Abstract
Mycobacterium tuberculosis infection is characterized by a strong inflammatory response whereby a few infected macrophages within the granuloma induce sustained cellular accumulation. The mechanisms coordinating this response are poorly characterized. We hypothesized that microparticles (MPs), which are submicron, plasma membrane-derived vesicles released by cells under both physiological and pathological conditions, are involved in this process. Aerosol infection of mice with M. tuberculosis increased CD45+ MPs in the blood after 4 wk of infection, and in vitro infection of human and murine macrophages with mycobacteria enhanced MP release. MPs derived from mycobacteria-infected macrophages were proinflammatory, and when injected into uninfected mice they induced significant neutrophil, macrophage, and dendritic cell recruitment to the injection site. When incubated with naive macrophages, these MPs enhanced proinflammatory cytokine and chemokine release, and they aided in the disruption of the integrity of a respiratory epithelial cell monolayer, providing a mechanism for the egress of cells to the site of M. tuberculosis infection in the lung. In addition, MPs colocalized with the endocytic recycling marker Rab11a within macrophages, and this association increased when the MPs were isolated from mycobacteria-infected cells. M. tuberculosis–derived MPs also carried mycobacterial Ag and were able to activate M. tuberculosis–specific CD4+ T cells in vivo and in vitro in a dendritic cell–dependent manner. Collectively, these data identify an unrecognized role for MPs in host response against M. tuberculosis by promoting inflammation, intercellular communication, and cell migration.
Introduction
Granuloma formation is a central component of the inflammatory response to Mycobacterium tuberculosis infection. Newly recruited leukocytes aggregate around infected macrophages, providing the microarchitecture of the granuloma necessary to control infection. The recruitment and retention of cells within the granuloma is regulated by the release of proinflammatory factors from infected cells; these include soluble cytokines and chemokines and may also include microvesicles, such as exosomes and microparticles.
Previous research has demonstrated that exosomes, which are derived from intraluminal vesicles and are less than 100 nm in size, carry Ag and have proinflammatory potential (1). Exosomes released by M. tuberculosis–infected macrophages contain MHC and can transfer mycobacterial Ags between cells (1–3). Microparticles (MPs) are distinct from exosomes, are larger in size (100–1000 nm), and arise from enzyme-regulated budding of the plasma membrane (4). MPs are released during normal physiologic conditions from multiple cell types. Elevated MP release has been observed after stimulation from infectious mediators, pathogens, mitogens, and stress conditions (5–7). Although the function of MPs during mycobacterial infection is unknown, MPs provide a means of communication between cells, transferring cargo from donor to recipient cells. This includes the transfer of surface receptors, integrins, and ligands, which reflect those of their cell of origin and their cytoplasmic contents, including biologically active cytokines, chemokines, mRNA, and microRNA (6, 8–11).
The composition of MPs released under varying conditions influences their capacity to modulate inflammation and contribute to both protective immunity and pathology following infection. Increased MPs in the serum has been associated with a number of systemic infections (12–14). Activated protein C treatment is known to reduce mortality during sepsis, having both anti-inflammatory and anti-apoptotic effects. A recent study demonstrated that treatment of sepsis with activated protein C drives the release of MPs, and these MPs themselves contain activated protein C (15). The authors suggest that these MPs may act to modulate inflammatory activity in recipient cells, possibly by disseminating activated protein C to distal sights. Plasma MP levels were also shown to be markedly increased in cerebral malaria (16, 17). In mice, MPs produced during cerebral malaria display procoagulant and proinflammatory properties, and inhibiting MP production protected mice against cerebral malaria (18). Furthermore, MPs released by activated neutrophils also have documented anti-inflammatory effects. When these MPs were added to human macrophages and dendritic cells (DCs) stimulated with TLR-2 and TLR-4 ligands, they inhibited the release of TNF and induced the expression of TGF-β (19, 20).
Several lines of study have also shown that MPs, like exosomes, can transfer Ag between cells. Apoptotic vesicles, likely to be a combination of apoptotic bodies and MPs, released from mycobacteria-infected macrophages, carried mycobacterial Ags that could stimulate CD8+ T cells in vivo, and importantly, provided protection against M. tuberculosis challenge (21). Mycobacteria themselves can release microvesicles, and these may contribute to mycobacterial virulence by accelerating local inflammatory responses. Indeed, mice pretreated with mycobacterial vesicles before M. tuberculosis infection demonstrated increased bacterial replication in vivo (22).
To study the influence of MPs during mycobacterial infection, we measured MP production during mycobacterial infection of human and murine macrophages and M. tuberculosis infection in mice. Mycobacterial infection upregulated MP release in vitro and in vivo. When taken up by uninfected recipient cells in vitro, these MPs promoted cell migration by inducing the release of proinflammatory cytokines and chemokines and by disrupting epithelial cell monolayers. In vivo, the injection of MPs released from M. tuberculosis–infected macrophages induced recruitment of neutrophils, macrophages, and DCs to the injection site. These MPs contained mycobacterial Ag, which stimulated Ag-specific CD4+ T cell proliferation in vivo and in vitro in a DC-dependent process. These data indicate a prominent role for MPs in amplifying the inflammatory and immune response to mycobacterial infection and may provide an important means of cellular communication within the granuloma.
