A Modified Peptide Derived from Goodpasture Autoantigen Arrested and Attenuated Kidney Injuries in a Rat Model of Anti-GBM Glomerulonephritis

Significance Statement

The main target antigen of autoantibodies against glomerular basement membrane (GBM) in Goodpasture disease is the noncollagenous domain 1 of the α3 chain (α3NC1) of type IV collagen. The authors previously identified a nephritogenic peptide, P14 (α3_127–148). In this study, they designed a modified peptide with one amino acid substitution in its core motif, replacing a pathogenic residue with a nonpathogenic one. Administering this modified peptide to rats with α3-P14–induced anti-GBM GN reduced inflammatory responses and crescent formation in the kidneys through inhibition of α3-P14 binding to antibodies and MHC II molecules, as well as through modulation of T cells, including inhibiting α3-P14–specific T cell proliferation and abating Th17 cell differentiation. This peptide modification approach may offer insights into treating Goodpasture disease and other autoimmune kidney disorders.

Visual Abstract

Abstract

Background

In Goodpasture disease, the noncollagenous domain 1 of the α3 chain (α3NC1) of type IV collagen is the main target antigen of antibodies against glomerular basement membrane (GBM). We previously identified a nephritogenic epitope, P14 (α3_127–148), that could induce crescentic nephritis in WKY rats, and defined its core motif. Designing a modified peptide, replacing critical pathogenic residues with nonpathogenic ones (on the basis of homologous regions in α1NC1 chain of type IV collagen, known to be nonpathogenic), might provide a therapeutic option for anti-GBM GN.

Methods

We synthesized a modified peptide, replacing a single amino acid, and injected it into α3-P14–immunized rats from day 0 (the early-treatment group) or a later-treatment group (from days 17 to 21). A scrambled peptide administrated with the same protocol served as a control.

Results

The modified peptide, but not the scrambled peptide, attenuated anti-GBM GN in both treatment groups, and halted further crescent formation even after disease onset. Kidneys from the modified peptide–treated rats exhibited reductions in IgG deposits, complement activation, and infiltration by T cells and macrophages. Treatment also resulted in an anti-inflammatory cytokine profile versus a proinflammatory profile for animals not receiving the modified peptide; it also reduced α3-P14–specific T cell activation, modulated T cell differentiation by decreasing Th17 cells and enhancing the ratio of Treg/Th17 cells, and inhibited binding of α3-P14 to antibodies and MHC II molecules.

Conclusions

A modified peptide involving alteration of a critical motif in a nephritogenic T cell epitope alleviated anti-GBM GN in a rat model. Our findings may provide insights into an immunotherapeutic approach for autoimmune kidney disorders such as Goodpasture disease.

Goodpasture disease, also known as anti-glomerular basement membrane (GBM) disease, is an autoimmune disorder characterized by the presence of anti-GBM autoantibodies, rapidly progressive glomerulonephritis (GN), and a high risk of pulmonary hemorrhage. The noncollagenous domain 1 of the α3 chain of type IV collagen [α3(IV)NC1] has been identified as the main target antigen of anti-GBM antibodies.^1,2 Plasma exchange, cyclophosphamide, and prednisolone arrests lung hemorrhage, but the kidney recovery always remains partial or absent because of the advancement of kidney destruction.¹ Alternative or additional therapies are necessary and the treatment would be improved by targeting the antigen-specific immune response.^3–5

Human GBM consists of five α chains of type IV collagen. Epitope mapping studies have defined major conformational epitopes of α3(IV)NC1 as E_A, α3_17–31, and E_B, α3_127–141.² In addition to reactivity to α3NC1, which is detectable in almost all patients, distinct autoantibody targeting α5NC1 has been recently found in circulating and in kidney-bound form in patients with Goodpasture disease, with the fourth conformational epitope within E_B region of α5NC1.⁶ The residues pattern presented by E_A/E_B regions of α3NC1 and α5NC1 differ from homologous regions in α1NC1, which has been proven to be nonpathogenic.^6–8 E_A and E_B epitope residues are not conserved among all human NC1 domains. These nonconserved residues are essential to the pathogenicity of these epitopes. On the other hand, these pathogenic critical amino acids give us unprecedented opportunities to design modified peptides for specifically blocking the pathogenic counterparts and developing potential immunotherapies.

Previous studies showed compelling evidence of T cell involvement in anti-GBM nephritis, when antigen-specific CD4⁺ T cells per se could initiate kidney injury^9,10 and a T cell epitope pCol(28–40) could induce crescentic nephritis.^11,12 Our previous study identified a nephrogenic linear peptide (P14, α3_127–148) that contains the epitope recognized by both T cells and B cells, and includes the conformational epitope E_B.^13,14 By sequential amino acid substitution, we found that tryptophan (W₁₃₆), isoleucine (I₁₃₇), leucine (L₁₃₉), and tryptophan (W₁₄₀) were crucial for inducing anti-GBM GN in Wistar Kyoto (WKY) rats.¹⁵ The immunodominant epitope α3_135–145 is presented by HLA-DRB1*1501, the predisposing allele for Goodpasture disease,^16,17 with the anchor residues I₁₃₇, W₁₄₀, phenylalanine (F₁₄₃), and F₁₄₅ in the binding pocket 1, 4, 7, 9 of DR15 molecule.¹⁸ T cells specific for this epitope from DR15⁺ transgenic mice were largely CD4⁺ Foxp3⁻ T_conv cells and produced the pathogenic T helper cell (Th) Th1 and Th17 cytokines.¹⁸ Thus, the core residue motif on P14 was W₁₃₆I₁₃₇_L₁₃₉W₁₄₀_F₁₄₃_F₁₄₅.

