Besnoitia besnoiti bradyzoite stages induce suicidal- and rapid vital-NETosis

Besnoitia besnoiti is an obligate intracellular apicomplexan protozoan parasite, which causes bovine besnoitiosis. Recently increased emergence within Europe was responsible for significant economic losses in the cattle industry due to the significant reduction of productivity. However, still limited knowledge exists on interactions between B. besnoiti and host innate immune system. Here, B. besnoiti bradyzoites were successfully isolated from tissue cysts located in skin biopsies of a naturally infected animal, and we aimed to investigate for the first time reactions of polymorphonuclear neutrophils (PMN) exposed to these vital bradyzoites. Freshly isolated bovine PMN were confronted to B. besnoiti bradyzoites. Scanning electron microscopy (s.e.m.)- and immunofluorescence microscopy-analyses demonstrated fine extracellular networks released by exposed bovine PMN resembling suicidal NETosis. Classical NETosis components were confirmed via co-localization of extracellular DNA decorated with histone 3 (H3) and neutrophil elastase (NE). Live cell imaging by 3D holotomographic microscopy (Nanolive®) unveiled rapid vital NETosis against this parasite. A significant increase of autophagosomes visualized by specific-LC3B antibodies and confocal microscopy was observed in B. besnoiti-stimulated bovine PMN when compared to non-stimulated group. As such, a significant positive correlation (r = 0.37; P = 0.042) was found between B. besnoiti-triggered suicidal NETosis and autophagy. These findings suggest that vital- as well as suicidal-NETosis might play a role in early innate host defence mechanisms against released B. besnoiti bradyzoites from tissue cysts, and possibly hampering further parasitic replication. Our data generate first hints on autophagy being associated with B. besnoiti bradyzoite-induced suicidal NETosis and highlighting for first time occurrence of parasite-mediated vital NETosis.

Typically, bovine besnoitiosis is characterized by an acute and a chronic phase with different clinical signs. In the acute phase, B. besnoiti-infected cattle present pyrexia, intensive respiratory disorders, increased heart rates, subcutaneous oedema, anasarca, swollen joints, conjunctivitis, nasal discharge, photophobia, reduced milk yield and orchitis associated with permanent infertility in bulls (Bigalke, 1981;Álvarez-García et al., 2013;Cortes et al., 2014). During the chronic phase of disease, B. besnoiti bradyzoites proliferate slowly within the epidermis, subcutaneous tissues, mucous membranes and/or sclera, and form characteristic cysts within mesenchymal host cells, related to dramatic thickening, hardening, folding, wrinkling of the skin (also termed 'elephant skin'), alopecia, and gradual deterioration of body condition and weight loss (Pols, 1960). Until now, the complete life cycle of B. besnoiti is not entirely known and final host species are unidentified carnivores. Nevertheless, direct contact between infected and non-infected animals (e.g. natural mating, naso-pharyngeal route) and insect-mediated transmission through biting flies (i. e. tabanids (Tabanus spp.), stable flies (Stomoxys calcitrans)) have been suggested as suitable transmission routes Gutiérrez-Expósito et al., 2017;Tainchum et al., 2018) and of epidemiological relevance (Sharif et al., 2019).
So far, very limited knowledge exists on early interactions between circulating polymorphonuclear neutrophils (PMN) of host innate immune system with B. besnoiti, although these cells are the first ones to be recruited to infection sites. As such, PMN are the most abundant granulocytes in the blood and being the first line of defence against invading pathogens including parasites (Weissmann et al., 1980;Behrendt et al., 2010;Villagra-Blanco et al., 2017a). Upon activation, and in addition to phagocytosis (Behrendt et al., 2008) and degranulation (Lacy, 2006), PMN also combat efficiently invading pathogens by releasing neutrophil extracellular traps (NETs) (Fuchs et al., 2007;Brinkmann and Zychlinsky, 2012;Brinkmann, 2018). These NETs are composed of nuclear DNA decorated with different histones (H1, H2A/H2B, H3, H4) and various antimicrobial granular effector molecules and commonly released via a novel cell death process known as suicidal NETosis. Suicidal NETosis is characterized by nuclear and cell membrane rupture and the loss of main PMN functions such as chemotaxis, degranulation and phagocytosis (Fuchs et al., 2007;Remijsen et al., 2011b;Yipp and Kubes, 2013). In contrast, vital NETosis can also occur by not affecting the continuation of mentioned PMN functions (Yipp and Kubes, 2013). Vital NETosis have been demonstrated in response to bacteria (Pilsczek et al., 2010), fungi (Byrd et al., 2013), LPS-activated platelets (Clark et al., 2007) and even the protozoan parasite Leishmania amazonensis (Rochael et al., 2015). A landmark of vital NETosis is its rapid induction, normally within 30 min after PMN stimulation (Yipp and Kubes, 2013) or as early as 10 min after neutrophil-L. amazonensis interaction (Rochael et al., 2015). In previous studies, it was shown that suicidal NETosis was able to efficiently trap B. besnoiti tachyzoites in vitro and that released suicidal NETosis was capable of hampering tachyzoites from active host cell invasion (Muñoz Caro et al., 2014a). Furthermore, also bovine monocyte-derived extraellular traps (METosis) occurred when these phagocytes have been exposed to vital and motile B. besnoiti tachyzoites (Muñoz Caro et al., 2014a), thereby expanding the spectrum of leukocytes undergoing ETosis (Villagra-Blanco et al., 2019).
Conversely, no data are available so far neither on interactions of bovine PMN with B. besnoiti bradyzoites nor the role of autophagy in parasite-induced NETosis. Autophagy has recently been indicated to play a crucial role not only influencing classical PMN-mediated effector mechanisms (e.g. phagocytosis) (Mitroulis et al., 2010;Skendros et al., 2018) but also actively regulating NETosis (Skendros et al., 2018). Thus, in the present study, we intended firstly to investigate rapid vital-as well as suicidal-NETosis in bovine PMN exposed to freshly isolated bradyzoites of B. besnoiti from subdermal tissue cysts and further to analyse the possible correlation of autophagy in B. besnoiti bradyzoite-mediated NETosis.

