Molecular pharmacology of the human prostaglandin D₂ receptor, CRTH2

Abstract

1. The recombinant human prostaglandin D₂ (PGD₂) receptor, hCRTH2, has been expressed in HEK293(EBNA) and characterized with respect to radioligand binding and signal transduction properties. High and low affinity binding sites for PGD₂ were identified in the CRTH2 receptor population by saturation analysis with respective equilibrium dissociation constants (K_D) of 2.5 and 109 nM. This revealed that the affinity of PGD₂ for CRTH2 is eight times less than its affinity for the DP receptor.

2. Equilibrium competition binding assays revealed that of the compounds tested, only PGD₂ and several related metabolites bound with high affinity to CRTH2 (K_i values ranging from 2.4 to 34.0 nM) with the following rank order of potency: PGD₂>13,14-dihydro-15-keto PGD₂>15-deoxy-Δ^12,14-PGJ₂>PGJ₂>Δ¹²-PGJ₂>15(S)-15 methyl-PGD₂. This is in sharp contrast with the rank order of potency obtained at DP : PGD₂>PGJ₂>Δ¹²-PGJ₂>15-deoxy-Δ^12,14-PGJ₂ >>>13,14-dihydro-15-keto-PGD₂.

3. Functional studies demonstrated that PGD₂ activation of recombinant CRTH2 results in decrease of intracellular cAMP in a pertussis toxin-sensitive manner. Therefore, we showed that CRTH2 can functionally couple to the G-protein G_αi/o. PGD₂ and related metabolites were tested and their rank order of potency followed the results of the membrane binding assay.

4. By Northern blot analysis, we showed that, besides haemopoietic cells, CRTH2 is expressed in many other tissues such as brain, heart, thymus, spleen and various tissues of the digestive system. In addition, in situ hybridization studies revealed that CRTH2 mRNA is expressed in human eosinophils. Finally, radioligand binding studies demonstrated that two eosinophilic cell lines, butyric acid-differentiated HL-60 and AML 14.3D10, also endogenously express CRTH2.

Introduction

Prostaglandin D₂ (PGD₂) is a prostanoid derived from arachidonic acid via the action of PGHS I & II and specific PGD synthases (Narumiya et al., 1999). PGD₂ has been shown to bind and activate two G-protein-coupled receptors, CRTH2 (Chemoattractant Receptor-homologous molecule expressed on T-Helper type 2 cells) and DP (Boie et al., 1995; Hirai et al., 2001). Interestingly, the two PGD₂ receptors share little sequence homology. The highest amino acid sequence identity for CRTH2 is found with members of the leukocyte chemoattractant receptor subfamily which includes the FMLP receptor, C3a receptor and C5a receptor (Nagata et al., 1999). In fact, PGD₂ was reported to have chemotactic/chemokinetic activity on T-helper type 2 cells, eosinophils and basophils (Gervais et al., 2001; Hirai et al., 2001; Monneret et al., 2001). DP shares the highest sequence identity with the other prostanoid receptors, TP, FP, EP_1–4 and IP (Hirai et al., 2001). Through the activation of the DP receptor, PGD₂ has been implicated in different physiological events including sleep induction, mucus production, cell survival, control of intraocular pressure and allergic responses (Gervais et al., 2001; Matsumura et al., 1994; Matsuoka et al., 2000; Woodward et al., 1993; Wright et al., 1999). The physiological role of CRTH2 is not as well characterized. However, it has been demonstrated that its activation by PGD₂ can increase eosinophil and Th2 cell motility and can also modulate eosinophil morphology and degranulation (Gervais et al., 2001; Hirai et al., 2001; Monneret et al., 2001).

DP activation by PGD₂ leads to the stimulation of adenylyl cyclase activity and an increase in intracellular cAMP levels (i[cAMP]) in a G-protein G_αs-dependent manner (Boie et al., 1995). In contrast, CRTH2 intracellular signalling is less well defined. CRTH2 activation by PGD₂ does not lead to a stimulation of cAMP production (Hirai et al., 2001). On the other hand, activation of CRTH2 by PGD₂ has been shown to lead to a pertussis toxin-sensitive increase in intracellular calcium (i[Ca²⁺]) mobilization (Hirai et al., 2001). This toxin sensitivity suggests that CRTH2 couples to the G-protein G_αi/o class which results in inhibition of adenylyl cyclase activity and, as a consequence, decreases intracellular cAMP levels. However, CRTH2-mediated decreases in i[cAMP] has not yet been reported.

