HKUST iGEM 2025 Team Slupeer is committed to addressing heavy metal pollution through an innovative, eco-friendly approach using bio-engineered bacteria for cadmium extraction. Guided by iGEM’s core engineering principles - design, build, test, and learn - our team has iteratively refined the project through multiple engineering cycles to achieve its current optimized form.
Through a series of Engineering cycles, Team Slupeer successfully constructed the plasmid for the expression of Lpp-OmA-Metallothionein (BBa_25OWO9RF) via Gibson assembly and transformed it in both E. coli BL21 Rosetta and P. putida KT2440 (please refer to the Cloning l Protocols and Experiments for more details). Upon successful transformation, 4 main experiments – cell viability assay, Western blot, TEV protease cleavage test, and ICP-OES – were performed, which confirmed the expression and functionality of the metallothionein, PflQ2MT, and showed high potential of using bio-engineered P. putida for cadmium extracting bioremediation.
Figure 1. Bar chart of OD600 of EC and PP at 30 µM Cd2+. Student’s t-test ∗∗P \< 0.01, ∗∗∗P \< 0.001, n \= 3.
Figure 1 shows that E. coli with MT (ECMT) has a higher OD600 than E. coli (EC), which justifies that MT does increase the survival of bacteria in cadmium solutions. Furthermore, at 30 µM Cd2+, P. putida has a significantly larger OD600 than E. coli. Hence, it is concluded that the metallothionein protein increases the survivability of the bacteria, and P. putida is a more advantageous candidate as a chassis to remove cadmium.
Figure 2. ICP-OES result of E. coli with different treatments in a 60 µM Cd2+ solution. n \= 2-3.
Moreover, we utilized ICP-OES to detect the cadmium concentrations after the solutions were incubated with ECMT overnight. As seen in Figure 2, the normalised value of the change in Cd2+ concentration to the OD showed that E. coli with MT causes a higher decrease in cadmium concentration, which proposes its potential to bind Cd2+ and extract it from the solutions. The data further affirms the engineering success of the basic part (BBa_25VCEOE5), and evidence supporting its binding function.
While our Experiments page documents a range of data with potential insights and promise, we believe these two figures most strongly demonstrate our success in achieving the goal of our original engineering success. To note, the goal of Engineering Cycle 0 was to produce Wet Lab data that affirms in-silico work surrounding the function of PflQ2MT as a potential bioremediating protein candidate. Subsequently, Engineering Cycle 1 aimed to utilise a more characterised system in E. coli to test the function of the MT circuit and confirm the subsequent added functions compared to the wild type strain. From our results, we believe these figures demonstrate strong engineering success and show we have managed to accomplish and undergo the two original goals set for these engineering cycles.
If you would like to know more about how we evaluated our Engineering Success for real-world implemntation, please take a look at ourApplication Development Page.
Our journey of engineering began with an in-depth literature review and discussions to identify effective metal-binding proteins for heavy metal extraction. Continuous engagement with field experts provided critical feedback, shaping the project’s direction and enhancing the design and functionality of our system.
The primary focus of the first design part was to narrow down our project’s scope. So to better understand cadmium pollution and determine cadmium as our dedicated primary goal, we conducted an interview with Dr. Xiyue JIA, who was an expert in cadmium contamination in rice.
Next, as we found EC20 and metallothionein (MT) to be strong candidates with their capabilities of binding to heavy metal ions, hence we also reached out to Dr. Weon Bae, the first author of multiple EC20-related articles to gain more insights. To learn more about the interviews, refer to integrated Human Practice page.
