Sepsis is one of the leading causes of mortality in intensive care units, and sedation in the intensive care unit during sepsis is usually performed intravenously. The inhalative anesthetic sevoflurane has been shown to elicit protective effects in various inflammatory studies, but its role in peritonitis-induced sepsis remains elusive. The hypothesis was that sevoflurane controls the neutrophil infiltration by stabilization of hypoxia-inducible factor 1α and elevated adenosine A2B receptor expression.
In mouse models of zymosan- and fecal-induced peritonitis, male mice were anesthetized with sevoflurane (2 volume percent, 30 min) after the onset of inflammation. Control animals received the solvent saline. The neutrophil counts and adhesion molecules on neutrophils in the peritoneal lavage of wild-type, adenosine A2B receptor −/−, and chimeric animals were determined by flow cytometry 4 h after stimulation. Cytokines and protein release were determined in the lavage. Further, the adenosine A2B receptor and its transcription factor hypoxia-inducible factor 1α were evaluated by real-time polymerase chain reaction and Western blot analysis 4 h after stimulation.
Sevoflurane reduced the neutrophil counts in the peritoneal lavage (mean ± SD, 25 ± 17 × 105vs. 12 ± 7 × 105 neutrophils; P = 0.004; n = 19/17) by lower expression of various adhesion molecules on neutrophils of wild-type animals but not of adenosine A2B receptor −/− animals. The cytokines concentration (means ± SD, tumor necrosis factor α [pg/ml], 523 ± 227 vs. 281 ± 101; P = 0.002; n = 9/9) and protein extravasation (mean ± SD [mg/ml], 1.4 ± 0.3 vs. 0.8 ± 0.4; P = 0.002; n = 12/11) were also lower after sevoflurane only in the wild-type mice. Chimeric mice showed the required expression of the adenosine A2B receptor on the hematopoietic and nonhematopoietic compartments for the protective effects of the anesthetic. Sevoflurane induced the expression of hypoxia-inducible factor 1α and adenosine A2B receptor in the intestine, liver, and lung.
Sevoflurane exerts various protective effects in two murine peritonitis-induced sepsis models. These protective effects were linked with a functional adenosine A2B receptor.
Sepsis is a major cause of mortality in critically ill intensive care unit patients because of host inflammatory injury.
Although inhaled sevoflurane is reported to have anti-inflammatory effects, its role in laboratory models is not well characterized.
In two murine peritonitis-induced sepsis models, sevoflurane reduced peritoneal lavage neutrophil counts by lower adhesion molecule expression, reduced cytokines, and protein extravasation. The mechanism of these reported protective effects was due to the expression of hypoxia-inducible factor 1α and adenosine A2B receptor in the intestine, liver, and lung.
Sepsis is one of the leading causes of mortality and critical illness worldwide. Despite decades of intensive research, the underlying mechanisms remain elusive.1 In the early stage of sepsis, neutrophils are the first cells of the innate immune system to be recruited to the site of infection.2 These cells are necessary for the clearance of the pathogen, but their excessive infiltration into the tissue leads to destruction and impairs organ function.3–5
Patients with severe sepsis are treated in the intensive care unit (ICU) and require sedation, which is usually performed with the intravenous anesthetic propofol. In a model of polymicrobial peritonitis, propofol enhanced morbidity and mortality in rats.6 Moreover, sevoflurane, a volatile anesthetic mainly used in the operation theater, can be applied to ICU units.7 Sevoflurane has been shown to elicit protective effects in terms of hepatic ischemia–reperfusion injury8 and acute pulmonary inflammation.9–11 In a septic model, sevoflurane reduced mortality,12 but the underlying mechanism has not been investigated thus far. A link between the protective effects of sevoflurane and adenosine A2B receptor is supposed.8 Therefore, adenosine A2B receptor is known to be crucial in terms of anti-inflammatory effects in acute inflammation, influencing neutrophil migration and microvascular permeability.13 Patients in the ICU have been shown to occasionally have reduced