Type I interferon limits mast cell–mediated anaphylaxis by controlling secretory granule homeostasis

Type I interferon (IFN-I) is a family of multifunctional cytokines that modulate the innate and adaptive immunity and are used to treat mastocytosis. Although IFN-I is known to suppress mast cell function, including histamine release, the mechanisms behind its effects on mast cells have been poorly understood. We here investigated IFN-I’s action on mast cells using interferon-α/β receptor subunit 1 (Ifnar1)-deficient mice, which lack a functional IFN-I receptor complex, and revealed that IFN-I in the steady state is critical for mast cell homeostasis, the disruption of which is centrally involved in systemic anaphylaxis. Ifnar1-deficient mice showed exacerbated systemic anaphylaxis after sensitization, which was associated with increased histamine in the circulation, even though the mast cell numbers and high affinity immunoglobulin E receptor (FcεRI) expression levels were similar between Ifnar1-deficient and wild-type (WT) mice. Ifnar1-deficient mast cells showed increased secretory granule synthesis and exocytosis, which probably involved the increased transcription of Tfeb. Signal transducer and activator of transcription 1(Stat1) and Stat2 were unexpectedly insufficient to mediate these IFN-I functions, and instead, Stat3 played a critical role in a redundant manner with Stat1. Our findings revealed a novel regulation mechanism of mast cell homeostasis, in which IFN-I controls lysosome-related organelle biogenesis.


Introduction
Type I interferon (IFN-I) is a family of pleiotropic cytokines that modulate the innate and adaptive immunity and play a critical role in antiviral responses [1]. The IFN-I family consists of 13 IFN-α subtypes and IFN-β, IFN-β, IFN-κ, and IFN-ω, and others in some species [2,3]. All IFN-I molecules bind to a receptor consisting of type-I interferon (IFN-α/β) receptor (IFNAR)1 and IFNAR2, and the resulting ternary complex forms an active signaling receptor [4]. Although IFN-I interacts with IFNAR1 and IFNAR2 with different affinities, structural analyses combined with site-directed mutagenesis demonstrated that IFNAR1 is necessary for the signaling of IFN-I [2,5]. After IFN-I binds to the receptor, the Janus family kinases (Jak) PLOS  tyrosine kinase 2 (Tyk2) and Jak1 are activated by reciprocal transphosphorylation [6]. These activated kinases then mediate the transcription of a large number of genes, the so-called IFNstimulated genes (ISGs), through the phosphorylation-dependent activation of the transcription factors, signal transduction and activator of transcription (STAT)1 and STAT2 [6,7]. This Jak-Stat pathway downstream of IFNAR is critical for the activation of various types of immune cells, which shapes the adaptive immunity, and therefore disorders observed in IFNAR1 or IFNAR2 deficiency largely overlap those observed in STAT1 or STAT2 deficiency [8,9]. IFN-I is also known to activate STAT3 through IFNAR, which confers negative effects on ISG expression [10][11][12][13], but the physiological role of STAT3 in IFN signaling is still largely unclear. IFN-I has been clinically used to treat mastocytosis [14][15][16]. This rare and intractable disease is primarily caused by disordered mast cell proliferation and activation, and exhibits diverse symptoms, including pruritus, flushing, and anaphylactic shock [15]. Many of these symptoms are due to mediators released from mast cells, such as histamine [14]. IFN-I treatment ameliorates the symptoms by inhibiting mast cell proliferation, degranulation, bone marrow infiltration, or mastocytosis-related ascites/hepatosplenomegaly [17][18][19]. Within this context, studies using neoplastic mast cell lines and cultured bone marrow-derived mast cells (BMMCs) have been performed and showed that IFN-I suppresses the histamine release from mast cells [15,[20][21][22]. However, the molecular mechanisms by which IFN-I affects mast cell functions are not well understood.
Mast cells are key players in the pathogenesis of allergic responses. These cells differentiate in the bone marrow and reside in peripheral tissues, where they functionally adapt to their environment and undergo terminal maturation. Mast cells secrete their granule contents and/ or cytokines when they are activated by stimuli such as high affinity immunoglobulin E receptor (FcεRI) cross-linking by an antigen bound to its specific immunoglobulin E (IgE). The mast cell's effector functions depend largely on the well-developed lysosome-related organelles (LROs), known as secretory granules [23], which contain histamine, proteases, and other inflammatory mediators [24]. Although the precise mechanisms of LRO biogenesis and functional adaptation are still largely unclear, one of the key molecules involved is the transcription factor EB (TFEB) [25]. TFEB regulates a set of gene expressions involved in lysosome biogenesis, known as the coordinated lysosomal expression and regulation (CLEAR) gene network, and is considered the master regulator of lysosomal biogenesis [26,27]. TFEB's transcriptional activity is regulated by the mechanistic target of rapamycin complex 1 (mTORC1) pathway, in which TFEB's nuclear translocation is controlled by its cytosolic sequestration, which is regulated by mTORC1-mediated phosphorylation [28,29]. In addition, TFEB's expression levels are positively correlated with its transcriptional activity, because TFEB induces its own transcription [26]. Furthermore, the increase in TFEB protein boosts its nuclear translocation. We previously reported that mast cell secretory granules are also under the control of TFEB in a manner depending on the lysosomal environment, which is regulated by a lysosome-resident amino acid transporter [30].
Here, we used IFNAR1-deficient mice to investigate how IFN-I influences mast cells, and revealed a critical role of IFN-I in controlling mast cell homeostasis in the steady state. We found that the loss of steady-state IFN-I signaling caused dysregulated secretory granule formation, which strongly promoted the onset of systemic anaphylaxis. Unexpectedly, STAT1 and STAT2 were not sufficient to elicit this IFN-I function, and instead, STAT3 played a critical role, in a manner redundant with STAT1. Our findings reveal new information about the action of IFN-I in mast cells and provide important insight for therapeutic strategies for mast cell-mediated allergic reactions.

