| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,,1
* Department of Pediatrics,
Institute of Experimental Dermatology,
Department of Dermatology, University of Münster, Germany;
Johnson & Johnson Pharmaceutical Research & Development, Spring House, Pennsylvania, USA;
|| Department of Pharmacology, University of Calgary, Canada;
¶ Departments of Surgery & Physiology, UCSF, San Francisco, California, USA;
** Department of Cell Biology, Center for Molecular Biology of Inflammation (ZMBE), University of Münster, Germany;
Department of Physiology, University of Erlangen, Germany; and
Department of Physiology, University of Calgary, Canada
1Correspondence: Department of Dermatology, University of Münster, von-Esmarch-Str. 58, 48129 Münster, Germany. E-mail: msteinho{at}uni-muenster.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: proteinase-activated receptors • G-protein-coupled receptors • immune response • knockout mouse • skin
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Acute or chronic inflammatory reactions could be induced by allergens (allergic contact dermatitis, ACD) or irritants (irritant toxic contact dermatitis, ICD) leading to eczema. Although the etiology and pathophysiology of various inflammatory dermatoses are known, several therapeutic regimens often have undesirable side effects (2)
. Thus, additional therapeutic tools for the treatment of inflammatory skin diseases are required. For example, therapeutic strategies preventing the early phase of vascular inflammatory responses such as edema, plasma extravasation, and recruitment of inflammatory mediators have become promising ways to fight excessive inflammatory reactions (3)
.
Proteinase-activated receptor-2 (PAR-2) belongs to a new subfamily of G-protein-coupled receptors with seven transmembrane domains that are activated by proteolysis. Four PARs have been cloned and characterized (4–6 and references). Since PAR-2 is highly expressed in the skin (7)
and PAR-2 agonists induce all the known characteristics of inflammation (8
, 9)
, we have hypothesized that PAR-2 mediates vascular responses in the inflamed skin. Moreover, serine proteases that activate PAR-2 have been demonstrated to play a crucial role during inflammation in various tissues, including the skin (ref 10
and references).
Recently we have shown that subdermal application of PAR-2 agonists induce acute inflammation, which is mediated at least in part by a neurogenic mechanism via release of inflammatory neuropeptides (8)
. Moreover, PAR-2 agonists induce leucocyte rolling, cell adhesion, and leucocyte extravasation in rats in vivo (9)
. However, little is known about the direct intradermal inflammatory effects of PAR-2 in vivo characteristic for inflammatory dermatoses and involvement of microvascular postcapillary venules. Since dermal microvascular endothelial cells express PAR-2 and directly contribute to inflammation after PAR-2 activation (11)
, one may hypothesize that PAR-2 may be a promising target for the treatment of inflammatory responses in the skin (ref 4
and references).
The experimentally induced CD model is a common and reliable animal model for the investigation of inflammatory responses in the skin, sharing clinical, histological, and immunological features with human CD. Thus, we used PAR-2-/- mice in the CD model to study the role of PAR-2 in the pathogenesis of cutaneous inflammation. We demonstrate that PAR-2 induces edema, plasma extravasation, leucocyte recruitment, up-regulation of cell adhesion molecules, and proinflammatory cytokines during the early phases of both toxic and allergic contact dermatitis. Thus, PAR-2 inhibitory agents may have a beneficial effect in treating CD.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
PAR agonists
Corresponding peptides to the tethered ligand domain of PAR-2 were used: SLIGRL-NH2 (mouse PAR-2 AP), SLIGKV-NH2 (human PAR-2-AP), and trans-cinnamyl-LIGRLO (modified mouse PAR-2 AP) as well as LRGILS-NH2 (reverse of mouse PAR-2 tethered ligand, negative control) and VKGILS-NH2 (reverse of human PAR-2 tethered ligand, negative control) as controls (8)
.
Ear swelling experiments during ACD and ICD
PAR-2-AP (SLIGRL) or RP was applied intradermally (0.5–50 µM) into mouse ears. CD was induced as described previously (12
, 13)
. After measuring ear thickness in untreated mice, ICD was induced by applying 5% benzalkonium chloride (Sigma, Taufkirchen, Germany) dissolved in olive oil:acetone (1:5 v/v) or 0.8% croton oil (13)
on the dorsal aspect of the ear. For induction of ACD, mice were sensitized on days 0 and 1 by applying 50 µL of 0.5% 2,4-dinitro-1-fluorobenzene (DNFB, Sigma) in olive oil:acetone (1:5 v/v) on the shaved abdomen. Mice were challenged on day 6 by epicutaneous application of 20 µL 0.3% DNFB in olive oil:acetone on the front and dorsal surface of the left ear (12)
. In all experiments, right ears were treated with vehicle alone. In both dermatitis models, responses were determined by the degree of ear swelling compared with the vehicle-treated contralateral ear by using a caliper (Oditest, Kroeplin, Schüchtern, Germany). To inhibit the NO synthesis (14)
, mice received N
-nitro-L-arginine methyl ester (L-NAME, 10 mg/kg) in 200 µL phosphate-buffered (PBS) i.p. at 12 and 1 h before challenge. Alternatively, N-nitro-L-arginine (L-NNA) was used at the same concentrations to confirm the specific effect of NOS inhibitors. To investigate the roles of neuropeptides in both dermatitis models, mice were treated with the sensory neurotoxin capsaicin (Zostrix HP cream; GenDerm, Lincolnshire, IL, USA) containing 1% trans-8-methyl-N-vanillyl-6-noneamide (capsaicin) to the left ear twice a day for 7 days, as described (13)
. The NF-
B inhibitor Bay-11-7082 (Calbiochem, Bad Soden, Germany) was used at a final concentration of 10-4 M and applied intradermally into mouse ears 30 min before challenge.
