Glycobiology, 2015, vol. 25, no. 7, 714–725
doi: 10.1093/glycob/cwv011
Advance Access Publication Date: 12 February 2015
Original Article
Original Article
A new sequencing approach for N-unsubstituted
heparin/heparan sulfate oligosaccharides
Qun Tao Liang, Xiao Mao Xiao, Jian Hui Lin, and Zheng Wei1
Institute of Glycobiochemistry, National Engineering Research Centre of Chemical Fertilizer Catalyst, Fu Zhou University, Fu Zhou 350002, PR China
1
To whom correspondence should be addressed: Tel: +86-591-83770818; Fax: +86-591-83738808;
e-mail: [email protected]
Received 2 July 2014; Revised 26 January 2015; Accepted 7 February 2015
Abstract
The rare N-unsubstituted glucosamine (GlcNH3 þ ) residues in heparan sulfate (HS) have important
biological and pathophysiological roles. Because of their low natural abundance, the use of chemically generated, structurally defined, N-unsubstituted heparin/HS oligosaccharides can greatly contribute to the investigation of their natural role in HS. However, the sequencing of mixtures of
chemically generated oligosaccharides presents major challenges due to the difficulties in separating isomers and the available detection methods. In this study, we developed and validated a simple
and sensitive method for the sequence analysis of N-unsubstituted heparin/HS oligosaccharides.
This protocol involves pH 4 nitrous acid (HNO2) degradation, size-exclusion HPLC and ion-pair
reversed-phase liquid chromatography-ion trap/time-of-flight mass spectrometry (IPRP-LC-ITTOF
MS). We unexpectedly found that absorbance at 232 nm (normally used for specific detection of
C4–C5 unsaturated oligosaccharides) was, in most cases, still sufficiently sensitive to also simultaneously detect saturated oligosaccharides during HPLC, thus simplifying the positional analysis of
GlcNH3 þ residues. The IPRP-LC-ITTOF MS system can supply further structural information leading
to full sequence determination of the original oligosaccharide. This new methodology has been used
to separate and sequence a variety of chemically generated, N-unsubstituted dp6 species containing
between 1 and 3 GlcNH3 þ residues per oligosaccharide in different positional combinations. This
strategy offers possibilities for the sequencing of natural N-unsubstituted oligosaccharides from
HS and should also be applicable, with minor modification, for sequencing at N-sulfated residues
using alternative pH 1.5 HNO2 scission.
Key words: heparin, HNO2 degradation, MS, N-unsubstituted disaccharide, sequencing
Introduction
Heparin and heparan sulfate (HS) are complex, sulfated glycosaminoglycans, which have important biological activities in developmental
processes, angiogenesis, blood coagulation, cell adhesion and tumor
metastasis (Perrimon and Bernfield 2000; Robinson and Stringer
2001; Liu, Shriver, Pope, et al. 2002; Liu, Shriver, Venkataraman,
et al. 2002; Couchman 2003). These involve interactions with a
wide variety of proteins, that is, enzymes, cytokines, growth factors
and extracellular matrix proteins (Salmivirta et al. 1996; Casu and
Lindahl 2001; Esko and Selleck 2002; Liu and Thorp 2002).
HS and heparin share a common biosynthetic pathway, being initially synthesized as repeating disaccharides of alternating β1 → 4 glucuronic acid (GlcA) and α1 → 4 N-acetylglucosamine (GlcNAc)
(Lindahl et al. 1998; Casu and Lindahl 2001; Esko and Lindahl
2001; Esko and Selleck 2002; ). This nonsulfated precursor is then
modified by coordinated N-deacetylation/N-sulfation of GlcNAc residues by N-deacetylase/N-sulfotransferase enzymes (NDST), forming
β-D-N-sulfoglucosamine (GlcNS) via an N-unsubstituted glucosamine
(GlcNH3 þ ) intermediate. Subsequently, further modifications occur,
i.e. C5-epimerization of GlcA to α1 → 4 iduronic acid (IdoA),
© The Author 2015. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: [email protected]
714
715
A new sequencing approach for N-unsubstituted Hep/HS
2-O-sulfation of hexuronic acid ( primarily IdoA), 6-O-sulfation of
GlcNS or GlcNAc and finally the rare, but functionally important,
3-O-sulfation of GlcNS by 3-O-sulfotransferases (HS 3-OST).
It has become clear that the GlcNH3 þ residue exists in mature HS/
heparin, which may be due to an occasional interruption of the
NDST-mediated, catalytic linkage between N-deacetylation and
re-N-sulfation, thereby trapping the intermediate GlcNH3 þ . This
may be exacerbated under conditions of limiting availability of
3′-phosphoadenosine 5′-phosphosulfate, which drives the final sulfation (Carlsson et al. 2008). In general, the content of GlcNH3 þ in HS
species is low, but variable across different tissues. It mostly ranges
from 0.2 to 4% of disaccharide units (Toida et al. 1997; Westling
and Lindahl 2002; Rees et al. 2005; Wei et al. 2005; Shi and Zaia
2009), but reaches a moderately high 12% in bovine kidney HS
(Wei et al. 2005). Recently, it was reported that HS chains from
adult mouse kidney have an exceptionally high GlcNH3 þ content
(30–57%, depending upon gender) (Murali et al. 2011). However,
this analysis was dependent upon the UV quantification of significant,
very early eluting peaks from strong-anion-exchange chromatography
of enzyme-generated disaccharides, which can easily be affected by
coelution of other UV-absorbing materials. Verification using other
detection methods, and/or analysis of nitrous acid-generated chain
fragmentation, would have been useful.
The rare GlcNH3 þ residues are implicated in important cellbiological and pathophysiological phenomena. Their presence correlates with the ability of bovine and human endothelial HS to bind
L-selectin (Norgard-Sumnicht and Varki 1995). GlcNH3 þ residues
have also been identified as additional targets for 3-O-sulfation. The
specific HS 3-OST-3A isoform generates a sequence that is utilized as a
binding site by the herpes simplex virus (HSV) glycoprotein D, thus
making cells susceptible to HSV-1 entry (Liu et al. 1999; Shukla
et al. 1999; Liu, Shriver, Pope, et al. 2002; Liu, Shriver, Venkataraman, et al. 2002). A highly sulfated hexasaccharide with an internal
3-O-sulfated GlcNH3 þ (i.e. GlcA±2S-GlcNS-IdoA2S-GlcNH3 þ 3S
±6S-GlcA±2S-GlcNS) is the binding site for the HSV gD protein
(Shukla et al. 1999). Similarly, 3-O-sulfation by the HS 3-OST-3B isoform provides specific binding sites for cyclophilin B (CyPB) on responsive cells. The GlcNH3 þ residue is specifically positioned two
saccharides from the nonreducing end of the CyPB-binding octasaccharide (Vanpouille et al. 2007). In addition, it has been proposed
that GlcNH3 þ residues provide cleavage sites in HS for the endogenous NO-derived nitroxyl anion, contributing to a recycling mechanism for glypican-1 (Cheng et al. 2012).
GlcNH3 þ residues in heparins are also now a particularly hot subject in the pharmaceutical industry. Additional GlcNH3 þ residues can
be generated by thermal stress and autoclave sterilization of heparin.
This can modify, and compromise, the highly specific, endogenous
antithrombin III-binding sites, resulting in a decrease in the anticoagulant potency that is the main medicinal application of heparin (Beaudet et al. 2011; Fu et al. 2014).
