Solid-phase synthesis of α-glucosamine sulfoforms with fragmentation analysis by tandem mass spectrometry

Article abstract of DOI:10.1021/jo800713m

(Chemical Equation Presented) Sulfated epitopes of α-glucosamine (GlcN sulfoforms) were prepared by solid-phase synthesis as models of internal glucosamines within heparan sulfate. An orthogonally protected 2′-hydroxyethyl GlcN derivative was immobilized on a trityl resin support and subjected to regioselective deprotection and sulfonation conditions, which were optimized with the aid of on-resin infrared or Raman analysis. The sulfoforms were cleaved from the resin under mild Lewis acid conditions without affecting the O- or N-sulfate groups and purified by re versed-phase high-performance liquid chromatography (HPLC). The α-GlcN sulfoforms and their 4-O-benzyl ethers were examined by electrospray ionization tandem mass spectrometry (ESI-MS/MS), with product ion spectra produced by collision-induced dissociation (CID). ESI-MS/MS revealed significant differences in parent ion stabilities and fragmentation rates as a function of sulfate position. Ion fragmentation by CID resulted in characteristic mass losses with strong correlation to the positions of both free hydroxyl groups and sulfate ions. Most of these fragmentation patterns are consonant with elimination pathways, and suggest possible strategies for elucidating the structures of glucosamine-derived sulfoforms with identical m/z ratios. In particular, fragmentation analysis can easily distinguish GlcN sulfoforms bearing the relatively rare 3-O-sulfate from isomers with the more common 6-O-sulfate.

Full text of DOI:10.1021/jo800713m

Solid-Phase Synthesis of r-Glucosamine Sulfoforms with
Fragmentation Analysis by Tandem Mass Spectrometry
Runhui Liu, Chamnongsak Chanthamontri, Hongling Han, Jesu´s M. Hernández-Torres,
Karl V. Wood, Scott A. McLuckey, and Alexander Wei*
Department of Chemistry, Purdue UniVersity, 560 OVal DriVe, West Lafayette, Indiana 47907-2084
alexwei@purdue.edu
ReceiVed March 31, 2008
Sulfated epitopes of R-glucosamine (GlcN sulfoforms) were prepared by solid-phase synthesis as models
of internal glucosamines within heparan sulfate. An orthogonally protected 2′-hydroxyethyl GlcN derivative
was immobilized on a trityl resin support and subjected to regioselective deprotection and sulfonation
conditions, which were optimized with the aid of on-resin infrared or Raman analysis. The sulfoforms
were cleaved from the resin under mild Lewis acid conditions without affecting the O- or N-sulfate
groups and purified by reversed-phase high-performance liquid chromatography (HPLC). The R-GlcN
sulfoforms and their 4-O-benzyl ethers were examined by electrospray ionization tandem mass spectrometry
(ESI-MS/MS), with product ion spectra produced by collision-induced dissociation (CID). ESI-MS/MS
revealed significant differences in parent ion stabilities and fragmentation rates as a function of sulfate
position. Ion fragmentation by CID resulted in characteristic mass losses with strong correlation to the
positions of both free hydroxyl groups and sulfate ions. Most of these fragmentation patterns are consonant
with elimination pathways, and suggest possible strategies for elucidating the structures of glucosamine-
derived sulfoforms with identical m/z ratios. In particular, fragmentation analysis can easily distinguish
GlcN sulfoforms bearing the relatively rare 3-O-sulfate from isomers with the more common 6-O-sulfate.
Introduction
least 21 disaaccharide sulfoforms have been identified within
HS, assembled from six out of eight possible sulfated glu-
cosamine derivatives (GlcNAc, GlcNAc-6S, GlcNS, GlcNS-3S,
GlcNS-6S, and GlcNS-3S,6S) and four different uronic acid
The biological functions of cell-surface carbohydrates and
related glycoconjugates are frequently regulated by postsynthetic
modifications, particularly via their derivatization as sulfate
esters.1,2 In many cases, protein-carbohydrate recognition is
determined by the presentation of specific sulfate epitopes, more
recently referred to as sulfoforms.2–5 A famous example of
sulfoform diversity is provided by the glycosaminoglycan
heparan sulfate (HS), a linear polysaccharide comprised of R-D-
glucosamine (GlcN) linked in alternating 1f4 fashion to either
ꢀ-D-glucuronic acid (GlcA) or R-L-iduronic acid (IdoA).6 At
derivatives (GlcA, GlcA-2S, IdoA, and IdoA-2S);7,8 another
+
three sulfoforms containing a free amine (GlcNH3+, GlcNH3
-
3S, GlcNH3+-3S,6S) have also been observed.9 These are already
sufficient to produce an enormous number of low molecular
weight HS fragments: for example, HS octasaccharides gener-
ated randomly from naturally occurring disaccharide epitopes
can result in well over 105 possible sulfoforms. Even greater
structural diversity can be attained by the incorporation of
unnatural (or yet to be identified) sulfoforms.
(1) Hooper, L. V.; Manzella, S. M.; Baenziger, J. U. FASEB J. 1996, 10,
1137–1146.
HS and other sulfated glycosaminoglycans have attracted
great interest because of their important and potentially thera-
(2) Bowman, K. G.; Bertozzi, C. R. Chem. Biol. 1999, 6, R9–R22.
(3) Lubineau, A.; Bonnaffe´, D. Eur. J. Org. Chem. 1999, 2523–2532.
(4) Lubineau, A.; Lortat-Jacob, H.; Gavard, O.; Sarrazin, S.; Bonnaffe´, D.
Chem. Eur. J. 2004, 10, 4265–4282.
(7) Esko, J. D.; Selleck, S. B. Annu. ReV. Biochem. 2002, 71, 435–471.
(8) Mochizuki, H.; Yoshida, K.; Gotoh, M.; Sugioka, S.; Kikuchi, N.; Kwon,
Y.-D.; Tawada, A.; Maeyama, K.; Inaba, N.; Hiruma, T.; Kimata, K.; Narimatsu,
H. J. Biol. Chem. 2003, 278, 26780–26787.
(5) (a) Barroca, N.; Jacquinet, J.-C. Carbohydr. Res. 2002, 337, 673–689.
(b) Jacquinet, J.-C. Carbohydr. Res. 2004, 339, 349–359.
(6) Conrad, H. E. Heparin-Binding Proteins; Academic Press: San Diego,
1998.
(9) Westling, C.; Lindahl, U. J. Biol. Chem. 2002, 277, 49247–49255.
10.1021/jo800713m CCC: $40.75
Published on Web 07/09/2008
2008 American Chemical Society
J. Org. Chem. 2008, 73, 6059–6072 6059

