GATA-Dependent Expression of the Interleukin-1 Receptor-Related T1 Gene in Mast Cells

Abstract

The murine delayed-early serum-responsive gene T1 encodes glycoproteins of the interleukin-1 receptor family. Transcriptional initiation in fibroblasts is regulated by c-Fos and gives rise to a rare 5-kb mRNA and an abundant 2.7-kb mRNA. These transcripts are translated into a receptor-like membrane-anchored protein and a secreted protein consisting only of the ectodomain. In mast cells, T1 gene transcription is initiated 10.5 kb further upstream than in fibroblasts and gives rise predominantly to the 5-kb transcript under normal growth conditions. Here we demonstrate that calcium ionophore stimulation of mast cells resulted in an upregulation of T1 gene expression and a switch from the long to the short T1 transcript. This was paralleled by the disappearance of the receptor-type T1 protein on the mast cell surface and the secretion of large amounts of the truncated T1 protein. c-Fos and a T1 enhancer, which have previously been identified to be essential for T1 expression in fibroblasts, were not required for calcium ionophore-mediated T1 gene upregulation. Overexpression of the transcription factor GATA-1 in mast cells caused elevated T1 synthesis. Three GATA elements were identified in the minimal GATA-responsive mast cell promoter. Mutational analysis revealed that all three GATA elements are involved in T1 gene expression. Point mutations within the middle GATA element eliminated promoter activity completely, while mutations of the distal and proximal GATA binding sites reduced promoter strength by factors of 2 and 5, respectively. Exogenous expression of GATA-1 was not sufficient to activate the mast cell-specific promoter in NIH 3T3 fibroblasts.

The T1 gene, also designated ST2 or DER4, was originally isolated as an oncoprotein and growth factor-responsive gene in murine fibroblasts (25, 28, 57, 63). T1 is transcribed into an abundant 2.7-kb and a rare 5-kb mRNA upon stimulation of fibroblasts with proliferation-inducing agents (serum, lysophosphatidic acid, platelet-derived growth factor, and fibroblast growth factor) (23), with proinflammatory cytokines (interleukin-1 [IL-1] and tumor necrosis factor alpha [TNF-α]) (27, 29), or in response to oncogene expression. Mitogen-triggered T1 gene induction in NIH 3T3 cells is mediated by transcription factors of the AP-1 family (23, 60). An essential tetradecanoyl phorbol acetate-responsive element (TRE), a binding site for AP-1 proteins, is located within the T1 enhancer 3.6 kb upstream of the transcription initiation site used in fibroblasts (60). Moreover, overexpression of c-Fos and FosB was sufficient for T1 gene induction in these cells (23). Likewise, the rat homolog of T1, fit-1, was identified as a c-Fos-responsive gene (2).

Both T1 transcripts are initiated at the same site in fibroblasts, and differential 3′ processing is the underlying mechanism for the generation of the two different transcripts (13). Mast cells synthesize mainly the 5-kb T1 mRNA under normal growth conditions. Transcription initiates 10.5 kb further upstream than in fibroblasts (13). The alternative, noncoding first exons used in fibroblasts and mast cells are spliced to the common second exon, where the translation start site is located. Thus, the proteins encoded by the short and long T1 transcripts share the same amino-terminal portion and diverge only eight amino acids before the carboxy terminus of the small T1 protein. The 5-kb mRNA is translated into a receptor-like, plasma membrane-spanning glycoprotein (T1M), whereas the 2.7-kb T1 transcript encodes a secreted glycoprotein (T1S) representing the ectodomain of T1M.

Both T1 proteins belong to the immunoglobulin (Ig) superfamily and in particular to the IL-1 receptor (IL-1R) family (41, 45, 67). The tight chromosomal linkage of the T1 gene and the genes encoding the type I and type II IL-1R on mouse chromosome 1 (59) and human chromosome 2 (58) as well as the highly conserved exon/intron structure of these three genes (53) strongly suggest a common ancestory. However, the cytokines IL-1α and IL-1β do not bind to the T1 protein (26, 49). Studies using recombinant chimeric receptor proteins consisting of the extracellular, ligand binding domain of IL-1R type I fused to the intracellular part of T1M suggest that the two receptors activate the same signal transduction cascades (26, 41, 47). The identification of putative T1 ligands has recently been reported by two groups (16, 27).

The expression patterns of the two T1 transcripts differ significantly. The 2.7-kb T1 mRNA has been detected in fibroblast cell lines (25, 28, 57, 63), in the skin, retina, and bone (49), in the developing mammary gland, and in Ha-ras-induced murine mammary adenocarcinomas (48). Abundant expression of the 5-kb T1 transcript is restricted to the major hematopoietic organs (fetal liver, spleen, and bone marrow) (49) and to the lung (2). Using T1-specific monoclonal antibodies (MAbs) (42), we have recently identified mast cells as the only cells of the hematopoietic system which express T1M (43). All developmental stages of mast cells were shown to express high levels of T1M.

Mast cell progenitors originate from the bone marrow (BM), migrate via the bloodstream, and invade mucosal and connective tissues, where they terminally differentiate into morphologically distinct mature mast cells. They are rich in cytoplasmic granules that store inflammatory mediators such as proteoglycans, histamine, serotonin, TNF-α, and proteases. Mast cell activation results in rapid degranulation and the synthesis and release of cytokines (e.g., TNF-α, IL-1, and IL-3) and lipid mediators, including prostaglandin D₂ and leukotriene C₄. One of the best-studied activation mechanisms of mast cells is triggered by multivalent antigens which bind to IgE on the cell surface and thereby mediate aggregation of the high-affinity Fcɛ receptor (FcɛRI). Some aspects of this mode of activation can be mimicked by calcium ionophore treatment, which results in elevated levels of cytoplasmic Ca²⁺, a hallmark of IgE-mediated mast cell activation. Mast cells participate in acute and persistent inflammatory responses. They play an important role in the pathogenesis of IgE-dependent allergic disorders and anaphylaxis (4, 14). Furthermore, mast cells can take part in acquired immune responses against parasites (36, 51), and it was recently shown that they are able to orchestrate life-saving host responses to bacterial infection (9, 15, 32).

GATA proteins belong to the family of zinc finger-containing transcription factors and play an important role in the regulation of hematopoiesis (44, 52, 62). They bind with high affinity and slightly different specificities to the consensus DNA sequence motif (A/T)GATA(A/G). Six proteins of the GATA transcription factor family have been identified, and each member has a distinct tissue distribution pattern. GATA-1 and GATA-2 are expressed in mast cells, erythroblasts, and megakaryocytes (34, 68), and they are instrumental for tissue-specific gene expression. Genes that are selectively expressed in mast cells and whose expression depends on GATA proteins include those encoding carboxypeptidase A, several mast cell proteases, and the IgE receptor β chain (3, 12, 31, 68). Sequence analysis of the mast cell-specific T1 promoter revealed the presence of three consensus GATA elements.

Here we demonstrate that GATA-1 is indeed involved in mast cell-specific T1 gene expression. We further show that calcium ionophore stimulation of mast cells leads to an upregulation of T1 gene expression which is paralleled by a switch from the production of the long to the short T1 transcript and followed by the secretion of T1S. This effect can be blocked by cyclosporin A, a potent immunosuppressive drug (20). In contrast to fibroblasts, c-Fos is not involved in basal and calcium ionophore-stimulated T1 expression in mast cells. In addition, we provide evidence that the enhancer which is essential for T1 gene activity in fibroblasts is dispensable for mast cell-specific T1 expression.