Drawing inspiration from successful heavy metal sequestration strategies documented within the iGEM Registry, we determined that surface display was the optimal approach compared to intracellular binding. This choice offers significant advantages, such as enhanced accessibility for binding target ions, simplified downstream separation of the biosorbent, and potential for regeneration and reuse. To achieve robust surface binding, we screened various membrane localization systems and ultimately selected a membrane fusion protein system designed for anchoring the binding protein onto the cell membrane. We eventually selected the well-documented Lpp-OmpA hybrid system. In this construct, the Lpp (Lipoprotein) domain serves as the membrane anchor, ensuring stable localization to the outer membrane, while the OmpA (Outer Membrane Protein A) domain functions as the linker protein, providing a stable scaffold for displaying the target binding domain on the cell surface ((Higham et al., 1985; INP-NC, n.d.; Krogsgaard et al., 2000; Narita et al., 2005; Parks et al., 2024)).
Figure 3. The optimal selection process of our anchor protein. From general meeting on 25th Apr.
As part of our iHP work, we developed our own criteria on the ideal metal-binding protein which involved:
In order to choose the best binding protein, instead of relying solely on Human Practices, we also wanted to screen through the proteins ourselves. Hence, we build a computational workflow to narrow down potential candidates and to try to predict the potential properties of said candidates.
To investigate the structural properties of EC20 and other MTs, phytochelatins, and functionally homologous proteins for cadmium binding, we utilized AlphaFold 3 to predict their folding, both in isolation and in complex with the Lpp-OmpA delivery system. The process of building the workflow and subsequent analysis is described thoroughly in the “MT Cycle 0” under our Model Page. As a general summary, it involved building a library of different parts from existing literature and the iGEM Registry and analysing pLDDT and pTM scores to assess the reliability of predicted structures. Candidates were first folded in isolation to ensure pLDDT values indicated medium to high confidence before proceeding with Lpp-OmpA folding simulations. Zinc ions, as cadmium analog, were used as a proxy in AlphaFold 3 simulations, given their chemical and structural similarities and documented affinity of MTs for both metals. As metal-binding proteins have dynamic structures, the presence of ions is crucial to provide context in the folding of the protein. Structural stability and binding potential were evaluated using pLDDT (>70 for reliable conformations), pTM (>0.6 for overall model quality), and ipTM scores (for protein-metal complex stability). This approach enabled the identification of well-defined protein structures with potential for effective cadmium binding under physiological conditions, providing a robust framework for further development of metal-binding proteins.
Due to limitations in AlphaFold 3 for predicting Cd²⁺ binding, we employed the MIB2 server to identify and model metal ion-binding sites in PflQ2MT, evaluating 18 metal ions with focuses on ions which are more commonly found in sludge or may compete with Cd². This included Cu²⁺, Fe³⁺, Mg²⁺, Mn²⁺, Zn²⁺, Cd²⁺, Fe²⁺, Ni²⁺, Hg²⁺, Co²⁺ and Pb²⁺. applications.
Due to various reasons, which were outlined in our Model page, we switched from EC20 to PflQ2MT. Our analysis of PflQ2MT revealed that up to 9 cysteine residues participate in binding, as determined by docking scores and structural analysis using MIB2. Several metal ions, including Cu²⁺, Fe³⁺, Zn²⁺, Hg²⁺, Co²⁺, and Au²⁺, showed affinity for individual cysteine residues, but only Mn²⁺, Cd²⁺, Fe²⁺, Ni²⁺, Cu⁺, Ba²⁺, and Pb²⁺ effectively bind within the protein’s three-dimensional structure. PflQ2MT exhibited the greatest preference for Cd²⁺ and Cu⁺, each binding to all 9 cysteines, followed by Fe²⁺ binding to 7 cysteines and Pb²⁺ to 5 cysteines. Although Cu⁺ binds effectively, its instability and tendency to oxidize to Cu²⁺, which does not bind any cysteines, makes it less competitive than Cd²⁺. Consequently, Cd²⁺ is the most preferentially bound metal ion by PflQ2MT due to its strong interaction with all 9 cysteines and minimal competition from Cu²⁺.
