Histone Peptide-Induced Nasal Tolerance: Suppression of Murine Lupus¹

Mucosal tolerance describes a state of lymphocyte hyporesponsiveness to protein Ags applied across mucosal surfaces by oral or nasal instillation. Inducing specific mucosal immune hyporesponsiveness to self Ags associated with autoimmune syndromes (1, 2, 3, 4, 5, 6) offers the basis of an appealing therapy with the potential to replace current nonspecific immunosuppressive drugs that may compromise normal effector functions of lymphocytes involved in immune surveillance and protective immunity.

The efficacy of mucosal tolerance induction can be dependent on a number of factors such as instilled Ag dose and solubility. It has been shown that high doses of self Ag are able to delete Ag-specific autoreactive T cells (7) or conversely, to break tolerance that has been observed at lower doses (8). Several studies of mucosal tolerance in animal models have demonstrated that this hyporesponsiveness is T cell mediated and typically results in a switch from an inflammatory Th1 T cell subset response (IFN-γ, IL-2) to a T cell response associated with the production of suppressive or noninflammatory cytokines (9, 10). T cell subsets that have been associated with mucosal immune hyporesponsiveness include the Th2 (IL-4, IL-10) (11), Th3 (TGF-β) (12), T regulatory 1 (IL-10, TGF-β) (13), and Th0 (IL-4, IL-10, IFN-γ) cells (14) and CD4⁺ CD25⁺ T regulatory (T reg)⁴ cells (15). Ag-specific serum Ab levels may remain relatively high after tolerization, but typically there is an IgG subclass switch to an effector isoform less likely to promote an inflammatory response (1). Finally, in animal models of mucosal tolerance in which the subjects are typically tolerized before disease induction, there is usually a significant reduction in the quality and measure of pathology. This advantage of preimmunization tolerance is, however, not applicable to the clinical situation in which researchers have attempted, to date with mixed results, to tolerize patients who present with chronic autoimmune disease, and this remains a significant challenge.

Mononucleosomes, single units of the nucleosome, consist of 147-bp dsDNA wrapped around a core octamer of histone proteins: two of each histones H2a, H2b, H3, and H4 (16). Interest in the nucleosome as a notable autoantigen of systemic lupus erythematosus (SLE) has been brought about by a series of individuals with convergent observations, which suggest that the nucleosome is an important adjuvant in the early development of pathogenic anti-DNA autoantibody response associated with this syndrome (17, 18). Nucleosomes are major components of apoptotic bodies, and defective clearance mechanisms (19, 20) may result in their presence as autoantigens in lupus models several weeks before the appearance of anti-DNA Abs (18). Kaliyaperumal et al. (21) have demonstrated that a panel of pathogenic autoantibody-inducing Th clones derived from lupus-prone mice (22) can be induced to proliferate upon presentation of histone-derived peptides. Furthermore, upon i.p. challenge into young prenephritic female SNF₁ mice, these peptides can induce and promote the production of pathogenic anti-DNA Abs and subsequently accelerate the onset of severe nephritis. In this study, we have synthesized one of the histone peptides identified by Kaliyaperumal et al., which we term H471 (derived from histone H4, position 71–93). In addition, mononucleosomes were purified from chicken RBCs. We have used the H471 peptide to induce tolerance to itself and mononucleosomes by nasal instillation of lupus-prone (NZB × SWR)F₁ (SNF₁) (H-2^d×q) and nonautoimmune, but H-2-identical (BALB/c × SWR)F₁ (BSF₁) mice with the peptide. We report that tolerance was induced in mice nasally dosed with the H471 peptide. Nasal tolerance to the H471 peptide can be achieved even after Ag priming. Furthermore, nasal instillation of mice with the H471 peptide can suppress the immune response to mononucleosomes. Chronic nasal instillation of lupus-prone SNF₁ mice with H471 can suppress the development of autoantibodies and reduce the incidence of severe glomerulonephritis. These findings demonstrate a clear potential for immune intervention in SLE with a peptide expressing a T cell epitope from a self protein.

