59 / 151
To test the method on second instrument calibrations and some spiked recoveries were performed on an Agilent 6530 q-TOF MS. The quantitation was performed on the
molecular ion within a 50 ppm mass window and 2 MS/MS full scan spectra for confirmation were acquired when that ion is present.
Figure 11 presents the calibrations for egg, hazelnut, milk and peanuts as well as the calibration residuals over the concentration range of 1-1000 ppm. The residuals for all
calibrations were between 85-120% and the linear correlation coefficient was greater than 0.996.
In Table 4 the authors presented the spike recoveries results from the q-TOF for the matrices in the AOAC SMPR 2016.002 at the 50 ppm level. The spike recoveries ranged
from 74% to 112% with an overall average of 89.4 % with a standard deviation of 12.6%. Due to the lower sensitivity of the q-TOF compared to the triple quadrupole, the 5 ppm
spike recovery was not evaluated.
I agree with the authors who suggested that a third platform (SCIEX API 3000) triple quadrupole would also be used to generate a set of data, which will then produce data from 3
instruments and 2 laboratories.
Specificity is one of most important analytical parameter
The specificity of the method was demonstrated by verifying the absence of the peptide markers for each allergen in the food matrices listed in the AOAC SMPR 2016.002.
Furthermore, the authors showed that there were no interferences detected for the MRM transition monitored from the matrices.
Specificity is another important analytical parameter. Peptide selection is based on selectivity and sensitivity. Peptides that are too small will have little selectivity, thus sequences
less than six amino acids are avoided. Larger peptides can offer more uniqueness, but result in reduced MS sensitivity. Too large a peptide and sensitivity decreases due to
multiple charging and costs of synthesis increase, thus peptides larger than 20 amino acids were usually not selected. Also, peptide sequences were selected based on
uniqueness to that protein. Each peptide sequence was searched against the NCBI nr protein database, in order to verify that the sequences were unique, and while, are
common ingredients in food products.
Peptide markers for detection of the four food allergens were selected from the allergenic proteins listed in Table 1. Synthesized peptide markers and stable labelled isotopes
were used as internal standards. Peptides that are representative for these allergen proteins are highlighted in the protein sequence and labelled (m=milk, H=hazelnut, p=peanut
and ew=egg whites and ey= egg yoke).
In Figure 4 the authors provide us with the information that LC/MS/MS MRM chromatograms for all the matrices (bread, cookie, dough, cereal, ice cream, milk chocolate, dark
chocolate, salad dressing, wine and infant formula) and a 3 ppm standard of peanut peptide NAQRPDNR (p2) and milk peptide HQGLPQEVL (m3). Figure 5 shows the
LC/MS/MS MRM chromatogram for all the marker peptides for egg, hazelnut, milk and peanut showing the 3 transitions monitored.
Conclusion
In conclusion, the proteomic approach described here and proposed by the authors uses LC–MS/MS to specifically detect allergens in one analysis. Proteins unique to eggs,
milk, peanut, and hazelnut have been extracted, subjected to trypsin digestion and analysis by liquid chromatography/quadrupole mass spectrometry, in order to find highly
conserved peptides that can be used as markers to detect components in the food.
This multiplexed approach provides the means to test food for hidden allergenic compounds with accuracy and sensitivity to satisfy both inspection and labelling purposes.
Analytical methodology can be implemented by specialist inspectorates. Innovative procedure of separating and quantitative analysis of hidden allergenic compounds can be
used for the correct identification and quantitative analysis of the allergens also by the European Food Safety Agency (EFSA).
6. Based on the supporting
information, what are the
cons/weaknesses of the
method?
Synthesized peptide markers and stable labelled isotopes as internal standards will
greatly enhance the robustness of any method but the cost is prohibitive.
Therefore, the cons/weakness of the method may be the costs (e.g., the costs of
synthesis peptide markers). Although, I think it is inevitable.
Tryptic digestions may result in missed cleavage sites that can result in interfering
peptides. But the authors of the method also confirmed the specificity of each "analyte"
peptide. Peptides were identified with their associated proteins searching NCBI nr and
then cross searches used to eliminate those peptides that appear in other plants or
animals (as with milk and eggs).
According to the authors, the proposed enzymatic digestion process results in samples
that contain some interferences decreasing MDL values. In order to overcome this
imitation additional SPE pre-concentration/clean-up step is required, which makes
whole sample preparation procedure more labour-intense.
A full method performance validation is still required for the all marker peptides. As the
authors declared, they also plan to submit data from analysis on a Sciex LC-QQQ
system in future experiments. I suppose that having done the experiments planned, the
results will be expressed as the reproducibility standard deviation (SDR); or %
reproducibility relative standard deviation (%RSDR).