Clinical and experimental evidence indicate that ANCA causepauci-immune necrotizing and crescentic glomerulonephritis (NCGN)and systemic small vessel vasculitis in humans. One of the majortarget antigens for ANCA is myeloperoxidase (MPO). An animalmodel that closely resembles the human disease is induced byintravenous injection of anti-MPO IgG into mice. The likelyprimary pathogenic targets for the anti-MPO IgG are circulatingneutrophils and monocytes, although other cells have been implicated,including endothelial cells and epithelial cells. Herein isreported a new model for anti-MPO–mediated glomerulonephritisand vasculitis that further documents the pathogenic potentialof ANCA and demonstrates that bone marrow (BM)-derived cellsare sufficient targets to cause anti-MPO disease in the absenceof MPO in other cell type. MPO knockout (Mpo–/–)mice that were immunized with mouse MPO were exposed to irradiationand received a transplant of Mpo+/+ or Mpo–/– BM.Engraftment in mice with circulating anti-MPO resulted in developmentof pauci-immune NCGN in all mice and pulmonary capillaritisand splenic necrotizing arteritis in some. Anti-MPO IgG alsowas introduced intravenously into chimeric mice by transplantationof Mpo+/+ BM into irradiated Mpo–/– mice or Mpo–/–BM into irradiated Mpo+/+ mice. Chimeric Mpo–/–mice with circulating MPO-positive neutrophils developed NCGN,whereas chimeric Mpo+/+ mice with circulating MPO-negative neutrophilsdid not, thereby indicating that BM-derived cells are not onlysufficient but also necessary for induction of anti-MPO disease.This novel animal model further documents ANCA IgG interactionswith neutrophils as a cause of ANCA-associated glomerulonephritisand vasculitis.
ANCA are associated with pauci-immune necrotizing and crescenticglomerulonephritis (NCGN) and small vessel vasculitis (SVV)(1,2). The major antigens that are recognized by ANCA in patientswith glomerulonephritis and vasculitis are myeloperoxidase (MPO)and proteinase 3 (PR3). Clinical observations as well as invitro and in vivo experimental studies indicate that ANCA notonly are disease markers but also are involved directly in diseasepathogenesis (2,3). The most compelling clinical evidence tosupport causality is the observation that transplacental passageof anti-MPO IgG seemed to induce glomerulonephritis and pulmonarycapillaritis in a single neonate (4,5); however, this is a rareevent that has not been reported in other births to motherswith ANCA disease. In vitro experiments demonstrated that ANCAIgG can activate neutrophils (2,3). ANCA bind to the antigensMPO and PR3 that are expressed on the membranes of neutrophilsand monocytes (6–8). These cells are activated by thebinding of ANCA, leading to degranulation and oxygen radicalproduction (9). Activation takes place via engagement of Fcreceptors and direct binding of antigen-binding domain of ANCAIgG (10–12). Hence, the circulating cell pool of neutrophilsand monocytes is believed to be the primary target for ANCA.However, there are conflicting reports of whether endothelialcells and epithelial cells can express the ANCA antigens andcould be the primary targets for disease induction (13–16).
Several animal models of MPO-ANCA disease have been established,but no convincing model of PR3-ANCA disease has been reported.Intravenous injection of mouse anti-mouse MPO antibodies intowild-type or immune-deficient mice causes pauci-immune necrotizingand crescentic glomerulonephritis (NCGN) and systemic SVV thatclosely resembles human ANCA (17). Neutrophil depletion studiesindicate that neutrophils are the main effector cells in thisanimal model (18). However, the glomerulonephritis that is inducedby anti-MPO injection is relatively mild compared with humandisease because typically fewer than 15% of glomeruli have necrosisor crescents. The severity of glomerular injury can be enhancedby concurrent injection of bacterial LPS, which induces an increasedlevel of circulating TNF- (19). In a different approach, severeNCGN that affects >80% of glomeruli and systemic SVV canbe induced in immune-deficient mice by transfer of splenocytesfrom MPO-deficient mice that have been immunized with murineMPO (18). After anti-MPO splenocyte transfer, these mice developedincreasing anti-MPO antibody titer. Pathologic examination after13 d shows severe NCGN as well vasculitis in various organs.However, the glomerulonephritis is not pauci-immune becausethere is a moderate amount of predominantly mesangial glomerulardeposition of immune complexes. Therefore, this is not a goodmodel of pauci-immune human disease, which often has a smallamount of glomerular Ig deposition but usually does not havethe extent of deposition that is seen in mice after splenocytetransfer.
A rat model of ANCA disease has been induced by immunizing ratswith human MPO, which results in the production of anti-MPOantibodies that cross-react with human and rat MPO (20). Thiscauses focal segmental pauci-immune glomerulonephritis and focalpulmonary capillaritis. However, not all rats developed glomerulonephritis,and, when present, <10% of glomeruli had crescents or necrosis.As in the mouse model that was induced by injection of anti-MPOIgG, the glomerulonephritis was not severe.
