One of key iGEM values are cooperation and good sportsmanship. In this page we have gathered information and work that may be useful for the future iGEM teams and help them succeed in their own projects. Other than adding new parts and user experience to iGEM registry, we described issues we came across while designing plasmids that could have been easily prevented had we thought of them before. We also wrote a beginner guide to Rosetta modelling aimed for people who have little to no experience in computational biology. In addition, we have summarized our research regarding functioning and designing Mtr pathway for E. coli which can be used for development of electrochemical biosensors, as well as an overview of possible methods of concentrating various compounds inside the cells.
PARTS
This year Aalto-Helsinki 2020 has incorporated five basic parts and one composite part to the iGEM Registry Parts page. The list and detailed descriptions can be found on our parts page. We have also added experience to the BBa_K2560001 part.
PLASMID DESIGN TROUBLESHOOTING
During laboratory work, many concerns regarding plasmid design for successful cloning became apparent to us. These problems could have easily been alleviated during the DNA designing stage, had we knew about these pitfalls beforehand. This page is a collection of issues that we did not consider during our DNA design, and solutions to them. We also included other tips and tricks that can be used while cloning.
Restriction
When designing restriction sites for cloning, remember to add a few extra base pairs to ensure that the restriction enzyme can bind efficiently to restriction recognition sites. Restriction enzymes require a few extra base pairs of the DNA around the restriction recognition site to bind efficiently, however different enzymes have different requirements so be sure to double check the requirement for your intended enzyme. New England biolabs have a convenient webpage where you can see the required number of base pairs from end for many restriction enzymes [1]. However, even the requirements for the same enzyme can vary between manufacturers, so be sure to have a look at what your specific enzyme manufacturer recommends. When designing restriction sites close to ends of the DNA, make sure to add a few extra base pairs of DNA for successful cleaving.
Methylation of restriction sites can become a problem if the sequence around the restriction site matches the specific required sequence of a naturally occurring DNA methyltransferases inside the cell. While a methylated restriction site does not always affect the restriction of DNA, it can block or inhibit the restriction enzyme. The effect of DNA methylation on restriction varies between different restriction enzymes, there are also different methyltransferases in different hosts. The most common bacterial methyltransferases in laboratory strains are Dam, Dcm, EcoKI and EcoBI [2]. While designing your DNA, make sure to check the sequence of DNA close to the restriction site to avoid methylation related problems during cloning.
PCR
We used benchling while designing our DNA constructs and planning our experiments [3]. One of the features in benchling is to quickly give the melting temperature of primers. However these melting temperatures are calculated for certain reaction conditions, and depending on the conditions in the polymerization reaction, the melting temperature for the primers can differ from those given. For example, we often used Thermo Fisher Scientific Phusion high-fidelity polymerase. In this PCR reaction setup, the melting temperature of our primes was usually around 10 °C higher than what benchling calculated without any modifications to the parameters in their algorithm [4]. When designing primers or primer binding sites, remember to take into account the conditions of your specific PCR reaction.
When ordering synthesized DNA sequences, where the ends of the sequences play little importance for the cloning application, it is a good idea to use the same exact ends for many different DNA. To clarify, these ends will be cut off during digestion and not be present in the final construct. When you have the same ends, you can use the same primers for amplifying most of your DNA fragments. This not only decreases the amount of primers needed, but also requires less work, since factors such as melting temperature, GC content and GC clamp at 3' end need to be considered only once. However, it is still necessary to check the specificity of the primers for each DNA sequence. When designing DNA sequences for which the sequence of the ends can be arbitrary, limit the amount of primers needed by using the same ends for all sequences. Melting temperature of primers should be calculated only for binding DNA and not the overhangs. Ideally, at least half of your primer should encompass the existing sequence, but we were successful even with one third. We would suggest both starting and ending the primer with 1-2 G/C pairs.
Codon Optimization
Seemingly arbitrary changes done to DNA can have unwanted consequences if they affect secondary structure of mRNA. For example when doing codon optimization, or when changing part of a gene such as RBS, the change in nucleotides may cause unwanted base pairing in the mRNA blocking ribosomes and thus inhibiting translation. There is a great tool for checking possible secondary structures available online [5]. Make sure that no obvious secondary mRNA structures arise when changing parts of the DNA in a gene (Fig. 1).
