花粉过敏患者在食用植物性食物(水果、蔬菜和香料)时也经常出现过敏症状。新的信息使我们能够将花粉-食物过敏综合征归因于IgE抗体对保守的植物过敏原的交叉反应，这些过敏原在相关和不相关的植物物种以及不同的植物组织中表达。保守蛋白是指具有重要生物学功能的蛋白质，其序列和/或结构是保守的。引起花粉-食物交叉反应的重要保守过敏原是肌动蛋白结合蛋白profilin，它首先在桦树花粉(Bet v 2)中被描述，而主要的桦树花粉过敏原Bet v 1,与致病相关的植物蛋白高度同源。我们提出的假设是，空气中花粉变应原致敏是导致IgE抗体产生的主要事件，然后IgE抗体能够与同源食物变应原发生交叉反应。这项研究总结了这一假设的证据。对多种过敏原编码的互补DNA进行了表征，并在异种表达体系中以重组过敏原的形式表达。这些过敏原的序列比较显示了与保守蛋白的显著同源性。主要的桦树花粉变应原Bet v1与植物受微生物侵袭时表达的蛋白(病原相关蛋白)具有显著的序列同源性，而另一种桦树花粉变应原Bet v2可被鉴定为profilin。profilin被称为保守的肌动蛋白结合蛋白，也参与了真核生物的磷酸肌醇途径。编码Bet v1和Bet v2的cDNAs以及重组过敏原的可用性使我们能够证明，水果、蔬菜和香料中的桦树花粉同源蛋白与IgE抗体发生交叉反应，导致花粉-食物过敏综合征。利用Bet v1特异性单克隆抗体和血清IgE，用Bet v1 cDNA进行Northern印迹，证实了苹果中存在Bet v1同源变应原。利用重组Bet v1和Bet v2进行IgE抑制实验，可以在许多水果、蔬菜和香料中发现Bet v1和Bet v2同源过敏原的存在。发现榛子过敏是由于与Bet v 1和Bet v 2有IgE抗体交叉反应的原因。虽然重组Bet v 1可以用来阻止IgE结合到相应的榛子过敏原,但食物过敏原无法完全阻止IgE结合到花粉过敏原上,表明可交叉反应的IgE抗体产生主要是由致敏花粉蛋白质引起的。艾草花粉中含有一种60 ~ 65KD的成分，可能与艾草的主要过敏原Art v 1相同，最近被报道与芹菜、苹果、花生、猕猴桃以及单子叶植物和双子叶植物花粉中的同源过敏原发生交叉反应。为了验证植物源性食物过敏主要是由于IgE抗体与花粉中存在的同源过敏原发生交叉反应的假设，我们在实验动物模型中以重组花粉过敏原为免疫原诱导了交叉反应的IgE抗体。在小鼠和恒河猴中，用重组花粉过敏原Bet v1和Bet v2免疫动物，可诱导IgE抗体与存在于水果和蔬菜中的同源过敏原发生交叉反应。
此外，免疫Bet v2的恒河猴对芹菜提取物表现出皮肤反应活性，提示交叉反应IgE抗体的临床意义。植物源性食物过敏是由对同源蛋白-花粉过敏原的初步致敏引起的，这一证据来自以下观察：只有在花粉和植物体细胞组织内表达的过敏原,如Bet v 1和Bet v 2,可以确定为花粉和植物源性食物交叉反应的过敏原。三种主要的草花粉变应原(I、II和V组)在花粉中特异性表达，而不在植物其他组织中表达。虽然95%以上对草花粉过敏的患者对这些过敏原有反应，但到目前为止，还没有一种过敏原可以被确定为植物源性食物中的过敏原蛋白。同样，一些其他过敏原，如Bet v3，一种钙结合桦树花粉过敏原，也只在花粉中表达，还没有被认为是食物过敏原。
相比之下，桦树花粉变应原Bet v1和Bet v2及其同源物可在植物组织细胞中检测到，而Bet v1和Bet v2在花粉-食物交叉反应中占很大比例。Bet v 1和Bet v 2在花粉中的表达是高度上调的,尤其是在花粉成熟期,虽然在植物组织体细胞中这类蛋白质含量很少,但还是认为花粉是花粉-食物过敏患者的主要致敏物。
了解花粉和植物源性食品中存在的类似的IgE表位是导致花粉-食品过敏综合征患者IgE抗体交叉反应的原因，这可能有助于诊断和治疗。用于诊断和治疗食物过敏的提取物制备非常难，因为水果和蔬菜的蛋白质和碳水化合物含量相对较低，而且这些提取物中致敏成分不稳定。事实上，在大量对食物过敏患者身上进行的实验证实，植物源性食物中针对蛋白质的IgE的高比例可以被同源花粉蛋白吸收。因此，考虑用与同源食物过敏原发生交叉反应的重组花粉过敏原用于诊断和治疗似乎是合理的。花粉过敏的成功治疗可能会显著减少相关的食物过敏症状。Bet v 1、Bet v 2或其他同源过敏原存在于某些植物源性食物中，这一知识可能还有助于临床医生将某些花粉和食物过敏的关联性解释为这两种来源中存在交叉反应决定因素。
Type I allergic reactions to plant-derived food: A consequence of primary sensitization to pollen allergens
Patients with pollen allergy also frequently experience allergic symptoms on ingestion of plantderived foods (fruits, vegetables, and spices). 1-3 The association of certain pollen allergies with food intolerance has led to the definition of clinical syndromes such as the celery-mugwort-birch pollen syndrome, the apple-birch pollen syndrome,and similar phenomena. More than 10 years ago, it was suggested that IgE antibodies that cross-react with pollen and food proteins could be responsible for the observed clinical phenomena.
