Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Apoptosis is a tightly controlled process of cell death that is important for development, host defense, and immune regulation (1). Following the death of unwanted cells, apoptosis can either proceed through silent corpse removal (4), or can elicit acute inflammation (7, 8). However, the factors that determine these opposite outcomes remain largely unknown. The immune system, in particular, relies on apoptosis to achieve homeostasis. Activation-induced cell death (AICD), mediated by Fas/Fas ligand (FasL) interactions, is a mechanism of T lymphocyte apoptosis that is important to maintain tolerance to self antigens, and to extinguish ongoing immune responses (5, 6).

Silicosis is a chronic lung disease characterized by granulomatous and fibrotic lesions due to accumulation of respirable silica mineral particles (9, 10). In experimental models, the lung response to silica deposition recruits host T lymphocytes, as well as macrophages and neutrophils (11, 12), and induces marked enlargement of draining lymph nodes (13), with increased production of type 1 cytokine interferon- in both thoracic lymph nodes (13, 14) and intrapulmonary lymphocytes (15). Therefore, exposure to silica stimulates lymphocyte populations and phagocytic leukocytes in a sustained manner.

Here, we employed murine silicosis to investigate whether exposure to silica promotes apoptosis in the lungs and draining lymph nodes, and whether apoptosis can have pathogenic implications for the host. We found exacerbated FasL- mediated CD4⁺ T cell AICD associated with deficient proliferative T cell responses to T cell receptor (TCR) stimulation in lymph nodes draining silica. In the lung parenchyma, we found large numbers of apoptotic cells in silica-induced nodular infiltrates. In vivo treatment of silica-exposed mice with caspase inhibitors significantly reduced intrapulmonary neutrophil accumulation and lung inflammation. Our results suggest that silica-induced apoptosis is proinflammatory in the lung parenchyma, but leads to immunologic abnormalities in the draining lymph nodes, with potential implications for regional immunity in the lungs.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Antibodies and Reagents

Anti-CD8 monoclonal antibody (mAb) 53.6.7, anti-Fas mAb Jo2, control hamster immunoglobulin G mAb, FITC-labeled anti-CD4 mAb GK 1.5, fluorescein isothiocyanate-labeled anti-Mac-1 mAb M1/70, phycoerythrin-labeled anti-CD8 mAb 53.6.7, PE- labeled anti-B220 mAb 6B2, and anti-FcII/III receptor mAb 2.4G2 were from PharMingen (San Diego, CA). Anti-rat Ig mAb MAR 18.5 was kindly donated by Dr. Ethan Shevach, National Institutes of Health (Bethesda, MD). Anti-TCR mAb H57.597 (16) was a gift from Dr. Maria Bellio, Federal University of Rio de Janeiro. Silica particles, o-phenylene diamine, tritiated thymidine, ethidium bromide, and Sirius Red F3BA were from Sigma Chemical Co. (St. Louis, MO). Caspase inhibitors N-benzyloxycarbonyl-Val-Ala-Asp-(O-methyl)-fluoromethyl ketone (zVAD-fmk), BOC-Asp-(O-methyl)-fluoromethyl ketone (BOC-Asp-fmk) and N-benzylcarbonyl-Phe-Ala-fluoromethyl ketone (zFA-fmk) were from Enzyme Systems Products (Livermore, CA). Other reagents and kits employed included OCT Tissue-Tek (Sakura Finetek Inc., Torrance, CA); Elite Vectastain ABC kit (Vector Laboratories, Burlingame, CA) for immunoperoxidase staining; and TdT-FragEL detection kit (Oncogen Research Products, Cambridge, MA) for in situ TUNEL assay.

Animals

Wild-type BALB/c (BALB.wt) and BALB gld/gld (BALB.gld) mice of both sexes (aged 6-8 wk, weighing 25-30 g) were from the Oswaldo Cruz Institute, Rio de Janeiro. FasL-deficient BALB.gld mice (17) were produced at the National Cancer Institute, NIH (Bethesda, MD) by serially backcrossing the gld gene onto a BALB/c background for 15 generations. All animal work in this study was conducted in accordance with humane institutional guidelines.

