Kill-switch

introduction

This project aims to protect the natural balance within the soil and provides maximal safety. We engineered a regulatory system that will only activate stable expression of fungal growth inhibitors when needed and will self-destruct the bacteria when the bacterium fulfilled its purpose. With this system we reassure that there are as little GMO bacteria in the soil as possible and that they are only active when BananaGuard is needed. The Kill-switch system will reassure regulated expression by providing a stable expression of the anti-fungal components that are activated only during exposure to fusaric acid. After exposure to fusaric acid the Kill-switch will have a stable expression of toxins that will kill the bacteria. This reassures that the bacterium removes itself when the threat of Fusarium oxysporum is gone.

To achieve these goals, we conducted the following projects :

• The regulative system with promoters from the registry.
• A new characterization method that is able to characterize non inducible promoters.
• Design, produce and test new promoters that are repressible by multiple repressors.

In the end we were able to construct a functional input/output plasmid, a toggle switch that is likely to be stable and develop a new way to characterize promoters. With this new method we characterized the Pci/lac promoter and determined its RPU. You can find the parts that we made here.

During the project there was a close collaboration with the modelling part of our project. The system was modelled and its conclusions where used to design the system and new promoters.

The Regulatory System

The Kill-switch regulatory system consist of a two plasmid system as can be seen in Figure 1. One will contain the input system, rhamnose promoter regulating the CIλ repressor gene, and the output pCI/Tet regulating the toxin to kill the bacterium, currently replaced by the reporter gene GFP. The other plasmid contains a genetic toggle switch based on the principle of Gardner et al 2000 [1].

The toggle switch functions as a memory system for the input signal. In the final system the rhamnose inducible promoter will be exchanged for the fusaric acid induced promoter and the reporter will express a toxin that will kill the cell. We used these repressors because they are proven to work and are well characterized. The promoters we used where the only promoters that could be double repressed by our selection of repressors. The ramose promoter was chosen due to its concentration specific expression and GFP due to its usefulness as reporter gene. The function of the Kill-switch can be explained in three states. These states are shown Figure 2-4. In the first state, only the repressor TetR will be expressed, repressing pCIλ/Tet and pTet. Initially, the toggle switch will be set to this state by IPTG inhibiting the binding of LacI to the promoter.

The rhamnose inducible promoter is induced by rhamnose therefore CIλ is expressed, switching the circuit to its second state. In the second state fusaric acid induced promoter is active upon F. oxysporum recognition, for research purposes replaced by rhamnose and a rhamnose induced promoter Therefore CIλ represses pCIλ/Tet and pCIλ/Lac. This leads to a switch within the toggle switch. TetR can no longer repress pTet, therefore LacI is expressed, keeping the toggle switch in its final state, in which TetR is no longer expressed.

When the organism cannot sense rhamnose anymore the third state of the Kill-switch is induced. In this final state the pCIλ/Tet promoter is not repressed by the CIλ repressor or the TetR repressor leading to the expression GFP. In the final system this GFP will be a toxin that will kill the cell. Besides the self-destruction system the Kill-switch circuit has lots of potential applications. The regulatory properties of this system can be utilized for any genetic system that needs to memorize the input signal and express the output gene when this signal is gone. The input promoter can be changed into the sensing promoter of choice. Likewise, the output gene can be changed giving the system a great flexibility. An even stronger improvement would be the addition intercellular communication, making a collective synchronized memory possible, which enables simple control over vast populations of bacteria equipped with this system.

The Input Output Plasmid

The input/output plasmid is the plasmid that can sense rhamnose and produce GFP or toxin, see Figure 5.

The input/output plasmid is the plasmid with the rhamnose promoter, the CIλ repressor, CItet promoter and the GFP or toxins see figure 5 and 6. It is assembled with the standard BioBrick assembly method. The end product is assembled into a low copy number plasmid to reduce the metabolic stress, enable faster reactions to input repressors, and lower the GFP production so that it can be measured more accurately, and to reduce the possibility that leakiness in toxin production will kill the cell. Another low copy plasmid than pSB3K3 can also be used.

