RasGRP4 Regulates the Expression of Prostaglandin D2in Human and Rat Mast Cell Lines*

Mast cells (MCs) are a major source of prostaglandin (PG) D2 in connective tissues, and the expression of this eicosanoid has been linked to asthma and other inflammatory disorders. While it is known that the surface receptor c-kit controls PGD2 expression in MCs by regulating the levels of a synthase that converts PGH2 to PGD2, the intracellular signaling proteins that act downstream of c-kit in this cyclooxygenase pathway have not been identified. We recently cloned a new cation-dependent, guanine nucleotide exchange factor/phorbol ester receptor (designated RasGRP4) that is required for the efficient expression of granule proteases in the human MC line HMC-1. GeneChip analysis of ∼12,600 transcripts in RasGRP4− and RasGRP4+ HMC-1 cells revealed a >100-fold difference in the levels of hematopoietic PGD2 synthase mRNA. No other transcript in the eicosanoid pathway was influenced by RasGRP4 in a comparable manner. As assessed by SDS-PAGE immunoblot analysis, RasGRP4+ HMC-1 cells contained substantial amounts of PGD2 synthase protein. RasGRP4+ MCs also produced ∼15-fold more PGD2 than did RasGRP4− MCs when both cell populations were activated by calcium ionophore. The induced transcript is therefore translated, and substantial amounts of functional PGD2 synthase accumulate in RasGRP4+ MCs. In support of the conclusion that RasGRP4 controls PGD2expression in MCs, inhibition of RasGRP4 expression in the rat MC line RBL-2H3 using a siRNA approach resulted in low levels of PGD2 synthase protein.

Mast cells (MCs) are a major source of prostaglandin (PG) D 2 in connective tissues, and the expression of this eicosanoid has been linked to asthma and other inflammatory disorders. While it is known that the surface receptor c-kit controls PGD 2 expression in MCs by regulating the levels of a synthase that converts PGH 2 to PGD 2 , the intracellular signaling proteins that act downstream of c-kit in this cyclooxygenase pathway have not been identified. We recently cloned a new cationdependent, guanine nucleotide exchange factor/phorbol ester receptor (designated RasGRP4) that is required for the efficient expression of granule proteases in the human MC line HMC-1. GeneChip analysis of ϳ12,600 transcripts in RasGRP4 ؊ and RasGRP4 ؉ HMC-1 cells revealed a >100-fold difference in the levels of hematopoietic PGD 2 synthase mRNA. No other transcript in the eicosanoid pathway was influenced by RasGRP4 in a comparable manner. As assessed by SDS-PAGE immunoblot analysis, RasGRP4 ؉ HMC-1 cells contained substantial amounts of PGD 2 synthase protein. RasGRP4 ؉ MCs also produced ϳ15-fold more PGD 2 than did RasGRP4 ؊ MCs when both cell populations were activated by calcium ionophore. The induced transcript is therefore translated, and substantial amounts of functional PGD 2 synthase accumulate in RasGRP4 ؉ MCs. In support of the conclusion that RasGRP4 controls PGD 2 expression in MCs, inhibition of RasGRP4 expression in the rat MC line RBL-2H3 using a siRNA approach resulted in low levels of PGD 2 synthase protein.
