Dominant protection from HLA-linked autoimmunity by antigen-specific regulatory T cells

Abstract

Susceptibility and protection against human autoimmune diseases, including type I diabetes, multiple sclerosis, and Goodpasture disease, is associated with particular human leukocyte antigen (HLA) alleles. However, the mechanisms underpinning such HLA-mediated effects on self-tolerance remain unclear. Here we investigate the molecular mechanism of Goodpasture disease, an HLA-linked autoimmune renal disorder characterized by an immunodominant CD4⁺ T-cell self-epitope derived from the α3 chain of type IV collagen (α3_135-145). While HLA-DR15 confers a markedly increased disease risk, the protective HLA-DR1 allele is dominantly protective in trans with HLA-DR15 (ref. 2). We show that autoreactive α3_135-145-specific T cells expand in patients with Goodpasture disease and, in α3_135-145-immunized HLA-DR15 transgenic mice, α3_135-145-specific T cells infiltrate the kidney and mice develop Goodpasture disease. HLA-DR15 and HLA-DR1 exhibit distinct peptide repertoires and binding preferences and present the α3_135-145 epitope in different binding registers. HLA-DR15-α3_135-145 tetramer⁺ T cells in HLA-DR15 transgenic mice exhibit a conventional T-cell phenotype (T_conv) that secretes pro-inflammatory cytokines. In contrast, HLA-DR1-α3_135-145 tetramer⁺ T cells in HLA-DR1 and HLA-DR15/DR1 transgenic mice are predominantly CD4⁺Foxp3⁺ regulatory T cells (T_reg cells) expressing tolerogenic cytokines. HLA-DR1-induced T_reg cells confer resistance to disease in HLA-DR15/DR1 transgenic mice. HLA-DR15⁺ and HLA-DR1⁺ healthy human donors display altered α3_135-145-specific T-cell antigen receptor usage, HLA-DR15-α3_135-145 tetramer⁺ Foxp3^- T_conv and HLA-DR1-α3_135-145 tetramer⁺ Foxp3⁺CD25^hiCD127^lo T_reg dominant phenotypes. Moreover, patients with Goodpasture disease display a clonally expanded α3_135-145-specific CD4⁺ T-cell repertoire. Accordingly, we provide a mechanistic basis for the dominantly protective effect of HLA in autoimmune disease, whereby HLA polymorphism shapes the relative abundance of self-epitope specific T_reg cells that leads to protection or causation of autoimmunity.

Extended Data Figure 1. HLA-DR15-α3_135-145 tetramer⁺ CD4⁺ T cells from humans and mice respond to α3_135-145 and to whole α3(IV)NC1

a, HLA-DR15-α3_135-145 tetramer⁺ CD4⁺ T cells isolated from patients with Goodpasture’s disease (n=8) respond to α3_135-145 and recombinant human (rh)α3(IV)NC1. HLA-DR15-α3_135-145 tetramer⁺ CD4⁺ T cells were isolated by magnetic bead separation from the blood of patients with Goodpasture’s disease, then cultured at a frequency of 400 HLA-DR15-α3_135-145 tetramer⁺ CD4⁺ T cells per well in the presence of mitomycin C treated HLA-DR15-α3_135-145 tetramer⁺ cell depleted white blood cells and either hα3_135-145 or rhα3(IV)NC1. Antigen specific responses were assessed by ELISPOTs for IFN-γ and IL-17A and expressed as numbers of spots per well. b, To induce experimental autoimmune anti-GBM disease, DR15^{+.Fcgr2b−/−} mice (n=5) were immunized with hα3_135-145 on days 0, 7 and 14. Disease was measured at day 42, by which time DR15^{+.Fcgr2b−/−} mice develop kidney disease similar to human Goodpasture’s disease including glomerular segmental necrosis, crescents (PAS stain), linear IgG deposition (direct immunofluorescence) and pathological albuminuria. Each dot represents one mouse. Photomicrographs taken at 400x depict a crescentic glomerulus with segmental necrosis (bottom left panel) and linear glomerular IgG deposition (bottom right panel) seen in DR15^{+.Fcgr2b−/−} mice immunized with hα3_135-145. Scale bars = 30μm. c, HLA-DR15-α3_135-145 tetramer⁺ CD4⁺ T cells isolated from DR15^+.Fcgr2b+/+ mice respond to mouse (m)α3_136-146, hα3_135-145, rmα3(IV)NC1 and rhα3(IV)NC1. HLA-DR15-α3_135-145 tetramer⁺ CD4⁺ T cells were isolated by magnetic bead separation from pooled spleen and lymph node cells of two DR15⁺ transgenic mice 10 days after immunisation with mα3_136-146 (DWVSLWKGFSF, hα3_135-145 GWISLWKGFSF) then cultured at a frequency of 400 HLA-DR15-α3_135-145 tetramer⁺ CD4⁺ T cells per well in the presence of mitomycin C treated HLA-DR15-α3_135-145 tetramer⁺ cell depleted spleen and lymph node cells. Antigen specific responses were assessed by ELISPOTs for IFN-γ and IL-17A and expressed as numbers of spots per well. ^*P<0.05; ^**P<0.01; ^***P<0.001 by Kruskal-Wallis test (a and c) or Mann-Whitney U test (b).

