ABSTRACT
Mycobacteria store triacylglycerols (TGs) under various stress conditions, such as hypoxia, exposure to nitric oxide, and acidic environments. These stress conditions are known to induce nonreplicating persistence in mycobacteria. The importance of TG accumulation and utilization during regrowth is not clearly understood. Here we specifically determined the levels of accumulated TG and TG lipase activity in Mycobacterium bovis bacillus Calmette-Guerin (BCG) in various different physiological states (logarithmic growth, aerated stationary phase, hypoxia-induced dormancy, and regrowth from dormancy). We found extensive accumulation and degradation of TGs in the bacilli during entry into and exit from hypoxia-induced dormancy, respectively. These processes are accompanied by dynamic appearance and disappearance of intracellular TG lipid particles. The reduction in TG levels coincides with an increase in cellular TG lipase activity in the regrowing bacilli. Tetrahydrolipstatin, an inhibitor of TG lipases, reduces total lipase activity, prevents breakdown of TGs, and blocks the growth of mycobacteria upon resuscitation with air. Our results demonstrate that utilization of TGs is essential for the regrowth of mycobacteria during their exit from the hypoxic nonreplicating state.
Mycobacterium tuberculosis is the causative agent of tuberculosis (TB), one of the major infectious diseases, which affects one-third of the world's population (http://www.who.int/mediacentre/factsheets/fs104/en/ ). The majority of TB patients carry a latent infection. However, reactivation leading to active disease (16, 27) often occurs once the host defenses are weakened. During the latency period, mycobacteria are tolerant to many conventional antibiotics (23, 28), thus making eradication of TB challenging.
In the human body, M. tuberculosis is believed to persist in lung lesions with hypoxic environments (6, 27). Wayne and Hayes established an “in vitro dormancy model” in which mycobacterial cultures are subjected to gradual depletion of oxygen, which causes the obligate aerobic cells to exit the cell cycle and enter into a nonreplicating persistent state (26). The bacilli in the nonreplicating persistent state are phenotypically drug resistant. Recent efforts to explore the mechanisms which allow the tubercle bacilli to enter into dormancy and survive in the host for a long period of time suggest that fatty acids (FAs) could be an important source of energy during the persistent state (1, 17, 20). In particular, triacylglycerols (TGs), a class of neutral lipids, are postulated to be a likely source of FAs (8). TGs are an efficient form of energy reserves in many organisms during long-term survival.
Recently, it was reported that tubercle bacilli in sputum from patients with TB contain lipid bodies (11). Moreover, TGs accumulate in hypervirulent W-Beijing strains of M. tuberculosis (19). It is interesting that TGs accumulate in these clinical strains of TB. However, no systematic study has been carried out yet to investigate the accumulation and degradation of TG species under various physiological conditions in mycobacteria. Here we used M. bovis bacillus Calmette-Guerin (BCG) to study the kinetics of TG buildup and breakdown during different growth phases, including logarithmic growth, aerated stationary phase, hypoxia-induced dormancy (Wayne model), and regrowth from dormancy upon reaeration of hypoxic cultures. Utilizing tetrahydrolipstatin (THL) as a chemical probe, we showed that mobilization of TGs is essential for the regrowth of mycobacteria during recovery from the hypoxia-induced dormant state.
MATERIALS AND METHODS
Bacterial strains and culture conditions.All experiments were carried out with the M. bovis BCG Pasteur strain (ATCC 35734) at 37°C. The liquid medium used was Dubos broth (supplemented with 10% Dubos medium albumin and 0.03% Tween 80). The solid medium used was Middlebrook 7H11 agar (supplemented with 10% oleic acid-dextrose-albumin-catalase enrichment and 0.5% glycerol). All media were purchased from Difco.
