A major Toxoplasma serine protease inhibitor protects the parasite against gut-derived serine proteases and NETosis damage

Abstract

Toxoplasmosis is a life-threatening opportunistic infection in immunocompromised patients, caused by the parasite Toxoplasma gondii. Infection is initiated through oral ingestion of Toxoplasma cysts that must survive the harsh environment of the gut to undergo excystation. Released parasites invade intestinal epithelial cells and then disseminate throughout tissues for encystation, mainly in the brain. How Toxoplasma escapes destruction mediated by gastrointestinal proteases is poorly understood. T. gondii has nine genes encoding serine protease inhibitor proteins (TgPIs). TgPI-1 is highly expressed across all Toxoplasma strains and developmental stages and contains three domains for binding to various serine proteases. Here, we explore the role of TgPI-1 in protecting Toxoplasma against serine proteases in the gut and neutrophil-derived proteases in the lamina propria. TgPI-1 localizes to the parasite plasma membrane and cyst wall. We generated ΔTgPI-1 parasites, and the mutant is more sensitive to neutrophil elastase (NE), trypsin and chymotrypsin than WT. Neutrophils exposed to Toxoplasma release neutrophil extracellular traps (NET) with strain-dependent morphologies, ranging from spiky to extended cloudy. TgPI-1 was detected on NET containing NE, and ΔTgPI-1 parasites are more susceptible to destruction by NETosis. In mice, ΔTgPI-1 parasites exhibit reduced infectivity, poor dissemination to abdominal organs, and lower cyst burden in the brain. These findings shed light on a strategy employed by Toxoplasma to counteract enzymatic antimicrobial defenses in gut tissues, highlighting potential avenues for controlling tissue dissemination of this medically significant parasite.

Pathogens that are orally ingested by a mammalian host confront harsh conditions: the acidic environment of the stomach, digestive enzymes secreted by the pancreas into the gastrointestinal (GI) tract, and local gut immune responses. In humans, Coccidia are gastrointestinal parasitic protozoa that infect a wide range of people and can cause severe diseases, particularly in patients with deficient immunity. Some coccidian parasites are capable of surviving in the GI system (e.g., Cryptosporidium, Isospora, and Sarcocystis) while others spread to other parts of the host to escape the hostile environment of the intestinal tract (e.g., Toxoplasma). Once ingested as cysts, both intestinal or nonintestinal coccidian parasites are dangerously exposed to proteolytic enzymes, such as pancreatic trypsin and chymotrypsin that are released into the duodenal lumen, yet some survive and invade intestinal epithelial cells to multiply within a parasitophorous vacuole (PV). Furthermore, Coccidia that exit the GI system, migrate to the lamina propria where they encounter neutrophils and monocytes that secrete serine proteases, such as neutrophil elastase (NE) and cathepsin G.

Expression and secretion of serine protease inhibitor proteins by invading microbes may represent a safeguarding mechanism for survival in the GI tract. These inhibitors are grouped into approximately 20 families, defined by sequence homology, reactive site location, structural characteristics, and mechanism of action (canonical, non-canonical inhibitors, and serpins) (1, 2). Serpins and proteases form irreversible complexes, associated with the disruption of the protease active site, while canonical Kazal-type inhibitors form tight but reversible interactions, by blocking the protease active site. Interestingly, all Coccidia contain at least one gene encoding a serine protease inhibitor (www.veupathdb.org): nine genes for Toxoplasma gondii (two serpins and seven Kazal); four genes for Bestoitia besnoiti (at least, one serpin and one Kazal); two genes for Sarcocystis neurona (one serpin and one Kazal), at least one gene for Neospora caninum (one putative Kazal), Cystoisospora suis, Hammondia hammondi, Cryptosporidium parvum (a Kazal), Eimeria brunetti and Cyclospora cayetanensis. Of note, Coccidia belongs to the phylum Apicomplexa and noncoccidial apicomplexan parasites that are not orally acquired and have no enteric stage in the host (such as Plasmodium sp. or Theileria bovis), do not have any genes coding for serine protease inhibitors, pointing to a potential protective role for these protein inhibitors against host serine proteases in the GI environment.

Infection by T. gondii occurs via oral ingestion of tissue cysts (in contaminating food) or environmental oocysts (shed by Felida) that rupture in the upper intestine, releasing bradyzoites and sporozoites, respectively (3, 4, 5). The excysted zoites invade epithelial cells of the small intestine within 30 min post-infection (p.i.), forming a PV. Over the next 24 h p.i., intravacuolar bradyzoites and sporozoites transform into fast-replicating tachyzoite forms. One to 2 days p.i., tachyzoites migrate through the basement membrane to colonize the lamina propria, the loose connective tissue beneath the epithelium (6). The lamina propria contains many immune cell types that play roles in the innate and adaptive defense against pathogens. Attracted by chemokines secreted by enterocytes, neutrophils are one of the first cells to arrive at the infection sites of T. gondii (7, 8). In general, neutrophils display a wide range of effector mechanisms to counteract pathogens, such as phagocytosis, oxidative burst, and degranulation releasing anti-microbial peptides and enzymes in the lamina propria (9, 10, 11). In addition, these cells can undergo a form of cell death termed “netosis” in which their DNA is released to form large, sticky, web-like structures outside the cell, known as NETs (12, 13). Netosis has evolved to amplify immune responses to pathogens by enabling neutrophils to capture and kill a variety of microbe invaders (14, 15, 16, 17, 18). Indeed, the NET scaffold contains various anti-microbial proteins, such as α-defensins, LL-37, histones, lactoferrin, and calprotectin, and several enzymes, such as lysozyme, NADPH oxidase and serine proteases (e.g., NE and cathepsin G). After reaching the lamina propria, T. gondii tachyzoites invade monocytes and neutrophils (6, 19), thus escaping NET destruction. From day 6 to 8 p.i., surviving tachyzoites enter blood vessels and lymphatics to disseminate through the body with a predilection for the brain, heart, and skeletal muscles where they form tissue cysts. Thus, Toxoplasma needs to be well-equipped to avoid degradation by many gut-derived serine proteases.

Among the seven serine protease inhibitors (TgPIs) of the Kazal-type expressed by T. gondii, TgPI-1 is highly expressed as two alternatively spliced isoforms (TgPI-1α and TgPI-1β) in all stages of the parasite development from Type I, II, and III strains (TGME49_217430 in www.ToxoDB.org). Functional studies reported that recombinant TgPI-1 isoforms inhibit a broad range of serine proteases, including trypsin, chymotrypsin, NE, and, to a lesser extent, pancreatic elastase (20). Recombinant TgPI-1β behaves as a tight-binding inhibitor since all K_i values are several orders of magnitude lower than the K_m values for the respective protease substrates (20). TgPI-2 mRNA levels are ∼10-fold lower than those of TgPI-1, and recombinant TgPI-2 (TGME49_208450) mainly inhibits trypsin (21). A previous study showed that ΔTgPI-1 parasites are viable but exhibit signs of stress associated with increased bradyzoite gene expression for differentiation (22). Of note, ΔTgPI-1 parasites show no changes in expression profiles for the eight other genes encoding serine protease inhibitors, suggesting no compensatory mechanism in the mutant.

At the onset of infection, Toxoplasma elicits the recruitment of neutrophils in gut submucosal tissues, and these cells are important immunomodulators, protecting the host from uncontrolled tachyzoite replication (23, 24, 25). Depletion of neutrophils in mice at the time of infection is associated with decreased levels of the cytokines IFN-γ, TNF-α, and IL-12, known to be important in parasite control. Indeed, neutrophil-depleted mice have a higher parasite burden in tissues from the liver, spleen, lung, and brain. Toxoplasma infection triggers the production of NET from neutrophils from diverse host species, as demonstrated in vitro (26, 27, 28, 29, 30, 31) and mouse models (32), resulting in their entrapment in NET scaffolds.

Toxoplasma secretes several proteins able to modulate host innate immunity (33) but, so far, the nature of the parasite effector/s that counteract the antimicrobial activities in the GI tract and confer resistance to neutrophil microbicidal activities such as NET-mediated entrapment and killing, remains largely unknown. In this study, we have investigated the contribution of TgPI-1 to protect T. gondii against deleterious effects of serine proteases present in the gut environment as the parasite traverses the GI tract and on NET structures as it enters the lamina propria. Our observations suggest that T. gondii mitigates gut serine protease activities by secreting TgPI-1. Furthermore, TgPI-1-deficient parasites lost their infectivity in the host, likely due to their destruction in the gut and impaired dissemination to systemic tissues.