Measurement of microvascular permeability
ACD and ICD were induced in left and right (vehicle control) ears as described (13
, 15)
. Mouse ears were harvested at 0, 1, 3, and 6 h after ICD induction or ACD challenge. To measure extravasation of Evans blue, anesthetized mice were injected with 30 mg/kg Evans blue (Sigma) in 0.9% NaCl into a femoral vein. Seven minutes after Evans blue injection, mice were transcardially perfused with 50 mL PBS saline (10 mM; NaCl, 120 mM; KCl, 2.7 mM, pH 7.4) containing 100 units/mL heparin), followed by 200 mL 1% paraformaldehyde in a 0.05% citrate buffer, pH 3.5. Ears were removed, rinsed in saline, blotted, and weighed. Half of each tissue was dried at 60°C for 48 h and reweighed; the other half was incubated in 1 mL formamide (Sigma) for 48 h at room temperature to extract the Evans blue. Evans blue was quantified by spectrophotometry (13
, 15)
. Extravasation is expressed as nanograms Evans blue per milligram of dry weight.
Immunofluorescence, immunohistochemistry, hematoxylin-eosin staining
In general, immunofluorescence studies were performed as described (7
, 8)
. Briefly, mice were anesthetized, transcardially perfused with 4% paraformaldehyde in 100 mM PBS, pH 7.4, and mouse ears were removed and immersion-fixed for 4 h. Tissue was washed, incubated in 25% sucrose in PBS for 24 h at 4°C, embedded in OCT (Miles, Slough, UK), and 10 µm frozen sections were prepared. Sections were preincubated in PBS containing 10% normal goat serum and 0.3% Triton X-100 for 30 min and incubated with primary antibodies in PBS with 2% normal goat serum and detergent for 12 h at 4°C. Sections were incubated with monoclonal primary antibodies: MoAb 10E9.6 (rat IgG2a) against murine E-selectin, MoAb RB40.34 (rat IgG2a) and against murine ICAM-1 overnight at 4°C (1:10 dilution, ref 16
). For immunostaining of E-selectin in human skin tissues, a mouse monoclonal antibody to human E-selectin (IgG, Santa Cruz, Santa Cruz, CA, USA) was used after a protocol recently described (15)
. Control staining was performed on tissue processed without primary antibodies and corresponding Ig control. No staining was obtained with either of the controls. For hematoxylin-eosin staining, 5 µm paraffin sections were generated from mice used for ear swelling studies, stained for hematoxylin-eosin, and examined using a Zeiss microscope.
Northern blot
The plasmid pcDNA3 containing full-length mouse E-selectin and ICAM-1 cDNA (Dr. Vestweber) was digested with Hind III and Xma I (E-selectin) or Pst I and Xma I (ICAM-1) to obtain a 766 bp (E-selectin) or 997 bp (ICAM-1) cDNA, respectively. cDNAs to ICAM-1, E-selectin, and GAPDH were labeled with 32P using a Ready Prime kit (Amersham, Braunschweig, Germany). Total RNA was prepared from right and left mouse ears from three mice per group using Trizol (Gibco-BRL, Grand Island, NY, USA). RNA hybridization was performed as described (7
, 8)
, and exposed for 3 h (E-selectin) or 6 h (ICAM-1) or comparative time points (GAPDH) at -70°C. RNA expression was quantitated by densitometry using a PhosphorImager and accompanying ImageQuant software (Molecular Dynamics, Sunnyvale, CA, USA). Relative amounts of E-selectin and ICAM-1 RNA were normalized to GAPDH/RNA content.
Enzyme-linked immunosorbent assay (ELISA)
To study whether PAR-2 agonists mediate release of proinflammatory mediators (IL-6), and cell surface expression of ICAM-1 in mouse ears during ACD, tissues were collected as described above (3 h, 6 h, 9 h, 12 h, 24 h), pulverized, and frozen until use for ELISA assays. To detect IL-6, ELISA kits from BioSource, Int. (Solingen, Germany) were used (sensitivity <2 pg/mL). ELISAs were performed with 10 times diluted samples according to the manufacturer’s instructions. ELISA for ICAM-1 was performed as described (17)
. All data were calculated from triplicate wells of n= 3 experiments and are presented as the mean ± 2 SE. The optical density of samples was measured with a Microplate Reader 3550 BioRad (München, Germany) at a wavelength of 450 nm. Release of ICAM-1 and IL-6 was normalized by total protein content.