Animal data
In early 2018, a 4-year-old Limousine heifer (502 kg BW) from South France presented inappetence, limb oedema with desquamation and emaciation. Natural B. besnoiti infection was confirmed by polymerase chain reaction investigation. For animal treatment, flunixine meglumine (2.2 mg/kg; Finadyne®) and sulfamethoxine (40 mg/kg; Sulfaron®) were given. Three months later, the same animal was admitted to the Ecole Nationale Veterinaire Toulouse (ENVT) because of weakness, hyperkeratosis, and multifocal alopecia, and presence of multiple visible cysts within the sclera.
One week later, the animal was euthanized due to severe emaciation. At necropsy, bovine besnoitiosis in the scleroderma phase was confirmed: multiple whitish punctuated cysts were observed in sclera and in mucocutaneous junctions of mouth and anus. Moreover, the skin of neck, shoulders, base of tail, hocks, and pasterns presented marked hyperkeratosis with crusty appearance. No other relevant clinical alterations were further noticed. Skin samples of affected areas have been collected, stored in sterile saline solution (4°C) and later on processed for histopathological evaluation as well as for parasite isolation.

Isolation of vital Besnoitia besnoiti bradyzoites
Skin biopsies were placed in a sterile Petri dish (Nunc) containing a small volume of sterile RPMI 1640 cell culture medium without phenol red (Sigma-Aldrich) supplemented with 1% penicillinstreptomycin (Sigma-Aldrich). A sterile tweezer was used to hold the skin and with a sterile scalpel, the skin surface was carefully scraped in order to release vital bradyzoites from these cysts. As soon as the RPMI 1640 cell culture medium became turbid, it was collected and filtered through a sterile gauze swab in a sieve into a 50-mL Falcon tube followed by centrifugation at 200 × g for 1 min at room temperature (RT). Then the supernatant was collected and transferred into a new 50-mL Falcon tube, centrifuged (400 × g, 12 min), and the pellet was washed again with RPMI 1640 medium to collect released bradyzoites. All supernatants were collected and centrifuged at 1500 × g for 10 min, and then the supernatant was discarded and the pellet containing bradyzoites was resuspended in sterile RPMI 1640 cell medium. Vital and extremely motile B. besnoiti bradyzoites were isolated and afterwards counted in a Neubauer haemocytometer chamber (Supplementary data video 1). Isolated B. besnoiti bradyzoites were firstly stored for 30 min at 4°C and afterwards at −80°C in RPMI 1640 cell medium supplemented with 10% DMSO (Merck).

