Munc18-2, but not Munc18-1 or Munc18-3, controls compound and single-vesicle–regulated exocytosis in mast cells

Mast cells (MCs) play pivotal roles in many inflammatory conditions including infections, anaphylaxis, and asthma. MCs store immunoregulatory compounds in their large cytoplasmic granules and, upon stimulation, secrete them via regulated exocytosis. Exocytosis in many cells requires the participation of Munc18 proteins (also known as syntaxin-binding proteins), and we found that mature MCs express all three mammalian isoforms: Munc18-1, -2, and -3. To study their functions in MC effector responses and test the role of MC degranulation in anaphylaxis, we used conditional knockout (cKO) mice in which each Munc18 protein was deleted exclusively in MCs. Using recordings of plasma membrane capacitance for high-resolution analysis of exocytosis in individual MCs, we observed an almost complete absence of exocytosis in Munc18-2–deficient MCs but intact exocytosis in MCs lacking Munc18-1 or Munc18-3. Stereological analysis of EM images of stimulated MCs revealed that the deletion of Munc18-2 also abolishes the homotypic membrane fusion required for compound exocytosis. We confirmed the severe defect in regulated exocytosis in the absence of Munc18-2 by measuring the secretion of mediators stored in MC granules. Munc18-2 cKO mice had normal morphology, development, and distribution of their MCs, indicating that Munc18-2 is not essential for the migration, retention, and maturation of MC-committed progenitors. Despite that, we found that Munc18-2 cKO mice were significantly protected from anaphylaxis. In conclusion, MC-regulated exocytosis is required for the anaphylactic response, and Munc18-2 is the sole Munc18 isoform that mediates membrane fusion during MC degranulation.


