GSDME promotes MASLD by regulating pyroptosis, Drp1 citrullination-dependent mitochondrial dynamic, and energy balance in intestine and liver

Abstract

Dysregulated metabolism, cell death, and inflammation contribute to the development of metabolic dysfunction-associated steatohepatitis (MASH). Pyroptosis, a recently identified form of programmed cell death, is closely linked to inflammation. However, the precise role of pyroptosis, particularly gasdermin-E (GSDME), in MASH development remains unknown. In this study, we observed GSDME cleavage and GSDME-associated interleukin-1β (IL-1β)/IL-18 induction in liver tissues of MASH patients and MASH mouse models induced by a choline-deficient high-fat diet (CDHFD) or a high-fat/high-cholesterol diet (HFHC). Compared with wild-type mice, global GSDME knockout mice exhibited reduced liver steatosis, steatohepatitis, fibrosis, endoplasmic reticulum stress, lipotoxicity and mitochondrial dysfunction in CDHFD- or HFHC-induced MASH models. Moreover, GSDME knockout resulted in increased energy expenditure, inhibited intestinal nutrient absorption, and reduced body weight. In the mice with GSDME deficiency, reintroduction of GSDME in myeloid cells-rather than hepatocytes-mimicked the MASH pathologies and metabolic dysfunctions, as well as the changes in the formation of neutrophil extracellular traps and hepatic macrophage/monocyte subclusters. These subclusters included shifts in Tim4⁺ or CD163⁺ resident Kupffer cells, Ly6C^hi pro-inflammatory monocytes, and Ly6C^loCCR2^loCX3CR1^hi patrolling monocytes. Integrated analyses of RNA sequencing and quantitative proteomics revealed a significant GSDME-dependent reduction in citrullination at the arginine-114 (R114) site of dynamin-related protein 1 (Drp1) during MASH. Mutation of Drp1 at R114 reduced its stability, impaired its ability to redistribute to mitochondria and regulate mitophagy, and ultimately promoted its degradation under MASH stress. GSDME deficiency reversed the de-citrullination of Drp1^R114, preserved Drp1 stability, and enhanced mitochondrial function. Our study highlights the role of GSDME in promoting MASH through regulating pyroptosis, Drp1 citrullination-dependent mitochondrial function, and energy balance in the intestine and liver, and suggests that GSDME may be a potential therapeutic target for managing MASH.

Fig. 1. Hepatic GSDME expression in MASH patients and mice.

Cell clustering (A) and gene marker heatmap (B) were generated using publicly available single-nuclei RNA-sequencing data obtained from the livers of individuals with MASH and healthy controls (GSE212837). C Odds ratio (OR) analysis was utilized to illustrate the transcriptional levels of GSDMD and GSDME in liver tissue from healthy control individuals and MASH patients. D The transcriptional levels of GSDME in cell clusters within liver tissue from healthy control individuals and MASH patients were examined. The red arrow points to the cell subcluster displaying the most significant changes. E GSDME activation was assessed through immunoblotting analysis of GSDME in liver tissues obtained from patients diagnosed with nonalcoholic steatohepatitis (MASH) and simple fatty liver (SFL). Both the full-length GSDME (GSDME-FL) and the cleaved N-terminal GSDME (GSDME-NT, active form) were detected. All patients were diagnosed histologically by pathologists based on biopsy results. F Immunohistochemistry staining of GSDME-NT was performed on liver tissues from MASH and SFL patients. Notably, lipid droplets were observed in liver sections from SFL patients. Additionally, hepatocellular ballooning was evident in liver sections from MASH patients. Arrows indicate immune cells. An antibody specifically against the NT of GSDME was used to stain GSDME-NT. G Histological analyses including H & E staining, Masson’s Trichrome staining, and Oil Red O staining demonstrated the presence of steatosis and fibrosis in two animal MASH models induced by a choline-deficient high-fat diet (CDHFD) and a high-fat/high-cholesterol diet (HFHC), respectively. The control group was fed a normal diet (ND). H Double-immunofluorescent staining and quantitative analysis on the colocalization of GSDME-NT and integrin α5 (a transmembrane adhesion protein) in liver sections from MASH mice fed CDHFD and HFHC. Mice fed ND were utilized as controls. I, J Immunoblotting analysis revealed GSDME activation in liver tissues from MASH mice fed CDHFD (E) and HFHC (F), with cleaved N-terminal GSDME (GSDME-NT, active form) detected exclusively in MASH mice livers. K, L Interleukin-1β (IL-1β) and IL-18, two downstream pro-inflammatory factors of GSDME, were upregulated in liver tissues from MASH mice fed CDHFD (G) and HFHC (H). P < 0.01, NS no significant difference. Unpaired Student’s t test was performed. N = 3 or 7 biological replicates.

