Oxygen Depletion-Induced Dormancy in Mycobacterium bovis BCG

Anaerobic survival and synchronized reactivation of BCG.

Figure 1 shows the growth and survival of BCG in the Wayne dormancy model, i.e., under sealed stirred culture conditions. After an initial aerobic growth phase (19-h generation time) the culture entered an NRP-1 phase which was characterized by a slow increase of turbidity without an increase in viable counts. The culture then entered an NRP-2 phase in which the turbidity stayed constant. As is the case for M. tuberculosis, a slow depletion of oxygen was observed (Fig. 1). Fading of the oxygen indicator methylene blue did not start until after termination of NRP-1. Complete decolorization took about an extra 8 days. These results show that growth, adaptation, and survival of BCG under the Wayne dormancy culture conditions are essentially identical to those of the tubercle bacillus. To determine whether the persistent NRP-2 culture of BCG was synchronized, reactivation experiments were carried out. Anaerobic sealed cultures (at 20 days) were diluted 1:100 in fresh oxygen-rich medium and incubated with aeration. Figure 2 shows synchronous cell division upon reactivation after a lag phase of about 13 h.

Induction of glycine dehydrogenase.

To analyze the temporal profile of glycine dehydrogenase activity during growth under sealed stirred culture conditions, protein extracts were prepared at various time points by using a Bead Beater (Biospec). Glycine dehydrogenase activity was determined by an assay based on optical measurement of the rate of oxidation of NADH to NAD, a reaction that accompanies the reductive amination of glyoxylate to glycine (19). Protein concentrations were determined with a Bio-Rad kit. Figure 3 shows that the specific enzyme activity stayed at a low baseline level during the initial aerobic exponential growth phase (0.08 U mg⁻¹ h⁻¹). After termination of aerobic growth, glycine dehydrogenase activity increased rapidly during NRP-1 to a level sixfold greater than the baseline level. During NRP-2, the enzyme activity declined until it reached a level about 40% of that of the peak value. These results show that the temporal profile of glycine dehydrogenase activity in BCG grown under the Wayne culture conditions is strikingly similar to the profile observed for M. tuberculosis.

An increase of isocitrate lyase activity, a marker of the glyoxylate shunt, has been reported for oxygen-deprived cultures of sedimented tubercle bacilli (19). However, determination of isocitrate lyase activity (16) in aerated growing (79 ± 12 μmol mg⁻¹ min⁻¹) and anaerobic persistent 20-day-old BCG cultures (7 ± 1 μmol mg⁻¹ min⁻¹) showed that this enzyme activity is down regulated in anoxic BCG grown under the Wayne culture conditions. Isocitrate lyase activity of anoxic tubercle bacilli grown under the Wayne culture conditions has not been reported yet.

Induction of α-crystallin-like protein.

To assess whether the 16-kDa α-crystallin-like protein was induced in BCG under the conditions of the Wayne dormancy model, protein extracts from aerated exponentially growing and anaerobic persistent 20-day-old cultures were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Figure 4 shows that a high level of a 16-kDa protein was detectable in the extract from anaerobic persistent cultures but not in the extract from aerated growing cultures. To confirm the identity of the 16-kDa protein, the Coomassie-stained band was excised from the gel and digested with trypsin, and one of the tryptic peptides was sequenced on an Applied Biosystems Procise Sequencer. With the resulting amino acid sequence DGQLTIKA, a search of the M. tuberculosis genome database (Sanger Centre [2]) was carried out. The only match with 100% sequence identity was the α-crystallin-like protein. The induction of α-crystallin-like protein in BCG grown under the Wayne dormancy culture conditions is consistent with recent findings showing induction of this protein in BCG under oxygen-limiting conditions (3, 17, 24).

Development of metronidazole sensitivity.

To determine whether anaerobic dormant BCG develops sensitivity to metronidazole, cultures were incubated in the presence of various concentrations of the drug under sealed stirred conditions. Metronidazole was added at the beginning of the experiment, and survival of the bacilli was determined after 20 days of incubation. Metronidazole was found to have a dose-dependent bactericidal effect on anaerobic BCG. Figure 5A shows that 10 and 100 μg ml⁻¹ metronidazole reduced viable counts threefold and more than 1,000-fold, respectively. Aerated, exponentially growing cultures were not affected by the drug (Fig. 5C).

Conclusions.

We recently described a dormancy response, induced by oxygen depletion, for Mycobacterium smegmatis (7). This physiological response, first observed in M. tuberculosis (21), appears to be conserved between the slow-growing pathogen and the fast-growing saprophyte. There are, however, several molecular differences between the dormancy response in M. smegmatis and that of the tubercle bacillus (11, 12, 23). In this paper, we report that BCG is capable of adapting to anaerobiosis by shifting down to a persistent state in a manner similar to the tubercle bacillus. While the vaccine strain is slow growing and therefore less convenient for molecular genetic analyses, we demonstrate that its physiological behavior is strikingly similar to that of the closely related pathogenic M. tuberculosis. Furthermore, the expression of two molecular markers, α-crystallin-like protein and glycine dehydrogenase, in BCG is very similar to that in M. tuberculosis. It has also recently been demonstrated that microaerobically and anaerobically cultured BCG developed a thickened cell wall similar to that of M. tuberculosis (3). These morphological, physiological, and molecular observations suggest BCG may be a useful nonpathogenic model for the in vitro analyses of dormancy of the tubercle bacillus.