ROS-inducible Catalase Results
Our investigation of the performance of the KatG, SoxS and TrxCP promoters revealed notable differences in their responses to varying concentrations of hydrogen peroxide (H2O2) (Figure 1). The native E. coli KatG promoter showed the lowest relative expression of RFP among the promoters tested and exhibited only minimal increases in expression as H2O2 levels rose. Due to its low baseline activity and weak inducibility, we decided that the KatG promoter is unsuitable for our ROS-inducible system, and is not the most promising candidate moving forwards.
Figure 1: Response of our ROS-sensitive biosensors to increasing hydrogen peroxide (H2O2) concentrations. The fluorescence of RFP, normalized to cell OD600, is shown under the control of three different ROS-sensitive promoters.
In contrast, both the SoxS and TrxCP promoters showed similar expression patterns with fluorescence peaking at 2500uM H2O2 before declining at 5000uM. Given that both these promoters are ROS inducible this decrease is unexpected, as we would typically anticipate further increases in fluorescence at higher concentrations. Additionally, we noted significant red fluorescence on plates containing the TrxCP plasmid even in the absence of added ROS. This observation indicates a leakiness of the TrxCP system over time, rendering it unsuitable for our aim to implement our system in the coral microbiome. Instead, we recommend retesting the SoxS promoter to confirm the decline in fluorescence at 5000uM. Despite this decrease the SoxS promoter still exhibited higher expression than KatG at these concentrations, positioning it as the most promising candidate for further testing.
Figure 2: Response of our ROS-sensitive biosensors in combination with catalase plasmids to increasing H2O2 concentrations. The fluorescence of RFP, normalised to cell OD600, is shown for each construct.
These findings are further supported by the fluorescence analysis of our three ROS-sensitive biosensors in conjunction with catalase plasmids exposed to varying concentrations of H2O2. Due to the leakiness of the TrxCP promoter, we decided to exclude it from these analyses and instead focused solely on the KatG and SoxS promoters. The KatG promoter exhibited a weak correlation between fluorescence and H2O2 concentration, reaffirming its unsuitability for a ROS induced system. In contrast, the SoxS promoter showed strong and consistent induction by ROS, with a clear activation threshold at 2500 µM across all four constructs. This highlights its suitability for ROS detection and solidifies it as the most promising candidate for our system.
As expected, cell survival analysis revealed that all cell lines displayed higher optical densities (OD) at lower H2O2 concentrations, reflecting enhanced viability under less toxic conditions (Figure 3). Interestingly the cells containing the SoxS promoter exhibited significantly better survival rates than those with the KatG promoter across all tested concentrations. In fact, cells expressing the KatG promoter alongside catalase plasmids displayed survival rates comparable to the biosensors alone when exposed to hydrogen peroxide. This observation suggests that the catalase plasmids in these cells do not significantly enhance cell survival. Instead, it suggests a potential detrimental interaction between the KatG promoter and all three of the catalase plasmids, which may hinder their protective effects against oxidative stress.
Figure 3: Effect of varying hydrogen peroxide (H2O2) concentrations on cell survival of E. coli cell lines expressing catalase plasmids. (a) The colored constructs represent cell lines containing one of our three catalase genes paired with a second plasmid expressing RFP under the control of the SoxS (red) or KatG (blue) promoter, with catalase expression induced by rhamnose. Constructs marked as negative were not induced. The catalase expression is inducible by rhamnose, and negatives are not induced while positives are. Black lines indicate the survival of control cell lines that do not express any catalase plasmids. (b) Colored bars represent OD600 of cell lines containing one of our three catalase genes paired with a second plasmid expressing RFP under the control of the SoxS or KatG promoter, with catalase expression induced by rhamnose. The results are grouped by construct.
