INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The number of fungal species occurring in the human environment is estimated to be at least 100,000. Most of these species can be propagated by airborne spores, and are therefore able to enter the respiratory tract of humans and animals. Despite this, endemic mycoses, caused mainly by dimorphic fungi of the genera Histoplasma, Blastomyces, and Coccidioides (1), and opportunistic mycoses caused by fungal species that are normally harmless to the immunocompetent host (i.e., Candida, Aspergillus, and Fusarium) (1) are relatively rare. A broad array of host-defense mechanisms, both immune and nonimmune, have evolved in warm-blooded animals to protect them against fungal infections (2). However, in recent years opportunistic fungal infections have become an important medical problem, in concomitance with advances in medical awareness (3). Furthermore, the incidence of mold sensitization in patients with asthma varies from 4% to 80% in different studies (4), illustrating the difficulties related to the diagnosis of fungal allergies. Many factors, including patient selection and seasonal variations in exposure, may be responsible for the discordance in the reported incidence of fungal sensitization. However, the most important factor hindering the reliable diagnosis of fungal infection is the lack of well-defined allergen preparations (5).

Molecular biology offers the possibility of cloning, producing, characterizing, and evaluating highly pure fungal antigens. We have cloned a panel of antigens/allergens from Aspergillus fumigatus (A. fumigatus) (6), a fungus considered to be an opportunistic pathogen associated with an impressive list of diseases ranging from benign colonization of the lung and allergic diseases to life-threatening diseases such as invasive aspergillosis or allergic bronchopulmonary aspergillosis (5, 7). The cloned proteins differ in molecular weight, amino-acid composition, and allergenicity (8). There is little doubt that significant immunologic cross-reactivity exists among various species of fungi (9), but its molecular basis is still unknown. One of the A. fumigatus allergens that we have cloned, formally termed Asp f 3, shows significant sequence similarity with two gene products of Candida boidinii (C. boidinii) that are considered to be peroxisomal proteins (12). Candida species are the most frequent cause of opportunistic fungal infections (13). Moreover, allergens from Candida have been reported to play a pathogenic role in allergic asthma and allergic rhinitis (14, 15), diseases also associated with A. fumigatus allergens (5, 7). Cross-reactivity has been demonstrated between a 58-kD antigen of A. fumigatus and an antigen of approximately 55 kD of C. albicans (16), although the biochemical nature of these antigens has not been elucidated.

In this study we show that a cloned major allergen of A. fumigatus (Asp f 3) shares extended amino-acid similarity with two peroxisomal proteins of C. boidinii. Moreover, these three proteins, produced in Escherichia coli, can bind IgE from sera of individuals sensitized to A. fumigatus and compete with each other for IgE binding, suggesting that the allergens share IgE-binding epitopes.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Sera from a group of 89 individuals (48 men and 41 women) sensitized to A. fumigatus, aged 48 ± 13 yr (mean ± SD), were selected according to case histories, skin tests and radioallergosorbent testing (RAST). All of the subjects had stable asthma, met the guidelines for the diagnosis and management of asthma (17), were free of Candida infections, and received no antihistamine medication when the serum sample was taken. Total IgE ranged from 118 to 6,456 kU/L (mean: 1,319 kU/L), and RAST to A. fumigatus ranged from 2 to 6 (mean: 3.6). A second group, of 11 healthy control subjects (four women and seven men), aged 29 ± 5 yr, without a history of atopy and with normal IgE levels, was also involved in the study.

