~Engineering~

Introduction: Pioneering Engineering in Biology

“Failure is a vital part of all scientific endeavor” -Entrapta, She-Ra, and the Princesses of Power

Synthetic biology is an interdisciplinary scientific research field that utilizes engineering principles to modify and design living organisms and systems. Throughout our iGEM journey, we consistently applied the Design-Build-Test-Learn cycle to refine and enhance our project. This involved characterizing bioparts and creating constructs with transcriptional units to establish a thoroughly characterized, measurable, and orthogonal engineering system. Trial and error played a key role in our journey. The setbacks we encountered helped us improve our designs and the way we think as future scientists. Each challenge taught us valuable lessons, leading to better solutions and a deeper understanding of our work. Our journey in engineering this system wasn't just about the end product—it was about embracing the process and learning from every step.

For engineering THAELIA and ensuring optimal outcomes, we divided our system into four key components for focused study and experimentation:

• 1st Engineering Cycle: Evaluation of the T7 polymerase system

• 2nd Engineering Cycle: Production of Outer Membrane Vesicles (OMVs)

• 3rd Engineering Cycle: Synthesis of the dsRNA in vitro and assessment of its silencing efficacy

This modular approach allowed us to focus on each element and improve the final design.

Cloning procedure:

As iGEM Thessaly 2024, for our cloning experiments, a variety of vectors were used to accommodate two different chassis. For E. coli strains (DH5α and BL21(DE3)), Golden Braid vectors were utilized, following the established Golden Braid grammar for modular cloning. The domestication process was carried out using the Golden Braid Domesticator tool to ensure compatibility and efficient assembly. Additionally, another cloning method, the Golden Standard method, was employed this year to address the needs of the second chassis, Pseudomonas putida.

By following the steps of the previous iGEM Thessaly teams, we decided to use the Golden Braid 2.0 cloning method, that allows the assembly of DNA parts in a standardized and efficient manner by using Type IIS restriction enzymes, that recognize specific DNA sequences but cleave at a defined distance away from the recognition site [1, 2].

The first step in GoldenBraid is to create Level 0 constructs. In this level, individual DNA parts (such as promoters, coding sequences, and terminators) are inserted into universal domesticator plasmids. For these parts to be compatible with the GoldenBraid system, a process called domestication must occur, where internal restriction sites that would interfere with the assembly process are removed, and specific 4nt overhangs are added to match the GoldenBraid grammar. These Level 0 constructs then serve as the foundational building blocks for higher-level assemblies in the cloning process [2].

For creating our Level 0 constructs, we used the BsmbI Type IIS enzyme and the pUPD2 vector.

Figure 1: Creating Level 0 constructs by inserting the Basic Parts with the overhangs in pUPD2 vector by using the BsmbI TypeIIS enzyme

To assemble functional devices (transcriptional units), we combined our basic parts using the Type IIS enzyme BsaI and the pDGB3 alpha1 and pDGB3 alpha2 vectors. In this process, multiple Level 0 parts were assembled into these vectors, resulting in a complete transcriptional unit, which is referred to as a Level a construct.

Figure 2: Creating Level a constructs by inserting the Basic Parts from pUPD2 vector into pDGB3a1 by using the BsaI TypeIIS enzyme

To move from Level a to Level ω using the Golden Braid cloning system, we utilized the BsmbI Type IIS restriction enzyme, similar to how BsaI is used in Level a. In this process, the two Level a constructs were assembled into the pDGB3ω vector, resulting in a complete Level Ω construct with two transcriptional units.

However, while working with Golden Braid vectors facilitated the straightforward assembly of DNA constructs, these vectors utilize the pBR322 origin of replication, which is incompatible with our proposed chassis, P. putida [3]. So, to successfully implement our constructs in P. putida, it was necessary to transition from Golden Braid to Golden Standard vectors. The Golden Standard vectors contain the pBBR1 origin of replication, which is compatible with P. putida, enabling stable replication and efficient gene expression within this target organism [4]. This transition ensured that our assembled constructs could function effectively in our chassis.

The structure of pUPD2, despite being a Golden Braid vector, remains compatible with various cloning methods. By utilizing the BsaI restriction enzyme, Level 0 parts from the pUPD2 vector can be efficiently integrated into the Level 1 Golden Standard vector pSEVA23g19[g1].

Figure 4: Assembly of genetic parts in pUPD2 and successful transfer of transcriptional units into pSEVA23g19[g1] Plasmid

1st Engineering Cycle : Evaluation of the T7 polymerase system

Our project, THAELIA, is centered on bmRNAi, where dsRNA is produced and delivered to the fungal pathogen V. dahliae. The first critical objective of our system is to generate dsRNA in sufficient quantities and efficiently encapsulate it into OMVs for delivery. A key factor in achieving this is the efficient production of dsRNA, which relies on the T7 polymerase system. Consequently, a major challenge we needed to address was how to effectively regulate the production of T7 polymerase to ensure optimal dsRNA synthesis without imposing undue stress on the bacterial chassis.

First iteration: Identification of the most suitable promoter for T7 polymerase production

Design

During our Designing phase, we were full of questions, such as; “How were we going to regulate the T7 polymerase production?” , And most importantly; “How could we isolate this function from other cellular processes?”

The T7 polymerase is a key component of our engineered bacterial system, driving the efficient production of double-stranded RNA (dsRNA). To determine the most suitable promoter for the T7 polymerase production system we tested three BG synthetic promoters (i.e. BG17, BG37, BG42) from the Zobel et al. library [6]. We selected the promoters BG17, BG42, and BG37 based on findings from Zobel et al., which demonstrated their significant activation during the exponential growth phase. Our objective was to evaluate their temporal activation profiles and relative strength. We used E. coli BL21 (DE3) cells, which are able to express the T7 polymerase system, so they can serve as a relevant model for our engineered P. putida chassis, allowing us to evaluate our constructs’ effectiveness in a system closely resembling our final design.

