elifeeLifeeLife2050-084XeLife Sciences Publications, Ltd9132910.7554/eLife.9132910.7554/eLife.91329.3version of recordShort ReportImmunology and InflammationInhibiting NINJ1-dependent plasma membrane rupture protects against inflammasome-induced blood coagulation and inflammationCuiJian1†LiHua1YeDien1ZhangGuoying2ZhangYan2YangLing2SimMartha MS3‡WoodJeremy P134WeiYinan2LiZhenyu2WuCongqinghttps://orcid.org/0000-0001-7725-7881cwu3@uky.edu156https://ror.org/02vm5rt34Saha Cardiovascular Research Center, College of Medicine, University of KentuckyLexingtonUnited Stateshttps://ror.org/02k3smh20Department of Pharmaceutical Sciences, Irma Lerma Rangel School of Pharmacy, Texas A&M UniversityCollege StationUnited Stateshttps://ror.org/02k3smh20Department of Molecular and Cellular Biochemistry, University of KentuckyLexingtonUnited Stateshttps://ror.org/01f5ytq51The Gill Heart and Vascular Institute, College of Medicine, University of KentuckyLexingtonUnited Stateshttps://ror.org/02k3smh20Department of Microbiology, Immunology, and Molecular Genetics, College of Medicine, University of KentuckyLexingtonUnited Stateshttps://ror.org/00ek9j693Department of Surgery, College of Medicine, University of KentuckyLexingtonUnited StatesAbu-AmerYousefReviewing Editorhttps://ror.org/01yc7t268Washington University in St. LouisUnited StatesRothlinCarla VSenior Editorhttps://ror.org/03v76x132Yale UniversityUnited States
Department of Medicine, Genetic Medicine Division, Vanderbilt University Medical Center, Nashville, United States
Bloodworks Northwest Research Institute, Seattle, United States
Systemic blood coagulation accompanies inflammation during severe infections like sepsis and COVID. We previously established a link between coagulopathy and pyroptosis, a vital defense mechanism against infection. During pyroptosis, the formation of gasdermin-D (GSDMD) pores on the plasma membrane leads to the release of tissue factor (TF)-positive microvesicles (MVs) that are procoagulant. Mice lacking GSDMD release fewer of these procoagulant MVs. However, the specific mechanisms coupling the activation of GSDMD to MV release remain unclear. Plasma membrane rupture (PMR) in pyroptosis was recently reported to be actively mediated by the transmembrane protein Ninjurin-1 (NINJ1). Here, we show that NINJ1 promotes procoagulant MV release during pyroptosis. Haploinsufficiency or glycine inhibition of NINJ1 limited the release of procoagulant MVs and inflammatory cytokines, and partially protected against blood coagulation and lethality triggered by bacterial flagellin. Our findings suggest a crucial role for NINJ1-dependent PMR in inflammasome-induced blood coagulation and inflammation.
NINJ1PyroptosisCoagulationMicrovesicleInflammasomeMacrophageSepsisInfectionResearch organismMousehttp://dx.doi.org/10.13039/100000050National Heart, Lung, and Blood InstituteR00 HL145117WuCongqinghttp://dx.doi.org/10.13039/100000050National Heart, Lung, and Blood InstituteR01 HL142640WeiYinanLiZhenyuhttp://dx.doi.org/10.13039/100000057National Institute of General Medical SciencesR01 GM132443WeiYinanLiZhenyuhttp://dx.doi.org/10.13039/100000050National Heart, Lung, and Blood InstituteR01 HL146744LiZhenyuhttp://dx.doi.org/10.13039/100000050National Heart, Lung, and Blood InstituteR00 HL129193WoodJeremy PThe funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.Author impact statementProcoagulant microvesicles derived from ruptured immune cells fuel blood coagulation.publishing-routeprcIntroduction
Inflammasomes are intracellular multiprotein complexes that detect pathogens by sensing pathogen-associated molecular patterns (PAMPs) (Schroder and Tschopp, 2010; Lamkanfi and Dixit, 2014). The activation of inflammasome and subsequent pyroptosis are critical in host defense against pathogens (Miao et al., 2010; Zhao et al., 2011; von Moltke et al., 2012). Pyroptosis exposes intracellular pathogens through plasma membrane rupture (PMR) (Nozaki et al., 2022). However, overactive inflammasomes, common in conditions like sepsis and COVID (Yap et al., 2020; Wu et al., 2021; Vora et al., 2021; Sefik et al., 2022; Junqueira et al., 2022), can trigger systemic blood coagulation and tissue and organ dysfunction (Levi and Ten Cate, 1999; Levi et al., 2020; Al-Samkari et al., 2020; Giannis et al., 2020; Bowles et al., 2020).
Tissue factor (TF) is a membrane protein and the primary initiator of the coagulation cascade, essential for hemostasis to stop bleeding upon injury (Morrissey et al., 1987; Mackman, 2004). Elevated TF activity is associated with consumption coagulopathy in sepsis and COVID (Taylor et al., 1991; Levi et al., 1994; Pawlinski et al., 2010; Rosell et al., 2021). Our previous study has shown that TF-positive MVs released by pyroptotic macrophages trigger systemic coagulation and lethality in experimental sepsis (Wu et al., 2019). However, the molecular mechanisms by which pyroptosis releases procoagulant MVs remain elusive. Pyroptotic cells undergo PMR at the late stage when the plasma membrane is beyond repair, where the blebbing and pinching of the plasma membrane releases MVs to remove membrane damaged by gasdermin-D (GSDMD) pores (Ding et al., 2016; Liu et al., 2016; de Vasconcelos et al., 2019; Babiychuk et al., 2011; Rühl et al., 2018). PMR could accelerate MV release as the entire plasma membrane is subject to being dissolved as MVs. Kayagaki et al. recently reported that Ninjurin-1 (NINJ1) oligomerization activates PMR, a process formerly thought to be passive, during lytic cell death like necrosis and pyroptosis (Kayagaki et al., 2021). However, the role of NINJ1 in releasing procoagulant MVs and initiating systemic coagulation after inflammasome activation remains unknown.
In this study, we show that Ninj1 haploinsufficiency inhibits the release of TF-positive MVs and provides protection against systemic coagulation. Reduced NINJ1 also limits the release of inflammatory cytokines. We find that a single dose of glycine injection has similar protective effects as they prevent PMR. Our findings underscore the essential role of NINJ1-dependent PMR in inflammasome-induced blood coagulation, thus expanding our understanding of the dual role that inflammasomes play in host defense.
Results and discussion<italic>Ninj1</italic> haploinsufficiency partially protects against flagellin-induced blood coagulation
We aimed to understand the role of NINJ1 on inflammasome-induced blood coagulation, using heterozygous Ninj1+/- mice with decreased NINJ1 protein abundance (Figure 1—figure supplement 1). This approach was justified by our observation that total NINJ1 deficiency in the mouse strain used in this study caused the reported frequent occurrence of hydrocephalus (Kayagaki et al., 2023).
To induce blood coagulation and lethality, we applied our established protocol that involves injection of mice with purified recombinant bacterial flagellin, fused with the cytosolic translocation domain of anthrax lethal factor (LFn). In the presence of anthrax protein protective agent (PA), this flagellin fusion protein (LFn-Fla) can be transported into the cytoplasm (Zhao et al., 2011; Milne et al., 1995). Flagellin is a PAMP known to activate the NAIP/NLRC4 inflammasome and trigger pyroptosis (Miao et al., 2010; Zhao et al., 2011; Lightfield et al., 2008; Kofoed and Vance, 2011). We confirmed robust inflammasome activation and pyroptosis in mouse primary bone marrow-derived macrophages (BMDMs) upon stimulating with purified Fla (LFn-Fla plus PA) (Figure 1—figure supplement 2A–B).
Our earlier work has shown that inflammasome-induced systemic coagulation is marked by prolonged prothrombin time (PT) and increased plasma thrombin-antithrombin (TAT) (Wu et al., 2019), frequently observed in patients with disseminated intravascular coagulation and COVID (Levi and Ten Cate, 1999; Levi et al., 2020; Al-Samkari et al., 2020; Giannis et al., 2020; Bowles et al., 2020; Iba et al., 2019). As expected, intravenous injection of Fla activated systemic coagulation in Ninj1+/+ mice, demonstrated by a significant increase of PT and plasma TAT (Figure 1A and B). Notably, when challenged with Fla, heterozygous Ninj1+/- mice showed lower PT and almost complete suppression of the elevation in plasma TAT compared with their wild-type littermates (Figure 1A and B). Our results suggest that a single copy deletion of Ninj1 is adequate for significant inhibition of inflammasome-induced blood coagulation.
NINJ1 is critical for flagellin-induced systemic coagulation, inflammation, and lethality.
