Team:Alma/Engineering

Engineering
Engineering
In order to begin our engineering process, the theoretical construction of a genetic circuit had to be undertaken. After consulting the literature, it became apparent that DDT is a known endocrine disrupting agent as well as a suspected carcinogen for humans. Additionally, it is described as a xenoestrogen. We took advantage of the fact that organochlorines such as DDT act as estrogen analogs, and decided to use an estrogen receptor in our circuit. This receptor can bind DNA in the presence of DDT, and can be used in a circuit to drive the expression of a reporter gene. Thus, the reporter gene expression will create a visual change which can be observed and indicate presence of DDT in a sample. This proposed circuit further piqued the interest of local field experts as our proposed implementation did not call for the introduction of any genetically modified organisms into the environment directly. A mobile lab set up at the superfund site could be used to avoid introducing new microbes into the environment. Alternatively, samples could even be brought back to the lab and studied in the safety of the lab, greatly decreasing the risk of escape of our modified organism.
Initial Design
Upon agreeing to the use of an estrogen receptor for binding of the DDT, our next question became which species of estrogen receptor to use. We elected to pursue the human estrogen receptor alpha for our genetic circuit, as this is a relatively well-characterized estrogen receptor and a good place to begin preliminary work with our biosensor. Indeed, this part already exists in the iGEM registry. Future plans may involve the use of estrogen receptors from a variety of other species (see Expanded Designs below).

The next step in construction of our genetic circuit was connecting the estrogen receptor to a reporting gene. For this, we elected to use an RFP read out upon the binding of the estrogen receptor. An additional twist we added to our circuit was actually the use of an inverter, which contained a segment of DNA known as the estrogen response elements, or ERE. As the estrogen receptor is an intracellular receptor, upon binding of the receptor by estrogen -- or an estrogen analog in our case -- the receptor undergoes a conformational change which allows it to bind the ERE. In order for the proper functioning of our circuit, our RFP readout is controlled through the use of a TeT repressible promoter that is turned off by the constitutive production of TetR. However, upon binding of the bound estrogen receptor complex to the ERE the transcription of TetR is inhibited. Thus, the production of RFP occurs and the fluorescent change can be observed and measured.
Assembly: First Steps
For the creation of our circuit we utilized 3 biobricks currently listed in the iGEM registry: I13521, K123002, and K123003, which acted as our RFP output, inverter, and estrogen receptor respectively, as can be seen in Fig. 1. Luckily for us, we were able to utilize our on campus lab to conduct scientific experiments in the last few weeks of this season’s competition. Therefore, to begin our work we looked at the 3 separate biobricks individually.
Figure 1. Our constructed genetic circuit for our DDT sensitive biosensor, which was created through the use of the software Pigeon.
The I13521 part consists of the TeT repressible promoter linked to the coding region for RFP, and functions as our RFP reporting gene, as previously mentioned. This piece was available to us in our leftover 2019 iGEM distribution kit, therefore it was transformed into NEB stable strain E. coli and successful transformation was confirmed through sanger sequencing. We proceeded to test this strain (see Preliminary Measurements below) and created our other part in parallel.

The next biobrick used was the inverter, K123002. This part is regulated by the constitutive LacI promoter, followed by the ERE site previously mentioned, and finally the coding region for TetR. Therefore, this piece acts to constitutively inhibit production of RFP; however, binding of the ERE by a bound estrogen receptor complex will block the transcription of TetR, which allows the expression of RFP, giving this piece its inverting ability. This BioBrick was synthesized by IDT with regions on the end of the DNA that are homologous to the pSB1C3 backbone, facilitating Gibson assembly. This assembly reaction was then successfully transformed into NEB stable strain E. coli. Confirmation of successful transformation was again obtained through sanger sequencing and colony PCR gel electrophoresis, which is seen in Fig. 2.
Figure 2. Colony PCR gel electrophoresis of K123002, which is 742 base pairs in length.
Finally, we began work on the coding sequence for the human estrogen receptor alpha, K123003. Like the previous part, this DNA was synthesized by IDT. However, due to the nature of this sequence, the resulting DNA was not a pure mixture. Nonetheless, we proceeded with Gibson assembly and transformation into NEB stable strain of E. coli. While our positive controls produced colonies, we obtained none for the K123003 assembly reaction as is seen in Fig. 3. Therefore, successful transformation of this biobrick has not yet been completed. One possibility is that the problems with the DNA synthesis interfered with the assembly and transformation. It is also possible that this part is cytotoxic - evidence for this was provided by other teams on the registry. Work to codon harmonize this specific DNA sequence to resolve its potential cytotoxicity has been undertaken and our preliminary results can be seen here.
Figure 3.
Left: Plate showing no colony growth following an attempted Gibson assembly and transformation of K123003 into NEB stable strain E. coli.
Right: Positive control of a Gibson assembly followed by transformation into NEB stable strain E. coli.


