There is more to creating genetically modified organisms than just lab work. Biosafety is one of the issues addressed by the Convention on Biological Diversity. They created the Cartagena Protocol, an international agreement on the need to protect human health and the environment from the possible adverse effects of the products of modern biotechnology [1]. We believe that everyone working in biotechnology should keep this protocol in mind to design safety in their research.

For this reason, we aimed to contribute to the incremental progress of synthetic biology becoming safer and more secure. We tried to achieve this by both talking with society and biosafety experts. To assess public opinion about our project and its safety, we performed a survey and spoke to many different stakeholders . Beside this, we wanted our project to be safe-by-design and we embedded biosafety in it.

Our living diagnostic is intended for human use and then it is excreted into the environment. For this reason, safety has to be taken into account at the very beginning of the project. Hence, a safe-by-design approach has been applied to Colourectal.

Safe-by-design is the approach in which safety is included at the earliest possible stage of product development. Safety of starting materials, processes, as well as the final products must be considered during the design phase. Safe-by-design does not only include technical lab safety, but also puts our project into a broader context with legal, moral and social aspects [2]. We thought of and implemented a safe-by-design approach before, during and after our project design.

The safe-by-design approach in Colourectal takes biosafety and biocontainment mechanisms into consideration. Apart from choosing parts without negative effects, we designed kill switches on three levels: confinement to the colon, biocontainment, and an inducible kill switch. Our living diagnostic should be taken as a pill that opens in the colon of the user. To avoid its spread to other parts of the body, we designed it to be dependent on mucin, a protein found in the lining of the colon.

As mentioned before, Colourectal will exit the user’s body and be able to enter the environment after use. To prevent escape of our genetically modified bacteria to the environment, a temperature-activated kill switch was designed. This kill-switch is inactive at the normal temperature of the colon, but activated when the temperature decreases, effectively stopping transcription in the bacterial cells. For more details, see the biosafety results section on this subject.

The third and final kill switch was designed to be inducible. In this way, the user can remove Colourectal from their gut with one pill, either at the end of the test or due to complications during the process. The metabolite in this pill is not harmful to the user or their microbiome but will induce a kill switch in our living diagnostic.

For more detail on the three kill switches, see the biosafety results section of these experiments. By implementing the kill switches, our living diagnostic should not pose a threat to the user’s health or the environment.

Following the safe-by-design principles, we chose Escherichia coli Nissle 1917 (EcN) as a safe chassis. EcN is a strain of E. coli that has probiotic properties. For almost 100 years, EcN has been used as an approved pharmaceutical ingredient. EcN was first discovered by prof. Alfred Nissle in 1917, who isolated the strain from a German soldier during the First World War. Since then, EcN has become the most intensively studied probiotic E. coli strain [3].

Several members of the E. coli family are known to possess the pks genomic island. This encodes colibactin, a toxin which can cause genotoxic effects in eukaryotic cells. This should be taken into account when using EcN as a living diagnostic. Literature reports that there is conflicting evidence on the genotoxicity of EcN [4–6]. Fortunately, the toxic effect can be shut down, while keeping the probiotic properties that EcN has, by substituting an amino acid in the clbP gene, which is responsible for the colibactin activation [5]. Alternatively, the whole clbP gene could be knocked out. Hence, we would like to use a clbP KO EcN strain in our design [7].

Another question that can be raised is whether the genetically engineered EcN is toxic. Of course, this depends on the specific parts and the goal that it is engineered for. Keeping the safe-by-design principles in mind, we chose intrinsically safe parts in our design. Additionally, using engineered EcN in medicine is not unheard of. EcN has already been genetically engineered to generate synthetic medicines for humans. These ‘living medicines’ aim to, for instance, treat tumours in vivo [8]. Finally, to make sure our living diagnostic will not be harmful to the microbiome, we modelled the effect that it would have on the gut microbiome, where a negligible effect was found.

Engineering EcN into a living diagnostic will ultimately require genome integration of the relevant sequences. However, currently we have designed and engineered most parts on plasmids. This makes them easier to work with during the early stages of engineering. However, plasmids can be a burden to bacterial cells. In order to maintain plasmids in bacteria, we must select for the antibiotic resistance marker on the plasmid by supplementing the relevant antibiotic in the growth medium. This is not favourable for the final product, as the end-goal is to introduce our living diagnostic in the human gut. Therefore, we designed all parts to be introduced into the genome of the bacteria in the end. By inserting our parts into the genome of our living diagnostic, our system is antibiotic free and has increased genetic stability.

When our living diagnostic detects cancer, it will secrete a blue chromoprotein to alarm the user. Chromoproteins are a class of small, coloured proteins that come in a variety of colours. They are derived from Green Fluorescent Protein (GFP), but with small alterations to make them absorb ambient light. This allows for instrument-free detection. We researched potential toxicity or allergenicity risks of GFP and its chromoprotein derivatives. Research performed in rats showed that when they were fed a diet containing high amounts of GFP there were no adverse reactions. Moreover, comparing GFP to known food allergens shows there are no significant matches to known food allergens [9].

Several of our parts contain Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)-Cas-based genome editing tools, which are a crucial part of synthetic biology. Biosafety concerns should be taken into account when introducing CRISPR systems. Will the pathogenicity, virulence, transmission or antibiotic resistance of the microbe be altered? We do not have the answers to these questions, which is why our living diagnostic should go through rigorous testing before being used to detect colorectal cancer in humans [10].