Mast cells (MCs)
derive from hematopoietic progenitors, circulate in immature form, and migrate into different tissues where they complete their differentiation. The widespread distribution of MCs throughout the body favors fast immune and inflammatory responses, including anaphylaxis (1,2). To accomplish this, MCs release a variety of mediators through several mechanisms, including regulated exocytosis. During exocytosis, the membrane of a secretory vesicle fuses with the plasma membrane, releasing its cargo into the extracellular space and translocating proteins associated with or integral to its membrane to the plasma membrane. Regulated exocytosis in MCs (MC degranulation) is characterized by the almost immediate release of mediators that are premade and stored in large secretory vesicles (MC granules), such as histamine, MC proteases, and other enzymes. This requires MC activation by stimuli that usually employ Ca 2ϩ and diacylglycerol as second messengers (3,4). Degranulation in MCs uses both single-vesicle and compound exocytosis (5,6). In single-vesicle exocytosis, the membrane of a single MC granule fuses with the plasma membrane. In multivesicular compound exocytosis, granules fuse with each other before fusing with the plasma membrane. In sequential compound exocytosis, granules fuse with granules already fused with the plasma membrane. Both forms of compound exocytosis allow the rapid discharge of granules located deep within the cell (5,7).
Most membrane fusion events in eukaryotes are mediated by soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNARE) proteins. The exocytic SNARE proteins include isoforms of synaptosomal-associated protein 25 (SNAP-25), Syntaxin (Stx), and vesicle-associated membrane protein (VAMP), proteins that form a highly stable coiled-coil structure (trans-SNARE complex) that pulls apposing membranes tightly together (8). The formation and function of the SNARE complex requires the intervention of the Sec1 (Saccharomyces cerevisiae secretory mutant 1)/Munc18 (mammalian homolog of Caenorhabditis elegans uncoordinated gene 18) (SM) family of proteins (9). Among the seven mammalian SM proteins, Munc18-1, -2, and -3 (encoded by the Stx-binding protein/Stxbp1, Stxbp2, and Stxbp3 genes) participate in several steps of exocytosis. Munc18 locks Stx into a "closed" conformation preventing the formation of productive trans-SNARE complexes (10,11). Afterward, interactions with Munc13 catalyze the transition of Stx to an open conformation (12,13) and promotes binding of Stx's SNARE domain with those of SNAP-25 and VAMP (12,14). Then Munc18 binds to the N-terminal peptide of Stx (15,16). This enables the interaction of Munc18 with the trans-SNARE complex to further stabilize it and facilitate efficient membrane fusion (17,18). In some forms of exocytosis this process is also controlled by Ca 2ϩ sensors such as Synaptotagmin (19).
We have shown that Synaptotagmin-2 and Munc13-4 control regulated exocytosis in MCs (20,21), but the role of Munc18 proteins in MC regulated exocytosis has not been studied in mature MCs. We found that cultured RBL-2H3 cells (a rat basophilic leukemia cell line) express the three Munc18 isoforms (22). Disagreement exists on the role of Munc18-1 in MCs: some could not document expression of Munc18-1 in RBL-2H3 cells (23), others found expression and a functional role of Munc18-1 in degranulation of RBL-2H3 cells (24,25), and others observed that the absence of Munc18-1 had no effects in fetal liver-derived MCs (26).
Overexpression and knockdown of Munc18-2 in RBL-2H3 cells inhibited degranulation (23,24,27). We found that bone marrow-derived MCs from a Munc18-2 hypomorphic mutant mouse had a partial defect in regulated exocytosis (28). However, the extent of the dependence of MC exocytosis on Munc18-2 has not been tested, given that it has never been completely deleted before. Munc18-3 regulates exocytosis in several immune (29 -31) and nonimmune cells (32)(33)(34). Taking into account that Munc18-3 interacts almost exclusively with Stx4 (35)(36)(37) and that Stx4 participates in MC cell degranulation (38 -43), it was postulated that Munc18-3 could play a role in MC exocytosis, but this role has yet to be examined. Finally, the role of any Munc18 protein in compound exocytosis has not been explored.
Given these controversies and knowledge gaps and the fact that findings in cell lines sometimes are not reproduced in mature MCs (20,44), we decided to study the three Munc18 proteins in fully developed MCs to determine whether their functions are supplementary, additive, or antagonistic to each other in the control of MC-regulated exocytosis. We conditionally deleted all three isoforms in MCs and observed a failure in regulated exocytosis only in Munc18-2-deficient MCs. The defect was severe-no exocytosis, single or compound, could be detected in mutant MCs-but it did not affect MC differentiation, number, distribution, or granule biogenesis. Finally, these abnormal MC responses had prominent effects on the reactions of mutant mice to systemic anaphylaxis.
RT-qPCR from freshly isolated peritoneal MCs from Munc18-2 and -3 mutant mice showed comparable expression between F/F and ϩ/ϩ animals and complete deletion in ⌬/⌬ mice. Also, deletion of these two paralogs did not interfere with the expression of the other Munc18 isoforms (Fig. S1). Immunoblots from peritoneal cell-derived MCs (PCMCs) showed that introduction of the loxP sequences did not alter expression of any Munc18 protein in the three F/F lines compared with ϩ/ϩ controls (Fig. 1, D-F). Also, all three ⌬/⌬ mouse lines lacked expression of the targeted Munc18 gene only in MCs, but in other tissues its expression was comparable with that of the corresponding F/F littermates. The disappearance of the immunoblot bands in ⌬/⌬ MCs proved the specificity of the three antibodies.

Effects of Munc18 proteins on MC exocytosis
The fact that MCs express the three Munc18 isoforms provided the opportunity to test their functions simultaneously in one system and to determine whether their roles are unique or redundant. The membrane capacitance (C m ) of a cell is proportional to its surface area, which increases as the secretory vesicle membrane is incorporated into the plasma membrane during exocytosis (47,48). We obtained time-resolved C m measure-4 B. F. Dickey, personal communication.