IFN-I signaling limits IgE-mediated systemic and local anaphylaxis
To investigate the mechanism by which IFN-I affects mast cell functions, we first examined the mast cell-dependent systemic anaphylaxis using IFNAR1-deficient mice (Ifnar1 −/− ). When mice were passively inoculated with trinitrophenyl (TNP)-specific IgE and TNP-conjugated bovine serum albumin (BSA), more severe hypothermia was observed in the Ifnar1 −/− mice than in the wild-type (WT) mice ( Fig 1A). Consistent with this observation, the serum histamine levels after FcεRI cross-linking were significantly higher in the Ifnar1 −/− mice than in WT mice (Fig 1B). These results were unexpected, because even though the mice did not receive exogenously administered IFN-I or agonistic reagents to induce IFN-I production, the IFNAR1 loss strongly influenced the onset of systemic anaphylaxis. The serum IgE concentration at steady state was similar between the Ifnar1 −/− and WT mice (Fig 1C), as was the binding of passively administered antigen-specific IgE to mast cells (Fig 1D). The peritoneal mast cell numbers were not significantly different between WT and Ifnar1 −/− mice ( Fig 1E). We also examined local anaphylaxis using the passive cutaneous anaphylaxis (PCA) model and found that the endothelial permeability was increased in the Ifnar1 −/− mice (Fig 1F and 1G). The number of ear-resident mast cells was similar between WT and Ifnar1 −/− mice ( Fig 1H). These results indicated that IFNAR1 plays a critical role in limiting mast cell-dependent anaphylaxis.

Mast cells respond to IFN-I in autocrine and paracrine manners
Constitutive IFN-I expression and secretion in the steady state are known to be important for maintaining the homeostasis of various types of cells and tissues [31]. The increased systemic and local anaphylaxis in Ifnar1 −/− mice strongly suggested that IFN-I in the steady state directly or indirectly influences mast cells. We therefore examined whether the properties of mast cells in peripheral niches were under the control of IFN-I. WT mast cells expressed IFNAR1 on their surface, although the expression level was low (Fig 2A). In contrast, the peak shift of IFNAR1 staining detected in WT mast cells was completely abolished in Ifnar1 −/− mast cells (Fig 2A). The IFNAR1 expressed on mast cells was functional, given that exogenously added IFN-I induced Stat1 tyrosine phosphorylation in WT but not in Ifnar1 −/− mast cells ( Fig  2B). This observation supported previous reports that IFN-I affects mast cell functions [21].
We next examined whether mast cells have access to IFN-I in the steady state. The transcriptions of a panel of IFN-I genes were detectable in ex vivo peritoneal cells and in fibroblasts used for a preparation of in vitro-differentiated mast cells (connective tissue-type mast cells; CTMCs) ( Fig 2C and S1 Fig). These results suggested that mast cells could be exposed to IFN-I in their environment. We also found that mast cells themselves transcribed IFN-I mRNA ( Fig  2D). However, IFN-I proteins were not detectable in CTMC culture supernatants in ELISA. Although we could not detect IFN-I secretion from cultured mast cells, we speculated that mast cells receive and respond to IFN-I continuously in their niche in the steady state, and the loss of IFN-I signaling in the Ifnar1 −/− mice altered the nature of the mast cells, which may have caused the exacerbated anaphylactic reaction.

Blockade of IFN-I signaling increases the severity of IgE-mediated systemic anaphylaxis
To examine the above possibility, we investigated the effect of an anti-IFNAR1 antibody that blocks the binding of the IFN-I that is present in the mast cell niche in the steady state. When WT mice were treated with the anti-IFNAR1 antibody, they exhibited more severe hypothermia in association with IgE-mediated systemic anaphylaxis compared with control mice, and  Fig 3A). Importantly, the anti-IFNAR1 antibody treatment also increased the histamine released into the serum of WT mice after challenging with IgE and antigen ( Fig 3B). These data collectively indicated that continuous signaling through IFNAR1 limits histamine secretion and systemic anaphylaxis. We also observed that the exogenous administration of IFN-I slightly but significantly ameliorated the systemic anaphylaxis-associated hypothermia was counted; n = 10 for WT and n = 14 for Ifnar1 −/− mice. n.s., not significant. Underlying data in A, B, C, E, G, and H can be found in S1 Data. BSA, bovine serum albumin; FcεRI, high affinity immunoglobulin E receptor; IFNAR, Interferon-α/β receptor; IgE, immunoglobulin E; IGELb4, Anti-TNP IgE monoclonal antibody (clone IGELb4); PCA, passive cutaneous anaphylaxis; TNP, trinitrophenyl; WT, wild-type. https://doi.org/10.1371/journal.pbio.3000530.g001