SNF₁ and BSF₁ mice were bred and maintained in the animal facility at King’s College London. SNF₁ mice used in the long-term study were purchased from Harlan (Bicester, Oxfordshire, U.K.) and maintained in our animal facility. Female mice between 8 and 10 wk old were used in all experiments, unless stated otherwise.

Synthetic peptide H471 based on the amino acid sequences of histone protein H4 at positions 71–93 (TYTEHAKRKTVTAMDVVYALKRQ) was synthesized as a peptide amide using a Rink linker 4′ methylbenzhydrylamine resin and Fmoc-based solid phase peptide synthesis automated by a PS3 peptide synthesizer (Anachem, Luton, U.K.). The purity of the H471 peptide was shown to be 96% by reverse-phase HPLC, and its composition was confirmed by mass spectrometry. Control peptides, OVA_323–339 and H416 (positions 16–39 of H4), were synthesized in the same way, and shown to be >95% pure. Individual core histone proteins were purchased from Boehringer (Sussex, U.K.), and the manufacturer guaranteed 99% purity. Mononucleosomes were prepared using a protocol kindly given by K. E. van Holde (Oregon State University, Corvallis, OR). The protocol has also been published previously (23) and briefly, chicken RBC lysate was treated with micrococcal nuclease (45 IU ml⁻¹; Sigma-Aldrich, Poole, U.K.) for 5 min at 37°C. The resulting nuclei were lysed by resuspension in 10 vol 0.25 mM EDTA, pH 7.5 (Sigma-Aldrich), followed by gentle stirring for 1 h in 35 vol 0.25 mM EDTA, and debris were eliminated by centrifugation for 20 min at 8000 × g. Histones H1 and H5 were removed by adding the supernatant containing chromatin to a Carboxymethyl Sephadex C25 column (Amersham Pharmacia, Buckinghamshire, U.K.). This was followed by redigestion of the chromatin with micrococcal nuclease, and mononucleosomes were obtained by running the redigested chromatin through an S-300 Sephacryl column (Pharmacia). Peaks containing mononucleosomes with 146-bp DNA were detected and retrieved by monitoring at A₂₆₀. The presence of the four core histones (H2a, H2b, H3, and H4) in the mononucleosome preparation was confirmed on a 15% SDS-PAGE. The length of DNA in the mononucleosome was shown to be between 100 and 150 bp on a 2% agarose gel (data not shown).

Mice were nasally dosed with peptide Ag dissolved in PBS for 5 consecutive days (days −12 to −8). In the control group, mice received OVA_323–339 peptide in PBS. On day 0, each mouse was immunized intradermally with 100 μg peptide Ag or mononucleosomes emulsified in CFA (Sigma-Aldrich). Alternatively, each mouse was immunized intradermally with 100 μg peptide Ag emulsified in CFA, and from day 7–11, mice were nasally dosed with peptide Ag dissolved in PBS. In long-term disease studies, autoimmune but prenephritic SNF₁ females at 4 wk old (25 mice per group) were nasally dosed with H471 or OVA_323–339 peptide (20 μg divided equally over 5 consecutive days) at 2-wk intervals until they reached 32 wk of age.

For anti-peptide Abs, 96-well Falcon 3912 microtiter plates (Marathon, London, U.K.) were coated with 50 μl/well H471 at 10 μg/ml in deionized water by evaporation overnight at 37°C. For anti-dsDNA and anti-mononucleosome Abs, dsDNA or mononucleosomes at 20 μg/ml in PBS were coated onto poly(l-lysine) (Sigma-Aldrich)-sensitized Nunc Maxisorb 96-well plates (Life Technologies, London, U.K.). Nonspecific Ab binding sites were blocked by adding 50 μl/well 2% Marvel in PBS. A total of 50 μl/well mouse serum diluted 1/100 in PBS containing 0.05% Tween 20 (Sigma-Aldrich) and 1% Marvel (P-T-M) was coated on plates by incubation at 37°C for 1 h. For IgG Ab detection, 50 μl/well HRP-conjugated IgG goat anti-mouse Ab (Serotec, Oxford, U.K.) diluted 1/1000 in P-T-M was added and incubated for 1 h. Abs bound to Ags coated on the plate were detected by the addition of 50 μl/well o-phenylenediamine (BDH, Poole, U.K.) at 1 mg/ml in 0.01 M citrate phosphate buffer (pH 6) containing 0.001% H₂O₂ (Sigma-Aldrich). Finally, the enzyme reaction was stopped by the addition of 50 μl/well 0.5 M citric acid, and absorbance reading was taken at 450 nm.

