Construct

Our goal is to engineer a genetic circuit within Pseudomonas sp. IsoF that allows the overproduction of biofilm only when the bacteria are in proximity to plant roots. To achieve this targeted biofilm production we integrated a sensing system that reacts to the plant exudate xylose. To prevent the spread of our engineered bacteria we incorporated a kill switch using a toxin-antitoxin system in our construct. Additionally, we built a positive feedback loop which further enhances the biofilm production. This feedback loop also serves as a regulatory switch that activates the kill switch as soon as the bacteria sense xylose. This way, our positive feedback loop is connected to both biofilm overproduction and the activation of the kill switch, ensuring the controlled bacterial activity. Thus, our construct consists of four main components and, accordingly, we have divided the engineering description into those four parts:

We chose Golden Gate Assembly for our cloning method. The iGEM distribution kit supplied us with basic parts, such as constitutive promoters, different ribosome binding sites (RBS), GFP and a terminator (see Parts for more details). We also utilized several of the provided backbones: pJUMP28-1A, pJUMP23-1A,pJUMP28-1B, pJUMP29-1C, pJUMP29-1D, pJUMP43-2A and pJUMP45-2A.

Additionally we received a level 0 backbone, the pBP, from this year’s iDEC team Evolution Suisse from the University of Zurich and ETH Zurich, for which we are very grateful. We researched all the gene sequences, attached the appropriate adaptors and ordered them through Twist.

Figure 1: This illustration represents our full construct to enhance biofilm formation only when the bacteria are in proximity to plant roots. Created with BioRender.com

Two state system: Connection of kill switch to feedback loop

By connecting the kill switch, feedback loop and the xylose-sensing module, we created a two-state system, in which the feedback loop acts as a switch. In the first state, before the plant gets inoculated with our engineered bacteria, neither biofilm production is increased nor the kill switch is active.

When the bacteria encounters plant roots, for example after we inoculate the plant with our engineered bacteria, xylose triggers the feedback loop. This initiates the overproduction of biofilm components and simultaneously the production of toxin.

Due to the self-activation and the self-amplifying characteristic of the feedback loop, it enhances biofilm production even further and also leads to a constant production of the toxin. This means that the feedback loop acts as a switch from the first state to the second state. When the feedback loop and therefore the second state is activated, the toxin production is sustained. However, the antitoxin is produced only in the presence of xylose and therefore only when the bacteria are close to the root. If the bacteria move away from the root and xylose is no longer sensed, toxin production continues without the neutralizing antitoxin, leading to the death of the bacteria.

Sensing

Since we aim for targeted biofilm production, our sensing system should be based on an interaction between plants and bacteria. We need a molecule, which is secreted by plants, recognized by bacteria and can simultaneously activate their gene expression. We decided on using a plant exudate from tomato plants as an inducer. After a lot of research we found a sensing mechanism, originally developed for Pseudomonas putida KT2440, which relies on the detection of xylose 1. Xylose is a sugar commonly exuded by tomato plants, as well as the roots of other plant species 5. Given the prevalence of xylose in root exudates, our sensing system and therefore our entire construct is broadly applicable across a wide range of plants.

In the presence of xylose, the XutR (Xylose-binding protein) binds to the sugar and forms a complex. This complex then binds to the XutR operator, which then activates the xylose-induced promoter PXut 1.

Cycle 1

Design and Build

Due to a misunderstanding when reading the paper, we thought the XutR complex would bind to the PXut directly. To test the functionality of the PXut, we therefore assembled the level 2 testing plasmid pT12Rhyz01 (see Parts for more details) without the operator sequence after the PXut. pT12Rhyz01 consists of pT11Rhyz01, pT11Rhyz02 and two linker transcription units. The xylose-binding protein XutR is constitutively expressed using the promoter BBa_J23119 on the first transcription unit, since we don’t want the promoter’s activation to be limited due to the insufficient expression of XutR but rather due to the presence or absence of xylose. The second transcription unit encodes the xylose-induced promoter (PXut) and a GFP (without the operator sequence). We introduced the pT12Rhyz01 plasmid into E. coli SY327 and then transferred pT12Rhyz01 into Pseudomonas sp. IsoF using Triparental Conjugation (see Experiments for more details).