Munc18 proteins in mast cell degranulation
ments from MCs using the whole-cell patch clamp configuration while stimulating exocytosis by intracellular dialysis of GTP␥S and Ca 2ϩ (49,50). The mean intracellular [Ca 2ϩ ] achieved, as measured with Fura-2, was 410 Ϯ 21 nM (means Ϯ S.E.). The baseline C m of all the MCs studied was 6.0 Ϯ 0.3 pF. There were no differences among all genotypes in these two values. We used the gain of C m over time (Fig. 2, A-C) to calculate the total gain in C m over baseline (⌬C m ) as a measure of the total amount of exocytosis (Fig. 2G). We normalized the curves (Fig. 2, D-F) and derived the rate of gain between 40 and 60% of total ⌬C m (Fig. 2H), which corresponds to the maximum speed of exocytosis. We also recorded the time between cell access and the beginning of the rise in C m to test whether there was any delay between stimulus and response (Fig. 2, I and J). Capacitance measurements can also identify individual fusion events, which register as almost instantaneous steps in ⌬C m , with the size of each step proportional to the surface area and thus the volume of each secretory vesicle (5,47,48,51). We measured the amplitudes of the fusion events to determine whether the sizes of the compartments being exocytosed were affected (Fig. 2, K and L), because a lack of large fusion events correlates with a failure in compound exocytosis (5,21).
The lack of expression of Munc18-1 (Fig. 2, A, D, G, H, I, and K) or Munc18-3 ( Fig. 2, C, F, G, H, J, and L) in MCs did not alter the total amount or speed of exocytosis, the time to initiation of the response, or the size of the vesicles fusing with the plasma membrane. Although we saw no differences among Munc18-2 ϩ/ϩ , Munc18-2 ϩ/Ϫ , and Munc18-2 F/F MCs, the deletion in Munc18-2 ⌬/⌬ MCs had a profound effect. These cells barely increased their C m after activation (Fig. 2, B and G), and what little they gained was at a very low speed (Fig. 2, E and H). The defect was so severe that we could not determine the interval to exocytosis because we could not detect an exocytic burst (21). Also, the number of steps recorded from Munc18-2 ⌬/⌬ MCs was not enough to obtain a reliable distribution histogram of their sizes, but all of the few steps we observed were Յ8 fF, which corresponds to the size of single granule fusion events, suggesting that compound exocytosis could be affected. After we found that Munc18-1 and -3 were not functionally required for exocytosis in MCs, we continued our studies using only the Munc18-2 mutants.

Effects of Munc18-2 on MC ultrastructural changes dependent on exocytosis
The whole-cell patch clamp configuration cannot detect intracellular fusion events; therefore we turned to the analysis of ultrastructural changes on stimulated MCs by EM and stereology (21). Others have previously reported that the plasma membrane of MCs undergoing degranulation increases its area and complexity (6), their granules increase in volume and lose electron density because of hydration and release of dense core material (52), and as granules fuse with each other inside the cell, they lose their membrane boundaries and create large compound compartments (5).

Munc18 proteins in mast cell degranulation
was compared with a scale unique for every cell profile, ranging from the most electron-dense material in the cell (value 0) to the lightness of the extracellular space (value 150). In these individual scales, all granules fell in the 0 -120 range, and we used these values to generate a frequency distribution of granule lucencies. Granules from activated Munc18-2 ϩ/ϩ MCs were widely distributed, whereas those from stimulated Munc18-2 ⌬/⌬ MCs had a narrow left-shifted distribution, almost identical to that of control unstimulated Munc18-2 ϩ/ϩ MCs (Fig. 3B). When we quantified the volume density (Vv) of hypodense granules (electron lucency Ն 60), they were present in Munc18-2-sufficient MCs but almost absent in Munc18-2deficient MCs, to a level similar to that of unstimulated WT controls (Fig. 3C). As the radius of a sphere increases, its surface-to-volume ratio (quantified as surface density; Sv) decreases. The reduction in the Sv of the granules of Munc18-2 ϩ/ϩ MCs after activation indicates an increase in the size of their granules (swelling), a change that was absent in Munc18-2 ⌬/⌬ MCs (Fig. 3D). Also missing from Munc18-2-deficient MCs were intracellular compound multivesicular compartments (Fig. 3E). Finally, the failure to add plasma membrane area via exocytosis explains the lack of increase in Sv of the cell in the absence of Munc18-2 ( Fig. 3F). All these results confirm that Munc18-2 is required for MC exocytosis and extends this requirement to compound exocytosis.

