Mast Cells Express Novel Functional IL-15 Receptor α Isoforms¹ Free

Interleukin-15 was discovered through its capacity to substitute IL-2-stimulatory functions on T cells (1, 2). IL-15 is a potent growth factor for T and B lymphocytes and NK cells (2, 3, 4, 5, 6), a chemoattractant for T cells (7), and an activator of the cytolytic program in T and NK cells (2, 8). In addition, IL-15 is widely recognized as a potent antiapoptotic agent, which inhibits apoptosis of activated T and B cells, keratinocytes, and melanoma cells in vitro, and protects mice from Fas-induced hepatic failure and multisystem apoptosis in vivo (9, 10, 11, 12). IL-15 can also serve as a growth factor for mast cells (13, 14, 15). In rather high concentrations in vitro, it supports the proliferation of bone marrow-derived mast cells (BMMCs)⁴ (14), protects them from apoptosis by inducing Bcl-x_L expression (13), and induces production of IL-4 (15). IL-15 and IL-15Rα knockout mice display a marked reduction in numbers of NK cells and memory phenotype CD8⁺ T lymphocytes, and a distinct population of intestine intraepithelial lymphocytes, thus reflecting an important role for IL-15 in the development and/or survival of these cells (16, 17).

Both IL-15 and IL-2 bind to a heterotrimeric receptor complex, which shares the IL-2Rβ and IL-2Rγ/common γ (γ_c) chains (18, 19, 20, 21), but has a unique α-chain that recognizes only its cognate cytokine (21, 22). Although IL-2 and its specific IL-2Rα chain have restricted expression on lymphoid cells, IL-15 and its specific high affinity receptor IL-15Rα are transcribed by a broad variety of different tissues and cells, including activated macrophages, keratinocytes, muscle cells, kidney, and endothelial and neural cells (8, 9, 23, 24, 25). Furthermore, IL-15 is expressed in a functionally active, membrane-bound form on monocytes (26). Because IL-15Rα alone binds IL-15 with high affinity (K_d ∼ 10⁻¹¹ M) (8, 22) compared with the low affinity (K_d ∼ 10⁻⁹ M) association between IL-2Rα and IL-2, it is reasonable to suggest that IL-15Rα subunit might be responsible for the differential functional effects of IL-15 and IL-2 on cells of the same type. It has been assumed that the IL-15Rα chain, like IL-2Rα, is not capable of signaling when it is expressed in the absence of IL-2Rβ or γ_c (8, 18, 19). However, we have recently shown that IL-15 can rescue murine fibroblasts from TNF-α-induced apoptosis by recruitment of TNFR-associated factor 2 to the IL-15Rα chain (27), and that IL-15Rα-mediated signaling in the human B cell line Raji occurs through association of IL-15Rα chain with Syk (28).

With respect to IL-15Rα-mediated signaling, mast cells have supposedly displayed unique features. Mast cells originate from bone marrow, maturate mostly in peripheral tissues, play a pivotal role in inflammatory and immediate allergic reactions, and have several additional functions, including roles in innate immunity, angiogenesis, and tissue remodeling (29, 30, 31). The development and proliferation of mast cells require proper signaling from multiple cytokines, among which the c-kit/stem factor system and IL-3 are best studied (29, 32). Reportedly, signaling pathways mediated by IL-15 in mast cells do not require IL-2β and γ_c chains and result in activation of Janus kinase 2 (JAK2)/STAT5 (14) or STAT6/Tyk2 (13, 15), whereas IL-15-mediated signaling in T lymphocytes recruits JAK1/JAK3 and STAT3/STAT5 (33). It is generally accepted that IL-15-mediated signaling in mast cells might make use of an alternative receptor provisionally designated as IL-15RX, while the role of the classical IL-15Rα receptor appears to be negligible (14). In this study, we show that murine mast cells express at least three novel isoforms of conventional IL-15Rα that result from an alternative splicing of IL-15Rα gene, and provide evidence for their functional importance.

Recombinant IL-2, IL-3, and IL-15 were purchased from PeproTech (London, U.K.). Anti-CD25/IL-2Rα PE (PC61), anti-CD122/IL-2Rβ PE (TMβ1), and anti-CD132/γ_c PE (4G2) mAbs were from BD PharMingen (San Diego, CA). Abs against IL-15Rα (N-19), Syk (N-19), STAT5 (L-20), phospho-STAT5 (pSTAT5, Tyr⁶⁹⁴), STAT6 (M-20), phospho-STAT6 (pSTAT6, Tyr⁶⁴¹), STAT3 (K-15), phospho-STAT3 (pSTAT3, Tyr⁷⁰⁵), JAK1 (H-106), JAK2 (C-20), Tyk2 (M-20), and JAK3 (C-21) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Abs that recognize JAK2 phosphorylated at Tyr^1007/1008 (pJAK2(pYpY^1007/1008)) were obtained from BioSource International (Solingen, Germany), and the mouse anti-phosphotyrosine (pTyr) Ab (RC20) was from BD Transduction Laboratories (Heidelberg, Germany). Goat anti-mouse, goat anti-rabbit, and rabbit anti-goat HRP conjugates (Amersham Life Science, Little Chalfont, U.K.) were used as secondary Abs. IL-15-IgG2b fusion proteins were produced, as previously described (10).

BMMCs were obtained from femoral bone marrow of 6-wk-old C57BL/6 mice, as previously described (34), and cultured in DMEM in the presence of 50 U/ml IL-3. Cells were collected after 4 wk of culture. The purity of mast cells was assessed by morphological (Giemsa stain) and by FACS analysis (98% of the cells were c-kit and FcεRI positive).

