Relative quantitation of glycans using stable isotopic labels 1-(d0/d5) phenyl-3-methyl-5-pyrazolone by mass spectrometry

Article abstract of DOI:10.1016/j.ab.2011.07.006

A deuterium reagent, 1-(d5) phenyl-3-methyl-5-pyrazolone (d5-PMP), has been synthesized and used for relative quantitative analysis of oligosaccharides by mass spectrometry (MS) using d0/d5-PMP stable isotopic labeling. Previously reported permethylation-based isotopic labels generate variable mass differences, and reductive amination-based isotopic labels cause a loss of some acid-labile groups in carbohydrates. In contrast, d0/d5-PMP stable isotopic labeling is performed at the reducing end of glycans under basic conditions without desialylation, and the mass difference (Δm = 10 Da) between the heavy form (d5-PMP derivative) and light form (d0-PMP derivative) of each glycan is invariable. When the two derivative forms of a glycan are mixed in equimolar amounts, a pair of peaks with a 10-Da mass differences is observed in the MS profile. The difference at relative intensity between the d0- and d5-PMP derivatives reflects the difference in quantity of glycans in two samples, making it possible to carry out both qualitative and relative quantitative analyses of glycans in glycomic studies. Application of this method on DP ₂ to DP₆ maltodextrin oligosaccharides and N-linked glycans released from ribonuclease B and bovine fetuin demonstrates a 10-fold relative quantitative dynamic range, a satisfying reproducibility (coefficient of variation [CV] ≤ 8.34%), and good accuracy (relative error [RE] ≤ 5.1%) of the method. The suggested technique has been successfully applied for comparative quantitative analysis of free oligosaccharides in human and bovine milk.

Full text of DOI:10.1016/j.ab.2011.07.006

Analytical Biochemistry 418 (2011) 1–9
Contents lists available at ScienceDirect
Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio
Relative quantitation of glycans using stable isotopic labels 1-(d0/d5)
phenyl-3-methyl-5-pyrazolone by mass spectrometry
a,
Ping Zhang a,b, Ying Zhang a, Xiangdong Xue a, Chenjian Wang a, Zhongfu Wang a, , Linjuan Huang
⇑
⇑
a Educational Ministry Key Laboratory of Resource Biology and Biotechnology in Western China, Life Science College, Northwest University, Xi’an 710069, China
b Chemistry and Chemical Engineering School, Xianyang Normal University, Xianyang 712000, China
a r t i c l e i n f o
a b s t r a c t
Article history:
A deuterium reagent, 1-(d5) phenyl-3-methyl-5-pyrazolone (d5-PMP), has been synthesized and used for
relative quantitative analysis of oligosaccharides by mass spectrometry (MS) using d0/d5-PMP stable
isotopic labeling. Previously reported permethylation-based isotopic labels generate variable mass differ-
ences, and reductive amination-based isotopic labels cause a loss of some acid-labile groups in carbohy-
drates. In contrast, d0/d5-PMP stable isotopic labeling is performed at the reducing end of glycans under
Received 14 December 2010
Received in revised form 27 June 2011
Accepted 5 July 2011
Available online 14 July 2011
basic conditions without desialylation, and the mass difference (Dm = 10 Da) between the heavy form
Keywords:
(d5-PMP derivative) and light form (d0-PMP derivative) of each glycan is invariable. When the two deriv-
ative forms of a glycan are mixed in equimolar amounts, a pair of peaks with a 10-Da mass differences is
observed in the MS profile. The difference at relative intensity between the d0- and d5-PMP derivatives
reflects the difference in quantity of glycans in two samples, making it possible to carry out both qualitative
and relative quantitative analyses of glycans in glycomic studies. Application of this method on DP2 to DP6
maltodextrin oligosaccharides and N-linked glycans released from ribonuclease B and bovine fetuin dem-
onstrates a 10-fold relative quantitative dynamic range, a satisfying reproducibility (coefficient of variation
[CV] 6 8.34%), and good accuracy (relative error [RE] 6 5.1%) of the method. The suggested technique has
been successfully applied for comparative quantitative analysis of free oligosaccharides in human and
bovine milk.
Mass spectrometry
Stable isotopic labeling
Oligosaccharide
Glycans
1-(d0/d5) Phenyl-3-methyl-5-pyrazolone
Ó 2011 Elsevier Inc. All rights reserved.
Glycosylation is one of the most common posttranslational
modifications of proteins in eukaryotic cells. It plays a pivotal role
in many biological functions of glycoproteins such as the cellular
recognition processes [1]. Recent studies have demonstrated that
many diseases, such as inflammation [2], immunologically medi-
ated disease [3], tumor [2,4], and CDG (congenital disorders of
glycosylation) syndrome [5], are related to alteration of glycans
attached to proteins [2–5]. Therefore, the qualitative and
quantitative analysis of glycans has considerable significance in
seeking useful information about some diseases.
area of qualitative and sequencing glycomics [6], whereas methods
for glycan quantitation are often problematic due to certain reasons
such as the discrepancy in the ionization efficiency of each analyte
and the instability of an instrument response [7]. Absolute quantita-
tion of glycans is hampered by the lack of defined pure glycan stan-
dards that are hardly available due to their structural high diversity
and microheterogeneity. Instead of it, the relative quantitation
method, in which an internal standard is added to the sample, pre-
sents a better choice. The response of the analyte relative to the
internal standard is measured by MS to compensate for the discrep-
ancy in ionization efficiency and the variation of instrument re-
sponse. One of the best internal standards is an isotopically (e.g.,
13C, D, 15N) labeled form of the corresponding analyte.
Biological mass spectrometry (MS)1 presents a powerful tool for
glycomic studies, and recently great progress has been made in the
⇑
The relative quantitation of glycans using stable isotopic label-
ing is typically performed by labeling of glycan samples isolated
from different physiological and/or pathological states with 12C
and 13C (or with 1H and D) followed by MS investigation of the
equimolar mixture of these separately labeled glycan derivatives.
The isotopic derivatives of the same glycans have the same
ionization efficiency because of their same chemical characteristics
and appear in an MS profile as a pair of peaks with a fixed mass dif-
ference based on their different molecular weights. Thus, relative
abundances of glycans can be determined by comparison of the
peak intensities of their derivatives with those of the isotopically
Corresponding authors. Fax: +86 29 88373079.
E-mail addresses: wangzhf@nwu.edu.cn (Z. Wang), huanglj@nwu.edu.cn (L.
Huang).