Figure 5: The BG promoters we tested for finding the most suitable promoter for the T7 polymerase production device. We used sfGFP as a reporter gene.

The promoter J23119 served as a positive control because it is a well-characterized constitutive promoter from the Anderson family, allowing for direct comparison of promoter strength and consistent activation throughout all growth phases [7]. Additionally, osmY is a well-characterized promoter specifically induced during the stationary phase, making it a valuable positive control for assessing time-dependent activation and promoter strength during this phase [8].

Figure 6: The J231119 and osmY promoters serve as positive controls. We used sfGFP as a reporter gene.

Build

We began by cloning the parts we needed (BG37, BG42, BG17, J23119, osmY, ter, sfGFP, T7pol, T7pro and T7hyb6ter, which included Golden Braid overhangs, into the pUPD2 vector to create the Level 0 parts. The promoters were synthesized by IDT with their respective RBSs already integrated. We used E. coli DH5α cells for the cloning procedure due to their high transformation efficiency.

Figure 7. Diagnostic digestion of the generated Level 0 parts by using EcoRV and NotI. A. in silico virtual digestion by GelSim Tool from SnapGene. B.1 % agarose gel ran at 110 V for 25 min stained with EtBr. The lanes order is as follows: (1) pUPD2 (control), (2) J23119-RBS2, (3) osmY-RBS2, (4)BG37-RBS2, (5) BG42-RBS2, (6) BG17-RBS2, (7) T7 pol.

Next, we transfered the basic parts from the pUPD2 vectors into the Golden Braid vectors: pDGB3a1 and pDGB3a2 vectors to assemble the level alpha constructs. This step enabled us to generate complete, functional genetic modules for further experimentation and analysis.

Figure 8. Diagnostic digestion of the generated Level a constructs by using EcoRV and NotI. A. in silico virtual digestion by GelSim Tool from SnapGene. B.1 % agarose gel ran at 110 V for 25 min stained with EtBr. The lanes order is as follows: 1) pDB3a1, (2) BG37-RBS2-sfGFP-ter, (3) osmY-RBS2 -sfGFP-ter, (4) J23119-RBS1 -sfGFP-ter, (5)BG37-RBS2-T7pol-ter , (6) , (7)P3.1-RBS2-sfGFO-ter , (8) P3.1-RBS1-sfGFP-ter, (9) BG37-RBS1-sfGFP-ter, (10) BG37-RBS1-sfGFP-ter, (11) T7pro-RBS2-sfGFP- T7terhyb6 , (12) BG17-RBS1,-sfGFP-ter (13) BG42-RBS1 -sfGFP-ter

Test

After assembling our plasmid constructs, we successfully transformed them into chemical competent E. coli BL21 (DE3). The following day, we prepared 10ml liquid cultures of LB medium, with the proper antibiotic, to promote the growth and acclimatization of the transformed bacteria, allowing them to incubate overnight for 16 hours at 37°C and 210 rpm. On the third day, we centrifuged the cultures and washed the resulting pellets twice with 0.8% saline solution (NaCl). Subsequently, we performed a 1:100 dilution to measure the optical density at 600 nm (OD600). To achieve an OD600 of 0.1 with a final volume of 2 mL of M9 culture with the corresponding carbon source, we applied the dilution equation "Cinitial x Vinitial= C final x V final". M9 medium was selected for this experiment due to the fluorescence interference associated with LB medium. We then transferred 200 μL of each diluted culture into the wells of a 96-well plate with a clear bottom. Each construct was tested in five technical repeats. The plate was incubated in a plate reader for 15 hours at 37°C, which is optimal for E. coli while shaking at 180 rpm. Measurements were automatically recorded every hour, monitoring OD at 600 nm and sfGFP fluorescence at 515 nm (emission peak of sfGFP, Roberts et al. )

Figure 9: Bacterial growth curve, with M9 medium with 0,2% glucose as a carbon source, based on Optical Density at 600 nm measurements taken every 1 hour. Data points represent the average of five technical replicates, with any values showing a standard deviation greater than 0.2 excluded for clarity. Error bars correspond to standard deviation of n=5 replicates. Blank was subtracted.

Figure 10: Normalized fluorescence intensity for J23119-RBS2-sfGFP-ter, osmy-RBS2-sfGFP-ter, BG37-RBS2-sfGFP-ter, BG17-RBS2-sfGFP-ter and BG42--RBS2-sfGFP-ter constructs during the time points: 3h, 6h, 10h and 15h. The RFUs (sfGFP measurements at 515 nm) are divided by cell growth (Optical Density at 600nm), in order to normalize all values. E. coli BL21 (DE3) cells with pDGB3a1 were used as the (-) control. Error bars correspond to standard deviation of n=5 replicates. Blank was subtracted.

Learn

By examining Figure 9 and analyzing the OD600 measurements, we learn that significant bacterial growth occurs between the 1-hour and 8-hour time points for all promoters except BG37. For E. coli cells containing the BG37 promoter, the growth continues to increase up to the 12-hour mark, with a steady rise in OD600 between the 6-hour and 12-hour time points indicating that the cells are in the exponential phase.

Further insights can be gained from Figure 10, which confirms that the cells remain in exponential growth at the 12-hour time point. BG37, in particular, demonstrates strong activity during this period. At the 6-hour time point, BG37 shows a notable increase in fluorescence, surpassing that of the other promoters, highlighting its heightened responsiveness during early exponential growth. By 10 hours, BG37 continues to exhibit superior regulation compared to BG17 and osmy. Even at 15 hours, when the cells are transitioning out of exponential growth, BG37 remains active, showcasing its sustained effectiveness.

