The accomplishment of our project has been a long process over the course of almost a year. We have thoroughly followed an experimental strategy to successfully reach our goals. The experiments we have performed are explained below:
The objective of our project was to create a more universal bacterial tracking system
than the one developed by Gregor et al. (2018). We wanted to:
The ilux operon was born from a mutated natural luminescence operon present in the bacteria Photorhabdus luminescens: the lux operon, and the addition of the frp gene from the Vibrio campbellii bacteria. These mutations were error-prone PCR induced according to Gregor et al.’s study in 2018, and are listed in the following table. The aim was to create a system of genes that produces more light.
| lux operon genes | Mutations to form ilux |
|---|---|
| luxA | K22E, T119A, S178A |
| luxB | S13P, V121A, N259D |
| luxC | N10T, N59D, E74D, S256P, M355T, N360D |
| luxD | |
| luxE | |
| frp | M213L, R242L, K256R |
Table 1 : Amino-acids mutations induced to the lux operon, by error-prone PCR, in order to create the improved version ilux.
The ilux operon is composed of 6 genes: iluxA, iluxB, iluxC, iluxD, iluxE and ilux-frp, that form an autonomous system producing bioluminescence thanks to the encoded luciferase enzymes. Each one of the ilux genes has a specific role, as you can see on the illustration above (Figure 1). According to Gregor et al.’s study in 2018, an FMN reductase is naturally already present in E.coli bacteria, making the frp gene not that essential for a durable luminescence production in E.coli. As for Dickeya solani, our proof of concept bacteria, the MAP laboratory has already used the lux operon without frp without any problem, and since ilux allows a bioluminescence production sevenfold more important than lux, we decided not to make it a priority to work with frp during our experiments. However, we think that adding frp to our tool would be an added value for future iGEM teams.
Initially, the ilux operon was available in a pGEX plasmid in an E.coli DH5α. The objective of this study was to adapt this ilux operon for use in the iGEM backbones. To assemble BioBricks for iGEM, it is necessary that EcoRI and XbaI restriction sites are placed upstream, and SpeI and PstI downstream of the operon. The ilux operon contains 3 PstI restriction sites, 1 EcoRI site and 1 XbaI site: these 5 sites need to be removed (Figure 2).
We attempted to remove the restriction sites in the iluxCDABE genes by following these steps: DNA extraction (1), PCR directed mutagenesis (11), agarose gel analysis with green gel (4), and gel purification (5). The restriction sites were indeed mutated, so removed, and 6 PCR fragments were obtained, according to Figure 2.A. However, we were not able to gather the fragments together with an overlap PCR, as the resulting product was not concentrated enough. We then tried a TEDA cloning to gather all the fragments together in the pWSK129 plasmid (a low copy plasmid), and also pGEX(-) but the restriction profiles of the obtained plasmids were both not correct (4). In another engineering cycle, we tried another method. As the problematic restriction sites were not present in the genes iluxA, B and E , we performed a classical PCR to reconstruct fiatluxABE in pSB1C3 (already in iGEM biobrick format) and pBAD18 (high-copy vector with an arabinose inducible promoter). After E.coli DH5α was transformed by both plasmids. Therefore, we obtained fiatluxABE without promoter in pSB1C3, and under the control of the arabinose inducible promoter pBAD in pBAD18. The part containing the iluxCD genes was synthesized by chemical synthesis (GeneCust), by removing directly the restrictions sites in the sequence. In the synthesis, we incorporated upstream iluxCD the BBa_J23117 promoter. It is a constitutive and weak promoter, so no inducer needs to be added.
Since mutations have been introduced in the ilux operon, the effect on bioluminescence had to be tested before reassembling fiatluxABE and fiatluxCD in the same plasmid. In order to check the bioluminescence, the two plasmids pACYDuet-1-fiatluxCD and pBAD18-fiatluxABE were inserted into E.coli DH5α by electrotransformation (8). After transformation, the bioluminescence of the transformants was well observed with ChemiDoc XRS+, Bio-Rad, indicating that the genes fiatluxABE and fiatluxCD are expressed and functional (Figure 3).
Besides, since fiatluxABE is under the control of the pBAD promoter, different arabinose concentrations were added to the culture to determine the link between inducer quantity and bioluminescence production. The mean of luminescence over the OD at 600nm ratio is shown on Figure 4:
This analysis shows that a low concentration of arabinose of 0.02 % is enough to induce bioluminescence production, but an arabinose concentration of 0.1% inhibits it. This would imply that high ilux gene transcription could result in an excessive energy cost. Therefore, we conclude that it is necessary to use a plasmid with a low-copy origin of replication in order to avoid depleting the bacterium of the energy needed to operate the luciferase. However, in any case, we decided to use a constitutive promoter to drive lux expression, since we cannot add an inducible substance during an in situ study.