1
Abbreviations used: MS, mass spectrometry; PMP, 1-phenyl-3-methyl-5-pyrazo-
lone; UV, ultraviolet; HPLC, high-performance liquid chromatography; CE, capillary
electrophoresis; LC, liquid chromatography; d5-PMP, 1-(d5)-phenyl-3-methyl-5-
pyrazolone; SPE, solid-phase extraction; SnCl2Á2H2O, tin(II) chloride; NaNO2, sodium
nitrite; PNGase F, peptide N-glycosidase F; NMR, nuclear magnetic resonance; DMSO,
dimethyl sulfoxide; ESI, electrospray ionization; SDS, sodium dodecyl sulfate; DTT,
dithiothreitol; TFA, trifluoroacetic acid; MALDI–TOF, matrix-assisted laser desorption/
ionization time-of-flight; MS/MS, tandem MS; DP, degrees of polymerization; CV,
coefficient of variation; RE, relative error; HMO, oligosaccharide isolated from human
milk; BMO, oligosaccharide isolated from bovine milk.
0003-2697/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.ab.2011.07.006

2
Relative quantitation of glycans / P. Zhang et al. / Anal. Biochem. 418 (2011) 1–9
labeled forms, with the latter serving as an internal standard. Cur-
rently, the development of the relative quantitative glycomics pro-
ceeds along two different paths. The first class of approaches is
based mostly on permethylation of glycans with heavy methyl io-
dide (12CD3I [8] or 13CH3I [9,10]) and light methyl iodide (12CH3I).
However, for complicated mixtures of unknown glycans, the num-
ber of methylation sites (NMe) is unknown for each participating
Materials and methods
Materials
d5-Aniline, maltodextrin, PMP, graphitized carbon solid-phase
extraction (SPE) column (150 mg/4 ml), and cellulose (microcrystal-
line) were purchased from Sigma–Aldrich (St. Louis, MO, USA).
Hydrochloric acid, tin(II) chloride (SnCl2Á2H2O), sodium nitrite
(NaNO2), sodium hydroxide, dichloromethane, ethyl acetoacetate,
glacial acetic acid, and 26–28% (v/v) ammonia were obtained from
Third Chemical Company of Tianjin (Tianjin, China). All of these re-
agents were of analytical grade and used without preliminary puri-
fication. Acetonitrile and methanol were HPLC grade and purchased
from Fisher Scientific (USA). Peptide N-glycosidase F (PNGase F) was
obtained from New England Biolabs (Ipswich, MA, USA), and ribonu-
clease B (4200 U/mg dry weight) was purchased from Worthington
Biochemical (USA). Bovine fetuin was obtained from Merck (Darms-
tadt, Germany). Milli-Q water (Millipore, Milford, MA, USA) was
used as a solvent. Human milk samples were provided by a healthy
mother volunteer (at 20 days after parturition) who was attended to
in the First Affiliated Hospital of Medical College (Xi’an Jiao Tong
University, China). Bovine milk was collected from healthy bovine
in Lantian Pasture (Shaanxi Province, China). Both milk samples
were stored at À20 °C until use.
glycan and the mass difference (Dm) between the heavy and light
forms varies and can be very large, presenting a serious limitation
to the permethylation method. Atwood and coworkers improved
the method employing 13CH3I and 12CH2DI [7,11]. The small
Dm
(NMe  0.002922 Da) between these two labeled, nearly isobaric
derivatives permits, on the one hand, locating the MS peaks related
to a corresponding glycan and, on the other, distinguishing be-
tween the components of each glycan originating from different
samples, which finally makes relative quantitative determination
of glycans possible. However, this method needs a high-resolution
mass spectrometer and is unsuitable for the analysis of glycans
with low NMe
.
The other group of methods in relative quantitative glycomics is
based on a reductive amination reaction. In this approach, only the
reducing end of glycans is labeled with isotopic labels such as d0/
d6-2-aminopyridine (derivatives
rium, and d6 denotes hexadeuterium) [12], 12C6/13C6-2-aminoben-
zoic acid ( m = 6 Da) [14].
m = 6 Da) [13], and 12C6/13C6-aniline (
Dm = 4 Da, d0 denotes nondeute-
D
D
Although 13C in place of deuterium could resolve the chromato-
graphic resolution problem produced by deuterium isotope [12–
14], these reagents are usually too expensive for their broad use.
Therefore, the relative quantitative method based on d0/d5-aniline
stable isotopic labeling was also developed in our laboratory [15].
In addition, Bowman’s group synthesized a group of tetraplex sta-
ble deuterium isotope labels for reductive amination reaction with
+0-, +4-, +8-, and +12-Da molecular weight differences, allowing
the direct comparison of glycan compositions from four samples
[16,17]. The major advantage of such methods is that the labeling
Synthesis of d5-PMP
Preparation of d5-phenyldiazonium chloride
A three-necked, round-bottom 500-ml flask was charged with
50 ml of 6 mol/L aqueous hydrochloric acid, and d5-aniline (10 g,
0.1019 mol) was slowly added under vigorous stirring. The mixture
was stirred for an additional 15 min at room temperature. The
solution was cooled down to 0–5 °C (icewater bath), and 21 ml of
a 35% aqueous solution of NaNO2 (0.1020 mol) was added drop-
wise for 40 min under vigorous stirring. The mixture was stirred
for an additional 30 min at 0–5 °C, and the orange yellow aqueous
solution of d5-phenyldiazonium chloride was formed.
happens at only one site of each glycan and
Dm between the heavy
and light forms of all glycans is invariable, simplifying the analysis
of the qualitative and quantitative variations of the same oligosac-
charides in different samples (e.g., disease and normal states, sam-
ples from a manufacturing process monitored at two different time
points). Unfortunately, under the reductive amination conditions,
an acid catalyst is required and the acid-labile groups, such as N-
and O-sulfate groups, and the sialic acid residues in carbohydrates
are usually degraded (partially or completely) [18]. Thus, it be-
comes a problem to quantitatively analyze the glycans with sialic
acids and sulfate residues.
Since first reported as a derivatization reagent for reducing car-
bohydrates by Honda et al. in 1989 [19], 1-phenyl-3-methyl-5-pyr-
azolone (PMP) has been widely used. Unlike a reductive amination,
PMP labeling is carried out in an alkaline medium and presents a
condensation of the active methylene group of PMP and the reduc-
ing end of a saccharide, allowing one to attach two pyrazolone
fragments to a saccharide molecule. It can be used to analyze car-
bohydrates with acid-labile groups and causes no desialylation and
desulfation during the derivatization [18]. The PMP derivatives
exhibit a strong ultraviolet (UV) absorption at 245 nm and a higher
hydrophobicity than those of nonmodified samples. They can be
separated and analyzed by various methods, such as high-perfor-
mance liquid chromatography (HPLC), capillary electrophoresis
(CE), and liquid chromatography (LC)/MS, and detected well even
in a low concentration [18–20]. In this article, we report on a
synthesis of a deuterium reagent, 1-(d5)-phenyl-3-methyl-5-pyr-
azolone (d5-PMP), using comparatively affordable d5-aniline as a
starting material and its further application for relative quantita-
tive analysis of reducing glycans by an MS method.
