Soluble Form of Fas and Fas Ligand in BAL Fluid From Patients With Pulmonary Fibrosis and Bronchiolitis Obliterans Organizing Pneumonia

Soluble Form of Fas and Fas Ligand in BAL Fluid From Patients With Pulmonary Fibrosis and Bronchiolitis Obliterans Organizing Pneumonia* Kazuyoshi Kuwano, MD, PhD; Masayuki Kawasaki, MD, PhD; Takashige Maeyama, MD; Naoki Hagimoto, MD, PhD; Norio Nakamura; Kamon Shirakawa; and Nobuyuki Hara, MD, PhD, FCCP

Study objectives: The Fas-Fas ligand (FasL) pathway is a representative system of apoptosissignaling receptor molecules. We previously described that this pathway may play an important role in the pathogenesis of fibrosing lung diseases. In this study, we hypothesized that soluble form of Fas (sFas) and FasL (sFasL) may also be associated with this disorder. Measurements and results: We measured sFas and sFasL levels in BAL fluid (BALF) from patients with idiopathic pulmonary fibrosis (IPF), interstitial pneumonia associated with collagen vascular diseases (CVD-IP), and bronchiolitis obliterans organizing pneumonia (BOOP), using enzymelinked immunosorbent assay. BALF from all patients was obtained before prednisolone therapy. sFasL levels were relatively increased in IPF patients (p ⴝ 0.084), and significantly increased in CVD-IP patients (p < 0.05) and BOOP patients (p < 0.05), compared with control subjects. BALF sFasL levels were elevated in the IPF or CVD-IP subgroups with an indication for prednisolone therapy, compared with those without an indication for therapy. The BALF sFasL level in IPF patients was correlated with the number of total cells and lymphocytes. The BALF sFasL level in BOOP patients was relatively or significantly correlated with the number of total cells or lymphocytes, respectively. The BALF sFas level was significantly increased in BOOP patients, but not in IPF or CVD-IP patients. Conclusions: We conclude that BALF sFasL levels may be associated with the accumulation of inflammatory cells and reflect the degree of lymphocyte alveolitis in IPF. The elevation of sFasL may be associated with the deterioration of IPF and CVD-IP. The elevation of the BALF sFas level may abrogate the cytotoxicity of FasL in BOOP patients, which may be associated with better prognosis of BOOP, compared with IPF or CVD-IP. (CHEST 2000; 118:451– 458) Key words: bronchiolitis obliterans organizing pneumonia; collagen vascular diseases; Fas; Fas ligand; idiopathic pulmonary fibrosis Abbreviations: BALF ⫽ BAL fluid; BOOP ⫽ bronchiolitis obliterans organizing pneumonia; CRP ⫽ C-reactive protein; CVD-IP ⫽ interstitial pneumonia associated with collagen vascular diseases; Dlco ⫽ diffusion capacity of the lung for carbon monoxide; ELISA ⫽ enzyme-linked immunosorbent assay; FasL ⫽ Fas ligand; IPF ⫽ idiopathic pulmonary fibrosis; mFasL ⫽ membrane-bound FasL; LDH ⫽ lactate dehydrogenase; sFas ⫽ soluble Fas; sFasL ⫽ soluble Fas ligand; VC ⫽ vital capacity

ligand (FasL) system is the represenT hetativeFas-Fas system of apoptosis-signaling receptor molecules. Fas is a cell surface protein consisting of *From the Research Institute for Diseases of the Chest (Drs. Kuwano, Kawasaki, Maeyama, Hagimoto, and Hara), Graduate School of Medical Sciences, Kyushu University, Fukuoka; and Bioscience Laboratory Research Center (Mr. Nakamura and Mr. Shirakawa), Mochida Pharmaceutical Company, LTD, Tokyo, Japan. Manuscript received October 11, 1999; revision accepted March 13, 2000. Correspondence to: Kazuyoshi Kuwano, MD, PhD, Research Institute for Diseases of the Chest, Faculty of Medicine, Kyushu University, 3–1-1 Maidashi, Higashiku, Fukuoka, 812 Japan; e-mail: [email protected]

