L’avocat

The avocado (Av) is a fruit that belongs to the Lauraceae family. We report 17 patients with immediate hypersensitivity to avocado. Clinical manifestations in relation to avocado ingestion were as follows: systemic anaphylaxis in seven patients, angioedema/urticaria in six, vomiting in two, bronchial asthma in one, and rhinoconjunctivitis in one. Skin prick test (SPT) with fresh avocado was positive in all patients with the Strong avocado variety (SAv) and in 14 patients with the Hass avocado variety (HAv). Our patient-associated sensitizations were as follows: 10 to latex, eight to chestnut, eight to banana, four to kiwi, and four to walnut. Avocado-sensitized patients with latex allergy were typically middle-aged women, professionally exposed to latex, who also exhibited frequent associated sensitizations to chestnut, banana, and other fruits. Specific IgE against avocado was demonstrated in 11 of our patients, by both commercial CAP and RAST with avocado extract coupled to nitrocellulose disks. Despite its lower protein content, SAv seems to be more allergenic than HAv, both in vivo and in vitro. On incubating a pool of sera from our patients with avocado, latex, chestnut, and banana extracts, a progressive RAST inhibition was obtained, with SAv- and chestnut-marked disks. This suggests the existence of common antigenic determinants among these allergens.

BACKGROUND: Latex-fruit cross-sensitization has been fully demonstrated. However, the antigens responsible for this "latex-fruit syndrome" have not been identified. We have recently shown that class I chitinases are relevant chestnut and avocado allergens. OBJECTIVE: We sought to evaluate the in vivo and in vitro reactions of purified chestnut and avocado chitinases in relation to the latex-fruit syndrome. METHODS: From a latex-allergic population, eighteen patients allergic to chestnut, avocado, or both were selected. Skin prick tests (SPTs) were performed with crude chestnut and avocado extracts, chitinase-enriched preparations, and purified class I and II chitinases from both fruits. CAP-inhibition assays with the crude extracts and purified proteins were carried out. Immunodetection with sera from patients with latex-fruit allergy and immunoblot inhibition tests with a latex extract were also performed. Eighteen subjects paired with our patients and 15 patients allergic to latex but not food were used as control groups. RESULTS: The chestnut class I chitinase elicited positive SPT responses in 13 of 18 patients with latex-fruit allergy (72%), and the avocado class I chitinase elicited positive responses in 12 of 18 (67%) similarly allergic patients. By contrast, class II enzymes without a hevein-like domain did not show SPT responses in the same patient group. Each isolated class I chitinase reached inhibition values higher than 85% in CAP inhibition assays against the corresponding food extract in solid phase. Immunodetection of the crude extracts and the purified class I chitinases revealed a single 32-kd band for both chestnut and avocado. Preincubation with a natural latex extract fully inhibited the IgE binding to the crude extracts, as well as to the purified chestnut and avocado class I chitinases. CONCLUSION: Chestnut and avocado class I chitinases with an N-terminal hevein-like domain are major allergens that cross-react with latex. Therefore they are probably the panallergens responsible for the latex-fruit syndrome.

BACKGROUND: In the medical literature immunoglobulin (Ig)E-mediated sensitization to avocado is rarely reported. On the other hand, more than 50% of subjects having IgE-mediated natural rubber latex allergy are sensitized to avocado fruit as demonstrated by skin-prick testing and/or specific IgE measurements and about 10-20% report hypersensitivity reactions after ingesting avocado. OBJECTIVE: The underlying pathomechanism of latex-associated avocado allergy is still unknown. The conserved hevein domain of the major latex allergen prohevein (Hev b 6.01) is a ubiquitous chitin-binding protein structure that can be found in several plant proteins and may be responsible for the observed cross-reactivity between latex and avocado fruit. METHODS: Chitin-binding avocado proteins (CBAPs) were isolated by affinity-chromatography and their IgE-binding characteristics were studied by immunoblotting using the sera from 15 avocado-sensitized latex patients. Inhibition experiments using isolated hevein and CBAPs as inhibitor solutions were performed to study the immunological cross-reactivity between both protein species and to assess the role of the CBAPs as mediators in latex-associated avocado allergy. RESULTS: In 80% of avocado-sensitized subjects (n = 15), IgE antibodies directed against a 31-kDa allergen were detected by immunoblotting. This IgE-binding protein was identified by protein sequencing to be a class I endochitinase containing a hevein domain at the N-terminus. Purified native and digested (using simulated gastric fluid) endochitinase were able to completely block all avocado-specific IgE antibodies in six out of seven avocado patients. CONCLUSIONS: Sensitization to endochitinase class I containing a hevein domain is the main underlying pathomechanism in latex-mediated avocado allergy.

