A SynBio-based SARS-CoV-2 therapeutic solution: orchestrating inducible interference using virus-like particles, toehold riboswitches and siRNAs.

Presented by Team AUC-EGYPT 2020

Ahmed Ghouneimy1, Rana Salah1 , Salma Abou Elhassan1, Abd-Elrhman Elsadany1, Sawsan Ali1, Norhan Abdeldayem1, Dalia El-Naggar1, Mohamed Hassan1, Abdurrahman Radwan1, Yousef Eldemerdash1, Mohammad Monir1, Kareem Kassab1 and Mina Ashraf1, Hana Abdelzaher2, Mahmoud Abdelgawad3, Mohammed AlFeky3, and Anwar Abdel Naser4.

1iGEM Student Team Member, 2iGEM Team Instructor, 3iGEM Team Advisors, 4iGEM Team Primary PI.

The COVID-19 pandemic is without doubt the biggest health crisis of the 21st century. Currently, there is no specific treatment or vaccine for the disease. In this phase-I project, we propose a SynBio-based therapeutic solution for COVID-19. Our approach consists of a SARS-CoV-2-like particle that is capable of targeting cells expressing the ACE2 receptor. Upon delivery, our sensing moiety (Toehold Riboswitch adjusted for mammalian systems) is constitutively expressed. If SARS-CoV-2 triggers are present within the cell, the sensor will unfold initiating a GAL4-VP16-dependent interference pathway. Consequently, we designed two de novo siRNAs that play a critical role in our antiviral therapeutic approach. They inhibit viral replication via RISC-mediated degradation of the replicase region of SARS-CoV-2 mRNA. We also created a deterministic model to predict the levels of siRNAs as well as a structural model to predict the thermodynamic stability of our toehold riboswitches.


Coronavirus disease 2019 (COVID-19) is caused by a novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). It is classified as a world health crisis and a fast-growing global pandemic. According to WHO (2020), the number of confirmed cases worldwide has exceeded 47 million so far, and it killed over 1.2 million individuals.

COVID-19 Choropleth Map, for the confirmed cases and deaths globally, reported to WHO.

The virus is known for its rapid cell entry and replication as it attaches its spike (S) protein to angiotensin-converting enzyme 2 (ACE2) receptors that are located on the surface of many human cells, including those in the lungs permitting the virus entry.

Following the virus RNA translation, replicase proteins will be generated from open reading frame 1a/b (ORF 1a/b), using the genome as a template to produce more RNAs and speed up the virus replication.

Despite how fast medical and pharmaceutical technologies are improving, there is currently still no cure for COVID-19, and vaccines are not yet available. However, some medications are prescribed to relieve the severity of the symptoms experienced by patients.


★ By the time we had our brainstorming, the COVID-19 pandemic had just started, and our university (AUC) was the leading university in Egypt to emphasize the importance of quarantine.

★ Moreover, some members of our team have lost their close ones due to the pandemic. Since then, we have realized how dangerous the novel coronavirus is, and the necessity of not downplaying it.

★ We wanted our project to have a maximal and a global impact in addition to its availability to every individual. Since we were concerned about the looming threat of the SARS-CoV-2 rapid spread and infection among people, we thought about uprooting the virus and inhibiting its replication and thereby its invasion to neighboring cells.

Our Solution

Our project proposes a synthetic biology approach for COVID-19 treatment through a designed circuit to control the expression of the hybrid transcriptional activator GAL4-VP16, using a toehold switch that allows the translation of a downstream mRNA upon base pairing with a complementary mRNA trigger (Green et al., 2014). The switch gets activated upon sensing SARS-CoV-2 mRNA, inducing the expression of GAL4-VP16 which in turn activates the siRNA expression through binding to its upstream UAS promoter.

Thus, we designed 2 de novo double-stranded siRNAs targeting the viral mRNA coding for replicase proteins. Those 2 siRNA sequences will bind to their complementary sequences on the ORF1ab gene encoding for replicase proteins. Then, a Dicer enzyme will come to cut these complementary sequences and thus knocking out ORF1ab gene.

