Although we weren’t able to go into the lab this year, due to the current COVID-19 pandemic, we planned our laboratory work in detail. Our aim was to learn about typical problems in the lab and how to solve them. Following the design-build-test engineering cycle, we came up with a plan how to evaluate our experimental outcome and deal with any unexpected results.
Degrading Diclofenac and Azithromycin by using Degradation Enzymes - Laccase and EreB
As we seek to degrade specific micropollutants in wastewater, we chose two bacterial laccases and one esterase to work with. Laccases are able to oxidize phenolic compounds, such as diclofenac, and the chosen esterase, EreB, is able to degrade erythromycin and also shows activity towards azithromycin, which is similar in structure. We want to express the three enzymes in E. coli strains with a strong promoter for high protein production rates. Cloning will be verified via (commercial) Sequencing. Purification of the target protein will be ensured by a fused StrepTag II affinity tag. After performing an SDS-PAGE analysis, a large band is expected at the protein size of the expressed enzyme (CueO: 57.81 kDA, CotA: 59.74 kDa, EreB: 49.21 kDa), since the enzymes are connected to a strong T7-promoter. Important kinetic data of our transformation enzymes can be measured after successful protein expression and purification.
Laccases are multicopper enzymes belonging to the group of oxidases that catalyse the oxidation of phenolic units in lignin as well as a wide variety of organic compounds including various types of phenols and aromatic amines. Due to this oxidoreductase activity laccases can react with 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) also known as ABTS as a substrate (figure 1). We would have used ABTS in a spectrophotometric assay to determine the activity of our laccases. This assay is based on the formation of a stable radical cation by enzymatic
Figure 1: Structure of 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid)
oxidation. The resulting cation can be measured photometrically[1,2]. Accordingly, the laccase activity would have been calculated from the optical density (OD) changes at linear range with 420 nm.
We would have performed the enzyme assay immediately at 30 °C after mixing the enzyme solution and substrate buffer (phosphate-citrate-buffer, pH 4.0). For the calculation of the enzyme activity we would have defined the transformation of 1 µmol substrate to product per minute as one unit of enzyme activity[2].
On the one hand, the gathered data could confirm the successful production of active enzyme and thus serves as a quality control. On the other hand, we would have a defined enzymatic activity of our produced laccases and could compare it to the activity of transformation of pharmaceutical compounds like diclofenac to evaluate a kind of efficiency. If we would not have been able to detect any enzymatic activity there is most likely a problem with the production or storage of our enzymes. Maybe the recombinant production of our laccase did not work properly, and only inactive forms were produced due to protein folding problems. In this case we would have to rethink our production process and change it accordingly.
As this assay only accounts for the laccases, we would have performed a different activity assay for EreB. Prior to EreB site saturation mutagenesis, a simple Kirby Bauer assay will be performed to roughly illustrate the activity of EreB for both erythromycin and azithromycin. Note that the results are relative, therefore the experiment would have been carried out in close proximity to the approach of iGEM Munich 2013. For that, the same reporter organism, which would be Micrococcus luteus, would need to be grown on an agar plate. Also note that the experiment would need to be repeated after mutagenesis of EreB, as the test has no absolute outcome and instead provides results for a qualitative comparison. This relative comparison prior/after mutagenesis can demonstrate improvements in azithromycin degradation ability. The best outcome would be decreased sizes of inhibitory zones, as that would mean increased degradation activity for the same amount of time. To preserve the amount of enzyme available, a general, qualitative-only incubation of EreB with erythromycin and azithromycin would have been performed with a negative control group, as no positive control is available for this enzyme