PCR Amplification, Cloning, and Characterization of rAsp f 3, rPMPA, and rPMPB

IgE-binding phages from a previously constructed A. fumigatus complementary deoxyribonucleic acid (cDNA) library were enriched as described (6). Phages were used to infect E. coli XL-1 Blue cells. The replicative form of phagemids was isolated and double-stranded sequenced with the dideoxy chain-termination method (18), using fluorescent primers. The sequence information was used to design specific polymerase chain reaction (PCR) primers for subcloning of inserts. One of these cDNAs, designated Asp f 3 (GenBank Accession No. U58050) showed sequence similarity with two genes from C. boidinii. Therefore, genomic DNA from C. boidinii, strain ATCC 32195, was isolated as described elsewhere (12). The published sequences of the genes encoding peroxisomal membrane protein A (PMPA) and PMPB (GenBank Accession Nos. J04984 and J04985, respectively) (12) were used to design specific primers for PCR amplification of the genes (Table 1). The same primers were used to amplify putative genes encoding peroxisomal proteins from C. albicans chromosomal DNA. PCR reactions were performed in a final volume of 50 µl containing 30 µl Amplimix (Anawa, Wangen/Zürich, Switzerland), 1 µM each of a 5'- and a 3'-primer (Table 1), 1 ng phagemid pJuFo (19) carrying Asp f 3 or 0.1 µg of genomic DNA from C. boidinii or C. albicans, and 1.5 U Taq polymerase (Stratagene, La Jolla, CA). PCR consisted of 30 cycles with the following steps: 1 min at 95° C, 1 min at 56° C, and 1 min at 72° C, followed by a terminal extension of 10 min at 72° C. The size of PCR products was verified by running 10-µl samples on 1.2% agarose gels. The remaining PCR samples were purified from primers, nucleotides, and polyemrase with the QIAquick PCR purification Kit (Qiagen, Hilden, Germany), digested with the appropriate restriction enzymes, and separated on 1.2% agarose gels. The desired DNA bands were purified according to the Geneclean procedure (Bio 101 Inc., La Jolla, CA), ligated into BamH I/Nhe I-restricted (for rAsp f 3) or BamH I/Hind III-restricted (for rPMPs) dephosphorylated pDS56/RBSII plasmids (20), and introduced into competent E. coli M15 pREP4 cells through electroporation with a Gene Pulser (Bio-Rad, Richmond, CA). Transformants were selected on Luria broth (LB)-agar-plates containing ampicillin (100 µg/ml) and kanamycin (25 µg/ml). Recombinant plasmids were isolated with commercial kits (Qiagen), and inserts were verified by sequence determination (18).

Sequence Analysis

Sequence comparisons were performed with the Genetics Computer Group programs FASTA and BESTFIT (Wisconsin Sequence Analysis Package, Madison, WI).

Production and Purification of Recombinant Proteins

E. coli M15 cells containing the repressor plasmid pREP4 (20) and a chimeric pDS56/RBSII plasmid carrying either Asp f 3 or one of the PMP-coding sequences were grown at 37° C, 220 rpm in 2× YT- medium to A₆₀₀ = 0.8. At this point, Isopropyl--D-thiogalactoside (IPTG, 2 mM) was added to induce expression of the recombinant proteins, and fermentation was continued for a further 5 h. The recombinant [His]₆-fusion-proteins, deposited in inclusion bodies, were purified under denaturing conditions over Ni²⁺-nitrilotriacetic acid (NTA)/Sepharose columns as described (20). Concentrations of the eluted protein solutions were determined according to the Bradford method, using a commercial kit (Bio-Rad) with bovine serum albumin (BSA) as standard. To obtain soluble proteins in a physiologic buffer, samples were dialyzed against phosphate-buffered saline (PBS) pH 7.4, sterile filtered, and stored at 20° C.

Sodium Dodecylsulfate-Polyacrylamide Gel Electrophoresis

Polyacrylamide gels were run under denaturing and reducing or denaturing and nonreducing conditions, using a Mini 2-D cell and 4 to 20% polyacrylamide gradient gels (Bio-Rad). Whole-protein lysates of E. coli were prepared by boiling pellets of 100 µl cell suspension for 7 min in sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer containing -mercaptoethanol (0.7 M), followed by sonication for 30 s (20). Polyacrylamide gels were stained with Coomassie brilliant blue R-250 (Merck, Darmstadt, Germany) according to the manufacturer's recommendations.