In contrast to DP, CRTH2 expression profile, signalling properties and pharmacology is poorly defined. Hirai et al. have established the expression of CRTH2 in various immune cells and its capacity to signal through increase in i[Ca²⁺]. They have also determined the affinity of PGD₂ (K_i=61±23 nM) and a limited number of ligands for the recombinant human CRTH2 using a whole cell radioligand binding assay (Hirai et al., 2001).

In the present study, we extensively characterized CRTH2 at the pharmacological level. We evaluated ligand : receptor interactions of recombinant CRTH2 using a membrane-based radioligand binding assay. The effect of CRTH2 activation on i[cAMP] and i[Ca²⁺] levels was also assessed. Finally, we determined CRTH2 expression profile by Northern blot analysis. This study provides new important information that should prove useful in understanding the role for the PGD₂ receptor CRTH2 in vivo.

Methods

Construction of pCEP4-hCRTH2 mammalian expression vector

RT–PCR was performed on human (h) eosinophil total RNA using oligo dT (Perkin Elmer RNA PCR core kit). The cDNA was amplified using Advantage cDNA PCR (Clontech) with gene specific primers containing restriction sites for cloning into pcDNA3.1neo(+) vector (Invitrogen). Sense primer 5′-GCATAAGGTACCATGTCGGCCAACGCCACACTG-3′. Reverse primer 5′-CGCTAGTCTAGACTAACTCGAGGTGCTGCTCA-3′. The nucleotide sequence of the cloned cDNA was identical to the previously published sequence (Genbank accession number AB008535) (Nagata et al., 1999). The human CRTH2 (hCRTH2) cDNA was subsequently subcloned in Kpn1/Nhe1 in the mammalian episomal expression vector pCEP4 for the generation of a cell line stably expressing CRTH2.

pCEP4-hCRTH2 expression in HEK293(EBNA) cells

The cell line HEK293(EBNA) was chosen based on its absence of endogenous expression of CRTH2. This was determined by in situ hybridization, [³H]-PGD₂ membrane binding assays and in functional cell-based assays (i.e changes in levels of cAMP and calcium following challenge with PGD₂) (data not shown; Figure 2C). HEK293(EBNA) cells were maintained in culture in a humidified atmosphere at 37°C (6% CO₂) in Dulbecco's modified Eagle's medium (DMEM) containing 10% heat-inactivated foetal bovine serum, 2 mM glutamine, 1 mM sodium pyruvate, 100 units ml⁻¹ of penicillin, 100 μg ml⁻¹ streptomycin and 250 μg (active) ml⁻¹ G418 (for EBNA selection). Stable expression of the hCRTH2 receptor in HEK293(EBNA) cells was achieved by transfection of the pCEP4-hCRTH2 plasmid using Lipofectamine-PLUS™ reagent according to the manufacturer's instructions (Invitrogen). Cells were maintained in culture for 48 h post-transfection and then grown in the presence of 250 μg ml⁻¹ hygromycin B (Calbiochem) for 2 weeks to select for resistant cells expressing the hCRTH2 receptor. Clonal cell lines were then generated by serial dilution. A clonal cell line, HEK-hCRTH2 no. 4, was selected for its high level of expression of recombinant hCRTH2 and the largest window of maximal inhibition of cAMP level achievable, as determined by receptor binding assays and cAMP functional assays, respectively.

Figure 2.

Inhibition of forskolin-stimulated i[cAMP] levels in HEK-hCRTH2 cells. (A) PGD₂ dose-response curve. (B) Pertussis toxin effect in the absence and presence of 100 nM PGD₂. Cells were treated overnight with 0 and 100 ng ml⁻¹ of PTX and harvested the next day for cAMP assay as described in the Methods. Curves for (A) and (B) were generated in the same experiment and are from 4–5 independent experiments where each data point was done in duplicate. Sigmoidal dose response curves were constructed and analysed as in Table 1 to determine the half maximum effective concentration, EC₅₀. (C) The release of i[Ca²⁺] was monitored over time by fluorometry in the cell lines indicated following a challenge with 100 nM of PGD₂ at time=10 sec. PTX was used at 100 ng ml⁻¹ to block the G_i/G_o-mediated response in HEK-CRTH2 cells. A representative experiment is shown.