MoClo using FastDigest enzymes
In our project we utilized the MoClo cloning technique to facilitate scar less cloning with multiple inserts at once [6]. However we did not have the specified reagents from the protocol, so we used BpiI (catalog number: FD1014) and EcoR31I (catalog number: FD0293) FastDigest restriction enzymes and T4 ligase (catalog number: EL0011), in FastDigest buffer (catalog number: B64) supplemented with 0.5mM ATP all from Thermo Fisher Scientific. According to instructions by Haddock et al., a total volume of 10 μL and 10 fmol of each DNA fragment should be used, but we had better success by following the advice of The Sainsbury Laboratory while doing our reactions [7]. Specifically, doubling reaction volume and only using half amount of destination vector DNA (10 fmol vector and 20 fmol insert DNA fragments) was very helpful. We also used the team's iGEM18_Marburg entry vector with RFP dropout, eliminating the need for IPTG and X-Gal as suggested by Haddock and colleges [6, 8]. Furthermore, we changed the conditions in the thermocycler, to be more suitable to the FastDigest restriction enzymes, and more cycles to help with the assembly. The thermocycler was cycling between the optimal restriction and ligation temperatures for a total of 50 cycles, 37 °C for 1.5 minutes for restriction, and 16 °C for 3 minutes for ligation. An extended optimal temperature restriction was performed at 37 °C for 5 minutes, to minimize the amount of re-ligated backbone plasmid. Finally, the enzymes were inactivated by heating the mixture to 80 °C for 10 minutes.
ROSETTA GUIDE FOR UNEXPERIENCED COMPUTATIONAL BIOLOGISTS
This year, Aalto-Helsinki 2020 iGEM team took part in a journal initiative organized by the Maastricht 2020 iGEM team, in which we wrote a step-by-step guide for designing ligand binding sites. This guide is based on a paper written by Moretti et al. [9] and a previous guide made by iGEM Technion 2016 and iGEM TU Eindhoven 2016 written during the iGEM 2016 Competition. Nonetheless, our guide is thought to be used by non-computational biologists, so it has more explanations and more concrete steps. We hope our Guide for Using Rosetta when Designing Ligand Binding Sites helps future iGEM participants to use this software even if they lack computational biologists in their teams.
The reader can have a look at our Rosetta modelling process by visiting our modelling page, but if they are interested in using the Rosetta software, they can download our step-by-step guide.
Mtr PATHWAY IN Escherichia coli
Why electrochemical biosensors?
Biosensors are a very common project topic for iGEM teams. As an example, all the Nordic teams participating in our ethics workshop were focusing on developing biosensors this year. We went through almost 70 previous iGEM biosensor projects to see how these sensors are usually built and found out that transducers are almost without an exception optical: colorimetric or fluorescent. Without surprise, these sensors are typically whole-cell biosensors engineered from E. coli.
When researching biosensors in general, it became apparent that in addition to optical biosensors, electrochemical biosensors are very popular. Electrochemical biosensors are capable of measuring turbid samples, which is an advantage in our case when working with effluent wastewater [10]. Electrochemical biosensors also tend to show higher sensitivity compared to optical biosensors, which is an important factor for determining low macrolide concentrations present in wastewater [11]. Based on these findings, we proceeded to search for a pathway that could provide E. coli with an ability to give an electrochemical output.
Mtr pathway
After searching for different options, we came across Mtr pathway of Shewanella oneidensis. Mtr pathway enables extracellular electron transport: movement of electrons from inside of the cell to outside. As E. coli is a model organism and our universities did not have experts working with S. oneidensis, we decided to research E. coli as a heterologous host for Mtr pathway. When cells are immobilized on an electrode, electrons from the Mtr pathway moving to extracellular space and further to the electrode will create an electric current. Additionally, if direct electron transport is not efficient, it can be enhanced with mediators via so called electron shuttling as occurs in S. oneidensis with flavins [12].
