ESX-1 dependent impairment of autophagic flux by Mycobacterium tuberculosis in human dendritic cells

Mtb and BCG have different impacts on the autophagic process in human primary DC

To set up our experimental model, we faced an issue related to the use of IL4 and granulocyte-macrophage colony stimulating factor (CSF2) for the in vitro differentiation of monocyte-derived DC. Given the inhibitory effect exerted by IL4 on autophagy²⁵ and the activation of MTOR pathway by CSF2 and IL4 along monocyte-derived DC,²⁶ we compared in preliminary experiments the effect of Mtb infection on the autophagic process in the presence of CSF2 and IL4 or following the replacement of the differentiating medium 20 h before infection. Having found that both the basal and Mtb-driven autophagic response was altered by the presence of IL4 and CSF2 (data not shown), we decided to perform our study in DC starved from these differentiating cytokines.

Autophagy levels were analyzed in primary human DC infected with either Mtb or BCG at a multiplicity of infection (MOI) of 1. Cells were harvested 24 h after infection and the occurrence of autophagy was monitored by both immunoblotting and immunostaining using an antimicrotubule-associated protein 1 light chain 3 (MAP1LC3) antibody. As shown in Figure 1A, Mtb induced a significant increase of MAP1LC3-II, the processed isoform of MAP1LC3 associated with autophagosomes. This increase was accompanied by an accumulation of MAP1LC3-positive vesicles observed by confocal microscopy (Fig. 1B). Notably, MAP1LC3-II increase was not observed when cells were infected with either BCG or stimulated with heat-inactivated Mtb, suggesting that only live, virulent mycobacteria are able to modulate MAP1LC3-II levels (Fig. 1A and B). Autophagosome accumulation observed during Mtb infection was prevented by incubating DC with the autophagy inhibitor wortmannin, confirming the specificity of MAP1LC3-positive vescicles observed in Mtb-infected DC (Fig. S1).

We then asked whether the accumulation of MAP1LC3-II was an early or late event during Mtb infection. MAP1LC3 levels were analyzed in DC harvested by immunoblotting (Fig. 1C) and confocal microscopy (Fig. 1D) at different time points after Mtb infection. MAP1LC3 II increase occurs at late times, being slightly induced at 8 h and increasing at 16 h, although Mtb was already detected intracellularly at earlier times (Fig. 1E), as revealed by confocal microscopy performed with a Cherry-expressing Mtb following the visualization of DC plasma membrane using a FITC-conjugated DR antibody.

Mtb causes a block of autophagosome maturation

The accumulation of MAP1LC3-II in Mtb-infected cells could be due to an enhanced autophagosome formation or a reduced degradation of autophagic material.^17,25 To discriminate between these events, we assessed the occurrence of a functional autophagic flux in Mtb-infected cells. To this aim, we analyzed the lysosome-mediated autophagosome degradation by monitoring either the effect of lysosomal inhibitors on MAP1LC3 levels or the colocalization of autophagosomes with lysosomal markers. Both analyses indicated that autophagosome accumulation in Mtb-infected cells was due to a defect in the late steps of the autophagic process. As shown in Figure 2A, treatment of Mtb-infected DC with the lysosomal protease inhibitors E64d and Pepstatin A (PepA) resulted in a reduced increase of MAP1LC3-II protein levels [1.3 fold, calculated as the ratio between MAP1LC3 levels in Mtb-infected DC treated with E64D+PepA (6.4) and those in untreated Mtb-infected DC (5.1)] as compared with the effect promoted by the same treatment in not-infected cells (3.8 fold), suggesting that MAP1LC3-II degradation by lysosomal enzymes is impaired during Mtb infection. Similar results were obtained when the levels of SQSTM1 protein, a well-characterized target of the autophagic process, were analyzed in the same experimental setting. Indeed, Mtb infection leads to the accumulation of SQSTM1 protein (5.7 fold) which is not further increased by treatment with lysosomal inhibitors (1.1 fold), different from what was observed in control cells (2.4 fold, Fig. 2A). It should be pointed out that SQSTM1 increase in Mtb-infected DC is not only due to an impairment of protein degradation, since an induction at the transcriptional level was also observed during infection (2.5 fold induction, Fig. S2A). Similar results were obtained when lysosome activity in Mtb-infected DC was interfered using bafilomycin A₁, an inhibitor of the vacuolar-type H⁺-ATPase (1.2 fold in Mtb-infected DC vs. 3.9 fold in not-infected cells, Fig. S2B). The impairment of autophagosome maturation was further investigated by confocal microscopy. We observed a reduced colocalization between MAP1LC3 and the lysosomal markers lysosomal-associated membrane protein 1 (LAMP1) and cathepsin B (CTSB) in Mtb-infected cells (Fig. 2B–D), confirming that Mtb hampers the autophagic flux.

