Neutrophil extracellular traps can activate alternative complement pathways

Supporting information

Figure S1. Neutrophil extracellular trap (NET) induced by anti-neutrophil cytoplasmic antibody (ANCA). Neutrophils were primed with tumour necrosis factor (TNF)-α and incubated with 250 µg/ml ANCA-positive immunoglobulin (Ig)G or normal human IgG. We also stimulated neutrophils with ANCA-positive IgG alone without TNF-α priming. Phorbol myristate acetate (PMA)-activated neutrophils were used as the positive control. (a) After 180 min of incubation, the DNA release was visualized by fluorescence microscopy of 4',6-diamidino-2-phenylindole (DAPI)-stained specimens. A representative example of three independent experiments is shown. Magnification ×200. (b) Quantitative assessment of the percentage of NETs-forming cells revealed robust NET formation after ANCA activation. After ANCA activation, 24·26 ± 5·62% of neutrophils produced NETs, compared to 8·90 ± 2·78% of normal IgG-treated neutrophils, 13·51 ± 3·50% neutrophils treated with ANCA-positive IgG alone without TNF-α priming. Incubation with PMA, known as a strong inducer of NETs, triggered NETs production in 40·91 ± 6·50% of all neutrophils.

Figure S2. Isolated anti-neutrophil cytoplasmic antibody (ANCA)-positive immunoglobulin (Ig)G could bind to neutrophil extracellular trap (NETs). Neutrophils were stimulated with 50 nM PMA for 3 h at 37°C. After fixation with 4% paraformaldehyde (PFA), the samples were washed with phosphate-buffered saline (PBS) and then incubated with 250 µg/ml isolated ANCA-positive immunoglobulin (Ig)G (a) or normal human control IgG (b) for 1 h at 37°C. After washing with phosphate-buffered saline (PBS), AF488-conjugated anti-human IgG antibodies (1 : 500 dilution; Jackson Immuno Research Laboratories) were applied for 1 h at 37°C. Isolated ANCA-positive IgG bound to the NETs was seen. Representative figures are shown. Magnification ×200.

Figure S3. Immunofluorescence identifying neutrophil extracellular traps (NETs) induced by anti-neutrophil cytoplasmic antibodies (ANCA). Neutrophils were primed with tumour necrosis factor (TNF)-α and incubated with 250 µg/ml ANCA-positive-IgG. NETs induced by ANCA were identified by co-localization of DNA (blue), histone (red) and MPO (green). A representative example of three independent experiments is shown. Magnification ×400.

Figure S4. Detection of deposition of Bb and properdin on neutrophil extracellular trap (NETs) induced by phorbol myristate acetate (PMA) or lipopolysaccharide (LPS) in vitro. Neutrophils isolated from healthy donors were stimulated with 50 nM PMA or 100ng/ml LPS. NETs were identified by co-localization of DNA (blue) and elastase (green). Deposition of Bb and properdin (red) on NETs was assessed by confocal microscopy. Magnification ×400. (a) Deposition of Bb (red) on NETs induced by PMA stained by DNA (blue) and elastase (green). (b) Deposition of Bb (red) on NETs induced by LPS stained by DNA (blue) and elastase (green). (c) Isotype control for staining antibody was perfomed. Primary antibodies were replaced by normal rabbit immunoglobulin (Ig)G and mouse IgG2a. (d) Deposition of properdin (red) on NETs induced by PMA stained by DNA (blue) and elastase (green). (e) Deposition of properdin (red) on NETs induced by LPS stained by DNA (blue) and elastase (green). (f) Isotype control for staining antibody was performed. Primary antibodies were replaced by normal rabbit IgG and mouse IgG2a.

Figure S5. Immunoblotting of complements on neutrophil extracellular trap (NETs). Human neutrophils were left untreated, treated with anti-neutrophil cytoplasmic antibody (ANCA)-positive immunoglobulin (Ig)G after tumour necrosis factor (TNF)-α priming. Supernatants were collected, and Bb, properdin were examined by Western blotting. Supernatants enriched in NETs and degraded NETs (treated with commercial DNase I) were incubated by normal human serum in the presence of ethyleneglycol teraacetic acid (EGTA), C5b and C3d were examined by Western blotting. Primary antibodies were used as below: mouse anti-human Bb antibody (Quidel), rabbit anti-human properdin antibody (Abcam), rabbit anti-human C5b antibody (Abcam) and rabbit anti-human C3d anibody (Abcam). (a) Detection of Bb on NETs by immunoblotting. (b) Detection of properdin on NETs by immunoblotting. (c) Detection of C5b on NETs incubated with serum in the presence of EGTA by immunoblotting. (d) Detection of C3d on NETs incubated with serum in the presence of EGTA by immunoblotting.