Intravital microscopy
Animals were anesthetized by i.p. injection of 10 mg/kg xylazine (MTC Pharmaceuticals, Cambridge, Ontario, Canada) and 200 mg/kg ketamine hydrochloride (Rogar/STB, London, Ontario, Canada). Intravital microscopy was performed on a skin flap, which does not allow for visualization of leukocyte/endothelial cell interaction by simple transillumination. Therefore, after anesthesia, mice received an intravenous injection of a fluorescent dye, rhodamine 6G (Sigma, St. Louis, MO, USA) (0.3 mg/kg). At this dose, rhodamine 6G labels leukocytes and platelets and has been shown to have no effect on leukocyte kinetics (18
, 19)
. A midline abdominal incision was performed from the diaphragm, extending to the pelvic region. The skin was carefully separated from the underlying tissue but remained attached laterally, so the blood supply to the skin flap remained intact (20)
. The skin flap was extended over a viewing pedestal to expose the dermal microvasculature and secured along the edges using 4.0 sutures. The exposed dermal tissues were superfused with a bicarbonate-buffered saline pH 7.4, to avoid tissue dehydration. The microcirculation was observed using a microscope (Nikon) with a x20 objective lens; rhodamine 6G allowed visualization and quantification of the number of rolling and adherent leukocytes by epi-illumination at 510–560 nm, using a 590 nm emission filter. Single unbranched venules (20–40 µm in diameter) were selected for the study. Images of the selected venule were recorded for 5 min after a 15 min equilibration period; the end of the 5 min interval was considered as time 0 (9)
. Then, contact dermatitis was induced by applying 5% benzalkonium chloride (Sigma), dissolved in olive oil:acetone (1:5 v/v) on the surface of the observed abdominal skin. Images of the selected venule were then recorded for 5 min beginning 15, 30, 45, and 60 min after the induction of contact dermatitis. Leukocyte adherence was determined upon video playback, on a 100 µm vessel length, a leukocyte being considered adherent to the endothelium if it remained stationary for 30 s or more. Leukocyte flux was defined as the number of leukocytes per minute moving at a velocity less than that of erythrocytes, which passed a reference point in the venule. The changes in flux of rolling leukocytes were evaluated as differences between the number of rolling leukocytes at each interval and the basal number of rolling leukocytes.
Measurement of protein extravasation in human skin by microdialysis
Subjects (healthy human beings age 18–50 years, n=8 per group) were seated comfortably on a reclining chair in a temperature-controlled laboratory (21°C; 60% rel. humidity). The forearm was supported by an arm rest at heart level and subjects acclimatized for 20 min. Five microdialysis catheters (0.4 mm in diameter, cutoff 3000 kDa; DermalDialysis, Erlangen, Germany) were inserted intracutaneously at a length of 1.5 cm in nonlesional skin of the volar forearm using a 25 G cannula as described previously (21)
. No local anesthesia was required. All microdialysis catheters were orientated transversally to the axis of the volar forearm at a distance of 4 cm between each capillary was used. They were perfused with Ringer's solution (Ringerlösung Fresenius, Germany) by a microdialysis pump (Pump 22, Harvard Apparatus, USA) at a constant flow rate of 4 µL per minute via a Tygon® tubing (Novodirect, Kehl/Rhein, Germany). Dialysate was sampled at 15 min intervals for a total period of 120 min. After a baseline of 60 min 100 µL of Ringer’s solution containing the PAR-2-agonist SLIGKV (5x10-4 to 10-2M) were injected intradermally at a distance of 2 mm at both sides of the fiber, such that the injection blebs covered the entire length of the intradermal microdialysis catheter. Concentration-independent relative recovery of albumin was confirmed in vitro using identical dialysis fiber specifications as used in vivo. Relative recovery was 25 ± 5% (mean±SE; albumin 1–30 mg/mL in Ringer’s solution). Relative recovery for the mediators was 24 ± 3% for histamine (10-6M) and 22 ± 5% (tryptase, 100 pg/mL).
Blood flow was assessed at 5 min intervals using a laser Doppler scanner (LDI, Moor Instruments, Devon, UK). To assess flare area, all pixels around each microdialysis catheter exceeding mean flux + 2*standard deviation of the baseline scan were analyzed using dedicated software (Moor LDI image processing).
Total protein content was measured photometrically (MRX reader, Dynatech, Denkendorf, Germany) according to a protocol by Bradford using Coomassie blue dye for the analysis and bovine serum albumin as a standard as described previously. Histamine concentration in the dialysate was measured by a fiber-based spectrofluometric assay as described before (22)
. Mast cell tryptase concentration in the dialysate was measured by using specific immunoassays (Uni CAP Tryptase; Pharmacia and Upjohn, Freiburg, Germany), according to the manufacturer’s instructions.
Statistical analysis
Results are expressed as mean ± SE, as indicated in figure legends. Comparisons between groups were assessed using a two-way ANOVA and Student-Newman-Kuels test (multiple groups). P 
0.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 9
, 15
, 23
, 24)
. However, precise data about a potential regulatory role of PAR-2 in an in vivo disease model of intracutaneous inflammation are rare (25)
and the underlying mechanisms of PAR-2-induced cutaneous inflammation are still unknown. To evaluate the role and mechanisms of intradermal PAR-2-mediated responses during skin inflammation, we used PAR-2-/- mice for experimentally induced CD. We also used NO inhibitors, the neurotoxin capsaicin, and a NF-B inhibitor to explore a potential NO-, NF-B-dependent, and/or neuronally mediated response.
