Evaluation of ash pollen sensitization pattern using proteomic approach with individual sera from allergic patients


  • Edited by: Marek Kowalski

Pascal Poncet, ESPCI LECA, CNRS UMR 7195, 10 rue Vauquelin, 75005, Paris, France.


To cite this article: Poncet P, Senechal H, Clement G, Purohit A, Sutra J-P, Desvaux F-X, Wal J-M, Pauli G, Peltre G, Gougeon M-L. Evaluation of ash pollen sensitization pattern using proteomic approach with individual sera from allergic patients. Allergy 2010; 65: 571–580.


Background:  In Europe, sensitization to ash pollen induces pollinosis with cross-reactivities with other pollen sources. The aim of the study was to identify the repertoire of ash pollen allergens and evaluate the extent of the diversity of the IgE response in ash allergic patients.

Methods:  The IgE reactivities of 114 ash pollen- and eight grass pollen-sensitized patients were screened by 1D immunoblot (SDS–PAGE) against ash pollen extract. The IgE reactivities of 13 ash pollen- and two grass pollen-sensitized patients were then evaluated in 2D immunoblots. Some IgE- and non-IgE-reactive proteins were identified by mass spectrometry.

Results:  In 1D analysis, 86% of sera showed binding to Fra e 1 (18–20 kDa), 23% to Fra e 2 (14 kDa), 3% to Fra e 3 (10 kDa) and 57% to High Molecular Weight allergens (HMW, >30 kDa). Individual analysis of 2D immunoblots showed several IgE-binding protein areas among which three were more often recognized: (i) Fra e 1 comprising, at least, 15 isoforms, (ii) a series of acidic spots (45 kDa), and (iii) Fra e 2, the ash profilin. HMW allergens could be resolved in four areas; two unidentified, one homologous to β-galactosidase and the other to sugar transport proteins. A malate deshydrogenase and calmodulin were shown to be IgE-binding proteins and 10 non-IgE reactive proteins were identified.

Conclusions:  No direct correlation was evidenced between IgE profile and the degree of sensitization even though 2 spectrotypes could be distinguished. Our data contribute to a better delineation of ash pollen allergens and patterns of sensitization.

Pollinosis represents 40–60% of the allergic disorders and, as observed for other allergic diseases, its prevalence has markedly increased in developed countries during last 20 years (1). Characterization of pollen allergens is therefore a necessary step for the understanding, diagnosis and treatment of pollinosis.

Ash (Fraxinus excelsior) is a wind-pollinated tree, widely distributed in the north and middle European temperate zones. It belongs to the Oleaceae family including Forsythia, Phillyrea, Syringa, Jasminum, Ligustrum and Olea. Ash pollen sensitization has long been underestimated because an overlapping pollination period with other trees such as birch and existence of cross-reactivities with some pollen allergens from trees (mainly birch) or grasses. Furthermore, ash pollen extracts were not systematically included in routine screening assays for inhalant allergies. An extract has recently been standardized (2). Studies in France and Italy reported that 18–34% of pollinosis could be attributed to ash pollen (3–6). The symptomatology is rhinoconjunctivitis associated to asthma in about half of the cases (7).

Several ash pollen allergens have been reported since the first published study (8). Three of them, from Fraxinus excelsior, have been named. The major allergen Fra e 1, sequenced and cloned (9), is a 18–21 kDa glycosylated protein. Fra e 2 (13–16 kDa) is a member of the pan allergen profilin family and Fra e 3 (6–10 kDa), a minor allergen belonging to the Ca-binding protein family. Another Ca-binding protein (20 kDa), homologous to Ole e 8, was suspected (10) and some allergens were indirectly identified in high molecular weight (HMW) proteins (>30 kDa) (4, 5, 11).

In this study we evaluated the pattern of sensitizing ash pollen proteins and also, at an individual level, the degree of sensitization of allergic patients by a proteomic approach. The localization of allergens by western blotting of 2D gel electrophoresis of ash pollen extract and identification of proteins by mass spectrometry (MS) allowed the characterization of numerous isoforms of Fra e 1, as well as the description of four, as yet not reported, ash allergens. Furthermore, although each patient exhibited a unique pattern of sensitization, two spectrotypes may be distinguished which might correspond to two different sensitization processes.

