Overview ​

Throughout the 2025 iGEM season, our HKUST (Guangzhou) team has made comprehensive contributions across wet lab, dry lab, and human practices. From constructing melatonin-based regulatory modules and engineering wax-coating biosynthesis pathways to developing AI-driven analysis tools and inclusive educational programs, every part of our work was collaboratively designed, tested, and integrated into a complete lychee preservation system. Our innovations span experimental, computational, and social dimensions, aiming not only to advance synthetic biology applications in food preservation but also to make scientific knowledge accessible to all. Below, we present the detailed contributions of each sub-team and individual members.

Contribution (WL) ​

Parts Contribution ​

Our contributions to iGEM Parts primarily fall into two categories: one is a universally applicable protocol for Parts, and the other is the Parts we have submitted-which can provide more Biobricks for future iGEM Teams.

Protocol: ​

Our team conducted wax detection on the parts for wax biosynthesis. We provide the protocol for wax detection, which facilitates direct use by subsequent iGEM teams.

Additionally, we have constructed lactose regulatory elements to achieve the regulation of wax ester synthesis. In this regard, we provide a protocol for the characterization of lactose regulatory elements. This protocol is based on fluorescent protein detection and has universal significance for verifying the efficacy of regulatory elements, which constitutes one of our contributions to future iGEM teams.

Wax Detection Protocol ​

(Deafult centrifuge setting: 4500rpm, 5min. Without extra instruction, the centrifuge operations uses the default setting)

1. Prepare the induction plate. Apply 0.02g sodium oleate powder to the liquid medium to reach the usual induction concentration of 0.2%. After applying sodium oleate, vortex for 5 minutes for full solution of the inducer. Incubate at 37°C in a shaker for the following 18h. (For full induction, consider extending incubation time to 48h)
2. Transfer 1.5 mL of bacterial culture into a 1.5 mL centrifuge tube. Centrifuge and aspirate the supernatant (culture medium).
3. Resuspend the pellet in 1.5 mL of PBS buffer, centrifuge, and aspirate the supernatant. Repeat this step once, but increase the centrifugation time to 10 minutes for the second wash.
4. Resuspend the pellet in 1.5 mL of PFA fixative solution, and let it stand for 1 hour to fix the cells. The fixation is conducted in cold stores at 4°C. Centrifuge and aspirate the supernatant.
5. Resuspend the pellet in 1.5 mL of PBS buffer, centrifuge, and aspirate the supernatant. Repeat this step once, but increase the centrifugation time to 10 minutes for the second wash.
6. Resuspend the pellet in 1.5 mL of staining wash buffer, centrifuge immediately, and aspirate the supernatant.
7. Add Oil Red O staining solution. Use a decolorizing shaker during our oil red staining step at 60rpm to improve this step for better results. Stain for 30min. (For better results, extend staining time would be a choice, but staining time should be within 60 min)
8. Resuspend the pellet in 1.5 mL of staining wash buffer and immediately centrifuge for 3 minutes. Increase the speed to 6000 rpm to shorten the time (to prevent excessive washing). Aspirate the supernatant.
9. Resuspend the pellet in 1.5 mL of PBS buffer, centrifuge, and aspirate the supernatant. Repeat this step once, but increase the centrifugation time to 10 minutes for the second wash.
10. After obtaining the wet cell pellet, add 200 µL of PBS to resuspend. Take cell suspension for following observation, e.g.

• Add 10 µL of cell suspension onto a glass slide to prepare a specimen
• Add 20 µL of cell suspension onto a confocal dish pre-treated with poly-L-lysine for observation
• Dispense 200 µL of cell suspension into a 96-well plate to measure the absorbance.