Our Dry Lab outputs highlighted the limitations of utilising prediction tools with static outputs such as the MIB2 algorithm and AlphaFold 3. Particularly, its application in predicting cadmium ion binding to dynamic proteins like LgiMT1 and LgiMT2 highlight the need for more robust computational approaches. MIB2's treatment of proteins as static structures fails to account for significant topological changes during metal ion binding, leading to inaccurate predictions of binding sites and affinities. Thus, this shifted our paradigm in future cycles.
Our dry lab screening results demonstrated that the PflQ2MT metallothionein is superior to EC20 in several key aspects. Specifically, PflQ2MT offers a better burden-effectiveness ratio, suggesting optimal performance relative to the metabolic load it imposes on the host cell. Furthermore, the gene encoding PflQ2MT is structurally simpler, featuring fewer repetitive sequences, which facilitates easier and more reliable synthesis and cloning. This selection aligns with our overall strategy of utilizing surface expression, which is inherently advantageous because it minimizes the metabolic burden placed upon the cell compared to high-level intracellular expression. By localizing the binding protein externally, we ensure that the host cell's internal machinery remains largely dedicated to growth and viability, optimizing overall biosorbent production efficiency.
Our focus on heavy metal remediation is crucial because cadmium is recognized as a serious and pervasive environmental pollutant. As a highly toxic heavy metal, cadmium poses significant risks to human health and ecosystems, necessitating the development of efficient and sustainable bioremediation solutions. In the ideation stage of our project development, we found challenges in project positioning and developing an optimal application model. As a result, we sought stakeholders from multiple fields including heavy metal pollution, cadmium and sludge treatment to better fit our project into the demands of the industry. The detailed process and feedback documentation can be found in our Application Development page.
From our research, we found that many parts we wanted to test were more well-characterised in E. coli, making it a better-documented and more appropriate system to initial our design and experiments. Moreover, we need to start constructing our plasmid from E. coli for easier transformation and selection.
The circuit we designed for expressing MT in E.coli contains 3 main parts, which are anchor protein fused with a protease cleavage site, downstreaming the cadmium binding protein PflQ2MT (BBa_25VCEOE5). You may refer to our Design | Description page for further detail. Our principle design was based upon maximising the function of PflQ2MT for cadmium adsorption. As the cell tolerance to cadmium for both E. coli and P. putida were not well-characterised, nor was the rate or mechanism which Cd2+ entered the cell, our team believed a surface-expression of PflQ2MT using the Lpp-OmpA system would provide greater potential for protein binding and cell stability.
To drive high and controlled protein synthesis, we selected the T7 promoter. This promoter system is widely utilized in E. coli expression strains (which carry the T7 RNA Polymerase) due to its ability to achieve high expression levels rapidly, simplifying the overall circuit design and leveraging readily available, popular plasmid backbones. For predictable translation initiation, the Ribosome Binding Site (RBS) was chosen from the well-documented Elowitz library, ensuring reliable and robust protein production. Furthermore, to enhance the functionality and modularity of the fusion protein, we inserted two critical sequences between the Lpp-OmpA anchor and the PflQ2MT cargo. The first is the TEV (Tobacco Etch Virus) protease cleavage site. This site was selected after screening several candidates because it serves a dual purpose: it is essential for verifying successful surface expression through protease digestion assays, and it introduces modularity, allowing the concept to be easily extended to display binding domains for other heavy metals.
Figure 4. The screening process of protease cleavage site. From general meeting on 9th May.
To ensure maximum accessibility and efficiency of the TEV protease, a flexible (Glycine)6 linker (GGGGGG) was strategically inserted between the anchor protein and the cleavage site. This linker provides spatial separation and conformational flexibility, which significantly enhances the cutting efficiency of the TEV protease (Chen et al., 2016). For the mean of simplest, we also select T7 terminator.
Plasmids play a significant role in successful genetic engineering. Specifically, we selected pET11a plasmid to express the main circuit optimized for E. coli. pET11a demonstrates high-level expression of recombinant proteins in E. coli, utilizing the strong T7 promoter system. As a widely used plasmid, it contains Ampicillin resistance, MCS, and pBR322 ori. We used it to assemble our Main Circuit (BBa_25OWO9RF) into plasmid (BBa_25JXOJIK) optimized for effective expression in E. coli DH5a and Rosetta.
