Martina Steurer and Martin Schläpfer contributed equally to the work.
Beatrice Beck-Schimmer, Institute of Anesthesiology, Institute of Physiology and Zurich Center of Integrative Human Physiology, University of Zurich, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland. E-mail: firstname.lastname@example.org
Acute lung injury (ALI) is a well-defined inflammation whereby alveolar macrophages play a crucial role as effector cells. As shown previously in numerous experimental approaches, volatile anaesthetics might reduce the degree of injury in pre- or post-conditioning set-ups. Therefore, we were interested to evaluate the effect of the application of the volatile anaesthetic sevoflurane on alveolar macrophages regarding the expression of inflammatory mediators upon lipopolysaccharide (LPS) stimulation in vitro. Alveolar macrophages were stimulated with LPS. Two hours later, cells were exposed additionally to air (control) or to sevoflurane-containing air for 4, 6, 8, 12 or 24 h. Tumour necrosis factor (TNF)-α, cytokine-induced neutrophil chemoattractant-1 (CINC-1), macrophage-inflammatory protein-2 (MIP-2) and monocyte chemoattractant protein-1 (MCP-1) proteins were determined and chemotaxis assays were performed. To evaluate possible cellular signalling pathways phosphorylation of the kinases extracellular-regulated kinase (ERK) and Akt was assessed. In the early phase of sevoflurane post-conditioning expression of TNF-α, CINC-1, MIP-2 and MCP-1 was attenuated, leading to a diminished chemotaxis reaction for neutrophils. Phosphorylation of ERK seems to be a possible cellular mechanism in the sevoflurane-induced protection in vitro. Pharmacological post-conditioning of alveolar macrophages with sevoflurane immunmodulates the inflammatory response upon stimulation with endotoxin. This might be a possible option for a therapeutical approach in ALI.
Acute lung injury is a frequent clinical problem . Despite improvements in supportive care and advances in ventilator management, the mortality rate in patients with acute respiratory distress syndrome (ARDS) still remains at about 30% . Several pharmacological interventions have been investigated in the last few years, however, without showing a convincing treatment effect .
The approach of cell or organ pretreatment with any agent to reduce the following injury is called ‘preconditioning’. Exposing myocytes to volatile anaesthetics shortly before ischaemia–reperfusion results in a decreased inflammatory response to the injury, which is known as pharmacological, anaesthetic-induced preconditioning in correspondence with the ischaemic preconditioning . The same protective effect of volatile anaesthetics is seen in renal ischaemia–reperfusion injury resulting in an attenuation of the inflammation . However, previous knowledge of the onset of the injury is required to render preconditioning efficacious. Volatile anaesthetics have also been shown to be protective not only in ischaemia–reperfusion settings, but in endotoxin-induced injury models [6,7]. Our group has recently shown an attenuation of production of inflammatory proteins in alveolar epithelial cells with sevoflurane preconditioning in an in vitro model of endotoxin-induced lung injury . Additionally, a protective effect was also demonstrated in a post-conditioning set-up . This approach of post-conditioning is of high interest concerning a possible clinical use of sevoflurane in the treatment of acute lung injury.
Alveolar macrophages (AM) as effector cells play a crucial role in a variety of models of acute lung injury [10,11]. They are a rich source of cytokines, such as tumour necrosis factor (TNF)-α and chemokines [cytokine-induced neutrophil chemoattractant-1 (CINC-1); macrophage-inflammatory protein-2 (MIP-2); monocyte chemoattractant-1 (MCP-1)], which lead to neutrophil recruitment, thereby inducing further injury [12,13]. TNF-α plays a key role in this inflammatory orchestration and represents one of the most important inflammatory mediators . In this context, the interaction of AM and sevoflurane might be crucial. The aim of this study was to investigate the effect of sevoflurane on AM in a post-conditioning setting in an in vitro model of endotoxin-induced lung injury.
Material and methods
In order to reduce the use of animals for harvesting primary cell cultures of AM, a cell line was used. The rat AM cell line CRL-2192 was obtained from the American Type Culture Collection (Manassas, VA, USA). AM were cultured in nutrient mixture F-12 Ham (Ham's F-12; Invitrogen Corporation, Carlsbad, CA, USA), completed with 15% fetal bovine serum (FBS), 5% penicillin/streptomycin (10 000 U/l) (Invitrogen Corporation) and 5% HEPES (Invitrogen Corporation). Overnight, before starting the experiment, cells were incubated with Ham's F-12, completed with 1% FBS.
