Major peptide autoepitopes for nucleosome-specific T cells of human lupus (original) (raw)
Response of an anti-DNA autoantibody–inducing lupus Th-cell clone to histone peptides. We first tested the response of a well-characterized CD4 T-cell clone, DD2, that was derived from a patient with active lupus nephritis (2, 3). DD2 induces the production of a pathogenic variety of anti-DNA autoantibodies when cocultured with autologous B cells, and it recognizes nucleosomes (3). Like other pathogenic autoantibody–inducing Th clones of human and murine lupus, DD2 bears charged residues in CDR3 loops of its T-cell receptor (TCR) α and β chains, and it uses a member of the Vα8 family (3, 22). As shown in Figure 1, DD2 can be stimulated by histone H4 peptides located in amino acid positions 67–99. The other nucleosomal histone peptides did not stimulate DD2, except for H355–69, which stimulated very weakly. These results confirmed autoantigenic specificity of DD2, as tested previously with whole antigens (3), and demonstrated that the peptide epitope scanning (pepscan) with synthetic histone peptides would work for human lupus T cells.
IL-2 production by anti-DNA autoantibody–inducing T-cell clone DD2 in response to nucleosomal histone peptides (10 μM each) presented by autologous EBV B-cell APCs. Bars represent mean of triplicate values. Values higher than 3 SD (horizontal line) above the mean of background (DD2 + APCs without added peptide produced 132 pg/mL IL-2) were considered as stimulatory. IL-2 production by DD2 in response to anti-CD3 stimulation was 554 pg/mL.
Localization of major peptide epitopes for autoimmune T cells by using short-term CD4 T-cell lines. To narrow down the histone regions, we used short-term CD4 lines first because a large number of cells were required to scan the entire library of overlapping peptides spanning all the nucleosomal core histones. Dose-response curves were determined after localizing stimulatory histone peptides for the respective T-cell lines. A representative example is shown in Figure 2a. On the basis of these studies, we used 10-μM peptides in further experiments. The short-term lupus T-cell lines were also tested for HLA-class II dependence for antigen recognition and were found to recognize their respective peptide epitopes presented by HLA-DR molecules. A representative example is shown in Figure 2b.
(a) IL-2 production by a representative short-term CD4 T-cell line L-EB, with different concentrations of histone H449–63 peptide-pulsed autologous EBV B-cell APCs. Stimulation with 0.01–100 μM peptide showed significant differences (*P < 0.05 at 0.01 μM, Student’s t test) from background (T cells plus APCs cultured without peptide). (b) An example of MHC class II–dependent recognition of nucleosomal peptides. Presentation of H449–63 peptide by autologous EBV B cells to short-term T-cell line L-EB was inhibited mainly by anti-DR mAb.
Altogether, 10 short-term CD4 T-cell lines derived from patients with lupus and 6 such lines derived from healthy donors were tested for IL-2 response to 154 histone peptides presented by autologous EBV B-cell lines. Response of a representative short-term T-cell line (L-SC) derived from a patient with lupus is shown in Figure 3, and the results of short-term T-cell line (N-JV) derived from a normal control donor are shown in Figure 4. All the short-term T-cell lines from normal donors showed very weak and insignificant responses to the histone peptides (Figure 4). By contrast, the lupus T-cell lines, which were also polyclonal, responded strongly to peptides corresponding to certain regions in the core histones (Figure 3). A summary of significant responses of all the short-term T-cell lines derived from 10 different patients with lupus is shown in Figure 5. These pepscan results localize the epitopes recognized consistently and recurrently by autoimmune T cells from different patients (Figure 5). There were no discernible differences between the responses of T-cell lines derived from patients with active lupus and from those with inactive lupus. Results of proliferative responses of the T cells to the nucleosomal peptides were consistent with their IL-2 production responses (data not shown).
Example of a pepscan of L-SC, a short-term CD4 T-cell line from a patient with lupus, with histone peptides presented by autologous EBV B-cell APCs. The baseline for y axis is set at 3 SD above the mean of background values obtained with T cells cultured with APCs alone. IL-2 production responses to peptides above that value are shown.
IL-2 production by N-JV, a short-term CD4 T-cell line from a representative normal donor to histone peptides using autologous EBV B-cell APCs. The horizontal line in each panel demarcates 3 SD above the mean of background values. None of the normal T-cell lines showed a response to any of the peptides that was 3 SD above their respective background values.