In comparison, while J23119 and BG42 also demonstrate strong performance, BG37 distinguishes itself with its early activation at 6 hours, making it particularly suited for systems that require robust expression during the initial stages of exponential growth. Although BG42 and J23119 provide more consistent expression across various phases, the results clearly indicate that BG37 is the optimal choice for targeted regulation during the exponential phase. Its early and sustained activity positions it as the most effective promoter for driving T7 polymerase production in scenarios where rapid and efficient expression is essential.
So, by completing this Design-Build-Test-Learn (D-B-T-L) cycle, we successfully integrated the BG37 promoter into our system to facilitate the production of T7 polymerase.

Figure 11: Production of T7 polymerase, regulated by BG37 promoter.

Second iteration: Identification of the most suitable promoter for T7 polymerase production

Design:Evaluation of T7 polymerase production device

After identifying BG37 as the most suitable promoter for T7 polymerase production, we aimed to validate the functionality of a Level Ω construct designed to efficiently express T7 polymerase under the control of BG37. Due to time constraints, we decided to test T7 polymerase production by using a sfGFP reporter under the control of a T7 promoter. For this reason, the J23119 Anderson and BG37 constitutive promoters were used to drive T7 polymerase gene expression. The J23119 promoter served as a positive control for continuous T7 production, while E. coli cells with the pDGB3omega plasmid were used as a negative control. Additionally, we aimed to assess the timing of BG37 promoter activation, focusing on whether it activates specifically during the exponential phase. We used E. coli DH5a cells for this experiment, which are optimized for maintaining plasmids, with specific mutations (such as recA1 and endA1) that prevent unwanted recombination and degradation of foreign DNA and lack the T7 RNA polymerase system [9].

Figure 12: Level omega constructs lead to T7 polymerase and sfGFP production.

Build:

Once we obtained our level A constructs, as it was described in the previous DBTL, in the pDGB3a1 and pDGB3a2 vectors, we proceeded to create our level omega constructs (BG37-RBS2-T7polymerase-ter + T7pro-sfGFP-T7ter, PJ23119-RBS2-T7polymerase-ter + T7pro-sfGFP-T7ter) by cloning them into the pDGB3omega vector.

Figure 13: Diagnostic digestion of the generated Level ω constructs by using EcoRV and NotI. A. in silico virtual digestion by GelSim Tool from SnapGene. B.1 % agarose gel ran at 110 V for 25 min stained with EtBr. The lanes order is as follows: (1) pDB3omega, (2) BG37-RBS2-T7pol-Ter + T7pro- RBS2-sfGFP-T7ter, (3) J23119-RBS2-T7pol-Ter + T7pro-RBS2-sfGFP-T7ter

Test

After assembling our plasmid constructs, we successfully transformed them into chemical competent E. coli DH5a. The following day, we prepared 10ml liquid cultures of LB medium, with the proper antibiotic, to promote the growth and acclimatization of the transformed bacteria, allowing them to incubate overnight for 16 hours at 37°C and 210 rpm. On the third day, we centrifuged the cultures and washed the resulting pellets twice with 0.8% saline solution (NaCl). Subsequently, we performed a 1:100 dilution to measure the optical density at 600 nm (OD600). To achieve an OD600 of 0.1 with a final volume of 2 mL of M9 culture with the corresponding carbon source, we applied the dilution equation "Cinitial x Vinitial= C final x V final". M9 medium was selected for this experiment due to the fluorescence interference associated with LB medium. We then transferred 200 μL of each diluted culture into the wells of a 96-well plate with a clear bottom. Each construct was tested in five technical repeats. The plate was incubated in a plate reader for 15 hours at 37°C, which is optimal for E. coli while shaking at 180 rpm. Measurements were automatically recorded every hour, monitoring OD at 600 nm and sfGFP fluorescence at 515 nm (emission peak of sfGFP, Roberts et al. )

Figure 14: Bacterial growth curve, with M9 medium with 0,2% glucose as a carbon source, based on Optical Density at 600 nm measurements taken every 1 hour. Data points represent the average of five technical replicates, with any values showing a standard deviation greater than 0.2 excluded for clarity. Error bars correspond to standard deviation of n=5 replicates. Blank was subtracted.

Figure 15: Normalized fluorescence intensity for J23119-RBS2-T7pol-ter + T7pro-RBS2-sfGFP-T7terhyb6, BG37-RBS2-T7pol-ter + T7pro-RBS2-sfGFP-T7terhyb6 level omega constructs during the time points:1h, 3h, 6h, 10h and 15h. The RFUs (sfGFP measurements at 515 nm) are divided by cell growth (Optical Density at 600nm), in order to normalize all values. E. coli DH5a cells with pDGB3omega1 were used as the (-) control. Error bars correspond to standard deviation of n=5 replicates. Blank was subtracted.

Learn

In figure 14, we observed that the cells exhibit suboptimal growth, likely due to the metabolic burden imposed by high plasmid copy numbers and the potential loss of plasmid over time. This metabolic strain can reduce overall system performance, as cells must allocate significant resources to plasmid replication and maintenance. To enhance the efficiency and stability of our system, we propose integrating the T7 polymerase production device directly into the bacterial chromosome. Chromosomal integration would alleviate the metabolic load by reducing plasmid-associated demands and ensuring stable, consistent expression of the T7 polymerase, leading to improved overall performance and growth.

