Isoflurane Activates Intestinal Sphingosine Kinase to Protect against Renal Ischemia–Reperfusion-induced Liver and Intestine Injury

• Acute kidney injury results in a systemic inflammatory condition that injures other organs, including the intestine and liver

• Potent volatile anesthetics have antiinflammatory effects and protect against renal ischemia–reperfusion injury

• The volatile anesthetic isoflurane protects the intestine and liver after renal ischemia–reperfusion injury by attenuating proinflammatory cytokine up-regulation and intestinal apoptosis through induction of the sphingosine kinase-1/sphingosine-1-phosphate pathway.

ACUTE kidney injury (AKI) continues to be a major clinical problem during the perioperative period.1The development of AKI often leads to multiorgan dysfunction and systemic inflammation and contributes to significant morbidity and mortality. In particular, hepatic dysfunction occurs frequently in patients with AKI and leads to other complications, such as intestinal barrier dysfunction, respiratory failure, and multiorgan failure (frequently complicated by sepsis).2We recently demonstrated that AKI leads to rapid hepatic and intestinal injury in mice with increased inflammation, apoptosis, and necrosis.3This hepatic and intestinal injury was mediated by up-regulation of the proinflammatory cytokines tumor necrosis factor (TNF)-α, interleukin (IL)-17A, and IL-6 after AKI.

Volatile anesthetics are an integral part of the perioperative period because they are administered to virtually all patients undergoing general anesthesia in the United States. It was previously reported that volatile anesthetics protected against renal ischemia–reperfusion injury (IRI) in vivo  4and had direct antiinflammatory and antinecrotic effects in cultured human kidney proximal tubule cells.5Most volatile anesthetics are lipophilic molecules6and have increased membrane fluidity and activated sphingomyelin hydrolysis in the renal cortex.7The lysophospholipid sphingosine-1-phosphate (S1P) is a product of sphingomyelin hydrolysis and functions as both an extracellular ligand for specific G protein–coupled receptors and an intracellular second messenger in promoting cell growth and survival and the inhibition of apoptosis.8 

In this study, we questioned whether volatile anesthetics would provide protection of the liver and intestine after renal IRI. We hypothesized that volatile anesthetics activate the sphingosine kinase-1 (SK1)/S1P pathway in the small intestine to protect against renal IRI-induced hepatic and intestinal injury. We demonstrate rapid hepatic and intestinal injury after renal IRI, with protective effects mediated by the representative volatile anesthetic, isoflurane. Isoflurane reduced hepatic and intestinal proinflammatory cytokine up-regulation and intestinal apoptosis via  induction of SK1 in small-intestinal crypts.

Isoflurane (2-chloro-2-[difluoromethoxy]-1,1,1-trifluoro-ethane) was purchased from Abbott Laboratories, North Chicago, IL; and the selective SK inhibitor (SKI-II; 4-[4-{4-chlorophenyl}-2-thiazolyl]amino phenol) was purchased from Tocris Bioscience, Ellisville, MO. Unless otherwise specified, all other reagents were purchased from Sigma, St Louis, MO.

All animal protocols were approved by the Institutional Animal Care and Use Committee of Columbia University, New York, NY. We used male C57BL/6 mice (Harlan, Indianapolis, IN) or SK1 knockout mice (provided by R. L. Proia, Ph.D., Chief, Genetics of Development and Disease Branch, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD); all mice weighed 20–25 g. The generation and initial characterization of SK1 knockout mice was previously described.9These mice were backcrossed to C57BL/6 mice for more than 10 generations.

In our model of renal IRI, mice were initially anesthetized with intraperitoneal pentobarbital, 50 mg/kg body weight or to effect (Henry Schein Veterinary Co, Indianapolis, IN); and subjected to right nephrectomy and 30 min of left renal ischemia. After closure of the abdomen in two layers, the mice were then exposed to an additional 4 h of equipotent doses of either pentobarbital via  intermittent intraperitoneal administration or isoflurane (1.2% or minimum alveolar concentration, defined as the concentration of volatile anesthetic in the lungs that is needed to prevent movement in 50% of subjects in response to a painful stimulus), as previously described.4Briefly, mice were placed in an airtight 10-l chamber on a warming blanket with inflow and outflow hoses located at the top and bottom of the chamber, respectively. Isoflurane was delivered in room air at 5 l/min using an isoflurane-specific vaporizer (Datex-Ohmeda, Madison, WI). The vaporizer was set to maintain chamber isoflurane concentration at 1.2%, monitored by an infrared analyzer sampling gas at the outflow hose. The mice were placed on a heating pad under a warming light to maintain body temperature at approximately 37°C.

