BALB/c mice were purchased from The Jackson Laboratory and Envigo and used as WT controls in all experiments. IL-10^-/- mice on the BALB/c background were purchased from The Jackson Laboratory. ST2^-/- mice on the BALB/c background are a kind gift of Dr. Andrew McKenzie, Medical Research Council, United Kingdom and Drs. Paul Bryce and Gurjit Khurana Hershey. IL-10^-/- and ST2^-/- mice were crossed to generate ST2^-/-/IL-10^-/- mice. All mice were bred in our facilities and all animal research was performed as approved by the IACUCs at the respective institutions, Western New England University (protocol no. 2019-S1) and the University of Connecticut (protocol no. A22-048).

BMMCs were generated from naïve WT BALB/c and ST2^-/- animals as previously described (14). Briefly, bone marrow cells were collected from the tibia and femurs of animals and cultured with 10 ng/ml of rIL-3 and rSCF (Shenandoah) for >4 weeks. Harvested BMMCs were positive for c-Kit and FcϵRI.

1 million BMMCs/ml were cultured in triplicate with 10 ng/ml IL-3 and SCF. Cells were activated by pre-sensitizing with 1 µg/ml DNP-IgE (clone SPE7, Sigma) or vehicle (medium), followed by treatment with 200 ng/ml DNP-BSA (14, 43). Some groups of cells were treated with 20 ng/ml of rIL-10 and/or rIL-33 (Biolegend) for various time periods (including 6 hours and 24hrs) prior to challenge with DNP-BSA. Thirty minutes to an hour after activation with DNP-BSA, cells were collected for isolation of RNA and cDNA was created. The cDNA was then used to assess the expression of various cytokine genes as described in the manuscript. In other experiments, supernatants were collected 6-24h later for the assessment of secreted cytokines by ELISA.

Quantitative RT-PCR was performed as previously described using Taqman probes (14, 43). The expression of cytokine genes (IL-4, IL-5, IL-13, IL-10, IL-33, IFN-γ) was calculated relative to GAPDH transcripts. ELISAs for mMCP-1 (Thermofisher), IL-4, IL-5, IL-6, TNF-α and IFN-γ (Biolegend), IL-13 (R&D Systems), and OVA-IgE were performed according to manufacturers’ protocols as previously described (14, 43).

BMMCs were grown in rIL-3 and rSCF as described above. Some groups of cells were treated with 20 ng/ml of rIL-10. Cells were counted daily for 1-3 days and live cells were enumerated on the basis of trypan blue exclusion or using a quantitative tetrazolium reduction cell proliferation assay (MTS assay kit by Abcam). Cell proliferation was calculated based on OD using the formula: (OD of samples – OD of untreated control)/(OD of untreated control × 100).

BMMCs were cultured in the presence or absence of 20 ng/ml rIL-10 for 24 hours. Cells were activated with IgE and antigen and β-hex activity was assessed as previously described (44).

Cultured BMMCs were resuspended in staining medium (SM) containing 1X HBSS, HEPES buffer, and 2% fetal calf serum and incubated with mAbs against mouse c-Kit, FcεRI, and ST2 (Biolegend). Stained cells were then washed and assessed phenotypically by flow cytometry as previously described (18).

To induce food allergy, WT, IL-10^-/-, ST2^-/- and IL-10/ST2^-/- mice were i.p. immunized with 50 μg chicken egg OVA in 1 mg alum twice (two weeks apart), as previously described (14, 43, 45). Four weeks later, mice were challenged i.g. with 50 mg OVA on 6 alternating days. Control animals were i.p. sensitized but not challenged with OVA. To supplement knockout data, some groups of mice were also treated with blocking antibodies for IL-10. In these experiments, mice were treated i.p. with 100 μg purified anti-IL-10 (Biolegend) 6 different times immediately prior to OVA challenges. Mice were sacrificed one hour after the 6^th challenge with OVA, and food allergy parameters were assessed as previously described (14, 43, 46). Blood was collected for evaluation of antibodies and mMCP-1 in serum. Jejunum was collected for histological assessment of MCs and evaluation of cytokine gene expression by RT-PCR as described above.

