We have reported that α1-acid glycoprotein (AGP) gene expression was induced in lung tissue and in alveolar type II cells during pulmonary inflammatory processes, suggesting that local production of this immunomodulatory protein might contribute to the modulation of inflammation within the alveolar space. Because AGP may also be secreted by other cell types in the alveolus, we have investigated the expression and the regulation of the AGP gene in human and rat alveolar macrophages. Spontaneous AGP secretion by alveolar macrophages was increased 4-fold in patients with interstitial lung involvement compared with that in controls. In the rat, immunoprecipitation of [35S]methionine-labeled cell lysates showed that alveolar macrophages synthesize and secrete AGP. IL-1β had no effect by itself, but potentiated the dexamethasone-induced increase in AGP production. RNase protection assay demonstrated that AGP mRNA, undetectable in unstimulated cells, was induced by dexamethasone. Conditioned medium from LPS-stimulated macrophages as well as IL-1β had no effect by themselves, but potentiated the dexamethasone-induced increase in AGP mRNA levels. In addition to cytokines, PGE2 as well as dibutyryl cAMP increased AGP mRNA levels in the presence of dexamethasone. When AGP expression in other cells of the monocyte/macrophage lineage was examined, weak and no AGP production by human blood monocytes and by rat peritoneal macrophages, respectively, were observed. Our data showed that 1) AGP expression is inducible specifically in alveolar macrophages in vivo and in vitro; and 2) PGE2 and cAMP act as new positive stimuli for AGP gene expression.
The α1-acid glycoprotein or orosomucoid (AGP)3 is one of the major acute phase proteins in humans, rats, mice, and other species (1). As most acute phase proteins, its serum concentration increases in response to systemic tissue injury, inflammation, or infection, and these changes in serum protein concentrations have been correlated with increases in hepatic synthesis (2). Expression of the AGP gene, like that of other genes encoding proteins involved in the acute phase reaction, is controlled by a combination of the major regulatory mediators, i.e., glucocorticoids (3, 4, 5) and a cytokine network involving mainly IL-1β, TNF-α, and IL-6 (6, 7, 8, 9).
AGP is considered a natural anti-inflammatory and immunomodulatory agent notably with respect to its antineutrophil and anticomplement activities (10). Indeed, AGP has been shown to act in vitro and in vivo as an immunomodulating molecule. In vitro, AGP inhibits polymorphonuclear neutrophil activation (11), modulates LPS-induced cytokine secretion by monocytes-macrophages (12), and increases the secretion of an IL-1 inhibitor by murine macrophages, probably the IL-1 receptor antagonist (13, 14). In vivo, AGP protects mice from TNF-α-induced lethality (15), and more recently, AGP has been shown to inhibit specifically TNF-α-induced, but not anti-Fas-induced, apoptosis of hepatocytes in mice (16). The immunomodulatory activity of different glycoforms of AGP has been shown to be dependent on carbohydrate composition (13, 17). AGP is a single polypeptide that contains three to five highly sialylated carbohydrate side chains (18). Williams et al. (10) have shown that the sialyl Lewis X form of AGP, which is induced during inflammation (19), ameliorates both complement- and neutrophil-mediated injuries, while a nonsialyl Lewis X form does not. In these respects, AGP can be considered as a natural anti-inflammatory and immunomodulatory agent, and local expression of AGP, at the site of the initial acute phase reaction, could protect against the deleterious effects of inflammation. This is particularly important in the alveolar space where the integrity of the structure is essential for the maintenance of lung function.
We have recently demonstrated that AGP gene expression was induced in vivo in human and rat lung tissue and notably in alveolar type II cells during acute and chronic pulmonary inflammatory processes. In vitro, we showed that rat alveolar type II cell primary cultures expressed AGP mRNA and secreted immunoreactive AGP when stimulated in the presence of dexamethasone (Dex) with the secretory products of alveolar macrophages (20). These results suggest that the local production of this immunomodulatory protein might play a role in the regulation of inflammation within the alveolar space.