Western Blotting

Recombinant proteins (0.5 µg rAsp f 3, 2 µg for both rPMPs) and the lysed pellet of 100 µl uninduced cell suspension as control were separated on homogenous 12% polyacrylamide gels (Novex, San Diego, CA) under denaturing and reducing conditions. For semidry blotting, proteins were transferred to Hybond^TM -ECL nitrocellulose membranes (Amersham, Buckinghamshire, UK), using a 2177 Multiphor II Electrophoresis system (LKB, Bromma, Sweden) with a discontinuous buffer system at 1.2 mA/cm² membrane for 2 h. Membranes were blocked at room temperature for 1 h with Buffer A (PBS, pH 7.4, containing 5% (vol/vol) Tween 20 and 5% (wt/vol) nonfat dry milk powder) and washed for 30 min with Buffer B (PBS, pH 7.4, containing 0.05% (vol/vol) Tween 20), with three changes of washing Solution. Patients' sera were diluted 1:100 in Buffer A and incubated with the membrane for 2 h at room temperature (RT). The binding of specific IgE antibodies to the proteins was detected with TN-142 mouse antihuman IgE monoclonal antibody (mAb) in Buffer A (130 ng/ml, 2 h, RT) and developed with peroxidase (PO)-labeled goat antimouse IgG heavy and light (H + L) chains (TAGO, Burlingame, CA) in Buffer A (170 ng/ml, 1 h, RT). Bands were visualized with the Renaissance Western blot chemiluminescence reagent (DuPont, Boston, MA) as recommended by the manufacturer.

Inhibition ELISA

Maxisorp-polystyrene microtiter plates (Nunc, Roskilde, Denmark) were coated overnight at 4° C with rAsp f 3 (10 µg/ml), washed with Buffer B, and blocked for 1 h at 37° C with Buffer A. In a separate, untreated 96-well plate (Petra Plastic, Chur, Switzerland), serially diluted rAsp f 3, rPMPA, rPMPB, or rAsp f 1/a as a control (21) were incubated for 2 h at 37° C with serum from individuals sensitized to rAsp f 3. Sera preincubated with antigens were transferred to the coated and washed plate and incubated for a further 2 h at 37° C. After washing and development (see specific IgG and IgE ELISA), percentage of inhibition was calculated by setting the absorbance obtained at 405 nm without antigen in the fluid phase as 0% inhibition. For graphics display, the calculated percentages were plotted against the antigen concentration in the fluid phase.

rAsp f 3- and rPMPA-specific IgE and IgG ELISA

Maxisorp polystyrene microtiter plates (Nunc) were coated overnight at 4° C with 10 µg/ml allergen in PBS, pH 8.0, washed, and blocked. After washing, sera were diluted 1:5 for specific IgE assay and 1:50 for specific IgG assay in Buffer A, further serially diluted twofold in the coated plate, and incubated for 2 h at 37° C. Subsequently, alkaline phosphatase (AP)-conjugated goat antihuman IgG (TAGO), diluted in Buffer A (1:10,000), or TN-142 mouse antihuman IgE mAb diluted in Buffer A (400 ng/ml), was added for 2 h at 37° C. TN-142 was detected with AP-labeled goat antimouse IgG H + L chains (Pierce, Rockford, IL) in Buffer A (500 ng/ml, 1 h, 37° C). ELISA plates were developed with 1.5 mg/ml 4-nitrophenylphosphate disodium salt hexahydrate (Merck) in diethanolamine buffer at pH 9.8 for 45 min. The enzyme reaction was stopped by adding 50 µl 2 M NaOH. Absorbance readings were measured at 405 nm with a Molecular Devices Reader (Menlo Park, CA) and converted to arbitrary ELISA units (EU) as described (21). From the serum dilution series for each patient, an average EU value was calculated from the values within the titratable region. Values below 1 EU/ml were set as 1 EU/ml for graphic display and nonparametric statistical analysis (Mann-Whitney U test).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isolation of Asp f 3 cDNA from a Phage Surface Display Library and Sequence Analysis

Biopanning of a fungal cDNA library displayed on phage surface (6, 8, 22), using IgE from serum of A. fumigatus-sensitized individuals as ligand, yielded a vast variety of phage displaying IgE-binding proteins. One of these cDNAs has a length of 616 bp (Figure 1). The correct reading frame for this cDNA was identified with start and stop codons (23), and spans a region of 507 nucleotides (nt), predicting a protein of 168 amino acids with a calculated molecular mass of 18,451 Da. The 3'-noncoding region includes a 53-nt-long poly(A) tail located 57 nt downstream from the TAA termination codon. A potential polyadenylation signal could be identified at position 512 (ATGAAA)a sequence that was also found in the Cladosporium herbarum allergens Cla h 3 and Cla h 4 (24).

