The Spectrum of Fungal AllergySimon-Nobbe B. · Denk U. · Pöll V. · Rid R. · Breitenbach M.
Department of Cell Biology, University of Salzburg, Salzburg, Austria Corresponding Author
Correspondence to: Dr. Birgit Simon-Nobbe
Department of Cell Biology, Division Genetics
University of Salzburg, Hellbrunnerstrasse 34
AT–5020 Salzburg (Austria)
Tel. +43 662 8044 5791, Fax +43 662 8044 144, E-Mail email@example.com
Fungi can be found throughout the world. They may live as saprophytes, parasites or symbionts of animals and plants in indoor as well as outdoor environment. For decades, fungi belonging to the ascomycota as well as to the basidiomycota have been known to cause a broad panel of human disorders. In contrast to pollen, fungal spores and/or mycelial cells may not only cause type I allergy, the most prevalent disease caused by molds, but also a large number of other illnesses, including allergic bronchopulmonary mycoses, allergic sinusitis, hypersensitivity pneumonitis and atopic dermatitis; and, again in contrast to pollen-derived allergies, fungal allergies are frequently linked with allergic asthma. Sensitization to molds has been reported in up to 80% of asthmatic patients. Although research on fungal allergies dates back to the 19th century, major improvements in the diagnosis and therapy of mold allergy have been hampered by the fact that fungal extracts are highly variable in their protein composition due to strain variabilities, batch-to-batch variations, and by the fact that extracts may be prepared from spores and/or mycelial cells. Nonetheless, about 150 individual fungal allergens from approximately 80 mold genera have been identified in the last 20 years. First clinical studies with recombinant mold allergens have demonstrated their potency in clinical diagnosis. This review aims to give an overview of the biology of molds and diseases caused by molds in humans, as well as a detailed summary of the latest results on recombinant fungal allergens.
© 2007 S. Karger AG, Basel
Fungi are eukaryotic, non-chlorophyllous and heterotrophic organisms that depend on external nutrients and therefore live as saprophytes, parasites or symbionts of animals and plants under nearly all environmental conditions. The phenotype of molds ranges from a unicellular to a dimorphic or filamentous appearance. Out of over 100,000 fungal species reported, a few hundred occur as opportunists and about 100 are known to elicit mycoses in man and animals . More than 80 mold genera have been shown to induce type I allergies in susceptible persons, whereas allergenic proteins have been identified in 23 fungal genera.
For decades, fungal spores and mycelial cells have been known to be a major health risk. In contrast to airborne pollen, fungal spores are not primarily associated with IgE-mediated type I allergies but also with a broad panel of other diseases, e.g. life-threatening primary and secondary infections in immunocompromised patients. Additionally, molds have been described to cause allergic bronchopulmonary mycosis (ABPM) and hypersensitivity pneumonitis, fungal sinusitis and toxic pneumonia, and a large number of mycotoxins have been listed [2, 3]. The broad panel of diseases results from the inhalation and ingestion of fungal spores and vegetative cells (hyphae) or the contact with fungal cells. In contrast to other allergenic sources, fungi are very common in the environment, and exposure to airborne spores is almost constant throughout the year. A major difference to other sources, e.g. house dust mite or pollen, is that fungi may colonize the human body, and they may damage airways by the production of toxins, proteases, enzymes  and volatile organic compounds . Thus, molds have a far greater impact on the patients’ immune system than pollen or other allergenic sources.
Fungi are eukaryotic, filamentous and mostly spore-bearing organisms representing a separate entity within living organisms. In general, a sexual generation is followed by an asexual generation during a life cycle. Each of these generations may propagate independently, exhibiting different morphologies (pleomorphism). The broad majority of allergy-causing molds belong to the divisions of ascomycota or basidiomycota. Ascomycota produce their ascospores in the course of sexual reproduction in the ascus, whereas basidiomycota produce their meiospores or basidiospores, respectively, in the basidium. About 30,000 species of ascomycota and 25,000 species of basidiomycota have been described. The size of fungal spores ranges from 2–3 µm (Cladosporium, Aspergillus and Penicillium) up to 160 µm (Helminthosporium). The average size lies between 2 and 10 µm, but spores of 500 µm (Alternaria longissima) have also been found.
Although optimal growth conditions vary among molds, their optimal growth temperature ranges from 18 to 32°C. For growth, they require oxygen, water and a carbohydrate source. Molds occur in outdoor and indoor environments, and they grow on virtually any substrate, including glass and plastic surfaces.
The outdoor spore concentration ranges from 230 to 106 spores/m3[7, 8]. Atmospheric fungal spore concentration exceeds mean pollen concentration 100–1,000 times . Spore concentration in the air varies substantially depending on climatic factors such as temperature, wind and moisture. The majority of the fungal species grow in the outdoor environment. Examples are Alternaria, Cladosporium, Epicoccum and Ganoderma.
Indoor fungi are a mixture of those growing indoors and those that have entered from outdoors . Their incidence is influenced by humidity, ventilation, the content of biologically degradable material, and the presence of pets, plants and carpets . In general, indoor spore concentration is less than half of the outdoor count (unless there is indoor mold growth) varying from 100 to 1,000 spores/m3[10, 12]. In a Danish study on 23 mold-infected buildings, the most frequent mold genera encountered were Penicillium (68%) and Aspergillus (56%), followed by Chaetomium, Ulocladium, Stachybotrys and Cladosporium (ranging from 22 to 15%) .
The immunological mechanisms underlying mold allergies are hypersensitivity reactions of types I, II, III and IV. The spectrum of allergic symptoms caused by these hypersensitivity reactions is very broad, including rhinitis, asthma, atopic dermatitis (AD) and ABPM. Since this review has its main focus on IgE-mediated type I allergies, only a short overview about allergic diseases of types II, III and IV is given.
Most frequently, ABPM is caused by Aspergillus fumigatus, which may grow in the bronchial lumen, leading to a persistent bronchial inflammation inducing bronchiectasis in asthmatic patients. Seven to 22% of asthmatic patients suffer from allergic bronchopulmonary aspergillosis (ABPA) . Besides A. fumigatus, ABPM is induced by Candida albicans, Curvularia, Geotrichum and Helminthosporium . Allergic reactions involved include types I, III and IV.