(A–E) Mice were injected intravenously with Ctrl (PBS) or Fla (500 ng LFn-Fla plus 3 μg protective agent [PA]). Blood was collected 90 min after Ctrl or Fla injection. Prothrombin time (A), plasma thrombin-antithrombin (TAT) (B), and plasma cytokines (C–E) were measured. Circles represent individual mice, with bars denoting means. **p<0.01 (two-way ANOVA with Holm-Sidak multiple comparisons). (F–G) Mice were injected intravenously with Ctrl or Fla. After 90 min, mice were euthanized and perfused with PBS, and tissues were isolated. (F) Lung sections were stained with the anti-fibrin monoclonal antibody (59D8). Scale bar denotes 20 μm. (G) Fibrin in the liver and lungs was detected by immunoblot with the anti-fibrin monoclonal antibody (59D8). (H) Mice were injected intravenously with a lethal dose of Fla (2.5 μg LFn-Fla plus 6 μg PA). Kaplan-Meier survival plots for mice challenged with Fla are shown. n=7–9. ****p<0.0001 versus WT (log rank test [Mantel-Cox]).
Excel file containing numeric values for panels A-E and H.
PDF file containing uncropped western blots with labeling for panel G.
Original tiff files of western blots for panel G.
NINJ1 protein abundance at baseline in different tissues.
Protein was extracted from fresh frozen tissues and detected by immunoblot. Band intensities are not comparable between different tissues.
PDF file containing uncropped western blots with labeling.
Original tiff files of western blots.
Flagellin-induced inflammasome activation and pyroptosis.
Bone marrow-derived macrophages (BMDMs) from Ninj1+/+ mice were incubated with LFn-Fla (1 μg/mL) and/or protective agent (PA) (1 μg/mL) for 90 min. (A) Plasma membrane rupture (PMR) was measured by lactate dehydrogenase (LDH) release. (B) Caspase-1, IL-1β, and gasdermin-D (GSDMD) in the cell lysates and culture supernatant was detected by immunoblot. Circles represent individual mice, with bars denoting means. **p<0.01 versus each of the other three groups (Student’s t-test; unpaired).
Excel file containing numeric values for panel A.
PDF file containing uncropped western blots with labeling for panel B.
Original tiff files of western blots for panel B.
Flagellin-induced tissue fibrin deposition (H&E staining) in <italic>Ninj1<sup>+/+</sup></italic> and <italic>Ninj1<sup>+/-</sup></italic> mice.
Mice were injected intravenously with Ctrl (PBS) or Fla (500 ng LFn-Fla plus 3μg protective agent [PA]). Tissues was collected 90 min after Ctrl or Fla injection. H&E staining was performed on PBS-perfused and paraffin-embedded lung tissue sections.
We also found that Ninj1 haploinsufficiency hindered the release of cytokines, as shown by decreased plasma concentrations of cytokines such as IL-1β, TNFα, and IL-6 (Figure 1C–E). The decrease in cytokine concentrations could be attributed to limited release of danger-associated molecular patterns (DAMPs) due to less effective PMR in Ninj1+/- mice. It’s worth noting that GSDMD facilitates IL-1β secretion into cell culture supernatant and NINJ1 deficiency does not affect pyroptosis-induced IL-1β release in vitro (Kayagaki et al., 2021; Borges et al., 2022). Given that Ninj1 haploinsufficiency did not change GSDMD expression (Figure 1—figure supplement 1), there are likely unrecognized mechanisms in pyroptosis-induced IL-1β secretion in vivo.
While our earlier research suggested that cell death, not the release of inflammatory cytokines like IL-1β, is crucial for the over-activation of blood coagulation triggered by inflammasomes (Wu et al., 2019), these cytokines are vital for maintaining inflammation. Our findings indicate their release is affected by NINJ1.
Fibrin deposition and lethality resulted from inflammasome activation depend on NINJ1
Fibrin deposition in tissues can restrict blood supply, potentially resulting in tissue and organ damage and even death (Levi and Ten Cate, 1999). Following the Fla challenge, we noted fibrin deposition in the lungs of wild-type mice by immunostaining with the fibrin-specific antibody 59D8 (Hui et al., 1983; Figure 1F), which was consistent with H&E staining (Figure 1—figure supplement 3). However, no signs of fibrin deposition were observed in the lung sections of Ninj1+/- mice, likely due to the technical limitations of detection on thin (5 μm) tissue sections, compounded by the potential of uneven fibrin distribution. Using immunoblot to examine a homogenized whole lobe of the lung or liver, we detected fibrin deposition in the liver and lungs of Ninj1+/- mice, but to a much less extent compared with their Ninj1+/+ littermates (Figure 1G). Furthermore, Ninj1 haploinsufficiency partially rescued flagellin-induced lethality (Figure 1H). These findings underscore the significant role of NINJ1 protein in inflammasome-triggered fibrin deposition in tissues and lethality.
NINJ1 plays a critical role in systemic coagulation in response to <italic>Escherichia coli</italic> infection
Systemic blood coagulation is a common complication of pathogen-triggered sepsis (Antoniak, 2018; Tang et al., 2021). To investigate the role of NINJ1 in bacterial infection-induced blood coagulation, Ninj1+/- mice were infected with cultured E. coli at a dose of 2×108 pfu per mouse. Consistently with the findings from the flagellin challenge, E. coli infection-induced blood coagulation was limited by Ninj1 haploinsufficiency. As shown in Figure 2A and B, Ninj1+/- mice exhibited a lower PT and decreased plasma TAT following E. coli infection. Furthermore, Ninj1+/- mice had less cytokine release in response to E. coli infection (Figure 2C–E). Collectively, these data demonstrate that reduced expression of NINJ1 confers protection against E. coli infection-induced blood coagulation. In addition, the reduction in cytokine release in Ninj1+/- mice supports the role of NINJ1 in modulating the inflammatory response to E. coli infection. These findings highlight a potential role of NINJ1 in the pathogenesis of bacterial infection, including blood coagulation and the cytokine storm.
<italic>E. coli</italic> infection-induced blood coagulation is limited in <italic>Ninj</italic>1<sup>+/-</sup> mice.
(A–E) Mice were injected intraperitoneally with Ctrl (saline) or E. coli (2×108 cfu per mouse). Blood was collected 6 hr afterward. Prothrombin time (A), plasma thrombin-antithrombin (TAT) (B), and plasma cytokines (C–E) were measured. Circles represent individual mice, with bars denoting means. *p<0.05, **p<0.01 (two-way ANOVA with Holm-Sidak multiple comparisons).
Excel file containing numeric values for panel A-E.
We have established that macrophage pyroptosis promotes the release of procoagulant MVs (Wu et al., 2019). Yet, the cellular processes coupling pyroptosis to MV release remain to be defined. We hypothesize that PMR during pyroptosis is the molecular event responsible for the release of these MVs. To test this hypothesis, we used BMDMs isolated from Ninj1+/+ and Ninj1+/- littermates. We observed a significant reduction in PMR, indicated by the decreased lactate dehydrogenase (LDH) in the cell culture medium of Ninj1+/- cells (Figure 3—figure supplement 1A), while cell death, measured by ATP levels, remained unaffected (Figure 3—figure supplement 1B). This is consistent with previous studies suggesting that NINJ1 mediates PMR but not cell death (Kayagaki et al., 2021; Borges et al., 2022).
Notably, we detected NINJ1 protein in MVs released from Ninj1+/+ BMDMs (Figure 3A). The protein level was decreased in MVs derived from Ninj1+/- BMDMs (Figure 3A), likely due to reduced MV release and/or less NINJ1 protein in Ninj1+/- cells (Figure 3B, Figure 3—figure supplement 1C). Consistently, NINJ1 was detected in plasma MVs from Ninj1+/+ mice challenged with Fla (Figure 3D). Ninj1+/- mice also had decreased amount of NINJ1 from MVs, likely due to fewer plasma MVs (Figure 3C). The reduced quantity of MVs from both plasma and BMDMs in Ninj1+/- mice correlated with lower TF activity (Figure 3E and F). Our data illustrate that a decline in NINJ1 expression significantly reduces PMR and procoagulant MV release. These findings provide compelling evidence supporting a critical role of NINJ1-dependent PMR in procoagulant MV release. Furthermore, MV shedding removes NINJ1 from the plasma membrane, hinting at a possible role of NINJ1 in cell repair.
Plasma membrane rupture (PMR) promotes the release of procoagulant microvesicles (MVs).
(A–B, E) Bone marrow-derived macrophages (BMDMs) were incubated with Ctrl (PBS) or Fla (1 μg/mL LFn-Fla plus 1 μg/mL protective agent [PA]). Cell culture supernatant and MVs were collected after 90 min of incubation. (A) NINJ1 in cell lysates and MVs was detected by immunoblot. (B) BMDM MVs were counted with NanoSight. (E) BMDM MV tissue factor (TF) activity. Circles represent individual mouse, with bars denoting means. **p<0.01 (two-way ANOVA with Holm-Sidak multiple comparisons). (C, D, F) Mice were injected intravenously with Ctrl (PBS) or Fla (500 ng LFn-Fla plus 3 μg PA). Blood was collected 90 min after Ctrl or Fla injection. (C) Plasma MVs were counted with NanoSight. (D) NINJ1 in plasma MVs isolated from equal volume of plasma was detected by immunoblot. (F) Plasma MV TF activity. Circles represent individual mice, with bars denoting means. **p<0.01 (two-way ANOVA with Holm-Sidak multiple comparisons).
PDF file containing uncropped western blots with labeling for panels A and D.
Original tiff files of western blots for panels A and D.
Excel file containing numeric values for panels B, C, E, and F.