Materials and Methods
Ethics statement
All mouse procedures performed in this study were conducted at the Centenary Institute, after protocol review and approval by the University of Sydney Animal Care and Ethics Committee, protocol number K75/11-2010/3/5436. All experiments complied with the National Health and Medical Research Council code of conduct for the use of animals for scientific purposes.
Cell culture
Human myelo-monocytic THP-1 and murine IC21 (H-2b) cells were grown in RPMI 1640 (Life Technologies) media supplemented with 2 mM l-glutamine, 25 mM HEPES, and 0.2 μm twice-filtered 10% FCS. Murine RAW 264.7 cells (H-2d) and airway epithelial C10 (H-2d) were grown in DMEM supplemented with 10% FCS. THP-1 cells were differentiated into monocytes by the addition of 100 nM PMA (Sigma-Aldrich) for 48 h, after which it was removed.
Mice
C57BL/6 mice were obtained from Animal Resources Centre (Perth, Australia), and p25 CD4 TCR transgenic mice (specific for residues 240-254 of M. tuberculosis Ag 85B) (23) were provided by Prof. K. Takatsu (Nihon University, Tokyo, Japan) and Dr. J. Ernst (New York University, New York, NY) and bred at the Centenary Institute animal facility. CD11c-EYFP mice (24) were provided by Prof. W. Weninger (Centenary Institute).
Bacterial cultures
Mycobacterium bovis bacillus Calmette–Guérin (BCG)-Pasteur and H37Rv M. tuberculosis were cultured in Middlebrook 7H9 liquid medium supplemented with 10% oleic acid, albumin, dextrose, and catalase, 0.5% glycerol, and 0.02% tyloxapol (Sigma-Aldrich). Mid-logarithmically grown, single bacterial cell cultures were sonicated for 10 s, and concentration was determined by measuring OD at 600 nm. CFU determination occurred using Middlebrook 7H11 agar supplemented with 10% oleic acid, albumin, dextrose, and catalase and 0.5% glycerol.
Generation of human monocyte-derived macrophages
Buffy coats were obtained from the Australian Red Cross (Sydney, Australia) and PBMCs separated on Ficoll Paque Plus (GE Healthcare), and cultured as described previously (25). PBMCs were cultured at 500,000 cells/well in 24-well plates. IFN-γ (100 U/ml; Roche Diagnostics, Indianapolis, IN) was added to the indicated wells 16 h before infection.
Generation of murine bone marrow–derived macrophages
Bone marrow–derived macrophages (BMDMs) were isolated from C57BL/6 wild-type mice and culture in RPMI supplemented with GM-CSF (10 μg/ml) for 5 d, before further stimulation or infection.
Macrophage infections with mycobacteria for MP quantitation
Macrophages were infected with BCG or M. tuberculosis H37Rv for 4 h at a multiplicity of infection (MOI) of 4 and 0.5 respectively, and then washed twice to remove extracellular bacteria. At the indicated times, 1 ml media was removed, centrifuged at 524 × g for 10 min, 750 μl supernatant was removed, and 180 μl sample was added to 20 μl 10× annexin V binding buffer (0.1 M HEPES/NaOH, pH 7.4, 1.4 M NaCl, 25 mM CaCl2) and 0.4 μl annexin V–AF488 or AF647 (Molecular Probes). CD40 and MHC class II (MHC-II) Ab were added at a volume of 0.2 μl per sample and incubated for 30 min at room temperature. Next, samples were analyzed on a Beckman Coulter FC500-MPL flow analyzer at a predetermined flow rate for 60 s. Flow-count beads (Beckman Coulter) were used to determine the volume measured during that time. Positive events were determined by annexin V positivity along with size restriction using microspheres measuring 1, 0.75, 0.5, and 0.2 μm (Molecular Probes, Invitrogen). Absolute numbers of MPs were calculated using the following formula:
Large scale MP purification from mycobacteria infected macrophages
THP-1 (human), RAW, and IC-21 (murine) cell lines (2 × 107) were infected with BCG or M. tuberculosis at an MOI of 4 and 0.5, respectively. Extracellular bacteria were removed at 4 h, and the supernatant was collected at 48–72 h and centrifuged at 524 × g for 10 min. This supernatant was collected and centrifuged at 3273 × g for 30 min, before the supernatant was again collected and centrifuged at 27,000 × g for 2 h with no brake. The MP pellet was resuspended in RPMI, and the MP concentration was determined by flow cytometry (size and annexin V positivity) or by the BCA protein assay. The sterility of the MP preparation was confirmed by plating 1–2 μl of the resuspended pellet onto 7H11 agar and incubating for 21 d at 37°C. Purified MPs were CFSE labeled with 1 μM CFSE for 10 min at 37°C, washed three times by centrifugation at 18,000 × g in RMPI containing 10% FCS before quantitation and use. Vybrant Cell Labeling Solution DiD dye (V-22887; Molecular Probes)–labeled MPs were incubated at 37°C for 20 min and washed three times before quantification and use. Whenever possible, freshly isolated MPs were used for all experiments.
Scanning electron microscopy
Purified MPs in PBS were seeded on polylysine-coated glass coverslips at 4°C overnight. MPs were washed in 0.1 M sodium cacodylate/0.1 M sucrose buffer solution before fixation in sodium cacodylate–buffered 2% (v/v) glutaraldehyde (Sigma-Aldrich) for 30 min, followed by 1 h in 1% osmium tetroxide/0.1M sodium cacodylate. Dehydration was performed in grading alcohols with a final step in hexamethyldisilazane (H4875; Sigma-Aldrich) for 3 min. Coverslips were mounted onto scanning electron microscopy specimen stubs, lined with silver dag, and coated with 3.75 nm platinum. Samples were imaged with a Zeiss Ultra Plus Field Emission Scanning Electron Microscope, using an accelerating voltage of 10 kV and an Everhart-Thornley detector for secondary electron detection.
