Trehalose-recycling ABC transporter LpqY-SugA-SugBSugC is essential for virulence of Mycobacterium tuberculosis Rainer Kalscheuera,1,2, Brian Weinricka, Usha Veeraraghavanb, Gurdyal S. Besrab, and William R. Jacobs, Jr.a,2 a The Howard Hughes Medical Institute, Department of Microbiology and Immunology, The Albert Einstein College of Medicine, Bronx, NY 10461; and bSchool of Biosciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom
Edited by Hiroshi Nikaido, University of California, Berkeley, CA, and approved November 5, 2010 (received for review September 29, 2010)
microbial pathogenesis carbon metabolism
| mycolic acid biosynthesis | cell wall formation |
uberculosis, caused by the bacterium Mycobacterium tuberculosis (Mtb), remains a major threat to global health, claiming the life of two million individuals each year (1). Mtb is an obligate human pathogen predominantly growing intracellularly within phagosomes of host phagocytes, although other cell types and niches might also be occupied during different phases of infection. Notwithstanding, there is strong evidence that host lipids provide the main carbon and energy sources for Mtb during infection, with carbohydrates being largely inaccessible for the bacilli (2–5). Support for this, among further findings, comes from the observed up-regulation of lipid catabolism genes of Mtb during intracellular replication in macrophages (4) and from the joint essentiality of the two isocitrate lyase isoforms, icl1 and icl2, for growth of Mtb in mice (6). It has to be mentioned that the importance of some lipid catabolic pathways for in vivo carbon metabolism of Mtb may be somewhat overestimated, as attenuation of mutants might be caused by accumulation of toxic intermediates of incomplete metabolism rather than by blocked utilization of a substrate (7). Nevertheless, the published literature strongly suggests that Mtb relies on metabolism of lipids from the host via the glyoxylate cycle in vivo. The nature of the lipid substrates used by Mtb during infection, however, remains largely unclear. Recently, there has been growing evidence that cholesterol is a host lipid used by Mtb as one carbon and energy source in vivo, although additional yetunspecified substrates are clearly also important, as blockage of
T
www.pnas.org/cgi/doi/10.1073/pnas.1014642108
cholesterol uptake and metabolism only partially attenuates Mtb virulence (8, 9). In contrast to lipids, Mtb has probably only highly restricted access to host sugars. The strongest evidence for this comes from studies demonstrating that gluconeogenesis is essential for Mtb virulence during all phases of infection in mice (10). Despite the lack of substrate in its niche, the Mtb genome encodes four carbohydrate ATP-binding cassette (ABC) importers and one import system belonging to the major facilitator super family. These Mtb sugar importers were identified based on reasonable homology to carbohydrate transporters characterized in other bacteria, but the substrate for none of them is known (11). It has been a paradox that even though there appear to be no exogenous sugars available to transport, genome-wide screens of saturated transposon mutant libraries have implicated sugar uptake systems in the virulence of Mtb. In particular, the LpqY-SugA-SugB-SugC ABC transporter was predicted to be critical for growth both in macrophages (12) and in mice during the acute infection phase (13). These findings were interpreted as indicating that, despite the prevalence of host lipids, as-yet-unidentified host sugars are also available to Mtb early during infection, and that metabolism of these sugars is crucial for pathogenesis (5, 12–14). Here we show, however, that the LpqY-SugA-SugB-SugC ABC transporter is highly specific for uptake of the disaccharide trehalose, a sugar not present in mammals, thus unlikely to be involved in nutrient acquisition from the host. In contrast, it is demonstrated that this importer plays a role in recycling of extracellular trehalose released from trehalose-containing molecules synthesized by the bacillus. The glycolipid trehalose monomycolate (TMM), serving as a transport form for mycolic acids, is used as a substrate by the antigen 85 complex during formation of the mycolate-containing cell wall layer. During this extracellular enzymatic process, the trehalose moiety is released. Our data indicate that the dedicated function of the LpqY-SugA-SugB-SugC transporter is retrograde recycling of the trehalose released from TMM, a process shown to be critical for Mtb to establish infection in mice. Results LpqY-SugA-SugB-SugC ABC Transporter Mediates Trehalose Uptake in Mycobacterium smegmatis and Mtb. We fortuitously obtained hints
as to the function of the LpqY-SugA-SugB-SugC ABC transporter in Mtb while studying suppressor mechanisms involved in trehalose-resistance of the ΔglgE mutant in the model organism
Author contributions: R.K. and W.R.J. designed research; R.K., B.W., and U.V. performed research; R.K., G.S.B., and W.R.J. analyzed data; and R.K. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. 1
Present address: Institute for Medical Microbiology and Hospital Hygiene, HeinrichHeine-University Duesseldorf, 40225 Duesseldorf, Germany.
2
To whom correspondence may be addressed. E-mail:
[email protected]. de or
[email protected].
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1014642108/-/DCSupplemental.
