Pentoxifylline inhibits the synthesis and IFN-γ-inducing activity of IL-18

Abstract

The effect of phosphodiesterase-inhibiting anti-inflammatory drug pentoxifylline (PTX) on LPS-induced IL-18 synthesis and IL-18-mediated IFN-γ-induction were investigated. In a dose-dependent manner PTX inhibited production of IL-18 in LPS-treated cultures of murine spleen cells and bone marrow-derived macrophages. Similarly, PTX treatment significantly reduced blood IL-18 levels and expression of spleen IL-18 mRNA in LPS-challenged mice. The inhibitory effect of PTX was specific for IL-18, since LPS-induced IL-12 p40 release was not suppressed either in splenocyte cultures or blood of LPS-injected animals. Synergistic induction of IFN-γ by combined IL-12/IL-18 treatment was also inhibited by PTX in vitro and in vivo. Experiments with IL-12 pretreatment of splenocytes, followed by IL-18 stimulation, revealed that PTX suppressed both IL-12 and IL-18 signals responsible for IFN-γ induction. These results suggest that interference with IL-18 synthesis and IFN-γ-inducing activity might contribute to anti-inflammatory actions of PTX.

INTRODUCTION

Interleukin-18 (IL-18) is a recently characterized member of the IL-1 family of cytokines, sharing with IL-1β a 15% homology at the amino acid level [1], as well as requirement for caspase 1 (IL-1β-converting enzyme — ICE) for production and release of mature protein [2]. IL-18 was originally described as an IFN-γ inducing factor (IGIF) in mice primed by P. acnes and challenged by lipopolysaccharide (LPS) [3]. Acting in synergy with another LPS-inducible macrophage cytokine, IL-12, IL-18 activates IFN-γ synthesis in T, B and natural killer (NK) cells [4–6]. Apart from important roles in the development of NK and Th1 cell activity [7] and host defence from infection and tumours [8–11], IL-18 is also involved in various pathological processes, including autoimmune diseases [12,13]. Recently, a role for IL-18 was proposed in the pathogenesis of endotoxin shock, since both anti-IL-18-treated and IL-18-deficient mice were less sensitive to LPS injection [14,15]. Most of the symptoms of LPS-induced shock — weight loss, diarrhoea, haemorrhagic colitis, splenomegaly, fatty liver and atrophic thymus — are also found after concomitant administration of IL-18 and IL-12 [16]. In accordance with the crucial role of IFN-γ in LPS toxicity [17–19], the deleterious effect of an IL-18/IL-12 combination was partly dependent on IFN-γ [16]. Therefore, interference with IL-18 and/or IL-12 synthesis and their IFN-γ-inducing activity could be a useful strategy in the therapy of sepsis.

Pentoxifylline (PTX) is a methylxantine derivative with phosphodiesterase-inhibiting activity, used widely in the therapy of microvascular disorders. In recent years, a broad range of anti-inflammatory properties of PTX became apparent, including protection in animal models of endotoxin shock [20,21], as well as beneficial effects in septic shock patients [22]. PTX-mediated alleviation of symptoms in sepsis was accompanied by reduced release of LPS-induced pro-inflammatory mediators TNF-α, nitric oxide (NO) and IFN-γ [20–24]. While the inhibitory effects of PTX on TNF-α and NO secretion are the consequence of direct drug interference with intracellular events involved in the induction of these mediators in monocytes and macrophages [25,26], the mechanisms behind PTX inhibition of LPS-triggered IFN-γ synthesis have not been fully elucidated thus far.

In the present study, we investigated the influence of PTX on production and IFN-γ-inducing activity of IL-18 in murine splenocyte cultures and in vivo. Our data indicate that inhibition of both IL-18 synthesis and action might be responsible for beneficial PTX-mediated suppression of LPS-induced IFN-γ release in sepsis.

MATERIALS AND METHODS

Animals

CBA female 6–8-week-old mice, obtained from the animal colony maintained at the Institute for Biological Research, Belgrade, Yugoslavia, were used in the experiments.