Additionally, the BG37 promoter demonstrates strong autoinducible activity during the exponential phase and beyond, reaching expression levels comparable to or higher than the constitutive J23119 promoter. This confirms the utility of BG37 for applications requiring high expression during growth.

Third iteration: Integration of T7 polymerase into the chromosome

Design

Building on insights gained from our previous Design-Build-Test-Learn (DBTL) cycle, we aimed to integrate the T7 polymerase transcriptional unit directly into the P. putida chromosome. This approach was selected to achieve stable, single-copy expression of the T7 polymerase, eliminating the potential variability and instability associated with plasmid-based systems[10]. Chromosomal integration ensures that the gene remains a permanent fixture within the bacterial genome, reducing the risk of loss during cell division and promoting consistent expression throughout the population [11].

So, to achieve stable expression of the T7 polymerase in P. putida, we set out to integrate the T7 polymerase transcriptional unit directly into the bacterial chromosome. After consulting with Dr. Esteban Garcia-Martinez, we identified the mini-Tn7 system as the most suitable tool for our goal.

The mini-Tn7 system is a broad host-range vector that allows for the stable integration of single-copy genes into bacterial chromosomes at a neutral site: attTn7. This system is particularly useful for bacteria with single attTn7 sites, such as P. aeruginosa and P. putida. The system operates through two key components. The first component is the Tn7-M delivery vector, contains an R6K replication origin, which makes it a suicide vector; it can replicate only in the presence of Pir proteins, ensuring that it remains stable and does not persist in the bacterial host after the desired integration has occurred [12]. Simultaneously, it has a multiple cloning site (MCS) flanked by Tn7 left (Tn7L) and Tn7 right (Tn7R) sequences.The second component is the helper plasmid (pTNS), which encodes the TnsABCD proteins (Tn7 transposase). These proteins are responsible for recognizing the Tn7 borders 9 on the delivery vector and facilitating site-specific integration into the attTn7 site. This precise integration occurs 25 nucleotides downstream of the glmS gene, ensuring that no essential functions of the host organism are disrupted. Together, these components allow for accurate and stable insertion of the T7 polymerase gene into the P. putida chromosome.

Figure 16: Components of the Tn7 cloning and integration system

In order to facilitate this process, we needed to construct the pTn-19[g1] plasmid. This can be accomplished by digesting the Golden Standard (GS) plasmid pSEVA23g19[g1] and the vector pTn7-M with the restriction enzymes PacI and SpeI, as described in Salgado et al [13]. The resulting fragments are then ligated together to generate the pTn-19[g1] plasmid, which carries both the Level 1 cargo and the necessary Tn7 borders for chromosomal integration.

Figure 17: Creation of pTN-19[g1]

Build- Test

Unfortunately, due to time constraints, we were unable to proceed with the construction of the pTN-19[g1] plasmid and the subsequent integration of the T7 polymerase into the chromosome.

2nd Engineering Cycle: Production of Outer Membrane Vesicles (OMVs)

The second critical component of developing the THAELIA project involved optimizing the timing of producing the Outer Membrane Vesicles (OMVs). In our design, the production of OMVs is initiated by the eTEV protease, which cleaves the TolB protein at engineered TEV protease recognition sites. We chose to utilize a tolB mutant and introduce a tolB variant containing TEV recognition sites because this particular mutant has demonstrated superior efficiency in producing Outer Membrane Vesicles (OMVs) in P. putida compared to other tol or Pal mutants [14]. To achieve this, we aimed for orthogonal production of eTEV during the stationary phase, allowing for precise control over the timing of protease expression.To achieve this, we designed the orthogonal production of eTEV protease specifically during the stationary phase, enabling precise control over the timing of protease expression. This approach ensures that the cleavage of TolB occurs at the most opportune moment, maximizing OMV production after dsRNA synthesis, allowing the dsRNA to be efficiently encapsulated within the vesicles. By finely tuning the expression of eTEV, we can enhance OMV production and ensure efficient loading of the dsRNA inside them.

Design

Evaluation of Various RBSs (BBa_B0030, and BBa_B0034) for the P3.1 Stationary Promoter

During our Designing phase, we hit a brick wall; ‘How were we going to control the stationary phase production of TEV protease?’ And most importantly; ‘How could we isolate this function from other cellular processes?’

The answer to our questions was found through a thorough literature review, where we encountered the work of Jaishankar et al. (2020) [15]. In their study, they described a synthetic auto-inducible promoter active in the stationary phase, which demonstrated very strong activity. To evaluate this promoter and assess its suitability for our project, we employed sfGFP as a reporter gene, enabling us to visualize and quantify the promoter's activity effectively. Furthermore, we designed experiments to test two different Ribosome Binding Sites (RBSs) to enhance the optimization of the system's performance. We used the promoter J23119 as a positive control because it is a well-characterized constitutive promoter from the Anderson family, allowing for direct comparison of promoter strength and consistent activation throughout all growth phases [16]. Additionally, osmY is a well-characterized promoter specifically induced during the stationary phase, making it a valuable positive control for assessing time-dependent activation and promoter strength during this phase [17].

Figure 18: The constructs we used for P3.1 - RBS_B0034 examination

Figure 19:The constructs we used forP3.1- RBS_B0030 examination

Build

We constructed the Level 0 and Level a constructs as described in Engineering Cycle 1.

