The experiments were conducted following the Justus Liebig University Giessen Animal Care Committee Guidelines. Protocols were approved by Ethics Commission for Experimental Animal Studies of Federal State of Hesse (Regierungspräsidium Giessen; A2/2016; JLU-No. 589_AZ and G16/2017, JLU-No. 835_GP) and in accordance to European Animal Welfare Legislation: ART13TFEU and current applicable German Animal Protection Laws.

Toxoplasma gondii (RH strain) tachyzoites were maintained by serial passages in MARC-145 (Meat Animal Research Center-145) layers. The infection rate in MARC was 40-50% in all experiments. Therefore, free-released T. gondii tachyzoites were harvested from MARC supernatants, pelleted (400 × g, 12 min), counted in a Neubauer chamber, suspended in HBSS and used for bovine PMN confrontation. All experiments were performed at an MOI of 1:2 (bPMN:parasites).

Healthy adult dairy cows (n = 4) served as blood donors. Animals were bled by puncture of the jugular vein and 30 ml blood were collected in heparinized sterile plastic tubes (Kabe Labortechnik). Twenty ml of heparinized blood were diluted in 20 ml sterile PBS with 0.02% EDTA (SigmaAldrich), layered on top of 12 ml Biocoll separating solution (density = 1.077 g/l; Biochrom AG) and centrifuged (800 × g, 45 min). After removal of plasma and PBMC, the cell pellet was suspended in 5 ml of Hank´s balanced salt solution (HBSS) and gently mixed with 1 volume of ice-cold phosphate-based lysis buffer (5.5 mM NaH₂PO₄, 8.4 mM HK₂PO₄, pH 7.2) for 1 min to lyse erythrocytes. Osmolarity was rapidly restored by adding 2 volumes of hypertonic buffer (5.5 mM NaH₂PO₄, 8.4 mM HK₂PO₄, 0.46 M NaCl, pH 7.2) and completing to 50 ml with HBSS. For full erythrocyte lysis, this step was repeated twice and PMN were washed twice in sterile HBSS. All centrifugation steps were performed using 600 × g for 6 min at 4 °C as previously described (23, 24). PMN were counted in a Neubauer haemocytometer. Finally, freshly isolated bovine PMN were allowed to rest (37 °C, 5% CO₂ atmosphere) for 30 min before the experiments were performed (25).

Non-stimulated bovine PMN (negative control) and PMN confronted with T. gondii tachyzoites (1:2) were fixed at 15, 30 or 60 min of co-incubation depending on the experiment with paraformaldehyde 4% for 15 min at room temperature (RT). As negative control HBSS was used. As a positive control of NET formation, the calcium ionophore A23187 was used at a concentration of 25 µ M. The samples were carefully washed thrice with sterile PBS and incubated in blocking/permeabilization solution (PBS containing 3% BSA, 0.3% Triton X-100; Sigma-Aldrich) for 1 h at RT. Thereafter, the samples were incubated with the primary antibodies indicated in the Table 1 diluted in blocking/permeabilization solution overnight at 4 °C in a humidified chamber. Thereafter, samples were washed thrice in sterile PBS (Sigma-Aldrich) and incubated in secondary antibody solutions ( Table 1 ) for 30 min at RT, protected from light. Nuclear counterstaining was achieved by 4′,6-diamidin-2-phenylindol (DAPI; Sigma-Aldrich) present in mounting medium (Fluoromount G, ThermoFisher).

Images were acquired with a Zeiss Confocal LSM 710 equipped with a motorized XY stage and oil 63× objective (numerical aperture of 1.4) or in a Nikon Eclipse Ti2-A inverted microscope equipped with ReScan confocal microscopic instrumentation (RCM 1.1 Visible, Confocal.nl) and a motorized z-stage (DI1500). Three channels were recorded for signal detection: Blue/DAPI/405-laser, AlexaFluor488/Green/Argon-laser, and AlexaFluor594/Red/HeNe-543 laser. Images were acquired with a digital camera controlled by Zeiss ZEN 2010 software or using the NIS-Elements v 5.11 software (Nikon). Samples were imaged by z-stack optical series with a step-size of 0.3-0.5 microns depth. The z-series were displayed as maximum z-projections, and gamma, brightness, and contrast were adjusted (identically for compared image sets) using Image J software, FIJI version (26).

Measurements of defined parameters (e.g. area, integrated density, number) were performed with FIJI/ImageJ software (version: 1.53c) (26). The percentage of cells releasing NETs was assessed as described by Brinkmann et al. (27) with minor modifications. Three random images per animal (n = 3) were taken at 10× magnification using a confocal microscope. Histone-DNA and DAPI signals were acquired at the same time for each image. A manual threshold was applied to each channel using the clustering algorithm of Otsu (28) and the total number of particles was counted. NETosis percentage was calculated using the following formula:

To determine the fluorescence intensity of cell cycle proteins used in this study, z-stack images of each channel were projected to obtain a single image using the maximum projection algorithm. The segmentation workflow consisted in two steps. First the cell boundaries were defined in the DIC channel to select the region of interest (ROI) which belongs to each cell using the following parameters: Gaussian blur 2 pixels, subtract background radius 30 pixels, threshold Otsu, fill holes binary process, analyzed particles over 20 pixels and circularity 0.5-1. Then, the fluorescence intensity in the green channel of the pre-defined ROI was determined as raw integrated density.

A schematic workflow of the image analysis is shown in the Supplementary Figure 1 .

