The main goal of our project is to use degradation-tuning RNAs to enhance protein concentrations without significantly affecting bacterial growth. First, we have to validate whether our assumption of changing mRNA degradation rate to control gene expression is correct, using in-silico approaches, say kinetic models. Then we choose several published sequences and test them in E.coli DH5α strains for experimental implementation. After detailed data analysis and learning, we re-designed a well-performed dtRNA using forward-engineering approaches, predicted its secondary structures using computational tools, and test it in our lab.
Design
At the design stage, we established a kinetic model of gene expression and bacteria growth to test the viability of our basic idea. Since we want to develop a dtRNA-based method for enhancing protein concentrations that surpasses conventional means such as adopting stronger promoters, the major goal of our model should somehow verify the effect of changing the mRNA degradation rate constant to bacterial growth and protein concentration, and compare it with changing the promoter strength. Therefore, we established a kinetic model that integrates gene expression-degradation and negatively feedbacked cell growth. We search the parameters of different parts from the parts registry and literature, substitute them into the model, and the results show that reducing the degradation rate has a negligible impact on cell growth compared to enhancing promoter strength if assuming the same gene expression levels.
Figure.1 Scheme for gene expression and degradation
Figure.2 effects of changing mRNA degradation rates on cell growth and protein concentration
To test the effectiveness of dtRNAs in DH5α strains, we choose six RNA structures in the literature that show the capability of manipulating RNA degradation in DH10B. Integrating dtRNA structures in GFP-expressing cassettes is well-suited for measurements. The fluorescence curve could indicate protein concentration, while OD600 indicates bacterial growth.
Build
At the build and test stage, we use the HiFi assembly cloning method to integrate dtRNAs in GFP-expressing cassettes. We have encountered a lot of problems in this section, and the most serious one is the unsuccessful transformation. After carefully considering the experimental protocol, we found several factors that would strongly affect the assembly process:
1. We use overlapping primers that contain dtRNA sequences and homology arms for HiFi assembly, so an annealing process is critical for the formation of duplexes as integration segments. At first, we use 1xTAE buffer for annealing, in which EDTA could become an inhibitory component for polymerases when mixing it with the HiFi mix. To avoid this situation, we changed the annealing buffer to NEBuffer 3 which is optimized for enzyme digestion and compatible with HiFi assembly systems.
2. We changed the overlapping primers to their reverse-complementary sequences so that the 3’ end of integration fragments would bond to the vector, facilitating elongation and integration, therefore the yield of HiFi assembly.
3. The initial ice-making conditions are poor in our lab so we are unable to get well-crushed ice for transformation experiments, which may affect the activity of competent cells. So we purchased a new ice-making machine and solve this problem.
Test
After making the aforementioned adjustments, we successfully transferred the plasmids into DH5α E.coli strains, then the transformants are sequencing-verified, and then collect the fluorescence and OD curve. Compared with the previous report from the literature, we observed comparable fluorescent signals, indicating the effect of degradation-tuning RNAs is indeed transferrable to DH5α strains, the only difference is that the introduction of dtRNA82 did not abolish the GFP fluorescence, which may come from the difference of Rnase targeting sites between the two strains.
Figure.3 GFP fluorescent curve under the control of J23106
Figure.4 Comparasion of dtRNA fluorescent fold change with different host strains
On the other hand, the measured OD600 curve shows a similar but even faster growth curve when using dtRNAs, no matter whether its function is to resist or facilitate degradation. This interesting observation deserves further investigation in the future. These experiment results solidified dtRNAs’ ability to change expressions and confirm our simulation results of not posting burden to the host.
Figure.5 Bacterial growth curve under the control of J23106
Learn
After the aforementioned experiment results, we firmly demonstrate the effect of dtRNAs. Still, we wonder if there is a limit to resistance to the degradation of dtRNAs. Since the dtRNA structures published in the article are only generated randomly in the first generation, there is space for rationally designing new dtRNA structures. Therefore, we wish to design a forward-engineered dtRNA according to the relationship between structure and function to further expand its dynamic range in the next cycle.
Next Cycle
In the next cycle, we firstly designed a second-generation dtRNA(dtRNA1v2) with optimal structural parameters based on the structure of dtRNA1 and used NUPACK to simulate the secondary structure.
Figure.6 Simulated secondary structure of dtRNA1v2
The structure parameters of dtRNA1v2:
1. stem length: 11 bp; structural factor: 100%
2. stem GC content: 54.5%; structural factor: 91%
3. loop size: 6 nt; structural factor:100%
Then we experimentally integrate dtRNA1v2 into GFP-expressing cassettes and measure the fluorescence and OD value, same as the first cycle.
Figure.7 Fluorescence curve of dtRNA1v2
The result shows a slightly higher fluorescence value compare to the original version of dtRNA1, with the GFP fold change up to 3.96. Although this value may not be statistically significant when adding repeating groups, it shows the potential of designing second-generation dtRNAs using forward-engineering principles.
All in all, we think we have demonstrated our engineering success by using in silico simulation to guide our design and experiments and applying rational design principles to further optimize our basic parts.