Graphical abstract

Main findings of the study. Intranasal therapeutic administration of receptor-interacting serine/threonine protein kinase 2 (RIPK2) inhibitors reduces house dust mite (HDM)-induced asthma features only in humanised nucleotide-binding oligomerisation domain 1 variant (hNOD1v) mice, and acts by downregulating thymic stromal lymphopoietin (TSLP) in epithelial cell air–liquid interface (ALI) cultures from asthma patients. WT: wild-type; Th2: T-helper type 2 cell; IL-33: interleukin 33; CXCL: C-X-C chemokine motif ligand.

Abstract

Background House dust mite is the most frequent trigger of allergic asthma, with innate and adaptive immune mechanisms playing critical roles in outcomes. We recently identified the nucleotide-binding oligomerisation domain 1 (NOD1)/receptor-interacting serine/threonine protein kinase 2 (RIPK2) signalling pathway as a relevant contributor to murine house dust mite-induced asthma. This study aimed to evaluate the effectiveness of a pharmacological RIPK2 inhibitor administered locally as a preventive and therapeutic approach using a house dust mite-induced asthma model in wild-type and humanised NOD1 mice harbouring an asthma-associated risk allele, and its relevance using air–liquid interface epithelial cultures from asthma patients.

Methods A RIPK2 inhibitor was administered intranasally either preventively or therapeutically in a murine house dust mite-induced asthma model. Airway hyperresponsiveness, bronchoalveolar lavage composition, cytokine/chemokine expression and mucus production were evaluated, as well as the effect of the inhibitor on precision-cut lung slices. Furthermore, the inhibitor was tested on air–liquid interface epithelial cultures from asthma patients and controls.

Results While local preventive administration of the RIPK2 inhibitor reduced airway hyperresponsiveness, eosinophilia, mucus production, T-helper type 2 cytokines and interleukin 33 (IL-33) in wild-type mice, its therapeutic administration failed to reduce the above parameters, except IL-33. By contrast, therapeutic RIPK2 inhibition mitigated all asthma features in humanised NOD1 mice. Results in precision-cut lung slices emphasised an early role of thymic stromal lymphopoietin and IL-33 in the NOD1-dependent response to house dust mite, and a late effect of NOD1 signalling on IL-13 effector response. RIPK2 inhibitor downregulated thymic stromal lymphopoietin and chemokines in house dust mite-stimulated epithelial cultures from asthma patients.

Conclusion These data support that local interference of the NOD1 signalling pathway through RIPK2 inhibition may represent a new therapeutic approach in house dust mite-induced asthma.

Shareable abstract

Intranasal therapeutic inhibition of RIPK2 attenuates crucial features of experimental asthma in humanised NOD1 mice, and acts on epithelial cells from asthma patients, by downregulating pro-allergenic chemokines and TSLP https://bit.ly/3WnfIZk

Introduction

House dust mite (HDM) is the most important and frequent allergenic trigger for type 2-high asthma [1]. Despite type 2 immune adaptive response being one hallmark feature of HDM-driven asthma, the onset, progression and outcome of this disease are orchestrated by a variety of cell types, including epithelial cells [2], and various innate immune mechanisms [3]. Among those, our group has recently identified the nucleotide-binding oligomerisation domain 1 (NOD1) signalling pathway as a highly relevant contributor to the development of HDM-induced asthma [4].

NOD1 is a pattern recognition receptor that recognises unique muropeptides from bacterial peptidoglycan [5], polymorphisms of which have been linked to asthma and high levels of IgE in genome-wide association studies [6]. Accordingly, our group previously showed that a NOD1 agonist used as a systemic adjuvant exacerbates type 2-high allergic asthma through dendritic cell activation [7]. Interestingly, HDM-derived microbiota are sensed by NOD1 and potentiate allergic airway disease severity through the downstream receptor-interacting serine/threonine protein kinase 2 (RIPK2) [4]. Other authors have provided further evidence for the role of RIPK2 in the promotion of HDM-driven asthma using knockout models [8].

