Aryl-functionalised α,α′-Trehalose 6,6′-Glycolipid Induces Mincle-independent Pyroptotic Cell Death

AF-2 [25] and TDB [28] were prepared according to previously published protocols and determined to be endotoxin free (≤ 0.1 EU/mL) using the Pierce^™ LAL Chromogenic Endotoxin Quantitation Kit (Thermo Scientific).

C57BL/6 WT mice and Mincle^−/− mice were bred and housed in a conventional animal facility at the Malaghan Institute of Medical Research, Wellington, New Zealand, or Osaka University, Japan. All animals used for the experiments were aged between 8–12 weeks.

α,α′-Trehalose 6,6′-glycolipids, AF-2 and TDB, were made up to 1 mM in CHCl₃/MeOH (2/1, v/v), then diluted to 400 μM or 50 μM in isopropanol to give 8 or 1 nmol of glycolipid per 20 μL, respectively. The solutions were then added to 96-well plates (20 µL/well) and the solvent was evaporated within a sterile hood for 18 h.

Bone marrow cells were collected from the tibia and femur of C57BL/6 or Mincle^−/− mice and cultured (250,000 cells/mL) in complete Roswell Park Memorial Institute medium [cRPMI-1640 (Gibco) with 10% heat inactivated fetal bovine serum (Gibco), 100 unit/mL penicillin-streptomycin (Gibco) and 2 mM Glutamax (Gibco)]. Macrophage differentiation was induced by 50 ng/mL GM-CSF (PeproTech) added to the cRPMI. Cells were incubated at 37 °C (5% CO₂) for 8 days (cells fed on days 3 and 6). On day 8, the media was removed, the cells were lifted using Accutase (StemCell), counted and re-seeded in 96-well plates coated with ligands. Where indicated Ac-YVAD-cmk (40 μM, Sigma) or KCl (50 mM) [12] were added to the cells an hour before, or CY-09 (20 or 40 µM) was added 30 min before adding the cells to ligand-coated plate according to the manufacturer’s protocol. 100 ng/mL LPS (Sigma) or nigericin (10 μM, ChemImpex) were used as controls [29].

Bone marrow cells were collected from the tibia and femur of C57BL/6 or Mincle^−/− mice and cultured as described above. On day 8, all media together with non-adherent cells were removed and 0.25 mL fresh media was added to each well (48 well plate). Stock solutions of AF-2 and TDB (1 mM, with 2% DMSO in sterile water) were prepared and kept sterile. From these stock solutions, 10 μL were added to each well to give a final ligand concentration of 40 μM in a total well volume of 250 μL. A positive control with 100 ng/mL LPS and a negative control with untreated cells were used. Stimulated cells were incubated for 24 h before the supernatants were analysed for IL-1β production.

The use of human leukocytes from healthy donors with written informed consent was approved by New Zealand Northern A Health and Disability Ethics Committee (approval number 15/NTA/178/AM04). Human monocytes were enriched from whole blood by negative selection [30] using RosettaSep Human Monocyte Enrichment Cocktail (StemCell) according to the manufacturer’s instructions. The cell concentration was adjusted to 1 × 10⁶ cell/mL in cRMPI and 100 μL of cells were added per well in a ligand-coated 96-well plate.

Levels of murine IL-1β, IL-23, IL-10, MIP-2 (R&D Systems), IL-18 (Sino Biological), IL-13 (Invitrogen), TNF-α, IL-6, IL-12p40, and human IL-1β (BD Biosciences) cytokines in the supernatants were determined by sandwich ELISA according to the manufacturer’s instructions.

Bone marrow cells from WT and Mincle^−/− mice were cultured in 10% FCS/RPMI supplemented with 50 ng/mL GM-CSF (BioLegend) for eight days. Attached cells were detached by 1 mM EDTA-supplemented PBS and were used as BMDMs. The cells were added into 96-well plates coated with 8 nmol/well of AF-2 or TDB, or 100 μg/mL zymosan (SIGMA), in the presence of 300 μM luminol (Nacalai tesque) for 60 min. Luminescence was quantified by POWERSCAN HT (DS Pharma Biomedical). Data are presented as mean ± SE of triplicate assays.