Rapamycin has been reported to promote autophagy-mediated clearance of Mtb in macrophages.¹² Thus, we decided to test whether rapamycin is able to overcome the impaired autophagic flux in Mtb-infected DC. Rapamycin was added 4 h after infection when the majority of cells showed the internalization of Mtb (see Fig. 1E). Western blot and confocal analyses showed that rapamycin treatment restores a functional autophagic flux as indicated by the reduction of SQSTM1 levels (Fig. 2A) as well as by the increased colocalization of MAP1LC3 with LAMP1 and CTSB, two markers of late endocytic/lysosomal compartments (Fig. 2B–D). Notably, rapamycin treatment led to a significant increase in Mtb-containing autophagosomes and autolysosomes, as evaluated by both electron and confocal microscopy (Fig. 3). Interestingly, confocal microscopy analysis showed not only an increased Mtb-MAP1LC3 colocalization following rapamycin treatment (Fig. 3B and E), but also a robust formation of Mtb-containing autolysosomes positive to LAMP1 (Fig. 3C and E). Mtb-MAP1LC3 colocalization induced by rapamycin is prevented by pretreating cells with wortmannin (Fig. S2C). To further characterize the intracellular Mtb localization after rapamycin treatment, the acidification of Mtb-containing phagosome was evaluated using LysoTracker, a fluorescent dye for labeling acidic organelles (Fig. 3D and E). Mtb was found within LysoTracker positive lysosome only after rapamycin treatment. Interestingly, the ability to rescue the inhibition of autophagic flux imposed by Mtb is not restricted to rapamycin, as it was also observed by treating cells with IFNG (Fig. S3), a known inducer of autophagosome formation.¹² Indeed, in IFNG stimulated DC Mtb localized with MAP1LC3 in LAMP1 positive compartment, likely autolysosome. Altogether these results indicate that, although a high number of autophagosomes accumulate in Mtb-infected DC, their maturation into the autolysosome is reached only after the treatment with autophagy inducers, such as rapamycin and IFNG.

ESX-1 is responsible for the inhibition of autophagosome maturation by Mtb

The different modulation of autophagy by Mtb and BCG prompted us to better characterize the mycobacterial factors responsible for this phenomenon. DC were infected with recombinant Mtb strains, which were deleted for either the entire RD1 region (MtbΔRD1::pYUB412),²⁷ or the promoter region of the CFP-10/ESAT-6 operon, preventing ESAT-6 and CFP-10 expression. Furthermore we tested a recombinant Mtb strain secreting an ESAT-6 variant lacking 12 aa of C-terminus (ESAT-6Δ84–95). These recombinant strains resulted attenuated in the TB mouse infection model.^28,29 Notably, all tested deletion mutants have lost the ability to accumulate MAP1LC3-II as well as to prevent MAP1LC3/LAMP1 colocalization (Fig. 4A–C), thus indicating that secretion of biologically functional ESAT-6/CFP-10 is a requirement for the inhibition of autophagosome maturation by Mtb.

