Team:Athens/Engineering
FOR THE FULL EXPERIENCE PLEASE USE A PC :)
Engineering Success
Design
The goal of the project was to create a biological system capable of producing dNTPs and Pfu polymerase in quantities that are on par with commercial standards, for PCR experiments and other NAA techniques. The process began with a systematic search of the literature, in order to find ways organisms in nature produce dNTPs. Through this search we managed to understand three key points:
- All organisms, prokaryotic and eukaryotic, share a very similar metabolic pathway for the de novo production of dNTPs [3]. When needed, cells can also use salvage pathways to produce essential quantities of the dNTPs [1-3].
- The rate limiting steps in these metabolic pathways were reported by many sources to be the enzyme ribonucleoside diphosphate reductase (RNR) and thymidylate synthase (TSase) [2-4]. In short, RNR catalyzes the formation of deoxyribonucleotides dADP, dGDP, dCDP, dUDP from ribonucleotides ADP, GDP, CDP, UDP respectively while TSase takes part in the catalyzation of the conversion of dUDP to dTDP. Kinases are responsible for the phosphorylation of the nucleotides, but their activity is far more rapid. As such nucleotides, di- and triphosphates are readily interconvertible and will be used interchangeably [4].
- The level of each dNTP pool inside the cell is heavily regulated through the allosteric regulation of the RNR enzyme. Failure to have balanced dNTP pools may lead to severe mutagenicity [1-3].
Based on these observations we derived these design decisions:
- We would create a biological system that overexpress the RNR and TSase enzymes through a well constructed, closely regulated genetic circuit. This system would be able to produce high levels of dNTPs and must be examined for possible mutagenicity that may lead to loss of function or the plasmid as a whole.
- Salvage pathways had a low yield and were used in cases where the cell was in unfavorable conditions. As such the design process would not include them.
- The relationship between the activities of RNR and TSase is not well documented so their overexpression levels should ideally be regulated by different inducers.
- The selected chassis was Escherichia coli BL21 (DE3) because its fast reproduction rates would aid the efficiency of the project.
Additionally, as stated above, we wanted the system to produce a DNA polymerase suitable for PCR reactions. After thorough conversations with Dr. Alissandratos, documented in the Human Practices page, we opted for the Pfu polymerase due to its high stability and low error rate [5]. However, we wanted our project to be applicable to other polymerases too, so the design process was generalized to fit any NAAT applicable polymerase, as shown in the Contributions page.
In order to test the function of the proposed system we designed, based on the literature, protocols that would purify and measure the dNTPs and the Pfu polymerase were adapted. This is possible through the use of HPLC for the measurement of dNTPs and the addition of N-His-tag on the Pfu polymerase which would be purified through affinity chromatography. More on the protocols can be found in the Protocols tab.
However, given the constraints involved with our designed system (mutagenicity and ribonucleotide levels) we wanted to create different approaches towards the desired solution:
Plan B
One of the main hypotheses of the project was that the system will have adequate, constant ribonucleotide (NTPs) levels that would be used for the formation of dNTPs. This statement was supported by the fact that biological organisms contain far more NTPs due to their use in many different applications. An example of our working hypothesis is based on the experimental results of Zhu et al. [2]. Despite that, we wanted to design a system that would be able to produce higher levels of NTPs given the off-chance that our hypothesis would fail. Searching through the literature, we pinpointed some enzymes able to catalyze the formation of NTPs [6-7]. By implementing those enzymes into a genetic circuit similar to that of RNR and TSase, we could resolve the issue of low quantities of substrates.
Plan C
Another concerning issue related to the stability of the biological system was that unstable levels of dNTP pools result in high mutagenicity [1-3]. In that scenario, we would use the original system to only overexpress RNR and TSase enzymes and immediately turn it into a lysate that would be used as “dNTP-formation master mix” when placed in a mixture with high levels of NTPs. This would diverge from some original idea of the project, but could provide a safe alternative for producing dNTPs in less expensive ways.
Model DBTL cycle
Due to the novel specifications and goals of our project we wanted to have some first estimations of its abilities. So we sought to create a kinetic model that would describe the enzymatic activities of RNR and TSase. Unfortunately, there has not been an extensive kinetic model describing the targeted metabolic path for over 30 years and the kinetic data documented in the literature are scarce [4]. Despite that, we used the available information to build a model that would give some rough estimates of the project’s capabilities regarding the rate of formations and time scales needed. Further documentation of the model’s structure and assumptions can be found in the Modeling page. Through the first results three qualitative observation could be made:
- The modeled system had immense changes on dNTPs’ concentrations and rate of formations when kinetics for overexpression of RNR and TSase were introduced. The time needed for the production of the desired quantities of dNTPs (i.e. 0.4 mM, a usual concentration in commercial PCR master mixes) was satisfying, meaning that the project’s goals of efficiency and cost effective production were plausible.