Fig. 2. Global knockout of GSDME ameliorates steatosis and steatohepatitis in CDHFD-induced MASH mouse model.

A H & E histological staining in liver sections of WT and GSDME-KO (KO) mice fed CDHFD or ND. Steatosis score, inflammation score, ballooning score, and total NAS were calculated and presented. N = 6 per group. B Oil Red O staining revealed lipid accumulation in livers of WT and KO mice fed CDHFD or ND. The positive area was quantified. N = 4 per group. C GTT assay in WT and KO mice fed CDHFD or ND, illustrating glucose and insulin concentrations in plasma at different time-points. N = 8 per group. D ITT assay in WT and KO mice fed CDHFD or ND, illustrating glucose concentrations in plasma at different time-points. N = 8 per group. E Expression of lipid synthesis genes (Fabp1, Fasn, Scd1, Cd36, and Pparg) and lipid oxidation genes (Cpt1α and Pparα) in liver tissue of WT and KO mice fed CDHFD or ND. N = 6 per group. F Serum ALT and AST levels in WT and KO mice fed CDHFD or ND. N = 6 per group. G Immunohistochemical staining of myeloperoxidase (MPO, a marker of neutrophils) and CD68 (a marker of monocytes/macrophages) in liver tissue of WT and KO mice fed CDHFD or ND. Representative images and quantification analyses are presented. N = 6 per group. H TUNEL analysis in liver tissue of WT and KO mice fed CDHFD or ND. Representative images and quantification analyses are presented. N = 6 per group. I Protein expression of pro-inflammatory factors (IL-18, IL-1β, and IL-8) and chemokine CXCL1 in liver tissue of WT and KO mice fed CDHFD or ND. N = 3 per group. *P < 0.05, **P < 0.01, ***P < 0.001. One-Way ANOVA followed by Sidak’s test was performed.

Fig. 3. Global Knockout of GSDME mitigates liver fibrosis, ER stress and mitochondrial dysfunction in CDHFD-induced MASH mouse model.

A Liver fibrosis was assessed using Sirius Red staining in the livers of WT and KO mice fed ND and CDHFD. The red-stained area was quantified. Representative images and quantitative analysis were provided. B Masson’s Trichrome staining was employed to evaluate liver fibrosis in WT and KO mice fed ND and CDHFD. The blue-stained area was quantified. Representative images and quantitative analysis were provided. C qPCR analyses determined the mRNA expressions of fibrosis-related genes (Col1α1, Col3α1, Col5α1, Tgf-β, and α-Sma) in the livers of WT and KO mice fed ND and CDHFD. D Immunoblotting analyses assessed the protein expressions of ER stress markers (VDUP and phospho-PERK/PERK ratio) in the livers of mice fed ND and CDHFD. E Representative electron microscopy images depicted and analyzed mitochondrial membrane rupture (indicated by arrows) in the livers of mice fed ND and CDHFD. F Representative immunoblotting images and quantitative analyses showed the release of cytochrome-C (Cyto-C) from mitochondrion to cytoplasm in the liver tissue of mice fed ND and CDHFD. G Representative immunoblotting images and quantitative analyses showed the protein expressions of Mfn 1 and Mfn 2 in the livers of mice fed ND and CDHFD. H ATP contents in the livers of WT and KO mice fed ND and CDHFD. I Citrate synthase activity in the livers of WT and KO mice fed ND and CDHFD. CSA, Citrate synthase activity. **P < 0.01, **P < 0.001. Unpaired Student’s t test was performed. N = 3–6 biological replicates.

