Safety and Security

Training

All wet lab work was carried out at [host institution and laboratory], under the supervision of [PI name] and [lab manager or safety officer]. Before any team member handled an organism or a reagent, they completed:

• [Institutional general laboratory safety training]
• [Biosafety and bloodborne pathogen training, if applicable]
• [Chemical hygiene and hazardous waste training]
• Hands-on induction with a supervising lab member, who signed off before independent work

No team member worked alone in the laboratory at any point in the season.

Personal protective equipment

• Lab coats, closed-toe shoes and nitrile gloves at all times in the lab space
• Safety glasses when handling chemicals, and a face shield during UV work
• Gloves changed between handling cultures and touching shared equipment, keyboards or door handles
• Lab coats removed and left in the lab on exit

Waste

No engineered organism left the laboratory at any point, and none was released into the environment.

Emergency equipment

Locations of the eyewash station, safety shower, fire extinguisher, first aid kit and spill kit were shown during induction and are known to every team member. Emergency contacts are posted at the lab entrance.

How we assessed risk

For each procedure we asked three questions: what could go wrong, how likely is it over a four-month season, and how bad would it be. We scored likelihood and severity on a three-point scale and used the product to decide whether a control was needed before proceeding.

Likelihood

Severity

Risk register

Near misses, and what we changed

The gel rig we did not use. On June 23 we had prepared a gel and loaded samples when the electrophoresis rig showed signs we did not trust. A damaged rig running at 160–190 V is both an electrical hazard and a route to ruined samples. We stopped, did not run the gel, and reported the equipment rather than trying it carefully. The cost was a day. We think this was the correct call and we would make it again.

Ampicillin satellite colonies. Not a safety incident, but a process risk with a real lesson. Colonies that grow on an ampicillin plate without being genuine transformants look identical to real ones. We lost a week assuming slow growth. The control we added was to require liquid culture establishment before any colony is treated as a transformant.

A tube swap. In August, sequencing revealed that two constructs had been placed in each other’s tubes. The error was invisible on a gel because both products were the same size. Nothing unsafe resulted, but it showed that our labelling discipline was not sufficient for parallel workflows. We introduced a formal clone tracker and a rule that tubes are labelled before they are filled.

Organisms used

All strains are Risk Group 1 and all work was performed under BSL-1 conditions, which is appropriate for organisms not known to cause disease in healthy adult humans.

Source of genetic material

Our FAAH coding sequence derives from Faecalibacterium prausnitzii, a commensal anaerobe that is one of the most abundant bacteria in the healthy human gut. We work with the synthesised coding sequence only, not with the organism itself, so no anaerobic culture of a gut commensal was required.

No sequence in any of our constructs comes from a pathogen, encodes a toxin, or appears on any controlled agent list.

Containment

• All culturing, transformation and lysis performed at the bench under BSL-1 practice, with a biosafety cabinet used where the host lab’s protocol requires it
• Cultures transported only in sealed, labelled containers and only within the building
• Benches disinfected before and after work
• All biological waste autoclaved on site

What we did not do

We did not perform any animal work this season. Our aims document an eventual in vivo appetite study in mice, and we want to be explicit that no such work has taken place and none will take place without full IACUC or equivalent ethics approval, a protocol reviewed in advance, and trained personnel. We list it as a future aim, not as work in progress.

We did not perform any work with human subjects, human tissue, or clinical samples.

The TAMRA-FP probe deserves emphasis because its mechanism is the hazard. It is designed to form a covalent bond with serine hydrolase active sites, and human serine hydrolases are not exempt. It will be used in a fume hood, in small volumes, with the hood sash low.

Most iGEM projects can treat dual use briefly. We do not think ours can, for two reasons: our chassis is designed to colonize the human gut, and our output is a change in human appetite. Both deserve honest discussion.

The engineered colonizer question

EcN is attractive precisely because it persists in the gut. That same property means an engineered EcN strain is not a self-limiting system. If released, it would be more likely to establish than a laboratory strain would.

Our mitigations at the current stage are procedural: BSL-1 containment, autoclaved waste, nothing leaves the lab. Those are adequate for benchtop research and inadequate for anything beyond it. A strain intended for human administration would need engineered biocontainment, and we consider this a requirement rather than an enhancement. The approaches we would expect to need are auxotrophy for a compound not available in the gut, or an inducible containment system, and these would have to be designed and validated before any translational step.

The antibiotic resistance marker

Our constructs carry chloramphenicol resistance. On a plasmid, in a gut-colonizing organism, this is a genuine concern rather than a theoretical one, because the gut is an environment where horizontal transfer between bacteria occurs.

This is acceptable for laboratory research and is not acceptable for a product. A translational version would need the resistance marker removed, with the construct integrated into the chromosome and maintained by a non-antibiotic mechanism. We flag this now rather than discovering it later.

The appetite modulation question

Our intent is to raise appetite in children recovering from malnutrition, where loss of appetite is itself a barrier to recovery. Appetite is not a neutral target. A technology that reliably changes how hungry a person feels could be directed at people who have not consented to it, or deployed against the interests of the person receiving it rather than for them.

We also note that the inverse application, suppressing appetite, has a large commercial market and a long history of harm when poorly regulated. A system that modulates endocannabinoid-like satiety signalling sits close to that space.

What we are not publishing

We describe our constructs, primers and protocols in full because reproducibility is a core iGEM value and because none of this information lowers a barrier to harm. Our work uses a non-pathogenic chassis, a commensal-derived enzyme, and standard cloning methods.

We have not and will not publish work on enhancing gut persistence beyond wild-type EcN, on evading host immune clearance, or on delivery routes to people without their knowledge. These are the directions where the dual-use concern becomes concrete, and they are outside our project scope by design rather than by omission.

We put this table on the page because the gap between our current work and a usable product is large, and we think a safety page that does not say so is not being straight with the reader.

• iGEM Safety and Security Policies, and the iGEM White List of approved parts and organisms
• Our host institution’s biosafety manual and chemical hygiene plan [add reference]
• Safety Data Sheets for every chemical listed in section 4, held in the laboratory and consulted before first use
• [Add institutional biosafety committee approval reference, if applicable]
• [Add any team safety form submission reference approved by iGEM HQ]

The reagent-level hazards for each protocol are repeated in context on the Experiments page.