PNAS | December 14, 2010 | vol. 107 | no. 50 | 21761–21766
MICROBIOLOGY
Mycobacterium tuberculosis (Mtb) is an exclusively human pathogen that proliferates within phagosomes of host phagocytes. Host lipids are believed to provide the major carbon and energy sources for Mtb, with only limited availability of carbohydrates. There is an apparent paradox because five putative carbohydrate uptake permeases are present in Mtb, but there are essentially no host carbohydrates inside phagosomes. Nevertheless, carbohydrate transporters have been implicated in Mtb pathogenesis, suggesting that acquisition of host sugars is important during some stages of infection. Here we show, however, that the LpqY-SugA-SugB-SugC ATP-binding cassette transporter is highly specific for uptake of the disaccharide trehalose, a sugar not present in mammals, thus refuting a role in nutrient acquisition from the host. Trehalose release is known to occur as a byproduct of the biosynthesis of the mycolic acid cell envelope by Mtb’s antigen 85 complex. The antigen 85 complex constitutes a group of extracellular mycolyl transferases, which transfer the lipid moiety of the glycolipid trehalose monomycolate (TMM) to arabinogalactan or another molecule of TMM, yielding trehalose dimycolate. These reactions also lead to the concomitant extracellular release of the trehalose moiety of TMM. We found that the LpqY-SugA-SugB-SugC ATP-binding cassette transporter is a recycling system mediating the retrograde transport of released trehalose. Perturbations in trehalose recycling strongly impaired virulence of Mtb. This study reveals an unexpected accessory component involved in the formation of the mycolic acid cell envelope in mycobacteria and provides a previously unknown role for sugar transporters in bacterial pathogenesis.
Mycobacterium smegmatis. GlgE is a maltosyltransferase involved in a recently described synthetic lethal trehalose-metabolizing α-glucan pathway in mycobacteria that utilizes maltose 1-phosphate (M1P) as a substrate (15). Although glgE is strictly essential in Mtb, inactivation of the glgE homolog in M. smegmatis results in a conditional lethal mutant strain that is highly sensitive to the exogenous presence of the disaccharide trehalose [α-D-glucopyranosyl-(1→1)-α-D-glucopyranoside]. In contrast to Mtb, the M. smegmatis ΔglgE mutant can tolerate the levels of M1P, which is formed from endogenous trehalose by the sequential catalytic action of the trehalose synthase TreS and the maltokinase Pep2. However, if exogenous trehalose is present in the growth medium, this disaccharide is efficiently taken up by the cells and channeled into the same pathway in addition to endogenous trehalose, leading to hyperaccumulation of M1P and causing bacteriostasis in the M. smegmatis ΔglgE mutant (15). During the screening for mutations abolishing the trehalose-sensitivity of the M. smegmatis ΔglgE mutant, we identified a suppressor mechanism based on the metabolic blockade of reactions leading to conversion of trehalose to M1P (i.e., mutations abolishing the function of the trehalose synthase TreS or the maltokinase Pep2) (15). In the present work, to identify additional suppressor mechanisms, we performed transposon mutagenesis of the M. smegmatis ΔglgE mutant, selecting for resistance to 1 mM trehalose. This approach led to insertions mapping to the locus lpqY-sugA-sugB-sugC (Fig. 1A). This locus encodes an ABC transporter highly conserved in mycobacteria comprised of the periplasmic sugar-binding lipoprotein LpqY, the ATP-binding protein SugC, and the transmembrane proteins SugA and SugB (Fig. 1A) (11, 16). In contrast to suppressor mutations in treS, transposon insertions in lpqY or sugC did not prevent conversion of endogenous trehalose to M1P, as revealed by TLC analysis (Fig. 1B), indicating a distinct mechanism of suppression. The ABC transporter LpqY-SugA-SugB-SugC has been speculated as being involved in the transport of the disaccharide maltose (16, 17). We hypothesized that the locus inactivated by our transposon insertions might be key for trehalose uptake. If so, it would follow that, although not affecting M1P formation from endogenous trehalose, the mechanism of suppression would be based on preventing uptake of exogenous trehalose and its conversion into M1P, thereby blocking hyperaccumulation of M1P
and bacteriostasis in the M. smegmatis ΔglgE mutant. To address the role of the locus in trehalose uptake, defined gene-deletion mutants in lpqY and sugC were generated in wild-type M. smegmatis (Fig. S1) and in Mtb (Fig. S2). The M. smegmatis ΔlpqY and ΔsugC mutants were unable to grow on trehalose as the sole carbon and energy source, whereas growth was not impaired on glucose or the disaccharide maltose (Fig. 1C). Given the structural similarity between the glucose dimers trehalose and maltose [α-D-glucopyranosyl-(1→4)-D-glucopyranoside], this indicates that uptake of these sugars in M. smegmatis is mediated by nonredundant transport systems with a high degree of specificity. Likewise, Mtb ΔlpqY and ΔsugC mutants also showed no growth on trehalose as the sole carbon and energy source, although growing normally on glucose (Fig. 1D). The growth defect of the Mtb mutants was genetically complemented by a single-copy integrative plasmid expressing the lpqY-sugC operon from its native promoter and the ΔlpqY mutant exhibited intermediate complementation (Fig. 1D). Growth of the mutants on maltose was not tested, as wild-type Mtb was unable to use this sugar. These findings demonstrate the essential role of the LpqY-SugA-SugBSugC ABC transporter in trehalose acquisition in M. smegmatis and Mtb. LpqY-SugA-SugB-SugC ABC Transporter Is a High-Affinity Permease Specific for Trehalose. Next, uptake rates were determined with
whole cells of Mtb wild-type and mutant strains using radiolabeled 14 C-trehalose. Compared with ABC-transporter–mediated import of sugar molecules in other bacteria [e.g., a rate of 2 nmol/ min/108 cfu has been reported for maltose uptake in Escherichia coli (18)], uptake of trehalose in Mtb wild-type was rather slow, with a rate of ∼25 pmol/h/109 cfu, consistent with the extremely slow growth rate of the bacillus. Inactivation of the transporter in the ΔlpqY and ΔsugC mutants completely abolished trehalose uptake (Fig. 2A). Transport was restored in the mutants by genetic complementation. Furthermore, trehalose uptake was fully abrogated in Mtb wild-type by heat inactivation, implying that it is an active process and ruling out any interference by the possible unspecific binding of trehalose to the cell surface (Fig. 2A). These findings, together with the inability of the mutants to use trehalose as a carbon source, clearly show that the LpqY-SugA-SugB-SugC
Fig. 1. The LpqY-SugA-SugB-SugC ABC transporter mediates trehalose uptake in mycobacteria. (A) Organization of the lpqY-sugA-sugB-sugC locus in M. smegmatis and Mtb. The triangles indicate the transposon insertion sites in trehalose-resistant M. smegmatis ΔglgE suppressor mutants. (B) M1P accumulation from endogenous trehalose in transposon-induced M. smegmatis suppressor mutants revealed by TLC analysis. Cells were cultivated in absence of exogenous trehalose.::Tn, transposon insertion. (C) Growth of M. smegmatis transporter mutants on defined carbon sources after 48 h of incubation. Data represent means of triplicates ± SD. (D) Growth kinetics of Mtb transporter mutants on defined carbon sources. Data represent means of triplicates ± SD. Comp., complemented mutant strains. Data in B to D are representative of at least two independent experiments.