Here we report a novel approach to the induction of ANCA glomerulonephritisand vasculitis that provides additional support for the pathogenicityof ANCA, sheds light on the pathogenic mechanisms that are involved,and provides a model that may be better suited than currentmodels for long-term studies of ANCA disease. We report theuse of transplantation of MPO-positive bone marrow (BM) intoMPO-deficient mice to induce pauci-immune NCGN and SVV.
Mice
Wild-type (WT) C57BL/6J (B6) mice were purchased from JacksonLaboratories (Bar Harbor, ME) and maintained by the Universityof North Carolina Division of Laboratory Animal Medicine. Micethat lacked MPO (Mpo–/– mice) originally were generatedby Aratani et al. (21). Table 1 details the 10 experimentalgroups of mice. Nine- to 10-wk-old mice were used as donorsof BM cells. Immunization of mice with MPO was begun at 8 to10 wk of age. All animal experiments were in accordance withNation Institutes of Health guidelines and approved by the Universityof North Carolina Institutional Animal Care and Use Committee.
Immunization of Mice
Purification of mouse MPO from WEHI-3 cells and immunizationof Mpo–/– mice were performed as described previously(17). Eight- to 10-wk-old MPO–/– mice were immunizedintraperitoneally with 10 µg of purified murine MPO (groups1 through 4 and 6) or BSA (group 5) in complete (first immunization)or incomplete (booster immunizations) Freund’s adjuvant(Table 1). Mice were immunized at day 0, receive a booster ondays 28 and day 35, and were irradiated on days 40 to 45. Developmentof antibodies was monitored by anti-MPO ELISA (17). Mice thatwere used for BM transplantation developed anti-MPO titers thatwere 80% or higher compared with a positive control serum poolthat was derived from MPO–/– mice that were immunizedby the regimen of Xiao et al. (17). Approximately 85% of immunizedmice met this criterion.
BM Transplantation in Mice
Mice were kept in sterile housing conditions with food and waterad libitum (sterile water with neomycin 2 g/L [pH 2.0]). Micein groups 1 to 9 were -irradiated with different whole-bodydosages that ranged from 450 to 900 rad (Table 1). BM cellswere harvested from femurs and tibia, erythrocytes were lysed,and 1.5 x 107 BM cells were injected intravenously and retro-orbitally.Four different combinations of BM donors and recipients wereused: WT into Mpo–/– (groups 1 through 5 and 9),Mpo–/– into WT (group 8), WT into WT (group 7),and Mpo–/– into Mpo–/– (group 6). Recoveryof peripheral blood leukocytes and peripheral blood neutrophilswas monitored after BM transplantation using the HESKA VeterinaryHematology System. Engraftment was measured as percentage ofperipheral blood neutrophils that were positive for MPO by enzymehistochemistry. Extent of engraftment stabilized after 4 to5 wk. Mice were killed for pathologic examination 8 wk aftertransplantation.
Induction of Glomerulonephritis by Intravenous Injection of Anti-MPO IgG
The IgG fraction from immunized Mpo–/– mice wasisolated from serum by 50% ammonium sulfate precipitation andprotein G affinity chromatography as described previously (17)Chimeric mice in groups 8 and 9 and control WT mice in group10 were received an intravenous injection of 50 µg/g bodywt anti-MPO IgG in PBS and killed on day 6. Just before themice were killed, serum and urine samples were collected toevaluate renal function.
Functional Evaluation of Renal Injury
Mice were placed in metabolic cages 1 d before being killed,and urine was collected for 12 h. Urine was tested by dipstick(Roche Diagnostics Corp., Indianapolis, IN) for hematuria andleukocyturia, and the extent of hematuria and leukocyturia isexpressed as the mean on a scale of 0 (none) to 4 (severe).The albuminuria was determined by Mouse Albumin ELISA QuantitationKit (Bethyl Laboratories, Montgomery, TX). Using the mean ±2 SD for the reference range established previously by our laboratoryusing 305 healthy B6 mice, abnormal hematuria was set at >0.9and leukocyturia at >0.2 (18). Serum creatinine and bloodurea nitrogen (BUN) were measured using a Johnson & JohnsonClinical Diagnostics VITROS 250 (Berkshire, UK).
Pathologic Evaluation of Renal Injury
Samples of kidney, lung, spleen, and liver tissue were collectedwhen the mice were killed and fixed in 10% formalin and embeddedin paraffin. Four-micrometer sections of specimens were stainedwith hematoxylin and eosin and periodic acid-Schiff (PAS). Theextent of glomerular crescents and necrosis was expressed asthe percentage of glomeruli with crescents and necrosis in eachmouse. For immunofluorescence microscopy to detect glomerularlocalization of immune determinants, 4-µm sections ofsnap-frozen kidney tissue were stained with fluoresceinatedgoat anti-mouse IgG (Molecular Probes Invitrogen, Carlsbad,CA); goat anti-mouse C3 (ICN/Cappel, Aurora, OH); and antibodiesthat were specific for mouse IgM, IgA, and MPO (ICN/Cappel,Aurora, OH). Immunofluorescence microscopy staining of glomeruliwas expressed as the intensity of staining on a scale of 0 to4+. For detection of leukocytes, sections of snap-frozen tissuewere stained with rat antibodies to neutrophils (anti–Gr-1,clone RB6–8C5; BD Pharmingen, Franklin Lakes, NJ) andmonocytes/macrophages (anti-CD68, clone FA11; Serotec, Raleigh,NC). Rat antibody binding was detected using peroxidase-labeledsecondary rabbit anti-rat IgG and tertiary goat anti-rabbitIgG antibodies (DAKO, Carpinteria, CA) followed by 3-amino-9-ethylcarbazoleand hydrogen peroxide. Sections were counterstained with hematoxylin.Leukocyte localization was expressed as the percentage of positiveglomeruli, the average leukocytes per cross-section of all glomeruli,and the average leukocytes per cross-section of positive glomerulion the basis of evaluation of an average of 48 glomeruli perspecimen (range 30 to 80 glomeruli).
