| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online July 18, 2003 as doi:10.1096/fj.02-0872fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Pathophysiology, University of Vienna, Vienna, Austria; and
Institute of Chemistry, University of Agriculture, Vienna, Austria
2 Correspondence: Department of Pathophysiology, University of Vienna, AKH-EBO 3Q, Währinger Gürtel 18-20, A-1090 Vienna, Austria. E-mail: Karin.Hoffmann{at}akh-wien.ac.at
SPECIFIC AIMS
High molecular weight (HMW) glycoproteins are major immunoglobulin E (IgE) binding components in pollen and plant food, but their capability of eliciting clinical symptoms of allergy has been questioned. Thus, we aimed to purify and characterize a glycosylated HMW allergen from celery, the major elicitor of food allergy in Central Europe.
PRINCIPAL FINDINGS
1. Api g 5 is a glycoprotein belonging to a family of flavin adenine dinucleotide (FAD)-containing oxidases
A fraction containing two IgE binding proteins with molecular weights of 53 and 57 kDa was purified from celery extract by ammonium sulfate precipitation, lectin affinity chromatography using concanavalin A sepharose, and anion exchange chromatography. Both proteins were subjected to N-terminal amino acid sequence analysis, which yielded identical N-termini corresponding to the N-terminal sequence of Api g 5, a previously described HMW celery allergen also isolated as a mixture of two proteins. Cleavage of the protein mixture by cyanogen bromide yielded several fragments with sequences highly similar to a family of FAD-containing oxygen-dependent oxidoreductases encompassing berberine bridge-forming enzymes and carbohydrate oxidases. The UV absorbance spectrum of purified Api g 5 showed maxima at 350 and 452 nm, indicative of flavin.
To confirm the presence of IgE binding N-glycans on Api g 5, the blotted protein was stained with an antihorseradish peroxidase (anti-HRP) antibody specific for N-linked glycan moieties containing ß1,2-xylosyl and 
1,3-fucosyl residues. Both Api g 5 bands gave very strong, positive reactions with these antibodies. The glycan structures were confirmed by mass spectrometry. The proteins were digested by trypsin, and the resulting peptide mixtures were treated with peptide-N-glycosidase (PNGase) A to release carbohydrate moieties. The mass spectra of the released glycans and the mass differences observed between the untreated and the PNGase A-digested tryptic maps revealed the presence of MUXF and MMXF, two types of fucose- and xylose-containing plant-specific N-glycans, repectively.
2. Deglycosylation of Api g 5 disrupts its IgE binding epitopes
The IgE binding capacity of Api g 5 was assayed in an immunoblot with 14 allergic patients’ sera containing IgE specific for HMW allergens. Both Api g 5 bands bound similar amounts of IgE. To determine the contribution of glycans to its IgE binding capacity, Api g 5 was deglycosylated by trifluoromethane sulfonic acid (TFMS). The electrophoretic mobility of Api g 5 changed significantly after removal of glycosyl groups, but both bands were still present. Immunostaining with a glycan-specific anti-HRP antibody confirmed that the deglycosylation procedure had been practically quantitative. The deglycosylated form of Api g 5 lacked detectable binding of IgE from allergic patients’ sera. These findings were corroborated by the results of a native IgE enzyme-linked immunosorbent assay (ELISA) with 12 individual allergic patients’ sera, in which only background levels of IgE bound to the deglycosylated protein were detected. To rule out that the lack of IgE binding to deglycosylated Api g 5 was caused by conformational differences between the native and the deglycosylated protein, Api g 5, irreversibly denatured by reduction and S-alkylation of disulfide bonds, was included in the proteins tested in this ELISA. No significant differences between the amounts of IgE bound to native and reduced Api g 5 were detected (Fig. 1
).
|
3. Native Api g 5 is capable of stimulating basophils derived from allergic patients
The allergenicity of native and deglycosylated Api g 5 was further examined by comparing their abilities to release histamine from basophils of a patient allergic to celery, mugwort pollen, and grass pollen. The blood cells from this patient gave rise to specific, dose-dependent histamine release after stimulation with native Api g 5 but not with deglycosylated Api g 5 (Fig. 1A
). In contrast, native or deglycosylated Api g 5 did not activate basophils from a patient sensitized to HMW glycoproteins but without clinical celery allergy (Fig. 1B
). The affinity of carbohydrate-specific IgE was similar compared with IgE directed to Api g 1, a nonglycosylated allergen with known clinical relevance (Fig. 1C
).
4. Cross-reactivity of Api g 5 with pollen allergens depends on its glycan moieties
We used a pool of sera from patients with pollen and/or celery allergy to identify allergens in celery, mugwort, and birch pollen extracts that share IgE epitopes with Api g 5. The serum pool was preincubated with Api g 5, MUXF-BSA (a fucose and xylose-containing glycan coupled to bovine serum albumin), and MM-BSA (a control glycan without fucose and xylose). Subsequently, IgE binding to blotted protein extracts was tested. MM-BSA did not significantly inhibit IgE binding to these extracts. Preincubation of the serum pool with Api g 5 almost completely abolished IgE reactivity to the 45- to 70-kDa allergens in celery extract and to all IgE-reactive bands in mugwort pollen but only partially inhibited binding to birch pollen extract. Inhibition by MUXF-BSA gave rise to effects equivalent to those achieved with native Api g 5. Cross-reactivity of native and deglycosylated Api g 5 was tested in an inhibition with three individual sera. Compatible with the results using the serum pool described above, complete inhibition of IgE binding to HMW allergens in celery and mugwort pollen extracts was achieved with native Api g 5, and deglycosylated Api g 5 brought about no effect.