21762 | www.pnas.org/cgi/doi/10.1073/pnas.1014642108
Kalscheuer et al.
Fig. 2. The LpqY-SugA-SugB-SugC ABC transporter is a highaffinity trehalose-specific importer. (A) Trehalose uptake rates in Mtb mutant strains. HI, cells heat-inactivated by incubation at 95 °C for 15 min before addition of the radiolabeled trehalose. Comp., complemented mutant strains. (B) Dependence of trehalose uptake rate on substrate concentration. (C) Trehalose uptake activity in Mtb wild-type does not require induction by the substrate. Wild-type cells were preincubated for 24 h in presence of 2 mM trehalose before uptake experiments. (D) Determination of the substrate specificity in trehalose uptake competition assays in presence of 1 mM of the indicated unlabeled competing sugars. (E) Comparison of the uptake rates for trehalose and maltose in Mtb. Data represent means of triplicates ± SD and are representative of two independent experiments.
ABC transporter provides the sole entry route for trehalose in Mtb and M. smegmatis. Enzymatic characterization revealed that this permease efficiently acquires trehalose at low nanomolar concentrations. No saturation of uptake was observed up to 6 μM trehalose (Fig. 2B). Assuming a Michaelis-Menten-like kinetic, a Vmax of ∼500 ± 130 pmol/h/109 cfu and an apparent KM of ∼19 ± 6 μM (Fig. 2B) were obtained, which is not unusual for carbohydrate ABC importers and which classifies the LpqY-SugA-SugB-SugC transporter as a highaffinity permease (18–20). Preincubation of the cells in the presence of trehalose did not alter the uptake rate, suggesting that the LpqYSugA-SugB-SugC importer is constitutively expressed (Fig. 2C). To assess the substrate specificity, uptake competition assays were performed with 14C-trehalose in the presence of an excess of unlabeled competing sugars. Except for the positive-control trehalose, no competition was observable using such structurally similar molecules as glucose, maltose [α-D-glucopyranosyl(1→4)-D-glucopyranoside], or even the nonnatural stereoisomer α,β-trehalose [α-D-glucopyranosyl-(1→1)-β-D-glucopyranoside] (Fig. 2D). This finding provides strong evidence that the LpqY-SugA-SugB-SugC ABC transporter exhibits a high degree of specificity for trehalose. To corroborate the specificity for trehalose versus maltose, uptake experiments were performed with 14C-labeled maltose. Under the test conditions, no uptake of maltose was measureable in Mtb cells (Fig. 2E), proving that maltose is not accepted as a substrate by the LpqY-SugA-SugBSugC importer or by any other import system in Mtb.
burdens at day 112 postinfection (P = 0.0281; unpaired t test). This finding might be because of partial compensation by the ATPhydrolyzing protein of another ABC transporter, as these transporter components typically exhibit a high degree of homology among each other in contrast to the transmembrane proteins and periplasmic binding proteins. Genetic complementation of the mutants fully restored virulence, unequivocally confining the mutant phenotype to the lpqY-
SugA-SugB-SugC ABC transporter in pathogenesis, mouse infection studies with mutant strains were performed. Trehalose transporter mutants exhibited a strong attenuation in the immunocompromised SCID mouse model. Mice infected via the aerosol route with the ΔlpqY (median survival time 77.5 d) or ΔsugC mutant (median survival time 79 d) survived significantly longer (P < 0.0001; Log-rank test) than mice infected with the wild-type (median survival time 46 d) or the complemented mutant strains (Fig. 3A). In immunocompetent C57BL/6 mice, trehalose transporter mutants showed a severe growth defect in lungs of infected animals during the acute phase of infection, but were able to persist during the chronic phase, albeit at a strongly reduced bacillary organ burden (Fig. 3B). Furthermore, dissemination of mutants to the spleen was greatly reduced. The degree of attenuation observed for the ΔsugC mutant was slightly lower compared with that for the ΔlpqY strain, although differences were only significant for lung Kalscheuer et al.
