Notch2 Directs aTreg Cell Fate Toward Immunoregulation or Inflammatory Pyroptosis

ABSTRACT

Activated regulatory T cells (aTregs) exhibit potent immunosuppressive functions and can migrate to tissues, playing a crucial role in attenuating inflammatory responses. The precise role and molecular mechanisms through which Notch2 regulates the immunoregulatory functions of aTregs remain incompletely elucidated. Here, we elucidate the mechanisms through which Notch2 influences aTreg dynamics and mitigates allergic rhinitis (AR) development. Mechanistically, the targeted knockout of Notch2 in aTregs resulted in decreased Foxo1 expression and elevated phosphorylated ASC (p‐ASC) levels, culminating in GSDMD‐N‐mediated pyroptosis in aTregs. Moreover, the Notch2 intracellular domain (NICD2) promoted the nuclear translocation of RREB1 through direct protein–protein interactions, thereby enhancing Foxo1 transcriptional activity. Importantly, the adoptive transfer of Notch2+ aTregs significantly reduced Th2 inflammatory responses in AR mice, providing an effective therapeutic strategy for managing AR‐related inflammation. Overall, our findings establish a novel paradigm in the pathogenesis of AR in which Notch2 expression dictates the functional dichotomy of aTregs in maintaining their immunoregulatory capacity or undergoing inflammatory pyroptosis.

Schematic illustration showing that the Notch2 intracellular domain (NICD2) facilitates pyroptosis resistance in activated Tregs through the RREB1/Foxo1 signaling pathway. In addition, Notch2‐mediated pyroptosis resistance in activated Tregs promotes immunoregulatory capacity, thereby attenuating Th2‐driven inflammatory responses in allergic rhinitis (AR). (AS1842856: Foxo1‐specific inhibitor).

1. Introduction

Allergic rhinitis (AR), which affects 4–38% of the global population, represents a growing health burden with limited treatment options [1, 2]. While conventional therapies (such as steroids, antihistamines, and leukotriene antagonists) provide symptomatic relief, 20% of patients remain refractory, and allergen immunotherapy faces practical limitations [2]. These findings underscore the need for a deeper understanding of the pathogenesis of AR and novel therapeutic strategies.

Regulatory T cells (Tregs) serve as master regulators of immune homeostasis through cytokine secretion and cell contact–dependent mechanisms [3]. Based on their functional and phenotypic differences, Tregs can be divided into two subpopulations: resting‐like Tregs (resting Tregs, rTregs) and activated‐like Tregs (activated Tregs, aTregs). Mouse aTregs (CD4+Foxp3+CD44+CD62L‐) and their human counterparts (CD4+CD25+CD127‐CD45RA‐) exhibit potent immunosuppressive capacity and tissue‐homing properties critical for controlling inflammation [4, 5, 6, 7, 8]. However, the immunosuppressive function of mouse rTregs (CD4+Foxp3+CD44‐CD62L+) and their human counterparts (CD4+CD25+CD127‐CD45RA+) is relatively weak, and under certain conditions, rTregs can transform into aTregs [4, 5, 6]. Tregs, particularly the activated subset (aTregs), play pivotal roles in immune regulation and the pathogenesis of AR [3, 8].

The Notch signaling pathway has emerged as a pivotal regulator of immune cell biology, influencing proliferation, survival, and functional differentiation [9]. Notably, the Notch pathway, especially through Notch1/2 isoforms, regulates Treg biology [9, 10, 11, 12, 13, 14]. Although the role of Notch1 in Treg autophagy has been established [12], our previous work revealed that Notch2 is crucial for Treg differentiation and function [13, 14], although its subtype‐specific mechanisms remain unclear.

Pyroptosis, which is mediated by GSDMD cleavage following NLRP3 inflammasome activation, represents an inflammatory cell death pathway [15, 16, 17, 18, 19]. The present study links aTreg pyroptosis to Notch2 downregulation. Notch2+ aTregs exhibit immunomodulatory properties, whereas Notch2‐ aTregs undergo pyroptosis. Mechanistically, Notch2 deficiency in aTregs reduces RREB1/Foxo1 expression while increasing the level of phosphorylated ASC, triggering GSDMD‐N‐mediated pyroptosis. Adoptive transfer of Notch2+ aTregs effectively inhibits aTreg pyroptosis and ameliorates AR by suppressing Th2‐mediated inflammatory responses. Our findings provide new insights and targets for the treatment of AR.

2. Materials and Methods

2.1. Subjects

In this study, patients who underwent allergy screening at the Department of Otolaryngology–Head and Neck Surgery of Renmin Hospital of Wuhan University outpatient clinic from September 2023 to April 2024 were recruited. The study included 41 healthy individuals and 46 AR patients from whom peripheral blood samples were collected for analysis (Table S1). The inclusion criteria for AR patients were in accordance with the literature [20]. This study was approved by the Ethics Committee of the Renmin Hospital of Wuhan University (WDRY2023‐K143), and all patients signed written informed consent forms.

2.2. Visual Analog Scale

Visual analog scale scores were assessed based on methods described in previous studies [21]. Scores ranged from 0 to 10, with increasing scores from 0 to 10 indicating increasing symptom severity.

2.3. Animals

B6.129(Cg)‐Foxp3tm4(YFP/icre)Ayr/J (Stock No: 016959) [22] mice and B6.129S‐Notch2tm3Grid/J (Stock No: 010525) [23] mice were obtained from the Jackson Laboratory (USA). The Foxp3^{Cre, YFP} Notch2^flox/flox mice (referred to as CKO mice) used in the study were obtained from the offspring of these two groups of mice. All experimental procedures were approved by the Institutional Animal Care and Use Committee of Renmin Hospital of Wuhan University (license numbers: WDRM20170310 and WDRM20231203B).