ligand affinity and distribution of adenosine receptors.14 The aim of our study was to investigate the detailed effects of sevoflurane in two peritonitis models and the link between the anesthetic and functional adenosine A2B receptor with the underlying mechanism. The primary goal of our study was to examine the effects of sevoflurane on the migratory behavior of neutrophils. The role of the hypoxia-inducible factor 1α/adenosine A2B receptor signaling pathway in this context served as a secondary goal. We hypothesized that sevoflurane had protective effects via hypoxia-inducible factor 1α/adenosine A2B receptor signaling. A link between sevoflurane and adenosine A2B receptor would allow analyses of the adenosine A2B receptor distribution in patients who would potentially benefit from the therapeutic use of sevoflurane in sepsis. Recent literature on inflammation has suggested the identification of subgroups of patients for a specific therapy.15
Materials and Methods
We used adenosine A2B receptor gene-deficient mice (adenosine A2B receptor −/−) and corresponding wild-type mice (C57BL/6) from Charles River Laboratories (Germany). The mice were male and between 8 and 12 weeks old. All animal protocols were approved by the Animal Care and Use Committee of the University of Tübingen (Tübingen, Germany) and the institutional animal protection board of the University of Würzburg (Würzburg, Germany). The small intestinal submucosal scaffold was generated from jejunal explants of piglets as described previously.16
Peritonitis Models and Application
All experiments were performed in deeply anesthetized mice (xylazine at 100 mg/kg body weight; ketamine at 10 mg/kg body weight), and all animals bled to death, as described previously.17 The mice inhaled sevoflurane in a custom-made chamber at 2 volume percent with an additional oxygen supply for 30 min breathed spontaneously but lost consciousness. To detect the optimal time point for the administration of anesthetic sevoflurane, we performed a time curve. Sevoflurane was applied simultaneously 1 and 3 h after zymosan (fig. 1A). Because the protective effects of the anesthetic were greatest 1 h after zymosan installation, we chose this time point for all subsequent experiments, including fecal-induced peritonitis experiments as described previously.17
In Vivo Migration Assay
Lungs and liver samples were homogenized. Cell counts of blood, peritoneal lavage, lungs, and liver samples were determined, and the samples were stained with a fluorescent antibody mix consisting of CD45 (Becton Dickinson, USA), 7/4-fluorescein isothiocyanate (AbD Serotec, USA), and Gr-1 (Lymphocyte Culture, USA) to detect neutrophils. A detailed description of the gating process is described in Supplemental Digital Content 1, fig. 1A (http://links.lww.com/ALN/C603). The samples were measured with a FACSCanto II flow cytometer (Becton Dickinson), and data analysis was performed by using FlowJo software (version 7.8.2; Ashland, USA). For the analyses of the expression of the adhesion molecules on neutrophils, additional staining was carried out with CD11a-phycoerythrin, CD11b-phycoerythrin, CD31-phycoerythrin, CD44-phycoerythrin, CD45-peridinin-chlorophyll-protein, CD162-phycoerythrin (all from Becton Dickinson), and CD54-phycoerythrin (Molecular Probes, USA).
The release of keratinocyte-derived chemokine, macrophage-inflammatory protein-2, tumor necrosis factor α, and interleukin-6 was measured in the peritoneal lavage of mice by enzyme-linked immunosorbent assays (R&D Systems, USA) according to the manufacturer’s protocol.
Total RNA was isolated from the murine intestine, lungs, and liver by using pegGOLD TriFast (Peqlab, USA). cDNA synthesis was performed by using a Bio-Rad iScript kit (Bio-Rad, USA) according to the manufacturer’s directions. Gene expression of murine adenosine A2B receptor and the transcription factor hypoxia-inducible factor 1α was evaluated by using the following primers: adenosine A2B receptor, 5´-gcattacagacccccaccaa-3´ and 5´-tttatacctgagcgggacgc-3´; and hypoxia-inducible factor 1α, 5´-atgttctgcccaccctgt-3´ and 5´-gagagcgcggaaaactct-3´. The 18s gene served as the housekeeping gene (5´-gtaacccgttgaaccccatt-3´ and 5´-ccatccaatcggtagtagcg-3´).