Loss of IFN-I signaling causes an increase in secretory granule formation
We next analyzed the characteristics of the WT and Ifnar1 −/− mast cells to determine the action point of the IFN-I. The expression levels of FcεRI and the cellular homolog of the transforming gene of a feline retrovirus (v-Kit) (c-kit) on the peritoneal mast cells in these mice were comparable (Fig 4A). However, the microscopic appearance of the secretory granules WT mice treated with an anti-IFNAR1 mAb that blocks type I IFN-I binding (n = 5) or isotype control mouse IgG1 (n = 5) on days −2 and −1 (left, 100 μg/administration; right, 50 μg/administration) and sensitized with IgE on day −1 were challenged with antigen on day 0. Hypothermia was monitored for 75 minutes. �� P < 0.01 Results are shown as the mean ± SEM of values representative of three independent experiments. (B) Serum histamine level upon IgE-mediated passive systemic anaphylaxis in mice pretreated with either the anti-IFNAR1 antibody or isotype control, as in the left panel of (A), was determined by competitive enzyme immunoassay (EIA). n = 4 for each group. �� P < 0.01 as determined by one-way ANOVA with Tukey's test. Results are shown as the mean ± SD of values representative of three independent experiments. (C) IgE-mediated passive systemic anaphylaxis in mice pretreated with IFN-α (n = 5 for each group). � P < 0.05 (A, C) Results are shown as the mean ± SEM of values representative of three independent experiments. Underlying data from A to C can be found in S1 Data. Ctrl, control; IFNAR1, interferon-α/β receptor subunit 1; IFN-I, type I interferon; IgG1, immunoglobulin G1; IgE, immunoglobulin E; Iso, isotype control; mAb, monoclonal antibody; rIFNα, recombinant interferon-α; WT, wild-type.  was different in the Ifnar1 −/− compared with the WT peritoneal mast cells. While the secretory granules in the WT mast cells were homogenous and of uniform size, those in the Ifnar1 −/− mast cells were of variable size and stainability, and included large granules ( Fig 4A). Transmission electron microscopic analyses also showed the enlarged tendency of granule diameters in Ifnar1 −/− CTMC compared with WT CTMC (Fig 4B and S2 Fig). Importantly, the Ifnar1 −/− peritoneal mast cells stored significantly more histamine compared to the WT peritoneal mast cells ( Fig 4C). In addition, the amount of stored β-hexosaminidase, a lysosomal enzyme, was also higher in the Ifnar1 −/− peritoneal mast cells ( Fig 4D). Similar results were obtained when we analyzed CTMCs (S3 Fig). These results strongly suggested that the secretory granule synthesis was increased by the loss of IFNAR1. This notion was supported by the finding that the expression of Tfeb, which is the master regulator of lysosome/LRO biogenesis, was increased in the Ifnar1 −/− CTMCs ( Fig 4E). Accordingly, the expression of the cathepsin A (Ctsa) gene, a TFEB target gene, was also up-regulated in the Ifnar1 −/− CTMCs ( Fig 4E). The elevated secretory granule formation in the absence of IFN-I signaling was confirmed by increased granuleassociated gene expressions, including that of histidine decarboxylase and secretory granule proteases in the Ifnar1 −/− CTMCs ( Fig 4F). Consistent with these observations, the FcεRIindependent systemic anaphylaxis, in which compound 48/80 administration triggers the granule exocytosis of mast cells in the absence of IgE and antigen, was more severe in Ifnar −/− than in WT mice ( Fig 4G). Taken together, these results indicated that IFN-I limits mast cell functions by restricting the secretory granule biogenesis.
The increased TFEB expression prompted us to compare the ability of LRO exocytosis between WT and Ifnar1 −/− CTMCs by evaluating the CD63 surface expression, because TFEB promotes lysosome fusion with the plasma membrane. A higher CD63 surface expression was detected in the Ifnar1 −/− mast cells compared with the WT ones after FcεRI cross-linking ( Fig  4H). This finding indicated that IFN-I also limits mast cell secretory granule exocytosis. Based on these observations, it appeared that increases in both the secretory granule formation and the granule exocytosis in Ifnar1 −/− mast cells contributed to the exacerbation of systemic anaphylaxis and the elevated histamine in the sera.

