Among the a-s1 casein peptides identified by auto-MS/MS analysis,
three peptides were then selected according to their MS/MS
signal intensities and the absence of missed cleavages and posttranslational
modification sites (Johnson et al., 2011). In order to
0
10000
20000
30000
40000
50000
60000
70000
80000
90000
100000
0 20 40 60 80 100 120 140 160
peak area
milk concentration (mg/kg)
data set 1
data set 2
data set 3
calibration curve 1
calibration curve 2
calibration curve 3
Fig. 3. Plot of MS response versus added milk concentration: data obtained from three replicates of the calibration curve (5–150 mg/kg).
L. Cristina et al. / Food Chemistry 199 (2016) 119–127 123
avoid false-positive results, peptide sequence specificity was veri-
fied by BLAST search (algorithm: blastp; database: UniprotKB;
matrix: PAM 30). Table 1 reports m/z signals of selected peptides
and corresponding product ions. Based on the data reported in
Table 1, the MRM method was developed. According to their signal
intensity, transition 692.86 ? 920.49 m/z was chosen for protein
quantification, while transition 634.35 ? 991.56 m/z was monitored
for confirmatory purposes. LC–MS/MS chromatographic pro-
files of the two monitored transitions and the fragmentation
pattern of the corresponding peptides are shown in Fig. 2.
In order to develop a robust and reliable analytical method,
studies of linearity, sensitivity, accuracy, precision and recovery
were performed on the chosen peptide (FFVAPFPEVFGK), by
monitoring the corresponding 692.86 ? 920.49 m/z transition.
Linear dynamic range was explored over two orders of magnitude.
Linearity was mathematically verified by residual graphical
analysis. Only data points with a percentage error of less than 25%
were accepted. Moreover, more than 75% of calibration curve data
points were required to comply with our accuracy threshold, while
lowest and highest data points could not be excluded. According to
these criteria, the calibration curve was linear from 5 mg/kg to
150 mg/kg (six data points; equation: y = 558.38 ± 20.10x;
correlation coefficient greater than 0.99) (Fig. 3)