For IgG subclass Ab detection, rat mAbs to mouse IgG1, IgG2a, IgG2b, and IgG3 (Serotec) diluted 1/750 in P-T-M were used. This was followed by the addition of goat anti-rat biotinylated Abs (Serotec) diluted 1/2000 in P-T-M. Finally, streptavidin complex with HRP (Sigma-Aldrich) diluted 1/1000 in P-T-M was added. Bound Abs were detected using o-phenylenediamine, and absorbance readings were measured as above.

Ten days following immunization, lymph nodes (LN; axial, bronchial, and inguinal) were removed aseptically from each mouse. A total of 3 × 10⁵/200 μl/well LN cells was cultured in triplicate with or without Ag in 96-well flat-bottom microtiter plates (Life Technologies) for 96 h at 37°C with 5% CO₂ in a humid incubator. Tissue culture medium used was IMDM supplemented with 1% l-glutamine, 2% penicillin and streptomycin (all from Life Technologies), and 1% normal mouse serum. Cultures were pulsed with 0.25 μCi tritiated thymidine ([³H]d Thd; Amersham, Buckinghamshire, U.K.) for the last 6 h. [³H]d Thd incorporation was measured using a liquid scintillation beta counter (Microbeta; Wallac, Milton Keynes, U.K.). Cell proliferation was expressed in Δ cpm or stimulation index.

LN cells (3 × 10⁵/200 μl/well) from mice of each experimental group were pooled and cultured in the presence of Ag at 37°C in a humid incubator. In vitro production of cytokines (IFN-γ, IL-2, IL-4, and IL-10) by cell cultures was measured using CelELISA technique adopted from Beech et al. (24). Briefly, 2 h before transfer of cell cultures, 50 μl/well anti-cytokine Abs (BD Biosciences, Oxford, U.K.) at 2 μg/ml in 0.05 M sodium bicarbonate buffer, pH 9.6, were coated onto sterile 96-well flat-bottom microtiter plates at 37°C for 1 h. This was followed by twice washing in PBS, then blocking of nonspecific sites using PBS containing 1% BSA at 37°C for 1 h. Following incubation, plates were washed twice with PBS, and 100 μl cell cultures were transferred to anti-cytokine Ab-sensitized plates and incubated for 24 h at 37°C. Two hours before harvest, doubling dilutions of cytokine standards (BD Biosciences) starting from 1200 pg/ml were made on the plate and incubated at 37°C. This was followed by washing in PBS containing 0.1% Tween 20 (PBS/Tween) and the addition of 50 μl/well biotinylated detecting Abs at 2 μg/ml in 1% BSA/PBS and incubation at 37°C for 1 h. Finally, bound Abs were detected by the addition of streptavidin-conjugated HRP. The enzyme reaction was developed using 100 μl/well tetramethylbenzene (Sigma-Aldrich) at 1 mg/ml in citrate phosphate buffer and stopped by the addition of 50 μl/well 2 M H₂SO₄. Absorbance readings were taken at 450 nm.

Mouse kidneys were processed through to paraffin wax. Kidney sections (5 μm) were mounted on glass slides, air dried, and dehydrated with 100% ethanol before being exposed to 1% periodic acid for 5 min. Sections were then stained with Schiff’s reagent (BDH) for 30 min, followed by brief washing with water and a gradient of ethanol (70–100%) before being mounted in xam. All slides were read by light microscopy (×400) as coded specimens by an independent observer. Glomerular lesions were graded as 1+, 2+, 3+, or 4+ (25). A 1+ lesion represents minor (>10%) thickening of the capillaries of the glomeruli; a 2+ lesion corresponded to focal or focal and diffuse thickening of the capillaries of 30–60% of the glomeruli; in a 3+ lesion, the capillaries of all glomeruli were affected; 4+ lesions were characterized by the preceding together with sclerosis of glomeruli, massive proliferation of epithelial cells, and numerous tubular casts. A grade of 0 was given to kidneys without glomerular lesions. Kidneys from 4-wk-old unmanipulated SNF₁ mice were used as negative controls.