Intestinal anaphylaxis was assessed in challenged mice by scoring the percentage of animals exhibiting allergic diarrhea for one hour after OVA challenge (14, 46).

Intestinal MCs were enumerated as we have previously described (14). Briefly, paraffin-embedded jejunal sections were stained with chloroacetate esterase (CAE) and MCs were counted in complete cross-sections. Data are represented as the average numbers of MCs in 3 high-powered fields (HPF).

WT and ST2^-/- were sensitized i.v. with 6 μg DNP-IgE (clone SPE7, Sigma). 24h later, they were challenged i.v. with 75 μg of DNP-BSA, and changes in core body temperature were recorded using subcutaneously placed transponders (Biomedic Data Systems). To assess the effects of anti-IL-10 on the development of passive anaphylaxis, some mice were injected i.p. with 400 μg rIL-10 concurrently with the DNP-IgE injection.

Data are expressed as mean plus or minus standard error of mean, unless stated otherwise. Statistical significance comparing two groups of mice was determined using the unpaired or paired Student’s t-test as appropriate. Two-way analysis of variance was used to calculate differences between multiple groups.

We have previously demonstrated that IL-10 can enhance FcεRI expression and promote IgE-mediated activation in MCs (14). Similarly, we recently observed that IL-10 can also significantly enhance the expression of the IL-33 receptor, ST2, and promote IL-33-induced type 2 cytokine production (18). As shown in Figure 1A , treatment with IL-10 enhanced ST2 expression on WT BMMCs ( Figure 1A ). This was further increased in cells that were cultured with IL-10 and activated with IgE/Ag, suggesting that IL-10 can enhance IL-33 responsiveness during IgE-mediated activation ( Figure 1A ).

IL-33 signaling also plays a critical role in the development of allergic responses and endogenously produced IL-33 has been shown to regulate IgE-mediated MC activation (47, 48). To therefore further evaluate the effects of IL-10, we assessed whether IL-10 can also enhance IL-33 production in IgE/Ag-activated cells. As observed in Figure 1B (and data not shown), while we could not detect any IL-33 protein secretion, IL-10 significantly enhanced the transcriptional levels of IL-33 in IgE-activated cells. We have previously shown that IL-10 can also enhance the production of cytokines such as IL-6 and IL-13 in IL-33-stimulated MCs (18). We therefore wondered whether IL-10 may have similar effects on IL-33 production in these cells. Interestingly, as observed in Figure 1C , IL-10 pre-treatment significantly enhanced IL-33 protein secretion in IL-33-treated BMMCs. Similarly, increased secretion of the cytokines IL-6 and IL-13 ( Figures 1D, E ) was also observed, suggesting that crosstalk between IL-10 and IL-33 may serve to further potentiate MC responses.