Material and methods

Ash pollen extract

Ash pollen (Allergon AB, Angelholm, Sweden) was rotated for 1 h at 20°C in distilled water (100 mg : 1 ml). After centrifugation (10 000 g, 15 min, 4°C), the supernatant was recovered and stored in aliquots at −20°C.

Human sera and reactives

Sera were obtained from 131 patients from various French regions. Among them, 60 were from north-east of France (Strasbourg). Patients were selected on the basis of symptoms, positive skin prick test results, and positive serum-specific IgE. Sera were provided by biological analysis laboratories and represented residues of IgE titre evaluations. One hundred and fourteen sera were from patients allergic to Oleaceae (ash or olive or privet) pollen, 8 to grass pollen, 5 to horse, 1 to latex, 2 to dust mite and 1 to cat. Four sera were from nonatopic donors.

One-dimensional (1D) gel electrophoresis

SDS–PAGE: samples were run into a thin 8–18% gradient polyacrylamide gel (Excel gel, GE) on a flat-bed electrophoretic chamber (Multiphor II, Amersham-Biosciences, Uppsala, Sweden) cooled at 12°C.

IEF: samples were separated in a polyacrylamide gel (4%T, 3%C) containing 5% v/v Servalyt pH 2–11 (Serva, Heidelberg, Germany) on a flat bed electrophoretic chamber (Multiphor II) cooled at 15°C.

Two-dimensional (2D) gel electrophoresis

Untreated ash pollen extract was first submitted to IEF separation. Five mm wide strips of the focused gel were cut, incubated in an equilibration buffer (Tris acetate pH 6.8, SDS 12%) and submitted, in second dimension, to an SDS–PAGE separation on an 8–18% gradient gel (ExcelGel; GE). Two-D gels were either silver-stained or coomassie blue-stained or blotted onto activated nitrocellulose membrane.

Western blot

For SDS–PAGE and 2D gels, electroblotting of proteins was performed onto a cyanogen-bromide activated nitrocellulose (NCa) sheet (12) with a semi-dry Novablot apparatus (AB). For IEF, blotting on NCa was passive under a 1 kg weight. NCa sheet was then dried and blocked with PBS containing 0.3% Tween 20. One tenth diluted sera were individually incubated with 2.5 mm wide strips NCa for 1D screening or with the whole NCa for 2D experiments, overnight at 20°C. For the inhibition study reported in Fig. 4, sera were preincubated for 1 h with bromelain (5 U/ml) (Sigma, St Louis, MO, USA) before contact with the blotted NCa (13). After washing, NCa was incubated with alkaline phosphatase (AP)-conjugated goat anti-human IgE (Sigma Biochemicals, St Louis, MO, USA) for 2 h at 20°C, followed by AP substrate 5-Bromo-4-Chloro-3-Indolyl Phosphate + Nitro Blue Tetrazolium (Sigma Biochemicals) in 0.1 M Tris buffer pH 9.5.

Figure 4.

 Inhibition of IgE reactivities against CCD determinants with bromelain. Sera were preincubated with (lanes 1, 3, 5 and 7) or without (lanes 2, 4, 6 and 8) bromelain (5 U/ml) prior incubation with IEF blots of ash extract. Lanes 1 and 2: control serum from a nonatopic patient, lanes 3 and 4: serum #43 from a grass pollen sensitized patient, lanes 5 and 6: serum #42 from a grass pollen sensitized patient, lanes 7 and 8: serum #5 from an ash pollen sensitized patient.

MALDI-TOF analysis

The spots of interest were analysed by peptide mass fingerprinting method (14).

