Abstract

Mast cell stabilizers, including disodium cromoglycate (DSCG), were found to have potential as the agonists of an orphan G protein–coupled receptor, GPR35, although it remains to be determined whether GPR35 is expressed in mast cells and involved in suppression of mast cell degranulation. Our purpose in this study is to verify the expression of GPR35 in mast cells and to clarify how GPR35 modulates the degranulation. We explored the roles of GPR35 using an expression system, a mast cell line constitutively expressing rat GPR35, peritoneal mast cells, and bone marrow–derived cultured mast cells. Immediate allergic responses were assessed using the IgE-mediated passive cutaneous anaphylaxis (PCA) model. Various known GPR35 agonists, including DSCG and newly designed compounds, suppressed IgE-mediated degranulation. GPR35 was expressed in mature mast cells but not in immature bone marrow–derived cultured mast cells and the rat mast cell line. Degranulation induced by antigens was significantly downmodulated in the mast cell line stably expressing GPR35. A GPR35 agonist, zaprinast, induced a transient activation of RhoA and a transient decrease in the amount of filamentous actin. GPR35 agonists suppressed the PCA responses in the wild-type mice but not in the GPR35−/− mice. These findings suggest that GPR35 should prevent mast cells from undergoing degranulation induced by IgE-mediated antigen stimulation and be the primary target of mast cell stabilizers.

SIGNIFICANCE STATEMENT The agonists of an orphan G protein–coupled receptor, GPR35, including disodium cromoglycate, were found to suppress degranulation of rat and mouse mature mast cells, and their antiallergic effects were abrogated in the GPR35−/− mice, indicating that the primary target of mast cell stabilizers should be GPR35.

Introduction

Mast cells play a pivotal role in IgE-mediated immediate responses, including allergic asthma (Galli and Tsai, 2012). Allergens activate tissue mast cells by crosslinking of FcεRI, which is mediated by the allergen-specific IgE. Stimulated mast cells produce a variety of proinflammatory mediators, including histamine, which is released through degranulation and acts on the H1 receptors of the endothelial cells, triggering the plasma extravasation. Disodium cromoglycate (DSCG) has long been used as a therapeutic agent, of which mode of action is different from the other antiallergic drugs (Cox, 1967). Although accumulating evidence suggests that DSCG should have a variety of anti-inflammatory actions, the primary target of DSCG might be mast cells (Sinniah et al., 2017). Using human lung and rat peritoneal mast cells, DSCG and its related compounds were found to suppress degranulation and were categorized into mast cell stabilizers. Previous studies demonstrated that DSCG could suppress degranulation induced by antigen stimulation and by cationic secretagogues (Assem and Richter, 1971; Orr et al., 1971; Leung et al., 1988; Jeong et al., 2006). Because DSCG was found to be a drug with a large safety margin, its target molecule(s) has been intensively explored to develop better therapeutic approaches.

In 2010, DSCG and its related compounds were reported to have potential as the agonist of rat and human GPR35, an orphan G protein–coupled receptor (GPCR), using the expression systems (Jenkins et al., 2010; Yang et al., 2010). Yang et al. (2010) reported that GPR35 was expressed in human cultured mast cells derived from the hematopoietic stem cells and was upregulated during the sensitization with IgE. These findings raised a hypothesis that mast cell stabilizers including DSCG should suppress the degranulation of mast cells by acting on GPR35 (Sinniah et al., 2017).

GPR35 was first cloned as a novel GPCR, which is abundantly expressed in the intestine (O’Dowd et al., 1998). It was postulated that a tryptophan metabolite, kynurenic acid (KA), served as a natural ligand of GPR35 (Wang et al., 2006). They demonstrated that GPR35 could be coupled with Gi/o and expressed in human leukocytes and gastrointestinal tissues. However, several concerns have been raised about the concept that KA is an endogenous ligand because the affinity of KA for GPR35 was quite low and the maximal activity was obtained at relatively high concentrations (∼10−3 M). Lodoxamide, another mast cell stabilizer, was found to be a potent agonist of GPR35 and to induce the coupling with G12/13 in addition to Gi/o in the expression system (Park et al., 2018). A 3D conformation of G13-coupled human GPR35 bound to lodoxamide has been recently demonstrated using cryo-electron microscopy (Duan et al., 2022).

We verified here whether GPR35 is involved in the pharmacological actions of mast cell stabilizers. Our findings suggest that GPR35 should be expressed in mature mast cells and prevent them from undergoing degranulation in response to antigen stimulation.

