Determining growth rates from bright-field images of budding cells through identifying overlaps.

Much of biochemical regulation ultimately controls growth rate, particularly in microbes. Although time-lapse microscopy visualises cells, determining their growth rates is challenging, particularly for those that divide asymmetrically, like Saccharomyces cerevisiae, because cells often overlap in images. Here, we present the Birth Annotator for Budding Yeast (BABY), an algorithm to determine single-cell growth rates from label-free images. Using a convolutional neural network, BABY resolves overlaps through separating cells by size and assigns buds to mothers by identifying bud necks. BABY uses machine learning to track cells and determine lineages and estimates growth rates as the rates of change of volumes. Using BABY and a microfluidic device, we show that bud growth is likely first sizer- then timer-controlled, that the nuclear concentration of Sfp1, a regulator of ribosome biogenesis, varies before the growth rate does, and that growth rate can be used for real-time control. By estimating single-cell growth rates and so fitness, BABY should generate much biological insight.

Diffusion-driven enhancement of the antibiotic resistance selection window.

The current crisis of antimicrobial resistance in clinically relevant pathogens has highlighted our limited understanding of the ecological and evolutionary forces that drive drug resistance adaptation. For instance, although human tissues are highly heterogeneous, most of our mechanistic understanding about antibiotic resistance evolution is based on constant and well-mixed environmental conditions. A consequence of considering spatial heterogeneity is that, even if antibiotics are prescribed at high dosages, the penetration of drug molecules through tissues inevitably produces antibiotic gradients, exposing bacterial populations to a range of selective pressures and generating a dynamic fitness landscape that changes in space and time. In this paper, we will use a combination of mathematical modelling and computer simulations to study the population dynamics of susceptible and resistant strains competing for resources in a network of micro-environments with varying degrees of connectivity. Our main result is that highly connected environments increase diffusion of drug molecules, enabling resistant phenotypes to colonize a larger number of spatial locations. We validated this theoretical result by culturing fluorescently labelled Escherichia coli in 3D-printed devices that allow us to control the rate of diffusion of antibiotics between neighbouring compartments and quantify the spatio-temporal distribution of resistant and susceptible bacterial cells.