Team:Nantes/Results

Results Results

Introduction

The goal of our project Bio’Clock is to control gene expression in time by using the sugar hierarchy naturally present in E.coli (see project description and design). To begin building this tool we designed 6 different constructs. Four simple insert constructs, each containing a sugar-specific promoter associated to a fluorescent reporter gene :

GFP : Green Fluorescent Protein
CFP : Cyan Fluorescent Protein
RFP : Red Fluorescent Protein
YFP : Yellow Fluorescent Protein
pLAC : Lactose specific promoter
pARA : Arabinose specific promoter
pSRL : Sorbitol specific promoter
pRIB : Ribose specific promoter

Two double insert constructs, each containing 2 sugar-specific promoters, each associated with their own fluorescent reporter gene :

The first part of our results demonstrate how we determined if the promoters in each of these constructs functioned properly. We then explain how we established the experimental link between the concentration of sugar in the medium and the promoter activity, before presenting how we tested the specificity of each promoter to it’s associated sugar. And finally, we present the work we did to optimize our part pARA, before finishing on the future plans for this project.


I - Testing of the functionality of our designed parts


After transforming K12 MG1655 E.coli with our 6 different constructs, we measured by spectrofluorometry the fluorescence of the transformed bacteria cultivated in M9 medium in the presence of saturating concentration (0.2%) of the associated sugar (for the simple insert) or the associated sugars (for the double inserts). The fluorescence studied in the cases below has been normalized with the Optical Density (OD).


a) Simple insert constructs


We transformed K12 MG1655 E.coli with 4 different simple insert constructs :
  • pLAC-GFP : plasmid containing a lactose specific promoter associated with a GFP reporter gene
  • pARA-CFP : plasmid containing a arabinose specific promoter associated with a CFP reporter gene
  • pSRL-RFP : plasmid containing a sorbitol specific promoter associated with an RFP reporter gene
  • pRIB-YFP : plasmid containing the ribose specific promoter associated with a YFP reporter gene

Our spectrofluorometric measurements were carried out on the transformed bacteria cultivated in a M9 medium in 2 different conditions : with 0.2% of the specific sugar, and without sugar.


● pLAC-GFP: E.coli:

Figure 1: GFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pLAC in the presence (green) or absence (grey) of Lactose at 0,2%. Experiments were conducted during 17 hours.

The activity of pLAC is relative to the expression of GFP fluorescence. In the condition without sugar we observe little increase in GFP fluorescence and therefore little activity of the pLAC promoter. This control shows that there is very little promoter activity in the absence of sugar. This validates our control. In the condition where 0.2% lactose was added, we observe significant GFP levels compared to the condition without sugar starting at hour 1. The level of GFP increases linearly over time. The reporter gene is expressed following the activation of the pLAC promoter in the presence of lactose. We observe an increase in GFP during the first 6 hours of the bacterial culture, but from the sixth hour this increase slows down. This verifies the proper functioning of our designed pLAC promoter.


● pARA-CFP: E.coli :


Figure 2:CFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pARA in the presence (cyan) or absence (grey) of arabinose at 0,2%. Experiments were conducted during 17 hours.

The activity of pARA is relative to the expression of CFP fluorescence. In the condition “without sugar”, there is no increase in CFP. There is therefore no promoter activity. This validates our control. In the presence of 0.2% arabinose, we observe a strong increase in CFP expression starting in the second hour of the bacterial culture. The CFP in this condition is significantly different than the CFP in the absence of sugar. This verifies the proper functioning of our designed pARA promoter.


● pSRL-RFP E.coli :


Figure 3: RFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pSRL in the presence (red) or absence (grey) of sorbitol at 0,2%. Experiments were conducted during 17 hours.

The activity of pSRL is relative to the expression of RFP fluorescence. We do not observe a significant difference of RFP fluorescence between both conditions, with or without sorbitol. Either (i) there is a RFP fluorescence in both conditions meaning that the promoter pSRL is active even in the absence of sugar, or (ii) we are in presence of an experimental error.


● pRIB-YFPE.coli :


Figure 4: YFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pRIB in the presence (yellow) or absence (grey) of ribose at 0,2%. Experiments were conducted during 17 hours.

