Gasdermin D-dependent neutrophil pyroptosis exacerbates Coxsackievirus A6-induced neuroinflammation through NET formation

Abstract

Background

Coxsackievirus A6 (CVA6) is a nonenveloped, single-stranded RNA virus linked to neurological complications. Emerging evidence suggests neutrophil pyroptosis drives inflammation. However, the role of neutrophil pyroptosis in CVA6 pathogenesis remains unexplored.

Methods

Ten-day-old wild-type (WT), Caspase-1 KO, and GSDMD KO mice were infected with a lethal dose of CVA6. For in vivo and in vitro studies, we used the caspase-1 inhibitor belnacasan, disulfiram, and anti-Ly6G antibody. We also generated neutrophil-specific PAD4-knockout mice (PAD4 Ne-KO) by deleting Padi4 under the S100A8 promoter. Post-infection, clinical scores, survival, and body weight were monitored. Brain tissues and bone marrow-derived neutrophils (BMDNs) were collected for analysis. Key methods included qPCR, Western blotting, histology/immunofluorescence, flow cytometry, and TEM to assess pyroptosis, inflammation, and immune cell infiltration. Findings were further validated using blood samples from HFMD patients.

Results

In this study, we investigated how the Caspase-1/GSDMD pathway mediates neutrophil extracellular trap (NET) release and drives CVA6-induced neuroinflammation. CVA6 infection increased neutrophil numbers in mouse brain and peripheral blood, along with elevated MPO-DNA—a NET marker. In BMDNs, degranulation and NET formation occurred by 24 hpi, accompanied by Caspase-1/GSDMD activation. Caspase-1 knockout prolonged survival and reduced GSDMD-N expression in brain neutrophils; pharmacological Caspase-1 inhibition decreased mature IL-1β and IL-18 in brain tissue and suppressed CVA6 replication in BMDNs. Together, in vitro and in vivo data indicate that Caspase-1/GSDMD activation and NETosis critically contribute to CVA6-induced brain injury. This was confirmed by GSDMD knockout or disulfiram-mediated GSDMD inhibition, both of which markedly reduced NET release and neuropathology. Notably, global neutrophil depletion worsened infection—suggesting a protective role—whereas neutrophil-specific PAD4 knockout improved survival. Clinically, GSDMD expression showed a significant positive correlation with NETosis markers in patient samples from CVA6-infected individuals.

Conclusion

These findings enhance understanding of enteroviral pathogenesis, identify GSDMD as a promising therapeutic target, and provide a novel framework for developing precision interventions that reduce excessive inflammation without impairing essential host defenses.

Graphical abstract

The proposed mechanism of CVA6-induced neuroinflammation: Following the invasion of the central nervous system by CVA6, neutrophil aggregation is induced, which triggers the activation of the classical pyroptosis pathway. During this process, Caspase-1 cleaves GSDMD, pro-IL-1β, and pro-IL-18 to produce active N-GSDMD, mature IL-1β, and IL-18, respectively. The pyroptosis mediated by N-GSDMD leads to the release of neutrophil extracellular traps (NETs) and is accompanied by a robust secretion of inflammatory cytokines. This cascade ultimately results in a cytokine storm and severe neuroinflammatory responses.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12974-026-03916-x.

Background

Coxsackievirus A6 (CVA6) is a type of nonenveloped, single-stranded RNA virus belonging to the genus Enterovirus within the Picornaviridae family [1]. As one of the most prevalent enteroviruses in the world, CVA6 seems to affect a wider range of children and causes more severe and prolonged illness than other enteroviruses [2]. In addition to its contribution to atypical cases of hand, foot, and mouth disease (HFMD) as well as respiratory illnesses, it also possesses the potential to induce severe complications [3]. These complications include myocarditis, encephalitis, and acute flaccid paralysis [4]. In recent years, CVA6 has frequently undergone genetic recombination and has been implicated in large-scale outbreaks worldwide [5–7]. The rising incidence of HFMD associated with CVA6 has emerged as a significant public health concern.

Our previous animal experiments have confirmed the neurotropic nature of CVA6 and its potential to cause neuroinflammation [8]. However, the precise mechanism through which CVA6 infection leads to neurological complications remains unclear. As the primary line of host defense against infections and a key regulator of both innate and adaptive immune cells, neutrophils play essential roles in the pathogenesis of CVA6 infection [9]. While neutrophils typically undergo apoptosis under homeostatic conditions, recent studies have demonstrated that they can also engage in alternative modes of regulated cell death, including pyroptosis and NETosis. Each of these alternative mechanisms of cell death has significant implications for inflammatory responses and the host’s defense against infections [10–12]. A mutually reinforcing positive feedback loop exists between neutrophil pyroptosis and the release of neutrophil extracellular traps (NETs), a phenomenon especially pronounced in inflammatory diseases. Recently, there has been a growing interest in the phenomenon of pyroptosis within neutrophils [13, 14]. The key executor protein of neutrophil pyroptosis—gasdermin D (GSDMD)—can be activated during NETosis, leading to the formation of plasma membrane pores, subsequent calcium ion influx, and activation of peptidyl arginine deiminase 4 (PAD4). This cascade drives histone citrullination and chromatin decondensation, thereby facilitating the release of NETs [15–18]. This suggests neutrophil pyroptosis and NETs release are not independent events but form a synergistic amplification network through shared molecular mechanisms (such as GSDMD, PAD4, Caspase-1), playing a dual role in inflammation and immune regulation. Such mechanisms have been proposed as significant pathological factors in CVA6 infection. Several studies have demonstrated that pyroptosis plays a significant role in the pathogenic mechanisms of various enteroviruses [19–21]. However, there is currently a lack of research investigating the upstream molecules involved in pyroptosis related to CVA6 infection. As a result, gaps remain in our understanding of the regulatory mechanisms governing neutrophil pyroptosis and its implications for CVA6 pathogenesis.

