Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
H I G H L I G H T S
G R A P H I C A L A B S T R A C T
A R T I C L E I N F O
A B S T R A C T
Article history:
Received 28 February 2014
Received in revised form 13 April 2014
Accepted 18 April 2014
Available online 26 April 2014
The aatoxin M1 (AFLAM1) is a mycotoxin that results from the hydroxylation of the aatoxin B1 (AFLAB1).
It contaminates the milk of animals fed with a diet containing its precursor. In this work, we determined
the occurrence of AFLAB1 and AFLAM1 in milk, as well as the chromatographic conditions to quantify
these mycotoxins. The extraction and quantication of AFLAB1 and AFLAM1 in naturally contaminated
and articially spiked milk samples which are produced and marketed in the state of RS were performed
using the AOAC ofcial method and UHPLC with uorescence detection. We obtained a separation factor
of 2.3 for AFLAB1 and AFLAM1 using a mobile phase consisting of 1% acetic acid:acetonitrile:methanol
(55:10:35). The analytical curves had a wide linearity range and the limit of quantication (LOQm)
concentrations of AFLAB1 and AFLAM1 were equal to 0.5 and 0.25 mg L1, respectively. Samples of
pasteurized and ultra-high-temperature processed (UHT) milk showed natural contamination, and the
levels for both aatoxins ranged from 0.7 to 1.5 mg L1. Raw and concentrated milk samples only
contained AFLAM1, with a maximum average concentration of 1.7 mg L1. These concentrations, higher
than permitted by legislation, conrm the existence of a health risk, as well as highlight the relevance of
searching for alternatives to reduce this contamination.
2014 Elsevier B.V. All rights reserved.
Keywords:
Aatoxin B1
Aatoxin M1
Analytical method
Milk
Liquid chromatography
1. Introduction
69
the southern Brazil (raw milk), local shops (pasteurized milk and
ultra-high temperature-UHT) and a beneciation industry of the
region (raw, pasteurized, concentrated, powder and UHT).
Samples of raw milk were stored in polyethylene terephthalate
packs provided by the producer and kept under cooling temperature (4 C) until the time of analysis, which did not exceed 24 h
after the acquisition thereof.
The samples obtained in the local commerce were kept under
refrigeration from the time of acquisition (pasteurized milk) or
after opening the package (UHT). The sampling unit consisted of
sealed containers without apparent deformation, within the
validity period, containing 1 L of sample and from different
manufacturing batches. They were chosen considering the
availability of lots in major supermarkets in the city of Rio Grande,
RS, Brazil. In view of the geographical distance between the factory
and the laboratory where the analysis were performed, the
samples derived from the processing industry were inoculated
with an specic preservative to maintain the characteristics of the
materials unchanged, namely 2-bromo-2-nitropropane-1,3-diol
(Bronopol1).
2.3. Extraction of AFLAB1 and AFLAM1 from milk
Using a method described by AOAC [11], the volumes of both the
sample and the solvents were reduced three times in order to
minimize the generation of waste. After that, the extraction was
carried out using methanol and celite. A subsequent partition was
carried out using 4% sodium chloride and hexane. The extraction
and the removal of any remaining water in the extracts were
performed, respectively, with chloroform and anhydrous sodium
sulfate. The nal extracts were dried using nitrogen current and
stored at freezing temperature until the time of quantication.
2.4. Establishment of the chromatographic conditions for quantifying
AFLAB1 and AFLAM1 by UHPLC- FL
Tests were carried out to choose the mobile phase that would
result in the best chromatographic separation of aatoxins with
the lowest interference of the matrix components. These tests
were based on the solvent mixture established by AOAC [11],
namely 1% acetic acid:acetonitrile:methanol (40:35:25).
The run was performed on a ultra high-efciency liquid
chromatograph (UHPLC) consisting of a LC-10 AT pump, a DGU
degasser, a CBM-20A controller, a SPD-20A uorescence detector, a
20 mL 7725i manual injector, a LC Solution-Shimadzu software and
a Kromasil1 C18 5 mm 250 4.6 mm column. The standards and
extracts of the samples were eluted at 35 C with a ow rate of
1.0 mL min1. The detection of mycotoxins was performed in
wavelengths of 360 nm to 450 nm for absorption and emission,
with a loop of 20 mL.
