Abstract
Prostaglandin (PG) D2 is an arachidonic acid metabolite that is released by allergen-stimulated mast cells. It is a potent in vitro chemoattractant for human eosinophils, acting through the DP2 receptor/chemoattractant receptor-homologous molecule expressed on Th2 cells (CRTH2). Furthermore, there is in vivo evidence that PGD2 contributes to allergen-induced pulmonary eosinophilia via its classic DP1 receptor. The PGD2-derived product 15-deoxy-Δ12,14-PGJ2 is widely used as a peroxisome proliferator-activated receptor γ agonist and has been shown to have anti-inflammatory properties. However, this substance can also activate eosinophils in vitro through the DP2 receptor. The objectives of the present study were to determine whether PGD2 and 15-deoxy-Δ12,14-PGJ2 can induce pulmonary eosinophilia, and, if so, to examine the abilities of selective DP1 and DP2 receptor agonists to induce this response. Brown Norway rats were treated by intratracheal instillation of PGs. Vehicle and 5-oxo-6,8,11,14-eicosatetraenoic acid were used as negative and positive controls, respectively. Lung eosinophils were identified by immunostaining of lung sections with an antibody to major basic protein. Both PGD2 and 15-deoxy-Δ12,14-PGJ2 induced robust eosinophilic responses that were apparent by 12 h and persisted for at least 48 h. Two selective DP2 receptor agonists, 15R-methyl-PGD2 and 13–14-dihydro-15-keto-PGD2, induced similar responses, the former being more potent than PGD2, whereas the latter was less potent. The selective DP1 receptor agonist BW245C [(4S)-(3-[(3R,S)-3-cyclohexyl-3-hydroxypropyl]-2,5-dioxo)-4-imidazolidineheptanoic acid] was completely inactive. We conclude that PGD2 and 15-deoxy-Δ12,14-PGJ2 induce eosinophil infiltration into the lungs through the DP2 receptor. The potent in vitro DP2 receptor agonist 15R-methyl-PGD2 is also very active in vivo and should be a useful tool in examining the role of this receptor.
Prostaglandin (PG) D2 is a metabolite of arachidonic acid that is formed by the action of PGD synthase on the cyclooxygenase product PGH2. There are two enzymes that catalyze this reaction: lipocalin-type PGD synthase, which is found in the central nervous system and hematopoietic PGD synthase, which is present in antigen-presenting cells, mast cells (Urade and Hayaishi, 2000), and Th2 cells (Tanaka et al., 2000). The association of PGD2 with mast cells has focused attention on its possible role in asthma and other allergic diseases. High levels of this prostaglandin are rapidly released into the airways after allergen challenge of asthmatic human subjects (Murray et al., 1986).
The recent finding by Narumiya's group that disruption of the classic Gs-coupled (DP1) receptor for PGD2 protects mice against asthma-like responses after antigen challenge (Matsuoka et al., 2000) has provoked considerable interest in the role of PGD2 in this disease. Mice lacking this receptor exhibit reduced pulmonary eosinophilia, hyperresponsiveness, and Th2 cytokine levels in response to antigen compared with wild-type mice (Matsuoka et al., 2000). Similarly, a DP1 receptor antagonist was reported to reduce pulmonary eosinophilia in a guinea pig model of asthma (Arimura et al., 2001). Conversely, exacerbated responses to antigen were observed in mice overexpressing lipocalin-type PGD synthase (Fujitani et al., 2002).
