Main

The major physiologic function of pulmonary surfactant (PS) is to confer mechanical stability to alveoli(1). However, there is growing evidence that PS also has a potential role in modulating inflammation in normal and injured lungs(2). In acute lung injuries, polymorphonuclear neutrophils and monocytes originating in the blood stream and accumulating within the lung become activated and participate in lung inflammation by producing ROS and releasing proinflammatory cytokines, AA metabolites and proteolytic enzymes(3). This pathway occurs in nRDS and ARDS, resulting in diffuse alveolar damage(4, 5).

Primary PS deficiency results in nRDS. However, secondary alterations of PS also have been described in ARDS(6). As a result of these findings, surfactant replacement therapy with either natural or synthetic exogenous surfactant is now currently used in nRDS(7). More recently it has been tested as a therapeutic modality in ARDS(8, 9).

The effects of the hydrophilic surfactant-associated proteins SP-A and SP-D, have been extensively studied in human and animal AM. SP-A, for example, has been shown to stimulate chemotaxis of AM, to enhance bacterial and viruses phagocytosis, and also to induce production of ROS, which suggests that it plays a significant role in the defense system of the lung against infiltrating microorganismes and viruses(10).

The role of the lipid components of PS has scarcely been studied. Incubation of rabbit, human, or guinea pig AM with the lipid components of PS has been shown to inhibit the stimulated production of ROS(11), to suppress the stimulated secretion and synthesis of proinflammatory cytokines, such as TNF-α, IL-1β (IL-1), and IL-6(12, 13), and also to inhibit AA metabolite production, such as PGE2 and TxB2(14).

Much less information is available about the role of the PS lipids on monocytes. In human monocytes, a 24-h preincubation with the lipid components of PS has been shown to suppress the stimulated secretion of TNF-α(15). However, in the same study, the stimulated production of ROS was not modified by a short preincubation with PS lipids(15).

The aim of the present study was to investigate the effects of PS lipids on proinflammatory functions in both resting and stimulated human peripheral blood monocytes after an overnight preincubation period with a well defined, modified natural surfactant of porcine origin (Curosurf; Chiesi Pharmaceutici, Parma, Italy), that is currently used in nRDS and testing for ARDS(16–18). The target functions examined in this initial study were: production of superoxide anions; AA metabolites, such as PGE2, TxB2, and LTC4; TNF-α; and cell adherence. We provide experimental evidence that favors the anti-inflammatory role of modified natural porcine surfactant (Curosurf) in human monocytes in vitro.

METHODS

Modified natural porcine surfactant. Curosurf is a modified natural surfactant that is isolated from minced pig lungs by chloroform-methanol extraction and liquid-gel chromatography. It contains 99% polar lipids, mainly phospholipids with >40% dipalmitoyl phosphatidylcholine, and about 1% hydrophobic surfactant-associated proteins(SP-B and SP-C). The extraction procedure removes SP-A(16). Curosurf was used in a concentration of 500 and 1000 μg/mL. These concentrations were similar to natural porcine and synthetic surfactant concentrations used previously in experiments with monocytes and AM(12, 13, 15).

Isolation and stimulation of monocytes. Human mononuclear cells from buffy coats taken from normal adult donors were isolated by Ficoll-Hypaque density gradient and monocytes purified by adherence as described previously(19). Adherent monocytes were cultured overnight in RPMI medium (Life Technologies, Inc., Paisley, Scotland) supplemented with 1% glutamine and 10% FCS (Life Technologies) at 37°C in a humidified atmosphere containing 95% air and 5% CO2 in the presence or absence of Curosurf (500 or 1000 μg/mL) dissolved in supplemented RPMI medium.

For stimulating the release of PGE2, TxB2, and TNF-α, monocytes were incubated overnight with OM-85 (1 mg/mL) (OM Laboratories S.A., Geneva, Switzerland) or LPS (10 μg/mL) (from Escherichia coli 055:B5, Difco, Detroit, MI) in the presence or absence of Curosurf (500 and 1000 μg/mL). OM-85 is a standard bacterial extract obtained by controlled alkaline hydrolysis of different strains of bacteria that have been shown to stimulate in phagocytes, cytokines production, calcium mobilization, NADPH oxidase activation and subsequent production of superoxide anions(20). Supernatants were recovered and centrifuged at 220× g; aliquots were kept at -20°C until assayed for TxB2, PGE2, and TNF-α content.