Among the catalases, the native KatG from E. coli demonstrated superior survival with minimal changes in OD as ROS levels increased when combined with the SoxS promoter. However, when paired with the KatG promoter, these plasmids demonstrated the poorest performance, indicating a potential detrimental interaction. This may be attributed to the introduction of both the native E. coli KatG and its corresponding promoter into E. coli, potentially leading to unexpected regulatory interactions. Future optimisation efforts should focus on determining whether these effects are specific to the tested organism (E. Coli) or whether a similar pattern is seen in other organisms.
Furthermore, the negative controls (lacking rhamnose induction) showed performance similar to that of the rhamnose-induced plasmids. This outcome contradicts our expectation that plasmid catalase expression would not occur without induction, and therefore these cells should exhibit decreased viability. This unexpected finding raises the possibility that even basal levels of catalase expression in E. coli may be sufficient to confer cell resilience under these conditions, warranting further investigation. Future experiments should explore a wider range of H2O2 concentrations to determine the overall resilience, and incoporate CRiSPR knockdown of the native catalase to better understand its role in these cells’ survival.
Overall, our results suggest that the most promising combination of promoter and catalase is the SoxS promoter with KatG from E. coli. However, experiments should first validate the performance of each of these parts individually in other organisms. After validation, these plasmids can be integrated into a single system, allowing the catalase to be regulated by the ROS-inducible promoter and eventually incorporated into the coral microbiome.
CRISPRi Results
Despite confirmation of successful gRNA primer insertion, via enzyme digest (Fig.1) and sanger sequencing, we were unable to replicate the levels of repression observed in the source paper1, with initial testing showing less than 20% percent reduction (Fig.2).
Figure 1: Gel electrophoresis following BamHI and SacII restriction enzyme digest of gRNA inserted plasmids. The three lanes on the left is a computational prediction of the gel electrophoresis, while the four wells on the right are the experimental results. MW refers to the DNA ladder. Well 1 of the computationally predicted gel represents the BsaI and SacII enzyme digest of the initial, unreacted, plasmid, whilst well 2 represents the digest of the plasmid following gRNA insertion. Computational well 2 represents the digest of the plasmid following gRNA insertion. Experimental well 1 is the plasmid backbone without an inserted gRNA. Experimental well 2 is the plasmid backbone with the KatG gRNA inserted. Experimental well 3 is the plasmid backbone with the mScarlet3 gRNA inserted.
Figure 2: Inducing production of dCas9, gRNA and mScarlet3 to measure CRISPRi activity. + indicates the presence of the inducer, while - indicates absence. Rhamnose induces mScarlet3 production. Anhydrotetracycline induces dCas9. L-arabinose induces gRNA. ++ indicates that the L-arabinose concentration was 26.6 mM. Constitutive in the rhamnose row indicates that a constitutive promoter was used and thus rhamnose was not added. Fluorescence was normalized to mScarlet3 induction only (condition C).
Following this, we attempted to optimize for [anhydrotetracycline] and [L-arabinose], but this was also unsuccessful, with changing concentrations resulting in decreased activity relative to the source paper1conditions (Fig.3). As such we have decided to discontinue the CRISPRi element of the project. Perhaps CRISPRi system would have found more success against the genomic E. coli KatG, due to the lower transcription rate (relative to the mScarlet3 plasmid), we were unable to test this with rigor. Furthermore, current scientific literature suggests that CRISPRi systems work better when the gRNA targets a promoter. As such, in the future it would be beneficial to design a plasmid system in which the gRNA targets the KatG/mScarlet3 promoter. However, this would require the use of a different CRISPR system, as the DH10B E. coli KatG promoter does not contain the dCas9 protospacer adjacent motif.
Figure 3: Varying anhydrotetracycline and rhamnose concentration. Rhamnose induces mScarlet3 and anhydrotetracycline induces dCas9. Results are normalized to concentrations used by the source paper1.
References
- Bradley, R. W. An easy-to-use CRISPRi plasmid tool for inducible knockdown in E. coli. Biotechnol. Rep. 32, e00680 (2021).