In the present study, we used Caspase-1 KO, GSDMD KO, Ly6G-DTR KI, neutrophil-specific PAD4 knockout mice and specific inhibitors, along with in vitro experiments, to investigate whether neutrophil pyroptosis contribute to CVA6-induced neuroinflammation. Meanwhile, we performed a comparative analysis of the levels of pyroptosis in peripheral blood leukocytes from mild and severe patients with HFMD. Finally, our findings indicate that GSDMD-dependent neutrophil pyroptosis significantly exacerbates CVA6-induced neuroinflammation through the formation of NETs. This study establishes a scientific foundation for the clinical management of neurological disorders associated with CVA6.

Materials and methods

Ethics statements

This study was carried out in line with the revised Declaration of Helsinki, and all experimental protocols involving animal and human specimens were approved by the Life Science Institutional Review Board of Zhengzhou University (permission no: ZZUIRB2025-190). Written informed consent was obtained in advance from the parents/legal guardians of all human participants.

Study participants

We conducted a case-control study involving pediatric patients of HFMD who were hospitalized at Henan Children’s Hospital from January 2018 and December 2023. The diagnosis of HFMD was based on the Guidelines for the Diagnosis and Treatment of Hand, Foot, and Mouth Disease (2018 edition) issued by the National Health Commission of the People’s Republic of China. Based on clinical manifestations, physical examination findings, and imaging results, CVA6 patients diagnosed with neurological complications were classified into the severe group (case, n = 137). Conversely, the remaining CVA6 cases were categorized as belonging to the mild group (control, n = 146). Plasma samples were collected using EDTA and sodium citrate as anticoagulants, then centrifuged at 3,000 rpm for 10 min at 4 ℃; following separation and erythrocyte lysis, peripheral blood leukocytes and plasma were stored at − 80 ℃. Demographic characteristics and clinical laboratory data for all participants are summarized in Supplementary Table 1. To reduce bias, we matched the severe group and the mild group by age and gender during the analysis of markers such as MPO, GSDMD, IL - 1β, and IL - 18.

Cells and CVA6 strain

Human rhabdomyosarcoma (RD) cells were obtained from the American Type Culture Collection and cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS). The CVA6 strain, identified by GenBank accession number OM179765, was previously described in our earlier study [8].

Animal infection experiments

The C57BL/6J wild-type (WT) mice and ICR mice utilized in this study were obtained from SiPeiFu Biotechnology Co., Ltd (Beijing). Gsdmd KO mice (C57BL/6NCya-Gsdmdcm1/Cya), C57BL/6NCya-Padi4em1flox/Cya (peptidyl arginine deiminase 4, PAD4^fl/fl), and S100A8 Cre (S100A8 is also known as migration inhibitory factor-related protein-8, associated with the activation and migration of neutrophils, C57BL/6JCya-S100a8em1Cre/Cya) mice were purchased from Cyagen (Suzhou) Biotechnology Co., Ltd. (Table S2). Caspase-1 KO mice (C57BL/6- CASP1tm201(-/-)/V) were purchased from Beijing Viewsolid Biotechnology Co., Ltd. Ly6g-DTR knock-in (KI) mouse strain (B6.129 S-Ly6g < tmX(DTR)> ) were provided by the Henan Key Laboratory of Immunology and Targeted Drugs (Table S3). To generate mice with deletion of Padi4 in neutrophils, PAD4^fl/fl mice were first crossed with S100A8-Cre/+ mice to obtain PAD4^fl/+; S100A8-Cre/+ offspring. These F1 mice were then backcrossed with PAD4^fl/fl mice to obtain PAD4^fl/fl; S100A8-Cre/+ mice (PAD4 Ne-KO). Littermate PAD4^fl/fl lacking S100A8-Cre mice were used as controls.

In this study, 10-day-old WT, Caspase 1 KO (Casp1-/-), GSDMD KO (Gsdmd-/-), Ly6G-DTR KI, and PAD4 Ne-KO mice were intraperitoneal (i.p.) inoculated with a lethal dose of CVA6 (10⁴ TCID₅₀ /mouse), and each group included 9∼12 mice. Mock-infected mice were inoculated with saline. For the intervention, 10-day-old ICR mice were i.p. inoculated with disulfiram (MCE, Cat: HY-B0240, 10 mg/kg) and belnacasan (MCE, Cat: VX-765, 10 mg/kg) at 2, 24, 48 and 72 h post infection (hpi) following a lethal challenge with CVA6. To deplete neutrophils 24–48 h prior to infection, 10-day-old WT mice received i.p. injections of either the anti-Ly6G neutralizing antibody (clone 1A8; 12.5 mg/kg) or a rat IgG2a isotype control antibody (12.5 mg/kg), administered once daily for two consecutive days before CVA6 infection.

The body weight, clinical manifestations, and survival rates of the mice were subsequently recorded daily until 15 days post-infection (dpi). The clinical scores were assessed in accordance with the methodologies outlined in our previous studies. At 3 and 5 dpi, the mice were euthanized with isoflurane anesthesia, and their brains were extracted for histopathological analysis and molecular biological experiments.

Animals were randomly allocated to groups (mock and infected), Due to the pronounced increase in clinical scores observed in CVA6-infected mice compared to the mock-infected controls, blinding was not feasible during the study.

Isolation of murine bone marrow-derived neutrophils (BMDNs)

Bone marrow single-cell suspensions were obtained from the femurs and tibias of mice following established protocols [22]. Neutrophils were purified using a mouse neutrophil isolation kit (Cat: TBD70907-10, Tianjin Haoyang Biological Products Co., Ltd., Tianjin, China), which employs biotin-conjugated monoclonal antibodies to label non-target cells (non-CD11b⁺ Ly-6G⁺ cells), followed by their removal via streptavidin-conjugated magnetic beads, thereby enabling the selective isolation of neutrophils. The purity of the isolated BMDNs was determined by FACS analysis of CD11b and Ly6G expression and found to be 95%. Cell viability was assessed utilizing the trypan blue exclusion method.