To stimulate LTC4 release, adherent monocytes were incubated overnight in a supplemented RPMI medium in the presence or absence of Curosurf(500 and 1000 μg/mL). Cells were then washed twice with PBS and stimulated with 10 μM calcium ionophore A23187 in 1 mL of fresh RPMI medium for 30 min(dissolved DMSO and diluted to 0.1% final vol/vol in RPMI medium, Sigma Chemical Co., St. Louis, MO). Supernatants were stored at -20°C until assayed.

Measurement of superoxide anions production. Superoxide anions production was measured by the superoxide dismutase-inhibitable reduction of ferricytochrome c. Adherent monocytes were cultured overnight in the presence or absence of Curosurf (500-1000 μg/mL). Cells were then washed twice with PBS, incubated in a superoxide buffer, and stimulated for 30 min at 37°C with either PMA (100 ng/mL) or the bacterial extract OM-85 (1 mg/mL). To determine the contribution of PKC on the observed effects on superoxide anions production, cells were preincubated for 10 min before stimulation with or without staurosporine (1 μM)(21).

Determination of arachidonic acid metabolites. LTC4, PGE2, and TxB2, the stable metabolite of TxA2, were measured in the supernatants by direct enzyme immunoassay. Tracers linked to acetylcholinesterase, specific rabbit antisera, mouse monoclonal anti-rabbit IgG, lyophilized synthetic standards and the enzyme substrate (Ellman's reagent) were purchased from Cayman, Ann Arbor, MI. Enzyme immunoassay, used to measure eicosanoids, was performed according to the instructions of the manufacturer and as reported by Pradelles et al.(22). The sensitivity of the assays was approximately 30 pg/mL for LTC4, 10 pg/mL TxB2, and 20 pg/mL for PGE2.

Determination of TNF -α. A two-site ELISA for human TNF-α was developed using a monoclonal-capture antibody (mouse IgG1k code M 300 A, Endogen, Boston, MA), a polyclonal-detection antibody(anti-human TNF-α, from rabbit serum, code P300A, Endogen) and recombinant TNF-α (code RTNF-α 10, Endogen). The lowest limit of detection was approximately 3 pg/mL.

Additional methods. Viability of the purified monocytes after overnight exposure to Curosurf (500 or 1000 μg/mL) was unchanged compared with control cultures assessed by excluding trypan blue and measuring of the extracellular lactate dehydrogenase in both supernatants and adherent cells, as described by Bergmeyer et al.(23) (data not shown).

At the end of each experiments, the monocyte monolayers were washed with sterile saline solubilized in 1 mL of 0.1% Triton X-100 (Sigma Chemical Co.) and scrapped off the dishes. Protein content of the cell lysates was determined using the BCA protein assay kit (Pierce, Rockford, IL) and BSA as a standard. The conversion factor used for correction of the number of adherent cells was 200 μg of protein corresponding to 1 × 106 cells.

This latter method was used to assess the effect of Curosurf on adherence after exposing or not exposing monocytes to Curosurf (500-1000 μg, overnight). Adherence was subsequently calculated by the ratio of adherent cells to total cells and expressed as a percent of adherence.

Data analysis. All data are expressed as a means, ± SEM. Between groups, differences were analyzed by analysis of variance using a Kruskal-Wallis test followed by a Mann-Withney t test, if appropriate. A p value of <0.05 was considered statistically significant.

RESULTS

Effects of surfactant on superoxide anions production. Curosurf at both concentrations tested (500 or 1000 μg/mL) did not influence the small basal level of superoxide anions production (0.036 ± 0.007 nmol/106cells/min, n = 6) of unstimulated monocytes incubated overnight.

As shown in Figure 1A, in response to PMA (100 ng/mL) monocytes generated significantly higher levels of superoxide anions (0.636± 0.123 nmol/106cells/min, n = 6) compared with unstimulated cells. However, PMA-induced superoxide anions production was not affected (n = 6) by the concentrations of Curosurf used (500 or 1000μg/mL).

Figure 1
figure 1

Effects of modified natural porcine surfactant(Curosurf) on superoxide anions production by stimulated human monocytes. Human monocytes were preincubated overnight both with and without Curosurf at the concentrations of 500 and 1000 μg/mL, and then stimulated with either(A) PMA (100 ng/mL) or (B) bacterial extract OM-85 (1 mg/mL). In some experiments monocytes were preincubated with staurosporine (1μM, 10 min), an inhibitor of PKC, before being stimulated. Vertical bars represent a mean ± SEM (n = 6). *p < 0.005 vs stimulated monocytes with either PMA or OM-85 in the absence of Curosurf.