The organs were removed and prepared for Western blot analysis 4 h after zymosan administration. Protein concentration was determined according to the standard protocol of the protein assay kit (Pierce; Thermo Fisher, USA). After the protein concentration was equalized, the samples were loaded on sodium dodecyl sulfate–polyacrylamide gels. After the proteins were blotted on polyvinylidene difluoride membranes, anti-adenosine A2B-R (H-40; Santa Cruz, USA) and anti-hypoxia-inducible factor 1α (ab1; Abcam, United Kingdom) were utilized. The rabbit monoclonal anti–glyceraldehyde 3-phosphate dehydrogenase antibody was used as a housekeeping protein (Santa Cruz).
Protein extravasation into the peritoneal lavage as a marker for capillary leakage was determined 4 h after the onset of inflammation by using a bicinchoninic protein assay kit according to the standard protocol (Pierce; Thermo Fisher).
Generation of Chimeric Mice
Chimeric mice that expressed adenosine A2B receptor on either hematopoietic (adenosine A2B receptor blood) or nonhematopoietic cells (adenosine A2B receptor tissue) were generated by transferring bone marrow between wild-type and adenosine A2B receptor −/− mice as previously described.18
Immunohistochemical Neutrophil Detection
With a Vectastain ABC kit, neutrophil accumulation in the intestine, lung, and liver tissue was visualized by immunohistochemistry (Vector Laboratories, Germany) as previously described.17 Tissue slides were processed with a Leitz DM IRB microscope (Leica, Germany) and analyzed with AxioVision, version 4.8.2 (Carl Zeiss, Germany).
In Vitro Neutrophil Migration
We performed an in vitro transmigration assay of human neutrophils through a monolayer of CaCo-2 cells (ATCC HTB-37, USA). Human epithelial cells were cultivated on inserts of a Transwell system (Costar, USA) until reaching confluence. Epithelial cells were seeded on the bottom side of the inserts so that cells migrated—comparable to the in vivo migration—from the basolateral side through the epithelial monolayer. Neutrophils from whole human blood were isolated. The cells were treated with sevoflurane vaporization of 2 volume percent continuously in a custom-made chamber for 30 min. Neutrophil migration through a monolayer of epithelial cells was initiated by the chemoattractant formyl-methionyl-leucyl-phenylalanine (20 ng/ml; Sigma-Aldrich, USA). After 1 h, migrated neutrophils in the lower chamber were quantified by the determination of myeloperoxidase (absorption length, 405 nm).
To verify our in vivo findings, we evaluated the effects of sevoflurane after adenosine A2B receptor inhibition with a specific antagonist, PSB1115 (Sigma–Aldrich). Human neutrophils and CaCo-2 cells were pretreated with the adenosine A2B receptor antagonist, and neutrophils were applied to the inserts. Neutrophil transmigration through the basolateral side of the intestinal epithelial monolayer was initiated by formyl-methionyl-leucyl-phenylalanine. To test our hypothesis that the protective effects of sevoflurane treatment depend on adenosine A2B receptor/hypoxia-inducible factor 1α signaling, we additionally depleted adenosine A2B receptor by small interfering RNA (20 nm; OriGene, USA) and evaluated the effects of overexpression of adenosine A2B receptor by a human cDNA plasmid (4 µg; OriGene). The effects of hypoxia-inducible factor 1α depletion on neutrophil migration were determined by human small interfering RNA hypoxia-inducible factor 1α (Santa Cruz). Nontargeting small interfering RNA (OriGene) and small interfering RNA control (Santa Cruz) were used as controls.