Preparation of N-(d5-phenyl)hydrazine
To 80 ml of aqueous solution of d5-phenyldiazonium chloride
prepared as described above, 100 ml of 2.0 mol/L SnCl2/HCl aque-
ous solution (45.2 g of SnCl2Á2H2O dissolved in 100 ml of 6 mol/L
aqueous HCl) was added dropwise. The temperature was main-
tained below 5 °C. After vigorous stirring for 3 h at 0–5 °C, the reac-
tion mixture was filtered under vacuum and a gray solid material,
N-(d5-phenyl)hydrazine hydrochloride, was obtained (yield
88.5%). M.p. 249–252 °C; 1H NMR (400 MHz, DMSO-d6) d, ppm:
10.37 (s, 3H, –NH3), 8.35 (s, 1H, –NH–). 13C {1H} NMR (100 MHz,
DMSO-d6) d, ppm: 145.56 (C1 in Ph), 128.42, 120.82 (o- and m-
CH in Ph), 114.12 (p-CH in Ph).
N-(d5-Phenyl)hydrazine was prepared by neutralizing its
hydrochloride filter cake, prepared as described above, with NaOH.
Thus, the filter cake was dissolved in 200 ml of distilled water.
Then 30 ml of 25.0% (w/w) NaOH aqueous solution was slowly
added to it until pH 8.0–9.0. After stirring for 3 h at ambient tem-
perature, the reaction mixture was subjected to exhaustive extrac-
tions with dichloromethane until no obvious fluorescence of the
aqueous phase was observed under a UV lamp. Removal of the
solvent under reduced pressure at 25 °C gave crude N-(d5-
phenyl)hydrazine (8.75 g, 0.0773 mol). Yield 77.4%.
Preparation of d5-PMP
A 100-ml round-bottom flask was charged with 8.75 g of crude
N-(d5-phenyl)hydrazine, and a mixture of 9.8 ml of ethyl acetoace-
tate and 1.2 ml of glacial acetic acid was added dropwise under

Relative quantitation of glycans / P. Zhang et al. / Anal. Biochem. 418 (2011) 1–9
3
vigorous stirring. The mixture was stirred at 30 °C for approxi-
mately an additional 12 h. Acetic acid was removed on a rotary
evaporator under reduced pressure. The mixture was subjected to
column chromatography (silica gel 200–300 meshes, Ø55 Â 170-
mm column; eluent: the mixture of petroleum ether and ethyl ace-
tate [3:2,v/v]). Fraction probes were monitored by electrospray ion-
ization (ESI)–MS. The fractions that exhibited a peak at m/z 180
([M+H]+) in full-scan MS spectra were collected. Removal of the sol-
vents under reduced pressure gave 8.50 g (46.6%) of the target com-
pound as a yellow powder. Purity: 98.8% (HPLC). M.p. 124–127 °C;
1H NMR (400 MHz, CDCl3) d, ppm: 2.19 (s, 3H, CH3), 3.45 (s, 2H, CH2
in pyrazolone); 13C {1H} NMR (100 MHz, CDCl3) d, ppm: 16.98
(CH3), 43.04 (CH2), 118.34 (–N@C < in pyrazolone), 128.50,
137.82, (CH in Ph), 156.28 (C in Ph), 170.50 (>C@O). ESI-MS detec-
tion in positive ion mode: m/z 180.17 [M+H]+.
content was heated at 70 °C for 90 min (water bath). After the reac-
tion was completed, the mixture was concentrated to dryness in a
stream of nitrogen to remove ammonia. The residue was redis-
solved in 200 ll of Milli-Q water and extracted with dichlorometh-
ane (3 Â 0.2 ml) to remove the excess d0/d5-PMP. The aqueous
layer, including d0/d5-PMP derivatives of glycans, was dried and
stored at À20 °C in the dark prior to analysis.
ESI-MS analysis of PMP-labeled glycans
The processed glycans labeled by d0-PMP and d5-PMP were dis-
solved in 1 ml of Milli-Q water, mixed in different molar ratios, and
finally subjected to ESI-MS analysis. The ESI-MS data were ac-
quired on a Thermo Scientific LTQ XL ion trap mass spectrometer
(USA) in positive ion mode (sheath gas N2, flow rate 30.0 arb, aux-
iliary gas flow rate 5.0 arb, spray voltage 5 kV, heated capillary
temperature 350 °C, capillary voltage 48 V, tube lens voltage
Release of N-glycans from standard glycoproteins (ribonuclease B and
bovine fetuin)
220 V). Samples were injected into the ion source by a 2-
quantity (Rheodyne) loop, and the mobile phase was the solution
of methanol/water (1:1, v/v) at a flow rate of 200 l/min.
ll fixed
Ribonuclease B and bovine fetuin (0.25 mg, 25
ll of 1 mg/
l
100 l) were denatured by heating at 100 °C for 10 min in 300 ll
l
of protein denaturation solution (containing 5% sodium dodecyl
sulfate [SDS] and 0.4 mol/L dithiothreitol [DTT]). The N-glycans
were released from the denatured proteins by treatment with
30–60 U/ml PNGase F in the presence of 30 ll of pH 7.5 sodium
phosphate buffer solution (1.9 g of sodium phosphate was dis-
solved in 10 ml of Milli-Q water and the pH was adjusted to 7.5
with phosphoric acid solution) and 30 ll of 10% NP-40 to protect
PNGase F according to the manufacturer’s instructions. The mix-
tures were incubated at 37 °C for 24 h and then boiled for 5 min
to stop the reaction. The sample was lyophilized and redissolved
in 200 ll of butanol/ethanol/water (4:1:1, v/v/v) solution and then
loaded onto a Cellulose (microcrystalline) column (1.5 ml) pre-
washed successively with 10 Â 2.5 ml of Milli-Q water and equili-
brated by 3.0 ml of butanol/ethanol/water (4:1:1, v/v/v) solution.
The column was then washed with 3 Â 5 ml of butanol/ethanol/
water (4:1:1, v/v/v) solution to remove the contaminants such as
proteins, buffer salts, surfactant, and hydrophobic impurities. The
free glycans were eluted with 4 Â 0.5 ml of Milli-Q water. The elu-
ates were collected and evaporated to dryness for d0/d5-PMP
labeling.
