sTAUbility

Proposed implementation

Introduction

The first few steps in project development include providing an innovative solution, detailed planning, and understanding its application in the real world.

In our Project Description Page, we described the inspiration for our idea and how we came up with our proposed solution for genetic stability. We propose creating a linkage between a target gene and an essential gene, such that mutations damaging the target gene will affect the essential gene as well, leading to the mortality of the mutated host. While this biological concept is important, we still had to define our exact design for this project and the steps that are involved in its application. Finally, we had to plan how to turn this design into a real-world implementation.

We decided to develop a recommendation engine that would generate a customized construct for each target gene, composed of the best matching essential gene and linker for improved stability and higher expression. The decision to focus on software development was not trivial for us. In order to verify the necessity of our solution and find the best way to implement it, we had several meetings with some of the leading pharma & biotech companies worldwide. These meetings helped us better grasp the needs of the industry. In addition, we refined our product again and again thanks to dozens of academic consultants we met.

Figure 1. Consultants from industry and academia

We used the AREA framework, listed on RRI website, for maintaining our focus in the right direction. It helped us define what are the most fundamental questions that we need to ask and how to optimally integrate their answers into our project:

What do we offer, a product or a service?
Where is our solution most impactful?
Who are our proposed end users?
How can we improve our suitability for real world applications?
Which safety aspects must we consider?

What Do We Offer, a Product or a Service?

Given that we could eventually optimize the genomic stability of engineered constructs, we had to choose between two approaches for our project’s implementation:

Service – Pharmaceuticals companies would send us the engineered circuit, and we would send back (physically) this circuit with an increased genomic stability.
Product – A recommendation engine that would return an optimized sequence as an output.

We realized that the second option would match our multidisciplinary team’s capabilities much better. Our group consists of members with biological knowledge, as well as members with bioinformatic and algorithmic capabilities. The combination of software based on complex biological experiments on the one hand, and bioinformatic methods on the other, is best suited for our group.

We envision biotechnology companies as well as Pharma companies using our product as a recommendation engine that would help them quickly and easily obtain an optimized sequence, ensuring genomic stability over time while maintaining the required expression level.

Where is Our Solution Most Impactful?

Recurring consultations with industry experts clarified that our solution is highly relevant in the following cases:

Development phase of Pharmaceutical product – at this phase, lengthy screenings of potentially stable cell lines are performed to determine if a lead clone is stable enough to support manufacturing. Screenings involve long-term sub-culture regimes with comparisons of initial and final outputs, testing for any loss of performance over time. Our solution, when implemented in the development phase, will ensure stability at the following production phase, by obtaining optimized sequences as an output.
Perfusion processes – the term "perfusion" refers to production processes in which proteins are produced in a bioreactor over time. The population of engineered cells (whether microorganisms or mammalian cells) produces the desired product over a much longer period than in fed batch processes. Therefore, genomic stability in these processes is crucial. Prolonging the time of efficient production could have a significant economic impact.
Products with low expression level – Our solution can increase the expression levels of products whose expression levels are low in the first place. Our software selects the conjugated gene as well as the right linker that will allow genomic stability over time while maintaining high expression levels.

Who are Our Proposed End Users?

Any company or lab that wants to easily and quickly ensure the genomic stability of an engineered construct is a potential end user. Our solution has the potential to promote any such research. However, from discussions with industry leaders we realized that product development processes, for which our product is highly relevant, rarely occur at Pharma companies. In contrast, Pharma and Biotechnology companies that provide consulting services to pharma companies (also called CDMO) are constantly engaged in the search for technologies that improve the development process. Therefore, such companies will make more frequent use of our software.

In addition, a mutational hotspots detector called EFM (developed by 2015 Austin UTexas iGEM team) is already implemented in Benchling, which offers a platform for sequence design and analysis tools. We believe that our tool, which offers not only identification but also optimization of constructs’ stability, is highly relevant for such applications and can aid academics and iGEMers in the design process of their construct.

How Can We Improve Our Suitability for Real World Applications?

There are some challenges that we should address before offering our product to the end users.

Biological challenges – there are many methods for linkage between the essential and target genes, all described in our Design page.
Our software outputs the best matching linker for the purpose of the construct's insertion. However, we still need to prove the feasibility of the linkers that our software offers that we haven’t tested yet, since we have only empirically tested a fusion linker.
In addition, a further proof of concept is required by using proteins other than GFP or RFP, as well as using different promoters with varying expression levels.
Bioinformatics challenges - The models we developed used some yeast-specific databases. Therefore, our learning algorithm is adapted to yeast. As our vision is to generalize the process to other organisms, we need to gather more information and create new features accordingly.
Industry related challenges - We need to increase the economic impact of our product. To do so, we strive to empirically prove that the use of the various linkers selected by our software will allow long-term high expression levels. In addition, in order to adapt our software to a wider range of products, we need to prove that it also provides a platform for secreted proteins.

Which Safety Aspects Must We Consider?

Conjugating the target gene to an essential gene significantly increases the genomic stability of the engineered construct and therefore maintains its functionality for longer periods. Therefore, using our solution increases the safety in working with transgenic cells, since strains with a mutation in the target gene will not survive. Prof. Ori Gophna described it best when asked about safety issues of our product - "I think your solution actually makes the transgenic microorganism safer, since it doesn't require the use of selective markers like antibiotics. That's the big advantage of your method. I don't recognize a safety problem. If the engineered yeast, for example, leaves the laboratory or industrial environment, it will be very difficult for it to survive under environmental conditions.”