Abstract
Mugwort (Artemisia vulgaris) pollen allergens represent the main cause of pollinosis in late summer in Europe. At least 95% of sera from mugwort pollen-allergic patients contain IgE against a highly glycosylated 24- to 28-kDa glycoprotein. Recently, this major allergen, termed Art v 1, was characterized, cloned in Escherichia coli, and produced in recombinant form. In the present study we characterized and compared the T cell responses to natural (nArt v 1) and recombinant Art v 1 (rArt v 1). In vitro T cell responses to nArt v 1 and rArt v 1 were studied in PBMC, T cell lines (TCL), and T cell clones (TCC) established from PBMC of mugwort-allergic patients. Stimulation of PBMC or allergen-specific TCL with either nArt v 1 or rArt v 1 resulted in comparable proliferative T cell responses. Eighty-five percent of the TCC reactive with rArt v 1 cross-reacted with the natural protein. The majority of the CD4+CD8−TCR αβ+ Art v 1-specific TCC, obtained from 10 different donors, belonged to the Th2 phenotype. Epitope mapping of TCL and TCC using overlapping peptides revealed a single immunodominant T cell epitope recognized by 81% of the patients. Inhibition experiments demonstrated that the presentation of this peptide is restricted by HLA-DR molecules. In conclusion, the T cell response to Art v 1 is characterized by one strong immunodominant epitope and evidently differs from the T cell responses to other common pollen allergens known to contain multiple T cell epitopes. Therefore, mugwort allergy may be an ideal candidate for a peptide-based immunotherapy approach.
The family of Compositae or Asteraceae is known to include several allergenic plants, among which Ambrosia (ragweed), Parthenium (feverfew), and Artemisia (mugwort) are of major importance. Pollen of mugwort (Artemisia vulgaris) is one of the main causes of allergic reactions in late summer and autumn in Europe. The weed is predominantly found in the temperate and humid climate zones of the Northern Hemisphere and the Mediterranean basin (1). In Central Europe mugwort pollination occurs at the end of July and through August; in Mediterranean areas it occurs in September and the beginning of October (2). Among patients suffering from pollinosis, the incidence of allergic disease caused by mugwort pollen is 10–14% (2, 3). Several allergens in mugwort pollen extracts have been characterized physicochemically and immunologically in the past. At least nine different allergens have been described to date, including two major allergens (4, 5, 6). The pan-allergen profilin is also present in mugwort pollen (7).
Only recently the first complete molecular structure of a mugwort pollen allergen has been reported (27). This highly glycosylated protein with an apparent molecular mass of 24–28 kDa reacts with IgE from >95% of patients allergic to mugwort. Being the major allergen in mugwort pollen, it was termed Art v 1.3 It does not belong to the proteins that cross-react in the so-called birch-mugwort-celery syndrome (8), but it cross-reacts with a homologous protein in ragweed (7) that is not yet defined on a molecular basis. The cDNA sequence of Art v 1 was determined, and the correlating protein was expressed as recombinant nonfusion protein in Escherichia coli and purified to homogeneity. The biochemical and immunological properties of the purified natural Art v 1 (nArt v 1) and the recombinant Art v 1 (rArt v 1) have been compared. Although the theoretical molecular mass of rArt v 1 is 10.8 kDa, in SDS-PAGE the migration of the molecule indicates an apparent molecular mass of 19 kDa, obviously due to a very unusual conformation and mobility. The carbohydrate structures that comprise 30–40% of the molecule seem to play an important role in IgE binding. Two groups of patients exist: one shows binding of serum IgE to both nArt v 1 and rArt v 1, and the other reacts with nArt v 1, but displays low or no binding to rArt v 1 (27).
The involvement of CD4+ T lymphocytes in the pathophysiology of atopic disease is well established (9, 10). In addition, T cells appear to contribute to mechanisms operative in specific immunotherapy. During the administration of increasing doses of allergen, a shift from a typical allergic Th2 response to a Th1 response (immunodeviation) and a suppression of allergen-specific T cell responses (tolerance induction) have been observed (11, 12, 13, 14). In the last decade, based on cDNA cloning, deduced amino acid sequences and recombinant expression of important allergens, new concepts of specific immunotherapy evolved (15). Hypoallergenic isoforms or mutants of allergenic proteins with reduced IgE binding may allow a safer and more efficacious specific immunotherapy in the future (16). To retain relevant T cell effector functions induced by these prospective molecules, it is necessary to determine T cell epitopes and HLA restriction of the respective allergens (17).
