Profiling of thiol-containing compounds by stable isotope labeling double precursor ion scan mass spectrometry

Article abstract of DOI:10.1021/ac5023315

Here we developed a novel strategy of isotope labeling in combination with high-performance liquid chromatography-double precursor ion scan mass spectrometry (IL-LC-DPIS-MS) analysis for nontargeted profiling of thiol-containing compounds. In this strategy, we synthesized a pair of isotope labeling reagents (?-bromoacetonylquinolinium bromide, BQB; ?-bromoacetonylquinolinium-d₇ bromide, BQB-d₇) that contain a reactive group, an isotopically labeled moiety, and an ionizable group to selectively label thiol-containing compounds. The BQB and BQB-d₇ labeled compounds can generate two characteristic product ions m/z 218 and 225, which contain an isotope tag and therefore were used for double precursor ion scans in mass spectrometry analysis. The peak pairs with characteristic mass differences can be readily extracted from the two precursor ion scan (PIS) spectra and assigned as potential thiol-containing candidates, which facilitates the identification of analytes. BQB and BQB-d₇ labeled thiol-containing compounds can be clearly distinguished by generating two individual ion chromatograms. Thus, thiol-containing compounds from two samples labeled with different isotope reagents are ionized at the same time but recorded separately by mass spectrometry, offering good identification and accurate quantification by eliminating the MS response fluctuation and mutual interference from the two labeled samples. Using the IL-LC-DPIS-MS strategy, we profiled the thiol-containing compounds in beer and human urine, and 21 and 103 thiol candidates were discovered in beer and human urine, respectively. In addition, 9 and 17 thiol candidates in beer and human urine were successfully identified by further comparison with thiol standards or tandem mass spectrometry analysis. Taken together, the IL-LC-DPIS-MS method is demonstrated to be a promising strategy in the profiling of compounds with identical groups in metabolomics study.

Full text of DOI:10.1021/ac5023315

Article
pubs.acs.org/ac
Profiling of Thiol-Containing Compounds by Stable Isotope Labeling
Double Precursor Ion Scan Mass Spectrometry
Ping Liu, Yun-Qing Huang, Wen-Jing Cai, Bi-Feng Yuan,* and Yu-Qi Feng*
Key Laboratory of Analytical Chemistry for Biology and Medicine, Ministry of Education, Department of Chemistry, Wuhan
University, Wuhan 430072, People’s Republic of China
S
* Supporting Information
ABSTRACT: Here we developed a novel strategy of isotope
labeling in combination with high-performance liquid
chromatography−double precursor ion scan mass spectrome-
try (IL−LC−DPIS-MS) analysis for nontargeted profiling of
thiol-containing compounds. In this strategy, we synthesized a
pair of isotope labeling reagents (ω-bromoacetonylquinoli-
nium bromide, BQB; ω-bromoacetonylquinolinium-d7 bro-
mide, BQB-d7) that contain a reactive group, an isotopically
labeled moiety, and an ionizable group to selectively label
thiol-containing compounds. The BQB and BQB-d7 labeled
compounds can generate two characteristic product ions m/z
218 and 225, which contain an isotope tag and therefore were
used for double precursor ion scans in mass spectrometry
analysis. The peak pairs with characteristic mass differences can be readily extracted from the two precursor ion scan (PIS)
spectra and assigned as potential thiol-containing candidates, which facilitates the identification of analytes. BQB and BQB-d7
labeled thiol-containing compounds can be clearly distinguished by generating two individual ion chromatograms. Thus, thiol-
containing compounds from two samples labeled with different isotope reagents are ionized at the same time but recorded
separately by mass spectrometry, offering good identification and accurate quantification by eliminating the MS response
fluctuation and mutual interference from the two labeled samples. Using the IL−LC−DPIS-MS strategy, we profiled the thiol-
containing compounds in beer and human urine, and 21 and 103 thiol candidates were discovered in beer and human urine,
respectively. In addition, 9 and 17 thiol candidates in beer and human urine were successfully identified by further comparison
with thiol standards or tandem mass spectrometry analysis. Taken together, the IL−LC−DPIS-MS method is demonstrated to be
a promising strategy in the profiling of compounds with identical groups in metabolomics study.
ith the rapid advancement of analytical methods,
metabolomics has been widely applied in various
polar and ionic metabolites that normally are poorly retained
on reversed-phase column, and the ionization efficiencies of
metabolites could also be improved upon isotope labeling due
to the addition of ionizable functional groups. Moreover, the
isotope-labeled analogues were often used as the internal
standards to overcome the matrix and ion suppression effect.11
The typical method of differential isotope labeling normally
introduces a light isotope tag to the analytes in one sample and
a heavy isotope tag to another comparative sample, followed by
mixing the light and heavy isotope-labeled samples for MS
analysis.11 The metabolites with common a functional group
can be easily recognized through determination of a chromato-
graphically coeluted pair of isotope-labeled analytes (MS
doublet peaks) with characteristic mass difference.13 The
isotope-labeled analytes pair can provide the information on
relative quantification as well as the identification of the
metabolites. For example, Guo et al. used 13C- and 12C-dansyl
W
research fields, including disease diagnosis,1,2 food science,3,4
and drug discovery.5,6 Mass spectrometry (MS) is one of the
most prominent platforms for metabolomics study due to its
high sensitivity and specificity.7 However, profiling of unknown
compounds remains to be a bottleneck of MS-based
metabolomics.8 First, certain compounds are inherently difficult
to analyze by liquid chromatography−electrospray ionization
mass spectrometry (LC−ESI-MS) and may thus escape
detection. For example, some compounds are too hydrophilic
to be sufficiently retained on common reversed-phase LC
columns, or have poor ionization yield in ESI due to the
absence of ionizable functional groups, which leads to the
inefficient detection by MS.9 Second, the quantitative profiling
of metabolites by MS is still challenging because the MS
responses of metabolites fluctuate and ionization efficiencies
alter even under the same liquid chromatography conditions.10
To circumvent these problems, the chemical isotope labeling
(IL) strategy has been developed for qualitative or quantitative
profiling of metabolites with the MS-based platform.11,12
Isotope labeling can alter the chromatographic behaviors of
Received: June 26, 2014
Accepted: September 15, 2014
Published: September 15, 2014
© 2014 American Chemical Society
9765
dx.doi.org/10.1021/ac5023315 | Anal. Chem. 2014, 86, 9765−9773

Analytical Chemistry
Article
chloride to derivatize amine, and phenolic hydroxyl group
where the differential isotope tags can be introduced for
qualitative and quantitative profiling of compounds in biological
samples.13,14 Yang et al. described an LC−MS method for γ-
tocopherol analysis in human A549 cell by chemical
derivatization with N-methyl-nicotinic acid N-hydroxysuccini-
mide ester (C1-NANHS) or (C1-NANHS- d3).15
The strategy of IL combined with MS facilitates the
identification of metabolites. However, there are several
disadvantages of this method performed under full scan
mode. First, the isotopic cluster peaks of unlabeled abundant
metabolites may overlap with the peaks of labeled metabolites,
which increases the difficulty of peak interpretation. In addition,
the dissociation of molecular ions of metabolites in the interface
followed by translation to the mass analyzer will cause
ambiguous assignment of mass spectra peaks under full scan
mode.16 Moreover, a number of peaks may arise from the
fragment ions of metabolites rather than the intact ions of
metabolites, and the numbers of ions detected in multiple
experiments may not be identical.16
Compared with full scan mode, precursor ion scan (PIS) that
monitors the characteristic fragment ions produced by the
fragmentation of precursor ions can provide better detection
sensitivity with lower noise due to the improved selectivity.17
PIS mode is powerful in the analysis of molecules which
possess one identical moiety and similar fragmentation
pattern.17 However, the requirement of the identical moiety
of analytes limits the application of PIS in metabolomics study.
