Our system can be divided into three biochemical steps: 1) transcription of the target sequence into mRNA, 2) reverse transcription of the target sequence into double-stranded cDNA, 3) recombination for the cDNA to replace the target sequence. We set out to separately verify each step before integrating together. Our team logo shows the mutants accumulated by varying blue colors.

Transcription

We put the target sequence and its flanking elements together under a T7 stable promoter for high expression level of target RNA. To verify our R-Evolution system, we constructed 8 nonsense mutant of chloramphenicol resistance gene (Chl), bearing the 8 base pair substitution from sense codon to nonsense mutant. We verified this construct through culturing bacteria carrying the original version or mutant on plates containing chloramphenicol. We found that bacteria carrying the original Chl grow naturally, while no colony was formed on the plates of mChl (Table. 1). After adding the flanking sequences on both ends, we used the C158X mutant and showed that the gene’s function has not been changed by our conduct (Figure 1).

In addition, we also constructed and verified nonsense mutants of fluorescent protein EGFP and mCherry at the 158^th and 159^th amino acid (Figure 2).

Reverse transcription (RT)

The successful expression of RT

RT is expressed under an IPTG controlled promoter, we constructed a series of promoters by placing a LacO fragment under a stable promoter, and hopes to determine under which we could achieve most stringent control. The constructs we tested are: T5, T5-LacO, T7-LacO, LacUV5-LacO, J23119-LacO.

We initially attempted to verify RT’s expression through an EGFP fusion protein. We fused the EGFP to the C’ end of pol protein, linked by a GS tag. But this construct proved unsuccessful (data not shown), possibly due to the length and complexity of the gag-pol polyprotein. Then we turned to directly expressing EGFP in the place of RT under the control of IPTG. If EGFP can be successfully expressed, so should RT. And results are obtained for each induction promoter construct (figure would be supplemented in presentation and poster).

We found through careful examination that this failure is due to problems with our plasmid construct, so we moved the RT to another tested plasmid, and through SDS-PAGE, verified its successful expression (Figure 3). Gag-pol polyprotein is expressed as a whole with its stop codon mutated into its readthrough product, glutamine. The polyprotein has three functional parts, capsid protein, protease and reverse transcriptase. We made Y586F mutation on reverse transcriptase to increase its mutation rate. From the PAGE gel, we can see that all three bands could be seen when induced.

Recombination

Cre expression is controlled by Tet operon

We placed Cre under the promoter ptetR, whose expression is controlled by its inhibitor TetR. Regarding where we should place the inhibitor gene to maximize its expression, we opted between 2 options, one is placing it downstream of the LacI inhibitor, the other is to place it downstream of the kanamycin resistance gene (KanR). We tested both construct by placing an EGFP in these two places and measuring its fluorescence emission. Results show that the expression level is almost the same in both construct.

Cre initiates excision between two homologous loxP site

Placing 2 wild-type loxP on both ends of the target sequence (mCherry) in the same direction, and expressing it under a stable promoter (J23101). By co-transforming the target plasmid with another plasmid carrying Cre recombinase, we verified that our Cre protein functions accordingly by excising the mCherry sequence from the promoter (Figure 4).

Through PCR amplification with the primers annealing to sequences outside the target, and subsequent electrophoresis, we found that the band from bacteria co-transforming Cre corresponds to the excision of mCherry.

lox5171 is most incompatible with wildtype loxP (wtlox)

When we were carrying out integrated human practice, we were warned by Prof. Wang that two homologous loxP would be excised at a much higher efficiency than performing recombination as we wished, so we searched the literature and selected 3 mutants that are said to be incompatible with wtlox but are compatible with themselves, they are lox511, lox2272 and lox5171.

We tested their incompatibility with wtlox by replacing one of wtlox into the mutant at the ends of mCherry, and co-transformed the plasmid with Cre (Figure 5 & 6). The result we obtained showed that lox5171-mCherry-wtlox performs best, and used it in further analysis (Figure 6).

Cre with degradation tags

When our modeling demonstrated to us that the expression level of Cre needs to be much lower than that of RT, we introduced degradation tags. By attaching them to the C terminal of Cre recombinase, the protein would be rapidly recognized and degraded by the E. coli’s native SsrA-SmpB degradation system. This construct could also solve the problem of basal leakage and continued existence after inducer removal.

Apart from the native AANDENYALAA tag, we also modified its last three or five amino acids into YALAV, YALVA, YALVV and WVLAA. We tested the stable expression level, as well as the degradation dynamic of each tag by attaching them to the C terminal of EGFP protein and measuring the change in fluorescence level (Figure 7). The stable state expression increases as the number of mutated amino acids increase, or the mutated site nears the N’ of the tag. Supported by our modeling result, we deemed that the XXX tag performs best and chose to use it in further experiments.

Modeling

Our modelling successfully demonstrated that our system could function and mutation could accumulate along bacteria growth (Figure 8). For detailed explanation of our system, please visit our Modeling page.