The objective of the present study was characterization and comparison of the T cell responses to natural and recombinant Art v 1. Besides PBMC, T cell lines (TCL) and T cell clones (TCC) were investigated for phenotype and function. Furthermore, T cell epitopes and HLA restriction were determined. Interestingly, differing from other known pollen allergens, only a restricted number of T cell epitopes dominated by a single major epitope was found. The characterization of specific epitopes and HLA restrictions will provide essential information for the development of directed immunotherapy in type I allergy against mugwort and cross-reacting pollen allergens.
Materials and Methods
Patients
Peripheral blood from 18 patients was collected for this study. Type I allergy to mugwort pollen was proven by typical case history, positive (>3) RAST/CAP test (Pharmacia Diagnostics, Uppsala, Sweden), and positive skin prick test to mugwort pollen extract. All patients displayed serum IgE against nArt v 1 and rArt v 1 as tested by immunoblotting (Fig. 1⇓).
IgE immunoblots. IgE reactivity of sera from mugwort allergic patients (lanes 1–18) was tested with blotted mugwort pollen extract (nArt v 1 indicated by arrow; A) and rArt v 1 (B). Negative controls are displayed in lane 19 (buffer alone) and lane 20 (normal human serum pool).
Allergens
Pollen from A. vulgaris was obtained from Allergon (Engelholm, Sweden). Complete mugwort pollen was extracted in PBS (10%, w/v) overnight at 4°C and centrifuged at 40,000 × g. The supernatant was filtered, lyophilized, and stored at −20°C. The nArt v 1 was purified from mugwort extract by cation exchange and size exclusion chromatography and characterized as previously described in detail (27). The recombinant protein (rArt v 1) was obtained from Biomay (Vienna, Austria).
Immunoblots
Immunoblotting of patient sera was performed as previously described (7). Briefly, rArt v 1 (2 μg/lane) mugwort pollen extract (33 μg/lane) was separated by 12% SDS-PAGE and blotted onto a nitrocellulose membrane. The membrane was incubated overnight at 4°C with a 1/4 dilution of patient sera, or, as a control, with a serum pool from 13 nonallergic individuals, or with buffer alone. After incubation with 125I-labeled anti-human IgE Ab (Pharmacia Diagnostics), bound IgE was visualized by autoradiography.
Proliferation assays
PBMC (2 × 105) were cultured in triplicate in 96-well plates (Nunclone; Nunc, Copenhagen, Denmark) in 200 μl of serum-free Ultra Culture Medium (BioWhittaker, Walkersville, MD) supplemented with 2 mM/l of glutamine and 2 × 10−5 M 2-ME in the presence of nArt v 1 or rArt v 1 for 6 days at 37°C in 5% CO2 with a humidified atmosphere. Art v 1 was titrated in concentration ranges from 1.5 to 25 μg/ml. During the last 16 h of culture [3H]thymidine (0.5 μCi/well) was added, and the incorporated radioactivity was measured by scintillation counting. Proliferation tests of TCL or TCC were performed accordingly using 2–5 × 104 T cells and 1 × 105 irradiated autologous APCs per well and an incubation time of 2 days.