Effects of Munc18-2 on MC degranulation
Next, to prove that the electrophysiologic changes we observed in the absence of Munc18-2 correspond to a defect in MC degranulation, we evaluated the secretion of mediators stored in MC granules: ␤-hexosaminidase and histamine (54). PCMCs were sensitized with a specific IgE and then exposed to the antigen (DNP) or PMA and ionomycin (PI). In Munc18-2 ϩ/ϩ , Munc18-2 ϩ/Ϫ , and Munc18-2 F/F MCs, we obtained the expected bell-shaped dose-response curve of released ␤-hexosaminidase to increasing doses of allergen, whereas Munc18- Color legend in A applies to all panels. G and H, average total C m gain (⌬C m ) (G) and rate of ⌬C m between 40 and 60% of total ⌬C m (H) from all mutant MCs. I and J, interval between start of dialysis of GTP␥S and Ca 2ϩ and beginning of the exocytic burst. The few fusion events recorded from Munc18-2 ⌬/⌬ MCs did not meet criteria to identify an exocytic burst. K and L, frequency distribution of C m step sizes between 1 and 15% of total ⌬C m . Signals of Ͻ4 fF have been removed for clarity. There were not enough events from Munc18-2 ⌬/⌬ MCs to perform an analysis. Number inside boxes in G, number of cells studied, obtained from five to nine animals of each genotype, applies to G-L; white line, mean; box, 25th to 75th percentile; whiskers, 5th to 95th percentile. *, p Յ 0.05; ***, p Յ 0.001; all compared with ϩ/ϩ unless otherwise specified.

Munc18 proteins in mast cell degranulation
2 ⌬/⌬ MCs exhibited a flat response (Fig. 4A). The almost absent secretion observed in Munc18-2 ⌬/⌬ MCs was not rescued by using PI as a stronger stimulus (Fig. 4B) and was confirmed when we measured secretion of histamine (Fig. 4C).

Characterization of Munc18-2-deficient MCs
We had two objectives: to determine whether a structural abnormality and not a pure functional exocytic defect could explain the strong phenotype observed in MCs lacking Munc18-2 and to establish whether an alteration in MC number, distribution, or development could affect the results of our in vivo studies. Fluorescence microscopy of ear sections stained with FITC-avidin were used to obtain the density of dermal MCs among all Munc18-2 genotypes, and no differences were noted ( Fig. 5A and Table 1). In cytospins of peritoneal lavages, we found the same fraction and number of MCs (Table 1), which had similar gross morphology and metachromasia (Fig.  5B). EMs of freshly harvested peritoneal MCs were obtained (Fig. 5C), and our stereological analysis (Table 1) revealed no differences in Vv and Sv of granules. The simultaneous constancy of these two values is only possible if the number and size of granules is unaffected. There were also no differences in cell profile area and Sv of the cells, indicating that the size and surface complexity of the MCs were similar at rest. In flow cytometry, we identified MCs as double-positive CD117 ϩ / Fc⑀RI␣ ϩ cells. The fractions of MCs in peritoneal lavages (Fig.  5D) and in PCMCs after 2 weeks in culture media enriched with IL3 and SCF (Fig. 5E) were similar, and the expression levels of these two receptors on the surface of these MCs were almost identical (Table 1). Overall, the level of expression of Munc18-2 did not affect the ultrastructure and number of MCs or their expression of receptors important for their development and activation.

Passive systemic anaphylaxis
It is postulated that MC degranulation plays a central role in anaphylaxis (55). Therefore, we tested the consequences of the MC exocytic abnormalities we identified above in a model of anaphylaxis. We chose passive sensitization because this model depends on IgE, MCs, and histamine (56). In this model, loss of body heat correlates with the severity of the anaphylactic response (57); thus we recorded the drop in body core temperature after systemic allergen challenge (Fig. 6A). The lack of response in MC-deficient Kit W-sh/W-sh mice confirmed the MC-dependence of this reaction. Our most impressive finding was a markedly reduced response in Munc18-2 ⌬/⌬ mice at early (Fig. 6B) and late (Fig. 6C) time points when compared with all other genotypes. This blunted response correlated with an almost absent histological evidence of MC degranulation in connective tissues (Fig. 6, D and E) and with a pronounced reduction in circulating histamine after the antigen challenge (Fig. 6F). The less severe anaphylactic response, faster recovery, and decreased number of degranulated MCs in tissues in the Munc18-2 ⌬/⌬ mice all correlated with our functional and morphological findings on defective MC regulated exocytosis in Munc18-2-deficient MCs.