Interestingly, the NOD1 signalling pathway, and RIPK2 in particular, are associated with other inflammatory diseases, e.g. inflammatory bowel diseases [9], and different types of cancer [10, 11], which prompted the development of therapeutic RIPK2 inhibitors. Miller et al. [12] have recently explored the oral use of one of these inhibitors in a murine model of HDM-induced asthma but merely as a preventive measure. However, a phase 1 clinical trial has shown nonclinical toxicity of such an orally administered inhibitor [13].

Therefore, this study aimed to evaluate the effectiveness and mechanism of action of a low dose of RIPK2 inhibitor administered locally as a preventive or therapeutic treatment using an experimental HDM-induced asthma model in wild-type (WT) mice [4, 14] and humanised NOD1 (hNOD1) mice, and on in vitro bronchial epithelial cells from asthma patients. Here we provide evidence that local interference through RIPK2 inhibition may represent a viable alternative or a complementary therapeutic approach to currently available asthma therapies.

Materials and methods

Additional descriptions of the transgenic mice, protocols, airway hyperresponsiveness (AHR), bronchoalveolar lavage (BAL), sera, histology, ELISA, quantitative PCR (qPCR) analyses, precision-cut lung slices (PCLS), air–liquid interface (ALI) cultures, reporter cell assay and statistical analyses are included in the supplementary material.

HDM-induced allergic airway inflammation and RIPK2 inhibitor interventions

WT female C57BL/6J mice or C57BL/6J hnod1+/card4−/– (hNOD1) transgenic mice [15] found to harbour a specific Nod1 mutation associated with asthma [6] were subjected to a preventive (supplementary figure S1a) or therapeutic (supplementary figure S1c) experimental asthma protocol based on a sensitisation (day 0) and five consecutive challenges (days 7–11) via 25 µL intranasal instillation of 5 index of reactivity (IR) HDM extract (Dermatophagoides farinae) or PBS alone, or 30 min after the intranasal administration of 25 µL of 0.1 mg·kg−1 optimal concentration in PBS (supplementary figure S2) of a chemically synthesised specific RIP2K inhibitor (GSK2983559) or PBS. RIPK2 inhibitor was administered starting from the sensitisation for the preventive protocol and only during challenges for the therapeutic intervention. Results of administration of RIPK2 inhibitor without HDM were similar to those from the PBS control group (supplementary figure S3) and were not included further. On day 13, mice were assessed for AHR. BAL fluid (BALF) and serum were recovered. Lung samples were collected for RNA isolation, protein extraction and histological analysis.

Preparation of PCLS and ex vivo culture

PCLS were prepared as described previously [16]. Briefly, lung lobes were filled with 2% low melting temperature agarose, separated and cut using a vibrating microtome. PCLS were stimulated for 24 h with 1 IR HDM alone or in combination with 5 μM of the RIPK2 inhibitor. Culture supernatants were collected for cytokine quantification by ELISA and PCLS were recuperated for protein extraction.

ALI bronchial culture, stimulation and analysis

Fully differentiated bronchial epithelial ALI cultures (MucilAir) reconstituted from primary human cells obtained from healthy donors and asthma patients were stimulated by nebulisation of 1 IR of HDM in 50 μL of PBS or with 50 μL of PBS in the presence or absence of 5 μM RIPK2 inhibitor in the culture medium. Culture supernatants were collected and analysed with the multiplex proximity extension assay (PEA) inflammation kit (Olink).

Statistical analyses

Normally distributed data were evaluated by one-way ANOVA with post hoc tests for multiple comparisons, and by two-tailed t-test for pairwise comparisons using GraphPad Prism 9.5 software. For airway resistance, two-way ANOVA was used. Non-normally distributed data were analysed using the Kruskal–Wallis H with Dunn's post hoc test. PEA data were evaluated by multiple t-test analysis for pairwise comparisons and matched two-way ANOVA with Fisher's least significant difference post hoc test for multiple comparisons.