For TEM analysis, 1 x 10⁷ bovine PMN were confronted with vital 2 x 10⁷ T. gondii tachyzoites, and after 2 h of co-incubation, cells were fixed in glutaraldehyde (2.5% final concentration, Merck). The cells were centrifuged at 1000 × g for 10 min and the pellet was stored at 4 °C until further use according to (7) and (29). Briefly, the cell pellet was washed and post-fixed in buffer containing 1% osmium tetroxide (Merck). After thoroughly washing in distilled water, the samples were incubated overnight in 2% aqueous uranyl acetate (Merck) at 4 °C, dehydrated in ethanol and embedded in Epon (all Merck). Ultrathin sections of the cured blocks were mounted on formvar-coated grids and stained with uranyl acetate and Reynolds lead citrate (both Merck). The sections were inspected in a ZEISS EM912 AB microscope (Oberkochem, Germany) at the Institute of Anatomy and Cell Biology of the Justus Liebig University Giessen, Germany.

Hypothesis testing was performed by a Mann-Whitney test with a confidence level of 95%. All graphs (mean ± SD) and statistical analyses were performed in the Graph Pad software (v. 7.03).

NET formation of bovine PMN being confronted with T. gondii tachyzoites was studied at 15, 30 and 60 min of interaction. The percentage of NET-positive cells was determined using an automated protocol and using NE- and DNA-related fluorescence as reference to define NET structures according to Brinkmann et al. (27). Therefore, immunofluorescence using anti-Histone-DNA and anti-NE antibodies was used ( Figure 1A ). After 60 min of interaction, NET structures showed co-localization of both parameters and resulted in typical NET-related characteristics, such as membrane-unbound and spread chromatin ( Figure 1A , 60 min). Quantification of NET-positive cells unveiled 9.25% of bovine PMN undergoing a NETotic process after 60 min of exposure to T. gondii tachyzoites in a 1:2 ratio ( Figure 1B ). In non-stimulated bovine PMN controls, 4% of NETs-positive cells were found using this quantification method. Stimulation of PMN with the calcium ionophore A23187 was used as positive control, inducing 33% of NET-positive cells after 60 min ( Supplementary Figure 2 ).

Signals of the proliferation marker Ki-67 in tachyzoite-exposed bovine PMN were monitored at 60 min of interaction by confocal microscopy. Non-stimulated cells were used as negative controls. After parasite exposure, an upregulation in abundance of Ki-67-derived fluorescent signals was observed. Unstimulated PMN presented either low Ki-67 related fluorescence signals (always associated with the PMN nuclei; Figure 2A , zoomed cells in white squares and Supplementary Video 1 ) or non at all. On the other hand, parasite-exposed PMN showed upregulation of Ki-67 (p ≤ 0.001) after 60 min of interaction ( Figure 2B , zoomed cells in white squares and Supplementary Video 2 ) thereby being correlated with the NET release-related time point.

Given that it was previously described that PMA-stimulated PMN duplicate centrosomes when forming NETs (22), we used the centrosome marker γ-tubulin to follow centrosomal structuring during activation of T. gondii-confronted PMN at 15, 30, 60 and 120 min of co-incubation ( Figure 3 , Supplementary Figure 3 ). The centrosome is visualized as a green dot in very close proximity to nuclear chromatin (stained with DAPI-mediated blue fluorescence). Current data show that centrosome duplication is absent in T. gondii-confronted bovine PMN ( Figure 3 ). Likewise, this finding was also observed when the co-incubation time was extended to 120 min. ( Supplementary Figure 3 ).

Next, we evaluated the abundance of the mitotic marker p-HH3S10 in NETotic PMN after 15, 30 and 60 min of co-incubation with T. gondii tachyzoites ( Figure 4A ). Here, the phosphorylated form of HH3 in bovine PMN was detected as early as 30 min after tachyzoite exposure. Quantification of p-HH3S10-derived fluorescent signals evidenced an increased neutrophil abundance of p-HH3S10 (p = 0.0024) at 60 min of T. gondii exposure ( Figure 4B ), thereby indicating that the mitotic machinery was indeed active in NETotic bovine PMN.

The abundance and distribution of the nuclear lamina marker lamin B1 was also studied at 60 min after PMN exposure to T. gondii tachyzoites. In control PMN, lamin B1 signal showed a homogeneous perinuclear pattern indicating intact nuclear membranes ( Figure 5A–C ). In contrast, when the PMN are confronted to T. gondii tachyzoites ( Figure 5D ), changes in lamin B1-derived fluorescence signals indicate nuclear membrane disruption since the signal are relocated ( Figure 5F ) or even lost when the PMN are forming NETs ( Figure 5E ).

To verify confocal microscopy-related observations and to obtain further ultrastructural details of T. gondii-induced NET formation in bovine PMN, we also performed TEM analysis ( Figure 6 ). The cell structures of interest are indicated in non-stimulated bovine PMN as follows, g: granules, cf: cytoskeleton filaments, n: nucleus, white asterisk (*): centrosome and nuclear membrane indicated by white arrows ( Figure 6A ). T. gondii-confronted PMN ( Figure 6B ) showed NETs (N) being extruded from PMN to the extracellular space where T. gondii tachyzoites (t) were present. Interestingly, the centrosome (asterisks) was present in NET structures in close proximity to the PMN (white square; zoomed in the right panel). Figure 6C illustrates a discontinuous PMN nuclear membrane (white arrows) in two different nuclear regions of a NETotic PMN (white rectangles).