Engineering
DBTL1. Concept Proof
DBTL2. Editing System Optimization
DBTL3. Strain Security
DBTL4. Splice Variants Detection
The core of our iGEM project is to build a highly efficient RNA editing system. Only when we achieve high and stable editing efficiency can we proceed with a series of subsequent applications. Therefore, we first constructed a relatively simple system to detect whether ADAR exhibits editing activity.
Through discussion, we have decided to optimize the editing system by focusing on both the ADAR enzyme itself and the design of the sensor. For ADAR, we will consider factors such as the type of ADAR, its concentration, and subcellular localization. For the sensor, we will look into the specific sequence that interacts with ADAR, all of which will be discussed in detail later. Additionally, we hope to 'quantify' the editing efficiency, allowing us to design a reporter system that directly reflects the levels of transcripts in the cell.
Strain security has always been a concern in the industrial biotechnology field. In our previous proof of concept, we were already able to flexibly regulate the expression levels of downstream genes by designing a reporter system. Therefore, we aim to design a specialized genetic circuit that can serve as a safeguard against strain theft.
Beyond the field of biomanufacturing, we have also turned our attention to human health. Most genes in the human body have splicing isoforms, and splicing errors can lead to various diseases, including cancer. Naturally, we wondered if these splicing isoforms, with their different 'junctions,' could be detected rapidly and efficiently using our system. Based on this idea, we have carried out a series of designs and experimental validations.
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 1-1: Transform yeast cells with ADAR
Determining the Types of ADAR: The team responsible for wet lab experiments conducted a thorough literature review and identified two types of ADAR for further study: one full-length (ADAR1-p150) and one truncated with the fusion of MCP domain that can specifically bind to the dsRNA sequence (MCP-ADAR2), both exhibiting enzymatic activity according to the paper.
Selecting the Method for Yeast Transformation: Typically, there are two primary methods for introducing foreign genes into yeast:
1.Plasmid Transformation: A simple and quick method with high transformation efficiency, making it suitable for short-term experiments. However, plasmids do not integrate into the genome and can be lost without selective pressure, leading to unstable gene expression.
2.Homologous Recombination: Allows for stable integration of foreign genes into the yeast genome, enabling precise study of gene function. However, the process is complex, time-consuming, and requires advanced molecular biology skills, making it less suitable for short-term experiments.
To find a balance between stability and convenience, we have decided to use and compare both methods.
We ordered the genes for ADAR1-p150 and MCP-ADAR2 from the company. After amplifying them by PCR, we performed Gibson assembly with the plasmid backbone mentioned in Cycle 1 and then transformed them into E. coli. Subsequently, we sent them for sequencing to verify whether the trnasformation was successful.
After successfully transforming E. coli with the two ADAR genes, we extracted the plasmids containing ADAR from the E.coli. We then performed yeast electroporation and colony PCR verification.
We ordered the genes for ADAR1-p150 and MCP-ADAR2 from the company. After amplifying them by PCR, we performed Gibson assembly with the plasmid backbone mentioned in Cycle 1 and then transformed them into E. coli. Subsequently, we sent them for sequencing to verify whether the trnasformation was successful.
After successfully transforming E. coli with the two ADAR genes, we extracted the plasmids containing ADAR from the E.coli. We then performed yeast electroporation and colony PCR verification.
Figure 1. The qPCR Results of ADAR1/2
Figure 2. The Western Blotting Results of ADAR1
In response to the results observed in the test, we conducted a literature review and discovered that ADAR2 is toxic to yeast cells. The higher copy number of ADAR introduced via plasmid exacerbates this inhibitory effect on yeast growth. Consequently, we will proceed with further experiments using ADAR1 and ADAR2 introduced through homologous recombination.
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 1-2: Transform Yeast Cells With The Sensor and gRNA (1)
Design of the Sensor: The sensor comprises an upstream green fluorescent protein (GFP) and a downstream blue fluorescent protein (BFP), separated by a stop codon (which serves as the ADAR editing site). Additionally, we incorporated a P2A self-cleaving peptide between the two fluorescent proteins to ensure that, after translation, they are cleaved and do not interfere with each other.
Design of the gRNA: For the initial proof-of-concept, we opted for a relatively simple design and did not include common ADAR binding sites such as MS2/BoxB on the gRNA. Instead, we designed a sequence complementary to the sensor, with an A-C mismatch at the stop codon to serve as the ADAR recognition and editing site.