The location of GlcNH3 þ residues and the nature of their surrounding sequences might contribute to their enzyme susceptibility,
antibody recognition and protein interactions. Previously, we demonstrated that the various heparinase enzymes have differential specificities toward GlcNH3 þ residues. Heparinase I cannot cleave at such
residues. In contrast, heparinases II and III can do so if the relevant
disaccharide is O-sulfated or nonsulfated, respectively (Wei et al.
2005). Thus, heparinases can be used specifically to generate a series
of heparin/HS oligosaccharides containing GlcNH3 þ residues, with
potential use in protein interaction and functional studies (Wei et al.
2011). The exact determination of the position of GlcNH3 þ residues
within oligosaccharide sequences is important for structure–function
studies. However, the heparinases have little value in direct sequencing
of such oligosaccharides as none of them specifically require GlcNH3 þ
residues.
The existing methods for the sequencing of heparin/HS oligosaccharides rely upon using exoglycosidase digestions of radioactive, or
fluorescent end-labeled, oligosaccharides with subsequent PAGE,
HPLC or LC/CE separation of the resulting fragments (Merry et al.
1999; Turnbull et al. 1999; Stringer et al. 2003; Li et al. 2012). The
established sequencing methods mainly focus on structural determination of the dominant structures within heparin/HS, whereas structural analysis of the rarer GlcNH3 þ residues has been much more
limited. Nitrous acid degradation has been used in the sequencing of
heparin/HS oligosaccharides by HPLC or electrophoresis; however, it
often requires radioisotopic labeling (Merry et al. 1999; Turnbull et al.
1999). At pH ≤ 1.5, scission occurs at GlcNS residues, whereas at pH
4.0 scission occurs at GlcNH3 þ residues. Using pH 4.0 nitrous acid
scission and radioisotope detection, Westling and Lindahl (2002) determined the distributions of GlcNH3 þ residues within HS chains.
Until now, the structural analysis of GlcNH3 þ residues in heparin/
HS has been mainly limited to disaccharide analysis, because of
their low abundance and lack of a useful separation and detection
method, which makes sequencing determination a very challenging
task.
Mass spectroscopy is now becoming a much more useful tool for the
structural analysis of heparin/HS oligosaccharides, because of its high
detection sensitivity and molecular specificity (Wolff et al. 2007;
Schenauer et al. 2009; Huang et al. 2013). LC–MS/MS, in particular,
clearly has the potential now to sequence most oligosaccharides, especially with the recent development of a computational framework for
dealing with the complex data generated (Hu et al. 2014).
In this study, we developed a new method for the sequencing
analysis of a heparin/HS oligosaccharide library containing N-unsubstituted residues. This method involves HNO2 degradation, size-exclusion
HPLC with UV detection at 232 nm, followed by the relatively simple
IPRP-LC-ITTOF MS. Using this new protocol, we successfully sequenced six N-unsubstituted dp6 oligosaccharides derived from partially de-N-sulfated heparin.
Results and discussion
Method development for sequencing of N-unsubstituted
oligosaccharides
Herein, we present an optimized method that enables the sequencing of
partially or fully N-unsubstituted heparin oligosaccharides involving
HNO2 degradation at pH 4, size-exclusion HPLC (SE-HPLC) and
LC–MS-ITTOF. This method was developed in three stages: (i) HNO2-generated dp2s, with or without an unsaturated bond between C4 and
C5, were prepared and their structures were analyzed by a combination
of SE-HPLC and LC–MS-ITTOF; (ii) N-unsubstituted heparin dp4s of
known structure were used to develop and optimize the sequencing
method; (iii) this method was then used to separate and sequence the
full range of chemically generated, N-unsubstituted dp6s.
SE-HPLC of HNO2-generated dp2
The majority of oligosaccharides used in this study were initially isolated from low-molecular-weight heparin, i.e. heparin that has already
been partially degraded by heparinase I, which introduces an unsaturated bond between C4 and C5 of the nonreducing terminal uronate
716
residues. Consequently, subsequent cleavage with HNO2 will generate
both Δ-unsaturated and saturated products.
The unsaturated, tri-sulfated dp2, ΔHexA(2S)-GlcNS(6S), was
purified from low-molecular-weight heparin, and then de-N-sulfated
to yield ΔHexA(2S)-GlcNH3 þ (6S). HNO2 at pH 4 specifically deaminates GlcNH3 þ residues, converting them to reducing terminal
2,5-anhydromannose (aMan) residues. Thus, the latter dp2 was treated with HNO2 and then borohydride reduced to obtain the unsaturated, deaminated dp2, ΔHexA(2S)-aMan(6S)R.
Conversely, to prepare the saturated dp2, HexA(2S)-aMan(6S)R,
intact heparin was de-N-sulfated, followed by HNO2 degradation at
pH 4 and then borohydride reduction.
QT Liang et al.
The separation of disaccharides by SE-HPLC is based on hydrodynamic volume, which is roughly proportional to the number of
sulfate groups present. The combination of SE-HPLC and MS has
been successfully used to separate and characterize disaccharides
(Shi and Zaia 2009). Because the initial dp2, ΔHexA(2S)-GlcNS
(6S), is unsaturated, the HNO2 reaction product is also detectable
by UV absorbance at 232 nm on SE-HPLC. As shown in Figure 1A,
a major product peak (designated as M1) was obtained at 30.5–
31.5 min. This is very close to the elution position of a known
ΔHexA(2S)-GlcNH3 þ (6S) standard ( peak at 30.9 min compared
with 29.5 min for a known ΔHexA(2S)-GlcNS(6S) standard; data
not shown), indicating that it contains two sulfate groups, i.e. a
Fig. 1. Preparation and characterization of HNO2-generated disaccharides. SE-HPLC profiles of (A) pH 4 HNO2-treated and reduced ΔHexA(2S)-GlcNH3 þ (6S) to
generate ΔHexA(2S)-aMan(6S)R, and (B) HNO2-degraded and reduced de-N-sulfated native heparin to generate HexA(2S)-aMan(6S)R. M1 and M2 fractions
indicated by the horizontal bars were pooled for IPRP-LC-ITTOF analysis. The mass spectra (C and D, respectively) and EIC profiles (E and F, respectively) are
shown. Note that M1 gave two peaks with an RT of 7.471 min (E), and M2 gave three peaks at an RT of 7.4807 min (F), both due to in-source losses of SO3. This
figure is available in black and white in print and in colour at Glycobiology online.
717
A new sequencing approach for N-unsubstituted Hep/HS
2-O-sufate and a 6-O-sulfate group, and consistent with it being
ΔHexA(2S)-aMan(6S)R.
The scission products from HNO2 cleavage of de-N-sulfated
native heparin were analyzed by SE-HPLC at absorbance 232 nm
(Figure 1B). A series of peaks were detected with one peak (designated
as M2) being eluted at 30.2–31.6 min, which is a similar elution
position as both the ΔHexA(2S)-GlcNH3 þ (6S) standard and peak
M1 (in Figure 1A), suggesting that this dp2 has two sulfate groups
as well. We unexpectedly found that this HNO2-scission product
has a significant absorbance at 232 nm, even though it lacks unsaturation between C4 and C5, suggesting that all HNO2-generated products could be detected by absorbance at 232 nm in this system,
though with different sensitivities.
