Neutrophils were isolated as previously described (3). Peripheral blood was collected from healthy volunteer donors and anticoagulated with ethylenediaminetetraacetic acid (EDTA, monovette, Sarstedt). The ethics committee of University Medical Center Hamburg-Eppendorf approved the blood collection from volunteer donors (PV 4616). Blood was layered onto Histopaque 1119 (Sigma-Aldrich). After centrifugation for 20 min at 800 g, the neutrophil-rich layer was collected. The cells were washed with Hanks-buffered salt solution without divalent cations (HBSS−, Life Technologies) supplemented with 5 mM EDTA and 0.1% bovine serum albumin (BSA, Sigma-Aldrich). Washed cells were further fractionated on a discontinuous Percoll gradient (GE Healthcare). After centrifugation for 20 min at 800 g, the neutrophil-rich layer was collected and washed with 0.1% BSA in HBSS−. All procedures were conducted at room temperature. Neutrophil viability was greater than 98%, as determined by trypan blue (Sigma-Aldrich) exclusion.

Isolated neutrophils were seeded into sterile 96 well optical plates (Falcon) coated with 0.001% poly-l-lysine (Sigma-Aldrich) at a concentration of 5 × 10⁴ cells per well in Dulbecco’s modified Eagle medium (Life Technologies). For non-activated cells, the naïve neutrophils were seeded into the plate and incubated for 15 min to allow the adherence of cells to the wells. To induce NET formation, 0.1 µM phorbol 12-myristate 13-acetate (PMA, Sigma-Aldrich) was added, and the cells were then incubated for 4 h. In selected experiments, diisopropylfluorophosphate (DFP, Sigma-Aldrich) was added at 20 mM or at indicated concentrations to the cells 5 min before addition of 0.1 µM PMA, and the cells were incubated for up to 15 h. All incubations were performed at 37°C with 5% CO₂. The cells were fixed with 2% paraformaldehyde (PFA, Sigma-Aldrich) overnight at 4°C in case of non-activated cells and NETs. NETotic cells were fixed for 2 h at 4°C.

Prior to immunostaining, PFA-fixed neutrophils were washed twice with phosphate-buffered saline (PBS). The cells were permeabilized with 0.5% Triton X-100 diluted in PBS for 5 min. For blocking, the wells were incubated with 2% BSA in PBS. Primary and secondary antibodies were used at the following concentrations: 5 µg/ml mouse antihuman MPO antibody (Hycult Biotech, Clone 266-6K1, product number: HM2164), 5 µg/ml of mouse antihuman PR3 antibody (Clone WGM2) (14), 2.5 µg/ml rabbit antihuman MPO antibody (DAKO, product number: A0398), 5 µg/ml rabbit antihuman lactoferrin (LF) antibody (Sigma-Aldrich, product number: L3262), 5 µg/ml mouse antihuman neutrophil elastase (NE) (DAKO, Clone NP57, product number: M0752), 5 µg/ml of Alexa Fluor-546 goat anti-rabbit IgG, 10 µg/ml of Alexa Fluor-488 goat anti-mouse IgG, and 10 µg/ml Alexa Fluor-555 goat-antihuman IgG (all Life Technologies). Primary antibodies were diluted in PBS with 0.05% Tween (PBST). Secondary antibodies were diluted in PBST with 2% BSA. Primary and secondary antibodies were incubated with cells at 37°C for 30 and 45 min, respectively. The wells were washed twice with PBST for 5 min before the addition of the secondary antibodies. DNA was stained with 5 µg/ml 4′,6-diamidino-2-phenylindole (Sigma-Aldrich). Fluorescent images were acquired on the day of the IIF staining using a confocal microscope (Leica TCS SP8) or if not stated otherwise with an inverted fluorescence microscope (Zeiss Axiovert 200 M). Images were analyzed using ImageJ software (NIH).

We used commercially available P-ANCA-positive serum (Euroimmun, product number: CA 1211-0105-3, titer 1/32), C-ANCA-positive serum (Euroimmun, product number: CA1200-0110-3, titer 1/32), and autoantibody-free serum (Euroimmun, product number: CA1000-0110). For Figures 7 and 8, we used a commercial kit with ethanol- and formalin-fixed neutrophils to detect neutrophil granular enzymes and ANCA by IIF (Euroimmun, product number: FA1201-1005-22). Sera were diluted 20-fold in PBST supplemented with 5 mM EDTA and 0.1% BSA before the analysis by IIF.

