Pollen from birch (Betula pendula) were obtained from Allergon (batches: 012518101, 0000468269) and extracted in Dulbecco’s Phosphate Buffered Saline (PBS, Sigma Aldrich) via constant agitation at 600 rpm for 1 hour at room temperature. Protein content was determined after sterile filtration through a 0.22 µm filter using a Pierce™ 660nm Protein Assay (Thermo Fisher Scientific) according to manufacturer’s instructions.

C57BL/6 wild-type (WT) mice were initially obtained from Janvier (France). Nlrp3^-/- (B6.129S6-Nlrp3^tm1Bhk /J, strain# 021302, RRID: IMSR_JAX:021302), Casp1^-/- (B6.Cg-Casp1^em1Vnce /J, strain# 032662, RRID: IMSR_JAX:032662) and B6 Thy1.1 (WT⁺, B6.PL-Thy1^a /CyJ, strain# 000406, RRID: IMSR_JAX:000406) mice were imported from The Jackson Laboratories (USA) via the Charles River Laboratories in Germany.

All mouse lines were housed and bred under specific pathogen-free conditions in individually ventilated cages at the local animal facility at the Paris Lodron University of Salzburg (PLUS) with unlimited access to water and food and maintained under a 12-hour light-dark cycle, constant temperature and humidity.

All animal experiments complied with EU guidelines (Directive 2010/63/EU and Regulation 2019/1010) and Austrian Regulations (TVG2012) and were approved by the Austrian Federal Ministry of Education, Science and Research (permit numbers: 66.012/0015-V/3b/2019, 2021-0.588.488). Both female and male littermates were used for the described experiments. Each data set was pooled from 2–3 individual experiments.

WT, Nlrp3^-/- and Casp1^-/- mice aged 8–14 weeks were sensitized either by i.p. injection of 100 µl PBS or BPE (40 µg protein) for three times every two weeks, or by i.n. instillation of 50 µl PBS or BPE (40 µg protein) under 2.5–3% isoflurane (Iso-Vet) anesthesia for 11 times, twice per week (with intervals of at least 3 days) over a course of 6 weeks. In week 7 all mice were challenged i.n. with 40 µg BPE on three consecutive days. One day after the last challenge, mice were euthanized by cervical dislocation, and the systemic response and local lung inflammation were analyzed ( Figure 1A ).

For bone marrow chimeras, 8- to 12-week-old B6 Thy1.1 (WT⁺, Thy1.1=CD90.1) or Nlrp3^-/- (CD90.2) mice were irradiated twice with 4.75 Gray (9.5 Gray in total) with an interval of 3 hours. Donor bone marrow was isolated from femurs and tibias of WT⁺ or Nlrp3^-/- littermates. After lysis of erythrocytes with Ammonium-Chloride-Potassium (ACK) lysis buffer (150 mM NH₄Cl, 10mM KHCO₃, 0.1 mM Na₂EDTA, pH 7.4) and two washing steps, cells were passed through a 70-µm cell strainer and counted. 1–2 hours after the last irradiation of recipient mice, 5×10⁶ bone marrow cells in PBS were injected intravenously into the tail vein. Mice were provided with a broad-spectrum antibiotic (Baytril, active ingredient: enrofloxacin) in their drinking water and general health status and weight were monitored closely over the first three weeks after irradiation. 10–12 weeks after engraftment, the i.n. sensitization protocol was performed as described above.