The aim was then to assemble fiatluxABE and fiatluxCD parts, to check that the operon was still functional after being mutated and reassembled, and to put it in an appropriate backbone to be able to transfer our fiatlux operon to other strains, even non-transformable ones, making our tool usable for more potential bacteria. Besides, we wanted to assemble J23117-fiatluxCD and fiatluxABE together in a conjugative plasmid, possessing an origin of transfer. An assembly of the J23117-fiatluxCD and fiatluxABE was thus done by HIFI assembly (7) in pSEVA521 and pSEVA531, two conjugative vectors with RK2 and pBBR origin of replication respectively (low replication strength). These two plasmids were transformed (8) in E. coli DH5α, and cloning was verified by a restriction profile (6;2;4). The ilux assembly of fiatluxABE and fiatluxCD was successfully done in both pSEVA521 and pSEVA531 vectors (Figure 5).
| ilux operon genes | Mutations to form fiatlux |
|---|---|
| iluxA | |
| iluxB | |
| iluxC | A→G(A949A) |
| iluxD | A→G(L167L) |
| iluxE | |
| iluxfrp | G→C(L60L), A→G(A234A) |
| non coding DNA | 2547_2548insG, T2689A |
Table 2: Directed mutagenesis on the ilux operon. Mutations in nucleotides in order to create the new Biobrick fiatlux. Mutations are silencious in coding regions. The assembling of iluxCD and iluxABE by digestion and ligature into the pSEVA plasmids lead to a guanine insertion between the nucleotide 2547 and 2548 (from the A of the start codon of iluxC). This insertion is not in a coding region or a ribosome binding site, and thus is not supposed to change the functionality of the lux system.
Our fiatlux tool was designed in order to achieve a perfect balance between luminescence output and regular bacterial growth. It is critical that the presence of our plasmid does not disrupt the regular functioning of the transformed bacteria; otherwise, an imaging study using our technology would be inaccurate. A bacterial growth defect due to our fiatluxCDABE system could impact bacterial processes such as virulence. In order to determine the appropriate culture and storage conditions, it is also necessary to characterize how the transformed strains react to temperature changes after cold storage and various antibiotic concentrations.
As an initial stage, optimization and characterisation were carried out in E. coli DH5. We have created 2 plasmids that differ only in their origin of replication. pSEVA521-fiatluxCDABE carries out the RK2 replication origin whereas pSEVA531-fiatluxCDABE has the pBBR1 replication origin. The number of copies of these plasmids were not evaluated in our studies but based only on previous studies performed in E. coli (Jahn et al. 2016) that evaluated a copy number between 2 and 7 for RK2-based plasmid and between 5 to 10 from pBBR1 replication origin. The effect on E. coli of the fiatluxCDABE expression was evaluated by comparing the phenotypes of E. coli strains expressing fiatluxCDABE with those carrying the empty vector pSEVA521 or pSEVA531.
First, we analyzed the behavior of the fiatluxCDABE strains on solid LB media. We streaked E. coli strains on LB agar containing various concentrations of tetracycline. Because we observed that DH5α with pSEVA531-fiatluxCDABE streaked on a LBagar plate with tetracycline at 10 µg/ml had difficulties to grow, we decided to determine the effect of tetracycline on the strains containing pSEVA521-fiatluxCDABE, pSEVA531-fiatluxCDABE or the empty vectors. We also decided to evaluate their growth depending on storage conditions. Strains were directly streaked from the -80°C glycerol stock, from colonies stored at 4°C or at room temperature. The plates were then incubated at 37°C overnight. Results are presented in Figure 6.
Transformed E. coli DH5α were able to grow after being stored at -80°C or +4°C, except in the presence of a 10 µg/mL tetracycline medium. When stored at room temperature, transformed E. coli DH5α were able to grow at any tetracycline concentration. We also observed that DH5α transformed with pSEVA521-fiatlux grew better in a 2 and 5 µg/mL tetracycline LBagar medium. In contrast, DH5α strains containing the empty vectors pSEVA521 or pSEVA531 grow very well on LBagar with 10 µg/ml tetracycline, even when the inoculum is from bacteria stored at 4°C or -80°C. Taken together, these observations indicate that expression of fiatlux from pSEVA531 certainly delays the ability of the E.coli DH5α to resist an usual concentration of tetracycline since the use of lower concentrations of tetracycline improves growth. It is likely that the production of light by luciferase has an energy cost that would prevent metabolically inactive bacteria (as stored at 4°C or -80°C) from adapting to the presence of tetracycline.
From now on, a tetracycline concentration between 2 and 5 µg/ml should be used for E.coli DH5α transformed strains culture.
