Carbon flux rerouting during Mycobacterium tuberculosis growth arrest

Abstract

A hallmark of the Mycobacterium tuberculosis life cycle is the pathogen's ability to switch between replicative and non-replicative states in response to host immunity. Transcriptional profiling by qPCR of ∼ 50 M. tuberculosis genes involved in central and lipid metabolism revealed a re-routing of carbon flow associated with bacterial growth arrest during mouse lung infection. Carbon rerouting was marked by a switch from metabolic pathways generating energy and biosynthetic precursors in growing bacilli to pathways for storage compound synthesis during growth arrest. Results of flux balance analysis using an in silico metabolic network were consistent with the transcript abundance data obtained in vivo. Similar transcriptional changes were seen in vitro when M. tuberculosis cultures were treated with bacteriostatic stressors under different nutritional conditions. Thus, altered expression of key metabolic genes reflects growth rate changes rather than changes in substrate availability. A model describing carbon flux rerouting was formulated that (i) provides a coherent interpretation of the adaptation of M. tuberculosis metabolism to immunity-induced stress and (ii) identifies features common to mycobacterial dormancy and stress responses of other organisms.

Fig. 1. Pathways and genes in central metabolism and lipid metabolism in M. tuberculosis analyzed in the present work

A schematic representation of the metabolic pathways of M. tuberculosis interrogated by qPCR is shown. In each pathway, gene names and/or Rv numbering (Cole et al., 1998) are indicated. Key intermediates and storage compounds are shown in bold. Abbreviations: PEP, phosphoenolpyruvate; DHAP, dihydroxyacetone phosphate, TAG, triacylglycerol. Fatty acid β-oxidation: fadE5 (Rv0244c) and fadE23 (Rv3140), acyl-CoA dehydrogenase; fadA4 (Rv1323), acyl-CoA acetyltransferase; fadB (Rv0860), fatty acid oxidation protein. Glycolysis: ppgK (Rv2702), glucokinase; pfkA (Rv3110c) and pfkB (Rv2029c), phosphofructose kinase; gap (Rv1436), glyceraldehyde 3-P dehydrogenase; pykA (Rv1617), pyruvate kinase; dlaT (Rv2215) and lpdC (Rv0462), E2 and E3 components of pyruvate dehydrogenase. PPP: devB (Rv1445), 6-phosphogluconolactonase; zwf1 (Rv1121) and zwf2 (Rv1447c), glucose 6-P 1-dehydrogenase; tkt (Rv1449c), transketolase. TCA cycle: citA (Rv0889c), citrate synthase II; gltA2 (Rv0896), citrate synthase I; acn (Rv1475), aconitase; icd1 (Rv3339c) and icld2 (Rv0066c), isocitrate dehydrogenase; kgd (Rv1248c), α-ketoglutarate decarboxylase; sdhA (Rv3318), flavoprotein of succinate dehydrogenase; sdhC (Rv3316), cytochrome B-556 of succinate dehydrogenase; mdh (Rv1240), malate dehydrogenase. Glyoxylate shunt: icl (Rv0467), isocitrate lyase; glcB (Rv1837c), malate synthase. Methylcitrate cycle and methylmalonyl pathway: prpDC (Rv1130–1131), methylcitrate dehydratase and methylcitrate synthase, respectively; mutAB (Rv1942–1943), methylmalonyl-CoA mutase, small and large subunit. Gluconeogenesis: mez (Rv2332), malic enzyme; pckA (Rv0211), PEP carboxykinase; glpX (Rv1099c), fructose 1,6-bisphosphatase. Glutamate and glutamine synthesis: gltB (Rv3859), glutamate synthase (large subunit); glnA1 (Rv2220), glutamine synthetase; gdh (Rv2476c), glutamate dehydrogenase. TAG synthesis: plsB2 (Rv2482c), glycerol 3-P acyltransferase; Rv2182c, 1-acylglycerol 3-P O-acyltransferase; tgs1 (Rv3130c), TAG synthase; Rv3734c and Rv3371, TAG synthase. Mycolic acid synthesis: fas (Rv2524c), fatty acid synthase; fabH (Rv0533c), β-ketoacyl-AcpM synthase III; kasA (Rv2245), β-ketoacyl-AcpM synthase; fabG1 (Rv1483), β-ketoacyl-AcpM reductase. Multimethyl-branched lipid synthesis: accD5 (Rv 3280), PropCoA carboxylase β chain 5; mas (Rv2940), mycocerosic acid synthase; fadD26 (Rv2930), fatty-acid-CoA ligase; fadD28 (Rv2941), fatty-acid-CoA synthetase; papA5 (Rv2939), conserved polyketide synthase; pks2 (rv3825c), conserved polyketide synthase.

Fig. 2. Changes in mRNA levels of M. tuberculosis genes involved in central metabolism and lipid metabolism during mouse lung infection

Lungs were harvested from 3 to 4 mice at multiple time points (up to day 100) post-infection, as indicated. Total RNA was extracted and qPCR enumeration of bacterial transcripts was performed. Bacterial mRNA copy numbers were normalized to 16S rRNA. Results are shown as ratios of the mean of normalized mRNA copy numbers at each time point relative to the mean of normalized copy numbers at day 12 post-infection. Raw data (means and SD) for each time point are shown in SI Table 2. Each panel represents the indicated metabolic pathways.

