Wet Lab
As part of our original plan, we designed all of our parts flanked with a set of M13 primers in order to have more DNA without the necessity of cloning in bacteria. We have successfully amplified by PCR most of our synthesized parts (toeholds, triggers and reporter genes). However some of our parts showed no amplification whatsoever, this led us to first set a small PCR reaction in order to confirm by electrophoresis before setting bigger reactions and potentially waste reagents.
We purchased the NEBExpress® cell-free E. coli protein synthesis system in order to characterize our parts, we tested the kit with a pET-28b plasmid provided by our instructor Chucho (who also sequenced the exact same plasmid) and tested one set of Agave tequilana toehold + trigger following the manufacturer's protocol but got no results, as we had two different kits we tested the second one being even more careful with thawing temperatures and handling in general, however the outcome was the same, we attribute it to bad shipping conditions. This situation led us to try to use our own E. coli lysate to test the kit (as we did not have access to the rest of the materials needed to set a cell-free protein expression reaction), we successfully prepared an IPTG induced E. coli BL21 DE3 lysate, we plated an aliquot of the extract and did not grow any colony, meaning that it would be safe to use in a non-controlled environment such in crop fields as no potentially harmful bacteria could be released.
However, as we tested using it instead of the provided E.coli extract in the kit we got the same results and did not achieve a noticeable nor measurable protein expression.
As we could not purchase nor prepare a working cell-free protein expression system we decided to clone our parts into plasmid so we could be able to test them in bacteria. First, our parts did not include any useful restriction sites in order to clone them, however, as all of them were flanked by M13 primers we just purchased M13 primers with EcoRI and PstI restriction sites overhangs so we could do PCR and have our parts with restriction sites. Secondly, we needed a set of compatible plasmids to cotransformate (different antibiotic resistance markers, ORIs from different compatibility groups and similar copy numbers). As we did not have resources nor time to purchase or synthesize any plasmid, we decided to use the plasmid pSB3C5 included in this year iGEM distribution kit along with the pUC19 plasmid we got as a part of a NEB sponsorship. Both plasmids fulfilled all of the previous requirements and included a MCS with EcoRI and PstI sites.
After planning all the protocols we quickly prepared E.coli competent cells from two different strains, DH5-α for cloning and BL21 DE3 for protein expression. We successfully digested both vectors as well as all of our pre-assembled toeholds and their corresponding triggers. We then ligated our toeholds into the pSB3C5 plasmid (low copy) and our triggers into the pUC19 plasmid (medium copy) and transformed them into DH5-α competent cells, performing in the process a Blue-white screening for the pUC19 plasmid ligations, confirming the insertion of our parts.
After picking white colonies we extracted the plasmidic DNA from the cultures and transformed the clones into Bl21 DE3 competent cells, both triggers and toeholds individually and cotransformed each toehold with its corresponding trigger, we successfully cotransformed only the Fusarium oxysporum toehold and trigger combination (although a colony PCR confirmed the absence of the toehold sequence, this absence was confirmed from the plasmid as its length corresponded only to the empty backbone, suggesting a ligation of the sticky ends left after the double digestion of the original plasmid), as the other cotransformations did not yield any colonies we produced competent cells from the single toehold transformations, and then transformed them with its corresponding trigger, having successfully grown colonies for the combination of Agave tequilana and Bois Noir toeholds and triggers. We only got to induce the culture of Agave tequilana with IPTG and saw no expression even after an overnight incubation that could be explained by the low concentration or absence of trigger.
Software
Although the toeholds used in our project were designed for Fusarium oxysporum and Agave tequilana, we decided to compare toeholds for different diseases reported in literature with the toeholds generated by Toehold Switch Creator, in order to demonstrate our software’s functionality. The comparisons were made with Bois Noir (caused by Candidatus phytoplasma solani ) [1], human orthopneumovirus (or Respiratory Syncytial Virus, RSV types A and B) [2] and Middle East Respiratory Syndrome (MERS) [3]. For this endeavour we used the same amplicon used by the original designers of these parts to generate our own structures and analyse 3 parameters: the score value, the minimum free energy (MFE), and the Gibbs free energy (ΔG). These parameters are extremely important for the prediction of a toehold switch viability (see more on Model) and by performing this analysis we aimed to prove the quality of our toehold’s structure.