MC/9 (ATCC CRL 8306) is a murine mast cell line derived from fetal liver cells of a (B6 × A/J)F₁ mouse. MC/9 cells were cultured in DMEM supplemented with 10% FCS, 4.5 g/L glucose, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 1 mM sodium pyruvate, 1.5 g/L sodium carbonate (all components from Invitrogen, Groningen, The Netherlands), 10% WEHI-3-conditioned medium, and 40% conditioned medium from Con A-activated splenocytes obtained from BALB/c mice. The rat basophilic leukemia cell line RBL-2H3 (ATCC CRL-2256) was cultured in DMEM supplemented with 10% FCS, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. The mastocytoma cell line P815 (ATCC TIB-64) derived from DBA/2 mice, mouse C3H/An connective tissue cells L929 (ECACC 85011425), and COS-7 cells (ATCC CRL-1651) were cultured in RPMI 1640 supplemented with 10% FCS, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin (complete RPMI 1640). IL-3-dependent pro-B cell line BAF/3 was cultured in complete RPMI 1640 supplemented with 10% WEHI-3 cell-conditioned medium.

Before stimulation, cells were washed twice with Dulbecco’s PBS and incubated in medium without FCS and cytokines at 37°C for 3 h. For each assay, 5 × 10⁶ cells/ml were stimulated with IL-15 or IL-2 (final concentration 100 ng/ml) for 15 min at 37°C. Activation was interrupted by adding 8–10 vol of ice-cold PBS with 10 mM EDTA and 100 mM sodium vanadate. Cells were pelleted and frozen at −80°C before electrophoresis.

BAF/3 cells were transfected by electroporation (960 μF, 350 V) using a Gene Pulser (Bio-Rad, Munich, Germany). Stable transfectants were selected by limiting dilutions in the presence of G418 (1 mg/ml) (PAA Laboratories, Coelbe, Germany) for 4 wk. COS-7 and RBL-2H3 cells were transiently transfected using the GenePORTER 2 transfection kit (Gene Therapy Systems, San Diego, CA), harvested, and analyzed after 24 or 48 h of the transfection. Transfection efficiency in all experiments was ∼30–35%.

RNA was extracted from cells using TRIzol reagent (Invitrogen), according to the manufacturer’s instructions. cDNA was synthesized from 5 μg of total RNA using oligo(dT) as a template and SuperScriptII kit (Invitrogen). cDNA was amplified by PCR in a reaction mixture (20 μl) containing 2 μl of 10-fold PCR buffer with 1.5 mM MgCl₂, 250 μM of each dNTPs, 200 nM 5′ and 3′ primers, and 1 U Taq DNA polymerase (Amplitaq; Applied Biosciences, Warrington, U.K.).

The following primers were used: murine IL-2Rβ, sense 5′-GTCGACGCTCCTCTCAGCTGTAGTGGCTACCATA-3′, antisense 5′-GGATCCCAGAAGACGTCTACGGGCCTCAAATCCCAA-3′; murine IL-2Rγ, sense 5′-GTCGACAGAGCAAGCACCATGTTGAAACTA-3′, antisense 5′-GGATCCTGGGATCACAAGATTCTGTAGGTT-3′; murine IL-15Rα (E1), sense 5′-CCATGGCCTCGCCGCAGCTC-3′, anti-sense (E7) 5′-GTTTCCATGGTTTCCACCTCAA-3′; murine IL-15Rα (E2), sense 5′-AACATCCACCCTGATTGAGTGT-3′; and β-actin, sense 5′-GTGGGGCGCCCCAGGCACCA-3′, antisense 5′-CTCCTTAATGTCACGCACGATTTC-3′. All primers were purchased from TIB MolBiol (Berlin, Germany).

Samples were amplified in a DNA Thermocycler (Eppendorf, Hamburg, Germany) for 33 cycles. Each cycle consisted of denaturation at 94°C for 30 s, annealing at 60°C for 30 s, and elongation at 72°C for 1 min, preceded by initial denaturation at 94°C for 5 min, and followed by a final extension step at 72°C for 5 min. Aliquots of PCR products were electrophoresed on 1.5% agarose gel and visualized by ethidium bromide staining. β-Actin message was used to normalize the cDNA amount to be used. A mock PCR (without cDNA) was included to exclude contamination in all experiments.

To obtain full-length isoforms of IL-15Rα, a 3′ RACE kit (Invitrogen) was used. Briefly, mRNAs were converted into cDNA using reverse transcriptase and an oligo(dT) adapter primer provided with the kit. Then an anchored PCR was performed, using E1 primer as an upper gene-specific primer and abridged universal amplification primer provided with the kit as a lower primer (5′-GGCCACGCGTCGACTAGTAC-3′). The resulting PCR products were electrophoresed on 1.5% agarose gel, visualized by ethidium bromide staining, excised separately, and eluted. The products resulting from RT-PCR using E1/abridged universal amplification primers were ligated into pCRII-TOPO vector, transfected into Escherichia coli (TOPO-TA cloning kit; Invitrogen), purified, and sequenced, leading to pCRII.IL-15RαΔ4, pCRII.IL-15RαΔ3,4, and pCRII.IL-15RαΔ3,4,5. BamHI/XhoI fragments from each construct were then inserted into the pcDNA3.1 plasmid (Invitrogen), yielding pcDNA3.IL-15RαΔ4, pcDNA3.IL-15Rα Δ3,4, and pcDNA3.IL-15RαΔ3,4,5, respectively. Cloning of mouse IL-15Rα wild type (IL-15Rα WT) in pcDNA3.1 was performed, as described earlier (27).