Molds (e.g. Aspergillus, Curvularia, Alternaria and Bipolaris) may cause allergic sinusitis and fungal ball production in the patients’ sinuses . In case of allergic sinusitis, multiple sinuses are affected, whereas tissue invasion does not occur. In the patients’ mucus, fungal hyphae are detectable. Additionally, patients may show a cutaneous hypersensitivity to specific allergens along with specific IgE and IgG antibodies and an elevated total IgE level . Immunologically, allergic sinusitis is a type I-, III- and IV-mediated allergic reaction.
Hypersensitivity pneumonitis (also known as extrinsic allergic alveolitis) is based on type III/IV allergic reactions to repeated inhalation of allergens and may lead to a chronic disease with irreversible lung damage. It is characterized by the presence of precipitating antibodies and an antigen-induced lymphocyte stimulation. The following molds have been associated with hypersensitivity pneumonitis: Aspergillus and Penicillium species, and the basidiomycetes Lentinus edodes, Merulius lacrymans and P. ostreatus[16, 17].
Molds not only cause various allergic reactions but they may also produce mycotoxins which affect the immune system.
Mycotoxins – non-volatile, secondary metabolites of low molecular weight produced by fungi – impair the immune system and have neurotoxic, mutagenic, carcinogenic and teratogenic effects. Diseases caused by mycotoxins are called mycotoxicoses. The severity of toxic effects depends on the type of mycotoxin, the duration and dose of exposure and the age, health and nutritional status of the individual affected. Mycotoxins may occur in spores, mycelia, and the matrix in which fungi grow. They are a health risk for farm workers, for persons living in houses with excessive mold growth and for persons exposed to moldy material at the workplace. So far, approximately 300 mycotoxins have been identified. Chronic exposure to mycotoxins causes immunosuppression of varying extent. Prominent examples for mycotoxins are aflatoxin (Aspergillus flavus and A. parasiticus), ergot alkaloids (Claviceps spp., A. fumigatus and Penicillium chermesinum), ochratoxins (A. ochraceus, A. alliaceus, A. terreus, P. niger and P. viridicatum) and trichothecenes (Fusarium sporotrichioides, Microdochium nivale and Stachybotrys atra)[18, 19].
Type I allergy is induced by a large number of fungal genera. The majority of them are members of the ascomycota or the basidiomycota. The most important allergy-causing fungal genera belonging to the ascomycota are Alternaria, Aspergillus, Bipolaris, Candida, Cladosporium, Epicoccum and Phoma, whereas Calvatia, Coprinus, Ganoderma, Pleurotus and Psilocybe are the most prominent genera of the basidiomycota (table 1). In table 1, all allergy-causing fungal genera belonging to the ascomycota, the basidiomycota and the zygomycota along with their prevalence reported in the literature are listed.
The incidence of mold allergy ranges from 6  to 24%  in the general population, up to 44% among atopics  and 80% among asthmatics . The incidence of mold allergy within asthmatic children is 45% whereas it is 70% in asthmatic adults .
Allergic rhinitis is characterized by sneezing, rhinorrhea, pruritus and nasal obstructions. It is induced by a large number of fungal species, with Alternaria, Aspergillus, Bipolaris, Cladosporium, Curvularia and Penicillium being the most prominent.
Comparing the size of pollen grains and fungal spores, it is obvious that fungal spores are smaller in general. Therefore, they may reach the alveolar surface of the lung inducing chronic inflammation of the lung tissue [26, 27].
In many studies, an apparent link between asthma and fungal sensitization was described . In children, fungal allergy was shown to be associated with increased bronchial reactivity [29,30,31], whereas in adults severe asthma, intensive care unit admission and even death was observed [32, 33]. In an US study performed in asthmatic patients, up to 80% of the subjects showed sensitization to molds . In a study on 981 4-year-old children from the Isle of Wight (UK), asthma was the most common disease in children sensitized to molds . Reed  stated that fungi have been considered an important cause of asthma for more than 60 years. In a Canadian study dealing with ‘thunderstorm asthma’, high spore (but not pollen) counts in the course of thunderstorms were strongly correlated with asthma exacerbations . Additionally, a strong association between fungal sensitivity, exposure to fungal spores and life-threatening asthmatic episodes was described [28, 36]. Taken together, the molds Alternaria, Aspergillus, Cladosporium, Helminthosporium, Epicoccum, Aureobasidium and Penicillium have frequently been implicated in allergic asthma [27,37,38,39].
AD is a chronic inflammatory disease of the skin that is associated with high levels of total and allergen-specific IgE .
In recent years, Malassezia furfur has been implicated in the pathogenesis of AD whereas 40–65% of AD patients either have a positive skin test, atopy patch test or radioallergosorbent test (RAST) with M. furfur extract . Sensitization to Malassezia allergens may be favored by impaired epidermal barriers, increased T-cell reactivity and distinctive features of antigen-presenting cells [42, 43]. Manganese-dependent superoxide dismutase (MnSOD) may be involved as an autoallergen in the pathogenesis of AD; 36% of patients with a positive Malassezia sympodialis skin test (n = 69) react with fungal and human MnSOD .
Saccharomyces cerevisiae is another yeast species showing a significant correlation between a positive skin prick test (SPT) and AD .
Among molds associated with allergic disorders, A. alternata is one of the most frequently encountered species, predominantly occurring in the outdoor environment. The incidence of A. alternata sensitization within atopics varies between 3.6 and 39.4% (table 1) depending on the climatic zone and the population tested.
Mari et al.  showed that in a cohort of 4,962 patients with respiratory symptoms, 65% were SPT positive to at least one allergenic source, and 19% of these allergics reacted to at least one fungal extract, whereas the incidence of sensitization to A. alternata was 66%. Interestingly, within the group of patients being sensitized to a single fungal species, Alternaria, Candida and Trichophyton were the most common.
In several studies, a strong association between an A. alternata sensitization and asthma severity was demonstrated [26, 27, 29, 31, 37, 38, 47]. In a cross-sectional study by Zureik et al. , asthma severity was not associated with sensitization to pollen and cats. According to a study by Halonen et al. , Alternaria sensitization at the age of 6 and 11 years, respectively, resulted in a statistically significantly increased risk to develop asthma in childhood. In a large scale study performed in the United States, 38.3% of 1,286 asthmatic children had positive skin test responses to Alternaria species .
