Engineering the ROS-inducible catalase system

Design and Build

The engineering of our ROS-inducible catalase system began with a comprehensive design and construction phase. A critical initial step in designing our system was selecting candidate catalase genes for testing, which would enable optimisation of our system’s ROS-quenching abilities. We chose our candidates based on the following criteria:

  • KatG from Escherichia coli served as our benchmark, as our testing was conducted in E. coli. We thought its established functionality and compatibility would provide a reliable reference point for evaluating the performance of other candidates.
  • KatG from Mycobacterium tuberculosis, selected for its well-studied properties and effectiveness in neutralizing ROS [1].
  • KatG from Bacillus, chosen as a wildcard candidate given previous reports of excellent ROS neutralising abilities [2].

After selecting our catalase genes, we had to slightly modify our KatG from E. coli plasmid to prevent recognition and targeting by our CRISPR system.

To facilitate the construction process, we employed Golden Gate Cloning, which allowed for the easy inclusion and swapping of genetic parts, streamlining the assembly of our plasmids and constructs. We utilised this cloning method to build cell lines containing catalase plasmids that are inducible by rhamnose, enabling us to switch the expression of these catlases on and off as needed.

In addition to constructing the catalase systems, we also decided to design and build biosensors to measure catalase effectiveness and evaluate suitable ROS-sensitive promoters. Using Golden Gate Cloning, we transformed plasmids containing different ROS sensitive promoters (Trxcp, SoxS and KatG) into E. coli. A second level of golden gate cloning enabled us to incorporate either oxyr or soxr in the ROS-sensitive plasmids, facilitating the construction of a ROS inducible system with RFP expression regulated by our promoters. This design allowed us to use fluorescence as a readout of promoter inducibility, enabling us to systematically evaluate which promoter would be most effective for our system.

Test

After selecting our candidates, the next step in our engineering cycle was to identify the most effective and suitable catalase and promoter for our system. To optimize our approach, we first aimed to measure these components individually, before combining them into a single cohesive system.

After completing the transformations of our level 2 Golden Gate reactions, we observed that the induction of the TrxCP promoter was insufficiently linked to ROS levels. The plasmid-containing plates exhibited significant red fluorescence even in the absence of added ROS, suggesting that the promoter was leaky. Because of this, we deemed the Trxcp promoter unsuitable for our system and particularly inappropriate for future applications in a marine environment.

Consequently, we focused our further experiments only on the remaining promoters SoxS and KatG. Unfortunately we encountered challenges during the testing phase; our attempts to follow established protocols and hydrogen peroxide concentrations from existing literature [1,2] did not yield favorable results, as our cells struggled to survive even low concentrations of ROS.

To address this, in our second iteration of testing we drew inspiration from the methods of the HKUST 2022 iGEM team, Fisherly. They recommended incubating the cells during their log phase rather than stationary phase, significantly enhancing the cells’ resilience and our ability to measure promoter activity. This second cycle yielded more promising outcomes: the cells demonstrated improved tolerance to high concentrations of hydrogen peroxide, and the promoters effectively regulated the expression of fluorophores in response to ROS.

Learn

Through these experiments, we identified the most optimal combination of promoters and catalases for our ROS-inducible system. We found that the SoxS promoter responded most effectively to ROS levels, demonstrating a strong baseline expression and robust and consistent responses to increasing concentrations of hydrogen peroxide. Additionally, cell survival was highest in the SoxS/KatG E. coli constructs, making this combination the most promising candidate for further engineering.

Moving forward, our focus will be on integrating these elements into a single engineered system. This integration will allow us to create a robust, responsive catalase system capable of efficiently mitigating ROS concentrations in a controlled manner. Once established, we can then test this system in other organisms, with the aim of implementing it in the coral microbiome.

  1. Manca, C., Paul, S., Barry, C. E., Freedman, V. H. & Kaplan, G. Mycobacterium tuberculosis Catalase and Peroxidase Activities and Resistance to Oxidative Killing in Human Monocytes In Vitro. Infect. Immun. 67, 74 (1999).
  2. Jia, X., Chen, J., Lin, C. & Lin, X. Cloning, Expression, and Characterization of a Novel Thermophilic Monofunctional Catalase from Geobacillus sp. CHB1. Biomed Res. Int. 2016, (2016).

Engineering the CRISPRi system

Earlier studies have found that bacteria favour the production of native genes over analogues of these genes inserted via plasmids. For example, our expression bacteria (Escherichia coli DH10B) would favour its native KatG over the Mycobacterium tuberculosis KatG. To optimise the production of such proteins, we looked for analogues of the potential catalases, where we noted that only M. tuberculosis KatG has an E. coli analogue. We then attempted to use a CRISPR interference (CRISPRi) system to knockdown native KatG production, hoping to increase the rate of M. tuberculosis KatG production.

Design and Build

We chose a CRISPRi system from Bradley1 for a few reasons. Firstly their plasmid utilised a p15A backbone, which would be compatible with the plasmids used for catalase production. Secondly, it uses a chloramphenicol antibiotic resistance (ABR) gene which is different from the ABR genes used by the catalase team. Finally, this system is run on one plasmid where both the dCas9 and gRNA are present on the same plasmid (unlike many other CRISPRi systems that use a seperate plasmid for the dCas9 and gRNA).

Test

When attempting to replicate an experiment by Bradley1 to confirm that the system works as expected. Bradley1 showed that their system resulted in a 40-fold repression of mScarlet 3 (mScar3), however we were unable to replicate this. To optimise this system to our needs, we increased the concentration of L-arabinose (Ara)- which induces gRNA production- resulting in a slight increase in repression. To further optimise the system, we attempted to alter the concentration of anhydrotetracycline (aTc)- which induces the dCas9- however we were unable to reach a 40-fold repression. We noticed that this plasmid contained a H840D in the dCas9 sequence (where other plasmids generally use a H840A to deactivate Cas9). We attempted to revert this using site-directed mutagenesis however this was unsuccessful. Finally, we performed sanger sequencing on the gRNA ensuring that the guide sequence was correct.

Learn

Through our engineering process we have produced two new plasmids as part of a CRISPRi system. However, our testing revealed that this system does not work. We were unable to pinpoint the exact reason, but it it is possible that the dCas9 is faulty. ChopChop2 was used to determine the optimal gRNA sequence, thus it is highly likely that the gRNA sequence is on-target. In future iterations we plan to use a different CRISPRi system with a similar gRNA sequence.

  1. Bradley, R. W. An easy-to-use CRISPRi plasmid tool for inducible knockdown in E. coli. Biotechnol. Rep. 32, e00680 (2021).
  2. Labun, K. et al. CHOPCHOP v3: expanding the CRISPR web toolbox beyond genome editing. Nucleic Acids Res. 47, W171–W174 (2019).