Rcan1 negatively regulates FcɛRI-mediated signaling and mast cell function

Suppression subtractive hybridization (SSH) allows us to identify target genes even without a prerequisite knowledge of these genes. SSH was used to identify genes that are differentially expressed in FcɛRI-activated mast cells. Total RNA from TNP-BSA–treated or untreated mouse bone marrow–derived mast cells (BMMCs) were used to generate double-stranded cDNAs. A subtracted cDNA library was constructed to enrich the transcripts that were differentially expressed by mast cells after TNP-BSA stimulation. The subtraction efficiency was evaluated by comparing GAPDH cDNA levels by PCR in serial dilutions of subtracted and nonsubtracted control cDNA (Fig. S1, available). The PCR products of the subtraction were cloned and sequenced after a differential hybridization procedure to exclude false positives. Surprisingly, among 33 positive clones sequenced, 8 clones matched the Rcan1 gene (exons 4, 5, 6, and 7), suggesting the abundance of this gene product in FcɛRI-activated mast cells (Table S1 and Fig. S2). One clone matched the Egr1 gene (Table S1). Reverse Northern blot analysis using one of the identified Rcan1 clones (2C3) as a probe confirmed the increased expression of Rcan1 in the TNP-BSA–stimulated sample (Fig. 1 A).

We further examined the kinetics of Rcan1 expression in mast cells in response to FcɛRI aggregation using real-time quantitative PCR. Given that the expression of any Rcan family members by mast cells has not been reported previously, we also examined Rcan2 and Rcan3. Two Rcan1 isoforms have been reported, the isoform Rcan1-4 encoded by exons 4, 5, 6, and 7, and the isoform Rcan1-1 encoded by exons 1, 5, 6, and 7 (17). Total RNA was isolated from TNP-BSA–treated mouse BMMCs and analyzed by real-time quantitative PCR for two Rcan1 isoforms, as well as for Rcan2 and Rcan3. Levels of Rcan mRNA expression were normalized to GAPDH in each sample. Rcan1-1 and Rcan3 showed low levels of expression under resting conditions, whereas Rcan2 was undetectable in both resting and activated cells (unpublished data). Among these Rcan family members tested, Rcan1-4 was the only member found to be up-regulated by TNP-BSA stimulation. The Rcan1-4 level in the unstimulated mast cells was undetectable. Rcan1-4 expression peaked at 60 min after TNP-BSA stimulation (Fig. 1 B). PCR-amplified Rcan1 products (Rcan1-4) were also separated in agarose gels and visualized by ethidium bromide staining. A representative gel is presented in Fig. 1 C. Strong Rcan1-4 expression can be seen at 60 min after TNP-BSA stimulation.

To determine Rcan1 expression at the protein level, IgE-sensitized mast cells were treated with TNP-BSA for various times and subjected to Western blot analysis. Rcan1 expression was observed in TNP-stimulated mast cells (Fig. 1 D).

To examine a role of Rcan1 in IgE-dependent mast cell activation, we chose to obtain mouse BMMCs from Rcan1-deficient mice. Exons 5 and 6 are deleted from the Rcan1 gene. Thus, these mice are deficient in Rcan1 products (Rcan1-1 and Rcan1-4) (18). Bone marrow cells from Rcan1-deficient and wild-type mice were cultured in conditioned medium and were analyzed by flow cytometry for c-kit and IgE receptor expression. No difference of the c-kit and IgE receptor expression after 5 wk of culture was observed between wild-type and Rcan1-deficient cells (Fig. S3 A, available). BMMCs were also used for Toluidine blue staining. Morphologically, no difference was observed between Rcan1-deficient and wild-type mast cells (Fig. S3 B). This result suggests that mast cells mature normally in the absence of Rcan1.

To determine the specific signaling pathways regulated by Rcan1 in FcɛRI-mediated mast cell activation, we examined activation of NFAT, NF-κB, and MAPK pathways in Rcan1-deficient BMMCs. For the study of NFAT pathway activation, an NFAT luciferase assay, EMSA, and an immunofluorescence assay were used. Rcan1-deficient and wild-type BMMCs were transfected with NFAT luciferase plasmid. Cells were then sensitized with anti-TNP IgE and activated with TNP-BSA for 18 h. A significant increase in NFAT luciferase activity was seen in Rcan1-deficient BMMCs after TNP-BSA stimulation compared with that in wild-type BMMCs (Fig. 2 A). These data suggests that Rcan1 is a negative regulator of IgE-dependent NFAT activation.