Preparation of free oligosaccharides from milk sample
The preparation of free oligosaccharides from milk was per-
formed according to the literature method [21]. The frozen stored
milk samples were completely thawed, and 1 ml of each sample
was centrifuged at 13,000 rpm in a 2-ml Eppendorf tube for
30 min at 4 °C. After the top lipid layer and the precipitated protein
were removed, a 4-fold volume of a dichloromethane/methanol
(2:1, v/v) mixture was added. After centrifugation at 13,000 rpm
for 30 min at 4 °C, the precipitate was removed and a 2-fold vol-
ume of ethanol was added. The solution was kept at 4 °C overnight,
followed by centrifugation for 30 min at 13,000 rpm to remove the
denatured proteins. The supernatant was further purified by SPE
using a porous graphitized carbon column. The oligosaccharides
were eluted with 25% acetonitrile containing 0.1% (v/v) trifluoro-
acetic acid (TFA). The eluates were collected and dried under
reduced pressure on a rotary evaporator and stored at À20 °C until
further use.
Fig.1. Strategies for the relative quantitation of glycans using d0/d5-PMP as stable
isotopic labels by MS. (A) Reducing glycans are derivatized with d0-PMP or d5-PMP.
(B) The d0-PMP and d5-PMP derivatives of glycans are mixed and analyzed by ESI-
MS or MALDI–TOF MS. (C) Mass spectrum of the mixture of d0/d5-PMP-labeled
glycans in an equimolar ratio. The same glycan from two samples is identified by a
Derivatization of free glycans with d0/d5-PMP
pair of peaks with
Dm = 10 Da, and the differences in its amount could be
d0/d5-PMP methanol solutions (0.5 mol LÀ1, 50
nia (28%, 30 l) were added to a 5-ml vial containing 0.5–10 nmol
of free glycans. The vial was sealed and shaken vigorously, and the
ll) and ammo-
determined by the intensity difference of the two peaks in the doublet. In contrast,
the glycan present in one sample but absent in the other would exhibit a single peak
in MS. The structure of each glycan is shown with symbols as indicated.
l

4
Relative quantitation of glycans / P. Zhang et al. / Anal. Biochem. 418 (2011) 1–9
charides, each with
D
m = 10 Da (molecular ion peaks [M+H]+ with
Results
m/z 673.00/683.00, 835.00/845.00, 997.00/1007.00, 1159.00/
1169.00, and 1321.00/1331.00 are assigned to d0/d5-PMP-labeled
DP2 to DP6 oligosaccharides, respectively), exhibit intensity ratios
similar to their theoretical (expected) molar ratios.
Strategy of d0/d5-PMP stable isotopic labeling method for qualitation
and relative quantitation of glycans
Fig. 3 demonstrates an evident linear dependency between
experimental and expected data (correlation coefficient
R P 0.9971) for five oligosaccharides taken in varied molar ratios
(10:1, 7:1, 5:1, 3:1, 2:1, 1:1, 1:2, 1:3, 1:5, 1:7, and 1:10) and the ob-
served ratios in ESI-MS spectra (measurements were performed
three times for each case). Expansion of the ratio ranges to 1:15
or 15:1 causes a drop of R down to 0.9905. At the ratio of 1:20 or
20:1, the minor peaks could not be detected (data not shown).
These data indicate that the limit of detection in this analysis is be-
tween 6.7% (1:15 or 15:1) and 10% (1:10 or 10:1) of the major peak.
The method clearly demonstrates a 10-fold linear dynamic range
for these measurements.
The reproducibility of the method was estimated by three re-
peated experiments for every ratio to give some coefficients of
variation (CVs) and then to obtain a mean CV for each oligosac-
charide. The accuracy of the method was estimated by the mea-
sured average ratio of d0/d5-PMP-labeled glycans versus the
theoretical ratio of d0/d5-PMP-labeled glycans to give some rela-
tive errors (REs) and then to obtain a mean RE. The statistical
analysis data are listed in Table 1. As shown in this table, the
mean CV does not exceed 6.86% and the mean RE does not exceed
4.51% for five studied DP2 to DP6 oligosaccharides. Such results
are indicative of rather good precision and acceptable accuracy
of the method.
This method is based on the quantitative derivatization of free
reducing glycans with d0-PMP and d5-PMP (see Fig. 1A). Condensa-
tion of an active methylene group in PMP with the reducing end of
saccharide in alkaline medium introduces two PMPs per one glycan.
Free glycans from two different samples were labeled by d0-PMP
and d5-PMP, respectively, and then mixed in an equimolar ratio
for MS detection. The same glycan from two samples exhibits a
doublet with a 10-Da mass difference in MS profile due to the sim-
ilar chemical characteristic ionization efficiency of the isotopic gly-
can derivatives. This can be used for qualitative analysis of glycans.
Meanwhile, the difference in relative abundance between the d0-
PMP derivative and the d5-PMP derivative corresponds to the quan-
tity variance of the glycans originating from two samples. It permits
the relative quantitative analysis of glycans in glycomic studies.
Fig. 1B and C show an example of the anticipated results of a
comparison between two glycoprotein samples (1 and 2) with
two identical N-linked glycans marked with A and B (glycan A
is at equal amount in two samples and glycan B is at unequal
amount in two samples) and one different N-linked glycan (gly-
can C is in sample 1 and glycan D is in sample 2). When the gly-
cans released from sample 1 (labeled with d0-PMP) are mixed
with glycans released from sample 2 (labeled with d5-PMP) in
an equimolar ratio and then analyzed by ESI-MS or matrix-as-
sisted laser desorption/ionization time-of-flight (MALDI–TOF)
MS, glycan 1A from sample 1 and the equivalent glycan 2A from
sample 2 are detected at an equal ion intensity with a 10-Da
mass difference (Fig. 1B and C). Thus, it can be concluded that
glycan A is present in samples 1 and 2 in equal amounts. Glycan
1B from sample 1 and glycan 2B from sample 2 are detected with
a 10-Da mass difference in different relative abundance; the rel-
ative amount of glycan B in sample 2 in comparison with sample
1 can be calculated as a ratio of their relative abundances. Struc-
tural identity of 1B and 2B can be easily confirmed by further tan-
dem MS (MS/MS)n analysis. In contrast, glycan 1C from sample 1
exhibits only a molecular ion peak of d0-PMP-labeled glycan,
with no peaks heavier by 10 Da observed (d5-PMP-labeled gly-
can). This indicates that glycan 1C is present in sample 1 but is
absent in sample 2. Similarly, glycan 2D from sample 2 exhibits
only a molecular ion peak of d5-PMP-labeled glycan, indicating
that glycan 2D is present in sample 2 but is absent in sample 1.
Such a simple procedure allows a direct quantitative comparative
glycomic analysis according to the intensity ratio between the
peaks of isotope-derivatized glycans. Furthermore, the glycans la-
beled with d0-PMP in sample 1 can be standard glycans taken in
known amounts. When they are mixed with unknown glycans la-
beled with d5-PMP and analyzed by MS, the exact concentration
of the unknown glycans can be determined.
