SEARCH

SEARCH BY CITATION

Keywords:

  • glucocorticoids;
  • inflammation;
  • interleukin-22;
  • sepsis

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

Interleukin (IL)-22 production triggered by innate immune mechanisms has been identified as key to efficient intestinal anti-bacterial host defence and preservation of homeostasis. We hypothesized that glucocorticoid therapy may impair IL-22 expression, which should promote intestinal epithelial damage with the potential of subsequent bacterial translocation. High-dose corticosteroid therapy in Crohn's disease has been associated with an increased rate of abscess formation and ultimately with a higher risk of developing postoperative infectious complications, including abdominal sepsis. Thus, we sought to investigate effects of the prototypic glucocorticoid dexamethasone on IL-22 production in the context of bacterial infection. Enhanced IL-22 plasma levels were detectable in rat sepsis. Moreover, heat-inactivated Staphylococcus epidermidis, used as a prototypic activator of innate immunity, induced robust production of IL-22 by human peripheral blood mononuclear cells (PBMC). Here, we report for the first time that dexamethasone mediates remarkable suppression of IL-22 as detected in S. epidermidis-activated PBMC and rat sepsis, respectively. The data presented herein suggest that insufficient IL-22 function may contribute to impaired intestinal host defence in the context of corticosteroid therapy.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

Interleukin (IL)-22 [1] has been identified as a prototypic constituent of the currently evolving T helper type 17 (Th17)-like cytokine response [2,3] that is produced mainly by activated T cells and natural killer cells [1] and, under specific conditions, by macrophages [4]. For reasons of structural similarity and shared usage of the R2 IL-10 receptor chain, IL-22 is considered a member of the IL-10 cytokine family [1]. With regard to function, IL-22 appears to connect specifically to IL-6 [5]. Activation of signal transducer and activator of transcription-3 (STAT3) and mitogen-activated protein kinases [1] is essential for signal transduction initiated by both cytokines. In addition, IL-22 is able to induce proinflammatory genes by activating nuclear factor-κB, at least in colonic subepithelial myofibroblasts [6]. Notably, bioactivity of both IL-22 and IL-6 is suppressed efficiently by action of suppressor of cytokine signalling-3 [7], a modulatory pathway not applying to IL-10 [8]. As a result of restrictive IL-22R1 expression, IL-22, but obviously not IL-6, fails to activate leucocytes [1], implying that IL-22 serves to signal specifically between the leucocytic and the non-leucocytic cell compartment in particular cells of epithelial origin.

Recent data obtained from animal models indicate that the function of IL-22 in inflammatory diseases is complex and dependent on the particular pathophysiological context. Specifically, IL-22 is protective in experimental hepatitis [9,10] and inflammatory bowel diseases (IBD) [11,12], but mediates dermal inflammation and disease in murine psoriasis [13]. Notably, IL-22 is detectable in areas of active disease in IBD patients [14]. Based on the murine model of oral infection by Citrobacter rodentium, IL-22 is suggested as being pivotal for efficient innate immunity and host defence in the intestinal microenvironment [15]. In fact, the IL-22-inducible anti-microbial proteins RegIIIβ, RegIIIγ[15], lipocalin-2 [16] and possibly IL-22-inducible anti-bacterial inducible nitric oxide synthase (iNOS) [17–19] may play a decisive role for pathogenesis seen in murine models of bacterial infection at epithelial interfaces. However, the role of IL-22 in intestinal homeostasis seems Janus-faced. By action on intestinal epithelial cells and colonic subepithelial myofibroblasts, IL-22 is similarly capable of up-regulating highly inflammatory genes, including IL-8 and matrix metalloproteinases [6,14]. Interestingly, the destructive potential of IL-22 has been recognized recently in a study demonstrating that the cytokine is able to disrupt blood–brain barrier tight junctions [20]. Proper tight junctions are obviously key to an efficient intestinal barrier [21]. Thus, it is tempting to speculate that IL-22, surpassing a certain threshold, converts from a predominantly protective to a pathological parameter of disease.

Usage of high-dose glucocorticoids in Crohn's disease patients has been associated with an increased rate of abscess formation [22] and ultimately with a higher risk for postoperative infectious complications, including abdominal sepsis [23,24]. Here, we set out to investigate effects of dexamethasone on IL-22 expression driven by activation of innate immunity in the context of bacterial infection.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

Materials

Dexamethasone and dexamethasone 21-phosphate disodium salt were from Calbiochem-Novabiochem (Bad Soden, Germany) and Sigma (Taufkirchen, Germany), respectively. Heat-inactivated Staphylococcus epidermidis was kindly provided by Professor Charles Dinarello (University of Colorado Health Sciences Center).

