Mechanisms of Latent Tuberculosis: Dormancy and Resuscitation of Mycobacterium tuberculosis

Abstract

Tuberculosis, caused by Mycobacterium tuberculosis , now kills more people in the world than any other single bacterial infection, and globally, one in three people is believed to harbor a persistent (latent) infection (1, 2) . The phenomenon of persistence has long been recognized (3–5) , but it remains poorly understood (6, 7) . It is generally agreed that the immune system plays an important role (8) . However, the microbiological aspect of a latent infection and especially the physiological status of persisting cells remain controversial. The subject has been extensively reviewed in recent years (6–7, 9–14) . Despite the intensive microbiological research the existence of specialized dormant mycobacterial cells is still being questioned. The main challenge in studying persisting mycobacteria is their low abundance in vivo . Therefore, great effort was made to develop conventional and reproducible in vitro models reflecting persistence state in vivo . Examples of such models include non-replicating persistence (NRP) under anaerobic conditions (7) , the starvation survival model (15) , and the stationary phase model (16) and its modification based on rifampin treatment of non-growing cells (17) . Despite the obvious similarities in survival patterns between NRP model and cell persistence in macrophages as confirmed by microarray analysis (18) , it is not clear if the NRP model reflects the formation of true dormant forms. Remarkably, attempts to treat tuberculosis with metronidazole (a drug effective against cells in the NRP state) were unsuccessful (19) . Moreover, recent studies on oxygen concentration in murine granulomas show significant oxygenation of granulomatous tissues. However, the degree of oxygenation of human granulomas is still not clear (20) . Mycobacteria surviving in vivo represent a highly heterogenic population, including growing, persisting, and dying cells. Some of them are probably exposed to starvation conditions. However, there is no clear evidence that the implication of starvation survival strategy will result in the production of specialized dormant forms. An antibiotic treatment model developed by Hu and colleagues (17) represents an interesting example of the so-called non-culturable (NC) forms. Nevertheless, it is not clear if the production of NC forms is a consequence of cell injury followed by treatment with rifampin or those cells were initially in the population and represented specialized forms. Therefore, a reliable dormancy model is a question of highest priority for studying the microbiological aspects of mycobacterial persistent infections. In a number of publications, we and other researchers have argued that dormant mycobacteria in vivo might adopt the so-called NC phenotype. Several important findings confirmed our suggestion, including the Cornell model of persistent infection (21, 22) and the isolation of NC cells from an in vivo environment (23) . In some cases, the isolated cells represented atypical forms with altered cell wall as judged by acid fast staining (24) . Recognizing the importance of reliable and reproducible models of mycobacterial persistence, we paid special attention in establishing specific conditions to obtain the NC forms of different bacteria, and in particular of mycobacterial NC forms.