Differentiation of HL60 cells and AML14.3D10 cells

HL-60/MF211 no. 7 is a pro-eosinophilic substrain of the human pro-myelocytic cell line that had been subcultured under alkaline conditions with the continuous selection of subtypes displaying eosinophilic properties when cultured with butyric acid (Scoggan et al., 1996). AML14.3D10 is an eosinophilic subline of the AML14 leukaemic cell line that does not require cytokine supplementation for eosinophilic differentiation (Paul et al., 1995). These cell lines were grown in a humidified atmosphere at 37°C (6% CO₂) in RPMI-1640 media (Invitrogen) supplemented with 10% heat-inactivated foetal bovine serum, 2 mM glutamine and 100 units ml⁻¹ of penicillin and 100 μg ml⁻¹ streptomycin. On day 1 of differentiation, cells were cultured at a cell density off 2×10⁵ cells ml⁻¹ and supplemented with 0.4 mM butyric acid. On day 4, the cells were diluted to 2×10⁵ cells ml⁻¹ and the concentration of butyric acid was adjusted to 0.4 mM. The cells were harvested on day 7.

Preparation of membranes

HEK-hCRTH2 cells were rinsed with phosphate buffered saline (PBS), detached using dissociation buffer (Specialty Media) and collected by centrifugation at 500 g for 10 min at 4°C. HL-60 and AML cells were collected by centrifugation for 15 min at 500 g at 4°C, washed with PBS and recovered by centrifugation. In all cases, the final cell pellet was resuspended in 10 mM HEPES/KOH pH 7.4, 1 mM EDTA, at approximately 10⁷ cells ml⁻¹ and frozen in liquid nitrogen or immediately processed for membrane preparation. The cells were disrupted by nitrogen cavitation (Parr) on ice (800 p.s.i. for 30 min) in the presence of protease inhibitors (2 mM AEBSF, 10 μM E-64, 100 μM leupeptin and 0.05 mg ml⁻¹ pepstatin). Cell membranes were isolated by differential centrifugation at 4°C, first at 1000×g for 10 min, then 160 000 g for 30 min. After centrifugation the pellet was resuspended in 10 mM HEPES/KOH pH 7.4, 1 mM EDTA, using Dounce homogenization (ten strokes), frozen in liquid nitrogen and stored at −80°C.

Radioligand binding assay

Radioligand binding assays were performed at room temperature in 10 mM HEPES/KOH pH 7.4, 1 mM EDTA containing 10 mM MnCl₂ and 0.4 nM [³H]PGD₂ (NEN, 172 Ci mmol⁻¹), in a final volume of 0.2 ml. Competing ligands (from Biomol and Cayman) were diluted in dimethylsulphoxide (Me₂SO) that was kept constant at 1% (v v⁻¹) of the final incubation volume. The reaction was initiated by the addition of 150 μg (for HL-60 no. 7 and AML14.3D10 cells) or 23 μg (for HEK-hCRTH2) of membrane protein. Total and non-specific binding were determined in the absence and the presence of 10 μM PGD₂, respectively. Under these conditions, specific binding (total minus non-specific) of the radioligand to the receptor reached equilibrium within 50 min and was stable up to 180 min. The reaction was routinely conducted for 60 min at room temperature and terminated by rapid filtration through GF/C filters (Brandel), prewetted in cold 10 mM HEPES/KOH pH 7.4, using a Brandel™ harvester (for HL-60 and AML cells) or through prewetted Unifilters GF/C (Packard), using a Tomtec MachIII semi-automated harvester (for HEK-hCRTH2). The filters were then washed with 4 ml of the same buffer and residual radioligand bound to the filter was determined by liquid scintillation counting following equilibration in 5 ml Ultima Gold™ (GF/C) or 50 μl Ultima Gold F™ (Unifilter) (Packard). Radioligand binding assays with HEK-hDP were performed as described (Abramovitz et al., 2000).

i[cAMP] measurements

HEK-hCRTH2 cells were grown to confluency on the day of the assay. The cells were washed with PBS, incubated for 3 min in cell dissociation buffer, harvested by centrifugation at 300 g for 6 min at room temperature and resuspended at 10⁶ cells ml⁻¹ in Hanks' balanced salt solution containing 25 mM HEPES pH 7.4 (HBSS/HEPES). The assay was performed in 0.2 ml HBSS/HEPES containing 5 μM forskolin (Sigma), 100 μM RO 20-1724 (Biomol) and 2 μl of test compound (from Biomol and Cayman). The reaction was initiated by the addition of 100 000 cells and left to proceed for 10 min at 37°C. The reaction was stopped by a 3 min incubation in a boiling water bath. The samples were centrifuged for 10 min at 500 g and the cAMP content in the supernatant was determined using a [¹²⁵I]-cAMP scintillation proximity assay (Amersham). Maximal inhibition of forskolin stimulated cAMP production by activation of CRTH2 was determined in the presence of 1 μM PGD₂. All compounds were prepared in Me₂SO kept constant at 1% (v v⁻¹) of the final incubation volume.

