Team:TU-Eindhoven/Proof Of Concept
Proof of Concept
Since the current methods for monitoring IBD patients are invasive, complex, non-specific, expensive, and time-consuming, we developed IBDetection. We envision a pill that deliberately releases Genetically Modified Organisms (GMOs), which sense tetrathionate and produces gas vesicles, into the intestine [1,2]. To proof IBDetection, we characterized the tetrathionate sensing TtrS/R system, reproduced the ultrasound-based ARG1 reporter system, and transformed the entire sensor required for IBDetection into three different E. coli strains, including BL21 (DE3) [3].
TtrS/R sensing system
The Background
The sensing part of IBDetection consists of the TtrS/R two-component system, designed by Daeffler et al. [5], which is designed to sense the presence of tetrathionate (Ttr). Additional to this system, they engineered and optimized a second orthogonal two-component sensing system to characterize the inflammation marker thiosulfate. Ideally, IBDetection would sense both inflammation markers, though due to the limited amount of time we will focus on the Ttr sensor. The motivation for this can be found in Human Practices segment The Science. The TtrS/R two-component system starts with the binding of Ttr to the transmembrane histidine kinase sensor—abbreviated to TtrS—that subsequently becomes phosphorylated. When this happens, TtrS will transfer its phosphoryl group to the cytoplasmic regulator protein, abbreviated to TtrR. Two phosphorylated TtrR proteins can homodimerize into a transcription factor. The phosphorylated TtrR dimer subsequently activates the TtrB185-269 promoter (pTtrB185-269) and recombinant super folder GFP (sfGFP) expression (Figure 1) [1]. The system is engineered based on computational analysis of the natural Ttr reduction system of Salmonella typhimurium, followed by experimental optimization of the promoter [1,4]. This optimization resulted in the sensitive and specific TtrB185-269 promoter, which is the truncated part of the natural TtrBCA operon [1,4]. Daeffler et al. [1] showed that the system with the pTtrB185-269 responds to micromolar concentrations of Ttr, with up to 30 ± 10 fold increase of super folder GFP levels. They tested the system in E. coli BL21 (DE3) and E. coli Nissle 1917 in both aerobic and anaerobic conditions and got similar results. This shows that the application of this Ttr sensor in gut-like environments is possible.
Figure 1: Schematic overview of the TtrS/R two-component system.
Before modifying the optimized sensing system to the application in IBDetection, we wanted to characterize the system by reproducing the results from Daeffler et al. [1]. Therefore, we ordered the plasmids they used for their experiments from Addgene (pKD227 and pKD233.7-3) and tested which inducer concentrations suit best for the cells and the sensor.
Plasmid Design
Both plasmids are required for the system, since the pKD227 plasmid (p15a vector) contains the IPTG inducible TtrS protein and the pKD233.7-3 plasmid (ColE1 vector) contains the doxycycline (dox, a tetracycline analog) inducible TtrR protein. The promoters used are the Tac promoter (pTac) and the LtetO-1 promoter (pLtetO-1) respectively (Figure 2). Other important proteins required for the characterization of the sensing part are the constitutively expressed mCherry, which is used for normalization purposes, and the sfGFP output expressed under the TtrB185-269 promoter (pTtrB185-269) (Figure 2). The pTtrB185-269 will later be introduced in front of the gas vesicle sequence. This promoter is the link between the sensing part of the system and eventually the reporter part and is, therefore, critical for IBDetection.
Figure 2: Schematic overview of the plasmid design for the TtrS/R two-component system. The top plasmid is a ColE1 vector with constitutive mCherry, pLtetO-1 activated TtrR and pTtrB185-269 activated sfGFP. The bottom line is a p15A vecor with a pTac activated TtrS
Experiments
For the bench-top characterization of the TtrS/R sensing system, we had to determine the concentrations of the different inducers in this system. Therefore, we did various experiments to prepare all solutions and transform the BL21 (DE3) cells with the corresponding plasmids. The transformed BL21 (DE3) cells were then induced with several concentrations of IPTG, Ttr, and dox and measured in a spectrophotometer. The GFP expression was normalized on the constitutively expressed mCherry. Subsequently, the sensing system was optimized for the detection of tetrathionate, which resulted in the optimal inducer concentrations of 0.1 mM IPTG and 0 ng/mL dox. The experiments which have been performed and the corresponding results, can be read in the Notebook, in the Results segment TtrS/R sensing system, and on the Part Registry Pages (Bba_K3972000 - Bba_K3972002). After discovering the best conditions for the system to respond to tetrathionate, a full characterization of the dose-response of the sensing part was performed. Concentrations of 0.1 mM IPTG and 0 ng/mL dox were used for induction of the large cultures, while the concentration of Ttr ranges over the region of interest (µM range). As can be seen in Figure 3, this experiment gave rise to a dose-response with an EC50 value of 49.1 ± 2.3 µM (n=1). When comparing this result with the EC50 of 50 ± 3 µM of Daeffler et al. [1], we showed a proper reproduction of the sensing system. Furthermore, we compared our results with the iGEM team SHSBNU 2017, who also made use of this system, in our Contributions.
