Some genetically modified microorganisms used in the production of engineered probiotics or industrial fermentation strains require special precautions for biosafety¹. It is important to prevent the unintentional release, multiplication, and spread of these genetically modified microorganisms into the environment, which could lead to unpredictable biological contamination². This project has designed a simple and user-friendly "safety lock" for engineered microorganisms. Under normal conditions at 37℃ (the temperature inside the human body, which is also the working temperature for probiotics and commonly used in industrial microbial fermentation), the "safety lock" remains inactive, allowing the host microorganism to reproduce and function normally. However, at 22℃ (a temperature closer to natural environmental conditions, excluding tropical regions and extremely hot summers), the "safety lock" becomes active, expressing a toxic protein that leads to the self-destruction of the host microorganism, thereby preventing the release of the engineered microorganisms^3-5. We constructed a temperature control system（pTRIP）, and introduced a visual reporting gene(EGFP), into the system to monitor its functionality and optimize its sensitivity. Lastly, we developed a plasmid, pTRIP-ccdB, which codes a toxic protein, (ccdB), that can kill bacteria upon expression, thereby preventing the leakage of engineered microorganisms(Figure 1).

 

.

Figure 1. The engineering design schematic diagram.

 

Results of Our Work

 

1.The construction of plasmid

1.1 pTRIP (pTrc99k-RIP)

We constructed pTRIP using homologous recombination. The RIP sequence was amplified by PCR, with a length of 2065 bp. The Figure 2A indicates the band consistent with the results. The pTrc99k plasmid was used as a template for PCR amplification, resulting in a fragment of 3408bp for pTrc. The Figure 2B shows a band consistent with the target size. It indicating successful amplification of RIP and linearization of the pTrc plasmid. After DpnI digestion and gel recovery, the pRIP fragment was obtained.

 

Figure 2. The gel electrophoresis validation of RIP and pTrc nucleic acids.

 

The plasmids were transformed into E.Coil DH5α . Figure3 A shows the presence of single colonies on the plate. We selected colonies 1-10 and sent them directly for sequencing. According to the results shown in Figure 3B and C, the RIP was successfully ligated to the pTrc vector without any apparent mutations, confirming the successful construction of the pTRIP plasmid.

 

 

Figure 3. The Monoclonal antibody validation and sequencing of pTRIP (DH5α) .

Note：

A. Transformation plate of pTRIP:

B.  Sequencing results of pTRIP

C.  Base comparison of a specific region within pTRIP

 

1.2 pTRIP-EGFP

We constructed the pTRIP-EGFP plasmid using homologous recombination. The EGFP sequence, which is 720bp in length, was amplified by PCR. The Figure 4 A showed that matched the expected size, indicating successful amplification of the EGFP sequence. Furthermore, using the pTRIP plasmid as a template, we performed PCR amplification and obtained 5473bp fragment of pTRIP-E. The Figure 4 B showed bands that matched the expected size, indicating successful amplification of the linearized pTRIP-E plasmid. We can conclude that we successfully amplified the EGFP sequence and linearized pTRIP-E plasmid.

 

Figure 4. The gel electrophoresis validation of EGFP and pTRIP-E .

 

 

We transformed the plasmid pTRIP-EGFP into DH5α, and Figure 5 A and B showed the growth of single clones. We selected clones 1-6 and performed antibody verification.The expected size of 720bp for the EGFP sequence. Additionally, Figure 5 C showed bands that matched the expected size, indicating successful transformation. Subsequently, we sent clones 1-6 for sequencing. The sequencing results in Figure 5 D showed a 100% match with the nucleotide sequence of EGFP, confirming the successful integration of the EGFP fragment into the pTRIP-E plasmid and further validating the successful construction of the plasmid.

 

 

 

Figure 5. The Monoclonal antibody validation and sequencing of pTRIP-EGFP (E.Coil DH5α) .

Note:

A.B shows the plate images of pTRIP-EGFP transformed into E.Coil DH5α cells.

C is the image showing the verification of single clone colonies.

D is the sequencing image of pTRIP-EGFP plasmid.

 

1.3 pTRIP-ccdB

We constructed the pTRIP-ccdB plasmid using homologous recombination. The PCR amplification of the ccdB sequence resulted in a fragment of 306bp in length. The figure 6 indicates that the amplified band matches the expected size, confirming the successful amplification of the ccdB sequence from the linearized plasmid.