BACKGROUND: Individuals with natural rubber latex allergy often have immediate reactions to plant-derived foods and fresh fruits, such as avocado and banana. IgE of these patients has been shown to bind endochitinases containing an N-terminal hevein-like domain (HLD). However, evidence on 31-kd endochitinase-induced reactions in vivo is lacking . OBJECTIVE: We sought to assess the clinical significance of 31-kd endochitinases and isolated HLDs in latex-fruit syndrome . METHODS: The 31-kd endochitinases and corresponding HLDs were purified or produced from avocado, banana, latex, and wheat germ. Skin prick test reactivities against purified proteins were examined in 15 patients with natural rubber latex allergy. The binding efficiency of IgE to purified proteins was studied by using an inhibition ELISA. Experimentally resolved or modeled structures of the proteins were compared to clarify the molecular basis of clinical reactions . RESULTS: Eleven (73%) patients had skin prick test reactions to isolated HLDs of avocado and banana, but only 1 (7%) patient reacted to their corresponding 31-kd endochitinases. HLDs from avocado and banana inhibited binding of IgE to prohevein (Hev b 6.01) in 59% and 38% of patients, respectively, whereas corresponding percentages for 31-kd endochitinases were 17% and 20%, respectively. Isolated HLDs of wheat germ agglutinin and 18-kd wheat germ agglutinin did not significantly inhibit IgE binding to hevein . CONCLUSION: The isolated HLD molecules alone, but not when linked to endochitinases, seem to be responsible for IgE-mediated clinical reactions in latex-fruit syndrome. Careful selection of relevant allergens in their proper molecular form is therefore crucial in forming a reliable diagnosis of latex-fruit syndrome.

Sera with IgE antibodies against grass pollen often contain IgE against vegetable foods. We investigated the role of the ubiquitous protein profilin in this cross-reactivity. Profilin was purified from Lolium perenne grass pollen by means of affinity purification with Sepharose-coupled poly(L-proline). This solid phase was also used as capturing agent for profilin from pollen and food extracts for application in a radioallergosorbent test. It was shown that profilin is an allergen in grass pollen and in a wide range of vegetable foods, like potato and celery. Within a grass-pollen-sensitive population, patients with IgE to vegetable foods have a high incidence of antibodies against profilin. IgE antibodies against grass pollen profilin were shown to be cross-reactive with respect to vegetable foods

BACKGROUND: Several foods associated with the latex-fruit syndrome present relevant allergens of around 30 kd. Neither these components nor any other responsible for the reported cross-reactions have been identified and purified. OBJECTIVE: We sought to isolate and characterize the 30 kd allergens from avocado fruit and chestnut seed, two of the main allergenic foods linked with latex allergy. METHODS: Sera from patients allergic to chestnut and avocado were selected according to clinical symptoms, specific IgE levels, and positive skin prick test responses. Class I and II chitinases were purified by affinity and cation-exchange chromatography and characterized by specific IgE and anti-chitinase immunodetection, immunoblot inhibition assays, enzymatic activity tests, and N-terminal sequencing. RESULTS: Relevant 32 kd allergens were detected by specific IgE immunodetection in both avocado and chestnut crude extracts. The same bands, together with others of 25 kd, were revealed by a monospecific antiserum against class II chitinases. Purification and characterization of the 32 kd allergens from both plant sources allowed their identification as class I chitinases with an N-terminal hevein-domain. The purified allergens fully inhibited IgE binding by the corresponding crude extract when tested in immunoblot inhibition assays. Highly related 25 kd class II chitinases that lack the hevein-like domain were also isolated from the same protein preparations. No IgE-binding capacity was shown by these class II enzymes. CONCLUSION: Class I chitinases are relevant allergens of avocado and chestnut and could be the panallergens responsible for the latex-fruit syndrome. The hevein-like domain seems to be involved in their allergenic reactivity.

