Main

Surfactant replacement therapy in severe neonatal RDS effectively lowers neonatal mortality and the incidence of acute lung damage in preterm infants. Despite this, however, a considerable number of infants still develop chronic lung disease or BPD(1). The pathogenesis of BPD has been only partly elucidated. Bronchoalveolar inflammation is apparently a key feature in the pathogenic developments related to this disease(2). It is particularly neutrophils and alveolar macrophages that are found in large numbers in the airways and in pulmonary interstitium of infants with RDS in whom BPD manifested at a later stage(3,4). Alveolar macrophages are thought to play a central role in the inflammatory response(5). They enhance the alveolar capillary leakage observed in infants with BPD(2) by releasing various chemotactic cytokines and other inflammatory mediators. These cytokines and mediators strongly affect intra-alveolar neutrophil recruitment. There is growing evidence that macrophage-derived TNF-α significantly provokes the inflammatory reaction(6,7). TNF-α is known to trigger the release of various inflammatory mediators in alveolar cells(8). TNF-α also activates neutrophils. This has been shown to cause severe oxidative and proteolytic tissue damage(3,9). In preterm infants with RDS, increased levels of TNF-α were detected in respiratory fluids(10).

We have previously shown that Curosurf® (Serono, Unterschleißheim, Germany), a biochemically well defined natural porcine surfactant, effectively suppressed TNF-α secretion by resting and through LPS-stimulated monocytes in a dose-dependent manner(11). Surfactant may, therefore, represent an important anti-inflammatory agent in the lung.

Monocytes, the alveolar macrophages precursors, which were co-cultivated with a purified phospholipid preparation, released even less TNF-α than those monocytes that were exposed to natural porcine surfactant(11). Lipid components of lung lining material are evidently related to the immunomodulating effects of surfactant(12,13). Reports have described that alveolar macrophages(14) can lower the secretion of proinflammatory substances such as superoxide anions, arachidonic acid derivatives(15), and TNF-α when influenced by phospholipids.

It is well known that different effects of TNF-α are mediated by different receptors(16). Although TNF-α type I receptor is evidently connected with the process of apoptosis, TNF-α-RII is involved mainly in the event of inflammatory response.

The aim of this study was to examine the cellular mechanisms involved in the modulation of TNF-α synthesis by means of Curosurf®. We were specifically interested in determining whether the surfactant exerts a direct effect on intracellular processes at the level of transcription from DNA into mRNA. We compared levels of mRNA for TNF-α and TNF-α-RII using a semiquantitative technique (PCR).

TNF-α-RII was examined as it predominates in the mediation of cellular activation induced by LPS and TNF-α. We examined the premise that the selective up-regulation of soluble and membrane-bound TNF-α-RII, which Leeuwenberg et al.(17) and Winzen et al.(18) described, would produce measurable changes on the mRNA level in stimulated monocytes.

METHODS

Enrichment of monocytes. Mononuclear cells from healthy adult donors (n = 6) were isolated by means of Ficoll-Hypaque density gradient. Cells were resuspended in RPMI 1640 medium (Seromed®, Biochrom, Berlin, Germany) containing 10% FCS, penicillin-streptomycin (20 U/mL), and L-glutamine (100 µg/mL). A number of mononuclear cells that contained 1× 106 monocytes (determination through nonspecific esterase stain; Technicon, Terrytown, NJ) were plated in tissue culture dishes with a surface of modified polystyrene (Becton Dickinson, Falcon® 3802, Lincoln Park, NJ). Purification of monocytes was performed through adherence during a 2-h period at 37°C in a humidified atmosphere containing 5% CO2. Nonadherent cells were removed after culture by gently aspirating the supernatant fluid and three washes using the medium. The resulting cell preparation contained approximately 90% monocytes as identified by nonspecific esterase stain. A trypan blue exclusion test revealed that more than 98% of the monocytes were viable.

TNF-α release by monocytes. Fresh medium was added to the adherent (enriched) monocytes. In vitro cultivation was done in quadruplicate over 24 h at 37°C in a humidified atmosphere containing 5% CO2, and by employing three different culture conditions (A, B, and C). Control cells were cultivated in RPMI 1640 (A). In contrast, monocytes were either stimulated by LPS at a concentration of 2.5 µg/mL(B), or by LPS combined with the natural porcine surfactant Curosurf® at a concentration of 500 µg/mL (C). LPS, which had been produced by means of the Escherichia coli strain 026:B6, was purchased at Sigma Chemical Co. (Deisenhofen, Germany). Earlier experiments have shown that the concentration of LPS quoted above induces a maximal release of TNF-α(11). The supernatant was removed and immediately frozen at -70°C. TNF-α protein concentrations were measured by means of the ELISA technique (T-Cell Sciences, Cambridge, MA).