The aim of this study was to determine whether AGP could also be secreted by alveolar macrophages, which play a key role in the regulation of the inflammatory intra-alveolar processes. We have thus investigated the expression and the regulation of the AGP gene by alveolar macrophages (AM) in human ex vivo and in rat in vitro, and we have compared this expression with extrapulmonary cells from the monocyte/macrophage lineage.
Materials and Methods
Recombinant murine cytokines were purchased from Immugenex (Los Angeles, CA). Escherichia coli (strain 026: B6)-derived LPS was obtained from Difco (Detroit, MI). N,N,N′,N′-tetra-methylethylenediamine (TEMED), ammonium persulfate, urea, Dex, PGE2, dbcAMP, protein A-Sepharose, and guanidine thiocyanate (GuSCN) were obtained from Sigma (La Verpilliere, France). Transcription reagents were purchased from Promega (Madison, WI). [α-32P]UTP (400 Ci/mmol) and ProMix, a mixture of l-[35S]methionine and l-[35S]cysteine (14.3 mCi/ml) were obtained from Amersham (Les Ulis, France). RNase-free DNase I, RNase A and T1, and brewer’s yeast transfer RNA were supplied by Boehringer Mannheim (Mannheim, Germany). Acrylamide/bisacrylamide, phenol, and proteinase K were purchased from Appligene (Illkirch, France). All restriction enzymes were obtained from New England BioLabs (Beverly, MA) or from Boehringer Mannheim. Tissue culture media, supplements, and FBS were obtained from Life Technologies (Cergy Pontoise, France); tissue culture plasticware was purchased from Costar (Cambridge, MA).
Isolation and culture of rat macrophages
Alveolar macrophages were recovered from adult pathogen-free male Sprague Dawley rats, weighing 250–280 g (Charles River Breeders, St. Aubin les Elbeuf, France), by bronchoalveolar lavage in calcium- and magnesium-free PBS (pH 7.4) and 0.2 mM EGTA as reported previously (21). Resident peritoneal macrophages were obtained by washing the peritoneal cavity with RPMI 1640 (Life Technologies) containing 2 mM l-glutamine, 5% (v/v) heat-inactivated FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.25 μg/ml amphotericin B (complete RPMI). The same medium was used for plating and stimulation of macrophages. The viability of recovered cells was >95% as assessed by the trypan blue exclusion test. Cells were plated at a concentration of 2 × 106 cells/well in six-well tissue culture plates for RNA analysis and ELISA, 5 × 106 cells/60-mm diameter tissue culture dishes, for [35S]methionine pulse labeling. After a 2-h adhesion period at 37°C under 5% CO2 in air (v/v), nonadherent cells were removed by two washes with PBS. Cells were immediately treated in fresh medium within 48 h as indicated in the text. In these experiments we stimulated rat cells with murine recombinant cytokines, because these cytokines have previously been shown to be active in rat cells (22, 23).
Alveolar macrophage-conditioned medium (AM-CM) was prepared as previously described (23). Rat alveolar macrophages recovered by bronchoalveolar lavage were resuspended in complete RPMI at a density of 106 cells/ml. Cells (106) were plated in each well of a 24-well cell culture plate and allowed to adhere for 2 h. Nonadherent cells were then removed, and fresh medium containing 10 μg/ml LPS was added. More than 98% of adherent cells were alveolar macrophages as assessed by nonspecific esterase stain (Sigma). Conditioned medium, consisting of LPS-activated rat alveolar macrophages (AM-CM), was recovered after a 24-h incubation period.
