Summary

We successfully assembled 5 of 6 of our planned transcriptional units (Level 1 parts). Due to time constraints, we weren’t able to build the final transcriptional unit (construct D in Figure 1) or the TPA-detection devices.

L1 Assemblies — Figure 1. All the level I parts we were working towards on assembly and which parts, whether from the distribution kit or new, we used. MoClo standards apply, but seams aren’t shown for simplicity. The letter of L1 transcriptional unit represents which JUMP plasmid it’s cloned into (e.g A into pJUMP29-1A, C1 into pJUMP29-1C, etc.)

We were able design, build, and test in silico our mutant phasin and show that it has increased binding affinity to PHB. A future direction would be to express the phasin and test it in vivo.

Repurposing System

Design

The CDS sequences for the TPH operon, referred to as TU A as it’s cloned into pJUMP29-1A, were assembled in order of tphC, tphA2, tphA3, tphB then tphA1. This order was selected because it’s the order in the native organism’s (Comamonas testosteroni) genome.

The coding sequences from the TPH degrading enzymes were isolated from Comamonas testosteroni, then optimized for expression with E.coli. Each enzyme within the TPH degrading transcriptional unit was interspersed with the BBa_B0030 RBS sequences, which allows for the expression of each enzyme in the pathway under a single promoter in the final construct (Figure 2).

Figure 2. Plasmid map of TU A, for expressing enzymes required for TPA degradation. The map is designed on benching.

Each enzyme along with the RBS sequence was synthesized into a level 0 ampicillin resistant GoldenGate compatible plasmid containing BSAI restriction sites. Each level 0 part was designed for assembly using the golden gate method and attempted to follow the MoClo standard overhang conventions. The additional overhangs that were needed between the enzymes were selected to maximize the efficiency of the ligation reaction that needed to occur between each DNA segment. The overhangs were selected using the “NEBridge Ligase Fidelity” by New England Biolabs.

The repurposing system intends on including the PETase and METase enzymes which are able to degrade PET plastic into TPH and ethylene glycol. PETase enzymes have been worked on in the past by other IGEM teams, specifically the University of Toronto 2019 IGEM team was able to determine specific residue mutations that increased the activity of the enzyme. Since this has already been determined by a previous team, we will not be improving upon this existing part.

Once TU A is assembled, we intend on combining it along with the PETase enzymes and the PHB synthesizing transcriptional unit into a single level 2 transcriptional unit that would include all the necessary enzymes for the repurposing of PETase into a biodegradable alternative, PHB. All of the necessary enzymes would be contained within one bacterium.

Build

TU A was assembled using GoldenGate assembly. The RBS+CDS parts were ordered from IDT by synthesis, and the vector backbone (pJUMP29-1A), promoter (J23100), eiraCFP, and terminator (B0015) parts were taken from the distribution kit.

Test

TU A colonies — Figure 3. Selection for positive colonies that were transformed with TU A. There were non-fluorescent colonies and fluorescent colonies that were difficult to distinguish between cyan or green.

We were expecting positive colonies to be cyan-fluorescent, as they would express the eiraCFP reporter. However, the fluorescent colonies appeared to be green and there were non-fluorescent colonies, as seen in the above figure. We cultured both cultured and plasmid prepped both colonies, then digested the plasmids with Esp3I.

TU A gel — Figure 4. Plasmids digested by Esp3I from a non-fluorescent colony (2), fluorescent colony (3), and pJUMP29-1C as a control (3).

Learn

The fluorescent colony’s digested plasmid matches the expected band lengths of TU A (Figure 4), and the fluorescent matches that of the JUMP plasmid control. This means eiraCFP is not being expressed in colonies with the TU A part, and thus other CDS inserts are not expressed either.

We reasoned that there was no expression of eiraCFP because of the spacing between the RBS and CDS start codon. In the original design, there was a 5 base pair spacing between the RBS and CDS, based off an iGEM page describing RBS design. A better design would be using the RBS with spacing from well characterized and well-used iGEM parts, such as the variation of B0030 found in the iGEM 2023 distribution kit, which has a “TACTA” spacer on the 3’ end of the original B0030 part.

