Team:Ionis Paris/Results


The objective behind our project is to develop a bioelectronic device able to quantify blood vitamin B12 (or cobalamin), thanks to the genetic modification of Escherichia coli and the use of Shewanella oneidensis, a metal-reducing bacteria. To develop our system, the experimental approach was divided in two main phases:

  1. Phase 1 : Sensing of vitamin B12 and production of lactate

  2. Phase 2 : Development of a bioelectronic sensor of lactate

Phase 1 : Sensing of vitamin B12 and production of lactate

Biobrick design for phase 1

In order to obtain a production of lactate corresponding to the amount of vitamin B12 sensed, our final plasmid for phase 1 was design as followed :

Figure 1 : Final construct for phase 1

As lactate production was not a common way to report riboswitch activity, this plasmid was split in two different ones:

Figure 2 : Plasmid A for the reporting of the Cobalamin riboswitch activity (containing biobrick 1 & 2 (BBa_K3933005))

Figure 3 : Plasmid B for the arabinose-induced production of lactate (containing biobrick 3 (BBa_K3933006))

Golden Gate Assembly of biobrick 1 and biobrick 2 :

At first, we tried to assemble biobrick 1 from iGEM kit plates, using Golden Gate Assembly. Unfortunately, we did not manage to transform our bacteria with plasmids of interest. A linear fragment corresponding to biobrick 1 and biobrick 2 was then ordered from Twist Bioscience to be assembled using Golden Gate assembly with Esp3I. Workflow for Golden Gate Assembly is shown below :

Figure 4 : Golden Gate Assembly workflow (created with Biorender)

A) Primer design and PCR

The Golden Gate Assembly technique mainly relies on the design of specific DNA overhangs, containing type II-S restriction enzyme sites. A more general structure for these overhangs is shown below :

Figure 5 : Golden Gate Assembly overhang structure (created with biorender)

  1. NNNNN (or (N)5) : Complementary ligation overhang sequences

  2. Restriction site : Recognition site for the enzyme

  3. NNNN (or (N)4) : Extra space needed for the enzyme to cut DNA

Tet operator sequences are sequences that can be difficult to synthesize. The fragment for biobrick 2 was therefore ordered with these sequences, and a first step of addition of Tet operator sequences by PCR was done. After this, specific overhangs were added to both biobricks 1 and 2 by PCR.

Figure 6 : Agarose gel after PCR (1: Addition of Tet operator on Biobricks 2. 2: Addition of Golden Gate overhangs of Biobricks 1 and Biobricks 2. L : ladder; A: Biobrick 2; B: Biobrick 1)

B) Backbone preparation

The idea behind this step was to modify a pSB1A3 containing a RFP (BBa_J04450), our main backbone for restriction enzyme clonings used in our lab, in order to obtain a vector usable both for restriction enzyme cloning and for Golden Gate Assembly using BsmBI/Esp3I. To do so, we ordered a 121 synthesized fragment from IDT containing random nucleotides and 4 specific sites :

Figure 7 : Ordered fragment with the different wanted cut sites (SnapGene representation)

As shown on Figure 7, the fragment contains Prefix (XbaI) and Suffix (SpeI) restriction sites for insertion inside a Biobrick RFC[10] compatible backbone.

Figure 8 : Esp3I recognition and cut sites (New England Biolabs)

Figure 8 shows the recognition and cut sites of Esp3I. In order to perform Golden Gate Assembly, Esp3I sites have to be oriented in a specific way :

  1. Site on the left needs to have the following sequence for the top strand : 3’ (N)₅GAGACG 5’

  2. Site on the right needs to have the following sequence for the top strand : 3’ CGTCTC(N)₅ 5’

When obtained, DNA was resuspended and both the pSB1A3_BBa_J04450 and the DNA fragment were digested with XbaI and SpeI and ligated using a T4 DNA ligase. E. coli DH5-α were transformed using the obtained DNA and the construction was verified using PCR on colonies and digestion.

1) PCR on colonies

As SpeI and XbaI use leave cohesive sticky ends, primers were designed in order to check for the presence of the insert (backbone can ligate on itself), but also for the correct orientation of the insert (incorrect orientation is possible) :

Figure 9 : Primer locations on the ligated backbone (SnapGene screenshot)


Figure 10 : Agarose gel after PCR on colonies (L : ladder; A : PCR fragments; B : linearized pSB1A3_RFP)

Results show that the fragment was correctly inserted into the backbone for nine colonies. Colonies were picked and grown into LB-Ampicilline liquid media for miniprep of plasmids.