In order to better elucidate the functional relationship between the RD1 encoded ESX-1 secretion system and the autophagic process, we decided to test the modulation of autophagy in DC infected with BCG complemented with a 32 kb fragment from Mtb spanning the extended RD1 region (BCG::RD1),³⁰ which actively secretes ESAT-6 and CFP-10. Cells infected with the BCG::RD1 strain depicted an accumulation of MAP1LC3-II protein similar to what was observed for Mtb (Fig. 4D). Finally, we tested the attenuated Mtb H37Ra strain, which shows an impaired secretion of ESAT-6 due to a point mutation in the DNA-binding region of the regulator PhoP. We analyzed the effect on autophagy of Mtb H37Ra and recombinant H37Ra strain complemented with a wild-type copy of phoP (Rv0757), which has been described to restore the ability of the H37Ra (H37Ra::PhoP) to secrete ESAT-6 and to elicit specific T-cell responses,³¹ in comparison with a strain complemented with an unrelated gene (H37Ra::RpsL). Strikingly, infection of DC with H37Ra::PhoP promoted MAP1LC3-II accumulation (Fig. 4D), whereas this effect was not observed with the Mtb H37Ra and H37Ra::RpsL strains. Taken together, these findings indicate that mycobacterial inhibition of autophagy in DC is dependent on the ability of mycobacteria strains to secrete intact ESAT-6/CFP-10.

Rapamycin enhances the capacity of Mtb-infected DC to promote a Th1-oriented response

Having found that rapamycin treatment overcomes the block in autophagosome maturation by Mtb, we sought to investigate the impact of this reversion on DC function. Initially, we evaluated whether rapamycin could modulate Mtb-induced DC maturation. Previous findings showed that rapamycin suppressed immunostimulatory molecules and the allostimulatory potential of LPS-stimulated DC.²⁶ However, when we investigated by flow cytometry analysis the immunophenotype of DC, no major differences were observed on the Mtb-induced expression of the molecules involved in T cell activation, such as CD86, CD83, HLA-DR and CD38 (Table 1). Also, no modulation of Mtb-induced tumor necrosis factor (TNF) production was observed upon rapamycin treatment (Fig. 5A). Conversely, the expression of cytokines involved in the polarization of Th1 lymphocytes was enhanced in Mtb-infected DC cultures upon rapamycin addition (Fig. 5B and C), namely interleukin (IL)12 p70 and interferon (IFN)B1. In particular, a significant induction of Mtb-stimulated production of interleukin (IL)12 p70 was observed following rapamycin treatment, (Fig. 5B). A similar result was obtained by studying interferon (IFN)B1 expression performed by real-time RT-PCR (Fig. 5C). Indeed, the induction of IFNB1 mRNA observed in Mtb-infected DC was further enhanced after exposure to rapamycin, although this increase was not statistically significant. Interestingly, also the treatment with another autophagy inducer, IFNG, enhances the IL12 production in Mtb-infected cultures (Fig. S4A), suggesting that the expression of this Th1 promoting cytokine could be linked to the autophagic process. In line with this result, E64D and PepA treatment abolishes Mtb-induced IL12 expression, while exerts a limited effect on TNF release (Fig. S4B).

To evaluate whether the effect induced by rapamycin on the secretion of Th1 cytokines from Mtb-stimulated DC resulted in an enhanced capacity to expand a Th1-oriented CD4⁺ T cell population, we studied T-cell polarization by mixed lymphocyte reaction (MLR) experiments. As shown in Figure 5D, the rapamycin treatment enhanced the production of IFNG. IL12 neutralization with specific monoclonal antibodies drastically reduced IFNG levels in co-cultures with DC infected with Mtb both in presence or absence of rapamycin, suggesting that the enhanced expression of IL12 induced by rapamycin may represent a key event for the expansion of an anti-Mtb Th1 response mediated by DC. No induction of IL4 and IL5 was observed when T cells were co-cultured for 5 d with Mtb-stimulated DC in the presence or absence of rapamycin (data not shown).

Finally, we evaluated the autophagy-mediated killing of the tubercle bacilli by CFU assay (Fig. S5). A bacteriostatic effect on mycobacterial growth rather than a reduction in Mtb viability was observed upon rapamycin treatment at 3 d, pointing at a major role of autophagy in DC in the induction of an efficient T cell response rather than Mtb killing, a function mainly exerted by macrophages.¹³

Collectively, these results provide evidence of the capacity of rapamycin to promote Mtb-infected DC to stimulate a Th1 response likely by overcoming the mycobacterial phagosome maturation block and enhancing IL12 production.