Neutrophil Reconstitution after Irradiation and BM Transplantation
MPO-deficient mice were immunized with murine MPO. After developinganti-MPO antibody titers >100% of a positive control standard,mice were irradiated with 450 to 900 rad of whole-body irradiation.Various donor and recipient combinations were used as shownin Table 1. The time course of repopulation of the circulatingleukocyte pool after irradiation was established (Figure 1).Both MPO-immunized and nonimmunized Mpo–/– micedisplayed a similar time course and amount of leukocyte repopulation(Figure 1A). There was a greater increase in the MPO-immunizedmice (group 1) after 4 wk compared with the nonimmunized mice(group 5), but this was only marginally significant (P = 0.03).Both group 1 and group 5 had a similar rate of engraftment (Figure 2A,Table 1), thereby ruling out an immune-mediated depletion ofrepopulating MPO-positive neutrophils. MPO-positive neutrophilswere identified by enzyme histochemistry (Figure 2, B and C).The slightly higher level of neutrophils in group 1 might bethe result of the inflammatory state that accompanies the glomerulonephritisand vasculitis in these mice but not the group 5 mice. As shownin Figure 2A, there was no significant difference in engraftmentwith MPO-positive neutrophils between the 900- and 750-rad irradiationgroups, but there was a progressive decline in engraftment withlower irradiation dosages.
Figure 1. Repopulation of peripheral blood leukocytes after irradiation (900 rad) and transplantation of myeloperoxidase (MPO)-positive bone marrow (BM) into Mpo–/– mice after immunization with mouse MPO (group 1; •) or into Mpo–/– mice without previous immunization (group 5; ). (A) Total leukocyte count. (B) Neutrophil count. There is a statistically significant difference in the neutrophil counts at 28 d (P < 0.01) and 35 d (P < 0.05).
Figure 2. Engraftment with MPO-positive BM in different irradiation groups. (A) Percentage of MPO-positive neutrophils in peripheral blood in individual mice and the average for each group. (B and C) Typical MPO-negative (B) and MPO-positive (C) neutrophils in a Mpo–/– mouse before and after transplantation with MPO-positive BM, respectively. (D) The upper four lines represent the time course of anti-MPO antibody titers as determined by ELISA in groups 1 through 4, and the bottom line demonstrates no significant titer of anti-MPO in group 5 mice that were not immunized with MPO before BM transplantation. See Table 1 for a description of the groups.
Anti-MPO Antibody Titer after Irradiation and BM Transplantation
After the irradiation, anti-MPO antibody titers were monitored.Mice that had been immunized with MPO before transplantation(groups 1 through 4) had no significant decrease in anti-MPOantibody titer (Figure 2D). Irradiated nonimmunized MPO-deficientmice showed no change in their baseline anti-MPO antibody levelafter engraftment with MPO-positive cells. Therefore, no newantibody response to MPO was mounted after irradiation, andthe circulating anti-MPO seems to be the result of residualcirculating antibody or residual committed plasma cells or both.
Urine Abnormalities after BM Transplantation
Eight weeks after irradiation and reconstitution, MPO-preimmunizedMpo–/– mice that were irradiated with 900 and 750rad before transplantation with Mpo+/+ BM developed comparablelevels of marked hematuria, leukocyturia, and albuminuria (Figure 3,A and B). This was in marked contrast to nonimmunized Mpo–/–mice that were irradiated with 900 rad before transplantationwith Mpo+/+ BM and developed no hematuria, leukocyturia, oralbuminuria. MPO-preimmunized Mpo–/– mice that wereirradiated with 600 or 450 rad before transplantation with Mpo+/+BM also failed to develop hematuria, leukocyturia, or albuminuria.
Figure 3. Induction of glomerular disease in the experimental groups. (A) Standard urinalysis results. (B) Urine albumin excretion as measured by ELISA and depicted as individual mice and the average for each group. (C) Extent of glomerular crescent formation and necrosis expressed as the average percentage of involved glomeruli in each mouse.
