Mycobacterium tuberculosis recovers slowly after TB drugs in mice

Treating tuberculosis relies not only on drugs that kill bacteria but also on the way bacteria recover when drug exposure stops. Scientists use the phrase post-antibiotic effect (PAE) to describe the delayed return of bacteria after antibiotic exposure; it helps explain why some TB regimens can tolerate missed doses. Jo Hendrix and colleagues set out to study this recovery process directly inside animals rather than relying only on in vitro counts. Using the BALB/c mouse model, they gave sub-curative courses lasting two or four weeks of the standard combination isoniazid, rifampin, pyrazinamide, ethambutol (HRZE). Instead of measuring only how many bacteria were present, the team examined Mycobacterium tuberculosis (Mtb) physiology during the weeks after treatment stopped. They measured rRNA synthesis with the RS ratio® and read the whole bacterial transcriptome with SEARCH-TB to capture which genes and processes were turned on or off during recovery. This approach was designed to reveal whether PAE reflects bacteria staying fewer in number or staying in a damaged state that takes time to repair.

The study paired two complementary molecular tools to follow Mtb after stopping HRZE. First, rRNA synthesis measured by the RS ratio® showed that ribosomal RNA production resumed within four days after treatment interruption, indicating a quick restart of at least one core activity. Second, whole-transcriptome profiling with SEARCH-TB showed a more complex picture: changes in gene expression tied to metabolic reactivation were delayed for more than two weeks. Over 28 days of drug-free recovery, Mtb burden did not increase, demonstrating prolonged PAE in vivo even though some molecular activity returned sooner. Key processes needed for replication—including expression of genes involved in protein and cell wall synthesis—remained suppressed throughout the full 28-day period. The longer, four-week HRZE courses caused deeper physiologic disturbance and were linked to slower and less complete recovery. Interestingly, processes typically induced by environmental stress (e.g., DosR regulon, universal stress proteins, and heat shock proteins) fell during drug treatment and rose during recovery, reversing expectations.

These results reframe PAE as a matter of drug-induced injury and staggered physiologic repair rather than just a pause in growth. After relatively short courses of HRZE, recovery in Mtb was slow, sequential and incomplete in this mouse model, with some functions restarting quickly and others, especially those tied to actual replication, staying repressed for weeks. The observation that longer treatment caused even slower and less complete recovery suggests that Mtb may progressively lose the capacity to bounce back as treatment depth increases. By using RS ratio® and SEARCH-TB together, Jo Hendrix and the team established a workable experimental framework to quantify how forgiving TB regimens are in vivo — meaning how well treatments tolerate missed doses — which could be applied to evaluate new regimens in animals before clinical testing.