Mitogen-Activated Protein Kinase Activation Through Fcε Receptor I and Stem Cell Factor Receptor Is Differentially Regulated by Phosphatidylinositol 3-Kinase and Calcineurin in Mouse Bone Marrow-Derived Mast Cells¹

Mast cells play a central role in inflammatory and immediate allergic responses. Aggregation of the high affinity FcR for IgE (FcεRI) triggers the activation of different signal transduction pathways. Activation of protein tyrosine kinases (PTKs),³ including Syk, Lyn, Btk, and Itk, is one of the earliest signaling events induced by aggregation of the FcεRI on mast cells 1, 2, 3, 4, 5 . PTK activation is thought to be proximal to the activation of phospholipase Cγ and protein kinase C, and appears to be essential for mast cell degranulation 6 since PTK inhibitors prevent the liberation of inositol trisphosphate and histamine release 7, 8 . In addition to the release of mast cell granule contents, these pathways lead to later responses, such as the modulation of cytokine gene expression. However, the downstream consequences of these early activation events are not well defined. We have shown recently that three members of the mitogen-activated protein (MAP) kinase family, designated c-Jun amino-terminal kinase (JNK), p38 MAP kinase (p38), and extracellular signal-regulated kinase (ERK), are activated following FcεRI aggregation in a mouse mast cell line, MC/9 9, 10 . Stem cell factor (SCF), the ligand for the Kit tyrosine kinase receptor (SCFR) encoded by the c-Kit proto-oncogene, also plays an important role in the development of mast cells and hemopoiesis 11, 12, 13 . In this study, we show that the same three MAP kinase family members are activated through both FcεRI and SCFR in mouse bone marrow-derived mast cells (BMMC), that the activation of these kinases is differentially regulated by upstream proteins such as PI3-kinase and calcineurin, and both PI3-kinase- and calcineurin-dependent pathways play an important role in the regulation of cytokine production and the degranulation response.

Bone marrow was obtained from the femurs of female BALB/c mice (The Jackson Laboratory, Bar Harbor, ME) and cultured in Iscove’s modified Dulbecco’s medium (IMDM; Life Technologies, Grand Island, NY) supplemented with 5% FBS (Summit Biotechnology, Ft. Collins, CO), 50 μM 2-ME (Life Technologies), 2 mM glutamine, 100 μg/ml streptomycin, 100 U/ml penicillin, 0.5 μg/ml amphotericin B, and IL-3 obtained from medium conditioned by X63 AG8-653 myeloma cells transfected with a vector expressing IL-3 14 . After 4 wk of culture, more than 95% of nonadherent cells contained granules that stained positively with toluidine blue. Wortmannin and rapamycin were purchased from Calbiochem (San Diego, CA). Cyclosporin A (CsA) and cyclosporin H (CsH) were provided by Sandoz Pharma (Basel, Switzerland), and FK506 was provided by Fujisawa Pharmaceutical (Osaka, Japan). The MEK1 inhibitor, PD98059, was purchased from New England Biolabs (Beverly, MA). Bovine myelin basic protein was obtained from Upstate Biotechnology (Lake Placid, NY). Goat polyclonal anti-ERK2 (C-14) Ab and anti-Akt1 (C-20) Ab were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Recombinant protein G agarose was purchased from Zymed Laboratories (San Francisco, CA). Kit ligand (SCF) was obtained from medium conditioned by CHO cells transfected with a kit ligand expression vector (kindly provided by Genetics Institute, Cambridge, MA). Medium conditioned by nontransfected CHO cells was used as a control and showed no activity in inducing kinase activation. Mouse rTNF-α, mouse rIL-4, purified rat anti-mouse TNF-α mAb, purified rat anti-mouse IL-4 mAb (ELISA capture), biotinylated rabbit anti-mouse TNF-α polyclonal Ab, and biotinylated rat anti-mouse IL-4 mAb (ELISA detection) were purchased from PharMingen (San Diego, CA). The PKA inhibitor (TTYADFIASGRTGRRNAIHD) and Crosstide (GRPRTSSFAEG) were made in the Molecular Resource Center, National Jewish Medical and Research Center (Denver, CO).