Histopathological examination
After successful bradyzoites isolation, parts of skin samples (5 × 5 mm 2 ) were stored in 10% phosphate-buffered formalin for histopathological examinations. Shortly, formalin-fixed samples were dehydrated using an ascending ethanol series, embedded in paraffin wax at 56°C and finally sectioned at 3 μm tissue samples at the Institute of Veterinary Pathology, Faculty of Veterinary Medicine, Justus Liebig University Giessen, Germany. Histological tissue samples have been stained using haematoxylin and eosin (HE), periodic acid-Schiff (PAS) and Giemsa staining according to routine protocols and pathological findings/changes of B. besnoiti-infected skin samples were then evaluated under a light microscope (Nikon Eclipse 80i) equipped with a DS-Fi1 digital camera (Nikon).

Isolation of bovine PMN
Healthy adult dairy cows (n = 3) were served as blood donors. Animals were bled by puncture of the jugular vein and 30 ml peripheral blood was collected in 12 ml heparinized sterile plastic tubes (Kabe Labortechnik). Approximately 20 ml of heparinized blood was re-suspended in 20 ml sterile PBS with 0.02% EDTA (Sigma-Aldrich), slowly layered on top of 12 ml Biocoll® separating solution (density = 1.077 g/L; Biochrom AG), and centrifuged (800 × g, 45 min). After extraction of plasma and peripheral mononuclear blood cells (PBMC), the pellet was washed in 25 ml distilled water and gently shaken during 40 s in order to lyse erythrocytes. Osmolarity was rapidly restored by Hank's balanced salt solution (4 ml, HBSS 10×; Biochrom AG). To complete erythrocyte lysis, this step was repeated twice and PMN were later re-suspended in sterile RPMI 1640 medium (Gibco). Finally, freshly isolated bovine
PMN were allowed to rest at 37°C and 5% CO 2 atmosphere for 30 min until further use (Behrendt et al., 2010). For each experiment, purity and viability of neutrophils were determined. Only samples with a purity of neutrophils higher than 93% and viability greater than 96% (tested by trypan blue exclusion assay (Sigma-Aldrich)) were used.

Scanning electron microscopy (S.E.M.) analysis
Bovine PMN were co-cultured with vital B. besnoiti bradyzoites (ratio 1:4) for 3 h on coverslips (10 mm diameter; Nunc) precoated with 0.01% poly-L -lysine (Sigma-Aldrich) in an incubator at 37°C and 5% CO 2 atmosphere. After incubation, cells were fixed in 2.5% glutaraldehyde (Merck), post-fixed in 1% osmium tetroxide (Merck), washed in distilled water, dehydrated, critical point dried by CO 2 -treatment and sputtered with gold. Finally, all samples were visualized via a Philips® XL30 scanning electron microscope at the Institute of Anatomy and Cell Biology, Justus Liebig University Giessen, Germany.

Live cell interactions between bovine PMN and B. besnoiti bradyzoites investigated by live cell 3d holotomographic microscopy
Isolated PMN (1 × 10 6 ) were centrifuged at 300 × g for 10 min at RT, the supernatant was carefully discarded and cells were suspended in 2 ml of pre-warmed RPMI 1640 cell medium. One ml of this PMN solution was placed in an Ibidi® cell plate 35 mm low profile, and the plate was incubated in an Ibidi® chamber at 5% CO 2 and 37°C. PMN were allowed to settle down (30 min) to bottom of the plate and then 2 × 10 6 B. besnoiti bradyzoites were added to the center of the plate. The acquisition was set for refractive index (RI; 3D tomography) for a time-lapse of 155 min every 30 s in a Nanolive Fluo-3D Cell Explorer® (Nanolive) microscope. At the end of the experiment, images were exported using Steve software v.2.6® (Nanolive). Using Image J software (Fiji version 1.7, NIH), every frame was exported using z-projection, maximum intensity algorithm and the video was constructed using 1 frame per 5 s of speed. For zoomed video, the region of interest was cropped and the same procedure described above was applied.
Digital staining and 3D rendering of activated PMN was performed by using Steve software v.2.6® (Nanolive).