Test

Since the xylose-induced promoter is specifically designed for P. putida KT2440, we decided to test the system exclusively in P. sp. IsoF without testing it first in E. coli SY327. We performed a xylose sensing assay to determine whether PXut can be induced by XutR in the presence of xylose, and to identify the concentration of xylose required to trigger promoter activation. The following xylose concentrations were tested: 0.01mM, 0.1mM, 0.2mM, 0.5mM, 1mM, 2mM, 5mM, 10mM and 20mM.

Overnight cultures were adjusted to an OD of 0.01 and 180µl of each adjusted culture were added to a 96-well-plate. They were induced with the different xylose concentrations and kept for 8 hours at room temperature. Each strain was tested in triplicate. The GFP expression was then measured every hour for a total of 15 hours. This sensing assay (see Experiments for more details) failed, there was no increase in GFP when higher xylose concentrations were present.

Learn

After re-reading the paper1 and re-evaluating Robert M. Q. Shanks (Co-Author of the paper) input, given when we contacted him in spring regarding his paper, we realized that the Xylose-XutR complex doesn’t bind directly to the PXut. Instead, it binds to a XutR operator sequence located between PXut and the ribosome binding site. Therefore we were missing a sequence between our xylose induced promoter and RBS.

Unfortunately the PCR-amplification and blunt-end ligation were unsuccessful in the tested colonies. As a result, we opted to implement the operator sequence using oligo annealing followed by digestion and ligation. We ordered the operator sequence as two oligonucleotides, with SalI and SphI recognition sites as overhangs. The sequences were annealed according to the oligo annealing protocol. Next, we designed primers specific to the sensing plasmid lacking the operator sequence (pT12Rhyz01). The forward primer was designed to overlap with the RBS on the plasmid, while also incorporating the SphI recognition site. The reverse primer overlapped with the PXut promoter on the plasmid and included the SalI recognition site. After amplifying the plasmid backbone using these primers, we performed digestion and ligation with the annealed oligos and the amplified backbone.

The newly assembled plasmid was transformed into E. coli SY327 and then introduced into P. sp. IsoF using Triparental Conjugation.

Figure 3: Plasmid map of p12Rhyz01

Test

Unfortunately the oligo annealing and digestion ligation failed, due to unspecific binding of the primer. For future experiments we will therefore repeat this step and then conduct the second sensing assay.

Diguanylate cyclase

Design and Build

We targeted the diguanylate cyclases (DGCs) by introducing a modified DGC into P. sp. IsoF. DGCs are either single-domain or multi-domain proteins that catalyze the production of c-di-GMP by cyclization of two guanosine triphosphate (GTP) molecules to one c-di-GMP molecule9. Various DGCs contain a conserved GG(D)EF domain and within this domain there is a so-called I-site (Inhibitory site) 8.

The I-site is an allosteric binding site specific for c-di-GMP. When c-di-GMP binds there, it reduces the enzyme’s activity and thereby prevents further production of c-di-GMP. C-di-GMP thus acts as a non-competitive inhibitor for the DGC. This negative feedback mechanism serves as a self-regulatory process of the intracellular c-di-GMP concentration. Specifically targeting this inhibitory site, or rather preventing c-di-GMP from binding to the I-site, may result in an increase in the enzyme’s activity even when c-di-GMP concentrations are already high and therefore a further increase in the c-di-GMP levels within the cells. Altering the inhibitory site and thus preventing c-di-GMP from binding to it has been successfully done in the DGC from Pseudomonas aeruginosa, WspR4.