Fig. 3. Metabolic model of M. tuberculosis infecting the mouse lung

In growing bacilli, carbon from fatty acid degradation and sugars is routed via the central metabolism toward generation of energy and biosynthetic precursors required for bacterial growth. Assimilation of AcCoA and PropCoA leads to the synthesis of major cell-wall components such as mycolic acids, PDIM, and poly-L-glutamine. The glyoxylate shunt is also active, as shown by a requirement for isocitrate lyase activity throughout mouse lung infection (Munoz-Elias & McKinney, 2005). In non-growing bacilli, glycolysis, PPP and TCA cycle are downregulated. AcCoA is preferentially assimilated through the glyoxylate shunt and gluconeogenetic reactions leading to PEP, while PropCoA is preferentially assimilated via the methylcitrate cycle leading to pyruvate. The resulting formation of PEP and pyruvate, which also results from reduced de novo synthesis of sugars, leads to rerouting of carbon flow toward glyceroneogenesis and TAG synthesis. Some AcCoA is also shunted into glutamate biosynthesis. Moreover, uncoupling of FAS-I and FAS-II activities leads to rerouting of FAS-I product toward TAG synthesis and utilization of AcCoA for meromycolate elongation rather than de novo synthesis. Together, reduced PDIM synthesis, slightly increased sulfolipid-1 (SL-1) synthesis, reduced poly-L-glutamine synthesis, increased poly-L-glutamate synthesis, and meromycolate elongation result in cell-wall remodeling.

Fig. 4. Metabolic flux changes during M. tuberculosis growth arrest predicted by in silico modeling

FBA of a genome-scale metabolic model of M. tuberculosis containing ~900 unique reactions (Beste et al., 2007) was conducted. The indicated substrates were used as input, and biomass of growing and non-growing tubercle bacilli cells was used as output (details in SI Tables 3 and 4A–B). Shown are ratios between the predicted flux values for growing and non-growing cells for 17 central metabolism reactions described in the text. Flux ratio calculations for all central metabolism reactions (~100) are shown in SI Table 4C–H. Positive ratios indicate fluxes in the same direction in growing vs. non-growing bacilli, and negative ratios indicate reactions in the opposite direction. Flux ratios that reach the +4/−4 value in the figure have values outside the range shown (see SI Table 4C–H). Reaction numbers are from the metabolic model (SI Table 4C–H). Abbreviations: Gly, glycolysis; GNG, gluconeogenesis; Pyr, pyruvate metabolism; TCA, tricarboxylic acid cycle; MCC, methylcitrate cycle; PPP, pentose phosphate pathway; PCK, PEP carboxykinase; GS, glyoxylate shunt.

Fig. 5. Expression changes of metabolic genes of M. tuberculosis during hypoxia and after treatment with NO in cultures grown in enriched media

M. tuberculosis cells were cultured in standard Dubos Tween-Albumin medium, and mid-log-phase cultures were subjected to gradual O₂ depletion (Wayne model) or treatment with 100µM DETA/NO. Culture aliquots were harvested at multiple times post-treatment. Transcripts were enumerated by qPCR and normalized to 16S rRNA. Shown are ratios between the means of normalized mRNA copy numbers [at hours 102 and 126 in hypoxia, or after 0.5-h of DETA/NO treatment] and the mean of the mid-log-phase culture data. The hypoxia data are from one of two independent repeats, which gave very similar results (measurements for all genes and time points tested are shown in SI Table 5A). The NO treatment data were obtained from triplicate samples [ratios for additional genes and time points tested are shown SI Table 5B; raw data (means and SD) for all genes and time points tested are shown in SI Table 5C].

Fig. 6. Accumulation of lipid bodies and loss of acid-fastness in M. tuberculosis cells during hypoxia

M. tuberculosis cultures were subjected to gradual O₂ depletion (Wayne model). Aliquots from mid-log cultures (panel A) and cultures at day 30 of the hypoxic time course (anaerobiosis) (panel B) were stained for acid-fastness with Auramine-O (green) and for lipid body accumulation with Nile Red (red). Cells were examined by fluorescence microscopy at the same intensity for all samples with Z stacking to get the depth of the scan field. Overlaid images of Auramine-O- and Nile-Red-stained cells are shown. Bar = 4µM. Panel C: Cells stained with Auramine-O (green line) or Nile-red (red line) were counted from three microscopic fields. Means (and SD) of % stain-positive cells are shown.

Fig. 7. Activity of key metabolic enzymes and corresponding mRNA profiles during hypoxia in vitro

Mid-log cultures of M. tuberculosis were subjected to gradual O₂ depletion (Wayne model). For the enzymatic assays, samples from triplicate cultures were harvested at multiple time points of the hypoxic time course corresponding to expression changes of the corresponding genes. Left panels: activity results for each enzyme, as indicated. Right panels: copy numbers of the corresponding transcripts. Data are shown as ratios between means of measurements at the indicated time point and those at time 0 (mid-log culture). Raw data (means and SD) for each time point are shown in SI Table 7.