Lactose Regulatory Element Characterization Protocol ​

1. Revival of glycerol stock: Prepare 4x 1.5mL configure tubes and inject 200µL Amp Liquid LB into it. Use tips to take 20μL bacteria each and inject into 4 1.5mL configure tubes. Incubate on a bench top shaker at 37°C and 220 rpm for 1 hour.
2. Prepare the induction plate. For each inducted group, apply 48µL 100mg/mL IPTG stock solution to the solid medium to reach the usual induction concentration of 1mM/L. (For more even distribution of IPTG, consider further diluting to 240μL)
3. Aftering IPTG applying, keep the medium for 5 minutes for fully absorb. Evenly spread 200µL culture onto the solid medium. Incubate at 37°C in a cell incubator for 18 hours.
4. Observe the spread plate using an inverted fluorescence microscope, and visualize the green/red fluorescence under blue/green excitation light(475nm/520nm).

New parts ​

In this year's iGEM project, we submitted 14 new parts to the Registry. They are a variety of enzymes that play crucial roles in the synthesis of melatonin and wax. Additionally, our parts also include regulatory elements. All of these can provide a powerful toolbox for subsequent iGEM teams.

Part Number	Type	Description	Category
BBa_25UAWIT4	Coding	m-hTPH1 (Tryptophan 5-hydroxylase 2), an enzyme which is involved in the first step of serotonin biosynthesis, converting L-tryptophan into 5-HTP.	Basic Part
BBa_25B8Y7Z3	Coding	hPCDB1 (Pterin-4-alpha-carbinolamine dehydratase), an enzyme which accelerates the formation of MH2 (Dihydromonapterin).	Basic Part
BBa_25FVZG9V	Coding	hQDPR-C-Flag (Dihydropteridine reductase), and enzyme which catalyzes the conversion of MH2 (Dihydromonapterin) into MH4 (5,6,7,8-tetrahydrobiopterin).	Basic Part
BBa_25IP3NXW	Coding	TDC-C-His (Aromatic-L-amino-acid decarboxylase), an enzyme which catalyzes the decarboxylation of L-tryptophan to tryptamine and L-5-hydroxytryptophan to serotonin.	Basic Part
BBa_25RJYKMY	Coding	SNAT-N-His (Serotonin N-acetyltransferase), enzyme which catalyzes the penultimate step in the synthesis of melatonin, converting N-acetylation of serotonin into N-acetylserotonin.	Basic Part
BBa_25J405LE	Coding	COMT-N-Flag (Flavone 3'-O-methyltransferase 1), enzyme which catalyzes the final methylation step in the melatonin synthesis pathway, converting N-acetylserotonin (NAS) into melatonin.	Basic Part
BBa_25TRZRAU	Device		Composite Part
BBa_25FYXD6F	Device	P43-SpoVR-RBS-mTPH1-SpoVR-RBS-hPCBD1-SpoVR-RBS-hQDPR-rrnB T1-T7-P43-SpoVR-RBS-TDC-C-His-T7-rrnB T1 P43-SpoVR-RBS-SNAT-N-His-rrnB T1-T7-P43-SpoVR-RBS-COMT-N-Flag-rrnB T1-T7	Composite Part
BBa_25IXNS8H	Coding	MBP-AcrB-His (Acridine resistance protein B), which can catalyze the direct, four-electron reduction of fatty acyl-CoAs to their corresponding fatty alcohols	Basic Part
BBa_25XVCT22	Coding	G6PD-HA (Glucose-6-phosphate dehydrogenase), to produce NADPH	Basic Part
BBa_2533I9RA	Coding	MBP-WS/DGAT-flag (Wax Ester Synthase / Acyl-CoA:Diacylglycerol Acyltransferase), a key enzyme in neutral lipid synthesis	Basic Part
BBa_25LEDVII	Coding	λ Repressor-Myc, which can encode the cI repressor protein from bacteriophage lambda.	Basic Part
BBa_25UYV4GP	Plasmid Backbone	pBluescript II sk(+), a high-copy number plasmid backbone	Basic Part
BBa_25T6P0HB	Coding	WS/DGAT-RBS-G6PD-RBS-AcrB, Wax Biosynthetic System starting from Acyl-CoA	Composite Part

Our main contribution lies in the design of three composite part and the improvement of their characterization, which provides solid experimental data for subsequent iGEM teams that need to use these parts.