Wild type and Mincle^−/− GM-CSF BMDMs were differentiated over 8 days, lifted using Accutase (StemCell) and stimulated on 96-well plates coated with AF-2 or TDB (8 nmol/well) or nigericin (10 μM, positive control) for 4 h. Whole cell lysate was collected with IGEPAL^® CA-630 (Sigma) buffer containing protease (cOmplete Tablets, Roche) inhibitors. Western blots were performed by 12% SDS-PAGE and wet blotting. Antibodies used were anti-GSDMD [EPR19828] (Abcam), anti-β-actin [D6A8], anti-rabbit IgG-HRP (Cell Signaling Technology). Blots were developed with Pierce^™ ECL Western Blotting Substrate (Thermo Scientific) and imaged with Amersham Imager 600.

WT and Mincle^−/− GM-CSF BMDMs were differentiated over 8 days and lifted using Accutase (StemCell). Sytox Green (1 μM, Invitrogen) was added to the media before adding the cells to 96-well plates coated with AF-2 or TDB (8 nmol/well). Nigericin (10 μM) was used as a positive control. Fluorescence of Sytox Green was measured at the indicated time points (cells kept in 37 °C, 5% CO₂ between measuring) using EnSpire plate reader (Perkin Elmer) at 523 nm; results are calculated relative to iPrOH control fluorescence.

LDH was measured from supernatant using CyQUANT^™ LDH Cytotoxicity Assay (Invitrogen). Percentage of cytotoxicity was calculated using spontaneous LDH activity and maximum LDH activity controls according to the manufacturer’s instructions.

WT and Mincle^−/− GM-CSF BMDMs were differentiated over 8 days, lifted using Accutase (StemCell) and seeded on cover slides coated with iPrOH as a control, 8 nmol/slide AF-2 or TDB, or treated with Nigericin (10 μM) as a control for pyroptosis. After 4 h, the cells were washed with D-PBS (Gibco) and stained with anti-myeloperoxidase (MPO) antibody [2D4] (FITC conjugated, Abcam), anti-histone H4 (acetyl K16) antibody [EPR1004] (Alexa Fluor 647 conjugated, Abcam) and DAPI (BD Pharmagen) for 30 min. After further washing with PBS, ProLong^™ Diamond Antifade Mountant (Invitrogen) was used to mount cover slips onto microscope slides. Confocal images were taken using the inverted microscope IX83 equipped with a confocal head FV3000, employing the blue filter to observe the DAPI dye, red filter for H4 and the green filter to observe the MPO. The BMMs with media only treatment was used to correct for autofluorescence of the cells.

BMDM cells were seeded on conductive silicon wafers, coated with AF-2 (8 nmol/slide) or left untreated (untreated control, staurosporine and nigericin controls), and allowed to adhere overnight. Nigericin/staurosporine control cells were first treated with 1 μg/mL of LPS for 4 h. LPS was removed and the cells were further treated with 10 μM nigericin for 1 h or 5 h with 1 μM staurosporine, respectively. At the end of each treatment the media was removed, and the cells were fixed overnight at 4 °C using Karnovsky’s EM fixative, with paraformaldehyde, glutaraldehyde, calcium chloride and cacodylate buffer, at 3% strength. Fixation buffer was removed, cells rinsed using PBS and dehydrated through a graded series of ethanol (30, 50, 70, 95 and 100%) and acetone (100%) followed by critical point drying with liquid CO₂. RAW264.7 cells were primed with LPS (1 μg/mL) and seeded on conductive silicon wafers coated with AF-2 or TDB (8 nmol/slide) or treated with staurosporine (1 μM) or with nigericin (10 μM). Following treatment, the cells were fixed overnight at 4 °C using Karnovsky’s EM fixative, rinsed briefly using water and ethanol, and snap frozen using liquid nitrogen to observe morphology using CryoSEM. Dried CryoSEM and SEM specimens were sputter coated with platinum and imaged with a JEOL JSM 6500F field emission scanning electron microscope operating at 20 kV (SEM) or 10 kV (CryoSEM).

Statistical significance of differences was assessed using two tailed 2-way ANOVA with Dunnett’s multiple comparison test or Multiple t-tests where appropriate, using Prism v8 software (GraphPad). A P value less than 0.05 was considered statistically significant.