BACKGROUND: Latex-fruit cross-sensitization has been fully demonstrated. However, the antigens responsible for this "latex-fruit syndrome" have not been identified. We have recently shown that class I chitinases are relevant chestnut and avocado allergens. OBJECTIVE: We sought to evaluate the in vivo and in vitro reactions of purified chestnut and avocado chitinases in relation to the latex-fruit syndrome. METHODS: From a latex-allergic population, eighteen patients allergic to chestnut, avocado, or both were selected. Skin prick tests (SPTs) were performed with crude chestnut and avocado extracts, chitinase-enriched preparations, and purified class I and II chitinases from both fruits. CAP-inhibition assays with the crude extracts and purified proteins were carried out. Immunodetection with sera from patients with latex-fruit allergy and immunoblot inhibition tests with a latex extract were also performed. Eighteen subjects paired with our patients and 15 patients allergic to latex but not food were used as control groups. RESULTS: The chestnut class I chitinase elicited positive SPT responses in 13 of 18 patients with latex-fruit allergy (72%), and the avocado class I chitinase elicited positive responses in 12 of 18 (67%) similarly allergic patients. By contrast, class II enzymes without a hevein-like domain did not show SPT responses in the same patient group. Each isolated class I chitinase reached inhibition values higher than 85% in CAP inhibition assays against the corresponding food extract in solid phase. Immunodetection of the crude extracts and the purified class I chitinases revealed a single 32-kd band for both chestnut and avocado. Preincubation with a natural latex extract fully inhibited the IgE binding to the crude extracts, as well as to the purified chestnut and avocado class I chitinases. CONCLUSION: Chestnut and avocado class I chitinases with an N-terminal hevein-like domain are major allergens that cross-react with latex. Therefore they are probably the panallergens responsible for the latex-fruit syndrome.

Family 28 belongs to the largest families of glycoside hydrolases. It covers several enzyme specificities of bacterial, fungal, plant and insect origins. This study deals with all available amino acid sequences of family 28 members. First, it focuses on the detailed analysis of 115 sequences of polygalacturonases yielding their evolutionary tree. The large data set allowed modification of some of the existing family 28 sequence characteristics and to draw the sequence features specific for bacterial and fungal exopolygalacturonases discriminating them from the endopolygalacturonases. The evolutionary tree reflects both the taxonomy and specificity so that bacterial, fungal and plant enzymes form their own clusters, the endo- and exo-mode of action being respected, too. The only insect (animal) representative is most related to fungal endopolygalacturonases. The present study brings further: (i) the analysis of available rhamnogalacturonase sequences; (ii) the elucidation of relatedness between the recently added member, the endo-xylogalacturonan hydrolase and the rest of the family; and (iii) revealing the sequence features characteristic of the individual enzyme specificities and the evolutionary relationships within the entire family 28. The disulfides common for the individual enzyme groups were also proposed. With regard to functionally important residues of polygalacturonases, xylogalacturonan hydrolase possesses all of them, while the rhamnogalacturonases, known to lack the histidine residue (His223; Aspergillus niger polygalacturonase II numbering), have a further tyrosine (Tyr291) replaced by a conserved tryptophan. Evolutionarily, the xylogalacturonan hydrolase is most related to fungal exopolygalacturonases and the rhamnogalacturonases form their own cluster on the adjacent branch.