Renal function was impaired in group 1 mice with a significantincrease of creatinine (0.38 ± 0.16 mg/dl; range 0.2to 0.7) and BUN (61.7 ± 56.9 mg/dl; range 18 to 178)compared with group 5 mice (creatinine 0.21 ± 0.06 mg/dl;BUN 31.6 ± 3.3 mg/dl; P < 0.05). All other irradiationgroups (groups 2 through 4) had no significant increase in creatinine(750 rad: 0.34 ± 0.05 mg/dl; 600 rad: 0.33 ± 0.05mg/dl; 450 rad: 0.24 ± 0.05 mg/dl) or BUN (750 rad: 31.4± 13.1 mg/dl; 600 rad: 35.6 ± 6.0 mg/dl; 450 rad:36.9 ± 8.5 mg/dl).
Histologic Evidence of NCGN
Mice were killed 8 wk after irradiation and BM transplantation.All MPO-preimmunized Mpo–/– mice that had been irradiatedwith 900 and 750 rad before transplantation with Mpo+/+ BM (groups1 and 2) developed glomerular necrosis and crescents, 71% ofpreimmunized mice that received 600 rad (group 3) developedglomerular necrosis and crescents, and none of the preimmunizedmice that received 450 rad (group 4) or the nonimmunized micethat received 900 rad (group 5) developed glomerular lesions(Table 1). No glomerular disease developed in preimmunized Mpo–/–mice that received 900 rad before Mpo–/– BM transplantation(group 6) or in Mpo+/+ WT mice that received 900 rad beforeMpo+/+ WT BM transplantation (group 7). Figure 3C and Table 2show that groups 1 and 2 mice had similar severity of glomerulonephritiswith crescents in approximately 31 to 32% of glomeruli and fibrinoidnecrosis in approximately 15 to 17% of glomeruli (Figure 4).Group 3 mice had very mild glomerulonephritis, and groups 4through 7 had no glomerular disease. These pathologic findingscorrelate extremely well with the urinary findings and renalfunction.
Figure 4. Pathologic findings in Mpo–/– mice transplanted with Mpo–/– or Mpo+/+ BM. None of the MPO-immunized Mpo–/– mice that received 900-rad irradiation followed by transplantation with Mpo–/– BM (group 6) developed glomerular lesions (A; hematoxylin and eosin [H&E]). All of the MPO-immunized Mpo–/– mice that received 900-rad irradiation followed by transplantation with Mpo+/+ BM (group 1) developed glomerular necrosis (B, arrow; H&E) and crescents (C, arrows; periodic acid-Schiff [PAS]); and some of these mice also developed necrotizing arteritis in the spleen with fibrinoid necrosis (D, arrow; PAS), pulmonary alveolar capillaritis with and septal influx of neutrophils (arrow) and intra-alveolar hemorrhage (E; H&E), and, less frequently, pulmonary granulomatous inflammation with multinucleated giant cells (F, arrows; H&E).
Immunofluorescence Microscopy Observations
Immunofluorescence microscopy demonstrated a paucity of stainingfor IgG, IgA, IgM, and C3 with no significant semiquantitativedifferences among the mice in the experimental or control groups(groups 1 through 7; Table 3, Figure 5). Groups 1 through 4had slight scattered glomerular staining for MPO that probablyrepresented infiltrating neutrophils and sites of neutrophildegranulation. Groups 1 and 2, which had high levels of engraftmentas evidenced by high percentages of circulating MPO-positiveneutrophils, had no staining of vessels away from sites of injuryand no staining of other tissue elements for MPO. Therefore,there was no evidence by immunohistology that BM-derived stemcells were giving rise to cells other than circulating myeloidleukocytes. Irregular foci of staining for fibrin seemed tocorrespond to areas of fibrinoid necrosis and crescent formation.Groups 5, 6, and 7 had no glomerular staining for MPO or fibrin.
Figure 5. Immunofluorescence microscopy findings in MPO-immunized Mpo–/– mice that received 900-rad irradiation followed by transplantation with Mpo+/+ BM (group 1). There is low-level staining for IgG, IgM, IgA, and C3 that is similar to staining that typically is seen in human pauci-immune crescentic glomerulonephritis. Scattered MPO staining seemed to correspond to infiltrating neutrophils and sites of neutrophil degranulation. Irregular staining for fibrin corresponded to foci of fibrinoid necrosis and crescent formation.
Glomerular Neutrophil and Macrophage/Monocyte Accumulation
Immunohistology demonstrated glomerular accumulation of neutrophils(anti-GR1) as well as monocytes/macrophages (anti-CD68) in micein the experimental groups, whereas mice in control groups showedno increase above normal (Table 4).
Table 4. Glomerular influx of neutrophils and macrophages in various experimental groups expressed as the percentage of glomeruli that were positive for the cell type, average cells per positive glomerulus, and overall average cells per glomerulusa
Systemic Manifestation of Vasculitis
Focal pulmonary alveolar capillaritis was identified in thelungs of mice in groups 1 (five of 13), group 2 (four of seven),and group 3 (two of seven; Figure 4G). One mouse in group 1and one mouse in group 2 also had focal pulmonary granulomatousinflammation with multinucleated giant cells that resembledlesions of Wegener’s granulomatosis (Figure 4F). Threemice in group 1 had necrotizing arteritis in the spleen (FigureD). No vasculitis was identified in groups 4 through 7.
