Apoptotic Cell–Induced, Antigen-Specific Immunoregulation to Treat Experimental Antimyeloperoxidase GN

Significance Statement

Current therapies for myeloperoxidase-ANCA–associated vasculitis (MPO-AAV), an autoimmune disease causing rapidly progressive GN, are nonspecific with considerable toxicities. Previous research defined the nephritogenic immunodominant myeloperoxidase (MPO) CD4⁺ T cell peptide, MPO_409–428, in a mouse model. In this study, the authors explored the therapeutic potential of generating endogenous MPO_409–428-specific regulatory T cells to achieve tolerance to MPO and regulate the anti-MPO autoimmune response driving GN. They created apoptotic MPO_409–428-conjugated splenocytes that were administered to mice before the animals had been immunized to MPO or after anti-MPO autoimmunity had been established. The resultant generation of antigen-specific type 1 regulatory T cells significantly attenuated GN. Defining the immunodominant MPO peptide(s) in patients with MPO-AAV offers the potential to restore tolerance to MPO via treatments focused on enhancing endogenous antigen-specific regulatory T cells.

Visual Abstract

Abstract

Background

Myeloperoxidase (MPO)-ANCA–associated GN is a significant cause of renal failure. Manipulating autoimmunity by inducing regulatory T cells is potentially a more specific and safer therapeutic option than conventional immunosuppression.

Methods

To generate MPO-specific regulatory T cells, we used a modified protein-conjugating compound, 1-ethyl-3-(3′dimethylaminopropyl)-carbodiimide (ECDI), to couple the immunodominant MPO peptide (MPO_409–428) or a control ovalbumin peptide (OVA_323–339) to splenocytes and induced apoptosis in the conjugated cells. We then administered MPO- and OVA-conjugated apoptotic splenocytes (MPO-Sps and OVA-Sps, respectively) to mice and compared their effects on development and severity of anti-MPO GN. We induced autoimmunity to MPO by immunizing mice with MPO in adjuvant; to trigger GN, we used low-dose antiglomerular basement membrane globulin, which transiently recruits neutrophils that deposit MPO in glomeruli. We also compared the effects of transferring CD4⁺ T cells from mice treated with MPO-Sp or OVA-Sp to recipient mice with established anti-MPO autoimmunity.

Results

MPO-Sp but not OVA-Sp administration increased MPO-specific, peripherally derived CD4⁺Foxp3⁻ type 1 regulatory T cells and reduced anti-MPO autoimmunity and GN. However, in mice depleted of regulatory T cells, MPO-Sp administration did not protect from anti-MPO autoimmunity or GN. Mice with established anti-MPO autoimmunity that received CD4⁺ T cells transferred from mice treated with MPO-Sp (but not CD4⁺ T cells transferred from mice treated with OVA-Sp) were protected from anti-MPO autoimmunity and GN, confirming the induction of therapeutic antigen-specific regulatory T cells.

Conclusions

These findings in a mouse model indicate that administering apoptotic splenocytes conjugated with the immunodominant MPO peptide suppresses anti-MPO GN by inducing antigen-specific tolerance.

Myeloperoxidase (MPO)-ANCA–associated GN is frequently severe and often results in end stage renal failure.¹ Despite current therapies, there is a 30% incidence of death or end stage renal failure over 3 years, with considerable therapeutic toxicities.² MPO-ANCA–associated vasculitis results from autoimmunity to MPO, a protein found in neutrophil lysosomes. Advances in the biologic manipulation of autoimmunity by inducing therapeutic T regulatory cells (Tregs) offer more specific, safer, therapeutic options.³ If the dominant disease-inducing autoantigen is known, it may be possible to induce antigen-specific therapeutic tolerance, avoiding unnecessary nonspecific immunosuppression.