2) Gel migration

After liquid growth, miniprep was done and plasmids were digested using Esp3I. Unfortunately, the Esp3I sample used was no longer viable : digestion was therefore done a second time later with a new enzyme sample on one clone. Figure 10 shows band sizes of undigested plasmids :

Figure 11 : Agarose gel after miniprep of plasmids (L: Ladder; A: Miniprep products)

On the expected gel, the supercoiled form of the plasmid appears between 1.2 and 1.5kb. This band size is found for all clones, except for clone 19 who shows a band at higher sizes. This unexpected band can be explained by the ligation of the BBa_J04450 part between the insert and the backbone, as XbaI and SpeI digestions leave the same sticky ends.

C) Golden Gate Assembly

For this step, inserts and the backbone are digested with Esp3I and ligated in the same tube, following the Golden Gate Assembly protocol.


Unfortunately, we did not manage to transform any bacteria after assembly. This could be explained by difficulties encountered to quantify low concentrations of DNA before assembly. We therefore changed our cloning strategy for CPEC, see more about this choice in engineering success.

Circular Polymerase Extension Cloning of biobrick 1 and biobrick 2 :

Figure 12 : General CPEC workflow (created with Biorender)

Step 1

    (A) Amplification of inserts by PCR
    (B) Addition of 30 base-pair overhangs to the insert by PCR
    (C) Opening of backbone in two split linear fragments by PCR
    (D) Addition of a 30 base-pair overhang to the backbone (fragment A) by PCR

Step 2

    (A) PCR purification of PCR products
    (B) Ligation of insert and resistance using a DNA polymerase (overhangs are used as primers) as in a PCR cycle

Step 3

    Transformation of the obtained plasmid in E. coli DH5-α competent cells.

Step 1 : preparation of PCR

A) Amplification of inserts by PCR

Using 20 base-pair primers, inserts were amplified by PCR. Results are shown on figure 12.

Figure 13 : Agarose gel showing amplification of fragments of interest (L : 1kb plus NEB ladder; A : amplified insert A (1106 bp); B : amplified insert B (887 bp)) and expected gel)

Bands are seen at the correct sizes, addition of overhang can be performed.

B) Addition of 30 base-pair overhangs to the insert by PCR

This step is done by using 50 base-pair primers containing the same 20 base-pair hybridization sequences as in step A) and 30 base-pair overhangs complementary to the backbone/insert sequences.

Figure 14 : Agarose gel showing addition of overhangs on fragments of interest after PCR purification (L : 1kb plus DNA ladder; A : amplified insert A with overhangs (1166 bp); B : amplified insert B with overhangs (917 bp)) with expected gel

Bands are seen at the correct sizes, fragments were correctly amplified with overhang primers.
Results show the correct band sizes for each insert. Fragments are therefore ready for CPEC reaction.

C) Opening of backbone in two split linear fragments by PCR

To make sure the reaction worked, this step was done on several miniprep preparations of the psB1A3 plasmid. Primer design follows the same rules for this step as in step A) Agarose gels in Figure 14 shows results after PCR.

Figure 15 : Agarose gel showing linearization of backbone fragments (L : NEB 1kb+ Ladder; A : Fragment A (1367bp); B : Fragment B (796 bp))

Results show that linearized fragments were amplified for most of the plasmids tested. Addition of CPEC overhang can be performed on fragment A.

D) Addition of a 30 base-pair overhang to the backbone (fragment A) by PCR

Addition of an overhang sequence is only done on fragment A to ligate back the ampicillin resistance inside the backbone. To do so, primer design is the same as step B.

Figure 16 : Agarose gel showing addition of CPEC overhang on backbone fragment A (L : NEB 1kb+ Ladder; A : Fragment A with overhang (1397bp))

Results show that fragment B was correctly amplified with overhang primers. backbone fragments were ready for cloning.

Step 2 : Assembly by CPEC

Assembly was done by following the CPEC protocol and transformed into E. coliDH5-α competent cells. Figure 16 shows results after overnight culture on ampicillin agar plates.

Figure 15 : Ampicillin agar plates with transformants after incubation (C+ : positive control (pSB1A3-RFP transformants); A : Ratio A CPEC; 3 : Ratio B CPEC)

After 24 hours of incubation, colonies grow on agar plates. PCR on colony was performed on some clones (Dish 1 on figure 15) to check the correct assembly of our biobricks. We made the mistake of plating only a part of our transformation product. More transformation product was therefore plated on agar plates in order to check assemblies (Dish 2 on figure 15). The next day, colonies were picked and grown overnight in LB media with ampicillin to perform miniprep.

Figure 16 : Agarose gel showing miniprep digestion of transformants (L : Ladder; A : undigested white colonies minipreps; B : undigested red colonies minipreps; C : digested white colonies minipreps; D : digested red colonies minipreps) and expected results

On figure 16, we can see two different sizes for digested plasmids : for white colonies, a band is seen between 2.0 kb and 3.0 kb. For red colonies, a band is seen between 3.0 kb and 4.0 kb.