Results

After secretion from dense granules, TgPI-1 localizes to the plasma membrane of extracellular tachyzoites and intracellular bradyzoites, in addition to the cyst wall

Kazal-type protease inhibitors consist of one or more domains characterized by highly conserved amino acid sequences and molecular conformations; each domain has six conserved cysteine residues that form three intra-domain disulfide cross-links, and in multidomain inhibitors, each domain is interconnected by peptide spacers of variable lengths (34, 35). TgPI-1 is a multidomain protein harboring four Kazal domains (36): domains one and two have reactive sites that specifically interact with the active site of trypsin and trypsin-like enzymes, and domains three and four are specialized for interaction with the active site of elastase and chymotrypsin (Fig. 1A, panel a) (20). Similar to other coccidian Kazal-type inhibitors, TgPI-1 deviates from typical Kazal inhibitors, with atypical distribution of cysteines (i.e., short spacing between cysteines one and 2, as depicted in Fig. 1A, panel b). This unique combination of Kazal domains suggests a modulatory function of TgPI-1, enabling interaction with different target proteases through its independent domains (37). In addition to harboring four non-classical Kazal domains, TgPI-1 exists as two isoforms, possibly for executing precise spatial and temporal inhibition of specific serine proteases. We confirmed that the two isoforms, TgPI-1α and TgPI-1β, are secreted by Toxoplasma by collecting material secreted by extracellular tachyzoites of the type I RH strain (more amenable for purification and manipulation than bradyzoites isolated from cysts) for immunoblot analysis using anti-TgPI-1 antibody, revealing two bands: 42-kDa and 38-kDa at the respective size of TgPI-1α and TgPI-1β isoforms (Fig. 1B). EM immunostaining on extracellular tachyzoites using anti-TgPI-1 antibody shows gold particles in secretory dense granule organelles, indicating that TgPI-1 is a dense granule (GRA) protein (Fig. 1C). Immunofluorescence assays (IFA) using antibodies against TgPI-1 and SAG1, a plasma membrane marker, to stain extracellular tachyzoites illustrate intraparasitic puncta, reminiscent of a dense granule staining (Fig. 1C, panel a). Interestingly, some TgPI-1 dots were observed to colocalize with SAG1 (arrows in panel a). To verify if TgPI-1 is associated with the parasite plasma membrane, IFA were conducted under non-permeabilized conditions, and TgP1-1 signal was clearly observed as puncta along the parasite periphery (Fig. 1C, panel b, arrows).

Figure 1.

Modeling, secretion, and localization of TgPI-1 in extracellular tachyzoites and intracellular bradyzoites.A, structural model of TgPI-1. In panel a: shown is a Model constructed with Phyre2 identifying four TgPI-1 Kazal-type inhibitory domains modeled from the leech-derived tryptase crystallographic structure. In panel b: Structure of a Kazal signature showing the position of the six CYS engaged in S-S bonds. B, secretion of TgPI-1. Immunoblot on Excreted-Secreted Antigen (ESA) preparations collected from extracellular tachyzoites using anti-TgPI-1 antibody, showing the two isoforms: TgPI-1α and TgPI-1β. C, localization of TgPI-1 in extracellular tachyzoites by immunoEM on extracellular tachyzoites using anti-TgPI-1 antibody coupled to gold particles, showing TgPI-1 in dense granules (DG). D, localization of TgPI-1 in extracellular tachyzoites by IFA under condition of permeabilization (panel a) or not (panel b) using antibodies against TgPI-1 and the plasma membrane marker SAG1, showing TgP1-1 in intraparasitic puncta for TgPI-1 and in discrete areas at the plasma membrane. Arrows show regions of TgPI-1 and SAG1 colocalization. Representative images are shown from 52 extracellular parasites. E and F, localization of TgPI-1 on differentiated bradyzoites in culture. Infected fibroblasts after 2 days of differentiation were fixed for IFA using antibody against TgPI-1 and staining for the cysts wall using TRICT-lectin, showing TgPI-1 at the plasma membrane (arrow) and on the cyst wall. Representative images are shown from 37 PV. G, localization of TgPI-1 on brain cysts. A mouse was i.p. infected with ME49 Toxoplasma expressing GFP and its brain was harvested 50 days p.i. to isolate parasite cysts for IFA using anti-TgPI-1 antibody, showing accumulation of TgPI-1 on the cyst wall. A representative image is shown from 29 isolated cysts.

We next examined the localization of TgPI-1 on bradyzoites in cysts, as a proteomic study reported that several GRA proteins are exported to the cyst wall 8 days after differentiation in vitro (38). We generated cysts containing mature bradyzoites using an in vitro tachyzoite-to-bradyzoite differentiation protocol, in which tachyzoites are exposed to low CO₂ and a medium at higher pH in cultured cells (39, 40). Bradyzoites are encased within a cyst matrix surrounded by a thick cyst wall up to 850 nm (41, 42). The cyst wall contains several glycoproteins, such as CST1, targeted by cyst wall-specific Dolichos biflorus agglutinin (DBA) (43). We performed IFA on cultured cysts and observed TgPI-1 localized at the bradyzoite plasma membrane (Fig. 1E, arrow) and on the cyst wall, stained with TRITC-DBA (Fig. 1, E and F). The cyst wall structure of in-vitro-derived cysts is similar to that of in-vivo cysts that have developed in an animal brain (44). We then examined the distribution of TgPI-1 on cysts collected from the brain of a mouse infected with a type II cystogenic CZ1 strain that was engineered to express cytosolic GFP, and a cyst wall staining for TgPI-1 was confirmed (Fig. 1G). These observations reveal that following secretion, TgPI-1 distributes at the cyst periphery, potentially preserving the cysts from degradation by gut proteases. Likewise, TgPI-1 located at the plasma membrane of bradyzoites may protect released bradyzoites from digestive enzymes in the intestinal lumen, and at the plasma membrane of tachyzoites, TgPI-1 may shield the parasite against serine proteases secreted by immune cells in the lamina propria.

In intracellular tachyzoites, TgP1-1 is secreted into the PV to localize to the intravacuolar network of tubules

Intracellular tachyzoites, despite being secluded in a PV, likely encounter survival challenges and may benefit from TgPI-1 to be protected against host cell-derived serine proteases. Intravacuolar tachyzoites secrete many GRA proteins that play diverse roles in transforming the PV into a replication-permissive niche (45). Some GRA are involved is the formation of an intravacuolar network (IVN) of membrane tubules that are distributed in the PV lumen (46, 47). Some IVN tubules connect to the PV membrane (PVM) (48), creating open gates that trap host macromolecules and organelles including lysosomes, to supply nutrients for Toxoplasma (49, 50, 51, 52). We analyzed the distribution of TgPI-1 inside the PV of replicating tachyzoites 24 h post-infection (p.i.) by performing double IFA using antibodies against TgPI-1 and GRA7, a GRA protein localized to the PVM and IVN (Fig. 2A, panel a). The fluorescence signal of TgPI-1 overlapped with that of GRA7 between parasites, forming a honey-comb mesh pattern for both proteins, which typifies the morphology of IVN structures (Fig. 2A, panel b). Unlike GRA7, no staining for TgPI-1 was seen at the PVM. EM immunogold staining on intracellular tachyzoites confirms the presence of gold particles on IVN tubules for TgPI-1; no gold particles were observed free in the PV or at the PVM (Fig. 2A, panels c and d). This suggests that the final destination of TgPI-1 in the IVN, where it may control activities of serine protease originating from host lysosomes (e.g., Cathepsin A or C) entrapped within IVN tubules.

Figure 2.

Localization of TgPI-1 in intracellular tachyzoites and viability of ΔTgPI-1 tachyzoites.A, localization of TgPI-1 in intracellular tachyzoites. Fibroblasts were infected for 24 before fixation for IFA using antibodies against TgPI-1 and the PVM/IVN GRA7, showing TgPI-1 secreted into the PV, not at the PVM (panel a). Panel b shows a higher magnification of the PV interior with IVN, illustrating an overlapped signal for TgPI-1 and GRA7. Representative images are shown from 44 PV. In panels c and d: intracellular parasites were processed for immunoEM 24 h p.i., showing gold particles on DG and IVN tubules. B, ΔTgPI-1 parasite replication. Quantitative measurement of the replication rate of WT and ΔTgPI-1 parasites 24 h p.i. assessed by parasite counting per PV, showing replication delay for the knockout. Data are means ± SD, n = 3 independent assays. PV sizes were statistically different between the WT and KO parasites (Chi-squared, p < 0.0001, all three biological replicates significant). C, ultrastructural examination of intravacuolar ΔTgPI-1 tachyzoites. EM studies on ΔTgPI-1 and WT parasites infecting HFF for 26 h before fixation. Panels a-c show 3 PV with a single parasite, illustrating abnormal granular parasites (asterisks) and abundant lipid droplets (LD) and amylopectin granules (AG). Panel d shows two parasites per PV, with aberrant structures in the PV lumen (inset). Panels e-f compare the morphology of four parasites per PV, looking more disorganized for the KO. Panels e’ and f’ are representative view of the IVN, with tubules attached to the PVM, having normal appearance (Bars is e’ and f; 50 nm). a, apicoplast; Go, Golgi; hc, host cell; hER, host endoplasmic reticulum; hmt, host mitochondrion; n, nucleus; P, parasite.