We demonstrated the existence of a second receptor (DP2 receptor) for PGD2 on eosinophils that mediates eosinophil migration in response to this prostaglandin (Monneret et al., 2001). The DP2 receptor seems to be identical to an orphan receptor cloned by Nagata et al. (1999) and named chemoattractant receptor-homologous molecule expressed on Th2 cells (CRTH2), which they subsequently found to be activated by PGD2 (Hirai et al., 2001). The selectivity of the DP2 receptor differs from that of the DP1 receptor, in particular with respect to the effects of modifications of the alkyl side chain, many of which have relatively little impact on DP2-mediated responses, but abrogate DP1-mediated responses. Examples of this are 15R-methyl-PGD2, the most potent known DP2 receptor agonist (Monneret et al., 2003); 15-deoxy-Δ12,14-PGD2 (Monneret et al., 2002); and 13,14-dihydro-15-keto-PGD2 (dhk-PGD2) (Hirai et al., 2001; Monneret et al., 2001).
15-Deoxy-Δ12,14-PGJ2 (15d-PGJ2) has been widely used as a PPARγ agonist and has a variety of anti-inflammatory properties, both in vitro and in vivo (Lawrence et al., 2002; Powell, 2003). In addition to activating PPARγ, it also covalently binds to certain proteins, and by this mechanism can inhibit the transcription factor nuclear factor-κB, which also contributes to its anti-inflammatory effects (Straus and Glass, 2001). However, relatively high (micromolar) concentrations are required to induce these responses. In contrast, we found that 15d-PGJ2 is a potent DP2 agonist (EC50 of ∼10 nM), which activates eosinophils in vitro through this receptor (Monneret et al., 2002).
The in vivo studies with DP1 receptor knockout mice and DP1 receptor antagonists discussed above would suggest a direct or indirect role for this receptor in antigen-induced pulmonary eosinophilia. In addition, the direct stimulatory actions of DP2 receptor agonists on eosinophils in vitro would suggest a role for this receptor in regulating eosinophil trafficking. Because it is not known whether in vivo administration of PGD2 and related compounds into the lungs can induce pulmonary eosinophilia, the present study was designed to examine the effect of intratracheal instillation of this prostaglandin as well as selective DP1 and DP2 receptor agonists on this phenomenon. In addition, we wanted to examine the effects of 15d-PGJ2 on this process to determine whether its proor anti-inflammatory effects would prevail.
Materials and Methods
Eicosanoids and Reagents. 5-Oxo-6,8,11,14-eicosatetraenoic acid (5-oxo-ETE) was synthesized chemically as described previously (Khanapure et al., 1998). All other eicosanoids as well as BW245C were purchased from Cayman Chemical (Ann Arbor, MI). A monoclonal antibody for eosinophil-derived MBP was kindly provided by Dr. R. Moqbel (University of Alberta, Edmonton, AB, Canada).
Animals and in Vivo Procedures. The animals used in this study were 6- to 8-week-old male Brown Norway rats (8–10 weeks old; 175–215 g; RijHsd substrain), which were purchased from Harlan (Indianapolis, IN). All experimental procedures were approved by an institutional animal care committee. The rats were housed in groups of three with food and water provided ad libitum. They were allowed to become acclimatized to laboratory conditions for 4 to 6 days before experimentation.
Before treatment with eicosanoids, the animals were anesthetized by intraperitoneal injection of xylazine (7 mg/kg) and pentobarbital (50 mg/kg). Different amounts of test compounds in 100 μl of vehicle (saline containing 0.5% ethanol) were administered by intratracheal instillation via a 6-cm-long polyethylene-90 tube connected to a 1-ml syringe containing the test compound followed by 0.6 ml of air to aid in its distribution throughout the lung. Control animals received vehicle alone.
Preparation of Tissue Sections. At various times after administration of eicosanoids, animals were sacrificed by intraperitoneal injection of an overdose of pentobarbital. After induction of anesthesia, rats were bled via the abdominal aorta to drain as much blood as possible. The lungs were removed, washed briefly in phosphate-buffered saline, and inflated with 3.5 ml 100% optimal cutting temperature compound. A section of the left lobe of the lungs around the hilum was removed and placed in an aluminum foil basket filled with optimal cutting temperature embedding medium. The section was then snap frozen in isopentane precooled in liquid nitrogen and stored at –80°C. Sections (6 μm) were cut in a cryostat, air dried for 1 h, fixed in acetone/methanol (60:40) for 7 min, and further air dried for 30 min. The slides were then wrapped in aluminum foil and stored at –20°C.