Design

We redesigned our new parts to use the B0030 variation (BBa_J428032). Unfortunately, we didn’t have enough base pairs to order the full gene. We had enough base pairs to order a new tphC CDS that’s flanked by BBa_J428032.

tphCfragmentv2 — Figure 5. Updated MoClo compatible tphC fragment, where tphC is flanked by the RBS part BBa_J428032. BsaI recogntion site is not shown.

Build+Test

As seen in the below figure, there were clearly both green and cyan fluorescent colonies, indicating that positive colonies are expressing eiraCFP. This confirms the RBS was the issue in our previous design, and a polycistronic construct works. This same stage is part of the second Build stage of TPA Biosensor.

TU C1 digest — Figure 6. pJUMP29-1 (2) and TU C1 (3) digested by Esp3I.

Learn

Future directions to further confirm that TphC, and other proteins involved in TPA-degradation, are being expressed would be an SDS-PAGE. After confirming protein expression, we would then test the engineered bacterium for TPA-degradation activity, ideally with the biosensor we developed.

TPA Biosensor

Design

We wanted to build a simple construct that can characterize the TPA-inducible promoter (P_tph) in E. coli DH5α, while using modular parts that are compatible with iGEM distribution kit parts.

First, we sketched out an initial design of our construct, as seen in Figure 7. tphC and tphR, which code for the activator protein of P_tph and TPA transporter, respectively, are expressed constitutively (we’ll call this operon tphCtphR). The fluorescent reporter gene, which is regulated by P_tph, is transcribed in the opposite direction. This is to mimic the gene cassette of Comamonas testosteroni, from which we took the sequences for P_tph, tphR, and tphC. C. testosteroni likely expresses tphR and the TPA-degrading enzymes in opposite directions to prevent leaky transcription. In other words, with regards to our construct, if tphCtphR were transcribed in the same direction and upstream of the reporter gene, the DNA polymerase may continue transcribing even after the terminator sequence. However, the JUMP plasmids’ overhangs are not compatible with bidirectional transcriptional units (TU). We reasoned that even if tphCtphR are expressed in the same direction, if it is placed upstream of the reporter gene, leaky expression can be avoided (Figure 8).

sketch of biosensor — Figure 7. Initial sketch of a TPA-detecting composite part. Terminator sequences are not drawn in but would be after tphR and the GFP CDS. Also note that later designs use CFP instead of GFP, as GFP is used as a selection indicator for cloning with JUMP plasmids found in the distribution kit

device design — Figure 8. Initial design for TPA-biosensor device.

Because transcriptional units built using traditional MoClo GGC only has one RBS and CDS, whereas our tphCtphR operon will have two of each, we designed tphR to have MoClo’s CDS overhangs (5’ AATG, 3’ GCTT) and tphC to be flanked by an RBS on each side, with a 5’ TACT and 3’ AATG overhang to make it a MoClo “RBS” (Figure 9).

ordered L0 fragments (biosensor) — Figure 9. Ordered fragments for P_tph (A), tphC (B), and tphR (C).

Build

P_tph TU (we’ll also refer as TU B) was assembled using GGC, into the pJUMP29-1B vector. All basic parts for TU B, except for the tph promoter (submitted as BBa_K4728000), come from the distribution kit.

Figure 11. Positive (non-fluorescent) and negative (green fluorescent) colonies for TU B.

TU B gel — Figure 12. Restriction digest of pJUMP29-1B (lane 2) and successfully assembled TU B (lane 3). There is no band at 3815bp in lane 2 as the JUMP plasmid sample was fully digested. Samples were digested by Esp3I

Learn

Now that TU B is assembled, the TPA-inducible promoter needs to be characterized with its responsiveness to TPA presence.

Upon reviewing the constructs and supporting papers, we realized that two other proteins, TpiB and TpiA, are required for TPA transport and were missing in our design. TphC is a periplasmic protein that binds to TPA, but tpiB and tpiA are the transporters that move the tphC-TPA across the membrane.

Additionally, we realized that tpaK, originating from the species Rhodococcus jostii, is responsible for encoding a transporter for the uptake of TPA. This protein serves to translocate TPA into our detection system and is a crucial element in our system that was missing.

Design

With the knowledge from the previous Learn stage, we designed two new TPA detection systems:

Figure 13.1 — Figure 13. To build those detection systems, we first need to assemble the L1 parts: C1, C2, C3, and D.

Figure 13.2 — Figure 13. To build those detection systems, we first need to assemble the L1 parts: C1, C2, C3, and D.

