|Home | About | Journals | Submit | Contact Us | Français|
Sarcoidosis is characterized by noncaseating granulomas containing CD4+ T cells with a Th1 immunophenotype. Although the causative antigens remain unknown, independent studies noted molecular and immunologic evidence of mycobacterial virulence factors in sarcoidosis specimens. A major limiting factor in discovering new insights into the pathogenesis of sarcoidosis is the lack of an animal model. Using a distinct superoxide dismutase A peptide (sodA) associated with sarcoidosis granulomas, we developed a pulmonary model of sarcoidosis granulomatous inflammation. Mice were sensitized by a subcutaneous injection of sodA, incorporated in incomplete Freund's adjuvant (IFA). Control subjects consisted of mice with no sensitization (ConNS), sensitized with IFA only (ConIFA), or with Schistosoma mansoni eggs. Fourteen days later, sensitized mice were challenged by tail-vein injection of naked beads, covalently coupled to sodA peptides or to schistosome egg antigens (SEA). Histologic analysis revealed hilar lymphadenopathy and noncaseating granulomas in the lungs of sodA-treated or SEA-treated mice. Flow cytometry of bronchoalveolar lavage (BAL) demonstrated CD4+ T-cell responses against sodA peptide in the sodA-sensitized mice only. Cytometric bead analysis revealed significant differences in IL-2 and IFN-γ secretion in the BAL fluid of sodA-treated mice, compared with mice that received SEA or naked beads (P = 0.008, Wilcoxon rank sum test). ConNS and ConIFA mice demonstrated no significant formation of granuloma, and no Th1 immunophenotype. The use of microbial peptides distinct for sarcoidosis reveals a histologic and immunologic profile in the murine model that correlates well with those profiles noted in human sarcoidosis, providing the framework to investigate the molecular basis for the progression or resolution of sarcoidosis.
The use of a microbial peptide, sodA, distinct for sarcoidosis revealed a histologic and immunologic profile in a murine model that correlated well with those profiles in human sarcoidosis. This murine model of sarcoidosis provides a framework for investigating the molecular basis of sarcoidosis progression or resolution.
Although sarcoidosis granulomatous inflammation can involve any organ, pulmonary involvement is evident in more than 90% of patients with sarcoidosis. Non-necrotizing granulomas comprise the histologic feature of the disease, with variable degrees of Th1-lymphocytic inflammation or fibrosis. The granuloma is compact and organized. It consists of a core of epithelioid and multinucleated giant cells, derived from the fusion of monocytes and macrophages. The macrophages are infiltrated and surrounded predominantly by T-helper (CD4+) cells, with some aggregates of T-suppressor (CD8+) cells, and a collagenous band with interspersed fibroblasts. The diagnosis is usually based on the presence of noncaseating granulomas, with a consistent clinical presentation and negative studies for active infection or other causes of granuloma formation.
Recent studies from independent laboratories revived the association of sarcoidosis with infectious antigens. In Japanese patients with sarcoidosis, quantitative differences in Propionibacterium acnes DNA among sarcoidosis and control subjects were detected (1). Among American patients with sarcoidosis, the presence of genetically distinct mycobacterial nucleic acids in sarcoidosis granulomas was reported (2). Song and colleagues (3) used matrix-assisted laser desorption/ionization time of flight mass spectrometry to demonstrate the presence of mycobacterial catalase-peroxidase (katG) proteins in sarcoidosis granulomas. Mycobacterial heat-shock proteins were also evident in Polish sarcoidosis granulomas (4). Major histocompatibility complex (MHC) Class II alleles are associated with protection or susceptibility in patients with sarcoidosis (5, 6). Th1 cellular immune responses to the mycobacterial virulence factors ESAT-6 and katG were reported in sarcoidosis peripheral blood mononuclear cells and bronchoalveolar lavage (7, 8). Mycobacterial antigens that are recognized in the context of MHC Class II human leukocyte antigen (HLA)-DR were reported (9). Most recently, antigen-specific immune responses to mycobacterial virulence factors were reported to be present in sarcoidosis bronchoalveolar lavage (BAL) at the time of diagnosis, as mediated via Toll-like receptor 2 signaling (10). The molecular analysis of sarcoidosis granulomas for microbial virulence factors led to the identification of genetically distinct sequences of mycobacterial superoxide dismutase A (sodA; GenBank accession number DQ768096) (11). Unlike other proteins reported in sarcoidosis granulomas, these sequences had not been previously associated with other mycobacterial infections such as tuberculosis. Taken together, these studies indicate a key role of mycobacterial proteins in the pathogenesis of sarcoidosis in American patients.