Before 1990, little was known about the relevant allergens of A. alternata. Meanwhile 13 allergens of A. alternata have been identified (table 2). Most of these allergens are intracellular housekeeping proteins. Nine of these allergens, e.g. NADP-dependent mannitol dehydrogenase, enolase, aldehyde dehydrogenase, flavodoxin (YCP4 homolog), acid ribosomal protein P1 and P2, heat shock protein (HSP) 70, nuclear transport factor 2 and glutathione-S-transferase (GST), have not only been identified in A. alternata but also in the closely related mold Cladosporium herbarum[48,49,50,51,52,53].
Most of the A. alternata allergens cloned so far are minor allergens except for Alt a 1, which is recognized by up to 98% of A. alternata-sensitized patients . Alt a 1 can be found as a predominant component in mycelial and culture filtrate extracts [55, 56]. A 20-mer peptide of Alt a 1 located at the N-terminal end showed weak binding of patients’ IgE antibodies and induced antibody synthesis in Balb/c mice indicating that this peptide harbors a linear B-cell and a T-cell epitope .
Two clinical studies using recombinant allergens of A. alternata have been performed. Unger et al.  tested seven A. alternata-allergic patients with Alt a 1 and Alt a 6 (enolase), which is recognized by 15–22% of A. alternata-allergic patients [52, 54]. In this study, all seven A. alternata-allergic patients reacted to the two recombinant allergens whereas commercially available fungal extracts partially failed to correctly diagnose the patients’ allergy. Asturias et al.  tested 42 A. alternata-allergic patients with natural and recombinant Alt a 1 (rAlt a 1), rAlt a 2 and rAlt a 6. Although the prevalence of Alt a 2 was previously determined to be 61% , none of the 42 patients reacted with rAlt a 2, but 41 of the 42 patients specifically reacted with rAlt a 6 (enolase) and rAlt a 1. Thus, the combination of Alt a 1 and Alt a 6 (maybe supplemented with one or two additional allergens) is a promising, molecule-based approach for the diagnosis and therapy of A. alternata allergy.
Alt a 1, the major allergen of A. alternata, was analyzed in respect to its B-cell epitopes. Kurup et al.  synthesized overlapping decapeptides (12 amino acids) spanning the entire Alt a 1 protein sequence and tested these peptides for their IgE reactivity with patient sera. They identified four linear IgE epitopes whereas two of them (K41-P50 and Y54-K63) showed strong IgE reactivity in all 4 A. alternata-sensitized patients tested.
Airborne spores of C. herbarum are prominent causes of fungal allergy and can be found indoors as well as outdoors.
In a study by Tariq et al. , 2.9% of 981 4-year-old children reacted to C. herbarum. In their study, C. herbarum together with A. alternata were the third most common causes of sensitization after house dust mite and grass pollen. Mari et al.  tested 4,962 patients having respiratory symptoms. The overall incidence of C. herbarum sensitization was 13%, but within the group of patients sensitized to more than two fungal sources,the prevalence of C. herbarum sensitization reached 84%. In other words, monosensitization to C. herbarum is rather seldom within mold-allergic patients.
So far, 14 allergens have been identified from C. herbarum, whereas seven of them have been cloned as recombinant proteins (table 2). Except for one, all of these allergens are minor allergens with a prevalence of about 20%. The only major allergen, Cla h 8, an NADP-dependent mannitol dehydrogenase, is recognized by 57% of the C. herbarum-allergic patients and represents a predominant component of the crude extract [49, 61, 62].
For some of the allergens (e.g. enolase and serine proteases), extensive cross-reactivity was demonstrated (see also Cross-Reactivity and Auto-Reactivity), making these proteins fungal pan-allergens [51, 52, 63].
IgE epitopes of C. herbarum enolase have been tested by a PCR-based approach. Ten different peptides spanning the entire protein sequence were tested for their IgE reactivity. Six peptides showed specific IgE reactivity in all patients tested (n = 10), whereas the smallest of them, with a length of 69 amino acids, corresponded to the overlapping region of the five other IgE-reactive peptides [52, 64].
The saprophytic genus Aspergillus includes 132 different species. It is distributed ubiquitously in our natural environment and represents a dominant indoor pathogen [65,66,67]. Aspergillus grows outdoors on decaying vegetation or indoors (e.g. in air conditioning systems) and has the ability to release large quantities of small conidiospores of 2–3 µm. In case of inhalation, they either reach terminal airways or are deposited in large clusters in the upper respiratory tract [14, 65, 68, 69]. Human disorders caused by Aspergillus range from colonization of the respiratory tract, hypersensitivity pneumonitis (extrinsic allergic alveolitis), allergic rhinitis, sinusitis and asthma, to life-threatening systemic invasive aspergillosis and ABPA [66, 68]. Very often aspergillosis is favored by an impaired immune status of the patient either caused by immunosuppressive treatment after transplantation surgery, HIV infection, certain leukemias or hospitalization under intensive care.
The biological characteristics of Aspergillus are its small spore size, its thermo-tolerance allowing growth at human body temperature, its resistance to oxidative killing and its ability to produce small metabolites and enzymes with proteolytic or even immunosuppressive activity [70,71,72].
Since A. fumigatus is implicated in about 80% of Aspergillus-related infections, a large number of allergens were cloned from cDNA and phage display libraries, and characterized and purified as recombinant proteins [70,73,74,75]. The spectrum of the more than 40 IgE-binding components of A. fumigatus that account for the complex, variable and heterogeneous pattern obtained in Western blot experiments includes for example acid ribosomal proteins, enzymes such as proteases, toxins, HSPs as well as several unique proteins exhibiting no significant sequence homologies to structures already deposited in the databases [69, 76]. At molecular level, all these molecules differ in their allergenicity and can be subdivided into two separate categories, namely secreted and cytoplasmic proteins.