Figure 5. pET-11a (BBa_25OZD053). Standard plasmid in E. coli T7 expression system (not compatible with P. putida), ampicillin resistance, pBR322 ori, high copy number, MCS, Donated to us by our advisor Changxin, Expresses Main–E. coli-Optimized–Circuit.
You may like to take a look at our Cycle 1 Circuit Design logbook below:
Our work involved assembling our composite part in the vector pET11a and expressing the plasmid in E. coli BL21. We also aimed to characterise its expression and downstream function in E. coli. Our goals for this cycle included confirming its expression and membrane localisation as well as its function as a metal-binding protein. For more details about the protocol design, please refer to the Experiments and Protocols page.
In this cycle, we also aimed to verify the correct folding, stability and binding under our desired conditions using Dry Lab. This would involve modelling the Lpp-OmpA-MT circuit into biologically accurate E. coli membranes and using GROMACs to generate trajectory files for further analysis. To see more details on this, please refer to the “MT Cycle 1” tab under our Model page.
With the addition of His-tag downstream our MT, we detected the expression of MT through Western Blotting with monoclonal and HRP-conjugated secondary antibodies. From the experiment, bands of 23 kDa showed that PflQ2MT was successfully expressed in E. coli. Further, we tested whether the presence of MT increases the survivability of the bacteria compared to the wild type by the growth curve of E. coli (EC) and E. coli with Metallothionein (ECMT). Moreover, a seven-amino acid-long sequence (Glu-Asn-Leu-Tyr-Phe-Gln-Gly/Ser) recognized specifically by TEV protease, was inserted between Lpp-OmpA (BBa_K1694010) and MT sequences and to verify the surface expression of the proteins by successful cleavage. In addition, Inductively-Coupled Plasma Optical Emission Spectrometry (ICP-OES) was utilized to test out the functionality of the MTs by measuring the concentrations of cadmium before and after ECMT treatment. Also, a survival test between wild-type E. coli and E. coli with our MT construct transformed. The details of the protocols and experiments are recorded in the Experiments page.
Higher final OD600 values from the growth curve suggest that E. coli with the metal-binding protein shows better survivability in the cadmium environment. However, both E. coli and E. coli with metallothionein did not show growth as reported in papers, while we suspected that it was due to their inability to adapt in an environment of heavy metals.
Generally readily available E. coli strains, like C+, K12, DH5ɑ, and Rosetta(DE3), were obtained from our lab stock. Eventually, we leveraged two common strains, including E. coli DH5ɑ for efficient plasmid transformation and cloning, and E. coli Rosetta(DE3) for high-level protein expression, particularly due to its compatibility with the T7 expression system, according to Du et al. (2021). The use of these standard strains allowed for rapid prototyping and verification of key genetic components before transitioning to a chassis better suited for the final application environment.
After more research, we noticed P. putida KT2440 as a better candidate as our chassis which exhibited more ideal characteristics and was more well-researched (Martínez-García & De Lorenzo, 2023). This decision was driven by several critical advantages that align with the project's real-world application in heavy metal remediation of industrial sludge. P. putida KT2440 offers superior environmental relevance, as it is a soil bacterium naturally found in environments similar to our target sludge (De Lorenzo et al., 2024). Furthermore, comparative studies demonstrated that P. putida exhibits significantly better resistance to the high-cadmium concentrations and acidic conditions characteristic of the target environment than E. coli (Higham et al., 1985). Experimentally, P. putida also showed a superior growth rate and higher viability at the relatively lower incubation temperature of 30 ℃, and demonstrated greater overall resistance to metabolic stress. These combined factors made P. putida the optimal host for developing our system.