Liu et al.
peutic roles in vascular diseases, inflammation, viral infections,
and cancer.6,10–12 However, only a handful of high-affinity HS
ligands have been characterized due to the challenges of
purification and structural determination,13,14 presenting a serious
bottleneck for target-oriented synthetic efforts. The structural
elucidation of HS sulfoforms with specific biological activities
may be particularly challenged by their low natural abundance
and their temporal or transient expression on cell-surface
proteoglycans. For example, 3-O-sulfate esters are generated
during a late stage of HS modification and considered to be
relatively rare, yet are found in high-affinity ligands for
extracellular proteins such as antithrombin and the herpes
simplex virus glycoprotein gD.15,16 Several recent studies have
shown that the upregulation of glucosamine 3-O-sulfotrans-
ferases(3-OSTs)isassociatedwithfetalandneuraldevelopment8,17
and also with several forms of human cancer.18 The presence
of 3-OSTs implies that the biosynthesis of HS ligands containing
GlcN-3S are critical in many important biological pathways;
nevertheless, the isolation and characterization of such ligands
remain elusive.
Mass spectrometry (MS) has become a leading method for
characterizing sulfated oligosaccharides, many of which are
available only in trace quantities.19 Three strategies have
emerged in recent years. In one approach, molecular ion signals
are generated by matrix-assisted laser desorption/ionization and
correlated with fragments produced by enzymatic digestion.14b,20,21
A second one is based on electrospray ionization tandem MS
(ESI-MS/MS) for characterizing fragments generated in situ by
collision-induced dissociation (CID)22 with the expectation that
isomers having identical m/z ratios can be distinguished by
uniqueionpeaksorbysignaturepatternsofmassfragments.19,23–26
A third strategy employs MS/MS and electron detachment
dissociation to produce signature mass fragment peaks, which
are likely generated via radical processes and can also be used
to differentiate isomeric sulfoforms.27 These approaches are
complementary and may even be performed in sequence, as the
latter two are applicable toward the compositional analysis of
mono- and disaccharide fragments generated from longer
oligosaccharide sequences. Regardless of the method, the
ionization conditions for producing mass spectra must be
optimized for a given molecular structure. Organic salts such
as sulfated carbohydrates are particularly sensitive to sampling
conditions, as their stabilities and volatilities may be affected
by counterions as well as by the parent ion structure.28
ESI-MS/MS with CID can be applied toward the quantitative
analysis of sulfoforms within HS. Saad et al. demonstrated that
the population distribution of isomeric disaccharides could be
determined to within 5% by a ratiometric analysis of fragment ion
peaks and validated their method by the compositional profiling
of several HS oligosaccharides.24 However, the sample set was
limited to commercially available HS disaccharides and did not
include rare sulfoforms containing GlcNS-3S or GlcNS-3S,6S.
Subsequent studies on sulfated monosaccharides bearing a 3-O-
sulfate ester indicate that these compounds may produce MS
fragmentation patterns and abundance ratios distinct from those
of isomeric sulfoforms.26,29 The growing significance of 3-O-sulfate
esters in biological signaling supports the value of including such
derivatives in the compositional profiling of HS oligosaccharides.
In this paper, we present the comprehensive synthesis and
ESI-MS/MS analysis of sulfoforms based on glucosamine, a
primary source of structural diversity within HS (Figure 1). GlcN
sulfoforms are also found in mucin-like glycoproteins and have
been shown to mediate binding to the cell adhesion protein
L-selectin.30 We elected to use a solid-phase approach to
generate R-GlcNAc 1 and sulfoforms 2-8 with up to three
sulfate esters at the C2, C3, and C6 positions, in overall yields
ranging from 35-88%. Solid-phase and polymeric supports have
been applied toward the linear synthesis of protected oligosac-
charides,31 but there are surprisingly few reports involving the
multistep deprotection of oligosaccharides32 and no verifiable
demonstrations involving sulfated carbohydrate derivatives.33
(10) (a) Arenberg, D. A.; Polverini, P. J.; Kunkel, S. L.; Shanafelt, A.; Strieter,
R. M. In Methods in Enzymology; Horuk, R., Ed.; Academic Press: San Diego,
1997; Vol. 288, pp 190-220. (b) Lever, R.; Page, C. P. Nature ReV. 2002, 1,
140–148.
(11) (a) Herold, B. C.; Gerber, S. I.; Polonsky, T.; Belval, B. J.; Shaklee,
P. N.; Holme, K. Virology 1995, 206, 1108–1116. (b) Feyzi, E.; Trybala, E.;
Bergstrom, T.; Lindahl, U.; Spillmann, D. J. Biol. Chem. 1997, 272, 24850–
24857. (c) Sasisekharan, R.; Schriver, Z.; Venkataraman, G.; Narayanasami, U.
Nature ReV. 2002, 2, 521–528.
(12) Linhardt, R. J. J. Med. Chem. 2003, 46, 2551–2564.
(13) Spillmann, D.; Witt, D.; Lindahl, U. J. Biol. Chem. 1998, 273, 15487–
15493.
(14) (a) Venkataraman, G.; Shriver, Z.; Raman, R.; Sasisekharan, R. Science
1999, 286, 537–542. (b) Keiser, N.; Venkataraman, G.; Shriver, Z.; Sasisekharan,
R. Nature Med. 2001, 7, 123–128.
(24) (a) Saad, O. M.; Leary, J. A. Anal. Chem. 2003, 75, 2985–2995. (b)
Saad, O. M.; Ebel, H.; Uchimura, K.; Rosen, S. D.; Bertozzi, C. R.; Leary, J. A.
Glycobiology 2005, 15, 818–826.
(15) (a) Petitou, M.; van Boeckel, C. A. A. Pur. Appl. Chem. 1997, 69, 1839–
1846. (b) Shriver, Z.; Raman, R.; Venkataraman, G.; Drummond, K.; Turnbull,
J.; Toida, T.; Linhardt, R. J.; Biemann, K.; Sasisekharan, R. Proc. Natl. Acad.
Sci. U.S.A. 2000, 97, 10359–10364.
(25) Zaia, J.; Costello, C. E. Anal. Chem. 2003, 75, 2445–2455.
(26) Minamisawa, T.; Hirabayashi, J. Rapid Commun. Mass Spectrom. 2005,
19, 1788–1796.
(16) (a) Shukla, D.; Liu, J.; Blaiklock, P.; Shworak, N. W.; Esko, J. D.; Cohen,
G. H.; Eisenberg, R. J.; Rosenberg, R. D.; Spear, P. G. Cell 1999, 99, 13–22.
(b) Liu, J.; Shriver, Z.; Pope, R. M.; Thorp, S. C.; Duncan, M. B.; Copeland,
R. J.; Raska, C. S.; Yoshida, K.; Eisenberg, R. J.; Cohen, G.; Linhardt, R. J.;
Sasisekharan, R. J. Biol. Chem. 2002, 277, 33456–33467.
(17) Kamimura, K.; Rhodes, J. M.; Ueda, R.; McNeely, M.; Shukla, D.;
Kimata, K.; Spear, P. G.; Shworak, N. W.; Nakato, H. J. Cell Biol. 2004, 166,
1069–1079.
(27) (a) Wolff, J. J.; Laremore, T. N.; Busch, A. M.; Linhardt, R. J.; Amster,
I. J. J. Am. Soc. Mass Spectrom. 2008, 19, 294–304. (b) Wolff, J. J.; Amster,
I. J.; Chi, L. L.; Linhardt, R. J. J. Am. Soc. Mass Spectrom. 2007, 18, 234–244.
(c) Wolff, J. J.; Chi, L. L.; Linhardt, R. J.; Amster, I. J. Anal. Chem. 2007, 79,
2015–2022.
(28) Gunay, N. S.; Tadano-Aritomi, K.; Toida, T.; Ishizuka, I.; Linhardt,
R. J. Anal. Chem. 2003, 75, 3226–3231.
(29) Tissot, B.; Salpin, J.-Y.; Martinez, M.; Gaigeot, M.-P.; Daniel, R.
Carbohydr. Res. 2006, 341, 598–609.
(18) Miyamoto, K.; Asada, K.; Fukutomi, T.; Okochi, E.; Yagi, Y.; Hasegawa,
T.; Asahara, T.; Sugimura, T.; Ushijima, T. Oncogene 2003, 22, 274–280.
(19) Pope, R. M.; Raska, C. S.; Thorp, S. C.; Liu, J. Glycobiology 2001, 11,
505–513.
(30) (a) Bruehl, R. E.; Bertozzi, C. R.; Rosen, S. D. J. Biol. Chem. 2000,
275, 32642–32648. (b) Hemmerich, S.; Rosen, S. D. Glycobiology 2000, 10,
849–856.
(20) (a) Venkataraman, G.; Shriver, Z.; Raman, R.; Sasisekharan, R. Science
1999, 286, 537–542. (b) Shriver, Z.; Raman, R.; Venkataraman, G.; Drummond,
K.; Turnbull, J.; Toida, T.; Linhardt, R. J.; Biemann, K.; Sasisekharan, R. Proc.
Natl. Acad. Sci. U.S.A. 2000, 97, 10359–10364.
(31) Seeberger, P. H.; Haase, W. C. Chem. ReV. 2000, 100, 4349–4393.
(32) (a) Zhu, T.; Boons, G. J. Tetrahedron: Asymmetry 2000, 11, 199–205.
(b) Tanaka, H.; Ishida, T.; Matoba, N.; Tsukamoto, H.; Yamada, H.; Takahashi,
T. Angew. Chem., Int. Ed. 2006, 45, 6349–6352.
(21) Tissot, B.; Gasiunas, N.; Powell, A. K.; Ahmed, Y.; Zhi, Z.-L.; Haslam,
S. M.; Morris, H. R.; Turnbull, J. E.; Gallagher, J. T.; Dell, A. Glycobiology
2007, 17, 972–982.
(33) Solid-phase supports have been used in the synthesis of sulfated
peptides: (a) Kitagawa, K.; Aida, C.; Fujiwara, H.; Yagami, T.; Futaki, S.; Kogire,
M.; Ida, J.; Inoue, K. J. Org. Chem. 2001, 66, 1–10. (b) Ueki, M.; Watanabe,
S.; Ishii, Y.; Okunaka, O.; Uchino, K.; Saitoh, T.; Higashi, K.; Nakashima, H.;
Yamamoto, N.; Ogawara, H. Bioorg. Med. Chem. 2001, 9, 477–486. (c) Campos,
S. V.; Miranda, L. P.; Meldal, M. J. Chem. Soc., Perkin Trans. 1 2002, 682–86.
(d) Young, T.; Kiessling, L. L. Angew. Chem., Int. Ed. 2002, 41, 3449–3451.
(22) (a) McLuckey, S. A. J. Am. Soc. Mass Spectrom. 1992, 3, 599–614. (b)
Shukla, A. K.; Futrell, J. H. J. Mass Spectrom. 2000, 35, 1069–1090.
(23) Karlsson, N. G.; Karlsson, H.; Hansson, G. C. J. Mass. Spectrom. 1996,
31, 560–572.
6060 J. Org. Chem. Vol. 73, No. 16, 2008