In this respect, chemical labeling may expand the application of
PIS. For example, O’Brien-Coker et al. used cyclohexanedione
to derivatize aldehydes, and the common fragment ion at m/z
216 was employed using PIS to detect all cyclohexanedione-
derivatized species.18 Boughton et al. used 6-aminoquinolyl-N-
hydroxysuccinimidyl carbamate to derivatize monoamines, and
the common fragment ion of derivatives at m/z 171 was used
by PIS to detect all monoamine-containing metabolites.19
Therefore, chemical labeling in combination with PIS could be
a promising strategy to qualitatively and quantitatively profile a
relatively wide range of metabolites that are normally
inefficiently detected by LC−ESI-MS.
of the BQB and BQB-d7 labeled thiol-containing compounds,
and the peak pairs with characteristic mass differences can be
extracted from the two PIS spectra and assigned as potential
thiol-containing candidates. The detected thiol-containing
candidates were further identified by product ion scan and
high-resolution mass spectrometry analysis.
MATERIALS AND METHODS
■
Chemicals and Reagents. N-Acetyl-cysteine (Nac),
glutathione (GSH), 3-mercaptohexan-1-ol (3-MH), glycine,
quinoline, quinoline-d7 were purchased from Sigma (St. Louis,
MO, U.S.A.). Chromatographic grade methanol and acetoni-
trile (ACN) were purchased from TEDIA Co. Inc. (Ohio,
U.S.A.). Formic acid and ethylenediaminetetraacetic acid
(EDTA) were purchased from Sinopharm Chemical Reagent
Co., Ltd. (Shanghai, China). All other solvents and chemicals
used were of analytical grade. The water used throughout the
study was purified by a Milli-Q apparatus (Millipore, Bedford,
MA). Tsingtao Beer (Qingdao, China) was obtained from a
local store in Wuhan, China. Stock standard solutions of Nac,
GSH, and 3-MH were prepared in 1.0 mmol/L ethyl-
enediaminetetraacetic acid (EDTA) solution containing 0.05%
formic acid at a concentration of 1.0 mmol/L.
Synthesis of BQB and BQB-d7. ω-Bromoacetonylquino-
linium bromide (BQB) and BQB-d7 were synthesized
according to our previous work.23 Briefly, bromine was added
dropwise to a round-bottom flask containing acetone and
methanol to form 1,3-dibromoacetone. Then, quinoline or
quinoline-d7 dissolved in toluene was added dropwise to the
solution of 1,3-dibromoacetone for 2 h at 0 °C with stirring.
The reaction mixture was then stirred overnight at room
temperature. The precipitate was collected and recrystallized
with absolute ethanol and dry ethyl ether. Stock solutions of
BQB and BQB-d7 were prepared in acetonitrile at the
concentration of 10 mmol/L, respectively, and stored at −20
°C.
BQB Labeling Reaction. The reaction for isotopic labeling
of thiol-containing compounds using BQB and BQB-d7 is
shown in Figure 1. Briefly, 30 μL of BQB or BQB-d7 (1 mmol/
Thiol-containing compounds, which belong to the group of
nonideal LC−ESI-MS analytes, are important for the various
aroma impacts of beer and wine.20,21 In addition, thiol-
containing compounds can potentially inhibit nonenzymatic
oxidative reactions, which cause the development of off-flavors
during beer aging.22 Therefore, profiling of thiol-containing
compounds is important to evaluate organoleptic characteristics
and nutrition of beer.
Here we developed a novel strategy for nontargeted profiling
of thiol-containing compounds in beer by isotope labeling
combined with high-performance liquid chromatography−
double precursor ion scan mass spectrometry (IL−LC−DPIS-
MS) analysis. In this strategy, we synthesized a pair of isotope
labeling reagents (ω-bromoacetonylquinolinium bromide,
BQB; ω-bromoacetonylquinolinium-d7 bromide, BQB-d7) that
contain a reactive group, an isotopically labeled moiety, and an
ionizable group, and they have been demonstrated to selectively
react with thiol-containing compounds.24 The major advantage
of this strategy is that two characteristic product ions (m/z 218
and 225 containing isotope tag) assigned to the fragmentation
at the C−S bond were generated from the BQB and BQB-d7
labeled thiol-containing compounds, respectively. In this
regard, DPIS (m/z 218 and 225) was used for the profiling
Figure 1. Derivatization reaction for thiol-containing compounds
using BQB and BQB-d7.
L) was added into a 1.5 mL tube and dried under nitrogen gas.
Subsequently 200 μL of Gly−HCl buffer solution (5.0 mmol/L,
pH 3.5) and 100 μL of beer were added. The mixture was
incubated at 60 °C for 60 min with shaking at 1500 rpm. Then
equal volumes of BQB and BQB-d7 labeled sample solutions
were mixed and 50 μL of the solution was used for analysis by
LC−DPIS-MS.
