Fig 1.
Evidence for an alternative route to M1P in M. smegmatis independent of TreS-Pep2 and trehalose.
(A) Heterologous expression of the M. tuberculosis glgC gene (Rv1213) restores M1P accumulation in the M. smegmatis ΔglgE Δpep2 double mutant harboring a spontaneous IS1096 element insertion in the endogenous glgC locus (i.e. M. smegmatis ΔglgE Δpep2 glgC:IS1096). Equivalent quantities of crude extracts of M. smegmatis strains were analyzed using 1H NMR spectroscopy. The assignment of the peaks was based on our previous studies [16, 18]. (B) Hot water extracts from 1 ml culture aliquots of M. smegmatis strains (normalized to OD600 nm = 0.5) were analyzed by TLC, demonstrating M1P accumulation in the M. smegmatis ΔglgE Δpep2 glgC:IS1096 strain expressing the M. tuberculosis glgC gene (Rv1213). M1P and trehalose (5 μg each) were used as standards. (C) M. smegmatis Δpep2(u) ΔglgE and ΔtreS(u) ΔglgE double mutants accumulate M1P. TLC analysis was performed as described in (B). (D) The M. smegmatis Δpep2(u) ΔglgE mutant is trehalose resistant despite accumulating M1P, indicating trehalose-independent M1P formation. Strains were grown on Middlebrook 7H10 agar plates with or without 1 mM trehalose and incubated at 37°C for 3 days. (E) Conditional silencing of the glgE gene in M. smegmatis mutant strains reveals the requirement of GlgC and GlgA for the alternative route to M1P synthesis. Cells of the indicated conditional c-glgE-tet-off mutant strains were cultivated for 24 h with or without 1 μg ml-1 ATc as indicated, and hot water extracts from 1 ml culture aliquots (normalized to OD600 nm = 0.5) were analyzed by TLC.
Table 1.
GlgA activity with glycogen and α-glucan.
Michaelis-Menten constants ± SE for recombinant M. tuberculosis GlgA with rabbit liver glycogen and S. venezuelae α-glucan were determined in triplicate in the presence of 1 mM ADP-glucose by monitoring the production of ADP with a spectrophotometric assay.
Fig 2.
Mycobacterial GlgA is an M1P synthesizing glucosyltransferase.
(A) The preferred reaction catalyzed by GlgA from M. tuberculosis. (B) Recombinant M. tuberculosis GlgA converts ADP-glucose and G1P to M1P. 1H NMR spectroscopy was used to monitor the anomeric protons of ADP-glucose and G1P (1 mM each) together with signals associated with the product in the presence of GlgA. The spectra show the concomitant consumption of ADP-glucose (~5.5 ppm) and appearance of resonances consistent with the formation of α-1,4 glycosidic linkages (~5.32 ppm). Given the lack of any resonances associated with free reducing ends (~5.1 ppm) and the retention of those associated with an α-glucosyl phosphate residue (~5.36 ppm), these observations are consistent with the formation of M1P. (C) Mass spectrometry of the product of the M. tuberculosis GlgA reaction with ADP-glucose and G1P. The dominant ion in the spectrum is consistent with the presence of M1P (m/z 421.0749 observed with 421.0752 expected for [M-H]-). The next most abundant ion is consistent with the presence of the co-product, ADP (m/z 426.0218 observed with 426.0221 expected for [M-H]-). (D) The dependence of M. tuberculosis GlgA activity on ADP-glucose. GlgA activity was determined spectrophotometrically by monitoring the production of ADP. A representative triplicate dataset with 1 mM G1P is shown as means and SEM. The data conform to the Michaelis-Menten eq 1 (fit shown as the solid line giving r2 = 0.82). (E) The dependence of M. tuberculosis GlgA activity on G1P. GlgA activity was determined spectrophotometrically by monitoring the production of ADP. A representative triplicate dataset with 0.1 mM ADP-glucose is shown as means and SEM. Significant inhibition by G1P at concentrations >1 mM is apparent and the dataset conforms to a simple substrate inhibition eq 2 (fit shown as the solid line giving r2 = 0.99). (F) M. tuberculosis GlgA (Rv1212c) was heterologously expressed in the M. smegmatis ΔtreS(u) ΔglgA(u) c-glgE-tet-off mutant. Cells were cultivated for 24 h with or without 1 μg ml-1 ATc as indicated, and hot water extracts from 1 ml culture aliquots (normalized to OD600 nm = 0.5) were analyzed by TLC. Conditional silencing of the glgE gene results in M1P accumulation, demonstrating that M. tuberculosis GlgA synthesizes M1P in vivo.