IFNAR1 loss exerts slight effects on FcεRI-proximal signaling events
We further examined whether IFNAR1 loss altered the FcεRI-mediated mast cell responses. For this analysis, we compared biochemical FcεRI-proximal signaling events between WTand Ifnar1 −/− -derived CTMCs. We detected an increase in spleen tyrosine kinase (Syk) phosphorylation in the Ifnar1 −/− CTMCs but little difference in other FcεRI-proximal signaling molecules such as phospholipase Cγ2 (PLCγ2) (Fig 5A). The pattern of FcεRI-dependent protein tyrosine phosphorylation in total cell lysates was also slightly affected by the IFNAR1 deficiency ( Fig 5B). While the FcεRI-dependent productions of tumor necrosis factor α (TNFα) and interleukin (IL)-13 were not affected by the absence of IFNAR1 (Fig 5C), the FcεRI-mediated IL-6 secretion was enhanced in the Ifnar1 −/− versus the WT CTMCs ( Fig 5C). However, the IL-6 gene transcription was comparable between the WT and Ifnar1 −/− CTMCs (Fig 5D), mast cells. The intracellular histamine level (C) or intracellular β-hexosaminidase activity (D) in WT or Ifnar1 −/− mast cells was quantified. �� P < 0.01. Results are representative of three independent experiments. (E, F) Gene expression in WT or Ifnar1 −/− CTMCs was analyzed by RT-qPCR. Hprt was used as an internal control. �� P < 0.01, � P < 0.05. (G) Compound 48/80-induced hypothermia in WT and Ifnar1 −/− mice (n = 5 for each group). � P < 0.05 as determined by one-tailed t test. (H) CD63 exposure on the cell surface upon FcεRI ligation in WT or Ifnar1 −/− CTMCs analyzed by flow cytometry. Histogram (left) or proportion of CD63 + cells at the indicated time points (right). Results are shown as the mean ± SD (B-F) or the mean ± SEM (G) of values representative of three independent experiments. Underlying data from B to H can be found in S1 Data. CTMC, connective tissue-type mast cell; Ctsa, cathepsin A; FcεRI, high affinity immunoglobulin E receptor; Hprt, hypoxanthine guanine phosphoribosyl transferase; IFN, interferon; RT-qPCR, quantitative reverse transcription PCR; Tfeb, transcription factor EB; WT, wild-type. https://doi.org/10.1371/journal.pbio.3000530.g004 Tonic IFN-I signaling limits mast cell homeostasis and systemic anaphylaxis indicating that the increased FcεRI-triggered IL-6 secretion in the absence of IFN-I signaling was not due to a transcriptional alteration but probably caused by the alteration of posttranscriptional regulations. This finding also suggested that the FcεRI-mediated signaling cascades related to these cytokine gene expressions were not affected by the absence or presence of IFNAR1-mediated signaling. In our hands, the number of CTMCs collected and the expression levels of surface molecules such as FcεRI and c-Kit were not significantly different between the Ifnar1 −/− and WT CTMCs (Fig 5E). Collectively, our results suggested that IFN-I in the steady state also limits FcεRI-dependent Syk activation but does not play an essential part of cytokine productions by mast cells.

IFN-I limits the terminal maturation of mast cells
The granule synthesis in mast cells is known to be correlated with their maturation process. Mast cell maturity can be evaluated by flow cytometry, by which peritoneal mast cells can be classified into two different populations, immature and mature mast cells [32]. Mature mast cells also possess a high side scatter (SSC) and low forward scatter (FSC) light profile, which probably reflects the mast cell granule state. Therefore, we examined whether a difference in SSC was detectable in the WT versus Ifnar1 −/− mast cells. Flow cytometric analyses of the peritoneal mast cells revealed that the ratio of mature mast cells to total peritoneal mast cells was significantly higher in Ifnar1 −/− mice than in WT mice and that the proportion of immature mast cells was conversely lower in the Ifnar1 −/− mice than in WT mice ( Fig 6A). In addition, significantly higher SSC mean fluorescence intensity (MFI) was detected in both mature and immature Ifnar1 −/− mast cells than in WT mast cells (Fig 6B). These data were consistent with the results shown in Fig 4. Peritoneal mast cell numbers and proliferation ability of Ifnar −/− BMMC/CTMC were not significantly different between WT and Ifnar −/− mice (Fig 1E). Furthermore, as far as we assessed peritoneal mast cell proliferation using ethynyl-deoxyuridine (EdU), we could not detect the proliferative fractions in peritoneal mast cells. Collectively, it is plausible that the increase of mature mast cell proportion in the Ifnar1 −/− peritoneal cavity is primarily caused by IFN-I's effect on mast cell maturation rather than on proliferation, and the absence of IFN-I signaling in mast cells promotes their maturation.

Stat1 and Stat2 are not sufficient to elicit IFN-I's effect on mast cell homeostasis
Both STAT1 and STAT2 play a central role in IFN-I-dependent signaling [6], and therefore Stat1 or Stat2 deficiency should phenocopy the IFNAR1 deficiency in various IFN-I-dependent cellular responses. Unexpectedly, however, both Stat1 −/− and Stat2 −/− mice exhibited comparable frequencies of mature and immature peritoneal mast cells to those of WT mice and did not show a progression of mast cell maturation (Fig 6C and 6D). Furthermore, the severity of hypothermia in the FcεRI-mediated systemic anaphylaxis in Stat1 −/− or Stat2 −/− mice was comparable to that in WT mice (Fig 6E and 6F). These results clearly indicated that Stat1 and Stat2 were not sufficient to mediate IFN-I's functions in mast cells.