The clearly characterized motif of P14 enabled us to design modified peptides to inhibit its pathogenic process. In this study, we aligned the primary sequences of P14 and its counterpart on α1NC1 to identify the discrepancy and designed a modified peptide (m-P14) using the nonpathogenic amino acid residue on α1NC1 for substitution. We found that this modified peptide could arrest anti-GBM GN in WKY rats and further explored its therapeutic mechanisms.

Methods

Synthesis of Peptides

All peptides (Figure 1) were synthesized on an automatic peptide synthesizer using F-moc chemistry (Beijing Scilight Biotechnology Ltd. Co., Beijing, China), and purified by reverse-phase CIS column on a preparative HPLC. Purified peptides were analyzed by HPLC for purity and mass spectrometry for correct sequence. Peptides with purity over 98% were used for further tests.

Figure 1.

The modified peptide was designed by replacing α3-I137 with α1-S137. Critical amino acids for α3-P14 (W₁₃₆I₁₃₇_L₁₃₉W₁₄₀_G₁₄₂F₁₄₃_F₁₄₅) are indicated by red arrows. Among them, the amino acids at 137 position and 143 position were different between α3-P14 and α1-P14. The substitution of α3-I₁₃₇ with α1-S₁₃₇ was designed for the modified peptide (m-P14). s-P14 was designed as control for m-P14 with the same amino acid residues but in scrambled sequence.

Induction and Treatment of Anti-GBM GN in WKY Rats

Animals and Immunization

Female WKY rats, aged 4 weeks, were immunized using α3-P14, α1-P14, or m-P14 at 200 μg/kg or m-P14 at 30 mg/kg by hind footpad injection, emulsified in complete Freund adjuvant (Sigma-Aldrich, St. Louis, MO). Twenty-four hour urine samples and blood samples were collected before and after immunization at each week. All rats were euthanized at the end of week 6 after immunization according to our previous experience,¹⁵ after which, renal tissues, spleens, and blood samples were collected.

Treatment with m-P14

For early-treatment groups, m-P14 was injected into rats intraperitoneally at 30 mg/kg (early-treatment group A) or 10 mg/kg (early-treatment group B) daily from day 0 to 14, and every other day from day 15 to 28. For the later-treatment group, m-P14 at 30 mg/kg was injected intraperitoneally into rats upon the detection of hematuria or proteinuria by urine dipsticks (Urit, Guilin, China) (approximately 3 weeks after immunization). Sterile PBS solution was used as negative control. The designs for early-treatment and later-treatment interventions are summarized in Figure 2.

Figure 2.

m-P14 interventions were applied in experimental anti-GBM GN. All of the WKY rats were immunized with 200 μg/kg α3-P14, administered subcutaneously at hind footpads at day 0. For early-treatment groups, m-P14 or s-P14 were injected intraperitoneally at 30 mg/kg (early-treatment group A) or 10 mg/kg (early-treatment group B) daily from day 0 to 14, and every other day from day 15 to 28. Then the rats were observed for another 2 weeks. For the later-treatment group, m-P14 or s-P14 at 30 mg/kg were injected intraperitoneally into rats every day, once the hematuria or proteinuria were detected. Negative control rats were injected subcutaneously with sterile PBS at day 0. All of the rats were euthanized at day 42.

Evaluation of Disease Severity

Urinary protein and BUN were analyzed by automatic biochemical analyzer (UniCel DxC 600 Synchron; Beckman Coulter, Inc.). Kidney tissue was examined by light microscopy and immunofluorescence. For light microscopy, kidney tissues were fixed in 10% buffered formalin and embedded in paraffin. Kidney sections (3 μm) was stained with periodic acid–Schiff. At least 60 glomeruli in each tissue were assessed. For immunofluorescence, frozen sections (5 μm) were fixed by acetone and incubated with FITC-conjugated anti-rat IgG (Jackson ImmunoResearch Laboratories Inc., Seattle, WA) at 1:50.

Antibody Detection by ELISA

IgG deposited in kidneys were eluted with glycine as described previously.¹⁵ Serum antibodies and kidney elutes against peptides and α3(IV)NC1 were detected by ELISA.¹⁵ Peptides (5 μg/ml) and α3(IV)NC1 (2 μg/ml) were coated overnight to the 96-well plates. Serum samples at 1:100 or kidney elutes at 1:5 in PBST (PBS and 0.1% Tween-20) with 1% BSA were incubated in the plate for 1 hour at 37°C. Then, alkaline phosphatase–conjugated goat-anti-rat IgG (Sigma) diluted in PBST at 1:5000 were added at 37°C for 30 minutes. P-nitrophenyl phosphate (1 mg/ml; Sigma) in substrate buffer (1mol/L diethanolamine, 0.5mmol/L MgCl₂, pH 9.8) was used as substrate, and color development was measured spectrophotometrically at 405 nm (Bio-Rad, Tokyo, Japan). For competitive inhibition ELISA, serum samples were preincubated with different concentrations of inhibiting peptides (0.1–10 μg/ml) at 37°C for 60 minutes before incubation with coated antigens.

