Overview
Welcome to the engineering success page! Here, we described how we developed and tested the pmrB-anti-mCherry receptor and mRNA switch . During this process of engineering, some of the experiments we planned did not work at the first attempt. Nevertheless, the drawbacks forced us to come up with new ideas and initiated new cycles of engineering, ultimately driving our project forward. On this page we summarized the underlying line of thinking that led to our final engineering success.
Anti-mCherry pmrB
We wanted to design a cheater-proof platform that consists of mutation-reporting workers and specifically-cytotoxic guards to improve the genetic stability. Once the workers transform into cheaters that don’t produce aimed protein, the mutation will be presented and then detected by guards, whose extermination system will therefore be activated and will secrete cytotoxin to selectively target cheaters. One of the most important steps is to recognize the killing signal from guards.
Design
There is a pmrBAC system in bacteria, which can be used as a bridge between extracellular signal and intracellular gene expression regulation. Originally, PmrB receptor on cell membrane uses an iron(III)-binding motif to sense the environmental ferric level[1].Based on the system, we wanted to replace the iron(III)-binding motif with an anti-mCherry motif, which can specifically bind mCherry protein and then facilitate PmrB to phosphorylate PmrA. The PmrC promoter will be activated by phosphorylated PmrA protein. Then it will initiate the expression of CcdB cytotoxin. In a word, mCherry acts as a killing signal to trigger the cheaters' death.
The anti-mCherry single-chain antibody binding region has six binding sites. We replaced the iron(III)-binding motif with a whole anti-mCherry single-chain antibody sequence, which contained 201bp in total. Besides, we removed some uncritical sites of anti-mCherry and added a connection sequence to make sure that the sequence can play a role. The following image is our preliminary design diagram.(Figure 1)
Figure 1 Preliminary design diagram
Test
1.Construct plasmids
In order to achieve verification, we constructed two plasmids, one was the original design with ccdB(plasmid-B), and the other was using GFP instead of ccdB(plasmid-G) so that we could observe the expression visually. All the fragments were connected to pET-Duet-1 plasmid by enzyme digestion and ligation and transformed into E. coli DH5α.
2.Validate plasmids
The products of colony PCR were analyzed by agarose gel electrophoresis and their sizes were verified to be correct. The plasmids were extracted and subjected to single enzyme digestion, and the results of electrophoresis were correct. We performed multiple double enzyme digestion and electrophoresis to verify that the sizes of the products were correct. Then the plasmids were validated by sequencing.
3.Conduct confirmatory experiments
We expanded the culture of bacteria containing plasmid-G(37℃,200rpm)for 12h,divided evenly into several parts. And we added mCherry solution to one of them and observed the fluorescence under the fluorescence microscope. However, we did not see our expected results under a fluorescence microscope.
Then we measured the fluorescence intensity of GFP produced by bacteria with different concentration of mCherry solution added by microplate reader. Not surprisingly, there wasno difference in GFP intensity at different concentrations of mCherry, and the absorbance values of the experimental groups were even similar to the control group.
Learn
The failure of the experiment led us to rethink our design. We suspected that the length of anti-mCherry sequence was so long that the extracellular polypeptide chain of the modified PmrB failed to form the correct structure, which led to the failure of the verification experiment. We checked the material of previous successful transformations and found that the length of their modified sequence was indeed much shorter than our modified sequence. That's why we finally decided to shorten the length of anti-mCherry sequence as much as possible while retaining its original functions.
Improve
The anti-mCherry single-chain antibodies binding region had six binding sites. Just as other antibody, anti-mCherry was also axisymmetric. The three binding sites at one side were symmetrical to the three binding sites at the other side. Therefore, we assumed that only three of the binding sites at one side were enough to function properly. [2]After learning from literature, we selected three binding sites from the same side of the antibody:CDR1, CDR2 and CDR3 (Figure 2) (Part:BBa_K3805238), which were only half of the single-chain antibody binding region to perform its function. We asked Professor Li whether our design scheme was feasible(Figure 3). He affirmed that our idea was theoretically feasible and worth a try.
Figure 2 Optimised design diagram
Figure 3 Our communication with professor Li
Test
Same as before, we conducted two confirmatory experiments after we constructed two plasmids and validated plasmids.
We examined our results with a fluorescence microscope. Images shot by fluorescence microscope before and after the addition of mcherry showed results that corresponded to our expectation(Figure 4,5).
Apparently, the higher the concentration of mCherry was, the stronger the fluorescence intensity was, which confirmed that the pathways we constructed properly performed the functions we expected(Figure 6).
Figure 6 Fluorescence intensity produced by cheaters with different concentrations of mCherry
Besides, we proved that the cheaters could accept the mCherry signal to commit suicide(Figure 7). Through comparison, our results confirmed that the killing efficiency of extermination pathway reached more than 90%.
