dermatitis
Ba¨umer W, Krekeler S, DeVries VC, Niedorf F, Tschernig T, Kietzmann M. Non-steroidal and steroidal anti-inflammatory drugs vary in their modulation of dendritic cell function in the elicitation phase of allergic contact dermatitis.
Exp Dermatol 2006: 15: 322–329. #2006 The Authors. Journal compilation#2006 Blackwell Munksgaard
Abstract: The role of dendritic cells (DCs) in allergic contact der-matitis has been clearly demonstrated for the induction phase. However, the situation during the elicitation phase is very complex within a distinct inflammatory response. This study was performed to exploit DC migration in the elicitation phase in a mouse model of allergic contact dermatitis and to evaluate the effects of steroidal and non-steroidal anti-inflammatory drugs (NSAIDs) on DC migration through skin in the elicitation phase of allergic contact dermatitis. Topically and systemically administered acetylsalicylic acid (ASA) did not reduce the inflammatory response. However, systemically administered ASA sig-nificantly reduced the DC migration to the draining lymph node. In contrast, topically administered indomethacin reduced the inflamma-tory response, but had only minor effects on DC migration, whereas diflorasone diacetate reduced both inflammatory reaction and DC migration. Thus, NSAIDs may differ in their inhibitory action in immunological inflammation.
W. Ba¨umer1, S. Krekeler1, V. C. DeVries2, F. Niedorf1, T. Tschernig2and M. Kietzmann1
1Department of Pharmacology, Toxicology, and Pharmacy, University of Veterinary Medicine Hannover, Foundation, Hannover, Germany
2Department of Functional and Applied Anatomy, Hannover Medical School, Hannover, Germany Key words: acetylsalicylic acid – allergic contact dermatitis – dendritic cells – diflorasone diacetate – indomethacin
Wolfgang Ba¨umer
Department of Pharmacology, Toxicology and Pharmacy
University of Veterinary Medicine Hannover, Foundation Buenteweg 17
30559 Hannover, Germany Tel.:þ49 511 9538732 Fax:þ49 511 9538581
e-mail: wolfgang.baeumer@tiho-hannover.de Accepted for publication 17 January 2006
Introduction
Non-steroidal anti-inflammatory drugs (NSAIDs) differ in their potential to modulate dermal inflammatory reactions mediated by immune and non-immune reactions (1–3). It was recently demonstrated that the immunosup-pressive agents tacrolimus and rapamycin and the phosphodiesterase 4-inhibitor cilomilast modulate the dendritic cell (DC) function in a murine model of allergic dermatitis (4). It was therefore of interest to determine whether classi-cal anti-inflammatory agents like steroidal and
NSAIDs could also modulate DC function in allergic dermatitis. As the results concerning the efficacy of NSAIDs in models of allergic contact dermatitis are contradictory (1, 2), a re-evalua-tion seemed reasonable. It was of particular interest to test acetylsalicylic acid (ASA), as that compound has been shown to have further anti-inflammatory properties in addition to its inhibition of cyclooxygenase-mediated reactions.
In high concentrations, ASA has been shown to inhibit nuclear factorkB (NFkB) (5,6). ASA also affects murine DC maturation and antigen pre-sentation, as is demonstrated by the impaired allergic reaction elicited by hapten-pulsed DCs incubated with high concentrations of ASA
Abbreviations:ASA, acetylsalicylic acid; BM, bone marrow; DCs,
den-Experimental Dermatology 2006: 15: 322–329 Copyright#2006 The Authors. Journal compilation#2006 Blackwell Munksgaard Blackwell Munksgaard . Printed in Singapore
EXPERIMENTALDERMATOLOGY ISSN 0906-6705
activation (7) and has also been reported for human monocyte-derived DC (8). Furthermore, it is well known that NSAIDs have an effect on PGE2 synthesis. Because DC migration seems to depend on prostaglandin (PG) E2-EP4 signaling, as has been demonstrated for the epidermal DCs [Langerhans cells (LC)] (9), this sequence of events could be a way to prevent LC from pre-senting the antigen in the draining lymph nodes.