Cell cultivation

For splenocyte suspension, spleens were aseptically removed, single-cell suspension was prepared using RPMI-1640 (Flow Labs, Irvine, UK) supplemented with 10% fetal calf serum (FCS), glutamine, penicillin and streptomycin (complete medium) and red blood cells were depleted by lysis in ammonium chloride. Cells (5 × 10⁶/ml) were seeded in 24-well plates (Flow Labs) in 1 ml of complete medium, with Escherichia coli LPS (Sigma, St Louis, MO, USA), in the presence or absence of pentoxifylline (PTX; Panfarma, Belgrade, Yugoslavia) or antimurine IL-18 antibody (Pharmingen, San Diego, CA, USA). Alternatively, cells were treated with IL-12, IL-18 (both kindly provided by Damo Xu, Department of Immunology, University of Glasgow, Glasgow, UK) or their combination, with or without PTX. After incubation for 48 h at 37°C in a humidified atmosphere with 5% CO₂, cell culture supernatants were collected for ELISA. To prepare bone marrow-derived macrophages, mouse femurs were flushed with complete medium and bone marrow cells were plated out in 10% L929 conditioned medium as a source of CSF-1. Cells were grown in 10-cm bacteriological plastic plates for 7 days in a 37°C incubator containing 5% CO₂. For the assessment of PTX effect on IL-18 production, macrophages were seeded in 24-well plates at 5 × 10⁵ cells per well in 1 ml complete medium containing LPS, in the presence or absence of PTX. In some experiments, macrophages were pretreated for 24 h with LPS, washed and incubated with protein synthesis inhibitor cycloheximide (CHX), or transcription inhibitor actinomycin D (Act D) (both from Sigma), in the presence or absence of PTX. After 24 h of cultivation, cells were lysed with 0·5% Triton X-100 (Sigma), centrifuged and supernatants were collected for ELISA.

In vivo LPS and IL-12 treatment

Animals were injected i.p. with drug vehicle (PBS) or 200 mg/kg PTX. Thirty minutes later, mice were challenged with 5 mg/kg of LPS administered i.p. and bled at various time-points after LPS treatment. Serum samples were pooled (n = 4 for each group) and stored at −20°C until cytokine estimation. For RNA isolation, spleens and livers were weighed, pooled (n = 4) and homogenized in RNA Isolator (Genosys, Woodlands, TX, USA). For IL-12-mediated induction of IFN-γ in vivo, mice were injected i.p. with 10 µg/kg of recombinant murine IL-12 (Pharmingen) or IL-12 and PTX (200 mg/kg) daily for 4 days. Four hours after the fourth injection blood was collected, centrifuged and sera pooled for cytokine determination by ELISA.

ELISA

Determination of cytokine levels in cell culture supernatants, cell lysates and serum was performed by ELISA using paired antibodies according to the manufacturer's instructions (IL-12 p40 and IFN-γ – Pharmingen; IL-18 – R&D Systems, Minneapolis, MN, USA). IL-18 antibodies were kindly provided by Damo Xu (Department of Immunology, University of Glasgow, Glasgow, UK). The assay sensitivity limit was < 20 pg/ml.

RT-PCR determination of IL-18 mRNA

Total RNA was isolated by RNA Isolator according to the manufacturer's instructions. RNA was reverse transcribed using Moloney leukaemia virus reverse transcriptase (Sigma) and random primers (Pharmacia, Uppsala, Sweden). PCR amplification of cDNA with primers specific for IL-18 and β-actin was carried out in the same tube in a Thermojet (Eurogentec, Seraing, Belgium) thermal cycler as follows: 30 s of denaturation at 95°C, 30 s of annealing at 53°C and 30 s of extension at 72°C. For the IL-18, primers were: sense, 5′-ACTGTACAACCGCAGTAATACGG-3′; antisense, 5′-AGTGAACATTAC AGATTTATCCC-3′, and the PCR product was 423 bp. For the β-actin, primers were: sense, 5′-TACTCCTGCTTGCTGAT-3′; antisense, 5′-TATTGGCAACGAGCGG-3′, and the PCR product was 355 bp. After 20 cycles of amplification samples were loaded on 1·2% agarose gel and the fragments were separated by electrophoresis. After the electrophoresis, gel was blotted on the positively charged Nylon membrane (Boehringer Mannheim, Mannheim, Germany). Prehybridization and hybridization with DIG-labelled IL-18 specific probe was performed according to Engler-Blum modification of the original hybridization and detection protocol [27]. After detection of IL-18 band, membrane was striped and rehybridized with DIG-labelled β-actin specific probe. Colorimetric detection of both IL-18 and β-actin signals was performed using DIG Nucleic Acid Detection Kit (Boehringer Mannheim). The membrane was scanned and the results were analysed by NIH Image 1·61 PPC software. Relative expression of IL-18 mRNA was calculated as a ratio of the densities of the IL-18 and β-actin bands.

