Trehalose dimycolate inhibits phagosome maturation and promotes intracellular M. tuberculosis growth via noncanonical SNARE interactions

Abstract

Mycobacterial cell envelopes are rich in unusual lipids and glycans that play key roles during infection and vaccination. The most abundant envelope glycolipid is trehalose dimycolate (TDM). TDM compromises the host response to mycobacterial species via multiple mechanisms, including inhibition of phagosome maturation. The molecular mechanism by which TDM inhibits phagosome maturation has been elusive. We find that a clickable, photoaffinity TDM probe recapitulates key phenotypes of native TDM in macrophage host cells and binds several host Soluble N-ethylmaleimide-Sensitive Factor Attachment Proteins Receptor (SNARE) proteins, including Vesicle Transport through Interaction with t-SNAREs 1B (VTI1B), Syntaxin 8 (STX8), and Vesicle-Associated Membrane Protein 2 (VAMP2). VTI1B and STX8 normally promote endosome fusion by forming a complex with VAMP8. However, in the presence of Mycobacterium tuberculosis, VTI1B and STX8 complex with VAMP2, which in turn decreases VAMP8 binding. VAMP2 acts together with mycolate structure to inhibit phagosome maturation and promotes intracellular M. tuberculosis replication. Thus one mechanism by which TDM constrains the innate immune response to M. tuberculosis is via noncanonical SNARE complexation.

Keywords: Mycobacterium tuberculosis; SNARE; cell envelope; click chemistry; phagosome.

Fig. 1.

Chemical probes can reveal interactions between (myco)bacterial lipids and host proteins. (A) The glycolipid TDM in the outer mycomembrane (MM) of M. tuberculosis (Mtb) inhibits fusion of lysosomes with Mtb-containing phagosomes. Other components of the mycobacterial cell envelope include arabinogalactan (AG), peptidoglycan (PG), and the inner plasma membrane (PM). (B) Workflow for chemical proteomics using the TDM-mimicking probe O-x-AlkTDM. (C) Structures of O-x-AlkTDM and control probes O-x-AlkTMM and x-AlkFA.

Fig. 2.

Mycobacterial glycolipid probe O-x-AlkTDM recapitulates key aspects of TDM activity and binding in macrophages. ELISA detection of proinflammatory cytokines (A) TNF-alpha and (B) IL-6 after 6 h incubation of iBMDMs from C57BL/6 mice with O-x-AlkTDM or commercial TDM mimic TDB. For (A and B) each color in the SuperPlots (37) represents an independent biological replicate. Smaller symbols represent technical replicates, and larger symbols are the means of the technical replicates per experiment. Statistical significance determined using a two-way ANOVA with a Šídák’s multiple comparisons test for post hoc analysis. Data obtained from three independent experiments. (C and D) Lysosome fusion of bead-containing phagosomes. THP-1 cells were incubated with BSA-, TDB-, TDM-, or O-x-alkTDM-coated fluorescent beads (green), followed by LysoTracker staining to assess acidification. Green and white arrows indicate LysoTracker-localized beads; white arrows highlight beads that have been magnified for enhanced visualization. Representative images shown in (C) and (SI Appendix, Fig. S5C). In (D), spatial coincidence between beads and LysoTracker was quantitated from 3 to 5 independent biological replicates. Images were blinded prior to analysis. White arrows indicate LysoTracker-localized beads that have been magnified for enhanced visualization, while green arrows highlight additional Lysotracker staining present within the image. (E and F) O-x-AlkTDM-mediated affinity enrichment of host interacting proteins. iBMDMs were incubated with O-x-AlkTDM (100 µM), UV-irradiated, and lysed. Lysates were reacted with AzTB via CuAAC and analyzed by Coomassie and in-gel TAMRA fluorescence (input). Clicked samples were incubated with NeutrAvidin agarose beads (output) to evaluate global enrichment of proteins (E) and specific enrichment of the known TDM receptor Mincle (F). Representative data shown for three independent biological replicates (E and F). Additional controls for (E and F) in SI Appendix, Fig. S5D.

Fig. 3.

Macrophage SNAREs interact with O-x-AlkTDM. (A) iBMDMs were incubated ±O-x-alkTDM for 6 h; exposed to UV or not; lysed; and subjected to CuAAC with AzTB as in Figs. 1B and 2 E and F. O-x-AlkTDM-interacting proteins were then enriched by NeutrAvidin beads; digested with trypsin; and analyzed by LC–MS/MS. ORA performed via WEBGestalt (38). Cut-off set at P < 0.05 and false discovery rate (FDR) < 0.25. Enrichment is relative to no probe samples across four independent biological replicates for each sample. iBMDMs (B) or THP-1s (C) treated 6 h ±O-x-AlkTDM and processed as in (A) and Fig. 2 E and F. Interacting proteins before (input) and after affinity enrichment (output) with NeutrAvidin agarose beads detected by in-gel fluorescence (TAMRA) or by immunoblotting with the indicated anti-SNARE antibodies. SNAREs identified by proteomics (A) highlighted in blue. Other SNAREs and housekeeping protein Glyceraldehyde-3-phosphate Dehydrogenase (GAPDH) in black. In (B), iBMDMs were also treated with control x-Alk-probes that lack either a lipid chain (TMM) or trehalose (FA; for structures, see Fig. 1C). Double bands for STX8 and VTI1B blots may be SNAREs interacting with one or two molecules of probe (SI Appendix, Fig. S7). Representative data shown for 2 to 5 independent biological replicates. (D) iBMDMs treated as in (B) but immunoprecipitated using the indicated anti-SNARE antibody. The presence of O-x-AlkTDM was confirmed by in-gel fluorescence (TAMRA) or immunoblotting with anti-biotin. There were no obvious changes in protein expression in samples prior to immunoprecipitation (input).

