<!DOCTYPE html> Site Igem

Experiments

Overview

Figure 1: General design for our ARBO-BLOCK project.

Small ConA quantity have been produced in E. coli

Cloning results

We couldn’t order the conA-cinker-conA sequence, because there was too much repetition in the sequence. That is why we have decided to do two successives, in order to get the awaited sequence.

conA-linker sequence
We have first amplified sequences ordered from IDT, with o211 and o213 primers in order to add, at the end of the sequence, BamHI : a restriction site. It will be needed for further cloning using SLIC method.

Figure 2: conA-linker sequence verification by PCR. Well 1 : Ladder ; Well 2 : PCR result for conA-linker sequence amplification ; Well 3 : Negative control.

Figure 2 allows us to verify that the first PCR, on the sequence ordered, worked properly, and amplified with the BamH1 restriction site. No bands are observed for the negative control. This shows that there was no contamination. For the second well, we may see a band at about 1000bp. We were awaiting a signal at 1020bp. Of this fact, the observed signal corresponds to the conA-linker sequence with BamH1.

This result allowed us to do a second PCR with the product of the first one as a matrix. PCR product has been amplified with the o211 and o214 primers in purpose to add the suffix needed to realise the cloning.

Figure 3: Second PCR verification for conA-linker. Photo of an agarose gel exposed to UV after a gelred bath. Well 1 : Ladder ; Well 2 : PCR result for ConA-Linker sequence amplification with suffix.

Figure 3 allows us to verify that the second PCR, on the product of the first one, worked properly and has properly added the suffix. We can see on the second well, a band at about 1000bp whereas we were awaiting a band at 1050bp. The observed signal corresponds to the conA-linker sequence with the suffix added. This result confirms the obtention of the conA-linker sequence with BamH1 and the suffix.

The product of these two consecutives PCR reactions has been used to do a cloning by digestion/ligation in the J04500 plasmid. The insert has been digested by Xba1 and Pst1, the vector has been digested by Spe1 and Pst1. It follows the RFC10 standard. We realized a ligation before transformation in competent K12 DH5𝛂 E. coli.

A colony PCR has been made on colonies that appeared on selection media, consisting of LB and chloramphenicol.

Figure 4: Colony PCR of the transformed colony conA-linker in J04500 plasmid and in K12 DH5𝛂 E. coli. Photo of agarose gel exposed to UV after a gelred bath. Well 1: Ladder, well 2 : negative control, well A to N : clones A to N.

Figure 4 shows the migration of PCR products on the colony of the different transformants obtained.The negative control, which does not contain DNA, has no signal in the corresponding well, indicating that contamination has not occurred. To transform it with the empty vector, a band of 600 Kb is observed, which does not correspond to the expected result of 300 Kb. Same result for wells D, E, H, I up to N which do not correspond to the expected result of 1543 Kb. For wells F and G, the band is at 800 Kb so not at the expected size. The results of this first colony PCR are inconclusive. We have tried to carry out this first cloning several times, but always without success.

That’s why we decided to change our approach by trying to get a conA without the linker, and cloned it into another vector. This changed the overall strategy for obtaining ConA-Linker-ConA.We decided to clone a single conA sequence without linker in a pBAD24 linker as we used this plasmid for other constructions, by using SLIC method to integrate it in the plasmid.

conA cloning
We chose to clone conA in pBAD24 vector by SLIC. To do this, the conA sequence is amplified by PCR from the order sequence with o2140 and o2141 primers.

Figure 5: PCR check for conA. UV detection after gelred bath. Well 1: ladder, well 2: PCR result of conA amplification.

Figure 5 shows the PCR product intended to amplify the conA sequence and prepare it for cloning by SLIC. There is a band at about 900 bp, the expected size is 900 bp. We have thus amplified the conA sequence with complementary ends of the vector needed for cloning.

SLIC and colony PCR were performed to discriminate recombinant clones of simple transformants.

Figure 6: Colony PCR following SLIC. Agarose gel exposed to UV after bathing in gelred. well 1: ladder, well 2: negative control, well 3: positive control, well 4 to 11: clones 1 to 8.

On figure 6, the negative control is contaminated. A band of 1100 bp was observed for all 8 clones as expected for a recombinant clone.These results show that the 8 colonies tested are potentially recombinant with the correct conA sequence.

To make sure we have performed a digestion profile on the first 3 clones isolated with the NcoI enzymes. If two bands appear, one of them at 900 bp, it means that they have conA sequence.

Figure 7: Digestion profile of clones 1,2 and 3 isolated from conA cloning in pBAD24. well 1: ladder, well 2 to 4: purified plasmid digestion profile of clones 1 to 3.

All clones tested were positive, with a signal of about 900 bp. The plasmid from clone 2 is then sent to be sequenced in order to ascertain the sequence of the clone.

conA sequencing

Figure 8: Result for conA sequencing in pBAD24. The orange-colored sequence corresponds to the concanavalin A sequence. The other is the result of sequencing.

Sequencing shows no mutation. This indicates that the cloning of conA in pBAD24 by SLIC is a success. This allows us to proceed to the next step for producing the concanavalin A protein in E. coli MG1655.

ConA production

To produce ConA, we transformed the plasmid obtained in E. coli MG1655. Since conA is controlled by a pBAD promoter, we must grow the bacterium in arabinose supplemented medium. We chose the conditions based on tests, performed for overproduction of toxins that are cloned in the same vector. The induction is done when the crop reaches the beginning of the exponential growth phase with 1% arabinose. After an unsuccessful first attempt because no corresponding signals have been observed on Western blot, we repeated the experiment.

ConA detection

In order to detect ConA by western blot, it was tagged with a 6His-Tag in the N-terminal position. We performed a western blot on the overnight-induced culture, with primary antibodies against His-tag.

Figure 9: Western blot on overnight culture of E. coli MG1655 pBAD24 ConA, with anti-His-tag antibodies. well 1: ladder, well 2: E. coli MG1655 pBAD24 ConA.

