. On this page, we describe the most important results that portray our journey while designing our project.

Quorum Sensingis a way for cells to regulate downstream gene expression based on their own density. The concentration of the signaling molecule - AHL - secreted by the cell increases as the cell density increases. When the concentration of AHL reaches a certain level, it can bind to the corresponding binding protein and alter the expression of downstream genes.

The Esa I/R system is quite special from traditional QS system. The EsaI/R QS system is homologous to the LuxI/R QS system and originates the maize pathogen--Pantoea stewartii subsp. stewartia. EsaR can act as both transcriptional activator and repressor.

Our project builds on what has been done before. In 2020, Fei Gu et. al succeeded in redirecting the metabolic flowof E. coli using the Esa I/R system[1]. And they have applied the QS switch in the production of PHB. The Stains that they used come from another study[2]. Our project goes one step further. We have re-characterized the system and added an auto-lysis system.This constitutes a three-layer dynamic regulation model that enables the separation of the bacterial growth phase, the production phase and the product-release phase. L19 and L31 were engineered E. coli MG1655, which already have Esa R/I system in them. They have different regulatory thresholds (Fig.2), and we chose L19 and L31 as our strains.

We transformed pCL-PesaS-GFP (LVA) into E. coli L19 and L31 in combination with pCL-PesaRwt-mkate, pCL-PesaRc-mkate, and pCL-PesaRp-mkate, respectively. We repeatedly characterized the Esa I/R QS system by detecting fluorescence intensity to reflect promoter expression intensity. Green fluorescence was used to reflect the transcriptional expression intensity of PesaS, and red fluorescence represented the transcriptional expression intensity of PesaR. In this way, the most suitable models for growth and production stages were screened. Here are the strains we constructed.

Through preliminary characterization, we can clearly see that PesaS and PesaR have a distinct sequential expression time sequence at different stages. The PesaS transcript expression peaked in the first 8 hours, especially in the 4th-6th hours, while PesaR peaked and stabilized after 10-12 hours (except for PesaR-p).

We hoped that by comparing the before and after differences in the expression of the two, we would be able to enter into the growth mode during the propagation of the strains faster, and the boundaries between the production mode and the growth mode would be clearer, with a quicker and more complete transition, and the results performed very well for both.

If artificial induction is used to make the bacteria lysed, the inducer is not only expensive, but also produces strong toxicity to the cells. If a general constitutive promoter is used to express the cleavage gene, it will cause the bacterial metabolic burden and great growth pressure.

So, we planned to use stationary phase promoter to control the lysis gene. Bacteria undergo significant changes in their physiological state after entering the plateau phase. The stationary phase promoter endogenously produces more σ factors required for the stationary phase when bacteria enter the exponential phase, and it can direct bacteriophage RNA polymerase to transcribe specific genes during the stationary phase. Because the signal required for the stationary phase promoter is produced endogenously, it is not toxic to the bacteria. It is also dynamically regulated and does not cause metabolic stress or growth problems.

We combined PACYC-Pflic-GFP, PACYC-P1.1-GFP, PACYC-P2.1-GFP, and PACYC-P3.1-GFP plasmids (these plasmids were given by Xin Jin, our advisor) with pCL-PesaRwt, pCL-PesaRc, and pCL-PesaRp plasmids (These plasmids were given by Fei Gu, our advisor), respectively, and electrotransformed them into L19 and L31. In total, we constructed 24 strains (Table 2)

We then used a Multi-Detection Microplate Reader (Synergy HT, Biotek, U.S.) to detect the red (excitation at 585 nm and emission at 640 nm) fluorescence and green (excitation at 485 nm and emission at 528 nm) fluorescence.

ITEM	MEANING	CONNT
L19	pCL-PesaRwt-mkate	PACYC-Pfic-GFP
		PACYC-P1.1-GFP
		PACYC-P2.1-GFP
		PACYC-P3.1-GFP
	pCL-PesaRp-mkate	PACYC-Pfic-GFP
		PACYC-P1.1-GFP
		PACYC-P2.1-GFP
		PACYC-P3.1-GFP
	pCL-PesaRc-mkate	PACYC-Pfic-GFP
		PACYC-P1.1-GFP
		PACYC-P2.1-GFP
		PACYC-P3.1-GFP
L31	pCL-PesaRwt-mkate	PACYC-Pfic-GFP
		PACYC-P1.1-GFP
		PACYC-P2.1-GFP
		PACYC-P3.1-GFP
	pCL-PesaRpt-mkate	PACYC-Pfic-GFP
		PACYC-P1.1-GFP
		PACYC-P2.1-GFP
		PACYC-P3.1-GFP
	pCL-PesaRc-mkate	PACYC-Pfic-GFP
		PACYC-P1.1-GFP
		PACYC-P2.1-GFP
		PACYC-P3.1-GFP

The expression time was found to be close to that of PesaR, with a maximum difference of only 2-4 hours to peak (Fig.4). Also, expression of this promoter is very rapid, with peak expression in less than 4 hours (Fig5, Fig.6). This means that by constructing an auto-lysis system with such a stationary phase promoter, our engineered bacteria will only have a few hours to produce PHB, not taking into account the metabolic disruption and reduced yield caused by the timing being too close to meet our expectations. : (

We also found that Pfic, P1.1, P2.1 and P3.1 have different levels of expression (Fig5, Fig.6)[3]. This difference is particularly evident in L19 (Fig5, Fig.6).

So, by reading the scientific articles, we selected the late stationary phase promoters of PYU3, PYU7, PYU6 and PYU92 for the second screening[4]. We combined them with PesaR and characterized them. In this batch we also constructed 24 bacteria (table 3).