Disease Induction by Passive Transfer of Anti-MPO IgG into Chimeric Mice
Chimeric mice with either Mpo–/– neutrophils andMPO+/+ tissue (group 8) or Mpo+/+ neutrophils and Mpo–/–tissue (group 9) were established to evaluate the pathogenictarget of passively administered anti-MPO IgG compared withdisease induction by the same dosage of anti-MPO IgG in WT Mpo+/+mice (group 10). All mice in group 9 that expressed MPO in BM-derivedcells but were otherwise Mpo–/– developed hematuria,leukocyturia, and albuminuria that were comparable to the urinaryfindings in the group 10 positive control mice (Figure 6). Incontrast, none of the group 8 mice with Mpo–/– BM-derivedcells and Mpo+/+ tissue developed urinary abnormalities.
Figure 6. Evidence for glomerular disease after intravenous injection of anti-MPO IgG into wild-type (WT) mice with no previous manipulations (group 10), into Mpo–/– mice that received a transplant of Mpo+/+ WT BM (group 9), or into WT Mpo+/+ mice that received a transplant of Mpo–/– BM (group 8). (A) Standard urinalysis results. (B) Urine albumin excretion measured by ELISA. (C) Extent of glomerular crescent formation and necrosis expressed as the average percentage of involved glomeruli in each mouse.
Histologic examination revealed that all mice in group 9 thatexpressed MPO in BM-derived cells but were otherwise Mpo–/–developed glomerular crescents (Figure 7B, arrows, PAS) andnecrosis (Figure 7C, arrow, hematoxylin and eosin), but noneof the group 8 mice with Mpo–/– BM-derived cellsand Mpo+/+ tissue developed glomerular lesions (Figure 7A, PAS).On average, positive control mice (group 10) had 11.0 ±2.0% of glomeruli with crescents and 5.0 ± 0.0% withnecrosis (Figure 6C). Group 9 mice that expressed MPO in BM-derivedcells but were otherwise Mpo–/– had crescents in14.0 ± 8.1% of glomeruli and necrosis in 7.6 ±5.0%. In contrast, group 8 mice with Mpo–/– BM-derivedcells and Mpo+/+ tissue had 0% crescents and 0% necrosis. Therefore,MPO-positive BM-derived cells and not tissue cells are the pathogenictargets of anti-MPO antibodies. In addition, these data demonstratethat BM-derived cells, most likely neutrophils and monocytes,are necessary and sufficient targets for induction of glomerulonephritisby anti-MPO antibodies.
Figure 7. Histologic findings in chimeric mice with either Mpo–/– neutrophils and Mpo+/+ tissue (group 8) or Mpo+/+ neutrophils and Mpo–/– tissue (group 9) that received an injection of a nephritogenic dose of anti-MPO IgG. None of the group 8 mice had glomerular lesions (A; PAS), whereas all of the group 9 mice had glomerular crescents (B, arrows; PAS) and necrosis (B, arrow; H&E) that was comparable to lesions in positive control mice.
ANCA vasculitis is the most common form of aggressive SVV inadults and is characterized in the acute phase by focal necrotizingvascular inflammation that is rich in neutrophils (22,23). Thereis compelling clinical and experimental evidence that ANCA antibodiesthat are specific for MPO or PR3 are involved directly in thepathogenesis of ANCA disease, but the pathogenic details havenot been elucidated fully (2,3). In vitro experimental studieshave shown that ANCA IgG can interact with and activate circulatingneutrophils, resulting in the release of injurious factors thatcould be involved in the pathogenesis of vasculitis and glomerulonephritis(2–12). However, for a better understanding of any humandisease, including ANCA vasculitis, good animal models are indispensable.Thus far, there are only a few convincing animal models of ANCAdisease. One involves passive transfer of murine anti-MPO antibodiesor anti-MPO lymphocytes from MPO-immunized MPO–/–mice into Mpo+/+ recipient mice (17). This approach does notinvolve breaking tolerance and therefore is not a model of theimmunogenesis of the ANCA autoimmune response, but it is a feasiblemodel of ANCA pathogenesis. Another model uses immunizationof rats with human MPO, resulting in the production of anti-MPOantibodies that cross-react with rat MPO (20). A model thatis more difficult to interpret uses anti-GBM antibodies to induceglomerulonephritis in mice that have been immunized with humanMPO (24). Additional animal models not only would help confirmthe pathogenic potential of ANCA but also would shed additionallight on details of the pathogenic mechanism and provide additionaloptions for future research, including research into the efficacyof treatment strategies.
In this report, we describe two new approaches for modelinghuman ANCA disease that use transplantation of Mpo+/+ BM intoMpo–/– mice. One approach involves active inductionof circulating anti-MPO antibodies in Mpo–/– miceby preimmunization with MPO before transplantation of MPO-positiveBM. The other approach involves passive intravenous injectionof anti-MPO antibodies into chimeric mice that have circulatingMPO-positive BM-derived cells but are otherwise Mpo–/–.