Isolation and cultivation of peripheral blood mononuclear cells (PBMC)

The study protocol and consent documents for isolation of PBMC from healthy donors were approved by the ‘Ethik-Kommission’ of the University Hospital Goethe-University Frankfurt. Informed consent was obtained from all donors. Donors abstained from using any drugs for 2 weeks before the study. PBMC were freshly isolated from peripheral blood using Histopaque®-1077, according to the manufacturer's instructions (Sigma). PBMC were resuspended in RPMI-1640 supplemented with 10 mM HEPES, 100 U/ml penicillin, 100 µg/ml streptomycin and 1% human serum (Invitrogen, Karlsruhe, Germany) and were seeded at 3 × 106 cells/1 ml in polypropylene tubes (Greiner, Frickenhausen, Germany). Cultivation of PBMC was performed at 37°C and 5% CO2.

Determination of IL-22 mRNA by reverse transcriptase–polymerase chain reaction (RT–PCR) techniques

After RNA isolation (TriReagent; Sigma) and reverse transcription (Applied Biosystems, Darmstadt, Germany), human IL-22 mRNA was determined semiquantitatively by standard PCR and real-time PCR. For standard PCR, the following sequence was performed for each PCR reaction: {95°C for 10 min (1 cycle); 95°C for 30 s, 60°C [glyceraldehyde-3-phosphate dehydrogenase (GAPDH); 23 cycles] or 58°C (IL-22; 40 cycles) for 1 min, and 72°C for 1 min; and a final extension phase at 72°C for 7 min}. The following primers were used: IL-22, forward: 5′-GCTAAGGAGGCTAGCTTG-3′, reverse: 5′-CAGCAAATCCAGTTCTCC-3′; GAPDH, forward: 5′-ACCACAGTCCATGCCATCAC-3′, reverse: 5′-TCCACCACCCTGTTGCTGTA-3′. Identity of amplicons was confirmed by sequencing (Abi Prism 310 Genetic Analyzer; Applied Biosystems).

For real-time PCR, changes in fluorescence are caused by the Taq-polymerase degrading the probe that contains a fluorescent dye (IL-22: FAM, GAPDH: VIC) and a quencher (TAMRA). For IL-22 (#Hs00220924_m1) and GAPDH (#4310884E) predeveloped assay reagents were obtained (Applied Biosystems). The assay mix was from Invitrogen (Karlsruhe, Germany). PCR (Abi Prism 7500 Sequence Detector; Applied Biosystems): one initial step at 50°C for 2 min and 95°C for 2 min was followed by 40 cycles at 95°C for 15 s and 60°C for 1 min. Detection of the dequenched probe, calculation of threshold cycles (Ct values) and data analysis were performed by the Sequence Detector software. Relative changes in IL-22 mRNA expression compared to unstimulated control and normalized to GAPDH were quantified by the 2-ddCt method.

Quantification of cytokines by enzyme-linked immunosorbent assay (ELISA)

Levels of human IL-22 (R&D Systems, Wiesbaden, Germany), IL-8 (BD Biosciences/Pharmingen, Erembodegem, Belgium) and IL-10 (Hoelzel Diagnostics/Diaclone, Cologne, Germany) were determined in cell-free supernatants by ELISA. Plasma concentrations of rat IL-22, IL-6, TNF-α (R&D Systems), IL-18 and macrophage inflammatory protein (MIP)-2α/CXCL2 (Biosource/Invitrogen, Karlsruhe, Germany) were determined by ELISA. Assays were performed according to the manufacturers' instructions.

