Metabolic modeling predicts unique drug targets in Borrelia burgdorferi

Abstract

ABSTRACT The Lyme disease bacterium Borrelia burgdorferi is extremely host dependent, with a small genome and correspondingly limited metabolism. As such, it is an excellent candidate for the development of targeted, narrow-spectrum antimicrobials. To accelerate drug discovery in this fastidious bacterium, in silico genome-scale metabolic modeling was used to construct a map of B. burgdorferi’s metabolism. This map was used to predict essential genes and enzymes; experimental data validated these predicted hits as viable drug targets. Repurposing existing small-molecule inhibitors, it is shown that inhibition of two predicted essential enzymes (pyridoxal kinase and serine hydroxymethyltransferase) selectively kills B. burgdorferi in culture. Thus, the essential processes identified here represent targets for the development of narrow-spectrum antimicrobials. This pipeline, pairing in silico discovery with validation in culture, may be useful for other genetically intractable pathogens.IMPORTANCELyme disease is often treated using long courses of antibiotics, which can cause side effects for patients and risks the evolution of antimicrobial resistance. Narrow-spectrum antimicrobials would reduce these risks, but their development has been slow because the Lyme disease bacterium, Borrelia burgdorferi, is difficult to work with in the laboratory. To accelerate the drug discovery pipeline, we developed a computational model of B. burgdorferi’s metabolism and used it to predict essential enzymatic reactions whose inhibition prevented growth in silico. These predictions were validated using small-molecule enzyme inhibitors, several of which were shown to have specific activity against B. burgdorferi. Although the specific compounds used are not suitable for clinical use, we aim to use them as lead compounds to develop optimized drugs targeting the pathways discovered here.

Keywords

• antimicrobial agents
• Lyme disease
• drug discovery
• host-pathogen interactions
• metabolic modeling