Sera from patients with GPA and MPA were diluted 40-fold in PBST supplemented with 5 mM EDTA and 0.1% BSA before the analysis by IIF. The ethics committee of Medical Faculty of the Technical University Dresden approved the use of the patient sera for this study (EK 226112006).

Statistical analysis was performed using Prism Software (GraphPad, USA) and one-way-ANOVA followed by Tukey’s multiple comparison test. Results were considered significant at p < 0.05.

To test whether ANCAs bind to NETs, we induced NETosis by activating human neutrophils for 4 h with PMA, a potent inducer of NETs (1, 3). DNA staining showed NETs as abundant lattices of extracellular DNA filaments covering the intercellular space between activated neutrophils (Figures 1A–C). Incubation of NETs with ANCA-positive sera revealed that P-ANCAs prominently and specifically stained NETs (Figure 1A). C-ANCA-positive sera did not bind to NETs but targeted the cell bodies of netting neutrophils (Figure 1B). As negative control, we used sera without detectable autoantibodies, which stained neither NETs nor netting neutrophils (Figure 1C). NETs exceed the size of netting neutrophils by several magnitudes (1). Consequently, quantification of human IgGs showed that P-ANCAs, but not C-ANCAs stained large areas of PMA-activated neutrophils (Figure 1D).

We used naïve, non-activated neutrophils to confirm the specificity of the tested ANCA sera for targets in neutrophil cytoplasmic granules. DNA staining of unstimulated neutrophils showed the characteristic lobulated nuclei (Figures 2A–C). Both P- and C-ANCA stained granules in the cytoplasm, but not the nucleus (Figures 2A,B). Autoantibody-negative sera did not stain neutrophils (Figure 2C). To exclude that the specificity of P-ANCAs for NETs is restricted to PMA-induced NETs, we analyzed neutrophils, which formed NETs spontaneously. Short-term incubation of freshly isolated neutrophils in serum-free conditions induced the formation of NETs in a few cells spontaneously, while the vast majority of neutrophils retain their lobulated-shaped nucleus (Figure 3A) (1, 3). P-ANCA-positive sera stained the spontaneously formed NETs as well as the cytoplasm of naïve neutrophils (Figure 3B), whereas C-ANCA-positive sera stained only naïve neutrophils (Figure 3C). Quantification of the stainings of spontaneously formed NETs confirmed the specificity of P-ANCA-positive sera for NETs (Figure 3D). In conclusion, these data show that NETs contain antigens targeted by P-ANCAs, but not C-ANCAs.

These results suggest that NETs may serve as an IIF substrate to discriminate patients with P-ANCAs from patients with C-ANCAs. To test this hypothesis, we analyzed sera from eight MPA patients and eight GPA patients. Eight sera from healthy donors were included as negative controls. We chose MPA and GPA sera, which stained ethanol-fixed neutrophils, the conventional substrate for ANCA detection, with a similar intensity (IIF titers, MPA: 420 ± 190.03, GPA: 640 ± 418.97, p > 0.05). We analyzed the patient and control sera by IIF using unstimulated naïve neutrophils and NETs as a substrate. Sera from MPA and GPA patients showed a similar cytoplasmic staining pattern on unstimulated neutrophils (Figures 4A,B). Consequently, autoreactive IgGs in MPA and GPA sera stained only the small area of the neutrophil cell bodies (Figure 4C). Control sera from healthy donors did not stain neutrophils (Figure 4C). Using NETs from PMA-activated neutrophils as a substrate, we observed a web-like staining pattern in the intercellular space with sera from MPA patients (Figure 4D), whereas sera from GPA patients stained the cell bodies of netting neutrophils (Figure 4E). In line with these results, quantification revealed larger stainings by autoantibodies in sera from MPA patients compared to GPA patients (Figure 4F). Control sera did not stain NETs or netting neutrophils (Figure 4F). In summary, IIF staining of NETs allows the discrimination of P- and C-ANCA in sera from patients with ANCA-associated vasculitis.

