15N-Cholamine - A smart isotope tag for combining NMR- and MS-based metabolite profiling

Article abstract of DOI:10.1021/ac401712a

Recently, the enhanced resolution and sensitivity offered by chemoselective isotope tags have enabled new and enhanced methods for detecting hundreds of quantifiable metabolites in biofluids using nuclear magnetic resonance (NMR) spectroscopy or mass spectrometry. However, the inability to effectively detect the same metabolites using both complementary analytical techniques has hindered the correlation of data derived from the two powerful platforms and thereby the maximization of their combined strengths for applications such as biomarker discovery and the identification of unknown metabolites. With the goal of alleviating this bottleneck, we describe a smart isotope tag, ¹⁵N-cholamine, which possesses two important properties: an NMR sensitive isotope and a permanent charge for MS sensitivity. Using this tag, we demonstrate the detection of carboxyl group containing metabolites in both human serum and urine. By combining the individual strengths of the ¹⁵N label and permanent charge, the smart isotope tag facilitates effective detection of the carboxyl-containing metabolome by both analytical methods. This study demonstrates a unique approach to exploit the combined strength of MS and NMR in the field of metabolomics.

Full text of DOI:10.1021/ac401712a

Article
pubs.acs.org/ac
15N‑CholamineA Smart Isotope Tag for Combining NMR- and MS-
Based Metabolite Profiling
Fariba Tayyari,† G. A. Nagana Gowda,‡ Haiwei Gu,‡ and Daniel Raftery*,†,‡,§
†Department of Chemistry, Purdue University, West Lafayette, Indiana 47907, United States
‡Northwest Metabolomics Research Center, Anesthesiology and Pain Medicine, University of Washington, Seattle, Washington
98109, United States
§Fred Hutchinson Cancer Research Center, Seattle, Washington 98109, United States
S
* Supporting Information
ABSTRACT: Recently, the enhanced resolution and sensitivity
offered by chemoselective isotope tags have enabled new and
enhanced methods for detecting hundreds of quantifiable
metabolites in biofluids using nuclear magnetic resonance
(NMR) spectroscopy or mass spectrometry. However, the
inability to effectively detect the same metabolites using both
complementary analytical techniques has hindered the correlation
of data derived from the two powerful platforms and thereby the
maximization of their combined strengths for applications such as
biomarker discovery and the identification of unknown metabolites. With the goal of alleviating this bottleneck, we describe a
smart isotope tag, 15N-cholamine, which possesses two important properties: an NMR sensitive isotope and a permanent charge
for MS sensitivity. Using this tag, we demonstrate the detection of carboxyl group containing metabolites in both human serum
and urine. By combining the individual strengths of the 15N label and permanent charge, the smart isotope tag facilitates effective
detection of the carboxyl-containing metabolome by both analytical methods. This study demonstrates a unique approach to
exploit the combined strength of MS and NMR in the field of metabolomics.
he metabolomics field has witnessed exponential growth
identification. The main contributing factors for this bottleneck
Tover the past decade due to its capabilities for systems are the limited NMR sensitivity, complex spectral signatures,
and variable MS ionization efficiency or suppression.