Full size image

By comparison, as Figure 1B shows, although OM-85 (1 mg/mL) also induced a significant production of superoxide anions (0.211± 0.039 nmol/106 cells/min, n = 6), an overnight preincubation with Curosurf at a concentration of 1000 μg/mL caused a significant inhibition of the OM-85-induced superoxide anions production(0.068 ± 0.008 nmol/106 cells/min, n = 6) (p< 0.005).

As expected, the PMA-induced superoxide anions production was suppressed by 10 min of preincubation with staurosporine (1 μM) (p < 0.005), which agrees with the involvement of PKC activation in the mechanism of superoxide production by PMA. In contrast, preincubation with staurosporine (1μM) did not affect the OM-85-induced superoxide production. These results indicate that the mechanisms involved in superoxide production are different for PMA and for OM-85, and that Curosurf interferes essentially with the latter.

Effects of surfactant on arachidonic acid metabolite production. The effect of Curosurf on the production of AA metabolites derived from the 5-lipoxygenase pathway, such as LTC4, was evaluated in supernatants taken from monocytes preincubated overnight with Curosurf (500 or 1000 μg/mL) and then stimulated, or not, by the 5-lipoxygenase stimulus Ca2+ ionophore 23187 (10 μM, 30 min) in a Curosurf-free culture medium(24). The concentrations of Curosurf tested did not influence the small basal level of LTC4 (140.0 ± 25.6 pg/106cells/30 min, n = 6), produced by unstimulated monocytes. However, as shown in Figure 2A, Curosurf reduced, in a dose-dependent manner, the release of LTC4 induced by A23187 (3435.9 ± 777.6 pg/106 cells/30 min, n = 8), resulting in a significant reduction of LTC4 at the concentration of 1000 μg/mL (1295.7 ± 201.7 pg/106 cells/30 min, n = 8, p < 0.05). Because LTC4 release was evaluated in supernatants from monocytes stimulated in a Curosurf-free medium, it was apparent that the inhibitory effects observed with Curosurf were not reversible, at least within the time allowed for these experiments.

Figure 2
figure 2

Effects of modified natural porcine surfactant(Curosurf) on the production of LTC4, PGE2, and TxB2 by stimulated human monocytes. LTC4 production (A) was evaluated in supernatants of human monocytes after overnight preincubation with Curosurf at the concentrations of 500 and 1000 μg/mL and then stimulated in a PS-free medium with Ca2+ ionophore 23187 (10 μM, 30 min). PGE2 (B) and TxB2 (C) production was evaluated after overnight preincubation with bacterial extract OM-85 (1 mg/mL), alone or combined with Curosurf at the concentrations of 500 and 1000μg/mL. Vertical bars represent a mean ± SEM (n = 8).*p < 0.05, vs stimulated monocytes in absence of Curosurf.

Full size image

The production of AA metabolites derived from the cyclooxygenase pathway, such as PGE2 and TxB2, was evaluated in supernatants taken from monocytes cultured overnight in supplemented RPMI medium alone (control condition), with OM-85 (1 mg/mL) (stimulated condition), or with Curosurf(500-1000 μg/mL) alone or combined with OM-85. Curosurf at both concentrations did not influence the basal levels of PGE2 (19.8 ng± 10.7/106 cells/18 h, n = 4) and TxB2 (151.8 ng ± 60.6/106 cells/18 h, n = 4). The production of PGE2 and TxB2 was significantly increased in monocytes exposed to the bacterial extract OM-85 (1 mg/mL), compared with control conditions(p < 0.05, n = 8). When cultures were prepared in the presence of combined OM-85 (1 mg/mL) and Curosurf (500-1000 μg/mL), compared with the OM-85-stimulated values, the release of both PGE2 and TxB2 was reduced. As shown in Figure 2,B and C, a dose-response inhibition was observed to have the most significant effect with the largest dose used (p < 0.05, n = 8).