Silica Instillation

Mice were anesthetized by sevoflurane and tracheotomyzed. Mice received an intratracheal injection of 50 µl of either sterile saline or silica suspension (SiO₂, particle size: 80% between 1-5 µm; 20 mg/50 µl in sterile saline; Sigma). The lung mechanics and extracellular matrix composition of this model of silicosis have been described (18). Animals lose weight 15 d after silica instillation but recover after 30 d, and enter the chronic stage of the disease. All experiments in the present study were conducted after 15 d of instillation, which is the peak of the inflammatory response.

Cell Proliferation Assay

After 15 d of silica exposure, animals were killed. Thoracic and mesenteric lymph node cells (LNC) from silica-exposed or control saline-exposed mice were obtained and cultured (2 × 10⁵ cells/ml) in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 2 mM glutamine, 5 × 10⁵ M 2-ME, 10 µg/ml gentamycin, sodium pyruvate, Eagle's minimum essential medium nonessential amino acids, and 10 mM Hepes buffer. Proliferation was assessed by uptake of tritiated thymidine. All cultures were treated with phorbol myristate acetate (2 ng/ml) in the presence or absence of anti-TCR mAb (10% vol/vol culture supernatant). Proliferation was assessed after 3 d by uptake of tritiated thymidine. Results are presented as the mean cpm values ± standard error of the mean of triplicate cultures.

T Cell Viability Assay

Partially purified CD4⁺ T cells were obtained by passage of LNC through a nylon wool column, followed by complement-mediated cytotoxicity of cells treated with anti-CD8 plus a second indirect antibody, and cultured as above. CD4⁺ T cells were cultured (3 × 10⁵ cells/ml) with anti-TCR-, anti-Fas, control mAb, or phorbol myristate acetate alone for 48 h at 37°C. Viable cell counts of triplicate cultures were performed by trypan blue exclusion. Results are expressed as percent cell loss, taking the viable cell count in unstimulated (PMA alone) cultures, as control, according to the following formula: % cell loss = 100  [(cell number with treatment) × 100/(cell number in unstimulated cultures)].

DNA Fragmentation Analysis

Apoptosis was detected by release of fragmented DNA into culture supernatants, as previously described (19). Briefly, LNC or CD4⁺ T cells, from silica-exposed and control saline-exposed mice, were cultured (2 × 10⁶ in 1 ml) in the absence or presence of anti-TCR mAb. After 18 h, supernatants were collected and further depleted of contaminating cells by centrifugation. An equal volume (500 µl) obtained from each sample was treated with 0.5% sodium dodecyl sulfate and centrifuged. The mixture was treated with 50% isopropanol/0.5 M NaCl at 20°C overnight, for DNA precipitation. Pellets were washed, allowed to dry in air, and reconstituted with 20 µl TE buffer (Tris 10 mM, ethylenediamine tetraacetic acid 1 mM, pH 7.8). Each DNA solution (3 µl) was applied to 11 × 14 cm horizontal agarose (1.5%) gel and separated by electrophoresis. Gels were stained with 5 µg/ml ethidium bromide and photographed under ultraviolet light.

Flow Cytometry Analysis

Thoracic LNC (10⁶/0.1 ml) from silica-exposed or control mice were directly stained with FITC-labeled anti-CD8 and anti-Mac-1, or PE-labeled anti-CD4 and anti-B220 mAbs, following pretreatment with anti-FcII/III mAb (Fc block). Antibodies were used at 1 µg/10⁶ cells. Cells were acquired (10,000/sample) and analyzed on a B-D Xcalibur flow cytometer (Becton-Dickinson Immunocytometry Systems, San Jose, CA).

