Exosomes released from macrophages infected with intracellular pathogens stimulate a proinflammatory response in vitro and in vivo

Abstract

Intracellular pathogens and the molecules they express have limited contact with the immune system. Here, we show that macrophages infected with intracellular pathogens Mycobacterium tuberculosis, M bovis BCG, Salmonella typhimurium, or Toxoplasma gondii release from cells small vesicles known as exosomes which contain pathogen-associated molecular patterns (PAMPs). These exosomes, when exposed to uninfected macrophages, stimulate a proinflammatory response in a Toll-like receptor- and myeloid differentiation factor 88-dependent manner. Further, exosomes isolated from the bronchoalveolar lavage fluid (BALF) of M bovis BCG-infected mice contain the mycobacteria components lipoarabinomannan and the 19-kDa lipoprotein and can stimulate TNF-alpha production in naive macrophages. Moreover, exosomes isolated from M bovis BCG- and M tuberculosis-infected macrophages, when injected intranasally into mice, stimulate TNF-alpha and IL-12 production as well as neutrophil and macrophage recruitment in the lung. These studies identify a previously unknown function for exosomes in promoting intercellular communication during an immune response to intracellular pathogens, and we hypothesize that extracellular release of exosomes containing PAMPs is an important mechanism of immune surveillance.

Figure 1

Activity and characterization of exosomes isolated from uninfected and M bovis BCG–infected J774 cells. (A) Morphologic characterization of exosomes by transmission EM. Representative EM of a sucrose gradient purified exosome (magnification, × 100 000; scale bar = 100 nm). Images were acquired with Microtek Scan Wizard 5 (Microtek Lab, Carson, CA). (B) Surface phenotype of exosomes by flow cytometry. Sulfate/aldehyde latex beads coated with indicated exosome preparations were probed for CD86, CD81, LAMP1, LAMP2, and MHC class II, and the mean fluorescence intensity (MFI) shown (filled peaks). Open peaks represent background MFI of exosome-coated beads probed with isotype control mAb. Data are the representative of 2 separate experiments. Exosomes derived from uninfected or M bovis BCG–infected J774 cells were added to BALB/c BMMs and TNF-α (C) or RANTES (D) production was evaluated 24 hours after exosome treatment by ELISA. (E) BALB/c BMMs were treated with 10 μg of exosomes isolated from uninfected or M bovis BCG–infected J774 cells. Western blot analyses were performed on macrophage cell lysates obtained 24 hours after exosome treatment using antibodies specific for iNOS. Total p38 was used as a loading control. Results are representative of 2 separate experiments. (F) BALB/c BMMs were incubated with the exosome-containing or exosome-depleted material from the sucrose gradient. TNF-α levels were measured by ELISA 24 hours after exosome treatment. ELISA data are representative of 2 independent experiments and are expressed as means plus or minus standard deviation of duplicate wells. RC indicates resting cells; and UI, exosomes from uninfected J774 cells.

Figure 2

Exosomes, not apoptotic vesicles, are responsible for the proinflammatory activity. (A) J774 cells were infected with M bovis BCG for 4 hours. As controls, macrophages were left uninfected or were starved of serum to induce apoptosis. Macrophages were treated with annexin V (green) and propidium iodide (red), fixed, and analyzed by confocal microscopy. Nuclei were stained using the bisalkylamino-anthraquinone fluorphore Draq5. Fluorescent images are representative of 2 independent experiments (scale bar = 10 μm). Images were acquired with a Nikon Diaphot 200 (Nikon, Tokyo, Japan), equipped with a 60× oil plan Apo160/0.17 NA lens with BioRad LaserSharp 2000 software (BioRad, Hercules, CA). (B) Exosomes were prepared from J774 cells infected with M bovis BCG treated with 50 μM of the caspase-3 inhibitor Ac-DEVD-CHO, or left untreated. As a control, apoptotic vesicles were also purified from serum-deprived J774 cells. Exosomes or apoptotic vesicles were added to separate, uninfected J774 cells; 24 hours after treatment, cell-conditioned media were assayed for TNF-α by ELISA. Data are representative of 2 independent experiments and are expressed as means plus or minus standard deviation for duplicate wells. (C) Cell lysates obtained 24 hours after treatment with exosomes or apoptotic vesicles were analyzed for iNOS expression by Western blot. Total p38 was used as a loading control. The results are representative of 2 separate experiments. (D) Sulfate/aldehyde latex beads coated with exosomes or apoptotic vesicles were probed by flow cytometry for annexin V binding or FcRII/III, and the percentage of beads positive for the markers is shown. Results are representative of 2 independent experiments and are exposed as means plus or minus standard deviation for duplicate wells. UI indicates uninfected J774 cells; BCG, J774 cells infected with M bovis BCG.

