Team:BHSF ND/Project/Bistable System
2019 IGEM BHSF_NDBistable System
Background
We are inspired by a lately published article written by Belén Calles, Angel Goñi-Moreno, and Víctor de Lorenzo to design a double-bistable system which achieves zero basal expression of gene of interest in the absence of any inducer while ensuring constant transcriptional capacities and induction rates among promoters.
“The mechanism of a bistable system is elaborated below: a repressor transcriptionally inhibits the promoter which is responsible for the transcription of sRNA, while the sRNA translationally inhibits the repressor. The gene of interest is translationally coupled to the repressor gene. That is to say, the repressor protein R targets the strong promoter where the inhibitory sRNA is produced, so that R and sRNA are mutually inhibitory. Such arrangement helps to minimize the basal activity of the inducible module and thus suppress leaky gene expression levels.”(1)
Design
In our system(Fig. 1), we have transcriptional repressor R1 expressed through the inducible promoter but translationally inhibited by MicC sRNA1 since the RBS of R1 is directly inhibited by sRNA while they initiates the transcription of the two repressors; the gene of interest is translationally coupled to the repressor gene while three promoters which produces the inhibitory sRNA are targeted by their corresponding repressors. Therefore, repressors and sRNAs are mutually inhibitory.
RBS1
The control of the recombinase in coordination with the upstream repressor gene, which is the target of the sRNA, requires a translational coupler cassette. “An efficient mechanism to reach this goal is based on the ability of translating ribosomes to unfold mRNA secondary structures. In our design, translational coupling is achieved by occluding the RBS of gene of interest by formation of a secondary mRNA structure, containing a His-tag sequence added to the 3 ́end of the repressor gene. The hairpin structure prevents the ribosomal recruitment to the recombinase and therefore its translation. In contrast, when the upstream repressor mRNA is actively translated, the 70s ribosome disrupts inhibitory mRNA secondary structure in the downstream gene translation initiation region thus allowing its expression.”【2】
However, during the experiment we tested that the RBS1 sequence provided by the original article fails to function in our circuit, we found RBS2 with a different working mechanism. The structure of AUGA in RBS2 acts as both the stop codon (AUG) of the former gene sequence, releasing the former peptide chain; takes a step (gene code) backward, pinpointing a new ribosome binding site, and initiated the translation of next gene sequence with the start codon (UGA). However, finally we still used RBS1 since the crux actually lied in another place.
SRNA
Small non‐coding RNAs (sRNAs) have important functions as genetic regulators in prokaryotes. “sRNAs act post‐transcriptionally through complementary pairing with target mRNAs to regulate protein expression. Antisense sRNAs negatively regulates proteins by destabilizing the target protein's mRNA. Antisense sRNAs prevent translation by binding to the target mRNAs in a process mediated by the RNA chaperone Hfq.”【3】 On binding, both the mRNAs and sRNAs are degraded, suggesting that prokaryotic sRNAs—unlike their eukaryotic counterparts—act stoichiometrically on their targets.
Repressor
“A DNA-binding repressor blocks the attachment of RNA polymerase to the promoter, thus preventing transcription of the genes into messenger RNA.”【4】
Double-layered bistable
We assume that when the whole system is applied to a factory, the presence of inducer is in the working condition of the bacterium. Within one single layer of the bistable system, the transcriptional factor is turned on, Pm initiates the expression of R1, which raises the concentration of R1 and thereby exerts greater inhibition on PR1 , therefore decreasing the concentration of sRNA1. As sRNA1 is inhibited, the inhibitory effect exerts on R1; therefore the recombinase is expressed through the first layer, and vice versa. Nevertheless, given that the expression of R1 inhibits PR1, even if the recombinase is expressed PR1 can be flipped over but cannot express the second layer. Since there is no further input signal to the second layer of the bistable system, the inhibitory effect of sRNA2 dominates over R2, therefore, both R2 and the toxin cannot be expressed. That is to say, the additional second layer ensures that there is no leakage of toxin when inducer is present.
