Gasdermin D drives focal crystalline thrombotic microangiopathy by accelerating immunothrombosis and necroinflammation Free

Thrombotic microangiopathy (TMA) is a heterogeneous group of diseases characterized by microvascular immunothrombosis and ischemic tissue injury, leading to organ failure. TMA may manifest in otherwise healthy individuals after exposure to bacterial toxins, such as Shiga toxin-associated hemolytic uremic syndrome. Furthermore, systemic TMA can arise either due to a deficiency in von Willebrand factor-cleaving protease (ie, thrombotic thrombocytopenic purpura), or owing to dysregulation of the alternative complement pathway (ie, atypical hemolytic uremic syndrome). These pathogenic conditions may be clinically expressed in either a hereditary or acquired manner following incidental triggers such as infections, certain drugs, pregnancy, or transplantation.^1,2 We recently described that injection of cholesterol crystals (CCs) into the kidney artery of mice induces focal TMA, particularly involving renal thrombotic arterial occlusions, and subsequent ischemic kidney infarction and failure.³ Although focal TMA may not manifest typical signs of systemic TMA, such as hemolytic anemia,⁴ the key elements of microvascular immunothrombosis are similar.

Gasdermin D (GSDMD) is a pore-forming protein.⁵ Upon activation, GSDMD translocates from the cytosol into the plasma membrane via its N-terminal domain, where it forms pores. GSDMD pores facilitate the secretion of mature interleukin-1β (IL-1β) and IL-18 but also promote membrane rupture, that is, pyroptosis, a highly inflammatory form of regulated necrosis in myeloid cells downstream of inflammasome activation.^6,7 In addition, GSDMD has been implicated in the release of neutrophil extracellular traps (NETs).^8-10 In neutrophils, neutrophil elastase can activate GSDMD in an inflammasome-independent manner, which in turn promotes the further release and activation of neutrophil elastase, histone cleavage, and chromatin decondensation, an early stage of NET formation.⁸ Although these proinflammatory cell death mechanisms contribute to host defense against pathogens, their dysregulation can cause unnecessary tissue damage in sterile diseases.¹¹ 

Circulating particles can activate the nucleotide oligomerization domain-like receptor protein 3 (NLRP3) inflammasome,^12,13 and induce neutrophils to release NETs.¹⁴ NETs also elicit a potent procoagulant response by recruiting and activating platelets.¹⁵ These processes can accelerate necroinflammation and immunothrombosis. Therefore, we hypothesized that GSDMD contributes to focal crystalline TMA by amplifying neutrophil necrosis, thus facilitating immunothrombosis within the vasculature and enhancing subsequent necroinflammation of the ischemic tissue.

A detailed description of the “Materials and methods” is provided in the supplemental Data, available on the Blood website.

C57BL/6N mice were obtained from Charles River Laboratories (Sulzfeld, Germany). Gsdmd^−/− mice were provided by Andreas Linkermann of the Technische Universität Dresden. All mice were housed in groups of 5 under specific pathogen-free conditions with enrichment and had access to food and water. All experimental procedures were approved by the local government authority Regierung von Oberbayern (reference no. ROB-55.2-2532.Vet_02-19-79) based on the European Union Directive for the Protection of Animals Used for Scientific Purposes (2010/63/EU), and reported according to the Animal Research: Reporting of In Vivo Experiments guidelines.¹⁶ 

Human blood neutrophils were isolated from healthy individuals using dextran sedimentation, followed by Ficoll-Hypaque density centrifugation as previously described.^17,18 The cells were suspended in Hanks’ balanced salt solution supplemented with 2% fetal calf serum. The study to obtain whole blood samples from healthy individuals received approval from the local ethical review board of the Medical Faculty at Ludwig-Maximilians-Universität, Munich, Germany (reference no. 21-0522), and written informed consent was acquired from all participants.

Induced pluripotent stem cells (iPSC) were maintained on a tissue culture dish coated with growth factor-reduced Matrigel (no. 356231; Corning) in mTeSR1 serum-free medium (no. 5850; STEMCELL). Differentiation toward neutrophil-like granulocytes was initiated as outlined in the supplemental Data.