Cell pellets were lysed for 15 min on ice in 1% Nonidet P-40 (Sigma-Aldrich, St. Louis, MO) cell extraction buffer: 20 mM Tris-HCl buffer, pH 8.0, 15 mM NaCl, 2 mM EDTA, 10 mM sodium fluoride, 1 μg/ml pepstatin A, 1 μg/ml leupeptin, 10 mM PMSF, and 100 μM sodium vanadate (all reagents from Sigma-Aldrich). The detergent-insoluble materials were removed by centrifugation for 15 min at 13,000 rpm at 4°C. Protein concentrations were determined using a BSA protein assay kit (Bio-Rad), and 50 μg of proteins was analyzed by electrophoresis in 10% SDS-PAGE. For analysis of glycosylation patterns, the samples (30 μg of protein) were treated with 250 μm of N-glycosidase F or with 200 μm of O-glycosidase (both enzymes are from Roche, Mannheim, Germany) for 3 h at 37°C, according to manufacturer’s instructions.

Membrane and cytoplasmic protein fractions were separated by repeated freezing/thawing cycles, followed by centrifugation at 13,000 rpm at 4°C for 15 min. Supernatant was collected as a fraction containing cytoplasmic proteins. Membrane proteins remaining in the pellet were solubilized using 1% Nonidet P-40 buffer. Nuclear proteins were isolated using the method described by Schreiber et al. (35). Three cellular fractions were analyzed in 10% SDS-PAGE.

For immunoprecipitation studies, lysates containing 500 μg of proteins were first precleared with the anti-mouse IgG bound to protein A/G-agarose and immunoprecipitated overnight at 4°C by incubation in 0.5% Nonidet P-40 buffer with 2 μg/ml of Abs. Immunocomplexes were captured on protein A/G-agarose. After washing, pellets were resuspended in SDS-PAGE sample buffer (62.5 mM Tris-HCl, pH 8.0, 1% glycerol, 2% SDS, 5% 2-ME, 0.01% bromphenol blue), boiled for 5 min, and analyzed in 10% SDS-PAGE. The resolved proteins were transferred onto nitrocellulose (Bio-Rad). Blots were blocked for 1 h in PBS with 0.05% Tween 20 (PBS-T) and 3% BSA (Sigma-Aldrich). After incubations with first and second Abs and washing with PBS-T, visualization of specific proteins was conducted by an ECL method using ECL Western blotting detection reagents (Amersham Life Science), according to the manufacturer’s instructions.

Human rIL-15 was iodinated, as described (36), with a sp. act. of ∼2000 cpm/fmol. Binding experiments were conducted for 4 h at 4°C using 2 × 10⁶ cells/sample and increasing concentrations of labeled IL-15 in a final volume of 100 μl. Nonspecific binding was determined in the presence of 200-fold excess of unlabeled cytokine. To maintain proper conditions of ligand excess cells were diluted in the incubation mixture with a carrier cell line, Daudi (ATCC CCL 213), which does not bind IL-15 (22). Data was plotted in the Scatchard coordinate system (37).

For cell survival assays, BA/F3 cells were washed three times with PBS to remove IL-3 and cultured at 1 × 10⁶ cells/ml in triplicate at 37°C for 1–3 days in RPMI medium in the presence of IL-15 or IL-3 (10 ng/ml). Cells were harvested and cell viability was determined by FACS analysis with 5 μg/ml of propidium iodide. The percentage of apoptotic cells was evaluated after 24 h of culture by ApoTarget annexin-V FITC apoptosis kit (BioSource International), according to the manufacturer’s protocol.

MC/9, P815, BA/F3, and BMMC were stained with mAbs, as previously described (4), and analyzed on a FACSCalibur (BD Biosciences, San Jose, CA) with CellQuest software. Negative controls consisted of isotype-matched, directly conjugated, nonspecific Abs (BD PharMingen).

COS-7 cells were seeded in concentration of 5 × 10⁴ cells/well in 12-well plates containing 18-mm glass coverslips the day before transfection and transfected the next day, using GenePORTER 2 reagent. Twenty-four hours after transfection, coverslips were fixed with 2% paraformaldehyde for 10 min at room temperature, permeabilized by 0.25% Triton X-100 for 10 min, and stained with 1/100 dilutions of Abs recognizing IL-15Rα in combination with GM130 (for Golgi complex; BD Biosciences) or Bip/GRR78 Abs (for endoplasmic reticulum (ER; BD Biosciences)) for 30 min at room temperature. Alexa Fluor-488 donkey anti-goat IgG (H + L) and Alexa Fluor-546 goat anti-rabbit IgG (H + L) (Molecular Probes, Leiden, The Netherlands) in dilution of 1/100 were used as second Abs. Nuclei were stained using TOTO-3 dye (Molecular Probes). To stain cell membranes, fixed cells were incubated with wheat germ agglutinin (Molecular Probes) for 15 min at room temperature, washed, and permeabilized with 0.25% Triton X-100, followed by staining with primary and secondary Abs. The specimens were mounted in 1, 4-diazabicyclo (2,2,2)octane antifading solution and analyzed by scanning confocal microscopy (Leica TCS SP, Bensheim, Germany).

All experiments were performed in at least three independent assays, which yielded highly comparable results. Data are summarized as mean + SD. Statistical analysis of the results was performed by Student’s t test for unpaired samples. A p value of <0.05 was considered statistically significant.