Our Proposed Genetic Circuit.

Having our circuit completed, it will be delivered to the targeted cells through a virus-like particle system (VLPs) which can mimic the cell entry of the actual virus (Naskalska et al., 2018). Here, we will produce baculovirus-mediated lentivirus vectors, which will be modified by changing them to a SARS-CoV-2 Pseudo-Virus, allowing the specific targeting of the cells expressing the ACE2 receptor. Consequently, through this delivery system, our designed sensing-interfering circuit will be successfully transduced into infected cells expressing the ACE2 receptor

The Mechanism of our Solution

Toehold Engineering
A new class of toehold switches

Toehold switches were originally designed to function in prokaryotic and cell free expression systems (Green et al., 2014; Pardee et al., 2016). The loop of the hairpin featured a prokaryotic RBS and the bulge in the middle of the stem domain featured the start codon. In Eukaryotes, the RBS, typically the Kozak consensus sequence, is not separate from the start codon. Wang et al., (2019) redesigned toehold switches by replacing the original RBS with a Kozak sequence that proved to be functional but with a low dynamic range. We hypothesize we can enhance the dynamic range of these toehold switches by following the series B design scheme (Fig 1) implemented by Pardee et al., (2016) which proved to be more efficient in prokaryotic systems.

Figure 1 Toehold Switches

Regions ‘a’ and ‘b’ are complementary to the trigger. Toeholds A and B are prokaryotic toehold switches with the RBS being separate from the start codon. B differs by the addition of a conserved domain of weak base pairs just below the RBS loop. C and D are mammalian toehold switches with the start codon being just downstream the Kozak sequence. D differs by the addition of a conserved domain of weak base pairs just below the Kozak loop.

Proposed Toehold Switch Characterization

The S2 domain is conserved among Sars-Cov-2 strains and, hence, is a good candidate for therapeutic development(Kaushal et al., 2020). All generated Toehold Switches will be characterized by measuring their dynamic range (ON/OFF ratio) (Fig 2).

Proposed ToeholdSwitch—GAL4BD-VP16—UAS (TGU) device characterization

Toehold switches that displayed the highest dynamic range will be ligated to GAL4BD-VP16 and the downstream promoter UAS will be fused to GFP.

siRNA Engineering

To test the efficacy of our siRNA, 2 vectors will be constructed at first: one for each siRNA sequence. Each vector will have the siRNA sequence, targeting the ORF1ab gene, attached to the downstream green fluorescent protein (GFP) reporter gene. One more vector containing the ORF1ab gene will be constructed to trigger siRNA as we do not have a biosafety level 3. The efficiency of knockdown by each candidate siRNA will be verified through using qPCR.

First Vector to Test the
1st siRNA (siRNA-R1)

Second Vector to Test the
2nd siRNA (siRNA-R2)

ORF1ab Plasmid

After verifying the efficiency of our candidate sequences, we will integrate the 2 siRNA sequences into one plasmid subsequently, followed by the GFP gene. By this, we anticipate our siRNA to silence SARS-CoV-2 replicase mRNA and thus decrease the number of those transcripts by the virus, inhibiting virus replication.

The Final Verified Vector to test our 2 siRNA Candidate Sequences (siRNA-R1 & siRNA-R2)

VLP Engineering

Analyzing prior delivery systems, we selected specific targeting, safety, high production, efficient cell entry and convenient loading capacity to be our design requirements. The engineering process included four iteration cycles involving SWOT qualitative analysis for evaluation. The primary design was decoy SARS-CoV-2 particles able to specifically target cells expressing the ACE2 receptor (Xu et al., 2020), however, they were poorly-experimented unsafe loading vector Accordingly, in the second iteration cycle, these particles were replaced by third-generation self-inactivating lentivirus vectors to have safer non-replicating vectors (Milone & O’Doherty, 2018). Again, for safety, the minimal needed genes to produce functional vectors were identified and separated into four plasmids (Crawford et al., 2020; Lesch et al., 2008). In the third iteration, we decided to have baculovirus-mediated lentivirus vectors’ production to magnify our production rates in the implementation phase (Lesch et al., 2008). Starting to design our plasmids, two of them, shown in figure (1), were responsible for the vectors’ packing coding for a transcript of gag, pol & RRE genes, and Rev gene respectively (Lesch et al., 2008).

