Achieving more Reliable Ingredient lists on Food Labels with LC-MS/MS-based Allergen Screens
For Research Use Only. Not for use in diagnostic procedures.
Allergens are invisible dangers, lurking in foods where least expected—in baked goods, a restaurant dish, or even a glass of wine. An estimated 1-3% of the population may be affected by food allergies, in which the immune system attacks an allergenic protein, a process mediated by the body’s own IgE antibodies. These immunogenic events can lead to anaphylaxis and include severe, life-threatening symptoms, such as swelling in the airways, respiratory distress, loss of consciousness, and even death. Currently the only treatments are self-injected epinephrine or avoidance of foods with allergenic ingredients, and many allergy sufferers must entrust their lives to the integrity of food labels in order to discern what is safe—and not safe—to eat.
The World Health Organization currently recommends that eight ingredients be listed on packaging labels: peanuts, tree nuts, eggs, milk, gluten-containing cereals, shellfish, fish and sulfites. Traditionally, allergenic proteins linked to these foods are found using enzyme-linked immunosorbent assays (ELISA) or polymerase chain reaction (PCR) experiments, although these detection methods are regularly prone to false positives/false negatives or limited to recognizing only a single epitope. Overlooking even one allergy-causing ingredient has an enormous impact on food safety, and undeclared allergens are one of the biggest reasons for food recalls, pushing the food industry to invest in methodologies that are sensitive enough to find trace levels of multiple allergens within a wide-variety of foods. To help avoid costly mistakes, researchers at SCIEX have developed increasingly more versatile and reliable food screening applications, utilizing LC-MS/MS-based approaches that can test for several allergens at once and verify their identities—even in the low parts-per-million (ppm) range.
Creating a highly specific LC-MS/MS workflow for detecting multiple allergens
To distinguish allergens from non-specific, background proteins at such a low level, a highly specific approach is needed. In the LC-MS/MS workflow presented here, intact proteins are extracted from food, digested, and the resulting tryptic peptides are separated using microflow chromatography prior to LC-MS/MS analysis on the QTRAP® system, a hybrid triple quadrupole/ion trap mass spectrometer with the sensitivity and processing speed necessary for multiplexed assays (Figure 1). To detect those peptides unique to allergenic proteins, pre-selected mass transitions are followed using a multiple reaction monitoring (MRM) method. MRM peaks corresponding to an allergenic peptide provide a means to identify and quantitate allergen levels. Moreover, the peptide’s product ions are further fragmented in the ion trap, and this array of secondary fragments provides detailed sequence information for validating the identity of the allergen.
"With mass spectrometry, what we want to do is provide a very specific way of detecting that allergenic protein, or, better yet, use the peptide signature that relates to that protein," explained Ashley Sage, Senior Manager, SCIEX Food Testing business. “Using MS/MS gives you the capability to measure multiple fragments, so you’re actually getting very specific measurements of that particular peptide." In turn, this additional degree of information ensures that non-specific peptides with overlapping MRM transitions are not inadvertently mistaken for allergen-related analytes.
To reduce the possibility of falsely detecting an allergen, researchers layered multiple levels of selectivity into the MRM workflow by tracking several distinctive peptides for each allergen of interest, each of which were followed using three to four MRM transitions. Having multiple points of identification reinforced each peptide’s identity, but determining such a large number of transitions—all of which needed to be unique—required some added method development. Researchers leaned on the MIDAS™ Workflow, where the allergenic protein was digested in silico and then MRM transitions were derived from projected tryptic sequences (Figure 2). Not all predicted peptides were good MRM candidates; those peptides not expected to fragment well or those that had post-translational modifications were excluded. After selecting three to four qualified peptides for each allergen, the predicted MRMs were verified in a survey scan. Once a peak was detected, a full scan MS/MS spectrum was triggered on the product ions, and the resulting fragment fingerprint was uploaded to a database to authenticate and identify the peptide. “So we get both confirmation and a measurement of the peptide in a single acquisition,” Sage said, “which allows us to detect the peptide with great certainty within a complex matrix.”
Screening for egg, milk, wheat and peanut with LC-MS/MS
Using this multiplexed workflow, bread or pasta extracts spiked with the allergens of interest were evaluated for peanut, milk, egg and wheat. The tracking of 55 MRM transitions—corresponding to 19 unique peptides originating from the 4 allergens—was enabled by the Scheduled MRM™ algorithm, which permits a particular MRM to be scanned only during the corresponding peak’s expected elution time, freeing up cycle time and dwell time for improved sensitivity (Figure 3). "A key aspect of this workflow is we can measure multiple allergens in one injection," explained Sage, commenting on how direct detection of the allergen benefits the accuracy of screening foods. "Measurement of multiple peptides helps to give more confidence and fewer false positive results."
