A Novel High Throughput Platform for Assessing Bacterial Growth and Morphology Enables Rapid Antibiotic Susceptibility Testing

Abstract

The alarming rise in antibiotic resistance necessitates a better understanding of the mechanisms of antibiotic action and calls for tools that enable rapid antibiotic susceptibility testing (AST). This thesis introduces the Multipad Agarose Plate (MAP): an innovative platform that can perform rapid AST in under three hours. The platform comprises 96 agarose pads and relies on single-cell timelapse microscopy to image bacteria as they develop into microcolonies. The platform can be manufactured with a laser cutter at a low cost based on off-the-shelf parts, making it easy to adopt for other labs. A robust open-source segmentation package, PadAnalyser, was developed to automatically extract this data from brightfield microscopy images. The MAP enables precise quantification of growth and morphological changes, offering significant applications in clinical and research settings. The primary application of the MAP is antibiotic susceptibility testing (AST), which determines what concentration of antibiotics is needed to inhibit the growth of a bacterial sample. Clinically, this is an important tool for selecting appropriate antibiotics to treat bacterial infections. The MAP can provide results in less than three hours by tracking changes in growth rate, offering an eightfold speedup in turnaround time compared to current clinical standard phenotypic methods. The MAP has also been demonstrated as a research tool, allowing us to study the effects of antibiotics on bacterial growth and morphology. Using three clinically relevant bacteria species (E. coli, S. aureus, and P. aeruginosa) and a broad test panel of antibiotics, I found that the antibiotics universally lead to increased growth rate heterogeneity as the dose approaches the minimum inhibitory concentration (MIC). Protein synthesis inhibitors are the exception, instead reducing heterogeneity towards the MIC. In addition, I show that all of the antibiotics tested cause morphological changes to the cells of the three species, with a dose dependency closely correlated to the MIC. This observation led to the development of the morphology 50 (MOR50) threshold, a new morphological parameter that enables the estimation of MIC with a single snapshot after 2.5 hours of incubation. This would drastically reduce imaging time per sample, making the MAP a truly high throughput tool. We also utilize the MAP to study how salts impact the mecillinam susceptibility of two E. coli strains: the wild type (WT) and a mutant with the cysB gene knocked out. The CysB protein enables cysteine synthesis, and I find that the ∆cysB strain has a lower growth rate and smaller cell size during normal growth on pads than the WT. We show that certain salts confer mecillinam resistance to both strains, increasing their MIC by several orders of magnitude. The mechanisms that drive this are not yet understood, and our hypothesis linking salt impact to morphology was disproved after demonstrating that salts do not significantly affect the morphology of either strain. Besides having proven itself useful as a tool for research, the MAP could have clinical applications in future. The validation done so far is promising, but more work with other bacteria species and patient samples will have to be performed. The speedup enabled by the MAP could improve patient care by expediting treatment initiation and alleviating the burden of antimicrobial resistance.