Impacts of pathogen and host factors on the dynamics of Escherichia coli bloodstream infection

Abstract

Bacterial bloodstream infections (BSIs) are leading causes of mortality and are poised to become increasingly difficult to treat due to rising rates of antimicrobial resistance. Developing new therapeutics is therefore of high importance and requires the use and analysis of animal models of infection. However, BSIs are highly complex infections, and we lack a thorough understanding of the host and pathogen factors that govern infection dynamics and control infection outcomes. Knowledge of several key facets of BSIs, including how microbes disseminate across the host, how different host/pathogen factors control bacterial clearance, replication, or dissemination, and how activation of innate immune factors is balanced to avoid deleterious systemic inflammation, is relatively sparse. Moreover, interactions at the host and pathogen interface yield tissue-specific phenotypes, contributing to the complexity of BSIs. A deeper understanding of the host factors and pathogen dynamics that govern BSIs in humans and animal models would inform the development of new therapeutics and provide a greater understanding of how microbes establish infection to cause disease.

This thesis describes my efforts toward expanding understanding of the host and pathogen factors that control infection dynamics during BSI. Chapter 2 describes a novel computational methodology known as STAMPR that uses barcoded bacteria to quantify bacterial replication, dissemination, and clearance within the host. In Chapter 3, we apply STAMPR to quantify the dynamics of Escherichia coli BSI following intravenous inoculation of mice. We defined several key features in this model, including identifying reservoirs of dissemination, quantifying tissue- specific patterns of clearance and bacterial replication, and characterizing host and bacterial factors that control infection dynamics.

Among the most surprising observations from these studies was the finding that E. coli replicates dramatically within liver-specific abscesses. In Chapter 4, we identify experimental, genetic, and molecular determinants of E. coli liver abscess formation. These studies suggest that hyperactivation of the innate immune response causes tissue necrosis in the liver, which serves as a site for bacterial replication, ultimately leading to abscess formation. Mice with diminished inflammatory responses, such as those lacking the lipopolysaccharide receptor TLR4, are resistant to abscess formation. Given the central role of TLR4 in sensing bacterial infection, in Chapter 5 we preform quantitative dose-response analysis using STAMPR to decipher the role of TLR4 during BSI. We describe a concept we term “dose scaling”, which relates changes in the inoculum size to the magnitude of host bottlenecks and the efficacy of the innate immune response. Our results demonstrate that during E. coli BSI, higher doses can lead to greater or reduced efficacy of innate immune responses in different tissues.

Despite the central role of TLR4 in controlling systemic inflammatory responses, TLR4 cannot solely explain variation in liver abscess susceptibility across mice. In Chapter 6, we further explore the mechanisms of liver abscess formation and find that expression of endogenous retroviruses (ERVs) correlates with abscess susceptibility. Administration of reverse transcriptase inhibitors, which may prevent accumulation of cytosolic DNA by ERV-encoded reverse transcriptases, prevents abscess formation. Reverse transcriptase inhibitors therefore represent a potential novel therapeutic for deleterious inflammatory consequences of BSIs. This dissertation concludes with a perspective on the use of barcoded bacteria to understand infection dynamics and a discussion of innate immune responses that control abscess formation.