Furthermore, we used an advanced model of the human intestinal barrier to study the effects of sevoflurane on neutrophil transmigration. This model is based on a porcine decellularized small intestinal submucosal scaffold. In this biologic scaffold, the architecture of the intestinal extracellular matrix is preserved. Human intestinal epithelial cells (CaCo-2) were cultured on the bottom side of the scaffold as described previously.19 Confluence was controlled by the permeability of fluorescein-dextran. Compared to Transwell-based models, in which neutrophils migrate through synthetic pores, neutrophils have to migrate through the extracellular matrix mostly composed of collagen and elastin, which mimics the situation in vivo. The inflammation-induced effects of formyl-methionyl-leucyl-phenylalanine on the accumulation of intracellular cyclic adenosine monophosphate (cAMP) in human epithelial intestinal cells were investigated by enzyme-linked immunosorbent assays following the manufacturer’s instructions (Enzo Biosciences, Germany).
Continuous variables are reported as the mean ± SD or median and interquartile range according to their distribution. The distribution was evaluated by investigating kurtosis, skewness, Q-Q plots, and histograms. The required number of animals needed for the experiments was based on our previous experience with this design. No a priori statistical power calculation was conducted. For all the animal experiments, the mice were randomly assigned to the groups and used in alternating sequential order, and experiments were performed at the same time point. Additionally, blinding of the experimenters took place in the animal experiments to avoid subjective bias. There were no missing data, and all animals were included in further analyses. However, in some in vitro experiments, sample sizes were adapted according to the observed effect strengths and based on previous experiences. No P value adjustments were made for these interim analyses of the data. Independent-sample t tests were used to compare numerical variables that were approximately normally distributed, whereas Mann–Whitney tests were used to evaluate skewed variables. For comparisons between more than two groups, one-way ANOVA was performed and adjusted by Bonferroni correction for multiple comparisons. Residuals were inspected; in the case of severe deviation from a normal distribution even after log transformation, nonparametric methods were used (Kruskal–Wallis test). All statistical tests were two-tailed, and the significance level was set at P ≤ 0.05 (corrected for multiple testing). Statistical analysis was performed using GraphPad Prism version 8.4.2 for Windows (GraphPad Software, USA).
Time-dependent Anti-inflammatory Effects of Sevoflurane
We determined neutrophil accumulation in the circulation, peritoneal lavage, and liver and lung tissue 4 h after zymosan-induced peritonitis by flow cytometry (fig. 1A; Supplemental Digital Content 2, table 1A, http://links.lww.com/ALN/C604). The anesthetic sevoflurane was applied simultaneously with the induction of peritonitis and 1 and 3 h after zymosan administration. In the circulation, neutrophilia was similar to baseline neutrophil counts without inflammation after the application of sevoflurane simultaneously with and 1 h after the induction of inflammation. In the peritoneal lavage, neutrophil influx was lower 1 h after zymosan administration. At this time point, a reduction in neutrophil migration into the liver and lung was also observed. To verify our findings, we determined myelo-peroxidase, which is an enzyme released upon neutrophil activation and is associated with leukocyte activity, in the peritoneal lavage (fig. 1B; Supplemental Digital Content 2, table 1B, http://links.lww.com/ALN/C604). In accordance with our flow-cytometric findings, myeloperoxidase was reduced 1 and 3 h after zymosan. The application of sevoflurane diminished the microvascular permeability when the anesthetic was applied 1 h after inflammation (fig. 1C; Supplemental Digital Content 2, table 1C, http://links.lww.com/ALN/C604). For all subsequent experiments, we chose sevoflurane application 1 h after the induction of zymosan.