Activated human and rodent mast cells (MCs) 1 generate and release substantial amounts of prostaglandin (PG) D 2 (1), and many of the vasodilation and hemodynamic problems that occur in patients with systemic mastocytosis are thought to be caused by the excessive production of this eicosanoid. PGD 2 is a neuromodulator/sleep-inducing factor in the central nervous system. In peripheral tissues, PGD 2 inhibits platelet aggregation (2) but activates eosinophils. PGD 2 is a potent chemotactic factor for eosinophils (3), and PGD 2 -treated eosinophils increase their calcium mobilization, actin polymerization, and surface expression of CD11b (4,5). This eicosanoid also enhances the rate of apoptosis of eosinophils if these granulocytes are cultured for ϳ20 h in the absence of a viability-enhancing cytokine such as interleukin (IL) 5 (6). Pulmonary MCs play important roles in the initiation and/or progression of asthma, and substantial amounts of PGD 2 are released into the lungs during asthma attacks (7,8). The observation that patients with asthma undergo bronchoconstriction when they inhale PGD 2 (9) documents the pathologic consequences of high levels of PGD 2 in the lung. PGD 2 exerts its biological actions via two seven-transmembrane, G protein-coupled receptors (designated PTGDR/DP and GPR44/CRTH2) (10 -12). Targeted disruption of the PTGDR gene in the mouse leads to a marked reduction in antigen-induced airway reactivity to acetylcholine (13), thereby supporting the earlier inhalation studies in humans and dogs that implicated an adversarial role for PGD 2 in the lung.
In the cyclooxygenase pathway that ultimately leads to PGD 2 expression, liberated arachidonic acid is converted to PGG 2 and then to PGH 2 . PG endoperoxide H synthase (PGHS) 1 (also known as cyclooxygenase 1) and PGHS-2 (also known as cyclooxygenase 2) are both able to carry out this two-step biosynthetic process. The resulting precursor eicosanoid is then metabolized by terminal synthases to form PGD 2 , PGE 2 , PGF 2␣ , PGI 2 /prostacylin, and thromboxane A 2 . Two PGD 2 synthases have been identified in mice, rats, and humans (14,15). The brain enzyme is a glutathione-independent member of the lipocalin family of proteins. The distinct hematopoietic enzyme that is expressed in MCs (16) is a sigma-class, glutathione S-transferase family member. PGH 2 can be metabolized inside cells to thromboxane A 2 and to a variety of PGs. Thus, the amount of PGD 2 produced by an Fc⑀RI-or calcium ionophore-activated MC is determined in a large part by the amount of PGD 2 synthase protein in the cell. MCs are heterogeneous in terms of what eicosanoids they produce. c-kit is a member of the type III receptor tyrosine kinase family. PGD 2 -expressing MCs contain abundant amounts of c-kit on their surfaces, and Murakami et al. (17) noted that c-kit ligand (KL) somehow regulates the levels of PGD 2 synthase in mouse MCs. To a lesser extent, IL-3 and IL-10 also influence the expression of PGD 2 synthase in MCs. Treatment of human megakaryocytic cell lines with phorbol esters results in a 2-5fold increase in the levels of PGD 2 synthase mRNA (18,19). While these findings suggest that one or more diacylglycerol/ phorbol ester-responsive proteins play an important role in the expression of PGD 2 synthase in hematopoietic cells, the intracellular proteins that act downstream of c-kit and other membrane receptors to control the levels of PGD 2 synthase in MCs have not been identified.
We recently cloned a new member of the Ras guanine nucle-* This work was supported by Grants AI-23483, HL-36110, and HL-63284 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ To whom correspondence should be addressed: Brigham and Women's Hospital, Dept. of Medicine, Smith Bldg., Rm. 616B, 1 Jimmy Fund Way, Boston, MA 02115. Tel.: 617-525-1231; Fax: 617-525-1310; E-mail: rstevens@rics.bwh.harvard.edu. 1 The abbreviations used are: MC, mast cell; mBMMC; mouse bone marrow-derived MC; IL, interleukin; LT, leukotriene; PG, prostaglandin; PGHS, PG endoperoxide H synthase; RT, reverse transcriptase; RBL, rat basophilic leukemia; TNF-␣, tumor necrosis factor ␣; KL, c-kit ligand; ELISA, enzyme-linked immunosorbent assay; siRNA, small interfering RNA. otide-releasing protein (RasGRP) family of intracellular signaling proteins (20). In contrast to the other three members of its family, RasGRP4 normally is restricted to mature MCs and their circulating progenitors. RasGRP4 functions as a cationdependent, guanine nucleotide exchange factor. It also is a diacylglycerol/phorbol ester receptor that appears to act downstream of c-kit. The hRasGRP4 gene resides on chromosome 19q13.1 (20) in the vicinity of a site that has been linked to bronchial hyperresponsiveness (21,22). RasGRP1 is essential for the final stages of T-cell development (23). Although human MCs do not express RasGRP1, RasGRP2, or RasGRP3, transfection studies carried out with the RasGRP4-defective HMC-1 cell line derived from a patient with a MC leukemia suggests that RasGRP4 is required for the final stages of MC development (20). Thus, at least two members of the RasGRP family of signaling proteins appear to control cellular differentiation and maturation. We previously noted that RasGRP4 influences the storage of varied neutral proteases in the secretory granules of a MC line. We now report that RasGRP4 also controls what eicosanoids this immune cell produces.

