Research article

Received: 20 January 2014 Revised: 17 April 2014 Accepted: 19 June 2014 Published online in Wiley Online Library

(wileyonlinelibrary.com) DOI 10.1002/jms.3418

Simultaneous determination of tenunderivatized biogenic amines in meat byliquid chromatography-tandem massspectrometry (HPLC-MS/MS)†

Veronica Sirocchi, Giovanni Caprioli, Massimo Ricciutelli, Sauro Vittoriand Gianni Sagratini*

Biogenic amines (BAs) are considered to be important indicators of freshness and quality in food. In this work, an analyticalmethod for analyzing ten underivatized BAs in meat by performance liquid chromatography-tandem mass spectrometryhas been developed. Comparison between ion trap and triple quadrupole as mass analyzers indicated that the latter providesgreater sensitivity and selectivity. The range of the correlation coefficients of the calibration curves of the analyzedcompounds was 0.987–0.999, and the limits of detection and limits of quantification were in the range of 0.002–0.1mg l�1

and 0.008–0.5mg l�1, respectively. Once validated, the method was used to analyze the concentrations of BAs in 16commercial meat samples, for evaluating the freshness of food through the study of BA indices, i.e. biogenic amine index(BAI) and the ratio spermidine/spermine (SPD/SPE). The results indicated that all the samples were fresh, with a BAI lower than1.49mgkg�1 and a SPD/SPE ratio lower than 0.41 in each case. This methodology for testing the freshness of meat has potentialfor quality control applications along the entire production chain of meat products. Copyright © 2014 John Wiley & Sons, Ltd.

Keywords: biogenic amines; meat; liquid chromatography; tandem mass spectrometry; triple quadrupole

* Correspondence to: Gianni Sagratini, University of Camerino, School of Phar-macy, via S. Agostino 1, 62032 Camerino, Italy. E-mail: gianni.sagratini@unicam.it

† This article is part of the Journal of Mass Spectrometry special issue entitled“3rd MS Food Day” edited by Gianluca Giorgi.

School of Pharmacy, University of Camerino, via S. Agostino 1, 62032,Camerino, Italy

819

Introduction

Biogenic amines (BAs) are basic nitrogenous compounds presentin food and produced by different mechanisms, such asdecarboxylation of amino acids (catalyzed by the decarboxylaseenzymes contained in various microorganisms) or by the normalcellular metabolism of tissues.[1] In low concentrations, BAs areessential for many physiological functions, but in high concentra-tions they can generate toxicological problems such as migraines,headaches, gastric and intestinal problems, and pseudo-allergicresponses[2] and thus are considered important indicators of foodfreshness and quality. Two key indicators, the biogenic amineindex, or BAI, which is the sum (mgkg�1) of putrescine, cadaverine,histamine and tyramine, and the spermidine/spermine ratio(SPD/SPE), are used to assess food quality.[3]

Such an important entity as the European Union, with itsRapid Alert System for Food and Feed, has affirmed the valueof BA assessment in food quality control. In 2013, it publishedits analysis of 10 years worth of notifications recorded in itsdatabase and reported that the average histamine concentra-tions in fish products, tracked from 2002 to 2012, wasbetween 1000 and 2000mg kg�1 in 12% of cases and above2000mg kg�1 in 11% of cases.[4] Recently, a study publishedby Erim F.B.[5] reported that, after fish products, meat is thefood matrix most studied for the presence of BAs, consideredmarkers of freshness and hygiene during storage of thisproduct. Another study emphasized that high levels of BAsand the nitrates and nitrites added as preservatives in meatproducts are associated with the potential for forming carcino-genic nitrosamines.[6]

