We have consistently adhered to the "Design-Build-Test-Learn" cyclic paradigm to advance this project. This year's project can be broadly divided into three cyclic phases: in the first cycle, we designed and constructed a biosensor responsive to salidroside; in the second cycle, we established the salidroside synthesis pathway in the Escherichia coli chassis strain; and finally, in the third cycle, we optimized the salidroside-producing performance of the engineered strain by overexpressing the endogenous genes.

This project developed a synthetic biosensor with the transcription factor mutant HucR^SD to detect salidroside. It has three DNA fragments working together for sensitive, visual detection.
HucR^SD (four mutations) in the sensing module specifically recognizes salidroside: binding HucO to block transcription without substrate, forming a complex to activate expression when substrate is present.
The reporting module links GFP (fluorescent quantification) and LacZ (enzyme checks). Controlled by strong promoter P23119 with RBS and terminators, it’s efficient and sensitive.


We constructed all target plasmids by Gibson Assembly, attaching Lacz and HucR on Vector. HucO is connected with HucR, acting as a repressor until it combines with salidroside and falls off.

Results for pRB1k-pHuc-eGFP-LacZ:

Measurements using a microplate reader revealed a linear relationship between salidroside concentration and biosensor absorbance. Absorbance increased proportionally with rising salidroside concentration, confirming the accuracy of our results. A sixfold difference in absorbance was observed between samples containing 0.0 g/L and 2.0 g/L salidroside. Testing with salidroside standards demonstrated that our plasmid specifically responds to salidroside.


Subsequently, the salidroside produced by the salidroside-synthesizing bacterial culture was divided into two batches. One batch was subjected to HPLC analysis, while the other was added to the biosensor strain for fermentation, followed by absorbance measurement using a microplate reader.
In conclusion,we've successfully built a system that can detect the presence and concentration of salidroside. With the mutant HucR^SD, we've achieved a regulatory response to salidroside. When salidroside is added, the fluorescent proteins GFP and LacZ are turned on, and they give quantifiable results, showing different shades of yellow to indicate the concentration differences.

Through an extensive literature review, we found that Escherichia coli is incapable of autonomously synthesizing salidroside. Therefore, we set out to construct its de novo synthesis pathway, with the biosynthesis of tyrosol as the core intermediate step. First, two heterologous genes, Kivd from Lactococcus lactis and PAR from Rosa damascena, were introduced to form an enzyme module, enabling the efficient biosynthesis of tyrosol using glucose. Salidroside is produced through the glycosylation of tyrosol catalyzed by UGTs (glycosyltransferases), and this reaction determines the yield and regioselectivity of the product. Hence, screening for highly efficient UGTs is crucial for enhancing the catalytic efficiency of the conversion from tyrosol to salidroside, thereby increasing the yield of salidroside.


We amplified the target genes KivD and PAR by PCR; digested the vector pY97a, and constructed the recombinant plasmid pY97a-P23107-KivD-PAR using Gibson assembly technology.

Results for pY97a-P23107-KivD-PAR:

We found 33 different UGT-protein series from different species on the NCBI website, and we analysed their similarity, we drew a phylogenetic tree, then we choosed UGT85A1 and UGT85A1^A21G which both from Arabidopsis thalinana, UGT72B14 which from Rhodiola sachalinensis, and RrUGT33 from Rhodiola rosea L. We let the company synthesize the pY97a-P23119-UGTs-P23107-KivD-PAR plasmids, each containing one of the four genes.

The successfully constructed recombinant plasmids were transformed into competent BW25113 cells for induced expression. After LB activation and ZY-induced expression for 16 hours, bacterial cultures with an OD₆₀₀ of 3 were collected in centrifuge tubes and centrifuged. Then, 200 µL of M9 fermentation medium was added for fermentation, and the cultures were incubated in a shaker for 16 hours. The cultures were centrifuged again, and the supernatant was collected using a 1 mL syringe fitted with a filter membrane. The filtered supernatant was transferred to liquid-phase vials and sent to a company for testing.
After we received the test results from the HPLC (high performance liquid chromatography), we constructed a graph to present the data. The graph is plotted in a way where the salidroside productivity(mg/L) is aligned on the Y-axis against the type of the UGT sample, which is shown on the X-axis.
The four types of UGT samples were each tested for three trials in order to guarantee the accuracy of the resulting data. The error and dispersion of the data is indicated on the graph by the error bar and dots located on top of the columns.