Figure 3: Dose-response curve of the TtrS/R two-component system measured in the lab. The EC50 is equal to 49.2 ± 2.3 µM Tetrathionate.
ARG1 Reporter system
The Background
The reporter part of IBDetection will make use of a novel approach called Acoustic Reporter Genes (ARG): genetic constructs that result in gas-filled protein nanostructures [5, 6]. Bourdeau et al. [5] and Lakshmanan et al. [6] successfully characterized several IPTG inducible ARG sequences, consisting of a specific ARG sequence containing 8 to 14 gas vesicle proteins (Gvp) under a single T7 promoter (Figure 4). Based on the basic structural protein GvpA, optional external scaffolding protein GvpC, and many secondary proteins that operate as minor constituents or chaperones, these proteins will create a hollow structure. This structure will be filled with dissolved bacterial gasses, which allow in vivo visualization of bacterial gene expression using ultrasound.
In nature, these vesicles are used to control buoyancy in bacteria, such as Bacillus megaterium and cyanobacterium Anabaena flos-aquae. Bourdeau et al.[5] combined and optimized the ARG sequences from both these bacteria to create larger gas vesicles which are ultrasound responsive called ARG1. ARG1 vesicles are relatively stable, however, can be collapsed with strong acoustic pulses, making it possible to filter out all background reflection from the ultrasound images. In addition, ARG2 was designed, which is similar to ARG1, but collapses by weaker acoustic pulses making multiplexed imaging possible [5]. More details about ARG and ultrasound detection can be found in Human Practices segment The Science. For the proof-of-concept, we focus on imaging ARG1, since we will only sense the presence of tetrathionate.
Before altering this reporter system to our desires, we wanted to characterize the system by reproducing the results from Bourdeau et al. [5]. Therefore, we ordered the plasmid they used for their experiments with ARG1 from Addgene (pET28a-T7_ARG1) and tested which setup suits best for the visualization of the vesicles.
Figure 4: Schematic overview of the ARG1 reporter system.
Plasmid Design
The 12 kb pET28a vector containing the ARG1 genes was used for the first characterization. The T7 promoter controls the expression of the entire ARG1 sequence, which is roughly a 7 kb sequence starting with several repeats of structural Gvp A and Gvp C, followed by other Gvps, which form the ultrasound detectable nanostructures (Figure 5). After characterization of the reporter part, the T7 promoter will be changed into the pTtrB185-269 promoter to combine the systems.
Figure 5: Schematic overview of the plasmid design for ARG1 reporter system. It is a single pET28a vector containing a pT7 activated ARG1 reporter gene. This gene consists of 12 different gas vesicle proteins.
Experiments
For the bench-top characterization of the reporter part, we had to get familiar with the purification protocol, as well as the ultrasound detection. Because the nanostructures are stable but cannot withstand high amounts of (hydrostatic)pressure, the purification of the vesicles is different from most protein purification methods. Long centrifuge steps, dilution with PBS buffer, and careful handling of the samples are required for the purification of the vesicles. We combined the knowledge about the purification of Bourdeau et al.[5] and Lakshmanan et al.[12], to create the protocol which can be found in Experiments. Besides, we went through the engineering cycle and adapted two standard scripts of the ultrasound department of the Eindhoven University of Technology, according to the details explained in the ARG articles [5,12], as described in Engineering Success.
With these final protocols and scripts, we were able to purify our proteins and visualize them on an SDS-PAGE, as well as visualize the gas vesicles using the ultrasound. The ultrasound measurement is performed on a phantom composed of a solid mixture of 0.5% agar-agar, 49.5% PBS, and 50 % induced LB large culture. The results below show a decreasing gradient in signal when reducing the concentrations of IPTG, which is in line with the expectations (Figure 6). Detailed results can be seen in Results segment TtrS/R sensing system. Based on these results, we can conclude that we were able to reproduce the ARG1 reporter system of Bourdeau et al.[5]
Figure 6: Processed ultrasound images obtained from phantoms containing BL21 (DE3) cells with the reporter plasmid. Each phantom contained cultures induced with different concentrations of IPTG. The white pixels represent the reflections created by the gas vesicles. The pixel density decreases in correlation with a reduction in concentration of IPTG (from left to right).