Serotonin Synthesis Part ​

Our team has submitted parts for serotonin synthesis. These parts are capable of utilizing TPH and TDC enzymes to catalyze the synthesis of serotonin from L-tryptophan. Furthermore, they implement the recycling of MH4 through hPCBD1 and hQDPR, thereby providing an adequate supply of cofactors for the TPH enzyme. Serotonin is a common drug and substrate, and our part is capable of synthesizing serotonin, which can provide support for subsequent iGEM teams with ideas related to this field. While utilising it in our project and submitting it into the Registry, we examined its characteristics. Here are our experiment results.

Western Blot: We have hQDPR-Flag and TDC-His in the plasmid, so we used western blot to detect them.

Since serotonin was detected using ELISA, we concluded that TDC-his was successfully expressed.

ELISA: Melatonin Synthesis PartWe generated a standard curve for the Serotonin ELISA Kit and subsequently detected the serotonin content in both the control group and the experimental group using this kit.

By using the t-test, we can find that there is a significant difference between the experimental group and the control group. We conclude that pWB980-ori-TPH-hPCBD1-hQDPR-TDC can convert L-tryptophan into serotonin.

Melatonin Synthesis Part ​

For our melatonin synthesis parts, it utilizes SNAT (serotonin N-acetyltransferase) and COMT (catechol-O-methyltransferase) enzymes to catalyze the synthesis of melatonin from serotonin.

While utilising it in our project and submitting it into the Registry, we examined its characteristics. Here are our experiment results.

Growth Analysis: We Expand culture of pHY300PLK-SNAT-COMT glycerol stock and pHY300PLK glycerol stock:

5ml liquid LB medium (Amp: 100ng/ml) + 10µl glycerol bacterial solution (4 replicate experiments).

Then, measure OD₆₀₀ at 0h, 1h, 2h, 3.5h, 4.5h, 6h, 7h, 10h, 14h and 16h during the expansion culture.

Figure 3. (a): Growth Curve of E. coli DH5α (OD₆₀₀) showing the growth dynamics of E. coli DH5α cultures containing pHY300PLK - SNAT - COMT and pHY300PLK plasmids over time, with OD₆₀₀ measured at different time points during expansion culture. (b): Data is presented as Mean ± SEM, with N = 4 biological replicates. Bar graph showing OD₆₀₀ values of E. coli DH5α cultures after 6 hours of cultivation, comparing those with pHY300PLK and pHY300PLK-SNAT-COMT plasmids. Statistical analysis via Mann - Whitney test (two-tailed) yielded p = 0.0286 < 0.05, significance denoted by *.

Western Blot: We have COMT-Flag and SNAT-His in the plasmids, so we used western blot to detect them.

COMT was successfully detected, as shown in Figure 1.1.

Results also show that COMT is expressed more in vector backbone pWB980-ori. We evaluated band gray-level values and performed one-way ANOVA to analyze the difference in COMT expression between the backbone vectors pWB980-ori and pHY300PLK. As shown in Figure1.2., there was a statistically significant difference between them. This result is expected since pWB980-ori is a high-copy plasmid while pHY300PLK is a middle-copy one.

SNAT-His was successfully expressed in both plasmid backbones, as shown in Figure2.1. and Figure2.2.

Resultes also shows that there is no significant difference between the level of expression of SNAT in plasmid backbone pWB980-ori and pHY300PLK. This is not expected. It might be because of the limited number of samples we tested.

ELISA: We generated a standard curve for the Melatonin ELISA Kit and subsequently detected the melatonin content in both the control group and the experimental group using this kit.

Due to the interference of bacterial proteins in ELISA detection, we control the OD .

This significant difference confirms that the biosynthetic pathway successfully enabled the experimental strain (pHY300PLK-SNAT-COMT) to produce a significantly higher amount of melatonin compared to the control.