ResultsLocal preventive administration of the RIPK2 inhibitor reduces AHR and prevents airway eosinophilia

The in vivo local preventive effect of the RIPK2 inhibitor was studied in a model of HDM-induced asthma, using a 0.1 mg·kg−1 intranasal administration during the whole duration of the protocol. The preventive administration of the RIPK2 inhibitor significantly reduced the increase in AHR measured in response to methacholine exposure in the HDM-exposed group (figure 1a), as well as BALF total cell counts including eosinophil (figure 1b) and lymphocyte recruitment (supplementary figure S1b). In addition, the serum levels of HDM-specific type 2-related IgG1, but not total IgE, were significantly reduced (figure 1b and supplementary figure S1b). The histological analysis [17] showed a reduction of mucus production with the preventive RIPK2 inhibitor compared to the HDM group (figure 1c). The intranasal preventive intervention was capable of averting the upregulation of several cytokines and chemokines traditionally related to type 2 immunity, including interleukin (IL)-4, IL-5, IL-13 and C-C motif chemokine ligand (CCL) 17 as well as IL-17, the alarmin IL-33 and the neutrophil attracting C-X-C motif chemokine ligand (CXCL) 1. By contrast, thymic stromal lymphopoietin (TSLP) level was not upregulated upon allergen challenge (figure 1d). Taken together, these data demonstrate that the local preventive inhibition of RIPK2 mitigates the cardinal features of HDM-induced asthma in this experimental model.

FIGURE 1

Receptor-interacting serine/threonine protein kinase 2 (RIPK2) inhibitor intranasal preventive administration reduces asthma features. a) Airway resistance of mice challenged with PBS, house dust mite (HDM) or HDM and RIPK2 inhibitor. b) Total and eosinophil bronchoalveolar lavage fluid cell counts in mice challenged with PBS, HDM or HDM and RIPK2 inhibitor. ELISA detection of HDM-specific IgG1 in serum of mice challenged with PBS, HDM or HDM and RIPK2 inhibitor. OD: optical density. c) Representative photographs and quantification of stained lung sections for mucus using periodic acid–Schiff (pink). Scale bars: 100 μm. d) mRNA relative expression (RE) of cytokines and chemokines assessed by quantitative real-time PCR in lung extracts from mice challenged with PBS, HDM or HDM and RIPK2 inhibitor. Data are presented as mean±sem of n=6–9 animals per group and for mucus as mean score±sem per bronchus of two to three lung sections from n=2–5 animals per group. IL: interleukin; CCL: C-C motif chemokine ligand; CXCL: C-X-C motif chemokine ligand; IFN-γ: interferon-γ; TSLP: thymic stromal lymphopoietin. *: p<0.05; **: p<0.01; ***: p<0.001.

Local therapeutic RIPK2 inhibitor intervention in WT mice fails to reduce AHR and cell recruitment but modulates IL-33 response

To assess the potential therapeutic efficacy of local RIPK2 inhibition using the same model, the RIPK2 inhibitor was administered only during the challenge phase of the experimental asthma protocol. The therapeutic intervention did not result in changes in AHR (figure 2a), cell recruitment in the BALF, levels of HDM-specific IgG1 and total IgE (figure 2b and supplementary figure S1d) or mucus secretion (figure 2c). Nonetheless, while the expression levels of IL-4, IL-5, IL-13, CCL17 and IL-17 remained unaffected, the expression of IL-33, CCL2 and CXCL1, related to the epithelial response, was significantly reduced and expression of interferon-γ (IFN-γ) was significantly increased. Additionally, TSLP expression was decreased in response to HDM challenge, suggesting a retro-control mechanism (figure 2d). Even though signalling through RIPK2 plays an important role in the early sensitisation phase, these results suggest that in this HDM-induced asthma model, the therapeutic inhibition only resulted in changes in some relevant asthma mediators, and not in lung function or cell recruitment.

