Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

We have recently demonstrated by immunohistochemistry the presence of leukemia inhibitory factor (LIF) in epithelial cells, mesenchymal cells, and nerve fibers in human airways (1). Using a guinea pig tracheal explant model, we showed that LIF specifically augmented the contractile responses to both endogenous and exogenous tachykinins, suggesting actions on the release or metabolism of tachykinins or actions on tachykinin receptor systems. On the basis of these findings, we have proposed that LIF is an important cytokine in the regulation of the inflammatory response in the lung in disorders such as asthma and bronchitis, particularly in mediating neuroimmune interactions. However, a detailed analysis of the structural cells in the lung that produce LIF, the processes involved in regulating its production, and effector roles are poorly understood.

LIF, a multifunctional, secreted glycoprotein that exists in both soluble and matrix-bound forms, displays biologic activities ranging from the differentiation of myeloid leukemic cells into macrophage lineage to effects on bone metabolism, inflammation, neural development, embryogenesis, and the maintenance of implantation (2). It is now clear that LIF is related in both structure and mechanism of action to the interleukin (IL)-6 family of cytokines, which also includes IL-11, ciliary neurotrophic factor, oncostatin M, and cardiotrophin 1 (2). The actions of these cytokines are mediated through specific cell-surface receptors that consist of a unique  chain and the shared signal transducing subunit gp130 (3). The family can be further divided into three subgroups on the basis of receptor subunit usage (4, 5).

LIF is released from a variety of cell types in vitro, including synovial and lung fibroblasts (6), endothelial cells (7), neural tissue (8), T lymphocytes (7), monocyte-macrophages (9), thymic epithelium (10), chondrocytes (11), and mesangial cells (12). LIF production is probably strictly controlled because in noninflamed tissue, LIF messenger RNA (mRNA) is not normally present. In the majority of the cell types listed above, LIF mRNA and protein are induced by the proinflammatory cytokine IL-1. In neural tissue, LIF acts as the key intermediate in producing effects of IL-1 such as induction of mRNA for the tachykinins substance P and neurokinin A and their receptors (8, 13, 14). LIF has also been shown to upregulate the expression and activity of the cytoplasmic form of phospholipase A₂ (15, 16), suggesting an interaction between this cytokine and eicosanoids.

The aim of this study was to determine whether LIF is released by human lung tissue in response to inflammatory stimuli, and to identify the predominant human pulmonary cell types expressing LIF and the LIF receptor (LIFR). Further, we sought to examine the effects of IL-1 and IL-6 treatment on the gene expression of LIF and LIFR, as well as on LIF release, from cultures of nontransformed human bronchial smooth-muscle cells (HBSMC), bronchial epithelial cells, and human lung fibroblasts.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Human Lung Tissue

Segments of lung tissue including small airways were obtained from 12 patients undergoing surgery for removal of lung tumors as previously described (17). In all cases, macroscopically normal lung tissue was taken from regions remote from the tumor. The tissue was placed in ice-cold Leibovitz L-15 medium (GIBCO, Burlington, OR) supplemented with antibiotics and was minced with fine dissecting scissors; approximately 100 mg of tissue was added to each well of a 24-well microtiter plate containing Dulbecco's modified Eagle's medium (DMEM) supplemented with L-glutamine, streptomycin, penicillin, and gentamicin (GIBCO).

Cell Cultures

Cultures of human nontransformed pulmonary artery smooth-muscle cells (PASMC), pulmonary artery endothelial cells (PAEC), bronchial microvascular endothelial cells (MVEC), HBSMC, and normal human bronchial epithelial (NHBE) cells were obtained from Clonetics (San Diego, CA) and were used between passages 2 and 6. The human fetal lung fibroblast cell line HFL and the transformed alveolar type II epithelial cell line A549 were obtained from American Type Culture Collection (ATCC, Rockville, MD). The transformed epithelial cell line BEAS-2B was kindly provided by Dr. Curtis Harris at the National Cancer Institute (Bethesda, MD). The tracheal glandular epithelium (HTG) line was a gift of Dr. Dharam Chopra, Institute of Toxicology, Wayne State University (Detroit, MI). Human alveolar macrophages and circulating neutrophils and eosinophils were obtained as previously described (18).