Strain	Plasmid 1	Plasmid 2
L19	pCL-PesaRwt-mkate	PACYC-PYU3-GFP
		PACYC-PYU7-GFP
		PACYC-PYU16-GFP
		PACYC-PYU92-GFP
	pCL-PesaRp-mkate	PACYC-PYU3-GFP
		PACYC-PYU7-GFP
		PACYC-PYU16-GFP
		PACYC-PYU92-GFP
	pCL-PesaRc-mkate	PACYC-PYU3-GFP
		PACYC-PYU7-GFP
		PACYC-PYU16-GFP
		PACYC-PYU92-GFP
L31	pCL-PesaRwt-mkate	PACYC-PYU3-GFP
		PACYC-PYU7-GFP
		PACYC-PYU16-GFP
		PACYC-PYU92-GFP
	pCL-PesaRp-mkate	PACYC-PYU3-GFP
		PACYC-PYU7-GFP
		PACYC-PYU16-GFP
		PACYC-PYU92-GFP
	pCL-PesaRc-mkate	PACYC-PYU3-GFP
		PACYC-PYU7-GFP
		PACYC-PYU16-GFP
		PACYC-PYU92-GFP

From the 24 characterization results, the above ten combinations with significant expression time differences were selected and prepared for further screening (Fig. 7).

In several of these combinations, we found, for example, L19-PesaRwt-PYU7, that the two promoters peaked in their respective expressions at times that differed by more than 10 hoursThe results show that the production and cleavage phases can be temporally uncoupledTherefore, the characterization results have met our requirement - the production phase is separated from the product-release phase. : )

In the end, the late steady state promoters PYU3, PYU7, PYU16 and PYU92 were selected as our ideal promoters. We also selected ten combinations that met our expectations to further build the three-layer dynamic regulation system.

We can see from this that the stationary phase begins to be expressed when the production phase is at its peak and the separation of the production phase from the desired lysis phase has been achieved. This means that the growth phase of the bacteria is barely affected and the bacteria have more than enough time for the production phase. It was only after 40 hours that the bacteria entered the expected lysis peak and lysed in droves.

We built three different modules,the growth module ,the production module , and the product release module, which form our Three-layer dynamic regulation model. we expressed them in the same bacterium, and the results were very good.

We cloned the backbone part on the PACYC plasmid except the GFP fragment by PCR using the Gibson method, leaving a short homologous sequence with the lysis gene. At the same time, we obtained the SRRz gene by company synthesis and cloned the SRRz gene fragment containing the homologous arm of PACYC; the E gene fragment containing the homologous arm of PACYC was cloned by PCR on Phi X174 plasmid, and the PACYC backbone fragment and the cleaved gene fragment were seamlessly joined under the action of recombinase. PACYC-PYU3, 7, 16, 92-SRRz and PACYC-PYU3, 7, 16, 92-E plasmids were constructed (Fig. 12). All gene recombinations in the experiment were performed by the Gibson method.

We transformed the constructed plasmids into L19 and L31 to observe its lysis effect. The left side of the figure shows L19 containing the cleavage plasmid, and the right side does not contain the cleavage plasmid. After 1h of resting, it can be seen that the left side has been cleaved and clarified, and the right side is turbid, and the bacterium is sinking (Fig. 13).

Due to incomplete lysis in L19 and L31, we utilized the Gibson self-assembly method and designed primers to add RBS (BBa_B0031, BBa_B0032, BBa_B0033, and BBa_B0034) of different intensities to the promoter regions of PYU3, PYU7, PYU16, PYU92 with intensities of 0.010. 0.07,0.3, and 1 (Fig. 14). Due to time and effort constraints, we did not succeed in constructing all plasmids (Fig. 15).

Due to primer design problems, PCR extension length setting, or assembly errors, etc., the RBS substitution was not completely done successfully, so the plasmids verified to be working successfully in the following table were used for the next experiments.

Next, we transformed them to L19,L31 for characterization:

PYU16-B0034-SRRz and PYU92-B0034-SRRz were finally selected as the ideal lysis systems (Fig. 16). They have the advantages of higher efficient lysis capacity and low impact on strain growth.

Nile red dye powder, prepared as a solution using dimethyl sulfoxide, was added to the medium at a working concentration of 0.5 μg/ml and plates were poured. We set up two groups: the first group used Nile Red plates, with PHB-producing strains and non-PHB-producing strains delineated on the plates; the second group used PHB-producing strains, delineated on Nile Red and non-Nile Red plates. Observations were made using a UV imager after 48 hours of incubation.

From the figure, we can see that the PHB-producing strains can be observed to fluoresce while the non-PHB-producing strains cannot fluoresce under UV irradiation. And Nile Red works very effectively because there is no fluorescence on the non-Nile Red plate. This is proof that our PHBcab gene is effective.

Single colonies were transferred to 5mL of Luria−Bertani (LB) Broth with appropriate antibiotics. And they were culturing at 220rpm/ min and 37℃for 8h. 1% seeds were inoculated in M9 medium supplemented with 22g/L glucose.

Material	Volume/Mass
ddH2O	1L
Yeast extraction	2g
NaCl	1g
NH4Cl	1g
Na2HPO4·12H2O	15.6g
KH2PO4	3g

Samples were taken at 3,7,11,14,24,30,38,48h for OD600 and glucose concentrations, and PHB concentrations were measured at the same time from the 11th hour onwards (Fig. 15).

• Glucose

After diluting the samples 50-fold (Fig. 16), the glucose concentration was measured using a glucose tester and glucose was replenished to 20 g/L at 14h and 25.5h of fermentation. Using our hardware can easily complete this work. The raw data is in the Notebook.

• OD600

Using spectrophotometer to measure OD600(Fig. 17). The raw data is in the notebook.