Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Jan 20.
Published in final edited form as: Cell Host Microbe. 2011 Jan 20;9(1):21–31. doi: 10.1016/j.chom.2010.12.004

Virulence of Mycobacterium tuberculosis depends on lipoamide dehydrogenase, a member of three multi-enzyme complexes

Aditya Venugopal 1,2, Ruslana Bryk 1, Shuangping Shi 1,2, Kyu Rhee 3, Poonam Rath 1,2, Dirk Schnappinger 1, Sabine Ehrt 1,2, Carl Nathan 1,2,4
PMCID: PMC3040420  NIHMSID: NIHMS261589  PMID: 21238944

SUMMARY

Mycobacterium tuberculosis (Mtb) adapts to persist in a nutritionally limited macrophage compartment. Lipoamide dehydrogenase (Lpd), the third enzyme (E3) in Mtb’s pyruvate dehydrogenase complex (PDH), also serves as E1 of peroxynitrite reductase/peroxidase (PNR/P), which helps Mtb resist host reactive nitrogen intermediates. In contrast to Mtb lacking dihydrolipoamide acyltransferase (DlaT), the E2 of PDH and PNR/P, Lpd-deficient Mtb is severely attenuated in wild type and immunodeficient mice. This suggests that Lpd has a function that DlaT does not share. When DlaT is absent, Mtb upregulates an Lpd-dependent branched chain keto-acid dehydrogenase (BCKADH) encoded by pdhA, pdhB, pdhC and lpdC. Without Lpd, Mtb cannot metabolize branched chain amino acids and potentially toxic branched chain intermediates accumulate. Mtb deficient in both DlaT and PdhC phenocopies Lpd-deficient Mtb. Thus, Mtb critically requires BCKADH along with PDH and PNR/P for pathogenesis. These findings position Lpd as a potential target for anti-infectives against Mtb.

INTRODUCTION

Most cases of tuberculosis are curable by drugs that have been available for over 40 years, yet Mtb is the single leading cause of death from bacterial infection (Bloom and Murray, 1992). This medical-societal failure stems in part from using anti-infectives that are chiefly active against replicating bacteria to treat an infection in which many of the bacteria are not replicating (McCune et al., 1966; Nathan, 2009; Nathan et al., 2008). During a multi-decade innovation gap in anti-infective drug development (Fischbach and Walsh, 2009), the inexorable progression of drug-resistance has made tuberculosis increasingly harder to treat and sometimes untreatable (Chan et al., 2008; Glaziou et al., 2009). Thus, a basic biologic question has acquired urgent medical relevance: on what gene products does Mtb depend for survival in the host? This is distinct from how the question is usually framed in anti-bacterial drug development: on what gene products does the pathogen depend for survival in vitro?

Whether Mtb infection is latent or active, many of the bacilli reside in macrophage phagosomes (Russell, 2001), where they can sustain net rates of population growth near zero for long periods (Gill et al., 2009; Munoz-Elias et al., 2005). Knowledge is growing about the metabolic pathways required for survival and persistence of intraphagosomal Mtb (Barry et al., 2009; Smith, 2000). Mtb requires the enzymes isocitrate lyases 1 and 2 (encoded by icl1 and icl2 and serving the glyoxylate shunt and methylcitrate cycle) to establish infection in the mouse (Munoz-Elias and McKinney, 2005), and genes required for cholesterol metabolism to persist (Pandey and Sassetti, 2008). Additionally, upregulation of gluconeogenic and glyoxylate shunt genes in Mtb recovered from mouse macrophages (Schnappinger et al., 2003), mouse lungs and patient specimens (Timm et al., 2003) suggests that to survive in the host, Mtb must metabolize acetyl CoA in an anaplerotic fashion. It has been assumed that this reflects the derivation of acetyl CoA from β-oxidation of fatty acids (Munoz-Elias and McKinney, 2006).

Anabolism of the branched chain amino acids leucine, isoleucine and valine has been well studied in Mtb, and pathways responsible for their synthesis are essential for virulence (Hondalus et al., 2000). However, little attention has been given to their catabolism, which might serve to prevent their over-accumulation, facilitate the generation of branched chain fatty acids, or provide energy.

One aspect of host immune chemistry that contributes substantially to Mtb’s non-replication in vivo is the generation of reactive nitrogen intermediates (RNI) by inducible nitric oxide synthase (iNOS) (MacMicking et al., 1997). RNI can kill bacteria by forming intrabacterial peroxynitrite (St John et al., 2001). Accordingly, we searched for enzymes that permit Mtb to survive exposure to RNI. We discovered a PNR/P consisting of four enzymes: Lpd, encoded by lpdC (Rv0462) (Argyrou and Blanchard, 2001); DlaT; a thioredoxin-like protein AhpD; and the peroxiredoxin AhpC (Bryk et al., 2000; Bryk et al., 2002). Enzymes that function in complexes are often designated “E1”, “E2”, etc. in the order of sequential catalysis. Lpd was originally predicted to serve as the last enzyme (E3) in two complexes, α-ketoglutarate dehydrogenase (αKGDH) and PDH in Mtb (Cole and Barrell, 1998). However, we found that Mtb lacks αKGDH (Tian et al., 2005a) and that the E1 and E2 of PDH are products of aceE and dlaT (Tian et al., 2005b), the latter originally mis-annotated as E2 of αKGDH. This left unexplained the functions of the proteins encoded by the genes pdhA (Rv2497c, annotated as the α subunit of E1 of PDH), pdhB (Rv2496c, annotated as the β subunit of E1 of PDH) and pdhC (Rv2495c, annotated as E2 of PDH). In many organisms Lpd is also the E3 of BCKADH, so we hypothesized that the pdhABC operon may encode the E1 and E2 of BCKADH (Tian et al., 2005b). However, we could detect no such activity in Mtb (Tian et al., 2005b).

Deletion of dlaT had a pronounced effect on Mtb’s growth in standard medium in vitro, sensitized Mtb to RNI, and attenuated Mtb in the mouse (Shi and Ehrt, 2006). Additionally, species-selective DlaT inhibitors selectively killed non-replicating Mtb (Bryk et al., 2008). We then focused on Lpd as a potential target, because it participates in the same two enzyme complexes as DlaT, namely, PDH and PNR/P (Bryk et al., 2000; Bryk et al., 2002). Structural differences between the active sites of human and mycobacterial Lpd (Rajashankar et al., 2005) were exploited in the development of species-selective inhibitors of mycobacterial Lpd (Bryk et al., 2010). In the course of these studies we were surprised to discover that Lpd-deficient Mtb was far more attenuated in vivo than DlaT-deficient Mtb. This led us to search for yet another function of Lpd. The findings add to our understanding of host-pathogen interactions in experimental tuberculosis.

RESULTS

Mtb Requires Lpd for Virulence

We disrupted lpdC in Mtb H37Rv by homologous recombination and confirmed the deletion by Southern blot (See Figure S1A online), mRNA analysis (See Figure S1B online) and western blot (Figure 1A). We reintroduced lpdC via complementation in an integrative vector driven by a constitutive hsp60 promoter and achieved wild type (WT) expression levels of Lpd mRNA (See Figure S1B online) and protein (Figure 1A). As will be shown subsequently, the Lpd-deficient strain grew, albeit poorly, in standard medium with dextrose, glycerol and fatty acids as carbon sources. To confirm Lpd’s role in metabolism of carbohydrates through its participation in PDH, we monitored the growth of WT, mutant and complemented bacteria with glycerol and dextrose as the only carbon sources. As predicted, the Lpd-deficient bacteria failed to grow on carbohydrates, while the WT and complemented strains grew normally (Figure 1B). The LpdC-containing PNR/P detoxifies peroxynitrite but not NO (Bryk et al., 2000; Bryk et al., 2002). Killing of bacteria by mildly acidified nitrite or other NO donors requires oxygen and involves oxidation of intrabacterial peptidyl methionine (St John et al., 2001), strongly suggesting that it proceeds through intrabacterial conversion of NO (a non-oxidizing species) to peroxynitrite (a strong oxidant). Thus, to test Lpd’s role in PNR/P, we exposed the three strains to mildly acidified nitrite. The Lpd-deficient bacteria lost 3 log10 colony forming units (CFU) at concentrations of nitrite that reduced the viability of the WT and complemented strains by only <0.5 log10 (Figure 1C), consistent with Lpd’s role in PNR/P. To determine whether Mtb lacking lpdC was specifically resistant to the stresses imposed, or merely sickly, we exposed WT, mutant and complemented strains to oxidative stress in the form of H2O2. Mtb lacking lpdC was more resistant than WT and complemented strains, a property shared by several other Mtb mutants that are more sensitive to RNI in vitro (Darwin et al., 2003; Gandotra et al., 2007) (Figure 1D).

