A Novel Role of the L-Type Calcium Channel α1D Subunit as a Gatekeeper for Intracellular Zinc Signaling: Zinc Wave

Recent studies have shown that zinc ion (Zn) can behave as an intracellular signaling molecule. We previously demonstrated that mast cells stimulated through the high-affinity IgE receptor (FcεRI) rapidly release intracellular Zn from the endoplasmic reticulum (ER), and we named this phenomenon the “Zn wave”. However, the molecules responsible for releasing Zn and the roles of the Zn wave were elusive. Here we identified the pore-forming α1 subunit of the Cav1.3 (α1D) L-type calcium channel (LTCC) as the gatekeeper for the Zn wave. LTCC antagonists inhibited the Zn wave, and an agonist was sufficient to induce it. Notably, α1D was mainly localized to the ER rather than the plasma membrane in mast cells, and the Zn wave was impaired by α1D knockdown. We further found that the LTCC-mediated Zn wave positively controlled cytokine gene induction by enhancing the DNA-binding activity of NF- κB. Consistent with this finding, LTCC antagonists inhibited the cytokine-mediated delayed-type allergic reaction in mice without affecting the immediate-type allergic reaction. These findings indicated that the LTCC α1D subunit located on the ER membrane has a novel function as a gatekeeper for the Zn wave, which is involved in regulating NF-κB signaling and the delayed-type allergic reaction.


Introduction
Zn is an essential trace element. Approximately 10% of all the genes in the human genome may contain Zn-binding motifs [1], and the dysregulation of Zn homeostasis is linked to a wide range of physiological defects, including those affecting growth, development, and the immune system [2,3].
Recent advances have revealed the existence and importance of free or labile Zn in living organisms [4], and Zn has been increasingly recognized as a potential biological signaling molecule [5]. It is well established that synaptic Zn acts as a neurotransmitter that can mediate cell-to-cell communication [6,7,8]. In addition to such intercellular communication, Zn can act as a second messenger [9], capable of transducing extracellular stimuli into intracellular signaling events. Intracellular Zn signaling is classified into two types: early and late [5,10,11]. Late Zn signaling, which occurs several hours after extracellular stimulation, depends on changes in the expression profile of Zn-related molecules, such as Zn transporters and metallothioneins, and leads to alterations in the intracellular Zn content and/or intracellular distribution of Zn [12,13,14,15,16]. On the other hand, early Zn signaling occurs several minutes after extracellular stimulation and does not involve transcriptional changes. It is mediated by extracellular Zn's influx into the cytoplasm and by intracellular Zn's detachment from metalloproteins and release from intracellular organelles.
FceRI stimulation induces a rapid elevation of the intracellular free Zn level in mast cells, and we named this phenomenon the ''Zn wave'' [9]. The Zn wave originates in the perinuclear region, which includes the endoplasmic reticulum (ER). Our evidence suggests that it is positively involved in FceRI-mediated cytokine production in mast cells. These findings indicated a novel function for the Zn released from intracellular organelles as an intracellular second messenger, like Ca 2+ [9]. However, the gatekeeper for the Zn wave remained unknown.
In addition to the FceRI-mediated Zn wave in mast cells, the rapid elevation of intracellular Zn by several stimuli for certain cellular functions has been reported [17,18,19]. However, the mechanism for the rapid intracellular induction of free Zn in those studies, as well as in the case of the Zn wave, has remained unclear.
L-type calcium channels (LTCCs) can conduct Zn [20] and act as Zn-permeable channels on the plasma membrane of neurons and pancreatic b cells [21,22]. However, it is unclear whether LTCCs can also function in Zn's release from intracellular organs. The LTCCs are complexes that include a 1 , b, and a 2 /d subunits.
The a 1 subunit functions as the voltage sensor, selective filter, and ion-conducting pore [23], and a 1 subunit on the cell surface is proposed to require an association with the b subunit, which masks one or more ER-retention signals [24,25]. Taken together, these characteristics of LTCCs make them potential candidates for performing the Zn wave gatekeeper function [21,22].
Transcription factors of the nuclear factor kB (NF-kB)/Rel family play pivotal roles in inflammatory and immune responses [26,27]. In unstimulated cells, NF-kB is sequestered in the cytoplasm by its inhibitory proteins, the IkBs. Stimulants that activate the NF-kB pathway induce the phosphorylation and degradation of IkBs through the ubiquitin-proteasome pathway, releasing NF-kB to enter the nucleus, where it binds specific DNA sequences [28]. Mast cells secrete cytokines in response to antigen stimulation and other activators [29,30]. NF-kB acts as a key regulator for inflammatory cytokines such as IL-6 and TNF-a [31]; in mast cells, FceRI stimulation induces the nuclear translocation of NF-kB to increase these cytokines [32].
Redox regulation of NF-kB's DNA-binding activity by Zn has also been demonstrated; the mechanism involves Zn's binding to cysteine residues in the DNA-binding region of NF-kB, as shown by site-directed mutagenesis experiments [33,34,35]. These findings suggest that Zn is a signaling molecule that modulates NF-kB, although the link between the regulation of NF-kB activation and the Zn wave is unknown.
Here, we show that the pore-forming a 1D subunit of LTCC on the ER membrane plays a novel role in generating the Zn wave, and that the NF-kB signaling pathway, a key regulator of allergic responses, is one of the targets of the LTCC-mediated Zn wave in mast cells.