J. Mass Spectrom. 2014, 49, 819–825

Over the years, various methods have been proposed foranalyzing BAs in food with the use of high performance liquidchromatography (HPLC) coupled to various detectors. The mostfrequently used methods are based on derivatization reactionsand separation processes followed by ultraviolet (UV) orfluorescence detection.[2] In these cases, the derivatization stepis fundamental because most BAs lack chromophore groups.The most used derivatizating agents are o-phtalaldeyde (OPA),[7]

benzoyl chloride,[8] dansyl chloride[9] and 6-aminoquinolyl-N-hydrosuccinimidyl carbamate.[10] HPLC can be used in combi-nation with mass spectrometry (MS) or a MS tandem system(MS/MS) for analyzing derivatized BAs in food matrices such ascheese, fermented meat,[11] beer[12] and wines.[13] But often thederivatization step is time consuming and, as an accessorychemical reaction, yields poor recoveries and low sensitivity. Forthese reasons, it is preferable to analyze underivatized BAs infood. To our knowledge, very few methods using an HPLCcoupled to an MS system for analyzing underivatized BAs in foodhave been published, and none of them have focused on BAs inmeat. Gosetti et al.[14] analyzed six BAs in typical Piedmontcheeses by HPLC-MS/MS (ion trap (IT)) and compared the results

Copyright © 2014 John Wiley & Sons, Ltd.

V. Sirocchi et al.

820

with the HPLC-UV method, while Calbiani et al.[15] proposed ananalytical method for analyzing histamine, β-phenylethylamineand tyramine in cheese by matrix solid-phase dispersion followedby HPLC-MS/MS (triple quadrupole (TQ)). Millàn et al.[16] proposeda method for determining eight BAs in wine by HPLC-MS/MS (IT)and Sagratini et al.[9] reported an HPLC-MS/MS (TQ) method fordetermining BAs in fish from Valencia (Spain).The aim of this work was to develop an analytical method

that simultaneously determines ten underivatized BAs, namely,spermidine, cadaverine, putrescine, histamine, agmatine, spermine,tyramine, serotonin, phenylethylamine, tryptamine, in fresh meatby using solid phase extraction (SPE) and an HPLC-MS/MS system.Moreover, the performances of two mass analyzers, the IT and theTQ, were studied and compared. Finally, the method was appliedto analyze the BA levels in 16 commercial meat samples, andthe results were used to formulate the BA quality indices of thesamples, as indicators of their freshness.

Experimental

Materials and standards

BAs, namely spermine tetrahydrochloride (SPE, C10H26N4·4HCl,>98%, CAS No. 306-67-2), spermidine trihydrochloride (SPD,C7H17N3 3HCl,>98%, CASNo. 334-50-9), cadaverine dihydrochloride(CAD, C5H14N2 2HCl, >98%, CAS No. 1476-39-7), putrescinedihydrochloride (PUT, C4H12N2 2HCl, >98%, CAS No. 333-93-7),histamine dihydrochloride (HIS, C5H9N3·2HCl, >99%, CASNo. 56-92-8), tyramine hydrochloride (TYR, C8H11NO·HCl, >98%,CAS No. 60-19-5), 2-phenylethylamine hydrochloride (PHE,C8H11N·HCl, >98%, CAS No. 156-28-5), tryptamine hydrochloride(TRY, C10H12N2·HCl, >99%, CAS No. 343-94-2), serotoninhydrochloride (SER, C10H12N2O·HCl, >98%, CAS No. 153-98-0),and agmatine sulfate salt (AGM, C5H14N4·H2SO4, >97%, CASNo. 2482-00-0) were supplied by Sigma-Aldrich (Milano, Italy).Trichloroacetic acid (TCA) for analysis was from Sigma-Aldrich(Milano, Italy). Individual stock solutions of BAs were preparedby dissolving 10mg of each compound in 10ml of HCl 0.1M(Merck Darmstadt, Germany) and were then stored in glass-stoppered bottles at 4 °C. Standard working solutions, at variousconcentrations, were prepared daily by appropriate dilution ofdifferent aliquots of the stock solutions with water. HPLC-grademethanol was supplied by Sigma-Aldrich (Milano, Italy). Deionizedwater (<8MΩ cm resistivity) was obtained from theMilli-Q SP ReagentWater System (Millipore, Bedford, MA, USA). All the solvents andsolutions were filtered through a 0.45-μm PTFE filter from Supelco(Bellefonte, PA, USA) before use. SPE STRATA X 33μ Cartridges,200mg/6ml, were purchased by Phenomenex (Bologna, Italy).

Sample collection

Fresh beef (13 samples), chicken (1 sample), lamb (1 sample) andrabbit (1 sample) purchased from local butchers in Camerino,Italy, were brought to the laboratory and immediately analyzed.