LC−DPIS-MS Analysis. Analysis of samples was performed
on the LC−ESI-MS/MS system consisting of an AB 3200
QTRAP mass spectrometer (Applied Biosystems, Foster City,
9766
dx.doi.org/10.1021/ac5023315 | Anal. Chem. 2014, 86, 9765−9773

Analytical Chemistry
Article
CA, U.S.A.) with an electrospray ionization source (Turbo
Ionspray) and a Shimadzu LC-20AD HPLC (Tokyo, Japan)
with two LC-20AD pumps, a SIL-20A auto sampler, a CTO-
20AC thermostated column compartment, and a DGU-20A3
degasser. Data acquisition and processing were performed using
AB SCIEX Analyst 1.5 Software (Applied Biosystems, Foster
City, CA, U.S.A.). The High-performance liquid chromatog-
raphy (HPLC) separation was performed on a Shimadzu VP-
ODS column (150 mm × 2.0 mm i.d., 5 μm, Tokyo, Japan)
with a flow rate of 0.2 mL/min at 30 °C. Formic acid in water
(0.1%, v/v, solvent A) and methanol (solvent B) were
employed as mobile phases. A gradient of 0−5 min 5% B, 5−
35 min 5% to 60% B, 35−40 min 60% to 5% B, and 40−55 min
5% B was used.
scans were detected in the mass range of m/z 50−550 using
dynamic fill and a scan rate of 4000 Da/s.
LC−QTOF-MS Analysis. High-resolution mass spectrome-
try experiments were performed on the LC−QTOF-MS system
consisting of a MicrOTOF-Q orthogonal-accelerated time-of-
flight (TOF) mass spectrometer (Bruker Daltonics, Bremen,
Germany) with an ESI source (Turbo Ionspray) and a
Shimadzu LC-20AB binary pump HPLC (Tokyo, Japan), a
SIL-20AC auto sampler, and a DGU-20A3 degasser. Data
acquisition and processing were performed using Bruker
Daltonics Control 3.4 and Bruker Daltonics Data analysis 4.0
software. The HPLC separation column and mobile phase
gradient were same as that of the LC−DPIS-MS method.
The mixture of BQB and BQB-d7 labeled samples (1/1, v/v)
were detected under positive ion mode. The optimized ESI
parameters were as follows: capillary voltage, −4.5 kV; dry gas,
5.0 L/min; dry temperature, 180 °C; funnel 1 rf, 200.0 Vpp;
funnel 2 rf, 200.0 Vpp; ISCID energy, 0.0 eV; hexapole rf, 200.0
Vpp; prepulse storage, 12.0 μs. Spectra were acquired by
summarizing 5000 single spectra. Full scan mode was used.
Extraction of centroid spectra peaks with a width of 0.01 Da
was used to pick up the extracted ion chromatograms (EICs)
from the total ion chromatogram (TIC).
The DPIS method consists of two PIS (m/z 218 and 225) in
the mass range of m/z 200−600 (Figure 2). A full scan was
The prospective molecular formulas of BQB-thiol derivatives
were generated based on the accurate mass and isotope
patterns of elemental composition using Bruker Daltonics Data
analysis 4.0 software. A mass tolerance of 5.0 mDa was set, and
a maximum elemental composition of C = 50, H = 100, N = 50,
O = 50, S = 10, P = 10, and Cl = 10 was used. The molecular
formulas of thiol-containing compounds were obtained by
subtracting the molecular formula of BQB (C12H10NO). The
molecular formulas obtained by TOF was further searched in
the database of METLIN (http://metlin.scripps.edu/index.
php) for putative identification.
Figure 2. Schematic diagram of the principle of IL−LC−DPIS-MS.
Effect of H2O2 Treatment on Thiol-Containing
Compounds in Beer. Thiol-type compounds, an important
class of antioxidants, could be easily oxidized by oxidizing
reagents. Here we used H2O2 to treat the beer to evaluate the
quantitative analysis of the DPIS method by examining the
contents change of thiol-containing compounds. Beer sample
was divided in two aliquots, then to one was added H2O2 (1%,
v/v, treated), and to the other one was added an equal volume
of H2O as a control. The two samples were labeled with BQB
and BQB-d7 to form the forward and reverse labeling sample
solutions. In the forward labeling, the untreated beer sample
was labeled with BQB and the treated sample was labeled by
BQB-d7. In the reversed labeling, the untreated beer sample was
labeled with BQB-d7 and the treated sample was labeled by
BQB. Then the two labeled samples were mixed (1:1, v/v) and
analyzed by LC−DPIS-MS. Triplicate measurements were
performed in each labeling strategy.
Profiling of Thiol-Containing Compounds by LC−
DPIS-MS in Human Urine. The first morning urine samples
were collected from 10 healthy volunteers. The urine samples
were pretreated according to a previously described method.23
Briefly, 200 μL of each urine sample was added to a prepared
screw-cap vial (1.0 mL) containing 18 μL of EDTA (10 mmol/
L) and 2 μL of formic acid. A pooled sample is prepared by
taking an equal volume of each urine sample. Then 100 μL
pooled sample was treated with 10 μL of tris(2-carboxyethyl)-
phosphine hydrochloride (TCEP, 10 mmol/L) under 45 °C for
60 min.
performed in the same mass range for comparison. DPIS and
full scan were carried out under positive ion mode. The source
and gas settings for both DPIS and full scans were identical.
IsoSpray voltage was set at 5.2 kV, and vaporizer temperature
was set at 550 °C. The mass spectrometer was operated with
gas settings of 40 psi for nebulizer gas, 30 psi for curtain gas,
and 60 psi for collision gas. Scan time per cycle was 2.0 s with a
pause of 5.0 ms for each scan. Resolution of Q1 and Q3 was set
to “low” and “unit”, respectively. Declustering potential,
entrance potential, cell entrance potential, collision energy,
and cell exit potential were set at 45, 7, 15, 38, and 3 V,
respectively.
For structural identification, IDA (information-dependent
acquisition) mode was performed under positive ion mode.