To test the effects of SK inhibition, SKI-II was administered to mice undergoing renal IRI. SKI-II (50 mg/kg) was administered subcutaneously 15 min before ischemia and 4 h after reperfusion. SKI-II is an SK-selective inhibitor with minimal effects on other kinases,10and this dose had effective inhibition of activity without significant toxicity.11 

For experiments involving a sham operation or renal IRI, all samples (including plasma and tissue) were collected from mice 24 h after a sham operation or renal IRI. For experiments involving sham exposure to pentobarbital or isoflurane (SK messenger RNA [mRNA] and protein expression, SK1 immunofluorescence, SK activity, and S1P measurement), mice were exposed to 4 h of pentobarbital or isoflurane (without a sham operation or renal IRI) and allowed to recover for 1 h. Samples were collected 5 h after initial anesthetic exposure.

Plasma alanine aminotransferase (ALT) activity and creatinine concentrations were measured using an ALT assay kit (Infinity) and an enzymatic creatinine reagent kit, respectively, according to the manufacturer's instructions (Thermo Fisher Scientific, Waltham, MA).

For histologic preparations, liver or small intestine (jejunum and ileum) tissues collected from mice were washed in ice-cold phosphate-buffered saline and fixed overnight in 10% formalin. After automated dehydration through a graded alcohol series, tissues were embedded in paraffin, sectioned at 5 μm, and stained with hematoxylin-eosin. Liver hematoxylin-eosin sections were graded for renal IRI-induced vacuolization injury (score range, 0–4) by a pathologist (V.D.D.'A.) blinded to the samples.

We detected small-intestinal apoptosis with terminal deoxynucleotidyl transferase biotin-dUTP nick end-labeling (TUNEL) staining. In situ  labeling of fragmented DNA was performed with TUNEL staining (green fluorescence) using a commercially available in situ  cell death detection kit (Roche, Indianapolis, IN), according to the instructions provided by the manufacturer.

Hepatic and intestinal inflammation was determined by measuring mRNA-encoding markers of inflammation, including IL-6, IL-17A, intercellular adhesion molecule (ICAM)-1, monocyte chemoattractant protein (MCP)-1, macrophage inflammatory protein (MIP)-2, and TNF-α (table 1). In addition, SK1 and SK2 mRNA levels were measured. Semiquantitative real-time reverse transcription–polymerase chain reaction was performed as previously described.12 

Changes in liver and small intestine vascular permeability were assessed by quantitating extravasation of Evans blue dye into the tissue, as previously described.13 

Small-intestinal tissues in mice were collected and homogenized in lysis buffer (20 mM HEPES, pH 7.4; 2 mM EGTA; 1 mM dithiothreitol; 1% Triton X-100; 10% glycerol; and protease inhibitor cocktail) (Calbiochem, La Jolla, CA) on ice with a glass homogenizer. The homogenates were centrifuged for 20 min at 16,000g . The supernatant was collected and used for immunoblotting as previously described.14The samples (50–100 μg protein/lane) were separated on 9 or 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and then transferred to membranes (Immobilon; Millipore, Bedford, MA). The membranes were blocked with blocking buffer (5% nonfat dry milk in tris buffered saline (TBS) containing 0.1% Tween 20) and incubated overnight with polyclonal anti-SK1 (3297; 1:1,000 dilution; Cell Signaling, Beverly, MA), anti-SK2 (ab37977; 1:2,000 dilution; Abcam, Cambridge, MA), or monoclonal anti–β-actin (A5316; 1:5,000 dilution; Sigma) antibodies diluted in blocking buffer at 4°C. After being washed, the membranes were incubated with horseradish peroxidase–conjugated donkey anti–rabbit or sheep antimouse (1:5,000 dilution; ECM Bioscience, Versailles, KY) antibodies for 1 h at room temperature. Finally, the membranes were detected with enhanced chemiluminescence immunoblotting detection reagents (Amersham, Piscataway, NJ) with subsequent exposure to a charge-coupled device camera coupled to an imaging device (UVP Bio-imaging System; UVP, Upland, CA). The band intensities of the immunoblots were within the linear range of exposure for all experiments.