ΔTgPI-1 parasites suffer from replication defects

We generated a Toxoplasma strain lacking the TgPI-1 gene to assess the physiological relevance of TgPI-1 for parasite survival and development. CRISPR/Cas9 technology was used to genetically ablate the PI-1 gene through its replacement with the HXGPRT selectable marker cassette in RH parasites (Fig. S1A); the deletion of PI-1 and insertion of the HXGPRT cassette in parasites were verified at the genomic level by PCR (Fig. S1B), and at the protein level using anti-TgPI-1 antibodies by immunoblotting on lysates from tachyzoites and mouse cysts (Fig. S1C, panels a and b) and by IFA, showing no signal in intracellular knockout tachyzoites (Fig. S1D) and bradyzoites (Fig. S1E). We assessed the ability of RHΔTgPI-1 parasites for replication in fibroblasts during 24 h, in comparison to parental parasites (hereafter referred as WT) by parasite enumeration per PV. While the majority of WT PV contained up to 4 to 16 parasites, the majority of RHΔTgPI-1 PV had only two parasites (Fig. 2B), suggesting a slower replication rate for the mutant. To analyze the ultrastructure of RHΔTgPI-1 parasites, we performed EM 26 h p.i. The first cytopathy observed for ∼25% of RHΔTgPI-1 PV was the accumulation of large amounts of granular material in the PV of single parasites (Fig. 2C, panels and b, asterisks), with some parasites dying as shown in panel b. In surviving mutant parasites, most organelles were identifiable, with no obvious defects (Fig. 2C, panels c to e), and the PVM was closely associated with host ER and mitochondria as observed for WT parasites. A second abnormality of RHΔTgPI-1 parasites was the abundance of large lipid droplets, up to three per section as shown in panel c, suggesting substantial storage of neutral lipids. A third cytopathy of RHΔTgPI-1 parasites is the presence of many amylopectin granules, which are energy reserves that fuel the transition from proliferative tachyzoites to slow-growing bradyzoites and thus a sign of metabolic stress in tachyzoites. A fourth defect observed in the PV of replicating RHΔTgPI-1 parasites is the presence of many tubules and vesicles of unknown origin (Fig. 2C, panels d, inset). For large PV with more than two tachyzoites, mutant parasites usually appear skinner than WT parasites (Fig. 2C, panels e and f); no noticeable differences were observed in the IVN between the two strains (Fig. 2C, panels e’ and f’). Overall, these observations highlight a substantial contribution of TgPI-1 to the optimal development of intravacuolar tachyzoites as its absence causes excessive accumulation of lipid and sugar stores within the parasite and the formation of abnormal structures inside the PV.

ΔTgPI-1 parasites are less resistant to serine proteases exogenously added to the medium

Next, we examined the sensitivity of extracellular ΔTgPI-1 tachyzoites towards serine proteases exogenously introduced into the medium. RHΔTgPI-1 and control tachyzoites were exposed to NE, trypsin, chymotrypsin, or a PBS control for 30 min, followed by plaque assays (7 days) to assess parasite growth in cultured fibroblasts. A minor difference in growth rate was observed between the mutant and control parasites incubated with PBS, with 181 ± 9 and 212 ± 11 plaques (n = 4 independent assays; p < 0.0248), respectively, as expected based on the slower replication of the mutant (Fig. 2B). ΔTgPI-1 parasites were significantly less resistant to NE, trypsin, or chymotrypsin than control parasites (Fig. 3). This suggests that secreted TgPI-1 that is exposed at the parasite plasma membrane (Fig. 1C) may contribute to neutralizing the proteolytic activity of serine proteases in the environment.

Figure 3.

Replication of ΔTgPI-1 parasites and sensitivity to serine proteases in vitro. Sensitivity of ΔTgPI-1 parasites to serine proteases. 5 × 10⁷ extracellular RHΔTgPI-1 and RHWT tachyzoites were exposed to 0.1% neutrophil elastase (NE), trypsin or chymotrypsin, or PBS for 30 min at 37 °C, collected and 150 parasites were seeded in monolayers of fibroblasts before counting the lysis plaques 7 days p.i. Data normalized to percentage of RHWT or RHΔTgPI-1 parasites incubated in PBS, are means ± SD of four independent assays (unpaired t test).

Upon Toxoplasma contact, neutrophils release NET structures containing NE, MPO, and MMP-9

Upon neutrophil stimulation, NE is released from cytoplasmic granules and is translocated to the nucleus where it modifies the histone core leading to chromatin decondensation (18, 53). Subsequently, myeloperoxidase (MPO) enters the nucleus to enhance the expansion of nuclear DNA (54). Eventually, the nuclear envelope disassembles, and the decondensed nuclear chromatin is released into the cytoplasm of intact cells, mixing with cytoplasmic and granule components. Within 3 to 8 h after neutrophil activation, NETs are extruded into the extracellular space following plasma membrane rupture. These NETs contain granule-derived components including NE, MPO, cathepsin G and matrix metalloproteinase 9 (MMP-9), with MMP-9 recruiting additional neutrophils to the sites of infection (55, 56). Depending on the nature of the stimulus, NET structures can adopt different morphologies: spiky NET resulting from increased chromatin pressure at certain regions of the plasma membrane followed by membrane local rupture and the “shooting” of a DNA network like a string (e.g., after simulation with C5a or many bacteria); cloudy NET resulting the flowing at low intracellular pressures of the DNA network from the lysed plasma membrane, creating a characteristic cloud around the cell (e.g., in the presence of bacterial LPS or the phorbol myristate acetate (PMA), an activator of PKC); or aggregated NET associated with high neutrophil densities and rapid cell disintegration (e.g., with bicarbonate or monosodium urate (MSU) crystals during gout), demonstrating the plasticity of these cells in different microenvironments (reviewed in (57)).

Based on the sensitivity of ΔTgPI-1 parasites to exogenous NE, we examined their ability and their resistance to serine proteases (e.g., NE and cathepsin G) on NET structures released by neutrophils. A previous study reported a 50% reduction in the population of WT Toxoplasma after 6 h of contact with mouse neutrophils (32). If TgPI-1 provides protection against NET, ΔTgPI-1 parasites would be more susceptible to NET damage than WT. Prior to testing this hypothesis, we analyzed the morphology and composition of NET structures released by human and mouse neutrophils co-cultured with Toxoplasma WT using IFA and DAPI staining. Resting neutrophils have multilobed nuclei, stores of NE, Cathepsin G, and MPO in large primary/azurophilic granules and of MMP-8 in tertiary/gelatinase granules, as observed in human blood neutrophils (Fig. 4A). Upon activation of human neutrophils with PMA (positive control) for 4 h, the granular content of neutrophils (e.g., NE) is released into the extracellular medium where it surrounds extruded DAPI-positive material (Fig. 4B). Exposure of human neutrophils to RH Toxoplasma for 4 h triggers NET formation and entrapment of the parasites (immunostained for SAG1) (Fig. 4C). The SAG1 signal often appeared diffuse or punctate, a sign of parasite clumping or disintegration. NE, MMP-9, and MPO also comprised the NETome upon stimulation with Toxoplasma, as their fluorescence signals were also observed on extracellular DAPI. The human promyelocytic leukemia cell line HL-60 can be differentiated into mature, neutrophil-like myeloid cells upon treatment with dimethyl sulfoxide (58). Next, we examined NET release by HL-60-derived neutrophils in response to Toxoplasma, and showed NE and Histone H3 molecules studded across the DNA fibers (Fig. S2A, panels and b). In parallel, we probed NET induction by RH parasites from mouse neutrophils isolated from bone marrow, and similarly, the red fluorescence signal for Toxoplasma SAG1 was noticed on extracellular DNA filaments (Fig. S2B, panels a and b). To determine whether parasite viability influences NET release, we exposed human neutrophils to dead (heat-killed) parasites for 3 h and analyzed NET formation by DAPI staining. Unlike neutrophils exposed to live parasites, those incubated with dead parasites kept condensed nuclei that remained inside cells, with no sign of chromatin release (Fig. S3A). Heat-killed parasites were dispersed throughout the field while live parasites were entrapped within the NET (Fig. S3B, panel a). Measurement of DAPI signals showed no significant difference in chromatin area between resting neutrophils and neutrophils exposed to dead parasites (Fig. S3B, panel b).

Figure 4.

Release of NET structures from human neutrophils induced by RHWT Toxoplasma.A, IFA of activated human neutrophils using antibodies against Matrix Metalloproteinase-9 (MMP-9) and NE, showing colocalization of the two enzymes in granules of resting neutrophils and DAPI staining for nuclei. B, IFA of neutrophils activated by 25 nM phorbol myristate acetate (PMA) in DMSO for 4 h, or DMSO control antibody against NE, showing NE on extracellular DNA. C, IFA of activated neutrophils by Toxoplasma RHWT for 4 h using antibodies against SAG1 and NE, MMP-9 or MPO, illustrating release of the three enzymes on extruded DNA from neutrophils in contact with the parasites (green).

Type I strain triggers the formation of spiky and cloudy NET while the type II strain elicits the release of extended cloudy NET

We next investigated if the types of NET released from Toxoplasma-activated human neutrophils would differ between strains by comparing NET morphology upon exposure to Type I (virulent and lethal) or Type II (avirulent and cystogenic) strains. Human neutrophils were incubated with RH tachyzoites in vitro for 4 h, fixed and stained for DAPI and neutrophil proteins released from secretory granules. In three independent experiments, two distinct types of NET were observed: cloudy NET (60% of cell preparations) containing clusters of dead neutrophils and parasites glued within extruded DNA material positive for NE and spiky NET (40%) consisting of thin and elongated (up to 50 μm) DNA strands positive for NE, MMP-9 and MPO, with parasite attached to the strands (Fig. 5A). In contrast, when human neutrophils were exposed to GFP-ME49 tachyzoites, the vast majority of NET structures were extensively cloudy (95% cell preparations, with 5% cloudy NET), with the massive accumulation of DNA material with NE dispersed between dead neutrophils (Fig. 5B). GFP signal denoting parasites was observed on the extended cloudy NET.

Figure 5.

Types of NET structures released by human neutrophils exposed to RHWT and ME49RH Toxoplasma.A, IFA of activated human neutrophils after 4 h exposure to RHWT parasites before fixation and immunostaining for SAG1 and NE, MMP-9 or MPO, illustrating cloudy or spiky NET. B, IFA of human neutrophils after 4 h exposure to ME49WT parasites expressing GFP before fixation and immunostaining for NE, showing extended cloudy NET.

