Biosafety#

Organisms and risk group#

All work was carried out in laboratory strains of Escherichia coli:

These are standard, non-pathogenic, well-characterised laboratory strains, attenuated by long domestication and unable to colonise the human gut. They are handled as Risk Group 1 organisms.

No pathogen, no virulence factor and no toxin gene is present anywhere in the project. Nothing constructed here is designed to survive outside a laboratory incubator.

The three things worth arguing about#

A generic risk statement would miss what is actually distinctive about this project. Three features deserve specific justification.

1. dCas9 is catalytically dead#

The Cas9 used throughout is nuclease-deficient. It binds DNA and obstructs transcription; it does not cut. The genome-editing hazard normally associated with Cas9 is absent by construction, not by containment.

This is a design property, not a precaution that could lapse — a mutation restoring nuclease activity would require reverting specific active-site substitutions, and would confer no advantage under our selection.

2. Hypermutation is targeted, not global#

MutaT7 raises the mutation rate of a defined cassette by fusing a deaminase to T7 RNA polymerase. Mutagenesis tracks along sequence behind a T7 promoter and stops at the terminator. The host genome is replicated by its own high-fidelity machinery and is not affected.

This matters for safety as well as for the experiment. We are not producing a general mutator strain, which would accumulate uncontrolled changes including potentially in stress-response and resistance pathways. The hypermutated sequence is one structural protein gene from a soil bacterium, and it is the only sequence under mutation in the cell.

3. Antibiotic resistance is the readout — and this needs stating carefully#

The selection works by repressing a kanamycin-resistance gene and rescuing it. This deserves an explicit argument rather than a reassurance.

• No new resistance is created. The kanamycin-resistance gene is a standard laboratory marker, already ubiquitous in cloning vectors worldwide. The selection modulates its expression; it does not evolve the resistance protein or broaden its spectrum. This is the opposite of the T7-ORACLE work we cite for context, which evolved a β-lactamase toward clinically relevant substrates — our mutagenised locus is a structural shell protein with no resistance function.
• The evolutionary pressure is on the shell, not the marker. The marker sits on the non-mutagenised plasmid, deliberately, for reasons that are simultaneously experimental and precautionary: mutations in the resistance gene would corrupt the selection and would be the only route by which this project could generate a novel resistance phenotype.
• Antibiotics used are standard laboratory selection agents — ampicillin, kanamycin, chloramphenicol — at ordinary working concentrations, disposed of as contaminated waste.

Containment practice#

Standard microbiological practice for Risk Group 1 work: work confined to the laboratory, cultures autoclaved before disposal, contaminated plasticware and plates treated as biological waste, benches disinfected, no organisms removed from the facility. Strains are archived as glycerol stocks at −80 °C.

Chemical hazards were handled alongside the biological ones. PMSF, used as a protease inhibitor during lysis, is acutely toxic and is handled with gloves, with attention to its persistence in the solutions it is added to.

Risk assessment#