yet. For this we would have incubated an equimolar amount of erythromycin and azithromycin respectively onto two filter plates, one containing EreB and one containing buffer only. With standard E. coli strains or Micrococcus luteus, we would have checked inhibitory effects of erythromycin and azithromycin. If the negative controls showed significant inhibitory zones while the second plate did not, we could qualitatively prove the degradation capability of EreB[3,4]. The exact concentration necessary for EreB to perform this assay can only be determined in the lab, as no literature is available. The concentration in the table listed below can be a good starting point and was used in a comparable assay from iGEM Munich 2013[3]:
Substance
Amount
Stock solution
EreB recombinant protein 40 nM
0.18 µl
11 µM EreB in PBS with 10 mM ß-mercaptoethanol, 2% glycerol (v/v) and 300 mM NaCl
Erythromycin 0.36 mM or Azithromycin 0,36 mM
3 µl
Erythromycin/Azithromycin in ethanol, 6 mM / 6 mM
Tris-HCl buffer pH 7.5 100 mM
10 µl
Tris-HCl buffer pH 7.5, 500 mM
NaCl 0.08 M
4 µl
NaCl in water, 1 M
ddH20
32.82 µl
ddH20
TOTAL:
50 µl
Table 1: Suggested concentrations of protein EreB and its substrates azithromycin and erythromycin for the Kirby-Bauer-Assay as used by iGEM Munich 2013[3]
To solve the various problems of our project we need appropriate and robust analytical methods. One of the main topics of our project is to enable and improve the degradation capabilities of diclofenac by the bacterial laccases CueO (BBa_K3429012) and CotA (BBa_K3429011), as well as improving the degradation of azithromycin by EreB (BBa_K1159000), as the latter is already able to degrade the structurally similar erythromycin. We therefore chose High Performance Liquid Chromatography (HPLC) as a powerful and robust analytical method. HPLC analysis would have enabled an evaluation as well as a comparison of the degradability. As for all three enzymes, we would thus determine the enzymatic degradation and degradation rate of the respective wastewater polluting drugs. Moreover, we could verify the educt transformation with HPLC analysis.
The detection method is based on the UV-absorbance of the substrate after its retention by a modified porous silica material. As for the laccases, we would have analysed a commercially available standard of the target substance diclofenac (Merck, Darmstadt, #38429)[5]. Thereby, we would have determined the method specific retention time and the UV-absorbance at different concentrations. Based on this initial test, we could analyse and estimate the substrate conversion in the reaction mixture containing our laccase in comparison to a reaction mixture lacking the laccase. On the one hand we could have observed the decrease of absorbance from diclofenac in presence of the laccase, which indicates its functionality. In this case, new absorbance signals according to the transformation products should be detectable. Based on this data we would have been able to estimate a degradation rate of the substance by the used laccase, which in turn would have allowed us to compare those degradation rates for the different enzymes and their modifications. This would greatly help to identify beneficial enzyme modifications.
On the other hand, we could have observed no significant change in the absorbance signal. In this regard, ”significant” would be a change clearly outside the defined error limits determined by the standard. No significant change would mean that no enzymatic degradation of the substrate has taken place. In this case, we would have gone through the whole assay step by step to identify possible sources of error. A possible problem might be inactive enzyme. Then we would have to determine the enzymatic activity by the ABTS assay again. When no error could have been found then maybe our enzyme could not have been able to degrade the substrate. Then we should try a different mutation approach or use another enzyme.
As we would have performed the same HPLC kinetic activity assay with EreB as we did for the Laccase, the discussion presented above can also be applied. The best outcome of the enzymatic assays for EreB would be an increased activity of EreB regarding azithromycin, and ideally, a constant or improved activity regarding erythromycin. This implies that enzymatic activity assays must be performed at least twice, prior and after site saturation mutagenesis, for both erythromycin and azithromycin.
