The Pex2 Mutant Mouse Strain used was 129S6.129-Pex2 ^tm1Plf /Mmmh(Null allele) (Faust and Hatten, 1997) and was obtained from the Mutant Mouse Resource and Research Centre (MMRRC) supported by the NIH. The mice used for this experiment were Pex2 ^+/+ , Pex2 ^−/− , Pex2 ^+/− . Homozygous null mutant strains showed no Pex2 transcript and protein. Homozygous mutants in this congenic strain show variable embryonic lethality, starting at ∼E11. Approximately 20% of homozygotes survive to birth but are hypotonic, do not feed and die on the day of birth. Homozygous mutants that survive in the postnatal period are obtained by mating congenic 129S6.129-Pex2 ^tm1Plf +/- mice with wild-type Swiss Webster strain mice. F1-Pxmp3 ^tm1Plf +/- hybrids (designated Sw129) are then intercrossed to obtain Sw129-Pxmp3 ^tm1Plf −/− (indicated in the text as Pex2 ^−/− ) mice.

Colonies were maintained as stable inbred lines in the Swiss Webster and 129SVEV background under approved animal protocol 21-023, abiding by the standards of the Canadian Council on Animal Care.

BMMCs were generated from SWR/J and Pex2 ^−/− mice according to the method of (Tertian et al., 1981). After at least 4 weeks of culture, the purity of mast cells was evaluated based on the expression of the high-affinity IgE receptor, also known as FcεRI and tyrosine-protein kinase cKIT (Cluster of differentiation, CD117). Cells were used at >98% of purity and consistently contained metachromatic granules.

Total RNA was extracted using the RNeasy Plus Mini Kit (Qiagen, Mississauga, Canada). Genomic DNA was depleted, and complementary DNA was amplified using the Platinum Taq Reverse Transcription Kit (Wisent). Pex2 gene was amplified using HiFi Platinum Taq DNA kit (Thermofisher) and the following primers forward (5′-TGA​AGG​AAC​CAC​TTA​GAA​ATT​ACA​GA) and reverse (5′-CCA​GGG​CCT​TAT​TCA​GTT​CA). Samples were loaded onto a 2.5% agarose gel (with ethidium bromide) in TAE and imaged using chemiDoc imaging system (Biorad).

Cytospins of mast cells were briefly fixed in Carnoy’s fixative then rinsed in water and 0.033N HCl. Cells were then stained with Toluidine blue (pH 0.3) overnight then rinsed before drying and mounting in DPX (Sigma) for imaging with Mantra 2TM at ×40 magnification.

BMMCs (2 × 10⁶/ml) in modified HEPES-Tyrode’s buffer were treated for 15 min with increasing doses of TNP-BSA (Trinitrophenyated-Bovine serum albumin) (Bioresearch Technologies) or calcium ionophore A23187 (Sigma) as a positive control. The level of degranulation was assessed via β-hexosaminidase release according to the method of Schwartz et al. (Schwartz et al., 1979). The percentage of β-hexosaminidase release was calculated as follow: %   o f   r e l e a s e   = { ( O . D   s u p e r n a t a n t − O . D   c o n t r o l ) ÷ [ ( O . D   s u p e r n a t a n t − O . D   c o n t r o l ) + ( O . D   p e l l e t − O . D   c o n t r o l ) ] } X 100

Prior to all activations, BMMC were ‘‘rested’’ overnight in modified mast cell growth medium, with 3 ng/ml mIL-3 (Peprotech) or without PGE₂ (Tocris). For analysis of cytokine production, cells were washed twice and resuspended in medium consisting of RPMI 1640 with 1% FBS, 15 mM HEPES and 3 ng/ml rmIL-3 and 100 μg/ml of soybean trypsin inhibitor. For all activations, cells at 1 × 10⁶/ml were incubated with Pam3-CSK4-KKKK (L2000, EMC microcollection) at 50 μg/ml or LPS (Sigma) at 50 μg/ml and A23187 (Sigma) at 0.5 µM for 24 h at 37°C. For IgE activation, BMMCs were sensitized with anti-TNP (Trinitrophenol phosphate) IgE overnight. Cells were then rinsed and treated with 10 ng/ml of TNP-BSA for 30 min, then supernatants were removed, and cells were cultured for a further 24 h. For mechanistic studies, cells were treated with 100 μg/ml of niacin (Sigma Aldrich) for 48 h or with 2.5 uM thioridazine for 1 h prior to IgE activation. Supernatants were removed, and cells were cultured for a further 24 h. Cell-free supernatants were collected and assayed for IL-6 (Peprotech), IL-13 (Peprotech) and TNF (Invitrogen) by ELISA from sources indicated.

