Eosinophils in anti-neutrophil cytoplasmic antibody associated vasculitis
Anti-neutrophil cytoplasmic antibodies associated vasculitides (AAV) are characterized by autoimmune small vessel inflammation. Eosinophils are multifunctional cells with both pro-inflammatory and immunoregulatory properties. Tissue activated eosinophils secrete cyto- and chemokines and form extracellular traps (EETs), they release free granules and produce reactive oxygen species. The role of eosinophils is well established in eosinophilic granulomatosis with polyangiitis (EGPA) but very little is known about their role in granulomatosis with polyangiitis (GPA) and microscopic polyangiitis (MPA).
The expression of surface markers CD11c, CD11b, CD16, CD35, CD62L, CD64, CD88, Siglec-8 and CD193 and reactive oxygen species production by peripheral blood eosinophils were studied using flow cytometry. Fluorescence microscopy was used to visualize the release of eosinophil extracellular DNA traps (EETs). 98 GPA and MPA patients and 121 healthy controls were included in the study.
Both GPA and MPA patients had decreased frequency of eosinophils in peripheral blood compared with healthy controls (p < 0.0001), which could not solely be explained by corticosteroid treatment. The patient’s eosinophils showed increased surface expression of the Fc receptors CD16 (p < 0.0001) and CD64 (p = 0.0035) as well as CCR3 (CD193) (p = 0.0022). Decreased expression was found of the complement receptors CD35 (p = 0.0022), CD88 (p < 0,0001) as well as CD11c (p < 0,0001), CD11b (p = 0.0061) and Siglec-8 (p = 0,0015). Moreover, GPA and MPA eosinophils, showed decreased capacity to produce ROS (p < 0.0001). ANCA stimulation of eosinophils from GPA and MPA patients after C5a priming enhanced EETosis (p = 0,0088).
The percentage of eosinophils were decreased in peripheral blood in GPA and MPA patients and showed altered surface marker expression and function. The enhanced EETosis after ANCA stimulation, suggests that eosinophil can contribute to the autoantibody driven inflammatory process.
KeywordsANCA Systemic vasculitis Eosinophils Reactive oxygen species Granulomatosis with polyangiitis Microscopic polyangiitis Extracellular traps
Anti-neutrophil cytoplasmic antibodies associated vasculitides
Anti-neutrophil cytoplasmic antibodies
Birmingham Vasculitis Activity Score version 3
Complement Component 5a
Complement Component 5a receptor
C-C chemokine receptor type 3
Cluster of differentiation
Complement receptor 1
Eosinophil extracellular traps
Eosinophilic granulomatosis with polyangiitis
Enzyme Linked immunosorbent assay
Extra cellular traps
Immunoglobulin gamma Fc receptor
Granulocyte monocyte-colony stimulating factor
Granulomatosis with polyangiitis
Healthy blood donor
Human serum albumin
Inter quartile range
Magnetic-activated cell sorting
Geographic mean fluorescence intensity
Nicotinamide adenine dinucleotide phosphate
Neutrophil extra cellular traps
Phosphate buffered saline
Programmed death-ligand 1
Reactive oxygen species
Roswell Park Memorial Institute medium
T helper cell 1
Tumor necrosis factor
ANCA (anti-neutrophil cytoplasmic antibody) associated vasculitides (AAV) is characterized by an autoimmune small vessel inflammation  and autoantibodies, ANCA. AAV can be divided into three clinical diagnoses: eosinophilic granulomatosis with polyangiitis (EGPA), granulomatosis with polyangiitis (GPA) and microscopic polyangiitis (MPA).
The antigens of the autoantibodies are proteinase 3 and myeloperoxidase, that primarily are found in primary granules in neutrophils and peroxidase positive lysosomes in monocytes. Both monocytes and neutrophils are frequently found around the inflamed vessel walls and are thought to be the main effector cells. Primed neutrophils in AAV patients can be stimulated by ANCA through binding to membrane bound PR3 or MPO and in response to this they produce reactive oxygen species (ROS), de-granulates and form neutrophil extracellular traps (NETs). However, it is not known why ANCA is formed or what primes PMNs in vivo. Since eosinophils express PR3  and eosinophil peroxidase (high structural homology to MPO) on their surface, ANCA might bind and activate also this cell type.