The IL-15R complex consists of three subunits, the high affinity IL-15Rα, IL-2Rβ, and IL-2Rγ or γ_c chain (20). The IL-2β subunit is shared between IL-2 and IL-15, while the γ_c chain serves as a common component of the receptors for IL-2, IL-4, IL-7, IL-9, and IL-15 (21, 38). However, it has been reported that in mast cells, IL-15 uses a different type of IL-15R, provisionally designated as IL-15RX (14). To study the nature of IL-15R complex on mast cells and to confirm the data that have been reported, we first analyzed the IL-15-binding ability of different murine mast cells, using a IL-15-IgG2b fusion protein (10). As shown in Fig. 1,A, BMMCs and MC/9 and P815 mast cells bind the IL-15-IgG2b fusion protein, although at a rather low level. Next, we examined the surface expression pattern of IL-2Rβ and IL-2Rγ (γ_c chain) on BMMCs and MC/9 and P815 cells using flow cytometry complemented by RT-PCR. Cytofluorometric analysis indicated the amount of γ_c protein on the cell surface of BMMCs to be lower in comparison with the mast cell lines (Fig. 1,A), even though RT-PCR confirmed that all cell types transcribe the IL-2Rγ chain (Fig. 1,B). In accordance with the earlier finding, IL-2Rβ chain was absent in BMMCs and MC/9 (14) and P815 cells. The murine T cell line CTLL-2, which expresses all components of the heterotrimeric IL-15R complex (14), was used as a positive control for RT-PCR (Fig. 1 B). Cytoskeletal β-actin products were amplified from each cDNA preparation to check the integrity of the RNA and to equalize the amount of the synthesized cDNA to be used. Thus, as previously described, BMMCs and MC/9 cells, and here extended to P815 mast cells, do not express the IL-2Rβ chain, carry the IL-2Rγ chain, and show IL-15-binding activity.

To examine the nature of the IL-15Rα chain on mast cells, two primer pairs, amplifying exons 1–7 (E1/E7) or exons 2–7 (E2/E7), were used, and RT-PCR was conducted on mRNAs extracted from primary BMMCs, or the mast cell lines MC/9 and P815. Both the E1/E7 primer pair (data not shown) and the E2/E7 primer pair amplify distinct IL-15Rα transcripts in these murine mast cells (Fig. 1 C), thus demonstrating that mast cells express a message for conventional IL-15Rα.

At the protein level, anti-IL-15Rα Abs do recognize several isoforms of the IL-15Rα chain. By Western blotting, these display a molecular mass of ∼20, 30, 42–45, and 60–65 kDa, suggesting the presence of several distinct isoforms (Fig. 1,D). However, by Western blotting, we were not able to detect a protein corresponding to the WT IL-15Rα (60–65 kDa) either in BMMCs or in the tested mast cell lines (Fig. 1 D), while the positive control (murine L929 fibroblasts) showed prominent expression of IL-15Rα WT protein. Thus, even if present at all, the amount of IL-15Rα WT appears to be low in all mast cell types tested.

To define the sequence and the genomic structure of the distinct IL-15Rα transcripts, the PCR products described above (Fig. 1,C) were cloned in the pCRII expression vector and sequenced. The sequence analysis revealed the presence of at least three novel transcripts corresponding to new mRNA species that lack exon 4; exons 3 and 4; or exons 3, 4, and 5 (designated IL-15RαΔ4, IL-15RαΔ3,4, and IL-15RαΔ3,4,5, respectively). In addtition, both MC/9 and P815, but not BMMCs, expressed the IL-15Rα WT transcript on the mRNA level. The deletion of the exon(s) did not shift the reading frame of the resulting proteins (data not shown). Because E2 primer is located in exon 2 (213–234 bp of murine IL-15Rα), and the E7 primer has the annealing site within the cytoplasmic domain of IL-15Rα in exon 7 (716–737 bp), the obtained isoforms were lacking full-length extracellular and cytoplasmic regions, and had sizes that were ∼260 bp (for E2/E7 primer pair) or ∼50 bp (for E1/E7 primer pair) smaller than those corresponding to the full-length sequence. The upper and middle bands corresponded to IL-15RαΔ4 (592 bp) and IL-15RαΔ3,4 (501 bp), respectively, while the lower band represented IL-15RαΔ3,4,5 (468 bp). All of these have lower molecular mass after amplification with the E2/E7 primer pair (Fig. 1 C).

To obtain full-length isoforms, 3′ RACE kit was used. By this method, we were able to obtain the full-length IL-15Rα isoforms. The schematic overview of the murine and human IL-15Rα genes and the transcription products of already known and novel IL-15Rα isoforms both in human and mouse systems are summarized in Fig. 2. Taken together, these results demonstrate the presence of at least three novel isoforms of conventional IL-15Rα in BMMCs and MC/9 and P815 mast cells on mRNA and protein level.

To understand the pattern of subcellular distribution of the IL-15Rα isoforms and to explore their ability to be translocated to the cellular membrane, we first analyzed their localization by confocal microscopy. Using the obtained cDNA sequences, the IL-15Rα isoforms were cloned in pcDNA3.1 expression vector. Then COS-7 cells were transiently transfected with the corresponding constructs, and the localization of the receptor isoforms was analyzed. IL-15Rα WT was used for comparison. All three novel isoforms were predominantly associated with the Golgi apparatus (Fig. 3,A) and the ER (data not shown). Furthermore, in addition to an association of IL-15Rα isoforms with the plasma membrane (Fig. 3,B), characteristic ring-like structures, as previously reported by Dubois et al. (39), could be detected in the perinuclear area (Fig. 3 A). A similar pattern was observed when endogenous IL-15Rα was stained in L929 cells, ruling out the possibility of a transfection-associated artifact (data not shown). Notably, transfection of the cells with IL-15Rα isoforms in a different expression vector (pEGFP.N1) gave similar results (data not shown). Because all three IL-15Rα isoforms displayed a pattern of subcellular distribution similar to that of IL-15Rα WT, these experiments indicate that the alternative splicing of the IL-15Rα gene had no influence on the localization of the resulting proteins.