To further examine NFAT activation, an NFAT DNA probe was synthesized based on the NFAT binding sequence in the mouse IL-13 promoter and labeled with ³²P. Nuclear extracts from Rcan1-deficient and wild-type BMMCs after TNP stimulation for various times (5–360 min) were subjected to EMSA using a ³²P-labeled NFAT probe. As shown in Fig. 2 B, NFAT activities were enhanced in Rcan1-deficient BMMCs at the later time points (180 and 360 min) after TNP stimulation. In contrast, little difference was observed at the early time points (5–60 min) between Rcan1-deficient and wild-type BMMCs. This result suggests that Rcan1 is a negative regulator by turning off the TNP-activated NFAT signal. Densitometry analysis of NFAT DNA binding activity is shown in Fig. 2 C. The binding specificity of nuclear proteins to the NFAT DNA sequence was verified through competitive binding by the nonradioisotope-labeled NFAT probe but not by a mutant probe (Fig. 2 D). The binding specificity was further verified by supershift assay. Nuclear protein binding to the NFAT probe was strongly blocked by anti-NFATc1 antibody or by anti-NFATc2 antibody (Fig. 2 E).

To examine whether Rcan1 regulates calcineurin activity in mast cells, Rcan1-deficient and wild-type BMMCs were sensitized with IgE and treated with TNP-BSA. BMMCs were then lysed and subjected to analysis for calcineurin activity. Increased calcineurin activity was observed in Rcan1-deficient BMMCs (Fig. 2 F).

To examine NFAT nuclear translocation, IgE-sensitized BMMCs were treated with TNP-BSA for 6 h. Cells were then fixed, permeabilized, and stained with anti-NFATc1 antibody. The cell nucleus was visualized by DAPI staining. TNP-induced NFAT nuclear translocation was visible in wild-type and Rcan1-deficient cells. Enhanced NFAT nuclear translocation was seen in Rcan1-deficient cells when compared with wild-type cells (Fig. S4, available).

In light of a critical role for transcription factor NF-κB in the regulation of gene expression (19), we examined whether Rcan1 is required in FcɛRI-mediated NF-κB pathway activation. Rcan1-deficient BMMCs and wild-type cells were transfected with NF-κB luciferase plasmid. After sensitization with anti-TNP IgE, BMMCs were stimulated with TNP-BSA for 5 h. As shown in Fig. 3 A, Rcan1-deficient BMMCs showed significantly enhanced NF-κB activity in response to TNP stimulation, suggesting that Rcan1 is needed for the negative regulation of FcɛRI-mediated NF-κB activation. NF-κB activity is regulated by IκB. The effects of Rcan1 on FcɛRI-mediated IκB phosphorylation and degradation were examined by Western blotting. Aggregation of FcɛRI induced a rapid phosphorylation of IκB (5 min), followed by IκB degradation (5 and 20 min) and subsequent regeneration of this molecule (60, 180, and 360 min; Fig. 3 B). Interestingly, the early events of IκB phosphorylation and degradation at 5 and 20 min were not affected by Rcan1 deficiency. In contrast, phosphorylation at the later time point (60 min) was significantly enhanced in Rcan1-deficient BMMCs (Fig. 3 B). This process was associated with an increased regeneration of IκBα at 60 min after TNP-BSA stimulation (Fig. S5, available). These results suggest that Rcan1 is likely involved in “turning off” FcɛRI-mediated activation signals, a notion consistent with that in NFAT signaling.

MAPKs represent a distinct mechanism in the regulation of mast cell function. Previous studies have established a major role of the MAPKs c-Jun N-terminal kinase (JNK) and p38 in FcɛRI-mediated mast cell activation (20, 21). Accordingly, we tested whether Rcan1 affects FcɛRI-mediated JNK and p38 signaling. Cell lysates from TNP-stimulated BMMCs were subjected to Western blotting for phospho-JNK, phospho-p38, and their total proteins. As shown in Fig. 3 C, TNP stimulation induced an increase of p38 phosphorylation (5 min) and JNK phosphorylation (20 min). Subsequently, levels of p38 and JNK phosphorylation returned to basal conditions (180–360 min). Similar patterns of TNP-induced phosphorylation (the increase in the early phase and decrease in the late phase) of p38 and JNK were observed for Rcan1-deficient and wild-type mast cells. These data suggest that Rcan1 is not required for FcɛRI-mediated MAPK activation and inactivation processes. We also tested Akt phosphorylation, which is involved in the activation of the PI3K pathway. Similarly, no difference in the Akt phosphorylation pattern was observed between Rcan1-deficient and wild-type BMMCs (Fig. 3 C). In addition, a similar pattern of FcɛRI-mediated Syk phosphorylation was observed between wild-type and Rcan1-deficient BMMCs (Fig. 3 C).