Primary AM (PAM) were isolated by bronchoalveolar lavage with phosphate-buffered saline (PBS) . Cells were centrifuged at 250 g for 5 min and plated in Dulbecco's modified Eagle's medium (DMEM; Invitrogen Corporation) with 1% FBS, 5% penicillin/streptomycin and 5% HEPES. As soon as the cells were adherent to the plates, experiments were started.
All experiments were performed with the cell line of AM except for the detection of cell signalling, where primary cultures of AM were used.
Mycoplasma detection assay
To exclude mycoplasm infection, mycoplasma detection assays were performed according to the manufacturer's protocol (Cambrex Bio Science, Rockland, ME, USA).
Stimulation with lipopolysaccharide
Macrophages were stimulated with lipopolysaccharide (LPS) from Escherichia coli, serotype 055:B5 (Sigma, Buchs, Switzerland) in a concentration of 20 µg/l [8,9]. Control cells were exposed to PBS.
Two hours after LPS stimulation, AM were placed in an airtight chamber (Oxoid anaerobic jar; Oxoid AG, Basel, Switzerland) and exposed to different vol% of sevoflurane (Sevorane(; Abott, Baar, Switzerland), provided by a sevoflurane vaporizer (Sevotec5®; Abbott) in air/5% CO2. The gas was humidified inside the chamber and the gas concentration was monitored (Ohmeda 5330 Agent Monitor; Abbott). The control group was exposed to air/5% CO2. The Oxoid chambers were sealed 5 min after reaching the desired sevoflurane concentration and incubated at 37°C for different time intervals up to 24 h. At the end of the experiment, supernatants were collected, aliquoted and stored at −20°C for further analysis.
Enzyme-linked immunosorbent assays
Enzyme-linked immunosorbent assays were performed according to the manufacturer's protocol assessing the cytokine TNF-α (BD Biosciences, San Diego, CA, USA), the chemokines CINC-1 (R&D Systems Europe Ltd, Abingdon, UK), MIP-2 (R&D Systems Europe Ltd) and MCP-1 (BD Biosciences). The detection range for TNF-α protein was 62·5–4000 pg/ml, for CINC-1 protein 20–1000 pg/ml, for MIP-2 15·6–2000 pg/ml and for MCP-1 31·3–2000 pg/ml.
Neutrophils [polymorphonuclear cells (PMN)] were isolated by gradient centrifugation over Ficoll-Paque (Amersham Pharmacia Biotech, Dubendorf, Switzerland) followed by 1% dextran sedimentation for 1 h to separate neutrophils from erythrocytes . After centrifugation of the supernatant, the contaminating erythrocytes were lysed with distilled water followed by the addition of 2·7% NaCl to stop hypotonic lysis. Neutrophils were incubated with 5 µM of the fluorescent indicator calcein acetoxymethyl ester (Calcein AM; Calbiochem®Biochemicals, Juro supply AG, Lucerne, Switzerland) for 30 min at 37°C. After washing neutrophils were resuspended at a total concentration of 6 × 106 PMN/ml in DMEM/1% FBS.
Cell supernatant, 600 µl, was filled into the lower wells of a transwell plate (Costar Transwell no. 3421, 5 µm; Corning Inc., Corning, NY, USA); 100 µl neutrophils at a concentration of 6 × 106/ml were added to the filter inlay. To determine basal migration, the lower well was loaded with PBS. After 1 h incubation in air/5% CO2, filters were removed and cells were lysed with Triton-X 0·1% (Sigma-Aldrich GmbH, Hamburg, Germany). Fluorescence was measured using an excitation filter at 485 nm and an emission filter at 535 nm.
Western blotting of extracellular signal-regulated kinase (ERK), phosphorylated ERK (pERK) and phosphorylated protein kinase B (pAkt)
To elucidate the intracellular signalling pathway of sevoflurane, Western blot analysis was performed, assessing pERK and pAkt, both known as pro-survival and anti-inflammatory kinases.
The PAM were harvested as described above. Cells were treated with or without LPS with an additional exposure to air/5% CO2 containing 3·3% of sevoflurane or to air/5% CO2. Cells were lysed with a buffer containing Hanks's balanced salt solution (Invitrogen Corporation), dithiothreitol 1 mM (DTT; Sigma-Aldrich GmbH), bestatin, aprotinin and leupeptin 200 nM (all from Sigma-Aldrich GmbH). Western blot analysis of 20 µg of cellular protein was performed on a 10% polyacrylamide gel. After separation, proteins were transblotted onto a nitrocellulose membrane. Incubation with a monoclonal mouse anti-pERK (Santa Cruz Biotechnology, Santa Cruz, CA, USA), a polyclonal rabbit ant-ERK (Cell Signaling Technology, Danvers, MA, USA) or a polyclonal rabbit anti-pAkt (Cell Signaling Technology) was performed overnight at 4°C. A secondary horseradish peroxidase-labelled anti-mouse or anti-rabbit IgG (1 : 5000) was added for 60 min at room temperature. Signals were detected by enhanced chemiluminescence and X-ray films were analysed with Scion Image (Scion Corporation, Frederick, MD, USA).