Statistical analysis was performed using the GraphPad Prism 7 software (GraphPad, La Jolla, CA). For in vivo data, the mean ± standard deviation is presented, and the Student t test was used to determine the significance between the 2 groups. For comparing 3 or more groups, a 1-way analysis of variance (ANOVA) with Tukey post hoc test was used. When 2 parameters with multiple groups were used, 2-way ANOVA with Bonferroni multiple comparisons test was performed. For in vitro data, the mean ± standard error of the mean is presented, and 1-way ANOVA with Dunnett multiple comparisons test was performed to compare 3 or more groups. When 2 or more parameters with multiple groups were used, 2-way ANOVA with Dunnett multiple comparisons test was performed. Statistical significance was determined by P values <.05, which were indicated as ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001.

To investigate the expression and activation of GSDMD in neutrophils during ischemic inflammation, we conducted a single-cell RNA sequencing analysis using a previously published data set.¹⁹ The data set was based on cells sorted from the kidney, blood, and spleen of C57BL/6J mice before and after unilateral kidney ischemia-reperfusion injury (Figure 1A) and contained 80 829 cells, which were grouped into 26 clusters after quality control (Figure 1B). Clusters 3 and 15 were identified as neutrophils, based on the expression of specific transcripts (supplemental Figure 1A). Further categorization of these neutrophils into 8 clusters (G0-4 and G5a-c) based on their differentiation stages²¹ (supplemental Table 1) revealed that clusters G0, G1, G2, and G3 correspond to granulocyte-monocyte progenitor, proneutrophil, preneutrophil, and immature neutrophil, respectively, and primarily derived from spleen and blood (Figure 1C; supplemental Figure 1B-D). Conversely, clusters G4 and G5a-c represent mature neutrophils from the spleen, blood, and kidney after injury, respectively. Gsdmd transcripts were mainly detected in immature and mature neutrophils from all organs (Figure 1D-E). Thus, neutrophil maturation involves constitutive expression of Gsdmd and Gsdmd-positive mature neutrophils that migrate to the injured kidney.

Next, to elucidate the involvement of neutrophils in CC-induced TMA, we performed MACSima imaging using markers for neutrophils, as well as cell death within the whole TMA kidneys and the contralateral sham kidneys from wild-type (WT) mice (Figure 1F). In the TMA kidney, smooth muscle actin-positive arteries showed the presence of CD41-positive platelets within thrombi (further supported by Figure 1G), together with neutrophils that formed NETs manifested by the colocalization of Ly6G and citrullinated histone 3 (CitH3). Significant accumulation of neutrophils was observed in the peri-infarct region. Moreover, the kidney cortex revealed widespread parenchymal cell death, as indicated by the presence of cleaved caspase-3-positive cells. These findings suggest that neutrophils actively participate in TMA lesions and ischemic tissue infarctions.

Furthermore, we investigated whether Gsdmd is expressed and activated in TMA kidneys. Immunohistochemistry revealed Gsdmd positivity in peritubular interstitial cells of TMA kidneys from WT mice compared with sham kidneys (Figure 1H). In contrast, no signal was detected in TMA kidneys of Gsdmd^−/− mice, demonstrating the specificity of the Gsdmd antibody. Immunoblot analysis of total kidneys further demonstrated increased expression of the cleaved N-terminal fragment of Gsdmd and cleaved caspase-1 in TMA kidneys from WT mice compared with kidneys from healthy controls (Figure 1I). Importantly, immunoblot analysis performed on immune cells isolated from WT TMA kidneys showed an increased expression of cleaved GSDMD (Figure 1J). Concurrently, flow cytometric analysis revealed increased levels of activated caspase1-positive neutrophils in WT TMA kidneys (Figure 1K-L). These findings suggest Gsdmd activation in the crystalline TMA kidney and substantiate its association with infiltrating immune cells including neutrophils.