MATERIALS AND METHODS

Cell culture.

NIH 3T3 cells were grown in Dulbecco’s modified Eagle’s medium (DMEM; Life Technologies, Inc.) supplemented with 10% fetal calf serum (FCS), penicillin (100 U/ml), and streptomycin (100 μg/ml). To arrest cell growth, the cells were incubated for 24 h in DMEM containing 0.5% FCS. This medium was replaced with fresh DMEM containing 10% FCS to stimulate cell cycle entry.

BM cultures were prepared by flushing femurs and tibias of mice with Iscove’s modified Dulbecco’s medium (IMDM; GibcoBRL). BM cells were cultured in IMDM containing GLUTAMAX I (l-alanyl-l-glutamine; 868 mg/liter; Gibco-BRL) supplemented with 10% FCS, 50 μM 2-mercaptoethanol, penicillin (100 U/ml), streptomycin (100 μg/ml), and 2% conditioned culture supernatant from murine IL-3-secreting X63/IL-3 cells (24) (complete IMDM). BM cultures were enriched for mast cells by repetitively subculturing the suspension cells in the presence of IL-3. After approximately 4 weeks in culture, >98% of the cells displayed a typical mast cell-like phenotype, as shown by double-positive flow cytometry staining for c-Kit and surface IgE receptor (66) as well as by Giemsa staining (42). To obtain BM-derived cultured mast cells (BMCMCs) from c-fos^−/− mice which are osteopetrotic, we cut the femur into small pieces and incubated the fragmented bone in complete IMDM.

The calcium ionophore A23187 (Calbiochem) was added at 1 μM unless otherwise indicated. Actinomycin D (Streptomyces sp.; Calbiochem) and cyclosporin A (Trichoderma polysporum; Calbiochem) were added at 5 and 0.5 μg/ml, respectively.

Plasmid constructions.

pBS-A was obtained by cloning an EcoRI/TaqI fragment containing 16 bp of the distal exon 1 and 314 bp of 5′ flanking sequence into the EcoRI and ClaI sites of pBluescript KS⁺ (pBSKS⁺; Stratagene). The XbaI/XhoI insert of pBS-A was subsequently cloned into the NheI/XhoI site of the promoterless secreted alkaline phosphatase (SEAP) expression vector (pSEAP; Tropix, Inc.) to give rise to pA. pBS-B and pBS-C were constructed by introducing a 2.6-kb SacI/EcoRI fragment and a 6-kb EcoRI fragment which flank the restriction fragment inserted into pBS-A into the SacI/EcoRI and EcoRI sites, respectively, of pBS-A. The SacI/XhoI insert of pBS-B whose SacI end was made blunt was cloned into the blunt-ended NheI and XhoI sites of pSEAP to obtain pB. To construct pC, we first introduced additional restriction sites into plasmid pSEAP. The oligonucleotides 5′-CTAGCGAATTCCTCGAGAGATCTGCGGCCGCGGGCCCACTAGTA-3′ and 5′-AGCTTACTAGTGGGCCCGCGGCCGCAGATCTCTCGAGGAATTCG-3′ were annealed and ligated into the NheI/HindIII site of pSEAP. The SpeI/XhoI insert of pBS-C was then introduced into the NheI and XhoI sites of this modified plasmid. The HincII site upstream of the ATG initiation codon in exon 2 was converted into an EcoRI site in previous experiments (60). A 0.8-kb HindIII/EcoRI fragment flanking this site was cloned into the HindIII/EcoRI site of plasmid pBCSK⁺ (Stratagene) to obtain pHelp-1. A 9.7-kb HindIII fragment (positions −12.9 to −3.2 kb with respect to the transcription start site used in fibroblasts) was inserted into the HindIII site of pHelp-1. The XhoI/NotI insert of this construct was introduced into the XhoI/NotI site of the modified SEAP vector to give rise to pD. To obtain the plasmid pE, we first destroyed the SacI restriction site in pBSKS⁺ by treatment with T4 DNA polymerase. The 9.7-kb HindIII fragment described above was cloned into this modified vector. The SacI insert (positions −10.1 to −3.9 kb) was deleted from this plasmid, and the HindIII insert from the resulting construct was introduced into the HindIII site of pHelp-1. The XhoI/SpeI fragment of this plasmid was thereafter ligated into the XhoI/SpeI site of the modified SEAP vector.

Plasmids pL-SH4.9 and pL-SH4.9 mut a, constructed previously (60), contain 4.9 kb of T1 sequence spanning the region 5′ of the translation start codon up to position −3.7 kb. A point mutation had been introduced within the TRE sequence in the enhancer in construct pL-SH4.9 mut a. The SmaI/EcoRI fragment of each of these plasmids was cloned into the blunt-ended HindIII and EcoRI sites of the promoterless SEAP expression vector pSEAP2-Basic (Tropix). The resulting plasmids contain 4.9 kb of T1 sequence 5′ of the open reading frame including the proximal exon 1 and the enhancer in its nonmutated (pHelp-wt) or mutated (pHelp-TREmut) form. A XhoI/ApaI fragment of pHelp-1 was then inserted into the XhoI/ApaI sites of these plasmids. Through this cloning step, T1 sequences harboring the distal exon 1 and 2.6 kb of 5′ flanking sequences were introduced into the constructs pHelp-wt and pHelp-TREmut to give rise to pF and pG, respectively.

To generate pS-G, the NdeI/BamHI fragment of pM1α-GH (33) containing a minimal rabbit β-globin promoter and a single binding site for the GATA transcription factors was cloned into the NheI and BglII sites of pSEAP2-Basic. The ends that were generated by restriction with NdeI and NheI were blunted for this cloning step.

The empty expression vector pXM was generated by digestion of pXM-mGATA-1 encoding the murine GATA-1 protein (61) with XhoI and religation.

Site-directed mutagenesis.

Introduction of point mutations into construct pA was performed by generating mutated EcoRI/XhoI fragments by PCR as described by Ho et al. (22). The mutated fragments were used to replace the EcoRI/XhoI fragment in pA. The subsequent insertion of the 6-kb EcoRI fragment (0.3 to 6.3 kb upstream of the distal exon 1) into the EcoRI site of the mutated plasmid pA resulted in the various mutated pC constructs. The mutation of each site was chosen such that a novel restriction site was introduced which allowed us to verify successful generation of the mutation. The following new restriction sites were introduced: dGATA, NheI; mGATA, BfrI; pGATA, ScaI; SP1, HindIII; and TRE, BamHI. The following primers were used to generate the mutations: 5′-pA (5′-GGATCCCCCGGGCTGCAGGAATTCGGTCTATCT-3′), pA-3′ (5′-AGATCTCTCGAGGTCGACGGTATCGAACCACCA-3′), dGmut-5′ (5′-CTTGAAGGTCCATGGCTAGCAGGGTAAAACTGGAAG-3′), dGmut-3′ (5′-CTTCCAGTTTTACCCTGCTAGCCATGGACCTTCAAG-3′), mGmut-5′ (5′-GTTCCTGTAAGTAACTCTTAAGGAACAGGAGGTGTT-3′), mGmut-3′ (5′-AACACCTCCTGTTCCTTAAGAGTTACTTACAGGAAC-3′), pGmut-5′ (5′-ACAGGAGGTGTTAGAAGTACTTGGCAACTGTATTGG-3′), pGmut-3′ (5′-CCAATACAGTTGCCAAGTACTTCTAACACCTCCTGT-3′), SP1mut-5′ (5′-CAGCCAGCAGGAATTAAGCTTGTTTTTTTGTTTTGA-3′), SP1mut-3′ (5′-TCAAAAGAAAAAAACAAGCTTAATTCCTGCTGGCTG-3′), TREmut-5′ (5′-GACAAACAGTAAAATGGATCCAGATGGTTAACAGCT-3′), and TREmut-3′ (5′-AGCTGTTAACCATCTGGATCCATTTTACTGTTTGTC-3′).