Extended Data Figure 2. Immune responses in DR15⁺DR1^+.Fcgr2b+/+ mice

a, Naïve FcγRIIb intact HLA transgenic mice have similar immune properties. Shown are the proportion of splenocytes expressing HLA-DR; HLA-DR15 and HLA-DR1 intensities (MFI) in naïve DR15⁺ (n=4), DR1⁺ (n=4) and DR15⁺DR1⁺ transgenic mice (n=5) showing half the amount of expression in DR15⁺DR1⁺ relative to its single transgenic counterpart; total number of splenocytes; proportion of CD4⁺ and CD4⁺Foxp3⁺ splenocytes; total number of lymph node cells retrieved from the brachial, axillary and inguinal lymph nodes; proportion of CD4⁺ and CD4⁺Foxp3⁺ lymph node cells and the overall Vβ repertoire of HLA transgenic mice showing no skewing of any one Vβ chain. Data expressed as mean±s.e.m., analysed by ANOVA. b, Co-expression of HLA-DR1 abrogates pro-inflammatory high autoreactivity to peptide mα3_129-148 but not to other parts of α3(IV)NC1. FcγRIIb intact DR15⁺, DR1⁺ and DR15⁺DR1⁺ mice were immunized with murine α3 peptide pools (see Methods for peptide pools) (n = 5 each group) and responses to individual 20-mer peptides measured by restimulating the draining lymph node cells ex vivo (x-axis numbering represents the N-terminus amino acid number of the individual 20-mer peptide) using [³H]-T proliferation assays, and IFN-γ and IL-17A ELISPOTs. Each dot represents the mean response from triplicate determinations in an individual mouse and each bar represents the mean response in each group. Blue bars indicate no reactivity (SI < 2; spots < 5), yellow bars indicate low reactivity (2 < SI < 5; 5 < spots < 25), and red bars indicate high reactivity (SI > 5, spots > 25).

Extended Data Figure 3. Self-peptide repertoires of HLA-DR15/DR51 and HLA-DR1 using human cell lines, with electron density maps for the α3_135-145 peptide

Peptide repertoire analysis of 9-mer core sequences of HLA-DR presented self-peptides in the human BLCLs IHW09013 (DR15⁺/DR51⁺) and IHW09004 (HLA-DR1⁺). Amino acid frequencies in each peptide position p1 to p9 were plotted using IceLogo, with the human proteome as the frequency reference. Relatively enriched amino acids are plotted above the horizontal bar, and depleted amino acids below. The scale of each letter is proportional to the frequency difference to the reference. Peptides were eluted from a, IHW09013 HLA-DR15⁺/DR51⁺ cells, and b, from IHW09004 HLA-DR1⁺ cells. For HLA-DR1, preferred amino acids at p1, p4, p6, p7 and p9 are similar between the human and mouse DR1⁺ cells. For DR15⁺/DR51⁺ human cells compared to DR15⁺ mouse APCs, the difference at p9, with Lys and Arg being the most frequently bound residues, reflects the known DR51-related motif. c, d, Simulated annealing 2F_obs-F_calc omit maps for the peptide of c, HLA-DR15 and d, HLA-DR1. e, f, Final 2F_obs-F_calc maps for the peptide in e, HLA-DR15 and f, HLA-DR1. Electron density maps contoured at 1σ are shown as blue mesh.