Discussion
We found that mature peritoneal MCs express all three Munc18 isoforms. The low level of expression of Munc18-1 in MCs (Fig. 1) could explain why some investigators failed to detect it in cell cultures (23). Although three other groups showed that MC lines express Munc18-1 (24 -26), they observed opposite effects of this protein on secretion. In one study, fetal liver-derived MCs cells activated via Fc⑀RI did not have any impairment in degranulation in the absence of Munc18-1 (26). Others showed that the knockdown of Munc18-1 expression in RBL-2H3 cells negatively affected degranulation and that expression of Munc18-1 rescued the severe secretory defect in Munc18-1/-2 double-knockdown RBL-2H3 cells (24). A third study showed that the positive effect of Munc18-1 on RBL-2H3 exocytosis depends on its phosphorylation by PKC (25). Although Munc18-1 may control exocytosis in this cell line, given that we observed no exocytic

Munc18 proteins in mast cell degranulation
abnormality in mature MCs lacking Munc18-1 using an extremely sensitive method (Fig. 2), we conclude that this isoform is not required for exocytosis in fully developed MCs.
The potential role for Munc18-3 in MC exocytosis was only hypothetical and was based on: (a) the high level of expression of Munc18-3 in MCs (Fig. 1); (b) the high affinity of Munc18-3 for Stx4 (36,37,58); (c) the proven role of this interaction in exocytosis in other cells (29, 36, 59 -63); (d) the interaction of Stx4 with other exocytic SNARES in RBL-2H3 cells (41, 64); and (e) the finding that altering the expression levels of Stx4 in RBL-2H3 cells (38,41) and MCs (39,40) modified their degranulation upon activation. Notwithstanding all that, we found that exocytosis is independent of Munc18-3 in mature MCs (Fig. 2) and conclude that this protein is dispensable for MC degranulation.
Given the discordant results on the roles of Synaptotagmin-2 (20, 44) and Munc18-1 (see above) obtained from RBL-2H3 cells versus mature MCs, we consider it important to study Munc18-2 in fully developed MCs. Also, although there was previous evidence that lowering expression levels of this protein decreases degranulation in RBL-2H3 cells (24,27,65), bone marrow-derived MCs (28,65), and mature MCs (28), the only approach to evaluate the degree of dependence of this process on this protein is the complete removal of Munc18-2 from the system. Different from the partial exocytic defect that we observed in Synaptotagmin-2-deficient (20) and Munc13-4deficient MCs (21), there was almost no residual exocytosis after deleting Munc18-2 by electrophysiologic (Fig. 2) and morphologic (Fig. 3) criteria. Moreover, this defect was entirely functional and not a consequence of morphologic abnormalities in the mutant MCs ( Fig. 5 and Table 1). Munc18-1 can interact with many of the Stxs proven to be functional partners of Munc18-2 (36, 66, 67), and it has been shown recently that Munc18-3 can bind to Stxs other than Stx4 in vitro (68), suggesting that Munc18-1 and -3 could compensate for the loss of Munc18-2. Nevertheless, we found an almost complete failure in exocytosis in mature MCs lacking Munc18-2 despite almost normal expression levels of Munc18-1 and -3 in these mutant MCs (Fig. S1), indicating that Munc18-1 and -3 do not play any role in MC exocytosis in vivo, not even a redundant one.
The residual exocytosis observed in MCs lacking Munc13-4 allowed us to document a significant delay between stimulus and response, and we interpreted that as a failure in vesicle priming (21). Because there were very few exocytic events in the absence of Munc18-2, we could not reliably perform this analysis. Therefore, we cannot draw any conclusions on the role of Munc18-2 in vesicle priming in MCs.
It has been estimated that the fusion of a single MC granule with the plasma membrane should increase its C m by ϳ7 fF (69). The analysis of the C m step sizes in Munc13-4 ⌬/⌬ MCs showed that the residual exocytosis was comprised mostly of steps of Յ8 fF, indicating that the small remaining response was mainly dependent on single-vesicle exocytosis (21). In the case of Munc18-2 deficiency, the scarcity of fusion events made it impossible to perform a formal distribution analysis, but no event of Ͼ8 fF was noted, suggesting a defect in compound exocytosis (5). Three possible explanations were that in the absence of Munc18-2 there is a failure in granule biogenesis,