Fig. 4. Impacts of rescuing GSDME gene transcription in different liver cell types on MASH development.

A Body weight curves of WT mice, S/S mice (Gsdme^S/S mice, in which Gsdme gene transcription is terminated), S/S-MR mice (Gsdme^S/S;Lyzm-Cre mice, in which Gsdme gene transcription is rescued in myeloid cells), and S/S-HR mice (Gsdme^S/S;Alb-Cre mice, in which Gsdme gene transcription is rescued in hepatocytes) with MASH induced by CDHFD for 6 months. The body weight curve of WT mice fed ND was utilized as a control. B Liver weight and liver/body weight ratio of WT, S/S, S/S-MR, and S/S-HR mice with MASH. Liver weight and liver/body weight ratio of WT mice fed ND were utilized as control measurements. C Representative images and quantitative analyses of H & E staining and Oil Red O staining in WT, S/S, S/S-MR, and S/S-HR mice with MASH. D Serum TG and TC levels in WT, S/S, S/S-MR, and S/S-HR mice with MASH. E qPCR analyses of the mRNA levels of lipid synthesis-related genes (Fasn, Cd36, and Pparg) in the livers of WT, S/S, S/S-MR, and S/S-HR mice with MASH. F Representative images and quantification analyses of Sirius Red staining and Masson’s Trichrome staining in the livers of WT, S/S, S/S-MR, and S/S-HR mice with MASH. G qPCR analysis of the mRNA expression of pro-fibrosis genes (Col1α1, Col3α1, and Col5α1) in the livers of WT, S/S, S/S-MR, and S/S-HR mice with MASH. H Serum ALT and AST levels in WT, S/S, S/S-MR, and S/S-HR mice with MASH. **P < 0.01, ***P < 0.001. NS no significance. One-Way ANOVA followed by Sidak’s test was performed. N = 6 biological replicates.

Fig. 5. Multiplexed Flow Cytometry Reveals the Regulatory Action of Myeloid GSDME on MΦ/MO Infiltration In MASH.

A High-dimensional analysis of multiplexed flow cytometry using an unbiased nonlinear dimensionality reduction algorithm (t-distributed stochastic neighbor embedding, t-SNE) to identify clustering of subpopulations in hepatic CD45⁺ leukocyte cells. The live CD45⁺ leukocyte cells were gated with macrophages/monocytes-related markers (CD11b, F4/80, Ly6C, CD86, CD163, TIM4, CCR2, CX3CR1 and MHC-II). The multiplexed flow cytometry results were concatenated, transformed, and plotted in 2D t-SNE plots using R software. The contour plots showing the relative expression of the indicated markers of MΦ/MO. ResKCs: F4/80^hiCD11b^intTIM4^hi; MoKCs: F4/80^hiCD11b^intTIM4^lo; Ly6C^hi IMs: CD11b^hiF4/80^intLy6C^hi; Ly6C^int IMs: CD11b^hiF4/80^intLy6C^int; Ly6C^low IMs: CD11b^hiF4/80^intLy6C^lo. B Comparison of t-SNE dimensionality reduction and embedding of CD45⁺ cells pooled from WT (red) and GSDME^S/S (S/S) mice fed with CDHFD. C Heatmap showing the differences in t-SNE dimensionality reduction and embedding of CD45⁺ cells from WT and S/S mice fed with CDHFD. Four different regions (R1 to R4) were indicated by dotted line. D Representative flow cytometry plots and quantification analyses on the proportions of Kupffer cells (KCs, F4/80^hiCD11b^int) and infiltrating monocytes (IMs, F4/80^intCD11b^hi) within CD45⁺ cells from liver tissues of WT-ND, WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. E Representative flow cytometry plots and quantification analyses on the proportions of embryo-derived resident KCs (ResKCs, F4/80^hiCD11b^intTIM4^hi) and monocyte-derived KCs (MoKCs, F4/80^hiCD11b^intTIM4^lo) within KCs of liver tissues of WT-ND, WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. F Representative flow cytometry plots and quantification analyses on the proportions of M1-like IMs (CD86⁺ IMs) and CD86⁻ IMs within IMs of liver tissues of WT-ND, WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. G Representative flow cytometry plots and quantification analyses on the proportions of M2-like ResKCs (CD163⁺ ResKCs) and CD163⁻ ResKCs within ResKCs of liver tissues of WT-ND, WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. H Representative flow cytometry plots and quantification analyses of three subclusters of IMs (Ly6C^hi IMs, CD11b^hiF4/80^intLy6C^hi; Ly6C^lo IMs, CD11b^hiF4/80^intLy6C^lo; Ly6C^lo;MHC-II^lo IMs, CD11b^hiF4/80^intLy6C^loMHC-II^lo) within IMs of liver tissues of WT-ND, WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. **P < 0.01 vs WT-ND, ^##P < 0.01 vs WT-CDHFD. One-Way ANOVA followed by Sidak’s test was performed. N = 3 biological replicates.