Sample preparation

The analytical procedure was a slightly modified version of apreviously published method used for fish.[9] Each slice of meatwas ground in a blender, and then 5 g was extracted by Ultra-Turrax S 18N-10G homogenizer (IKA-Werke Gmbh & Co.,Germany) and 5% TCA; after the centrifugation step, the

wileyonlinelibrary.com/journal/jms Copyright © 2014 Jo

supernatants were collected and purified on SPE STRATA Xcartridges (conditioned with 4ml of methanol followed by 4mlof Milli-Q water). Then, 2ml of the sample (supernatant) with apH adjusted to 11 by adding 200μl of NH4OH 28% was passedthrough the cartridges. After sample loading was complete,cartridges were rinsed with 2ml of a mixture of MeOH/H2O(5/95 v/v), then dried under vacuum to remove excess of water.Analytes were eluted from the STRATA X sorbents with 2 + 2mlof a solution of methanol/acetic acid (99/1 v/v). The eluting solu-tion was dried with nitrogen gas, and the residue was collectedwith 2ml of HCl 0.1M, filtered and injected into the LC-MS/MS.

Liquid chromatography-tandem mass spectrometry

The separation of the BA was achieved using a Synergi HydroRP-80A analytical column (250×4.6mm I.D., particle size 4μm) fromPhenomenex (Torrance, CA, USA). The columnwas thermostatted at25 °C. The mobile phase for LC-MS/MS analysis was ammoniumformiate 5mM and formic acid in water (pH= 3.3) (A) andmethanol (B), at a flow rate of 0.5mlmin�1.The gradientprogram was: 0min 30% B, 0–15min 90% B, 15–20min 30% Band 20–25min 30% B. The injection volume was 10μl for HPLC-MS/MS analysis using the IT and 5μl for HPLC-MS/MS analysisusing the TQ.

IT

HPLC-MS/MS studies were performed using an Agilent 1100 seriesinstrument (Santa Clara, CA, USA), made up of an autosampler, aquaternary solvent pump and a diode-array detector inconjunction with an MSD Trap SL mass spectrometer equippedwith an ESI source operating in positive ionization mode. Themass spectrometer was tuned for each compound, optimizingionization source parameters, voltages of the lenses and trapconditions in the ExpertTune mode of the Daltonic EsquireControl software, while infusing a standard solution (10μgml�1)via a syringe pump at a flow rate of 5μl min�1, which was mixedwith the mobile phase at 0.5mlmin�1 by means of a T piece.The optimized parameters of the ESI interface were vaporizertemperature, 325 °C; nebulizer gas (nitrogen) pressure, 50 psi;drying gas (nitrogen) flow rate, 10mlmin�1; capillary voltage,3500 V. The mass spectrometer was run in MS/MS and full scan-extract ion. The trap parameters were set in ion charge controlusing rolling averaging set at 2 with a target of 10,000, andmaximum accumulation time of 200ms at m/z range from 40 to300 u. MS/MS acquisition parameters used for the identificationof the target compounds are reported in Table 1.

TQ

HPLC-MS/MS studies were performed using an Agilent 1290Infinity series coupled to a Triple Quadrupole 6420 LiquidChromatography/Mass Spectrometry System from Agilent Tech-nology (Santa Clara, CA) equipped with an ESI source operatingin positive ionization mode.

Optimization of the LC-MS/MS conditions was carried out byvarying them in flow injection analysis of the analytes (5μl of a50μgml�1 individual standard solutions). Operating ESIconditions were gas temperature 300 °C, nebulizer gas 40 psi,drying gas (nitrogen) flow rate 12mlmin�1, and capillary voltage4000 V. The MS/MS acquisition parameters (MRM mode) used forthe quantification of the target compounds are provided in

hn Wiley & Sons, Ltd. J. Mass Spectrom. 2014, 49, 819–825

Table 1. MS/MS acquisition parameters (ion trap) used for the identification of the target BAs

Biogenic amines Precursor iona [M+H]+ Product ion Cutoff Amplitude Time window (min)

SPD 146 129 39 0.4

CAD 103 86 28 0.3 0–4.7

PUT 89 72 24 0.3

HIS 112 95 30 0.4

AGM 131 114 35 0.4

SPE 203 129 55 0.4

TYR 138 121 37 0.4

SER 177 160 48 0.4

PHE 122 105 33 0.4 8.0–25

TRY 161 144 43 0.4

aIn MS/MS experiments (quadrupole ion trap), in all cases the width parameter was equal to 4.0.