View larger version (45K): [in this window] [in a new window]   Figure 1.   DNA and the deduced amino-acid sequence of Asp f 3. DNA sequence of Asp f 3 (A) with the corresponding amino acids (B). Numbers on the right side of each line represent the position of the last nucleotide or amino acid, respectively. The potential adenylation signal sequence is underlined.

Relation of Asp f 3 to Published Sequences

The amino acid sequence deduced from the verified nucleotide sequence of Asp f 3 was used for a homology search in the SWISS-PROT data base, and revealed significant sequence similarities between Asp f 3 and two peroxisomal membrane proteins of C. boidinii, termed PMPA and PMPB (SWISS PROT Accession Nos. P14292 and P14293, respectively; Figure 2). The calculated molecular weights of the three proteins were 18.5 kD for rAsp f 3, 18.0 kD and 18.1 kD for PMPA and PMPB, respectively. The sequence identity between Asp f 3 and both PMPs was 36.1%, and the degree of similarity reached 57.6% for PMPA and 58.2% for PMPB, respectively. The small discrepancy in similarity between Asp f 3 and the two Candida proteins is due to 35 nucleotide differences in the coding sequences of PMPA and PMPB, which, however, result in only five amino acid changes (Figure 2).

View larger version (54K): [in this window] [in a new window]   Figure 2.   Deduced amino-acid sequence of Asp f 3 compared with PMPA and PMPB. (B) The deduced amino-acid sequence of Asp f 3, (A) amino-acid sequence of PMPA, and (C) amino-acid sequence of PMPB. Identities are marked by vertical bars and conservative exchanges by colons. Gaps due to alignment are indicated by dots. Numbers on the right side represent the positions of the amino-acid residues. Differences between PMPA (A) and PMPB (C) are marked by asterisks.

Production, Purification, and Analysis of rAsp f 3, rPMPA, and rPMPB

Selected transformants harboring the expression vector pDS56 with inserts encoding Asp f 3, PMPA, or PMPB were used to produce recombinant proteins. Lysates of small fractions of the cell cultures (100 µl) were analyzed with SDS- PAGE, the bulk cultures were lysed, and the recombinant proteins purified over Ni²⁺-NTA/Sepharose columns under denaturing conditions. The yield was about 30 mg rAsp f 3, 90 mg rPMPA, and 100 mg rPMPB per liter of induced E. coli culture. Total cell lysates and the purified proteins were analyzed with a denaturing and reducing, 4 to 20%-gradient SDS- PAGE. Lanes 3 and 5 in Figure 3A show the pattern of induced E. coli M15 cells with rAsp f 3 and rPMPA. The induced bands are visible in the region around 20 kD as compared with the pattern produced by a noninduced culture (Lane 2). Lanes 4, 6, and 7 demonstrate that all three proteins were essentially depleted from E. coli proteins by affinity chromatography. The molecular weights of the recombinant [His]⁶-tagged proteins predicted from the nucleotide sequences correspond to 19.9 kD for rAsp f 3, 19.4 kD for rPMPA, and 19.5 kD for rPMPB, which are slightly lower than the observed values in the gel, probably reflecting the inaccuracy of molecular-weight estimation with reducing SDS-PAGE (25). When run under denaturing and nonreducing conditions, all three proteins showed the same molecular weight of about 20 kDa (Figure 3B), and dimers of both rPMPs are clearly visible. Oligomeric forms of rAsp f 3 were not detectable under these conditions. However, immunoblot analyses with sera of patients sensitized to A. fumigatus revealed an IgE-binding dimer of rAsp f 3 (data not shown).

View larger version (57K): [in this window] [in a new window]   View larger version (36K): [in this window] [in a new window]   Figure 3.   (A) Analysis of purified recombinant Asp f 3, PMPA, PMPB, and lysates of induced and uninduced E. coli cultures with reducing SDS-PAGE. Pellets of 100 µl E. coli cell culture were lysed and subjected to electrophoresis. Clear bands of induced rAsp f 3 (Lane 3) and rPMPA (Lane 5) are visible, as compared with their absence in uninduced E. coli M15 cells harboring rAsp f 3 (Lane 2). Lanes 4, 6, and 7 show rAsp f 3, rPMPA, and rPMPB, each at 2 µg, after purification with affinity chromatography. The molecular weight marker in Lane 1 is indicated (kDa) on the left side. (B) Analysis of purified recombinant Asp f 3, PMPA, and PMPB with nonreducing SDS-PAGE. rAsp f 3 (Lane 2), rPMPA (Lane 3), and rPMPB (Lane 4) at 2 µg each were separated with a 4 to 20%-gradient SDS-PAGE. Monomeric forms of all three proteins are visible at about 20 kDa, dimers of both rPMPs at about 40 kDa. Dimers of rAsp f 3 were not detectable under these conditions.