In addition to HPLC analysis, we planned to do a LC-MS to identify the erythromycin and azithromycin degradation products. As the mechanism of reaction starts with the cleavage of the macrolide ring, in sequence to which the added water molecule is eliminated during the acetal formation, the substrate and the product show identical molecular mass. We must therefore detect and interpret the open lactone intermediate which is expected to be 18 g/mol heavier due to H2O addition. The molecular mass of erythromycin is 733,5 g/mol, while the hydrolyzed hemiketal is 751,6 g/mol[6]. For azithromycin the respective molecular masses are 748,98 g/mol for the macrolide and 768,98 g/mol for its degradation product. Unless the antibiotics are fully degraded during incubation, the LC is expected to yield three successive peaks with partly different molecular masses and retention times for the cleaved and the non-degraded molecules. It is therefore possible, through comparison of the calculated to the measured molecular weight, to determine whether the molecule is cleaved by EreB the way it is expected from literature. The exact protocol for the suggested LC-MS gradient can be found on our experiments page. As the product does not differ from the intermediate substantially in terms of molecular weight, the gradient of water and formic acid used might need adjustments until the peaks are clearly differentiable based on their retention times. We expect to see a single peak in the control group with untreated erythromycin or azithromycin, and three peaks for the samples containing EreB, of which the molecular mass of the substrate and the final product vary only in their retention times and are otherwise indistinguishable during LC.
By site directed mutagenesis we hope to achieve an increased affinity of our laccase-oxireductases towards diclofenac (figure 2). We chose the targeted amino acids based on already published discussions for directed mutagenesis on the corresponding enzymes. Further experiments following our laccase-activity assays are needed to quantify the increase/decrease in activity. An ABTS assay is appliable for measurement of unspecific transformation activity and HPLC analysis for target specific measurements of substrates that do not absorb visible light spectra. The same conditions as with the wildtype enzymes should be used here to increase comparability.
Figure 2: Site directed mutagenesis. The plasmid containing the GOI is replicated using a primer pair with a built in mutation to replicate nicked plasmids containging the desired mutations. The plasmids are then transformed into competent cells for ligation and expression of the engineered protein.
Not only positive mutants showing an increase in activity should be used in further experiments but also primary lower efficiency laccase variants, as there might be hidden epistase effects (the mutation at one site can influence the effect of other mutations). This way even unfavorable seeming mutants can cause an increase in enzymatic activity by changing the impact of other mutations by specific epistasis. Therefore, mutation sites should also be combined to check their compatibility and mutual effects. Experimental results can be predicted/verified by small molecule ligand docking.
Since there are no publications on directed mutagenesis of EreB, we consider a site saturation mutagenesis (SSM) approach. Because wild type EreB already shows promiscuous activity to azithromycin, we hope to improve the enzyme to bind and degrade azithromycin more effectively. As shown before, enzymes with promiscuous activities towards other substrates are a great starting point for the directed evolution of enzymes for new substrate specificities. For SSM we want to focus on the residues close to EreB’s active site to improve substrate affinity and accessibility. Mutants can be screened by either Kirby Bauer Assay or HPLC analysis. Unfortunately both assays are no methods capable of high throughput screening for positive mutants so we want to focus on mutating residues close to EreB's active site to generate a limited, screenable amount of mutants for low-throughput screening towards improved activity on azithromycin.
The Zebrafish Embryo Acute Toxicity Assay would have been used to determine the acute fish toxicity as well as the developmental and reproductive toxicity (DART) of the enzymatic transformation products from wastewater polluting drugs such as diclofenac and the antibiotic azithromycin. The comparison of the enzymatic transformation products with the original substance provides an evaluation of whether the laccase/EreB treatment reduces the toxic impact of these drugs on the aquatic environment.
For this purpose, embryos of the zebrafish (Danio rerio) are used, which are an established model in ecotoxicology and a good alternative to animal experiments on adult fish species (figure 3). This assay is based on the OECD Test Guideline 236 for the Fish Embryo Acute Toxicity (FET) test and is used to determine the lethal concentration LC50 value of diclofenac, azithromycin and their transformation products. The calculation of the LC50 value is based on four apical observations as indicators of lethality within a period of 96 hours post fertilization.
Figure 3: By using a Zebrafish Embryo Acute Toxicity Assay, we efficiently prove the decrease of toxicity for aquatic environment. The assay allows measurement of toxicity to a tailored model-organism, without doing actual animal testing.
Using negative controls consisting of a water and a solvent control as well as a positive control with the substance 3,4-dichloroaniline we would have gotten defined limits.
In addition, we would have tested the initial substance. With this experimental setup we would have been able to compare the caused toxicity of the enzymatic transformation products with the toxicity of the initial substance.