New information allows us to attribute pollen-food allergy syndromes to the cross-reactivity of IgE antibodies to conserved plant allergens that are expressed in related and unrelated plant species and different plant tissues. The term conserved protein refers to proteins that fulfill important biologic functions and therefore are conserved in their sequence and/or structure. Prominent conserved allergens responsible for pollen-food cross-reactivities are the actin-binding protein profilin, first described in birch pollen (Bet v 2), 9"1~ and the major birch pollen allergen, Bet v 1,11 which is highly homologous to pathogenesis-related plant proteins. We have developed the hypothesis that sensitization to airborne pollen allergens represents the primary event leading to the induction of IgE antibodies that are then capable of cross-reacting with homologous food allergens. This study summarizes evidence for this hypothesis.
Complementary DNAs coding for a number of allergens were characterized and expressed as recombinant allergens in heterologous expression systems. Sequence comparison of several of these allergens has revealed significant homology with conserved proteins. The major birch pollen allergen Bet v 1 showed significant sequence homology with proteins (pathogenesis-related proteins) that are expressed in plants on microbial attack, ~1 whereas another birch pollen allergen, Bet v 2, could be identified as profilin. 9 Profilins are known as conserved actin-binding proteins, ~2 which also participate in the phosphoinositide pathway in eukaryotes. 13 The availability of the cDNAs coding for Bet v 1 and Bet v 2 and the recombinant allergens has allowed us to demonstrate that homologous proteins in fruits, vegetables, and spices cross-react with IgE antibodies to the birch proteins, leading to pollen-food allergy syndromes.
The presence of a Bet v 1 homologous allergen in apples was demonstrated by Northern blotting with the Bet v 1 cDNA by using Bet v 1 specific monoclonal antibodies and serum IgE. 14 By using recombinant Bet v 1 and Bet v 2 for IgE inhibition experiments, as well as specific antibodies, the presence of Bet v 1 and Bet v 2 homologous allergens could be demonstrated in numerous fruits, vegetables, and spices. 15"17 Allergy to hazelnuts was found to be due to IgE antibodies that cross-react with Bet v 1 and Bet v 2.18 Although recombinant Bet v 1 could be used to absorb IgE binding to the homologous hazelnut allergen, the food allergens were not able to completely block IgE binding to the pollen allergens, indicating that the cross-reactive IgE antibodies originate primarily from a sensitization to pollen proteins. A 60 to 65 kd component present in mugwort pollen, possibly identical to the major mugwort allergen Art v 1, was recently described to cross-react with homologous allergens present in celery, apple, peanut, kiwi, and pollens from monocot and dicot plants. 19 To test the hypothesis that allergy to plant-derived food is primarily due to IgE antibodies that cross-react with homologous allergens present in pollen, we induced cross-reactive IgE antibodies in experimental animal models by using recombinant pollen allergens as immunogens. In mice, as well as in rhesus monkeys, immunization of animals with recombinant pollen allergens, Bet v 1 and Bet v 2,induced IgE antibodies that cross-reacted with homologous allergens present in fruits and vegetables.
2~ Moreover, Bet v 2-immunized rhesus monkeys displayed skin reactivity in response to celery extracts, indicating the clinical relevance of the cross-reactive IgE antibodies. 22 Further evidence that allergy to plant-derived food originates from a primary sensitization to homologous pollen allergens comes from the observation that only allergens, such as Bet v 1 and Bet v 2, which are expressed in pollen and in somatic plant tissues, 23 could be identified as crossreactive allergens in pollen and plant-derived food. The three major grass pollen allergens (groups I, II, and V) are specifically expressed in pollen and not in other plant tissues. 24, 25 Although more than 95% of patients allergic to grass pollen react to these allergens, so far none of the allergens could be identified as allergenic proteins in plant-derived food. 26 Likewise, several other allergens such as Bet v 3, a calcium-binding birch pollen allergen, z7 are expressed exclusively in pollen and are not yet described as food allergens.
~ In contrast, the birch pollen allergens Bet v 1 and Bet v 2 and their homologues, which are responsible for a great proportion of pollen-food cross-reactivities, can be detected in somatic plant tissues. The fact that the expression of Bet v 1 and Bet v 2 is highly upregulated in pollen, particularly during pollen maturation, 28, 29 whereas only tiny amounts of the proteins could be detected in somatic plant tissues, also argues in favor of the idea that pollen is the primary sensitizing agent in patients with pollen-food allergy. The knowledge that similar IgE epitopes present in pollen and plant-derived foods are responsible for the cross-reactivity of IgE antibodies in patients with pollen-food allergy syndromes might be useful for diagnostic and therapeutic purposes. Extracts for use in the diagnosis and treatment of food allergies are extremely difficult to prepare because of the rather low protein versus carbohydrate content of fruits and vegetables and because of the instability of allergenic components in such extracts. In fact, experiments performed in a large number of patients allergic to food corroborate that a high proportion of IgE directed against proteins in plant-derived foods can be absorbed with homologous pollen proteins.
It therefore appears reasonable to consider recombinant pollen allergens that cross-react with homologous food allergens for use in diagnosis and perhaps in treatment. It might be that successful treatment of the pollen allergy would result in a significant reduction of the associated food allergy symptoms. The knowledge that Bet v 1, Bet v 2, or other homologous allergens are present in certain plant derived foods may additionally help the clinician to interpret the association of certain pollen and food allergies as being due to the presence of crossreactive determinants in both sources.
Rudolf Valenta, MD, and Dietrich Kraft, MD Vienna, Austria