Solid-Phase Synthesis of R-Glucosamine Sulfoforms
FIGURE 1. 2′-Hydroxyethyl R-N-acetylglucosaminoside 1 and GlcN sulfoforms 2-8 (a: R ) Bn; b: R ) H).
SCHEME 1. Synthesis and Loading of 2′-Hydroxyethyl
r-GlcN Derivatives onto a Trityl-PS Resina
Here, we examine some practical issues enabling the solid-phase
synthesis of GlcN sulfoforms from a common precursor and
provide a systematic analysis of molecular ion stability and
fragmentation behavior as a function of sulfonation pattern. The
synthetic GlcN sulfoforms reveal important differences between
positional isomers and offer new insights for structural analysis
by MS/MS techniques.
Results and Discussion
Synthesis. Sulfoform diversity can be generated from com-
binations of carbohydrate building blocks having protecting
groups at predesignated sulfate sites or by developing synthetic
precursors with orthogonal protecting group systems.3,34–36 Both
approaches have their merits, but the latter offers the benefit
of using a universal intermediate to generate an entire set of
sulfoforms, unnatural as well as natural. The challenges of
developing a unified sulfation strategy increase geometrically
with the number of possible sulfate positions, as each additional
protecting group reduces the chemical space for further or-
thogonality. We have recently addressed the issue of generating
diverse sulfoforms from a common intermediate by introducing
a heparan disaccharide (GlcN(R1f4)GlcA) with six orthogonal
protecting groups.35 A high density of sulfate groups also raises
some concerns for the handling of highly charged intermediates
in a multistep synthetic sequence. Here, we address this
important practical issue by using a solid-phase approach to
produce sulfated carbohydrates in deprotected forms.
R-Thioethyl 2-azidoglucoside 9 was prepared in three steps
and 57% overall yield from glucosamine hydrochloride by Cu-
mediated diazo transfer,37 peracetylation, and TMSOTf-mediated
glycosylation using TMSSEt,38 added in two portions (Scheme
1). Triacetate 9 was converted into 4,6-benzylidene acetal 10
in 85% yield over three steps, followed by reductive cleavage
to the 4-O-benzyl ether using BH3/Bu2BOTf and 2,6-lutidine
as a proton scavenger.39 The C6 alcohol was protected as a tert-
a Select abbreviations: DBU ) 1,8-diazabicyclo[5.4.0]undec-7-ene; HE-
BriB ) 2′-hydroxyethyl 2-bromoisobutyrate; TBDPS ) tert-butyldiphe-
nylsilyl.
butyldiphenylsilyl (TBDPS) ether to afford compound 11 in 65%
overall yield from 10. A glycol linker was installed by NIS/
TfOH-promoted coupling of R-thioethyl glycoside 11 with 2′-
hydroxyethyl 2-bromoisobutyrate, followed by treatment with
DBU to yield a 6:1 R/ꢀ mixture of glycol-linked GlcN
derivatives.40 Careful separation by silica gel chromatography
afforded 2′-hydroxyethyl R-GlcN derivative 12 in 66% isolated
yield from 11. It is worth mentioning that the coupling of
ꢀ-thioglycosides with monoprotected glycol linkers generally
favored the formation of ꢀ-O-glycosides,40 contrary to expecta-
tions of glycosyl donors with nonparticipatory C2 groups.41–43
Orthogonally protected R-GlcN 12 was tethered onto a trityl-
PS resin in dry pyridine at 65 °C, assisted by 4-dimethylami-
nopyridine (DMAP) to produce R-GlcN-loaded resin 13. A
loading ratio of 0.8-0.9 mmol/g was attained using stoichio-
metric amounts of 12; a higher ratio of 1.3 mmol/g could be
obtained by using 1.5 equiv of monosaccharide relative to the
maximum loading density. The loading ratios of 13 were
determined by the amount of recovered 12 after cleavage from
the resin.
(34) (a) Herna´ndez-Torres, J. M.; Liew, S.-T.; Achkar, J.; Wei, A. Synthesis
2002, 487–490. (b) Liew, S.-T.; Wei, A. Carbohydr. Res. 2002, 337, 1319–
1324.
Orthogonal cleavage conditions for the TBDPS, Ac, and N3
groups were first optimized for homogeneous (solution) condi-
tions, in accordance with our earlier studies of an orthogonally
(35) Fan, R. H.; Achkar, M.; Hernandez-Torres, J. M.; Wei, A. Org. Lett.
2005, 7, 5095–5098.
(36) Other selected examples of carbohydrates with orthogonally protecting
group systems: (a) Wong, C.-H.; Ye, X.-S.; Zhang, Z. J. Am. Chem. Soc. 1998,
120, 7137–7138. (b) Wunberg, T.; Kallus, C.; Opatz, T.; Henke, S.; Schmidt,
W.; Kunz, H. Angew. Chem., Int. Ed. 1998, 37, 2503–2505. (c) Prabhu, A.;
Venot, A.; Boons, G.-J. Org. Lett. 2003, 5, 4975–4978. (d) Haller, M. F.; Boons,
G.-J. Eur. J. Org. Chem. 2002, 2033–2038. (e) Misra, A. K.; Agnihotri, G.;
Madhusudan, S. K.; Tiwari, P. J. Carbohydr. Chem. 2004, 23, 191–199.
(37) Alper, P. B.; Hung, S. C.; Wong, C. H. Tetrahedron Lett. 1996, 37,
6029–6032.
(40) Glycosylation using an R-thioethyl derivative with a 4-O-p-methoxy-
benzyl ether afforded a 9:1 R/ꢀ ratio in 90% combined yield. Herna´ndez-Torres,
J. M. Ph.D. Thesis, Purdue University, 2005.
(41) (a) Paulsen, H.; Stenzel, W. Angew. Chem., Int. Ed. 1975, 14, 558. (b)
van Straten, N. C. R.; Kriek, N. M. A. J.; Cziria, Z. A. C.; van der Marel, G. A.;
van Boom, J. H. Tetrahedron 1997, 53, 6539–6554.
(42) Lohman, G. J. S.; Seeberger, P. H. J. Org. Chem. 2004, 69, 4081–
4093.
(38) Pozsgay, V.; Glaudemans, C. P. J.; Robbins, J. B.; Schneerson, R.
Tetrahedron 1992, 48, 10249–10264.
(39) Herna´ndez-Torres, J. M.; Achkar, J.; Wei, A. J. Org. Chem. 2004, 69,
7206–7211.
(43) Cid, M. B.; Alfonso, F.; Mart´ın-Lomas, M. Chem. Eur. J. 2006, 11,
928–938.
J. Org. Chem. Vol. 73, No. 16, 2008 6061

Liu et al.
TABLE 1. Deprotection Conditions and Characteristic Vibrational Bands of Protecting Groups
a
group
deprotection conditions
IR, Raman bands (cm-1
)
N3 (to -NH2)
N3 (to -NHAc)
6-O-TBDPSc
(i) Bu3P (5 eq), CH2Cl2, 5 h; (ii) 95% aq DMF, 10 h, rtb
as above, then 1:1 pyridine/Ac2O, 1 day, rt
I: 0.9 M TBAF in THF (pH g10), 1 day, rt
II: 0.9 M TBAF in THF (pH 7) 1-2 days, rt
NaOMe or Bu4NOH (5 eq), 1:1 MeOH/CH2Cl2, 5 h, rt
IR: 2107 (-)
IR: 1687 (+)
Raman: 1407, 1462, 1568 (-)
3-O-Ac
IR: 1753 (-)
a Vibrational bands used to monitor the disappearance (-) or appearance (+) of functional groups. b Reference 35. c Condition I preferred for efficient
cleavage; condition II preferred for chemoselectivity.
FIGURE 2. Raman and IR spectra of R-GlcN derivatives on trityl-PS resin. (a, b) IR and Raman spectra of fully protected R-GlcN (13).44 (c) IR
spectrum following N3 reduction to free amine. (d) Raman spectrum following 6-O-TBDPS deprotection. (e) IR spectrum following conversion of
N3 to -NHAc. (f) IR spectrum following 3-O-Ac deprotection.
protected heparan disaccharide.35 These deprotection conditions
could also be employed in solid-phase reactions, but the order
of functional group transformations and the responsivity of the
supporting PS resin had some influence on reaction efficiency.
In some cases, reaction speed was reduced by changes in local
chemical environment; in others, a loss of chemoselectivity was
observed. For example, deprotection of the TBDPS ether with
unneutralized TBAF (pH g 10) did not affect the C3 acetate in
azidoglucose 13, but performing the same reaction on the
corresponding N-acetyl derivative resulted in partial C3 deacety-
lation. We thus employed a dual set of deprotection conditions,
one optimized for speed but which could be slightly aggressive
toward other functionalities, and a slower but milder condition
which could be used without any concern for chemoselectivity
(Table 1).
The efficacy of each deprotection step could be monitored
on-resin by Fourier transform infrared (IR) or Raman spectros-
copy.31 Each protecting group exhibited one or more charac-
teristic vibrational bands, whose presence could be correlated
with the end point of the deprotection reaction (Table 1). Typical
IR and Raman spectra following a deprotection reaction starting
from immobilized R-GlcN derivative 13 are shown in Figure
2.44 It is interesting to note that the IR frequency corresponding
to the amide carbonyl stretch of NHAc appears at 1687 cm-1
while immobilized on resin, but is restored to a more typical
value of 1652 cm-1 after cleavage from the resin support.
Partially deprotected R-GlcN derivatives were directly con-
verted on-resin into sulfate esters using SO3 ·pyridine, then
(44) The origin of the IR stretch at 1679 cm-1 in Figure 2a is unknown.
6062 J. Org. Chem. Vol. 73, No. 16, 2008