Allergen-specific TCL and TCC
Allergen-specific, short term TCL and TCC were obtained as previously described (18). Briefly, 1.5 × 106 PBMC were stimulated with 20 μg/ml of purified nArt v 1 (TCLn) or rArt v 1 (TCLr) in 24-well, flat-bottom culture plates (Costar, Cambridge, MA). After 5 days suboptimal doses of human rIL-2 (10 U/ml; Roche, Mannheim, Germany) were added, and cultures were continued for an additional 7 days. Thereafter, monoclonal T cell cultures were established by limiting dilution, and the remaining T cell blasts were used for epitope-mapping experiments. Cells (0.3 cells/well) from Art v 1-specific TCL were seeded into 96-well, round-bottom plates (Nunclone) in the presence of 2 × 105 irradiated (60 Gy) allogeneic PBMC, 0.25% (v/v) PHA (Life Technologies, Grand Island, NY), and rIL-2 (4 U/well) in the medium mentioned above. After 14–21 days, growing microcultures were expanded at weekly intervals with fresh allogeneic irradiated feeder cells and rIL-2. The specificity of TCC was assessed in proliferation assays as soon as the cell number reached 2 × 105. When the stimulation index (SI; ratio of cpm obtained in cultures containing TCC, autologous APC, and Ag to cpm obtained in cultures containing TCC and APC alone) was >10, responses were considered positive. Art v 1-specific TCC were expanded by alternating turns of stimulation with autologous irradiated APC and Art v 1 or with allogeneic feeder cells and rIL-2.
Analysis of the phenotype of TCC
The phenotype of TCC was analyzed by flow cytometry, using a FACScan (BD Biosciences, Mountain View, CA) and the FITC-labeled mAbs anti-Leu 4/CD3, anti-Leu 3a/CD4, anti-Leu 2a/CD8, anti-TCR αβ WT 31, and anti-TCR γδ (BD Biosciences) as previously described (19).
Measurement of cytokines
TCC were washed and incubated with irradiated autologous APC in the presence of Art v 1 (5 μg/ml) for 24 h. Cytokine levels in the resulting supernatants were measured in ELISA using matched Ab pairs (Endogen, Woburn, MA) according to instructions by the manufacturer (sensitivity limits: IL-4, 9 pg/ml; IFN-γ, 9 pg/ml). Cultures containing TCC and APC alone served as negative controls. TCC with a ratio of IFN-γ/IL-4 of >10 were classified as Th1, those with a ratio of 0.1–10 were classified as Th0, and those with a ratio <0.1 were classified as Th2.
Epitope mapping
A panel of 33 peptides was synthesized according to the Art v 1 amino acid sequence (4) by Mimotopes/Biotrend (Koln, Germany). Peptides were 12 residues long and overlapped for three amino acids, i.e., neighboring peptides shared nine amino acids. T cells (5 × 104) of TCL or TCC were tested with each of the 33 peptides (5 μg/ml) in the presence of 1 × 105 autologous irradiated APC. An SI > 3 was considered positive for TCL; an SI >10 was considered positive for TCC.
HLA restriction and typing
Blocking Abs directed against HLA-DP (B7/21, IgG1), HLA-DQ (SK10, IgG1), and HLA-DR (L243, IgG2a; BD Biosciences) were used in a final concentration of 10 μg/ml in the presence of 2.5 μg/ml Art v 125–36, 2.5 × 104 autologous APC, and Art v 125–36-specific TCC to perform proliferation assays. Molecular HLA typings for HLA-DRB and DQB alleles were performed according to methods previously described (20).
TCRβ gene usage
Total RNA was isolated from ∼1 × 106 T cells using the RNeasy kit (Qiagen, Hilden, Germany) and was subjected to first-strand cDNA synthesis using an oligo(dT)16 primer (PerkinElmer, Norwalk, CT). A panel of 25 specific primers complementary to TCRBV and a respective primer for the constant region were used to determine the expression of TCRβ gene families by PCR as recommended by the provider (Clontech, Heidelberg, Germany). The PCR products were visualized in 2% agarose gels with ethidium bromide under UV light.
Three-dimensional structural modeling
Sequence similarity search was performed using PSI-Blast version 2.1 (National Center for Biotechnology Information, Bethesda, MD) (21) vs the NCBI-NR non-redundant database and the second run vs domain-based sequence database of proteins with known structures. The ProFIT program (Proceryon BioSciences, Salzburg, Austria) was used for fold recognition calculation (22). The three-dimensional structural models were generated by a threading procedure, which aligns the query sequence with the template structure according to pair and surface mean potentials. The structural models were evaluated according to their z-scores. Data analysis and interpretation were performed using ProHIT package facilities.
Statistics
Statistical significance was assessed by Mann-Whitney U test, and Pearson’s correlation coefficients were determined using SPSS software (SPSS, Chicago, IL).