BMMC (5 × 10⁶/ml) were cultured with 500 ng/ml anti-OVA IgE 15 for 2 h. This concentration of IgE was shown in initial experiments to be optimal in stimulation of JNK in BMMC. The cells were washed with medium three times and cultured with fresh medium for an additional 2 h. OVA dissolved in PBS was added to the passively sensitized cells or PBS was used as a control. In some experiments, BMMC (3 × 10⁶/ml) were incubated with fresh medium for 2 h and SCF was added to the medium in a final volume of 1%. This concentration of conditioned medium was shown in initial studies to lead to optimal activation of JNK and was comparable with 100 ng/ml rSCF 9 .

Glutathione S-transferase-c-Jun 1–79(1–79) fusion protein was prepared as described previously 16 , and kinase activity was measured as described 9, 10 .

p38 kinase activity was assayed as described using ATF-2 as substrate and quantified with a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) 10 .

In vitro kinase assay of ERK2 was conducted as described previously 17 using myelin basic protein as substrate.

Cells (3 × 10⁶) were lysed in a buffer (10 mM KPO₄ (pH 7.4), 0.1% Nonidet P-40, 1 mM EDTA, 5 mM EGTA, 10 mM MgCl₂, 20 mM β-glycerophosphate, 0.5 mM Na₃VO₄, 2 mM DTT, 1 mM PMSF, 10 μg/ml aprotinin, and 5 mg/ml leupeptin). The lysates were incubated with 0.8 μg goat anti-Akt1 Ab for 2 h at 4°C. Recombinant protein G agarose was added to the lysates and incubated for an additional 1 h at 4°C. The immunoprecipitates were washed twice with lysis buffer, and once with kinase buffer (20 mM Tris-HCl (pH 7.5), 10 mM MgCl₂, 0.1 mg/ml BSA, and 1 mM DTT). After the final wash, 50 ml of a kinase assay buffer containing 5 μCi of [γ-³²P]ATP (DuPont, Wilmington, DE), 100 μM cold ATP, 1 μg/ml PKA inhibitor, and 5 μg Crosstide 18 was added per sample. The samples were incubated for 15 min at 30°C and the reaction was stopped by adding 10 μl of 0.5 M EDTA. A total of 25 μl of each sample was loaded on phosphocellulose paper (Whatman, Clifton, NJ), and phosphorylation of Crosstide was determined by liquid scintillation counting. The phosphocellulose paper was washed five times with 0.425% (v/v) phosphoric acid.

BMMC (2 × 10⁶/ml) were cultured in medium containing 1 μCi/ml 5-[1,2-³H](N)-hydroxytryptamine creatinine sulfate (DuPont) and cultured for 15 h. After washing, cells (5 × 10⁶/ml) were cultured with 500 ng/ml anti-OVA IgE for 2 h and washed three times with culture medium. BMMC (1 × 10⁶/ml) were incubated in medium for 2 h and stimulated by addition of OVA. The reaction was stopped by adding ice-cold medium and centrifugation. After centrifugation, supernatants and cell pellets were loaded in the scintillation counter. The percentage of release was calculated by dividing the net supernatant counts by the total counts.

Purified rat anti-mouse TNF-α mAb or purified rat anti-mouse IL-4 mAb was diluted to 2 μg/ml or 1 μg/ml in coating solution (0.1 M NaHCO3, pH 8.2), and 50 μl was added to wells of an ELISA plate (Dynatech Laboratories, Chantilly, VA). After overnight incubation at 4°C, wells were washed twice with washing solution (0.05% Tween-20/PBS) and blocked with PBS containing 10% FCS at room temperature for 2 h. After washing two times, standards (30 pg/ml-2 ng/ml mouse rTNF-α or 40 pg/ml-2.5 ng/ml mouse rIL-4) and samples were added at 100 μl/well and incubated overnight at 4°C. After washing four times, biotinylated rabbit anti-mouse TNF-α polyclonal Ab (1 μg/ml) or biotinylated rat anti-mouse IL-4 mAb (0.5 μg/ml) was added to the wells and incubated at room temperature for 45 min, and the wells were washed six times. Avidin-peroxidase (2 μg/ml) was added to the wells and incubated at room temperature for 30 min, and the wells were washed eight times. 2,2′-Azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS, 30 mg/ml 0.1 M citric acid, pH 4.35) containing 0.03% H₂O₂ was added at 100 μl/well, and the color reaction was allowed to develop at room temperature for 30 min. The plates were read at OD 410 nm and analyzed by Microplate Manager (Bio-Rad, Hercules, CA).