- The overexpression of RNR and TSase levels showed no considerable changes if their regulation originated from two different promoters or one common promoter.
- The system was very sensitive to the values of the initial concentrations of dNTPs, resulting in vastly different final concentrations. This should be taken into consideration when performing the wet lab measurements (measurements should be taken to estimate the starting dNTP concentrations of our system).
To resolve some of the uncertainties contained in the model, we would use dNTPs' concentration measurements to pinpoint some kinetic parameters' values. Moreover, we designed an additional plasmid containing transcriptional units for RNR and TSase, linked with GFP and RFP proteins respectively. This would allow us to have experimental data for the quantification of the protein expression and denaturation kinetic parameters. Through this enhancement, the model could provide better estimates of the time scale of the system and better optimize the quantities used for the inducers.
Build
Here are presented the originally designed steps to be followed in order to complete the AdAPTED project.
After extensive research and discussions with Dr Alissandratos, we decided to split the production of dNTPs and Pfu polymerase into two plasmids, since having a smaller insert would drastically increase the yield of ligation and transformation. At the same time, the project would not be compromised, as those two products would need different processing post production anyway. Therefore, it was decided to construct two genetic circuits using Snapgene. The first genetic circuit includes two TUs that encode enzymes RNR and TSase, which are to be inserted in the destination plasmid pSB1C3. The second genetic circuit corresponds to Pfu polymerase and is to be inserted in the destination plasmid pGGA. For enzymes RNR and TSase, another version was designed that also contains a G4SAS linker of 21 bp to connect to a different fluorescent protein for each, in order to help with the measurements needed for the modeling. Through the literature it was ensured that the active site is not close to the N-terminal of the protein, so the linker will not affect the enzyme's activity [8]. The parts that make each TU (promoter, RBS, gene, fluorescent protein, terminator) were ordered separately, as presented in Table 1.
| TU | Fragment |
|---|---|
| RNR | T7-LacO Promoter |
| RBS | |
| nrdA (or nrdA-linker and linker-GFP) | |
| RBS | |
| nrdB | |
| Terminator | |
| TSase | AraC Promoter |
| RBS | |
| TSase (or TSase-linker and linker-mRFP) | |
| Terminator | |
| Pfu | T7 Promoter |
| RBS | |
| Pfu | |
| Terminator |
The decision to split the parts of each TU in this way was made due to the size of the genes and limitations related to ordering the DNA fragments, as well as the desire to be able to change promoters, RBS and terminators, in order to have better measurements for the modeling and more flexibility during the experiments. These limitations also lead to separately ordering the two genes (nrdA and nrdB) that make RNR. In addition, Pfu polymerase was built with a polyhistidine-tag; a six histidine (His) residues motif in the N-terminus of the protein. This is done to ensure that it is possible to identify and isolate the polymerase, as well as measure its concentration.
The ends of all fragments in the RNR and TSase are built to ligate using BsaI and T4 DNA Ligase, in order to create the respective TU. The TUs are designed to ligate with one another using SapI and T4 DNA Ligase, and with the plasmid using NotI and EcoRI restriction enzymes and T4 DNA Ligase. This decision was made in order to provide the wet lab members with different options, as for the combinations of the parts, when ligating the system. The final design of the plasmid can be seen in Figure 1.
The process was designed as follows: after the ligation of the TUs with their respective plasmid, transformation of two different Escherichia coli BL21(DE3) takes place, followed by the isolation and purification of the desired products, the the four dNTPs and Pfu polymerase. To isolate and purify Pfu polymerase, the transformed cells are grown in liquid cultures, centrifuged and broken with ultrasound. Then a gravity flow column chromatography with Nickel resin is used to collect the protein in imidazole. The protein is identified using sodium dodecyl sulfate polyacrylamide gel electrophoresis method (SDS-PAGE) and the concentration is calculated using Lambert-Beer law. For the isolation and purification of dNTPs, high-performance liquid chromatography (HPLC) is used with a C18 column. The four dNTPs are eluted at different times (as documented in literature measurements, the elution time for dCTP is 16.9 min, dGTP 23.8 min, dTTP 24.8 min and dATP 25.6 min [5]). The detailed description of the protocols used for the AdAPTED project can be found in the Protocols page.