Transfections.

For transient transfections, NIH 3T3 cells were plated at a density of 1.5 × 10⁴ cells/cm² in 3.5-cm-diameter tissue culture dishes and grown for 24 h. Cells were transfected with 5 μg of total plasmid DNA by the Ca²⁺ phosphate precipitation method (64). Six hours after addition of the precipitate, cells were washed twice with phosphate-buffered saline (PBS) and incubated in DMEM containing 10% FCS for 40 h. For serum stimulation, transfected cells were grown for 16 h in DMEM–10% FCS and subsequently starved in DMEM–0.5% FCS for 24 h. Thereafter the medium was replaced with 1.5 ml of fresh DMEM–10% FCS, and the cells were maintained in this medium for an additional 24 h; 100 μl of the medium was then removed for SEAP activity measurements. The cells were rinsed twice with cold PBS, scraped from the plate in 1.5 ml of cold PBS, pelleted, resuspended in 100 μl of cold 40 mM Tris-HCl (pH 7.5)–1 mM EDTA–150 mM NaCl, and disrupted by three cycles of freeze-thawing. Cell debris were removed by centrifugation, and the supernatant was used to measure β-galactosidase activity as specified by Miller (40).

For stable transfections, NIH 3T3 cells were grown on 10-cm-diameter plates and transfected with 19 μg of the expression vector pXM-mGATA-1 and 1 μg of plasmid pSV₂neo, encoding the neomycin phosphotransferase gene (54). The precipitate was removed after 16 h, and neomycin-resistant clones were selected in medium supplemented with G418 (1 mg/ml).

Mast cells were washed once with cold electroporation buffer (IMDM-GLUTAMAX), pelleted, and resuspended at 2 × 10⁷ to 4 × 10⁷ cells/ml in electroporation buffer. Then 0.5-ml aliquots of the cell suspension were transferred to the electroporation cuvettes, 65 to 85 μg of total plasmid DNA was added and mixed with the cell suspension, and the mixture was incubated for 10 min at room temperature. The cells were electroporated once at 320 V and 960 μF, incubated for 15 min at 37°C, transferred into 5-cm-diameter culture dishes, and grown in 6 ml of complete IMDM for 42 h. Subsequently, the cells were harvested by centrifugation and resuspended in 4 ml of fresh medium; 1 ml of this cell suspension was removed and washed once with cold PBS, and β-galactosidase activity was determined as described above. The remaining cells (3 ml) were grown in 3.5-cm-diameter dishes for an additional 20 h. For Ca²⁺ ionophore stimulation, the cell suspension was split into two equal parts. One half was left untreated, while the other was stimulated with 1 μM A23187 for 20 h; 120 μl of the culture supernatant was then taken to measure SEAP activity.

SEAP assay.

SEAP activity was determined with a Phospha-Light chemiluminescent reporter gene assay kit as specified by the manufacturer (Tropix). The compositions of the buffers in this kit have not been specified. A 100- or 120-μl aliquot of cell culture medium was harvested and cleared by centrifugation at 12,000 × g for 10 s; 80 μl of the supernatant was then added to 240 μl of 1× dilution buffer, and endogenous alkaline phosphatase was heat inactivated at 65°C for 30 min. The tubes were cooled to room temperature, and 150 μl of the medium was transferred to fresh tubes (this was always done in duplicate); 100 μl of the assay buffer was added, and the mixture was incubated for 5 min at room temperature. Subsequently, 100 μl of reaction buffer was added, and the solution was mixed and immediately transferred into a small scintillation vial. After 30 min, the chemiluminescence was measured in a scintillation counter (Packard 1900 TR; TriCarb).

Northern blot analysis.

RNA was prepared as described by Chomczynski and Sacchi (6). Five or 7.5 μg of total RNA was denatured by glyoxylation, separated on 1% agarose gels (38), and transferred onto nylon membranes (GeneScreen Plus; Du Pont-NEN). The following DNA fragments were radiolabeled as described by Feinberg and Vogelstein (10, 11), purified over Nick columns (Sephadex G-50; Pharmacia), and used as hybridization probes: T1, a 1-kb HindIII fragment from plasmid pMV7TORF (23), spanning the entire T1 open reading frame present on the 2.7-kb mRNA; c-fos, a 1-kb PstI fragment from the plasmid pv-fos, harboring the cDNA of the FBJ murine leukemia virus-derived v-fos gene (8); c-jun, a 1.5-kb EcoRI/BamHI fragment from plasmid pTZ jun, containing 1.5 kb of c-jun cDNA; and mGATA-1, a 1.8-kb XhoI fragment from plasmid pXM-mGATA-1 (61).

Primer extension.

End-labeled oligonucleotide TG2 (GCTCTCTGAGGTAGGGTCCAGAAGAGAAATCAC) (1.5 × 10⁵ to 2 × 10⁵ cpm) was mixed with 10 μg of total RNA, and primer extension reactions were performed as described previously (13).

Immunoprecipitation of calcium ionophore-treated BMCMCs.

For metabolic labeling, two identical cultures of 10⁶ BMCMCs were starved for methionine and cysteine by incubation for 30 min at 37°C in labeling medium (DMEM without methionine and cysteine, 2% FCS, 2 mM l-glutamine). After starvation, the medium was replaced with 5 ml of fresh labeling medium, and 0.5 mCi of Tran³⁵S-label (ICN) was added per culture. For the stimulation with calcium ionophore, one BMCMC culture was supplemented with 0.5 μM A23187 (Calbiochem). The control culture was left untreated. After a 16-h incubation, the cell-free conditioned cell culture supernatants were collected. The remaining cells were washed three times with PBS and lysed in Nonidet P-40 (NP-40) lysis buffer (1% NP-40, 150 mM NaCl, 50 mM Tris-HCl [pH 8.0], 5 mM EDTA, 10 mM iodoacetamide) supplemented with a cocktail of protease inhibitors (Complete; Boehringer). The postnuclear lysates as well as the conditioned culture supernatants were precleared by incubation with 100 μl of protein G-Sepharose beads (Pharmacia) for 1.5 h to decrease nonspecific binding. Immunoprecipitation was carried out by adding 30 μl of protein G-Sepharose beads with 3 μg of anti-T1 MAb DJ8 (42) or IgG1 isotype control MAb and incubation for 6 h at 4°C. Immunoprecipitates were washed twice in NET-TON (0.65 M NaCl, 5 mM EDTA, 50 mM Tris-HCl, 0.5% Triton X-100, 1 mg of bovine serum albumin per ml, 0.05% sodium azide), twice in NET-T (150 mM NaCl, 5 mM EDTA, 50 mM Tris-HCl, 0.5% Triton X-100, 0.05% sodium azide), and once in distilled water. Beads were boiled in reducing Laemmli sample buffer, and the eluted proteins separated by sodium dodecyl sulfate–10% polyacrylamide gel electrophoresis (SDS-PAGE). The gel was rinsed in water, soaked for 20 min in 1 M sodium salicylate, dried, and fluorographed at −80°C.