Intradermal injection of SLIGRL (10–100 µM) into PAR-2+/+ mice induced a marked dermal edema, plasma extravasation and recruitment of leucocytes into the skin that was not observed in PAR-2-/-mice (Fig. 1
A, ear-swelling; see Fig. 3I, K
, histological changes; concentration of 10-4 M AP shown). L-NAME abrogated these effects, showing that PAR-2-induced proinflammatory effects in PAR+/+ mice is a NO-mediated event. L-NAME also slightly decreased mild ear swelling responses in PAR-2-/- mice at 1 h (data not shown).
|
|
In PAR-2+/+ mice, ACD was associated with a significantly increased dermal edema within the early phase of inflammation that was maximal at 48 h (Fig. 1 B
). The largest difference of dermal edema between PAR+/+ mice (0.56 mm±0.05 mm SE) and PAR-2-/- mice (0.37 mm±0.04 SE) was observed at 3 h. Between 6 h and 48 h, no significant differences were observed in PAR-2-/- and PAR-2+/+ mice (Fig. 1B
). All control groups (RP, which does not activate PAR-2; only vehicle used; -/- mice; no sensitization) showed a similar response with minimal effects (Fig. 1A-E
). During ICD, ear swelling responses were also maximal after 6 h (Fig. 1F-I
). The NO synthase (NOS) inhibitors L-NAME (Fig. 1C
) and L-NNA (not shown) strongly inhibited PAR-2-induced ear edema by 62% (1 h), 83% (3 h) and 90% (6 h), respectively, during ACD. Similar effects were observed during ICD (Fig. 1G
). Capsaicin pretreatment slightly but not significantly diminished PAR-2-induced ear swelling responses during the first 6 h in PAR-2+/+ mice in the ACD (Fig. 1D
) and ICD model (Fig. 1H
). Effectiveness of the capsaicin treatment was verified by eye blinking experiments, as described (8)
. In contrast, the NF-B inhibitors Bay 11-7082 (Calbiochem; Fig. 1E, I
) did not influence ear swelling responses in PAR-2-/- or PAR-2+/+ mice. Thus, PAR-2-induced ear edema was not dependent on release of neuropeptides or activation of NF-B from dermal sensory nerve fibers, but NO mediated the edema responses to PAR-2 agonists during CD.
PAR-2 is involved in intradermal plasma extravasation during CD in a NO-dependent manner
Because edema requires the formation of gaps between endothelial cells of postcapillary venules and subsequent extravasation of plasma into tissues, we quantified extravasation of plasma proteins using Evans blue, which permitted simultaneous evaluation of PAR-2-mediated plasma extravasation in murine skin during inflammatory ACD and ICD. We compared the effects of allergic (DNFB) and toxic agents (benzalkonium chloride 5%, 0.8% croton oil in acetone) on plasma extravasation in the skin of PAR+/+ mice and PAR-2-/- mice.
Intradermal injection of SLIGRL (10–100 µM) into PAR-2+/+ mice induced a marked plasma extravasation, with a maximum at 1 h, which was not observed in PAR-2-/-mice (Fig. 2
A). L-NAME abrogated the SLIGRL-induced effects showing that PAR-2-induced plasma extravasation was mediated by a NO-dependent mechanism (Fig. 2B
).
|
During ACD, PAR-2 agonist stimulated Evans blue extravasation in the skin maximally by 325% +/- 12% SE over the basal response to vehicle control in wild-type mice and by 62.5% +/- 5% SE compared with PAR-2-/-, with a maximum at 1 h. Pretreatment of mouse ears in PAR-2+/+ mice with the NOS inhibitor L-NAME before challenge significantly attenuated ACD-induced intracutaneous extravasation of Evans blue by 27.5% +/- 3% SE (Fig. 2B
). During ICD, Evans blue extravasation was similarly induced maximally by 5-fold over the basal response (vehicle) in wild-type mice and by 2.4-fold compared with PAR-2-/- at 3 h (not shown). Differences between PAR-2-/-mice and controls were not statistically significant (Fig. 2A, B
). Pretreatment of mouse ears with capsaicin and NF-B inhibitor (Bay 11-7082) before ACD and ICD induction did not result in statistically significant reduction of plasma extravasation in the skin (not shown). Effectiveness of the capsaicin treatment was verified by eye blinking experiments, as described (8)
. Thus, our results show that PAR-2-induced plasma extravasation in the skin requires release of NO but not of neuropeptides or NF-B.