Flagellin-induced pyroptosis and microvesicle (MV) release in <italic>Ninj1<sup>+/+</sup></italic> and <italic>Ninj1<sup>+/-</sup></italic> bone marrow-derived macrophages (BMDMs).
BMDMs from Ninj1+/+ and Ninj1+/- mice were incubated with Ctrl (PBS) or Fla (1 μg/mL LFn-Fla plus 1 μg/mL protective agent [PA]) for 90 min. Lactate dehydrogenase (LDH) release (A) and ATP (B) were measured. Circles represent individual mice, with bars denoting means. **p<0.01, n.s. denotes not significant (two-way ANOVA with Holm-Sidak multiple comparisons). (C) BMDM MV quantity and size were analyzed with NanoSight.
Excel file containing numeric values for panels A-C.
Glycine inhibition of NINJ1 oligomerization reduces pyroptosis-induced blood coagulation
It has long been recognized that glycine buffering protects against PMR during cell death, such as pyroptosis and necrosis (Weinberg et al., 1987; Estacion et al., 2003). The exact mechanisms underlying the glycine protective effects remained unclear until Borges et al. revealed that glycine inhibits NINJ1 oligomerization and shields cells from rupture (Borges et al., 2022). Yet, whether glycine buffering would block the release of PMR-induced TF-positive MVs or limit pyroptosis-induced coagulation in vivo is unknown. To complement our genetic approach, we injected a glycine solution into wild-type mice, prior to Fla injection, aiming to inhibit NINJ1.
Intriguingly, glycine injection led to a significant decrease in PT and plasma TAT (Figure 4A and B), demonstrating significant protection against flagellin-induced blood coagulation. Consistently, mice injected with glycine exhibited lower plasma MV TF activity post the Fla challenge (Figure 4C). In vitro, glycine buffering on isolated BMDMs protected against PMR, indicated by the reduction in LDH release (Figure 4—figure supplement 1A) and decreased MV TF activity (Figure 4—figure supplement 1B). However, we observed a discrepancy in MV counts between the in vitro and in vivo studies. Contrary to isolated BMDMs (Figure 4—figure supplement 1C), glycine-injected mice had higher MV counts irrespective of the Fla challenge (Figure 4D). We speculate that glycine might enhance MV release from cell types other than monocytes and macrophages, independently of inflammasome activation. Lastly, glycine administration reduced the release of proinflammatory cytokines IL-1β and TNFα, consistent with previous research (Yang et al., 2001), but not IL-6 (Figure 4E–G). In conclusion, the cytoprotective effect of glycine buffering resulted in improved outcomes in inflammasome-induced blood coagulation and inflammation.
Glycine inhibition of NINJ1 blocks pyroptosis-induced blood coagulation.
Mice were injected intravenously 50 μL of 0.5 M glycine 2 hr before administrating Ctrl (PBS) or Fla (500 ng LFn-Fla plus 3 μg protective agent [PA]). Blood was collected 90 min after Ctrl or Fla injection. Prothrombin time (A), plasma thrombin-antithrombin (TAT) (B), plasma tissue factor (TF) microvesicle (MV) activity (C), plasma MV counts (D), and plasma cytokines (E–G) were measured. Circles represent individual mice, with bars denoting means. *p<0.05, **p<0.01, n.s. denotes not significant (two-way ANOVA with Holm-Sidak multiple comparisons).
Excel file containing numeric values for panels A-G.
Glycine treatment inhibits bone marrow-derived macrophage (BMDM) microvesicle (MV) release.
(A–C) BMDMs from Ninj1+/+ mice were treated with 5 mM glycine for 30 min before incubation with Ctrl (PBS) or Fla (1 μg/mL LFn-Fla plus 1 μg/mL protective agent [PA]). Cell culture supernatant was collected after 90 min incubation. Lactate dehydrogenase (LDH) release (A), BMDM MV tissue factor (TF) activity (B), and MV counts (C) were measured. Circles represent individual mouse, with bars denoting means. **p<0.01, n.s. denotes not significant (two-way ANOVA with Holm-Sidak multiple comparisons).
Excel file containing numeric values for panels A-G.
In summary, our study employed both genetic and pharmacological approaches to inhibit NINJ1-dependent PMR. Our findings suggest that PMR serves as a pivotal molecular event responsible for the release of procoagulant MVs and the induction of systemic coagulation and inflammation. This novel mechanism sheds light on the potential for developing NINJ1 inhibition as a promising therapeutic strategy for reducing acute inflammation and coagulation-associated tissue and organ damage. In fact, an anti-NINJ1 monoclonal antibody (but not the NINJ126-37 peptide) has been shown to prevent PMR, limiting tissue inflammation in mice (Kayagaki et al., 2023). We do, however, recognize the limitations of our main model, as a single injection of bacterial flagellin does not replicate the evolving nature of a polymicrobial infection. Consequently, it cannot offer insights into extended timelines beyond the brief period of flagellin-induced pyroptosis.
Materials and methods
Reagent type (species) or resource
Designation
Source or reference
Identifiers
Additional information
Genetic reagent (Mus musculus)
Ninj1 knockout
Genentech, Inc;Kayagaki et al., 2021
C57Bl/6 strain background
Biological sample (Mus musculus)
Primary bone marrow-derived macrophages (Ninj1+/+ BMDMs)
This paper
Freshly prepared from Ninj1+/+ male mice at 8–12 weeks of age, described in the Materials and methods section
Biological sample (Mus musculus)
Primary bone marrow-derived macrophages (Ninj1+/- BMDMs)
This paper
Freshly prepared from Ninj1+/- male mice at 8–12 weeks of age, described in the Materials and methods section
Strain and strain background (Escherichia coli)
E. coli strain (EC10)
Dr. Kwang Sik KimDivision of Pediatric Infectious Diseases, Department of Pediatrics, Johns Hopkins University School of Medicine
Gift from Dr. Hartmut Weiler (Medical College of Wisconsin) and Dr. Rodney M Camire (University of Pennsylvania)
59D8
Western blot, 1:2000;Immunohistochemistry, 1:250
Commercial assay or kit
CellTiter-Glo Luminescent Cell Viability Assay
Promega
G7572
Commercial assay or kit
CytoTox 96 Non-Radioactive Cytotoxicity Assay
Promega
G1780
Commercial assay or kit
Mouse TAT ELISA kit
Abcam
ab137994
Commercial assay or kit
Mouse IL-1β ELISA kit
Thermo Fisher Scientific
88-7013A-88
Commercial assay or kit
Mouse IL-6 ELISA kit
Thermo Fisher Scientific
88-7064-22
Commercial assay or kit
Mouse TNFα ELISA kit
Thermo Fisher Scientific
88-7324-22
Peptide, recombinant protein
Protective agent (PA)
Zhenyu Li LaboratoryYinan Wei Laboratory
Described in the Materials and methods section
Peptide, recombinant protein
LFn-flagellin
Zhenyu Li LaboratoryYinan Wei Laboratory
Described in the Materials and methods section
Chemical compound and drug
HisPur Ni-NTA resin
Thermo Fisher Scientific
88222
Chemical compound and drug
Thromboplastin-D
Pacific Hemostasis
100357
Chemical compound and drug
RGR-XaChrom
Enzyme Research Laboratories
100-03
Chemical compound and drug
Glycine
Sigma
G7126
Chemical compound and drug
LPS (E. coli O111:B4)
Sigma
L4130
Chemical compound and drug
T-PER tissue protein extraction reagent
Thermo Fisher Scientific
78510
Chemical compound and drug
Cocktail inhibitor
Sigma
P8340
Software and algorithm
ImageStudio 5.0
Li-COR
Western blot image
Software and algorithm
Prism 9
GraphPad Software Inc
Other
M.O.M. (Mouse on Mouse) ImmPRESS HRP (Peroxidase) Polymer Kit
Vector Laboratories
MP-2400
Immunohistochemistry
Mice
C57BL/6J, Ninj1+/+, and Ninj1+/- mice were housed in the University of Kentucky Animal Care Facility, following institutional and National Institutes of Health guidelines after approval by the Institutional Animal Care and Use Committee.
Recombinant protein purification
Recombinant proteins were expressed using LPS-free E. coli strain ClearColi BL21(DE3) (Lucigen Corporation, Cat#60810) at 37°C for 4 hr with 500 μM IPTG after OD600 reached 0.6–0.8. Bacteria were collected and lysed in 50 mM Tris-HCl and 300 mM NaCl. Proteins containing a His-tag were purified by affinity chromatography using HisPur Ni-NTA resin (Thermo Fisher Scientific, Cat#88222). Proteins were then eluted with 250 mM imidazole in 50 mM Tris-HCl and 300 mM NaCl, and subsequently dialyzed against PBS to remove imidazole. Proteins concentrations were determined by A280 before sterile filtration.
In vivo challenges
For flagellin challenge, purified PA and LFn-Fla in PBS were administered via retro-orbital injection. Blood was collected at 90 min post-injection, followed by PBS perfusion. Subsequently, tissue samples were harvested. For E. coli challenge, mice were injected intraperitoneally with E. coli in saline solution. Blood was collected at 6 hr post-injection. For glycine in vivo experiment, mice were injected intravenously with glycine saline solution 2 hr before flagellin challenge. Blood was collected at 90 min after PBS or flagellin injection.
