Uncovering the spectrum of single-cell physiological states in bacteria.

One pronounced feature of bacteria is their plasticity and dynamic nature. For example, they rapidly transition between growth stages as their population grows and matures, assuming markedly different physiologies and antibiotic tolerance levels in each stage. Yet, whereas population growth phases are generally described via a few broad labels (e.g., the lag, exponential, and stationary phases), a closeup of individual cells suggests a spectrum of nuanced stages may define this process. Furthermore, the fact that individual cells divide asynchronously means that different sub-populations can reach different growth stages at different times. Yet, current bulk population analysis masks these cell-cell differences, prohibiting their discovery and characterization.
In this work, we explore this hidden complexity and its relationship with phenotypic heterogeneity, using novel computational image analysis approaches and machine learning. We quantitatively map these nuanced transitions during growth at the individual bacterium level using combined measurements of cell morphology, DNA content, cell wall deposition, and ribosome abundance. Our results reveal a wide spectrum of previously uncharacterized physiological cell states and pave a new path toward understanding bacterial physiology across scales, from the collective to the individual.