We also compared the growth over time (O.D.) of E. coli DH5a containing pSEVA521, pSEVA531, pSEVA521-fiatluxCDABE and pSEVA531-fiatluxCDABE, with a TECAN microplate reader. Luminescence was also measured over time. Bacteria were cultivated in LB at 37°C with shaking microplates. To note that we concentrated our bio modelisation analysis on Dickeya solani behavior and not on E.coli’s. Thus, all of our E.coli graphics and results are not statistically confirmed but based on observation, owing to time constraints.
In this graphic, only one replicate is shown among the 16 performed replicates, for each condition. The represented replicate is representative of the global results.
Concerning pSEVA521 transformed strains, fiatlux seems to slightly
affect the maximum O.D., but a statistical analysis would be necessary to confirm its
significance.
Concerning pSEVA531 transformed strains, pSEVA531-fiatlux transformed bacteria and empty
pSEVA531 transformed bacteria in Tet 0µg/mL seems to have similar maximum O.D., whereas a
growth lag seems to occur in Tet 10µg/mL transformed bacteria. Indeed, the beginning of
growth is visible starting from 7 000 secs for empty pSEVA531 transformed bacteria, and from
75 000 secs for pSEVA531-fiatlux transformed bacteria. Therefore, it seems that pSEVA521 is
a better vector to use in E.coli.
Then, the luminescence over growth ratio is analyzed to understand how fiatlux transformed E.coli bacteria behave, in function of the used vector (pSEVA531 or 521) and in function of the antibiotic concentration (tet 10, 5 or 0 µg/mL).
These graphics show that, whatever the tetracycline concentration is, E.coli DH5alpha transformed with fiatlux emits luminescence, and E.coli bacteria that do not have our fiatlux operon do not produce any light. It is interesting to note that, in a tet 0µg/mL growth medium, there is no selective pressure on the transformed bacterias anymore, but these bacterias were still shown to produce luminescence. Besides, for all antibiotics concentration, pSEVA521-fiatlux transformed bacteria’s behavior seems similar, whereas pSEVA531-fiatlux transformed bacteria’s behavior is irregular, hence not reproducible.
To note that the luminescence is comparable between pSEVA521-fiatlux transformed bacteria and pSEVA531-fiatlux transformed bacteria. Besides, by comparing both O.D and luminescence/O.D. ratio graphics, the maximum luminescence emission is observed at the end of the exponential growth phase, which confirms that luminescence is potentially dependent on the internal bacterial metabolism.
In the end, both liquid and solid media showed that the plasmid pSEVA521 seemed more suitable than pSEVA531 in E.coli DH5α strains. We believe that the reason could be the following: a higher replication rate could tire the bacteria and thus slow down its growth, which could be due to the potential toxic effect the high production of luminescence could have on the bacteria.
After transforming E. coli bacteria with our fiatlux operon, we transferred the plasmids into phytopathogenic bacteria Dickeya solani, to validate our proof of concept. After having characterized the stability of this new strain, we infected plants (chicory and Saintpaulia) in order to visualize the luminescence of the bacteria directly in situ, thus validating our proof of concept.
After successfully transforming E.coli DH5α bacteria with our plasmids and characterizing the tool, our next goal was to insert it in D. solani. We used the WT strain D. solani D s0432-1. The four plasmids used in the experiment are pSEVA521, pSEVA521-fiatlux, pSEVA531 and pSEVA531-fiatlux. The two empty vectors (pSEVA521 and pSEVA531) act as control and are used to compare the effects of the fiatlux operon in D.solani. We performed two methods to introduce our tool in D.solani D s0432-1: an electrotransformation, and a conjugation.
For the electrotransformation, we used Petri dishes with 5µg/mL of tetracycline to only select bacteria with the plasmid. As a result, we got a few colonies that grew on Tetracycline 5µg/mL, those transformed with pSEVA 531, and those transformed with pSEVA531-fiatlux. However, none of the colonies transformed with pSEVA531-fiatlux were luminescent. Although our results were not conclusives, we did not push our investigation further, because our second method of DNA transfer worked.
For the conjugation, we used an intermediate E. coli strain, MFDpir, because of several advantages it presented for the conjugation:
We thus used a transformation (8) to insert our four plasmids in E.coli MFDpir. After 48h of incubation at 37°C, some colonies grew on the Petri dishes (with 5µg/mL of tetracycline) and the luminescence of colonies was observed with a highly-sensitive camera.
At this point, it was possible to carry out the conjugation step (14), between E.coli MFDpir and D.solani, for the four plasmids. After 48h of incubation at 30°C, some colonies grew on the Petri dishes (with 5µg/mL of tetracycline) and the luminescence of the colonies was observed with a ChemiDoc XRS+, Bio-Rad highly-sensitive camera. For each conjugation, 7 colonies were re-isolated on solid LB.