BMDM isolation and differentiation
Mouse femur and tibia from one leg were harvested and rinsed in ice-cold PBS, followed by a brief rinse in 70% ethanol for 10–15 s. Both ends of the bones were cut open, and the bone marrow was flushed out using a 10 mL syringe with a 26-gauge needle. The marrow was passed through a 19-gauge needle once to disperse the cells. After filtering through a 70 μm cell strainer, cells were collected by centrifugation at 250×g for 5 min at 4°C, then suspended in two 150 mm Petri dishes, each containing 25 mL of L-cell conditioned medium (RPMI-1640 supplemented with 10% fetal bovine serum [FBS], 2 mM L-glutamine, 10 mM HEPES, 15% LCM, and penicillin/streptomycin). After 3 days, 15 mL of LCM medium was added to each dish. The cells typically reached full confluency after 5–7 days.
BMDM culture and stimulation
BMDMs were seeded onto 12-well cell culture plate or 96-well cell culture plate at a density of 1×106 cells/mL of RPMI-1640 medium (Thermo Fisher Scientific, Cat#61870036) containing 10% of FBS (Thermo Fisher Scientific, Cat#A3160502). BMDMs were allowed to settle overnight and refreshed with Opti-MEM (Thermo Fisher Scientific, Cat#51985034) before purified proteins were added. To measure cell cytotoxicity and viability and MVs, supernatant was collected 90 min after stimulation.
Cell viability and cytotoxicity
Cell viability of BMDMs was determined using CellTiter-Glo Luminescent Cell Viability Assay (Promega, Cat#G7572). Luminescence was recorded as an indicator of ATP levels in metabolically active cells. BMDMs cell cytotoxicity, as measured by LDH in the cell culture supernatant, was determined using CytoTox 96 Non-Radioactive Cytotoxicity Assay (Promega, Cat#G1780) according to the manufacturer’s instruction.
Prothrombin time
Blood was collected from ketamine/xylazine-anesthetized mice by cardiac puncture with a 23-gauge needle attached to a syringe prefilled with 3.8% trisodium citrate as anticoagulant (final ratio at 1:10). Blood was centrifuged at 2,000×g for 20 min at 4°C to obtain plasma. PT was determined with Thromboplastin-D (Pacific Hemostasis, Cat#100357) in a manual setting according to the manufacturer’s instruction, using CHRONO-LOG #367 plastic cuvette.
Plasma TAT
Plasma was collected as mentioned above in PT. Plasma TAT concentrations were determined using a mouse TAT ELISA kit (Abcam, Cat#ab137994) at 1:50 dilution according to the manufacturer’s instruction.
ELISA
IL-1β, IL-6, and TNFα levels in culture supernatant and plasma were measured with ELISA kits from Thermo Fisher Scientific (#88-7013A-88 for IL-1 β, #88-7064-22 for IL-6, and #88-7324-22 for TNFα) according to the manufacturer’s instruction. Plates were read on a Cytation 5 at 450 and 570 nm.
Tissue preparation and immunohistochemistry
Mice were perfused via both right and left ventricles with PBS. Tissues were collected and embedded in paraffin, then sectioned at 5 μm. Anti-fibrin antibody 59D8 (kindly provided by Dr. Hartmut Weiler at Medical College of Wisconsin and Dr. Rodney M Camire at the University of Pennsylvania) at 4 mg/mL was used to detect fibrin deposition, with M.O.M. (Mouse on Mouse) ImmPRESS HRP (Peroxidase) Polymer Kit (Vector, Cat#MP-2400) from Vector Laboratories according to the manufacturer’s instruction for developing positive staining.
Fibrin extraction for immunoblot
Frozen tissues were homogenized in 20 volumes (mg:μL) of T-PER tissue protein extraction reagent (Thermo Fisher Scientific, Cat#78510) containing cocktail inhibitor (Sigma, Cat#P8340). After centrifugation at 10,000×g for 10 min, supernatant was collected for β-actin detection. Pellets were then homogenized in 3 M urea and vortexed for 2 hr at 37°C. After centrifugation at 14,000×g for 15 min, pellets were suspended in LDS sample buffer and vortexed at 65°C for 30 min and ready for fibrin detection.
Fluorescent immunoblot
For detection of Caspase-1, IL-1β, GSDMD, and NINJ1 by immunoblot, cells were washed with cold PBS and lysed with LDS sample buffer. Culture supernatant was precipitated with 1/10 volume of 2% sodium deoxycholate and 1/10 volume of 100% trichloroacetic acid, and then dissolved in LDS sample buffer. Caspase-1 was detected using anti-Caspase-1 (Adipogen, Cat#AG-20B-0042-C100) at 1:1000 dilution. IL-1β was detected using anti-IL-1β (GeneTex, Cat#GTX74034) at 1:1000 dilution. GSDMD was detected using anti-GSDMD (Abcam, Cat#ab219800) at 1:1000 dilution. NINJ1 was detected using anti-NINJ1 (kindly provided by Genentech, Cat#Ninj1-rbIgG-25:10363). Tissue fibrin was detected using anti-fibrin (59D8) at 1 mg/mL. Images were acquired with LI-COR Odyssey Imager. Western blot experiments were replicated three times with representative images shown in the figures (Figures 1G and 3A).
Isolation of MVs from mouse plasma
Plasma was collected as mentioned above in PT. Then, 50 μL of mouse plasma was diluted with 1 mL of HBSA (137 mM NaCl, 5.38 mM KCl, 5.55 mM glucose, 10 mM HEPES, 0.1% bovine serum albumin, pH 7.5). MVs were pelleted at 20,000×g for 60 min at 4°C, washed once with 1 mL of HBSA and resuspended in 100 μL HBSA.
Isolation of MVs from cell culture medium
Cell debris was removed from cell culture supernatant by centrifugation at 1000×g for 10 min. MVs were then pelleted at 20,000×g for 60 min at 4°C and resuspended in HBSA.
MV counting
To count MVs collected from cell culture supernatant, cell debris was removed by centrifugation at 1000×g for 10 min. MVs in cell culture supernatant or plasma were analyzed directly without further centrifugation. All samples were diluted to optimal conditions (about 108 particles/mL) for analysis in PBS. Video recordings were made for 10 intervals at a length of 30 s each using NanoSight. Nanoparticle tracking analysis was performed to measure the number and size of the plasma and culture supernatant MVs.
Plasma MV TF activity assay
Plasma (25 μL) was incubated with the 1H1 anti-TF antibody (Genentech) at 100 mg/mL or rat IgG controls for 15 min at room temperature. Next, 25 μL of HBSA containing 10 nM mouse FVIIa, 300 nM human FX, and 10 mM CaCl2 was added to the samples and incubated for 2 hr at 37°C in a half area 96-well plate. Finally, 12.5 μL of the FXa substrate RGR-XaChrom (4 mM, Enzyme Research Laboratories#100-03) was added and the mixture was incubated at 37°C for 15 min. Absorbance at 405 nm was measured on Cytation 5. The relative TF activity was calculated with absorbance after subtracting the TF-independent activity in the presence of TF blocking antibodies from the total activity in the presence of the IgG control.
Statistical analysis
Data are represented as individual dots, with bars denoting means. All data represent biological replicates unless stated otherwise. For multiple group with two independent factors, two-way ANOVA with Holm-Sidak multiple comparisons was used for normally distributed variables. A p-value<0.05 was considered significant. Statistical analyses were performed in GraphPad Prism 9.
Additional informationCompeting interests
No competing interests declared
has an investigator-initiated grant through Pfizer, which is unrelated to this project
Author contributions
Conceptualization, Investigation, Writing – original draft, Writing – review and editing
Investigation, Writing – review and editing
Investigation, Writing – review and editing
Investigation, Writing – review and editing
Investigation, Writing – review and editing
Investigation, Writing – review and editing
Investigation, Writing – review and editing
Writing – review and editing
Writing – review and editing
Writing – review and editing
Conceptualization, Supervision, Funding acquisition, Investigation, Writing – original draft, Writing – review and editing
Ethics
This study was performed in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All of the animals were housed and handled according to approved Institutional Animal Care and Use Committee (IACUC) protocol (#2021-3751) at the University of Kentucky.
Additional filesData availability
All data generated or analyzed during this study are included in the manuscript and supporting files, which includes the source data for the manuscript figures.
Acknowledgements
This work was supported by the National Institutes of Health R00 HL145117 to CW, R01 HL142640 and GM132443 to YW and ZL, R01 HL146744 to ZL, and R00 HL129193 to JPW. Dr. Wendy Katz provided help with tissue paraffin embedding and sectioning and was supported by NIH/NIGMS Institutional Development Award P20GM103527. Dr. Hartmut Weiler at Medical College of Wisconsin and Dr. Rodney M Camire at the University of Pennsylvania provided fibrin antibody. Ninj1+/- mice and NINJ1 antibody were provided by Genentech, Inc. Dr. Chris Richard provided help with MV counting on NanoSight. Dr. Xu Fu at the University of Kentucky Light Microscopy Core provided help on image acquisition.