Statistical analysis

Data from the representative of 2–5 independent experiments with similar results are presented. To analyse the significance of the differences between various treatments performed in triplicate we used the analysis of variance (anova), followed by Student–Newman–Keuls test. A P-value less than 0·05 was considered significant.

RESULTS

PTX inhibition of LPS-induced IFN-γ synthesis in vitro is accompanied by suppression of IL-18 production

While unstimulated splenocytes did not produce measurable levels of IFN-γ (data not shown), a significant amount of this cytokine was detected in the supernatants of splenocyte cultures upon LPS treatment (Fig. 1). Recently, it was reported that LPS-mediated induction of IFN-γ in mononuclear cells depends on IL-18 secretion [2]. Indeed, neutralization of IL-18 activity with specific antibody confirmed that IL-18 was involved in IFN-γ induction by LPS in our experimental system (Fig. 1a). Addition of PTX markedly inhibited LPS-induced IFN-γ release in a concentration-dependent manner (Fig. 1b). The suppression of IFN-γ release by PTX was accompanied by a significant reduction of IL-18 levels in LPS-stimulated splenocyte cultures (Fig. 1b). Cell viability was not affected by PTX treatment, as judged by trypan blue exclusion (not shown). The inhibitory effect of PTX was specific for IL-18, since IL-12 p40 (Fig. 1b) and IL-6 production (not shown) were slightly enhanced upon PTX treatment. Thus, it seems that inhibition of IL-18 production might be partially responsible for PTX down-regulation of LPS-induced IFN-γ synthesis.

Fig. 1.

PTX inhibition of LPS-induced IL-18-dependent IFN-γ production is accompanied by decrease of IL-18 levels. (a) Spleen cells (5 × 10⁶/ml) were incubated with LPS (5 µg/ml), in the presence or absence of 10 µg/ml αIL-18 or isotype-matched irrelevant antibody as a control. (b) LPS-stimulated splenocytes were incubated with different concentrations of PTX. After 48 h, cytokine concentration in the cell culture supernatants was analysed by ELISA. Results are presented as mean ±s.d. from triplicate observations (*P < 0·05).

PTX suppresses LPS-induced IL-18 synthesis in macrophages

To investigate whether PTX inhibits intracellular accumulation of IL-18, or perhaps specifically interferes with its release to the extracellular compartment, we took advantage of an interesting feature observed in bone marrow-derived macrophages: while LPS strongly enhanced IL-18 synthesis in these cells, as assessed by measuring IL-18 concentration in cell lysates (Fig. 2a), no IL-18 release was detected upon LPS or LPS/IFN-γ treatment (not shown). PTX inhibited this LPS-induced intracellular accumulation of IL-18 in a dose-dependent manner (Fig. 2a), while IL-18 was still undetectable in the supernatants of macrophage cultures. Treatment with PTX did not affect the constitutive expression of IL-18 in macrophages (Fig. 2a). Similar results were also observed in thioglicolate-elicited peritoneal cells (not shown). Viability of macrophages was not affected by PTX, as determined by trypan blue staining (not shown). Similarly to results observed in splenocyte cultures, IL-6 production was potentiated by PTX (not shown), indicating that macrophage function was not inhibited in a non-specific fashion. When macrophages were pretreated with LPS to induce IL-18, and further IL-18 accumulation was blocked by CHX or ActD at the translational or transcriptional level, respectively, PTX failed to affect intracellular IL-18 levels (Fig. 2b). Therefore, it appears that PTX inhibits IL-18 synthesis, rather than its release, probably by interfering with IL-18 transcription.

