17Persisters constitute a subpopulation of bacteria that can tolerate lethal concentrations 18 of antibiotics. Multiple mechanisms have been suggested for bacterial persistence 19 against antibiotics. With mycobacteria being no exception to this behaviour, we had 20 reported the de novo emergence of genetically antibiotic-resistant Mycobacterium 21 tuberculosis from persister cells upon prolonged exposure to microbicidal 22 concentrations of the anti-tuberculosis drugs, rifampicin and moxifloxacin. Here, we 23 present evidence for reduced permeability to rifampicin as a mechanism for 24 persistence of Mycobacterium tuberculosis in vitro. We observed that rifampicin 25 persistent M. tuberculosis cells developed a thick outer layer (TOL) capsule. The TOL 26 restricted the entry of fluorochrome-conjugated rifampicin, 5-carboxyfluorescein-27 rifampicin (5-FAM-rifampicin), which retained only 2.5% of its original bactericidal 28 activity, but high levels of permeability, on actively growing mid-log phase cells. Gentle 29 mechanical removal of TOL significantly enhanced 5-FAM-rifampicin entry into the 30 persister cells. The level of 5-FAM-rifampicin in the persister cells was not affected by 31 the pre-incubation of the cells with verapamil, a drug efflux pump inhibitor, ruling out 32 the involvement of efflux pumps in the reduced intracellular concentration of 5-FAM-33 rifampicin. GC-MS analysis of TOL showed the presence of ~7-fold, ~5-fold and ~2-34 fold higher levels of α-D-glucopyranoside, 1,2,5-linked-mannitol, and 3,4-linked 35 mannose, respectively, among ~2-fold higher levels of derivatives of several other 36 types of sugars such as arabinose and galactose. Taken together, the present study 37 reveals that rifampicin-persistent M. tuberculosis cells develop TOL that enables the 38 bacilli to restrict entry of rifampicin and thereby remain tolerant to the antibiotic in vitro.39 40 65 4unknown strategy to ensure sub-lethal concentration of antibiotics inside persister cells 66 even though they remain exposed to lethal concentrations of the antibiotics. In order 67 to verify this possibility, we examined whether the permeability of rifampicin into M. 68 tuberculosis persister cells was affected by any ultrastructural changes. The present 69 study reports the ultrastructural changes of rifampicin persistent M. tuberculosis cells 70 and their contribution to the restricted permeability of rifampicin, in comparison with 71 the entry of rifampicin into actively growing mid-log phase cells. 72 73 MATERIALS AND METHODS 74 75 Rifampicin treatment of Mycobacterium tuberculosis 76 Mycobacterium tuberculosis H37Ra (National JALMA Institute of Leprosy and Other 77 Mycobacterial Diseases, Agra, India) was used in all the experiments. Bacteria were 78 grown in Middlebrook 7H9 broth (BD) supplemented with 0.2% glycerol, 0.05% 79 Tween-80 and 10% ADS (albumin, dextrose, sodium chloride) until mid-log phase 80 (MLP, 0.6 at OD600 nm) at 37ºC under shaking at 170 rpm. Cultures were then exposed 81 to 1 µg/ml (10x MBC; 12) of ...