Microscopic examination of skin tissues from mouse ears after ACD revealed increased edema, plasma extravasation, and emigration of neutrophils at 12 h in PAR-2+/+ mice (Fig. 3
A, C, E) in vivo. These results were maximally increased when mice were pretreated with SLIGRL 30 min before challenge (Fig. 3A, C, E
). This effect was significantly diminished in PAR-2-/- mice with (Fig. 3B, D, F, K
) or without SLIGRL pretreatment (Fig. 3K
). PAR-2+/+ mice with allergen challenge but without SLIGRL pretreatment (Fig. 3I
) demonstrated less inflammatory characteristics compared with SLIGRL-pretreated and challenged PAR-2+/+ mice (Fig. 3A, C, E
) but enhanced inflammation compared with L-NAME pretreated and challenged PAR-2+/+ mice (Fig. 3L
) or PAR-2-/- mice (Fig. 3B, D, F, H, K, M
).
PAR-2 is involved in up-regulation of cell adhesion molecules E-selectin and ICAM-1 in experimentally induced CD
To determine whether PAR-2 is involved in cell adhesion molecule regulation during CD, we localized E-selectin and ICAM-1 in sections of mouse skin by immunofluorescence. E-selectin (Fig. 4A-D
) and ICAM-1 (Fig. 4E-H
) immunoreactivity were detected in microvascular dermal endothelial cells of PAR-2-/- mice and PAR-2+/+ mice. However, changes for E-selectin and ICAM-1 immunoreactivity were observed during experimentally induced ACD. In PAR-2+/+ mice, immunoreactivity was enhanced for both E-selectin and ICAM-1 after antigen challenge at 3 h (E-selectin, Fig. 4
C) and 6 h (ICAM-1, Fig. 4G
). In contrast, only slight effects on E-selectin or ICAM-1 immunoreactivity were observed in PAR-2-/- mice (Fig. 4D, H
) compared with controls (Fig. 4A, B, E, F
). These semiquantitative results indicate that cell adhesion molecules ICAM-1 and E-selectin may be regulated at least in part by PAR-2 during experimentally induced ACD.
|
These observations are supported by RNA gel blotting experiments. Primary transcripts of E-selectin and ICAM-1 (Fig. 5
A) were detected in mouse ears of ACD and control animals after 3 h (E-selectin) and 6 h (ICAM-1), respectively. In PAR-2+/+ mice, significantly higher levels of E-selectin (375±55%) and (to a lesser extent, but statistically significant) ICAM-1 (251±37%) RNA expression were observed compared with PAR-2-/- mice after induction of ACD (Fig. 5B
). Thus, E-selectin and ICAM-1 mRNA expression is up-regulated in PAR-2+/+ mice during ACD whereas RNA for these CAMs was only slightly enhanced in PAR-2-/- mice, supporting the idea that PAR-2 is involved in the up-regulation of adhesion molecules.
|
PAR-2 stimulates up-regulation of ICAM-1 protein and release of IL-6 during murine ACD
To study potential up-regulation of PAR-2-mediated ICAM-1 protein and of proinflammatory cytokines (IL-6) in murine ears during ACD, tissues from PAR-2+/+ and PAR-2-/- mice were collected at the indicated intervals (1 h, 3 h, 12 h, 24 h, 48 h) and processed for ELISA assays (Fig. 6
). In accordance with immunofluorescence and RNA data, ICAM-1 protein was enhanced in PAR-2+/+ mice compared with PAR-2-/- mice (165%±38% at 12 h and 412%±35% at 0 h) and to vehicle (492%±40%, not shown), with a maximum at 12 h (Fig. 6A
). IL-6 release was increased by three- to sevenfold in PAR-2+/+ mice compared with PAR-2-/- mice in ACD animals with a maximum 3 h after challenge, suggesting PAR-2-induced up-regulation of proinflammatory IL-6 during CD (Fig. 6B
).
|
PAR-2 activation is involved in contact dermatitis-induced leukocyte rolling and adherence
Under basal conditions (time 0, before the induction of CD), an average of 30 leukocytes per minute was observed rolling onto the vessel wall and no difference was observed in the flux of rolling leukocytes between the groups (PAR-2+/+ or PAR-2-/- mice). No change in flux of rolling leukocytes was observed throughout the experiment (over the 60 min period) in mice that did not have ACD (Fig. 7
A). Induction of ACD in PAR-2+/+ mice significantly increased the flux of rolling leukocytes from 15 to 60 min after exposure to the inflammatory agent (Fig. 7A
). However, no significant increase in leukocyte rolling was observed in PAR-2-/- mice after the induction of ACD compared with the group of PAR-2+/+ mice with no ACD. Under basal conditions (time 0), an average of three leukocytes per 100 µm vessel length were adherent to the vessel wall of the mouse dermal tissues in all groups. From 15 to 60 min after the induction of ACD in PAR-2+/+ mice, the number of leukocytes adhering to the vessel wall was significantly increased compared with basal (Fig. 7B
). In PAR-2-/- mice, no significant increase in the number of adherent leukocytes compared with basal was observed at any time point, and the number of adherent leukocytes after the induction of ACD did not differ from PAR-2+/+ mice without induction of ACD.