Several cytokines such as IL-6, IL-13, and TNF play important roles in allergy and are regulated by the transcription factors NFAT and NF-κB. To determine whether Rcan1 is involved in the regulation of FcɛRI-mediated cytokine production, Rcan1-deficient and wild-type BMMCs were sensitized with anti-TNP IgE and stimulated with TNP-BSA for various times. IL-6, IL-13, and TNF production in cell-free supernatants was determined by ELISA. Expression of these cytokine mRNAs was analyzed by real-time quantitative PCR using total RNA isolated from cell pellets. Production of IL-6 was significantly enhanced by 87, 170, and 115% at the time points of 3, 6, and 18 h, respectively, in Rcan1-deficient BMMCs compared with wild-type BMMCs (Fig. 4 A). Importantly, real-time PCR analysis revealed that TNP-BSA–induced IL-6 mRNA expression was prolonged in Rcan1-deficient BMMCs (Fig. 4 B). PCR-amplified products separated by agarose gel electropheresis also revealed a prolonged TNP-BSA–induced IL-6 expression in Rcan1-deficient BMMCs (Fig. 4 C). Similarly, TNP-BSA–induced IL-13 and TNF production was enhanced at the protein level, whereas the mRNA expression of these cytokines was prolonged in Rcan1-deficient cells (Fig. 4, D–I). Enhanced production of IL-6, IL-13, and TNF at the protein level is likely caused by the prolonged mRNA expression of these cytokines. The prolonged mRNA expression of these cytokines supports the notion that Rcan1 functions as a negative regulator by turning off FcɛRI-activated signals. This is consistent with the negative role of Rcan1 in the regulation of FcɛRI-induced NFAT and IκB–NF-κB activation.

To further determine a role of Rcan1 in FcɛRI-mediated cytokine production, wild-type and Rcan1-deficient BMMCs were transfected with Rcan1-pCMV vector or empty vector pCMV. These cells were then sensitized with anti-TNP IgE and stimulated with TNP-BSA. Cell supernatants were collected for testing cytokine production. Both wild-type and Rcan1-deficient cells transfected with Rcan1-pCMV showed decreased IL-6 and TNF production after TNP-BSA stimulation, suggesting that forced expression of Rcan1 reduced FcɛRI-mediated cytokine production (Fig. 5).

Mast cell–derived cytokines contribute to the development of late-phase allergic reactions (22, 23). To examine whether Rcan1 contributes to FcɛRI-mediated late-phase cutaneous reaction in vivo, mice were sensitized with anti-DNP IgE 24 h before a solution of 0.3% dinitrofluorobenzene (DNFB; hapten) was applied epicutaneously to the ear skin or footpad. IgE/antigen-specific edema was determined by tissue weight and thickness. Rcan1-deficient mice exhibited increased IgE/antigen-specific edema when measured by tissue weight (Fig. 6, A and B). Tissue thickness, measured by a digital micrometer, also showed similar results (Fig. 6, C and D). These findings are consistent with the increased IL-6, IL-13, and TNF production by Rcan1-deficient BMMCs in vitro.

To examine whether Rcan1 plays a role in IgE-dependent mast cell degranulation, wild-type and Rcan1-deficient BMMCs were sensitized with anti-TNP IgE and stimulated with TNP-BSA for 20 min. Mast cell degranulation was determined by testing β-hexosaminidase release. Increased mast cell degranulation was observed in Rcan1-deficient mast cells (Fig. 6 E). Mast cell mediators released by degranulation contribute to the anaphylactic reactions in vivo. To examine whether Rcan1 deficiency affects passive cutaneous anaphylaxis in vivo, the left ears of mice were passively sensitized with 20 ng anti-DNP IgE, whereas the right ears were injected with saline as a control. 24 h later, mice were injected with 100 µg DNP-BSA containing Evans blue dye via the tail vein. 30 min later, tissues from both ears were collected for extraction of Evans blue dye to determine vascular permeability. Increased vascular permeability (passive cutaneous anaphylaxis) was observed in Rcan1-deficient mice (Fig. 6 F).