Stat3 is required for IFN-I's negative regulation of mast cell functions
Previous studies demonstrated that IFN-I's binding to IFNAR1 induces STAT3 phosphorylation [11][12][13], although the significance of this effect was unclear. As expected, the Stat3 in WT mast cells was tyrosine-phosphorylated by IFN-I treatment (Fig 7A). Neither Stat1 nor Stat3 was phosphorylated by the same treatment in Ifnar1 −/− CTMCs (Fig 7A), indicating that the phosphorylation of these two STATs was indeed IFNAR-dependent. Stat3's tyrosine phosphorylation was still detectable in Stat1 −/− CTMCs and appeared to increase (Fig 7B), consistent with a previous study [33]. Therefore, we tested Stat3's contribution to the IFN-Imediated mast cell regulation using a STAT3 inhibitor, S3I-201. WT mice treated with S3I-201 showed a similar degree of hypothermia as control-treated WT mice (Fig 7C). Interestingly, however, when Stat1 −/− mice were treated with the STAT3 inhibitor, the IgE-mediated systemic anaphylaxis was exacerbated, although the severity of the hypothermia in S3I-201treated Stat1 −/− mice appeared to be milder than that of the Ifnar −/− mice (Fig 7D). Consistent with this observation, S3I-201-treated Stat1 −/− but not WT mice tended to have a higher serum histamine concentration (Fig 7E). In addition, Stat1 −/− CTMCs secreted significantly higher amounts of histamine than WT CTMCs after S3I-201 treatment (Fig 7F). Importantly, the S3I-301-treated Stat1 −/− CTMCs also showed increased Tfeb transcription compared with vehicle-treated Stat1 −/− CTMCs (Fig 7G). These results strongly suggested that Stat1 and Stat3 act redundantly downstream of IFNAR and that the loss of both STAT molecules releases mast cells from the IFN-I-mediated limitations on secretory granule biogenesis and functional maturation.

