Abstract
The inoculum effect (IE) reduces the efficacy of antibiotics both in vitro and in clinical settings. However, we do not understand, nor can we predict, when IE will occur. Recent work has shown that interactions between bacterial metabolism and growth explain IE in bacteria that lack acquired resistance, such as β-lactamases. Owing to the energetic costs associated with production, the expression of β-lactamases can perturb metabolism and growth rate, which may impact their relationship with IE. Indeed, the presence of a β-lactamase does not guarantee IE, suggesting additional aspects of bacterial physiology are at play. Using Escherichia coli expressing the NDM-1 β-lactamase, we quantified [ATP], growth rate, β-lactamase expression, and IE across multiple environments defined by the chemical composition of the growth medium. We found that expression of β-lactamases nonlinearly impacted both [ATP] and growth rate, but neither alone could predict IE. Instead, we found that the ratio of [ATP] to growth rate formed a biphasic relationship with the strength of IE. Mathematical modeling suggests that IE initially increases with the ratio of [ATP] to growth rate due to increased β-lactamase expression, which increasingly protects the population from antibiotics. However, once metabolism becomes sufficiently high, the ability of β-lactamases to protect the population is reduced, which decreases IE. Our results suggest that the chemical composition of the environment influences IE in β-lactamase-expressing bacteria, potentially resolving conflicting reports on the clinical relevance of IE. They may also aid in developing strategies to predict, prevent, and minimize IE.