Antibodies and other reagents

Monoclonal antibodies (BD Bioscience) specific for CD1a (555806), CD14 (555398), CD38 (555460), CD86 (555657), HLA-DR (555811), CD83 (556855), IgG1 (555748), IgG2a (553457) were used as direct conjugates to FITC (Fluorescein Isothiocyanate) or PE (Phycoerythrin). Rabbit anti-MAP1LC3 for immunoblotting analysis (Cell Signaling, 2775), rabbit anti-MAP1LC3 for confocal microscopy (Sigma-Aldrich, L7543), mouse anti-LAMP1 (Abcam, ab25630), mouse anti-actin (Sigma-Aldrich, A2066), mouse anti-SQSTM1 (SQSTM1 D3, Santa Cruz Biotechnology, sc28359), mouse anti-cathepsin B (Calbiochem, IM27L). For autophagy induction, rapamycin (0.2 μM; Sigma-Aldrich, R0395) or with 10 ng/ml of IFNG (Pepro Tech EC LTD, 300-02) was added to DC cultures. To inhibit lysosomal activity, cells were incubated with 10 μg/ml E64d (Sigma-Aldrich, E8460) together with 10 μg/ml PepA (Sigma-Aldrich, P5318), bafilomycin A₁ (Sigma-Aldrich, B1793), while wortmannin 1 μM (Sigma-Aldrich, W1628) was used as a pharmacological inhibitor of autophagy. The optimal concentration of rapamycin and E64d plus PepA was evaluated in preliminary dose-response experiments, while the treatment time is specified in each figure legend.

DC preparation

DC were prepared as previously described.⁴⁹ Briefly, peripheral blood mononuclear cells were isolated from freshly collected buffy coats obtained from healthy voluntary blood donors (Blood Bank of University “La Sapienza,” Rome, Italy) by density gradient centrifugation using Lympholyte-H (Cedarlane, CL 5020). Monocytes were purified by positive sorting using anti-CD14 conjugated magnetic microbeads (Miltenyi, 130-050-201). The recovered cells were > 99% CD14⁺ as determined by flow cytometry with anti-CD14 antibody. DC were generated by culturing monocytes in 6-well tissue culture plates (Costar Corporation, 3516) with 50 ng/ml CSF2 (R&D Systems, 215-GM) and 1000 U/ml of IL4 (R&D Systems, 204-IL) for 5 d at 0.5 × 10⁶ cells/ml in RPMI 1640 (BioWhittaker, BE12-167F) supplemented with 2 mM L-glutamine (BioWhittaker, BE17-605E) and 15% FCS (BioWhittaker, DE14-801F). No antibiotics were added to the cultures. At day 5 the cells were 90% CD1a⁺ and 95% CD14^-. The differentiation medium containing IL4 and CSF2 was removed and replaced with RPMI supplemented with 2 mM L-glutamine and 15% FCS before Mtb infection to avoid cytokine-driven modulation of the autophagic and MTOR pathway.

Bacteria preparation and infection of DC

Mtb H37Rv (ATCC 27294; American Type Culture Collection) and M. bovis BCG (ATCC 27291) were used to infect DC cultures. As regards the recombinant strains used in this study, they were derived from Mtb H37RvΔRD1²⁷ or BCG Pasteur (1173P2) by integrating either the empty vector pYUB412, or pRD1-2F9 that carries a 32 kb fragment from Mtb H37Rv overlapping the extended RD1 region, or modified versions of pRD1-2F9 deleted for the promoter region of the CFP-10/ESAT-6 operon (ΔCFP-10prom) or the 11 C-terminal codons of ESAT-6 (ESAT-6-Δ84-95).^28,29 Mtb H37Ra, the recombinant strain H37Ra::PhoP containing the wild-type PhoP gene and the strain complemented with an unrelated gene (H37Ra::RpsL) were previously described in Frigui et al.³¹ The plasmid pSMT3-DsRed (kindly by Wilbert Bitter, VU University, Amsterdam), and containing a cassette with the gene encoding the red fluorescent protein cherry under the control of the hsp60 promoter, was electroporated in Mtb H37Rv following standard procedures.⁵⁰ The cloning of the recombinant Mtb H37Rv GFP was previously described by Delogu and collaborators.⁵¹ Recombinant Mtb H37Rv Ds-Red cherry as well as Mtb H37Rv GFP were cultured on 7H11 agar (BD Bioscience, 283810) supplemented with OADC (BD Bioscience, 211886) and with 50 μg/ml of hygromycin (Sigma-Aldrich, H9773) or in 7H9 liquid medium (BD Bioscience, 271310) supplemented with ADC (BD Bioscience, 211887), glycerol (Sigma-Aldrich, G5516) (0.5% vol/vol), 0.05% Tween 80 (Sigma-Aldrich, P1754), and containing 50 μg/ml of hygromycin until late log phase. One-ml of culture aliquots containing 10% of glycerol were stored at -80°C until use.