The activity of pRIB is relative to the expression of YFP fluorescence. In the condition “without sugar”, there is little increase in YFP. There is therefore no significant promoter activity in the absence of sugar. This validates our control. With 0.2% ribose, starting from the 4.5th hour, we observe a strong increase in YFP which is significantly different than the YFP observed in the condition without sugar. Therefore, we can confirm that ribose activates the expression of YFP by increasing the activity of pRIB. This verifies the proper functioning of our designed pRIB promoter.


b) Double insert constructs


We transformed K12 MG1655 E.coli with 2 different double insert constructs :
  • “pLAC-GFP + pSRL-RFP” : plasmid containing 2 sugar specific promoters pLAC and pSRL associated with GFP and RFP reporter genes respectively
  • “pARA-CFP + pRIB-RFP” : plasmid containing 2 sugar specific promoters pARA and pRIB associated with CFP and RFP reporter genes respectively

Our spectrofluorometric measurements were carried out on the transformed bacteria cultivated in a M9 medium in 3 different conditions : with 0.2% of the first specific sugar, with 0.2% of the second specific sugar, and without sugar.


● pLAC-GFP/pSRL-RFP E.coli:

The activity of pLAC is relative to the expression of GFP fluorescence, and the activity of pSRL to RFP fluorescence.

Measurements done in the presence of 0.2% lactose and without sugar:


Figure 5: GFP(a) and RFP(b) fluorescence in K12 bacteria with the double insert (pLAC-GFP + pSRL-RFP), in the absence (grey) or presence of 0.2% lactose (green for GFP or red for RFP). Experiments were conducted during 17 hours

In figure 5a, as expected, we observe a significant difference between GFP levels in the presence of lactose and without the sugar. The pLAC in our double insert construction is not activated in the absence of sugar (as previously shown for the single insert construction, Figure 1).
Likewise, in Figure 5b, in the absence or presence of lactose there is no difference in the emission due to RFP. Lactose did not induce an increase in the pSRL promoter activity.
Therefore, pLAC is significantly activated in the presence of lactose, however the presence of lactose does not affect the activity of pSRL

Measurements done in the presence of 0.2% sorbitol and without sugar:


Figure 6:GFP(a) and RFP(b) fluorescence in K12 bacteria with the double insert (pLAC-GFP + pSRL-RFP), in the absence (grey) or presence of 0.2% sorbitol (green for GFP or red for RFP). Experiments were conducted during 17 hours

In Figure 6a we notice an increase in GFP fluorescence with 0.2% sorbitol and no increase of the GFP fluorescence in the absence of the sugar. This means that sorbitol is able to activate pLac promoter and this corroborates the data published in the paper of Dr. Guy Aidelberg et al (1).

In figure 6b, we notice that sorbitol even in saturating concentrations (0.2%), do not activate pSrl promoter and hence RFP fluorescence in the context of a double insert with the pLac promoter. There could be two explanations for this observation. First, it could be that our pSrl promoter is not functioning. This would explain our data in figure 6b and 3. Indeed, the promoter in the single insert construction (see above) is not activated by sorbitol

The fact that sorbitol strongly activates the pLac promoter as shown by GFP fluorescence, could also mean that sorbitol would preferentially activates pLac promoter in place of pSrl promoter. This result has never been shown before because no binary combination of both of these promoters were tested before. Our data could suggest that there is indeed a hierarchy in the activation of promoters. A certain amount of Sorbitol will activate pLac if the latter is present. We cannot so far say if both promoters could be activated and to which level one compared to the other because our pSrl insert does not seem to function


● pARA-CFP/pRIB-RFP E.coli :

The activity of pARA is relative to the expression of CFP fluorescence, and the activity of pRIB to RFP fluorescence.

Measurements done in the presence of 0.2% arabinose and without sugar:

Figure 7: CFP(a) and RFP(b) fluorescence in K12 bacteria with the double insert (pARA-CFP + pRIB-RFP), in the absence (grey) or presence of 0.2% arabinose (cyan for CFP or red for RFP). Experiments were conducted during 17 hours.

We observe in Figure 7a an increase of CFP fluorescence in the presence of 0.2% arabinose after a latency of about 4.5h and no increase of the same fluorescence in the absence of the sugar. This validates our control (without sugar) but also indicates that the pARA promoter in the context of the double insert is indeed activated by arabinose.

Figure 7b shows that for the first 8 hours, we do not have a significant difference in the fluorescence of RFP between the two conditions tested, i.e with or without 0.2% arabinose. After 8h, there is an increase in RFP fluorescence relative to that observed in the absence of the sugar, indicating an increase in the pRIB promoter activity induced by the sugar. As detailed in the Dr. Guy Aidelberg et al (1) paper, arabinose can indeed very slightly activate pRIB promoter. This hence corroborates their observations which was done with a single pRIB insert. It is noteworthy that in our case we have performed the measurements every 30 min during 17h while Aidelberg et al established activation level of pRIB by arabinose on a single data point (at mid-exponential growth). Interestingly, the delay for the activation of pRIB by arabinose is 8h, twice the time needed for it’s activation with ribose only (see figure 4). This is an original result. Surprisingly, we note a sharp increase in RFP fluorescence in the microplate wells with the double insert (pARA/pRIB) very early during bacterial growth from t=0h to t=2h. This increase may be artifactual due to some experimental error

Measurements done in the presence of 0.2% ribose and without sugar:

Figure 8: CFP(a) and RFP(b) fluorescence in K12 bacteria with the double insert (pARA-CFP + pRIB-RFP), in the absence (grey) or presence of 0.2% ribose (cyan for CFP or red for RFP). Experiments were conducted during 17 hours

We observe a significant linear increase of CFP in the presence of ribose when compared to the experiment without sugar, validating our control.