Myeloperoxidase and PR3 are the predominant antigens recognized by P- and C-ANCAs, respectively (5, 11). We therefore speculated that NETs are rich in MPO, but not PR3. Indeed, staining of NETs from PMA-activated neutrophils with anti-MPO antibodies showed the co-localization of MPO and NET-DNA fibers (Figure 5A), whereas anti-PR3-antibodies targeted the cell bodies of netting neutrophils (Figure 5B). Additional known antigens of P-ANCAs are LF (15) and NE (16, 17). Staining with specific antibodies against LF and NE showed that both enzymes are present in NETs (Figures 5C,D), confirming previous reports (1). Using unstimulated neutrophils and spontaneously formed NETs as a substrate, IIF for antibodies against MPO, PR3, LF, and NE showed a cytoplasmic pattern in cells with lobulated nuclei (Figures 5E–H), whereas the spontaneously formed NETs stained only for MPO (Figure 5E), LF (Figure 5F), and NE (Figure 5G), but not for PR3 (Figure 5H). Control IgG did not stain neutrophils or NETs. Taking together, these data suggest that during NETosis, NETs are selectively loaded with enzymes targeted by P-ANCAs, while the C-ANCA antigen PR3 resides in the cell bodies of netting neutrophils.

This conclusion is supported by a previous study, which showed that NE and MPO, but not PR3, translocate to the nucleus in the initial phases of NETosis (4). Within the nucleus, NE mediates the unfolding of chromatin by cleaving histones, and NE is therefore required for NET formation (4). We activated neutrophils with PMA in the presence of DFP, which inhibits the enzymatic activity of neutrophil serine proteases, including NE (6, 18). DFP blocked the generation of NETs and nuclei of PMA-activated neutrophils retained their lobulated shape (Figures 6A,B). Quantification of the area covered by DNA from PMA-activated neutrophils showed that the inhibition of NET formation by DFP was dose dependent (Figure 6C) and effective even at prolonged activation times (Figure 6D). Localization of MPO by immunostaining showed abundant MPO in NETs (Figure 6E), whereas MPO was localized to nucleus in neutrophils activated with PMA in the presence of DFP (Figure 6F). The nuclear translocation of MPO in the presence of DFP was dose and time dependent and detectable in vast majority of neutrophils (Figures 6G,H). DFP did not cause a nuclear translocation of PR3, which was detected within the cytoplasm even at the highest DFP concentration (Figure 6F). The data suggest that DFP inhibits the nuclear breakdown during NETosis, but not the translocation of neutrophil enzymes such as MPO into the nucleus. Therefore, activation in the presence of DFP generates neutrophils with characteristic morphological features of NETosis, which we termed “NETotic neutrophils.”

NETotic neutrophils resemble the appearance of ethanol-fixed neutrophils, which are a commonly used and commercially available substrate for ANCA testing. Ethanol fixation of neutrophils causes the translocation of P-ANCA antigens from cytoplasmic granules toward the nucleus, while PR3 remains in the cytoplasm (7). Immunostaining of ethanol-fixed neutrophils showed a prominent nuclear staining of MPO (Figure 7A) and a staining in or at the periphery of the nucleus for LF (Figure 7C) and NE (Figure 7D). Staining for PR3 showed a cytoplasmic staining only (Figure 7B). Staining of unstimulated neutrophils fixed with formalin showed a cytoplasmic and granular staining pattern for all neutrophil enzymes tested (Figures 7F–I). In conclusion, these data illustrate that the fixation with ethanol selectively translocates enzymes to nucleus, which are found in NETs, suggesting that ethanol fixation mimics the process of NETosis. Indeed, direct comparison of ethanol-fixed neutrophils and “NETotic neutrophils” showed that on both substrates P-ANCA-positive sera stain the nucleus (Figures 8A,D), whereas C-ANCA-positive sera stained the cytoplasm (Figures 8B,E). Control sera did not stain NETotic neutrophils (Figure 8C) or ethanol-fixed neutrophils (Figure 8F). Taking together, these data show that NETs and NETotic neutrophils provide a physiological substrate for ANCA testing.