Test

After assembling our plasmid constructs, we successfully transformed them into chemical competent E. coli BL21 DE3. The following day, we prepared 10ml liquid cultures of LB medium, with the proper antibiotic, to promote the growth and acclimatization of the transformed bacteria, allowing them to incubate overnight for 16 hours at 37°C and 210 rpm. On the third day, we centrifuged the cultures and washed the resulting pellets twice with 0.8% saline solution (NaCl). Subsequently, we performed a 1:100 dilution to measure the optical density at 600 nm (OD600). To achieve an OD600 of 0.1 with a final volume of 2 mL of M9 culture with the corresponding carbon source, we applied the dilution equation "Cinitial x Vinitial= C final x V final". M9 medium was selected for this experiment due to the fluorescence interference associated with LB medium. We then transferred 200 μL of each diluted culture into the wells of a 96-well plate with a clear bottom. Each construct was tested in five technical repeats. The plate was incubated in a plate reader for 15 hours at 37°C, which is optimal for E. coli while shaking at 180 rpm. Measurements were automatically recorded every hour, monitoring OD at 600 nm and sfGFP fluorescence at 515 nm (emission peak of sfGFP, Roberts et al. )

1. Examining RBS1

Figure 20: Normalized fluorescence intensity for J23119-RBS1-sfGFP-ter, osmy-RBS1-sfGFP-ter, and P3.1--RBS1-sfGFP-ter constructs during the time points: 3h, 6h, 10h and 15h. The RFUs (sfGFP measurements at 515 nm) are divided by cell growth (Optical Density at 600nm), in order to normalize all values. E. coli BL21 (DE3) cells with pDGB3a1 were used as the (-) control. Error bars correspond to standard deviation of n=5 replicates. Blank was subtracted.

2. Examining RBS2

Figure 21: Normalized fluorescence intensity for J23119-RBS2-sfGFP-ter, osmy-RBS2-sfGFP-ter, and P3.1--RBS2-sfGFP-ter constructs during the time points: 3h, 6h, 10h and 15h. The RFUs (sfGFP measurements at 515 nm) are divided by cell growth (Optical Density at 600nm), in order to normalize all values. E. coli BL21 (DE3) cells with pDGB3a1 were used as the (-) control. Error bars correspond to standard deviation of n=5 replicates. Blank was subtracted.

Comparing the performance of P3.1 under the regulation of different RBSs: BBa_B0030 (RBS1) vs BBa_B0034 (RBS2)

Figure 22: Bacterial growth curve, with M9 medium with 0,2% glucose as a carbon source, based on Optical Density at 600 nm measurements taken every 1 hour. Data points represent the average of five technical replicates, with any values showing a standard deviation greater than 0.2 excluded for clarity. Error bars correspond to standard deviation of n=5 replicates. Blank was subtracted.

Figure 23: Normalized fluorescence intensity for P3.1-RBS1-sfGFP-ter, and P3.1--RBS2-sfGFP-ter constructs during the time points: 3h, 6h, 10h and 15h. The RFUs (sfGFP measurements at 515 nm) are divided by cell growth (Optical Density at 600nm), in order to normalize all values. Error bars correspond to standard deviation of n=5 replicates. Blank was subtracted.

Learn

By observing Figure 22 and analyzing the OD600 measurements, we notice significant bacterial growth between the 2h and 6h time points. The steady rise in OD600 during this period indicates that the cells are in the exponential phase.

The results indicate that the P3.1 promoter is significantly stronger than any other promoter tested, exhibiting a continuous increase in expression levels as the bacteria grow. Notably, after 15 hours of incubation, the fluorescence data demonstrate a marked enhancement in sfGFP production, underscoring the P3.1 promoter's effectiveness in driving gene expression during the stationary phase. It was evident that under the regulation of RBS2, the P3.1 promoter drives a higher level of expression. Based on our current data, P3.1 performs well even with RBS1, suggesting that a stronger RBS may not be necessary for optimal system performance. However, further experiments with our actual eTEV production system are needed to determine which RBS provides the best regulation.

Figure: Production of eTEV regulated by P3.1

3rd Engineering Cycle: Synthesis of the dsRNA in vitro and assessment of its silencing efficacy

Project Thaelia was developed to safeguard olive trees against V. dahliae, hence during the engineering process our main objective was to find the best way to achieve it. It was made clear that one of the most important aspects of our design is the successful activation of V. dahliae’s RNAi mechanism. To achieve that, three conditions should be met:

The dsRNA should be produced in sufficient quantities to ensure its availability for uptake by the fungal cells

The fungal cells must possess the ability to efficiently internalize the OMVs that carry the dsRNA molecules

The internalized dsRNA must effectively induce gene silencing within the fungus, leading to a targeted suppression of essential genes.

While the first condition is multifactorial and difficult to prove, we made the first steps by examining basic components of our system. The second condition has already been validated in the literature [18], leaving the efficiency of gene silencing as the main focus—an exciting and open area for exploration in our design. We consulted with PhD candidate Afrodite Katsaouni, who has hands-on experience in RNAi experiments, and Prof. Kalliopi Papadopoulou , expert in plant biotechnology, and they suggested evaluating gene silencing in vitro, as it is a fast and informative first level of evaluation.

1st iteration: Evaluating the effects of dsRNA on V. dahliae

Design

Our first task was to identify the ideal target genes for silencing that play a critical role in fungal growth or virulence. Through extensive literature review we documented genes of V. dahliae involved in various essential functions. We engaged with experts PhD candidate Afrodite Katsaouni and Dr. Athanasios Dalakouras, both of whom have worked with RNAi and V. dahliae, and identified three promising genes: Regulator of G signaling 1 (RGS1), ADP/ATP Carrier (AAC) and Thiamine transport protein (THI20). To validate our selection we consulted with Prof. Aliki Tzima and Prof. Kalliopi Papadopoulou, experts in plant and fungal biology.They confirmed our choice, noting that in literature deletion of these genes exhibit pronounced phenotypic effects, suggesting that they influence both growth and virulence. [19,20,21].