Munc18 proteins in mast cell degranulation
that fusion between granules was impaired, and that intergranular fusion was taking place, but the resulting large multivesicular compound compartments were unable to fuse with the plasma membrane. Because C m recordings are unable to distinguish among these options, we drew upon EM analysis of resting (Table 1) and activated (Fig. 3) MCs. Once again, stere-

Munc18 proteins in mast cell degranulation
ology of profiles of activated MCs proved to be an excellent complement to MC electrophysiology (21). We observed that although the MC granules were normal in size and number, there was an almost total failure in granule-to-granule fusion. Hence, like Munc13-4, Munc18-2 is required for the heterotypic fusion of vesicle to plasma membrane and the homotypic fusion between vesicle membranes, and consequently it controls both single-vesicle and compound exocytosis. The failure in MC degranulation in the absence of Munc18-2 documented in vitro was confirmed in vivo when we observed a marked reduction in plasma histamine levels and histologic markers of MC degranulation in Munc18-2 ⌬/⌬ mice subjected to a model of systemic anaphylaxis (Fig. 6). Relative to WT controls, the loss in hypothermic response in this model is more complete in Munc18-2 ⌬/⌬ mice (Fig. 6) than in Munc13-4 ⌬/⌬ mice (21). We think that the more severe defect in exocytosis in the absence of Munc18-2 (the residual exocytosis observed in Munc13-4 ⌬/⌬ MCs was not present in Munc18-2 ⌬/⌬ MCs) accounts for the difference.
On the other hand, compared with controls, the anaphylactic response is significantly decreased but not absent. The residual anaphylactic response in Munc18-2 ⌬/⌬ is more likely due to residual MC function than to the participation of other allergic effector cells (e.g. basophils and platelets) because MC-deficient mice (70) are completely protected from anaphylaxis (Fig. 6A). MCs could be grouped into many subpopulations depending upon different criteria (71). A useful classification (72) distinguishes between constitutive/connective tissue MCs, evolutionarily the most primitive and the most abundant in naïve mammals (73), and inducible/mucosal MCs, which depend on T-cell-derived cytokines for development (74,75). Cma1 (mMCP-5) is expressed mainly in the former (71,76,77). Hence, it is possible that a small population of mucosal MCs that do not undergo recombination and do not lose Munc18-2 expression accounts for the weak reaction. Another possibility is that MC responses independent of exocytosis, such as eicosanoid generation, are responsible. In conclusion, MC exocytosis is an important component of the anaphylactic response, and Munc18-2 is the sole, nonredundant Munc18 isoform that mediates both homotypic and heterotypic membrane fusion during MC degranulation. In the Munc18-1 cKO line, exon 2 of Stxbp1 is flanked by two loxP sites, and the frameshift induced by Cre recombination results in nonsense mutations and absence of the protein (78). In the Munc18-2 cKO mouse, the targeted exon was exon 1, and recombination removes the translation initiation codon. 4 To create a Munc18-3 cKO by homologous recombination, we built a targeting vector to insert a loxP upstream of exon 1 (GRCm38; Chr3:1088440680) of the mouse Stxbp3 gene and a second loxP followed by the phosphoglucokinase promoterneomycin resistance gene (PGK-Neo) flanked by two Flp recognition target sites in intron 1 (Chr3:108840293-108840125). The herpes simplex virus thymidine kinase gene was introduced outside the homology arms (Fig. 1B). The construct was electroporated into C57BL/6 ES cells, and genetically modified mice were created as described (21). The mutant mice were crossed with B6.129S4-Gt(ROSA)26Sor tm1(FLP1)Dym /RainJ mice to remove PGK-Neo, leaving exon 1 flanked by two loxP sites (F allele). We crossed Munc18-1 F/F , Munc18-2 F/F , and Munc18-3 F/F mice with B6.C-Tg(CMV-cre)1Cgn/J CMV-Cre to delete the targeted gene in the germ line and produce global heterozygote (ϩ/Ϫ) lines, and with Tg(Cma1-cre)ARoer mice to generate MC-specific deletant mice (⌬/⌬) (21,45).