To investigate the role of GSDMD in focal TMA, we used Gsdmd^−/− mice and WT mice (supplemental Figure 2A). TMA was induced by injecting CC (20 mg/kg) into the left kidney artery and the mice were euthanized 24 hours after induction (Figure 2A). Gsdmd deficiency reduced the number of arteries and glomerular microvessels obstructed by crystal clots (Figure 2B-F). Compared with WT mice, Gsdmd^−/− mice were partially protected from the sudden drop in glomerular filtration rate, that is, acute kidney injury (AKI; Figure 2G). The kidney infarct size was consistently reduced (Figure 2H-I), similar to the tubular injury score (Figure 2J-K) and terminal deoxyribonucleotidyltransferase–mediated 2´-deoxyuridine, 5´-triphosphate–biotin nick end labeling positivity of kidney cells (supplemental Figure 2B-C). Furthermore, Gsdmd deficiency reduced the levels of mature IL-1β in the kidneys (Figure 2L), as well as the levels of circulating IL-1β and histone (Figure 2M; supplemental Figure 2D) after TMA. These data indicate that GSDMD contributes to focal crystalline TMA and its consequences, such as tissue infarction and organ failure.

To investigate the mechanisms underlying the improvement of the TMA phenotype in Gsdmd^−/− mice, we focused on neutrophils, which are the first leukocytes recruited to inflammatory sites²² (Figure 1F) and exhibit GSDMD expression during maturation and sterile kidney inflammation (Figure 1D-E). Consequently, we examined the abundance of circulating and kidney-infiltrating neutrophils using flow cytometry (Figure 3A; supplemental Figure 3A-B). In contrast to WT mice, Gsdmd^−/− mice after focal TMA showed reduced percentages and absolute numbers of neutrophils, recognized as CD45⁺ CD11b⁺ Ly6G⁺ cells in both the blood (Figure 3B-C) and kidney (Figure 3D-E). Consistent with the flow cytometric analysis, immunohistochemistry displayed reduced numbers of Ly6G-positive neutrophils infiltrating the TMA kidney in Gsdmd^−/− mice compared with WT mice (supplemental Figure 3G-H). In contrast, the number of monocytes (CD45⁺ CD11b⁺ Ly6C⁺) in the kidney and blood was unaffected (supplemental Figure 3C-F).

To comprehend the difference in neutrophil recruitment from the bone marrow to the blood and kidney, we analyzed the percentages and absolute numbers of neutrophils in the bone marrow, spleen, and lungs (supplemental Figure 4A-C). However, neither neutrophil nor monocyte numbers showed any significant differences (supplemental Figure 4D-O). Nevertheless, we examined neutrophil maturation in circulation and hematopoietic organs (Figure 3F). Interestingly, Gsdmd deficiency reduced the fraction of mature neutrophils identified as CD45⁺ CD11b⁺ Ly6G⁺ CD101⁺ cells in the bone marrow, spleen, and blood compared with WT mice after focal TMA (Figure 3G-I). Correspondingly, the number of mature neutrophils in each organ was significantly decreased in Gsdmd^−/− mice after focal TMA (Figure 3J-L). The baseline levels of neutrophils and monocytes (supplemental Figure 5A-C), as well as the fraction of mature neutrophils (supplemental Figure 5D-E) were comparable in both groups under normal conditions. Additional in vitro experiments revealed that Gsdmd^−/− bone marrow mature neutrophils showed reduced expression of CXCR2 in the presence of the differentiation factor granulocyte colony–stimulating factor or tumor necrosis factor alpha after 24 hours in culture (Figure 3M), which is crucial for neutrophil egress from the bone marrow to blood.²³ 

Furthermore, we performed flow cytometric analysis to examine the expression of β₂ integrin macrophage-1 antigen (MAC-1), which plays a crucial role in facilitating neutrophil migration into inflamed tissues,^24,25 in bone marrow neutrophils. We observed increased expression of MAC-1 in bone marrow neutrophils from WT mice after focal TMA compared with that in healthy WT mice, whereas MAC-1 expression was diminished significantly in Gsdmd^−/− mice with focal TMA (Figure 3N). These findings indicate that neutrophil maturation and β₂-integrin activation are impaired in Gsdmd^−/− mice with focal TMA, leading to reduced neutrophil migration from the bone marrow to the blood and kidneys, thus contributing to an attenuated TMA phenotype.