ReferencesAl-SamkariHKarp LeafRSDzikWHCarlsonJCTFogertyAEWaheedAGoodarziKBendapudiPKBornikovaLGuptaSLeafDEKuterDJRosovskyRP2020COVID-19 and coagulation: bleeding and thrombotic manifestations of SARS-CoV-2 infectionBlood13648950010.1182/blood.202000652032492712AntoniakS2018The coagulation system in host defenseResearch and Practice in Thrombosis and Haemostasis254955710.1002/rth2.1210930046760BabiychukEBMonastyrskayaKPotezSDraegerA2011Blebbing confers resistance against cell lysisCell Death and Differentiation18808910.1038/cdd.2010.8120596076BorgesJPSætraRSRVolchukABuggeMDevantPSporsheimBKilburnBREvavoldCLKaganJCGoldenbergNMFloTHSteinbergBE2022Glycine inhibits NINJ1 membrane clustering to suppress plasma membrane rupture in cell deatheLife11e7860910.7554/eLife.7860936468682BowlesLPlattonSYarteyNDaveMLeeKHartDPMacDonaldVGreenLSivapalaratnamSPasiKJMacCallumP2020Lupus anticoagulant and abnormal coagulation tests in patients with covid-19The New England Journal of Medicine38328829010.1056/NEJMc201365632369280de VasconcelosNMVan OpdenboschNVan GorpHParthoensELamkanfiM2019Single-cell analysis of pyroptosis dynamics reveals conserved GSDMD-mediated subcellular events that precede plasma membrane ruptureCell Death and Differentiation2614616110.1038/s41418-018-0106-729666477DingJWangKLiuWSheYSunQShiJSunHWangDCShaoF2016Pore-forming activity and structural autoinhibition of the gasdermin familyNature53511111610.1038/nature1859027281216EstacionMWeinbergJSSinkinsWGSchillingWP2003Blockade of maitotoxin-induced endothelial cell lysis by glycine and L-alanineAmerican Journal of Physiology. Cell Physiology284C1006C102010.1152/ajpcell.00258.200212477666GiannisDZiogasIAGianniP2020Coagulation disorders in coronavirus infected patients: COVID-19, SARS-CoV-1, MERS-CoV and lessons from the pastJournal of Clinical Virology12710436210.1016/j.jcv.2020.10436232305883HuiKYHaberEMatsuedaGR1983Monoclonal antibodies to a synthetic fibrin-like peptide bind to human fibrin but not fibrinogenScience2221129113210.1126/science.66485246648524IbaTLevyJHWarkentinTEThachilJvan der PollTLeviMScientific and Standardization Committee on DIC, and the Scientific and Standardization Committee on Perioperative and Critical Care of the International Society on Thrombosis and Haemostasis2019Diagnosis and management of sepsis-induced coagulopathy and disseminated intravascular coagulationJournal of Thrombosis and Haemostasis171989199410.1111/jth.1457831410983JunqueiraCCrespoÂRanjbarSde LacerdaLBLewandrowskiMIngberJParryBRavidSClarkSSchrimpfMRHoFBeakesCMargolinJRussellNKaysKBoucauJDas AdhikariUVoraSMLegerVGehrkeLHendersonLAJanssenEKwonDSanderCAbrahamJGoldbergMBWuHMehtaGBellSGoldfeldAEFilbinMRLiebermanJ2022FcγR-mediated SARS-CoV-2 infection of monocytes activates inflammationNature60657658410.1038/s41586-022-04702-435385861KayagakiNKornfeldOSLeeBLStoweIBO’RourkeKLiQSandovalWYanDKangJXuMZhangJLeeWPMcKenzieBSUlasGPayandehJRoose-GirmaMModrusanZRejaRSagollaMWebsterJDChoVAndrewsTDMorrisLXMiosgeLAGoodnowCCBertramEMDixitVM2021NINJ1 mediates plasma membrane rupture during lytic cell deathNature59113113610.1038/s41586-021-03218-733472215KayagakiNStoweIBAlegreKDeshpandeIWuSLinZKornfeldOSLeeBLZhangJLiuJSutoELeeWPSchneiderKLinWSeshasayeeDBhangaleTChalouniCJohnsonMCJoshiPMossemannJZhaoSAliDGoldenbergNMSayedBASteinbergBENewtonKWebsterJDKellyRLDixitVM2023Inhibiting membrane rupture with NINJ1 antibodies limits tissue injuryNature6181072107710.1038/s41586-023-06191-537196676KofoedEMVanceRE2011Innate immune recognition of bacterial ligands by NAIPs determines inflammasome specificityNature47759259510.1038/nature1039421874021LamkanfiMDixitVM2014Mechanisms and functions of inflammasomesCell1571013102210.1016/j.cell.2014.04.00724855941LeviMten CateHBauerKAvan der PollTEdgingtonTSBüllerHRvan DeventerSJHackCEten CateJWRosenbergRD1994Inhibition of endotoxin-induced activation of coagulation and fibrinolysis by pentoxifylline or by a monoclonal anti-tissue factor antibody in chimpanzeesThe Journal of Clinical Investigation9311412010.1172/JCI1169348282778LeviMTen CateH1999Disseminated intravascular coagulationThe New England Journal of Medicine34158659210.1056/NEJM19990819341080710451465LeviMThachilJIbaTLevyJH2020Coagulation abnormalities and thrombosis in patients with COVID-19The Lancet. Haematology7e438e44010.1016/S2352-3026(20)30145-932407672LightfieldKLPerssonJBrubakerSWWitteCEvon MoltkeJDunipaceEAHenryTSunY-HCadoDDietrichWFMonackDMTsolisRMVanceRE2008Critical function for Naip5 in inflammasome activation by a conserved carboxy-terminal domain of flagellinNature Immunology91171117810.1038/ni.164618724372LiuXZhangZRuanJPanYMagupalliVGWuHLiebermanJ2016Inflammasome-activated gasdermin D causes pyroptosis by forming membrane poresNature53515315810.1038/nature1862927383986MackmanN2004Role of tissue factor in hemostasis, thrombosis, and vascular developmentArteriosclerosis, Thrombosis, and Vascular Biology241015102210.1161/01.ATV.0000130465.23430.7415117736MiaoEALeafIATreutingPMMaoDPDorsMSarkarAWarrenSEWewersMDAderemA2010Caspase-1-induced pyroptosis is an innate immune effector mechanism against intracellular bacteriaNature Immunology111136114210.1038/ni.196021057511MilneJCBlankeSRHannaPCCollierRJ1995Protective antigen-binding domain of anthrax lethal factor mediates translocation of a heterologous protein fused to its amino- or carboxy-terminusMolecular Microbiology1566166610.1111/j.1365-2958.1995.tb02375.x7783638MorrisseyJHFakhraiHEdgingtonTS1987Molecular cloning of the cDNA for tissue factor, the cellular receptor for the initiation of the coagulation protease cascadeCell5012913510.1016/0092-8674(87)90669-63297348NozakiKLiLMiaoEA2022Innate sensors trigger regulated cell death to combat intracellular infectionAnnual Review of Immunology4046949810.1146/annurev-immunol-101320-01123535138947PawlinskiRWangJ-GOwensAPWilliamsJAntoniakSTencatiMLutherTRowleyJWLowENWeyrichASMackmanN2010Hematopoietic and nonhematopoietic cell tissue factor activates the coagulation cascade in endotoxemic miceBlood11680681410.1182/blood-2009-12-25926720410508RosellAHavervallSvon MeijenfeldtFHisadaYAguileraKGroverSPLismanTMackmanNThålinC2021Patients with covid-19 have elevated levels of circulating extracellular vesicle tissue factor activity that is associated with severity and mortality-brief reportArteriosclerosis, Thrombosis, and Vascular Biology4187888210.1161/ATVBAHA.120.31554733267656RühlSShkarinaKDemarcoBHeiligRSantosJCBrozP2018ESCRT-dependent membrane repair negatively regulates pyroptosis downstream of GSDMD activationScience36295696010.1126/science.aar760730467171SchroderKTschoppJ2010The inflammasomesCell14082183210.1016/j.cell.2010.01.04020303873SefikEQuRJunqueiraCKaffeEMirzaHZhaoJBrewerJRHanASteachHRIsraelowBBlackburnHNVelazquezSEChenYGHaleneSIwasakiAMeffreENussenzweigMLiebermanJWilenCBKlugerYFlavellRA2022Inflammasome activation in infected macrophages drives COVID-19 pathologyNature60658559310.1038/s41586-022-04802-135483404TangDWangHBilliarTRKroemerGKangR2021Emerging mechanisms of immunocoagulation in sepsis and septic shockTrends in Immunology4250852210.1016/j.it.2021.04.00133906793TaylorFBJrChangARufWMorrisseyJHHinshawLCatlettRBlickKEdgingtonTS1991Lethal E. coli septic shock is prevented by blocking tissue factor with monoclonal antibodyCirculatory Shock331271342044206von MoltkeJTrinidadNJMoayeriMKintzerAFWangSBvan RooijenNBrownCRKrantzBALepplaSHGronertKVanceRE2012Rapid induction of inflammatory lipid mediators by the inflammasome in vivoNature49010711110.1038/nature11351VoraSMLiebermanJWuH2021Inflammasome activation at the crux of severe COVID-19Nature Reviews. Immunology2169470310.1038/s41577-021-00588-x34373622WeinbergJMDavisJAAbarzuaMRajanT1987Cytoprotective effects of glycine and glutathione against hypoxic injury to renal tubulesThe Journal of Clinical Investigation801446145410.1172/JCI1132242890657WuCLuWZhangYZhangGShiXHisadaYGroverSPZhangXLiLXiangBShiJLiX-ADaughertyASmythSSKirchhoferDShiroishiTShaoFMackmanNWeiYLiZ2019Inflammasome activation triggers blood clotting and host death through pyroptosisImmunity501401141110.1016/j.immuni.2019.04.00331076358WuRWangNComishPBTangDKangR2021Inflammasome-dependent coagulation activation in sepsisFrontiers in Immunology1264175010.3389/fimmu.2021.64175033796108YangSKooDJChaudryIHWangP2001Glycine attenuates hepatocellular depression during early sepsis and reduces sepsis-induced mortalityCritical Care Medicine291201120610.1097/00003246-200106000-0002411395604YapJKYMoriyamaMIwasakiA2020Inflammasomes and Pyroptosis as Therapeutic Targets for COVID-19Journal of Immunology20530731210.4049/jimmunol.200051332493814ZhaoYYangJShiJGongY-NLuQXuHLiuLShaoF2011The NLRC4 inflammasome receptors for bacterial flagellin and type III secretion apparatusNature47759660010.1038/nature105102191851210.7554/eLife.91329.3.sa0eLife AssessmentAbu-AmerYousefReviewing Editorhttps://ror.org/01yc7t268Washington University in St. LouisUnited StatesSolidUseful
The authors aim to elucidate the mechanism by which pyroptosis (through the formation of Gasdermin D (GSDMD) pores in the plasma membrane) contributes to increased release of procoagulant Tissue Factor-containing microvesicles. The data offers solid mechanistic insights as to the interplay between pyroptosis and microvesicle release with NINJ1. The study provides useful insights into the potential of targeting NINJ1 as a therapeutic strategy.