A positive outcome would have been that the transformation product is less toxic than the initial substance or does not show any toxicity at all, by comparison with the solvent control. If enzymatic degradation has been occurred but the transformation products had the same toxic effect as the initial substance, we would have used those transformation products in a new experiment to show the biodegradability within bioactive sludge. It could have been that the transformation products are still toxic but show better biodegradability and could therefore be better removed in sewage treatment plants.
This method would have enabled us to observe a direct effect of performed enzyme modification to the resulting degradation ability. By taking this into account we could have identified enzyme variants for further optimization and fine-tuning.
Figure 4: Diclofenac transformation products. Confirmed enzymatic transformation products of diclofenac by the laccase from T. Versicolor.
Like the laccases CotA/CueO, EreB is intended to be surface displayed in the matrix of B. subtilis. To guarantee its functionality, in vitro assays should be under oxidizing conditions (air) as they are present in the biofilm environment. The display itself can only be tested in vivo (correct biofilm-formation and orientation of the fusion protein can hardly be predicted). It is expected, that B. subtilis can appropriately fold EreB, CueO and CotA.
Both for E. coli and B. subtilis, in vivo assays on survival in erythromycin/azithromycin or diclofenac contaminated environments would be of interest. However, the low concentrations of erythromycin/azithromycin or diclofenac in wastewater are not expected to inhibit bacterial growth, and measurement of susceptibility of bacteria for higher pharmaceutics concentration is not representative for the final application. The impact of exceptionally low concentrations is hardly measurable in in vivo assays.
For B. subtilis expression, the plasmid encoding EreB and CueO were also codon optimized within the sequence (CotA is orginated from B. subtilis), however this would require an additional cloning step.
Displaying our degradation Enzymes in the Biofilm Matrix - Fusion Proteins with TasA
Based on our research our goal was to degrade toxic compounds in wastewater by enzymes. These enzymes should be stable and work as efficient as possible. Therefore, we imagined different solutions. Besides secreting our enzymes by microorganisms into the wastewater or add purified enzymes, immobilization would be a possible solution. Immobilization can be realized on various materials but also by using biological systems. Huang et al. reached immobilization of proteins and enzymes in the biofilm matrix of Bacillus subtilis[7]. They successfully built fusion proteins of the matrix protein TasA with other proteins or enzymes[7].
Considering the manifold advantages of immobilization and the research of Huang et al. we decided to display our degradation enzymes in the biofilm matrix of B. subtilis as fusion proteins with TasA. The matrix protein TasA fits best as a fusion protein because it is a main compound of the biofilm matrix leading to a high number of enzymes displayed. By adjusting the fusion proteins, we could display different enzymes, making our project more modular. Since we use a living system, we avoid elaborate purification of enzymes and can easily adapt our biofilm.
Huang et al. already showed that fusion proteins consisting of TasA and another protein (e.g. mCherry or SpyTag proteins) are successfully exported into the biofilm matrix. Even a fusion protein of TasA with the enzyme MHETase (63.1 kDa), that is bigger than the enzymes we want to use, was localized in the biofilm matrix of B. subtilis.
We conceptualized fusion proteins with our degradation enzymes and the protein TasA (BBa_K3429001) based on the research of Huang et al The genes of our laccases CotA and CuO are fused to the 3' end of the tasA gene with a gene fragment encoding a flexible glycine-serine linker (ARGGGGSGGGGS)(BBa_K3429003), that was also used by Huang et al.
Figure 5: The degradation enzymes are displayed in the biofilm matrix as fusion proteins with the matrixprotein TasA. The gene for the fusion protein is integrated into the genome of B. subtilis. TasA is fused to the enzyme via a flexible glycine-serine linker.
The fusion protein will be inserted into an appropriate plasmid via Gibson Assembly. Overhangs and primers were designed successfully in silico with the software SnapGene. As a plasmid we use the shuttle vector pSEVA3b67Rb[8]. This high copy plasmid is compatible as reporter for Escherichia coli and B. subtilis and has a chloramphenicol resistance[8]. This resistance will be exchanged against an ampicillin resistance via Gibson Assembly, because the strain we would like to use for our proof of concept experiments also owns a chloramphenicol resistance.