In order to make sure the colonies we obtained after conjugation were D.solani, we performed a pectinase test (25). We spread a drop of our 7 clones (putting in liquid culture the day before) on a Petri dish containing 0.2% of dipecta, and after 24h, we revealed with copper acetate. According to the stains around the drops, all of our colonies are D.solani. The four plasmids were successfully transfered into D.solani, and the colonies were luminescent.
2.1. Growth and Luminescence of D. solani fiatluxCDABE Strains on Solid Media
Growth and luminescence of the Dickeya solani strains in function of tetracycline concentration in LB agar medium were evaluated as it was done previously with E. coli. The strains were stored at a temperature of -80°C, and plated on Petri dishes. Four conditions were tested, for each strain: 0, 3, 5 and 10µg/mL of tetracycline. The Petri dishes were incubated at 30°C. Photos of the Petri dishes were taken after 24h and 48h of incubation with a ChemiDoc XRS+, Bio-Rad highly-sensitive camera, to see the growth of colonies. In order to compare the behavior of our tool in D.solani with our results for E. coli, we did the same test in parallel with both D.solani and E. coli DH5α strains.
The bacteria containing the plasmid with fiatlux (pSEVA521-fiatlux and pSEVA531-fiatlux) and the bacteria with the empty vectors (pSEVA521 and pSEVA531) look similar (same shape and same size), for a same concentration of tetracycline. This means that the expression of our fiatlux operon from J23117 promoter in pSEVA521 and pSEVA531 does not affect the growth of Dickeya solani . This is a big difference from what was observed in E. coli DH5α.
Luminescence of D.solani (with pSEVA521-fiatlux and pSEVA531-fiatlux) is much stronger with a higher concentration of tetracycline. This is an unexpected result since fiatlux is under the control of a constitutive promoter. We suggest two hypothesis to explain this result:
As for E.coli DH5⍺, D.solani with the plasmid pSEVA531-fiatlux emits more luminescence than the strain with pSEVA521-fiatlux. pSEVA531-fiatlux being a higher copy number plasmid than pSEVA521-fiatlux, the luciferase is more produced and more light is emitted.
We did a second experiment in order to characterize our tool: the growth and luminescence of colonies on solid medium were observed every 30 min for 48h thanks to a NightShade LB 985 (Berthold) optical imaging system, equipped with a high resolution CCD camera. The experiment with D.solani and E.coli DH5alpha containing the plasmids (pSEVA521, pSEVA521-fiatlux, pSEVA531 and pSEVA531-fiatlux) on Petri dishes in LB medium with different concentrations of tetracycline (0, 5 and 10 µg/mL)
Video of D.solani and E.coli DH5alpha containing the plasmids (pSEVA521, pSEVA521-fiatlux, pSEVA531 and pSEVA531-fiatlux) on Petri dishes in LB medium with different concentrations of tetracycline (0, 5 and 10 µg/mL)
What we can observe in the time laps: Dickeya transformed with pSEVA531-fiatlux with LB media and tet 5µg/mL media start to emit luminescence before all the others.
We can clearly see the effect on antibiotic concentration on the time laps: for both LB medium and tetracycline 5µg/mL medium, luminescence appears earlier than with tetracycline 10 µg/mL medium. This may be due to the growth delay we observed for higher antibiotic concentrations. Luminescence reaches a peak, before fading. It is fading faster on LB medium than on medium with antibiotics. This can be explained by the selective pressure of tetracycline that encourages the bacteria to keep the plasmid. In order to quantify the speed of the loss of the plasmid, we will perform a stability test on D.solani with our four plasmids.
Luminescence evolves over time: for both D.solani and E. coli DH5alpha, the luminescence starts to increase until it reaches a maximum, before decreasing. For a same concentration of tetracycline and the same plasmids, D.solani emits the maximum of luminescence earlier than E. coli DH5⍺. These maxima of luminescence also reach higher levels.
This video allows us to observe the dynamics of bacterial growth on solid medium and the luminescence of colonies over time. We observed that the luminescence is emitted longer for isolated colonies than for the bacterial mat. After 48h, the bacterial mats are almost no longer luminescent. It can be assumed that the luminescence is affected by the growth state of the bacterium which varies according to the cell density on the solid culture medium. Bacteria in the bacterial mat are stuck to each other. They have to share the nutrients of the medium and the oxygen. While isolated bacteria have a greater availability of nutrients and oxygen.
Based on the results, we could conclude that pSEVA531-fiatlux is more efficient than pSEVA521-fiatlux with D.solani because it generates more light without affecting bacterial growth.