Loading control was performed by measuring β-actin levels (monoclonal mouse anti-β-actin antibody; Sigma-Aldrich GmbH) and densitometry was performed, calculating the ratio of pERK/β-actin. This ratio of LPS-stimulated cells was defined as 100%.
All experiments were repeated at least three times, each conducted in quartets, the controls in doublets. Results are expressed as means ± standard error of the mean.
Statistical significance of the differences between means was determined at a 5% level. Statistical calculations were performed using GraphPad Prism 4·0 software (San Diego, CA, USA).
Production of inflammatory proteins by AM
Because macrophage-derived TNF-α is a key chemoattractant upon endotoxin stimulation, in a first step we focused upon the post-conditioning-induced modification of this cytokine. Sevoflurane concentrations of 1·1, 2·2, 3·3 and 4·4 vol% were used in order to identify the most effective concentration regarding possible suppression of the production of TNF-α. After 4 h of post-conditioning, TNF-α protein concentration was assessed in the supernatant (Fig. 1). With 1·1% sevoflurane, a down-regulation of 0% (P = 0·972) was observed, with 2·2 vol% 6% (P = 0·011), 3·3 vol% 35% (P = 0·002) and 4·4 vol% 18% (P < 0·001).
To elucidate further the interaction of sevoflurane and AM at an optimal level, 3·3 vol% of sevoflurane was used for 4 h of post-conditioning.
The production and release of the inflammatory mediators TNF-α, CINC-1, MIP-2 and MCP-1 upon LPS stimulation and sevoflurane post-conditioning were assessed for different time-periods.
The TNF-α concentration in the supernatant decreased significantly after sevoflurane post-conditioning in comparison with the LPS control group after 2 h by 18% (P = 0·015), after 4 h by 35% (P < 0·001) and after 6 h by 30% (P < 0·001) (Fig. 2). A non-significant decrease was seen after 8, 12 and 24 h.
Expression of CINC-1 was modified by sevoflurane in the following way (Fig. 3): after 2 h of sevoflurane exposure production of CINC-1 decreased by 11% (P = 0·002), after 4 h by 30% (P < 0·001) and again after 24 h by 20 = 0·029).
Application of sevoflurane in combination of LPS exposure attenuated MIP-2 expression by 19% at 4 h only (P < 0·001), while no impact was seen at all other time-points (Fig. 4).
The MCP-1 production was attenuated in a similar manner (Fig. 5): the down-regulation was significant at 2 h with 30% (P < 0·001), at 4 h with 56% (P < 0·001), at 6 h with 34% (P = 0·001) and at 8 h with 34% (P < 0·001). No difference was seen at the 12-h and 24-h time-points.
Because neutrophil recruitment is an important step in the inflammatory pathway of acute lung injury, we assessed the biological function of the chemokines released by AM, performing chemotaxis assays (Fig. 6). Supernatants of 4 h of post-conditioning with 3·3 vol% sevoflurane were used with the appropriate controls. In the presence of sevoflurane, neutrophil migration was attenuated by 27% (P = 0·010) in comparison with the LPS control group. Chemotaxis was also decreased in the control group upon sevoflurane application (P = 0·015).
Intracellular signalling pathways
To determine a possible cellular mechanism of protection we focused upon the anti-inflammatory and anti-apoptotic kinases pERK and pAkt, as they are known to be involved in protective signalling pathways in stressed cells . pERK levels were increased significantly by 26% (P = 0·020) in the LPS-sevoflurane group compared with the LPS-air group (Fig. 7). No differences in the expression of pAkt were found between the two groups.
Our results show that sevoflurane administration after the onset of endotoxin injury decreases the inflammatory response of AM in vitro by attenuating cytokine and chemokine production. As a consequence of this, neutrophil recruitment is diminished as assessed by chemotaxis assay, demonstrating an important biological consequence of post-conditioning.