The tolerogenicity of the H471 peptide was tested in lupus-prone SNF₁ mice. Three dosages (5, 20, and 100 μg) of the H471 peptide were tested. Mice in the control group were nasally dosed with 20 μg OVA_323–339. Mice were nasally dosed with peptide Ags before receiving intradermal immunization with H471. Results show that LN cells from mice nasally dosed with 20 μg H471 produced significantly (p < 0.04, t test) lower proliferative responses to 10 μg/ml H471 in vitro than LN cells from control mice (Fig. 1)_. However, LN cell hyporesponsiveness was not seen with mice nasally dosed with 5 μg H471 and less apparent with mice nasally dosed with 100 μg H471. Furthermore, the hyporesponsiveness of LN cells was overcome by increasing Ag stimulation to 100 μg/ml (Fig. 1). Thus, the 20 μg peptide dosage was the most effective in inducing hyporesponsiveness of LN cells, and this was reiterated in a second experiment with groups of six SNF₁ mice (Fig. 2,A). It is demonstrated in this study that the mean proliferative response of the H471 nasally dosed group was significantly reduced (p < 0.003, t test) compared with the control group (OVA_323–339). Again, this apparent reduction in cell proliferation was overcome, but still significantly (p < 0.02, t test) lower than the control group, by increasing Ag stimulation in vitro (Fig. 2 B). Therefore, a dose-dependent nasal tolerance was induced in SNF₁ mice, and the level of LN cell hyporesponsiveness was dependent on the Ag concentration in vitro. Experiments presented in this work were repeated twice in SNF₁ mice as well as BSF₁ mice (data not shown), and similar observations were obtained on each occasion.

In subsequent experiments, we addressed the question of whether tolerance can be induced in Ag-primed mice. Groups of six SNF₁ mice were immunized intradermally with 100 μg H471 in CFA before being nasally dosed with a total of 20 μg of either H471 or OVA_323–339 peptide. Mouse LN cells were cultured with H471 or OVA_323–339 peptide. After 96 h, LN cells from H471 nasally dosed mice proliferated significantly (p < 0.02, t test) less than LN cells from OVA_323–339 peptide nasally dosed mice (Fig. 3,A). Like preimmunization tolerance, LN cell hyporesponsiveness was overcome by increasing Ag concentration in vitro (Fig. 3 B; p < 0.003, t test). Data shown in this work are highly reproducible in repeat (3×) experiments in SNF₁ as well as BSF₁ mice (data not shown). This is the first recorded observation to show that ongoing autoimmune responses in lupus-prone mice can be suppressed by nasal treatment with a self Ag-derived peptide.

The H471 peptide is derived from the core histone protein H4 of mononucleosome. Therefore, we addressed the question of the effect of nasal dosing of mice with H471 on the immune response to mononucleosomes. Groups of six SNF₁ mice were nasally dosed with a total of 20 μg H471 or OVA_323–339 peptide before receiving intradermal injection of 100 μg mononucleosomes emulsified in CFA. Mouse LN cells were cultured in the presence of 10, 100, or 200 μg/ml mononucleosomes, H471, or OVA protein (fraction V; Sigma-Aldrich). Optimal cell proliferation was observed at 100 μg/ml mononucleosomes in vitro. In fact, increasing Ag concentration to 200 μg/ml inhibited cell proliferation (data not shown). Suppression of cell proliferative response to mononucleosome in vitro was seen in the H471 nasally dosed animals (Fig. 4,A). Suppression of cell proliferation was shown to be significant (p < 0.05, t test) when the mean proliferation counts of the experimental groups were plotted (Fig. 4 B). Similar data were obtained in two repeat experiments. Thus, nasal instillation of SNF₁ mice with H471 can suppress the immune response to mononucleosomes that carry the H471 epitope.