For this cycle, we need to transform our chassis from E. coli to P. putida. Within such a transformation, we need to redesign the nearly whole workflow, including our circuit design, wet lab protocols, and more, while remaining our fundamental concept unchanged, to better improve our Main Circuit for a better fit of practical use, and to verify our concept on our new chassis.
P. putida KT2440 is a bacterium found in sludge. Due to its natural habitat, it has resistance towards high concentrations of heavy metals and extreme pH levels. P. putida showed characteristics of a chassis that was capable of fulfilling our goals of extracting cadmium within sludge.
To further ensure that P. putida was a suitable chassis, we planned another cell viability assay to compare its survival compared to E. coli. Within sludge, there exists a vast array of heavy metal ions, contaminants, and slightly acidic conditions (pH 4-5). Thus, this required a more robust chassis that was resistant to such conditions. Also, we determined the optimal P. putida growth environment with 30℃, as well as constructed a P. putida transformation protocol, documented in our Protocols.
Our new chassis P. putida KT2440 was kindly provided by Prof. Conghui You from ShenZhen University, whose group is working on catabolic pathways in Pseudomonas putida. In order to ensure the expression of protein could occur in sludge, our goal was to determine whether the bacteria would survive in the concentration range where Cd2+ was found in sludge. Furthermore, we aimed to characterise the extent of resistance P. putida had in sludge. For further details, please refer to the Experiments page.
Similarly to before, we prepared different concentrations of Cd2+ solution and allowed P. putida and E. coli to grow. Their absorbances were read periodically every 30 minutes. Their end OD600 absorbances were recorded and compared analytically. Particularly, P. putida was able to survive significantly better at 30 micromoles concentration of Cd2+ and persisted well up to 60 micromoles concentration of Cd2+. For more insights on the results, please refer to the Experiments page.
From the assay, we observed that P. putida indeed had a better survivability in cadmium compared to E. coli. From the previous literature review, Cd2+ concentration in sludge was found to be around 4 mg/L. Thus, this showed that P. putida was able to endure the desirable range and was thus a more suitable circuit to express the MT.
After confirming P. putida has better resistance than in our practical environment, we started to transform our Main Circuit, which was already verified in E. coli, to our new selected chassis. This transformation includes new parts, new vectors, new protocols, and a new mindset.
In this cycle, our work was to transform the MT into P. putida. We integrated our MT main circuit into pSEVA441 and transformed the whole plasmid into P. putida. Compared to E. coli, P. putida KT2440 is not a chassis as popular in synthetic biology, and the T7 expression system does not function. We obtained the RBS, promoter, and terminator libraries of P. putida KT2440 from different articles (Elmore et al., 2017; Benedetti et al., 2021; Silva-Rocha & de Lorenzo, 2023) and utilized compatible Standard European Vector Architecture (SEVA) vectors for protein expression in P. putida KT2440 from previous studies. Additionally, rather than an IPTG-induced promoter, we also aimed to test whether P. putida could survive the metabolic burden if the MT was placed downstream to a constitutive promoter. Under such circumstances, we selected EM7 promoter (BBa_K4278705), pJL1-sfGFP RBS (BBa_25TMR0E8), and eT1 terminator (BBa_25RQDKM3) to construct our circuit (BBa_250SPQMJ).
pSEVA441 is a standard cloning plasmid with spectinomycin resistance, MCS, and pRO1600 ori. It functions in both E. coli (DH5a and Rosetta (DE3)) and P. putida KT2440
We used pSEVA441 to express the Main Circuit optimized for P. putida KT2440.
Figure 6. pSEVA441 (BBa_2543S1TA). Cloning Plasmid Works in both E. coli and P. putida, spectinomycin resistance, MCS, pRO1600 ori, low copy number, expresses Main–P. putida KT2440-Optimized–Circuit. Gratefully obtained from SEVA.