Transmission Electron Microscopy and Energy-Dispersive X-Ray Analysis

Thoracic LNC from silica-exposed or control mice were obtained and fixed with 4% paraformaldehyde in 0.1 M cacodylate buffer, pH 7.2 for 2 h at room temperature. Cells were washed and post-fixed for 1 h in 1% osmium tetroxide, 0.8% potassium ferricyanide, 1 mM CaCl₂ in 0.1 M cacodylate buffer, washed in the same buffer, and cell pellets were dehydrated in acetone and embedded in Epon. Thin sections were stained with uranyl acetate/lead citrate and observed in a Zeiss 900 microscope (Carl Zeiss, Inc., Oberkochen, Germany). Energy dispersive X-ray analysis (20) of macrophages bearing cytoplasmic electron-dense inclusions was performed for detection of silicon dioxide (SiO₂). Ultrathin sections (~ 100 nm) were placed on copper grids and analyzed in a JEOL 1200-EX transmission electron microscope operated at 80 kV equipped with a Noran-Voyage analytical system. Control analyses were performed in additional cytoplasmatic and background resin areas.

Immunohistochemistry, Picrosirius Staining, and In Situ Detection of Apoptosis

Animals were killed and the pulmonary circulation was flushed by intracardiac injection of saline, followed by 4% paraformaldehyde and 7.5% sucrose in 0.1 M phosphate buffer. Lung cryostat sections (4-10 µm) were embedded in OCT Tissue-Tek, and mounted on poly-L-lysine-coated glass microscope slides. Activated caspase-3 was detected with anti-activated caspase-3 mAb CM1 (21), kindly donated by Dr. Anu Srinivasan (Idun Pharmaceutics, La Jolla, CA), and developed by Elite Vectastain ABC kit. Negative controls lacked the primary CM1 antibody and gave negative staining. DNA fragmentation associated with apoptosis was detected by terminal deoxynucleotidyl transferase (TdT)- mediated dUTP nick-end labeling (TUNEL), using TdT-FragEL detection kit according to the manufacturer's instructions. Slides were developed with streptavidin-horseradish peroxidase conjugate, counterstained with methyl green, and mounted for photomicroscopy. Negative controls developed without TdT gave negative staining. In addition, both activated caspase-3 and TUNEL-positive and -negative stainings were tested in developing rat retina sections, where apoptosis is induced in specific cellular layers (22). For morphometric analysis of apoptotic cells, profiles stained with either TUNEL or activated caspase-3 were counted at ×1,000 magnification using an eyepiece graticule of 0.0148 mm². Five fields were counted in one section of the lung of each of two animals treated with either saline or silica. In the lungs of silica-treated mice, the fields were classified as containing either high amounts of silica (positive), or evidence of fibrosis without traces of silica (negative), and these two types of fields were quantified separately. For specific staining of collagen fibers, we employed the picrosirius red technique modified for fluorescence microscopy (23). The pulmonary circulation was flushed by intracardiac injection of neutral buffered formalin. Lung paraffin sections (10 µm) were stained by the picrosirius red technique, modified by treatment with phosphomolybdic acid necessary for fluorescence microscopy (23). Slides were observed using a Zeiss confocal laser scanning microscope.

Myeloperoxidase Activity

Neutrophil content in pulmonary parenchyma was evaluated by measuring myeloperoxidase (MPO) activity, as described (24). After removing the bronchoalveolar lavage fluid by three lavages, individual lungs were homogenized in 1.0 ml phosphate-buffered saline-0.1% Triton X100. Homogenized samples were centrifuged at 10,000 × g for 15 min at 4°C, and the supernatants (50 µl) were added to 50 µl substrate solution containing 5 mM o-phenylene diamine in 10 mM citrate buffer, pH 5.0, and 8.8 mM H₂O₂. The reaction was stopped after 15 min with 50 µl H₂SO₄ 4 N, and the absorbance was read at 492 nm. Results shown are from individual mice, and each point represents the mean of triplicate readings.

Administration of Caspase Inhibitors In Vivo

Caspase inhibitors zVAD-fmk, BOC-Asp-fmk, and the inactive compound zFA-fmk were dissolved in dimethylsulfoxide, and further diluted in Hanks' balanced salt solution. A total of 0.75 mg of inhibitor was given to BALB.wt mice over 10 d, divided into three intraperitoneal injections of 0.25 mg/mouse on Days 0, 5, and 10 after silica instillation. At Day 15, animals were killed and analyzed for silicosis induction by cellular composition of BAL fluid, lung MPO activity, and lung histopathology.