BLAST analysis

Before finalizing our gene selections, we needed to ensure that targeting them would not adversely affect other microorganisms, plants, or humans, nor compromise biodiversity. For this purpose, we performed a blast analysis, looking for homology between these three fungal genes in V. dahliae and human, olive or other microorganisms’ genes.

While some regions of the genes showed homology with genes from other fungal species, no homology was detected with olive tree or human genes, suggesting that targeting them wouldn’t pose a human health risk. At the beginning of our journey, our main focus was to identify sequences of the AAC, RGS1 and THI20 genes that lead to efficient silencing and hinder Verticilllium’s growth. We did not prioritize designing dsRNAs that avoid homology with other organisms but one of our future goals is to improve our system by targeting sequences with higher species-specificity.

Target sequences selection

After selecting our target genes, the next step was to choose what gene regions we would target. We began by searching in the literature for gene regions that were successful targets for RNAi. We also talked with our PI, Prof. Antonis Giakountis, expert in long non-coding mediated transcriptional regulation, who informed us that depending on the RNAi mechanism of the fungus, different gene regions could be more effective targets. If the RNAi mechanism leads to mRNA degradation, a better target would be the 3' UTR, while if it prevents translation, the 5' UTR would be more effective. Since it’s not clear how V. dahliae’s RNAi mechanism works, we decided to target regions confirmed by literature as successful targets for gene silencing [22,23].

Moreover, we consulted with Dr. Athanasios Dalakouras to determine a length that would both fit within the OMVs and be effective for RNAi and we concluded that 200 bp dsRNA molecules would be the ideal choice. So, after taking in consideration the optimal length and the sequences we needed to target, we proceeded on producing the dsRNA.

Figure 24: Target gene regions selected for dsRNA sequence design (1:RGS1, 2:AAC, 3:THI20)

Primer design

To produce the dsRNAs targeting these regions, we needed to design primers with T7 promoter overhangs, as we would be using T7 polymerase for their in vitro synthesis. In our design our bacteria will produce dsRNA 200 bp long, so we tailored our primers to amplify a region of this size. Additionally, we designed primers for RT-qPCR to assess gene expression levels. We designed our primers using the Geneious software with the guidance of PhD candidate Dimitris Rallis who provided us with useful tips on how to design primer pairs suitable for our experiments.

Build

Synthesis of dsRNA Sequences

To evaluate whether we can induce gene silencing in Verticillium dahliae with our designed dsRNAs, we needed to apply these molecules to developing fungal hyphae. The first task was to generate the DNA template required for in vitro transcription of the dsRNA sequences.

To achieve this, we first cultivated V. dahliae on Potato Dextrose Agar (PDA) to obtain hyphal tissue. From the harvested tissue, we extracted total RNA, which was used to synthesize cDNA, providing the essential starting material for the following steps.

As illustrated in Figure 25, we used primers containing T7 promoter overhangs to amplify the target gene sequences (Fig.25-1) and incorporate the T7 promoter sequence necessary for in vitro transcription (Fig.25-2). The initial PCR product was then used as a template for a second round of PCR amplification (Re-PCR) to increase the yield of the desired DNA fragments (Fig.25-3).

After confirming the correct PCR amplification through gel electrophoresis, the DNA bands corresponding to our target sequences were extracted from the gel (Fig.25-4). With the purified DNA template in hand, we proceeded to in vitro transcription using T7 RNA polymerase to synthesize the dsRNA molecules (Fig.25-5,6). This dsRNA was ultimately used for gene silencing experiments in V. dahliae.

Figure 25: Steps for dsRNA production. 1-2) PCR with T7 promoter overhang primers to incorporate T7 sequence into the sequence of interest, 3) Re-PCR to increase yield, 4) Gel electrophoresis and extraction, 5-6) In vitro transcription for dsRNA synthesis.

Test

Monitoring fungal growth

Next, we proceeded to examine whether these dsRNA molecules effectively silenced these key genes, operating under the hypothesis that successful gene silencing could have a visible effect on fungal growth and gene expression. We cultivated V. dahliae in Potato Dextrose Broth (PDB) and isolated its conidia, which are asexual reproductive structures that generate new hyphae [24]. Then, we prepared a 96-well plate, as illustrated in Figure 26, by adding PDB, fungal conidia and dsRNA to each well. The plate was incubated for 10 days, during which we measured OD595 every 24 hours to monitor the growth of V. dahliae.

Figure 26: 96 well plate layout. Row A serves as the blank control, containing only PDB. Row B contains PDB and conidia, acting as our experimental control. Rows C, D, and E are designated for the dsRNA treatment, each containing PDB, conidia, and dsRNA targeting a specific gene.

Analysis of Gene Expression Levels

In order to have a more holistic view of the effects of the dsRNAs to the fungus we performed an RT-qPCR to assess expression levels of the target genes. We expected that the dsRNA would activate the fungal RNAi machinery leading to the silencing of the targeted genes and thus causing reduced expression levels.

We pooled the hyphae from six wells in each row into a single tube, resulting in two technical replicates for each treatment. We then extracted RNA from these samples and synthesized cDNA. Finally, we performed RT-qPCR on one of the technical replicates from each row to assess expression levels of RGS1, AAC and THI20.

Learn

The OD595 measurements (Fig. 27,28) indicated that there was little variation in fungal growth across all conditions. This limited response could be attributed to several factors:

• Given the small volume of dsRNA used, inaccuracies in pipetting may have resulted in insufficient dsRNA being added

• It may be necessary to optimize the dsRNA concentration to achieve a more pronounced effect

• The dsRNA sequences may not have been ideal for activating the RNAi mechanism of the fungus

• These gene regions may not be ideal targets for impacting fungal growth

• Silencing these genes may have no effect in fungal growth but in virulence

• This experiment may not be suitable to observe the effects of gene silencing in the fungus. In this case we would also need in planta experiments to acquire more prominent results.