319 amino acids with a single transmembrane domain, belonging to the nerve growth factor receptor/ tumor necrosis factor receptor family.1 FasL is a type II membrane protein that belongs to the tumor necrosis factor family.2 FasL is dominantly expressed in activated T cells, whereas Fas is expressed in various cells and tissues.3 Analysis of mice lacking Fas or FasL have indicated that FasL is one of the major effector molecules of cytotoxic T lymphocytes, such as CD8⫹ T cells, CD4⫹ T helper 1-type T cells, and natural killer cells.3–5 The role of cytotoxic T lymphocytes is to remove virally infected or cancerous cells to prevent the spreading of viruses or CHEST / 118 / 2 / AUGUST, 2000

451

cancer cells in animals. However, when this system overfunctions, it causes tissue destruction. Although the etiology of idiopathic pulmonary fibrosis (IPF) is still uncertain, it is believed that the initial lesion prior to the formation of fibrosis is probably alveolitis, which is characterized by the loss of type I epithelial cells, and type II pneumocyte hyperplasia.6 – 8 Interstitial pneumonia associated with collagen vascular diseases (CVD-IP) is generally indistinguishable clinically and pathologically from IPF. Bronchiolitis obliterans organizing pneumonia (BOOP) is characterized pathologically by inflammation of the bronchial wall, obliteration of the bronchiolar lumen by connective tissue, and patchy areas of organizing pneumonia. Although the clinical symptoms, pulmonary function tests, and radiographic findings of BOOP overlap with those of IPF, patients with BOOP usually achieve a complete recovery after corticosteroid therapy and subsequently have a much better prognosis, compared with patients with IPF.9 Molecular cloning and nucleotide sequence analysis have revealed a human Fas messenger RNA variant that encodes a soluble Fas (sFas) molecule lacking the transmembrane domain because of the deletion of an exon encoding this region.10 sFas blocks apoptosis in cells by inhibiting the binding of FasL to Fas on the cell membrane.10,11 Membrane-bound FasL (mFasL) is converted to soluble Fas ligand (sFasL) by a matrix metallopro-

teinase-like enzyme.12,13 Reports indicate that sFasL is present in the supernatant of activated human peripheral blood T cells3 and is increased in large granular lymphocytic leukemia and natural killer cell leukemia.11,13 Since the patients affected by these leukemia also suffer from hepatitis and neutropenia, it has been postulated that sFasL may cause systemic tissue damage. We previously demonstrated that the expression of Fas was upregulated in bronchiolar and alveolar epithelial cells and that FasL protein was upregulated mainly in infiltrating T lymphocytes in lung tissues from patients with IPF.14 Therefore, we hypothesized that Fas-FasL pathway may be involved in the pathogenesis of IPF. Since sFas and sFasL have not been studied in IPF, CVD-IP, and BOOP, we examined BAL fluid (BALF) levels of sFas and sFasL in patients with these disorders. Materials and Methods Study Population The study population, sex, age, smoking history, lung functions, and BAL and serum parameters are shown in Table 1. Table 2 shows characteristics of the IPF and CVD-IP subgroups, which were classified by the clinical indication for prednisolone therapy at the time of BAL. BALF was obtained before treatment. Patients with acute exacerbation were treated with prednisolone. Acute exacerbation was defined as increased dyspnea (an increase of more than one grade in the Hugh-Johns classification

Table 1—Characteristics of the Study Populations* Parameters Patients, No. Sex, male/female Age, yr Smoker/nonsmoker, No. Lung functions, % predicted VC Dlco Pao2, mm Hg BAL parameters Total cell count, ⫻ 105/␮L Macrophages, % Lymphocytes, % Neutrophils, % Eosinophils, % CD4/CD8 Serum parameters LDH, U/L (range, 261 to 483) CRP, mg/dL (normal ⬍ 0.2) WBC, ␮L (range, 4 to 9 ⫻ 103/L)