Fig. 4.

Noncanonical SNARE complexation is triggered by M. tuberculosis or native TDM. (A) iBMDMs were infected ±M. tuberculosis mc²6206 (Mtb; estimated MOI ~5) then immunoprecipitated and immunoblotted with indicated anti-SNARE antibodies. The VTI1B–STX8 interaction is M. tuberculosis–independent, whereas VAMP2 and VAMP8 interactions with VTI1B-STX8 occur preferentially when M. tuberculosis is present or absent, respectively. (B) SNARE expression (10 µg/lane) ±M. tuberculosis. iBMDMs infected as in (A) and lysates were immunoblotted with indicated anti-SNARE antibodies. (C–G) iBMDMs were incubated ±latex beads coated with BSA, IgG, TDB, native TDM, or O-x-AlkTDM then processed as in (A). Native TDM is sufficient to trigger VAMP2 interaction with VTI1B and STX8 (C and D), although a concomitant change in VAMP8 interactions with these proteins was not detected (F and G) in contrast to M. tuberculosis (A). (D and G) Quantitation of immunoprecipitations shown in (C) and (G) (and replicates thereof shown in SI Appendix, Fig. S8A), respectively. Mock, isotype control for each primary antibody. UI, uninfected (A and B) or unincubated (C–G). Representative data shown for 2 to 3 independent biological replicates. (E) SNARE expression (10 µg/lane) ±latex beads. iBMDMs lysates were immunoblotted with indicated anti-SNARE antibodies prior to immunoprecipitations in (C) and (F).

Fig. 5.

VAMP2 inhibits phagosome acidification and promotes M. tuberculosis survival during macrophage infection. WT and VAMP2 KO THP-1s were infected +/- M. tuberculosis mc²6206 (estimated MOI ~5) in (A–C) or fully virulent parent strain H37Rv (estimated MOI ~1 or 3) in (D–F). Lysates were immunoblotted with anti-SNAREs after (A, D) or before (C, F) immunoprecipitation. (B) and (D) show quantitation of immunoprecipitations from (A) and (E), respectively, as well as additional replicates from SI Appendix, Fig. S8. VAMP8 complexation with VTI1B and STX8 in the presence of high MOI M. tuberculosis is enhanced in the absence of VAMP2 for both strains. Mock, isotype control for each primary antibody. UI, uninfected. (G) iBMDMs expressing eGFP-VAMP2 were infected +/- mCherry-expressing M. tuberculosis (estimated MOI ~1). Additional images available in SI Appendix, Fig. S10. (H) Loss of macrophage VAMP2 or remodeled M. tuberculosis mycolates are nonadditively associated with enhanced phagosome acidification. THP-1s were infected with mEmerald-expressing M. tuberculosis that were untreated (Ctrl) or pretreated with Dimethyl sulfoxide (DMSO) carrier or DMSO plus O-TMM-C16 for ~5 doublings prior to infection (growth curves of treated M. tuberculosis in SI Appendix, Fig. S12). Representative images show spatial coincidence between M. tuberculosis and LysoTracker (denoted with white arrow) in THP-1 cells. Blinded quantification, Right, from 3 to 6 independent biological replicates each with n = 13 to 26 individual bacteria. Statistical significance was determined using unpaired Student’s t test. Additional images available in SI Appendix, Fig. S12. (I and J) VAMP2 promotes intracellular M. tuberculosis survival. iBMDMs expressing eGFP-VAMP2 or not (I) and WT and VAMP2 KO THP-1s (J) were infected ±M. tuberculosis mc²6206 (estimated MOI ~1). Macrophages were lysed and plated for colony-forming units at the indicated time points after infection. Representative data shown for 2 to 3 independent biological replicates. Statistical significance was determined using a two-way ANOVA with a Šídák’s multiple comparisons test for post hoc analysis. (K) Proposed model for TDM–SNARE interactions. TDM released from the mycomembrane of M. tuberculosis (Fig. 1A) intercalates into macrophage membranes, including those of the phagosome and other organelles and vesicles. TDM directly or indirectly promotes noncanonical, VAMP2-containing SNARE complexation at the expense of canonical, VAMP8-containing SNARE complexation. Fusion of lysosome with the M. tuberculosis–containing phagosome is suppressed as in Model 1 or Model 2 (details in text), enabling the pathogen to evade destruction by naive macrophages.