This report marks 20 years since the online version of the Structural Database of Allergenic Proteins (SDAP) was first established. SDAP's original purpose was to provide a cross referenced website to classify allergens according to their names, structure and function, a need that had been building for over 100 years. Although allergic reactions are described in print as early as the 16th century, or even ancient times (1, 2), the word allergy, and the beginning of treatments for allergy with plant extracts, began in the first years of the 20th century (3). Extracts of ragweed helped many individuals with “hay fever,” but injecting patients with whole pollen extracts could also induce dangerous anaphylaxis. Thus researchers, aided by modern protein chemistry, set out to make safer, simpler, “component resolved” extracts. Molecular studies of the protein components in the extracts that could contribute to reactivity began about 60 years ago. During the standardization of these extracts, it was found that there were several components that bound IgE in sera from hypersensitized individuals, leading to modern tests for categorizing the types of allergens a patient might react to (4). The first isolated allergenic proteins, Amb a 1 and Amb a 2 of ragweed (5, 6) were soon followed by many others from many different sources.

A standardized nomenclature based on the Latin names of their plant, animal, insect or venom source was first published in 1984 (7). The nomenclature they agreed upon was to designate highly purified allergenic proteins by the first three letters of the genus followed by a space, the first letter of the species name, followed by a space, both italicized, followed by a Roman numeral to indicate the order of importance (or isolation) of the protein. For example, the first perennial rye grass allergen, from the plant with the Latin name Lolium perenne was called Lol p I.

The list of proteins in the paper, mostly aeroallergens from dander and pollen, covered less than a journal page. The only “ingested” allergens were parvalbumin from cod (Gad c 1), three egg white proteins (Gal d 1-3) and a surface protein of round worm (Asc s 1).

The allergen field expanded rapidly (some called it a “data explosion”), thanks to innovations in immunology, protein sequencing and the recognition that IgE in serum bound specifically to allergens (Figure 1). Multiple allergens were isolated from peanuts (8) and other sources (9, 10).

The nomenclature was simplified a few years later (11), in that names were no longer italicized, and Arabic, not Roman numbers were used (e.g., Ara h 8 from peanuts instead of VIII). The list of allergens in the 1994 paper stretched over more than four journal pages, with additional recommendations, not just for the sequences of the whole protein, but for peptides from the sequence that were epitopes for IgE from patient sera. But numbering remained a problem. The small table in the 1984 publication shows that Api m I, Ves g I, Pol a I all clearly referred to phospholipases of honey bee, yellow jacket and wasps, respectively, while Api m II, Ves g II, and Pol a II refer to hyaluronidases. But new allergens of different families, such as pectate lyases and from various pathogen related (PR) groupings (12, 13) were named as they were isolated, so the numbers no longer reliably predicted similarity in function or structure. For example, the vicilin allergen of peanut, Ara h 1, corresponded best in structure to Jug r 2 of walnut. Later identified allergens, Ara h 6 and Ara h 7, had similarity to Ara h 2 and other 2S albumins.

The first job of SDAP was thus to provide a cross-referenced list of the sequences and associated information of all the proteins acknowledged to be allergenic by the IUIS (16). This was done by a series of cross referenced MySQL lists, most of which were assembled tediously by human effort (17). Later versions of the database could use automatic identification (14), but many proteins were found to be allergenic or IgE_binding only after their isolation and naming.

The need for bioinformatic tools to identify such potential allergens was brought to the forefront by an attempt to enhance the methionine content of grains, by inserting a newly identified gene from a brazil nut (18, 19). These projects were terminated when allergic responses were found to the “genetically modified” foods. Thus, a sequence FASTA search in SDAP was implemented to rapidly show whether a test protein had significant identity to any known allergen, using a set of rules established by the IUIS (20). The user could then decide whether to proceed with using the protein or drop projects before problems arose.

The information on allergens' structures and epitopes continues to grow at a rapid rate. SDAP was created to help understand the similarities and differences in these proteins. Twenty years after its start, there is now a major push to update its software and list of allergenic proteins and their isomers, to be a tool for researchers, regulatory agencies and patients.

CHS wrote and edited the manuscript with help from WB. SN prepared Figure 1B. All authors contributed to the article and approved the submitted version.

SDAP was originally funded by a Research Development Grant (#2535-01) from the John Sealy Memorial Endowment Fund for Biomedical Research, grants from the US-Food and Drug Administration (FD-U-002249), and the Texas Higher Education Coordinating Board (ATP grant 004952-0036-2003) to WB. Our work with allergenic proteins and SDAP upkeep has been supported by grants from the U.S. Environmental Protection Agency under STAR Research Assistance Agreements (RD-833137, RD-834823-01 to WB, RE-83406601 to CHS), the US Department of Agriculture (ARS 58-6435-9-40 to CHS), the National Institute of Health (R01 AI 064913 to WB, 1R01AI165866-01 to Stephen Dreskin; subcontract to CHS), and funding for this article was provided by the Margaret Maccallum Gage and Tracy Davis Gage Professorship in Biochemistry and Allergies to WB.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.