Unfortunately, due to constraints in time and technology, we were unable to obtain more precise results.

However, We can preliminarily speculate that pHY300PLK-SNAT-COMT is capable of converting serotonin into melatonin.

Wax Synthesis Part ​

We have submitted a composite part that provides the iGEM community with a novel, convenient, and practical pathway for wax synthesis. We first transform fatty acyl-CoA into fatty alcohols by Acridine resistance protein B (ArcB), and then synthesize waxes from fatty alcohol and fatty acids by wax ester synthase/acyl-CoA:diacylglycerol acyltransferase(WS/DGAT). At the same time, we also express glucose-6-phosphate dehydrogenase(G6PD) to promote the synthesis of NADPH, providing sufficient sources for the reaction of first step.

The common pathway from fatty acid synthesis to wax production is: acyl-CoA → fatty acid → fatty aldehyde → fatty alcohol → wax (synthesis of macromolecular fatty acids and fatty alcohols). However, in typical bacteria (such as Escherichia coli and Bacillus subtilis), the genes encoding the individual enzymes in this pathway are missing. Therefore, completely reconstructing this pathway would require an enormous amount of work.

It is worth noting that, we discovered a unique fatty acyl-CoA reductase, AcrB, in the novel strain Marinobacter aquaeolei VT8, capable of directly catalyzing the reduction of fatty acyl-CoA to fatty alcohols(Willis et al., 2011).

Below are some experiments we conducted to measure specific characteristics of the protein within this composite part, which we hope will be helpful to the iGEM community.

Western Blot:

Due to technical and experimental limitations, we performed only one Western blot assay. In this experiment, we successfully validated that the molecular weight of the modified WS/DGAT protein is 98.6 kDa. Additionally, we conducted a T-test analysis of gray values. The results indicate that our band exhibits a significant difference compared to the negative control group.

Product Detection: We established five groups for full-wavelength absorbance detection. Since the study focused on the wax synthesis pathway initiated by acyl-CoA, oleic acid was added as the reaction substrate in all experimental groups. The groups were ultimately categorized as follows: PBS control group, PBS + Oil Red O control group, empty plasmid group, empty plasmid + oleic acid group, and target plasmid + oleic acid group. We employed normalization to ensure experimental results were unaffected by irrelevant factors such as bacterial culture concentration and background values.

Oil Red O exhibits a strong absorption peak at 500 nm and demonstrates absorbance exceeding the control across the 400-650 nm range. Our experimental results corroborate this observation.

To eliminate extraneous factors and better analyze curve variations, we normalized the absorbance curves during analysis. Comparing the expression vector + oleic acid group with the other two control groups, we observed that the experimental group exhibited significantly higher absorbance at 500 nm than the other two groups. Additionally, it demonstrated higher absorbance below 650 nm, with the overall curve closely matching the absorbance curve of Oil Red O. However, due to experimental constraints, we did not replicate the experiment and thus could not perform a significant analysis.

Contributions in Intergrated Human Practices ​

In the process of planning and implementing Human Practices (HP) activities, we have developed a set of effective methods and accumulated valuable experience, which we are eager to share with everyone here.

1. Stakeholder Identification ​

One of our main contributions to Human Practices this year was establishing a structured method for identifying and analyzing stakeholders across the entire lychee industry chain. To understand how different groups interact within the production and distribution network, we created a comprehensive stakeholder map covering farmers, distributors, wholesalers, and retailers. Rather than limiting ourselves to single interviews, we intentionally selected multiple representatives from each stakeholder type to capture diverse perspectives and avoid overgeneralization.

Our team conducted field visits and comparative interviews across Guangzhou—a region with a mature lychee industry—to gather first-hand information from both small family farms and large-scale commercial orchards. This comparative design provided a more balanced and realistic understanding of the challenges in lychee preservation, including post-harvest decay, transportation losses, and market priorities.