Immunocytochemistry. To identify eosinophils in the lung sections, immunostaining was performed with the use of the alkaline phosphatase-anti-alkaline phosphatase technique as described previously (Frew and Kay, 1988). Briefly, sections were incubated overnight at 4°C with a mouse anti-human monoclonal antibody to MBP, followed by treatment with rabbit anti-mouse immunoglobulin as a secondary antibody and then alkaline phosphatase-anti-alkaline phosphatase. The reaction was visualized with Fast Red. The slides were then coded and read in a blind manner by two independent observers at a magnification of 200×. Eosinophils were observed in clusters around both large and small airways. For each slide, positive cells were counted in at least four nonoverlapping eosinophil-containing areas (1 mm2) around the airways using a squared eyepiece graticule. In each case, the numbers of cells in the four regions with the highest numbers of cells were averaged.
Data Analysis. Mean values of the cell counts determined for each slide by the two independent observers were used for all further analyses. All values are expressed as means ± S.E. of the numbers of immunoreactive cells per square millimeter. The statistical significance of differences among groups was assessed using either one-way or two-way analysis of variance as appropriate, with the Student-Newman-Keuls test as a multiple comparison method. Differences with P values of less than 0.05 were considered to be statistically significant.
Results
PGD2 Elicits Pulmonary Infiltration of Eosinophils. Brown Norway rats were treated with either vehicle or PGD2 (5 μg), which were administered by intratracheal instillation. Twenty-four hours later, the rats were sacrificed, the lungs were removed, and sections were taken for immunostaining for MBP using a mouse anti-human antibody that cross-reacts with rat MBP. PGD2 induced a dramatic increase in the number of eosinophils in the lung (Fig. 1B) compared with vehicle-treated controls (Fig. 1A). The MBP-positive cells were found principally around the airways. No staining was observed in controls in which the antibody was omitted (data not shown). The response to PGD2 was highly reproducible (P < 0.001) and was only slightly less than that to the potent eosinophil chemoattractant 5-oxo-ETE, which we used as a positive control (Fig. 2).
Pulmonary Eosinophilia Is Induced by DP2 but Not by DP1 Receptor Agonists. Because PGD2 activates both DP1 and DP2 receptors, we wished to determine which of these receptors is responsible for the effect of this prostaglandin on eosinophil infiltration. To address this issue, we compared the effects of the selective DP1 receptor agonist BW245C with those of a series of selective DP2 receptor agonists, including dhk-PGD2, 15R-methyl-PGD2, and 15d-PGJ2 (Fig. 3). Unlike PGD2, BW245C had no effect on the numbers of eosinophils present in the lung (Fig. 2). In contrast, all of the DP2 receptor agonists tested induced eosinophil infiltration, with 15R-methyl-PGD2 (Fig. 1C; P < 0.001) being the most effective, followed by 15d-PGJ2 (Fig. 1D; P < 0.001), and dhk-PGD2 (P < 0.05).
Time Courses for the Effects of DP2 Receptor Agonists on Eosinophil Infiltration. To investigate the time courses for the effects of DP2 receptor agonists on eosinophil infiltration, sections were taken from lungs immediately after administration of vehicle and 4, 12, 24, and 48 h after intratracheal instillation of PGD2, 15R-methyl-PGD2, and 15d-PGJ2 (Fig. 4). Because we have previously shown that lung eosinophil numbers did not change during this time after instillation of vehicle alone (Stamatiou et al., 1998), we did not evaluate responses to vehicle at all of the time points tested. However, the numbers of eosinophils in lungs were the same immediately after administration of vehicle (4.0 ± 0.7 eosinophils/mm2) and 24 h after vehicle instillation (4.1 ± 0.6 eosinophils/mm2). Although none of the substances tested altered eosinophil numbers by 4 h, PGD2 (P < 0.05) and the two selective DP2 receptor agonists (P < 0.005) significantly increased their numbers by 12 h. The maximal response to 15R-methyl-PGD2 was observed after 24 h. PGD2 and 15d-PGJ2 elicited near maximal responses by this time, but the responses to these agonists continued to rise slightly up to 48 h.