Discussion
We here investigated the effect of IFN-I on mast cell functions using Ifnar1 −/− mice, and revealed that IFN-I has a critical role in mast cell homeostasis to limit mast cell-dependent anaphylaxis. IFNAR1 deficiency caused an exacerbated systemic anaphylaxis that was associated with increased histamine in the circulation. This outcome was primarily due to the increased storage and exocytosis of secretory granule components, including histamine, in the mast cells. This scenario was supported by the finding that compound 48/80 also induced a more severe systemic anaphylaxis in Ifnar1 −/− than in WT mice, which is known to be a mast cell-dependent but FcεRI-independent response [34]. We currently speculate that mast cells are continuously exposed to IFN-I in their peripheral environment because IFN-I genes were transcribed in mast cells and their niche cells in the steady state. This steady-state IFN-I exposure appears to confine mast cells to less mature states and to limit the onset of anaphylactic responses. Consistent with this scenario, the administration of an anti-IFNAR1 antibody that blocks IFN-I binding also exacerbated systemic anaphylaxis and increased the histamine level in the serum. Although several previous studies have demonstrated the suppressive effect of IFN-I treatment on the FcεRI-mediated histamine release by mast cells [21,22], our current study revealed a novel role of IFN-I in mast cell homeostasis, in particular, a significant contribution of the IFN-I signaling axis to LRO biogenesis. analyzed by flow cytometry; WT versus Stat1 −/− mice (C), and WT versus Stat2 −/− mice (D). Each symbol represents the value from one mouse; horizontal bars show the mean values. n.s., not significant. (E, F) IgE-mediated passive systemic anaphylaxis. WT and Stat1 −/− mice (E) or WT and Stat2 −/− mice (F) were sensitized with anti-TNP IgE followed by challenge with TNP 4 -BSA, and the body temperature was monitored for the indicated periods. n = 5 for each group. There were no statistically significant differences. Results are shown as the mean ± SEM of values representative of three independent experiments. Underlying data from A to F can be found in S1 Data. BSA, bovine serum albumin; c-kit, cellular homolog of the transforming gene of a feline retrovirus (v-Kit); FcεRI, high affinity immunoglobulin E receptor; IFN-I, type I interferon; IgE, immunoglobulin E; MC, mast cell; MFI, mean fluorescence intensity; SSC, side scatter; Stat, signal transduction and activator of transcription; TNP, trinitrophenyl; WT, wild-type. In our experiments, the histamine levels in CTMC were appreciably lower than those in peritoneal mast cells. One possible reason for the difference in histamine levels between these two types of mast cells is that CTMC obtained after 4 days' culture with IL-3/SCF plus feeder cells is still not fully matured. The observation that mature and immature mast cell subsets detected in peritoneal mast cells (Fig 6A) were not detectable in the CTMC population may support this notion. Furthermore, in electron microscopic analyses, we observed obvious differences in the electron density of secretory granules between peritoneal mast cells and CTMC; i.e., the granules in peritoneal mast cells showed high electron density, while those in CTMC showed low electron density. Although in vitro-differentiated CTMC seems to be less mature compared with peritoneal mast cells, the effect of IFN-I on mast cells could be detectable, and therefore it is conceivable that IFN-I has an effect on mast cells before they undergo full maturation.
Our current study also provides the novel and unexpected observations that STAT1 and STAT2 were insufficient to elicit the IFNAR1-mediated regulation of mast cell homeostasis. Instead, we revealed that Stat3 has an important role in the IFN-I-mediated regulation of mast cell homeostasis downstream of IFNAR1. Previous studies reported that in addition to Stat1 and Stat2, IFN-I induces Stat3 phosphorylation [11][12][13], although the significance of the Stat3 phosphorylation downstream of IFNAR1 was unclear. Our findings that Stat1 −/− mice showed normal systemic anaphylaxis and that the administration of a Stat3 inhibitor in a Stat1 −/− but not WT background caused the exacerbation of anaphylaxis strongly suggested that Stat1 and Stat3 function redundantly in this process. It is currently unclear whether Stat1/3's nuclear translocation and transcriptional activity are required for IFN-I's effect on mast cells.
It should be noted that the mast cell TFEB expression was increased both in the absence of IFN-I signaling and under Stat3 inhibition in the Stat1-deficient background. TFEB is a master regulator of lysosome biogenesis [25], and we have reported that this transcription factor regulates the secretory granule biogenesis in mast cells [30]. Therefore, it is possible that Stat1/3 mediate IFN-I's effects in cooperation with TFEB, nonetheless, other transcription factors besides TFEB are possibly involved in this machinery. Although whether and how Stat3 and/ or Stat1 participate in the regulation of TFEB expression are currently unclear, two recent studies demonstrating Stat3's involvement in lysosome homeostasis may provide a clue for future experiments aimed at clarifying this mechanism [35,36]. One study showed that Stat3 associates directly with TFEB in the nucleus and partly suppresses TFEB's function [35]. These observations are not inconsistent with our results, because a decreased Stat3 activation caused by the absence of IFN-I signaling could contribute to the restoration of TFEB activity. Another experiments. (C) WT mice treated with the Stat3 inhibitor S3I-201 (n = 5) or vehicle (n = 5) on day −1 and 2 hours before antigen challenge on day 0 (100 μg/administration), and sensitized with IgE on day −1 were challenged with antigen on day 0. Hypothermia was monitored for 75 minutes. Two groups were not significantly different over all time points. (D) Stat1 −/− mice were treated with S3I-201 (n = 3) or vehicle (n = 4), subjected to antigen challenge, and monitored for hypothermia as described for panel C. �� P < 0.01, � P < 0.05. (E) Serum histamine level upon IgE-mediated passive systemic anaphylaxis in mice pretreated with either S3I-201 or vehicle, as in the left panel of (A), was determined by competitive EIA. Each symbol represents the value from one mouse; horizontal bars and error bars show the mean and SD. Statistical significance was determined by one-way ANOVA with Tukey's test. Results are representative of three independent experiments. (F) Histamine release upon FcεRI ligation from WT or Stat1 −/− CTMCs treated with S3I-201 was determined by EIA. �� P < 0.01, by one-way ANOVA with Tukey's test. Results are shown as the mean ± SD of values representative of three independent experiments. (G) Tfeb expression in WT or Stat1 −/− CTMCs treated with S3I-201 was analyzed by RT-qPCR. � P < 0.05; n.s., not significant. Results are shown as the mean ± SD of values determined from three independent experiments. Original images in A and B can be found in S3 Data. Underlying data from C to G can be found in S1 Data. BMMC, bone marrow-derived mast cell; c-kit, cellular homolog of the transforming gene of a feline retrovirus (v-Kit); CTMC, connective tissue-type mast cell; Ctrl, control; EIA, enzyme immunoassay; ERK, Extracellular signal-regulated kinase; FcεRI, high affinity immunoglobulin E receptor; IFN-I, type I interferon; IgE, immunoglobulin E; pY, phosphotyrosine; RQ, relative quantity; RT-qPCR, quantitative reverse transcription PCR; Stat, signal transduction and activator of transcription; Tfeb, transcription factor EB; WT, wild-type. https://doi.org/10.1371/journal.pbio.3000530.g007 Tonic IFN-I signaling limits mast cell homeostasis and systemic anaphylaxis study reported that lysosomal stress induces Stat3 activation, which regulates lysosome hydrolase expression under the conditions of nutritional sufficiency [36]. This Stat3 function is TFEB-independent and was suggested to be one of the mechanisms for lysosome adaptation to lysosomal stress, such as oxidative stress. Thus, Stat3 could be involved in the regulation and adaptation of lysosomes. However, our current results showing that Stat3 apparently negatively regulates mast cell granule synthesis seem inconsistent with the positive regulatory role of Stat3 for lysosome regulation under stress. Because lysosome biogenesis and adaptation are probably regulated in a context-dependent manner, i.e., under fed or starved, or inflammatory or homeostatic conditions [25], it is conceivable that Stat3, TFEB, and other unknown molecules contribute differentially to these fundamental cellular responses in non-mutually exclusive manners.
A previous study showed that the Stat3 inhibitor C188-9, which blocks the phosphorylation of Stat3's tyrosine 705 residue, efficiently ameliorates mast cell-mediated systemic anaphylaxis through a regulation of vascular permeability, based on a Stat3-transcriptional activity-dependent tight junction stabilization in endothelial cells [37]. These results are apparently inconsistent with our data obtained by treating WT mice with the Stat3 inhibitor S3I-201, in which we did not observe an amelioration of the systemic anaphylaxis. There are several possible reasons for the apparently different behavior of Stat3, such as C188-9's potential inhibition of Stat1 and Stat5 in addition to Stat3, differences in the modes of action between these two Stat3 inhibitors, and different experimental conditions, including the period of Stat3 inhibitor administration. Although the effect of these two Stat3 inhibitors on systemic anaphylaxis appeared to be different, it is interesting that both C188-9 and S3I-201 caused an increase in histamine release in the circulation [37]. Thus, these two independent observations suggest that Stat3 is functionally linked to secretory granule homeostasis.
STAT3's role in mast cells and allergic responses may be helpful for understanding the pathology of autosomal-dominant hyper-IgE syndrome (AD-HIES) [38][39][40]. AD-HIES patients carry a dominant-negative STAT3 mutation such as a missense mutation or in-frame deletion. These patients exhibit elevated serum IgE and frequently develop allergic or nonallergic inflammation. Food allergies and anaphylaxis are reported to be markedly diminished in AD-HIES patients with STAT3 mutations compared with other hyper-IgE patients with no STAT3 mutations [20]. In addition, transgenic mice expressing AD-HIES-associated STAT3 mutations show an amelioration of mast cell-dependent systemic anaphylaxis [37]. It was also demonstrated that STAT3 knock-down in a human mast cell line, LAD2, decreased the cells' FcεRI-triggered degranulation [37]. Although hyper IgE may decrease the contribution of antigen-specific IgE to FcεRI cross-linking, these previous findings suggest that STAT3 positively regulates mast cell function, including degranulation, and support the idea that STAT3 inhibition is beneficial for controlling anaphylaxis [37]. It should be emphasized that our current observations do not contradict these previous findings, because all of the observations concerning AD-HIES reflect the loss of multiple STAT3-dependent signaling axes [41]. In contrast, our interpretation was based on data focusing on the STAT3 activity downstream of IFNAR1. Interestingly, the phosphorylation, nuclear translocation, and target gene expression of the AD-HIES-associated mutant STAT3 molecules are not different from those of WT STAT3 [40]. Therefore, how the mutant STAT3 behaves and causes AD-HIES-associated phenotypes is still largely unclear. Thus, even though there is abundant evidence that STAT3 contributes to mast cell functions and anaphylactic responses in the context of AD-HIES, we still do not fully understand its mechanism.
From a therapeutic point of view, even though mast cells in the steady state were under the control of IFN-I, it was meaningful that the exogenous administration of IFN-I could ameliorate the mast cell-dependent systemic anaphylaxis. Notably, this amelioration was achieved by a relatively short-term treatment with IFN-I, i.e., two pretreatments for two consecutive days before antigen challenge. In addition, the blockade of IFN-I signaling by an anti-IFNAR1 antibody in WT mice or STAT3 inhibition in the Stat1 −/− background significantly exacerbated the onset of systemic anaphylaxis and increased the serum histamine concentration when provided for only 2 days before sensitization. Taken together, it is plausible that IFN-I exerts a rapid effect on mast cells to alter the cells' properties, including granule storage. Such a quick effect of IFN-I may be explained by the rapid mode of TFEB-mediated lysosome regulation [26]. However, the mechanisms of IFN-I action on systemic anaphylaxis need to be interpreted with caution; considering the broad targets of IFN-I in vivo, besides the suppression of LRO biogenesis in mast cells, other possible mechanisms in non-mast cells should be further investigated.
In conclusion, we demonstrated that IFN-I is important for maintaining mast cell homeostasis in the steady state. This study revealed a novel, homeostatic role of IFN-I in LRO biogenesis. Furthermore, we revealed that IFNAR1-mediated Stat3 activation is required for this function of IFN-I. Although our study raises many questions that still need to be answered, it also provides the promising possibility that targeting the machinery of LRO biogenesis or adaptation may represent a novel strategy for controlling mast cell functions.