Figure 1.

Figure 1

Open in a new tab

Characterization of Lpd-deficient Mtb in vitro. A) Western blot analysis of WT (H37Rv), mutant (ΔlpdC) and complemented (H37RvΔlpdC::lpdC) strains with antiserum against Lpd. Anti-DlaT was used to assess equal loading. B) Growth curve in bacteriologic medium containing dextrose and glycerol as the sole carbon sources. C) Survival of strains after 4 days exposure to NaNO2 at pH 5.5. Means ± SD of triplicate experimental samples from 1 experiment, representative of 3. D) Survival of strains upon exposure to increasing concentrations of H2O2 for 4 hours. Means ± SD of triplicate experimental samples from 1 experiment, representative of 2.

Next we assessed the role of lpdC in Mtb’s ability to grow and persist in vivo by infecting C57BL/6 mice with Lpd-deficient Mtb or the WT or complemented strains. The mutant strain was rapidly cleared from the lungs, (Figure 2A) and either never seeded the spleen or failed to survive there (Figure 2B). Because we plated the organs in their entirety, the detection limit in our assays was 1 CFU/organ. Non-detection of CFU did not result from toxicity of lung homogenates to the bacteria, because the expected number of CFU was observed using the same homogenization procedure on the first day after infection. Lungs contained no detectable CFU in most of the mice from day 7 through termination of the experiment at day 164, and spleens contained no CFU at any time tested (days 28, 64 and 164). Moreover, the lungs of mice infected with the lpdC mutant displayed no gross or microscopic pathology (Figure 2C, See Figure S2 online). Complementation of the mutant restored virulence in vivo, confirming that the severe attenuation of the mutant in vivo was due to loss of lpdC.

Figure 2.

Figure 2

Open in a new tab

Mtb requires Lpd for virulence in both WT and IFNγ−/− mice and in macrophages. Fate of Mtb strains in C57Bl/6 mice in A) lungs and B) spleens after aerosol infection. Means ± SD for 4 mice per point in 1 experiment representative of 3. C) Gross pathology of lungs was recorded at day 164 post-infection. D) Survival of Mtb strains in mice lacking IFNγ. Means ± SD for 4 mice per point in 1 experiment. E) Survival in naïve and F) IFNγ-activated macrophages of the above Mtb strains, as well as Mtb lacking dlaT. Means ± SD of triplicate experimental samples in 1 experiment representative of 2.

Infection with Mtb elicits IFNγ production first by NK cells and then by T lymphocytes. IFNγ is critically required for control of Mtb infection in both mice and humans (Flynn et al., 1993; Jouanguy et al., 1997; Zhang et al., 2008). To test whether Mtb lacking lpdC requires IFNγ-dependent immunity for clearance of the bacterium, we infected IFNγ−/− mice. IFNγ−/− mice infected with the lpdC mutant rapidly cleared the infection, while those infected with WT Mtb were highly susceptible (Figure 2D).

Consistent with the rapid clearance of Lpd-deficient Mtb in vivo, bone marrow derived macrophages from C56BL/6 mice killed the lpdC mutant by ~5 fold even without immunologic activation, while the WT and complemented strains grew ~1.5 log10 over the same period (Figure 2E). When macrophages were activated with IFNγ and then infected, the lpdC mutant was killed by ~1 log10 while the WT and complemented strains still grew, albeit more slowly than in naïve macrophages (Figure 2F). The lesser effect of IFNγ in these experiments than earlier (e.g., (Shi and Ehrt, 2006)) was associated with use of different IFNγ preparations and our current practice of omitting all antibiotics during the differentiation of the macrophages.

Metabolic Differences Between DlaT-deficient and Lpd-deficient Mtb

Characterization of DlaT-deficient Mtb in the same mouse model had revealed only a 1.5 log10 reduction in CFU in the lungs at the onset of the chronic stage of the infection compared to WT Mtb (Shi and Ehrt, 2006). The ≥6 log10 attenuation that we observed here with Lpd-deficient Mtb suggested that Lpd serves an important function in Mtb in which DlaT does not participate. Lpd-deficient and DlaT-deficient mutants were almost indistinguishable in their survival during exposure to a cell wall-perturbing agent (SDS) (See Figure S3A online) or lipophilic antibiotics (See Figures S3B - S3D online). Both mutants, however, were more resistant to ethambutol and rifimpicin than WT Mtb, possibly due to the impaired replication of these strains.

However, compared to DlaT-deficient Mtb, Lpd-deficient Mtb had an inferior capacity for extended growth in 7H9 medium (Figure 3A). Along with carbohydrates, 7H9 medium provides fatty acids when it is supplemented in the standard manner with the dispersal agent Tween 80 (an oleic acid polymer) and bovine serum albumin (a fatty acid carrier). The lpdC mutant did not merely cease growth at later time points, but died, as reflected by ~1 log loss of CFU, while WT, DlaT-deficient and complemented Mtb continued to survive in stationary phase (Figure 3B). Removing dextrose and glycerol impaired the growth of WT Mtb as well as the dlaT mutant so that their growth rates became almost indistinguishable from that of the lpdC mutant (Figure 3C). This suggested that when carbohydrates were present, DlaT-deficient Mtb may have metabolized carbohydrates along with the fatty acids, while Lpd-deficient Mtb may have subsisted only on the fatty acids.

Figure 3.

Figure 3

Open in a new tab

Mtb lacking dlaT grows and survives better than Mtb lacking lpdC. A) Growth of WT, mutants (ΔlpdC, ΔdlaT) and complemented strains (ΔlpdC::lpdC, ΔdlaT::dlaT) in complete 7H9 medium. B) Survival of the same strains as measured by change in CFU. Means ± SD of triplicate cultures. C) Growth in 7H9 containing no carbohydrate carbon sources. All results are representative of at least 2 independent experiments.