Extended Data Figure 4. HLA-DR1-α3_135-145 tetramer⁺ CD4⁺ T cells have differential TRAJ usage and the CD4⁺CD25⁺ Tregs gene transcription profile is similar to that of thymically derived Tregs (tTregs)

a, The TRAJ and TRBJ usage of α3_135-145-specific CD4⁺ T cells was compared using the HLA-DR15-α3_135-145 tetramer (in naïve DR15^{+ Fcgr2b+/+} mice) and the HLA-DR1-α3_135-145 tetramer (in naïve DR1^+.Fcgr2b+/+ mice) by single cell sequencing. α3_135-145-specific CD4⁺ tetramer⁺ T cells from naïve DR15⁺ (n=3) and DR1⁺ (n=3) transgenic mice were pooled then single cell sorted, TCR genes amplified by multiplex PCR, then sequenced to determine TRAJ (DR15, n=81; DR1 n=84) and TRBJ (DR15, n=100; DR1 n=87) usage. For each TCR type/region (TRAV, TRBV, TRAJ, TRBJ; TRAV and TRBV shown in Fig. 2g), we compared the TCR distribution (frequencies of different TCRs) between DR15 and DR1, ^*** P<0.001 by Fisher’s exact test. b, HLA-DR1-α3_135-145 tetramer⁻ CD4⁺CD25⁻ Tconv, HLA-DR1-α3_135-145 tetramer⁻CD4⁺CD25⁺ Tregs and HLA-DR1-α3_135-145 tetramer⁺ CD4⁺CD25⁺ Tregs were isolated from naïve FcγRIIb intact DR1⁺ or DR15⁺DR1⁺ transgenic mice by flow cytometry, RNA extracted and the relative expression of tTreg genes expressed relative to 18S. The HLA-DR1-α3_135-145 tetramer⁺ CD4⁺CD25⁺ Tregs express genes consistent with a tTreg origin, similar to other Tregs that are not α3_135-145-specific from the same mice.

Extended Data Figure 5. Responses and effects of antigen specific DR1-associated Tregs

a, Measurement of pro- and anti-inflammatory mouse (m)α3_136-146-specific CD4⁺ T cell responses by ex vivo stimulation. CD4⁺ T cells, isolated and pooled from naïve DR15^+.Fcgr2b+/+ (n=2), DR1^+.Fcgr2b+/+ (n=2) and DR15⁺DR1^+.Fcgr2b+/+ (n=2) transgenic mice were cultured for 8 days in the presence of mα3_136-146 and mitomycin C treated syngeneic CD4⁺ cell-depleted splenocytes. Tregs were depleted by removing CD4⁺CD25⁺ T cells by cell sorting. IFN-γ, IL-17A, IL-6 and IL-10 were measured in the cultured supernatant by cytometric bead array and TGF-β by ELISA. The experiment was performed in triplicate and the data presented as mean±s.e.m. To measure α3_136-146-specific CD4⁺ Treg proliferation, CD4⁺ T cells were labelled with CTV then stained with Foxp3 on day 8. Results were similar to those performed using human (h)α3_135-145 presented in Fig. 3b. b, HLA-DR1-α3_135-145-specific Tregs are potent suppressors of HLA-DR15-α3_135-145 induced pro-inflammatory responses. CD4⁺CD25⁻ T cells were isolated and pooled from naïve DR15⁺DR1^+.Fcgr2b+/+ mice (n=2) and co-cultured with either CD4⁺CD25⁺ Tregs (which included HLA-DR1-α3_135-145 tetramer⁺ Tregs) or CD4⁺CD25⁺ HLA-DR1-α3_135-145 tetramer⁻ from naïve DR1^+.Fcgr2b+/+ mice (n=4) in the presence of mα3_136-146 and CD4⁺ cell-depleted spleen and lymph node cells from DR15⁺DR1⁺ mice. Cells were cultured for 8 days. α3_135-145-specific CD4⁺CD25⁻ cell proliferation was measured by labelling only the DR15⁺DR1⁺ derived naïve CD4⁺CD25⁻ cells with CTV; IFN-γ, IL-17A, IL-6 and IL-10 were measured in the cultured supernatant by cytometric bead array. In the absence of CD4⁺CD25⁺ HLA-DR1-α3_135-145 tetramer⁺ cells, the capacity of Tregs to prevent the induction of autoreactivity to α3_135-145 was impaired. These experiments were performed in triplicate and the data presented as mean±s.e.m. c, Treg depletion expands HLA-DR15-α3_135-145-specific T follicular helper (Tfh) cells. Anti-CD25 mAbs (or control Rat IgG) were administered 2 days prior to hα3_135-145 immunization and boost (n=4 per group), then the number of α3_135-145-specific PD-1⁺CXCR5⁺ Tfh cells enumerated in the draining lymph nodes. FACS plots show the expansion of PD1⁺CXCR5⁺ Tfh cells after Treg depletion and the detection of HLA-DR15-α3_135-145-specific cells within that population. ^*P < 0.05; ^** P < 0.01; ^*** P < 0.001 by Mann-Whitney U test (a and b) or Kruskal-Wallis test (c).