However, when the bacterium is stolen and thereby inducer isn’t present, the transcriptional factor is turned off, promoter Pm doesn’t work. Therefore, sRNA1 dominates over R1 and then the recombinase is not expressed in the first layer. Since there is a decrease in the expression of R1, the inhibitory effect it exerts on PR1 decreases, therefore PR1 which has already been flipped over when inducer is present (working condition) initiates the expression of R2. Applying the same logic, finally R2 dominates over sRNA2 and expresses the toxin to kill the cell. (Fig. 4)
Why double layer?
The reason why we designed such double-bistable system can be attributed to the consideration that the double-decked bistable systems of both recombinase and toxin could further reinforce the insurance of zero expression of toxin.
When the second bistable system coupling the toxin with the first layer of recombinase, double insurance of zero expression of recombinase and toxin is achieved. However, as the diagram shown below, a single bistable system where toxin is coupled without the second layer of insurance could collapse if the mutual inhibitory effect of repressor and sRNA within the first layer malfunctions. Whereas such concern could be eliminated by the second layer of toxin with a bistable system around.
Experiment & Results
After the circuit(Fig.1) including only one layer of the bistable system is constructed with recombinase being replaced by sfGFP, LacI & tetR serving as repressor, and used 3MBz & AraC as inducers, we qualitatively tested the fluorescence of the reporter protein under two situations when excited blue light. Theoretically, sfGFP should be expressed when inducer is present since LacI, which transcribes sfGFP, dominates over the inhibitory force of sRNA. However, in the experiments when inducers are present and absent, there is no fluorescence of sfGFP (BN027&28&29&30). We deduced that there must be defects lying in our circuit.
Figure 2
A: BN006(without GFP) when inducer is absent and present
B: BN007(AraC + sfGFP) when inducer is absent and present
C: BN027 (AraC + LacI + sfGFP) when inducer is absent and present
D: BN029 (AraC + tetR + sfGFP) when inducer is absent and present
Figure 3
A: BN006(without GFP) when inducer is absent and present
B: BN007(Xyls + pBAD + sfGFP) when inducer is absent and present
C: BN028(Xyls + LacI + sfGFP) when inducer is absent and present
D: BN030(Xyls + tetR) + sfGFP when inducer is absent and present
Figure 4
As shown in the graph from flow cytometry experiments: in the presence of a fixed concentration of inducer (1.0 mM of Ara) at the indicated time points (5 to 240 min after induction), BN006, which carries no sfGFP, has low fluorescence level under 100 a.u. being expressed in inducer’s presence. BN027(under the control of bistable system) shows similar pattern as time increases as that of BN006. BN007(without control of bistable system) showed higher fluorescence level throughout the experiment.
Figure 5
As shown in the graph from flow cytometry experiments: at same time period, as the conc. of Ara increases from 0 to 1.0 mMol, BN006( no sfGFP) and BN027(under the control of bistable system) showed both fluorescence level around 100 a.u., revealing a lack of expression of sfGFP; while BN007 displayed increase in fluorescence level as inducer concentration increases.
Analyzing through the whole circuit, we deduced that the crux might lie in RBS1 for the sake of its overwhelmingly strong hair-pin structure that even if the 3’ end of LacI is translated the ribosomes produced still couldn’t melt sRNA secondary structure, thereby sfGFP cannot be translated. Therefore, we replaced RBS1 with a commonly used RBS B0034 with weaker strength in order to ensure that there is no problem lying in other sites in the circuit. We tested the sfGFP’s fluorescence in inducers’ presence under excitation of blue light. The expression of sfGFP in the second attempt (Fig. 6B, Fig. 7B & 7D) proved to us that there’s nothing wrong with the rest of the circuit except for RBS1.
Figure 6
A: BN040(Xyls + tetR + RBS B0034 + sfGFP) when inducer is present and absent
B: BN009(Xyls + msfGFP) when inducer is present and absent
Figure 7
A: BN007(Arac + pBAD + sfGFP) when inducer is absent
B: BN036(pBAD + LacI +RBS B0034 + sfGFP) when inducer is absent
C: BN007(Arac + pBAD + sfGFP) when inducer is present
D: BN036(pBAD + LacI +RBS B0034) + sfGFP when inducer is present
Figure 8
As shown in the graph from flow cytometry experiments: at same time period, in the absence and presence(1mMol) of inducers, both plasmids (BN036&BN040) have their fluorescence level 10 times lower in inducer’s absence than that in inducer’s presence.