MethodsAnimals

Specific-pathogen–free, 8-week-old male BALB/cCrSlc mice, 8-week-old C57BL/6NCrSlc mice, and 8- to 10-week-old male Wistar rats were supplied by Japan SLC (Hamamatsu, Japan). The mutant mice that lack functional GPR35 were generated by electrophoretic injection of two CRISPR RNAs (5′-CCG CAU CUA UAU GAC CAA CC-3′ and 5′-UAG UGA CUG UGC UAU CGA GA-3′) and Alt-R Cas9 endonuclease (Integrated DNA Technologies, Coralville, IA) into fertilized eggs of a C57BL/6NCrSlc mouse using NEPA21 (NEPA GENE, Ichikawa, Japan). The established strain was found to have a truncated GPR35 gene, which encodes the amino-terminal 39 residues of GPR35 and the segment generated by a frameshift (NH2-GQP LMW GRQ RPP KRP PTW SGP TWL CLS SAS CPC MWS-COOH). Animal use was approved by the Animal Care and Use Committee of Okayama University (OKU-2018178 and OKU-2018274) and by the Committee on the Ethics of Animal Research of Kyoto Pharmaceutical University (A22-034 and PCOL-20-007), conforming to the Guidelines for the Proper Conduct of Animal Experiments of Science Council of Japan and the Policy on the Care and Use of Laboratory Animals of Okayama University and Kyoto Pharmaceutical University.

Materials

The following materials were purchased from the sources indicated: compound 48/80, p-nitrophenyl-β-D-2-acetoamide-2-deoxyglucopyranoside, kynurenic acid (KA), zaprinast, ketotifen fumarate, pamoic acid, histamine dihydrochloride, oleoyl-L-α-lysophosphatidic acid, 5-hydroxyindole-3-acetic acid (5-HIAA), dinitrophenyl human serum albumin (DNP-HSA), an anti-dinitrophenyl IgE antibody (clone SPE-7), and Histodenz from Sigma-Aldrich (St. Louis, MO); disodium cromoglycate (DSCG), Evans blue, toluidine blue, and Safranin-O from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan); PEI: Polyethylenimine “Max” from Polysciences (Warrington, PA); pertussis toxin from Bordetella pertussis from List Biologic Laboratories (Campbell, CA); an fluorescein isothiocyanate (FITC)-conjugated anti-mouse IgE antibody (553415), a phycoerythrin (PE)-conjugated anti-mouse c-kit antibody (553355), an anti-trinitrophenyl IgE antibody (clone IgE-3), and an anti-mouse CD16/CD32 antibody (clone 2.4G2) from BD Biosciences (San Diego, CA); trinitrophenyl bovine serum albumin (TNP-BSA) from LSL (Tokyo, Japan); thapsigargin from Merck Millipore (Billerica, MA); 1-oleoyl-2-hydroxy-sn-glycero-3-[phospho-L-serine] (lysophosphatidylserine) from Avanti Polar Lipids, Inc. (Alabaster, AL); recombinant mouse stem cell factor, Hoechst 33342, and an anti-hemagglutinin (HA) antibody (clone HA124) from Nacalai Tesque (Kyoto, Japan); a PE-conjugated F(ab’)2-goat anti-mouse IgG (H+L) antibody (12-4010-82) from Thermo Fisher Scientific (Waltham, MA); phalloidin-iFluor 488 from Abcam (Cambridge, UK); and recombinant mouse interleukin 3 (IL-3) from R&D Systems (Minneapolis, MN). All other chemicals were commercially available and of reagent grade.

Purification of Rat Peritoneal Mast Cells

Male Wistar rats were euthanized by cervical dislocation followed by a peritoneal injection with 20 ml/capita Tyrode-HEPES buffer (10 mM HEPES-NaOH, pH 7.3, containing 137 mM NaCl, 2.7 mM KCl, 0.41 mM NaH2PO4, 1.6 mM CaCl2, 1 mM MgCl2, and 5.6 mM glucose) containing 0.05% gelatin. The peritoneal cells were harvested from the peritoneal lavage and resuspended in Tyrode-HEPES buffer containing 0.05% gelatin. The cell suspension was layered onto 22.5% Histodenz solution (10 mM HEPES-NaOH, pH 7.3, containing 137 mM NaCl, 2.7 mM KCl, 0,41 mM NaH2PO4, 1.6 mM CaCl2, 1 mM MgCl2, 5.6 mM glucose, 1% gelatin, and 0.29 g/ml Histodenz) and centrifuged at 45 × g for 15 minutes at room temperature. After careful aspiration of the upper layers of cells, the precipitated cells were collected and resuspended in PIPES buffer (25 mM PIPES-NaOH, pH 7.4, containing 125 mM NaCl, 2.7 mM KCl, 5.6 mM glucose, 1 mM CaCl2, and 0.1% bovine serum albumin). Greater than 90% of the cells were positively stained with Safranin-O.