ΔTgPI-1 parasites have increased vulnerability to NETosis

To investigate NETosis susceptibly for RHΔTgPI-1 parasites, we first exposed human neutrophils to RHΔTgPI-1 parasites and used fluorescence microscopy to visualize released NET structures; in the presence of RHΔTgPI-1 parasites, neutrophils were activated and dispersed DNA and NE (Fig. 6A). Second, we compared the amount of extracellular DNA (measured by Picogreen binding) and of NE (determined by enzymatic activity) from neutrophils co-incubated with either mutant or WT parasites. No significant differences in DNA or elastase values were observed for either mouse or human neutrophils exposed to the two parasite strains (Fig. 6B, panels a-c). Third, as the recombinant TgPI-1 protein binds to NE (20), we performed IFA on NET structures to examine whether secreted TgPI-1 could interact and bind to NE on NET. We observed regions with TgPI-1 signal on NE-positive NET (Fig. 6C), suggesting that TgPI-1 either at the tachyzoite surface or secreted may be able to inhibit NE via specific binding.

Figure 6.

Release of NET structures from neutrophils induced by RHΔTgPI-1 parasites and NETosis sensitivity.A, IFA of activated human neutrophils after 4 h exposure to RHΔTgPI-1 parasites before fixation and immunostaining for SAG1 and NE, as compared to non-activated neutrophils distant from parasites. B, quantification of NETosis upon parasite contact. Upon 4 h coincubation of neutrophils and RHΔTgPI-1 or WT parasites (or PMA as control), extracellular DNA released from human (panel a) or mouse neutrophils (panel b) was measured using Quant-iT Picogreen assay for double-stranded DNA, and human elastase released from neutrophils (panel c) was measured enzymatically using N-methoxysuccinyl-Ala-Ala-Pro-Val-ρ-nitroanilide as substrate, showing no values in dsDNA and NE difference in values between mutant and WT. Data are means ± SD of three independent assays (unpaired t test). C, IFA using anti-TgPI-1 and NE antibodies, showing TgPI-1 on NE-containing NET structures. D and E, quantification of parasite viability upon neutrophil contact. RHΔTgPI-1 or WT parasites were exposed for 4 h to human neutrophils in the presence of cytochalasin D to block Toxoplasma invasion. Parasite viability was assessed by enumerating the live/dead parasites using the live/dead fixable Orange viability dye (D) and by counting the lytic areas in infected monolayers of fibroblasts in plaque assays (E) after seeding 1000 parasites for 4 days (panel a) or 200 parasites for 7 days (panel b). Data are means ± SD of three independent assays (unpaired t test).

To examine the sensitivity of RHΔTgPI-1 parasites to NETosis, we assessed the viability of parasites entrapped within NET using Orange Viability Dye, a live-cell exclusion dye that binds to dsDNA specifically within dead cells. Staining with the orange dye revealed that after 4h of contact between human neutrophils and tachyzoites, about 20% of RHWT parasites were dead versus 80% of RHΔTgPI-1 parasites (Fig. 6D). Next, we used plaque assays to confirm the differential killing of RHΔTgPI-1 and WT tachyzoites by human neutrophils releasing NET. Parasites were cultured with or without neutrophils for 6h in the presence of cytochalasin D, to prevent both invasion and phagocytosis for 6 h. Afterwards, 1000 parasites were collected and added to fibroblast monolayers for 4 days. The co-culture of parasites and neutrophils led to a significant reduction in the number of lysis plaques, with a 51% decrease for RHWT and a 73% decrease for RHΔTgPI-1 parasites (Fig. 6E, panel a). Plaque assays were repeated using 200 RHΔTgPI-1 and WT tachyzoites preincubated with neutrophils. After 7 days of fibroblast monolayer infection, there was a ∼70% reduction in plaque numbers with the mutant, compared to WT parasites (Fig. 6E, panel b). These data confirm the sensitivity of tachyzoites to NET, with a higher vulnerability of RHΔTgPI-1 parasites.

We also examined the morphology of the NET structures triggered by RHΔTgPI-1 parasites. Interestingly, RHΔTgPI-1 parasites elicited a distinctive starry-like NET morphology, characterized by short DNA spikes extruding from many areas of the plasma membrane, giving the neutrophils a star-like appearance (70% cell preparations with 30% cloudy NET) (Fig. 7A). This differs from the longer spiky NET structures induced by RHWT parasites (Fig. 5A). We also analyzed NET morphology upon the exposure of neutrophils to parasites lacking TgPI-1 in the type II strain ME49. NET structures were formed and NE was detected on released DNA filaments. However, while 95% of NET induced by ME49WT had extended cloudy morphology (Fig. 5B), the NET induced by ME49ΔTgPI-1 parasites had a spiky/starry morphology 40% of the time and the extended cloudy morphology 60%.

Figure 7.

Types of NET structures released by human neutrophils exposed to RH ΔTgPI-1 and ME49 ΔTgPI-1 Toxoplasma.A, IFA of activated human neutrophils after 4 h exposure to RHΔTgPI-1 parasites before fixation and immunostaining for SAG1 and NE, illustrating starry NET. B, FA of human neutrophils after 4 h exposure to ME49ΔTgPI-1 parasites before fixation and immunostaining for SAG1 and NE, showing extended starry/spiky or extended cloudy NET.

After mouse intraperitoneal infection, CZ1ΔTgPI-1 parasites are less infectious and form fewer brain cysts

In Toxoplasma-infected mice, a massive influx of neutrophils toward the peritoneal cavity is observed after mast cell degranulation from 12 h to 48 h p.i., reaching levels 30 times higher than in uninfected mice (59). Neutrophils play a critical role in restricting tachyzoite growth by lysing extracellular tachyzoites and retarding the division of intracellular tachyzoites (25, 60). By 48 h p.i., any surviving parasites spread from the peritoneal cavity, and invade the mesentery and adjacent organs, including the ileum. While mouse infection with the type I strain is typically fatal within 8 to 12 days post-inoculation, mice infected with tachyzoites from a cystogenic type II strain survived due to parasite dissemination throughout the body from approximately 10 days post-inoculation, followed by conversion to bradyzoites that form semi-dormant tissue cysts (61). Type II strains of Toxoplasma include ME49 (isolated from a Californian sheep) and CZ1 (isolated from a Czech Republic tiger): ME49 is more prevalent but in comparison, CZ1 grows and replicates slower in vitro, has reduced virulence and no mortality in mice at high infection doses (2000 tachyzoites) while producing a comparable number of Toxoplasma cysts than ME49. In addition, the CZ1 strain retains its cat competency for sexual reproduction, able to go through its entire life cycle. Because mice infected with CZ1 parasites routinely survive, we selected the CZ1 strain to compare the virulence of ΔTgPI-1 and WT parasites over time. CZ1ΔTgPI-1 parasites were generated and CD-1IGS mice were intraperitoneally (i.p.) infected with 5 × 10³ tachyzoites from CZ1ΔTgPI-1 and CZ1WT strains expressing luciferase; the weight of the two groups of mice was recorded daily and the parasite burden monitored by bioluminescence imaging. As shown in Figure 8A, mice infected with CZ1ΔTgPI-1 parasites maintained a stable weight up to 34 days p.i. whereas mice infected with WT experienced weight loss at the beginning of the infection, and then progressively recovered. Bioluminescence imaging of mice infected with CZ1WT tachyzoites illustrated an increase in the luminescence intensity within the peritoneal cavity from day 5 to 10, followed by a decline from day 12 to day 15. In contrast, mice infected with the CZ1ΔTgPI-1 parasites showed peak luminescence at day 7 before it declined (Fig. 8B with two representative mice shown). Quantification of bioluminescence values from six mice performed in a separate assay than shown in Figure 8A revealed statistically significant lower intensity values by ∼ 2 to three for the mutant (Fig. 8C).

Figure 8.

Growth features of CZ1ΔTgPI-1 parasites in mice after intraperitoneal infection.A, 800 tachyzoites from CZ1WT or CZ1ΔTgPI-1 strain expressing luciferase, or PBS alone were used to infect intraperitoneally CD-1IGS outbred mice (n = 6 mice for each strain) and the weight of the mice were monitored daily. Mean ± SEM plotted. B, Bioluminescence imaging of the infected mice as described in A using the IVIS system on days 5, 7, 10. 12 and 15 p.i. showing chemiluminescence signals in the peritoneal cavity of two representative mice for each strain (panel a). C, quantification of chemiluminescence signals in the peritoneal cavity. In an independent assay than shown in (B), CD-1IGS mice were infected as described in A, to record the intensity values expressed as total flux on photons per second. Data show a decline after day 10 more pronounced with the KO strain. Data are means ± SD with six mice per group of CZ1WT or CZ1ΔTgPI-1 tachyzoites (p values using 2-way ANOVA). D, Brain cyst morphology and abundance in the infected mice 50 dpi as described in A. Panel a: representative brightfield images of the mouse brain after staining of sections with H and E stain, identifying the cysts as spherical structures, nuclei (in blue-purple), showing two brain cysts for each strain. Panel b: quantification of cyst diameters using ImageJ, showing no signification difference in size between the WT and KO. Panel c: enumeration of cysts 50 dpi, showing a significant difference between WT and KO. Data for panels b and c are means ± SD from 26 brain slides with cysts.

