigm

We adopted an evidence-based closed-loop paradigm.

We first developed basic and professional versions of a Q&A system. Using Node.js, we built a web-based interactive interface that supports retrieval and explanation of domain-specific questions, thereby enabling online data querying and AI-assisted Q&A services on a unified platform. This provided an innovative approach for "disease-probiotic" association analysis.

1. Raw Data

Gathering 211 bacterial strains

DETAILS

During the data collection phase, we internally compiled and integrated data on edible probiotics, gathering 211 bacterial strains. In the data cleaning process, we strictly adhered to the principle of evidence prioritization, first evaluating each strain's safety for consumption, potential benefits, and relevance to target diseases. We then systematically searched professional literature databases such as PubMed, selecting only entries with clear theoretical foundations and peer-reviewed evidence while excluding records lacking reliable support. This minimized bias and ensured high credibility and traceability of the data.

2. Literature Support

(1) Extracted "strain-disease" relationships from papers → The knowledge graph

(2) used gutSMASH and antiSMASH → predict the static metabolic potential and specialized pathways

DETAILS

In the subsequent data processing and functional annotation stage, we extracted "strain-disease" relationships from papers to lay the foundation for node connections in the knowledge graph. Additionally, we used tools such as gutSMASH and antiSMASH to predict the static metabolic potential and specialized pathways of the strains. We also integrated professional databases to supplement phylogenetic, phenotypic, and safety information, thereby forming comprehensive functional profiles of the strains.

3. Database Construction

We constructed a structured triple database centered on "strain-disease-paper" as the knowledge foundation for the ProbiEase online Q&A platform. This database supports both the general retrieval and the professional version.

DETAILS

Based on this, we constructed a structured triple database centered on "strain-disease-paper" as the knowledge foundation for the AI-PPDD online Q&A platform. This database supports both the general retrieval and interpretation functions of the basic version and the evidence strength evaluation, mechanism inference, and comparative analysis of solutions in the professional version. It also enables bidirectional queries, allowing users to "search for diseases by strain" and "search for strains by disease." In practical applications, our system can provide targeted inquiries and recommendations from diseases to strains, outputting literature-verified candidate strain sets along with their potential mechanisms of action. It clearly specifies applicable scopes and uncertainties, delivering comprehensive professional responses to queries.

4. Modeling & Simulation

We employed dFBA methods to simulate community dynamics of candidate strains.

DETAILS

Subsequently, we employed dFBA methods to simulate community dynamics of candidate strains, quantitatively analyzing resource competition, metabolic interactions, and potential conflicts to evaluate the feasibility of co-culture. This provided parameter guidance for designing combination strategies and optimizing experimental conditions, thereby preemptively excluding high-risk options at the computational level and prioritizing pathways with strong synergistic potential.

5. Experimental Validation

DETAILS

We validated co-culture of the two engineered strains and made a preliminary assessment of their coexistence using fluorescence-based cell counting. Targeting Parkinson’s disease as the indication, we further modularized and engineered the consortium to enhance therapeutic performance: strengthening target-specific colonization, increasing therapeutic output, implementing division-of-labor control between strains, and incorporating biosafety safeguards to prevent escape.

The Application of the ProbiEase

Addressing the Parkinson's Challenge

Characteristics of Parkinson's Disease

A neurodegenerative disease
Progresses slowly and silently
Mainly attacks the body's motor system
Particularly plagues elderly groups
The morbidity and mortality situation is grim

Strategy 1

Current Approach

Oral Administration

Convenient and non-invasive
High patient compliance
Applicable to various PD drugs
Blood-brain barrier
First-pass hepatic metabolism
Frequent dosing

Strategy 2

Alternative Route

Intranasal Delivery

Bypassing the blood-brain barrier
Rapid onset of action
Efficient absorption
Short residence time
Unstable and poorly sustained drug levels
Low targeting specificity

Strategy 3

Our Solution

Intranasal Delivery of Engineered Probiotics

More than just:

Convenient and non-invasive
High patient compliance
Applicable to various PD drugs
Bypassing the blood-brain barrier
Rapid onset of action
Efficient absorption

We further added:

Precise Targeting
Sustained and Controlled Drug Release

Implement of Our ProbiEase

Engineered Nasal Probiotics for Brain-Targeted Therapy.

Overview

We introduce a novel therapeutic strategy for Parkinson's Disease based on a three-strain nasal-adhesive probiotic consortium comprising two engineered starins of E. coli Nissle 1917 and one Lactiplantibacillus plantarum WCFS1. These strains were co-engineered with four functional modules that enable sustained release of two therapeutic molecules and promote colonization of the nasal mucosa.

1. Adhesion Module

Enables stable consortium residence in the nasal cavity.

(1) Utilizes the OppA protein for specific host receptor binding.

(2) Enhances bacterial cohesion via an antigen-antibody system, forming a robust community.

DETAILS

2. Drug Delivery Module

Enables local synthesis of Parkinson's therapeutics.

(1) Produces L-DOPA to replenish dopamine levels in the brain.

(2) Simultaneously delivers glutathione to combat oxidative stress and protect neurons.

DETAILS

3. Control Module

Achieves on-demand drug release via bacterial communication.

(1) Establishes a cross-species signaling circuit using AHL and AIP.

(2) Tightly couples therapeutic production with successful colonization.

DETAILS

4. Safety Module

Incorporates multiple biocontainment strategies for enhanced safety.

(1) Includes thermo- and hypoxia-sensitive suicide switches to prevent escape.

(2) Features an inducible kill switch as an emergency brake.

DETAILS