Autophagosome detection by immunofluorescence analysis
LC3 has been used as a classical marker for autophagosomes (Karim et al., 2007), being LC3-I cytosolic and LC3-II membrane bound and enriched in the autophagic vacuole. Therefore, we tested whether LC3 expression might be present during B. besnoiti bradyzoite-induced NETosis as described elsewhere (Zhou et al., 2019). Briefly, bovine PMN (n = 3) were added on 0.01% poly-L -lysine pre-coated coverslips (15 mm diameter; Nunc), then stimulated by B. besnoiti bradyzoites for 1 h at RT. After incubation, cells were fixed with 4% paraformaldehyde for 10 min, permeabilized with cold methanol (Merck) for 3 min and blocked by using the following blocking buffer (5% BSA (Sigma-Aldtich), 0.1% Triton X-100 (Sigma-Aldrich) in sterile PBS) for 60 min at RT. After removing blocking buffer, cells were incubated overnight at 4°C with rabbit anti-LC3B antibodies (Cat#2775, 1:200, Cell Signaling Technology) diluted in blocking buffer, washed three times with PBS, incubated with goat antirabbit IgG conjugated with Alexa Fluor 488 (Invitrogen) for 1 h in the dark at RT. After being washed three times with PBS, coverslips were mounted by prolonged anti-fading reagent with DAPI (Invitrogen) on glass slides (Nunc), and five randomly images were taken per condition using an inverted epifluorescence microscope IX 81® (Olympus) and/or by using confocal microscopy analysis (LSM 710®; Zeiss).

Statistical analysis
Results are illustrated as means ± S.E.M. of at least three independent experimental settings. One-way analysis of variance and Dunnett's multiple comparison test and Spearman correlation test were performed here by using GraphPad Prism 7®. Differences were considered significant at a level of P ⩽ 0.05.

Histopathological examination of B. besnoiti-infected skin
Histopathological examination revealed multifocal large-sized round to ovoid B. besnoiti-cysts present in the dermis, panniculus and underlying muscle layer (Fig. 1A, highlighted exemplary by black arrows). Rare early tissue cysts were small, approximately 10-20 μm and contained a parasitophorous vacuole (PV) with few banana-shaped 3-5 μm structures (bradyzoites). Mature B. besnoti-cysts (up to 400 μm) were filled with thousands of bradyzoite stages (Fig. 1A-C, asterisks). Cysts containing numerous typical banana-shaped B. besnoiti bradyzoites had a three-layered wall ( Fig. 1B and C, thickness of 10-30 μm thick): the outer wall composed of compressed collagen type I fibres ( Fig. 1B and C, black arrows), the middle layer representing a thick hyaline capsule composed of extracellular matrix ( Fig. 1B and C, clear arrows) and the inner layer composed of a small rim of host cell cytoplasm with often multiple flattened nuclei containing the PV (Fig. 1B and C, arrowheads). The outer wall of compressed collagen, as well as the bradyzoites, stained mildly and the middle hyaline layer were brightly stained with PAS (Fig. 1A). Using Giemsa stain, the inner rim of the middle hyaline layer of mature cysts stained purple, while the outer rim of the hyaline layer was translucent (Fig. 1C, clear arrow) and the inner layer containing the host cell cytoplasm as well as the bradyzoites themselves stained blue (Fig. 1C, arrowhead and asterisk). Surrounding the tissue cysts there was a mild to moderate multifocal to coalescing infiltrate (Fig. 1D) composed of macrophages (Fig. 1D, black arrow), fewer lymphocytes, plasma cells (Fig. 1D, black arrowhead), neutrophils (Fig. 1D, red arrowhead) and eosinophils (Fig. 1D, clear arrow) and rare multinucleated giant cells (Fig. 1D, asterisk). Few tissue cysts were found ruptured and surrounded by abundant macrophages including multinucleated giant cells, numerous neutrophils and eosinophils, which were sometimes arranged in clusters as well as fewer lymphocytes and plasma cells. Also, throughout the affected skin, there was mild to moderate diffuse epithelial hyperplasia and moderate orthokeratotic hyperkeratosis.