Isolation and culture of human alveolar macrophages and blood monocytes
Six patients (diseased group) with interstitial lung involvement as assessed by chest × ray and computed tomography scan, for whom fiberoptic bronchoscopy with bronchoalveolar lavage (BAL) was needed for diagnostic procedure, were studied (Table I). They were compared with a control group constituted of 10 patients with mild chronic bronchitis. The latter had normal chest x ray and computed tomography scan, normal lung function data, and no respiratory infection within the last 3 mo. All of them were smokers or ex-smokers. Lung macrophages were isolated from BAL fluid as previously described (12). After centrifugation (4°C, 600 × g, 10 min), cells were resuspended in complete RPMI, and the alveolar macrophages population (2 × 106) was purified by a 45-min adherence period to the plastic of a 60-mm diameter dish. Spontaneous ex vivo secretion of AGP in culture supernatants was determined at 24 h by ELISA and expressed as nanograms of AGP secreted per million cells.
|Patient .||Age (yr) .||Sex .||Smoker .||Diagnosis .||BAL .||.||.||.||.||TLC (% pred) .||DLCO/AV (%) .|
|.||.||.||.||.||TC/ml .||Ly (%) .||AM (%) .||Neut (%) .||Eo (%) .||.||.|
|Patient .||Age (yr) .||Sex .||Smoker .||Diagnosis .||BAL .||.||.||.||.||TLC (% pred) .||DLCO/AV (%) .|
|.||.||.||.||.||TC/ml .||Ly (%) .||AM (%) .||Neut (%) .||Eo (%) .||.||.|
IPF, idiopathic pulmonary fibrosis, BAL, bronchoalveolar lavage; TC/ml, total cells/ml; Ly, lymphocytes, AM, alveolar macrophages; Neut, neutrophils; Eo, eosinophils; TLC (% pred), total lung capacity (% of predicted); DLCO/AV, diffusing lung capacity/alveolar ventilation.
Peripheral blood from healthy, medication-free volunteers was collected into EDTA-coated tubes. Mononuclear cells were separated by means of Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) density sedimentation and cultured in complete RPMI as previously described (12). Cells were plated at a concentration of 6 × 106 /well in six-well plates, and the monocyte population was purified by a 2-h adherence period before stimulation. AGP production in culture supernatants was measured at 24 h by ELISA.
Quantification of AGP concentration in cell supernatants
The amount of rat AGP secreted by isolated macrophages was measured in the supernatant using sandwich-type immunoenzymatic methods as reported previously (24). Results were expressed as nanograms of AGP secreted per 106 cells or per micrograms of total protein at 48 h. Macrophage monolayers were scraped in 500 μl of 0.5 N sodium hydroxide, and total protein was determined according to the method of Lowry et al. (25) using a Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA).
Immunoprecipitation of [35S]methionine-pulse labeled cell lysates
Rat alveolar macrophages were pulse labeled with 100 μCi of [35S]methionine as previously described (26) and then lysed in 1 ml of a solution composed of: 20 mM Tris base (pH 8), 20 mM NaCl, 0.25% (w/v) sodium deoxycholate, 0. 1% (v/v) Triton X-100, 10 mM l-methionine, 2 mM PMSF, 1 mM benzamidine, and 10 μg/ml leupeptin. Cell culture supernatants were recovered and adjusted to correspond to the immunoprecipitation buffer (20 mM Tris base (pH 8), 5 mM EDTA, 0. 1% (v/v) Triton X-100, 10 mM l-methionine, 2 mM PMSF, 1 mM benzamidine, and 10 μg/ml leupeptin). Isolated rat hepatocytes were similarly treated to compare the apparent AGP m.w. of macrophages and hepatocytes. Immunoprecipitation was performed as described by Poüs et al. (27). Lysates and supernatants were precleared for 2 h at room temperature with 100 μl of Sepharose-protein A and 10 μl of rabbit nonimmune serum, and then immunoprecipitated overnight at 4°C with 100 μl of Sepharose-protein A and 5 μl of rabbit anti-AGP immune serum (provided by Prof. G. Durand, Laboratory of Biochemistry, Hopital Bichat, Paris, France). The immune complexes containing the AGP Ag adsorbed to Sepharose-protein A were washed four times with 1 ml of ice-cold immunoprecipitation buffer containing 0.15 M NaCl and 0.1% (w/v) SDS. Following the final wash, the Sepharose-protein A was pelleted in a microfuge for 5 min and solubilized in 30 μl of Laemmli sample buffer containing 5% (v/v) 2β-ME. The radiolabeled AGP was separated by SDS-PAGE in a 10% (w/v) polyacrylamide gel under reducing conditions. Radioactive bands were localized by autoradiography of the dry gel, and the blots were quantified using an electronic autoradiography device (Instant Imager, Packard, Groningen, The Netherlands). Incorporation of [35S]methionine into total TCA-precipitated protein was determined by precipitation with 10% (w/v) ice-cold TCA. The results were expressed as disintegrations per minute incorporated into AGP per 107 dpm of TCA-precipitable protein. Individual experiments were conducted in duplicate a minimum of three times.