The authors demonstrated that NINJ1 promotes TF-positive MV release during pyroptosis and thereby triggers coagulation. Coagulation is one of the risk factors that can cause secondary complications in various inflammatory diseases, making it a highly important therapeutic target in clinical treatment. This paper effectively explains the connection between pyroptosis and MV release with Ninj1, which is a significant strength. It provides valuable insight into the potential of targeting Ninj1 as a therapeutic strategy.
Although the advances in this paper are valuable, several aspects need to be clarified. Some comments are discussed below.
(1) Since it is not Ninj1 directly regulating coagulation but rather the MV released by Ninj1 playing a role, the title should include that. The current title makes it seem like Ninj1 directly regulates inflammation and coagulation. It would be better to revise the title.
(2) Ninj1 is known to be an induced protein that is barely expressed in normal conditions. As you showed in "Fig1G" data, control samples showed no detection of Ninj1. However, in "Figure S1", all tissues (liver, lung, kidney and spleen) expressed Ninj1 protein. If the authors stimulated the mice with fla injection, it should be mentioned in the figure legend.
(3) In "Fig3A", the Ninj1 protein expression was increased in the control of BMDM +/- cell lysate rather than fla stimulation. However, in MV, Ninj1 was not detected at all in +/- control but was only observed with Fla injection. The authors need to provide an explanation for this observation. Additionally, looking at the MV β-actin lane, the band thicknesses appear to be very different between groups. It seems necessary to equalize the protein amounts. If that is difficult, at least between the +/+ and +/- controls.
(4) Since the authors focused Ninj1-dependent microvesicle (MV) release, they need to show MV characterizations (EM, NTA, Western for MV markers, etc...).
(5) To clarify whether Ninj1-dependent MV induces coagulation, the authors need to determine whether platelet aggregation is reduced with isolated +/- MVs compared to +/+ MVs.
(6) Even with the authors well established experiments with haploid mice, it is a critical limitation of this paper. To improve the quality of this paper, the authors should consider confirming the findings using mouse macrophage cell lines, such as generating Ninj1-/- Raw264.7 cell lines, to examine the homozygous effect.
(7) There was a paper reported in 2023 (Zhou, X. et al., NINJ1 Regulates Platelet Activation and PANoptosis in Septic Disseminated Intravascular Coagulation. Int. J. Mol. Sci. 2023) that revealed the relationship between Ninj1 and coagulation. According to this paper, inhibition of Ninj1 in platelets prevents pyroptosis, leading to reduced platelet activation and, consequently, the suppression of thrombosis. How about the activation of platelets in Ninj1 +/- mice? The author should add this paper in the reference section and discuss the platelet functions in their mice.
The authors main goal is to understand the mechanism by which pyroptosis (through the formation of Gasdermin D (GSDMD) pores in the plasma membrane) contributes to increased release of procoagulant Tissue Factor-containing microvesicles (MV). Their previous data demonstrate that GSDMD is critical for the release of MV that contains Tissue Factor (TF), thus making a link between pyroptosis and hypercoagulation. Given the recent identification of NINJ1 being responsible for plasma membrane rupture (Kayagaki et al. Nature 2011), the authors wanted to determine if NINJ1 is responsible for TF-containing MV release. Given the constitutive ninj1 KO mouse leads to partial embryonic lethality, the authors decide to use a heterozygous ninj1 KO mouse (ninj1+/-), and demonstrate that Ninj1 plays a role in release of TF-containing MV.
10.7554/eLife.91329.3.sa3Author responseCuiJianAuthorVanderbilt UniversityNashvilleUnited StatesLiHuaAuthorUniversity of KentuckyLexingtonUnited StatesYeDienAuthorUniversity of KentuckyLexingtonUnited KingdomZhangGuoyingAuthorTexas A&M UniversityCollege StationUnited StatesZhangYanAuthorTexas A&M UniversityCollege StationUnited StatesYangLingAuthorTexas A&M UniversityCollege StationUnited StatesSimMartha MSAuthorBloodworks NorthwestSeattleUnited StatesWoodJeremy PAuthorUniversity of KentuckyLexingtonUnited StatesWeiYinanAuthorTexas A&M UniversityCollege StationUnited StatesLiZhenyuAuthorTexas A&M UniversityCollege StationUnited StatesWuCongqingAuthorUniversity of KentuckyLexingtonUnited States
The following is the authors’ response to the current reviews.
Public Reviews:
Reviewer #1 (Public Review):
The authors demonstrated that NINJ1 promotes TF-positive MV release during pyroptosis and thereby triggers coagulation. Coagulation is one of the risk factors that can cause secondary complications in various inflammatory diseases, making it a highly important therapeutic target in clinical treatment. This paper effectively explains the connection between pyroptosis and MV release with Ninj1, which is a significant strength. It provides valuable insight into the potential of targeting Ninj1 as a therapeutic strategy.
Although the advances in this paper are valuable, several aspects need to be clarified. Some comments are discussed below.
(1) Since it is not Ninj1 directly regulating coagulation but rather the MV released by Ninj1 playing a role, the title should include that. The current title makes it seem like Ninj1 directly regulates inflammation and coagulation. It would be better to revise the title.
Thanks for the thoughtful comments. We show that the release of procoagulant MVs by plasma membrane rupture (PMR) is a critical step in the activation of coagulation. In addition, the release of cytokines and danger molecules by PMR may also contribute to coagulation. In choosing the title, we are trying to emphasize NINJ1-dependent PMR as a common trigger for these biological processes.
(2) Ninj1 is known to be an induced protein that is barely expressed in normal conditions. As you showed in "Fig1G" data, control samples showed no detection of Ninj1. However, in "Figure S1", all tissues (liver, lung, kidney and spleen) expressed Ninj1 protein. If the authors stimulated the mice with fla injection, it should be mentioned in the figure legend.
We respectfully disagree with the comment that “Ninj1 is known to be an induced protein that is barely expressed in normal conditions”. NINJ1 protein is abundantly expressed (without induction) in tissues including liver, lung, kidney, and spleen, as shown in Fig S1. Consistently, other groups have shown abundant NINJ1 expression at baseline in tissues and cells such as liver (Kayagaki et.al. Nature 2023) and BMDM (Kayagaki et.al. Nature 2021; Borges et.al. eLife 2022). Fig 1G shows fibrin deposition as an indicator of coagulation, not NINJ1 protein.
(3) In "Fig3A", the Ninj1 protein expression was increased in the control of BMDM +/- cell lysate rather than fla stimulation. However, in MV, Ninj1 was not detected at all in +/- control but was only observed with Fla injection. The authors need to provide an explanation for this observation. Additionally, looking at the MV β-actin lane, the band thicknesses appear to be very different between groups. It seems necessary to equalize the protein amounts. If that is difficult, at least between the +/+ and +/- controls.
Thanks for the valuable comments. In Fla-stimulated Ninj1+/- BMDMs, most of the NINJ1 is released in MVs, therefore, not in the cell lysate, as shown in Fig 3A. The difference in beta-actin band intensity correlated with MV numbers shown in Fig 3B. We ensure consistency by using the same number of cells.
(4) Since the authors focused Ninj1-dependent microvesicle (MV) release, they need to show MV characterizations (EM, NTA, Western for MV markers, etc...).
Thanks for the suggestion. We now add NTA analysis of MV for BMDMs in Fig S4C.