Validation of the method by maltodextrin oligosaccharide
Maltodextrin mixture was first chosen as an oligosaccharide
standard to validate the feasibility of the above strategy.
Maltodextrin was labeled with d0-PMP and d5-PMP, respectively,
and the two derivatives were mixed in different molar ratios, fol-
lowed by ESI-MS analysis. Five oligosaccharides of degrees of poly-
merization (DP) from 2 to 6 were selected to evaluate the method
in terms of linearity, precision, and accuracy. Fig. 2 shows the ESI-
MS spectra of the mixture of the five maltodextrin oligosaccharides
labeled with d0-PMP and d5-PMP in five different molar ratios (10:1,
2:1, 1:1, 1:2, and 1:10). As shown in Fig. 2, the peaks of five oligosac-
Fig.2. ESI-MS spectra of the maltodextrin oligosaccharide mixture of DP2–DP6
labeled with d0/d5-PMP in five different molar ratios: (A) 10:1; (B) 2:1; (C) 1:1; (D)
1:2; (E) 1:10. The molecular ion peaks ([M+H]+) of oligosaccharides derivatized with
d0/d5-PMP are assigned to maltose (m/z 673.00/683.00), maltotriose (m/z 835.00/
845.00), maltotetrose (m/z 997.00/1007.00), maltopentaose (m/z 1159.00/1169.00),
and maltohexaose (m/z 1321.00/1331.00), respectively.

Relative quantitation of glycans / P. Zhang et al. / Anal. Biochem. 418 (2011) 1–9
5
Fig.3. Linear graph of the molar ratios of d0/d5-PMP-labeled DP2–DP6 maltodextrin oligosaccharides versus the ratios of their relative intensities observed in ESI-MS profiles. The
inset represents an expansion of the linear graph at the molar ratios below 1.0 of d0/d5-PMP derivatives. Two equal parts of maltodextrin samples were labeled with d0-PMP and
d5-PMP, respectively, and then were mixed in different molar ratios (10:1, 7:1, 5:1, 3:1, 2:1, 1:1, 1:2, 1:3, 1:5, 1:7, and 1:10), followed by ESI-MS analysis. The linear dependency
between the molar ratios and the calculated intensity ratios of the d0-PMP and d5-PMP derivatives of DP2–DP6 oligosaccharides is characterized by a correlation coefficient R.
Table 1
Linearity, reproducibility, and accuracy of d0/d5-PMP stable isotopic labeling method for relative quantitation of maltodextrin oligosaccharide.
Glycan
Regression
equationa
Correlation
Mixed molar ratios of
Measured molar ratios of d0/
Mean reproducibility Mean relative
coefficient (R) d0/d5-PMP-labeled glycans d5-PMP-labeled glycans (mean SD, n = 3) (% CV, n = 3)
error (%)
Maltose (DP2)
y = 0.9657x + 0.0346 0.9992
0.10
0.20
0.50
1.00
3.00
7.00
10.00
0.098 0.002
0.196 0.013
0.507 0.007
0.989 0.032
2.974 0.060
7.265 0.641
9.521 0.363
1.61
6.40
1.40
3.20 (3.90)
2.03
2.00
2.00
1.40
1.10 (2.32)
0.87
8.82
3.82
3.78
4.79
Maltotriose (DP3)
y = 0.9908x À 0.0219 0.9994
0.10
0.20
0.50
1.00
3.00
7.00
10.00
0.094 0.010
0.172 0.008
0.507 0.005
0.984 0.021
2.841 0.365
6.836 0.199
9.949 0.965
10.27
4.91
1.40
2.10 (6.31)
12.85
2.91
6.00
14.0
1.40
1.60 (4.51)
5.31
2.34
0.51
9.70
Maltotetrose (DP4) y = 0.9548x + 0.0631 0.9971
Maltopentaose (DP5) y = 1.0006x + 0.0041 0.9996
Maltohexaose (DP6) y = 0.9401x + 0.0758 0.9983
0.10
0.20
0.50
1.00
3.00
7.00
10.00
0.089 0.007
0.202 0.002
0.496 0.010
0.998 0.055
3.232 0.196
7.035 0.506
9.364 0.567
7.65
0.94
2.01
5.46 (5.05)
6.05
11.40
1.23
0.76
0.15 (4.02)
7.75
0.49
7.20
6.06
6.36
0.10
0.20
0.50
1.00
3.00
7.00
10.00
0.099 0.015
0.199 0.011
0.507 0.006
1.021 0.016
3.099 0.175
7.065 0.741
9.259 0.407
15.61
5.34
1.28
1.57 (6.36)
5.65
10.69
4.39
0.87
0.63
1.35
2.12 (2.37)
3.30
0.92
7.41
0.10
0.20
0.50
1.00
3.00
7.00
10.00
0.103 0.012
0.203 0.001
0.533 0.029
0.981 0.017
3.036 0.174
6.881 0.888
9.685 0.982
11.38
0.63
5.48
1.71 (6.86)
5.75
12.91
10.14
3.10
1.38
6.64
1.88 (2.72)
1.21
1.70
3.15
Note: SD, standard deviation.
a
x denotes theoretical ratios, and y denote measured ratios.
Verification of the method by N-glycans released from glycoproteins
(ribonuclease B and bovine fetuin)
verified by a quantitative study of high mannose-type glycans from
ribonuclease B and N-linked glycans bearing sialic acid residues
from bovine fetuin. These objects are commonly used as standards
whose glycan structures have been well-characterized in previous
studies [8,14,22]. The N-glycans in glycoproteins were released by
The feasibility of the method of the question for relative quan-
titative analysis of the glycans released from glycoproteins was

6
Relative quantitation of glycans / P. Zhang et al. / Anal. Biochem. 418 (2011) 1–9
Fig.4. Mass spectrum of the equimolar mixture of d0-PMP- and d5-PMP-labeled N-glycans released from ribonuclease B by PNGase F. Two identical samples of high
mannose-type N-glycans (containing GlcNAc2Man5, GlcNAc2Man6, GlcNAc2Man7, GlcNAc2Man8, and GlcNAc2Man9) released from ribonuclease B by PNGase F were labeled
with d0-PMP and d5-PMP, respectively, mixed in an equimolar ratio, and then subjected to ESI-MS analysis. The masses of the glycans correspond with their most plausible
structures. The peaks of all the glycans are assigned to [M+H]+. The intensity ratio between the d0-PMP and d5-PMP derivatives of each glycan is shown as R. The inset
represents an amplified profile of the doublets for d0-PMP and d5-PMP derivatives of the GlcNAc2Man5. j, GlcNAc; s, Man.
