Anti-neutrophil cytoplasmic antibodies stimulate release of neutrophil microparticles

Abstract

The mechanisms by which anti-neutrophil cytoplasmic antibodies (ANCAs) may contribute to the pathogenesis of ANCA-associated vasculitis are not well understood. In this study, both polyclonal ANCAs isolated from patients and chimeric proteinase 3-ANCA induced the release of neutrophil microparticles from primed neutrophils. These microparticles expressed a variety of markers, including the ANCA autoantigens proteinase 3 and myeloperoxidase. They bound endothelial cells via a CD18-mediated mechanism and induced an increase in endothelial intercellular adhesion molecule-1 expression, production of endothelial reactive oxygen species, and release of endothelial IL-6 and IL-8. Removal of the neutrophil microparticles by filtration or inhibition of reactive oxygen species production with antioxidants abolished microparticle-mediated endothelial activation. In addition, these microparticles promoted the generation of thrombin. In vivo, we detected more neutrophil microparticles in the plasma of children with ANCA-associated vasculitis compared with that in healthy controls or those with inactive vasculitis. Taken together, these results support a role for neutrophil microparticles in the pathogenesis of ANCA-associated vasculitis, potentially providing a target for future therapeutics.

Figure 1.

Flow cytometric analysis of NMPs derived from ANCA-stimulated primed neutrophils. (A) Dot plot demonstrating the gating strategy for MPs. Annexin V⁺ MPs derived from stimulation of primed neutrophils from a healthy adult donor with 12.5 μg/ml IgG1 anti-PR3 chimeric antibodies. The NMP gate was defined by forward-scatter characteristics corresponding with a size <1.1 µm and positive annexin V labeling. (B–G) Histograms demonstrating detection of NMP expression of CD18 (B), MPO (C), CD66b (D), PR3 (E), total CD11b (F), and active CD11b (G). The same flow cytometry strategy was used to quantify NMPs in platelet-poor plasma from healthy donors or from patients with AAV.

Figure 2.

Human polyclonal ANCAs and anti-PR3 chimeric ANCA stimulate NMPs release from primed neutrophils. (A) Neutrophils were primed with 2 ng/ml TNF-α for 15 minutes and then treated with 200 μg/ml polyclonal PR3-ANCA, polyclonal MPO-ANCA, or control polyclonal IgG for 60 minutes. Both PR3-ANCA and MPO-ANCA stimulation of primed neutrophils resulted in a five- or nine-fold (respectively) increase in NMPs, an effect not observed with priming alone or using control polyclonal IgG with or without priming. *P<0.05. (B) This experiment was repeated using 12.5 μg/ml chimeric IgG1 or IgG3 PR3-ANCA or control chimeric IgG3 antibody for 60 minutes. Chimeric PR3-ANCA again resulted in NMP release from primed neutrophils, an effect not observed with chimeric control IgG3 (P=0.24). There was no difference between IgG1 and IgG3 PR3-ANCA in their ability to induce NMP production (P=0.79). ***P<0.001 and **P<0.01. (C) Dose-response curve for NMP release in response to chimeric IgG3 PR3 ANCA demonstrated a sharp rise in NMP release from primed neutrophils from a concentration up to 5 µg/ml, which then plateaued at higher concentrations. Annexin V⁺ NMPs expressed several neutrophil markers including CD18, CD11b, and the ANCA antigens PR3 and MPO. Data are expressed as mean and SEM of three experiments. (D) Comparison of the phenotype of NMPs spontaneously released from resting neutrophils with that of the NMPs derived from ANCA stimulation (anti-PR3 IgG3, 12.5 µg/ml). PR3-ANCA stimulated NMPs had higher expression of active CD11b (P=0.04), MPO (P=0.01), and PR3 (P=0.01) but no significant difference in CD66b (P=0.9). In contrast, compared with resting NMPs, fMLP-NMP had comparable expression of MPO (P=0.60), PR3 (P=0.48), active CD11b (P=0.08), and CD66b (P=0.9). MFI, median fluorescence intensity.

Figure 3.

NMPs bind to endothelial cells using CD18. HUVECs were incubated with NMPs for 30 minutes at 37°C. (A–C) After washing the HUVECs twice, NMP binding was measured by flow cytometry and was represented as the percentage of HUVECs expressing specific neutrophil markers CD66b (A), CD11b (B), and MPO (C). This NMP binding to HUVECs was attenuated by addition of a blocking CD18 antibody (A–C).

Figure 4.