The figure 9 shows a weak signal at about 32KDa, which is the expected size for ConA. Therefore, we can conclude that we have been successful in producing concanavalin A in E. coli MG1655.

T5SS a promising system to address ConA in the outer membrane

N.B.1: We forgot to change the corresponding codon when we ordered the DNA sequence on iDT so we obtained a wild-type version of the Aida-I autotransporter. (BBa_K3788010)
N.B.2: We realized afterward that our construction wouldn’t be functional due to the presence of the cleavage site. We ordered a new DNA sequence that is the same but integrating the D33N mutation (BBa_K3788011)

Export to the periplasm: cloning strategy

Our first choice was to integrate our sequence in the Part BBa_K3788013 by cloning it in RFC10. However, we realized that the RFC10 standard creates a scar after ligation that integrate a STOP codon, and so, even if we insert the Part BBa_K3788010 in Part BBa_K3788013, we couldn’t see any fluorescence at the outside of the cell due to the absence of production of the Aida-I autotransporter.

Consequently, we decided to use another plasmid to assemble both parts and we used pBAD24 plasmid, which was given by researchers in our lab and already used for other parts of our project.

The assembly of BBa_K3788013 and BBa_K3788010 is named BBa_K3788026

First, we have ordered on IDT the coding sequence for pelBsfgfp and DNA primers adapted to amplify pelBsfgfp with prefix and suffix compatible with RFC10 cloning. We adapted our amplification program with the optimal temperature for primer annealing and the right elongation time to ensure that we obtain the complete sequence.

Once the PCR amplification was over, we performed an agarose gel migration to verify if our product size corresponds to the expected size of our sequence; then, we purified our results with [kit PCR clean-up] to retain only purified DNA. After purification, our sample contained 50µL at 41,9ng/µL.

In parallel, we extracted J04500 from the iGEM distribution kit 2019 plates. Then we digested both plasmids (with SpeI and PstI restriction enzymes) and insert (with XbaI and PstI restriction enzymes) to allow their assembly following the RFC10 standard, before their ligation and transformation into DH5a E. coli strains.

Figure 10: Visualisation of a Petri dish containing bacteria transformed with J04500 (on the left) and J04500_pelBsfgfp (on the right) with no induction.

We can observe that bacteria containing J04500_pelBsfgfp express our protein even without induction. We made the hypothesis that it’s due to a lack of our promoter.

Figure 11: Schematic representation of the molecular biology construction and verifications made.

Export to the periplasm: experiments and results

The quantitative measure of fluorescence

Induction with IPTG
To follow the fluorescence intensity variations across time of our bacteria, we induced the lac promoter in liquid culture with different concentrations of IPTG. The first step was to prepare 10mL starters of bacteria, containing our plasmid or not, the day before to allow bacterial growth overnight. In the morning, we measured the 600OD of our starters and then took a volume equal to 600OD=0,05 in a final volume equal to 100mL in 8 Erlenmeyer.

Once the 600OD is higher than 0,4 we add a certain quantity of IPTG and let it incubate for 3 hours at 37°C.

Figure 12: Schematic representation of the protocol followed to measure the IPTG induction created on Biorender).

**Table 1: Conditions of IPTG induction**
Condition	IPTG concentration (µM)
Bacteria containing J04500_PelBsfGFP	0	10	20	50	100	1000
Empty bacteria	0	N/A				1000

After the 3 hours of induction, we measured fluorescence intensity to determine the effect of IPTG induction on protein production.

Figure 13: Fluorescence intensity obtained for each induction condition. Induced culture of cells containing the BBa_J04500 plasmid cloned with the sfgfp coding sequence studied at OD 0.02 (green), OD 0.2 (yellow) and OD 2 (orange); Induced culture of cells containing the empty BBa_J04500 plasmid studied at OD 0.02 (light blue), OD 0.2 (grey) and OD 2 (dark blue)

We concluded that adding IPTG to our bacteria doesn't have any effect on our protein production. We can correlate these results with the fact that our bacteria, containing J04500_pelBsfgfp, express our protein at a level high enough to see the green color with naked eyes.

A hypothesis could be that the Lac repression in our bacteria is not sufficient to see a difference with or without IPTG induction or that the lac promoter is not working well.

Testing the repression with glucose
To determine whether the Lac repression is not sufficient or if the promoter is not working, we incubated bacteria containing J04500_pelBsfgfp with different glucose concentrations, as described in the followed table:

**Table 2: Concentrations of glucose used for incubation.**
Sample	1	2	3	4	5	6	7
Glucose content ratio (% v/v)	0	0.01	0.02	0.05	0.1	0.2	0.5

The first step was to prepare 10mL starters of bacteria containing our plasmid, or not, on the day before to allow bacterial growth overnight. In the morning, we measured the 600OD of our starters and then took a volume equal to 600OD=0,01 in a final volume equal to 100mL in 7 Erlenmeyer.

Once the 600OD is higher than 0,1 we add a certain quantity of glucose and let it incubate at 37°C. We measured 600OD for each conditions at t=0 ; t =60 ; t=120 and t=170.

Figure 14: Schematic representation of the protocol followed to prove the glucose repression (created on Biorender).

Figure 15: Fluorescence intensity obtained for cultures expressing the sfGFP protein (bacteria holding the cloned vector) at 0 ; 60 ; 120 and 170 minutes for each condition (0,5% not shown). Are presented in purple, results for a 0.2% glucose ratio; in red, results for a 0.1% glucose ratio; in orange, results for a 0.05% glucose ratio; in yellow, results for a 0.02% glucose ratio; in green, results for a 0.01% glucose ratio and in blue, results for a 0% glucose ratio.

Here we can see that, the more there is glucose in the growing medium, the lowest the fluorescence intensity is. This experiment gives us a clue on how to interpret the non-induction of protein expression: it looks like the lac promoter is functional and the non-induction by IPTG could be due to a too low level of LacI in the cell to repress the promoter.

Visualization of protein location with a confocal microscope
To help us determine if the fluorescence is located in the periplasm or the cytoplasm, we used a confocal microscope to visualize our cells.