Test & Learn
Here, we present the process that was carried out, based on the previous step, in order to complete the AdAPTED project.
After the build process, all elements of each TU were ordered, obtained, amplified with PCR and the ligation protocol for each TU was followed. During this process, the wet lab team came across two problems.
First, all genes were electrophoresed before and after amplification. However, both nrdA and nrdA-linker amplified fragments were found to have an extra band after the electrophoresis at approximately 500 bp, as seen in Figure 2. This observation also explained why later, when ligation was attempted for the RNR TU, the experiment was unsuccessful, as either the nrdA or nrdA-linker gene were not able to ligate with the rest of the TU elements. That was likely because the impurities were also able to be amplified using the same primers and thus ligated, since they had similar ends. In addition, before obtaining these genes, the company that synthesized these elements informed us that there were problems with the synthesis of those particular fragments and could achieve a product yield that would pass their QC requirements. After several attempts on their end, we were able to obtain the fragments, but unfortunately there were impurities that did not allow for that segment of our project to blossom.
The second issue our wet lab team came across was the complexity of the ligation. Since the initial design required the ligation protocol to be performed using four different restriction enzymes for the RNR and TSase TUs, the ligated DNA was significantly decreased after each consecutive step (Creation of TUs and Ligation of the two TUs). This process was very laborious, but at the same time, an important learning experience that familiarized the wet lab members with the techniques. Moving forward, we decided to simplify the design of our project and to attempt the original idea once again.
Design
After the previously described learning experience, our team decided to study and contact more professionals with year-long experience who could help with the design of a more simplified system that would still achieve the original goal of augmenting the dNTPs and Pfu polymerase production in E. coli. A change that we decided to incorporate to the project was to use the same promoter, T7-LacO, for both the RNR and TSase enzymes, to simplify the genetic circuit. Moreover, we decided to edit the ends of each one of the TUs, so as to enable a one pot ligation, that is more streamlined and that would not reduce the ligation yield. Finally, to recover the nrdA and nrdA-linker genes without the impurities that would impede the TU ligation process, our advisor, Anastasios Galanis, suggested to proceed with iterative gene amplification using PCR, electrophoresis and gel extraction to obtain only the desirable DNA fragment.
Build
After contacting experts and redefining our project approach, our team had to alter the originally built genetic circuit to reflect the new decisions. For the new system, the united TU that consists of RNR and TSase is to be inserted in the destination plasmid pGGA, while the TU for Pfu polymerase is designed to be inserted in a separate pGGA plasmid. For that reason, the 5’ end of the T7 lacO and T7 promoter as well as the 3’ end of the terminators used are edited to enable the ligation with restriction enzyme BsaI. In that way, each one of the two systems can be constructed in a one pot ligation reaction with BsaI and T4 DNA Ligase. However, the new system, that is shown in Figure 3, has a few limitations, compared to the previous one, for the sake of simplicity. The new TU that contains RNR and TSase does not allow for exchanging terminators, as the last fragment includes the gene for TSase and the newly designed terminator, along with the RBS. This decision was made in order to decrease the number of separate fragments that have to be ligated, in order to simplify the project. With the same goal in mind, the new T7-LacO promoter is connected with RBS and does not have to be ligated separately. As it is seen in Figure 3, it was also decided to not proceed with the ligation of nrdA with the fluorescent protein, as that particular fragment was more contaminated and the final RNR protein could be tracked by the fluorescent protein connected to TSase, that existed in the same TU.
In this way the system’s fragments that create the insert (with the fluorescent protein for detecting the production) are reduced from 16 to 8, split in two plasmids with 4 fragments each. In this way, the system is severely simplified enabling for an easier ligation process.
| TU | Fragment |
|---|---|
| RNR & TSase | *T7-LacO Promoter + RBS |
| nrdA | |
| RBS + nrdB | |
| *RBS + TSase-linker + mRFP + Terminator | |
| Pfu | *T7 Promoter |
| RBS | |
| Pfu | |
| *Terminator |
Test & learn
After designing and building the new genetic circuit, the new parts were ordered, obtained, amplified with PCR and the ligation protocol for each TU was performed. During this process, the wet lab team came across a new problem. The newly ordered RBS + TSase-linker + mRFP + Terminator, came at a very small concentration, (lower than 1ng/μL), as it did not appear in the electrophoresis of the new fragments (Figure 3) or in the concentration measurements with NanoDrop. After several attempts to increase the concentration of DNA through PCR, this part of the project was put to a halt, as the deadlines of the competition were approaching and the team decided that it would not be wise to proceed with re-ordering the fragment.