You may like to take a look at our Cycle 2 Circuit Design logbook below:
As a slight alteration to Cycle 1, we transformed the circuit in the vector pSEVA441 which appeared to show greater native compatibility to P. putida. Following Gibson Assembly, transformation into E. coli DH5a was first done, and the miniprep product was extracted. Finally, the plasmid was transformed into P. putida using electroporation. For more details, please refer to the Experiments page.
Similarly to before, we prepared different concentrations and left P. putida and E. coli growing in them and read their absorbances every 30 minutes. Their end OD600 absorbances were recorded and compared analytically. Western Blot was done and confirmed that the circuit was able to constitutively express in P. putida. TEV protease cleavage assay was performed, with confirmed capability of cleavage site. Regrettably, time restrictions resulted in an inability to definitively confirm its surface expression in P. putida, nor was a metal-binding assay using ICP-OES done, which will be implemented in our future plans. For more insights and results, please refer to the Experiments page.
Our results with P. putida showed it provided more robust characteristics while still being able to successfully express the MT. Thus, our work with E. coli shows promising potential that the MT could carry out the same functionality in P. putida. With future plans, our team hopes our results show future potential for a cadmium-specific bioremediation solution. Also, the enhanced capabilities of P. putida allowed us to not only transfer the Main Circuit but also to design and integrate two additional, application-critical circuits: a dedicated cadmium Biosensor and a Kill Switch for safety.
After transitioning to a more optimal chassis P. putida and successfully retrieving our Main Circuit for cadmium surface binding, we took a deeper look into how to fit into the need of a practical use as a one-stop solution towards cadmium remediation from sludge. Eventually, we came up with an extraordinary idea, which is to add a cadmium biosensor into our chassis, for the indication of the remediation progress.
To construct a biosensor, we followed a fundamental logic, starting from an input, then processing, eventually transferring to a readable output. Within this logic, our Biosensor circuit has gone through multiple times of iteration for optimal design and part selection.
In our project, we aimed to develop a biosensor capable of monitoring cadmium concentrations in real time. Existing cadmium biosensors typically face two major limitations:
This led to our ideation of creating a biosensor which reacts in real-time. Achieving a genuine real-time response requires optimizing the kinetics of the fluorescent reporter protein in two critical ways: fast expression (to rapidly generate a large signal) and rapid degradation (to ensure quick signal clearance, preventing carry-over from previous measurements and allowing for accurate, dynamic monitoring).
We initially focused on the degradation aspect by incorporating a degradation tag at the C-terminus of our reporter protein. Literature reviews, which highlighted the effectiveness of tags like the LVA degradation tag confirmed this strategy. Implementing a faster degrading reporter system is essential for making our biosensor suitable for real-world, continuous monitoring applications.
Concurrently, to control and fine-tune the expression level, we utilized the Ptrc-lacO promoter. This IPTG-inducible system allows us to precisely adjust the expression rate, balancing the need for high output with the metabolic load on the cell. Furthermore, for the reporter component itself, we introduced a new, improved variant of the fluorescent protein mScarlet-I3-NCwt into the iGEM Registry (BBa_25IVGUN2). Characterization and testing of this improved reporter are vital steps in validating the performance of our real-time biosensor system.
For details of the biosensor design, please refer to our Design | Description page.
pSEVA234 with the lacI-Ptrc system was chosen as an expression vector for P. putida KT2440. pSEVA234 has Kanamycin resistance, MCS, and pBBR1 ori, demonstrating high expression efficiency in both P. putida KT2440 and E. coli. We used pSEVA234 to express the biosensor circuit and the kill-switch circuit.
Figure 7. pSEVA234 (BBa_25XDR5JO). Expression plasmid, lacI-Ptrc system, works in both E. coli and P. putida, kanamycin resistance, MCS, pBBR1 ori, medium-high copy number, expresses the Biosensor Circuit and the Kill Switch. Gratefully obtained from SEVA.
