Differential Proliferative Characteristics of Alveolar Fibroblasts in Interstitial Lung Diseases: Regulative Role of IL-1 and PGE2

Fibroblasts (Fb) from patients with sarcoidosis (SA) and hypersensitivity pneumonitis (HP) exhibited a lower proliferative capacity compared with Fb obtained from control (CO) and diffuse interstitial fibrosis patients (DIF). Proliferation of Fb from SA or lip patients was suppressed by autologous LPS-stimulated alveolar macrophages (AM) supernatants but not by those from CO patients. Similarly, alveolar macrophages (AM) derived supernatant, obtained from CO, did not suppress the proliferation of SA and HP Fb. AM from SA and HP patients secreted higher amounts of IL-1α and β compared with controls and compared with Fb from SA and HP patients. Steady levels of IL-1α and βmRNA were expressed in unstimulated and stimulated cultures. Fb from SA and HP patients could be stimulated by LPS to secrete significantly higher levels of PGE2 than those detected in supernatants from LPS stimulated Fb of DIF patients. Only the proliferation of Fb from SA and HP patients was sensitive to amounts of IL-1 equivalent to those detected in the lung of these diseases. As SA and HP are two diseases where irreversible deterioration occurs in only 20% of the patients, we hypothesize that mediators in the lung may modulate Fb proliferation. IL-1 of AM origin and PGE2 of Fb origin secreted at high levels, may be candidates for this suppression because it was abrogated by anti IL-1β and indomethacin.


Introduction
One of the most prominent histologic features of granulomatous and interstitial diseases in the lung is the close proximity of mononuclear cells with fibroblasts or the matrix secreted by them. Several studies characterized mononuclear cell-derived factors that stimulate 1," or inhibit 3,4 fibroblast growth and secretory functions in vitro. Other researchers have shown that cytokines are released spontaneously by macrophages isolated from lungs of animals exposed to a variety of stimulants, including cytotoxic drugs and mineral dusts, 5,6 as well as from patients with idiopathic pulmonary fibrosis (IPF) and sarcoidosis (SA). Furthermore, alveolar lining fluid collected by pulmonary lavage from patients with pulmonary fibrosis has been shown to contain fibrogenic cytokines, such as TGF and TNFoq 1 pointing to the role of cytokines in vivo. These data suggest that cytokines are present in situ and may .mediate. the pathological manifestations of interstitial lung diseases. Lately, efforts have been directed towards the identification of macrophage-derived factors which affect fibroblast growth-function, while fewer studies were oriented towards the investigation of the immunoregulatory and pro-inflammatory role of fibroblasts. Recently, it was shown that murine fibroblasts, stimulated by cytokines and LPS, are able to generate IL-Iz activity ha2 and that rIL-1 and TNF stimulate normal adult human lung fibroblast to accumulate not to secrete IL-1 [3. 3 Moreover, other studies demonstrated that fibroblast strains secrete inflammatory mediators, such as prostaglandin E,. (PGE2), interleukin 6 (IL-6), interleukin 8 (IL-8), monocyte chemoattractant peptide (MCP-1) and colony-stimulating factor. [4][5][6][7][8][9] As we have previously shown that alveolar fibroblasts can be obtained from long-term cultures of bronchoalveolar cells recovered from patients with SA, we decided to further characterize the interactions between alveolar macrophages and alveolar fibroblasts in interstitial lung diseases in an autologous system. Thus, in the present study we assessed the differential secretion of IL-1 and PGE2 by these cells, in comparison with alveolar macrophages and the possible role of these mediators as suppressive agents of fibroplasia and fibrosis in these diseases. Study population. Eighteen patients, belonging to three groups were included in this study.
Pulmonary sarcoidosis (SA). Diagnosis was made in six untreated patients (three males and three females, mean age 37+ 7 years), by clinical and roentgenological presentation, a positive Kveim test, or a positive biopsy of non-caseating granuloma. All patients were in Stage II sarcoidosis. None of them was a smoker.
Diffuse interstitial fibrosis (DIF). Three patients (two males and one female, mean age 61 + 10 years). Diagnosis of DIF was made by roentgenological evidence of reticular infiltration and different degrees of interstitial fibrosis, demonstrated by transbronchial biopsy. None of them was a smoker.
Hypersensitivity pneumonitis (HP). Three patients belonged to this group (two males, one female, mean age 48 + 6 years). Roentgenological evidence (X-rays and CT scan) showed reticular nodular pattern with predominant upper zone involvement. Bronchoalveolar lavage (BAL) analysis demonstrated features of a cell-mediated immune response with lymphocytosis. Lung histology was compatible with HP but no attempt was made to characterize the sensitizing antigens. None of them was a smoker.
Control. Six patients (three males and three females, mean age 44 + 18 years) were admitted for investigation due to persistent cough. All of them presented chest roentgenograms within normal limits. None of them was a smoker.
None of the patients received any medicaments prior to the study. Written consent was obtained from each subject before bronchoscopy. Characterization of cell population present in BAL and pulmonary function test parameters of all patients examined are summarized in Table 1.