RNA extraction and quantitative PCR (qPCR)

Total RNA was extracted from brain tissues using RNAiso Plus (Takara Biomedical Technology (Beijing) Co., Ltd., China). The concentration and purity of the extracted RNA were assessed with a NanoDrop ND-2000 spectrophotometer (Thermo Fisher Scientific Co., Ltd., USA). Subsequently, cDNA was synthesized from the RNA samples using a reverse transcription kit (Yeasen BioTechnologies Co., Ltd., China). Regarding the qPCR reaction composition, the 10 µL reaction system consisted of the following components: 5 µL of SYBR Green Premix Ex Taq (TaKaRa, Japan), 0.2 µL of each forward and reverse primer (10 µM), 2 µL of cDNA template, and 2.6 µL of nuclease-free water. The concentrations of all reagents were carefully optimized to ensure efficient and specific amplification. The corresponding primer sequences used in this study are listed in Table S4.

Western blotting

Mice brains and BMDNs were homogenized in RIPA lysis buffer containing protease inhibitors. The homogenates were placed on ice for 1 h, then centrifuged at 12,000×g for 15 min at 4 °C. Protein concentrations were determined using a BCA kit. Proteins were separated by 8% or 12% SDS-PAGE and then transferred to PVDF membranes. Membranes were blocked with 5% skim milk, incubated with the primary antibodies, washed three times with TBST, and then incubated with the appropriate secondary antibody. Protein bands were visualized using ECL substrate with the Amersham Imager 600 imaging system (General Electric Co., Ltd. USA). Each band on the blot was quantified using Image J software.

Histology and immunofluorescence (IF) staining

At 5 dpi, brain tissues were harvested, rinsed with PBS, and fixed in 4% paraformaldehyde at room temperature overnight. Paraffin-embedded brain Sects.  (4–5 μm) were stained with hematoxylin and eosin (H&E) and Nissl staining or subjected to immunohistochemical (IHC) staining for the evaluation of brain pathology. Additionally, specific antibodies were employed for immunolabeling to assess pyroptosis. Images were captured using the 3DHISTECH Slide Viewer software, and each image was quantitatively analyzed using Image J software.

BMDNs were seeded in 6-well plates and subsequently infected with CVA6 at a multiplicity of infection (MOI) of 1 for a duration of 24 h or treated with PMA (Cat: P8139-1MG, Sigma-Aldrich) at 400 nmol/L for a duration 4 h. Cells were fixed, permeabilized, blocked with 5% BSA in PBST, incubated with primary then secondary antibodies, and mounted with DAPI. Paraffin-embedded tissues of 5 μm in thickness were dewaxed and washed three times with 1× PBS. Before being permeabilized with 1× PBS containing 0.5% Triton X-100, the sections were blocked for 30 min with goat serum at room temperature, incubated with primary then secondary antibodies, and mounted with DAPI. And the number of positive stained cells or integrated fluorescence signal was quantified by Image J software.

Determination of viral loads

The determination of viral loads in brain tissues and BMDNs was conducted according to our previous study [8].

Quantification of myeloperoxidase (MPO) and MPO-DNA

The concentration of MPO in brain tissue was quantified using the Myeloperoxidase Assay Kit (Nanjing Jiancheng Bioengineering Institute, China) following the protocols provided by the manufacturer. And the concentration of MPO-DNA in brain tissue was quantified using the Mouse MPO-DNA ELISA Kit (Cat: ZC-56424, ZCIBIO Co., Ltd., China).

Transmission electron microscopy (TEM)

BMDNs were inoculated in cell culture dish, and then the cells were infected with CVA6 (MOI = 1) for 24 h. Subsequently, the cell suspension was transferred to a centrifuge tube and centrifuged at low speed for 5 min. The supernatant was carefully removed, and the cell pellet was resuspended in electron microscope fixative. To ensure uniform fixation, cell clumps were gently dispersed by pipetting, followed by incubation at room temperature in the dark for 30 min. The fixed samples were then stored at 4 °C and transported under refrigerated conditions to Servicebio Biotech Co., Ltd. (Wuhan, China) for further processing, sectioning, and imaging using a transmission electron microscope (TEM, HT7800, Hitachi, Japan).

Analysis of brain cells by FACS

Single Cell preparation from the peripheral blood, lungs, or brains of the WT mice or Ly6G-DTR mice were prepared and stained with antibodies as described in our previous studies [23]. Samples were analyzed on an Agilent NovoCyte flow cytometer, and data were processed using NovoExpress software.

Antibodies

The following antibodies were used for Western blotting, IHC and IF staining (Table S5): anti-IL-1β, anti-IL-18, anti-MPO anti-β-actin (Proteintech Co., Ltd., China); anti-pro-Casp1 (Cell Signaling Technologies Inc., Danvers, Massachusetts, USA), anti-Cleaved-Casp1 (p20) (Affinity Biosciences Co., Ltd., USA); anti-GSDMD/-N (HUABIO Co., Ltd., China); anti-GFAP, anti-NeuN and anti-IBA1 (Servicebio Biotech Co. Ltd., Wuhan, China); anti-CitH3 (Abcam plc, Cambridge, UK); anti-CVA6 VP1 (GeneTex Inc., California, USA); anti-mouse immunoglobulin G (IgG) and anti-rabbit IgG (Proteintech Co., Ltd., China).

The antibodies used in FACS analysis were purchased from Biolegend Inc., CA, USA: FITC-conjugated Ly6G antibody (#128006), PerCP/Cy5.5-conjugated Ly-6G antibody (#127615), APC-conjugated CD3 antibody (#100312), PE-Cyanine7-conjugated CD11b antibody (#101215), Brilliant Violet 510-conjugated CD19 antibody (#115546). LIVE/DEAD Fixable Violet Dead Cell Stain Kit (Thermo Fisher Scientific Co., Ltd, Waltham, USA) was used to distinguish dead cells from living ones.

Statistical analysis

Data were analyzed using SPSS version 31.0. For measurement data, normality of distribution was assessed first: normally distributed data were presented as mean ± standard deviation (SD) and compared between two groups using the independent samples t-test; for non-normally distributed data, median and interquartile range were reported, and group differences were analyzed using non-parametric tests (e.g., Mann-Whitney U test). Count data were expressed as frequencies and percentages. Pearson correlation analysis was performed to assess linear relationships between continuous variables. Survival analysis was conducted using the Kaplan-Meier method, and any disparities in survival rates were evaluated with the log-rank test. For all analysis, difference was considered significant at a P < 0.05.