Progress toward the development of an animal model has been hindered by a lack of antigens that induce the sarcoidosis Th1 immunophenotype. Using the distinct sodA peptides isolated from patients with sarcoidosis only (GenBank accession number ABG68026), we developed a pulmonary model of sarcoidosis granulomatous inflammation. At present, no clear understanding exists of the factors that influence the natural history of sarcoidosis or govern its relative degrees of inflammation, granuloma formation, or fibrosis. The development of an animal model of sarcoidosis immunopathogenesis, using distinct microbial peptides associated with sarcoidosis granulomas, will enable us to identify molecular events that contribute to clinical outcomes in sarcoidosis. Some of the results of these studies were previously reported in the form of an abstract (12).
Six-week-old female C57Bl/6 mice were purchased from Jackson Laboratory (Bar Harbor, ME). Mice were maintained under specific pathogen–free conditions, and all in vivo manipulations were performed under protocols approved by the Institutional Animal Care and Use Committee of the Cleveland Clinic.
The sodA peptide, AAAIAGAFGSFDKFR, was synthesized as described previously (11). Each peptide was synthesized by solid-phase F-moc chemistry (Genemed Synthesis, San Diego, CA) to a purity of greater than 70%. Identity was confirmed by mass spectroscopy, and purity was assessed according to high-performance liquid chromatography.
We used the timeline of the murine lung granuloma model described by Chensue and colleagues (13), with some modifications. Mice were sensitized by a subcutaneous injection of 20 μg of Mycobacterium sodA peptide, incorporated into 0.25 ml of incomplete Freund's adjuvant (IFA; Sigma, St. Louis, MO) on Day 1. Control groups included untreated mice (baseline), mice with no sensitization (ConNS), mice sensitized with IFA only (ConIFA), or mice injected intraperitoneally with 3,000 Schistosoma mansoni eggs suspended in 0.5 ml of PBS. Fourteen days later, ConNS and ConIFA mice were challenged by tail-vein injection with 6,000 Sepharose 4B naked beads (in 0.5 ml of PBS; Sigma). The sodA-sensitized and Schistosoma egg–sensitized mice were challenged in the same manner with beads covalently coupled to sodA or soluble Schistosome egg antigens (SEAs; Schistosome Biological Supply Center, Theodor Bilharz Research Institute, Giza, Egypt), respectively.
After BAL, PBS-perfused lungs were excised, excluding the trachea and major bronchi. The left lung was fixed in Histochoice (Amresco, Solon, OH), processed, and paraffin-embedded. Paraffin sections (5-μm thickness) were stained with hematoxylin and eosin (H&E) for histologic examination and determination of granuloma size. Granulomas were measured blindly, using computerized morphometry with an Olympus BX41 microscope (Olympus America, Inc., Center Valley, PA). A minimum of 10 granulomas per lung, sectioned through the central bead nidus, was used in measurements.