Cells of the B. subtilis strain GP1622 will be transformed with this plasmid using a protocol obtained from Prof. Stülke (Georg-August-Universität Göttingen)[9]. The strain also obtained from Prof. Stülke lacks the genes sinR and tasA[9]. With the sinR knockout we improve the biofilm formation and the tasA knockout makes sure that only our fusion proteins are integrated into the biofilm matrix[10].
As a proof of concept, we built a fusion protein of TasA and superfolder green fluorescent protein (sfGFP) (BBa_K3429002). The sequence of the sfGFP gene is codon optimized for B. subtilis and fused to tasA as described above. The display of the TasA-sfGFP fusion protein in the biofilm matrix can be verified using fluorescence microscopy. Therefore, we would use the measurement kit of iGEM to standardize our results.
We hypothesize that TasA fusion proteins with our targeted degradation enzymes will succeed, if TasA-sfGFP is successfully expressed in the biofilm matrix.
In the following we discuss possible outcomes and possible adaptions of our experimental setup.
If our biofilm is built correctly but we are unable to measure any fluorescence we can conclude that there has to be a problem with the sfGFP. It can mean that it was not even expressed or lost its function due to the fusion with TasA. To react on the former, we would check the sequences of our ordered sfGFP-TasA gblock as well as doing a plasmid preparation with following sequencing of our PCR amplified fusion protein parts. Approaching to a possible loss of function due to the fusion with TasA we would try to use a different linker. A potential reason for the loss of function can be too less space between the proteins, leading to misfolding or aggregation. This can be solved with a longer and more rigid linker[11]. Another reason could be that sfGFP cannot be folded fast enough before secretion into the matrix with TasA. This can also be adapted by changes in the linker sequence[12]. We would try to use amino acids that slow down the translation of the fusion protein, for example proline. By testing different protein architectures, we can learn more about the TasA-fusion platform in general, helping other researchers who also want to use this technology.
If the biofilm we grow seems to be less stable and we are able to detect fluorescence in our cells the export of TasA-sfGFP into the biofilm matrix is detracted by the fusion of TasA to sfGFP. In this case we would redesign our fusion protein especially regarding to the used linker as well as checking the secretion sequence for the transport of TasA.
A less stable biofilm combined with lack of any fluorescence could be explained with an unsuccessful transformation of our B. subtilis cells, a mistake in the sequence of our fusion protein or a mistake in our plasmid. Promoter, ribosome binding site or selection marker could be defective. Further steps would be the repetition of the described approach with TasA-CueO, TasA-CotA and TasA-EreB fusion proteins. The successful display of these enzymes in the biofilm matrix could be verified using our flow chamber and the ABTS assay. We already modeled our new composite part, the TasA-EreB fusion protein (BBa_K3429013) and the TasA-CotA fusion protein, via Molecular-Dynamic simulation. Furthermore, for the implementation of our biofilm the gene for our fusion proteins should be integrated into the genome of the B. subtilis strain via homologous recombination using integrative vectors[13].
Bacterial On-/Off-switch - How to contain B-TOX
As already previously described, this year's iGEM project is a solely theoretical project due to the COVID-19 pandemic. Therefore, the experimental setup was only theoretically designed, and we elaborate some expected results based on literature references.
The Cre recombinase allows the inversion of the DNA region between its two head-to-head alignedrecognition sites lox 66& 71, and was previously used by the iGEM team of Tuebingen 2015[14]. For our kill switch, we use the Cre recombinase to inverse the promoter region controlling the expression of an essential gene in B. subtilis. The expression of the genomically integrated Cre recombinase cassette is controlled by the xylose-inducible promoter PXylA. As the recombinase has a very important function in our kill switch, we first aim to test the recombinase activity and the correctness of our cassette design.