(5) To clarify whether Ninj1-dependent MV induces coagulation, the authors need to determine whether platelet aggregation is reduced with isolated +/- MVs compared to +/+ MVs.
Thanks for the suggestion. We agree that platelet aggregation is closely linked to blood coagulation but would argue that one does not directly cause the other. While it would be interesting to examine whether MVs induce platelet aggregation, we hope the reviewer would agree that the outcome of this experiment would neither significantly support nor challenge our statement that NINJ1-dependent PMR promotes coagulation.
(6) Even with the authors well established experiments with haploid mice, it is a critical limitation of this paper. To improve the quality of this paper, the authors should consider confirming the findings using mouse macrophage cell lines, such as generating Ninj1-/- Raw264.7 cell lines, to examine the homozygous effect.
Thanks for the valuable comments. We acknowledge the limitation of using haploid mice in this study. However, our data provides strong evidence supporting the role of NINJ1-dependent plasma membrane rupture in blood coagulation using primary macrophages.
(7) There was a paper reported in 2023 (Zhou, X. et al., NINJ1 Regulates Platelet Activation and PANoptosis in Septic Disseminated Intravascular Coagulation. Int. J. Mol. Sci. 2023) that revealed the relationship between Ninj1 and coagulation. According to this paper, inhibition of Ninj1 in platelets prevents pyroptosis, leading to reduced platelet activation and, consequently, the suppression of thrombosis. How about the activation of platelets in Ninj1 +/- mice? The author should add this paper in the reference section and discuss the platelet functions in their mice.
Thanks for the valuable comments. We examine PT time, plasma TAT, and tissue fibrin deposition as direct evidence of blood coagulation in this manuscript. We acknowledge that platelets play a key role in thrombosis; however, we hope the reviewer would agree that tissue factor-induced blood coagulation and platelet aggregation are linked yet distinct processes. Therefore, the role of NINJ1 in platelet aggregation falls beyond the scope of this manuscript.
The following is the authors’ response to the original reviews.
Public Reviews:
Reviewer #1 (Public Review):
Referring to previous research findings, the authors explain the connection between NINJ1 and MVs. Additional experiments and clarifications will strengthen the conclusions of this study.
Below are some comments I feel could strengthen the manuscript:
(1) The authors mentioned their choice of using heterozygous NINJ1+/- mice on page 4, because of lethality and hydrocephalus. Nonetheless, there is a substantial number of references that use homozygous NINJ1-/- mice. Could there be any other specific reasons for using heterozygous mice in this study?
Thanks for the thoughtful comments. We are aware that a few homozygous NINJ1-/- mouse strains were used in several publications by different groups, including Drs. Kayagaki and Dixit (Genentech), from whom we obtained the heterozygous NINJ1+/- breeders. We do not have experience with the homozygous NINJ1-/- mice used by other groups. It’s reasonable to assume that homozygous NINJ1-/-, if healthy, would have even stronger protection against coagulopathy than heterozygous NINJ1+/-. The only reason for not using homozygous mice in this study is that a majority of our homozygous NINJ1-/- develops hydrocephalus around weaning and these mice are required to be euthanized by the rules of our DLAR facility. Although our homozygous NINJ1-/- mice develop hydrocephalus (the same reported by Drs. Kayagaki and Dixit, PMID: 37196676, PMCID: PMC10307625), heterozygous NINJ1+/- mice remain healthy.
(2) Figure S2 clearly shows the method of pyroptosis induction by flagellin. It is also necessary as a prerequisite for this paper to show the changes in flagellin-induced pyroptosis in heterozygous NINJ1+/- mice.
Thanks for the valuable suggestions. We agree that a plasma LDH measurement as an indicator of pyroptosis in vivo would add to the manuscript. Therefore, we have made several attempts to measure plasma LDH in flagellin-challenged WT and NINJ1+/- mice using CytoTox96 Non-Radioactive Cytotoxicity Assay (a Promega kit commonly used for LDH, Promega#G1780). Flagellin-challenged WT and NINJ1+/- mice develops hemolysis, which renders plasma red. Because plasma coloring interferes with the assay, we could not get a meaningful reading to make an accurate comparison. We also tried LHD-Glo Cytotoxicity Assay (Luciferase based, Promega#J2380) with no luck on both plasma and serum. We hope the reviewer would agree that reduced plasma MV count (Fig 3C) would serve as an alternative indictor for reduced pyroptosis.
(3) IL-1ß levels controlled by GSDMD were not affected by NINJ1 expression according to previous studies (Ref 37, 29, Nature volume 618, pages 1065-1071 (2023)). GSDMD also plays an important role in TF release in pyroptosis. Are GSDMD levels not altered in heterozygous NINJ1 +/- mice?
Thanks for raising these great points. It’s been reported that IL-1β secretion in cell culture supernatant were not affected by NINJ1 deficiency or inhibition when BMDMs were stimulated by LPS (Ref 29, 37, now Ref 29, 35) or nigericin (Ref 29). As GSDMD pore has been shown to facilitate the release of mature IL-1β, these in vitro observations are reasonable given that NINJ1-mediated PMR is a later event than GSDMD pore-forming. However, we observed that plasma IL-1β (also TNFα and IL-6) in Ninj1+/- mice were significantly lower. There are a few differences in the experimental condition that might contribute to the discrepancy: 1, there was no priming in our in vivo experiment, while priming in BMDMs were performed in both in vitro observations before stimulating with LPS or nigericin; 2, the flagellin in our study engages different inflammasome than either LPS or nigericin. Priming might change the expression and dynamics of IL-1β. More importantly, there might be unrecognized mechanisms in IL-1β secretion in vivo. We now add discussion on this in the main text.
We examined GSDMD protein levels in liver, lung, kidney, and spleen from WT and NINJ1+/- mice by Western blotting. The data is now presented in the updated Fig S1, we did not observe apparent difference in GSDMD expression between the two genotypes.
(4) In Fig 1 F, the authors used a fibrin-specific monoclonal antibody for staining fibrin, but it's not clearly defined. There may be some problem with the quality of antibody or technical issues. Considering this, exploring alternative methods to visualize fibrin might be beneficial. Fibrin is an acidophil material, so attempting H&E staining or Movat's pentachrome staining might help for identify fibrin areas.
Thanks for the valuable suggestions. The fibrin-specific monoclonal antibody in our study is mouse anti-fibrin monoclonal antibody (59D8). This antibody has been shown to bind to fibrin even in the presence of human fibrinogen at the concentration found in plasma [Hui et al. (1983). Science. 222 (4628); 1129-1132]. We apologize that we did not cite the reference in our initial submission. We obtained this antibody from Dr. Hartmut Weiler at Medical College of Wisconsin and Dr. Rodney M. Camire at the University of Pennsylvania, who were acknowledged in our initial submission.
We performed H&E staining on serial sections of the same tissues for Figure 1F. The data is now presented as Fig S3.
Reviewer #2 (Public Review):
Summary:
The author's main goal is to understand the mechanism by which pyroptosis (through the formation of Gasdermin D (GSDMD) pores in the plasma membrane) contributes to increased release of procoagulant Tissue Factor-containing microvesicles (MV). Their previous data demonstrate that GSDMD is critical for the release of MV that contains Tissue Factor (TF), thus making a link between pyroptosis and hypercoagulation. Given the recent identification of NINJ1 being responsible for plasma membrane rupture (Kayagaki et al. Nature 2011), the authors wanted to determine if NINJ1 is responsible for TF-containing MV release. Given the constitutive ninj1 KO mouse leads to partial embryonic lethality, the authors decided to use a heterozygous ninj1 KO mouse (ninj1+/-). While the data are well controlled, there is limited understanding of the mechanism of action. Also, given that the GSDMD pores have an ~18 nm inner diameter enough to release IL-1β, while larger molecules like LDH (140 kDa) and other DAMPs require plasma membrane rupture (likely mediated by NINJ1), it s not unexpected that large MVs require NINJ1-mediated plasma cell rupture.
Strengths:
The authors convincingly demonstrate that ninj1 haploinsufficiency leads to decreased prothrombin time, plasma TAT and plasma cytokines 90 minutes post-treatment in mice, which leads to partial protection from lethality.
Weaknesses:
- In the abstract, the authors say "...cytokines and protected against blood coagulation and lethality triggered by bacterial flagellin". This conclusion is not substantiated by the data, as you still see 70% mortality at 24 hours in the ninj1+/- mice.
Thanks for the thoughtful comments. We corrected the text to “partially protected against blood coagulation and lethality triggered by bacterial flagellin”.
- The previous publication by the authors (Wu et al. Immunity 2019) clearly shows that GSDMDdependent pyroptosis is required for inflammasome-induced coagulation and mouse lethality. However, as it is not possible for the authors to use the homozygous ninj1 KO mouse due to partial embryonic lethality, it becomes challenging to compare these two studies and the contributions of GSDMD vs. NINJ1. Comparing the contributions of GSDMD and NINJ1 in human blood-derived monocytes/macrophages where you can delete both genes and assess their relevant contributions to TF-containing MV release within the same background would be crucial in comparing how much contribution NINJ1 has versus what has been published for GSDMD? This would help support the in vivo findings and further corroborate the proposed conclusions made in this manuscript.