Choice of Plasmid Backbone: To avoid imposing an additional burden on the yeast by integrating into the genome, we plan to introduce the sensor and gRNA in plasmid form. We will use AmpR and BleoR as antibiotic resistance genes for E. coli and yeast, respectively.
We amplified eBFP, eGFP, and the plasmid backbone from our lab's plasmid library using PCR, then ligated them into the plasmid through three-fragment Gibson Assembly. The constructs were then transformed into E. coli for amplification, resulting in plasmids without gRNA.
Next, we used site-directed mutagenesis primers to amplify the gRNA gene that pairs with the sensor. We then used two-fragment Gibson Assembly to construct the plasmids, which were subsequently transformed into E. coli, yielding plasmids containing both gRNA and the sensor.
After successfully transforming E. coli with the sensor and gRNA, we extracted the plasmids from the E. coli. We then performed yeast electroporation and colony PCR verification.
Although the plasmids containing both the sensor and gRNA were successfully constructed and the transformation into E. coli was successful, the yeast can not be screened on the plate containing bleomycin, because wild type yeast can grow on the plate.
We hypothesize that the experiment may have failed due to the following two reasons:
1.The bleomycin in the lab may have degraded, resulting in non-toxic to the yeast.
2.There may have been an error in the plasmid library registration.
Therefore, we plan to change the selection gene for yeast in future experiments and repeat the experiment.
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 1-3: Transform Yeast Cells With The Sensor and gRNA (2)
Due to issues with bleomycin, we switched to using a uracil auxotrophic strain for selection. The yeast strain was changed to a uracil synthesis-deficient type, while all other conditions remained the same.
We amplified the uracil synthesis gene from a plasmid, then used Gibson Assembly to integrate it with the original fragment backbone. Subsequent steps followed the same protocol as in Cycle 1-2.
We performed colony PCR on the newly transformed yeast to verify the presence of the sensor. The band sizes matched the expected results.
From the Test results, we predict that the sensor has been successfully introduced into the yeast. Therefore, we can proceed with experiments to assess ADAR enzyme activity.
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 1-4: Examing Editing Activity of ADAR
We have engineered a construct featuring two linked fluorescent proteins in Cycle 1-2 and 1-3: an upstream green fluorescent protein (GFP) and a downstream blue fluorescent protein (BFP), separated by a stop codon. The hypothesis is that a high editing efficiency of ADAR will result in a higher probability of the stop codon being replaced by a sense codon, thereby allowing the ribosome to continue translation and produce increased blue fluorescence intensity. To evaluate the editing efficiency of ADAR, we employed confocal fluorescence microscopy and flow cytometry and established the following four experimental groups:
1.With ADAR and A-C mismatch: This group should exhibit downstream blue fluorescence, as it represents the standard condition for ADAR editing.
2.With ADAR, without gRNA: This group should not exhibit downstream blue fluorescence because there is no double-stranded RNA substrate for ADAR to recognize.
3.With ADAR and gRNA, without A-C mismatch: This group should not exhibit downstream blue fluorescence because ADAR cannot edit non-mismatched double-stranded RNA.
4.Without stop codon: This group serves as a positive control to confirm that the downstream BFP gene can be normally expressed.
The students responsible for the wet lab experiments were divided into two groups, each focusing on learning and testing confocal fluorescence microscopy and flow cytometry, respectively.
Confocal Fluorescence Microscopy: The results indicated that when both ADAR and gRNA were present, there was weak downstream blue fluorescence in the dark field, suggesting that ADAR1/2 exhibits a low level of editing activity.
Figure 3. The Results of Confocal Fluorescence Microscopy
Flow Cytometry: The results appeared contradictory to those from confocal fluorescence microscopy. The group with ADAR/gRNA did not show significant differences compared to the two negative control groups.
Figure 4. The Results of Flow Cytometry
We hypothesize that the discrepancies in flow cytometry results might be due to leakage expression in the two negative control groups and the low editing efficiency of ADAR, resulting in a lack of significant differences. Therefore, further optimization of ADAR and gRNA design is necessary to enhance editing efficiency.
Considering that the fluorescence intensity of BFP is inherently low, it may not be detectable if it falls below the detection threshold. Hence, in future experiments, it would be beneficial to replace BFP with a fluorescent protein of higher intensity to facilitate detection.
The results from the Concept Proof section indicate that ADAR does exhibit some enzymatic activity, and the enzymatic activity of ADAR2 is higher than that of ADAR1. However, the system we designed might be too simple, leading to very low observed enzyme activity. Therefore, further improvements to the system are necessary to enhance the editing efficiency of ADAR.