MICROBIOLOGY
Trehalose Uptake Is Critical for Virulence of Mtb During the Acute Infection Phase in Mice. To investigate the role of the LpqY-
Fig. 3. Trehalose acquisition by the LpqY-SugA-SugB-SugC ABC transporter is essential for virulence of Mtb in mice. (A) Survival of SCID mice after lowdose aerosol infections with Mtb trehalose transporter mutants. n = 10 mice per group. (B) Growth kinetics of Mtb trehalose transporter mutants in lungs and spleens of C57BL/6 mice after low-dose aerosol infections. Values represent means of triplicates ± SD. Comp., complemented mutant strains.
PNAS | December 14, 2010 | vol. 107 | no. 50 | 21763
sugA-sugB-sugC locus (Fig. 3B). These data, which are in agreement with previous studies (13), demonstrate the crucial importance of the LpqY-SugA-SugB-SugC transporter-mediated uptake of trehalose for the virulence of Mtb. LpqY-SugA-SugB-SugC ABC Transporter Is a Recycling System for Cell Wall-Released Trehalose. In light of the observed high specificity
for trehalose, a disaccharide not present in mammals, the observed importance for virulence of the LpqY-SugA-SugB-SugC ABC transporter appeared paradoxical. Thus, it is not reasonable that the LpqY-SugA-SugB-SugC ABC transporter is involved in the acquisition of host sugars. Alternatively, we hypothesized that this importer is rather involved in the uptake of a carbohydrate derived from the bacillus itself. Among other possible cellular functions, trehalose might play a carrier role in the form of the glycolipid TMM in the biosynthesis of the mycolic acid cell envelope, the hallmark of mycobacteria. According to the current model of mycolic acid processing, conjugation of mycolic acids and trehalose to form TMM occurs in the cytoplasm by means of an asyet-unknown enzymatic activity. Subsequently, TMM may function as the mycolate transport form that is secreted by an as-yet-unidentified export process. Extracellularly, TMM serves as the substrate for the mycolyl transferases of the antigen 85 complex (i.e., FbpA, FbpB, FbpC), which transfer the mycolic acid moiety of TMM either to the arabinogalactan cell wall layer or to another molecule of TMM, yielding trehalose dimycolate (TDM). These enzymatic reactions also lead to the concomitant release of the trehalose moiety of TMM in the periplasm (21–23). Whereas the essential role of TMM as the extracellular substrate in the biosynthesis of the mycolic acid cell envelope in mycobacteria is generally accepted (21–23), the fate of the released trehalose has never been addressed. We therefore hypothesized that the function of the LpqY-SugA-SugB-SugC ABC transporter is the recycling of this cell wall-released trehalose. Consequently, impairment of recycling should lead to the release of free trehalose by the cells. To test this hypothesis, we analyzed cell-free culture supernatants from Mtb wild-type and trehalose transporter mutants for the presence of trehalose. This prediction was confirmed: Mtb ΔlpqY and ΔsugC mutants secreted substantial amounts of trehalose (accumulating to more than 120 μM in stationary-phase cultures) during growth on glycerol, whereas virtually no trehalose was detected in culture supernatants of Mtb wild-type and the complemented mutant strains (Figs. 4 A and B). The presence of trehalose in culture supernatants was confirmed by coupled GC/MS, validating the release of trehalose by the Mtb ΔlpqY and ΔsugC mutants and its absence in supernatants of Mtb wild-type and the complemented mutant strains (Fig. 4C and Fig. S3). Aside from trehalose, no signs of the secretion of other carbohydrate metabolites were detected by comparison of the sugar profiles of supernatants of these Mtb strains (Fig. 4C). Impaired Trehalose Recycling Does Not Alter the Glycolipid Profile in Mtb. The mycobacterial cell envelope is rich in glycolipids containing
trehalose as a sugar moiety, such as TMM, TDM, di-, tri, and pentaacyltrehalose, or sulfolipid, some of which play an important role in virulence and possess potent immunomodulatory properties (24, 25). We therefore asked whether the presence of free trehalose might interfere with biosynthesis of trehalose-containing Mtb glycolipids, as alterations in the lipid profile could have contributed to the observed attenuation of the transporter mutants in mice. Lipid profiling using 14C-acetate labeling of cells followed by 2D TLC analyses, however, revealed no discernable differences in the glycolipid spectrum of trehalose transporter mutants grown in vitro. Prominent trehalose-containing glycolipids, such as TMM, TDM, and sulfolipid, as well as other abundant glycolipids, such as phosphatidylinositol mannosides, were unaltered (Fig. S4). These findings rule out direct trehalose-induced differences in cell wall lipids as a cause of attenuation of the Mtb transporter mutants. Labeling of cells with 14C-trehalose resulted only in poor incorporation of radioactivity into Mtb lipids. Surprisingly, no detectable labeling of the prominent trehalose-containing glycolipids, TMM, TDM, and 21764 | www.pnas.org/cgi/doi/10.1073/pnas.1014642108
Fig. 4. The LpqY-SugA-SugB-SugC ABC transporter is a trehalose recycling system. (A) Growth kinetics of Mtb strains on 40 mM glycerol as sole carbon source. (B) Trehalose accumulation in cell-free culture supernatants of Mtb strains during growth on glycerol as sole carbon source, as indicated in A. Trehalose was determined using an enzymatic quantification assay. Data in A and B represent means of triplicates ± SD and are representative of two independent experiments. (C) Verification of trehalose secretion by Mtb transporter mutants, as revealed by GC/MS analyses. Samples were taken from 26-d-old cultures, as shown in A and B. The peak at 14.7 min corresponded to that of an authentic trehalose standard and showed an identical mass spectrum (Fig. S3). Comp., complemented mutant strains.