RNA isolation. RNA was isolated using the reagent system available in the InViSorb™ RNA Kit II (InViTek, Berlin, Germany). Lysing solution was added to culture dishes and transferred to 1.4-mL Eppendorf reaction cups (Eppendorf-Netheler-Hinz, Hamburg, Germany). Adsorbin solution was added, and DNA-adsorbin complexes were precipitated. Extraction of total monocyte RNA was performed by applying a phenol/chloroform extraction technique. RNA concentration and protein contamination in each sample (n = 4) was quantified by UV absorbance at 260 and 280 nm.

Semiquantitative reverse transcriptase PCR. A volume corresponding to 1 µg of RNA was used for transcription into cDNA using 9 U of a RAV-2 reverse transcriptase (Amersham Life Science, Cleveland, OH). In a second step, 25 ng of cDNA, in triplicate from each sample, served as source material for PCR. Specific primer pairs for human TNF-α(5′ primer GAG TGA CAA GCC TGT AGC CCA TGT TGT AGC A, 3′ primer GCA ATG ATC CCA AAG TAG ACC TGC CCA GAC T), human TNF-α-RII (5′ primer GAA TAC TAG GAC CAG ACA GCT CAG ATG TGC, 3′ primer TAT CCG TGG ATG AAG TCG TGT TGG AGA ACG) were used. The PCR product of humanβ2-microglobulin amplification (5′ primer ACC CCC ACT GAA AAA GAT GA, 3′ primer ATC TTC AAA CCT CCA TGA TG) served as an internal standard. Primer pairs were purchased at Clontech (Palo Alto, CA). For the polymerase chain reaction, 2.5 U of Taq-DNA polymerase (Amersham Life Science) were used on a Biomed60 thermocycler (Biomed, Oberschleißheim, Germany). With the aim of detecting PCR signals in the linear phase of product amplification, 20-33 PCR cycles were performed. PCR products were analyzed using a polyacrylamide gel (8%). Gels were stained with ethidium bromide and examined under a UV transilluminator. Semiquantitative analysis was done using a video-based densitometric system(Wincam®, Cybertech, Berlin, Germany). Gene expression for TNF-α and TNA-α-RII was examined in a semiquantitative way. PCR signals obtained from specific amplification of TNF-α and TNA-α-RII were standardized to the constitutive gene expression ofβ2-microglobulin (β2m) (relative signal intensities). Relative signal intensitiesTNF-α = (PCR-signalTNF-α)× (PCR-signalβ2m)-1 Relative signal intensitiesTNF-α-RII = (PCR-signalTNF-α-RII)× (PCR-signalβ2m)-1 Relative signal intensities in stimulated monocytes were compared with the findings related to unstimulated monocytes.

Nuclear extract preparation. To measure NF-κB activation, nuclear extracts prepared from freshly isolated human monocytes(106 cells/probe) incubated with media alone, LPS (2.5 µg/mL), and LPS + Curosurf® (500 µg/mL) for different time intervals were used for mobility shift assays as described previously(19). Briefly, incubation was terminated by cell lysis with 10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, and 0.25 mM phenylmethylsulfonyl fluoride (all purchased from Sigma Chemical Co.). Samples were kept on ice for 15 min, then Nonidet P-40 was added to a final concentration of 0.6%, samples were immediately centrifuged, and the pellets were incubated with 20 mM HEPES, pH 7.9, 0.4 M KCl, 1 mM EDTA, 1 mM EGTA, and 10% glycerol; nuclear proteins were eluted for 1 h at 4°C. The amount of extracted protein was determined photometrically (Bio-Rad, Munich, Germany) with BSA as a standard.

Labeling of oligonucleotides. The NF-κB specific double-stranded oligonucleotide (Santa Cruz, Heidelberg, Germany) was labeled with [γ-32P]ATP (Amersham, Braunschweig, Germany) using T4 polynucleotide kinase (Promega, Heidelberg, Germany).

