Date Published: August 11, 2016
Publisher: Public Library of Science
Author(s): Hendrik Koliwer-Brandl, Karl Syson, Robert van de Weerd, Govind Chandra, Ben Appelmelk, Marina Alber, Thomas R. Ioerger, William R. Jacobs, Jeroen Geurtsen, Stephen Bornemann, Rainer Kalscheuer, Helena Ingrid Boshoff.
Mycobacterium tuberculosis synthesizes intra- and extracellular α-glucans that were believed to originate from separate pathways. The extracellular glucose polymer is the main constituent of the mycobacterial capsule that is thought to be involved in immune evasion and virulence. However, the role of the α-glucan capsule in pathogenesis has remained enigmatic due to an incomplete understanding of α-glucan biosynthetic pathways preventing the generation of capsule-deficient mutants. Three separate and potentially redundant pathways had been implicated in α-glucan biosynthesis in mycobacteria: the GlgC-GlgA, the Rv3032 and the TreS-Pep2-GlgE pathways. We now show that α-glucan in mycobacteria is exclusively assembled intracellularly utilizing the building block α-maltose-1-phosphate as the substrate for the maltosyltransferase GlgE, with subsequent branching of the polymer by the branching enzyme GlgB. Some α-glucan is exported to form the α-glucan capsule. There is an unexpected convergence of the TreS-Pep2 and GlgC-GlgA pathways that both generate α-maltose-1-phosphate. While the TreS-Pep2 route from trehalose was already known, we have now established that GlgA forms this phosphosugar from ADP-glucose and glucose 1-phosphate 1000-fold more efficiently than its hitherto described glycogen synthase activity. The two routes are connected by the common precursor ADP-glucose, allowing compensatory flux from one route to the other. Having elucidated this unexpected configuration of the metabolic pathways underlying α-glucan biosynthesis in mycobacteria, an M. tuberculosis double mutant devoid of α-glucan could be constructed, showing a direct link between the GlgE pathway, α-glucan biosynthesis and virulence in a mouse infection model.
More than 100 years after its discovery by Robert Koch, Mycobacterium tuberculosis, the etiologic agent of tuberculosis, still remains an unresolved global public health threat. New and more effective chemotherapies for the treatment of tuberculosis are urgently required. Findings in recent years suggest that α-glucan biosynthesis in M. tuberculosis may be an attractive process that offers several vulnerable steps that could be exploitable in the development of novel treatment options. Many bacteria produce α-1,4/α-1,6-glucan. However, while this glucose polymer is typically assembled intracellularly in a glycogen-like storage form, M. tuberculosis also deposits α-glucan extracellularly as a major component of the capsule [1, 2]. Anecdotal reports have suggested the presence of a capsular layer surrounding mycobacteria for a long time [3–5], but only recently has this layer been visualized in a near native state . In vitro experiments using purified capsular α-glucan demonstrated that it can interact with complement receptor 3, thus mediating binding of M. tuberculosis to phagocytic cells [1, 7]. Capsular α-glucan also blocks dendritic cell functions  and interacts with the C-type lectin receptor DC-SIGN . Furthermore, consistent with the general importance of the capsule for virulence of many bacterial and fungal pathogens, an M. tuberculosis mutant with a somewhat reduced amount of capsular α-glucan showed impaired virulence . Collectively, these findings indicate that capsular α-glucans may be important for M. tuberculosis pathogenesis by interacting with mammalian host cells and influencing the immune response to M. tuberculosis. However, its precise role in virulence is unclear because the genes involved in its biosynthesis and export have not been conclusively elucidated, precluding the generation of capsule-deficient mutants.
Prior to this study, it was thought that three partially redundant routes collectively contribute to α-glucan production in M. tuberculosis and other mycobacteria: the GlgC-GlgA pathway for classical glycogen, the recently discovered TreS-Pep2-GlgE pathway for trehalose-to-glucan conversion, and the Rv3032 pathway. This multiplicity of pathways for the production of one type of molecule was puzzling, so it was unclear how these pathways interrelate in the biosynthesis of intracellular and capsular α-glucans. Refuting previous assumptions, we have now demonstrated that both cytosolic and capsular α-glucan polymers in M. tuberculosis, M. smegmatis and probably all other mycobacteria are predominantly, if not exclusively, synthesized by the maltosyltransferase GlgE together with the branching enzyme GlgB. In addition, the activated M1P donor substrate of GlgE is unexpectedly generated by two alternative routes: TreS-Pep2 as described previously  and GlgC-GlgA (Fig 6). These findings on the central role of the GlgE pathway in global glucan production in mycobacteria support our very recent observation that purified recombinant GlgE and GlgB together, using M1P as the sole substrate, are sufficient to generate α-glucans in vitro possessing the same distinctive structural features as the polymers isolated from the cytosol and capsule of M. tuberculosis .