Ethics statement
All animal experiments were approved by the President of the National Center for Global Health and Medicine (NCGM) with the consideration of the institutional Animal Care and Use Committee (approval No. 18115) and were conducted in accordance with institutional procedures, national guidelines, and the relevant national laws on the protection of animals.

Mast cell preparation
BMMCs were prepared as follows. Bone marrow cells taken from the femur or tibia of WT, Ifnar1 −/− , or Stat1 −/− mice were cultured with IL-3 for 6-8 weeks, and the purity of the BMMCs was confirmed by flow cytometry in which cells were stained with PE-anti-FcεRIα (MAR-1, Tonbo Biosciences, San Diego, CA, 50-5898-U100) and FITC-anti-c-kit (2B8, BioLegend, San Diego, CA, 105806) monoclonal antibodies (mAbs). For in vitro maturation, the BMMCs were cocultured with 3T3 fibroblasts in the presence of IL-3 and SCF [42]. Briefly, the BMMCs were cultured with BALB/c 3T3 fibroblasts in the presence of IL-3 and SCF for 4 days (with medium replacement on day 2), and the floating cells were collected and used as mature BMMCs (namely CTMCs). Peritoneal mast cells were enriched by the following procedure. Cells collected from the peritoneal lavage were stained with PE-anti-FcεRIα and FITC-anti-c-kit antibodies, and the double-positive cells were sorted on a FACSAria II (BD Biosciences, San Jose, CA) cell sorter.
To measure serum histamine levels, the mice were killed 90 seconds after the TNP 4 -BSA challenge, and the serum was immediately isolated from blood obtained by cardiac puncture. The histamine level in mast cell culture medium or serum was determined by a competitive enzyme immunoassay (EIA) (Bertin Pharma, Montigny le Bretonneux, France, #A05890). For the measurement of histamine levels in peritoneal mast cells, 6-8 × 10 4 sorted mast cells were used, and in CTMC, 3 × 10 5 CTMC were used in each experiment. For the compound 48/80induced anaphylaxis model, compound 48/80 (Nacalai tesque, Kyoto, Japan, #09347-94) was i. p. administered into mice at 4 mg/kg, and the rectal temperature of the mice was monitored for 75 minutes using a digital thermometer.