Fig. 2.

PTX inhibits LPS-induced IL-18 synthesis in macrophages. (a) Bone marrow-derived macrophages (5 × 10⁵/ml) were incubated alone or with LPS (5 µg/ml), in the presence or absence of PTX. Alternatively, macrophages were pretreated with LPS for 24 h, washed and then incubated in fresh medium containing 2 µg/ml CHX (b) or Act D (c), with or without 200 µg/ml PTX. Concentration of IL-18 in cell lysates was determined after 24 h and results were presented as mean ±s.d. from triplicate observations (*P < 0·05 refers to treatment with LPS alone). ○, Medium; ▪, LPS.

PTX inhibits IL-18 production and mRNA expression in LPS-challenged mice

Next, we tested the ability of PTX to influence IL-18 production in vivo after intraperitoneal injection of LPS. PTX completely prevented elevation of IL-18 blood levels caused by LPS injection (Fig. 3). In contrast, LPS-augmented IL-12 p40 production was not affected by PTX treatment (Fig. 3). As expected, PTX-mediated down-regulation of IL-18 release coincided with a significant decline of IFN-γ concentration in the blood of LPS-injected mice (Fig. 3). The analysis of IL-18 mRNA showed elevation of its expression in both spleen and liver of LPS-challenged mice (Fig. 4). PTX reduced expression of IL-18 message in the spleen (Fig. 4a), but not in the liver of LPS-injected animals (Fig. 4b), indicating tissue-specific drug interference with LPS-induced accumulation of IL-18 mRNA.

Fig. 3.

PTX reduces blood IL-18 levels in LPS-treated mice. Mice were injected i.p. with 200 mg/kg PTX or PBS (control), and then 30 min later LPS (5 mg/kg) was administered by i.p. injection. At indicated time-points (for IL-18 and IL-12), or after 8 h (for IFN-γ), mice were bled and sera pooled (n = 4) for ELISA detection of cytokines. Results are presented as mean ±s.d. from triplicate observations (*P < 0·05 refers to control treatment). ▪, Control; ○, PTX.

Fig. 4.

PTX inhibits IL-18 mRNA expression in the spleen of LPS-challenged mice. Mice were injected i.p. with 200 mg/kg PTX or PBS, and then 30 min later LPS (5 mg/kg) was administered by i.p. injection. Control animals received PBS only. After 3 h, mice were sacrificed and the expression of IL-18 mRNA in the pooled (n = 4) spleen (a) and liver (b) samples was determined by Southern blot. Membranes were scanned and IL-18/β-actin signal intensity ratio was calculated. Results were presented as relative expression of IL-18 mRNA, compared to the control.

PTX reduces IL-12 and IL-18- induced IFN-γ production in vitro and in vivo

Our data suggested that inhibition of IL-18 synthesis could be responsible for PTX suppression of LPS-induced IFN-γ generation. However, PTX-mediated decrease of IL-18 production was considerably less pronounced, compared to reduction of IFN-γ levels (Fig. 1). Furthermore, LPS-mediated IFN-γ production in PTX-treated splenocytes was only partially restored in the presence of exogenous IL-18 (not shown), suggesting the existence of additional mechanisms for PTX suppression of IFN-γ release. Since LPS-stimulated secretion of another IFN-γ-inducing cytokine, IL-12, was not reduced by PTX, we investigated the possibility that PTX might interfere with signalling pathways involved in IFN-γ induction by IL-12 and IL-18. Indeed, PTX abolished IL-12 and IL-18-mediated production of IFN-γ in splenocyte cultures (Fig. 5a). Similar to the results obtained in vitro, PTX administration significantly reduced concentration of IFN-γ in the blood of IL-12-injected mice (Fig. 5b). Since in vivo induction of IFN-γ by IL-12 is dependent on endogenous IL-18 [28], we concluded that PTX inhibits IL-12 and IL-18-triggered IFN-γ synthesis both in vitro and in vivo.

