We thought we should first understand the existing avian influenza countermeasures and identify areas that synthetic biology should solve. We initially hypothesized that the problems were the weak virus resistance of the chickens themselves, the failure to prevent the virus from entering the chicken houses, and inadequate monitoring of unhealthy chickens.

Options

In Japan, HPAI prevention is guided by the administration. Therefore, we believed that interviewing the administrative body for domestic animal infectious diseases would be the most efficient way to grasp the on-site needs.

Strengthening the virus resistance of chickens

• Vaccines are a method for strengthening HPAIV resistance in chickens, but they have many drawbacks, and an effective technique is not yet available.
• Disinfection and hygiene maintenance around chicken houses are thoroughly implemented at a very high level, so the need to strengthen existing countermeasures using synthetic biology is low.
• Health monitoring is performed daily by visual inspection, and blood collection is also carried out periodically. We felt that the existing countermeasures are sufficient.

Initially, we considered constructing a further defense by using other immune actions besides cell death-induced immunity. As an additional defense, we considered using humoral immunity through antibodies.

To recognize the challenges in avian influenza virus prevention in the poultry industry, we interviewed the administrative body for domestic animal infectious diseases.

We decided not to implement additional infection defense measures and to focus on improving infection defense by cell death.

• If HPAIV antibodies are expressed and produced for infection defense, the infection history of HPAI cannot be determined by the presence or absence of antibodies, and food safety cannot be guaranteed.

We considered that limiting the expression site to non-edible parts could enhance the safety of the GM food.

To determine the necessary degree of virus infection defense against HPAI, we interviewed a researcher involved in avian influenza virus research.

Express in all chicken cells.

• HPAI has extremely strong infectivity and virulence, causing the virus to multiply throughout the body. Therefore, limiting the expression site could greatly compromise the viral infection defense.

Even if we create chickens with high HPAI resistance, the possibility of social implementation is low if they are not accepted by consumers. Therefore, we needed to know how to make the GM food safe and easily accepted. One safety assurance method we initially considered was using only genes from closely related species.

This decision required listening to the actual voices of consumers. Therefore, we attended a lecture by a civic group researching the safety of GM food and investigated what kind of GM foods they held opposing views on.

Limit the introduced gene to proteins found in chickens and mallards. Furthermore, limit it to those constantly expressed in cells.

• When introducing and expressing a protein not present in closely related chicken species, even if the protein itself is safe, its interaction with other native proteins is unknown, making it difficult to guarantee safety.
• Even chicken-derived proteins may have an impact if their expression timing or location is changed.
• Introducing genes taken from sources other than chickens or mallards is likely to provoke strong consumer backlash.

To induce a rapid defense response in animal cells that is dependent on viral infection, accurately sensing the viral infection stimulus was essential. As a sensing method, we considered using parts of the influenza virus that are unlikely to mutate (localized dsRNA regions of the genomic RNA, NP, Polymerase) as infection indicator substances.

To understand the life cycle of the influenza virus and the virus-specific substances produced during the process, we interviewed researchers of the influenza virus.

Sense the localized double-stranded RNA (dsRNA) region of the genomic RNA formed inside the cell upon influenza virus infection.

• Many immune factors recognize dsRNA in a structure-specific manner, making it easy to use as an infection stimulus.
• Since it recognizes a double-stranded structure, it functions even if the base sequence mutates.
• Previous research used the same method [2].
• NP was a useful candidate, but we decided not to adopt it for reasons such as the difficulty of proving the concept at the iGEM level of experimentation.
• It was found that the function of the influenza virus receptor is largely unknown, making it difficult to utilize.

dsRNA exists not only virus-dependently but also endogenously in cells (miRNA), etc. Therefore, it was necessary to use a sensor that could specifically detect viral dsRNA. An additional condition was that it should be easily connectable to the apoptosis pathway after dsRNA detection. Thus, we considered proteins that trigger synthesizable biological actions, such as dsRNA-dependent polymerization (PKR and RIG-I), as candidates.

To determine whether RIG-I, which was a candidate dsRNA sensor, could actually be used, we sought opinions from an expert on RLR (RIG-I Like Receptor).

Use PKR and RIG-I as dsRNA sensors.

• PKR has a track record as a dsRNA sensor, as it was used in a previous study [1].
• RIG-I preferentially recognizes the localized double-stranded RNA region of the influenza virus genome RNA. It also has a high ability to distinguish between endogenous dsRNA and viral dsRNA, making it easy to handle. Additionally, since it polymerizes upon dsRNA binding, it is easy to connect to the downstream apoptosis pathway.
• We judged that LGP2 is difficult to handle as a sensor because many parts of its function in the cell are unknown.
• MDA5 preferentially recognizes longer dsRNA than the influenza virus, making it inferior to RIG-I for the current application.
• NLR and TLR3 were not adopted because connecting them to the downstream apoptosis pathway was judged to be difficult.

In this infection defense system, rapid cell death needed to occur while suppressing virus outflow and without affecting surrounding uninfected cells. Therefore, we needed to select the optimal one among the three major types of cell death.