The M1 and M2 fractions were individually pooled, and structurally analyzed further by ion-pair reversed-phase liquid chromatography
(IPRP-LC) coupled with ion trap/time-of-flight mass spectrometry
(ITTOF), i.e. IPRP-LC-MS-ITTOF.
IPRP-LC-MS-ITTOF analysis
Sample analysis utilized both mass spectra and extracted ion chromatography (EIC) profiles. The separation mechanism of IPRP-LC differs
from that of SE-HPLC, being based upon a combination of the size,
sulfation and isomerization differences of scission products. The
ion-pair reagent in the LC–MS system achieves a highly efficient
separation, together with only a limited sulfate group loss during
MS analysis (Doneanu et al. 2009), allowing IPRP-LC-MS-ITTOF
to provide a more detailed structural analysis of the scission products.
IPRP-LC-MS-ITTOF analysis of the M1 fraction gave a peak at m/z
480.9916 (Figure 1C), eluting at a retention time (RT) of 7.471 min in
the EIC profile (Figure 1E), indicating that M1 is the unsaturated ΔHexA
(2S)-aMan(6S)R (see structure in Supplementary data, Figure S4). A
peak of m/z 401.0359, corresponding to the [M1−SO3−H]− ion, was
also detected with an RT of 7.471 min (Figure 1E), indicating the loss
of one sulfate from M1 in the MS ion source.
The mass spectrum of the M2 fraction (Figure 1D) showed a peak at
m/z 498.9975, corresponding to the [M2−H]− ion that eluted at an RT
of 7.4807 min in EIC (Figure 1F), indicating that M2 is the saturated
HexA(2S)-aMan(6S)R (see structure in Supplementary data, Figure S4).
A further two peaks, corresponding to m/z 419.0466 and 339.0861,
were also observed with an RT of 7.4807 min, due to in-source loss of
sulfate groups from M2. EIC also revealed two peaks, of m/z 419.0466,
with RT values from 5.0 to 6.1 min, which might be saturated HexA
(2S)-2,5-anhydromannitol and HexA-aMan(6S)R (data not shown)
contaminants from the SE-HPLC fractionation.
The combination of SE-HPLC analysis with the above result confirmed that the saturated HexA(2S)-aMan(6S)R is able to be detected
at UV 232 nm during SE-HPLC. HNO2 specifically cleaves N-sulfated
and GlcNH3 þ residues at different pH optima to generate saturated
disaccharides; however, the disadvantage of HNO2 degradation is
that it often requires the use of radioisotopes to yield detectable products (Merry et al. 1999; Westling and Lindahl 2002). Here we show
that simple UV detection at 232 nm can be satisfactory for analysis of
HNO2-generated products if sufficient, relatively clean material is
available, together with the use of relatively non-UV-absorbing eluents. Absorbance at 206 nm (or at both wavelengths) can be alternatively employed to give greater sensitivity. Absorbance at 232 nm can,
however, be useful for helping to distinguish between saturated and
unsaturated oligosaccharides, as the latter have a higher molar absorption at this wavelength. It may also be less susceptible to interference
from other organic molecules than absorbance at 206 nm.
Sequencing of model N-unsubstituted tetrasaccharides
In our previous research, we prepared N-unsubstituted tetrasaccharides containing 1 or 2 GlcNH3 þ residues by partial de-N-sulfation
of the fully sulfated heparin dp4. NMR analysis (Wei et al. 2011)
showed that their structures were ΔHexA(2S)-GlcNH3 þ (6S)-IdoA
(2S)-GlcNS(6S) (Figure 2A) and ΔHexA(2S)-GlcNH3 þ (6S)-IdoA
(2S)-GlcNH3 þ (6S) (Supplementary data, Figure S1A), named dp4-1
and dp4-2, respectively. Herein, these two oligosaccharides were
used as model oligosaccharides to develop the sequencing method
involving HNO2 degradation, SE-HPLC and IPRP-LC-MS-ITTOF.
Dp4-1 was degraded by HNO2 at pH 4, and then reduced with
NaBH4. After removing the major chemical reagents, the scission products were analyzed by SE-HPLC at 232 nm. Theoretically, as dp4-1
has one internal GlcNH3 þ , two disaccharides, ΔHexA(2S)-aMan
(6S)R and HexA(2S)-GlcNS(6S), should be generated by HNO2 depolymerization. In the SE-HPLC profile of the scission products
from dp4-1 (Figure 2B), one broad peak centered at 30.9 min was obtained (labeled as dp2), indicating the presence of more than one dp2
structure. The dp2 fraction was collected, and further analyzed by
IPRP-LC-MS-ITTOF (Figure 2C and D). Two peaks were observed
at m/z 480.9911 and 401.0359 (eluting at RT of 7.459 min) corresponding to the molecular ions [M1−H]− and [M1−SO3−H]− derived
from ΔHexA(2S)-aMan(6S)R. Two further peaks appeared at m/z
593.9746 and 514.0167, with an RT of 8.0757 min in the EIC profile
(Figure 2D), which correspond to the molecular ions [M3−H]− and
[M3−SO3−H]−, indicating that the saturated HexA(2S)-GlcNS(6S)
(M3 in Supplementary data, Figure S4) is also present. Thus MS analysis experimentally confirmed the presence of the two expected disaccharides (namely M1 and M3) in the dp2 fraction, even though they
eluted as a single broad peak in SE-HPLC. This suggests a poor resolution of these two species, and/or a relatively poorer UV232 nm absorbance for the saturated dp2 compared with the unsaturated one.
The dp4-2 species contains two GlcNH3 þ residues, so two identically sulfated disaccharides should be generated by HNO2 cleavage,
though one would be unsaturated (i.e. ΔHexA(2S)-aMan(6S)R) and
the other saturated (i.e. HexA(2S)-aMan(6S)R). The SE-HPLC and
IPRP-LC-MS-ITTOF analyses confirmed that two dp2s, corresponding to unsaturated M1 (ΔHexA(2S)-aMan(6S)R) and saturated M2
(HexA(2S)-aMan(6S)R), existed in the HNO2-scission products (for
details, see Supplementary data, Figure S1). However, SE-HPLC was
clearly not able to separate M1 and M2, presumably because of their
minimal size difference.
HNO2 degradation has previously been used for the sequencing of
radiolabeled heparin/HS oligosaccharides, by using HPLC and electrophoretic separations with radioisotopic detection (Merry et al.
1999; Turnbull et al. 1999). However, these methods required complex exoglycosidase digestions to deduce the sequences. Herein, the
above results demonstrate that MS is a powerful alternative technique
that can resolve species that were not resolved by SE-HPLC, thereby
providing full structural information for the HNO2-scission products
leading to a sequence determination.
Preparation and characterization of N-unsubstituted
hexasaccharide library
It has been demonstrated that the sulfated disaccharide sequence IdoA
(2S)-GlcNH3 þ (±6S) is a target for the HS 3-OST-3 enzyme, which
generates binding sites for the gD glycoprotein of HSV-1 (Liu, Shriver,
Pope, et al. 2002; Liu, Shriver, Venkataraman, et al. 2002) and for
CyPB (Vanpouille et al. 2007). This disaccharide could be primarily
located within the highly sulfated domains of HS. Therefore, highly
718
QT Liang et al.
Fig. 3. Subfractionation of the partially de-N-sulfated dp6 mixture by
SAX-HPLC. The fractions indicated by the horizontal bars were individually
pooled for disaccharide and sequencing analyses.