Figure 3

Proinflammatory activity of exosomes is dependent on the time of isolation from M bovis BCG–infected J774 cells. (A) Exosomes were isolated from the cell-conditioned media of uninfected (UI) J774 cells or at indicated times after infection with M bovis BCG and the isolated exosomes (10 or 20 μg) added to BALB/c BMMs. Cell-conditioned media collected 24 hours after exosome treatment were assayed for TNF-α production by ELISA. Data are representative of 2 independent experiments and are expressed as means plus or minus standard deviation for duplicate wells. (B) Exosomes (10 μg) isolated from the cell-conditioned media at the times indicated were analyzed for the mycobacterial components LAM and the 19-kDa lipoprotein, and the host components hsp70 and MHC class II by Western blot using antigen-specific antibodies. (C) Densitometry in arbitrary units for the bands shown in the Western blot for the 19-kDa lipoprotein and LAM. (D) J774 cells were left uninfected or were infected with M bovis BCG for 24, 48, 72, and 96 hours. Exosomes were isolated and resuspended in equal volumes of PBS and protein concentration (mg/mL) determined by Micro BCA assay. The data are representative of 2 independent experiments and expressed as means plus or minus standard deviation for triplicate wells.

Figure 4

Activity and characterization of exosomes isolated from uninfected, M bovis BCG–infected and M tuberculosis H37Rv–infected THP-1 cells. (A) Morphologic characterization of exosomes by transmission EM. Representative EM of a sucrose-gradient–purified exosome (magnification, × 80 000; scale bar = 100 nm). Images were acquired with BioRad LaserSharp 2000 (BioRad). (B) Surface phenotype of exosomes by flow cytometry. Sulfate/aldehyde latex beads coated with indicated exosome preparations were probed for the indicated proteins. The MFI is shown. Light-penciled peaks represent background MFI of exosome-coated beads probed with isotype control mAb. The data are representative of 2 separate experiments. (C) Exosomes (10 μg) isolated from the cell-conditioned media of uninfected or THP-1 cells infected for 72 hours with M. bovis BCG were analyzed by Western blot for the indicated host proteins as well as the mycobacteria 19-kDa lipoprotein and LAM. (D) Exosomes isolated from uninfected, M bovis BCG–, or M tuberculosis H37Rv–infected THP-1 cells were added to separate uninfected THP-1 cells; 24 hours after treatment, cell-conditioned media were assayed for TNF-α by ELISA. Data are representative of 3 independent experiments and are expressed as means plus or minus standard deviation for duplicate wells. (E) Cell lysates obtained 24 hours after exosome treatment were analyzed for iNOS expression by Western blot. Total p38 was used as a loading control. Results are representative of 2 independent experiments

Figure 5

Exosomes isolated from mycobacteria-infected THP-1 cells can induce a TLR/MyD88-dependent proinflammatory response in naive macrophages. (A) C57BL/6 BMMs were treated with 10 μg of exosomes isolated from uninfected, M bovis BCG–infected, or H37Rv-infected THP-1 cells. Western blot analyses were performed on cell lysates obtained 24 hours after exosome treatment using antibodies specific for phosphorylated p38 or IκBα. Total p38 was used as a loading control. Data are representative of 2 separate experiments. TLR2^−/− (B), TLR4^−/−, (C) or MyD88^−/− (D) and the corresponding BALB/c or C57BL/6 WT BMMs were treated with the indicated concentrations of the different exosome preparations, and TNF-α levels were measured 24 hours after exosome treatment by ELISA. Data are representative of 2 independent experiments and are expressed as means plus or minus standard deviation of duplicate wells.