Mycobacterial cultures.Culture conditions, including logarithmic, aerated stationary, and hypoxia-induced dormancy conditions, were adapted from the conditions described by Boon et al. (4). Inocula used for starting various cultures were prepared using growing cells. Essentially, aerobic M. bovis BCG cultures were grown in roller bottles in which the initial optical density at 600 nm was 0.05. The logarithmic and aerated stationary cultures were rotated at 1 rpm for 3 and 10 days, respectively. Mycobacterial cultures were subjected to slow withdrawal of oxygen as described by Wayne and Hayes (26) (Wayne model) and grown using an initial optical density at 600 nm of 0.005 in screw-cap test tubes (20 mm by 125 mm) with a final volume of 17 ml. The Wayne model cultures were sealed by tightly screwing on solid caps with latex liners and were stirred gently at 170 rpm on magnetic stirring platforms for 20 days to allow the bacilli to enter into dormancy. Fresh air was reintroduced into hypoxic cultures by loosening the caps for exit from dormancy. This exposure to air allowed dormant bacilli to regrow.
Growth and survival were monitored by enumeration of CFU on Middlebrook 7H11 agar after plating of appropriate dilutions.
Exogenous FA treatment of mycobacterial cultures.A fluorescent FA, 1-pyrenedecanoic acid (PDA), was acquired from Marker Gene Technologies Inc. (Eugene, OR). PDA (excitation wavelength, 340 nm; emission wavelength, 377 nm) at a final concentration of 100 μM was added to the Wayne model cultures at the start of an experiment. Cells were harvested after 20 days of incubation for analysis.
Lipase inhibitor treatment of mycobacterial cultures.Tetrahydrolipstatin (THL) (Sigma Chemicals, St. Louis, MO) was added to mycobacterial cultures (in the logarithmic, aerated stationary, hypoxia-induced dormancy, and regrowth phases) for a 3-day treatment at a fixed concentration of 80 μM. Cells were harvested at the beginning and end of the treatment for analysis.
Lipid extraction.Total lipids were extracted from M. bovis BCG cells or supernatant using a modified Bligh-Dyer method (2). Bacilli (4.0 × 108 cells) were pelleted by centrifugation (3,000 × g for 10 min) and washed twice with phosphate-buffered saline containing 0.05% Tween 80 (PBST). Chloroform-methanol (6 ml; 2:1, vol/vol) was added to the cell pellet. Lipids were extracted at 4°C for 24 h before 4 ml of water was added for separation of the phases by centrifugation at 3,000 × g for 20 min. Cell culture supernatants were used for extraction of lipids from media. Chloroform-methanol (2:1, vol/vol) was added to the supernatant at a ratio of 3:2, and this was followed by centrifugation at 3,000 × g for 20 min. The lower organic phase was dried under a stream of nitrogen and stored at −20°C until analysis.
Nile red staining and confocal microscopy.Nile red (Invitrogen, Carlsbad, CA) was used according to previously described protocols (12). Briefly, 7.0 × 107 cells were washed twice with PBST before fixation with 4% paraformaldehyde for 10 min. Next, the cells were labeled with 10 μg/ml of Nile red stain (excitation wavelength, 550 nm; emission wavelength, 650 nm) for 5 min, and the excess dye was washed off. The cells were then mounted on slides in Fluorsave reagent (Calbiochem, Gibbstown, NJ). The slides were observed by confocal microscopy using a Zeiss LSM 510 Meta confocal microscope equipped with a 543-nm, 1-mW HeNe laser as the excitation source. Images were viewed using an EC Plan-Neofluar 100×/1.30 oil objective lens (Zeiss) and were captured using a photomultiplier tube with an LP615nm filter.
Thin-layer chromatography (TLC).Dried lipid films were resuspended in 300 μl of chloroform-methanol (1:1, vol/vol). Ten-microliter portions of the resuspended lipid extracts were spotted onto Silica Gel 60-coated glass plates (20 by 20 cm; Merck, Darmstadt, Germany). The plates were developed using hexane-diethyl ether-formic acid (45:5:1, vol/vol/vol) as the solvent system. Iodine vapor was used to reversibly stain the lipids. Triolein (Sigma Chemicals, St. Louis, MO) was used as a standard for TGs. Defined amounts of triolein were spotted to obtain calibration curves using densitometry (Image J v1.37; National Institutes of Health, Bethesda, MD).