Expression of the LTCC a 1D Subunit on the ER Membrane in Mast Cells
LTCCs are Zn-permeable [20] and are expressed by various cell types, including non-excitable cells [36]. To determine whether this ion channel could be involved in the Zn wave in mast cells, we examined the expression of the four pore-forming a 1 subunits. RT-PCR analysis showed that cacna1d, the a 1 subunit for the Cav1.3 LTCC (a 1D ), was the dominantly expressed a 1 subunit in bone marrow-derived mast cells (BMMCs) ( Figure 1A). Furthermore, a 1D protein was observed in BMMCs by western blot analysis with an antibody against a 1D ; the signal was abolished by prior incubation of the antibody with an antigenic peptide ( Figure 1B).
We next examined the intracellular distribution of the a 1D in BMMCs. In confocal microscopic analysis, the immunofluorescent signal of a 1D was observed in the intracellular area, partially merged with that of the ER marker calnexin, but not with that of F-actin, which accumulated beneath the plasma membrane ( Figure 1C). In addition, in a discontinuous sucrose density gradient ultracentrifugation experiment, a 1D was distributed in fractions 4 and 5, which corresponded with fractions containing the ER marker SERCA2 (fractions 3 to 5), but only partially overlapped with fractions containing the plasma membrane marker LAT (fractions 1 to 4) ( Figure 1D). These results indicated that a 1D localized preferentially to intracellular organelle membranes, such as the ER membrane, rather than to the plasma membrane in mast cells.
The b subunit is known to be required for a 1 subunits to be localized to the plasma membrane and for their full activity as a channel [24,25]. As shown in Figures 1E and F, the expression levels of b subunits were very low in BMMCs, consistent with a 1D 's main localization to the ER in these cells. This intracellular localization of a 1D in BMMCs suggested that it plays a different role from that observed in other cells, in which it is located on the plasma membrane and acts as a calcium channel.

