Plant diseases are an increasing threat to global food security. As plant infections can rarely be visualized at early phases of plant-pathogen encounter, we sought to devise a bioassay kit that could identify an infection before phenotypic symptoms arise. For this purpose we developed a low-cost cell-free system to be used on-site, based on exchangeable toehold switches that respond to pathogen-specific gene sequences. After thorough research we got inspired by the toehold publications by Arce et al (2021) and Pardee et al (2016) and partially based our project on those. As a demonstration, we built a novel toehold switch for detecting barley yellow dwarf virus (BYDV), and compiled a library of alternative switches towards other pathogens to showcase system modularity.
As the ultimate demonstration of our engineering success we managed to design and build a working cell-free expression system that activates when a pathogen-specific trigger sequence is present in the reaction. We also confirmed the modularity of the detection kit and identified subsequent steps in the process that are necessary to develop the system towards commercialization. There is still work to do and therefore future iGEM teams can continue brainstorming and developing our modular system.
The project was subdivided into a number of consecutive design-build-test-learn (DBTL) cycles, as presented in a simplified sequence below. The entire process included seven completed interconnected iterations (Cycles 0-6), and two uncompleted cycles, which represent the next steps of the future work.
Design
Novel toehold switches were designed using design principles, shown to produce viable toehold switches. In an effort to detect the target pathogen, we designed several toehold switches that were outlined at the sites of conserved regions of pathogen genomes. More about the design of the toehold switches is presented on our design page.
In our project, we wanted to create a novel system by using a single target pathogen, which we decided to be the barley yellow dwarf virus (BYDV). We partnered with the iGEM teams TAU and IISER-Tirupati for designing more efficient toeholds for the virus. The partnerships were essential when designing our system. Along with BYDV, we designed toehold switch sequences for other target pathogens, such as the tomato brown rugose fruit virus (TBRFV) and the potato virus Y (PVY). Total of 12 pathogens including BYDV were compiled into a library of alternative switches. Rest of the pathogens can be found on the design page. In order to exemplify the modularity, we also partnered with iGEM teams Patras and TecCEM, gathering information about other possible detectable pathogens. This partnership with Patras and TecCEM was an intrinsic part of building the pathogen library, thus successfully demonstrating the global impact of our project. More about the partnership can be found on our partnership page.
Build
Toehold switches were generated using an original design algorithm. The algorithm can be accessed on our GitLab page.
Test
The performance of the designed toehold switches was estimated using computational modeling.
Learn
Based on the modeling, the best performing BYDV toehold switches were selected for laboratory testing. This cycle was performed multiple times, more information can be seen on Design, Results and Model.
Design
We selected Golden Gate cloning as the assembly platform for generating the toehold sensor plasmids (see Design). We designed our sensor plasmids so that the toehold switch is easily exchangeable with Golden Gate assembly.
Build
The reporter gene fragments for β-galactosidase (β-gal) and mScarlet-I were generated by PCR, while the rest of the parts were ordered as synthetic fragments from IDT. Components arrived in IDT Golden Gate compatible kanamycin A plasmids and we transformed them into the DH5α-cells. T7 promoter and T7 terminator were ordered from IDT as ready fragments, whereas the toehold switch sequences were self-constructed. The plasmid backbone, pOdd1 (BBa_J428381), was provided in the iGEM distribution kit. The constructs were assembled using Golden Gate cloning, and transformed into the E. coli DH5α strain.
The modularity of the detection kit is based on the standardized junction sites used in Golden Gate assembly. These junction sites correspond to the iGEM Type IIS standards, with slight modifications. Using this approach, we produced several variants of our sensor plasmid using toehold switches and different reporters. More information can be seen on our Design page.
Test
Clone candidates harboring the successfully assembled constructs were selected based on kanamycin resistance and the color of the colonies on the transformant plates, white indicating a plasmid carrying an insert fragment. The presence of the construct with a correct-sized insert, indicating successful Golden Gate assembly, was further verified using colony PCR.
Learn
The Golden Gate assembly was successful based on the acquired fragment sizes, and allowed the isolation of all fourteen toehold sensor plasmids used in the successive analysis (Cycle 6). This also demonstrated the modularity of the assembly strategy, which allows the pathogen-specific toehold sequences in the reporter plasmid to be flexibly replaced for the identification of other plant pathogens of interest. More information can be found on the Results page.
Design
To confirm the functionality of the reporter system and the detection method used in the work, we set to analyze cells expressing β-gal in a well plate format.
Build
E. coli BL21 (DE3) strain harboring the plasmid construct for ꞵ-gal expression was acquired from our university.
Test
The β-gal strain was cultured in the presence of IPTG (isopropyl β-D-1-thiogalactopyranoside) and colorimetric substrate (ONPG, 2-Nitrophenyl β-D-galactopyranoside) to induce expression, and analyzed spectrophotometrically at 420 nm using a plate reader.