For this purpose, we prepared 3mL 2YT growing medium with chloramphenicol and then put MG1655 containing either J04500_pelBsfgfp or J04500 plasmids.After 3 hours of incubation at 37°C, we prepared the plates and visualized our cells.

Figure 16: Visualization of MG1655 E. coli strains containing J04500 (left) or J04500_pelBsfgfp (right) under confocal microscope.

We can see that bacteria containing J04500 do not emit light, contrary to the ones containing J04500_pelBsfgfp, which are green under a confocal microscope. However, we are not able to determine whether the light is located at the surface of the bacteria or if it is located in all the cells.

Export to the external membrane: cloning strategy

As we were not using an iGEM plasmid anymore, we had to adapt our strategy for cloning our sequences in our plasmid. We decided to use the SLIC method to assemble both of the parts with a single restriction of the vector.

Figure 17: Retrieved from M. Gearing. Plasmid 101: Sequence and Ligation Independent Cloning (SLIC).(2017).

To use the SLIC method we have to :
1) Amplify our sequences with specific primer to create sequences with extremities that will be degraded by T4 DNA Polymerase, so that there will be a complementarity with the ss-DNA extremities.

We designed four primers:

Amplification of BBa_K3788025 sequence with extremities compatible with SLIC in pBAD24 and with BBa_K3788010.

Fw: o2125
Rev: o2127 (also suppresses the 2xSTOP codon at the end of the sfgfp sequence in BBa_K3788025)

Amplification of BBa_K3788010 sequence with extremities compatible with SLIC in pBAD24 and with BBa_K3788025.

Fw: o2128
Rev: o2129

(2) Find a restriction site in the vector where we will integrate our sequences. We used a NcoI restriction site that is near the SD sequence so our sequence will be well expressed.

We followed the SLIC protocol to assemble the 2 parts in the pBAD24 vector.

Figure 18: Schematic representation of the protocol followed to measure the IPTG induction.

Export to the external membrane: experiments and results

Quantitative measurement of fluorescence.

To follow the fluorescence intensity’s variations across time of our bacteria, we induced the protein expression in liquid culture with different concentrations of Arabinose.

The first step was to prepare 10mL starters of bacteria containing our plasmid on the day before to allow bacterial growth overnight. In the morning, we measured the 600OD of our starters and then took a volume equal to 600OD=0,05 in a final volume equal to 100mL in 2 different Erlenmeyer. Once the 600OD is higher than 0,2 in one of the two Erlenmeyer, we take 10mL that we put in another Erlenmeyer (7 times) and add a certain quantity of Arabinose and then let it incubate at 37°C. Once the 600OD is higher than 1 in the other Erlenmeyer, we did the same manipulation as before.

**Table 3: Concentrations of arabinose in a final volume of 10mL, used for induction of protein expression.**
Sample	1	2	3	4	5	6	7
Arabinose content ratio (%)	0	0.2	0.5	1	2	5	10

We measured the fluorescence every 60 minutes during 5 hours with the TECAN machine, a 96-well plate reader.

Figure 19: Schematic representation of the protocol followed to measure the Arabinose induction.

Figure 20: Fluorescence intensity across time in different conditions of induction.Left : induction at 600OD= 0,2 ; Right : induction at 600OD = 1.

Both figures show that the highest level of fluorescence is obtained with Arabinose induction 1%. However, the fluorescence intensity levels measured are very low, compared with BBa_K3788013 fluorescence intensity levels, which may be due to a low protein expression that could be a problem for further immunodetection.

Visualization with a confocal microscope
To help us determine if the fluorescence is located in the periplasm or in the cytoplasm, we used a confocal microscope to visualize our cells.

In order to do that, we prepared 3mL 2YT growing medium with ampicillin and then put MG1655 bacteria containing either empty pBAD24 (induced with 1% arabinose) or pBAD24 with BBa_K3788026 plasmids (induced with 1% arabinose/non-induced). After 3 hours of incubation at 37°C, we prepared the plates and visualized our cells.

Figure 21: Visualization with a confocal microscope of bacteria containing pBAD24 with BBa_K3788026 induced with 1% Arabinose (left), pBAD24 with BBa_K3788026 not induced (middle), empty pBAD24 (right)

We can see that bacteria containing pBAD24 induced with 1% arabinose, and bacteria-containing pBAD24 with BBa_K3788026 without induction do not emit light, but the bacteria containing pBAD24 with BBa_K3788026 with 1% arabinose induction does emit light under a confocal microscope.

However, we are not able to determine whether the light is located at the surface of the bacteria or if it is located in all the cells. We can’t conclude about the localization of our protein.

At this point, we realized that our construction doesn’t integrate the D33N mutation and so that the protein won’t be functional as we expect. We decided to order a new aida-I sequence integrating the mutation (BBa_K3788011). With the corrected sequence, we realized the same cloning protocol: BBa_K3788027 in pBAD24.

Immunodetection
To ensure the localization of our protein we made an immunodetection with anti-GFP antibodies. As we realized that our fluorescent protein should be cleaved and secreted in the medium because we forgot to integrate the mutation D33N that clears the cleavage site, we decided to centrifugate our cells and to charge on an SDS-PAGE cell and supernatant before revelation with antibodies.

Figure 22: Nitrocellulose membrane showing the results of the immunodetection led on cells and supernatant of bacteria containing pBAD24 and BBa_K3788026 induced or not. Controls: bacteria containing empty pBAD24, induced with 1% arabinose for 5 hours; charge control for anti-GFP antibodies; Rule page Ladder molecular weights (kDa) (lowest to highest band) : 10 ; 15 ; 25 ; 35 ; 55 ; 70 ; 100 ; 130 ; 250.

We can observe a band at a molecular weight between 25 and 35 kDa; according to ApE, sfGFP protein has a molecular weight equal to 26,8 kDa, so we could think that this band corresponds to sfGFP that has been cleaved from Aida-I autotransporter, but not secreted.

