The Effect of Midazolam and Propofol on Interleukin-8 from Human Polymorphonuclear Leukocytes

Abstract

Anesthetics and sedatives contribute to postoperative immunosuppression.Interleukin-8 (IL-8) is a chemotactic and activating factor that mediates neutrophil adhesion and margination and is essential for host defense. We investigated the effect of anesthetics on isolated human polymorphonuclear leukocyte production of IL-8. Healthy human polymorphonuclear leukocytes were isolated using a single-step density gradient and stimulated with lipopolysaccharide in the presence of varying concentrations of propofol or midazolam for up to 20 h. IL-8 was measured in both culture supernatants and cell lysates using enzyme immunoassay, and IL-8 mRNA in cells was measured using Northern blotting and phosphorimaging. Data were analyzed using Kruskal-Wallis analysis of variance or the Mann-Whitney U-test as appropriate. Lipopolysaccharide increased extracellular accumulation of interleukin-8, which was suppressed by both propofol (P = 0.025) and midazolam (P = 0.028). However, intracellular IL-8 increased with exposure to lipopolysaccharide (P = 0.028) and remained increased with both anesthetics. Northern blot analysis also revealed increased IL-8 mRNA levels in the presence of both midazolam and propofol, which was confirmed by molecular imaging. These data strongly suggest that the anesthetics modulate transport or secretion of IL-8 protein from the cell. Suppression of IL-8 by anesthetics and sedatives may predispose postoperative and intensive care patients to infection. Implications: Anesthesia causes immune suppression and alters neutrophil function. We investigated the effect of propofol and midazolam on interleukin-8, a neutrophil chemotactic agent in human neutrophils. Both anesthetics decreased extracellular interleukin-8 accumulation, but intracellular levels and mRNA remained high. This suggests that propofol and midazolam alter interleukin-8 secretion from cells.

(Anesth Analg 1998;86:1289-93)

Trauma, surgical stress, and anesthesia are associated with postoperative immune suppression and increased susceptibility to infection. The direct effects of several anesthetics on immune function have been described ^[1]. Such immunomodulation may be particularly relevant in intensive care unit (ICU) patients, in whom anesthetics such as propofol and midazolam are used for long periods by infusion as sedatives. These drugs may contribute to the predisposition of such patients to nosocomial infections.

Interleukin-8 (IL-8) is a potent chemotactic and activating factor produced by many cell types, including neutrophils, in response to proinflammatory stimuli, such as endotoxin, interleukin-1 beta (IL-1 beta), tumor necrosis factor alpha (TNF-alpha), and viruses ^[2,3]. Neutrophils themselves further augment neutrophil recruitment and activation through the local production of IL-8 at sites of infection, thus amplifying antimicrobial defense. IL-8 levels are increased in inflammatory conditions that involve neutrophil infiltration, including acute respiratory distress syndrome ^[4]. The action of anesthetics on IL-8 release has not been studied.

We hypothesized that the immunosuppressive actions of anesthesia may be explained by altered IL-8 production or release in addition to the reported effects on chemotactic responses. We investigated the effect of two IV anesthetics often used as sedatives in ICU patients on both IL-8 release and IL-8 mRNA expression from isolated healthy human polymorphonuclear leukocytes (PMN).

Methods

Normal PMN were isolated from healthy volunteers using a single-layer density gradient procedure ^[5] after institutional review board approval and informed consent. Blood was added to sterile, preservative-free heparin (10 U/mL of blood) and layered onto Polymorphprep[trade mark sign] separation medium (Nycomed UK Ltd., Birmingham, West Midlands, UK) at a rate of 5 mL of blood per 3.5 mL of medium. After centrifugation at 475g for 30 min at 20[degree sign]C, the mononuclear cell band was discarded and PMN were retrieved and washed in Dulbecco's modified Eagle's medium (ICN Biomedicals Ltd, Thame, Oxon, UK). Contaminating erythrocytes were removed by hypotonic shock, and cells were washed three times in Dulbecco's modified Eagle's medium before being resuspended, counted, and adjusted to a density of 3 x 10⁶ cells/mL.