Underivatized biogenic amines by HPLC-MS/MS

Table 2. The ESI interface and mass spectrometer parameterswere optimized to obtain maximum sensitivity.

Results and discussions

Liquid chromatography tandem mass spectrometry (IT)

The performances of the HPLC-MS/MS method for analyzing tenunderivatized BAs in meat were evaluated by using two differentmass analyzers, namely, the IT and the TQ. First, the IT was used inMS/MS mode, by selecting for each BA the precursor ion that wasthe base peak of the mass spectrum, corresponding in all cases tothe protonated molecule [M+H]+. Table 1 reports the MS/MSacquisition parameters used for analyzing BAs: precursor ions,product ions, cutoff, amplitude, and time windows. Eachprecursor ion was then subjected to tandem mass spectrometryexperiments in which we tested various values of amplitude; forall BAs, and especially for spermine the sensitivity in detectingsignals was poor, probably due to the difficulty of isolating inthe same time window various ions subjected to fragmentation.After these preliminary experiments, we decided to use the ITmass analyzer in full scan mode and then to perform theextraction of ions characteristic for each molecule. For all BAs, thiswas the protonated molecule [M+H]+ reported in Table 1.Figure 1 reports an HPLC-MS chromatogram of a standardmixture of BAs at concentration of 5mgkg�1 obtained in full scanMS condition followed by the extraction of a single characteristicion for each molecule. As can be seen, there was poor sensitivityin detecting signals for both cadaverine and putrescine, and avery bad peak shape was obtained for spermine. Better results

Table 2. MS/MS acquisitions parameters (triple quadrupole in MRM

Biogenic amines Precursor ion [M+H]+ Product ion Fragme

SPD 146 72 91

CAD 103 86 72

PUT 89 72 72

HIS 112 95 76

AGM 131 72 83

SPE 203 129 96

TYR 138 121 58

SER 177 160 68

PHE 122 105 67

TRY 161 144 58

J. Mass Spectrom. 2014, 49, 819–825 Copyright © 2014 John W

in terms of sensitivity were obtained for histamine, agmatineand tyramine. Limits of detection (LODs) and limits of quantifica-tion (LOQs) obtained by HPLC-MS (IT) experiments for theanalyzed BAs are reported in Table 3.

Liquid chromatography tandem mass spectrometry (TQ)

Performance of the IT as mass analyzer was tested in comparisonwith that shown by the TQ mass analyzer. Table 2 reports theMS/MS acquisition parameters in MRM mode used for analyzingBAs: precursor ions, product ions, fragmentor, collision energy(CE), retention time and retention windows. All BAs presentmore than one transition. The most intense transitions, reportedin Table 2, were used for the quantitative analysis and refer toquantifier transition, while the others were employed in theidentification step.

For each BA, the precursor ion was the protonated molecule[M+H]+. Table 2 reports the proposed fragmentation pattern ofeach analyzed BA. As can be seen, for most BAs (cadaverine,putrescine, histamine, tyramine, serotonine, phenylethylamineand tryptamine), the product ion was generated by the loss ofa fragment of ammonia, [M+H-NH3]

+.[9] According to theliterature,[14] the product ion of spermine at m/z 129 wasgenerated by the loss of a fragment at m/z 74, while in the caseof spermidine and agmatine the same heterocyclic product ionwas generated at m/z 72, in agreement with the findings ofHäkkinen et al.[17] Figure 2 reports an HPLC-MS/MS chromato-gram in MRM mode of a spiked meat sample at a concentrationof 0.25mg kg�1 in which each monitored analyte can bedetected. All BAs showed excellent signals and a very good peak

mode) used for the identification of the target BAs

ntor (V) CE (V) Retention time (min) Retention Window

12 3.9 1

4 4.1 1

4 4.1 1

12 4.1 1

16 4.2 2.5

16 4.6 1

4 5.8 1

9 6.1 1

8 8.4 1

12 9.0 2.5

iley & Sons, Ltd. wileyonlinelibrary.com/journal/jms

821

Figure 1. HPLC-MS (ion trap) chromatogram acquired in MS full scan/extract ion of a standard mixture of BAs at concentrations of 5mg kg�1.