We used the same fragment that’s composed of tphC and the RBS BBa_J428032 as in the “Repurposing System” cycle. See Figure 1 for visualization of the ordered parts/fragments that are required to assemble TU C1, C2, C3, and D.

Test

TU C1 colonies — Figure 14. TU C1 positive (cyan) and negative (green) colonies.

gel of TUC1-3,D — Figure 15. Agarose gel of plasmids digested by Esp3I. Plasmids are from positive colonies transformed with pJUMP29-1C (1), TU C1 (2), TU C2 (3), TU C3 (4), TU D (5+6).

The cyan fluorescent colonies as shown in Figure 14 were a positive indication to us that expression was not impacted by the addition of coding sequences before the eiraCFP reporter. We also performed a restriction digest on the plasmids prepped from positive colonies, and ran the samples on an agarose gel. Unfortunately, we were unsuccessful with assembling TU D. The yellow lines indicate what the successful assembly of TU D would have resembled on the gel.

Learn

We did not have enough time to attempt another assembly of TU D. Similarly, we were limited in time to assemble any level 2 constructs which would house the TPA-detection system. Another additional step would be to perform a protein extraction and confirm there is protein expression of our new parts, as we are currently under the assumption that non-distribution kit CDSs (e.g. tphC) are being expressed because the eiraCFP reporter is.

Phasin Mutant

Research

Our initial step in the engineering cycle involved conducting thorough research on the Phasin F (PhaF) protein structure and its established interaction with PHB. Our team's approach was built upon an extensive review of theoretical models proposed in previous studies regarding their interaction.

What distinguished our project was the treating PhaF as a biosurfactant to PHB, which positively impacted our engineering process. This perspective afforded us a degree of flexibility, such that enhancing its biosurfactant role would also improve its chaperone activity (2018, et al. Mato, A).

Design

After formulating our project's concept, we aimed to create a design that was not only scalable but also amenable to future enhancements. Recognizing the potential for subsequent teams to build upon our work, we realized an in-silico test was invaluable.

To gain a better understanding of PhaF’s structure, we conducted a BLAST alignment to identify highly conserved regions (Maestro et al., 2013). We then found the highest alignment score for a segment in the N-terminal domain. In literature, this region is commonly referred to as BioF, which consists of 4 short segments: Bi1, Bi2, Bi3, and Bi4. Furthermore, the literature highlights residues 26-32 (WLAGLGI) located in the Bi1 segment to be responsible for interacting with PHB (Jendrossek & Pfeiffer, 2014). This region also demonstrated the lowest disordered protein score, depicted by our Dispredict_v1 analysis (Figure 16).

Fig.16 — Figure 16. Disordered protein analysis conducted on Dispredict_v1 (Iqbal & Hoque, 2015).

The Bi1 segment interacts with PHB in non-specific hydrophobic interactions. In light of this, our team decided to clone residues 26-32 to nearby residues 33-39. We then ran a heliquest analysis which indicated the amphiphilicity (moment) and hydrophobicity at each region in the protein. We found an increase in hydrophobicity which was expected, while amphiphilicity was maintained. Additionally, we carried out Dispredict_v1 to ensure the low protein disordered region spanned for longer in our mutant.

Below you can find a depiction of our engineering process, produced in BioRender (BioRender, 2023).

Building and Testing

Phase I: Homologous structure

Our first challenge was constructing a high-fidelity homologous structure for PhaF, since there exists no available crystallized model, in part a consequence of its complex structure. We utilized alphafold 2 as a way of acquiring a protein that presented PhaF’s main properties; i.e. an amphiphilic composition, a highly conserved N-terminal region, and supercoils in its C-terminus. We then acquired the following structure (Figure 17.) (Jumper et al., 2021):

Fig.17 — Figure 17. Mutant PhaF structure generated by alphafold2, in red, is the N-terminus region, while in blue it's the C-terminus.

After subjecting our Mutant PhaF to an energy minimization through GROMACS, we observed an increase in van der Waal forces, leading to misfolding. This prompted a reevaluation of our approach to identify potential misfolding and its impact on its interaction with PHB. This iterative process guided us in refining our structure.