The IDA cycle comprised a survey scan of a PIS (m/z 218 or
225) followed by three consecutive EPI (enhanced product ion,
MS/MS) scans with different collision energy offsets of 25, 35,
and 45 V. The same precursor was fragmented in three EPI
scans, and then the IDA cycle ended. Upon completing the
cycle, a new IDA cycle started, in which the next precursor
candidate was taken from an inclusion list. An inclusion list was
compiled for all BQB-labeled or BQB-d7-labeled precursor ions,
and the m/z and retention times were generated from the
aforementioned LC−DPIS-MS analysis. The criteria were set as
that EPI was triggered when signals of the preselected
compounds by PIS exceeding 1000 counts/s at their retention
times. The mass tolerance was set to 250 mDa, and retention
time tolerance was set to 60 s. Fragments formed in the EPI
9767
dx.doi.org/10.1021/ac5023315 | Anal. Chem. 2014, 86, 9765−9773

Analytical Chemistry
Article
For the derivatization reaction, 50 μL of BQB or BQB-d7 (1
mmol/L) was added into a 1.5 mL tube and dried under
nitrogen gas. Subsequently, 100 μL of Gly−HCl buffer solution
(5.0 mmol/L, pH 3.5) and 10 μL of urine were added. The
mixture was incubated at 60 °C for 60 min with shaking at 1500
rpm. Then equal volumes of BQB and BQB-d7 labeled sample
solutions were mixed and 50 μL of the solution was used for
LC−DPIS-MS analysis.
FMEA, which limits their applications. Due to the easy
preparation of BQB and BQB-d7 as well as their favorable
generation of the common characteristic fragment ion, we used
BQB and BQB-d7 for the IL−LC−DPIS-MS strategy.
Enhancement of Detection Sensitivity upon Chemical
Labeling. The thiol-containing compounds naturally occurring
in human, animals, plants, and foods, can be ionized in ESI-MS
with either negative or positive ion mode. However, the
ionization efficiency is normally poor. In this respect, chemical
labeling of the thiol-containing compounds by BQB that
harbors a precharged moiety of quaternary ammonium could
largely enhance the ionization efficiencies of thiol-containing
compounds during MS analysis.
RESULTS AND DISCUSSION
■
Principle of the IL−LC−DPIS-MS Strategy. In this IL−
LC−DPIS-MS strategy, a pair of isotope labeling reagents
(BQB and BQB-d7) containing a reactive group, an isotopically
labeled moiety, and an ionizable group were used to label thiol-
containing compounds in beer and human urine. Then BQB
and BQB-d7 labeled sample solutions were mixed (1/1, v/v)
and detected by LC−DPIS-MS. Two characteristic product
ions (m/z 218 and 225 containing the isotope tag) were
generated from the BQB and BQB-d7 labeled thiol-containing
compounds, respectively. In this respect, two PIS (m/z 218 and
225) could be carried out simultaneously in the MS analysis to
record the signals of BQB and BQB-d7 labeled thiol-containing
compounds.
Here we used three thiol-containing standards (Nac, GSH, 3-
MH) to evaluate the enhancement of detection sensitivity upon
chemical labeling by BQB. The mixture of Nac (100 μmol/L),
GSH (100 μmol/L), and 3-MH (100 μmol/L) was analyzed by
direct infusion under either positive or negative ion mode, and
the BQB-labeled mixture (1.0 μmol/L for each analyte) was
analyzed in the positive ion mode. The results showed that the
signal intensities of BQB-labeled GSH increased by 500- and
80-fold compared to the unlabeled GSH detected under
negative and positive ion mode, respectively (Supporting
Information Figure S1). As for unlabeled Nac and 3-MH,
they were barely detected by ESI-MS in either positive ion or
negative ion mode due to their low ionization efficiencies,
whereas after BQB labeling, they can be clearly observed even
the concentration decreased 100-fold (Supporting Information
Figure S1). And it was worth noting that these thiol-containing
compounds were relatively stable under the reaction conditions
(data not shown).
Fragmentation Behavior of BQB-Labeled Thiol-Con-
taining Compounds. Three thiol-containing standards (Nac,
GSH, 3-MH) were used to investigate the fragmentation
behavior of BQB-labeled thiol-containing compounds. The
fragmentation behaviors of BQB-labeled thiols are shown in
Supporting Information Figure S2. The results showed that two
common product ions at m/z of 130.1 and 218.1 were observed
for all the BQB-labeled thiols (Supporting Information Figure
S2A−C). The product ion of m/z 130.1 was assigned to the
fragment ion generated from the quinoline moiety, and m/z
218.1 was assigned to the fragmentation of the C−S bond in
derivatives. Similarly, BQB-d7-labeled thiol-containing com-
pounds generate the other two common product ions at m/z
137.1 and 225.1 with the same fragmentation behavior as that
of BQB-labeled thiols (Supporting Information Figure S2D−
F). In theory, either the ion pair of 130/137 or 218/225 can be
used in the qualitative and quantitative profiling of thiol-
containing compounds in the mixture of BQB and BQB-d7
labeled samples by DPIS strategy.
The advantages of this IL−LC−DPIS-MS strategy includes
(1) BQB labeling can significantly enhance the ESI efficiency,
which therefore leads to the enhanced detection sensitivity of
thiol-containing compounds; moreover, the detection sensitiv-
ity also can be further improved by the elevated selectivity due
to PIS; (2) BQB and BQB-d7 isotope labeling facilitates the
identification of analytes among multiple MS spectral features
by selecting the peak pair with characteristic mass difference;
(3) with DPIS analysis, BQB and BQB-d7 labeled thiol-
containing compounds can be clearly distinguished by
generating two individual ion chromatograms. Thus, thiol-
containing compounds from two samples labeled with different
isotope reagents are ionized at the same time but recorded
separately by mass spectrometry, which can offer accurate
identification and quantification by eliminating the MS
response fluctuation and mutual interference from the two
labeled samples.
Some other thiol-derivatizing reagents, such as sulfhydryl-
reactive reagents, 2-bromo-4′-chloroacetophenone (p-CPB),24
2-bromo-4′-bromoacetophenone (p-BPB),24 isopropylchloro-
formate (IPCF),25 2,3,4,6-tetra-O-acetyl-β-glucopyranosyl iso-
thiocyanate (GITC),26 p-bromo-phenacyl-bromide (p-BMP),27
Ellman’s reagent,28 and popylchloroformate (IPCF)29 were
used to derivatize thiols and disulfides followed by analysis
using LC−MS. The alkylation of −SH and the introduction of
an aromatic group reduce the polarity of the target analytes,
enabling a good retention on a reversed-phase column, whereas
the fragment ions of these derivatives were complex, and it is
difficult to obtain the common characteristic fragment ions for
untargeted profiling of thiol-containing compounds using
precursor ion scanning.