Results

The activation of the Caspase-1/GSDMD pathway and NETs formation in neutrophils are involved in CVA6-induced neuroinflammation

Based on the animal model established in our previous study [8], we observed that CVA6 infection resulted in neurological symptoms (Fig. S1A) and mortality (Fig. S1B) in mice. Viral replication, neuronal damage, and inflammatory cell infiltration were detected in brain tissues (Fig. S1C-E). The transcription levels of TNFA and IL-6 mRNA exhibited a significant increase at 5 dpi (Fig. S1F). Our previous research indicated that neutrophils may play a role in the brain injury associated with CVA6 infection [8, 23]. To elucidate the function of neutrophils, we employed FACS to assess neutrophil populations in both peripheral blood and brain tissue. Our findings revealed a significant increase in the proportion of neutrophils (Fig. 1A). Importantly, the levels of MPO in both brain tissues and serum were significantly elevated at 3 dpi and 5 dpi (Fig. S1G). Moreover, the levels of MPO-DNA complexes, indirect biomarkers of NETs release, were significantly elevated at 3 dpi and 5 dpi in brain tissues (Fig. 1B). Next, we employed BMDNs to establish an in vitro infection model, and observed the reticular structures composed of DNA and histones (CitH3) that were released by BMDNs (Fig. 1C and D). Likewise, TEM results also revealed the release of extracellular DNA degranulation as well as NETs formation 24 hpi (Fig. 1E). Using confocal microscopy and Western blotting, we detected elevated levels of GSDMD-N, IL-1β, and IL-18 in BMDNs infected with CVA6 at different time points (Fig. 1F, Fig. S1H-J). Further in vivo experiments revealed the activation of Caspase-1/GSDMD pathway in CVA6-infected brains (Fig. 1G, Fig. S1K). This activation predominantly occurred in activated neutrophils (Fig. S2A-C). Then, we observed elevated levels of IL-1β and IL-18 in brain slices (Fig. 1H-I).

Fig. 1.

The activation of the Caspase-1/GSDMD pathway and NETs formation in neutrophils are involved in CVA6-induced neuroinflammation. Ten-day-old ICR mice were i.p. inoculated with a lethal dose of the CVA6 strain (10⁴ TCID₅₀ /mouse) or saline, and sacrificed at 3 dpi and 5 dpi. (A): The percentage of neutrophils in brain tissues and blood samples from Mock- or CVA6-infected mice by using FACS (n = 4 per group); (B): The concentrations of MPO-DNA in brain lysis from Mock- or CVA6-infected mice at 3 dpi and 5 dpi (n = 4 per group); (C): BMDNs infected by CVA6 or stimulated by PMA to release NETs under bright field microscopy, and the black arrows represent the NETs, bar = 20 μm; (D): Ties with NETs (CitH3/DAPI) in BMDNs infected by CVA6 or stimulated by PMA (n = 12 per group), and the white arrows represent the NETs, bar = 50 μm; (E): NETs formation after CVA6 infection in BMDNs under TEM; (F): Expression profiles of key pyroptosis-associated proteins, namely pro-Casp1, cleaved Casp1 (p20), full-length GSDMD, and GSDMD-N, alongside the mature forms of IL-1β (17 kD) and IL-18 (18 kD) in CVA6-infected BMDNs. (G): The expression profiles of key pyroptosis-associated proteins, namely pro-Casp1, cleaved Casp1 (p20), full-length GSDMD, and GSDMD-N, alongside the mature forms of IL-1β (17 kD) and IL-18 (18 kD) in the brains tissues from Mock- or CVA6-infected mice at 3 dpi and 5 dpi (n = 3 per group); (H-I): Co-localization between neutrophils (MPO) and IL-1β, IL-18 in brain slices from Mock- or CVA6-infected mice at 5 dpi (n = 3 per group), and marked by white arrows, bar = 20 μm; ns, no significance; * P < 0.05; ** P < 0.01; *** P < 0.001

To further explore the mechanism of GSDMD activation in neutrophils during CVA6 infection, we performed experiments using Caspase-1 KO mice (Casp1-/-). Caspase-1 deficiency significantly prolonged the survival time of CVA6-infected mice, yet it did not improve overall survival rates (Fig. 2A-B). Moreover, compared with CVA6-infected WT mice, Casp1-/- mice exhibited reduced expression levels of GSDMD-N, IL-1β, and IL-18 in brain tissues (Fig. 2C, Fig. S3A). Immunofluorescence analysis further corroborated these findings in brain neutrophils (Fig. 2D-G, Fig. S3B-D). To investigate the impact of Caspase-1 inhibition on downstream of GSDMD-mediated pyroptosis, we administered the selective Caspase-1 inhibitor belnacasan to CVA6-infected mice. We observed that belnacasan significantly suppressed the maturation and secretion of IL-1β and IL-18—key cytokines downstream of GSDMD cleavage (Fig. 2H, Fig. S3E). Furthermore, in an in vitro model using BMDNs infected with CVA6, belnacasan treatment markedly reduced viral load (Fig. 2I). Collectively, these results collectively indicate that the activation of the Caspase-1/GSDMD pathway, along with the formation of NETs, plays a significant role in brain injury caused by CVA6.

Fig. 2.

Blocking Caspase-1 suppressed GSDMD expression and NETs release in brain of mice infected with CVA6. Ten-day-old C57BL/6J and Casp1-/- mice were i.p. inoculated with a lethal dose of the CVA6 (104 TCID50 /mouse) or saline. Mean clinical scores (A) and survival rates (B) were monitored until 15 dpi (n = 9–12 per group). Brain tissues were harvested for analysis at 5 dpi. (C): Expression profiles of key pyroptosis-associated proteins—including full-length GSDMD, GSDMD-N, and the mature forms of IL-1β (17 kDa) and IL-18 (18 kDa)—in brain tissues from WT or Casp1-/- mice following CVA6 infection (n = 3 per group). (D): Co-localization between neutrophils (MPO) and pyroptosis (GSDMD-N) in brain slices from WT or Casp1-/- mice following CVA6 infection, marked by white arrows. Bar = 20 μm. (E): Co-localization between neutrophils (MPO) and IL-1β, and marked by white arrows. Bar = 20 μm. (F): Co-localization between neutrophils (MPO) and IL-18, and marked by white arrows. Bar = 20 μm. (G): Quantification of IL-1β and IL-18 in brain slices (n = 3 per group). (H): Expression profiles of key pyroptosis-associated proteins in brain tissues of mice infected with CVA6 and treated with either PBS or belnacasan (n = 3 per group). (I): Viral load in CVA6-infected BMDNs treated with belnacasan was quantified by qPCR at 24 hpi (n = 4 per group). ns, no significance; * P < 0.05; ** P < 0.01; *** P < 0.001