Figure 27: Monitoring of fungal growth in the span of 10 days. The graph illustrates fungal growth in every condition.

Figure 28: Growth curves for V. dahliae in each treatment. From the slope we can see that in all conditions the fungus develops normally.

To have a glimpse of what’s happening inside the fungus, we attempted to assess gene expression with RT-qPCR. Unfortunately, due to our inexperience and the sensitivity of the technique we weren’t able to produce reliable data, while the limited time prevented us from repeating the experiment. We observed variability in Cq values between our technical replicates, which could be the result from pipetting errors or even from degradation of RNA before cDNA synthesis. In the future we want to repeat the experiment to obtain more accurate results.

To be able to draw conclusions about the effect of these dsRNA sequences we would need to repeat both the plate reader assay and the RT-qPCR. Due to limited time we couldn’t repeat these experiments but we hope future iGEM teams will further develop this project.

Proposed redesign

While we couldn’t see the genes’ expression levels due to unreliable data, we also didn’t see any significant difference in fungal growth. To address this in the future we will utilize bioinformatic tools to predict siRNAs with a high likelihood of effective gene silencing. We will design 200 bp dsRNA sequences that incorporate these predicted siRNAs to increase the chances of successful gene silencing. Such tools are siDirect and BLOCK-iT™ RNAi Designer from Thermo Fisher Scientific. Both tools are suitable to design siRNAs [25] for gene silencing, since they use algorithms to check for knockdown effectiveness and thermodynamic stability. For the efficacy they use empirically based rules that rely on the fact that the knockdown efficiencies of siRNAs are dependent on their sequences [26]. Using tools like that will allow us to test sequences that have theoretically better chances for inducing gene silencing. We will also be able to compare our results taking us a step closer to understanding the RNAi mechanism of V. dahliae.

Additional planned DBTLs

The first cycle didn’t result in conclusive data, however, we identified flaws in our design and generated new questions that necessitated further investigation.To address them, we discussed with Dr. Athanasios Dalakouras, who made us realize the need to enhance our experimental approach. He recommended that we target multiple genes simultaneously to determine whether a multifaceted attack on the fungus would have a more pronounced effect on fungal growth and gene expression. Accordingly, we devised a plan for two additional DBTL cycles, which will progressively provide us with insights into the effects of dsRNA on V. dahliae.

2nd iteration: Evaluating silencing when targeting two genes at a time

In this cycle we’ll examine the effect that dsRNA has on the fungus when targeting two genes at a time. Will fungal growth be more affected than when targeting one gene? Would we observe any change in gene expression?

We will apply dsRNA to fungal conidia, targeting two genes simultaneously and by measuring OD595 and performing RT-qPCR we would evaluate its effects in hyphal growth and gene expression. Then, we will compare the results between the two cycles to determine whether the effects of dsRNA treatment are enhanced with this approach.

3rd iteration: Evaluating silencing when targeting all three genes at the same time

In this cycle, we will investigate the effect of dsRNA on the fungus by targeting all three genes simultaneously. Our objective again is to evaluate whether applying greater pressure to the fungus through dsRNA treatment leads to a more significant impact on its growth and gene expression.

We will apply dsRNA to fungal conidia, targeting two genes simultaneously and by measuring OD595 and performing RT-qPCR we would evaluate its effects in hyphal growth and gene expression. Then, we will compare the results with the two previous cycles to determine whether the effects of dsRNA treatment are enhanced with this approach.

Conclusion

In conclusion, utilizing the D-B-T-L cycle throughout various aspects of our project has proven to be highly beneficial, allowing us to systematically explore and enhance each component. We gained valuable insights into the evaluation of the T7 polymerase system, the synthesis of dsRNA in vitro and its subsequent assessment for silencing efficacy. This modular approach not only facilitated our focus on each individual element but also significantly improved the overall design and functionality of our final construct.

References

[1] Sarrion-Perdigones A, Vazquez-Vilar M, Palací J, Castelijns B, Forment J, Ziarsolo P, Blanca J, Granell A, Orzaez D. GoldenBraid 2.0: a comprehensive DNA assembly framework for plant synthetic biology. Plant Physiol. 2013 Jul;162(3):1618-31. doi: 10.1104/pp.113.217661. PMID: 23669743; PMCID: PMC3707536.

[2] Pingoud, A., & Jeltsch, A. (2001). Structure and function of type II restriction endonucleases. Nucleic acids research, 29(18), 3705–3727.

[3] Pei, L., Garfinkel, M., & Schmidt, M. (2022). Bottlenecks and opportunities for synthetic biology biosafety standards. Nature communications, 13(1), 2175. https://doi.org/10.1038/s41467-022-29889-y

[4] Müller, K. M., & Arndt, K. M. (2012). Standardization in synthetic biology. Methods in molecular biology (Clifton, N.J.), 813, 23–43. https://doi.org/10.1007/978-1-61779-412-4_2

[5] Zobel, S., Benedetti, I., Eisenbach, L., de Lorenzo, V., Wierckx, N., & Blank, L. M. (2015). Tn7-Based Device for Calibrated Heterologous Gene Expression in Pseudomonas putida. ACS synthetic biology, 4(12), 1341–1351. https://doi.org/10.1021/acssynbio.5b00058

[6] Taş, H. (2020). Upgrading Pseudomonas putida as a Synthetic Biology chassis through inter-operativity of genetic devices [Doctoral dissertation, Universidad Autónoma de Madrid].