To complement the confocal data and to confirm the subcellular distribution of IL-15Rα isoforms by Western blotting, subcellular fractions were prepared by biochemical means. As shown in Fig. 4,A, the nuclear fraction of transfected COS-7 cells contains only a very small amount of IL-15Rα WT, whereas all isoforms as well as WT receptor were mostly detected in the cellular membrane fraction, indicating the predominant association of the IL-15Rα isoforms with the plasma membrane. However, the IL-15RαΔ3,4 isoform and, to a lesser extent, the IL-15RαΔ4 and IL-15RαΔ3,4,5 isoforms were also detected in the cytosolic fraction, while the amount of WT receptor detected was very low (Fig. 4 A).

Taken together, these experiments show the association of the IL-15Rα isoforms with the plasma membrane, the Golgi complex, the ER, and the perinuclear area, closely resembling that of WT IL-15Rα.

Because the detection of multiple bands with variable intensities and slightly diverse molecular mass, as shown in Fig. 1 D, could be due to a different glycosylation state (22) of the novel IL-15Rα isoforms, we investigated their glycosylation pattern. Protein glycosylation assures protein stabilization and correct folding, serves as a lipid anchor for attaching of the protein to a membrane, and provides the protein with certain recognition characteristics (40). Two types of carbohydrate addition are commonly found: the O-glycosidic linkage involves attachment of the carbohydrate to the hydroxyl group of serine, threonine, and hydroxylysine, while the N-glycosidic linkage uses the amide group of asparagines (40). IL-15Rα has single N-glycosylation and multiple O-glycosylation sites in the extracellular domain of the receptor (22, 41). To study the role of glycosylation in posttranslational processing of the IL-15Rα isoforms, the constructs were transiently transfected into COS-7 cells. The whole cell extracts were treated with N- or O-glycosidases, and the expression products were analyzed by Western blotting.

As shown in Fig. 4,B, the transfection of IL-15RαΔ4 led to the expression of a product with molecular mass of ∼38–42 kDa. Treatment with N-glycosidase induced a shift from ∼38- to 42-kDa to ∼33- to 35-kDa band corresponding to nonglycosylated IL-15RαΔ4, whereas treatment with O-glycosidase did not affect the mobility of the protein in the gel. Transfection of IL-15RαΔ3,4 resulted in the expression of a band with molecular mass of ∼25–30 kDa. Treatment with N-glycosidase induced a shift to ∼18- to 20-kDa band corresponding to nonglycosylated IL-15RαΔ3,4 isoform, whereas treatment with O-glycosidase again had no effect. The same results were obtained with IL-15RαΔ3,4,5, reducing the size of the protein from ∼20–22 to ∼18–20 kDa only after treatment with N-glycosidase (Fig. 4 B). Because most (up to 74%) of the serine and threonine residues that are potential targets for O-linked sugar addition in the extracellular part of the human and mouse receptors are located in the exon 3/4-encoded domains (23), the absence of the effect of O-glycosidase in our experiments appears to result directly from the deletion of these exons. Taken together, our data show the usage of N-glycosylation site for all three novel isoforms of murine IL-15Rα in COS-7 cells, while no O-glycosylation was observed.

A number of JAK/STAT factors have been implicated in IL-15 signaling in BMMCs and PT-18 and MC/9 cells (13, 14, 15). To confirm the available data and to investigate the mast cells activation in more detail, we studied the ability of IL-15 to induce activation of signaling molecules first in BMMCs, and then in MC/9 and P815 cells. BMMCs were stimulated with 100 ng/ml of each IL-15, IL-2, or IL-3 for comparison. The cells were lysed, and tyrosine phosphorylation status of STAT5 and STAT6 was examined by immunoblotting using corresponding anti-phospho-specific STAT5 and STAT6 Abs. As shown in Fig. 5 A, both IL-15 and IL-3 were able to stimulate phosphorylation of STAT5 in BMMCs, while IL-2 had no effect. We also found that, in accordance with the previous finding (15), IL-15 was able to trigger phosphorylation of STAT6. Notably, STAT6 protein was 65-kDa size. This is in agreement with the recent study of Sherman et al. (42), who reported that murine mast cells preferentially express 65-kDa STAT6 isoform with truncated C-terminal region, despite the presence of conventional 94-kDa molecule.

For immunoprecipitation, BMMCs treated as described above were lysed, and JAK2 was immunoprecipitated by the specific Abs and analyzed by SDS-PAGE. Then the proteins were transferred to nitrocellulose membrane, and the tyrosine phosphorylation was evaluated by the use of anti-pTyr mAb RC20. In agreement with the earlier study of Tagaya et al. (14), IL-15 was able to induce the phosphorylation of JAK2 (Fig. 5 B), sharing this ability with IL-3, while IL-2 was without effect. The similar pattern of phosphorylation was detected in MC/9 and P815 cells (data not shown). We were not able, however, to detect phosphorylation of Tyk2 (data not shown). Thus, IL-15 can trigger phosphorylation of STAT5, STAT6, and JAK2 in BMMCs as well as in MC/9 and P815 cells.