Nabanita De. et al.4 discovered that c-di-GMP binds to two arginine residues (R242 and R198) within the GGDEF domain of the WspR diguanylate cyclase. They created two site-directed mutants of the enzyme by substituting one of the arginine residues by alanine. The purified enzymes showed higher activity and no detectable c-di-GMP was bound to the I-site, indicating that the binding of c-di-GMP and therefore inhibition of the enzyme was disrupted in both mutations 4. Since both mutated DGCs showed an increased activity we decided to focus on testing one of these mutations, specifically the R242A variant. To compare it to the wildtype sequence, we ordered the WspR DGC in two variants: one retaining the wildtype sequence and the other modified, so that the arginine at position 242 is replaced with alanine.

Although this strategy has been applied to DGCs in other bacterial species, such as the PleD DGC of Caulobacter crescentus3, it has never been tried on DGCs native to Pseudomonas sp. IsoF. Therefore we decided to explore this approach in one of the DGCs naturally occurring in P. sp. IsoF. Hailing Nie et al. conducted a comparative analysis of various DGCs and PDEs from Pseudomonas putida KT2440 with the main aim of understanding their role in biofilm formation and swimming motility2. P. putida KT2440, a well-established model organism with similarities to P. sp. IsoF, served as a reference point for our project. In their research, the DGC PP_1494 in P. putida KT2440 showed the highest increase in biofilm formation when overexpressed in a mutant. After examining a gene bank for P. sp. IsoF, we identified its homologous DNA sequence, PisoF_00565 and chose to focus on this DGC. As mentioned, this I-site is typically conserved across different DGCs with GGDEF domains and thus we assume this is also true for the PisoF_00565 DGC of P. sp. IsoF. Despite the conservation of the GGDEF domain and the I-site3 across different DGCs, the exact sequence and structure of these regions vary, which presented us with a challenge, especially since there is not a lot of data on P. sp. IsoF.

Nabanita De. et al. performed a sequence alignment of the GGDEF domains of WspR proteins across various Pseudomonas species to identify the locations of the I-site within these sequences. In both the WspR and PleD diguanylate cyclases, c-di-GMP binds to an arginine residue. When this arginine is substituted with alanine, it disrupts the binding site of c-di-GMP3.

Figure 6: Sequence Aligment. Created with BioRender.com

The initial conjugation attempt was unsuccessful and after thorough research, we discovered that our plasmids are not replicative in P. sp. IsoF due to an incompatible origin of replication. Instead of re-cloning them into a replicative plasmid using Golden Gate assembly, we chose to perform a digestion ligation into the pBBRMCS5 plasmid, which was provided by the lab. The pBBRMCS5 plasmid is commonly used for gene expression in P. sp. IsoF in the lab and therefore replicative in P. sp. IsoF.

To begin, we designed primers that would partially bind to our transcription units on the plasmids while also incorporating SacI and SpeI recognition sites on each end. We then carried out a PCR to amplify the transcription unit from the level 1 plasmid and simultaneously extended it with the SacI and SpeI sites for later implementation into the multiple cloning site of the pBBRMCS5 plasmid, which contains the same recognition sites in their multiple cloning site. We confirmed the amplification by checking the amplified sequence length on a gel. After DNA purification, we proceeded with a digestion of both the backbone and transcription units at 37 °C for one hour, followed by ligation. Alongside the five ligation reactions we included two controls where we replaced the T4 DNA ligase with water. After transformation, the controls showed numerous colonies, suggesting that our digestion was inefficient. A colony PCR on three colonies per construct revealed that the transcription unit had not been successfully inserted. We repeated the experiment, extending the digestion by 30 minutes. The controls still showed more colonies than expected but colony PCR on 16 colonies per construct confirmed that the DGC transcription unit had been successfully integrated into the pBBRMCS5 plasmid in at least one of the tested colonies per construct.

These five newly cloned plasmids pBBR pT21Rhyz01, pBBR pT21Rhyz02,pBBR pT21Rhyz03,pBBR pT21Rhyz04 and pBBR pT21Rhyz05 were then conjugated into P. sp. IsoF using triparental conjugation.