Interestingly, the sevoflurane concentration which induced the maximum alteration of cytokine production was 3·3 vol%, which is higher than the concentration used for alveolar epithelial cell pre- and post-conditioning, where sevoflurane concentrations of 1·1–2·2 vol% were used [8,9]. It seems likely that AM need a higher sevoflurane concentration to decrease their inflammatory response in comparison with epithelial cells. However, little is known about the ideal sevoflurane concentration for pre- or post-conditional approaches. Lee et al. found a plateau effect for sevoflurane ischaemic protection in kidney proximal tubule cells at 1·1 vol% . It is not known if the interaction of several cell types interferes with the concentration of the volatile anaesthetic needed for organ conditioning, which might lead to different results in vivo.
The TNF-α production by AM was attenuated strongly in the early phase of our experimental setting (after 2, 4 and 6 h of sevoflurane post-conditioning). As shown previously, early TNF-α release from pulmonary macrophages is an important step in the endotoxin-induced lung injury model , as well as in the model of ischaemia–reperfusion injury . TNF-α protein is secreted by macrophages during the first few hours after the onset of injury, orchestrating the early events which lead to acute lung injury . A down-regulation of the cytokine TNF-α by sevoflurane in the early phase of acute injury might be a crucial step in lung protection in vivo.
The same pattern of reduction at the early time-points is seen for the chemokines CINC-1, MIP-2 and MCP-1, which are also key chemoattractants during endotoxin-induced lung injury. Mounting evidence suggests that MCP-1 and its cell receptor (CCR2) are involved in numerous inflammatory disorders of the lung. This receptor acts by several mechanism including recruitment of regulatory and effector leucocytes such as macrophages and neutrophils . Our data show a significant reduction of MCP-1 production up to 50% after 4 h of sevoflurane administration. CINC, especially CINC-1 and MIP-2, play an important role in the recruitment of neutrophils to the lung in LPS-induced acute lung injury . A maximal attenuation of these two chemokines is also seen after 4 h of sevoflurane administration.
It is known that the endotoxin model of acute lung injury is biphasic. Early pulmonary oedema and alterations in lung function probably contribute to TNF-α, as it can cause both pulmonary dysfunction and pulmonary oedema . Therefore, we can expect a less severe course of the first phase of the disease in vivo using a sevoflurane post-conditioning model. The second phase, at about 6 h after injury onset, consists of an infiltration of the lung with inflammatory cells, predominantly neutrophils . This neutrophil recruitment is guided by chemokines such as CINC-1 and MIP-2. Applying our results concerning these chemokines, we can also anticipate alleviation of this phase of disease in vivo.
The results of this study might be of clinical importance. Because the development of an ARDS is not a planned event, it is of interest that the application of sevoflurane might be initiated after the onset of injury as post-conditioning. A previous study has shown a protective effect of isoflurane in endotoxin-induced injury when administered before the onset of injury . Our data demonstrate protective effects with sevoflurane administration 2 h after endotoxin challenge. In which period after injury onset the sevoflurane treatment should be started and how long the duration should be is the topic of further investigation.
A special focus of this study was the cellular signalling, which could be involved in pharmacological post-conditioning. Based on previous results, it was highly likely that the protective effects of sevoflurane might be mediated through phosphorylation of the anti-apoptotic and anti-inflammatory kinases ERK and Akt, both involved in the nuclear factor kappa B pathway . Lee et al. showed this effect in a preconditioning setting in kidney cells. In our experiments, using the cell line of AM, determination of the two kinases was not conclusive even under control conditions, as several non-specific bands were found with the antibodies used for Western blots. Therefore, some of the post-conditioning studies were repeated with primary culture of AM, resulting in the same findings regarding attenuation of the expression of inflammatory mediators. Increased pERK expression upon post-conditioning with sevoflurane was observed in PAM. Sevoflurane-induced cytoprotection might be extremely complex with a differential temporal regulation by ERK .
Although the data presented are promising we are aware of the fact that they represent an in vitro situation, which does not necessarily reflect in vivo mechanisms.
In conclusion, these data provide evidence of the protective effect of post-conditioning with the volatile anaesthetic sevoflurane on endotoxin-induced injury in AM. Pharmacological post-conditioning prevents enhanced expression of inflammatory mediators and attenuates increased chemotaxis. Phosphorylation of ERK might be a possible pathway in the protection. This strategy, although needing further investigation, may provide a new and easily applicable therapeutic option to protect the lung upon acute injury.
The authors thank Irene Odermatt, art designer, Institute of Anesthesiology, University of Zurich, Switzerland, for the development of illustrations. This study was supported by the Swiss National Science Foundation, grant no. 3200B0-109558, the Swiss Society of Anesthesiology and Resuscitation, Berne, Switzerland, Abbott AG, Baar, Switzerland and the Lungenliga, Switzerland.