Figure(1): the packaging plasmids.

In the fourth iteration cycle, we decided to pseudo-type our vectors to express the SARS-CoV-2 spike to maintain the specific targeting ability of our vectors (Crawford et al., 2020). Our third plasmid is shown in figure (2) with the three variations we designed for the spike gene to test their effect on pseudo-typing and receptor binding efficiency. In this manner, we successfully designed a modular delivery system as future teams can edit only on the third plasmid to target different cells.

Figure (2): the variants of the SARS-CoV-2 spike gene.

Finally, our fourth plasmid, shown in figure (3), will be responsible for loading our construct of interest into the vectors. The GFP gene will be used for testing the transduction and expression efficacy of the vectors (Lesch et al., 2008). In all plasmids, the newly designed composite promoter (polh-pSel) was used to enhance the DsRed2 gene expression offering higher quality results in our testing phase (Martínez-Solís et al., 2016).

Figure (3): the transfer plasmid.

Access Github code Here


We troubleshot and improved Toeholder, a software designed by Ulaval iGEM 2019. Toeholder 2.0 can generate series B-like mammalian toehold switches (Figure 1).

Figure 1 Workflow of Toeholder 2.0

We improved the processing time of Toeholder by redefining the trigger region. Briefly, instead of passing the whole input sequence as a trigger region, the sequence is parsed, and only a region of 200 nucleotides spanning the 30-nucleotide trigger is passed to NUPACK minimal free energy (mfe) function(Zadeh, Steenberg, et al., 2011). This assumes that nucleotides in close proximity will more likely interact with each other. Similar approaches are utilized to determine the secondary structure of conserved regions in SARS-CoV-2 genome(Huston et al., 2020; Rangan et al., 2020). This step was of complexity O(N3), where N is the length of the sequence (Zadeh, Wolfe, et al., 2011). Now that we are passing a trigger region of constant length, the complexity becomes O(1). In other words, irrespective of the size of the input sequence, the processing time of this step is constant.

We compared the processing time of Toeholder and Toeholder 2.0 by doing a benchmarking experiment (Fig 2).

Figure 2 Benchmarking
Toeholder did not accept files beyond 1 kb, while Toeholder 2.0 accepted larger files. The maximum file size we tested was 30 kb (SARS-CoV-2 Genome), and it took 80 minutes to process.


In the original publication, Pardee et al., (2016) hypothesized that the more accessible the trigger region is, the better the toehold is going to bind. We modeled the relationship between the number of accessible positions and the free energy of the bound state (ΔGbound). Remarkably, there was no correlation between the two variables (Fig 1).

Figure 1 Correlation between accessible positions and free energy of binding In this sample of 10827 toehold switches, there seems to be no correlation between the number of accessible (unpaired) nucleotides and the binding energy of the toehold to the trigger. There was no restriction on the minimum number of unpaired positions when designing the toehold switches.

Reflections from the structural model

The linear regression model predicted no correlation between the number of unpaired bases and the free energy of binding between the toehold switch and the trigger. Accordingly, we redesigned our toehold switches by eliminating the criteria for unpaired bases.