After euthanasia, lungs of mice were lavaged by introducing 2× 1 ml cold PBS into the trachea using a Venflon™ Pro Safety venous catheter (BD) and a syringe. Lavage fluid was screened for secreted cyto- and chemokines via Luminex Multiplex Technology as described below. Cells in broncho-alveolar lavage fluid were blocked with mouse serum and stained with PerCP/Cyanine5.5 anti-mouse CD45 Antibody (clone: 30-F11; Biolegend, RRID: AB_893340), Ly-6G Monoclonal Antibody APC (clone: 1A8-Ly6g, eBioscience, RRID: AB_2573307), CD4 Monoclonal Antibody, eFluor™ 450 (clone: GK1.5, eBioscience, RRID: AB_10718983), CD8a Monoclonal Antibody FITC (clone: 53–6.7; eBioscience, RRID: AB_464916), BD Pharmingen™ PE Rat Anti-Mouse Siglec-F (clone: E50–2440; BD Biosciences, RRID: AB_394341), and PE/Cyanine7 anti-mouse CD19 Antibody (clone: 6D5, RRID: AB_313655, Biolegend), and for bone marrow chimeras additionally with Brilliant Violet 605™ anti-rat CD90/mouse CD90.1 (Thy-1.1) Antibody (clone: OX-7, RRID: AB_2562644, BioLegend) and CD90.2 (Thy-1.2) Monoclonal Antibody APC-eFluor™ 780 (clone: 53–2.1, RRID: AB_1272187, eBioscience). After a washing step with FACS buffer (PBS, 5% (v/v) heat-inactivated FCS (Merck), 3 mM EDTA (Merck)) and ACK lysis, cells were resuspended in FACS buffer. Fluorescence of WT, Nlrp3^-/- and Casp1^-/- cells was measured via BD FACSDiva software and a BD FACSCanto II (BD Biosciences), and infiltration into the lung of WT⁺ or Nlrp3^-/- mice was analyzed using CytExpert software and CytoFLEX S (Beckman Coulter). Gating was performed as described before by Neuper et al (21) except for doublet exclusion. Due to size differences of lymphocytes and granulocytes, single cell gating was performed for each gated population individually. CD90.1 or CD90.2 expression was analyzed within the CD4⁺ T cell group.

Blood was isolated from the heart. 4 µl were added to 20 µl of antibody staining mix diluted in PBS containing 10 mM EDTA. After washing once, lysing erythrocytes with ACK-lysis buffer and two further washing steps, cells were resuspended in PBS/10 mM EDTA and analyzed using CytExpert software and CytoFLEX S.

Spleens were processed and erythrocytes lysed with ACK- lysis buffer. After 2 washing steps with cell culture medium (MEM with Earle’s salts (Sigma Aldrich) supplemented with 1% (v/v) heat-inactivated FCS, 20 mM HEPES (Sigma Aldrich), 4 mM L-Glutamine (Sigma Aldrich), 1 mM Sodium pyruvate (Sigma Aldrich), 100 U/ml Penicillin/100 µg/ml Streptomycin (Sigma Aldrich), 2 µM 2-Mercaptoethanol (Gibco) and 1% (v/v) MEM- non-essential amino acids (Sigma Aldrich)), 1×10⁶ splenocytes were seeded per well to a 48-well plate and stimulated with 10 µg/ml BPE in a total volume of 500 µl for 3 days at 37°C, 5% CO₂. Supernatants were harvested and secreted cytokines and chemokines were analyzed by Luminex Multiplex Technology.

Cytokine and chemokine levels in broncho-alveolar lavage fluid (BALF) and secreted by restimulated splenocytes were measured by a bead-based multiplex technology using ProcartaPlex Mo Cytokine/Chemokine Panel 1A 36plex (Lot #: 304658–002, Thermo Fisher Scientific) and Mouse ProcartaPlex Mix&Match 30-plex (Lot #: 320797–000, Thermo Fisher Scientific). After washing beads with PBS containing 0.05% Tween-20 (Sigma Aldrich), they were resuspended in assay buffer and distributed to a 96-well V-bottom plate. 15 µl/well of standards, BALF or splenocyte supernatant were incubated overnight at 4°C with constant agitation (600 rpm) on an orbital shaker. Subsequently the plate was washed three times with 150 µl/well of wash buffer and incubated with detection antibody solution (15 µl/well) for 30 minutes at room temperature (RT) on an orbital shaker followed by three further washing steps and incubation with 20 µl/well of Streptavidin-PE solution for 30 minutes at RT on an orbital shaker. Afterwards, beads were washed again three times, resuspended in 80 µl reading buffer and measured on a Luminex MAGPIX® instrument (Luminex). Data were evaluated using ProcartaPlex Analyst 1.0 software (Thermo Fisher Scientific).