Results

Figure 7 shows cells carrying the pSB3K3 plasmid containing the Kill-switch input output plasmid (BBa_K1493560 ) that were incubated on LB agar plates containing 0.2% rhamnose and no rhamnose overnight at 37°C.

On plates without rhamnose the construct should express GFP since the rhamnose promoter is not active so there is no repression of the promoter expressing GFP. On the plates with rhamnose colonies should not express GFP since the rhamnose promoter should be active and producing repressors that repress the promoter that produces GFP. Under UV light the colonies on the plate with no rhamnose shows green fluorescence. The colonies induced by rhamnose don’t show a green colour. From the picture in Figure 7 we can conclude that the production of GFP is repressed by the repressor protein CIλ.

Since this plasmid is also used to characterize the pCI/Tet promoter (see rhamnose mediated characterization) it also gives a first suggestion that this method works.

The Toggle Switch Plasmid

The toggle switch is a memory system consisting out of the PCIlac promter, TetR gene, Tet promoter and LacI repressor, see figure 8 and 9.

It is assembled with the standard BioBrick assembly method. The end product is assembled into a low copy number plasmid. Another low copy plasmid than pSB3K3 can also be used. To test the toggle switch a GFP was added behind the TetR gene.

In the active state the promoter pTet is on expressing the lac repressor protein LacI, which binds to pCI/lac and inhibits the production of TetR, keeping it in the same state. Besides the lac gene, a second gene is arranged downstream of the tet promoter: the gfp gene. We expect to see high fluorescence values for the active state, since both LacI and GFP will be expressed. In the resting state only TetR will be produced, repressing pTet, thus no GFP is expressed resulting in low fluorescence. The active state can be induced by adding the chemical anhydrotetracycline (aTc). aTc is an inhibitor of TetR, which induces the toggle switch to start expressing LacI and GFP. The resting state is induced by the chemical Isopropyl β-D-1-thiogalactopyranoside (IPTG), which is an inhibitor of LacI.

Results

To determine if the toggle switch is functional and the repressor proteins produced under the promoters inhibit the gene expression of each other the system was characterized. Three cultures were prepared in M9 medium in duplo, inoculated with Escherichia coli strain NEB5α carrying the toggle switch with reporter. The first containing 500 ng/ml aTc to get the cells into the active state and the second inducing the resting state by 2 mM IPTG as described by Gardner et al[1]. A third culture is grown without inducer, evoking a random state. E. coli carrying pSB3K3 with CI and lacI with no promoter was used as an auto fluorescence control. Cultures were grown overnight and fluorescence was measured the next morning using a plate reader (395 excitation, 509 emission). Cells did not grow well in M9 with 2mM IPTG, as the OD was low (OD 600 nm of 0.3) compared to OD 600nm 0.6-0.7 of the cultures grown with ATC or with no inducer, after overnight incubation.

Figure 10 shows that the cultures grown in M9 with 500 ng/ml aTc have an average relative fluorescence unit of 7300. The cultures grown with 2 mM IPTG give an average fluorescence unit of 1000. The cells grown in M9 with no inducer give a fluorescence of 5200 RFU. These values indicate that the toggle switch is functional as it has a high fluorescence when grown with aTc, which means it is in its active state. The resting state is reached with low fluorescence when grown with IPTG. We can conclude from the RFU value of the cultures grown with no inducer that the culture has a mix of both states. Thus, the toggle switch is able to choose a random state and no state is significantly more preferred than the other . This is in contrast to what the model predicted.

Promoter Design

The model of the system we build predicted that the system would not be stable with the promoter configurations we had been using so far. These promoters, see figure 11, where promoters from the registry with the same operator sites that we needed, but configured in such a way that it was likely they would not work. This was predicted by the model of the kilswitch . Therefore promoters with another configuration of the operator sites of the promoter would be necessary. With the results of the model we build 6 new promoters based on the sequences of Elowiz et al 2007 [2].