Immunofluorescence to detect SK1 was performed as previously described.15Small-intestinal tissues in mice were collected and embedded in oxytetracycline compound and frozen, and cryosections were incubated with anti–SK1 antibody (AP7237; Abgent, San Diego, CA) overnight at 4°C. After washing with phosphate-buffered saline, the sections were incubated with Texas red–conjugated goat anti–rabbit IgG (Vector Laboratories, Burlingame, CA) for 1 h at room temperature. For nuclear staining, 4′,6-diamidino-2-phenylindole (blue) was placed on sections for 1 min. The sections were mounted with antifade reagent (Prolong Gold; Invitrogen, Carlsbad, CA) and observed under a fluorescence microscope (model IX81; Olympus, Center Valley, PA).

Small intestines were collected and homogenized mechanically in cell lysate buffer (100 mM sucrose; 1 mM EGTA; 20 mM 4-morpholinepropanesulfonic acid (MOPS), pH 7.4; 5% Percoll; 0.01% digitonin; and protease [Calbiochem] and phosphatase inhibitors) on ice. After a 1,000 g  spin for 15 min to pellet cellular debris, protein concentrations were determined. SK activity was measured as described by Vessey et al .16using 20 μg protein, with some modifications, as previously described.17 

S1P concentrations were measured in the small intestines of mice using high-performance liquid chromatography, as previously described.18 

Protein content was determined with the bicinchoninic acid protein assay reagent (Thermo Scientific), with bovine serum albumin as a standard.

The data were analyzed with a two-tailed Student t  test when comparing means between two groups. A one-way ANOVA, plus a Tukey post hoc  multiple comparison test, was used when comparing multiple groups. The ordinal values of the liver injury scores were analyzed by the Mann–Whitney nonparametric test. In all cases, P < 0.05 was considered significant. All data are expressed throughout the text as mean ± SEM.

Twenty-four hours after renal IRI, mice exposed to pentobarbital anesthesia developed significant renal dysfunction, as indicated by an increase in plasma creatinine concentration (2.39 ± 0.05 mg/dl, n = 10, P < 0.001 vs . sham) higher than sham values (0.47 ± 0.03 mg/dl, n = 6). Isoflurane exposure after renal IRI protected the kidneys, as evidenced by a significant decline in plasma creatinine concentrations (1.61 ± 0.17 mg/dl, n = 8, P < 0.001 vs . pentobarbital renal IRI).

Twenty-four hours after renal IRI, mice developed acute hepatic injury with pentobarbital exposure, as indicated by an increase in plasma ALT concentration higher than sham concentrations (fig. 1A). In contrast, isoflurane exposure after renal IRI protected against liver injury, with a significant reduction in plasma ALT concentrations (fig. 1A).

To evaluate the role of SK in mediating the protective effects of isoflurane after renal IRI, we treated some animals with SKI-II before the induction of renal IRI. SKI-II administration had no detrimental effects on renal function (creatinine concentration, 0.43 ± 0.04 mg/dl; n = 3) in sham–operated mice. There were no differences in plasma creatinine values with SKI-II treatment before renal IRI with pentobarbital exposure (creatinine concentration, 2.53 ± 0.18 mg/dl; n = 6) compared with isoflurane exposure (creatinine concentration, 2.32 ± 0.09 mg/dl; n = 5; P = 0.47 vs . SKI-II pentobarbital renal IRI). Plasma ALT concentrations increased in SKI-II–treated mice exposed to pentobarbital after renal IRI compared with sham mice (fig. 1B). There was no reduction in plasma ALT concentrations after isoflurane exposure.