Cell Culture

A human embryonic kidney cell line, HEK293 (Japanese Collection of Research Bioresources Cell Bank, Ibaraki, Japan; < 20 passage number of cells were used), and HEK293/ΔGαq/11/12/13 cells (described below) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% heat-inactivated fetal bovine serum, 100 U/ml penicillin, and 0.1 mg/ml streptomycin. A rat mast cell line, RBL-2H3 (Japanese Collection of Research Bioresources Cell Bank), was transfected with pApuro (a generous gift from Dr. Tomohiro Kurosaki, Osaka University) (Mock) or pApuro containing cDNA encoding rat GPR35 with hemagglutinin tag at its amino terminus using PEI: Polyethyleneimine “Max,” and a permanent clone was established respectively. The surface expression levels of rat GPR35 were determined using Gallios (Beckman Coulter, Brea, CA) with a PE-conjugated F(ab’)2-goat anti-mouse IgG (H+L) antibody (described below). These clones were cultured in RPMI-1640 medium containing 10% heat-inactivated fetal bovine serum, 100U/ml penicillin, 0.1 mg/ml streptomycin, and 1 μg/ml puromycin. The cell culture was performed at 37°C in the presence of 5% CO2 in a fully humidified atmosphere. Murine IL-3–dependent bone marrow–derived cultured mast cells (BMMCs) were prepared as previously described (Takano et al., 2008). Briefly, BMMCs were prepared by culturing the bone marrow cells in the presence of 10 ng/ml IL-3 for 1 month. A majority of the cells (>95%) were found to be in a single c-kit+FcεRI+ population. Connective tissue-type mast cell–like cultured mast cells [CTMC-like mast cells (MCs)] were obtained by coculturing BMMCs in the presence of 100 ng/ml c-kit ligand and Swiss 3T3 fibroblasts for 16 days. Greater than 95% was found to be positively stained by Safranin-O.

Degranulation of Mast Cells

Mast cells were first suspended in PIPES buffer and stimulated for 30 minutes at 37°C in the presence of various compounds. In case of antigen-induced degranulation, purified rat peritoneal mast cells were sensitized with 10 μg/ml anti-DNP IgE (clone SPE-7) in Tyrode-HEPES buffer containing 0.1% bovine serum albumin at 37°C for 3 hours and then stimulated with 100 ng/ml DNP-HSA and 2 μM lysophosphatidylserine. RBL-2H3 cells sensitized with an anti-DNP IgE (clone SPE-7, 50 ng/ml) in RPMI-1640 medium described above at 37°C for 3 hours were stimulated with DNP-HSA in PIPES buffer. For BMMCs and CTMC-like MCs, they were sensitized with an anti-TNP IgE (clone IgE-3, 1 μg/ml) for 3 hours and followed by stimulation with TNP-BSA (10 ng/ml). After stimulation, aliquots were centrifuged at 800 × g for 10 minutes at 4°C. The supernatant fractions were collected, and the remaining cells were lysed in PIPES buffer containing 0.5% Triton X-100. The lysates were centrifuged at 15,000 × g for 10 minutes at 4°C to extract the cellular fractions. The reaction of β-hexosaminidase was performed at 37°C for 30 minutes in 67 mM sodium citrate, pH 4.5, containing 3.3 mM p-nitrophenyl-β-D-2-acetoamide-2-deoxyglucopyranoside. The enzymatic reaction was terminated by adding 1.22 volume of 250 mM glycine-NaOH, pH 10.7. The amount of liberated p-nitrophenol was calculated by measuring OD405. The levels of degranulation were expressed as the ratio of β-hexosaminidase activity in the supernatant fractions and the sums of the supernatant and cellular fractions (total activities).

Quantitative PCR Analysis

Messenger RNA levels of Gpr35 were determined by quantitative reverse-transcription polymerase chain reaction (RT-PCR). Total RNAs were extracted using QIAGEN RNeasy Kit (QIAGEN, Hilden, Germany), treated with DNase, and reverse transcribed using Takara Bio PrimeScript RT Reagent Kit (Takara Bio Inc. Kusatsu, Japan). PCR was performed using StepOnePlus (Thermo Fisher Scientific, Waltham, MA) with KOD SYBR qPCR Mix (TOYOBO, Osaka, Japan) in the presence of the specific primer pairs for rat Gpr35 (forward: 5′-GGA AAC ATC TTC AGC CGT GC-3′; reverse: 5′-ATC TTG GCT CTT GTG GGG TG-3′); mouse Gpr35 (5′- ACA ACC TGT AAC AGC ACC CTC -3′, 5′- GCG ATA GCA GAA TAC CCA GAG T -3′); rat Gapdh (5′-TGA ACG GGA AGC TCA CTG G -3′, 5′- TCC ACC ACC CTG TTG CTG TA -3′); and mouse Actb (5′-CAT CCG TAA AGA CCT CTA TGC CAA C-3′, 5′-ATG GAG CCA CCG ATC CAC A-3′). Forty cycles of the reaction were performed for rat Gpr35 and Gapdh (each cycle consisted of three conditions: 98°C 10 seconds, 60°C 10 seconds, and 68°C 30 seconds); for mouse Gpr35 (each cycle consisted of three conditions: 98°C 10 seconds, 60°C 10 seconds, and 68°C 1 minute); and for mouse Actb (each cycle consisted of two conditions: 95°C 15 seconds and 60°C 1 minute).

It is difficult to precisely compare the mRNA expression levels among different kinds of cells and tissues. The expression levels of GPR35 were quantified as fold changes relative to those in the small intestine, in which GPR35 was found to be abundantly expressed.