Thanks for the valuable question. We have shown that plasma MV TF activity was reduced in both GSDMD deficient mice (Ref 23) and Ninj1+/- mice (present manuscript). Given that TF is a plasma membrane protein, MV TF most likely comes from ruptured plasma membrane. In flagellin-induced pyroptosis, both GSDMD and NINJ1 deficiency equally blocked LDH release (plasma membrane rupture) in BMDMs (Ref 29). Further, in pyroptosis glycine acts downstream of GSDMD pore formation for its effect against NINJ1 activation (Ref 35). Therefore, GSDMD pore-forming should be upstream of NINJ1 activation in pyroptosis (which may not be the case in other forms of cell death) and there are likely equal effects of GSDMD and NINJ1 on MV release in flagellin-induced pyroptosis. As the reviewer suggested, experiments using human blood-derived monocytes/macrophages will enable a direct comparison to determine the relative contribution. However, this approach presents a few technical difficulties: it’s not easy to manipulate gene expression on primary human monocytes/macrophages (in our experience); variable efficiency in gene manipulation of GSDMD and NINJ1 will complicate the comparison. I hope the reviewer would agree that a direct comparison between GSDMD and NINJ1 is not required to support our conclusion that NINJ1-dependent membrane rupture is involved in inflammasome-pyroptosis induced coagulation and inflammation.
- What are the levels of plasma TAT, PT, and inflammatory cytokines if you collect plasma after 90 minutes? Given the majority (~70%) of the ninj+/- mice are dead by 24 hours, it is imperative to determine whether the 90-minute timeframe data (in Fig 1A-G) is also representative of later time points. The question is whether ninj1+/- just delays the increases in prothrombin time, plasma TAT, and plasma cytokines.
Thank for the valuable question. The time point (90 min) was chosen based on our in vitro observation that flagellin-induced pyroptosis in BMDMs largely occurs within 60-90 min.
Because our focus on the primary effect of flagellin in vivo, potential secondary effects at later points may complicate the results and are hard to interpret. As the reviewer suggested, we have measured plasma PT, TAT at 6 hours post-flagellin challenge. The significant difference in PT sustained between Ninj1+/+ and Ninj1+/- (Fig A), suggesting coagulation proteins remained more depleted in Ninj1+/+ mice than in Ninj1+/- mice. However, plasma TAT levels were diminished to baseline level (refer to Fig 1B in main text) in both groups and showed no significant difference between groups (Fig B), which could be explained by the short half-life (less than 30 min) in the blood. Since flagellin challenge is a one-time hit, there might not a second episode of coagulation after the 90-minute time point, at least not triggered by flagellin, supported by the plasma TAT levels at 6 hours. We now comment on this limitation at the end of the main text.
Based on our previous studies, plasma IL-1β and TNFα peaked at early time point and diminished over time, but plasma IL-6 levels maintained. As shown below, plasma IL-6 appeared higher in Ninj1+/+ compared with Ninj1+/-, but not statistically significant (partly because one missing sample, n = 4 not 5, in Ninj1+/+ group decreased the statistical power of detecting a difference).
Mice were injected with Fla (500 ng lFn-Fla plug3 ugPA).
Blood was collected 6 hours after Fla injection. Prothrombin time (A), plasma TAT (B), and plasma IL-6 (C) were measured. Mann-Whitney test were performed.
Recommendations for the authors:
Reviewer #1 (Recommendations For The Authors):
- Fig 1F: are there lower magnification images that capture the fibrin deposition? The IHC data seems at odds with the WB data in Fig. 1G where there is still significant fibrin detected in the heterozygous lungs and liver. Quantitating the Fig. 1G Western blot would also be helpful.
IHC surveys a thin layer of tissue section while WB surveys a piece of tissue, therefore fibrin deposition may be missing from IHC and but found in WB. That is why we used two methods. Below we provide lower mag images of fibrin deposition (about 2 x 1.6 mm area).
- Fig1H - lethality study uses 5x dose of Fla used in earlier studies. In the lethality data where there is a delay in ninj1+/- mortality, are the parameters (prothrombin time, plasma TAT, and plasma cytokines) measured at 90 minutes different between WT and ninj+/- mice? This would be critical to confirm that this is not merely due to a delayed release of TF-containing MVs.
We used 5x lower dose of Fla in coagulation study than lethality study because it’s not as easy to draw blood from septic mouse with higher dose of flagellin. We need to terminate the mice to collect blood for plasma measurement and therefore the parameters were not measured for mice in lethality study.
- What is the effect of ninj+/- on E. coli-induced lethality in mice? How do these data compare to E. coli infection of GSDMD-/- mice?
We did not examine the effect of Ninj1+/- on E. coli-induced lethality. After the initial submission of our manuscript, we have focused on Ninj1 flox/flox mice instead of Ninj1+/- for NINJ1 deficiency. We are using induced global Ninj1 deficient mice for polymicrobial infectioninduced lethality in our new studies.
- Fig 2 - in the E. coli model, the prothrombin time, plasma TAT, and plasma cytokines are measured 6 hours post-infection. How were these time points chosen? Did the authors measure prothrombin time, plasma TAT, and plasma cytokines at different time points?
The in vivo time point for flagellin and E. coli were chosen based on our in vitro observation of the timelines on BMDM pyroptosis induced by flagellin and bacteria. This disparity probably arises from distinct dynamics between purified protein and bacterial infections. Purified proteins can swiftly translocate into cells and take effect immediately after injection. Conversely, during bacterial infection, macrophages engulf and digest the bacteria to expose their antigens. Subsequently, these antigens initiate further effects, a process that takes some time to unfold.
Our focus is on the primary effect of flagellin in vivo, potential secondary effects at later points may complicate the results and are hard to interpret. As the reviewer suggested, we have measured plasma PT, TAT at 6 hours post-flagellin challenge. The significant difference in PT sustained between Ninj1+/+ and Ninj1+/- (Fig A), suggesting coagulation proteins remained more depleted in Ninj1+/+ mice than in Ninj1+/- mice. However, plasma TAT levels were diminished to baseline level (refer to Fig 1B in main text) in both groups and showed no significant difference between groups (Fig B), which could be explained by the short half-life (less than 30 min) in the blood. Since flagellin challenge is a one-time hit, there might not a second episode of coagulation after the 90-minute time point, at least not triggered by flagellin, supported by the plasma TAT levels at 6 hours. We now comment on this limitation at the end of the main text.
Based on our previous studies, plasma IL-1β and TNFα peaked at early time point and diminished over time, but plasma IL-6 levels maintained. As shown below, plasma IL-6 appeared higher in Ninj1+/+ compared with Ninj1+/-, but not statistically significant (partly because one missing sample, n = 4 not 5, in Ninj1+/+ group decreased the statistical power of detecting a difference).
- Fig 3 - the sequence of figure panels listed in the legend needs to be corrected. Fig 3A requires quantitation of NINJ1 levels compared to beta-actin. Fig 3C - needs a control for equal MV loading.
Thanks for the recommendations. The figure sequence has been corrected. There remain no common markers or loading controls for MV, so we use equal plasma volume for loading control.
Additional comments:
(1) In Fig 3A, the size of NINJ1 appears to be increased in the NINJ+/- group.
This discrepancy is likely attributed to a technical issue when running the protein gel and protein transfer, which makes the image tilt to one side.
(2) Describe the method of BMDM isolation.
Thanks for the recommendations. We now include the method of BMDM isolation. In brief, mouse femur and tibia from one leg are harvested and rinsed in ice-cold PBS, followed by a brief rinse in 70% ethanol for 10-15 seconds. Both ends of the bones are then cut open, and the bone marrow is flushed out using a 10 ml syringe with a 26-gauge needle. The marrow is passed through a 19-gauge needle once to disperse the cells. After filtering through a 70-μm cell strainer, the cells are collected by centrifugation at 250 g for 5 minutes at 4 °C, then suspended in two 150 mm petri dish, each containing 25 ml of L-cell conditioned medium (RPMI-1640 supplemented with 10% FBS, 2mM L-Glutamine, 10mM HEPES, 15% LCM, and penicillin/streptomycin). After 3 days, 15 mL of LCM medium is added to each dish cells. The cells typically reach full confluency by days 5-7.
(3) According to this method, BMDMs are seeded without any M-CSF or L929-cell conditioned medium. How many macrophages survive under this condition?
BMDMs are cultured and differentiated in medium supplemented with 15% L929-cell conditioned medium. For the experiment, the cells were seeded in Opti-MEM medium (Thermo Fisher Scientific, Cat# 51985034) without M-CSF or L929-cell conditioned medium. BMDMs can survive under this condition, as evidenced by low LDH and high ATP measurement (Fig S5).
Reviewer #2 (Recommendations For The Authors):
- There is significant information missing in the methods and this makes it unclear how to interpret how some of the experiments were performed. For example, there is no detailed description or references in the methods on how the in vivo experiments were performed. The methods section needs significantly more details so that any reader is able to follow the protocols in this manuscript. References to previous work should also be included as needed.
Thanks for the recommendations. We had some of the details in the figure legend. We now add details in the methods for better interpretation of our data.
- Line numbers in the manuscript would be helpful when resubmitting the manuscript so that the reviewer can easily point to the main text when making comments.
Thanks for the recommendations. We now add line numbers in the manuscript.