In addition, we used a strain of mice deficient in the SK1 enzyme. There were no differences in plasma creatinine values in SK1 knockout mice after renal IRI with pentobarbital exposure (creatinine concentration, 2.33 ± 0.18 mg/dl; n = 8) compared with isoflurane exposure (creatinine concentration, 2.25 ± 0.38 mg/dl; n = 6; P = 0.82 vs . SK1 knockout pentobarbital renal IRI). SK1 knockout mice exposed to pentobarbital after renal IRI had increased plasma ALT concentrations compared with sham–operated mice with no reduction with isoflurane exposure (fig. 1B).

In figure 2and figure 3, the protective effects of isoflurane anesthesia are supported further by representative histologic slides. Compared with sham-operated mice (figs. 2A, 3A, and 3B), pentobarbital exposure after renal IRI led to marked hepatocyte vacuolization (fig. 2B) and profound epithelial villous swelling and apoptosis in the small intestine (fig. 3, C and D). Isoflurane exposure after renal IRI dramatically attenuated these injuries in the liver (fig. 2C) and small intestine (fig. 3, E and F).

After renal IRI, the predominant component of hepatic injury was vacuolization. Pentobarbital exposure after renal IRI resulted in severe hepatic vacuolization, as demonstrated by the vacuolization score (3.4 ± 0.2; n = 5; score range, 0–4). In contrast, hepatic vacuolization was less severe with isoflurane exposure after renal IRI (1.6 ± 0.4, n = 4, P < 0.05 vs . pentobarbital renal IRI). Sham mice had normal hepatic histologic features, as demonstrated by a vacuolization score of 0.0 ± 0.0 (n = 4).

We failed to detect significant TUNEL-positive cells in small-intestinal sections from sham–operated on mice exposed to pentobarbital (fig. 4A) or isoflurane (fig. 4B). Mice exposed to pentobarbital after renal IRI (fig. 4C) showed many TUNEL-positive cells in the small intestine (representative of 4 experiments). Mice exposed to isoflurane after renal IRI (fig. 4D) had a reduction in TUNEL-positive cells in the small intestine.

We found increased mRNA expression of TNF-α, ICAM-1, IL-17A, MCP-1, MIP-2, and IL-6 in the livers and small intestines of mice exposed to pentobarbital after renal IRI compared with sham mice.3When mice were exposed to isoflurane after renal IRI, there were significantly reduced expressions of most proinflammatory mRNAs (TNF-α, IL-17A, MCP-1, ICAM-1, and MIP-2) in both the liver and intestine compared with pentobarbital exposure (fig. 5). Isoflurane exposure decreased IL-6 expression in the liver, but not intestine, after renal IRI.

We measured liver and small-intestinal vascular permeability using Evans blue dye, which binds to plasma proteins; its appearance in extravascular tissues reflects an increase in vascular permeability.19Renal IRI caused significant increases in vascular permeability, as measured by increased Evans blue dye content compared with sham mice in the liver, jejunum, and ileum (fig. 6). Vascular permeability was significantly decreased with isoflurane exposure after renal IRI in the liver, jejunum, and ileum.

Isoflurane anesthesia increased small-intestinal SK1 mRNA expression (fig. 7) and protein expression (fig. 8) in sham mice compared with mice anesthetized with pentobarbital. However, there were no changes in SK2 mRNA or protein expression in the small intestine after isoflurane exposure (figs. 7 and 8).

Compared with pentobarbital anesthesia (fig. 9A), isoflurane anesthesia increased staining for SK1 in the small intestine, specifically in intestinal crypts (red fluorescence, fig. 9B, representative of 4 experiments).

The small intestines of sham mice anesthetized with isoflurane demonstrated higher SK activity compared with mice anesthetized with pentobarbital (fig. 10A). Correspondingly, small-intestinal S1P concentrations were higher in mice after isoflurane anesthesia than after pentobarbital anesthesia (fig. 10B).

The major finding of this study is that a clinically relevant concentration of isoflurane (1 minimum alveolar concentration) administered after renal IRI reduced the degree of renal and hepatic dysfunction and hepatic and intestinal injury by reducing inflammation and apoptosis while improving vascular permeability. The protective effects of isoflurane were mediated by the SK1/S1P pathway because isoflurane failed to protect mice treated with an inhibitor of SK activity (SKI-II) or in mice lacking the SK1 enzyme. Moreover, isoflurane induced small-intestinal SK1 mRNA, protein expression, and enzymatic activity, leading to higher S1P concentrations in the small intestine.

