New test reveals which antibiotics actually kill bacteria
Researchers at the University of Basel have developed a method that measures whether antibiotics truly kill bacteria, not just inhibit their growth. The technique, called Antimicrobial Single-Cell Testing, tracks millions of individual bacterial cells in real time and has demonstrated that a drug’s killing power – rather than growth inhibition alone – predicts treatment success in tuberculosis and other mycobacterial infections.
Beyond growth inhibition: measuring bacterial death
Current antibiotic testing primarily assesses a drug’s ability to prevent bacterial growth under laboratory conditions. However, this approach fails to capture whether antibiotics actually eliminate pathogens in the body – a critical distinction that helps explain why some infections persist despite treatment with apparently effective drugs.
Dr Lucas Boeck and colleagues at the University of Basel’s Department of Biomedicine have addressed this gap with Antimicrobial Single-Cell Testing (ASCT). The method uses microscopic imaging to track millions of individual bacteria over several days, recording precisely when each cell dies following antibiotic exposure.
“We use it to film each individual bacterium over several days and observe whether and how quickly a drug actually kills it,” explains Dr Boeck. This approach reveals what proportion of the bacterial population is eliminated and how efficiently killing occurs.
Predicting treatment outcomes in tuberculosis
The research team tested 65 drug combinations against Mycobacterium tuberculosis, examining both actively growing bacteria and those in a dormant, starved state. The findings, published in Nature Microbiology on 9 January 2026, revealed a striking pattern: whilst standard drugs like isoniazid, rifampicin and ethambutol effectively killed growing bacteria, only killing activity against starved bacteria predicted treatment success in mice and humans.
Drug combinations containing bedaquiline, clofazimine or pretomanid proved particularly effective under starvation conditions. These findings align with clinical data demonstrating that these regimens can shorten tuberculosis treatment duration – a critical advancement given that standard therapy requires four to six months.
The authors note in their paper that “drug-specific killing dynamics in starved bacteria, rather than growth inhibition or killing of growing cells, predict regimen efficacy in mice and humans.” This observation underscores the importance of testing antibiotics under conditions that mirror infection sites, where bacteria often exist in metabolically less active states.
Bacterial genetics influence antibiotic effectiveness
Extending their analysis to Mycobacterium abscessus, a related pathogen causing difficult-to-treat lung infections, the researchers examined 405 clinical isolates from patients across Europe and Australia. They discovered substantial variation in how different bacterial strains responded to identical antibiotic exposures – a phenomenon termed drug tolerance.
Crucially, bacterial strains that survived longer during antibiotic treatment were associated with worse clinical outcomes in individual patients. The research demonstrated that combining tolerance measurements with conventional resistance testing improved prediction of treatment success from 69% to 78% accuracy.
“The better bacteria tolerate an antibiotic, the lower the chances of therapeutic success are for the patients,” says Dr Boeck.
The study also revealed that drug tolerance is largely genetically determined and heritable, with 32–97% of variation across different antibiotics attributable to bacterial genetics. The team identified multiple genes associated with tolerance, including a phage protein that modulates antibiotic killing when targeting protein synthesis.
Clinical implications
Dr Boeck suggests the method could eventually benefit both patients and drug development: “Our test method allows us to tailor antibiotic therapies specifically to the bacterial strains in individual patients.” He adds that understanding the underlying genetics could enable simpler tolerance tests and improve efficacy estimates for new drugs during development.
Reference
Boeck, L., et al. (2026). Large-scale testing of antimicrobial lethality at single-cell resolution predicts mycobacterial infection outcomes. Nature Microbiology. https://doi.org/10.1038/s41564-025-02217-y