AGP mRNA analysis
Quantitative molecular hybridization with cRNA probes and RNase protection assay were performed on cell lysates without prior RNA extraction, as previously described for hepatocytes (28) with minor modifications. Macrophages (1.5 × 106 cells) were scraped in 500 μl RNase-free PBS and centrifuged (5 min, 12,000 × g, 4°C), and the pellet was solubilized in 100 μl of 5 M GuSCN and 0.1 M EDTA, pH 7.0, then stored at −80°C until use. Cellular lysates (20 μL) were mixed with 2 μl of a mixture containing 105 dpm of labeled AGP cRNA probes and 2.5 pg of unlabeled sense AGP mRNA as an internal standard. AGP cDNA inserts (pAGP663) (29) were subcloned into pBluescript II SK+ phagemid vectors and provided by Dr. B. Lardeux (Institut National de la Santé et de la Recherche Médicale, Unité 327, Hopital Bichat). Riboprobe synthesis (286 nt that recognize exons 4–6 of the gene) was performed in the presence of [α-32P]UTP (50 μCi, 400 Ci/mmol) and T3 RNA polymerase after linearization of the vector by EcoO109I digestion. Unlabeled sense mRNA (164 nt) synthesis was performed in the presence of unlabeled nucleotides and T7 RNA polymerase after linearization by SalI digestion. After overnight hybridization at 37°C, the samples were treated with A and T1 RNases and subsequently exposed to proteinase K. After extraction with phenol/chloroform/isoamyl alcohol (50/49/1, v/v/v), the protected RNA:RNA hybrids were precipitated and loaded on a 6% acrylamide/bisacrylamide (19:1), 7.8-M urea denaturing gel. Quantitative analysis was performed from gels exposed to phosphorimager screens by direct counting of the gel through an electronic autoradiography device (Instant Imager, Packard, Groningen, The Netherlands). The signal ratio between protected fragments (intracellular mRNA/internal standard or sense mRNA) was calculated and then normalized for DNA content in identical aliquots of hybridized lysates. The amount of DNA was directly measured in guanidine thiocyanate-solubilized cells by fluorometry using the fluorochrome Hoechst 33258 as described by Labarca and Paigen (30). Isolated rat hepatocytes were used as positive controls for AGP gene expression (28).
The results are expressed as the mean ± SEM for at least three separate cultures or values. Statistical significance between groups was made using Student’s paired t test or the Mann-Whitney U test to compare unpaired groups.
Production of AGP by human alveolar macrophages is increased during inflammation
As shown in Fig. 1, human alveolar macrophages recovered by BAL from subjects with no pulmonary involvement spontaneously secreted AGP in culture supernatant (3 ng/million cells) during the 48-h culture period. This constitutive ex vivo production was increased in alveolar macrophages from patients with ongoing lung interstitial processes (12 ng/million cells at 48 h). It is interesting to note that this 4-fold increase in alveolar macrophage AGP production is comparable to that observed in plasma during systemic inflammation.