Immunohistochemistry of sodA-elicited lung granulomas was performed for the presence of T cells, macrophages, and B cells. Paraffin sections of left lungs of mice were deparaffinized, rehydrated, and permeabilized with Trilogy (Cell Marque Corporation, Rocklin, CA) for 30 minutes in a steamer, rinsed in fresh hot Trilogy for 30 minutes, and cooled for 5 minutes. In situ peroxidase activity was quenched with a methanol/30% H2O2 solution for 20 minutes. After a brief rinse in water, nonspecific binding was prevented by incubating tissue sections in serum-free, ready-to-use Protein Block (DAKO, Carpinteria, CA) for 5–10 minutes at room temperature. Purified rat anti-mouse Mac-2 monoclonal antibody (1:500, clone M3/38 by Cedarlane; Accurate Chemical and Scientific Corp., Westbury, NY), polyclonal rabbit anti-mouse CD-3 (1:1,000, ab5690; Abcam, Inc., Cambridge, MA), and polyclonal rat anti-mouse CD45R (also B220; 1:50, clone RA3-6B2; B.D. Biosciences, San Jose, CA) were used for the identification of macrophages, general T cells, and general B cells, respectively. Sections were incubated in primary antibody diluted in AB Diluent (DAKO) for 60 minutes at room temperature and overnight at 4°C, and then rinsed twice in PBS for 5 minutes. Sections were incubated in Biotin-SP–conjugated AffiniPure Mouse Anti-Rat IgG or Goat Anti-Rabbit IgG (1:1,000; Jackson ImmunoResearch Laboratories, West Grove, PA) secondary antibodies diluted in AB Diluent for 60 minutes at room temperature, and rinsed as already described. Sections were incubated in AB Complex (Vectastain Elite ABC Kit; Vector Laboratories, Burlingame, CA) for 30 minutes at room temperature, and rinsed as already described. Peroxidase activity was developed by incubation in DAB (Vector) according to the manufacturer's instructions. Sections were counterstained in hematoxylin, dehydrated, and mounted in Richard-Allan Mounting Medium (Fisher Scientific, Houston, TX).
For the identification of CD4+ T cells, 5-μm sections of frozen tissues embedded in optimal cutting temperature (OCT; Tissue-Tek, Sakura Japan) were placed on charged slides, fixed for 5 minutes in acetone, rinsed in distilled water, and placed in PBS. Endogenous peroxidase was quenched with 0.03% hydrogen peroxide, and samples were treated with a casein block before addition of the primary antibody. Tissues were incubated with biotinylated rat anti-CD4 (catalogue number 553649; B.D. Pharmingen, San Jose, CA), diluted 1:50, and incubated overnight at 4°C. The Vectastain ABC Elite System (Vector Laboratories, Inc.) and DAB+ (DAKO) were used to produce localized, visible staining. Slides were lightly counterstained with Mayer's hematoxylin, dehydrated, and coverslipped.
Immunohistochemistry of a sarcoidosis granuloma containing the unique sodA sequence was performed for the presence of CD3+cells, CD4+ cells, CD8+ cells, macrophages, and B cells. The CD4+ histologic stain worked suboptimally on the tissue specimen and the control, and was not included in the analysis. For CD3, CD8, CD68, and CD20 staining, 5-μm sections were placed on charged slides and rehydrated. Sections were rehydrated and placed in heated Target Retrieval Solution (DAKO). Endogenous peroxidase was neutralized with 0.03% hydrogen peroxide, followed by a casein-based protein block (DAKO) to minimize nonspecific staining. Sections were incubated with CD3 (sc-1127; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), diluted 1:250 for 60 minutes. The Vectastain Elite ABC Kit and DAB were used to produce localized, visible staining. When staining for CD8+ T cells, sections were incubated with CD8 (sc-7970; Santa Cruz Biotechnology, Inc.), diluted 1:300 for 60 minutes When staining for CD68+ cells, sections were incubated with CD68 (M 0814; DAKO), diluted 1:300 for 60 minutes. For the staining for CD20+ cells, sections were incubated with CD20 (ab9475; Abcam, Inc.), diluted 1:35 for 60 minutes. When staining for CD8, CD68, and CD20, the Envision+ HRP/DAB System (DAKO) was used to produce localized, visible staining. All slides were lightly counterstained with Mayer's hematoxylin, dehydrated. and coverslipped.