BPE-specific IgG subtypes in serum isolated from heart blood were examined by ELISA. In brief, plates were coated with 4 µg/ml BPE overnight at 4°C. After washing with PBS containing 0.1% (w/v) Tween-20 (PBS/Tween), wells were blocked with 1% BSA (w/v, Sigma Aldrich) in PBS/Tween and subsequently incubated with a serial dilution of sera (1:100 – 1:100,000). After addition of secondary antibodies (HRP Goat anti-mouse IgG Antibody (Biolegend, RRID: AB_315009, 1:2000), Goat anti-mouse IgG1 (γ1) horseradish peroxidase conjugate (Invitrogen, RRID: AB_2534048,1:1000), Goat Anti-Mouse IgG2c heavy chain (HRP) (Abcam, RRID: AB_10680258, 1:5000)), the assay was developed with ABTS substrate (2.2 mg/ml ABTS (Sigma Aldrich) in 0.1 M Citrate buffer (pH 4.25, Bio-Rad), 0.12% (v/v) H₂O₂ (Merck)) and absorbance measured at 405 nm using a Tecan infinite® M200 or M200 PRO plate reader.

BPE-specific IgE levels were determined via a cell-based assay using rat basophilic leukemia cells expressing the murine Fcϵ-receptor (RBL-2H3, RRID: CVCL_0591, CRL-2256™, ATCC) as described before (22–25). The adherent cell line was cultured in 70% (v/v) MEM with Earle’s Salts, 20% (v/v) RPMI-1640 (Sigma Aldrich), 10% (v/v) heat-inactivated FCS, 4 mM L-glutamine, 2 mM sodium pyruvate and 100 U/ml penicillin/100 µg/ml streptomycin. No mycoplasma contamination was detected in periodic tests. 8×10⁴ cells were seeded in 100 µl cell culture medium per 96-well. The following day, cells were incubated with mouse sera diluted 1:10 in culture medium for 2 hours at 37°C. After three washing steps with Tyrode’s buffer (Sigma Aldrich) supplemented with 0.1% BSA (w/v), cells were stimulated with 100 µl of Tyrode’s buffer/BSA containing 3 µg/ml BPE for 30 minutes at 37°C. To induce total lysis, 10 µl of 10% Triton X-100 (Sigma Aldrich) was added to control wells. Cross-linking of Fcϵ-receptors upon antigen-binding and subsequent degranulation was determined by means of β-hexosaminidase release. Therefore, 50 µl of cell culture supernatant were transferred to a new plate and incubated with 50 µl of 80 µM 4-Methyl umbelliferyl-N-acetyl-b-D-glucosaminide (Sigma Aldrich) in 0.1 M citrate buffer (pH 4.25) for 1 hour at 37°C. After addition of 100 µl of glycine buffer (0.2 M glycine (Sigma Aldrich), 0.2 M NaCl (Roth), pH 10.7), fluorescence was measured with a Tecan plate reader at an excitation wavelength of 360 nm and an emission wavelength of 465 nm, and β-hexosaminidase release was calculated as percentage of total lysis.

Histomorphology of lung specimens was performed as described before (21). Briefly, after fixation in 4% phosphate-buffered formalin, lung samples were marked with tissue-marking dye (General Data Company Inc, Cincinnati, OH, USA) and embedded in paraffin prior to processing several samples in one tissue cassette. H&E staining was applied to assess basic histomorphology. In combination with Giemsa staining, lymphocytes and eosinophils were highlighted. Periodic acid–Schiff (PAS) staining was performed to visualize mucopolysaccharides. Trichrome staining was used to illustrate the collagen of the basal membrane (colored blue) which allows measurement of the thickness (in μm) of the respiratory epithelium in subsegmental bronchi. All stained slides were completely scanned using the Aperio GT 450 DX whole slide image (WSI) scanner (Leica Mikrosysteme Vertrieb GmbH, Wetzlar, Germany) at 40x scan magnification with a resolution of 0.26 µm/pixel. Subsequently, all analyses of the digitized slides were performed using the Aperio Imagescope software (Leica Mikrosysteme Vertrieb GmbH, Wetzlar, Germany) and NIS Elements (Nikon CEE GmbH, Vienna, Austria) by an investigator blinded to the mouse groups.