Construction of pCDNA3-G_α15 mammalian expression vector

G_α15 cDNA (Aurora Biosciences) was subcloned into the HindIII and XbaI sites of pCDNA3.1-Zeo(+) (Invitrogen). The resulting plasmid was transfected into HEK-hCRTH2 no. 4 using Lipofectamine™ reagent. Cells were selected in presence of 250 μg (active) ml⁻¹ G-418 and hygromycin B and 10 μg ml⁻¹ Zeocin (InVitrogen) (for G_α15 selection). The clonal cell line HEK-hCRTH2-G_α15 no. 14 was selected for further characterization based on release of intracellular calcium (i[Ca²⁺]) following PGD₂ challenge.

i[ Ca⁺²] measurements

Cells were seeded onto Biocoat poly-D lysine 96-well plates (Becton Dickinson) at 40 000 cells per well and incubated overnight in supplemented DMEM as indicated above. Cells were washed once with HBSS supplemented with 20 mM HEPES pH 7.4 and then incubated for 50 min at 37°C (6% CO₂) in the same buffer containing the cytoplasmic Ca⁺² indicator (Calcium assay kit; Molecular Devices). Changes in fluorescence were monitored before and after the addition of various agonists using a FLIPR™ (Molecular Devices) at λ_ex=488 nM and λ_em=540 nM.

Northern blot analysis

Human multiple tissue Northern (MTN™) blot I, III and IV (Clontech) were probed with a [³²P]-dCTP radiolabelled probe generated from full length CRTH2 using the T7 QuickPrime kit according to the manufacturer's specifications (Pharmacia). The blots were first incubated with 10 ml of ExpressHyb solution (Clontech) for 30 min at 68°C. Hybridization of the probe was carried out in ExpressHyb solution (1×10⁶ c.p.m. per ml) at 68°C overnight. Following hybridization, Northern blots were washed twice with 2×SSC-0.1% (w v⁻¹) SDS at room temperature for 10 min and then twice with 0.2×SSC-0.1% (w v⁻¹) SDS at 50°C for 20 min. Autoradiography was carried out overnight using MR Kodak film. β-actin was used as a control probe.

In situ hybridization

Freshly isolated eosinophils (2×10⁵ cells) were layered onto a poly-lysine coated glass slide by centrifugation (Cytospin). Cells were then fixed in 4% (v v⁻¹) paraformaldehyde solution prepared in PBS, pH 7.4, for 20 min at room temperature. The slides were then processed as follows: 2 min in 3×PBS, 2 min in 1×PBS twice and incubations of 5 min in 50, 70, 95 and 100% ethanol : water (v v⁻¹) solutions. Slides were air dried and stored at −80°C. The slides were equilibrated to room temperature and washed for 5 min in diethylpyrocarbonate (DEPC)-treated water and twice in PBS. The purity of the eosinophil population was assessed by means of flow cytometry (CELL-DYN 3700 system) on the basis of size, complexity, granularity and lobularity and found to be >95% pure. A CRTH2 cDNA fragment representing nucleotides 360 to 676 (corresponding to amino acids 120 to 225; Genbank accession no. AB008535) was amplified by PCR and subcloned into the PCR II dual promoter vector (Invitrogen). DIG-labelled riboprobes were synthesized using DIG-RNA labelling kit from Boehringer Mannheim. In situ hybridization was carried out as previously described (Wright et al., 1999). Colorimetric detection of the hybridized riboprobe was performed using an alkaline phosphatase-linked anti-DIG antibody (Boehringer Mannheim). All subsequent steps were carried out at room temperature. The slides were washed in detection buffer (DB; 100 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% (v v⁻¹) Tween), incubated with SuperBlock buffer (Biogenex) for 10 min and then incubated for 2 h with anti-DIG antibody (1 : 75 dilution) in DB and then washed three times in DB. The chromagen solution Fast Red (Sigma Biochemicals) was then added and the slides were left to incubate for 30 min. The reaction was stopped by washing in 10 mM Tris pH 8.0, 1 mM EDTA. The cells were mounted using SlowFade (Molecular Probes) and examined on a fluorescent microscope connected to a CCD camera.