2.4. Mouse Gene Identification

Mouse genomic DNA was extracted according to the instructions of the DNA extraction kit. The extracted mouse DNA, Taq polymerase, dNTPs, 2.5 mM MgCl2, and 1 µM Notch2 primers (5'–3' CAACCCCAGATAGGAAGCAG and GAGCCTTTTCCCCATATTCC) were used for PCR (94°C for 5 min; 94°C for 30 s, 65°C for 30 s, and 68°C for 30 s × 34 cycles; and 72°C for 10 min). The PCR products were identified by 1.5% agarose gel electrophoresis. The same method was used to analyze whether Foxp3 carried the Cre gene, with the Cre primer sequences (5′‐3′ GTGTGACTGCATGACTAATTTGA and TGGCTGGACCAATGTGAAC).

2.5. Cell Lines

Treg‐like cells (MoT cells) were obtained from the American Type Culture Collection (ATCC, No: CRL‐8066, RRID: CVCL_1439). Jurkat (No: ORC0830, RRID: CVCL_0065) and THP‐1 (No: ORC0194, RRID: CVCL_0006) cells were obtained from China AORUICELL Biological (Shanghai, China). All the cell lines were verified, with reports indicating no contamination. Additionally, MoT oeNICD2 and MoT shNICD2 cells, and their control cells, MoT oeVec and MoT shNC cells, were constructed. Different groups of MoT cells were subjected to various treatments, including house dust mite extract (HDM) and AS1842856 (a specific inhibitor of Foxo1), lipopolysaccharide (LPS), IL‐4, GM‐CSF, nigericin, disulfiram (a pyroptosis‐specific inhibitor), necrostatin‐1 (a necroptosis inhibitor), and Z‐VAD‐FMK (a panapoptosis inhibitor).

2.6. Flow Cytometry and Sorting

Surface‐staining antibodies were added and incubated with the lymphocytes for 30 min at 4°C. For intracellular staining, we used a permeabilization buffer kit (BD, 562574) according to the manufacturer's instructions; intracellular antibodies were added and incubated with the cells at 4°C for 50 min. Antibody details for flow cytometry are provided in Table S2. Additionally, we used a cell death detection kit (annexin V + and 7‐AAD +; Hang Zhou Lianke Biotechnology Co., Ltd.) to detect cell death. The levels of cytokines in human serum were measured using a cytokine microbead array (CBA). Data were acquired on a Beckman flow cytometer.

Mouse spleen lymphocytes and human PBMCs were stained, and a previously described method [24] was used to sort mouse Tregs, rTregs, human aTregs, and rTregs with an S3e Cell Sorter.

2.7. RNA‐Seq

Activated Tregs were sorted from mice according to the method described by Kadja T et al. [25]. A KC‐Digital Stranded mRNA Library Prep Kit for Illumina (Catalog no. DR08502; Wuhan Seqhealth Co., Ltd.) was used to prepare a single‐stranded RNA sequencing library from 2 µg of total RNA. The product was enriched, quantified, and ultimately sequenced on a NovaSeq 6000 sequencer (Illumina) at 200–500 bps. The edgeR package (version 3.12.1) was used to identify genes that were differentially expressed between the groups. A p‐value threshold of 0.05 and a fold change threshold of 2 were used to judge the statistical significance of differences in gene expression.

2.8. Generation of Knockdown and Overexpression Cell Lines

We constructed viruses and transfected them into cells using our previously described approach [13, 26]. 293T cells were used for lentiviral packaging; viral supernatants were collected after 48 h. At 24 h after the transfection of MoT cells, the cells were selected with puromycin (MedChemExpress, HY‐B1743A) or G418 (MedChemExpress, HY‐K1056) for 14 days prior to expansion.

2.9. Western Blot

Specific steps for protein extraction, detection, and analysis to evaluate protein expression were conducted with reference to the literature [27, 28]. The WB antibodies used in this research are listed in Table S3.

2.10. SEM and TEM

Cells were placed on a cover slide and fixed with electron microscopy fixation solution (Servicebio, GP2001) for 2 h. Images were captured with a scanning electron microscope (Hitachi SU8100) and a transmission electron microscope (Hitachi HT7800) and analyzed. The method reported by Chang L et al. was used to count pyroptotic cells [29].

2.11. Real‐Time Quantitative‐PCR

TRIzol (TIANGEN Biochemical Technology, G3013) was used to extract total cellular RNA, the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, USA) was then used to reverse transcribe the RNA into cDNA, and FastStart Universal SYBR Green Master Mix (Rox; Roche) was used for real‐time quantitative PCR. The experimental methods have been previously described [30]. mRNA expression was normalized to that of Notch1 in rTregs. The primers used are shown in Table S4.

2.12. Animal Models

2.12.1. AR Animal Model

Thirty‐two female Foxp3^{Cre, YFP} mice were selected and divided into a control group, a 3‐day nasal drip group, a 7‐day nasal drip group, and a 14‐day nasal drip group, with 8 mice in each group. The models were constructed according to our previously described methods [13, 27]. The mice in the 3‐day nasal drip group received nasal drops for 3 days, those in the 7‐day nasal drip group received nasal drops for 7 days, and those in the 14‐day nasal drip group received nasal drops for 14 days.

2.12.2. Adoptive Transfer Mouse Model

The adoptive transfer mouse model was established based on our previous studies [13, 27]. Briefly, aTregs were transduced with a lentiviral vector overexpressing the mouse Notch2 intracellular domain (NICD2) or the corresponding empty vector control as described in our previous studies [13, 27]. Subsequently, these transduced aTregs were activated via CD3/CD28 stimulation. Fifteen female Foxp3^{Cre, YFP} mice were divided into four groups: the control group, the AR group, and the AR group with adoptive transfer of Notch2‐overexpressing aTregs (oeNotch2‐aTreg) or empty vector‐transfected aTregs (EV‐aTreg). In the adoptive transfer group, 4 × 10⁵ aTregs overexpressing Notch2 or empty vector‐transfected aTregs; these cells were adoptively transferred to each AR mouse on the seventh day following nasal drip stimulation.

2.13. Symptom Scores

The scoring methods were as previously described [13, 27]. The higher the symptom score in the mice is, the more severe the allergic rhinitis (AR) symptoms are.

2.14. Histological Observation

After anesthesia, the mice were euthanized, and the liver, lungs, kidneys, skin, duodenum, nasal cavity, spleen, lymph nodes, and thymus were removed and fixed with formaldehyde [27]. The nasal bone was decalcified with a decalcification solution. All paraffin‐embedded sections were subsequently subjected to H&E or PAS staining to observe tissue morphology. The spleen, lymph nodes, and thymus were treated with an anti‐Foxp3 antibody (1:1000; Abcam; ab253297) for immunohistochemical staining. The counting method was as previously described [13, 27].