Table 3. Concentration range, linearity (R2), limits of detection (LODs) and limits of quantification (LOQs) of analyzed BAs using ion trap (IT) andtriple quadrupole (TQ) as mass analyzers

HPLC-MS (ion trap) HPLC-MS/MS (triple quadrupole)

Name Concentration range(five points) mg l �1

Correlationcoefficient (R2)

LOD mg l�1 LOQ mg l�1 Concentration range(five points) mg l�1

Correlationcoefficient (R2)

LOD mg l�1 LOQ mg l�1

SPD 1–10 0.999 0.1 0.5 0.01–0.5 0.997 0.002 0.008

CAD 1–10 0.993 0.3 1 0.01–0.5 0.997 0.002 0.008

PUT 1–10 0.999 0.3 1 0.01–0.5 0.987 0.002 0.008

HIS 1–10 0.980 0.03 0.1 0.01–0.5 0.997 0.002 0.008

AGM 0.1–1 0.999 0.03 0.1 0.01–0.5 0.994 0.002 0.008

SPE 5–15 0.993 0.2 0.8 0.5–10 0.994 0.1 0.5

TYR 0.1–15 0.970 0.03 0.1 0.01–0.5 0.997 0.002 0.008

SER 0.1–15 0.960 0.1 0.5 0.01–0.5 0.996 0.002 0.008

PHE 1–10 0.999 0.1 0.5 0.01–0.5 0.997 0.002 0.008

TRY 1–10 0.999 0.1 0.5 0.01–0.5 0.998 0.002 0.008

V. Sirocchi et al.

822

shape except spermine, which gave a quite large signal, probablyinfluenced by some coeluted substance. The TQ used as massanalyzer for the concentrations of BAs in meat showed highersensitivity than did the IT; in fact, as reported in Table 2, the LOQsobtained with the TQ were from 20 to 60 times lower (thusgreater sensitivity) than those obtained with the IT mass analyzer.

Method validation

The analytical performances of the HPLC-MS/MS (IT and TQ)method are reported in Table 3. Calibration curves of theanalyzed compounds were constructed by injecting into theHPLC-MS/MS (IT) 10μl of each BA standard solution at differentconcentrations, from 0.1 to 15mg l�1. In the case of the HPLC-MS/MS (TQ), 5μl of each BA standard solution was injected at

wileyonlinelibrary.com/journal/jms Copyright © 2014 Jo

different concentrations, from 0.01 to 10mg l�1. The correlationcoefficients of calibration curves of the analyzed compoundswere in the range of 0.960–0.999 (IT) and 0.987–0.999 (TQ). Fivereplicates of standard mixtures of BAs at 0.25 and 0.5mg l�1 ofconcentration were performed in both the IT and the TQ systems.With the IT, the related standard deviations (RSDs) ranged from3.9 to 10.1% for run-to-run precision, and from 7.1 to 18.2% forday-to-day precision, while with the TQ, they ranged from 0.8 to3.0% for run-to-run precision, and from 0.5 to 12% for day-to-dayprecision. The recoveries obtained by spiking meat samples atconcentrations of 0.25 and 0.5mg kg�1 with standard mixtureof BAs were in a range of 47.4–120.0% and 48.4–127.1%, respec-tively. The repeatability of the method, evaluated 5 times, wasexpressed by % RSDs, lower than 12.3% in all cases. LODs andLOQs were estimated on the basis of 3:1 and 10:1 signal-to noise

hn Wiley & Sons, Ltd. J. Mass Spectrom. 2014, 49, 819–825

Figure 2. HPLC-MS/MS chromatogram (triple quadrupole) acquired in MRM mode of a spiked meat sample at a concentration of 0.25mg kg�1. SingleBAs are reported in the figure.

