Table 1: Allergenicities of salt-soluble and -insoluble (gluten)
fractions of wheat flour. Wheat flour proteins were divided
into salt-soluble and gluten fractions in the usual manner, and subjected
to ELISA. IgE-binding is expressed in arbitrary units based on the
absorbance at 490 nm.
|
|
|
|
|
2 3 |
0.15 >2.0 |
<0.05 0.07 |
|
5 |
0.05 |
0.23 |
|
7 8 9 |
>2.0 >2.0 >2.0 |
>2.0 >2.0 >2.0 |
1 IgE-BINDING TO AMINO ACID SEQUENCE BASED STRUCTURES OF GLUTENIN
1.1 IgE-BINDING OF CHYMOTRYPTIC PEPTIDES FROM LOW-MOLECULAR-MASS GLUTENIN
Since gluten was insoluble in aqueous media, it was hydrolyzed with alpha-chymotrypsin to obtain soluble peptide fragments with allergenicity. Food allergens are often characterized by their high stability against digestive enzymes, with their epitope structures remaining unchanged (Taylor et al. 1987). Thus digested peptides derived from the gluten fraction were expected to be still capable of IgE-binding. The resulting hydrolytic reaction product was centrifuged, and the supernatant was subjected to gel filtration and reversed-phase HPLC. The allergenicity of the fractionated elute was evaluated by ELISA, and the peak with the highest allergenicity was subjected to a primary structure determination.
The primary structure of the purified compound was a 30-mer peptide
and determined to be (Ser-Gln-Gln-Gln-(Gln-)Pro-Pro-Phe)4.
This allergenic peptide showed high similarities (almost 90%) to low-molecular-mass
glutenin precursors (Pitts et al. 1988, Colot et al. 1989). Therefore,
we concluded the peptide originated from low-molecular-mass glutenin. Similarities
of about 70% were also obtained between the sequence of the allergenic
peptide and those of the low-molecular-mass glutenin precursors from durum
wheat (Cassidy & Dvorak 1990, D'Ovidio et al. 1992.). Surprisingly,
a high degree of similarity (53.6%) was also found between the allergenic
peptide and a Saccharomyces cerevisiae protein (Rasmussen 1994).
Thus, it should be noted that wheat allergic patients are also suspected
to be sensitive to yeast used for bread making.
The repeated sequence in allergenic peptides such as (Ser-Gln-Gln-Gln-(Gln-)Pro-Pro-Phe)4
may be favourable for cross-linking IgE antibodies and triggering the release
of chemical mediators from mast cells in our body. There exists a very
famous allergen, that is cod allergen (allergen M) which contains three
homologous IgE-binding tetrapeptides in the residues 41-64 (Elsayed et
al. 1982).
1.2 ANALYSIS OF IgE-BINDING EPITOPES WITH SYNTHETIC PEPTIDES
In order to identify IgE-binding epitopes on the 30-mer amino acid sequence, peptides listed in Table 2 were synthesized according to the solid phase method. The N-terminal amino acid of each peptide was acetylated to mimic the condition under which each peptide existed in intact form. The allergenicity of each peptide was evaluated by ELISA (Tanabe et al. 1996). As shown in Table 2-A, (Ser-Gln-Gln-Gln-(Gln-)Pro-Pro-Phe)4, (Ser-Gln-Gln-Gln-(Gln-)Pro-Pro-Phe)2, and Ser-Gln-Gln-Gln-(Gln-)Pro-Pro-Phe bound to IgE almost equally. There was no difference between the relative ELISA values of Ser-(Gln)4-Pro-Pro-Phe and Ser-(Gln)3-Pro-Pro-Phe. These data suggest that the Ser-Gln-Gln-Gln-Pro-Pro-Phe motif is involved in binding to IgE antibodies.
To examine which amino acid residues in the motif are essential for binding to IgE, we replaced each constituent amino acid residue by Gly. When any of the asterisked amino acid residues in the sequence Ser-Gln*-Gln-Gln-Pro*-Pro*-Phe was replaced, the ELISA value dropped below the limit of detection (Table 2-B). These amino acid residues are therefore thought to be indispensable for IgE-binding. Tables 2-B and C also show that Gln-Gln-Gln-Pro-Pro, which lacks the N- and C-terminals of Ser-Gln-Gln-Gln-Pro-Pro-Phe, gave an ELISA value equal to that obtained with the full peptide.
We further examined which amino acid residues of Gln-Gln-Gln-Pro-Pro are essential for binding to IgE (Table 2-C) and found that the N-terminal glutamine residue and the two proline residues are essential. It was thus concluded that the IgE-binding epitope of the allergenic peptide comprised Gln-X-Y-Pro-Pro, where X and Y were replaceable amino acid residues. Indeed the inhibition ELISA assay showed that Ac-Gln-Gln-Gln-Pro-Pro bound to wheat-specific IgE in the serum of patients. To analyze the binding between Ac-Gln-Gln-Gln-Pro-Pro and IgE antibody, Fukushi et al. (1998) made the first NMR analysis of Ac-Gln-Gln-Gln-Pro-Pro. Their data showed that the configurations of the amide bonds of the peptide backbone were all-trans.