The LTCC a 1D Subunit is a Gatekeeper for the Zn Wave in Mast Cells
To examine whether LTCCs expressed on the ER membrane are involved in the Zn wave, we examined the effect of the LTCC antagonist Verapamil on the FceRI-induced Zn wave in BMMCs. Verapamil-treated BMMCs showed an impaired Zn wave compared to control cells (Figure 2A), without disturbing cell survival or FceRI expression ( Figures S1A and B). The FceRImediated Ca 2+ elevation, however, was not inhibited by Verapamil in BMMCs ( Figure 2B). The Zn wave was also inhibited in BMMCs treated with a lower concentration of Verapamil (1 mM) or with another type of LTCC antagonist, Diltiazem ( Figures S2A  and B), and Diltiazem did not affect the FceRI-mediated Ca 2+ elevation ( Figure S2C). On the other hand, treatment with the LTCC agonist (s)-(-)-BayK8644, without antigen stimulation, induced an elevation in the intracellular Zn level, but not in the Ca 2+ level (Figures 2C, D, S3A, and B). The LTCC agonistinduced increase in intracellular Zn was observed even in the absence of Ca 2+ , and it was inhibited by Verapamil (Figures 2C, D, and S3C). To reveal whether the FceRI-induced Zn wave and LTCC-mediated Zn elevation were regulated by a similar mechanism, BMMCs were stimulated with antigen in the presence of the LTCC agonist. The level of the FceRI-induced Zn elevation was similar in BMMCs with or without the LTCC agonist, indicating that the FceRI-induced Zn wave and LTCC agonistinduced Zn elevation probably occur by a similar mechanism ( Figure S3D).
All these results were consistent with the idea that a 1D , one of the a 1 subunits of LTCCs, might be a gatekeeper for the Zn wave in mast cells. To examine this possibility further, we knocked down a 1D in BMMCs by siRNA. The expression level of the mRNA for Cacna1d, but not for other Cacna1 family members, such as Cacna1f, and the protein level of a1 D were reduced in the a 1D -knockdown BMMCs compared with control cells (Figures S4). The FceRIinduced Zn wave was significantly reduced in the a 1D -knockdown BMMCs compared with control cells ( Figure 3A). On the other hand, the a 1D knockdown did not affect the FceRI-induced Ca 2+ elevation ( Figure 3B), similar to the results of Verapamil treatment. Moreover, the ectopic expression of wild-type a 1D rescued the inhibitory effect of siRNA knockdown on the Zn wave ( Figure 3C). These results indicated that the LTCC a 1D subunit is a gatekeeper for the Zn wave.

Requirement of the Zn Wave for FceRI-induced Cytokine Gene Induction, but not for Degranulation in Mast Cells
FceRI stimulation activates several downstream pathways that initiate immediate allergic inflammatory responses by eliciting mast-cell degranulation, accompanied by the rapid release of preformed chemical mediators, such as histamine and serotonin. In contrast, the mast cell-mediated delayed-type responses are mainly dependent on cytokine production. To examine whether the LTCC-mediated Zn wave could play a role in these mast-cellactivation events, we first investigated the effect of Verapamil treatment and a 1D knockdown on FceRI-mediated cytokine gene induction and degranulation. Inhibition of the Zn wave by Verapamil reduced the FceRI-mediated gene induction of Il6 and Tnfa in BMMCs ( Figure 4A). The a 1D -knockdown BMMCs also showed impaired FceRI-mediated gene induction of Il6 and Tnfa ( Figure 4B). In these siRNA experiments, the Cacna1d mRNA expression level in the a 1D -knockdown BMMCs was 20.567.1% of the control level. On the other hand, neither Verapamil treatment nor a 1D knockdown inhibited the FceRI-mediated degranulation in these cells ( Figures 4C and D). Taken together, these results indicated that the Zn wave is involved in the FceRImediated cytokine gene induction, but not the degranulation, of mast cells.