To exclude the possibility that the increased cutaneous allergic responses is caused by the abnormal mast cell development in vivo, the ears, tongues, and back skin from Rcan1-deficient and wild-type mice were used to examine the presence of mast cells. A similar number and morphology of mast cells were observed in tissues from Rcan1-deficient and wild-type mice (Fig. 6, G and H).

Next, we examined how the negative Rcan1 signal is activated during FcɛRI activation. Our SSH assay using RNAs from TNP-BSA–stimulated BMMCs identified one clone (2A10) that matched the Egr1 gene (Table S1). Real-time quantitative PCR analysis showed that Egr1 gene expression peaked at 15 min after TNP-BSA stimulation. In contrast, Rcan1 expression began to increase at 15 min and peaked at 60 min (Fig. 7 A). These data suggested a sequential gene expression relationship between Egr1 and Rcan1 in mast cells and prompted us to analyze the promoter sequence of Rcan1, where we identified a putative Egr1 binding domain (Fig. 7 B).

To determine whether the putative Egr1 binding sequence on the Rcan1 promoter between −158 and −142 is functional, we generated a series of constructs for a luciferase assay containing various regions of the Rcan1 promoter. All constructs start at −67 bp from the ATG codon. The 5′ end of each construct varies to include different parts of the putative Egr1 binding sequence (Fig. 7 C, left). Mouse BMMCs were transfected with these constructs, sensitized with anti-TNP IgE, and stimulated with TNP-BSA. Luciferase activity was determined. As shown in Fig. 7 C, both constructs pR-luc283 and pR-luc158, which contain the full length of the Egr1 binding sequence from −158 to −142, showed strong luciferase activity. The construct pR-luc154, which contains a partial Egr1 binding sequence from −154 to −142, showed decreased luciferase activity. In contrast, the construct pR-luc141, which does not contain an Egr1 binding region, and the empty plasmid pGL4-Basic showed little luciferase activity. This result suggests that the Egr1 binding sequence in the Rcan1 promoter is functional.

To further confirm that the Egr1 binding sequence is functional in response to FcɛRI aggregation, we synthesized a DNA probe containing the Egr1 binding sequence in the Rcan1 promoter for a gel shift assay. Nuclear proteins were isolated from BMMCs after TNP-BSA stimulation and subjected to EMSA. Strong DNA binding was seen in TNP-BSA–stimulated cells (Fig. 7 D). The binding specificity was confirmed by competition with the nonradiolabeled probe (E, unlabeled probe) but not with a mutant probe (Em, unlabeled probe; Fig. 7 D).

Furthermore, a chromatin immunoprecipitation (ChIP) assay was performed to confirm FcɛRI-induced binding of Egr1 to the Rcan1 promoter. Nuclei from TNP-BSA–stimulated and unstimulated BMMCs were subjected to enzymatic digestion followed by immunoprecipitation using anti-Egr1 or control antibodies. Immunoprecipitates were subjected to PCR using a pair of primers that amplify the Rcan1 promoter region encompassing the Egr1 binding sequence. Specific Egr1 binding to the Rcan1 promoter was observed in TNP-BSA–stimulated cells (Fig. 7 E, lane 8) but not in unstimulated cells (Fig. 7 E, lane 7). This result suggested that TNP-BSA stimulation induced Egr1 binding to the Rcan1 promoter.

To evaluate whether Egr1 is required for FcɛRI-induced Rcan1 expression, we generated Egr1-deficient and wild-type BMMCs. Egr1-deficient BMMCs mature normally and express similar levels of FcɛRI compared with wild-type BMMCs. BMMCs were sensitized with anti-TNP IgE and stimulated with TNP-BSA. Total RNA was isolated and analyzed by real-time quantitative PCR for Rcan1 (isoform Rcan1-4) expression. TNP-BSA–induced Rcan1 expression was significantly reduced in Egr1-deficient BMMCs (Fig. 8 A). A PCR-amplified Rcan1 product (Rcan1-4) was also separated on an agarose gel. A representative gel is presented in Fig. 8 B. Reduced Rcan1 expression in Egr1-deficient BMMCs can be seen after TNP-BSA stimulation. Thus, Egr1 is required for FcɛRI-induced Rcan1 expression.

