Dynamic imaging reveals surface exposure of virulent Leishmania amastigotes during pyroptosis of infected macrophages

Leishmania amazonensis is a protozoan parasite, causing localized cutaneous and anergic diffuse cutaneous leishmaniasis in South America (Barral et al., 1991; Silveira et al., 2004). Restriction of intracellular L. amazonensis replication in vitro and in vivo has been shown to depend on the NOD-, LRR- and pyrin domain-containing protein 3 (NLRP3) inflammasome (Zamboni and Sacks, 2019). This intracellular sensor is induced in response to a ‘priming signal’ represented by cytokines or ligands of Toll-like receptors, and is further activated by damage-associated molecular patterns (DAMPs), such as ATP. Inflammasome activation triggers caspase-1 activity, which cleaves pro-IL-1β and pro-IL-18 into mature cytokines – promoting an anti-microbial, pro-inflammatory immune response (Swanson et al., 2019) – and cleaves gasdermin D, which forms pores inside the macrophage plasma membrane, sustaining IL-1β release (Lieberman et al., 2019) and leading to pyroptosis (Bergsbaken et al., 2009; Fink and Cookson, 2006; Jorgensen et al., 2017; Shi et al., 2015). Importantly, pyroptotic cell death contributes to a protective response against intracellular pathogens by removing the niche for infection (Bergsbaken et al., 2009), by exposing microbes to the immune system (Jorgensen and Miao, 2015) and by rendering pathogens more susceptible to anti-microbial agents (Jorgensen et al., 2016) or direct microbial killing (Liu et al., 2016).

Even though the interaction of L. amazonensis with the NLRP3 inflammasome has attracted considerable attention (Lecoeur et al., 2020; Lima-Junior et al., 2013), how pyroptosis affects parasite survival and virulence remains to be elucidated. Here, we deployed a real-time, high-content single-cell analysis that uncovers unique features of Leishmania amastigotes during host macrophage pyroptosis.

To analyze the dynamics of pyroptosis in L. amazonensis-infected macrophages, we established a new imaging protocol combining high-content and single-cell analyses using the OPERA™ QEHS confocal plate reader (Fig. 1A). Bone marrow-derived macrophages (BMDMs) were seeded into 96-well plates (macrophage settlement). They were subsequently infected for 3 days with mCherry transgenic virulent amastigotes isolated from lesions of infected Nude mice, and allowed to form characteristic communal parasitophorous vacuoles (PVs; parasite settlement). Then, lipopolysaccharide (LPS) was added for 225 min (NLRP3 priming) followed by staining with fluorescent reporters. Pyroptosis was triggered after NLRP3 activation by treatment with ATP (pyroptosis induction). Dynamic cellular changes were monitored at high-content and single-cell levels by using Hoechst 33342 (to determine cell number; Fig. 1B1), YO-PRO-1 [to determine loss of plasma membrane (PM) integrity, Fig. 1B2] (Adamczak et al., 2014), LysoTracker green (LTG, to determine PV acidity, Fig. 1B3) (Aulner et al., 2013) and mCherry (to determine parasite localization, Fig. 1B4) (Aulner et al., 2013). Macrophage morphology was analyzed by transmission light microscopy (TL, Fig. 1B5). High-content assay (HCA) analyses were carried out in real-time up to 240 min post ATP addition. Analyses at the population level were performed using segmentation procedures (Fig. S1A,B). The analysis of fluorescent read-outs at 240 min documented that (1) no macrophages were lost during the analyses (Fig. 1C1), (2) pyroptosis induction was successful (Fig. 1C2), and the PV integrity was lost in all pyroptotic cells (Fig. 1C3). The induction of pyroptosis after LPS and ATP treatment in L. amazonensis-infected macrophages was validated by monitoring cell morphology (Fig. S2A), caspase-1 cleavage and release (Fig. S2B), and IL-1β secretion (Fig. S2C). Single-cell analyses permitted to determine macrophage PV area, and parasite location (Fig. 1D).