Figure 7: Colony PCR revealed that the digestion ligation into the pBBR1MCS5 backbone of the transcription units containing the different DGCs was successful.

Test

To compare the different enzymes and their mutants, we measured the c-di-GMP concentration of the five different P. sp. IsoF strains containing the respective testing plasmid, using the c-di-GMP assay kit from Lucerna Technologies.

Preliminary tests were conducted to identify the optimal conditions for the assay. After adjusting the protocol based on the preliminary results, the c-di-GMP assay was performed three times (see Experiments for details).

The overnight cultures were adjusted to an OD of 0.03, then induced by adding 125ul of 40% Rhamnose to achieve a final concentration of 1% in a 5ml tube. This led to the expression of the enzyme. The cultures were incubated for 5 hours and then washed with NaCl solution to eliminate background interference from the LB medium. The absorbance of each overnight culture was measured with a spectrophotometer and adjusted to an optical density (OD) of 0.25. A black 96-well plate was set up according to the c-di-GMP assay protocol. As negative controls, we used P. sp. IsoF dCas9 (contains no plasmid but dCas9 is implemented on genome), P. sp. IsoF pBBR1MCS (P. sp. IsoF containing the empty plasmid) and as a positive control we used P. sp. IsoF pBBR1MCS YedQ (constitutively overexpressed DGC). Into each well, 50ul of each of the adjusted and diluted cultures, were added. Each strain was tested in triplicate (3 replicates per strain). The plate was incubated in the dark for 14 hours. Then the fluorescence was measured with a plate reader, with the first excitation set at 469nm and emission at 501nm, followed by a second measurement with excitation set at 482nm and emission at 505nm.

After the c-di-GMP assay, we performed a biofilm staining assay (see Experiments for more details). A square plate containing 50ml LB medium with agar and 1.25ml of a Congo-Red derived dye (2mg/ml) was prepared and dried. This dye stains polysaccharides and therefore the biofilm produced by our P. sp. IsoF strains. The cultures were adjusted to an OD of 1 and 10ul of each adjusted strain were plated onto the prepared plate. The plate was incubated at 30°C for two days.

Additionally we conducted a swimming motility assay, to assess whether the introduction of a DGC would affect the bacterial motility. Elevated levels of c-di-GMP are associated with reduced motility, which in turn promotes biofilm formation, since a lower motility is beneficial for forming a robust biofilm2.

Figure 8:Plasmid map of pBBR pT21Rhyz01

Figure 9: Plasmid map of pBBR pT21Rhyz02

Figure 10: Plasmid map of pBBR pT21Rhyz03

Figure 11: Plasmid map of pBBR pT21Rhyz04

Figure 12: Plasmid map of pBBR pT21Rhyz05

Previous Next

Learn

The c-di-GMP assay, the biofilm staining assay and the motility assay show that introducing certain diguanylate cyclases (DGCs) can significantly increase intracellular c-di-GMP levels, promote polysaccharide production and reduce bacterial motility. These three results all indicate that the different P. sp. IsoF strains containing different DGCs do indeed produce and form a stronger biofilm.

Our mutated DGC PisoF_00565 R196A resulted in a significantly higher concentration of c-di-GMP compared the wildtype sequence of PisoF_00565. This indicates that we successfully modified the enzyme’s I-site, which usually reduces enzyme activity when c-di-GMP binds. By altering the sequence of the DGC we prevented c-di-GMP from binding to the negative allosteric site, which effectively enhanced the enzyme’s activity. This has never been done and showed before for any DGC native to P. sp. IsoF. For further details on the results, please see the Results page.

Given that the mutated WspR diguanylate cyclase (DGC WspR R242A), native to P. aeruginosa, resulted in the greatest increase in c-di-GMP levels and polysaccharide production, we chose to incorporate this mutated WspR DGC into our final plasmid construct.