Selection of the Ion Pair for DPIS. As aforementioned,
either ion pair of 130/137 or 218/225 can be used in the LC−
DPIS-MS method for qualitative and quantitative profiling of
thiol-containing compounds. Here we used the mixture
solution of Nac, GSH, and 3-MH standards spiked in beer to
evaluate the quantitative analysis by these two ion pairs, i.e., ion
pair of 130/137 or 218/225. To this end, BQB and BQB-d7
labeled standards were mixed at the concentration ratios of
1:20, 1:10, 1:5, 1:1, 5:1, 10:1, and then the mixture was
measured by DPIS using the ion pair of 130/137 or 218/225.
The calibration curves were constructed by plotting mean peak
intensity ratios of light/heavy versus the mean concentration
ratios of light/heavy based on data obtained from triplicate
N-(2-Ferroceneethyl)maleimide (FEM) and ferrocenecar-
boxylic acid-(2-maleimidoyl)ethylamide (FMEA) also has been
synthesized to derivatize thiols.30,31 The fragmentation behavior
of the FEM and FMEA derivatives (FEM derivatives can
generate neutral loss of 66 amu; FMEA can generate one
common characteristic fragment ion of m/z 213) enables the
detection of unknown thiols by neutral loss scan or precursor
ion scan of derivatives. But the synthesis of FEM and FMEA is
tedious, and it is also difficult to obtain the isotopic FEM and
9768
dx.doi.org/10.1021/ac5023315 | Anal. Chem. 2014, 86, 9765−9773

Analytical Chemistry
Article
Figure 3. BQB and BQB-d7 labeled beer analyzed by LC−DPIS-MS. (A) Total ion chromatogram of DPIS using the ion pair of 218 and 225. (B)
Extracted ion chromatograms by m/z 303 and 310 from BQB and BQB-d7 labeled ion chromatograms, respectively. (C) Extracted ion
chromatograms by m/z 462 and 469 from BQB and BQB-d7 labeled ion chromatograms, respectively. (D) Extracted ion chromatograms by m/z 439
and 446 from BQB and BQB-d7 labeled ion chromatograms, respectively.
Table 1. Measured m/z of BQB and BQB-d7 Labeled Thiol-Containing Compounds from Beer and Their Prospective Molecular
Formulas Obtained by QTOF Analysis
BQB-labeled
(m/z)
BQB-d7-labeled
(m/z)
no. T/min
prospective formulas
compound names
1
16.1
8.6
347.1083
491.1597
303.1138
333.0876
375.0989
377.1142
384.1394
407.1305
407.1283
375.0922
413.1270
416.1240
434.1410
441.2030
449.1310
449.1354
449.1306
462.1980
467.1409
513.2011
524.1739
354.1452
498.2001
310.1502
340.1285
382.1432
384.1578
391.1749
414.1729
414.1733
382.1356
420.1610
423.1658
441.1795
448.1787
456.1749
456.1762
456.1771
469.2301
474.1853
520.2313
531.2150
C5H9NO3S (163.1910)
C10H17N3O6S (307.3210)
C4H9NOS (119.1820)
C4H7NO3S (149.1640)
C6H9NO4S (191.2010)
C6H11NO4S (193.2170)
C8H12N2O2S (200.2560)
C7H13NO5S (223.2430)
C7H13NO5S (223.2430)
C8H5N3OS (191.2098)
C8H11N3O3S (229.2562)
C13H13PS (232.2811)
N-acetylcysteine
glutathione
2
3
16.5
14.4
23.5
17.9
18.6
15.1
16.7
17.9
27.1
16.5
17.1
24.2
14.8
15.6
17.6
20.8
13.0
19.5
14.0
3-aminobutanethioic S-acid
formylcysteine
4
5
(Z)-3-((carboxymethyl)amino)-2-(mercaptomethyl)acrylic acid
2-((2-hydroxyethyl)amino)-3-mercapto-4-oxobutanoic acid
6
7
(2Z,3Z)-2-(aminomethylene)-N-ethyl-3-formyl-4-mercaptobut-3-enamide
3-((1-carboxy-2-mercaptoethyl)amino)-2-(hydroxymethyl)propanoic acid
4-((1-carboxy-2-mercaptoethyl)amino)-2-hydroxybutanoic acid
8
9
10
11
12
13
14
15
16
17
18
19
20
21
C9H18N2S3 (250.4476)
C13H23NS2 (257.4584)
C10H12N5PS (265.2745)
C13H15NOS2 (265.3943)
C8H15N3O3S2 (265.3530)
C10H22N4OS2 (278.4379)
C9H18NO5PS (283.2817)
C22H19NS (329.4580)
C16H20O6S (340.3914)
measurements. As shown in Supporting Information Figure
S3A, the slopes of linear regressions of GSH, Nac, and 3-MH
were 0.37, 0.52, and 0.80 by DPIS using the ion pair of 130/
137, suggesting the ratios of the chromatographic peak intensity
inadequately matched with the concentration ratios of the
different isotope-labeled analytes, whereas the slopes of linear
regressions of GSH, Nac, and 3-MH were 0.88, 0.94, 0.91,
which were approximate to 1.00 by DPIS using the ion pair of
218/225 (Supporting Information Figure S3B), indicating the
ratios of the chromatographic peak area closely matched with
9769
dx.doi.org/10.1021/ac5023315 | Anal. Chem. 2014, 86, 9765−9773

Analytical Chemistry
Article
the concentration ratios of the different isotope-labeled
analytes. The coefficients of determination (R2) were also
found to be close to 1.00, demonstrating good linearities by
DPIS using the ion pair of 218/225. The better correlation
obtained by DPIS using the ion pair of 218/225 rather than
130/137 may be attributed to that the product ions of 218 and
225 assigned to the fragmentation of the C−S bond in BQB
and BQB-d7 derivatives were more specific for the selection of
the thiol-containing compounds. Moreover, the product ions of
130 and 137 are at the low-mass region that typically exhibits
significant background noise from fragment ions of solvent or
contaminants.
We also examined the ion pairs of 218 and 225 in the
qualitative analysis of the mixed BQB and BQB-d7 labeled
standards (1/1, v/v). The results showed that the peak areas of
the three thiol standards (GSH, Nac, and 3-MH) were similar,
and the isotopic effect was approximate 0.1 min on the
reversed-phase LC separation (Supporting Information Figure
S4); the slight retention time difference did not influence the
identification of thiol-containing compounds by LC−DPIS-MS
method. Therefore, the ion pairs of 218 and 225 were used in
the subsequent DPIS analysis.
