Anal Bioanal Chem (2008) 391:26632672
DOI 10.1007/s00216-008-2167-9
ORIGINAL PAPER
Analysis of amino acids without derivatization in barley
extracts by LC-MS-MS
Bjrn Thiele & Kerstin Fllner & Nadine Stein &
Marco Oldiges & Arnd J. Kuhn & Diana Hofmann
Received: 16 November 2007 / Revised: 25 April 2008 / Accepted: 28 April 2008 / Published online: 28 May 2008
# Springer-Verlag 2008
Abstract A method has been developed for quantification
of 20 amino acids as well as 13 15N-labeled amino acids in
barley plants. The amino acids were extracted from plant
tissues using aqueous HClethanol and directly analyzed
without further purification. Analysis of the underivatized
amino acids was performed by liquid chromatography
(LC)electrospray ionization (ESI) tandem mass spectrom-
etry (MS-MS) in the positive ESI mode. Separation was
achieved on a strong cation exchange column (Luna 5
SCX 100) with 30 mM ammonium acetate in water
(solvent A) and 5% acetic acid in water (solvent B).
Quantification was accomplished using d2-Phe as an
internal standard. Calibration curves were linear over the
range 0.550 M, and limits of detection were estimated to
be 0.13.0 M. The mass-spectrometric technique was
employed to study the regulation of amino acid levels in
barley plants grown at 15 C uniform root temperature (RT)
and 2010 C vertical RT gradient (RTG). The LC-MS-MS
B. Thiele (*)
Institute for Chemistry and Dynamics of the Geosphere,
Institute-3: Phytosphere/BioSpec,
Forschungszentrum Jlich GmbH,
52425 Juelich, Germany
e-mail: b.thiele@fz-juelich.de
K. Fllner : A. J. Kuhn
Institute for Chemistry and Dynamics of the Geosphere,
Institute-3: Phytosphere,
Forschungszentrum Jlich GmbH,
52425 Juelich, Germany
N. Stein : M. Oldiges
Institute of Biotechnology 2,
Forschungszentrum Jlich GmbH,
52425 Juelich, Germany
D. Hofmann
Central Division of Analytical Chemistry/BioSpec,
Forschungszentrum Jlich GmbH,
52425 Juelich, Germany
results demonstrated enhanced concentration of free amino
acids in shoots at 2010 C RTG, while total free amino
acid concentration in roots was similarly low for both
RT treatments. 15NO3 labeling experiments showed lower
15 14
N/ N ratios for Glu, Ser, Ala and Val in plants grown at
2010 C RTG compared with those grown at 15 C RT.
Keywords Liquid chromatographymass spectrometry .
Amino acids . Strong cation exchange column .
Plant metabolism
Introduction
Amino acids as one main class of primary plant metabolites
play not only a key role in the fixation of nitrogen from
nitrate but also in intracellular and intercellular nitrogen
transport and in the biosynthesis of numerous other
metabolites [14]. Plant reactions to biotic and abiotic
factors cause changes in amino acid levels [5]. Therefore,
plant physiology investigations of the stress-induced
regulation of amino acid levels afford powerful analytical
methods. Analysis of amino acids has been commonly
performed by precolumn derivatization with ortho-
phthaldialdehyde and/or 9-fluorenylmethyl chloroformate
followed by reversed-phase high-performance liquid chro-
matography (HPLC) and fluorescence detection [68].
Spectrophotometric detection systems, however, are un-
suitable for the simultaneous determination of 14N- and
15
N-labeled amino acids, which is essential for 15N
labeling experiments. For that reason only mass spectrom-
etry (MS) can be considered as a detection system.
Numerous methods have been published for the determi-
nation of amino acids using gas chromatography (GC)
MS [912], capillary electrophoresis (CE)MS [1315] or
liquid chromatography (LC)MS [9, 1620]. GC affords
derivatization of the amino acids, which has been
2664
effectively performed with chloroformates [21, 22]. Al-
though this GC method shows high sensitivity and
excellent resolution it has the drawback that a laborious
manual derivatization procedure is necessary and the
method fails in the determination of Arg. The advantage
of CE-MS and LC-MS is the fact that all nonvolatile polar
amino acids can be analyzed without derivatization. A
CEelectrospray ionization (ESI)MS method has been
developed for determination of free amino acids using an
electrolyte at low pH [14, 15]. Short analysis times, good
selectivity and almost no matrix interferences have been
reported for this CE method. Reversed-phase LC is
commonly used to analyze nonvolatile components in
biological matrices; however, polar low molecular weight
compounds such as hydrophilic amino acids are not
sufficiently retained and consequently alternatives are