Briefly, protein spots were excised from a Coomassie blue stained gel. Proteins were then reduced with di-thio-threitol, alkylated with iodoacetamide and in gel-digested with porcin trypsin (Promega, Charbonnières-les-Bains, France). Resulting peptides, in solution, were added to the matrix α-cyano-hydroxy-cinamic acid, and submitted to laser shot in the mass spectrometer Applied Biosystem Voyager DE-STR instrument. The instrument was calibrated using two trypsin autodigestion peaks (m/z 842.5099 and m/z 2211.1045) to draw a calibration curve. Depending upon whether the fragments identifying a protein were in this range, the tolerance was adjusted from 4 to 500 parts per million (ppm). For example Fra e 1, which is the only protein identified with certainty, was better identified with a rather high tolerance of 200 ppm. Peptide listings were then submitted to databases using either ProFound or Protein Prospector as proteomics tools (http://www.expasy.org/tools/).


Screening of sera on 1D gel (SDS–PAGE) of water-soluble ash pollen extract

For choosing sera for 2D-analysis, 135 sera including 114 from patients allergic to Oleaceae pollen and 21 controls were screened using immunoblotting in 1D (SDS–PAGE). Representative IgE reactivity patterns on ash pollen extract of 41 sera from Oleaceae-sensitized patients and 6 from grass-sensitized patients are depicted in Fig. 1. IgE reactivities were observed against proteins with relative molecular masses (Mr) ranging from 80 to 10 kDa. Sera from nonatopic patients (n = 4) and allergic patients sensitized to grass (n = 8) or horse (n = 5) or latex (n = 1) or dust mite (n = 2) or cat (n = 1) were used as controls. None of the non atopic patient’s sera as well as sera from cat- and latex-sensitized patient showed IgE reactivity (data not shown). Two of five and one of the two sera from horse- and dust mite-sensitized patients, respectively, exhibited faint IgE reactivities around 20, 45 and 16 kDa, respectively (data not shown).

Figure 1.

 1D screening analysis using Western blot after SDS–PAGE of a water soluble extract of ash pollen. Examples of results with 41 sera from Oleaceae-sensitized patients and six grass-sensitized patient are shown. Strips were ordered according to the intensity of binding to the 20 kDa band, supposed to be Fra e 1, the major ash pollen allergen. Expected positions, according to the literature, of Fra e 2 and Fra e 3 are indicated. Arrows indicate the sera used for 2D analysis.

Table 1 summarizes the results obtained for all sera tested with ash pollen extract in 1D. Among sera from Oleaceae-sensitized patients showing a reactivity to at least one band, 86% (n = 78) exhibited an IgE reactivity around 20 kDa, probably the major allergen Fra e 1. For 18% of them, a 14 kDa band (profilin) was revealed and for 3% of them, a 10 kDa band (calcium-binding protein). Fifty-seven per cent of the sera showed IgE reactivity to HMW allergens. The ash proteins recognized by IgE from six grass-sensitized patients (of eight tested) did not include Fra e 1 but Fra e 2 and HMW allergens (Fig. 1).

Table 1.   Immunoreactivities of seric IgE from atopic and non atopic patients to ash pollen extract. Water-soluble extract was separated by SDS–PAGE followed by Western blot
PatientDiagnostic allergenNo. serum testedNo. serum positive to
At least one bandFra e 1Fra e 2Fra e 3HMW Ag
 only only only only
  1. *Ash and/or olive and/or privet

  2. †The 23 negative sera contain low levels of Oleaceae-reactive IgE (<7.7 ng/ml).

  3. ‡Percentage of total serum tested.

  4. §Percentage of the number of serum positive to at least one band.

  5. HMW, high molecular weight, >30 kDa.

Pollen allergyOleaceae*11491† (80%)‡78 (86%)§4316 (18%)‡53 (3%)§052 (57%)§1
Other allergyHorse52110011
Dust mite210000
Non atopic 40

2D gel analysis of ash pollen allergens

About 200 spots were revealed upon silver staining of a 2D gel (Fig. 2). Ash pollen proteins ranged from pI 4.0 to 10, often in series of spots with same Mr, typical of protein isoforms that are not the result of protein degradation as no change in protein pattern was observed in presence of protease inhibitors (data not shown). Two main protein-rich regions are distinguished according to Mr, a region between 10 and 20 kDa and another between 40 and 60 kDa.

