Results


Overview

LuxR and pLux are well-established transcriptional regulatory proteins and their corresponding promoters. In bacteria, once LuxR binds with AHL, the resulting complex associates with the pLux promoter region, thereby initiating the expression of downstream genes. The concentration of AHL is related to the bacterial population density; when it exceeds a certain threshold, the LuxR-AHL complex activates gene expression. pLux is the promoter regulated by LuxR protein, responsible for controlling the expression of downstream genes. When LuxR binds with AHL, the LuxR-AHL complex binds to the pLux promoter, initiating transcription of downstream genes. The pLux promoter regulates the expression of the luxICDABE gene cluster in natural bacterial systems, which is responsible for the production of luminescent proteins.

This year, we aim to utilize this mature regulatory system to detect spoilage in milk products. As milk spoils, the bacterial concentration within increases. As bacterial concentration increases, the detection of AHL also increases, allowing bacteria to gauge the presence of other bacteria nearby based on AHL concentration. Since AHL is a lipophilic molecule, it can freely diffuse between bacteria, indicating a quantitative relationship between the number of bacteria in space and the detectable AHL quantity. Therefore, we can detect the corresponding bacterial quantity by measuring AHL.

Based on this design, we conducted system validation and regulation.


Part I: Fluorescence Signal Levels Under Varying Concentrations of AHL

Three monoclonal colonies from the engineered bacterial agar plates were picked and cultured overnight. On the following day, the overnight culture was diluted to the logarithmic phase, and AHL was added to the culture system to achieve different final concentrations. The bacteria were then cultivated overnight in a TECAN plate reader, where their growth and fluorescence expression were monitored. It was observed that as the concentration of AHL increased, the fluorescence signal level of the system also increased, aligning with our expectations for the system's response. Additionally, the time for the fluorescence level to reach a plateau and stabilize increased with higher concentrations of AHL. This may indicate that at higher AHL concentrations, the system requires more time to mobilize transcriptional and translational resources to produce more fluorescent proteins and ultimately reach a steady state (Fig. 1). We fitted the response curve, and the R² value was 0.9315, indicating a high degree of fit between the experimental data and the model used. The model effectively describes the relationship between AHL concentration and the response of the biosensor. We found that the system reaches half of its maximum response to AHL at a concentration of 14.22 mM (Fig. 2), and it has a broad detection range, capable of detecting AHL at both low and high concentrations.

Fig. 1 Time-course response curves of strains with the LuxR and pLux regulatory system to varying concentrations of AHL (3O-C6-HSL).
Fig. 2 Strains with LuxR and pLux regulatory system response curves to varying concentrations of AHL (3O-C6-HSL).

Part II: Real Milk Sample Testing

To further validate the system, we tested expired milk samples from two sources: Yue Xian Huo(milk1) and Wens Dairy(milk2). The samples were kept at room temperature for 7 days. For comparison, we employed three methods: (1) colony counting using a 10-fold serial dilution on LB plates without antibiotics, (2) our synthetic biology detection system to measure 3O-C6-HSL concentration based on the standard response curve, and (3) a traditional method involving smell and visual observation to assess milk spoilage.

Fig. 3 Two expierd milk, the left one is yue xian huo and right one is wens dairy

We provided sensory observations from three team members (table1), recorded on days 5 to 7 after purchase (3 to 5 days post-expiration). For safety reasons, we did not taste spoiled milk, but it was interesting to note differences in the physical properties of spoiled milk from different sources.

Tab. 1 The sensory observation of 2 expired milk during 3 to 5 days post-expiration

The colony counting results for the expired milk samples showed significant differences between the two sources. For Wens Dairy (Milk2), the 10⁶ diluted plate yielded 15 colonies, while the 10⁵ diluted plate resulted in 124 colonies. In contrast, Yue Xian Huo (Milk1) showed only 5 colonies on the 10⁵ diluted plate and 20 colonies on the 10⁴ diluted plate. These results indicate a much higher bacterial load in the Wens Dairy sample compared to the Yue Xian Huo sample, suggesting that the milk from Wens Dairy underwent more rapid spoilage or had a higher initial contamination level. This also partly explains why the sensory observations regarding the smell differed between the samples. The higher bacterial load in the Wens Dairy sample likely contributed to more pronounced spoilage, leading to stronger odor changes, whereas the lower colony count in the Yue Xian Huo sample might correspond to less severe spoilage, resulting in subtler changes in smell. These results highlight the correlation between bacterial contamination levels and sensory indicators of spoilage, further emphasizing the importance of microbial load in determining milk freshness and spoilage characteristics.