However, we know that the protein transported to the periplasm with the Sec pathway doesn’t take their final conformation before being in the periplasm, so we can make a different hypothesis:

Our construction is localized to the periplasm, but the Aida-I autotransporter can’t encroach to the external membrane but still conserve its proteolytic activity to release the passenger protein, which explains that the molecular weight corresponds to a free-sfGFP.
Our construction is not taken in charge by the Sec Pathway, and so it isn’t directed to the periplasm and takes its conformation in the cytoplasm; and due to the mutation that we forgot to integrate into aida-I sequence, the passenger protein is cleaved inside of the cell.

We had the time to construct the Part BBa_K3788027 in pBAD24, and to realize an immunodetection in the same conditions as for BBa_K3788026 immunodetection.

Figure 23: Nitrocellulose membrane showing the results of the immunodetection led on cells and supernatant of bacteria containing pBAD24 and BBa_K3788027 induced or not. Controls: bacteria containing empty pBAD24, induced with 1% arabinose for 5 hours; charge control for anti-GFP antibodies; Rule page Ladder molecular weights (kDa) (lowest to highest band) : 10 ; 15 ; 25 ; 35 ; 55 ; 70 ; 100 ; 130 ; 250. Two clones were tested at the same time because we hadn’t received the sequencing results yet and were at the end of our lab time.

We can’t see any specific band (no bands either at 25kDa or76 kDa for induced and non-induced conditions). We can’t conclude if our passenger proteins stay linked to the Aida-I autotransporter.

Perspectives

If we had more time, we would have done cell-fractionation experiments and then an immunodetection with anti-GFP antibodies to ensure that our construction is well located in the bacteria’s periplasm and further, on the external membrane.

Engineered E. coli timer lysis device permits asynchronous toxins production and release

Initial study of the original Colicin A synthesis and release system

Since we planned to use the asynchronous production of toxin and lysis protein of the Colicin A system as a timer-lysis device for our project, we first needed to characterize the original colicin system in order to understand its functioning, and to be able to engineer it. Colicin A and the lysis protein CaL, responsible for colicin release in the environment, are transcribed as an operonic structure caa-cal from the pRL1 plasmid. The genetic region also includes the cai gene, encoding an immunity protein CaI, which is constitutively transcribed in the opposite direction. Previous publications have reported the identification of regulatory elements in the operon, including the -35 and -10 promoter elements, the +1 transcription site and LexA binding sites in the promoter region of the caa-cal operon, as well as several transcriptional terminators. Overall, regulation of the colicin operon is complex and not totally elucidated (1).

Figure 24: Operonic structure and regulation of the colicin A system. caa: colicinA gene; cai: immunity protein CaI gene; cal: lysis protein.cal gene.

The Colicin A and lysis protein CaL synthesis are under the control of the LexA repressor and respond to the SOS response. In the literature, this proteins are commonly produced in response to the addition of the DNA-damaging agent mitomycin C at the final concentration of 300 ng.ml^-1 for 180 min (2,3). We decided to test if the promoter was responding in an on/off manner or in a dose-dependent manner to mitomycin C.

To characterize the LexA-dependent promoter of the ColicinA plasmid, we first grew cells carrying the original pRL1 plasmid with various doses of mitomycin, ranging from 0 to 800 ng.ml-1. The OD (600 nm) was recorded over time using a TECAN microplate reader. As expected, after mitomycin induction, the OD decreased, which indicates the synthesis of CaL, and the resulting lysis of the producing cells. We observed that there was a dose-dependent effect of mitomycin C on cell lysis, the minimal concentration being 50 ng.ml^-1 (at T = 170 min). Higher concentrations resulted in an earlier (around T = 90 min) and more drastic cell lysis.

Figure 25: OD(600) across the time for bacteria containing pRL1 with or without mitomycin C. The experiment was made with pRL1 plasmids in which Cal was disrupted (pRL1 Cal-) (up) or with WildType plasmids (pRL1) (down). The bacterial colonies were incubated with 800ng.mL^-1 mitomycin C or 0 mitomycinC.

This result was not due to a toxic effect of mitomycin on the cells, as we verified that cultures carrying the pRL1 mutant plasmid producing a non-functional lysis protein (disrupted cal gene) grew similarly in the presence of 0 to 800 ng.ml^-1 of mitomycin C.

Timer-lysis device: characterization of CaL synthesis

In order to characterize how the colicin release system could be used in our project, we studied the previously reported lag period occurring between the induction of the system, the production of Colicin A and CaL lysis protein. Because both caa and cal genes are described as under the regulation of the same promoter, we rationalized that the delay between Colicin A and CaL synthesis was dependent on regulatory elements present in the sequence between caa and cal genes. The BBa_K378818 part composed of this proposed regulatory sequence and of the cal gene was designed “timer-lysis device” part (BBa_K3788018).

Figure 26: Representation of BBa_K3 788 018. cai: CaI immunity protein gene; cal: CaL lysis protein gene. The pink segment is the 3’ caa region containing the cai terminator.

To study its activity, a shorter version with a truncated (and non-functional) cal gene (=Part BBa_K378815 “Modified lysis device”) was fused to the RFP encoding gene (Part BBa_E1010) to construct the BBa_K378820 composite part.

Figure 27: Representation of BBa_K3 788 020. caa: colicin gene, cai: CaI immunity protein gene. The 5’ region of callus containing the cai promoter is fused with the RFP gene.

Cells carrying the construct were treated with mitomycin C and the red fluorescence signal was recorded over time using a TECAN microplate reader.
Following iGEM Measurement recommendations, the fluorescence intensity was converted to a concentration of TexasRed protein thanks to a standard curve of TexasRed dilution series.

Figure 28: Delayed synthesis of the CaL::RFP protein production. The experiment was done in triplicate. Standard deviation is small and not visible on the graph.

We observed a strong dose-dependent synthesis of our reporter protein upon mitomycin C treatment. The signal was detected after 330 min to 420 min of induction, for the highest and the lowest tested doses, respectively.

This is in contrast to what we observed when GFP was under the control of the LexA-dependent promoter (BBa_K378817) where we were able to observe the accumulation of green fluorescence as early as 40 minutes after induction.