In order to elucidate a potential mechanism by which the dlaT mutant, despite lacking PDH, was able to use carbohydrates, we quantified several intracellular metabolites in WT and mutant bacteria. Inhibition of PDH by genetic deletion of either lpdC or dlaT is expected to lead to a build up of glycolytic intermediates and pyruvate (Bryk et al., 2008). We grew the WT, mutant and complemented strains on filter discs on 7H9 agar plates to facilitate rapid harvest and to avoid losing the metabolites that leak during handling (de Carvalho, 2010) and analyzed metabolites in their lysates by liquid chromatography-mass spectrometry. At week 2 of in vitro culture, by which time the lpdC and dlaT mutants had grown to a similar extent, pyruvate was comparably elevated in both mutant strains compared to WT and complemented strains (Figure 4A). Alanine, a metabolite of pyruvate, was also elevated in both mutant strains (See Figure S4A online). The branched chain amino acids valine, leucine, and isoleucine can also arise from pyruvate, and their levels were also elevated at day 14 in both mutant strains compared to WT and complemented strains (See Figures S4B - S4D online). Metabolism of branched chain amino acids results in generation of branched chain keto acids, which are direct substrates for the BCKADH complex. The intracellular levels of ketovaline, ketoleucine, and ketoisoleucine were similarly elevated in both mutant strains compared to WT and complemented strains (Figures 4B, C). In contrast, histidine, whose metabolism is independent of PDH, was equally abundant in all five strains (Figure 4D). At day 28, when the dlaT mutant had continued to grow but the lpdC mutant had stopped growing, intracellular pyruvate, leucine, isoleucine, valine and their keto acid metabolites returned to near normal levels in the dlaT mutant, suggesting that they had been metabolized. However, these metabolites all remained elevated in Mtb lacking lpdC (Figures 4A - 4C, See Figures S4A - S4D online), while histidine remained constant (Figure 4D). The reduction in levels of ketoleucine/ketoisoleucine in the dlaT mutant from day 14 to day 28 (p = 0.0026 for the difference between the two time points) brought the level nearly to that of WT. In contrast, there was a much smaller reduction in levels in the lpdC mutant (p = 0.01 for the difference between the two time points). As noted, complementation of the mutant strains partially or completely restored levels of metabolites to WT levels (Figure 4A - 4D, see Figures S4A - S4D online). Thus, early in culture, both DlaT-deficient Mtb and Lpd-deficient Mtb gave metabolic evidence of their PDH deficiency. However, between the 2nd and 4th week of culture, the metabolic profile of DlaT-deficient Mtb and Lpd-deficient Mtb diverged. DlaT-deficient Mtb appeared to acquire the ability to overcome the block in metabolite flux, while Lpd-deficient Mtb did not.

Figure 4.

Figure 4

Open in a new tab

Dysregulation of intracellular metabolite pools in WT, mutant, and complemented strains. A) Levels of intracellular pyruvate, B) ketovaline, C) ketoleucine + ketoisoleucine, and D) histidine in WT, mutant, and complemented strains. Metabolite data are means ± SD from triplicate experimental samples in 1 experiment representative of 2. *, p < 0.02. E) Growth of strains in 7H9 medium containing 0.2% (w/v) leucine or F) 0.2% (w/v) isoleucine as the sole carbon sources. Growth curves are representative of at least 3 experiments.

Consistent with this, Mtb lacking DlaT grew on leucine (Figure 4E) and isoleucine (Figure 4F) even better than WT and complemented strains. In contrast, Mtb lacking Lpd was unable to grow on leucine or isoleucine. In this respect, Lpd-deficient Mtb phenocopied Mtb lacking both Icl1 and Icl2 (See Figures 4E, 4F online). Thus, both Lpd and the methylcitrate cycle and/or glyoxylate shunt are required for growth on branched chain amino acids. In contrast, absence of DlaT augmented the ability of Mtb to grow on these substrates, perhaps by forcing Mtb to express an Lpd-dependent enzyme that metabolizes them. Neither the mutants nor WT Mtb was able to grow on valine.

Appearance of BCKADH in PDH-deficient Mtb

To identify proteins that complex with Lpd in different circumstances, we generated FLAG-HA-tagged Lpd and introduced it into both the lpdC and dlaT mutants (ΔlpdC::FLAGlpdC and ΔdlaT::FLAGlpdC). FLAG-HA-tagged Lpd restored growth of the lpdC mutant in 7H9 medium to WT levels, suggesting that the tag did not interfere with Lpd’s functional associations. Introduction of FLAG-HA-tagged Lpd into the dlaT mutant had no discernible effect on its growth (See Figure S5A online). After 28 days of growth in 7H9 medium, we lysed the bacteria and immunoprecipitated Lpd using α-FLAG antibody. E2’s of ketoacid dehydrogenase complexes require lipoic acid as a prosthetic group (Perham, 2000). In earlier work, DlaT was the only lipoylated protein we could detect in WT Mtb (Bryk et al., 2002). In ΔlpdC::FLAGlpdC, DlaT was the major interacting partner of Lpd and was again the only lipoylated protein observed in the immunoprecipitate (Figure 5A). In contrast, in the ΔdlaT::FLAGlpdC strain, Lpd co-immunoprecipitated with PdhA, PdhB and PdhC, as identified by peptide mass fingerprinting (See supplementary Table 1 online). An antibody we raised against pure, recombinant PdhC immunoblotted a protein in the latter immunoprecipitate that migrated at 42 kDa, as expected for PdhC. The anti-PdhC antibody reacted with nothing in a control immunoprecipitate prepared from lysates of ΔdlaT Mtb in which native LpdC did not carry the FLAG epitope (Figure 5A). The results with anti-PdhC antibody strongly reinforced the mass spectroscopic evidence that PdhC associates specifically with LpdC. Finally, in the anti-LpdC-FLAG immunoprecipiates from lysates of ΔdlaT Mtb, a unique protein became lipoylated instead of DlaT, and it co-migrated with PdhC (Figure 5A).

Figure 5.

Figure 5

Open in a new tab

Characterization of BCKADH complex in Mtb. A) Immunoblot (IB) of the IP from the ΔlpdC::FLAGlpdC and the ΔdlaT::FLAGlpdC strains using antisera against DlaT, lipoic acid (LA), and PdhC. B) Western blot of lysates from WT and mutant strains collected at day 28 post inoculation at OD580 0.02 using antisera against Lpd, DlaT, LA, PdhC and PrcB (proteasome component B, used as a loading control). Immunoprecipitations and western blots were performed at least twice. C) Levels of pdhA and pdhC transcripts assessed by qRT-PCR at indicated time points in WT and mutant strains. Results show means ± SD of one sample from 3 individual experiments, normalized to fold above sigA. D) In vitro BCKADH activity in lysates prepared from DlaT-deficient Mtb, E) WT Mtb, and F) Lpd-deficient Mtb. Results are representative of 2 experiments.

Next, we monitored the status of protein lipoylation in the total lysates of Mtb strains rather than only in Lpd-interacting proteins, again focusing on 28-day cultures. In DlaT-deficient Mtb, we detected only one lipolylated protein. It migrated at ~42 KDa and co-migrated with the upper band in the doublet immunoblotted by anti-PdhC (Figure 5B). The lower band in the doublet thus corresponds to PdhC that was not lipoylated. Although there was a non-lipoylated fraction of PdhC in Mtb lysates, the PdhC recovered from anti-Lpd immunoprecipitates migrated as a singlet (Figure 5A). These observations suggest that only lipoylated PdhC binds LpdC. In contrast, only DlaT was lipoylated in WT and Lpd-deficient Mtb. Moreover, the level of PdhC in DlaT-deficient Mtb appeared to be higher than in WT Mtb or Lpd-deficient Mtb. Consistent with this, transcripts for pdhA and pdhC were 8-fold higher in DlaT-deficient Mtb than in WT and Lpd-deficient strains (Figure 5C). Deep sequencing of the transcriptomes of WT and DlaT-deficient Mtb revealed that pdhA, pdhB and pdhC were the most highly upregulated genes in the dlaT mutant, being expressed at levels 14.3-, 12.5- and 12.5-fold, respectively, over the levels in WT Mtb (R.B., unpublished observations). We amplified cDNA across the junctions of pdhA-B and pdhB-C and verified that pdhABC are transcribed as an operon (See Figure S5B online).

Finally, BCKADH activity was evident in lysates from 28-day cultures of DlaT-deficient Mtb with 3-methyl 2-oxobutyrate (3M2OB) as a substrate (Figure 5D). Lysates from WT, Lpd-deficient, and DlaT-complemented strains showed no BCKADH activity (Figures 5E, 5F; see Figure 5C online). Lysates from 14-day WT and ΔlpdC cultures lacked BCKADH activity while lysates from 14-day ΔdlaT cultures showed very low levels of BCKADH activity (See Figure S5D online). As a control, we also measured PDH activity in WT and mutant lysates. As expected, only the WT strain carried out this reaction (See Figure 5E online).