Additionally, we discovered that the origins of replication OriV9 (pBBR322), p15A and pUC are not replicative in P. sp. IsoF. This insight needs to be taken into consideration when designing and introducing other plasmids into P. sp. IsoF.

Unlike the standard Cas9 protein, which cleaves DNA, the CRISPRi system utilizes a catalytically inactive variant known as dead Cas9 (dCas9). DCas9 can still be guided to a specific DNA sequence by a guide RNA, but it can no longer cut the DNA. Instead, it binds to the target sequence and physically blocks the RNA polymerase and therefore prevents the transcription of the targeted gene. This system thus allows knockdown of a gene encoding for a Phosphodiesterase.

We again utilized the comparative analyses of Hailing Nie et al. to identify a suitable PDE for knockdown. Among the P. putida KT2440 mutants they created, the one lacking PDE_0914 showed the highest increase in biofilm formation. After examining the gene bank for Pseudomonas sp. IsoF, we identified the homologous gene in P. sp. IsoF, PisoF_02645. We chose to target this PDE for our knockdown.

Since the research group we were working with already has a well-established functioning CRISPRi system that is not based on Golden Gate assembly but on PCR amplification and blunt-end ligation, we decided to initially test the knockdown of our selected PDE using their cloning method. After testing, we will then clone it into our final plasmid using Golden Gate assembly.

The CRISPRi system requires two key components: the modified dCas9 protein and a guide RNA (sgRNA). We were provided with a Pseudomonas sp. IsoF strain, which already contains the gene encoding dCas9 integrated on its chromosome. We will refer to this strain as P. sp. IsoF dCas9. The dCas9 gene sequence is under the control of a rhamnose-inducible promoter. The CRISPRi system is therefore rhamnose inducible, which allows us to control the expression of the dCas9 and our knockdown.

The sgRNA consists of two main parts: a tracr-cr sequence and a guide sequence. The tracr-cr sequence is 35 base pairs long and forms a hairpin structure, which is essential for the dCas9 to remain bound to the target DNA sequence and stops the DNA polymerase to function. The sequence for this tracr-cr is: GTTTTAGAGCTAGAAATAGCAAGTTAAAATAAGGC.

For an efficient knock-down the non-template strand should be targeted, meaning that the guide RNA should have the same sequence as the template strand. The guide RNA itself should be 20-22 nucleotides long and again consists of two parts: the first 10 nucleotides should be unique within the Pseudomonas sp. IsoF genome to avoid off-target effects. The remaining sequence (10-12 nucleotides) does not necessarily need to be unique. Additionally, those 10 unique nucleotides must be located adjacent to a PAM motif (Protospacer Adjacent Motif), which in our system is CCN on the non-template strand. The PAM sequence is essential because it is recognized by the Cas nucleases and dictates where the enzyme binds.

Figure 14: Our sgRNA to knockdown PDE PisoF_02645. Created with BioRender.com

This sequence was then synthesized as a primer by Microsynth. Following the CRISPR interference protocol (see Experiments detailed protocol), the sgRNA was cloned into the plasmid using PCR amplification and blunt-end ligation. We performed PCR amplification to insert our sgRNA into the pSCB2-sgRNA-Gm plasmid. This plasmid already contains the tracr-crRNA sequence, which allows the tracr-crRNA part of our sgRNA to anneal directly to the template plasmid. The guide RNA sequence, however, is not yet present on the plasmid. Since the guide RNA is physically linked to the tracr-crRNA, the DNA polymerase extends the sequence by adding nucleotides during the PCR reaction, resulting in a plasmid that includes both the guide RNA and the tracr-crRNA. To eliminate the original template plasmid (pSCB2-sgRNA-Gm) that does not contain our guide RNA, we treated the reaction mixture with DpnI. DpnI is a restriction enzyme that specifically digests GATC motifs where adenine residues are methylated. Since only the template plasmid is methylated, DpnI selectively digests it, resulting in a product which only contains our desired plasmid, called p1DRhyz04.

