Simultaneous Analysis of Wine Biogenic Amines and Amino Acids
J. Agric. Food Chem., Vol. 55, No. 3, 2007 609
polyamines after derivatization with N-(9-fluorenylmethoxy-
carbonyloxy)-succinimide (FMOC-Su); however, the excess
reagent and its hydrolysis product (FMOC-OH) are still present
as majority peaks in the chromatogram, and the method does
not determine the aromatic biogenic amines histamine and
tyramine, which are the most important given their effects for
consumer health, or γ-aminobutyric acid (GABA), a charac-
teristic amino acid in grapes. Another simultaneous analysis
based on OPA derivatization quantifies 20 amino acids and 17
amines, but the reagent does not serve to determine proline,
quantitatively the most important amino acid in grape and wine
(14). A third simultaneous method requires dual OPA/FMOC
derivatization and also uses a fluorescence detector (15).
The present paper reports the development of a new method
for simultaneous analysis of biogenic amines, amino acids, and
ammonium ion in wines and beers. It is based on a method that
was initially developed for analysis of protein hydrolysates (16,
17) and other biological samples (18) and subsequently used to
analyze free amino acids in foods (19, 20) but had never hitherto
been used to analyze biogenic amines. It consists of reversed-
phase separation by HPLC and UV-vis detection of the
aminoenones formed by the reaction of amino acids, biogenic
amines, and ammonium ion with the derivatization reagent
diethyl ethoxymethylenemalonate (DEEMM).
Table 1. Eluent Gradient for HPLC Determination of Aminoenone
Derivatives of Amino Acids, Biogenic Amines, and Ammonium Ion
time (min)
eluent A (%) 90
eluent B (%) 10
0.0 20.0 30.5 33.5 65.0 73.0 78.0
82.0
0
100
85.0
0
100
90
10
83
17
83
17
60
40
28
72
18
82
min. For detection, a photodiode array detector monitored at 280, 269,
and 300 nm was used. In the proposed conditions, 34 compounds were
separated, identified, and quantified in a single injection: 24 amino
acids (plus the internal standard), the ammonium ion, and nine biogenic
amines.
The target compounds were identified according to the retention
times and UV-vis spectral characteristics of the derivatives of the
corresponding standards and were quantified using the internal standard
method. Detection limits were calculated according to the OIV Oeno
7/2000 (21) method as 3 times the baseline noise.
Statistical Analysis. Statistical analysis was performed using SPSS
12 statistical software (SPSS Inc., Chicago, IL).
RESULTS AND DISCUSSION
Method Development. On the basis of an original method
designed for the analysis of only free amino acids (19, 20), we
developed a new method that improved amino acid separation
and allowed us to extend the analysis to other nitrogenated
compounds. The amino acids asparagine and serine, which
coeluted in the original method, were successfully separated,
and a new compound, hydroxyproline, was also introduced with
the amino acids analyzed. At the same time, further changes
were necessary to separate the aminoenone derivatives of nine
biogenic amines in the same chromatogram. To do this, it was
necessary to extend the chromatographic running time from 40
to 85 min and also to change the composition of the mobile
phase B, which acts as an organic modifier, from 100%
acetonitrile to a mixture of 80% acetonitrile and 20% methanol.
With the analytical method developed, correct chromato-
graphic separation of 24 amino acids, nine biogenic amines,
and ammonium ion was successfully achieved (Figure 1). The
maximum absorption wavelengths in the UV of the aminoenone
derivatives of amino acids, biogenic amines, and ammonium
ion were found between 269 nm (aminoenone of the ammonium
ion) and 292 nm (aminoenones of proline and hydroxyproline),
and the intermediate maximum absorption wavelengths were
279-284 nm (aminoenones of the primary amino acids), 277-
278 nm (aminoenones of lysine and ornithine), and 278-280
nm (aminoenones of the polyamines or biogenic amines). On
the basis of these data, we selected 280 nm as the wavelength
for quantifying all the biogenic amines and most of the amino
acids, with the exception of asparagine, serine, and hydroxy-
proline (peaks 3, 4, and 5); as shown in Figure 1, these
displayed better separation at 300 nm since hydroxyproline only
appeared as a shoulder of serine in the chromatogram recorded
at 280 nm. The ammonium ion could also have been quantified
at 280 nm, but it was decided to use the response at 269 nm to
increase the intensity of its signal.