BAL was performed as described previously (14), by infusing 1 ml of PBS through a tracheal cannula. The BAL was centrifuged (700 × g, 5 min, 4°C), and the fluid phase was aliquoted and frozen at −70°C for cytokine measurements. After the hypotonic lysis of red blood cells in the pellet, BAL cells were resuspended in 1 ml of PBS and counted using a hemocytometer. Differentials were performed in BAL cells by the cytocentrifugation of 100 μl containing 103 cells/ml onto glass slides, using Wright stain (EM Science, Gibbstown, NJ), and counting by light microscopy. The percentage of different leukocyte populations was determined in duplicate by counting 200 cells in fields selected randomly under a low-power magnification (×2) at which leukocytes cannot be distinguished, and the total number of each leukocyte population was calculated according to the volume of the BAL.
BAL was performed as described above. After centrifugation, BAL cells were counted and analyzed by flow cytometry. Intracellular staining was performed to identify IFN-γ–secreting T cells in response to microbial stimulation. For the intracellular cytokine assay, BAL was pooled from 3–6 mice, and 0.5–1 × 105 BAL cells were incubated in RPMI-1640–supplemented medium with or without sodA (20 μg/ml). Staphylococcal enterotoxin B (SEB, 10 μg/ml; Sigma) stimulation served as a positive control at 37°C under 5% CO2 for 60 minutes before the addition of B.D. GolgiStop (B.D. Biosciences). To demonstrate that the immune response was MHC class II–specific, BAL cells from sodA-treated mice were first incubated at room temperature for 30 minutes with a monoclonal anti-MHC II antibody (1 μg/ml of blocking antibody, clone m5/114.15.2; eBioscience, San Diego, CA) or the corresponding isotype control (rat IgG2b, κ) before antigen stimulation. After 6 hours of incubation at 37°C under 5% CO2, cells were washed and stained with the surface antibodies anti-CD3 and anti-CD4 (B.D. Biosciences) at 4°C for 30 minutes. After washing, fixation, and permeabilization, using a Fix and Perm Kit according to the manufacturer's instructions (B.D. Biosciences), anti–IFN-γ (B.D. Biosciences) was added at 4°C for 45 minutes. Cells were washed and analyzed via flow cytometry. IFN-γ frequencies were defined as the percentage of stimulated CD3+CD4+ T cells minus their unstimulated background frequency. The medium used in all experiments was RPMI-1640 (Cellgro, Mediatech, Inc., Manassas, VA) supplemented with 10% FBS (Gemini Bio-Products, West Sacramento, CA), penicillin (50 U/ml; Cellgro), streptomycin (50 mg/ml; Cellgro), sodium pyruvate (1 mM; Cellgro), and glutamine (2 mM; Cellgro).
To measure Th1 (IL-2 and IFN-γ) and Th2 (IL-4 and IL-5) cytokine levels in BAL, fluid was collected from 5–6 mice from each group: baseline, SEA-treated, or sodA-treated mice. Cytokine levels were measured using a cytometric bead array (CBA; B.D. Biosciences) according to the manufacturer's instructions.
Statistical analyses of differences between groups were performed using a Kruskal-Wallis test or ANOVA as appropriate, followed by Dunnett's multiple comparison test (JMP Statistical Discovery Software; SAS, Cary, NC). Statistical analyses of distinctions in cytokine expression between baseline, SEA-treated, and sodA-treated mice were performed using the Wilcoxon rank sum test at an α-level of 0.05.