Pyroptosis dynamics, PV integrity and parasite localization were analyzed every 5 min after ATP addition for a duration of 240 min in infected LPS-primed macrophages. HCA analysis of YO-PRO-1 incorporation revealed that pyroptosis was asynchronous, as judged by the progressive increase in the number of dead cells (Fig. 2A, black curve). Pyroptosis occurred with a constant rate (+18.2±4.2% of cell death per hour; mean±s.e.m.). A loss of LTG staining revealed that PVs rapidly decayed (Fig. 2A, green curve), following three distinct empirical rates: (1) a fast decay rate during the first 45 min (−4.2×10⁶±1.2×10⁶/30 min, stage 1) leading to a 47.6% decrease of LTG fluorescence intensity, (2) an intermediate rate (45 to 120 min, −1.2×10⁶±6.3×10⁵/30 min, stage 2), and (3) a slow rate thereafter (0.58×10⁶±0.37×10⁶/30 min, stage 3). This decay may result from ATP-induced changes in the vacuolar pH and of the PV integrity that could be caused by osmotic changes similar to those described for lysosomes (Guha et al., 2013; Takenouchi et al., 2009). Single-cell analyses revealed differences in the PV decay kinetics between macrophages (Fig. S3A–D) as well as between PVs inside a same macrophage (Fig. 2B1–B3). Monitoring individual PVs revealed progressive reduction of PV area and loss of about half of the PVs at the final time point of 240 min (Fig. S3B). Interestingly, the decrease of PV area is more pronounced in large PVs (area of more than 2000 square pixels) than in smaller ones (area less than 1000 square pixels) (Fig. S3C).

HCA analysis revealed that amastigotes either remained intra-vacuolar (62% of macrophages) or were exposed at the cell surface (38% of macrophages), as shown in representative cells (Fig. 2B1,B2; Fig. S3A). Importantly, different decay rates and parasite localizations were observed for individual PVs, even inside the same macrophage as indicated in Fig. 2B2, showing three independent vacuoles (a, b and c) that rapidly lost LTG fluorescence (stage 1). PVs a and b displayed a moderate decrease in size and retained intracellular amastigotes (Fig. 2B2,B3), whereas PV c collapsed (at 60 min) and the luminal side of its membrane was exposed to the extracellular milieu (at 120 min, stage 2). This PV lumen exposure could be triggered by ATP, as judged by previous reports implicating ATP in the extracellular discharge of secretory exosomes (Andrei et al., 2004; Bergsbaken et al., 2011), in exosome exocytosis (Qu et al., 2007), or in the release of auto-phagolysosomes and phagocytosed intracellular particles (Bergsbaken et al., 2011; Takenouchi et al., 2009). Amastigote exposure occurred preferentially in macrophages displaying a rapid PV decay (Fig. 2;,Fig. S3). Additionally, parasite exposure does not correlate with PV size, PV number or amastigote number per vacuole (data not shown). In contrast, PVs that display a high area value and that rapidly expose parasites – such as PV c (Fig. 2B2) – decayed faster than other PVs (Fig. S3C,D).

Pyroptosis has been recognized as an anti-microbial strategy during bacterial infection that is capable of trapping pathogens inside dying host cells (Jorgensen et al., 2016) and directly killing bacteria (Liu et al., 2016). Our observation that amastigotes remain attached on the surface of pyroptotic cells primed us to investigate parasite integrity and the molecular structures underlying parasite surface retention.

Scanning electron microscopy analyses showed that 38.8% of pyroptotic macrophages displayed intact parasites at the cell surface (Fig. 3A1,A2; Fig. S3E,F), thus confirming our results obtained with the OPERA™ system (Fig. 2). Ultrastructural analysis demonstrated that exposed amastigotes remained attached at macrophage membranes. Significantly, exposed amastigotes showed a highly polarized orientation with only the anterior pole (flagellar side) being exposed toward the extracellular milieu. This feature was also observed for pyroptotic macrophages infected with L. donovani amastigotes that reside in individual PVs (Fig. S3G). We next investigated the posterior attachment zone by transmission electron microscopy (TEM), which revealed an electron-dense membranous junction formed between parasite and PV membranes both in non-stimulated and pyroptotic macrophages (Fig. 3B1,B2). This junction corresponds to a defined attachment site – showing features similar to gap junctions – that permits amastigotes of communal Leishmania species to be anchored to PV membranes (Benchimol and de Souza, 1981). These results demonstrate that amastigotes remain strongly attached to pyroptotic macrophages through their attachment site (see Movie 1), which represents a new type of interaction observed in pyroptotic cells, different from the one previously described in the literature (Jorgensen et al., 2016).