Extended Data Figure 6. In vivo Treg depletion in HLA transgenic mice

a, Efficiency and duration of Treg depletion using anti-CD25 mAb (clone PC61) in DR15⁺DR1^{+.Fcgr2b−/−} mice. Timeline showing the administration of anti-CD25 mAb 2 days before immunizing mice with human (h)α3_135-145. Detection of CD4⁺foxp3⁺ Tregs in the blood showing depletion of Tregs at days 7 and 14 in mice that received the anti-CD25 mAb (n=5 each group). White bars represent mice that received control antibodies and solid bars represent mice that received anti-CD25 mAb. b, DR15^{+.Fcgr2b−/−} (n=6 each group), DR1^{+.Fcgr2b−/−} (n=4 each group) and DR15⁺DR1^{+.Fcgr2b−/−} (n=6 each group) mice immunized with a control peptide, OVA_323-339 do not develop renal injury. Functional injury measured by albuminuria and blood urea nitrogen; and histological injury assessed by scoring of PAS stained histological sections for segmental necrosis and glomerular crescents. Representative PAS stained histological sections. White bars represent mice that received control antibodies and solid bars represent mice that received anti-CD25 mAb. c, Treg depletion leads to cell mediated injury in DR15⁺DR1^{+.Fcgr2b−/−} mice (further data collected in the experiment presented in Fig. 4a). Anti-CD25 mAbs (or control Rat IgG) were administered 2 days prior to the induction of experimental autoimmune anti-GBM GN by hα3_135-145 immunization in DR15^{+.Fcgr2b−/−} (n=6 per group), DR1^{+.Fcgr2b−/−} (n=4 per group) and DR15⁺DR1^{+.Fcgr2b−/−} (n=8 [Treg intact], 9 [Treg depleted]) mice. Cell mediated injury was assessed by quantifying glomerular fibrin deposition and enumerating and inflammatory cell infiltrates (CD4⁺ T cells, macrophages and neutrophils). Renal inflammation was measured by RT-PCR of inflammatory cytokines (TNF, IL-6 and IL-1α) on kidney digests. Scale bars = 30 μm. d, Repeating the experiment presented in Fig. 4a using mouse (m)α3_136-146 instead of hα3_135-145 as the immunogen showed similar results (DR15^{+.Fcgr2b−/−}, n=6 [Treg intact], 7 [Treg depleted]; DR1^{+.Fcgr2b−/−}, n=4 [Treg intact], 5 [Treg depleted]; and DR15⁺DR1^{+.Fcgr2b−/−}, n=6 per group), with the emergence of autoimmune anti-GBM glomerulonephritis in DR15⁺DR1^{+.Fcgr2b−/−} mice only after Treg depletion. Values are mean±s.e.m.; ^*P < 0.05; ^** P < 0.01; ^*** P < 0.001 by Mann-Whitney U test (a–d).

Extended Data Figure 7. Single cell sequencing data from healthy humans and Goodpasture’s patients validate and corroborate findings in HLA transgenic mice

a, Comparing the TCR usage of α3_135-145-specific CD4⁺ T cells using the HLA-DR15-α3_135-145 tetramer and the HLA-DR1-α3_135-145 tetramer by TCR single cell sequencing, TRAV (DR15, n=20; DR1 n=28), TRBV (DR15, n=20; DR1 n=24), TRAJ (DR15, n=20; DR1 n=28) and TRBJ (DR15, n=20; DR1 n=24). For each TCR type/region (TRAV, TRBV, TRAJ, TRBJ), we compared the TCR distribution (frequencies of different TCRs) between DR15 and DR1. ^*P < 0.05 by Fisher’s exact test. The full HLA-types of the DR15 homozygous donor (HD1) and the DR1 homozygous donor (HD7) are listed in Extended Data Table 3. b, In anti-GBM patients, α3_135-145-specific T cells clonally expand. α3_135-145-specific CD4⁺ T cells from the blood of two anti-GBM patients were single cell sorted using HLA-DR15-α3_135-145 tetramer by flow cytometry. TCR genes were amplified by multiplex PCR, then sequenced to determine TRAV (n=12 and 45) and TRBV usage (n=16 and 44) and CDR3 amino acid sequence. Differently coloured slices within the pie chart highlights the TRAV or TRBV in which repeated CDR3 sequences were found. All CDR3 sequences within the highlighted TRAV or TRBV are shown in coloured letters (1 row = 1 CDR3 sequence).