Promoters have 3 different regions that are separated by the -34 and -10 RNA polymerase binding sites, see figure 12. In the article Elowiz et al 2007 [2] these are called the distal, core and proximal region.

Binding sites on the core, distal and proximal regions have different strengths, generally the core region is the strongest followed by the proximal region and finally the distal region [2] [3]. By putting repressor operator sites on different positions you can influence the strength of the repression [2, 3] and place the most important repressor sites on the strongest regions. In theory systems would be tuneable with this method. Here we give this theory a try by designing promoters for a toggle switch that have operator sites at the same places so that the toggle switch is as balanced as possible. The most stable setup was modelled and the results can be found here. The promoters from the registry where not balanced in this way, so we made new promoters with the sequences provided in the article of Elowiz et al 2007 [2]. In the article a set of different sequences for the proximal, core and distal regions was used that can be combined in different combinations. We used these sequences to design 6 different promoters in silico/. The operator configurations of the 6 promoters are listed below:

• Empty site, TetR site, TetR site (-- Tet Tet)
• TetR site, empty site, TetR site (Tet – Tet)
• CIλ site, LacI site, LacI site (CI Lac Lac)
• TetR site, CIλ site, TetR site (Tet CI Tet)
• CIλ site, TetR site, TetR site (CI Tet Tet)
• LacI site, CIλ site, Lac site (Lac CI Lac)

We then ordered two single stranded sequences that had an overlapping region, and used these to make a double stranded promoter with prefix and suffix so they can be assembled according to BioBrick Standard 10. These promoters where ten assembled into pSB1C3 and other plasmids to make the Kill-switch system.

Results

We were able to assemble the promoters with GFP, to characterize their functionality after rhamnose induction. All promoters except for promoter 3 (CI Lac Lac) showed GFP expression. See also the plates depicted in Figure 13.

Figure 13 shows all the promoters (P1-6) with GFP on ampicillin plates exposed to UV light. On each plate are colonies that are fluorescent suggesting correct constructs in these bacteria, except for promoter 3. After sequencing all parts had the expected sequences so promoter 3 is either a very weak promoter or the sequence used is not working. Future work with a plate reader could give some more conclusive results.

Rhamnose Mediated Characterization

Some well-known promoters like the Lac promoter and the Tet promoter have a repressor that can be inactivated by a chemical, making characterization relatively easy. In this way the promoter can be induced by this chemical. By treatment of different concentrations of this chemical there will be different expression levels of this promoter. With this information you can compare different promoters. The drawback of this is that you need a promoter that is inducible with a chemical. There are also repressors and promoters where that is not possible such as the CIλ promoter and repressor. The CI lambda repressor is not inhibited by a chemical so the CI lambda promoter is not inducible. To characterize non inducible promoters we developed a new system to characterize promoters based on a two plasmid system with a ramose promoter and GFP, see Figure 14.

The final design to characterize promoters includes the following two components:

• A repressor plasmid with a rhamnose promoter and a repressor of choice
• A promoter GFP plasmid with your favoured promoter and GFP

The double repressible promoters used in the kill switch system are taken from the iGEM Registry. Since these promoters pCI/lac and pCI/tet were not characterized we could not estimate if a functional kill switch could be constructed with these promoters. The double repressible promoter pCI/lac from the toggle switch, is characterized. pCI/lac was combined with a gfp reporter gene under Elowitz RBS (BBa_I13504 ) to make the new BioBrick pCI/lac gfp . As a reference pRha and the tetracyclin promoter pTet are also assembled to the gfp gene (BBa_K1493501, BBa_K1493504 ). Since the promoter strength of pTet is known in relative promoter unit (RPU) it can be used as a reference promoter to estimate promoter strength [4]. In this way, we can compare the constitutive promoter strength at one time point and give the characterized promoter a RPU value. The rhamnose promoter was successfully assembled to the CI (BBa_P0451) and lac (BBa_P0412) repressor protein genes under Elowitz RBS creating new BioBricks BBa_K1493510 and BBa_K1493520.