In silico Thermodynamic Modeling

The generated toehold switches were ranked based on five criteria:

1) Minimal Free Energy (MFE) Difference:

MFE = ΔGbound - (ΔGtoehold + ΔGtrigger)

MFE < 0

2) Free Energy of the toehold switch:

ΔGtoehold< 0

3) ΔGRBS-Linker: The RBS-Linker is the sequence of nucleotides starting from the RBS loop to the end of the 21-nucleotide linker. Green et al., (2014) inferred a strong correlation between the ΔGRBS_Linker and the dynamic range:

ΔGRBS-Linker ~ 0

4) Perfect Matching: from the structural model of the bound state predicted by NUPACK (Zadeh et al., 2011), we can draw conclusions about the base matching between the toehold and the trigger.

5) Orthogonality to the human genome: Ideally, Toehold switches feature a stable RBS loop flanked by a stable stem. Figure 2 displays one the structure predicted by NUPACK.

Figure 4 The structure of one toehold switch featuring the Kozac loop atop of a stable stem and a free toehold domain (Unpaired bases to the left).


In our project, we specified three core values: integrity, opportunity, and equity. As integrated human practices, the social science expert, Dr. Hoda Rashad, highlighted the integrity in abiding by the “no harm” concept, and the equity in facilitating the accessibility for our final product. Also, the health expert, Dr. Rami Gharieb, highlighted our project’s dual therapeutic and diagnostic potential. Additionally, the Pylon CMP team, which developed a platform for healthcare data centralization, helped us identify the firmographic and behavioral characteristics of our market segments.

Following our core values, we conducted sessions to underprivileged children mediated by the charitable organization, Volunteers in Action, in which we raised awareness about COVID-19 and explained the basic virus structure through an interactive drawing contest. Also, believing in the value of education, we conducted an online webinar session about biosafety and biosecurity where 86 students from different STEM high schools in Egypt attended. It was a TOT workshop that introduced the dual-use dilemma in synthetic biology and possible strategies for both crisis prevention and management.

Communicating our project, we gave a virtual talk in Hult prize National camp where 30 representatives from 23 private and public universities in Egypt attended. We presented the beneficial applications of GMOs in different sectors and got feedbacks about our business model and marketing plans. Additionally, we had an interview with Molecular Cloud company. It was a great chance to propose our idea to a scientific community and get useful technical feedback on possible improvements.

For our collaborations, we started two initiatives. First, the “SDGs from home” initiative that shows the capability of working on sustainable development through our daily routines. Second, the “Thank your mentor” initiative where we spread the importance of gratitude. Many iGEM teams participated in our collaborations as TU Delft, NCKU Tainan, Thessaly, IIT Roorkee and AFCM Egypt. Also, we have attended two meetups: the MIT mammalian meetup, where we won their graphical abstract contest judged by MIT BE communication lab, and the first African Meetup, where we presented our project and got fruitful feedback.


One of our missions was raising this generation knowing the importance of science and eager to contribute to it. After research, we concluded that the best interactive way of science communication is game-based learning.
Play the game here.

The game starts with the basics of biology like levels of organization then moves on to more complex topics like Transcription and Translation.

Afterwards, the player is introduced to the topic of Recombinant DNA. Finally, the player has to use all what was learned to construct a circuit to make the little firefly glow.

Business Model

As a B2B business model, our primary potential customers are the Egyptian military hospitals as they market themselves as quality sensitive. Also, there is no evidence of any direct competition in the Egyptian which sets the scene for the release of our product. Furthermore, the most relevant stakeholders are universities, research centers, the Egyptian drug authority, and physicians working in military hospitals (Fig 1).

Figure 1 Fig 1 Importance of Stakeholders

Moreover, our milestones involve three stages during 3 years: proof of concept and animal testing for 8 months and funded by a research grant, clinical trials for 18 months and funded by seed funding, and finally proof of market and partnership for 6 month (Fig 2). Our main product's impact is our circuit design and delivery method can serve for a potential future expansion for our bio-venture or as an alternative revenue stream in case of market stagnation as they can be utilized for diagnostics and therapeutics of many other diseases.