Flow cytometry of calcium ionophore-treated BMCMCs.

For flow cytometry, BMCMCs were grown in six-well plates and stimulated with different concentrations of the calcium ionophore A23187 for 16 h. Cells were harvested, washed, and stained by adding 0.5 μg of MAb DJ8 per 10⁶ cells for 30 min at 4°C. This was followed by the addition of fluorescein isothiocyanate (FITC)-labeled goat anti-rat Ig serum (Southern Biotechnology). Background staining was determined by addition of 0.5 μg of IgG1 isotype control as the first antibody. Dead cells were detected by propidium iodide staining and excluded from the analysis. After washing and resuspending the cells in fluorescence-activated cell sorting buffer (PBS, 10% FCS, 0.1% sodium azide), 10,000 forward scatter/side scatter-gated viable cells were acquired and analyzed on a Becton Dickinson FACS Calibur flow cytometer.

IgE-DNP stimulation.

Mast cells (2 × 10⁶ cells/ml) were incubated in complete IMDM with anti-2,4-dinitrophenol (DNP) IgE MAbs (0.5 μg/ml; Sigma Immuno Chemicals) for 24 h. For determination of serotonin release, 1 μCi of 5-hydroxy [G-³H]tryptamine creatine sulfate (Amersham Corp.) per ml was added for the final 5 h of IgE sensitization. The cells were then washed twice with complete IMDM, resuspended at 2 × 10⁶ cells/ml, and exposed to DNP-derivatized human serum albumin (0.05 μg/ml; (Sigma Immuno Chemicals) at 37°C for the times indicated. Released radioactivity in supernatant fractions (100 μl) was measured in a scintillation counter. To determine the total amount of incorporated radioactivity, cells were lysed with 0.1% Triton X-100.

RESULTS

Calcium influx results in the downregulation of T1M and in the strong induction of T1S in mast cells.

To test whether T1 gene expression is affected by treatments which promote the degranulation of mast cells, we stimulated BMCMCs with different concentrations of the calcium ionophore A23187 for 6 h. RNA was extracted and subjected to Northern blot analysis (Fig. 1A). The long 5-kb transcript predominates in untreated cells. However, upon A23187 treatment, the 5-kb mRNA disappears while large amounts of the 2.7-kb T1 mRNA accumulate. The strongest effect was observed at 1 μM A23187. A23187 concentrations above 3 μM are toxic to mast cells. This toxicity is the likely cause for the low amount of accumulated T1 mRNA in mast cells which were treated with high concentrations of A23187.

FIG. 1.

Northern blot analysis of mast cells stimulated with the calcium ionophore A23187 or IgE-DNP. (A) Aliquots of 5 μg of total RNA from BMCMCs incubated for 6 h with the indicated concentrations of A23187 were analyzed on a Northern blot. The filter was hybridized with a probe representing the whole T1 open reading frame of the 2.7-kb mRNA. The positions of long and short T1 transcripts are indicated. The ethidium bromide-stained gel is shown at the bottom to demonstrate integrity of the RNA and equal loading. (B) BMCMCs were stimulated for the indicated times (hours) with either 1 μM A23187 or DNP-human serum albumin following preincubation for 24 h with a monoclonal mouse anti-DNP IgE. NIH 3T3 cells were either exponentially growing (exp.) or serum stimulated (s.-stim.) for 6 h (see Materials and Methods). Total RNA was collected, and 5 μg was subjected to Northern blot analysis with a T1 probe. The bottom panel depicts the ethidium bromide-stained gel. (C) To control for efficient FcɛRI cross-linking in the experiment represented in panel B, we measured mast cell degranulation. ³H-serotonin-labeled BMCMCs either were incubated with monoclonal mouse anti-DNP IgE and subsequently stimulated with DNP-human serum albumin (HSA) (IgE/DNP) or were treated with A23187. Release of ³H-serotonin was measured and compared to the total ³H-serotonin content of cells which was determined in a Triton X-100 cell lysate. Stimulation with either IgE or DNP-HSA alone resulted in very low ³H-serotonin secretion.

Since calcium ionophore treatment can mimic some aspects of the physiological IgE-dependent activation of mast cells, we next compared T1 mRNA levels of BMCMCs which were stimulated either with A23187 or by cross-linking their FcɛRI. To this end, BMCMCs were preincubated with an IgE MAb directed against the hapten DNP. The subsequent addition of DNP-conjugated human serum albumin mediated FcɛRI cross-linking and resulted in mast cell degranulation. As seen earlier, calcium ionophore stimulation evoked a dramatic upregulation of the 2.7-kb transcript and downregulation of the 5-kb mRNA (Fig. 1B). The level of the short T1 transcript in mast cells exceeded even that in serum-stimulated fibroblasts and remained high for at least 16 h. Similarly, cross-linking of the FcɛRI resulted in accumulation of the 2.7-kb mRNA. However, the induction was weaker and transient in nature.

To confirm that antigen stimulation and Ca²⁺ ionophore treatment indeed activated the BMCMCs, we measured the capacity of these cells to release preloaded ³H-serotonin (7) in response to these treatments (Fig. 1C). We measured the total amount of ³H-serotonin taken up by the mast cells after lysing the cells with the detergent Triton X-100. A23187 and antigen stimulation released 80 and 47%, respectively, of the total amount of incorporated serotonin. Thus, both modes of activation mediated degranulation. However, the more efficient stimulation of mast cells by calcium ionophore treatment might be partially responsible for the stronger upregulation of the short T1 transcript under these conditions (Fig. 1B).

We next tested whether the accumulation of the short T1 transcript and the disappearance of the long T1 transcript in response to Ca²⁺ ionophore treatment of BMCMCs is reflected by a corresponding change in the pattern of T1 protein synthesis and performed an immunoprecipitation experiment using the recently generated T1-specific MAb DJ8 (42). From lysates of unstimulated mast cells, DJ8 but not the IgG1 isotype control antibody precipitated the 110- to 120-kDa membrane-associated T1M protein (Fig. 2A). In contrast, the cell lysates of A23187-treated mast cells contained barely detectable amounts of T1M. No T1 protein was present in the cell-free supernatant of untreated BMCMCs. However, the stimulation with calcium ionophore resulted in stimulation of T1 synthesis, as evidenced by the detection of large amounts of T1S which appear as a broad band of about 45 to 65 kDa (Fig. 2A). Smaller amounts of T1S were also detectable in lysates of A23187-stimulated but not of unstimulated BMCMCs. This probably represents the intracellular pool of T1S in the endoplasmic reticulum and the Golgi apparatus.

FIG. 2.