Statistics

Data were analyzed by Student's t test for independent samples, using a SigmaPlot for Windows (Version 4.01) package. Differences with a P value of  0.05 were considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Silica Exposure Leads to Accumulation of Macrophages and T and B Lymphocytes in Thoracic Lymph Nodes

We determined the number of macrophages and lymphocyte populations in thoracic lymph nodes of both saline- and silica-treated BALB.wt mice by flow cytometry (FCM). The relative percentages of B cells and CD8⁺ T cells did not change significantly between control and silica-exposed lymph nodes (Figure 1A). However, the percentage of CD4⁺ T cells was reduced and that of macrophages was increased in silica-exposed lymph nodes (Figure 1A). There was a marked enlargement of lymph nodes draining silica, leading to a considerable increase in absolute numbers of macrophages, B cells, CD4⁺, and CD8⁺ T cells (Figure 1B). CD4⁺ and CD8⁺ T cells were also analyzed for expression of the activation marker CD25 by double-staining FCM. Although the relative frequencies of activated CD4⁺CD25⁺ and CD8⁺CD25⁺ T cells did not significantly change in lymph nodes draining silica (not shown), there was a marked increase in the absolute number of activated CD4⁺ and CD8⁺ T cells expressing CD25 (Figure 1C).

View larger version (20K): [in this window] [in a new window]   Figure 1.   Phenotypic analysis of leukocyte populations in lymph nodes draining silica. Relative (A) and absolute (B) cell counts of the indicated subsets in thoracic LNC from silica-exposed (filled bars) or control saline-exposed (open bars) BALB.wt mice. (A) *P < 0.05 compared with untreated control. (B and C) All parameters from silica-treated LN cells gave significant (P < 0.01) differences compared with untreated controls. (C) Absolute number of activated CD4+ and CD8+ T cells expressing CD25. Data represent mean and SE of four animals/group.

Reduced Mitogenic Responses and Increased FasL-Mediated AICD in T Cells from Silica-Exposed Lymph Nodes

LNC from thoracic lymph nodes draining silica were compared with control LNC from saline-treated lymph nodes in an in vitro polyclonal mitogenesis assay using anti-TCR mAb stimulation. Silica-exposed LNC were significantly hyporesponsive to anti-TCR stimulation, compared with control LNC, and to mesenteric LNC from silica-exposed mice (Figure 2A). This deficient mitogenic response could be the result of increased lymphocyte AICD. We therefore assessed induction of DNA fragmentation in LNC following TCR stimulation. TCR stimulation of saline- exposed LNC did not induce release of fragmented DNA in the supernatant (Figure 2B). However, TCR stimulation of silica-exposed LNC induced detectable DNA fragmentation (Figure 2B). Mesenteric LNC from silica-exposed mice did not undergo DNA fragmentation following TCR stimulation (Figure 2B). These results demonstrate exacerbated AICD only in lymph nodes draining silica.

View larger version (45K): [in this window] [in a new window]   Figure 2.   AICD and deficient mitogenesis in LNC exposed to silica. (A) Cell proliferation. Thoracic (open and filled bars) or mesenteric (hatched bars) LNC, from silica-exposed (hatched and filled bars) or control, saline-exposed (open bars) mice, were cultured in the absence or presence of anti-TCR mAb. Proliferation was assessed by [3H]TdR uptake after 3 d in culture. Results are mean ± SE of triplicate cultures. *P < 0.05 compared with positive control. (B) Activation-induced apoptosis. Detection of fragmented DNA in LNC supernatants. Lane 1: molecular weight markers; lanes 2, 4, and 6: supernatants from unstimulated cells (); lanes 3, 5, and 7: supernatants from TCR-stimulated cells (+). Lanes 2 and 3: control saline-exposed thoracic LNC; lanes 4 and 5: silica-exposed thoracic LNC; lanes 6 and 7: silica-exposed mesenteric LNC.