Results
IgE immunoblots
Eighteen patients allergic to mugwort pollen with IgE against both nArt v 1 and rArt v 1 were selected for our study. The immunoblots in Fig. 1⇑ demonstrate that serum IgE from all patients reacted with the natural protein in the 24–28 kDa range present in mugwort extract. Due to the heterogeneous glycosylation of nArt v 1, a typical broad double band was observed. The recombinant allergen was recognized as a single band at 19 kDa. No signal was observed with buffer or normal human serum, which were used as negative controls.
Proliferation assays: PBMC and TCL
The proliferative responses observed in the patients’ PBMC (n = 18) after stimulation with nArt v 1 (SI (mean ± SD), 2.99 ± 2.30) or with rArt v 1 (2.89 ± 2.66) revealed a correlation coefficient of r = 0.96 (p = 0.01), indicating comparable T cell responses for both forms of allergen. PBMC of healthy, nonallergic donors showed a significantly (p = 0.013 and p = 0.019, respectively) lower response for both allergens (Fig. 2⇓). TCL obtained after initial stimulation with either nArt v 1 (TCLn) or rArt v 1 (TCLr) also did not result in conspicuously different T cell responses to secondary stimulation with either Ag (n = 16; TCLn, 4.07 ± 2.99 with nArt v 1 and 4.44 ± 2.99 with rArt v 1; TCLr, 2.75 ± 1.54 and 3.81 ± 3.32), with r = 0.76 (p = 0.01) and r = 0.82 (p = 0.01).
Proliferation of PBMC induced by nArt v 1 and rArt v 1. The results are shown in box plots. Each box represents the interquartile range containing 50% of the values. The line across the box indicates the median, and the whiskers extent the highest and lowest values. The baseline counts varied from 660 to 67,797 cpm for allergic donors and from 2,449 to 36,141 for normal donors.
Art v 1-specific TCC
A total of 54 TCC were obtained from 10 different patients, 25 from TCLn (Table I⇓) and 29 from TCLr (Table II⇓). All TCC expressed the CD4+CD8− TCRαβ+ Th phenotype. The majority of TCC (TCLn-derived, 72%; TCLr-derived, 66%) produced Th2 cytokines after specific stimulation. A high degree of cross-reactivity between rArt v 1 and nArt v 1 was observed when the TCC were tested for proliferation; 85% of TCC reactive with rArt v 1 (n = 41) also recognized nArt v 1. Similarly, 73% of nArt v 1-reactive TCC (n = 48) also proliferated in response to the recombinant protein.
Characterization of Art v 1-specific TCC derived from initial stimulation with nArt v 1
Characterization of Art v 1-specific TCC derived from initial stimulation with rArt v 1
T cell epitopes and TCR Vβ usage
T cell epitopes were determined in TCLn and TCLr established from 15 patients and in Art v 1-specific TCC (Tables I⇑ and II⇑) by proliferation assays using 33 overlapping 12-mer peptides spanning the complete amino acid sequence of rArt v 1. Peptide specificity was defined by an SI >3 for TCL and >10 for TCC. An overview of all epitopes obtained is depicted in Fig. 3⇓. In general, only one or two epitopes were detectable in each patient. Fourteen of 17 patients recognized an Art v 1 epitope in the range of aa 22–36. Five patients recognized an epitope contained in Art v 143–54. The dominance of epitope Art v 122–36 was also reflected at the clonal level, where in seven of 10 individuals a response(s) to an epitope(s) in this region was observed. Epitope mapping of 41 TCC reactive with rArt v 1 obtained from either TCLn (Table I⇑) or TCLr (Table II⇑) resulted in a predominant reactivity to Art v 125–36 (32 of 41 TCC; i.e., 78%). Six TCC from patient KRE reacted with the neighboring peptide Art v 122–33, and one TCC (KRE N 129) reacted with two consecutive peptides (Art v 122–33 and Art v 125–36), indicating the possibility of two closely neighboring epitopes. In Fig. 4⇓ the immunodominant region is highlighted within a computer model of the Art v 1 molecule. Further T cell-stimulating peptides could be identified in the sequence area Art v 11–18 and Art v 140–55. The nArt v 1-specific TCC that did not cross-react with rArt v 1 could not be stimulated with any of the peptides synthesized according to rArt v 1.