Results

Saturation analyses of [³H]PGD₂ specific binding to HEK-hCRTH2

Radioligand binding assays to recombinant hCRTH2 were initially optimized with respect to divalent cation, pH dependence and incubation time as described in the Methods section. Under optimal conditions, [³H]PGD₂ binding to CRTH2 was at equilibrium in 50 min and stable for 3 h, and could be fully dissociated by excess unlabelled PGD₂. Saturation analysis of [³H]PGD₂ specific binding to hCRTH2 was then performed and revealed the presence of two populations of specific binding sites (Figure 1). Tritiated PGD₂ bound with high and low affinity to hCRTH2 with equilibrium dissociation constants (K_D) of 2.5±1.1 nM (n=3; mean±s.e.mean throughout) and 109±68.4 nM (n=3) respectively. The maximal number of high and low affinity binding sites (B_max) detected were 7.8±2.9 pmol mg⁻¹ (n=3) and 29.5±9.5 pmol mg⁻¹ (n=3) of membrane protein respectively.

Figure 1.

Saturation analysis of [³H]PGD₂ specific binding to recombinant HEK293(EBNA) cell membranes expressing hCRTH2. The radioligand binding assay was performed essentially as described in Methods. The incubation also contained increasing concentrations of radioligand from 0.2 to 77 nM, and 100 μM of unlabelled PGD₂ for non-specific determination. The experiment was conducted over 90 min. Inset: Non-linear transformation of the deduced specific binding isotherm showing the presence of two affinity binding sites. Analysis was done using GraphPad Prism (GraphPad Software Inc.). This is representative of three independent experiments where each data point was determined in duplicate.

Competition for [³H]PGD₂ specific binding to HEK-hCRTH2

Five groups of ligands were evaluated for their ability to compete with [³H]PGD₂ specific binding to HEK293(EBNA) cell membranes expressing hCRTH2 in equilibrium competition assays (Table 1). The first group comprised the preferred ligands for the prostanoid receptors: PGD₂ for DP, PGE₂ for the EP subtypes 1 to 4, PGF_2α for FP, the stable prostacyclin analogue iloprost for IP and the thromboxane mimetic U46619 for TP (Coleman et al., 1994). Out of all the prostanoid receptor preferred ligands, only PGD₂ competed with high affinity with [³H]PGD₂ for CRTH2 specific binding with a K_i value of 2.4±0.2 nM (n=17). The overall rank order of affinities for this group of ligands was: PGD₂>PGF_2α>U46619>PGE₂>iloprost. Since PGD₂ binds with the highest affinity to hCRTH2, we evaluated various PGD₂ metabolites. Most of the metabolites tested retain high affinity for CRTH2. Specifically, 13,14-dihydro-15-keto PGD₂ (DK-PGD₂) and 15-deoxy-Δ^12,14-PGJ₂ were comparable to PGD₂ in binding affinities to hCRTH2 with K_i values of 2.91±0.29 nM (n=10) and 3.15±0.32 nM (n=3), respectively. The third group of ligands consisted of selective compounds that are routinely used in pharmacological studies to classify the prostanoid receptors (Coleman et al., 1994), while the fourth group included various eicosanoid ligands. These ligands did not display appreciable affinity. It is important to note that DP selective agonists such as BW245C and L-644698 (Wright et al., 1998) were inactive at hCRTH2. Finally, the fifth group of compounds tested were a selection of NSAIDS. Only indomethacin showed high affinity (K_i=25.0±3.6 nM, n=4) for hCRTH2.

Table 1.

Affinities of eicosanoid ligands and related compounds for recombinant hCRTH2 and hDP expressed in HEK293(EBNA) cell membranes

Cyclic AMP functional assay with HEK-hCRTH2

The signal transduction properties resulting from activation of recombinant CRTH2 in HEK293(EBNA) were investigated. PGD₂ was unable to stimulate cAMP production (data not shown). PGD₂, however, reduces forskolin-elevated i[cAMP] levels in a dose-dependent manner with an EC₅₀=1.8 nM±0.4 nM (Figure 2A and Table 2, n=16). Activation of CRTH2 by PGD₂ reduced forskolin-stimulated i[cAMP] levels from 16.3±1.9 pmol to 3.0±0.5 pmol (n=16). Maximal potency was achieved with 100 nM PGD₂. There was no PGD₂-mediated reduction in i[cAMP] in forskolin-treated HEK293(EBNA) wild-type cells (data not shown). Moreover, the decreased cAMP response mediated by 100 nM of PGD₂ was reduced in the presence of 100 ng ml⁻¹ pertussis toxin (PTX) (Figure 2B). This suggests that CRTH2 couples through the PTX-sensitive G_αi/o class of G-proteins. Additional prostanoid-like ligands with high affinity for hCRTH2 were tested for their potency to stimulate the receptor to reduce forskolin-stimulated i[cAMP]. All the ligands shown in Table 2 are full agonists at hCRTH2, as compared with PGD₂. Overall, the EC₅₀ for these ligands followed the same rank order as their affinities determined in radioligand binding assays, with the highest potencies observed for PGD₂ and its metabolites. Indomethacin was the only NSAID tested in the radioligand binding assay that had high affinity for CRTH2. Indomethacin was also a full agonist at hCRTH2 and decreased forskolin-stimulated i[cAMP] with an EC₅₀ of 14.9±4.9 nM (n=3). The other NSAIDS tested in membrane binding assays were unable to efficiently reduce cAMP under the same assay conditions (data not shown).