Gasdermin D deficiency inhibits the release of NETs and alleviates neuroinflammation induced by CVA6

To determine whether GSDMD is involved in NETs release and the physiopathology of neuroinflammation, we induced neuroinflammation using CVA6 in WT and Gsdmd-/- mice (Fig. 3A). As shown in Fig. 3B-E and Fig. S4A-E, GSDMD deficiency markedly enhanced the survival rates of mice infected with CVA6, significantly suppressed the expression levels of IL-1β and IL-18, and substantially mitigated the pathological alterations in brain tissues as well as reduced the viral load. Notably, the release of NETs (MPO-DNA) and the transcription levels of IL-6 and TNFA mRNA were abrogated in Gsdmd-/- mice (Fig. 3G). Using confocal microscopy, we demonstrated that the quantity of cells exhibiting co-localization of MPO (neutrophils) with IL-1β and IL-18 was significantly diminished (Fig. 3H-I, Fig. S4F-H). Taken together, our findings reveal that GSDMD participates in NET production and consequent neuroinflammation, and mortality during CVA6 infection.

Fig. 3.

Gasdermin D deficiency inhibits the release of NETs and alleviates neuroinflammation induced by CVA6. Ten-day-old C57BL/6J and Gsdmd-/- mice were i.p. inoculated with a lethal dose of the CVA6 strain (10⁴ TCID₅₀ /mouse) or saline (A). A daily record was kept on the body weight (B), mean clinical score (C) and percent survival (D) of mice (n = 9–12 per group). Tissue analysis was performed at 5 dpi. (E): Expression profiles of key pyroptosis-associated proteins, namely full-length GSDMD, and GSDMD-N, alongside the mature forms of IL-1β (17 kD) and IL-18 (18 kD) in the brains tissues from Mock- or CVA6-infected mice (WT, Gsdmd-/-) (n = 3 per group). (F): Pathological damage and viral replication of brain were measure by H&E staining, Nissl staining and IHC (CVA6 VP1) of brain from Mock- or CVA6-infected mice (WT, Gsdmd-/-) (n = 3 per group), and black arrows indicate inflammatory cell infiltration in H&E, viral replication in IHC, and neuronal damage in Nissl staining, bar = 20 μm. (G): MPO concentrations in brain lysates and serum, MPO–DNA complex levels, and relative mRNA expression of TNFA and IL-6 in brain lysates from mock- or CVA6-infected mice (WT, Gsdmd-/-) (n = 4 per group). (H-I): Co-localization between neutrophils (MPO) and IL-1β (H), IL-18 (I) in brain tissues from Mock- or CVA6-infected mice (WT, Gsdmd-/-) (n = 3 per group), and marked by white arrows, bar = 20 μm; * P < 0.05; **** P < 0.0001

Pharmacologic inhibition of GSDMD prevents NET release and neuroinflammation induced by CVA6

A recent study demonstrated that disulfiram effectively inhibits pyroptosis mediated by the GSDMD [24]. Herein, we investigated whether the pharmacological inhibition of GSDMD using disulfiram could prevent NETosis and subsequent neuroinflammation during CVA6 infection (Fig. 4A). We observed that treatment with disulfiram improved the survival rates of animals subjected to CVA6 (Fig. 4B-D). We also observed a reduction in GSDMD activation, as well as decreased expression levels of IL-1β and IL-18 (Fig. 4E, Fig. S5A), alongside a mitigation of brain injury (Fig. 4F, Fig. S5B-D). Notably, the release of NETs (MPO-DNA) and the transcription levels of IL-6 and TNFA mRNA were abrogated in disulfiram-treated group compared with that in the vehicle control group (Fig. 4G). Using confocal microscopy, we demonstrated that the number of co-localized cells expressing MPO (neutrophils) with GSDMD (Fig. 4H), IL-1β (Fig. 4I), and IL-18 (Fig. 4J) was significantly reduced in disulfiram-treated group (Fig. S5E-H). Thus, these findings suggest that pharmacologic inhibition of GSDMD with disulfiram prevents NETs release and neuroinflammation induced by CVA6.

Fig. 4.

Pharmacologic inhibition of gasdermin D prevents NET release and neuroinflammation induced by CVA6. Ten-day-old ICR mice treated with PBS or Dissulfiram (10 mg/kg) were i.p. inoculated with a lethal dose of the CVA6 strain (10⁴ TCID₅₀ /mouse) or saline (A). A daily record was kept on the body weight (B), mean clinical score (C) and percent survival (D) of mice (n = 9–12 per group). Tissue analysis was performed at 5 dpi. Tissue analysis was performed at 5 dpi. (E): Expression profiles of key pyroptosis-associated proteins, namely full-length GSDMD, and GSDMD-N, alongside the mature forms of IL-1β (17 kD) and IL-18 (18 kD) in the brains tissues from Mock- or CVA6-infected mice treated with PBS or Dissulfiram (n = 3 per group). (F): Pathological damage and viral replication of brain were measure by H&E staining, Nissl staining and IHC (CVA6 VP1) of brain from Mock- or CVA6-infected mice treated with PBS or Dissulfiram (n = 3 per group), and black arrows indicate inflammatory cell infiltration in H&E, viral replication in IHC, and neuronal damage in Nissl staining, bar = 20 μm. (G): MPO concentrations in brain lysates and serum, MPO–DNA complex levels, and relative mRNA expression of TNFA and IL-6 in brain lysates from mock- or CVA6-infected mice treated with PBS or Dissulfiram (n = 4 per group). (H): Co-localization between neutrophils (MPO) and pyroptosis (GSDMD) in brain tissues from Mock- or CVA6-infected mice treated with PBS or Dissulfiram (n = 3 per group), and marked by white arrows, bar = 20 μm; (I-J): Co-localization between neutrophils (MPO) and IL-1β (I), IL-18 (J) in brain tissues from Mock- or CVA6-infected mice treated with PBS or Dissulfiram (n = 3 per group), and marked by white arrows, bar = 20 μm. * P < 0.05; ** P < 0.01; **** P < 0.0001