In mice infected with the type II strain, a significant proportion of tachyzoites reach the brain around day 8 p.i. and progressively transform into cyst forms (61). We next examined the ability of CZ1ΔTgPI-1 parasites to disseminate and form tissue cysts in the brain. CD-1IGS mice i.p. infected with CZ1ΔTgPI-1 parasites were sacrificed at day 50, and their brains were harvested for cyst examination. Cysts containing CZ1ΔTgPI-1 and CZ1WT parasites were detected in the brain (Fig. 8D, panel a). Quantification of cyst size showed no significant difference in the diameter between the cysts of the two strains (Fig. 8D, panel b). However, the number of CZ1ΔTgPI-1 cysts was significantly lower (∼3-fold) compared to CZ1WT cysts (Fig. 8D, panel c). These observations suggest that the absence of TgPI-1 in type II strains results in the attenuation of virulence in mice, marked by reduced brain infection during the chronic stage of infection.

After mouse peroral infection, CZ1ΔTgPI-1 parasites exhibit poor dissemination in abdominal organs, resulting in lower cyst burden in the brain

Next, we examined whether TgPI-1 could protect parasites in the GI tract following oral infection of mice with infective cysts, e.g., containing bradyzoites. Ten tissue cysts with CZ1ΔTgPI-1 or CZ1WT parasites were collected from mice (as described in Fig. 8) and administrated to C57BL/6 inbred mice via oral gavage. The weight of mice was monitored daily post-gavage. Mice infected with the mutant maintained their initial weight while WT-infected mice gradually lost weight from day 11 until day 16 when they started to recover (Fig. 9A). On Day 6 or Day 8, the small intestine, spleen, and liver were harvested from mice to analyze the GFP fluorescence signal in each organ. Organs containing CZ1ΔTgPI-1 parasites showed less fluorescence as recorded for two representative mice (Fig. 9B). Quantification of bioluminescence data in a separate assay than shown in Figure 9B likewise revealed significant differences in bioluminescence values between organs infected with CZ1ΔTgPI-1 vs. CZ1WT parasites, with two to three reduction fold (Fig. 9C). Brains were harvested on day 45 post-gavage from infected mice, revealing significantly reduced fluorescence or bioluminescence signals in CZ1ΔTgPI-1-infected brains as shown in Figure 9, B and C, respectively. These data were further confirmed by brain cysts enumeration on day 50, showing a ∼90% reduction in cysts number in mice infected with CZ1ΔTgPI-1 parasites, with an average of 1800 vs. 210 cysts for WT and mutant, respectively (Fig. 9D). These observations point to a protective role of TgPI-1 during natural infection, as evidenced by poor dissemination in the GI tract of mice and lower brain infectivity.

Figure 9.

Growth features of CZ1ΔTgPI-1 parasites in mice after oral infection.A, Ten infectious cysts from CZ1WT or CZ1ΔTgPI-1 strain expressing luciferase, isolated from the brain of CD-1IGS mice, were used to orally infect C57BL/6 mice inbred mice (n = 9 mice for each strain) and the weight of the mice was monitored daily. Mean ± SEM plotted. B, fluorescence imaging on isolated organs from the infected mice was performed as described in A using the IVIS system with fluorescence filters in place: small intestine six dpi, spleen and liver eight dpi, and brain 45 dpi, showing less chemiluminescence signal for the KO strain. C, quantification of chemiluminescence signals in the small intestine six dpi, spleen and liver eight dpi, and brain 45 dpi. In an independent assay than shown in B, CD-1IGS mice were infected as described in A, to record the intensity values expressed as total flux on photons per second, showing lower values for the mutant. Data are means ± SD with nine mice per group of CZ1WT or CZ1ΔTgPI-1 tachyzoites (p values using 2-way ANOVA). D, enumeration of cysts 50 dpi, showing a significant difference between WT and KO. Data are means ± SD from eight mice (p value using 2-way ANOVA).

Discussion

Many parasites express a diverse array of proteinase inhibitors that play crucial roles in parasite development and survival. Reasonably, we could assume that the primary function of these inhibitors is to regulate aberrant proteolytic events within the parasite itself. However, many of these inhibitors are detected in the secretome or on the parasite surface, pointing out that their targets lie at the host-parasite interface and that they have been specifically adopted to modulate the degradative activities of proteases derived from the host. For instance, T. gondii expresses eight serine protease inhibitors, and among them, TgPI-1 has a broad spectrum of activities, for binding to trypsin, chymotrypsin, and elastase in vitro whileTgPI-2 specifically targets trypsin (20). Our study highlights two potential functions for TgPI-1 in vivo: (i) protection from digestion in the GI tract by inactivating digestive enzymes, e.g., trypsin, and chymotrypsin, and (ii) counteraction of the host anti-parasite response, particularly NETosis in the lamina propria by neutralizing serine proteases released by neutrophils, e.g., NE and cathepsin G. Interestingly, it has been reported that acute and chronic T. gondii infections can reduce the development of allergic airway inflammation by promoting a Th1 phenotype and inducing regulatory T cells (62, 63). Serine proteases such as trypsin or NE may contribute to asthma pathophysiology through the activation of Protease-activated Receptor 2 (PAR-2) leading to inflammation (64, 65, 66). Through the exploration of immunomodulatory proteins in Toxoplasma, the anti-allergic role of TgPI-1 has been investigated in a mouse asthma model: intranasal administration of the recombinant TgPI-1 protein induced a reduction of asthma hallmarks, suggesting a tolerogenic property by inhibiting serine proteases involved in immune responses and inflammation (67). Thus, this finding expands the role of TgPI-1 beyond protection against gut serine proteases.

Many secreted serine protease inhibitors have been identified in parasitic helminths, either secreted or localized at the tegument of the worm (reviewed in (68, 69, 70)). For intestinal worms, such as Ascaris suum in the gut and A. lumbricoides in the small intestine, these inhibitors protect against their proteolysis by pancreatic proteinases. Hematophagous worms, such as Schistosoma mansoni or Ancylostoma spp. that live in blood vessels, secrete small serine proteases inhibitors (i.e., smapins) to inhibit host serine proteases involved in the blood coagulation cascade (e.g., Factor Xa) thereby preventing blood-clotting and enabling feeding. Similarly, many worms secrete inhibitors that counteract serine proteases involved in the immune response. For example, mast cells secrete various serine proteases, such as chymase and tryptase that promote vascular permeability and recruit monocytes and neutrophils, contributing to the expulsion of intestinal nematodes. Other serine proteinases, such as thrombin, NE, and cathepsin G, released from mast cells and neutrophils, trigger inflammatory processes in response to infection. Worms like Brugia malayi (in lymphatic vessels) or Trichuris suis (in the cecum) secrete their serine protease inhibitors to dampen intestinal mucosal cell-associated, protease-mediated, host immune responses (71, 72). Downregulation of immune reactions by serpins from parasitic helminth is not only critical for parasite survival but also enhances the long-term survival of the host by modulating immunopathogenic inflammatory responses. Overall, in the broad context of parasitism and the adaptability of a parasite to a host (or vice versa), numerous examples point to a co-evolution of proteinases and proteinase inhibitors, reflecting an evolutionary ‘arms race’ in host-parasite relationships.

Like serine protease inhibitors localized on the cuticle of the nematode to trap host enzymes, TgPI-1 is distributed along the cyst wall encapsulating bradyzoites and is located at the plasma membrane of both tachyzoites and bradyzoites. In the gut lumen, prior to cyst wall rupture, TgPI-1 may form covalent complexes with trypsin and chymotrypsin, inhibiting their activities and ensuring bradyzoite protection against the detrimental effects of these intestinal proteases. It would be interesting to examine whether TgPI-1 also localizes at the oocyst wall of environmental cysts containing sporozoites, in anticipation of oral ingestion by a host. Like other dense granule proteins, TgPI-1 is constitutively secreted, and at the parasite surface, it may continue to protect free bradyzoites released in the gut lumen upon cyst wall rupture. During mammalian cell invasion, TgPI-1 may also shield Toxoplasma from plasma membrane-anchored serine proteases (73). Intracellular tachyzoites export TgPI-1 to IVN tubules. Attached to the PVM, IVN tubules sequester host organelles, such as endolysosomes as sources of nutrients (e.g., lipids) (44, 45). This scenario involves a Toxoplasma phospholipase A2 (TgLCAT) (74) localized to the IVN (44) that presumably disrupts the lipid bilayer of intra-PV host organelles to promote the release of their content. In this context, the localization of TgPI-1 on the IVN may serve as a safeguard strategy to protect the PV environment against lysosomal serine proteases, such as cathepsins A and G. This scenario is supported by our EM observations of ΔTgPI-1 tachyzoites that identify abnormalities in the PV content with accumulation of structures surrounding damaged or stressed parasites. The relocalization of TgPI-1 from inside the PV of tachyzoites to the cyst wall during differentiation into bradyzoites reflects an adaptative strategy of Toxoplasma to neutralize host serine proteases in different environments during stage transition.