Fig. 4 Plate of milk2 on 5 days post-expiration

These observations led to a discussion with the quality control manager from Yantang Dairy, one of Guangzhou's largest and earliest fresh milk suppliers, established in 1956. She suggested that the differences in spoilage might be attributed to several factors, including contamination from the cows themselves, environmental conditions in the packaging facility, and variations in transport conditions. Further testing for specific contaminants or monitoring environmental factors during transportation and storage could provide more clarity on these variations(more details in HP). The colony count data also aligns with the sensory observations made by the team, supporting the idea that different contamination levels and spoilage rates might explain the sensory differences noted between milk from different sources.

We also tested the previously designed system in 5-fold diluted boiled milk supernatant samples and observed significant induction of fluorescence expression after adding the milk (Figure 3a,b). Based on the calibrated response curve, the AHL concentration in the original expired milk samples was estimated to be between 5-10 nM. Our circuit shows significant differences compared to control group even at 5 hr after coculture(Figure 3b), suggesting that our reporting system can detect the concentration of AHLs not only in standard solutions but also in complex matrices.

Fig. 5 The milk supernatant after heating at 100°C for 15 minutes and centrifuging, the sample on the left is labeled "milk1" and the one on the right is labeled "milk2."
Fig. 6 Response Curves and Quantitative Analysis of Strains with LuxR and pLux Regulatory System to Milk Samples. a. Time-course Response Curves of Strains with LuxR and pLux Regulatory System to Two Spoiled Milk Samples and Control; b. Quantitative Analysis Comparing the Significant Differences in Response between Strains with LuxR and pLux Regulatory System to Spoiled Milk and Control.

Part III: High-throughput Screening of Promoters

Utilizing the Golden Gate Assembly technique, a variety of promoters were inserted into the system for testing to evaluate their regulatory performance under diverse conditions. The iGEM promoter library, featuring Golden Gate standard cloning sites, was procured from Xinzhu Bio. The Golden Gate products were transformed into DH5α competent cells via chemical transformation, and the transformed bacteria were then spread onto agar plates containing the appropriate antibiotics to accomplish high-throughput library construction.

The transformed strains were cultured on plates to yield multiple monoclonal colonies, which were subsequently selected for overnight cultivation in liquid medium. The fluorescence signals of 16 distinct promoters were tested under various concentrations of AHL, employing a uniform methodology. It was observed that 15 of the 16 promoters exhibited a response to AHL, with the level of response escalating in tandem with increasing AHL concentrations. One promoter, however, demonstrated no significant response to AHL solution.

Fig. 7 Maximum Response Signals of LuxR-pLux Genetic Circuits with Different Promoters at Various AHL Concentrations.

Subsequently, we compared the response levels of strains containing these 16 distinct promoters to 0 mM and 50 mM AHL solutions to further quantify the presence or absence of a response, that is, the difference in induction strength of the responsive promoters. The results indicated that at a concentration of 50 mM, with the exception of one promoter that showed no significant response, the induction strengths corresponding to the other promoters varied. However, the system response levels under their regulation were markedly higher than those of the corresponding 0 mM control groups.

Fig. 8 Sequence-Response Relationship Analysis

We exposed the strains containing these 16 distinct promoters to spoiled milk, and the results similarly demonstrated that within the spoiled milk environment, except for one promoter that showed no significant response, the induction strengths corresponding to the other promoters varied. However, the system response levels under their regulation were all markedly higher than those of the corresponding control groups without spoiled milk.

Fig. 9 Significant differences in the responses of the LuxR-pLux gene circuit with various promoters to spoiled milk versus the control group.

After sequencing, we obtained the promoter sequences of these experimental groups and explored the relationship between different sequences and response multiples. Further, we plan to use machine learning methods for modeling and prediction to facilitate subsequent optimization(For more detailed information, please refer to our model page).