There are four main components needed to construct the Mtr pathway into E. coli: MtrC, MtrA, MtrB and CymA (Fig. 2). CymA is non-native cytochrome for E. coli, which is beneficial in biosensing applications because it has a faster re-reduction rate than NapC, a native E. coli cytochrome [13]. In addition to these, research groups have added cytochrome maturation genes into their constructs [12-14]. It is proposed that engineering of these maturation genes will increase biosensing abilities of E. coli [15]. Lactate functions as an electron donor for the Mtr pathway in E. coli [13]. Electrons released from lactate oxidation will transfer to the electrode via multiple intramolecular electron transfer events. First major component of this chain is CymA from which electrons are transferred to MtrA. From MtrA electrons move to MtrC via membrane pore MtrB. Electrons from MtrC will move to extracellular space and to the electrode. Lactate is a good substrate for Mtr pathway because the cell is forced to use Mtr pathway in anaerobic conditions whereas glucose can be fermented by E. coli, not leading to utilization of Mtr pathway. Many of the papers discussed Mtr pathway in E. coli as being a promising option for biosensing, which is why we decided to look at the idea of electrochemical biosensor further.
Our electrochemical biosensor design
The aim of our project is to create a biosensor from E. coli to detect and quantify macrolide antibiotics, such as erythromycin, from wastewater. Our idea was to express all the components needed for the Mtr pathway in E. coli independently of the concentration of erythromycin except for mtrB, which would have been under the erythromycin inducible pMphR promoter (Fig. 3). This way no current would have been produced when there are no macrolides, but increasing the amount of them would induce current production. In the ideal case, our cells would be immobilized to an electrode so that at the same time the electron flux would be efficient for macrolide concentration determination.
We found out that TU Delft iGEM team 2014 had a very similar idea on utilizing Mtr pathway for biosensor application, but instead of macrolide antibiotics they wanted to detect landmines. When reading through their wiki, we realized how much effort should be put into the Mtr pathway itself, even without considering the inducibility of mtrB. We contacted Michael Lienemann from VTT (Technical Research Centre of Finland) who has worked with Mtr pathway in E. coli to discuss feasibility of a biosensing project with Mtr pathway. The main identified obstacles were found to be immobilization and anoxic environment.
It is important for cells to attach to an electrode and transfer electrons as efficiently as possible. As team TU Delft, we considered using conductive curli or, as another option, conductive pili, but it seemed there was not enough research on them for us to be able to utilize them properly. Lienemann also mentioned that we should consider the immobilization technique carefully, as conductivity might decrease because of the properties of certain materials. In order for E. coli to utilize Mtr pathway, there should be an anaerobic or anoxic environment, which we also found to be a challenge. When cells are in an aerobic environment, oxygen will be reduced so there is no pressure for the cell to utilize the Mtr pathway to get rid of electrons. Considering also the final biosensor product, we should have thought how to make the measurement possible in wastewater treatment plants where there are no anaerobic tents available.
To get more perspectives, we also had a discussion with Lasse Murtomäki, professor of physical chemistry and electrochemistry in Aalto University, about biosensors and various electrochemical outputs. Wastewater contains so many different compounds that there is a risk of unknown compounds getting oxidised on the electrodes and a risk of electrode fouling as there many compounds that might be able to attach to the electrode better than cells. In a discussion with Silvan Scheller, professor of biochemistry, and Norman Adlung, postdoctoral researcher, we concluded that Mtr pathway should work properly in E. coli before it would be sensible to use it for a biosensor application. Scheller also noted that a bigger issue was not sensitivity of the output signal but the low concentration of macrolides, which could be tackled by other means.
Conclusion
To conclude, we decided that either we should focus solely on improving Mtr pathway in E. coli and contributing for its development for biosensing applications in the future or developing a macrolide-sensing circuit in E. coli. Due to the limited time of three months in the lab and our lack of electrochemical background, we chose the latter, as it more likely gives us concrete results. As low concentration of macrolides is our primary challenge, it also seemed more urgent to try to concentrate macrolides instead of focusing on the output signal, which would have a minimal impact on sensitivity.
CONCENTRATING MACROLIDES INSIDE OUR BIOSENSOR
One of the problems we had to consider when it came to applying our biosensor to wastewater was the low concentration of macrolide antibiotics. Our research had suggested that the amount of macrolide antibiotics in wastewater is very low, in the magnitude of 10-9 to 10-10 moles per litre, which meant that one key thing to improve the function of our biosensor would be to concentrate the macrolides inside our whole-cell biosensor. We considered both methods of synthetic biology and microfluidics. Here we will detail concentration efforts with means of synthetic biology.