AKI continues to be a significant clinical problem in the perioperative period, and the development of postoperative AKI requiring renal replacement therapy has a 60% mortality rate.20Extrarenal dysfunction associated with AKI, including liver and respiratory failure, was predictive of in-hospital mortality.21AKI is an inflammatory process involving multiple cellular and systemic responses, including activation of proinflammatory cytokines and chemokines22and infiltration by leukocytes, such as neutrophils, macrophages, and T cells.23Indeed, modulation of the inflammatory cascade (e.g ., with adenosine generation and signaling modulation via  ecto-5′-nucleotidase24and the A2B-adenosine receptor25) provides powerful protection against AKI after renal IRI. There is growing interest in the extrarenal manifestations of AKI because it is becoming clear that AKI leads to a systemic inflammatory state affecting distant organs. AKI caused increases in pulmonary capillary leak and interstitial neutrophil infiltration via  IL-6,26led to worsening of cardiac function and induced cardiomyocyte apoptosis,27increased neuronal pyknosis and microgliosis in the brain, and increased the proinflammatory cytokines keratinocyte-derived chemokine (KC) and granulocyte-colony stimulating factor (G-CFS).28 

Previous studies29have shown that the gut plays an important role in mediating the hyperdynamic response early in sepsis. Hepatic dysfunction after AKI has been described30; recently, Paneth cells, located in small-intestinal crypts, were identified as the source of the inflammatory mediator, IL-17A, seen in mice after AKI.31The release of IL-17A from Paneth cells led to hepatic dysfunction and a cascade of inflammation, including generation of TNF-α and IL-6. Furthermore, mice deficient in TNF-α, IL-17A, or IL-6 or mice treated with antibodies to TNF-α, IL-17A, or IL-6 had attenuation of hepatic and small-intestinal inflammation.3 

In our model, isoflurane, decreased the expression of IL-17A, TNF-α, MIP-2, MCP-1, and ICAM-1 in the liver and small intestine after renal IRI, reflecting the direct antiinflammatory effects of isoflurane. The antiinflammatory effects of volatile anesthetics are well described because they decreased TNF-α–mediated release of heme oxygenase-1 and IL-8 in human monocytic THP-1 cells32and reduced KC and MIP-2 in a model of endotoxin-induced lung injury.33 

IL-17A is involved in innate immune defense and inflammation and is mainly produced by a subset of T cells known as T helper (Th) 17 cells.34Th17 cells can be found in the intestinal lamina propria  of healthy mice35but rarely in the spleen, mesenteric lymph nodes, or Peyer patches.36IL-17A derived from intestinal Paneth cells mediated shock induced by TNF-α37; we demonstrated that AKI induces production of IL-17A in Paneth cells, causing increased concentrations of additional cytokines, including TNF-α,31a potent mediator of the inflammatory cascade seen after AKI.38Herein, we show that isoflurane up-regulated SK1 in intestinal crypts, revealing a potential mechanism for the decrease in IL-17A and TNF-α, both produced in intestinal crypts, seen after isoflurane exposure. MIP-2 is a chemokine involved in inflammation and immunoregulation and is a potent regulator of neutrophil chemotaxis.39MCP-1 mediates inflammation40and ICAM-1 regulates neutrophil retention41after AKI. Isoflurane exposure reduced IL-6 expression after renal IRI in the liver but not in the small intestine, suggesting a divergent role of this cytokine in the liver and intestine. Indeed, IL-6 plays a role in liver inflammation and progression to hepatocellular carcinoma42but protects enterocytes against cell death and apoptosis and protected mice against intestinal IRI.43 