In the presence of ribose (Figure 8a), CFP fluorescence increases linearly throughout the experiment and has a maximum level at around 17 000 a.u obtained at hour 17. This means that ribose is able to activate the pARA promoter. This corroborates with the experimental data published by Guy Aidelberg et al (1).

When pRIB activity is monitored in the presence or absence of 0.2% ribose in the context of a double insert with pARA, there is a difference that is observed between the RFP fluorescence as from 4.5h. This confirms that pRIB promoter is indeed activated by ribose. Interestingly seems to be an immediate activation of pARA by ribose while there is a delay in the activation of pRIB by the same and similar to what was observed in figure 7b (activation of pRIB with arabinose that exhibited a delay of 4.5h).

Summary

In the simple insert constructs, we conclude that 3 out of 4 of our parts (pLAC, pARA and pRIB) are significantly activated in the presence of their associated sugar compared to conditions without sugar, verifying the proper functioning of these parts. Unfortunately, we were not able to have pSRL construct to work properly.
In the case of the double insert constructs, we observed an activation of pLAC, pARA and pRIB in the presence of their respective sugars.
They are all activated but to a lesser degree in the presence of the other sugar, thus confirming the existence of cross-activation and a sugar hierarchy in E.coli.



II - The link between sugar concentration and promoter activity


For these tests we measured the fluorescence in K12 MG1655 E.coli transformed with each of our 6 constructs. The bacteria were cultivated in M9 medium with various different concentrations of the associated sugar (for simple insert constructs), or associated sugars (for double insert constructs). The fluorescence studied in the graphs below has been normalised with the Optical Density (OD).


a) Simple insert constructs


As a reminder, in our 4 simple insert constructs :
  • pLAC is associated to GFP
  • pARA is associated to CFP
  • pSRL is associated to RFP
  • pRIB is associated to YFP

● pLAC-GFP E.coli :

Figure 9: GFP Fluorescence normalised with OD for E.coli K12 MG1655 transformed with pLAC in selected concentrations of Lactose

The activity of pLAC is relative to the expression of GFP fluorescence. We observe a slight increase of GFP fluorescence in the condition without sugar, but the fluorescence level remains relatively insignificant when compared to the fluorescence in conditions where lactose is present. However we do not observe a significant difference of GFP expression between the different conditions containing lactose. The concentration of lactose does not seem to have a significant effect on the activity of our designed pLAC in the simple insert construct. However for this conclusion to be decisive we would need to carry out more experiments with different sugar concentrations


● pARA-CFP E.coli :

Figure 10:CFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pARA in selected concentrations of arabinose

The activity of pARA is relative to the expression of CFP fluorescence.
Without sugar, we observe a very low level of CFP fluorescence which allows us to validate the proper functioning of our pARA promoter in this condition. The difference of CFP expression between conditions with arabinose and the condition without sugar becomes significant around the fourth hour of bacterial culture.
From 14.5h, we notice a significant difference of CFP expression amongst the conditions containing 0.1 %, 0.05% and 0.01% of arabinose. However, the curve of the condition containing 0.02% of arabinose follows the same tendency as the previous curves, but it has enormous error bars. If we consider the results of this condition to be distorted, we can conclude that the intensity of the activation of pARA can be regulated proportionally to the concentration of arabinose in the medium


● pSRL-RFP E.coli :

Figure 11: RFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pSRL in selected concentrations of Sorbitol

The activity of pSRL is relative to the expression of RFP fluorescence. This figure shows no significant difference of RFP expression between all the different conditions This is coherent with our previous observations whereby we observed that pSRL activity was equivalent in absence and in presence of 0.2% sorbitol (see figure 3). The pSRL does not seem to be affected by the sugar in the medium, therefore we assume a dysfunctionality of this promoter in our constructs


● pRIB-YFP E.coli :

Figure 12: YFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pRIB in selected concentrations of Ribose.

The activity of pRIB is relative to the expression of YFP fluorescence.