Fig.5. Mass spectrum of the equimolar mixture of d0-PMP- and d5-PMP-labeled N-glycans released from bovine fetuin by PNGase F. Two identical samples of the N-glycans
released from fetuin by PNGase F were labeled with d0-PMP and d5-PMP, respectively, mixed in an equimolar ratio, and then subjected to ESI-MS analysis. The description of
the information in the figure is similar than in Fig. 4. j, GlcNAc; s, Man; h, Gal; ꢀ, NeuNAc; N, Fuc.
PNGase F and then divided into two equal parts for labeling with
d0-PMP and d5-PMP, respectively, according to the methods as
described above. The d0/d5-PMP-derivatized samples were mixed
in an equimolar ratio and diluted to a suitable concentration prior
for the observed ratio versus the theoretical mixing ratio is 4.4%
for the set of these five glycans. The low CV and RE values indicate
the high precision and good accuracy of this method. Such results
demonstrate the feasibility of the method in relative quantitation
of the neutral glycans released from glycoproteins.
to ESI-MS analysis. The mass spectra of glycan samples from 0.5
of glycoprotein are depicted in Figs. 4 and 5. Five pairs of peaks,
each with m = 10 Da (m/z 1565.33/1575.33, 1727.42/1737.42,
lg
Fig. 5 presents the mass spectrum of d0/d5-PMP-labeled N-gly-
cans released from bovine fetuin taken in an equimolar ratio. Seven
D
1889.33/1899.33, 2051.42/2061.42, and 2213.50/2223.42), are
observed (Fig. 4). These pairs of peaks correspond to [M+H]+ peaks
of d0-PMP- and d5-PMP-labeled glycans (GlcNAc2Man5, Glc-
NAc2Man6, GlcNAc2Man7, GlcNAc2Man8, and GlcNAc2Man9) from
ribonuclease B. The average value of the peak intensity ratios (in
Fig. 4, presented as R = mean value [n = 3] standard deviation be-
low each structure and the mass) between the light and heavy gly-
cans in five pairs of peaks is 0.98 (CV = 5.8%), which is close to the
theoretical ratio of 1.0 for an equimolar mixture. The average RE
visible pairs of peaks, each with Dm = 10 Da (m/z 2118.09/2128.36,
2263.09/2273.36, 2554.18/2564.36, 2628.36/2638.45, 2919.45/
2929.45, 3210.36/3220.64, and 3501.36/3511.82), are observed.
They correspond to [M+H]+ peaks of d0/d5-PMP-labeled N-linked
glycans (assigned to GlcNAc4Man3Gal2Fuc1, GlcNAc4Man3Gal2
NeuNAc1, GlcNAc4Man3Gal2NeuNAc2, GlcNAc5Man3Gal3NeuNAc1,
GlcNAc5Man3Gal3NeuNAc2, GlcNAc5Man3Gal3NeuNAc3, and Glc-
NAc5Man3Gal3NeuNAc4 in order) in fetuin. Such a qualitative result
is convincing due to the fact that the six glycans with NeuNAc

Relative quantitation of glycans / P. Zhang et al. / Anal. Biochem. 418 (2011) 1–9
7
residue all are consistent with previous reports [22]. The only
exception is the neutral glycan assigned to GlcNAc4Man3Gal2Fuc1.
This difference is probably due to the fetuin sample originating
from different batches or sources. The average value of the mea-
sured d0/d5-PMP intensity ratios in the seven pairs of peaks is
1.01 (CV = 8.34%) (in Fig. 5, presented as R = mean value
[n = 3] standard deviation above each pair of peaks), which is
close to the theoretical ratio of 1.0 for an equimolar mixture. The
average RE of the observed ratios versus the theoretical ratio is
5.1% for the set of these seven glycans. Such satisfying results dem-
onstrate that the method can also be applied for relative quantita-
tion of the glycans bearing sialic acid residues.
sample, free oligosaccharide mixtures from human and bovine
milk. The free oligosaccharides (HMOs from human milk and BMOs
from bovine milk) were prepared by the method as described
above and then labeled with d0-PMP and d5-PMP. The tagged
products named as d0/d5-PMP HMOs and d0/d5-PMP BMOs were
mixed in equimolar ratios in three combinations: (i) d0-PMP HMOs
and d5-PMP HMOs, (ii) d0-PMP BMOs and d5-PMP BMOs, and (iii)
d0-PMP HMOs and d5-PMP BMOs. The mixtures were analyzed by
ESI-MS in positive ion mode. The obtained MS spectra are depicted
in Fig. 6. The analysis results are listed in Table 2.
Fig. 6A presents the ESI-MS spectrum of the equimolar mixture
of d0-PMP HMOs and d5-PMP HMOs. In total, 14 pairs of peaks,
each with
Dm = 10 Da, were detected. They are assigned to
[M+H]+ peaks of d0-PMP and d5-PMP derivatives of glycans. The
average peak intensity ratios (n = 3) of the light versus heavy gly-
cans in each of the 14 pairs of peaks ranges from 0.92 to 1.10,
which is close to the theoretical ratio of 1.0 for an equimolar mix-
ture. The proposed sugar compositions are listed in Table 2. This
Analysis of free oligosaccharides derived from milk samples with the
new method
The validity of the suggested method for the qualitative and
relative quantitative analysis of glycans was also tested on a real
Fig.6. Three mass spectra demonstrating the qualitation and quantitation differences of free oligosaccharides isolated from human milk (HMOs) and bovine milk (BMOs) by
the d0/d5-PMP stable isotopic labeling method. (A) ESI-MS spectrum of an equimolar mixture of d0-PMP- and d5-PMP-labeled HMOs. (B) ESI-MS spectrum of an equimolar
mixture of d0-PMP- and d5-PMP-labeled BMOs. (C) ESI-MS profile of an equimolar mixture of d0-PMP-labeled HMOs and d5-PMP-labeled BMOs. The m/z values of these
oligosaccharides are associated with the corresponding doublets. All of the m/z values are assigned to [M+H]+ of the d0/d5-PMP-derivatized oligosaccharides. The intensity
ratio of the d0-PMP and d5-PMP derivatives is shown as R. The inset represents an expansion of a portion of the mass spectrum.