Figure 2.

 2D gel electrophoresis of a water soluble extract of ash pollen. First dimension (IEF) was performed in a 4–11 pH gradient and the second dimension in an 8–18% acrylamide gradient gel. The gel was subsequently silver stained.

Immunoblotting of 2D gel electrophoresis was performed with 13 selected sera from Oleaceae- and two sera from grass-sensitized patients (Table 2). Criteria of selection were based on 1D IgE-binding profiles, i.e. complementary heterogeneity and intensity of reaction (Fig. 1).

Table 2.   Clinical data of selected patients for 2D analysis
Patient no.AllergenClassAllergenClassAllergenClassAllergenClassAllergenClass
  1. Class 0: IgE < 0.77 ng/ml; Class 1: 0.78 < IgE < 1.54; Class 2: 1.55 < IgE < 7.72; Class 3: 7.73 < IgE < 38.50; Class 4: 38.52 < IgE < 110.00; Class 5: 110.02 < IgE < 220.00; Class 6: IgE >220.02 ng/ml.

  2. rBet v 1, recombinant Bet v 1, the major birch allergen; rBet v 2: recombinant Bet v 2, the birch profilin.

  3. NA, not available (see Material and methods for sensitization assessment).

15Olive4Orchard3Cottonwood2Cypress3Dust mite0
16Olive4Festuca3Ryegrass2Parietaria2Dust mite1
32AshNArBetv13GrassNADust miteNA  

Six examples of individual patterns, four obtained with Oleaceae- and two from grass-sensitized patient sera are depicted in Fig. 3. Figure 3A displays a pattern of 35 IgE-reactive proteins distributed in four areas: (i) a series of mainly doublets around a Mr 20 kDa with heterogenous pI, (ii) a faint series of 4–5 spots around 35 kDa, pI from 5.8 to 7.0, (iii) a series of six spots around 45 kDa, pI 5.0–6.0 and (iv) an acidic spot, Mr 14.4 kDa. The series of spots around 20 kDa is also observed on the IgE immunoblot depicted in Fig. 3B together with a very acidic spot at Mr < 14.4 kDa. Partial or complete IgE binding against this former series of 20 kDa spots, is found for 12 of the 13 sera from Oleaceae-sensitized patients tested in 2D. In Fig. 3C, the serum IgE revealed two series of spots, a first one acido-neutral, partially similar to the series exhibited in Fig. 3A and a second one, basic, with approximately the same Mr, 45 kDa. In Fig. 3D, IgE from a grass pollen-sensitized patient revealed acido-neutral protein spots of Mr 45 kDa partially similar to those observed in Fig 3A and C. The 2D IgE immunoblot obtained with the second grass sensitized patient (Fig. 3E) showed reactivities with numerous proteins in HMW and an acidic protein (Mr about 14 kDa) presumed to be the profilin. These diffused reactivities to HMW allergens are similar to the reactivity pattern displayed with a serum from an ash-sensitized patient (Fig. 3F), suggesting that these patients might react with ash pollen proteins because of carbohydrate cross-reactive structures. Indeed, as shown in Fig. 4 on IEF immunoblots, preincubation of serum #42 with bromelain, reported to bear relevant carbohydrate determinants, totally inhibited the broad and diffused IgE reactivities (4.7 < pI < 8.5) and partially the profilin reactivity (pI < 4.75) in contrast to the profilin reactivity of serum #43 that was retained in presence of bromelain. Fra e 1 IgE reactivity of the serum #5 was not affected by the presence of bromelain suggesting the recognition of peptidic epitopes.

Figure 3.