Carbohydrates have been suggested to account for some IgE cross-reactions between various plant, insect, and mollusk extracts, while some IgG antibodies have been successfully raised against plant glycoproteins. A rat monoclonal antibody raised against elderberry abscission tissue (YZ1/2.23) and rabbit polyclonal antiserum against horseradish peroxidase were screened for reactivity in enzyme-linked immunosorbent assay against a range of plant glycoproteins and extracts as well as neoglycoproteins, bee venom phospholipase, and several animal glycoproteins. Of the oligosaccharides tested, Man3XylFucGlcNAc2(MMXF3) derived from horseradish peroxidase was the most potent inhibitor of the reactivity of both YZ1/2.23 and anti-horseradish peroxidase to native horseradish peroxidase glycoprotein. The reactivity of YZ1/2. 23 and anti-horseradish peroxidase against Sophora japonica lectin was most inhibited by a neoglycoconjugate of bromelain glycopeptide cross-linked to bovine serum albumin, while the defucosylated form of this conjugate was inactive as an inhibitor. A wide range of plant extracts was found to react against YZ1/2.23 and anti-horseradish peroxidase, with particularly high reactivities recorded for grass pollen and nut extracts. All these reactivities were inhibitable with the bromelain glycopeptide/bovine serum albumin conjugate. Bee venom phospholipase and whole bee venom reacted weakly with YZ1/2.23 but more strongly with anti-horseradish peroxidase in a manner inhibitable with the bromelain glycopeptide/bovine serum albumin conjugate, while hemocyanin from Helix pomatia reacted poorly with YZ1/2.23 but did react with anti-horseradish peroxidase. It is concluded that the alpha1, 3-fucose residue linked to the chitobiose core of plant glycoproteins is the most important residue in the epitope recognized by the two antibodies studied, but that the polyclonal anti-horseradish peroxidase antiserum also contains antibody populations that recognize the xylose linked to the core mannose of many plant and gastropod N-linked oligosaccharides.

The N-glycans from 27 "plant" foodstuffs, including one from a gymnospermic plant and one from a fungus, were prepared by a new procedure and examined by means of matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS). For several samples, glycan structures were additionally investigated by size-fractionation and reverse-phase high-performance liquid chromatography in conjunction with exoglycosidase digests and finally also (1)H-nuclear magnetic resonance spectroscopy. The glycans found ranged from the typical vacuolar "horseradish peroxidase" type and oligomannose to complex Le(a)-carrying structures. Though the common mushroom exclusively contained N-glycans of the oligomannosidic type, all plant foods contained mixtures of the above-mentioned types. Apple, asparagus, avocado, banana, carrot, celery, hazelnut, kiwi, onion, orange, pear, pignoli, strawberry, and walnut were particularly rich in Le(a)-carrying N-glycans. Although traces of Le(a)-containing structures were also present in almond, pistachio, potato, and tomato, no such glycans could be found in cauliflower. Coconut exhibited almost exclusively N-glycans containing only xylose but no fucose. Oligomannosidic N-glycans dominated in buckwheat and especially in the legume seeds mung bean, pea, peanut, and soybean. Papaya presented a unique set of hybrid type structures partially containing the Le(a) determinant. These results are not only compatible with the hypothesis that the carbohydrate structures are another potential source of immunological cross-reaction between different plant allergens, but they also demonstrate that the Le(a)-type structure is very widespread among plants.