Synthesis and secretion of AGP by rat alveolar macrophages in vitro
We next examined, in vitro in the rat, the regulation of AGP gene expression in alveolar macrophages. Because hepatic AGP gene expression in rats is induced by glucocorticoids and cytokines, mainly TNF-α, IL-1β, and IL-6, we investigated whether these inflammatory mediators could modulate AGP gene expression in alveolar macrophages. We first analyzed de novo AGP synthesis and secretion by immunoprecipitation of pulsed-labeled protein. Fig. 2 shows a representative autoradiograph of a 10% PAGE that revealed a 45- to 47-kDa band and a 48- to 52-kDa band in rat alveolar macrophage cell lysates and supernatants, respectively. These sizes are similar to those we obtained with hepatocytes and to those described in the literature and correspond to a mature and sialylated form of AGP (27). Constitutive synthesis of AGP by unstimulated macrophages was low (control) and remained undetectable in culture supernatants after a 26-h incubation period including 8 h of pulse labeling. This synthesis was induced by the glucocorticoid Dex (1 μM) and by 20 ng/ml IL-1β used alone or in combination. Results obtained by immunoprecipitation of labeled protein in supernatants showed that AGP from Dex-treated cells was found in culture supernatants after an 8-h labeling period, while the IL-1β-induced increase in AGP synthesis was not followed by increased secretion at this time point. When cells were stimulated with the combination of Dex and IL-1β, most of the de novo synthesized AGP was secreted in culture supernatants. Incubation of macrophages with the two other cytokines involved in the up-regulation of hepatic AGP gene expression (TNF-α and IL-6) did not affect AGP production compared with that in the controls (data not shown).
Fig. 3 depicts the ratio of intracellular AGP vs secreted AGP at two different times of pulse labeling. After an 8-h pulse period most of the de novo synthesized AGP is localized within the cells (Fig. 3,A), and after a longer pulse period (20 h) the totality of the labeled AGP is found in supernatants (Fig. 3 B).
Inducible AGP mRNA levels in rat alveolar macrophages
Our data demonstrate that alveolar macrophages synthesize and secrete AGP similarly to hepatocytes and that AGP production is up-regulated by glucocorticoids and IL-1β. To determine the basis of the increase in macrophage AGP synthesis induced by the inflammatory modulators, AGP mRNA levels in whole cell lysates were analyzed by RNase protection assay. Fig. 4 shows an autoradiograph of a representative experiment of quantitative molecular hybridization with a 286-base AGP cRNA probe (lane 1). The band at 147 bases corresponds to the protected fragment for the sense mRNA or internal standard and was used to normalize hybridization ratio between samples (lane 2, internal standard alone). AGP mRNA from unstimulated cells (lane 3) or from 20 ng/ml IL-1β-treated cells (lanes 7–9) remained undetectable after 24 h of culture. Dex (1 μM) induced an increase in AGP mRNA steady state levels (lanes 4–6) as confirmed by the presence of a band at 230 bases corresponding to the protected fragments of the riboprobe with intracellular AGP mRNA. This enhancement was potentiated by IL-1β as illustrated in lanes 10–12.
Fig. 5 depicts the effects of inflammatory mediators on AGP gene expression in AM. AGP mRNA steady state levels were quantified as described in Materials and Methods and normalized to intracellular DNA content. AGP mRNA levels were barely detectable in unstimulated as well as in AM-CM- or cytokine-treated cells in the absence of Dex. Incubation of cells with 1 μM Dex induced an increase in AGP mRNA levels. This enhancement was significantly potentiated by AM-CM or IL-1β (2-fold increase; p < 0.05), while the 1.5-fold increase obtained with IL-6 remained statistically not significant. Fig. 6 depicts the time course of variations in AGP mRNA steady state levels in response to Dex and cytokines. The results are consistent with the data in Fig. 4 and show that Dex in the presence or the absence of cytokines induced AGP mRNA expression as early as 8 h and reached a maximum at about 30 h followed by a plateau over the subsequent 20 h. As previously discussed IL-1β in the presence of Dex induced a 2-fold increase in AGP mRNA levels (from 24–48 h) compared with Dex alone, while IL-6 had no significant effect.