“This Colourectal diagnostic tool you are building could empower people to take charge of their own health, and there is an opportunity for the doctors as well. You have thought about the angles of biosafety and the containment, and it is exciting.” - Elizabeth Vitalis, Biosecurity Specialist at Inscripta.

Biosecurity aims to prevent misuse by loss, theft, diversion or intentional release of pathogens, toxins and any other biological materials. Before, during and after the design of an engineered bacterium it is important to think about the biosecurity measures that should be taken. Specific to our project, we should think about the dual use of our genetically modified organism (GMO) and how to prevent misuse. Dual use usually refers to technologies that can be used by civilians and the military, but in life sciences this term slightly changes meaning. Dual use in life sciences refers to the dilemma of creating something meant to for instance improve public health, but could also be used to impair public health if fallen into the wrong hands [11]. In the last centuries, bioterrorism has become a more tangible problem, which emphasizes the need for education on and considerations of biosecurity. An example of this can already be found in the 18th century. The smallpox virus was being researched, which led to the first viral vaccine, causing a great improvement of public health. However, this research was also used to impair public health, leading to the use of a virus as a bioweapon [12].

Biosecurity risks associated to the Colourectal project are mainly found in implementing a landing path for genetic insertions in EcN’s genome. This has the dual use possibility of being used to insert other dangerous genes or swap the chromoprotein with something else that could be used as a bioweapon (e.g., a toxic molecule for humans). It is especially important to consider non-toxic insertions that could inadvertently become dangerous. For this reason, when engineering a strain, it is important to look at the origin and predicted function of the genes that we are inserting.

The concerns regarding the safety of EcN already arose when pathogenic genes were found in its genome, presumably acquired by horizontal gene transfer (HGT) [13,14]. These genes do not confer pathogenicity to the bacteria unless the complete pathway is present.

Apart from considering biosafety and biosecurity measures, ethics should also be taken into account when designing a living diagnostic tool meant for the general public. During our project we were guided by Dr. Zoë Robaey, Assistant Professor in Ethics of Technology at Wageningen University.

First, we took the ethical considerations of working with GMOs into account. At the beginning we were aware that EcN poses no harm to human health or the health of the gut microbiome. However, we considered the possibility of our final product, an engineered EcN, being harmful to human and microbiome health due to unexpected changes. Additionally, our living diagnostic could pose a threat to other organisms and the environment. We took these considerations into account when designing our tool, which is why it has several built-in safety mechanisms. More on this can be found in safe-by-design, under biosafety. However, we then found genes with potential genotoxic effects in EcN’s genome. Luckily these genes do not confer pathogenicity unless the complete pathway is present. Still, this possible pathogenicity should be taken into account during the production of our living diagnostic.

Next to these technical considerations, we thought about the ethics of our tool once it is in use, focussing on access to the test, inclusion, and privacy issues. Firstly, concerning access: who will be able to take this test? Can everyone afford it? In order to deal with a fair access, and based on our stakeholder workshop, we decided that our tool should be implemented into the current colorectal cancer screening programme. In this way, people will not have to buy the test themselves, instead receiving one at home when they are in the risk group for colorectal cancer. Second, concerning inclusion, we took further measures against discrimination in our living diagnostic. We used a colourblind-friendly colour scheme and an easy-to-read font in all our information. Additionally, we used a gender-neutral persona of no specific race in all figures explaining the usage of our living diagnostic. See the inclusivity page for more choices we made against discrimination.

Third, we thought about privacy issues. We developed an app to give people less uncertainty when checking the colour of their stool after using our diagnostic tool. This means that the app will handle sensitive medical data of users. The app should thus handle privacy data and medical information with care, and security should be our top priority by following regulations on handling medical data by being HIPAA (Health Insurance Portability and Accountability Act) compliant and the GDPR (General Data Protection Regulation ) compliant in Europe.

Not only the safety of our tool itself, but also the safety of designing and making our tool should be taken into account. During our iGEM journey we were able to use the labs of the Laboratory of Microbiology, Systems & Synthetic Biology, Human Nutrition and Health and Bioprocess Engineering at Wageningen University & Research. Before we started any research all our group members followed the obligatory lab safety tour. Additionally, we finished obligatory online lab and office safety modules. In the lab tour and modules, we learned and were tested on lab rules, lab etiquette, safety equipment, cleaning and waste disposal, how to handle chemicals and other dos and don’ts in the lab.

The following is a list of all the things we and all workers in our university were taught to ensure a safe work environment. We followed the lab safety rules described by the World Health Organization [15].

All our group members that performed lab work, worked in ML-1 labs. In ML-1 labs, we do not work with pathogens. A lab coat and safety glasses are mandatory in the lab. Gloves may be, this differs per lab. Lab coats are forbidden in common areas (toilet, canteen etc.). Open-toed shoes, eating, drinking, smoking, applying makeup or handling lenses is forbidden in the lab. Contaminated samples must be sterilized e.g., with an autoclave.

Some of our experiments had to be performed in an ML-2 lab, and the members of our group that worked there followed an additional lab safety tour for this. In ML-2 labs, we work with organisms that could cause diseases that have an available treatment. In an ML-2 lab gloves are mandatory, and only authorized people have access to the lab.