To examine the mechanism(s) of nasal tolerance, we measured the level of cytokines produced in vitro by cells from mice nasally dosed with H471 or OVA_323–339. Therefore, LN cells were pooled from mice nasally dosed with H471 or OVA_323–339 before being immunized with H471 or mononucleosomes. The levels of IFN-γ, IL-2, IL-4, and IL-10 were measured on day 3 of culture. Fig. 5 A shows that cells from mice nasally dosed with OVA_323–339 before being immunized with H471 produced significantly higher levels of IL-2 (p < 0.005, t test) and IFN-γ (p < 0.001, t test) in response to 100 μg/ml H471 challenge compared with cells from mice nasally dosed with H471. The levels of IL-4 and IL-10 produced by cells were similar in both peptide nasal dose groups and not significantly above the background levels. These findings were also observed at 10 μg/ml H471 stimulation in vitro (data not shown).

However, cytokines produced by cells from mice nasally dosed with peptide before being immunized with mononucleosomes were different from their peptide immune counterparts. Fig. 5 B shows that a significantly higher level of IL-10 (p < 0.001, t test) was produced by cells from mice nasally dosed with H471, but not with OVA_323–339 in response to 100 μg/ml mononucleosome challenge. This was not seen at 10 μg/ml mononucleosomes in vitro (data not shown). Cells from mice nasally dosed with OVA_323–339 produced predominantly IFN-γ. The levels of IL-2 and IL-4 produced by cells from either peptide nasal dose group were similar and not significantly above the background levels. Results in this study demonstrate that the nature of the Ag encountered by lymphocytes and consequently the immune response generated can determine the form of tolerance in terms of cytokine production. Student’s t tests were performed on triplicate data points. The experiments shown in this work were repeated once in SNF₁ mice, and similar results were obtained.

Blood was collected from mice nasally dosed with peptide before being immunized with H471, and the levels of total IgG as well as IgG subclass anti-H471 peptide Abs were examined. No significant difference in the level of serum IgG anti-H471 peptide Abs was found between the two experimental groups. However, mice nasally dosed with H471 produced significantly (p < 0.03, t test) higher level of IgG1 (Th2) anti-peptide Abs than mice nasally dosed with OVA_323–339, which produced predominantly IgG2a (Th1) Abs (p < 0.05, t test) (Fig. 6). Therefore, the IgG1-IgG2a ratios of the H471 and OVA_323–339 nasal treatment groups are 2 and 0.6, respectively. This is a 3.3-fold shift in IgG subclass Ab response from a Th1 to Th2 phenotype as a result of nasal dosing of SNF₁ mice with the H471 peptide.

It was observed that nasal dosing of mice with H471 can suppress not only cellular and Ab responses, but spontaneous autoimmunity was also halted by chronic exposure of lupus-prone SNF₁ mice to H471. For each test Ag, proliferation assays were performed on LN cells of five mice randomly chosen from each peptide treatment group (Fig. 7). Cells from mice that received the irrelevant OVA_323–339 peptide intranasally produced significantly (p < 0.001, t test) higher response to mononucleosomes in vitro (Fig. 7,A). This response was 7-fold higher then cells from mice nasally dosed with H471 in the presence of 100 μg/ml mononucleosomes. The suppression was not seen at 10 μg/ml mononucleosomes in vitro (Fig. 7,B). Proliferative responses to H471, OVA_323–339, and a peptide representing a second T cell epitope in H4 (21, 26), which we named H416, were minimal and did not reach statistically significant levels. The presence of autoantibodies was also examined in mouse serum. Results suggest that nasal dosing of mice with H471 can halt or at least delay the appearance of IgG anti-mononucleosomes (p < 0.03, t test) and anti-dsDNA autoantibodies (p < 0.002, t test) (Fig. 8,A). Examination of the IgG subclass anti-mononucleosome Abs did not reveal significant differences between the two nasal treatment groups (Fig. 8 B). Data shown in this work demonstrate that chronic nasal dosing of lupus-prone mice with the H471 peptide can suppress the development of autoimmune responses in these mice.