Induced AGP mRNA expression is dependent on protein and RNA synthesis
To further investigate the molecular mechanisms involved in the regulation of AGP expression, macrophages were treated with 1 μM Dex with or without 40 ng/ml IL-1β in the presence or the absence of the protein synthesis inhibitor cycloheximide (CHX; at 10 μg/ml) or the polymerase II inhibitor dichlororibosidebenzymidazol (DRB; at 5 μg/ml). AGP mRNA levels were analyzed as previously described after a 24-h incubation period. Both the Dex-induced and the Dex- plus IL-1β-induced increase in AGP mRNA levels were abolished when the cells were concomitantly treated with CHX or DRB (Fig. 7). These results indicate that 1) de novo protein synthesis is required for the Dex-mediated and Dex- plus IL-1β-mediated increase in AGP mRNA level; and 2) these mediators up-regulate AGP gene expression by activating gene transcription rather than by stabilizing mRNA.
PGE2 and dbcAMP induced AGP mRNA expression in rat alveolar macrophages
As illustrated in Fig. 5, our results clearly indicate that maximal expression of the AGP gene in rat alveolar macrophages requires both glucocorticoids and inflammatory mediators contained in the supernatant of stimulated alveolar macrophages (AM-CM). Among the mediators secreted by alveolar macrophages, the cytokine IL-1β appeared to be the most effective in inducing AGP gene expression. We next investigated the effect of another major inflammatory product of macrophages, the lipid mediator PGE2. Stimulation of rat AM with PGE2 alone did not lead to any detectable AGP gene expression (data not shown). However, when used in combination with 1 μM Dex, PGE2 stimulated AGP gene expression in a dose-dependent fashion (Fig. 8,A). The value observed in the absence of PGE2 corresponds to the effect of Dex alone. We have previously shown that in macrophages PGE2 elevated intracellular cAMP concentrations and enhanced PKA activity (26). When the cell-permeable cAMP analogue dbcAMP was used, a similar dose-response curve was obtained (Fig. 8 B). These results indicate that in the presence of Dex, PGE2 induced an increase in AGP mRNA levels through a cAMP pathway.
AGP gene expression in human monocytes and rat peritoneal macrophages
To investigate whether the inducible AGP expression obtained in AM was observed in other cells from the monocyte/macrophage lineage, we measured the secretion of AGP in human monocyte culture supernatants and the steady state levels of AGP mRNA in rat peritoneal macrophages. Fig. 9 shows that human monocytes constituitvely secreted AGP (0.5 ng/106 cells/48 h) and that this basal production was not modulated by any of the stimuli used. It is interesting to note that this ex vivo spontaneous production by unstimulated monocytes was low compared with the basal levels obtained with human AM (3 ng/106 cells/48 h; Fig. 1). Fig. 10 depicts steady state levels of AGP mRNA in rat peritoneal macrophages compared with those in rat AM. AGP mRNA, analyzed by RNase protection assay as described previously, remained undetectable in rat peritoneal macrophages even after incubation of cells with IL-1β, PGE2, or dbcAMP in the presence of 1 μM Dex. These results suggest that AGP gene expression is inducible in AM by inflammatory mediators, but not in monocytes or in peritoneal macrophages.
In the present study we show that AGP gene expression is inducible by inflammatory mediators in AM recovered from human and rat cells, but not in extrapulmonary cells from the monocyte/macrophage lineage. Furthermore, we demonstrated that in rat AM, AGP gene expression is up-regulated by PGE2, another major inflammatory mediator, which acts in macrophages by activating cAMP-dependent protein kinase.
We have previously demonstrated that AGP gene expression was inducible in the rat lung in vivo during localized or systemic inflammation and that immunoreactive AGP was produced in the human lung in the course of a local inflammatory reaction. Immunohistochemistry and in situ hybridization results showed that alveolar epithelial cells were the primary source of AGP production in human and rat lung. However, while immunoreactivity was never detected in bronchial epithelium or in endothelial cells, positive signal was observed in cells located in the alveolar lumen consistent with desquamated alveolar epithelial cells or alveolar macrophages (20). Indeed, we show in the present work that immunoreactive AGP is spontaneously secreted by human AM ex vivo and is increased in patients with interstitial lung involvement. Thus, alveolar macrophages, which play an important role in the regulation of inflammatory and immune responses, produce and secrete AGP, an acute phase reactant with immunomodulatory properties.