Surgical procedures

All animal experiments in this prospective study were approved by the governmental board for the care of animal subjects (Regierungspräsidium Darmstadt, Germany, 54-19c20/15-143/12) and were in accordance with the National Institute of Health guidelines (National Academy of Sciences, Washington DC, USA, 1996). Fourteen male Sprague–Dawley rats (Harlan-Winkelmann, Borchen, Germany) were assigned randomly to two groups and used to analyse effects of dexamethasone on IL-22 production in the rat sepsis model of caecal ligation and incision (CLI) (CLI group 2 and dexamethasone group with body weight at 507 ± 44 g, for each group: n = 7). In addition, plasma samples of 24 randomly assigned male Sprague–Dawley rats with body weight at 538 ± 58 g (Harlan-Winkelmann; sham group and CLI group 1, for each group: n = 12) were used to determine IL-22 production in rat sepsis (CLI) compared to sham-treated animals. The sepsis model of CLI as performed herein has been described recently in detail [25]. Briefly, after intraperitoneal (i.p.) injection of pentobarbital (10 mg/kg body weight; Narcoren, Halbergmoos, Germany) and fentanyl (0·05 mg/kg; Janssen-Cilag, Neuss, Germany), rats of all four groups received tracheotomy and mechanical ventilation. Intravenous infusion (i.v.) of 0·9% NaCl (12 ml/kg/h; B. Braun, Melsungen, Germany), pentobarbital (0·6 mg/kg/h) and fentanyl (0·03 mg/kg/h) was continued permanently throughout the entire observation time of 390 min. After establishment of sufficient anaesthesia and baseline arterial blood gas analysis, rats in the CLI group 2 received 250 µl of ddH2O (B. Braun, Melsungen, Germany) intravenously. Rats in the dexamethasone group received i.v. 5 mg/kg body weight dexamethasone 21-phosphate disodium salt dissolved in 250 µl ddH2O. Animals of the sham group and CLI group 1 received no drugs/ddH2O at this point. Subsequently, a median laparotomy was performed in all rats. The caecum was exteriorized carefully in all animals by means of cotton sticks that had been placed in 0·9% NaCl before. In the sham group, the caecum was replaced into the abdomen after gentle manipulation. In the other three groups (CLI group 1, CLI group 2, dexamethasone group) severe sepsis was induced by CLI. Briefly, the caecum and the mesenteric blood vessels were ligated distal of the ileocaecal valve, and a 1·5 cm caecal blade incision was performed. Immediately before closure of the abdominal wall, 2 ml/kg of 0·9% NaCl solution was given i.p. for fluid resuscitation. At the end of the observation time or in case of early death, plasma samples (Heparin-Natrium, Ratiopharm, Ulm, Germany) were obtained.

Statistical analyses

Data obtained from PBMC are shown as absolute concentrations or as means ± standard error of the mean (s.e.m.) and are presented as fold-induction compared to unstimulated control, as pg/ml, as ng/ml or as (% of S. epidermidis alone). Data were analysed by one-way analysis of variance (anova) with post-hoc Bonferroni correction. For the animal experiments, data are shown as absolute IL-22 concentrations or as box plots (median, 25% and 75% quartile with whiskers denoting maximal and minimal data points) and are expressed as pg/ml or as ng/ml. Raw data were analysed by Mann–Whitney rank sum test (GraphPad 4·0).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

Dexamethasone is a potent suppressor of IL-22 and IL-8 but not of IL-10 secretion in human PBMC stimulated with heat-inactivated S. epidermidis

In order to investigate the modulatory potential of dexamethasone on IL-22 production, heat-inactivated S. epidermidis was used to mimic robust activation of innate immunity in the cell culture model of human PBMC. Exposure to S. epidermidis mediated robust IL-22 release by PBMC, which concurs with previous observations on S. aureus[26]. The results showed potent suppression of IL-22 by co-incubation with the glucocorticoid dexamethasone. Real-time (Fig. 1a) and standard PCR (Fig. 1a, inset) revealed IL-22 modulation on the level of mRNA expression. Moreover, suppression of mRNA resulted in significantly impaired IL-22 protein secretion (92·5% inhibition, P < 0·05, n = 5) by activated PBMC (Fig. 1b). Under these same conditions, release of IL-8, a prototypic proinflammatory cytokine, was reduced to a similar extent (Fig. 1c). In contrast, dexamethasone at 50 nM, a concentration that blocked secretion of IL-22 and IL-8 most efficiently, was unable to affect IL-10 release by PBMC (Fig. 1d). The latter observation is in accord with previously reported effects of dexamethasone on IL-10 secretion by lipopolysaccharide-stimulated human whole blood [27]. After having shown that dexamethasone inhibits IL-22 production in human PBMC, this modulatory activity was investigated in vivo by using a rat model of sepsis.