Thus, we concluded that is possible to use a colicin plasmid-inspired device in order to program an asynchronous production of a toxin and of a lysis protein for its release in the environment.

Characterization of the LexA-dependent promoter responsible for ColicinA synthesis. (proposed as a Best basic part BBa_K3788017)

To go further in the characterization of the system, we amplified the promoter region of the caa-cal operon from the pRL1 plasmid in order to construct the promoter part BBa_K3788017. It was then fused to the GFP encoding part BBa_E0040 and to the timer-lysis device part BBa_K3788018 (composite part BBa_K3788019) in the pRL1 plasmid.

Figure 29: Representation of the part BBa_K3 788 019. caa fused to gfp by reserving its 5’ region with its RBS and its 3’ region with the cai terminator. cai: gene of the immunity protein CaI; cal: gene of the lysis protein CaL. The pink segment is the 3’ region of caa containing the terminator of cal.

The cells carrying the BBa_K3788019 construct were grown with various concentrations of mitomycin, ranging from 0 to 800 ng.ml^-1. The OD (600) and the amount of green fluorescence were recorded over time.

Following iGEM Measurement recommendations, the fluorescence intensity was converted to a concentration of fluorescein protein thanks to a standard curve of fluorescein dilution series.

Figure 30: Activity of the LexA-dependant promoter fused to GFP.The experiment was done in triplicate. Standard deviation is small and not visible on the graph.

Our data confirmed that the promoter part BBa_K3788017 responds to mitomycin in a dose dependent manner from 50 to 800 ng.ml^-1. The signal is significantly detected 90 min to 130 min after the beginning of the experiment, depending of the inducer tested concentration. Moreover, we showed that the promoter is strongly repressed in the absence of the inducer, as no fluorescence signal is detected for the untreated control during the time course of the experiment (450 min).

Finally, thanks to the composite part BBa_K3788019, we demonstrated that during the 90 to 210 min window after induction, there is a linear relationship between the accumulation of GFP signal and Mitomycin concentration. We established titration curves for this time period (R2>0.975).

Figure 31: Linear relationship between mitomycin concentration and GFP signal 90 to 210 min after induction.

We are enthusiastic about presenting the BBa_K3788017 promoter for the Best Basic Part special prize – as it represents a new and strongly regulated promoter.

Engineering design cycle : how we introduced the « Optimized Timer-Lysis device »

Interestingly, when we recorded the growth of cells carrying the composite part BBa_K3788019 (composed of BBa_K3788017, E0040, BBa_K3788018), we did not observe mitomycin-induced cell lysis, suggesting that our initial “Timer-lysis” device (part BBa_K3788018) was not adequate for inducing CaL dependent lysis. Unfortunately, we were not able to assemble the BBa_K3788021 composite part (including BBa_K3788017, BBa_E0040, BBa_K3788015, BBa_E1010) to experimentally verify this hypothesis.

Figure 32:

However, the data suggested to us that other regulatory elements might exist inside the caa-cal operon. We got back to the design of our constructs and re-analysed the caa-cal sequence with various in silico tools in order to identify potential hidden regulatory element. Indeed, we detected an alternative LexA binding site in the caa coding sequence (CTGATGGTACAGTCAG; position 781 to 796) using Virtual Footprint (https://www.prodoric.de/vfp). The prediction score was as high as for the published LexA binding site in front of the caa-cal operon (CTGTATATAAACACAT; position 270 to 285). This suggests the presence of an alternative and so far unpublished promoter responsible for cal transcription. However, preliminary bioinformatic analysis did no point out obvious -10 and -35 boxes in the same region.

Figure 33: Virtual footprint results (4).

Figure 32: Operonic structure and regulation of the colicin A system with an alternative LexA box. caa: colicin A gene; cai: CaI immunity protein gene; cal: CaL lysis protein gene.

From this new data, we decided to rethink our timer-lysis device in order to include this potential regulatory element (Part BBa_K3788016 « optimized timer-device »). In order to test this new design, we plan to fuse it to the LexA-dependent promoter (BBa_K3788017) and GFP (BBa_E0040) in the composite part (BBa_K3788023), and to test it using a similar methodology as described for the initial design.

Figure 34: Representation of BBa_K3 788 016. cai: CaI immunity protein gene; cal: CaL lysis protein gene. The pink segment is the 3’ caa region containing the cai terminator and the LexA box alternative.

Figure 35: Representation of the part BBa_K3 788 023. caa fused to gfp by retaining its 5’ region with its RBS and its 3’ region with the cai terminator and the alternative LexA box. cai: gene of the immunity protein CaI; cal: gene of the lysis protein CaL. The pink segment is the region 3’ of caa containing the cai terminator.

Bacillus thuringiensis toxins can lead to mosquitoes death

Molecular biology: plasmid construction

Figure 36: General scheme. Overview of the molecular biology part.

Two steps are distinct: 1) the design of 6hisp20_flagcry11Aa (BBa_BBa_K3788003), flagcry11Aa (BBa_K3788002), strepcyt1Aa (BBa_K3788000) and 6hisp20 (BBa_K378800) in pBAD24 MSC3 plasmid, and 2) the design of toxto (pBAD24_6hisp20_flagcry11Aa_strepcyt1Aa) in pBAD24 MSC3. To build toxto, pBAD24_6hisp20_flagcry11Aa should contain the correct DNAs sequence (every verification step has to be validated).

To obtain our constructions, we amplified our DNAs sequence with primers as shown in the design part.

Figure 37: Verification PCR amplification of DNAs sequence and these extensions for SLIC cloning.a. Control – (Absence of DNAs, no band expected); b, c, d, e. Amplification of 6hisp20_flagcry11Aa (expected 2500 bp), flagcry11Aa (expected 1847 bp), strepcyt1Aa (expected 775 bp) and 6hisp20 (expected 568 bp); L (SMART Ladder, its size is shown on the right side).