Figure 7 Comparison of growth of cheaters with and without mCherry
No mCherry was added on the left; mCherry was added on the right
mRNA switch
Design
mRNA switch is a new part proposed by BNU-China. We designed a pathway with a certain sequence behind its promotor to act as an additional switch to adjust the expression of downstream GFP gene, which represents the extermination system in our project. This unique 12bp mRNA sites in Pathway 2 are complementary with continuous 12bp insulin proteinogen mRNA.[3][4]
Figure 8 Preliminary design diagram
Test
At the beginning, in order to test our Pathway 1+Pathway 2 and the expression of insulin proteinogen, we designed a series of bacteria internal（in vivo） experiments to verify the combination of certain 12bp sites normal or changed in insulin proteinogen with complementary sequences in Pathway 2. According to our assumption, it should not be fluorescent when the targeted protein is expressed normally because the specific 12bp sites in insulin proteinogen mRNA will bind to the complementary sequences in Pathway 2. However, if the targeted protein was incorrectly expressed (as in Mutation Ⅰ and Mutation Ⅱ), the cell should show fluorescent. But unfortunately, our results show that all cells could be fluorescence, which didn’t meet our expectations.
Figure 9 GFP expression (from the left to the right: normal, mutationⅠ, mutationⅡ)
Learn
We have analyzed the results, discussed the design of experiments, searched for information and asked experts for help. After these, we proposed two possible causes for the unexpected results. (1)The 12bp mRNA binding site is not enough for mRNAs’ interaction. (2)The efficancy of the switch is poorly guaranteed. So, we continued to design a series of experiments to test for our proposed ideas.
Figure 10 Our group meeting
Improve
Then, we designed an in vivo experiment to test whether mRNAs can serve as a switch to control the expression of downstream genes. In our experiment, we set three different concentrations for 12bp mRNA in bacterial culture, hoping that the bacteria could internalize the mRNA and indirectly reflect mRNA binding through the expression of GFP. Interestingly, the results showed that the fluorescence intensity of the bacteria with two pathways became lower and lower with the prolongation of incubation time, indicating that the bacteria could absorb short mRNA under normal conditions, and the interaction of mRNA could inhibit the expression level of GFP. 12 bp mRNA can be absorbed by generations of bacterial cells after cell divisions, which causes the ratio of the green fluorescent /cell density (indicated by OD600) to decline continuously over a period of time.
Figure 11 Optimised design diagram
Re-learn
However, our team members suggested that the reduced GFP expression in the in vivo experiment might also be due to absorption efficiency caused by strain propagation. After discussion, most of us think that this may exist.
Re-design
In addition to the in vivo experiment, we also designed a series of mRNA transformation experiments, which is a method that can directly allow mRNA to enter cells, and can record the GFP expression of a specific generation under the constraint of mRNA binding. We used a uniformed 12bp mRNA concentration gradient and the results showed that our mRNA switch is feasible.[5][6]
Figure 12 The result of mRNA transformation experiments
Further Exploration
After demonstrating the feasibility of our switch, we tried to further explore an optimized binding site length. We added a group of 24bp mRNA experiments to compare with the one in the previous experiments. The results showed that gene length, gene concentration and the time after administration and transformation may all have impact on the expression of downstream genes.
Figure 13 The result of further mRNA transformation experiments
Overall, we have successfully verified the feasibility of our mRNA switch and found several factors that can be adjusted to improve the efficiency of our switch.
References
[1] Liang H, Deng X, Bosscher M, Ji Q, Jensen MP, He C. Engineering bacterial two-component
system PmrA/PmrB to sense lanthanide ions. J Am Chem Soc. 2013 Feb 13;135(6):2037-9.
[2] Mehdi M , Akshay B , Laura G , et al. A Split-Luciferase Reporter Recognizing GFP and mCherry Tags to Facilitate Studies of Protein–Protein Interactions[J]. International Journal of Molecular Sciences, 2017, 18(12):2681-.
[3]Li JH, Liu S, Zhou H, Qu LH, Yang JH. starBase v2.0: decoding miRNA-ceRNA, miRNA-ncRNA and protein-RNA interaction networks from large-scale CLIP-Seq data. Nucleic Acids Res. 2014 Jan;42(Database issue):D92-7. doi: 10.1093/nar/gkt1248. Epub 2013 Dec 1. PMID: 24297251; PMCID: PMC3964941.
[4]Pushparaj PN, Aarthi JJ, Manikandan J, Kumar SD. siRNA, miRNA, and shRNA: in vivo applications. J Dent Res. 2008 Nov;87(11):992-1003. doi: 10.1177/154405910808701109. PMID: 18946005.
[5]Perez Ruiz de Garibay A. Endocytosis in gene therapy with non-viral vectors. Wien Med Wochenschr. 2016 May;166(7-8):227-35. English. doi: 10.1007/s10354-016-0450-5. Epub 2016 May 3. PMID: 27141862.
[6]Kumar P, Nagarajan A, Uchil PD. Electroporation. Cold Spring Harb Protoc. 2019 Jul 1;2019(7). doi: 10.1101/pdb.top096271. PMID: 31262965.
