To demonstrate that a single nucleotide mutation in the start codon could result in ON to OFF change in EGFP expression in cells, we cloned both the EGFP-T2A-mCherry-WT (BBa_K2807012) and the EGFP-T2A-mCherry-ACG mutant (BBa_K2807013) into C1 mammalian expression vector for expression in HEK293T cells. We transfected HEK293T cells with both wild type (WT) reporter and ACG mutated (ACG) reporter and thereafter, imaged them via microscopy and harvested them for flow cytometry analysis.

As shown in Figure 1, the cells transfected with EGFP-T2A-mCherry-WT reporter expressed both GFP and mCherry proteins in the same cells, showing that T2A self cleavage peptide is effective. We then further quantified the level of fluorescence by flow cytometry and our results showed that at least 36.6% of the transfected cells are both EGFP+/mCherry+ (Figure 3). There are about 7.9% of the transfected cells with EGFP signal only. This is expected because it is possible for the ribosome to fall off from the mRNA when it encounters the T2A signal peptide and cannot continue with translation. As a result, only EGFP protein is produced but not mCherry.

On the other hand, HEK293T cells transfected with EGFP-T2A-mCherry-ACG mutant showed red fluorescence expression, with very low or undetectable green fluorescent protein levels as visualised via microscopy (Figure 2). Our experiment showed that the mutation of the start codon in EGFP is indeed effective in abolishing the expression of EGFP without affecting the expression of the mCherry protein. Compared to the cells transfected with WT reporter, the percentage of EGFP+/mCherry+ positive cells in cells transfected with the EGFP-T2A-mCherry-ACG mutant is almost undetectable (Figure 3). In addition, the mean fluorescence intensity for mCherry in cells transfected with EGFP-T2A-mCherry-WT and cells transfected with EGFP-T2A-mCherry-ACG mutant are similar, while there is significant reduction in EGFP expression between the two groups of cells.

An important feature of a reliable reporter system is to have an internal control signal such that the reporter signal can be normalized to account for stochastic processes in cells. Expression of both EGFP and mCherry reporter on separate plasmids may lead to uneven gene expression due to stochastic processes. Thus, it is more advantageous to have both EGFP and mCherry on the same plasmid. Therefore, we designed our reporter system with this intention in mind. We then carried out experiments to compare our reporter system with the reporter system described by the Worcester 2016 team. We evaluated the system in terms of background noise, as well as the reliability of the internal control.

We performed double transfection of the mCherry plasmid with either ATG EGFP (BBa_K2083009) or ACG EGFP (BBa_K2083010) plasmid in HEK293T cells. As shown in Figure 4, WPI reporters showed the expected OFF to ON change from ACG mutant to ATG. However, the fluorescence intensity of EGFP and mCherry does not correlate well with one another, and there were some leaky expression of EGFP even in the mutant form (Figure 5A). Moreover, 2.3% of the cells are double positive for EGFP in cells expressing the WPI ACG mutant reporter (Figure 5B & 5C), which is higher than our ACG bicistronic reporter construct (Figure 3B).

On the other hand, our bicistronic reporter construct showed a strong linear correlation between EGFP and mCherry fluorescence intensity in WT reporter (Figure 2A), and the number of cells positive for EGFP in our ACG mutant construct is effectively non-detectable. Therefore, our dual reporter system provides higher signal to noise ratio and allows for the quantification of relative editing efficiency between different cells and in different transfection experiments.

Although the EGFP-T2A-mCherry Reporter was initially designed for RNA editing, we reasoned that this construct could also be used to report Cas9 editing efficiency in DNA strands. Base editor 3 and 4 (BE3, BE4), fused with the nuclease-deficient dCas9n to the enzyme APOBEC can deaminate C to U in DNA templates, which restores the DNA sequence to T after replication (Komor et al., 2016, Komor et al., 2017).

As such, we tested out our reporter system using BE3 and BE4 plasmids. HEK293T cells were transfected with base editor, ACG reporter and gRNA concurrently and the number of DNA editing events was quantified using flow cytometry. We observed that among the mCherry expressing cells, 60% of them have EGFP fluorescence as a result of base correction from ACG to ATG (Figure 6,7). The EGFP fluorescence intensity was also restored to a level comparable to the wild type reporter (Table 1), confirming the occurrence of editing events. On the other hand, transfecting cells with only the base editors and ACG reporter showed no EGFP fluorescence, indicating that the editing event is specific and directed by the gRNA sequence.

In conclusion, our EGFP-T2A-mCherry reporter system is able to achieve a clear OFF-to-ON switch upon base editing from C to U. It improved the part from a previous iGEM team’s project by adding mCherry (as transfection control) in the same reporter plasmid. In cells expressing our constructs, the expression efficiency, EGFP and mCherry expression showed strong correlation with transfection rates. In addition, our reporter has been shown to work for Cas9-mediated DNA editing, thus providing an alternative tool for real time, in vitro visualization and reporting of base editing efficiency.