After purifying the PCR product, we proceeded with ligation and transformed the p1DRhyz04 plasmid into E. coli SY327. We then performed colony PCR and sent the amplified sequence for sequencing to ensure that the correct plasmid was introduced into our E. coli strain. Finally, we performed a triparental conjugation to introduce p1DRhyz04 into P. sp. IsoF dCas9.

Figure 15: Plasmid map of p1DRhyz04

Test

Since the gene knockdown is only effective in P. sp. IsoF, we were only able to test our system in P. sp. IsoF dCas9 and could not perform parallel experiments in E. coli SY327. To evaluate our gene knockdown, we initially conducted a c-di-GMP assay, following the same procedure as the assay for our different DGCs and also using the same controls. We induced the overnight cultures with 125ul of 40% rhamnose after adjusting them to an OD of 0.02. Then we washed the cells, let them grow and adjusted them again to an OD of 0.25. A black 96-well plate was set up according to the c-di-GMP assay protocol. We added 50ul of the adjusted cultures to each well and each strain was tested in triplicate. The plate was then incubated in the dark for one hour. Then the fluorescence was measured using a plate reader and the first measurement was taken with excitation at 469nm and emission at 501nm, the second measurement with excitation set at 482nm and emission at 505nm.

After the c-di-GMP assay we conducted a biofilm staining assay (see Experiments for more details), again following the same procedure as the biofilm staining assay for our different DGCs. The overnight culture of each strain was induced with rhamnose and adjusted to an OD of one. 10ul of each adjusted strain was plated onto the prepared dyed plate. The plate was incubated at 30°C for two days. The absorbance of each strain was then measured using a fluorescence microscope.

Additionally we conducted a swimming motility assay, to assess whether the knockdown of a PDE would affect the bacterial motility. Elevated levels of c-di-GMP are associated with reduced motility, which in turn promotes biofilm formation, since a lower motility is beneficial for forming a robust biofilm2.

Learn

The biofilm staining assay confirms that PisoF_02645 is knocked down when implementing our sgRNA and dCas9 into P. sp. IsoF. This knockdown then leads to a higher c-di-GMP concentration in the cell and therefore an increased production of polysaccharides. The motility of the P. sp. IsoF strain with the knocked-down PDE was also reduced. These results all indicate that this strain produces a stronger biofilm.

Since our CRISPRi system works as intended we proceeded to order the transcription unit of our sgRNA (T7 promoter and sgRNA) with the appropriate overhangs. We then cloned it directly into one of our level 2 plasmids.

To verify the functionality of this feedback loop and confirm that higher concentrations of T7 RNAP lead to increased gene expression, we designed three level 2 testing plasmids: pT32Rhyz01, pT32Rhyz02 and pT32Rhyz03 (see "Parts" for more details) using the Golden Gate assembly.

The pT32Rhyz01 contains two transcription units (and two linker transcription units). The first one is a constitutive promoter and a T7 RNAP (pT32Rhyz01). The second one is a T7-like promoter and a GFP (pT31Rhyz02). The aim of this plasmid is to test whether the T7-like promoter is effectively activated by the T7 RNAP.

The pT32Rhyz02 is designed to test the impact of T7 RNAP without a feedback loop. It includes three transcription units: a constitutive promoter drives the expression of T7 RNAP (pT31Rhyz01), a T7-like promoter controls the expression of GFP (pT31Rhyz02). Additionally, a separate T7 promoter controls the expression of RFP, a gene unrelated to the feedback loop (pT31Rhyz03).

The pT32Rhyz03 simulates the feedback loop. It also contains three transcription units: a constitutive promoter drives T7 RNAP (pT31Rhyz01), a T7-like promoter leads to the expression of an additional T7 RNAP and a T7-like promoter controlling GFP. The aim is to compare the GFP expression of pT32Rhyz02 and pT32Rhyz03 to identify whether the RNA polymerase is suitable as a self-amplifying feedback molecule.