Eosinophils have for a long time been considered as nonspecific cytotoxic cells playing a role in parasitic infections and allergy. However, during the last couple of years a more complex view has evolved, stating eosinophils as multifunctional cells with for instance immunoregulatory properties . At baseline, eosinophils are present in several tissues, as bone marrow, adipose tissue and gastrointestinal tract . They contain more than 30 different pre-synthesized and stored proteins in their cytoplasmic granule and express receptors for proinflammatory cytokines, chemokines, lipid mediators, complement factors and immunoglobulins [5, 6]. Eosinophils could also selectively suppress Th1 cells via a constitutive expression of indoleamine 2,3-dioxygenase, an enzyme important for tryptophan catabolism  and they have proven to be essential for the survival of plasma cells by supplying necessary cytokines into the plasma cell niches .
Extracellular DNA traps formation was first described in neutrophils but is now considered as a common mechanism for the innate immune system. A main difference between NETs and EETs (Eosinophil Extracellular DNA Traps) is the association of intact granules to DNA in EETs . Moreover, viable eosinophils can form EETs by releasing mitochondrial DNA and granule derived proteins .
We hypothesize that eosinophils are hyper reactive in situations of chronic inflammation, e.g. AAV, and can rapidly be recruited to tissues by for example innate lymphoid cell in response to microbes or other triggers. The activated eosinophil will accentuate the inflammation by releasing cyto- and chemokines, ROS production, deposition of free granules and EET formation. Moreover, neutrophils, monocytes and cells within the adaptive immune system will be recruited and give rise to additional tissue damage and pathology.
EGPA is characterized by eosinophilia and necrotizing eosinophil inflammation but very little is known about the role of eosinophils in GPA and MPA. Hence, the aims of this study are to characterize eosinophils from GPA and MPA patients, in regard to function and activation in order to understand their role in the disease.
Patients and controls
Patient characteristics and demographics
All patients (n = 98)
Remissiona (n = 76)
Active diseasea (n = 22)
Age, years, median (range)
Disease duration, years, median (range)
GPA, %, (n)
MPA, %, (n)
PR3-ANCA, %, (n)
MPO-ANCA, %, (n)
BVAS3, median (range)
Leukocytes, 109/L, median (range)c
p-creatinine, μmol/L, median (range)
p-CRP, mg/L, median (range)
2.9 (< 0.6–92)
2.6 (< 0.6–27)
5.7 (< 0.6–92)
Prednisone, %, median dose of treated
59 (6.25 mg/d)
53 (5 mg/d)
82 (17.5 mg/d)
Azathioprine, % (n)
Mycophenolate, % (n)
Rituximab, % (n)
Methotrexate, % (n)
Cyklophosphamide, % (n)
No current treatment, % (n)
The expression of selected surface markers on phagocytes was analyzed using flow cytometry. Briefly, heparinized peripheral blood (4-6 mL) was lysed, by adding 45 mL 0.84% ammonium chloride and incubated for 10 min. The lysed blood was centrifuged for 10 min at 250 g. The cells were washed once with PBS and after centrifugation resuspended in 100 μl PBS with 0.5% BSA. The cells were divided into two tubes and incubated for 20 min with antibody mix 1 and 2 respectively (Mix 1: CD10-PECy7, CD14-V500, CD16-APC-H7, CD88-PE, CD49d-APC, CD62L-FITC, CD11b-v450, CD11C-PerCPCy5.5 and Mix 2: CD10-PECy7, CD14-PerCPCy5.5, CD16-APC-H7, CD35-FITC, CD49D-APC, CD64-v450, CD193-v500, Siglec-8-PE. All antibodies were from BD Biosciences except CD11c and Siglec-8 that were purchased from BioLegend. The cells were then washed by adding 3 mL PBS and centrifuged for 3 min at 250 g and resuspended in 25 μL PBS and analyzed using a FACSCanto II and the DIVA software (Becton Dickinson Biosciences, New York, USA). Doublet cells were excluded by plotting FSC height against FSC area and the single cells were divided into monocytes, lymphocytes and granulocytes based on FSC and SSC plots. CD14 positive cells were excluded from the granulocytes and eosinophils were selected as CD16−/CD10− and CD49d+, Siglec-8+ and/or CD193+ cells. 30,000 events in the granulocyte gate were recorded. Surface expression were measured as mean fluorescence intensity (MFI) by calculating the geographic mean for the respective peak.