Besnoitia besnoiti bradyzoite-triggered NETosis was unveiled via S.E.M.-and immunofluorescence microscopy-analysis
To investigate whether B. besnoiti bradyzoites were capable to induce NETosis, bovine PMN exposed to bradyzoites were analysed by S.E.M. Fine network structures were observed in B. besnoiti bradyzoites-stimulated bovine PMN (Fig. 2), and many bradyzoites were trapped by those structures (Fig. 2) visualized in S.E.M. analysis. Alongside, different morphologies of bovine PMN were observed around these fine networks. Whilst typical smooth rounded PMN have been found in close proximity to bradyzoites indicating a rather inactivation status, other exposed PMN showed disrupted cell membrane surfaces and thereby releasing extracellular filaments entrapping firmly bradyzoites by cell death. Former PMN status corresponded well to previously described suicidal (lytic) ETosis against this apicomplexan protozoa where extracellular fibres mainly derived from dead PMN and monocytes (Muñoz-Caro et al., 2014a,b).
In order to confirm whether bovine PMN were undergoing suicidal NETosis, main components of NET formation (i.e. DNA, histones (H3) and NE) were visualized via immunostaining. In the control group (non-exposed PMN), no NETosis-like structures were observed by co-localization of H3 and NE (Fig. 3A). In contrast, the classical characteristics of suicidal NETosis were demonstrated in bovine PMN exposed to B. besnoiti bradyzoites by co-localization of extracellular DNA adorned with H3 and NE (Fig. 3B), and several bradyzoites being firmly trapped by NETosis as indicated by white arrows in Fig. 3B.

3D-holotomographic microscopy live cell imaging of B. besnoiti bradyzoite-triggered vital NETosis
Activation of bovine PMN and NETosis were additionally analysed by live cell 3D-holotomographic microscopy technology (Nanolive®). Activation of PMN occurred within the first 5 to 30 min of interaction with motile bradyzoites thereby showing pseudopod formation and rapid migration and crawling activities of PMN into the vision field showing bradyzoites. Noteworthy to mention was the observation of an elongated structure being

404
Ershun Zhou et al.  rapidly tossed out from PMN after 30 min of parasite interaction. Due to the time point of occurrence and the non-lytic PMN phenotype of this 'chameleon tongue-like' reactions we interpreted this response as vital NETosis (Fig. 4A-B; please refer to Video S2). The digital staining and 3D reconstruction of vital NETosis showed clearly that neither the overall cell phenotype nor crawling activities were altered by the protrusion of this elongated structure (Video S3; Fig. 4C).

Autophagy occurred during B. besnoiti-triggered suicidal NETosis
Autophagy is a highly conserved intracellular degradation process not only to keep homeostasis or energy source of mammalian cells but also pivotal in several host innate immune functions (Germic et al., 2019). During autophagy, LC3 (microtubule-associated protein 1A/1B-light chain 3) is an important protein being involved in autophagosome formation, and it has been used as a classical marker of autophagosomes. As shown in Fig. 5, the majority of bovine PMN (see Fig. 5A) were still round and inactive without stimulation of B. besnoitia bradyzoites. In contrast, most of bovine PMN exposed to B. besnoitia bradyzoites (see Fig. 5B) were undergoing autophagy alongside with NETosis resulting in bradyzoite entrapment (Fig. 5B), indicating an association of these two cellular processes.
To investigate in more detail concomitant autophagy while B. besnoiti-induced NETosis, the percentages of 'NETotic cells' and LC3B-positive cells were calculated, and thereafter a Spearman correlation test was performed. As seen in Fig. 6A, more cells were undergoing suicidal NETosis in B. besnoiti bradyzoite-stimulated bovine PMN when compared to non-stimulated PMN (not statistically significant). Moreover, a significant low positive correlation (r = 0.3735, P = 0.042) was found between suicidal NETosis and autophagy in B. besnoiti bradyzoite-stimulated bovine PMN (Fig. 6B) compared to negative controls.