Liquid chromatography (LC)-MS of TGs.A sensitive high-performance liquid chromatography (HPLC)-electrospray ionization (ESI)-mass spectrometry (MS) method was developed to separate TGs from polar lipids, using an Agilent Zorbax Eclipse XDB-C18 column (4.6 by 150 mm). Analysis of TGs was performed with an Agilent 1100 HPLC system coupled to a 4000 Q-Trap mass spectrometer (Applied Biosystems, Foster City, CA). In brief, the HPLC conditions were as follows: mobile phase, chloroform-methanol-0.1 M ammonium acetate (100:100:4) at a flow rate of 0.25 ml·min−1; column temperature, 25°C; and injection volume, 20 μl. MS was performed using positive ESI modes. The ESI conditions were as follows: turbospray source voltage, 5,000 V; source temperature, 250°C; scan rate, 1,000 atomic mass units/s; declustering potential, 30.00 V; and scan range, 300 to 1,200 Da. Dried lipid extracts were resuspended in the HPLC mobile phase. A total run time of 30 min was utilized to elute both polar lipids and nonpolar TGs from the column, and the elution times of TGs were averaged for comparisons of TG profiles for mycobacteria under different growth conditions. The parameters used for HPLC-ESI-tandem MS were the same as those described above except that the collision energy was set at 40 V. For quantitation of the total TG content, the intensities (counts) of different TG peaks were added using a Matlab program.
TG lipase assay.Mycobacterial cells were washed twice with PBST and resuspended in lipase buffer (100 mM glycine, 19 mM sodium deoxycholate; pH 9.5). Cell homogenates were prepared by subjecting the cells to three cycles of 1 min of bead beating separated by 3-min intervals on ice. The resulting cell homogenates were stored at −80°C until analysis.
Fluorescent lipase assay kits were purchased from Marker Gene Technologies, Inc. (Eugene, OR). Measurement of TG lipase activity was performed according to the manufacturer's protocol. Briefly, 50 μg of lysate was incubated with a fluorescent TG substrate, 1,2-dioleoyl-3-pyrenyldecanoyl-rac-glycerol. TG lipase activity was measured by monitoring the formation of PDA (excitation wavelength, 341 nm; emission wavelength, 377 nm), which is released from 1,2-dioleoyl-3-pyrenyldecanoyl-rac-glycerol upon enzymatic degradation, over a 30-min period. TG lipase activity was determined by determining the increase in the number of relative fluorescence units of PDA over a fixed period of time. A blank without any protein was used as a control.
RESULTS
Accumulation and degradation of TG lipids under various physiological conditions.The kinetics of TG accumulation under various growth conditions (logarithmic, aerated stationary, hypoxia-induced dormancy, and regrowth) was analyzed using TLC (Fig. 1A). During aerobic exponential growth, very little TG was present in the bacilli (4.9 ng per 106 cells), while no TG was detected in late stationary phase. On the other hand, TGs accumulated during hypoxia-induced dormancy, as has been reported previously for M. tuberculosis H37Rv (8). The pool of TGs was reduced dramatically once dormant cells underwent regrowth upon exposure to oxygen.
TG metabolism in M. bovis BCG during logarithmic, aerated stationary, hypoxia-induced dormancy, and regrowth phases. (A) TLC analysis of TGs in M. bovis BCG in different physiological states. Equal numbers of cells were harvested, and lipids were extracted using chloroform-methanol, separated on silica gel plates, and stained with iodine vapor. The values are the amounts of TGs as determined by densitometry using a triolein standard calibration curve. (B) TLC analysis of neutral lipids extracted from supernatants of M. bovis BCG under different growth conditions. Note that the disappearance of TGs in the cells during the regrowth phase is not due to export of the lipids into extracellular media.
We next used LC-MS to determine the predominant TG species and their FA compositions. This sensitive approach revealed at least nine major species of TGs in M. bovis BCG (Table 1). The predominant species were TG with mass/charge (m/z) ratios of 875 and 905. FA compositions were confirmed using tandem MS (Table 1). C16:1, C18:1, and C26:0 are the major TG species in dormant M. bovis BCG (Table 1). The total TG intensities (sum for all nine TG species) as determined by LC-MS (data not shown) correlated well with the results obtained by TLC (Fig. 1A).
Predominant TG species which accumulated in M. bovis BCG during dormancy
An attempt was made to monitor the presence of lipid droplets (consisting mainly of TGs) in M. bovis BCG in different physiological stages. In our studies, all of the dormant M. bovis BCG cells were positively stained with the Nile red dye, indicating the presence of TG lipid bodies. None of the bacilli under aerated conditions contained any Nile red-positive lipid droplets. These data thus further demonstrate the accumulation and disappearance of TGs in dormant and regrowing cultures of M. bovis BCG, respectively.