Production of reactive oxygen species (ROS) in peripheral blood eosinophils was investigated using the PhagoBurst assay, (Glycotope Biotechnology, GmBH, Germany), according to the manufacturer’s protocol, after ex vivo activation with phorbol-12-myristate-13-acetate (PMA) or opsonized E. coli. After the fixation of the cells, they were labelled by a Siglec-8-PE antibody (BioLegend) and analyzed by flow cytometry. At least 15.000 PMN were collected based on forward and side scatter properties. Eosinophils were defined as Siglec-8+ granulocytes (Additional file 1). After fixation and lysing (according to the manufacturer’s protocol), it was possible to select eosinophils also by their forward and side scatter properties. No patient with ROS deficiency was observed.
Isolation of blood eosinophils and neutrophils and detection of extracellular DNA traps
Eosinophils and neutrophils were purified from peripheral blood collected in EDTA tubes for ETosis experiments. The cells were isolated using Histopaque 1119 (Sigma) followed by Percoll (GE Healthcare) gradient following the manufacturers protocols . The eosinophils were thereafter separated from the granulocytes using MACS Eosinophil Isolation Kit (Miltenyi Biotech) according to manufacturer’s instruction. The cells that were removed during the eosinophil purification step were regarded as neutrophils. Viability at isolation was determined with Trypan blue. Cytospin preparations stained with May-Grünwald Giemsa were used to determine the cell purity. (Additional file 2A and B). The purified neutrophils and eosinophils were used to measure NET/EET production after stimulation with PBS, TNFα, C5a and PMA (see below).
100 μL of purified eosinophils or neutrophils at a concentration of 1 × 106 /mL in 0.5%HSA/RPMI were seeded on Poly-L-lysin treated coverslips (Sigma). Neutrophils and eosinophils from healthy controls (n = 7) were stimulated with PBS, C5a (150 ng/mL), TNFα (5 ng/mL) or PMA (25 ng/mL) for 180 min at 37 °C and 5% CO2. The cells were fixed with 2% PFA (BD Cytofix, BD Biosciences) for 10 min at room temperature. The coverslips were carefully washed three times with washing buffer (3% BSA in PBS) before the cells were permeabilized for 5 min (PBS containing 3% goat serum, 3% cold water fish gelatin, 1% BSA, 0,05% Tween20, 0,5% TritonX100) at room temperature. The coverslips were then washed carefully three times with washing buffer. To block unspecific binding coverslips were incubated for 30 min (PBS containing 3% goat serum, 3% cold water fish gelatin, 1% BSA, and 0,05% Tween) at room temperature and then washed twice with washing buffer. The cells were incubated with an anti-nucleosome-Alexa 594 (final concentration 5μg/ml, (B6.SLE-1) a kind gift from Anders Bengtsson, Lund University, labelled using Alexa Flour 594 Antibody Labeling Kit from Molecular Probes) for 1 h, 37 °C 5% CO2, and then washed three times with washing buffer. Finally, the coverslips were removed from the wells and mounted on slides with mounting medium (Prolong Gold antifade reagent with DAPI, Molecular Probes) and stored in dark at 4 °C until analysis. The slides were analyzed using a Zeiss AX10 fluorescence microscopy. Five representative photographs were taken from each slide and the percentage of cells that had released NETs/EETs were calculated by two independent persons.
Isolation and stimulation of blood eosinophils for EET analysis after ANCA stimulation
To investigate if ANCA can stimulate eosinophils to produce EET we purified eosinophils using the MACSXpress® Eosinophil Isolation kit (Miltenyi Biotech) according to the manufacturer’s instructions. By using this kit, the eosinophils can be isolated without the density centrifugation steps and resulted in less pre-activation of the eosinophils. Briefly, erythrocytes are aggregated and sedimented, while non-targeted cells are removed by immunomagnetic depletion. The eosinophils remain in the supernatant and are carefully collected into another tube. Viability at isolation was determined with Trypan blue. Cytospin preparations stained with May-Grünwald Giemsa were used to determine the cell purity. (Additional file 2C).