Mice were considered to have severe glomerulonephritis when their kidneys showed a 3+ to 4+ grade of nephritis by histopathology. Mice chronically exposed to the H471 peptide intranasally had a 10% incidence of severe glomerulonephritis at 32 wk of age, whereas mice nasally dosed with the OVA_323–339 peptide had a 50% incidence of severe glomerulonephritis. This demonstrates that nasal tolerance induced with the H471 peptide can reduce the incidence of severe glomerulonephritis in lupus-prone SNF₁ mice.

The establishment of nucleosome as a major autoantigen involved in the etiology of lupus (18), and the identification of critical T cell epitopes in histones (21) have shed new light on the development of peptide-based therapeutic treatment for SLE. In agreement with previous observations (21), we also found that the H471 peptide expressing a dominant T cell epitope in histone H4 induced strong Ab and cell proliferative responses upon intradermal injection in lupus-prone and H-2 identical, but nonautoimmune mice (data not shown). Immunization of SNF₁ mice with the H471 peptide augmented the production of IgG anti-dsDNA Ab (21). We have confirmed these observations in our experiments (data not shown). These findings are consistent with the notion that a single pathogenic lupus Th clone can provide help for a dsDNA-, ssDNA-, histone-, and nucleosome-specific B cell (22). The H471 peptide may represent one of the T cell epitopes involved in this T and B cell cognate interaction that can lead to disease pathology. This led us to believe that suppression of responses to the H471 peptide may alter the course of disease development and/or reduce the severity of pathology.

We attempted to induce tolerance in SNF₁ mice by delivering the H471 peptide across the nasal mucosa. Initially, we tested three peptide dosages, 5, 20, and 100 μg, given intranasally over 5 consecutive days. Initial results showed that a suboptimal dose of 5 μg does not induce tolerance (Fig. 1) and in fact, in some experiments, cells from mice that received this small dose of peptide were shown to be more readily stimulated by the H471 peptide than cells from control mice (data not shown). Derry et al. (27) and others in our lab also recently showed that suppression of collagen-induced arthritis by nasal instillation of DBA/1 mice with type II collagen was dose dependent, and the lowest dose of type II collagen (5 μg) in fact aggravated disease in mice. These findings are inconsistent with previous observations made by Weiner et al. (28), which suggested that low doses of Ag administered across mucosal surfaces induce active suppression of immune responses through the production of TGF-β by Th3 cells. However, tolerogenicity may vary greatly between different Ags, and this could be a reason for the discrepancies observed in these studies.

The induction of tolerance was also less effective with a high dose of peptide (100 μg) (Fig. 1). We analyzed the 100 μg peptide dosage with scanning electron microscope. Although each peptide solution was visually soluble, peptide aggregates were observed in the 100 μg peptide solution, while no molecules were observed in the other (5 and 20 μg) peptide solutions or PBS (data not shown). This demonstrated that the H471 peptide formed aggregates at high concentrations. Therefore, when the high concentration H471 solution was applied across the nasal mucosa, the peptide aggregates may obstruct peptide uptake or even induce T cell activation. Hence, the physical properties of the peptide Ag and its formulation are important factors that can influence its tolerogenic efficacy.