Role of the Zn Wave in the NF-kB Pathway
Since NF-kB is a master transcription factor that controls the expression of proinflammatory cytokines such as IL-6 and TNF-a in mast cells [37], the Zn wave was likely to be involved in the FceRI-induced NF-kB-signaling pathway. Inhibiting the Zn wave with Verapamil did not affect FceRI-induced IKK phosphorylation, IkB phosphorylation, or its degradation in BMMCs ( Figure 5A). Even though the upstream activation pathway of NF-kB was intact, the frequency of NF-kB p65 accumulation in the nuclei upon FceRI stimulation was reduced in the Verapamiltreated BMMCs ( Figure 5B). These results indicated that the Zn Figure 1. The a 1D subunit of LTCC is primarily expressed on the ER membrane in mast cells. (A) RT-PCR of mRNA encoding the a 1 subunit of LTCC family members (cacna1s, cacna1c, cacna1d, and cacna1f), and Gapdh in BMMCs. (B) Western blot for a 1D in BMMCs. Total cell lysates were blotted with an anti-a 1D polyclonal antibody or the same antibody pre-incubated with an antigenic peptide. Arrowhead indicates the putative a 1D signal; this signal detected by the antibody pre-incubated with antigenic peptide was 15.1611.1% of that detected by the anti-a 1D polyclonal antibody. (C) Intracellular distribution of the LTCC a 1D subunit in BMMCs examined by confocal microscopy. Representative images are shown. Staining with an anti-a 1D monoclonal antibody is in green, anti-Calnexin (ER marker) in red, or phalloidin-Alexa 546 (for F-actin beneath the plasma membrane) in red, and 49 6-diamidino-2-phenylindole, dihydrochloride (DAPI; for nuclei) in blue. (D) The postnuclear supernatant obtained from BMMCs was fractionated by ultracentrifugation in a discontinuous sucrose gradient. The collected fractions were then separated by SDS-PAGE, and the protein distributions were detected by immunoblotting with antibodies against a 1D , SERCA2 (ER), and LAT (plasma membrane). (E) The mRNA expression of LTCC b-subunit family members (cacnb1 to cacnb4) in the brain and BMMCs was examined by RT-PCR. (F) The protein levels of the b2 and b4 subunits in the brain and BMMCs were examined by western blots. doi:10.1371/journal.pone.0039654.g001 wave might be required for NF-kB's localization to nuclei, but not for its upstream activation pathway. We further investigated whether the Zn wave was required for the nuclear import step. For this, we used the exportin inhibitor leptomycin B (LMB). The frequency of NF-kB in nuclei was elevated by LMB treatment, and this effect was observed even in the presence of Verapamil ( Figure  S5A), suggesting that the Zn wave might be required for post nuclear translocation events, rather than for the nuclear import step.
Therefore, we next examined the DNA-binding activity of NF-kB p65 in nuclei. Whereas the DNA-binding activity of NF-kB in nuclei was elevated after FceRI stimulation in BMMCs, it was reduced in Verapamil-treated cells. This reduction in DNAbinding activity was recovered by adding Zn with the FceRI stimulation ( Figure 5C). Consistent with this result, the reduction in FceRI-mediated cytokine gene induction in Verapamil-treated BMMCs was recovered by Zn supplementation ( Figure 5D). These results indicated that the Zn wave might participate in the signal transduction for cytokine gene induction by enhancing the DNAbinding activity of NF-kB.

Role of the Zn Wave in the Regulation of Allergic Responses in vivo
Mast cells are a major player in allergic responses such as the immediate-and delayed-type hypersensitivity reactions [30]. Passive cutaneous anaphylaxis (PCA) and contact hypersensitivity (CHS) are mouse models for the immediate-type and delayed-type allergic responses, respectively. We therefore examined the effect of Verapamil treatment on the PCA and CHS reactions. The allergic response in the PCA model was evaluated by the extravasation of Evans blue dye in the ears of mice that had been sensitized and challenged with an antigen. There was no notable difference in the extravasation of Evans blue dye in the ear of the vehicle-versus Verapamil-treated mice ( Figure 6A), indicating that the mast cell-mediated PCA reaction occurred normally in the presence of Verapamil.
We then examined the CHS response to the experimental hapten FITC, by assessing the amount of tissue swelling at the site of hapten challenge. While the vehicle-treated mice developed a robust CHS response 24 h after stimulation, the Verapamil-treated mice showed a greatly reduced response ( Figure 6B). In addition, mice treated with the other LTCC antagonist, Diltiazem, showed a decreased FITC-induced CHS response ( Figure S6). The effects of Verapamil on mast-cell activation in vitro ( Figure 4) and the allergic response in vivo together indicated that the Zn wave plays a role in FceRI-induced cytokine production, but not in degranulation, and is involved in regulating the delayed-type but not the immediatetype allergic response.