The eosinophils from five healthy controls and five GPA or MPA patients were seeded on coverslips as described above. As the number of eosinophils were limited, especially in samples taken from patients, we chose C5a over TNFα to prime the eosinophils prior to IgG stimulation. The eosinophils were primed for 15 min with PBS or C5a (150 ng/mL) at 37 °C and 5% CO2. After priming, the eosinophils were stimulated by addition of purified IgG from ANCA positive patients (250μg/mL) (one MPO-ANCA and one PR3-ANCA), purified IgG from a healthy control (250μg/mL), PBS (negative control) and PMA (positive control – PMA was not used in combination with C5a) and incubated for 180 min at 37 °C and 5% CO2. After the incubation the glass were treated as described above.
All statistical analyses were performed on GraphPad Prism 8.0 software (GraphPad Software, San Diego, CA, USA). Correlations were determined by Spearman’s correlation test and linear regression analysis. Mann-Whitney U-test was used for two group comparisons. Kruskal-Wallis and Dunn’s multiple comparisons test for three or more groups. When comparing NET and EET from the same donor Wicoxon matched-pair signed rank test was used. All p-values were considered significant at p < 0.05.
AAV patients have decreased numbers of eosinophils
Ninety-eight patients with GPA (n = 74, 76%) and MPA (n = 24, 24%) were included in the study and clinical and demographic data at time of sampling are shown in Table 1. Most patients were PR3-ANCA positive (n = 62, 63%), 30 patients were MPO-ANCA positive (30%) and five were ANCA negative. ANCA specificity from one patient was missing and one was double positive. The majority of the patients were in remission (n = 76). Twenty-two patients had an active disease, with a median activity score according to the BVAS3 of 5 (range 1 to 16).
To see if disease activity or diagnosis influenced the frequencies the patients were divided into active disease (BVAS3 ≥ 1) or inactive disease (BVAS3 = 0). Patients with active disease had significantly lower frequencies of both eosinophils and basophils compared to patients with inactive disease and HBD, but no difference were found between GPA and MPA patients (Additional file 3).
Expression of eosinophil surface markers
To characterize the status of eosinophils in peripheral blood, the expression of CD16 (FcγRIII), CD64 (FcγRI), CD35 (CR1), CD193 (CCR3), CD88 (C5aR), CD11c, CD11b, Siglec-8 and CD62L were investigated.
Decreased ROS production in eosinophils from AAV patients
Eosinophils form extracellular traps more easily than neutrophils
ANCA stimulation enhanced extracellular trap formation in eosinophils from patients
AAV are characterized by ANCA autoantibodies that binds mainly to PR3 or MPA. Eosinophils express PR3  and eosinophil peroxidase (high structural homology to myeloperoxidase) on their surface, indicating that ANCA could bind and stimulate also this cell type.
To investigate if ANCA could affect the release of EETs, eosinophils were purified using the MACSExpress eosinophil kit from five AAV patients (3 GPA and 2 MPA) and healthy controls (n = 5). The eosinophils were primed with either PBS or C5a followed by incubation with PBS (negative control), PMA (positive control), IgG from healthy controls or IgG from ANCA patients (one PR3-ANCA and one MPO-ANCA). Stimulation of eosinophils from the patients with C5a followed by ANCA gave a higher level of EETosis (p = 0,0088, Fig. 6c). This was not seen among healthy controls. EETs from these stimulations are shown in Fig. 6d. Eosinophil purification using the MACSExpress kit generated less pre-activated eosinophils from healthy blood donors as the percentage of EETs in the negative control (PBS incubation) was much lower (22% versus 5.5% in Fig. 6a and c respectively). Nonetheless, stimulation of eosinophils with C5a alone generated more EETs (61% in Fig. 6a and 24% in Fig. 6c) compared to NETs in experiments done on neutrophils (0.4% in Fig. 6a).