2.15. Determination of Serum Cytokine Levels in Mice

Changes in the concentrations of several cytokines were detected in mouse serum using OVA‐IgE (Bioswamp; MU30065), IL‐4 (MU30385), IL‐5 (MU30011), IL‐13 (MU30012), IFN‐γ (Pyram; PM16903), IL‐10 (PM16842), and IL‐17A (PM16822) kits according to the manufacturers’ instructions.

2.16. Bioinformatic Analysis

Foxo1‐associated transcription factors were computationally predicted through integrated analysis of multiple bioinformatic databases, including (1) the NCBI Gene database (https://www.ncbi.nlm.nih.gov/gene), (2) JASPAR (v2022; https://jaspar.genereg.net/) for transcription factor binding site prediction, (3) the UCSC Genome Browser (GRCh38/hg38 assembly; http://genome‐asia.ucsc.edu/) for genomic context analysis, and (4) GEPIA2 (http://gepia.cancer‐pku.cn/) for expression correlation analysis.

2.17. Chromatin Immunoprecipitation (ChIP)

ChIP assays were performed according to the established protocol [28] to determine whether RREB1 could bind to the promoter region of Foxo1. ChIP‐enriched DNA levels in the specific antibody‐precipitated samples were quantitatively compared with those in the normal rabbit IgG control group through input DNA normalization. The primer sequences are detailed in Table S5.

2.18. Luciferase Reporter Assay

The human Foxo1 promoter region was PCR‐amplified using primers (F: 5′‐GAAAACATTAATTCCCCACGTCGTTCAGC‐3′; R: 5′‐CGTGGGGAATTAATGTTTTCTTTCGCTGCGATC‐3′) and cloned and inserted into the BglII/HindIII sites of the pGL3‐basic vector (Promega, Madison), generating pGL3‐phFoxo1. Bioinformatic analysis was conducted to predict the binding sites of RREB1 within the Foxo1 promoter. Using fusion PCR with mutagenic primers (F: 5′‐AAAGAGAAAACCTCCCCGTGGAAAACCGG‐3′; R: 5′‐CACGGGGAGGTTTTCTCTTTCACACACTCACCTCC‐3′), we created a deletion mutant (pGL3‐phFoxo1‐del) lacking the RREB1 binding site. For the reporter assays, MoT cells were cotransfected with 100 ng of either pGL3‐phFoxo1 or pGL3‐phFoxo1‐del along with 1 ng of pRL‐TK‐Luc (internal control). The cells were harvested and lysed with 100 µL of passive lysis buffer (Promega, E1941. After cell debris was removed by centrifugation at 3000 × g for 5 min, the supernatant was detected with a single‐mode SpectraMax microplate reader in accordance with the manufacturer's (Molecular Devices) instructions.

2.19. Co‐IP Assay

The interaction between the intracellular segment of Notch2 (NICD2) and the transcription factor RREB1 was analyzed using coimmunoprecipitation (co‐IP). MoT cells were lysed in RIPA lysis buffer supplemented with protease inhibitor cocktail. For each reaction, 2 mg of whole‐cell lysate was precleared with 30 µL of protein G magnetic beads (Abcam, ab174816) for 1 h at 4°C. The precleared lysates were then incubated with 2 µg of either anti‐NICD2 antibody (Gene Tex, GTX101593) or isotype control IgG (Abcam, ab172730) overnight at 4°C with gentle rotation, followed by the addition of fresh protein G beads for 2 h. The immunocomplexes were subsequently washed five times with lysis buffer, eluted in 2× Laemmli buffer, and analyzed by Western blotting using anti‐RREB1 antibody (Abcam, ab113287).

2.20. GST Affinity‐Isolation Assay

The GST‐NICD2 and HA‐REBB1 genes were first cloned and synthesized, and then inserted into pGEX‐6P‐1. After ligation, transformation into recipient cells, and clonal expression, protein samples containing 500 µg of GST (control group) or GST‐NICD2 (experimental group) were added to glutathione‐agarose resin (Thermo Scientific, 16100) and mixed for 3 h. Five hundred micrograms of HA‐NICD2 protein was added to the control and experimental groups and then mixed overnight. The two groups of samples were centrifuged, and an appropriate amount of protein loading buffer was added, and then incubated at 100°C for 5 min. Finally, WB experiments were performed with anti‐GST (Abcam, ab111947) and anti‐HA (Abcam, ab236632) antibodies.

2.21. Cellular Immunofluorescence Colocalization Assay

After fixation, cells were incubated overnight at 4°C with the following primary antibodies: anti‐GSDMD‐N (1:200; Affinity, DF13758), anti‐Foxo1 (1:200; Servicebio, GB11286‐1), anti‐p‐ASC (1:200; Affinity, AF3515), anti‐NICD2 (1:200; Affinity, AF5296) and anti‐RREB1 (1:200; Proteintech, CL488‐84048). After being washed with PBS, the cells were incubated for 1 h at room temperature with species‐matched secondary antibodies labeled with CY3 (red fluorescence) and 488 (green fluorescence). Nuclei were counterstained with DAPI (5 mg/mL) for 15 min. Coverslips were mounted and imaged using a confocal microscope (Nikon Eclipse ci, Japan).

2.22. Statistical Analysis

Statistical analyses were systematically conducted using dedicated platforms: SPSS v20.0 (IBM Corp.) and GraphPad Prism v9.0 for general analyses, FlowJo v10.8 for flow cytometry data, and ImageJ for Western blot quantification. Prior to testing, the data underwent rigorous preprocessing, with normality assessed via the Shapiro–Wilk test and variance homogeneity evaluated by Levene's test. Conforming datasets (presented as the mean ± SD) were then analyzed using specific comparative methods: unpaired two‐tailed t tests for two‐group comparisons, one‐way ANOVA with Tukey's post hoc test for multigroup analyses, and Pearson's correlation for continuous variable relationships. The threshold for statistical significance was set at two‐sided p < 0.05. The corresponding sample sizes (n) for each experiment are explicitly documented in the figure legends.