Calcium ionophore stimulation of BMCMCs leads to the downregulation of T1M and the induction of T1S expression. (A) Immunoprecipitation. NP-40 cell lysates of BMCMCs which were either treated with 0.5 μM A23187 for 16 h or left untreated, as well as cell-free supernatants (SN) of A23187-treated and untreated BMCMCs, were used for immunoprecipitation. For metabolic radioactive labeling, cells were grown in the presence of [³⁵S]methionine and [³⁵S]cysteine. Immunoprecipitations (IPP) were performed with the anti-T1 MAb DJ8 or an IgG1 isotype control MAb. Precipitated proteins were separated by SDS-PAGE. Positions of T1M and T1S are indicated. (B) Increasing concentrations of the calcium ionophore A23187 induced the downregulation of T1M on BMCMCs in a dose-dependent manner. BMCMCs were incubated for 16 h with different concentrations of A23187. Surface expression levels of T1M protein were determined by flow cytometry after staining with MAb DJ8-FITC. Mean geometric fluorescence intensities are plotted against the indicated concentrations of A23187. Background binding of IgG1-FITC isotype control MAb is shown as a dotted line.

This finding was further substantiated by flow cytometric analysis (Fig. 2B). BMCMCs were incubated with increasing concentrations of the calcium ionophore A23187, and the surface expression levels of T1M were quantified by staining with MAb DJ8. The level of T1M surface expression decreased with increasing concentrations of A23187 in a dose-dependent fashion. The cells treated with the highest A23187 concentration (0.5 μM) contained only about 1/10 the amount of T1M present in the untreated cells. We conclude that calcium ionophore-activated mast cells downregulate membrane-associated T1M and, correspondingly, secrete large amounts of T1S.

Accumulation of the short T1 mRNA in BMCMCs results from transcriptional activity and can be partly blocked by the immunosuppressant cyclosporin A.

The accumulation of the 2.7-kb T1 mRNA in A23187-treated mast cells could be due to an enhanced transcription rate of the T1 gene or to the stabilization of low amounts of this transcript that escape detection in untreated cells (65). We performed a pulse-chase experiment to distinguish between these two regulatory mechanisms. The 2.7-kb T1 mRNA was allowed to accumulate in BMCMCs in response to the treatment with A23187 for 3 h. Subsequently, actinomycin D was added to block further transcription, and the levels of T1 mRNA were measured by Northern blot analysis in cells that were harvested 1, 3, and 5 h later (Fig. 3, lanes 9 to 11). In parallel, the same experiment was performed, except that A23187 was removed prior to the addition of actinomycin D (lanes 12 to 14). The amounts of T1S mRNA detected were similar in the two groups. The result indicates that the short T1 transcript is stable in the presence as well as in the absence of A23187. Thus, it seems unlikely that Ca²⁺ ionophore-mediated accumulation of the short T1 transcript is caused by mRNA stabilization. This is concordant with the observation that actinomycin D pretreatment blocked A23187 mediated increases of T1 mRNA (lanes 15 and 16). From this, we conclude that the accumulation of the 2.7-kb T1 mRNA in response to Ca²⁺ ionophore treatment is the result of transcriptional activation.

FIG. 3.

Accumulation of the 2.7-kb T1 transcript in A23187-treated mast cells requires ongoing transcriptional activity. BMCMCs were either left untreated (lane 1) or treated for the indicated times (hours) with A23187 (lanes 2 and 3), actinomycin D (Act.D; lanes 4 and 5), or cyclosporin A (CsA; lanes 6 to 8). Lanes 9 to 14, cells were pretreated with A23187 for 3 h before the addition of actinomycin D for 1 h (lane 9), 3 h (lane 10), or 5 h (lane 11). In parallel, BMCMCs were treated identically except that the calcium ionophore was removed by washing the cells twice with PBS before the addition of actinomycin D (lanes 12 to 14). Lanes 15 to 18, mast cells were preincubated for 15 min with actinomycin D (lanes 15 and 16) and cyclosporin A (lanes 17 and 18), followed by calcium ionophore stimulation for 3 and 6 h. Total RNA was harvested, and 7.5-μg aliquots were subjected to Northern blot analysis using T1 cDNA as the hybridization probe. Bottom panel, ethidium bromide-stained rRNA.

Cyclosporin A is an immunosuppressant which operates through the inhibition of Ca²⁺-induced dephosphorylation of proteins such as NF-AT, IκB, Bcl-2, and NO synthase by calcineurin (46, 50). Cyclosporin A partially blocked A23187-induced accumulation of the 2.7-kb T1 mRNA (Fig. 3, lanes 17 and 18). Hence, we conclude that the calcineurin pathway is involved in the calcium ionophore-dependent T1 gene expression in mast cells.

Transcription of both T1 mRNAs is initiated at the distal promoter in mast cells.

T1 gene expression is initiated at one of two alternative first exons (Fig. 4A). Fibroblasts synthesize predominantly the 2.7-kb T1 mRNA, but substantial amounts of the 5-kb T1 mRNA are observed under some conditions. Both transcripts are exclusively initiated at the proximal exon 1 in these cells. In contrast, unstimulated mast cells produce mainly the 5-kb T1 mRNA, which is initiated at the distal exon 1. As shown in Fig. 1B, stimulation of these cells with calcium ionophores evokes a strong upregulation of the short T1 transcript, and we wondered whether transcriptional initiation still occurred at the distal promoter. A primer extension experiment was performed to analyze which of the two alternative exons 1 is used for the synthesis of the 2.7-kb T1 transcript in A23187 stimulated BMCMCs. Oligonucleotide TG2, which is complementary to the 5′ region of exon 2 (Fig. 4A), was extended on RNA derived from BMCMCs that were either left untreated (Fig. 4B, lane 1) or stimulated with calcium ionophore (lanes 2 and 3). The extension product indicative for transcriptional initiation at the proximal promoter was obtained with RNA from serum-stimulated NIH 3T3 cells (lane 4). Primer extension on all mast cell-derived RNA resulted exclusively in the two extension products characteristic for initiation at the distal promoter (lanes 1 to 3). Thus, transcription of both the long and short T1 mRNAs is initiated at the distal promoter in mast cells. Hence, transcription initiation occurs in a strictly cell-type-specific manner at the proximal and distal promoters in fibroblasts and mast cells, respectively.

FIG. 4.

Exclusive usage of the distal T1 promoter in mast cells. (A) Genomic organization of the 5′ part of the T1 gene. Only exons 1 to 5 are shown. Empty, striped, and filled boxes represent the distal exon 1 (d1), the proximal exon 1 (p1), and the next four exons, respectively; the ellipse depicts the enhancer (Enh) (60). The translation start codon (ATG), the position of the oligonucleotide used for primer extension (TG2), and the expected products of primer extension reactions are indicated. (B) Products of primer extension reactions with oligonucleotide TG2 and 10 μg of total RNA isolated from serum-stimulated NIH 3T3 fibroblasts (lane 4), untreated BMCMCs (lane 1), and BMCMCs treated for 3 h (lane 2) and 6 h (lane 3) with 1 μM A23187. A sequencing reaction of an unrelated DNA fragment was run in parallel on the polyacrylamide gel as a size marker. Open and filled arrowheads indicate the positions of the extension products which are characteristic for transcription start at the proximal and distal promoters, respectively.

Mast cell-specific T1 gene expression is not dependent on the T1 enhancer and c-Fos.

We have previously characterized an enhancer element located at a position 3.6 kb upstream of the proximal and 6.9 kb downstream of the distal promoter (60). Within this enhancer, we have identified a centrally located TRE, a binding site for transcription factors of the AP-1 family. We demonstrated by mutational analysis that this sequence motif is essential for T1 gene expression in fibroblasts. Moreover, elevated levels of c-Fos or FosB were shown to be sufficient to strongly induce the T1 gene in these cells.