Effects of surfactant on TNF -α. The effect of Curosurf on the production of TNF-α was evaluated in supernatants of monocytes cultured overnight either in a supplemented RPMI medium alone, control conditions, with OM-85 (1 mg/mL) or LPS (10 μg/mL), stimulated conditions, or with Curosurf (500-1000 μg/mL) alone or combined with OM-85 or LPS. Curosurf did not influence the basal value of TNF-α release(444.1 ± 126.9 pg/106 cells/18 h, for OM-85 assays, n= 9, and 565.6 ± 201.7 pg/106 cells/18 h for LPS assays,n = 5). The release of TNF-α by monocytes stimulated with OM-85 or LPS was significantly increased over the basal value (2850.2 ± 1020.9 pg/106 cells/18 h, n = 9, p < 0.005 and 7906.4 ± 1654.9 pg/106 cells/18 h, n = 5, p< 0.01, respectively). As shown in Figure 3,A and B, the OM-85 or LPS-stimulated release of TNF-α also was significantly inhibited in a dose-dependent manner with both concentrations of Curosurf(with p < 0.05 (500 μg/mL) and <0.01 (1000 μg/mL) for OM-85, n = 9, and p < 0.05 (500-1000 μg/mL) for LPS,n = 5).

Figure 3
figure 3

Effects of modified natural porcine surfactant(Curosurf) on the production of TNF-α by stimulated human monocytes. Human monocytes were preincubated overnight with OM-85 (1 mg/mL) (A) or LPS (10 μg/mL) (B), alone or combined with Curosurf at the concentrations of 500 and 1000 μg/mL. Vertical bars represent a mean± SEM. *p < 0.05, †p < 0.01vs stimulated monocytes with OM-85 (n = 9) or LPS(n = 5) in absence of Curosurf.

Full size image

Effects of surfactant on adherence. As shown in Figure 4, adherence calculated as the ratio of adherent cells to the total unstimulated monocytes cells preincubated overnight with 500 or 1000 μg/mL of Curosurf decreased from 90.8 ± 3.0% to 71.3%± 5.0 and 63.4 ± 4.2%, respectively (p < 0.001)(n = 8), compared with controls in a dose-dependent fashion. However, as shown in the above results, this small decrease in adherence did not influence, either the basal values of the measurements or the capacity of the cells to respond to subsequent stimulations.

Figure 4
figure 4

Effects of modified natural porcine surfactant(Curosurf) on the adherence of human monocytes. Unstimulated monocytes were preincubated overnight both, with (500 and 1000 μg/mL) and without Curosurf(controls). Adherence was calculated as the ratio of adherent to total cells and expressed in percent. Vertical bars represent a mean ± SEM (n = 8). *p < 0.001 vs controls.

Full size image

DISCUSSION

The present study shows that in stimulated human monocytes, an overnight preincubation with a modified natural porcine surfactant (Curosurf) was able to inhibit the production of superoxide anions in a dose- and stimulus-dependent manner, and the release of AA metabolites (PGE2, TxB2, and LTC4) and TNF-α in a dose-dependent fashion. In addition, Curosurf also decreases the spontaneous adherence of monocytes to plastic culture wells, again in a concentration-dependent fashion.

Previous studies on the effect of PS in regard to superoxide anions production by AM or monocytes have yielded conflicting results. Several reports indicate that PS can either enhance(25), suppress(11, 14), or not modify(15) the production of superoxide by AM or monocytes. We believe that these contradictory results can be explained by differences either in the composition (i.e. natural, natural modified, or synthetic surfactants) or concentrations of PS used, variations in interspecies or cell type, or differences in experimental conditions. The present study indicates that the type of stimulus used to evoke the production of superoxide anions by monocytes may be also important. We have previously shown that the mechanisms involved in the production of superoxide anions by monocytes are different for PMA and for the bacterial extract OM-85 and that OM-85 increases Ca2+ mobilization, which suggests that NADPH oxidase activation is in this case a receptor-mediated event(20). The experiments we performed using staurosporine, an inhibitor of PKC, indicate that, in contrast with PMA, the production of superoxide anions resulting from stimulation with bacterial extract OM-85 (1 mg/mL) is not related to PKC activation(21). Consequently, we propose that the mechanism involved in the suppressive effect of Curosurf is PKC-independent.

The observed suppressive effect of Curosurf on the release of PGE2 and TxB2 by OM-85-stimulated monocytes also agrees with a recent study reporting a dose-dependent suppressive effect of natural guinea pig surfactant, or several of its lipid components, on the production of cyclooxygenase metabolites (PGE2 and TxB2) in guinea pig AM stimulated by various proinflammatory stimuli(14).