2.2. Effect of tetracycline on the growth and luminescence of D. solani fiatluxCDABE strains on liquid media
We then performed the characterization of luminescent strains of Dickeya solani in liquid culture in the same way as it was done with E. coli. The objectives were to:
In order to reach these goals, we followed the OD (at 600 nm) and the luminescence of the bacterial cultures for 48 hours, thanks to a microplate reader. Bacteria containing each of our four plasmids are put in a liquid culture with different concentrations of tetracycline (0, 3, 5, 7 and 10 µg/mL). All the cultures were added at the same OD, and we did 4 to 8 replicates for each condition, so we can perform statistical analysis. Thanks to these data, we performed a mathematical modelisation of the bacterial growth, in order to compare the different conditions
The results of this experiment can be found on the page Results : Model.
To test the stability of our plasmid in D.solani, we performed a stability test (20) for each plasmid. In the absence of selection pressure (the antibiotic), the plasmid could be lost during cell divisions. This control is essential for iGEM teams wishing to ensure that the plasmid remains stable in their bacteria.
The objective is to determine if after a certain number of generations without selective pressure, enough bacteria still possess the plasmid. A generation of bacteria is defined as the division of one bacteria into two bacteria. At every generation, the number of bacteria is multiplied by two. For example, in a chicory leaf, after beginning of infection, between 10 and 15 generations are sufficient to infect all the leaf. In our experiment in vitro in LB liquid culture, we set up the experiment to obtain 9 cell divisions of D. solani per day. This can be calculated by measuring the OD of the culture at 600 nm at the beginning of the culture (DO(0) = 0,0012) and after 24 of incubation (DO(f) = 0,600). We performed the test over three days, to obtain a total of 27 generations. Every day, we enumerated the bacteria cells of the liquid culture that grew on two different media (LB and tet), but we also monitored the evolution of luminescence (produced by pSEVA531-fiatlux and pSEVA521-fiatlux). Bacteria growing in LB medium represent the total number of bacteria in the liquid culture and bacteria growing in LB with 5µg/mL of tetracycline represent only bacteria of the culture still containing the plasmid.
This experiment was realized with two replicates that are consistent. In order to do statistics, a third replicate has to be realized (not done here because of a lack of time).
The results of this experiment can be found on the page Results : Wetlab.
After characterizing D.solani with our four plasmids, we performed infections of chicory leaves and potatoes (23). The objectives are multiples:
Firstly, we infected chicory (23), because this plant can be easily and rapidly infected, and its small size enables us to make a lot of replicats. Dickeya solani carrying pSEVA521, pSEVA521-fiatlux, pSEVA531 or pSEVA531-fiatlux were used to infect chicory leaves, with eight replicates for each. Four chicories were used: their leaves were mixed to have homogeneous leaf sizes between the replicates and to avoid a disturbance of results due to more resistant chicory. The size of the infection sites were measured after 30 and 48h, and the luminescence of the leaves was observed with a ChemiDoc XRS+, Bio-rad highly-sensitive camera.
D. solani with plasmids pSEVA531 or pSEVA531-fiatlux showed more important signs of infection than D.solani with pSEVA521 and pSEVA521-fiatlux. Indeed the average size of infection is significantly higher for pSEVA531 and pSEVA531-fiatlux (after both 30h and 48h).
Moreover, we can observe a significant difference of the size of the rotten area between D.solani with pSEVA521 and pSEVA521-fiatlux. On the other hand, there is no significant difference between the size of the rotten area between D.solani with pSEVA531 and pSEVA531-fiatlux. The behavior of the infection by D.solani is impacted by fiatlux when used in pSEVA521, but not when used in pSEVA531. All the test has been performed with the software R using R studio. In order to calculate the pvalue, we used the multiple tukey test, enabling us to compare each single pair of means. For further details about the code, please refer to: Model Data
Moreover, infection with bacteria containing pSEVA531-fiatlux produces much more luminescence than with bacteria containing pSEVA521-fiatlux. This observation confirmed that for D.solani, the plasmid pSEVA531-fiatlux is more efficient than pSEVA521-fiatlux (regarding symptoms of infection and luminescence).
Timelapse of D. solani containing pSEVA531-fiatlux or pSEVA531 done with the NightSHADE LB 985 In Vivo Plant Imaging System was realized, by taking pictures every 30 minutes to follow the evolution of both the luminescence and the symptoms.
On this video, we can observe the evolution of symptoms over time: the rot is growing all over the leaf, and reaches the edge of the leaf for one of them. As far as luminescence is concerned, we observe a front of luminescence where the bacteria are degrading the tissues. We can suppose there are more bacteria at this front, or more nutrients available, so more energy for the bacteria, and therefore more luminescence. After some time, luminescence is fading in the most rotten parts of the leaf. All those observations are complemented by a further analysis thanks to the software designed by our team.