Sensor
As explained in the Description, our sensor should be able to sense tetrathionate using the TtrS/R two-component system and produce ARG1 proteins as an output. By combining the TtrS/R sensing system with the ARG1 reporter system, this concept can be achieved. As can be seen in Figure 7, the TtrS/R system senses the Ttr and activates the promoter which is introduced in front of the ARG1 genes. Below, we will explain our approach and briefly discuss the results.
Figure 7: Schematic overview of the sensor system (IBDetection).
Plasmid Design
As explained, the pTtrB185-269 is the key between coupling the TtrS/R sensing system to the ARG1 reporter system. To do this, we engineered two different plasmid designs:
The first design—a tetrathionate sensor with pTtrB185-269 activated ARG1, i.e. design A—only requires the substitution of the T7 promoter in front of ARG1 for the TtrB185-269 promoter and transform all three plasmids (pKD227, pKD233.7-3, and pET28a_pTtrB-ARG12) into a single cell (Figure 8). Since the transformation of three plasmids into a single bacteria is challanging and additional problems with the origins of replication, we prepared a second design. Details about the problems are further explained on our Human Practices page segment The Application.
In contrast, the second design—a tetrathionate sensor with transferred TtrR and pTtrB activated ARG1, i.e. design B—only requires two plasmids (pKD227 and pET28a_TtrR-pTtrB-ARG13). For the system to work with two plasmids, we had to transfer the TtrB185-269 promoter to the pET28a vector, together with the TtrR protein. Since we want to keep using the same inducers as for the TtrS/R sensing system, we also had to transfer the pLtetO-1 in front of the TtrR and the sequence for the tetR inhibitory protein (Figure 9).
Figure 8: Schematic overview of the plasmid design for the tetrathionate sensor with pTtrB activated ARG1 (Design A). It contains the ColE1 vector and the p15A vector from the TtrS/R sensing system and a pET28a vector with a pTtrB185-269 activated ARG1.
Figure 9: Schematic overview of the plasmid design for the tetrathionate sensor with transferred TtrR and pTtrB activated ARG1 (Design B). It contains the p15A vector from the TtrS/R sensing system and a pET28 vector with a pLtetO-1 activated TtrR and a pTtrB185-269 activated ARG1.
Plasmid Construction
To create the plasmids required for both designs, we researched several assembly methods. Restriction ligation techniques appeared to be the best option. To substitute the promoter, it is preferential that the restriction sites flank the T7 promoter present in the ordered plasmid. Due to the long length of the ARG1 plasmid, several endonucleases have multiple restriction sites on the plasmid, making them unreliable. The best option was to cut the sequence in the non-coding DNA between the GvpA and GvpC proteins and include the sequence for GvpA into the insert. As a result, we were able to cut the pET28a_T7-ARG1 plasmid with the restriction enzymes SgrAI and NheI and restricted the T7 promoter and GvpA DNA sequence. To create the desired systems for the sensor, we ligated a pTtrB185-269-GvpA insert, flanked with the proper restriction sites for design A, and additionally, we inserted tetR-TtrR-pTtrB185-269-GvpA with the desired restriction sites for design B (Figure 10).
Figure 10: Schematic overview of the plasmid construction of design A (left) and design B (right). The top row consists of the three plasmids used for the characterization of the sensing system and reporter system separately. The middle row shows the restriction of these plasmids and the bottom row displays the (constructed) plasmids required for each design.
Co-transformation
The restriction-ligation product of the plasmid for design A was successfully transformed into XL10-Gold cells. Subsequently, the DNA was amplified and purified for the next challenge, which is the transformation of all three plasmids required for design A into a single BL21 (DE3) cell. As can be read on our Results page, various transformations have been performed, however, none has resulted in successfully transformed BL21 (DE3) cells containing design A.
Since design A did not result in co-transformed cells, we continued with design B. For design B, we also performed restriction-ligation experiments of the required plasmid and transformed the plasmid into XL10-Gold cells. Since the first time failed, we repeated the experiment and transformed the DNA into various other E.coli competent cells. This resulted in colonies of TOP10 and XL10-Gold cells, containing the design B plasmid. Since there was no time left to characterize the complete design B, we transformed the design B plasmids into BL21 (DE3), TOP10, and XL10-Gold cells to show the potential of this design (Figure 11). All three types of bacteria were successfully transformed with the plasmids required for IBDetection to work.