M. bovis BCG incorporates exogenous FAs into its intracellular TG store.We used LC-MS to examine the FA composition of TGs and found that m/z 875 and m/z 905 ions represented major TG species under our culture conditions (Table 1). Both of these lipids contain mainly C18:1. Since oleic acid is present in the culture medium (Dubos and Tween 80), we postulate that the bacilli could use exogenous FAs as a substrate for TG synthesis. To validate this hypothesis, a fluorescent FA, PDA, was added to the Dubos media during hypoxia-induced dormancy. Confocal microscopy revealed that the lipid droplets were labeled with PDA (Fig. 2A). Closer examination of LC-MS data also showed the appearance of various novel TG species when M. bovis BCG was treated with PDA (Fig. 2B). These TG species have m/z of 1022, 1051, 1081, and 1107, which correspond to addition of m/z 90 to endogenous m/z 932, 961, 991 and 1017 TG species, respectively. It is noteworthy that native m/z 818, 846, 875, and 905 TG species also have an m/z 90 addition in the PDA-treated bacilli. However, these species are less obvious in the mirror plot as the resulting m/z ions are m/z 908, 936, 965, and 995 ions, which are almost identical to the endogenous TG species. Collectively, it was observed that PDA (molecular weight, 372.5) can competitively replace oleic acid (molecular weight, 282.5) as one of the FA substrates for TG synthesis.
Incorporation of exogenous fluorescent FA into intracellular stored TGs in M. bovis BCG during the hypoxia-induced dormancy phase. (A) Confocal images of dormant M. bovis BCG cells. Bacilli were treated with or without PDA (100 μM) for 20 days (Wayne model) before they were stained with Nile red dye. The panels (from top to bottom) show fluorescent images of Nile red signals (excitation wavelength, 550 nm; emission wavelength, 650 nm), PDA signals (excitation wavelength, 341 nm; emission wavelength, 377 nm), dual signals (overlay), and transmission microscope images of M. bovis BCG. Scale bar = 2 μm. The images are representative of the bacilli in at least 10 different microscopic fields. (B) Mirror plot showing LC-MS analysis of TG species in dormant M. bovis BCG cultured in the presence (top spectrum) or absence (bottom spectrum) of PDA (100 μM). Novel TG species corresponding to replacement of oleic acid (molecular weight, 282.5) with PDA (molecular weight, 372.5) are indicated. The TG species (m/z 829.9) which served as an internal standard (IS) was added during lipid extraction in order to normalize intensities for two different treatments.
Fate of TGs during regrowth from hypoxia-induced dormancy.Since the size of intracellular TG pools decreases during the regrowth period, we investigated if the disappearance of TGs could be related to export of TGs into the extracellular media. Supernatants from hypoxic nongrowing and regrowing cells were analyzed to determine their TG contents. TLC profiles revealed that these supernatants contained hardly any detectable amounts of TGs (Fig. 1B). Thus, if intact TG lipids were excluded from the cells, they would have appeared in the extracellular media. This suggested that TGs were not exported out of the cells during regrowth.
If intracellular TG pools are hydrolyzed by enzymatic degradation, the activity of cellular TG lipases should increase accordingly. Thus, we measured total TG lipase activity under each growth condition using a fluorescent lipase assay. Indeed, the total TG lipase activity (Fig. 3) is inversely correlated with the amount of TGs in M. bovis BCG (Fig. 1A), indicating that the bacilli upregulate their lipase activities to utilize TGs upon regrowth.
Reduction in cellular TG levels is correlated with increased lipase activity: total TG lipase activities of mycobacterial extracts in the logarithmic, aerated stationary, hypoxia-induced dormant, and regrowth phases. Blank refers to a control incubated in the absence of mycobacterial proteins. The bars indicate means, and the error bars indicate standard deviations (n = 3). *, P < 0.05 for a comparison with the lipase activity in the logarithmic phase; **, P < 0.05 for a comparison with the lipase activity in the hypoxia-induced dormant phase. RFU, relative fluorescence units.