Through these investigations, we identified the distinct core concerns of each stakeholder group—farmers focused on storage conditions, distributors on shelf life, and sellers on product appearance. We transformed these findings into a practical framework for stakeholder engagement, featuring tailored interview outlines and categorized feedback documentation. This framework guided our later HP activities and can serve as a reference for other iGEM teams aiming to conduct industry-based Human Practices.

By systematizing the stakeholder identification process, our contribution lies not only in gathering information but in building a replicable model for mapping and comparing industry actors, ensuring that social and economic factors are fully integrated into project design.

2. Our Innovative HP Methodology ​

We believe that every iGEM team should develop its own HP logic that genuinely connects science with society. From our 2025 practice, we formed an iterative, evidence-based methodology that supported our work in three key areas: activity design, field engagement, and reflection-driven improvement.

At the beginning of the project, we established clear Integrated Human Practices (iHP) goals focused on understanding real-world needs within the lychee preservation industry chain. Our methodology started from stakeholder engagement—including farmers, distributors, and consumers—to identify the social, economic, and ethical dimensions of our problem. As the project advanced, feedback from experts and communities continuously reshaped our direction, ensuring that every design decision was both scientifically sound and socially relevant.

This process naturally evolved into a dynamic cycle of research–reflection–redesign, distinct from the conventional DBTL model. It allowed us to refine not only our experimental plan but also our educational and outreach strategies. By treating every activity—whether expert consultation, public education, or policy discussion—as a source of feedback, we built a living framework that grew alongside the project itself.

We encourage other teams to adopt a similar reflective mindset: instead of following fixed templates, build HP methods that evolve from genuine interaction with your communities and stakeholders. Our approach demonstrates that meaningful Human Practices emerge not from structure, but from continuous learning and integration between scientific innovation and human values.

3. Approach to Leverage Researchers' Expertise in iHP ​

Our team contributed a structured and transparent approach to incorporating expert insights into Human Practices. Instead of conducting one-off interviews, we developed a feedback-based communication model that connected expert advice directly to project decisions.

During the project, we consulted researchers such as Professor Stanley Brul and other specialists in microbiology and food safety. Their feedback—especially on the limitations of using pH as a biological signal—was systematically summarized, discussed within the team, and recorded in our design documentation. This internal reflection mechanism helped us identify weak points early and adjust our focus toward more feasible and safe directions.

By documenting how expert opinions were collected, evaluated, and translated into project choices, we aimed to make the process reproducible for future teams. Our contribution lies not only in obtaining expert advice, but in establishing a method for integrating scientific consultation into iterative design, bridging communication between researchers and iGEM practitioners in a traceable way.

Contributions to Education ​

This year, our educational work focused on developing adaptable and evidence-based approaches to science education, aiming to make synthetic biology more inclusive and engaging for diverse audiences. Instead of listing activities, we summarized and refined our practices into a set of practical frameworks that future teams can adopt or adapt.

1. Feedback-Driven Iteration ​

Each activity served as a testbed for the next. We systematically collected feedback from participants and instructors, evaluated the effectiveness of each session, and used these insights to improve later ones. This cyclical improvement model proved especially effective for engaging younger and special-needs audiences, forming a replicable feedback-based education loop.

2. Theory-Guided Design ​

Our curriculum design incorporated three key learning approaches—active learning, problem-centered learning, and project-based learning—inspired by consultation with educational and sociological experts. We demonstrated how these theories can guide activity structure and participant engagement, offering a model for iGEM teams to design educational content with pedagogical grounding.

3. Audience Stratification Framework ​

We proposed a structured age-stratified education model, summarizing the most effective communication strategies for each group—from kindergarten children to university students and the general public. This framework enables other teams to quickly identify suitable formats, levels of difficulty, and interaction methods for their target audiences.

4. Inclusive Education for Special Needs ​

Our collaboration with Qihui School led to the development of scenario-based teaching methods tailored for students with autism and mild intellectual disabilities. These adaptive designs demonstrated how synthetic biology education can be made accessible and enjoyable for all learners.