In the condition without sugar, we observe a very low, stable and significantly different YFP fluorescence compared to the YFP in the other conditions (YFP max is around 10 000 a.u compared to 40 000 and 70 000 for the other conditions).
In the conditions with sugar, we notice the same curve tendencies with an increase in YFP fluorescence around the third hour, however there is no significant difference of YFP levels between these different conditions.
We do notice one anomaly for the condition 0.01% ribose. At hour 12, the YFP fluorescence strongly increases (from around 40 000 au to 70 000 a.u.) during the last 4 hours of the culture.
We suspect either a measurement anomaly on these plate wells, an error during the preparation of the medium for this condition, or an unidentified phenomenon (this experience was repeated twice and the same curve was observed both times).
The promoter pRIB was indeed activated in the presence of ribose during this experiment, however, does not seem sensitive to the change of concentration of this sugar. The intensity of the activity of pRIB in our construct does not seem to depend on the concentration in ribose present in the medium.


b) Double insert constructs


As a reminder, in our 2 double insert constructs :
  • pLAC-GFP-pSRL-RFP ; pLAC is associated to GFP, and pSRL to RFP
  • pARA-CFP-pRIB-YFP ; pARA is associated to CFP, and pRIB to RFP

● pLAC-GFP/pSRL-RFP E.coli :

The activity of pLAC is relative to the expression of GFP fluorescence, and the activity of pSRL to RFP fluorescence.

Measurements done in the presence of different concentrations of lactose and without sugar:

Figure 13:GFP(a) and RFP(b) normalized with OD for E.coli K12 MG1655 transformed with pLAC/pSRL in selected concentrations of Lactose

Without sugar, the GFP fluorescence does not significantly change (very slight increase from 400 au to 500 a.u.), and remains significantly inferior to the GFP fluorescence in presence of lactose.
The GFP signal is lower with 0.01% lactose in comparison to the signals observed with higher concentrations of the sugar (figure 13a). There is no significant difference in GFP fluorescence levels between experiments with lactose concentrations at 0.02%, 0.05% and 0.1%. Maximal activation of pLAC promoter seems to be reached as from 0.02% lactose.
The RFP fluorescence does not seem to have any increase over time, or to have any significant difference between the different conditions. Therefore, there is no significant pSRL activity due to the presence of lactose. It might be because the pSRL promoter is not functional as it was seen in Figures 3, 5b and 6b. So we can conclude that we were not successful to have a function pSRL promoter.

Figure 14:(a)GFP and (b)RFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pLAC/pSRL in selected concentrations of Sorbitol

In Figure 14a we observe that the GFP without sugar and with 0.01% sorbitol are not significantly different. 0.01% sorbitol is not enough to activate pLAC promoter. Significant activation of this promoter is only observed after 6 hours for concentration of sorbitol of 0.02% or above and would be concentration dependent. Though maximal activation is reached with 0.05% sorbitol
In figure 14b we observe no significant difference between any of the sorbitol concentrations. As mentioned above for Figure 13b, our results seem to indicate that pSRL promoter is unfortunately not functional


● pARA-CFP/pRIB-RFP E.coli :

The activity of pARA is relative to the expression of CFP fluorescence, and the activity of pRIB to RFP fluorescence.

Measurements done in the presence of different concentrations of arabinose and without sugar:

Figure 15: CFP (a) and RFP (b) Fluorescence normalized with OD for K12 MG1655 transformed with pARA/pRIB in selected concentrations of Arabinose

We observe in Figure 7a an increase if CFP fluorescence in the presence of 0.2% arabinose after a latency of about 4.5h and no increase of the same fluorescence in the absence of the sugar. This validates our control (without sugar) but also indicates that the pARA promoter in the context of the double insert is indeed activated by arabinose.

In the conditions without sugar for both figures, we observe a stable fluorescence. Both promoters do not seem to be significantly activated in the absence of arabinose.
In both graphs, the fluorescence in the presence of arabinose become significantly different from the fluorescence in the absence of sugar from around hours 7 and 8 of the culture.
Our data do not show a concentration dependant variation of GFP or RFP fluorescence. Both pARA and pRIB promoters are very active at low concentrations of arabinose. This is an original result

Measurements done in the presence of different concentrations of ribose and without sugar:

Figure 16: CFP (a) and RFP (b) Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pARA/pRIB in selected concentrations of Ribose

We observe a significant linear increase of CFP in the presence of ribose when compared to the experiment without sugar, validating our control.