When recipient mice are preimmunized with MPO, the presenceof anti-MPO antibodies did not prevent successful engraftmentwith MPO-positive BM because preimmunized and nonimmunized MPO-deficientmice had similar engraftment with MPO-positive BM (Figures 1and 2). It is interesting that preimmunized mice developed aslight increase in the peripheral neutrophil count comparedwith nonimmunized counterparts. This likely was the consequenceof the inflammatory events that are involved with the developingANCA disease process and thus would be analogous to observationsin patients who have ANCA disease and develop an elevated peripheralneutrophil count (25).
Engraftment with donor MPO-positive BM depended on the dosageof irradiation. Engraftment was more successful at higher irradiationdosages than at lower dosages (Table 1, Figure 2A). The efficacyof engraftment and the resultant proportion of peripheral bloodneutrophils that were MPO positive correlated with the inductionof glomerulonephritis and systemic vasculitis (Table 1). Allmice that received the highest irradiation dosages (900 or 750rad) and had approximately 90% circulating MPO-positive neutrophilsdeveloped glomerulonephritis, and most had evidence for systemicvasculitis, especially pulmonary capillaritis. A total of 71%of mice that received 600 rad and had approximately 50% circulatingMPO-positive neutrophils developed glomerulonephritis; however,when present, the glomerulonephritis was less severe as determinedby urinalysis and histopathology (Figure 3). None of the micethat received 450 rad and had only 16.4% circulating MPO-positiveneutrophils developed glomerulonephritis, although there wasa slight increase in glomerular staining for MPO (Table 3) thatprobably reflected slight subclinical neutrophil accumulation.
There was no difference in the anti-MPO antibody titer amongthe preimmunized experimental Mpo–/– mouse groupsthat received different dosages of radiation (Figure 2D). Thus,the differences in disease induction correlate directly withthe availability of MPO-positive neutrophils in the circulationto act as pathogenic targets for anti-MPO antibodies. Mpo–/–mice that were not preimmunized with MPO did not develop circulatinganti-MPO antibodies (Figure 2D) and did not develop glomerulonephritis(Figure 3) even though they had excellent engraftment with approximately90% MPO-positive peripheral blood neutrophils. This documentsthe requirement for anti-MPO antibodies for disease inductionin this model. The basis for this failure to mount an immuneresponse against the MPO in the engrafted BM is not known. Alikely possibility is that the irradiation caused so much ablationto the immune system that it did not recover during the experimentalinterval. Alternatively, sequestration of the MPO antigen withinneutrophils and monocytes and their precursors may have reducedimmunogenicity, although this is unlikely because of the degranulationthat occurred at the sites of inflammation. Another, even lesslikely possibility is tolerization of the reemerging immunesystem by exposure to high levels of MPO from the engraftedBM. The relatively stable level of anti-MPO antibodies in thepreimmunized mice (Figure 2D) suggests that these mice alsodid not developed an augmented response to the engrafted MPO-positivecells but rather that the anti-MPO antibodies were derived fromresidual circulating IgG, persistent plasma cells that survivedirradiation, or both. Future studies will investigate eventsat later time points and will include additional booster immunizationsin an attempt to perpetuate or augment the anti-MPO immune response.This may allow the development of a model of long-term ANCAdisease that will be a better mimic for human disease and abetter model for testing therapeutic regimens.
The most common hypothesis about anti-MPO antibody–inducedvasculitis is that the disease is caused by binding of ANCAto neutrophils, with subsequent activation leading to adherenceto and injury of vessel walls (2,3). Activation by ANCA of monocytes,which also contain MPO and PR3, may contribute to disease induction,but neutrophils seem to be the dominant effector cells (18).Some investigators, however, have raised the possibility thatother cells, such as endothelial cells, renal epithelial cells,or pulmonary alveolar cells, might produce ANCA antigens andcould be targets for pathogenic events (13–16). We usedtwo approaches to document the importance of circulating MPO-positiveneutrophils in the pathogenesis of ANCA and to rule out a rolefor other cell types. One approach generated circulating anti-MPOantibodies by active immunization with MPO, and the other approachintroduced circulating anti-MPO by passive intravenous injection.Both approaches caused pauci-immune NCGN in mice that had MPOonly in BM-derived cells. This demonstrated that MPO+/+ BM-derivedcells, primarily neutrophils, are sufficient pathogenic targetsfor anti-MPO to cause disease. In contrast, generation of circulatinganti-MPO antibodies either by active immunization with MPO orby passive injection of anti-MPO IgG did not cause glomerulonephritisin mice that had Mpo–/– BM-derived cells but otherwisehad Mpo+/+ tissue. This demonstrated that Mpo+/+ BM-derivedcells are required pathogenic targets for anti-MPO to causedisease.
These novel animal models further confirm that anti-MPO antibodiescause pauci-immune NCGN and systemic SVV. Furthermore, the dataindicate that MPO-positive neutrophils are sufficient and necessarypathogenic targets for the induction of glomerulonephritis andsystemic SVV by anti-MPO antibodies.