Discussion
Bovine besnoitiosis is caused by the parasite B. besnoiti which is a cyst forming apicomplexan parasite closely related to Toxoplasma gondii and Neospora caninum (Ellis et al., 2000). Chronic bovine besnoitiosis is characterized by tissue cyst formation, especially in the skin and in the mucosa of diverse organs (e.g. eyes, genitals) with associated progressive thickening, folding, hardening, wrinkling lesions of affected skin or mucosa. Characteristic large-sized cysts of B. besnoiti containing thousands of bradyzoites were identified in the skin samples of this naturally infected heifer by histopathological examinations in accordance to previous reports (Cortes et al., 2005;Jacquiet et al., 2010;Rostaher Ana et al., 2010;Gentile et al., 2012;Frey et al., 2013). The affected animal did not benefit from implemented treatments and its deteriorated clinical status determined euthanasia. However, even if the clinical status of this animal had remained unaltered, culling would have been the better control measurement for cattle besnoitiosis in the farm.
NETosis is an effective defence process of activated PMN to ensnare and eliminate invading pathogens by releasing web-like extracellular traps which consist of DNA as a backbone, histones (H1, H2A/H2B, H3, H4), and diverse anti-microbial peptides/ proteases such as cathepsin G, α-defensin, pentraxin, cathelicidin (LL37), lactoferrin, calprotectin and others (Amulic and Hayes, 2011;Hermosilla et al., 2014). Recently, more attention has been paid on pivotal role of NETosis against protozoan-and metazoan-parasites in various terrestrial and marine mammalian species (Behrendt et al., 2010;Silva et al., 2014;  Consequently, PMN-derived NETosis and monocyte-derived METosis exerted potent entrapment capacities against B. besnoiti tachyzoites indicating that these two leukocyte populations might reduce parasite replication during the acute phase of infection as previously postulated. Nevertheless, no data in the literature are still available on early NETosis against B. besnoiti bradyzoites although bradyzoites are released from tissue cysts in vivo . Here, for the first time, we demonstrated that bradyzoites of B. besnoiti were also able to induce NETosis in a similar manner as tachyzoites, and thus proving that B. besnoiti-triggered NETosis is a rather parasite stage-independent effector mechanism. Bradyzoites and tachyzoites of B. besnoiti are known to exhibit different antigens (Fernandez-Garcia et al., 2009;Schares et al., 2010), and whether bradyzoite-specific antigens induced bovine NETosis could not be answered here Muñoz-Caro et al., 2015, 2018. In line with these findings, other reports have also shown that different apicomplexan parasitic stages of the same species are able to induce NETosis Muñoz-Caro et al., 2015;Villagra-Blanco et al., 2017b). S.E.M. analysis unveiled the presence of classical web-like structures released by bovine PMN exposed to B. besnoiti bradyzoites, as previously observed for tachyzoites (Muñoz Caro et al., 2014a;Maksimov et al., 2016). Additionally, the main components of NETosis (i.e. DNA, histones (H3), NE) were identified here and visualized via Fig. 5. Autophagy occurs in Besnoitia besnoiti-triggered suicidal NETosis. Bovine PMN (n = 3) were exposed to B. besnoiti bradyzoites on coverslips for 1 h at 37°C, 5% CO 2 . Samples were fixed and thereafter permeabilized for LC3B-based immunostaining in order to determine autophagosome formation by confocal microscopy analysis. Bradyzoites were indicated by arrows in merged images. Scale bar = 10 μm.

