NAD+ auxotrophy is bacteriocidal for the tubercle bacilli

Abstract

The human tubercle bacillus Mycobacterium tuberculosis can synthesize NAD(+) using the de novo biosynthesis pathway or the salvage pathway. The salvage pathway of the bovine tubercle bacillus Mycobacterium bovis was reported defective due to a mutation in the nicotinamidase PncA. This defect prevents nicotinic acid secretion, which is the basis for the niacin test that clinically distinguishes M. bovis from M. tuberculosis. Surprisingly, we found that the NAD(+)de novo biosynthesis pathway (nadABC) can be deleted from M. bovis, demonstrating a functioning salvage pathway. M. bovisDeltanadABC fails to grow in mice, whereas M. tuberculosisDeltanadABC grows normally in mice, suggesting that M. tuberculosis can acquire nicotinamide from its host. The introduction of M. tuberculosis pncA into M. bovisDeltanadABC is sufficient to fully restore growth in a mouse, proving that the functional salvage pathway enables nicotinamide acquisition by the tubercle bacilli. This study demonstrates that NAD(+) starvation is a cidal event in the tubercle bacilli and confirms that enzymes common to the de novo and salvage pathways may be good drug targets.

Fig. 1

Deletion of the NAD⁺ de novo pathway genes, nadABC, in M. bovis and M. tuberculosis. A. NAD⁺ biosynthetic and salvage pathways. NadA, NadB, and NadC enzymes are unique to the NAD⁺ de novo biosynthesis pathway, while the two last enzymes, NadD and NadE, are common to both the de novo and the salvage NAD⁺ pathways. The salvage pathway decomposes NAD⁺ into nicotinamide (NAm) or NMN, or converts NAm from the media to nicotinic acid mononucleotide (NAMN) using 2 enzymes, PncA and PncB. The enzymes and intermediates involved in this study are indicated in bold. B.Southern analysis of M. bovis ΔnadABC and M. tuberculosis ΔnadABC transductants. The diagram shows the replacement of the nadABC operon in M. bovis by a hygromycin resistance gene (hyg) using specialized transduction. Southern analysis of the genomic DNA from wt and ΔnadABC mutants, digested with BamH1, confirmed the replacement of the nadABC genes by hyg in M. bovis and M. tuberculosis.

Fig. 2

Metabolization of NAm by M. bovis. M. bovis strains were labeled with ¹⁴C-NAm for 72 hr. The cell pellets were washed intensively and lyzed. The ¹⁴C-labeled metabolites extracted from the cell pellets were separated by reverse-phase HPLC and detected with a beta-gamma radiation detector. The metabolites were identified by comparison with chromatograms of standards. The position of the signal for NAm is indicated by an arrow. Comp stands for complemented.

Fig. 3

Growth of M. bovis ΔnadABC and M. tuberculosis ΔnadABC in vitro. A. The strains were grown with NAm (20 mg/l) to log phase, spun down, washed 5 X with PBS, resuspended in media without NAm, and inoculated into media containing either NAm at different concentrations or NAc at 20 mg/l. The complemented strains were grown without any supplement. The cultures were incubated with shaking at 37°C for 3 weeks and growth was followed by measuring OD_600nm. B. The strains were grown, washed, diluted, and inoculated in media containing or not containing NAm (20 mg/l). At each time point, samples were taken, diluted, and plated onto Middlebrook 7H10 plates containing OADC, glycerol, and NAm (20 mg/l). The plates were incubated at 37°C for 4-5 weeks and colonies were counted. The concentrations are in mg/l.

Fig. 4

Growth of M. bovis ΔnadABC pMV261::pncA^TB in vitro. M. bovis ΔnadABC was transformed with M. tuberculosis pncA cloned into the replicative plasmid pMV261. A. M. bovis ΔnadABC pMV261::pncA^TB and its parent strain M. bovis ΔnadABC were grown in NAm-containing media, washed 5 X in PBS, diluted, and inoculated in media containing NAm (20 and 0.1 mg/l). Growth was followed by measuring OD_600nm at diverse time points. B. M. bovis ΔnadABC pMV261::pncA^TB and M. bovis ΔnadABC were grown and washed as described above, diluted 1-50, and inoculated in media without NAm (20 mg/l). At each time, samples were taken, diluted, and plated onto Middlebrook 7H10 plates containing NAm (20 mg/l). The plates were incubated at 37°C for 4-5 weeks and colonies were counted.

Fig. 5

NAD⁺ and NADH concentrations in ΔnadABC strains. A. The strains were grown to log-phase (OD_600nm ≈ 1) before NAD⁺ and NADH were extracted, as described in Experimental Procedures. M. bovis ΔnadABC, M. bovis ΔnadABC pMV261::pncA^TB and M. tuberculosis ΔnadABC were grown in media containing NAm (20 mg/l). B. M. tuberculosis ΔnadABC, M. bovis ΔnadABC, and M. bovis ΔnadABC pMV261::pncA^TB were grown to early log phase (OD_600nm ≈ 0.5), spun down, washed 5 X with PBS, and resuspended in media without NAm. The resulting OD_600nm after washes were 0.45 for M. tuberculosis ΔnadABC, 0.3 for M. bovis ΔnadABC, and 0.2 for M. bovis ΔnadABC pMV261::pncA^TB. NAD⁺ and NADH concentrations were measured after 2, 7, and 14 days of NAm starvation. The experiments were done in triplicate and the average is plotted with the standard deviation.

Fig. 6

Growth in the lungs of immunocompetent C57Bl/6 mice, following low-dose aerosol infection, of: A. M. bovis, M. bovis ΔnadABC and M. bovis ΔnadABC complemented; B. M. tuberculosis, M. tuberculosis ΔnadABC and M. tuberculosis ΔnadABC complemented; and C. M. bovis ΔnadABC, M. bovis ΔnadABC pMV261::pncA^TB, and M. bovis ΔnadABC complemented