It was thus hypothesised that ASA modulates DC function in vivo in a well-characterised model of allergic contact dermatitis provoked by toluene diisocyanate (10,11). The inhibition of PGE2synthesis would modulate the migration and the inhibition of NFkB, could further affect DC function like T cell stimulation, as has been demonstratedin vitro (7). The focus of this study was DCs, as both LC and dermal DCs transport haptens like toluene-2,4-diisocyanate (TDI) from the skin through afferent lymphatic vessels to draining lymph nodes in the induction phase and present these haptens to hapten-specific T cells in the elicitation phase. The DCs are there-fore the major antigen-presenting cells in both the induction and the elicitation phase of allergic con-tact dermatitis. The effects of topically and sys-temically administered ASA were compared with that/those of indomethacin, which has shown some efficacy in models of allergic contact dermatitis (1), and of diflorasone diacetate, a strong topical corti-costeroid which displayed a strong inhibitory effect in the allergy model used here (11).
Materials and methods Mice
Female BALB/c mice were obtained at 8 weeks of age (20 g body weight) from Charles River (Sulzfeld, Germany). The mice were housed in groups of six per cage at 22C with a 12-h light/dark cycle. Water and a standard diet (Altromin, Lage/Lippe, Germany) were available ad libitum. The animal experiment was approved by the regional government of Hannover, Germany (Reference no. 509.6-42502-03/711).
DC generation from bone marrow cultures
Bone marrow (BM)-derived DCs were generated according to the protocol of Lutz et al. (12) with slight modifications as previously described (13). Briefly, BM was cultured with RPMI 1640 (Biochrom, Berlin, Germany), 10% fetal calf serum (Biochrom) and 50mmol/l 2-mercaptoethanol (Sigma, Deisenhofen, Germany). The medium contained 20 ng/ml gran-ulocyte macrophage colony-stimulating factor (R&D Systems, Wiesbaden, Germany). Fresh medium supplemented with gran-ulocyte macrophage colony-stimulating factor was added on days 3, 6 and 8. FACS analysis of the day-10 cell suspension demonstrated a high yield of CD 11c and major histocompat-ibility complex class II (MHC-II)-positive cells (13).
Incubation with ASA, indomethacin and diflorasone diacetate
After 9 days of cultivation, the BM-derived DCs were placed in a 12-well plate (4105cells/well) and incubated for 24 h with 0.5, 1, 2.5 and 5 mmol/l ASA (Sigma, Deisenhofen, Germany), 0.1, 0.5 and 5mmol/l indomethacin (Sigma) and 2, 20 and 200 nmol/l diflorasone diacetate (Basotherm, Biberach an der Riss, Germany). These concentrations did not reduce cell via-bility as determined by CellTiter#AQueousOne Solution Cell Proliferation Assay (Promega, Mannheim, Germany, data not shown). The higher concentrations of ASA were acidic and had to be buffered to pH 7.4 with HEPES (Sigma).
At day 10, the cells were placed in a 96-well plate (5104 cells/well) and stimulated by addition of lipopolysaccharide (Escherichia coli, O127:B8, Sigma) working solution (1mg/ml lipopolysaccharide, final concentration). The lipopolysacchar-ide (LPS)-positive control was treated with the vehicle (ethanol) in the highest final concentration used here. The supernatant was collected 24 h later and stored at20C until determination of tumor necrosis factor-a(TNFa) and PGE2by enzyme-linked immunosorbent assay (ELISA, R&D Systems). Three indepen-dent assays were performed.
Mouse ear swelling test and LLNA
The mouse ear swelling test was performed as previously described (4,11), and the local lymph node assay (LLNA) was performed with non-radioactive endpoints as recently described (14). After the mice had settled in for 1 week, the skin of their abdomens was shaved and depilated with Veet1. The abdom-inal skin was then stripped with adhesive tape. For active sensitisation, 100ml of 5% TDI in acetone were administered to the abdominal epidermis. This was repeated with 50ml of TDI for 3 consecutive days.
The allergic reaction was elicited 21 days later by administra-tion of 20ml of 0.5% TDI in acetone on both the inner and the outer surfaces of the left ears to determine the sensitisation status. The amount of swelling was determined 24 h after chal-lenge by comparing the thickness of the ear at that time with the individual basal value (c. 240mm) determined earlier. A cuti-meter (Mitutoyo, Neuss, Germany) was used for these measure-ments. Animals with a mean swelling difference less than 20%
of their basal value 24 h after challenge were then excluded as being not sensitised; this applied to less than 5% of all mice in this study. Equal numbers of the mice included in the study were distributed to the treatment groups (n¼6) according to the degree of swelling, so that each group contained animals responding to varying extents. The mice were allowed to rest until 7 days after the ear thickness returned to almost the normal level. To exclude contamination by residues of the allergen on the ears, the untreated right ears were used for the main experiment as follows.