EXPERIMENTAL PROCEDURES
Transcript Analysis of RasGRP4 Ϫ and RasGRP4 ϩ HMC-1 Cells-RasGRP4 ϩ and RasGRP4 Ϫ HMC-1 cells (20) were cultured in enriched medium (Iscove's modified Dulbecco's medium (BioWhittaker) containing 10% heat-inactivated fetal calf serum (Sigma), 2 mM L-glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, and 10 M monothioglycerol (Sigma) with or without 200 -500 g/ml G418) in the absence of human cytokines. Total RNA was isolated from the two populations of cells with TRIzol (Invitrogen), and comparative transcript profiling was carried out at the Gene Array Technology Center (Brigham and Women's Hospital, Boston, MA) with HG-U95A GeneChips (Affymetrix, Santa Clara, CA) and the experimental protocol recommended by Affymetrix. Each GeneChip contains ϳ12,600 probe sets. In these analyses, 8 g of total RNA from RasGRP4 Ϫ and RasGRP4 ϩ HMC-1 cells were reverse-transcribed using the GeneChip T7-oligo(dT) promoter primer kit. Biotinylated complementary RNAs, generated from the resulting cDNAs, were fragmented and incubated with the GeneChips for 16 h. The resulting GeneChips were incubated with streptavidin-  (Table I) was obtained from the first transfection experiment (lane 2). B, a separate semiquantitative RT-PCR approach also was used to evaluate the levels of PGD 2 synthase mRNA in these same four populations of HMC-1 cells. Gel electrophoresis confirmed that the 100-bp product generated in each case corresponds to PGD 2 synthase. Size markers are shown on the left.  a The "fold change" values represent the levels of the indicated transcripts in RasGRP4 ϩ HMC-1 cells relative to that in RasGRP4 Ϫ HMC-1 cells. In each case, data were normalized to the ubiquitously expressed transcripts that encode ␤-actin and glyceraldehyde-3-phosphate dehydrogenase.
phycoerythrin staining solution. The obtained signals were then amplified by sequentially incubating the GeneChips with goat IgG, biotinylated goat anti-streptavidin antibody, and staining solution. Hybridization to the array was quantified with a Hewlett-Packard gene array laser scanner. In separate studies, the generated RT-PCR products were subjected to gel electrophoresis to confirm that they were derived from the authentic PGD 2 synthase transcript.
