Selection and optimization of the mass spectrometry conditions
The mass spectrometer parameters were determined and optimized in the ion spray mode at a flow rate of 5 μL/min, and the standard solution of the BAs (250 μg mL-1) was introduced by a continuous pump. To confirm the quasi-molecular ion [M+H]+ peak, the BA standard solution(Table 1) was used for the parent ion full scan in the ESI positive ion mode. Subsequently, the molecular ion was selected as the precursor ion for analysis through two-stage mass spectrometry. Multiple reaction monitoring (MRM) scanning in the positive mode was performed on the BAs to identify ion transition. The ions with a stronger abundance and less interference were selected as the qualitative ions. In addition, the mass parameters, such as the source voltage capillary, source temperature, desolvation temperature, source gas flow and desolvation, were adjusted to optimize the characteristic ion and to determine the qualitative ions and appropriate mass spectra parameters (Table 2). The optimized ESI source conditions were as follows: source voltage capillary, 3.52 kV; cone, 25 V; source temperature desolvation temperature, 350°C; and source gas flow desolvation, 650 L/h. The MRM mass spectrum of the BAs is shown in Figure. 1.
Table 1 Target compounds and relevant data
Analyte
|
Abbreviation
|
CAS
|
Structure
|
Formula
|
Molecular mass
|
Histamine
|
HIS
|
51-45-6
|
|
C5H9N3
|
111.15
|
Cadaverine
|
CAD
|
462-94-2
|
|
C5H14N2
|
102.18
|
Putrescine
|
PUT
|
110-60-1
|
|
C4H12N2
|
88.15
|
2-phenylethylamine
|
PEA
|
64-04-0
|
|
C8H11N
|
121.18
|
Tyramine
|
TYR
|
51-67-2
|
|
C8H11NO
|
137.18
|
Tryptamine
|
TRY
|
61-54-1
|
|
C10H12N2
|
160.21
|
Serotonin hydrochloride
|
5-HT
|
153-98-0
|
|
C10H13ClN2O
|
212.676
|
Urocanic Acid
|
UCA
|
104-98-3
|
|
C6H6N2O2
|
138.12
|
Agmatine
|
AGM
|
2482-00-0
|
|
C5H14N4
|
130.19
|
Octopamine hydrochloride
|
OA
|
770-05-8
|
|
C8H12ClNO2
|
189.64
|
Table 2 MS detection parameters for the 10 Bas
BA
|
Parent Ion(m/z)
|
Daughter(m/z)
|
Dwell(s)
|
Cone(V)
|
Collision Energy(V)
|
Putrescine
|
88.99
|
71.94*
|
0.005
|
20
|
8
|
29.98
|
0.005
|
18
|
18
|
Cadaverine
|
103.02
|
68.99
|
0.005
|
20
|
16
|
85.95*
|
0.005
|
19
|
8
|
Histamine
|
111.92
|
82.94
|
0.005
|
26
|
16
|
94.91*
|
0.005
|
26
|
16
|
2-phenylethylamine
|
121.93
|
79.55
|
0.005
|
24
|
22
|
104.92*
|
0.005
|
24
|
10
|
Agmatine
|
130.91
|
59.92
|
0.005
|
26
|
16
|
71.95*
|
0.005
|
26
|
16
|
113.96
|
0.005
|
26
|
14
|
Tyramine
|
137.96
|
93.06
|
0.005
|
20
|
22
|
121.00*
|
0.005
|
20
|
10
|
Urocanic Acid
|
138.83
|
120.90*
|
0.005
|
34
|
20
|
92.88
|
0.005
|
34
|
28
|
Octopamine
|
153.92
|
90.93
|
0.005
|
12
|
26
|
135.92*
|
0.005
|
12
|
10
|
Tryptamine
|
160.94
|
117.00
|
0.005
|
18
|
26
|
143.98*
|
0.005
|
18
|
12
|
Serotonin hydrochloride
|
176.96
|
159.93*
|
0.005
|
18
|
18
|
Note:* Represents qualitative ions
Optimization of the chromatographic conditions
The aim of this study was to separate the BAs. To determine the best analytical performance of this method, the BA separation performance was optimized, and various chromatographic conditions, including different mobile constitutions, different stationary phases, various flow rates, various back pressures, and different column temperatures, were tested.
Influence of stationary phases
Achieving excellent separation of the target analytes using a suitable chromatographic column is crucial. A prominent challenge is screening an appropriate column for acquiring a good peak shape and satisfactory resolution and sensitivity for the compound. Thus, the three columns, namely, (1) Waters ACQUITY UPC2 BEH (hybrid silica without bonding), (2) Waters ACQUITY UPC2 HSS C18 SB (classical silica bonded with C18) and (3) Waters ACQUITY UPC2 Torus 2-PIC (hybrid silica with 2-picolylamine bonding), were tested for the separation of ten Bas (Figure.2). The results suggested that under identical conditions, there was no baseline separation of the analytes on the Torus 2-PIC column, and the target substances on the HSS C18 SB column were not completely separated. Some analytes were not retained or were retained weakly, but the separation of the ten BAs could be successfully accomplished in 5 min using the BEH columns. Hence, the BEH column was selected in the present study.