Identification of the LTCC a 1D Subunit as a Gatekeeper for the Zn Wave
It is well established that LTCCs function as voltage-gated calcium channels on the plasma membrane. In this study, we showed that the LTCC a 1D subunit was expressed in mast cells, but was localized to the ER rather than to the plasma membrane. We also showed that the expression level of LTCC b subunits, which are required for the localization of a 1 subunits to the plasma membrane [24,25], was very low in mast cells. Furthermore, the expression of ZnT-1 in mast cells [16] might support the ER The intracellular labile Zn level upon antigen stimulation was examined in control and a 1D siRNA-treated BMMCs. The difference in Newport Green intensity at 15 min between the control and a 1D siRNA-treated BMMCs was statistically significant. *P,0.05, Student's t-test. (B) The intracellular Ca 2+ level upon antigen stimulation was examined in control and a 1D siRNA-treated BMMCs. The difference in Fluo-4 intensity between the control and a 1D siRNA-treated BMMCs was not statistically significant. (C) The FceRI-mediated Zn wave was examined in a 1D siRNA-treated BMMCs with or without transfection of human a 1D . All data represent the mean + SEM. N.S., not significant, *P,0.05, ***P,0.001, Bonferroni's multiple comparison test. NPG, Newport Green. doi:10.1371/journal.pone.0039654.g003 localization of a 1D subunits, because ZnT-1 is reported to interact with b subunits on the plasma membrane, reducing their availability to bind a 1 , and thus inhibiting a 1 -subunit trafficking to the plasma membrane [38].
An important finding in the present study was that the a 1D subunit expressed on the ER membrane has little effect on FceRIinduced Ca 2+ influx, and instead plays a novel role as a gatekeeper for the Zn wave. Our data showed that LTCC antagonist treatment or a 1D knockdown inhibited the Zn wave, but did not affect the FceRI-mediated Ca 2+ elevation or FceRI-mediated degranulation, which requires an increase in intracellular Ca 2+ . In addition to the lack of effect on FceRI-mediated Ca 2+ elevation, Ca 2+ -mediated signaling was not disturbed in the Verapamiltreated BMMCs, as shown by the normal nuclear translocation of NFAT2 in these cells ( Figure S7). In addition, LTCC agonist treatment increased the level of intracellular free Zn but not of Ca 2+ in mast cells. These results showed that LTCC is not involved in the FceRI-mediated Ca 2+ regulation in mast cells. This might be because mast cells, like lymphocytes, utilize storeoperated calcium (SOC) entry as their main mode of Ca 2+ influx [39]. However, we could not rule out the ability of a 1D subunit expressed on the ER membrane to conduct Ca 2+ from the ER to the cytoplasm.
Most importantly, our finding that the LTCC a 1D subunit, when expressed on the ER membrane, has a novel function as a gatekeeper for the Zn wave also made it possible for us to address the physiological roles of the Zn wave.

Regulation of the LTCC Activation for Zn Wave Generation
The a 1 subunits of LTCC contain a voltage-sensor domain, and the channel activity is elevated after membrane depolarization. The plasma membrane potential in BMMCs is hyperpolarized after FceRI stimulation [40,41], but we found that inhibition of the FceRI-mediated plasma membrane hyperpolarization by high KCl treatment did not impair induction of the Zn wave ( Figures  S8A and B). We then examined the intracellular membrane potential using tetramethyl rhodamine methyl ester (TMRM). Treatment with the ADP/ATP transporter inhibitor bongkrekic acid inhibited the FceRI-mediated intracellular membrane depolarization, but it did not inhibit the induction of the Zn wave ( Figures S8C and D). This result suggested that intracellular membrane depolarization does not affect the Zn wave generation, although we cannot exclude the possibility that depolarization of the ER inner membrane has an effect.
Modification of the pore-forming a 1 subunit by phosphorylation has an additional effect on channel activity; in fact, cAMPmediated channel activity is reduced by site-directed mutagenesis of the PKA consensus sites of a 1D [42]. However, we did not observe a negative effect on the Zn wave by PKA inhibitor treatment ( Figure S9); therefore, PKA may not participate in the regulation of the Zn wave, at least in mast cells. As-yet unidentified regulatory proteins on the ER membrane may control this event.