Quantitation of PGD 2 Synthase mRNA Levels in Cells by Real-time RT-PCR-The GeneChip data obtained with the PGD 2 synthase probe set were confirmed by real-time RT-PCR. The PCR primers and fluorogenic probes for measuring PGD 2 synthase mRNA levels were designed with the use of "Primer Express" (Applied Biosystems, CA). TaqMan's 18 S rRNA control reagents were used to normalize RNA levels in each HMC-1 sample. Fluorescent probes were selected such that their T m was ϳ10°C higher than the matching primer pair. Each high performance liquid chromatography-purified fluorescent probe contained a 6-carboxyfluorescein (FAM) reporter dye covalently attached at its 5Ј end and a black hole quencher 1 quencher dye covalently attached at its 3Ј end. The forward primer 5Ј-GGGCAGAGAAAAAGCAAGATGT-3Ј, the reverse primer 5Ј-CCCCCCTAAATATGTGTCCAAG-3Ј, and the dual-labeled fluorescent probe 5Ј-(FAM)-CAATGAGCTGCTCACGTATA-ATGCGCC-(BHQ-1)-3Ј were used to quantitate PGD 2 synthase mRNA levels in these assays. Reactions were carried out using an iCycler IQ real-time detection system (Bio-Rad). SuperScript one-step RT-PCR with Platinum Taq kits (Invitrogen) were used. Each 50-l reaction contained 200 ng of total RNA, 5 mM MgSO 4 , 500 nM forward and reverse primers, and 200 nM fluorescent probe. Samples were analyzed in triplicate. Negative-control reactions were carried out on replicate samples that had not been subjected to the reverse transcriptase step. Additional negative-control reactions were carried out in wells lacking HMC-1 cellular RNA. The reaction conditions were as follows: 15 min at 50°C and 5 min at 95°C, followed by 45 two-temperature cycles (15 s at 95°C and 1 min at 60°C). The standard curve method (24,25) was used to analyze the obtained data.
Calcium Ionophore Activation of RasGRP4 Ϫ and RasGRP4 ϩ HMC-1 Cells-RasGRP4 Ϫ and RasGRP4 ϩ HMC-1 cells were washed, suspended at a concentration of 10 6 cells/ml in calcium/magnesium-free phosphate-buffered saline, and stimulated with 0.5 M calcium ionophore A23187 (Sigma) at 37°C for 30 min as done in other eicosanoid studies of MCs (26). The generated eicosanoids PGD 2 , PGE 2 , and leukotriene C 4 (LTC 4 ) in the supernatants were quantitated using the relevant ELISA kits (Cayman Chemical). Each reaction was read at 450 nm using an ELISA plate reader (Molecular Device). Data are given as mean Ϯ S.D. Significance was defined as p Ͻ 0.05 by the Student's t test.
siRNA-mediated Inhibition of RasGRP4 Expression in RBL-2H3 Cells-A siRNA approach similar to that described by Elbashir et al. (27) was used to evaluate the consequences of decreased expression of RasGRP4 in the rat MC line RBL-2H3. The coding sequence of rat RasGRP4 (28) was scanned to identify a gene-specific 21-nucleotide sequence downstream of an "AA" sequence that possesses a 55% GC content. A BLAST search confirmed that the selected sequence (corresponding to residues 27-47 in GenBank TM accession number AF465263) is not present in another transcript in GenBank TM data bases. The RasGRP4-specific oligonucleotide 5Ј-GUCUCAUCAGGAA-UGCUCUGGdTdT-3Ј and its corresponding oligonucleotide 5Ј-CCAG-AGCAUUCCUGAUGAGACdTdT-3Ј were synthesized and purified (Dharmacon Research, Lafayette, CO) and then annealed to form the final siRNA duplex with its TT overhangs. The resulting siRNA duplex was introduced into RBL-2H3 cells (line CRL-2256; American Type Culture Collection, Manassas, VA) using a liposome transfection approach. Liposome/siRNA complexes were formed at room temperature using 3 l of 20 M siRNA, 2 l of LipofectAMINE TM 2000 (Invitrogen), and 100 l of Opti-MEM I serum-free culture medium (Invitrogen). The resulting solution was added dropwise to each culture dish containing ϳ5 ϫ 10 4 adherent MCs. The cells were incubated 3-4 h at 37°C. One ml of serum-enriched medium was then added, and the cells were cultured for an additional 24 -48 h. The transiently transfected cells were harvested, and the levels of PGD 2 synthase and ␤-actin protein were measured using the above SDS-PAGE immunoblot approach. In these assays, each protein blot was incubated ϳ17 h with anti-PGD 2 synthase antibody and then for 1 h with the anti-␤-actin antibody (Sigma) before final development.