Thus, in the absence of DlaT, Mtb upregulated pdhABC, expressed readily detectable quantities of lipoylated PdhC in complex with Lpd and acquired the ability to metabolize branched chain keto acids.

Recapitulation of Lpd Deficiency by Double Deficiency of DlaT and PdhC

To help confirm the foregoing interpretations, we deleted pdhC from H37Rv, generating ΔpdhC, as well as from H37RvΔdlaT, generating ΔdlaTΔpdhC. We confirmed the disruption of pdhC by western (Fig. 6A) and Southern blotting and qRT-PCR (See Figures S6A, S6B online). The strain ΔdlaTΔpdhC was severely attenuated in its ability to grow in 7H9 containing glycerol and dextrose, as well as on leucine or isoleucine, thus phenocopying ΔlpdC. The strain lacking only PdhC did not appear to have a growth defect in any of these conditions, suggesting that when PdhC is missing, these substrates may be metabolized via PDH. Only when the E2’s of PDH and BCKADH were both missing did it appear that glycolytic and branched chain carbon sources could no longer be metabolized. Introduction of a single copy of the dlaT gene into ΔdlaTΔpdhC, generating ΔdlaTΔpdhC::dlaT, complemented all growth phenotypes to WT levels (Figures 6B – 6D).

Figure 6.

Figure 6

Open in a new tab

Characterization of an Mtb mutant lacking both dlaT and pdhC. A) Western blot of WT, ΔdlaT, ΔpdhC, ΔdlaTΔpdhC, and the complemented strain (ΔdlaTΔpdhC::dlaT) using antisera against DlaT, PdhC, Lpd and PrcB. B) Growth curve of the above strains, as well as Mtb lacking lpdC, in complete 7H9 medium or C) medium containing leucine or D) isoleucine as the sole carbon source. E) Survival of WT, ΔdlaTΔpdhC, ΔlpdC, ΔdlaT, and ΔdlaTΔpdhC::dlaT strains in lungs of C57Bl/6 mice infected via aerosol. Means ± SD 4 mice per Mtb strain from 1 experiment representative of at least 2.

Finally, we infected C57Bl/6 mice with both the ΔpdhC and the ΔdlaTΔpdhC mutants. As controls, we also infected mice with ΔdlaTΔpdhC::dlaT, ΔlpdC or ΔdlaT. The ΔpdhC mutant resembled WT Mtb (See Figures S6C online) in its virulence. Reproducibly, the ΔdlaTΔpdhC strain could not be recovered at day 1, despite equal input into the aerosolizer. However, by day 7, the ΔdlaTΔpdhC and ΔlpdC mutants were both recovered at ~100 CFU’s. This suggests that the ΔdlaTΔpdhC mutant was viable but non-culturable at day 1, for unknown reasons. At day 35, the ΔlpdC strain could no longer be recovered. At that point, and at day 190, the ΔdlaTΔpdhC mutant was 2-3 log more attenuated than the ΔdlaT mutant, but more virulent than the ΔlpdC mutant. Complementation of the double knockout with dlaT restored WT virulence levels (Figure 6E). Thus, both PDH and BCKADH complexes are jointly required for virulence of Mtb. However, the finding that the ΔdlaTΔpdhC mutant was not as attenuated as the ΔlpdC mutant suggests Lpd may play a fourth important role over and above its contributions to PDH, BCKADH and PNR/P.

DISCUSSION

This work sheds light on three facets of our understanding of host-pathogen interactions in experimental tuberculosis: how we assess the nutritional composition of the host niche; how a pathogen adapts to nutritional challenges; and how targets are validated in the search for anti-infectives that hold promise to be effective against Mtb in the host.

One major avenue in the study of host-pathogen relationships is to infer the properties of the pathogen’s survival and growth niches in the host by assessing the phenotype of the pathogen when specific enzymes have been knocked out. Attenuation by knockout of a metabolic enzyme is often interpreted to mean that the pathogen must subsist in the host on that enzyme’s substrate(s). This line of reasoning requires two key inputs: an accurate understanding of what reactions the enzyme catalyzes, which may be different from what was expected; and an assessment of whether the phenotype may reflect not lack of the enzyme’s products, but rather, intoxication by substrates that accumulate in its absence. The extent to which metabolic substrates will actually accumulate in the absence of a given enzyme is difficult to predict. Most metabolites lie on multiple enzymatic pathways. We have insufficient knowledge of alternative pathways and their rate constants, particularly in the setting where one major route is inactivated and others may then be adaptively modulated, not only by transcription but also by changes in negative and positive feedback by substrates and products.

For example, Mtb is severely attenuated in the mouse by combined inactivation of icl1 and icl2 (Munoz-Elias and McKinney, 2005). While this was initially interpreted as reporting that Mtb subsists in vivo on fatty acids, subsequent in vitro studies with mutants of M. smegmatis cultured on various carbon sources led to an alternative interpretation, that poor survival of Mtb lacking Icl1 and Icl2 may reflect intoxication by accumulation of propionate and its metabolites (Upton and McKinney, 2007). That disruption of a gene encoding an enzyme can prove lethal from accumulation of toxic precursors was recently demonstrated for the Mtb maltosyltransferase encoded by glgE (Kalscheuer et al., 2010). Similarly, Pethe et al. identified pyrimidine-imidazoles that killed Mtb in vitro by promoting accumulation of glycerol-phosphate in a standard growth medium containing glycerol (Kevin Pethe, 2010).

Although Mtb requires leucine to survive, excess leucine in the growth medium proved toxic (data not shown). This is reminiscent of maple syrup urine disease, an autosomal recessive disorder of humans in which insufficiency of BCKADH leads to accumulation of toxic levels of branched chain amino keto acids (Mackenzie and Woolf, 1959). In the present work, disruption of lpdC led to extraordinary accumulations of pyruvate and branched chain amino and keto acids. These accumulations may cause this strain’s marked attenuation. Metabolomic analysis was invaluable for allowing us to recognize this possibility. Thus, the first implication of our findings is that metabolomic analysis provides an important complement to gene knockout studies in studies aimed at inferring the nutritional composition of the pathogen’s niche in the host.

Another fundamental goal of studying host-pathogen interactions is to understand the molecular basis of the pathogen’s adaptability to changing conditions. Adaptability is critical to pathogenesis, both when the pathogen changes niches upon infecting a new host and when the host responds to infection in ways that alter the niche. The present work has demonstrated a remarkably slow, extremely specific and functionally important adaptation by Mtb that consists in switching lipoylation from one protein substrate to an alternative protein substrate. Thus, when the E2 of PDH was genetically deleted, Mtb upregulated pdhABC, accumulated lipolyated PdhC and expressed a BCKADH whose function was evident in the distribution of metabolites and the activity of lysates. In light of these findings, Rv2497c, Rv2496c and Rv2495c can be re-annotated as bkdA, bkdB and bkdC, respectively. In contrast, WT and Lpd-deficient Mtb grown on a standard rich medium in vitro either did not express a functional BCKADH complex or expressed it at levels below our ability to detect it.

Yet another important goal of studying host-pathogen interactions is to guide the development of anti-infectives that are effective in vivo because they cripple pathways on which the pathogen actually depends in vivo. This has been difficult to predict from studies in vitro. A pathway can be essential in vitro but not in vivo (Kevin Pethe, 2010), or essential in vivo but not in vitro (Munoz-Elias and McKinney, 2005). Thus, it is important to learn that ΔlpdC Mtb was markedly attenuated in vivo. This could not have been predicted from our limited understanding of central carbon metabolism in Mtb.

It has commonly been thought that intermediary metabolism in Mtb mirrors that in other organisms and is therefore well understood. In fact, Mtb’s core intermediary metabolism diverges markedly from the standard model. For example, the tricarboxylic acid cycle is interrupted by the lack of αKGDH (Tian et al., 2005b). A unique anaerobic-type but aerobically-active ferredoxin oxidoreductase acts on α-ketoglutarate (Baughn et al., 2009), but metabolomic analysis does not support the interpretation that the oxidative and reductive arms of the tricarboxylic acid cycle are fully joined (de Carvalho et al., 2010). Additionally, the pentose phosphate pathway and tricarboxylic acid cycle in Mtb are extensively compartmentalized from each other (de Carvalho et al., 2010). In short, core intermediary metabolism in Mtb is vitally important but incompletely understood. Combining genetics, biochemistry, metabolomics and infections can help fill this gap.