The results above were obtained. Expected bands were visualized for 6hisp20_flagcry11Aa, flagcry11Aa, strepcyt1Aa and 6hisp20 constructions. Amplification of DNAs sequences with primers were validated, thus it is possible to do the SLIC cloning in the vector pBAD24 MSC3 at the NcoI restriction site. The SLIC cloning was done as shown in the protocol.

After the SLIC cloning, E. coli DH5⍺ were transformed. Obtained clones were selected first on LB agar plate containing ampicillin. Then, colony PCR were carried out with primers used to amplify our DNAs sequences and with pBAD24 primers. It helped to verify the correct integration of DNAs sequences in the pBAD24 vector.

Figure 38: Colony PCR results. Amplifications of sequences were done on the vector and insert sequences. DNAs were separated on agarose gel 1% for 30min. DNAs were revealed using the RedGel solution. Different clones were tested, the SMART Ladder (L) was used and the empty pBAD24 vector was used as a negative control. We had some issues with the SMART Ladder.

(Top-left). pBAD24_6hisp20 construction: amplification was done with Ebm158 and o2116. The expected size is 630bp. Every clone had the expected size.

(Top-right). pBAD24_flagcry11Aa construction: amplification was done with Ebm158 and o2112. The expected size is 2000 bp. 9 clones had the expected size.

(Bottom-left). pBAD24_strepcyt1Aa construction: amplification was done with Ebm158 and o2114. The expected size is 837 bp. Every clone had the expected size.

(Bottom-right). pBAD24_6hisp20_flagcry11Aa construction: amplification was done with Ebm158 and o2116. The expected size is 630 bp. 7 clones had the expected size.

Thus, out of all the clones obtained from all the constructions, only 2 or 3 clones were selected to carry out a second verification by digestion profile.

Figure 39: Digestion profile results. After plasmid digestion using EcoRI and PstI, the digestion profil was obtained after D NAs were separated on agarose gel 1% for 30min. DNAs were revealed using the RedGel solution. Different clones were tested (restricted with 1 or 2 enzymes), the SMART Ladder (L) was used and the empty pBAD24 vector was used as a negative control.

(Top-left). pBAD24_6hisp20 construction:The expected size of the digested vector with one enzyme is 5113 bp and 4500 bp and 609 bp using two restriction enzymes. The clone II has the expected size in both cases.
(Top-right). pBAD24_6hisp20_flagcry11Aa construction: The expected size of the vector restricted with one enzyme is 7 kbp and restricted with both we expected for 4500 bp and 2500 bp bands. The clone a and u has the expected size in these both cases.
(Bottom). pBAD24_flagcry11Aa (clones 2 & 3), pBAD24_strepcyt1Aa (clones K & N) and pBAD24_6hisp20_flagcry11Aa (clones i & j) constructions:

This figure combines many different restriction profil of different constructions. The pBAD24_6hisp20_flagcry11Aa had negative results because of somes issues.

Clones K and N : The expected size of the vector restricted with one enzyme is 5320 bp and restricted with both we expected for 4500 bp and 816 bp bands. The clone K has the expected size in these both cases.
Clones 2 and 3 : The expected size of the vector restricted with one enzyme is 6500 bp and restricted with both we expected for 4500 bp and 2000 bp bands. The clone j has the expected size in these both cases.

Thus out of all the clones obtained from all the constructions, only 1 clone was selected to carry out the sequencing, to analyse if there is no mutation. Sequencing results are shown below:

Clone K of pBAD24_strepcyt1Aa: Results from sequencing with pBAD24 primer in forward allows sequencing the all gene cyt1Aa and there is no mutation in the gene: Fw cyt1Aa.
Clone u of pBAD24_6hisp20_flagcry11Aa: Result from sequencing with pBAD24 primer in forward allows sequencing all the p20 gene and a bit of the cry11Aa gene. Reverse sequencing permits to obtain all DNAs sequences. A mutation at the 1090nt seems to change the A to C nt. But the spectre analysis shows there is no mutation. Obtaining this clone permits us to make the toxto construction in pBAD24.
Sequence p20cry11Aa.
Fw p20cry11Aa.
Rev p20cry11Aa.
Clone II of pBAD24_6hisp20: Results from sequencing with pBAD24 primer in forward allows sequencing the all gene p20 and there is no mutation in the gene. A mutation at the 1090nt seems to add a G nt. But the spectre analysis shows there is no mutation, it is only a long G signal and no 2 Gnt.
Sequence p20.
Fw p20.
Clone 2 of pBAD24_flagcry11Aa : Results sequencing is not shown, on every sequencing result a mutation is visible at the end of the cry11Aa gene and generates a STOP codon. We tried many times to obtain a clone of this construction (repeat every step, from the SLIC cloning) but we never obtained a clone without mutation. The pBAD24_flagcry11Aa construction was abandoned.

To create toxto (pBAD24_6hisp20_strepcyt1Aa_flagcry11Aa) construction, we needed the pBAD24_6hisp20_flagcry11Aa. This plasmid pBAD24_6hisp20_flagcry11Aa was used as vector to adding strepcyt1Aa gene with adequates extensions (as shown in the design). The cyt1Aa DNA sequences was added at the KpnI restriction site using the SLIC cloning method. As for the other constructions, we transformed the toxto plasmid in E. coli DH5a and the same verification and selecting step were done.

Figure 40: Verification step of toxto construction:

(Top-left). Verification PCR amplification of DNA sequences and these extensions for SLIC cloning:Amplification was done with o2139 and o2135 primers. Expected size for “for SLIC toxto cyt1Aa” is +/- 840 bp. Smart Ladder (L) was used.
(Top-right). PCR on colony result: DNAs sequences are amplified with o2122 and Ebm159 primers. The negative control (C-) is the vector empty and the positive control (C+) is pBAD24_6hisp20_flagcry11Aa. Expected size of toxto is 1053bp.
(Bottom). Digestion profile results: After plasmid restriction using KpnI, NcoI and PstI enzymes, the digestion profil was obtained after DNAs were separated on agarose gel 1% for 30min. DNAs was revealed with the RedGel solution. The negative control (C-) is the vector empty and the positive control (C+) is pBAD24_6hisp20_flagcry11Aa. The expected size for toxto is : 4500bp (vector empty size), 2500bp (6hisp20_flagcry11Aa sequence) and 816 pB (strepcyt1Aa sequence).