The Protective Effects of Sevoflurane Depend on Functional Adenosine A2B Receptor Signaling
Zymosan-induced peritonitis was induced in wild-type and adenosine A2B receptor −/− animals, and the effects of sevoflurane on neutrophil migration were evaluated. Confirming our previous results and demonstrating reproducibility, the application of the anesthetic reduced neutrophil accumulation in the circulation and neutrophil migration into the peritoneal lavage and lung and liver tissues in the wild-type animals. This effect was not observed in the adenosine A2B receptor −/− animals (fig. 2A; Supplemental Digital Content 3, table 2A, http://links.lww.com/ALN/C605). To visualize the effects of sevoflurane on neutrophil migration, we performed immunohistochemical detection of neutrophils in all three organs (fig. 2B). Neutrophils were marked with a specific antibody so that they appeared brown in the tissue. Sevoflurane reduced the detected neutrophils in the lung, liver, and intestine in the wild-type animals but had no effect on neutrophil influx into the organs of the adenosine A2B receptor −/− animals. The concentration of myeloperoxidase affirmed this finding with a reduction after sevoflurane in the wild-type mice and no effect in the adenosine A2B receptor −/− mice (fig. 2C; Supplemental Digital Content 3, table 2B, http://links.lww.com/ALN/C605). Capillary leakage was also lower in the wild-type mice after the inhalation of sevoflurane (fig. 2D; Supplemental Digital Content 3, table 2C, http://links.lww.com/ALN/C605). The anesthetic failed to influence microvascular permeability in the adenosine A2B receptor −/− animals.
Because the release of cytokines triggers neutrophil migration, we determined the concentration of the two main neutrophil chemoattractants in the mice in peritoneal lavage: keratinocyte-derived chemokine and macrophage-inflammatory protein-2, which resemble human interleukin-8. Both chemoattractants were lower in the wild-type animals after sevoflurane administration (fig. 2E; Supplemental Digital Content 3, table 2D, http://links.lww.com/ALN/C605). In addition, the cytokines tumor necrosis factor α and interleukin-6 were diminished by the administration of sevoflurane in the wild-type animals. Confirming our previous finding, sevoflurane had no effects in the adenosine A2B receptor −/− animals in this setting.
To increase the clinical impact of our study, we evaluated the effects of sevoflurane in a fecal-induced model of peritonitis. The application of the anesthetic 1 h after inflammation reduced neutrophil migration into the peritoneal lavage and both organs (fig. 2F; Supplemental Digital Content 3, table 2E, http://links.lww.com/ALN/C605). This finding was confirmed by the determination of myeloperoxidase (fig. 2G; Supplemental Digital Content 3, table 2F, http://links.lww.com/ALN/C605), where sevoflurane also reduced the release. Microvascular permeability was also lower after the animals inhaled sevoflurane, highlighting the protective properties of sevoflurane in sepsis (fig. 2H; Supplemental Digital Content 3, table 2G, http://links.lww.com/ALN/C605).
The Influence of Sevoflurane on Adhesion Molecules on Neutrophils
We determined the expression of specific adhesion molecules on neutrophils. The investigated adhesion molecules all play distinct roles in the different steps of neutrophil migration, and the expression patterns change during migration.20
We investigated the expression of CD44, CD31, CD162, CD54, CD11b, and CD11a. Sevoflurane diminished the expression of CD44 and CD31 on intravascular neutrophils in the wild-type animals (fig. 3A; Supplemental Digital Content 4, table 3A, http://links.lww.com/ALN/C606). In these animals, neutrophils in the peritoneal lavage showed lower expression of CD162, CD54, CD44, and CD31 on their surfaces after inhalation of sevoflurane (fig. 3B; Supplemental Digital Content 4, table 3B, http://links.lww.com/ALN/C606). In the livers of the wild-type animals, CD162, CD54, CD44, CD31, CD11a, and CD11b were reduced on neutrophils (fig. 3C; Supplemental Digital Content 4, table 3C, http://links.lww.com/ALN/C606). In the lung, an effect of sevoflurane on the expression of CD11b and CD11a was detectable (fig. 3D; Supplemental Digital Content 4, table 3D, http://links.lww.com/ALN/C606). According to our finding that the effect of sevoflurane depends on functional adenosine A2B receptor, the expression of adhesion molecules on neutrophils was not affected in the adenosine A2B receptor −/− animals.