An inducible mouse model for specific neutrophil depletion was reproduced

Given that hyperactive neutrophils play a significant role in tissue damage following infection, we subsequently employed Ly6G-DTR KI mice to investigate the function of neutrophils. The strategy of hDTR gene KI was exhibited in Fig. 5A. The hDTR template DNA and CRISPR/Cas9 sgRNA for Ly6G-DTR KI have been described in previous study [25]. Ten-day-old KI and WT mice were i.p. injected with DT at a dosage of 20 ng/g bw. The KI mice appeared normal, similar to the WT mice (Fig. 5B). Genotyping analysis showed that heterozygous mice exhibited electrophoretic bands at both 364 bp and 558 bp (Fig. 5C), which were used in subsequent experiments. Following DT injection, neutrophils in the lungs and blood of Ly6G-DTR KI mice were nearly completely depleted (Fig. 5D-G), as reflected in their proportions (Fig. 5E and G). Although neutrophil levels slightly recovered by 48 h post-injection, they remained significantly lower than those in WT mice. No significant changes were observed in T (CD3ε+) or B (CD19+) lymphocytes in either KI or WT mice after DT injection. These data indicate that neutrophils can be specifically depleted for up to 48 h using the highly sensitive DTR system in this inducible mouse model.

Fig. 5.

An inducible mouse model for specific neutrophil depletion was reproduced. A: Strategy for hDTR gene knock-in using the CRISPR/Cas9 system. B: Macroscopic observations of Ly6G-DTR knock-in and wild-type C57BL/6 mice at 6 weeks old. C: Agarose gel electrophoresis images for mouse genotyping; red labels indicate heterozygous mice with bands at 364 bp and 558 bp. D-G: Flow cytometry plots show CD3ε-CD19-CD11b + Ly6G+ neutrophils in peripheral blood (D) and lung tissues (F), collected 24 and 48 h post-DT treatment (n = 3 per group). Neutrophil percentages in blood (E) and lung tissues (G)

Complete depletion of neutrophils exacerbates CVA6 infection

As illustrated in Fig. 6A, following the incubation of Ly6G-DTR KI and WT mice with CVA6 (10⁴ TCID₅₀) or an equivalent volume of PBS, we administered DT to these mice for either a single-dose intervention at 1 dpi or a two-dose intervention at 1 dpi and 3 dpi. Compared to WT mice, KI mice exhibited significant weight loss and all succumbed by 7–8 dpi, regardless of single or multiple DT doses (Fig. 6B-D). Notably, escalating DT doses accelerated clinical deterioration and mortality in CVA6-infected mice. Although the clinical scores among the three CVA6-infected groups were not statistically significant, a marked difference in survival rates was observed between the “WT+CVA6 + DT” and “KI+CVA6 + DT” groups (Fig. 6D). To further assess specific neutrophil depletion effects, we conducted histopathological analysis of brains and lungs. KI mice showed more severe inflammatory pneumonia (massive erythrocyte leakage, excessive infiltrated inflammatory cells, and serious alveolar collapse) and viral encephalitis post-CVA6 infection compared to WT mice (Fig. 6E). A significant correlation existed between pathological scores and corresponding changes (Fig. 6F). Additionally, viral loads in brains, lungs, and muscles (Fig. 6F) were higher in CVA6-infected KI mice. In addition, anti-Ly6G-mediated neutrophil depletion also accelerated mortality following CVA6 infection (Fig. 7A-B). Collectively, these findings suggest that complete neutrophil depletion facilitates rather than mitigates CVA6 infection.

Fig. 6.

Effects of complete specific depletion of neutrophils after CVA6 infection. Ten-day-old WT or Ly6G-DTR KI mice were administered i.p. with either a lethal dose of the CVA6 strain (10⁴ TCID₅₀ /mouse) or PBS. DT was administered to mice at the indicated time points. A: The CVA6 infection and DT intervention were performed. Body weight (B), clinical scores (C), and survival rates (D) were monitored until 15 dpi (n = 6–8 per group). E: H&E staining of brain and lung tissues showed characteristic lesions, marked by black arrows or circles, bar = 20 μm. Histopathological scores for the brain (F) and lung (G) were evaluated with n = 3/group. Viral loads in the brain (H), lung (I), and muscle (J) were measured, with n = 8/group. no significance; *P < 0.05, ***P < 0.001, ****P < 0.0001

Fig. 7.

The effects of anti-Ly6G-mediated neutrophil depletion and genetic NETs ablation on CVA6-induced mortality in mice. Ten-day-old ICR mice were administered i.p. with either the anti-Ly6G neutralizing antibody (clone 1A8; 12.5 mg/kg) or a rat IgG2a isotype control antibody (12.5 mg/kg), beginning 48 h prior to CVA6 infection and repeated once daily for two consecutive days (i.e., at 48 and 24 h before infection). A daily record was kept on the mean clinical score (A) and percent survival (B) of mice (n = 9–12 per group). Ten-day-old PAD4^fl/fl and PAD4 Ne-KO were administered i.p. with either a lethal dose of the CVA6 strain (10⁴ TCID₅₀ /mouse) or saline. A daily record was kept on the mean clinical score (A) and percent survival (B) of mice (n = 9–12 per group). ns, no significance; * P < 0.05; ** P < 0.01; *** P < 0.001

Genetic NETs ablation reduces CVA6-induced mortality in mice

Given that NET formation followed increased neutrophil recruitment during CVA6-induced neuroinflammation, we next examined whether selective targeting of NETs alone could improve survival. To this end, we generated neutrophil-specific PAD4 knockout mice (PAD4 Ne-KO) by crossing S100A8 Cre mice with PAD4^fl/fl mice. Neutrophil-specific ablation of PAD4 markedly ameliorated CVA6-induced neurological disease, as evidenced by attenuated clinical symptoms (Fig. 7C) and significantly improved survival (Fig. 7D). Collectively, these findings indicate that NET formation contributes to CVA6-induced neuropathogenesis.