To determine if Mycobacterium sodA peptides isolated from human sarcoid granulomas could elicit granuloma formation in mouse lungs, mice were sensitized with an emulsion of IFA and sodA on Day 1 and challenged on Day 14 with sodA-coupled beads, which embolized to the lungs (Figure 1). Lungs and BAL fluid were collected 4 days after the challenge, on Day 18 (Figure 1). A gross examination of the euthanized mice detected the presence of extensive hilar lymphadenopathy and parenchymal infiltrates. No gross abnormalities of the heart, liver, or spleen were evident, and no beads or granulomas were found in these tissues. Histologic analyses of H&E-stained lung sections showed extensive granuloma formation throughout the lungs of sodA-treated mice (Figure 2A). Most granulomas were concentric to the bead (Figure 2A, inset), and localized to the broncho-arterial bundle. The presence of hilar lymphadenopathy and noncaseating granulomas consisting of multinucleate giant cells and lymphocytic infiltrate correlates well with observations in human sarcoid lung histology (Figure 2B). Multinucleated giant cells and lymphocytic infiltration were present in the granulomas of sodA-treated mice (Figures 2A, inset, and 2C). Consistent with previous reports, SEA produced large granulomas with extensive lymphocytic infiltrate (13), as well as extensive extracellular matrix deposition and fibrosis (Figure 2D). In contrast to both sodA-treated and SEA-treated mice, untreated mice had no granulomas. No sensitization (ConNS; Figure 2E) or sensitization with IFA alone (ConIFA; Figure 2F), followed by challenge with naked beads, led to the formation of significantly smaller foreign-body granulomas consisting of a few cell layers (Figure 2E, inset). An examination of the size of granuloma formation between ConNS, ConIFA, sodA-treated, and SEA-treated mice revealed significant differences between treatments. The median size of sodA-elicited granulomas was 32,045 μm2 (25th and 75th percentiles, 23,061 μm2 and 39,634 μm2, respectively; Figure 3), which was larger than the size elicited in ConNS mice (median, 11,247 μm2; 25th and 75th percentiles, 6,425 μm2 and 17,510 μm2, respectively; P < 0.001, according to Kruskal-Wallis test; Figure 3) and ConIFA mice (median, 15,199 μm2; 25th and 75th percentiles, 12,414 μm2 and 21,714 μm2, respectively; P < 0.001, according to Kruskal-Wallis test; Figure 3). In contrast, SEA-elicited granulomas (median area, 61,215 μm2; 25th and 75th percentiles, 42,423 μm2 and 95,263 μm2, respectively) were approximately twofold larger than sodA-elicited granulomas (P < 0.001 according to Kruskal-Wallis test; Figure 3). Dunn's multiple comparison test revealed no differences between the lung granuloma sizes of ConNS and ConIFA mice. However, significant differences were found between sodA-elicited and SEA-elicited granulomas versus those in ConNS mice (P < 0.001 for both) and ConIFA mice (P < 0.01 and P < 0.001, respectively).
The cellular profile of sarcoidosis granulomas is well-characterized. To compare the cellular profile of sodA-elicited mouse lung granulomas with those of patients with sarcoidosis, we assessed for the presence of macrophages, T cells, and B cells. Analogous to what occurs in human sarcoidosis granulomas containing the unique sodA peptide, murine macrophages were interspersed throughout the granuloma (Figures 4A and 4B). However, in the murine model, they were most abundant in the innermost layers, closest to the beads (Figure 4A). In addition, general staining for T cells showed them to be interspersed in the middle layers of both human and mouse granulomas, as expected (Figures 4C and 4D). Specific staining for CD4+ T cells showed them to be abundant in the middle layers of mouse granulomas (Figure 4C, inset). Although B cells were present in both murine and human sarcoidosis granulomas, they did not comprise a significant proportion of the lymphocytic population (Figures 4E and 4F, respectively). Control, serially cut sections, treated with secondary antibody only, showed no staining.
BAL was performed to identify and quantify the different cell types present. No significant difference was evident between the total numbers of leukocytes in the BAL of ConNS, ConIFA, or sodA-treated mice (Table 1). To determine if sodA affected a specific leukocyte population, differentials were performed on BAL cells. Although the number of macrophages and neutrophils was similar among all groups studied, sodA caused a significant increase in lymphocytes compared with the control groups (P < 0.025; Table 1). The ratio of CD4:CD8 T cells in sodA-treated BAL cells was determined by FACS analysis to be 2.3:1. Eosinophils did not constitute a significant cell type in any of the treatments used.