Lymphocyte apoptosis is an ongoing process in lymph nodes draining silica. First, there is a marked increase in spontaneous cell death of CD4⁺ T cells placed in culture (not shown). Furthermore, transmission electron microscopy of cells directly explanted from lymph nodes draining silica demonstrated numerous apoptotic lymphocytes at distinct stages of chromatin condensation (Figure 3A). We also identified apoptotic lymphocytes physically interacting with macrophages bearing cytoplasmatic silica inclusions (Figure 3B). The X-ray spectroscopic analysis of these cytoplasmatic inclusions confirmed that they were indeed silicon dioxide particles (Figure 3C). These data indicate that silica particles are transported to the draining lymph nodes, and that silica-laden macrophages become closely associated with lymphocytes undergoing apoptosis.

View larger version (133K): [in this window] [in a new window]   Figure 3.   Ongoing apoptosis in lymph nodes draining silica. (A) Transmission electron micrograph of LNC freshly explanted from thoracic lymph nodes of silicotic mice. The picture shows numerous apoptotic lymphocytes at distinct stages of chromatin condensation. (B) Transmission electron micrograph of a silica-laden macrophage (M) physically interacting with apoptotic lymphocytes (L). The picture shows electron-dense cytoplasmatic inclusions of silica (arrow). (C) Energy dispersive X-ray analysis of a macrophage cytoplasmatic inclusion (denoted by arrow in B) showing peaks of silicon and oxygen characteristic of SiO2. Additional peaks of copper and carbon derived from grid and biologic material, respectively, and were also present in the analysis of other unrelated cytoplasmic areas. Silicon peak was absent in control areas (not shown). Original magnification: A, ×5,250; B, ×14,000.

We investigated the role of the Fas/FasL molecular pathway in silica-induced AICD. We compared silica-exposed CD4⁺ T cells from either wild-type BALB (BALB.wt) or FasL-deficient BALB.gld (17) mice. TCR stimulation induced AICD in CD4⁺ T cells from wt mice, but failed to kill CD4⁺ T cells from gld mice (Figure 4A). As expected, an agonist anti-Fas mAb (Jo2) induced cell loss in CD4⁺ T cells from both wt and gld mice (Figure 4A). Furthermore, addition of a neutralizing anti-FasL mAb reduced AICD in wt CD4⁺ T cells (not shown). Induction of T cell apoptosis was also analyzed by DNA fragmentation. TCR stimulation induced marked DNA fragmentation over the background in the supernatants of silica-exposed, wt CD4⁺ T cells, but failed to increase production of fragmented DNA in the supernatants of gld CD4⁺ T cells (Figure 4B). These results indicate that silica induces FasL- dependent AICD in CD4⁺ T cells from draining lymph nodes.

View larger version (48K): [in this window] [in a new window]   Figure 4.   AICD in lymph nodes draining silica is FasL-dependent. (A) Percentage of cell loss in purified thoracic lymph node CD4+ T lymphocytes from silica-exposed BALB.wt and BALB.gld mice. Data indicate % cell loss ± SE induced by anti-TCR (solid bars), anti-Fas (shaded bars), or an isotype control mAb (hatched bars). Open bars indicate no cell loss. Viability was measured after 48 h in triplicate cultures. (B) DNA fragmentation in the supernatants of TCR-stimulated thoracic lymph node CD4+ T cells from silica-exposed BALB.wt (lanes 1-4) and BALB.gld (lanes 5-8) mice. Results of two individual mice per group are shown. Purified CD4+ T cells were cultured in the absence () or presence (+) of anti-TCR mAb. *P < 0.05 compared with unstimulated cultures.

Silica-Induced Apoptosis in Lung Inflammatory Infiltrates and the In Vivo Effect of Caspase Inhibitors