NHBE cells were grown in serum-free conditions in defined growth medium (SAGM; Clonetics) supplemented with epidermal growth factor (0.5 ng/ml), hydrocortisone (0.7 mg/ml), triiodothyronine (65 ng/ml), epinephrine (0.5 mg/ml), retinoic acid (0.1 ng/ml), insulin (5 mg/ml), transferrin (10 mg/ml), bovine pituitary extract (BPE; 1 ng/ml), and gentamicin (50 mg/ml). PAEC and MVEC were cultured in EGM (Clonetics) medium supplemented as described above. Cultures of HBSMC and PASMC were grown in SmGM (Clonetics) supplemented with gentamicin (50 mg/ml) and 10% fetal bovine serum (FBS), whereas HFL cultures were grown in Ham's F-12 medium (GIBCO) supplemented with gentamicin (50 mg/ml) and 10% FBS.

LIF Release

Lung tissue. For studies examining the release of LIF from human lung tissue, duplicate wells were incubated with recombinant human (rh)IL-1 (100 pg/ml) (Pharmingen, San Diego, CA) or medium as a vehicle control, for periods of 1, 3, 6, 24 and 48 h. At each of these times, an aliquot of culture medium was taken for measurement of LIF by enzyme-linked immunosorbent assay (ELISA) (Amersham, Chicago, IL). The microtiter well was then emptied and the tissue was rinsed in phosphate-buffered saline (PBS), snap-frozen in liquid N₂, and stored at 65°C until required. LIF release was expressed per milligram of protein. Immunoglobulin (Ig)E receptor triggering was performed in DMEM without FBS. Resected lung was divided into two equal proportions, and 6 µg/ml of anti-IgE (Sigma Chemical Co., St. Louis, MO) or vehicle was added. Explant lung was cultured for 24 h, and LIF release was quantified per milliliter of supernatant.

Cell cultures. For studies examining the release of LIF from NHBE cells, cells were grown to 80% confluence and then quiesced by incubation in basal medium without hydrocortisone and BPE for 48 h. For HBSMC, HFL, and A549 cultures, cells were also grown to 80% confluence and then quiesced by removal of FBS for 48 h. After this period of time the medium was changed and either IL-1 (100 pg/ml), IL-6 (100 U/ml, smooth-muscle cells only) (Pharmingen), or vehicle was added for a further 1, 3, 6, 24, or 48 h. At each of these times, an aliquot of culture medium was taken and stored at 65°C for subsequent measurement of LIF by ELISA. Cells were lysed directly in the flask by direct addition of Trizol reagent (GIBCO). The resulting cell lysates were stored at 65°C until required for protein and gene expression studies.

Protein quantification. Total soluble protein concentration was determined by a modification of the Bradford method as per the manufacturer's (Bio-Rad, Mississauga, ON) instructions.

Reverse transcription (RT). Total cellular RNA concentration was determined spectrophotometrically, and integrity was judged by inspection of 28S and 18S ribosomal RNA bands after electrophoresis on a 1% agarose-5% formaldehyde denaturing gel. Samples of total RNA (1 µg each), together with random hexamers, were heated to 70°C for 10 min and then placed on ice. First-strand complementary DNA (cDNA) synthesis was then performed using 200 U of Moloney Murine Leukemia Virus (MMLV) reverse transcriptase in 1× cDNA buffer, 2 U of RNAsin, and 500 µM of each deoxyribonucleoside 5'-triphosphate (dNTP) in a total volume of 20 µl (Promega, Madison, WI). The samples were incubated at 37°C for 60 min, and the reaction was then terminated by heating to 65°C for 10 min. A volume of this cDNA preparation equating to 250 ng of starting RNA was then amplified by polymerase chain reaction (PCR).