Underivatized biogenic amines by HPLC-MS/MS

J. Mass Spectrom. 2014, 49, 819–825 Copyright © 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jms

823

Table

4.Quan

tificationoftenBAsin

commercial

meatsamplesbytheHPLC-M

S/MS(triple

quad

rupole)method

.Theconcentrationvalues

areexpressed

inmgkg

�1

Biogen

icam

ines

12

34

56

78

910

1112

1314

1516

SPD

0.66

0.56

1.64

1.03

0.98

0.93

1.09

0.80

1.18

1.86

2.88

0.83

0.78

2.86

1.43

8.00

CAD

nda

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

PUT

nd

nd

nd

nd

nd

0.12

0.40

0.21

0.36

0.17

0.75

0.12

0.27

nd

nd

nd

HIS

0.89

0.67

1.49

1.01

1.23

0.59

0.57

0.47

0.69

0.83

0.55

0.36

0.51

0.23

0.37

0.04

AGM

nd

nd

nd

nd

nd

nd

nd

nd

nd

0.15

nd

nd

nd

nd

nd

nd

SPE

3.28

8.23

11.35

16.21

8.58

10.31

6.19

10.93

20.93

7.65

11.43

11.41

11.07

22.22

9.30

19.40

TYR

nd

nd

nd

nd

nd

nd

nd

nd

nd

0.29

nd

nd

0.01

nd

nd

nd

SER

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

PHE

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.03

0.01

0.01

0.01

0.01

0.01

0.01

TRY

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

Total

4.84

9.47

14.49

18.26

10.80

11.96

8.26

12.42

23.17

10.98

15.62

12.73

12.65

25.32

11.11

27.45

BAIb

0.89

0.67

1.49

1.01

1.23

0.71

0.97

0.68

1.05

1.29

1.30

0.48

0.79

0.23

0.37

0.04

SPD/SPE

c0.20

0.07

0.14

0.06

0.11

0.09

0.18

0.07

0.06

0.24

0.25

0.07

0.07

0.13

0.15

0.41

and=notdetectable,thevalueofconcentrationislower

than

thelim

itofdetection(LOD).

bTh

eBAI=

biogen

icam

ineindex,isthesum

ofconcentrationsofputrescine,cadaverine,histaminean

dtyramine.

c TheSP

D/SPE

=spermidine/spermineindex,istheratiobetweentheconcentrationsofspermidinean

dspermine.

V. Sirocchi et al.

824

ratios obtained with standards containing the compounds ofinterest at low concentration levels. In particular, in the caseof the HPLC-MS/MS method (IT), the LODs were in the range of0.03–0.3mg l�1 and LOQs were in the range of 0.1–1mg l�1,while in the case of the HPLC-MS/MS (TQ), the LODs were inthe range of 0.002–0.1mg l�1 and the LOQs were in the rangeof 0.008–0.5mg l�1.In the preliminary experiments of recovery using the IT as mass

analyzer, trichloroacetic extracts of meat were put in the refriger-ator at �20 °C for 24 h before the SPE step in order to eliminatelipidic substances that can interfere with the HPLC-MS/MSanalysis. Unfortunately, in this way the recovery percentagesobtained for most of the BAs were too poor and we decidednot to use this purification step.One of the main problems of MS/MS was that the presence of

matrix components can affect the ionization of the analytes,reducing or enhancing their response compared to standards insolvents. For this reason we evaluated how the matrix affectsthe analysis of BAs in meat. In our case, especially for spermidine,putrescine, agmatine and tyramine, the matrix effect showedsignal suppression in a range of 12–32%, while cadaverine,spermine and phenylethylamine showed enhancement of thesignal in a range of 4–50%.