References and Acknowledgements

★ Crawford, K. H. D., Eguia, R., Dingens, A. S., Loes, A. N., Malone, K. D., Wolf, C. R., Chu, H. Y., Tortorici, M. A., Veesler, D., Murphy, M., Pettie, D., King, N. P., Balazs, A. B., & Bloom, J. D. (2020). Protocol and Reagents for Pseudotyping Lentiviral Particles with SARS-CoV-2 Spike Protein for Neutralization Assays. Viruses, 12(5), 513.

★ Green, A. A., Silver, P. A., Collins, J. J., & Yin, P. (2014). Toehold Switches: De-Novo-Designed Regulators of Gene Expression. Cell, 159(4), 925–939.

★ Huston, N. C., Wan, H., de Cesaris Araujo Tavares, R., Wilen, C., & Pyle, A. M. (2020). Comprehensive in-vivo secondary structure of the SARS-CoV-2 genome reveals novel regulatory motifs and mechanisms. BioRxiv.

★ Kaushal, N., Gupta, Y., Goyal, M., Khaiboullina, S. F., Baranwal, M., & Verma, S. C. (2020). Mutational Frequencies of SARS-CoV-2 Genome during the Beginning Months of the Outbreak in USA. Pathogens, 9(7), 565.

★ Lesch, H. P., Turpeinen, S., Niskanen, E. A., Mähönen, A. J., Airenne, K. J., & Ylä-Herttuala, S. (2008). Generation of lentivirus vectors using recombinant baculoviruses. Gene Therapy, 15(18), 1280–1286.

★ Martínez-Solís, M., Gómez-Sebastián, S., Escribano, J. M., Jakubowska, A. K., & Herrero, S. (2016). A novel baculovirus-derived promoter with high activity in the baculovirus expression system. PeerJ, 2016(6), e2183.

★ Milone, M. C., & O’Doherty, U. (2018). Clinical use of lentiviral vectors. Leukemia, 32(7), 1529–1541.

★ Naskalska, A., Dabrowska, A., Nowak, P., Szczepanski, A., Jasik, K., Milewska, A., Ochman, M., Zeglen, S., Rajfur, Z., & Pyrc, K. (2018). Novel coronavirus-like particles targeting cells lining the respiratory tract. PLoS ONE, 13(9).

https:// doi: 10.1371/journal.pone.0203489

★ Pardee, K., Green, A. A., Takahashi, M. K., Braff, D., Lambert, G., Lee, J. W., Ferrante, T., Ma, D., Donghia, N., Fan, M., Daringer, N. M., Bosch, I., Dudley, D. M., O’Connor, D. H., Gehrke, L., & Collins, J. J. (2016). Rapid, Low-Cost Detection of Zika Virus Using Programmable Biomolecular Components. Cell, 165(5), 1255–1266.

★ Rangan, R., Zheludev, I. N., & Das, R. (2020). RNA genome conservation and secondary structure in SARS-CoV-2 and SARS-related viruses [Preprint]. Genetics.

★Wang, S., Emery, N. J., & Liu, A. P. (2019). A Novel Synthetic Toehold Switch for MicroRNA Detection in Mammalian Cells. ACS Synthetic Biology, 8(5), 1079–1088.

WHO Coronavirus Disease (COVID-19) Dashboard. Retrieved from

★ Xu, R., Shi, M., Li, J., Song, P., & Li, N. (2020). Construction of SARS-CoV-2 Virus-Like Particles by Mammalian Expression System. In Frontiers in Bioengineering and Biotechnology(Vol. 8, p. 862).

★ Zadeh, J. N., Steenberg, C. D., Bois, J. S., Wolfe, B. R., Pierce, M. B., Khan, A. R., Dirks, R. M., & Pierce, N. A. (2011). NUPACK: Analysis and design of nucleic acid systems. Journal of Computational Chemistry, 32(1), 170–173.

★ Zadeh, J. N., Wolfe, B. R., & Pierce, N. A. (2011). Nucleic acid sequence design via efficient ensemble defect optimization. Journal of Computational Chemistry, 32(3), 439–452.