Fig. 5.

PTX blocks IL-12/IL-18-induced IFN-γ synthesis in vitro and in vivo. (a) Spleen cells (5 × 10⁶/ml) were incubated for 48 h with 20 ng/ml IL-12 and 50 ng/ml IL-18, in the absence (control), or presence of 100 µg/ml PTX. After 48 h of cultivation, IFN-γ levels were assessed by ELISA. (b) Mice were challenged for 4 consecutive days with daily intraperitoneal injections of IL-12 (10 µg/kg), in combination with PTX (200 mg/kg) or PBS (control). At day 4, blood was collected and sera pooled (n = 4) for IFN-γ detection. Results (a, b) are presented as mean ±s.d. from triplicate observations (*P < 0·05).

PTX blocks both IL-12 and IL-18 signalling responsible for IFN-γ induction

Finally, we wanted to examine whether PTX blocks IL-12 or IL-18-triggered signalling pathway responsible for IFN-γ production. This could not be inferred from single IL-12 or IL-18 stimulation in vitro, since IL-18 alone did not induce IFN-γ production, and IL-12-mediated IFN-γ release, although inhibited by PTX (not shown), is exerted in synergy with endogenous IL-18 [28]. Interestingly, it appears that synergistic cooperation of IL-12 and IL-18 is mediated primarily at the level of the IL-18 receptor, the expression of which is strongly induced by IL-12 [29,30]. Accordingly, cells pretreated with IL-12 and extensively washed, produced high amounts of IFN-γ in response to IL-18 (Fig. 6). In these conditions, IFN-γ production was completely mediated by IL-18 signalling machinery, since IL-12 was undetectable during IL-18 stimulation (not shown). Administration of PTX either in IL-12 pretreatment, or subsequent IL-18 stimulation, significantly reduced IFN-γ release in splenocyte cultures (Fig. 6), indicating that PTX can interfere with both IL-12 and IL-18-triggered signals that lead to IL-18R or IFN-γ expression, respectively.

Fig. 6.

PTX interferes with both IL-18 and IL-12 signalling. Spleen cells (5 × 10⁶/ml) were incubated with 20 ng/ml IL-12 for 24 h, washed and stimulated for IFN-γ production with 50 ng/ml IL-18 (control). PTX (100 µg/ml) was added either during IL-12 pretreatment (IL-12/PTX) or IL-18 stimulation (IL-18/PTX). Production of IFN-γ in culture supernatants was determined 48 h after IL-18 addition and results were presented as mean ±s.d. from triplicate observations (*P < 0·05 refers to control).

DISCUSSION

In the present report, we have shown that PTX inhibits synthesis, as well as IFN-γ-inducing activity of IL-18, both in vitro and in vivo. This property of PTX might account for its known ability to suppress detrimental IFN-γ secretion in endotoxin shock.

In contrast to non-fractionated splenocytes ([28], and the present study) or macrophages from P. acnes-infected mice [3], peritoneal or bone marrow-derived macrophages from healthy animals accumulated, but did not release IL-18 upon LPS stimulation in our study. Similar results were obtained after LPS stimulation of human monocyte line THP.1 [31], and murine macrophage lines RAW264 and J774 (our unpublished observation). Therefore, it seems that some additional signals, triggered by infectious agents or contact with other cells, are required for macrophage IL-18 release. We used this macrophage inability for IL-18 secretion to show that PTX interferes with LPS-induced IL-18 synthesis, rather than with its release. Since the ELISA system we used had a greater affinity for ICE-cleaved, bioactive IL-18 than for its immature form [28], there was a possibility that the putative inhibition of IL-18 synthesis was actually an artefact resulting from PTX-mediated ICE inhibition and a shift of pro/mature IL-18 balance towards the inactive form. However, the inability of PTX to affect intracellular IL-18 levels when protein synthesis or transcription was blocked argues strongly against PTX suppression of ICE activity. It seems more likely that drug interferes with the transcription of the IL-18 gene and reduces the intracellular pro-IL-18 pool from which the bioactive form is generated. Reduced expression of LPS-induced IL-18 mRNA in the spleens of PTX-treated mice supports trancriptional regulation of IL-18 expression by PTX. Interestingly, PTX modulated IL-18 mRNA in a tissue-specific manner, not influencing its expression in the liver. Since IL-18 is one of the mediators of endotoxin-induced liver pathology [32], a beneficial effect of PTX in this model of liver damage [33] is probably mediated through mechanisms unrelated to inhibition of IL-18 synthesis.