IPF 33 25/8 65 ⫾ 6 21/12

CVD-IP 21 4/17 59 ⫾ 9 5/16

BOOP

Control

9 3/6 64 ⫾ 10 5/4

13 8/5 58 ⫾ 14 9/4

74 ⫾ 20‡ 62 ⫾ 14 75 ⫾ 9†

75 ⫾ 18 68 ⫾ 26 77 ⫾ 6

93 ⫾ 27 86 ⫾ 14 78 ⫾ 8

98 ⫾ 15 NE 85 ⫾ 5

4.9 ⫾ 4.4 62 ⫾ 16‡ 22 ⫾ 16㛳 11 ⫾ 10㛳 4 ⫾ 5§ 1.8 ⫾ 1.4

5.2 ⫾ 3.5 56 ⫾ 20‡ 28 ⫾ 18‡ 7.7 ⫾ 9§ 7.1 ⫾ 9§ 1.1 ⫾ 1.4

3.9 ⫾ 1.9 53 ⫾ 18‡ 39 ⫾ 14㛳 4 ⫾ 2‡ 4 ⫾ 7‡ 1.6 ⫾ 1.5

3.4 ⫾ 1.9 92 ⫾ 5 7⫾4 1⫾1 1⫾2 2.1 ⫾ 1.2

487 ⫾ 109㛳 1.9 ⫾ 4.6 7.3 ⫾ 2.2

587 ⫾ 298§ 0.9 ⫾ 2.1 6.9 ⫾ 2.2

377 ⫾ 61 2.2 ⫾ 3.1 7.1 ⫾ 1.5

335 ⫾ 50 0.0 ⫾ 0.0 6.5 ⫾ 2.2

*Data are presented as mean ⫾ SD unless otherwise indicated. NE ⫽ not examined. †Less than control subjects, p ⬍ 0.05. ‡Less than control subjects, p ⬍ 0.01. §More than control subjects, p ⬍ 0.05. 㛳More than control subjects, p ⬍ 0.01. 452

Clinical Investigations

Table 2—Characteristics of the IPF and CVD-IP Subgroups* IPF Characteristics Patients, No. Sex, male/female Age, yr Smoker/nonsmoker, No. Lung functions, % predicted VC Dlco Pao2, mm Hg BAL parameters Total cell count, ⫻105/␮L Macrophages, % Lymphocytes, % Neutrophils, % Eosinophils, % CD4/CD8 Serum parameters LDH, U/L (range, 261–483) CRP, mg/dL (normal, ⬍ 0.2) WBC, ␮L (range, 4 to 9 ⫻ 103 ␮L)

⫺

CVD-IP ⫺

⫹

6 2/4 61 ⫾ 8 3/3

10 1/9 59 ⫾ 7 2/8

11 3/8 60 ⫾ 10 3/8

76 ⫾ 19 61 ⫾ 15 76 ⫾ 9

52 ⫾ 6‡ NE 65 ⫾ 11‡

84 ⫾ 16 83 ⫾ 16 78 ⫾ 6

67 ⫾ 17 44 ⫾ 17§ 76 ⫾ 6

4.2 ⫾ 3.1 66 ⫾ 11 18 ⫾ 11 11 ⫾ 11 5⫾5 1.8 ⫾ 1.3

8.4 ⫾ 7.8‡ 40 ⫾ 17† 46 ⫾ 19† 10 ⫾ 8 5⫾3 0.9 ⫾ 0.8

4.1 ⫾ 2.2 60 ⫾ 22 27 ⫾ 21 10 ⫾ 11 4⫾7 1.1 ⫾ 1.0

6.3 ⫾ 4.3 51 ⫾ 18 32 ⫾ 13 6⫾7 11 ⫾ 10 1.0 ⫾ 1.8

463 ⫾ 89 0.9 ⫾ 1.2 7.3 ⫾ 2.2

603 ⫾ 132† 7.0 ⫾ 9.5† 8.0 ⫾ 3.5

441 ⫾ 111 0.2 ⫾ 0.3 5.3 ⫾ 1.4

700 ⫾ 352 1.5 ⫾ 2.7 8.3 ⫾ 1.8§

27 23/4 66 ⫾ 7 18/9

⫹

*Data are presented as mean ⫾ SD unless otherwise indicated. See Table 1 for abbreviation. †Significant difference between subgroup IPF (⫺) and IPF (⫹), p ⬍ 0.001. ‡Significant difference between subgroup IPF (⫺) and IPF (⫹), p ⬍ 0.05. §Significant difference between CVD-IP (⫺) and CVD-IP (⫹), p ⬍ 0.01.

