Tolerogenic Immunomodulation by PEGylated Antigenic Peptides

Abstract

Current treatments for autoimmune disorders rely on non-specific immunomodulatory and global immunosuppressive drugs, which show a variable degree of efficiency and are often accompanied by side effects. In contrast, strategies aiming at inducing antigen-specific tolerance promise an exclusive specificity of the immunomodulation. However, although successful in experimental models, peptide-based tolerogenic "inverse" vaccines have largely failed to show efficacy in clinical trials. Recent studies showed that repetitive T cell epitopes, coupling of peptides to autologous cells, or peptides coupled to nanoparticles can improve the tolerogenic efficacy of peptides, suggesting that size and biophysical properties of antigen constructs affect the induction of tolerance. As these materials bear hurdles with respect to preparation or regulatory aspects, we wondered whether conjugation of peptides to the well-established and clinically proven synthetic material polyethylene glycol (PEG) might also work. We here coupled the T cell epitope OVA^323-339 to polyethylene glycols of different size and structure and tested the impact of these nano-sized constructs on regulatory (Treg) and effector T cells in the DO11.10 adoptive transfer mouse model. Systemic vaccination with PEGylated peptides resulted in highly increased frequencies of Foxp3⁺ Tregs and reduced frequencies of antigen-specific T cells producing pro-inflammatory TNF compared to vaccination with the native peptide. PEGylation was found to extend the bioavailability of the model peptide. Both tolerogenicity and bioavailability were dependent on PEG size and structure. In conclusion, PEGylation of antigenic peptides is an effective and feasible strategy to improve Treg-inducing, peptide-based vaccines with potential use for the treatment of autoimmune diseases, allergies, and transplant rejection.

Keywords: autoimmunity; immune tolerance; nanoparticles; peptide vaccination; regulatory T cells.

Figure 1

Stimulatory capacity of PEGylated peptides in vitro. (A) Representative histograms of CFSE dilution following stimulation with 6 nM pOVA or equimolar amounts (based on peptide amount) of PEGylated peptides. (B) Dose-response curves of peptide and conjugates. CFSE labeled OVA-specific CD4⁺ T cells from DO11.10 mice were cultured with irradiated APCs (1:3) for four days. pOVA and equimolar amounts of pOVA-PEG conjugates (based on peptide amount) were added as indicated. For analysis of proliferation, cells were gated on OVA-TCR (KJ1.26)⁺ CD4⁺ cells and the geometrical mean of the fluorescence intensity (GMFI) of CFSE was determined. Fold CFSE dilution was calculated as: GMFI (control)/GMFI (sample); log₂-scale. One log₂ step is equivalent to one cell division. Data are means of triplicates ± SD and representative for at least three independent experiments. (C) To exclude unspecific inhibitory effects of the PEG polymer, OVA-specific CD4⁺ T cells were stimulated in vitro with 100 nM pOVA in presence or absence of 100 nM of the irrelevant pMOG-PEG20. After 6 days, proliferation was assessed by flow cytometry. n = 3–6 from two independent experiments. Statistical testing was performed using the nonparametric Mann Whitney test and Holm-Bonferroni correction for multiple comparisons. ns, non-significant. For (B), dose response curve fitting and statistical analysis is provided in Supplementary Figure 2 .

Figure 2

PEGylated peptides efficiently activate antigen-specific CD4⁺ T cells in vivo. CFSE-labeled CD4⁺ T cells from DO11.10 mice were transferred into BALB/c mice. After 24 h recipients received i.v. PBS (control), indicated amounts of pOVA or equimolar amounts (based on peptide amount) of pOVA-PEG20, pOVA-PEG40 or pOVA-PEG-tetramer. 5 µg of pOVA are equivalent to 2.8 nM (and 250 µg/kg) per mouse. After 6 days, proliferation was assessed by flow cytometry (for representative dot plots, see Figure 4 ). Displayed is the mean x-fold CFSE dilution ± SD gated on OVA-TCR⁺ CD4⁺ splenocytes from at least two independent experiments. n = 4–6; PBS, n = 10.

Figure 3

Conjugation of pOVA to the macromolecular carrier PEG extends peptide bioavailability. (A) Set-up for analysis of the bioavailability. (B) OVA-specific CD4⁺ CFSE-labeled T cells were transferred into BALB/c mice three days after i.v. peptide vaccination with unconjugated pOVA or equimolar amounts of pOVA-PEG conjugates. After 6 days proliferation was assessed by flow cytometry. Cells from spleen and pLN were gated on OVA-TCR⁺ CD4⁺ T cells. Data show mean x-fold CFSE dilution ± SD from at least two independent experiments. n = 6–9. Statistical testing was performed using the nonparametric Mann Whitney test and Holm-Bonferroni correction for multiple comparisons. ns, non-significant; (*) p < 0.05; (**) p < 0.01; (***) p < 0.001.