Among the most important A. fumigatus allergens identified through molecular approaches is Asp f 1, a non-glycosylated 18-kDa major allergen originally detected in the urine of patients suffering from invasive aspergillosis. It is related to ribotoxins, which are known to inhibit protein translation by cleaving a conserved region of the 28S acid ribosomal RNA . Asp f 1, which was considered to be a kind of virulence factor promoting colonization as well as infection of human tissue, seems to be abundantly secreted after spore germination and during early phases of fungal growth [71, 78]. Although it is recognized by 85% of ABPA patients as well as A. fumigatus SPT-positive asthmatics, its effectiveness in diagnosis and therapy is still controversial because of its high toxicity [68, 69, 79]. Asp f 1 is one of the A. fumigatus allergens which have been analyzed regarding B- and T-cell epitopes. Kurup et al.  synthesized 13 linear decapeptides spanning the whole Asp f 1 molecule and tested them for their IgE reactivity and their potency to stimulate peripheral blood mononuclear cells from ABPA patients. They revealed several peptides harboring B- and T-cell epitopes, whereas the C-terminal region (aa 115–149) was shown to be involved in humoral as well as in cell-mediated immunoresponses in ABPA. Most of the Asp f 1-specific T-cell clones reacted with the peptides aa 46–65 and aa 106–125 restricted by HLA-DR2 and HLA-DR5 alleles .
Banerjee et al. [81 ]performed two studies on the B-cell epitopes of Asp f 2, identifying nine epitopes located in hydrophilic regions , with a putative major B-cell epitope at the N-terminus . T-cell clones were generated from ABPA patients using synthetic peptides from Asp f 2, identifying aa 54–74 as a major T-cell epitope .
The 19-kDa Asp f 3, which shares common IgE-binding epitopes with the peroxisomal membrane proteins A and B from Candida boidinii,can be regarded as the second major allergen of this fungus (94% IgE reactivity), with clinical relevance being already demonstrated in vivo by the provocation of mediator release [67, 73, 84, 85]. B-cell epitopes were analyzed using synthetic peptides and constructing Asp f 3 mutants. Ramachandran et al.  identified seven linear IgE-binding regions spanning the entire protein sequence. They identified 12 amino acids at the N-terminus and 8 amino acids at the C-terminus to be critical for IgE binding.
In case of Asp f 4, three cysteine deletion mutants were generated by selectively deleting cysteine residues. These mutants reacted differently with the IgE antibodies from ABPA patients. The authors concluded that the N-terminal IgE-epitope regions of the protein are crucial for the maintenance of the proper three-dimensional structure whereas the C-terminal cysteines play a significant supporting role in IgE binding .
Asp f 6, an MnSOD, represents a phylogenetically highly conserved protein belonging to the metalloenzyme superfamily, which is required for the conversion of superoxide radicals to hydrogen peroxide and oxygen . Since Asp f 2, Asp f 4, whose biological function still is unresolved, and the MnSOD Asp f 6 are strictly intracellular proteins and thus very unlikely to be available as aeroallergens under normal conditions, sensitization against these two marker molecules seems to be sufficient to allow a precise diagnosis of ABPA [70, 89, 90]. ABPA is the result of fungal proliferation in the respiratory tract, exposing especially atopic asthmatics and patients suffering from cystic fibrosis to non-secreted A. fumigatus allergens due to cellular defense mechanisms and fungal damage .
A. fumigatus acid ribosomal protein P2, Asp f 8, shows a high degree of conservation among eukaryotic organisms and is characterized by the presence of cross-reactive epitopes shared with the homologous allergens from C. herbarum and A. alternata[48, 91].
Asp f 12, a HSP90 protein, may play a major role during stress response and possesses considerable homology to the HSP90 molecules from C. albicans, S. cerevisiae, Trypanosoma, housefly, mouse and homo sapiens. Asp f 12 is also thought to play a role in ABPA and other Aspergillus-induced diseases .
Furthermore, alkaline as well as vacuolar serine proteases have been identified to be major allergens in case of A. fumigatus (Asp f 13 and Asp f 18), A. flavus (Asp fl 13 and Asp fl 18) and A. oryzae (Asp o 13) sharing IgE and IgG epitopes with each other as well as with fungal serine proteases from Penicillium spp. (Pen b 13, Pen c 13, Pen n 13, Pen n 18 and Pen o 18) [68, 93, 94]. In order to analyze the B-cell epitopes from Asp f 13, the protein was chemically and enzymatically cleaved and subsequently the N-terminal sequences were determined. At the end, 3 of 13 linear epitopes located at the C-terminus were proven to be immunodominant .
Another important A. fumigatus allergen is enolase (Asp f 22), a protein of 47 kDa, whose cross-reactivity with Pen c 22 (Penicillium citrinum), Alt a 6 (A. alternata) and Cla h 6 (C. herbarum) has been proven by inhibition immunoblotting [52, 96].
Recently, Bowyer and Denning  compared previously published A. fumigatus allergen sequences with A. fumigatus genomic sequences and revealed that Asp f 15 is identical to Asp f 13. Additionally, they observed partial homology between Asp f 16 and Asp f 9, whereas the Asp f 16 sequence, in contrast to the Asp f 9 sequence, could not be localized on two different A. fumigatus genomic sequences. Assuming either sequencing errors or the existence of an isoform, the authors concluded that the Asp f 9 sequence is more reliable and that Asp f 16 also should be termed Asp f 9. In case of the Asp f 56-kDa allergen, the authors could not find any corresponding genomic sequence. Since these new results have not been included into the WHO allergen list so far, the respective allergens were kept in the list of fungal allergens (table 2) but were parenthesized.
Additionally, A. oryzae α-amylase (Asp o 21) and A. niger β-xylosidase (Asp n 14), which are used as baking additives in the food industry as well as in the starch industry, show allergenic activity [65, 67].
Recombinant Asp f 1, rAsp f 4, rAsp f 6 (MnSOD) and rAsp f 8 (acid ribosomal protein P2), have been tested in several clinical studies [84, 85, 88, 91] involving patients suffering from asthma, ABPA and AD. In these studies, the diagnostic specificity was better in case of recombinant A. fumigatus allergens, and additionally no adverse reactions have been reported.