Local therapeutic RIPK2 inhibition mitigates asthma features in hNOD1 mice

The human NOD1 and mouse Nod1 orthologues exhibit different affinity and recognition of peptidoglycan moieties [18]. Furthermore, a mutation in hNOD1 is associated with asthma [6]. Hence, we decided to explore the efficacy of the local therapeutic RIPK2 inhibitor intervention in a mouse expressing the hNOD1 transgene and deficient for murine Nod1 [15]. Sequencing of the hNOD1 transgene revealed that the BAC cloned hNOD1 gene [15] corresponded to the asthma-associated risk allele variant [6] (referred to as hNOD1v). The same HDM-induced asthma protocol used in our hNOD1v and co-housed WT mice in both cases induced significant increases in AHR. However, while the RIPK2 therapeutic intervention failed to modify AHR in the WT mice, there was a significant reduction in AHR in the RIPK2 inhibitor-treated hNOD1v mice (figure 3a). The BALF total cell counts showed that HDM-treated hNOD1v exhibited more cells than WT mice, including mostly eosinophils, and that these cell increases were significantly reduced by the RIPK2 inhibitor (figure 3b). Moreover, RIPK2 inhibition resulted in reduced BALF neutrophil and lymphocyte numbers (supplementary figure S4). While no significant changes were observed in the specific humoral response (figure 3b and supplementary figure S4), the histological analysis showed a reduction in mucus production (figure 3c). Importantly, hNOD1v mice displayed higher expression of IL-4, IL-5, IL-13 and CCL17 than WT mice. Furthermore, while the RIPK2 inhibitor failed to modify the expression of the above-mentioned mediators in WT mice, all of them except CCL17 were reduced in hNOD1v mice when compared to the HDM-treated group. Moreover, the higher levels of IL-33 and IL-17 in hNOD1v mice were significantly reduced. Again, TSLP expression was reduced in HDM-challenged hNOD1v mice (figure 3d). Beyond HDM [4], cat and dog but not Alternaria allergens were capable of activating a hNOD1 cell reporter assay (supplementary figure S5). These data underline the higher sensitivity of the human NOD1 variant to HDM stimulation, and the beneficial effect of RIPK2 inhibitors in a therapeutic setting, through a decreased Th2-type profile and lower IL-33 expression. Moreover, these results demonstrate that the therapeutic local inhibition of RIPK2 in a human NOD1 signalling context mitigates the main asthma hallmarks of this HDM-induced model, and potentially of other allergens that are composed of NOD1 agonists.

FIGURE 2

Therapeutic receptor-interacting serine/threonine protein kinase 2 (RIPK2) inhibitor intervention only modulates mediators related to the epithelial response. a) Airway resistance of mice challenged with PBS, house dust mite (HDM) or HDM and RIPK2 inhibitor. b) Total and eosinophil bronchoalveolar lavage fluid cell counts in mice challenged with PBS, HDM or HDM and RIPK2 inhibitor. ELISA detection of HDM-specific IgG1 response in serum of mice challenged with PBS, HDM or HDM and RIPK2 inhibitor. OD: optical density. c) Representative photographs and quantification of stained lung sections for mucus using periodic acid–Schiff (pink). Scale bars: 100 μm. d) mRNA relative expression (RE) of cytokines and chemokines assessed by quantitative real-time PCR in lung extracts from mice challenged with PBS, HDM or HDM and RIPK2 inhibitor. Data are presented as mean±sem of n=4–10 animals per group, and for mucus as mean score±sem per bronchus of two to three lung sections from n=4–5 animals per group. IL: interleukin; CCL: C-C motif chemokine ligand; CXCL: C-X-C motif chemokine ligand; IFN-γ: interferon-γ; TSLP: thymic stromal lymphopoietin. *: p<0.05; **: p<0.01; ***: p<0.001.