PCR. Because large amounts of total cellular RNA are required for mRNA isolation, RT-PCR and Southern blot analysis were used in experiments involving cell lines. In all cases, the PCR was optimized such that measurements of cDNA were taken during the exponential phase of amplification. A standard curve generated with each experiment provided a semiquantitative estimate of changes in gene expression. Gene-specific, intron-spanning primers for LIF, LIFR, and -actin, based on previous publications, were synthesized at the University of Calgary (Calgary, AB, Canada) (Table 1). PCR amplifications were performed using 50-µl reactions containing MgCl₂ at a final concentration of 1.5 mM for both LIF and LIFR, 200 µM of each dNTP, 25 pmol of specific primers, 5 µl of cDNA template, and 2.5 U of Taq DNA polymerase added after heating the reaction to 94°C for 3 min. LIF or LIFR primers were added with primers for -actin in the same tube containing cDNA, and were co-amplified for 28 cycles. PCR products generated by primers for LIF and LIFR were sequenced using an ABI prism automated sequencer at the University of British Columbia (Vancouver, BC, Canada). The sequences generated were 98.5 and 97.8% homologous with LIF and LIFR, respectively, and had no homology with any other human sequences contained in Genbank. Standard curves to allow quantification of product were made from serial dilutions of recombinant complementary RNA (rcRNA) for LIF and LIFR. These rcRNA products were made by synthesizing PCR primers containing both a T7 RNA polymerase recognition site and either the LIF or LIFR sequence of interest. The resulting PCR product was used to generate rcRNA transcripts using T7 RNA polymerase (GIBCO). Following PCR amplification, aliquots were electrophoresed on a 2% agarose gel and visualized by staining with ethidium bromide.

Southern blot analysis. Following denaturation and neutralization, PCR products were transferred onto a Hybond N+ membrane by capillary transfer and the DNA was irreversibly cross-linked to the membrane by baking for 2 h at 80°C. Membranes were then prehybridized in hybridization buffer consisting of 6× saline sodium citrate (SSC), 0.5% sodium dodecyl sulfate (SDS), 50 µg/ml heparin, and 0.1% Na(PO₄)₂ for 2 h before the addition of random primer-labeled ³²P-labeled LIF PCR product and overnight hybridization at 65°C. Following hybridization, the membranes were washed twice in 2× SSC/0.1% SDS for 30 min at 68°C, followed by 1× SSC/0.1% SDS for 30 min, and finally 0.1× SSC/0.1% SDS for 30 min. Membranes were then exposed to X-ray film for 1 h in a light-tight X-ray cassette with an intensifying screen.

Northern blot analysis. Total cellular RNA was isolated from tissue via acid guanidium thiocyanate/phenol/chloroform extraction. Polyadenylated RNA was isolated using the PolyATract mRNA isolation system III (Promega). A total of 3 µg mRNA was electrophoresed in a 1.2% agarose gel containing 2.2 M formaldehyde and was transferred to a Hybond N+ nylon membrane (Amersham, Oakville, ON, Canada), and irreversibly cross-linked. The blot was prehybridized for 4 h in 50% formamide, 5× saline sodium phosphate ethylenediaminetetraacetic acid (SSPE), 5× Denhardt's, and 100 µg/ml denatured salmon sperm DNA at 42°C. The blot was then hybridized with 1 × 10⁶ counts per minute (cpm)/ml of ³²P-labeled LIF cDNA probe (481 base pairs) (specific activity 7 × 10⁸ cpm) overnight at 42°C. The membrane was then subjected to high-stringency washes in 2× SSPE-0.1% SDS, twice for 15 min at 42°C; 1× SSPE- 0.1% SDS for 30 min at 42°C; and 0.1× SSPE-0.1% SDS for 20 min at 42°C; and was then placed in a light-tight X-ray cassette opposed to X-ray film with intensifying screens at 70°C for 20 h. For analysis of glyceraldehyde-3-phosphate dehydrogenase (GAPDH), the hybridized probe was stripped by boiling the blot in 0.1% SDS and, after being checked for residual activity, the blot was rehybridized using a 1.2-kb PstI fragment of rat GAPDH cDNA, which cross-hybridizes with human GAPDH mRNA.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

