Team:UPF Barcelona/Proof Of Concept


As a proof of concept for our device, which regulates a correct dosing depending on a measured hormonal level, we designed a biological circuit capable of producing and sensing lactone in a controlled way. This circuit is formed by two kinds of cells, ones responsible for producing lactone, and others responsible for sensing the presence of lactone and producing super-folder GFP (sfGFP) in response. With this circuit, we can simulate and understand how the thyroid axis works and how our device would work.

Why develop a bacterial lactone proof of concept?

Figure 1. Diagram for a general closed circuit.

To prove that our PID controller can correct an abnormal hormonal level, a closed biological circuit is needed. With our bacterial lactone proof of concept we are able to close the circuit and assess the proper functioning of our PID controller using a system that has been used extensively in sythetic biology [1][2]. This closed circuit (Fig.1) is composed of three parts.

On the one hand, we have the PID controller, which reads the fluorescence value and converts it into the hormone concentration. This concentration is then compared with our target concentration, which is the concentration we would ideally want to preserve. Finally, it responds by releasing effector molecules or not.

On the other hand, we have the biological part. This part is composed of two different cells: the producer cells and the sensor cells. The producer cells produce hormones or not depending on the concentration of the effector molecules. The sensor cells produce sfGFP when the concentration of the hormone is large enough for it to detect it.

In the T3 hormone context, this closed circuit cannot be implemented in vitro, as T3 is a complex protein that requires multiple enzymes that are not present in bacteria. This means that the production of T3 in bacteria is not a feasible option. This lack of a producer cell leaves the circuit open and unable to control the synthesis of T3 (Fig.2).

Although a closed circuit cannot be implemented in bacteria using T3, the PID controller can be tested using a molecule which can be easily produced and sensed by bacteria, the bacterial lactone. Using this molecule, we can test and prove the ability of the PID controller to regulate the hormonal levels without the risks associated with a non-rigorous control of hormones in the body (Fig.2).

Figure 2. Comparison between an open T3 bacterial circuit and a closed lactone bacterial circuit.

Minimal hormonal regulation model

The circuits pictured in the section above show a minimal hormonal regulation model, composed of two different cells: the producer cells (BBa_K3484005) and the sensor cells (BBa_K3484004).

The producer cells (Fig.3) synthesize lactone via the LuxI enzyme. This enzyme is regulated by the pBad promoter, which is induced by arabinose but repressed by glucose. This means that the production of lactone can be controlled in two ways, increasing the production with arabinose and decreasing it with glucose. These cells were provided by the Synthetic Biology for Biomedical Applications laboratory in the Barcelona Biomedical Research Park (PRBB).

Figure 3. A. Scheme of the interactions on a producer cell. B. Truth table that explains how the producer cell reacts to high and low concentration of arabinose and glucose.

The sensor cells (Fig.4) produce sfGFP when a large enough concentration of lactone is present in the medium. In this case, the cell constitutively produces LuxR, which binds to the lactone present in the medium and together, they form a complex that activates the pLux promoter, generating the sfGFP whose fluorescence can be detected.

Figure 4. A. Scheme of the interactions on a sensor cell. B. Truth table that explains how the sensor cell reacts to high and low concentrations of lactone (AHL).

Why using super-folder GFP?

The sfGFP is a variant of the Green Fluorescent Protein which folds faster, thus giving faster responses [3]. This fact, in conjunction with the insertion of an ASV degradation tag at the end of its sequence, allows our sensor to react faster to variations in lactone concentration [4].

Analogy with the thyroid axis

The analogy between a hormone regulation model and our bacterial lactone proof of concept has been stated above, but the analogy goes beyond that. If we put the lactone circuit in context with the thyroid axis as template, the analogy between both systems arises.

For instance, the lactone-producer cells can be compared to the peripheral tissue cells, which in hypothyroidism cases are responsible for metabolizing levothyroxine into T3 hormones [5]. This process can have similar dynamics to the lactone-producer cells, where arabinose leads to the metabolization of lactone.

Moreover, this system can also be used to simulate other endocrine disorders such as hyperthyroidism. In this case, the injected antithyroid drug is mimicked by the glucose, inhibiting the activity of the producer cell, either the thyroid gland or the lactone-producer cell. This leads to a decrease in the levels of T4-T3 hormones, which are modelled by the lactone. At the same time, TSH, which simulates the activity of the thyroid gland, would be modelled by the arabinose [5].

Figure 5. Comparison between the minimal hormonal regulation model and different thyroid disorders.

Cloning strategy

The lactone-producer cell was provided by the Synthetic Biology for Biomedical Applications laboratory, but in order to close the circuit we needed to build the lactone biosensor. For this aim, the same laboratory provided us with two different parts: the sfGFP and the LuxR-pLux constructs.

The first step towards our lactone biosensor was to include the ASV degradation tag into the sfGFP sequence. This was done via a PCR, where one of the primers included an overhang which introduced this tag’s sequence at the end of the sfGFP coding region.

Once the tag had been added, the next step was the cloning of the lactone biosensor itself. The cloning was done using the Biobricks assembly. The backbone was digested using the EcoRI and PstI cutting sites, the LuxR-pLux construct in EcoRI and SpeI and the sfGFP+TAG construct in XbaI and PstI. The products of these digestions were then ligated and transformed into E. coli cells. These lactone biosensor cells were then tested to check their viability.

Figure 6. Scheme of the cloning approach used to produce the sensor cells.

*All the information regarding the different attempts at the cloning and its details can be found on the proof of concept notebook.

Further work

To enquire into the analogy between the lactone circuit and the thyroid axis, a turbidostat was developed. Although we didn’t manage to join both the lactone circuit and the turbidostat, this can be seen as the next step in the assessment of our proof of concept.

The idea is to use the turbidostat to keep a constant OD in our lactone cell system, so we can actively regulate the presence of lactone in a long period of time, similar to what Hormonic is bound to do. To do that, we would need a system composed of three different parts connected by four pumps (Fig.7).

Figure 7. Scheme of the final proof of concept design for Hormonic.

First, we need a turbidostat with the lactone-producer cells. This turbidostat should have two different incoming pumps, one for glucose and another for arabinose. This way, we can control the lactone production. The produced lactone should then be sent through an outcoming pump to a second turbidostat with the sensor cells. Finally, this second turbidostat would get rid of the sensed media and the used cells by an external pump that discharges waste to a biological container.


[1] Ballestero, M. C., Duran-Nebreda, S., Monta, R., Solé, R., Macía, J., Rodríguez-Caso, C. A bottom-up characterization of transfer functions for synthetic biology designs: lessons from enzymology. Nucleic Acids Research, 2014, Vol. 42, No. 22

[2] Garcia-Ojalvo, J., Elowitz, M.B., Strogatz, S.H. Modeling a synthetic multicellular clock: repressilators coupled by quorum sensing. Proc. Natl Acad. Sci. U.S.A., 101, 10955–10960

[3] Pédelacq, J., Cabantous, S., Tran, T. (2006). Engineering and characterization of a superfolder green fluorescent protein. Nat Biotechnol, 24, 79–88, doi: 10.1038/nbt1172.

[4] Andersen, J.,Sternberg, C., Poulsen, L. K., Bjørn, S. P., Givskov, M. New Unstable Variants of Green Fluorescent Protein for Studies of Transient Gene Expression in Bacteria. Applied and environmental microbiology, June 1998, p. 2240–2246

[5] Brent G. A. (2012). Mechanisms of thyroid hormone action. The Journal of clinical investigation, 122(9), 3035–3043.