In figure 16a, CFP fluorescence seems not to vary significantly between any of the conditions. There does not seem to be any particular activity of the pARA promoter in the presence of ribose in the medium.
In figure 16b, there is a non significant variation in RFP fluorescence in the absence of ribose. In presence of the sugar, the RFP fluorescence is increasing significantly after 6.5-7h. The increase is concentration dependant.
This experimental setup where the two promoters were combined and tested against ribose is original. Here the highest concentration of ribose used was 0.1%. It seems to be insufficient to activate the pARA promoter while it was able to do it at a concentration of 0.2%. We can hypothesize that in the context of this double insert, ribose preferentially activates its own promoter pRIB and not pARA, when it is at a concentration of 0.1% or below. At higher concentrations (like at 0.2%), the pRIB promoter is saturated and some ribose is available for activating pARA (Figure 8).

Summary

In our simple insert constructs, we showed that pARA is the only promoter that responded to the concentration level of its cognate sugar. For two other promoters, pLAC and pRIB, a high level of activation was rapidly reached with a low concentration of their cognate sugars. We were not successful to demonstrate that our pSRL construct was functioning correctly. In our double constructs, pLAC/pSRL and pARA/pRIB, our results corroborates with known hierarchy of the sugars. Sorbitol indeed was able to activate pLAC promoter with concentrations higher than 0.02%. Arabinose is able to cross-activate pRIB and reciprocally, ribose is able to activate pARA but only at high concentrations and will preferentially activate its cognate promoter.



III - The specificity of each promoter to their associated sugar


To complete our experiments we tested the promoter activity of the simple insert constructs pLAC-GFP and pRIB-YFP, in K12 MG1655 E.coli, in the presence of saturation concentration of different sugars. Due to a lack of time, we were unable to carry out these experiments on our other constructs pARA and pSRL. The primary purpose of these measurements is to confirm already published cross-activation results.


● pLAC-GFP E.coli :

Figure 17: GFP fluorescence normalized with the OD in E.coli K12 MG1655 bacteria transformed with pLAC-GFP construct in the presence of saturating amounts of different sugars

The activity of pLAC is relative to the expression of GFP fluorescence.
In absence of any sugar, we observe no increase of GFP fluorescence during the 15 hours of culture . This validates our negative control.
On the other hand, all tested sugars were able to induce an increase of GFP fluorescence hence activate the pLAC promoter. This corroborates already published results. At this stage we cannot conclude on the differences between the sugars regarding the fluorescence levels reached after 5, 10 or 15 hours since we did not quantify the amount of proteins and only normalized the data with bacterial growth OD values.


● pRIB-YFP E.coli :

Figure 18: YFP fluorescence normalized with the OD in E.coli K12 MG1655 bacteria transformed with pRIB-YFP construct in the presence of saturating amounts of different sugars.

The activity of pRIB is relative to the expression of YFP fluorescence.

In the condition without sugar we notice an increase in YFP over the first 5 hours which then stagnates after hour 5. This is similar to what was observed in Figures 7b and 8b where without sugar, there is a slight but sharp increase of RFP fluorescence corresponding to pRIB activity over the first 5 hours before stabilization of the fluorescence. We can hypothesize that this increase could be due to internal ribose produced by the bacterie and that enters in the composition of its DNA or RNA.
Only three sugars, arabinose, sorbitol and ribose were able to induce an YFP fluorescence higher than the condition without sugar. Maximum activation of pRIB is reached with its cognate sugar ribose. On the other hand, activation of pRIB by lactose follow a very similar pattern as the negative control (in the absence of sugar). Hence, lactose does not seem to induce pRIB promoter activity. This last observation corroborates already published data by Guy Aidelberg et al (1).

Summary


pLAC seems to be activated by all 4 sugars. pRIB is only activated by 3 sugars (arabinose, sorbitol and ribose) and is not activated by lactose. These confirm cross-activation data published previously by Guy Aidelberg et al (1).
We hypothesize that internal bacterial ribose produced for DNA and RNA may explain the sudden and slight activation of pRIB promoter in the absence of any added sugar in the medium.



IV - The optimisation of pARA


One of our goals was to improve a part present in the iGEM “part registry”. In doing so, we wanted to continue working on sugar activated promoters. Therefore we chose 2 different promoters associated with the consumption of arabinose.

  • part:BBA I13453 → that we call “1” from this point
  • part:BBA K206000 → that we will call “3” from this point

We ordered these parts associated with a CFP reporter gene, as well as a second version of these parts with the consensus sequence of CRP (AAATGTGATCTAGATCACATT).
After transforming top10 E.coli we obtained 4 different strains to be tested :

  • “1” : pIDT pARAI13453 - CFP
  • “2” : pIDT pARAI13453 optimized - CFP
  • “3” : pIDT pARAK206000 - CFP
  • “4” : pIDT pARAK206000 optimized -CFP
  • “5” : pIDT pARAI13453 - GFP
  • “6” : pIDT pARAI13453 optimized - GFP


We then cultivated our bacteria in M9 medium with 0.2% arabinose. Our 4 different bacterial cultures were followed by spectrofluorometry for 17 hourss. The fluorescence observed in the graphs below has been normalized with the Optical Density (OD).