Values were compared by Student’s t test or Welch’s t test.

BMMC previously sensitized with anti-OVA IgE were challenged in the presence or absence of OVA. Fig. 1, A–C, shows that addition of 10 μg/ml OVA induced the activation of three members of the MAP kinase family, JNK, p38, and ERK2. Similarly, addition of SCF (1%) to BMMC induced the activation of JNK, p38, and ERK2 (Fig. 1, D–F). JNK activity reached maximum levels at 15 min, p38 activity at 5 min, and ERK2 activity at 5–15 min after the addition of either OVA or SCF. These data indicate that aggregation of FcεRI receptors or signaling through the SCFR rapidly activates all three MAP kinases in BMMC.

The Akt proto-oncogene encodes a serine/threonine kinase, Akt1 19 , which is rapidly and specifically activated by growth factors such as platelet-derived growth factor (PDGF) 20 . It is well known that Akt1 activation is mediated through PI3-kinase signaling 21 and Akt1 activation is predicted to parallel PI3-kinase activity. Akt1 was activated both by aggregation of FcεRI and ligation of the SCFR (Fig. 2). SCFR-induced Akt1 activation was stronger than FcεRI-mediated activation in BMMC. Activation was maximal 1–5 min after addition of the ligand (Fig. 2,A). Under both conditions, Akt1 activation was inhibited by wortmannin, a PI3-kinase inhibitor 22 , in a dose-dependent manner (Fig. 2 B). These results confirm that PI3-kinase is activated following aggregation of FcεRI and ligation of the SCFR in mast cells 23 .

Addition of wortmannin demonstrated the sensitivity of JNK activation via FcεRI in BMMC; inhibition of JNK was dose dependent, and in the presence of 100 nM wortmannin, JNK activation was inhibited by 90% (Fig. 3,A). Wortmannin also inhibited p38 activation via FcεRI in BMMC in a dose-dependent manner, although the degree of inhibition in p38 activity was less (50%) than observed for JNK activation (Fig. 3,B). Wortmannin failed to inhibit ERK2 activation via FcεRI in BMMC (Fig. 3,C). These results indicate that FcεRI-mediated PI3-kinase activation is involved in JNK activation, and to some extent, p38 activation in BMMC, whereas ERK activation is independent. Virtually identical results were obtained previously in MC/9 cells 9, 10 . Preincubation of BMMC with wortmannin before the addition of SCF revealed that JNK activation was inhibited in a dose-dependent manner, similar to activation of JNK through FcεRI, but SCF-induced activation of p38 and ERK2 was insensitive in these mast cells (Fig. 3, D–F).

CsA is known to affect mast cell function, particularly the degranulation response and cytokine production 24, 25, 26, 27 . Addition of CsA (1 μg/ml) strongly inhibited JNK activation (80% inhibition) and partially inhibited p38 activation (40% inhibition) through FcεRI; ERK2 activation via FcεRI was not affected. CsH, a derivative of CsA that does not bind cyclophilin A or B and is not immunosuppressive 28, 29, 30, 31 , did not affect JNK activation (Fig. 4,A). FK506 and RAP similarly did not affect JNK activation. It is known that BMMC lack FKBP12 32 , and since binding to FKBP12 is required for FK506 to inhibit calcineurin 33 , the absence of the protein accounts for the lack of FK506-inhibitory activity. By contrast, CsA did not affect SCF-induced JNK, p38, or ERK2 activation (Fig. 4, D–F).