The present study demonstrates an as of yet unknown pathway linking activation to inhibition in FcɛRI-mediated signaling. Notably, we found that an FcɛRI aggregation–induced activation signal initiates de novo synthesis of Rcan1. Rcan1 was identified as a negative regulator, required for turning off FcɛRI-activated signals. FcɛRI-mediated Rcan1 inhibits calcineurin activity, which leads to down-regulation of NFAT and NF-κB activation, and subsequent cytokine production and inhibition of late-phase allergic reactions. Moreover, Rcan1 is also required for the negative regulation of mast cell degranulation and acute passive cutaneous anaphylaxis. Thus, both the preformed mediator release and the production of newly synthesized cytokines are negatively regulated by Rcan1.

It is noteworthy that Rcan1 selectively regulated FcɛRI-mediated NFAT and NF-κB pathway but not MAPK and Akt activation. The activity of NFAT is regulated by its phosphorylation state, which is controlled by the opposing action of the serine/threonine phosphatase, calcineurin, and the serine/threonine kinase, glycogen synthase kinase 3 β. Under resting conditions, NFAT is phosphorylated and localized in the cytoplasm. Upon stimulation, it is dephosphorylated by calcineurin and translocates to the nucleus. Studies have indicated that Rcan1 can bind to and inhibit calcineurin (24–26). Thus, it is possible that FcɛRI-induced Rcan1 blocks NFAT activation through inhibition of calcineurin. Consistent with this model, we found that calcineurin activity was increased in Rcan1-deficient mast cells. Various pharmaceutical molecules targeting the calcineurin–NFAT pathway have successfully been used for the treatment of allergic inflammation (16). This highlights the importance of the NFAT pathway in allergic diseases.

In addition to NFAT activation, we also found that FcɛRI-induced IκB phosphorylation and NF-κB activity were enhanced in Rcan1-deficient mast cells. Rcan1 is able to regulate the NF-κB pathway through direct interaction with NF-κB–inducing kinase or IκB (27, 28). In addition, calcineurin can dephosphorylate IκB, leading to the subsequent nuclear translocation of NF-κB (15). Thus, Rcan1 could also indirectly regulate the NF-κB pathway through calcineurin. Enhanced FcɛRI-induced IκB–NF-κB pathway activation in Rcan1-deficient mast cells likely contributes to the increased cytokine production and enhanced late-phase cutaneous allergic reactions.

FcɛRI-mediated mast cell production of cytokines such as IL-6, IL-13, and TNF may be beneficial or harmful in the setting of innate or allergic responses. Efficient production of mast cell mediators would be beneficial for the effective development of host defenses, whereas overproduction of mast cell mediators could be harmful. Accordingly, a temporal regulatory mechanism is necessary that allows the initial FcɛRI-mediated signaling to proceed in the early phase and permits the termination of the activated signal subsequently in a timely manner. We found that the new synthesis of Rcan1 was dependent on Egr1, which is synthesized de novo after FcɛRI stimulation (6). These data suggest that activation signals, including Egr1, proceed to initiate inflammatory cytokine expression, but subsequently Rcan1 is induced, which serves as a negative regulator to inhibit FcɛRI-mediated signals by inhibiting calcineurin activity. The new link of Egr1 and Rcan1 may represent a transition from activation to inhibition in FcɛRI signaling.

Rcan1 deficiency not only up-regulates cytokine production, it also enhances IgE-dependent mast cell degranulation in vitro and passive cutaneous anaphylaxis in vivo. Calcineurin activity has been associated with mast cell exocytosis (29). Thus, it is possible that Rcan1 modulates IgE-dependent mast cell degranulation and passive cutaneous anaphylaxis through inhibiting calcineurin activity. However, a recent paper demonstrated that Rcan1 regulates vesicle exocytosis and fusion pore kinetics through novel unknown mechanisms other than calcineurin (30). Indeed, Rcan1 also directly interacts with several additional targets other than calcineurin. These include Raf-1 (31), 14-3-3 (32), and NF-κB–inducing kinase (28). All of these molecules have been directly or indirectly associated with exocytosis.

RCAN1 was first identified as an expressed sequence on a yeast artificial chromosome clone from human chromosome 21 (33). The Rcan1 gene consists of seven exons, of which exons 1–4 can be alternatively transcribed to produce different mRNA isoforms (17). Members of this new Rcan family also include Rcan2 and Rcan3 (17). The functional distinctions among these family members are unclear. It is interesting to note that the isoform of Rcan1-4, which is encoded by exons 4, 5, 6, and 7, was the only member to be up-regulated by FcɛRI aggregation. This result suggests that this isoform may be functionally distinct from the other family members.