BACKGROUND: Cross-reactive carbohydrate determinants (CCD) are well known interferants in specific IgE assays (sIgE). Glyco-epitopes are not restricted to CCD and extracts used to prepare in vitro tests contain many other glycoproteins able to bind glycan-specific IgE. The overall amounts of IgE-bindable glycan structures in allergen sources are unknown . OBJECTIVE: We aimed at quantifying the influence of N-glycan structures on IgE reactivity to commonly tested allergen sources . METHODS: IgE reactivity to 51 allergen extracts, one purified natural allergen and 10 recombinant allergens was measured on Phadia UniCAP system using 2 sera demonstrating significant levels of glycan-related IgE reactivity. Immobilized bromelain and horseradish peroxidase (HRP) were used to capture N-glycan-specific IgE from these sera. Residual IgE reactivity was measured for 42 allergen sources and 4 recombinant/purified allergens . RESULTS: An obviously excessive number of positive CAP-results were obtained with both sera, especially for plant-based allergen sources. Capture of glycan-specific IgE led to a decrease of serum IgE ractivity, variable among allergen sources and between sera. Among others, peanut results were proven largely interfered by the presence of glycan-specific IgE. Unexpectedly some allergen sources showed a slight influence of glycan-related reactivity, such as cockroach, mosquito, mussel, shrimp and domestic mites . CONCLUSION: In patients sensitized to pollens or to Hymenoptera venoms sIgE results should be interpreted with caution. One cannot substract the result of a glyco-reporter test (bromelain and/or HRP) in order to compute glycan-free slgE results for common allergen sources like peanuts. As long as the demonstration of a significant role for glycan structures in clinical allergic reactions is lacking, a simple pre-treatment able to discard glycan-specific IgE from serum would be useful to improve accuracy of in vitro diagnostic tests.

Avocado, the fruit of the tropical tree Persea americana, is a source of allergens that can elicit diverse IgE-mediated reactions including anaphylaxis in sensitized individuals. We characterized a 32-kDa major avocado allergen, Prs a 1, which is recognized by 15 out of 20 avocado- and/or latex-allergic patients. Natural Prs a 1 was purified, and its N-terminal and two tryptic peptide sequences were determined. We isolated the Prs a 1 encoding cDNA by PCR using degenerate primers and 5'-rapid amplification of cDNA ends. The Prs a 1 cDNA coded for an endochitinase of 326 amino acids with a leader peptide of 25 amino acids. We expressed Prs a 1 in the yeast Pichia pastoris at 50 mg/liter of culture medium. The recombinant Prs a 1 showed endochitinase activity, inhibited growth and branching of Fusarium oxysporum hyphae, and possessed IgE binding capacity. IgE cross-reactivity with latex proteins including a 20-kDa allergen, most likely prohevein, was demonstrated, providing an explanation for the commonly observed cross-sensitization between avocado and latex proteins. Sequence comparison showed that Prs a 1 and prohevein had 70% similarity in their chitin-binding domains. Characterization of chitinases as allergens has implications for engineering transgenic crops with increased levels of chitinases.

Class I chitinases have been identified as the major panallergens in fruits associated with the latex-fruit syndrome, such as avocado, banana, and chestnut. However, other plant foods containing these enzymes have not been related to this syndrome. OBJECTIVE: We sought out class I chitinases in the green bean, a legume that is known to express chitinases but is not associated with latex allergy, and examined whether the content or allergenic activity of chitinases can be modified by physical or chemical treatments. METHODS: IgE-binding proteins in untreated bean samples, as well as in ethylene- and heat-treated samples, were detected by using a pool of sera from patients with latex-fruit allergy. Putative allergens were purified by cation-exchange chromatography and characterized by N-terminal sequencing, enzymatic activity assays, immunodetection with sera and antichitinase antibodies, and immunoblot inhibition tests. Skin prick tests with untreated and heated purified allergens were also carried out. RESULTS: An IgE-binding protein of 32 kd that was also recognized by antichitinase antibodies was detected in green bean extracts. This reactive component was strongly induced by ethylene treatment. The protein, designated PvChI, was identified as a class I chitinase closely related to the major avocado allergen Prs a 1. Immunoblot inhibition assays demonstrated cross-reactivity between both allergens. Purified PvChI induced positive skin prick test responses in 7 of 8 patients with latex-fruit allergy. Heat treatment of both Prs a 1 and PvChI produced a full loss of their allergenic capacities both in vitro and in vivo. No IgE-binding component was detected in the white mature bean in which the main isolated 32-kd protein corresponded to a nonreactive phytohemagglutinin. CONCLUSIONS: Ethylene treatment induces the expression of plant class I chitinases. The allergenic activity of plant class I chitinases seems to be lost by heating. This fact could explain why plant foods containing these putative allergens that are consumed after cooking are not usually associated with the latex-fruit syndrome.