We were surprised to discover that Mtb was more attenuated by lack of Lpd than by lack of DlaT or even by lack of both DlaT and PdhC. Lpd clearly has important functions that it does not share with DlaT. At present, participation in BCKADH is the only such function identified. Transcription of bkdABC was detected in Mtb recovered from lungs of mice, but only at the same low level as in Mtb grown in rich medium in vitro (Talaat et al., 2004). This leaves open the question whether BCKADH is expressed at the protein level to an extent that can contribute to the metabolism of PDH-replete Mtb during infection of the mouse, or only plays a role when it is upregulated, for example, in the context of insufficient PDH activity. In either setting, BCKADH might be critical to Mtb’s survival for several reasons. Above, we stressed the role in avoiding accumulation of pyruvate, branched chain amino acids and branched chain ketoacids. Additionally or alternatively, in glucose-limiting conditions in vivo, Mtb may operate a phosphoenol pyruvate (PEP)-glyoxylate pathway, whose activity is dependent on PDH (Fischer and Sauer, 2003; Munoz-Elias and McKinney, 2006). In the absence of PDH, BCKADH may be able to shunt metabolites from pyruvate to acetyl coA, as is required for the functioning of the PEP-glyoxylate pathway. In vitro, carbon-starved Mtb upregulated transcripts for bkdA, bkdB, bkdC and icl1 (Betts et al., 2002). Thus, when otherwise facing starvation, Mtb may metabolize branched chain keto and amino acids via BCKADH as a source of energy. Finally, BCKADH may help regulate the levels of branched-chain lipids. Inactivation of the BCKADH complex in S. aureus and M. xanthus led to a decrease in quantity of branched chain fatty acids in the cell membrane and rendered the mutants more susceptible to a variety of stresses (Singh et al., 2008; Toal et al., 1995). Changes in Mtb’s branched chain fatty acids likewise affect Mtb’s virulence (Glickman and Jacobs, 2001). Thus, the present work establishes the contribution of LpdC to BCKADH in Mtb and the essentiality of LpdC for Mtb’s virulence in the mouse. However, it is not yet clear how BCKADH contributes to Mtb’s virulence.

The essentiality of mycobacterial Lpd in the mouse and the participation of this one protein in at least three different enzyme complexes (Figure 7) suggest that Lpd might be a promising target for chemotherapy of tuberculosis. The triazospirodimethoxybenzoyl inhibitor that inhibits Mtb’s Lpd while sparing human Lpd was not able to enter intact Mtb but its identification demonstrates that species-selective Lpd inhibition is attainable (Bryk et al., 2010).

Figure 7.

Figure 7

Open in a new tab

Schematic showing roles of Lpd in PDH, BCKADH, and PNR/P complexes in Mtb.

Experimental Procedures

Strains and Culture Conditions

WT Mtb H37Rv, mutants and complemented strains were cultured at 37 °C in Middlebrook 7H9 with 0.2% glycerol, 0.5% bovine serum albumin fraction V (BSA), 0.05% Tween 80, 0.2% dextrose and 0.085% sodium chloride (standard medium). Studies on defined carbon sources (0.2%) used fatty acid-stripped BSA (Roche) without glycerol and dextrose with tyloxapol (0.02%) and vitamin B12 (10 μg/ml) (Savvi et al., 2008). CFU were determined on 7H11 agar plates after 4 weeks at 37 °C. Strains bearing antibiotic resistance cassettes were cultured in the presence of hygromycin (50 mg/ml), kanamycin (30 μg/ml), streptomycin (30 μg/ml) or zeocin (25 μg/ml). For growth on solid medium, Middlebrook 7H11 was supplemented with 0.5% glycerol and 10% oleic acid-dextrose-catalase (7H11-OADC).

Mutant Strains

Efforts to delete lpdC (Rv0462, lpdC) from the genome of Mtb H37Rv were initially unsuccessful. We cloned lpdC into an integrative complementation vector (pMV306k) downstream of a constitutively active hsp60 promoter. pMV306k-hsp60lpdC was transformed into WT H37Rv to create a merodiploid strain carrying lpdC at its native site as well as at the attB site (H37Rv::lpdC). LpdC was deleted from the native site of the H37Rv merodiploid strain via a single step recombination using the mycobacteriophage pHAE87 (Bardarov et al., 1997). 550 base pairs upstream and 500 base pairs downstream of lpdC were cloned into the plasmid pJSC284 flanking a hygromycin resistance cassette. The recombinant mycobacteriophage was used to infect H37Rv::lpdC. Knockout colonies were screened by PCR and confirmed by Southern blotting. The complementing plasmid was switched out of the attB site with an empty plasmid bearing a streptomycin resistance cassette, pTCS-mcs1, creating H37RvΔlpdC pTCS-mcs1 (H37RvΔlpdC).

To construct FLAG-HA tagged lpdC, an N-terminal FLAG-HA tag (codon optimized for use in Mtb) was fused to lpdC by PCR, deleting the N terminal valine of Lpd (FLAG-Lpd). This was cloned into an episomal plasmid, pTEK-hsp60, downstream of the hsp60 promoter. pTEK hsp60 nFLAG-HAlpdC was transformed into H37RvΔlpdC (ΔlpdC::FLAGlpdC) and H37RvΔdlaT (ΔdlaT::FLAGlpdC).

PdhC (Rv2495c) was deleted via single step recombination using the mycobacteriophage pHAE87 with a streptomycin resistance cassette. Regions flanking pdhC (727 bp upstream, 717 bp downstream) were cloned on either side of the streptomycin cassette. The recombinant mycobacteriophage infected H37Rv and H37RvΔdlaT::dlaT, creating H37RvΔpdhC and H37RvΔpdhCΔdlaT::dlaT. Colonies were screened by PCR and confirmed by Southern blotting. To create a strain lacking both DlaT and PdhC, the complementing plasmid was switched out of the attB site of H37RvΔpdhCΔdlaT::dlaT with an empty plasmid bearing a zeocin resistance cassette, pTCZ-MCS1, creating H37RvΔpdhCΔdlaT pTCZ-mcs1 (ΔpdhCΔdlaT).

To confirm deletion of lpdC, genomic DNA from WT H37Rv and knockout candidates was digested with XmaI and NcoI, separated by agarose gel electrophoresis and transferred to a nylon membrane that was probed with a 500-bp digoxygenin-labeled fragment containing the upstream lpdC 5′ flanking fragment, revealing a 4.5 Kb band for WT Mtb, and a 2.7 Kb band for lpdC deficient Mtb.

Genomic DNA from H37RvΔpdhC and H37RvΔpdhCΔdlaT::dlaT was digested with EcoRV and XmaI and transferred to a nylon membrane that was probed with a fragment containing 300 bp of pdhB yielding a 1.5-Kb fragment for WT and a 3.5-kb fragment for pdhC deficient Mtb upon Southern blotting with the ECL Direct Nucleic Acid Labeling and Detection System (Amersham).

Antisera to Lpd and PdhC

Recombinant Lpd purified as reported (Argyrou and Blanchard, 2001) was used to immunize chickens. Recombinant PdhC purified as reported (Tian et al., 2005b) was used to immunize rabbits.