LIF Release from Human Lung Tissue

Unstimulated lung tissue (n = 7) did not release detectable levels of LIF protein at 1 and 3 h (Figure 1, upper panel). However, when lung tissue was exposed to rhIL-1 (100 pg/ml), the rate of LIF release was markedly enhanced. LIF release was quantifiable after 1 h and continued to increase in a time-dependent manner to a maximum of 15.2 ± 7 ng/mg of total protein at 48 h. IL-1-treated lung LIF values were significantly greater than control values at 1, 3, and 6 h, P < 0.05, by a repeated-measures analysis of variance with sequential Bonferroni corrections for multiple comparisons. In contrast, in lung tissue that was exposed to medium alone, detectable amounts of LIF were not observed until at least 6 h of incubation, although after 24 and 48 h of culture these tissues appeared to release similar amounts of LIF to the IL-1-treated tissue (Figure 1, upper panel). IgE receptor triggering also yielded significant increases in LIF release over control values (P < 0.05); the increase in LIF release at 24 h is illustrated in Figure 1, lower panel.

View larger version (22K): [in this window] [in a new window]   Figure 1.   LIF release from human lung tissue. Lung tissue was incubated with IL-1 (100 pg/ml) (solid bars) or vehicle (hatched bars) for 0 to 48 h (upper panel), or anti-IgE (6 U/ml) or vehicle (control) for 24 h (lower panel), as described in MATERIALS AND METHODS. A specific ELISA assessed LIF generation. Results are expressed as the mean ± SE of seven experiments for IL-1; paired results from five different lungs are shown for anti-IgE or vehicle (*P < 0.05). Note that the ordinate in the upper panel is discontinuous, changing to ×1,000 after the break in scale.

LIF and LIFR mRNA Expression in Human Lung Tissue

To assess whether the increase in LIF generation induced by IL-1 was associated with a specific increase in LIF mRNA expression, samples of human lung tissue mRNA were subjected to Northern blot analysis. A single band of hybridization at the expected size of 4.1 kb was observed in unstimulated lung tissue, albeit at a very low level (Figure 2). A similar low level of gene expression was observed after 3 h exposure to culture medium alone. In contrast, in lung tissue exposed to IL-1, the expression of LIF mRNA was substantially upregulated at 3 h (Figure 2). However, in agreement with ELISA data, after 24 h, LIF gene expression was similar in both control and IL-1-treated samples of lung tissue.

View larger version (56K): [in this window] [in a new window]   Figure 2.   IL-1 induces expression of LIF mRNA. Human lung tissue was incubated with IL-1 (100 pg/ml) for 0, 3, or 24 h. At these times, the expression of LIF mRNA or GAPDH mRNA was assessed by Northern blot analysis using 3 µg mRNA. A representative blot of three separate experiments is shown. C = control.

LIF and LIFR mRNA Expression in Human Lung Cells

Because experiments in whole lung tissue do not provide information regarding the precise cellular localization of LIF or LIFR mRNA, experiments were performed to localize and quantify LIF and LIFR gene transcripts in a variety of primary and transformed cell cultures representative of those cells found in lung tissue. As can be seen in Figure 3, a single specific hybridization band corresponding to the expected size of the LIF transcript was observed in all cell types examined, with greatest intensity observed in HFL cells. High levels of expression were also observed in samples of bronchial smooth-muscle cells as well as pulmonary artery smooth muscle. The rank order of cellular LIF gene expression was HFL  HBSMC > PASMC > A549 > BEAS-2B > NHBE = HTG > MVEC > PAEC as judged by reference to standard curves, with unstimulated HFL cells containing approximately 200 pg of LIF mRNA.