You may like to take a look at our Cycle 3.1 Circuit Design logbook below:
To make sure that our biosensor design actually works, we first designed a testing circuit for the mScarlet and the LVA degradation tag. The circuit is regulated by IPTG, allowing us to alter the expression level in order to test fluorescence and degradation at ease.
Figure 8. Demonstration of our Biosensor Degradation Test Circuit.
The main focus of the first Biosensor cycle is to test the parts that are reviewed in the literature but not yet tested extensively. In the preliminary testing, we tested where the LVA tag added to the mScarlet works with IPTG induction. After the induction, IPTG was washed away with PBS to test out the degradation rate of fluorescence. Please refer to our Experiment page and the ‘Biosensor Degradation Test’ protocol.
From the results of the degradation test, we were confident that the fluorescent protein (mScarlet) worked as expected. Though the degradation of the fluorescent signal was not obvious after washing away IPTG, we believed that the main reason might be the leakage of the promoter (since the fluorescence signal was detected even before adding IPTG). This urges us to look for other effective degradation systems. Despite such flaws in our report system, we still went forward, to design a Biosensor with other properties facilitating real-time detection.
After a few iterations, we ultimately discovered and selected the Toehold Switch system as the processing system, and the cadR system as our input. We successfully constructed our full Biosensor, and made efforts in its testing and characterization. With a well-documented future plan on testing the sensitivity of the regulatory system, our Biosensor became a powerful tool for our genetic modified machine to be applied practically.
We went on to further test the whole biosensing system. In addition to the LVA tag, the RiboSwitch was designed from Toehold Switch, which will lead to mRNA encoding the mScarlet continuously being transcribed in the absence of cadmium. With the upstream regulatory and input system of tetR (BBa_259S6D4J, regulated by anhydrotetracycline (aTc)) and cadR (BBa_251WLXJ9), we successfully composed our full Biosensor circuits (BBa_256ORVBQ).
Figure 9. Demonstration of our full Biosensor circuit, excelling in real-time capability and sensitivity.
Our full Biosensor circuit was built with three fragments, and divided into two inserts. With the pSEVA234 Gibson Assembly overhangs (BBa_25I4ZGKJ, BBa_2506U5S6), in addition to the insert overhang i (BBa_2534D5XF), the Biosensor was assembled into pSEVA234 and our Biosensor plasmid (BBa_25EHA43F) was built.
From the preliminary test, we detected the red mScarlet fluorescence, when the P. putida with a full Biosensor circuit (PPfBS) was exposed to 50 µM cadmium solutions. E. coli with the same circuit (ECfBS) was transformed successfully and incubated with the same concentration of cadmium, but it did not glow. Furthermore, we measured the fluorescence of PPfBS incubated in 50 µM cadmium solutions overnight by FlexStation. You may check our results and interpretations via our Experiment page.
We suspect that the reason for not detecting fluorescence from ECfBS was due to growth conditions and circuit design, as both bacteria were incubated at 30°C and circuit codons are optimized for P. putida, which are not ideal for E. coli to show reproducible results. We plan to repeat the experiment with lower cadmium concentrations to test out the regulatory system of aTc and the sensitivity of the whole circuit.
We still could not detect the degradation of the fluorescence from the overnight measured data, which suggests the LVA degradation tag was not working well. A wired but characterized result was obtained and documented in our Experiment page, which indicates the fluorescence intensity decreases then increases with cadmium concentration rising. We also inferred that a complicated construct might have burdened the bacteria, which led us to design the circuit into a more simplified version.
In order to implement our genetic engineered machine into practical use, a mechanism to ensure safety and trigger cell death after leaking into the environment is vital. Upon this concept, we designed a Kill Switch to ensure biosafety for our chassis.
Besides building the biosensor, we also designed a kill switch to prevent the persistence of genetically modified bacteria in case of leakage. For details, please refer to our Design | Description page.