system), or a decrease in Pao2 (⬎ 10 mm Hg in the same condition), accompanied with remarkable exacerbation of the findings on chest radiograph or CT during the prior 2 months. All control subjects were healthy volunteers. The diagnosis of diseases was established by a combination of medical history, physical examination, laboratory tests, high-resolution CT scans, chest radiographs, pulmonary function tests, and the results of pulmonary biopsies, according to previously described criteria.6 –9 In 10 of 33 patients with IPF, the diagnosis of IPF was confirmed by thoracoscopic lung biopsy. In BOOP patients, at least three specimens were obtained by transbronchial biopsy. Two patients had histologic findings of BOOP, and seven patients had organizing pneumonia with mononuclear cell infiltration in the interstitium. In all cases of BOOP, BAL cytology and phenotypical profile were consistent with a typical BAL profile of BOOP. All BOOP patients responded dramatically to prednisolone treatment. In CVD-IP patients, there were 12 cases of rheumatoid arthritis, 6 cases of polymyositis/dermatomyositis, 2 cases of Sjo¨gren’s syndrome, and 1 case of progressive systemic sclerosis. A transbronchial lung biopsy was performed in all patients to rule out competing diagnoses. In all patients, a current infection with bacteria, mycobacteria, or fungi was excluded by negative cultures of BALF and biopsy specimens. Analysis of BAL Cells BAL was performed using a total of 150 mL of sterile physiologic saline solution. The recovered fluid was filtered through a single layer of gauze to remove mucus. Cells in the lavage fluid were counted using a hemocytometer. Differential counts were performed on a total of 100 cells stained with Wright and Giemsa stain. Pulmonary Function Pulmonary function was assessed by vital capacity (VC), diffusion capacity for carbon monoxide (Dlco), and Pao2. Dlco was

not measured in control subjects. The lung function tests were performed according to American Thoracic Society criteria.15,16 sFasL and sFas Measurement by Enzyme-Linked Immunosorbent Assay BALF sFasL levels were measured with a newly developed enzyme-linked immunosorbent assay (ELISA) system. In brief, microtiter plates were coated with an anti-human FasL antibody (F918 –20-2 antibody). The purified human recombinant FasL produced in pichia pastoris was used as a standard. After an incubation for 2 h at 37°C, the wells were washed with saline solution containing 0.05% Tween 20. The wells were incubated with the poly peroxidase-labeled anti-human FasL antibody F919 –9-18 for 1 h at 37°C, and washed with saline solution containing 0.05% Tween 20 and distilled water. The wells were incubated with tetramethylbenzine for 20 min. Peroxidase activity was developed in proportion to the amount of sFasL. The reaction was stopped by adding 1 normal H2SO4. The color generated was determined by measuring the optimal density at 450 nm in a spectrophotometric microtitor plate reader. The limit of detection (negative control, 0 pg/mL ⫹ 2 SD) was 5 pg/mL in BALF. Intra-assay and inter-assay coefficient variations were ⬍ 2.4 to 11.7% and ⬍ 8.3 to 15.7%, respectively. We performed all assay in duplicate, and the mean of two data was determined for individual samples. BALF sFas levels were also measured by a sandwich ELISA Kit (Medical & Biological Laboratories; Nagoya, Japan). This ELISA kit is capable of detecting the secretory form of sFas (not the proteolytic membrane form). As the principle of measurement, the assay uses Fas antibodies against two different epitopes: one is polyclonal antibody that recognizes the intracellular domain (No. 305–309 amino acid), and the other is a monoclonal antibody that recognizes the extracellular domain (No. 110 –120 amino acid). In order to measure BALF sFas levels, we determined the limit of detection of BALF sFas. The limit of detection (negative control, 0 pg/mL ⫹ 2 SD) in BALF CHEST / 118 / 2 / AUGUST, 2000