Eosinophils are multifunctional cells with both pro-inflammatory and immune-regulatory properties that have been suggested to regulate local tissue immunity, repair and remodeling . Moreover, eosinophils seem to be important for B cell activation and the homing and survival of plasma cells . There is some evidence that eosinophils have a protective role in autoimmunity e.g. Finlay et al. showed that activated eosinophils conferred protection against experimental allergic encephalomyelitis . Here we show that the percentage of eosinophils were decreased in peripheral blood in GPA and MPA patients and showed altered surface marker expression and function. Moreover, ANCA stimulation enhanced EETosis in GPA and MPA patients a phenomenon not seen in healthy controls. The altered surface expression and low ROS production in eosinophils is true for both patients in remission and with active disease and no difference were seen between GPA and MPA patients. The results are in line with our previous findings for neutrophils in AAV patients .
Glucocorticosteroid treatment has been reported to affect the number of eosinophils in peripheral blood, by inducing apoptosis and inhibiting pro-survival signals by cytokines e.g. IL-5 and GM-CSF [20, 21]. In line with earlier observation we saw a weak correlation between corticosteroid treatment and the number of eosinophils in blood (r2 = 0.088, p = 0.008). However, the decreased levels of eosinophils could not completely be explained by corticosteroid treatment as no significant correlation was observed between patients with active disease with or without corticosteroid treatment (p = 0.27). Two patients were newly diagnosed and both of them had low levels of eosinophils already before treatment was started. Even if the pro-apoptotic effect of steroids on eosinophils is widely accepted, there are reports suggesting that it depends on the activation status of the eosinophils . This could be one explanation to why we could not detect any correlation in patients with active disease. Moreover, dexamethasone has been reported to increase eosinophil differentiation and proliferation from CD34+ hematopoietic stem cells . Why GPA and MPA patients have decreased frequencies of eosinophils in peripheral blood is not known. A possible explanation could be that activated eosinophils are recruited to sites of inflammation and play an active role in the pathogenesis of AAV. Mononuclear cells have been reported to be the most frequent cell type in interstitial infiltration of ANCA-positive renal biopsies from patients with suspected systemic vasculitis, however eosinophils and neutrophils were found in 20 and 27% of the cases respectively .
An important feature of eosinophils is the binding of antibodies and of complement proteins via specific receptors, events that will induce degranulation and antibody mediated cellular cytoxicity and eventually killing of invading microbes or host cells . Here we showed that AAV patients have increased surface expression of the low affinity FcγRIII (CD16) as well as the high affinity FcγRI (CD64), a phenomenon described earlier in patients with allergic conditions . In several conditions, including EGPA, an increased expression CD69 and CD11b and shedding of CD62L are markers of activated eosinophils [25, 27]. In this study we found less CD62L on eosinophils from patients with active disease but lower levels of CD11b. Recently, Lingblom et al. described a naturally occurring CD16+ eosinophil subset in peripheral blood, able to suppress T cell proliferation . In line with our observation, these CD16+ eosinophils also showed decreased expression of some surface markers including CD88 and CD11a, whereas surface molecules associated with T cells suppression e.g. PD-L1, CD54 were up-regulated. The importance of this eosinophil subset in AAV needs further investigation.
NET depositions have been observed in kidney biopsies from AAV patients. We and others, have previously shown that ANCA could stimulate NET formation by neutrophils. Moreover, sera from AAV patients seem to degrade NETs more slowly than healthy controls . Interestingly, Kraaij et al. recently showed that ANCAs did not influence NET formation by neutrophils from a healthy donor, but autologous serum induced increased formation of NETs in neutrophils from three out five studied patients , suggesting that it is not ANCA per se that induces enhanced NETs formation rather the combination of ANCA and neutrophils from AAV patients. In line with these results, we show that addition of ANCA increased EETs formation in eosinophils from AAV patients. How ANCA induce formation of extra cellular traps in neutrophils and eosinophils is not known. The high affinity IgG receptor FcγRI (CD64) has been shown to be induced on blood eosinophils by IFN-γ  and human FcγRI expressed in mice was sufficient to trigger autoimmune arthritis . The high CD64 expression on eosinophils, that is further increased in AAV patients, may explain the increased EET formation in eosinophils from AAV patients compared to healthy controls when stimulated with ANCA. An increased influx and following EET formation in the inflamed tissue could partly contribute to the lower levels of circulating eosinophils. EETosis in the tissue could also mean deposition of intact eosinophilic granules (Additional file 5) in the tissue further increasing the inflammation . The presence of free eosinophil granules in tissue from AAV patients still needs to be proven.