|
PAR-2 agonists induce plasma extravasation in human skin in vivo
Baseline levels of protein concentrations 60 min after insertion of the microdialysis catheters did not differ between AD patients and controls. SLIGKV-NH2 injection dose-dependently provoked a prolonged increase of protein extravasation as measured by total protein concentration in the dialysate (Fig. 8
A). After injections of SLIGKV-NH2 at concentrations of 5 and 10 mM, protein extravasation was significantly enhanced in AD patients. Similarly, the injections dose-dependently induced the release of mast cell tryptase (Fig. 8B
). In AD patients, tryptase concentrations were significantly higher than with controls after injection of SLIGKV-NH2 at concentrations of 5 and 10 mM. In parallel to the release of mast cell tryptase, tethered ligand provoked histamine release upon intradermal injection at concentrations of 5 and 10 mM (Fig. 8C
). For the highest stimulatory concentration of SLIGKV-NH2, histamine concentrations were significantly increased in AD patients. Tethered ligand injection dose-dependently caused an axon reflex erythema ("flare"). However, flare areas did not differ between AD patients and controls (Fig. 8D
).
|
SLIGKV-mediated E-selectin induction in normal human skin
We also examined the in vivo effects of PAR-2 agonist SLIGKV-NH2 (AP) on dermal endothelial cell E-selectin expression. SLIGKV-NH2 or VKGILS (RP, negative control) was applied into the dermis of human volunteers (n=3). Biopsies were obtained at 0 h and 6 h and evaluated for E-selectin immunoreactivity associated with microvascular endothelial cells in the dermis. Low constitutive levels of microvascular E-selectin immunoreactivity were observed in untreated (Fig. 9
A) or RP-treated skin (Fig. 9C
) at 0 h. A marked increase in microvascular endothelial cell E-selectin staining was observed 6 h after application of SLIGKV-NH2 agonist (Fig. 9B
) but not after application of RP (Fig. 9D
). Thus, PAR-2 agonists induce increased in vivo microvascular endothelial E-selectin protein expression in humans in vivo.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Accumulating evidence suggests that PAR-2 is an important mediator of inflammation (4
5
6
, 26
27
28)
. However, the exact role of PAR-2 in different organs and its value as a therapeutic strategy for the treatment of inflammatory diseases in humans are still unclear. For example, PAR-2 appears to exert pro- as well as anti-inflammatory effects depending on cell type and disease state (8
, 9
, 27
28
29)
. Thus, disease models in which the PAR-2 gene is deleted by homologous recombination are important in solving the problem of PAR-2 agents as potential therapeutic tools in different diseases (30)
.
Our results indicate that PAR-2 is involved in NO-dependent intradermal edema and extravasation since these effects were dramatically reduced by specific NOS inhibitors during experimentally induced CD. Pretreatment of mouse ears with PAR-2-AP before challenge potentiated the PAR-2-induced proinflammatory reaction in mouse skin supporting the idea that PAR-2 mediates a proinflammatory response during ACD and ICD. On the other hand, pretreatment using capsaicin only slightly diminished these effects suggesting a minor role of neuropeptides in this mechanism. No significant changes between PAR-2+/+ mice and PAR-2-/- mice were observed using NF-B inhibitors before challenge. Thus, NO is an important mediator of PAR-2-induced inflammatory responses during dermatitis. This may have a therapeutic consequence since NO inhibitors were shown to be effective in regulating vascular responses during ACD and ICD. This agrees with observations in rats under physiological conditions. Compton et al. (31)
found that activation of PAR-2 was blocked by L-NAME in a rat model. The involvement of nitric oxide in PAR-2-induced effects was also shown in rats during hyperalgia and for mucus secretion (32
33)
. Finally, Wallengren et al. (34)
as well as Sahin et al. (35)
have already demonstrated a role of NO during allergic contact dermatitis in humans.
Infiltration of leucocytes is a key feature of contact dermatitis. Cell adhesion molecules mediate initial leukocyte adhesions, followed by subsequent trans-endothelial migration and have been suggested to play a crucial role during CD (36
37
38)
. Various serine proteases as well as PARs have been demonstrated to modulate cell adhesion molecule expression in vitro (9
, 27
, 39)
. Here we demonstrate in a model of cutaneous inflammation as well as human skin biopsies that PAR-2 agonists are sufficient to mediate E-selectin up-regulation and, to a lesser extent, cell adhesion molecule (ICAM-1) up-regulation, indicating an additional role of PAR-2 in modulating dermal microvascular endothelial cell function during dermatitis. However, PAR agonists are not always selective, and there are no PAR-2 antagonists that would permit evaluation of agonist selectivity or investigation of the role of PAR-2 in experimental models of human disease.
Here we have shown for the first time in a disease model that endothelial cells of PAR-2+/+ mice express higher amounts of cell adhesion molecules both at the mRNA and protein level during inflammation in vivo. This correlates well with our finding in human skin. Additional in vivo experiments in mice give evidence that PAR-2 functionally modulates leucocyte-endothelial cell interactions during cutaneous inflammation. During experimentally induced ACD, we demonstrated significantly attenuated rolling and adhesion of leucocytes in PAR-2-/- mice compared with PAR-2-/- mice. These data agree with results from other groups who found PAR-2 involvement in leucocyte rolling and adhesion after induction of PAR-2 agonists in noninflammatory conditions (8
, 9
, 27)
. As seen in Fig. 8
, in patients with atopic dermatitis a disease in which the concentration mast cell tryptase is enhanced (M. Steinhoff et al., unpublished results) show a marked increase of plasma extravasation and tryptase release due to PAR-2 agonists compared with controls, as demonstrated by microdialysis studies. Tryptase is a good candidate to facilitate mast cell-mediated effects during contact dermatitis via activation of PAR-2. Interesting in that context is that mast cells have recently been shown to be critically involved in the regulation of T cell responses during CD by releasing TNF-
(40)
. Moreover, PAR-2 and tryptase are coexpressed on human mast cells. Thus, tryptase may activate PAR-2 not only on dermal endothelial cells, but also on mast cells in an autocrine manner during dermatitis (M. Steinhoff et al., unpublished results).