Intrakidney Infiltrates and Deposits by Immunohistochemistry

Immunohistochemistry was performed for staining intrakidney CD4⁺ T cells, CD8⁺ T cells, and macrophage infiltrates, and C3d, C4d, and fibrin deposition. First, 3 μm paraffin-embedded, formalin-fixed sections underwent either heat-mediated (for CD4, CD8, C4d, and Foxp3) or enzyme-mediated (CD68 with 0.04% pepsin, or C3d and fibrin with 0.5 mg/ml proteinase K) antigen retrieval. Then, 3% H₂O₂ was applied for blocking the endogenous peroxidase at room temperature for 10 minutes and 3% BSA was used for blocking the unspecific binding sites at 37°C for 1 hour. The incubation of monoclonal antibodies (anti-CD4; Cell Signaling Technology, Danvers, MA and anti-Foxp3; Abcam, Cambridge, MA) or polyclonal antibodies (anti-CD8; Santa Cruz Biotechnology, Dallas, TX; anti-CD68; Abcam; anti-fibrin; DAKO, Santa Clara, CA; anti-C3d; R&D Systems, Minneapolis, MN; and anti-C4d; Biomedica, Budapest, Hungary) was performed overnight at 4°C. Ultrasensitive two-step kit and 3,3′-diaminobenzidine (ZGSB-BIO, Beijing, China) were used for following procedures and color development.

At least 20 consecutive glomeruli were analyzed by Image-Pro Plus (version 6.0; Media Cybernetics, Dallas, TX) for the staining of each marker. CD4⁺, CD8⁺ and Foxp3⁺ T cells were displayed as cells per glomerular cross section. Macrophages, C3d, C4d, and fibrin were displayed as integrated OD per glomerular cross section.

Intrakidney Cytokines by Quantitative Real-Time PCR

Kidney RNA was extracted by RNAprep Pure Tissue Kit (TIANGEN BIOTECH, Beijing, China). Complementary DNA was acquired from extracted RNA with the High Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, Waltham, MA). Real-time quantitative PCR was performed in triplicate with the PowerUpTM SYBR Green Master Mix system (Thermo Fisher Scientific) to quantify the mRNA expression levels of rat IFN-γ, IL-4, IL-17, and IL-10 through the 2^-ΔΔCT method, with β-actin as reference gene. Primer sets are shown in Supplemental Table 1.

Peptide-MHC Binding Assay

The binding capacity of α3-P14 and m-P14 with rat MHC II molecules RT1B and RT1D and human MHC II molecules HLA-DRB1*1501 were detected by flow cytometry. The gating strategies are shown in Supplemental Figure 1.

Lymphocytes extracted from rat spleens without red blood cell lysis buffer were incubated with biotinylated peptides α3-P14, m-P14, and hemagglutinin (HA_307–319, negative control). Phycoerythrin (PE)-mouse-anti-rat RT1B IgG (1:1000), FITC-mouse-anti-rat RT1D IgG (1:200,) and allophycocyanin (APC)-streptavidin (1:4000) (BioLegend, San Diego, CA)¹² were incubated with the cells for 20 minutes at room temperature to label RT1B and RT1D molecules and peptides, respectively. PE-mouse IgG1 (1:1000) and FITC-mouse IgG1 (1:200) were used as isotype controls. After washing twice, cells were detected by FACS Verse flow cytometer (Becton Dickinson, California). All experiments were repeated thrice. Binding percentage >75% was considered as high affinity, 25%–75% was medium affinity, and <25% was low affinity.

A cell line of EBV-transformed human B cells homozygous for HLA-DRB1*1501 (sample identifier: MGar, International Histocompatibility Working Group, Seattle, WA) was used as previously reported.¹⁹ Cells were incubated with biotinylated peptides α3-P14, m-P14, and HA_307–319 at 10 μg/ml for 30 minutes at room temperature. After washing, cells were incubated with mouse anti–HLA-DRB1 (1:40) IgG (Abnova, Taipei, Taiwan). Normal mouse IgG (Sigma) was used as control. PE-goat-anti-mouse IgG (1:200) (BioLegend) and APC-streptavidin (1:1000) (BioLegend) were added to stain the cells and peptides, respectively. Cells were detected by FACS Verse flow cytometer.

For competitive inhibition, 10–50 μg/ml m-P14 or HA_307–319 were coincubated with rat splenocytes or with HLA-DRB1*1501⁺ cells with 5 μg/ml biotinylated α3-P14 for 20 minutes at room temperature. Cells were stained and detected following the above procedures.

Enzyme Linked Immunospot

IFN-γ secretion was detected by Enzyme Linked Immunospot (ELISpot) (Mabtech AB, Nacka Strand, Sweden). Rats were immunized with 10 μg/rat α3-P14 by hind footpad injection and emulsified in complete Freund adjuvant. Ten days later, splenocytes were collected (1×10⁵ cells per well in triplicate) and incubated for 20 hours in wells coated with anti–IFN-γ-mAb together with 5 μg/ml Con A (positive control), 10–50 μg/ml α3-P14, m-P14, OVA_223–239 (negative control), or s-P14. For competitive inhibition assay, 10 μg/ml α3-P14 was added to the cells together with α3-P14, m-P14, or s-P14 (50 μg/ml). Plates were read by ELISpot Reader (AID, Strasberg, Germany).