Neutrophils serve as the first line of immune defense against microbial pathogens. The innate immunity functions of these cells are mainly mediated through phagocytosis, degranulation, and NET formation. Neutrophils can ‘sense’ the size of microbes and adjust their response accordingly: small single bacterium and small yeast are internalized by phagocytosis for degradation upon fusion of the phagosome with azurophilic granules, while larger preys, such as yeast hyphae (∼3 up 100 μm) or bacterial aggregates (>5 μm) are entrapped into the extracellular DNA-web like filaments of NET (75, 76). Interestingly, some bacteria, such as Streptococcus pneumoniae and Haemophilus influenza that are protected by a capsule allowing evasion of complement-dependent neutrophil phagocytosis (i.e., opsonophagocytosis), trigger NET release by neutrophils (77, 78). For these encapsulated bacteria, the NET induction factor has been identified: α-enolase for S. pneumonias that binds to the Myoblast antigen 24.1D5 at the surface of neutrophils (79) and specific subsets of lipooligosaccharides for H. influenza that activate toll-like receptor signaling in neutrophils (80). To circumvent NET-mediated destruction, many pathogens including S. pneumonia secrete nucleases that degrade the NET DNA scaffold, allowing escape from the antimicrobial molecules within NET (81). Toxoplasma, regardless of its size (3 × 7 μm), has evolved strategies to avoid phagocytic killing by actively invading neutrophils and forming a nonfusogenic PV. In response, neutrophils then release NET to kill the parasite extracellularly (26–32; this study). Whether is the ‘failure’ to phagocytose Toxoplasma or the recognition of specific parasite effectors by neutrophils that elicit NETosis, the extent of NET structures ejected to target the parasite, is impressive. The distinct NET morphologies observed upon contact with type I and II strains of Toxoplasma may be attributed to different, strain-specific factor/s present in the secretome or associated with the plasma membrane. To this point, neutrophils are able to discriminate between the lipopolysaccharides of different bacterial sources, with the types of LPS-induced NET influenced by the sugar composition of the O-antigen; this specificity is not only bacterial species-specific but also serotype-specific (82).

It would be interesting to explore whether the release of DNA spikes or clouds triggered by type I and type II strains respectively, correlates with the ability of a strain to migrate rapidly away from a focus of infiltrated neutrophils. Differences in speed and migration have been observed among Toxoplasma strains in vitro, with type I parasites exhibiting enhanced migration ability due to greater motility, as compared to types II or III parasites (83). This phenotype has also been observed in vivo wherein a higher number of type I parasites are found in deeper tissues (e.g., the lamina propria and the submucosa) and the submucosal blood vessels as early as 12 h p.i., whereas type II parasites require to 2 to 3 days to reach similar locations. Strategically, neutrophils, which are relatively slow-moving cells, and thus unable to chase highly motile pathogens may expel very long DNA spikes to efficiently trap RH parasites, in contrast to spreading DNA clumps over short distances to target slower moving ME49 parasites. In support of this hypothesis, flagellated Pseudomonas aeruginosa displaying strong swimming motility, induce robust NET formation while nonmotile flagellum-deficient or flagellar motor-deficient strains trigger only minimal extracellular DNA release (84). The flagellum in P. aeruginosa has been identified as the primary component responsible for NET extrusion, likely through the binding to an as-yet unidentified protein at the surface of neutrophils. In parallel, neutrophils do not form NET when exposed to dead Toxoplasma, possibly due to the immobility of the parasites. Alternatively, the signal for NET release by neutrophils in the presence of (live) Toxoplasma may originate from the parasite secretome, which comprises approximately 100 proteins (85), rather than the parasite surface or a cell contact-dependent mechanism, as obviously dead parasites lacking secretory activity, fail to induce NETosis.

For Toxoplasma, the specific component/s responsible for inducing NET formation remains to be identified. Preliminary studies on the mechanism of NET release induced by Toxoplasma suggest the involvement of the ERK1/2 MAPK signal pathway (32), with the production of reactive oxygen species (ROS). In this model, ROS facilitates the oxidative dissociation of the NE from the azurosome complex at the limiting membrane of azurophilic granules (28, 29, 31). The dissociated NE subsequently translocates to the nucleus for chromatin decondensation, as demonstrated in other NET-inducing parasitic protozoa, such as Leishmania sp., Entamoaeba histolytica, Eimeria sp., and N. caninum, that induce NET (14, 86). ROS production in Toxoplasma-activated neutrophils seems to depend on store-operated calcium entry (SOCE) (26), which leads to the breakdown of the nuclear envelope, a hallmark of suicidal NETosis (30).

To counteract the detrimental effects of NETs, pathogens have developed various strategies (reviewed in (87)): inhibition of NET release by down-regulating host inflammatory responses (e.g., induction of the NET-suppressive IL-10 to block TLR-induced ROS generation) (88); dismantlement of NET structures (e.g., secretion of nucleases) (81); or resistance to the microbicidal components of NETs (e.g., formation of a physical barrier: bacterial film, capsule, glycocalyx of leishmanial lipophosphoglycan) (89, 90). In humans, uncontrolled NET production is known to exacerbate the pathology of autoinflammatory and autoimmune diseases (91). However, neutrophils contain intrinsic cytosolic regulators of NETosis, such as serpins (e.g., SerpinB1), which modulate the activity of NE, cathepsin G and proteinase-3 (92, 93). Analogous to mammalian serpins, TgPI-1 has been identified in this study as a protective factor that confers resistance to Toxoplasma against neutrophils and their arsenal of serine proteases. Our findings highlight that TgPI-1-deficient Toxoplasma are more vulnerable in the gut environment as evidenced by a reduced number of mutant parasites in GI organs. In the absence of TgPI-1, the eight other TgPIs are not upregulated (22). If secreted, however, these inhibitors may provide an additional layer of protection by targeting specific host serine proteases. We reason that the large repertoire of TgPIs may correlate with their physiological relevance for Toxoplasma virulence and establishment in the host. From an evolutionary perspective, proteinase inhibitors are one of the most actively evolving proteins, making them fascinating model proteins for studying host-pathogen interactions, particularly in the context of immune evasion and pathogen survival.

Experimental procedures

Chemicals, reagents and antibodies

All reagents were obtained from Sigma-Aldrich, unless otherwise stated. Human neutrophil elastase was from Calbiochem. d-Luciferin potassium salt was from Xenogen. Phorbol myristate acetate (PMA), S7 Nuclease, and methoxysuccinyl-Ala-Ala-Pro-Val-ρ-nitroanilide were from Cayman Chemical. BD Vacutainer EDTA tubes were from Becton, Dickinson and Company and the EasySep Direct Human Neutrophil Isolation Kit was from StemCell Technologies. The PrestoBlue Cell Viability Reagent, the Quant-iT PicoGreen dsDNA Reagent, and the LIVE/DEAD Fixable Orange (602) Viability kit were from Thermo Fisher Scientific Inc. The following primary antibodies used for the immunofluorescence assays (IFA) or immunoblotting included: rabbit anti-TgPI-1 (36); rabbit anti-TgPI-1 (gift from V, Carruthers, University of Michigan); rabbit anti-GRA7 (94), rabbit anti-SAG1, mouse anti-SAG1 (TG05–54) (95), and mouse anti-GRA3 (T6-2Hll) (96) were gifts from J.-F. Dubremetz, Université of Montpellier; mouse anti-neutrophil elastase (Abcam); rabbit anti-neutrophil elastase (Thermo Fisher Scientific Inc.); mouse anti-histone H3 tri-methyl K27 (Abcam); rabbit anti-myeloperoxidase (Abcam), mouse anti-matrix metalloproteinase-9 (Thermo Fisher Scientific Inc.) and mouse anti-αtubulin (Abcam). Secondary antibodies against host animals used for antibody production included: IgG Alexa-488, Alexa-555, Alexa-568, or Alexa-647.

Mammalian cell lines, parasite strains, and culture conditions

Human foreskin fibroblasts (HFF) and hTERT-immortalized fibroblasts obtained from the American Type Culture Collection were grown as monolayers and cultivated in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2 mM glutamine and penicillin/streptomycin (100 units/ml per 100 μg/ml), and maintained at 37^oC in 5% CO₂ unless specified otherwise. The HL-60 cell line, kindly provided by J.S. Dumler (Uniformed Services University of the Health Sciences), was cultivated in RPMI-1640 with 10% FBS, 2 mM Glutamine, and penicillin-streptomycin (100 U/ml and 100 μg/ml, respectively) and maintained at 37 °C in a humidified atmosphere containing 5% CO₂. Differentiation of HL-60 cells was induced by treatment with 1.25% dimethyl sulfoxide (DMSO) for 7 days. Post-differentiation, cells were washed with phosphate-buffered saline (PBS), resuspended in RPMI 1640 medium, and prepared for NETosis assays. The tachyzoites from the RH strain (Type I lineage) or ME49 and CZ1 (Type II lineage) were propagated in vitro by serial passage in monolayers of fibroblasts as described (97). For the engineering of RHΔTgPI-1-deficient parasites, the strain RHΔK80 obtained from the NIAID Reagent Program was used. Parasite strains expressing GFP or GFP-luciferase were generously provided by D.S. Roos (University of Pennsylvania) and ME49 or CZ1ΔTgPI-1-deficient parasites generated in the lab of M.E. Grigg (NIAD) in the background of these strains were kindly provided to us.

Mice

Male or female CD-1IGS and C57BL/6 mice were purchased from the Jackson Laboratory or Taconic Farms and used at 5 to 8 weeks of age. All procedures were performed in compliance with the Public Health Service Policy on Humane Care and Use of Laboratory Animals and the Association for the Assessment and Accreditation of Laboratory Animal Care guidelines. The animal protocol on infected mice with T. gondii was approved by the Institutional Animal Care and Use Committee from both the Johns Hopkins University and the National Institutes of Health, NIAID.