Since bacteria can be damaged during host cell pyroptosis (Liu et al., 2016), we next investigated whether parasite viability and infectivity were affected during macrophage pyroptosis. Amastigotes isolated from pyroptotic macrophages (A-pm) were compared to amastigotes isolated from unstimulated control macrophages (A-cm). Macrophage pyroptosis did not reduce parasite viability (99% of A-pm remained YO-PRO-1 negative; Fig. 4A1). In addition, when pyroptotic macrophages were cultured in promastigote-specific medium (48 h, 27°C), amastigotes rapidly transformed into ovoid cells with a long flagellum, motile and proliferating promastigotes that detached from macrophage remnants (Fig. 4A2). These promastigotes showed normal growth (visual parasite counting, Fig. 4A3) and metabolic activity (resazurin reduction data, Fig. 4A4). Finally, A-pm maintained their capacity to establish efficient infection of new macrophages (Fig. 4B). In addition, amastigotes associated to macrophage remnants were efficiently phagocytosed by naïve BMDMs and established normal infection levels as shown by the formation of typical large communal PVs housing numerous parasites after a 3-day co-culture (Fig. 4C3).

We analyzed dynamic interactions between Leishmania amastigotes and PVs in pyroptotic BMDMs at the population, single-cell and ultrastructural levels. Real-time HCA analysis uncovered three distinct stages of the pyroptotic process in our experimental system (Fig. S4): During stage 1, the PV membrane decays (rapid loss of PV acidity and reduced PV size). In stage 2, the PV lumen is externalized in 38% of cells, resulting in exposure of membrane-anchored amastigotes. In stage 3, cellular alterations are further sustained (80% of cell death). L. amazonensis amastigotes retained their viability and infectivity during macrophage pyroptosis. We therefore can speculate that parasites attached to pyroptotic cell debris in vivo may favor parasite spreading – as shown for other forms of host cell death – allowing for uptake by macrophages newly recruited to inflammatory infection sites (de Menezes et al., 2016). On the other hand, the extracellular exposure of amastigotes could expose parasites to complement-dependent cytotoxicity, which is known to control parasite load in vivo (Laurenti et al., 2004), or to anti-leishmanial antibodies that could promote cell-mediated cytotoxicity via Fc receptor-dependent phagocytosis. Our results will stimulate future studies designed to assess the role of macrophage pyroptosis in Leishmania dissemination, transmission, and immuno-pathology.

Animals were housed at the Institut Pasteur animal facilities accredited by the French Ministry of Agriculture for performing experiments on live rodents. Work on animals was performed in compliance with French and European regulations on care and protection of laboratory animals (EC Directive 2010/63, French Law 2013-118, February 6th, 2013). All experiments were approved by the Ethic Committee for animal experimentation (CETEA#89) and authorized by the French ministry of higher education, research and innovation under the reference 2013-0092 in accordance with the Ethics Charter of animal experimentation that includes respect of the 3Rs principles, appropriate procedures to minimize pain and animal suffering.

Female C57BL/6 mice were obtained from Janvier (Saint Germain-sur-l'Arbresle, France). Bone marrow cell suspensions were recovered from tibias and femurs as described previously (Courret et al., 1999). Bone marrow cells were plated at 1.5×10⁷ cells/ml in hydrophobic Petri dishes (#664161, Corning Life Science) and cultured for 6 days at 37°C in a 7.5% CO₂ air atmosphere in complete Dulbecco's modified Eagle's medium (DMEM; #P04-03500, Pan Biotech) containing 4.5 g/l glucose, 2 mM L-glutamine, 1 mM sodium pyruvate and 3.7 g/l NaHCO₃ and supplemented with 15% fetal calf serum (FCS; #A3160801, Gibco), 10 mM HEPES (# 15630080, Gibco), 50 µg/ml penicillin/streptomycin (#P4333, Sigma-Aldrich), 50 µM 2-mercaptoethanol (# M6250, Sigma-Aldrich) and with 75 ng/ml recombinant mouse CSF-1 (#234311, ImmunoTools). Adherent bone marrow-derived macrophages (BMDMs) were recovered and seeded in complete DMEM supplemented with 30 ng/ml recombinant mouse CSF-1 (rmCSF-1, #12343115, ImmunoTools) into various culture-treated supports, including (1) flat bottom 96-well black µClear^® plates (# 655090, Greiner Bio-One) for OPERA analyses, (2) 24-well plates (# 353047, Corning, Falcon^®) with glass coverslips inside for scanning electron microscopy analyses, and (3) 6-well plates (#353046, Dutscher, Falcon^®) for amastigote isolation from infected BMDMs.