Intracellular Cytokine Staining

The T helper cell (Th cell) subsets of rat splenocytes were detected by flow cytometry. The gating strategies are shown in Supplemental Figure 2.

Splenocytes collected from rats of m-P14 immunization and from rats of the early-treatment or later-treatment groups were stimulated with phorbol-12-myristate 13-acetate (0.081 µM), ionomycin (1.3386 µM), and brefeldin A (0.005 mg/ml) (BioLegend) at 37°C for 5 hours. Cell surface markers were stained with FITC-anti-CD3, APC-Cy7 anti-CD4, peridinin-chlorophyll-protein complex-anti-CD8a, and PE-anti-CD25 antibodies (BioLegend) for 20 minutes at room temperature. Subsequently, cells were fixed with 1% formaldehyde for 30 minutes at room temperature, followed by permeabilization.

For intracellular staining, cells were resuspended in diluted Intracellular Staining Perm Wash Buffer (BioLegend) and centrifuged at 350 g for 5 minutes. For intranuclear staining, diluted Transcription Factor Perm Buffer (BioLegend) was added to cell samples and centrifuged at 350×g for 5 minutes. Both steps were repeated. Either buffer was applied for subsequent intracellular or intranuclear staining respectively as staining buffer.

Antibodies against intracellular cytokines were used to stain the IFN-γ (Alexa Fluor 647 anti-rat IFN-γ; BioLegend), IL-4 (PE anti-rat IL-4; BioLegend), and IL-17 (PE-Cy7 anti-mouse/rat IL-17A; eBioscience). For Treg detection, there was no stimulation of cells and the fixation time was prolonged to 45 minutes. Antibodies against Foxp3 (Alexa Fluor 647 anti-mouse/rat/human Foxp3; BioLegend) were used for endonuclear staining. Cells were detected by FACS Verse flow cytometry.

Molecular Docking

Modeling of P14 and m-P14 presented by the HLA-DRB1*1501 molecule was performed by the software RosettaDock 3.8 with docking procedures described by Selvaraj et al.²⁰ The α3_135–145-HLA-DRB1*1501 (Protein Data Bank code: 5V4M) complex was chosen as the template for docking the three-dimensional models of peptides (α3-P14 and m-P14) and HLA-DRB1*1501 molecule. The satisfying complementarity of molecular shape was generated from the docked transformation predicted by the software. In cluster analysis, 4Å was set as a default value for clustering and root-mean-square deviation clustering was used for discarding redundant solutions.

Statistical Analyses

SPSS version 13.0 (SPSS Inc., Chicago, IL) was used for statistical analysis. Results were expressed as mean±SEM. Differences of quantitative parameters were assessed using separate individual t test for data that were normally distributed or Mann–Whitney U test for data that were not normally distributed. P values <0.05 were considered statistically significant.

Study Approval

All animal experiments were approved by the Experimental Animal Ethics Committee of Peking University First Hospital (Beijing, China).

Results

Peptide Alignment of α3-P14 and α1-P14 Sequences and Design of m-P14

Peptide sequence alignment was conducted via Geneious version 6.1.8.⁶ The amino acid sequences of α3-P14 (α3_127–148) and the counterpart on α1 chain (α1-P14, α1_127–148) were compared (Figure 1). In the critical amino acid motif of P14 (W₁₃₆I₁₃₇_L₁₃₉W₁₄₀_F₁₄₃_F₁₄₅), the 137 position and 143 position were different between α3-P14 and α1-P14. I₁₃₇ of α3-P14 is of the highest hydrophobicity, whereas serine (S₁₃₇) of α1-P14 is neutral, which is significantly different. F₁₄₃ of α3-P14 and tyrosine₁₄₃ of α1-P14 are both hydrophobic and have similar binding characteristics to HLA-DR1 molecule.²¹ Therefore, we designed a modified peptide (m-P14) by the substitution of I₁₃₇ with S₁₃₇ (Figure 1).

To ensure the specificity of m-P14’s effects, we designed a scrambled peptide (s-P14, IWPTGCSIPDSKMPFSHFWGLF) as a control, which has the same amino acids as m-P14 but in randomized sequence. s-P14 was administered with the same protocols as m-P14.

Immunization of m-P14 and α1-P14 Did Not Induce Kidney Injury in WKY Rats

The pathogenicity of m-P14 and α1-P14 was investigated by immunization to WKY rats with a single dose of 200 μg/kg (Supplemental Figure 3). All rats developed antibodies toward the immunogen m-P14/α1-P14. None of them manifested any kidney injury. The urinary protein and BUN were comparable to the negative controls (P>0.05). No linear IgG deposit or crescent was observed in the kidneys.