required. Ion-pairing reagents such as perfluorinated
carboxylic acids have been presented to improve the
separation of amino acids on C18 columns and for use in
LC coupled with ESI-MS [1619]; however, the use of
ion-pairing reagents in combination with ESI-MS might
lead to problems such as ion suppression, memory effects
and contamination of the MS source [23, 24]. Hydrophilic
LC columns in combination with hydrophobic mobile phases
offer an opportunity for separation of polar analytes. This
variant of LC was termed hydrophilic interaction LC (HILIC)
[25]. HILIC-ESI-MS-MS was successfully employed to
separate and quantify amino acids in different matrices [26,
27]. A further alternative is the separation on ion-exchange
columns, which has been carried out only with electrochem-
ical detectors so far [28, 29].
Within the scope of research on the response of barley
(Hordeum vulgare Barke) to spatially heterogeneous soil
temperatures, the regulation of amino acid levels was
studied at different root temperatures. It has recently been
shown that plants grown at a vertical gradient in soil
temperature (2010 C from the top to the bottom of the
plant pot) developed faster and produced more biomass
within a certain time frame (30 days) than plants grown at
uniform root temperatures (10, 15 and 20 C) [30]. To cast
light on possible mechanisms involved in the beneficial
effect of a vertical gradient in soil temperature on plant
development and growth, plant nitrogen metabolism was
looked at more closely as it is known that temperature
influences enzymatic activity. Results of free amino acids as
well as 15N-labeled amino acids (15NO3 labeling experi-
ments) in barley shoots and roots grown at 15 C uniform
root temperature and 2010 C vertical gradient in root
temperature are presented here.
For these studies we have developed a method of strong
cation exchange LC for use with ESI-MS-MS which
enables the simultaneous determination of 20 14N-labeled
and 13 15N-labeled amino acids without derivatization.
Anal Bioanal Chem (2008) 391:26632672
Optimization of the mobile phase and column temperature
is reported as well as multiple reaction monitoring (MRM)
of each single amino acid in order to enable sensitive and
selective detection.
Materials and methods
Reagents and material
All amino acids (Table 1) as well as DL-phenylalanine-
,-d2 (d2-Phe) were purchased from Sigma-Aldrich
(Taufkirchen, Germany). All 15N-labeled amino acids
(Table 2) were donated by M. Gehre, UFZ Leipzig,
Germany. Ammonium acetate (Fractopur), acetic acid
(100%, Suprapur), ethanol (LiChrosolv) and hydrochloric
acid (25%, ultrapure) and all salts of the nutrient solution
[p.a., 2.5 mM KCl, 2.5 mM NaNO3, 2.5 mM Ca(NO3)2,
1.0 mM MgSO4, 0.5 mM KH2PO4, 0.5 mM trace elements
(MnCl2, CuSO4, ZnSO4, H3BO3, NaMoO4) and 0.5 mM
Fe-EDTA] were obtained from Merck (Darmstadt, Ger-
many). Na15NO3 was purchased from ISOTEC (St. Louis,
MO, USA). Water obtained from a Milli-Q-Plus ultrapure
water purification system (Millipore, Bedford, MA, USA)
was used to prepare all samples and mobile phases.
Plant growth and treatment
Barley plants were germinated and grown at 15.00.5 C
uniform root temperature and 20.010.01.0C (top to
bottom of the plant pot) vertical gradient in root tempera-
ture in specially designed temperature boxes [30]. Shoot
temperature (22.01.0 C) and light intensity (photon flux
density of 240250 mol m-2 s-1 at 30 cm above the
substrate surface) were the same for both temperature
treatments. Sand of defined substrate density (1.65 g cm-3)
was used as a growth medium. A constant water content of
10.4 wt% in average and constant nutrient supply was
adjusted in the plant pots. The nutrient solution used was
added by continuous percolating drip irrigation (40 ml h-1).
Plants were harvested 30 days after germination. Root
and shoot were immediately separated by scalpel. Roots
were additionally separated into four horizontal layers each
of 5-cm depth allowing analysis of root segments of
different depths. The biomass fresh weight of shoots, roots
from different depths and the second leaf developed was
determined. The second developed leaf of each plant as
well as about 100 mg root from 05-cm and 510-cm depth
were frozen in liquid nitrogen and stored at -80 C for
further analyses, such as of amino acids. Two plants from
each temperature treatment were sampled and analyzed.
The same procedure was performed with plants spiked
with 15N tracer (Na14NO3 of Hoagland solution substituted
Anal Bioanal Chem (2008) 391:26632672 2665