image

Figure 1. Suppression of interleukin (IL)-22 by dexamethasone as detected in peripheral blood mononuclear cells (PBMC) under the influence of heat-inactivated Staphylococcus epidermidis. (a) PBMC from healthy volunteers were either kept as unstimulated control, incubated with dexamethasone alone or stimulated with heat-inactivated S. epidermidis (final concentration: 2·1 × 106 organisms/ml or 0·7 organisms/PBMC) in the presence or absence of the indicated concentrations of dexamethasone. After 16 h, cells were harvested and IL-22 mRNA expression was evaluated by real-time polymerase chain reaction (PCR). IL-22 mRNA expression was normalized to that of glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) and is shown as fold-induction compared to unstimulated control ± standard error of the mean (s.e.m.) (n = 6); ***P < 0·005 compared to unstimulated control; ###P < 0·005 compared to S. epidermidis. Inset: these same RNA isolates were also evaluated by standard PCR. One representative experiment is shown. (b) PBMC obtained from healthy volunteers (n = 5) were either kept as unstimulated control (filled square) or stimulated with heat-inactivated S. epidermidis (final concentration: 2·1 × 106 organisms/ml or 0·7 organisms/PBMC) in the presence or absence of the indicated concentrations of dexamethasone. In addition, PBMC were incubated with dexamethasone alone (open square). After 16 h, cells were harvested and IL-22 concentrations in cell-free culture supernatants were determined by enzyme-linked immunosorbent assay. Representative raw data of two donors are shown as pg/ml (left and middle panel). Right panel, data obtained on modulation of IL-22 release by dexamethasone (50 nM) obtained from five different donors are summarized and expressed as (% of S. epidermidis alone). Mean percentage change in secretion of IL-22 ± s.e.m. (n = 5) is depicted. *P < 0·05 compared to S. epidermidis alone. (c,d) Supernatants obtained from four of these same experiments were analysed further for IL-8 (c) and IL-10 (d) secretion in the presence and absence of dexamethasone at 50 nM. Data are shown as means ± s.e.m. (n = 4). **P < 0·01 compared to unstimulated control; ##P < 0·01 compared to S. epidermidis.

Download figure to PowerPoint

Elevated levels of plasma IL-22 detected in rat bacterial peritonitis

In order to investigate whether IL-22 is a parameter that can be determined easily in the blood compartment during rodent sepsis, levels of this cytokine were verified by ELISA analysis of plasma either obtained from sham-treated rats or from rats that had undergone the CLI procedure. In agreement with data obtained from S. epidermidis-activated human PBMC, we report for the first time on significantly enhanced plasma levels of IL-22 (P < 0·005) that are associated with severe sepsis in rats (Fig. 2). This present observation agrees with strong IL-22 induction as detected by mRNA analysis of spleen and kidney tissues during the course of murine peritonitis [28]. We recently reported on elevated levels of IL-6 under these same experimental conditions, similarly confirming robust activation of systemic inflammation in rats undergoing the CLI procedure [25].

image

Figure 2. Up-regulation of plasma interleukin (IL)-22 in rat bacterial peritonitis; 390 min after laparotomy, plasma levels of IL-22 obtained from animals belonging either to the sham group (n = 12) or to the group undergoing the caecal ligation and incision (CLI) procedure (CLI group 1, n = 12) were determined by enzyme-linked immunosorbent assay analysis. (a) Data are shown as absolute IL-22 concentrations in pg/ml; median plasma IL-22 levels are depicted. (b) These same data are shown as box plots (median, 25% and 75% quartile with whiskers denoting maximal and minimal data points) and expressed as pg/ml. ***P < 0·005 compared to sham-treated animals.

Download figure to PowerPoint

Pronounced inhibition of proinflammatory cytokine and IL-22 production after application of dexamethasone 21-phosphate in vivo

To address the question of whether glucocorticoids impair production of IL-22 in rat abdominal sepsis, modulatory effects of dexamethasone 21-phosphate disodium salt were investigated. This is a water-soluble inorganic ester of dexamethasone that is converted rapidly to dexamethasone in vivo. For that purpose, dexamethasone 21-phosphate disodium salt at a concentration of 5 mg/kg body weight was applied i.v. to rats immediately before initiation of sepsis by CLI. Administration of dexamethasone 21-phosphate did not affect survival in CLI in our experimental set-up (data not shown). Rat abdominal sepsis is characterized by systemic induction of inflammatory cytokines, among others IL-6 [25], TNF-α[29], MIP-2α/CXCL2 [30] and IL-18 [31]. In order to verify the experimental set-up applied herein, the effects of dexamethasone 21-phosphate on production of these parameters were assessed. In fact, we observed significant reduction of plasma IL-6 (P < 0·005, Fig. 3a), TNF-α (P < 0·005, Fig. 3b) and MIP-2α (P < 0·005, Fig. 3c) in response to the drug. Furthermore, the median plasma concentration of IL-18 was diminished similarly by 71·9% in the group of rats treated with the glucocorticoid. However, this reduction did not reach statistical significance in the set of experiments performed (Fig. 3d). Low but detectable amounts of TNF-α and IL-18 agree with previous reports on rat sepsis [29,31].