3. Results

3.1. Notch2+ aTregs Exhibit Immunomodulatory Properties, Whereas Notch2⁻ aTregs Undergo Pyroptosis

AR patients exhibited distinct cytokine alterations, with elevated IL‐4 and IL‐6 levels and reduced IL‐2 and IL‐10 levels (Figure S1A, P<0.05), whereas TNF‐α/IFN‐γ levels remained unchanged. Flow cytometry revealed an increase in Th2 cells and a decrease in Th1 cells (Figure S1B, C, P<0.05), whereas the frequencies of Th17/Th9/Tfh cells and total CD4+ T cells were comparable to those in the control group. These findings demonstrate a characteristic Th1/Th2 imbalance, confirming the presence of Th2‐hyperactive immunopathology in AR.

Comparative analysis revealed significant clinical and immunological alterations in AR patients versus controls. AR patients had elevated VAS scores and serum IgE levels (Figure 1A, P<0.05), along with Treg cell dysregulation characterized by a reduced total Treg number (Figure 1B, P<0.05), a decreased aTreg number but increased rTreg number (Figure 1C, P<0.05), and increased aTreg death, as confirmed by electron microscopy. Additionally, these patients had significantly more pyroptotic aTregs than the controls did (Figure 1D, E, Figure S2C; P<0.05).

FIGURE 1.

Clinical parameters, Treg characterization, and Notch signaling in allergic rhinitis (AR). (A) Visual analog scale (VAS) scores and serum IgE levels in the control and AR groups. (B) Circulating Treg frequencies. (C) Proportions of aTreg and rTreg subsets. (D) Cell death rate of aTregs measured by flow cytometry. (E) Representative scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images of aTreg cell membranes. (F) Protein expression of Notch1, Notch2, and GSDMD‐N in aTregs and rTregs. (G) Frequency of Notch2‐positive Tregs. (H) Pearson correlation analysis between Notch2⁺ aTreg counts and clinical parameters. Data are presented as bar graphs with individual data points (control, n = 41; AR, n = 46). Each symbol in the correlation plots (H) represents an individual patient.

Moreover, compared with those in control cells, the Notch2 mRNA expression in both aTregs and rTregs from AR patients was significantly lower (Figure S2D, P<0.05), whereas the Notch1, Notch3, and Notch4 mRNA levels remained unchanged (Figure S2D, P>0.05). At the protein level, aTregs in the AR group exhibited significantly reduced Notch2 but elevated GSDMD‐N expression (Figure 1F, P<0.05), with no alteration in Notch1 expression. Similarly, the expression of Notch2 in rTregs was decreased in the AR group (Figure 1F, P<0.05), whereas the expression of Notch1 and GSDMD‐N did not significantly differ between the groups (P>0.05). These results indicate that Notch2 deficiency promotes GSDMD‐N‐mediated susceptibility to pyroptosis in aTregs.

Flow cytometry confirmed the significant depletion of Notch2+ aTregs in AR patients (Figure 1G, P<0.05). The number of these cells was strongly negatively correlated with the levels of disease severity markers, including the VAS scores, serum IgE level, Th2 IL‐4, and the number of Th2 cells (Figure 1H, Figure S2E; P<0.001) but was positively correlated with the total Treg and IL‐10 levels (Figure 1H, P<0.001). Notably, the strongest association was observed between the number of Notch2+ aTregs and the IL‐10 level (Figure 1H, r = 0.6456; P<0.0001). No significant correlations were detected between the number of Notch2+ aTregs and the levels of proinflammatory cytokines (IL‐6, TNF‐α, and IFN‐γ) or the size of other Th‐cell subsets (Th1/Th9/Th17/Tfh cells; Figure S2E; P>0.05). Further stratification of AR patients by the sensitization profile revealed consistently lower Notch2⁺ aTreg levels across all subgroups (mite‐only, mite plus food, and mite plus grass) than those in the control group (Figure S2G, P<0.0001), with no statistically significant differences among the AR subgroups themselves (P > 0.05).

These findings indicate that Notch2+ aTregs act as key immunomodulators, while Notch2‐ aTregs are predisposed to pyroptosis. The depletion of circulating Notch2+ aTregs appears to be central to the pathogenesis of AR, driving Th2 polarization, possibly via impaired IL‐10 regulation.

3.2. Notch2 was Specifically Knocked out in Tregs, Leading to the Induction of Pyroptosis in these Cells and Impairing Their Immunoregulatory Function

To investigate the role of Notch2 signaling in Tregs, we generated Treg‐specific Notch2 knockout mice by crossing Foxp3^Cre‐YFP mice with Notch2^flox/flox mice (designated the CKO group), with Foxp3^{Cre, YFP} mice serving as controls (Figure 2A). Flow cytometric analysis revealed that Notch2 deficiency in Tregs resulted in a significant reduction in splenic Treg and aTreg populations, a marked increase in rTregs, and a substantial decrease in Notch2 expression in Tregs (Figure 2B, P<0.05). Immunohistochemical staining revealed comparable Foxp3+ cell numbers in the thymus but significant reductions in the spleen and lymph nodes of CKO mice compared with those of controls (Figure S3A, P<0.05). Phenotypically, CKO mice developed spontaneous multiorgan inflammation characterized by hepatocyte hydropic degeneration and disruption of nasal mucosal integrity (Figure 2C, Figure S3B; P<0.05).

FIGURE 2.

Phenotypic and molecular characterization of Treg‐specific Notch2 knockout mice. (A) PCR genotyping for Treg‐specific Notch2 knockout. (B) Frequencies of splenic Tregs, aTregs, rTregs, and Notch2⁺ Tregs. (C) H&E staining depicting inflammatory infiltration in multiple organs. (D) Assessment of the purity of the sorted Treg cells; n = 4. (E) SEM images illustrating aTreg cell membrane morphology; n = 3. (F) Flow cytometric quantification of annexin V⁺7AAD⁺ dead aTregs; n = 3. (G) Heatmap of the results of the RNA‐seq analysis showing differentially expressed genes in aTregs. (H) Western blot analysis of the expression of Notch2, Foxo1, NLRP3, p‐ASC, and GSDMD/GSDMD‐N in aTregs. (I) Serum cytokine profiles of control and conditional knockout (CKO) mice. Data are presented as bar graphs with individual data points; n = 6.