sulfolipid, occurred under the tested conditions. In the Mtb ΔlpqY mutant, there was no incorporation of radioactivity from 14C-trehalose into Mtb cell wall lipids whatsoever (Fig. S4). Discussion There is an increasing body of evidence that Mtb relies on gluconeogenic carbon sources, such as fatty acids and amino acids, for growth and persistence in mice, and that host sugars are largely inaccessible for the bacillus (10). Nevertheless, the Mtb genome encodes five transporters implicated in sugar uptake. The LpqY-SugA-SugB-SugC ABC transporter was predicted to be of critical importance for virulence of Mtb, leading to the speculation that unidentified host sugars are also available to Mtb, at least early during infection (5, 12–14). Our studies, however, now convincingly show that this permease exhibits a high specificity for trehalose, a disaccharide not synthesized by mammals. Mammals routinely take up small quantities of trehalose with their diets, as this disaccharide is commonly present in yeast- and fungi-derived food products, as well as in gut flora bacteria. However, the ingested trehalose is normally hydrolyzed quantitatively to glucose by the enzyme trehalase, which is present in the intestinal mucosa as well as in the blood serum, the kidneys, and the liver (26). Therefore, free trehalose would not reach tissue levels relevant to mycobacterial growth in mammals unless, perhaps, they were fed a specific highly trehalose-enriched diet (27). Thus, we conclude that the LpqY-SugA-SugB-SugC ABC transporter is unlikely involved in acquisition of host sugars. In contrast, our data suggest that this permease functions as an efficient and specific retrograde recycling system for trehalose, which may be extracellularly released by Mtb and other mycobacteria from trehalose-containing molecules as a by-product of cell wall biosynthesis. Although the involvement of other processes cannot be excluded, a likely source of extracellular trehalose are the secreted mycolyl transferases of the antigen 85 complex, which transfer the mycolate moiety of TMM, either to arabinogalactan or to another TMM molecule, yielding TDM, with concomitant extracellular release of the treKalscheuer et al.
Fig. 5. Model of trehalose recycling as an accessory component in mycolic acid processing. AG, arabinogalactan layer; Ag85, antigen 85 complex; CM, cytoplasmic membrane; PG, peptidoglycan layer; TMM, trehalose monomycolate; TDM, trehalose dimycolate.
Kalscheuer et al.
a unique function for an ABC sugar transporter and a previously unappreciated aspect of mycobacterial pathogenesis. Our data do not exclude other mechanisms of attenuation. For example, although trehalose secretion has no direct effect on the glycolipid profile of Mtb in vitro (Fig. S4), the mutation might lead to an altered trehalose metabolism during in vivo growth, which might cause a specific preceding depletion of trehalosecontaining glycolipids before general starvation responses will manifest. Subtle changes in these potent immunomodulatory lipids, although difficult to detect during in vivo growth, could play a significant role for the observed attenuation of the mutants. Furthermore, it is conceivable that the released trehalose could induce innate immune mechanisms in the infected host cells capable of restricting growth of the bacilli. As an example, it has been demonstrated that trehalose can induce autophagy in various cell types (31), an innate defense mechanism known to limit intracellular replication of mycobacteria (32). However, because relatively high concentrations are necessary, it is questionable whether Mtb trehalose transporter mutants would be effective in triggering this process. Because of its importance for virulence, the LpqY-SugA-SugBSugC ABC transporter may represent a new target for chemotherapeutic treatment of acute Mtb infections, although inhibition of this process would probably have little effect against persistent bacilli for treatment of latent infections. As a surface-exposed system that includes an extracytoplasmic component (i.e., the extracellular substrate-binding lipoprotein LpqY), this ABC transporter could be targeted by compounds that manifest their inhibitory activity on the surface of the bacterial cell. Bypassing the need for uptake to reach an intracellular target avoids a major obstacle that limits the antibacterial potency of many inhibitors. Although close mechanistic relationships exist to human ABC exporters, the absence from humans of ABC importers and extracellular substrate-binding proteins strengthens the potential of this system as a target for tuberculosis chemotherapy, particularly when focusing on the lipoprotein LpqY as the primary target. Materials and Methods Strains and Growth Conditions. All strains were derived from M. smegmatis mc2155 and M. tuberculosis H37Rv. The M. smegmatis ΔglgE mutant has been described previously (15). Cells were grown aerobically at 37 °C in Middlebrook 7H9 medium supplemented with 10% (vol/vol) OADC enrichment (Becton Dickinson Microbiology Systems), 0.5% (vol/vol) glycerol and 0.05% (vol/vol) Tyloxapol. For growth on defined carbon sources, the following minimal medium was used: 1 g/L KH2PO4, 2.5 g/L Na2HPO4, 0.5 g/L (NH4)2SO4, 0.15 g/L asparagine, 50 mg/L ferric ammonium citrate, 0.5 g/L MgSO4 × 7 H2O, 0.5 mg/L CaCl2, 0.1 mg/L ZnSO4, 0.05% (vol/vol) Tyloxapol (pH 7.0). Sugars were added at the indicated concentrations. Hygromycin (50 mg/L) and kanamycin (20 mg/L) were added for selection for appropriate strains. Generation of Transposon-Induced and Targeted Gene-Deletion Mutants. Mutants of M. smegmatis and Mtb were generated by allelic exchange or random mutagenesis employing a Himar-1 mariner transposon, respectively, using specialized transduction (33, 34). Details on the generation of mutants and genetic complementation are described in SI Materials and Methods and Table S1. Determination of Trehalose-Uptake Rates. Trehalose uptake was measured using exponentially growing Mtb strains suspended in PBS containing 0.05% (vol/vol) Tyloxapol using [U-14C]-α-D-trehalose (specific activity 600 mCi/mmol; American Radiolabeled Chemicals) at a concentration of 167 nM, unless indicated otherwise. In one experiment, [U-14C]-α-D-maltose (specific activity 600 mCi/mmol; Sigma) was used at a concentration of 167 nM. After incubation at 37 °C, cells were washed twice with PBS containing 0.05% (vol/vol) Tyloxapol and heat inactivated at 100 °C for 60 min. Incorporation of radioactivity into biomass was quantified by scintillation counting. Cell numbers were estimated by measuring OD at 600 nm by calculating with 3 × 108 cfu at OD = 1. Under these conditions, uptake rates were linear for at least 180 min. Typically, cells were incubated for 120 min for uptake experiments. TLC Analysis of Lipids and Carbohydrates. Lipid extracts from cells labeled for 24 h with [1-14C]-acetate (specific activity 50 mCi/mmol; Perkin-Elmer; final