For immunofluorescence staining, 4-μm sections mounted on glass slides were deparaffinized in xylene and rehydrated in decreasing concentrations of ethanol (100–70%). Antigen retrieval was performed in 1x citrate buffer (IHC Select Citrate Buffer pH 6.0, 10x, #21545, Merck Millipore) for 20 minutes in a steamer, followed by 15 minutes at RT. After 3 washes with PBS, tissue slices were surrounded with a Hydrophobic Barrier Pap Pen (#R3777, Thermo Fisher Scientific). Samples were permeabilized with 0.1% Triton-X100 (Sigma-Aldrich) and blocked in 2% (w/v) BSA (Sigma-Aldrich) plus 5% (v/v) donkey serum (RRID: AB_2810235, Sigma-Aldrich) diluted in 1x Fish Gelatin Blocking Agent (#22010, Biotium) for 1 hour at RT. Primary antibody incubation was performed overnight at RT. The following primary antibodies were used: ASC/TMS1 (D2W8U) Rabbit mAb (1:200, RRID: AB_2799736, Cell Signaling Technology), Goat Anti-Mouse CD45 Polyclonal antibody (1:100, RRID: AB_442146, R&D Systems). On the next day, samples were washed with PBS and incubated for 2 hours at RT in the dark with secondary antibodies and the nucleus was counterstained with 4′,6-Diamidino-2-phenyl-indol-dihydrochlorid (DAPI, 1:2000, #MBD0015, Sigma-Aldrich). The following secondary antibodies were used: Donkey anti-Rabbit IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 568 (1:1000, RRID: AB_2534017, Invitrogen) and Donkey anti-Goat IgG (H+L) Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 647 (1:1000, RRID: AB_2535864 Invitrogen). After washing again with PBS, autofluorescence was quenched by treatment with TrueBlack (#23007, Biotium) for 3 minutes according to the manufacturer’s instructions. Tissue sections were extensively washed with PBS and mounted with a microscope glass cover in ProLong Gold Antifade Mountant (Invitrogen #P36934).

As positive control for NLRP3 inflammasome formation, we analyzed ASC speck formation in BMDCs/BMDMs and digested lung tissue of healthy mice upon stimulation with LPS and Nigericin. For BMDCs/BMDMs, bone marrow was isolated from femurs and tibias of C57BL/6 mice using a 0.4×50mm needle. The cell suspension was transferred to a new tube through a 70-µm cell strainer and centrifuged for 7 minutes at 300 × g. After two washing steps with PBS, cells were taken up in BMDC medium (RPMI-1640, 10% (v/v) FCS heat-inactivated, 2 mM L-Glutamine, 100 U/ml Penicillin, 100 µg/ml Streptomycin, 50 µM 2-Mercaptoethanol) supplemented with 20 ng/ml GM-CSF, counted and 1×10⁷ cells in 4 ml seeded to 6-well plates. After two days of incubation at 37°C, 5% CO₂, half of the medium was replaced with fresh medium containing 40 ng/mL GM-CSF. Medium was completely exchanged on day 3 and cells harvested after 6 days of differentiation and frozen in 80% BMDC medium, 10% additional heat-inactivated FCS and 10% DMSO first at -70°C and afterwards in liquid nitrogen. After thawing the cells, they were washed 3 times with BMDC medium without antibiotics and 2-Mercaptoethanol, counted, seeded at a concentration of 1×10⁶ cells/24-well on 0.01% Poly-L-Lysine-coated glass cover slips and left to recover at 37°C, 5% CO₂ overnight.