 Examples of IgE immunoblots of 2D-separated ash pollen proteins. Rectangle areas are drawn to help the comparison of the various IgE immunoblots. Nitrocellulose membranes were incubated with (A) serum #1 from an Oleaceae-sensitized patient. The lower rectangle around 20 kDa corresponds to Fra e 1 reactivity. The arrow indicates IgE reactivity to Fra e 2, ash profilin. (B) serum #9 from an Oleaceae-sensitized patient. The arrow indicates IgE reactivity to Fra e 3. (C) serum #16 from an Oleaceae-sensitized patient. The basic group was identified as sugar transport proteins (see text). (D) serum #44 from a grass-sensitized patient. (E) serum #42 from a grass-sensitized patient. Note a diffused IgE profile suggestive of carbohydrate reactivities (see Fig. 4). The arrow indicates IgE reactivity to Fra e 2, ash profilin. (F) serum #31 from an Oleaceae-sensitized patient. Note a reactivity pattern closer to the one shown in panel E (grass-sentized patient) than in panels A or B (ash-sensitized patient).

Each immunoblotting IgE pattern was unique even if a given protein or a series of proteins may be recognized by two or more sera. Globally, the number of spots recognized by IgE per serum ranged from 10 to 40. No obvious correlation could be made between the 2D IgE pattern and the degree of sensitization, mono or polysensitization, of a group of patients.

Each of the 13 immunoblot patterns was schematically interpreted (an ellipse for one protein) and normalized to overlap on a standard silver stained 2D gel. Allergenic proteins were thus underlined. Such an analysis allows an evaluation of the relative frequency of recognition for a protein as well as the association of spots in a series (Fig. 5). Proteins may be delineated in several groups (see Fig. 5A–G). It can be concluded that proteins in groups C and F and the only one protein G are more often recognized as allergens by Oleaceae-sensitized patients.

Figure 5.

 The repertoire of allergens of a water soluble extract of ash pollen analysed with 13 sera from Oleaceae sensitized patients. Allergens are marked out by ellipses and overlaid on a reference 2D silver stained gel of ash pollen proteins. White ellipses are allergens recognized by more than three sera, black ellipses by two sera and dotted ellipses are allergens recognized by a single patient serum. Proteins are arbitrarily grouped from A to G. One individual protein, G, delimited in white, i.e. very often recognized, corresponds to the ash profilin.

Mass spectrometry analysis of excised protein spots

Forty spots, among the richest in proteins, were submitted to MS analysis for protein identification (Fig. 6). Only 30 of them, 24 IgE-binding and six non-IgE-binding proteins, gave consistent results. Results are summarized in Table 3. Only Fra e 1 is referenced in data banks. Therefore, the results of data bank queries reported in Table 3 include, first, the proteins exhibiting the highest probability of homology (assessed by ‘mowse’ and ‘Z’ scores given by Protein Prospector and ProFound, respectively) and second, the Viridiplantae species and genus in which this homologous protein was sequenced. Spots #24 to #39 were identified as Fra e 1, or its homologues Ole e 1 or Lig v 1, major allergens in olive tree and privet pollen, respectively. The percentages of matched peptide masses varied from 19 to 46%. All of them, except the very acidic spot 39, were recognized by IgE from the various Oleaceae sensitized patients, although a given patient can ignore some of them. The isoform around 6.7–7.0 was invariably recognized when a reactivity with Fra e 1 was observed. The spot protein #40 (Mr < 14.4, pI < 4.75), appearing as 3 very close pI variants, was identified as profilin and named Fra e 2 in ash pollen. Of the 10 processed allergenic proteins in group C, only the spot #7 was identified as a malate deshydrogenase. The spot #23 was also recognized by IgE and corresponded to calmodulin, a 4-EF hand Ca2+ binding protein. The only protein processed in group B can be referred to a β-galactosidase and the 4 proteins in group D, to isoforms of the same sugar transport protein than the one found in Beta vulgaris. Finally, besides the 24 IgE reactive proteins identified, six non-IgE binding proteins were determined: two reductases (#2, #18), a pectin methylesterase (#19), a mitochondrial glycoprotein (#20) and a protein related to Ole e 1, the major olive pollen allergen (#39). Spot #22 could not be determined.

Figure 6.

