To the editor,

Pyroptosis is a lytic cell death induced by various stimuli, including allergens, which is induced by inflammasome assembly and subsequent caspase cleavage [1] to remove inhibitory gasdermin D (GSDMD)-C domain and release pore-forming GSDMD-N domain to break down cell membrane [2,3,4]. A GSDMD-target strategy that limits airway inflammation from allergen-induced cellular pyroptosis would hold promise for alleviating allergic asthma. Herein, we demonstrate this precise therapeutic inhibitor of allergen-induced pyroptosis, heme oxygenase-1 (HO-1), in house dust mite (HDM)-induced allergic asthma model of mice.

HO-1 is a rate-limiting enzyme that catalyzes heme into iron, carbon monoxide and biliverdin, resulting in anti-inflammatory and anti-oxidant activity [5]. We previously reported an inhibitory effect of HO-1 by targeting the NF-κB p65 RHD domain and NLRP3-RXR axis in allergic asthma mice [6, 7]. Given the protective effects [6, 7], we set out to explore how HO-1 limits GSDMD-mediated pyroptosis. First, we assessed the expression and distribution of NLRP3, caspase-1 and GSDMD in human bronchial epithelial airway cell (BEAS-2B) stimulated by HDM. In accordance with our previous study [7], HDM-stimulated BEAS-2B displayed significantly increased NLRP3, caspase-1 and GSDMD expression (Fig. 1A). Distribution of NLRP3 and GSDMD overlapped at the margin of cell and displayed as a bubble and caspase-1 distributed at the margin of the bubble. After induction of HO-1 by hemin, scattered expression pattern of NLRP3 became condensed and caspase-1 expression significantly decreased while GSDMD expression stayed unchanged (Fig. 1A). In addition, we observed HDM led to the shift of NLRP3, GSDMD and caspase-1 distribution to the central but not the margin area of the cell, forming a bubble-like structure (white arrow), which was totally disappeared after HO-1 induction. This indicated existence of trimer structure formed by NLRP3, caspase-1 and GSDMD and importantly this process is dynamic. The finding of the inhibitory role of HO-1 in NLRP3 6, 7 and in caspase-1 (Fig. 1A) drive us to investigate whether HO-1 interacts with GSDMD. We then transfected BEAS-2B with Flag-GSDMD plasmid and found that HO-1 directly binds to GSDMD, specifically GSDMD-C, to display its inhibitory effect (Fig. 1B).

Fig. 1
figure 1

Heme oxygenase-1 binds gasdermin D to inhibit airway epithelium pyroptosis in allergic asthma. (A) BEAS-2B cells were cultured with stimulation of HDM in the presence and absence of hemin. Expression of NLRP3 (green), caspase-1 (purple) and GSDMD (red) was immunofluorescence stained. White arrows indicate expression of each protein or their overlap expression (yellow). (B) BEAS-2B cells were transfected with HO-1 plasmid and Flag-GSDMD (or GSDMD-C or GSDMD-N) plasmid and co-immunoprecipitated with Flag-antibody. Expression of Flag and HO-1 was determined by western blot. (C) Allergic airway inflammation model was established by intranasal administration of HDM for 14 consecutive days. Mice were sacrificed at day 16. Immunofluorescence staining of GSDMD-N in lung section of mouse or in human lung tissue. DAPI was stained to display nuclei. (D) HE staining of mouse lung tissue and analysis of ILC2 and eosinophils in BALF by flow cytometry. n = 5 mice each group. Bar represents 20 μm and 50 μm

To verify the above findings, allergic asthma model of mice was established by intranasal administration of HDM and hemin was intraperitoneally administered to induce HO-1 7. Increased GSDMD-N was observed in airway epithelium in HDM group mice, which was sharply reversed after HO-1 induction. In human lung tissue, immunostaining verified increased GSDMD-N in the airway epithelium of an allergic asthma patient compared to a lung cancer patient who was excluded of asthma and other allergic diseases (Fig. 1C). We also observed that HO-1 reversed HDM-induced increases in inflammatory cell infiltration, including innate lymphoid cell type 2 (ILC2) and eosinophils in the bronchoalveolar lavage fluid (BALF) (Fig. 1D). Our data coincided with a previous study showing GSDMD-mediated allergic airway inflammation [8] and further solidified HO-1 as a potential inhibitor of allergy. Since cleavage of GSDMD relies on the exosite interaction between auto-processed caspase-1 and GSDMD-C-terminus, our observation may identify a novel molecular inhibitor of GSDMD.

Collectively, this study demonstrates a near-absolute anti-pyroptotic function through HO-1 inhibiting pyroptosis corpuscle formation in allergen-stimulated airway epithelium. The therapeutic potential of HO-1 is linked to blockade of trimer formation via binding NLRP3 and GSDMD-C, major contributors to membrane porforin formation. The HO-1 biology may therefore provide a frontline therapy for organ preservation in allergy and for other pathogen-associated organ damage for which therapeutics are unavailable. Thus, our findings provide the first proof of concept supporting further basic and clinical development of HO-1 as a biologic to treat allergen-driven inflammatory disease.