As we revisited the literature, we recognized the pivotal role of Bi1 in its interaction ability, as demonstrated in past wet-lab experiments (Mezzina & M. Julia Pettinari, 2016). Accordingly, we decided to truncate the protein up to the 52nd amino acid for subsequent phases. The decision was driven by our exploratory mindset, guided by comprehensive literature .

Phase II: Homologous structure for Truncated proteins

With our insight, we set about constructing our Mutant and wild-type proteins. Once again we turned to alphafold2 while refining its properties in Pymol. Through this process, we assessed their electrostatic forces to discern potential conformations. Notably, we observed potential inward folding, due to an increase in negative potential and heightened hydrophobicity for our mutant. This can be observed below in a molecular dynamic simulation run by GROMACS.

Phase III: Autodock Vina

We then employed AutoDock Vina—a tool that leverages genetic algorithms and local optimization approaches to predict ligand-protein interactions with a high degree of accuracy (Eberhardt et al., 2021). This allowed us to gain a deeper understanding of the interaction between the wild type and mutant. We then utilized the PHB monomer as our ligand, as it possesses lower degrees of freedom, ensuring more efficient results compared to docking its polymer.

The docking revealed notable differences in binding affinity. As observed, the predicted model scores a binding affinity greater for the Mutant: Monomer= -2.8 kcal/mol, while for Wildtype: Monomer= -2.0 kcal/mol at an RMSD score of 0.0 for both respectively.

Fig.18 — Figure 18. PHB monomer binding to Bi1 domain of mutant, binding affinity is indicated, it interacts with LEU, GLY & TRP.

Learning and Improving

As discussed earlier, the pursuit of learning and improving was not something we only did at the end of our project, but rather an ongoing process, considering the challenges that continually arose. We consulted resources for the software tools in use, in addition to seeking guidance from our advisors. However, we quickly learned from our project that success in the learning and improvement step is only the beginning of another engineering cycle and that many rounds will eventually lead to a finished project.

Despite time constraints preventing us from conducting molecular dynamic simulations for both the wildtype and mutant about PHB. We gathered the knowledge gained for this part of the project and optimized our methodology and script for our molecular dynamics so future IGEM teams working with MDs, particularly in simulating biopolymer and protein interaction, can have a guided process.

References

Eberhardt, J., Santos-Martins, D., Tillack, A. F., & Forli, S. (2021). AutoDock Vina 1.2.0: New Docking Methods, Expanded Force Field, and Python Bindings. Journal of Chemical Information and Modeling. https://doi.org/10.1021/acs.jcim.1c00203

Iqbal, S., & Hoque, M. T. (2015). DisPredict: A Predictor of Disordered Protein Using Optimized RBF Kernel. PLoS ONE, 10(10), e0141551. https://doi.org/10.1371/journal.pone.0141551

Jendrossek, D., & Pfeiffer, D. (2014). New insights in the formation of polyhydroxyalkanoate granules (carbonosomes) and novel functions of poly(3-hydroxybutyrate). Environmental Microbiology, 16(8), 2357–2373. https://doi.org/10.1111/1462-2920.12356

Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., Bridgland, A., Meyer, C., Kohl, S. A. A., Ballard, A. J., Cowie, A., Romera-Paredes, B., Nikolov, S., Jain, R., Adler, J., & Back, T. (2021). Highly accurate protein structure prediction with AlphaFold. Nature, 596(7873), 583–589.

Maestro, B., Galán, B., Alfonso, C., Rivas, G., Prieto, M. A., & Sanz, J. M. (2013). A New Family of Intrinsically Disordered Proteins: Structural Characterization of the Major Phasin PhaF from Pseudomonas putida KT2440. PLoS ONE, 8(2), e56904. https://doi.org/10.1371/journal.pone.0056904

Mato, A., Tarazona, N. A., Hidalgo, A., Cruz, A., Jiménez, M., Pérez‐Gil, J., & Prieto, M. A. (2018). Interfacial Activity of Phasin PhaF from Pseudomonas putida KT2440 at Hydrophobic–Hydrophilic Biointerfaces. Langmuir, 35(3), 678–686. https://doi.org/10.1021/acs.langmuir.8b03036

Mezzina, M. P., & Pettinari, M. J. (2016). Phasins, Multifaceted Polyhydroxyalkanoate Granule-Associated Proteins. Applied and Environmental Microbiology, 82(17), 5060–5067. https://doi.org/10.1128/aem.01161-16