Transforming Growth Factor Alpha Shedding Assay

Transforming growth factor alpha (TGF-α) shedding assay was carried out as previously described (Inoue et al., 2012). Rat, human, and mouse GPR35 cDNAs were cloned into pCAGGS vector and subjected to the assay. In verification of G protein coupling, HEK293/ΔGαq/11/12/13 cells in which the Gαq, Gα11, Gα12, and Gα13 genes were deleted using the CRISPR/Cas9 method were used instead of parental HEK293 cells because TGF-α shedding should occur when exogenously expressed GPCRs were coupled with the endogenously expressed Gαq, Gα11, Gα12, and Gα13 in HEK293 cells (Devost et al., 2017). Briefly, HEK293 and HEK293/ΔGαq/11/12/13 were transiently transfected with a series of expression vectors using PEI: Polyethyleneimine “Max.” Twenty-four hours after the transfection, they were stimulated with various GPR35 agonists in Hank’s balanced salt solution containing 5 mM HEPES-NaOH, pH 7.4, for 30 minutes at 37°C. Cellular and extracellular fractions were separated and incubated in 40 mM Tris-HCl, pH 9.5, containing 40 mM NaCl, 10 mM MgCl2, and 10 mM p-nitrophenylphosphate at 37°C for 1 hour. The enzymatic activity was determined based on the values of OD405.

Synthesis of a Series of GPR35 Agonists

1H nuclear magnetic resonance (NMR) (400 MHz) spectra were recorded with JEOL JNM-ECZ400S (JEOL, Akishima, Japan) or JEOL HNM-ECX400 NMR spectrometers in the indicated solvent. Electron ionization-mass spectrometry (EI-MS) spectra were recorded on a JEOL JMS-GCmate II. The purities of compounds were determined based on the results of high-performance liquid chromatography (HPLC) analysis on Shimadzu LC-20AD separations module with SPD-20A photodiode array detector (Shimadzu, Kyoto, Japan). The sample was applied on a YMC-Triart C18 column (2.1 mm × 50 mm, 5 μm) and eluted at 1 ml/min for 0.5 minutes (20% B), then 5 minutes gradient (20% B to 100% B), and then maintaining the final condition for 3.5 minutes, where solvent A was water (0.1% H3PO4 buffer) and solvent B was acetonitrile.

6-(3-Carboxyphenyl)-8-(4-Methoxybenzamido)-4-Oxo-4H-Chromene-2-Carboxylic Acid (KGP-7).

A suspension of methyl 6-bromo-8-(4-methoxybenzamido)-4-oxo-4H-chromene-2-carboxylate (Funke et al., 2013) (300 mg, 0.728 mmol); PdCl2 (dppf) (61 mg, 0.084 mmol); 3-carboxyphenylboronic acid (209 mg, 1.26 mmol); and K3PO4 (309 mg, 1.46 mmol) in 1,4-dioxane (6.5 ml) and DMF (2 ml) was refluxed under Ar atmosphere overnight. After removal of solvent, the residue was partitioned with CHCl3-sat.NH4Claq. The organic phase was dried over Na2SO4 and evaporated. The resulting residue was purified with silica gel column chromatography (CHCl3/MeOH = 75/1), providing 3-[8-(4-methoxybenzamido)-2-methoxycarbonyl-4-oxo-4H-chromen-6-yl]benzoic acid (164 mg, 0.346 mmol) as brown solid in 48% yield. A suspension of 3-[8-(4-methoxybenzamido)-2-methoxycarbonyl-4-oxo-4H-chromen-6-yl]benzoic acid (123 mg, 0.260 mmol) and LiI (312 mg, 2.33 mmol) in THF (4.0 ml) was heated for 25 hours at 100°C under microwave irradiation. After the reaction mixture was evaporated, 1M HClaq was slowly added to the residue. The resulting precipitate was collected, washed with water, dried, and crystallized with MeOH, providing the title compound (106 mg, 0.231 mmol) as pale yellow solid in 89% yield. 1H-NMR (400 MHz, DMSO-d6) δ: 10.22 [1H, broad singlet (brs)]; 8.48 (1H, d, J = 2.0 Hz); 8.27 (1H, s); 8.13 (1H, d, J = 2.4 Hz); 8.06 (2H, d, J = 8.7 Hz); 8.01 (1H, d, J = 8.3 Hz); 7.96 (1H, d, J = 8.3 Hz); 7.67 (1H, t, J = 7.9 Hz); 7.12 (2H, d, J = 9.1 Hz); 7.00 (1H, s); 3.87 (3H, s). EI-MS m/z: 459 [M]+. Purity by HPLC: 97%, RT 4.78 minutes.

6-(3-Hydroxyphenyl)-8-(4-Methoxybenzamido)-4-Oxo-4H-Chromene-2-Carboxylic Acid (KGP-18).