by 98% 15N). The experimental setting was the same as
described above. Labeled nutrient solution (50 ml) was
added before harvest (after 30 days of plant growth) for 9 h
in total. The tracer (50 ml) was added 1 cm below the
substrate surface from the top by a syringe regularly
distributed around the shoot allowing tracer uptake of each
part of the root system. This treatment was conducted three
times in 3-h intervals.
ethanol (1:1, v/v) and 10 l of 1.0 mM d2-Phe as an internal
standard were added to the leaf powder and grinding was
continued with the pestle. The homogenate was transferred
to an Eppendorf vial with additional 0.5 ml extraction
solvent. After centrifugation at 14,000 rpm for 15 min at
4 C, the supernatant was directly used for analysis.
LC-ESI-MS-MS analysis

Sample preparation LC-ESI-MS-MS analysis was performed with an Agilent

The whole process of homogenization of leaves was carried
out with tools chilled with liquid nitrogen prior to use.
Approximately 100 mg of a leaf part was weighed in a
mortar and homogenized to a fine powder in liquid nitrogen
by use of a pestle. A 0.5-ml aliquot of 0.05 M aqueous HCl
(Santa Clara, CA, USA) 1100 series HPLC instrument
equipped with a binary pump, a vacuum degasser system
and a thermostated column oven. The LC system was
coupled with a Thermo Electron (Waltham, MA, USA)
TSQ Quantum triple quadrupole mass spectrometer. The
mass spectrometer was operated in the positive ESI mode.

Table 1 Precursor and product ions for liquid chromatographyelectrospray ionization tandem mass spectrometry (LC-ESI-MS-MS) analysis of
20 underivatized amino acids and d2-Phe (internal standard) with their optimized values for collision energy and tube lens voltage

Compound Precursor ion [M+H]+ (m/z) Product ion (m/z) Collision energy (eV) Tube lens (V)

Asp 134 74 [HO2C-CH=NH2]+ 22 151

Glu 148 84 [C4H6NO]+a 22 158
Ser 106 60 [R-CH=NH2]+ 20 131
Asn 133 74 [HO2C-CH=NH2]+ 22 158
Thr 120 74 [R-CH=NH2]+ 16 160
Gln 147 84 [C4H6NO]+a 24 151
Tyr 182 136 [R-CH=NH2]+ 20 157
Gly 76 30 [R-CH=NH2]+ 16 140
Pro 116 70 [C4H8N]+b 20 165
Ala 90 44 [R-CH=NH2]+ 14 168
Met 150 104 [R-CH=NH2]+ 16 156
Val 118 72 [R-CH=NH2]+ 16 151
Phe 166 120 [R-CH=NH2]+ 20 180
d2-Phe 168 122 [R-CH=NH2]+ 18 152
Leu 132 86 [R-CH=NH2]+ 14 170
Ile 132 86 [R-CH=NH2]+ 14 170
Trp 205 146 [M + H 59]+ 24 180
CC 241 152 [M + H 89]+ 20 155
His 156 110 [R-CH=NH2]+ 20 145
Lys 147 84 [C5H10N]+c 24 166
Arg 175 70 [C4H8N]+b 28 156
a

O N
H
m/z 84
b

N
H
m/z 70
c

N
H
m/z 84
2666 Anal Bioanal Chem (2008) 391:26632672

15
Table 2 Precursor and product ions for LC-ESI-MS-MS analysis of 13 underivatized N-labeled amino acids with their optimized values for
collision energy and tube lens voltage

Compound Parent ion [M+H]+ (m/z) Product ion (m/z) Collision energy (eV) Tube lens (V)
15
N-Glu 149 85 [C4H615NO]+a 20 138
15
N-Ser 107 61 [R-CH=15NH2]+ 16 113
15
N-Gln 148 84 [C4H6NO]+a 24 150
15
N-Tyr 183 137 [R-CH=15NH2]+ 10 169
15
N-Gly 77 60 [M + H 15NH2]+ 10 121
15
N-Ala 91 45 [R-CH=15NH2]+ 18 193
15
N-Met 151 92 [M + H 59]+ 18 117
15
N-Val 119 73 [R-CH=15NH2]+ 16 142
15
N-Phe 167 121 [R-CH=15NH2]+ 18 143
15
N-Leu 133 87 [R-CH=15NH2]+ 16 156
15
N-Ile 133 87 [R-CH=15NH2]+ 16 156
15
N-Lys 148 131 [M + H 15NH2]+ 10 139
15
N2-Arg 177 70 [C4H8N] b 36 150
a