Stress Assays

For nitrosative stress, Mtb strains were grown to mid log phase and washed twice in 7H9 acidified to pH 5.5 with HCl. Cultures were centrifuged at 120 g for 10 minutes to remove clumps. Supernatants were adjusted to an OD580 = 0.1, NaNO2 was added and CFU were determined after 4 days. For oxidative stress, Mtb strains prepared as above but in non-acidified medium were adjusted to an OD580 = 0.1, H2O2 was added and CFU were determined after 4 hours. For other stresses, Mtb strains prepared as above were adjusted to OD580 = 0.1 and SDS or antibiotics were added. CFU were determined after 1 hour with to SDS and after 4 hours with lipophilic antibiotics. All experiments were performed in triplicate with serial dilution and plating on 7H11 agar, counting CFU after 4 weeks at 37 °C.

Mouse Infections

C57BL/6 WT and IFNγ−/− mice (Jackson Labs) were infected using an Inhalation Exposure System (Glas-Col). Early-mid log phase Mtb was washed once in PBS containing with 0.05% Tween 80 (PBST) and centrifuged at 120 g to remove clumps. The inoculum (6 ml, OD580 0.002 – 0.005) was nebulized for 40 minutes. Lungs and spleen were homogenized in PBS, serially diluted and plated on 7H11 agar. For mice infected with Lpd-deficient or DlaT/PdhC-deficient Mtb, entire organ homogenates were plated on 7H11-OADC plates, except in mice euthanatized on days 64 and 164, where the upper left lobe was reserved for histology. Plates were incubated at 37 °C for 5 weeks if no CFU were detected earlier. In most experiments, the upper lobe of the left lung was fixed in 4% paraformaldehyde for histopathology.

Macrophage Infections

Marrow cells were flushed from femurs of 8-10 week old C57BL/6 mice and incubated for 6 days at 37 °C, 5% CO2 in DMEM (GIBCO) with 10% heat-inactivated fetal calf serum, 20% L929 fibroblast-conditioned medium (LCM), 1 mM sodium pyruvate (Mediatech Inc.), 10 mM Hepes (GIBCO) and 0.58 g/L L-glutamine (Mediatech Inc.). Cells were fed with the same medium at day 4 and seeded in 48 well plates on day 6 (Ehrt et al., 2001). Where indicated, macrophages were stimulated with IFNγ (10 ng/ml, R&D Systems). Mid-log Mtb cultures were washed twice in PBS + 0.05% Tween 80 (PBST). Clumps were removed by centrifugation at 120 g and the macrophages were infected with 0.1 Mtb per macrophage. CFU were assessed over 6 days (Ehrt et al., 2001).

Quantitative Real Time PCR

We grew Mtb in standard medium starting at an OD580 of 0.05 for 14 and 28 days and added an equal volume of buffer containing guanidium thiocyanate (4 M), sodium lauryl sulfate (0.5%), trisodium citrate (25 mM), and 2-mercaptoethanol (0.1 M). Cultures were pelleted, resuspended in TRIzol and bead-beaten 3 times. RNA was extracted and qRT-PCR performed using gene-specific Taqman probes and primers (Biosearch Technologies).

Analysis of Intracellular Metabolites

Mtb was grown to mid-log phase in standard medium and diluted to an OD580 of 0.2. One ml of each culture was filtered through a nitrocellulose membrane (0.22 μM, Millipore GSWP 02500). Mtb-bearing filters were placed on individual 7H9 agar plates supplemented with 0.5% glycerol, 0.5% BSA, 0.2% dextrose, and 0.085% sodium chloride with the bacterial surface on top. Plates were incubated at 37 °C in 5% CO2 for 14 and 28 days. Filters were flipped onto a 1-ml ice-cold solution of 40% acetonitrile, 40% methanol and 20% water. Bacteria were scraped off the filter and lysed by bead beading 3 times. Intracellular metabolites and those that had leaked into the lysis solution were pooled as described (de Carvalho, 2010). Metabolite abundances were quantified using a calibration curve generated with chemical standards spiked into homologous mycobacterial extracts to correct for matrix-associated ion suppression effects. Protein was measured using the Biorad DC assay (Biorad). Ion counts of specific metabolites were normalized to total protein. Statistical significance was assessed using the Mann-Whitney T test. *, p < 0.02.

Immunoblotting and Immunoprecipitation

Mtb was grown to late stationary phase (28 days in culture) in standard medium, pelleted, washed twice in PBST, and resuspended in potassium phosphate buffer (KPi, 25 mM) containing EDTA (1 mM) and PMSF (1 mM). Lysates were prepared by bead-beating 3 times and pelleting at 20,000 g. Proteins (15-30 μg) were separated by SDS-PAGE and transferred to a nitrocellulose membrane. Immunoblots used: anti-Lpd (1:5000); anti-PdhC (1:5000); anti-DlaT (1:10,000) (Bryk et al., 2002); and anti-PrcB (1:10,000) (Gandotra et al., 2007); or anti-lipoic acid (Calbiochem, 1:2500). Secondary antibodies used were rabbit anti-chicken IgY (Promega, 1:5000) and donkey anti-rabbit IgG (Amersham, 1: 10,000). Lysates of late log phase cultures of ΔlpdC::FLAGlpdC, ΔdlaT::FLAGlpdC, H37Rv and H37RvΔdlaT were adjusted to 5 mg/ml with the lysis buffer, pre-cleared with Sepharose beads for 1 hour and incubated with anti-FLAG M2 beads (SIGMA) at 4 °C overnight. Beads were washed 3 times using wash buffer (WB300) containing Tris pH 7.4 (20 mM), NaCl (300 mM), glycerol (10%), EDTA (0.2 mM), Triton X-100 (0.2%), PMSF (1 mM) and DTT (1 mM). The beads were then washed with WB100 buffer containing only 100 mM NaCl and eluted in WB100 containing 2x FLAG peptide (SIGMA) for SDS-PAGE and western blotting.

Enzyme Assays

Lysates of late log phase cultures (1 mg protein/mL) were incubated in cuvettes at 37 °C in KPi (25 mM) with NAD (1 mM), thiamine pyrophosphate (0.2 mM), coenzyme A (0.17 mM) and MgCl (1mM) and 3-methyl 2-oxobutyrate (2 mM) (Fluka). Production of NADH was monitored at 340 nm in a Uvikon XL spectrophotometer.

HIGHLIGHTS.

  • Mtb lacking Lipoamide dehydrogenase (Lpd) is highly attenuated in mice

  • Lpd-deficient Mtb cannot metabolize branched chain amino acids

  • Lpd functions in Mtb’s branched chain keto acid dehydrogenase complex (BCKADH)

  • Disruption of both pyruvate dehydrogenase and BCKADH phenocopies loss of Lpd

Supplementary Material

01

ACKNOWLEDGEMENTS

We thank Xuiju Jiang, Jean Schneider, Charlie G. Buffie and Sheetal Gandotra for assistance, Luiz Pedro Sorio De Carvalho for anti-PdhC antiserum, John McKinney for the Δicl1/2 strain, and Haiqiang Yu and Haiteng Deng of the Rockefeller University Mass Spectrometry Core facility for mass spectrometric analysis. The authors declare no competing financial interests. Supported by NIH RO1 AI064768. K.R. is supported by Burroughs Wellcome Career Award in the Biomedical Sciences, the William Randolph Hearst Foundation Clinical Scholar Award and NIH R21 AI081094. The Department of Microbiology and Immunology is supported by the William Randolph Hearst Foundation.