Sevoflurane Increases the Expression of Adenosine A2B Receptor and Hypoxia-inducible Factor 1α during Acute Peritonitis
Next, we investigated the impact of sevoflurane on the expression of adenosine A2B receptor. The gene expression of adenosine A2B receptor in the intestine, liver, and lung was greater after the application of sevoflurane (fig. 4A; Supplemental Digital Content 5, table 4A, http://links.lww.com/ALN/C607). Western blots confirmed these results at the protein concentration (fig. 4B; Supplemental Digital Content 6, fig. 2A, http://links.lww.com/ALN/C608; Supplemental Digital Content 5, table 4B, http://links.lww.com/ALN/C607). Under inflammatory conditions, hypoxia-inducible factor 1α has been shown to be a transcription factor of adenosine A2B receptor. We investigated the impact of sevoflurane on this transcription factor in our setting (fig. 4C; Supplemental Digital Content 5, table 4C, http://links.lww.com/ALN/C607). The therapeutic use of sevoflurane enhanced the gene and protein expression of hypoxia-inducible factor 1α in all three organs (fig. 4D; Supplemental Digital Content 6, fig. 2B, http://links.lww.com/ALN/C608; Supplemental Digital Content 5, table 4D, http://links.lww.com/ALN/C607), explaining our findings on the impact of sevoflurane on adenosine A2B receptor.
The Role of Adenosine A2B Receptor in Hematopoietic and Nonhematopoietic Tissue
To investigate whether adenosine A2B receptor on hematopoietic or nonhematopoietic cells is crucial for the protective effects of sevoflurane, we determined the effects on neutrophil migration in chimeric mice (fig. 5A). Adenosine A2B receptor expression in the different tissues of the chimeric mice was confirmed by real-time polymerase chain reaction (Supplemental Digital Content 1, fig. 1B, http://links.lww.com/ALN/C603). Neutrophil influx was determined in the lavage, liver, and lung samples of the chimeric mice (fig. 5A), in which sevoflurane did not show any protective effects. The anesthetic had no effect on the release of myeloperoxidase in either chimeric mouse (fig. 5B). In addition, a functional adenosine A2B receptor was necessary in both compartments to prevent capillary leakage by sevoflurane (fig. 5C).
The Protective Effects of Sevoflurane In Vitro Depend on a Functional Adenosine A2B Receptor
To further verify our in vivo findings, we determined the effect of sevoflurane on the migration of human neutrophils through a monolayer of human intestinal epithelium in a Transwell system in vitro. Migration was reduced by the anesthetic, and the blockade of adenosine A2B receptor by a specific antagonist reversed this finding (fig. 6A; Supplemental Digital Content 5, table 4E, http://links.lww.com/ALN/C607). The elevation of adenosine A2B receptor expression by a plasmid did not show a synergistic effect when it was combined with sevoflurane in terms of neutrophil migration through an epithelial monolayer (fig. 6B; Supplemental Digital Content 5, table 4F, http://links.lww.com/ALN/C607). Adenosine A2B receptor depletion by small interfering RNA abolished the protective effects of the anesthetic on neutrophil migration (fig. 6C; Supplemental Digital Content 5, table 4G, http://links.lww.com/ALN/C607), further confirming our in vivo results. The impact of sevoflurane on the migration of human neutrophils through intestinal tissue was further investigated by the use of a three-dimensional human intestinal tissue model based on porcine decellularized small intestinal submucosal scaffolds.19 This migration was again lowered by sevoflurane (fig. 6D; Supplemental Digital Content 5, table 4H, http://links.lww.com/ALN/C607).
We determined the impact of sevoflurane on the release of the intracellular signaling protein cAMP, a component of the downstream signaling pathway of adenosine A2B receptor. The anesthetic elevated the release of cAMP, further confirming our in vivo findings on the impact of sevoflurane on the expression of adenosine A2B receptor (fig. 6E; Supplemental Digital Content 5, table 4I, http://links.lww.com/ALN/C607).
After depletion of the transcription factor hypoxia-inducible factor 1α by small interfering RNA, a Transwell migration assay of human neutrophils through a monolayer of human intestinal cells was performed. Hypoxia-inducible factor 1α depletion abolished the anti-inflammatory effects of sevoflurane on neutrophil migration (fig. 6, F and G; Supplemental Digital Content 5, tables 4J and 4K, http://links.lww.com/ALN/C607). A schematic overview of the effects of the anesthetic in our model of acute inflammation is shown (fig. 6H).