NMPs derived from IgG1 or IgG3 anti-PR3 chimeric antibody–stimulated neutrophils cause upregulation of endothelial ICAM-1 and cytokine production. HUVECs were incubated with NMPs derived from the indicated experimental conditions for 24 hours. The expression of ICAM-1 (CD54) was measured by flow cytometry. (A) Representative flow cytometric profile of ICAM-1 on HUVEC expression after stimulation with IgG3 PR3-ANCA–derived NMPs compared with that of control. (B) ICAM-1 expression on HUVECs increased in response to NMPs derived from stimulation of primed neutrophils using chimeric anti-PR3: summary of three to four experiments.

Figure 5.

Removal of NMPs by filtration or blockade with anti-CD18 attenuated the stimulatory effects of NMP on HUVECs. (A and B) Representative flow cytometry analysis (A) and summary of three experiments (B) of NMPs derived from supernatants of neutrophils stimulated with IgG1 PR3-ANCA stimulation before and after 0.2-μm filtration demonstrating 90% removal of annexin V⁺ events. (C and D) Removal of NMPs from supernatants by filtration (0.2 μm) abolished the effect of the supernatant to upregulate ICAM-1 on HUVECs (C), as did blockade of NMP binding to HUVECs using anti-CD18 (D). (E and F) Chimeric IgG1 PR3-ANCA NMPs caused increased HUVEC release of IL-6 (P=0.03) and IL-8 (P=0.01); this response was again attenuated by preincubation of HUVECs with neutralizing mAb to CD18 or filtration. Incubation of HUVECs with TNF-α at 100 ng/ml for 24 hours served as a positive control in these experiments. (G and H) Dose-response curves were plotted comparing NMPs derived from chimeric IgG3 PR3-ANCA stimulation versus NMPs spontaneously released from resting neutrophils. Using endothelial CD54 upregulation (G) and IL-8 production (H), there was higher endothelial activation using NMPs derived from IgG3-PR3 ANCA stimulation in a dose-dependent manner (P=0.02 for ICAM-1; P=0.02 for IL-8 production at the highest NMP concentration).

Figure 6.

NMPs induce endothelial ROS production, and antioxidants can inhibit the effects of NMP-mediated endothelial activation. (A) HUVECs were loaded with H₂DCFDA in HBSS and exposed to NMPs derived from different experimental conditions (as indicated) or H₂O₂ (positive control). Oxidation-dependent fluorescence of H₂DCFDA was determined in a plate reader at 525 nm, and rates of fluorescence increase were determined over 40 minutes and normalized to untreated, control HUVECs. Both IgG1 and IgG3 PR3-ANCA–derived NMPs caused increased production of ROS in HUVECs (^§P<0.01). (B) HUVECs were pretreated for 1 hour with 100 mM apocynin or 3 μM MnTBAP, after which the NMPs were added and incubated for an additional 24 hours. Apocynin or MnTPAB inhibited upregulation of ICAM-1 induced by NMPs. Data are presented as mean ± SE of three experiments. ^††P<0.05 versus untreated cells.

Figure 7.

NMPs derived from stimulation of primed neutrophils with PR3-ANCA generate thrombin in MP-depleted human plasma. (A) Representative example of a thrombin-generation assay curve that measures four parameters: lag phase, velocity index, peak thrombin, and area under curve (also referred to as endogenous thrombin potential). (B) Summary of peak thrombin (nM) of the thrombin generation assay: NMPs derived from stimulation of primed neutrophils with PR3-ANCA generated thrombin in MP-depleted human plasma.

Figure 8.

NMPs are increased in plasma of children with active AAV. (A) Patients with active AAV (n=9) had significantly higher total AnV⁺ MPs (median 642 × 10³/ml, range 447 × 10³/ml to 1981 × 10³/ml) than that of children with inactive AAV (n=4) (total AnV⁺ 237 × 10³/ml, range 57 × 10³/ml to 503 × 10³/ml), P=0.02. There was no difference in total AnV⁺ MPs between active AAV and active ANCA-negative vasculitides. (B) Breakdown of specific NMP markers in the active AAV versus inactive AAV patients: median active CD11b 144 (21 × 10³/ml to 575 × 10³/ml) versus 19 (14 × 10³/ml to 62 × 10³/ml), P=0.03; median CD66b 79 (37 × 10³/ml to 207 × 10³/ml) versus CD66b 32 (16 × 10³/ml to 48 × 10³/ml), P=0.02; median MPO NMPs 52 (26 × 10³/ml to 207 × 10³/ml) versus 14 (3 × 10³/ml to 59 × 10³/ml), P=0.04. Healthy controls and children with ANCA-negative vasculitis had fewer plasma NMPs of all subtypes. ^‡P<0.03 and ^§P<0.01.