A.V. designed and performed experiments, analyzed data and prepared figures. R.B. and P.R. helped with experiments. S.S. and S.E. constructed the lpd mutant bacteria. D.S. provided constructs and advice. K.R. helped with metabolomic experiments. C.N. guided the study. A.V. and C.N. wrote the paper. All authors discussed results and commented on the manuscript.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

REFERENCES

  1. Argyrou A, Blanchard JS. Mycobacterium tuberculosis lipoamide dehydrogenase is encoded by Rv0462 and not by the lpdA or lpdB genes. Biochemistry. 2001;40:11353–11363. doi: 10.1021/bi010575o. [DOI] [PubMed] [Google Scholar]
  2. Bardarov S, Kriakov J, Carriere C, Yu S, Vaamonde C, McAdam RA, Bloom BR, Hatfull GF, Jacobs WR., Jr. Conditionally replicating mycobacteriophages: a system for transposon delivery to Mycobacterium tuberculosis. Proc Natl Acad Sci U S A. 1997;94:10961–10966. doi: 10.1073/pnas.94.20.10961. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barry CE, 3rd, Boshoff HI, Dartois V, Dick T, Ehrt S, Flynn J, Schnappinger D, Wilkinson RJ, Young D. The spectrum of latent tuberculosis: rethinking the biology and intervention strategies. Nat Rev Microbiol. 2009;7:845–855. doi: 10.1038/nrmicro2236. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Baughn AD, Garforth SJ, Vilcheze C, Jacobs WR., Jr. An anaerobic-type alpha-ketoglutarate ferredoxin oxidoreductase completes the oxidative tricarboxylic acid cycle of Mycobacterium tuberculosis. PLoS Pathog. 2009;5:e1000662. doi: 10.1371/journal.ppat.1000662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Betts JC, Lukey PT, Robb LC, McAdam RA, Duncan K. Evaluation of a nutrient starvation model of Mycobacterium tuberculosis persistence by gene and protein expression profiling. Mol Microbiol. 2002;43:717–731. doi: 10.1046/j.1365-2958.2002.02779.x. [DOI] [PubMed] [Google Scholar]
  6. Bloom BR, Murray CJ. Tuberculosis: commentary on a reemergent killer. Science. 1992;257:1055–1064. doi: 10.1126/science.257.5073.1055. [DOI] [PubMed] [Google Scholar]
  7. Bryk R, Arango N, Venugopal A, Warren JD, Park YH, Patel MS, Lima CD, Nathan C. Triazaspirodimethoxybenzoyls as selective inhibitors of mycobacterial lipoamide dehydrogenase. Biochemistry. 2010 doi: 10.1021/bi9016186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bryk R, Gold B, Venugopal A, Singh J, Samy R, Pupek K, Cao H, Popescu C, Gurney M, Hotha S, et al. Selective killing of nonreplicating mycobacteria. Cell Host Microbe. 2008;3:137–145. doi: 10.1016/j.chom.2008.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bryk R, Griffin P, Nathan C. Peroxynitrite reductase activity of bacterial peroxiredoxins. Nature. 2000;407:211–215. doi: 10.1038/35025109. [DOI] [PubMed] [Google Scholar]
  10. Bryk R, Lima CD, Erdjument-Bromage H, Tempst P, Nathan C. Metabolic enzymes of mycobacteria linked to antioxidant defense by a thioredoxin-like protein. Science. 2002;295:1073–1077. doi: 10.1126/science.1067798. [DOI] [PubMed] [Google Scholar]
  11. Chan ED, Strand MJ, Iseman MD. Treatment outcomes in extensively resistant tuberculosis. N Engl J Med. 2008;359:657–659. doi: 10.1056/NEJMc0706556. [DOI] [PubMed] [Google Scholar]
  12. Cole ST, Barrell BG. Analysis of the genome of Mycobacterium tuberculosis H37Rv; Novartis Found Symp; 1998; pp. 160–172. discussion 172-167. [PubMed] [Google Scholar]
  13. Darwin KH, Ehrt S, Gutierrez-Ramos JC, Weich N, Nathan CF. The proteasome of Mycobacterium tuberculosis is required for resistance to nitric oxide. Science. 2003;302:1963–1966. doi: 10.1126/science.1091176. [DOI] [PubMed] [Google Scholar]
  14. de Carvalho LP, Fischer SM, Marrero J, Nathan C, Ehrt S, Rhee KY. Metabolomics of Mycobacterium tuberculosis reveals compartmentalized co-catabolism of carbon substrates. Chem Biol. 2010;17:1122–1131. doi: 10.1016/j.chembiol.2010.08.009. [DOI] [PubMed] [Google Scholar]
  15. de Carvalho LPS, Zhao H, Dickinson CE, Arango NM, Lima CD, Fischer SM, Ouerfelli O, Nathan C, Rhee KY. Activity-based metabolomic profiling of enzymatic function: Identification of Rv1248c as a mycobacterial 2-hydroxy-3-oxoadipate synthase. Chemistry and Biology. 2010 doi: 10.1016/j.chembiol.2010.03.009. In Press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Ehrt S, Schnappinger D, Bekiranov S, Drenkow J, Shi S, Gingeras TR, Gaasterland T, Schoolnik G, Nathan C. Reprogramming of the macrophage transcriptome in response to interferon-gamma and Mycobacterium tuberculosis: signaling roles of nitric oxide synthase-2 and phagocyte oxidase. J Exp Med. 2001;194:1123–1140. doi: 10.1084/jem.194.8.1123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fischbach MA, Walsh CT. Antibiotics for emerging pathogens. Science. 2009;325:1089–1093. doi: 10.1126/science.1176667. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Fischer E, Sauer U. A novel metabolic cycle catalyzes glucose oxidation and anaplerosis in hungry Escherichia coli. J Biol Chem. 2003;278:46446–46451. doi: 10.1074/jbc.M307968200. [DOI] [PubMed] [Google Scholar]
  19. Flynn JL, Chan J, Triebold KJ, Dalton DK, Stewart TA, Bloom BR. An essential role for interferon gamma in resistance to Mycobacterium tuberculosis infection. J Exp Med. 1993;178:2249–2254. doi: 10.1084/jem.178.6.2249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gandotra S, Schnappinger D, Monteleone M, Hillen W, Ehrt S. In vivo gene silencing identifies the Mycobacterium tuberculosis proteasome as essential for the bacteria to persist in mice. Nat Med. 2007;13:1515–1520. doi: 10.1038/nm1683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gill WP, Harik NS, Whiddon MR, Liao RP, Mittler JE, Sherman DR. A replication clock for Mycobacterium tuberculosis. Nat Med. 2009;15:211–214. doi: 10.1038/nm.1915. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Glaziou P, Floyd K, Raviglione M. Global burden and epidemiology of tuberculosis. Clin Chest Med. 2009;30:621–636. vii. doi: 10.1016/j.ccm.2009.08.017. [DOI] [PubMed] [Google Scholar]
  23. Glickman MS, Jacobs WR., Jr. Microbial pathogenesis of Mycobacterium tuberculosis: dawn of a discipline. Cell. 2001;104:477–485. doi: 10.1016/s0092-8674(01)00236-7. [DOI] [PubMed] [Google Scholar]
  24. Hondalus MK, Bardarov S, Russell R, Chan J, Jacobs WR, Jr., Bloom BR. Attenuation of and protection induced by a leucine auxotroph of Mycobacterium tuberculosis. Infect Immun. 2000;68:2888–2898. doi: 10.1128/iai.68.5.2888-2898.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jouanguy E, Altare F, Lamhamedi-Cherradi S, Casanova JL. Infections in IFNGR-1-deficient children. J Interferon Cytokine Res. 1997;17:583–587. doi: 10.1089/jir.1997.17.583. [DOI] [PubMed] [Google Scholar]