Qualitative Analysis of Thiol-Containing Compounds
by LC−DPIS-MS in Beer. For the qualitative profiling of thiol-
containing compounds from beer, an equal volume of beer was
labeled with BQB or BQB-d7, respectively. Then the light- and
heavy-labeled samples were mixed and analyzed by LC−DPIS-
MS. The LC−DPIS-MS method consisted of two PIS of m/z
218 and 225, which generated two individual ion chromato-
grams corresponding to the precursor ion of BQB and BQB-d 7
labeled thiol-containing compounds, respectively. Peak-pair
data were extracted from the two ion chromatograms according
Figure S5). These results demonstrated that the LC−DPIS-MS
method is reliable to profile thiol-containing compounds.
We then further examined the 21 thiol-containing com-
pounds by QTOF-MS analysis. The detected peak was analyzed
according to ion pair recognition and retention time
information from the DPIS. A major advantage of the stable
isotope labeling method is that the target analytes can be easily
identified by extracting the peak-pair data even if the retention
time changed due to the use of a different chromatography
column or analytical instrument. The accurate molecular weight
of BQB and BQB-d7 labeled thiol-containing compounds and
the 21 prospective molecular formulas of the thiol-containing
compounds determined using Bruker Daltonics data analysis
4.0 software are shown in Table 1. Beer is a fermented
beverage, and the thiol-containing compounds were normally
considered as the metabolites of yeast.32 We then searched the
molecular formulas of the thiol-containing compounds in the
METLIN database, and compounds 1 (C5H9NO3S) and 2
(C10H17N3O6S) were identified to be Nac and GSH. The
retention times (Supporting Information Figure S6) and MS/
MS spectra (Supporting Information Figure S7) of BQB and
BQB-d7 labeled compounds 1 and 2 from beer are consistent
with the standards Nac and GSH spiked in beer (GSH, 1
μmol/L; Nac, 0.05 μmol/L) in the DPIS analysis, which further
supports the identified compounds 1 and 2. It is worth noting
that Nac existence in beer was first reported by our method. As
for the other thiol-containing compounds, their formulas were
not found in the METLIN database, indicating they could be
some new compounds. And we further determined additional
seven thiol-containing compounds in beer (Table 1, com-
pounds 3−9) by their prospective molecular formulas and MS/
MS spectra analysis (Supporting Information Figure S8).
Method Validation. We investigated the detection limits of
labeled thiol standards (GSH, Nac, and 3-MH) under LC−
DPIS−MS conditions. The results showed that the detection
limits (calculated at a signal-to-noise ratio of 3) were found to
be 37.5, 2.3, and 22.5 nmol/L for GSH, Nac, and 3-MH,
respectively.
To evaluate the accuracy of the relative quantification of the
method, BQB and BQB-d7 labeled beer samples were mixed at
different volume ratios (1:10, 1:5, 1:2, 1:1, 2:1, 5:1, and 10:1).
The samples were analyzed using LC−DPIS-MS with triplicate
measurements. Fifteen BQB and BQB-d7 labeled peak pairs
with high intensities were extracted, and their ratios were
calculated. The average isotopic ratios calculated from each pair
of BQB/BQB-d7 labeling thiols were determined to be 0.10,
0.22, 0.51, 1.02, 2.10, 5.11, and 9.72 for the 1:10, 1:5, 1:2, 1:1,
2:1, 5:1, and 10:1 mixtures, respectively. The results show that
the chromatographic peak intensity ratios highly matched with
the concentration ratios of the different isotope-labeled analytes
with relative standard deviations (RSDs) being less than 15.6%
(Figure 4).
On the basis of the peak area ratios of BQB/BQB-d7-labeled
derivatives, we can determine the absolute concentration of
each metabolite in the samples. BQB-d7-labeled standards were
spiked into the BQB-labeled beer sample, and the concen-
tration of relevant compounds can be determined based on the
measured peak area ratios of BQB/BQB-d7-labeled derivatives.
The GSH and Nac were quantified to be 2.09 and 0.31 μmol/L
in beer, respectively.
to a mass shift of n × 7 Da (i.e., MBQB‑d ‑labeled − MBQB‑labeled = n
7
× 7 Da), and only peak pairs with the same retention time and
intensity were assigned to be the candidates of thiol-containing
compounds. The structures of the assigned candidates were
further elucidated by product ion scan (MS/MS) and high-
resolution mass spectrometry (QTOF-MS) analysis.
Figure 3A shows the total ion chromatograms of beer sample
analyzed by DPIS using the ion pair of 218/225. The
chromatograms derived from the BQB and BQB-d7 labeled
beer samples display almost identical peak patterns. Peak-pair
data were extracted from the BQB and BQB-d7 labeled
chromatograms according to the aforementioned criteria, and
21 ion pairs were detected (Table 1). Taking compound 1 and
17 as the examples, the peak intensities and retention times
were the same in the extracted ion chromatograms at m/z 303/
310 or m/z 462/469 from BQB and BQB-d7 labeled samples
(Figure 3, parts B and C), suggesting these candidate
compounds could be thiol-containing compounds. On the
contrary, in the extracted ion chromatograms at m/z 439 and
446, the peak was only found in the BQB-labeled sample and
no peak was observed in BQB-d7-labeled sample, indicating that
this compound should not be a thiol-containing compound
(Figure 3D).
For the MS/MS analysis, similar to the thiol standards (GSH,
Nac, and 3-MH), all the BQB-labeled thiol candidates possess
two common product ions of m/z 130 and 218, and BQB-d7-
labeled thiol candidates possess two common product ions of
m/z 137 and 225, indicating that all the 21 thiol candidates
were thiol-containing compounds (Supporting Information
Effect of H2O2 Treatment on Thiol-Containing
Compounds in Beer. Thiol-containing compounds are
important for the antioxidation capability of beer.33 As reducing
9770
dx.doi.org/10.1021/ac5023315 | Anal. Chem. 2014, 86, 9765−9773

Analytical Chemistry
Article
mode under the same chromatography conditions. Compared
to full scan mode (Figure 6A), DPIS has two major advantages.
