Abstract
Modulation of growth rate in Mycobacterium tuberculosis is key to its survival in the host; particularly with regard to its adaptation during chronic infection when the growth rate is very slow. The resulting physiological changes will influence the way this pathogen interacts with the host and responds to antibiotics. Therefore, it is important that we understand how growth rate impacts antibiotic efficacy, particularly with respect to recovery/relapse. This is the first study that has asked how growth rates influence the mycobacterial responses to combinations of frontline antimycobacterials, isoniazid (INH), rifampicin (RIF), and pyrazinamide (PZA), using continuous cultures. Time-course profiles of log-transformed total viable counts for cultures, controlled at either a fast growth rate (23.1. mean generation time (MGT)) or slow growth rate (69.3h MGT), were analysed with the fitting of a mathematical model by nonlinear regression that accounted for the dilution rate in the chemostat, and profiled kill rates and recovery in culture. Using this approach, we show that populations growing more slowly were generally less susceptible to all treatments. We observed a higher kill rate associated with INH (compared to RIF or PZA) and the appearance of re-growth. In line with this observation, re-growth was not observed with RIF-exposure, which provided a slower bactericidal response. The sequential additions of RIF and PZA did not eliminate re-growth. We consider here that faster, early bactericidal activity is not what is required for successful sterilisation of M. tuberculosis, but instead slower elimination of bacilli followed by reduced recovery of the bacterial population.