More than 150 Penicillium species exist, some of which have been described to be common indoor molds. Wei et al.  analyzed 88 homes in the Taipei area in order to isolate and identify the indoor Penicillium species. Their results showed that P. citrinum is the most common Penicillium species in this area. Muilenberg et al.  have reported that P. citrinum, P. oxalicum and Penicillium chrysogenum (former P.notatum) are the five most frequently encountered species of Penicillium in Topeka (Kans., USA). Penicillium can cause atopic asthma in sensitive persons after inhalation of their spores . In Taiwan, 22% of the asthmatic children showed a positive reaction in intracutaneous skin tests for Penicillium species . Shen et al.  showed that IgE antibodies against components of P. citrinum, P. notatum, P. oxalicum and P. brevicompactum could be detected in the sera of 16–24% of asthmatic patients. In 100 patients, P. chrysogenum had the highest positive intradermal skin test reactivity (68%). Therefore, P. chrysogenum is the most frequent Penicillium species used for the clinical diagnosis of fungal allergy.
Results from Shen et al.  showed that 80–93% of asthmatics displayed IgE reactivity to the 32- to 34-kDa serine proteases from P. citrinum, P. chrysogenum, P. oxalicum, P. brevicompactum, A. fumigatus, A. flavus, A. oryzae and A. niger, suggesting a role as major allergens. Alkaline and vacuolar serine proteases from Aspergillus and Penicillium were termed group 13 and group 18 allergens, respectively, by the World Health Organization-International Union of Immunological Societies Allergen Nomenclature Subcommittee , whereby there also exist homologous and partially cross-reactive alkaline and serine proteases in other fungal species (table 3; see also Cross-Reactivity and Auto-Reactivity). Serine proteases are expressed as large precursor molecules which are posttranslationally cleaved forming the mature enzymes. Besides N-terminal cleavage of a pre-pro-sequence, which has been described for all serine proteases during maturation [94,102,103,104], Pen c 18 and Pen o 18 also undergo C-terminal processing .
The alkaline serine protease Pen ch 13 was analyzed for linear IgE epitopes. Eleven peptide fragments spanning the whole molecule were generated and tested for their IgE reactivity in dot blot immunoassays. Determination of the IgE reactivity  revealed that peptide f-2n (aa 31–61) showed the highest frequency (77.1%, n = 35). Three further peptides were IgE reactive with incidences ranging from 31 to 51%. The B-cell epitope analysis was refined by narrowing down peptide f-2n and site-directed mutagenesis of Pen ch 13. Finally, one major linear B-cell epitope was identified to be located within aa 48–55.
In case of Pen ch 18, a dominant linear IgE epitope was mapped within aa 73–95 of the N-terminally processed allergen . A similar result was observed by Yu et al.  who located nine different IgE-binding epitopes distributed throughout the whole protein. One peptide, peptide C12 (V44-W62), was also located at the N-terminal end and was recognized by 75% (n = 8) of the patients tested.
Besides the highly cross-reactive serine proteases, several other Penicillium allergens have been identified. In case of P. citrinum, six allergens have been identified. One of them is Pen c 3, a peroxisomal membrane protein. Thirteen out of 28 (46.4%) sera of Penicillium-sensitized asthmatic patients demonstrated IgE binding to Pen c 3. Immunoblot inhibition experiments showed cross-reactivity between Pen c 3 and Asp f 3, which share 82.6% sequence identity .
Another P. citrinum-allergen was described to be HSP70. Members of the 70-kDa heat shock gene family are highly conserved across a wide range of organisms. They assist the proper folding of polypeptides, inhibit protein aggregation and target misfolded proteins for degradation. The new allergen was designated Pen c 19, and 14 out of 34 (41%) allergic patients showed IgE binding to the recombinant and natural allergen .
A 47-kDa IgE-reactive component was shown to be an enolase (Pen c 22) being cross-reactive with enolases from A. fumigatus and A. alternata. Seven out of 23 (30.4%) sera of Penicillium-sensitized asthmatic patients reacted with a 47-kDa P. citrinum protein from the extract and the recombinant Pen c 22, respectively [96, 109].
Pen c 24, elongation factor 1β (EF-1β), shows a sequence identity of 53% with its yeast (S. cerevisiae) homolog . The N-terminal (aa 1–118) half of the protein was recognized by 2 out of 7 Pen c 24-reactive patient sera, whereas 5 out of 7 sera reacted to the C-terminal half (aa 119–228) , indicating that on both halves B-cell epitopes are present.
An acid ribosomal protein, P1, was characterized to be an allergen of P. brevicompactum by Sevinc et al. . It was designated Pen b 26, and only sera of individuals who were sensitized to this mold reacted with the protein. It is a polypeptide of 11 kDa, rich in acidic residues (> 20%) and its isoelectric point is 3.87.
Cand a enolase, for example, was isolated and analyzed for its B-cell epitopes by testing six proteolytic fragments for their IgE reactivity . Ito et al. [113 ]identified a C-terminal fragment (F-171-I-399), which reacted to 90% IgE antibodies examined (n = 10). A similar result was obtained by Eroles et al. , who also demonstrated the high immunogenicity of the C-terminus.
Among fungi, the basidiomycota are a very large phylum comprising approximately 20,000 species including puffballs, bracket fungi, toad stools, jelly fungi, plant rusts, smuts and mushrooms like the edible Boletus, Cantharellus and Coprinus. Of the large number of basidiomycete species, about 25 species have been shown to be allergenic . Basidiospores contribute most of all to the airborne fungal spore load ranging from 5 to 30% [8, 65, 116]. They particularly occur outdoors, but can also be found indoors, e.g. on wet decaying wood or as infiltrates from outdoors. In temperate zones, seasonal peaks of basidiospores are observed in spring and autumn . The diameter of basidiospores ranges from 3 to 15 µm enabling them to reach the lower respiratory tract . In contrast to ascomycota, basidiomycota do not have vegetative spore production. Since not only the spores but also the fruiting bodies of Ganoderma, Coprinus and Pleurotus contain allergens, they may induce food allergy in sensitized patients upon consumption of these mushrooms [118, 119]. Hence, basidiomycota as well as ascomycota are known to cause atopic asthma in susceptible persons . The incidence of basidiomycota-caused allergy ranges from 3.5  to 25.4% .