View larger version (71K): [in this window] [in a new window]   Figure 3.   Distribution of LIF (upper panel) and LIFR mRNA (lower panel) in human lung cells. Various cell types normally resident in human lung tissue were cultured as described in MATERIALS AND METHODS. RNA was reverse transcribed into cDNA and subjected to PCR. A representative blot of three separate experiments is shown. Upper panel shows results of Southern blot analysis; lower panel is ethidium bromide-stained gel of PCR products. B2B = BEAS-2B.

mRNA transcripts for LIFR were also observed in all cell types examined (Figure 3). The basal level of expression of LIFR mRNA was greater than that for LIF gene expression in the corresponding sample (Figure 3) and relatively uniform between cell types, despite expression levels within the exponential range of amplification as judged by standard curves. Gene transcripts for LIF and the LIFR were also detected in alveolar macrophages, peripheral blood neutrophils, and eosinophils (data not shown).

Effect of IL-1 and IL-6 on LIF Release from Cultured Cells

The pattern of LIF released from NHBE, HBSMC, and HFL in response to IL-1 was similar to that observed for whole lung tissue. Unstimulated cells did not release detectable amounts of LIF. Similarly, cells incubated with culture medium alone for the duration of the experiments (48 h) did not release detectable amounts of LIF. However, exposure to IL-1 (100 pg/ml) produced a rapid increase of LIF protein that increased with time to 48 h (n = 4; Figure 4). When normalized for total soluble protein, the rank order of maximum release of LIF was HFL > HBSMC >> NHBE. IL-6 (100 U/ml), but not LIF (5 ng/ ml) itself, caused a time-dependent increase in LIF release from bronchial smooth-muscle cells (Figure 5).

View larger version (19K): [in this window] [in a new window]   Figure 4.   IL-1 induces LIF release from cultured cells. HFL lung fibroblasts (solid circles), HBSMC (solid triangles), and non-transformed NHBE (solid squares) were cultured as described in MATERIALS AND METHODS and stimulated with IL-1 (100 pg/ml) for 0 to 48 h. LIF generation was assessed by a specific ELISA. Results are expressed as the mean ± SE of four separate experiments. No LIF was detectable in unstimulated time-control cells.

View larger version (13K): [in this window] [in a new window]   Figure 5.   Effect of IL-6 (100 U/ml) or LIF (5 ng/ml) on LIF release in cultured HBSMC (n = 4).

Effects of IL-1 on LIF and LIFR mRNA in Cultured Cells

Epithelial cells. In these cells, very low levels of LIF mRNA were observed under basal conditions (Figure 6). Exposure to IL-1 (100 pg/ml) induced a rapid 8-fold increase in LIF gene expression (from 50 to 400 pg) that was maximal at 3 to 6 h. However, this increase was short-lived; LIF mRNA expression declined to approach control levels by 24 h. In contrast, the expression of LIFR mRNA was unaltered by exposure to IL-1 at any time up to 48 h.

View larger version (77K): [in this window] [in a new window]   Figure 6.   (Upper panel ) IL-1 induces expression of LIF mRNA in cultured normal human epithelial cells. Cells were incubated in the presence of IL-1 (100 pg/ml) for 0 to 48 h. RNA was reverse transcribed into cDNA and subjected to PCR. A representative ethidium bromide-stained gel of LIF and -actin products and an LIF rcRNA standard curve are shown. (Lower panel ) Southern blot for LIF from same experiment.

Bronchial smooth-muscle cells. As was observed with bronchial epithelial cells, cultures of nontransformed HBSMC also expressed low levels of LIF mRNA under resting conditions. Similarly, exposure to IL-1 (100 pg/ ml) produced a rapid 7-fold increase in LIF gene expression in these cells. However, the kinetics of LIF mRNA expression differed from that in epithelial cells, with maximal gene expression maintained for a longer period. However, by 24 h, LIF mRNA expression had returned to approach baseline expression. As was observed for epithelial cells, exposure of HBSMC to IL-1 did not alter the expression of LIFR mRNA for the duration of the experiment (Figure 7).