Our data suggest that PTX-mediated down-regulation of LPS-induced IFN-γ release in vitro and in vivo could be partly dependent on inhibition of IL-18 production. However, another macrophage cytokine, IL-12, can synergize with IL-18 for induction of IFN-γ in LPS-cultured mononuclear cells and endotoxin shock [34,35]. Interestingly, LPS-induced IL-12 p40 synthesis in our experiments was unaffected, or even slightly enhanced by PTX in vivo and in vitro, respectively. Similarly, Marcinkiewicz et al. recently reported that PTX can potentiate in vitro production of IL-12 p40 subunit in murine macrophages [36]. In the same study, PTX inhibited the expression of p35 chain of IL-12 [36], as described previously by Moller et al. in human blood mononuclear cells [37]. Accordingly, the production of functional IL-12 heterodimer p70 was blocked by PTX in both human monocytes and mouse macrophages in vitro, as well as in the blood of SEB-challenged mice [37,38]. Thus, it is quite possible that PTX, in addition to decreasing IL-18 release, could prevent IFN-γ induction in endotoxic mice partly through selective impairment of p35 subunit synthesis and a subsequent reduction of functional IL-12 heterodimer assembly.

The mechanisms responsible for the induction of IFN-γ in T cells stimulated by TCR engagement or IL-12/IL-18 combination seem to be different, since TCR-derived signals down-regulate IFN-γ production induced by the latter stimulus [30]. While PTX is a well-known inhibitor of IFN-γ release in TCR-activated T cells [38–40], the present study shows for the first time its ability to suppress TCR-independent IFN-γ induction mediated by IL-18 in IL-12-pretreated splenocytes. Interestingly, cells pretreated with IL-12 in the presence of PTX were less responsive to subsequent IL-18 stimulation, indicating that PTX might also affect IL-12 induction of the IL-18 receptor. While the putative attempt to suppress IL-18 release in endotoxin shock would require early treatment, the inhibition of IL-12/IL-18-mediated IFN-γ release by PTX could presumably widen the time-window during which drug administration might yield a protective effect. Similarly, in spite of the inability to inhibit transcription of liver IL-18, PTX could exert protection from LPS-induced liver destruction partly through inactivation of already synthesized IL-18. Although PTX influence on IL-18 synthesis induced by non-infectious stimuli still remains to be investigated, drug ability to affect IL-12/IL-18 action broadens the range of pathological conditions in which its administration could be beneficial, including Th1-mediated autoimmune diseases such as multiple sclerosis and its animal model, experimental allergic encephalomyelitis (EAE). Recently, the autoimmune destruction of CNS in EAE was found to be dependent on both IL-12 and IL-18 action [13,41]. Although PTX-mediated suppression of EAE development was attributed mainly to inhibition of T cell activation and TNF-α production [42], our data indicate that interference with IL-12 and IL-18 activity might present an additional mechanism responsible for protective PTX action in EAE.

In conclusion, the present study suggests that inhibition of IL-18 synthesis and activity could contribute to the anti-inflammatory properties of PTX. Recently, the elevation of IL-18 in human sepsis and Crohn's disease has been reported [43,44]. If confirmed in humans, an IL-18-blocking effect of PTX might be relevant for therapy of septic shock and other IL-18-mediated inflammatory diseases. However, further studies are needed to determine PTX cellular targets and to shed some light on intracellular mechanisms responsible for these PTX actions, as well as to explore its effects on other IL-18-induced pro-inflammatory events, such as promotion of TNF-α synthesis and expression of adhesion molecules [45,46]. This might be important in view of the findings that IL-18/IL-12 treatment induces adverse effects through not only IFN-γ-dependent, but also IFN-γ-unrelated mechanisms [16].