Mouse lung cell suspension was generated by digesting lung tissue as described before (24, 25). Briefly, lung lobes were minced in RPMI containing 0.125 mg/ml Liberase™ (Roche), 0.25 mg/ml Liberase^TL (Roche), 0.03 mg/ml DNAse I (Sigma Aldrich) and 1 mg/ml Hyaluronidase (Sigma Aldrich) and digested for 2x30 min, 300 rpm, 37°C. Single cell suspension was seeded on 0.01% Poly-L-Lysine-coated glass cover slips.

BMDCs/BMDMs and lung cells were stimulated with 100 ng/ml LPS for 3 hours prior to stimulation with 10 µM Nigericin. The plate was briefly centrifuged (800 rpm, 10 seconds) and incubated for 45 minutes at 37°C to allow inflammasome formation. Subsequently to fix cells, 800 µl of medium was replaced with 4% paraformaldehyde to a final concentration of 3.2% for 15 minutes at RT. After washing three times with PBS, cells were permeabilized using 0.1% Triton-X100 (Sigma-Aldrich) for 10 minutes, followed by two washes with PBS. Unspecific binding sites were blocked with 2% BSA (Sigma-Aldrich) and 5% donkey serum (RRID: AB_2810235, Sigma-Aldrich) for 1 hour at RT prior to incubation with primary ASC/TMS1 (D2W8U) Rabbit mAb (1:200, RRID: AB_2799736, Cell Signaling Technology) and Goat Anti-Mouse CD45 Polyclonal antibody, (1:100, RRID: AB_442146, R&D Systems) overnight at 4°C. Afterwards, samples were extensively washed with PBS and incubated for 2 hours at RT in the dark with secondary antibodies Donkey anti-Rabbit IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 568 (1:1000, RRID: AB_2534017, Invitrogen), Donkey anti-Goat IgG (H+L) Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 647 (1:1000, RRID: AB_2535864 Invitrogen) and the nucleus was counterstained with 4′,6-Diamidino-2-phenyl-indol-dihydrochlorid (DAPI 1:2000, MBD0015, Sigma-Aldrich). After washing the samples three times with PBS, glass cover slips were semi-dry mounted in ProLong Gold Antifade Mountant (Invitrogen #P36934) on a microscope slide.

Samples were analyzed using a Zeiss Observer Z1 fluorescence microscope equipped with an Abberior Instruments STEDYCON unit enabling widefield epifluorescence and confocal and super-resolution STED microscopy. Representative confocal images from single focal z-planes were taken using the 100× oil objective lens (NA 1.46) to visualize and investigate the ASC staining pattern in lung epithelial and immune cells. For quantification of ASC and CD45 positive cells (ASC⁺/CD45⁺) in lung tissue, 1 widefield image of mouse lung sections from 3 different mice per experimental group (n=3/group) was taken using a 20x objective lens. The number of ASC⁺/CD45⁺ cells per image was manually counted using the cell counter plugin from Fiji (ImageJ 1.53t). For image acquisition and image export, the ZEN 2 blue edition (Zeiss) or the STEDYCON Web interface software (Abberior) was used. Images were post-processed with the image processing software Fiji (ImageJ 1.53t) and Microsoft PowerPoint.

Data were analyzed using Flow Jo 10, ProcartaPlex Analyst and GraphPad Prism 10.1.1. Statistical differences were examined using One-Way ANOVA with Tukey’s post hoc test, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Commonly used mouse models of allergic sensitization often depart from physiological relevance because they involve intraperitoneal (i.p.) or subcutaneous injection of antigens usually adsorbed to an adjuvant. Therefore, we first developed an adjuvant-free intranasal (i.n.) model of allergic lung inflammation. We then used this model to induce an immune response to BPE and to investigate the role of NLRP3 in BPE-induced allergic inflammation.