We needed to thoroughly understand the influenza virus, its budding, and its spread, so we interviewed a researcher of the influenza virus.

Apoptosis.

• Apoptosis is a relatively quiet form of cell death, so it is predicted to have less impact on surrounding uninfected cells.
• We judged that the spread of replicating viruses inside the cell during apoptosis is unlikely, so its performance as an infection defense would also be guaranteed.

We needed to consider how to connect to apoptosis as quickly as possible when the dsRNA sensor recognizes the dsRNA region. Therefore, we set proteins that induce apoptosis when polymerized as candidates.

We attempted to interview an expert on apoptosis, but we obtained a good conclusion from the consultation with the expert already conducted, so we did not conduct additional research.

Fuse CARD of APAF1 and ΔCaspase9 to the dsRNA sensor.

• CARD of APAF1 and ΔCaspase9 are proteins that can induce apoptosis when polymerized, making them compatible with PKR and RIG-I, which polymerize upon dsRNA binding.
• Since the dsRNA sensor polymerizes in a dsRNA-dependent manner, and these Apoptosis Inducers also polymerize at that time, rapid apoptosis induction without transcription and translation is possible.
• Split proteins such as split-Caspase have self-polymerization activity, and the risk of toxic leakage is large, so they were not adopted.
• Toxins such as diphtheria toxin were also considered, but were not adopted because their toxicity was too strong.

Although the system is designed so that apoptosis occurs dsRNA-dependently upon HPAIV infection, unintended apoptosis may occur even in the absence of dsRNA. To further enhance the utility of COCCO, we needed to devise a way to reduce this unintended apoptosis. We assumed that unintended apoptosis is likely to occur when COCCO is highly expressed, so we thought that reducing the expression level might be the solution.

We needed to know how much self-polymerization capacity the planned dsRNA sensor and Apoptosis Inducer have, and under what conditions dsRNA-independent apoptosis might occur. Therefore, we sought opinions from a RIG-I expert and an influenza virus expert.

Suppress cell death induction even if RIG-I polymerizes by not exposing Apoptosis Inducers such as CARD in the absence of dsRNA.

• RIG-I is structured such that its CARD is not exposed in the absence of dsRNA. Therefore, it is efficient to modify the protein to have a similar function even if the CARD of RIG-I is replaced with an Apoptosis Inducer.
• Excessive suppression of the expression level to prevent dsRNA-independent apoptosis may conversely reduce the apoptosis induction function in the presence of dsRNA.

To reduce the background activity of COCCO, it was necessary to make modifications such as giving RIG-I a binding ability to a protein different from its native target. Please refer to the Model Page for details. To achieve this modification, we needed to consider which method or software to use for protein modification. As a result of our own research and selection of usable methods and software, we considered three options before the interview.

Since we had already designed the variant protein using directed evolution, we needed to confirm if that approach was correct. We also interviewed an expert in protein modification for advice for the second time.

Variant protein design using Directed Evolution. Simultaneously generate and evaluate candidate proteins using ProteinMPNN.

• ProteinMPNN allows specifying a structure and outputs sequences that are likely to form that structure using machine learning. Therefore, efficient mutation introduction is possible using ProteinMPNN based on the Hel2i-CARD2 complex structure.

In iGEM projects, it is very difficult to conduct proof-of-concept using actual animals. Therefore, we needed to show whether COCCO could suppress HPAIV in chickens through simulation. Since there were several models for infection simulation in animal bodies, we chose the optimal one for this simulation.

We wanted advice on constructing infection dynamics models both inter-cell and intracellular. We interviewed an expert who authored a book on the inter-cell dynamics model of viruses, which we were referencing in our modeling process.

Construct a cell-free infection model and calculate the basic reproduction number to investigate the effectiveness of a defense system that controls the mortality rate of infected cells.

To confirm the effectiveness of the defense system, it is sufficient to evaluate the value indicating the possibility of infection. For this claim, considering virus inflow from the cell contact surface, cell-to-cell infection, and detailed virus behavior inside the cell constitutes excessive conditions for constructing the minimum model.

We initially thought that if we could show data that GM chickens equipped with COCCO have very strong resistance to HPAIV, significantly suppressing the amount of virus released and resulting in the convergence of avian influenza in the chicken house, large-scale mass culling could be avoided. We also understood that HPAI prevention guidelines were somewhat independently determined by each country, and that each country could revise them independently.

During an interview with the administrative staff related to livestock in Kyoto Prefecture, we heard that the HPAI prevention law is discussed by the Ministry of Agriculture, Forestry and Fisheries Council. Therefore, we investigated the discussion records regarding HPAI-related laws published on the MAFF website and interviewed a person who participated as a committee member.

Any evidence data will make the complete elimination of mass culling very difficult to revise.

• The HPAI prevention guidelines are determined by the World Organisation for Animal Health (WOAH) (OIE), and each country ratifies them. Therefore, revising the law requires the consensus of all WOAH member countries, making revision very difficult.