Results

The graphs containing all data points can be found in here. The characterization protocol can be found here.

The measurements in figure 15 indicate an activation of pRha by L-rhamnose. The RFU values of 0% and 0.001% rhamnose are not significant taking into account the high standard deviation for these measurements as can be seen in figure 15 B. From a rhamnose concentration of 0.01% to 0.2% a significant increase in fluorescence is measured. Fluorescence from figure 15 can be used to predict the concentration of repressor protein produced. Data points for time 8.15 were chosen for the graph in figure 15 B due to the peak at time for 0.2% rhamnose, which is visible in figure 15 A. This peak can be explained by the rhamnose depletion, as it is consumed by E. coli, causing the stop of promoter induction. We assume that at this time point the highest concentration of repressor proteins are present in the cells inhibiting the expression of GFP.

The constitutive promoter strength of pCI/lac , is determined using the results of the fluorescence measurements. The RFU values of NEB5α E. coli strains containing pCI/lac GFP and pTet GFP are shown in figure 16. As can be seen in this figure pTet is a stronger promoter than pCI/lac. Using the results shown in figure 16 the RPU value of pCI/Lac is determined using the known value of pTet as stated by Kelly et al. 2009 [4]. Data from time point 8.13 was chosen for calculation. Compared to the pTet RPU value 1.5, pCI/lac has a RPU of 1.0.

Figure 17 shows that a higher concentration of L-rhamnose gives a lower RFU value. This means that the GFP expression is repressed by the repressor protein produced under the rhamnose promoter. As shown in figure 15 a low concentration of rhamnose (0.001%) does not have any substantial effect on the expression rate of the protein. This is also visible in figure 17 in which for both CIλ (A) and LacI (B) 0.001% rhamnose doesn’t show a lower fluorescence compared to 0% rhamnose. We expected the same RFU values for cultures grown in medium with 0.2% glucose and 0% rhamnose, since glucose functions as a repressor for pRha. Though, cultures grown in medium with glucose show lower RFU values in both graphs in figure 17. This can be explained by the higher growth rate of the cells in glucose containing medium, since the RFU value is OD dependant (notebook). In figure 15 0.01% rhamnose shows a higher RFU compared to the culture grown on 0% rhamnose, which indicates that pRha is active at that concentration at that time point. However, at the concentration of 0.01% rhamnose both CIλ (A) and LacI (B) don’t show any substantial repression. We can conclude that the concentration repressors produced at the promoter strength of pRha in 0.01% rhamnose is not high enough to perform repression of the promoter pCI/lac. Though, cultures grown in 0.05% and 0.2% rhamnose show a lower RFU value for both constructs as the RFU in figure 15 increases in higher rhamnose concentrations. We can conclude that repression of pCI/lac is stronger in higher concentrations rhamnose, since higher concentrations rhamnose lead to a higher concentration of repressor proteins. We expected the same RFU values for cultures grown in medium with 0.2% glucose.

Future work

Future work: In the last month we made some characterization plasmids to characterize the registry promoters and the new set of promoters, but due to a lack of time we were not able to finish them all or to characterize them all. In future work they could be finished and the new and registry promoters could be fully characterized. More experiences with this new type of characterization could also further optimize the protocol and the rhamnose concentrations that should be used. When fully developed this could become a standard method to characterize repressible promoters, since it is able to characterize non chemically inducible promoters with different levels of rhamnose. The final regulatory system was only a few steps away from completion, but is due to a lack of time not finished. In future work this system could be finished, tested and characterized. When fully finished this system could become a mayor basis for safety systems and regulated expression. It can be implemented as a Kill-switch as would be the case for BananaGuard or as regulator for production in a bioreactor where the bacteria only produce products when the rhamnose is cleared out of the media.