One million BMMCs per genotype and under each condition was sensitized with anti-TNP as described above and treated with 10 ng/ml of TNP-BSA for 24 h. Supernatants were removed, and cells were analyzed for fatty acid accumulation using free fatty acid quantification kit (Sigma) according to manufacturer’s recommendations.

ELISA: Levels of IL-6 (Peprotech), IL-13 (Peprotech) and TNF (Invitrogen) in supernatants were assessed according to the manufacturer recommendations.

Cells were fixed in 4% paraformaldehyde in PBS for 30 min and then incubated for 1 h at room temperature in 5% normal goat serum (Sigma) and for 16 h at 4°C with primary antibody at 1:100 dilution in 5% normal goat serum. Appropriate Alexa Fluor secondary antibodies (anti-rabbit secondary antibodies, were from Abcam) were then used at 1:1000 dilution in 5% normal goat serum. After 4 washes in PBST (PBS +0.1% (v/v) Triton X- 100), cells were mounted in DAPI Pro-Gold Antifade Reagent (Thermo Fisher) and imaged using a ×100 oil immersion objective (NA = 1.4) mounted on an Zeiss800 confocal microscope (Zeiss) or using a Zeiss AxioObserver LSM 880, 100 × 1.4 oil plan-Apochromat lens. Primary antibody was rabbit anti-SKL antibody was previously described (Szilard et al., 1995).

Flow cytometry: Maturation of Pex2 ^−/− and WT BMMC were assessed by flow cytometry. Antibodies to CD117 (Clone 2B8, Biolegend), FceR1(Clone MAR-1, Invitrogen) were used to assess the maturation of BMMC cells. Fc receptors were first blocked with anti CD16/CD32 (Clone 93, eBisocience) for 10 min. Combinations of the two anti-CD117 and anti-FceR1 fluorescently tagged antibodies (at manufacturers recommended dilutions) were added to each sample for 30 min at 4°C. For assessment of IgE binding, BMMCs were sensitized with anti-TNP (Trinitrophenol phosphate) IgE overnight. Cells were then rinsed then stained with fluorescently tagged anti-IgE antibody (ClonePME-1, Biolegend) or with isotype control rat IgG2b (Clone G0114F7, Biolegend) for 30 min at 4°C.

After stainings, cells were washed twice with PBS supplemented with 2% Fetal Calf Serum (FCS) (Gibco) then fixed with PBS containing 1% paraformaldehyde for 30 min 4°C before analysis on BD FACSCelesta™ (BD). Data were analyzed using FlowJo Version 10 software (BD).

BMMCs were resuspended at a density of 1 million cells per mL in activating media and seeded into a 24-well plate. BMMCs were stimulated in duplicates with either activation media, LPS (50 μg/ml), Pam3CSK4 (50 μg/ml), or thiorizi dine (2.5 uM) for 24 h. BMMCs were then washed in PBS prior to staining with fixable viability dye Efluor 450 for 20 min at 4°C then rinsed before fixation in 1% paraformaldehyde, and acquired on the FACS Canto II flow cytometer. Data were analyzed using FlowJo Version 10 software (BD).

Statistical analyses were performed using a non-parametric t-test comparing between Pex2 ^−/− and WT BMMCs. * p < 0.05, ** p < 0.001, *** p < 0.0001 ns: not significant. All results are represented as the mean of at least 3 independent experiment ± Standard error of the mean.

Average number of puncta per cell were calculated using ImageJ software, applying the following steps to each image:

1—We opened stacks image.

File -> Open…

2—We filtered to remove noise.

Process -> Filters -> Gaussian Blur…

3—We subtracted background.

Process -> Subtract Background…

(the box marked “Light Background” was unticked).

4—We clicked on “Image”

Color -> Split the channels …

In this step, brightness and contrast were adjusted, and setting were applied to all stacks.

5—We performed threshold image.

Image -> Adjust -> Threshold…

(We selected: Apply it to all stacks).

Box labeled “Dark Background” was ticked. We adjusted the sliders so that features were red colored, but the rest of the image was not. Then we clicked “Apply” button. This replaced grayscale image with an “8-bit binary image.” All “red” pixels were converted to a value of “255,” while all non-red pixels were given a value of “0.”

6—We filled in any holes in the nuclei.

Process -> Binary -> Fill Holes

Process -> Binary -> Watershed.

Process -> Find Edges.

8—We performed the analysis.

Analyze -> Analyze Particles…

In this dialog box the algorithm started to include or exclude puncta based on their attributes. “Size” smaller than 0.1 mm and larger than 1 mm “Circularity” set range: 200-1.