453

was 8 pg/mL. We performed all assays in duplicate, and the mean of two data was determined for individual samples. Serologic Parameters Total serum lactate dehydrogenase (LDH; reference range, 280 to 440 U/L) and C-reactive protein (CRP; reference range, ⬍ 0.2 mg/dL) were analyzed by standard methods. WBC counts were analyzed by a hematology analyzer (reference range for adults, 4.0 to 9.0 ⫻ 109 cells/L). Statistical Analysis Data are expressed as mean ⫾ SD. Comparisons between diseases and control subjects, and those between subgroups of patients were assessed by one-way analysis of variance using Abacus Concepts Statview J4.5 package (Abacus Concepts; Berkeley, CA). Correlations between different parameters were determined by Spearman’s rank correlation coefficient. All p values ⬍ 0.05 were regarded as significant.

Results Clinical Parameters and sFas/sFasL Table 1 showed that there was a significant decrease in percent VC (p ⬍ 0.01) and Pao2 (p ⬍ 0.05) in IPF patients, compared with control subjects. There was a significant increase in serum LDH in IPF (p ⬍ 0.01) and CVD-IP (p ⬍ 0.05) patients, compared with control subjects. In BALF, there were significant increases in the percentage of lymphocytes, neutrophils, and eosinophils in all groups

compared with control subjects. There was a significant decrease of percentage of macrophages in all groups. There was no significant difference in the ratio of CD4 to CD8 between diseases and control subjects. Table 2 shows characteristics of the IPF and CVD-IP subgroups with or without a clinical indication for prednisolone therapy at the time of BAL. There were significant differences in percent VC, Pao2, LDH, and CRP between IPF subgroups. In BALF, there were significant increases in the number of total cells and the percentage of lymphocytes in the IPF subgroup with a clinical indication for prednisolone therapy, compared with another subgroup without an indication for therapy. BALF sFas and sFasL Levels BALF sFas levels (mean ⫾ SD) were 21 ⫾ 54 pg/mL in control subjects. BOOP patients exhibited significant higher levels of sFas (215 ⫾ 215 pg/mL), compared with control subjects (p ⬍ 0.05). There was no significant difference in BALF sFas levels between IPF or CVD-IP patients and control subjects (Fig 1). There was also no significant difference in BALF sFas levels between the IPF or CVD-IP subgroups with and without indication for therapy (Fig 2). BALF sFasL levels (mean ⫾ SD) were 4.7 ⫾ 6.0 pg/mL in control subjects. CVD-IP and BOOP

Figure 1. BALF sFas levels in patients with IPF, CVD-IP, BOOP, and control subjects. The mean is indicated by the horizontal bar; each circle represents one individual. There is a significant difference between BOOP patients and control subjects. 454

Clinical Investigations

Figure 2. BALF sFas levels in patients with IPF and CVD-IP subgroups. The mean is indicated by the horizontal bar; each circle represents one individual. There was no significant (NS) difference between IPF or CVD-IP subgroups as indicated.

patients exhibited significant higher levels of sFasL, 32.9 ⫾ 42.5 pg/mL (p ⬍ 0.05) and 47.0 ⫾ 49.2 pg/mL (p ⬍ 0.05), respectively, compared with control subjects. BALF sFasL levels were relatively higher in IPF patients (40.1 ⫾ 69.1 pg/mL; p ⫽ 0.084) than control subjects (Fig 3). There was a significant difference in BALF sFasL levels between IPF (p ⬍ 0.05) or CVD-IP subgroups (p ⬍ 0.01) with and without a clinical indication for prednisolone therapy (Fig 4).

Figure 4. BALF sFasL levels in patients with IPF and CVD-IP subgroups. The mean is indicated by the horizontal bar; each circle represents one individual. BALF sFasL levels were significantly elevated in patients with IPF or the CVD-IP subgroup with an indication for therapy, compared with those without an indication, respectively.