The target compound was obtained in the same manner as 6-(3-carboxyphenyl)-8-(4-methoxybenzamido)-4-oxo-4H-chromene-2-carboxylic acid except that the corresponding 3-hydroxyphenylboronic acid was used instead of 3-carboxyphenylboronic acid. Yield 54% (two steps). 1H-NMR (400 MHz, DMSO-d6) δ: 10.01 (1H, brs); 9.53 (1H, brs); 8.42 (1H, d, J = 2.0 Hz); 8.04 (2H, d, J = 8.8 Hz); 8.01 (1H, d, J = 2.4 Hz); 7.31 (1H, d, J = 7.6 Hz); 7.16 (1H, m); 7.14 (1H, d, J = 2.4 Hz); 7.10 (2H, d, J = 8.4 Hz); 6.95 (1H, s); 6.84 (1H, dd, J = 8.0 Hz, 2.4 Hz); 3.87 (3H, s). EI-MS (m/z): 431 [M]+. Purity by HPLC: 97%, RT 4.77 minutes.

6-(2-Carboxyphenyl)-8-(4-Methoxybenzamido)-4-Oxo-4H-Chromene-2-Carboxylic Acid (KGP-20).

The target compound was obtained in the same manner as 6-(3-carboxyphenyl)-8-(4-methoxybenzamido)-4-oxo-4H-chromene-2-carboxylic acid except that the corresponding 2-methoxycarboxyphenylboronic acid was used instead of 3-carboxyphenylboronic acid. Yield 2% (two steps). 1H-NMR (400 MHz, DMSO-d6) δ: 10.04 (1H, brs); 8.21 (1H, d, J = 2.0 Hz); 8.03 (2H, d, J = 8.7 Hz); 7.79 (1H, d, J = 7.1 Hz); 7.70 (1H, d, J = 2.0 Hz); 7.65 (1H, dt, J = 7.5 Hz, 1.2 Hz); 7.52 (1H, m); 7.48 (1H, d, J = 7.5 Hz); 7.09 (2H, d, J = 8.3 Hz); 6.69 (1H, s); 3.86 (3H, s). EI-MS (m/z): 459 [M]+. Purity by HPLC: 95%, RT 4.76 minutes.

6-(4-Hydroxyphenyl)-8-(4-Methoxybenzamido)-4-Oxo-4H-Chromene-2-Carboxylic Acid (KGP-27).

The target compound was obtained in the same manner as 6-(3-carboxyphenyl)-8-(4-methoxybenzamido)-4-oxo-4H-chromene-2-carboxylic acid except that the corresponding 4-hydroxyphenylboronic acid was used instead of 3-carboxyphenylboronic acid. Yield 48% (two steps). 1H-NMR (400 MHz, DMSO-d6) δ: 10.20 (1H, brs); 9.76 (1H, brs); 8.32 (1H, d, J = 2.4 Hz); 8.05 (2H, d, J = 8.7 Hz); 8.00 (1H, d, J = 2.4 Hz); 7.60 (2H, d, J = 8.7 Hz); 7.11 (2H, d, J = 8.7 Hz); 6.96 (1H, s); 6.91 (2H, d, J = 8.3 Hz); 3.87 (3H, s). EI-MS (m/z): 431 [M]+. Purity by HPLC: 96%, RT 4.57 minutes.

Flow Cytometry

RBL-2H3/HA-rGPR35 and its control were incubated with an anti-CD16/CD32 antibody (clone 2.4G2, 12.5 μg/ml) for 10 minutes at 4°C. They were rinsed in PBS(−) containing 2% FBS and 0.05% NaN3 and incubated with an anti-HA antibody (clone HA124, 1:1,600) for 30 minutes at 4°C. The cells were stained with a PE-conjugated F(ab’)2-goat anti-mouse IgG (H+L) antibody (1:100) for 30 minutes at 4°C. For detection of FcεRI, the cells were incubated with an anti-DNP IgE (clone SPE-7, 12.5 μg/ml) for 50 minutes at 4°C. The cells were then rinsed and incubated with an FITC-conjugated anti-mouse IgE (1:500). Viability of the cells was monitored by the propidium iodide staining, and the dead cell population was gated out. Control experiments were performed using fluorescent dye–conjugated isotype control antibodies.

Measurement of Filamentous Actin Formation

RBL-2H3 clones were fixed with 4% paraformaldehyde in phosphate buffer for 20 minutes followed by permeabilization using PBS containing 0.2% Triton X-100 for 10 minutes at room temperature. Blocking was performed in PBS containing 1% bovine serum albumin. They were incubated with Phalloidin-iFluor 488 and Hoechst 33342 for 1 hour. The fluorescence images were obtained using LSM800 (Carl Zeiss AG, Oberkochen, Germany) and were analyzed with Image J (version 1.53e; https://imagej.net/ij/index.html).

Measurement of RhoA Activity

The amounts of GTP form of RhoA were determined using G-LISA RhoA Activation Assay (Cytoskeleton, Inc., Denver, CO). Briefly, GTP-bound RhoA was captured by an immobilized effector protein of RhoA and detected by an anti-RhoA antibody.