Solid-Phase Synthesis of R-Glucosamine Sulfoforms
TABLE 2. Generation of 4-O-Benzyl GlcN 1a and Sulfoforms
2a-8a (Bu4N Salts) from 13a
treated with a neutral Bu4NHSO4 solution (30-50 mM) in 1:1
MeOH/DMF to produce the Bu4N salt. This procedure mini-
mized any negative effects of the sulfate esters on the swelling
properties of the resin, which is essential for maintaining access
to the reaction medium. Mono-O-sulfates were readily formed
at room temperature in DMF, whereas N-sulfate formation
required basic conditions and was best achieved using 5:1
pyridine/Et3N.45 Multiple sulfates could be installed in a single
step in some instances but more often required prolonged
reaction times at 55 °C using pyridine/Et3N or an intermediate
ion-exchange step with Bu4NHSO4 to restore resin swelling
followed by a second round of SO3 ·pyridine to achieve
saturation. We note that while direct sulfonation is suitably
efficient, one might also consider the use of protected sulfate
diesters for installing multiple sulfate groups while maintaining
charge neutrality.46 However, this approach introduces an
additional deprotection step, which imposes another condition
for orthogonality.
The 2′-hydroxyethyl R-GlcN derivatives were cleaved from
the resin as their 4-O-benzyl ethers for characterization and to
determine overall yield (Table 2). We first established the
efficiency of the loading and deprotection steps by cleaving the
unsulfated glucosamine using 3% CF3COOH in CH2Cl2, and
obtained R-GlcNAc 1a in 88% yield over six operations after
purification by reversed-phase HPLC. We note that this acidic
condition did not cause the decomposition of R-GlcNAc-6S 2a,
R-GlcNAc-3S 3a, or R-GlcNAc-3S,6S 5a, which were recovered
intact by adding acetic acid to the filtrate prior to concentration
under reduced pressure. However, N-sulfated R-GlcN derivatives
were more labile to acid and typically experienced partial or
complete desulfation during recovery.
Prior experience suggested that Lewis acid conditions might
be used to cleave the GlcN sulfoforms from the resin without
sulfate ester decomposition.35 After evaluating several possibili-
ties we were pleased to find that both O- and N-sulfates were
stable in the presence of 0.3 M BF3 ·Et2O at -20 °C. This
condition was also efficient at cleaving the glycol linker from
the trityl-PS resin and proved to be of general utility for
obtaining 4-O-benzyl GlcN sulfoforms 2a-8a as their Bu4N
salts (Table 2). The C4 benzyl ether facilitated HPLC purifica-
tion by reducing the sulfoforms’ polarity and provided a
convenient UV-active chromophore for detection during elution.
The benzyl ethers were then cleaved by catalytic hydrogenation,
followed by treatment with an ion-exchange resin in aqueous
methanol to produce the fully deprotected GlcN 1b and
sulfoforms 2b-8b as Na salts (Scheme 2). The efficient sorption
of Bu4N ions is noteworthy as it can be used to purify polar
organic compounds at the final stage of synthesis,47 if care is
taken to remove residual polyelectrolyte from the resin
beforehand.
a Reagents and conditions: (a) Bu3P, CH2Cl2, rt; (b) 95% aq DMF, rt;
(c) Ac2O, pyridine, rt; (d) TBAF (pH g10), THF, rt; (e) NaOMe,
MeOH/CH2Cl2, rt; (f) TFA, CH2Cl2, rt; (g) TBAF (pH 7), THF, rt; (h)
SO3•py, DMF, rt; (i) Bu4NHSO4 (pH 7), DMF/MeOH, rt; (j) Bu4NOH,
MeOH/CH2Cl2, rt; (k) BF3•Et2O, CH2Cl2, -20 °C; (l) SO3•py, pyridine/
Et3N, rt; (m) SO3•py, pyridine/Et3N, 55 °C.
SCHEME 2. Synthesis of 2′-Hydroxyethyl r-GlcN 1b and
Sulfoforms 2b-8ba
Fragmentation Analysis. ESI-MS/MS analyses were per-
formed in negative ion mode on 2′-hydroxyethyl R-GlcN
sulfoforms with 4-O-benzyl ethers (2a-8a) as well as on fully
deprotected sulfoforms (2b-8b), with each set respectively
serving as models of internal and terminal R-GlcN residues.
Preliminary analyses were performed on sulfoforms 2a-8a
a R1,R2 ) H or SO3-; R3 ) Ac or SO3
.
-
using an ion trap mass analyzer with the CID energy set to 40%
of the resonance excitation rf voltage (5 Vpp), in order to identify
differences in mass losses or fragment patterns as a function of
the parent ion (Table 3). Among monosulfate derivatives 2a-4a,
the loss of SO3 [M - H - 80]- was responsible for the major
fragment produced from GlcNAc-3S, but was a minor frag-
mentation pathway for GlcNS or GlcNAc-6S. Among mono-
protonated disulfate derivatives 5a-7a, the loss of SO3 was
produced by all sulfoforms but the spectrum of GlcNS-6S also
featured a significant loss of glycol [M - H - 62; B ion]-.
(45) Noti, C.; de Paz, J. L.; Polito, L.; Seeberger, P. H. Chem. Eur. J. 2006,
12, 8664–8686.
(46) (a) Liu, Y.; Lien, I. F. F.; Ruttgaizer, S.; Dove, P.; Taylor, S. D. Org.
Lett. 2004, 6, 209–212. (b) Gunnarsson, G. T.; Riaz, M.; Adams, J.; Desai, U. R.
Bioorg. Med. Chem. 2005, 13, 1783–1789. (c) Simpson, L. S.; Widlanski, T. S.
J. Am. Chem. Soc. 2006, 128, 1605–1610. (d) Ingram, L. J.; Taylor, S. D. Angew.
Chem., Int. Ed. 2006, 45, 3503–3506.
(47) Kaburagi, Y.; Kishi, Y. Org. Lett. 2007, 9, 723–726.
J. Org. Chem. Vol. 73, No. 16, 2008 6063

Liu et al.
TABLE 3. ESI-MS2 Fragments of 4-O-Benzyl GlcN Sulfoforms
2a-8a
MS2 fragment mass loss
compda
M (amu)b
(% intensity)c
2a (GlcNAc-6S)
435
M - H (12)
M - H - 62 (100)
M - H - 80 (3)
M - H (100)
3a (GlcNAc-3S)
4a (GlcNS)
435
393
M - H - 62 (15)
M - H - 80 (67)
M - H (100)
M - H - 62 (2)
M - H - 80 (<1)
M - H - 108 (5)
M - H (5)
5a (GlcNAc-3S,6S)
515
M - H - 62 (1.4)
M - H - 80 (100)
M - H - 62 (<1)
M - H - 80 (100)
M - H - 62 (14)
M - H - 80 (100)
M - H (3)
FIGURE 3. Parent ion dissociation rate data of 6a, with each data
point based on the average of 50 individual mass analysis scans. CID
conditions: 100 mVpp, 90.806 kHz; bath gas temperature ) 298 K;
total gas pressure ) 1.4 × 10-4 Torr (uncorrected for ion gauge detector
response to helium).
6a (GlcNS-3S)
7a (GlcNS-6S)
8a (GlcNS-3S,6S)d
473
473
553
M - H - 80 (36)
M - H - 160 (11)
M - 3H + Na, Bu4N (100)
observed between isomeric disulfates 6a and 7a (GlcNS-3S and
GlcNS-6S, respectively). The faster dissociation rate of 6a can
be attributed to the destabilizing effect of electrostatic repulsion
between neighboring N- and O-sulfates at C2 and C3.
a Bu4N counterions. b Mass of neutral (acidic) form. c Negative ions.
d ESI-MS data; no MS2 was obtained from the parent ion [M - H]-.
High-accuracy ion-trap CID experiments performed on a
quadrupole/time-of-flight (QqTOF) tandem mass spectrometer
generated characteristic daughter ions and fragmentation patterns
from 4-O-benzyl GlcN sulfoforms under basic electrospray
conditions (Figure 4 and Table 5).48 Most fragments can be
derived from their parent ions via elimination pathways and
assigned according to Domon-Costello nomenclature.49 The
3-O-sulfate ester again plays a defining role in fragment
production: hydrogen sulfate (HOSO3-, m/z 97) is generated
as the dominant fragment from GlcNAc-3S 3a, whereas the mass
spectra of GlcNAc-6S 2a and GlcNS 4a also feature ion peaks
generated by ring fragmentation pathways leading to A and X
ions. In these cases the C3 hydroxyl is likely to promote the
formation of 3,5A and 0,2X fragments,26 whereas the C2 sulfamate
may contribute toward the formation of the 1,3A fragment.
Sulfamate (H2NSO3-, m/z 96) is also produced as a major
fragment ion in the spectrum of GlcNS 4a but in lower relative
abundance from N,O-disulfates 6a and 7a. The appearance of
a small HOSO3- peak in the GlcNS spectrum indicates that N-to-
O-sulfate migration is possible, but minor compared with other
pathways. GlcNS-6S 7a produces fragments with m/z 374 (M
- 2H - 97) and m/z 375 (M - 2H - 96), corresponding to the
TABLE 4. Dissociation Rate Constants of 4-O-Benzyl GlcN
Sulfoforms
compda
kP,b s-1
2a (GlcNAc-6S)c
3a (GlcNAc-3S)c
4a (GlcNS)d
151 ( 12
189 ( 11
100 ( 9
741 ( 24
239 ( 11
6a (GlcNS-3S)e
7a (GlcNS-6S)e
a Bu4N counterions. b 95% confidence limits. c CID conditions: 100
mVpp, 49.016 kHz. d CID conditions: 100 mVpp, 54.273 kHz. e CID
conditions: 100 mVpp, 90.806 kHz. Bath gas temperature ) 298 K; gas
pressure ) 1.4 × 10-4 Torr.
The parent ion generated from diprotonated trisulfate 8a [M -
H]- was too weak for MS/MS analysis under these conditions,
although standard ESI-MS revealed a substantial loss of one or
two SO3 groups. These losses are well-known to be produced
from sulfonic acids and indicate the importance of performing
electrospray under basic conditions.21 Overall, the results suggest
that sulfoforms bearing a 3-O-sulfate are significantly more
prone to fragmentation, relative to their positional isomers.
Several 4-O-benzyl GlcN sulfoforms were also examined by
ESI-MS/MS using a dueling-ion trap at lower CID voltages (100
mVpp) to measure the rates of parent ion dissociation and
fragment ion formation (Table 4). First-order dissociation rate
constants (kP) were obtained from the exponential decay of
normalized parent ion density P during activated collision
-
-
monovalent parent anion with loss of HOSO3 and H2NSO3
respectively, whereas GlcNS-3S 6a does not produce the latter.
The sulfation patterns have considerable influence on the
generation of neutral mass losses, i.e. glycosidic and aglyconic
cleavage pathways leading to B and Z ions, respectively. B ion
formation (mass loss of 62) is especially pronounced in GlcNAc-
6S 2a, whereas GlcNS derivative 4a has a capacity to form Z ions
(mass loss of 108). In the case of GlcNS-3S 6a, aglycon (C4)
elimination may operate in conjunction with other fragmentation
processes, not easily described by conventional nomenclature.
Mechanistically speaking, “Z elimination” may be combined with
“1,5A fragmentation” and various eliminations to produce conjugated
( )
P t
kPt ) -ln
(1)
( )
[
]
F + P t
∑
where ∑F is the sum of the normalized fragment peak
intensities. A typical semilog plot of parent ion dissociation
versus collision time t is shown in Figure 3. Differences in
dissociation rates among monosulfated GlcN derivatives were
modest: the kP values of O-sulfated GlcNAc isomers 2a and 3a
were greater than that of GlcNS 4a by 1.5 to 2 times,
respectively. On the other hand, a 3-fold rate difference was
(48) It should be noted that differences in the mass analyzers and CID
conditions used in this study had a significant impact on their respective ESI-
MS/MS spectra. In particular, preliminary MS/MS analyses using the ion trap
mass analyzer were performed with a cutoff at 30% of the parent ion m/z ratio;
therefore, no ions below this threshold could be observed.
(49) Domon, B.; Costello, C. E. Glycoconjugate J. 1988, 5, 397–409.
6064 J. Org. Chem. Vol. 73, No. 16, 2008