By inserting the exognous gene KivD, PAR,and UGT to the E.coli,we find a effective way to produce biosynthesis salidroside. Indeed, the test results demonstrate that all the exognous genes we select have contribute to the synthesis of salidroside. What's more, we discovered that UGT72B14 has the highest productivity of biosynthesis salidroside. Average amount of UGT85A1 was 50.95mg/L, and average amount of UGT72B14 was 72.96mg/L, which means that the productivity of UGT72B14 increased 43% more than productivity of UGT85A1. In conclusion, we decide to use UGT72B14 as our enzyme to catalyze the UDP-Glucose and the tyrosol produce salidroside.

Our goal is to enhance the biosynthetic pathways of tyrosol and uridine diphosphate glucose (UDPG), both of which are precursors required for salidroside production. To improve the efficiency of microbial salidroside production by optimizing the biosynthesis of these two primary precursors, we overexpress the endogenous Escherichia coli genes GalU and Pgm to increase intracellular UDPG levels. Meanwhile, to boost tyrosol production, we introduce feedback-resistant mutant genes AroG^fbr and TyrA^fbr. These mutants can maintain metabolic flux under high-load production conditions and avoid inhibition by downstream metabolites.

We amplified the target genes GalU, Pgm, AroG^fbr, and TyrA^fbr by PCR; digested the vectors pLB1s and pSB1c; and constructed the recombinant plasmids pLB1s-GalU-Pgm and pSB1c-AroG^fbr-TyrA^fbr using Gibson assembly technology.

Results for pLB1s-GalU-Pgm:

Results for pSB1c-AroG^fbr-TyrA^fbr:

The successfully constructed recombinant plasmids were transformed into competent BW25113 cells for induced expression. After LB activation and ZY-induced expression for 16 hours, bacterial cultures with an OD₆₀₀ of 3 were collected in centrifuge tubes and centrifuged. Then, 200 µL of M9 fermentation medium was added for fermentation, and the cultures were incubated in a shaker for 16 hours. The cultures were centrifuged again, and the supernatant was collected using a 1 mL syringe fitted with a filter membrane. The filtered supernatant was transferred to liquid-phase vials and sent to a company for testing.
In order to evaluate the effect of our biosynthesis pathway enhancments, we submitted four different samples for testing the salidroside yield (mg/L) using HPLC technique, including:
●No enhancement, use as control data (UGT72B14)
●Only the shikimate biosynthesis pathway enhancement plasmid (UGT72B14-AroG^fbr-TyrA^fbr)
●Only the UGP-glucose biosynthesis pathway enhancement plasmid (UGT72B14-GalU-Pgm)
●Both enhancement plasmids (UGT72B14-AroG^fbr-TyrA^fbr-GalU-Pgm).


We performed a side-by-side comparison of the detection results from the biosensor fermentation and the HPLC analysis. The consistent trend between the two methods confirms that the trend we detected using the biosensor, although approximate, is accurate.

According to HPLC results, comparing our results to the control data bar, we are able to visualize that the enhancement of shikimate and UDP-glucose biosynthesis pathway lead to the increase in yield of salidroside. When both shikimate and UDP-glucose biosynthesis pathway are enhanced, yield of salidroside has multiplied about 2.86 times, achieving 208.32mg/L.
Following a rigorous "Design-Build-Test-Learn" methodology, our project successfully engineered an E. coli strain for the efficient biosynthesis of salidroside. Key engineering successes include the construction of a novel, low-cost biosensor, and the establishment of a functional de novo synthesis pathway. Through systematic optimization, we achieved a final salidroside yield of 208.32 mg/L. This work not only provides a viable platform for sustainable salidroside production but also the utility of our custom-built biosensor for accelerating future metabolic engineering projects.