As in Table 2-C, the ELISA value obtained with Gln-Gly-Gln-Pro-Pro was lower by almost 30% than that obtained with Gln-Gln-Gln-Pro-Pro, and the value with non-acetylated Gln-Gln-Gln-Pro-Pro was almost half of that with acetylated Gln-Gln-Gln-Pro-Pro. From these data, the second glutamine residue of Gln-Gln-Gln-Pro-Pro and acetylation of the N-terminal amino group are both advantageous for binding to IgE.
Also, recombinant low-molecular-mass glutenin, which contained many
Gln-Gln-Gln-Pro-Pro motifs, were expressed in Escherichia coli by
a pET vector system and confirmed its IgE-binding ability (Maruyama et
al. 1998).
Table 2: IgE-binding abilities of synthetic peptides. Peptides
were synthesized according to the solid phase method. The peptide-bound
multipins in a solid state were subjected to ELISA using sera of wheat
allergic patients. Amino acids are denoted in the single-letter code. (Tanabe
et al. 1996)
(A) Peptide | Relative ELISA value |
Ac-SQQQQPPF SQQQPPF SQQQQPPF SQQQPPF | 1.0 |
Ac-SQQQQPPF SQQQPPF | 1.1 |
Ac-SQQQQPPF | 1.1 |
Ac-SQQQPPF | 1.0 |
(B) Peptide | Relative ELISA value |
Ac-GQQQPPF | 1.1 |
Ac-SGQQPPF | nd |
Ac-SQGQPPF | 0.8 |
Ac-SQQGPPF | 1.0 |
Ac-SQQQGPF | nd |
Ac-SQQQPGF | nd |
Ac-SQQQPPG | 0.9 |
(C) Peptide | Relative ELISA value |
Ac-QQQPP | 0.9 |
Ac-GQQPP | nd |
Ac-QGQPP | 0.7 |
Ac-QQGPP | 1.0 |
Ac-QQQGP | nd |
Ac-QQQPG | nd |
QQQPP | 0.6 |
2 IgE-BINDING TO STRUCTURES CONTAINING ASN-LINKED GLYCOCHAINS
Wheat alpha-amylase inhibitors (AIs) have been studied as allergens for over 20 years. There is a family of wheat AIs with a number of differing monomeric, dimeric and tetrameric proteins. As described above, IgE-binding epitope structures of AI 0.28 have already been determined on the amino acid sequence (Walsh & Howden 1989). Moreover, a glycan moeity of one subunit from the tetrameric AI (CM16) was also capable of IgE-binding from sera of patients with baker’s asthma (Sánchez-Monge et al. 1992). Although, other AIs, such as AI 0.19, AI 0.28, and AI 0.53, were also reported as allergens, involvement of glycans in allergic responses has not been fully proven. As for the AI, James et al. (1997) showed that AI was an allergen for both asthma and wheat allergy.
In addition, Asn-linked glycochains have received recent attention in
the studies on cross-reactivity between pollen, insects, and food allergens
(Batanero et al. 1996, Garcia-Casado et al. 1996). Garcia-Casado et al.
(1996) reported that the presence of a beta-1,2-xylosyl residue, which
was attached to the beta-linked mannose of the glycochain core in bromelain
and peroxidase, constituted an IgE-reactive determinant. Indeed, we previously
reported that patients sensitive to salt-soluble fraction of wheat flour
cross-reacted to bromelain (Tanabe et al. 1997). We thus examined IgE-binding
glycoproteins in wheat flour and clarified whether any new glycoprotein
occurred or not (Watanabe et al. 2001).
Wheat flour was extracted with 10 mM sodium dihydrogenphosphate followed by addition of ammonium sulfate to 50% saturation at pH 7.0. The precipitate was dialyzed against running water, and then dissolved with 10 mM acetate buffer (pH 4.5) containing 0.5 M NaCl. The solution was submitted to Carboxymethyl-(CM)-cellulose and DEAE-cellulose column chromatography. The IgE-binding crude fraction thus obtained was lyophilized. SDS-PAGE was carried out using a 7.5% gel, and proteins in the gel were electrotransferred onto a PVDF membrane. The same procedure was repeated three times. One of the membranes was immunoassayed using sera of wheat-sensitive allergic patients (Figure 1, lane A). Another membrane was submitted to immunodetection with a rabbit anti-HRP (horse radish peroxidase) as a primary antibody, which recognizes peroxidase type N-linked glycochains (Batanero et al. 1996) (lane B). The other membrane was stained non-specifically with coomassie blue (CBB R-250) (lane C). There was one unknown IgE-binding protein in all three lanes detected at about 60 kDa (asterisked band). The protein reacted with the anti-peroxidase antibody (lane B), indicating that it contained N-linked glycochain(s). The reactivity of the glycan moiety in the 60 kDa allergen is under investigation. Bands at about 40 kDa and 16 kDa are probably peroxidase (Sánchez-Monge et al. 1997) and AIs (James et al. 1997, Sánchez-Monge et al. 1997), respectively. |
|
The N-terminal amino acid sequence of the asterisked band was determined
to be LDPDESEXVTRYFRIR. The 8th amino acid residue would be Asn to which
a glycochain attaches. The glycoprotein reacted both with the anti-horse
radish peroxidase IgG antibody and sera from several wheat-allergic patients.