In this part, for the design of the toehold switches by our software, first we identified the genome section that the authors selected as a target for the toeholds that they have designed, then we performed a BLAST search in order to find the nucleotide sequences related to these toehold switches. Next, our team selected one with the lowest E-Value and higher nucleotide identity. Once we got the desired sequence, we performed the toehold design with our software. It is important to note that the sequences in which our toeholds will link are not the same as those chosen by the other authors, however both belong to the same organism that will be detected. Just in the case of the detection of MERS, the accession number of the FASTA file was provided by the authors, in such a way that the BLAST search was not needed.
The accession numbers in NCBI database for the sequences taken from BLAST analysis were
- AF447593.1 for the creation of toeholds based on [1]
- JF920069.1 and MG642082.1 for the creation of toeholds based on [2] (RSV-A and RSV-B, respectively)
- JX869059.2 for the creation of toeholds based on [3]
The first analysis we perform is the MFE, this parameter is important because more negative values will represent higher switch-trigger concentration than switch in equilibrium [4], after performing the calculations for the MFE we found that the value obtained by our toeholds is comparable with the values for the structures generated by the other authors. In fact, we found that in the case of our MERS toehold, the MFE value of our toehold is superior to the toehold found in literature. We found that our toeholds have a variation of 5 Kcal/mol on average which demonstrates the quality of our toeholds, having 50% of them better than their literature counterpart.
On the other hand we also performed analysis of the score value obtained from the consideration of the toeholds single-strandness, the target-single-strandness and the ensemble defect, where lower values represent a preferred toehold structure (for a more detailed explanation of the score value see Model). We found for this analysis that 50% (2 out of 4) of the toeholds generated by our software obtained better results than their counterparts reported in the literature. Additionally, one of our toeholds (RSV_A) has a significantly better score than its counterpart. Since this parameter represents the linearity of our parts, it predicts that our toeholds will be as available for the toehold-trigger complex to happen, this is because if the target dimerizes during the reaction the toehold would not be able to form the complex and allow the expression of the reporter protein, having both of the parts as linear as possible will ensure that trigger will bind effortlessly to the switch.
Finally we analysed the ΔG of the toeholds with their respective triggers. Having more negative values of this parameter will make the unwinding process more difficult, leading to lower translation rates [4], the results of this analysis where in favour of the toeholds generated by our software, we obtained that for 3 out of the 4 toeholds analysed the results of the Gibbs free energy were more positive than negative, therefore foreseeing higher translation rates. In Addition to this analysis, in order to have a more graphical approach of the comparison of the structures, we obtained an image of the secondary structure of the toeholds using the NUPACK package. The structures of the left are the original toehold sequences reported in literature, and the structures of the right are the ones generated by our software.
References
- Ragios, K. & iGEM19_EPFL. (2019). Part:BBa_K2916055:Hard Information. Registry of Standard Biological Parts. Available at: http://parts.igem.org/cgi/partsdb/part_info.cgi?part_name=BBa_K2916055
- Cao, M., Sun, Q., Zhang, X., Ma, Y., & Wang, J. (2021). Detection and differentiation of respiratory syncytial virus subgroups A and B with colorimetric toehold switch sensors in a paper-based cell-free system. Biosensors and Bioelectronics, 182, 113173. https://doi.org/10.1016/j.bios.2021.113173
- Park, S., & Lee, J. W. (2021). Detection of Coronaviruses Using RNA Toehold Switch Sensors. International Journal of Molecular Sciences, 22(4), 1772. https://doi.org/10.3390/ijms22041772
- The Chinese Hong Kong University. (2017). Modeling: Designing Toehold Switch. Available at: https://2017.igem.org/Team:Hong_Kong-CUHK/Model