406
Ershun Zhou et al. immunostaining and proving that these web-like structures were mainly suicidal NETosis. In addition, live cell imaging by 3D holotomographic microscopy showed rapid vital NETosis within the first 30 min without compromising the general structure of PMN cell membrane as well as crawling activity. In the past, it has been proposed that PMN subpopulations are able to elicit different types of NETosis and that only 20-30% undergo suicidal NETosis (Fuchs et al., 2007;Yipp and Kubes, 2013). Interestingly, suicidal NETosis seems to be more related to chemical stimuli as PMA, requiring hours to occur, meanwhile, vital NETosis might be more related to biological triggering agents of NETosis such as bacteria, fungi (Yipp and Kubes, 2013) or parasites (Rochael et al., 2015). Currently, it is not clear how vital NETosis could contribute to the control of infection, but it has been suggested that this process is the first effector mechanism against invasive pathogens to occur (Yipp and Kubes, 2013) and afterwards followed by suicidal NETosis, which results in stronger DNA release. Additionally, it has been speculated that this phenomenon could be initialized by TRL mediated responses (de Buhr and von Köckritz-Blickwede, 2016), but further investigations are necessary to better understand such an intriguing feature in bovine besnoitiosis. To our knowledge, this is the first time that vital NETosis is evidenced by 3D live cell imaging as a PMN response to motile parasite stages. As stated above, bradyzoites can be released from tissue cysts either after host induced-or after mechanical rupture (Schulz, 1960;Langenmayer et al., 2015). Further, Langenmayer et al. (2015) suggested that during chronic bovine besnoitiosis intravascular circulation of 'zoites' might be possible after mechanical rupture of cysts located directly underneath vascular endothelium or after reactivation of tissue cysts and stage conversion into tachyzoite stages. Irrespective of these in vivo scenarios, released bradyzoites would be immediately in close contact to PMN and extruded NETosis might ultimately hamper bradyzoite dissemination. Released bradyzoites might not be immediately identified in the bloodstream of infected animals due to pro-inflammatory host innate immune reactions as proposed elsewhere  and in vivo PMN are among the first ones to be recruited to inflammation/infection sites (Fuchs et al., 2007;Villagra-Blanco et al., 2019;Zhou et al., 2019).
Our present results also demonstrated that autophagy was associated with bradyzoite-triggered NETosis. These findings corresponded well to recent data on B. besnoiti tachyzoite-mediated suicidal NETosis with concomitant autophagy (Zhou et al., 2019). While autophagy process has recently been tightly associated with major neutrophil functions, including degranulation, reactive oxygen species production, and release of neutrophil extracellular traps, the exact molecular mechanisms and autophagy pathways are still not completely clear (Remijsen et al., 2011a;Ullah et al., 2017;Skendros et al., 2018). Also, autophagy is an essential intracellular degradation mechanism to regulate protein and organelle turnover in many living cells thereby maintaining homeostasis and intracellular energy balance (Levine and Kroemer, 2008). During the process of autophagy, intracellular autophagosomes ultimately fuse with lysosomes to degrade and recycle the inside cargo (Bernard and Klionsky, 2013). LC3 is a small soluble protein, which is distributed ubiquitously in mammalian tissues and known to form stable associations with the membrane of autophagosomes (Tanida et al., 2008). Thus, LC3 is widely used as a classical marker for microscopical detection of autophagosomes (Koukourakis et al., 2015;Park et al., 2017). Previous studies have revealed that autophagy is required for NETosis (Remijsen et al., 2011a;Ullah et al., 2017), and that autophagy induction significantly increased NETosis (Park et al., 2017). Accordingly, LC3B-stained autophagosomes were detected concomitantly in PMN extruding suicidal NETosis towards B. besnoiti bradyzoites. These findings confirm that autophagy is required in bovine NETosis not only against B. besnoiti tachyzoite- (Zhou et al., 2019) but also against bradyzoite-stages as Spearman test revealed a significant low positive correlation between these two processes. In agreement to our findings, autophagy has also been reported to prime PMN not only for increased NETosis but also for increased phagocytosis during sepsis (Park et al., 2017). PMN-derived phagocytosis through autophagy is expected to occur in the chronic phase of cattle besnoitosis in vivo, however this process needs further investigations.
In summary, we describe for the first time the ability of bovine PMN to cast NETosis against motile B. besnoiti bradyzoites evidencing the importance of this ancient and well-conserved effector mechanism of early host innate immune system in cattle. Furthermore, LC3B-stained autophagosomes were detected in B. besnoiti bradyzoite-exposed PMN casting NETs resulting in a significant low positive correlation of autophagy and parasite-induced suicidal NETosis. Nevertheless, further autophagy-related investigations should elucidate other molecular mechanisms in this cell pathway, such as potential membrane changes of mitochondria and oxygen consumption rates, as well as the role of AMPK in autophagy (Zhou et al., 2019). Finally, B. besnoiti-mediated vital NETosis resulted in a rapid extrusion and retraction of a 'chameleon tongue-like' structure, which is the first hint for this type of NETosis against these apicomplexan parasites. Exact machinery, B. besnoitispecific antigens and PMN receptors leading to fast parasitetriggered vital NETosis need further investigations. Fig. 6. Autophagy has a significant correlation with Besnoitia besnoiti bradyzoite-triggered suicidal NETosis. Five images were randomly taken from each sample, the number of NETotic (A) and LC3B-positive PMN were counted using ImageJ and the percentages over total cells was calculated. Positive correlation between B. besnoiti-induced LC3B expression and NETotic cells was analysed by Spearman test (B). Results are represented as a before-after graph with data derived from three different animals (n = 3). P values of <0.05 were considered as statistically significant.
Supplementary material. The supplementary material for this article can be found at https://doi.org/10.1017/S0031182019001707