Potato tuber were infected, with only D.solani containing plasmid pSEVA531 and pSEVA531-fiatlux. Five potatoes were infected for each plasmid. After 24h and 48h after the infection, half of potatoes were cut to observe the putrefaction in them, and the luminescence, thanks to the ChemiDoc XRS+, Bio-Rad and the NightSHADE LB 985 In Vivo Plant Imaging System.
As far as symptoms are concerned, we can observe a rotten area of approximately the same size for each tuber. For one of the control potatoes, infected with the empty vector, the symptoms are more severe. Other replicates will be performed in order to study those symptoms. Luminescence is well observed for potatoes infected with bacteria containing fiatlux operon, and not with empty vectors, as desired.
We succeeded in infecting chicories and potatoes with our plasmids. We could observe luminescence only when D.solani, used for infections, owns the fiatlux operon on its plasmid. The next step would be to infect potato plants or another living plant, to observe how bacteria spread themself into tissues and the different parts of the plants. It would be also interesting to infect a potato tuber and let it grow, to observe the bacteria during the whole life cycle of the plant.
fiatlux is made as an analyzing tool, to help scientists to better understand the behavior of new or unwell known pathogens. We value the adaptativity of our tool, and in order to offer our tool to researchers, we want to make sure it can adapt to as many bacteria as possible. This is why we inserted our plasmid pSEVA531-fiatlux in other bacterium genus. We chose pathogens with different assets, in order to open opportunities for other studies. Here are the ones we tried so far:
We decided to insert our plasmid with the conjugation method, because it is easy to implement and has a great chance of success. Furthermore, we already had the bacteria E.coli MFDpir with our four different plasmids (discussed in the preceding section), which sped up the experiment.
The conjugation gave us colonies for all of the four bacterial strains we tested. Then, the bacteria were observed with the ChemiDoc XRS+ Bio-rad highly-sensitive camera. Luminescence was observed only for two strains: Citrobacter rodentium and Pseudomonas putida. For those two strains, luminescence was oberserved for bacteria containing plasmid pSEVA531-fiatlux, and not for those containing the empty vector. This means that our fiatlux operon is well responsible for the luminescence observed.
These results show that our tool can be inserted in other bacteria than D.solani (in this case Citrobacter rodentium and Pseudomonas putida). However, the characterisation and the stability of pSEVA531-fiatlux have to be determined in future experiments.
After having a proof of concept by infecting chicory with D.solani containing our plasmids, we wanted to create new biobricks, in order to improve our tool.
2.1. Offer a Biobrick without Promoter
The objective was to create a biobrick with fiatlux, that we could offer to other igem team or researcher, by enabling the use of our tool with any promoter. For example, they can add promotor with a higher copy number, or an inducible promoter (unlike the promoter used in the project).
A digestion (2) enables us to remove the promoter J23117 and an annealing of two primers (20) was made to ligate both plasmid ends. The new vector has a unique BamHI restriction usptream iluxC that will be useful for further cloning of new promoters.
2.2. Toxin/antitoxin system
We then wanted to couple the fiatlux operon with a toxin/antitoxin system in the same plasmid. This enables a constitutive selection of the plasmid to be carried out: this method of selection ensures that all the bacteria present on the plant have the plasmid, and can hence emit light. Toxins have a longer life span in cells than antitoxins, so if the bacteria lose their plasmid, the toxin is no longer destroyed by the antitoxin (Van Melderen 2010). This results in the death of the bacteria, thus assuring the death of all fiatlux non-carriers. As a consequence, by involving a selective pressure, the toxin/antitoxin system increases the stability of the plasmid in the bacteria.
We tried to clone 2 differents toxin/antitoxin systems : hok/sok and axe/txe. Both systems should be inserted in pSEVA531-fiatlux, which was digested (2), before the fiatlux operon. axe/txe was inserted in the opened plasmid thanks to a TEDA cloning (9). hok/sok, contained in pGEX, was transfered in the plasmid pSB1C3 thanks to a TEDA cloning (9). Then hok/sok was inserted in pSEVA531 thanks to a ligation (3). E. coli DH5α were transformed by both plasmids.
After verification thanks to a PCR on colonies (10), we can conclude that only the plasmid pSEVA53-fiatlux-Axe/Txe worked, but not the pSEVA53-fiatlux-Hok/Sok. In order to make sure that our transformation with pSEVA53-fiatlux-Axe/Txe worked, we also did a restriction map, which showed coherent results. We also plan to sequence our plasmid.