Duplicate 0.5-mL volumes of cells were placed into 24-well tissue culture plates (1.5 x 10⁶), and 2 [micro sign]g/mL lipopolysaccharide (LPS; Sigma-Aldrich Chemical Company Ltd., Poole, Dorset, UK) or isotonic sodium chloride solution was added as appropriate (50 [micro sign]L/well). To each well was also added 50 [micro sign]L of the anesthetic, either midazolam or propofol. The anesthetics were initially prepared by diluting an appropriate volume in 100 [micro sign]L of 70% (vol/vol) ethanol, then diluting 1:10 with phosphate-buffered saline (PBS; Dulbecco's A, pH 7.4). Lower concentrations were prepared by doubling the dilution in PBS. Zero concentrations comprised similar dilutions of Intralipid (10%; Phamacia-Upjohn, Milton Keynes, Hertshire, UK), the solvent for propofol, or PBS in appropriate concentrations of ethanol. The concentrations to which the cells were exposed (1-20 [micro sign]M) were chosen to mimic circulating sedative to anesthetic concentrations ^[6,7].

Cells were incubated at 37[degree sign]C in an atmosphere of 5% CO_2/95% air for 20 h. Culture supernatants were removed and stored at -80[degree sign]C for IL-8 measurement. Cells were disrupted by sonication on ice, and lysates were stored at -80[degree sign]C for protein and IL-8 measurement.

IL-8 was measured using solid-phase sandwich enzyme-linked immunosorbent assay (R&D Systems Europe Ltd., Abingdon, Oxon, UK). The assay is highly specific for IL-8 and shows no cross-reactivity for any of the human cytokines tested. The within-assay precision (coefficient of variation) was 3.2% (n = 20), and the between-assay precision was 9.5% (n = 20) for cell culture supernatants at a concentration of 500 ng/mL. Results are expressed in terms of protein content, measured using a modification of the Lowry et al. technique ^[8].

In some experiments, total RNA was isolated from cells after 4 h using guanidinium thiocyanate and phenol extraction (Totally RNA kit; Ambion Inc., Austin, TX). Northern blotting was performed using standard techniques with digoxigenin-labeled probes for human IL-8 mRNA (R&D Systems), with detection using the substrate CDP-Star (Boehringer Mannheim, East Sussex, UK). Results were quantified by phosphorimaging using Molecular Analyst software (Bio-Rad Laboratories, Hemel Hempstead, Hertshire, UK). Constant mRNA loading was confirmed by the analysis of mRNA for the housekeeping gene beta actin (R&D Systems). Cell viability was determined using trypan blue exclusion.

Data are expressed as median (range) and represent results from six to eight separate experiments. Data were assessed using Microsoft Excel 5.0. Kruskal-Wallis one-way analysis of variance was used to assess changes across a concentration range. The Mann-Whitney U-test was used to assess differences between the two groups and for post hoc testing as appropriate. Differences were taken to be significant when P < 0.05.

Results

Midazolam and propofol had no effect on cell viability, as assessed by exclusion of the vital stain trypan blue, at any of the concentrations studied compared with cells incubated without anesthetic (data not shown). Incubation of cells with 2 [micro sign]g/mL LPS for 20 h resulted in increased IL-8 concentrations in the culture medium (Figure 1). The median (range) IL-8 was 0.08 (0.02-0.19) ng/mg protein without LPS and 11.28 (6.08-11.64) ng/mg protein with LPS (P = 0.002). Intracellular IL-8 concentrations increased from 5.19 (3.68-9.00) ng/mg protein to 12.52 (7.88-21.3) ng/mg protein with exposure to LPS (P = 0.025; Figure 2). Cells incubated without LPS, but with the highest concentration of anesthetic, showed IL-8 concentrations similar to those without anesthetic (Figure 1). Extracellular IL-8 accumulation was slightly decreased in the presence of intralipid alone, but this was not significant (P = 0.061) and was much less than that of propofol (Figure 1).