RESULTS AND DISCUSSION
All nontransformed rodent and human MCs that have been examined to date preferentially metabolize arachidonic acid via the cyclooxygenase pathway to PGD 2 rather than to PGE 2 . Nevertheless, Macchia et al. (29) discovered that HMC-1 cells produce ϳ20-fold more PGE 2 than PGD 2 . This surprising finding allowed us to use the c-kit ϩ HMC-1 cell line to further elucidate the intracellular signaling pathways that control PGD 2 production in MCs. Transcript analysis (Fig. 1) revealed that the failure of HMC-1 cells to generate large amounts of PGD 2 is a consequence of a low rate of transcription of the PGD 2 synthase gene and/or a high rate of catabolism of its transcript.
The RasGRP4 transcript was initially cloned from IL-3-developed mouse bone marrow-derived MCs (mBMMCs). While all mouse, rat, and human MCs appear to express RasGRP4 mRNA and/or protein, the amount of RasGRP4 protein in a   FIG. 3. Generation of PGD 2 , PGE 2 , and LTC 4 in calcium ionophore-activated RasGRP4 ؊ and RasGRP4 ؉ HMC-1 cells. Ras-GRP4 Ϫ and RasGRP4 ϩ HMC-1 cells were exposed to calcium ionophore A23187 for 30 min. The amounts of PGD 2 (left bars), PGE 2 (middle bars), and LTC 4 (right bars) generated in each experiment were determined by separate ELISAs.
FIG. 4. PGD 2 synthase levels in control and siRNA-treated RBL-2H3 cells. SDS-PAGE immunoblots, prepared from the lysates of RBL-2H3 cells before and after these cells were transfected with a RasGRP4-specific siRNA for 48 h, were probed for with anti-PGD 2 synthase (PGDS) and anti-␤-actin antibodies. Similar data were obtained in a second siRNA experiment. mouse peritoneal MC greatly exceeds that in a mBMMC as assessed by SDS-PAGE immunoblot analysis. 2 Calcium ionophore-or Fc⑀RI-activated mBMMCs produce ϳ25-fold more LTC 4 than PGD 2 , whereas peritoneal MCs activated in a similar manner produce Ͼ40-fold more PGD 2 than LTC 4 (1,26). The cumulative data raised the possibility that RasGRP4 regulates arachidonic acid metabolism in MCs. Thus, we evaluated whether or not RasGRP4 controls PGD 2 and/or LTC 4 expression in HMC-1 and RBL-2H3 cells.
Comparative transcript analysis of RasGRP4 Ϫ and Ras-GRP4 ϩ HMC-1 cells using an Affymetrix GeneChip approach revealed a dramatic difference in the steady-state levels of the transcript that encodes hematopoietic PGD 2 synthase in the two populations of cells (Table I). RasGRP4 ϩ HMC-1 cells contained Ͼ100-fold more PGD 2 synthase mRNA than did the starting population of HMC-1 cells that express nonfunctional forms of RasGRP4. No transcript was induced to a comparable level, including the transcripts that encode brain-type PGD 2 synthase and LTC 4 synthase. Table I shows profile data relating to the levels of the transcripts that encode different proteins that participate in arachidonic acid metabolism. The PGD 2 synthase GeneChip data were confirmed by real-time RT-PCR (Fig. 1A) and by semiquantitative RT-PCR (Fig. 1B) analyses in three separate populations of RasGRP4-expressing cells. In a control experiment, HMC-1 cells transfected with the expression vector pcDNA3.1 lacking the RasGRP4 cDNA contained barely detectable amounts of PGD 2 synthase transcript (data not shown).