We next examined AGP gene expression in extrapulmonary cells of the monocyte/macrophage lineage in vitro. Our results revealed that human monocytes in vitro produced AGP constitutively, whereas AGP mRNA levels remained undetectable in cell lysates from rat peritoneal macrophages cultured in vitro even when stimulated with the most potent inducer of AGP gene expression (Dex plus IL-1β). It is interesting to note that none of the modulators of the inflammatory response we used to induce AGP gene expression in AM was able to stimulate significantly human monocyte AGP basal production.
Thus, the inducible expression of AGP by AM seems tissue specific (no expression in peritoneal macrophages) and appears to be acquired during cell differentiation from blood monocytes into AM.
We were interested in examining the molecular basis of the increased secretion of AGP from patients with interstitial lung disease. To accomplish this we studied, in rat AM in vitro, the effect of inflammatory mediators that have already been shown to modulate AGP gene expression in hepatocytes. Immunoprecipitation of AGP from [35S]methionine-labeled rat AM cells indicated that these cells synthesized and secreted a protein that corresponded to the molecular mass of AGP from hepatocytes (45 and 50 kDa from cell lysates and supernatants, respectively, which represents the mature protein) (31). Stimulation of rat AM with Dex increased both the synthesis and the release of AGP, and this increase was about 20-fold more than that seen in rat ATII cells (70 ± 8 ng/24 h/μg protein) (20). This Dex-induced increase was potentiated by treatment of the AM with IL-1β, with most of the newly synthesized protein found in the supernatant. The analysis of AGP mRNA by RPA showed a similar pattern, i.e., up-regulation of the transcript by Dex and potentiation of this effect by IL-1β- and LPS-stimulated macrophage-conditioned media. These results are in agreement with those reported by others who have demonstrated that IL-1β modulates acute phase protein expression at both the transcriptional and post-translational levels by increasing protein secretion (32).
The expression of AGP in response to cytokines was strictly dependent on treatment with Dex. Similar results have been observed in rat hepatocytes and hepatoma cell lines (33, 34) as well as in ATII cells (20). In vivo, we also found that glucocorticoids are potent inducers of AGP gene expression by rat AM. Indeed, AM isolated from Dex-injected rats expressed AGP mRNA (data not shown).
The regulation of AGP gene expression has been extensively studied in vivo in rats in various hepatoma cell lines and in hepatocyte primary cultures. It has been well established that AGP expression is increased by IL-1, TNF-α (35), IL-6 (and related cytokines) (6, 36, 37), and glucocorticoids (5). In AM, our results showed that in the presence of Dex, IL-1β is a more potent inducer of AGP expression than IL-6, which is in agreement with previous studies (38). Surprisingly, TNF-α had no effect (data not shown).
Hepatic AGP mRNA levels are regulated in vivo at both transcriptional (5) and post-transcriptional (39) levels by glucocorticoids and acute phase mediators. In AM, results obtained using the RNA polymerase II inhibitor DRB indicated that the induced increase in AGP gene expression was at least partially regulated at the level of transcription. Treatment of cells with cycloheximide was also able to block the expression of AGP mRNA, indicating the involvement of protein factors. In hepatocytes, similar results have been described, and the de novo synthesis of nuclear proteins belonging to the C/EBP family (CCAAT/enhancer binding protein), especially the inducible isoforms of C/EBP (C/EBPβ and C/EBPδ), was shown to be required for the induction of AGP gene expression under acute phase conditions (40, 41). Baumann et al. (42) demonstrated that trans-activation of C/EBP by glucocorticoids is necessary for AGP gene expression. Thus, the involvement of C/EBP isoforms in the mediation of AGP gene up-regulation in alveolar macrophages is likely, because the up-regulation requires both new protein synthesis and the presence of glucocorticoids.