Figure 1: Extracellular interleukin-8 (IL-8) accumulation from isolated human polymorphonuclear leukocytes. Cells were exposed to propofol or midazolam with and without lipopolysaccharide (LPS) for 20 h, and IL-8 was measured in the culture medium. Box and whisker plots indicate median, 25th and 75th percentiles, with ranges as vertical lines. Data represent six to eight separate experiments and were assessed using the Mann-Whitney U-test or Kruskal-Wallis one-way analysis of variance as appropriate.

Figure 2: Intracellular interleukin-8 (IL-8) concentrations from isolated human polymorphonuclear leukocytes. Cells were exposed to propofol or midazolam with and without lipopolysaccharide (LPS) for 20 h, and IL-8 was measured in cell lysates. Box and whisker plots indicate median, 25th and 75th percentiles, with ranges as vertical lines. Data represent six to eight separate experiments and were assessed using the Mann-Whitney U-test or Kruskal-Wallis one-way analysis of variance as appropriate.

Both propofol and midazolam caused significant decreases in LPS-stimulated extracellular IL-8 accumulation (P = 0.025 for propofol and P = 0.028 for midazolam, Figure 1). Levels of mRNA for IL-8 were increased when cells were exposed to LPS for 4 h (Figure 3) and further increased with exposure to either propofol or midazolam, as seen with both Northern blotting and quantitative phosphorimaging. Exposure of cells to anesthetics alone had no effect on IL-8 mRNA.

Figure 3: A representative Northern blot of interleukin-8 mRNA from isolated human polymorphonuclear leukocytes. Cells were exposed to propofol (P) or midazolam (M) with and without lipopolysaccharide (LPS) for 4 h.

Intracellular IL-8 concentrations in LPS-stimulated cells were increased in the presence of propofol and midazolam, and they tended to increase as the anesthetic concentration increased (P = 0.0039 for propofol and P = 0.1 for midazolam; Figure 2).

Discussion

The anesthetics/sedatives propofol and midazolam have marked effects on IL-8 release from stimulated human PMN. The results of this study suggest that the mechanism of the effect of anesthetics on IL-8 is at the posttranslational level.

The first line of cellular host defense to inflammation is the neutrophil. It is a pivotal effector cell, able to respond to mediators and recruit additional neutrophils to sites of inflammation. The directed migration of neutrophils in response to specific neutrophil chemotactic and activating factors, such as IL-8, is essential to their margination from the circulation. Participation of neutrophils at the site of inflammation results in phagocytosis and removal of antigenic material. Activation and margination of neutrophils in response to IL-8 leads to the release of a variety of damaging species, including proteases, oxygenderived free radicals, hypochlorous acid, and other cytokines. In addition, neutrophils have an important role in autocrine and paracrine regulation of immune cells via production of cytokines ^[2].

IL-8 is released by neutrophils, which therefore augment their own recruitment and activation through further local IL-8 production, resulting in the accumulation of neutrophils at sites of inflammation. Patients in the ICU are often sedated with propofol and midazolam for several days. We found that these often used sedatives significantly decreased IL-8 release from LPS-stimulated neutrophils. Previous studies have shown that neutrophils do not constitutively express mRNA for IL-8 but that they rapidly respond in a dose-dependent manner to LPS stimulation with maximal steady-state mRNA expression at four to eight hours ^[3,9]. IL-8 protein is released from neutrophils in response to LPS in two distinct phases; an initial release directly responding to LPS at approximately 6 hours, followed by a later release at up to 20 hours, which is thought to be in response to LPS-stimulated IL-1 beta and TNF-alpha release ^[2]. We therefore measured mRNA expression at 4 hours and immunoreactive IL-8 protein accumulation at 20 hours.