Correlation Between BALF sFasL Levels and the Number of BAL Cells There was a significant correlation between the total number of BAL cells and BALF sFasL levels in IPF patients (r ⫽ 0.45; p ⬍ 0.05). BALF sFasL levels in IPF patients were also significantly correlated with the number (r ⫽ 0.57; p ⬍ 0.005) and percentage (r ⫽ 0.64; p ⬍ 0.001) of lymphocytes in BALF (Fig 5). BALF sFasL levels in BOOP patients tended to correlate with the total number of BAL cells (r ⫽ 0.65; p ⫽ 0.087). BALF sFasL levels in BOOP patients were significantly correlated with the number (r ⫽ 0.74; p ⬍ 0.05) but not percentage (r ⫽ 0.55; p ⫽ 0.14) of lymphocytes in BALF. There was no correlation between BALF sFasL levels in CVD-IP patients and the number or percentage of BAL cells. There was no correlation between BALF sFas levels and the number of BAL cells in any group. Discussion

Figure 3. BALF sFasL levels in patients with IPF, CVD-IP, BOOP, and control subjects. The mean is indicated by the horizontal bar; each circle represents one individual. The BALF sFasL levels in IPF patients were relatively elevated, and levels in CVD-IP and BOOP patients were significantly elevated as indicated, compared with control subjects.

This study is the first to demonstrate the levels of sFas and sFasL in BALF in IPF patients. Although the mechanism of alveolitis and fibrosis in IPF is unclear, autoimmune mechanism against bronchiolar and alveolar epithelial cells, especially T cellmediated immunity, is considered to play an important role in the pathogenesis. There was a relative CHEST / 118 / 2 / AUGUST, 2000

455

Figure 5. Correlation between the BALF sFasL level and the number of BAL cells in patients with IPF. There were significant correlations between the BALF sFasL level and the number of total cells, the number of lymphocytes, or lymphocytes in percentage of BAL cells as indicated.

increase in BALF sFasL levels in IPF patients compared with control subjects, and the BALF sFasL level in the IPF subgroup with a clinical indication for prednisolone therapy was significantly increased, compared with those without an indication for therapy. There were significant decreases in 456

percent VC and Pao2 and significant increases in LDH and CRP in the IPF subgroup with an indication for therapy, compared with those without it. Furthermore, there was a significant correlation between the BALF sFasL level and the number of total cells or the number and percentage of lymphocytes in IPF patients. Therefore, the increase of the BALF sFasL level may reflect the accumulation of inflammatory cells in lung tissues and the disease exacerbation in IPF patients. Although the number of cases are small (n ⫽ 5), sFas but not sFasL levels in BALF were increased in bacterial pneumonia (data not shown). Therefore, the increase of sFasL in BALF may be associated with alveolitis, which subsequently leads to fibrosis. Recently, Hashimoto et al17 demonstrated that BALF sFasL levels in patients in the acute phase of septic ARDS are significantly increased, compared with patients in the late phase of septic ARDS or control subjects. They also showed that Fas messenger RNA expression in BAL cells was significantly increased in the acute phase compared with the late phase of septic ARDS, and FasL messenger RNA expression in BAL cells was detected in the acute phase but not in the late phase of septic ARDS. In contrast to ARDS, Fas or FasL messenger RNA in BAL cells was not detectable in control subjects. Therefore, the reason why so many of each group have low levels of sFas and sFasL may be that these molecules are inducible molecules, and most of the patients were not in the acute phase of the disease. mFasL is converted to sFasL by a matrix metalloproteinase-like enzyme.12,13 When the human recombinant sFasL was injected into mice, a large amount of sFasL was necessary to exhibit its lethal effect, compared with anti-Fas antibody.18 Since human Jurkat T cells, which are highly sensitive to apoptosis induced by anti-Fas antibody, remained viable at sFasL concentrations ⬎ 1 ␮g/mL, it was suggested that sFasL is only poorly active in contrast to mFasL.19 Furthermore, it was described that the metalloproteinase inhibitor potentiates concanavalin A-induced hepatitis in mice.20 These results suggested that mFasL is the functional form of FasL in vivo, and that mFasL expressed on the cell surface kills the local target cells and is then downregulated by shedding.21,22 This mechanism may prevent the killing of the healthy bystander cells. We previously demonstrated that FasL was detected in predominantly infiltrating T lymphocytes in lung tissues from patients with IPF, but was not detectable in normal lung parenchyma, using immunohistochemistry.14 Therefore, mFasL in activated T lymphocytes may damage the lung tissues in IPF patients. In this study, patients with an indication for therapy, which means that these patients had an Clinical Investigations

