Population Size-Dependent Dynamics of Bacterial Response to Antibiotics

Abstract

Antibiotic resistance is on the rise throughout the world and poses an increasing risk to our health systems. Understanding how bacteria respond to drugs in complex environments can help us manage our current arsenal of drugs in a more effective way. Past work has primarily focused on mechanistic responses by bacteria that confer resistant to specific drugs, and this molecular level understanding is essential. Yet bacteria live in large and possibly heterogeneous populations, and it's often not clear how known molecular scale events lead to large scale behavior like survival or extinction.

In this thesis I make use of quantitative in-vitro experiments and mathematical models to understand and predict the population level dynamics of bacterial communities in the presence of drugs. Traditional methods for investigations at this scale suffer from some experimental limitations which I am able to overcome using new custom hardware. Using such tools, my experiments can include both precise measurements of a bacterial population over time while also including precise control of the growth environment. I can use this control to respond to the population in some way, such as holding a size threshold for it, or use it to stress the population in a specific manner, such as using differing drug dosing protocols. This versatility has allowed me to perform new and interesting investigations about the bacterial population respond to drugs in various settings, three of such experiments make up this thesis. In the first chapter, I show that the efficacy of many common drugs is dependent upon the density of the bacterial population in E. faecalis. I am able to quantify the amount to which several common drugs inhibit the growth rate of a bacterial population at different densities within exponential phase growth. In general, if such a density dependence for a drug exists, the drug is less effective at growth inhibition the denser the population is. I also investigate the cause of this effect, and find that the resulting change in pH of the environment from cellular growth can explain the effect for a couple of drugs. These results are then used to create a mathematical model that shows that a treatment regime could possibly lead to treatment failure when populations are very dense. In the presence of resistant sub-populations this density-dependent inhibition then leads to counter-intuitive dynamics, and this is the subject of Chapter 2. Using ampicillin and constant drug influx, these counter-intuitive dynamics include preferential survival of low-density populations over high-density counterparts. Using a model to understand this system, I show that this result comes from the pH driven density effect, and disappears when pH modulation of the environment is not possible. Finally, I show how competitive suppression can limit growth of drug resistant populations for a sufficiently high-density population. Using sub-inhibitory amounts of drug, I show that a mixed population of bacteria can be held at a high density far past when a non-suppressed resistant sub-population would have completely taken over the population. par As a whole, my work in this thesis helps to underscore the importance of density-driven community level interactions in determining the fate of bacterial populations exposed to antibiotics.