We have shown that both propofol and midazolam cause decreased release of IL-8 from LPS-stimulated neutrophils. Propofol also suppresses other aspects of neutrophil function, including respiratory burst activity ^[10], polarization ^[11], and chemotaxis ^[12]. Midazolam, however, had no effect on neutrophil respiratory burst or polarization ^[10,11] at clinically relevant concentrations, but perhaps the stimulant used may determine the effects seen ^[10]. Other studies have shown a variety of immunomodulatory effects for several anesthetics. Halothane, isoflurane, and enflurane decreased, and desflurane increased, neutrophil function in terms of hydrogen peroxide production ^[14]. Methohexital and thiopental decreased neutrophil respiratory burst activity ^[13]. Mitogen stimulation of T-cell proliferation was decreased by thiopental, methohexital, and etomidate, but propofol had no effect ^[15]. There have been no studies on cytokine production, nor has the effect of anesthetics on IL-8 been investigated.

Propofol is supplied prepared in a 10% soybean emulsion (Intralipid), which was included as a control drug. Previous in vitro studies have suggested that Intralipid itself may cause immunosuppression in terms of neutrophil respiratory burst activity ^[10], neutrophil chemotaxis ^[15], and IL-2-dependent lymphocyte responses ^[16]. However, Intralipid had no effect on in vitro neutrophil polarization ^[11] or T-cell proliferation ^[15,17]. The in vivo infusion of Intralipid, however, caused increased ex vivo T-cell proliferation ^[18], but decreased the chemotactic migration of leukocytes ^[19]. The present study showed that Intralipid caused a small decrease in extracellular IL-8 accumulation, but the effect of propofol exceeded that of Intralipid alone.

Physiological regulation of IL-8 expression occurs at the transcriptional level, mediated through binding of transcription factors including NF kappa B and NF-IL-6 to sequences located within the IL-8 promoter ^[20]. Although some anesthetics inhibit protein kinase C, a key component in signal transduction ^[21], we have shown that neutrophil IL-8 mRNA expression was increased by propofol and midazolam despite reduced extracellular IL-8 accumulation. This suggests that the effect on IL-8 was not the result of altered transcriptional regulation, because effects at this level should have reduced mRNA levels. Intracellular IL-8 remained increased in the presence of increasing concentrations of propofol, which indicates that the effect on IL-8 occurs at the posttranslational level. Anesthetics affect cell membrane fluidity, receptor function, intracellular calcium, and a variety of transport activities ^[22-25]. In eukaryotic cells, proteins destined for secretion are assembled in the endoplasmic reticulum, packaged into transport vesicles, and conveyed through the Golgi apparatus to the plasma membrane. This is termed the classical secretory pathway. However, some proteins, including the cytokine IL-1 beta, are not transported from the cell by the classical endoplasmic reticulum-Golgi route ^[26]. It is not known whether IL-8 protein is secreted via the classical or nonclassical route. The classical secretory mechanism may be inhibited via blockage of transport through the Golgi apparatus, failure of glycosylation, altered calcium transport, or inadequate vesicle packaging. The nonclassical pathway may be inhibited via nonspecific membrane damage or calcium-mediated effects. We suggest that propofol and midazolam inhibit secretion of LPS-stimulated IL-8 from neutrophils, which is in line with the finding that propofol inhibits neuropeptide secretion from mouse pituitary cells ^[27]. Further study is required to elucidate the mechanisms involved in IL-8 secretion and the point at which anesthetics exert their effects.

Immunosuppression is common postoperatively due to the combined effects of surgery and anesthesia, and patients may be predisposed patients to infection. Adequate early responses of neutrophils to an inflammatory or infective insult are essential to host defense, and blunting of IL-8 production by these cells may contribute to postoperative immunosuppression. ICU patients receiving sedation may be additionally predisposed to compromised immune function as a result of the immunosuppressive effect of anesthetics. In addition, in vitro studies show that endothelialderived IL-8 may inhibit neutrophil adhesion to endothelial cells ^[28]. Clearly, the modulatory effects of anesthetics and sedatives on cellular host defense are an important consideration and require further study.