Assaying single inversion of our promoter region by Cre recombinase:
For testing the Cre recombinase activity in B. subtilis, we use a plasmid designed by iGEM Team Paris 2007 with the lox sites 66 & 71 aligned head-to-head [15]. Between the two recombinase sites the constitutive promoter Pveg is located on the antisense strand, but if the DNA region is inversed, Pveg controls the expression of a reporter gene for quantification of protein production, e.g.green fluorescent protein (GFP). Figure 6 shows the schematic layout of the cassette for testing Cre recombinase activity.
Figure 6: Schematic representation of the designed GFP expression cassette to assay recombinase activity and genomically integrated the recombinase expression cassette. GFP expression is under the control of a upstream DNA region(red) flanked by the two recombinase sites (lox 66 and 71, grey arrows) aligned head-to-head. The sense strand of the DNA region controlling GFP is empty, but the antisense strand encodes the constitutive promoter Pveg. The addition of xylose induces the expression of the Cre recombinase. As a result, the DNA region is inversed and the constitutive promoter Pveg starts expressing GFP as a quantifiable signal.
The cassette is cloned into a replicative plasmid with a selection marker, and the plasmid is transformed into a B. subtilis strain containing a genomically integrated Cre recombinase expression cassette. As a control, we use the same B. subtilis strain not harboring the plasmid in the following experiment. The test culture and the control culture are supplemented with xylose in the exponential growth phase to induce Cre recombinase expression. If the Cre recombinase is active and the system correctly designed, Cre recombinase inverses the promoter region on the plasmid a single time. Consequently, the reporter gene encoding GFP is expressed by the constitutive promoter Pveg and the GFP signal can be qualitatively determined by visual observation or quantitatively by fluorescence spectroscopy measurement. In the control culture no GFP expression is expected and it can therefore be used as negative control for GFP quantification. This experimental setup additionally allows the testing of necessary inducer concentrations to achieve sufficient expression levels and enables the investigation of further parameters to optimize the system, e.g. media composition or cell density.
Assaying multiple inversion of our promoter region by Cre recombinase:
In the literature, the potential double inversion between the lox sites is, to our knowledge, not described. Among others, the iGEM Team Paris 2007 characterized the lox site 66 and 71 as sites for "irreversible" rotation[15]. Nevertheless, we want to assure that double inversions are not possible. For this test the cassette described in figure 6 needs to be slightly modified by fusing a C-terminal SsrA degradation tag to the reporter protein[16]. This is necessary as GFP has a half-life period which is too long to be used within our developed assay, but the fused degradation tag leads to a faster degradation of the tagged protein in the cell[16,17]. The procedure is initially identical to the procedure described above for assaying single turns by Cre recombinase. The plasmid is transformed, the cells are allowed to grow and the expression of the Cre recombinase is induced by adding xylose in the exponential growth phase. As already described, a GFP signal is expected, which indicates a single inversion of the cassette. The xylose-containing media is subsequently exchanged to fresh media with or without supplemented xylose. In the absence of xylose, Cre recombinase expression stops and recombinase activity shrinks over time. Consequently, in the absence of xylose another inversion is less likely due low recombinase activity. In the presence of xylose, the Cre recombinase is further on expressed and a second inversion may occur, if Cre recombinase is capable of its catalysis. Consequently, if the Cre recombinase is able to inverse the promoter more than once, the GFP expression in presence of xylose is expected to shrink over time. Addition of the degradation tag shall help to decrease the incubation times in this assay.
This setup only allows a qualitative assay. For a quantitative measurement, the sequencing of single colonies is likely inevitable[18].
Assaying functionality of the kill switch design:
When eventually testing our final kill switch design, it is important to quantify the deadliness of our kill switch. We have to ensure that our kill switch does what it is supposed to do: Killing our genetically modified Bacillus subtilis cells upon leaving the biofilm environment. It is well known that, some antibiotics are able to kill cells by blocking the activity of protein biosynthesis or ribosomal proteins, but we don't know if this also applies to our kill switch model[19,20,21].
For this purpose, we compare the growth behavior of a B. subtilis strain harboring the genomically integrated kill switch (figure 7) with a control strain not harboring the kill switch cassette.