Following nasal instillation of SNF₁ mice with 20 μg H471, a marked lymphocyte hyporesponsiveness to in vitro H471 stimulation was shown. The reduction of cell proliferation was apparently overcome by increasing the concentration of Ag in vitro. This suggested that the underlying mechanism responsible for this form of peptide-induced tolerance was most likely to be T cell anergy. By definition, anergy represents a state of abolished proliferation and cytokine responses that can be reversed by certain stimuli (29, 30) such as IL-2 (31). The ability of IL-2 to reverse suppression of cell proliferation in mice nasally dosed with H471 was not examined; however, the pattern of H471-induced suppression was similar, in that reduction of cell proliferation was overcome by increasing Ag stimulation in vitro. Similar to our findings, in the study of IL-10-dependent suppression of the proliferative response of PBMC from staphylococcal enterotoxin B-sensitized individuals to staphylococcal enterotoxin B was not observed at low (0.04 ng/ml) or high (3.3 ng/ml) Ag concentrations in vitro (31). Thus, it is hypothesized that following H471 immunization in SNF₁ mice nasally dosed with H471, two distinct H471-specific T cell populations, one anergic and the other pathogenic, were primed. In the presence of a low level (10 μg/ml) of H471 in vitro, competition for H471 and various cell growth factors such as IL-2 between the two T cell populations meant that the reactivity of pathogenic H471-specific T cells was obscured by the action of the anergic T cells; hence, the net result was suppression. However, at a high Ag concentration (100 μg/ml) in vitro, competition for H471 was less, and both pathogenic and anergic H471-specific T cell populations were activated, but the expression of hyporesponsiveness of H471-specific anergic T cells was restricted; hence, reduction of cell proliferation was overcome. There is no published data to support this hypothesis; however, our analysis of cytokine production by cells from mice tolerized to H471 indicates that anergic T cells are indeed involved. It was revealed that while cells from mice nasally dosed with OVA_323–339 produced predominantly Th1 cytokines (IFN-γ and IL-2), cells from mice nasally dosed with H471 did not produce above background levels of any of the cytokines tested (Fig. 5 A).

In animal models of mucosal tolerance in which the subjects are typically tolerized before disease induction, there is usually a significant reduction in the quality and measure of pathology. The advantage of preimmunization tolerance is, however, not applicable to clinical situations such as lupus because it is, as yet, impossible to make a diagnosis before the patient has already developed disease. Thus, suppressing autoimmune reactions in patients who have already developed disease will be required if peptide-induced nasal tolerance is to become a realistic form of therapeutic treatment for humans. With this aim in mind, experiments were conducted to study nasal tolerance in Ag-primed animals. Thus, SNF₁ mice were primed with H471 by intradermal injection before being nasally dosed with H471. Similar to tolerance demonstrated in Ag-naive animals, suppression of cell proliferation was also demonstrated in Ag-primed animals (Fig. 3). This is the first recorded observation to show that ongoing autoimmune responses in lupus-prone mice can be suppressed by nasal instillation of a self Ag-derived peptide. The successful demonstration of tolerance in Ag-experienced mice may have implications for treatment of patients with SLE and other autoimmune conditions.

One of the major obstacles in developing treatments for systemic autoimmune syndromes such as lupus has been to design a therapy that allows the suppression or elimination of a widespread autoimmunity against multiple self Ags and/or tissues. The early appearance of an autoimmune response to nucleosomes in lupus has been suggested to represent the initiation of a wider breakdown of immune regulation against self Ags (18). If this is true, then the suppression of cell proliferative response to nucleosomes induced by nasal dosing of mice with the H471 peptide (Figs. 4 and 7) may prevent or delay the development of autoreactivity against other Ags, hence a reduction of disease pathology. This was indeed proven to be the case in the histopathological study shown in this work. SNF₁ mice chronically treated with the H471 peptide intranasally have shown a significantly lower incidence of severe glomerulonephritis (>3+) compared with mice treated with the control OVA_323–339 peptide (Table I). The significant improvement in the severity of glomerulonephritis in SNF₁ mice as a result of peptide nasal treatment is likely to be associated with the suppression of autoantibody production shown in the disease study (Fig. 8). A recent publication by Kaliyaperumal et al. (26) also showed that suppression of disease in SNF₁ mice could be achieved by i.v. injection of a peptide bearing a different autoepitope in histone H4 (16–39). They showed that i.v. injection with the histone peptide in 3-mo-old prenephritic SNF₁ mice that were already producing pathogenic autoantibodies markedly delayed the onset of severe lupus nephritis. Furthermore, chronic i.v. therapy in mice with established glomerulonephritis prolonged survival and even halted the progression of renal disease. An important disadvantage associated with tolerance induction by i.v. injection is its invasiveness. One would argue that mucosal administration is superior to and less invasive than i.v. as a potential route to reprogram the immune system in autoimmune conditions.