After the SLIC cloning and the transformation of the pBAD24_toxto into E. coli DH5a, only one clone was selected on the agar plate added to ampicillin. Bands observed on the PCR on colony and digestion profil of toxto are expected bands. The positive and negative controls also have the expected bands.

Thus, we carry out the toxto sequencing in forward and reverse. Results are shown below. Unfortunately there is a gap in the sequencing result, 538bp are not covered by the sequencing, and results are not reliable.

The 6hisp20 gene is all sequenced, there is no mutation.
2 mutations are observed in the strepcyt1Aa gene who created no modification of the codon. Indeed, the first mutation (3954G:A) generated no mutation the codon 1318Pro:Pro the same for the second mutation (4141T:C) the 1381 Leu:Leu.
The flagcry11Aa gene were not all sequenced. Was sequenced from the beginning 1904 bp (START) to 3037bp and from 3575 to 3859 bp (STOP codon). But the sequenced parts have no mutation. Thus, to be sure of the sequencing result we have to sequence the middle part of flagcry11Aa, the primer has to be designed.
Fw toxto
Middle toxto
Rev toxto

Every plasmid constructions obtained were transformed in E. coli MG1655 to allow the right expression of these genes.

Biochemestry

Figure 41: General scheme. Overview of the biochemestry part.

Expression tests

First, we wanted to highlight protein productions in all of our constructions. For that we produced our proteins in E. coli MG1655.

Figure 42: Protein production.Strep-Cyt1Aa, Flag-Cry11Aa and 6his-P20 were expressed by MG1655 E. coli strain hold pBAD24_strepcyt1Aa, pBAD24_6hisp20 and pBAD24_6hisp20_flagcry11Aa plasmids. When bacteria were in exponential phase (OD(600) = 0.4), pBAD24 was induced with Arabinoses 0.1%. E. coli is cultivated at 37°C stirring in LB Ampicillin medium. These proteins were revealed by Westernblot using antibody anti-strep II (a), antibody anti-histidine (b, c.) or antibody anti flag tag (c.). Molecular size markers are indicated in kDa “RulePAGE Ladder”.

(Top-left). pBAD24_strepcyt1Aa induced or no induced sample. OD 1 was taken after 4h of induction. The expected size for Cyt1Aa is 28kDa. T+ Strep is a positive control of antibodies.
(Top-right). pBAD24_6hisp20 induced or no induced sample. OD 1 was taken after 4h of induction. The expected size for P20 is 20kDa. 6hisT7 is a positive control of antibodies and pBAD24 a negative control.
(Bottom). pBAD24_6hisp20_flagcry11Aa induced (Ni) and induced at different times samples (2h, 3h or 4h). The expected size for P20 is 20kDa and 70kDa is the expected size for Cry11Aa. This experiment was not done with the clone u, but with a clone with a mutation in its cry11Aa sequence, it was done to illustrate.

In the construction of pBAD24_strepcyt1Aa, Strep-Cyt1Aa is expressed and have the expected size. For the pBAD24_6hisp20 no 6his-P20 proteins are visibles, and the positive control reacts to the antibody anti 6his. The expression tests of 6his-P20 is not the first one, and everytime P20 expression in pBAD24_6hisp20 is not shown. The expression of 6his-P20 and Flag-Cry11Aa in the pBAD24_6hisp20_flagcry11Aa construction was not shown in the clone u because the westenblot failed, thus a mutated version of the construction was expressed. When gene transcription is induced, we can see huge bands of 6his-P20 at the expected size (20kDa) and bands at +/- 60kDa corresponding to the mutated form of Cry11Aa. In this condition we can see P20 and Cry11Aa.

SDS PAGEs were not shown because we cannot see bands corresponding to an overproduction of protein. Thus we had to find the best conditions for overproduction to have the maximum protein and allow to have a good protein yield. For the rest of the experimentations we were focused on the Strep-Cyt1Aa production, because its gene construction was the only well advanced.

Optimization of the overproduction

Several conditions were tested to optimised the Strep-Cyt1Aa production:

Different concentration of arabinose : 0%, 0.2%, 0.5% and 1%.
Bacteria stage of development : induction OD(600) = 0.4 or 1.
Time of induction : overnight or 5 hours.
Temperature of growth : 37°C, 28°C or 16°C.

After testing theses conditions, the proteins could thus be revealed by SDS PAGE migration and a Western Blot using anti strep tag II antibodies were done. Results are shown below.

A

B

C

Figure 43: Optimization of Strep-Cyt1Aa production. Strep-Cyt1Aa toxins were expressed by E. coli MG1655 holding pBAD24_strepcyt1Aa plasmid with different growth conditions.

A. Overnight production. [Arabinose] 0% to 1%, 16°C or 28°C temperature and OD(600) induction stage of0.4.
B. Overnight production. [Arabinose] 0% to 1%, 16°C or 28°C temperature and OD(600) induction stage of 1.
C. 5h of induction. [Arabinose] 0% to 1%, 37°C temperature and OD(600) induction stage of 0.4 or 1.
Strep-Cyt1Aa expressed was revealed on SDS-PAGE colored by Coomassie Blue (letf) or by Western Blot using antibody anti-strep II (right).

First, if we are comparing every test condition, we do not have significant production of Strep-Cyt1Aa, the expected size for Strep-Cyt1Aa is 28kDa. Also, the Western Blot revelation suggests that our antibodies are not specific for the Strep-tag, it is difficult to well analyse the Western Blot. On a nitrocellulose membrane there are 2 conditions, surrounded in green is the most significant Strep-Cyt1Aa production of the both conditions.

Among all conditions, we determine the “28°C overnight production with [Arabinose] 1% at OD600 induction stage = 0.4” as the best condition to express our proteins.

For the rest of the project we stayed in these conditions of production.