biology research and potential applications in numerous
disciplines including biomedicine, toxicology, food and
nutrition, drug development, and environmental science.1−5
Commonly used analytical techniques such as nuclear magnetic
resonance (NMR) spectroscopy and/or mass spectrometry
(MS) have evolved in response to the high demand for
resolving the complexity of biological mixtures and identifying
the large pool of quantifiable metabolites. However, despite
numerous advances, the biological complexity still often
outweighs the capabilities of these advanced analytical methods;
no single technique currently is capable of detecting all
metabolites in a single experiment. Each analytical method is
sensitive to certain classes of metabolites, and depending on the
nature of the metabolites of interest, generally one or
sometimes a combination of NMR or MS techniques are
used to profile as many metabolites as possible and thereby
derive the biological meaning. A major hurdle of such an
approach is that the metabolite data obtained from NMR and
LC-MS or GC-MS methods for the same or similar samples are
often not directly comparable. The inability to compare and
correlate data from different analytical techniques for the same
or similar samples is a significant challenge that prevents
drawing meaningful conclusions from the vast amount of
metabolite data existing in the literature and exploiting the
combined strength of NMR and MS for unknown metabolite
The use of chemo-selective tags provides an avenue to
improve the sensitivity of metabolite detection by both NMR
and MS methods. For example, the sensitivity of MS detection
is shown to be enhanced by three orders of magnitude or more
by tagging metabolites with chemoselective tags containing a
permanent charge.6−10 Because of the permanent charge, the
tagged metabolites are effectively detected with high sensitivity
and better quantitative accuracy, irrespective of the pH or
nature of the solvents used to separate metabolites before
detection by MS. Separately, based on differential dansylation
using 12C/13C-dansyl chloride, absolute or relative quantitation
of amine and phenol containing metabolites has been achieved
with a sensitivity enhancement of three orders of magni-
tude.11,12 Similarly, NMR-sensitive isotope based chemo-
selective tags have been shown to detect many quantifiable
metabolites with high sensitivity and resolution by NMR.13−17
Using 15N-ethanolamine as the tag, for example, over a hundred
carboxyl-containing metabolites have been detected by 1H−15N
two-dimensional NMR with high resolution and sensitivity.13
However, while metabolites can be detected with high
Received: June 7, 2013
Accepted: August 9, 2013
Published: August 9, 2013
© 2013 American Chemical Society
8715
dx.doi.org/10.1021/ac401712a | Anal. Chem. 2013, 85, 8715−8721

Analytical Chemistry
Article
Figure 1. Schematic figure illustrating the “smart isotope tag” approach used to detect the same metabolites using NMR and MS with high
sensitivity. Tagging carboxyl-containing metabolites with 15N-cholamine enables their enhanced detection by both NMR and MS.
sensitivity by both MS and NMR separately using chemo-
selective tags, the inability to compare and correlate the data
from the two methods is a major bottleneck in the
metabolomics field.
modifications as described below to yield the pure product.18,19
The first step involved reacting potassium 15N-phthalimide with
(2-bromoethyl)trimethylammonium bromide in DMF to
obtain the 15N-substituted phthalimide intermediate (Scheme
1). The second step involved alkaline and acid hydrolyses of the
15N-substituted phthalimide to yield the smart isotope tag, 15N-
cholamine (Scheme 2).
The ability to more easily detect the same metabolites by
both NMR and MS methods would offer new avenues to
compare data between MS and NMR platforms and to exploit
the combined strength of the complementary methods. Toward
this goal, we introduce a new “smart isotope tag” approach,
using 15N-cholamine in this case, which possesses the
characteristics of high NMR sensitivity and resolution through
its isotope enrichment and high MS sensitivity through its
permanent positive charge (see schematic Figure 1). The tag
combines the strengths of individual chemoselective tags,
demonstrated previously and separately for NMR and MS
detection,6,13 and offers news avenues to exploit the combined
strength of these powerful and complementary techniques for
areas such as metabolite profiling and unknown metabolite
identification.
Briefly, for the synthesis of 15N-substituted phthalimide
(Scheme 1), (2-bromoethyl)trimethylammonium bromide (9.5
mmol, 2.35 g) was mixed with 15N-phthalimide potassium (10
mmol, 1.86 g) in a 250 mL round-bottom flask and dry DMF
(100 mL) was added under nitrogen gas. The mixture was then
refluxed at 100 °C with stirring for 12 h. The supernatant from
the reaction mixture was separated, and the solvent was
removed using a rotary evaporator.18 The resulting crude
residue was washed thrice using acetonitrile (5 mL each time),
twice with acetone (2 mL each time) followed by washing again
once with acetonitrile (3 mL) to obtain the pure 15N-
1
substituted phthalimide. H NMR spectra in D2O at each
step were monitored to assess the purity of the intermediate
product. For the synthesis of 15N-cholamine, in the second
step, the 15N-substituted phthalimide (538 mg) (Scheme 1)
was dissolved in DI water (24 mL); 1 N NaOH (2.69 mL) was
added to the solution, and the mixture was left at room
temperature with stirring for 30 min to complete the alkaline
hydrolysis (Scheme 2).19 Subsequently, 12 N HCl (1.8 mL)
was added to the solution, the temperature was raised to 70 °C,
and left for 12 h with stirring to complete the acid hydrolysis
(Scheme 2).19 The solvent was then removed using a rotary
evaporator. The resulting crude residue was washed thrice with
acetonitrile (4 mL each time) followed by washing thrice with
25:75 water/acetone mixture (2 mL each time) to yield the
pure product, 15N-cholamine. 1H NMR spectra in D2O at each
step were monitored to assess the purity of the final product.