Solid-Phase Synthesis of R-Glucosamine Sulfoforms
FIGURE 4. ESI-MS/MS spectra with CID of 4-O-benzyl GlcN sulfoforms. (a) GlcNAc-6S (2a); (b) GlcNAc-3S (3a); (c) GlcNS (4a); (d) GlcNS-
6S (7a); (e) GlcNS-3S (6a). Parent ion [M-H]- for 2a-4a; [M-2H]2- for 7a and 6a. See Table 5 for CID conditions.
FIGURE 5. ESI-MS/MS spectra with CID of fully deprotected GlcN sulfoforms: (a) GlcNAc-6S (2b); (b) GlcNAc-3S (3b); (c) GlcNS (4b); (d)
GlcNS-6S (7b); (e) GlcNS-3S (6b). Parent ion [M - H]- for 2b-4b; [M - 2H]2- for 7b and 6b.51 See Table 6 for CID conditions.
J. Org. Chem. Vol. 73, No. 16, 2008 6065

Liu et al.
TABLE 5. Major Fragments of 4-O-Benzyl GlcN Sulfoforms by
ESI-MS/MS with CID
SCHEME 3. Proposed Fragmentation Mechanisms in the
ESI-MS/MS of GlcNS-3S 6aa
a Domon-Costello nomenclature has been applied purely for descriptive
purposes; fragmentation pathways may be acting in concert.
sometimes at the expense of other ring fragments such as the
3,5A ion. The C4 hydroxyl also appears to promote the generation
of SO3-, most likely by serving as a hydrogen bond donor to
the adjacent O-sulfate. In the case of GlcNS 4b, the relative
intensity of the HOSO3- peak (16.8%) is much higher than that
produced from 4a (2.9%), which suggests that hydrogen bonding
can also promote N-to-O-sulfate migration to GlcN-4-O-sulfate
(Scheme 4).
The 3,4-diols in 4b and 7b can participate in [1,2]-hydride
shifts (semipinacol rearrangements), a mechanism previously
reported for vicinal diols in carbohydrate rings.52 The rear-
rangement of B ions with loss of H2O (B′) can lead to
fragmentation by retroaldol or retro-Diels-Alder reaction
mechanisms (Scheme 5a). In addition, vicinal diols and espe-
cially enediols are easily deprotonated, due to electronic
stabilization of the corresponding planar anion.53 This can
explain the observation of additional dianion fragments from
GlcNS-6S 7b, which are not observed in the spectra of the
corresponding 4-O-benzyl sulfoform 7a (Scheme 5b).
a Derived from parent ion [M - H]- or [M - 2H]2- as described in
Figure 4. b Relative abundance based on parent ion peak, except 3a and
6a. c CID at 380 mV, 80.39 KHz (10 ms). d CID at 330 mV, 80.39 KHz
(10 ms). e CID at 350 mV, 89.04 KHz (10 ms). f CID at 220 mV,
149.22 KHz (10 ms). g See Scheme 3 for mechanistic details.
fragments with m/z 174 and 176 (Scheme 3a), with elimination of
H2NSO3 and “C elimination” (loss of ethanol) to produce an
R-pyrone with m/z 221 (Scheme 3b), or with loss of HSO3- and
ring contraction with extrusion of HCN and loss of H2 to produce
furans with m/z 191 and 253 (Scheme 3c).
ESI-MS/MS of fully deprotected sulfoforms 2b-4b, 6b, and
7b using similar CID conditions also produces characteristic
fragmentation patterns, but with somewhat greater complexity
(Figure 5 and Table 6).50 The C4 hydroxyl adds considerable
diversity to the possible types of fragmentation pathways. It
significantly alters the ring cleavage pathways for monosulfates
2b and 4b with a higher production of 0,nA and 0,nX fragments,
Conclusions
R-GlcN sulfoforms can be synthesized by orthogonal depro-
tection and sulfonation of a common intermediate tethered on
a solid support, enabling their systematic analysis by tandem
mass spectrometry. The sulfoforms were efficiently cleaved from
trityl-PS resins as 4-O-benzyl ethers using Lewis acid conditions
(51) Isotopic parent ion peaks were observed in the MS/MS spectra of 4b,
7b, and 6b. These were higher than expected after CID, possibly due to
perturbations in rf resonance frequencies corresponding with the expected parent
ions.
(52) Robins, M. J.; Nowak, I.; Wnuk, S. F.; Hansske, F.; Madej, D. J. Org.
Chem. 2007, 72, 8216–8221.
(53) (a) Wit, G. D.; Kieboom, A. P. G.; Bekkum, H. V. Carbohydr. Res.
1979, 74, 157–175. (b) Calle, P.; Sanchez, A.; Sieiro, C. J. Chem. Soc., Perkin
Trans. 2 1990, 1181–1185.
(50) Control experiments with Bu4N salts suggest that the counterion does
not have a significant effect on the negative-mode ESI-MS/MS spectra.
6066 J. Org. Chem. Vol. 73, No. 16, 2008

Solid-Phase Synthesis of R-Glucosamine Sulfoforms
TABLE 6. Major Fragments of Fully Deprotected GlcN
Sulfoforms by ESI-MS/MS with CID
SCHEME 4. Elimination and Migration Pathways in the
ESI-MS/MS of GlcNS Sulfoforms 4b
SCHEME 5. Proposed Fragmentation Pathways in the
ESI-MS/MS of GlcNS 4b and GlcNS-6S 7b
with a 3-O-sulfate ester eliminate hydrogensulfate ion more
readily than other sulfoforms, whereas derivatives with a C3
hydroxyl are more prone to ring fragmentation. GlcNS deriva-
tives with a 4-O-benzyl ether tend to form Z ions, whereas those
with a C4 hydroxyl produce additional 0,nA and 0,nX fragments.
The increased number of fragmentation pathways introduced
by the C4 hydroxyl suggests that GlcN sulfoforms with O4
protecting groups are better suited for modeling the MS
fragmentation of internal GlcN units within HS derivatives.
Lastly, the unique ESI-MS/MS spectra produced by GlcN
a Derived from parent ion [M - H]- or [M - 2H]2- as described in
Figure 5. b Relative abundance based on ions with highest intensity.
c CID at 350 mV, 101.51 KHz (10 ms). d CID at 400 mV, 115.60 KHz
(10 ms). e See Scheme 5a for mechanistic details. f CID at 250 mV,
184.88 KHz (10 ms). g See Scheme 5b for mechanistic details.
at low temperature, with overall yields ranging from 88% (six
operations) to 35% (nine operations) and were readily converted
to the fully deprotected sulfoforms by catalytic hydrogenation.
ESI-MS/MS analysis with CID revealed that GlcN derivatives
J. Org. Chem. Vol. 73, No. 16, 2008 6067