Application to real samples

The selected and validated method using the TQ as mass analyzerwas applied to analyze the BA content in 16 commercial samplesof fresh meat purchased from various local butchers in Camerino,Italy. Results are reported in Table 4. Samples from n.1 to n.13were beef, while samples n.14, 15 and 16 were chicken, lamband rabbit, respectively. Total content of BAs found in thesamples ranged from 4.84 to 27.45mg kg�1, in line with previousdata reported in the literature for fresh meat.[18] Spermidine,spermine and histamine were present in all analyzed samples;in particular, the highest level of spermidine was found in samplen. 16 (rabbit meat, 8.00mg kg�1), while the highest content ofhistamine was in sample n. 4 (beef meat, 1.01mg kg�1).Putrescine was found in 8 of the 16 samples in concentrationsranging from 0.12 to 0.75mg kg�1; agmatine was found only insample n.10, while tyramine was found in samples n.10 and 13.Cadaverine, serotonine and tryptamine were not found in thefresh meat samples, while phenylethylamine was found in allsamples at low concentrations ranging from 0.01 to 0.03mg kg�1.A number of authors have proposed various chemical indices forevaluating fresh meat quality according to the content of BAs.Hernàndez-Jover et al. proposed calculating the BAI, which isthe sum of concentrations of putrescine, cadaverine, histamineand tyramine expressed in mg kg�1.[19] They established thefollowing values as indicators of meat quality: BAI <5mg kg�1

indicates that the meat is of good quality; the range of5–20mg kg�1 indicates that the meat is acceptable but withinitial signs of spoilage; the range of 20–50mg kg�1 indicatespoor quality, and >50mg kg�1 indicates that meat is spoiled. Inour case, all analyzed samples had BAI values lower than 1.49mgkg�1, indicating their good quality. Another proposed index formeat freshness uses the ratio between the concentrations ofspermidine and spermine (SPD/SPE). Silva et al.[2] reported thatif the SPD/SPE ratio is lower than 0.50, the meat sample can beconsidered fresh, while values in the range of 0.50–0.70 indicatethat the product must be consumed immediately, and valueshigher than 0.70 denote that the product is in an advanced stage

wileyonlinelibrary.com/journal/jms Copyright © 2014 Jo

hn Wiley & Sons, Ltd. J. Mass Spectrom. 2014, 49, 819–825

Underivatized biogenic amines by HPLC-MS/MS

of deterioration. The advantage of using their index is that itdepends only on the levels of polyamines produced by microor-ganisms and is independent of the specific type of bacterialflora.[19,20] In our case, the SPD/SPE ratio for all analyzed sampleswas lower than 0.50, (the highest value was 0.41, for sample n.16)confirming the high quality of these local meat samples.

Conclusions

An analytical method for determining ten underivatized BAs inmeat by HPLC-MS/MS has been developed. Comparison betweenperformance of the IT and TQ as mass analyzers in the study ofBAs in meat samples showed that the TQ is unequivocallysuperior because of its greater sensitivity and selectivity. Thismethodology for determining the BAI and SPD/SPE qualityindices is quick and highly sensitive, and has potential forquality control applications along the entire production chain ofmeat products.

Acknowledgements

The authors would like to thank Paola Beltrami (School ofPharmacy, University of Camerino) for helping in samplepreparation and HPLC-MS/MS analysis. They would also liketo thank Sheila Beatty for her editing of the English usage inthe manuscript.

References[1] V. Sirocchi, G. Caprioli, C. Cecchini, M. M. Coman, A. Cresci, F. Maggi,

F. Papa, M. Ricciutelli, S. Vittori, G. Sagratini. Biogenic amines asfreshness index of meat wrapped in a new active packaging systemformulated with essential oils of Rosmarinus officinalis. Int. J. FoodSci. Nutr. 2013, 64, 921.

[2] A. Önal, S. E. K. Tekkeli, C. Önal. A review of the Liquidi chromato-graphic methods for the determination of biogenic amines in foods.Anal. Methods 2013, 138, 509.

[3] C. M. G. Silva, M. B. A. Gloria. Bioactive amines in chicken breast andthigh after slaughter and during storage at 4±1 °C and in chicken-based meat products. Food Chem. 2002, 78, 241.

[4] R. G. K. Leuschner, A. Hristova, T. Robinson, M. Hugas. The rapid alertsystem for food and feed (RASFF) database in support of risk analysisof biogenic amines in food. J. Food Compos. Anal. 2013, 29, 37.

[5] F. B. Erim. Recent analytical approaches to the analysis of biogenicamines in food samples. Trends Anal. Chem. 2013, 52, 239.