Tagging Metabolites Using the Smart Isotope
EXPERIMENTAL SECTION
■
Chemicals and Biofluids. A total of 48 carboxyl-containing
metabolite standards (Table I), (2-bromoethyl)-
trimethylammonium bromide, dimethylformamide (DMF),
methanol, acetonitrile, acetone, hydrochloric acid (HCl),
sodium hydroxide (NaOH) (all from Sigma-Aldrich, St.
Louis, MO), 4-(4,6-dimethoxy[1,3,5]triazin-2-yl)-4-methylmor-
pholinium chloride (DMTMM) (Acros Organic, Pittsburgh,
PA), 15N-phthalimide potassium, and deuterium oxide (Cam-
bridge Isotope Laboratories, Andover, MA) were used without
further purification. Human serum samples were obtained from
Innovative Research, Inc., (Novi, MI) and urine from healthy
volunteers, in accordance with the Internal Review Board at
Purdue University. Deionized (DI) water was from in-house
Synergy Ultrapure Water System from Millipore (Billerica,
MA).
15
Tag N-Cholamine. 15N-Cholamine (5 mg, 50 μmol) was
15
added to 500 μL sample in an eppendorf tube, and the pH of
the mixture was adjusted to 7.0 with 1 M HCl or NaOH. A 21
mg portion of DMTMM was added to initiate the
Design and Synthesis of the Smart Isotope Tag N-
Cholamine. Synthesis of 15N-cholamine involved a two-step
reaction and followed the Gabriel synthesis procedure with
8716
dx.doi.org/10.1021/ac401712a | Anal. Chem. 2013, 85, 8715−8721

Analytical Chemistry
Article
1
Table I. H and 15N-NMR Chemical Shifts for 15N-Cholamine Tagged Carboxyl-Containing Metabolites That Were Measured
with Reference to Smart Tagged Formic Acid
label
name
1H (ppm)
15N (ppm)
label
name
1H (ppm)
15N (ppm)
1
cis-aconitic acid
8.5
118.24
121.47
119.49
120.21
116.00
120.81
120.57
114.39
111.35
115.96
116.03
120.01
115.27
115.6
23
24
2-hydroxyisobutyric acid
DL-isocitric acid
7.95
8.40
8.11
8.37
8.07
8.69
8.63
8.23
8.49
8.34
8.33
8.39
8.28
8.29
8.08
8.19
8.36
8.47
8.35
8.63
8.21
8.35
7.95
8.29
8.63
8.35
8.17
7.96
8.01
8.03
8.11
8.34
7.98
8.27
8.38
117.51
117.15
120.77
118.19
121.92
116.34
111.84
114.18
114.45
115.74
115.88
120.39
122.83
122.15
115.14
121.44
116.08
117.13
112.67
111.40
118.85
115.58
118.85
115.88
111.39
112.72
117.63
119.16
119.64
119.17
119.67
117.79
119.37
118.05
118.20
8.14
8.06
8.07
8.23
8.14
8.05
8.30
8.25
8.34
8.31
8.15
8.38
8.31
25
26
27
isoleucine
isovaleric acid
α-ketoglutaric acid
2
3
4
5
6
7
adipic acid
DL-alanine
28
lactic acid
4-aminobenzoic acid
arginine
29
30
31
32
leucine
asparagine
lysine
aspartic acid
maleic acid
malic acid
8.16
8.55
8.20
8.07
7.87
8.35
8.5
121.35
122.69
121.46
123.95
121.88
115.93
115.22
123.93
122.68
124.24
119.54
115.99
120.42
33
34
35
36
malonic acid
methionine
8
9
betaine
citric acid
oxalic acid
oxaloacetic acid
10
11
12
13
cysteine
37
38
39
40
41
L-phenylalanine
L-proline
cystine
formic acid
fumaric acid
8.05
8.42
8.56
8.38
8.28
8.05
propionic acid
pyroglutamic acid
Pyruvic acid
14
15
glucuronic acid
glutamic acid
42
43
serine
succinic acid
16
glutamine
8.35
115.90
44
succinyl-COA
17
18
glycine
8.2
115.45
114.97
115.19
115.62
116.60
122.20
115.89
117.62
45
46
47
48
L-threonine
L-tryptophan
tyrosine
glycolic acid
8.22
8.37
8.2
19
20
21
22
hippuric acid
valine
histidine
8.36
8.07
8.5
3-hydroxybutyric acid
4-hydroxy-L-proline
8.36
Scheme 1. Synthesis of 15N-Substituted Phthalimide
Scheme 2. Alkaline and Acid Hydrolyses of the 15N-Substituted Phthalimide to Yield 15N-Cholamine
8717
dx.doi.org/10.1021/ac401712a | Anal. Chem. 2013, 85, 8715−8721

Analytical Chemistry
Article
Scheme 3. General Reaction for Tagging Carboxyl-Containing Metabolites with the Smart Isotope Tag 15N-Cholamine
reaction.13,20,21 The mixture was stirred at room temperature
for 4 h to complete the reaction. The general reaction for
tagging metabolites with the smart isotope tag is shown in
Scheme 3. To maintain 15N amide protonation, the pH was
adjusted to 5.0 by adding 1 M HCl or 1 M NaOH, and the
solution volume was adjusted to 580 μL by adding DI water.