Liu et al.
116 mmol). The mixture was stirred for 2 h at 0 °C, extracted with
CH2Cl2 (2 × 35 mL), washed with saturated aqueous Na2CO3
solution (75 mL), dried over Na2SO4, and used immediately.
A solution of glucosamine hydrochloride (10 g, 46.4 mmol) in
66% aqueous MeOH (231 mL) was treated with K2CO3 (9.6 g, 70
mmol), CuSO4 ·5H2O (1.2 g, 4.6 mmol), and freshly prepared TfN3
(70 mL, 1.66 M solution in CH2Cl2). An additional 30 mL of MeOH
was added to homogenize the mixture. The mixture was stirred for
12 h at rt, concentrated, and azeotroped with toluene and then dried
under reduced pressure. The crude azide was dissolved in pyridine
followed by addition of Ac2O at 0 °C. The mixture was warmed to
rt, stirred for another 12 h, concentrated, azotroped with toluene,
and dried under reduced pressure. The product was purified by
recrystallization in 25% EtOAc in hexanes to afford the desired
tetraacetate as a white solid and a mixture of anomers (14.3 g; 2.4:1
R/ꢀ). These were dissolved in CH2Cl2 (60 mL) and treated with
TMSSEt (18.6 mL, 15 mmol, added in two portions) and TMSOTf
(2.76 mL, 15.3 mmol). The mixture was heated to reflux for 20 h
under Ar, quenched with saturated NaHCO3 solution (50 mL),
extracted with CH2Cl2 (3 × 35 mL), washed with brine (50 mL),
dried over Na2SO4, and concentrated. The crude product was
purified by recrystallization in 95% EtOH to afford R-thioethyl
glycoside 9 as an amorphous white solid (9.92 g, 57% over three
derivatives with identical m/z ratios support the use of frag-
mentation analysis to distinguish HS sulfoforms with 3-O-sulfate
esters from those with 6-O-sulfate esters.
Experimental Section
General experimental methods. See the Supporting Informa-
tion.
Solid-Phase Generation of r-GlcN Sulfoforms on Trityl-
PS Resin. Trityl-PS resins were obtained from Novabiochem and
used directly without activation. Standard workup conditions for
solid-phase reactions involved (i) filtering and washing the resin
(50 mg) three times alternatively with DMF (3 × 3 mL) and CH2Cl2
(3 × 3 mL) and then with additional CH2Cl2 (2 × 3 mL) and (ii)
drying the resin under reduced pressure in the presence of KOH
pellets prior to the next operation. The dry loaded resin (50 mg)
was swollen in CH2Cl2 (3 mL) for 5 min under an inert atmosphere,
cooled to -20 °C, and treated dropwise with BF3 ·Et2O (114 µL,
0.9 mmol). The mixture was stirred for 1 h and then quenched
with pyridine (3 mL) and slowly warmed to rt. The resin was filtered
and washed with MeOH (4 × 3 mL), and the extracts were
concentrated under reduced pressure at rt. Compound 1a and
sulfoforms 2a-8a were purified by reversed-phase HPLC (Hydro-
RP C18). The 4-O-benzyl group was removed using standard
hydrogenation conditions in MeOH (see below); reaction mixtures
were filtered then passed through an ion-exchange column (Dowex
Marathon MSC, Na form) to yield fully deprotected sulfoforms
2b-8b as Na salts. These were purified by reversed-phase HPLC
to remove residual polystyrenesulfonate.
Mass Spectrometry. Preliminary ESI-MS/MS analyses were
carried out on a mass spectrometer with an ion-trap mass analyzer.
Electrospray experiments were typically performed with the needle
voltage at 4.0 kV and the capillary voltage at 10 V with the capillary
heated to 207 °C, the background source pressure at 1.2 × 10-2
mTorr as read by an ion gauge, and a sample flow rate of
approximately 8 µL/min with N2 as the drying gas. The sample
was administered as an unbuffered aqueous solution and scanned
to 1000 amu. CID experiments were conducted using helium as
the bath gas, at approximately 1 mTorr for optimal trapping
efficiency. The collision energy was set to 40% of the maximum
resonance excitation voltage (5 Vpp).
High-accuracy ion-trap CID experiments were performed on
parent ions using a quadrupole/time-of-flight (QqTOF) tandem mass
spectrometer coupled with a nanoelectronspray ionization source.54
Nanospray capillaries were prepared from borosilicate glass (1.5
mm o.d., 0.86 mm i.d.) using a micropipette puller. The sample
was administered as a 0.5 mM solution in 98% aqueous CH3CN
containing 1% NH4OH, with parent ions isolated by the ion-trap
filter in mass-resolving mode, followed by injection into a collision
chamber at 5 mTorr with nitrogen as the bath gas. The ions were
cooled prior to collisional activation (10 ms) followed by QqTOF
mass analysis; the final mass spectra are averages of 50 individual
scans.
Parent ion dissociation rates were measured on a dueling ion-
trap MS with nanoelectrospray ionization.55 Ions were accumulated
over several hundred milliseconds, followed by isolation of the
parent ion and collisional activation for periods ranging from 2 to
50 ms. CID was performed by applying resonance excitation
voltages with an amplitude of 100 mVpp and optimized for each
substrate based on the signal intensities in the product ion spectra.
Thioethyl 3,4,6-O-Acetyl-2-azido-2-deoxy-r-D-glucopyrano-
side (9). Procedure for Synthesis of TfN3 (Caution: Potentially
Explosive!). A solution of NaN3 (37.6 g, 580 mmol) in water (48
mL) and CH2Cl2 (80 mL) at 0 °C was treated with Tf2O (19.6 mL,
1
steps). H NMR (300 MHz, CDCl3): δ 5.43 (d, 1 H, J ) 6 Hz),
5.27 (t, 1 H, J ) 9.6 Hz), 4.99 (t, 1 H, J ) 9.8 Hz), 4.44 (ddd, 1
H, J ) 1.8, 4.5, 10.2 Hz), 4.29 (dd, 1 H, J ) 4.8, 12.3 Hz), 4.05
(dd, 1 H, J ) 2.1,12.3 Hz), 3.99 (dd, 1 H, J ) 6, 10.5 Hz), 2.6 (m,
2 H), 2.06 (s, 6 H), 2.02 (s, 3 H), 1.30 (t, 3 H, J ) 7.2 Hz). 13C
NMR (75 MHz, CDCl3): δ 170.3, 168.5, 82.8, 71.9, 68.5, 67.7,
61.7, 61.3, 24.5, 20.4, 14.4. IR (KBr): 2965, 2112, 1749, 1367,
1225, 1054, 1016 cm-1. [R]20D ) +180.2 (c 1.04, CH2Cl2). HRESI-
MS: m/z calcd for C14H21N3O7SNa [M + Na]+ 398.0998, found
398.1000.
Thioethyl 3-O-Acetyl-2-azido-4,6-O-benzylidene-2-deoxy-r-
D-glucopyranoside (10). A solution of triacetate 9 (7.7 g, 20.5
mmol) in MeOH (60 mL) was treated with 1 M NaOMe solution
in MeOH (8.2 mL, 8.2 mmol). The mixture was stirred at rt for
3 h, neutralized with activated Dowex 50X-W-H+ ion-exchange
resin, filtered, concentrated, and dried under reduced pressure. A
solution of the crude triol in THF (48 mL) was treated with
benzaldehyde dimethyl acetal (9.3 mL, 61.5 mmol) and TsOH·H2O
(390 mg, 2.05 mmol). The mixture was refluxed for 10 h, quenched
with saturated NaHCO3 solution (30 mL), extracted with CH2Cl2
(3 × 50 mL), dried over Na2SO4, and concentrated under reduced
pressure. The crude acetal was dissolved in pyridine (64 mL)
followed by addition of Ac2O (32 mL) at 0 °C. The mixture was
stirred for 10 h at rt and then concentrated under reduced pressure.
Purification by recrystallization in 95% EtOH yielded 4,6-O-
benzylidene acetal 10 as an amorphous white solid (6.6 g, 85%
1
over three steps). H NMR (300 MHz, CDCl3): δ 7.45 (m, 1 H),
7.35 (m, 1 H), 5.50 (s, 1 H), 5.46 (d, 2 H, J ) 5.7 Hz), 5.46 (t, 1
H, J ) 9.9 Hz), 4.36 (m, 1 H), 4.25 (dd, 1 H, J ) 4.5, 9.9 Hz),
3.97 (dd, 1 H, J ) 5.7, 9.9 Hz), 3.78 (t, 1 H, J ) 9.9 Hz), 3.62 (t,
1 H, J ) 10.2 Hz), 2.63 (m, 2 H), 2.12 (s, 3 H), 1.31 (t, 3 H, J )
7.6 Hz). 13C NMR (75 MHz, CDCl3): δ 169.6, 136.9, 129.2, 128.3,
126.2, 101.7, 84.1, 79.8, 70.6, 68.7, 63.3, 62.4, 24.9, 20.9, 14.9.
IR (KBr): 2966, 2933, 2901, 2108, 1750, 1370, 1218, 1098, 969
cm-1. [R]20 ) +126.7 (c 1.03, CH2Cl2); HRESI-MS: m/z calcd
D
for C17H21N3O5SNa [M + Na]+ 402.1100, found 402.1103.
Thioethyl 3-O-Acetyl-2-azido-4-O-benzyl-6-O-(tert-butyldiphe-
nylsilyl)-2-deoxy-r-D-glucopyranoside (11). Compound 10 (2.49
g, 6.55 mmol) was dissolved in a 1 M BH3 solution in THF (66
mL) and then treated with a 1 M Bu2BOTf solution in THF (6.88
mL) at 0 °C in the presence of 2,6-lutidine (2.29 mL, 19.66 mmol).
The mixture was stirred for 10 h at 0 °C, cooled to -78 °C, and
quenched with Et3N (1.82 mL, 13.11 mmol) followed by the slow
addition of cold MeOH (30 mL). The mixture was slowly warmed
to rt, concentrated, and then purified by silica gel chromatography
(20% EtOAc in hexanes) to yield the intermediate C6 alcohol (2.04
(54) Shevchenko, A.; Chernushevich, I.; Ens, W.; Standing, K. G.; Thomson,
B.; Wilm, M.; Mann, M. Rapid Commun. Mass Spectrom. 1997, 11, 1015–1024.
(55) Wells, J. M.; Chrisman, P. A.; McLuckey, S. A. J. Am. Soc. Mass
Spectrom. 2002, 13, 614–622.
6068 J. Org. Chem. Vol. 73, No. 16, 2008