Together, our results are clearly in favor of a proinflammatory effect of PAR-2 during early stages of experimentally induced CD. This is also supported by the findings that proinflammatory agents markedly up-regulate PAR-2 on endothelial cells (39
, 41)
and that PAR-2 agonists induce endothelial cell activation by stimulating vasodilatation, plasma extravasation, neutrophil adhesion, neutrophil infiltration, and secretion of proinflammatory cytokines and neuropeptides (8
, 9
, 29
, 42
43
44
45
46
47
48)
. However, results about a proinflammatory vs. anti-inflammatory role of PAR-2 in vivo are contradictory and clear effects of PAR-2 agonists in an inflammatory disease model are so far lacking. There are also observations of an anti-inflammatory effect of PAR-2 activation in different tissues and species (49)
. For example, PAR-2 agonists induce relaxation of the intestine (50)
, up-regulation of cytoprotective prostanoids in the jejunum (51)
and gastric mucosa (52)
, and exert anti-inflammatory responses during experimentally induced colitis (28)
. In the airways, PAR-2 may play a protective role and partially reverse bronchoconstriction by inducing eicosanoid release (29)
. In contradiction, other findings show that PAR-2 agonists intensify the contraction of airway smooth muscle cells and potentiate the contractions to histamine. Additionally, induced proliferation of human lung cells was observed (53
54
55)
. Thus, mucosal surfaces may respond differently from squamous epithelium to PAR-2 stimulation, e.g., by generating protective mediators for the mucosa (56)
. Another explanation is that local effects of PAR-2 activation have been demonstrated to be often proinflammatory while systemic effect were protective (52
, 56)
. Others suggest a proinflammatory role of PAR-2 agonists during airway inflammation by up-regulating release of MMP-9 and GM-CSF, which promotes eosinophil activation and survival (57
, 58)
, and by inducing bronchoconstriction in human airway smooth muscle cells (59
, 60)
. Finally, differences in the glycosylation state of PAR-2 may be involved in the susceptibility of PAR-2 to agonist activation (56)
.
From our findings and other in vivo and in vitro observations, we conclude that under physiological conditions in certain tissues, PAR-2 is a receptor with a diverse role during different phases of inflammation. In the early phase, PAR-2 activation induces early proinflammatory responses such as vasodilatation (9
, 11
, 27
, 61)
, recruitment of leucocytes (8
, 9
, 11)
, and up-regulation of proinflammatory mediators such as neuropeptides (8)
, NO, IL-6 (Fig. 6)
(11)
, IL-8 (11
, 61)
, MMP-9, and eotaxin (57)
. PAR-2 also induces NF-B activation in human keratinocytes (62)
and release of proinflammatory cytokines (11)
. This agrees with PAR-2 induced activation of NF-B with a maximum after 1 h. Maximal stimulation after 1 h reflects proinflammatory effects of NF-B whereas maximal stimulation of NF-B at 4 h is associated with an anti-inflammatory effect of this transcription factor (62
, 63)
. During later phases of inflammation, PAR-2 may exert its anti-inflammatory effects by up-regulating prostanoids (49
, 51)
and secretion (52)
. For example, in a model of chronic inflammation, PAR-2 induced down-regulation of inflammatory cytokines and T cells (28)
, suggesting an anti-inflammatory effect of PAR-2 during certain phases of inflammation in the gut. However, we did not observe significant differences between PAR-2+/+ and PAR-2-/- mice at later stages, suggesting a minor role of PAR-2 for T cell regulation during contact dermatitis in our model. One also has to consider that more lipophilic synthetic PAR-2 agonists may provoke a longer proinflammatory response compared with SLIGRL (48)
. Since no PAR-2-deficient mice were used in the colitis model, effects of other PARs except PAR-2 cannot be excluded. Moreover, Balb/c mice were used in the colitis model (28)
, which differ in their immune response to C57/B6 mice during inflammation (64)
. In our model, we used PAR-2-/- mice to verify a specific proinflammatory effect of PAR-2 during early stages of CD.