During our attendance of iGBA, a regional iGEM conference, a senior member of BNUZH-China Team, iGEM 2024, introduced to us an inducible promoter PopdH (BBa_K5291021) and its corresponding expression system needed to adjust the expression of proteins for our Kill Switch circuit. When targeting the kill system,, we selected lysis protein E (BBa_K2152003) originating from bacteriophage Phi X 174, to maximise lysis effectiveness and ensure no leakage.
In the beginning ,we weren’t entirely sure whether PodpH is compatible with our chassis P. putida due to limited literature, therefore we compared the structure of key proteins involved in the regulation of PodpH. Through protein structure predictions and protein alignment, we determined that PodpH is cross-compatible. See “Kill Switch Modelling” in our Model page.
Figure 10. Demonstration of our Kill Switch, preventing the persistence of the genetic engineered machine after leaking into natural environment.
Our Kill Switch circuit was built with four fragments, and divided into three inserts. With the pSEVA234 Gibson Assembly overhangs (BBa_25I4ZGKJ, BBa_2506U5S6), in addition to the insert overhang i (BBa_2534D5XF) and the unique site (BBa_25ZHOKVG), we could selectively adding the lysis gene in and assemble the whole device (BBa_25CTXIKQ) into pSEVA234 and construct our Kill Switch plasmid (BBa_258P32V0).
Initially, 1 mM IPTG and 2 mM citric acid were used to test the Kill Switch construct, as suggested by the literature. We plated the transformed bacteria on an agar plate with Kanamycin, which led to no growth of bacteria. Then, we switched the method of testing, making IPTG and citrate-containing agar plates instead of adding them to the liquid bacterial culture. This showed a promising result, where E. coli with the Kill Switch construct grew.
With the promising result, we will test out with different concentrations of IPTG and citric acid to see the expression levels and characterize the circuit. Our preliminary result shows that the Kill Switch works logically as expected, while still pending to implement a detailed characterization to determine if it aligns well with the safety needs of the practical application.
Due to time constraints within the initial project cycle, several key validation steps and advanced design implementations remain slated for future work. Our continued efforts will focus on fully characterizing our latest constructs and translating our engineered mechanism into a practical, real-world application pipeline.
The next phase of design focuses on rigorous characterization and modular improvement:
Driven by insights from our Human Practices efforts, the most critical next step is comprehensive characterization. This includes assessing the efficiency of our biosorbent when applied to real sludge samples and validating the proposed application within a functional pipeline, details of which are further elaborated on our Project Description page.
Following consultation with Prof. BAE and based on dry lab screening results, we designed an improved version of our cadmium surface binding circuit (BBa_25310AKD). This design replaces the previously used PflQ2MT with EC20, a phytochelatin we were initially unable to synthesize. You may like to take a look at our EC20 Circuit Design logbook below:
Subsequent to the successful implementation and characterization of EC20, we plan to further enhance the cadmium binding capacity and specificity by integrating a specifically designed peptide sequence (Li et al., 2025).
A core feature of our design is its modularity, allowing the basic concept to be extended to remediate other heavy metals:
In the current cycle, we successfully built the following constructs intended for future testing:
Figure 11. Demonstration of EC20 cadmium surface binding design, which shows promising capability for a more specific and effective binding.
Figure 12. Demonstration of simplified Biosensor design, showing similar reacting behavior, despite losing part of the real-time ability and quick response.
The next phase is dedicated to rigorous testing and moving toward real-world application:
While we successfully assembled the fragments for the EC20 circuit (BBa_25310AKD), we unfortunately did not have sufficient time to conduct the necessary expression experiments and binding characterization. Thorough testing of this improved construct is a priority.
We have a detailed plan to validate the proposed column pipeline through small-scale experiments and characterization studies. This will involve:
The ultimate goal is to implement our technology in a real-life setting. This will involve utilizing real sludge samples to characterize the system's efficiency under complex environmental conditions. We aim to collaborate with end-users, such as the local facility T-Park, to conduct large-scale trials. This collaboration will be crucial for fine-tuning our pipeline and processing protocol to specifically target the composition of local sludge.