The release of NETs from neutrophils can contribute to immunothrombosis and necroinflammation; hence, we performed immunofluorescence staining for markers of NETs. Gsdmd deficiency reduced the formation of intravascular NETs in the kidneys, identified as a CitH3-positive area originating from Ly6G-positive neutrophils within the alpha smooth muscle actin-positive arteries (Figure 4A,C). Additionally, Gsdmd deficiency led to a decrease in intravascular occlusion involving neutrophils, platelets, and the area containing both neutrophils and platelets (Figure 4B,D-F). Furthermore, Gsdmd deficiency reduced NET formation in the kidney interstitial tissue (Figure 4G-H). This observation was further supported by a decrease in the absolute number of NETing neutrophils in the kidneys, undergoing cell death identified by flow cytometry as Ly6G and CitH3 double-positive cells (Figure 4I; supplemental Figure 6). These data indicated that GSDMD plays a crucial role in neutrophil- and NET-mediated immunothrombosis and necroinflammation in TMA.

We postulated that neutrophils could potentially function as a cellular source of GSDMD in TMA. This is supported by the observation that neutrophils predominantly infiltrated the kidney within 24 hours after crystal injection (Figure 3D), the presence of activated GSDMD and caspase1 in immune cells, including neutrophils within TMA kidneys (Figure 1J-L), and by our finding that the absence of Gsdmd in tubular epithelial cells did not impact on the levels of hypoxia-induced necrosis (supplemental Figure 7). Next, we examined the role of GSDMD in CC–induced neutrophil activation and death in vitro. Human neutrophils, primed with lipopolysaccharide (LPS) and exposed to increasing doses of CC, dose-dependently released IL-1β (Figure 5A). Immunoblot analysis showed that LPS/CC stimulation induced GSDMD cleavage, resulting in a subsequent increase in mature IL-1β in the supernatant (Figure 5B). Disulfiram (DSF), an inhibitor of GSDMD pore formation without affecting other inflammasome components,^6,26 dose-dependently reduced this effect, whereas receptor-interacting protein kinase 1 and mixed lineage kinase domain-like inhibitors, and a caspase-1 inhibitor had no effect on IL-1β release (Figure 5C). These results suggest that GSDMD, rather than neutrophil necroptosis, is involved in IL-1β release. The same was noted for lactate dehydrogenase release, a marker of cell necrosis, indicating that CC triggers GSDMD–dependent neutrophil pyroptosis rather than necroptosis (Figure 5D). Notably, the IL-1 receptor antagonist anakinra did not affect IL-1β release or GSDMD cleavage in CC-stimulated neutrophils (supplemental Figure 8), suggesting that the IL-1β pathway does not reciprocally influence GSDMD cleavage and the subsequent release of IL-1β.

Experiments using SYTOX Green (SG), a marker of cell death, showed an increased SG signal in CXCL8-primed human neutrophils in a CC dose-dependent manner (supplemental Figure 9A). Furthermore, fluorescence microscopy confirmed that CC exposure promoted NET formation, which was characterized by web-like structures of extracellular DNA decorated with granule proteins (Figure 5E). However, despite DSF treatment of CC-stimulated neutrophils, NET formation was still observed (Figure 5E), and SG positivity was not reduced, unlike the effects seen with diphenyleneiodonium chloride, an inhibitor of nicotinamide adenine dinucleotide phosphate oxidase, which served as a control for CC–induced NET inhibition (Figure 5F-G). Thus, these findings indicate that GSDMD contributes to CC–induced neutrophil pyroptosis but not CC-induced NETosis.

Notably, histone, a known soluble NETs inducer,¹⁷ showed reduced NETs induction, as indicated by decreased levels of SG positivity in DSF-treated human neutrophils (supplemental Figure 9B-D). Histone stimulation did not induce IL-1β release from neutrophils (supplemental Figure 9E). Consistently, bone marrow neutrophils derived from Gsdmd^−/− mice showed comparable levels of CC–induced NET formation but diminished histone–induced NET formation, in contrast to those from WT mice (supplemental Figure 10). Furthermore, Gsdmd deficiency prevented neutrophil adhesion and NET formation in CC-activated platelet–induced thrombi (Figure 5H-K; supplemental Figure 11). Together, these findings suggest that GSDMD does not modulate CC-induced but histone-induced NETosis and NETosis triggered by CC-activated platelets.