Compared with CD4+YFP‐ non‐Tregs, flow‐sorted CD4+YFP+ Tregs exhibited significantly greater Foxp3 expression (Figure 2D, P<0.05). Ultrastructural analysis of sorted aTregs revealed distinctive pore formation on cell membranes in CKO mice (Figure 2E, P<0.05), accompanied by an increased proportion of annexin V+7AAD+ double‐positive cells (Figure 2F, P<0.05), indicative of pyroptotic cell death.

Transcriptomic profiling of aTregs revealed that the expression of Notch2 and Foxo1 was downregulated in the CKO group (P<0.05), whereas that of the pyroptosis‐related genes GSDMD and ASC was upregulated (P<0.05). However, the expression of NLRP3 inflammasome components (NLRP3 and caspase‐1) and other cell death markers (GSDMA, DFNA5, caspase‐3, and BAX) remained unchanged (Figure 2G). These findings were corroborated at the protein level, with aTregs in the CKO group showing reduced expression of Notch2 and Foxo1 (P<0.05) and elevated expression of GSDMD‐N and phosphorylated ASC (P<0.05) but unaltered NLRP3 expression (Figure 2H, Figure S3C).

Serum cytokine analysis revealed decreased levels of the Treg cytokine IL‐10 (P<0.05) and increased levels of Th1 (IFN‐γ) and Th2 (IL‐4/5/13) cytokines in CKO mice (Figure 2I, P<0.05), with no significant change in IL‐17A levels. Collectively, these data establish that Notch2 deficiency induces GSDMD‐mediated pyroptosis in aTregs, compromising their immunosuppressive capacity. This loss of regulatory function results in uncontrolled Th1/Th2 responses and spontaneous inflammatory pathology.

3.3. HDM Induces Dose‐Dependent Pyroptosis in shNICD2 Treg‐Like MoT Cells

Ho et al. [31, 32] established a Treg‐like cell line, termed MoT cells, which exhibit high expression of immunosuppressive markers and functional properties comparable to those of conventional Tregs. To validate the Treg‐like phenotype of MoT cells, we assessed Foxp3 expression via Western blotting. Consistent with previous reports, MoT cells demonstrated significantly greater Foxp3 levels than Jurkat (T‐cell leukemia) and THP‐1 (monocytic) cells (Figure 3A, P < 0.05), confirming their suitability as a Treg model.

FIGURE 3.

NICD2 modulates Foxp3 expression and HDM‐induced pyroptosis in MoT cells. (A) Foxp3 protein levels across different cell lines. (B) NICD2 expression in oeNICD2‐ and shNICD2‐transduced MoT cells. (C) Foxo1 mRNA levels measured by qRT–PCR. (D, E) Western blot analysis of pyroptosis‐related proteins following HDM stimulation. (F) Analysis of GSDMD‐N expression in HDM‐stimulated shNICD2 MoT cells alongside a positive control for pyroptosis (LPS+nigericin macrophages) and following disulfiram inhibition. (G) Cell death rates and p‐ASC speck formation upon HDM stimulation. (H) Effect of various cell death inhibitors on shNICD2 MoT cells. Data are presented as bar graphs with individual data points; n = 3; ns, not significant (p > 0.05).

To investigate the role of Notch2 signaling in pyroptosis regulation, we generated two modified MoT cell lines, namely, oeNICD2 MoT cells, which exhibit stable overexpression of NICD2, and shNICD2 MoT cells, which are characterized by stable knockdown of NICD2 (Figure 3B, P < 0.01). Notably, NICD2 overexpression upregulated Foxo1 mRNA levels, whereas NICD2 knockdown downregulated Foxo1 (Figure 3C, P < 0.01), suggesting that Foxo1 is regulated by a Notch2‐dependent mechanism.

Upon house dust mite (HDM) stimulation, shNICD2 MoT cells exhibited progressive increases in the expression of pyroptosis‐related markers, including elevated p‐ASC, caspase‐1 p20, and GSDMD‐N expression (Figure 3D, Figure S4A; p < 0.01); a dose‐dependent increase in the number of annexin V+7‐AAD+ double‐positive cells; a distinct pyroptotic morphology featuring membrane pore formation, structural collapse, and loss of membrane integrity; and an increased number of p‐ASC specks (Figure 3G, P < 0.05). In contrast, oeNICD2 MoT cells showed no significant changes in pyroptotic markers or morphology upon HDM exposure (Figure 3E, Figure S4B; p > 0.05).

Furthermore, the GSDMD‐N band generated in HDM‐stimulated cells was similar to that observed in pyroptosis‐positive controls (LPS+nigericin‐treated macrophages), confirming GSDMD cleavage. This band was markedly attenuated by disulfiram, a specific GSDMD inhibitor (Figure 3F, Figure S4C; p < 0.05). In shNICD2 MoT cells, disulfiram significantly suppressed cell death, whereas the necroptosis inhibitor necrostatin‐1 had no significant effect, and the apoptosis inhibitor Z‐VAD‐FMK only partially inhibited death (Figure 3H, Figure S4D; p < 0.05). Collectively, these results demonstrate that HDM stimulation promotes dose‐dependent, GSDMD‐mediated pyroptosis in shNICD2 MoT cells.

3.4. NICD2‐Foxo1 Signaling Attenuates Pyroptosis via SYK/JNK‐ASC Pathway Suppression

RNA‐seq and qRT–PCR confirmed the significant downregulation of Foxo1 mRNA expression following Notch2 knockdown (Figures 2G and 3C), indicating that NICD2 is a positive regulator of Foxo1 transcription. Western blotting revealed reduced NICD2 and Foxo1 protein levels in HDM+shNICD2 cells compared with those in control cells (p < 0.05), accompanied by increased SYK, p‐SYK, JNK, p‐JNK, p‐ASC, and GSDMD‐N expression. NICD2 overexpression (HDM+oeNICD2) reversed these effects, with elevated Foxo1 expression and a concomitant reduction in SYK/JNK pathway activation. Pharmacological Foxo1 inhibition (AS1842856) in oeNICD2 cells abolished this protective effect, with restored SYK/JNK signaling despite the maintenance of NICD2 expression (Figure 4A, Figure S5A).