Calcium ionophore treatment leads to the induction of the c-fos and c-jun genes in many cell systems (5). Likewise, we found that c-fos and c-jun mRNA levels strongly and rapidly increase in A23187-treated BMCMCs (Fig. 5A). The extent of c-fos and c-jun mRNA accumulation exceeds even that in serum-stimulated NIH 3T3 cells. This observation and the fact that c-Fos regulates T1 gene expression in fibroblasts prompted us to investigate whether the large increase of the short T1 transcript in A23187-induced mast cells is due to AP-1-mediated activation of the T1 enhancer. Therefore, we cloned several T1 gene fragments into a promoterless reporter plasmid encoding SEAP (Fig. 5B). These constructs contain 0.3, 2.6, and 6 kb of 5′-flanking promoter sequence and 16 bp of the distal exon 1. Constructs pD and pE additionally harbor a portion of intron 1 including the enhancer element and sequences upstream of the translation initiation site in exon 2. These plasmids were transiently introduced into BMCMCs by electroporation. Promoter activity was evaluated in untreated and A23187-stimulated cells (Fig. 5C). We observed that 310 bp of the 5′ region flanking the distal exon 1 (pA [Fig. 5B]) was sufficient to stimulate reporter gene expression and in particular to mediate calcium ionophore-triggered gene upregulation. The longer T1 5′-flanking sequences in pB and pC gave rise to stronger gene expression. However, the extents of A23187-mediated induction of transcriptional activity were similar for all three constructs (Fig. 5C, bottom). Likewise, the presence of the enhancer element in the first intron (pD and pE [Fig. 5B]) influenced neither A23187-mediated T1 reporter gene expression nor basal promoter activity.

FIG. 5.

The T1 enhancer does not affect transcription from the distal promoter. (A) Total RNA from BMCMCs and NIH 3T3 cells stimulated for the indicated times (hours) with 1 μM A23187 and serum, respectively, was subjected to Northern blot analysis. Two identical blots were prepared with 7.5 μg of RNA. The filters were independently hybridized with a c-fos- and a c-jun-specific probe. Bottom panel, ethidium bromide-stained rRNA from one of the gels. (B) Genomic organization of the 5′ part of the T1 gene as depicted in Fig. 4A. The seven reporter gene constructs pA to pG are shown below. Open bars represent T1 sequences fused to SEAP. The splicing patterns of the distal and proximal exon 1 to exon 2 are depicted for pD to pG. Filled and open ellipses represent the enhancer (Enh) in its unmutated form and with a mutation in the TRE (TRE mut), respectively. (C) Mast cells were cotransfected with equimolar amounts of the empty vector (−), the T1-SEAP reporter constructs pA to pE (B), and plasmid CMV-LacZ; 44 h later, the transfected cells were split into two equal parts and incubated in fresh medium for an additional 20 h either in the presence or in the absence of 1 μM A23187. SEAP activity was measured and normalized for β-galactosidase activity to correct for variations in transfection efficiency. Results are the averages of three independent experiments, and the standard deviations are given by the error bars. Fold stimulation in response to Ca²⁺ ionophore treatment is shown below the histogram. (D) pF and pG (B) and the empty vector were transfected into mast cells and NIH 3T3 fibroblasts. A23187 treatment of BMCMCs and SEAP assays were done as described for panel C; 16 h after transfection, the fibroblasts were serum starved for 24 h and subsequently serum stimulated for 24 h. SEAP activity was determined and normalized for β-galactosidase activity. The values obtained with the stimulated unmutated construct pF were taken as 100%. The results are the averages of at least three independent experiments.

To further substantiate the finding that the enhancer does not influence T1 gene expression in BMCMCs, we generated a reporter construct which contains both the distal and the proximal first exons, which allowed us to study the influence of the enhancer element in mast cells as well as in fibroblasts (pF [Fig. 5B]). In addition, we produced the same construct with an inactivating TRE mutation within this enhancer (pG [Fig. 5B]). These two plasmids were transiently transfected into BMCMCs and fibroblasts. In accordance with previous results (60), we observed that the mutation of the TRE resulted in a considerable reduction of reporter gene expression in fibroblasts. In contrast, the TRE mutation affected basal gene expression in mast cells only marginally and had no influence on calcium ionophore-mediated gene induction (Fig. 5D). We therefore conclude that the upregulation of the T1 gene in response to Ca²⁺ ionophore treatment is not mediated through the enhancer element. The enhancer acts cell type and promoter specifically, exerting its influence selectively on the proximal promoter in fibroblasts.

Sequence analysis of the 310-bp minimal T1 distal promoter element which is sufficient to confer Ca²⁺ ionophore stimulation of the T1 gene revealed the presence of a sequence element that resembles a TRE (Fig. 6A). We mutated this putative TRE in pC and analyzed the effect of the mutation in BMCMCs in the presence or absence of A23187. As shown in Fig. 6B, the mutation abrogated neither basal nor Ca²⁺ ionophore-induced T1 reporter gene expression. Hence, neither the enhancer-TRE nor this putative TRE in the 5′ region of the distal exon 1 is involved in Ca²⁺ ionophore-triggered T1 gene stimulation in mast cells.

FIG. 6.

c-Fos is not required for T1 gene expression in A23187-stimulated or unstimulated mast cells. (A) Mutational analysis of putative transcription factor binding sites in the 5′ region of the distal exon 1. The wild-type (wt) and mutated (mut) putative DNA binding sequences are depicted (capital letters, core sequences; lowercase letters, flanking nucleotides; superscript letters, mutated nucleotides). dGATA, mGATA, and pGATA, distal, medial, and proximal GATA binding sites; +1, transcription start site. (B) BMCMCs were transfected with the empty SEAP reporter vector (−), the T1-SEAP reporter construct pC (B), and pC with a mutated TRE (A). A23187 stimulation of BMCMCs and SEAP assays were done as described in the legend to Fig. 5C. The calcium ionophore-induced increases of the SEAP activity for the three constructs are given. (C) A 5-μg aliquot of total RNA extracted from BMCMCs which were obtained from a wild-type mouse (lane 1) and a c-fos^−/− mouse (lanes 2 to 7) was analyzed on a Northern blot. The c-fos knockout mast cells were calcium ionophore treated for the time periods (hours) indicated. The filter was hybridized with a T1-specific probe. Bottom panel, ethidium bromide-stained rRNA.

If this assumption is correct, the accumulation of the short T1 transcript in A23187-treated BMCMCs derived from c-fos knockout mice should still occur. Such mice are osteopetrotic and have only a rudimentary bone marrow (18). Nevertheless, it was possible to obtain pure mast cell cultures by incubating suspension cells derived from whole meshed femurs in IL-3-containing medium. Northern blot analysis revealed that the 2.7-kb T1 mRNA accumulated in response to A23187 with similar kinetics and to comparable levels in c-fos^−/− and in c-fos^+/+ BMCMCs (compare Fig. 6C and 1B). Hence, basal as well as A23187-induced T1 synthesis can occur in the absence of c-Fos in mast cells. However, we cannot exclude that another member of the fos family is involved in T1 gene expression in mast cells.