FIGURE 3

Therapeutic local receptor-interacting serine/threonine protein kinase 2 (RIPK2) inhibition mitigates asthma features in asthma-associated humanised nucleotide-binding oligomerisation domain 1 (NOD1) variant (hNOD1v) mice. a) Airway resistance in co-housed wild-type (WT) (upper) and hNOD1v (lower) mice challenged with PBS, HDM or HDM and RIPK2 inhibitor. b) Total and eosinophil bronchoalveolar lavage fluid cell counts in co-housed WT (left) and hNOD1v (right) mice challenged with PBS, HDM or HDM and RIPK2 inhibitor. ELISA detection of HDM-specific IgG1 response in serum from co-housed WT and hNOD1v mice challenged with PBS, HDM or HDM and RIPK2 inhibitor. OD: optical density. c) Representative photographs and quantification of stained lung sections for mucus using periodic acid–Schiff (pink) in hNOD1v mice. Scale bars: 100 μm. d) mRNA relative expression (RE) of cytokines and chemokines assessed by quantitative real-time PCR in lung extracts from co-housed WT and hNOD1v mice challenged with PBS, HDM or HDM and RIPK2 inhibitor. Data are presented as mean±sem of n=7–9 (WT) and n=5–10 (hNOD1v) animals per group, and for mucus as mean score±sem per bronchus of two to three lung sections from n=4–5 animals per group. IL: interleukin; CCL: C-C motif chemokine ligand; CXCL: C-X-C motif chemokine ligand; IFN-γ: interferon-γ; TSLP: thymic stromal lymphopoietin. *: p<0.05; **: p<0.01; ***: p<0.001.

RIPK2 inhibition reduces cytokine production in HDM PCLS obtained from hNOD1v mice

To get more insight into the early versus late events occurring in this humanised model, we next evaluated the ex vivo effect of the RIPK2 inhibitor using PCLS. Ex vivo HDM stimulation of PCLS obtained from both WT and hNOD1v naïve mice promoted significant increases in the production of TSLP and IL-33, with significantly higher levels in hNOD1v PCLS. The addition of the RIPK2 inhibitor reduced the HDM-induced TSLP and IL-33 production in both WT and hNOD1v PCLS (figure 4a). In contrast, HDM stimulation of PCLS prepared from WT or hNOD1v mice subjected to the HDM-induced asthma protocol did not result in significant changes in their TSLP and IL-33 production. Additionally, the RIPK2 inhibitor did not change the TSLP and IL-33 supernatant concentrations (figure 4b).

FIGURE 4

Receptor-interacting serine/threonine protein kinase 2 (RIPK2) inhibition reduces cytokine production in house dust mite (HDM)-stimulated precision-cut lung slices (PCLS). a) ELISA detection of thymic stromal lymphopoietin (TSLP), interleukin (IL) 33 and IL-13 in supernatants of PCLS obtained from two naïve wild-type (WT) mice and one humanised nucleotide-binding oligomerisation domain 1 (NOD1) variant (hNOD1v) mouse stimulated ex vivo with PBS, HDM, the RIPK2 inhibitor alone or HDM and the RIPK2 inhibitor. b) ELISA detection of TSLP, IL-33 and IL-13 in supernatants of PCLS obtained from three co-housed WT mice and one hNOD1v mice after completion of the HDM-induced asthma protocol and ex vivo stimulation with PBS, HDM, the RIPK2 inhibitor alone or HDM and the RIPK2 inhibitor. Data are presented as mean±sem of n=3–9 PCLS per group. ND: not detected. *: p<0.05; **: p<0.01.