Choice of organic modifiers
In view of the fact that the mobile phase used in the UPC2 system is mainly non-polar CO2, to enhance the polarity of the mobile phase and to ensure that the polar component could be eluted, it was necessary to add a modifier to promote separation. The role of the addition of the organic modifier was to cover the active sites on the surface of the column packing material and change the density of the mobile phase(Berger, 1997). Hence, four different modifiers, namely, methanol, ethanol, isopropanol and methanol containing 0.2% (V/V) ammonia, were investigated. The results suggested that methanol was an ideal modifier for the separation of the target, and the introduction of 0.2% ammonia into methanol increased the peak capacity and improved the peak symmetry. Thus, methanol containing 0.2% ammonia was selected as the best modifier in the mobile phase.
Effects of the back pressure and temperature
The back pressure directly interfered with the separation efficiency of the analyte. We also investigated the effect of CO2 supercritical fluid on the sample separation in the range from 1300 to 2000 psi. At the optimal back pressure of 1500 psi, the retention time was shortened, and the best separation effect for the UPC2 analysis was observed. The column temperature was also an important factor that affected analyte retention. Increasing the temperature changed the mobile phase density and increased the retention. As the temperature increased past the maximum, the retention declines(Zou, et al., 2000). The column temperature was tested from 30°C to 50°C. The best resolution was acquired at 40°C. Thus, 40°C was selected as the optimal column temperature, and a flow rate of 1.2 mL/min was optimal.
We detected spermine and spermidine, but these did not elute. This finding was potentially obtained because spermine and spermidine have more amino groups than other analytes. As the number of amino groups increases, the polarity increases(Camacho-Munoz, Kasprzyk-Hordern, & Thomas, 2016), which causes the analyte polarity to be higher than that of the column. Therefore, the separation of these two compounds was not ideal.
Analytical performance of this strategy
Under the optimal conditions, the sensitivity and the detection range of the chromatograph were quantified by measuring the quantitative ion peak area. The concentrations of the standard solution of each BA in the range of 0.05 to 5 μg mL-1 was prepared from stock solutions with a concentration of 100 µg mL-1. The calibration curves of the BAs were established using the mixed standard solution with an injection volume of 1 µL, and the linear regression equation is shown in Table 3. Linear relationships with acceptable correlation coefficients of R2>0.9920 were achieved. The sensitivity of the target analytes rely not only on the different chemical structures of the BAs but also on the ionization efficiency of the target in the MRM mode. For verification of the LOD and LOQ, the above 10 BA standards were added to the blank samples. The results indicated that the proposed method had good sensitivity, with detection limits of 0.0010-0.0086 μg mL-1, and the LOQs were in the range of 0.0028 to 0.028 μg mL-1 (S/N=3). The LODs and LOQs in this work were 10-fold lower than those reported by Simat et al. (2011) and Tahmouzi et al. (2011). There was no interference in the detected sample after selecting 20 channels of UHPSFCUHPSFC-MS. The detection accuracy for BAs in fermented fish samples was also measured by determining the recovery of the BAs using the standard addition method. The samples were spiked with the mixed BA standards with concentrations of 0.4 and 2.0 μg mL-1. The samples were routinely pretreated. The results are shown in Table 4, and the recoveries for the BAs were all between 95.10% and 103.78%. The recovery results were acceptable for all BAs, and the results indicated that this method is suitable for determining the BA content in fish samples. Furthermore, this strategy might be useful for the detection of BAs in complex food matrices.
The repeatability of the recovery test was estimated for all spiked levels (n=5). The relative standard deviation (RSD) was between 0.15% and 1.42% (RSD<5%) (Table 4). Thus, the analytical method employed in this work was sensitive enough to detect and analyze the targets (BAs) in real fish samples.