We have recently shown that Syk kinase associates with the intracellular part of IL-15Rα and becomes phosphorylated and activated upon IL-15 treatment in lymphoid cells (28). To prove the involvement of Syk in IL-15-mediated signaling in mast cells, BMMCs were stimulated for 15 min with IL-15. Then the cells were lysed and Syk kinase was precipitated from the lysates using specific Abs. After SDS-PAGE, the blots were probed by anti-pTyr Abs. As shown in Fig. 5,C, IL-15, but not IL-2, induced phosphorylation of Syk kinase. To detect the association of Syk with IL-15Rα, cell extracts were immunoprecipitated with anti-IL-15Rα Abs and analyzed by Western blotting using Abs against Syk. Fig. 5 C illustrates that stimulation of BMMCs with IL-15, but not IL-2, for 15 min induced coprecipitation of Syk with the IL-15Rα. Noteworthily, IL-15 was able to induce phosphorylation and precipitation of Syk kinase with IL-15Rα also in MC/9 and P815 cells (data not shown). Taken together, these results demonstrate that IL-15 is able to induce phosphorylation and precipitation of Syk with IL-15Rα in BMMCs and MC/9 and P815 mast cell lines.

BA/F3 is an IL-3-dependent pro-B cell line that reportedly expresses IL-2Rα and IL-2Rγ subunits, but does not express IL-2Rβ chain (43). The stable transfection of functional receptors for several cytokines, such as IL-2Rβ, IL-4R, or IL-21R, into BA/F3 cells results in the establishment of transfectants that are able to respond to IL-2, IL-4, or IL-21, respectively, because these cytokines acquired the ability to substitute the growth-promoting effect of IL-3 (20, 43, 44, 45). Therefore, this cell line was chosen as a suitable model for the functional characterization of the IL-15Rα isoforms derived from murine mast cells.

First, the expression of IL-2R/IL-15R complex subunits on these cells was evaluated. As shown in Fig. 6, parental BA/F3 cells express the transcripts coding for IL-2Rγ, but not IL-2Rβ and IL-15Rα, when tested by RT-PCR (Fig. 6,A). Western blotting with anti-IL-15Rα Abs (Fig. 6,B) and FACS analysis (Fig. 6,C) confirmed the absence of IL-15Rα (60–65 kDa) protein expression. CTLL-2 or L929 cells were used as positive controls for RT-PCR and Western blotting, respectively. To investigate the functional properties of each of the IL-15Rα isoforms, BA/F3 cell lines stably transfected with the corresponding constructs were established. RT-PCR analysis confirmed the presence of messages for the IL-15Rα isoforms in such genetically modified BA/F3 cells (Fig. 7 A).

Next, we evaluated the IL-15-binding properties of the transfected BAF/3 cells by using radioiodinated IL-15. Scatchard analysis of ¹²⁵I-labeled IL-15-binding data demonstrated that each of the transfected BA/F3 clones that expressed the respective IL-15Rα isoform at the cell surface was able to bind labeled IL-15 (Fig. 7 B). The cells expressed ∼1500 high affinity binding sites for IL-15 upon the cell surface (K_d ∼ 10 pM). Parental cells did not bind IL-15 (data not shown). Formation of the IL-15/IL-15Rα complex was inhibited by competing with unlabeled IL-15 (data not shown).

It has been reported that IL-15 has little, if any, growth-promoting effects on MC/9 mast cells at low concentration (10 ng/ml), but can prolong survival of the cells after growth factor withdrawal (13). To test whether IL-15 is able to prevent apoptosis of transfected BA/F3 cells stably expressing the IL-15Rα isoforms, IL-3-deprived transfectants as well as parental (nontransfected) BA/F3 cells were maintained in the presence of IL-15 (10 ng/ml) for 1–3 days. Cell viability was assessed by exclusion of propidium iodide. The number of apoptotic cells was evaluated by flow cytometry analysis using FITC-conjugated annexin V staining. Cell culture in the presence of IL-3 was used for comparison. As shown in Fig. 8, IL-15 significantly enhanced the survival rate of BA/F3 cells expressing IL-15Rα WT, IL-15RαΔ4, and IL-15RαΔ3,4 isoforms, keeping 80–90% of the cells alive (Fig. 8,A) and significantly diminishing the percentage of apoptotic cells (Fig. 8 B). Such property of IL-15 was reduced in the cells transfected by the IL-15RαΔ3,4,5 isoform, resulting in survival of ∼50% of the cells. The effect of IL-15 on cell viability was comparable with that of IL-3, but IL-15 failed to support survival of the parental cells.

Taken together, these experiments show the ability of IL-15 to promote survival of BA/F3 cells expressing murine mast cell-derived IL-15Rα isoforms at the cell surface, attesting to the functional activity of these newly discovered receptor isoforms.

To further characterize the downstream signaling events after IL-15RαΔ4, IL-15RαΔ3,4, and IL-15RαΔ3,4,5 stimulation, these receptor isoforms were transiently transfected into the RBL-2H3 rat mast cell line using pcDNA3.1 vector (46). The choice of these rat mast cells was validated by the fact that, in our preliminary experiments, murine IL-15 was not able to induce any considerable changes either in the phosphorylation pattern of STATs and JAK2, or of Syk in these cells. Therefore, parental RBL-2H3 cells might serve as an internal control for their transfected counterparts, thereby allowing us to evaluate the contribution of each murine IL-15Rα isoform to the signaling mediated by murine IL-15. IL-15Rα WT was used for comparison. To confirm the efficacy of transfection, the cells carrying the respective constructs were analyzed by Western blotting using anti-IL-15Rα Abs (Fig. 9 A). After 48 h, the cells were serum starved for 3 h, stimulated with 100 ng/ml of IL-15 for 15 min, and subjected to SDS-PAGE, and the membranes were probed with Abs against phosphorylated STAT3, STAT5, STAT6, and JAK2.