A hallmark of sarcoidosis immunology is a Th1 immunophenotype. To determine if the murine model was similar immunologically to human sarcoidosis, we assessed for the immune recognition of sodA by T cells from the BAL of untreated, SEA-treated, and sodA-treated murine populations. BAL cells stimulated by staphylococcal enterotoxin B (SEB) served as the positive control, and BAL cells alone served as the negative control (Figure 5). Appropriate responses occurred in all three murine populations to SEB, and none to a lack of peptide. BAL cells from sodA mice displayed a twofold CD3+CD4+ IFN-γ response to sodA, compared with no response in baseline or SEA-treated mice (Figure 5). Strong CD3+CD4− T-cell responses were also evident, suggesting that the animal model demonstrates CD8+ T-cell responses to mycobacterial antigens that were also in samples of human sarcoidosis (8, 10). Cytokine analyses of the BAL fluid demonstrated statistically significant differences in the expression of Th1 and Th2 cytokines between untreated, SEA-treated, and sodA-treated mice. The BAL fluid from sodA-treated mice contained significantly higher concentrations of IL-2 compared with baseline mice and SEA-treated mice (P = 0.008 and P = 0.02, respectively; Wilcoxon rank sum test) (Figure 6A). IFN-γ concentrations were also significantly different compared with baseline and SEA control mice (both P = 0.008, Wilcoxon rank sum test) (Figure 6B). No significant differences in IFN-γ concentrations were evident between untreated and SEA-treated murine populations. No significant expression of Th2 cytokines occurred among untreated and sodA-treated murine populations. However, a significant expression of both IL-4 and IL-5 cytokine responses was evident in SEA-treated mice (P = 0.008 for IL-4 and IL-5, Wilcoxon rank sum test) (Figures 6C and 6D).
Previous reports noted an association between MHC Class II alleles and susceptibility to sarcoidosis. According to a recent report, the mycobacterial antigens presented by these MHC Class II alleles are recognized by sarcoidosis CD4+ T cells (15). Although BAL cells from ConNS or ConIFA mice did not elicit a response against sodA peptide, BAL cells from sodA-treated mice showed CD4+ IFN-γ responses that were 5.3 and 6.5 fold higher, respectively, than those of BAL cells from these mice (Figure 7). To assess whether the antigen-specific recognition exhibited by CD4+ T cells from sodA-treated mice was secondary to sodA antigens presented in the context of MHC Class II alleles, BAL cells were incubated with α-MHC II or isotype antibody for 30 minutes before antigen stimulation. The blocking of MHC II abolished the immune recognition of sodA by CD4+ T cells, whereas the isotype antibody did not (Figure 7).
Sarcoidosis involves the lung in more than 90% of patients. The disease is characterized by Th1 alveolitis, accompanied by a CD4+ T-cell predominance, leading to the formation of granulomas. No animal model, to the best of our knowledge, uses peptides that are unique to sarcoidosis. The mounting evidence for the involvement of mycobacterial antigens in the pathogenesis of sarcoidosis, and the specific identification of mycobacterial virulence factors in the BAL and granulomas of patients with sarcoidosis, led us to develop a murine model of pulmonary sarcoidosis. The administration of antigen-coated beads to elicit a delayed-type hypersensitivity reaction in mice has been widely used to study the cellular and molecular mechanisms for the selective recruitment of leukocytes to the airways (13, 16–18). After the embolization (via tail-vein injection) of antigen-coated beads, mice presensitized with the same antigen develop many features similar to human granuloma formation. For example, Mycobacterium bovis purified protein derivative (PPD)–sensitized and challenged mice develop lung granulomas that are largest 4 days after challenge, and that resolve 8 days after challenge (13). Similar to the situation in human pulmonary granulomas, sodA induced the formation of large granulomas throughout murine lungs. The murine model of sarcoidosis possesses the strongest pathologic hallmarks of sarcoidosis: hilar lymphadenopathy, lymphocytic infiltrates, the formation of giant cells, and a Th1 immunophenotype (Figures 2 and and6).6). This model demonstrates an additional strength, in that peptides isolated only from sarcoidosis granulomas are used to incite granulomatous inflammation. This unique Mycobacterium sodA peptide sequence demonstrates the specificity of this peptide for eliciting the histologic features observed in human sarcoidosis.
