Overview

  In response to the growing prevalence of metabolic diseases such as obesity and diabetes caused by high-sugar, high-fat diets worldwide, the promotion of low-calorie functional sugars has become a crucial direction for improving public health. D-allulose, a natural rare sugar with about 70 % of the sweetness of sucrose but nearly zero calories, exhibits various physiological benefits such as regulating lipid metabolism and combating obesity. It has been approved as a food additive in many countries and officially recognized as a new food ingredient in China in July 2025, demonstrating broad application prospects. However, its industrial-scale production remains limited by the catalytic efficiency and stability of the key enzyme, D-tagatose-3-epimerase (DTE). Conventional enzyme engineering strategies like directed evolution are time-consuming and often yield limited improvements, highlighting the urgent need for new approaches integrating computational design and experimental validation. To address this, we developed a complete AI-powered enzyme design-evaluation-validation pipeline, incorporating tools such as LigandMPNN, AlphaFold3, molecular docking, and dynamic simulations, enabling rapid rational design of DTE enzymes and resulting in two novel catalytically active DTE enzyme parts. We also constructed an allulose biosensor based on the RNA-Pepper system, allowing for fast and quantitative detection of environmental allulose concentrations, thereby enriching the iGEM parts library. Furthermore, we established a multi-scale, closed-loop modeling system that integrates enzyme kinetics, fluorescent signal readouts, and population dynamics to achieve systematic quantitative analysis of enzyme function within living cells. Through expert interviews, public science services, and campus promotions, the project also strengthened the integration of scientific research, education, and social responsibility, forming an innovative platform that combines research, service, and education.

Contributions of Wet Lab

Enriches the iGEM parts library

  We have introduced an allulose biosensor based on RNA-pepper (BBa_25N1O2HD) into the iGEM parts library. Through the repression–derepression system of the allulose operon, the concentration of environmental allulose is linked to the expression level of the pepper RNA. The fluorescence intensity emitted by the addition of HBC dye reflects the concentration of allulose in the environment. The RNA-pepper system allows for faster expression and superior temporal resolution in detection compared to traditional GFP-based systems. Quantitative characterization of the system has been completed, demonstrating that the sensor can respond significantly to allulose concentrations ranging from 10 to 1000 mM/L in the environment. We invite readers to consult our Parts section for comprehensive details. Utilizing this system, we re-measured the relevant data of a several part (BBa_K2791019) and supplemented our experimental findings, which further validates the system's reliability.

  Moreover, by incorporating different repression-derepression systems, this framework can be further engineered to create other biosensors. In the era of advancing quantitative biology, systems capable of precise quantitative responses to input signals are particularly crucial.

Novel and catalytically active DTE enzymes were obtained

  Two novel and catalytically active DTE enzymes (BBa_25XZ56RI、BBa_25WB3AMN) were obtained through a combination of AI design, modeling, and experimental verification. This enzyme catalyzes the conversion of fructose to D-allulose, thereby contributing to the further development and progress of the alternative sweetener industry.

  More importantly, this workflow integrating AI design, modeling/screening, and experimental validation can be generalized to the screening of other enzyme types and the mining of novel iGEM parts.

Contributions of Dry Lab

A Complete AI-Powered Enzyme Design-Evaluation-Validation Pipeline

  We provide a comprehensive AI-driven enzyme design and validation pipeline that integrates LigandMPNN for sequence design, AlphaFold3 for structure prediction, multi-tool molecular docking, and GROMACS for molecular dynamics simulations, enabling rapid iterative optimization of enzyme proteins. Through wet-lab validation, we have successfully obtained engineered enzymes with novel structural features and catalytic potential, demonstrating the effectiveness of this design scheme. The pipeline employs LigandMPNN for structure-based global sequence redesign, Expasy ProtParam for initial theoretical parameter screening, AlphaFold3 to verify protein folding reliability, CB-Dock2, DiffDock, and SwissDock to evaluate substrate binding modes, PLIP for atomic interaction analysis, Gromacs for dynamic simulations, and PyMOL & VMD for structure visualization and trajectory analysis. This solution significantly reduces the experimental costs associated with traditional directed evolution, overcomes the limitations of point mutations, and utilizes exclusively open-source or free online tools, allowing iGEM teams to directly reuse and extend the pipeline. It forms a complete closed loop from computational design to experimental validation, providing a generalizable rational design pathway for the field of enzyme engineering.

The pipeline and tools we have established are as follows:

Enzyme Design Program: LigandMPNN

Initial Sequence Screening: Expasy ProtParam

3D Structure Visualization/Prediction: AlphaFold3

Multiple Docking Tools: CB-Dock2. DiffDock. SwissDock.