To validate the findings observed with pharmacological inhibitors, we used a genetic approach and generated GSDMD–deficient, human iPSC–derived neutrophils using CRISPR-Cas9–mediated gene editing. The absence of GSDMD in iPSC-derived neutrophils was confirmed through immunoblot and Sanger sequencing of the 2 distinct clones (Figure 6A; supplemental Figure 12A). Notably, GSDMD protein expression increased in differentiated control iPSC-derived neutrophils toward mature neutrophils (Figure 6B). The number of live floating neutrophils and their maturation states were comparable between control and GSDMD-knockout clones (supplemental Figure 12B-C). Fluorescence microscopy revealed cytoplasmic expression of GSDMD colocalized with myeloperoxidase in control iPSC-derived neutrophils (supplemental Figure 12D). GSDMD-knockout clones primed with LPS and exposed to CC showed reduced levels of IL-1β and lactate dehydrogenase in the cell culture supernatants compared with control iPSC-derived neutrophils (Figure 6C-D; supplemental Figure 13). Furthermore, flow cytometric analysis demonstrated reduced activation of β₂ integrins, lymphocyte function–associated antigen 1, and MAC-1 in the GSDMD-knockout clones (Figure 6E-F). GSDMD-knockout clones formed NETs upon CC stimulation, similar to control iPSC-derived neutrophils (Figure 6G), and did not show a decrease in the inducibility quantified by SG positivity (Figure 6H-6I). Collectively, these data confirm the involvement of GSDMD in CC–induced neutrophil pyroptosis rather than NETosis and β₂-integrin activation.

To investigate the therapeutic potential of targeting GSDMD, we administered DSF to WT mice either 4 hours prior (prophylactic) or 3 hours after (therapeutic) CC injection and assessed AKI and ischemic infarction (Figure 7A; supplemental Figure 14A). Compared with vehicle-treated controls, both prophylactic and therapeutic DSF treatment reduced the number of arteries affected by focal crystalline TMA (Figure 7B-C; supplemental Figure 14B-C). Both treatment regimens protected mice from the sudden drop in glomerular filtration rate, preventing AKI (Figure 7D; supplemental Figure 14D). Consistently, both prophylactic and therapeutic DSF treatments reduced the ischemic infarction (Figure 7E-F; supplemental Figure 14E-F) and tubular injury scores (Figure 7G-H; supplemental Figure 14G-H). Flow cytometric analysis demonstrated a reduction in the absolute numbers of blood neutrophils and monocytes in both treatment groups (Figure 7I-K; supplemental Figures 14I-K and 15), as well as a decrease in the percentages and absolute numbers of kidney-infiltrating neutrophils (Figure 7L-M; supplemental Figure 14L-M). Moreover, both treatments reduced NET formation in the TMA kidney (Figure 7N; supplemental Figure 14N). Taken together, both prophylactic and therapeutic DSF treatment improved focal TMA-induced AKI and kidney infarction by ameliorating clot formation and neutrophil-mediated necroinflammation.

We hypothesized that GSDMD plays a role in the pathogenesis of (crystalline) TMA by promoting intravascular immunothrombosis and its subsequent outcomes, including ischemic necroinflammation and organ failure. Our experiments revealed that Gsdmd deficiency alleviated CC–induced focal TMA, AKI, and ischemic kidney infarction. We observed that GSDMD contributed to CC–induced neutrophil necrosis via pyroptosis rather than necroptosis or NETosis. Moreover, GSDMD plays a role in the activation of β₂ integrin and the maturation process of neutrophils, both of which promote neutrophil migration into inflamed tissues. Both prophylactic and therapeutic administration of DSF protected CC-injected mice from crystal clot formation, AKI, and kidney infarction. These findings identified GSDMD as a key mediator of focal crystalline TMA and its consequences in ischemic tissue infarction and organ failure.