We discussed whether the introduction of GM chickens equipped with COCCO could potentially limit culling during an HPAI outbreak to only the infected chicken house.

During an interview with the administrative staff related to livestock in Kyoto Prefecture, we heard that the HPAI prevention law is discussed by the Ministry of Agriculture, Forestry and Fisheries Council. Therefore, we investigated the discussion records regarding HPAI-related laws published on the MAFF website and interviewed a person who participated as a committee member.

COCCO is a potential HPAI countermeasure that can mitigate mass culling.

• Culling for Newcastle disease is mitigated because an excellent vaccine that is strong against viral mutation and stably effective has been developed. Existing HPAI vaccines have not achieved this because they are weak against HPAIV mutation.
• However, GM chickens equipped with COCCO are expected to maintain stable and powerful viral infection defense and onset suppression effects even if HPAIV mutates.

Gene editing tools have been rapidly evolving in recent years, and various tools exist. Therefore, we needed to consider which editing tool to use when practically implementing chickens with COCCO. Initially, we considered adopting an editing method with fewer off-target effects and without double-strand breaks, and PASTE [4] was the most promising option.

To consider the selection of a gene editing tool for commercialization, we interviewed an expert who is involved in a genome-edited fish sales company and is knowledgeable about gene editing tools.

Use tools whose patents will expire in a few years (Crisper-Cas9, etc.) or self-developed tools.

• Commercial use of gene editing tools requires paying very high licensing fees to the licensor, so it is necessary to be mindful of licensing matters first when commercializing.
• Tools whose patents will expire in a few years or self-developed tools can significantly reduce costs such as licensing fees, making them optimal for use in commercialization.
• We initially considered using high-performance third-generation editing methods (PASTE, etc.), but decided to avoid using them due to the high risk of extremely high licensing fees during commercialization.

Through lectures and dialogues with consumer groups, we felt that there is a strong backlash against concealed research systems. Therefore, we needed to consider the best way to communicate about the GM chickens to society. For this question, no options were rejected, but rather new tasks were added based on stakeholder opinions.

We decided that it would be ideal to interview people with experience in developing and implementing genome-edited foods and those from a consumer group researching the safety of GM food, and to comprehensively consider both perspectives. We also sought the opinions of on-site poultry farmers to comprehensively incorporate feedback from various stakeholders.

Clearly define and promote why GM chicken implementation is necessary for current society. Constantly open and publish research data. Limit media dissemination to reliable media only to prevent misunderstandings from spreading to the public. Express a sincere attitude toward dialogue and research for people who have doubts about the research and implementation.

• Whether genetically modified technology is accepted is heavily influenced by how necessary the technology is to people.
• Making research data as open as possible allows consumers to make appropriate judgments about whether to use GM chickens, leading to increased peace of mind.
• If misleading information is disseminated on social media, etc., credibility can be lost quickly, so it is necessary to carefully select the media used for information dissemination.
• Giving a name that is easily accepted as an alien substance increases the likelihood of it being perceived as dangerous, so a soft impression name is important.
• Consumer doubts about GM food often stem from aversion. Sincerely dialoguing and responding to doubts and communicating a sincere research attitude that thoroughly considers safety can potentially mitigate this aversion.

When conducting a business, it is common to protect core technology with patents, but this is not always the right answer. Patents have the disadvantage that the details of the technology are fully disclosed upon application. Keeping it confidential, like the Coca-Cola recipe, can sometimes be the correct approach. This time, we considered how to protect COCCO as the core technology of the business.

We decided to hear the experiences of people who have developed and implemented GM food. We also consulted with a venture capitalist as a business planning expert.

Acquire a patent with the widest possible scope of effect and gain an advantage over competitors until the M&A stage.

• Since there is no IP that can be kept confidential without relying on patents (such as chicken raising know-how), acquiring a patent that is as difficult as possible for competitors to bypass is the best option.

Planning the initial implementation is extremely important for commercialization. Therefore, we considered which region the GM chickens should be initially implemented in. Initially, we considered implementation starting from Japan, and also assumed North America, where HPAI damage is frequent.

We interviewed a researcher who is well-informed about the current situation of the avian influenza problem in each country. We also consulted with a venture capitalist regarding the legal aspects.

North America, South America, China, and the Middle East.

• The four selected regions all suffer large losses from HPAI, resulting in high demand for HPAI countermeasures.
• While demand for HPAI countermeasures is high in Europe and Japan, for example, regulations on the use of biotechnology are strong, making legal implementation difficult.

We needed to include an exit strategy in the business plan to receive aid from investors. The exit strategy options were M&A (Mergers & Acquisitions) or IPO (Initial Public Offering), and we needed to select one.

We decided to consult with a venture capitalist as a business planning expert.

M&A with a major chicken breeding stock sales company.

• A few large companies monopolize chicken breeding stock worldwide, making new entry extremely difficult.
• To deliver HPAI-resistant GM chickens worldwide, M&A, which allows utilizing the resources (funds and distribution network) of such companies, is the ideal format.