In contrast, no sensitization (ConNS) or sensitization with IFA alone (ConIFA), followed by challenge with naked beads, led to the formation of small, foreign-body granulomas. To contrast the granuloma formation elicited by sodA to that elicited by a sarcoidosis-unrelated antigen, we used the SEA model developed by Chensue and colleagues (13). As was the case in sodA-treated mice, SEA elicited granulomas throughout the lungs, with the formation of giant cells. In contrast to sodA-elicited granulomas, SEA granulomas were larger, they developed fibrosis, and they were associated with a Th2 immunophenotype (Figures 2 and and6)6) (13). Taken together, these studies show that sodA elicits lung granulomas in mice, with the histologic features of human sarcoidosis.
Next, we determined the leukocyte components of sodA murine lung granulomas by immunohistochemistry, and compared them with those in human sarcoid pulmonary granulomas. Macrophages were the most abundant cell type in the granulomas of both sodA-treated mice and human sarcoidosis. Macrophages are known to be key players in sarcoidosis pathogenesis, because of their ability to present foreign antigens to innate and adaptive cellular components. Other cell types that have been the focus of intense investigation for their possible role in granuloma formation include T lymphocytes. We used CD3, a general marker for both CD4+ and CD8+ T cells. T cells were found in the middle layers of granulomas in both murine and human lungs. In addition, we localized abundant CD4+ T cells to the same region as the CD3+ cells in the sodA-elicited murine lung granulomas. A recent study of sarcoidosis BAL reported on CD4+ and CD8+ T-cell responses to mycobacterial antigens. We found similar responses in the murine model (Figure 5). Although the formation of granulomas was evident in SEA-sensitized mice, T cells isolated from their BAL did not recognize sodA peptides (Figure 5). Finally, B cells were also found sporadically in human and murine sarcoid lung granulomas. These observations are consistent with those of Minami and colleagues (19), Nishiwaki and colleagues (20), and McCaskill and colleagues (21), who studied murine immune pulmonary responses to Propionibacterium acnes. They found peribronchovascular granulomatous inflammation, composed of T and B cells and histiocytes, in response to P. acnes. Although P. acnes was implicated as a putative agent in Japanese and European patients with sarcoidosis, to the best of our knowledge, the molecular and immunologic evidence of its involvement in American patients with sarcoidosis was not previously reported.
Increasing evidence suggests that the degree of inflammation and fibroblast action/proliferation may be dependent on a balance of Th1-like and Th2-like cytokines expressed during the evolution of sarcoidosis (22). A study by Moller and colleagues (23) of BAL cells and fluid from sarcoid patients showed a predominant Th1 cytokine response, with an elevated gene and protein expression of IFN-γ and IL-12, but not IL-4 or IL-10. Moller and colleagues (23) suggested that a chronic deregulation of IL-12 (which induces the production of IFN-γ, and directs T-cell expansion down the Th1 pathway) is the driving force of granuloma formation in sarcoidosis. The importance of these cytokines in the formation of granulomas was shown when IL-12 and IFN-γ knockout mice failed to develop granulomas after being challenged with granulomatous agents (24, 25). The lymphocytes from sodA-treated mice also demonstrated the Th1 immunophenotype seen in patients with sarcoidosis, compared with the Th2 granulomatous inflammation in SEA-treated mice (Figure 6). Other early mediators include TNF-α, IL-1β, and IL-6, which both amplify and maintain the formation of granulomas (26). The flow cytometry of murine BAL fluid demonstrates that no immune responses to microbial antigens occur at baseline, with an appropriate positive response to the positive control, SEB. BAL CD4+ T cells demonstrate antigen-specific responses to sodA-sensitized mice. A Th1 immunophenotype was demonstrated by the secretion of IL-2 and IFN-γ. A lack of recognition of sodA was evident in SEA-sensitized, ConNS, and ConIFA mice, as expected (Figures 5 and and77).
