No abstract

The enzyme-linked immunosorbent assay (ELISA) offers many advantages for the detection of potentially hazardous allergenic food residues that might become adventitious components of other foods during the course of food production and processing. ELISAs detect proteins, and food allergens are proteins. ELISAs are sufficiently sensitive and specific for detection of food allergen residues. ELISAs can also be produced in formats that are compatible with the industrial food processing environment. However, ELISAs also have disadvantages that should be carefully evaluated and widely recognized. Various food-processing operations can have profound effects on the detectability of allergenic food residues. ELISAs detect intact proteins but protein hydrolysates evade detection in some ELISA formats. The residual proteins present in some ingredients derived from commonly allergenic sources may also not be easily detected with ELISAs because of the nature of the protein residues remaining, e.g. lipophilic. Processing operations can dramatically lower the solubility of proteins. In some food formulations, heat processing, in particular, induces chemical modifications that can affect antibody binding to epitopes in the ELISA. The use of naturally incurred standards where allergenic food residues are incorporated into various representative food matrices and then processed in a manner similar to "real-world" food processing can reveal some of the limitations of allergen ELISAs. Methods for the preparation of naturally incurred standards in chocolate, cookie, muffin, ice cream, pasta, frankfurter, and cream of potato soup are provided as examples.

Cows' milk is a commonly allergenic food. Cross-contamination of milk proteins into nondairy, kosher-pareve foods prepared on shared processing equipment can cause severe, life-threatening reactions in milk-allergic individuals. A sandwich-type enzyme-linked immunosorbent assay (ELISA; 96-well plate format) was developed for the detection of undeclared casein in foods. Rabbit anti-casein antibodies were used as the capture reagent. Food samples and standards were ground, extracted in 0.01 M phosphate-buffered saline, clarified by centrifugation, and added to the wells. Goat anti-casein antibodies were employed as the detector antibody, and the amount of antibody bound was determined with a commercial rabbit anti-goat immunoglobulin conjugated to alkaline phosphatase, with subsequent substrate reaction. Antibodies developed were specific to casein, with no cross-reaction observed with 30 foods and food ingredients. Non-milk-containing products such as fruit juices, fruit juice bars, sorbets, and dark and pareve-labeled chocolate were purchased from June 2002 through June 2003. In addition, samples allegedly causing eight milk-allergic consumer complaints were analyzed. The ELISA had a detection limit of less than 0.5 ppm of casein. The casein content in the analyzed foods ranged from less than 0.5 ppm to more than 40,000 ppm casein; undeclared casein residues were found in all of the samples implicated in allergic reactions. The levels of milk contamination in some of the other surveyed products could also be hazardous for milk-allergic consumers. This ELISA method provides a useful quality control tool for the food industry and could also be used as a validation of kosher-pareve status.

Enzyme-linked immunosorbent assay (ELISA) is a commonly used method for the detection of trace amounts of potentially allergenic protein residues in foods. However, food matrices and processing conditions can affect the detection of protein residues. The effects of acidity on the detectability of several allergenic proteins commonly found in salad dressing using ELISAs was investigated. First, recovery experiments were performed on salad dressing formulated with 0 to 1000 ppm mustard flour (mustard). The mean percent recovery for mustard from the salad dressing was only 7.7%+/- 1.6%. When the pH of the salad dressing was adjusted to pH 7 prior to spiking with mustard, recovery improved to 94.1%+/- 7.6%. However, if the pH was adjusted to pH 7 after spiking and extraction, the recovery was only 11.1%+/- 1.7%. When vinegar was spiked with mustard flour at pH 3, 3.5, and 4, detectability of mustard was lowest at pH 3. Basic extraction of mustard proteins from salad dressing did not improve the mustard detection. Acidic salad dressing matrices reduced the detectability of mustard by the mustard ELISA probably because of acid precipitation of mustard proteins that renders them insoluble and nonextractable. Commercial salad dressings containing 100 ppm (mg/kg) of egg, milk, or gluten were analyzed every 2 to 4 d for 90 d using 3 commercially available ELISAs. A decrease in the detection of the egg, milk, and gluten in the salad dressing upon storage was observed. Our study highlighted the importance of evaluating the utility of various ELISAs for specific food matrices and the recovery as a function of product storage.

No abstract