Summary of IL-2 production responses of short-term T-cell lines to nucleosomal histone peptides in 10 patients with lupus. The results are expressed as percentages of maximal responses of the T-cell lines stimulated by anti-CD3 mAb. Bars represent mean values from 10 lupus T-cell lines, with error bars in SEM. Values of T cells cultured with autologous EBV B-cell APCs alone were considered as background. The baseline for y axis is set at 22%, which is 3 SD above the mean of background values (14.5%). A peptide was considered stimulatory when it elicited a positive response above this baseline. For instance, the average of responses of all 10 T-cell lines to H416–30 peptide was 23 ± 2.6. All the stimulatory peptides in the panels elicited positive responses in all 10 lupus T-cell lines to levels greater than 3 SD above their respective background values. The extreme right bar in the bottom right (H4) panel was the mean value of responses of the 10 lines to whole nucleosome preparation.
On the basis of the consensus stimulatory epitope regions identified by using the lupus T-cell lines and clone DD2 (Figures 1–5), we selected sixteen 15-mer peptides (Figure 6) for further testing of T cells directly in PBMCs of patients with lupus and of normal subjects. For further studies, we also used 3 longer (24-mer) peptides that overlapped some of these epitopes and were found to be relevant to pathogenic Th and B cells of lupus-prone mice (refs. 7, 23; Figure 6).
A library of histone peptides that correspond to the stimulatory epitope regions identified with lupus T-cell lines and clone DD2. In addition, three 24-mer peptides that overlapped some of these epitopes were used for further studies: H2B10–33 (PKKGSKKAVTKAQKKDGKKRKRSR), H416–39 (KRHRKVLRDNIQGITKPAIRRLAR), and H471–94 (TYTEHAKRKTVTAMDVVYALKRQG). All peptide sequences are identical to those in human histones (GenBank).
Histone peptide epitopes for CD4 T cells in PBMCs of patients with lupus. We wanted to confirm that the peptide autoepitopes (Figure 6) were also relevant to unmanipulated CD4 T cells in PBMCs of patients with lupus. Flow cytometry for newly synthesized, intracellular cytokines is ideal for rapidly detecting antigen-specific responses and determining the frequency of CD4 T cells response to a particular epitope (19–21). Because autoimmune T cells are preactivated in vivo in patients with active disease, we found that activation markers, such as CD69, CD25, or CD79, were not as reliable. We also optimized the intracellular cytokine assay by providing costimulatory signal via CD28, which did not cause any nonspecific background stimulation, as described (19). Moreover, this method could be applied directly to PBMCs without T-cell purification. Intracellular production of newly synthesized IFN-γ, IL-2, IL-10, and IL-4 was assayed in response to the panel of peptide epitopes in Figure 6. As an additional background control, we used the H383–97 peptide, which, although being partly within a stimulatory region in H3 (Figure 6), was not recognized by any of the T-cell lines or clone DD2 (Figures 1–5). Two representative cytokine staining profiles from patients R-WG and R-SC are shown in Figure 7. Cells positive for a particular cytokine form a discrete population; single cells producing more than one cytokine were rare in these fresh PBMCs.
Representative 2-color intracellular cytokine staining of gated CD4+ T cells freshly obtained from 2 patients with lupus who were in remission (R-WG and R-SC). IFN-γ and IL-10 response (a) and IL-2 and IL-4 production (b) in response to stimulation by anti-CD3, nucleosomes, or the histone peptide epitopes. Demarcation of the quadrants was set based on background staining of the T cells cultured in medium alone. Percentage values for the respective cytokine stainings can be found in Figure 8.
The results of intracellular cytokine production response of CD4 T cells in PBMCs from 12 patients with lupus are shown in Figure 8. Eight patients in remission and 4 patients with active disease, designated by prefixes, R- and A-, respectively, were studied. Results from 7 normal controls, designated by N-, are shown in Figure 9. The absolute numbers of CD4 T cells were comparable in the PBMCs of patients with lupus in remission and those from normal subjects. An equal number of events from every sample were analyzed. After doing preliminary studies with test antigens, such as tetanus toxoid, mumps, and candida, and taking into account previous experience in other systems (19–21), we considered a response to the nucleosomal antigens to be unequivocally positive when the frequency of positive cells was greater than 0.2% and the values were at least 2-fold higher than respective background values (“medium”; Figures 8 and 9). Repeat assays with aliquots of cells from the same sample gave almost identical results (SD <5%). Some of the nucleosomal peptides caused impressive stimulation of T cells from the patients with lupus; by contrast, positive responses were rare in the normal subjects (compare Figure 8 with Figure 9). IFN-γ and IL-4 responses to nucleosomes and some of their peptides were greater than 10-fold above background values, and corresponding IL-10 responses were up to 20-fold higher in some of the patients (Figure 8). These results are highly significant because the frequency of autoantigen-specific responder CD4 T cells was measured by this assay (19). Remarkably, the ability to respond to the nucleosomal peptides was still present in the patients in long-term remission. In fact, responses as a whole were considerably attenuated in the PBMCs of some of the patients with active lupus (A-KJ and A-WB), possibly owing to prior activation in vivo and exhaustion or desensitization. Interestingly, IL-10 responses to certain nucleosomal peptides were in many cases higher than corresponding anti-CD3 responses, even in the patients in remission; and in some patients, responses to certain peptides were stronger than to the whole nucleosome particle (Figure 8).