Serum was deproteinized using methanol prior to metabolite
tagging and urine was used with no pretreatment.13
ethanolamine tag including the solubility of the tagged
metabolites in aqueous media, large one-bond J-coupling
between 1H and 15N of ∼90 Hz for efficient polarization
1
transfer between H and 15N nuclei, and wide chemical shift
dispersion for different metabolites in the resulting 2D NMR
spectra.13 In addition, and importantly, 15N-cholamine
possesses a permanent positive charge, which enables efficient
positive mode detection of the same carboxyl-containing
metabolites by MS, irrespective of the pH or solvent conditions
of the eluting media, commonly used for chromatographic
separation before detection by MS.6
NMR Spectroscopy. For each sample, 580 μL was mixed
with 30 μL D2O and placed in a 5 mm NMR tube. NMR
experiments were performed on a Bruker DRX 500 MHz or
Avance III 800 spectrometer equipped with a room temper-
Synthesis of 15N-cholamine involved a two-step reaction and
followed the Gabriel synthesis procedure with suitable
modifications to yield the pure product.18,19 As seen in the
1H NMR spectrum (Supporting Information Figure S1), the
pure intermediate compound, 15N substituted phthalimide, was
obtained after the first step of the synthesis. Hydrolysis of this
compound yielded the 15N-cholamine in pure form as can be
ascertained from its 1H NMR spectrum (Supporting
Information Figure S2; peaks at 3.16; 3.48; 3.64 ppm). The
accurate MS and MS/MS spectra for 15N-cholamine, shown in
Supporting Information Figure S3, help further verify the
identity and purity of the synthesized smart isotope tag (m/z =
104.120).
The compound was then used to tag 48 metabolites that
were selected for their prominence as carboxyl-containing
metabolites in biofluids that represent a variety of metabolic
pathways. The 1H and 15N chemical shift data derived from the
2D NMR experiments, after tagging with 15N cholamine, are
shown in Table I. Because the 15N-cholamine and the
previously used 15N-ethanolamine differ only in their terminal
group, the tagging efficiency, reproducibility and chemical shift
values for metabolites with 15N-cholamine tag were similar to
those obtained using the 15N-ethanolamine tag.13
1
ature probe or cryoprobe, respectively, suitable for H inverse
detection with Z-gradients at 298 K. A one pulse sequence with
or without solvent signal suppression using presaturation was
1
used for H 1D NMR experiments. The sensitivity-enhanced
1H−15N 2D heteronuclear single quantum coherence (HSQC)
experiments employed an INEPT transfer delay of 6 ms
1
corresponding to the JNH of 90 Hz. Spectral widths for the H
and 15N dimensions were approximately 8 and 3 kHz,
respectively. Here, 128 free induction decays of 1024 data
points each were collected in the indirect (t1) dimension with 1
or 4 transients per increment. Nitrogen decoupling during the
direct acquisition (t2 dimension) was achieved with the GARP
(globally optimized alternating-phase rectangular pulses)
sequence. The resulting 2D data were zero-filled to 1024
points in the t1 dimension after forward linear prediction to 256
or 512 points. A 45° shifted sine-bell window function was
applied to both dimensions before Fourier transformation.