 

Figure 6. The gel electrophoresis validation of ccdB.

 

By using the pTRIP plasmid as a template, we performed PCR amplification to obtain a 5473kb fragment referred to as pTRIP-C. The gel electrophoresis image in Figure 7 shows that the amplified band matches the expected size, indicating the successful amplification of the linearized pTRIP-C plasmid.

 

Figure 7. The gel electrophoresis validation of pTRIP-C.

 

The pTRIP-ccdB plasmid was transformed into E. coli DB3.1. The single clone colony growth on plates is shown in Figure 8 A and B. Clones 1-8 were selected for antibody verification, and the results in Figure 8 C demonstrate clear bands, confirming the presence of the ccdB sequence with a length of 306bp. The gel electrophoresis image in Figure 8C matches the expected band, indicating the successful transformation.

Next, colonies 1-8 were sent for sequencing, and the sequencing results in Figure 8 D showed a 100% match with the ccdB nucleotide sequence. This confirms the successful integration of the ccdB fragment into the pTRIP-E plasmid. It further validates the successful construction of the pTRIP-ccdB plasmid.

 

Figure 8. The monoclonal antibody validation and sequencing of pTRIP-ccdB (E.Coil DB3.1) .

 

2. Protein expression

2.1 pTRIP-EGFP

The target protein EGFP has a size of 26.9kDa. Protein expression was induced at a concentration of 0.6mmol AI (AutoInducers, herein it refers to N-(3-oxohexanoyl)-L-homoserine lactone), and induction was performed at 37 ^oC and 22 ^oC. According to the figure 9, in the control group (line 1), no EGFP protein is present. Under the condition of 37 ^oC (line 1 to line 3), EGFP protein was not observed. However, under the condition of 22 ^oC (line 4 to line 5),  there is a clear presence of EGFP protein. This indicates that EGFP expression is not express or occurring at a lower level at 37 ^oC, while there is substantial expression at 22 ^oC. Therefore, it can be concluded that our temperature control system is activated state at 22^oC and deactivated state at 37 ^oC.

 

 

Figure 9.The SDS-PAGE protein gel of EGFP at different temperatures

Note:

line 1: pTRIP(DH5α)

line 2：pTRIP-EGFP-37^oC(DH5α)

line 3：pTRIP-EGFP-37^oC(DH5α)

line 4：pTRIP-EGFP-22^oC(DH5α)

line 5：pTRIP-EGFP-22^oC(DH5α)

 

2.2 pTRIP-ccdB

The size of the ccdB protein, the target protein, is 11.7 kDa. Protein expression was induced at a concentration of 0.6mmol AI, and induction was performed at 37 ^oC and 22 ^oC. According to the figure 10, in the control group (line 1 -line 2 and line 9 -line 10), ccdB protein was not observed. Under the condition of 37 ^oC (line 3 to line 4), ccdB protein was not observed. However, under the condition of 22 ^oC (line 5 to line 8), a weak presence of ccdB protein is detected. This indicates that ccdB expression does not occur at 37 ^oC, while there is limited expression at 22 ^oC. Therefore, it can be concluded that our temperature control system is in the activated state at 22 ^oC and in the deactivated state at 37 ^oC.

 

Figure 10 : The SDS-PAGE of ccdB protein

Note:

1-2:  E.coil DB3.1(control)

3-4：37^oC-pTRIP-ccdB(E.coil DB3.1)

5-8：22^oC-pTRIP-ccdB(E.coil DB3.1)

9-10: E.coil DB3.1(control )

 

3. Functional Test

3.1 pTRIP-EGFP

3.1.1 Detection of fluorescence intensity EGFP

Protein expression was induced by AI at a concentration of 0.6 mmol and at different temperatures. According to Table 1 and Figure 11, the fluorescence intensity of pTRIP-EGFP increased first and then decreased with the increase of temperature. At 22 ^oC, the fluorescence intensity of pTRIP-EGFP was the strongest. At 37 ^oC, the fluorescence intensity of pTRIP-EGFP was the weakest. However, no fluorescence signal was detected in the control group pTRIP. This shows that the temperature control system of our subject is controlled by temperature. At 22 ^oC, in the “open” state ; At 37 ^oC, it is close to the “closed” state.