GSDMD is associated with NET release in patients with HFMD

To demonstrate the clinical relevance of our findings, we conducted an analysis of serum MPO levels, as well as the transcription levels of GSDMD, IL-1β, and IL-18 in peripheral blood leukocytes obtained from patients with CVA6 infection exhibiting either mild or severe symptoms. First, we confirmed that severe patients with CVA6 infection showed higher levels of circulating M than mild patients at ≥ 5 days after onset (Fig. 8A), a finding that was consistent with our findings in the mouse model. Using qPCR, we also observed significantly elevated mRNA levels of GSDMD (Fig. 8B), IL-1β (Fig. 8C), and IL-18 (Fig. 8D) in peripheral blood leukocytes at ≥ 5 days after onset. Furthermore, the expression of GSDMD exhibited a significant positive correlation with both IL-1β (Fig. 8E) and IL-18 (Fig. 8F). Collectively, these results indicate that the GSDMD is involved in the release of NETs during CVA6 infection.

Fig. 8.

Gasdermin D is associated with NET release in patients with HFMD. A: The concentration of MPO in mild (n = 83) and severe cases (n = 83) with CVA6 infection was determined by MPO kit; B-D: The relative mRNA level of GSDMD (B), IL-1β (C) and IL-18 (D) in mild (n = 79) and severe cases (n = 79) with CVA6 infection was determined by qPCR; E-F: Correlation analysis between GSDMD mRNA level and IL-1β (E) / IL-18 (F) among all CVA6 infected cases (n = 158); * P < 0.05

Discussion

This study systematically reveals a previously unrecognized key mechanism through which enterovirus CVA6 infection induces neuroinflammation: neutrophils release NETs via the Caspase-1/GSDMD pathway. Using gene KO and pharmacological inhibition approaches, we demonstrated that GSDMD activation is essential for CVA6-induced NETs release and subsequent neuroinflammation. Notably, our findings highlight a dual “double-edged sword” role of neutrophils in CVA6 infection—while GSDMD-dependent NETs release contributes to tissue pathology, neutrophils themselves are indispensable for controlling viral replication, as their complete depletion exacerbates infection outcomes. Furthermore, we validated the link between GSDMD activation and NETs formation in clinical samples from HFMD patients. These findings enhance understanding of enteroviral pathogenesis, identify GSDMD as a promising therapeutic target, and provide a novel framework for developing precision interventions that reduce excessive inflammation without impairing essential host defenses.

NETs have garnered important attention as a key antimicrobial mechanism among the arsenal of neutrophils [26, 27]. This distinct extracellular structure not only confines pathogen dissemination but also directly eliminates microorganisms. However, excessive immune activation may lead to dysregulated NETs release, which can amplify inflammatory responses and even result in host tissue injury [28, 29]. MPO, a well-established neutrophil marker, exhibits a pathological role in inflammatory diseases that parallels neutrophil activity [30]. Upon overstimulation, MPO is released into surrounding tissues, where its intrinsic cytotoxicity contributes to exacerbated inflammation and cellular damage [31, 32]. Our study demonstrates that the inflammatory response triggered by excessive activation of NETs occurs in a GSDMD-dependent manner. In human populations, GSDMD is involved in NETs release during CVA6 infection. The underlying mechanism may involve GSDMD cleavage by neutrophil elastase (NE), followed by its translocation to the plasma membrane, thereby promoting NET release. Furthermore, GSDMD has been shown to play a regulatory role in the upstream activation of NE, suggesting a positive feedback loop between GSDMD and NE during this process [17, 33].

Pyroptosis is a form of inflammatory cell death primarily functioning to trigger robust inflammatory responses and defend the host against microbial infections. However, excessive pyroptosis can contribute to the development of various inflammatory disorders, including human immunodeficiency virus (HIV) infection and sepsis [34, 35]. The execution of pyroptosis depends on the cleavage of GSDMD by Caspase-1 in the classical pathway and by Caspase-4/5/11 in the non-classical pathway, resulting in the release of N-terminal and C-terminal fragments. Following proteolytic activation, the GSDMD N-terminal domain oligomerizes and inserts into the plasma membrane to form transmembrane pores, facilitating the extracellular release of proinflammatory cytokines such as IL-1β and IL-18, thereby driving intense inflammation and lytic cell death [36]. Our study indicates that Caspase-1 is in an activated state within the brain neutrophils infected by CVA6. Similarly, Santoni et al. reported that Pseudomonas aeruginosa lacking exotoxins U or S induces enhanced neutrophil death through Caspase-1-dependent pyroptosis [37]. Furthermore, in the context of SARS-CoV-2 infection, GSDMD activation in neutrophils has been shown to require active Caspase-1 and Caspase-4 [38].

Disulfiram is a U.S. FDA-approved drug originally developed treat alcoholism [39, 40]. It has been shown to inhibit pyroptosis by covalently binding to the catalytic cysteine residues of Caspase-1 and Caspase-11, thereby suppressing the activation of the Caspase-1/Caspase-11–GSDMD-N signaling axis. Notably, in a lipopolysaccharide-induced mouse model of sepsis, disulfiram covalently modifies human/mouse Cys191/Cys192 in GSDMD, effectively blocking pore formation and subsequent cell lysis [24]. Silva et al. also demonstrated that disulfiram-mediated inhibition of GSDMD in neutrophils significantly reduces NETosis, attenuates systemic inflammation and organ dysfunction, and improves survival outcomes [41]. Our study further reveals that pharmacological inhibition of GSDMD with disulfiram prevents CVA6-induced NET release and neuroinflammation.