These results suggest that white colonies were backbones without insert (around 2200 bp) and red colonies pSB1A3-RFP (contamination with positive controls).


Inserts were prepared correctly. However, Our first CPEC reaction did not result in any assembly. Due to time limitations, the experiment was not reproduced. Ways of improving this step can be found in engineering success.

Functional assays for biobricks 1 and 2

If assembly of biobrick 1 and 2 would have worked, we would have followed the following workflow to check riboswitch activity :

Figure 17 : Workflow for the evaluation of the cobalamin riboswitch activity (created with Biorender)

By following this workflow, we would have been able to evaluate both the activity of our riboswitch and its kinetics in one assay.

Riboswitch activity : Expected results

If the cobalamin riboswitch is effective, we expect the increase of fluorescence with both cobalamin concentration and time. For improvement of this part, see engineering success.

Assembly of biobrick 3

Initially, biobrick 3 should have been assembled with Golden Gate Assembly using kit plates. As mentioned above, we were not able to transform any bacteria from kit plates. We therefore chose to order a linearized fragment from IDT. Unfortunately, we never managed to clone or amplify this fragment.

Functional assay for biobrick 3

Lactate production by the lactate dehydrogenase would have been measured under stimulation with arabinose, for different lactose concentrations.

Phase 2 : Development of a bioelectronic sensor for lactate

As explained before in Description and Contribution, we aim to use the bacteria Shewanella oneidensis (S. oneidensis) as a reporter system.

Experiments and results

To do so, a lab-scale prototype was setup (figure 18), following experimental setups described in the article from Zeng et al. (2019) [1]. 48 hours culture of Shewanella oneidensis were diluted to obtain a DO600nm = 0.600 and 5 mL were inoculated into 45 mL of LB media in the prototype. The experiment was performed with various amounts of lactate in the media (40 mM, 20 mM, and 0 mM). Measures of voltage generated by Shewanella oneidensis were done on 500 minute periods.

Figure 18: Lab-scale prototype for bacterial electrical measures

Figure 19: Electrical activity of S. oneidensis over time (Blue : 40mM L-lactate; Grey : 20mM L-lactate; Red : No lactate)

Overall, the three curves show an increase in voltage over time. The rise in electricity production by our culture is higher and faster as concentration increases, until reaching a plateau around 300 mV. As rising in potential seems to be linked to lactate concentration, a curve was plotted using the No lactate condition as a control (figure 20).

Figure 20: Lactate specific electrical activity by Shewanella oneidensis over time

By looking at this curve, we can more easily see the specific impact of lactate on electrical activity without the impact of the media. This specific impact seems to be detectable in a few hours only, with a peak of specific activity at around 3 hours. After 500 minutes, the specific electrical activity from lactate seems to decrease.

Unfortunately, issues with S. oneidensis stocks blocked our experiments for one month. We were therefore not able to produce more results with these experiments.

Perspectives and expected results

In order to confirm the link between lactate concentration and electricity production by S. oneidensis, more experiments should be made. Firstly, more experiments should be done by testing a great variety of lactate concentrations, as we aim to precisely evaluate lactate production by phase I. Moreover, other parameters should be evaluated : varying S. oneidensis DO, testing different co-culture ratios with E. coli, or changing LB media to minimal media without a carbon source to avoid background signal due to LB media. By testing all this parameters, we aim to fine tune the accuracy of our system.


Overall, results suggest that our system could be used to report lactate concentrations thanks to the production of electricity by Shewanella oneidensis in a few hours. Even though this reporting seems to reach a plateau after several hours, the fact that a signal difference can be detected early is promessing, suggesting that our system could be used as a fast and efficient way of reporting genetic reporter systems. By combining our system with Phase I designs, we expect that vitamin B12 could be quantified precisely and in a few minutes to hours in the context of a personal quantifying system.

Project achievements

    + A first lab-scale prototype for the measurement of Shewanella oneidensis electrical activity was successfully built

    + First measures suggest that different amounts of lactate could be measured using our lab-scale prototype

    + Attempts of automatisation using different hardware/software solutions were made

    + A final prototype was modelled and 3D printed

    - We spent a lot of time on different cloning strategies that did not work due to equipment and time limitations

    - Due to difficulties encountered in Shewanella oneidensis growth, only a few experiments were done with the prototype


    [1] Zeng, J., Banerjee, A., Kim, J. et al. A Novel Bioelectronic Reporter System in Living Cells Tested with a Synthetic Biological Comparator. Sci Rep 9, 7275 (2019).

Big thanks to our sponsors :