All of these frameworks, along with documentation templates and feedback forms, are openly shared in our Wiki to support other teams in building educational programs that are both effective and inclusive.

Contribution to Community ​

1. iGEMer Community Responsibility ​

As one of the cohost of iGBA (an iGEMer community in Great Bay Area), we actively coordinated exchange activities for various iGEM teams, providing a platform for communication and continuously sharing experiences with different teams. Throughout the whole community activity this year, our team and HKUST are responsible for the stream media promotion and our team actively participated in the preparation of the iGBA event.

2. Promoting Inclusive Science Education ​

We extended community engagement beyond universities through our collaboration with Qihui School, where we designed tailored courses for students with autism and mild intellectual disabilities. This program introduced an adaptive, scenario-based learning framework, demonstrating how synthetic biology can be taught inclusively. The approach has since become a model for our team's educational outreach and a reference for other iGEM teams aiming to engage special-needs communities.

Contributions in Dry Lab ​

Overview ​

Our dry lab team contributed to the iGEM community by developing computational tools, datasets, and modeling pipelines that go beyond our own project needs. We designed our dry lab work to be open, reusable, and adaptable, so that future iGEM teams and researchers can benefit directly.

SynbioMCP – AI Agent Framework for Dry Lab Tasks ​

What we built: SynbioMCP is an AI-agent software tool that connects LLMs with common bioinformatics and modeling tools.

Why it matters: Many dry lab tasks (e.g., translating sequences, visualizing protein structures, optimizing DNA sequences) are repetitive and technically demanding. SynbioMCP reduces the learning curve by automating workflows while leaving important biological decisions to human researchers.

How others can use it: The software itself is already useful for future iGEM teams' work. Plug-in architecture allows future teams to add new tools with minimal effort. Proof-of-concept implementation with visualization, sequence optimization, and simulation calls. Documentation and demo video included on our wiki and GitHub.

User Experience Research: Performed grounded theory–based analysis of user willingness to engage with offline experiences, providing a framework for future iGEM teams conducting human practices.

BactaGenome – Genome-Scale Modeling ​

What we built: A deep learning pipeline inspired by AlphaGenome, tailored to bacterial genome regulation (tested on E. coli).

Why it matters: Gene expression prediction is a fundamental challenge. Our model investigates transcriptional regulation patterns by integrating RNA-seq (PRECISE dataset) and regulatory networks (RegulonDB).

How others can use it: Our training code, model configurations, and evaluation scripts are openly documented. We share insights about loss function choice (combined vs. MSE) and practical training observations that may help other teams starting genome-scale modeling. Outreach to experts (e.g., DeepMind researchers, Prof. Pan at HKUST(GZ)) enriches reproducibility and interdisciplinary understanding.

Molecular Dynamics & Thermodynamic Simulations ​

What we built: A molecular dynamics (MD) simulation workflow and analysis script for enzyme thermal stability testing.

Why it matters: Many teams use GROMACS or similar software but face steep setup difficulties. We provide annotated protocols and scripts for: System preparation Running MD with GPUs (including handling CUDA, NVML, and maxwarn issues) Energy and radius of gyration (Rg) analysis

How others can use it: Our reusable scripts can accelerate future teams’ simulation studies.

Lychee Recognition Dataset & Model ​

What we built: A dataset and machine learning model for non-destructive evaluation of lychee freshness and variety using lychee imaging.

Why it matters: Freshness evaluation is often destructive (peeling, enzyme assays). Our work explores image-based evaluation to support both consumers and industry.

How others can use it: The dataset structure and model pipeline (variation classification, days-after-harvest) can be adapted to other fruits or crops. Demonstrates how to combine agricultural needs with accessible ML workflows.

Datasets, UI, and Tools ​

Lychee Dataset Plan: Designed and implemented a dataset collection pipeline, including a UI for structured metadata input.

Collected Data: Version 1 and Version 2 dataset completed. Data include image, and days-after-pick.

Reusability: Documentation ensures future teams can build upon or extend our datasets.