Zn Wave Regulates the DNA-binding Activity of NF-kB and Cytokine Gene Induction
We found that the LTCC-mediated intracellular Zn signal upregulates the DNA-binding activity of NF-kB and the transactivation of inflammatory cytokines. NF-kB-mediated transactivation can be divided into the following three steps. First, NF-kB dissociates from IkB after IkB's phosphorylation and degradation. Second, NF-kB translocates from the cytosol to the nucleus, and finally, NF-kB binds to its target sequences. We found that the frequency of NF-kB p65 nuclear translocation was reduced in LTCC antagonist-treated cells, even though the upstream regulators were unaffected. Treatment with the exportin inhibitor LMB enhanced the frequency of NF-kB in nuclei, and this effect was observed in LTCC antagonist-and LMB-treated cells, suggesting that the Zn wave is not involved in the nuclear translocation step. Rather, our evidence indicates that the Zn wave is required for the DNA-binding activity of NF-kB.
That the LTCC antagonist treatment reduced the DNAbinding activity of NF-kB further supports this scenario. Moreover, the DNA-binding activity of NF-kB was enhanced by supplementing the cell lysate with Zn ( Figure S5B). These findings suggest that the elevated intracellular Zn caused by the Zn wave positively regulates the DNA-binding activity of NF-kB. However, although we showed that the Zn wave is involved in FceRI- mediated cytokine gene induction, LTCC agonist treatment, which induced an increase in free zinc, like the Zn wave, did not increase the mRNA induction and protein synthesis of IL-6 and TNF-a ( Figure S10). These results suggest that the Zn wave is required, but not sufficient, for the FceRI-induced cytokine productions in mast cells. Nevertheless, taken together, our results indicate that the Zn wave is a novel modulator of NF-kB activation.

The Zn Wave is Involved in the Delayed-type Allergic Response in vivo
The mast cell is one of the effector cells for allergic responses in vivo. Mast cell-derived cytokines, which are induced by FceRImediated activation of the PKC/Bcl10/Malt1/NF-kB signaling pathway, are known to be involved in delayed-type allergic responses, such as CHS [16,43]. Mast cell-derived TNF is required for the maximum CHS response; it induces the infiltration of leukocytes at the site of inflammation [44], enhances the elongation of cutaneous nerves [45], and enhances the dendritic cell migration to draining lymph nodes [46]. In this study, we revealed that treating mice with the LTCC antagonist Verapamil inhibited the CHS reaction, a delayed-type immune response, without affecting the PCA, an immediate-type response. Consistent with these results, we showed that Verapamil treatment inhibited the FceRI-mediated activation of NF-kB's DNA-binding activity and cytokine gene inductions, but not Ca 2+ elevation or degranulation in BMMCs. Thus, the inhibitory effect of Verapamil on CHS might depend at least in part on the reduction of mast cell-derived cytokine production, independent of other mediators, such as histamine. Although we do not exclude the possibility that Verapamil affects the function of dendritic cells and T cells in vivo, all our results suggest that one of the in vivo roles of the Zn wave is to regulate the allergic response by controlling cytokine production in mast cells.
In summary, we identified a novel function of the pore-forming a 1D subunit of LTCC, when it is expressed on the ER membrane, as the gatekeeper for the Zn wave in mast cells. In addition, the LTCC-mediated Zn wave may function as a positive regulator for inflammatory cytokines by enhancing NF-kB's DNA-binding activity. These findings will help us understand the regulation and importance of intracellular Zn signaling in a variety of biological responses in which Zn-susceptible proteins are involved.