Table 2.

Determination of agonist potency in inhibiting i[cAMP] production and stimulating i[Ca²⁺] release in HEK-hCRTH2 and HEK-hCRTH2-G_α15 cells

Calcium functional assay with HEK-hCRTH2-Gα₁₅

We also investigated whether activation of CRTH2 resulted in mobilization of i[Ca²⁺]. PGD₂ stimulated a slight increase in intracellular calcium in HEK-hCRTH2 cells compared to the HEK parental cell line (Figure 2C). This increase in intracellular calcium was fully inhibited by 100 ng ml⁻¹ of PTX. The release of i[Ca²⁺] provoked by PGD₂ was dramatically increased in HEK-hCRTH2 cells stably transfected with the promiscuous G-protein G_α15. Using the HEK-hCRTH2-G_α15 cell line, the EC₅₀ value for PGD₂ in mobilizing i[Ca²⁺] was determined as 22.1±4.4 nM (n=6, Table 2). This is approximately 10 fold higher than the EC₅₀ determined for reduction of i[cAMP]. A limited number of CRTH2 ligands were evaluated for their ability to mobilize i[Ca²⁺]. These ligands were generally 10 fold less potent in this paradigm compared with their effect on i[cAMP], in agreement with the result observed with PGD₂. However, the rank order of potency observed in the calcium assay was consistent with the ligand binding data (see Table 1).

Tissue distribution of hCRTH2

The tissue distribution of CRTH2 was determined by Northern blot analysis on RNA from 23 different human tissues. A single transcript of approximately 3.5 Kb was detected in a variety of tissues (Figure 3). Higher levels of CRTH2 expression were seen in stomach, small intestine, heart and thymus. Intermediate levels were detected in colon, spinal cord and peripheral blood. Lower levels were observed in brain, skeletal muscle and spleen.

Figure 3.

Distribution of CRTH2 mRNA in human tissues. A Northern blot of total human RNA from adult tissues shows an approximately 3.5-kb transcript hybridized with the CRTH2 probe (upper panel). The lower panel represents the same blot hybridized with a β-actin probe to confirm RNA integrity and relative loading. Molecular markers are indicated in kb on the right.

Detection of CRTH2 mRNA in human eosinophils by in situ hybridization

It was previously shown, by RT–PCR, that human eosinophils express CRTH2 (Gervais et al., 2001; Hirai et al., 2001). In situ hybridization studies were performed on purified eosinophils to ensure that CRTH2 mRNA detected by means of RT–PCR was not produced by a non-eosinophil contaminating cell type. Human eosinophils (Figure 4a) but not neutrophils (Figure 4c) were positively stained with a CRTH2-specific antisense riboprobe. Staining of eosinophils was not observed with the negative control sense riboprobe (Figure 4b).

Figure 4.

In situ hybridization of hCRTH2 in human eosinophils (a and b) and granulocytes (c and d). CRTH2 mRNA was revealed using CRTH2-specific antisense probe (a and c). A complementary sense probe was used as a negative control (b and d).

Expression of hCRTH2 on eosinophilic cell lines

The eosinophilic cell lines HL-60/MF211 no. 7 (Scoggan et al., 1996) and AML14.3D10 (Paul et al., 1995) were also evaluated for expression of CRTH2. Membranes prepared from HL-60 no. 7 showed negligible binding to [³H]PGD₂ (Figure 5a). However, membranes prepared from AML14.3D10 showed significant binding to [³H]PGD₂. Binding of [³H]PGD₂ to membranes from both cell lines was increased by differentiation in the presence of butyric acid, 800 fold and 3 fold for HL-60 no. 7 and AML14.3D10 cells, respectively. Competition assays revealed that DK-PGD₂, a CRTH2-selective ligand, but not BW245C, a DP-selective ligand, efficiently competed for [³H]PGD₂ from the eosinophilic cell line membranes (Figure 5b). These results indicate that CRTH2 is predominantly expressed by AML14.3D10 and that the expression of this receptor by both AML14.3D10 and HL-60/MF211 no. 7 can be significantly increased in the presence of the pro-eosinophilic differentiating agent, butyric acid. These cell lines represent valuable sources of endogenous CRTH2.