An emerging strategy for inducing tolerance to autoantigens after the development of autoimmunity has been to re-present the autoantigenic immunodominant peptides into physiologic homeostatic pathways of apoptotic cell removal.⁴ This pathway allows the removal of senescent cells from tissues with high cell turnover (hematopoietic cells) without the risks of inflammation and potential autoimmunity associated with cell death by necrosis.⁵ Apoptotic cell death exposes cell components, such as phosphatidylserine, that induce the generation of immunomodulatory molecules, including TGF-β,⁶ IL-10,⁷ and Prostaglandin E2, by the phagocytes that engulf them.⁸ In the spleen, specialized phagocytes and immature tolerogenic dendritic cells induce tolerance by first inducing anergy and then generating a variety of Tregs.⁹ These homeostatic senescent cell-clearing pathways can be used to reinduce tolerance in established autoimmunity by coupling the dominant autoantigen to apoptotic cells. A major technical advance in achieving this goal was the use of a modified protein-conjugating compound, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (ECDI).¹⁰ ECDI concomitantly conjugates peptides to splenocytes and induces a high frequency of apoptosis in the conjugated cells. It has been demonstrated that administration of these cells induces antigen-specific tolerance in several experimental autoimmune diseases, including autoimmune encephalomyelitis,¹¹ arthritis,^12,13 and type 1 diabetes.¹⁴ ECDI peptide–conjugated cells have been safely administered to human subjects in phase 1/2 clinical trials.¹⁵

Patients with MPO-ANCA–associated GN have many characteristics that suggest that this technique may be applicable. Patients often present with sufficient residual renal function for prolonged renal sufficiency, and emerging evidence suggests that we have knowledge of the immunodominant peptides in humans¹⁶ and mice.¹⁷ In mice, we have defined the immunodominant MPO T cell peptide, MPO_409–428, and demonstrated that it induces nephritogenic autoimmunity, resulting in focal segmental necrotizing GN.¹⁷ The murine MPO epitope has substantial homology with a dominant disease-associated MPO peptide recognized by ANCA from patients with acute GN,¹⁶ making these murine studies highly relevant to humans. Moreover, we have provided proof of concept that this epitope, MPO_409–428, can induce tolerance to the whole MPO protein when re-presented into the tolerogenic upper airways (nasal tolerance).¹⁸ Nasal insufflation of MPO_409–428 to mice with established anti-MPO autoimmunity protected mice from the development of focal segmental necrotizing GN.¹⁸ In this context, we coupled MPO_409–428 to splenocytes and induced splenocyte apoptosis by ECDI in vitro. These ECDI myeloperoxidase-conjugated apoptotic splenocytes (MPO-Sps) were injected to promote the induction of tolerance to MPO in mice with anti-MPO autoimmunity and GN. The results of these studies showing that MPO-Sp can treat established anti-MPO autoimmunity and its associated GN in this model provide evidence suggesting that this method may be effective in the human disease.

Methods

Mice

C57BL/6 (wild-type [WT]) mice were purchased from Monash University Animal Services, and Foxp3-GFP mice were from Alexander Rudensky (University of Washington, Seattle, WA). Mice were housed/bred at Monash Medical Centre, and experiments were approved by the Monash Medical Centre Animal Ethics Committee.

Peptides and Reagents

Synthetic peptides MPO_409–428 (PRWNGEKLYQEARKIVGAMV) and ovalbumin_323–339 (OVA_323–339; ISQAVHAAHAEINEAGR) were purchased from Mimotopes (Victoria, Australia). Recombinant murine MPO (rMPO) was generated using a baculovirus system as previously described.¹⁹

Experimental Design

Anti-MPO GN

To induce autoimmune anti-MPO GN, 6- to 8-week-old mice were immunized on day 0 with 100 μg of MPO_409–428 in Freund complete adjuvant (Sigma-Aldrich) intraperitoneally. On day 7, mice received a booster injection of 100 μg of MPO_409–428 in Freund incomplete adjuvant intraperitoneally (Sigma-Aldrich). On days 16 and 17, mice were administered an intravenous injection of 1.5 mg of sheep anti-mouse glomerular basement membrane globulin and humanely killed 4 days later.²⁰ For studies assessing immune responses, mice were immunized on day 0 with 100 μg MPO_409–428 in Freund complete adjuvant, and they were culled 10 days later.