To further investigate whether IL-33 signaling is necessary for IL-10’s effects, we next assessed cytokine production in unactivated and IgE-activated ST2^-/- BMMCs. As previously observed (14, 18) and shown in Figures 2A-C , IL-10 significantly enhanced the production of IL-6 and IL-13 in both unactivated and IgE-activated WT BMMCs. Similarly, IL-10 pre-treatment also enhanced IL-6 and IL-13 production in unactivated ST2^-/- BMMCs, suggesting that IL-33 signaling is not required for its effects. In general, ST2^-/- BMMCs exhibited reduced cytokine responses after IgE-mediated activation ( Figures 2A-C ), although some variability was observed depending on experimental conditions ( Supplementary Figures S1A-C ). Pre-treatment with IL-10 enhanced the production of both IL-6 and IL-13 but not TNF-α in IgE-activated ST2^-/- cells. A similar pattern was also observed after long-term culture (three days) with rIL-10 as we have previously shown ( Supplementary Figures S1A-C ). Taken together, these data suggest that IL-10 may further modulate the function of MCs in vivo, and that its effects on IgE-activated MCs are independent of IL-33 signaling. Next, we also assessed the effects of IL-10 on MC degranulation and murine MC protease (mMCP)-1 secretion. As observed in Figure 2D , both WT and ST2^-/- BMMCs exhibited comparable levels of β-hex activity in response to IgE activation. IL-10 pre-treatment enhanced β-hex release in both cell types and this was higher in ST2^-/- BMMCs compared to their WT counterparts. This suggests that while IL-10 can promote MC degranulation independently of IL-33 signaling, endogenous IL-33 may regulate some of its effects. Similarly, as previously shown by us (18), IL-10 also enhanced mMCP-1 secretion in both WT and ST2^-/- resting BMMCs ( Figure 2E ). However, lower mMCP-1 levels were observed in IL-10-treated ST2^-/- BMMCs compared to WT cells. Interestingly, no mMCP-1 secretion was observed in either WT or ST2^-/- BMMCs after activation with IgE and antigen ( Figure 2E ). Instead, IL-10 stimulation led to mMCP-1 secretion in these cells as well ( Figure 2E ). Finally, we have previously shown that IL-10 can also promote MC proliferation during cell culture (14, 18). To assess whether IL-33 signaling may be required in this process, we next also assessed the proliferation of both WT and ST2^-/- BMMCs. As observed in Figure 2F , IL-10 enhanced the proliferation of BMMCs from both strains. Furthermore, greater proliferation was observed in ST2^-/- BMMCs treated with IL-10. While the mechanism for this is unclear, these data further underscore the role of IL-10 and suggest that IL-33 signaling is not required for its effects.

Recently, IL-33 has emerged as a critical mediator of MC responses during food allergy (8, 26, 49–51). To further investigate the crosstalk between IL-10 and IL-33, and whether IL-10’s effects on IgE-induced MC responses may depend on IL-33, we assessed the development of food allergy in WT and IL-10-depleted ST2^-/- mice using an ovalbumin (OVA)-induced model of intestinal anaphylaxis. We and others have previously shown that the development of food allergy in this model is IgE and MC-dependent (9, 14, 43, 46, 52). Similarly, we recently also demonstrated a critical role for IL-33 in inducing MC responses in this model (26). Lastly, using both IL-10^-/- mice (14) as well as pharmacological blockade of IL-10 (17), we have shown that MC responses and the development of intestinal anaphylaxis in this model are also IL-10-dependent. We therefore hypothesized that this would be a good system to assess the roles of IL-10 and IL-33 and their interrelated effects on MCs.

Briefly, WT and ST2^-/- mice were i.p. sensitized with chicken egg OVA and alum as previously described and subsequently challenged orally with OVA to induce the development of intestinal anaphylaxis. As demonstrated in Figure 3A and Supplementary Figure S2A , WT BALB/c mice developed profuse diarrhea after the sixth oral gavage, accompanied by a robust OVA-specific IgE-mediated response ( Figure 3B , Supplementary Figure S2B ). Histological analysis using chloroacetate esterase staining revealed a significant recruitment of mature degranulating MCs to the small intestine of WT OVA mice compared to unchallenged controls ( Figure 3C , Supplementary Figure S2C ). Furthermore, assessment of MC activation in WT mice revealed the presence of elevated levels of serum mMCP-1, a marker correlated with the degranulation of mucosal MCs ( Figure 3D , Supplementary Figure S2D ). In contrast to these positive markers of food allergy in WT mice and as previously observed by us (26), ST2^-/- animals did not develop allergic diarrhea or exhibit MC-mediated activation, suggesting that the IgE and MC-dependent effects of food allergy require IL-33 signaling ( Figures 3A-D ).