Chemical shifts were referenced to the 1H signal of TSP for the
1D spectra or the derivatized formic acid signal (1H 8.05 ppm;
15N 123.93 ppm) in the HSQC spectra. Bruker Topspin
versions 3.0 or 3.1 software packages were used for NMR data
acquisition or processing.
Importantly, as anticipated based on the 15N-ethanolamine
tagging approach shown earlier in our laboratory,13 the 15N-
cholamine tagging of metabolites in human serum provided a
rich NMR spectrum due to the large number of carboxyl-
containing metabolites normally present in blood (Figure 2).
The low natural abundance of 15N (0.37%) ensures that none
of the nitrogen containing metabolites interferes with the
detection of carboxyl-metabolites through the 15N-cholamine
tag. Each peak in the spectrum corresponds to different
metabolite from the carboxylic acid class. However, metabolites,
which contain more than one carboxyl group, provide
additional peaks depending on the number of carboxyl groups
and molecular symmetry. In addition, metabolites such as
lactate, which possess α-hydroxyl groups, show more than one
peak for the same metabolite as we described earlier using the
15N-ethanolamine tag.13 Some of the peaks assigned based on
the chemical shift values for the standard compounds are
marked with the corresponding number shown in Table I and
Figure 2b. Similarly, tagging of metabolites in human urine with
15N-cholamine also enabled the detection of peaks correspond-
Mass Spectrometry. LC-MS and LC-MS/MS experiments
were performed using an Agilent 1200 SL-LC system coupled
online with an Agilent 6520 Q-TOF mass spectrometer
(Agilent Technologies, Santa Clara, CA). The sample (8 μL)
was injected onto an Agilent Poroshell 120 EC-C18 column
(30 mm × 50 mm, 2.7 μm), which was heated to 50 °C. The
flow rate was 0.5 mL/min. Mobile phase A was 5 mM
ammonium acetate in water, and mobile phase B was 0.1%
water in ACN. The mobile phase composition was initially kept
isocratic at 3% B for 1 min, then increased to 90% B over 4
min; after another 4 min at 90% B, the mobile phase
composition returned to 3% B. Electrospray ionization (ESI)
was used in positive mode, and the voltage was 3.5 kV. The
mass analyzer was scanned over a range of 50−1000 m/z. The
collision energy for auto LC-MS/MS experiments was fixed at
10 V, targeting preselected compounds.
RESULTS AND DISCUSSION
■
The smart isotope tag, 15N-cholamine, designed, developed,
and used in this study retains all the characteristics of the 15N-
8718
dx.doi.org/10.1021/ac401712a | Anal. Chem. 2013, 85, 8715−8721

Analytical Chemistry
Article
1
Figure 3. Portion of the H−15N HSQC spectrum of human urine
tagged with 15N-cholamine: (1) aconitic acid; (2) adipic acid; (3)
alanine; (5) arginine; (6) asparagine; (7) aspartic acid; (9) citric acid;
(12) formic acid; (15) glutamic acid; (18) glycolic acid; (19) hippuric
acid; (24) isocitric acid; (28) lactic acid; (33) malonic acid; (39)
propionic acid; (40) pyroglutamic acid; (43) succinic acid; (45)
threonine; (46) tryptophan.
The 15N-cholamine tagging of metabolites in aqueous media
enabled a sensitivity enhancement of up to 3 orders of
magnitude or more in the MS detection of carboxyl
metabolites. The derivatized metabolites could be detected
easily in positive ion mode as compared to the same
metabolites detected in negative ion mode without the tag.
For example, the sensitivity for pyruvic acid detected in positive
ion mode after 15N-cholamine tagging was enhanced by a factor
of about 1500 when compared to that detected for the same
metabolite without the 15N-cholamine tag, in negative ion
mode. Figure 4 shows typical mass spectra for formic acid and
pyruvic acid after tagging with 15N-cholamine. The enhance-
ment in sensitivity is primarily due to the high ionization
efficiency imparted by the permanent positive charge of the
15N-cholamine and is in agreement with results by Smith and
co-workers for fatty acid analysis using the heavy and light
forms of cholamine.6 In that study, reactions of metabolites
with cholamine were made in organic solution in contrast to
the aqueous media used here. The 15N-cholamine derivatized
serum samples were then analyzed by LC-MS. As anticipated,
due to the presence of the permanent positive charge, tagged
metabolites could be readily detected in positive ion mode with
high sensitivity. Sensitivity enhancement by a factor of up to
nearly 3000 could be achieved for tagged acids. The extracted
ion chromatograms for a few typical carboxylic acids detected in
serum with 15N-cholamine tag are shown in the Supporting
Information Figure S4.