PNAS | December 14, 2010 | vol. 107 | no. 50 | 21765
MICROBIOLOGY
halose moiety of TMM (21–23). Whereas the role of TMM as an extracellular substrate in biosynthesis of the mycolate cell wall layer in mycobacteria is generally accepted, the enzymes involved in biosynthesis of TMM and their cellular localization, as well as the potential mechanism of export, are unknown (23). Our findings are most consistent with the concept that conjugation of mycolic acids and trehalose to form TMM occurs in the cytoplasm and that TMM serves as the mycolate transport form, which is secreted by a yet unidentified export process (Fig. 5). It has to be mentioned that an alternative model for the formation of the mycolic acid cell envelope has been proposed based on studies in Corynebacterium, with trehalose being secreted in free form by the cells followed by the subsequent extracellular transfer of mycolyl residues onto the sugar molecule (28). However, our findings in mycobacteria do not agree with this model. First, in light of the presence of a trehalose-specific importer, the secretion of free trehalose seems counterproductive, as this would result in a futile cycle. Second, we found that extracellular radiolabeled trehalose was not efficiently incorporated into trehalose-containing glycolipids in Mtb, as would be expected if the conjugation indeed occurred outside the cell (Fig. S4). Third, any labeling of lipids in Mtb strictly depended on the uptake of radiolabeled trehalose via the LpqY-SugA-SugB-SugC importer, proving that incorporation takes place in the cytoplasm (Fig. S4). Furthermore, intracellular dilution of the label by the high concentration of endogenous trehalose typically present in the cytoplasm of mycobacteria (29) can explain the poor incorporation into lipids. The molecular basis for the importance of the sugar transporter in pathogenesis has been elusive. The scenario for Mtb now appears to be that impaired trehalose recycling successively leads to the loss of a sugar that is required for constructing the extracellular mycolic acid coat. Trehalose can also serve as a carbon source and its loss may contribute to starvation during growth in nutrient-limited microenvironments in vivo. Thus, trehalose recycling may constitute an accessory component of mycolic acid processing, which evolved as a specific adaptation to Mtb’s nutrient-restricted intracellular lifestyle by maintaining energy efficiency of cell wall biosynthesis (Fig. 5). Trehalose recycling is most important during the acute infection phase, where the bacillus is exponentially growing and replicating. The observed dispensability of trehalose recycling for Mtb persistence in mice during the chronic infection phase is consistent with the assumption that this phase is characterized by limited replication, and hence minimal requirement for cell wall biosynthesis, of the bacilli, and hence minimal coupled secretion of trehalose. The described recycling strategy to preserve scarce resources suggests a previously unexplored context for understanding the function of many importers for sugars and other metabolites in Mtb and in other microbes occupying nutrientdeficient environments as saprophytes, pathogens or commensals, resembling recycling of cell wall peptidoglycan breakdown products by bacteria (30). However, to our knowledge, this is
concentration 1 μCi/mL) or [U-14C]-trehalose (specific activity 600 mCi/mmol; American Radiolabeled Chemicals; final concentration 0.1 μCi/mL) were separated by 2D TLC systems, as described (35), by applying equal amounts of lipids per TLC plate and by using the solvent system D for separation of apolar lipids and E for separation polar lipids, respectively. Lipids were visualized by autoradiography. Carbohydrates were extracted from equal amounts of cells with hot water (95 °C for 4 h) and analyzed by TLC on Silica gel 60 (EMD Chemicals) using the solvent system 1-propanol:ethyl acetate: water (6:1:3, vol/vol), as described (15). Trehalose Determination. For enzymatic trehalose quantifications, 25 μL aliquots of cell-free culture supernatants were mixed with 15 μL acetic acid (1 M), 60 μL sodium acetate (0.2 M; pH 5.2), and 5 μL trehalase solution (3.7 U/mL; Sigma) and incubated at 37 °C for 15 h. For quantification of the liberated glucose, 10 μL of this trehalase-treated sample was mixed with 100 μL glucose oxidase/peroxidase/o-dianisidine solution (Sigma) and incubated at 37 °C for 30 min. Reactions were stopped by the addition of 75 μL sulphuric acid (6 M), and absorbance was measured at 531 nm. Controls without trehalase addition were done to correct for the presence of free glucose in the samples. Trehalose concentrations were estimated using a standard curve. Trehalose in lyophilized cell-free culture supernatants was acetylated and analyzed by coupled GC/MS. The samples were acetylated as reported (36) with modifications. Briefly, 250 μL of acetic anhydride and 250 μL of pyridine (1:1 vol/vol) were added to the lyophilized culture supernatant or trehalose standard, respectively, and stirred at 80 °C for 2 h followed by stirring at room temperature overnight. Next, 2 mL of chloroform and 2 mL of water were added and mixed, and the chloroform phase was separated by centrifugation (37). The chloroform phase was removed and washed twice with 1 M HCl and once with water, then dried, and the resulting acetylated trehalose was solubilized in chloroform and analyzed by GC/MS. GC/MS analysis was performed using an Agilent 7890 Gas chromatograph attached to a Waters GCT time-of-flight mass spectrometer.