To explore the mode of action of AF-2, we investigated the ability of AF-2 to induce cytokine and chemokine production using WT and Mincle^−/− BMDMs. The commonly used plate-bound assay was performed, whereby the AF-2 and TDB were first dissolved in an organic solvent, added to the bottom of a tissue culture plate, and the solvent then allowed to evaporate to dryness in a sterile hood. Cytokine and chemokine production was then measured by ELISA. AF-2 led to the production of IL-1β, IL-6, IL-23, MIP-2, TNF-α, IL-12p40 and IL-10 by WT BMDMs (Fig. 2a). Levels of IL-18 and IL-13 were too low to be detected. All responses were Mincle-dependent except for IL-1β production (P ≤ 0.001) and IL-12 production (P ≤ 0.05), with AF-2, but not TDB, leading to a significant increase in these cytokines in the absence of Mincle. We also measured the production of reactive oxygen species (ROS) following the stimulation of WT and Mincle^−/− BMDM with AF-2 and TDB. However, in the absence of priming, ROS production was not observed by either AF-2 or TDB (SI, Fig. 1).

The strong AF-2-mediated IL-1β response in the absence Mincle was unexpected and prompted us to reassess the ability of AF-2 to lead to IL-1β production by WT and Mincle^−/− BMDMs at two different glycolipid concentrations (1 and 8 nmol/well) and to compare these responses to those elicited by TDB and the NLRP3 inflammasome activator, nigericin [29]. WT and Mincle^−/− BMDMs produced IL-1β in response to AF-2 in a concentration-dependent manner, while TDB showed only Mincle-dependent IL-1β production with little change in the quantity of IL-1β being induced at both ligand concentrations (Fig. 2b). Like AF-2, nigericin led to a strong IL-1β response in both cell types. When AF-2 and TDB were added in PBS, the Mincle-dependent production of IL-1β was observed (Fig. 2c), thus highlighting how glycolipid presentation can affect the immune response [31‐33]. We then assessed whether plate-coated AF-2 was able to lead to IL-1β production in human monocytes. In the absence of LPS priming, TDB was unable to induce detectable IL-1β production in human monocytes after 24 h (Fig. 2d). In contrast, at a concentration of 1 nmol/well, AF-2 induced IL-1β production by human monocytes, although this response was abrogated at higher concentrations of AF-2.

The production of IL-1β can be associated with cell death [34]. Accordingly, the ability of plate-coated AF-2 to induce cell death was investigated using the Sytox Green assay [35]. In this assay, detection of a cell membrane impermeable nucleic acid stain is proportionally related to the cell nuclear contents, and thus, simultaneously provides a measure of membrane integrity and cell death. For murine WT and Mincle^−/− BMDMs, AF-2, but not TDB, induced cell death, with this reaching a plateau after ca. 4–6 h when using 8 nmol/well of glycolipid (Fig. 3a, SI Fig. 2a). Cell death by AF-2 was dose dependent, with an increasing concentration of AF-2 (0.01, 0.1, 1, 8 10 nmol/well) increasing membrane degradation, as measured at the 24 h time point in WT BMDMs (SI, Fig. 2b). AF-2 also induced cell death in murine RAW264.7 cells (SI, Fig. 2c).

Further evidence for cell death and a loss of membrane integrity upon the treatment of BMDMs with AF-2 was observed by measuring the release of cytoplasmic lactate dehydrogenase (LDH) [36]. AF-2 led to a concentration dependent leakage of intracellular LDH for both WT and Mincle^−/− murine BMDMs, with cell death being enhanced when using a higher concentration of AF-2 (Fig. 3b). No extracellular LDH was detected upon the treatment of WT and Mincle^−/− BMDMs with TDB at either glycolipid concentration. Rather, TDB stimulation impaired LDH leakage from WT and Mincle^−/− BMDMs resulting in a negative percentage of cytotoxicity compared to the isopropanol control. Cell death was also observed when treating human monocytes with AF-2, as indicated by the Sytox Green assay (Fig. 3c) and by LDH release (Fig. 3d). Stimulation of human monocytes with TDB did not result in cell death. Scanning electron microscopy (SEM) also revealed that BMDMs remained viable and were numerous after being seeded on coverslips overnight (Fig. 3e). Treating the plates with AF-2 led to striated ring-like deposits on the plate, and the BMDMs were observed to undergo cell death (Fig. 3e). TDB did not lead to any obvious striations when coated on the cover-slips. The cells treated with TDB were as abundant as the untreated control cells but were more highly adherent and extracellular membrane bound vesicles could be observed (SI, Fig. 3).