Leishmania amazonensis amastigotes (LV79 strain, MPRO/BR/72/M1841) expressing mCherry (Lecoeur et al., 2020) were isolated from footpad lesions of infected Swiss nu/nu mice and purified as described previously (Courret et al., 1999). Infections were carried out at 34°C at a ratio of four amastigotes per macrophage. Leishmania donovani amastigotes (Ld1S2D strain, MHOM/SD/62/1S-CL2D) were isolated from infected spleens of female RjHan:AURA Golden Syrian hamsters and purified as described previously (Pescher et al., 2011; Prieto Barja et al., 2017). Infections were carried out at 37°C at a ratio of eight amastigotes per macrophage.

All pyroptosis experiments were performed after 3 days of infection using a sequential treatment of 500 ng/ml LPS (# LPS11-1, Alpha Diagnostic) for 4 h and 5 mM ATP (# A26209, Sigma-Aldrich) for different periods up to 4 h.

The following fluorescent reporters were added to the cell cultures 15 min before ATP stimulation: Hoechst 33342 (10 µg/ml) (# H3570, Invitrogen), LysoTracker Green DND-26 (1 µM) (LTG, # L7526, Invitrogen) and YO-PRO-1 (1 µM) (# Y3603, Invitrogen). The pyroptotic process was monitored at 34°C and 7.5% CO₂ using a fully automated spinning disk confocal microscope OPERA™ quadruple Excitation High Sensitivity (QEHS, PerkinElmer Technologies) with a 40× water immersion objective (Aulner et al., 2013). Image acquisition was performed every 5 min after ATP addition using the following sequential acquisition settings: (1) 405 nm laser line excitation, filter 450/50 for Hoechst 33342 detection, (2) 488 nm laser line excitation, filter 540/75 for LTG or YO-PRO-1 detection and (3) 561 nm laser line excitation, filter 600/40 for mCherry detection. Additionally, transmission light microscopy images were taken for analyses of the macrophage PV area and amastigote localization. A total of 15 fields at the same focal plane were taken every 5 min for every channel and for each sample. Images were transferred to the Columbus Conductor™ Database (Perkin Elmer Technologies) for storage and further analysis using specific scripts (Fig. S1). Single-cell analyses were performed using the ImageJ software package (https://imagej.nih.gov/ij/docs/faqs.html).

Pyroptotic BMDMs were fixed overnight with 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.2) at 4°C and post-fixed in 0.1 M cacodylate buffer (pH 7.2) containing 1% OsO₄. After serial dehydration, samples were critical-point dried (Emitech K850 or Balzers Union CPD30) and coated with gold using a sputter coater (Gatan Ion Beam Coater 681). Scanning electron microscopy observations were made with the JEOL 7600F microscope. Images were colorized using Adobe Photoshop CS software.

For transmitted electron microscopy analysis, cells were cultured on coverslips and were fixed with 2.5% of glutaraldehyde (Sigma-Aldrich) in PHEM buffer (120 mM PIPES, 50 mM HEPES, 20 mM EGTA, 4 mM MgCl₂, pH 7.3). Post fixation was performed with 1% osmium tetroxide (EMS 19152 2% aqueous solution, Merck) and 1.5% ferrocyanide (# 8131, Sigma-Aldrich) in PHEM. After dehydration with a graded series of ethanol from 25 to 95% grade 1 (99.8% pure ethanol), the samples were infiltrated with epoxy resin. The 70 nm sections obtained by thin sectioning using a Leica UC 7 microtome (Leica Microsystems), were collected on formvar coated slot grids (EMS 215-412-8400) and were contrasted with 4% uranyl acetate and Reynolds lead citrate (#11300, Delta Microscopies). Stained sections were observed with a Tecnai spirit FEI operated at 120 kV. Images were acquired with FEI Eagle digital camera.