2.3. Creation of a Plasmid by replacing fiatlux by lux
The objective of this plasmid was to replace the fiatlux gene by the lux gene (Gregor et al. 2018), to compare the difference in emitted luminescence between fiatlux and lux. Indeed, although we could show the efficiency of our fiatlux system to make bacteria luminescent, we could not compare if fiatlux was more luminescent than the luxCDABE system currently commonly used in laboratories. This plasmid construction did not work. We leave room for future reworkings and improvements of our project for future iGEM teams.
2.4. Insertion of fiatlux in a Plasmid with a Kanamycin Resistance Gene (did not work: to be developed further by future iGEM teams…)
The objective of this plasmid is to see the effect of the new antibiotic resistance gene on the bacterial growth and on the luminescence. Indeed, we observed during our project that the level of luminescence varies according to the concentration of tetracycline used in the medium to maintain a selecting pressure. Changing the type of resistance gene would allow us to determine if the observed phenomenon is only related to tetracycline resistance. Kanamycin is a bacteriostatic antibiotic, which means that it kills bacteria rather than inhibiting their growth as a bacteriostatic antibiotic such as tetracycline does. Unfortunately, this plasmid construction did not work. We leave room for future reworkings and improvements of our project for future iGEM teams.
The aim of this part was to purify the luciferase encoded by fiatluxAB and the FMN-reductase encoded by frp, and to characterize the enzymatic kinetics of the luciferase encoded by fiatluxAB.
3.1. Extraction of our luciferase
Our luciferase was extracted from E.coli MFDpir containing the plasmid pSEVA531-fiatlux. E.coli MFDpir transformed with pSEVA531 (high copy number plasmid) was chosen and not with pSEVA521 (low copy number type) because we wanted the highest possible concentration of our luciferase. The lysis of the bacteria (16) was done using the “French Press” method. This method was chosen because it disrupts cell walls without altering cytoplasmic proteins. The solution obtained is centrifuged and the proteins remain in the supernatant. The supernatant is collected: this is our solution A.
3.2. Purification of our Luciferase
Our solution A contains all the proteins from our E.coli MFDpir x pSEVA531-fiatlux. Our luciferase encoded by fiatluxAB has to be separated from the other proteins. It is an enzymatic complex that weighs 80kDa with two subunits of ~40kDa (Meighen et al., 1996). Because we want to use our luciferase after the purification, we have to purify it without denaturing it.
A Sephadex G-100 column was used to realize a gel exclusion chromatography (17). The exclusion size of the Sephadex G-100 column is 100kDa. In that way, all proteins heavier than 100kDa will be excluded from the column and eluted first.
A column of 20cm tall was built, dissolving the gel in a solution of distilled water containing 0.03% of azide sodium (17). The O.D. 280 nm of the solution A is of 3.301. This measure is necessary to have a global idea of the quantity of proteins contained in it.
The column is loaded with 1mL of the solution A (17). The fraction collector is set to collect 2.5mL of eluent in each tube.
The O.D. 280 nm is measured for each collected tube in order to draw the chromatogram. The chromatogram showed a single peak meaning that the column was overloaded. The separation of the proteins was unsuccessful.
A new 20cm tall Sephadex G-100 column was built as previously. 350µL of the solution A is loaded. 30mL of the eluent were collected in 12 tubes of 2.5mL. The chromatogram obtained is shown below.
The O.D. 280nm of a solution of 350µL of the solution A added to 2.5mL of the eluent is 0.394. The first tube collected contains the majority of the proteins (O.D. 280nm = 0.214). After the 6th tube, the O.D. 280nm drops to almost 0, meaning that all proteins have been eluted. The first six tubes were conserved at -20°C.
A sample of each of the 6 tubes collected after the Sephadex G-100 column must be used in a polyacrylamide gel electrophoresis (PAGE) (18) in order to know the size of proteins contained in each tube.
Because the luciferase is a complex composed of two subunits, the gel must be non-denaturing. 10µL of each tube were deposited in the gel with a loading buffer (18).
No bands other than the protein ladder (molecular-weight size marker) appeared. Either our tubes had no proteins or the protein concentration was too weak. The detection threshold is at 50µg of proteins deposed. Hence, the protein concentration of each tube was assayed using the Bradford method (19).
A standard range of BSA was created (19) (Fig.X). The R² is 0.9953 which is satisfactory. The equation is “O.D.(595nm)=0.0193*concentration” with the concentration in µg/mL.
The protein concentration of two samples (n=2) of each tube are dosed using the Bradford method (19). The O.D. 595 nm is measured and used to calculate the protein concentration using the equation of the standard range of BSA. In order to have the real concentration in each tube, the factor 250 must be applied to the concentration calculated because a dilution of 250 was done on each sample in the protocol (19).