The role of genetics in sarcoidosis pathogenesis demonstrates that MHC Class II alleles are associated with susceptibility to, or protection from, sarcoidosis (5, 27, 28). We were interested in the role of MHC Class II in the antigen presentation of sodA peptides in the murine model. A blockade of MHC Class II inhibited recognition of the sodA peptide, a result similar to what was described the recognition of mycobacterial antigens in patients with sarcoidosis (9). This particular feature of the model will allow us to identify the molecular basis for genetic associations with sarcoidosis outcomes.
Experimental models of granulomatous inflammation, characterized by either a Th1 or Th2 response (22), are also useful in delineating the mechanisms that maintain and resolve chronic granulomatous lung disease (29). For example, in vivo studies of the granuloma formation induced by Mycobacterium infection showed that IFN-γ and TNF-α are necessary for the progression of lesions (13). Moreover, McCaskill and colleagues (21) revealed a Th1 cytokine profile in the BAL fluid and lungs associated with P. acnes–induced granulomatous inflammation in murine lungs. In the present study, both IFN-γ and IL-2 were present in the BAL fluid of sodA-treated mice, consistent with the Th1 cytokine pattern observed in sarcoidosis BAL. On the other hand, the formation of granulomas induced by Schistosoma mansoni was dependent on IL-4 (17, 30). In the SEA model, a shift in the cytokine profile occurred from Th1 to Th2 (i.e., IL-4, IL-5, and IL-10), which eventually resulted in fibrosis (13, 16). The activation of tissue remodeling, including the increased deposition and degradation of extracellular matrix, occurs simultaneously with the inflammatory response.
Thus, this murine model of sarcoidosis contains the gross, histologic, and immunologic features seen in active sarcoidosis. The work to date demonstrates that (1) mice can develop noncaseating granulomas from peptides unique to sarcoidosis; (2) the gross pathology and histology closely parallel those seen in human patients with sarcoidosis; (3) according to cell type and cytokine pattern, the immunology reflects observations in patients with sarcoidosis at presentation; and (4) MHC Class II alleles are also shown to be important in generating immune responses in this model. This model will serve as an important vehicle to facilitate the identification of mechanistic and immunologic contributors to the resolution of sarcoidosis or progression to fibrosis.
The authors thank Nina Volokh, from the Department of Immunology at the Cleveland Clinic; Terri O'Brian, Matt Kohlmann, and Joseph A. Pangrace from the Center for Medical Arts and Photography at the Cleveland Clinic; and Melissa Downing from the Vanderbilt University Medical Center Immunohistochemistry Core Laboratory, for their technical assistance.
This work was supported by National Institute of Health grants HL080953 (C.M.S.), HL069765 (K.O.-R.), HL081538 (D.A.C.), HL083839, and AI065744, and Vanderbilt University grant RR-00095 (W.P.D.).
Originally Published in Press as DOI: 10.1165/rcmb.2009-0350OC on March 26, 2010
Author Disclosure: D.A.C. has received lecture fees from Takeda Corporation (less than $5,000), industry-sponsored grants from the Takeda Corporation ($10,001–$50,000), Actelion ($10,001–$50,000), the Bard Corporation ($50,001–$100,000), Ortho-McNeil ($50,001–$100,000), and the Hollister Corporation ($10,001–$50,000). None of the other authors have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.