THL inhibits the TG metabolism of M. bovis BCG during regrowth from hypoxia-induced dormancy.THL (commonly known as Orlistat; marketed as Xenical by Roche, Nutley, NJ) is approved for treating obesity. This drug acts as an irreversible inhibitor of pancreatic and gastric lipases (5, 13). THL has also been shown to have bactericidal effects on some mycobacteria (14) and some activity against an extracellular lipase encoded by Rv0183 in M. tuberculosis (7).
To determine if THL has inhibitory activity against TG lipases in M. bovis BCG, we determined its inhibitory effect in total-cell lysates. The concentration of THL required to inhibit 50% of TG lipase activity was found to be approximately 40 μM (Fig. 4). In contrast, isoniazid (INH), a standard anti-TB drug, did not inhibit the lipase activity (data not shown).
THL inhibits total TG lipase activity. Cell lysate of a regrowing mycobacterial culture was incubated with THL for 30 min prior to determination of the total TG lipase activity. The data are means ± standard deviations (n = 3). RFU, relative fluorescence units.
The effect of THL on TG metabolism in different physiological states was also studied. Cultures of M. bovis BCG in different growth states were incubated with THL (80 μM) for 3 days. It was found that THL prevents the breakdown of TGs during exponential growth, hypoxic phases, and regrowth (Fig. 5A and 5B). The dynamics of TG metabolism were correlated with the presence of Nile red-positive intracellular lipid deposits (Fig. 5C). Collectively, these results demonstrate that THL prevents the degradation of TGs during regrowth of M. bovis BCG.
THL prevents the breakdown of TGs and lipid bodies during regrowth. The cultures used contained bacilli in the logarithmic (d3A and d6A), aerated stationary (d10A and d13A), hypoxia-induced dormant (d20 and d23), and regrowth (d3R) phases (d, day; A, aeration; R, regrowth). All cultures were treated with or without THL (80 μM) for 3 days. (A) TLC analysis of TG levels in M. bovis BCG under different growth conditions and with different drug treatments. The values are the amounts of TGs as determined by densitometry. (B) LC-MS analysis of the total TG content in lipid extracts of M. bovis BCG under different conditions (n = 3). *, P < 0.05 for a comparison with the TG level in the regrowth phase without THL treatment. (C) Confocal images of M. bovis BCG cells stained with Nile red dye. The panels show transmission microscope images of M. bovis BCG integrated with a fluorescent signal (red). Scale bar = 2 μm. The percentages indicate the percentages of positively stained bacilli among 500 bacilli counted in at least 10 different microscopic fields. a.u., arbitrary units.
Regrowth of hypoxia-induced dormant M. bovis BCG is attenuated by THL treatment.The effect of THL on the growth and survival of M. bovis BCG was determined by incubating BCG cells in various growth conditions in the presence or absence of THL. This drug had a mild effect on growth in the logarithmic phase (Fig. 6). It had no inhibitory effect on aerated stationary cultures, which is consistent with the finding that TG metabolism seems to be minimal in this physiological state. Conversely, a significant effect on regrowing bacilli was observed, as the cells were unable to replicate when they were treated with the lipase inhibitor. Furthermore, THL also affected the survival of hypoxic dormant bacilli. These results show that THL is able to attenuate regrowth of mycobacteria.
THL inhibits regrowth of M. bovis BCG from hypoxia-induced dormancy. The numbers of CFU were determined under various growth conditions, and then appropriate dilutions of cells were plated. Note the approximately 10-fold reduction in the number of CFU upon THL treatment under regrowth and dormant conditions. The bars indicate the means, and the error bars indicate the standard deviations (n = 3). *, P < 0.05 for a comparison with the number of CFU in the regrowth phase without drug treatment. d, day; A, aeration; R, regrowth.
