In the present work, a method for the quantification of milk
contamination in bakery products by mass spectrometry is
proposed. To validate this method, incurred cookies were spiked
with different amounts of skimmed powdered milk (SRM-1549)
and baked. The method was developed on baked cookies to take
into account any contamination occurring either before, during
or after processing and moreover in order to consider any allergens
modification occurring during processing to better simulate what
happens to the equivalent commercial products. Proteins were
extracted and digested by trypsin and finally the tryptic peptide
mixtures were analysed by LC–MS/MS (Fig. 1).
After a preliminary characterisation of milk test material (SRM-
1549) by auto-MS/MS analyses, the same experiment was set up
on contaminated cookies spiked with 150 ppm of milk (FC150ppm).
The MS/MS data were fed to the Mascot engine to screen the NCBI
database. Protein identification was achieved by setting ‘‘Other
mammalian” as the taxonomy parameter to force the system to
search among animal proteins. Milk proteins were easily identified
both in the test material and in the cookies at this level of contamination
(150 ppm). Surprisingly no peptides from the milk blactoglobulin,
the best candidate from the SRM-1549 analysis in
terms of peptide abundance (data not shown), were detected in
the spiked cookies. This evidence suggests that the heat treatment
could affect extraction and/or digestibility of the b-lactoglobulin,
resulting in the masking of its presence (Chevalier & Kelly, 2010).
Among the identified milk proteins, a-s1 casein gave the best
MS/MS signals, especially in terms of sensitivity, being the most
abundant protein in milk (32% of all caseins, a class counting 80%
of the total proteins in cow milk) (Wal, 2004). For this reason, as1
casein was selected as the target protein. Furthermore, a-s1
casein lacks cysteines, making the digest protocol extremely easy
because no reduction or alkylation steps are needed. Actually, the
alkylation step may reduce reproducibility and protein recovery
by increasing the peptide variability that could increase MS signal
scattering.