Figure 17: Plasmid map of pT32Rhyz01

Figure 18: Plasmid map of pT32Rhyz02

Figure 19: Plasmid map of pT32Rhyz03

Previous Next

Test

We encountered some challenges during the cloning of our level 2 plasmids. Initially, all the colonies we obtained were green, suggesting that the GFP either hadn’t been properly excised or had re-ligated with the backbone. To address this issue, we increased the insert-to-backbone ratio to 5:1 (instead of 3:1). However, after another unsuccessful attempt, we decided to pre-digest the backbone and perform gel extraction to remove the GFP before adding the inserts. This adjustment resulted in all white colonies, but due to time constraints, we were unable to test the feedback loop unfortunately.

Kill Switch

Design and Build

To ensure that the bacteria die when they move away from the plant roots, we incorporated a kill switch into our construct using the gyrase-inhibiting Toxin/Antitoxin system CcdB/CcdA. The ccdB gene encodes the toxic protein CcdB, which inhibits the DNA gyrase and therefore leads to cell death. The ccdA gene encodes the antitoxin protein CcdA that neutralizes the toxin by forming a CcdA-CcdB complex and therefore protecting the cell10.

Our first goal was to determine the amount of CcdB required to effectively kill the cell. Therefore we designed three testing plasmids pT41Rhyz01, pT41Rhyz02 and pT41Rhyz03 (see Parts for more details) with Golden Gate Assembly and transformed into E. coli SY327. Each of these plasmids contains the toxin gene but with a different ribosome binding site (RBS) strength. pT41Rhyz01 has the weakest RBS leading to little toxin production, while pT41Rhyz03 has the strongest RBS leading to high toxin production. The expression of the toxin is controlled by an rhamnose promoter to ensure the toxin is produced only when we induce it with rhamnose.

The first toxin test (see Experiments for more details) in E. coli SY327 showed that the bacteria does not get killed by the toxin, indicating that E. coli SY327 is resistant to CcdB. Those three plasmids were then transformed into E. coli DH5alpha and the same toxin test was conducted. Again, the experiment failed and the bacteria weren’t killed. After in-depth research we found that E. coli SY327 as well as E. coli DH5alpha had a mutation in their gyrase and were therefore resistant to our toxin 6.

We identified three suitable E. coli strains6: E. coli HB101, E. coli MC1061 and E. coli CSH50. We prepared competent cells following the protocol (see Experiments protocol for more details) and transformed our plasmids into this new strain.

Since these three testing plasmids also contain the pJUMP28-1A backbone, we assumed that they would also not replicate in P. sp. IsoF. We therefore performed the same digestion ligation to insert them into the pBBRMCS5 plasmid, as we did with the DGC testing plasmids described earlier. The initial ligation attempt was unsuccessful but after prolonging the digestion the second attempt was successful.

Additionally we constructed pT41Rhyz04 with Golden Gate Assembly, which contains a rhamnose inducible promoter and the antitoxin gene ccdA. Level 2 testing plasmid pT42Rhyz01 was then assembled with pT41Rhyz04 and one of the three toxin testing plasmids to test whether the antitoxin CcdA is potent enough to rescue the cell and transformed into E. coli. The assembly was not successful. The toxin assays failed.

Since the bacterial growth was not inhibited, there is no reason for testing whether the antitoxin can block the toxin. Therefore, we have decided not to proceed with recloning.

Test

We performed the same toxin assay 1 (see Experiments for more details) for all different E. coli strains. The toxin assay consists of an OD measurement overnight. The overnight cultures were adjusted to an OD of 0.01 and induced with different rhamnose concentrations (0%, 1%, 1.5%). Then they were incubated for four hours. 195 ul of bacterial culture was added to a 96-well plate and the OD600 was measured overnight.