To investigate whether our i.n. sensitization model induces similar levels of allergic inflammation as commonly used adjuvant-free i.p. protocols, C57BL/6 mice were sensitized with BPE or mock-treated over a course of six weeks via i.p. or i.n. according to the scheme shown in Figure 1A . One week after the last immunization, all groups were challenged i.n. with BPE on three consecutive days, followed by the isolation of serum, broncho-alveolar lavage fluid (BALF), lung tissue and restimulation of splenocytes after another 24 hours.

Both i.p.- and i.n.-sensitized mice showed slightly higher serum IgG titers than mock-treated animals but barely produced Th1-associated IgG2c. Strikingly, however, the Th2-related antibody subtypes IgG1 and particularly IgE were significantly higher in the i.n. group ( Figure 1B ).

Restimulation of splenocytes with BPE revealed high amounts of Th1-associated cytokines and chemokines in all groups ( Figure 1C ), which can be explained by the considerable amounts of bacterial PAMPs, as described by McKenna and colleagues (8), in the pollen extract that we used. Remarkably, Th2-associated cytokines, in particular IL-4, IL-5 and IL-13, were only produced by BPE-sensitized mice but not by mock-treated animals via both sensitization routes ( Figure 1C ; Supplementary Figure 1A ). This set of data indicates that sensitization via the i.n. route induces robust BPE-specific, Th2-associated systemic inflammation.

As BP is an inhalant allergen, we also studied local lung inflammation by histological staining of fixed tissue sections and analyzed cell and cytokine composition in BALF. In general, sensitization with BPE resulted in highly increased cell infiltration and chemokine secretion into the lungs irrespective of the sensitization route ( Figures 2A, B, E ; Supplementary Figures S1B, C ). However, consistent with the increase in Th2-associated antibodies, the numbers of eosinophils as well as peribronchial and perivascular lymphocytes were significantly increased upon sensitization via the nasal route compared to the i.p. route ( Figures 2A, B ). Additionally, mucin production and epithelial thickness were significantly higher after i.n. sensitization ( Figures 2C, D ). As shown in Supplementary Figure S1C , we also observed higher infiltration by B cells in the i.n. sensitization group, whereas similar numbers of CD4⁺ T cells and neutrophils migrated into the lungs of the i.p. and i.n. groups. These findings show that the i.n. route induces a highly allergic lung pathology.

Taken together, i.n. sensitization resulted in a robust systemic and local allergic immune response, while the i.p. route failed to induce an IgE antibody response and provoked significantly less lung inflammation, confirming the physiological relevance of our i.n. model.

Based on the stronger allergic outcome of the adjuvant-free i.n. sensitization model, we used this procedure to elucidate the role of NLRP3 in BP allergy.

As described before, C57BL/6 WT and Nlrp3-deficient mice (whole-body knockout) were mock-treated or sensitized i.n. with BPE over a course of 6 weeks. Interestingly, after three days of challenge with BPE, serum-IgG1 and -IgE levels were significantly decreased in sensitized Nlrp3^-/- mice compared to WT ( Figure 3C ). In line with this, cytokine secretion by restimulated splenocytes was strongly attenuated in Nlrp3-deficient animals ( Figure 3D , Supplementary Figure 2A ), indicative of a significantly reduced systemic response to BPE in Nlrp3^-/- mice.

Consistent with these findings, Nlrp3-deficient animals showed a massive decrease in local lung inflammation, characterized by lower or significantly less chemokine levels in BALF ( Figure 3E ; Supplementary Figure 2B ) and reduced infiltration by eosinophils, B cells and CD4⁺ and CD8⁺ T cells compared to WT mice ( Figures 3F, G ; Supplementary Figure 2C ). Additionally, we observed significantly attenuated signs of allergic inflammation, including reduced lymphocyte accumulation in the perivascular space, mucin production and epithelial thickening ( Figures 3H, I ; Supplementary Figure 2D ).

These data clearly show significantly reduced signs of systemic and local allergic inflammation in Nlrp3-deficient mice compared to WT animals, suggesting that the absence of NLRP3 partially protects against allergic sensitization to BP and raising the question of the specific role of NLRP3 in this process.