O N
H
m/z 84 and 85
b

N
H
m/z 70

The ionspray voltage was set at 4 kV. Nitrogen was used as
the sheath gas (25 psi) and the auxiliary gas (1.0 arbitrary
units). The ion transfer capillary was heated to 300 C.
Injections were carried out using an HTC Pal autosampler
(CTC Analytics, Zwingen, Switzerland) equipped with a
20-l sample loop. MRM was used for specific quantifica-
tion of the amino acids and the internal standard applying a
scan time of 0.4 s. Argon was used as the collision gas at a
pressure of 1.0 mTorr. MS-MS parameters of the amino
acids were determined by flow injection analysis of amino
acid standard solutions using the inbuilt syringe pump. The
amino acid standard solutions were infused in the flow of
the HPLC system via a tee piece under the following
conditions: flow rate of the syringe pump, 5 l min-1; flow
rate of the HPLC system, 0.2 ml min-1 (30 mM ammonium
acetate5% acetic acid, 12.5:87.5).
LC separations were carried out with a Phenomenex
(Torrance, CA, USA) Luna 5 SCX 100 column,
150 mm2.0-mm internal diameter, at 40 C. For elution
of the amino acids 30 mM ammonium acetate (pH 6.0,
adjusted with 100% acetic acid) (solvent A) and 5% acetic
acid (solvent B) was applied. Two consecutive isocratic
elution steps were programmed as follows: starting with a
solvent A to solvent B ratio of 12.5:87.5, we held the
concentration for 42 min. Then it was changed to 100%
solvent A within 2 min and kept constant for 19 min. Within
2 min it was changed to the initial composition and kept
constant for 10 min. The overall flow rate was adjusted to
0.2 ml min-1. Prior to use, the new SCX column was flushed
with 150 mM ammonium acetate solution overnight.
A WTW(Weilheim, Germany) inoLab pH meter was
used for measuring the pH of the mobile phases.
Method validation
Calibration curves were constructed by analysis of standard
solutionsprepared both in solvent and in matrix solution
at four (0.550 M) and three (550 M) concentrations,
respectively. Instrumental detection limits (IDLs) corre-
sponded to a concentration with a signal-to-noise ratio of 3.
Method accuracy and precision were evaluated by
analyzing unspiked and spiked barley leaf samples at
concentrations equivalent to the natural ones. Leaf samples
were spiked after homogenization with a mixed standard
working solution prepared in 5% acetic acid, then left to
stand for 24 h at -80 C before extraction. The method
detection limit (MDL) was defined as the lowest concen-
tration of the analyte in a sample matrix resulting in a
signal-to-noise ratio of 3.
Calibration and quantification
Two 200 M stock solutions of a mixture of 20 amino acids
and d2-Phe were used for the preparation of calibration
standards with concentrations of 5, 20 and 50 M each
containing 10 M d2-Phe as an internal standard. These
Anal Bioanal Chem (2008) 391:26632672 2667

standards were used to generate response factors fi in
relation to the internal standard. Quantification of the 20
amino acids in barley leaf extracts was carried out by the
internal standard method (spiking of the leaf samples with
d2-Phe as described above) using the response factors.
The areas of those peaks with the same retention times as
the calibration standard as well as concurrent MS-MS
spectra were integrated and subsequently used for
calculation of the concentrations of each individual amino
acid.
Calibration of 13 15N-labeled amino acids was also
carried out with 5, 20 and 50 M calibration standard
solutions each containing 10 M d2-Phe. The peak area of
d2-Phe had to be corrected because of its coelution with
isobaric 13C,15N-Phe. The percentage of 13C,15N-Phe in
15
N-Phe was determined by LC-MS-MS analysis.
Results and discussion
LC-MS-MS analysis
Detection of the amino acids in full-scan and single ion
monitoring mode caused severe problems because of strong

Fig. 1 Selected ion chromato-

grams of 20 amino acids (20 M Asp
134 74
each) and d2-Phe (10 M)
Glu
obtained in multiple reaction Glu
148 84
monitoring mode. The numbers
on the right are m/z ions of Q1 Ser
(protonated precursor ion) and 106 60
Q3 (product ion) in multiple
reaction monitoring for each Asn
133 74
analyte
Thr
120 74

Gln
147 84

Tyr
182 136

Gly
76 30

Pro
116 70

Ala
90 44

Met
150 104

Val
118 72

Phe
166 120

d2-Phe
168 122
Ile
Leu Leu/Ile
132 86

Trp
205 146

C-C
241 152

His
156 110

Lys
147 84

Arg
175 70

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
Time (min)
2668 Anal Bioanal Chem (2008) 391:26632672