Mycobacteria were grown with gentle agitation (80 rpm) in 7H9 liquid medium supplemented with 0.05% Tween 80 and 10% OADC. Logarithmically growing cultures were centrifuged at 800 rpm for 10 min to eliminate clumped mycobacteria and then washed three times in RPMI 1640. Mycobacteria were resuspended in RPMI 1640 containing 10% FCS and then stored at -80°C. Vials were thawed and bacterial viability was determined by counting the number of colony forming units (CFU) on Middlebrook 7H10 agar plates (BD Bioscience, 262710). All bacteria preparations were analyzed for LPS contamination by the Limulus lysate assay (BioWhittaker, 50-647U) and contained less than 1 EU/ml. DC cultures were infected with a MOI of 1 bacteria/cell. Where indicated, Mtb was heat killed at 80°C for 1 h.

CFU assay

Four h after infection, the cell cultures were gently washed (three times) with RPMI 1640. DC were centrifuged at 150 × g for 10 min to selectively spin down cells while extracellular bacteria remain in the supernatants. Cells (~300.000) were resuspended in complete medium and cultured for the times indicated. Triplicate samples were assayed for CFU. At each time point cells were lysed with water containing 0.05% saponin (Sigma-Aldrich, 84510). Serial dilutions of the bacterial suspensions were plated (in duplicate) on 7H10 agar plates and viable bacteria were evaluated after 3 weeks of incubation at 37°C. Values are expressed as CFU/well.

Flow cytometry analysis

Approximately 1–2 × 10⁵ cells were aliquoted into tubes and washed once in PBS containing 2% FCS. The cells were incubated with indicated antibodies (BD Bioscience) at 4°C for 30 min. The DC were then washed and fixed overnight with 4% paraformaldehyde (Pro-Labo, 20909290) before analysis on a FACSCan using CellQuest software (BD Bioscience).

Cytokine determination

Supernatants from DC cultures infected with Mtb and treated with rapamycin (0.2 μM) or E64d/PepA (10 μg/ml each) were harvested at 24 h after infection, filtered (0.2 μm) and stored at -80°C. The production of TNF (A) and IL12 (B) was measured by CBA Flex Set (BD Bioscience, 558472 for TNF, 558459 for IL12p70). The production of cytokines in MLR experiments was measured with human Th1/Th2 CBA Flex Set (BD Bioscience).

RNA isolation and real-time PCR quantifications

RNA was extracted from DC with RNeasy kit (Qiagen Inc., 74104) according to the manufacturer’s instructions. A phenol/chloroform extraction was performed to inactivate residual mycobacterial particles. Reverse transcriptions and quantitative PCR assays were performed as previously described.⁴⁹ Primer pairs have been previously described.⁴⁹ The primers used for IFNB1 and GAPDH were previously described³³ while the oligonucleotides for SQSTM1 mRNA quantification were: 5′-ACAGATGCCAGAATCCGAAG-3′ (forward) and 5′-TGGGAGAGGGACTCAATCAG -3′ (reverse). Levels of IFNB1 and SQSTM1 mRNA are normalized to the GAPDH level using the Equation 2^-ΔCt; the values are means ± SD of triplicate determinations