FIGURE 4.

Notch2 signaling inhibits GSDMD‐mediated pyroptosis in Tregs. (A) Representative Western blots and quantification of NICD2, Foxo1, SYK/p‐SYK, JNK/p‐JNK, p‐ASC, and GSDMD‐N expression under different Notch2/foxo1 signaling conditions. (B) Quantitative analysis of cell death rates and p‐ASC speck‐positive cells. (C) Representative SEM and TEM images depicting the characteristic pyroptotic morphology of Treg cell membranes.

In addition, cell death assays and p‐ASC speck analysis revealed a significant increase in pyroptosis in the HDM+shNICD2 group compared with the control group (p < 0.05), marked protection against pyroptosis in HDM+oeNICD2 cells, and a reversion to the pyroptotic phenotype upon Foxo1 inhibition (Figure 4B). Ultrastructural analysis (SEM/TEM) revealed prominent pyroptotic features (membrane pores and cellular swelling) in the HDM+shNICD2 group, protection against pyroptosis in the HDM+oeNICD2 group, and re‐emergence of pyroptotic morphology upon treatment with AS1842856 (Figure 4C, Figure S5B, C). Notably, in Notch2‐deficient cells, Foxo1 inhibition with AS1842856 did not significantly alter GSDMD‐N expression (Figure S5D), indicating that the protective effect of Foxo1 is contingent upon Notch2 expression.

These findings are consistent with those of previous reports indicating that Foxo1 modulates SYK/JNK signaling [33, 34, 35, 36] and further demonstrate that the NICD2‐mediated upregulation of Foxo1 expression inhibits ASC phosphorylation, a critical checkpoint in pyroptosis, through the suppression of SYK/JNK signaling. The complete rescue of pyroptosis in oeNICD2 cells and its reinstatement upon Foxo1 inhibition confirm that Foxo1 is an essential downstream effector of the antipyroptotic function of NICD2.

3.5. RREB1 Functions as a Transcription Factor for Foxo1, and NICD2 Interacts with RREB1

Through comprehensive bioinformatic analysis, we predicted that RREB1 may serve as a transcription factor for Foxo1, with four potential binding sites (P1–P4) identified in the Foxo1 promoter region (Figure 5A, C). To validate this prediction, we first examined the correlation between RREB1 and Foxo1 expression in lymphocytes and observed a significant positive association (Figure 5B, P < 0.05). Subsequent chromatin immunoprecipitation (ChIP–PCR) assays confirmed that RREB1 specifically binds to the P2 region (829–848 bp) of the Foxo1 promoter (Figure 5D, P < 0.05).

FIGURE 5.

RREB1 functions as a transcription factor for Foxo1, and NICD2 interacts with RREB1. (A) Bioinformatic prediction of RREB1 binding sites within the Foxo1 promoter region. (B) Correlation between RREB1 and Foxo1 expression levels in lymphocyte populations. (C) Schematic diagram of four predicted RREB1 binding sites (P1–P4) on the Foxo1 promoter. (D) ChIP–PCR validation of RREB1 enrichment at the Foxo1 P2 site. (E, F) Luciferase reporter assays assessing RREB1‐dependent Foxo1 promoter activity. (G) Immunofluorescence staining showing nuclear colocalization of NICD2 and RREB1. (H) Coimmunoprecipitation (co‐IP) analysis of the NICD2‐RREB1 protein interaction. (I) Results from the results of the GST pull‐down assay confirm direct binding between NICD2 and RREB1. Data are presented as the mean ± SD from three independent experiments; n = 3; ns, not significant (p > 0.05).

Luciferase reporter assays further revealed that the relative luciferase activity in the AdRREB1 + p‐Foxo1 group was significantly greater than that in the p‐Foxo1, m‐pFoxo1, and AdRREB1 + m‐pFoxo1 groups (Figure 5E, P < 0.05). Additionally, compared with pLVX‐Puro + FOXO1‐PGL3‐Basic, pLVX‐RREB1‐Flag + Foxo1‐pGL3 resulted in significantly greater relative luciferase activity (Figure 5F, P < 0.05). These results confirm that RREB1 acts as a transcription factor for Foxo1, enhancing its transcription by binding to specific promoter regions.

We next explored whether NICD2, the intracellular domain of Notch2, modulates RREB1 activity. Immunofluorescence colocalization assays revealed strong colocalization of NICD2 and RREB1 in MoT cells (Pearson's Rr = 0.98887) (Figure 5G). Coimmunoprecipitation (co‐IP) revealed a direct protein–protein interaction between NICD2 and RREB1 in MoT cells (Figure 5H). This interaction was further corroborated by the results of the GST affinity isolation assays (Figure 5I). These data provide compelling evidence that NICD2 physically associates with RREB1, suggesting a potential regulatory mechanism involving these two proteins.

3.6. NICD2 Modulates the Subcellular Distribution of RREB1 in MoT Cells

To investigate whether NICD2 influences the intracellular distribution of RREB1, we first performed immunofluorescence colocalization assays. In control cells (shNC group), RREB1 was predominantly localized in the nucleus. In contrast, NICD2 knockdown (shNICD2 group) resulted in a marked shift in RREB1 localization to the cytoplasm, suggesting that NICD2 depletion impaired RREB1 nuclear translocation. Conversely, NICD2 overexpression (oeNICD2 group) increased the nuclear accumulation of RREB1, indicating that elevated NICD2 levels promote RREB1 nuclear import (Figure 6A, Figure S6A, B, p<0.0001).

FIGURE 6.

NICD2 modulates the subcellular distribution of RREB1 in MoT cells. (A) Immunofluorescence analysis of RREB1 subcellular localization under different Notch2 signaling conditions. (B, C) Western blot quantification of RREB1 protein levels in cytoplasmic (B) and nuclear (C) fractions. (D) Total RREB1 expression in shNICD2 and oeNICD2 MoT cells. Data are presented as bar graphs with individual data points; n = 3.