Cycle 1-1
Cycle 1-2
Cycle 1-3
Cycle 1-4
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 2-1: Increase Transcription Level of ADAR
We aim to enhance ADAR expression by changing the promoter, as higher ADAR concentrations might lead to better editing outcomes. After reviewing the literature, we selected the TEF1 promoter, which is the most active in yeast, to replace the previous PDC-1 promoter. Apart from the promoter change, the rest of the ADAR plasmid remains the same as in previous experiments.
We amplified the TEF1 promoter and the part of the plasmid excluding the PDC1 promoter using PCR. After Gibson assembly, we transformed the construct into E. coli and sent it for sequencing. Upon successful sequencing, we extracted the plasmid from E. coli and performed yeast transformation.
Colony PCR: After yeast transformation, we performed colony PCR for validation, the result showed that ADAR1/2 were successfully transformed to yeast.
qPCR: We extracted total RNA from yeast transformed with either PDC1-ADAR and TEF1-ADAR, then measured ADAR gene transcription levels using qPCR.
The validation results indicate that both ADAR1 and ADAR2 were successfully introduced into yeast. Moreover, using the TEF1 promoter significantly increased the ADAR gene transcription levels compared to the original PDC-1 promoter (showing significant differences).
Figure 5. The comparison of transcription level between TEF1-ADAR and PDC1-ADAR
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 2-2: Add MS2 sequences on ogRNA
Through literature review, we have found that the target transcript and ogRNA do not simply complement each other. There exists one to several “stem-loop structures” on the ogRNA, among which MS2 is one type. Since ADAR itself has MS2 recognition sites, adding the MS2 recognition sequence to ogRNA may help the binding of ADAR to the RNA heteroduplex, thereby further improving the editing efficiency of ADAR.
Therefore, we designed ogRNA sequences containing 0, 2, and 4 MS2 sequences, respectively, and intend to compare the impact of different numbers of MS2 on the editing efficiency of ADAR.
We ordered ogRNA containing 0, 2, and 4 MS2 sequences and the corresponding target transcripts from the company. The method for constructing the plasmids is similar to that previously described and will not be elaborated here. Subsequently, these three plasmids were introduced into yeast expressing PDC/TEF-ADAR1 and PDC/TEF-ADAR2, respectively, for further test.
We conducted yeast colony PCR validation, and the results indicated that the sensors were successfully introduced into both types of yeast.
In addition, we also performed confocal fluorescence microscopy imaging and sorted the cells using flow cytometry.
Figure 6. FACs Results of Editing Systems Containing Different MS2 sequences
Figure 7. Confocal Results of Editing Systems Containing Different MS2 sequences
Figure 8. FACs Results of Editing Systems Containing Different ADAR Types and Promoters
Figure 9. Confocal Results of Editing Systems Containing Different ADAR Types and Promoters
We could clearly see from Figure 6 and Figure 7 that adding MS2 sequences effectively enhances the editing efficiency of the whole system. However, it is strange that increasing the number of MS2 sequences leads to a lower editing efficiency for TEF-ADAR1 and TEF-ADAR2, which is not consistent with the results observed for PDC-ADAR2. We hypothesize that MS2 sequences might obstruct the transcription process, potentially leading to premature termination of the mScarlet-EGFP fusion protein. When the promoter of ADAR1/2 is PDC, the binding affinity plays a role: more MS2 sequences result in stronger binding, ultimately leading to higher editing efficiency. However, when the promoter is changed to TEF, ADAR1/2 may become saturated due to its high expression level. In this case, the concentration of sensor RNA becomes crucial. Considering that MS2 sequences might cause a high frequency of abortive transcription, the concentration of sensor RNA could be lower if more MS2 sequences are added, leading to lower editing efficiency.
Furthermore, TEF-ADAR2 clearly exhibits the highest editing efficiency, as indicated by Figure 8 and Figure 9. This is partly due to its stronger expression level, which was identified in Cycle 1.
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 2-3: Measuring Editing Efficiency Quantitatively
In order to quantitatively characterize the editing efficiency of ADAR, we need to measure the relationship between transcription level of the target transcript and its editing efficiency. Therefore, we changed the promoter of target transcript to a xylose-induced promoter, which could be used to adjust the transcription level by adding different amounts of xylose.