It has been shown that wortmannin inhibits mast cell degranulation at the same concentration in which inhibition of PI3-kinase is observed 34, 35 . Wortmannin inhibited FcεRI-mediated serotonin release from BMMC in a dose-dependent manner; at a concentration of 100 nM wortmannin, serotonin release was inhibited by more than 75% (Fig. 5,A). CsA also inhibits FcεRI-mediated degranulation in rat basophilic leukemia cells and basophils 24, 25, 26 , and a role for calcineurin in the degranulation process has been proposed 36 . CsA (1 μg/ml) partially (56%) but significantly inhibited FcεRI-mediated serotonin release in these BMMC. In concert with the data on kinase activation, FK506 and RAP did not affect mast cell degranulation in BMMC (Fig. 5,B). The MEK1 inhibitor PD98059 37, 38 also inhibited serotonin release and ERK2 activation in Ag-stimulated BMMC in a dose-dependent fashion (Fig. 5, C and D). These data imply that MEK1 and ERK2 activation may be involved in mast cell degranulation; however, the exact mechanism whereby PD98059 inhibits mast cell degranulation is not defined.

BMMC sensitized with anti-OVA IgE were incubated with 10 μg/ml OVA for 3 h, and cytokine secretion in the medium was measured by ELISA. Aggregation of FcεRI on BMMC induced both TNF-α and IL-4 production (TNF-α, 1.44 ± 0.44 ng/10⁶ cells; IL-4, 1.14 ± 0.80 ng/10⁶ cells). As shown in Fig. 6,A, wortmannin inhibited both TNF-α and IL-4 production in a dose-dependent manner; the production of IL-4 appeared to be more sensitive to the drug. CsA, but not FK506 and RAP, inhibited both TNF-α and IL-4 production in a similar manner (Fig. 6, B and C). The inhibition by CsA was complete at 1 μg/ml, suggesting that inhibition of calcineurin completely blocks cytokine production in BMMC. In contrast, addition of the MEK1 inhibitor failed to alter TNF-α or IL-4 production (Fig. 6,D), although it significantly inhibited ERK2 activation (Fig. 5 D). These data not only indicate the role of calcineurin in BMMC cytokine production, but also imply the absence of a role for ERK or MEK1 activation in TNF-α or IL-4 production in these cells. Addition of SCF to BMMC did not trigger any release of IL-4 or TNF-α.

Mast cells are responsible for the initiation of immediate hypersensitivity responses in allergic diseases in large part because they release preformed mediators such as histamine and generate lipid mediators. It is also known that mast cells produce cytokines in rapid response to Ag cross-linking of IgE bound to FcεRI, and that release of a number of cytokines, including TNF-α and IL-4, may play major roles in triggering and sustaining the allergic inflammatory response 39 . However, little is known about the signal transduction pathways, other than the role of calcineurin/NF-AT, that regulate cytokine gene expression in mast cells. Clues are evident from the promoters for the TNF-α and IL-4 genes, which contain nuclear factor-κB-, AP-1-, AP-2-, NF-AT-, Ets-, AP-1/ATF-, NF-AT/AP-1, and c-Maf-related elements, implicating additional signal transduction pathways that regulate cytokine gene expression 40, 41, 42, 43, 44, 45, 46, 47, 48, 49 .

In our previous studies, we demonstrated aggregation of FcεRI resulted in the rapid activation of JNK, p38, and ERK2 in the mouse mast cell line, MC/9 9, 10 . In MC/9 cells, wortmannin, at concentrations that inhibit PI3-kinase activity 19, 50, 51, 52, 53, 54 , strongly inhibited JNK and partially inhibited p38 activation, but not ERK2 activation 9, 10 . As shown in this study, these three members of the MAP kinase family were also activated following aggregation of FcεRI of BMMC. The effects of wortmannin on FcεRI-mediated JNK, p38, and ERK2 activation in BMMC were virtually identical to those observed in MC/9 cells 9, 10 . Mechanistically, these results indicate that there is an early separation in the signaling pathways activated through FcεRI to differentially regulate the ERK, p38, and JNK sequential protein kinase pathways.