Consistent with the immunofluorescence data, compared with the controls, NICD2 knockdown significantly increased cytoplasmic RREB1 levels while reducing nuclear RREB1 accumulation (Figure 6B, C, P < 0.05). Notably, total cellular RREB1 expression remained unchanged (Figure 6D, P > 0.05), confirming that NICD2 depletion specifically affects RREB1 localization rather than overall expression. Conversely, NICD2 overexpression had the opposite effect, significantly decreasing cytoplasmic RREB1 expression but increasing nuclear RREB1 expression (Figure 6B, C, P < 0.05) without altering total RREB1 expression (Figure 6D, P > 0.05).

3.7. In the AR Model, the Number of Notch2+ aTregs Decreased, and Notch2– aTregs Underwent GSDMD‐N‐mediated Pyroptosis

In this study, an AR mouse model was established according to the flowchart shown in Figure 7A. Compared with control mice, AR mice exhibited significantly elevated nasal symptom scores (Figure 7B, P<0.05), accompanied by increased serum OVA‐specific IgE levels (Figure 7C, P<0.05). Histopathological examination revealed pronounced nasal inflammation characterized by substantial eosinophilic infiltration (Figure 7D, Figure S7A; P<0.05), which is consistent with typical allergic rhinitis pathology. Flow cytometric analysis demonstrated significant alterations in Treg subpopulations during disease progression (Figure 7E–I). Notably, we observed a time‐dependent reduction in total Treg numbers (Figure 7E, P<0.05), progressive depletion of aTreg populations (Figure 7F, I; P<0.05), and concurrent expansion of rTregs (Figure 7F, I; P<0.05).

FIGURE 7.

Allergic rhinitis progression drives Notch2+ aTreg depletion and GSDMD‐mediated pyroptosis in Notch2⁻ aTreg cells. (A) Schematic of the OVA‐induced AR mouse model. (B) Nasal symptom scores. (C) Serum OVA‐specific IgE levels. (D) H&E staining of nasal mucosa. (E) Total splenic Treg frequencies. (F) Proportions of splenic aTreg and rTreg subsets. (G, H) Notch2 expression in sorted aTregs (G) and rTregs (H). (I) Statistical summary of Treg subpopulations and Notch2 expression. (J) Western blot analysis of Notch2, Foxo1, p‐ASC, and GSDMD‐N protein levels in aTregs vs. rTregs. Data are presented as bar graphs with individual data points; n = 6; ^* p < 0.05; ns, not significant.

Additionally, distinct patterns of Notch2 expression were detected in different Treg subsets. In aTregs, we observed a progressive downregulation of Notch2 expression (Figure 7G, I; P<0.05), whereas in rTregs, stable Notch2 levels were maintained throughout AR disease progression (Figure 7H, I; P>0.05). Western blot analysis demonstrated cell–type–specific molecular changes: aTregs exhibited coordinated downregulation of the expression of Notch2 and Foxo1, concomitant with upregulation of the expression of the pyroptotic markers p‐ASC and GSDMD‐N (Figure 7J, Figure S7B; P<0.05). In contrast, rTregs showed no significant alterations in the expression of these molecular markers (Figure 7J, Figure S7B; P>0.05). These findings collectively demonstrate that AR progression leads to selective depletion of Notch2+ aTreg populations and upregulation of GSDMD‐N‐mediated pyroptotic cell death in Notch2‐ aTreg populations, whereas rTregs remain molecularly and functionally stable.

3.8. Adoptive Transfer of Notch2+ aTregs Significantly Reduces Th2 Inflammatory Responses in AR Mice

We established an aTreg adoptive transfer model in AR mice using previously described methods (Figure 8A). Compared with those in the control group, the nasal symptom score, serum OVA‐IgE level, nasal eosinophil count, PAS‐positive cell count in the nasal mucosa, and aTreg pyroptosis, IL4, IL‐5, and IL‐13 levels were significantly greater in the mice in the AR group. Notably, adoptive transfer of Notch2+ aTregs markedly attenuated clinical symptoms (nasal scores), OVA‐IgE levels, tissue inflammation (eosinophil counts and PAS+ cells), aTreg pyroptosis, and Th2 cytokine levels (IL‐4/5/13). The empty vector‐transferred aTregs failed to produce a statistically significant improvement in key disease parameters—including clinical symptom scores, OVA‐specific IgE levels, eosinophil counts, PAS‐positive cell numbers, and Th2 cytokine levels—compared to the untreated AR group (Figure 8B–F, P<0.05).

FIGURE 8.

Adoptive transfer of Notch2+ aTregs significantly reduces Th2 inflammatory responses in AR mice. (A) Schematic of the aTreg adoptive transfer experiment and treatment timeline. (B) Nasal symptom scores. (C) Serum OVA‐specific IgE levels. (D) Representative H&E and PAS staining of nasal mucosa tissue sections. (E) SEM images showing the membrane integrity of aTregs. (F) Quantitative analysis of inflammatory parameters and Th2 cytokine levels. Data are presented as bar graphs with individual data points; n = 6.

In aTregs isolated from AR mice, RREB1 was predominantly localized in the cytoplasm. Adoptive transfer of Notch2‐overexpressing aTregs promoted the nuclear translocation of RREB1, upregulated Foxo1 expression, and reduced GSDMD‐N levels, whereas empty vector transfer failed to elicit these effects (Figure S8A, B). Furthermore, AR mice that received Notch2⁺ aTregs exhibited no significant inflammation in the liver, kidneys, skin, or small intestine (Figure S9), supporting the safety of this cellular intervention. Together, these results underscore the functional role of the Notch2/RREB1/Foxo1 axis in modulating aTreg homeostasis and its therapeutic relevance in allergic rhinitis.