Cytometric bead array
IL-8, MIP-1α, MCP-1, and TNF were determined using a cytometric bead array (BD Bioscience) according to the manufacturer’s instructions, and the data were evaluated using BCA FCAP Array software.
Impedance measurement
The effect of MPs on epithelial integrity was determined using the Electrical Cell-Substrate Impedance System (ECIS) Model 1600R in an eight-well plate format (Applied BioPhysics). C10 cells were cultured in ECIS 8W10E plates and coated with a 50-mM solution of l-cysteine until cells were confluent and showed a plateau in their impedance. RAW cells were added at a ratio of 1:20 C10 to RAW cells. An MP concentration of 30 μg, as determined by the BCA assay, was used where indicated. The impedance was measured every 10 min for at least 72 h.
Murine intradermal delivery of MP, tissue processing, and flow cytometry
Intradermal injection of C57BL/6 mice ears were performed as described previously (26, 27). Mice were anesthetized by i.p. injection with ketamine/xylazine (80/10 mg/kg), and given 105 MPs isolated from uninfected cells (MP-UI), BCG-infected cells (MP-BCG), M. tuberculosis-infected cells (MP-Mtb) or 2.5 × 105 BCG in 4μL using 35-gauge Hamilton syringe. After 24 h the ears were processed as described (27
Analysis of MP quantities in the blood of mice
C57BL/6 mice were infected with ∼100 CFU of M. tuberculosis H37Rv by aerosol exposure using a Middlebrook airborne infection apparatus (Glas-Col, Terre Haute, IN) (28). Four weeks after infection, blood was collected by cheek puncture, place into trisodium citrate (1:9 vol/vol), and processed as previously described (29). MPs were stained for CD41 (MWReg30) and CD45.2 (BD Biosciences). Plasma MP quantification was performed using an LSRII flow analyzer for small particle detection (BD Biosciences).
Murine bone marrow neutrophil purification and confocal microscopy
Murine bone marrow–derived neutrophils were purified from 6–8-wk-old C57BL/6 mice as described previously (30). Neutrophils (2 × 107) were labeled with CellTrace Violet according to the manufacturer’s instructions (Molecular Probes, Invitrogen) and injected i.v. 24 h prior to imaging. CD11c-EYFP C57BL/6 mice were injected intradermally in the ear with 5 × 104 DiD-labeled MP-UI or MP-Mtb. After 24 h, mice were euthanized, and the ears were removed and fixed in formalin. Images were captured using a Leica SP5 confocal microscopy and ×20 objective.
Fixed cell imaging using confocal microscopy
Cells were fixed with 4% paraformaldehyde overnight at 4°C, washed with PBS +1% FCS, treated for 1 h with 0.25% TritonX-100 in PBS + 1% FCS, blocked with 5% FCS in PBS for 4 h, and incubated with anti-Rab5a Ab (sc309), anti-Rab7 Ab (sc10767), or anti-Rab11a (sc26590; Santa Cruz Biotechnology) for 1 h at 4°C. Cells were washed and incubated with either anti-goat or anti-rabbit secondary Abs (Invitrogen, Molecular Probes) for 45 min at 4°C and washed. Next, Hoechst 33342 trihydrochloride, trihydrate (1 μg/ml) was added for 10 min at room temperature, followed by four washes. Isotype controls without primary Ab were performed to ascertain nonspecific secondary Ab binding. Imaging was performed using a Leica SP5 confocal microscope and a ×63/N.A. 1.4 objective with individual Z-planes separated by 1 μm.
T cell proliferation assay
For the in vitro assays, DCs were generated from bone marrow as described previously (31). MPs purified from 3-d–non-infected, BCG, or M. tuberculosis–infected IC21 cells were added to the DCs at a ratio of 10:1. After 24 h, CD4+ T cells from the spleens of p25 transgenic mice were purified and CFSE labeled as described previously (31). CFSE-labeled p25 T cells (2 × 106) were added to the DC-MP cultures. The p25 peptide was synthesized by GenScript and was used at a concentration of 10 μg/ml. Seventy-two hours later, cells were stained for CD3, CD4, CD8, CD44, CD45.1, CD62L (BD Biosciences), and the Live/Dead Fixable Violet Dead cell stain. For the in vivo assays, 106 MPs or 5 × 104 M. tuberculosis cells were injected into the footpad of C57BL/6 mice, followed by i.v. injection of 2 × 106 CFSE-labeled splenic p25 T cells. After 5 d, mice were culled and the popliteal lymph node was removed, separated into single cells, and stained as described above. Samples were acquired using an LSRFortessa or LSR-II (BD Bioscience) flow analyzer.
Microscopic image analysis, computer software used, and statistical analysis
Improvision Volocity 5.5 software and Imaris 7.2 BitPlane Scientific Software were used to analyze image data. Microscopic images were treated equally in terms of their contrast and brightness settings for each experiment. Flow cytometric analysis was performed using FlowJo analysis software (Tree Star). Prism (GraphPad Software) was used for graphical representation of data and statistical analysis by Student t test with pairwise comparisons between samples sets. Differences with p < 0.05 were considered significant.