PCA model
For the PCA model, 100 ng of anti-TNP IgE was intradermally administered into the ear pinna using an Ito Microsyringe (Ito corporation, Japan, MS-N50), and the mouse was challenged 24 hours later with an i.v. injection of 200 μL of PBS containing 200 μg of TNP 4 -BSA and 0.5% Evans blue (Tokyo Chemical Industries, Tokyo, Japan, E0197), as an indicator of vascular leakage. Thirty minutes after the antigen challenge, the mice were euthanized, and the Evans blue dye was extracted by incubating the dissected ear in 700 μL of formamide (Nacalai tesque, 16228-05) at 63˚C overnight. The absorbance of Evans blue in the extract was measured with a spectrophotometer at 620 nm.

Degranulation assay and cytokine quantification
To measure degranulation upon FcεRI ligation, CTMCs were incubated overnight with anti-TNP IgE at 0.5 μg/mL (clone IGELb4) in medium containing IL-3. The cells were washed twice with prewarmed Tyrode's buffer (10 mM HEPES [pH 7.4], 130 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , 1.4 mM CaCl 2 , 5.6 mM D-glucose, and 0.1% BSA) and stimulated with TNP 4 -BSA in Tyrode's buffer for the indicated periods. The culture supernatant was incubated with 4-MUAG (final concentration 0.5 mM) in citrate-phosphate buffer (pH 4.5; 52 mM citric acid, 94 mM Na 2 HPO 4 ) for 15 minutes at 37˚C, followed by the addition of Stop solution (2 M Na 2 CO 3 , 1.1 M glycine) to end the reaction. Fluorescence (excitation, 365 nm; emission, 450 nm) was measured on a microplate reader (VarioScan Flash, Thermo Fisher). The degranulation rate was calculated by dividing the β-hexosaminidase activity in the culture supernatant by that in the cell lysate, which was the supernatant of cells lysed with Tyrode's buffer containing 0.5% Triton-X 100. Cytokine levels in the BMMC culture supernatants were measured using the Mouse ELISA Max Kit (BioLegend Mouse IL-6 ELISA MAX, 431302) or the Ready-Set-Go ELISA Kit (Thermo Fisher, 88-7324-88 for TNFα, 88-7137-88 for IL-13).

Immunoblot analyses
The whole-cell lysate of BMMCs was subjected to SDS-PAGE, and the separated proteins were transferred to a membrane. The blots were then incubated with the target antibody, followed by the appropriate HRP-conjugated secondary antibody, and visualized by the LAS-3000 detection system (Fuji Photo Film, Japan). The following antibodies were used for immuno-

Gene expression assays
The RNA from mast cells was extracted with Isogen (Nippon Gene, Japan, 311-02501) and reverse-transcribed using a SuperScript III cDNA Synthesis Kit (Thermo Fisher 11752050); otherwise, the CellAmp Direct Lysis and RT set (Takara, Japan, 3737S) was used. Quantitative RT-PCR was conducted using a StepOnePlus Real-Time PCR System (Thermo Fisher) with TaqMan probes for mouse Il6, Ifna4, Ifna5, Ifnb1, or Hprt (Thermo Fisher, Mm00446190_m1 for Il6, Mm00833969_s1 for Ifna4, Mm00833976_s1 for Ifna5, Mm00439552_s1 for Ifnb1, and Mm03024075_m1 for Hprt). The expression of Hdc, a gene involved in histamine synthesis, was quantified with the THUNDERBIRD SYBR qPCR mix (Toyobo, Japan, QPS-201) on a StepOnePlus Real-Time PCR System. Gene expressions were normalized to that of Hprt. The primers used in this study are listed in S1 Table.

Flow cytometry
Single-cell suspensions were prepared from the peritoneal lavage, the spleen, or the bone marrow of mice, and the cells were stained with the indicated mAbs and analyzed by flow cytometry using a BD FACSVerse or BD FACSCalibur (BD Biosciences). The following antibodies were used for flow cytometry: PE-or APC-conjugated anti-FcεRI (Thermo Fisher 17-5898-82), Alexa Fluor647-(BioLegend 105818) or FITC-conjugated anti-c-kit (BioLegend 105806), and PE-anti-mouse IFNAR1 (BioLegend 127311). To detect the degranulation of mast cells after FcεRI ligation, APC-anti-mouse CD63 (Thermo Fisher 17-0631-80) was used. For the in vivo IgE-binding assay, 10 μg of the IGELb4 IgE antibody labeled with Alexa Fluor 647 using the Alexa Fluor 647 Microscale Protein Labeling Kit (Thermo Fisher A30009) was i.v. administered into mice, and 24 hours later, the mast cells in the peritoneal lavage were analyzed on a FACSVerse.

Electron microscopy
Peritoneal MCs were separated as c-kit + , FcεRIα + cells on FACSAria II and fixed in 2% glutaraldehyde in 0.1 M HEPES for overnight, attached on a poly-L-Lysine-coated plastic dish, postfixed with 1% OsO 4 for 60 minutes at RT, followed by dehydration in ethanol, and finally embedded in epoxy resin. Sections were analyzed by electron microscope (JEM-1400; JEOL, Japan). Sizes of the intracellular granules were determined by measuring the length of the granules on TEM images using ImageJ software.

Statistics
The statistical significance of differences in the mean ± SD of various groups was calculated with a two-tailed t test unless otherwise noted. For the comparison between more than two groups, Tukey's test was applied as post hoc analyses in ANOVA. P values less than 0.05 were considered statistically significant. All animal data are presented as the mean ± SEM of at least 3 independent experiments.