Table 3 Linear range, LODs and LOQs for the 10 Bas
BA
|
RT
|
Calibration curve
|
R2
|
LOD
(μg mL-1)
|
LQD
(μg mL-1)
|
Putrescine
|
3.41
|
y=2.24x-448.59
|
0.9964
|
0.0077
|
0.025
|
Cadaverine
|
3.45
|
y=0.59x-33.02
|
0.9968
|
0.0086
|
0.028
|
Histamine
|
2.91
|
y=3.42x-1913.72
|
0.9977
|
0.0037
|
0.012
|
2-phenylethylamine
|
1.56
|
y=6.13x+397.09
|
0.9969
|
0.0010
|
0.0028
|
Agmatine
|
3.50
|
y=6.19x+362.53
|
0.9920
|
0.0029
|
0.0096
|
Tyramine
|
2.14
|
y=5.28+294.24
|
0.9973
|
0.0037
|
0.012
|
Urocanic Acid
|
2.53
|
y=2.39+174.19
|
0.9968
|
0.0014
|
0.0046
|
Octopamine
|
2.32
|
y=2.10+75.74
|
0.9952
|
0.0020
|
0.0066
|
Tryptamine
|
2.23
|
y=6.88+443.54
|
0.9971
|
0.0015
|
0.0048
|
Serotonin hydrochloride
|
2.63
|
y=1.74x-56.54
|
0.9953
|
0.0033
|
0.011
|
Table 4 BA content (mg/L) in fish sauce, fermented fish and shrimp samples (n=5)
Compound
|
Spiked level(μg/mL)
|
Recovery(%)
|
%RSD
|
Putrescine
|
0.4
|
98.85
|
1.42
|
2
|
102.28
|
1.33
|
Cadaverine
|
0.4
|
100.70
|
1.03
|
2
|
97.34
|
0. 91
|
Histamine
|
0.4
|
100.15
|
1.18
|
2
|
95.54
|
0.96
|
2-phenylethylamine
|
0.4
|
99.45
|
0.16
|
2
|
98.87
|
0.59
|
Agmatine
|
0.4
|
103.78
|
1.10
|
2
|
96.50
|
1.02
|
Tyramine
|
0.4
|
97.95
|
0. 61
|
2
|
100.29
|
0.35
|
Urocanic Acid
|
0.4
|
99.61
|
0.93
|
2
|
100.275
|
0.52
|
Octopamine
|
0.4
|
98.75
|
0.37
|
2
|
97.82
|
0.29
|
Tryptamine
|
0.4
|
101.05
|
0.60
|
2
|
95.10
|
0.39
|
Serotonin hydrochloride
|
0.4
|
95.36
|
0.39
|
2
|
96.88
|
0.15
|
Analytical application
The proposed method was successfully used to detect BAs in dried small shrimp and fermented fish samples. As shown in Figure. 3, the chromatogram of the ten BA standard solutions was obtained using the gradient profile described in section 2.4. All ten BAs were effectively separated using a total run time of 5 min with excellent peak resolution and sharpness. This result indicated that this method offers good separation among the BAs and a faster analysis speed compared with HPLC (Abdulhussain, et al., 2021; Toribio, et al., 2021). The feasibility of the applied strategy was evaluated using dried small shrimp, fermented fish and fish sauce samples obtained from a local supermarket in China. The results are presented in Figure. 4, Figure. 5, Figure. 6 and Table 5. The total BA content ranged from 3.31 to 763.36 μg mL-1, and the amine concentrations of the analyzed samples of dried small shrimp and fermented fish were far below the toxicity levels that pose a health risk. The most important BAs, including histamine, tyramine, and cadaverine, were all detected. Extremely high levels of urocanic acid were found. This result can probably be explained because the dried small shrimp and fermented fish were susceptible to microorganisms that continuously generate urocanic acid during storage. Agmatine and octopamine hydrochloride were not found in the fermented fish. The levels were in accordance with the results reported in a previous publication(Lee, et al., 2015). In summary, the BA concentrations in the dried small shrimp and fermented fish samples did not exceed toxic levels, which illustrated that the two samples were of good quality. In the fish sauce sample, the only BAs that were detected were histamine, 2-phenylethylamine, serotonin hydrochloride and tryptamine. The histamine content was 763.36 μg mL-1, which was significantly higher than the tolerance level suggested by the Food and Drug Administration for aquatic products. In previous studies, Stute and Jiang(Jiang, et al., 2014; Stute, et al., 2002) determined the histamine content in commercially produced fish sauce samples and found histamine contents of 783.9 μg mL-1 and 729 μg mL-1 in selected samples, which are higher than the standard range. The results indicated that consuming fish sauce for a long time might pose a threat to human health and that the BA content in fish sauces should be monitored.
Table 5 Spike recovery tests
Samples
|
fermented fish(μg mL-1)
|
dried small shrimps(μg mL-1)
|
Fish sauces
(μg mL-1)
|
PUT
|
7.60 ±0.010
|
9.87±0.032
|
ND
|
CAD
|
12.52±0.050
|
5.85±0.013
|
ND
|
HIS
|
30.05±0.019
|
12.88±0.060
|
763.36±0.053
|
2-PEA
|
6.42±0.021
|
19.78±0.040
|
38.48±0.018
|
AGM
|
ND
|
ND
|
ND
|
TYR
|
15.65±0.023
|
7.88±0.010
|
ND
|
UCA
|
50.66±0.014
|
101.83±1.42
|
ND
|
OA
|
11.07±0.096
|
59.96±0.015
|
ND
|
TRY
|
19.69±0.015
|
25.68±0.030
|
60.13±0.022
|
5-HT
|
ND
|
3.31±0.039
|
30.10±0.032
|