It was noted that, unlike tolerance to the H471 peptide, in which a reduction of cell proliferation was most apparent at 10 μg/ml Ag in vitro, tolerance to mononucleosomes as a result of nasal instillation of mice with H471 was most apparent at 100 μg/ml Ag in vitro. The expression of suppression at a relatively high (100 μg/ml), but not at a low (10 μg/ml) Ag concentration suggests that an active suppression mechanism rather than T cell anergy or deletion may be involved in the tolerance to mononucleosomes. This is because suppression would have been seen at 10 μg/ml Ag stimulation if anergic T cells were involved due to competition between the H471-specific anergic T cells and the mononucleosome-specific pathogenic T cells for Ag and growth factors. The fact that suppression was not seen at 10 μg/ml Ag in vitro suggested that the availability of the H471 epitope was restricted; therefore, only a small number of H471-specific T cells that have suppressive properties can be activated, and this was not enough to suppress a polyclonal T cell reactivity to mononucleosomes, hence no suppression shown overall. However, at a higher Ag concentration (100 μg/ml), a substantial number of H471-specific T suppressor cells were activated, which in turn actively suppress, perhaps through the production of suppressive cytokines such as IL-10 (32), the polyclonal T cell proliferative response to mononucleosomes; hence, the net result was suppression. Suppression would not have been seen at 100 μg/ml Ag in vitro if anergic T cells were involved because T cell proliferative response to mononucleosomes would mask the hyporesponsiveness of H471-specific anergic T cells.

This hypothesis was supported by the cytokines produced by cells from mice tolerized to mononucleosomes. It is shown in this study that H471-induced nasal tolerance to mononucleosomes was associated with an increased production of the antiinflammatory cytokine, IL-10 (Fig. 5 B). This was not observed in control mice nasally dosed with OVA_323–339, which predominantly produced IFN-γ when injected with mononucleosomes. Tolerance to the whole protein by nasal instillation of a peptide derived from the protein was also demonstrated in the study of diabetes in nonobese diabetes mice. It was demonstrated in the study that repeated nasal instillation of mice with a peptide derived from insulin protected them from developing diabetes when challenged with insulin (33). The fact that cytokine production by cells from SNF₁ mice nasally dosed with H471 is different in response to the peptide or the whole protein suggests that the availability of the H471 epitope to T cells determines the phenotype (anergic or regulatory) of such T cells in terms of cytokine production and, consequently, the expression and form of suppression of cellular and Ab responses in SNF₁ mice.

Recently, much attention has been placed on T cells that have regulatory functions in both autoimmune diabetes (34) and intestinal inflammatory disease (35, 36). These functionally specialized T reg cells exist as part of the normal immune repertoire and become anergic and/or suppressive (by producing TGF-β in vivo and IL-10 in vitro) when encountering self Ags. They may thus control the appearance of autoimmunity by keeping the pathogenic responses to self Ags in check (37). The T reg cells are CD4⁺ and CD25⁺ and are dependent on CTLA-4 for their suppressive function (36). The CD4⁺ CD25⁺ CTLA-4-dependent T reg cells may bear similarities with T cells involved in H471-induced nasal tolerance, in that both T cell subsets are able to become anergic and/or produce suppressive cytokines. Studies of CD4 and CD25 expression on H471-specific T cells from mice nasally tolerized to H471 and the effect of CTLA-4 blockade on H471-induced nasal tolerance are being conducted presently in our lab. These studies will help to address whether nasal dosing of the H471 peptide can activate and expand (38) the T reg cell repertoire in lupus-prone mice.

In conclusion, we have demonstrated dose-dependent nasal tolerance to an autoreactive Th cell epitope and to a major autoantigen, nucleosome, in lupus. The induction of nasal tolerance can be achieved in both Ag-experienced and inexperienced mice. The mechanisms of nasal tolerance are likely to include active suppression by T reg cells as well as anergic T cells. Finally, chronic nasal instillation of a peptide representing an autoreactive Th cell epitope in histone H4 can suppress the development of autoantibodies and reduce the incidence of severe glomerulonephritis in lupus-prone mice. These intriguing observations may have important therapeutic implications on the future development of treatments for lupus.

We thank Dr. Stephen J. Thompson for his critical assessment of the manuscript and Alan Maple for his excellent technical assistance in histopathology.