1. Dye C (2006) Global epidemiology of tuberculosis. Lancet 367:938–940. 2. Boshoff HI, Barry CE, 3rd (2005) Tuberculosis—metabolism and respiration in the absence of growth. Nat Rev Microbiol 3:70–80. 3. Muñoz-Elías EJ, McKinney JD (2006) Carbon metabolism of intracellular bacteria. Cell Microbiol 8:10–22. 4. Schnappinger D, et al. (2003) Transcriptional adaptation of Mycobacterium tuberculosis within macrophages: Insights into the phagosomal environment. J Exp Med 198:693–704. 5. Eisenreich W, Dandekar T, Heesemann J, Goebel W (2010) Carbon metabolism of intracellular bacterial pathogens and possible links to virulence. Nat Rev Microbiol 8: 401–412. 6. Muñoz-Elías EJ, McKinney JD (2005) Mycobacterium tuberculosis isocitrate lyases 1 and 2 are jointly required for in vivo growth and virulence. Nat Med 11:638–644. 7. Upton AM, McKinney JD (2007) Role of the methylcitrate cycle in propionate metabolism and detoxification in Mycobacterium smegmatis. Microbiology 153:3973–3982. 8. Pandey AK, Sassetti CM (2008) Mycobacterial persistence requires the utilization of host cholesterol. Proc Natl Acad Sci USA 105:4376–4380. 9. Yam KC, et al. (2009) Studies of a ring-cleaving dioxygenase illuminate the role of cholesterol metabolism in the pathogenesis of Mycobacterium tuberculosis. PLoS Pathog 5:e1000344. 10. Marrero J, Rhee KY, Schnappinger D, Pethe K, Ehrt S (2010) Gluconeogenic carbon flow of tricarboxylic acid cycle intermediates is critical for Mycobacterium tuberculosis to establish and maintain infection. Proc Natl Acad Sci USA 107:9819–9824. 11. Titgemeyer F, et al. (2007) A genomic view of sugar transport in Mycobacterium smegmatis and Mycobacterium tuberculosis. J Bacteriol 189:5903–5915. 12. Rengarajan J, Bloom BR, Rubin EJ (2005) Genome-wide requirements for Mycobacterium tuberculosis adaptation and survival in macrophages. Proc Natl Acad Sci USA 102:8327–8332. 13. Sassetti CM, Rubin EJ (2003) Genetic requirements for mycobacterial survival during infection. Proc Natl Acad Sci USA 100:12989–12994. 14. Cook GM, et al. (2009) Physiology of mycobacteria. Adv Microb Physiol 55:81–182. 15. Kalscheuer R, et al. (2010) Self-poisoning of Mycobacterium tuberculosis by targeting GlgE in an alpha-glucan pathway. Nat Chem Biol 6:376–384. 16. Braibant M, Gilot P, Content J (2000) The ATP binding cassette (ABC) transport systems of Mycobacterium tuberculosis. FEMS Microbiol Rev 24:449–467. 17. Borich SM, Murray A, Gormley E (2000) Genomic arrangement of a putative operon involved in maltose transport in the Mycobacterium tuberculosis complex and Mycobacterium leprae. Microbios 102:7–15. 18. Szmelcman S, Schwartz M, Silhavy TJ, Boos W (1976) Maltose transport in Escherichia coli K12. A comparison of transport kinetics in wild-type and lambda-resistant mutants as measured by fluorescence quenching. Eur J Biochem 65:13–19. 19. Brunkhorst C, Schneider E (2005) Characterization of maltose and maltotriose transport in the acarbose-producing bacterium Actinoplanes sp. Res Microbiol 156:851–857. 20. Hülsmann A, Lurz R, Scheffel F, Schneider E (2000) Maltose and maltodextrin transport in the thermoacidophilic Gram-positive bacterium Alicyclobacillus acidocaldarius is
21766 | www.pnas.org/cgi/doi/10.1073/pnas.1014642108
Animal Infections. C57BL/6 and SCID mice (BALB/c background) (4- to 6-wk-old females; National Cancer Institute, Bethesda, MD) were infected via the aerosol route using a whole-body aerosolization chamber. A suspension of exponentially growing Mtb strains suspended at a density of 106 cfu/mL in PBS containing 0.05% (vol/vol) Tween 80 and 0.004% (vol/vol) antifoam (Sigma) was used as the inoculum. For assessment of growth kinetics, three mice per group were killed at the indicated time points, and bacterial burden was determined by plating serial dilutions of lung and spleen homogenates onto Middlebrook 7H10 agar plates supplemented with 10% (vol/ vol) OADC enrichment and 0.5% (vol/vol) glycerol. For survival experiments, 10 mice per group were infected and survival of mice was monitored. Mouse protocols were approved by the Animal Care and Use Committee of the Albert Einstein College of Medicine. Statistical Analysis. Statistical significance of survival rates was