Results
Mycobacterial infection upregulates MP production by macrophages
We examined the effect of mycobacterial infection on macrophage MP production and compared these MPs to those released spontaneously by uninfected macrophages. In addition, because IFN-γ–activated macrophages are less permissive to intracellular mycobacterial growth, the role of IFN-γ in MP production during infection was also determined.
IC21 cells, a C57BL/6 macrophage line, were infected with BCG. After 48 h, the supernatant was removed, centrifuged, and stained with annexin V. Using flow cytometry and microsphere beads ranging from 0.2–1.0 μm, the forward scatter size gate was determined, and the MP population was identified and quantified by size and annexin V binding (Fig. 1A). Scanning electron microscopy demonstrated that the MPs were 0.1–1 μm in size, with spherical morphology (Fig. 1B). Particles of the size and morphology of exosomes or apoptotic bodies were not observed with the purified MPs, and no mycobacteria were detected by bacterial culture of the MP preparations used.
MPs derived from macrophages. (A) Flow cytometry gating strategy to identify MPs. Microspheres (0.2, 0.5, 0.75, and 1 μm) were analyzed using a Beckman Coulter FC500-MPL flow cytometer. A forward scatter gate ranging from 0.2 to 1.0 μm was applied to identify events based on size. Annexin V was used to identify MPs in the supernatants of IC21 macrophages 48 h after infection. (B) Scanning electron micrograph images of IC21-derived MPs or BCG-infected IC21-derived MPs. Left panels, Original magnification ×50,000 (scale bar, 1 μm); right panels, original magnification ×100,000 (scale bar, 200 nm). (C–H) MP concentrations per 5 × 105 cells are shown. (C) Annexin V+ MPs from human PBMCs with or without IFN-γ, infected with M. tuberculosis. (D) MHC-II+ MPs from human PBMCs infected with BCG. (E) Annexin V+ MPs from M. tuberculosis–infected BMDMs. (F) Murine IC21 macrophages with or without IFN-γ infected with M. tuberculosis. Annexin V+ (G) or CD40+ (H) MPs from BCG-infected IC21 macrophages. Each time point represents the means and SEM of triplicate cultures from one of two independent experiments except for the PBMCs where six individual donors were analyzed with a representative experiment displayed. *p ≤ 0.05 (Student t test) for comparisons with the untreated control group in each experiment.
Using the staining and gating strategy from Fig. 1A, different primary macrophages and cell lines were examined for MP production after infection with either M. tuberculosis or BCG. PBMC-derived macrophages and murine BMDMs infected with M. tuberculosis produced significantly more MPs than uninfected macrophages (Fig. 1C, 1E), as did murine IC21 macrophages infected with M. tuberculosis or M. bovis BCG (Fig. 1F, 1G). Regardless of the macrophage origin, we found a reproducible pattern of increased MP production following mycobacterial infection. This effect was more pronounced when cells were infected with virulent M. tuberculosis, despite an MOI 8-fold less than that used for the BCG infection. Treatment of macrophages with IFN-γ did not affect the level of MPs released from cells during infection (Fig. 1C, 1F). Human PBMCs and murine BMDMs were infected with BCG, and MHC-II or CD40 expression on the released MPs determined (Fig. 1D, 1H). MPs from both infected and uninfected macrophages expressed MHC-II and CD40.
MPs derived from mycobacteria-infected macrophages stimulate the release of proinflammatory cytokines and chemokines from uninfected macrophages
During mycobacterial infection IL-8 (32), TNF (33), and MCP-1 (34) are upregulated in the lung. To determine whether MPs released by infected macrophages can contribute to these responses, MPs from BCG-infected and uninfected human (THP-1) or murine (RAW cells) macrophages were added to recipient macrophages. IL-8 and MIP-1α levels increased when THP-1 cells were incubated with MP-BCG compared with MPs from uninfected cells or cells without MP stimulation (Fig. 2A, 2B). Similarly, RAW macrophages stimulated with MP-BCG produced significantly higher levels of TNF and MCP-1 compared with unstimulated cells alone or cells exposed to MP-UI (Fig. 2C, 2D).
Microparticles isolated from mycobacteria-infected macrophages stimulate cytokine and chemokine production in recipient macrophages. Listed macrophages were seeded at 5 × 105 cells/ml and were untreated or treated at a ratio of 1:1 with MP-UI or MP-BCG originating from the same cell type. Supernatants were collected at indicated time points, and bead arrays were performed to ascertain cytokine and chemokine levels. (A) IL-8, (B) MIP-1α from THP-1 cells, (C) TNF, and (D) MCP-1 from RAW cells. Data represent the means and SD of triplicate cultures. Each assay was performed independently twice with similar results. (E) MP-BCG incubated with recipient macrophages disrupts an epithelial cell monolayer. MPs (30 μg/well) derived from uninfected or BCG-infected RAW cells were coincubated with fresh RAW cells over a monolayer of murine epithelial C10 cells. MP-UI and MP-BCG were also added to C10 cells without the presence of RAW cells. The zero time point indicates the addition of MPs to the indicated wells. Three independent experiments were performed in a similar manner with a representative result displayed. *p < 0.05, **p < 0.01, ***p < 0.005.