IgE Western Blots with Purified rAsp f 3, rPMPA, and rPMPB

We analyzed the IgE-binding capacity of rAsp f 3 and of the two rPMPs with Western blots, using serum from patients sensitized to A. fumigatus. Lanes 2 to 4 in Figure 4 show examples of blots obtained with purified recombinant proteins, with all three proteins exhibiting clear, background-free IgE binding properties. No IgE against E. coli proteins could be detected in the patients' sera (Lane 1).

View larger version (35K): [in this window] [in a new window]   Figure 4.   IgE-binding properties of rAsp f 3 and the two rPMPs in Western blotting. The lysed pellet of 100 µl uninduced cell culture (Lane 1, Control), 0.5 µg rAsp f 3 (Lane 2), 2 µg rPMPA (Lane 3), and 2 µg of rPMPB (Lane 4) were separated on a 12% single-percentage reducing SDS-PAGE and blotted onto a nitrocellulose membrane. IgE-binding capacity of the proteins was shown with serum from a patient sensitized to rAsp f 3. Lane 1 shows that the serum is free of IgE directed against E. coli proteins.

IgE Inhibition ELISA with rAsp f 3 and rPMPs in Fluid Phase

In a further step, we investigated whether rAsp f 3 and the two PMPs shared common IgE-binding epitopes. The binding of serum IgE from rAsp f 3 sensitized patients to rAsp f 3 coated on a solid phase was inhibited by increasing amounts of rAsp f 3 in the fluid phase (Figure 5A through C). Analysis of the inhibition properties of rPMPA and rPMPB revealed that both Candida proteins were also capable of inhibiting binding of IgE to rAsp f 3. In all three patients investigated, the inhibitory ability of rPMPA was slightly greater than that of rPMPB. However, as compared with rAsp f 3, a greater amount of fluid-phase antigen was required to achieve inhibition for both Candida proteins. Preincubation of serum with any concentration of rAsp f 1/a failed to inhibit the binding of IgE to rAsp f 3. These inhibition experiments suggest that rAsp f 3 and rPMPA and rPMPB share some IgE-binding epitopes.

View larger version (26K): [in this window] [in a new window]   Figure 5.   Inhibition of IgE binding to rAsp f 3 coated on a solid phase by rAsp f 3, rPMPA, and rPMPB in the fluid phase. Sera from three patients (A through C ) were preincubated with different amounts of rAsp f 3 (solid triangles), rPMPA (solid circles), rPMPB (solid squares), or rAsp f 1/a (open circles, control) in fluid phase. Preincubated samples were transferred to rAsp f 3-coated wells and IgE-binding was analyzed with rAsp f 3-specific ELISA.

Incidence of Sensitization to rAsp f 3 and rPMPA

To analyze the incidence of sensitization to rAsp f 3 and the rPMPs, we selected 89 patients sensitized to A. fumigatus and 11 control individuals without a history of atopy, and analyzed their sera by ELISA for the presence of rAsp f 3- and rPMP-specific IgE. As shown in Figure 6A, the group of sensitized patients (Mean value: 293 EU/ml) exhibited significantly higher levels of specific IgE against rAsp f 3 than did the control group (mean value: 2 EU/ml; p < 0.001). The study data show that rAsp f 3 is a major allergen of A. fumigatus, since 72% of the patients had significant amounts of specific IgE against rAsp f 3 in their serum, whereas in the control group none of the subjects had specific IgE values above the cutoff value, which was defined as 10 EU/ml according to skin tests with rAsp f 3 (data not shown). Figure 6B shows the ELISA results for rPMPA. Compared with the controls (mean value: 37 EU/ml), patients sensitized to A. fumigatus displayed significantly higher levels of rPMPA-specific IgE in their serum (mean value: 100 EU/ml; p < 0.005). Comparable results were obtained for the rPMPB-specific IgE serum levels (data not shown).