Through further searching we found that RBS2 has a weaker hair pin structure while it could support the translation of repressor. We replaced RBS B0034 with RBS2. Results are shown in the two graphs below. Plasmid with RBS2 expressed sfGFP with quite a high expression leakage when both inducers are absent. We thought that such high expression leakage might be the result of the weak hair-pin structure of RBS2, which means that we still need a strong RBS to reduce expression leakage. Hence, by we were inspired to realize that we might wrongly include the terminator of LacI before RBS. If we include the terminator, the transcription of LacI stops there and therefore the LacI protein produced by translation ends in front of RBS and leaves the mRNA secondary structure before melting it, sfGFP failed to be expressed.
Figure 9
A:BN009(Xyls + sfGFP) when inducer is absent and present
B:BN045(TetR + Xyls +RBS2) when inducer is absent and present
C:BN049(LacI+ Xyls + RBS2) when inducer is absent and present
Figure 10
A:BN006 (without GFP) when inducer is absent and present
B:BN043 (pBAD + LacI +RBS2 + sfGFP) when inducer is absent and present
C:BN007 (Arac + pBAD + sfGFP) when inducer is absent and present
D:BN051 (pBAD + tetR + RBS2 + sfGFP) when inducer is absent and present
Figure 11&12
As shown in the graph from flow cytometry experiments: at the same time period, the conc. of inducer increases from 0.000 to 1.000 mMol, BN006, which carries no sfGFP, has a low fluorescence level around 0.60 a.u. in inducer’s absence. BN007 and BN009 without the control of bistable system had their fluorescence value at a relatively higher level in inducer’s absence and their fluorescence value increase as conc. of inducers increases. BN043&51&45&49 (under the control of bistable system) have their fluorescence value lower than that of BN007 and BN009 in inducer’s absence and have their fluorescence value as conc. of inducer increases.
Figure 13&14
As shown in the graph from flow cytometry experiments: in the absence of inducer at the indicated time points (4h and 24h after induction), BN006, which carries no sfGFP, has a low fluorescence level around 0.60 a.u. in inducer’s absence. BN043&51&45&49 (under the control of bistable system) have their fluorescence value higher than that of BN006 while lower than that of BN007&BN009 (without the control of bistable system) in inducer’s absence.
We improved our gene circuit upon that reflection, with exclusion of the terminator of repressors and replacement of BRS2 using RBS1, and made the third trial. The results are shown below.
Figure 15
A: BN007(Arac + pBAD + sfGFP) when inducer is absent and present
B: BN055(Xyls + tetR + RBS1 + sfGFP) when inducer is absent and present
C: BN054 (Arac + LacI +RBS1 + sfGFP) when inducer is absent and present
Figure 16&17
As shown in the graph from flow cytometry experiments: in the absence of inducer at the indicated time points (4h and 24h after induction), BN006, which carries no sfGFP, has a low fluorescence level around 0.60 a.u. in inducer’s absence. BN007 and BN009 are two plasmids without the control of bistable system; they displayed high fluorescence level as time passed. BN054 and BN055 have similarly no fluorescence level as BN006.
Figure 18&19
As shown in the graph from flow cytometry experiments: at same time period, as the conc. of inducers increases from 0.0001 (nearly no inducer) to 10 mMol, BN006, which carries no sfGFP, always has a low fluorescence level around 0.60 a.u. While both two plasmids (BN054, induced by Ara; BN055, induced by 3MBz) have their fluorescence level increase from around 10 a.u. (lack of expression of sfGFP) when inducer is absent to a higher value.
The plasmid PSB4K5 which carries our bistable system till now has a low copy number in order to accustom the recombinase we utilized. In order to augment the application range, we transferred the whole system to plasmid P15A, with a medium copy number and tested its usability. Two qualitative results came out under the excitation of blue light shown below verified that the system is also applicable to plasmid with medium copy number.