Figure 7: Schematic representation of the designed final kill switch cassette. RpsB expression is under the control of a upstream DNA region (red) flanked by the two recombinase sites (lox 66 and 71, grey arrows) aligned head-to-head. The sense strand encodes the constitutive promoter Pveg and the antisense strand encodes the inducible PdegQ. The addition of xylose induces the expression of the Cre recombinase. As a result, the DNA region is inversed and the inducible PdegQ starts expressing rpsB.
Both cultures are grown to a defined cell density until the xylose is added to induce expression the Cre recombinase cassette. This will cause the inversion of the kill switch cassette, as described above and the expression of rpsB will then be under the control of the cell density sensitive degQ promoter PdegQ. (figure 7 below ) Both cultures are then strongly diluted with fresh medium and the cells are incubate to grow. The growth behaviour of both cultures can be determined by measuring the cells absorbance at 600 nm (OD600) at defined time points. In addition, we will perform live/dead staining with SYTO 9 or propidium iodide stain to compare the ratio of dead to alive bacteria in the kill switch-harboring strain and the control strain[22]. We expect that the growth curves of the two strains will differ significantly, if the kill switch design is functional. The strain with the established kill switch should not grow after dilution as the quorum sensing signal is not strong enough to induce expression of rpsB. Ideally, the majority of the cells of this strain are dead to confirm high deadliness of the kill switch mechanism. Using the experimental setup, it is also possible to determine the necessary cell density to activate the PdegQ by diluting the bacteria in various ratios. Cultures not passing the threshold will show bacterial growth similar to the control. However, the functionality of the degQ promoter can also be verified by another method.
Assaying the degQ promoter behavior :
For this alternative assay, the kill switch-harboring cells are grown in separate test flasks up to a defined cell density and recombinase expression is induced. We hypothesize that if the cell density is not high enough to trigger a quorum sensing signal, the induced inversion of the promotor region leads to the death of the cells. If the cells however continue growing, the density during induction is high enough. As positive growth control a B. subtilis strain not harboring the kill switch is tested. Cell viability and consequently the expression of the essential gene can be determined by live/dead staining, as already described above.
Assaying the Pveg promoter strength :
In our kill switch design, we use a different constitutive promoter Pveg than the naturally encoded promoter controlling rpsB expression in B. subtilis. Consequently, it is possible that the used promoter is stronger than the original and RpsB is overexpressed. We want to test whether 1) the rpsB expression controlled by Pvegdiffers from the original expression level, and 2) whether a shifted level has an influence on the B. subtilis fitness. For this purpose, we grow our B. subtilis strain haboring the established kill switch cassette and a strain not harboring the cassette to a certain cell density and then compare the expression level of RpsB by reverse transcriptase qPCR[23]. A qPCR assay templated by the reverse transcripted rpsB-encoding mRNA allow the comparison between the two used promoters and the quantification of RNA levels by measuring a calibration curve. In addition, we compare the growth curves of the two cell cultures and perform live/dead staining assays to investigate the impact of Pveg controlling rpsB expression. This allows us to determine whether the constitutive promoter used, has an influence on the growth or viability of the cells. If the intended Pvegpromoter has a negative influence, two solutions are possible:
The first possibility is to use an inducible promoter system of which the optimal inducer concentrations needs to be determined by titrating the inducer concentration[24]. The second possibility is to use a weaker or stronger constitutive promoter from a library or various ribosome binding site strengths[25]. In both cases, the same experimental setup as used for the investigation on Pveg is feasible to verify success of measures taken.
Our kill switch design contains a second promoter, the quorum sensing promoter PdegQ, which may also exhibit a different expression strength than the original rspB-controlling promoter. In the case of PdegQ the expression levels is depend on the cell density and thus on the growth phase. This complicates the previously described assay to investigate Pveg impact on rspB expression. Nevertheless, the same assay can be used but has to be carefully planned with particular attention on various cell density conditions in the samples assayed.
With the MATLAB Simbiology toolbox, we created an in silicomodel of our kill switch. The model relies on ordinary differential equations, which helped us to understand the kill switch even further.