The protective effects of volatile anesthetics have been shown in multiple organ systems, including the brain,44heart,45and lung.33In our model, mice were protected from AKI-induced liver and intestine injury with exposure to isoflurane after renal IRI (i.e ., postconditioning). Clinically, volatile anesthetics can be administered outside of the operating room in the intensive care unit,46and this may potentially have therapeutic benefits in patients with AKI. We demonstrated that the protective effects of volatile anesthetics on the kidney and cultured human proximal tubule cells were mediated via  the SK/S1P pathway17,18; recently, isoflurane postconditioning protected neonatal rats from hypoxic–ischemic brain injury via  a mechanism dependent on SK/S1P signaling.44Given this knowledge, we tested whether there was a role for SK/S1P signaling in protecting against extrarenal organ dysfunction after renal IRI. We found that administration of SKI-II reversed the isoflurane-mediated protection from renal IRI. In addition, we demonstrated that mice lacking the SK1 enzyme were not protected against liver and intestinal injury after renal IRI with isoflurane. Because of the inherent concerns regarding the use of genetic knockout mice (e.g ., alterations in the expression of unrelated proteins), we used both a pharmacologic inhibitor and genetic knockout mice to study the role of SK in renal IRI. By using immunofluorescence imaging, we determined that the small-intestinal crypts had the greatest increase in staining for SK1. Taken together, these data indicate that activation of the SK1/S1P pathway in small-intestinal crypts is involved in mediating the protective effects of isoflurane on the liver and intestine after renal IRI.

The lysophospholipid S1P has multiple roles in cellular signaling and balances against the proapoptotic effects of sphingosine and ceramide via  the “sphingolipid rheostat.”47S1P protected intestinal cells from apoptosis via  serine/threonine-specific protein kinase (Akt) activation48and mediated the protective effects of glucocorticoids on renal mesangial cells.49S1P reduced IRI injury in a rat model of lung transplantation50and protected against renal IRI injury in mice.51S1P binds to specific G protein–coupled receptors, of which five are known (S1P1–5).8FTY720, a sphingosine analog that is phosphorylated in vivo , has produced lymphopenia through its actions on the S1P1 receptor by reducing lymphocyte egress from lymph nodes and has shown protection in models of liver,52bowel,53and renal54ischemia.

SK, the enzyme catalyzing the conversion of sphingosine to S1P, is an important regulator of the sphingolipid rheostat. Many agents are known to stimulate SK activity, including agonists of growth factor receptors (e.g ., platelet-derived growth factor [PDGF], vascular endothelial growth factor [VEGF], nerve growth factor [NGF], and epidermal growth factor [EGF]), transforming growth factor β, and TNF-α.55Of the two known isoforms, SK1 mediates cytoprotection, whereas SK2 is generally thought to be proapoptotic.56Indeed, SK1 mediated ischemic postconditioning in isolated mouse hearts57and was protective via  down-regulation of c-Jun N-terminal kinase (JNK) activity.58Mice lacking the SK1 enzyme had poor recovery from anaphylaxis and delayed histamine clearance, whereas mice lacking the SK2 enzyme had rapid recovery from anaphylaxis.59In certain models, such as Crohn disease, SK had a deleterious effect because SK inhibition60or mice lacking the SK1 enzyme61had reduced inflammation and colon damage.

Anesthetics can cause hemodynamic changes and regional disturbances in blood flow. Previously, it was demonstrated that, compared with pentobarbital, volatile anesthetics did not significantly alter systemic blood pressure or renal blood flow.17Certain volatile anesthetics, such as methoxyflurane, undergo renal metabolism with direct nephrotoxic effects due to inorganic fluoride. However, isoflurane is minimally metabolized and has not been linked to fluoride nephrotoxicity.62 

A limitation of our study is that volatile anesthetics may have both local and systemic effects for reducing the severity of hepatic and intestinal injury after renal IRI; therefore, it is difficult to determine in an in vivo  study whether volatile anesthetics mediate their protective effects directly on small-intestinal crypts or via  systemic effects on the migration and infiltration of lymphocytes and neutrophils. The effects are likely multifactorial, including both direct cytoprotective effects on hepatocytes and intestinal epithelial cells with activation of prosurvival signaling pathways, such as the extracellular-signal-regulated kinase (ERK) and Akt pathways, and systemic antiinflammatory mechanisms, including peripheral lymphopenia, preservation of endothelial barrier integrity, and reduction of proinflammatory cytokines.

In conclusion, we demonstrated that isoflurane activates the SK1/S1P signaling pathway in small-intestinal crypts to reduce hepatic and intestinal injury, apoptosis, and proinflammatory mRNA up-regulation after renal IRI. Further elucidation of the mechanisms of protection may lead to advancements in the treatment of extrarenal organ dysfunction after renal IRI.