Antigen-Coupled Cell Tolerance

Tolerance was induced by intravenous injections of chemically ECDI-treated MPO-Sps or ovalbumin-coupled splenocytes (OVA-Sps). Donor spleens from naïve female mice were removed, and red blood cells were lysed using lysis buffer. The splenocytes were incubated with ECDI (Calbiochem; 46.88 mg) and MPO_409–428 (or OVA_323–339; 200 μg/1×10⁸ cells) and shaken on ice for 1 hour (Figure 1). Cells were washed three times and filtered through a 70 μM cell strainer to remove cell debris. On the basis of published literature, MPO-Sps or OVA-Sps were resuspended in sterile saline, and each mouse received 5×10⁷ MPO-Sps or OVA-Sps in 200 μl of PBS intravenously 7 days before induction of MPO autoimmunity (Figure 1).²¹ Following this protocol, ECDI induced apoptosis in >70% of isolated naïve splenocytes within 4 hours (Supplemental Figure 1).

Figure 1.

Schematic illustration of the induction of chemically treated, antigen-coupled splenocytes which enhances T regulatory cell (Treg)-mediated immunoregulation. Donor spleens were harvested, and splenocytes were isolated. Splenocytes were incubated with 1-ethyl-3-(3′dimethylaminopropyl)-carbodiimide (ECDI) and the nephritogenic myeloperoxidase (MPO) T cell peptide, MPO_409–428. To enhance Treg numbers and function, 5×10⁷ myeloperoxidase-conjugated apoptotic splenocytes (MPO-Sps) were administered to mice 7 days before MPO in Freund complete adjuvant (FCA) immunization. MPO-Sp treatment has the capacity to prevent the development of anti-MPO GN and treat mice with established anti-MPO autoimmunity. pMPO, peptide MPO; Teff, T effector cells.

Adoptive Transfer of CD4⁺ or CD4⁺Foxp3⁺ Cells from ECDI-Conjugated Apoptotic Splenocytes

WT mice were administered MPO-Sp or OVA-Sp, and spleens were harvested 5 days later. CD4⁺ T cells were isolated by magnetic cell sorting according to the manufacturer’s instructions (MACS CD4⁺ T cells isolation kit; Miltenyi Biotec); 5×10⁶ CD4⁺ T cells were injected into mice intravenously on day 14 of the anti-MPO GN model (after established autoimmunity).

To obtain CD4⁺Foxp3⁺ cells, Foxp3-GFP mice were treated with MPO-Sp or OVA-Sp, and spleens were harvested 5 days later. CD45⁺ cells were isolated using mouse CD45 microbeads (Miltenyi Biotec) and sorted for the Foxp3-GFP⁺ cells on the FACS Aria Fusion Cell Sorter (BD Biosciences); 1×10⁶ Foxp3⁺ Tregs were injected intravenously to donor mice with established anti-MPO autoimmunity (day 14).

In Vivo Depletion of Tregs

Intraperitoneally, 1.5 mg rat anti-mouse CD25 mAb (clone PC61)²² was administered on days −4 and −2 (n=8). On day 0, CD25-depleted mice received 5×10⁷ MPO-Sps, and 7 days later, autoimmunity to MPO was induced as described above. A group of mice received the isotype control: rat IgG1. The degree of anti-CD25 mAb depletion/blocking in blood and spleen cells was determined by using CD4 (RM4–4; BD Biosciences), Foxp3 (FJK-16s, eFluor 450; eBioscience), CD8a (53–6.7, PE; Biolegend), and NK1.1 (PK136, APC/Cy7; Biolegend). Samples were analyzed on the Navios flow cytometer, and data were analyzed using FlowJo software.

Ex Vivo Assessment of MPO-Sp’s Capacity to Enhance Treg Proliferation

MPO-Sps or OVA-Sps were administered to WT mice 7 days before they were MPO_409–428 immunized in Freund adjuvant, and 10 days later, spleens were harvested. Splenocytes were stained with CellTrace Violet dye (CellTrace Violet Cell Proliferation Kit; Life Technologies), stimulated with 10 μg/ml rMPO, and seeded at 5×10⁵ cells per milliliter for 72 hours; 10 μg/ml Brefeldin A (Sigma-Aldrich) was added during the last 8 hours of culture. After 72 hours, cells were surface stained with anti-CD4 (GK1.5, APC/H7; BD Biosciences) in 1% BSA/PBS for 30 minutes at 4°C, and then, they were fixed and permeabilized with fixation/permeabilization buffer as per the manufacturer’s protocol (eBioscience) before labeling with anti-mouse IL-10 (JES5–16E3, Alexa 647; eBiosciences), anti-mouse Foxp3 (FJK-16s, APC; eBiosciences), anti-mouse IL-17a (eBio17b1, FITC; eBiosciences), and anti-mouse IFN-γ (XMG1.2, PE; BD Biosciences).