Figure 20
A: BN057 (P15A + Xyls + tetR + RBS2 + sfGFP) when inducer is absent and present
B: BN058 (P15A + LacI + Xyls + RBS2 + sfGFP) when inducer is absent and present
C: BN073 (P15A + Xyls +sfGFP) when inducer is absent and present
D: BN067 (P15A + Xyls + tetR + RBS1 + sfGFP) when inducer is absent and present
Figure 21
A:BN056(P15A + pBAD + LacI + RBS2 + sfGFP) when inducer is absent and present
B:BN059 (P15A + pBAD + tetR + RBS2 + sfGFP) when inducer is absent and present
C:BN072 (P15A + pBAD + sfGFP) when inducer is absent and present
BN070 (P15A + Xyls + LacI + RBS1 + sfGFP)
BN070 (P15A + Xyls + LacI + RBS1 + sfGFP)
Figure 22&23
As shown in the graph from flow cytometry experiments: in the absence of inducer at the indicated time points (4h and 24h after induction), BN006, which inherently has no expression of sfGFP and therefore low fluorescence level around 5 a.u. in inducer’s absence, has similar fluorescence level as BN067 and BN070(both with RBS1) as time passes; while as we expected, BN056 &57 & 58 & 59 showed similar pattern of fluorescence as BN073 & 72, which have great expression leakage of sfGFP.
Figure 24&25
As shown in the graph from flow cytometry experiments: at same time period, as conc, of inducers increases from 0.000 (no inducer) to 1.000 mMol, fluorescence of BN006 without sfGFP remains at a level below 10 a.u. , while the fluorescence of BN073 & 72 increases as conc. increases at a relatively higher level. BN067 and BN070 (both with RBS1) have similar fluorescence value as BN006 and have their fluorescence value increase as conc. of inducers increase. The rest of the plasmids (BN056 &57 & 58 & 59) that have RBS2 have much higher fluorescence value compared with others and have their fluorescence increase as conc. of inducers increase.
Discussion
Through our design and experimental analysis, our bistable finally achieved the desired result: a state of zero leakage through mutual inhibition. This way we form a zero-and-one-converted module.
After mathematical modeling guidelines and standard and modular operations, we finally digitalize our bistable system. Its good experimental results ensure the scientificity, feasibility, and efficiency of our overall design. The basic characteristics of preventing leakage also allow us to put toxin to complete the construction of our complete suicide system.
Our digitalized bistable module can be used efficiently in other synthetic biology and related iGEM projects. This zero-expression leakage effect will effectively promote the development of synthetic biology engineering.
References
【1】Aiba H (2007) Mechanism of RNA silencing by Hfq-binding small RNAs. Current Opinion in Microbiology
【2】Anthony LC, Suzuki H, Filutowicz M (2004) Tightly regulated vectors for the cloning and expression of toxic genes. Journal of Microbiological Methods 58: 243-250
【3】Aparicio T, de Lorenzo V, Martínez-García E (2017) Broadening the SEVA Plasmid Repertoire to Facilitate Genomic Editing of Gram-Negative Bacteria. In Hydrocarbon and Lipid Microbiology Protocols: Genetic, Genomic and System Analyses of Pure Cultures
【4】McGenity TJ, Timmis KN, Nogales B (eds), pp 9-27. Berlin, Heidelberg: Springer Berlin Heidelberg Balzer S, Kucharova V, Megerle J, Lale R, Brautaset T, Valla S (2013) A comparative analysis of the properties of regulated promoter systems commonly used for recombinant gene expression in Escherichia coli. Microbial cell factories
【5】Bervoets I, Charlier D (2019) Diversity, versatility and complexity of bacterial gene regulation mechanisms: opportunities and drawbacks for applications in synthetic biology.
【6】Aiba H (2007) Mechanism of RNA silencing by Hfq-binding small RNAs. Current Opinion in Microbiology
【7】Anthony LC, Suzuki H, Filutowicz M (2004) Tightly regulated vectors for the cloning and expression of toxic genes. Journal of Microbiological Methods 58: 243-250 Aparicio T, de Lorenzo V, Martínez-García E (2017) Broadening the SEVA Plasmid Repertoire to Facilitate Genomic Editing of Gram-Negative Bacteria. In Hydrocarbon and Lipid Microbiology Protocols: Genetic, Genomic and System Analyses of Pure Cultures