To next examine the effects of IL-10 depletion, some groups of mice were treated with anti-IL-10 during the acute, MC-dependent, challenge phase of the model as described above. As expected and consistent with our previous findings (17), anti-IL-10 treatment in WT mice significantly attenuated MC responses including decreased allergic diarrhea ( Figure 3A ), OVA-specific IgE production ( Figure 3B ), intestinal MC numbers ( Figure 3C ) and mMCP-1 levels ( Figure 3D ). Interestingly, however, treatment with anti-IL-10 had no additional effects in ST2^-/- mice, including changes in allergic diarrhea, OVA-IgE or MC numbers, or intestinal cytokine expression in ST2^-/- mice ( Figures 3A-J ). mMCP-1 levels were below the limits of detection in ST2^-/- animals, except in the case of one mouse ( Figure 3D ). These data suggest that IL-33 signaling is not required for IL-10’s effects on MCs in the food allergy model.

To next assess the effects of IL-10 on intestinal type 2 cytokine expression, we examined jejunal tissue from experimental animals for various cytokine transcripts. As we have previously reported, while the expression of IL-4, IL-5, IL-13, and IL-10 was increased in the jejunae of allergic WT mice, the induction of these cytokines was significantly decreased in both ST2^-/- and anti-IL-10-treated WT mice ( Figures 3E-H ). In contrast, no significant differences were observed between control and anti-IL-10-treated ST2^-/- mice. Furthermore, no significant differences were observed in the expression of IFN-γ or IL-33 between any of the groups ( Figures 3I, J ).

We have previously shown that ST2^-/- mice exhibit reduced IL-4 responses during food allergy which may affect the development of antigen-specific IgE (26). Similarly, IL-10 has also been shown to play an important role in the development of IgE production (53–55). To therefore ascertain whether the effects of anti-IL-10 in our system may be related to reduced IgE responses as opposed to functional defects in MCs, we next assessed the development of IgE-mediated passive anaphylaxis in anti-IL-10-treated naïve WT and ST2^-/- mice. As observed in Supplementary Figure S3 , IgE-sensitized WT mice exhibited significant drops in core body temperature after intravenous antigen administration. In contrast, the induction of hypothermia was delayed in ST2^-/- mice. Anti-IL-10 treatment attenuated the development of passive anaphylaxis in WT mice, suggesting that basal IL-10 levels may prime MC responsiveness to IgE-mediated activation. This is also consistent with our previous observation demonstrating that exogenous IL-10 priming can enhance IgE-mediated passive anaphylaxis (14). However, similar effects were not observed in ST2^-/- mice. While anti-IL-10 treatment initially had no effect in these mice, a more sustained hypothermic response was observed in anti-IL-10-treated ST2^-/- mice, suggesting that other factors may also be involved in these animals, including effects on basophils and other cell types. Taken together, these data suggest that the reduced MC responses in both ST2^-/- mice and anti-IL-10-treated WT mice during food allergy occur irrespectively of IgE levels and may result directly as a consequence of functional deficiencies in MCs.

Considering that ST2^-/- mice exhibit profoundly attenuated MC responses during food allergy and to rule out any variability that may be associated with the pharmacological depletion of IL-10, we next generated mice with a genetic deficiency in both IL-10 and ST2, and examined their susceptibility to the development of food allergy. As observed in Figure 4 , neither ST2^-/- nor ST2^-/-/IL-10^-/- OVA-sensitized mice developed diarrhea in response to OVA challenges ( Figure 4A ). As expected, OVA-IgE levels ( Figure 4B ) and MC responses including intestinal MC numbers and mMCP-1 levels were decreased in the absence of IL-33 signaling ( Figures 4C, D ). To our surprise, however, genetic deletion of IL-10 in these animals further decreased the levels of OVA-IgE and mMCP-1 levels ( Figures 4B, D ). This was also accompanied by a decrease in the number of intestinal MCs ( Figure 4C ). These data suggest that mice with an intrinsic deficiency in both ST2 and IL-10 are further protected from food allergy development and that IL-10 can regulate the IgE-mediated activation of MCs even in the absence of IL-33 signaling.