Immunoblotting analysis

Proteins were separated on 12% NuPAGE Bis-Tris gel (Invitrogen, NP0341) and electroblotted onto polyvinylidene fluoride membranes (Millipore, IPVH20200). Blots were incubated with primary antibodies in 5% nonfat dry milk in TBS plus 0,1% Tween20 (Sigma-Aldrich, P1379) overnight at 4°C. Detection was achieved using horseradish peroxidase-conjugate secondary antibody anti-mouse (Jackson Immunoresearch, 711-035-152), anti-rabbit (Jackson Immunoresearch, 715-035-150) and anti-goat (Jackson Immunoresearch, 705-035-147) and visualized with Enhanced Chemiluminescence plus (GE Healthcare, RPN2132). MAP1LC3-II/actin and SQSTM1/actin ratios were quantified by densitometric analysis using the ImageQuant software according to Rubinsztein et al.⁵²

Confocal microscopy

Cells were fixed with 4% paraformaldehyde (Sigma-Aldrich, P6148) in PBS followed by permeabilization with 0.1% triton X-100 (Sigma-Aldrich, 93443) in PBS. Primary antibodies were incubated for 1 h at room temperature and visualized by means of Cy2-conjugated secondary antibodies (Jackson Immunoresearch, 711-226-152) and Cy3-conjugated secondary antibodies (Jackson Immunoresearch, 715-166-150). Coverslips were mounted in SlowFade-Anti-Fade (Invitrogen, S36939) and examined under a confocal microscope (Leica TCS SP2). For lysosome staining 500 nM Lysotracker Red (Invitrogen, L7528) was added to cells 30 min before fixation. Digital images were acquired with Leica software, and processed and quantified with ImageJ (National Institutes of Health) using a set of defined intensity thresholds that were applied to all images. A minimum of 50 cells per sample was counted for triplicate samples per condition in each experiment.

Transmission electron microscopy

DC cultures were analyzed 24 h after infection. Cells cultured in different conditions were fixed with 2.5% glutaraldehyde (Assin Spa, R1012) in 0.1 M cacodylate buffer for 1 h at 4°C (sodium cacodylate trihydrate, Sigma-Aldrich, C4945), and postfixed in 1% osmium tetroxide (Sigma-Aldrich, 75632) in 0.1 M cacodylate buffer for 1 h. The cells were then dehydrated in graded ethanol and embedded in Epon resin (AGAR 100, Agar Scientific R1045). Ultrathin sections were stained with 2% uranyl acetate (Sigma-Aldrich, 73943) and observed under a Zeiss EM900 transmission electron microscope. Images were captured digitally with a Mega View II digital camera (SIS; Zeiss).

MLR experiment

MLR was conducted as previously described.⁴⁹ CD4⁺ T cells were purified from cord blood cells obtained from healthy voluntary donors (Cord Blood Bank of UOC Transfusion Medicine, S. Eugenio Hospital-ASL Roma C, Rome, Italy) by indirect magnetic sorting with a CD4⁺ T-cell isolation kit (Miltenyi, 130-091-155). DC were infected with Mtb and 4 h after the infection rapamycin was added for additional 20 h. Immature DC, Mtb-infected DC with or without a rapamycin treatment were resuspended in 5% human AB serum (Euroclone, ECS0219D) complete medium prior to co-culture with naïve allogeneic cord blood CD4⁺ T cells. Where indicated, neutralizing mouse monoclonal anti-IL12 antibody (R&D Systems, AB-219-NA) and its isotype control antibody (R&D Systems, MAB002) were added to DC/CD4⁺ T cell co-culture at 20 μg/ml to neutralize the IL12. Supernatants from T cell/DC co-culture (ratio 10:1) were harvested at day 5 and analyzed for IFNG (560111), IL4 (558462) and IL5 (558463) release by CBA Flex set (BD Bioscience).

Ethics statement

Human blood was collected at the Blood Bank of University “La Sapienza” (Rome, Italy) and the Cord Blood Bank of UOC Transfusion Medicine, S. Eugenio Hospital (Rome, Italy) from healthy volunteers after written informed consent. Both blood centers operate under license from the Italian Ministry of Health. Therefore, the use of blood from these sources is exempt from our institutional review board.

Statistical analysis

Statistical analysis was calculated using a two-tailed for paired-data Student’s t-test. A p value < 0.05 was considered statistically significant.