Preliminary Measurements
Upon determining successful transformation, this part was tested through use of the HL2483 strain of E. coli, which constitutively expresses TetR, in conjunction with the DH5ɑ E. coli strain. The two strains were both separately transformed with I13521 and J04450, and were then streaked onto a plate. J04450 served as our positive control and showed expression of RFP in both strains, while I13521 only showed RFP expression in DH5ɑ, indicating repression of I13521 in the presence of TetR, as seen in Fig. 4. Additionally, expression was tested in the same format previously mentioned through the use of liquid cultures and a plate reader, providing the same result as the streaked plates.
Figure 4. Streaked plate showing I13521 and J04450 transformed into both DH5ɑ and HL2483 E. coli.
We also carried out preliminary measurements of expression of K123002, measuring the production of TetR by harvesting cells and loading protein onto an SDS PAGE gel. During this measurement of expression, no appreciable amounts of TetR were seen on the gel. TetR is approximately 25 kDa in size, which is pointed out on the protein ladder by the arrow on the gel below in Fig. 5. Therefore, further characterization of the expression of TetR by this part will need to be performed in the future.
Figure 5. SDS PAGE gel of K123002 in comparison to the Sigma Marker. The arrow indicates 25 kDa, the approximate size of translated TetR.


Assembly: A New Part is Born
While further work is being conducted on the assembly of the estrogen receptor coding region, we proceeded with assembly to combine the two BioBricks we had. Our first attempt to quickly take advantage of what was present in our lab was cloning via a Universal Golden Gate approach. While the PCR using Universal Golden Gate primers appeared successful, we obtained a high number of background colonies. Even after screening through several clones, we did not find our desired part. We also attempted to create this part via a modified form of 3A assembly, but also failed to achieve the desired result.

Finally, we ordered additional primers and were able to successfully assemble the I13521 and K123002 BioBricks together by Gibson assembly to make a new BioBrick: K3445001. Confirmation of successful assembly of these two pieces was done through sanger sequencing and colony PCR gel electrophoresis, as can be seen below in Fig. 6. Background bands are seen in each lane on the gel below and were used as a standard for the determination of the DNA size within each of our analyzed bacterial clones.
Figure 6. Colony PCR gel electrophoresis of multiple clones containing the K3445001 part, which is approximately 1700 base pairs in length. Various primers we used in the different reactions to amplify different regions of the newly assembled BioBrick.
However, colonies containing this assembled piece had an interesting visual appearance. We would expect these colonies would not be expressing any RFP as it should be repressed by TetR; however, some of our transformed colonies were all red, as seen below in Fig. 7.
Figure 7. Streaked plate containing both HL2483 and DH5ɑ E. coli transformed with K3445001, I13521, and J04450.
It is seen that the expression of TetR by the HL2483 E. coli was able to successfully inhibit our K3445001 RFP production in Fig. 7. However, Fig. 8 below shows that RFP was still produced by K3445001 when not transformed in HL2483 E. coli, even though it should have been repressed by the production of TetR within the circuit. RFP corresponds to a band of approximately 25 kDa in size, which is alluded to by the arrow on figure 8. Therefore, in our constructed genetic circuit, our inverter was not properly functioning to inhibit the production of RFP. This SDS PAGE also shows that expression of TetR (which is also approximately 25 kDa in size) by our K123002 part occurs at such a low level that it cannot be appreciated on the gel, as no band is seen. This was additionally seen in Fig. 5 above. In fact, I13521 in HL2483 (in lane 9) shows no band corresponding to RFP or TetR. Therefore, this result likely indicates that the production of TetR needed to repress the TeT repressible promoter occurs at very low levels that cannot be appreciated at this level of separation. Additionally, this problem is exacerbated by the fact that the TetR transcribed by K123002 has a strong degradation tag on it and is likely being degraded before it can create an effect on the TeT promoter, causing the K3445001 part to be uninhibited and produce RFP; causing its red appearance. Two separate theories have been constructed in order to correct this complication in the future. Firstly, we can use a strong promoter (such as J23100) in K123002, rather than the currently used LacI promoter, in order to produce TetR at much higher levels in the cell. Secondly, we could remove, or use a less powerful, degradation tag on the produced TetR in order to allow minor accumulation of the TetR in the cell.
Figure 8. SDS PAGE gel analyzing K3445001 in NEB stable strain E. coli, I13521 in HL2483 and DH5ɑ, and J04450 in Hl2483 and DH5ɑ.
The next step, upon showing expression of a functional estrogen receptor in E. coli, will be assembling all three biobricks into one circuit. Ideally, this step will be performed through use of Gibson assembly between our new K3445001 piece and the estrogen receptor coding region. The circuit is planned to be assembled in order of RFP output, followed by the inverter, and finally the estrogen receptor coding sequence. This order was selected as a way to minimize any RFP bleed through we may experience from the LacI promoter if we had shifted the RFP output to the end of the circuit. The way we plan to construct the circuit will allow for RFP to only be controlled by the TeT promoter, and therefore will only be present when the estrogen receptor is bound.
Planned Experiments
Upon successful assembly of our circuit, we can begin the testing phase of the circuit to demonstrate that it is properly working. This can be undertaken by testing the circuit in several different ways. Firstly, we would test the circuit when exposed to normal media lacking xenoestrogens. We would expect these cells to display no RFP output. We could additionally test the circuit upon exposure to DDT, where we would expect an RFP output. Finally, to confirm control of our RFP output through the TeT promoter, we could induce the produced TetR with anhydrotetracycline (aTc); therefore, cells containing our circuit that are exposed to aTc should produce the maximum output of RFP. If we see no RFP output in the presence of DDT, we could also test our circuit within cells exposed to estrogen to determine if we are experiencing issues with the production of our estrogen receptor, or if we are experiencing issues specifically with the human estrogen receptor binding DDT, which will guide future changes made to our circuit.