signals from the mobile phase and the sample matrix. times in the LC-MS system were evaluated by 20 sample
Therefore, specific MS-MS transitions for each amino acid measurements (calibration standards and biological sam-
had to be determined also in order to achieve sufficient ples). The relative standard deviations (RSDs) were lower
sensitivity for their quantification. Selection criteria for the than 1.0% in most cases, except for Trp (1.3%) and CC
MS-MS transitions were specificities and intensities of the (2.1%), which were eluted last in the first isocratic elution
product ions. Tables 1 and 2 show the collision induced step. The basic amino acids showed stable retention times
dissociation (CID) product ions of 20 amino acids and 13 with only 0.1% RSDs again. Replacement of the SCX
15
N-labeled amino acids together with their corresponding column, however, led to significant drifts in retention times
collision energies obtained in positive ESI mode. The most up to 4 min in case of CC. The retention-time drifts could
prominent fragmentation of amino acids was the loss of be minimized by flushing the column with 150 mM
H2O and CO leading to abundant immonium ions [R ammonium acetate overnight. Prior to data processing, an
CH=NH2]+, where R is the residue of the amino acid [17, adjustment of retention times in the processing file for
18]. Asp and Asn preferably formed the immonium ion automatic identification and integration of the peaks was
[HO2C-CH=NH2]+ (m/z 74) under elimination of H2O and performed.
CH2CO. The CID mass spectra of Glu, Gln and the basic
amino acids Arg and Lys showed intensive cyclic product Method validation
ions as is also well known from mass spectra of these amino
acids obtained by electron impact ionization [31]. In these The performance of the analytical method was evaluated
cases the loss of the functional group at the -position (Glu, with regard to linearity, accuracy, precision and limits of
Gln, Arg) and the -position (Lys), respectively, initiated a detection.
reaction sequence leading to stable pyrrolidone ions and
cyclic immonium ions, respectively (Tables 1, 2, footnotes).
Experiments with labeled 15N-Glu (-15NH2) and 15N-Gln
Table 3 Calibration linearity (R2), precision and instrumental
(CO15NH2) proved that the nitrogen atoms of the -amino detection limit (IDL) for amino acids dissolved in water + 5% acetic
groups were incorporated in the rings. Our observations of acid
CID fragmentations of protonated amino acids are consis-
Compound R2a Intraday Interday IDL
tent with those reported in the literature [17, 32].
RSD (%)b RSD (%)b (M)c
For chromatographic separation of the amino acids we
chose a strong cation exchange column owing to the fact Asp 0.9996 4.3 9.0 1.0
that all amino acids exist in their ammonium form at pH<3. Glu 0.9988 4.4 8.6 0.3
A chromatographic profile of a solution of 20 amino acids Ser 0.9960 7.3 7.7 1.0
and d2-Phe as an internal standard on the Luna 5 SCX Asn 0.9943 6.2 9.6 0.2
Gln 0.9984 11.7 16.7 0.3
100 column is reported in Fig. 1. Elution of the amino
Tyr 0.9994 3.2 6.8 0.3
acids was performed in two consecutive isocratic steps. Gly 0.9963 18.8 12.9 3.0
During the first 42 min at pH 2.5, acidic and neutral amino Pro 0.9996 1.9 4.0 0.2
acids were separated. After adjustment to pH 6.0, separa- Ala 0.9958 4.1 8.0 0.9
tion of basic amino acids was achieved. The concentration Met 0.9987 5.8 10.0 0.3
of NH4+ in the mobile phase had a strong impact on the Val 0.9992 3.5 7.1 0.1
retention times of all amino acids. A decrease of NH4+ Thr 0.9995 6.0 4.4 0.1
Phe 0.9992 1.9 3.3 0.1
concentration by 50% caused an increase of the retention
Leu 0.9947 6.1 17.6 0.2
time of, e.g., Trp by 20 min. The column temperature was a
Ile 0.9989 4.1 16.8 0.2
second effective variable that shortened analysis times. An Trp 0.9998 4.4 9.7 0.1
increase of the column temperature from 20 to 40 C CC 0.9997 3.0 15.7 0.1
caused a decrease of the retention time of d2-Phe from 21.8 His 0.9976 14.0 11.5 0.3
to 18.1 min. The resolution of the chromatographic Lys 0.9981 17.5 15.0 0.3
separations remained constant upon reduction of the Arg 0.9978 10.0 13.2 0.3
retention times, with peak widths at half height b1/2 in the a
Calibration curves were determined in the concentration range 0.5
range from 0.25 (Asp) to 0.95 min (CC) . According to 50 M.
b
Rey and Pohl [33] the sulfonic groups of the benzene The precision is indicated as the relative standard deviation (RSD)
sulfonic acid resin are less ionized at higher temperatures defined as the ratio of the standard deviation to the mean response
factor of each amino acid.
and therefore its cation-exchange capacity is lowered. For c
Extrapolated from calibration curves acquired in multiple reaction
that reason the column oven temperature was set to 40 C monitoring (MRM) mode. IDLs correspond to concentrations with
for all subsequent analyses. The stabilities of the retention signal-to-noise ratio of approximately 3.
Anal Bioanal Chem (2008) 391:26632672 2669