Acknowledgments
This research was supported by National Institute of Diabetesand Digestive and Kidney Diseases grant 2 P01 DK58335-06. A.S.was supported by a fellowship grant from the Deutsche Forschungsgemeinschaft(SCHR 771/1-1).
Some of the data in this article were published previously inabstract form (J Am Soc Nephrol 16: 51A, 2005; and Kidney BloodPress Res 28: 160–161, 2005).
We thank Dr. Bei Zhang and Jue Yao for expert technical assistance.
Footnotes
Published online ahead of print. Publication date availableat www.jasn.org.
Franssen CF, Stegeman CA, Kallenberg CG, Gans RO, De Jong PE, Hoorntje SJ, Tervaert JW: Antiproteinase 3- and antimyeloperoxidase-associated vasculitis.
Kidney Int 57
: 2195
–2206, 2000[CrossRef][Medline]
Morgan MD, Harper L, Williams J, Savage C: Anti-neutrophil cytoplasm-associated glomerulonephritis.
J Am Soc Nephrol 17
: 1224
–1234, 2006[Abstract/Free Full Text]
Jennette JC, Xiao H, Falk RJ: The pathogenesis of vascular inflammation by antineutrophil cytoplasmic antibodies.
J Am Soc Nephrol 17
: 1235
–1242, 2006[Abstract/Free Full Text]
Bansal PJ, Tobin MC: Neonatal microscopic polyangiitis secondary to transfer of maternal myeloperoxidase-antineutrophil cytoplasmic antibody resulting in neonatal pulmonary hemorrhage and renal involvement.
Ann Allergy Asthma Immunol 93
: 398
–401, 2004[Medline]
Schlieben DJ, Korbet SM, Kimura RE, Schwartz MM, Lewis EJ: Pulmonary-renal syndrome in a newborn with placental transmission of ANCAs.
Am J Kidney Dis 45
: 758
–761, 2005[CrossRef]
Csernok E, Ernst M, Schmitt W, Bainton DF, Gross WL: Activated neutrophils express proteinase 3 on their plasma membrane in vitro and in vivo.
Clin Exp Immunol 95
: 244
–250, 1994[Medline]
Muller Kobold AC, Kallenberg CG, Tervaert JW: Leukocyte membrane expression of proteinase 3 correlates with disease activity in patients with Wegener’s granulomatosis.
Br J Rheumatol 37
: 901
–907, 1998[Abstract/Free Full Text]
Schreiber A, Busjahn A, Luft FC, Kettritz R: Membrane expression of proteinase 3 is genetically determined.
J Am Soc Nephrol 14
: 68
–75, 2003[Abstract/Free Full Text]
Falk RJ, Terrell RS, Charles LA, Jennette JC: Anti-neutrophil cytoplasmic autoantibodies induce neutrophils to degranulate and produce oxygen radicals in vitro.
Proc Natl Acad Sci U S A 87
: 4115
–4119, 1990[Abstract/Free Full Text]
Mulder AH, Heeringa P, Brouwer E, Limburg PC, Kallenberg CG: Activation of granulocytes by anti-neutrophil cytoplasmic antibodies (ANCA): A Fc gamma RII-dependent process.
Clin Exp Immunol 98
: 270
–278, 1994[Medline]
Kettritz R, Jennette JC, Falk RJ: Crosslinking of ANCA-antigens stimulates superoxide release by human neutrophils.
J Am Soc Nephrol 8
: 386
–394, 1997[Abstract]
Williams JM, Ben-Smith A, Hewins P, Dove SK, Hughes P, McEwan R, Wakelam MJ, Savage CO: Activation of the G(i) heterotrimeric G protein by ANCA IgG F(ab')2 fragments is necessary but not sufficient to stimulate the recruitment of those downstream mediators used by intact ANCA IgG.
J Am Soc Nephrol 14
: 661
–669, 2003[Abstract/Free Full Text]
Mayet WJ, Csernok E, Szymkowiak C, Gross WL, Meyer zum Buschenfelde KH: Human endothelial cells express proteinase 3, the target antigen of anticytoplasmic antibodies in Wegener’s granulomatosis.
Blood 82
: 1221
–1229, 1993[Abstract/Free Full Text]
Pendergraft WF, Alcorta DA, Segelmark M, Yang JJ, Tuttle R, Jennette JC, Falk RJ, Preston GA: ANCA antigens, proteinase 3 and myeloperoxidase, are not expressed in endothelial cells.
Kidney Int 57
: 1981
–1990, 2000[CrossRef][Medline]
Brockmann H, Schwarting A, Kriegsmann J, Petrow P, Gaumann A, Muller KM, Galle PR, Mayet W: Proteinase-3 as the major autoantigen of c-ANCA is strongly expressed in lung tissue of patients with Wegener’s granulomatosis.
Arthritis Res 4
: 220
–225, 2002[CrossRef][Medline]
Hattar K, Grandel U, Bickenbach A, Schwarting A, Mayet WJ, Bux J, Jessen S, Fischer C, Seeger W, Grimminger F, Sibelius U: Interaction of antibodies to proteinase 3 (classic anti-neutrophil cytoplasmic antibody) with human renal tubular epithelial cells: Impact on signaling events and inflammatory mediator generation.