● Top10 E.coli 1 and 2

As a reminder, in this case the top10 E.coli was transformed with plasmids containing the pARA part I13453 either optimized with the consensus sequence of CRP (AAATGTGATCTAGATCACATT) (“2”), and non optimized (“1”).

Figure 19: CFP Fluorescence normalized with OD for E.coli T10 transformed with the part 1 and 2 in absence of any sugar (◇), in the presence of 0.2% arabinose (△) , 0,1% arabinose (☐), 0,05 % arabinose (+), 0.02% arabinose (○) and M9 rich medium (x)

Both strains express CFP, but we observe no difference between the optimised pARA part CFP (2) and the non optimised part (1). We can conclude that the optimisation by addition of the consensus sequence of CRP (AAATGTGATCTAGATCACATT) does not seems to have significant effect on the activity of the pARA I13453.


● Top10 E.coli3 and 4

As a reminder, in this case the top10 E.coli was transformed with plasmids containing the pARA part K206000 either optimized with the consensus sequence of CRP (AAATGTGATCTAGATCACATT) (“4”), and non optimized (“3”).

Figure 20: CFP Fluorescence normalized with OD for E.coli T10 transformed with the part 3 and 4 in absence of any sugar (◇), in the presence of 0.2% arabinose (△) , 0,1% arabinose (☐), 0,05 % arabinose (+), 0.02% arabinose (○) and M9 rich medium (x)

Both strains express CFP, but we observe no difference between the optimised pARA part CFP (3) and the non optimised part (4). We can conclude that the optimisation by addition of the consensus sequence of CRP (AAATGTGATCTAGATCACATT) does not seems to have significant effect on the activity of this promoter too.


● Top10 E.coli 5 and 6

As a reminder, in this case the top10 E.coli was transformed with plasmids containing the pARA part I13453 either optimized with the consensus sequence of CRP (AAATGTGATCTAGATCACATT) (“5”), and non optimized (“6”).

Figure 21: GFP Fluorescence normalized with OD for E.coli T10 transformed with the part 5 and 6 in the absence of any sugar (◇), in the presence of 0.2% arabinose (△) , 0,1% arabinose (☐), 0,05 % arabinose (+), 0.02% arabinose (○) and M9 rich medium (x).

Both strains express GFP, but we observe no difference between the optimised pARA part GFP (5) and the non optimised part (6). We can conclude that the optimisation by addition of the consensus sequence of CRP (AAATGTGATCTAGATCACATT) seems to not have a significant effect on the activity.

Summary


Here we have shown that the consensus sequence of CRP (AAATGTGATCTAGATCACATT) seems to not have changed the level of expression of our different pARA promoters. However, it seems that the M9 rich medium have an impact on the activation level of pARA. We do not have an explanation to why the M9 rich medium seems to enhance the activation of our promoters, an experimental error may be possible.



V - Binary combinations of sugars with simple and double inserts

Here we tested the activities of single or double inserts with binary combinations of sugars. The tested single inserts are pLAC-GFP, pARA-CFP and pRIB-YFP. The only tested double insert was pARA-CFP/pRIB-RFP. The primary purpose of these experiments is to analyse the activity of each promoter in the presence of 2 sugars at a time: it’s own specific sugar and another. We will see if the sugar-hierarchy is still present in our constructions in these conditions, and how a sugar,higher or lower in the hierarchy impacts the expression of the different promoters.


● pLAC-GFP E.coli :

As a reminder, here E.coli K12 MG1655 was transformed with pLAC-GFP. We monitored pLAC activity by measuring GFP fluorescence in absence or presence of combinations of sugars. Results are provided below (Fig. 22) .

Figure 22: Figure 22. GFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pLAC-GFP in absence of any sugar (■), in presence of 0.2% lactose combined with M9 rich medium (●), 0.2% arabinose (○), 0.2% sorbitol (△) and 0.2% ribose (+). Experiments were conducted during 17 hours

Our results show that pLAC promoter is activated by all tested binary combinations of sugars which all contained 0.2% lactose. The increase in activity over time is mainly imputable to lactose. Addition of other sugars known to be lower in the hierarchy with respect to lactose did not induced an additional effect on pLAC activity. This confirms that pLAC is on the top of the hierarchy


● pARA-CFP E.coli :

As a reminder, here E.coli K12 MG1655 was transformed with pARA-CFP. We monitored pARA activity by measuring CFP fluorescence in absence or presence of combinations of sugars. Results are provided below (Fig. 23)

Figure 23: CFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pARA-CFP in absence of any sugar (■), in the presence of 0.2% arabinose combined with M9 rich medium (●), 0.2% lactose (○), 0.2% sorbitol (△) and 0.2% ribose (+). Experiments were conducted during 17 hours.