 

Table 1. The fluorescence intensity of reporter gene EGFP at different temperatures 

 

 

Figure 11. The fluorescence intensity of reporter gene EGFP at different temperatures

 

In order to observe the fluorescence signal of pTRIP-EGFP more clearly, we made the glass slides of pTRIP-EGFP bacteria and control group ( pTRIP ) cultured at 22 ^oC and 37 ^oC. According to the Figure12, A (colonies under white light) and B (pTRIP-EGFP at 37 ^oC) have no fluorescence signal, C and D (pTRIP-EGFP at 22 ^oC) have obvious fluorescence signal. It shows that there is no fluorescence signal at 37 ^oC, and obvious fluorescence signal can be seen at 22 ^oC.

 

Figure12.The fluorescence signal of report gene EGFP under microscope

Note:

A : pTRIP-EGFP colonies under white light

B. Fluorescence reporter gene of pTRIP-EGFP plasmid at 37 °C

C. 22 ° fluorescence reporter gene of pTRIP-EGFP plasmid ( 10 × 40 )

D. 22 ° fluorescence reporter gene of pTRIP-EGFP plasmid ( 10 × 100 )

 

3.2 pTRIP-ccdB

3.2.1 The growth ability of pTRIP-ccdB (DH5α)

A.The pTRIP-ccdB was transform into DH5α

In order to verify the sensitivity of our temperature control system, we transferred the plasmid pTRIP-ccdB into DH5α. The length of ccdB is 306 bp, and Figure 13A proves that the colony growth is successful. The line 1-9 in Figure 13B are consistent with the expected results, which proves that pTRIP-ccdB has been successfully transformed into DH5α.

 

Figure 13. Verification of the pTRIP-ccdB (E.coil) monoclonal antibody

 

B.The pTRIP-ccdB(E.coil DH5α) growth ability

A. The growth ability of pTRIP-ccdB (E.coil DH5α) at 22 °C

Firstly, it can be seen from these 5 graphs in Figure 14 that basically AI induction concentration at 0.2mmol to 1mmol has very little effect on the growth trend of the E. coli in the control group and experimental group.

According to Figure 14 A to E, given the temperature 22 °C, the OD600 of E.coil in the control group increased first and then gradually tended to be stabilized over time. While there was no significant increase in OD600 of pTRIP-ccdB at the experimental group over time. Therefore, compared with the wild DH5α (the control group), we can conclude that the ccdB did kill bacteria, the host at 22 °C upon expression.

 

Figure 14.The growth ability of pTRIP-ccdB ( E.coil DH5α ) in different AI concentrations at 22 °C

 

C. The growth ability of pTRIP-ccdB (E.coil DH5α) at 37 ° C

According to Figure15A, when the temperature was 37 °C and AI was 0.2 mmol, the OD600 of the control group DH5α and the experimental group pTRIP-ccdB increased first and then gradually stabilized over time. The control group was slightly higher than the experimental group after 20 minutes but following they went to a similar level after another 10 minutes.

According to Figure 15B and 15C, when the temperature was 37 °C and AI was 0.4 mmol and 0.6 mmol, respectively, these two groups have very similar values and trend upon time duration that the OD600 increased first and then gradually stabilized over time.

According to Figure15D, when the temperature was 37 °C and AI was 0.8 mmol, the OD600 of the control group DH5α and the experimental group pTRIP-ccdB increased first and then gradually stabilized over time, and the experimental group was slightly lower than the control group.

According to Figure15E, when the temperature was 37 °C and AI was 1 mmol, the OD600 of the control group DH5α and the experimental group pTRIP-ccdB increased first and then gradually stabilized over time. The control group was slightly higher than the experimental group after 10 minutes but following they went to a similar level after another 20 minutes.

 

In summary, the pTRIP-ccdB E.coil DH5α has a very similar growth trend with the control group, the wild E.coil DH5α at 37 °C which means the “switch” pTRIP-ccdB has little effect on the strain’s growth since no expression of ccdB at 37 °C.

 

Figure 15.The growth ability of pTRIP-ccdB ( E.coil DH5α ) in different AI concentrations at 37 ° C.

Note: in C, two curves coincide.