Munc18 proteins in mast cell degranulation
weeks (37°C, 5% CO 2 , biweekly media changes). For flow cytometry, 10 6 peritoneal lavage cells or PCMCs were blocked with anti-mouse CD16/CD32 and labeled with phycoerythrincyanine-7 anti-mouse CD117 and allophycocyanin anti-mouse Fc⑀RI␣ antibodies (all from BD Biosciences) and then analyzed, recording the number of double-positive CD117 ϩ /Fc⑀RI␣ ϩ cells and their mean fluorescence intensity in each channel. To isolate MCs from peritoneal lavages by FACS, the cells were labeled as above, and the double-positive ones were sorted.

Histology
The ears were processed as described (28). In brief, 5-m cross-sections embedded in paraffin were stained with 2.5 mg/ml of FITC-avidin and 20 mM Hoechst (both from Thermo Fisher) for 1 h at room temperature. Cells with FITC ϩ granules surrounding a Hoechst ϩ nucleus were counted as MCs. The density of MCs was expressed as number of MCs per mm 2 of dermis, which was delineated as the area between the epidermis and the auricular cartilage or muscle using the autofluorescence of these tissues under a Texas Red filter. For quantification of MC degranulation in tissues, the mice were euthanized 90 min after the anaphylactic challenge. The lips were excised, fixed (0.5% paraformaldehyde, pH 7.0, 4°C, overnight), and embedded in OCT compound. Sections of 5 m were stained with toluidine blue (pH 2.0; MP Biomedicals). A MC was classified as degranulated if most (Ͼ 50%) of its metachromatic granules were found extracellularly (21).

EM and stereology
5 ϫ 10 6 peritoneal cells per sample were activated with PMA (20 ng/ml) plus ionomycin (1 g/ml) in 1 ml of PBS for 5 min at 37°C. Activation was stopped by adding cold glutaraldehyde (final concentration, 2%) on ice, and then EM samples were handled as described (21). Sections of 100-nm thickness were stained with uranyl acetate and lead citrate and then imaged on a Tecnai 12 transmission electron microscope. MC profiles containing a nuclear profile were recorded. Stereology was performed using a line-segments-and-points grid in STEPanizer (80). Points inside cell profiles and granules and line intercepts with plasma and granule membranes were counted to obtain A, Vv, and Sv (20). Agglomerates of multiple granule profiles not separated by a membrane were counted as multigranular compartments (21). The relative electron lucency of MC granules was computed using an internal scale adjusted for each cell profile. Circles with an area of 0.0366 m 2 within granules selected randomly with a cycloid grid were digitally graded using a 0 -150 point gray scale, where 0 represents the most electron-dense structure in the cell profile, and 150 represents the lightness of the extracellular space. Granules with relative electron lucency of Ͻ60 were categorized as dense, and those of Ն60 as were categorized hypodense.

MC secretion assays
Assays were performed as described (21). Briefly, 3 ϫ 10 4 PCMCs were sensitized with 100 ng/ml of mouse anti-DNP IgE (clone SPE-7; Sigma-Aldrich) for 5 h. They were washed and resuspended in 90 l of degranulation buffer. Then 10 l of degranulation buffer containing no extra compounds (negative control) or DNP-HSA (final concentrations 0.1, 1, 10, 100 and 1000 ng/ml) or PMA (final concentration 50 ng/ml) with ionomycin (final concentration, 1 M) was added. After 30 min, supernatants were collected, and the remaining cells were lysed. For ␤-hexosaminidase, 50 l of supernatant or lysates were mixed with 100 l of 4-nitrophenyl N-acetyl-␤-D-glucosaminide (3.5 mg/ml) in citrate buffer and incubated at 37°C for 90 min. The reaction was stopped by adding 100 l of glycine (pH 10.7; 400 mM). Then the absorbance at ϭ 405 nm was recorded (reference ϭ 620 nm), and ␤-hexosaminidase activity was reported as a fraction of the total cell content. ELISA (Oxford Biochemicals) was used to measure histamine in supernatants.

