Overview

One of the main threats when releasing genetically modified organisms (GMO) in the soil is the possibility of horizontal gene transfer [1]. This is the transfer of plasmid DNA and thereby non-natural genetic content to native bacteria in the environment. Therefore, in this project a double plasmid interdependent Toxin-Antitoxin (T-A) system has been designed to stop horizontal gene transfer between our system and the native soil bacteria. Both plasmids of the designed system contain a toxin and an antitoxin. The toxin has its corresponding antitoxin on the other plasmid, making the system interdependent (Figure 1). Through this interdependence a bacterium will need both plasmids (or none) to survive. If a native soil bacterium received a single plasmid it would die. The possibility of the two plasmids to be transferred simultaneously is very improbable yet possible; however, this system makes gene transfer much less likely to occur than any other known system. An additional advantage of this system is the ability to maintain the GMO population by positive selection. As losing a plasmid is metabolically attractive for the GMO but also deadly, the system would make the population retain their plasmids.

Two sets of T-A combinations, Kid-Kis and Zeta-Epsilon which have been studied for biosafety purposes before [2], were selected for this project. Both are type II T-A systems, meaning that antitoxin protein prevents the toxin protein from functioning by binding.

Epsilon and Zeta

The Epsilon (antitoxin) and Zeta (toxin) were first found on pSM19035, a plasmid originating from Streptococcus pyogenes [3]. The T-A cassettes in plasmids such as pSM19035 have the biological function to stabilize plasmids, maintaining the plasmid in non-selective conditions. The Zeta toxin converts an essential building block of the bacterial cell wall into a form that prevents normal cell wall growth. Due to the distorted cell wall the bacterial cell becomes prone to hydrostatic pressure, leading to cell lysis with comparable effects to penicillin [4]. Epsilon antitoxin forms a complex with the Zeta toxin disabling the toxins function, with an expected ratio of 1:1 [4]. Zeta toxin has been proven to be bactericidal for gram-positive Bacillus subtillis and bacteriostatic for gram-positive E. coli [3]. Nonetheless, more recent works show that Zeta toxins are also bactericidal for E. coli [4].

Kis and Kid

Kis (antitoxin) and Kid (toxin) are part of the parD stability system of plasmid R1 acting as a pre-segregation mechanism to ensure plasmid proliferation. Kid breaks down single stranded messenger RNAs. By degradation of something as vital as messenger RNA, this toxin does not differ in function with gram positive and gram negative bacteria. In nature, Kid causes cells to divide slower to ensure that every daughter cell receives the R1 plasmid. In our system an excess of Kid will result in a bactericidal effect. Additionally, the Kis-Kid interaction has proven to be a 1:2 ratio [5, 6].

Final design and system specifications

The T-A system is intertwined with both the Kill-Switch and our host Pseudomonas putida. The toxins were designed to be part of the Kill-Switch, making use of the required balance between toxins and antitoxins. The activation of the Kill-Switch after induction and subsequent absence of fusaric acid will cause the toxins to be produced, disrupting the T-A balance, resulting in cell death. There are different theories on the T-A system putative functions that are not mutually exlusive. Besides plasmid retention, increasing evidence suggests that the T-A systems can act like a hysteric switch, in which a fraction of the population can enter a toxic state. In this toxin state the free toxins inhibit cell growth transforming the cells into persistent cells. In this slow growing persistent state the organisms become more robust to external fluctuations allowing it to survive in unfavorable environments [7,8]. Many micro-organisms contain multiple toxin-antitoxin systems, six T-A systems are predicted for P. putida. As free toxins have an impact on the growth of our host this characteristic has been taken up in the system performance model in order to describe the system as accurately as possible.

Approach, constructing BioBricks

Cloning the T-A systems in pSB1C3

The first step to build the toxins and antitoxin BioBricks was a PCR amplification and elongation to add a Ribosomal Binding Sites (RBS) to the individual toxins and antitoxins and incorporate standard overhangs for RFC10 assembly, containing the BioBrick prefix and suffix. Two plasmids, pHT286 and pAB24, containing Zeta-Epsilon and Kis-Kid respectively, were obtained from Imperial Imperial College London (ICL) ACS Synthetic Biology [2]. After successful amplification and elongation, the toxins and antoxins were cloned in the standard iGEM backbone pSB1C3 with and without the LacI promoter BBa_R0010.

Toxins and antitoxins cloned in pSB1C3 containing and not containing the LacI promoter were transformed into commercially DH5a cells. Moreover, toxins and antitoxins containing the LacI promoter were transformed into E. coli JM109, that expresses the LacI repressor, and therefore can inhibit the production of toxin, a desirable effect to avoid the death of the cell by their toxic effect.

Cloning the toxins in a low copy number plasmid in combination with the antitoxins

An alternative approach was taken, cloning the two toxins in low copy number plasmids in electro competent cells containing the antitoxins in the pSB1C3 backbone. To chose the most appropriate backbone we made a complete study of the different Origin of Replication (Ori) of the iGEM standard backbones becasue our team members was experiencing problems in finding an actual low copy number plasmid in the registry. Results indicated that the only real low copy number plasmid is pSB3K3, which contains the p15A Ori (clik here for detailed results). Also different resistance marker was required to select for two plasmids in the same cell, being chloramphenicol and ampicillin.