We constructed two plasmids named xyl-Sensor-C1 and xyl-Sensor-C2. After completing the plasmid construction, we amplified the C1 and C2 fragments by PCR, and then introduced them into yeast expressing TEF-ADAR1 and TEF-ADAR2, respectively. The gene were integrated to the yeast genome by homologuous recombination.
We performed colony PCR verification on the transformed yeast, and the results indicated that both fragments were successfully introduced into the yeast with the correct orientation.
We plotted a standard curve of target transcript concentration (by qPCR) versus editing efficiency (by FACs) by adding different concentrations of xylose.
Figure 10. The Relationship Between Relative Transcription Level and Editing Efficiency
From the test results, there seemed no relationship between target transcript concentration and editing efficiency, that was maybe because induction time was insufficient. So we decided to extend the induction time from 12 hours to 24 hours in the future experiment.
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 2-4: Transform yeast cells with ADAR Liuzruiuiuei cbsdu hfsi disbsdb vks
We noticed that the editing efficiency of ADAR2 is much higher than ADAR1 in Cycle 3, so we speculated that subcellular localization of the ADAR may affect its editing efficiency. If ADAR is localized in the nucleus rather than the cytoplasm, it would be difficult to edit ogRNA, which is located in the cytoplasm. Therefore, we fused ADAR with mScarlet to construct the detection system, which were linked by a flexible linker to prevent protein misfolding.
We constructed two plasmids: pLocate-ADAR1 and pLocate-ADAR2, which was co-transferred to yeast with pUC19-C1 by homologous recombination.
Figure 11. Plasmid Maps of pLocate-ADAR1/2 and pUC19-C1
We observed the fluorescence by LSCM (Laser Scanning Confocal Microscope), the result is shown below:
Figure 12. Confocal Result of ADAR1 and ADAR 2
For the ADAR1, one or more red bright spots could be observed in some cells, which are likely to be the locations of the yeast nuclei. Additionally, a portion of the fluorescence is dispersed throughout the entire cell.
For the ADAR2, it was clear that there is a region in the middle of the cell that lacks fluorescence, whereas the rest of the cell contained red fluorescence.
Most ADAR1 located in the yeast nucleus, with a small portion in the cytoplasm. This accounts for the lower editing efficiency of ADAR1 in the pre-experiment. ADAR2, on the other hand, is primarily located in the cytoplasm rather than nucleus. This might be related to the NES (nuclear export signal) sequence in ADAR2. Look back on the experiments in Cycle 1-3, ADAR2 also had high expression level and higher editing efficiency based on FACs and confocal results, we finally decided to are use ADAR2 for further application.
The research on optimizing ADAR has been successful. By using a stronger promoter to increase ADAR expression and adding more MS2 sequences that bind to ADAR, the editing efficiency has significantly improved. We also observed that ADAR2 has a higher editing efficiency than ADAR1, which can be explained by the higher expression level of ADAR2 and the fact that some ADAR1 is localized in the nucleus, where it cannot function. Therefore, we plan to use yeast with TEF-ADAR2 as the chassis for subsequent applications. However, the experiments to quantify or semi-quantify transcript levels using the reporter system were not very successful, likely due to insufficient induction time, which will require further improvement.
Cycle 2-1
Cycle 2-2
Cycle 2-3
Cycle 2-4
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 3-1: Gene Screening
To achieve the anti-theft effect, we naturally thought of designing a downstream gene that is expressed under specific conditions. This could be a suicide gene, and it would ideally respond very sensitively to certain conditions. We plan to conduct research in this direction.
Through searching papers on the Internet, we found that it is not easy to find genes that satisfy both of the above criteria simultaneously. Therefore, we searched for the two types of genes separately. Ultimately, we identified two genes that respond to external stimuli: HSP26 and GLC3, and two suicide genes: GSDMD and BAX.
We further searched the literature to confirm the accuracy of our findings. The literature shows that HSP26 is expressed at 160 times the level at 40°C compared to 30°C, while GLC3 is highly expressed in the absence of external glucose. Additionally, there is literature indicating that GSDMD and BAX promote pyroptosis and apoptosis, respectively. These genes are described in detail in the Design section with attached references.
Click here to know more about our Design.
Since the above four genes are the most responsive stress-related and suicide genes we found, we plan to design our system based on these four genes.
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 3-2: Plasmid Construction of Strain Security System
Connecting stress response genes with suicide genes is not a straightforward task. Initially, we had several simple ideas, such as placing the suicide gene directly downstream of the stress response gene. However, these were eventually discarded because they did not require ADAR's involvement and the regulatory process was not sensitive enough, leading to issues with leakage expression.