CONCLUSIONS AND SIGNIFICANCE
HMW proteins from celery have previously been shown to be major IgE binding components when tested with sera from patients with confirmed celery allergy. IgE specific for these allergens cross-reacts with proteins from birch and mugwort pollen, which in part, explains the clinical correlation between birch, mugwort, and celery allergy. During recent years, it became clear that IgE directed to HMW allergens binds to glycoproteins carrying N-glycans containing 1,3-fucosyl and ß1,2-xylosyl residues, which are found in plants and invertebrates but not in mammals. The importance of celery as the prevailing food allergen source in Central Europe, the cross-reactivity of HMW allergens, and the lack of conclusive data on the clinical significance of IgE specific for CCDs prompted us to purify and characterize a glycoprotein allergen from celery.
The purified glycoprotein turned out to be identical to the allergen described previously, Api g 5. Glycan structures of the protein were determined by an immunoblot assay using an antibody reactive to plant-specific N-glycans and by enzymatic removal of N-glycans combined with mass spectrometric analysis of the cleaved sugars. The glycans bound to Api g 5, MUXF and MMXF, two types of fucosylated and xylosylated complex N-glycans, are two examples of a range of structures previously found in total celery extract.
N-terminal and internal sequencing of Api g 5 revealed that the protein belongs to a family of flavoproteins with diverse functions comprising carbohydrate oxidases, berberine bridge-forming enzymes, and a drought-inducible protein from cowpea. The biochemical functions of all these proteins are related to the plant’s response to pathogen attack and abiotic stress. A stress-related function of Api g 5 is compatible with the observation that many plant allergens are homologous to pathogenesis-related proteins.
Western blotting and ELISA confirmed the IgE binding capacity of purified Api g 5. The chemical nature of the allergenic epitopes on Api g 5 was examined by destroying the covalently linked carbohydrates by acid cleavage with TFMS, which led to a quantitative elimination of the affinity for IgE as determined by immunoblot and ELISA. Possible effects of the TFMS treatment on IgE binding capacity via disruption of the protein structure were excluded by additional experiments: Mass spectrometric analysis of tryptic digests of native and deglycosylated Api g 5 showed no differences other than those caused by cleavage of the sugar moieties, indicating that TFMS had no effects on the primary structure of Api g 5. Possible disruption of the tertiary structure cannot be excluded but was shown to have no impact on IgE binding capacity, as irreversible denaturation of Api g 5 had no significant influence on IgE binding in a native ELISA. In conclusion, we present strong evidence that IgE binding to Api g 5 is mediated exclusively by its N-glycans. A similar conclusion can be drawn for the cross-reactivity of Api g 5 with other HMW allergens from celery and with proteins from birch and mugwort pollen. IgE immunoblot inhibition experiments showed that native Api g 5 and the fucosylated and xylosylated N-glycan MUXF coupled to BSA completely inhibited IgE binding to HMW proteins from celery, birch, and mugwort in a similar manner. No cross-reactivity was observed with a neoglycoprotein carrying N-glycans lacking fucose and xylose as well as with deglycosylated Api g 5.
IgE binding capacity and cross-reactivity of plant-specific N-glycans have previously been shown using various model compounds as well as pollen and food allergens. In contrast, only a few publications dealt with the clinical significance of carbohydrate-specific IgE. Initially, most researchers shared the opinion that glycan-specific IgE merely leads to a false-positive allergy diagnosis without being capable of eliciting an allergic reaction. This view was supported by a study of grass pollen-allergic patients with a positive peanut radioallergosorbent test but lacking clinical symptoms of peanut allergy whose peanut-specific IgE exclusively bound to CCDs. Conversely, carbohydrate-specific IgE was shown to be capable of activating basophils derived from allergic patients via binding to bromelain glycopeptides coupled with BSA, free N-glycans cleaved off from the major olive pollen allergen Ole e 1, and the cypress glycoallergen Cup a 1. Compatible with these data, our experiments showed that Api g 5 activates basophils derived from a patient allergic to celery and mugwort pollen via binding to glycan-specific IgE (Fig. 2
).
|
In conclusion, we present the biochemical and immunological characterization of the first HMW glycoprotein allergen that contributes to the birch-mugwort-celery-spice syndrome. The celery allergen, Api g 5, is capable of binding human IgE and thus activating basophils derived from an allergic patient exclusively via its cross-reactive, plant-specific N-glycans. The possible clinical relevance of N-glycan-specific IgE may have consequences for the practice of allergy diagnostics. However, it remains to be elucidated if there are additional factors determining the clinical outcome of a glycan-specific IgE response.
FOOTNOTES
1 To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.02-0872fje; doi: 10.1096/fj.02-0872fje 
This article has been cited by other articles:
|
|
 |
|
  G. Poltl, O. Ahrazem, K. Paschinger, M. D. Ibanez, G. Salcedo, and I. B.H. Wilson Molecular and immunological characterization of the glycosylated orange allergen Cit s 1 Glycobiology, February 1, 2007; 17(2): 220 - 230. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Jin, M. Bencurova, N. Borth, B. Ferko, E. Jensen-Jarolim, F. Altmann, and B. Hantusch Immunoglobulin G specifically binding plant N-glycans with high affinity could be generated in rabbits but not in mice Glycobiology, April 1, 2006; 16(4): 349 - 357. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Bencurova, W. Hemmer, M. Focke-Tejkl, I. B.H. Wilson, and F. Altmann Specificity of IgG and IgE antibodies against plant and insect glycoprotein glycans determined with artificial glycoforms of human transferrin Glycobiology, May 1, 2004; 14(5): 457 - 466. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), TFMS-deglycosylated Api g 5 (
), and recombinant Api g 1 (
). Histamine release is expressed as a percentage of total histamine content determined after cell lysis.
RI, High-affinity receptor for IgE.