Drug administration
For systemic administration, one group of six mice received 150 mg/kg ASA orally in 0.5% methylcellulose 24, 9 and 1 h before TDI challenge.
For topical administration, one group received 8 mg ASA in 20ml acetone/DMSO (9:1) applied topically onto the right ear 9 and 1 h before challenge; a third group was treated topically with 1 mg indomethacin in 20ml acetone/DMSO (9:1); the fourth, with 20ml diflorasone diacetate 0.05% (in acetone/
DMSO 9:1) [11], also 9 and 1 h before challenge. The fifth group received the vehicle (20ml acetone/DMSO 9:1). One group was not sensitized and challenged and served as basal
Non-steroidal and steroidal anti-inflammatory drugs
control for the ear swelling test and the LLNA. The dosages chosen for the topically administered indomethacin and sys-temically administered ASA have already been shown to be effective in arachidonic acid and/or experiments in which dermal inflammation was induced by tetradecanoylphorbol acetate (TPA) (3,15). The dosage for the topical administration of ASA was limited by solubility. Diflorasone diacetate has been shown in previous experiments to be effective at the indicated dose (11). The determination of ear thickness was performed as described for the boost reaction. Directly after determination of ear thickness, the mice were sacrificed, and the draining lymph nodes were excised for determination of weight and cell count.
Skin DC migration assay
The mice were sacrificed by cervical dislocation 16 h after chal-lenge and measurement of ear thickness. The cartilage-free dorsal halves of split mouse ear skin were cultured in 24-well microtiter plates based on a method of Ortner et al. (16) The halves were laid epidermal side up onto racks made of sieves (13). The racks were changed daily to new wells and new media. Migrated cells from each ear (day 2 and day 3) were pooled and counted with a hemocytometer (Neubauer, VWR), the ear halves were weighed, and migrated cells/mg ear tissue was calculated. Migrated cells from day 1 have been discarded, as the DC purity is very low (approximately 30%) (16).The viability of the cells was assessed by trypan blue exclusion. The quantity of trypan blue positive cells was low, and there was no increase in any treatment group.
Determination of DCs in lymph nodes
The auricular lymph nodes were excised, carefully separated from remaining fatty tissue and stored on ice in PBS/BSA 0.5%. Single cell suspensions were prepared by mechanical tissue disaggregation by means of a glass potter (VWR), counted and stained for CD 11c-positive cells without further enrichment (4). Because it has recently been reported that DCs in draining lymph nodes which have migrated from skin are characterised by both CD11chighand CD40high, we performed double staining for CD11c (hamster IgG1, PE-conjugated, HL3, Pharmingen, Hamburg, Germany) and CD40 (rat IgG2a, FITC-conjugated, 3/23, Pharmingen) by FACS analysis (FACSCalibur, Becton Dickinson, Mountain View, CA, USA).
Preparation of epidermal sheets for immunhistochemistry
Epidermal sheets were prepared and evaluated as previously described (4). In short, skin was floated on 0.5 M ammonium thiocyanate (Riedel de Hae¨n, Hannover, Germany) for 10 min at 37C. The epidermis was separated from the dermis and immediately fixed in cold acetone. The DCs were detected with biotinylated monoclonal anti-mouse MHC-II (I-A/I-E, rat IgG2b, Beckton Dickinson, Heidelberg, Germany). Labeling of the antibodies was visualized using streptavidine-fluorochrome (carbocyanin 3, Jackson Immunoresearch Laboratories, West Grove, PA, USA). Analyses were performed using the Kontron KS 400 image analysis system. The density of LCs was analysed
(at20 magnification on a calibrated grid). Sixteen randomly chosen areas were counted per ear. Six ears of each group were analysed. Due to accidental loss, only three ears were evaluated from the mice treated with indomethacin.