Characterization of CD4⁺ T Cells from ECDI-Conjugated Apoptotic Splenocytes

CD4⁺ T cells from MPO-Sp– or OVA-Sp–treated mice were stimulated ex vivo with rMPO in supplemented RPMI media (10% FCS, 2 mM l-glutamine, 2-ME, 100 U/ml penicillin, and 0.1 mg/ml streptomycin; Sigma-Aldrich) to determine if there is a change in the proportion of T effector (Teff) cells and Tregs. Cells were stained with CD4, IFN-γ, IL-17A, IL-10, and Foxp3. To determine proliferation, cells were labeled with CTV before culture. Cells were cultured for 72 hours in 96-well plates.

No difference was observed in the proportion of either Tregs (type 1 regulatory T [Tr1] cells: OVA-Sp 0.46%±0.04% versus MPO-Sp 0.31%±0.06% IL-10⁺/CD4⁺Foxp3⁻CTV^lo, P=0.07 or peripheral Tregs: OVA-Sp 0.065%±0.02% versus MPO-Sp 0.12%±0.03% IL-10⁺/CD4⁺Foxp3⁺CTV^lo, P=0.12) or effector T cell subsets (Th1: OVA-Sp 4.82%±0.4% versus MPO-Sp 4.55%±0.6% IFN-γ⁺/CD4⁺CTV^lo, P=0.73 or Th17: OVA-Sp 0.48%±0.09% versus MPO-Sp 0.61%±0.06% IL-17A⁺/CD4⁺CTV^lo, P=0.21) between groups (n=7 per group).

Assessment of Renal Disease

Glomerular pathology was assessed on formalin-fixed, paraffin-embedded 3-μm-thick sections stained with periodic acid–Schiff reagent. A minimum of 30 glomeruli per mouse were assessed and scored for glomerular segmental necrosis.²³ Albuminuria was assessed by housing mice in individual metabolic cages to collect urine over 24 hours before the end of the experiment. Urinary protein concentration was measured by ELISA using an albuminuria kit (Bethyl Laboratories). Glomerular leukocytes (neutrophils, macrophages, and CD4⁺ T cells) were assessed by staining immunoperoxidase-stained periodate lysine paraformaldehyde frozen 6 μm cryostat-cut kidney sections. The primary mAbs used were RB6–8C5 for neutrophils (anti-GR-1; DNAX), CD68 for macrophages (FA/11 from Gordon L. Koch, Cambridge, England), and GK1.5 for CD4⁺ T cells (anti-CD4⁺; American Type Culture Collection). The secondary antibody used was rabbit anti-rat biotin (BD Bioscience) followed by 3,3′-diaminobenzidine tetrachloride (Sigma-Aldrich). A minimum of 30 glomeruli were scored per mouse, and results were expressed as positive cells per glomerular cross-section.

Assessment of Systemic Responses to MPO

Delayed-Type Hypersensitivity Responses to MPO

Mice were challenged with 6 μg of rMPO in 30 μl in the left hind footpad, whereas the contralateral footpad received saline 24 hours before termination of the experiment. Delayed-type hypersensitivity (DTH) was quantified by measuring the difference between footpad thickness (Δmillimeter) using a micrometer.

Serum MPO-ANCA IgG (ELISA)

Next, 96-well polystyerene microplates (Invitrogen Technologies) were coated with 5 μg/ml recombinant murine MPO, and sera were added in a 1:100 serial dilution. Detection was by horseradish peroxidase–conjugated sheep anti-mouse IgG (1:2000; Amersham Biosciences).

IFN-γ Secretion Detected by ELISPOT

Splenocytes were cultured for 18 hours at 5×10⁵ cells per well with OVA_323–339 or MPO_409–428 (10 μg/ml) in supplemented RPMI media. Spots were developed according to the manufacturer’s instructions and enumerated by an automated ELISPOT reader ELR06 (AID), and results are expressed as the number of IFN-γ⁺ cells per spleen.