In α3-P14 group by the same immunization method, all rats developed severe crescentic nephritis with 97.5%±1.5% of crescent formation, elevated urinary protein, and BUN, together with circulating anti–α3-P14 antibodies and glomerular linear deposits.

m-P14 Arrested α3-P14–Induced anti-GBM GN

In early-treatment groups, m-P14 was injected to α3-P14 immunized rats from day 0 to day 28, at 30 mg/kg and 10 mg/kg, respectively (Figure 3). Serum antibodies against α3-P14 were significantly lowered in the two early-treatment groups from week 3 (30 mg/kg: 0.2±0.1 versus 1.3±0.1, P=0.002; 10 mg/kg: 0.4±0.1 versus 1.3±0.1, P=0.009) (Figure 3A). The antibody levels were decreased more remarkably in the group with higher dosage of m-P14 (30 mg/kg) at week 4 (0.2±0.03 versus 0.6±0.1, P=0.03). Crescent formation was scarcely seen in the two groups (30 mg/kg: 0.5±0.4 versus 68.8%±15.4%; P=0.002; 10 mg/kg: 6.3±5.6 versus 68.8%±15.4%, P=0.009) (Figure 3B). No linear IgG deposition was observed (Figure 3, E2 and E3). The proteinuria and BUN levels were significantly decreased and even comparable with that in negative group (Figure 3, C and D).

Figure 3.

Early-treatments of m-P14 arrested α3-P14-induced anti-GBM GN. WKY rats were immunized with α3-P14. m-P14 was injected intraperitoneally at 30 mg/kg (ET-30) or 10 mg/kg (ET-10) daily from day 0 to 14 and every other day from day 15 to 28. Compared with the group immunized with α3-P14, (A) the levels of antibodies against α3-P14, (B, F1–F4) glomerular crescents, (C) 24-hour proteinuria, and (D) BUN were significantly lower in the two early-treatment m-P14 groups. Linear IgG deposit along GBM was found in (E1) the α3-P14 group, but not observed in (E2 and E3) the two early-treatment groups and (E4) negative control group. In the s-P14 groups, (E5, E6) linear IgG deposition and (F5, F6) crescent formation were comparable with the α3-P14 group (B).

In s-P14 intervention groups, antibody levels against α3-P14 were comparable with that in the α3-P14 immunized group (P>0.05). All of the rats showed crescent formation (Figure 3, E5 and E6), with the percentages of crescents, proteinuria, and BUN comparable with the α3-P14 group (P>0.05) and significantly higher than m-P14 groups (Figure 3, B–D).

m-P14 Attenuated Kidney Injuries after Disease Presentation

In the later-treatment group, α3-P14 immunized rats were injected by m-P14 (30 mg/kg) after disease onset with proteinuria and hematuria (Figure 4). The level of serum antibodies against α3-P14 was comparable between the later-treatment group and untreated α3-P14 immunized rats (the disease group; P>0.05). However, the percentage of crescents was significantly lower in the later-treatment group (20.1±8.4 versus 68.8%±15.4%, P=0.03). Much weaker IgG deposition along GBM was observed in the later-treatment group. And the rats showed significantly lower level of BUN (22.0±1.7 versus 73.3±7.0 mg/dl, P=0.002).

Figure 4.

Treatment of m-P14 attenuated kidney injuries after anti-GBM GN presentation. WKY rats were immunized with α3-P14. m-P14 was injected intraperitoneally daily at 30 mg/kg (T-30) after the detection of hematuria or proteinuria (approximately 3 weeks after immunization). (A) The antibody levels against α3-P14 and (C) the levels of proteinuria were comparable between the later-treatment group and α3-P14 group. However, the (B, F1–F3) percentage of crescents and (D) BUN were significantly reduced. Linear IgG deposit along the GBM were weaker in (E2) the later-treatment group than (E1) the α3-P14 group, but stronger than (E3) the negative group. In the s-P14 group, the antibody levels, linear IgG deposit, and crescent formation were all comparable to the α3-P14 group (B, E4, F4).

In the s-P14 group, the antibody level against α3-P14, the percentage of crescents, proteinuria, and BUN were comparable with those in the α3-P14 group (P>0.05; Figure 4).

Kidney injury severity was evaluated at the time of m-P14 treatment initiation, at 3 weeks after α3-P14 immunization (Supplemental Figure 4). The rats presented with elevated urinary protein (24.8±15.8 versus 0.3±0.1 mg/24 h, P=0.002) and BUN (21.5±1.4 versus 14.5±1.4 mg/dl, P=0.02) compared with day 0, and a few crescents in the kidneys (5.5±2.6 versus 0%±0%, P=0.02). The proteinuria (24.8±15.8 versus 179.0±37.4 mg/24 h, P=0.009), BUN (21.5±1.4 versus 73.3±7.0 mg/dl, P=0.002), and crescents (5.5±2.6 versus 68.8%±15.4%, P=0.002) increased continuously in the α3-P14 group, but remained stable at the end of m-P14 treatment (P>0.05).

m-P14 Inhibited α3-P14 Binding to MHC II Molecules

There was no difference on the binding capacities between m-P14 and α3-P14 to both rat MHC II molecules RT1B (P=0.3) and RT1D (P=0.64) (Figure 5, A and B). The competitive inhibition assay showed that the binding of α3-P14 at 5 μg/ml to RT1D could be inhibited by m-P14 at 50 μg/ml, from 16.9%±0.9% to 11.4%±0.9% (P=0.01), but not by HA at the same concentration. The binding of α3-P14 at 5 μg/ml to RT1B could not be inhibited by m-P14 at any concentration (P>0.05) (Figure 5, D and E).