Granuloma formation is the hallmark of tuberculosis and requires reorganization of the extracellular matrix to facilitate movement of the infiltrating cells. Therefore, the capacity of MPs to aid cell migration by disrupting epithelial monolayers was assessed. Mouse epithelial C10 cells were cultured on ECIS tissue culture slides until confluent, as demonstrated by stable electrical resistance. RAW cells and MP-UI or MP-BCG (purified from infected RAW cells) were added to the monolayer, and the epithelial integrity of the C10 cells was monitored by measuring changes in electrical resistance across the monolayer for 60 h (Fig. 2E). The addition of RAW cells or MPs alone had no effect on the electrical resistance of the monolayer, nor did the addition of MP-UI with RAW cells. However, the addition of MP-BCG with the RAW cells caused a continuous decrease in the resistance of the monolayer, resulting in its complete disruption after 60 h (Fig. 2E). This finding demonstrates that MPs released from mycobacteria-infected macrophages may act in concert with the uninfected macrophages to induce disruption of epithelial cell tight junctions, and this could aid migration of inflammatory cells across the respiratory epithelium.
MP-BCG or MP-Mtb induce immune cell recruitment in vivo
Having shown that MPs can aid disruption of epithelial monolayers in vitro, we proceeded to quantitate the early cellular response to MP-BCG or MP-Mtb injection in vivo. We injected DiD-labeled IC21 macrophage-derived MPs, MP-UI, MP-BCG, MP-Mtb, or BCG intradermally into the ears of C57BL/6 mice and analyzed the leukocyte populations in the skin 24 h later. To examine the effect of TNF in modulating cellular recruitment, TNF−/− mice were also injected with MP-Mtb.
The injection of MP-BCG induced a significant influx of neutrophils (identified as CD45+, Ly6G+, CD11b+, MHC-II–, CD11c–) after 24 h compared with the injection of MP-UI (Fig. 3A). The injection of MP-Mtb induced an even more pronounced increase in infiltrating neutrophils, with an ∼40-fold increase compared with MP-UI. This recruitment of neutrophils was not TNF dependent, because TNF−/− mice injected with MP-Mtb also showed increased neutrophil numbers compared with MP-UI. As a positive control, intradermal injection of BCG also induced a 10- to 20-fold increase in infiltrating neutrophils compared with the untreated and RPMI-injected control mice (data not shown). Interestingly, independent of the origin of the MPs (infected or uninfected macrophages) or the number of neutrophils within the tissues, the proportion of neutrophils with associated MPs remained constant (Fig. 3B).
Injection of MP-BCG or MP-Mtb induces neutrophil recruitment into the skin. (A) Wild-type C57BL/6 and TNF−/− mice were injected intradermally with 105 MP-UI per 4 μl, 105 MP-BCG per 4 μl, or 105 MP-Mtb per 4 μl into both ears. At 24 h after injection, cells were isolated from both ears, pooled, stained, and analyzed for neutrophils gated as CD45+, CD11b+, Ly6G+, CD11c−, and MHC-II−. Absolute cell numbers and the means and SEM of four to five mice per group are displayed. Two independent biological experiments were performed in a similar manner with representative result displayed. MP-UI versus other conditions is shown. (B) Percentage of neutrophils from (A) that were positive for DiD-labeled MP. (C) Visualization of increased accumulation of neutrophils into the ears of mice by MPs from M. tuberculosis–infected macrophages using confocal microscopy. Purified bone marrow–derived neutrophils from C57BL/6 mice were labeled with CellTrace Violet, and 2 × 107 labeled neutrophils were injected i.v. into CD11c-EYFP C57BL/6 mice. These mice were then injected intradermally with 4 μl 5 × 104 DiD-labeled MP-UI or MP-Mtb into the ears. At 24 h after injection, mice were euthanized; their ears were removed, fixed in formalin, and imaged using confocal microscopy. Images are Z-projections of a 40–50-μm total Z-plane captured every 3 μm. Dotted circles are hair follicles, and the surrounding scale’s major increment is 50 μm. The lower panels show a close-up of the square region present in the upper panels. The experiment was performed twice independently, with a representative field of view slightly askew from the injection site being displayed. *p < 0.05 (Student t test).
To visualize the cellular interactions quantitated in Fig. 3A, we labeled MP-UI or MP-Mtb with membrane-binding DiD dye and injected these intradermally into CD11c-EYFP+ mice directly after adoptive transfer of neutrophils labeled with CellTrace Violet. After 24 h, the mice were euthanized; their ears were removed, fixed, and imaged using confocal microscopy. Injection of MP-Mtb resulted in a significant increase in fluorescently labeled neutrophils and CD11c-EYFP–positive cellular recruitment to the ear. This pattern of recruitment was not seen after injection of MP-UI (Fig. 3C), which is consistent with the flow cytometry results in Fig. 3A.
Finally, the influence of MP-Mtb on the recruitment of other inflammatory cells, specifically macrophages and dermal DCs and the number of MPs associated with these cells, was also examined. Only mice receiving MP-Mtb showed an increase in total macrophages, defined as the CD45+, CD11b+, CD11c−, Ly6G−, MHC-II−/lo cells, within the dermis (Fig. 4A). This effect was dependent on TNF, because the injection of MP-Mtb into TNF−/− mice did not increase macrophage recruitment. In addition to the observed increase in macrophage numbers, an increased proportion of macrophages was associated with MP-Mtb (Fig. 4B). This macrophage association was also TNF dependent, because TNF−/− mice injected with MP-Mtb recruited fewer macrophages to the injection site, with a lower proportion of these macrophages associated with MPs.