The protein concentration of each collected tube and of the solution A range between 0.96 and 1.20 µg/µL. Because 10µL was deposed in the gel, the amount of protein in the gel ranged from 9.6 to 12.0 µg, which is under the threshold value of 50µg. Hence, proteins had to be concentrated in each tube. Because our solvent is aqueous, we thought of using Amicon ® Ultra centrifugal filters on our collected tubes to concentrate our solutions. However, even if this will allow us to see proteins on a PAGE, the amount of proteins collected in each tube will be too low for further use.
3.3. Conclusion
We didn’t have the time to continue our experiments. If we had more time, we would have changed our plan and switched the promoter in front of fiatluxA and fiatluxB for a high-level expression inducible promoter. This and the following plan for the enzymatic characterisation is presented in below. However, this experiment shows that the amount of luciferase present in our bioluminescent bacteria is too low for direct purification.
In order to purify and verify the presence of our luciferase in the collected fractions in a more optimal way, we thought about a new series of experiments that we could have performed if we had enough time. The first thing to do would be to place the gene encoding for the luciferase under the control of a strong inducible promoter, such as the T7 promotor, in order to overexpress the protein. Then we would have performed a PAGE to check this overexpression. This way, we would obtain highly concentrated fractions after the column purification.
The next step would be to produce an in vitro reaction that would allow us to characterize the luciferase’s kinetics by measuring the disappearance of only one reagent (n-decanal), as it satisfies the required conditions to measure the activity of our protein of interest. To follow the disappearance of the n-decanal, we planned to do a spectrometric monitoring by choosing the wavelengh at which the n-decanal has the best absorbance.
Then, we would draw the evolution of the OD for different initial concentrations of n-decanal throughout time. From these graphs we would acquire information about the reaction speed (represented by the curve’s slope). Then, if we assemble all these datas we could plot a Michaelis-Menten hyperbole (reaction speed depending on initial n-decanal concentration). This representation would allow us to deduce different constants (Km, Vmax) that characterize our luciferase.
Here are all the protocols that we have followed to successfully conduct the experiments of our project!
References
Gregor C, Gwosch KC, Sahl SJ, Hell SW. Strongly enhanced bacterial bioluminescence with the ilux operon for single-cell imaging. Proc Natl Acad Sci U S A. 2018 Jan 30;115(5):962-967. doi: 10.1073/pnas.1715946115. Epub 2018 Jan 16. PMID: 29339494; PMCID: PMC5798359.
Jahn, M., Vorpahl, C., Hübschmann, T. et al. Copy number variability of expression plasmids determined by cell sorting and Droplet Digital PCR. Microb Cell Fact 15, 211 (2016). https://doi.org/10.1186/s12934-016-0610-8
Jousset, Alexandre, Joerg Schuldes, Christoph Keel, Monika Maurhofer, Rolf Daniel, Stefan Scheu, et Andrea Thuermer. 2014. « Full-Genome Sequence of the Plant Growth-Promoting Bacterium Pseudomonas Protegens CHA0 ». Genome Announcements 2(2):e00322-14. doi: 10.1128/genomeA.00322-14.
MEI CHONG, RICKY LEUNG, CARMEN WONG, AND ALEX YUEN. 2003. The Effects of Ampicillin versus Tetracycline on the Plasmid Copy Numbers of pBR322. Journal of Experimental Microbiology and Immunology (JEMI), Vol. 3:87-95,
Meighen E. A. 1993. Bacterial bioluminescence: organization, regulation, and application of the lux genes. FASEB journal : official publication of the Federation of American Societies for Experimental Biology, 7(11), 1016–1022. https://doi.org/10.1096/fasebj.7.11.8370470
Wang RF, Kushner SR. Construction of versatile low-copy-number vectors for cloning, sequencing and gene expression in Escherichia coli. Gene. 1991 Apr;100:195-9. PMID: 2055470.
Weimer, Anna, Michael Kohlstedt, Daniel C. Volke, Pablo I. Nikel, et Christoph Wittmann. 2020. « Industrial Biotechnology of Pseudomonas Putida: Advances and Prospects ». Applied Microbiology and Biotechnology 104(18):7745 66. doi: 10.1007/s00253-020-10811-9.
Wiles, Siouxsie, Karen M. Pickard, Katian Peng, Thomas T. MacDonald, et Gad Frankel. 2006. « In Vivo Bioluminescence Imaging of the Murine Pathogen Citrobacter Rodentium ». Infection and Immunity 74(9):5391 96. doi: 10.1128/IAI.00848-06.
Weinstock et al. (2016) Vibrio natriegens as a fast-growing host for molecular biology. Nature methods VOL.13 (NO.10) doi:10.1038/nmeth.3970
↑