Zymography
Ear halves were cultured for 3 days (for the migration assay) and homogenised in liquid nitrogen. Each sample was diluted to a concentration of 500mg/ml of protein, and 25ml of each sample was subjected to electrophoresis as described earlier (4). Gels were stained with Coomassie Brilliant Blue (VWR, Darmstadt, Germany) and de-stained in methanol and acetic acid (all salts and solutions from VWR). Gelatinase activities appeared as clear bands against a blue background. Molecular weights of gelatinolytic bands were estimated using recombi-nant protein molecular weight markers (10–250 kDa; Bio-Rad, Munich, Germany). Six samples of epidermis and dermis were evaluated from each treatment group. The gelatinolytic bands were quantified by densitometry. The images were acquired out with an HP Scan Jet 7400c scanner using the PHOTOIMPACT 4.0 programme (Ulead Systems, Kaarst, Germany). The densi-tometric evaluation was made using SCION IMAGE software for PC (Scion Corporation, Frederick, MD, USA). The zymo-grams were inverted, and gray background values were sub-tracted from those of the gelatinolytic band. Each gel consisted of two vehicle controls, which were compared with two ears treated with ASA, indomethacine and diflorasone diacetate. Three gels from each treatment group were per-formed, resulting inn¼6 per group.
Determination of PGE2 in skin tissue
A part of the zymography protein sample (500mg/ml of pro-tein) was further diluted (1:10) with a specific assay buffer, and PGE2 was determined by ELISA (R&D Systems) as described above.
Penetration of ASA through murine abdominal skin
A standard diffusion experiment using Franz-type diffusion cells was performed for the estimation of the amount of topi-cally administered ASA penetrating through murine skin.
Excised murine abdominal skin was carefully shaved with a hair clipper and set onto the diffusion cell, so that 1.5 cm (diameter) of the dermal side were in contact with warmed (34C) PBS buffer. ASA was administered in analogy to thein vivo experiment. Eight milligram (dissolved in 20ml acetone/
DMSO 9:1) was administered twice, 1 and 9 h after the begin-ning of the diffusion experiment. Samples were taken before and at the indicated times (Table 1). ASA and salicylic acid (SA) were determined on an RP 18 column by HPLC with UV detection (237 nm) and without further sample treatment. The retention time for ASA and SA was 3.2 min and 4.5 min, respec-tively. The calibration was performed with ASA and SA dis-solved in PBS.
Table 1. Reabsorbed amount of acetylsalicylic acid (ASA) and its metabolite salicylic acid (SA) (mmol/cm2) after two topical administrations of 8 mg (total amount:
88mmol) ASA onto murine abdominal skin (1.5 cm diameter) in analogy to thein vivoexperiment. Samples were taken from the acceptor medium at indicated times. MeanSD. The estimated amount of absorbed ASA (for 1.5 cm diameter) is 1.1mmol 24 h after the first administration
Hours 1 2 4 6 8 12 22 24 26
Ba¨umer et al.
Statistical evaluation
Figures are presented as mean (+SEM). Statistically significant differences between the drug treatments and the controls were assessed by a one-way ANOVA followed by Dunnett’spost hoctest.
Results
In vitro, LPS induced an increased PGE2 secre-tion in murine DCs derived from BM 24 h after stimulation. At the doses tested, preincubation with ASA and indomethacin lowered the PGE2 synthesis to basal control values (Table 2).
Diflorasone diacetate reduced the PGE2secretion at 20 nmol/l and 200 nmol/l. However, there was no inhibition at 2 nmol/l.
The LPS-induced TNFa secretion was decreased dose dependently by ASA at 2.5 and 5 mmol/l, whereas indomethacin had no inhibi-tory effect at up to 5mmol/l. Diflorasone diace-tate reduced TNFa secretion dose dependently from 2 to 200 nmol/l (Table 2).
Mouse ear swelling test and LLNA
Topical administration of TDI to sensitised mice resulted in a mean increase in ear thickness of 115mm (+16) 16 h after challenge. Neither topically nor systemically administered ASA inhibited the allergic inflammatory response.
However, topically administered indomethacin significantly reduced ear swelling to 23mm (+5), and diflorasone diacetate completely blocked the swelling. Untreated mice served as a basal control (Fig. 1). Directly after determina-tion of ear thickness, the mice were sacrificed,
and the draining lymph nodes were excised for determination of weight and cell count. Although there was no effect on the ear thickness, all sub-stances significantly inhibited the increase in lymph node weight and cell count, except for the lymph node cell count in mice treated topic-ally with ASA (Fig. 2a,b). Furthermore, systemic ASA and topical diflorasone diacetate treatment reduced the amount of CD11c and CD40 double-positive DCs in the draining lymph node significantly in comparison with mice treated with vehicle (Fig. 2b). The CD11c/CD40 double staining was performed, as DCs in draining lymph nodes which have migrated from skin are characterised by both CD11chighand CD40high.