Separating toxins and antitoxins.

Once having obtained the toxins cloned in pSB3K3, our aim was to separate the toxins and antitoxins using agarose gel electrophoresis and showing that toxins and antitoxins sequences were completely separated.

For more specific details about the materials and methods used in our project, please, go our journal and protocols section.

Results and discussion

Cloning the T-A systems in pSB1C3

Building BioBricks for the double interdependent T-A system was a challenging task because the toxins, as the name implies, are toxic to the cell. Colonies were obtained both for toxins and antitoxins, but when verified by restriction digestion, only the antitoxins were obtained in both types of cells E. coli DH5a and JM109 (Figure 2). The most probable reason for this results is that the LacI promoter is leaky and that even without a promoter the toxins are transcribed and translated being toxic for the cell, causing the cloned DNA to rearrange and lose the toxin coding part. Kis and Epsilon antitoxins, can be found in the registry under the code BBa_K1493602 and BBa_K1493603.

On the other hand, when cloning the antitoxins in pSB1C3 without the LacI promoter, only the Kis antitoxin was successfully obtained (BBa_K1493601).

Cloning the toxins in a low copy number plasmid in combination with the antitoxins

Following the previous results, the toxins (Zeta and Kid) were transformed in E. Coli containing the antitoxins, Epsilon and Kid. When taking this approach, successful restriction digestion of the combination of both plasmids, showed the presence of the toxin and antitoxins (Figure 3) in JM109 E. Coli cells, that contain the LacI repressor.

Separating toxins and antitoxins.

Once obtained the cells with the combination of toxins and antitoxins, we aimed to separate them by gel electrophoresis. We evaluate the most optimal setup to determine the optimum agarose percentage, voltage and time of electrophoresis. The plasmids were physically separated, as can be seen in Figure 4, the high copy number plasmid (pSB1C3) is smaller in size, therefore, runs further and has a higher intensity. Toxin bands for Kid and Zeta were extracted and purified from the gel and a PCR to check for the presence of the antitoxins was performed. The results showed that the isolated toxins still had amplification bands corresponding to the antitoxins. Due to lack of time no further steps were performed to continue building the double plasmid interdependent T-A system. Nonetheless, we introduced in the iGEM registry the antitoxin parts of the Kis-Kid and Epsilon-zeta in three different BioBricks BBa_K1493601, BBa_K1493602 and BBa_K1493603.

Conclusions

This project aimed to build a double interdependent system to prevent the horizontal gene transfer between BananaGuard and other native bacteria in the soil. The designed system is an innovative approach to avoid horizontal gene transfer. The two main advantages over other biosafety existing systems are that the GMO dies if the plasmid is lost and that the system itself regulates when passing it after division. Moreover, cell viability depends mainly on the balance between two T-A systems and no genome integration is needed. As working with toxins is a challenging work because cells die immediately after toxin cloning, several repetitions and setups were required to successfully built BioBricks for Kis and Epsilon, the antitoxins of the system. Interestingly, we could only clone the toxins in low copy number plasmid (pSB3K3) in combination with the antitoxins, but no successful separation was achieved.

Future work

Future steps should focus on cloning the toxins in a low copy number plasmid without the promoter and, later, clone them behind the promoter and antitoxins of the same T-A pair and with the other, in order to test the whole final system (Figure 1). Moreover, it would be useful to use a less leaky promoter than LacI, like the Rhamnose promoter, especially for the toxins which will decrease cell toxicity. Moreover, as a final step, this system should also be transformed in P. putida and integrated with all the other BananaGuard modules.

BioBricks

BBa_K1493601, BBa_K1493602 and BBa_K1493603

1. Wright, O., G.B. Stan, and T. Ellis, Building-in biosafety for synthetic biology. Microbiology, 2013. 159(Pt 7): p. 1221-35.
2. Wright, O., et al., GeneGuard: A Modular Plasmid System Designed for Biosafety. ACS Synthetic Biology, 2014.
3. Zielenkiewicz, U. and P. Cegłowski, The toxin-antitoxin system of the streptococcal plasmid pSM19035. Journal of bacteriology, 2005. 187(17): p. 6094-6105.
4. Mutschler, H., et al., A novel mechanism of programmed cell death in bacteria by toxin–antitoxin systems corrupts peptidoglycan synthesis. PLoS biology, 2011. 9(3): p. e1001033.
5. de la Cueva-Méndez, G. and B. Pimentel, Gene and cell survival: lessons from prokaryotic plasmid R1. EMBO reports, 2007. 8(5): p. 458-464.
6. Ruiz-Echevarría, M.J., et al., Kid, a small protein of the parD stability system of plasmid R1, is an inhibitor of DNA replication acting at the initiation of DNA synthesis. Journal of molecular biology, 1995. 247(4): p. 568-577.
7. R.A. Fasani, M.A. Savageau Molecular mechanisms of multiple toxin-antitoxin systems are coordinated to govern the persister phenotype (2013) . Proc. Natl. Acad. Sci. USA, 110, pp. E2528–E2537
8. Etienne Maisonneuve, Kenn Gerdes (2014), Molecular Mechanisms Underlying Bacterial Persisters. Cell Volume 157, Issue 3, Pages 539–548