Dose-Response Relationships for PGD2, 15d-PGJ2, and 15R-Methyl-PGD2. The effects of different doses of PGD2 and selective DP2 receptor agonists on eosinophil infiltration were examined (Fig. 5). Of the three compounds tested, 15R-methyl-PGD2 was the most potent (P < 0.005) with an ED50 of about 0.6 μg. PGD2, which had an ED50 of about 1.5 μg, induced a smaller response at all doses tested. 15d-PGJ2 seemed to have a potency similar to that of PGD2, although the responses to lower doses of this compound were somewhat variable. Interestingly, diminished responses to all three compounds were observed at the highest dose tested.
Discussion
This study demonstrates for the first time that intratracheal administration of PGD2 alone induces pulmonary eosinophilia. This response was almost as intense as that induced by the potent eosinophil chemoattractant 5-oxo-ETE (Powell et al., 1995), which acts through the recently cloned OXE receptor (Hosoi et al., 2002; Brink et al., 2004) and elicits a robust eosinophilic response in Brown Norway rat lungs after intratracheal administration (Stamatiou et al., 1998). Our results are consistent with an in vivo study in dogs demonstrating that superfusion of tracheal segments with PGD2 induced the accumulation of eosinophils in the superfusion fluid (Emery et al., 1989).
The response to PGD2 is clearly mediated by DP2 rather than DP1 receptors, because it was not shared by BW245C, a potent and highly selective DP1 receptor agonist (Town et al., 1983). In contrast, all of the selective DP2 agonists tested induced pulmonary eosinophilia with relative potencies similar to those predicted by their in vitro effects on human eosinophils, with 15R-methyl-PGD2 being the most potent. This compound, which has the unnatural R configuration at C15, should be a very useful tool for examining DP2 receptor-mediated responses in animal models, because it is more potent than dhk-PGD2, which is presently the most frequently used selective DP2 receptor agonist.
15d-PGJ2 was approximately equipotent with PGD2 in inducing pulmonary eosinophilia, consistent with its potent stimulatory effects on human eosinophils in vitro (Monneret et al., 2002). This is interesting in view of the many reports in the literature on the anti-inflammatory effects of this compound (Powell, 2003). 15d-PGJ2 is formed by the albumin-catalyzed degradation of PGD2 (Fitzpatrick and Wynalda, 1983). Although this clearly occurs in vitro, the presence of this substance in vivo in amounts compatible with its anti-inflammatory effects remains controversial (Powell, 2003). Although 15d-PGJ2 is measurable in urine and cell supernatants by mass spectrometry, its levels are very low and are not enhanced under conditions associated with PPARγ activation or inflammation. For example, 15d-PGJ2 was not augmented in humans in joint fluid from arthritic subjects or after administration of LPS (Bell-Parikh et al., 2003).
The anti-inflammatory effects of 15d-PGJ2 seem to be mediated by activation of PPARγ or by covalent binding to components of the NF-κB system or other cellular proteins (Straus and Glass, 2001). The reactivity of 15d-PGJ2 is due to its cyclopentenone ring structure, which can form adducts with protein thiol groups (Fig. 3). Although it can induce various responses by these mechanisms, including suppression of the expression of cyclooxygenase-2, inducible nitricoxide synthase, and cytokines, the concentrations required are considerably higher than those needed for DP2 receptor-mediated responses (Monneret et al., 2002) and are probably higher than those attained in vivo (Bell-Parikh et al., 2003).