Results

Apoptotic MPO-Sp Induces Antigen-Specific Attenuation of the Development of Anti-MPO Autoimmunity and GN

The effect of administering MPO-Sp intravenously on the development of experimental autoimmune anti-MPO GN was compared with control ECDI-treated splenocytes conjugated with an irrelevant peptide, OVA_323–339 (OVA-Sp). Apoptotic splenocytes were injected before the induction of anti-MPO autoimmunity by immunization and boosting of WT mice with peptide MPO_409–428 in Freund’s adjuvant. Compared with control OVA-Sp–treated mice, MPO-Sp significantly attenuated cell-mediated anti-MPO autoimmunity with reduced MPO-induced skin DTH recall responses but had no effect on the MPO-ANCA titers (Figure 2, A and B). The extent of GN induced by the anti-MPO autoimmunity was significantly less in MPO-Sp–treated mice than in control OVA-Sp–treated mice. Pathologic albuminuria was attenuated, there was less frequent glomerular segmental necrosis, and there were fewer infiltrating glomerular effector leukocytes: CD4⁺ T cells, macrophages, and neutrophils (Figure 2, C–E). When MPO-Sps were administered to mice with established anti-MPO autoimmunity (10 days after immunization before inducing GN), there were significant reductions in MPO-specific DTH responses, albuminuria, glomerular leukocyte infiltration, and the frequency of glomerular segmental necrosis (Figure 2H) compared with untreated WT mice. Comparisons between OVA-Sp and MPO-Sp treatments of established autoimmunity showed that MPO-Sp treatment was also more effective in the attenuation of DTH (0.14±0.02 versus 0.05±0.01 ∆mm, P<0.01), glomerular segmental necrosis (44%±5% versus 26%±4%, P<0.05), and glomerular CD4⁺ T cells (0.25±0.03 versus 0.13±0.02 positive cells per glomerular cross-section, P<0.01). Collectively, these data suggest that treatment with apoptotic splenocytes conjugated to the immunodominant nephritogenic peptide, MPO_409–428, can suppress anti-MPO autoimmunity and GN.

Figure 2.

Administration of myeloperoxidase-conjugated apoptotic splenocytes (MPO-Sps) prevents the development of and treats established antimyeloperoxidase (anti-MPO) autoimmunity and GN. (A) Administration of MPO-Sp before the development of anti-MPO autoimmunity (day −7, n=8) significantly reduced cell-mediated MPO-specific delayed-type hypersensitivity (DTH) responses compared with control ovalbumin-coupled splenocyte (OVA-Sp)–treated mice (n=8). (B) Preventative MPO-Sp treatment did not affect humoral MPO-ANCA IgG titers. (C and D) Renal injury assessed by albuminuria and percentage glomerular segmental necrosis was significantly attenuated in MPO-Sp–treated mice. (E) Glomerular leukocyte accumulation of CD4⁺ T cells, macrophages (Mɸ), and neutrophils (PMN) was all reduced in an antigen-specific manner when administered with MPO-Sp. Administration of MPO-Sp in mice with established anti-MPO autoimmunity (day 10, n=8) reduced (F) MPO-specific DTH responses and (G) albuminuria compared with the untreated wild type (WT) mice with established anti-MPO autoimmunity (n=8). (H) Structural injury assessed by percentage glomerular segmental necrosis was significantly lower in MPO-Sp–treated mice. (I) Reduced numbers of glomerular CD4⁺ T cells, macrophages, and neutrophils were observed in MPO-Sp–treated mice compared with controls. Error bars represent mean ± SEM with statistical analysis by unpaired t test. c/gcs, cells per glomerular cross-section; FCA, Freund complete adjuvant; FIA, Freund incomplete adjuvant; GBM, glomerular basement membrane. *P<0.05; **P<0.01; ***P<0.001.