Figure S6. Immunofluorescence detection for C4d on neutrophil extracellular trap (NETs). Neutrophils were primed with TNF-α and incubated with 250µg/ml anti-neutrophil cytoplasmic antibody (ANCA)-positive immunoglobulin (Ig)G. The induced NETs were then incubated with normal human serum (NHS) at the presence of ethyleneglycol teraacetic acid (EGTA) or not for 40 min at 37°C. After incubation with serum, coverslips with NETs were washed, and C4d was detected using rabbit anti-human C4d polyclonal antibodies (1 : 100 dilution; Abcam) as primary antibodies with secondary antibodies AF488-labelled donkey anti-rabbit IgG (1 : 500 dilution, Jackson ImmunoResearch Laboratories). NETs were visualized using 4',6-diamidino-2-phenylindole (DAPI) (blue). (a) C4d deposited on NETs during incubation with normal serum. (b) Little or no deposition of C4d on NETs during incubation with normal serum at the presence of EGTA could be observed. Representative images of three independent experiments are shown. Magnification ×200.

Figure S7. Quantification of mean fluorescence intensity of C3b staining on neutrophil extracellular trap (NETs) after incubation with serum. The mean fluorescence intensity (MFI) of C3b staining on NETs incubated with normal human serum (NHS), heat inactivated (Hi)-NHS (lack of active complements), magnesium salt-ethyleneglycol teraacetic acid (Mg-EGTA)-treated NHS, Mg-EGTA-treated Hi-NHS (lack of active complements). Data are presented as mean of three independent experiments ± standard deviation (s.d.). The MFI of C3b on NETs incubated with NHS was 49·74 ± 8·36, while the MFI of C3b on NETs incubated with Hi-NHS was 11·95 ± 2·85 (P<0·001). The MFI of C3b on NETs incubated with Mg-EGTA-treated NHS was 30·96 ± 5·37, while the MFI of C3b on NETs incubated with Mg-EGTA-treated Hi-NHS was 7·14 ± 1·40 (P<0·001).

Figure S8. Neutrophil extracellular trap (NETs) induced by phorbol myristate acetate (PMA) or lipopolysaccharide (LPS) could activate the alternative complement cascade in the serum. Neutrophils were stimulated by 50 nM PMA or 100 ng/ml LPS. Bars represent mean ± standard deviation (s.d.). Each measured on neutrophils of five independent experiments and donors. Supernatants from the cell suspensions were incubated for 40 min with normal serum or magnesium salt-ethyleneglycol teraacetic acid (Mg-EGTA)-treated normal serum, and C3a, C5a, SC5b-9 generation was measured by specific enzyme-linked immunosorbent assay (ELISA). *P < 0·05, **P < 0·01, ***P < 0·001.

Figure S9. Quantification of the generation of C5a in serum after specific inhibition of tissue factor or thrombin. Neutrophils were incubated with anti-TF antibody (10 μg/ml) for 30 min in 37°C or 1 μM D-Phe-Pro-Arg-choro-methylketone dihydrochloride (PPACK) for 40 min at room temperature, and were then stimulated by ANCA-positive immunoglobulin (Ig)G after tumour necrosis factor (TNF)-α-priming or 50 nM phorbol myristate acetate (PMA). Bars represent mean ± standard deviation (s.d.). Each measured on neutrophils of four independent experiments and donors. Supernatants from the cell suspensions were incubated for 40 min with normal serum or magnesium salt-ethyleneglycol teraacetic acid (Mg-EGTA)-treated normal serum, and C5a generation was measured by specific. D-Phe-Pro-Arg-choro-methylketone dihydrochloride (PPACK) was used to inactivate thrombin. Anti-TF antibody (murine monoclonal antibody against human tissue factor) was used to inhibit the pro-coagulant activity of tissue factor; n.s. no significant difference.