The retention factor (k) was calculated dividing the corrected
retention time of each aatoxin by the dead time. The separation
factor (a) was determined by the kA/kB ratio, knowing that the
solute B is more retained in the column than the solute A [12,13].
Two standard curves were built, one for each aatoxin, using the
standard solutions of AFLAB1 and AFLAM1, which were diluted in the
solvent mixture that makes up the mobile phase, at increasing
concentrations (4.5 mg L1; 9 mg L1; 18 mg L1; 27 mg L1; 54 mg
L1; 81 mg L1; 108 mg L1; 162 mg L1 and 270 mg L1). Two other
curves, one for each aatoxin, were made from the fortication (in
triplicate) of milk samples with eight increasing concentrations of
aatoxins (0.25 mg L1; 0.5 mg L1; 4.5 mg L1; 18 mg L1; 54 mg L1;
81 mg L1; 108 mg L1 and 162 mg L1). These samples went through
the same extraction process as the others, according to the AOAC
method, and were subsequently quantied by UHPLC-FL. Then, the
70
2.6. Occurrence of the AFLAB1 and AFLAM1 in milk samples and their
distribution in fractions of the raw material
A natural contamination by aatoxins was observed in milk
samples that underwent different processes such as, raw milk,
pasteurized, UHT, concentrated and powdered.
To investigate the distribution of aatoxins in raw material
during the processing of milk for cheese production, an acid
coagulation of milk was performed, by the method determined by
the Ministry of Agriculture, Livestock and Food Supply [15]. To
separate the milk serum from its clot, a solution of 10% glacial
acetic acid was added to a sample of pasteurized milk preheated to
35 C until the sample reached a pH of 4.6. The levels of AFLAB1 and
AFLAM1 and the protein content in both fractions were determined. This procedure was performed in samples of control milk
(without contamination by addition of the standard) and on
samples spiked with both aatoxins standards in the nal
concentration of 5 mg L1. For these determinations, we used
the extraction method of AOAC [11], as well as a subsequent
quantication by UHPLC-FL.
3. Results and discussion
To obtain an optimal chromatographic separation between
AFLAB1 and AFLAM1 and a reasonable run duration, different ratios
of mobile phase were tested (Table 1). Tests were made for the
retention factor, which indicates the relation between the time
each mycotoxin was retained in the stationary phase and the time
that it was being carried by the mobile phase. The separation factor
was also tested, indicating the selectivity of the chromatographic
system of mycotoxins eluting at adjacent peaks, both estimated for
the different elution gradients of their stationary phase (C18).
The proportion of the mobile phase used in Experiment 1 was
indicated by the ofcial method [11] using a mixture of patterns of
both aatoxins, at the same concentration (0.01 mg mL1), and
resulted in peaks eluted in less than 5 min of run, which could be
affected by other components of the sample eluted in the
beginning of the run. To increase the retention time, the plots of
1% acetic acid and methanol were increased, considering that the
polar character of these solvents can help a better retention of
aatoxins in the column. In Experiment 2, standards with
concentrations of 0.5 mg mL1 were used, resulting in a small
retention time. In the Experiments 4 and 5, the retention time of
AFLAB1 was too long, which promoted the generation of waste and
the reduction of the analytical efciency, resulting in a 25 min run.
Experiment 3 had the most suitable retention times for both
aatoxins, with a decreased risk of the initial interference in a total
running time of 15 min. To conrm the efciency of the separation
under these conditions, we tested it with the AFLAB1 standard
(0.5 mg mL1), the AFLAM1 standard (0.1 mg mL1) and a mixture of
both aatoxins (0.05 mg mL1), obtaining the same retention time
in all cases. In this gradient also occurred the largest separation
factor, indicating that the peaks are presented as far as possible
from each other, ensuring the best selectivity among the
performed tests and a sufciently small interval not to result in
peak broadening (Fig. 1).