Because NLRP3 is well described as an inflammasome-forming NLR but has also been described to act as a transcription factor, we next asked whether the induction of BP allergy in our sensitization model depends on NLRP3 as an inflammasome driver.

Immunofluorescence staining as a method to detect NLRP3 expression has serious disadvantages, as available NLRP3 antibodies very often also bind non-specifically. However, since NLRP3 activation leads to the oligomerization of ASC proteins, the so-called ASC pyroptosome (26), we analyzed ASC-speck formation in lung sections of C57BL/6 and Nlrp3^-/- mice as a surrogate for inflammasome activation. ASC was significantly higher expressed in lung tissue of sensitized WT animals and evenly distributed within CD45⁺ immune cells irrespective of genotype. However, even after three days of BPE challenge, no ASC-specks were observed in CD45⁺ cells within the lungs of sensitized mice ( Figures 4A, B ), whereas BMDCs/BMDMs and CD45⁺ cells of healthy digested lungs treated with LPS and nigericin, used as positive control, showed clear ASC-speck formation ( Supplementary Figure 3A ). The fact that lung epithelial cells also did not show ASC-speck formation suggests that inflammasome formation may play only a minor role at best in these particular cells in BPE-induced allergic inflammation.

To further analyze the potential inflammasome-independent role of NLRP3 in BP allergy, we next sensitized Casp1-deficient animals in addition to C57BL/6 WT and Nlrp3^-/- mice as described before. Interestingly, absence of Caspase 1 did not resemble NLRP3 deficiency. Instead, Casp1^-/- mice showed amounts of serum IgE and IgG1, IL-4-, IL-5- and IL-13-secretion by splenocytes, and CXCL1/CCL11 in the lungs similar to those of the WT control group ( Figures 4C–E ; Supplementary Figures 3B–D ). This suggests that Caspase 1 deficiency does not result in reduced allergic inflammation. In addition, we did not observe a significant increase in IL-1β secretion by splenocytes or increased IL-1β levels in the BALF ( Supplementary Figures 3C, D ) after sensitization with BPE in either WT, Nlrp3^-/- or Casp1^-/- mice, further confirming that NLRP3 inflammasome formation is dispensable in BPE-induced allergic responses.

To understand whether NLRP3 exerts its key function during allergic sensitization primarily in hematopoietic cells that migrate to the lung or rather in the lung epithelium, we irradiated C57BL/6- CD90.1 (WT⁺) and Nlrp3^-/– CD90.2 mice and reconstituted their hematopoietic system with bone marrow from either the same or the other mouse strain ( Figure 5A ). In both models, we achieved high engraftment rates, with 98–99% of donor cells after engraftment of Nlrp3-deficient bone marrow and 58–92% of donor cells after engraftment of WT⁺-cells ( Supplementary Figure 3E ).

After BPE sensitization and challenge, mice lacking NLRP3 in the hematopoietic system showed reduced IgE levels compared to animals receiving WT bone marrow ( Figure 5B ). In line with this, enhanced levels of IL-4 and IL-13 secreted by restimulated splenocytes and CCL11 in the lungs were observed in mice expressing NLRP3 in the hematopoietic system, compared to animals expressing NLRP3 only in the epithelium ( Figures 5C, D ).

Consistent with these findings, in the absence of NLRP3 in the hematopoietic compartment, animals showed reduced levels of infiltration by eosinophils and lymphocytes, in particular CD4⁺ T cells, into the lungs and perivascular areas ( Figures 5E–G ; Supplementary Figure 3G ). In addition, mucin production and epithelial thickness were slightly higher in mice expressing NLRP3 in immune cells compared to a deficiency of NLRP3 in the hematopoietic system ( Figures 5F, H ).

Taken together, these findings imply that while NLRP3 inflammasome formation and subsequent activation of Caspase 1 is dispensable in our model of BP allergy, NLRP3 expression in the hematopoietic system is essential for BPE-induced allergic inflammation.