While IL-13 was not detectable in the supernatants of PCLS obtained from naïve WT mice, PCLS from naïve hNOD1v and both asthmatic WT and hNOD1v mice produced IL-13 in detectable concentrations at baseline. Whereas the ex vivo HDM stimulation of the asthmatic WT PCLS did not result in a significant IL-13 change, naïve and asthmatic hNOD1v PCLS presented with significantly increased IL-13 levels. In these three groups, RIPK2 inhibitor treatment decreased the IL-13 concentration in HDM-stimulated PCLS, with only a trend in asthmatic hNOD1v mice (figure 4b). IL-25 was not detected, while IL-5 and CCL17 production was not modified regardless of the PCLS group (data not shown). These results underline the differences in the murine and human NOD1 orthologue-driven responses and emphasise the early role of TSLP and IL-33 in the NOD1-dependent response to HDM, and a late effect of NOD1 signalling on IL-13 effector response.

RIPK2 inhibition modulates cytokine and chemokine production in human asthmatic bronchial ALI cultures with and without concomitant HDM stimulation

To further explore the mechanistic effect of RIPK2 inhibitors in a NOD1 human cell signalling setting and the role of the epithelium, we used fully differentiated bronchial epithelium ALI cultures obtained from healthy individuals and asthma patients, stimulated or not with HDM, and comprehensively analysed them using a 96 inflammation-related protein PEA. HDM stimulation of the bronchial ALI cultures from healthy individuals induced an increased expression of several mediators, including pro-inflammatory interleukins and chemokines (IL-6, leukaemia inhibitory factor (LIF), monocyte chemotactic protein 3 (MCP-3), CXCL5, CXCL10), fibroblast growth factor 19 (FGF-19), the scaffold protein axis inhibition protein 1 (AXIN1), the signalling molecules eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1) and CD244, and the enzyme adenosine deaminase (ADA) (figure 5a). Conversely, RIPK2 inhibitor treatment of these cultures led to decreased production of a group of mediators including mainly chemokines, with HDM-induced production of CXCL5 and CXCL10 mediated by RIPK2 (figure 5a).

FIGURE 5

Receptor-interacting serine/threonine protein kinase 2 (RIPK2) inhibition modulates cytokine and chemokine production in bronchial air–liquid interface (ALI) cultures with and without concomitant house dust mite (HDM) stimulation. All data were obtained from n=2 healthy controls and n=3–4 asthma patients. a) Volcano plot of PBS- versus HDM- and RIPK2 inhibitor-treated or non-treated HDM-stimulated healthy bronchial ALI supernatants, highlighting significantly different proteins calculated as p<0.05 by multiple t-test. b) Volcano plot of PBS- and HDM-stimulated healthy versus asthmatic bronchial ALI supernatants, highlighting significantly different proteins calculated as p<0.05 by multiple t-test. c) Mean normalised protein expression (NPX) values of bronchial ALI culture supernatants obtained from n=3–4 asthma patients, stimulated or not with HDM in the presence or absence of RIPK2 inhibitor and evaluated twice on different dates. Data are presented as mean±sem of two replicates per patient per group. *: p<0.05; **: p<0.01; ***: p<0.001.

At baseline, ALI cultures from asthma patients exhibited higher levels of IL-6 and the repair proteins glial cell derived neurotrophic factor (GDNF) and matrix metallopeptidase 10 (MMP10) compared to ALI from healthy individuals. Moreover, HDM stimulation of asthmatic ALI cultures further differentiated them from healthy individuals, showing increased levels of the repair proteins vascular endothelial growth factor A (VEGFA) and MMP10, and of TSLP, IL-4 and CCL23, with decreased levels of CXCL10 (figure 5b). In asthma patients, RIPK2 treatment of PBS-stimulated ALI cultures inhibited the production of some chemokines, including CCL8, CXCL6, CXCL10 and CXCL11, but increased IL-6. Treatment of HDM-stimulated asthma ALI cultures with the RIPK2 inhibitor resulted in reduced levels of TSLP, programmed death ligand 1 (PDL1), IL-17C, CCL8, CXCL9, CXCL10 and CXCL11 (figure 5c). Among these molecules, only TSLP was significantly increased by HDM stimulation as compared with baseline levels. Conversely, HDM stimulation and the addition of the RIPK2 inhibitor resulted in augmented levels of IL-6.