Several previous studies have implicated the involvement of GSDMD and pyroptosis in driving AKI, but these findings have yielded conflicting results.^27-29 However, various conditions in a model disease context may account for this virtual discrepancy. In CC-induced TMA, we observed upregulation of activated caspase-1 and GSDMD protein expression in immune cells in the kidneys, and the NLRP3 inhibitor MCC950 exhibited a protective effect against ischemic necroinflammation.³ Furthermore, CCs themselves have been identified as inducers of NLRP3 inflammasome.^12,13 Hence, focal crystalline TMA involves robust inflammasome activation, a major upstream event in GSDMD activation, and subsequent pathogenesis. Importantly, previous studies^27,28 have predominantly focused on tubular epithelial cells, in which GSDMD and pyroptosis are activated downstream of caspase-4, 5, and 11. In contrast, our findings indicate that the predominant expression of GSDMD in peritubular interstitial cells in the TMA kidney and the absence of Gsdmd in tubular epithelial cells did not affect the levels of hypoxia-induced necrosis. Moreover, it has been reported that protein expression of GSDMD is undetectable in lysates of specifically isolated mouse kidney tubules.²⁹ Based on our observations, we speculate that infiltrating leukocytes, particularly neutrophils, which massively evolve only hours after the embolic event, serve as the cellular origin of GSDMD in the development of TMA.

It has been reported that CC induces neutrophil necrosis as a form of necroptosis³⁰ and NETosis.¹⁴ In our investigation, GSDMD regulated CC–induced neutrophil pyroptosis and the release of IL-1β. Importantly, this observation indicates an active and specific process facilitated by GSDMD, rather than a passive event associated with necroptosis. IL-1β is a highly potent proinflammatory mediator that stimulates the recruitment of neutrophils to the site of inflammation.^31,32 Moreover, platelets further boost the inflammasome and subsequent release of IL-1β from neutrophils.³³ Hence, neutrophil-derived IL-1β, in response to CC, serves to amplify the pathogenesis of focal crystalline TMA, exacerbating both immunothrombosis and necroinflammation. In contrast, our findings suggest that CC–induced NET formation occurs independently of GSDMD, whereas both histone and CC-activated platelets trigger NET formation in a GSDMD-dependent manner. Consequently, the observed reduction in kidney tissue and intravascular NETs detected in Gsdmd^−/− mice with focal crystalline TMA could arise from the decreased levels of histone released by necrotic tubular cells, reduced neutrophil infiltration, and limited NET formation. Previous studies have reported that caspase-11 or neutrophil elastase can activate GSDMD, thereby promoting NET formation,^8,9 and that GSDMD pores can elevate cytoplasmic calcium concentrations and activate protein arginine deiminase 4, leading to histone citrullination.⁹ However, because CC induces NETs in a protein arginine deiminase 4-independent manner,¹⁴ it is possible that GSDMD does not influence this process. Similarly, neutrophils lacking GSDMD remained capable of forming NETs when stimulated with phorbol 12-myristate 13-acetate or calcium ionophores.^18,34 Therefore, the role of GSDMD in NETosis is stimulus-dependent and requires further investigation.

We observed a decrease in the population of circulating and kidney-infiltrating mature neutrophils in response to focal TMA in Gsdmd^−/− mice, whereas the number of neutrophils in the bone marrow and spleen, sites for emergency granulopoiesis during inflammatory conditions,^35,36 and lungs, in which neutrophils emigrate from inflammation sites before redirecting to the bone marrow for withdrawal,³⁷ remained unchanged. This suggests that the reduced neutrophil numbers in the blood and kidneys of Gsdmd^−/− mice with focal TMA did not arise from an alteration in neutrophil development or homing. Indeed, under homeostatic conditions, GSDMD deficiency had no impact on neutrophil numbers and maturation in vivo, as well as during iPSC–derived neutrophil differentiation. Nevertheless, Gsdmd deficiency reduced the number of mature neutrophils in the bone marrow, spleen, and blood after focal TMA. The expression of CXCR2 was diminished in Gsdmd^−/− bone marrow neutrophils in response to maturation factor granulocyte colony–stimulating factor or necrosis factor alpha after 24 hours of in vitro stimulation, indicating an active involvement of GSDMD in neutrophil maturation under inflammatory conditions. This impaired neutrophil maturation could ultimately lead to reduced CXCR2 and β₂ integrin-mediated²³ migration of mature neutrophils into the blood and kidneys of Gsdmd^−/− mice after focal TMA. However, several questions remain unanswered: when do neutrophils acquire GSDMD proteins? How does GSDMD regulate β₂-integrin activation? Does GSDMD also modulate aging or reverse migration processes? Given the complex and heterogenous nature of neutrophil maturation,^20,38 further investigation is needed to elucidate the precise role of GSDMD in this process.