Egr1 can be induced by a variety of stimuli such as sepsis (34) and tissue injuries (35). Rcan1 has also been associated with several diseases (36). The link between Egr1 and Rcan1 identified in this study will offer unforeseen potential for the regulation of signaling in a wide range of Egr1- and Rcan1-associated physiological processes and human diseases.

Rcan1-deficient mice were generated as previously described (18). Egr1-deficient mice and control C57BL/6NTac mice were purchased from Taconic. The protocols were approved by the University Committee on Laboratory Animals, Dalhousie University, in accordance with the guidelines of the Canadian Council on Animal Care.

Antibody to Rcan1 (AP6315c) was purchased from Abgent Inc. Antibodies to phospho-JNK (Thr 183/Tyr 185), JNK, phospho-p38 MAPK (Thr 180/Tyr 182), phospho-Syk (Tyr 352), phospho-Akt (Ser 473), and Akt were purchased from Cell Signaling Technology. Antibodies to p38 MAPK and actin were purchased from Santa Cruz Biotechnology, Inc. FITC-conjugated rat anti–mouse CD117 (c-kit) mAb and FITC–rat IgG2a were purchased from Cedarlane Laboratories Ltd. FITC-conjugated rat anti–mouse IgE (IgG1) and FITC–rat IgG1 were purchased from BD.

Mouse primary cultured BMMCs were cultured as previously described (6). BMMCs were passively sensitized with IgE from TIB-141 cells (American Type Culture Collection). Cells were then stimulated with 10 ng/ml TNP-BSA (Biosearch Technologies). Mast cell degranulation was determined by measuring β-hexosaminidase release.

Nuclear protein extracts were obtained using a nuclear extract kit (Active Motif) according to the manufacturer's protocol. EMSA was performed as previously described (6). The following synthesized double-stranded oligonucleotides (Sigma-Aldrich) were used: NFAT binding concensus sequence (N) on mouse IL-13 promoter 5′-AAGGTGTTTCCCCAAGCCTTTCCC-3′ and the mutant sequence (mutant Nm) 5′-AAGGTGTCCATCCAAGCCTCCTAC-3′; and Egr1 binding concensus sequence (E) on Rcan1 promoter 5′-TCCCCGCCCCCAGGGGGCTGGCTCTCC-3′ and the mutant sequence (Em) 5′-TCCCCTACCCCAGGGGGCTGGCTCTCC-3′ (italics indicate mutated sequences). For competition assays, 1 µl nonradiolabeled wild-type or mutant oligonucleotides (50-fold excess of radiolabeled probe) were added and incubated for 15 min before the addition of the radiolabeled probe. For supershift assay, 4 µg NFATc1 (Thermo Fisher Scientific) or NFATc2 (Santa Cruz Biotechnology, Inc.) was added and incubated for 30 min before the addition of the radiolabeled probe.

The mRNA levels of various genes were quantified using TaqMan MGB probes and TaqMan Master Mix on a sequence detection system (ABI Prism 7000; Applied Biosystems). GAPDH was used as an endogenous reference. Data were analyzed using the relative standard curve method according to the manufacturer's protocol. A mean value of each gene after GAPDH normalization at the time point showing highest expression was used as a calibrator to determine the relative levels of Rcan1, IL-6, TNF, IL-13, IκBα, or Egr1 at different conditions. In addition, PCR products were resolved on a 2% agarose gel and stained with ethidium bromide.

Cells lysates (20–30 µg) were subjected to electrophoresis in 12% SDS-polyacrylamide gels. Gels were transferred to polyvinylidene difluoride membrane, blotted with primary and secondary antibodies, and detected by an enhanced chemiluminescence detection system (Western Lightning Plus-ECL; PerkinElmer).

Rcan1 expression vector was constructed by inserting the Rcan1 gene into pCMV plasmid. BMMCs were resuspended at 4 × 10⁶ cells/transfection in Amaxa nucleofector solution and electroporated with 8 µg DNA using Amaxa Nucleofector Device (program U-023). After 24 h, mast cells were sensitized with IgE for 18 h and challenged with 10 ng/ml TNP-BSA for 6 h.