In previously reported work we have shown that in macrophages, gene expression of insulin-like growth factor I, a growth factor implicated in normal wound healing and pathologic tissue fibrosis, was up-regulated by PGE2 and cAMP by a TNF-α-independent pathway. We concluded that PGE2 acts as a new positive stimulus for insulin-like growth factor I synthesis through a cAMP/protein kinase A pathway (26). Herein, we demonstrated that in addition to cytokines another major inflammatory product of macrophages, the lipid mediator PGE2, increased AGP mRNA levels in the presence of Dex through a cAMP/PKA signal transduction pathway. Furthermore, we provided evidence in a previous study (43) that PGE2 was the arachidonic acid metabolite preferentially secreted by macrophages during inflammation (stimulation by TNF-α), whereas control cells mainly produced PGD2. It is interesting to note that PGD2, which does not increase PKA activity (43), does not increase AGP gene expression (data not shown). The up-regulation of AGP gene expression by PGE2 and cAMP has not been previously described in any cell type. When the modulation of AGP production by PGE2 and cAMP was studied in hepatocytes, no changes were observed (data not shown).
It is known that C/EBPβ requires phosphorylation at several functional domains, such as nuclear translocation and DNA binding. The cAMP-induced increase in AGP gene transcription may be mediated by phosphorylation and trans-activation of a transcription factor belonging to the C/EBP family. In this connection, Metz and Ziff showed that cAMP and the protein kinase A activator (forskolin) stimulate the C/EBP-related transcription factor NF-IL-6 to trans-locate to the nucleus and induce transcription of the cellular proto-oncogene c-fos (44).
The physiological roles of many acute phase proteins remain unclear, but most of them prevent tissue damage associated with inflammation. It has been shown that AGP protected mice from lethal shock induced by TNF or endotoxin when given at least 2 h before the challenge at doses similar to AGP serum concentrations obtained during an acute phase response (15). It has been suggested that the potent platelet aggregation-inhibitory activity of AGP (45) and its potent inhibition of neutrophil chemotaxis and oxidative metabolism (11) underlie its protective properties. The immunosuppressive effect of AGP has also been linked to its highly sialylated carbohydrate moieties (46). Inflammation induces the expression of sialyl-Lewis X-containing glycan structures on AGP in human serum (19). Sialyl-Lewis X is the ligand for the cell adhesion molecules E-selectin and P-selectin, which are involved in the inflammation-dependent adhesion of neutrophils, monocytes, or resting T cells to endothelial cells or platelets (47). Because sialylated oligosaccharides have a protective effect in vivo in immune complex-induced lung injury in rats (48), the inflammation-induced increase in sialyl-Lewis X-substituted glycans on AGP might represent a mechanism for feedback inhibition of granulocyte extravasation into inflamed tissues as prevention of lung injury. The immunomodulatory properties of AGP extends to the control of cytokines and cytokine antagonist secretion by monocytes-macrophages in vitro (12, 13, 14). Due to these properties, AGP production in the alveolar space during pulmonary inflammation may exert a local protective effect by limiting the inflammatory reaction and its potentially deleterious effect on alveolar structures.
Taken together, our data suggest that AGP expression in AM is inducible in vivo and in vitro during inflammation, while no variations are observed in monocytes or in peritoneal macrophages. Furthermore, PGE2 and cAMP represent a new activation pathway for AGP gene expression that seems specific to AM.
We thank Bernard Lardeux for helpful discussion and technical support, Anne Barnier for human AGP ELISA, and Naima Viires for helpful review of the manuscript.
This work was supported by Institut National de la Santé et de la Recherche Médicale (National Institute of Health and Medical Research) and a grant from Zeneca Pharma.
Abbreviations used in this paper: AGP, α1-acid glycoprotein or orosomucoid; dbcAMP, dibutyryl cAMP; Dex, dexamethasone; GuSCN, guanidine thiocyanate; PKA, cAMP-dependent protein kinase; AM, alveolar macrophages; BAL, bronchoalveolar lavage; AM-CM, LPS-stimulated alveolar macrophage-conditioned medium; CHX, cycloheximide; DRB, dichlororibosidebenzymidazol.