Figure 4

Increased frequency of regulatory Foxp3⁺ CD4⁺ T cells upon pOVA-PEG20 treatment. CFSE-labeled CD4⁺ cells from DO11.10 mice were transferred into BALB/c mice. After 24 h, mice received either PBS (control), indicated doses of pOVA or equimolar amounts of PEGylated peptides. After 6 days, Foxp3 expression was assessed using flow cytometry. Spleen cells were gated on OVA-TCR⁺ CD4⁺ T cells. (A) Representative FACS dot plots of Foxp3 expression following tolerization with pOVA or pOVA-PEG conjugates at a dose of 5 µg (based on peptide amount). (B) Dose response curve in vivo. Data displayed as mean ± SD of % Foxp3⁺ cells from at least two independent experiments. n = 4–8; PBS, n = 10. (C) Pooled data from treatment with 5 µg of pOVA or equimolar amounts of PEGylated pOVA. Mean ± SD from at least two independent experiments. Statistical testing was performed using the nonparametric Mann Whitney test and Holm-Bonferroni correction for multiple comparisons. (*) p < 0.05; (**) p < 0.01; (***) p < 0.001.

Figure 5

Administration of pOVA-PEG20 leads to de novo induction of Foxp3⁺ Tregs. Treg-depleted CD4⁺ T cells from DO11.10 mice were transferred into BALB/c mice 24 h prior to treatment with PBS (control), 5 µg of pOVA or equimolar amounts of pOVA-PEG20. Six days post peptide vaccination, spleen, pLN, mLN and livLN were isolated and Foxp3 expression was analyzed using flow cytometry. Data were pooled from two experiments and represent mean ± SD of % Foxp3⁺ cells gated on OVA-TCR⁺ CD4⁺ cells. n = 3–7. Statistical testing was performed using the nonparametric Mann Whitney test and Holm-Bonferroni correction for multiple comparisons. (ns) and all comparisons not marked were non-significant. *p < 0.05; **p < 0.01.

Figure 6

Reduced frequency of TNF⁺ among OVA-TCR⁺ T cells upon treatment with PEGylated pOVA. 24 h after transfer of OVA-specific CFSE-labeled CD4⁺ T cells, recipients were treated i.v. with PBS (control), pOVA or equimolar amounts of pOVA-PEG conjugates. After 6 days, TNF production was assessed by intracellular staining of restimulated cells. Splenocytes were gated on OVA-TCR⁺ CD4⁺ cells. (A) Representative FACS dot plots of TNF expression following tolerization with pOVA or pOVA-PEG conjugates at a dose of 5 µg (based on peptide amount). (B) Dose response curve. Mean ± SD of % TNF⁺ from at least two independent experiments. n = 4–6; PBS, n = 10. (C) Data for treatment with 5 µg of pOVA or equimolar amounts of PEGylated pOVA. Mean ± SD of % TNF⁺ cells from at least two independent experiments. Statistical testing was performed using the nonparametric Mann Whitney test and Holm-Bonferroni correction for multiple comparisons. ns, non-significant, **p < 0.01; ***p < 0.001.

Figure 7

Reduced TNF/Foxp3 ratio upon treatment with pOVA-PEG conjugates. 24 h after transfer of OVA-TCR⁺ CFSE-labeled CD4⁺ T cells, recipients received PBS (control), 5 µg of pOVA or equimolar amounts of different pOVA-PEG conjugates. After 6 days, TNF production and Foxp3 expression were assessed using flow cytometry for spleen cells. The TNF/Foxp3 ratio gated on OVA-TCR⁺ CD4⁺ cells was calculated, reflecting the effector T cell/Treg ratio of responding cells. Mean ± SD of the ratio % TNF⁺/% Foxp3⁺cells from at least two independent experiments; n = 5–8; PBS, n = 10. Statistical testing was performed using the nonparametric Mann Whitney test and Holm-Bonferroni correction for multiple comparisons. **p < 0.01; ***p < 0.001.