To study the requirement of peroxisomes for MCs function, we assayed BMMCs from wildtype and Peroxin2 null mutant mice (Pex2 ^−/− ) (Supplementray Figure S1A) (Faust and Hatten, 1997). We confirmed the presence of peroxisomes by performing indirect immunofluorescence (IF) using an antibody against the C-terminal Peroxisome Targeting Sequence Type 1 Ser-Lys-Leu (SKL), the canonical marker for peroxisomal matrix proteins (Szilard et al., 1995).

WT MCs showed SKL-positive puncta while intense diffuse staining was observed in Pex2 ^−/− MCs confirming that the MCs from Pex2 mutant mouse do not form functional peroxisomes due to defects in peroxisomal protein import into the matrix (Smith and Aitchison, 2013) (Figures 1A,B, Supplementary Figure S1A; Supplementary Data Sheet S1, S2). Assessment of maturation based on the expression of FceR1 and CD117 by flow cytometry showed no differences between the two cell types (Supplementary Figure S1B). Pex2 ^−/− BMMCs exhibited a similar morphology and granularity compared to the WT BMMCs (Supplementary Figure S1C). All together these observations indicated that absence of peroxisomes did not alter maturation, morphology, or granule content of MCs.

Through IgE-mediated degranulation, MCs hold a key role in allergic disease and host defence against several parasites. Peroxisomes are known to proliferate and increase during responses to viral infection (Cook et al., 2019; Knoblach et al., 2021). To assess the role of IgE-mediated activation and peroxisomes we measured peroxisome numbers in WT BMMCs when stimulated with IgE. Indirect IF followed by automated quantification showed an increase in SKL-positive puncta in stimulated MCs (Figures 1C,D, Supplementary Data Sheet S3, S4) indicating that an increase in cellular peroxisomes occurs during the IgE mediated MC response. We next assessed whether an absence of peroxisomes altered IgE/antigen-induced degranulation, assessed via a β-hexosaminidase release. Peroxisome biogenesis defects caused by mutations in Pex2 have been linked to lipid metabolic defects (Faust and Hatten, 1997; Faust, 2003) thus, affecting the lipid milieu of the cell membrane (Schrader et al., 2020). In fact, an altered membrane lipids environment was reported to compromise signaling (Lodhi et al., 2015) in multiple cell types including immune cells (Lodhi et al., 2015; Di Cara et al., 2019). Thus, we first assessed the ability of WT and Pex2 ^−/− BMMCs to bind IgE by flow cytometry. WT and Pex2 ^−/− BMMCs demonstrated equivalent IgE binding (Figure 1E and Supplementary Figure S1D). Next, BMMCs were loaded with Anti-TNP IgE overnight then crosslinked with a dose range of TNP-BSA antigen or treated with calcium ionophore A23187 allowing degranulation for 15 min. Pex2 ^−/− BMMCs displayed similar degranulation potential compared to their WT counterparts (Figure 1F). When treated with calcium ionophore A23187, Pex2 ^−/− BMMCs demonstrated a higher percent degranulation at lower A23187 doses (Figure 1G). Overall, lack of functional peroxisomes did not adversely affect mast cell degranulation.

It is well established that MCs have degranulation-independent pathways which allow the production of cytokines and chemokines independent of classical degranulation (Leal-Berumen et al., 1994; Moon et al., 2014). Beyond their role in allergy, MCs are key player in responses to pathogens (Marshall, 2004; Abraham and St John, 2010). MCs express and respond via Toll like receptors (TLRs) to several bacterial or viral products (McCurdy et al., 2003; Agier et al., 2018). We thus asked whether peroxisome function is necessary for cytokine release upon TLR2 or TLR4 stimulation induced by Pam3CSK4 (Pam3) and E.coli lipopolysaccharide (LPS) respectively. We assessed whether the lack of functional peroxisomes affected TLR and IgE-mediated interleukin 6 (IL-6), interleukin 13 (IL-13) and Tumor Necrosis Factor (TNF) production after 24 h stimulation. Our data showed that IL-6 (Figure 2A) and IL-13 (Figure 2B) were produced in response to TLR2 and TLR4 stimulation in WT BMMCs but production of both cytokines was significantly reduced in Pex2 ^−/− cells. On the other hand, TNF production was significantly reduced only upon LPS treatment in Pex2 ^−/− compared to WT MCs (Figure 2C). Interestingly, while IgE-mediated degranulation was not affected by the absence of functional peroxisomes, cytokine production was markedly reduced in Pex2 ^−/− BMMCs upon IgE stimulation followed by TNP-BSA treatment (Figures 2D–F). Additionally, IL-13 secretion was lower in Pex2 ^−/− BMMCs after IgE stimulation followed by treatment with both TNP-BSA or with the calcium ionophore A231287 (Figure 2E). We tested the secretion of other cytokines such as interleukin 5 (IL-5), Granulocyte-macrophage colony-stimulating factor (GM-CSF) and Chemokine (C-C motif) ligand 3 (CCL3) (Supplementary Figure S2A–C). A23187 selectively induced CCL3 production which was significantly decreased in Pex2 ^−/− compared to the WT BMMCs (Supplementary Figure S2C). All together these results indicated a requirement for peroxisomes in MCs activation in response to IgE-mediated, TLR2 or TLR4 stimulation.