We investigated the detailed effect of anesthetic sevoflurane in the acute phase of zymosan- and polymicrobial-induced peritonitis. We detected reduced neutrophil migration into the peritoneal lavage, liver, and lung with altered expression of adhesion molecules on neutrophils and anti-inflammatory effects of the therapeutic use of sevoflurane on microvascular permeability.
In a model of cecal ligation and puncture, the inhalation of sevoflurane and the intravenous application of the primary sevoflurane metabolite improved survival and reduced tissue damage. Cellular infiltration, histology, or capillary leakage was not assessed in this study.12 Lucchinetti et al.21 investigated the effects of sevoflurane in a model of ischemia–reperfusion injury. The authors detected a downregulation of CD11b, which is considered an activation marker of leukocytes, on neutrophils. Similarly, in the current study, the expression of CD11b on neutrophils was reduced in both organs—liver and lung. We detected a significant reduction in adhesion molecules on neutrophils in all investigated compartments and organs (circulation, peritoneal lavage, liver, and lung tissue) in the wild-type animals after sevoflurane treatment, indicating the pivotal effect of the anesthetic on neutrophils. Because peritonitis and subsequent sepsis are major problems in ICUs, we tried to increase the clinical impact of our study by administering sevoflurane after the onset of inflammation so that the anesthetic was used as a therapeutic.
Extracellular adenosine signaling has anti-inflammatory functions in the acute phase of diverse diseases, such as acute lung injury and intestinal inflammation.22 During inflammation, extracellular adenosine predominantly stems from the breakdown of precursor nucleotides such as adenosine triphosphate.23 Experimental data showed that a reduced extracellular adenosine concentration exacerbated the inflammatory response during acute pulmonary inflammation.24 Patients in ICUs may have altered adenosine receptor expression or ligand affinity.14,25 Additional studies on these adenosine receptors in diverse clinical settings are needed to further define their exact role, expression, and function in different states of inflammation and critical illnesses. The findings of this study may provide a molecular link between the protective effects of sevoflurane and functional adenosine A2B receptor during peritonitis-induced sepsis. Similarly, Kiers et al.26 observed an increased extracellular adenosine concentration, reduced endotoxin-related inflammation, and augmented adenosine A2B receptor expression in a murine model of short-term hypoxia. The authors added experiments with healthy volunteers in an endotoxemia analysis, in which short-term hypoxia also elevated adenosine expression and reduced inflammatory cytokine concentrations. Therefore, inflammation is linked with hypoxemic conditions.27 In humans, sevoflurane has been shown to reduce cytokine expression during lung resection and cardiopulmonary bypass surgeries.28,29 For a few years, sevoflurane has been available for application in ICUs by using the AnaConDa device. Jerath et al.30 indicated in their meta-analysis that sevoflurane may promote faster weaning and extubation of ICU patients, showing beneficial aspects for the use of this anesthetic in the ICU.