Compared with saline-exposed lungs, inflammatory infiltrates of silica-exposed lung parenchyma contained large numbers of apoptotic cells, identified by immunostaining for activated caspase-3, or by in situ TUNEL assay (Figure 5A). There was a strong association between the presence of apoptotic cells in the parenchyma and the presence of silica-laden macrophages, as determined by morphometric analysis (Figure 5B). We then investigated a role of apoptotic cell death upon inflammation. BALB.wt mice were instilled with silica and treated with caspase inhibitors zVAD-fmk (25, 26), BOC-asp-fmk (26), or the control inactive compound zFA-fmk (25, 26), given as a total dose of 0.75 mg divided into three intraperitoneal injections of 0.25 mg at 5-d intervals. Treatment with zVAD-fmk and BOC-asp-fmk reduced neutrophil accumulation in lung parenchyma by 50% (Figure 5C), and the effect was significant (P < 0.01). Treatment with zFA-fmk had no effect on pulmonary inflammation (Figure 5C). Treatment with zVAD-fmk also reduced the numbers of neutrophils in BAL fluid from silica-treated mice (34.3 ± 2.1% neutrophils, n = 3), compared with mice treated with the control compound zFA-fmk (79.0 ± 1.0% neutrophils, n = 3; P < 0.01 for the difference between the two groups). The reduction of pulmonary inflammation was confirmed by histopathologic analysis (not shown) and by assessment of collagen fiber content in pulmonary sections. In the model we employed, collagen fiber content in parenchyma increases 15 d after instillation, and remains at the same levels 30 d after instillation (18). As shown in Figure 5D, treatment with zVAD-fmk reduced the extent of collagen deposition in the parenchyma of silica-treated mice, compared with controls.

View larger version (71K): [in this window] [in a new window]   Figure 5.   Caspase inhibitors block silica induced inflammation. (A) Immunohistochemistry for apoptotic cells. BALB.wt mice were instilled with saline (upper panels) or silica (lower panels). After 15 d, apoptotic cells in lung sections were stained for activated caspase-3 (left panels) or for fragmented DNA by TUNEL assay (right panels). Silica exposure induced large numbers of apoptotic cells in inflammatory infiltrates. Original magnification: ×630. (B) Apoptotic cells locate close to silica particles. Morphometric analysis of the number of apoptotic cells, evaluated either by activated caspase-3 (open bars), or TUNEL staining (filled bars), in microscopic fields that were either negative or positive for the presence of silica particles. (C) Caspase inhibitors block neutrophil accumulation. BALB.wt mice were instilled with saline, silica, or silica plus treatments with zFA-fmk, zVAD-fmk, or BOC-asp-fmk. After 15 d, lung MPO activity was measured after BAL removal. Treatments with zVAD-fmk (P < 0.01, n = 4) and BOC-asp-fmk (P < 0.01, n = 4), but not with zFA-fmk (NS, n = 6), decreased MPO activity. (D) Caspase inhibitor reduces collagen fiber deposits. BALB.wt mice were instilled with saline (PBS), or with silica, and treated with zVAD-fmk (SIL/zVAD), or with control compound zFA-fmk (SIL/zFA). After 15 d of instillation, lung sections were stained for collagen fibers by the picrosirius technique, and examined by confocal fluorescence microscopy. Original magnification: ×200. Note the intense staining for collagen fibers in the parenchyma of mice treated with SIL/zFA, and the decrease in collagen fiber staining of mice treated with SIL/zVAD.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We used a murine model of acute silicosis to demonstrate that silica exposure induces T cell unresponsiveness in draining pulmonary lymph nodes concomitant with exacerbated CD4⁺ T cell AICD. Furthermore, silica induces extensive apoptosis in cells associated with inflammatory infiltrates in the lung parenchyma. Treatment with caspase inhibitors in vivo reduced intrapulmonary neutrophil infiltration, suggesting that caspase activity is proinflammatory in murine silicosis. Therefore, silica-induced apoptosis plays important roles in both lung inflammation and regional immunologic abnormalities associated with silicosis.

Silica-induced immunostimulation is an intriguing phenomenon that has been previously described (11, 12). Massive increase in lymph node size (13) and sustained activation of distinct T lymphocyte populations (13) suggest that silica is mitogenic for T lymphocytes. It has been reported that silica polyclonally activates human T cells in vitro (27). Because T cells only recognize peptide/MHC complexes on antigen-presenting cells (APC), silica must activate T cells indirectly. One possibility is that reactive oxygen intermediates produced by silica, chemically modify self proteins on APC, thus creating neoantigenic determinants. A second possibility is that apoptotic cells produced by exposure to silica become immunogenic for T cells after uptake and processing by APC. Apoptotic cells are rapidly taken up by APC and their antigens can be presented to T cells in the presence of appropriate costimuli (28). Furthermore, it has been proposed that apoptosis generates neoantigenic determinants due to the attack of self proteins by oxidants and activated caspases (29). However, it should be noted that although our TEM data demonstrated physical interactions between apoptotic lymphocytes and macrophages, we did not observe phagocytosis of the dead cells.