J Immunol 168
: 3057
–3064, 2002[Abstract/Free Full Text]
Xiao H, Heeringa P, Hu P, Liu Z, Zhao M, Aratani Y, Maeda N, Falk RJ, Jennette JC: Antineutrophil cytoplasmic autoantibodies specific for myeloperoxidase cause glomerulonephritis and vasculitis in mice.
J Clin Invest 110
: 955
–963, 2002[CrossRef][Medline]
Xiao H, Heeringa P, Liu Z, Huugen D, Hu P, Maeda N, Falk RJ, Jennette JC: The role of neutrophils in the induction of glomerulonephritis by anti-myeloperoxidase antibodies.
Am J Pathol 167
: 39
–45, 2005[Abstract/Free Full Text]
Huugen D, Xiao H, an Esch A, Falk RJ, Peutz-Kootstra CJ, Buurman WA, Cohen Tervaert JW, Jennette JC, Peter Heeringa P: Aggravation of anti-myeloperoxidase antibody induced glomerulonephritis by bacterial lipopolysaccharide: role of tumor necrosis factor alpha.
Am J Pathol 167
: 47
–58, 2005[Abstract/Free Full Text]
Little MA, Smyth CL, Yadav R, Ambrose L, Cook HT, Nourshargh S, Pusey CD: Anti-neutrophil cytoplasm antibodies directed against myeloperoxidase augment leukocyte-microvascular interactions in vivo.
Blood 106
: 2050
–2058, 2005[Abstract/Free Full Text]
Aratani Y, Koyama H, Nyui S, Suzuki K, Kura F, Maeda N: Severe impairment in early host defense against Candida albicans in mice deficient in myeloperoxidase.
Infect Immun 67
: 1828
–1836, 1999[Abstract/Free Full Text]
Jennette JC, Falk RJ: Small-vessel vasculitis.
N Engl J Med 337
: 1512
–1523, 1997[Free Full Text]
Jennette JC: Rapidly progressive and crescentic glomerulonephritis.
Kidney Int 63
: 1164
–1172, 2003[CrossRef][Medline]
Ruth AJ, Kitching AR, Kwan RY, Odobasic D, Ooi JD, Timoshanko JR, Hickey MJ, Holdsworth SR: Anti-neutrophil cytoplasmic antibodies and effector CD4+ cells play nonredundant roles in anti-myeloperoxidase crescentic glomerulonephritis.
J Am Soc Nephrol 17
: 1940
–1949, 2006[Abstract/Free Full Text]
D’Cruz D, Direskeneli H, Khamashta M, Hughes GR: Lymphocyte activation markers and von Willebrand factor antigen in Wegener’s granulomatosis: Potential markers for disease activity.
J Rheumatol 26
: 103
–109, 1999[Medline]
Related Article
This Month’s Highlights
J. Am. Soc. Nephrol. 2006 17: 3265-3266.
[Full Text][PDF]
This article has been cited by other articles:
D. O. Bunch, J. S. Silver, M. C. Majure, P. Sullivan, D. A. Alcorta, H. Chin, S. L. Hogan, Y. I. Lindstrom, S. H. Clarke, R. J. Falk, et al. Maintenance of Tolerance by Regulation of Anti-myeloperoxidase B Cells
J. Am. Soc. Nephrol.,
September 1, 2008;
19(9):
1763 - 1773.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
(19). In a different approach, severe NCGN that affects >80% of glomeruli and systemic SVV can be induced in immune-deficient mice by transfer of splenocytes from MPO-deficient mice that have been immunized with murine MPO (18). After anti-MPO splenocyte transfer, these mice developed increasing anti-MPO antibody titer. Pathologic examination after 13 d shows severe NCGN as well vasculitis in various organs. However, the glomerulonephritis is not pauci-immune because there is a moderate amount of predominantly mesangial glomerular deposition of immune complexes. Therefore, this is not a good model of pauci-immune human disease, which often has a small amount of glomerular Ig deposition but usually does not have the extent of deposition that is seen in mice after splenocyte transfer. 
-irradiated with different whole-body dosages that ranged from 450 to 900 rad (Table 1). BM cells were harvested from femurs and tibia, erythrocytes were lysed, and 1.5 x 107 BM cells were injected intravenously and retro-orbitally. Four different combinations of BM donors and recipients were used: WT into Mpo–/– (groups 1 through 5 and 9), Mpo–/– into WT (group 8), WT into WT (group 7), and Mpo–/– into Mpo–/– (group 6). Recovery of peripheral blood leukocytes and peripheral blood neutrophils was monitored after BM transplantation using the HESKA Veterinary Hematology System. Engraftment was measured as percentage of peripheral blood neutrophils that were positive for MPO by enzyme histochemistry. Extent of engraftment stabilized after 4 to 5 wk. Mice were killed for pathologic examination 8 wk after transplantation. 
). (A) Total leukocyte count. (B) Neutrophil count. There is a statistically significant difference in the neutrophil counts at 28 d (P < 0.01) and 35 d (P < 0.05).