Fig. 2. Sequencing of dp4-1 possessing one GlcNH3 þ residue. ΔHexA
(2S)-GlcNH3 þ (6S)-IdoA(2S)-GlcNS(6S) (A) was cleaved by HNO2 at pH 4, and
then analyzed by SE-HPLC (B). The dp2 fraction indicated by the horizontal
bar was pooled, and further analyzed by IPRP-LC-ITTOF with MS (C) and EIC
(D) detection. Note that two peaks at an RT of 8.0757 min (D) were due to
in-source losses of SO3. This figure is available in black and white in print
and in colour at Glycobiology online.
sulfated oligosaccharides containing potentially sulfated GlcNH3 þ
residues in different positions could be particularly valuable substrates
for testing protein-binding activities and also enzyme susceptibilities.
In our previous study, a series of oligosaccharides with internal
GlcNH3 þ residues were generated by de-N-sulfation of heparin
followed by heparinase I digestion (Wei et al. 2011). In order to obtain
a full range of dp6 oligosaccharides containing a variety of both
number and sequence of GlcNH3 þ residues, the fully sulfated dp6
(ΔHexA(2S)-GlcNS(6S)-HexA(2S)-GlcNS(6S)-HexA(2S)-GlcNS(6S))
was partially de-N-sulfated, and then subfractionated by SAX-HPLC
on a ProPac column (Figure 3). The disaccharide compositions of the
major separated dp6 species were analyzed (summarized in Table I),
and used to calculate the constituent disaccharide ratios for each
species.
As shown in Figure 3, six major peaks, designated as dp6-3,
dp6-2a, dp6-2b, dp6-2c, dp6-1a and dp6-1b (in order of elution),
were identified in the modified dp6 mixture. As SAX-HPLC separation
is based on charge density differences, dp6 species having shorter retention times possess less sulfation, i.e. more GlcNH3 þ residues. A further peak, eluting at ∼36 min, is a fraction of the original dp6 that has
not been de-N-sulfated at all, i.e. ΔHexA(2S)-GlcNS(6S)-[HexA
(2S)-GlcNS(6S)]2.
Disaccharide analysis of the recovered dp6-3 showed that only one
disaccharide, ΔHexA(2S)-GlcNH3 þ (6S), was present, indicating that
dp6-3 contains three GlcNH3 þ residues. In contrast, disaccharide analyses of dp6-2a, dp6-2b and dp6-2c revealed the presence of two disaccharides, namely ΔHexA(2S)-GlcNH3 þ (6S) and ΔHexA(2S)-GlcNS
(6S), in proportions of 65–67 and 32–34% respectively, i.e. essentially
a 2 : 1 ratio (Table I), suggesting that two GlcNH3 þ residues are present
in all three of these oligosaccharides. As these three, dp6 species have the
same disaccharide compositions, but different retention times on
SAX-HPLC, it suggests that they are isomers containing the three possible sequence permutations of two GlcNH3 þ residues. On the other
hand, both dp6-1a and dp6-1b comprised ΔHexA(2S)-GlcNH3 þ (6S)
and ΔHexA(2S)-GlcNS(6S), in a ratio of 1 : 2, with different retention
times on SAX-HPLC, indicating that they are isomers varying in the
position of a single GlcNH3 þ residue. What is surprising is that, in theory, there should be three possible sequence variants of dp6 species that
contain just one GlcNH3 þ residue, and yet only two peaks were
resolved by SAX-HPLC.
The five partially N-unsubstituted dp6-1 and dp6-2 species detailed above were collected and sequenced by the combination of
HNO2 scission, SE-HPLC and IPRP-LC-MS-ITTOF. Dp6-3 was not
analyzed further as its disaccharide composition indicated a predictable, repetitive sequence.
Sequencing of dp6-2a with two GlcNH3 þ residues
The sequencing results for HNO2-treated dp6-2a are shown in Figure 4.
The SE-HPLC profile of the scission products (Figure 4A) shows two
major peaks (labeled dp2 and dp4), which have the same retention
719
A new sequencing approach for N-unsubstituted Hep/HS
Table I. Disaccharide compositions of the various N-unsubstituted dp6 species purified by SAX-HPLC (Figure 3)
Disaccharide structures
þ
ΔHexA (2S)-GlcNH3 (6S)
ΔHexA (2S)-GlcNS (6S)
Frequency (%) (disaccharide numbera)
dp6-3
dp6-2a
dp6-2b
dp6-2c
dp6-1a
dp6-1b
100 (3)
0 (0)
67.1 (2)
32.9 (1)
65.9 (2)
34.1 (1)
65.6 (2)
34.4 (1)
32.9 (1)
67.1 (2)
31.3 (1)
68.7 (2)
a
Number of constituent disaccharides in the oligosaccharide calculated from the percent disaccharide composition.
Fig. 4. Sequencing of dp6-2a fraction. The dp6-2a fraction was cleaved by HNO2 at pH 4, and then analyzed by SE-HPLC (A). The dp4 and dp2 products were
separately pooled, as indicated by the horizontal bars, for subsequent IPRP-LC-ITTOF analysis. Mass spectra and EIC profiles of the dp4 (B and D) and dp2 (C
and E) fractions allowed the sequence of dp6-2a to be determined (F). This figure is available in black and white in print and in colour at Glycobiology online.
times as standard dp2 and dp4 species, indicating that dp6-2a was
degraded into dp2 and dp4 fragments by pH 4 HNO2. Interestingly,
the UV 232 nm absorbance of the dp4 is higher than that of the dp2,
though they should be present in equimolar proportions. This suggests
that the dp4 may be unsaturated while the dp2 may not be. If so, then
the two GlcNH3 þ residues would be expected to be located in the
internal and reducing end disaccharides (see Figure 4F).
To confirm the detailed structure, the dp4 and dp2 scission products were analyzed by IPRP-LC-MS-ITTOF. The dp4 gave four
major ion peaks of which the peak at m/z 528.4753 (see Figure 4B)
corresponds to the unsaturated dp4, ΔHexA(2S)-GlcNS(6S)-HexA
(2S)-aMan(6S)R (designated as M4; see Supplementary data,
Figure S4). This result accords with that of the SE-HPLC, i.e. one
GlcNH3 þ residue is located in the central disaccharide of dp6-2a.
The m/z 488.4978, 448.5205 and 408.5429 peaks represent the
[M4−SO3−2H]2−, [M4−2SO3−2H]2− and [M4−3SO3−2H]2−
molecular ions, respectively, that all elute with the same RT of
10.035 min in EIC (Figure 4D), i.e. corresponding to loss of sulfate
720
QT Liang et al.
Fig. 5. Sequencing of dp6-2b fraction. The dp6-2b fraction was cleaved by HNO2 at pH 4, and then analyzed by SE-HPLC (A). The dp4 and dp2 products were
separately pooled, as indicated by the horizontal bars, for subsequent IPRP-LC-ITTOF analysis. Mass spectra and EIC profiles of the dp4 (B and D) and dp2
(C and E) fractions allowed the sequence of dp6-2b to be determined (F). This figure is available in black and white in print and in colour at Glycobiology online.
groups from the intact molecular ion [M4−2H]2− in the MS ion
source. As the RP-LC-MS system contains the ion-pair reagent
n-pentylamine (PTA) in the analysis solution, it can also produce
peaks corresponding to an adduct of PTA with the intact molecular
ion (Doneanu et al. 2009); this is seen as a fifth, minor peak at m/z
572.0289 corresponding to [M4+PTA−2H]2−.