Figure 2. (a) Portion of the 1H−15N HSQC spectrum of human
serum tagged with 15N-cholamine: (1) aconitic acid; (2) adipic acid;
(3) alanine; (7) aspartic acid; (8) betaine; (9) citric acid; (11) cystine;
(12) formic acid; (15) glutamic acid; (17) glycine; (20) histidine; (21)
3-hydroxybutyric acid; (24) isocitric acid; (28) lactic acid; (29)
leucine; (32) malic acid; (37) phenylalanine; (40) pyroglutamic acid;
(45) threonine; (46) tryptophan; (47) tyrosine; (48) valine. (b)
1
Portion of the H−15N HSQC spectrum of a mixture of standard
compounds at various concentrations obtained after tagging with 15N-
cholamine. The peak numbers correspond to the compounds shown in
Table I.
One potential issue is the effect on chromatographic
retention time caused by the addition of the cholamine tag.
However, separation of the tagged metabolites using HILIC
columns offers an opportunity to effectively separate before
detection using MS. For example, the results of separation of a
mixture of standard carboxylic and amino acids performed
ing to well over a hundred carboxylic acid group containing
metabolites (Figure 3). Peaks tentatively assigned based on the
values for the standard compounds are marked by their
numbers shown in Table I and Figure 2b.
8719
dx.doi.org/10.1021/ac401712a | Anal. Chem. 2013, 85, 8715−8721

Analytical Chemistry
Article
Figure 4. Typical LC-QTOF-MS spectra for formic acid and pyruvic acid obtained after tagging with the smart isotope tag, 15N-cholamine. The
permanent charge on the tagged metabolites enables their sensitive detection; the observed peak is from the intact tagged metabolite.
using a HILIC column, without attempting to optimize
chromatography conditions, indicate that 15N-cholamine tagged
metabolites can be separated effectively (Supporting Informa-
tion Figure S5). More broadly, we can contemplate the use of
dual purpose smart tags for other NMR-MS combinations. For
GC-MS, the addition of a charged species will likely cause
problems related to reduced volatility; however, a different tag,
such as 13C or even 29Si labeled silyl-type tags can be
contemplated.22 Another alternative is the use of smart tags
for capillary electrophoresis (CE) coupled to MS, which is
increasingly of interest in metabolomics.23 In fact, positively
charged derivatization agents (based on pyridinum containing
compounds) have been demonstrated for the use of metabolite
profiling of carboxylic acids by CE-MS.24 Thus, the potential
for the use of smart tags such as cholamine for CE-MS and
NMR is quite promising.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: draftery@uw.edu. Tel.: (206) 543-9709.
Notes
The authors declare the following competing financial interest:
D.R. is an officer at Matrix-Bio.
ACKNOWLEDGMENTS
■
This work was supported by a grant from the National
Institutes of Health (2R01GM085291). We thank the reviewer
who suggested we think more broadly about possible
applications of smart tags. The authors thank Dr. David
Thompson, Purdue University, for discussions and Agilent
Technologies for the generous donation of a Poroshell 120
HILIC column.
In conclusion, we have developed a smart isotope tag, 15N-
cholamine, which possesses dual characteristics for metabolite
profiling in complex biological mixtures using the powerful
analytical techniques of NMR and MS. By combining the
individual strengths of the 15N label and permanent charge, the
smart isotope tag facilitates detection of carboxyl-containing
metabolome by both NMR and LC-MS techniques with high
sensitivity. Detection of the same metabolites by both NMR
and MS (Supporting Information Figure S6), effectively opens
unique opportunities for identification of unknown metabolites
and direct comparison of metabolite data from the two
powerful analytical platforms.
REFERENCES
■
(1) Gowda, G. A. N.; Zhang, S.; Gu, H.; Asiago, V.; Shanaiah, N.;
Raftery, D. Expert Rev. Mol. Diagn. 2008, 8 (5), 617−633.