Cells were lysed in RIPA buffer (# R0278, Sigma-Aldrich) supplemented with a cocktail of anti-protease and anti-phosphatase inhibitors (MS-SAFE, Sigma-Aldrich). Proteins were resolved by SDS-PAGE on NuPAGE gels (4–12% Bis-Tris) in MOPS buffer and electroblotted onto polyvinylidene difluoride (PVDF) membranes in transfer buffer (Lecoeur et al., 2020). Membranes were blocked with 5% fat-free milk in 1× Tris-buffered saline containing 0.25% Tween 20 and then probed overnight at 4°C with anti-capase-1 antibody (# AG-20B-0042, Adipogen Life Sciences) and further incubated incubation with anti-mouse-IgG peroxydase conjugate secondary antibodies. Membrane signals were revealed by SuperSignal West Pico reagent (#10481945, Thermo Fisher Scientific) in a high-resolution PXi machine (Syngene).

Supernatants of BMDM cultures (pyroptotic and control BMDMs) were carefully replaced by fresh medium without LPS or ATP. After detaching adherent macrophages using a plastic cell scraper (#541070, Greiner bio-one) and multiple passages through a 27-gauge needle to free amastigotes from cell remnants, isolated amastigotes were centrifuged (1500 g for 10 min), washed in PBS, counted and used to assess parasite viability, the capacity of amastigotes to differentiate into proliferating promastigotes in vitro, and amastigote virulence by infection of naïve BMDMs.

Amastigotes isolated from Leishmania-infected, pyroptotic and control BMDMs were seeded in a 96-well plate (# 353072, Dutcher, Falcon^®) at a final concentration of 5×10⁶ parasites per ml and incubated for 10 min with YO-PRO-1™ iodide at 0.1 µM final concentration (# Y3603, Thermo Fisher Scientific). Samples were then immediately analyzed on the CytoFLEX cytometer (Beckman Coulter) in a BSL2 containment to evaluate YO-PRO-1 incorporation. Data were analyzed using the Kaluza 1.5 software (Beckman Coulter Life Sciences). Lesion-derived amastigotes were treated with 70% ethanol for 10 min and were included as a positive control for cell death and YO-PRO-1 incorporation.

Amastigote to promastigote differentiation was analyzed by scanning electron microscopy (as described above) directly on amastigotes exposed on pyroptotic macrophages after 24 h incubation in 1 ml of promastigote culture medium at 27°C. For growth analysis in culture, purified amastigotes were incubated at 27°C in promastigote culture medium at 10⁵ parasites/ml. Promastigote culture density was determined daily during 6 consecutive days by visual counting using a Malassez chamber. The parasite metabolic status was analyzed at days 2 and 3 through a resazurin assay. Briefly, 200 µl of parasite culture was transferred into wells of a 96-well plate (#655090, Sigma-Aldrich) and incubated with 2.5 µg/ml of resazurin (#R7017, Sigma-Aldrich) for 4 h at 27°C. The fluorescence intensity of the resazurin-derived resorufin was determined using a Tecan Safire2 plate reader (558 nm excitation, 585 nm emission) (Lamotte et al., 2019).

Amastigote virulence was assessed using two protocols. First, isolated amastigotes from Leishmania-infected, pyroptotic and control BMDMs were added at a ratio of four parasites per macrophage on naive BMDMs seeded on glass coverslips placed in 24-well plates. At 1 and 2 days after incubation at 34°C, amastigote-infected BMDMs were fixed with 4% paraformaldehyde (PFA) (#15710, Electron Microscopy Sciences) and incubated for 10 min in PBS containing 10 µg/ml of Hoechst 33342. Image acquisition was performed with the Axio Imager (Zeiss) (405 nm laser line excitation, filter 450/50). The number of parasites per macrophage was determined by visual counting, considering at least 100 macrophages. Second, the culture medium of pyroptotic macrophages was carefully replaced by pre-warmed BMDM medium containing freshly differentiated, naïve BMDMs (1:1 ratio between naïve and pyroptotic BMDMs). After 3 days later, live-images combining bright field and mCherry fluorescence were acquired using the EVOS microscope (Thermo Fisher Scientific) and analyzed with the FIJI software for infection of the naïve BMDMs.

We thank Drs Jacomina Krijnse Locker and Geneviève Milon for scientific discussions, and Dr Nathalie Aulner for help with the Opera system.