In a study performed in Europe and the USA , a total of 701 adults were tested for their reactivity to eight basidiomycete species. The majority (70%) of the individuals tested were classified to be atopic. Out of these 701 persons, 25.4% reacted to at least one basidiomycete extract, whereas Psilocybe cubensis elicited most of the positive skin reactions (13.7%) followed by Pleurotus ostreatus (10.6%), Ganoderma meredithae (9.3%) and Coprinus quadrifidus (5.4%). In a study by Helbling et al. , 9.8% of atopic subjects, who were not preselected with regard to mold allergy, were sensitized to at least one basidiomycete species. Within 457 atopic patients, 8.3% reacted to Pleurotus pulmonalis, 6.2% to Coprinus comatus and 5.4% to Boletus edulis. Moreover, they found that only 4% of the basidiomycete-sensitive subjects were exclusively skin test positive to basidiomycete extracts.
Up to now, the knowledge about basidiomycete allergens lags behind the information about ascomycete allergens. One of the reasons is the lack of source material since cultivation of basidiomycetes is much more complicated and in some cases even impossible.
Among basidiomycota-sensitized patients, C. comatus shows a sensitization rate of 58% . In 1999, Cop c 1 was cloned. It harbors two leucine zipper motifs. Its biologic function is unknown, and it represents a minor allergen being recognized by 25% of C. comatus-sensitized patients . In sensitized individuals, Cop c 1 is skin test reactive in the picomolar range, making it a clinically relevant allergen . Six further allergens with an open reading frame between 68 and 342 amino acids were isolated, whereas only in case of Cop c 2 (thioredoxin) any homology to previously isolated proteins was observed .
M.furfur, previously also known as Pityrosporum ovale or Pityrosporum orbiculare, is a member of the normal cutaneous flora, preferentially colonizing the skin of the head-neck-face region as single-cell yeast, normally being non-pathogenic . Nevertheless, this yeast can act as a pathogen causing pityriasis versicolor and seborrheic dermatitis [41, 126].
IgE reactivity to M. furfur, as shown in skin tests and radioallergosorbent tests, has frequently been observed in patients with AD . M. furfur contains several IgE-reactive proteins ranging from 14 to 94 kDa .
Mala f 2 and Mala f 3 are peroxisomal proteins forming homodimers with an apparent molecular weight of 21 and 20 kDa, respectively, under reducing conditions in SDS-PAGE. They have a sequence identity of 51% and exhibit a high sequence similarity with Asp f 3 from A. fumigatus and two peroxisomal membrane proteins from C. boidinii [84, 129]. In a study of Yasueda et al. , 64 of 127 AD patients reacted with M. furfur extract, and 71.9 and 70.3% were IgE reactive to Mala f 2 and Mala f 3, respectively, making these proteins major allergens. Lindborg et al.  published the isolation of Mala f 5, which again has a high sequence identity with Mala f 2 (57%) and Mala f 3 (58%) and is recognized by 48% of M. furfur extract-reactive patients. Additionally, Mala f 6, a putative cyclophilin, was isolated, having an incidence of IgE reactivity of 48% .
Further allergens identified are Mala f 4, a mitochondrial malate dehydrogenase, with 83.3% of patients having elevated serum IgE levels to purified Mala f 4 .
M. sympodialis as well as M. furfur are associated with AD. Several allergens were cloned, including MnSOD (Mala s 11) and HSP88 (Mala s 10) with IgE reactivities of 75 and 69%, respectively [97, 133]. First, Mala s 10 was published to be an HSP70 protein , but Nierman et al.  compared the published allergen sequences with the genomic sequences obtained recently and concluded that this allergen is actually an HSP88 protein.
Skin test reactivity to P. cubensis spore extract is the highest (13.7%) among basidiomycetes in Europe and the USA . More than ten allergens have been identified by SDS-PAGE immunoblots. Psi c 2, the first recombinant basidiomycete allergen (molecular weight: 16 kDa) shows high homology to cyclophilins and is recognized by 82%, representing a major allergen [135, 136].
Rhodotorula mucilaginosa,also known as R. rubra, is one of the most frequently encountered yeast species in our environment. Chang et al.  published the isolation of an enolase (Rho m 1) which shows high sequence identity with other fungal IgE-reactive enolases. Rho m 1 is recognized by 21.4% of R. mucilaginosa-sensitized patients and cross-reacts with several fungal enolases. Rho m 2, a vacuolar serine protease, is the second cloned allergen, which also cross-reacts with other fungal vacuolar serine proteases .
Cross-reactivity can be seen when IgE antibodies originally directed against a given allergen also bind to a structurally related allergen from another allergen source , thus it is the result of shared B-cell epitopes among homologous proteins. A sequence identity of more than 50% between homologous allergens seems to be necessary in order to exhibit cross-reactivity . Cross-reactivity may be analyzed by various techniques, e.g. immunoblots, RAST and ELISA inhibition. Cross-reactivity between two allergens of different molds has to be distinguished from ‘co-sensitization’ of an allergic person to an allergen originating from another allergenic source. Co-sensitization and cross-reactivity may be differentiated by inhibition experiments between two extracts originating from distinct fungal species, where the degree of inhibition is determined. Cross-reactivity has been described for about 20 fungal allergens. Partly, the cross-reactivity observed may be ascribed to the close phylogenetic relationship of some fungal species. O’Neil et al.  performed skin tests with selected ascomycota and basidiomycota species demonstrating an association between P. ostreatus, A. alternata, Fusarium solani and Epicoccum purpurascens, as well as between Calvatia cyathiformis, A. alternata and F. solani. C. quadrifidus was associated with F. solani and P. cubensis with A. fumigatus. Thus, cross-reactivity is widespread within the two phyla and is one explanation for the clinical observation that the majority of mold-allergic patients react with several fungal species in vitro and/or in vivo . Interestingly, very often cross-reactive fungal allergens represent intracellular proteins, whereas some species-specific mold allergens tend to be secreted, as it was shown for Asp f 1  from A. fumigatus and Cop c 1  from C. comatus.