First, DPIS generates two individual ion chromatograms
corresponding to BQB and BQB-d7 labeled samples (Figure
6B). In this respect, the analytes from two samples labeled with
different isotope reagents are ionized at the same time but
recorded separately by mass spectrometry, which can eliminate
the MS response fluctuation and mutual interference. The ion
signals between the light- and heavy-labeled samples can be
clearly distinguished; therefore, the accuracy to select the peak-
pair data from the light- and heavy-labeled sample chromato-
grams was significantly improved, whereas only one chromato-
gram was obtained under full scan mode, so the mutual
interference from the light- and heavy-labeled ion signals was
inevitable. In fact, it is difficult to select the peak-pairs using
either low-resolution (QTRAP) or high-resolution (QTOF)
mass spectrometry under full scan mode.
Figure 4. Regression line of the measured peak intensity ratios vs the
mean concentration ratios (1:10, 1:5, 1:2, 1:1, 2:1, 5:1, 10:1) of 15
BQB/BQB-d7-labeled peak pairs with high intensities from beer.
Second, DPIS monitors the characteristic ion fragment
produced by the fragmentation of precursor ions, which can
effectively eliminate the interference from the unlabeled
compounds or other fragments. Thus, DPIS can significantly
improve the detection selectivity and sensitivity. Figure 6C
shows the extracted ion chromatograms of BQB and BQB-d7
labeled Nac at the m/z 347 (S/N, 14) and 354 (S/N, 16) in
beer under full scan mode, which have strong background noise
signals. Nevertheless, the extracted ion chromatograms
obtained under DPIS mode are much clearer, and the S/N
was 116 and 130 for the BQB and BQB-d7 labeled Nac,
respectively (Figure 6D). The same results was also observed in
MS spectra (Figure 6, parts E and F). Taken together, the DPIS
method can significantly improve the detection accuracy,
selectivity, and sensitivity, and the distinctly identified quasi-
molecular ions of the thiol-containing compounds are beneficial
to further elucidate their structures.
agents, thiol-containing compounds could be oxidized upon
H2O2 treatment. In this respect, we used a forward and reverse
labeling strategy to examine the contents change of the thiol-
containing compounds in beer by H2O2 treatment. The
measured peak intensity ratios of treated/untreated of 21
thiol-containing compounds in the beer are shown in the
Supporting Information Table S1. As described in Figure 5, the
Profiling of Thiol-Containing Compounds by LC−
DPIS-MS in Human Urine. To extend the LC−DPIS-MS
method to more challenging biological samples, we further
profiled thiol-containing compounds in human urine. The
results showed that 103 thiol candidates in urine were identified
(Supporting Information Table S2). Their structures were
further elucidated by MS/MS and high-resolution mass
spectrometry analysis. Among the 103 thiol candidates, 66
had distinct MS/MS spectra. Through the comparison of
retention times (Supporting Information Figure S10) and MS/
MS spectra with the standards, compounds 1, 2, 3, 4, and 5
from human urine were identified to be Cys, Hcys, Nac, γ-Glu-
Cys, and GSH with concentrations of 12.3, 3.06, 1.92, 0.13,
0.68 μmol/L, respectively. Another 12 thiol candidates were
identified by their prospective molecular formulas and MS/MS
spectra (Supporting Information Table S2, compounds 6−17).
The proposed compound structures and fragmentation path-
ways are shown in Supporting Information Figure S11. It was
worth noting that Cys, Hcys, Nac, γ-Glu-Cys, and GSH have
been previously reported in human urine,23,34 while the other
thiol-containing compounds were first identified in human
urine in the current study.
Figure 5. Effect of 1% H2O2 treatment on the contents change of 21
thiol-containing compounds in beer. Forward labeling, the untreated
beer sample was labeled with BQB and the treated sample was labeled
by BQB-d7. Reversed labeling, the untreated beer sample was labeled
with BQB-d7 and the treated sample was labeled by BQB.
results obtained by forward labeling and reversed labeling were
similar. And all the ratios of treated/untreated were less than 1,
demonstrating that thiol-containing compounds were reduced
after the H2O2 treatment. We also performed the studies at
three different concentrations of H2O2 (0.5%, 1%, 2%) with
different treatment time periods (1 and 2 h) (Supporting
Information Figure S9). The results showed that the contents
of the thiol-containing compounds decreased as the concen-
tration of H2O2 and treatment time increased. However, even
at the condition of 2% H2O2 treatment with 2 h, it was difficult
to oxidize all the thiol-containing compounds. We reason that
the insufficient oxidation of thiol-containing compounds may
be due to the existence of other reduction reagents in beer.
Comparison of Precursor Ion Scan with Full Scan
Method. The BQB and BQB-d7 labeled beer sample (1/1, v/
v) was analyzed by LC−MS with either full scan mode or DPIS
CONCLUSION
■
In the current study, we developed an IL−LC−DPIS-MS
method for the profiling of thiol-containing compounds in beer
and human urine. Two characteristic product ions m/z 218 and
225 that contain an isotope tag were generated from the BQB
and BQB-d7 labeled compounds, respectively, which were used
9771
dx.doi.org/10.1021/ac5023315 | Anal. Chem. 2014, 86, 9765−9773

Analytical Chemistry
Article
Figure 6. Comparison of the DPIS with full scan mode. (A) Total ion chromatogram obtained under full scan mode. (B) Total ion chromatogram
obtained under DPIS mode. (C) Extracted ion chromatograms of BQB and BQB-d7 labeled Nac at m/z 347 and 354 from beer under full scan
mode. (D) Extracted ion chromatograms of BQB and BQB-d7 labeled Nac at m/z 347 and 354 from beer under DPIS mode. (E) MS spectrum of
BQB and BQB-d7 labeled Nac from beer under full scan mode. (F) Mass spectrum of BQB and BQB-d7 labeled Nac from beer under DPIS mode.
for DPIS in MS analysis. The method significantly improved
the detection accuracy, selectivity, and sensitivity in the analysis
of thiol-containing compounds. Using this method, 21 and 103
thiol candidates were discovered in beer and human urine,
respectively. In addition, 9 and 17 thiol candidates in beer and
human urine were successfully identified by further comparison
with standards or tandem mass spectrometry analysis. This IL-
DPIS-MS method can also be extended for the efficient
identification and quantification of a relatively wide group of
metabolites that contain an identical group in metabolomics
study.
ACKNOWLEDGMENTS
■
The authors are thankful for the financial support from the
National Natural Science Foundation of China (21475098,
91217309, 91017013), and we thank Mr. Shu-Jian Zheng for
the help on the fragmentation pathway analysis for the
identified compounds.
REFERENCES
■
(1) Laborde, C. M.; Mourino-Alvarez, L.; Posada-Ayala, M.; Alvarez-
Llamas, G.; Serranillos-Reus, M. G.; Moreu, J.; Vivanco, F.; Padial, L.