The mass spectra of the dp2 (Figure 4C and E) show three peaks at
m/z 499.0013, 419.0448 and 339.0868 corresponding to the [M2
−H]−, [M2−SO3−H]− and [M2−2SO3−H]− ions, which, being eluted
at the same RT of 7.490 min, indicate the saturated dp2, HexA
(2S)-aMan(6S)R. This confirms that the other GlcNH3 þ is indeed
located in the reducing terminal disaccharide of dp6-2a. In summary,
SE-HPLC and IPRP-LC-MS-ITTOF analyses prove that dp6-2a is
ΔHexA(2S)-GlcNS(6S)-HexA(2S)-GlcNH3 þ (6S)-HexA
(2S)-GlcNH3 þ (6S) (see Figure 4F), with two GlcNH3 þ residues
located in the central and reducing terminal positions.
Sequencing of dp6-2b with two GlcNH3 þ residues
The SE-HPLC profile of the scission products of dp6-2b (Figure 5A)
shows two major peaks, labeled dp2 and dp4, with the same retention
times as standard dp2 and dp4. In this case, the UV 232nm absorbance of the dp2 is higher than that of the dp4, suggesting that the
dp2 is unsaturated while the dp4 is not. This would indicate that the
two GlcNH3 þ residues are located in the two terminal disaccharides
(see Figure 5F).
IPRP-LC-MS-ITTOF analysis of the dp4 gave four major ion
peaks (Figure 5B). The peak at m/z 537.4779, with an observed RT
of 10.213 min (Figure 5D), indicated a saturated dp4 of sequence,
HexA(2S)-GlcNS(6S)-HexA(2S)-aMan(6S)R (M5 structure in Supplementary data, Figure S4). This result accorded with the indication
from SE-HPLC that the remaining single N-sulfation was centrally located. The peaks at m/z 497.5008, 457.5230 and 417.5445 and the
peaks, being eluted at the same RT of 10.213 min, are the [M5−SO3−2H]2−, [M5−2SO3−2H]2− and [M5−3SO3−2H]2− ions, respectively, due to the in-source loss of sulfate groups (Figure 5D). Again, a
minor peak at m/z 581.0310 is a PTA adduct of the M5 ion, i.e.
[M5+PTA−2H]2−.
The mass spectra of the dp2 (Figure 5C and E) showed two peaks at
m/z 480.9925 and 401.0351, corresponding to the [M1−H]− and [M1
−SO3−H]− ions, indicating the unsaturated dp2, ΔHexA(2S)-aMan
(6S)R. In summary, SE-HPLC and IPRP-LC-MS-ITTOF analyses
721
A new sequencing approach for N-unsubstituted Hep/HS
revealed the sequence of dp6-2b to be ΔHexA(2S)-GlcNH3 þ (6S)-HexA
(2S)-GlcNS(6S)-HexA(2S)-GlcNH3 þ (6S) (see Figure 5F), with the two
GlcNH3 þ residues being at the reducing and nonreducing termini.
Sequencing of dp6-2c with two GlcNH3 þ residues
Dp6-2c was sequenced by the same protocol as dp6-2a and dp6-2b.
The SE-HPLC and IPRP-LC-MS-ITTOF analysis showed that three
major HNO2-scission products were obtained (detail shown in Supplementary data, Figure S2), i.e. unsaturated M1 (ΔHexA(2S)-aMan
(6S)R), saturated M2 (HexA(2S)-aMan(6S)R) and saturated M3
(HexA(2S)-GlcNS(6S)). The sequence of dp6-2c was thus determined
to be ΔHexA(2S)-GlcNH3 þ (6S)-HexA(2S)-GlcNH3 þ (6S)-HexA(2S)GlcNS(6S), with two GlcNH3 þ residues present in the central and
nonreducing terminal positions.
Sequencing of dp6-1a with one GlcNH3 þ residue
The SE-HPLC profile of dp6-1a after pH 4 HNO2 scission (Figure 6A)
shows three major product peaks, labeled dp2, dp4 and dp6 (based on
the same retention times as standard dp2, dp4 and dp6 species). The
presence of a dp6 fraction would suggest that a GlcNH3 þ residue is
located in the reducing terminal disaccharide (see Figure 6H). However, this is in contradiction with the additional presence of both
dp4 and dp2 fragments, suggesting that there are two other possible
locations, i.e. either the central or nonreducing disaccharides. The
only possible explanation for this is that the dp6-1a fraction is not
pure, and contains more than one dp6 species. The structures of
these three fragments were subsequently analyzed by IPRP-LC-MSITTOF.
MS analysis of the dp6 product gave a series of ion peaks corresponding to both intact molecular ion (M7 structure in Supplementary
data, Figure S4) and the PTA adduct, and also various losses of sulfate
(Figure 6B, and sulfate losses confirmed by the EIC in Figure 6E). The
diagnostic ion peaks show that the structure of M7 is ΔHexA
(2S)-GlcNS(6S)-HexA(2S)-GlcNS(6S)-HexA(2S)-aMan(6S)R, indicating that this original dp6 possessed one GlcNH3 þ residue located at
the reducing end (Figure 6H).
Mass spectra of the dp4 fraction (Figure 6C) revealed a peak at m/z
584.9693, which eluted at an RT of 10.655 min (Figure 6F), indicating that the saturated dp4 species, HexA(2S)-GlcNS(6S)-HexA
(2S)-GlcNS(6S) (M6 in Supplementary data, Figure S4), is present.
A further three peaks, observed at m/z 544.9898, 505.0111 and
465.0324, were again due to the in-source loss of sulfate groups
(Figure 6F). This suggests that a dp6 with a GlcNH3 þ residue in the
nonreducing disaccharide must also exist in the dp6-1a fraction
(Figure 6I).
The mass spectra of the dp2 fraction (Figure 6D and G) reveal two
molecular ions [M1−H]− (m/z 480.9902) and [M1−SO3−H]− (m/z
401.0326), corresponding to the unsaturated ΔHexA(2S)-aMan
(6S)R. This is also consistent with the dp4 analysis, indicating one
GlcNH3 þ residue located in the nonreducing disaccharide of a dp6,
that upon HNO2 cleavage yields a saturated dp4 (M6) and an unsaturated dp2 (M1).
In summary, combined SE-HPLC and IPRP-LC-MS-ITTOF analyses reveal that the dp6-1a fraction is actually a mixture of two different, unresolved dp6s, each possessing one GlcNH3 þ residue, namely
ΔHexA(2S)-GlcNS(6S)-HexA(2S)-GlcNS(6S)-HexA
(2S)-GlcNH3 þ (6S) (Figure 6H) and ΔHexA(2S)-GlcNH3 þ (6S)-HexA
(2S)-GlcNS(6S)-HexA(2S)-GlcNS(6S) (Figure 6I). This resolves the
original puzzle that only two dp6 species containing only one
GlcNH3 þ were obtained, when there are theoretically three potential
sequence combinations. As the GlcNH3 þ residues of the above two
dp6 species are located at either end of the oligosaccharides, presumably their “mirror image” similarities in linear charge distribution
make them difficult to resolve by SAX-HPLC.
It should be pointed out here that significant in-source sulfate loss
is apparent in the mass data. This can be reduced by appropriate tuning of the instrument, though with a concurrent loss in sample ionization and thus sensitivity. Such sulfate losses can potentially complicate
interpretation of the mass spectra, though in these examples the
combined MS and EIC did allow a consistent interpretation of the
sequence.
Sequencing of dp6-1b with one GlcNH3 þ residue
The sequencing results for dp6-1b show that two major HNO2-scission products, i.e. saturated M3 (HexA(2S)-GlcNS(6S)) and unsaturated M4 (ΔHexA(2S)-GlcNS(6S)-HexA(2S)-aMan(6S)R), were
obtained, revealing the unique sequence of dp6-1b to be ΔHexA
(2S)-GlcNS(6S)-HexA(2S)-GlcNH3 þ (6S)-HexA(2S)-GlcNS(6S) (for
details, see Supplementary data, Figure S3).
Conclusions
A new method, comprising pH 4 HNO2 scission followed by
SE-HPLC and IPRP-LC-ITTOF analyses, was developed and validated
for the sequence analysis of heparin/HS oligosaccharides containing
GlcNH3 þ residues.
We also demonstrated that HNO2-generated species, both saturated and unsaturated, were able to be detected with sufficient sensitivity by absorbance at 232 nm on an HPLC system, allowing a simple
UV detection method to be used for sequence analysis. The location of
GlcNH3 þ residues can be rapidly analyzed by SE-HPLC with 232 nm
detection after pH 4 HNO2 scission, and the detailed sequences can be
elucidated by IPRP-LC-ITTOF.
This new method was successfully applied to the sequencing of
chemically generated, N-unsubstituted heparin dp6s. In these cases,
the oligosaccharides are fully sulfated, with no variation in the location of O-sulfates, or the presence of N-acetyl groups. In dealing
with potentially lesser sulfated species, mass data could reveal the likely level of sulfation, and distinguish between N-sulfation and
N-acetylation. Indeed, our protocol could also be applied to the
sequence analysis of N-sulfation in HS/heparin oligosaccharides by
the alternative use of pH 1.5 HNO2 degradation. However, it cannot
directly reveal the positions of O-sulfation, which is dependent upon
the availability of additional disaccharide compositional data. This
method could be useful in the structural characterization and application of oligosaccharide libraries for protein screening and other interaction studies in vitro, and could lead to a greater understanding of the
biological roles of GlcNH3 þ residues in HS/heparin. However, the use
of a UV detection method may prove limited in the analysis of
less-purified biological samples, where the possibility of significant
interference from the presence of other UV-absorbing organic materials is more likely.
Materials and methods
Materials
Heparin was obtained from Sinopharm Chemical Reagent Co. Ltd
(Shanghai, China). An enzymatically depolymerized low-molecularweight heparin (Innohep) was obtained from Leo Laboratories Ltd
(Princes Risborough, Bucks, UK). Twelve HS disaccharide standards
722
QT Liang et al.
Fig. 6. Sequencing of dp6-1a fraction. The dp6-1a fraction was cleaved by HNO2 at pH 4, and then analyzed by SE-HPLC (A). The dp6, dp4 and dp2 products were separately
pooled, as indicated by the horizontal bars, for subsequent IPRP-LC-ITTOF analysis. Mass spectra and EIC profiles of the dp6 (B and E), dp4 (C and F) and dp2 (D and G)
fractions were obtained. These indicated the presence of two distinct dp6 species within the original dp6-1a fraction, and their sequences were determined (H and I). This
figure is available in black and white in print and in colour at Glycobiology online.
were purchased from Iduron (Manchester, UK). Heparinase I (Flavobacterium heparinum; heparin lyase EC 4.2.2.7) and heparinase II
(F. heparinum; no EC number assigned) were purchased from
Sigma-Aldrich (St. Louis, MO, USA). Bio-Gel P-10 (fine grade) was
obtained from Bio-Rad Laboratories (Hemel Hempstead, Herts,
UK). ProPac PA-1 SAX-HPLC columns were obtained from Dionex
723
A new sequencing approach for N-unsubstituted Hep/HS
(Camberley, Surrey, UK). Size-exclusion chromatography columns,
Superdex™ Peptide 10/300 GL, Sephadex™ G-10 and Sephadex™
G-25 were purchased from GE Healthcare Life Sciences (Uppsala,
Sweden). Acetonitrile gradient grade for liquid chromatography was
provided by Merck KGaA (Darmstadt, Germany). Acetic acid
(99%), sodium nitrite (99%), sodium carbonate (99%), formic acid
(98%), PTA (99%), trifluoroacetic acid, sodium hydroxide, Amberlite
IR-120 (H+ form) and reagents used for de-N-sulfation were
purchased from Sigma-Aldrich (Heysham, Lancashire, UK). Sodium
borohydride (96%) and sulfuric acid (AR) were from Sinopharm
Chemical Reagent Co. Ltd (Shanghai, China). All HPLC solutions
were prepared using MilliQ (Millipore, Watford, Herts, UK) ultra
pure water.
Preparation of fully sulfated heparin tetrasaccharide
(dp4) and hexasaccharide (dp6)
The fully sulfated dp4 and dp6, comprising two and three tri-sulfated
disaccharides, respectively, were prepared as previously described
(Wei et al. 2011). Briefly, heparin dp4 and dp6 were isolated from
low-molecular-weight heparin (Innohep) by Bio-Gel P-10 chromatography. The fully sulfated dp4 and dp6 species were then resolved from
the individual dp4 and dp6 mixtures, respectively, by stronganion-exchange HPLC (SAX-HPLC).
changed to the sodium form by passage through an Amberlite
IR-120 (H+ form) column and titration of the resulting protonated
form with NaOH. HexA(2S)-aMan(6S)R was prepared by HNO2 deamination of de-N-sulfated high-molecular-weight heparin and then
depolymerization using nitrous acid at pH 4, as described on “Nitrous
acid depolymerization” section. The recovered cleavage products were
separated by SE-HPLC and the purified dp2 was characterized by
LC-MS-ITTOF.
Nitrous acid depolymerization
Deamination of oligosaccharides with nitrous acid was done at pH 4
to effect specific cleavage at N-unsubstituted GlcN residues. Fifty microliters of freshly prepared HNO2, pH 4 (0.5 M H2SO4 and 5.5 M
NaNO2 in a volume ratio of 2 : 5), were added to the dry oligosaccharide sample, and incubated for 15 min at room temperature (Gill et al.
2012). The reaction was stopped by the addition of 1 M Na2CO3, taking it to pH 8–9, and the cleavage products were then immediately reduced with 0.5 M NaBH4 for 8 h at 55°C. Afterward, the sample was
acidified to ∼pH 4 with 4 M acetic acid, and then neutralized to pH 7
by addition of ammonium hydroxide. The sample was then desalted
on a Sephadex™ G-10 column (10 × 400 mm) eluted with distilled
water at a flow rate of 1.3 mL/min.
Strong anion-exchange HPLC
Partial de-N-sulfation of oligosaccharides
The procedure for partial de-N-sulfation of fully sulfated dp4 and dp6
was slightly modified from that previously described by Wei et al.
(2005). The pyridinium salt of dp4 was obtained by passage through
an Amberlite IR-120 (H+ form) column and titration of the resulting
protonated form with pyridine. The pyridinium salt was then treated
with 95% dimethyl sulfoxide (DMSO), 5% water at 18–20°C for
15 min followed by extensive dialysis and freeze-drying.
The pyridinium salt of the dp6, was treated with 95% DMSO,
5% water at 27–30°C for 20 min followed by extensive dialysis and
freeze-drying.
Preparation of unsaturated, deaminated disaccharide
ΔHexA(2S)-aMan(6S)R was prepared from the tri-sulfated disaccharide, ΔHexA(2S)-GlcNS(6S), by de-N-sulfation followed by HNO2 deamination. De-N-sulfation of the dp2 was as described for dp4/dp6
above, except that the pyridinium salt was treated with 95%
DMSO, 5% water at 50°C for 90 min. Instead of using dialysis, the
dp2 was separated from the reaction chemicals by chromatography
on a Sephadex™ G25 gel filtration column (16 × 450 mm) run in distilled water, and the eluent monitored by UV at 232 nm. The recovered, de-N-sulfated dp2 mix was fractionated by SAX-HPLC (using
the conditions described for disaccharide analysis; see below). The
ΔHexA(2S)-GlcNH3 þ (6S) fraction was dried on a rotary vacuum concentrator (Christ, Germany).
Deamination of ΔHexA(2S)-GlcNH3 þ (6S) with nitrous acid at
pH 4, followed by borohydride reduction, was done as described on
“Nitrous acid depolymerization” section. The cleavage products were
separated by size-exclusion HPLC on a Superdex™ Peptide PC 10/300
column, and the deaminated dp2 was recovered and characterized by
LC-MS-ITTOF.
Preparation of saturated, deaminated disaccharide
High-molecular-weight heparin was de-N-sulfated, as described by
Wei et al. (2005). After de-N-sulfation, the pyridinium salt was
SAX-HPLC separation of dp6 fractions was performed on an Agilent
Technologies-1200 HPLC with an analytical ProPac PA-1 column (4.0
× 250 mm). Samples were applied in 1 mL of water adjusted to pH 3.5
with HCl, washed with pH 3.5 water for 2 min, and then eluted using a
biphasic, linear gradient of NaCl, pH 3.5 from 0 to 0.8 M (2.1–
7.1 min) and then 0.8–1.5 M (7.1–47.1 min), at a flow rate of 1 mL/
min. Eluates were monitored by on-line absorbance at 232 nm.
Disaccharide analysis
Oligosaccharides were completely digested to disaccharides using a
mixture of heparinases I (1 Unit) and II (0.5 Unit) in 0.1 mL of
0.1 M sodium acetate buffer, pH 7.0, containing 0.1 mM calcium
acetate and 100 μg/mL of bovine serum albumin, at 37°C overnight.
The reaction was terminated by heating at 100°C for 2 min.
Disaccharides were separated by SAX-HPLC on an analytical ProPac PA-1 column (4.0 × 250 mm). Disaccharides were applied in 1 mL
of water, pH 3.5, at a flow rate of 1 mL/min. After a 2 mL wash with,
pH 3.5, H2O, a biphasic linear gradient of NaCl was applied from 0 to
0.5 M (2.1–35.1 min) and then 0.5 to –1.0 M (35.1–57.1 min). The
elution profile was monitored by on-line absorbance at 232 nm, and
compared with that of standard disaccharides.
Size-exclusion HPLC
The HNO2-treated oligosaccharides were analyzed by SE-HPLC on a
Superdex™ Peptide PC 10/300 column, eluted with 0.2 M NH4HCO3
at a flow rate of 0.5 mL/min. The elution profile was monitored by online absorbance at 232 nm and individual saccharide fractions were
collected and concentrated on a rotary vacuum concentrator.
LC–MS-ITTOF analysis
A hybrid ion trap/time-of-flight mass spectrometer coupled to a
high-performance liquid chromatography system (LC–MS-ITTOF)
(Shimadzu Corp., Kyoto, Japan) was used to analyze HNO2-treated
oligosaccharides.
724
The liquid chromatography system was equipped with a binary gradient pump (LC-20AD), autosampler (SIL-20AC), degasser (DGU-20A3),
photodiode array detector (SPD-M20A), communication base module
(CBM-20A) and a column oven (CTO-20AC). Separations were performed by ion-pair reversed-phase chromatography (RP-HPLC) on a
3.5 μm Eclipse Plus C18 column (2.1 × 100 mm) at a flow rate of
0.2 mL/min at 35°C. Both eluent A (water) and eluent B (75% acetonitrile, v/v) contained 15 mM PTA ion-pairing reagent, and were adjusted
to pH 8.8 with formic acid. The linear elution gradient was 5–100% B
over 20 min. The photodiode array detection was performed from 190
to 800 nm.
The mass spectrometer was equipped with an electrospray ionization source and was operated in the negative mode. Mass spectroscopic analyses were carried out on a full-scan mass spectrometer with a
mass range 200–1800 m/z. Liquid nitrogen was used as the nebulizing
gas at a flow rate of 1.5 L/min. The curved desolvation line and heat
block temperatures were both 200°C. The interface voltages were set
at −3.5 kV, and the detector voltage was 1.7 kV. The IT and TOF area
vacuums were maintained at 1.8e−002 and 1.6e−004 Pa, respectively.
External mass calibration was carried out prior to data acquisition
using direct infusion of a reference standard from 150 to 2100 m/z.
The reference standard consisted of 0.25 mL/L trifluoroacetic acid
and ∼0.1 g/L sodium hydroxide. The flow rate of the infusion pump
was 0.2 mL/min. All calculated mass errors were <5 ppm after mass
calibration with the reference standard.
Supplementary Data
Supplementary data for this article are available online at http://glycob.
oxfordjournals.org/.
Funding
This work was supported by Grants 20773023, 21343013 and
J2013-019 from National Natural Science Foundation of China.
Acknowledgments
We greatly thank Dr Malcolm Lyon (Manchester, UK) for critical reviewing of
the manuscript. We also greatly thank Shimadzu Corp. for supporting our LCMS-ITTOF installation as a “construction laboratory”.
Conflict of interest statement
None declared
Abbreviations
aMan, 2,5-anhydromannose; CyPB, cyclophilin B; dp, degree of polymerization
(i.e. number of monosaccharide units), e.g. dp2 is disaccharide; EIC, extracted ion
chromatography; GlcA, β-D-glucuronic acid; GlcNAc, β-D-N-acetylglucosamine;
GlcNH3 þ , β-D-N-unsubstituted glucosamine; GlcNS, β-D-N-sulfoglucosamine;
ΔHexA, Δ4–5 unsaturated hexuronic acid; HS, heparan sulfate; HS 3-OST, heparan sulfate 3-O-sulfotransferases; HSV, herpes simplex virus; IdoA, α-L-iduronic
acid; IPRP-LC-ITTOF MS, ion-pair reversed-phase liquid chromatography-ion
trap/time-of-flight mass spectrometry; NDST, N-deacetylase/N-sulfotransferase;
PTA, n-pentylamine; RT, retention time; SAX-HPLC, strong-anion-exchange
HPLC; SE-HPLC, size-exclusion HPLC.
References
Beaudet JM, Weyers A, Solakyildirim K, Yang B, Takieddin M, Mousa S,
Zhang F, Linhardt RJ. 2011. Impact of autoclave sterilization on the activity
and structure of formulated heparin. J Pharm Sci. 100:3396–3404.
QT Liang et al.
Carlsson P, Presto J, Spillmann D, Lindahl U, Kjellén L. 2008. Heparin/heparan
sulfate biosynthesis: Processive formation of N-sulfated domains. J Biol
Chem. 283:20008–20014.