View larger version (17K): [in this window] [in a new window]   Figure 6.   Specific IgE against rAsp f 3 and rPMPA in sera from individuals sensitized to A. fumigatus and control individuals. Eighty-nine sera of patients sensitized to Aspergillus fumigatus and 11 sera of control subjects were analyzed for their content of (A) rAsp f 3- and (B) rPMPA-specific IgE with an ELISA. Units are expressed as arbitrary units calculated according to a reference serum pool of two ABPA patients, which was set as 100 EU/ml. The cutoff value of  10 EU/ml (hatched area) was defined according to skin-test results with rAsp f 3 (data not shown). Mean values are indicated by bars.

Specific IgG levels raised against PMPA and PMPB in Aspergillus-sensitized patients and controls reached mean values of about 70 EU/ml and 50 EU/ml, respectively, and showed no statistical differences. In contrast, IgG raised against Asp f 3 reached a mean of 109 EU/ml in Aspergillus-sensitized patients, and was significantly increased as compared with the values obtained for control individuals (9 EU/ml).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We report the cDNA sequence and the deduced amino-acid sequence of an allergen of A. fumigatus formally designated Asp f 3 (26). At the protein level, Asp f 3 showed significant similarity with two peroxisomal membrane proteins (PMPA and PMPB) of C. boidinii (approximately 58% similarity and 36% identity with each). Cross-inhibition experiments demonstrated that rAsp f 3 shares common IgE-binding epitopes with both rPMPs, since addition of recombinant PMPs to serum from Aspergillus-sensitized individuals significantly inhibited the binding of IgE antibodies to rAsp f 3. As compared with the inhibition obtained with rAsp f 3 in the fluid phase, greater amounts of rPMPA and rPMPB were necessary to inhibit IgE binding. However, all three proteins strongly bound IgE in Western blots. From these results we conclude that the two Candida proteins share some, but not all, of their IgE-binding epitopes with rAsp f 3. As demonstrated by an extended serologic investigation, 72% of sera from individuals showing an immediate skin reaction to A. fumigatus extracts exhibit significant amounts of IgE directed against rAsp f 3, demonstrating that rAsp f 3 is a major allergen of the fungus (26). The combination of rAsp f 1/a, a major allergen of A. fumigatus described earlier (21), and rAsp f 3 allows serologic detection of 90% of individuals sensitized to A. fumigatus extracts (data not shown), which is a high percentage considering that the fungus can produce more than 40 distinct allergenic proteins (27). The clinical relevance of rAsp f 1/a (21) and rAsp f 3 has been demonstrated in skin challenges showing that both recombinant allergens can provoke mediator release in vivo.

IgE-mediated cross-reactivity between products of different fungal species has been described (9, 28). Apart from sensitizations to species-specific antigens, which certainly occur, a portion of these polyvalent sensitizations is probably due to cross-reactivity between homologous proteins distributed among different fungal species (11). Cross-reactivity between Aspergillus and Candida has been shown (16), but in the studies in which this was done the molecular nature of the components responsible for the cross-reactivity was not identified. The IgE-binding data for rAsp f 3 from A. fumigatus and the rPMPs from C. boidinii represent a first example of cross-reactivity between fully characterized fungal products. Moreover, PCR amplification using chromosomal DNA from C. albicans as target and the PMP-specific primers yielded amplification products of the expected size, indicating that rAsp f 3-related genes are also present in this clinically important yeast (data not shown). Sequencing and production of the proteins encoded by these PCR products will allow clarification of whether peroxisomal proteins play a role also in sensitization to clinically important C. albicans strains.

The fungal extracts currently used for in vivo and in vitro diagnosis of sensitization, prepared from raw material (29), are not altogether satisfactory, and need to be improved for dependable diagnoses (5, 27, 30). The availability of pure, fully characterized recombinant allergens will allow investigation of patient-specific reactivity patterns (6, 31) and identification of cross-reactive IgE-binding proteins at the molecular level. However, in order to fully understand the molecular basis of cross-reactivity and to study the pathophysiologic mechanisms underlying cross-reactive sensitizations, more allergens from different sources need to be cloned and characterized, since only a limited number of fungal antigens/allergens have so far been identified at the molecular level (21, 24, 32).