Fig. 6. GSDME controls energy homeostasis and intentisne nutrition absorption.

A The protein level of citrullinated histone 3 (citrullinated H3) was measured to evaluate the formation of neutrophil extracellular traps (NETs) in liver tissue of MASH. B, C Mitochondrial mass was assessed by ATP contents and citrate synthase activity in the liver from WT-ND, WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. D Measurement of oxygen consumption rate (OCR) using Seahorse XF Analyzers in the liver tissues of WT-ND, WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. The basal and maximal respiration were calculated and presented. E Accumulative food intake in WT-ND, WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. F Comparison of energy expenditure (EE) in WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice, adjusted by body mass and lean mass. G UCP1 protein expression in brown adipose tissue of WT-ND, WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. H Fecal output of WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. I Energy in the feces of WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. J Serum LPS levels in WT-ND, WT-CDHFD, S/S-CDHFD, S/S-MR-CDHFD and S/S-HR-CDHFD mice. K Serum LPS levels in WT and KO mice fed with ND or CDHFD. **P < 0.01 vs WT-ND. NS, no significance. Analysis of covariance (ANCOVA) followed by Tukey test was performed in F. ANOVA followed by Sidak’s test was performed in other panels. N = 3 biological replicates. S/S, Gsdme^S/S; S/S-MR, Gsdme^S/S-MR; S/S-HR, Gsdme^S/S-HR.

Fig. 7. MASH Promotes Loss of Drp1 R114 Citrullination, Which Is Reversed by GSDME Deletion.