Mouse neutrophil isolation

Male mice were administered an intraperitoneal injection of 1 ml of 7.5% casein solution (Cayman Chemical). Approximately 24 h post-injection, the animals were euthanized and subjected to a second intraperitoneal injection consisting of 5 ml of neutrophil isolation medium (PBS supplemented with 10% bovine serum albumin (BSA)). Peritoneal lavage was collected, and neutrophils were purified by overlaying the lavage onto a 63% Percoll gradient (Cayman Chemical), followed by centrifugation at 1000g for 20 min without applying a brake. The neutrophil-rich layer was then isolated, mixed with neutrophil isolation medium, and centrifuged at 500g for 10 min. The resultant cell pellet was resuspended in a neutrophil isolation medium and utilized for the Neutrophil Extracellular Trap (NET) assay. Neutrophil viability was assessed using the PrestoBlue Cell Viability Reagent.

Human neutrophil source and isolation

Anonymous human blood used for neutrophil isolation was obtained under IRB protocol NA 00019050 approved by the Johns Hopkins School of Public Health Ethics Committee. Human neutrophils were isolated from whole blood using the EasySep Direct Human Neutrophil Isolation Kit (StemCell Technologies) following the manufacturer’s protocol with slight modifications. Whole blood, freshly collected in vacutainer EDTA tubes was transferred to a 50 ml conical tube, in which Isolation Cocktail (50 μl/ml) and RapidSpheres (50 μl/ml) were added. The mixture was incubated for 10 min at room temperature. Following incubation, the tube was placed on a magnet for 10 min to facilitate magnetic separation. The cell suspension was carefully transferred to a new conical tube, supplemented with an additional 50 μl/ml of RapidSpheres, and incubated for another 10 min. Magnetic separation was repeated as described, and the steps were performed once more to enhance cell purity. Isolated neutrophils were centrifuged at 250g for 5 min and resuspended in NET buffer (RPMI 1640 medium supplemented with 1% BSA and 0.1% calcium chloride (Cayman Chemical), and cell viability was assessed using the PrestoBlue Cell Viability Reagent and cell concentration by performing a tenfold dilution of the sample with Trypan Blue, followed by a hemocytometric count.

Generation of Toxoplasma from the RH strain lacking the TgPI-1 gene

The CRISPR/Cas9 system was used to delete the TgPI-1 gene from the RH strain. The protospacer sequence aatctcgaattaaactgcgg was incorporated into the pSAG1::Cas9-U6::sgUPRT construct using the Q5 Site-Directed Mutagenesis Kit (New England Biolabs, Ipswich, MA), resulting in the modified construct pSAG1::Cas9-U6::sgTgPI411. Homologous sequences corresponding to the upstream region (gaatttacggcttcggttagtgctaacccgtcgtaaaa) and downstream region (cgcaagcaaagatgcctttgaaaacaagatcctcta) of the TgPI-1 gene were integrated into the primers DHFR-F1 (cggtttttgtcttttttgagtg) and DHFR-R1 (cgatcccccggtttgcaggaattc), respectively. These custom-designed primers were used to amplify the HR-DHFR fragment from the pUPRT-DHFR plasmid. Subsequently, 700-ng of linearized HR-DHFR and 6 μg of the pSAG1::Cas9-U6::sgTgPI411 plasmid were co-transfected into RHΔKU80 parasites by electroporation. Selection of transfected parasites was initiated 24 h post-transfection using 10 μM pyrimethamine. Genomic DNA was extracted from the subclones, and PCR screening was performed using TgPI-1-specific primers TgPI-1F (atgggaaagaatcctcttttgtttcttgc) and TgPI-1R (ttattggtcatcccagatctcttcggtg), as well as DHFR-related primers 3HDF4 (gagcaaaaggaactgattcgggc). To enhance the precision of screening, the UTR regions upstream (FaUTR5F: gcaccgaagttacgttgttaaacaag) and downstream (FarUTR3R: gcgaagacacctcaaaatcacaagt) of the TgPI-1 gene were also amplified. Successful deletion of the TgPI-1 gene was validated through sequencing, and the absence of TgPI-1 protein was further confirmed by Western blot analysis and IFA using anti-TgPI-1 antibody.

In vitro assays for parasite growth and replication

To monitor the growth of ΔTgPI-1 and parental parasites, standard plaque assays were performed as described (97) using hTERT-immortalized cells grown until confluence in 6-well plates, infected with parasites, and incubated at 37 °C for 7 days in culture medium (unless otherwise indicated). Some assays included 30-min of pretreatment of parasites with 0.1% Trypsin (bovine pancreas), 0.1% α-chymotrypsin (bovine pancreas), and 0.2 U human NE. The plates were then washed twice with PBS and fixed with 4 ml of 95 to 100% ethanol for 5 min at room temperature. The ethanol was removed, and the plates were washed twice with PBS. Subsequently, the samples were stained with crystal violet for 5 min at room temperature, followed by two additional PBS washes. Lysis plaques were enumerated and data graphed in Prism software. To assess parasite number per PV, coverslips with confluent HFF were infected with parasites for 4 h, thoroughly washed with PBS to remove extracellular parasites. Coverslips were fixed with 4% formaldehyde in PBS for 15 min, stained for DAPI and the number of parasites was recorded for at least 150 to 260 PV on the coverslip for each condition by fluorescence microscopy and graphed as a percentage of all PV recorded with standard deviations using Excel (Microsoft). To assess parasite replication rate, the number of ΔTgPI-1 and parental parasites per PV was determined 24 hpi. Confluent HFF monolayers grown on coverslips were infected with parasites and after 24 h, the coverslips were thoroughly washed with PBS to remove extracellular parasites and fixed with 95 to 100% ethanol for 1 min at room temperature, followed by two washes with PBS. The coverslips were stained with Giemsa solution (Sigma, USA) for 10 min. Following staining, the coverslips were washed once with PBS. ProLong Diamond Antifade Mountant (Life Technologies) was used to mount the coverslips onto microscope slides, and the number of parasites per PV was recorded for at least 150 to 260 PVs per coverslip for each experimental condition. The mean parasite counts for each condition were calculated and represented as a percentage of all recorded PVs. Data are presented as mean ± standard deviation, and statistical analysis and graphing were performed using Prism software.

NETosis assays

This section includes methods for examination of NET structures, quantification of NET formation, and quantitative evaluation of parasite viability under NETosis.

Morphological analysis of NETosis

To visualize NETotic events by fluorescence microscopy, coverslips were pre-coated with poly-L-lysine for 3 to 5 h, washed twice with PBS, and covered with 200 μl of NET buffer. Freshly isolated neutrophils (5 × 10⁵ cells) were seeded to each coverslip and incubated at 37 °C for 10 min. Freshly egressed Toxoplasma parasites were added to the coverslip at a multiplicity of infection (MOI) of 1:5 or 100 nM PMA as a positive control. The plate was centrifuged at 500g for 3 min to facilitate cell-parasite contact and incubated at 37 °C for 4 h to allow NET formation. The coverslips were then washed once with PBS and fixed with 1% paraformaldehyde for 12 min at room temperature. When required, samples were permeabilized with 0.2% Triton X-100. The coverslips were then blocked with 3% BSA for 1 h at room temperature. Subsequently, samples were incubated overnight at 4 °C with appropriate primary antibodies, followed by three washes with PBS. Secondary antibodies were applied for 1 h at room temperature, followed by another three PBS washes. DNA materials were stained with DAPI (1:1000 dilution) for 5 min, washed once with PBS, and mounted with ProLong Antifade mounting medium. Images were acquired using a Thunder fluorescence microscope equipped with a K8 scientific CMOS camera and analyzed using LAS X Premium software. In separate assays, dead parasites were used to assess whether NET was released. To kill the parasites while preserving intact their plasma membrane, 10⁷ Toxoplasma cells were collected from culture supernatant, filter-purified, and subjected to heat inactivation at 60 °C for 15 min. The cells were subsequently centrifuged at 500g for 10 min, washed, and resuspended in PBS. The integrity of the cell membrane was assessed using Trypan Blue exclusion. Heat-killed parasites were then added to neutrophils at a MOI of 1:5 for 4 h before staining for DAPI. DAPI fluorescence for DNA on coverslips was measured with ImageJ software.

Quantification of NET formation for extracellular DNA

The NET formation was quantified by measuring extracellular DNA as previously described (32), with slight modifications. Neutrophils were freshly isolated from either the peritoneal cavity of mice or human whole blood. The cells were resuspended at a concentration of 1 × 10⁶ cells/ml in the NET buffer. Isolated neutrophils were seeded at 5 × 10⁵ cells/ml in well in 24-well plates. Parasites or PMA were added to the coverslip as described above. Subsequently, 15 U/ml S7 nuclease was added to each well to digest extracellular DNA, and the plates were incubated for an additional hour at 37 °C. The supernatants were collected into 1.5 ml microcentrifuge tubes, and 10 μl of EDTA (500 mM) was added to each sample. The tubes were centrifuged at 300g for 5 min to remove debris. The supernatants were transferred to new tubes, and 100 μl of each solution was added to a black 96-well plate. Each sample was mixed with 100 μl of Quant-iT PicoGreen dsDNA Reagent. The plate was incubated at room temperature for 2 to 5 min, and fluorescence was measured using a microplate reader at an excitation wavelength of 480 nm and an emission wavelength of 520 nm.