Figure 5.

m-P14 in high dose inhibited α3-P14 binding to MHC II molecules. (A and B) There was no difference on the binding capacities between m-P14 and α3-P14 to both RT1B and RT1D molecules. (E) The binding of α3-P14 at 5 μg/ml to RT1D could be inhibited by m-P14 at 50 μg/ml, but not by HA at the same concentration. (D) The binding of α3-P14 at 5 μg/ml to RT1B could not be inhibited by m-P14 at any concentration. (C) Both m-P14 and α3-P14 could bind to HLA-DRB1*1501. (F) The binding of α3-P14 at 5 μg/ml to HLA-DRB1*1501 could be inhibited by m-P14 at 50 μg/ml, but not by HA at the same concentration.

Both m-P14 (35.4%±2.0%, median affinity) and α3-P14 (87.1%±0.6%, high affinity) could bind to human HLA-DRB1*1501 (Figure 5C). The binding of α3-P14 at 5 μg/ml to HLA-DRB1*1501 could be inhibited by m-P14 at 50 μg/ml, from 49.7%±0.3% to 39.8%±0.9% (P<0.001), but not by HA at the same concentration (Figure 5F).

m-P14 Inhibited the Activation of α3-P14–Specific T Cells

The splenocytes collected at day 10 from α3-P14 immunized rats showed activation to the restimulation of α3-P14 at 10 μg/ml in vitro, compared with the splenocytes from PBS immunized rats (204.0±24.8 versus 54.7±16.0 spot forming cells/10⁶ cells, P=0.002), whereas they showed no response to the restimulation of m-P14 or s-P14 (P>0.05; Figure 6A1). In the competitive inhibition assay, α3-P14–specific splenocyte activation could be inhibited by the preincubation of m-P14 at 50 μg/ml (367.3±44.0 versus 169.1±59.0 spot forming cells/10⁶ cells, P=0.02), but not by OVA or s-P14 (P>0.05; Figure 6A2).

Figure 6.

m-P14 inhibited the activation of α3-P14-specific T cells and decreased the differentiation of Th17 cells. (A1 and A2) The activation of α3-P14–specific T cells were detected by ELISpot for the secretion of IFN-γ. The splenocytes collected from α3-P14 immunized rats showed activation after α3-P14 stimulation at 10 μg/ml, but had no response to m-P14 or s-P14 stimulation (A1). The α3-P14 induced splenocytes activation could be inhibited by m-P14 at 50 μg/ml, but not by OVA or s-P14 (A2). (B1–B5) The differentiation of splenocytes from m-P14 immunized rats were detected by flow cytometry. Compared with α3-P14 group, the percentage of Th1 cells in m-P14 group was decreased (B1); the percentage of Th2 cells were comparable among the groups (B2); the proportion of Th17 cells was decreased in m-P14 groups 200 μg/kg and 30 mg/kg immunization, and that of the m-P14 30 mg/kg group was even lower than that of the PBS immunization group (B3). The percentage of Treg cells was comparable between the α3-P14 group and m-P14 200 μg/kg group, and both were higher than that in the negative controls, which was comparable with that in the m-P14 30 mg/kg group (B4). The ratio of Treg cells/Th17 cells was significantly elevated in m-P14 groups compared with that in the α3-P14 group and negative controls (B5). (C) In both the early-treatment groups (30 mg/kg and 10 mg/kg from day 0) and the later-treatment group (30 mg/kg after disease onset) of m-P14, the levels of Th17 cells were significantly decreased, compared with that in α3-P14 group and those in s-P14 intervention groups.

m-P14 Decreased the Differentiation of Th17 Cells

In m-P14 immunized rats at week 3, the percentage of Th17 (CD3⁺CD4⁺IL-17⁺) cells in total CD3⁺CD4⁺ T cells was significantly lower than that in the α3-P14 immunized group (200 μg/kg: 3.5±0.1 versus 5.4%±0.5%, P=0.004; 30 mg/kg: 2.0±0.3 versus 5.4%±0.5%, P=0.002) (Figure 6B). The percentage of Th17 cells in the m-P14 30 mg/kg immunized group was even lower than that in the PBS immunization group (2.0±0.3 versus 3.0%±0.3%, P=0.03). The ratio of Treg/Th17 cells in m-P14 immunized groups was higher than that in α3-P14 group (200 μg/kg: 3.5±0.1 versus 2.5±0.2, P=0.009; 30 mg/kg: 5.4±0.9 versus 2.5±0.2, P=0.002), although the percentages of Treg cells were not prominent among all of these groups.