Administration of MPO-Sp Re-Establishes Tolerance to MPO via the Generation of Tregs

MPO-Sps or control OVA-Sps were administered 7 days before MPO_409–428 immunization, and the generation of anti-MPO Treg immunomodulation and Teff responses was assessed ex vivo 10 days later. Detectable MPO- and OVA-specific ex vivo immune responses were confirmed by measuring the proportion of splenocyte cells producing IFN-γ by ELISPOT in response to OVA or MPO immunization (Supplemental Figure 2). MPO-Sp treatment increased the frequency of MPO-specific Tr1 cells that are CD4⁺Foxp3⁻IL-10⁺ (Figure 3A) compared with in OVA-Sp–treated and nontreated (saline) mice. Increased numbers of peripheral Tregs, phenotyped as CD4⁺Foxp3⁺IL-10⁺, were also observed in MPO-Sp–treated mice compared with untreated mice; however, no difference was observed with OVA-Sp–treated mice (Figure 3B). MPO-SP treatment increased the frequency of Tregs and reduced the frequency of antigen-specific IL-17A and IFN-γ producing CD4⁺ Teff cells compared with untreated mice. Although OVA-Sp–treated mice did not reduce Th17 cells, there was a significant reduction in the proportion of Th1- and IFN-γ–producing cells (Figure 3, C and D).

Figure 3.

The 1-ethyl-3-(3′dimethylaminopropyl)-carbodiimide peptide–conjugated splenocytes enhance myeloperoxidase (MPO)-specific immunoregulation. Naïve mice were treated with saline, ovalbumin-coupled splenocyte (OVA-Sp), or myeloperoxidase-conjugated apoptotic splenocyte (MPO-Sp; n=5, 8, and 8, respectively) 7 days before MPO_409–428 immunization, anti-MPO–specific immune responses were measured in spleens 10 days later. (A) MPO-Sp–treated mice significantly enhanced the proportion of splenic-induced type 1 T regulatory (Tr1) cells (CD4⁺Foxp3⁻IL-10⁺) compared with both untreated and OVA-Sp–treated mice. (B) Expansion of peripheral T regulatory cells (pTregs; CD4⁺Foxp3⁺IL-10⁺) was observed between MPO-Sp–treated mice and untreated controls. (C and D) MPO-specific proliferation of both CD4⁺IL-17A (Th17) and CD4⁺IFN-γ (Th1) cells was significantly reduced in MPO-SP–treated mice compared with untreated controls, whereas OVA-Sp–treated mice only suppressed Th1 responses. Error bars represent mean ± SEM with statistical analysis by one-way ANOVA. CTV, cell trace violet. *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001.

Therapeutic MPO-Sp Immunomodulation Is Dependent on Tregs

Confirmation that MPO-Sp induces Treg immunomodulation was achieved using mouse anti-CD25 monoclonal antibody (mAb) to deplete/block the suppressive functionality of Tregs before the induction of anti-MPO GN (days −4 to −2). Anti-CD25 mAb treatment significantly reduced the numbers of CD4⁺Foxp3⁺ Tregs in the blood and spleen 7 and 30 days after, respectively. CD25 depletion only significantly reduced the proportion of Tregs (CD4⁺Foxp3⁺) in blood and spleen. No differences in CD4⁺, CD8⁺, and NK⁺ cells were observed (Supplemental Figure 3, A–H).

Compared with untreated diseased mice, we assessed the effect of CD25 cell depletion on the efficacy of MPO-Sp treatment. In mice given isotype rat IgG1 (as control for anti-CD25 mAb), MPO-Sp significantly reduced cell-mediated anti-MPO autoimmunity (MPO-specific dermal DTH responses) compared with in untreated diseased mice, whereas MPO-Sp administration to mice injected with anti-CD25 mAb did not result in reduced MPO-specific DTH (Figure 4A). Similarly, MPO-Sp treated with concomitant administration of anti-CD25 mAb did not protect from glomerular injury (glomerular leukocyte accumulation, glomerular focal and segmental necrosis, and albuminuria) by MPO-Sp treatment (Figure 4, B–F). Additionally, administration of anti-CD25 mAb to mice not receiving MPO-Sp treatment caused the development of similar cellular anti-MPO autoimmunity and GN to those in mice given isotype rat IgG1 (Supplemental Figure 3, I–O). Therefore, MPO-Sp treatment is ineffective in the absence of Tregs.

Figure 4.