[7] Blázquez, B., et all. (2023). Golden Standard: a complete standard, portable, and intraoperative MoClo tool for model and non-model proteobacteria. In Nucleic Acids Research (Vol. 51, Issue 19, pp. e98–e98). Oxford University Press (OUP). https://doi.org/10.1093/nar/gkad758 )

[8] Mirkin, E. V., & Mirkin, S. M. (2005). Mechanisms of transcription-replication collisions in bacteria. Molecular and cellular biology, 25(3), 888–895.

[9] Martin-Pascual, M., Batianis, C., Bruinsma, L., Asin-Garcia, E., Garcia-Morales, L., Weusthuis, R. A., van Kranenburg, R., & Martins dos Santos, V. A. P. (2021). A navigation guide of synthetic biology tools for Pseudomonas putida. In Biotechnology Advances (Vol. 49, p. 107732). Elsevier BV.

[10] Yan, Q., & Fong, S. S. (2017). Study of in vitro transcriptional binding effects and noise using constitutive promoters combined with UP element sequences in Escherichia coli. Journal of biological engineering, 11, 33.

[11] Yim HH, Brems RL, Villarejo M. Molecular characterization of the promoter of osmY, an rpoS-dependent gene. J Bacteriol. 1994 Jan;176(1):100-7. doi: 10.1128/jb.176.1.100-107.1994. PMID: 8282684; PMCID: PMC205019.

[12] Sarrion-Perdigones A, Vazquez-Vilar M, Palací J, Castelijns B, Forment J, Ziarsolo P, Blanca J, Granell A, Orzaez D. GoldenBraid 2.0: a comprehensive DNA assembly framework for plant synthetic biology. Plant Physiol. 2013 Jul;162(3):1618-31. doi: 10.1104/pp.113.217661. PMID: 23669743; PMCID: PMC3707536.

[13] Salgado, S., Hernández‐Herreros, N., & Prieto, M. A. (2024). Controlling the expression of heterologous genes in Bdellovibrio bacteriovorus using synthetic biology strategies. In Microbial Biotechnology (Vol. 17, Issue 6). Wiley.

[14] JLlamas, M. A., Ramos, J. L., & Rodríguez-Herva, J. J. (2000). Mutations in each of the tol genes of Pseudomonas putida reveal that they are critical for maintenance of outer membrane stability. Journal of bacteriology, 182(17), 4764–4772. https://doi.org/10.1128/JB.182.17.4764-4772.2000

[15] Jaishankar J, Srivastava P. Strong synthetic stationary phase promoter-based gene expression system for Escherichia coli. Plasmid. 2020 May;109:102491. doi: 10.1016/j.plasmid.2020.102491. Epub 2020 Feb 5. PMID: 32035079.

[16] Martínez-García, E., Goñi-Moreno, A., Bartley, B., McLaughlin, J., Sánchez-Sampedro, L., del Pozo, H. P., Hernández, C. P., Marletta, A. S., De Lucrezia, D., Sánchez-Fernández, G., Fraile, S., & de Lorenzo, V. (2020). SEVA 3.0: an update of the Standard European Vector Architecture for enabling portability of genetic constructs among diverse bacterial hosts. In Nucleic Acids Research (Vol. 48, Issue 6, pp. 3395–3395). Oxford University Press (OUP). https://doi.org/10.1093/nar/gkaa114

[17] Carbonelli, D. L., Corley, E., Seigelchifer, M., & Zorzópulos, J. (1999). A plasmid vector for isolation of strong promoters inEscherichia coli. In FEMS Microbiology Letters (Vol. 177, Issue 1, pp. 75–82). Oxford University Press (OUP).

[18] Cai, Q., He, B., Wang, S., Fletcher, S., Niu, D., Mitter, N., … Jin, H. (2021). Message in a bubble: Shuttling small RNAs and proteins between cells and interacting organisms using extracellular vesicles. Annual Review of Plant Biology, 72(1), 497–524. doi:10.1146/annurev-arplant-081720-010616

[19]Xu, J., Wang, X., Li, Y., Zeng, J., Wang, G., Deng, C., & Guo, W. (2018). Host-induced gene silencing of a regulator of G protein signalling gene (VdRGS1) confers resistance to Verticillium wilt in cotton. Plant Biotechnology Journal, 16(9), 1629–1643. doi:10.1111/pbi.12900

[20] Su, X., Rehman, L., Guo, H., Li, X., Zhang, R., & Cheng, H. (2017). AAC as a Potential Target Gene to Control Verticillium dahliae. Genes, 8(1), 25. doi:10.3390/genes8010025

[21] Qin, T., Hao, W., Sun, R., Li, Y., Wang, Y., Wei, C., … Wang, Q. (2020). Verticillium dahliae VdTHI20, involved in pyrimidine biosynthesis, is required for DNA repair functions and pathogenicity. International Journal of Molecular Sciences, 21(4), 1378. doi:10.3390/ijms21041378

[22] Ayabe, S., Kimura, Y., Umei, N., Takikawa, Y., Kakutani, K., Matsuda, Y., & Nonomura, T. (2022). Real-time collection of conidia released from living single colonies of Podosphaera aphanis on strawberry leaves under natural conditions with electrostatic techniques. Plants, 11(24), 3453. doi:10.3390/plants11243453

[23] Dang, Y., Yang, Q., Xue, Z., & Liu, Y. (2011). RNA interference in fungi: pathways, functions, and applications. Eukaryotic Cell, 10(9), 1148–1155. doi:10.1128/EC.05109-11

[24] Naito, Y., & Ui-Tei, K. (2012). SiRNA design software for a target gene-specific RNA interference. Frontiers in Genetics, 3, 102. doi:10.3389/fgene.2012.00102