As shown in Fig. 9,B, the treatment of all transfectants with IL-15 induced the phosphorylation of STAT5 and JAK2. The phoshorylation of STAT5 and JAK2 was more prominent in the cells expressing IL-15RαΔ4 construct. In addition, we were able to detect the phosphorylation of STAT6 in cells bearing IL-15RαΔ4 and IL-15RαΔ3,4,5 isoforms, but not in cells transfected with IL-15Rα WT or IL-15RαΔ3,4. Importantly, IL-15 also triggered the phosphorylation of STAT3 in the cells transfected with IL-15Rα WT, IL-15RαΔ3,4, or IL-15RαΔ3,4,5 isoforms. Furthermore, Syk kinase was phosphorylated in response to IL-15 treatment and associated with all IL-15Rα isoforms, including IL-15Rα WT (Fig. 9 C). Taken together, this suggests that all isoforms and IL-15Rα WT can trigger the phosphorylation of STAT5, JAK2, and Syk kinase, as well as the precipitation of Syk with IL-15Rα in response to IL-15 treatment. However, only the cells transfected with IL-15RαΔ4 and IL-15RαΔ3,4,5 show the phosphorylation of STAT6. Given the ability of IL-15 to stimulate STAT6-dependent expression of IL-4 and Bcl-x_L both in BMMCs and MC/9 cells (13, 15), we also tested whether the activation of these signaling molecules may take place in response to IL-15 in the RBL-2H3 cells expressing the IL-15RαΔ4 and IL-15RαΔ3,4,5 isoforms. However, Western blotting experiments revealed that IL-15 had no capacity to stimulate the expression of IL-4 and Bcl-x_L in these cells (data not shown).

Taken together, these results show that, even after transfection into mast cells from a different species, the IL-15Rα isoforms still preserve the ability to trigger activation of distinct intracellular signaling molecules.

In this study, we have shown the existence of at least three novel isoforms of IL-15Rα in BMMCs and MC/9 and P815 mast cell lines, which may explain the selective effects exerted by IL-15 on mast cells. These isoforms result from an alternative splicing of murine IL-15Rα mRNA and correspond to the deletion of exon 4; deletion of exons 3 and 4; and deletion of exons 3, 4, and 5, showing the same pattern of subcellular distribution as IL-15Rα WT. Thus, contrary to the previously held concept (14), murine mast cells do indeed express conventional IL-15Rα chains, yet are novel isoforms of the receptor. That these isoforms are fully functional is suggested by our finding that IL-15 was able to mediate the survival of the genetically modified BA/F3 cells stably expressing the IL-15Rα isoforms upon the cell surface. Finally, we show that IL-15 is able to activate a number of signaling molecules in BMMCs and MC/9 and P815 cells, such as STAT5, STAT6, JAK2, and Syk, and preserves the capacity to mediate the signaling events in RBL-2H3 rat mast cells expressing the novel IL-15Rα isoforms.

Mast cells play a central role in the initiation of acute allergic and pseudoallergic reactions due to the rapid secretion and secondary generation of proinflammatory mediators (29, 30). A number of studies have focused on the consequences of IL-15 action on BMMCs and mast cell lines and have yielded somewhat contradictory results (13, 14, 15). Tagaya et al. (14) reported that IL-15 treatment of BMMCs and PT-18 and MC/9 cells resulted in the phosphorylation of JAK2, and of STAT5 (in parental PT-18 cells), but not in the phosphorylation of JAK1, JAK3, STAT3, and Tyk2. Conversely, Masuda et al. (15) have shown that IL-15 was able to induce tyrosine phosphorylation of Tyk2 and a STAT6-dependent expression of IL-4 in MC/9 cells and BMMCs. In addition, IL-15 can promote survival of BMMCs and MC/9 cells through STAT6-mediated Bcl-x_L expression in the absence of IL-3, or upon treatment with anti-Fas Abs (13). It has been assumed that IL-15 in mast cells uses a novel receptor, provisionally designated as IL-15RX, as well as distinct signaling pathways, despite the presence of mRNA for a conventional IL-15Rα in PT-18 cells (14). The fact that BMMCs as well as MC/9 and P815 cell lines are able to express novel, alternatively spliced IL-15Rα isoforms raises the possibility that these isoforms, rather than a unique IL-15RX, may mediate IL-15 signaling in mast cells. Therefore, studies aimed at understanding the functional significance of conventional IL-15Rα isoforms in mast cells are likely to provide particularly helpful insights into the role of IL-15 in mast cell biology, and may explain the specific responses of selected mast cell subtypes to IL-15 stimulation. Furthermore, additional studies will focus on investigating the feasibility to use IL-15 antagonists, such as a soluble IL-15Rα, for blocking mast cell activities.

Three differentially spliced human IL-15Rα variants capable of a high affinity binding of IL-15 were already reported (6, 41). Recently, the existence of a novel type of IL-15Rα mRNA lacking exon 2 and a putative nuclear localization signal has been shown in human tissues (39). The new alternatively spliced products of IL-15Rα gene from mouse mast cells can be added to the ones already described in human system (39, 41). However, we cannot exclude that murine mast cells may also express, although presumably at rather low levels, messages for other IL-15Rα isoforms than the ones identified in this work.