Electrophoretic mobility shift assay. Nuclear extracts (5-10µg/probe) were incubated with 2 µg of poly(dI-dC) (Boehringer Mannheim, Mannheim, Germany), 20 µg of BSA, 4% Ficoll 400 in a final volume of 20 µL (5 mM HEPES, pH 7.9, 50 mM KCl, 5 mM MgCl2, 0.5 mM DTT, 1% glycerol, 3 µCi of labeled, double-stranded NF-κB oligonucleotide) for 25 min at room temperature.

As controls for the specificity of the reaction a 100-fold excess (100 ng) of unlabeled NF-κB oligonucleotide or an unrelated oligonucleotide, respectively, were used. Complexes were separated on a native 4% polyacrylamide gel in 0.5× Tris-borate-EDTA. The gel was vacuum-dried and developed by autoradiography.

Statistical methods. Statistical analysis was done by using the t test for unpaired samples. Values are given as mean ± SEM.

RESULTS

TNF-α secretion of monocytes exposed to surfactant. Our findings show that unstimulated monocytes secreted only small amounts of TNF-α during a 24-h assay. Monocytes stimulated with 2.5 µg/mL LPS secreted very high levels of TNF-α compared with unstimulated monocytes (mean 31 376 pg/mL, p < 0.001). Additional exposure of LPS-stimulated monocytes to surfactant (500 µg/mL) led to a significant reduction of TNF-α measured in culture supernatants (mean 3628 pg/mL; p < 0.001). This is synonymous with an 89% suppressive effect of surfactant (Fig. 1).

Figure 1
figure 1

Concentrations of TNF-α protein(pg/mL) in monocyte culture supernatants (n = 6) after 24 h of incubation. Sample description: (A) unstimulated monocytes;(B) monocytes cultivated with 2.5 µg/mL LPS; (C) monocytes cultivated with 2.5 µg/mL LPS and 500 µg/mL Curosurf®. Values are given as mean ± SEM.

Full size image

An assessment through exclusion of trypan blue revealed that the viability of unstimulated and LPS-stimulated monocytes as well as monocytes exposed to surfactant was >90% in all experiments.

Verification of primer-specific PCR products. Specificity of PCR products was verified through PAGE. As shown in Figure 2, specific gene products were amplified through PCR amplification. These products with different lengths(β2-microglobulin, 345 bp; TNF-α, 444 bp; and TNF-α-RII, 403 bp) were clearly identified according to their migrational behavior. No unspecific side products were observed.

Figure 2
figure 2

Representative electrophoresis gel used for quantification of PCR signals. Lane description: 1, molecular weight marker; 2-4, PCR products of unstimulated monocytes in triplicate; 5-7, PCR products of LPS-stimulated monocytes;8-10, PCR products of LPS-stimulated monocytes exposed to surfactant.

Full size image

Levels of TNF-α mRNA and TNF-α-RII mRNA in monocytes exposed to surfactant. Monocytes stimulated with LPS expressed 6.7 times(range, 3.8-9.3) more TNF-α mRNA than unstimulated monocytes(p < 0.001). The exposure to surfactant resulted in a reduction of TNF-α mRNA expression by 69% in LPS-stimulated monocytes [2.1 times more TNF-α mRNA (range, 1.7-2.5) than in unstimulated monocytes, respectively; p < 0.05] (Fig. 3). In addition, LPS-stimulated monocytes were found to contain 86% less TNF-α-RII mRNA when exposed to surfactant (p < 0.001). Expression of TNF-α-RII mRNA was 6.3 times higher (range, 5.8-6.7) in LPS-stimulated monocytes when compared with unstimulated controls. In contrast, LPS-stimulated monocytes exposed to surfactant produced only 0.9 times the amount of TNF-α-RII mRNA (range, 0.3-1.4) found in unstimulated monocytes (Fig. 3).

Figure 3
figure 3

Relative expression of TNF-α mRNA and TNF-α-RII mRNA after 24 h of incubation (n = 4) as determined by reverse transcription-PCR. Basal expression of TNF-α mRNA TNF-α-RII mRNA in unstimulated monocytes (control) was set as 1 on an arbitrary unit scale. Sample description: (A) unstimulated monocytes, (B) monocytes cultivated with 2.5µg/mL LPS, and (C) monocytes cultivated with 2.5 µg/mL LPS and 500 µg/mL Curosurf®. Values are given as mean ± SEM.