8
Relative quantitation of glycans / P. Zhang et al. / Anal. Biochem. 418 (2011) 1–9
Table 2
Comparison of detected oligosaccharides from human milk and bovine milk using d0/d5-PMP stable isotopic labeling method by ESI-MS.
m/z value ([M+H]+) of d0/d5-PMP-labeled oligosaccharides
Proposed composition
Calculated molecular
weight of oligosaccharides
Observed in the equimolar mixture
of d0-PMP HMOs/d5-PMP BMOs
Observed doublets
in human milk
Observed doublets
in bovine milk
m/z ([M + H]+)
Intensity ratio
(d0/d5 SD)
673.01/683.00
819.00/829.03
835.11/845.13
876.23/886.23
673.05/683.13
–
835.13/845.17
876.33/886.33
Lactose
[Hex]2[Fuc]
[Hex]3
[Hex]2[HexNAc]
[Hex]2[NeuAc]
[Hex]3[HexNAc]
[Hex]3[HexNAc][Fuc]
[Hex]3[HexNAc][Fuc]2
[Hex]4[HexNAc]2
[Hex]3[HexNAc]1[Fuc]3
[Hex]4[HexNAc]2[Fuc]
[Hex]4[HexNAc]2[Fuc]2
[Hex]4[HexNAc]2[Fuc]3
[Hex]5[HexNAc]3[Fuc]1
342.11
488.18
504.26
545.18
633.20
707.23
853.30
999.39
1072.37
1145.47
1218.45
1364.52
1510.60
1583.58
673.01/683.00
819.00/–
835.01/845.00
876.13/886.13
964.10/–
1038.02/–
1184.33/–
1330.00/–
1403.01/–
1476.12/–
1549.05/–
1695.10/–
1841.00/–
1914.00/–
1.62 0.02
–
1.10 0.10
4.55 0.06
–
–
–
–
964.16/974.24
964.16/974.24
1038.02/1048.02
1184.13/1194.13
1330.03/1340.05
1403.10/1413.11
1476.10/1486.10
1549.03/1559.01
1695.02/1705.12
1841.04/1851.00
1914.00/1924.00
–
–
–
–
–
–
–
–
–
–
Note: SD, standard deviation.
table demonstrates that all glycans contain at least two hexose res-
idues (2Hex), similar to lactose-core mentioned in the literature
[23], and nine of them contain one to three fucose residues, which
is consistent with previous studies [21,24]. The gained result
further verified the accuracy of this method. Owing to the partial
loss of lactose in purification of oligosaccharide samples using an
SPE column, the relative abundance of the peak couple (m/z
673.01/683.00) of lactose is not the highest one in those of all
the oligosaccharides observed in the MS profile of human milk.
Fig. 6B depicts the ESI-MS spectrum of the equimolar mixture of
d0-PMP BMOs and d5-PMP BMOs. Under the same sample prepro-
cessing conditions, the variety of oligosaccharides in bovine milk
is much smaller than that in human milk. Only four oligosaccharides
(m/z 673.05/682.13, 835.13/845.17, 876.33/886.33, and 964.16/
974.24) assigned to [M+H]+ peaks of d0/d5-PMP derivatives of lac-
tose, Hex3, Hex2HexNAc, and Hex2NeuAc were detected in bovine
milk (cf. with 14 oligosaccharides detected in human milk). Lactose
shows the highest relative abundances in the BMO MS profile even
though part of free lactose was lost in the purification of milk
samples by the SPE column. This is perhaps due to the fact that the
bovine milk sample we used was taken at the end of a lactation per-
iod when the fraction of oligosaccharides is less abundant than lac-
tose. In addition, the high content of lactose made the peak
intensities of other minor oligosaccharides depressed or even unde-
tected. The average value of the measured abundance ratios of d0-
PMP and d5-PMP derivatives in these peak couples is 0.97, which
is also close to the theoretical ratio of 1.0 for an equimolar mixture.
Fig. 6C represents the ESI-MS spectrum of an equimolar mixture
of d0-PMP HMOs and d5-PMP BMOs. Three pairs of peaks (m/z
673.01/683.00, 835.01/845.00, and 876.13/886.13), each with
Discussion
A deuterium reagent, d5-PMP, has been synthesized and used for
the qualitative and relative quantitative analysis of reducing glycans
by MS with d0/d5-PMP stable isotopic labeling. The relative quanti-
tation by this method was proved to be reliable if the ratio of labeled/
nonlabeled components is in the range of 0.1 through 10 (analysis of
maltodextrin oligosaccharides DP2–DP6). Analysis of the same test
mixture and N-linked glycans released from ribonuclease B and bo-
vine fetuin demonstrated the satisfying reproducibility (CV 6
8.34%) and accuracy (RE 6 5.1%). As a test application, the qualita-
tive and relative quantitative differences of free oligosaccharides
isolated from human and bovine milk were explored. The obtained
satisfying results demonstrate the utility of this approach for de-
tailed comparative investigation of glycans from different samples.
The suggested method is based on the PMP labeling technology
employing aqueous ammonia as a catalyst and causes no desialyla-
tion, so it can be applied to the analysis of various types of glycans,
especially those with acid-labile groups such as N- and O-sulfate
groups, and the sialic acid residues. Moreover, the volatile base
can be removed easily after the reaction, and no inorganic salts
are produced in the sample prior to an ESI-MS analysis. In addition,
glycan peaks can be easily identified in MS profiles because of the
fixed
Dm = 10 Da of the isotopic doublets. Therefore, this method is
expected to be useful for comparative glycomic studies.
However, to assign a structure to an unknown glycan or to its
isomer, besides its molecular mass, the HPLC separation, perme-
thylation. and MSn analysis are also necessary. Fortunately, the
PMP derivatives of glycans are suitable for analysis by all of these
techniques. Recently, our research team was engaged in an analy-
sis of the PMP derivatives of glycans by on-line reversed-phase
HPLC–MS/MS2. Now we are also trying to apply this method in
the relative quantitation of O-linked glycans derived from compli-
cated biological specimens such as blood serum to obtain abnor-
mal glycosylation information on some major diseases.
D
m = 10 Da, were found. They are assigned to [M+H]+ peaks of
d0/d5-PMP-labeled lactose, Hex3, and Hex2HexNAc. The single
peak at m/z 964.10 is assigned to [M+H]+ ions of the d0-PMP deriv-
ative of Hex2NeuAc from human milk. The absence of the corre-
sponding d5-PMP derivative at m/z 974.24 in Fig. 6C along with
the presence of it in Fig. 6B indicates that the content of this glycan
in bovine milk is much lower than in human milk, so that their ra-
tio with HMOs exceeds the detection limit of this method. On the
basis of the peak intensity ratios between d0-PMP- and d5-PMP-
derivatized glycans, the content of lactose in human milk is 1.62
times that in bovine milk, the content of Hex3 in human milk is
1.10 times that in bovine milk, and the content of Hex2HexNAc1
in human milk is 4.55 times that in bovine milk. All of these results
are the average values of three experiments.