In the skin, keratinocytes seem to be important "sensors" to respond to exogenous (bacterial, house dust mites, etc.) (65
, 66)
or endogenous proteases such as extrapancreatic trypsin (67)
, tryptase (7)
, SCTE (68
; M. Steinhoff, unpublished observation), or factor Xa/VII from the bloodstream (69)
. Mast cells are closely associated with blood vessels as a microvascular unit and tryptase has been shown to play an important role during early phases of inflammation (70
71
72
73
74)
. In human skin, PAR-2 is up-regulated in keratinocytes during atopic dermatitis, a disease in which activation by serine proteases from house dust mite or bacteria may play an important pathophysiological role (66
; M. Steinhoff, unpublished observation). Thus, allergens, irritants and pathogens may activate PAR-2 during cutaneous inflammation leading to a proinflammatory effect by up-regulating cell adhesion molecules, induction of edema and plasma extravasation, and recruitment of leucocytes to the site of inflammation.
Potential endogenous ligands for endothelial PAR-2 during dermatitis may be tryptase (23
, 75
, 76)
endothelial or epithelial trypsin (77)
, or factor Xa/VII from the bloodstream (69)
. Mast cells are closely associated with blood vessels as a microvascular unit and tryptase has been shown to be a potent vasodilator that can cause increased microvascular permeability and protein extravasation (46)
. Tryptase appears to play an important role during the immediate type phase of hypersensitivity (70
, 71)
and airway inflammation (49
, 78)
. Moreover, tryptase modulates leukocyte effectors activities such as cell cytotoxicity (72)
mast cell degranulation, macrophage and polymorphonuclear leukocyte activation (73)
, angiogenesis (79
80
81
82)
, and cytokine production by monocytes and cell adhesion molecule regulation (74)
.
Our present data on the effect of PAR-2 agonists on cutaneous microvascular endothelial cell regulation further supports the role of PAR-2 during inflammatory skin diseases. PAR-2 is capable of influencing ear swelling responses, plasma extravasation, and adhesion molecule expression on dermal microvascular endothelial cells. This is supported by the finding that PAR-2 AP is capable of inducing E-selectin and, to a lesser extent, ICAM-1 up-regulation in human skin tissues. We further demonstrate that the induction of PAR-2-mediated ear swelling responses can be diminished in vivo after application of NOS inhibitors and the neurotoxic agent capsaicin. NO seems to be the major contributor of PAR-2-mediated effects during dermatitis. NO has long been demonstrated as an important mediator of cutaneous inflammatory responses and PAR-2-mediated effects in various tissues and blood vessels (43
, 60)
. Here we give additional evidence that NO is involved in PAR-2-mediated inflammatory responses in the skin in vivo. Thus, NOS inhibitors may be beneficial therapeutical tools for the treatment of PAR-2-mediated inflammation.
PAR-2 is also known to induce release of cytokines from various endothelial cells in vitro (11
, 61)
. Our results indicate that PAR-2 agonists directly induce IL-6 release in mouse ears during CD in vivo. Thus, cytokine release may play a role in endothelial adhesion molecule up-regulation in vivo, although cytokine release is neither necessary nor required for PAR-2-mediated induction of ICAM-1 or E-selectin in cultured microvascular endothelial cells.
Our studies prove that PAR2 is capable of directly regulating cutaneous inflammation in a dermatitis model in mice and humans. Very recently it was demonstrated that PAR-2 participates in the pathophysiology of arthritis (48)
. Using PAR2-deficient mice, the authors showed that PAR-2 deficiency results in a marked reduction of swelling responses, synovial hyperplasia and inflammatory infiltrate during experimentally induced monoarthritis. This also supports a role of PAR-2 as a mediator of proinflammatory responses in the joints. Moreover, a novel lipophilic synthetic agonist PAR-2 peptide (ASKH95, phenylacetyl-LIGKV-OH) revealed proinflammatory effects after intra-articular injection like synovial hyperemia and joint swelling. These results clearly demonstrate that the PAR-2 agonist ASKH95 induces signs of chronic inflammation (48)
.
These findings further support a regulatory role of serine proteases and PARs during skin inflammation and immune response in vivo. Our results are in favor of an early effect of PAR-2 in orchestrating the inflammatory response in the skin and suggest a NO-dependent over a neuronally or NF-B-mediated pathway during this disease. Further studies investigating the endogenous cutaneous proteases, protease inhibitors, and secondary messengers involved in the PAR-2-mediated inflammatory responses may provide a novel basis for controlling inflammatory skin disorders such as contact dermatitis, urticaria, psoriasis, and atopic eczema.
|
ACKNOWLEDGMENTS |
|---|
Received for publication November 14, 2002. Accepted for publication June 10, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
PAR-2-/- with PBS, 
PAR-2-/- with SLIGRL, 
PAR-2-/- with LRGILS, 
PAR-2+/+ with PBS, 
amp; PAR-2+/+ with SLIGRL, 
PAR-2+/+ with LRGILS. B) Mice were treated with agents to induce ACD as described in Materials and Methods. Ear swelling responses were determined before and at various time points after elicitation. 
= 1, PAR-2+/+ ACD with SL; 2, PAR-2+/+ ACD with SL and L-NAME; 3, PAR-2-/- ACD with SL; 4, PAR-2-/- with SL and L-NAME; 5, PAR-2+/+ AC; 6, PAR-2+/+ ACD with L-NAME; 7, PAR-2-/- ACD; 8, PAR-2-/- with L-NAME). *Significant effect of strain and peptide treatment, P 