determined with the Log-rank (Mantel-Cox) test and of differential organ burdens with the unpaired t test at P < 0.05 level of significance using GraphPad Prism 5 software. ACKNOWLEDGMENTS. We thank B. Chen, M. Chen, and J. Kim for technical support during the animal experiments, T. Hsu for the provision of plasmid p0004S and J. Kriakov for the temperature-sensitive phage phAE159 (both Albert Einstein College of Medicine, Bronx, NY), N. Veerapen and P. Ashton for help with the acetylation of samples and in the GC/MS experiments, respectively, and D. Thaler, S. Bornemann, and D. Hopwood for helpful discussions and editorial suggestions. W.R.J. acknowledges generous support from the Albert Einstein College of Medicine Center for AIDS Research Grant AIO-51519. R.K. acknowledges support from the Juergen Manchot Foundation. G.S.B. acknowledges support in the form of a Personal Research Chair from Mr. James Bardrick, Royal Society Wolfson Research Merit Award, as a former Lister Institute Jenner Research Fellow, the Medical Research Council, and the Wellcome Trust (081569/2/06/2). This work was supported by National Institutes of Health Grants AI26170 and AI59877 (to W.R.J.).
21. 22. 23. 24. 25. 26. 27. 28.
29.
30. 31.
32.
33.
34. 35. 36. 37.
mediated by a high-affinity transport system that includes a maltose binding protein tolerant to low pH. J Bacteriol 182:6292–6301. Belisle JT, et al. (1997) Role of the major antigen of Mycobacterium tuberculosis in cell wall biogenesis. Science 276:1420–1422. Ronning DR, et al. (2000) Crystal structure of the secreted form of antigen 85C reveals potential targets for mycobacterial drugs and vaccines. Nat Struct Biol 7:141–146. Takayama K, Wang C, Besra GS (2005) Pathway to synthesis and processing of mycolic acids in Mycobacterium tuberculosis. Clin Microbiol Rev 18:81–101. Daffe M, Lemassu A (2000) Glycomicrobiology, ed Doyle RJ (Plenum Press, New York), pp 225–273. Karakousis PC, Bishai WR, Dorman SE (2004) Mycobacterium tuberculosis cell envelope lipids and the host immune response. Cell Microbiol 6:105–116. Gudmand-Høyer E, Skovbjerg H (1996) Disaccharide digestion and maldigestion. Scand J Gastroenterol Suppl 216:111–121. Tanaka M, et al. (2004) Trehalose alleviates polyglutamine-mediated pathology in a mouse model of Huntington disease. Nat Med 10:148–154. Tropis M, et al. (2005) The crucial role of trehalose and structurally related oligosaccharides in the biosynthesis and transfer of mycolic acids in Corynebacterineae. J Biol Chem 280:26573–26585. Elbein AD, Mitchell M (1973) Levels of glycogen and trehalose in Mycobacterium smegmatis and the purification and properties of the glycogen synthetase. J Bacteriol 113:863–873. Park JT, Uehara T (2008) How bacteria consume their own exoskeletons (turnover and recycling of cell wall peptidoglycan). Microbiol Mol Biol Rev 72:211–227. Sarkar S, Davies JE, Huang Z, Tunnacliffe A, Rubinsztein DC (2007) Trehalose, a novel mTOR-independent autophagy enhancer, accelerates the clearance of mutant huntingtin and alpha-synuclein. J Biol Chem 282:5641–5652. Deretic V, et al. (2009) Autophagy in immunity against Mycobacterium tuberculosis: A model system to dissect immunological roles of autophagy. Curr Top Microbiol Immunol 335:169–188. Bardarov S, et al. (2002) Specialized transduction: An efficient method for generating marked and unmarked targeted gene disruptions in Mycobacterium tuberculosis, M. bovis BCG and M. smegmatis. Microbiology 148:3007–3017. Bardarov S, et al. (1997) Conditionally replicating mycobacteriophages: A system for transposon delivery to Mycobacterium tuberculosis. Proc Natl Acad Sci USA 94:10961–10966. Besra GS (1998) Preparation of cell-wall fractions from mycobacteria. Methods Mol Biol 101:91–107. Besra GS, et al. (1992) Structural elucidation of a novel family of acyltrehaloses from Mycobacterium tuberculosis. Biochemistry 31:9832–9837. Besra GS, et al. (1995) A new interpretation of the structure of the mycolyl-arabinogalactan complex of Mycobacterium tuberculosis as revealed through characterization of oligoglycosylalditol fragments by fast-atom bombardment mass spectrometry and 1H nuclear magnetic resonance spectroscopy. Biochemistry 34:4257–4266.
Kalscheuer et al.