The same toxin test was done on P. sp. IsoF containing the three toxin testing plasmids on the pBBRMCS5 backbone. Additionally we performed toxin assay 2 on two strains: E. coli CSH50 and P. sp. IsoF.

Learn

For E. coli CSH50 we see a significant decrease in growth when inducing it with more rhamnose. This is not the case for P. sp. IsoF, the CFU remains consistent across all tested rhamnose concentrations. We can thus conclude that the toxin CcdB does inhibit the growth of E. coli CSH50 in this second toxin assay but it does not inhibit the growth of P. sp. IsoF. We don’t know why the first toxin test didn’t lead to the same result as the second toxin test for E. coli strains. Since our aim is to implement this Kill Switch into P. sp. IsoF, our main focus was on determining whether this toxin was effective in P. sp. IsoF. For this bacterial species, the result is clear. In both experiments there is no inhibition in growth for P. sp. IsoF. We can thus conclude that the toxin CcdB doesn’t effect the growth of P. sp. IsoF significantly. For the future, we would therefore research more on toxin antitoxin systems specifically designed for P. sp. IsoF and implement one of those into our genetic circuits.

Design of final construct plasmids

Based on our test results, we designed two final construct plasmids p21Rhyz02 and p22Rhyz02 with eight transcription units in total. We chose to implement the sgRNA, to knockdown phosphodiesterase PisoF_02645, and the diguanylate cyclase of Pseudomonas aeruginosa that is mutated at position 242 (WspR DGC R242A), since this mutation resulted in the highest intracellular c-di-GMP levels and highest polysaccharide production. The toxin assay indicated that a sufficient level of toxin is required to impact bacterial growth. Therefore, we incorporated the toxin with a strong RBS into our final construct.

Instead of implementing a linker transcription units to fill the second level 2 plasmid, we introduced a second DGC transcription unit (T7 promoter and WspR DGC R242A). This even further enhances biofilm formation when the feedback loop is induced.

To clone the two level 2 plasmids p21Rhyz02 and p22Rhyz02, we used pJUMP43-2A(sfGFP) and pJUMP45-2A backbones from the iGEM distribution kit, as they are replicative in P. sp. IsoF. Since we need all eight transcription units for our construct to work, we needed to introduce both plasmids into P. sp. IsoF. However both backbones carry a spectinomycin resistance, which posed a challenge in selecting bacteria that carried both plasmids, rather than just one. Therefore we performed Gibson Assembly to change the resistance marker on the pJUMP45-2A(sfGFP). We amplified the kanamycin resistance of the pJUMP29-1B(sfGFP) with primers overlapping our backbone and simultaneously amplified the backbone without the spectinomycin resistance with primers overlapping the kanamycin resistance gene. One of the plasmids that showed GFP expression and grew on kanamycin plates was sent for sequencing, but unfortunately, the sequence was incorrect. The sequence we received matched that of pJUMP29-1B, indicating that the template had been transformed into E. coli SY327. Due to time constraints, we did not repeat the Gibson Assembly and decided to focus on working on individual components rather than repeating the Gibson assembly to integrate our final level 2 plasmids into P. sp. IsoF. Had we chosen to repeat the Gibson assembly, we could have performed a DpnI treatment after amplification to eliminate the template and potentially improve the efficiency of the cloning.

We also encountered challenges while cloning the level 2 plasmids and were ultimately only able to clone one, p21Rhyz02. Unfortunately, due to time limitations we could neither introduce the final level 2 plasmids into P. isoF nor conduct any testing.

plasmids with backbone pBBR1 MCS5 (replicative in P. sp. IsoF)

Figure 26: Plasmid map of p21Rhyz02

Figure 27: Plasmid map of p22Rhyz02

Previous Next

Conclusions

Our goal was to engineer a genetic circuit within Pseudomonas sp. IsoF that allows the overproduction of biofilm only when the bacteria are in proximity to plant roots. Our construct consists of four components and we were able to successfully clone and implement following parts of the construct: