Enhanced detection of sialylated and sulfated glycans with negative Ion mode nanoliquid chromatography/mass spectrometry at high pH

Article abstract of DOI:10.1021/ac902602e

Negative ion mode nanoliquid chromatography/mass spectrometry (nano-LC/MS) on porous graphitic carbon columns at pH 11 was studied and compared to capillary LC/MS at pH 8 for the analysis of neutral and acidic glycan alditols. Oligosaccharides were chromatographed with an acetonitrile gradient containing 0.04% ammonium hydroxide and analyzed with a linear ion trap mass spectrometer (LTQ) equipped with a modified nanospray interface. Analysis of acidic N-and O-glycan standards revealed that good quality MS/MS spectra could be obtained when loading 1-3 fmol, a 10-fold increase in sensitivity compared to capillary-LC/MS at pH 8. Analysis of a complex mixture of O-glycans from porcine colonic mucins with nano-LC/MS and MS/MS at high pH revealed 170 oligosaccharides in one analysis, predominantly corresponding to sulfated glycans with up to 11 residues. Analysis of the same sample with capillary-LC/MS showed a lower sensitivity for multiply sulfated glycans. Nano-LC/ MS of O-linked oligosaccharides on MUC2 from a human colon biopsy also illustrated that the ionization of oligosaccharides with multiple sialic acid groups was increased compared to those with only one sialic acid residue. Nano-LC/MS at high pH is, thus, a highly sensitive approach for the analysis of acidic oligosaccharides.

Full text of DOI:10.1021/ac902602e

Anal. Chem. 2010, 82, 1470–1477
Enhanced Detection of Sialylated and Sulfated
Glycans with Negative Ion Mode Nanoliquid
Chromatography/Mass Spectrometry at High pH
Kristina A. Thomsson,* Malin Ba¨ ckstro¨ m, Jessica M. Holme´ n Larsson, Gunnar C. Hansson, and
Hasse Karlsson
Department of Medical Biochemistry, University of Gothenburg, Box 440, 405 30 Gothenburg, Sweden
Negative ion mode nanoliquid chromatography/mass
spectrometry (nano-LC/MS) on porous graphitic carbon
columns at pH 11 was studied and compared to capillary
LC/MS at pH 8 for the analysis of neutral and acidic
glycan alditols. Oligosaccharides were chromatographed
with an acetonitrile gradient containing 0.04% ammonium
hydroxide and analyzed with a linear ion trap mass
spectrometer (LTQ) equipped with a modified nanospray
interface. Analysis of acidic N- and O-glycan standards
revealed that good quality MS/MS spectra could be
obtained when loading 1-3 fmol, a 10-fold increase in
sensitivity compared to capillary-LC/MS at pH 8. Analysis
of a complex mixture of O-glycans from porcine colonic
mucins with nano-LC/MS and MS/MS at high pH revealed
170 oligosaccharides in one analysis, predominantly
corresponding to sulfated glycans with up to 11 residues.
Analysis of the same sample with capillary-LC/MS showed
a lower sensitivity for multiply sulfated glycans. Nano-LC/
MS of O-linked oligosaccharides on MUC2 from a human
colon biopsy also illustrated that the ionization of oli-
gosaccharides with multiple sialic acid groups was in-
creased compared to those with only one sialic acid
residue. Nano-LC/MS at high pH is, thus, a highly
sensitive approach for the analysis of acidic oligosac-
charides.
large heterogeneity and technical difficulties to analyze these
molecules, the full repertoire and upper size limits of mucin
oligosaccharides are not known. Mucins are overexpressed in
many adenocarcinomas, and documented glycosylation changes
accompany the shift to the cancer associated form of the protein,
such as for the breast cancer associated MUC1.2 Altered mucin
O-glycosylation has also been associated with infection and
inflammation in intestine and lungs.3-5 Consequently, methodol-
ogy for analyzing mucins and mucin glycosylation, and performing
comparative studies between the health and disease state are,
therefore, of medical importance.
In our aim to simultaneously characterize more complex
mixtures of O-glycans with varying number of acidic residues from
mucins and other O-glycosylated proteins and also from limited
sources, we have turned to negative ion mode capillary-LC/MS
with porous graphitized (PGC) carbon columns, which has been
used for the analysis of glycans since the beginning of the 1990s.6
The chromatography step is essential in order to separate glycan
isoforms with the same mass, before these are fragmented with
CID. The occurrence of isoforms with the same mass in many
mucin-derived glycan samples also makes the use of sensitive
approaches such as, for example, MALDI-MS of permethylated
sulfated glycans7 less attractive. Several groups have demonstrated
the use of nano-LC/MS for N- and O-glycans using reversed and
normal phase HPLC and in conjunction with both positive
and negative ion mode mass spectrometry with detection limits
around 10-20 fmol8-10 and down to 1-7 fmol.11 In our nanospray
system, we have used an interface constructed in-house, which
Mucins constitute a subgroup of large glycoproteins and are
characterized by dense O-glycosylation on serine and threonine
in repeated amino acid sequences called PTS domains. Secreted
mucins are major protein components in the mucosal secretions
of the body, protecting and lubricating the underlying epithelia,
whereas membrane bound mucins are expressed on epithelial
surfaces, involved in intracellular signaling.1,2 The mucin O-glycans
can constitute 50-90 wt % of the mass of the glycoprotein and
can be highly diverse, with hundreds of different glycans on a
single mucin. Branches protrude from the core GalNAc residues,
consisting of repetitive Galꢀ1-3/4GlcNAc units with fucose, sialic
acid, and sulfate with different linkages and often terminated with
blood group epitopes of the ABO and Lewis system. Due to the
(3) Karlsson, N. G.; Olson, F. J.; Jovall, P.; Andersch, Y.; Enerback, L.; Hansson,
G. C. Biochem. J. 2000, 350, 805–814
(4) Holmen, J. M.; Karlsson, N. G.; Abdullah, L. H.; Randell, S. H.; Sheehan,
.
J. K.; Hansson, G. C.; Davis, C. W. Am. J. Physiol. Lung, Cell and Mol. Physiol.
2004, 287, L824–L834
(5) Schulz, B. L.; Sloane, A. J.; Robinson, L. J.; Sebastian, L. T.; Glanville, A. R.;
.
Song, Y. L.; Verkman, A. S.; Harry, J. L.; Packer, N. H.; Karlsson, N. G.
Biochem. J. 2005, 387, 911–919
(6) Davies, M. J.; Smith, K. D.; Carruthers, R. A.; Chai, W.; Lawson, A. M.;
Hounsell, E. F. J. Chromatogr. 1993, 646, 317–326
(7) Yu, S. Y.; Wu, S. W.; Hsiao, H. H.; Khoo, K. H. Glycobiology 2009, 19 (10),
1136–1149
(8) Wuhrer, M.; Koeleman, C. A. M.; Deelder, A. M.; Hokke, C. H. Anal. Chem.
2004, 76 (3), 833–838
(9) Barroso, B.; Dijkstra, R.; Geerts, M.; Lagerwerf, F.; Van Veelen, P.; de Ru,
A. Rapid Commun. Mass Spectrom. 2002, 16 (13), 1320–1329
(10) Bereman, M. S.; Williams, T. I.; Muddiman, D. C. Anal. Chem. 2009, 81
(3), 1130–1136
(11) Karlsson, N. G.; Wilson, N. L.; Wirth, H. J.; Dawes, P.; Joshi, H.; Packer,
N. H. Rapid Commun. Mass Spectrom. 2004, 18, 2282–2292
.
.
.
.
* To whom correspondence should be addressed. Tel:+46-31-7861000. Fax:
+46-31-416108. E-mail: kristina.thomsson@medkem.gu.se.
(1) Thornton, D. J.; Rousseau, K.; McGuckin, M. A. Annu. Rev. Physiol. 2008,
.
.
70 (1), 459
.
(2) Hattrup, C. L.; Gendler, S. J. Annu. Rev. Physiol. 2008, 70 (1), 431
.
.
1470 Analytical Chemistry, Vol. 82, No. 4, February 15, 2010
10.1021/ac902602e  2010 American Chemical Society
Published on Web 01/21/2010

is a simplified version of the one described by Shen and
collegues.12 Our approach has been used for the analyses of
samples of different biological origin.13-15 We have used am-
monium bicarbonate at pH 8 as mobile phase for capillary-LC/
MS,16 but using this for nano-LC/MS was unsuccessful due to
the fact that the emitter tip is situated close to the heated capillary
inlet and buffer salt precipitation on the tip inhibited the formation
of a Taylor cone. In order to achieve a strong basic environment
for ESI, we then explored nano-LC/MS at high pH using 0.04%
ammonium hydroxide as additive and found it to be compatible
with the nanospray conditions. Here, we have compared capillary-
LC/MS at pH 8 with nano-LC/MS at pH 11 for the analysis of O-
and N-linked oligosaccharide standards as well as complex
mixtures of O-glycans from porcine and human colonic mucins.
This shows the advantage of negative ion mode nano-LC/MS at
high pH for the characterization of large acidic glycans.
alkylation of the insoluble mucin-containing fraction. The
detailed protocol is published elsewhere.15 The oligosaccha-
rides were released as alditols as described above.
HPLC Columns. Porous graphitic carbon HPLC columns
were packed in-house at 500 bar pressure in fused silica capillaries
(nano-LC/MS: 25 cm × 75 µm i.d. or 30 cm × 100 µm i.d.; capillary-
LC/MS: 20 cm × 180 µm i.d.; o.d. 375 µm; Polymicro Technolo-
gies, Phoenix, AZ). The columns were packed with 5 µm graphite
particles (Hypercarb, Thermo Hypersil-Keystone, Bellefonte, PA)
using tetrahydrofurane/isopropanol (9:1) as slurry solvent and
methanol as pusher. For connection of the fused silica columns,
sleeves were made of PEEK tubing, 1/16 in. o.d., 400 µm i.d.
(Upchurch Scientific, Oak Harbor, WA).
Nano-LC/MS Interface. From the packed fused silica capil-
lary column, a nanoESI-emitter tip was connected via a through-
bore 1/16 in. union (ZU1T, Vici AG, Switzerland). Two steel
screens with pore sizes of 1 and 0.5 µm and thicknesses of 50
and 40 µm, respectively (Vici AG International), were placed
between the column and the emitter. Double steel screens are
needed in order to prevent graphite particles to penetrate into
the emitter. The tapered nanoESI-emitter tips (around 30 mm
length) were made from 20 µm i.d. and 150 µm o.d. fused silica
capillary and held in position by 20 mm PEEK tubing (150 µm
i.d.) as a sleeve. This connection produces a negligible dead
volume for nanocolumn flow rates. The high voltage (-1.6 kV)
was connected via the bore union. The interface holder was made
in-house of a nonconductive polymer (Delrin, Dupont, DE).
Nano-LC/MS and MS/MS. The HPLC pump (binary pump
1100 series, Agilent Technologies, Palo Alto, CA) delivered a flow
of 200 µL/min which was split down in a 1/16 in. microvolume
connector T with a 0.15 mm bore, by a 50 cm × 50 µm i.d. fused
silica capillary before the injector, allowing approximately 400 nL/
min through the analytical column. The sample (1 µL) was injected
using an HTC-PAL autosampler (CTC Analytics AG, Zwingen,
Switzerland). A fused silica capillary (30 cm × 75 µm i.d.) was
used as transfer line from the injector to the column connected
via a 1/16 in. through-bore union with a steel screen (1 µm pore
size). As graphite is electrically conductive, unions on both sides
of the column will be at high voltage. The voltages at either end
were measured to be -1.600 and -1.608 kV (probe union),
respectively. The union at the column inlet was electrically isolated
with silicone rubber tubing. The gradient [5-50% B; A ) 0.04%
NH3 (equivalent to 21.4 mM, ionic strength of approximately 2
to 3 mM) in mpH2O; B ) 100% acetonitrile] was started 5 min
after injection and eluted for 43 min with a subsequent washing
step of 4 min in 80% B and equilibrated for 16 min. The mass
spectrometer was a LTQ linear quadrupole ion trap mass
spectrometer (Thermo Electron, San Jose´, CA). The heated
capillary was kept at 200 °C; the capillary voltage was -50 V,
and the electrospray voltage was -1.6 kV. Full-scan (m/z
380-2000, 2 microscans, maximum 100 ms, target value of
30 000) was performed, followed by data dependent MS2 scans
of the three most abundant ions in each scan (2 microscans,
maximum 100 ms, target value of 10 000). Signal threshold for
MS2 was set to 500 counts. Normalized collision energy was
35%; an isolation window of 3 u, an activation q ) 0.25, and an
activation time of 30 ms were used. The samples were evaluated
and integrated manually using the Xcalibur version 2.0,
EXPERIMENTAL SECTION
Materials and Methods. Deionized water (18.2 MΩ) was
obtained from a Milli-Q water system (Millipore). HPLC grade
acetonitrile and analytical grade acetic acid, methanol, and
ammonium hydroxide (25% solution) were from Merck. All
chemicals, unless otherwise stated, were purchased from Sigma.
O-glycan standards T-antigen (Galꢀ1-3GalNAcol, Mr ) 385),
sialyl-T antigen (NeuAcR2-3Galꢀ1-3GalNAcol, Mr ) 676), and
disialyl-T antigen (NeuAcR2-3Galꢀ1-3(NeuAcR2-6)GalNAcol, Mr
) 967) were purified as described.17 Su-Lea (Sulfo Lewis a, SO3
-
-
3Galꢀ1-3(FucR1-4)GlcNAc, Mr ) 611), A1 (monosialo(2,6)
biantennary N-glycan, Mr ) 1933), and A2 (disialo(2,6) bian-
tennary N-glycan, Mr ) 2224) were purchased from Dextra
Laboratories (Reading, UK). LNF1 (FucR1-2Galꢀ1-3GlcNAcꢀ1-
3Galꢀ1-3Glcol Mr ) 855) from milk was a gift from David Zoft.
Oligosaccharide standards were reduced to their alditol form
with 0.5 M NaBH4 in 50 mM KOH, 50 °C, 2 h, followed by
desalting as described elsewhere.17 Complete sialylation of the
N-glycans standards was confirmed by analysis with capillary
LC/MS and MS/MS.
Purification of Mucin Oligosaccharides. Mucin oligosac-
charides from porcine colonic mucins were purified as described
elsewhere.18 Briefly, oligosaccharides from 0.5 mg of mucin were
chemically released by ꢀ-elimination and reduced into alditols with
NaBH4/KOH, followed by desalting on cation exchange media,
and elimination of borate by repetitive evaporation with 1%
acetic acid in methanol. MUC2 derived oligosaccharides from
a human biopsy (sigmoid colon) from a healthy individual were
purified by extraction with 4 M guanidinium hydrochloride
containing protease inhibitors, followed by reduction and
(12) Shen, Y.; Zhao, R.; Berger, S. J.; Anderson, G. A.; Rodriguez, N.; Smith,
R. D. Anal. Chem. 2002, 74, 4235–4249
(13) Karlsson, H.; Holmen Larsson, J.; Thomsson, K. A.; Hard, I.; Backstrom,
M.; Hansson, G. C. Methods Mol. Biol. 2009, 534, 1–15
(14) Backstrom, M.; Thomsson, K. A.; Hansson, G. C. J. Proteome Res. 2009,
8, 538–545
(15) Holme´n Larsson, J. M.; Karlsson, H.; Sjo¨vall, H.; Hansson, G. C. Glycobiology
2009, 19, 756–766
(16) Andersch-Bjo¨rkman, Y.; Thomsson, K. A.; Holme´n Larsson, J. M.; Ekerhovd,
E.; Hansson, G. C. Mol. Cell. Proteomics 2007, 6, 708–716
(17) Olson, F. J.; Backstrom, M.; Karlsson, H.; Burchell, J.; Hansson, G. C.
Glycobiology 2005, 387, 177–191
(18) Thomsson, K. A.; Karlsson, N. G.; Hansson, G. C. J. Chromatogr., A 1999,
854, 131–139
.
.
.
.
.
.
.
Analytical Chemistry, Vol. 82, No. 4, February 15, 2010 1471

Figure 1. Structures of oligosaccharides analyzed with capillary and nano-LC/MS and MS/MS.
(Thermo Electron, San Jose´, CA). No smoothing was applied.
For the porcine colonic mucin oligosaccharides, one MS/MS
spectrum of at least one glycoform of every oligosaccharide
was required for positive identification. The MS/MS spectrum
should contain characteristic fragment ions that could be
assigned as saccharide fragments, even though the full se-
quence in many cases could not be assigned. Fragmentation
annotations applied are based on the nomenclature suggested
,20
by Domon and Costello19
Capillary-LC/MS and MS/MS. The HPLC pump delivered
a flow of 250 µL/min and splitted down in a Valco-T by a 50 cm
× 50 µm i.d. fused silica capillary before the injector, allowing
approximately 3 µL/min through the column. The oligosaccha-
rides (2 µL) were injected onto the column and eluted with an
acetonitrile gradient (A: 8 mM ammonium bicarbonate; B: 100%
acetonitrile). The gradient (0-45% B) was started 3 min after
injection and eluted for 43 min, followed by a wash step with 80%
B and equilibration of the column for 14 min. A 30 cm × 50 µm
i.d. fused silica capillary was used as transfer line to the ion source.
The IonMax standard ESI source on the LTQ mass spectrometer
was equipped with a stainless steel needle kept at -3.5 kV.
Compressed air was used as nebulizer gas. The heated capillary
was kept at 270 °C, and the capillary voltage was -50 V.
Instrument scan settings and MS2 experiment settings were the
same as for nano-LC/MS.
Figure 2. Capillary-LC/MS at pH 8 of standard neutral and acidic
N- and O-glycans. Peak areas of the molecular ions plotted against
the amount injected on the column. Each sample was analyzed three
times. Symbols: b,(A1); +, (A2); ], (sialyl-T antigen); O, (disialyl-T
antigen); 0, (LNF1); (, (Su-Lea); 2,(T-antigen). The vertical order of
individual lines is marked by symbols adjacent to the plot. R2 values
for the standard curves were between 0.98 and 0.99.
curves for the seven saccharides were linear with R2 values
between 0.98 and 0.99. The lowest amounts required to
generate good quality MS/MS spectra of the different com-
pounds was 25 fmol, except for T-antigen, where 50 fmol was
required. The signal responses were different for the individual
glycans, with the large acidic N-glycans ionized most efficiently,
followed by the acidic O-glycans, by the neutral pentasaccharide
LNF1 displaying responses in the range of 25-40% of that of
the N-glycans, and finally by the neutral disaccharide T-antigen
with approximately 5%. The relative responses for the three
sugars T-antigen, sialyl-T, disialyl-T presented here are in line
with the response factors that we previously have published
(0.4:1:1.3, respectively).17 It is not completely clear why the
N-glycans were ionized much more efficiently than the O-
glycans. The large N-glycans have many more hydroxyl groups
which are sites for deprotonation and, therefore, are probably
easier ionized in negative mode MS as compared to the smaller
O-glycans. The presence of one or two acidic residues (sulfate
or carboxyl groups) did not improve ionization significantly
(compare signal response for A1 and A2 or sialyl-T and disialyl-
T), which was expected as these groups are fully deprotonated
at pH 8.
RESULTS
Capillary-LC/MS at pH 8 of Oligosaccharide Standards.
Seven O- and N-glycan alditol standards (T-antigen, LNF 1, sialyl-
T, disialyl-T, Su-Lea, A1, and A2, see Figure 1) were analyzed
with carbon column capillary-LC/MS with an acetonitrile gradient
containing 8 mM ammonium bicarbonate. Increasing amounts
(12.5, 25, 50, 100, and 200 fmol, triple analyses) were injected on
the HPLC column, and the signal response for each saccharide
was plotted against the injected amount (Figure 2). Neutral and
acidic glycans consisting of up to five residues (T-antigen, LNF
1, sialyl-T, disialyl-T, Su-Lea) were detected as singly charged
deprotonated molecular ions ([M - H+]-), whereas the two
larger sialylated biantennary N-glycans A1 and A2 were
detected as doubly charged ions [M - 2H+]2-. The standard
(19) Domon, B.; Costello, C. E. Glycoconj. J. 1988, 5, 387–409
(20) Karlsson, N. G.; Karlsson, H.; Hansson, G. C. J. Mass Spectrom. 1996, 31,
560–572
.
.
1472 Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

and disialyl-T, A1, and A2 were predominantly observed as
doubly charged ions ([M - 2H+]2-) and with a minor amount
of the singly charged form of disialyl-T. Combined full scan
MS and MS/MS spectra from the analysis of disialyl-T (3 fmol)
and Su-Lea (1.5 fmol) are shown in Figure 4. Despite low
amounts, there are sufficient fragment ions to confirm the
monosaccharide units and partial sequence information. The MS/
2-
MS spectrum of the [M - 2H+
]
precursor ion of disialyl-T
(Figure 4B) revealed B-, Y-, and Z-ions, generated by the loss of
NeuAc monosaccharide units, as well as a doubly charged cross
ring fragment 0,2X1R at m/z 372, diagnostic of NeuAc linked R2-6
to the core GalNAc residue.23 In the MS/MS spectrum of the
trisaccharide Su-Lea (Figure 4D), two fragment ions were formed
by cleavage of the two glycosidic bonds, and of which the fragment
at m/z 241 (B1ꢀ) was indicative of sulfate linked to the Gal
residue. LNF1 was not consistently detected in all analyses,
and an MS/MS spectrum was first obtained when 25 fmol was
analyzed (Figure 4F). The mass chromatograms of three sac-
charides analyzed with nano-LC/MS and capillary-LC/MS are
shown in Figure 5. The peak shapes of disialyl-T and Su-Lea were
considerably improved when analyzed with ammonium hy-
droxide at pH 11 (nano-LC/MS) compared to ammonium
bicarbonate at pH 8, although this was not found as a general
phenomena during the analysis of larger oligosaccharides.
Increasing amounts (2, 5, 10, and 20 fmol, two analyses) of
the acidic glycan standards were injected on the HPLC column,
and the signal response for each saccharide was plotted against
the injected amount (Figure 6). The standard curves were not as
linear as those obtained with capillary-LC/MS (Figure 2), and
consequently, the R2 values were lower (0.76-0.96). These
results highlight that the signal response with the nano-LC/
MS setup is not as stable as for capillary LC/MS. The lowest
amounts required to generate MS/MS spectra of the different
compounds was in the range of 1-3 fmol for the acidic glycans,
which is approximately 10 times more sensitive than with
capillary LC-MS/MS, and the improvements are in the same
range as previously reported when negative mode capillary-
and nano-LC/MS of glycans at pH 8 was compared.11 A rough
estimation of the signal response based on the standard curves
indicated that sugars with two acidic residues (disialyl-T, A2)
and A1 were ionized most efficiently, followed by the smaller
glycans with one acidic residue (sialyl-T and Su-Lea). Compared
to capillary LC/MS at pH 8 (Figure 2), the relative ionization of
disialyl T was improved, it was mainly detected as a doubly
charged precursor ion, and the signal response compared to
sialyl-T had nearly doubled. It appeared most likely that the
increase of pH from 8 to 11 in the mobile phase supported the
deprotonation of the acidic groups on disialyl-T in the ionization
process and, thus, subsequently improved its ionization. It is,
therefore, hypothesized that the high pH could be most favorable
when analyzing oligosaccharides with multiple acidic residues.
This idea was confirmed by the analysis of complex mixtures of
multiply sialylated and sulfated mucin O-glycans from porcine
colon and from human biopsies.
Figure 3. Nano-LC/MS interface constructed in-house, modified from
the interface described by Shen et al.12 The fused silica HPLC column
is connected via a steel screen to the emitter tip in a Valco union
(upper panel). The emitter tip is situated less than 1 mm from the
heated capillary (lower panel).
Nano-LC/MS at pH 11 of Oligosaccharide Standards. To
optimize the nanospray interface, we designed a new nanospray
probe for the LTQ mass spectrometer (Figure 3), inspired by the
work of Shen et al.12 Dead volumes are kept low, and the emitter
tip is positioned less than 1 mm from the heated capillary, which
enables almost all of the analyte to enter the mass spectrometer
compared to the commercially available interfaces. The shape of
the Taylor cone is monitored by a stereo microscope mounted
above the nanospray housing. This interface is currently also used
on a routine basis for positive ion mode nano-LC/FTICR MS
analyses for proteomics peptide analyses.21,22
Dilution series of oligosaccharide standards were analyzed with
nano-LC/MS with an acetonitrile gradient containing 0.04% am-
monium hydroxide generating an approximate pH of 11. The
T-antigen standard was not analyzed, as this is sometimes not
retarded on the carbon column at this pH. Su-Lea and sialyl-T
were observed as singly charged molecular ions ([M - H+]-),
(21) Johansson, M. E. V.; Phillipson, M.; Petersson, J.; Holm, L.; Velcich, A.;
Hansson, G. C. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 15064–15069
.
(22) Bell, A. W.; Deutsch, E. W.; Au, C. E.; Kearney, R. E.; Beavis, R.; Sechi, S.;
Nilsson, T.; Bergeron, J. J.; Beardslee, T. A.; Chappell, T.; Meredith, G.;
Sheffield, P.; Gray, P.; Hajivandi, M.; Pope, M.; Predki, P.; Kullolli, M.;
Hincapie, M.; Hancock, W. S.; Jia, W.; Song, L.; Li, L.; Wei, J.; Yang, B.;
Wang, J.; Ying, W.; Zhang, Y.; Cai, Y.; Qian, X.; He, F.; Meyer, H. E.;
Stephan, C.; Eisenacher, M.; Marcus, K.; Langenfeld, E.; May, C.; Carr,
S. A.; Ahmad, R.; Zhu, W.; Smith, J. W.; Hanash, S. M.; Struthers, J. J.;
Wang, H.; Zhang, Q.; An, Y.; Goldman, R.; Carlsohn, E.; van der Post, S.;
Hung, K. E.; Sarracino, D. A.; Parker, K.; Krastins, B.; Kucherlapati, R.;
Bourassa, S.; Poirier, G. G.; Kapp, E.; Patsiouras, H.; Moritz, R.; Simpson,
R.; Houle, B.; Laboissiere, S.; Metalnikov, P.; Nguyen, V.; Pawson, T.; Wong,
C. C.; Cociorva, D.; Yates Iii, J. R.; Ellison, M. J.; Lopez-Campistrous, A.;
Semchuk, P.; Wang, Y.; Ping, P.; Elia, G.; Dunn, M. J.; Wynne, K.; Walker,
A. K.; Strahler, J. R.; Andrews, P. C.; Hood, B. L.; Bigbee, W. L.; Conrads,
T. P.; Smith, D.; Borchers, C. H.; Lajoie, G. A.; Bendall, S. C.; Speicher,
K. D.; Speicher, D. W.; Fujimoto, M.; Nakamura, K.; Paik, Y. K.; Cho, S. Y.;
Kwon, M. S.; Lee, H. J.; Jeong, S. K.; Chung, A. S.; Miller, C. A.; Grimm,
R.; Williams, K.; Dorschel, C.; Falkner, J. A.; Martens, L.; Vizcaino, J. A.
Nat. Methods 2009, 6, 423–430.
(23) Robbe, C.; Capon, C.; Coddeville, B.; Michalski, J. C. Rapid Commun. Mass
Spectrom. 2004, 18 (4), 412–420
.
Analytical Chemistry, Vol. 82, No. 4, February 15, 2010 1473

Figure 4. Nano-LC/MS and MS/MS of oligosaccharide standards. Combined full scan mass spectra of disialyl-T (3 fmol) detected as a doubly
charged precursor ion ([M - 2H+]2-) at m/z 482.8 eluting at 17.3 min (A), Su-Lea (1.5 fmol) detected as a singly charged precursor ion ([M -
H+]-) at m/z 610.3 eluting at 16.6 min (B), and LNF1 (25 fmol) detected as a singly charged precursor ion ([M - H+]-) at m/z 854.3 eluting at
21.8 min (C). The corresponding MS/MS spectra of the parent ions are shown in (B), (D), and (F), respectively. *: Background ion.
Figure 6. Nano-LC/MS at pH 11 of standard acidic N- and
O-glycans. Peak areas of the molecular ions plotted against the
amount injected on the column. Each sample was analyzed twice.
Symbols: ], (sialyl-T antigen); (, (Su-Le a); b, (A1); +, (A2); O,
(disialyl-T antigen). R2 values for the standard curves were 0.76-0.96.
Figure 5. Mass chromatograms of standard oligosaccharides
analyzed with nano-LC/MS at pH 11 and capillary-LC/MS at pH 8.
Details are described in the Experimental Section.
dominated by different oligosaccharides (Figure 7). The two
largest peaks in the nano-LC/MS chromatogram in the upper
Nano- and Capillary-LC/MS and MS/MS of Porcine
Colonic Mucin Oligosaccharides. Oligosaccharides from ap-
proximately 1.25 µg of purified porcine colonic mucin were
analyzed with capillary- and nano-LC/MS in ammonium bicarbon-
ate at pH 8 and ammonium hydroxide at pH 11, respectively. The
samples were analyzed twice and at different occasions, giving
reproducible results. The base peak chromatograms of the
O-glycans analyzed with the two methods were found to be
panel corresponded to two monosulfated oligosaccharides at m/z
708 and 813, followed by a number of peaks corresponding to
mono-, di-, and trisulfated oligosaccharides.
The base peak chromatogram obtained with capillary-LC/MS
(lower panel) was on the contrary dominated by neutral and
monosulfated saccharides, and di- and trisulfated glycans were
observed as minor peaks with intensities less than 10% of the
largest glycan in the chromatogram. As the same amount of
1474 Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

Figure 7. Base peak chromatograms of mucin oligosaccharides from porcine colon mucins analyzed with nano-LC/MS and MS/MS at pH 11
and with capillary-LC/MS at pH 8. The most abundant components in major peaks are annotated, and deduced sequences of major
oligosaccharides are displayed. Sulfated [M - nH+]n- ions are highlighted in bold. Detected oligosaccharides and their monosaccharide composition
are summarized in Table 1 and given in detail in Table S-1 in the Supporting Information.
generation Q-TOF mass spectrometer when we detected only 38
Table 1. Porcine Colonic O-linked Mucin
monosulfated and two disulfated components.24
Oligosaccharides Analyzed with Nano-LC/MS at pH 11
and Capillary LC/MS at pH 8a
Capillary-LC/MS at pH 8 revealed 45 neutral oligosaccharides
number of sequencesc
nano-LC/MS capillary-LC/MS
consisting of 2-7 residues, compared to only 21 detected with
nano-LC/MS at pH 11. These results are consistent with the
analyses of the neutral oligosaccharide LNF1 (Figure 5), suggest-
ing that there is no improved sensitivity when analyzing neutral
glycans with nano-LC/MS at high pH, instead it is rather the
opposite. Approximately the same number of oligosaccharides
with one or two acidic residues were detected with the two
methods (75/77 with one acidic residue and 55/48 with two acidic
residues with capillary and nano-LC/MS, respectively). However,
the MS/MS spectra coverage with capillary-LC/MS of these
glycans was not as complete as with nano-LC/MS (not shown).
Only five oligosaccharides with three acidic residues were
detected with capillary-LC/MS, whereas 19 were observed with
nano-LC/MS. The triply sulfated oligosaccharide at m/z 445 (Mr
1339) was observed as a prominent component with nano-LC/
MS but was not detected at all with capillary-LC/MS (Figure
number of acidicb residues
0
1
2
3
21
75
55
19
45
77
48
5
a Number of detected components corresponding to individual
oligosaccharide sequences. b Sulfate and/or sialic acid residues.
c Complete list with number of glycan isoforms and their deduced
monosaccharide composition is found in Table S-1 in the Supporting
Information.
oligosaccharides were injected on the columns, these results
supported the hypothesis that multiply charged glycans are ionized
more efficiently with nano-LC/MS at high pH. The oligosaccha-
rides detected with the two methods are summarized in Table 1
and given in detail in Table S-1 in the Supporting Information. A
total of 45 neutral and 156 mono-, di-, and triply sulfated and/or
sialylated oligosaccharides were identified, of which most were
sulfated. This is far more than our previously published analysis
of the same porcine colonic mucin oligosaccharides using a first
7
). The MS/MS spectrum of the hexasaccharide at m/z 445 is
presented in Figure 8. The three sulfate groups, which strongly
promoted charge remote fragmentation, gave rise to singly,
doubly, and triply charged fragment ions. The doubly charged
B-ions at m/z 261.42- (B2ꢀ/Y3ꢀ′) and m/z 334.42- (B2ꢀ) confirmed
the presence of a disulfated Hex-HexNAc branch with an
additional fucose, an interpretation also supported by a promi-
(24) Thomsson, K. A.; Karlsson, H.; Hansson, G. C. Anal. Chem. 2000, 72,
4543–4549
.
Analytical Chemistry, Vol. 82, No. 4, February 15, 2010 1475

indicative of a type 2 configuration (Galꢀ1-4GlcNAc), as well
as for assigning the sulfate to C-6 of the HexNAc residue.
Finally, the ions at m/z 241 (B1R or B1ꢀ) and m/z 547.62- (Y2R/
Y2ꢀ) supported a doubly branched oligosaccharide with the
sequence: SO3--Hex-4(Fuc-3)(SO3--6)HexNAc-(SO3--Hex-Hex-
NAc)HexNAcol.
Nano- and Capillary-LC/MS and MS/MS of the MUC2
Mucin Oligosaccharides from a Single Human Colon Mil-
limeter Sized Biopsy. The MUC2 mucin oligosaccharides from
a human colon biopsy were analyzed with capillary and nano-LC/
MS and MS/MS as shown in Figure 9. The mucin O-glycans were
highly sialylated in contrast to the highly sulfated porcine colon
glycans. The base peak chromatogram from capillary-LC/MS at
pH 8 (Figure 9, lower panel) was dominated by two isoforms of
HexNAc-3(NeuAc-6)GalNAcol at m/z 716, and oligosaccharides
with two or more acidic residues were small components. In
contrast, the largest peaks in the nano-LC/MS base peak chro-
matogram were the doubly charged molecular ions at m/z 551.82-
and m/z 685.82-, interpreted as glycans with two sialic acid
residues or one sialic acid and one sulfate, respectively. In
addition, numerous minor components corresponding to gly-
cans with multiple acidic residues were detected. Many of these
glycans have not been observed before in a human colon as
recently described.15 A total of 17 triply and one tetra sialylated
and sulfated glycan were found, of which a majority could be
sequenced by MSn experiments. The formation of multiply
charged molecular ions was more prominent with nano-LC/
MS at pH 11 than with capillary LC/MS at pH 8. An example
Figure 8. CID spectrum from nano-LC/MS/MS of a major trisulfated
oligosaccharide from porcine colonic mucins (Figure 7, upper panel)
and with the molecular mass of 1339.3 Da. CID was performed on
the triply charged parent ion at m/z 445.6, and the sequence was
deduced to SO3--Hex-4(Fuc-3)(SO3--6)HexNAc-(SO3--Hex-HexNAc-)
HexNAcol.