The suppressive effect of Curosurf on LTC4 production, elicited by the Ca2+ ionophore A23187, has not been described previously. The fact that LTC4 release was suppressed without the continued presence of Curosurf indicates that the mechanism involved in the suppression does not result from the sequestration of the agonist by lipid micelles derived from Curosurf lipids.

Taken together, these data suggest that, in monocytes, Curosurf could interfere with intracellular signaling that leads to the activation of phospholipase A2 and NADPH oxidase and consequently could suppress the Ca2+-dependent release of AA and superoxide anions, respectively. However, the precise mechanisms by which Curosurf mediates these suppressive effects remains unknown. Interferences with membranous phospholipids, resulting from incorporations of PS lipids by monocytes, might explain the alterations in transmembrane signaling. Accordingly, perturbations of the membrane receptors' responsiveness from exposure to PS have been recently reported in lymphocytes. Yarbrough and colleagues(26) reported that natural human surfactant suppresses the CD3-mediated intracellular calcium influx in lymphocyte T cells and Roth and colleagues(27) reported that modified natural bovine surfactant reduces the binding of lymphokine-activated killer cells to tumor targets and induces a parallel reduction in the expression of adhesion molecules involved in inflammation such as CD2, LFA-1, LFA-3, and ICAM-1.

The dose-dependent inhibition by Curosurf of the TNF-α-stimulated release generated by OM-85 or LPS that we observed agrees with previous results obtained with natural modified or synthetic surfactant in both AM and monocytes of human origin(12, 13, 15, 28). This suppressive effect is not specific to TNF-α, because the stimulated production of other proinflammatory cytokines, such as IL-1 and IL-6, were also suppressed by modified or synthetic surfactant(12, 13). Moreover, this suppressive effect is not a stimulus- or cell-specific, as it is observed for the production of cytokines by AM and monocytes stimulated with various stimuli, such as LPS, IL-1, bacterial extract or phagocytosis of Staphylococcus aureus(12, 13, 15, 28). Additionally, this suppressive effect is not specific to the type of surfactant used, because it was described either with a synthetic surfactant (Exosurf), which consists of dipalmitoyl phosphatidylcholine, cetyl alcohol and tyloxapol; or with modified natural bovine (Survanta), or porcine (Curosurf) surfactant, which contains several different phospholipids and the surfactant-associated proteins SP-B and C; or with the purified phospholipids fraction of Curosurf(12, 13, 15, 28). However, the precise mechanism for the suppressive effects of PS remain hypothetical and deserve further study. Recent studies suggests that the suppressive effects of PS lipids may, in part, involve transcriptional regulation through inhibition of nuclear factor-κB activation(13, 29).

The slight decrease we observed in the basal values of monocytes' spontaneous adherence to plastic wells after overnight preincubation of monocytes with Curosurf has not been reported by Thomassen et al.(12) and Speer et al.(15). However, these latter and the present studies are not easily comparable. In Speer et al.(15) the time allowed for preincubation with Curosurf (30 min) and the concentration used (1-100 μg/mL) were, respectively, shorter and 10-fold less than the present study, and may have masked the inhibitory effects of Curosurf on adherence. In Thomassen et al.(12), both the cell type (AM) and the surfactant used(Exosurf) were different from the present study. The mechanisms involved are still unknown. However, the previously reported down-regulating effect of surfactant on the expression of adhesion molecules might be an explanation, and is currently being explored in our laboratory(27). Although we cannot completely rule out the possibility that the slight decrease in adherence that we report explains the suppressive effect of Curosurf on the proinflammatory mediators that we observed in our present study. As a matter of fact, adherence to plastic has been shown to induce selective mRNA expression of monocyte mediators, such as TNF-α, IL-1, and IL-6, and to prime monocytes(30–32). The specificity of the down-regulating effect of Curosurf renders this hypothesis unlikely. Indeed, cells preincubated with Curosurf still respond to PMA, as normal cells do, although there adherence is less.

In summary, the present study provides experimental evidence in favor of an anti-inflammatory role of Curosurf in human monocytes in vitro. Further studies are necessary to define precisely the mechanisms involved. If confirmed in vivo such an anti-inflammatory property may explain, at least in part, the beneficial therapeutic effects of modified natural porcine surfactant in nRDS and ARDS.