T regulatory cells (Tregs) are essential in the induction of myeloperoxidase-conjugated apoptotic splenocyte (MPO-Sp)–specific immunoregulation. We induced antimyeloperoxidase (anti-MPO) GN in mice given anti-CD25 mAb known to inhibit Treg capacity or treated mice with a control rat IgG1. (A) Pretreatment with anti-CD25 mAb (gray bars) abrogated the capacity of MPO-Sp (black bars) to block MPO-specific delayed-type hypersensitivity (DTH) response and was similar to controls (checked bars; n=8). (B–D) Coadministration of anti-CD25 mAb and MPO-Sp increased the numbers of infiltrating glomerular leukocytes compared with rat IgG1 and MPO-Sp–treated mice. (E) MPO-Sp–treated mice significantly attenuated the proportion of glomerular segmental necrosis and were abrogated by anti-CD25 mAb treatment. (F) No difference in albuminuria was observed between groups. Error bars represent mean ± SEM with statistical analysis by one-way ANOVA. c/gcs, cells per glomerular cross-section; FCA, Freund’s complete adjuvant; FIA, Freund’s incomplete adjuvant; GBM, glomerular basement membrane. *P<0.05; **P<0.01; ***P<0.001.

Adoptive Transfer of CD4⁺ T Cells from Mice Treated with MPO-Sp to Recipient Mice with Established Anti-MPO Autoimmunity Is Therapeutic

To evaluate the capacity of MPO-Sp to induce antigen-specific tolerance mediated by CD4⁺ Tregs in the face of established anti-MPO autoimmunity, we transferred splenic CD4⁺ T cells from MPO-Sp– and control OVA-Sp–treated mice to mice with established anti-MPO autoimmunity before triggering GN (day 14). Transferred CD4⁺ cells from MPO-Sp–treated mice induced significant suppression of DTH responses to MPO and MPO-ANCA titers in mice with established anti-MPO autoimmunity compared with mice receiving OVA-Sp–treated CD4⁺ T cells (Figure 5, A and B). Glomerular injury was significantly reduced in mice that received CD4⁺ T cells from mice treated with MPO-Sp compared with GN in control mice receiving CD4⁺ T cells from mice treated with OVA-Sp (Figure 5, C–G). This is consistent with the hypothesis that the mice were protected by induced antigen-specific tolerance.

Figure 5.

Adoptive transfer of CD4⁺ T cells from myeloperoxidase-conjugated apoptotic splenocyte (MPO-Sp)-treated mice to recipient mice with established anti-MPO autoimmunity is therapeutic. (A and B) Administration of CD4⁺ T cells treated with MPO-Sp to mice with established anti-MPO autoimmunity significantly reduced MPO-specific delayed-type hypersensitivity (DTH) responses and MPO-ANCA IgG production compared with mice receiving CD4⁺ T cells from OVA-Sp–treated mice (n=8 mice per group). (C–E) Infiltrating glomerular CD4⁺ T cells, macrophages, and neutrophils were significantly reduced in mice receiving CD4⁺ T cells from MPO-Sp–treated mice. (F) Glomerular segmental necrosis was significantly reduced in mice receiving CD4⁺ T cells from MPO-SP–treated mice compared with controls. (G) No significant difference in albuminuria was observed between groups. Error bars represent mean ± SEM with statistical analysis by unpaired t test. c/gcs, cells per glomerular cross-section; FCA, Freund’s complete adjuvant; FIA, Freund’s incomplete adjuvant; GBM, glomerular basement membrane. *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001.

To determine if Foxp3⁺ Tregs alone from MPO-Sp–treated mice are responsible for the observed adoptive transfer of tolerance in mice with established anti-MPO autoimmunity, splenic CD4⁺Foxp3⁺ Tregs were isolated from MPO-Sp– or OVA-Sp–treated mice and transferred to mice with established anti-MPO autoimmunity. Transfer of Foxp3⁺ Tregs from MPO-Sp–treated mice significantly reduced anti-MPO autoimmune DTH responses; however, these cells did not improve structural and functional kidney injury (Supplemental Figure 4).