Cyt1Aa purification

We wanted to know the role of Cyt1Aa : verified the hemolytic effect, and verified its toxicity against mosquitoes. To do it, we tried to overproduce the Strep-Cyt1Aa protein and purified it. The purification was according to the purification protocol and different parts were analysed by SDS PAGE and Western Blot methods.

Figure 44: Strep-Cyt1Aa production.

Strep-Cyt1Aa toxins was expressed by E. coli MG1655 strain holding pBAD24_strepcyt1Aa plasmid. When bacteria were in exponential phase (OD(600) = 0.4), pBAD24 was induced with Arabinoses 1%. E. coli was cultivated at 28°C stirring, overnight in 2 liters of LB.

Strep-Cyt1Aa expressed was revelated on SDS-PAGE colored by Coomassie Blue (a) or by Westernblot using antibody anti-strep II (c), transfert quality preview is shown with red ponceau coloration (b) . Molecular size markers are indicated in kDa “RulePAGE Ladder”. The “No induced”, “total fraction”, “soluble part” and “unsoluble part” samples represent, respectively, E. coli MG1655 strain holding pBAD24_strepcyt1Aa plasmid no induced by arabinose, induced part holding 1UDO of bacteria, the cytoplasmic material after lysis and the membrane material solubilised. A positive sample of antibody anti-strep was deposited on the gel for the Westernblot (b., c.).

The Strep-Cyt1Aa waiting band is 28kDa. On the SDS-PAGE colored by Coomassie Blue (a) no band can be clarely identify as our toxin. In fact, the no induced part got similar bands on gel than induced fractions.

Despite the bad quality of red ponceau coloration (b) the transfer on nitrocellulose membrane is correctly visible.

After revelation of the westernblot using the antibody anti-strep II, are visibles 2 bands at 25 kDa corresponding respectively “total fraction” and “soluble part”. The positive control is not visible, possibly a too low quantity of control was deposited on gel. These 2 bands are Cyt1Aa (supposed 28kDa), this test highlights the fact that Cyt1Aa can be expressed in our strain, and Cyt1Aa is soluble in the cytoplasmic part.

From the experimentation, bacteria from these 2L of LB was separated from the medium and was lysed as described in the material and method to obtain firstly the cytosol that will be purified on affinity chromatography strep-tactin column and on the other hand membranes.

Figure 45: Strep-Cyt1Aa purifaction.Strep-Cyt1Aa toxins was purified thanks to a affinity chromatography strep-tactin column and revelated on SDS-PAGE colored by Coomassie Blue (1) or by Westernblot using antibody anti-strep II (3), transfert quality preview is shown with red ponceau coloration (2) . Molecular size markers are indicated in kDa to the left of the gel. The “Soluble part”, “Not retained”, “Wash” and “Elution” (E0 to E3) samples represent, respectively, what is loaded on the affinity chromatography column, what is not retained before and after washes, and what is eluted at 500 mM biotin. A positive sample of antibody anti-strep was deposited on the gel for the Westernblot (2., 3.).

On the SDS-PAGE colored by Coomassie Blue (1) no band can be clarely identify as our toxin for the soluble part. We are waiting for a 28kDa band. The no retained fraction have the same profil. First as shown with the expression test, the SDS-PAGE colored by Coomassie blue is not an element that can clearly show the presence of our toxin. Plus, we cannot clearly see a difference between total protein in the cytosol and no retained protein on the column. After the wash, different the different elution show that the column is not washed enough. Few bands are still visible after the final elution. But, is visible on the 4 elutions, a 20kDa bands. It is not the waiting bands for Cyt1Aa. Two hypothesis can be done : Cyt1Aa have a lower molecular weight than expected or this is an other protein. The first hypothesis seems false because on the expression test westernblot gel we can watch Cyt1Aa at his expected size (+/- 28kDa).

If we are looking at the red ponceau coloration (2), the transfer on nitrocellulose membrane is correct, and this 20kDa band is still visible.

But after revelation of the Western blot using the antibody anti-strep II (3), are visibles bands at 25 kDa corresponding to every sample, except the wash. The positive antibody control is also visible. Finally, bands visibles are bands expected. Two observation can be done. First, the purification have to be improved. In fact, almost all proteins wasn’t retained on the column, just some proteins are eluted. But the purification is working, we are able to obtain purified toxins. Second, bands saw on SDS-PAGE colored by Coomassie Blue (1) are not the same on westernblot. That’s mean, we are expressed our toxin in low part and an other protein that can’t be revelated with antibody anti-strep tag. As explain before, the expression of our toxin is too low to be able to see it by only using SDS-PAGE colored by Coomassie Blue.

Next steps

Figure 6: General scheme. Overview of the next step part.

The next step of our project in the short term is to improve the purification protocol and more precisely the overproduction procedure. There is still a low concentration of protein visible on the SDS PAGES gel (1 UDO was deposed). This improvement can allow us to try the hemolytic effect of Cyt1Aa on blood cells or test its toxicity against mosquitoes.

Other, we can raise that there always was a low OD(600nm) after strep-cyt1Aa production compared to the pBAD24 empty vector (OD +/- 3 compared to 7) a fter the overnight production. If we referred to literature, Cyt1Aa is toxic for E. coli sp. productrice cell, and P20 chaperon warns the toxics effects.

In the long term, we planned to obtain the toxto construction and clone this construction into the timer lysis device instead of caa gene coding for ColA as presented in the schema and tried again to characterize the timer process with our toxins.

References

Cascales, E. et al. Colicin biology. Microbiol. Mol. Biol. Rev. MMBR71, 158–229 (2007).
Cavard, D. Role of Cal, the colicin A lysis protein, in two steps of colicin A release and in the interaction with colicin A-porin complexes. Microbiol. Read. Engl. 150, 3867–3875 (2004).
Lloubes, R., Baly, D. & Lazdunski, C. The promoters of the genes for colicin production, release and immunity in the ColA plasmid: effects of convergent transcription and Lex A protein. Nucleic Acids Res . 14, 2621–2636 (1986).
VirtualFootprint - PRODORIC

Team:Aix-Marseille/experiments