Figure 5.

Tritiated PGD₂ specific binding to HL-60 no. 7 and AML14.3D10 eosinophilic cell lines. (A) Specific radioligand binding on HL-60 no. 7 and AML14.3D10 cell membranes before (−) and after (+) treatment with the eosinophilic differentiating agent butyric acid at 0.4 mM. Results are from three independent experiments performed in duplicate. (B) Specific radioligand binding to butyric acid-treated HL-60 no. 7 and untreated AML14.3D10 cell membranes in the presence of either vehicle (Me₂SO) only (control), unlabelled PGD₂, the DP-selective ligand BW245C and the CRTH2-selective ligand DK-PGD₂ at 1 μM. The results are expressed as a percentage of the maximal [³H]PGD₂ specific binding observed in the absence of unlabelled competitor (control). Results are from three independent experiments done in duplicate.

Discussion

We have characterized pharmacologically the recombinant human CRTH2 on HEK293(EBNA) cells. Two populations of hCRTH2 specific binding sites for PGD₂ were identified by saturation analysis: a high and low affinity site both expressed at high levels (Figure 1). The affinity of PGD₂ for hCRTH2 high affinity binding sites is significantly lower than the affinity of PGD₂ for hDP high affinity binding sites as determined in a similar recombinant system, with equilibrium dissociation constants (K_D) of 2.5 and 0.3 nM for hCRTH2 and hDP, respectively. The affinity of PGD₂ for CRTH2 low affinity binding sites is also significantly lower than for DP low affinity binding sites, with K_D values of 109 nM and 13 nM for hCRTH2 and hDP, respectively (Figure 1 and Wright et al., 1998). Saturation analysis reveals, therefore, that the affinity of PGD₂ for CRTH2 is approximately eight times inferior to its affinity for DP.

Equilibrium competition binding assays revealed that, of all compounds tested, only PGD₂, several of its metabolites and the NSAID indomethacin bind efficiently to CRTH2 with K_i values in the low nanomolar range and the following rank order of affinities: PGD₂⩾DK-PGD₂⩾15-deoxy-Δ^12,14-PGJ₂>PGJ₂⩾Δ¹²-PGJ₂>indomethacin⩾15(s)-15 methyl-PGD₂. This is in sharp contrast with the rank order of affinities obtained for PGD₂ at DP: PGD₂>PGJ₂>Δ¹²-PGJ₂>15-deoxy-Δ^12,14-PGJ₂>>>DK-PGD₂ (Wright et al., 1998). DK-PGD₂ has previously been shown to be selective for CRTH2 over DP with K_i values of 160 and >30 000 nM for CRTH2 and DP, respectively, as determined in a whole cell binding assay (Hirai et al., 2001). In our membrane-based binding assay, DK-PGD₂ shows a 2000 fold selectivity for CRTH2 over DP (Table 1). Whether this metabolite is stable in vivo and exerts important physiological effects remains to be determined. However, it represents a valuable tool to evaluate the role of endogenously expressed CRTH2. 15-deoxy-Δ^12,14-PGJ₂ is another PGD₂ metabolite which preferentially binds CRTH2 over DP with 89 fold selectivity. It is a high affinity ligand for the nuclear receptor PPARγ through which it exerts anti-inflammatory effects (Straus & Glass, 2001). Recently generation of this metabolite has been shown in vivo during inflammatory processes using immunohistochemical methods (Shibata et al., 2002). It has also been shown to mediate both pro- and anti-inflammatory effects independently of PPARγ (Harris et al., 2002; Straus & Glass, 2001). Whether these effects are mediated through CRTH2 needs to be established. Finally, in contrast to other NSAIDS tested, indomethacin is a high affinity and potent agonist at hCRTH2, approximately 10 fold less potent than PGD₂ itself. Indomethacin was previously reported to have potent agonist effects on hCRTH2 in both recombinant and naturally expressing cell lines and it was suggested that some of the therapeutic and/or adverse effects of indomethacin could be mediated through CRTH2. Gastrointestinal injury is one example of such indomethacin-induced adverse effects (Shen & Winter, 1977). An early stage of indomethacin-induced intestinal injury was shown to be the infiltration and degranulation of eosinophils in the epithelium (Anthony et al., 1993). It is possible that agonism of CRTH2 by indomethacin contribute to the inflammation of the gastrointestinal tract through the recruitment and activation of inflammatory cells such as eosinophils. However, gastrointestinal toxicity is a property associated with a number of NSAIDs that are not ligands for CRTH2. The use of CRTH2-selective antagonists or CRTH2-deficient mice will be useful in evaluating the potential contribution of CRTH2 to this side effect of indomethacin. Hirai et al. (2002) reported that the K_i value for indomethacin at CRTH2 was 8.1 μM which differs significantly from the K_i of 20 nM that we determined (Table 1). A possible explanation for this discrepancy could be that Hirai et al. determined the binding affinity of PGD₂ using a cell-based binding assay in which both the radioligand and the competitor could potentially be internalized.