Our data demonstrated that THL was able to inhibit TG breakdown, thereby preventing the bacilli from regrowing after hypoxia-induced dormancy. However, it is possible that the lack of TG degradation could be just a side effect of the growth inhibition and not the primary reason for this effect. In order to address this issue, we used an effective antimycobacterial drug with a known mechanism of action as a reference control. INH inhibits the biosynthesis of FAs and mycolic acids very well (14, 24). Moreover, based on work done with Mycobacterium kansasii, Kremer and colleagues concluded that INH and THL are not likely to share the same target(s) (14). INH can inhibit growing and regrowing BCG very well (data not shown). This is expected as INH affects cell wall synthesis that occurs only during the growth phase. When we examined TG profiles, we observed that INH does not prevent any TG hydrolysis during the regrowth phase (data not shown). On the other hand, THL inhibits both cell growth and TG degradation during the regrowth phase. Therefore, our data demonstrated that the inhibition of TG breakdown significantly affects the regrowth of M. bovis BCG after a hypoxia-induced nonreplicating phase.
DISCUSSION
The success of M. tuberculosis as a human pathogen lies in its capacity to remain dormant after an initial infection. The tubercle bacillus reactivates when the host immune defenses are compromised (16). This pathogen is believed to reside in granulomas during latency, and hypoxic environments within these lesions could act as stimuli for bacilli to change their metabolic states (27). In view of this, oxygen limitation was used as a means to induce tubercle bacilli into a state of dormancy (26). On the other hand, to compel the cells to resume replication, fresh air can be reintroduced into an hypoxic dormant mycobacterial culture. This method may mimic the reactivation of tubercle bacilli during reinfection in the host.
TGs were shown to accumulate in M. tuberculosis H37Rv during hypoxia-induced dormancy and after acidic and NO treatments (8, 22). Importantly, upon nutrient starvation, hypoxic dormant bacilli hydrolyze deposits of TGs (9), indicating that these lipids have an important role as a carbon reserve during periods of nongrowing persistence and reactivation. Furthermore, stored TG builds up in exponentially growing hypervirulent W-Beijing family strains (19). This suggests that TGs could provide a growth advantage to the W-Beijing family strains over other strains during latency since they have ready access to stored FAs. Intracellular TG particles are found in acid-fast stained bacilli derived from sputum of TB patients (10). These lipid particles are also proposed to be a biomarker for nonreplicating persistence (11) as a strong correlation between the intracellular TGs and nonreplicating dormancy has been established. However, in spite of all these findings, the importance of TGs for survival or growth during different growth stages has not been clearly defined yet.
Our results demonstrate that there is extensive accumulation and degradation of TGs in M. bovis BCG during entry into and exit from hypoxic conditions, respectively (Fig. 1). Lipid droplets, the likely form of TGs inside bacilli, correlate with TG dynamics (Fig. 5C), and such kinetics do not occur in aerobic conditions. The accumulation of TGs during dormancy has led workers to postulate that these neutral lipids are important for survival during latency and reactivation. Importantly, if TGs are utilized for energy consumption during regrowth, their hydrolysis would be required for release of FAs. Indeed, the reduction of intracellular TG particles during regrowth is accompanied by increased TG lipase activity (Fig. 3), suggesting that TGs are utilized as a source of energy or as a resource by the reactivating bacilli. Our findings thus indicate that TGs could act as a “reservoir” of FAs for utilization during regrowth from hypoxia-induced dormancy.
Accumulation of TGs has been reported previously in dormant mycobacteria (8, 10, 11, 22). The exact identity of TG species, however, remains poorly characterized. TGS1 (Rv3130c) seems to preferentially incorporate C26:0 into TGs (22), and it has been reported that the C26:0 FA is a major component of TGs under stress conditions (22). Using LC-MS, our study showed that m/z 875 and m/z 905 TG ions represent approximately 50% of all TG species during hypoxia-induced dormancy (Table 1). The main FA species is C18:1, which is also esterified as Tween 80 found in the media. Furthermore, we found that the supplemented fluorescent FAs are able to compete with C18:1 for incorporation into the stored TG of dormant bacilli (Fig. 2). These results thus indicate that M. bovis BCG is able to take up exogenous FAs for the synthesis of intracellular TGs during hypoxia-induced dormancy. Our data are in line with the view that mycobacteria do require host lipids as a source of energy for survival during persistent infection (1, 20). Pertinent to this, an extracellular lipase, Rv0183, has been characterized (7), suggesting that the tubercle bacilli may hydrolyze lipids from the host tissue. Upon hydrolysis of host lipids, liberated FAs could be taken up by the bacilli and incorporated into intracellular TGs for storage as an energy or carbon source. Such a lipid degradation pathway might contribute to the viability of mycobacteria during infection, since these bacilli prefer to utilize FAs when they grow in vivo (3, 21).