Taken together, these results lead us to conclude that in contrast to fibroblasts, neither the enhancer element nor c-Fos protein influences T1 synthesis in mast cells.

The mast cell-specific promoter is regulated by GATA transcription factors.

Sequence analysis of the minimal, 310-bp-long mast cell-specific T1 promoter revealed the presence of three GATA elements (Fig. 6A). The GATA consensus motif (T/A)GATA(A/G) is the recognition site of GATA transcription factors. GATA-1 (34, 68), GATA-2, and GATA-3 (68) are expressed in several mouse and rat mast cell lines and were shown to be instrumental in mast cell-specific expression of several genes. Hence, GATA factors are possibly involved in the regulation of the T1 gene transcription in mast cells.

To investigate the role of GATA transcription factors in mast cell-specific T1 gene expression, we transiently transfected the T1 reporter constructs pA and pC (Fig. 5B) either alone or together with a GATA-1 expression vector or the empty vector pXM into BMCMCs. GATA-1 overexpression strongly enhanced T1 reporter gene activity (Fig. 7A). The extents of stimulation were similar for pA and pC, indicating that the essential GATA binding sites are within the 310 bp of the T1 promoter present in pA. Reporter gene expression was not influenced by the control vector pXM.

FIG. 7.

GATA-1 is a key regulator of the distal promoter. (A) Cotransfection of the T1-SEAP reporter constructs pA and pC (Fig. 5B) as well as the empty vector with the expression plasmid pXM-mGATA-1 (encoding the murine GATA-1 transcription factor) or the empty vector pXM. BMCMCs were electroporated, the medium was changed 44 h later, and SEAP activity was measured after an additional 20 h. Plasmid CMV-LacZ was included in each transfection, and β-galactosidase activity was used to correct for variations in transfection efficiency. Results are the averages of three independent experiments. Fold induction of promoter strength caused by cotransfection of the reporter constructs with the expression vector pXM-mGATA-1 compared with transfection of the reporter constructs alone is given. (B) Transient transfections of mast cells with the T1-SEAP reporter construct pC (Fig. 5B) either unmutated or with the indicated mutations (Fig. 6A). The medium was changed 44 h after transfection; SEAP activity was determined 20 h later and normalized as described before. Values are the averages of three independent experiments. The fold reduction of the distal promoter strength caused by the individual mutations is shown at the bottom.

This finding suggests that one or several of the three putative GATA elements identified in the 310-bp minimal promoter mediate T1 gene expression. To evaluate the contribution of each GATA element, we introduced point mutations into all three GATA sites in pC (Fig. 6A). Analysis of T1 reporter gene expression in transiently transfected BMCMCs revealed that a mutation of the middle GATA element completely inactivated the distal promoter (Fig. 7B). Mutations of the distal and proximal GATA sites reduced promoter strength as well, although to a lesser extent. This finding taken together with the previous observation of enhanced T1 reporter gene expression in the presence of increased GATA-1 levels demonstrates that mast cell-specific T1 gene expression is dependent on a GATA transcription factor(s). Not only constitutive T1 gene activity but also calcium ionophore-stimulated T1 gene expression is dependent on GATA proteins (data not shown).

Others have shown that SP-1, binding in close proximity to GATA transcription factors, is essential for efficient gene expression. A sequence element resembling an SP-1 binding site can be found centered at position −251 of the mast cell-specific T1 promoter (Fig. 6A). We introduced point mutations to evaluate whether this element is necessary for T1 synthesis. The mutation of the putative SP-1 site reduced basal T1 gene activity twofold (Fig. 7B). Thus, the SP-1 element seems to be involved in but not critical for T1 gene expression.

GATA-1 expression is not sufficient to activate the mast cell-specific distal promoter in NIH 3T3 fibroblasts.

In view of the finding that GATA transcription factors are essential for T1 gene expression in mast cells, the complete lack of distal promoter usage in fibroblasts could be explained by the absence of GATA factors in these cells. To test this hypothesis, we stably introduced the GATA-1 expression plasmid pXM-mGATA-1 (61) into NIH 3T3 cells. Several transfected cell clones which express the introduced gene under normal growth conditions at levels comparable to that in BMCMCs were identified (Fig. 8A). In contrast, no GATA-1 mRNA was found in untransfected NIH 3T3 cells.

FIG. 8.

GATA-1 expression in fibroblasts is not sufficient to activate the mast cell-specific promoter. (A) Northern blot analysis of different fibroblast cell clones stably transfected with the GATA-1 expression vector (pXM-mGATA-1 [61]). Aliquots of 5 μg of total RNA from transfected cell clones, mast cells, and exponentially growing (exp.) as well as serum-stimulated (s.stim.) NIH 3T3 cells were analyzed on a Northern blot by subsequent hybridization with a GATA-1-specific probe and after stripping with a T1-specific probe. The lower panel depicts the ethidium bromide-stained gel. (B) The empty vector (pS2/B) and the construct pS-G were transiently transfected into NIH 3T3 cells and the pXM-mGATA-1-transfected cell clones 13 and 33. SEAP activities and β-galactosidase activities were measured as described in Materials and Methods. The results are the averages of three independent experiments. (C) Products of primer extension reactions with oligonucleotide TG2 (Fig. 4A) and 10 μg of total RNA isolated from the GATA-1-overexpressing clones 13, 27, and 33 (lanes 1 to 3), from exponentially growing and 6-h serum-stimulated NIH 3T3 cells (lanes 4 and 5), and from BMCMCs (lane 6). Open and filled arrowheads mark the positions of the extension products which are characteristic for initiation of transcription at the proximal and distal promoters, respectively.

To ascertain that functional GATA-1 protein is present in these clones, we transiently transfected two of them as well as the parental NIH 3T3 cells with a reporter plasmid harboring a minimal human β-globin promoter containing a GATA-1 binding site in front of the SEAP gene (pS-G). Reporter gene activity was approximately four times higher in clones 13 and 33 than in parental NIH 3T3 cells, indicating that functional GATA-1 protein is present in these cells (Fig. 8B).

We next performed a primer extension experiment to investigate whether expression of GATA-1 protein is sufficient to activate the distal T1 promoter in fibroblasts. The extension products indicative for initiation of transcription at the distal promoter are found only with RNA from BMCMCs (Fig. 8C, lane 6), whereas primer extension on all fibroblast-derived RNA resulted exclusively in the extension product characteristic for initiation at the proximal promoter (lanes 1 to 5). Thus, no transcription start occurs at the mast cell-specific promoter in NIH 3T3 clones expressing functional GATA-1 protein.

DISCUSSION

Ca²⁺ ionophore-induced shift from T1M to T1S in mast cells.