Ultimately, we solved this problem by introducing Cre recombinase and loxP sequences. We incorporated the Cre recombinase gene downstream of the sensor and designed loxP sequences on either side of the suicide gene. Under specific conditions, the concentration of stress response gene transcipts is high, which can form duplex with sensor. thus, ADAR is able to edits the stop codon of the ogRNA sequence in the sensor, allowing for the expression of Cre recombinase. The Cre recombinase can knockout the suicide gene, preventing the death of the strain. When conditions change, Cre recombinase is not expressed in large quantities, so the suicide gene is continuously expressed and accumulates, eventually leading to the death of the strain. This method avoids the drawbacks of persistent leakage expression of the suicide gene, which could otherwise damage the cells.
We plan to use a method similar to the one used previously to construct the plasmids, and we hope to ultimately introduce them into yeast.
During the plasmid construction process, we found that the gene amplification products consistently contained a large amount of nonspecific bands, which led to significant side reactions in the Gibson assembly. As of September 15, we have not yet constructed all the required plasmids.
Upon inspection, we found that the the Gibson assembly reaction always failed. That was maybe because mutli-fragments Gibson assambly was not easy to perform. As a condequence, we needed to change our scheme of plasmid construction. For example, we can use four-fragments Gibson assembly followed by a three-fragments one instead of performing six-fragments Gibson assembly at one time, which might improve the chance of success.
Unfortunately, due to operational errors and challenging protocols, this part of the work did not achieve a high level of completion. However, we believe that the design itself is fundamentally sound and that strain protection has significant application value. Therefore, we still plan to continue advancing this project and explore further validation processes.
Cycle 3-1
Cycle 3-2
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 4-1: Gene Screening
To find out the most suitable candidate for the research of splice variants, we combined several factors into consideration and chose three possible genes：
1. The length: We hope that the length of genes can be shorter than 6000bp. If possible, The shorter the gene, the better. This criterion is set because plasmids in yeast are generally less than 10,000 base pairs (bp), and including other auxiliary genes, we hope to keep the entire plasmid size below 6,000-7,000 bp. Plasmids that are too large not only pose difficulties in manipulation but may also lead to a decrease in stability.
2. The sequence: The sequence of genes must include CCA. This criterion is set because our RNA detecting system is based on ADAR which detects A-C mismatched base pair of CCA.
3. The relationship between the gene and cancer: The splicing isoforms of this gene should be closely related to the occurrence and progression of cancer. The expression and abundance of different splicing isoforms can affect the proliferation, metabolism, and structural changes of cancer cells. This criterion is set because we hope that the detection of the content of splicing isoforms has certain academic value. In particular, we aim to use our detection system to inspire the academic community.
4. The mechanism of splicing isomerism: We are looking for a set of splicing isoforms with the characteristic that different isoforms can be identified at the junction of a particular segment. For example, this could be due to mechanisms such as exon exclusion or the inclusion of specific exons leading to alternative splicing. This criterion is set because our detection system is best suited to utilize the sequence differences at the segment junctions to identify different splicing isoforms, and thereby analyze their presence and relative abundance.
We finally chose three candidates: RPS6KB1, PK-M and Chk1.
Considering applying our RNAssay system into monitoring these splice variants, we tried to design different sensors to sense different splice variants and we found Chk1 most suitable for our system, due to the following reasons:
1. Suitable length: 4128bp with CCA sequence
2. The ratio of Chk1 and its splice variant Chk1-s are signals for a series of cancer, and Chk1 are cell-cycle relavant protein so it means a lot to understand the dynamic change of proteins expression.
After the whole process of gene selection, we learned thatAfter the whole process of gene selection, we learned that
1. Many splice variants are key signals of disease like cancer.
2. Not all genes are suitable for our system.
3. RNAssay due possess the potential for *in vivo* monitoring of splice variants but need more improvement to expand university
Back to Last Cycle
Design
Build
Learn
Test
Guide to Next Cycle
Cycle 4-2: Detecting System Construction
Design of Sensor: Using the same fluorescence protein as the optimization experiments built before, we design a mScarlet in the upstream and eGFP in the downstream of our sensor, and put the gRNA in the middle. We used the program from the research article we based on to design our gRNA. gRNA for Sensor Chk1 mainly bind to the sequence which Chk1s don't have and gRNA for Sensor Chk1s mainly bind to the connection part of the upstream and downstream area of its lacking sequence compared to Chk1.