View larger version (49K): [in this window] [in a new window]   Figure 7.   (Upper panel ) Southern blot of IL-1 induces expression of LIF mRNA in cultured normal HBSMC. Cells were incubated in the presence of IL-1 (100 pg/ml) for 0 to 48 h. (Middle panel ) Results of densitometric analysis. (Lower panel ) Effects of IL-1 on LIFR mRNA in bronchial smooth-muscle cells. The standard curve is also shown (0.4 to 250 pg). Note that the samples are all within the range of the standard curve.

Lung fibroblasts. The pattern of LIF mRNA expression in HFL cells was similar to that observed in HBSMC; that is, low basal expression that is rapidly increased by exposure to IL-1. The level of amplification was maintained for 6 h before returning toward control levels of expression. As was observed in other cell cultures, expression of LIFR mRNA was not influenced by exposure to IL-1.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The results of this study demonstrate for the first time substantial liberation of LIF from human lung following proinflammatory stimuli. Furthermore, the results demonstrate that numerous lung structural cell types express LIF and LIFR, suggesting previously unrecognized roles for this cytokine in lung biology. Because IL-6 was shown to release LIF, our data suggest that some biologic effects previously attributed to IL-6 may in fact be due to LIF. The results thus confirm our recent immunohistochemical data in normal human airways (1), in which immunoreactive LIF was primarily located in the epithelial cells of both central and peripheral airways with some staining in mesenchymal cells (probably fibroblasts) and nerve fibers. These results suggest potentially important roles for LIF in the local regulation of airway inflammation. Our studies in airway explants (1) and the results of experiments with transgenic mice (2) suggest a role for LIF in the regulation of neurotransmitter and neurotransmitter receptor phenotype, particularly for tachykinin and muscarinic receptors.

In the current study, we have demonstrated that freshly obtained lung tissue stimulated with the proinflammatory cytokine IL-1 releases detectable amounts of LIF after 1 to 3 h that increase in a time-dependent manner up to 48 h. The pattern of release suggests that LIF is not primarily present in stored form, but is synthesized de novo. This finding is supported by the observation that under normal conditions, circulating levels of LIF are not readily measurable (19). Carlson and associates demonstrated that LIF gene transcription is highly active even when mRNA levels are low or undetectable, suggesting that LIF gene transcription may be constitutive but that under normal conditions, nascent nuclear RNA is degraded rather than processed into mRNA (20). Data from the current study support this finding because exposure of lung tissue to IL-1 produced a rapid increase in gene expression for LIF that was observable after 1 h and was maintained for up to 24 h. Consistent with our results, high levels of LIF are also found in bronchoalveolar lavage fluid obtained from patients with the acute respiratory distress syndrome, in which there is diffuse pulmonary inflammation (21). Control lung tissues also liberated substantial amounts of LIF during explant culture, although levels were not detectable until at least 6 h of culture. It is likely that the release of LIF from unstimulated tissue is the result of paracrine stimulation, because this phenomenon was not seen in unstimulated isolated cells in culture over a 48-h period. This finding may reflect cytokines or arachidonic acid products released from resident lung cell types or infiltrating cells during the culture period following tissue manipulation, or could reflect preexisting lung inflammation, because most of these individuals were tobacco smokers.