  26. Kalscheuer R, Syson K, Veeraraghavan U, Weinrick B, Biermann KE, Liu Z, Sacchettini JC, Besra G, Bornemann S, Jacobs WR., Jr. Self-poisoning of Mycobacterium tuberculosis by targeting GlgE in an alpha-glucan pathway. Nat Chem Biol. 2010;6:376–384. doi: 10.1038/nchembio.340. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Pethe Kevin, S. P, Agarwalla Sanjay, Rhee Kyu, Kuhin Kelli, Phong Wai Yee, Patel Viral, Beer David, Walker John R, Duraiswamy Jeyaraj, Jiricek Jan, Keller Thomas, Chatterjee Arnab, Tan Mai Ping. A Chemical Genetic Screen in Mycobacterium tuberculosis identifies carbon-source-dependent growth inhibitors devoid of in vivo efficacy. Nature Communications. 2010;10:1060. doi: 10.1038/ncomms1060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Mackenzie DY, Woolf LI. Maple syrup urine disease; an inborn error of the metabolism of valine, leucine, and isoleucine associated with gross mental deficiency. Br Med J. 1959;1:90–91. doi: 10.1136/bmj.1.5114.90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. MacMicking J, Xie QW, Nathan C. Nitric oxide and macrophage function. Annu Rev Immunol. 1997;15:323–350. doi: 10.1146/annurev.immunol.15.1.323. [DOI] [PubMed] [Google Scholar]
  30. McCune RM, Feldmann FM, Lambert HP, McDermott W. Microbial persistence. I. The capacity of tubercle bacilli to survive sterilization in mouse tissues. J Exp Med. 1966;123:445–468. doi: 10.1084/jem.123.3.445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Munoz-Elias EJ, McKinney JD. Mycobacterium tuberculosis isocitrate lyases 1 and 2 are jointly required for in vivo growth and virulence. Nat Med. 2005;11:638–644. doi: 10.1038/nm1252. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Munoz-Elias EJ, McKinney JD. Carbon metabolism of intracellular bacteria. Cell Microbiol. 2006;8:10–22. doi: 10.1111/j.1462-5822.2005.00648.x. [DOI] [PubMed] [Google Scholar]
  33. Munoz-Elias EJ, Timm J, Botha T, Chan WT, Gomez JE, McKinney JD. Replication dynamics of Mycobacterium tuberculosis in chronically infected mice. Infect Immun. 2005;73:546–551. doi: 10.1128/IAI.73.1.546-551.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Nathan C. Taming tuberculosis: a challenge for science and society. Cell Host Microbe. 2009;5:220–224. doi: 10.1016/j.chom.2009.02.004. [DOI] [PubMed] [Google Scholar]
  35. Nathan C, Gold B, Lin G, Stegman M, de Carvalho LP, Vandal O, Venugopal A, Bryk R. A philosophy of anti-infectives as a guide in the search for new drugs for tuberculosis. Tuberculosis (Edinb) 2008;88(Suppl 1):S25–33. doi: 10.1016/S1472-9792(08)70034-9. [DOI] [PubMed] [Google Scholar]
  36. Pandey AK, Sassetti CM. Mycobacterial persistence requires the utilization of host cholesterol. Proc Natl Acad Sci U S A. 2008;105:4376–4380. doi: 10.1073/pnas.0711159105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Perham RN. Swinging arms and swinging domains in multifunctional enzymes: catalytic machines for multistep reactions. Annu Rev Biochem. 2000;69:961–1004. doi: 10.1146/annurev.biochem.69.1.961. [DOI] [PubMed] [Google Scholar]
  38. Rajashankar KR, Bryk R, Kniewel R, Buglino JA, Nathan CF, Lima CD. Crystal structure and functional analysis of lipoamide dehydrogenase from Mycobacterium tuberculosis. J Biol Chem. 2005;280:33977–33983. doi: 10.1074/jbc.M507466200. [DOI] [PubMed] [Google Scholar]
  39. Russell DG. Mycobacterium tuberculosis: here today, and here tomorrow. Nat Rev Mol Cell Biol. 2001;2:569–577. doi: 10.1038/35085034. [DOI] [PubMed] [Google Scholar]
  40. Savvi S, Warner DF, Kana BD, McKinney JD, Mizrahi V, Dawes SS. Functional characterization of a vitamin B12-dependent methylmalonyl pathway in Mycobacterium tuberculosis: implications for propionate metabolism during growth on fatty acids. J Bacteriol. 2008;190:3886–3895. doi: 10.1128/JB.01767-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Schnappinger D, Ehrt S, Voskuil MI, Liu Y, Mangan JA, Monahan IM, Dolganov G, Efron B, Butcher PD, Nathan C, et al. Transcriptional Adaptation of Mycobacterium tuberculosis within Macrophages: Insights into the Phagosomal Environment. J Exp Med. 2003;198:693–704. doi: 10.1084/jem.20030846. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Shi S, Ehrt S. Dihydrolipoamide acyltransferase is critical for Mycobacterium tuberculosis pathogenesis. Infect Immun. 2006;74:56–63. doi: 10.1128/IAI.74.1.56-63.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Singh VK, Hattangady DS, Giotis ES, Singh AK, Chamberlain NR, Stuart MK, Wilkinson BJ. Insertional inactivation of branched-chain alpha-keto acid dehydrogenase in Staphylococcus aureus leads to decreased branched-chain membrane fatty acid content and increased susceptibility to certain stresses. Appl Environ Microbiol. 2008;74:5882–5890. doi: 10.1128/AEM.00882-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Smith H. Questions about the behaviour of bacterial pathogens in vivo. Philos Trans R Soc Lond B Biol Sci. 2000;355:551–564. doi: 10.1098/rstb.2000.0597. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. St John G, Brot N, Ruan J, Erdjument-Bromage H, Tempst P, Weissbach H, Nathan C. Peptide methionine sulfoxide reductase from Escherichia coli and Mycobacterium tuberculosis protects bacteria against oxidative damage from reactive nitrogen intermediates. Proc Natl Acad Sci U S A. 2001;98:9901–9906. doi: 10.1073/pnas.161295398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Talaat AM, Lyons R, Howard ST, Johnston SA. The temporal expression profile of Mycobacterium tuberculosis infection in mice. Proc Natl Acad Sci U S A. 2004;101:4602–4607. doi: 10.1073/pnas.0306023101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Tian J, Bryk R, Itoh M, Suematsu M, Nathan C. Variant tricarboxylic acid cycle in Mycobacterium tuberculosis: identification of alpha-ketoglutarate decarboxylase. Proc Natl Acad Sci U S A. 2005a;102:10670–10675. doi: 10.1073/pnas.0501605102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Tian J, Bryk R, Shi S, Erdjument-Bromage H, Tempst P, Nathan C. Mycobacterium tuberculosis appears to lack alpha-ketoglutarate dehydrogenase and encodes pyruvate dehydrogenase in widely separated genes. Mol Microbiol. 2005b;57:859–868. doi: 10.1111/j.1365-2958.2005.04741.x. [DOI] [PubMed] [Google Scholar]
  49. Timm J, Post FA, Bekker LG, Walther GB, Wainwright HC, Manganelli R, Chan WT, Tsenova L, Gold B, Smith I, et al. Differential expression of iron-, carbon-, and oxygen-responsive mycobacterial genes in the lungs of chronically infected mice and tuberculosis patients. Proc Natl Acad Sci U S A. 2003;100:14321–14326. doi: 10.1073/pnas.2436197100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Toal DR, Clifton SW, Roe BA, Downard J. The esg locus of Myxococcus xanthus encodes the E1 alpha and E1 beta subunits of a branched-chain keto acid dehydrogenase. Mol Microbiol. 1995;16:177–189. doi: 10.1111/j.1365-2958.1995.tb02291.x. [DOI] [PubMed] [Google Scholar]
  51. Upton AM, McKinney JD. Role of the methylcitrate cycle in propionate metabolism and detoxification in Mycobacterium smegmatis. Microbiology. 2007;153:3973–3982. doi: 10.1099/mic.0.2007/011726-0. [DOI] [PubMed] [Google Scholar]
  52. Zhang SY, Boisson-Dupuis S, Chapgier A, Yang K, Bustamante J, Puel A, Picard C, Abel L, Jouanguy E, Casanova JL. Inborn errors of interferon (IFN)-mediated immunity in humans: insights into the respective roles of IFN-alpha/beta, IFN-gamma, and IFN-lambda in host defense. Immunol Rev. 2008;226:29–40. doi: 10.1111/j.1600-065X.2008.00698.x. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS261589-supplement-01.pdf (2.6MB, pdf)

RESOURCES