Collectively, these data demonstrate that HDM-induced asthma can be modified by therapeutic interventions through epithelial RIPK2 in the context of human NOD1 activation.

Discussion

Our results show that local RIPK2 inhibition interventions can effectively mitigate asthma development in HDM-induced asthma murine models. In concordance with Miller et al. [12], RIPK2 inhibitor administered as a preventive treatment reduced eosinophil recruitment and Th2 and Th17 cytokine production. Equally of importance we have demonstrated that it also decreases AHR and HDM-specific IgG1 levels. While Miller et al. [12] used an oral route to administer the RIPK2 inhibitor, we took advantage of intranasal administration with a very low dose in order to minimise systemic exposure and potential toxicity [19] and to achieve targeted RIPK2 inhibition in the lung. We presume that the chosen local administration route might be the reason for the higher efficacy of the inhibitor treatment in our preventive model.

Therapeutic RIPK2 inhibitor intervention showed only a limited effect on the asthma features in WT mice, yet significantly reduced the levels of some of the mediators related to the epithelial response and previously linked to the activation of NOD1 and RIPK2, e.g. IL-33, CCL2 and CXCL1 [4, 8, 20]. This discrepancy between the effects in the preventive and therapeutic interventions could indicate that RIPK2 signalling plays a more relevant role in the early phase of sensitisation, during which the epithelial response to HDM through NOD1 and RIPK2 helps shape the subsequent adaptive responses. In agreement, our group has previously used chimeric mice to show that structural cells play the most relevant role in the response to local HDM through NOD1 and RIPK2 [4].

The reduction in all the main asthma features with the therapeutic RIPK2 intervention in the transgenic mouse expressing the asthma-associated hNOD1 variant, together with the stronger induction of asthma features, supports the existence of a divergent response between the murine and the human NOD1–RIPK2 signalling axis to HDM, and the potential higher relevance of RIPK2 inhibition in susceptible human asthma. Indeed, hNOD1 preferentially detects tri-diaminopimelic acid (DAP) ligands whereas mNOD1 detects tetra-DAP ligands [18]. Furthermore, our group has previously shown that the main NOD1 peptidoglycan ligand present in HDM extracts is MTriDAP [4] and that peptidoglycans present in HDM extracts aggravate experimental asthma. These data show that murine and human NOD1 orthologues present with different responses to HDM-derived peptidoglycan moieties, and that the preferential and stronger activation of hNOD1 by MTriDAP and other muramyltripeptides may explain the differential therapeutic effect of RIPK2 inhibitors.

To get further insight into the early versus late events modulated by RIPK2 inhibitors, we used PCLS obtained from WT and hNOD1v mice. At baseline, PCLS from both strains responded to in vitro HDM stimulation by producing the epithelium-derived TSLP and IL-33, production of which was abrogated by RIPK2 inhibitors, suggesting that epithelial NOD1 may be a driving force in the initiation of sensitisation in response to HDM. In already sensitised and challenged PCLS, RIPK2 inhibition altered IL-13 production. Th2 polarisation is achieved through dendritic cells that also strongly express NOD1 [7], which suggests that RIPK2 inhibitors may also act on the effector phase of the allergic asthma reaction, as observed in the therapeutic approach in hNOD1v mice.