 

 

D. Comparison of the growth ability of pTRIP-ccdB (E.coil DH5α) at 37 °C and 22 °C

 

Since we can see that the control group has the same growth trend at 22 °C and 37 °C, we decided to compare the growth trend of pTRIP-ccdB (E.coil DH5α) at 22 °C and 37 °C, respectively, to evaluate the engineering of ccdB.

According to Figure 16, in two groups with AI concentration of 0.6 mmol, the OD600 of pTRIP-ccdB at 37 °C increased firstly and then tended to be stabilized over time, while there was almost no significant change of that at 22 °C. It is seen that pTRIP-ccdB (E.coil DH5α) grew much better at 37 °C than 22 °C where it indicated little growth over time. This further supports the conclusion that the bacterial strain grows normally at 37°C, while the presence of ccdB at 22°C leads to bacterial cell death.

 

 

Figure16. Comparison of OD600 at 37°C and 22°C pTRIP-ccdB (E.coil DH5α ) with AI concentration of 0.6 at different times

 

 

3.2.2 The growth ability of pTRIP-ccdB in E.coil BL21 (DE3).

A. The pTRIP-ccdB was transform into E.coil BL21(DE3).

In order to verify the sensitivity of our temperature control system, we transferred the plasmid pTRIP-ccdB into E.coil BL21 (DE3). The length of ccdB is 306 bp, and Figure 17A proves that the colony growth is successful. The line 1-5 in Figure 17 B are consistent with the expected results, which proves that pTRIP-ccdB has been successfully transformed into E.coil BL21 (DE3).

 

 

Figure 17. Verification of the pTRIP-ccdB (E.coil BL21 (DE3)) monoclonal antibody

 

B.The Growth ability of pTRIP-ccdB in E.coil BL21 ( DE3 )

Comparisons of growth capabilities were made at 37°C and 22°C with an AI concentration of 0.6 mmol.The E.coil BL21 was the control group. According to Figure 18, it is evident that the OD600 of BL21 (the control group) is significantly higher than that of BL21(pTRIP-ccdB) at 22°C while the OD600 of BL21 and BL21(pTRIP-ccdB) have similar value at 37°C. This test result is consistent with that for DH5α as discussed previously and this also back up the engineering success of our temperature-based “kill switch”, the plasmid pTRIP-ccdB in E. coli host.

 

 

Figure 18. The Growth ability of pTRIP-ccdB in E.coil BL21comparing to the control group BL21

 

Future plans

1. In the experiment, we only conducted sensitivity testing of the temperature-controlled switch in E. coli BL21(DE3) and E. coli DH5α. In the future, we need to transform it into more bacterial strains to perform sensitivity tests and further optimize the sensitivity of our temperature-controlled switch based on the test results. We also need to explore the temperature range of 22-37°C for the temperature-controlled switch and validate its states through additional verification methods.

2. Our temperature-controlled switch requires the addition of exogenous inducers (N-(3-oxohexanoyl) -L-homoserine lactone) for better expression. When no inducer is added, the expression level is relatively low, and this limitation needs to be addressed in practical applications.

 

[1] Oscar Z,B. I C,E. Y K, et al. Traditional fermented foods of Benin Republic: microbiological safety and health benefits[J]. Discover Food,2023,3(1).

[2] Obeng A P,Awere E,Obeng A P, et al. Usage and Microbial Safety of Shared and Unshared Excreta Disposal Facilities in Developing Countries: The Case of a Ghanaian Rural District[J]. Sustainability,2023,15(13).

[3] Bazhenov, S.V., Scheglova, E.S., Utkina, A.A. et al. New temperature-switchable acyl homoserine lactone-regulated expression vector. Appl Microbiol Biotechnol 107, 807–818 (2023). https://doi.org/10.1007/s00253-022-12341-y

[4] Nocadello, S., Swennen, E.F. The new pLAI (lux regulon based auto-inducible) expression system for recombinant protein production in Escherichia coli. Microb Cell Fact 11, 3 (2012). https://doi.org/10.1186/1475-2859-11-3

[5] Hoffmann SA, Diggans J, Densmore D, Dai J, Knight T, Leproust E, Boeke JD, Wheeler N, Cai Y. Safety by design: Biosafety and biosecurity in the age of synthetic genomics. iScience. 2023 Feb 10;26(3):106165. doi: 10.1016/j.isci.2023.106165.