A Single Dose Kills 97% of Superbugs

Antibiotics are running out of road. For every drug that works, bacteria evolve to survive it — and the pipeline of new candidates has been thin for decades. A team at the University of Oxford is taking a different approach: instead of designing a new chemical, they built a living bacterial cell that cannot reproduce but knows how to kill.

The platform, described in a paper published this week in PNAS, uses engineered E. coli cells stripped of their chromosomes — rendered incapable of replication — and fitted with surface nanobodies that recognize specific pathogens. When a SimCell (short for simple cell) encounters its target, it docks and delivers a two-part kill signal: a needle-like injection system fires toxic effectors directly into the bacterial cytoplasm, while an engineered enzyme converts aspirin into hydrogen peroxide at the site of contact.

In tests against E. coli ST131 — a multidrug-resistant clinical strain that causes urinary tract and bloodstream infections and has spread globally — a single dose of the smaller mini-SimCell variant eliminated more than 97% of the target population within 24 hours. Four sequential doses in a mixed microbial community achieved a 103-fold reduction of the targeted strain while leaving off-target bacteria largely untouched.

The mechanism matters more than the number. Traditional antibiotics work by diffusing through tissue, which means the dose is limited by what healthy cells can tolerate. The SimCell approach is contact-dependent: the nanobody anchors the hunter cell to its prey, then the T6SS needle fires at close range. The aspirin-triggered hydrogen peroxide adds a secondary, sustained antimicrobial effect at the same location. Bacteria can evolve resistance to a single drug. Evolving resistance to a system that simultaneously punctures the membrane, injects toxins, and floods the local environment with reactive oxygen species is a taller order.

AMR is projected to cause more than 10 million deaths annually by 2050, with cumulative GDP losses of $100 trillion. Every approved antibiotic class currently on the market has seen resistance emerge somewhere in the world, the authors note.

The SimCell concept originated in Wei E. Huangs lab at Oxford, which first described the platform in a 2020 PNAS paper as a programmable chromosome-free cell system. The platform was subsequently developed for cancer therapy. This new work extends it to pathogens — the first direct demonstration against a clinically relevant multidrug-resistant isolate.

The choice of E. coli ST131 as the test case is deliberate. ST131 is one of the most successful drug-resistant bacterial lineages in the world, responsible for a large fraction of fluoroquinolone-resistant and extended-spectrum beta-lactamase-producing E. coli infections in hospitals and communities across the US, Europe, and Asia. It is a WHO priority pathogen. The Oxford team targeted it with nanobodies against OmpA, an outer membrane protein abundant on the ST131 surface.

Safety is built into the design. SimCells cannot replicate — their chromosomes have been removed — and the measured escape frequency is below 10 to the minus eighth, meeting NIH guidelines for clinical recombinant microorganisms. This addresses a long-standing concern with engineered bacteria: the risk of uncontrolled proliferation inside or outside the body. The prior minicell platform that underlies this work has already been tested in a Phase I clinical trial and received FDA Fast-Track designation. The SimCell approach adds targeting precision to that established safety baseline.

The approach is modular by design. The nanobody targeting layer can be swapped: different surface antigens would, in theory, allow the same platform to be redirected at different pathogens. This is the plug-and-play claim in the papers title, and it is what makes the system potentially scalable across a range of drug-resistant infections rather than being a one-off therapy.

Oxford SimCell Ltd., a spinout from the university, is the commercial vehicle. The team includes former Siemens Healthcare business development leadership and a CTO who has been developing the SimCell patent portfolio since graduate work at Oxford. The company is currently pursuing clinical development of the antimicrobial platform.

The step from a PNAS paper to a clinical candidate is long, and in vitro efficacy does not guarantee in vivo success. The paper notes that the work has not yet been tested in animal models of active infection. But the platforms prior clinical validation, the dual-mechanism design, and the specificity of the targeting system represent a combination that the field has not had before — a drug that uses a living cell as its delivery vehicle and cannot grow out of control. Whether that translates into a real clinical option is the question the next several years will answer.

https://www.pnas.org/doi/10.1073/pnas.2517118123