Acknowledgments
Financial support from the National Natural Science Foundation
of China (30870548), the Program for New Century Excellent Talents
in Universities (NCET-08-0893), the Hi-Tech Research and Develop-
ment Program of China (2006AA02Z146), the Natural Science Foun-
dation of Shaanxi Province (2010JM2009), and the Natural Science

Relative quantitation of glycans / P. Zhang et al. / Anal. Biochem. 418 (2011) 1–9
9
[13] J.M. Prien, B.D.Q. Qin, S.L. Cockrill, Mass spectrometric-based stable isotopic 2-
aminobenzoic acid glycan mapping for rapid glycan screening of
biotherapeutics, Anal. Chem. 82 (2010) 1498–1508.
Foundation of Shaanxi Province Educational Department (2010
JK894) is gratefully acknowledged.
[14] B.Y. Xia, C.L. Feasley, G.P. Sachdev, D.F. Smith, R.D. Cummings, Glycan
reductive isotope labeling for quantitative glycomics, Anal. Biochem. 387
(2009) 162–170.
References
[15] X.D. Xue, P. Zhang, Z.F. Wang, L.J. Huang, A method for relative quantitation
and qualitation of reductive glycans via aniline stable isotopic labeling and
ESI-MS analysis, Chem. J. Chin. Univ. 31 (2010) 1–4.
[16] M.J. Bowman, J. Zaia, Tags for the stable isotopic labeling of carbohydrates and
quantitative analysis by mass spectrometry, Anal. Chem. 79 (2007) 5777–
5784.
[1] P. Rudd, T. Elliott, P. Cresswell, I. Wilson, R. Dwek, Glycosylation and the
immune system, Science 291 (2001) 2370–2376.
[2] J.N. Arnold, R. Saldova, U.M.A. Hamid, P.M. Rudd, Evaluation of the serum N-
linked glycome for the diagnosis of cancer and chronic inflammation,
Proteomics 8 (2008) 3284–3293.
[3] J.D. Marth, P.K. Grewal, Mammalian glycosylation in immunity, Nat. Rev.
Immunol. 8 (2008) 874–887.
[17] M.J. Bowman, J. Zaia, Comparative glycomics using a tetraplex stable-isotope
coded tag, Anal. Chem. 82 (2010) 3023–3031.
[4] J.A. Goetz, Y. Mechref, P. Kang, M.H. Jeng, M.V. Novotny, Glycomic profiling of
invasive and non-invasive breast cancer cells, Glycoconj. J. 26 (2009) 117–131.
[5] M. Butler, D. Quelhas, A.J. Critchley, H. Carchon, H.F. Hebestreit, R.G. Hibbert, L.
Vilarinho, E. Teles, G. Matthijs, E. Schollen, P. Argibay, D.J. Harvey, R.A. Dwek, J.
Jaeken, P.M. Rudd, Detailed glycan analysis of serum glycoproteins of patients
with congenital disorders of glycosylation indicates the specific defective
glycan processing step and provides an insight into pathogenesis, Glycobiology
13 (2003) 601–622.
[6] A. Dell, H.R. Morris, Glycoprotein structure determination by mass
spectrometry, Science 291 (2001) 2351–2356.
[7] J.A. Atwood III, L. Cheng, G. Alvarez-Manilla, N.L. Warren, W.S. York, R. Orlando,
Quantitation by isobaric labeling: applications to glycomics, J. Proteome Res. 7
(2008) 367–374.
[18] S. Honda, S. Suzuki, A. Taga, Analysis of carbohydrates as 1-phenyl-3-methyl-
5-pyrazolone derivatives by capillary/microchip electrophoresis and capillary
electrochromatography, J. Pharm. Biomed. 30 (2003) 1689–1714.
[19] S. Honda, E. Akao, M. Okuda, S. Suzuki, K. Kakehi, J. Nakamura, High-
performance liquid chromatography of reducing carbohydrates as strongly
ultraviolet-absorbing and electrochemically sensitive 1-phenyl-3-methyl-5-
pyrazolone derivatives, Anal. Biochem. 180 (1989) 351–357.
[20] S. Suzuki, R. Tanaka, K. Takada, N. Inoue, Y. Yashimaba, A. Honda, S. Honda,
Analysis of sialo-N-glycans in glycoproteins as 1-phenyl-3-methyl-5-
pyrazolone derivatives by capillary electrophoresis, J. Chromatogr.
(2001) 319–329.
A 910
[21] N. Tao, E.J. DePeters, S. Freeman, J.B. German, R. Grimm, C.B. Lebrilla, Bovine
milk glycome, J. Dairy Sci. 91 (2008) 3768–3778.
[22] Y.P. Huang, Y. Mechref, M.V. Novotny, Microscale nonreductive release of O-
linked glycans for subsequent analysis through MALDI mass spectrometry and
capillary electrophoresis, Anal. Chem. 73 (2001) 6063–6069.
[23] K. Nakajima, M. Kinoshita, N. Matsushita, T. Urashima, M. Suzuki, A. Suzuki, K.
Kakehi, Capillary affinity electrophoresis using lectins for the analysis of milk
oligosaccharide structure and its application to bovine colostrum
oligosacchardes, Anal. Biochem. 348 (2006) 105–114.
[24] P. Chaturvedi, C.D. Warren, M. Altaye, A.L. Morrow, G. Ruiz-Palacios, L.K.
Pickering, D.S. Newburg, Fucosylated human milk oligosaccharides vary
between individuals and over the course of lactation, Glycobiology 11
(2001) 365–372.
[8] P. Kang, Y. Mechref, Z. Kyselova, J.A. Goetz, M.V. Novotny, Comparative
glycomic mapping through quantitative permethylation and stable-isotope
labeling, Anal. Chem. 79 (2007) 6064–6073.
[9] W.A. Tao, R. Aebersold, Advances in quantitative proteomics via stable isotope
tagging and mass spectrometry, Curr. Opin. Biotechnol. 14 (2003) 110–118.
[10] G. Alvarez-Manilla, N.L. Warren, T. Abney, J.A. Atwood III, P. Azadi, W.S. York,
M. Pierce, R. Orlando, Tools for glycomics: relative quantitation of glycans by
isotopic permethylation using 13CH3I, Glycobiology 17 (2007) 677–687.
[11] J.C. Botelho1, J.A. Atwood III, L. Cheng, G. Alvarez-Manilla, W.S. York, R.
Orlando, Quantification by isobaric labeling (QUIBL) for the comparative
glycomic study of O-linked glycans, Int. J. Mass Spectrom. 278 (2008) 137–142.
[12] J. Yuan, N. Hashii, N. Kawasaki, S. Itoh, T. Kawanishi, T. Hayakawa, Isotope tag
method for quantitative analysis of carbohydrates by liquid chromatography–
mass spectrometry, J. Chromatogr. A 1067 (2005) 145–152.