4. Discussion

Our study elucidates the critical role of Notch2 in regulating activated Treg (aTreg) fate and function in allergic rhinitis (AR) pathogenesis. An imbalance between Th1/Th2 responses and Treg dysfunction has been well established in AR, with allergen immunotherapy (AIT) demonstrating efficacy through Treg induction [2, 3]. Building upon previous reports [37], we identified a specific reduction in Notch2+ aTreg populations in AR patients that correlated with decreased IL‐10 levels and enhanced Th2 responses. The therapeutic potential of these findings was confirmed by adoptive transfer experiments showing that Notch2+ aTregs effectively ameliorated Th2‐mediated inflammation in AR models, likely through both cellular expansion and cytokine‐mediated suppression [22]. Our study also provides direct evidence that Notch2⁺ aTreg transfer does not induce off‐target inflammation or disrupt tissue homeostasis under our experimental conditions, supporting its safety profile in this therapeutic context.

The Notch signaling pathway, particularly through the Notch1 and Notch2 isoforms, plays a pivotal role in allergic diseases [38]. While our prior work revealed the involvement of these genes in AR development [13, 30], the current study revealed the unique role of Notch2 in maintaining aTreg homeostasis. Unlike the expression of Notch1/3/4, the expression of Notch2 was specifically reduced in both aTregs and rTregs from AR patients. These observations were further supported by data from conditional knockout models in which Notch2 deficiency led to peripheral Treg depletion and multiorgan inflammation. We subsequently found that the number of aTregs that play an immunosuppressive role in tissue inflammation was significantly reduced in mice in which Notch2 was specifically knocked out in Tregs, confirming that Notch2 plays an important role in maintaining the number of aTregs.

Mechanistically, we revealed a novel pathway linking Notch2 deficiency to aTreg pyroptosis. Notch2 knockout aTregs exhibited characteristic pyroptotic features (membrane pores and cellular swelling), decreased Foxo1 expression, and increased p‐ASC/GSDMD‐N levels. These findings were replicated in AR patient‐derived aTregs and in vitro models, in which dust mite allergens exacerbated pyroptosis in Notch2‐deficient cells. Importantly, we demonstrated that the protective effect of Notch2 is mediated through Foxo1, as pharmacological Foxo1 inhibition abolished the ability of Notch2 to suppress pyroptosis.

This study revealed that after Notch2 was knocked out in aTregs, the level of Foxo1 significantly decreased. This result aligns with the findings of Fujimaki et al. [39], which indicate that Notch2 regulates Foxo1 expression at the upstream level but not at the transcriptional level. Notably, no binding between the intracellular segment of Notch2 (NICD2) and either Foxo1 or AKT kinase upstream of Foxo1 was detected via immunoprecipitation [39]. Integrated bioinformatic screening revealed RREB1 as an upstream transcriptional regulator of Foxo1, showing a strong correlation with expression in lymphocytes. Computational prediction and experimental validation revealed that RREB1 binds specifically to the Foxo1 promoter P2 site to increase transcriptional activity. Mechanistic investigation revealed NICD2‐mediated nuclear translocation of RREB1. Thus, in contrast to the study by Fujimaki et al. [39], this research elucidates how Notch2 regulates Foxo1 transcription: NICD2 interacts with the upstream transcription factor RREB1 of Foxo1, promoting its nuclear translocation and binding to the Foxo1 promoter, thereby enhancing Foxo1 transcription. Downstream of Foxo1, we elucidated a complete signaling cascade in which Foxo1 suppression activates SYK/JNK signaling, leading to ASC phosphorylation, NLRP3 inflammasome assembly, and ultimately GSDMD‐mediated pyroptosis [15, 17, 33, 34, 35]. Central to this cascade is our finding that Foxo1 itself is a potent suppressor of the SYK/JNK/ASC pathway. This aligns with established mechanisms whereby Foxo1 directly represses NLRP3 inflammasome activation [36, 40, 41, 42, 43]; however, the molecular mechanism by which NICD2 facilitates RREB1 nuclear translocation requires further validation. While we propose that NICD2 acts as a nuclear chaperone for RREB1 via its functional NLS [44], definitive evidence through structural mapping and NLS mutagenesis is needed.

Despite identifying the role of the Notch2/RREB1/Foxo1–pyroptosis axis in aTregs, our study has several limitations. First, the specific dependency of aTregs versus rTregs on Notch2 suggests the presence of distinct regulatory programs in these subsets [4, 5, 6, 7, 8, 9], although the underlying mechanisms require further exploration. Second, our experimental approaches did not permit precise titration of Notch2 expression levels; future studies using inducible or graded expression systems should define the quantitative thresholds governing aTreg fate decisions between immunoregulation and pyroptosis. Finally, the expression profiles of canonical Notch ligands (e.g., JAG1/2 and DLL1/4), which critically regulate receptor activation [39, 45, 46], were not characterized, and peripheral eosinophil/neutrophil counts were not available for correlative analysis in our clinical cohort. Addressing these aspects in future work will strengthen the mechanistic and translational relevance of our findings.

In conclusion, our work establishes Notch2 as a molecular switch determining aTreg fate in AR. Notch2+ aTregs maintain a potent immunoregulatory capacity, whereas their Notch2‐ counterparts undergo inflammatory pyroptosis. These findings not only advance our understanding of the immunopathology of AR but also reveal multiple therapeutic targets, from Notch2+ aTreg therapy to small‐molecule modulators of the Notch2‐RREB1‐Foxo1 axis, offering new avenues for the development of AR treatments.

Author Contributions

Y. L. Q. and S. M. C. made the conception and design of the work, performed the experiments, analyzed the data, and drafted the work. S. M. C., Y. Y. Z., W. E. J., and S. X. made the conception, revised the manuscript, and acquired the funding. Y. G. K. revised the manuscript. H. Z., J. Y. L., H. M. F., Y. T. Z., and H. Y. W. performed the experiments. R. Y. and Y. Z. revised the manuscript and acquired the funding. All authors have read and agreed to the contents of the manuscript and consent to its publication.

Declarations Ethics approval

All experiments described herein were approved by the Institutional Animal Care and Use Committee of Renmin Hospital of Wuhan University (licence no.: WDRM20170310 and WDRM20231203B). This study was approved by the Ethics Committee of the Renmin Hospital of Wuhan University (ethical approval number: WDRY2023‐K143), and all patients signed written informed consent forms.

Conflicts of Interest

The authors declare no conflict of interest.

Supporting information

Supporting File: advs73368‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.