ROS are major effector molecules in inflammatory processes and tightly linked to EET formation [9, 10]. During the last decade, an increasing amount of data support an immune modulating role for monocyte and granulocyte produced ROS [15, 32, 33, 34], as ROS can affect redox sensitive pathways . We have previously reported that PMN and monocytes from AAV patients had decreased capacity to phagocytose and impaired ROS production, which was associated with disease activity .
In this study we show that AAV eosinophils have decreased capacity to produce ROS compared with healthy controls. Even though eosinophils seem to produce sufficient amounts of ROS to form EETs, it may have an impact on immune regulation.
The frequency of eosinophils was decreased in peripheral blood in AAV patients and they showed altered surface marker expression and function. They also produce less ROS when stimulated with opsonized E.Coli or PMA. Moreover, eosinophils produce EETs when stimulated with TNFα or C5a and addition of ANCA further increase the number of EETotic cells, suggesting that eosinophil can contribute to the autoantibody driven inflammatory process in AAV.
This work was funded by Reumatikerfonden (Swedish Rheumatism association), Anna-Greta Crafoord’s Foundation for rheumatological research, Konung Gustaf V:s 80-årsfond, The Swedish Kidney Foundation, Royal Physiographic Society of Lund and Alfred Österlunds Stiftelse. All of the funding bodies support the project financially as it was described in the project plan but have not been involved in the design of the study, data collection, analysis of data, interpretation of data or writing the manuscript.
Availability of data and materials
Raw data files from flow cytometry datasets used in the current study are available from the corresponding author on reasonable request.
This study was designed by all the authors. TH, ÅJ and ÅP performed the experiments, SO collected the clinical data, TH and ÅJ analyzed the flow cytometry data, TH, MH and ÅJ interpreted the experimental data. ÅJ, ÅP and TH were major contributors in writing the manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate
The Regional Ethics Board in Lund, Sweden (EPN2008/110) approved the study and written informed consent was obtained from all participants.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 3.Liao W, Long H, Chang CCC, Lu QJ. The eosinophil in health and disease: from bench to bedside and Back. Clin Rev Allerg Immu. 2016;50(2):125–39.Google Scholar
- 11.Watts R, Lane S, Hanslik T, Hauser T, Hellmich B, Koldingsnes W, Mahr A, Segelmark M, Cohen-Tervaert JW, Scott D. Development and validation of a consensus methodology for the classification of the ANCA-associated vasculitides and polyarteritis nodosa for epidemiological studies. Ann Rheum Dis. 2007;66(2):222–7.PubMedGoogle Scholar
- 13.Brinkmann V, Laube B, Abu Abed U, Goosmann C, Zychlinsky A. Neutrophil extracellular traps: how to generate and visualize them. J Vis Exp. 2010;36:e1724.Google Scholar
- 14.Johansson AC, Ohlsson S, Pettersson A, Bengtsson AA, Selga D, Hansson M, Hellmark T. Impaired phagocytosis and reactive oxygen species production in phagocytes is associated with systemic vasculitis. Arthritis Res Ther. 2016;18(1):1–9.Google Scholar
- 30.Kraaij T, Kamerling SWA, van Dam LS, Bakker JA, Bajema IM, Page T, Brunini F, Pusey CD, Toes REM, Scherer HU, et al. Excessive neutrophil extracellular trap formation in ANCA-associated vasculitis is independent of ANCA. Kidney Int. 2018;94(1):139–49.Google Scholar
- 32.Hansson M, Hermodsson S, Brune M, Mellqvist UH, Naredi P, Betten A, Gehlsen KR, Hellstrand K. Histamine protects T cells and natural killer cells against oxidative stress. J Interf Cytokine Res. 1999;19(10):1135–44.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.