nent ion at m/z 196.82-, formed by a cleavage in the sulfated
HexNAc residue (3,5A2ꢀ). This fragment ion was found to be
diagnostic for disulfated Hex-HexNAc residues and was also
Figure 9. Base peak chromatograms of mucin oligosaccharides from the MUC2 mucin from a millimeter sized human colon biopsy analyzed
with nano-LC/MS at pH 11 and with capillary-LC/MS at pH 8. Major components are annotated with the observed molecular ions ([M - nH+]n-),
and some deduced oligosaccharide sequences are shown. For all details, see Larsson et al.15 Some of the deduced glycan structures have
also been found by others in human colon.27
1476 Analytical Chemistry, Vol. 82, No. 4, February 15, 2010

of this is a nonasaccharide containing three acidic residues (3
HexNAc, 2 Gal, 3 NeuAc, GalNAcol) with the mass of 2029.7
Da. This was detected as a triply charged ion at m/z 675.63-
([M - 3H+]3-) with nano-LC/MS and as a doubly charged ion
at m/z 1013.92- ([M - 2H+]2- with capillary LC/MS (Figure
nanospray voltage (-1.6 kV) between the two unions. Despite
this, we have not observed technical difficulties with this design
as been discussed by others.25,26 This may be due to a combination
of several factors, the low ionic strength of the mobile phase, the
high pH, the use of negative mode MS, and also the relatively
low voltage used in nanospray.
9). In nano-LC/MS, some oligosaccharides with three acidic
residues were observed both as doubly and triply charged
precursor ions, something that could be a disadvantage as it will
slightly increase the detection limit of an individual oligosaccha-
ride. However, we have also found this useful, as the MS/MS
spectra of the two parent ions can be quite different and
consequently aid in the sequence deduction (not shown).
The enhanced detection of the acidic glycans by nano-LC/
MS allows for detailed structural characterization of these com-
ponents with MSn experiments.15 For multiply sialylated oli-
gosaccharides, sequential MSn experiments are often required
to eliminate the sialic acid residues before the remaining part
of the saccharide is fragmented. This is often not possible to
achieve with capillary-LC/MS but is easily obtained with nano-
LC/MS. Our nano-LC/MS approach is a valuable alternative
to negative ion mode capillary LC/MS for highly sensitive
analyses of complex negatively charged oligosaccharide
mixtures.
DISCUSSION AND CONCLUSIONS
Porous graphitic carbon (PGC) chromatography of O-glycans
using negative ion mode nano-LC/MS with a mobile phase at high
pH was studied and compared to capillary LC/MS at pH 8. The
results highlight that the signal response of individual glycans
do vary with size and number of acidic residues when analyzed
with both methods. Acidic oligosaccharides were ionized more
efficiently than neutral species with both methods. This should
be taken into account when profiling complex oligosaccharide
mixtures and makes it necessary to use correction factors to obtain
a better estimation of relative amounts, as we have done before.17
Nano-LC/MS at high pH was even more selective toward the
detection of glycans with multiple acidic residues. This was
demonstrated by the analysis of oligosaccharide standards, as well
as complex mixtures of mucin oligosaccharides from porcine and
human colonic mucins.
ACKNOWLEDGMENT
This work was supported by the Swedish Research Council
(Nos. 7461, 21027, and 342-2004-4434), The Swedish Cancer
Foundation,TheKnutandAliceWallenbergFoundation(KAW2007.0118),
IngaBritt and Arne Lundberg Foundation, Sahlgren’s University
Hospital (LUA-ALF), EU-FP7 IBDase, Wilhelm and Martina
Lundgren’s Foundation, and The Swedish Foundation for Strategic
Research-The Mucosal Immunobiology and Vaccine Center (MI-
VAC).
SUPPORTING INFORMATION AVAILABLE
Additional information as noted in text. This material is
available free of charge via the Internet at http://pubs.acs.org.
The graphitized carbon nanocolumn is in electrical contact with
the emitter tip in our interface design and is floating at the
(25) Tornkvist, A.; Nilsson, S.; Amirkhani, A.; Nyholm, L. M.; Nyholm, L. J. Mass
Spectrom. 2004, 39 (2), 216–222
(26) Pabst, M.; Altmann, F. Anal. Chem. 2008, 80 (19), 7534–7542
(27) Robbe, C.; Capon, C.; Coddeville, B.; Michalski, J. C. Biochem. J. 2004,
384, 307–316
.
Received for review November 13, 2009. Accepted
January 4, 2010.
.
.
AC902602E
Analytical Chemistry, Vol. 82, No. 4, February 15, 2010 1477