Solid-Phase Synthesis of R-Glucosamine Sulfoforms
g, 5.34 mmol). This was dissolved in DMF (7 mL) and treated
with imidazole (1.1 g, 16.02 mmol) and TBDPS-Cl (2.78 mL, 10.68
mmol). The mixture was stirred for 12 h at rt, quenched with
saturated NaHCO3 solution (40 mL), extracted with CH2Cl2 (2 ×
50 mL), washed with brine (40 mL), dried over Na2SO4, and then
concentrated under reduced pressure. Purification by silica gel
chromatography (6.25% EtOAc in hexanes) yielded C6 TBDPS
ether 11 as an amorphous white solid (2.52 g, 62% over two steps).
1H NMR (300 MHz, CDCl3): δ 7.77 (m, 4 H), 7.45 (m, 6 H), 7.34
(m, 3 H), 7.25 (m, 2 H), 5.51 (d, 1 H, J ) 6 Hz), 5.49 (t, 1 H, J
) 9.75 Hz), 4.69 (m, 2 H), 4.26 (dd, 1 H, J ) 1.2, 9.6 Hz), 4.35
(dd, 1 H, J ) 3.6, 11.1 Hz), 3.94 (m, 2 H), 3.86 (t, 1 H, J ) 9.3
Hz), 2.61 (m, 2 H), 2.08 (s, 3 H), 1.31 (t, 3 H, J ) 7.7 Hz), 1.16
(s, 9 H). 13C NMR (75 MHz, CDCl3): δ 169.7, 137.7, 135.8, 135.6,
133.4, 132.9, 129.8, 128.5, 127.9, 127.8, 127.6, 82.4, 76.5, 74.6,
73.6, 72.0, 62.5, 62.4, 26.9, 24.3, 20.9, 19.3, 14.6. IR (KBr): 3070,
and washed with 5% diisopropylethylamine (DIPEA) in 8:1 CH2Cl2/
MeOH (2 × 4 mL) and then alternatively with DMF (2 × 4 mL)
and CH2Cl2 (2 × 4 mL), and then finally with additional CH2Cl2
(3 × 4 mL). The resin was dried under reduced pressure in the
presence of KOH pellets to afford fully resin-bound R-GlcN
derivative 13 with a loading ratio ranging from 0.83 mmol/g to
0.9 mmol/g. IR (KBr): 3026, 2929, 2107, 1753, 1600, 1493, 1448,
1224, 1040, 701 cm-1. To determine the loading ratio, 50 mg of
13 was treated with a solution of 3% trifluoroacetic acid (TFA) in
CH2Cl2 (3 mL). The mixture was stirred for 1 h at rt, and then the
resin was filtered and washed with CH2Cl2 (3 × 3 mL). AcOH (3
mL) was added to the combined washings, which was then
concentrated under reduced pressure to give compound 12.
2′-Hydroxyethyl 2-Acetamido-4-O-benzyl-2-deoxy-r-D-glu-
copyranoside (1a). Op. 1: Resin-bound R-GlcN derivative 13 (40
mg, 0.022 mmol sugar) was swollen in CH2Cl2 (3 mL) for 5 min
and then treated with Bu3P (34 µL, 0.137 mmol). The mixture was
shaken for 5 h at rt, filtered, and then subjected to standard workup
conditions to give the intermediate N-ylide. Op. 2: The resin was
treated with 95:5 DMF/H2O (4 mL), shaken for 1 day at rt, washed
with THF (2 × 3 mL), and then subjected to standard workup
conditions to give the resin-bound primary amine. Op. 3: The resin
was swollen in pyridine (3 mL) for 5 min, treated with Ac2O (1.5
mL), and shaken for 1 day at rt. The resin was filtered, washed
with 4:1 DMF/pyridine (2 × 3 mL), and then subjected to standard
workup conditions to give the N-acetylated intermediate. Op. 4:
The resin was swollen in THF (0.3 mL) for 5 min, treated with 1
M TBAF in THF (2.7 mL), and shaken for 1 day at rt. The resin
was filtered, washed with THF (2 × 3 mL), and then subjected to
standard workup conditions to give the C6 alcohol. Op. 5: The resin
was swollen in CH2Cl2 (2 mL) for 5 min, treated with 1 M NaOMe
in MeOH (66 µL), and shaken for 5 h at rt. The resin was filtered,
washed with 1:1 MeOH/CH2Cl2 (2 × 3 mL), and then subjected to
standard workup conditions to give the fully deprotected resin-bound
R-GlcN. Op. 6: The resin was treated with 3% TFA in CH2Cl2 (3
mL) for 1 h at rt, filtered, and washed with CH2Cl2 (3 × 2 mL).
AcOH (3 mL) was added to the combined washings, which was
then concentrated under reduced pressure at rt. Purification by
reversed-phase HPLC yielded glycol-linked R-GlcN 1a as an
amorphous white solid (7 mg, 88% over six operations). 1H NMR
(300 MHz, CD3OD): δ 7.32 (m, 5 H), 4.96 (d, 1 H, J ) 11.1 Hz),
4.76 (d, 1 H, J ) 3.9 Hz), 4.66 (d, 1 H, J ) 11.1 Hz), 4.00 (dd, 1
H, J ) 3.3, 10.8 Hz), 3.76 (m, 7 H), 3.47 (m, 2 H), 2.00 (s, 3 H).
13C NMR (100 MHz, CD3OD): δ 173.7, 140.0, 129.3, 129.0, 128.7,
98.9, 79.9, 76.0, 73.7, 73.0, 70.3, 62.3, 62.0, 55.5, 22.6. IR (neat):
2931, 2858, 2103, 1752, 1428, 1217, 1112 cm-1. [R]20 ) +77.6
D
(c 1.10, CH2Cl2). HRESI-MS: m/z calcd for C33H41N3O5SSiNa [M
+ Na]+ 642.2434, found 642.2435.
2′-Hydroxyethyl 3-O-Acetyl-2-azido-4-O-benzyl-6-O-(tert-bu-
tyldiphenylsilyl)-2-deoxy-r-D-glucopyranoside (12). Synthesis
of 2′-Hydroxyethyl 2-Bromoisobutyrate. A solution of 2-bro-
mopropionic acid (500 mg, 3.3 mmol) and ethylene glycol (1.8
mL, 33 mmol) in CH2Cl2 (16 mL) was treated at 0 °C with
dicyclohexylcarbodiimide (DCC; 1.0 g, 4.95 mmol) and DMAP
(37 mg, 0.3 mmol). The mixture was stirred for 5 h at 0 °C,
quenched with saturated NaHCO3 solution (20 mL), extracted with
EtOAc (3 × 30 mL), washed with water (30 mL) and brine (30
mL), dried over Na2SO4, and then concentrated under reduced
pressure. Purification by silica gel chromatography (20% EtOAc
in hexanes) afforded 2′-hydroxyethyl 2-bromoisobutyrate in 67%
yield as a light yellow oil.
A mixture of thioglycoside 11 (1.24 g, 2 mmol) and 2′-
hydroxyethyl 2-bromoisobutyrate (465 mg, 2.2 mmol) in 1:1
(CH2Cl)2/Et2O (14 mL) was stirred with activated 4A molecular
sieves for 1 h at rt under Ar. Then the mixture was cooled to 0 °C
and treated with N-iodosuccinimide (495 mg, 2.2 mmol) and TfOH
(53 µL, 0.6 mmol). The mixture was stirred continuously for another
5 h at 0 °C, quenched with Et3N (167 µL, 1.2 mmol), warmed to
rt, filtered through Celite, extracted with CH2Cl2 (3 × 40 mL),
washed with brine (40 mL), dried over Na2SO4, and then concen-
trated under reduced pressure to yield the corresponding O-glycoside
as a 6:1 mixture of anomers. These were separated by silica gel
chromatography (10% EtOAc in hexane) to afford the major
R-glycoside as a pure compound.
3307, 2923, 1656, 1637, 1558, 1130, 1029 cm-1. [R]20 ) +73.4
A solution of R-glycoside (1.20 g, 1.56 mmol) in 1:1 MeOH/
THF(10mL)wastreatedat-5°Cwith1,8-diazabicyclo[5.4.0]undec-
7-ene (DBU) (233 µL, 1.56 mmol). The mixture was stirred for
15 h at -5 °C, quenched with saturated NH4Cl solution (20 mL),
extracted with CH2Cl2 (3 × 40 mL), washed with brine (40 mL),
dried over Na2SO4, and then concentrated under reduced pressure.
Purification by silica gel chromatography (20% EtOAc in hexanes)
yielded R-glycol GlcN derivative 12 as a colorless oil (0.82 g, 66%
D
(c 0.63, MeOH). HRESI-MS: m/z calcd for C17H25NO7Na [M +
Na]+ 378.1529, found 378.1532.
2′-Hydroxyethyl 2-Acetamido-4-O-benzyl-2-deoxy-6-O-sul-
fonato-r-D-glucopyranoside Tetrabutylammonium Salt (2a). Op.
1: Resin-bound R-GlcN derivative 13 (40 mg, 0.022 mmol sugar)
was swollen in CH2Cl2 (3 mL) for 5 min and then treated with
Bu3P (34 µL, 0.137 mmol). The mixture was shaken for 5 h at rt,
filtered, and then subjected to standard workup conditions to give
the intermediate N-ylide. Op. 2: The resin was treated with 95:5
DMF/H2O (4 mL), shaken for 1 day at rt, washed with THF (2 ×
3 mL), and then subjected to standard workup conditions to give
the resin-bound primary amine. Op. 3: The resin was swollen in
pyridine (3 mL) for 5 min, treated with Ac2O (1.5 mL), and shaken
for 1 day at rt. The resin was filtered, washed with 4:1 DMF/
pyridine (2 × 3 mL), and then subjected to standard workup
conditions to give the N-acetylated intermediate. Op. 4: The resin
was swollen in THF (0.3 mL) for 5 min, treated with 1 M TBAF
in THF adjusted to pH 7 with AcOH (2.7 mL), and shaken for 1
day at rt. The resin was filtered, washed with 5:1 THF/pyridine (2
× 3 mL), and then subjected to standard workup conditions to give
the C6 alcohol. Op. 5: The resin was swollen in dry DMF (3 mL)
for 5 min, treated with SO3 ·Py (22 mg, 0.137 mmol), and shaken
for 10 h at rt. Op. 6: The resin was filtered, washed with DMF (2
1
over two steps). H NMR (300 MHz, CDCl3): δ 7.77 (m, 4 H),
7.46 (m, 6 H), 7.32 (m, 3 H), 7.22 (m, 2 H), 5.67 (dd, 1 H, J )
8.7, 10.8 Hz), 5.09 (d, 1 H, J ) 3.6 Hz), 4.69 (s, 2 H), 3.89 (m, 7
H), 3.68 (m, 1 H), 3.26 (dd, 1 H, J ) 3.3, 11.1 Hz), 2.73 (t, 1 H,
J ) 6 Hz), 2.09 (s, 3 H), 1.17 (s, 9 H). 13C NMR (75 MHz, CDCl3):
δ 169.8, 137.5, 135.6, 135.4, 133.2, 132.7, 129.6, 128.3, 127.7,
127.6, 127.5, 76.0, 74.6, 72.1, 71.6, 70.1, 62.4, 61.5, 61.4, 26.7,
20.7, 19.1. IR (KBr): 2931, 2858, 2109, 1752, 1428, 1225, 1113.
1037 cm-1. [R]20D ) +81.5 (c 0.69, CH2Cl2). HRESI-MS: m/z calcd
for C33H41N3O7SiNa [M + Na]+ 642.2612, found 642.2607.
2′-Hydroxyethyl 2-Azido-3-O-acetyl-4-O-benzyl-6-O-(tert-bu-
tyldiphenylsilyl)-2-deoxy-r-D-glucopyranoside, Bound to Tri-
tyl PS Resin (13). Trityl-PS resin beads (1.07 g, 1.61 mmol) were
suspended in dry pyridine for 5 min followed by addition of
compound 12 (1.0 g, 1.61 mmol) and DMAP (79 mg, 0.645 mmol).
The mixture was shaken for 3 days at 65 °C. The resin was filtered
J. Org. Chem. Vol. 73, No. 16, 2008 6069