R.; Barderas, M. G. Metabolomics 2014, 10, 414−424.
(2) Zhang, A. H.; Sun, H.; Yan, G. L.; Wang, P.; Han, Y.; Wang, X. J.
Cancer Lett. 2014, 345, 17−20.
ASSOCIATED CONTENT
* Supporting Information
Additional information as noted in text. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
(3) Castro-Puyana, M.; Herrero, M. TrAC, Trends Anal. Chem. 2013,
52, 74−87.
(4) Shepherd, L. V. T.; Fraser, P.; Stewart, D. Bioanalysis 2011, 3,
1143−1159.
AUTHOR INFORMATION
Corresponding Authors
*Phone: +86-27-68755595. Fax: +86-27-68755595. E-mail:
bfyuan@whu.edu.cn.
(5) Kell, D. B.; Goodacre, R. Drug Discovery Today 2014, 19, 171−
182.
■
(6) Robertson, D. G.; Frevert, U. Clin. Pharmacol. Ther. 2013, 94,
559−561.
(7) Dettmer, K.; Aronov, P. A.; Hammock, B. D. Mass Spectrom. Rev.
2007, 26, 51−78.
*Phone: +86-27-68755595. Fax: +86-27-68755595. E-mail:
yqfeng@whu.edu.cn.
Notes
(8) Peironcely, J. E.; Rojas-Cherto, M.; Tas, A.; Vreeken, R.;
Reijmers, T.; Coulier, L.; Hankemeier, T. Anal. Chem. 2013, 85,
3576−3583.
The authors declare no competing financial interest.
9772
dx.doi.org/10.1021/ac5023315 | Anal. Chem. 2014, 86, 9765−9773

Analytical Chemistry
Article
(9) Siegel, D.; Meinema, A. C.; Permentier, H.; Hopfgartner, G.;
Bischoff, R. Anal. Chem. 2014, 86, 5089−5100.
(10) Huang, Y. Q.; Wang, Q. Y.; Liu, J. Q.; Hao, Y. H.; Yuan, B. F.;
Feng, Y. Q. Analyst 2014, 139, 3446−3454.
(11) Bruheim, P.; Kvitvang, H. F.; Villas-Boas, S. G. J. Chromatogr. A
2013, 1296, 196−203.
(12) Qi, B. L.; Liu, P.; Wang, Q. Y.; Cai, W. J.; Yuan, B. F.; Feng, Y.
Q. TrAC, Trends Anal. Chem. 2014, 59, 121−132.
(13) Guo, K.; Bamforth, F.; Li, L. J. Am. Soc. Mass Spectrom. 2011, 22,
339−347.
(14) Guo, K.; Li, L. Anal. Chem. 2009, 81, 3919−3932.
(15) Yang, W.-C.; Regnier, F. E.; Jiang, Q.; Adamec, J. J. Chromatogr.
A 2010, 1217, 667−675.
(16) Guo, K.; Li, L. Anal. Chem. 2010, 82, 8789−8793.
(17) Yamashita, S.; Honjo, A.; Aruga, M.; Nakagawa, K.; Miyazawa,
T. J. Oleo Sci. 2014, 63, 423−430.
(18) O’Brien-Coker, I. C.; Perkins, G.; Mallet, A. I. Rapid Commun.
Mass Spectrom. 2001, 15, 920−928.
(19) Boughton, B. A.; Callahan, D. L.; Silva, C.; Bowne, J.; Nahid, A.;
Rupasinghe, T.; Tull, D. L.; McConville, M. J.; Bacic, A.; Roessner, U.
Anal. Chem. 2011, 83, 7523−7530.
(20) Vermeulen, C.; Lejeune, I.; Tran, T. T. H.; Collin, S. J. Agric.
Food Chem. 2006, 54, 5061−5068.
(21) Inoue, K.; Nishimura, M.; Tsutsui, H.; Min, J. Z.; Todoroki, K.;
Kauffmann, J. M.; Toyo’oka, T. J. Agric. Food Chem. 2013, 61, 1228−
1234.
(22) de Almeida, N. E. C.; Lund, M. N.; Andersen, M. L.; Cardoso,
D. R. J. Agric. Food Chem. 2013, 61, 9444−9452.
(23) Huang, Y. Q.; Ruan, G. D.; Liu, J. Q.; Gao, Q.; Feng, Y. Q. Anal.
Biochem. 2011, 416, 159−166.
(24) Duan, X.; Chen, X.; Yang, Y.; Zhong, D. Rapid Commun. Mass
Spectrom. 2007, 21, 3234−3244.
(25) Suh, J. H.; Kim, R.; Yavuz, B.; Lee, D.; Lal, A.; Ames, B. N.;
Shigenaga, M. K. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci.
2009, 877, 3418−3427.
(26) Ito, S.; Ota, A.; Yamamoto, K.; Kawashima, Y. J. Chromatogr. A
1992, 626, 187−196.
(27) Vancea, S.; Imre, S.; Donath-Nagy, G.; Bela, T.; Nyulas, M.;
Muntean, T.; Borka-Balas, R. Talanta 2009, 79, 436−441.
(28) Guan, X.; Hoffman, B.; Dwivedi, C.; Matthees, D. P. J. Pharm.
Biomed. Anal. 2003, 31, 251−261.
(29) Suh, J. H.; Kim, R.; Yavuz, B.; Lee, D.; Lal, A.; Ames, B. N.;
Shigenaga, M. K. J. Chromatogr. B 2009, 877, 3418−3427.
(30) Seiwert, B.; Karst, U. Anal. Bioanal. Chem. 2007, 388, 1633−
1642.
(31) Seiwert, B.; Karst, U. Anal. Chem. 2007, 79, 7131−7138.
(32) Swiegers, J. H.; Kievit, R. L.; Siebert, T.; Lattey, K. A.; Bramley,
B. R.; Francis, I. L.; King, E. S.; Pretorius, I. S. Food Microbiol. 2009,
26, 204−211.
(33) Lund, M. N.; Andersen, M. L. J. Am. Soc. Brew. Chem. 2011, 69,
163−169.
(34) Shen, C. C.; Tseng, W. L.; Hsieh, M. M. J. Chromatogr. A 2009,
1216, 288−293.
9773
dx.doi.org/10.1021/ac5023315 | Anal. Chem. 2014, 86, 9765−9773