Effects of dexamethasone on TNF-alpha-induced release of cytokines from purified human blood eosinophils

Background TNF-alpha is an important mediator in allergy also for its effects on eosinophils. Methods The effect of dexamethasone on TNF-alpha induced eosinophils survival, degranulation (ECP), cytokines release (IL-8, GM-CSF) and adhesion to VCAM-1, ICAM-1 and IgG coated wells (EPO release) were evaluated. Results The drug inhibited IL-8 and GM-CSF production, but not viability, degranulation or adhesion in human peripheral blood eosinophils. Conclusion These results indicate that part of the activity of glucocorticosteroids on eosinophils may be mediated by their ability to inhibit cytokine secretion that in turn is important for the perpetuation of the allergic inflammation.


Background
Eosinophils are bone marrow-derived granulocytes that play a crucial role in allergic inflammation. TNF-α is a pro-inflammatory cytokine synthesized by many inflammatory and structural cells. We previously demonstrated that mast cell-derived TNF-α induced eosinophil survival by autocrine production of GM-CSF [1]. TNF-α is also involved in eosinophil adhesion to endothelial cells and induces eosinophil activation, degranulation, and cytokines production. Glucocorticosteroids (GCS), the main anti-inflammatory drugs in allergic diseases, have been demonstrated to decrease circulating and tissue eosi-nophils. In vitro dexamethasone can inhibit eosinophil survival [2], expression of adhesion molecules [3], and cytokines production [4]. However, the effect of GCS on TNF-α induced eosinophil activation has only been partially investigated. The present study evaluated the effect of dexamethasone on TNF-α induced eosinophil degranulation, cytokines release and adhesion to VCAM-1, ICAM-1 and IgG.

Eosinophils degranulation and cytokines release
ECP level was measured in the culture supernatants by a RIA kit (ECP, Pharmacia Upjohn, Milton Keynes, UK). GM-CSF and IL-8 content was detected in the culture supernatants by ELISA kit (R&D Systems, Abingdon, UK). Effect of dexamethsone on TNF-α-induced ECP release from eosinophils Figure 2 Effect of dexamethsone on TNF-α-induced ECP release from eosinophils. Eosinophils were cultured in medium alone (control) or with dexamethsone (medium+Dexa) or in the presence of TNF-α alone (50 ng/ml) (TNF-α) or TNF-α with dexamethsone (1 µM) (TNF-α+Dexa). ECP release was evaluated by RIA. Values are mean ± SEM (n = 3).

Effect of TNF-α on eosinophil viability in vitro
Effect of TNF-α on IL-8 release from eosinophils Figure 3 Effect of TNF-α on IL-8 release from eosinophils. Eosinophils were cultured with different concentrations of TNF-α (0 -100 ng/ml). IL-8 release was evaluated by ELISA. Values are mean ± SEM (n = 5).

Statistical analysis
Statistical analysis was performed by Student's t paired test. A p value of <0.05 was considered statistically significant.

Results
TNF-α significantly increased eosinophil viability in a concentration-dependent fashion, compared to culture in medium alone, with a maximal effect at 20 ng/ml ( Figure  1). This effect was not influenced by the addition of dexamethasone in the culture medium (13.3% vs 11.5%).
Incubation of eosinophils with TNF-α induced a significant release of ECP compared to eosinophils cultured in medium alone (24.5 ± 8.9 vs 6.9 ± 1.2 pg/10 6 ; p < 0.05). However, addition of dexamethasone in the cultures did not affect TNF-α-induced ECP release ( Figure 2). The release of IL-8 by TNF-α treated eosinophils was dosedependently proportional to the concentrations of TNF-α. A maximal release was achieved at 100 ng/ml (1770,49 ± 129 pg/ml; p < 0.05) ( Figure 3). TNF-α also induced GM-CSF release by eosinophils although to a lesser extent than that of IL-8 (data not shown). Treatment of the cultures with dexamethasone completely blocked the TNF-αinduced release of both IL-8 and GM-CSF (p < 0.05) (Figures 4A-B). TNF-α enhanced significantly the percentage of eosinophil adhesion to VCAM-1, ICAM-1 and IgG in comparison to medium alone, by 177%, 205% and 169%, respectively. However, this effect was not inhibited by dexamethasone ( Figure 5).

Discussion
We have shown that dexamethasone inhibits the release of IL-8 and GM-CSF in TNF-α activated human peripheral blood eosinophils from non-atopic volunteers.
eosinophils after different stimuli has been demonstrated [4]. However, its effect on eosinophils activated by TNF-α has not been fully investigated as yet. In our system we used dexamethasone to study the effect of GCS on TNF-αinduced eosinophil activation. It is important to note that different GCS have similar effects on inflammatory cells. Several reports have demonstrated that the inhibition of dexamethasone on eosinophil survival and activation parallels the one of inhaled GCS in vitro. For example, budesonide reduced the number of peripheral blood eosinophils by suppressing both their progenitors in the blood and colony-forming unit production in the bone marrow [6,7]. It is also known that inhalation of high doses of fluticasone reduced the number of blood eosinophils by increasing their apoptosis in vivo [8].
Although dexamethasone has been shown to induce eosinophil apoptosis, we have found that it did not decrease eosinophil survival after 18 h of treatment. Therefore, its effects on cytokine release observed in our system can not be attributed to the eosinophil death. Dexamethasone, as other GCS, inhibits cytokines release by eosinophils by interference with transcription factors such as NF-kB and AP-1 [9]. Since TNF-α is a potent inducer of NF-kB in eosinophils [10] we can hypothesize that dexamethasone inhibits GM-CSF and IL-8 release in TNF-α activated eosinophils by blocking NF-kB (genomic mechanism). In our study dexamethasone was unable to inhibit TNF-αinduced ECP release and their adhesion to immobilised VCAM-1, ICAM-1 and IgG. These data are in accordance with previous works in which dexamethasone did not affect the C5a-and IL-5 enhanced immunoglobulininduced eosinophil release of EDN [11].
Recent evidence supports a direct and extremely rapid inhibitory effect of GCS on some activated inflammatory cells (i.e. basophils) via a non-genomic effect that results from the interaction of the GCS with biological membranes [12]. However, we have not observed any effect of dexamethasone on eosinophils degranulation and adhesion (rapid events).
In conclusion, from our data we can speculate that GCS exert beneficial effects in allergic inflammation also by selectively inhibiting TNF-α-induced eosinophils release of GM-CSF and IL-8, but not their survival, degranulation and adhesion.