A number of in vivo studies have reported anti-inflammatory effects of 15d-PGJ2 in animal models, including carrageenan-induced pleurisy, in which it inhibited the infiltration of both neutrophils (Cuzzocrea et al., 2002) and mononuclear cells (Gilroy et al., 1999) in the early and late phases, respectively, of this response. The present study clearly demonstrates that 15d-PGJ2 can also induce proinflammatory effects in vivo through activation of the DP2 receptor, suggesting that caution should be used in the development of such compounds as therapeutic agents (Lawrence et al., 2002).
The absence of a response to the selective DP1 receptor agonist BW245C is interesting in view of the inhibitory effects of both disruption of the gene for the DP1 receptor (Matsuoka et al., 2000) and the selective DP1 antagonist S-5751 ([((Z)-7-[(1R,2R,3S,5S)-2-(5-hydroxybenzo[b]thiophen-3-ylcarbonylamino)-10-norpinan-3-yl]hept-5-enoic acid)) (Arimura et al., 2001) on pulmonary eosinophilia in animal models of asthma. The lack of in vitro eosinophil chemotactic activity of BW245C (Monneret et al., 2001) suggests that activation of the DP1 receptor affects lung eosinophil numbers by an indirect mechanism, possibly due to the release of macrophage-derived chemokine (MDC/CCL22) (Honda et al., 2003). Activation of DP1 receptors might also augment the responses to inflammatory mediators by increasing blood flow and vascular permeability in the affected tissue (Flower et al., 1976), consistent with the expression of these receptors on endothelial cells (Nantel et al., 2004). Alternatively, stimulation of DP1 receptors on eosinophils could increase their survival in the lung, as shown for BW245C (although not for PGD2 itself) in vitro (Gervais et al., 2001). On the other hand, the DP1 receptor also has the potential to play an inhibitory role in allergic responses. Both PGD2 and BW245C, when coadministered intratracheally with fluorescein isothiocyanate-labeled ovalbumin, inhibited the migration of dendritic cells from the lungs to draining lymph nodes (Hammad et al., 2003). Activation of the DP1 receptor can also attenuate DP2 receptor-mediated responses (Monneret et al., 2005).
The response to PGD2 and selective DP2 receptor agonists was delayed, because it was absent after 4 h and only about half-maximal by 12 h. This is similar to the time course for 5-oxo-ETE-induced pulmonary eosinophilia (Stamatiou et al., 1998). This may be due to a requirement for mobilization of eosinophils from the bone marrow. The DP2 receptor agonist Δ12-PGJ2 has been shown to induce the release of eosinophils from the bone marrow in the guinea pig isolated perfused hind limb preparation (Heinemann et al., 2003). Moreover, intravenous injection of dhk-PGD2 into Brown Norway rats resulted in increased numbers of circulating eosinophils (Shichijo et al., 2003). Thus, it is possible that the DP2 agonists used in the present study first acted on the bone marrow to induce the release of eosinophils, and then promoted their accumulation in the lung through their chemoattractant properties. Alternatively, we cannot rule out the possibility of an indirect effect due to the release of another mediator that induces eosinophil mobilization and/or migration into the lung.
All three DP2 agonists elicited diminished responses at the highest dose tested. Although in the case of PGD2 this could potentially be explained by an inhibitory effect of the DP1 receptor, this would not apply to the selective DP2 agonists. PGD2 has some agonist activity at TP receptors for thromboxane A2, resulting in bronchoconstrictor responses to higher doses (Coleman and Sheldrick, 1989). It is possible that 15R-methyl-PGD2 and 15d-PGJ2 could also activate these receptors, and thereby elicit bronchoconstrictor responses, which could limit the distribution of agonist throughout the airways, resulting in reduced eosinophil infiltration at high doses. Another possibility is that high doses of the agonists used could have induced eosinophil apoptosis, because both PGD2 and 15d-PGJ2 (10 μM) were reported to have this effect (Ward et al., 2002). Interestingly, PGD2 was selective for eosinophils, whereas 15d-PGJ2 accelerated apoptosis in both eosinophils and neutrophils, through inhibition of nuclear factor-κB activation. Although the proapoptotic effects on eosinophils could potentially have been mediated by the DP2 receptor, this would not explain the effect of 15d-PGJ2 on neutrophils, which do not express this receptor. In another study, lower concentrations of PGD2 (1 μM) had no effect on eosinophil survival, in contrast to the selective DP1 agonist BW245C, which prolonged survival (Gervais et al., 2001), presumably due to stimulation of adenylyl cyclase (Monneret et al., 2001).