However the ligation and transformation of Pfu in a separate pGGA plasmid was successful. Afterwards, we proceeded with the transformation in E. coli BL21 (D3), protein isolation and purification, as described in the build section of the first circle. The isolated Pfu also served as the proof of concept for the original design of the project, since similar steps are needed to be followed for the isolation of dNTPs. The results can be found in more detail in the Proof of Concept page.
Learn & Future Plans
After all the previously described attempts, the deadlines of the competition did not allow for a new circle of designing, building and testing a new approach of our project. However, the whole experience provided the team with plenty of new knowledge and herein we document in detail all the steps that are to be followed in the future, in order to complete the project. The obvious solution for the contaminated RBS + TSase-linker + mRFP + Terminator fragment is to obtain the fragment again and ensure careful handling of the DNA during resuspension, as it is described in our Contributions page. However, in our case the particular fragment was resuspended at the same time as other fragments which were not found to be contaminated.
The next steps of the project include the ligation of the TUs that produce dNTPs with the plasmid, the transformation in E. coli, the production induction for dNTPs, the isolation and concentration measurements for each one of the four dNTPs. These measurements will be incorporated in the modeling and then the project will be repeated with different promoters for the two enzymes, RNR and TSase, as it was planned in the original design of the project. The final designed plasmid for the production of dNTPs that is intended to be distributed to the end users is shown in Figure 4. The system is controlled by two promoters and does not include the fluorescent proteins. As for the future of Pfu polymerase, a simpler approach for the isolation of the protein from the bacteria is planned, using thermal shock, since the polymerase is thermotolerant and it is expected to be the only molecule that will not be deactivated by the heat shock. Finally, performing a PCR with the newly produced Pfu polymerase, would be the last step, that would both give an estimate of the enzyme activity and also prove that our project is correctly designed and executed.
After these steps, it would be beneficial for the implementation of the project for multiple experiments to be performed with different concentrations of the inducer for the production of dNTPs and the culture growth time for the production of Pfu, in order to standardize the system. More information on why this would benefit the project can be found in the Proposed Implementation page.
1. Pezo V, Hassan C, Louis D, Sargueil B, Herdewijn P, Marlière P. Metabolic Recruitment and Directed Evolution of Nucleoside Triphosphate Uptake in Escherichia coli. ACS Synth Biol. 2018 Jun 15;7(6):1565-1572. doi: 10.1021/acssynbio.8b00048. Epub 2018 May 18. PMID: 29746092
2. Zhu M, Dai X, Guo W, Ge Z, Yang M, Wang H, Wang YP. Manipulating the Bacterial Cell Cycle and Cell Size by Titrating the Expression of Ribonucleotide Reductase. mBio. 2017 Nov 14;8(6):e01741-17. doi: 10.1128/mBio.01741-17. PMID: 29138305; PMCID: PMC5686538.
3. Torrents E. Ribonucleotide reductases: essential enzymes for bacterial life. Front Cell Infect Microbiol. 2014;4:52. Published 2014 Apr 28. doi:10.3389/fcimb.2014.00052
4. Jackson RC. A kinetic model of regulation of the deoxyribonucleoside triphosphate pool composition. Pharmacol Ther. 1984;24(2):279-301. doi: 10.1016/0163-7258(84)90038-x. PMID: 6379685.
5. Loan, T.D., Easton, C.J. & Alissandratos, A. DNA amplification with in situ nucleoside to dNTP synthesis, using a single recombinant cell lysate of E. coli. Sci Rep 9, 15621 (2019). https://doi.org/10.1038/s41598-019-51917-z
6. Ding, Y., Ou, L. & Ding, Q. Enzymatic Synthesis of Nucleoside Triphosphates and Deoxynucleoside Triphosphates by Surface-Displayed Kinases. Appl Biochem Biotechnol 190, 1271–1288 (2020). https://doi.org/10.1007/s12010-019-03138-3
7. Nahálka, J., Liu, Z., Gemeiner, P. et al. Nucleoside triphosphates production using recombinant Escherichia coli entrapped in calcium pectate gel. Biotechnology Letters 24, 925–930 (2002). https://doi.org/10.1023/A:1015508611935
8. Kolberg M, Strand KR, Graff P, Andersson KK. Structure, function, and mechanism of ribonucleotide reductases. Biochim Biophys Acta. 2004 Jun 1;1699(1-2):1-34. doi: 10.1016/j.bbapap.2004.02.007. PMID: 15158709.