In both early-treatment groups, the ratio of Th17 cells (intervention groups/negative group) were significantly decreased (30 mg/kg: 1.0±0.1 versus 1.8±0.2, P=0.004; 10 mg/kg: 1.0±0.2 versus 1.8±0.2, P=0.02) compared with that in the disease group and those in s-P14 intervention groups (Figure 6C). It was also decreased in the m-P14 later-treatment group (0.8±0.1 versus 1.8±0.2, P=0.002). Th17 cells were reduced remarkably in all m-P14 intervention groups, to the extent that they were comparable with that in negative group. There was no difference in other Th cell subsets between the disease group and intervention groups (P>0.05).

m-P14 Inhibited Antibody Binding to α3-P14

Using competitive inhibition ELISA, we found that the binding of serum IgG to α3-P14 could be inhibited by m-P14 in a dose-dependent manner. The inhibition capacity of m-P14 (91.1%) was even stronger than the immunogen α3-P14 itself (59.4%) (Figure 7A).

Figure 7.

m-P14 inhibited antibody binding to α3-P14 and confined the intramolecular epitope spreading induced by α3-P14. The binding of serum IgG to α3-P14 coated onto an ELISA plate could be inhibited by soluble α3-P14 and m-P14. (A) The inhibition capacity of m-P14 (91.1%) was even stronger than α3-P14 (59.4%). (B) Serum (n=6) and (C) kidney eluted (n=3) antibodies against the intact α3NC1 protein could be detected in α3-P14 group, but not in the two early-treatment groups, and were much lower in the later-treatment group.

m-P14 Impeded Intramolecular Epitope Spreading Induced by α3-P14

In the α3-P14 immunization group, serum antibodies against intact α3NC1 could be detected 1 week after the occurrence of anti-P14 antibodies, indicating intramolecular epitope spreading from linear peptide α3-P14 to conformational epitopes on intact α3NC1 protein. In the two early-treatment groups, no antibody toward α3NC1 was detectable from the circulation (Figure 7B), nor from the kidney elute (Figure 7C). In the later-treatment group, antibodies against α3NC1 were detectable, but of a much lower level compared with that in the disease group (circulation: 0.1±0.1 versus 0.5±0.1, P=0.03; kidney elute: 0.1±0.02 versus 0.3±0.02, P=0.001).

The Intrakidney Inflammatory Responses

The intrakidney infiltrate of CD4⁺ and CD8⁺ T cells and macrophages, and the depositions of C3d, C4d, and fibirin, were significantly reduced in all three m-P14 intervention groups, compared with the α3-P14 immunized group and s-P14 intervention groups (Figure 8, A–F).

Figure 8.

m-P14 suppressed the intrakidney inflammatory responses. (A–F) The IHC staining showed significantly reduced CD4⁺, CD8⁺, and CD68⁺ cells and weakened C3d, C4d, and fibrin deposits in all of the three m-P14 intervention groups, compared with α3-P14 or s-P14 groups. (G) mRNA detection showed significantly reduced expression of IFN-γ in both early-treatment groups (G1), IL-17 reduction in the -P14 30 mg/kg early-treatment group (G3), and IL-10 enhancement in the m-P14 10 mg/kg early-treatment group and the later-treatment group (G4). There was no difference among the groups for IL-4 (G2).

The intrakidney cytokines of IFN-γ, IL-4, IL-17, and IL-10 were detected by RT-PCR. The expression of proinflammatory cytokine IFN-γ was reduced in both early-treatment groups. IL-17 expression was also reduced in the 30 mg/kg early-treatment group. However, the expression of IL-10 was enhanced in the 10 mg/kg early-treatment group and later-treatment group. There was no difference among the groups for IL-4 (Figure 8G).

Foxp3⁺ cells were increased in kidney tissues of m-P14 intervention groups compared with the α3-P14 immunized group, s-P14 intervention groups, and negative control group (Supplemental Figure 5).

The Molecular Docking of Peptide-HLA Interactions

Molecular docking was applied to investigate the interactions between α3-P14/m-P14 and HLA-DRB1*1501 molecule. The amino acid mutation on α3₁₃₇ of P14 (from isoleucine to serine) disarranged the binding pattern, with serine difficult fitting into pocket 1 and fewer hydro bonds formed between the peptide and HLA molecule (Supplemental Figure 6).

Discussion

In this study, we modified the nephrogenic T cell epitope from α3(IV)NC1 by the substitution of one amino acid residue in its critical motif from the counterpart of α1NC1, and constructed a new peptide which could arrest and attenuate the kidney injuries in experimental autoimmune GN (EAG) model of anti-GBM GN. All of the three intervention groups with m-P14 revealed improvement of clinical features, milder kidney pathologic injury, and less aggressive inflammatory responses. The rats treated at the same day after α3-P14 immunization showed no signs of kidney injuries. When m-P14 was administrated after disease onset with proteinuria and hematuria, further crescent formation was halted and the rats showed milder kidney injuries. To the best of our knowledge, this study is the first investigation of immunotherapy using a modified peptide on Goodpasture disease.