Figure 1. α3_135-145 induces nephritogenic autoimmunity, but not when DR1 is co-expressed

a, α3_135-145-specific Foxp3⁻ effector CD4⁺ T cells in DR15⁺ healthy humans (n=7), and Goodpasture’s (GP) patients (n=8). b, α3_135-145-specific effector CD4⁺ T cells in kidneys from α3_135-145-immunized DR15^{+.Fcgr2b−/−} mice (n=5 each group). c, HLA-DR1 co-expression prevents pro-inflammatory autoreactivity to the GP antigen in α3_135-145-immunized DR15⁺DR1^+.Fcgr2b+/+ mice (n=4 each group). d, Peptide repertoire analyses of HLA-DR presented self-peptides in DR15⁺ and DR1⁺ transgenic mice. Enriched and depleted amino acids are above and below the horizontal bar, respectively, scale of letter proportional to frequency difference. Values are mean±s.e.m.; ^*P<0.05; ^**P<0.01; ^***P<0.001 by Mann-Whitney U test (a, b) or Kruskal-Wallis test (c).

Figure 2. Presentation of α3_135-145 by HLA-DR15 and HLA-DR1

a, Polymorphisms (red) with the β-chains of HLA-DR15 (light blue) and HLA-DR1 (light yellow). b–f, HLA-DR15-α3_135-145 and HLA-DR1-α3_135-145 with HLA-DR α-chain (light green). Black, red and brown dashes represent H-bonds, salt-bridges, and peptide pocket positions, respectively. b, HLA-DR15-α3_135-145 c, HLA-DR-α3_135-145. d, HLA-DR15 and HLA-DR1 alignment. e, HLA-DR15 and DR1 polymorphisms (f) and interactions with α3_135-145. g, TRAV and TRVB chain usage of α3_135-145-specific CD4⁺ T cells in HLA-DR transgenic mice (pooled samples, n=3 mice each group), and phenotyping for Vα2 (TRAV14) and Vβ6 (TRBV19). Frequencies of TCRs between DR15 and DR1 were compared by Fisher’s exact test ^*** P<0.001. For flow cytometry, n=4 individual mice per group, ^*P<0.05, ^*** P<0.001 by unpaired two-tailed t-test.

Figure 3. DR15 selects α3_135-145-specific Tconvs but DR1 selects protective Tregs

a, α3_135-145-specific CD4⁺ T cells from naïve HLA-DR transgenic mice are predominantly Tconv (Foxp3⁻) in the context of HLA-DR15, but Tregs (Foxp3⁺) are selected by HLA-DR1. b, Naive CD4⁺ T cells from DR15⁺ mice with intact Tregs stimulated by α3_135-145 secrete pro-inflammatory cytokines, cells from DR1⁺ and DR15⁺DR1⁺ mice secrete tolerogenic cytokines, but without Tregs, T cells from DR15⁺ mice become autoreactive. Only in cells from mice bearing HLA-DR1 does α3_135-145 induce Treg proliferation. Cells were pooled from three mice of each strain, representative of three independent experiments performed in triplicate. c, In vivo depletion of Tregs results in autoreactivity in immunized DR15⁺DR1^+.Fcgr2b+/+ mice (n=4 each group). Values are mean±s.e.m.; ^*P<0.05, ^**P<0.01, ^***P<0.001 by Mann Whitney U test.

Figure 4. Treg depletion unmasks disease in DR15⁺DR1^{+.Fcgr2b−/−} mice and autoimmunity in humans in vitro

a, Treg depletion before α3_135-145 immunization leads to anti-GBM glomerulonephritis in DR15⁺DR1^{+.Fcgr2b−/−} mice (DR15^{+.Fcgr2b−/−} n=6 each group, DR1^{+ Fcgr2b−/−} n=4 each group and DR15⁺DR1^{+.Fcgr2b−/−} n=8 [Treg intact], 9 [Treg depleted]). Scale bars = 30 μm. b, In the blood of healthy HLA-typed humans, α3_135-145-specific CD4⁺ Tconv cells predominate in the setting of DR15 and Tregs with DR1. c, In humans, only CD4⁺ T cells from healthy DR15 homozygotes (n=6) exhibit pro-inflammatory autoreactivity to α3_135-145 when Tregs are present, while Tregs in DR1 homozygotes (n=5) and DR15/DR1 heterozygotes (n=5) proliferate and secrete tolerogenic cytokines. Values are mean±s.e.m.; ^*P<0.05; ^**P<0.01; ^***P<0.001, by unpaired two-tailed t-test (a, c top panels), Mann Whitney U test (b) and Kruskal-Wallis test (c, bottom panel).