Quantification of NET formation for extracellular elastase

The NET formation was induced as described above. Following a 4-h incubation of neutrophils with either parasites or PMA, each well was washed twice with NET buffer to remove unbound neutrophil elastase. NETs were then digested by adding S7 nuclease, as detailed above. The digestion reaction was terminated by the addition of EDTA, and cellular debris were removed by centrifugation. To measure neutrophil elastase activity, a 100 μl aliquot of the resulting supernatant was transferred to a clear 96-well plate. Each sample was mixed with 1:30 diluted 15 mM of the elastase substrate N-methoxysuccinyl-Ala-Ala-Pro-Val-p-nitroanilide. The plate was incubated for 2 h at 37 °C, after which absorbance at 405 nm was measured using a microplate reader to assess neutrophil elastase activity.

Evaluation of parasite viability under NETosis-inducing conditions

To assess the impact of NETosis on parasite viability, two assays were performed: live/dead cell viability assays and plaque assays following neutrophil exposure. For viability assays, 5 × 10⁵ parasites were exposed to 1 × 10⁵ human neutrophils seeded on pre-coated coverslips and incubated at 37 °C in the presence of 1 μM cytochalasin D (Calbiochem) to inhibit parasite invasion and neutrophil phagocytosis. After 3 h, the medium was removed by aspiration and the coverslips were washed with PBS before incubation with LIVE/DEAD Fixable Orange Viability Dye according to the manufacturer's instructions. The coverslips were imaged 10 fields using an Apo 40 × objective on a Zeiss AxioImager M2 microscope equipped with a Hamamatsu Orca-R2 digital camera. For plaque assays, 1 × 10⁶ neutrophils/well were seeded in a 6-well plate. Freshly egressed Toxoplasma parasites or PMA were added to the neutrophils to trigger NETosis as described above. To inhibit parasite invasion of neutrophils, 100 nM cytochalasin D was included in each well. The plate was incubated at 37 °C for 6 h. Following incubation, the content of each well was collected and centrifuged at 500g for 10 min. The resulting pellet was washed twice with PBS, and the parasites were counted using a hemocytometer. Two different parasite concentrations, 200 or 1000 cells were transferred to hTERT-immortalized cell plates at 37 °C for either four or 7 days. The plates were then washed, fixed and stained as described above. Lysis plaques were enumerated to assess parasite viability and data were graphed in Prism software.

Bioluminescence and fluorescence imaging of mice

Female mice infected with CZ1-GFP-luc (WT and ΔTgPI-1) were imaged for bioluminescence or fluorescence using the in vivo imaging system (IVIS) from Xenogen Corporation using IVISbrite D-Luciferin Potassium Salt Bioluminescent Substrate (MediLumine) or GFP fluorescence as described (98). Briefly, mice were injected with 200 μl of luciferin (stock solution: 15 mg/ml) before anesthesia, then imaged in a dorsal position with 1- to 5-min integration times depending on the intensity of the bioluminescent signal. To assess parasite dissemination throughout the mouse body, mice were sacrificed after imaging, and individual organs were excised from the mice (6, 99), washed in sterile PBS, and maintained in DMEM at room temperature, and imaged ex vivo. Data acquisition and analysis were performed by using the LivingImage (Xenogen) software with the IgorPro image analysis package (WaveMetrics).

Cyst purification for oral gavage

Cysts were isolated from brains of CD-1 IGS mice chronically infected with CZ1-GFP-luc (WT and ΔTgPI-1) strains for at least 6 weeks, using the Percoll gradient method as described (100) and the number of cysts was microscopically quantified. For oral gavage of infectious cysts, C57BL/6 mice were forced-fed with 100 μl of brain homogenate containing 10 to 40 cysts using a ball-tipped feeding needle.

Cyst counting and morphological analysis

Mouse brains were harvested at 45 or 50 days p.i. and homogenized in 2 ml of PBS by syringe passage through a 19-gauge needle. The numbers of cysts in ten aliquots (200 μl each) of the brain suspensions were counted microscopically. For histological analysis of cysts in the brain, brains were removed from mice, embedded in a paraffin wax block, and cut in 5-μm-thick layers using a microtome. Cysts were identified by Hematoxylin and Eosin stain (H&E stain).

Fluorescence microscopy and image analysis

Immunofluorescence assays (IFA) on infected HFF were performed as described previously (101) with a 5-min permeabilization step with 0.3% TritonX-100 in PBS (or no permeabilization) following fixation with 4% formaldehyde (Polysciences, Warrington, PA) plus 0.02% glutaraldehyde in PBS for 15 min. For cyst wall labeling, fixed infected HFF with differentiating bradyzoites were incubated with TRITC-lectin (1:250 dilution in PBS containing 3% BSA) performed after the secondary incubation step for 30 min, followed by three 5-min PBS washes before staining with DAPI. Coverslips were mounted using ProLong Diamond Antifade Mountant (Life Technologies) to minimize bleaching during microscopy. For IFA on extracellular T. gondii, coverslips pre-coated with poly-L-lysine and freshly egressed parasites were collected, centrifuged at 500g for 10 min, and the resulting pellet was washed twice with PBS. The parasites were resuspended in PBS, and 1 × 10⁶ cells were transferred to the coated coverslips. The coverslips were incubated at 37 °C for 10 min to facilitate attachment, followed by centrifugation at 500g for 5 min to ensure adherence of the cells to the coverslip surface. The PBS was gently removed from each coverslip and replaced with 1% formaldehyde for fixation. Subsequent immunofluorescence staining was performed as described above. For NET structure observations, isolated neutrophils were incubated on poly-L-lysine-treated glass coverslips in a 24-well plate with medium, PMA, or Toxoplasma and centrifuged for 5 min at 200g to initiate infections, followed by incubation at 37 °C for 4 h. Samples were collected by gently removing coverslips, fixing them with 3% paraformaldehyde (20 min at room temperature), and were then blocked in BSA for 1 h at room temperature. Coverslips were incubated with the indicated antibodies and then washed and mounted with ProLong Antifade containing DAPI (Molecular Probes). Slides and coverslips were viewed using an oil immersion plan Apo 100x (NA1.4) objective, a Zeiss AxioImager M2, and a Hamamatsu Orca-R2 digital camera.

Excreted/secreted antigen (ESA) collection

Large-scale preparation of ESA proteins was performed by incubating 3 × 10⁹ RH strain tachyzoites in 10 of DMEM containing 2 mM glutamine, 10 mM HEPES, and 1% ethanol at 37 °C for 20 min followed by cooling on ice for 5 min. Parasites were removed by centrifugation (1000g, 10 min, 4 °C), and the supernatant was concentrated to ∼0.5 ml using C-20 concentrators according to the manufacturer’s instructions.

Western blotting

For immunodetection of TgPI-1 in Toxoplasma cells, cultured tachyzoites or mouse cysts were lysed by suspension in SDS gel loading buffer (50 mM Tris-HCl (pH 6.8), 50 mM 2-mercaptoethanol, 2% SDS, 0.1% bromophenol blue, 10% glycerol) followed by boiling in a water bath. For immunodetection of TgPI-1 in the Toxoplasma secretome, supernatants containing the ESA proteins were filtered through 0.22 μm Millipore membrane, concentrated 3 times by centrifugation at 3000g for 60 min in Centricon 10 tubes (Amicon) with 100 U/ml of aprotinin added to the ESA preparation. The samples (parasite lysates and ESA) were subjected to SDS-PAGE, and the proteins were then electrophoretically transferred to a membrane (Immobilon Transfer Membranes, Millipore, Bedford, MA). The membrane was immersed in a blocking buffer (PBS containing 3% skim milk) for 60 min, and then incubated with rabbit anti-TgPI-1 (1:5000) or mouse anti-a-Tubulin (1:2000) in the blocking buffer for 60 min. Unbound antibody was removed by washing the membrane six times with a blocking buffer. Next, the membrane was incubated with horseradish peroxidase-conjugated goat anti-mouse IgG antibody (Amersham Pharmacia Biotech; dilution, 1:10,000) in blocking buffer for an additional hour, before detection by chemiluminescence using ECL-Plus.

Electron microscopy

For ultrastructural observations of T. gondii-infected cells by thin-section transmission electron microscopy, infected cells were fixed in 2.5% glutaraldehyde (Electron Microscopy Sciences) in 0.1 mM sodium cacodylate (pH 7.2) for 1 h at room temperature and processed as described (94). For immunostaining of TgPI-1, extracellular and intracellular Toxoplasma in HFF for 24 h cells were fixed in 4% paraformaldehyde (Electron Microscopy Sciences) in 0.25 M HEPES (pH 7.4) for 1 h at room temperature, then in 8% paraformaldehyde in the same buffer overnight at 4 °C. They were infiltrated, frozen, and sectioned as previously described (102). The sections were immunolabeled with rat anti-TgPI-1 antibody (1:50 in PBS/1% fish skin gelatin), then with mouse anti-rat IgG antibodies, followed by 10 nm protein A-gold particles. Ultrathin sections and cryosections of infected cells were examined with a Hitachi 7600 EM under 80 kV.

Statistical analysis

All statistical analyses were performed using GraphPad Prism. Unless otherwise noted, all error bars are presented as the SD and from a minimum of three independent trials. Differences were considered significant if p values were < 0.05. ANOVA was used to compare larger than two sets of data. Comparison between two sets was done using paired or unpaired two-tailed t-tests. Statistical tests performed with corresponding statistical significance are specified in each figure legend.

Data availability

All the data described in the manuscript are contained within the manuscript and shown on Figures, except for Figure 3 for which the raw data are deposited as a separate file named “Supporting information (Not for publication)” with this manuscript.

Supporting information

This article contains supporting information.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.

Reviewed by members of the JBC Editorial Board. Edited by Clare E. Bryant