Figure 11: Agar plates with bacteria containing both plasmids required for design B and thus IBDetection to work. a) BL21 (DE3) b) XL10-gold c) TOP10.
Conclusion
Even though the characterization cannot be performed, we are able to predict that the system should be successful. At first, we were able to reproduce and characterize the dose-response of the TtrS/R sensing system. These results show controlled induction of the TtrB185-269 promoter in the presence of micromolar concentrations of tetrathionate. Furthermore, like Bourdeau et al.[5], we were able to show that the gas vesicles produced under a certain promoter (pT7 for the ARG1 reporter system) can be visualized by ultrasound imaging and a SDS-Page. Supplementary, we made a Model which is able to simulate similar GFP output as measured in the lab. Even though the experiments for design B were not executed inside the lab, the model was able to give a prediction for the number of gas vesicles that are produced after induction with various Ttr concentrations. Since the modeled amount of gas vesicles for design B is in the same range as measured by Bourdeau et al. [5], we expect the system to give similar results in real-life as in the model. Therefore, we can state that IBDetection has a high chance of successfully measuring IBD (tetrathionate), giving an ultrasound detectable gas vesicle output.
1. All normalized GFP emissions are signals measured at 512 nm and normalized on the mCherry emission at 610 nm.
2. pET28a_pTtrB-ARG1 is the pET28a_T7-ARG1 vector of which the T7 promoter was substituted by a TtrB185-269 promoter.
3. pET28a_TtrR-pTtrB-ARG1 is the pET28a_T7-ARG1 vector of which the T7 promoter was substituted by a sequence containing TetR-TtrR-pTtrB185-269.
-
Daeffler, K., Galley, J., Sheth, R., Ortiz‐Velez, L., Bibb, C., Shroyer, N., Britton, R. and Tabor, J., 2017. Engineering bacterial thiosulfate and tetrathionate sensors for detecting gut inflammation. Molecular Systems Biology, [online] 13(4), p.923.
-
Riglar D, Giessen T, Baym M, Kerns S, Niederhuber M, Bronson R et al. Engineered bacteria can function in the mammalian gut long-term as live diagnostics of inflammation. Nature Biotechnology [Internet]. 2017;35(7):653-658.
-
BL21(DE3) Competent Cells [Internet]. Thermofisher. 2021 [cited 9 September 2021]. Available from: https://www.thermofisher.com/order/catalog/product/EC0114#/EC011
-
Hensel M, Hinsley A, Nikolaus T, Sawers G, Berks B. The genetic basis of tetrathionate respiration in Salmonella typhimurium. Molecular Microbiology [Internet]. 1999;32(2):275-287.
-
Bourdeau R, Lee-Gosselin A, Lakshmanan A, Farhadi A, Kumar S, Nety S et al. Acoustic reporter genes for noninvasive imaging of microorganisms in mammalian hosts. Nature [Internet]. 2018 [cited 9 May 2021];553(7686):86-90.
-
Errico C, Pierre J, Pezet S, Desailly Y, Lenkei Z, Couture O et al. Ultrafast ultrasound localization microscopy for deep super-resolution vascular imaging. Nature [Internet]. 2015;527(7579):499-502.
-
Chu J, Oh Y, Sens A, Ataie N, Dana H, Macklin JJ, et al. A bright cyan-excitable orange fluorescent protein facilitates dual-emission microscopy and enhances bioluminescence imaging in vivo. Nat Biotechnol. 2016;34(7):760–7.
-
Daniel C, Poiret S, Dennin V, Boutillier D, Pot B. Bioluminescence imaging study of spatial and temporal persistence of Lactobacillus plantarum and Lactococcus lactis in living mice. Appl Environ Microbiol. 2013;79(4):1086–94.
-
Foucault M-L, Thomas L, Goussard S, Branchini BR, Grillot-Courvalin C. In vivo bioluminescence imaging for the study of intestinal colonization by Escherichia coli in mice. Appl Environ Microbiol. 2010;76(1):264–74.
-
Foster FS, Lockwood GR, Ryan LK, Harasiewicz KA, Berube L, Rauth AM. Principles and applications of ultrasound backscatter microscopy. IEEE Trans Ultrason Ferroelectr Freq Control. 1993;40(5):608–17.
-
Lakshmanan A, Lu GJ, Farhadi A, Nety SP, Kunth M, Lee-Gosselin A, et al. Preparation of biogenic gas vesicle nanostructures for use as contrast agents for ultrasound and MRI. Nat Protoc. 2017;12(10):2050–80.