We further explored the mechanisms involved by using human bronchial ALI cultures that recapitulate the donor's epithelium characteristics and the transcriptional changes and responses to external stimuli [21]. Our results show that ALI cultures obtained from healthy donors and asthma patients showed different protein profiles that match some changes previously observed in asthma patients, such as higher levels of IL-6 and other mediators related to eosinophil recruitment, airway inflammation and remodelling (e.g. MMP10) [22] or to increased AHR (e.g. GDNF) [23, 24]. Similarly, HDM stimulation reproduced some of the changes typically associated with asthma responses, such as changes in MMP10 between healthy subjects and asthma patients. Furthermore, HDM stimulation resulted in reductions of the pro-Th1 chemokine CXCL10 [25] and increases in the Th2 cytokine IL-4 [26], pro-eosinophilic chemokine CCL23 [27] and epithelial alarmin TSLP [28] between the two groups, further validating the representativeness of the ALI model for Th2 human asthma responses. Mechanistically, RIPK2 inhibitor in asthma ALI cultures reduced the levels of the alarmin TSLP, a pleiotropic cytokine able to induce Th2 differentiation through dendritic cells, eosinophil recruitment and bronchial remodelling [29], as well as PDL1, which is associated with Th2 responses and increased IL-4 production [30]. RIPK2 inhibition concomitantly with HDM stimulation also resulted in reduced levels of different chemokines known to modulate the recruitment of eosinophils, neutrophils, basophils, monocytes and T cells to the airway [31]. Unexpectedly, HDM stimulation in the presence of the RIPK2 inhibitor resulted in increased levels of IL-6, a cytokine with pleiotropic functions shown to be fundamental for airway integrity, promoting lung repair and epithelial cell survival [32], but also capable of promoting T regulatory cell development [33]. Moreover, in murine asthma the absence of IL-6 leads to increased lung inflammation and eosinophil recruitment [34], suggesting a protective role.

Altogether these results on human epithelium suggest that the RIPK2 inhibitor acts by inhibiting some eosinophil- and Th2-attracting chemokines, as well as on TSLP, a major mediator in asthma, whose systemic or local inhibition is associated with successful management of asthma in clinical trials [35, 36].

The main strength of our study is the translational perspective, using first human ex vivo analyses, second in vivo humanised mice and third a therapeutic pharmacological approach. A limitation of our study is the lack of clinical information on the asthma patients who provided the ALI cultures for this study; a second limitation is the as yet unavailable non-asthma-associated hNOD1 mouse.

The results presented here demonstrate that RIPK2 inhibition downregulates asthma responses in a humanised asthma model even after sensitisation is already established, and impairs TSLP in asthma ALI and naïve humanised PCLS cultures. The fact that our local intranasal administration reduced the effective inhibitor concentration needed to efficiently block RIPK2 activity and the current development of new generation RIPK2 inhibitors [37, 38] makes the therapeutic targeting of RIPK2 a very appealing novel strategy for fighting HDM asthma.

Supplementary materialSupplementary Material

Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.

Supplementary material ERJ-02288-2023.Supplement

Footnotes

Conflict of interest: All authors report support for the present study from Inserm and grants from Institut Pasteur (PTR (18-16)) and ANR (18-CE14-0020). M. Chamaillard reports grants from CoPOC grant entitled RIFFLE from Inserm-Transfert. M. Chamaillard, I. Gomperts Boneca, M. Fanton d'Andon, S. Ait Yahlia, D. Alvarez-Simon and A. Tsicopoulos report a patent filed on 11 October 2023 (Nber EP23306767.7). I. Gomperts Boneca and M. Fanton d'Andon report support for the present study from Laboratoire d'Excellence “Integrative Biology of Emerging Infectious Diseases” (ANR-10-LABX-62-IBEID).

This article has an editorial commentary: https://doi.org/10.1183/13993003.01372-2024

Support statement: The work was supported by grants from PTR (18–16) from Institut Pasteur (to A. Tsicopoulos) and from ANR 18-CE14-0020 (to A. Tsicopoulos, M. Chamaillard and I. Gomperts Boneca). Funding information for this article has been deposited with the Crossref Funder Registry.

Received December 20, 2023.Accepted July 9, 2024.Copyright ©The authors 2024. For reproduction rights and permissions contact permissionsersnet.org