We did some ideation and research on the subject and came up with several different possible strategies concentrating macrolide antibiotics in our cells. First, we focused on the efflux pumps found in our E. coli BL21 strain. An easy way to increase antibiotic concentration inside the cell would have been to remove macrolide efflux pumps from our bacteria. However, we quickly realised that our strain does not have efflux pumps specific to macrolide antibiotics, therefore there was nothing to remove. Another fun idea we considered was to reverse macrolide efflux pumps to turn them into influx pumps. This way we could have potentially actively pumped macrolides inside our biosensor. Unfortunately, there was little to no research about this topic, and we thought it to be too time-consuming for our time in the laboratory.
The first feasible option we came upon in our research was removing TolC, an outer membrane protein that is a part of several multidrug resistance pumps. Removing TolC has been shown to increase the concentrations of various antibiotics inside E. coli, including erythromycin. Without TolC, many multidrug resistance efflux pumps of E. coli cannot pump out macrolide antibiotics as efficiently and their concentration inside the cell grows [16][17]. We also considered utilizing an E. coli with a modified LPS. Mutant “deep-rough” LPS makes it easier for antibiotics to access bacterial cells, and an E. coli strain with this kind of LPS profile would have likely been much more susceptible to macrolide antibiotics [18][19].
Another way to increase concentration inside a cell is to increase the influx of macrolides to the cell. Mohammad et al. describes in a paper from 2010 a redesigned plugged β-barrel membrane protein. The protein, originally called FhuA, was modified by removing its cork and a few of its inner loops. This way they were able to turn the protein from a selective, plugged membrane protein to a protein that forms a large open pore [20]. The modified protein is known as FhuAΔC/Δ4L. Krishnamoorthy et al. utilised this modified pore protein to hyperporinate E. coli outer membrane. Hyperporination of the outer membrane allows big hydrophobic antibiotics such as macrolide antibiotics to easily pass through it. This increases the amount of macrolide antibiotics inside the cell, since of the main limitations of macrolide antibiotics entering gram-negative bacteria is their poor ability to cross the outer membrane [17].
The best result however, seemed to result from combining the two aforementioned techniques. Removal of TolC and hyperporination of outer membranes together yield a better result than either of them on their own. Concentration of intracellular erythromycin was monitored by using carbon-14, a radioactive carbon isotope. When erythromycin is marked with carbon-14 the faint radioactivity can be used to monitor levels of erythromycin inside the cell.
Concentrations of [14C] erythromycin were highest in cells where TolC was removed and they had been hyperporinated, even significantly exceeding external concentration [17]. This would be ideal for our biosensor, as this way our biosensor could work in lower concentrations of macrolide antibiotics in wastewater.
In order to test the effects of TolC deletion and hyperporination we contacted the authors of Krishnamoorthy et al. in Oklahoma. They agreed to give us the strain GKCWT104, in which TolC has been deleted and a gene encoding the modified FhuAΔC/Δ4L added to the bacterial chromosome [17]. We made a material transfer agreement with them and received the strain in the mail. Our objective is to test whether the increased macrolide uptake to the E. coli will increase the sensitivity of our biosensor. We are very grateful for Krishnamoorthy and the colleagues for their contribution to our project.
REFERENCES
1. Biolabs, N. E. (n.d.). Cleavage Close to the End of DNA Fragments | NEB. Retrieved 3 September 2020, from https://www.neb.com/ tools-and-resources/ usage-guidelines/ cleavage-close-to-the-end-of-dna-fragments
2. Restriction Enzyme Key Considerations—FI. (n.d.). Retrieved 3 September 2020, from //www.thermofisher.com/uk/en/home/ life-science/ cloning/ cloning-learning-center/ invitrogen-school-of-molecular-biology/ molecular-cloning/ restriction-enzymes/ restriction-enzyme-key-considerations.html