Electrophysiology
Whole-cell recordings from individual MCs were made using 3-8 M⍀ Sylgard-coated patch pipettes. The internal solution defined intracellular [Ca 2ϩ ] and induced degranulation and contained 0.05 mM GTP␥S, 135 mM potassium gluconate, 10 mM HEPES, 0.1 mM Fura-2, 7 mM MgCl 2 , 3 mM KOH, 0.2 mM Na-ATP, and either 10 mM K 2 EGTA with 8 mM CaCl 2 or 2.5 mM Cs-EGTA with 7.5 mM Ca-EGTA (pH 7.21, 0.304 osmoles) (49,50,81). The intracellular [Ca 2ϩ ] was determined ratiometrically from the emitted Fura-2 fluorescence at two different wavelengths (82)(83)(84), and Ca 2ϩ calibration constants determined in vitro using a multipoint calibration kit (Thermo Fisher Scientific/Molecular Probes) and analyzed in IgorPro (Wavemetrics). For recordings of membrane capacitance (C m ), membrane conductance (G m ), and series conductance (G s ), an 800 Hz sinusoidal, 30 mV peak-to-peak stimulus was applied around a holding potential of Ϫ70 mV, and the resultant signal was analyzed using the Lindau-Neher technique (85,86). For each 100-ms sweep, the average value was recorded, yielding a temporal resolution for C m , G m , and G s of ϳ7 Hz. Cells selected for analysis met the criteria of G m Յ 1,200 pS, G s Ն 35 nS, and steady-state intracellular [Ca 2ϩ ] ϭ 400 Ϯ 100 nM. The total change in C m (⌬C m ) was calculated as the difference between the prestimulus baseline and the maximum plateau in C m . Each trace of C m over time was fitted to a four-parameter sigmoid Munc18 proteins in mast cell degranulation function, normalized to its total ⌬C m and then used to obtain the rates of C m from 40 -60% of total ⌬C m . The exocytic burst was identified as successive step changes in C m Ն 8 fF sustained for Ͼ4 s, and the time between whole-cell access and the first fusion event with C m Ն 8 fF within this 4-s region was noted (21). We also measured the sizes of all the steps between 1 and 15% of total ⌬C m after eliminating any signal of Ͻ4 fF to decrease noise.

Passive systemic anaphylaxis
The mice were sensitized intraperitoneally with 20 g of mouse monoclonal anti-DNP IgE in 100 l of PBS. After 24 h, the mice were challenged intraperitoneally with 1 mg of DNP-HSA in 100 l of PBS. Body core temperature was monitored using a rectal probe (DT80 -1; General). At 90 min, the mice were euthanized, and their lips were processed as described above. Another set of mice were manipulated as above, but 15 min after the challenge, they were anesthetized with isoflurane, and blood was collected via inferior vena cava puncture with a 21-gauge needle into a citrated syringe, slowly and avoiding bubbles to prevent platelet activation. An equal volume of Tyrode's buffer (5.6 mM glucose, 140 mM NaCl, 12 mM NaHCO 3 , 2.7 mM KCl, and 0.46 mM NaH 2 PO 4 ) was added, plasma was separated by centrifugation, and histamine was measured by ELISA.

Statistical analysis
For continuous variables, we compared the means of all groups by one-way analysis of variance; if a significant difference was detected, we applied Tukey's test for multiple pairwise comparisons, Dunnett's test for multiple comparisons against the control group, and two-tailed unpaired Student's t test for simple comparisons. For categorical data, we used Pearson's 2 test or Fisher's exact test. Significance was set at p Ͻ 0.05.