Parts summary

New promoter parts

The promoters from the registry

New promoter parts

BioBrick code	Feature	Operator Configuration	Code
BBa_K1493801	Promoter repressed by TetR	-- Tet Tet	P1
BBa_K1493802	Promoter repressed by TetR	Tet -- Tet	P2
BBa_K1493803	Promoter repressed by CIλ and LacI	CI Lac Lac	P3
BBa_K1493804	Promoter repressed by CIλ and TetR	Tet CI Tet	P4
BBa_K1493805	Promoter repressed by CIλ and TetR	CI Tet Tet	P5
BBa_K1493806	Promoter repressed by CIλ and LacI	Lac CI Lac	P6

The promoters from the registry

BioBrick code	Feature	Used for?
BBa_K909012	Promoter repressed by LacI and CIλ	In-output plasmid
BBa_K909013	Promoter repressed by TetR and CIλ	Toggle switch plasmid
BBa_R0040	Promoter repressed by TetR	Toggle switch plasmid

Rhamnose mediated characterization parts

BioBrick code	Feature	Used for?
BBa_K1493520	Rhamnose promter with lacI	To characterize promoters with a LacI operator
BBa_K1493540	Rhamnose promter with CIλ and lacI	To characterize promoters with an CIʎ operator and an LacI operator
BBa_K1493703	Promter represed by CIʎ and LacI with TetR	To characterize promoters an cIʎ operator and an TetR operator
BBa_K1493504	Tet promter with GFP	To characterize the Tet promoter
BBa_K1493501	Rhamnose promter with GFP	To characterize the Rhamnose promoter
BBa_K1493560	Rhamnose promter with CIʎ and pCI/Tet with GFP	To characterize the pCI/Tet promoter for the CIʎ repressor
BBa_K1493530	Rhamnose promter with TetR	To characterize promoters with a TetR operator
BBa_K1493502	pCI/Lac with GFP	To characterize the pCI/Lac promoter
BBa_K1493503	pCI/Tet with GFP	To characterize the pCI/Tet promoter
BBa_K1493510	Rhamnose promoter with CIʎ	To characterize promoters with a CIʎ operator
BBa_K1493701	pCI/Lac with TetR and GFP	To characterize the toggle switch
BBa_K1493702	Tet promoter with LacI and GFP	To characterize the toggle switch

Parts we used

BioBrick code	Feature	Used for?
BBa_K909012	Promoter repressed by LacI and CIλ	In-output plasmid
BBa_K909013	Promoter repressed by TetR and CIλ	Toggle switch plasmid
BBa_R0040	Promoter repressed by TetR	Toggle switch plasmid
BBa_K914003	Rhamnose inducible promoter	In-output plasmid
BBa_I13504	CIλ repressor	in-output plasmid
BBa_P0440	TetR repressor	toggle switch plasmid
BBa_P0412	LacI repressor	toggle switch plasmid
BBa_I13504	GFP	Reporter
BBa_I13504	CIλ	Repression of promoter CI/Tet and CI/Lac
BBa_P0440	TetR	Repression of promoter CI/Tet and pTet
BBa_P0412	LacI	Repression of promoter CI/Lac

References

1. Gardner, T.S., C.R. Cantor, and J.J. Collins, Construction of a genetic toggle switch in Escherichia coli. Nature, 2000. 403(6767): p. 339-42.
2. Cox, R.S., 3rd, M.G. Surette, and M.B. Elowitz, Programming gene expression with combinatorial promoters. Mol Syst Biol, 2007. 3: p. 145.
3. Lanzer, M. and H. Bujard, Promoters largely determine the efficiency of repressor action. Proc Natl Acad Sci U S A, 1988. 85(23): p. 8973-7.
4. Kelly, J.R., et al., Measuring the activity of BioBrick promoters using an in vivo reference standard. J Biol Eng, 2009. 3: p. 4.