Our data demonstrated that lymphocytes from silica-exposed lymph nodes undergo AICD following TCR stimulation. AICD is FasL-dependent, because CD4⁺ T cells from FasL-deficient gld mice did not apoptose in response to TCR stimulation. Moreover, TEM of freshly explanted cells showed that lymphocyte apoptosis is an ongoing process in lymph nodes draining silica. As in the case of T cells chronically activated by antigen (5, 6), it is likely that FasL-mediated AICD is the consequence of sustained lymphocyte activation due to silica exposure. Lymph nodes draining silica have increased numbers of activated lymphocytes, increased rates of spontaneous and activation-induced apoptosis, and mount deficient T cell mitogenic responses. Similar characteristics were described for human T cells from secondary lymphoid organs subject to chronic immunostimulation (30). The presence of these T cell abnormalities raises the possibility that regional pulmonary defenses are compromised in the course of silicosis. FasL-mediated AICD preferentially targets effector Th1 T cells (31), which are important in cell-mediated responses against pathogens. Moreover, contact with apoptotic lymphocytes deactivates macrophages, blocks NO production, and increases the growth of a pathogenic parasite (32). There is increased incidence of mycobacterial infection among silicotic patients (33), suggesting that silicosis induces regional immunosuppression. In this context, it is important to investigate the response of silica-exposed lymph nodes to mycobacterial infection.

Our data showed extensive apoptosis in cells associated with silica-induced inflammatory infiltrates in lung parenchyma. We tested the role of apoptosis by treating silica-exposed mice with caspase inhibitors. Caspase inhibitors reduced both the amplitude of intrapulmonary inflammation and the extent of collagen fiber deposition in the parenchyma, compared with the inactive control compound, zFA-fmk. These results suggest that silica-induced apoptosis is required for eliciting pulmonary inflammation in lung parenchyma. We employed pan-caspase inhibitors (25, 26), thus precluding the assignment of specific roles to any activated caspase in the inflammatory process. However, zFA-fmk is a cysteine proteinase inhibitor inactive for caspases (25, 26), and it had no effect on the extent of silica-induced inflammation. These results suggest the specific involvement of caspases. The anti-inflammatory effect of caspase inhibitors could result either from blockade of apoptosis, or from inhibition of caspase-1 activation, which is required for cleavage of IL-1 and IL-18 precursors (34). In any case, proapoptotic signaling initiated by silica appears to be involved. In agreement, we recently found that silica induces FasL-dependent apoptosis in pulmonary macrophages, and that FasL-deficient mice do not develop pulmonary inflammation following silica exposure (35). Moreover, treatment of wt mice with anti-FasL antibody in vivo prevented pulmonary inflammation elicited by silica (35). Therefore, FasL plays a pivotal role in murine silicosis. There is also indirect evidence for involvement of Fas and FasL in human silicosis (36, 37). Apoptosis can elicit acute inflammation (7, 8) characterized by neutrophil infiltration (38). Macrophages undergoing FasL-mediated apoptosis release potent chemotactic factors for neutrophils (39). Furthermore, FasL induces caspase 1-independent release of IL-1 (7) and IL-18 (8), leading to acute tissue injury.

We demonstrated intense apoptosis in both pulmonary parenchyma and draining lymph nodes in acute silicosis. Except for the requirement of FasL, it is unlikely that the two phenomena are functionally related. Neutrophil infiltration in the parenchyma is a common reaction to other forms of tissue injury (39). However, extensive apoptosis and increased AICD in lymph nodes appear to be unique immunostimulatory properties of silica. The consequences for systemic Th1/Th2 polarization, and for the immune status of the host, remain to be determined. Finally, we did not investigate chronic silicosis in the present study, but it is likely that chronic fibrotic lesions start from acute injury. It will be important to evaluate the role of apoptosis in the chronic fibrotic lesions induced by silica.