Utilization of TGs as a carbon or energy source requires the activation of TG lipases. Genes encoding approximately 24 TG lipases have been predicted to be present in the genome of M. tuberculosis (9), and the lipY (Rv3097c) and Rv0183 products have been experimentally shown to possess TG lipase activity (7, 9). THL, an inhibitor of pancreatic lipases, reduced total lipase activity and prevented breakdown of TGs and regrowth of mycobacteria when they were resuscitated with oxygen (Fig. 4 and 5). Intriguingly, dormant bacilli also have a lower survival rate when they are treated with a TG lipase inhibitor during hypoxia. Close examination showed that the TG content increased slightly after THL treatment (Fig. 5). This increase in the TG pool could be attributed to prevention of TG hydrolysis during dormancy, which might be required for viability (Fig. 6). This effect is minor compared to regrowth, which, at least in part, can be explained by the slower metabolism of dormant M. bovis BCG.
In order to validate our pharmacological approach for probing the importance of TG utilization during the regrowth period, genetic complementation was carried out. LipY has been reported to be an enzyme for utilization of stored TG in a nutrient-deprived environment (9). Therefore, we knocked out lipY (BCG3122) in M. bovis BCG in an attempt to generate a mutant deficient in TG hydrolysis. Our findings revealed that a lipY-deficient mutant has a significantly higher level of stored TG and decreased TG lipase activity compared to wild-type bacilli during the first day of the regrowth phase (data not shown). The regrowth rate of the mutant was also affected as we observed lower numbers of CFU compared to the numbers of wild-type CFU (data not shown). The lipY-deficient mutant exhibited a more pronounced phenotype only during the initial regrowth period, and this indicates that there may be other lipases that contribute to the degradation and utilization of TG in M. bovis BCG (9). From the observations described above, it is apparent that the utilization and degradation of TGs are essential for the bacilli to exit from hypoxia-induced dormancy.
We observed that THL at a concentration of 80 μM was able to block regrowth of nonreplicating bacilli. However, the concentration that blocked 50% of TG lipase activity was 40 μM. Furthermore, a microarray analysis also showed that THL treatment induced upregulation of a few putative lipases in the tubercle bacilli (25). It is thus conceivable that THL inhibits multiple lipases. These enzymes are likely to complement one another during metabolism as gene redundancy is common in tubercle bacilli.
Although we obtained evidence that THL prevents regrowth of mycobacteria via inhibition of TG lipases, we cannot exclude the possibility that there are off-target effects. Indeed, THL inhibits a thioesterase encoded by Rv3802c, an enzyme involved in mycolic acid biosynthesis (18), and was also shown to inhibit FA synthase in tumor cells (15). However, THL only partially inhibits overall mycolic acid formation compared to a cell wall synthesis inhibitor, INH (14). In addition, we did not observe major changes in overall levels of mycolic acids and phospholipids using LC-MS technology with THL-treated bacilli during regrowth (data not shown). Hence, we attribute inhibition of mycobacterial regrowth by THL to prevention of TG hydrolysis.
In summary, this study shows that TG accumulation, maintenance, and hydrolysis occur as bacilli enter, remain in, and exit hypoxia-induced dormancy, respectively. At these stages, TG lipase activity is regulated to utilize TGs for cell survival. The bacteriostatic effect of THL on mycobacteria implies that enzymes involved in TG hydrolysis could represent a potential drug target for reactivating bacilli.
ACKNOWLEDGMENTS
This research was supported in part by the Singapore National Research Foundation under CRP award 2007-04, by the Academic Research Fund (grant R-183-000-160-112), and by the Novartis Institute for Tropical Diseases (grant R-183-000-166-592). L.K.L. was a recipient of a scholarship from the NUS Graduate School for Integrative Sciences and Engineering.
We thank Yong Xiao and Candy Zhuang for their technical advice on confocal microscopy.
FOOTNOTES
- Received 21 April 2009.
- Accepted 2 June 2009.
- Copyright © 2009 American Society for Microbiology