acute exacerbation of the disease, demonstrated high levels of BALF sFasL. Although the shedding mFasL converted to sFasL may have a protective effect on lung tissues against the cytotoxicity of mFasL, these results suggest that high levels of sFasL in BALF reflect the severity of inflammation and exacerbation of the disease. Although sFasL fundamentally downregulates the cytotoxicity of mFasL, a high concentration of sFasL could damage cells in which Fas is highly expressed. Since we previously demonstrated that Fas was upregulated in bronchiolar and alveolar epithelial cells in lung tissues from patients with IPF,14 these epithelial cells may be sensitive to sFasL as well as mFasL. Additionally, BALF sFas levels were not significantly increased in IPF patients compared with control subjects, and were not correlated with the number of inflammatory cells, whereas sFasL levels were correlated with the number of lymphocytes. Therefore, sFas levels in lung tissues may be not sufficient to protect epithelial cells from the cytotoxic effect of FasL in IPF patients. In contrast to IPF or CVD-IP patients, BALF sFas levels in patients with BOOP were significantly higher compared with control subjects, along with BALF sFasL levels. Patients with BOOP have a much better prognosis than patients with IPF, and most patients with BOOP achieve a complete clinical and functional recovery without remaining pulmonary fibrosis after steroid treatment. We speculate that epithelial cell damage induced by mFasL may be abrogated by both sFas and the shedding of mFasL in BOOP. The regulated increase of sFasL and sFas may protect epithelial cells from cytotoxic effect of mFasL, which may be associated with better prognosis of BOOP than that of IPF. There was a significant increase in BALF sFasL levels in CVD-IP patients compared with control subjects. As in IPF patients, Fas was upregulated in bronchiolar and alveolar epithelial cells in CVD-IP patients,23 and BALF sFas levels were not significantly increased compared with control subjects. Therefore, the shedding of mFasL and the production of sFas may be insufficient to protect epithelial cells from cytotoxicity of FasL in CVD-IP patients as well as IPF patients. In contrast to IPF and BOOP patients, there was no correlation between sFasL levels and the number of total cells or lymphocytes in BALF. The group of CVD-IP is heterogeneous. There was no significant difference in the number or percentage of BAL cells between the patients with an indication for prednisolone therapy. Therefore, there may be some differences in the mechanism of lymphocyte accumulation to the lung tissues and of the expression of sFasL in inflammatory cells between the CVD-IP and IPF groups.

In conclusion, we measured sFas and sFasL levels in BALF obtained from patients with IPF, CVD-IP, and BOOP. The increased levels of BALF sFasL may reflect the accumulation of inflammatory cells and the exacerbation of the disease in IPF patients. The shedding of mFasL to sFasL and the production of sFas may have an important role in the protection of epithelial cell damage from the cytotoxicity of FasL. Although the relationship between sFas and mFasL and FasL appears to be complex, sFas and sFasL may have an important role in the pathophysiology of fibrosing lung diseases.