Methods:
Bronchoalveolar lavage (BAL). BAL was performed using a flexible fibre optic bronchoscope (BF-B2; Olympus Optical Co., Ltd, Tokyo, Japan) as previously described. 21 The cells were recovered by gentle aspiration. The percentage of lavage fluid (+SD) recovered from each group of patients was as follows: 67 + 10% from CO and 58 + 8% from ILD cases. The average total cells recovered was 5 + 1 x 106 cells from CO and 17 + 7 x 106 cells from ILD patients.
Preparation of AM and AM supernatants. AM were prepared as previously described. 21 Differential counts were performed on a Giemsa stained cytocentrifuge preparation (Cytospin; Shandon, Southern Products Ltd, Runcorn, Cheshire, UK), by counting a minimum of 500 cells. Cells were adjusted to a final concentration of 106 cells/ml in RPMI 1640 medium, supplemented with 10% heat" inactivated FCS, 1% t-glutamine, and 1% streptomycin, penicillin, mycostatin complete medium (Biological Industries, Beit Haemek, Israel). The AM were purified by adherence in a 5% CO2 humidified atmosphere for 1 h at 37C. Identification of macrophages was done by morphology and nonspecific esterase staining and counted with an objective micrometer (Olympus Optical Co., Ltd, Tokyo, Japan). The adherent cell population contained more than 90% AM. AM were cultured in 3-cm diameter plastic Petri dishes (Sterilin, Hounslow, Middlesex, UK) for either 24h or 72 h. The 24 h period was found to be optimal for testing the production if IL-1 in the presence of lipopolysaccharide (LPS-Escherichia coli 055:B5; Difco Laboratories, Detroit, USA; 10 btg/ml) stimulated cultures. The 72 h period was chosen as the optimal time for release of PGE2 in unstimulated cultures. Supernatants were recovered, filtered (Acrodisc 0.2 bt; Gelman Sciences) and stored at -70C until determination for IL-1, and not longer than 2 weeks for PGE2. Preparation of alveolar fibroblasts. The fibroblast line was derived from the bronchoalveolar cells as previously described. 2 First clones of proliferating fibroblasts appeared after 2-3 weeks of incubation in 3-cm diameter plastic Petri dishes (Sterilin, Hounslow, Middlesex, UK). After reaching confluence, usually 5-6 weeks later, the explant tissue was removed with trypsin-EDTA Type B (Biological Industries, Beit Haemek, Israel) for 10 min and cells were split 1:2 at confluency in 25 cm tissue plastic culture flasks (Sterilin, Hounslow, Middlesex, UK). In all experiments the cells used were from passages 4-7.
Preparation of pulmonary fibroblasts. Stable lines of human pulmonary fibroblasts were used a control cells. Lung specimens from pneumonectomy specimens from patients with thoracic malignancies or benign lesions were minced into pieces of 2-4 mm and incubated in i x 5 cm Petri dishes (Sterilin, Hounslow, Middlesex, UK) containing 3 ml of complete DMEM. Every 72 h the non-adherent cells were removed by washing with PBS and fresh media was added. After 2 weeks, cultures reached confluence and the cells were split and used as described above.