[6] C. Ruiz-Capillas, F. Jimenez-Colmenero. Biogenic Amines in Meat andMeat Products. Crit. Rev. Food Sci. 2004, 44, 489.

J. Mass Spectrom. 2014, 49, 819–825 Copyright © 2014 John W

[7] S. Moret, D. Smela, T. Populin, L. Conte, A survey on free biogenicamine content of fresh and preserved vegetables. Food Chem.2005, 89, 355.

[8] C. C. Balamatsia, E. K. Paleologos, M. G. Kontominas, I. N. Savvaidis.Correlation between microbial flora, sensory changes and biogenicamines formation in fresh chicken meat stored aerobically or undermodified atmosphere packaging at 4 _C: Possible role of biogenicamines as spoilage indicators. Antonie van Leeuwenhoek. Int. J.Gen. Mol. Microbiol. 2006, 89, 9.

[9] G. Sagratini, M. Fernández-Franzón, F. De Berardinis, G. Font, S.Vittori, J. Mañes. Simultaneous determination of eight underivatizedbiogenic amines in fish by solid phase extraction and liquidchromatography-tandem mass spectrometry. Food Chem. 2012,132, 537.

[10] A. Pena-Gallego, P. Hernandez-Orte, J. Cacho, V. Ferreira. Biogenicamine determination in wines using solid-phase extraction: Acomparative study. J. Chromatogr. A 2009, 1216, 3398.

[11] C. M. Mayr, P. Schieberle. Development of Stable Isotope DilutionAssays for the Simultaneous Quantitation of Biogenic Amines andPolyamines in Foods by LC-MS/MS. J. Agric. Food Chem. 2012,60, 3026.

[12] S. Jia, Y. P. Kang, J. H. Park, J. Lee, S. W. Kwon. Simultaneous determi-nation of 23 amino acids and 7 biogenic amines in fermented foodsamples by liquid chromatography/quadrupole time-of-flight massspectrometry. J. Chromatogr. A 2011, 1218, 9174.

[13] S. Jia, Y. P. Kang, J. H. Park, J. Lee, S. W. Kwon. Determination ofbiogenic amines in Bokbunja (Rubus coreanusMiq.) wines using anovel ultra-performance liquid chromatography coupled withquadrupole-time of flight mass spectrometry. Food Chem. 2012,132, 1185.

[14] F. Gosetti, E. Mazzucco, V. Gianotti, S. Polati, M. C. Gennaro. Highperformance liquid chromatography/tandem mass spectrometrydetermination of biogenic amines in typical Piedmont cheeses.J. Chromatogr. A 2007, 1149, 151.

[15] F. Calbiani, M. Careri, L. Elviri, A. Mangia, L. Pistarà, I. Zagnoni. Rapidassay for analyzing biogenica amines in cheese: matrix solid-phasedispersion followed by liquid chromatography-electrospray-tandemmass spectrometry. J. Agric. Food Chem. 2005, 53, 3779.

[16] S. Millàn, M. C. Sampedro, N. Unceta, M. A. Goicolea, R. J. Barrio.Simple and rapid determination of biogenic amines in wine by liquidchromatography-electrospray ionization ion trap mass spectrometry.Anal. Chim. Acta 2007, 584, 145.

[17] M. R. Häkkinen, T. A. Keinänen, J. Vepsäläinen, A. R. Khomutov, L.Alhonen, J. Jänne, S. Auriola. Analysis of underivatized polyaminesby reversed phase chromatography with tandem mass spectrometry.J. Pharmaceut. Biomed. 2007, 45, 625.

[18] G. Vinci, M. L. Antonelli. Biogenic amines: quality index of freshnessin red and white meat. Food Control 2002, 13, 519.

[19] T. Hernàndez-Jover, M. Izquierdo-Pulido, M. T. Veciana-Nogues, M. C.Vidal-Carou. Biogenic amines sources in cooked cured shoulderpork. J. Agric. Food Chem. 1996, 44, 3097.

[20] A. R. Shalaby. Significance of biogenic amines to food safety andhuman health. Food Res. Int. 1996, 29, 675.

iley & Sons, Ltd. wileyonlinelibrary.com/journal/jms

825