Secondary fragments from the MS/MS spectra also added certainty to the detection of allergens in bread and pasta extracts. These fingerprint spectra stemming from further fragmentation of product ions in the ion trap provided detailed sequence information on egg or milk peptides, and these spectra were the same for allergens from either bread or pasta, meaning that signals were specific enough not to be impacted by matrix effects (Figure 4). "So we are getting a high degree of specificity when measuring the allergens, as well as when getting the actual sequence coverage," Sage said regarding the advantages of secondary fragmentation.
Once their identity was confirmed, allergenic peptides could be used with confidence to quantitate the allergen levels within a food sample. Although internal standards were not available, standard curves created from the method of standard additions, with different levels of milk and egg proteins spiked into bread or pasta, proved to be highly reproducible, giving CVs between 3.9 and 9.3% for different levels of milk proteins as well as providing low LLOQs for detection down in the low ppm range (Table 1).
LC-MS/MS applications for food allergen testing stand out in a crowd
Optimization of several aspects of the workflow has greatly improved the accessibility of LC-MS/MS for many food testing labs. For one, the multi-step extraction of food samples was time-consuming and labor-intensive. SCIEX scientists focused on improving efficiency, shrinking tryptic digest assay times and incorporating an online solid phase extraction (SPE) column, for faster prep times. Other improvements included developing a microflow chromatography method that maintained good resolution of peptides while increasing sensitivity and reducing overall separation time compared to normal flow techniques. These advancements have enabled LC-MS/MS to outshine the other techniques currently used for allergen testing.
Using an LC-MS/MS solution overcomes the limitations of traditional screening methods that are typically less efficient, less sensitive, and less dependable. LC-MS/MS approaches allow for direct detection of multiple representative peptides and also supply confirmatory sequence information, delivering a highly reliable way to locate and measure food allergens. “Mass spectrometry does offer the capability to give you improved selectivity above the traditional ELISA and PCR tests and fits in very nicely as a complementary and confirmatory tool to these standard technologies," noted Sage. "We are working closely with customers to develop the next range of assays, to improve workflows, to streamline those processes, and to improve the capabilities of doing multiple species as we move forward."
Figure 1: At the core of the LC-MS/MS approach to allergen detection is the direct detection of tryptic peptides obtained from digests of food extracts that are separated using microflow chromatography and analyzed using a hybrid triple quadrupole/linear ion trap mass spectrometer. Detection of peptides by a multiple reaction monitoring (MRM) transition in Q1 provides an identifying mass. MS/MS fragmentation of product ions collected in the linear ion trap provides a unique fragment fingerprint that delivers peptide sequence information.
Figure 2: The MIDAS Workflow (MRM-initiated detection and sequencing) is used to identify several, unique peptides from each allergenic protein and predict their expected multiple reaction monitoring (MRM) transitions in silico. This information allows for multiple MRM transitions to be used to identify multiple allergens from within a food sample. A peptide corresponding to an MRM transition is characterized further with MS/MS information obtained through ion trap fragmentation of product ions.
Figure 3: This survey scan shows the results of a bread sample spiked with milk and egg proteins (100 ppm) that was scanned for 55 multiple reaction monitoring (MRM) transitions using the Scheduled MRM Workflow.
Figure 4: A survey scan from the MIDAS Workflow reveals several peaks corresponding to multiple reaction monitoring (MRM) transitions of milk and egg allergens spiked into pasta (top panel) or bread (not shown). MS/MS spectra were automatically acquired for each MRM transition detected permitting each peak’s identity to be verified. Shown here are the MS/MS spectra for a peptide from egg (lower left panel) and from milk (lower right panel). The collected MS/MS spectra for egg and milk in the bread sample were identical to the MS/MS library spectra collected for the peptide standards, providing added confirmation of positively detected peptides (data not shown).
Table 1: The concentrations of milk allergens spiked into a bread sample at two different levels (10 and 100 ppm) were determined by integrating the multiple reaction monitoring (MRM) peak corresponding to a milk allergen peptide. Two separate bread samples were extracted and run in triplicate, displaying excellent reproducibility at both concentrations.