No change in CFP fluorescence is observed in the absence of sugar. This validates our negative control.
Contrasting results were obtained for the different combinations of sugars. In presence of saturating arabinose (0.2%) and M2 rich medium, CFP fluorescence slightly increased over time with a fluorescence level that was slightly above that of the negative control after 17 hours. A similar pattern is observed when 0.2% arabinose is combined with 0.2% lactose. This would suggest that M9 rich medium likewise lactose would have an inhibitory effect on pARA activity. Indeed, when arabinose is combined with others sugars lower in the hierarchy, i.e sorbitol or ribose, the activity of pARA promoter was much higher. With sorbitol, the CFP fluorescence increased steadily between 5 hours and 10 hours of culture to reach a plateau after the 10th hour. When combined with 0.2% ribose, the fluorescence pattern was different as it exhibited a steady linear increase over time.
These results show that ribose and sorbitol have a cumulative effect with arabinose that leads to a bigger activation of the pARA promoter.


● pRIB-YFP E.coli :

As a reminder, here E.coli K12 MG1655 was transformed with pRIB-YFP. We monitored pRIB activity by measuring YFP fluorescence in absence or presence of combinations of sugars. Results are provided below (Fig. 24).

Figure 24: YFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pRIB-YFP in absence of any sugar (■), in the presence of 0.2% ribose combined with M9 rich medium (●), 0.2% lactose (○), 0.2% arabinose (△) and 0.2% sorbitol (+). Experiments were conducted during 17 hours.

Here, for all tested conditions, there is a sharp increase in YFP fluorescence during the first 2.5 hours. This is similar to our previous results as shown in Figures 7b, 8b and 18. As mentioned above, we hypothesize that this could result from endogenous ribose produced by the bacteria for DNA and RNA that could activate pRIB promoter.
When 0.2% ribose was combined with M9 rich medium or 0.2% lactose, no additional increase of YFP fluorescence was observed when compared to the fluorescence in absence of sugar, indicating that these may strongly inhibit the action of ribose on pRIB promoter. On the other hand, sorbitol and arabinose when combined with ribose did not inhibit pRIB promoter activity. Combined with arabinose, YFP fluorescence levels reached slightly higher levels than when combined with sorbitol.


● pARA-CFP/pRIB-RFP E.coli :

As a reminder, here E.coli K12 MG1655 was transformed with a double insert containing pARA-CFP/pRIB-RFP. We monitored both pARA and pRIB activities by measuring CFP and RFP fluorescence respectively in absence or presence of combinations of sugars. Results are provided below (Fig. 25).

Figure 25: CFP Fluorescence normalized with OD for K12 MG1655 transformed with pARA-CFP/pRIB-RFP double insert in absence of any sugar (■), in the presence of 0.2% arabinose + 0.2% ribose (●), 0.2% arabinose + 0.2% lactose (○), 0.2% arabinose + 0.2% sorbitol (△), 0.2% ribose + 0.2% lactose (+) and 0.2% ribose + 0.2% sorbitol (◇). Experiments were conducted during 17 hours.

Results in figure 25 show that in absence of sugar, there is no significant change in CFP fluorescence. This validates the negative control for CFP fluorescence.
In presence of 0.2% arabinose combined with 0.2% lactose or 0.2% ribose, CFP fluorescence that is supposed to monitor pARA activity showed a steady increase as from 6 hours of culture. This suggests that neither lactose nor ribose inhibits pARA promoter activity.
Similarly, when 0.2% arabinose was combined with 0.2% sorbitol, a steady increase in pARA activity as from the 7.5th hour was also observed but with much lower levels of CFP fluorescence reached when compared to the combination with lactose or ribose.

The combination of ribose with lactose also induced a steady increase in CFP fluorescence as from 7.5 hours that reaches a level that is roughly 50% to that reached when arabinose was combined with lactose or ribose. Similarly, when 0.2% ribose was combined with 0.2% sorbitol, a steady increase in pARA activity as from the 7.5th hour was also observed but with much lower levels of CFP fluorescence as for arabinose + sorbitol.

Figure 26: RFP Fluorescence normalized with OD for E.coli K12 MG1655 transformed with pARA-CFP/pRIB-RFP double insert in absence of any sugar (■), in the presence of 0.2% arabinose + 0.2% ribose (●), 0.2% arabinose + 0.2% lactose (○), 0.2% arabinose + 0.2% sorbitol (△), 0.2% ribose + 0.2% lactose (+) and 0.2% ribose + 0.2% sorbitol lactose (◇). Experiments were conducted during 17 hours.