With the help of recently generated MAbs directed against the extracellular domain of T1, we have previously investigated the expression pattern of T1M (42, 43). These studies revealed that mast cells are the only cells which express T1M within the hematopoietic system. The present study was undertaken to analyze the molecular mechanism of T1 gene expression in mast cells and to test whether the activation of mast cells influences T1 gene activity. We observed that Ca²⁺ ionophore treatment, which mimics some aspects of antigen-mediated mast cell activation and leads to efficient degranulation, results in the depletion of T1M and secretion of the soluble receptor. Recently, we and others have observed that the proinflammatory cytokines IL-1 and TNF-α stimulate the synthesis and secretion of T1S in fibroblasts (27, 29). It is interesting that mast cells store in their granules TNF-α which is released upon activation. Similarly, they synthesize and secrete IL-1 in response to stimulation. The release of these proinflammatory cytokines by activated mast cells could possibly trigger the secretion of soluble T1 from neighboring fibroblasts. Thus, mast cell activation might result in the local accumulation of T1S and the disappearance of T1M from mast cells. We assume that T1S acts antagonistically to cell-bound T1 by binding the ligand without inducing intracellular signaling. Our studies do not provide evidence that T1 is functionally involved in mast cell stimulation. However, they lead to the speculation that T1 signaling needs to be dampened after the onset of mast cell activation. This is reminiscent of the IL-1 system, where several mechanisms have evolved to counteract ligand stimulation (35). A decoy IL-1R which binds IL-1α and IL-1β without stimulating intracellular signaling exists in membrane-anchored and secreted forms. Moreover, many cells produce and secrete the IL-1 antagonist IL-1ra, which binds with high affinity to the IL-1R type I but does not stimulate signaling. Vaccinia viruses secrete an IL-1R-like protein which binds IL-1β with high affinity and presumably acts as a ligand sink. This results in a diminished systemic acute-phase response to infection and modulates the severity of the disease (1, 55).

Accumulation of the 2.7-kb mRNA in response to enhanced intracellular Ca²⁺ is independent of c-Fos and the enhancer and is sensitive to cyclosporin A.

We have investigated whether Ca²⁺ ionophore-mediated T1 gene upregulation is a consequence of elevated c-Fos levels. We considered this mode of T1 gene activation likely because we and others have previously observed that the murine T1 gene and its rat homolog fit-1 are subject to regulation by c-Fos (2, 23) and because Ca²⁺ ionophore stimulation results in strongly enhanced c-Fos levels. However, we found that the TRE within the T1 enhancer which is essential for gene activation in fibroblasts as well as a putative AP-1 binding site in the minimal mast cell-specific promoter are not required for T1 gene upregulation in response to Ca²⁺ ionophore treatment. This is in line with the finding that T1 gene expression in mast cells is not altered in c-fos^−/− mice.

One of the enzymes activated by elevated cytoplasmatic Ca²⁺ is the Ca²⁺-calmodulin-dependent phosphatase calcineurin, a primary target of the immunosuppressive drug cyclosporin A (46). Activated calcineurin dephosphorylates the transcription factor NF-AT and thereby allows its translocation into the nucleus, where it recognizes specific DNA sequence elements and triggers gene activation. We found that cyclosporin A reduced Ca²⁺ ionophore-mediated T1 gene activation. There is no sequence motif in the mast cell-specific promoter which resembles a classical binding site for NF-AT. However, we cannot exclude that NF-AT in a heteromultimeric complex with other transcription factors recognizes a novel sequence motif in the T1 promoter. Thus, further mutational analyses are needed to determine whether NF-AT is involved in T1 gene expression.

Differential 3′ processing is not dependent on promoter usage.

Unstimulated mast cells almost exclusively produce the 5-kb T1 transcript, whereas fibroblasts predominantly synthesize the 2.7-kb T1 mRNA. Studying the expression of the rat T1 gene homolog, fit-1, Bergers et al. proposed the interesting model that transcript size, i.e., poly(A) site selection, correlates with transcription initiation at the mast cell- or fibroblast-specific promoters (2). We have recently questioned this model by showing that in fibroblasts, both the 2.7-kb T1 mRNA which predominates in these cells and the 5-kb T1 transcript which accumulates under certain conditions are initiated exclusively at the proximal, fibroblast-specific promoter (13). Here, we have further substantiated this finding by demonstrating that the 5-kb T1 transcript which predominates in untreated mast cells as well as the 2.7-kb T1 mRNA which is the only form of the T1 transcript in long-term Ca²⁺ ionophore-treated mast cells are initiated at the distal, mast cell-specific promoter. Thus, at least in mice, we find no correlation between transcript size and the site of transcription initiation.

The 2.7-kb T1 mRNA arises by transcription termination after exon 8. In untreated mast cells, this poly(A) site is ignored and transcription proceeds efficiently past exon 9, which results in splicing within exon 8. Differential poly(A) site usage has been well studied in the B-lymphocyte lineage, where a switch from the membrane-anchored to the secreted form of Igs is dictated by poly(A) site selection. The regulatory component in this system is the polyadenylation factor CstF-64, whose level of expression determines the choice of poly(A) site usage (56). Another regulatory mechanism has recently been described. U1 snRNP binding to a splice site adjacent to the late polyadenylation site of bovine papillomavirus inhibits RNA processing at this site and thereby represses late gene expression at early times of infection (19). We have shown here that the choice of T1 gene poly(A) site usage can be easily manipulated in mast cells. T1 gene expression in these cells therefore represents an attractive model system to study the regulatory mechanisms of differential poly(A) site selection.

GATA proteins are required for T1 gene expression in mast cells.

GATA-1 overexpression in mast cells resulted in enhanced T1 gene expression. The complete inactivation of the T1 promoter through the introduction of a point mutation in the middle GATA sequence element further demonstrated that a GATA transcription factor is crucial for T1 gene expression in mast cells. GATA proteins are key regulators of hematopoiesis and play an important role in the control of gene expression in erythrocytes, megakaryocytes, and mast cells (62). Among the GATA transcription factors, GATA-1 and GATA-2 are abundantly expressed in mast cells (34, 68) and are critical for the expression of the genes encoding carboxypeptidase A (68), IL-4 (21), and baboon chymase (31) in mast cells. The promoter regions of other mast cell-specific genes such as those encoding mouse mast cell proteases and IgE receptor β chain contain putative GATA binding sites, but their importance in tissue-specific expression has not yet been demonstrated. The observation that the experimental overexpression of GATA-1 results in enhanced T1 gene expression suggests that the level of GATA-1 limits the extent of T1 promoter activity in untreated mast cells. T1 gene upregulation in response to Ca²⁺ ionophore treatment is unlikely due to increased GATA-1 levels, as we have not observed an elevation of GATA-1 mRNA amounts under these conditions (data not shown). However, we cannot exclude the possibility that calcium ionophore treatment results in elevated GATA-1 protein levels in the nucleus due to increased protein stability or enhanced nuclear translocation.

Ectopic expression of GATA-1 in fibroblasts is insufficient to redirect transcription initiation to the mast cell-specific promoter. Several mechanisms could explain this finding. Additional transcription factors might be required which are present in mast cells but not in fibroblasts. Alternatively, a specific repressor might prevent distal promoter usage in fibroblasts.

GATA proteins were shown to cooperate with other transcription factors, including SP-1 and AP-1, in promoter activation (17, 30, 37, 39). Putative binding sites for these two transcription factors were found near the GATA sites in the mast cell-specific T1 promoter. However, point mutations within the putative TRE had no effect on promoter strength, and mutations within the putative SP-1 site reduced promoter activity only slightly.

We have previously identified a TRE and three E boxes within the enhancer element which are essential for T1 gene expression in fibroblasts (23, 60). Our present study revealed that this enhancer is not required in mast cells. Thus, T1 gene activity in fibroblasts and mast cells at a proximal and a distal promoter, respectively, is regulated very differently. c-Fos is an important mediator of T1 gene expression in fibroblasts but not in mast cells, whereas a GATA transcription factor is essential for T1 gene activity in mast cells but not in fibroblasts.