Adenosine A2B receptor is well known to be protective in terms of acute lung inflammation and liver ischemia.8,18,31 In cecal ligation and puncture-induced peritonitis, the loss of adenosine A2B receptor enhanced chemokine release and dampened survival.13 Inflammation increases the expression of adenosine A2B receptor, most likely as part of natural defense mechanisms, but an increase in adenosine A2B receptor stimulation by a specific agonist dampens acute inflammation even further.18,32,33 Furthermore, it seems that the activation of adenosine A2B receptor on regulatory T cells can induce protective effects during acute inflammation.34 Specific inhibition of equilibrative nucleoside receptor 2, a cellular adenosine transporter, reduced the uptake of adenosine and raised adenosine expression, which induced protective effects during acute pulmonary inflammation through adenosine A2B receptor.35
Various experimental studies in mice have shown that BAY60-6583, the most examined activator of adenosine A2B receptor, has protective effects in different models of inflammation.36,37 Unfortunately, to the best of our knowledge, there are no clinical data available.38 Adenosine itself has a short half-life and undesirable side effects, such as atrial fibrillation.39 In contrast, various hypoxia-inducible factor–prolyl hydroxylase inhibitors, which prevent the degradation of hypoxia-inducible factor and induce a transient increase in hypoxia-inducible factor–related genes, exist. These hypoxia-inducible factor–prolyl hydroxylase inhibitors are currently under investigation in clinical trials. Elevated hypoxia-inducible factor–related genes can increase renal erythropoietin. In chronic kidney disease–related anemia, hypoxia-inducible factor inhibitors were used and successfully passed phase 3 of clinical development.40,41
The inflammatory process is linked with hypoxia and hypoxic mechanisms.27 In inflammatory colitis, hypoxia-inducible factor 1α was elevated,42 and the mucosa became hypoxic.43 Hypoxia-inducible factor 2α, an immunomodulatory isoform of the transcription factor, demonstrated cardioprotective effects in a cardiac ischemia–reperfusion model.44 To the best of our knowledge, there is no connection between hypoxia-inducible factor 2α and adenosine A2B receptor signaling. Therefore, we focused our experiments on hypoxia-inducible factor 1α/adenosine A2B receptor signaling.
Our findings detected a link between sevoflurane and hypoxia-inducible factor 1α/adenosine A2B receptor signaling, but the possible direct mechanism remains elusive. In a recently published study, sevoflurane reduced the immune response by decreasing cytokine expression via elevated activation of signal transducer and activator of transcription 3 signaling in a hepatic ischemia–reperfusion model in rats.45 Signal transducer and activator of transcription 3 is involved in inflammation-induced hypoxia-inducible factor 1α gene expression.46 Based on these studies and our findings, it is tempting to speculate that sevoflurane induced elevated gene expression and stabilization of hypoxia-inducible factor 1α, providing the observed beneficial effects during peritonitis-related sepsis via activation of signal transducer and activator of transcription 3 signaling. Further studies are needed to prove exactly how the anesthetic sevoflurane contributes to the stabilization of hypoxia-inducible factor 1α.
In this study, we utilized only male mice, and this study did not highlight possible sex-specific differences. As limitations of our study, the potential effects of the other adenosine receptors, adenosine A1 receptor, adenosine A2A receptor, and adenosine A3 receptor, were not considered. Furthermore, it would be interesting to investigate the effects of sevoflurane on the activation and stabilization of hypoxia-inducible factor 2α in future projects.
The presented results demonstrate a protective role of the anesthetic sevoflurane in two murine peritonitis-induced sepsis models. Sevoflurane reduced the release of inflammatory cytokines, neutrophil migration, and protein release in two different peritonitis models. Functional adenosine A2B receptor and hypoxia-inducible factor 1α appear to be involved in these effects. Although the exact mechanism of sevoflurane in mitigating inflammation related to sepsis is still not understood, these findings further support the idea that clinical sevoflurane administration should be considered for sepsis and systemic inflammation.
Adenosine A2B receptor gene–deficient mice were kindly provided by Katya Ravid, Ph.D. (Department of Biochemistry, School of Medicine, Boston University, Boston, Massachusetts). The authors thank Lina Maria Serna Higuita, M.D. (Institute of Clinical Epidemiology and Applied Biometry, University of Tübingen, Tübingen, Germany), for statistical counseling and work-up of the results.
Supported by the Else Kröner-Fresenius-Stiftung 2014_A148 (to Dr. Konrad) and German Research Foundation grant No. KO 4280/2-1 (to Dr. Konrad).
Dr. Straub has indicated the following financial relationships: speaker honoraria and travel support from CSL Behring (Hattersheim, Germany); speaker honoraria from Aspen Germany (Munich, Germany); and speaker honoraria from Schöchl Medical Education (Mattsee, Austria). The other authors declare no competing interests.