We plan to integrate the results we obtain with the math model that we are currently creating. Our initial foray into this approach is available as a PDF.
Expanded Designs
Ensuring there is not an overproduction of TetR (thus inhibiting our RFP output), but also not an underproduction (which would allow RFP production to be uninhibited) as control of our reporting gene by our inverter is crucial to the proper functioning of our biosensor. Proper functioning of the biosensor, as listed above, would allow for qualitative analysis of superfund site samples, where the presence of red indicates the presence of DDT in a sample. However, upon speaking with field experts such as the EPA and local Alma College researching professors, it became apparent that a semi-quantitative approach may be of even greater benefit. This approach would offer a screening method to indicate if certain superfund site samples exceed threshold limits for DDT, as determined by the EPA. We were informed that levels of DDT from 1-5 ppm indicate the necessity for cleanup and levels greater than 40 ppm indicate the necessity for an emergency response by the EPA. This semi-conservative methodology would allow for a rapid and cost-efficient way to influence the allocation of remediation efforts and resources to the areas of the superfund site most in need of them.

To achieve this type of semi-quantitative response, we expanded our initial design to include another fluorescent output, GFP. This output is controlled through the use of an additional inverter. This inverter is also repressed by the estrogen receptor, but at a different concentration than the original RFP circuit. One theorized way to achieve this is by applying several EREs in tandem to affect the level of repression in the presence of estrogen receptors bound by DDT.

In the example design illustrated in Fig. 9, the TetR inverter system is subjected to more sensitive repression compared to the cI repressor. When no DDT is present, the Estrogen receptor will be unable to prevent expression of either repressor, and thus no color will be produced. When a low level of DDT is present (such as 1-5ppm), some estrogen receptor will be activated. This activated receptor will have a greater chance of repressing the cI inverter versus the TetR inverter because of the greater number of ERE sequences controlling the former. Thus, at low levels, GFP will be produced. This indicates to the user that the sensor is working, that DDT is present, but that the site might be a low priority target for cleanup.
Figure 9. Modified genetic circuit, generated by Pigeon, that theorizes a way in which we can perform semi-quantitative analysis in the field.
By contrast, at high concentrations of DDT (i.e. 40ppm), both cI and TetR inverters will be repressed, and both GFP and RFP will be produced. A yellow or red color will result (depending on the relative efficiency of GFP versus RFP expression and stability). This will indicate that the area sampled requires immediate cleanup, and can allow the EPA to access additional emergency funding for cleanup.

Another variant on our biosensor design may help discriminate DDT and its derivatives from other xenoestrogens. To create this biosensor, we would take advantage of the fact that Estrogen receptors from different species display different binding affinities for DDT, estradiol, and other xenoestrogens (Matthews 2000). For example, the rainbow trout and human estrogen receptor both bind estradiol tightly, but rainbow trout can also bind DDT at much higher affinities than the human receptor. Thus, a circuit was designed in which the human receptor acts to filter out background signals.

In this circuit, shown below as Fig. 10, the rainbow trout estrogen receptor is capable of repressing the inverter, while the human estrogen receptor represses the output. If DDT is present, the inverter will be repressed to a much greater degree than the output – with the inverter repressed, the output will be activated. Thus, the bacteria will appear red. If estradiol or a xenoestrogen other than DDT is present, then both the inverter and the output will be repressed – the repression of the output by the human estrogen receptor will negate any activation afforded by silencing the inverter. Thus, no red color will be produced.
Figure 10. Modified genetic circuit, generated by pigeon, that theorizes a way in which we can differentiate between various xenoestrogens using different species-specific estrogen receptors and different EREs.
This general approach to generating a discriminating biosensor can be adapted to detecting a particular type of xenoestrogen depending on the species of receptors used. It additionally offers a gateway to the detection of multiple different endocrine-disrupting pollutants, where varying types of species specific endocrine receptors can be used to bind environmental EDCs. This work forms the basis of a synthetic biology tool that can have major implications in remediation efforts and hopefully can serve as a way to enhance environmental remediation efforts not only in our local community, but across the globe.