Phosphatase activity was measured using a calcineurin assay kit (BIOMOL International L.P.). Wild-type and Rcan1-deficient BMMCs were sensitized with IgE and stimulated with 10 ng/ml TNP-BSA for 6 h. Lysates were prepared, and phosphatase activity was measured according to the manufacturer's instructions.

ChIP assays were performed using the ChIP-IT kit (Active Motif) according to the manufacturer's instructions. In brief, BMMCs were fixed with 1% formaldehyde, and the nuclei were subjected to an enzymatic digestion with 5 U of enzymatic shearing cocktail solution for 25 min at 37°C. Sheared chromatin was immunoprecipitated with 4 µg Egr1 (Santa Cruz Biotechnology, Inc.) or control IgG (Active Motif). 0.4% of the input DNA and 5% of the precipitated DNA were then used as templates for each PCR, consisting of 36 cycles of 20 s at 94°C, 30 s at 59°C, and 30 s at 72°C. PCR products were separated by a 2% agarose gel. Primers for amplification of the Rcan1 promoter region were 5′-AGCAAACCTTAGCGCCTTTT-3′ (forward) and 5′-AAGAGAACGAGCGAGACCAC-3′ (reverse).

The plasmid pGL4.10 [Luc2] (Promega) was used to construct pR-Luc plasmids containing Rcan1 promoter fragments. A fragment of the Rcan1 promoter (−283 to −67, relative to the transcription start site) was generated by PCR with the forward primer 5′-GCTCTAGACTAAGGTGTTGACGTCACC-3′ (XbaI site underlined) and the reverse primer 5′-TTTAAGCTTCCTTTGCAAGAGAACGAGCG-3′ (HindIII site underlined). The PCR product was purified, digested with XbaI and HindIII, and ligated into the NheI/HindIII sites of pGL4.10 using T4 DNA ligase and sequenced to verify orientation. It was designated as pR-Luc283. To generate progressive 5′-unidirectional truncations by PCR, a series of forward primers was used in combination with the same reverse primer and pR-Luc283 as the template. The forward primers began with −158 (5′-CTGGAAGCCAAGATCTCCCCG-3′; BglII site underlined), −154 (5′-GAAGCCAAGCGGTACCGCCC-3′; KpnI site underlined), and −141 (5′-CCAGGCAGCTGGCTCTCCGCG-3′; PvuII site underlined). The purified PCR products were digested with corresponding restriction enzymes, treated with mung bean exonuclease, and subcloned into the EcoRV/HindIII sites of pGL4.10. They were designated pR-Luc158, pR-Luc154, and pR-Luc141, respectively. Each construct was verified by sequencing the insert and plasmid-flanking region.

4 × 10⁶ BMMCs were cotransfected with 3 µg of various luciferase plasmids and 1 µg of control plasmid (pRL-TK; Promega) using a mouse T cell nucleofector kit (Amaxa) and the Amaxa Nucleofector Device (program U-023). Luciferase plasmids used include pNFAT-Luc plasmid (Agilent Technologies), pNF-κB-Luc, and pGL4-Luc containing different Rcan1 promoter regions. After electroporation, BMMCs were plated in culture medium and allowed to recover for 24 h. Subsequently, the cells were sensitized with anti-TNP IgE for 18 h and challenged with TNP-BSA for 5 or 18 h. Firefly and Renilla activities were sequentially quantified using a dual-luciferase reporter assay system (Agilent Technologies).