T cell epitope mapping. TCL from 15 patients and TCC from 10 patients were established with nArt v 1 or rArt v 1, and proliferation was tested with overlapping 12-mer peptides spanning the Art v 1 sequence. Proliferations with SI >3 (TCL) or >10 (TCC) are indicated (▪).
Computer-derived model of Art v 1. A, Model of the total Art v 1 molecule; B, N-terminal domain in two views representing a rotation of 180° around the vertical (z) axis as indicated at the top. The T cell epitope Art v 122–36, indicated in dark blue, including aa side chains, encompasses part of the α-helix and the loop connecting the α-helix to the β-sheet. C, Electrostatic surface potential plot of the N-terminal domain of Art v 1 (calculated using the program MOLMOL) in the same orientations as the structures in B. Electrostatic potential is indicated in red (negative charge), white (neutral), and blue (positive charge). Art v 122–36 encompasses a positive patch of the Art v 1 surface.
The TCR Vβ families of TCC from several patients were determined (Tables I⇑ and II⇑). In three individuals (MAP, KRE, and PHS) multiple TCR Vβ families were detected. This diversity indicated monoclonality of our TCC and, in addition, that TCC specific for Art v 125–36 were not independent isolates of a single T cell clone that dominated the original TCL.
HLA restriction and HLA-DR/DQ typing
Inhibition experiments using anti-HLA-DR, -DP, and -DQ Abs indicated that the peptide Art v 125–36 was presented by autologous HLA-DR molecules (Fig. 5⇓). HLA-DR/DQ typing was performed for 12 of the study patients (Table III⇓). Interestingly, the results revealed a disproportionately high expression of HLA-DRB1*01 (67%) and HLADRB1*16 (33%) in these patients compared with expected frequencies (25 and 5%) in an adequate population (23).
HLA-DR restriction of the T cell response to the immunodominant epitope of Art v 1. Anti-HLA-DR, but not -DP or –DQ, Abs inhibit the proliferation of TCC induced by Art v 125–36 presented by irradiated autologous APCs (three representative TCC from different patients are shown).
HLA-DR, -DQ typing of mugwort pollen allergic patients with Art v 1-specific IgE
Discussion
In the present study we investigated the T cell response of mugwort allergic patients to Art v 1, the major allergen of mugwort pollen. The nArt v 1 and its recombinant form elicited similar T cell responses in PBMC and TCL. All Art v 1-specific TCC established from these TCL expressed the CD4+CD8− TCRαβ+ phenotype, and the majority exhibited a Th2 cytokine profile. The most prominent finding was that a single immunodominant epitope emerged from peptide mapping experiments; >80% of the patients studied possessed T cells specific for Art v 122–36. Inhibition experiments demonstrated that this peptide was presented by HLA-DR molecules. Interestingly, HLA typing of patients with T cell reactivity to Art v 122–36 revealed a remarkably high frequency of HLA DRB1*01 and HLADRB1*16 haplotypes.
In late summer mugwort pollen is a major cause of allergy in Europe. Recently, the first molecular structure of a major mugwort allergen has been reported (27): Art v 1 is recognized by IgE in >95% of patients allergic to mugwort. Art v 1 has an unusual tertiary head and tail structure (Fig. 4⇑) that results in unorthodox biochemical behavior. The deduced amino acid sequence of the mature protein revealed a modular structure with a signal peptide followed by an N-terminal, cysteine-rich, defensin-like domain and a C-terminal domain rich in prolines. Seventy-eight percent of the prolines appear to be hydroxylated and provide an anchor for one or two O-linked, highly branched carbohydrate structures containing galactose and arabinose. The binding of IgE to nArt v 1 is partly due to the extensive carbohydrate structures that comprise 30–40% of the molecule. However, with respect to T cell responses, similar proliferations to nArt v 1 and rArt v 1 were observed in PBMC and TCL obtained after initial stimulation with either nArt v 1 (TCLn) or rArt v 1 (TCLr). In addition, congruent epitopes for TCLn or TCLr and comparable cloning frequencies of Art v 1-specific TCC derived from these TCL were observed (not shown). Therefore, the extensive presence of carbohydrates on the natural protein apparently is not essential for the T cell response to Art v 1. The TCC do not show 100% cross-reactivity between recombinant and natural protein. A possible explanation may be the presence of different isoforms in the natural allergen source. In fact, we have isolated 30 DNA clones coding for 10 different isoforms of Art v 1 (F. Ferreira, unpublished observations).