Table 4 Calibration linearity (R2), precision and IDL for 15N-labeled Linearity was studied for each amino acid and 15N-
amino acids dissolved in water + 5% acetic acid
labeled amino acid in the concentration range 0.550 M.
Compound R2a Intraday Interday IDL Linear regression analysis from calibration curves showed
RSD (%)b RSD (%)b (M)c correlation coefficients R2 between 0.9943 and 0.9999
15
(Tables 3, 4). Intraday precision expressed as the RSD of
N-Glu 0.9991 15.2 14.0 0.1
15 the response factors was found to be lower than 10% for 16
N-Ser 0.9999 7.2 8.7 0.1
15
N-Gln 0.9987 5.7 6.2 1.4 amino acids and nine 15N-labeled amino acids. For the
15
N-Tyr 0.9994 19.3 13.5 0.3 remaining compounds intraday precisions were below 19%.
15
N-Gly 0.9982 10.2 9.8 1.5 Ion suppression from buffer solutions caused poor signal-
15
N-Ala 0.9974 10.5 10.9 1.0 to-noise ratios for Gly, Lys, Arg, 15N-Glu and 15N-Tyr and
15
N-Met 0.9998 9.3 7.3 1.4 as a consequence bad precision. The interday precision of
15
N-Val 0.9996 5.7 6.1 0.2 most of the amino acids was less than 10%. In addition to
15
N-Phe 0.9999 9.9 11.2 0.2
15 the abovementioned amino acids CC and Leu/Ile showed
N-Leu 0.9991 7.0 8.1 0.3
15 RSDs of more than 10%. Leu/Ile were poorly separated
N-Ile 0.9987 5.6 6.3 0.2
15
N-Lys 0.9992 15.6 16.6 0.5 (Fig. 1), resulting in a less precise integration of the peaks.
15
N2-Arg 0.9993 19.3 15.1 0.1 The IDLs ranged generally between 0.1 and 0.3 M.
Higher IDLs for Asp, Ser, Gly and Ala (1.03.0 M) as
a
Calibration curves were determined in the concentration range 0.5 well as for 15N-Gly and 15N-Met (1.5 M) were attributed
50 M.
b
The precision is indicated as the RSD defined as the ratio of the to matrix effects. To study the influence of the biological
standard deviation to the mean response factor of each amino acid. matrix on linearity and MDLs, the supernatants of barley
c
Extrapolated from calibration curves acquired in MRM mode. IDLs leaf extractions were spiked with all amino acids at three
correspond to concentrations with signal-to-noise ratio of approxi- different levels (Table 5). Linear regression of the calibra-
mately 3.

Table 5 Calibration linearity (R2), method detection limit (MDL), precision and recovery for amino acids in barley leaf extracts

Compound R2a MDL Spike amount Barley leaf Barley leaf + spike
(M])b (mol/g)c
Average Interday Average Interday Spike
concentration RSD (%) concentration RSD (%) recovery
(mol/g)c (mol/g)c (%)d

Asp 0.9998 1.5 2.4 1.682 5.3 4.448 3.5 115.2

Glu 0.9987 0.6 4.8 4.573 3.7 9.431 5.6 101.2
Ser 0.9993 1.0 2.4 2.023 5.6 4.910 2.6 120.3
Asn 0.9991 0.2 0.6 0.080 8.8 0.585 3.6 84.1
Gln 0.9984 0.4 2.4 1.738 5.1 4.086 3.2 97.8
Tyr 0.9983 0.3 0.6 0.030 8.8 0.510 2.2 80.0
Gly 0.9978 3.0 2.4 2.393 5.6 4.656 4.4 94.3
Pro 0.9984 0.4 0.6 0.172 5.8 0.821 3.6 108.2
Ala 0.9998 0.9 2.4 1.382 5.0 3.645 4.5 94.3
Met 0.9998 0.6 0.6 <MDL 0.661 3.7 110.2
Val 0.9994 0.3 0.6 0.137 4.1 0.711 2.4 95.7
Thr 0.9998 0.4 2.4 0.731 4.5 3.463 2.4 113.8
Phe 0.9999 0.2 0.6 0.112 4.6 0.76 2.0 108.5
Leu 0.9987 0.3 0.6 0.102 4.3 0.681 4.2 96.4
Ile 0.9986 0.3 0.6 0.074 8.8 0.656 3.3 96.9
Trp 0.9996 0.1 0.6 0.022 9.4 0.504 3.3 80.3
CC 0.9998 0.1 0.6 <MDL 0.452 2.4 75.3
His 0.9988 0.6 0.6 0.348 4.0 0.768 8.8 70.0
Lys 0.9987 1.0 0.6 0.373 6.3 0.755 7.4 63.7
Arg 0.9988 0.7 0.6 0.288 7.3 0.784 3.5 82.7
a
Calibration curves were determined by spiking supernatants of barley leaf extractions in the concentration range 550 M.
b
Extrapolated from chromatograms acquired in MRM mode. MDLs corresponded to concentrations with signal-to-noise ratio of approximately 3.
c
Concentration with regard to grams of fresh weight of barley leaf homogenate
d
Recoveries based on three replicate analyses of spiked barley leaf homogenates on three different days
2670 Anal Bioanal Chem (2008) 391:26632672