The amino acid sequence similarity between the peptide fragment and other
naturally occurring proteins was checked using a sequence database. As
a result, no similarity was obtained between the sequence of the 60 kDa
glycoprotein and any other proteins including wheat allergens. Thus,
the glycoprotein was identified as a new wheat allergen.
3 IgE-BINDING TO POLYSACCHARIDE STRUCTURES: MANNOGLUCAN
In the meantime, it remained unclear whether a non-proteinaceous constituent in wheat also acts as an allergen. Unlike proteinaceous allergens, some non-proteinaceous substances would be more stable in our body, possibly acting as a remaining allergen to cause a longer-lasting allergic reaction. Thus, the existence of such a non-proteinaceous allergen would explain why wheat allergy is difficult to treat. Next, we aimed to isolate a polysaccharide allergen from a water-soluble fraction of wheat flour and to clarify its chemical structure and immunological properties (Tanabe et al. 2000).
The water-soluble fraction of wheat flour was first subjected to DEAE-cellulose
column chromatography to remove the proteinaceous substances. The unretained
fraction was then subjected to ConA-agarose affinity column chromatography
and gel filtration HPLC to isolate the fraction with
IgE-binding activity. The mean molecular mass was estimated to be approximately
50,000 kDa.
ConA is a specific adsorbent with an affinity
for mannose (Man)- and/or glucose (Glc)-containing polysaccharides and
glycoproteins. To clarify whether the IgE-binding compound consisted of
polysaccharide or glycoprotein, it was examined by IR spectrometry. The
IR spectrum of the allergenic compound suggested the presence of OH-groups,
with no characteristic absorption for amide groups being apparent. Therefore,
the compound appeared to consist mainly of polysaccharide. The polysaccharide
was hydrolyzed with 2 M TFA (trifluoroacetic acid), and the sugar composition
of the hydrolysate was analyzed by HPLC. The result revealed that the polysaccharide
consisted of Glc and Man in a molar ratio of 4.4 : 1, while no other common
sugars such as xylose, galactose, fucose, N-acetyl glucosamine, or N-acetyl
galactosamine were detected. Furthermore the polysaccharide allergen was
converted to oligosaccharides by hydrolysis with cellulase, suggesting
that the polysaccharide had beta-1,4-glycosidic linkages.
Judging from our detailed analysis, the polysaccharide was a novel allergen with linear beta-1,4 linkages composed of Glc and Man. While some studies have shown the presence of arabinoxylan and arabinogalactan in water extracts of wheat flour, our report was the first that clearly demonstrated the occurrence of mannoglucan in wheat flour (Tanabe et al. 2000).
The IgE-binding ability of wheat mannoglucan was confirmed by inhibition ELISA. The water-soluble fraction of wheat flour was coated on a microplate. Separately, patients’ sera were incubated with wheat mannoglucan. After blocking unoccupied sites with bovine serum albumin (BSA), the preincubated sera were used as the antibody for ELISA, and untreated sera were used as controls. This procedure was followed by the addition of biotinylated anti-human IgE, streptavidin-peroxidase conjugate, and o-phenylenediamine. As a result, wheat mannoglucan inhibited antigen-antibody binding by approximately 30% in all (four) patients allergic to the water-soluble fraction of wheat flour.
While the orally administered mannoglucan allergen would be excreted
because of its indigestible nature, it could be absorbed by the inhalation
of wheat flour. In this case, it would not be degraded, and would remain
longer in the body as a remaining allergen. This would be the probable
reason why patients sensitive to the water-soluble fraction of wheat flour
are found to possess mannoglucan-specific IgE antibodies.
As I described, there are three classes of allergens in wheat flour;
1) proteins such as alpha-amylase inhibitors, low-molecular-mass glutenin,
acyl-CoA oxidase, peroxidase, fructose-bisphosphate aldolase, and so on,
2) Asn-linked glycochains in alpha-amylase inhibitor, peroxidase, and newly
found 60 kDa protein (however it should be noted that the reactivity of
the glycan moiety in these glycoprotein has not been fully proven), and
3) polysaccharide (mannoglucan). The knowledge of allergens will contribute
to the countermeasures against the worldwide social problem, intractable
wheat allergy. For example, hypoallergenic rice and wheat flour have been
produced for patients by our group. Such products should be of great benefit
to patients as was reported for hypoallergenic rice in Nature with
the title "Japan explores the boundary between food and medicine" (Swinbanks
& O’Brien 1993).
Moreover, recent reports indicate that strategies aiming specific immunotherapy
at the level of specific T cells are promising (Secrist et al. 1993, Bellinghausen
et al. 1997, Ebner et al. 1997). We aim at the identification
of T cell epitope structures of food allergens.