Table 2.
Dependence of GlgA activity on ADP-glucose and G1P.
Kinetic constants ± SE for GlgA with ADP-glucose and G1P were determined in triplicate using the spectrophotometric assay monitoring the production of ADP. Data where ADP-glucose and G1P concentrations were varied were fitted to the Michaelis-Menten equation either without or with substrate inhibition, respectively.
Fig 3.
Compensatory flux of ADP-glucose through GlgA and OtsA links the GlgC-GlgA and TreS-Pep2 routes for α-glucan production.
(A) α-Glucan visualization in M. smegmatis mutant strains. Cells were cultivated on Middlebrook 7H10 agar plates for 3 days and exposed to iodine vapor for staining of α-glucans. Branched α-glucans give a pale red-brown color. In the absence of branching enzyme GlgB, long linear glucans are produced resulting in a dark blue color of cells with the intensity of staining correlating with the amount of total cellular α-glucans. (B) Analysis of nucleotide sugar diphosphates in cell extracts of M. smegmatis mutant strains. Cells were cultivated for 24 h in Middlebrook 7H9 liquid medium. Due to trehalose auxotrophy of the M. smegmatis ΔglgA(u) ΔotsA mutant, 50 μM trehalose was added to all cultures. Cell suspensions were normalized to OD600 nm, washed with PBS, concentrated 50-fold and disrupted by bead beating. Cell-free extracts were heat inactivated for 15 min at 100°C, and 1H NMR spectroscopy was used to detect the anomeric protons of ADP-glucose.
Fig 4.
The central importance of the GlgE pathway in intracellular and capsular α-glucan synthesis in M. tuberculosis.
Quantification of intracellular (A) or extracellular (i.e. capsular) (B) α-glucan in M. tuberculosis H37Rv mutant strains. Cells were grown in Middlebrook 7H9 liquid medium for 7 days with shaking. Intracellular glucans were measured in hot water extracts of washed cells. Capsular glucans were measured from cell-free culture supernatants. Intracellular and capsular glucans were assayed by sandwich ELISA employing an α-glucan specific monoclonal antibody. Similar results for intracellular glucan content were also obtained using an enzymatic assay with cells from independent biological replicates (S3 Fig). Values were normalized based on OD 600 nm of cultures. Values in (A) and (B) represent means of triplicates ± SEM. (C) Visualization of the α-glucan capsule in the M. tuberculosis WT and the ΔglgC(u) ΔtreS mutant by immunogold labelling. Cells were grown in liquid medium without shaking, fixed, labelled with an α-glucan specific monoclonal antibody, and analyzed by electron microscopy (scale bar 0.5 μm). (D) Quantitative evaluation of α-glucan capsule visualization as shown in (C), plotted as anti-α-glucan specific gold particles per cell. Values represent means ± SEM (WT n = 27, ΔglgC(u) ΔtreS n = 28). Negative controls were not treated with the primary anti-α-glucan antibody (n = 32).
Fig 5.
Importance of GlgE-mediated α-glucan synthesis for virulence of M. tuberculosis in mice.
BALB/c mice were challenged by intravenous infections with 5 × 104 colony forming units of M. tuberculosis α-glucan mutant strains (mean ± SD, n = 3 per time point). The growth of the double mutant was significantly attenuated in lung and spleen as compared to wild-type at all time-points (p < 0.05), except for the earliest 21 days post-infection time point in the lung.
Fig 6.
Revised model of GlgE-mediated intracellular and capsular α-glucan synthesis in mycobacteria.
The M1P building block for the maltosyltransferase GlgE is formed via two alternative routes, TreS-Pep2 and GlgC-GlgA. Both routes are connected via the shared use of ADP-glucose by GlgA and OtsA, the latter providing the trehalose substrate for the TreS-Pep2 pathway. GlgA, like OtsA, is also capable of using UDP-glucose as a donor, which in turn is produced from G1P by GalU, but this appears to be less significant in vivo. GlgE is essential for intracellular and capsular α-glucan synthesis and generates linear maltooligosaccharides as the substrates for the branching enzyme GlgB. α-Glucans are produced intracellularly with partial coupling to secretion by unknown transport mechanisms. Steps highlighted in red are new findings as described in this report. G6P, glucose 6-phosphate; G1P, glucose 1-phosphate; M1P, α-maltose 1-phosphate; T6P, trehalose 6-phosphate; ADPG, ADP-glucose; UDPG, UDP-glucose.