Considering the role of GSDMD as the final downstream effector in the inflammasome and pyroptosis/IL-1β pathways, targeting GSDMD is a potentially effective and precise strategy. DSF, originally used in the management of chronic alcohol addiction by acting on aldehyde dehydrogenase, has also been recognized as an inhibitor of GSDMD through high-throughput biochemical screening. It covalently modifies Cys191 and blocks GSDMD pore formation, thereby preventing pyroptosis and the release of IL-1β.³⁹ Notably, DSF protected mice against sepsis induced by LPS³⁹ as well as cecal ligation and puncture.¹⁰ Our study demonstrated the prophylactic and therapeutic effects of DSF on CC-induced TMA and its consequences, suggesting a potential therapeutic approach, including administration of the drug before catheterization, a major trigger for CC embolism as a form of focal crystalline TMA. Given that CC-induced TMA and various forms of TMA share a common pathogenesis initially induced by immunothrombosis, inhibiting GSDMD holds potential for ameliorating TMA in a broader context.

Our study has a series of limitations. Although we focused on investigating the role of GSDMD in neutrophils, Gsdmd^−/− mice do not fully exclude the possibility that GSDMD also contributes to the overall phenotype of other cell types. Using cell type-specific tools will be necessary to address these aspects in detail. Moreover, DSF, used as a tool to study GSDMD effects in both in vitro and in vivo settings, may not possess absolute selectivity as an inhibitor, thus allowing for potential off-target effects. Notably, the animal model used in this study represents a focal TMA within the kidney, which may not fully replicate all hematological features of systemic TMA, such as hemolytic anemia.

Together, GSDMD contributes to the development of crystalline TMA in the kidney and its consequences, ischemic kidney infarction, and kidney failure (supplemental Figure 16). Because GSDMD actively contributes to neutrophil maturation and the induction of inflammatory pyroptosis in neutrophils, which prominently evolve within only a few hours after an embolic event, prophylactic and therapeutic GSDMD blockade can effectively prevent focal TMA and its associated consequences. The same process should be applied to other organs in the body affected by the same disease and may applied to systemic forms of TMA, which share a common pathogenesis that initially triggers immunothrombosis with subsequent ischemic necroinflammation and organ failure.

The authors thank Yvonne Minor and Uschi Kögelsperger for animal husbandry, and Janina Mandelbaum and Anna Anfimiadou for expert technical support.

This study was supported by the Deutsche Forschungsgemeinschaft (DFG) (TRR332, projects A2, A7, B1, Z1, AN372/20-2, and 30-1). K.W.-K. was supported by the Takeda Science Foundation. The experiments were performed using a Nikon Eclipse Ti2 microscope system (funded by DFG [GZ:INST 86/1851-1 FUGG] to Martha Merrow).

Contribution: H.-J.A. and S.S. provided funding and supervision; K.W.-K., S.S., and H.-J.A. conceptualized the study; K.W.-K. and S.S. developed the methodology; K.W.-K. conducted the investigation; K.W.-K., C.L., J.K., H.L., C.H., and M.K. analyzed the data; K.W.-K., E.M.-B., K.V.A., M.R., and O.S. contributed to the visualization of the data; M.I.L. and C.K. contributed to the differentiation of induced pluripotent stem cell–derived neutrophils with gene editing; K.W.-K., C.H., A.B., and E.M.-B. performed the flow chamber experiments; T.S.B.H., D.Z., and Y.K. provided the protocol and suggestions; A.L. provided resources; K.W.-K. and H.-J.A. wrote the original draft of the manuscript; and all authors read and revised the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Hans-Joachim Anders, Medizinische Klinik und Poliklinik IV, LMU Klinikum, Ziemssenstr 5, 80336 Munich, Germany; email: hjanders@med.uni-muenchen.de.