MATERIALS AND METHODS
Reagents. Super-gradient HPLC grade acetonitrile and methanol
were obtained from Labscan (Dublin, Ireland), and ultrapure water
generated by the Milli-Q system Millipore (Bedford, MA) was used.
L-Cysteine, L-leucine, L-phenylalanine, L-lysine, ammonium chloride,
L-histidine, agmatine sulfate, cadaverine, L-arginine, histamine,
L-proline, L-R-alanine, spermidine, glycine, â-alanine, L-aspartic acid,
L-glutamic acid, L-tyrosine, L-valine, and L-serine were from Fluka
Chemie (Buchs, Switzerland); isoamylamine, diethylethoxymethylen-
emalonate (DEEMM), putrescine, L-glutamine, tyramine, and trans-4-
hydroxy-L-proline were from Aldrich Chemie (Steinhein, Germany);
L-2-aminoadipic acid, L-ornithine monohydrochloride, L-tryptophan,
L-asparagine, L-threonine, γ-aminobutyric acid (GABA), L-isoleucine,
L-methionine, phenylethylamine, and sodium azide were from Sigma
Chemie (Steinhein, Germany). Solutions of amino acids and biogenic
amines were prepared with HCl 0.1 N.
Materials. Twenty-eight red wines and 14 white wines from the
2005 harvest were kindly supplied by 10 different wineries from the
Spanish region of Castilla-La Mancha. Beer was purchased in a retail
store.
Enzymatic Determination of Ammonium Ion. Ammonia was
determined with an enzymatic kit for the determination of urea and
ammonium ion in foodstuffs from Boehringer Mannhein/R-Biopharm
(Darmstadt, Germany). Wine samples were previously treated with
polyvinylpolypyrrolidone (PVPP) to avoid the interference of tannins
as specified by the manufacturer.
Reaction of Derivatization. Aminoenone derivatives were obtained
by reaction of 1.75 mL of borate buffer 1 M (pH ) 9), 750 µL of
methanol, 1 mL of target sample without any pretreatment, 20 µL of
internal standard (L-2-aminoadipic acid, 1 g/L), and 30 µL of DEEMM
in a screw-cap test tube over 30 min in an ultrasound bath. The sample
was then heated at 70 °C for 2 h to allow complete degradation of
excess DEEMM and reagent byproducts.
HPLC Analysis. The analyses were performed on a Varian ProStar
HPLC (Varian Inc., Walnut Creek, CA) comprising a ProStar 240
ternary pump, a ProStar 410 autosampler, and a ProStar 330 array
photodiode detector.
Chromatographic separation was performed in an ACE HPLC
column (5 C18-HL) particle size 5 µm (250 mm × 4.6 mm)
thermostatized at 16 °C in an MFE-01 oven (Ana´lisis V´ınicos,
Tomelloso, Spain) through the binary gradient shown in Table 1 (phase
A, 25 mM acetate buffer pH ) 5.8 with 0.02% sodium azide; phase
B, 80:20 mixture of acetonitrile and methanol) and a flow rate 0.9 mL/
This analytical procedure has various advantages over the
methods reported in the literature for simultaneous analysis of
amino acids and biogenic amines (13-15). An important feature
common to all these other methods is the use of fluorescent
derivatizing agents, which requires HPLC equipment with this
type of detector. However, the aminoenone derivatives formed
by reaction with DEEMM can be detected with a photodiode
array detector and also with an ultraviolet detector, the two most
common detectors in HPLC equipment. Some ultraviolet
detectors can be set at several simultaneous detection wave-