Preparation of Fb supernatants and Fb proliferation
test. Fb derived supernatants were obtained from 24 h LPS-stimulated Fb cultures cultured for 24 h with or without LPS. Aliquots of the supematants were frozen at-70C until used. Proliferation test was performed as previously described. 2 Briefly, 100 t.tl of Fbs suspended at 105 Fb/ml were allowed to attach for 1-2 h. Aliquots (50 l.tl) of supernatants of LPSstimulated AM were added. Proliferation was assessed after 72 h, and pulsed with 1 laCi H-Tdr (48 l.tCi/nmol specific activity, Rotem Industries Ltd, Beer-Sheva, Israel) for the last 16 h of culture. The proliferation of Fb in the presence of AM supernatants was compared with that of Fbs DMEM with a final concentration of 100 l.tg/ml LPS.
Assay of prostaglandin and IL-l production. Aliquots of AM and Fb supernatants (24 h production) were assayed for PGE,. by a radioimmunoassay ( Statistics. Student's t-test was used for statistical evaluations using the Epistat Software, (C) 1984, T.L. Gustafson. The results are expressed as mean + SD and values less than 0.05 were considered to be significant.

Results
Effect of AM supernatants on the proliferation of fibroblasts: The basic proliferation rate of Fb from SA and HP patients was shown to be significantly lower than that of Fb in the CO group (Table 2).
AM-derived supernatants were tested for effects on the proliferation of Fb in an autologous culture set-up and in presence of AM supematant of CO patients (Fig. 1) and on normal Fb lines (Fig. 2). The AM supernatants of SA and HP suppressed the prolifera-     Detection of IL-lc and IL-I mRNA transcripts in Fb: In order to assess whether the IL-l(z and IL-I genes are expressed in Fb of these diseases, we assessed the mRNA transcripts by PCR. IL-10t and IL-I transcripts were constitutively found in stimulated, as well as non-stimulated, Fbs (Fig. 3a, b, c and d).

Effects of exogenous IL-I on Fb proliferation:
Exogenous IL-113 (concentrations in the range of 0.3-1000 ng/ml), were added to Fb cultures and proliferation was assessed by the tritiated thymidine incorporation. IL-1]3, at concentrations of 0. 35-62.5 ng/ml (Fig. 4a) and 0.35-125 ng/ml (Fig. 4b) significantly suppressed the basic proliferation rate of Fb by 46 + 1.4% and 39 + 0.9% (p < 0.001) compared with proliferation of Fb in complete medium (Fig. 4a and   b). These concentrations include the range of IL-113 by AM of SA and HP patients (4-0.8 ng/ml). As a general trend in Fbs of the CO group, no suppressive activity was observed at minute concentrations of ILl[3 (Fig. 4c). The modulatory effects of IL-13 were abrogated by anti-IL-1 antibodies or indomethacin ( Table 6). A clear reversion of the suppression was achieved in the SA and HP group, but not in the CO and DIF groups.