Figure 6

Exosomes isolated from THP-1 cells infected with intracellular pathogens but not with heat-killed BCG or LPS-treated cells contain proinflammatory activity. (A) Exosomes were isolated from THP-1 cells activated for 72 hours with LPS (200 ng/mL) or infected for 72 hours with heat-killed M bovis BCG (HI-BCG), live M bovis BCG, or S typhimurium SL1344. Exosomes were added at 1 or 5 μg to uninfected THP-1 cells, and 24 hours after treatment, cell-conditioned media were assayed for TNF-α by ELISA. Data are represented as means plus or minus standard deviation for duplicate wells. (B) S typhimurium–infected THP-1 cells were incubated with N-rhodamine-phosphoatidylethanolamine (Rh-PE) to label the MVBs. Trafficking of LPS in the infected cells was measured over time using permeablized cells stained with a primary antibody to LPS and FITC-labeled secondary antibody. Representative confocal images are shown for the different time points; coincident staining between LPS and Rh-PE appear yellow in the merged images. Nuclei were stained with the bisalkylaminoanthraquinone fluorphore Draq5 (scale bar = 5 μm). (C) Flow cytometry of exosomes isolated from uninfected or Salmonella-infected THP-1 cells after coating on sulfate/aldehyde beads and probing with a monoclonal antibody against LPS (closed peaks). Data are shown as MFI. Open peaks represent background MFI of exosome-coated beads probed with isotype control monoclonal antibody. Data shown are representative of 2 independent experiments. (D) TLR4^−/− and WT C57BL/6 BMMs were treated with 10 μg of exosomes isolated from uninfected or S typhimurium–infected THP-1 cells. TNF-α levels were measured 24 hours after exosome treatment by ELISA. (E) THP-1 cells were left uninfected or infected with T gondii or M bovis BCG for 72 hours. Exosomes were isolated from the cell-conditioned media. Uninfected C57BL/6 BMMs were treated for 24 hours with the different exosome preps, and the amount of secreted TNF-α was measured by ELISA. The ELISA data are representative of 2 separate experiments and are expressed as the means plus or minus standard deviation of duplicate wells.

Figure 7

Exosomes isolated from mycobacteria-infected THP-1 cells or from M bovis BCG–infected mice induce a proinflammatory response in vitro and in vivo. IL-12 p40 (A) and TNF-α (B) protein levels in lysates of mouse lungs 1 day after intranasal injection with 20 μg exosomes derived from uninfected, M bovis BCG–infected, or M tuberculosis H37Rv–infected THP-1 cells as quantified by ELISA. Background cytokine levels (ie, PBS injection) are subtracted from all values. Shown are the means plus or minus standard deviations for 3 mice and are representative of 2 separate experiments. (C) Neutrophil infiltration in the BALF of mice injected with the various exosome preparations or PBS control at 1 day after injection. Data are representative of 2 experiments. (D) Exosomes (3 μg) isolated from the BALF of uninfected (UI) or M bovis BCG–infected C57BL/6 mice were analyzed by Western blot using specific antibodies for the host protein hsp70 and for the mycobacteria 19-kDa lipoprotein and LAM. (E) C57BL/6 BMMs were treated with 200 ng of exosomes isolated from the BALF of uninfected or M bovis BCG–infected mice. Cell-conditioned media were assayed for TNF-α by ELISA 24 hours after exosome treatment. Data are representative of 2 experiments and are expressed as means plus or minus standard deviation for duplicate wells.