(2) Shintu, L.; Banudin, R.; Navratil, V.; Prot, J. M.; Pontoizeau, C.;
̀
Defernez, M.; Blaise, B. J.; Domange, C.; Pery, A. R.; Toulhoat, P.;
Legallais, C.; Brochot, C.; Leclerc, E.; Dumas, M. E. Anal. Chem. 2012,
84, 1840−1848.
(3) Wishart, D. S. Trends in Food Sci. Technol. 2008, 19, 482−493.
(4) Lindon, J. C.; Holmes, E.; Nicholson, J. K. Pharm. Res. 2006, 23,
1075−1088.
(5) Veldhoen, N.; Ikonomou, M. G.; Helbing, C. C. Ecotoxicol.
Environ. Saf. 2012, 76, 23−38.
(6) Lamos, S. M.; Shortreed, M. R.; Frey, B. L.; Belshaw, P. J.; Smith,
L. M. Anal. Chem. 2007, 79, 5143−5149.
ASSOCIATED CONTENT
■
S
* Supporting Information
Figures S1−S6.
8720
dx.doi.org/10.1021/ac401712a | Anal. Chem. 2013, 85, 8715−8721

Analytical Chemistry
Article
(7) Yang, W. C.; Adamec, J.; Regnier, F. E. Anal. Chem. 2007, 79,
5150−5157.
(8) Yang, W. C.; Sedlak, M.; Regnier, F. E.; Mosier, N.; Ho, N.;
Adamec, J. Anal. Chem. 2008, 80, 9508−9516.
(9) Yang, W. C.; Regnier, F. E.; Silva, D.; Adamec, J. J. Chromatogr. B.
2008, 870, 233−240.
(10) Yang, W. C.; Regnier, F. E.; Jiang, Q.; Adamec, J. J. Chromatogr.
A. 2010, 1217, 667−675.
(11) Guo, K.; Li, L. Anal. Chem. 2009, 81, 3919−3932.
(12) Wu, Y.; Li, L. Anal. Chem. 2013, 85, 5755−5763.
(13) Ye, T.; Mo, H.; Shanaiah, N.; Gowda, G. A. N.; Zhang, S.;
Raftery, D. Anal. Chem. 2009, 81 (12), 4882−4888.
(14) Ye, T.; Zhang, S.; Mo, H.; Tayyari, F.; Nagana Gowda G, A.;
Raftery, D. Anal. Chem. 2010, 82 (6), 2303−2309.
(15) Shanaiah, N.; Desilva, M. A.; Nagana Gowda G, A.; Raftery, M.
A.; Hainline, B. E.; Raftery, D. Proc. Natl. Acad. Sci. USA 2007, 104,
11540−11544.
(16) DeSilva, M. A.; Shanaiah, N.; Gowda, G. A. N.; Rosa-Perez, K.;
Hanson, B. A.; Raftery, D. Magn. Reson. Chem. 2009, 47, S74−S80.
(17) Gowda, G. A. N.; Tayyari, F.; Ye, T.; Suryani, Y.; Wei, S.;
Shanaiah, N.; Raftery, D. Anal. Chem. 2010, 82, 8983−8990.
(18) Iida, K.; Ohtaka, K.; Kajiwara, M. J. Label. Compd. Radiopharm.
2007, 50, 251−253.
(19) Khan, M. N. J. Org. Chem. 1996, 61 (23), 8063−8068.
(20) Kunishima, M.; Kawachi, C.; Monta, J.; Terao, K.; Iwasaki, F.;
Tani, S. Tetrahedron 1999, 55, 13159−13170.
(21) Kunishima, M.; Kawachi, C.; Hioki, K.; Terao, R.; Tani, S.
Tetrahedron 2001, 57, 1551−1558.
(22) Schraml, J. Prog. NMR Spect. 1990, 22, 289−348.
(23) Ramautar, R.; Somsen, G. W.; de Jong, G. J. Electrophoresis
2013, 34 (1), 86−98.
(24) Yang, W.-C.; Regnier, F. E.; Adamec, J. Electrophoresis 2008, 29
(22), 4549−4560.
8721
dx.doi.org/10.1021/ac401712a | Anal. Chem. 2013, 85, 8715−8721