Cell Culture and Mice
All animal experiments were conducted in accordance with animal protocols approved by the Animal Research Committee at RIKEN (permit number 22-013 (3)). Bone marrow-derived mast cells (BMMCs) were prepared as described previously [47]. Briefly, 8-week-old C57/BL6 mice were sacrificed, and their bonemarrow cells were cultured in RPMI 1640 supplemented with 10% FCS, 10 mU/mL penicillin, 0.1 mg/mL streptomycin, 40 mM 2-ME, and IL-3, in a 5% CO 2 and 95% humidified atmosphere at 37uC. After 4-5 weeks of culture, the cell-surface expression of FceRI and c-Kit was confirmed, and the cells were used for experiments.

Plasmid Construction and Transfection
The coding region of the human CACNA1D gene was isolated from a cDNA library of human brain (Clontech). To construct Nterminally FLAG-tagged CACNA1D, the CACNA1D fragment was amplified by PCR, followed by sequencing and cloning into the NotI and XbaI site of the expression vector p3XFLAG-Myc-CMV26 (Sigma Aldrich). BMMCs were transfected with expression vectors or siRNA using a two-step electroporator CUY21Pro-Vitro (Nepa Gene, Japan). For the electroporation, 1610 6 BMMCs were resuspended in 100 ml of OPTI-MEM, and 10 mg of plasmid DNA or 400 pmol of siRNA was added. Electroporation was carried out with 1 pore-forming pulse (275 V for 3 msec) and 10 driving pulses (20 V for 50 msec), and then the BMMCs were diluted in 1 ml of BMMC culture medium. After a 48-h incubation, the cells were used for experiments.

Microscopy
The intracellular Zn or Ca 2+ level was measured as described previously [9]. Briefly, sensitized BMMCs were allowed to adhere to a poly-L-lysine-coated glass-bottom dish or glass-bottom dish. After being incubated with 10 mM Newport Green or 5 mM Fluo-4 for 30 min at 37uC, the cells were stimulated with 100 ng/ml dinitrophenylated human serum albumin (DNP-HSA; Sigma) or 10 mM (s)-(-)-BayK8644 at 37uC. The images of fluorescent signals were captured every 10 or 30 sec with an inverted microscope (Axiovert 200 MOT, Carl Zeiss), CCD camera (Cool Snap HQ, Roper Scientific), and the system control application SlideBook (Intelligent Imaging Innovation).

Sucrose Gradient Fractionation Assay
Mature BMMCs (5610 7 ) were harvested and washed with PBS and homogenized with a Dounce homogenizer in 0.5 ml of icechilled HES buffer (250 mM sucrose, 1 mM EDTA, 20 mM HEPES, pH 7.5). The homogenate was spun at 5006g for 5 min to sediment the nuclei, and the postnuclear supernatant was used for further fractionation. The postnuclear supernatant was mixed with 0.5 ml of 0.8 M sucrose in 50 mM Tris-HCl and loaded onto the top of a discontinuous sucrose gradient (0.6, 1, 1.35, 1.65, 2 M) prepared in the same buffer. The gradient was spun in a SW 55 Ti rotor for 16 h at 100,0006g (32,100 rpm) in a Beckman ultracentrifuge, and fractions of 300 ml each were collected from the top of the tube. Proteins from each fraction were separated by SDS-PAGE.

Measurement of Cytokines
Cells were sensitized with 1 mg/mL IgE for 6 h at 37uC. After sensitization, the cells were washed twice with Tyrode's buffer (10 mM HEPES pH 7.4, 130 mM NaCl, 5 mM KCl, 1.4 mM CaCl 2 , 1 mM MgCl 2 , 5.6 mM glucose), then suspended in the same buffer containing 0.1% BSA and stimulated with polyvalent dinitrophenyl-human serum albumin (DNP-HSA, Sigma) for 30 min. TNF-a and IL-6 in the cell culture supernatants were measured with an ELISA kit (Biosource), following the manufacturer's recommendation.