Peroxisomes are highly conserved organelles and play a pivotal role in lipid metabolism and ROS such as hydrogen peroxide (H₂O₂) catabolism (Liu et al., 2019). Both ROS and lipids are important mediators in cellular signaling in immune cells (Di Cara, 2020; Di Cara et al., 2017; 2019). H₂O₂ is a permeable and diffusible molecule involved in inter- and intracellular signaling during host defense (Bedard and Krause, 2007; Blander and Sander, 2012). We measured the cellular amount of H₂O₂ in Pex2 ^−/− and WT BMMCs. At rest, BMMCs lacking peroxisome function exhibited similar amounts of H₂O₂ (Figure 3A) as WT BMMCs.

Different studies have shown that the metabolism of fatty acids (FAs) is a major source of biological lipids that form cell membranes and regulate inflammatory functions (Dowds et al., 2014; Hubler and Kennedy, 2016; Nath et al., 2022; O'Neill et al., 2016; Puertollano et al., 2001; Sadik and Luster, 2012). Peroxisomes contribute to the homeostasis of FAs in the cell (Wanders and Waterham, 2006; Lodhi and Semenkovich, 2014) and we probed whether free fatty acids (FFAs) are altered in MCs in absence of peroxisomes, affecting cytokine release. We measured cellular FFAs and observed a significant accumulation of FFAs in Pex2 ^−/− BMMCs compared to WT BMMCs, at rest (Figure 3B). Interestingly, we observed that IgE stimulation triggers a significant increase of FFAs in WT BMMCs while the level remained unchanged in IgE stimulated Pex2 ^−/− (Figure 3B). These results indicated that IgE-mediated stimulation triggers an increase in cellular FFAs in MCs while lack of functional peroxisomes affects FFA metabolism and turnover at rest and during IgE-mediated activation.

To explore the link between FFAs and MCs activation, we used Niacin a vitamin B3 shown to reduce FFAs in plasma, macrophages and adipocytes by inducing anti-lipolytic effects (Tunaru et al., 2003; Nath et al., 2022). Human MCs have been shown to respond to niacin treatment by prostaglandin D2 (PGD₂) production (Papaliodis et al., 2008). We hypothesised that the accumulation of FFAs in WT BMMCs was hindering cytokine release under IgE-mediated stimulation conditions. To test this hypothesis, we treated Pex2 ^−/− BMMCs with niacin for 48 h to reduce FFAs (Figure 3B) and then we stimulated the cells. Treatments with niacin reduced FFAs in Pex2 ^−/− BMMCs at rest to the amount observed in WT. Moreover, after IgE stimulation, niacin treatment recapitulated the increase FFAs observed in IgE stimulated WT BMMC (Figure 3B). Intriguingly, the amounts of released IL-6 and IL-13 after IgE-mediated activation was rescued to WT levels in Pex2 ^−/− BMMCs upon 48 h treatment with niacin (Figures 3C,D). This result provides a link between the cellular FFAs milieu, regulated by peroxisomes, and cytokine release by MCs.

We affected cellular FFAs by treatment with thioridazine, a small molecule that causes accumulation of FFAs (Van den Branden and Roels, 1985; Shi et al., 2012). Similar to Pex2 ^−/− BMMCs, thioridazine-treated WT BMMCs present high amount of cellular FFAs. Upon IgE stimulation, thioridazine-treated WT BMMCs secreted lower amount of IL-6 (Figure 3E) and IL-13 (Figure 3F) compared to untreated WT BMMCs. Of note, all the treatments used to stimulate MCs and/or to manipulate cellular FFAs did not affect cell viability (Supplementary Figure S2D).

These results indicated that peroxisomal control of cellular FFAs is required in MCs to respond to and participate in inflammatory signaling cascades (Figure 4).