3. Sign In · Benchling. (n.d.). Retrieved 3 September 2020, from https://benchling.com/signin/welcome
4. PhusionTM High-Fidelity DNA Polymerase (2 U/µL). (n.d.). Retrieved 3 September 2020, from https://www.thermofisher.com/order/catalog/product/F-530XL
5. Reuter, J. S., & Watson, R. M. (2017). Welcome to the Predict a Secondary Structure Web Server. Predict a Secondary Structure Web Server. https://rna.urmc.rochester.edu/ RNAstructureWeb/ Servers/ Predict1/ Predict1.html
6. Haddock, T. L., Densmore, D. M., Appleton, E., Carr, S., Iverson, S., De Freitas, M., Jin, S., Awtry, J., Desai, D., Lozanoski, T., Shah, P., Agarwal, Y., Lewis, K., & Pacheco, A. (2015). BBF RFC 94: Type IIS Assembly for Bacterial Transcriptional Units: A Standardized Assembly Method for Building Bacterial Transcriptional Units Using the Type IIS Restriction Enzymes BsaI and BbsI. https://dspace.mit.edu/handle/1721.1/98267
7. TSL SynBio. (n.d.). Retrieved 9 September 2020, from http://synbio.tsl.ac.uk/docs/item/f5dffe61-b762-4ee1-b8a3-eb90f965e49c
8. Stukenberg, D. (2018). Part:BBa K2560001—Parts.igem.org. Phytobrick Entry Vector with RFP Dropout. http://parts.igem.org/ Part:BBa_K2560001
9. Moretti, R., Bender, B. J. & Allison, B. (2016). Methods in Molecular Biology: Design and Creation of Ligand Binding Proteins, 1414, 47–62. https://doi.org/10.1007/978-1-4939-3569-7
10. Lagarde, F., Jaffrezic-Renault, N. (2011). Cell-based electrochemical biosensors for water quality assessment. Analytical and Bioanalytical Chemistry, 400, 947-964, doi:10.1007/s00216-011-4816-7
11. Yi, H., Li, M., Huo, X., Zeng, G., Lai, C., & Huang, D. et al. (2019). Recent development of advanced biotechnology for wastewater treatment. Critical Reviews In Biotechnology, 40(1), 99-118. doi: 10.1080/07388551.2019.1682964
12. Lienemann, M., TerAvest, M. A., Pitkänen, J.-P., Stuns, I., Penttilä, M., Ajo-Franklin, C. M., Jäntti, J. (2018). Towards patterned bioelectronics: facilitated immobilization of exoelectrogenic Escherichia coli with heterologous pili. Microbial Biotechnology, 11(6), 1184-1194. doi: 10.1111/1751-7915.13309
13. TerAvest, M. A., Zajdel T. J., Ajo-Franklin, C. M. (2014). The Mtr Pathway of Shewanella oneidensis MR-1 Couples Substrate Utilization to Current Production in Escherichia coli. ChemElectroChem, 1, 1874-1879. doi: 10.1002/celc.201402194
14. Jensen, H. M., TerAvest, M. A., Kokish, M. G., Ajo-Franklin, C. M. (2016). CymA and Exogenous Flavins Improve Extracellular Electron Transfer and Couple It to Cell Growth in Mtr-Expressing Escherichia coli. ASC Synthethic Biology, 5, 679-688. doi: 10.1021/acssynbio.5b00279
15. Su, L., Fukushima, T., Prior, A., Baruch, M., Zajdel, T. J., Ajo-Franklin, C. M. (2020). Modifying Cytochrome c Maturation Can Increase the Bioelectronic Performance of Engineered Escherichia coli. ASC Synthetic Biology, 9, 115-124. doi: 10.1021/acssynbio.9b00379
16. Zhong, P., & Shortridge, V. D. (2000). The role of efflux in macrolide resistance. Drug Resistance Updates, 3(6), 325-329. doi:10.1054/drup.2000.0175
17. Krishnamoorthy, G., Wolloscheck, D., Weeks, J. W., Croft, C., Rybenkov, V. V., & Zgurskaya, H. I. (2016). Breaking the permeability barrier of escherichia coli by controlled hyperporination of the outer membrane. Antimicrobial Agents and Chemotherapy, 60(12), 7372-7381. doi:10.1128/AAC.01882-16
18. Tsujimoto, H., Gotoh, N., & Nishino, T. (1999). Diffusion of macrolide antibiotics through the outer membrane of moraxella catarrhalis. Journal of Infection and Chemotherapy, 5(4), 196-200. doi:10.1007/s101560050034
19. Møller, A. K., Leatham, M. P., Conway, T., Nuijten, P. J. M., de Haan, Louise A. M., Krogfelt, K. A., et al. (2003). An escherichia coli MG1655 lipopolysaccharide deep-rough core mutant grows and survives in mouse cecal mucus but fails to colonize the mouse large intestine. Infection and Immunity, 71(4), 2142-2152. doi:10.1128/iai.71.4.2142-2152.2003
20. Mohammad, M. M., Howard, K. R., & Movileanu, L. (2011). Redesign of a plugged β-barrel membrane protein. The Journal of Biological Chemistry, 286(10), 8000-8013. doi:10.1074/jbc.M110.197723