BMMC Degranulation Assay
Cells were activated as described above, then spun down at 60006g for 1 min, and 50 ml of the culture supernatant or the cell pellet solubilized with 1% Triton X-100 in Tyrode's buffer was combined with 100 mL of 1.3 mg/mL p-nitrophenyl-N-acetyl-Dglucosamide and developed for 60 min at 37uC. The enzyme reaction was stopped by adding 150 mL of 0.2 M glycine-NaOH (pH 10.2), and the absorbance at 405 nm was measured with a microplate reader (Bio-Rad).

FITC-induced Contact Hypersensitivity
The FITC-induced CHS procedure was performed as described previously [46]. Briefly, mice were sensitized by applying 200 ml of 2% FITC isomer-I (FITC; Sigma-Aldrich) in a vehicle consisting of acetone-dibutylphthalate (1:1) to the skin of the back. Five days after the sensitization with FITC, the mice were pretreated with 20 ml of acetone-EtOH (1:1) or 50 mg/ml Verapamil in acetone-EtOH (1:1) and then challenged with 20 ml of vehicle alone on the right ear (10 ml on each side of the ear) and 1% FITC on the left ear (10 ml on each side). The ear thickness was measured before and at various times after FITC challenge, with an engineer's microcaliper (Ozaki).

Passive Cutaneous Anaphylaxis
A total of 2 mg IgE in 20 ml was injected subcutaneously into the ears over a period of 12 h. After the sensitization, the mice were challenged with an intravenous injection of 50 ml polyvalent dinitrophenyl-bovine serum albumin (DNP-BSA: Cosmobio, Japan) in 250 ml of saline-5 mg/mL Evans blue dye (Sigma, Japan). The extravasation of Evans blue into the ear was monitored for 30 min. The mice were then sacrificed, both ears were dissected, and the Evans blue dye was extracted in 700 ml of formamide at 63uC overnight. The absorbance of the Evans bluecontaining formamide was measured at 620 nm.

NF-kB DNA-binding Assay
The DNA-binding activity of NF-kB p65 was examined with the TransAM TM NFkB p65 kit (Active Motifs), according to the manufacturer's protocol. In brief, sensitized BMMCs (2610 6 ) were rinsed twice with Tyrode's buffer and incubated with or without 100 mM Verapamil for 30 min at 37uC, then stimulated with 10 ng/ml DNP-HSA for 30 min at 37uC with or without 1 mM pyrithione and ZnSO 4 . The cytosolic and nuclear proteins were separated using the Nuclear Extraction Kit (TransAM) according to the manufacturer's protocol. Nuclear proteins were prepared in a 50 ml volume, and 20 ml of each sample was used to estimate the amount of DNA-bound NF-kB p65, and 15 ml of each sample was subjected to SDS-PAGE to determine the total amount of NF-kB protein in the nuclear fraction. The DNA-binding activity of the NF-kB p65 in the nuclear fraction was estimated by dividing the amount of NF-kB p65 bound to the target sequence by the amount of NF-kB p65 protein in the nuclear fraction.

Statistical Analysis
All statistical analyses were performed using Statcel software. Data were analyzed by two-tailed Student's t-test or Student's ttest with Bonferroni's correction for multiple comparison. Data were considered statistically significant when the P value was less than 0.05. N.S., not significant, *P,0.05, **P,0.01, ***P,0.001.  Figure S9 Effect of PKA-mediated signaling on the FceRI-induced Zn wave. The FceRI-mediated Zn wave was examined in BMMCs treated with a PKA inhibitor at the indicated concentration. The intracellular labile Zn level was determined by staining the BMMCs with Newport Green and analyzed by flow cytometry. Data represent the mean fluorescent intensity of Newport Green 6 S.D. ***P,0.001 Student's t-test, two-tailed. NPG, Newport Green. (TIF) Figure S10 Effect of LTCC agonist-induced Zn elevation on cytokine production. The production of IL-6 and TNF-a upon antigen stimulation for 3 hours was measured by ELISA in