Flow cytometry results of intracellular cytokine staining of viable, CD4+-gated T cells from peripheral blood of all the patients with lupus. T-cell response to a peptide epitope was considered positive when the percentage of positive cells was twice the background (medium alone) and at least 0.2% of the total viable CD4+ T cells were stained positive. Positive response is shown in shaded and bold numbers. nd = not done.
Flow cytometry results of intracellular cytokine staining of CD4 T cells from all the normal subjects. A positive response is displayed, as in Figure 8. The normal T cells did not show a positive response to the longer (24-mer) peptides (data not shown).
In Figure 10, a summary of the percentage of positive responders to a particular nucleosomal peptide among the patients with lupus is shown. Although some cytokine responses showed spreading to many epitopes in a few cases, such as R-DS, R-JE, and A-MC (Figure 8), certain nucleosomal peptides turned out to be recurrent epitopes for the autoimmune T cells of patients with lupus. Among the patients whose PBMCs were tested and whose T cells were not desensitized, as determined by anti-CD3 stimulation, the incidence of positive responders ranged from 50% to 100% to the following autoantigens in the panel: nucleosomes and their longer 24-mer peptides H2B10–33, H416–39, and H471–94, as well as the 15-mer nucleosomal peptides H2A 34–48, H391–105 and overlapping H3100–114, H414–28, which is overlapped by H416–39, and H449–63 (Figure 10).
Frequency of positive responders to the histone autoepitopes among the patients with lupus shown in Figure 8. Bold and shaded numbers highlight when 50% or more responded to a particular epitope; bold and italics represent at least 40% responders. These epitopes were also recognized by T-cell lines and clones from other lupus patients (Figures 1–6). AAssays could not be done in all cases owing to lack of sufficient PBMCs. The responses of certain lupus T cells were poor, judging from their anti-CD3 response, and those are also excluded from the total.
The type of cytokines that were predominantly produced was influenced in some instances by the autoantigen and in others by the patient’s T cells. For example, the ratio of IFN-γ/IL-10–positive cells responding to H391–105 was 70 in patient R-SS, but the opposite was the case in patient R-HH, in whom the IL-10/IFN-γ ratio of response to the same peptide was 10 (derived from Figure 8, after subtraction of respective background values). Similarly, H449–63 induced a stronger IFN-γ response in R-SS, but a stronger IL-10 response in A-MC. Overall, 4 of the patients (R-DS, R-DW, R-SS, and R-SC) had a higher percentage of IFN-γ–producing cells in response to the peptides; 5 others (R-JE, R-HH, A-MC, A-VS, and A-WB) were skewed more toward IL-10– and/or IL-4–producing cells; 1 patient (R-WG) appeared to favor both IFN-γ– and IL-10–positive cells (depending on the peptide used for stimulation); and 2 others (R-FN and A-KJ) were poor responders in general. Despite these variations, anti-CD3 stimulation overwhelmingly favored IFN-γ–producing CD4 T-cell response in all cases (12/12). After subtracting background values (“medium”; Figure 8), the ratios of IFN-γ/IL-10–positive CD4 T cells ranged from 2.7 to 171 (mean ± SEM = 58.6 ± 15.9), IFN-γ/IL-4 from 2.1 to 114.5 (20.3 ± 10.8), and IL-2/IL-4 from 0.6 to 24.8 (9.7 ± 2.8) upon anti-CD3 stimulation of PBMCs from the 12 patients. Using similar analysis, an IFN-γ–positive T-cell response was favored in the cases of the following autoantigens: nucleosomes, H416–39, H449–63, H471–94, and H391–105. The H391–105 peptide also stimulated IL-10– and/or IL-4–positive T-cell response among 7 of the 8 responders (Figure 10), but the latter type of cytokine-positive cells was dominant in only 2 patients. In the case of H2B10–33 and H2A34–48, the majority of the responders were in favor of a relatively stronger IL-10– and/or IL-4–positive T-cell response.