Phenotypic and functional analyses of individual Art v 1-specific TCC revealed a typical Th2-biased T cell response similar to many other allergens characterized in the past (9). In striking contrast to other pollen allergens, Art v 1 displayed a remarkably restricted epitope specificity. The Art v 1 epitope in the range of aa 22–36 was recognized by TCL or TCC from 14 of 17 patients (82%), and 10 of 17 (59%) reacted with this epitope exclusively. Our finding is in contrast with published data showing that atopic allergens, in general, harbor multiple T cell epitopes scattered over the complete amino acid sequence and that these patterns differ also interindividually (17). In this study, several dispersed epitopes for Art v 1 were detected in only two of 17 patients (Fig. 3⇑, AW and KOG). In theory, the apparent restricted response to Art v 122–36 might be caused by the outgrowth of repetitive isolates from one dominating TCC in the original TCL. However, the diversity of TCR Vβ families expressed in TCC indicated that in each patient the majority of TCC is derived from a distinct T cell ancestor. In total, 18 TCC specific for Art v 122–36 were found to express 11 different TCR Vβ families, demonstrating that the restricted epitope recognition in Art v 1 is also not due to a restricted TCR repertoire.
The presentation of Art v 122–36 to T cells is apparently HLA-DR mediated (Fig. 5⇑). Due to the observed epitope diversity, associations of HLA haplotypes and allergy have not been confirmed, with rare exceptions (17, 24). Perhaps because of the present restricted epitope situation, special HLA specificities could be observed in our study. However, algorithms to predict T cell epitopes such as syfpeiti (25) or tepitope (26) did not classify the sequence of Art v 122–36 as a highly probable HLA-binding peptide for any of the most common HLA-DR haplotypes. We found a skewed expression of HLA-DRB1*01 (73%) and HLA-DRB1*16 (36%) in patients with T cell reactivity to Art v 122–36 compared with the normal distribution of these HLA-DR types in the Austrian population (25 and 5%, respectively) (23). This finding may indicate that HLA molecules encoded in DRB1*01 or DRB1*16 act as preferred restriction elements for the Art v 122–36 peptide in our patient sample. The results from inhibition experiments with anti-HLA class II Abs supports the hypothesis that HLA-DR molecules act as restriction elements.
The characterization of T cell epitopes has implications for future developments in specific immunotherapy of type I allergy. Concepts have been developed that suggest the use of hypoallergenic recombinant molecules, e.g., low/no IgE-binding isoforms, mutants (15), or even peptides (14), instead of allergen extracts. These molecules should, however, retain relevant T cell epitopes to enable T cell-dependent immune regulatory mechanisms. For most allergens containing multiple individually diverse epitopes, peptide immunotherapy is not promising. Considering the distinguished finding of an immunodominant T cell epitope in the major mugwort pollen allergen Art v 1, mugwort allergy may be a suitable candidate for peptide treatment. Whether there exists a possible association between mugwort pollen allergy and certain HLA restriction molecules has to be addressed in larger epidemiological studies.
Footnotes
-
↵1 This work was supported by grants from the Fonds zur Förderung der Wissenschaftlichen Forschung, Austria (S8808-MED, S8802-MED), and Biomay, Austria.
-
↵2 Address correspondence and reprint requests to Dr. Christof Ebner, Department of Pathophysiology, University of Vienna, AKH-3Q, Waehringer Guertel 18-20, A-1090 Wien, Austria. E-mail address: christof.ebner{at}akh-wien.ac.at
-
↵3 Abbreviations used in this paper: Art v 1, major mugwort pollen allergen; n/r, natural/recombinant; TCC, T cell clone; TCL, T cell line; TCLn(r), TCL obtained after initial stimulation with n(r) Artv 1; SI, stimulation index.
- Received June 11, 2002.
- Accepted September 18, 2002.
- Copyright © 2002 by The American Association of Immunologists
References
In this issue