The presence of such differentially spliced isoforms in murine mast cells of IL-15Rα poses the intriguing question as to which biological reasons may underlie such receptor diversity. Interestingly, the deletions of the exon(s) described in the current study affected linker and/or Ser/Thr-rich regions, thus leaving extracellular and cytosolic parts intact, thereby allowing the resulting proteins to preserve cytokine-binding and signaling properties. The expression of the alternatively spliced IL-15Rα proteins in COS-7 cells showed that they were predominantly associated with the Golgi, the ER, the cellular membrane, and the perinuclear space. The localization of IL-15Rα isoforms in the Golgi and the ER might likely reflect the necessity for the posttranslational modifications of the expressed proteins in COS-7 cells. Confocal microscopy showed that all isoforms have the same localization pattern as IL-15Rα WT, indicating that such deletions do not affect the subcellular routing of the proteins. Both nonglycosylated and glycosylated forms of the proteins appeared to be present. The murine IL-15Rα has a potential N-glycosylation site in the sushi domain (Asn⁵¹) within the exon 2-encoded domain (22). Because the new IL-15Rα isoforms all lack exon 4 or exons 3 and 4, containing most of the serine and threonine residues (22, 41), which are the potential targets for O-linked sugar addition, the size increase appears to be due to N-glycosylation.

The ability of IL-15 to promote the survival of BA/F3 cells transfected with the IL-15RαΔ3,4,5, IL-15RαΔ3,4, IL-15RαΔ4 isoforms, and IL-15Rα WT indicates that IL-15Rα is capable of triggering intracellular signaling upon ligand binding, which delivers an effective survival message. In fact, the property of IL-15 to substitute the growth-promoting effect of IL-3 of the genetically engineered BA/F3 cells, diminishing the number of apoptotic cells and resulting in survival of 80–90% of the cells, appears to be in accordance with its well-documented antiapoptotic function (10, 27, 28).

Nevertheless, in contrast to IL-3, IL-15 was not able to induce proliferation of the BA/F3 transfectants. Here, the limited effect of IL-15 might reflect the inability of IL-15Rα isoforms to provide a mitogenic stimulus for BA/F3 cells or, alternatively, indicates that IL-15 has no capacity to trigger the proliferation of these cells. The transfected BA/F3 cells were capable of binding IL-15 with a high affinity (K_d ∼ 10 pM). All isoforms were able to bind labeled IL-15 almost equally, clearly demonstrating that deletion of exon 4; exons 3 and 4; and exons 3, 4, and 5 does not affect the essential elements involved in IL-15 binding, such as the sushi domain of IL-15Rα (47). These data are supported by the work of Anderson et al. (41), showing that the exon 3-encoded linker sequence is dispensable for IL-15 binding and signaling.

In accordance with previous reports (14, 15), we found that both STAT5 and STAT6 are phosphorylated in response to IL-15 treatment in BMMCs. Notably, the detected STAT6 protein was 65 kDa in size, corresponding to the earlier reported C-terminal truncated isoform predominantly found in mast cells (42). IL-15 also induced the phosphorylation of JAK2 and Syk, and association of Syk with IL-15Rα, which is in agreement with the previous findings (14, 28). Recent studies have also underlined the involvement of the γ_c chain and JAK3 in activation, proliferation, and survival of mast cells (48, 49, 50). Importantly, treatment of cells with specific inhibitory Abs to γ_c chain (48) or an inhibitor of JAK3, WHI-P131 (49, 50), had no effect on the ability of IL-15 to trigger the phosphorylation of STAT5, STAT6, JAK2, or Syk in BMMCs as well as in MC/9 and P815 cells (data not shown). However, despite the observed inability of the inhibitory Abs to γ_c chain to affect signaling events in response to IL-15, further studies are required to address the possible participation of γ_c chain to the IL-15 signaling in mast cells, and its potential interactions with the novel IL-15Rα isoforms.

Furthermore, IL-15 was able to trigger activation of JAK2, STAT5, STAT6, and, interestingly, of STAT3 as well as Syk kinase, in the genetically modified rat RBL-2H3 mast cell line, thus providing an unambiguous evidence for the ability of the isoforms to mediate downstream signaling. It should be noted that activation of STAT6 was detectable in the cells transfected with the IL-15RαΔ4 and IL-15RαΔ3,4,5 isoforms, but not in IL-15RαΔ3,4- and IL-15Rα WT-expressing cells. Notwithstanding, Western blotting experiments showed that the activation of STAT6 in the RBL-2H3 cells after IL-15 treatment had no effect upon the level of expression of IL-4 and Bcl-x_L, despite the well-documented ability of IL-15 to stimulate STAT6-dependent expression of IL-4 and Bcl-x_L both in BMMCs and MC/9 cells (13, 15) These results, however, require cautious interpretation, and future experiments are necessary to understand in more detail the molecular mechanisms underlying the capacity of different IL-15Rα isoforms to induce activation of distinct signaling molecules shown here in the rat mast cells used.

Taken together, our results provide the evidence that murine mast cells express novel, alternatively spliced isoforms of IL-15Rα, whose presence proves the functional importance of the conventional IL-15Rα. Further studies, which investigate IL-15-mediated mast cell activities in IL-15Rα- and IL-2Rγ-deficient mice, will surely provide helpful insights into the role of the distinct IL-15R complex components in this particular cell system.

We are grateful to Martina Hein and Katrin Streeck for excellent technical assistance.