Total RNA (0.6 µg) from no treatment and TNP-treated BMMCs were reverse transcribed using the SMART PCR cDNA Synthesis Kit (Clontech Laboratories, Inc.) and amplified in 19 cycles of PCR, according to the manufacturer's protocol. These cDNAs were subjected to two rounds of subtractive hybridization using the PCR-Select cDNA Subtraction Kit (Clontech Laboratories, Inc.). Forward-subtracted cDNA (no treatment cDNA as driver and TNP-treated adaptor-ligated cDNA as tester) was amplified by two rounds of nested PCR using adaptor-specific primers and Platinum High Fidelity Taq Polymerase (Invitrogen). The resulting PCR products were treated with Taq polymerase (Invitrogen) for 10 min and purified using a MinElute column (QIAGEN) before subcloning into the pCR4-TOPO TA cloning vector (Invitrogen). The ligations were transformed into TOP10 chemically competent cells and plated on lysogeny broth (LB) agar (Invitrogen) containing 100 µg/ml ampicillin (Sigma-Aldrich). Transformed colonies were picked into separate wells of 96-well plates containing 100 µl of LB ampicillin, grown for 4 h, arrayed onto four Hybond-N+ membranes (GE Healthcare), and grown overnight on LB agar supplemented with ampicillin. The colonies were lysed and fixed to the Hybond-N+ membranes using 0.5 M NaOH/1.5 M NaCl, neutralized with 1.5 M NaCl/1 M Tris (pH 7.4), and UV cross-linked in a UV Stratalinker (Agilent Technologies). The forward- and back-subtracted nested PCR products, as well as nonsubtracted tester control and nonsubtracted driver control PCR products, were digested with EagI (New England Biolabs, Inc.) to remove the adaptor sequences, purified with a MinElute column, and labeled with α-[³²P]dCTP (GE Healthcare) using the RediPrime labeling kit (GE Healthcare). Unincorporated nucleotide was removed using MicroSpin G-25 columns (GE Healthcare). Each denatured probe (2.5 × 10⁶ dpm) was hybridized with one membrane in 20 ml of Church and Gilbert's buffer overnight at 70°C. The membranes were washed and exposed to Hyperfilm (GE Healthcare). Clones that only hybridized with the forward-subtracted probe were picked into LB AMP, and plasmid DNA was extracted using the Qiaprep Spin Kit (QIAGEN). DNA sequencing was performed on a capillary sequencer (CEQ8000; Beckman Coulter) using the M13rev primer.

DNA sequence analysis was performed using the DNAStar Megalign program. Sequence homologies were determined at the National Center for Biotechnology Information using the BLAST program.

Mice were sensitized by intradermal injection of 20 ng anti-DNP IgE mAb (Sigma-Aldrich) into the left ears, whereas the right ears received saline as a control. After 24 h, mice were challenged by i.v. injection of 100 µg DNP-BSA in 200 µl of Evan's blue dye (1% wt/vol; Sigma-Aldrich). 30 min later, an 8-mm ear punch was collected in 300 µl of formamide and incubated at 80°C for 2 h in water bath to extract the Evan's blue dye. The absorbance was determined at 620 nm.

Mice were passively sensitized by i.v. injection of 2 µg anti-DNP IgE mAb (Sigma-Aldrich). After 24 h, a cutaneous reaction was elicited by the application of 20 µl DNFB (0.3% wt/vol; Sigma-Aldrich) in acetone/olive oil (4:1) to both sides of the left hind paw or left ear, and 20 µl of acetone/olive oil to the right hind paw or right ear as a control. The thickness of the foot pad or ear was measured using a digital micrometer after 24 h. The weight of the hind paw or ear punch (5 mm) was also determined. The thickness and weight of the right ear or right hind paw (treated with acetone/olive oil only) were used as baseline values. The DNFB-induced increment of tissue thickness and weight was expressed as a percentage of the baseline values.

Analysis of variance and the paired Student's t test were used for statistical evaluation of data. Results were considered significant when P < 0.05. Throughout the paper, data are expressed as means ± SEM.

Fig. S1 shows the subtraction efficiency of the SSH method using RNAs from TNP-stimulated and unstimulated mast cells. Fig. S2 shows the gene location of the clones identified by SSH. Fig. S3 shows normal maturation of BMMCs from Rcan1-deficient mice. Fig. S4 shows increased nuclear translocation of NFAT in Rcan1-deficient mast cells compared with wild-type mast cells after TNP-BSA stimulation. Fig. S5 shows increased IκBα synthesis (particularly at 60 min) in Rcan1-deficient mast cells after TNP-BSA stimulation. Table S1 shows eight Rcan1 clones and one Egr1 clone identified by SSH.

We thank F. Liu for her assistance in cell culture and passive cutaneous anaphylaxis assays.

T.-J. Lin is supported by grants from the Canadian Institutes of Health Research (MOP 68815 and MOP 81355), the Canadian Cystic Fibrosis Foundation, and the Izaak Walton Killam (IWK) Health Center. J. Molkentin is supported by a grant from the National Institutes of Health. J. Berman is supported by grants from the Canadian Institutes of Health Research (ROP 85505), the Nova Scotia Health Research Foundation, and the IWK Health Center. Y.J. Yang is supported by a postdoctoral fellowship from the IWK Health Center.

The authors have no conflicting financial interests.