 Proteins analysed by MALDI-TOF. The locations of 40 analysed proteins are shown on the reference silver stained gel and numbered from 1 to 40.

Table 3.   Results of MALDI-TOF analysis of 40 ash proteins
SampleDatabase researchProteins
GroupNo.Mr (Da)/pIProteomics toolTol.† (ppm)Score Z (MOWSE)No. (%) masses matchedNamePlant species and genusMr (Da)/pIAccession no.
  1. Spots have been collected (see Fig. 6) and processed by an in-gel proteolytic digestion method. The collection of resulting peptides were then analysed using MALDI-TOF mass spectrometry and proteins were identified on the basis of peptide mass fingerprinting data with either proFound or Protein Prospector as proteomics tools (ExPaSy Proteomics Server).

  2. *IgE-binding protein.

  3. †Tolerance expressed as party per million (ppm).

B1*73000/7.5Profound50.913/12 (4)β-galactosidaseCITRUS SINENSIS82 700/9.0AAK31801.1
242 000/4.8Profound1922.3411/35 (25)Ascorbate reductaseMESEMBRYANTHEMUM51 960/6.4CAC82727
C3*–6* Undetermined       
7*42 000/5.8Profound2001.913/15 (14)Malate deshydrogenaseCICER ARIETINUM35 820/5.9CAC10208.1
8*–12* Undetermined       
1340 500/7.3Profound892.437/28 (23)Translocation proteinARABIDOPSIS THALIANA42 150/6.1NP 566671.1
D14*49 000/8.5Profound82.434/15 (5)Sugar transport proteinBETA VULGARIS60 060/9.5T14606
15*50 000/8.9Profound82.434/15 (5)Sugar transport proteinBETA VULGARIS60 060/9.5T14606
16*49 000/8.9Profound82.434/15 (5)Sugar transport proteinBETA VULGARIS60 060/9.5T14606
17*50 000/9.2Profound82.434/15 (5)Sugar transport proteinBETA VULGARIS60 060/9.5T14606
1834 000/5.4Profound202.439/47 (29)ReductaseFORTHYSIA X INTERMEDIA34 060/5.9AAF64174.1
Prospector15(12060)8/47 (25)ReductaseFORTHYSIA X INTERMEDIA33 900/6.47578897
1938 000/9.8Profound1031.614/27 (15)Pectin methylesteraseVITIS RIPARIA35 650/9.4AAD51853
2027 000/5.3Profound831.43/15 (23)Mitochondrial glycoproteinARABIDOPSIS THALIANA22 000/5.0NP 565244.1
2127 000/5.5Profound251.664/13 (17)Putative MURBZCZEA MAYS27 180/4.6AAN40033
2227 000/5.8Undetermined       
23*19 000/4.7Profound42.434/44 (42)CalmodulineWHEAT16 870/4.1MCWT
F24*21 000/5.3Profound600.162/10 (23)Allergen Lig v 1LIGUSTRUM VULGARE16 720/5.9O82015
Prospector40(237)4/10 (37)Allergen Fra e 1FRAXINUS EXCELSIOR16 423/5.9133327133
25*19 000/5.3Profound401.675/27 (30)Allergen (Lig v 1)LIGUSTRUM VULGARE16 790/5.9CAA54819.1
Prospector30(2322)6/27 (22)Allergen Fra e 1FRAXINUS EXCELSIOR16 423/5.9133327133
26*17 000/5.3Profound5000.775/22 (46)Allergen OLE3c (Ole e1)OLEA EUROPAEA16 650/5.7A53806
27*20 000/5.7Profound42.375/66 (31)Allergen OLE17(Ole e 1)OLEA EUROPAEA15 830/6.3E53806
Prospector7(6177)5/67 (34)Allergen OLE3c (Ole e 1)OLEA EUROPAEA16 329/5.71362135
Prospector3(577)4/67 (30)Allergen Fra e 1FRAXINUS EXCELSIOR16 423/5.9133327133
28*18 000/5.7Profound741.13/16 (19)Allergen Lig v 1LIGUSTRUM VULGARE16 790/5.9CAA54819.1
29*20 000/6.0Prospector200(8770)7/21 (33)Allergen Fra e 1FRAXINUS EXCELSIOR16 423/5.9133327133
30*20 000/6.4Prospector200(3260)6/18 (33)Allergen Fra e 1FRAXINUS EXCELSIOR16 423/5.9133327133
31*17 000/6.5Profound81.684/24 (29)Ole e 1.0103 (Ole e 1)OLEA EUROPAEA16 750/5.7CAA73037.1
32*19 000/6.8Prospector200(8760)7/24 (29)Allergen Fra e 1FRAXINUS EXCELSIOR16 423/5.9133327133
33*20 000/7.0Prospector200(21900)8/21 (38)Allergen Fra e 1FRAXINUS EXCELSIOR16 636/6.7234978692
34*18 000/7.0Prospector200(79000)9/27 (33)Allergen Fra e 1FRAXINUS EXCELSIOR16 636/6.7234978692
35*16 000/6.8Prospector200(1990)4/22 (18)Allergen Fra e 1FRAXINUS EXCELSIOR16 423/5.9133327133
36*19 000/7.8Prospector200(8760)7/17 (41)Allergen Fra e 1FRAXINUS EXCELSIOR16 423/5.9133327133
37*20 000/8.0Prospector200(594000)10/32 (31)Allergen Fra e 1FRAXINUS EXCELSIOR16 423/5.9133327133
38*17 000/8.0Prospector200(8770)6/21 (28)Allergen Fra e 1FRAXINUS EXCELSIOR16 423/5.9133327133
3919 000/<4.75Profound41.243/31 (34)Allergen OLE3c (Ole e 1)OLEA EUROPAEA16 650/5.7A53806
G40*14 200/<4.75Profound480.33/20 (11)Profilin (allergen Gly m 3)SOYBEAN14 200/4.6O65809