Learn
The observed signal in the induced cells confirmed that the β-gal detection system was functional, and that the well plate system could be applied for the analysis for the toehold constructs. More information can be found on the Results page.
Design
We decided to use crude cell extracts obtained from the lysis of E. coli BL-21-Gold-dLac (DE3) as the basis of the cell-free system (see Design). The lysates were to be tested first with the reporter, which was shown to be functional in the previous Cycle 2.
Build
The cells were grown, lysed and used in the construction of a cell-free expression (CFE) system. The CFE system was combined with the rest of the components required for the detection, including the β-gal expression plasmid, IPTG and ONPG substrate (see Experiments).
Test
The β-gal signal was measured spectrophotometrically the same way as in the previous in vivo assay (Cycle 2).
Learn
The observed β-gal signal confirmed that the crude lysate CFE system was functional, and could be applied in the successive cycles in the work (Cycles 4-6). However, the signal from the system decreased over time, suggesting the CFE system should be further optimized (Cycle 8). More information can be seen on the Results page.
Design
Next we evaluated two different lysis methods to optimize the CFE system tested in Cycle 3 (see Design).
Build
E. coli BL-21-Gold-dLac (DE3) cells were cultured and lysed using either a probe sonicator or a phage λ-based autolysis system. The crude cell extracts from two different lysing methods were used in the construction of two distinct CFE systems and combined with expression plasmids for β-gal and mScarlet-I (see Experiments) .
Test
The acquired parallel lysates were tested for the expression of two alternative reporters, β-gal (Cycles 2 and 3) and mScarlet-I.
Learn
Both lysis systems were functional. The test set-up revealed differences between the methods; sonication appeared to be more effective for both β-gal and mScarlet-I. More information can be found on our Results page.
Design
Next, the CFE system generated by sonication in Cycle 4 was to be tested in context with a known functional toehold system with β-gal as the reporter (Pardee et al, 2016)(Cycle 3-5). For this, we found suitable toehold switch plasmid that had previously been shown to work in a cell-free context (see Design).
Build
The analyzed toehold construct (ZIKV_Sensor_27B_LacZ) was a gift from James Collins & Alexander Green, acquired from Addgene. The plasmid was amplified and extracted from E. coli DH5α.
Test
The toehold construct was used for setting up a CFE reaction evaluated in previous Cycles 3 and 4. The system was analyzed with and without the toehold-specific trigger for the expression of ꞵ-gal reporter spectrophotometrically at 420 nm using a plate reader. The trigger was provided as either ssDNA or RNA sequences in parallel reactions.
Learn
The analyzed toehold reporter system was shown to be functional in the developed CFE system. Importantly, the measured ꞵ-gal signal was induced in the presence of the trigger, with a lower nonspecific background signal in its absence. However, the activity of the toehold sensor could not be measured with RNA triggers, necessitating further testing and optimization of the CFE system (Cycle 8). More information can be found on the results page.
Design
Next, we wanted to test the BYDV-specific toehold constructs assembled in Cycle 1 in the established CFE system (Cycle 5) (see Design).
Build
Four of the 14 assembled plasmids (Cycle 1) were amplified and extracted, and in order to use them for parallel CFE reactions as in the previous successful test reaction (Cycle 5).
Test
The parallel reactions were monitored for ꞵ-gal signal (spectrophotometrically as in previous Cycle 5) in the presence and absence of the toehold-specific triggers, against negative and positive controls (see Results).
Learn
One of the tested BYDV toehold sensor plasmids was clearly functional and responded to the presence of the specific trigger (i.e. specific BYDV DNA fragment), as seen in the increase of the ꞵ-gal signal as compared to the negative control. The observed trigger-speficity was less pronounced than in the case of the control toehold tested in Cycle 5. The other three tested BYDV toehold systems did not produce the desired β-gal response in the presence of the trigger sequences in our reaction setup. Further testing is necessary for evaluating the applicability of the toeholds and for system optimization, as discussed in Cycle 8.
Design
To obtain conclusive results on the functionality of the designed BYDV sensors, additional testing is required. This will include activity screens for the remaining untested BYDV toeholds, as well as appropriate repetitions of the trials in Cycle 6.
Build / Test / Learn
To be done.
Design
Once all the functional BYDV toeholds have been identified, the reaction needs to be optimized for sensitivity and maximal signal levels. This will include evaluation of the CFE reaction conditions, as well as additional fine-tuning of the toeholds at sequence level, before the system can be optimized for the bioassay format.
Build / Test / Learn
To be done.