Neutrophils are among the first immune cells recruited to sites of inflammation following infection or tissue injury. They employ a broad array of effector mechanisms, including phagocytosis, production of reactive oxygen species (ROS), release of proteases, and formation of NETs. However, excessive or dysregulated neutrophil activity at infectious foci can contribute to tissue damage, facilitating pathogen dissemination into the bloodstream and vital organs, thereby promoting systemic infection and potentially leading to fatal outcomes [42, 43]. In viral infections, neutrophils are generally recognized as key contributors to host defense. As the primary innate immune cells rapidly recruited to infection sites, they play a critical role in antiviral immunity. During influenza A virus (IAV) infection, neutrophils constitute the predominant infiltrating leukocyte population and are associated with a protective functional role in the inflammatory response [44, 45]. Notably, impaired neutrophil recruitment in IAV-infected mice deficient in inflammasome components—such as Nlrp3, ASC, or Caspase-1—is linked to increased mortality [46]. Our study further demonstrates that neutrophil depletion exacerbates CVA6 virus infection. Consistently, in a murine model of hepatitis virus infection, neutrophil depletion resulted in enhanced viral replication and higher mortality rates [47].

This study has several limitations. First, the upstream regulatory mechanisms governing GSDMD activation remain incompletely elucidated. Although our findings indicate involvement of the Caspase-1/GSDMD pathway in CVA6-induced neutrophil pyroptosis, they do not conclusively establish that GSDMD-mediated pyroptosis is strictly dependent on caspase-1. Future studies will investigate whether other inflammasome components—particularly NLRP3—as well as alternative pyroptotic pathways (e.g., Caspase-3/GSDME) contribute to neutrophil pyroptosis. Second, the human cohort component of this study employed peripheral blood leukocytes, which provide a useful but imperfect proxy for neutrophil-specific responses. Moreover, quantification of IL-1β and IL-18 mRNA levels in total peripheral blood leukocytes may not accurately reflect the abundance or activity of their mature, bioactive protein forms. The presence of p20 and GSDMD cleavage fragments in uninfected brain and cells was indeed unexpected. However, basal GSDMD processing in unstimulated cells and tissues has been recently reported [48, 49], likely reflecting low-level inflammasome activity associated with normal cellular turnover. Additionally, caspase-1 self-cleavage is an intrinsic mechanism to terminate inflammasome activity, which may further account for the observed p20 fragment [49].

Conclusion

This study demonstrates that CVA6-induced neuroinflammation is critically mediated by Caspase-1/GSDMD-dependent NET release from neutrophils. Genetic ablation of GSDMD significantly suppressed NET formation and alleviated neuroinflammatory damage, while pharmacological inhibition of GSDMD yielded comparable protective effects. Intriguingly, although GSDMD-driven NETosis exacerbates pathology, complete neutrophil depletion worsened CVA6 outcomes, revealing their indispensable role in viral control. Clinically, elevated GSDMD activation and NET markers in HFMD patients corroborated this pathway’s relevance. Collectively, these findings establish the Caspase-1/GSDMD axis as a pivotal driver of NET-mediated neuroinflammation in CVA6 infection and highlight GSDMD as a promising target for therapies that selectively mitigate pathological inflammation without compromising essential neutrophil antiviral functions.

Abbreviations

CVA6

Coxsackievirus A6

WT

Wild-type

GSDMD

Gasdermin D

KO

Knock-out

KI

Knock-in

PAD4

Peptidyl arginine deiminase 4

BMDNs

Bone marrow-derived neutrophils

HFMD

Hand, foot, and mouth disease

NETs

Neutrophil extracellular traps

hpi

Hours post infection

RD

Human rhabdomyosarcoma

DMEM

Dulbecco’s Modified Eagle Medium

FBS

Fetal bovine serum

i.p.

Intraperitoneal

dpi

Days post-infection

qPCR

Quantitative PCR

IF

Immunofluorescence

H&E

Hematoxylin and eosin

IHC

Immunohistochemical

MOI

Multiplicity of infection

MPO

Myeloperoxidase

TEM

Transmission electron microscopy

SD

Standard deviation

NE

Neutrophil elastase

HIV

Human immunodeficiency virus

FDA

Food and Drug Administration

ROS

Reactive oxygen species

IAV

Influenza A virus

Authors’ contributions

Y.F Jin and S.Y Chen designed the study. Q.M Hu conducted literature search, analytical screening, and manuscript writing. D. Li, F. Wang, YQ. Xie, TT. Sun, WG. J, PY. Z, JZ. L, HY. Y and GC. D refined the detailed research questions and modification in the manuscript. Y.F Jin, S.Y Chen and G.C Duan helped wrote the manuscript and secondary modification. Y.F Jin, S.Y Chen and G.C Duan supervised the study. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (NO.82574168, NO. 82372229, NO. 82572546, NO. 82002147, NO. 82273695, NO. 82073618), supported by Henan Province Science and Technology Research Project (NO. 242102311147), supported by China Postdoctoral Science Foundation (No. 2024T170246, No. 2024M750815), supported by the Open Grant from the Pingyuan Laboratory (No. 2023PY-OP-0202), supported by Open Project of Henan Province Engineering Research Center of Diagnosis and Treatment of Pediatric Infection and Critical Care (No. ERC202302), supported by the Zhengzhou Science and Technology Beneficiary Program Project (No. 2022KJHM0004), supported by the Open Project of Key Laboratory of Infection and Immunity of Anhui Higher Education Institutes (No. I&I-2024-R01), supported by the Open Project of Key Laboratory of Children’s Disease Research in Guangxi’s Colleges and Universities (No. GXCDR202401), supported by Program for Innovative Talents in Higher Education Institutions of Henan Province (No. 25HASTIT055), and supported by Outstanding Youth Science Foundation of Henan Province (No. 252300421122).

Data availability

All data generated or analyzed during this study are included in this article.

Declarations

Ethics approval and consent to participate

All experimental procedures carried out strictly and approved by the Life Science Ethics Review Committee of Zhengzhou University (protocol code: ZZUIRB2025-190).

Consent for publication

Written informed consent was obtained from parent/legal guardian(s) included in the study. The Author confirms: (1) that the work described has not been published before (except in the form of an abstract or as part of a published lecture, review, or thesis); (2) that it is not under consideration for publication elsewhere; (3) that its publication has been approved by all co-authors, if any; (4) that its publication has been approved (tacitly or explicitly) by the responsible authorities at the institution where the work is carried out.

Competing interests

The authors declare no competing interests.