The knowledge of the repertoire of allergens from an allergenic source allows the classification of allergens in biochemical families and may help to distinguish between cross-reactivities and polysensitization. Data reported in this study is a tentative approach to solve this question: ‘genuine’ sensitization vs cross-reactivities by exploring, for individually studied allergic patients, the profile of all seric IgE reactivities against all ash pollen proteins of a water soluble extract. A proteomic approach applied to allergy is, for this purpose, a powerful global approach (15–17).

Results showed, with 13 patients sensitized to ash, that more than 100 proteins of ash pollen among the 200 visualized by silver staining on a 2D-gel were allergens.

A main reactivity, in 86% of sera from ash-sensitized patients, was found against 20 kDa proteins that were shown, by MS, to be the major allergen, Fra e 1. Fra e 1 was distributed in 15 isoforms, heterogeneous in Mr because of differential glycosylations (18), and exhibiting a broader range of pI (5.3–8.0) than Ole e 1 (5.0–7.2) (19). Some isoforms were sometimes identified as Ole e 1 (Olea) or Lig v 1 (Ligustrum) because of 87 and 91% identity with Fra e 1 (20). Interestingly, some patient sera did not show reactivities against all isoforms of Fra e 1.

Eighteen per cent of patients exhibited IgE reactivity against the pan allergen profilin Fra e 2 (Table 1), already described as an allergen (4, 11, 21). This frequency is close to what is reported for this allergen in various plants (22). Sera of grass-sensitized patients used in our study also showed IgE binding to ash profilin because of significant sequence homologies among plant profilins (23).

Three per cent of tested sera showed reactivity towards a 10 kDa protein described as Fra e 3, a 2-EF hand Ca-binding protein. Protein assessment in 2D (Fig. 3B) was based on Mr and pI analogies with published data (4, 11, 23, 24).

The 4-EF hand Ca-binding protein (Mr 17 kDa, pI 4.4), calmodulin of ash, was identified and found allergenic as previously suggested on the basis of cross-reactivity with Ole e 8, the calmodulin of olive (10). Ash calmodulin might be called Fra e 8.