LC density
The ventral ear halves were taken after determina-tion of ear thickness, and the epidermis was stained for MHC-II-positive cells. The density of LCs (i.e.
MHC-II-positive cells) in the epidermis of murine ear skin was reduced in TDI challenged animals.
The reduction was highly significant (P¼0.004), when the TDI challenged mice were directly com-pared with untreated controls (but it failed to be significant by the one-way ANOVA). The density could be restored to values of untreated controls by all tested substances (Table 3).
Skin DC migration assay
The remaining excised dorsal ear halves were cultured for 3 days, and the emigrated cells from days 2 and 3 were pooled and counted, as
Table 2. Concentration of PGE2 and tumor necrosis factor-a (TNFa) (meanSD) in supernatants of murine dendritic cells derived from bone marrow and stimulated with lipopolysaccharide (LPS). Cells were preincu-bated with acetylsalicylic acid (ASA), indomethacin (Indo) and diflorasone diacetate (Diflo) in doses as indicated 24 h before LPS stimulation.
Supernatants were collected 24 h after LPS stimulation
Treatment PGE2 TNFa
Control 137241 9281
1P<0.05. Representative data from three independent experiments.
** **
Control Vehicle ASA top ASA sys
Indomethacin Diflorasone
Ear swelling (µm)
Figure 1. Effect of topically (top) and systemically (sys) adminis-tered acetylsalicylic acid, and of topically adminisadminis-tered indometha-cin and diflorasone diacetate on the inflammatory response (ear swelling) determined 16 h after toluene-2,4-diisocyanate challenge.
n¼6; *P<0.05; **P<0.01 compared with vehicle.
Non-steroidal and steroidal anti-inflammatory drugs
these cells represent DCs with a high purity (16).
Treatment with steroidal and non-steroidal anti-phlogistic drugs resulted in a diminished emigra-tion of skin DCs. There was a moderate, but significant reduction of migration in ear explants from mice treated topically with ASA (Table 3).
There was no distinct alteration of the morphol-ogy of DC by any drug treatment as far as it could be evaluated under the hemocytometer.
Directly after cultivation, PGE2 was measured in the skin tissue. The PGE2concentration in ear tissues correlated to the inhibition of DC
migra-ASA, whereas systemically administered migra-ASA, topically administered indomethacin, and diflor-asone diacetate inhibited the PGE2 synthesis in ear tissue almost completely (Table 3).
Zymography
A part of the skin tissue was also taken for the determination of matrix metalloproteinase-9 (MMP-9) activity after cultivation. Tissue extracts of the mice treated with vehicle showed marked white bands at approximately 80 and 90 kDa, which represent a distinct gelatinolytic band of the active and pro-MMP-9, respectively [as shown in earlier studies (4, 13)]. Neither topi-cally administered ASA nor indomethacin reduced this gelatinolytic activity significantly, whereas systemically administered ASA and topi-cally administered diflorasone diacetate almost completely diminished this activity (Table 3).
Penetration of ASA through murine skin
As topically administered ASA lead to only marginal reduction of DC migration and no inhibition of the inflammatory response, it was investigated how much of the topical administered ASA could penetrate murine skin in our experi-mental setting to determine the bioavoailability of topically administered ASA. When ASA was administered twice onto murine abdominal skin, approximately 1% of the administered ASA was found in the acceptor medium 24 h after the first administration (Table 1). A large part of the ASA was metabolised to SA within the first 24 h (the molarity ratio of SA to ASA was about 5:1).
Discussion
The role of DCs in the sensitisation phase of allergic contact dermatitis has been studied inten-sively (17–19); very recent studies deal with the particular role of LCs in the sensitisation phase (20, 21). However, not as much is known about the function of DCs (either LCs or dermal DCs) in the elicitation phase, and the available findings
The role of DCs in the sensitisation phase of allergic contact dermatitis has been studied inten-sively (17–19); very recent studies deal with the particular role of LCs in the sensitisation phase (20, 21). However, not as much is known about the function of DCs (either LCs or dermal DCs) in the elicitation phase, and the available findings