image

Figure 3. Inhibition of interleukin (IL)-6, tumour necrosis factor (TNF)-α, macrophage inflammatory protein (MIP)-2α and IL-18 by application of dexamethasone 21-phosphate disodium salt in vivo. For each group [caecal ligation and incision (CLI)/water: CLI group 2, n = 7; CLI/Dexa: dexamethasone group, n = 7], median plasma levels of IL-6 (a), TNF-α (b), MIP-2α (c) and IL-18 (d) as detected by enzyme-linked immunosorbent assay at 390 min after laparotomy/CLI are indicated. Data are shown as box plots (median, 25% and 75% quartile with whiskers denoting maximal and minimal data points) and expressed as pg/ml (b,d) or ng/ml (a,c); ***P < 0·005 compared to CLI/water; n.s.: not significant compared to CLI/water.

Download figure to PowerPoint

Next, we determined effects of dexamethasone 21-phosphate on IL-22 production by septic rats in these same samples. Here, we report for the first time that application of the glucocorticoid at a dosage that markedly weakens production of prototypic inflammatory cytokines resulted similarly in a striking significant reduction of plasma IL-22 levels (P < 0·005, Fig. 4).

image

Figure 4. Inhibition of interleukin (IL)-22 by application of dexamethasone 21-phosphate disodium salt in vivo. For each group [caecal ligation and incision (CLI)/water: CLI group 2, n = 7; CLI/Dexa: dexamethasone group, n = 7], the median IL-22 plasma levels as detected by enzyme-linked immunosorbent assay at 390 min after laparotomy/CLI are indicated. Data are shown as box plots (median, 25% and 75% quartile with whiskers denoting maximal and minimal data points) and expressed as pg/ml; ***P < 0·005 compared to CLI/water.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

In this study, we demonstrate that the prototypic glucocorticoid dexamethasone is able to suppress significantly IL-22 production mediated by fulminant activation of innate immunity. Inhibition was evident in the cell culture model of freshly isolated human PBMC under the influence of heat-inactivated S. epidermidis. Those results coincided with similar reduction of inflammatory IL-8, but notably deviated from effects on anti-inflammatory IL-10. For analysis of IL-22 driven by bacterial infection in vivo, effects of dexamethasone 21-phosphate were evaluated in experimental rat sepsis. Under the influence of the drug, robust reduction of IL-22 plasma levels became apparent which was comparable to effects on prototypic inflammatory parameters such as TNF-α, IL-6 and MIP-2α/CXCL2. Our current data on human PBMC and rat sepsis that aim at acute bacterial infections with strong activation of innate immunity contrast with data on isolated in vitro polarized murine Th17 cells, which actually displayed dexamethasone insensitivity in culture [32]. Those disparate observations apparently reflect diverse molecular mechanisms directing IL-22 production in the context of innate and adaptive immunity.

Inadequate mucosal host defence, potentially as a result from defective IL-22 production, certainly promotes bacterial epithelial damage, systemic inflammation and sepsis [33]. Tissue protective properties of IL-22 have been observed in murine hepatitis models [9,10] and may relate to anti-apoptotic functions of IL-22-induced STAT3 activation [9]. Thus, IL-22 may not only increase mucosal host defence by up-regulating, e.g. anti-microbial proteins [15] or iNOS [18], but beyond that apparently has the potential to counteract infectious complications by preserving epithelial vitality upon an infectious challenge. In this context it is also noteworthy that IL-22 has been implicated in intestinal restitution and wound healing. Specifically, IL-22 mediates proliferation and migration of intestinal epithelial cells. Moreover, in vitro wound-healing assays using cultured intestinal epithelial cells suggest clearly that IL-22 has the capacity to promote repair processes [14].

The data presented herein suggest that steroids may impair crucial aspects of tissue protection provided by IL-22. Those aspects are probably of significance under conditions of epithelial immunoactivation coupled with enhanced IL-22 expression. Such unwanted consequences of high-dose glucocorticoid therapy may relate to reports on higher incidence of postoperative infectious complications in Crohn's disease patients that had undergone this treatment regimen [23,24].

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

This work was supported by a grant from the Deutsche Forschungsgemeinschaft DFG (GK1172 ‘Biologicals’) to H. M.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References