Passive Cutaneous Anaphylaxis

Mice received an intradermal injection of 30 ng of IgE (clone SPE-7) in the ear pinnae 24 hours before being challenged with an intravenous injection of 60 μg DNP-HSA and 1 mg Evans blue dye in 0.2 ml saline. Thirty minutes after the challenge, the mice were euthanized by cervical dislocation. Their ear tissues were respectively collected and dissolved in 3 N KOH. The amounts of acetone-extracted dye were determined through measuring the value of OD620.

Statistical Analysis

Data are presented as the independent values or the means ± S.E.M. Statistical significance was determined using one-way ANOVA or two-way ANOVA. Further comparisons were performed with Dunnett multiple comparison test for comparison with the control groups or Tukey-Kramer multiple comparison test for all pairs of column comparison. Holm-Sidak multiple comparison test was performed in case of two-way ANOVA. Two-tailed unpaired Student’s t test was used for comparison between two independent groups.

ResultsSuppression of Peritoneal Mast Cell Degranulation by GPR35 Agonists.

Accumulating evidence indicates that the therapeutic effects of disodium cromoglycate (DSCG) should be based on its actions on mast cells. Degranulation of rat peritoneal mast cells upon IgE-mediated antigen stimulation was significantly suppressed by DSCG (Fig. 1A). Pretreatment with DSCG at 37°C, not at 4°C, canceled its suppressive effects on degranulation, raising the possibility that the target molecule(s) of DSCG should be rapidly desensitized. We further investigated the effects of the other GPR35 agonists, such as KA, and zaprinast. These compounds significantly suppressed degranulation as well as DSCG (Fig. 1B). Suppressive effects of KA and zaprinast were alleviated in the cells pretreated with DSCG at 37°C, raising the possibility that these compounds should share the same target(s) with DSCG. We then investigated the internalization of rat GPR35 using HEK293 cells stably expressing rat GPR35 with hemagglutinin (HA) epitope tag at its amino terminus. Rapid decreases in the surface expression levels of rat GPR35 were found in the cells stimulated with DSCG or zaprinast (Fig. 1C). GPR35 was found to have a potential to couple with several trimeric G proteins including Gi/o (Taniguchi et al., 2006; Wang et al., 2006; Zhao et al., 2010) and to induce chemotaxis of neutrophils. Pertussis toxin (PT), which could inactivate Gαi/o through its ADP-ribosylation, had no significant effects on DSCG-mediated suppression of degranulation (Fig. 1D). Zaprinast was found to be a more potent stabilizer for rat peritoneal mast cells compared with DSCG (calculated IC50 values: DSCG 0.425 μM; zaprinast 14.8 nM) (Fig. 1E).

Fig. 1.

(A) Purified rat peritoneal mast cells sensitized with an anti-DNP IgE (10 μg/ml, clone SPE-7) for 3 hours were stimulated by the antigen (100 ng/ml, DNP-HSA) and 2 μM lysophosphatidylserine. DSCG (200 μM, CG) was added simultaneously with the antigen (0 minutes) or pretreated 5 minutes before the antigen stimulation (5 minutes). Pretreatment with DSCG was performed either at 4°C or at 37°C. Values with **P < 0.01 are regarded as significant. (B) Purified rat peritoneal mast cells sensitized as described above were pretreated without (None) or with 200 μM DSCG (CG) for 5 minutes at 37°C and then stimulated with the antigen in the presence of various GPR35 agonists (CG, 200 μM DSCG; K, 300 μM kynurenic acid; Z, 100 μM zaprinast). Values with *P < 0.05 and **P < 0.01 are regarded as significant. (C) HEK293 cells stably expressing HA-tagged rat GPR35 were stimulated without (open circles), with vehicle alone (open squares), with 100 μM DSCG (closed circles) or with 10 μM zaprinast (closed squares) for the periods indicated. The surface expression levels of GPR35 were measured using flow cytometry as described in Methods. Values are presented as the means ± S.E.M. (n = 3). Values with **P < 0.01 are regarded as significant (vs. vehicle alone). (D) Purified rat peritoneal cells were sensitized with IgE as described above in the presence or absence of pertussis toxin (100 ng/ml, PT) and then stimulated with the antigen with or without 100 μM DSCG (CG). Values with **P < 0.01 are regarded as significant. (E) Degranulation induced by IgE-mediated antigen stimulation was induced as described above. Concentration-dependent suppression of antigen-induced degranulation was investigated using DSCG (open circles) or zaprinast (closed circles). Values are presented as the means ± S.E.M. (n = 3). The levels of degranulation were determined by measuring β-hexosaminidase activity.

In Vitro Characterization of GPR35.

We then investigated the G protein coupling of rat, human, and mouse GPR35 using TGF-α shedding assay. Because increases in the cytosolic Ca2+ concentrations are often associated with stimulated exocytosis and suppression of degranulation mediated by Gq and/or Gi has never been reported, we paid attention to the other G protein coupling. Because accumulating evidence suggests that G12/13-mediated cytoskeletal reorganization could modulate the process of degranulation and that increases in cytosolic cAMP levels could suppress degranulation (Peachell, 2006; Ménasché et al., 2021), a coupling with G12/13 or Gs was investigated as the responsible G protein in mast cells using HEK293/ΔGαq/11/12/13 cells, which genetically lack four endogenous Gα proteins involved in triggering TGF-α shedding. Rat, human, and mouse GPR35 were found to be coupled with exogenously added Gαq/Gα12/13 chimeric proteins, whereas mouse GPR35, not rat and human GPR35, was coupled with exogenously added Gαq/Gαs chimeric proteins (Fig. 2A).