Macrophages and DCs are recruited by MP-Mtb after intradermal injection. Total macrophage (A) and DC (C) populations were quantitated after intradermal injection of DiD-labeled MP-UI, MP-BCG, or MP-Mtb into the ears of C57BL/6 mice. TNF−/− mice were similarly injected with MP-Mtb. Macrophages were defined as the CD45+, CD11b+, CD11c−, Ly6G−, MHC-II−/lo cells, whereas dermal DCs were CD45+, CD11b+, CD11c+, MHC-IImid/hi, Ly6G−, CD326−. (B and D) Percentage of macrophages and dermal dendritic cells from their respective individual total cell counts that were also positive for DiD fluorescently labeled MP. MP-UI versus other conditions. Two independent biological experiments were performed with a representative result displayed. *p < 0.05 (Student t-test). ns, Not significant.
Similar to the recruitment of macrophages, dermal DC populations, defined as CD45+, CD11b+, CD11c+, MHC-IImid/hi, Ly6G−, CD326− were also significantly increased after MP-Mtb injection into wild-type mice, but not TNF−/− mice (Fig. 4C). There were also significantly fewer MPs associated with dermal DCs in the TNF−/− mice than in the wild-type mice (Fig. 4D). These results demonstrate that MPs derived from mycobacteria-infected macrophages contribute to the inflammatory response to injection by stimulating the recruitment of leukocytes to the site of infection. Furthermore, this activity of MP function is influenced by the virulence of the infecting agent, because MP-Mtb created greater cellular recruitment than MP-BCG, and this effect is modulated, in part, by TNF.
Aerosol infection with M. tuberculosis increases leukocyte-derived MPs in the blood of mice
The results of Figs. 3 and 4 reveal a role for MP-Mtb in the local recruitment of leukocytes to the site of injection; therefore, we quantitated the systemic changes in blood borne MPs 4 wk after aerosol infection with M. tuberculosis. Blood was obtained from infected and uninfected mice, and the MPs were isolated and stained for CD41, to measure platelet-derived MPs or CD45 to measure leukocyte-derived MPs (Fig. 5). There was a significant increase in the number of circulating CD45+ MPs in the blood of M. tuberculosis–infected mice. In contrast, the levels of CD41+ MPs between the two groups remained unchanged.
CD45+ MP increase in blood during acute infection with M. tuberculosis. C57BL/6 mice were infected with M. tuberculosis by aerosol infection, and after 4 wk their circulating blood MPs levels were compared with uninfected mice. Plasma MPs were stained for CD41 or CD45 and analyzed with flow cytometry. Two independent biological experiments yielded similar results with a representative example shown. *p < 0.05 (Student t test), MP-UI versus MP-Mtb.
MPs colocalize with Rab11a recycling endosomes
To determine the intracellular location of MPs taken up by macrophages, CFSE-stained MP-UI and MP-BCG were added to fresh IC21 macrophages and incubated for 24 h, when MPs were observed within the cells. Intracellular staining of Rab5a, an early endosome marker, and Rab7, a late endosomal maker, showed no colocalization with MP-UI or MP-BCG (Fig. 6A). Similarly, LAMP-1 (a lysosomal marker) did not colocalize with MPs. In contrast, MPs were found associated with the recycling endocytic cell protein marker Rab11a (Fig. 6A–C). MP-BCG had a higher number of colocalizing events by Pearson’s correlation coefficient calculation than MP-UI (Fig. 6D), indicating a greater association between Rab11a and MP-BCG than with MP-UI.
MPs interact with the Rab11a recycling compartment. (A) MPs labeled with CFSE were added to IC21 cells at a 10:1 ratio. After 24 h, cells were labeled with anti-Rab5a, anti-Rab7, anti-Rab11a, or anti-LAMP-1 and examined with confocal microscopy. Extended focus images are displayed. Scale bar, 20 μm. (B) Representative close-up extended focus images of Rab11a recycling compartment with MP-UI or MP-BCG. Scale bar, 10 μm. (C) Orthographic and planar projection images showing colocalization of MPs with Rab11a. (D) Pearson’s correlation coefficient for the colocalization between CFSE-labeled MPs and the recycling endosome marker Rab11a in IC21 macrophages. Experiments were performed in triplicate, and multiple random fields of view were analyzed (minimum 10 per condition) and performed independently twice.
MPs derived from mycobacteria-infected macrophages stimulate Ag-specific CD4+ T cell proliferation in vitro and in vivo
We hypothesized that MPs may represent a means for transferring mycobacterial Ag between cells. To address this hypothesis, MPs were purified from BCG-infected or M. tuberculosis–infected IC21 macrophages and incubated with bone marrow–derived dendritic cells (BMDCs) before the addition of CFSE-labeled T cells (p25) specific for M. tuberculosis Ag 85B. After 3 d, both MP-BCG and MP-Mtb induced significant p25 CD4+ T cell proliferation demonstrating the presence of Ag 85B in both MP preparations (Fig. 7). MP-Mtb induced a greater response than did MP-BCG, indicating that comparatively more mycobacterial Ag or costimulatory factors were associated with these MPs. T cell activation required the processing of MPs by a recipient APC, because the addition of MP-BCG to p25 CD4+ T cells in the absence of BMDC failed to stimulate T cell proliferation (Fig. 7B).