Discussion
AM are present within the alveoli and bronchioli while the fibrosis occurs in the interstitium. However, it is feasible to assume that AM affect the process of fibrosis because of their close proximity. The outcome of a number of interstitial lung diseases (ILD) results in severe pulmonary fibrosis. In contrast, patients with SA or HP, manifesting inflammatory or granulomatous diseases, usually heal without excessive scarring. In the present study, we assessed the fibroblast-macrophage interactions in ILD, in an attempt to understand the differential outcome of the diseases. We assessed the proliferative capacity of Fb and the secretion of cytokines/inflammatory products by AM and Fb, the latter being the target cells in these diseases.
We showed ( Table 2) that Fb recovered from SA and HP, display significantly lower rates of background proliferation, compared with Fb recovered from normal tissue specimens even after serial passages for up to 2 months. The existence of Fb populations from fibrotic lungs which retain enhanced proliferative potential in culture, have already been reported '2,23 together with reports on decreased 4 or increased 25 cytokine production by fibroblasts in chronic GVHD (graft versus host dis-   We showed also that AM-derived supernatants suppressed the proliferation of autologous Fb, recovered from patients with SA and HP, but not of Fb from DIF patients or CO ( Fig. la and b). However, as described also by others, iv AM supernatants from SA patients, when tested on normal lung fibroblasts, either inhibit or enhance their proliferation, and no clear pattern of response could be concluded. These results may indicate that our autologous culture setup may approximate the in vivo situation. As SA and HP are two diseases where irreversible deterioration in pulmonary function occurs in only 20% of patients, we hypothesize that inflammatory mediators secreted by AM or Fb may mediate the suppression in the proliferative capacity of autologous Fb in SA and HP. We found here and in previous studies '1,28 that IL-1 is actively secreted from AM in SA and HP patients. Thus, the secretion of ILl together with secretion of TNFo, 29-31 which may possibly mediate the suppression of Fb proliferation in these diseases. This observation was also confirmed by Elias et al. ,4,2 In order to investigate the role of IL-1 in the Fb suppression we examined first whether exogenous stimuli can potentiate Fb to generate soluble IL-I and [. We could detect IL-10t and IL-I[ gene transcription in CO or LPS/IL-1 stimulated Fb from CO and SA patients. Subsequently, we found that stimulated and non-stimulated Fb secrete low detectable levels of both species of IL-1 (0.12-0.17 ng/ml). The same stimulus induced AM to produce high levels of IL-I and [. These results are in agreement with previous findings ,4 demonstrating that stimulated Fb contain more (1-10%) IL-1 mRNA than LPS-stimulated monocytes, but they produce less detectable IL-1[ (<0.04%). The mechanism of IL-1 secretion is still obscure, as it lacks a signal peptide..A rote for plasmin or plasmin-like proteases for the release of IL-1 has been reported in macrophages. 35 It may be that this enzymatic activity is lacking in Fb. This dichotomy between IL-1 gene transcription and expression of biological activity has already been demonstrated in stimulated macrophages and Fb 6,7 and may suggest that IL-1 activity may be regulated at multiple levels, such as transcription, stability of mRNA, translation and post-translational events. In our case, we postulate that one of these regulatory post-transcriptional events may be mediated by prostaglandins, as this mechanism was already demonstrated for normal AM to suppress cytokine production. 8 In fact, we showed that only fibroblasts recovered from SA and HP patients may be induced to release significant amounts of PGE upon stimulation with LPs.
Two points have to be considered concerning the possible involvement of IL-1 in suppressing the proliferative capacity of lung Fb: (i) Fb from SA and HP patients in vivo are exposed to high levels of IL-1 of AM origin; (ii) the exogenous effect of this cytokine may be differential on quiescent versus rapidly proliferating fibroblasts. In fact, we tested the effects of exogenous IL-I[ on the proliferation of normal Fb, as well as Fb cells obtained from SA and HP patients.
Our results indicate that IL-I[, at the levels equivalent to those present in the microenvironment in the lungs of SA and HP patients (0-4 ng/ml), exerts a suppressive effect on slowly subconfluent proliferating alveolar Fb from these diseases, whereas at levels detected in normal lungs (0-0.2 ng/ml), enhance the replication of actively proliferating Fb recovered from control lungs (Fig. 3a-c). Similar dichotomic effects have already been demonstrated for the regulation of fibroblast proliferation by recombinant interferons , 0t and [.8 Indeed, using Fb, as well as other cell types, contrasting effects of IL-1 on cell proliferation were reported.  This may result from differential sensitivity of various cells to the direct mitogenic effects of IL-1 or to a distinct cytokine cascade induced by IL-1. In our case, it may be possible that IL-1 induces in SA and HP, Fb suppressive cytokines (such as TGF-[, IL-6 and possibly IL-l0) or other mediators (such as PGE2).
In conclusion, we postulate that the benign course of the disease of most patients with SA and HP involves a downregulation in the proliferative capacity and possibly alters the secretory function of Fb.