Static Atomic Interaction Analysis: Protein-Ligand Interaction Profiler

Dynamic Simulation Process: Gromacs 2025.2

Visualization Tools： PyMOL. VMD.

 The figure shows the protein generated by AlphaFold3 and subsequent sequence design, demonstrating the well-preserved overall protein framework.

 The figure shows the protein-ligand atomic interaction analysis generated by PLIP, indicating a clear docking effect.

 The figure shows the trajectory plot from molecular dynamics simulations processed by VMD, displaying the dynamic changes of the ligand and substrate.

Multi-Scale, Closed-Loop Modeling System

  We propose an integrated computational framework based on "Four Models," deeply integrating multi-dimensional data including enzyme kinetics, fluorescent signal readouts, and cellular population burden to achieve systematic quantitative analysis of enzyme function within living cells. This framework first employs a Catalytic Circuit Model to precisely quantify the kinetics of enzyme expression processes, encompassing the formation, turnover, and functional conformational changes of both homo- and heterodimers, laying the foundation for understanding enzyme dynamic behavior in real cellular environments. Building upon this, the Reporting Circuit Model, relying on the efficient Pepper-HBC RNA aptamer system, establishes a linear mapping relationship from substrate concentration to transcriptional activation and subsequently to fluorescent signal output, ensuring that changes in enzyme activity can be rapidly and stably converted into detectable optical signals.

  To overcome interference from population heterogeneity and metabolic burden, we introduce a Population Dynamics Model, which simultaneously tracks the dynamic balance between mutation accumulation effects, growth rate changes, and product synthesis yield, thereby evaluating the performance and stability of engineered strains at the system level. Concurrently, a Half-Life Correction Module models and compensates for the temporal degradation behavior of key biosensor components (including the RNA aptamer and the enzyme protein itself), effectively eliminating signal drift caused by differences in molecular stability.

  This multi-scale, closed-loop modeling system achieves real-time, quantitative monitoring and prediction of enzyme activity at single-cell resolution. It features high modularity and portability, enabling rapid adaptation to various metabolic pathway optimization and biosensor development scenarios. It provides a powerful theoretical tool and computational platform for the rational design and high-throughput screening of synthetic biology components.

Contributions of Human Practice

I. Guide Project Education and Innovation Capacity Building

  Through interviews with Professor Lu Xiaoyun and Professor Li Jianjun, we gained deeper clarity on the educational philosophy that student-led research projects should uphold: "capability cultivation over short-term outcomes." Professor Lu emphasized that the true value of such projects lies in practical training, which helps students accumulate professional skills applicable to cutting-edge fields. Professor Li highlighted the central importance of market orientation in application-driven projects. These insights have enriched the team's understanding of the educational essence of scientific research, transforming students from passive executors into innovative actors with strategic thinking. Consequently, the team has reframed the process of technical optimization as a platform for cultivating problem-solving skills.

II. Multi-Dimensional Health Science Services

  Operating within an integrated "science-service-research-policy" framework, we have conducted a series of health service initiatives in communities and senior care institutions across six provinces and municipalities in China. The team developed three themed micro-lectures — "Scientific Nutrition," "Chronic Disease Prevention" and "Early Cancer Prevention" — to accurately deliver knowledge on nutrition planning, diabetes management, and disease prevention to middle-aged and elderly audiences. Through these service activities, the team not only strengthened their professional expertise but also cultivated the ability to translate theoretical knowledge into practical social service, effectively enhancing public awareness of chronic disease prevention and healthy sweeteners.

III. Enhance Public Awareness

  Through booth promotions, poster distribution, and questionnaire surveys on campus, the team systematically promoted the iGEM competition philosophy and the health benefits of D-allulose. Analysis of 200 valid responses revealed a 70 % acceptance rate of D-allulose among young adults aged 18-35, with its "minimal impact on blood glucose" and "near-zero calorie" properties identified as core advantages. Beverages were ranked as the most promising application scenario. This comprehensive "design-execution-analysis" process enabled the team to master social survey methodologies and data-driven decision-making skills. The findings not only provided empirical support for the project's future application directions but also enhanced the relevance and effectiveness of synthetic biology outreach in campus settings.

IV. Enhancement of Educational Innovation and Team Capabilities

  Throughout the process of public science outreach and team collaboration, members progressively developed an integrated skill system encompassing "scientific research, science communication, and social service." This model effectively cultivated students' critical thinking and self-adaptive learning abilities, achieving an educational cycle of "learning through service, optimizing through communication." Consequently, the project has evolved into a comprehensive educational platform that integrates professional education, general education, and character development.