Receptors coupled to the PTX-sensitive G_αi/o subunit typically decrease i[cAMP] through the inhibition of adenylyl cyclase. We have demonstrated that CRTH2 activation can lead to a pertussis toxin-sensitive decrease in i[cAMP] and, therefore, presumably couples to the G_αi/o subunit. We have also observed a PTX-sensitive mobilization of i[Ca²⁺] in HEK-CRTH2 upon PGD₂ challenge. This slight mobilization of i[Ca²⁺] may be generated by G_βγ subunits that have previously been shown to activate phospholipase C when released from G_i-coupled receptors (Smrcka et al., 1993). A more robust mobilization of i[Ca²⁺] was achieved when the promiscuous G_α15 subunit was co-expressed with CRTH2. G_α15 couples to many receptors that are not activators of the G_αq subunit and allows these receptors to strongly activate the phospholipase C/calcium pathway (Offermans et al., 1995; Zhu et al., 1996). The EC₅₀ values for the various ligands tested for mobilization of i[Ca²⁺] are approximately one log unit higher than the values obtained for these ligands in reducing i[cAMP]. This probably reflects a less efficient coupling of CRTH2 to the transfected G_α15 compared to endogenous G_αi. However, the overall rank order of potencies of the various agonists tested for mobilization of i[Ca²⁺] is comparable to the one obtained for reducing i[cAMP]. The coupling to G_α15 may prove useful for calcium-based high-throughput screening for CRTH2 ligands.

Our Northern blot analysis revealed that the expression of CRTH2 is not limited to cells of haemopoietic origin such as Th2 cells, eosinophils and basophils. Similarly to DP, CRTH2 is expressed in tissues of haemopoietic origin, but also in the digestive system and in the central nervous system. Data suggests that PGD₂ might differentially affect both receptors when expressed in the same tissue. For example, PGD₂ has been shown to produce both increases and decreases in short circuit current across the canine proximal colon. Interestingly, the DP selective agonist BW245C only produced increases in current while the CRTH2 selective agonist DK-PGD₂ elicited only decreases (Rangachari et al., 1995). We have previously shown that PGD₂ could increase or decrease the extent of eosinophil apoptotic cell death depending on the blood donor. Again, a DP-selective agonist increased eosinophil survival while a CRTH2-selective agonist mainly decreased it (Gervais et al., 2001). This suggests that DP and CRTH2 can antagonize each other in a given tissue. This agrees with the data showing that activation of CRTH2 can lead to a decrease in i[cAMP] whereas DP stimulation activates adenylyl cyclase and causes an accumulation of i[cAMP] in the cell (Boie et al., 1995). Thus, for a given cell type, the relative abundance of functional CRTH2 and DP might determine how this cell responds to PGD₂.

PGD2 has been reported to be a potent coronary vasoconstrictor agent in guinea pig isolated heart (Schror, 1978). Interestingly, CRTH2 mRNA was highly expressed in heart tissues (Figure 3), although the cell type expressing CRTH2 in the heart has not been identified and the physiological role for this receptor in the heart has not been defined. It would be of interest to determine whether the vasoconstrictor effect of PGD₂ on heart is mediated through CRTH2.

In summary, we have characterized the ligand binding and signal transduction properties of CRTH2, a novel PGD₂ receptor (also known as DP2). We have shown that CRTH2, a chemokine family member, has distinct pharmacological properties when compared to the classic PGD₂ receptor DP, a prostanoid family member, with respect to both ligand interactions and cell signalling. These data bring forward the intriguing possibility that these two receptors will have distinct and interactive functions when regulating PGD₂ effects in vivo.

Abbreviations

13,14-dihydro-15-keto-PGD₂

DK-PGD₂

cAMP

cyclic AMP

GPCR

G-protein-coupled receptor

HEK

human embryonic kidney

i[cAMP]

intracellular cyclic AMP

i[Ca²⁺]

intracellular calcium

NSAID

non-steroidal anti-inflammatory drug

PGD₂

prostaglandin D₂

PTX

pertussis toxin