Fig. 2.

(A) HEK293/ΔGαq/11/12/13 cells were transiently transfected with the expression vectors (pCAGGS) encoding alkaline phosphatase (AP)-conjugated TGF-α and rat-, human-, and mouse-GPR35, and with those encoding the chimeric Gαq with various trimeric Gα protein tails (Gαs, open squares; Gα12, open circles; Gα13, closed circles) as described in Methods. The cells were stimulated with the indicated concentrations of zaprinast, and the shedding levels of TGF-α were monitored by measuring the enzyme activity of AP. Values are presented as the means ± S.E.M. (n = 3). (B) HEK293 cells were transiently transfected with the expression vectors encoding alkaline phosphatase (AP)-conjugated TGF-α, Gα16 and the chimeric Gαq with various trimeric G protein tails, and GPR35 of rat, human, or mouse as described in Methods. The cells were stimulated with the indicated concentrations of various GPR35 agonists (CG: DSCG, open circles; K: kynurenic acid, closed circles; Z: zaprinast, open squares; PA: pamoic acid, closed squares) and the shedding levels of TGF-α were monitored by measuring the enzyme activity of AP. Values are presented as the means ± S.E.M. (n = 3).

Affinities of GPR35 agonists were found to be varied among the species. Zaprinast might be a good agonist, which could act on rat, human, and mouse GPR35 with high affinities (Fig. 2B). DSCG was less potent as a mouse GPR35 agonist. Although KA is regarded as a candidate for the endogenous agonist, its agonistic activities were found only at higher concentrations. Pamoic acid could act only on human GPR35.

Development of Novel GPR35 Agonists.

Funke et al. (2013) developed a series of 8-amido-chromen-4-one-2-carboxylic acid derivatives, some of which showed high affinity for human GPR35 in the β-arrestin recruitment assay but lower affinity for rat and mouse GPR35. MacKenzie et al. (2014) demonstrated that the other mast cell stabilizers, bufrolin and lodoxamide, should be equipotent agonists of rat and human GPR35. We designed a series of novel compounds by reference to these studies (Fig. 3) and verified their potential as the GPR35 agonists. Newly synthesized compounds KGP-18 and KGP-27 exhibited equal or higher affinities compared with zaprinast in the cells expressing rat or mouse GPR35, whereas they were good agonists with much higher affinities in the cells expressing human GPR35 (Fig. 3; Table 1). KGP-7 exhibited similar good profiles as the agonist in the cells expressing rat or human GPR35, whereas it was a partial agonist in the cells expressing mouse GPR35. KGP-20, a structurally related compound of KGP-7, was found to be a weak agonist in the cells expressing rat or mouse GPR35, whereas it had no effects on human GPR35.

Fig. 3.

Chemical structures of the novel compounds used in this study (A) and zaprinast (B). HEK293 cells were transiently transfected with the expression vectors encoding alkaline phosphatase (AP)-conjugated TGF-α, Gα16 and the chimeric Gαq with various trimeric G protein tails and GPR35 of rat (C), human (D), or mouse (E) as described in Methods. The cells were stimulated with the indicated concentrations of various GPR35 agonists (Z, zaprinast; 7, KGP-7; 18, KGP-18; 20, KGP-20; 27, KGP-27), and the shedding levels of TGF-α were monitored by measuring the enzyme activity of AP. Values are presented as the means ± S.E.M. (n = 3).

TABLE 1

TGF-α shedding assay of rat, human, and mouse GPR35 was performed as described in the legend to Fig. 3. The amount of EC50 (nM) of each compound is presented.

We then verified the potential of these novel compounds as mast cell stabilizers using rat peritoneal mast cells. KGP-7 and KGP-18 significantly suppressed degranulation induced by IgE-mediated antigen stimulation (Fig. 4, A and B), whereas KGP-20 did not suppress it (Fig. 4C). DSCG, zaprinast, and KGP-18 could also suppress degranulation induced by compound 48/80, which has been identified as an agonist of Mas-related G protein–coupled receptor (Mrgpr) subtypes such as MRGPRX2 and Mrgprb2, although their effects might be partial (Fig. 4D). These GPR35 agonists had no effects on degranulation induced by thapsigargin (Fig. 4E).

Fig. 4.

Degranulation induced by IgE-mediated antigen stimulation was induced as described in the legend to Fig. 1. The effects of KGP-7 (A), −18 (B), and −20 (C) on degranulation were investigated. The concentration of KGP-20 and zaprinast was 1 μM. Values with *P < 0.05 and **P < 0.01 are regarded as significant. (D and E) Purified rat peritoneal mast cells were stimulated with 100 ng/ml compound 48/80 (D) or 1 μM thapsigargin (E) in the absence (C) and presence of DSCG [CG, 100 μM (D) and 200 μM (E)]; zaprinast (Z, 1 μM); KGP-18 (18, 1 μM); and KGP-20 (1 μM). The levels of degranulation were determined by measuring β-hexosaminidase activity. Values with *P < 0.05 and **P < 0.01 are regarded as significant.