Signaling through the tyrosine kinase receptor, SCFR, also resulted in activation of these three MAP kinase family members in BMMC. Wortmannin inhibited SCFR-induced JNK activation, but did not affect p38 or ERK2 activation in BMMC (Figs. 2 and 7). The effects of wortmannin on SCFR-induced JNK activation in BMMC contrast with the resistance demonstrated in MC/9 cells 55 . These results suggest that in BMMC, PI3-kinase activation is required for JNK activation through either FcεRI or SCFR, and that, at least in MC/9 cells, the SCFR can utilize a wortmannin-insensitive pathway to activate JNK. One of the downstream events following PI3-kinase is the activation of Akt1 20, 21 . As shown in this study, addition of Ag to passively sensitized cells or SCF stimulated Akt1 activity, and this activity was completely blocked in the presence of wortmannin in BMMC.

Addition of the immunosuppressant CsA inhibited FcεRI-mediated JNK activation and partially inhibited p38 activation without affecting ERK2 activation. Another inhibitor of calcineurin activation, FK506, did not affect FcεRI-mediated JNK activation in BMMC, almost certainly due to our demonstration that BMMC do not express the binding protein FKBP12 (data not shown). In T cells, CsA has been demonstrated to inhibit cytokine production by inhibiting calcineurin activity and the nuclear translocation of NF-AT 56 . However, based on the data presented in this work and our studies of MC/9 cells 55 , CsA can also inhibit the FcεRI-mediated induction of JNK activation and JNK-activated transcription factors. These results indicate an additional mechanism whereby CsA can regulate cytokine production in mast cells. In contrast to FcεRI signaling, SCFR-induced activation of JNK and p38 was insensitive to CsA, further indicating that signaling through FcεRI and SCFR can utilize different upstream pathways for the activation of JNK and p38.

CsA inhibits the FcεRI-mediated degranulation of rat basophilic leukemia cells and human basophils without affecting phosphatidylinositol hydrolysis or Ca²⁺ fluxes 30 . Similarly, CsA, but not FK506, inhibited serotonin release in BMMC. CsA did not affect FcεRI-mediated Ca²⁺ fluxes in these BMMC (data not shown). Although the mechanism underlying the inhibition of the degranulation response remains to be defined, the inhibition of calcineurin is suspected to play a role 36 . The inhibitor PD98059 also abolished FcεRI-mediated degranulation, suggesting that MEK1 and ERK activation are involved in the degranulation response. PD98059, on the other hand, had no effect on cytokine production, indicating that MEK1 or ERK2 activation is not required for this response. These data clearly distinguish the downstream signaling requirements for the degranulation response and cytokine production following FcεRI aggregation.

We previously showed that the wortmannin-sensitive pathways play an important role in TNF-α production in MC/9 cells 10 . It appears that calcineurin and NF-AT, the MEK kinase/JNK kinase/JNK pathway, and PI3-kinase may play a role in the synthesis and secretion of cytokines such as TNF-α and IL-4 in response to the activation through FcεRI in BMMC. In addition to cytokine production, wortmannin also inhibited FcεRI-mediated degranulation, as previously reported 34, 35 . Thus, the activation of PI3-kinase is important for both the secretion of preformed mediators and the induction of cytokine synthesis. However, neither activation of PI3-kinase alone nor JNK alone is sufficient to induce cytokine production, since addition of SCF, which can trigger both of these responses, cannot induce TNF-α or IL-4 production (data not shown). SCF induced little serotonin release (2% net release) or consistent arachidonic acid release (6% net release) at the concentrations used to activate JNK. Our findings define the importance of the calcineurin and PI3-kinase pathways in IgE-mediated signaling through FcεRI in BMMC.

We thank Drs. Akihiro Oshiba, Gregory Cieslewicz, Anthony Joetham, and Saiphone Webb for their assistance.