Fig. 2 Concentrations of amino

free amino acid concentration [g g-1 fresh weight]

acids in shoots of barley plants
800
grown at 15 C root temperature 15 C
and 2010 C root temperature 20-10 C
gradient (A) and 15N/14N ratios 700
of amino acids in shoots of
barley plants grown at 15 C 600
root temperature and 2010 C A
root temperature gradient (B) 500

400

300

200

100

0
0.16

0.14
N/14N ratio of free amino acids

0.12 B
0.10

0.08

0.06

0.04
15

0.02

0.00
GLY

VAL
GLU
SER
ASN
GLN
TYR

PRO

LEU

ARG
ASP

ALA

MET

PHE

ILE
TRP
LYS
HIS
tion curves showed good linearities, with R2 >0.9983. Influence of root temperature on amino acid metabolism
MDLs were in the range 0.13.0 M. In comparison with in shoots and roots
IDLs, the MDLs were on the same order of magnitude,
indicating no significant effects of matrix compounds from Amino acid concentrations were analyzed in leaf and root
barley leaf extracts on the sensitivity. extracts of barley which had been grown at different root
Precision and accuracy of the method were verified by temperatures (n=2). Irrespective of root temperature treat-
measuring recovery from barley leaf samples spiked with ment, total free amino acid concentrations in shoots (here,
all amino acids at levels equivalent to natural ones. Leaf second developed leaf) of barley were 38 times higher
samples were spiked and frozen at -80 C for at least 24 h than in roots. The concentration of total free amino acids in
before thawing, preparing and analyzing them. Three the shoot for plants grown at 2010 C root temperature
separate aliquots of the unspiked/spiked leaf homogenates gradient was twice that for plants grown at 15 C root
were prepared and analyzed on three different days. The temperature. Therefore, most of individual free amino acids
interday precision of all amino acids was less than 10% attained higher concentrations in shoots of plants grown at
RSD for unspiked samples and less than 8% RSD for 2010 C root temperature gradient (Fig. 2, panel A).
spiked samples. Spike recoveries for most amino acids were Exceptions were Asp and Glu, both showing no concen-
within 80120%, with the exception of CC (75%), His tration differences between the two temperature treatments.
(70%) and Lys (64%). The result for Glu is at least in agreement with results of
Anal Bioanal Chem (2008) 391:26632672
Fritz et al. [5], who showed that the Glu concentration is
quite stable independent of the CN status of the plants. It
was hypothesized that the level of Glu is exclusively
regulated by 2-oxogluterate.
No differences of total free amino acid concentration
occurred between roots of different depths and temper-
atures. This was also valid for the individual free amino
acids (data not shown). The 15N/14N ratios of the individual
free amino acids in shoots revealed lower ratios for Glu,
Ser, Ala and Val in plants grown at 2010 C root tempe-
rature gradient compared with plants grown at 15 C root
temperature (Fig. 2, panel B). That means the fraction of 15N
tracer was lower in these amino acids at 2010 C root
temperature gradient compared with 15 C root temperature
after 9 h of labeling. The differences in 15N/14N ratios and in
concentrations of individual free amino acids in shoots
supported the hypothesis that a vertical gradient in root
temperature affects nitrogen metabolism. This might be due
to differences in nitrogen uptake rates or the amount of NO3
delivered to the shoot, since NO3 induces nitrate reductase
and therefore the synthesis of Glu as a key amino donor in
the synthesis of many other amino acids. However, further
research is necessary to study the mechanisms responsible
for the change in nitrogen metabolism under a root
temperature gradient.
The enhanced concentration of free amino acids in
shoots at 2010 C root temperature gradient correlates
with the high shoot dry weight production occurring at
this root temperature treatment (3 times higher than that at
15 C root temperature ): high free amino acid concen-
trations in shoots indicate a high ability to initiate growth of
new leaves [34], since most amino acids will be used locally
during vegetative growth [35]. Glu and Asp, however, are
important molecules in nitrogen transport and signaling
within the plant [36, 37], resulting in similar concentrations
in shoots independent of the root temperatures (Fig. 2, panel
A). Constant levels of Glu and Asp in the shoots and of total
free amino acids in roots might point to a quite stable amino-