BACKGROUND: In the medical literature immunoglobulin (Ig)E-mediated sensitization to avocado is rarely reported. On the other hand, more than 50% of subjects having IgE-mediated natural rubber latex allergy are sensitized to avocado fruit as demonstrated by skin-prick testing and/or specific IgE measurements and about 10-20% report hypersensitivity reactions after ingesting avocado. OBJECTIVE: The underlying pathomechanism of latex-associated avocado allergy is still unknown. The conserved hevein domain of the major latex allergen prohevein (Hev b 6.01) is a ubiquitous chitin-binding protein structure that can be found in several plant proteins and may be responsible for the observed cross-reactivity between latex and avocado fruit. METHODS: Chitin-binding avocado proteins (CBAPs) were isolated by affinity-chromatography and their IgE-binding characteristics were studied by immunoblotting using the sera from 15 avocado-sensitized latex patients. Inhibition experiments using isolated hevein and CBAPs as inhibitor solutions were performed to study the immunological cross-reactivity between both protein species and to assess the role of the CBAPs as mediators in latex-associated avocado allergy. RESULTS: In 80% of avocado-sensitized subjects (n = 15), IgE antibodies directed against a 31-kDa allergen were detected by immunoblotting. This IgE-binding protein was identified by protein sequencing to be a class I endochitinase containing a hevein domain at the N-terminus. Purified native and digested (using simulated gastric fluid) endochitinase were able to completely block all avocado-specific IgE antibodies in six out of seven avocado patients. CONCLUSIONS: Sensitization to endochitinase class I containing a hevein domain is the main underlying pathomechanism in latex-mediated avocado allergy.

BACKGROUND: Resistance to pepsin digestion has been claimed to be a characteristic of food allergens that can induce severe adverse reactions. Moreover, pepsin treatment is included in protocols to evaluate the potential allergenicity of transgenic foods. Allergenic plant class I chitinases, such as avocado Prs a 1, are the panallergens involved in the latex-fruit syndrome. Previous reports indicated their susceptibility to simulated gastric fluid (SGF) digestion. OBJECTIVE: We sought to evaluate the IgE-binding capacity and the in vivo reactivity of the SGF products of the avocado allergen Prs a 1. METHODS: Patients with a clinical history of latex-fruit allergy syndrome, a positive skin prick test (SPT) response to Prs a 1, and specific IgE to avocado were selected. Untreated and SGF-digested Prs a 1 samples were analyzed by means of IgE and IgG immunoblotting, IgE immunoblotting and ELISA-inhibition assays, and SPTs. Peptides from SGF-digested samples were fractionated by means of HPLC, characterized by N-terminal amino acid sequencing and matrix-assisted laser desorption/ionization analysis, and tested for in vivo reactivity with SPTs. RESULTS: Neither protein staining nor IgE immunoblotting with a pool of sera from allergic patients resulted in the detection of any band after SDS-PAGE separation of an SGF-digested sample of Prs a 1. However, this sample showed a similar inhibitory potency to that of untreated Prs a 1 in both immunoblot- and ELISA-inhibition assays (up to 70% inhibition of the IgE binding to crude avocado extract) and induced positive SPT responses in 5 of 8 allergic patients. Peptides from SGF-digested Prs a 1 were separated by means of HPLC, and 4 of them reached more than 50% inhibition values when using avocado extract as the solid phase in ELISA-inhibition assays. Reactive peptides were located both in the N-terminal hevein-like domain and in the catalytic domain of Prs a 1. Those corresponding to the hevein-like domain (approximately 5100 d) produced positive SPT responses in 5 of 8 allergic patients, whereas 2 peptides located in the catalytic domain (approximately 1400 and 2500 d) were reactive in 2 or 3 of the 8 patients. CONCLUSION: Prs a 1 was extensively degradated when subjected to SGF digestion. However, the resulting peptides, particularly those corresponding to the hevein-like domain, were clearly reactive both in vitro and in vivo.