Casu B, Lindahl U. 2001. Structure and biological interactions of heparin and
heparan sulfate. Adv Carbohydr Chem Biochem. 57:159–206.
Cheng F, Svensson G, Fransson L-Å, Mani K. 2012. Non-conserved,
S-nitrosylated cysteines in glypican-1 react with N-unsubstituted glucosamines in heparan sulfate and catalyze deaminative cleavage. Glycobiology.
11:1480–1486.
Couchman JR. 2003. Syndecans: Proteoglycan regulators of cell-surface microdomains? Nat Rev Mol Cell Biol. 4:926–937.
Doneanu CE, Chen WB, Gebler JC. 2009. Analysis of oligosaccharides derived
from heparin by ion-pair reversed-phase chromatography/mass spectrometry. Anal Chem. 81:3485–3499.
Esko JD, Lindahl U. 2001. Molecular diversity of heparan sulfate. J Clin Invest.
108:169–173.
Esko JD, Selleck SB. 2002. Order out of chaos: Assembly of ligand binding sites
in heparan sulfate. Annu Rev Biochem. 71:435–471.
Fu L, Li L, Cai C, Li G, Zhang F, Linhardt RJ. 2014. Heparin stability by determining unsubstituted amino groups using HILIC-MS. Anal Biochem.
461:46–48.
Gill VL, Wang Q, Shi XF, Zaia J. 2012. Mass spectrometric method for determining the uronic acid epimerization in heparan sulfate disaccharides generated using nitrous acid. Anal Chem. 84:7539–7546.
Hu H, Huang Y, Mao Y, Yu X, Xu Y, Liu J, Zong C, Boons GJ, Lin C, Xia Y,
et al. 2014. A computational framework for heparan sulfate sequencing
using high-resolution tandem mass spectra. Mol Cell Proteomics.
13:2490–2502.
Huang RR, Liu J, Sharp JS. 2013. An approach for separation and complete
structural sequencing of heparin/heparan sulfate-like oligosaccharides.
Anal Chem. 85:5787–5795.
Li LY, Ly M, Linhardt RJ. 2012. Proteoglycan sequence. Mol Biosyst.
6:1613–1625.
Lindahl U, Kusche-Gullberg M, Kjellén L. 1998. Regulated diversity of heparan
sulfate. J Biol Chem. 273:24979–24982.
Liu J, Shriver Z, Blaiklock P, Yoshida K, Sasisekharan R, Rosenberg RD. 1999.
Heparan sulfate D-glucosaminyl 3-O-sulfotransferase-3A sulfates
N-unsubstituted glucosamine residues. J Biol Chem. 274:38155–38162.
Liu J, Shriver Z, Pope M, Thorp SC, Duncan MB, Copeland RJ, Sasisekharan R.
2002. Characterization of a heparan sulfate octasaccharide that binds
to herpes simplex virus type 1 glycoprotein D. J Biol Chem. 277:
33456–33467.
Liu J, Shriver Z, Venkataraman G, El Shabrawi Y, Sasisekharan R. 2002.
Tumor cell surface heparan sulfate as cryptic promoters or inhibitors of
tumor growth and metastasis. Proc Natl Acad Sci USA. 99:568–573.
Liu J, Thorp SC. 2002. Cell surface heparan sulfate and its roles in assisting viral
infections. Med Res Rev. 22:1–25.
Merry CLR, Lyon M, Deakin JA, Hopwood JJ, Gallagher JT. 1999. Highly sensitive sequencing of the sulfated domains of heparan sulfate. J Biol Chem.
274:18455–18462.
Murali S, Leong DFM, Lee JJL, Cool SM, Nurcombe V. 2011. Comparative assessment of the effects of gender-specific heparan sulfates on mesenchymal
stem cells. J Biol Chem. 286:17755–17765.
Norgard-Sumnicht K, Varki A. 1995. Endothelial heparan sulfate proteoglycans that bind to L-selectin have glucosamine residues with unsubstituted
amino groups. J Biol Chem. 270:12012–12024.
Perrimon N, Bernfield M. 2000. Specificities of heparan sulphate proteoglycans
in developmental processes. Nature. 404:725–728.
Rees MD, Pattison DI, Davies MJ. 2005. Oxidation of heparan sulphate by
hypochlorite: Role of N-chloro derivatives and dichloramine-dependent
fragmentation. Biochem J. 391:125–134.
Robinson CJ, Stringer SE. 2001. The splice variants of vascular endothelial
growth factor (VEGF) and their receptors. J Cell Sci. 114:853–865.
Salmivirta M, Lidholt K, Lindahl U. 1996. Heparan sulfate: A piece of information. FASEB J. 10:1270–1279.
Schenauer MR, Meissen JK, Seo YJ, Ames JB, Leary JA. 2009. Heparan sulfate
separation, sequencing, and isomeric differentiation: Ion mobility
A new sequencing approach for N-unsubstituted Hep/HS
spectrometry reveals specific iduronic and glucuronic scid-containing hexasaccharides. Anal Chem. 81:10179–10185.
Shi X, Zaia J. 2009. Organ-specific heparan sulfate structural phenotypes. J Biol
Chem. 284:11806–11814.
Shukla D, Liu J, Blaiklock P, Shworak NW, Bai X, Esko JD, Cohen GH,
Eisenberg RJ, Rosenberg RD, Spear PG. 1999. A novel role for 3-O-sulfated
heparan sulfate in herpes simplex virus 1 entry. Cell. 99:13–22.
Stringer SE, Kandola BS, Pye DA, Gallagher JT. 2003. Heparin sequencing.
Glycobiology. 13:97–107.
Toida T, Yoshida H, Toyoda H, Koshiishi I, Imanari T, Hileman RE,
Fromm JR, Linhardt RJ. 1997. Structural differences and the presence of
unsubstituted amino groups in heparan sulphate from different tissues
and species. Biochem J. 322:499–506.
Turnbull JE, Hopwood JJ, Gallagher JT. 1999. A strategy for rapid sequencing
of heparan sulfate and heparin saccharides. Biochemistry. 96:2698–2703.
725
Vanpouille C, Deligny A, Delehedde M, Denys A, Melchior A, Liénard X,
Lyon M, Mazurier J, Fernig DG, Allain F. 2007. The heparin/heparan sulfate sequence that interacts with cyclophilin B contains a 3-O-sulfated
N-unsubstituted glucosamine residue. J Biol Chem. 282:24416–24429.
Wei Z, Lyon M, Gallagher JT. 2005. Distinct substrate specificities of bacterial
heparinases against N-unsubstituted glucosamine residues in heparan sulfate. J Biol Chem. 280:15742–15748.
Wei Z, Deakin AJ, Blaum BS, Uhrín D, Gallagher JT, Lyon M. 2011. Preparation
of heparin/heparan sulfate oligosaccharides with internal N-unsubstituted
glucosamine residues for functional studies. Glycoconj J. 28:525–535.
Westling C, Lindahl U. 2002. Location of N-unsubstituted glucosamine residues
in heparan sulfate. J Biol Chem. 277:49247–49255.
Wolff JJ, Chi LL, Linhardt RJ, Amster J. 2007. Distinguishing glucuronic from
iduronic acid in glycosaminoglycan tetrasaccharides by using electron
detachment dissociation. Anal Chem. 79:2015–2022.