Full size image

NF-κB activation in monocytes exposed to surfactant. Human blood monocytes stimulated with LPS showed activation of NF-κB after only 30 min (data not shown), lasting for at least 2 h(Fig. 4). Incubation of monocytes with LPS and Curosurf® for the same time periods exhibited similar activation of NF-κB. As controls, an oligonucleotide with a random sequence and the unlabeled NF-κB oligonucleotide were added in excess to two control probes. The random oligonucleotide did not not affect the mobility shift, whereas the unlabeled specific NF-κB oligonucleotide abolished complex formation with the labeled probe, indicating the specificity of the reaction(Fig. 4).

Figure 4
figure 4

Nuclear extracts were prepared from monocytes and analyzed for NF-κB binding activity by electrophoretic mobility shift assay. Human monocytes were left untreated (lane 1), incubated with either LPS for 1 or 2 h (lanes 2 and 3) or with LPS and Curosurf® for 1 or 2 h (lanes 4 and 5), respectively. As controls for the specificity of the NF-κB binding complexes either unlabeled oligonucleotides with a random sequence (lane 6) or unlabeled NF-κB oligonucleotides(lane 7), respectively, were used in excess for competition. The result presented is representative of three independent experiments.

Full size image

DISCUSSION

TNF-α, a potent cytokine generated by macrophages, is a crucial factor involved in the occurrence of intra-alveolar and pulmonary inflammation in preterm infants with RDS and BPD. This potent cytokine, which is generated mainly by macrophages(5), stimulates different pulmonary cells, thereby generating chemotactic factors and a variety of inflammatory mediators. These substances induce various changes in pulmonary tissue, vascular endothelium(6–8), and the surfactant system(20). In addition, TNF-α has been shown to activate the generation of toxic oxygen radicals by phagocytic cells and to release potent proteases. This activation clearly enhances the destruction of pulmonary tissue(2,9). Moreover, it has recently been shown that TNF-α interferes with the synthesis of surfactant. Murch et al.(21) were able to demonstrate that the accumulation of TNF-α-positive macrophages in the lungs of infants with BPD was strongly associated with the destruction of sulfated glycosaminoglycans, which are essential to the integrity of the endothelium and other pulmonary structures.

The present study confirms our previous findings that surfactant suppresses the secretion of TNF-α in LPS-stimulated monocytes and can, therefore, prevent inflammation in the lung(11). Our present data provide new insights on the surfactant-mediated mechanisms. Our findings show that Curosurf® interferes with the action of TNF-α at an intracellular site by down-regulating TNF-α specific mRNA expression. The same effect was observed in the mRNA expression of the TNF-α-RII which transduces proinflammatory signals of LPS and TNF-α(17,18). Surfactant strongly inhibits both the secretion as well as the reception of TNF-α signals in activated human monocytes.

In the recent past, remarkable progress has been made in elucidating the intracellular signaling pathways of TNF-α and the immunomodulating effects of surfactant. There is evidence that phospholipid components in surfactant preparations regulate different immunosuppressive actions(11,12,22). One possible reason for the interference of phospholipids during intracellular signaling can be attributed to the second messenger pathway in which diacylglycerol is an intermediate product activating protein kinase C. Extracellular phospholipids of the microenvironment have been shown to affect the membrane phospholipid pattern of cells, which is crucial for the transmission of receptor-mediated signals through cell membranes(23). Phosphatidylcholine and sphingomyelin, which together represent more than 80% of the Curosurf® phospholipid composition(24), are able to diminish diacylglycrol-mediated protein kinase C activation(25), the latter being involved in the priming of macrophages and their activation by LPS(26,27). Phosphatidylcholine, for example, is also known to lower the proliferative lymphocyte response(12).

Antal et al.(28) recently reported that two other surfactant preparations [Survanta® (Ross) and Exosurf®(Burroughs-Wellcome)], which also induced diminished levels of TNF-α mRNA in an LPS-stimulated human monocytic cell line, inhibited activation of NF-κB, a ubiquitous transcription factor. In contrast to our findings, no suppression of TNF-α mRNA was found after more than 3 h of incubation. Because NF-κB can be activated by TNF-α and LPS, and as NF-κB plays a pivotal role in the transcription of genes involved in inflammatory responses (i.e. TNF-α, IL-1, IL-6)(29), we tested whether the given surfactant's suppressive effects are mediated through inhibition of NF-κB activation. However, incubation of freshly isolated human monocytes with LPS and Curosurf® resulted in a nearly identical activation of NF-κB when compared with monocytes exposed to LPS alone. Therefore, the clearly suppressive effect of Curosurf® on transcription and protein synthesis of TNF-α seems to be regulated by other, yet undefined signaling cascades.