Although studies using whole tissue are valuable, they provide little data on the cellular localization of LIF or LIFR within the lung. Therefore, we have used a variety of cell types that are resident in the lung in order to determine the cellular localization of the genes for LIF and its receptor. Gene transcripts for LIF and LIFR were observed in all cell types examined. This widespread distribution of LIF and LIFR mRNA suggests that LIF and related cytokines play an important role in a variety of airway functions. The greatest levels of LIF expression were observed in the cultured fibroblast cell line HFL. High levels of LIF gene expression were also observed in cultured bronchial smooth-muscle cells and PASMC, with lower levels of expression observed in airway epithelial cells. Interestingly, expression of LIF gene transcripts in microvascular and PAEC was extremely low. LIFR mRNA was also detected in all cell types examined. Exposure of nontransformed human fibroblasts, bronchial smooth-muscle cells, and bronchial epithelial cells to IL-1 produced a rapid but transient increase in LIF mRNA levels. In contrast, IL-1 did not influence gene expression for LIFR in these cells. The release of LIF from these cells correlated well with the cellular mRNA expression. Thus, HFL cells generated the most LIF in response to IL-1, followed by HBSMC and NHBE in the rank order HFL > HBSMC > NHBE. LIF has been shown to be released from nonpulmonary tissues (15, 20, 22, 23) and from fibroblasts (6) and mast cells (24).

Exposure to IL-1 produced a rapid, time-dependent increase in LIF mRNA accumulation that was maximal by 6 h. However, the kinetics of mRNA induction appeared to differ between cell types. In HBSMC and HFL cells, after the rapid increase in mRNA accumulation there was a rapid decline in mRNA levels, such that by 24 h, LIF gene expression was similar to unstimulated conditions. This finding is consistent with data from Carlson and colleagues (20) in which exposure of superior cervical ganglia to IL-1 significantly increased the LIF gene transcription rate between the time of exposure and 1 h, and thereafter returned to the basal rate. The rapid increase in mRNA expression is also consistent with findings in mesangial cells in which maximal LIF mRNA expression was observed at 8 h and declined to background after 20 h (12). Similar kinetics were also seen in cultured lung fibroblasts where LIF mRNA levels were increased after 2 h exposure to IL-1 (6). However, in contrast to the present study, LIF mRNA accumulation in fibroblasts remained elevated for up to 16 h. The reasons for the difference in the kinetics of LIF mRNA accumulation may reflect the use of different fibroblast lines (HFL versus CCL-202) or the concentration of IL-1 used. In the present study, the concentration of IL-1 was 100 pg/ml, which was based on studies by De and coworkers (25) and Ludlam and associates (8). In contrast, Elias and colleagues (6) used a 25-fold higher concentration of IL-1. However, Hartner and coworkers (12) reported that the maximum stimulatory concentration of IL-1 was 5 ng/ml. On this basis it is unlikely that the difference in the IL-1 concentrations used in these studies is a significant factor.

In NHBE cells, LIF mRNA expression remained significantly upregulated after 24 h exposure to IL-1, although after this time LIF mRNA accumulation appeared to decrease toward basal levels. When comparing the results between cell types, it appears that the level of LIF gene expression remains upregulated for longer periods in epithelial cells.

LIF itself was unable to induce LIF mRNA or LIF release in HBSMC, consistent with experiments using rat mesangial cells (12), and exposure to IL-1 or LIF itself did not significantly influence the levels of LIFR mRNA expression in any of the cell types examined. The reasons underlying this finding are not clear, although it may reflect high constitutive expression of LIFR mRNA, which was observed in unstimulated cells, and a noninducible gene. In contrast, IL-6 caused a sustained elevation in LIF release in bronchial smooth-muscle cells, a finding that suggests that effects attributed to IL-6 (such as smooth-muscle proliferation) require reexamination in the presence of neutralizing LIF or LIFR antibodies.

In conclusion, we have demonstrated novel sites of synthesis and release of LIF in human lung. LIF gene expression is upregulated in human lung tissue exposed to the proinflammatory cytokine IL-1 and following IgE receptor triggering. The highest level of basal expression was observed in fibroblasts, followed by HBSMC and NHBE. Exposure of these cells to IL-1 also resulted in a rapid induction and accumulation of LIF mRNA, correlated with the cellular release of LIF protein. LIF may be involved in many facets of the pulmonary response to inflammation. In particular, on the basis of the cellular sources of LIF we have detected, we suggest roles for LIF in repair processes and airways responsiveness in addition to the airway neuromodulatory roles we have already found.