We also monitored the RFP fluorescence when the double insert construct was tested with combinations of sugars (Figure 26).
The results show that pRIB activity measured as RFP fluorescence was not significantly changed in all tested conditions. We can hypothesized that we did not measured any pRIB activity probably due to the fact that the sugars would either have preferentially promoted pARA activity (e.g ribose or lactose) or inhibited pRIB activity.

Summary


We can conclude that in the simple constructs, a sugar higher up in the hierarchy will inhibit promoters associated to sugars lower in the hierarchy. For example, when pARA is placed in the presence of 0,2% arabinose and 0.2% lactose, pARA is not activated. Lactose, being higher up in the hierarchy than arabinose, inhibits pARA.
A sugar which is lower in the hierarchy has no effect on the activity of a promoter associated with a sugar of higher hierarchy.
For the double construct, we notice that every combination of sugar in a lower hierarchy than the arabinose produced a CFP fluorescence which results of the induction a pARA activity. It does not modified the activation of pARA. The combination of 0,2% of lactose and 0,2% of Arabinose shows an activation of pARA. It means that the Lactose did not inhibit the pARA, as we expected. It can be a cross-activation because theses sugars are close in the hierarchy.When we study the RFP fluorescence (showing the activity of pRIB), we can’t observe any significant results because every sugar are higher than Ribose in the hierarchy. All sugars will induce an inhibition of the pRIB. This results were expected.



VI - Duration of the promoter expression


As our tool is mainly based on controlling gene expression in time, it is important for us to study the activity of these promoters in time. This will allow us to better understand the behavior of our promoters, enabling us to take into account any time differences we might discover in the design of our tool.


● pLAC-GFP E.coli :

Figure 27: Rate of GFP Fluorescence/hours for E.coli K12 MG1655 transformed with pLAC-GFP simple insert in the absence or in the presence of 0,2% of Lactose

The activity of pLAC is relative to the expression of GFP fluorescence.
We do not observe the production of GFP fluorescence in the condition without sugar. It confirms our control. With 0,2% of lactose, we can observe a linear increase of GFP production per hour that begins around hour 2,5. At the 5th hour We observe a peak of production, and then the production of GFP sharply decreases until the 7th hour.
We can conclude that the activity of the promoter pLAC, observed here as the rate of GFP production, starts 2,5 hours after the presence of sugar in the medium and it’s activity lasts around 4,5 hours.


● pARA-CFP E.coli :

Figure 28:Rate of CFP Fluorescence/hours for E.coli K12 MG1655 transformed with pARA-CFP simple insert in the absence or in the presence of 0,2% of Arabinose

The activity of pARA is relative to the expression of CFP fluorescence.
We do not observe an increase in the rate of production of CFP fluorescence in the condition without sugar. It confirms our control. With 0,2% of Arabinose, we observe a steady increase of the CFP production rate beginning at hour 1,5. At hour 4,5 the rate of CFP production reaches a plateau and remains in this state throughout the end of the experiment.
We can conclude that the activity of the promoter pARA starts around 4,5 hours after the presence of sugar, and the activation remains constant.

● pRIB-YFP E.coli :

Figure 29: Rate of YFP Fluorescence/hours for E.coli K12 MG1655 transformed with pRIB-YFP simple insert in the absence or in the presence of 0,2% of Ribose

The activity of pRIB is relative to the expression of YFP fluorescence.
We do not observe an increase in theYFP production rate in the condition without sugar. It confirms our control. With 0,2% of Ribose, we can observe a linear increase of YFP production rate starting at the third hour. We have a peak in the YFP production rate at the 6th hour, and then the speed of YFP production declines until the 10,5th hour.
We can conclude that the activity of the promoter pRIB starts 3 hours after the presence of sugar in the medium and it is activated during 7,5 hours.

Summary


We can conclude that the activation and the activity of promoters vary in time, from one to the other. In saturated sugar conditions, pLAC is activated around hour 2,5 after the addition of sugar and it’s activity lasts for around 4,5 hours, pARA is activated around hour 4,5 after the addition of sugar, and the activity withheld throughout the end of the experiment, and finally pRIB was activated around hour 3 and lasted around 7,5 hours.
This information would need to be taken into account to increase the precision of sequential gene expression in time of our tool.

(1)“Hierarchy of non-glucose sugars in Escherichia Coli.” Aidelberg Guy et al, BMC Systems Biology., 2014 , PMID: 25539838