Cross-reactive allergens may be subdivided according to the origin of their cross-reactive partners. In table 3, all cross-reactive fungal allergens are listed, along with the name of the allergen and whether or not the respective cross-reactive allergen can be found within one fungal phylum, all fungal phyla or even non-fungal species. In case of a few allergens, homologous human cross-reactive proteins have also been identified, which may give rise to auto-reactivity. The allergens showing only cross-reactivity within one fungal phylum are Alt a 1, flavodoxin (YCP4-homolog), mannitol dehydrogenase, nuclear transport factor 2 and the acid ribosomal protein P1. Cross-reactivity between fungal phyla in general has been obtained in case of peroxisomal proteins and vacuolar serine proteases. More than half of the cross-reactive fungal allergens (aldehyde dehydrogenase, alkaline serine protease, serine protease, enolase, GST and HSP70) have got homologous IgE-reactive proteins in non-fungal species. In four of them (thioredoxin, cyclophilin, MnSOD and ribosomal protein P2), cross-reactivity with the human homolog has been observed. Taken together, it is obvious that within the last years the picture has changed in a way that meanwhile more than half of the cross-reactive fungal allergens show cross-reactivity to non-fungal species, raising the importance of fungal allergens in general.
Recently, several fungal species were tested for Alt a 1 homologues using a rabbit-anti-rAlt a 1 serum . The authors could show that cross-reactive proteins were detectable in Stemphylium botryosum, Ulocladium botrytis, Curvularia lunata and Alternaria tenuissima, but not in C. herbarum, P. chrysogenum and A. fumigatus.
Vacuolar Serine Protease. Vacuolar serine proteases have been isolated from Aspergillus, Cladosporium, Penicillium, Rhodotorula and Trichophyton. Lin et al.  generated monoclonal antibodies against culture medium and/or crude extract from P. citrinum and A. fumigatus. They obtained five monoclonal antibodies directed against serine proteases. Two of them (FUM20 and PCM39) were shown to be cross-reactive with the vacuolar serine proteases from P. notatum, P. oxalicum and A. fumigatus. From our work  we know that these mAbs are also cross-reactive with Cla h 9, the vacuolar serine protease from C. herbarum. Chou et al.  demonstrated cross-reactivity for the native and recombinant vacuolar serine proteases from R. mucilaginosa and P. chrysogenum.
Peroxisomal Membrane Protein. In a cross-inhibition study, Asp f 3 shared common IgE epitopes with Cand b 2, previously called peroxisomal membrane proteins A and B (PMPA and PMPB) .
Enolase. Enolase represents an allergen in many fungal species, e.g. C. herbarum, A. alternata, C. albicans, S. cerevisiae, A. fumigatus, F. solani, C. lunata, R. mucilaginosa, Beauveria bassiana and P. citrinum. Preliminary data also indicate that E. purpurascens  and Stachybotrys chartarum have got IgE-reactive enolases. Cynodon dactylon and Hevea brasiliensis are the non-fungal species where enolase has been described to be an allergen. The enolases of C. herbarum, A. alternata, A. fumigatus and C. albicans were shown to be cross-reactive by inhibition experiments . Wagner et al.  demonstrated cross-reactivity between A. alternata, C. herbarum and Hevea brasiliensis by pre-incubating a serum pool with rHev b 9 and testing this depleted serum with rCla h 6 and rAlt a 6, where there was no IgE-binding detectable.
Glutathione-S-Transferase. The crude extracts of A. alternata, A. fumigatus, C. herbarum, C. lunata and E. purpurascens were proven to have GST-enzymatic activity. Additionally, in all extracts a 26-kDa protein reacted with anti-GST antibodies. Using these anti-GST antibodies in ELISA inhibition experiments revealed inhibition in case of C. herbarum, A. alternata, C. lunata, A. fumigatus and E. purpurascens.
There is evidence that fungal sensitization also contributes to auto-reactivity against self-antigens due to shared epitopes between fungal and human proteins. The underlying mechanism seems to be molecular mimicry perpetuating severe chronic allergic diseases.
Cross-reactivity between fungal and human proteins has been demonstrated for MnSOD [148, 149], cyclophilin , acid ribosomal protein P2  and thioredoxin . Based on our own research on C. herbarum and A. alternata allergens, we could show that intracellular fungal proteins are presented to the immune system. Intracellular human proteins are normally not presented to the immune system. However, in case of chronic inflammation, tissue may be damaged and as a consequence these proteins may be accessible for the immune system. Thus human proteins like MnSOD or acid ribosomal protein 2 may sustain allergic symptoms. In a recent study on the pathogenesis of AD, 36% of the patients exhibiting M. sympodialis colonization of the skin had specific IgE antibodies against human MnSOD . These patients were skin test positive to M. sympodialis extract, to human recombinant MnSOD and to structurally related MnSODs. In an atopy patch test with patients suffering from severe atopic eczema, the application of human recombinant MnSOD on healthy skin elicited an eczematous reaction . The release of intracellular self-antigens as a consequence of inflammation processes causing tissue damage is also proposed to be involved in the pathogenesis of ABPA .
Asp f 8, the acid ribosomal protein P2 from A. fumigatus, cross-reacts with its human homologue P2. In skin tests, a humoral autoimmune response to the human P2 protein was seen in ABPA patients and patients with severe AD .
For decades, the diagnosis of mold allergy has based on the patient’s history, and on in vivo (e.g. SPT, intradermal test or inhalation challenge) and in vitro tests (e.g. RAST, ELISA and Western blot). However, the accuracy and reliability of in vivo and in vitro assays is very highly dependent on the quality of the fungal extracts used. Unfortunately, the correlation of the results obtained with skin tests and serological tests is very poor. A direct comparison between in vitro and in vivo results is hampered by the fact that extracts immobilized on testing devices, e.g. ImmunoCAPs, are not available as SPT solution and vice versa.