Because the levels of a transcript do not always correlate with the levels of its translated product, an SDS-PAGE immunoblot approach was used to compare the levels of PGD 2 synthase protein in RasGRP4 Ϫ and RasGRP4 ϩ HMC-1 cells. The amount of PGD 2 synthase protein in RasGRP4 Ϫ HMC-1 cells was nearly below detection (Fig. 2). In contrast, RasGRP4 ϩ HMC-1 cells contained substantial amounts of an intracellular 25-kDa protein that was recognized by the anti-PGD 2 synthase antibody. The induced PGD 2 synthase transcript is therefore translated and the appropriately sized biosynthetic enzyme accumulates in the transfectants. As assessed by SDS-PAGE immunoblot analysis, RasGRP4 did not induce HMC-1 cells to increase their accumulation of PGHS1, PGE 2 synthase, or 5-lipoxgenase protein (Fig. 2). Thus, RasGRP4 induces a selective accumulation of PGD 2 synthase mRNA and protein in this MC line.
As assessed by ELISA, calcium ionophore-activated Ras-GRP4 ϩ HMC-1 cells produced 12-20-fold more PGD 2 (p Ͻ 0.05) than did RasGRP4 Ϫ HMC-1 cells (Fig. 3). The levels of PGE 2 and LTC 4 were modestly increased and decreased, respectively, in the calcium ionophore-treated RasGRP4 ϩ cells. However, the variations in the amounts of these eicosanoids were not statistically significant. The fact that HMC-1 cells express nonfunctional forms of RasGRP4 indicates that RasGRP4 is not essential in the early stages of MC development, including the c-kit/KL-mediated proliferation of its progenitors. Nevertheless, the observation that HMC-1 cells are unable to granulate (20) and to produce substantial amounts of PGD 2 (Fig. 3) implies that RasGRP4 is required for the efficient expression of the cassette of genes that encode a number of the granule and lipid mediators of MC. The siRNA data obtained from transiently transfected RBL cells (Fig. 4) support this conclusion. RBL cells contain PGD 2 synthase protein, and these rat MCs (30) produce substantial amounts of PGD 2 when exposed to calcium ionophore (31). RBL-2H3 cells also contain RasGRP4 mRNA. 2 Thus, a siRNA approach was used to evaluate the consequences of decreased expression of RasGRP4 in RBL-2H3 cells. As noted in Fig. 4, inhibition of RasGRP4 expression in the MC line resulted in a transient  h) inhibition of PGD 2 synthase expression. As far as we are aware, no one has examined eicosanoid production in transgenic mice that lack RasGRP1 or in cultured cells that have been induced to express varied forms of the other RasGRP family members. Nevertheless, the finding that RasGRP4 regulates PGD 2 expression in two populations of cultured MCs raises the possibility that RasGRP1, RasGRP2, and/or RasGRP3 regulate eicosanoid production in other cell types.
Earlier in vitro studies suggested that KL is required for maximal expression of PGD 2 synthase in mouse MCs. HMC-1 cells are able to proliferate in the absence of exogenous human cytokines, because these transformed cells possess an activating mutation in c-kit (32). The inability of HMC-1 cells to produce large amounts of PGD 2 supports the conclusion that RasGRP4 acts downstream of c-kit. Murakami et al. (17) identified a number of cytokines that influence the KL-mediated expression of PGD 2 synthase in cultured mouse MCs either in a positive or negative manner. As assessed by GeneChip analysis (data not shown), HMC-1 cells express the transcripts that encode the surface receptors for IL-4, IL-10, IL-13, and KL. This MC line also expresses three distinct receptors that recognize TNF-␣ and its family members. RasGRP4 Ϫ and Ras-GRP4 ϩ HMC-1 cells were therefore cultured for 5 days in the presence of varied combinations of IL-3, IL-4, IL-10, IL-13, KL, and TNF-␣. None of these cytokines were able to induce PGD 2 synthase expression in RasGRP4 Ϫ HMC-1 cells (data not shown). In addition, none of these cytokines were able to inhibit the expression of PGD 2 synthase in RasGRP4 ϩ HMC-1 cells. These data imply that RasGRP4 is the dominant intracellular signaling protein that controls PGD 2 expression in MCs no matter what extracellular cytokine environment this immune cell encounters in tissues.