References 1 Itoh N, Yonehara S, Ishii A, et al. The polypeptide encoded by the cDNA for human cell surface antigen Fas can mediate apoptosis. Cell 1991; 66:233–243 2 Suda T, Takahashi T, Golstein T, et al. Molecular cloning and expression of the Fas ligand, a novel member of the tumor necrosis factor family. Cell 1993; 75:1169 –1178 3 Suda T, Okazaki T, Naito Y, et al. Expression of the Fas ligand in cells of T cell lineage. J Immunol 1995; 154:3806 –3813 4 Lowin B, Hahne M, Mattmann C, et al. Cytotoxic-T cell cytotoxicity is mediated through perforin and Fas lytic pathways. Nature 1994; 370:650 – 652 5 Kagi D, Vignaux F, Ledermann B, et al. Fas and perforin pathways as major mechanisms of T cell-mediated cytotoxicity. Science 1994; 265:528 –530 6 Liebow A. Definition and classification of interstitial pneumonias in human pathology. In: Baset F, George R, eds. Progress in respiration research. Vol. 8. New York, NY: Karger, 1975; 1–33 7 Crystal RG, Bitterman PB, Rennard SI, et al. Interstitial lung diseases of unknown cause: disorders characterized by chronic inflammation of the lower respiratory tract (first of two parts). N Engl J Med 1984; 310:154 –166 8 Crystal RG, Fulmer JD, Roberts WC, et al. Idiopathic pulmonary fibrosis: clinical, histologic, radiographic, physiologic, scintigraphic, cytologic, and biochemical aspects. Ann Intern Med 1976; 85:769 –788 9 Epler GM, Colby TV, McLoud TC, et al. Bronchiolitis obliterans organizing pneumonia. N Engl J Med 1985; 312: 152–158 10 Cheng J, Zhou T, Liu C, et al. Protection from Fas-mediated apoptosis by a soluble form of the Fas molecule. Science 1994; 263:1759 –1762 11 Tanaka M, Suda T, Haze K, et al. Fas ligand in human serum. Nature Med 1996; 2:319 –322 12 Kayagaki N, Kawasaki A, Ebata T, et al. Metalloproteinasemediated release of human Fas ligand. J Exp Med 1995; 182:1777–1783 13 Tanaka M, Suda T, Takahashi T, et al. Expression of the functional soluble form of Fas ligand in activated lymphocytes. EMBO J 1995; 14:1129 –1135 14 Kuwano K, Miyazaki H, Hagimoto N, et al. The involvement of Fas-Fas ligand pathway in fibrosing lung diseases. Am J Respir Cell Mol Biol 1999; 20:53– 60 15 American Thoracic Society. Standardization of spirometry: 1994 update. Am J Respir Crit Care Med 1995; 152:1107– 1136 16 American Thoracic Society. Single-breath carbon monoxide CHEST / 118 / 2 / AUGUST, 2000

457

diffusing capacity (transfer factor): recommendations for a standard technique; 1995 update. Am J Respir Crit Care Med 1995; 152:2185–2198 17 Hashimoto S, Kobayashi A, Kooguchi K, et al. Upregulation of two death pathways of perforin/granzyme and FasL/Fas in septic acute respiratory distress syndrome. Am J Respir Crit Care Med 2000; 161:237–243 18 Tanaka M, Suda T, Yatomi T, et al. Lethal effect of recombinant human Fas ligand in mice pretreated with propionibacterium acnes. J Immunol 1997; 158:2303–2309 19 Schneider P, Holler N, Bodmer J, et al. Conversion of membrane bound Fas (CD95) ligand to its soluble form is associated with down regulation of its proapoptotic activity and loss of liver injury. J Exp Med 1998; 187:1205–1213

458

20 Solorzano CC, Ksontini R, Pruitt JH, et al. Involvement of 26-kDa cell-associated TNF alpha in experimental hepatitis and exacerbation of liver injury with a matrix metalloproteinase inhibitor. J Immunol 1997; 158:414 – 419 21 Tanaka M, Itai T, Adachi M, et al. Down regulation of Fas ligand by shedding. Nature Med 1998; 4:31–36 22 Suda T, Hashimoto H, Tanaka M, et al. Membrane Fas ligand kills human peripheral blood T lymphocytes and soluble Fas ligand blocks the killing. J Exp Med 1997; 186:2045–2050 23 Kunitake R, Kuwano K, Miyazaki H, et al. The expression of p53, p21 (Waf1/Cip1/Sdi1) and Fas antigen in interstitial pneumonia associated with collagen vascular diseases and granulomatous lung diseases. Eur Respir J 1999; 13:1329 – 1337

Clinical Investigations