MPs purified from mycobacteria-infected IC21 cells elicit Ag-specific CD4+ T cell proliferation in vitro and in vivo. (A) MPs were purified from IC21 cells 3 d after infection with either BCG or M. tuberculosis and were added to murine BMDCs at a ratio of 10 MPs per DC. p25 splenic CD4+ T cells labeled with CFSE were added to each sample at a ratio of 10 T cells to 1 DC. Cells were stained for CD3, CD4, CD44, and CD45.1. Representative examples of uninfected IC21 MPs, BCG-infected IC21 MPs, and M. tuberculosis–infected IC21 MPs incubated with DCs and CFSE-labeled p25 CD4+ T cells. (B) The percentage of CFSElo p25 CD4+ T cells present after 3 d incubation for the conditions listed. Each condition was performed in triplicate, and the means and SEM are presented from one of two independent experiments. In all cases, similar results were observed between biological replicates. (C) In vivo injection of MPs and CFSE-labeled p25 CD4+ T cells demonstrates that MPs induce T cell proliferation in vivo. The percentage of p25 CD3+ CD4+ T cells having proliferated 5 d after footpad injection of MP-UI or MP-Mtb is shown. The means and SEM are presented from one of two independent experiments. *p < 0.05 (Student t test), (B) T cells alone versus conditions listed, not significant (ns), (C) MP-UI versus MP-Mtb.
To characterize further the Ag-carrying potential of MPs, the capacity of MP-Mtb to stimulate p25 CD4+ T cell proliferation was examined in vivo. CFSE-labeled p25 CD4+ T cells were adoptively transferred into C57BL/6 mice after footpad injection of MP-UI, MP-Mtb, or viable high-dose M. tuberculosis. After 5 d, p25 CD4+ T cells were recovered from the popliteal lymph node, and proliferation of the T cells was quantitated. Approximately 10% of the Ag-specific CD4+ T cells proliferated in response to MP-Mtb (Fig. 7C). In comparison, MP-UI stimulated no T cell proliferation, whereas infection with 5 × 104 M. tuberculosis caused ∼90% of the CD4+ p25 T cells to proliferate.
Discussion
This study demonstrates that mycobacterial infection stimulates the release of MPs by infected macrophages, and these MPs induced distinct immunologic responses compared with MPs secreted under normal physiologic conditions. These effects included the MP-mediated exchange of mycobacterial Ag between cells, the induction of cytokine and chemokine release, and the stimulation of macrophages to disrupt respiratory epithelial cell monolayers. Aerosol infection of mice with M. tuberculosis increased the number of CD45+ MPs in the blood, and injection of MPs from M. tuberculosis–infected macrophages enhanced leukocyte recruitment, indicating their proinflammatory potential in vivo.
It is well documented that even under normal physiologic conditions, human plasma contains a significant number of circulating MPs of various cellular origin (16). Infection with mycobacteria increased numbers of MPs released from primary human macrophages. Taken together with our findings that M. tuberculosis infection increased the number of CD45+ MPs in the plasma, this suggests that MPs may have a role as a biomarker of infection.
M. tuberculosis infection caused greater MP release than BCG infection, suggesting that bacterial virulence is also a modulating factor in MP release. MP production was independent of IFN-γ; in addition, MP-Mtb provoked greater and more diverse cellular recruitment in vivo than did MP-BCG (Figs. 3, 4). This effect may be caused by differences in the Ags within the MPs, different cellular receptors or ligands expressed on their surface, or proinflammatory cytokines and chemokines packaged within the MPs, or all of the above. The increased recruitment of APCs might explain the greater effects of MP-Mtb on CD4+ T cell proliferation observed in Fig. 7. In addition, MP-Mtb was associated with increased recruitment of macrophages and increased MP uptake. It is possible that a feedback loop exacerbating and amplifying the effect is occurring where MPs are taken up by macrophages, stimulating them to produce additional MPs or release proinflammatory cytokines that assists in the recruitment of additional leukocytes.
MPs released from mycobacteria-infected macrophages were proinflammatory, inducing TNF, IL-8, MIP-1α, and MCP-1 release from recipient macrophages. Furthermore, the addition of MPs to uninfected macrophages resulted in the disruption of respiratory epithelial cell junctions, as measured by a decrease in membrane resistance over time (Fig. 2E). This effect was dependent on the concomitant presence of both MPs from infected donor macrophages and effector macrophages, as macrophages or MPs alone did not disrupt the epithelial monolayer. In vivo, MPs released from infected alveolar macrophages might contribute to the disruption of the epithelial cell membrane and the recruitment of leukocytes to the site of the infection in the alveolus and thus facilitate the subsequent formation of the granuloma. This hypothesis is strengthened by our findings that MPs generated from M. tuberculosis–infected macrophages induced neutrophil, macrophage, and DC infiltration when injected in vivo (Figs. 3, 4). Both neutrophil (35) and migratory DCs (36–38) transport intracellular mycobacteria to draining lymph nodes, leading to activation of naive CD4+ T cells and the development of effector T cell responses against M. tuberculosis (38). Mycobacterial Ags may be transported by extracellular MPs or by MPs internalized by migratory APCs, and this might amplify the cellular immune response at a time when small numbers of M. tuberculosis are present at the initial site of infection. The continued production of MPs from infected macrophages within the granuloma could signal the presence of infection to noninfected cells, stimulating the production of TNF, which is essential for granuloma formation and maintenance (28, 39), and the recruitment of monocytes into the granuloma. These MPs may also deliver Ag to APCs within infected tissues and more distant lymphoid organs, aiding the activation and recruitment of effector T cells. Thus, MPs might represent a mechanism to facilitate intercellular communication within the granuloma.