Peptide immunotherapy has been identified as a potential intervention for many autoimmune diseases because of its capability to modulate and restore immune homeostasis.^22–24 Several peptides originated from autoantigens have been adopted in clinical trials for multiple sclerosis, type 1 diabetes, and celiac disease.²³ Therapeutic peptides could alleviate autoimmune diseases mostly via two mechanisms.²⁵ One is through interfering antigen presentation process by competitive binding to the groove of MHC molecules, which blocks the presentation of pathogenic self-antigens on the MHC molecules. The other is to apply proper substitutions of TCR contact residues of the original pathogenic peptide to alter the subsequent T cell responses, including inhibiting antigen-specific T cell activation, increasing the percentage of protective Th2 cells and inducing regulatory T cells to suppress the inflammatory responses.^24,26–32

In this study, m-P14 showed nearly the same binding affinity as α3-P14 to RT1D and RT1B molecules. The competitive inhibition effect of m-P14 was displayed at a relatively high concentration (ten times the concentration of α3-P14) and worked only on RT1D but not on RT1B. We speculated that m-P14 in vivo might prevent the pathogenic peptide α3-P14 from being presented by blocking its binding to MHC molecules; however, this mechanism was not the major one. We also detected the binding between m-P14 and HLA-DRB1*1501 molecule, and found a lower affinity compared with α3-P14. This could be explained by the molecular docking that the amino acid mutation on α3₁₃₇ of P14 disarranged its binding pattern. Thus, m-P14 at a high dosage was required to exert competitive inhibition effects during the antigen presentation process.

The major mechanism of m-P14 therapeutic role may be the alteration of subsequent T cell responses. m-P14 could inhibit the IFN-γ secretion of splenocytes induced by α3-P14, serving as a potential antagonist to P14-TCR interactions. Antigen-TCR contacting is critical for immune modulation.²⁵ Modified peptide was administrated to experimental autoimmune encephalomyelitis mice models and the relief of disease progression was observed. Through T cell proliferation assay^30,31 and other experiments, the modified peptide was considered to play as TCR antagonist which brought inhibitive effects to the disease.³⁰

TCR recognition of different antigens may subsequently induce discrepant Th cell differentiation.³³ In this study, we found that Th17 cells were remarkably decreased in all m-P14 treatment groups and even in m-P14 immunized rats, indicating that the recognition of m-P14 does impede the differentiation of Th17 cells. We also observed the reduced m-RNA levels of IFN-γ and IL-17 and less local inflammatory cells infiltration in the kidneys of m-P14 intervention groups. Th17 cells play critical proinflammatory roles in the pathogenesis of autoimmune kidney diseases.^34–37 A previous study showed that a lower level of IL-17 was associated with ameliorated kidney damage of experimental GN.³⁵ The absence of IL23/Th17 axis lowered the autoimmune responses and displayed a protective pattern for proliferative and crescentic GN.^34,36 We speculated that the modified peptide m-P14 may affect TCR contact residues by the altered affinity and disrupt the pathways for Th17 cells differentiation.³³

Moreover, we found that the ratio of Treg/Th17 cells was elevated in m-P14 immunized rats in a dose-dependent manner. Increased expression of IL-10 and more Foxp3⁺ cells were found in the kidneys of m-P14 intervention groups, implying that Treg cells may be upregulated. A detailed description on T cell subsets in the kidneys is necessary in the future. Previous studies have indicated that certain epitopes like “Tregitopes” could induce Treg cell activation when coincubated with PBMCs.³⁸ Endogenous Treg cells were proved by Ooi et al.³⁹ to suppress inflammation by infiltrating the kidney in the later phase of an EAG mice model. Thus, the higher ratio of Treg cells induced by m-P14 may also contribute to its therapeutic effects.

Besides the change in cellular immunity, we also noticed that the level of antibodies toward α3-P14 was decreased significantly in early-treatment groups, and epitope spreading from linear α3-P14 to conformational α3NC1 was impeded in the later-treatment group. One possible explanation for these humoral immunity alterations is that m-P14 may inactivate the α3-P14–specific T cells and produce less signals to stimulate B cells for anti-P14 antibody production.⁴⁰ Another process may be attributed to the competitive inhibition of m-P14 to the binding between α3-P14 and its circulating antibodies. The inhibition capacity of m-P14 was even stronger than the immunogen α3-P14 itself. Thus, m-P14 could competitively block the existing pathogenic anti-P14 antibodies directly, arrest the antibody deposit on GBM, and restrain kidney destruction.

In conclusion, we designed a modified peptide derived from the nephrogenic T cell epitope on α3NC1 by one single amino acid substitution from α1NC1, which could arrest and attenuate the kidney injuries of anti-GBM GN in rat model, through mechanism on cellular and humoral immunity regulation. This approach confirmed the feasibility of modulating T cell activation for the treatment of Goodpasture disease and may shed new insights on the treatment of autoimmune kidney diseases in future.

Disclosures

None.

Funding

This work is financially supported by grants from Natural Science Foundation of China to the Innovation Research Group (81621092), the Outstanding Young Scholar (81622009), and the general programs (81870482, 81870486).

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2019010067/-/DCSupplemental.

Supplemental Table 1. The sequences of RT-PCR primers.

Supplemental Figure 1. Histogram data and gating strategies for affinity binding of MHC II molecules in flow cytometry.

Supplemental Figure 2. Histogram data and gating strategies for T cell subsets in flow cytometry.

Supplemental Figure 3. Immunization of m-P14 and α1-P14 did not induce kidney injury in WKY rats.

Supplemental Figure 4. The clinical and pathologic features of α3-P14 immunized rats at the time of m-P14 treatment initiation.

Supplemental Figure 5. Foxp3⁺ staining in the kidney tissues.

Supplemental Figure 6. Structural model of DRB1*1501-peptide complex.