A Initial screening of post-translational modification (PTM) using immunoblotting. The liver samples of WT-ND, WT-CDHFD and KO-CDHFD mice were loaded in SDS-PAGE and probed by antibodies against ubiquitination, succinylation, lysine 2-hydroxyisobutyrylation, lysine lactylation and citrullination. Protein citrullination in these sample was significantly altered (indicated as red arrows). B Citrullination process that occurs in an arginine residue of protein. PAD, peptidyl arginine deiminase. C Citrullination mapping employed a 4D label-free quantitative (4D-LFQ) proteomics strategy in liver tissues of WT-ND, WT-CDHFD, and KO-CDHFD mice. D Volcano plot illustrates differentially expressed citrullinated proteins influenced by MASH stress (WT-CDHFD vs WT-ND). E Volcano plot illustrates differentially expressed citrullinated proteins influenced by GSDME deletion (KO-CDHFD vs WT-CDHFD). F Protein-protein interaction analysis of the differentially expressed citrullinated proteins by MASH and GSDME deletion. Three clusters were identified, and the enriched pathways within these clusters were presented. G Screening strategy for potential citrullination site regulated by MASH and GSDME deletion. Six sites (Hadhb^R280/283, Akr1b1^R70, Nop58^R117, Gtpbp1^R189, Hsd11b1^R137, and Drp1^R114) were identified. Drp1^R114 was selected for further investigation. H The MS/MS spectrum showcases the citrullinated peptide (LYTDFDEIRQEIENETER) from mouse Drp1, indicating the citrullinated arginine (R) in pink. This MS/MS spectrum was obtained through 4D-LFQ proteomics with an Orbitrap mass spectrometer. I Quantitative R114 citrullinated Drp1 levels in liver tissue were extracted from 4D-LFQ mass spectrum results of WT-ND, WT-CDHFD, and KO-CDHFD mice. J Quantitative total Drp1 levels in liver tissue were extracted from 4D-LFQ mass spectrum results of WT-ND, WT-CDHFD, and KO-CDHFD mice. **P < 0.01. One-Way ANOVA followed by Sidak’s test was performed. N = 3 biological replicates.

Fig. 8. R114 Citrullination of Drp1 Is Critical for Its Stability and Mitochondrial Function.

A Representative immunoblotting image and quantitative analysis of total Drp1 in the livers of WT and KO mice fed with ND and CDHFD. B Representative immunoblotting image and quantitative analysis of citrullinated Drp1 in the livers of WT and KO mice fed with ND and CDHFD. C Alignment of Drp1 amino acid sequences across different species. The R114 site of mouse Drp1 is conserved and corresponds to the R108 site of human DRP1. D Position of the R114 site in the mouse Drp1 sequence and its corresponding R108 site in the human DRP1 sequence within the entire Drp1 amino acid sequence. E Visualization of the R108 site within the 3D structure of human DRP1, obtained from the PDB database (No. 4H1U) and viewed using PyMol software. F Immunoblotting depicting the citrullination of Flag-tagged WT Drp1 and Flag-tagged mutant-R114K Drp1 in BMDMs. Plasmids carrying these constructs were overexpressed in BMDMs, and immunoprecipitation with anti-Flag followed by anti-citrullination immunoblotting was performed. G Immunoblotting displaying the protein expression of Drp1 in BMDMs overexpressing Flag-tagged WT Drp1 and Flag-tagged mutant-R114K Drp1. The cells were immunoprecipitated with anti-Flag and then immunoblotted using anti-citrullination antibodies. H Immunoblotting showing the protein expression of Drp1 in BMDMs overexpressing Flag-tagged WT Drp1 and Flag-tagged mutant-R114K Drp1 under an in vitro MASH model induced by a methionine- and choline-deficient medium (MCDM). The cells were cultured in either control medium or MCDM for an additional 48 h. I Immunoblotting revealing the protein expression of mitochondrial transcription factor A (TFAM) in BMDMs overexpressing Flag-tagged WT Drp1 and Flag-tagged mutant-R114K Drp1 under an in vitro MASH model. J Mitochondrial dynamics of BMDMs overexpressing Flag-tagged WT Drp1 and Flag-tagged mutant-R114K Drp1 were evaluated by staining with Mito-tracker Deep Red probe under an in vitro MASH model. The average mitochondrial length was calculated based on intact mitochondria. DAPI was used for nuclear staining. K Mitochondrial ROS in BMDMs overexpressing Flag-tagged WT Drp1 and Flag-tagged mutant-R114K Drp1 under an in vitro MASH model. L Mitochondrial complex I and complex IV activities in BMDMs overexpressing Flag-tagged WT Drp1 and Flag-tagged mutant-R114K Drp1 under an in vitro MASH model. *P < 0.05, **P < 0.01, ***P < 0.001. One-Way ANOVA followed by Sidak’s test was performed. N = 6 biological replicates.