Our results differ from a previous report that inhalation of aerosolized PGD2 by antigen-sensitized mice does not increase eosinophil numbers in bronchoalveolar lavage fluid 24 and 48 h later (Honda et al., 2003). However, exposure of these mice to low-dose antigen 24 h after administration of PGD2 induced a strong eosinophilic response that could be blocked by an antibody to MDC. Furthermore, PGD2 was shown to increase the expression of MDC by airway epithelial cells (Honda et al., 2003). The identity of the receptor responsible for these effects is not known, although the DP1 receptor would seem to be the most likely candidate, because it is expressed on airway epithelial cells, in contrast to the DP2 receptor, which is not (Hirai et al., 2001; Nantel et al., 2004). There are several possible explanations for the differences between the two studies. It is possible that the rat may respond more strongly than the mouse to PGD2. Alternatively, the dose of PGD2 used in the mouse may have been too high to observe a direct effect of PGD2 on eosinophil infiltration (cf. Fig. 5). Finally, eosinophil numbers were evaluated by a more selective method (immunocytochemistry) in lung tissue in the present study. Another study that was published while the present manuscript was under review found that PGD2, but not BW245C, induced pulmonary eosinophilia in Brown Norway rats. However, this response was observed only after pretreatment of the rats with interleukin-5 (Shiraishi et al., 2005).
In conclusion, we have shown that PGD2 and selective DP2 receptor agonists induce pulmonary eosinophilia in vivo, in contrast to a selective DP1 receptor agonist, which is inactive. 15R-Methyl-PGD2 is the most potent among these compounds, and should serve as an excellent selective DP2 receptor agonist for in vivo studies. 15d-PGJ2 also elicits pulmonary eosinophilia, demonstrating that in this context it induces a proinflammatory response. These results are consistent with an important role for PGD2 and the DP2 receptor/CRTH2 in allergic diseases such as asthma and suggest that this receptor may be an important therapeutic target in these conditions.
Acknowledgments
We are grateful to Dr. Redwan Moqbel for providing the monoclonal antibody to MBP.
Footnotes
-
This study was supported by Canadian Institutes of Health Research Grants MOP-6254 (to W.S.P.), MOP-13273 (to Q.H.), and MOP-10381 (to J.G.M.); the J. T. Costello Memorial Research Fund; and National Institutes of Health Grants DK44730 and HL69835 (to J.R.). J.R. also acknowledges the National Science Foundation for an AMX-360 NMR instrument Grant CHE-90-13145.
-
doi:10.1124/jpet.104.079079.
-
ABBREVIATIONS: PG, prostaglandin; Th2, T helper 2; CRTH2, chemoattractant receptor-homologous molecule expressed on Th2 cells; dhk-PGD2, 13,14-dihydro-15-ketoprostaglandin D2; 15d-PGJ2, 15-deoxy-Δ12,14-prostaglandin J2; PPARγ, peroxisome proliferator activated receptor γ; 5-oxo-ETE, 5-oxo-6,8,11,14-eicosatetraenoic acid; BW245C, (4S)-(3-[(3R,S)-3-cyclohexyl-3-hydroxypropyl]-2,5-dioxo)-4-imidazolidineheptanoic acid; MBP, major basic protein; MDC, macrophage-derived chemokine (CCL22).
- Received October 14, 2004.
- Accepted December 6, 2004.
- The American Society for Pharmacology and Experimental Therapeutics