nitrogen pool cycle through the plant [38].
However, further research on amino acids, their appear-
ance in phloem and xylem sap as well as in their use in
proteins has to be conducted, to cat light on explore the role
of nitrogen metabolism in stimulating plant development
and growth at vertical temperature gradients in soil.
Conclusion
In this study a new method for the analysis of 20
underivatized amino acids and 13 15N-labeled amino acids
by LC-ESI-MS-MS using a strong cation exchange column
was described. Matrix interferences with the target analytes
2671
in MS instrument were eliminated by use of the highly
specific MRM mode. With two consecutive isocratic
elution steps at pH 2.5 and 6.0 good separation of acidic/
neutral amino acids and basic amino acids, respectively,
was achieved. NH4+ concentration and column temperature
appeared to be critical for obtaining short retention times
and were optimized in this respect. The method was
successfully validated by determining linearity, sensitivity,
precision and accuracy. The LC-ESI-MS-MS technique
developed for measurement of amino acids was successful-
ly employed to study the regulation of amino acid levels in
barley plants at different root temperatures.
Acknowledgements The authors are grateful to Matthias Gehre
(Helmholtz Centre for Environmental Research, UFZ Leipzig,
Germany) for providing the 15N-labeled amino acids. We also wish
to thank Bernd Kastenholz for his assistance in the optimization of the
leaf extraction procedure.
References
1. Hewitt EJ (1975) Annu Rev Plant Physiol 26:73100
2. Guerrero MG, Vega JM, Losada M (1981) Annu Rev Plant
Physiol 32:169204
3. Miflin BJ, Lea PJ (1977) Annu Rev Plant Physiol 28:299329
4. Atkins C, Beevers L (1990) In: Abroyl YP (ed) Nitrogen in higher
plants. Wiley, New York, p 223
5. Fritz C, Mueller C, Matt P, Feil R, Stitt M (2006) Plant Cell
Environ 29:20552076
6. Roth M (1971) Anal Chem 43:880882
7. Einarsson S, Josefsson B, Lagerkvist S (1983) J Chromatogr
282:609618
8. Blankenship DT, Krivanek MA, Ackermann BL, Cardin AD
(1989) Anal Biochem 178:227232
9. Chace DH (2001) Chem Rev 101:445478
10. Husek P, Simek P (2006) Curr Pharm Anal 2:2343
11. Calder AG, Garden KE, Anderson SE, Lobley GE (1999) Rapid
Commun Mass Spectrom 13:20802083
12. Wood PL, Khan MA, Moskal JR (2006) J Chromatogr B
831:313319
13. Monton MRN, Soga T (2007) J Chromatogr A 1168:237246
14. Soga T, Kakazu Y, Robert M, Tomita M, Nishioka T (2004)
Electrophoresis 25:19641972
15. Soga T, Heiger DN (2000) Anal Chem 72:12361241
16. Chaimbault P, Petritis K, Elfakir C, Dreux M (1999) J Chromatogr
A 855:191202
17. Petritis K, Chaimbault P, Elfakir C, Dreux M (2000) J Chromatogr
A 896:253263
18. Qu J, Chen W, Luo G, Wang YM, Xiao SY, Ling Z, Chen GQ
(2002) Analyst 127:6669
19. Piraud M, Vianey-Saban C, Petritis K, Elfakir C, Steghens J-P,
Bouchu D (2005) Rapid Commun Mass Spectrom 19:15871602
20. Armstrong M, Jonscher K, Reisdorph NA (2007) Rapid Commun
Mass Spectrom 21:27172726
21. Husek P, Sweeley CC (1991) J High Resolut Chromatogr 14:751
753
22. Husek P (1991) J Chromatogr 552:289299
23. Bonfiglio R, King RC, Olah TV, Merkle K (1999) Rapid Commun
Mass Spectrom 13:11751185
2672
24. Hsieh Y, Chintala M, Mei H, Agans J, Brisson J-M, Ng K,
Korfmacher WA (2001) Rapid Commun Mass Spectrom 15:2481
2487
25. Alpert AJ (1990) J Chromatogr A 499:177196
26. Schlichtherle-Cerny H, Affolter M, Cerny C (2003) Anal Chem
75:23492354
27. Langrock T, Czihal P, Hoffmann R (2006) Amino Acids 30:291297
28. Welch LE, LaCourse WR, Mead DA, Johnson DC, Hu T (1989)
Anal Chem 61:555559
29. Luo P, Zhang F, Baldwin RP (1991) Anal Chem 63:17021707
30. Fllner K (2007) The influence of spatially heterogeneous soil
temperatures on plant structure and function. Dissertation,
Heinrich Heine University, Dsseldorf
Anal Bioanal Chem (2008) 391:26632672
31. Huang ZH, Wang J, Gage DA, Watson JT, Sweeley CC, Husek P
(1993) J Chromatogr 635:271281
32. Dookeran NN, Yalcin T, Harrison AG (1996) J Mass Spectrom
31:500508
33. Rey MA, Pohl CA (1996) J Chromatogr A 739:8797
34. Rufty TW, Raper CD, Jackson WA (1981) New Phytol 88:607619
35. Imsande J, Touraine B (1994) Plant Physiol 105:37
36. Aslam M, Travis RL, Rains DW (2001) Plant Sci 160:219228
37. Vidmar JJ, Zhuo D, Siddiqi MY, Schjoerring JK, Touraine B,
Glass ADM (2000) Plant Physiol 123:307318
38. Touraine B, Clarkson DT, Muller B (1994) In: Roy J, Garnier E
(eds) A whole plant perspective on carbon-nitrogen interactions.
SPB, The Hague