Figure 8

Administration of pOVA-PEG20 reduces the frequency of OVA-TCR⁺ CD4⁺ T cells among total CD4⁺ splenocytes. OVA-specific CD4⁺ cells were transferred into BALB/c mice. After 24 h, mice received i.v. PBS (control), 5 µg of pOVA or equimolar amounts of pOVA-PEG conjugates. After 6 days, the frequency of OVA-TCR⁺ CD4⁺ cells among total CD4⁺ T cells was assessed using flow cytometry. Mean ± SD of % KJ1.26⁺ cells (OVA-TCR⁺ T cells) among total CD4⁺ splenocytes from at least two independent experiments. n = 5–11; PBS, n = 13. Statistical testing was performed using the nonparametric Mann Whitney test and Holm-Bonferroni correction for multiple comparisons. ns, non-significant, *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 9

pOVA-PEG20 vaccination in presence of pre-existing T effector cells. In vitro generated OVA-specific Th1 cells were labeled with CFSE and transferred into BALB/c mice. After 24 h, recipients received i.v. PBS (control), 5 µg of pOVA or equimolar amounts of pOVA-PEG20. After four days, splenocytes were re-isolated and proliferation, IFN-γ and TNF production as well as Foxp3 expression was assessed using flow cytometry. (A) Proliferation (x-fold CFSE dilution). (B) OVA-TCR⁺ T cells among total CD4⁺ cells. (C) % IFN-γ producers among OVA-specific CD4⁺ cells. (D) % Foxp3⁺ cells among OVA-specific CD4⁺ cells. (E, F) Absolute number of antigen-specific IFNγ or Foxp3⁺ cells expressed as % of total CD4⁺. (A–F) Pooled data with mean ± SD from two independent experiments; n = 6. Non-marked comparisons are non-significant. Statistical testing was performed using the nonparametric Mann Whitney test and Holm-Bonferroni correction for multiple comparisons. (*) p < 0.05; (**) p < 0.01.

Figure 10

Administration of 50 µg of LPS does not enhance secretion of pro-inflammatory cytokines but inhibits induction/expansion of Foxp3⁺ Tregs upon treatment with pOVA-PEG20. 24 h after adoptive transfer of OVA-specific CD4⁺ T cells, recipients received i.v. PBS (control), 5 µg of pOVA or equimolar amounts of pOVA-PEG conjugates with or without additional LPS. After 6 days, splenocytes were isolated and Foxp3 expression as well as TNF production was assessed by flow cytometry. (A) % TNF⁺ cells gated on OVA-specific CD4⁺ cells, mean ± SD of data from three independent experiments. (B) % KJ1.26⁺ (OVA-specific) splenocytes among total CD4⁺ population; mean ± SD of pooled data from three independent experiments. (C) % Foxp3⁺ cells gated on OVA-specific CD4⁺ cells; mean ± SD of pooled data from three independent experiments. n = 5–6; pOVA-PEG20, n = 8. Statistical testing was performed using the nonparametric Mann Whitney test and Holm-Bonferroni correction for multiple comparisons. ns, non-significant, **p < 0.01; ***p < 0.001.

Figure 11

Vaccination by either pOVA or pOVA-PEG20 confers partial tolerance in a Th1-driven DTH model. OVA-specific CD4⁺ cells were adoptively transferred into BALB/c mice, which were treated i.v. with PBS (control), 5 μg of pOVA or equimolar amounts of pOVA-PEG20 24 h later. On day 7, in vitro generated OVA-specific Th1 cells were transferred into the recipients. After 24 h, pOVA in IFA or the control, PBS/IFA, were injected into the left or right hind footpad, respectively. The difference in footpad thickness after challenge was monitored for eight consecutive days. Statistical testing was performed using the nonparametric Mann Whitney test and Holm-Bonferroni correction for multiple comparisons. Difference between pOVA-PEG20 and pOVA was significant with adjusted p = 0.03. Data represent mean of Δ footpad thickness ± SEM (n = 15).

Figure 12

Non-hematopoietic APCs present pOVA-PEG20 and induce activation of OVA-specific T cells at high doses. OVA-specific CD4⁺ MHCII^-/^- cells, isolated from OTIIxB6.PL (CD90.1⁺) mice, were labeled with CFSE and transferred into controls or MHCII deficient BM (bone marrow) chimeras. After 24 h, recipients received 5 or 50 µg (based on peptide amount) of pOVA-PEG20. Proliferation as well as Foxp3 expression was assessed using flow cytometry on day 7. Splenocytes were gated on donor (CD90.1⁺) OVA-specific CD4⁺ T cells. Representative FACS dot plots from at least two independent experiments. wt: wild type.