More than half of the tested serum (57%) showed IgE reactivities against proteins with Mr >30 kDa (HMW allergens). Indeed, ash allergens were observed all along the pI range and up to 90 kDa Mr (Fig. 5). Cross inhibition experiments, direct and indirect binding with known specificity borne-antibodies and comparisons of physico-chemical characteristics with identified proteins gave rise to some hypothetical allergen identifications (4, 5, 11).

A Bet v 5-like allergen (isoflavone reductase) was suspected at Mr 35 kDa (11). We also found a neutral series of IgE-reactive proteins (group E, Fig. 5) at 35 kDa but no MS analysis was carried out on these spots because they were not visible in Coomassie blue stained gels.

IgE reactivities were also frequently found against a series of acido-neutral proteins around 43–45 kDa with ash- and grass-sensitized patient sera (Fig. 3). Only the most abundant spot, a malate deshydrogenase, also reported to be an allergen in Senecio pollen (25), was characterized. Some of the unidentified proteins in this series might correspond to the β-1.3 glucanase found in olive pollen repertoried as Ole e 9 allergen, Mr 46 kDa and pI 4.8–5.4 (26, 27). One of the anti-grass pollen serum used showed IgE reactivities against these acido-neutral proteins (see Fig. 3D). When this serum was tested in the same conditions, firstly, on orchard pollen, IgE reactivities were found on almost all allergenic groups of proteins, except Dac g 4 (28) and, secondly, on oilseed rape pollen, only pectinase (polygalacturonase, neutral, Mr 45 kDa) (29) was revealed. It remains to be clarified whether proteins in region C (Fig. 5) are ash pectinase.

The second serum from grass sensitized patient showed IgE binding to about 70 proteins, including profilin (Mr 14 kDa, acidic) and HMW allergens. These latter reactivities, broad and diffuse, were also found for some Oleaceae-sensitized patients and reported to be associated with carbohydrate cross-reactivities (4) of poor clinical relevance, in contrast to reactivities against Fra e 1. Indeed these reactivities were inhibited with bromelain, a relevant carbohydrate borne-plant enzyme (Fig. 4). It has to be noted that the two anti-grass pollen sera used in 2D study did not bind to Fra e 1 supporting the idea that a genuine sensitization to ash pollen is correlated with the presence of anti-Fra e 1 IgE.

Besides the malate deshydrogenase, two other allergens or family of allergens were newly identified in HMW ash pollen proteins: β-galactosidase (Mr 75 kDa, pI 7.5) and sugar transport proteins (Mr 50 kDa, pI 8.5–9.0). β-galactosidases have been reported to be allergenic in some Mediterranean trees (Cupressus and Olea) with various Mr and also in tomato and apple as 75–78 kDa proteins, like in ash pollen (30).

Unexpectedly, no IgE reactivity was found against ash pectin methylesterase (basic, Mr 38–40 kDa) in contrast to what was reported concerning the allergenicity of this protein (11, 31–33).

It is likely that more minor allergens would be revealed by using higher number of patient sera in agreement with the uniqueness of the individual IgE immune response. In the same line, only water soluble proteins have been studied here and it has been reported that allergens are also present in nonwater soluble extract (34, 35).

To conclude, using a proteomic approach, we analysed in this study the diversity of ash pollen allergens at a molecular level and showed that Fra e 1, the major allergen, exhibits more isoforms than it was previously reported. Fra e 1, Fra e 2 and a series of acido-neutral allergens with Mr of 43–45 kDa are frequently recognized and several minor allergens are also evidenced. Although no direct correlation can be made with clinical symptoms, two IgE profiles may be distinguished, the first one focused on Fra e 1 reactivity and the second one more broadly directed against HMW allergens that are cross-reactive mainly through carbohydrate moieties. Four ash pollen allergens such as β-galactosidase, calmodulin, sugar transport proteins and malate deshydrogenase were newly identified. The clinical relevance of the two last allergens remains to be assessed.


This work was partly supported by a grant from ‘Legs Poix’, Chancellerie des Universités de Paris, France. We thank Prof. Jean Bousquet for instigating the project.