Design of target gene: In our cycle 1, we choose Chk1 and Chk1s as the targets we want to detect. We buy these two genes in their cDNA sequence from a company. We designed to construct target gene and the sensor into one plasmid. There are four plasmids we need to build:
1. Sensor-target_Chk1-Chk1
2. Sensor-target_Chk1s-Chk1
3. Sensor-target_Chk1-Chk1s
4. Sensor-target_Chk1s-Chk1s
(Sensor-target_A-B means a plasmid contain Sensor A and target B together, Sensor A means the Sensor we designed to bind with A only)
Apart from the Sensor and target change, the rest of the plasmid remains the same as the pSensor plasmids in previous experiments.
strategies to construct our plasmids: First we decided to do a six fragments Gibson to build two of our plasmids and based on these two plasmids we will build another two with two fragments Gibson. However, we failed to build for at least three times. The results were weird as we did colony PCR successfully for the first time but next day we do the same experiments again but couldn't see the right bands.
We analyzed the possible reasons, and do some improvement.
1. There are two short fragments have the length about 200bp and it might be too short for Gibson, so we do overlap extending PCR to connect them together as one fragment.
2. By analyzing our colony PCR results we found a interface was hard to connect. So we change the design of primer for this interface simple by extend the gibson overlap or move the overlap right or left for several base pair.
Finally we did two overlap extending PCR and two four fragments Gibson, two two fragments Gibson to construct all four of our plasmids. After Gibson assembly, we transformed the construct into E. coli and sent it for sequencing. Upon successful sequencing, we extracted the plasmid from E. coli and performed yeast transformation.
Colony PCR: After yeast transformation, we performed colony PCR for validation, the result showed that each Sensor-target plasmid was successfully transformed to yeast.
LSCM (laser scanning confocal microscope): To evaluate the success of the system we designed, we assessed the fluorescence intensity of each group. Initially, we utilized LSCM for detection. By measuring the fluorescence intensity and calculating the ratio of eGFP to mScarlet, we can determine the editing efficiency of the ADARs.
Figure 13. Confocal Results of Sensor-target_Chk1(s)-Chk1(s)
Flow cytometry analysis: We got nice datas from Flow cytometry analysis, though the results were not exactly as what we expected. The gates we had drawn to distinguish different cell signals are shown in Fig.14. Then we processed the data into bar chart following the formula:
Edit Efficiency = $\bf{\it{\frac{\text{Contain Both mScarlet and EGFP}}{\text{Contain mScarlet}}}}$
Figure 14. FACs Results of Sensor-target_Chk1(s)-Chk1(s)
IntaRNA testing: Though the results were promising, they were out of our expectations. So we put the sequences into IntaRNA to test if there will be an evidence to explain.
Figure 15. IntaRNA Simulation of Binding Energy of Chk1(s)-Chk1(s)
TA2 has the higher edit efficiency at about 80-95%, but this causes no evidence of selectivity. So we couldn't evaluate whether our sensor function well or not according to the editing effect cause by TA2.
The extremely high efficiency of TA2 in yeast shows the potential of becoming a powerful gene edit tool in yeast.
TA1 has the edit efficiency >15% according to the datas from our Optimization group, which is significantly higher than the stop codon readthrough or off target effect (edit efficiency is around 5%). So we could infer that TA1 has a strong selective edit effect for Sensor-Chk1 with Target-Chk1. But TA1 only has the edit effect on Sensor-target_Chk1-Chk1, that was beyond our expectations.
Design of Sensor is important for the selectivity of ADAR. After the experiment, we put the sequences into IntaRNA and predict the binding energy and binding area of sensor and its corresponding target. According to the results differences, we inferred that the sensor might not have more than one potential binding area and the base around the A-C mismatch might be strictly complementary pairing (not pairing like CAU-AUC).
There are several problems we need to address, which can be focused in futured experiments.
1. Whether PDC1-ADAR1 and PDC1-ADAR2 will work as the same as TEF-ADAR1.
2. Why group TA1-SS do not have the same results as TA1-KK as we expected.
3. Why TEF-ADAR2 has no selectivity for all four plasmids.
4. Whether the CAU-AUC sequence cause the decreasing edit efficiency of ADAR.
Due to the time limited, further experiments will be designed later.
Overall, this part has initially validated the potential for detecting splicing isoforms. However, due to certain design flaws in the sensor, some of the control group results did not meet expectations. Therefore, we plan to redesign the sensor and conduct subsequent validations.
Cycle 4-1
Cycle 4-2