The analytical curves for both aatoxins, as well as their
parameters (Table 2), show that the procedure was suitable for the
quantication of both aatoxins. Their indicators meet the
recommendations of analysis reliability, considering the recommendations of ANVISA [16] (a correlation coefcient of 0.99) and of
INMETRO [17] (values above 0.9). Furthermore, the linearity of the
curves showed a wide range of application, and the matrix curves
Table 1
Separation efciency of the mycotoxins eluted by different mobile phases.
Exp.
1
2
3
4
5
40:35:25
50:15:35
55:10:35
60:05:35
65:10:25
Retention factor
(k)
AFLAM1
AFLAB1
AFLAM1
AFLAB1
AFLAM1 AFLAB1
3.4
5.2
6.6
11.8
13.5
4.4
8.0
11.8
20.0
25.0
0.36
1.08
1.64
3.72
4.40
0.76
2.20
3.72
7.00
9.00
2.11
2.04
2.27
1.88
2.05
71
Fig. 1. Elution chromatograms of the AFLAB1 and AFLAM1 in UHPLC-FL using different mobile phase gradients composed of 1% acetic acid:acetonitrile:methanol. (1)
40:35:25; (2) 50:15:35; (3) 55:10:35; (4) 60:05:35; (5) 65:10:25.
72
Table 2
Analytical parameters evaluated in UHPLC-FL.
Analytical parameters
AFLAB1
AFLAM1
Solvent curves
Linearity (mg L1)
Coefcient of correlation
Coefcient of determination
LODi (mg L1)
LOQi (mg L1)
y = 2,947,476x 8261.30
4.5 a 108
0.9987
0.9974
1.5
4.5
y = 2,246,144 x 6973.83
4.5 a 270
0.9991
0.9983
1.5
4.5
Work curve
Linearity (mg L1)
Coefcient of correlation
Coefcient of determination
LODm (mg L1)
LOQm (mg L1)
y = 1,506,689x 14,579.84
0.5 a 162
0.9918
0.9836
0.2
0.5
y = 693,877.2x 5680.73
0.25 a 162
0.9933
0.9866
0.09
0.25
AFLAB1-aatoxin B1, AFLAM1-aatoxin M1, LOD1-instrument detection limit, LOQ1-instrument quantitation limit, LODm-method detection limit, LOQm-method quantitation
limit.
Table 3
Recovery of the AFLAB1 and AFLAM1 extracted from milk by the AOAC method.
Fortication (mg L1)
0.25
0.5
3.0
6.0
8.0
10.0
Average
AFLAM1
83 (3)
78 (16)
73 (14)
76 (21)
98 (12)
82
75 (29)
91 (15)
114 (15)
111 (33)
94 (11)
77 (19)
94
Table 4
Incidence of AFLAB1 and AFLAM1 in different samples of milk.
Milk sample
AFLAB1
% incidence
Raw
0 (0/7)
Pasteurized
41.7 (5/12)
UHT
13.3 (2/15)
Concentrated 0 (0/3)
Powdered
0 (0/3)
AFLAM1
Average contamination (mg L1)
(CV%)
% above the
legislationa
%
incidence
ND
1.476 (51)
0.690 (20)
ND
ND
100
100
28.6
58.3
66.7
66.7
0 (0/3)
0.835 (9)
0.884 (42)
1.168 (27)
1.718 (8)
ND
100
100
100
100
(2/7)
(7/12)
(10/15)
(2/3)
Fig.
73
2. Chromatograms of pasteurized milk without (1) and with (2) natural contamination by aatoxins.
Table 5
Distribution of AFLAB1 and AFLAM1 between coagulated milk fractions.
Sample
Control
Spiked
Milk
Serum
Clot
Milk
Serum
Clot
Humidity
(%)
(CV%)
Proteins (%)
(CV%)a
88.9 (0.5)
94.8 (0.1)
27.3 (1.9)
2.7 (7.4)
0.6 (17.0)
11.7 (15.0)
ND
ND
ND
5.11 (9.5)
ND
5.42 (8.2)
1.20 (11.1)
ND
1.22 (6.4)
5.86 (5.7)
ND
6.48 (11.5)
74