Discussion

These studies build on a growing body of evidence that apoptotic cells are anti-inflammatory and immunomodulatory.^4,24 The spleen plays a major role in homeostatic senescent cell clearance.^21,25 Splenic macrophages and dendritic cells have surface receptors that recognize phosphatidylserine-rich remnant peptides of apoptotic cells. Phagocytosis of apoptotic cell generates mediators that favor anergy and the generation of Tregs.^26,27 This pathway can be co-opted for therapeutic purposes. By conjugating dominant autoantigenic peptides to splenocytes with ECDI ex vivo while concurrently inducing apoptosis, it has been possible to create a “Trojan horse” that can reintroduce autoantigenic peptides into this homeostatic pathway to induce antigen-specific tolerance even in the setting of established autoimmunity.^21,28 The safety of this technique was demonstrated in human subjects in the treatment of multiple sclerosis.¹⁵

The capacity to determine the immunodominant nephritic MPO peptide (MPO_409–428) in mice has allowed for the testing of this method of inducing antigen-specific tolerance in this relevant model of anti-MPO GN. The relevance to humans comes from the pathologic (development of focal and segmental necrotizing GN) and immunologic (development of ANCA staining neutrophils in a pANCA pattern and the homology of human and mice MPO epitopes) similarities of the murine model to the human disease. In this study, we demonstrated that MPO-Sp prevented the development of experimental anti-MPO GN and re-established tolerance to MPO in mice with established anti-MPO autoimmunity. Administration of ECDI-conjugated apoptotic splenocytes is more effective in attenuating anti-MPO autoimmunity and GN when conjugated to the MPO_409–428 compared with conjugating with a control irrelevant antigen, OVA_323–339. Compared with mice treated with OVA-Sp, administration of MPO-Sp before induction of MPO immunization revealed that MPO-Sp treatment results in the expansion of Tregs, specifically Tr1 cells. However, both MPO-Sp and OVA-Sp reduced anti-MPO–specific Th1 and Th17 cells, revealing that administration of apoptotic splenocytes induces a component of nonspecific immunomodulation. This is not surprising, because harnessing the apoptotic senescent pathway by therapeutic intravenous administration of apoptotic cells has been demonstrated to be effective in the treatment of numerous autoimmune diseases.²⁹ Nevertheless, we have demonstrated proof of concept that re-presentation of the nephritogenic MPO peptide through the apoptotic senescent pathway induces more effective, prolonged, and specific tolerance to mice with anti-MPO GN.

Tolerance induction by means of administration of antigen-specific ECDI-treated splenocytes has been demonstrated to be dependent on CD4⁺CD25⁺ Tregs.¹⁴ In this study, we used an anti-CD25 mAb around the time of preventative MPO-Sp administration to specifically deplete CD4⁺Foxp3⁺ Tregs. In the absence of CD4⁺Foxp3⁺ Tregs, MPO-Sp was ineffective in suppressing anti-MPO autoimmunity and GN. Because MPO-Sp treatment expands both Tr1 cells and peripheral Tregs, we then adoptively transferred CD4⁺ T cells from mice receiving either MPO-Sp or OVA-Sp to donor mice with established anti-MPO autoimmunity. Only CD4⁺ T cells from MPO-Sp–treated mice had the capacity to reduce anti-MPO autoimmunity and GN. Transfer of peripheral Tregs (CD4⁺Foxp3⁺) after MPO-Sp or OVA-Sp treatment, which excludes Tr1 cells (CD4⁺Foxp3⁻), reduced anti-MPO T cell autoimmunity but did not reduce functional or histologic glomerular injury.

Collectively, these studies demonstrate that MPO-Sps induce antigen-specific immunoregulation via Tregs capable of attenuating anti-MPO autoimmunity and nephritogenic injury. This discovery provides proof of concept that administration of peptide-conjugated apoptotic cells may be therapeutic in patients with MPO-ANCA GN.

Disclosures

Dr. Kitching reports grants from the National Health and Medical Research Council of Australia during the conduct of the study, personal fees from CSL Limited, and personal fees from Ablynx outside the submitted work. All of the remaining authors have nothing to disclose.

Funding

These studies were supported by National Health and Medical Research Council of Australia grant 1147388.

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2018090955/-/DCSupplemental.

Supplemental Figure 1. ECDI induces >70% splenocyte apoptosis.

Supplemental Figure 2. Splenic antigen-specific recall immune responses.

Supplemental Figure 3. Anti-CD25 mAb depletes Tregs, and pretreatment of anti-CD25 mAb does not aggravate anti-MPO GN.

Supplemental Figure 4. Adoptive transfer of CD4⁺Foxp3⁺ Tregs from MPO-SP–treated mice does not transfer tolerance to recipient mice with established anti-MPO autoimmunity.