Within macrophages, M. tuberculosis bacilli interact with multiple Rab proteins (40, 41). Rab proteins are involved in endosome trafficking within the cell and have an integral role in the pathogenesis of bacteria like M. tuberculosis. One virulence mechanism expressed by M. tuberculosis is the capacity to interrupt phagosomal maturation from Rab5 expressing early endosomes to Rab7 expressing late endosomes, thereby providing access to the transferrin-receptor recycling pathway and enhancing replication within macrophages (42–45). Our MPs showed no colocalization with Rab5 or Rab7, suggesting that MPs are not processed through this pathway; however, they showed a strong association with the recycling endosome marker Rab11a, which was higher for MPs from infected cells. Rab11a is also associated with the trans-Golgi, indicating that MPs may be transported to the Golgi directly after uptake by macrophages (46). MPs and their contents would then be able to interact with the exiting components of the trans-Golgi network, including MHC-II and secreted cytokines (47, 48).
TLR4 has also been shown recently to colocalize with Rab11, revealing a new mechanism of NF-κB stimulation whereby, in the presence of bacterial Ags, TLR4 is recruited to the endocytic-recycling compartment (49, 50). The association of MPs with recycling endosomes suggests a potential mechanism whereby MPs can interact with TLR4 internally, present mycobacterial TLR4-dependent ligands, and in turn stimulate NF-κB activation, which then induces TNF production. Therefore, infection of the donor cell influences MP surface receptor and ligand expression as well as MP contents, and this can influence the intracellular processing of MPs within the recipient cells. The fate and pattern of colocalization of the non–Rab11a-associated MPs within recipient macrophages remains to be determined, because they did not associate with early or late endosomes or lysosomes at the time points examined.
Our data demonstrate that MPs, like exosomes, carry mycobacterial Ags (1–3). Our MP preparations were screened with scanning electron microscopy, and the absence of exosomes confirmed that MPs express phosphatidylserine and were quantified by size and annexin V binding. Furthermore, because MPs are formed from the plasma membrane, they contain surface markers of their cell of origin, including MHC-II and CD40, as demonstrated for MPs from mycobacteria-infected macrophages (12, 17). MPs from M. tuberculosis– and BCG-infected macrophages induced the proliferation of Ag-specific CD4+ T cells in a DC-dependent manner. MPs added directly to T cells in the absence of DCs were unable to induce T cell activation. This finding contrasts with a recent study suggesting that the addition of MPs without APCs to T cells was sufficient to induce mycobacterial Ag-specific T cell proliferation (51). The reason for this difference might be the origin of the MPs. The MPs in the current study were released physiologically from infected macrophages over the course of 72 h, whereas in the study by Ramachandra et al. (51), macrophages were infected with mycobacteria, washed to remove endogenously released MPs, and stimulated with ATP to generate microvesicles. ATP stimulation induces apoptosis of macrophages with the release of apoptotic bodies and microvesicles of different sizes (25), and this may have influenced the activity of the MP preparations. Our results show unequivocally that purified MPs require Ag processing by APCs for T cell activation.
In summary, our work demonstrates that MPs released by mycobacteria-infected cells can have multiple roles during infection. Following acute aerosol infection with M. tuberculosis, MPs released by the small number of infected alveolar macrophages can aid the disruption of the tight junctions of the epithelial cells lining the alveolus, facilitating the recruitment of leukocytes and the formation of the nidus of infection. At later stages within granulomas, MPs could provide long-range and complex intercellular communication between macrophages and epithelioid cells, stimulating the production of TNF, and transferring Ags to APCs, resulting in the persistent antigenic stimulation required for long-term immunologic containment of infection.
Disclosures
The authors have no financial conflicts of interest.
Acknowledgments
We thank the staff at the Australian Microscopy and Microanalysis Research Facility (University of Sydney) for the use of facilities and scientific and technical assistance; Prof. K. Takatsu and Dr. J. Ernst for providing the p25 TCR transgenic mice; Prof. W. Weninger for the CD11c-EYFP mice; Dr. A. Abtin, Dr. B. Roediger, and C. Gillis for assistance with the mice intradermal injections and flow analysis; E. Shklovskaya for advice on flow analysis; and Dr. N. West and Dr. J. Huch for critical reading of the manuscript.
Footnotes
This work was supported by National Health and Medical Research Council Project Grant 570771 and an infrastructure grant from the New South Wales Government to the Centenary Institute.
Abbreviations used in this article:
- BCG
- bacillus Calmette–Guérin
- BMDC
- bone marrow–derived dendritic cell
- BMDM
- bone marrow–derived macrophage
- DC
- dendritic cell
- ECIS
- Electrical Cell-Substrate Impedance System
- MHC-II
- MHC class II
- MP
- microparticle
- MP-BCG
- BCG-infected cell
- MP-Mtb
- M. tuberculosis-infected cell
- MP-UI
- MP isolated from uninfected cell.
- Received July 6, 2012.
- Accepted November 12, 2012.
- Copyright © 2013 by The American Association of Immunologists, Inc.
References
In this issue