GPR35-Mediated Suppression of Degranulation in a Rat Mast Cell Line.

We performed RNA interference experiments but failed to achieve sufficient genetic silence in purified rat mast cells. We instead performed a gain-of-function study. RBL-2H3 cells, which have been frequently used as a mast cell model, were found to express negligible levels of GPR35 when compared with rat peritoneal mast cells (Fig. 5A). We established an RBL-2H3 clone that stably expresses rat GPR35 with the HA-tag at its amino terminus (RBL-2H3/HA-rGPR35) (Fig. 5B). The surface expression levels of FcεRI were quite comparable between RBL-2H3/HA-GPR35 and its control (RBL-2H3/mock) (Fig. 5C). Similar levels of thapsigargin-induced degranulation were observed between these two clones (Fig. 5D), whereas significantly lower levels of degranulation were observed in RBL-2H3/HA-rGPR35 cells when they were subjected to IgE-mediated antigen stimulation (Fig. 5E).

Fig. 5.

(A) Total RNAs were extracted from purified rat peritoneal mast cells (pMC, open circles) and RBL-2H3 cells (RBL, closed circles). The expression levels of mRNA of GPR35 were determined by quantitative RT-PCR analyses. (B) An RBL-2H3 clone stably expressing HA-tagged rat GPR35 (RBL-2H3/HA-rGPR35) was established. The amino-terminal HA tag was detected by flow cytometry using an anti-HA antibody (an open cytogram, 1:1600) and a PE-conjugated F(ab’)2-goat anti-mouse IgG (H+L) antibody (1:100). An RBL-2H3 clone transfected with the empty vector (Mock) was also investigated as the control (a closed cytogram). The cytogram with the isotype control is shown with a dashed line. (C) The surface expression levels of FcεRI of the IgE-sensitized cells were measured using an FITC-conjugated anti-IgE antibody by flow cytometry. (D) RBL-2H3/HA-rGPR35 (GPR35, closed circles) or its control (Mock, open circles) was stimulated with the indicated concentrations of thapsigargin. The levels of degranulation were determined by measuring the enzyme activity of β-hexosaminidase. (E) RBL-2H3/HA-rGPR35 (GPR35, closed circles) or its control (Mock, open circles) was sensitized with an anti-DNP IgE (50 ng/ml, clone SPE-7) for 3 hours, washed, and then stimulated with the indicated concentrations of antigen (DNP-HSA). The levels of degranulation were determined by measuring the enzyme activity of β-hexosaminidase. Values are presented as the means ± S.E.M. (n = 3). Values with *P < 0.05 and **P < 0.01 are regarded as significant. (F and G) RBL-2H3/HA-rGPR35 (closed circles) or its control (open circles) was activated by thapsigargin (F) or the antigen (G) as described above for the indicated periods. The cells were stained with phalloidin-iFluor 488 (1:2000) and Hoechst 33342 (1:5000) to visualize the filamentous actin as described in Methods. The mean fluorescent intensities (MFI) were calculated using Image J and presented in the right graphs. Values are presented as the means ± S.E.M. (n = 3). Values with *P < 0.05 and **P < 0.01 are regarded as significant.

Because accumulating evidence indicates that actin reorganization should play a critical role in degranulation (Ménasché et al., 2021; Lazki-Hagenbach et al., 2021), we then investigated the changes in the amount of filamentous actin (F-actin) during degranulation. It was significantly higher in RBL-2H3/HA-rGPR35 cells under unstimulated conditions (Fig. 5F) and rapidly dropped to similar levels to RBL-2H3/mock cells when they were stimulated with thapsigargin. The amount of F-actin was also decreased in RBL-2H3/HA-rGPR35 cells upon IgE-mediated antigen stimulation, but it was significantly higher in comparison with RBL-2H3/mock cells all through the process of degranulation (Fig. 5G).

We then investigated the effects of zaprinast on RBL-2H3/HA-rGPR35 cells. The amount of GTP-form of RhoA was higher by approximately 1.4-fold in RBL-2H3/HA-rGPR35 cells than in RBL-2H3/mock cells under unstimulated conditions (Fig. 6A). Zaprinast induced a transient but significant increase in the amount of GTP-form of RhoA in RBL-2H3/HA-rGPR35 cells sensitized with an anti-DNP IgE and not in RBL-2H3/mock cells (Fig. 6B). The amount of F-actin was also transiently decreased and then restored in RBL-2H3/HA-rGPR35 cells stimulated with zaprinast (Fig. 6C). An agonist-induced rapid internalization of GPR35 was reproduced in RBL-2H3/HA-rGPR35 cells as well as in HEK293 cells expressing HA-rGPR35 (Fig. 6D).