The quality of crude extracts for diagnosis and therapy is very unsatisfactory in case of fungal extracts. Currently, the quality of mold extracts varies dramatically between commercial suppliers in Europe and the USA since no standardized extracts are available [46, 58, 153]. The reasons for the insufficient quality are manifold. On the one hand, crude extracts from ascomycota as well as basidiomycota were shown to vary considerably in their protein composition [154, 155]. These problems are caused by strain variabilities  and batch-to-batch variations [10, 74]. Additionally, mold extracts may be produced from mycelial cells and/or spores, which may vary in their protein pattern [157, 158]. On the other hand, growth conditions, protein extraction methods and storage conditions are critical with respect to the quantity and even existence of individual allergens [61, 65, 157]. Finally, degradation of the extracted proteins may occur, too . In case of A. alternata, different allergens had different optimal extraction times, whereas the composition of the extraction buffer did not significantly affect the quantity of allergens extracted (with the exception that a low pH which resulted in a low protein yield). The diagnosis of mold allergy is also hampered by the fact that patients might not be aware of the mostly perennial fungal exposure, thus molds may not be taken into account for medical history. Moreover, the panel of allergy-causing molds exceeds by far the number of extracts that reasonably can be used in routine assessments .
To some extent, the problems with fungal extracts may be overcome by the use of recombinant allergens. The major advantages of recombinant proteins over crude fungal extracts are threefold. Firstly, the protein preparations are reproducible and can be standardized for biochemical and immunological tests, e.g. mass spectrometry, circular dichroism, inhibition ELISAs, determination of T-cell reactivity and histamine release assays, and thus will give a batch-to-batch consistency. Secondly, the production of large quantities of pure proteins is possible. Thirdly, using recombinant allergens, it is possible to differentiate among co-exposure, co-sensitization and cross-reactivity. This differentiation is important since primary sensitizing molds have to be known for a successful immunotherapy. Although recombinant allergens have got major advantages, they also have some properties which have to be taken into account for their expression. A few allergens undergo secondary modifications such as glycosylation, phosphorylation, and N- and/or C-terminal processing. Although these modifications may not directly be involved in IgE binding, they nevertheless may have a large impact on the three-dimensional structure and thus on the formation of IgE epitopes of a given protein. Therefore, the choice of the expression system is very important. Routinely, bacterial systems such as Escherichia coli are employed, but since proteins may not be folded properly and eukaryotic posttranslational modifications are not accomplished, alternative eukaryotic systems like Pichia pastoris, S. cerevisiae, Yarrowia lipolytica, Baculovirus and tobacco plant may be used [162, 163]. The P. pastoris system, for example, has been used for the expression of Alt a 1, the major allergen of A. alternata.
In order to use a high throughput test, an allergogram may be generated using a microarray format enabling a large number of allergens to be tested in duplicate or triplicate with a small amount of patient sera, in order to receive a profile of the patient’s IgE reactivity pattern .
Since the total number of relevant IgE-reactive allergens in molds is mostly higher than in pollens or foodstuff, a panel of recombinant allergens may be necessary in order to cover the patients’ allergen profile. Major allergens of all fungal phyla like Alt a 1 [48, 171], Cla h 8 , Asp f 1 , Pen n 18 , Mala f 6 , Mala s 11  and Psi c 2  have been described. These major allergens combined with minor allergens are promising candidate molecules for molecular-based, patient-tailored immunotherapy.
In the last years, the first diagnostic studies have compared recombinant fungal allergens and crude mold extracts with respect to their negative and positive predictability of mold sensitization. In case of A. alternata two clinical studies were performed [54, 58] in which two allergens (Alt a 6 and Alt a 1) were promising candidate molecules. For A. fumigatus, a large number of allergens have been published. Since A. fumigatus is particularly known for its broad spectrum of human disorders, some groups aimed to find a link between a given disease and the patients’ reactivity pattern to individual recombinant allergens. Hemmann et al.  and Kurup et al.  showed that individual recombinant allergens can be used to discriminate between ABPA (Asp f 2, Asp f 4 and Asp f 6) and fungal allergy (Asp f 1 and Asp f 3).
Specific immunotherapy is defined as the repeated administration of increasing doses of an allergen extract. For successful treatment, effective therapeutic doses are required, which often cannot be reached, especially in the case of mold allergy, since side effects due to a large number of non-allergenic components may occur. Several drawbacks have been ascribed to the use of crude protein extract. Since protein extracts contain a vast number of allergenic and non-allergenic components, in the course of immunotherapy a patient might develop IgE antibodies against additional components present in crude extracts, as was shown in case of specific immunotherapy of grass- and birch pollen-allergic patients using crude extracts [172, 173].
Immunotherapy with fungal extracts is possible, but in most countries not recommended because of problems with the standardization of extracts (see also Diagnosis of Fungal Allergy)  and the frequent occurrence of side effects . Additionally, the use of fungal extracts for immunotherapy is hampered by the vast number of fungal species and the lack of knowledge on the degree of exposure to many molds. In the last years, only very few studies reporting a moderate reduction in symptoms have been conducted [25, 175, 176]. In a high-dose (maximal dose of 100,000 biological units), placebo-controlled, double-blind study , 81% of the C. herbarum-allergic patients hyposensitized with C. herbarum extract improved their clinical symptoms, whereas 19% showed a deterioration in their symptoms. In a 3-year clinical study including 79 children with asthma and rhinitis showing Alternaria sensitization, Cantani et al.  reported a successful immunotherapy (doses were higher than 80,000 protein nitrogen units) in 80% of their children.
Using a defined panel of allergenic molecules instead of crude extracts, a patient-tailored immunotherapy may be a future aim.
Taken together, a large number of fungal allergens have been isolated and characterized in the last years. Some of them have already been tested in clinical trials, demonstrating their benefit in the diagnosis of mold allergy [54, 58] and other fungal diseases such as ABPA [89, 90]. It has been shown that the specificity of recombinant allergens in serology and skin tests is clearly superior to the specificity obtained with commercial extracts .
Nevertheless, there is still a long way to go until immunotherapy of mold allergy will be safe and successful.
This work was supported by project S8812-MED given to B. Simon-Nobbe and M. Breitenbach by the Austrian Science Fund (FWF).
Correspondence to: Dr. Birgit Simon-Nobbe
Department of Cell Biology, Division Genetics
University of Salzburg, Hellbrunnerstrasse 34
AT–5020 Salzburg (Austria)
Tel. +43 662 8044 5791, Fax +43 662 8044 144, E-Mail firstname.lastname@example.org
Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher or, in the case of photocopying, direct payment of a specified fee to the Copyright Clearance Center.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.