We then sought to provide further insight into the mode of action of cell death elicited by plate-coated AF-2. A hallmark of pyroptosis is the cleavage of gasdermin D (GSDMD) [34, 37]. This leads to the formation of pores in the plasma membrane that allow for the release of cytoplasmic contents including IL-1β and LDH [34, 38]. The cleavage of GSDMD was investigated by Western Blot following the treatment of WT and Mincle^−/− BMDMs with AF-2 and nigericin (positive control) (Fig. 4a). In both cell types, AF-2 led to GSDMD cleavage, as did nigericin, while TDB did not. GSDMD is a substrate for Caspase-1 [39]. The addition of the Caspase-1 inhibitor Ac-YVAD-cmk prevented GSDMD cleavage in both cell types, thus confirming Caspase-1 mediated pyroptotic cell death. Caspase-1 cleaves pro-IL-β [39]. We observed that the production of IL-1β by AF-2 was also dependent on Caspase-1, with the addition of Ac-YVAD-cmk leading to a significant decrease in IL-1β production by WT and Mincle^−/− BMDMs (Fig. 4b). Caspase-1 inhibition prevented IL-1β production in response to TDB, which was consistent with the observation that the NOD-like receptor family pyrin domain containing 3 (NLRP3) inflammasome and Caspase-1 activation is required for TDB-mediated IL-1β production [12, 40]. The addition of Ac-YVAD-cmk to AF-2-treated BMDMs led to a significant decrease in LDH release at 4 h at both concentrations of AF-2 used, and a significant decrease in LDH release when using the lower concentration of AF-2 at 24 h for both WT and Mincle^−/− BMDMs (Fig. 4c, d, SI Fig. 4a, b). The addition of Ac-YVAD-cmk to wild type BMDMs also decreased AF-2-mediated cell death, as evidenced using the Sytox Green assay (SI Fig. 4c).

Caspase-1 containing inflammasomes consist of NLR inflammasomes, NLRP3 and NLRC4, and the absent in melanoma 2 (AIM2) inflammasomes [41, 42], whereby the latter binds to cytosolic dsDNA [41]. NLRP3 and NLRC4 are differentially affected by K⁺ efflux, with the concentration of extracellular K⁺ required to inhibit NLRC4 being reported to be higher (> 90 mM KCl) than that required for NLRP3 inhibition [43]. Using 50 mM KCl, we observed a decrease in the AF-2- and nigericin-mediated deaths of WT (Fig. 5a) and Mincle^−/− BMDMs (SI Fig. 5a), but there was no significant effect on TDB treated cells. The addition of extracellular KCl decreased AF-2 mediated LDH release from both WT and Mincle^−/− BMDMs (SI Fig. 5b, c), and led to the abolition of IL-1β production by WT BMDMs after stimulation with AF-2 (3) and nigericin (Fig. 5b, SI Fig. 5d). To confirm the role of the NLRP3 inflammasome in AF-2-mediated cell death, we added titrated quantities of the NLRP3 inflammasome inhibitor CY-09 [44] to BMDMs treated with AF-2 and measured IL-1β production at 4 h and 24 h (Fig. 6c). A significant concentration dependent decrease in IL-1β was observed for both concentrations of AF-2 at both time points following the addition of CY-09. A similar response was observed for nigericin (positive control).

Pyroptosis is morphologically and mechanistically distinct from other forms of cell death [34, 45]. AF-2 stimulation caused cell swelling and the spillage of intracellular material that is characteristic of pyroptotic cell death [34, 45], as indicated by DAPI (double stranded DNA) and Histone 4 (chromatin) staining imaged by confocal microscopy (Fig. 6a). In contrast, in control and TDB treated cells a clear distinction between nuclear and cytoplasmic contents (MPO lysosomal stain) was observed, with TDB-treated cells exhibiting cell clustering. Cells treated with nigericin dramatically decreased in size and showed some membrane disruption. SEM of untreated control BMDMs showed intact adherent cells, while staurosporine treated cells revealed classic apoptotic bodies, and treatment with nigericin revealed corpses of pyroptotic cells with a ‘fried egg’ morphology (Fig. 6b, c) [46]. Treatment with AF-2 induced the formation of large networks of cellular connections and smaller cells with cellular blebs. Similar findings were observed via scanning electron cryomicroscopy (CryoSEM) of RAW264.7 cells treated with AF-2, staurosporine, or nigericin (Fig. 6d, e). Treatment with AF-2 led to a reduced number of adherent cells, a reduction in cell size, and the expulsion of intracellular material, which was not membrane bound.