Liu et al.
× 3 mL), and then treated twice with Bu4NHSO4 (47 mg, 0.137
mmol) in 1:1 DMF/MeOH (2 × 2 mL, adjusted to pH 7 with 1 M
Bu4NOH in MeOH). The mixture was shaken for 3 h each time at
rt. The resin was filtered, washed with 8:1 MeOH/pyridine (2 × 3
mL), and then subjected to standard workup conditions to give the
resin-bound 6-O-sulfate. Op. 7: The resin was swollen in CH2Cl2
(2 mL) for 5 min, treated with 1 M Bu4NOH in MeOH (137 µL,
0.137 mmol), and shaken for 5 h at rt. The resin was filtered, washed
with 1:1 MeOH/CH2Cl2 (2 × 3 mL), and then subjected to standard
workup conditions to give the fully deprotected resin-bound 6-O-
sulfate. Op. 8: The resin was treated with 0.3 M BF3 ·OEt2 in
CH2Cl2 under standard cleavage conditions. Purification by reversed-
phase HPLC yielded glycol-linked 6-O-sulfate 2a as an amorphous
white solid (10.5 mg, 58% over 8 operations). 1H NMR (300 MHz,
CD3OD): δ 7.43 (dd, 2 H, J ) 1.8, 8.4 Hz), 7.29 (m, 3 H), 4.92 (d,
2 H, J ) 10.5 Hz), 4.76 (d, 2 H, J ) 10.5 Hz), 4.75 (d, 1 H, J )
3.3 Hz), 4.25 (m, 2 H), 4.02 (dd, 1 H, J ) 3, 10.5 Hz), 3.86 (m, 2
H), 3.78 (dt, 1 H, J ) 3, 9.3 Hz), 3.7 (m, 2 H), 3.49 (m, 2 H), 3.24
(m, 8 H), 2.00 (s, 3 H), 1.66 (m, 8 H), 1.42 (m, 8 H), 1.03 (t, 12
H, J ) 6.9 Hz). 13C NMR (100 MHz, CD3OD): δ 173.7, 139.9,
129.4, 129.2, 128.6, 98.9, 79.8, 76.1, 73.8, 70.9, 70.3, 67.8, 62.0,
59.5, 55.4, 24.8, 22.6, 20.7, 13.9. IR (neat): 3401, 2962, 2876, 1656,
MeOH). HRESI-MS: m/z calcd for C17H24NO10S [M - NBu4]-
434.1121, found 434.1118.
2′-Hydroxyethyl 2-Amino-4-O-benzyl-2-deoxy-2-N-sulfonato-
r-D-glucopyranoside Tetrabutylammonium Salt (4a). Op. 1:
Resin-bound R-GlcN derivative 13 (40 mg, 0.022 mmol sugar) was
swollen in CH2Cl2 (3 mL) for 5 min and then treated with Bu3P
(34 µL, 0.137 mmol). The mixture was shaken for 5 h at rt, filtered,
and then subjected to standard workup conditions to give the
intermediate N-ylide. Op. 2: The resin was treated with 95:5 DMF/
H2O (4 mL), shaken for 1 day at rt, washed with THF (2 × 3 mL),
and then subjected to standard workup conditions to give the resin-
bound primary amine. Op. 3: The resin was swollen in dry pyridine
(3 mL) for 5 min, treated with Et3N (0.6 mL), SO3 ·Py (87 mg,
0.548 mmol), and shaken for 10 h at rt. Op. 4: The resin was filtered,
washed with 1:2 MeOH:pyridine (2 × 3 mL), and then treated twice
with Bu4NHSO4 (47 mg, 0.137 mmol) in 1:1 DMF/MeOH (2 × 2
mL, adjusted to pH 7 with 1 M Bu4NOH in MeOH). The mixture
was shaken for 3 h each time at rt. The resin was filtered, washed
with 8:1 MeOH/pyridine (2 × 3 mL), and then subjected to standard
workup conditions to give the resin-bound 2-N-sulfate. Op. 5: The
resin was swollen in THF (0.3 mL) for 5 min, treated with 1 M
TBAF in THF (2.7 mL), and shaken for 1 day at rt. The resin was
filtered, washed with THF (2 × 3 mL), and then subjected to
standard workup conditions to give the C6 alcohol. Op. 6: The resin
was swollen in CH2Cl2 (2 mL) for 5 min, treated with 1 M Bu4NOH
in MeOH (137 µL), and shaken for 5 h at rt. The resin was filtered,
washed with 1:1 MeOH/CH2Cl2 (2 × 3 mL), and then subjected to
standard workup conditions to give the fully deprotected resin-bound
N-sulfate. Op. 7: The resin was treated with 0.3 M BF3 ·OEt2 in
CH2Cl2 under standard cleavage conditions. Purification by reversed-
phase HPLC yielded glycol-linked N-sulfate 4a as an amorphous
1630, 1251, 1223, 1016 cm-1. [R]20 ) +34.4 (c 0.43, MeOH).
D
HRESI-MS: m/z calcd for C17H24NO10S [M - NBu4]- 434.1121,
found 434.1124.
2′-Hydroxyethyl 2-Acetamido-4-O-benzyl-2-deoxy-3-O-sul-
fonato-r-D-glucopyranoside Tetrabutylammonium Salt (3a). Op.
1: Resin-bound R-GlcN derivative 13 (40 mg, 0.022 mmol sugar)
was swollen in CH2Cl2 (3 mL) for 5 min and then treated with
Bu3P (34 µL, 0.137 mmol). The mixture was shaken for 5 h at rt,
filtered, and then subjected to standard workup conditions to give
the intermediate N-ylide. Op. 2: The resin was treated with 95:5
DMF/H2O (4 mL), shaken for 1 day at rt, washed with THF (2 ×
3 mL), then subjected to standard workup conditions to give the
resin-bound primary amine. Op. 3: The resin was swollen in
pyridine (3 mL) for 5 min, treated with Ac2O (1.5 mL), and shaken
for 1 day at rt. The resin was filtered, washed with 4:1 DMF/
pyridine (2 × 3 mL), and then subjected to standard workup
conditions to give the N-acetylated intermediate. Op. 4: The resin
was swollen in CH2Cl2 (2 mL) for 5 min, treated with 1 M NaOMe
in MeOH (82 µL), and shaken for 5 h at rt. The resin was filtered,
washed with 1:1 MeOH/CH2Cl2 (2 × 3 mL), and then subjected to
standard workup conditions to give the C3 alcohol. Op. 5: The resin
was swollen in dry DMF (3 mL) for 5 min, treated with SO3 ·Py
(22 mg, 0.137 mmol), and shaken for 10 h at rt. Op. 6: The resin
was filtered, washed with DMF (2 × 3 mL), and then treated twice
with Bu4NHSO4 (47 mg, 0.137 mmol) in 1:1 DMF/MeOH (2 × 2
mL, adjusted to pH 7 with 1 M Bu4NOH in MeOH). The mixture
was shaken for 3 h each time at rt. The resin was filtered, washed
with 8:1 MeOH/pyridine (2 × 3 mL), and then subjected to standard
workup conditions to give the resin-bound 3-O-sulfate. Op. 7: The
resin was swollen in THF (0.3 mL) for 5 min, treated with 1 M
TBAF in THF (2.7 mL), and shaken for 1 day at rt. The resin was
filtered, washed with THF (2 × 3 mL), and then subjected to
standard workup conditions to give the fully deprotected resin-bound
3-O-sulfate. Op. 8: The resin was treated with 0.3 M BF3 ·OEt2 in
CH2Cl2 under standard cleavage conditions. Purification by reversed-
phase HPLC yielded glycol-linked 3-O-sulfate 3a as an amorphous
1
white solid (11 mg, 63% over seven operations). H NMR (300
MHz, CD3OD): δ 7.25-7.39 (m, 5 H), 5.12 (d, 1 H, J ) 3.6 Hz),
4.98 (d, 1 H, J ) 11.1 Hz), 4.64 (d, 1 H, J ) 11.1 Hz), 3.63-3.84
(m, 8 H), 3.55 (m, 1 H), 3.42 (t, 1 H, J ) 9 Hz), 3.24 (m, 8 H),
1.66 (m, 8 H), 1.42 (m, 8 H), 1.03 (t, 12 H, J ) 7.5 Hz). 13C NMR
(100 MHz, CD3OD): δ 140.1, 129.2, 129.1, 128.6, 99.1, 79.8, 75.6,
74.5, 72.5, 70.4, 62.5, 62.0, 60.0, 59.4, 24.8, 20.7, 13.9. IR (neat):
3383, 2962, 2876, 1630, 1455, 1382, 1151, 1029 cm-1. [R]20
)
D
+31.5 (c 0.41, MeOH). HRESI-MS: m/z calcd for C15H22NO9S [M
- NBu4]- 392.1015, found 392.1010.
2′-Hydroxyethyl 2-Acetamido-4-O-benzyl-2-deoxy-3,6-di-O-
sulfonato-r-D-glucopyranoside Tetrabutylammonium Salt (5a).
Op. 1: Resin-bound R-GlcN derivative 13 (40 mg, 0.022 mmol
sugar) was swollen in CH2Cl2 (3 mL) for 5 min and then treated
with Bu3P (34 µL, 0.137 mmol). The mixture was shaken for 5 h
at rt, filtered, and then subjected to standard workup conditions to
give the intermediate N-ylide. Op. 2: The resin was treated with
95:5 DMF/H2O (4 mL), shaken for 1 day at rt, washed with THF
(2 × 3 mL), and then subjected to standard workup conditions to
give the resin-bound primary amine. Op. 3: The resin was swollen
in pyridine (3 mL) for 5 min, treated with Ac2O (1.5 mL), and
shaken for 1 day at rt. The resin was filtered, washed with 4:1 DMF/
pyridine (2 × 3 mL), and then subjected to standard workup
conditions to give the N-acetylated intermediate. Op. 4: This resin
was swollen in THF (0.3 mL) for 5 min, treated with 1 M TBAF
in THF (2.7 mL), and shaken for 1 day at rt. The resin was filtered,
washed with THF (2 × 3 mL), and then subjected to standard
workup conditions to give the C6 alcohol. Op. 5: The resin was
swollen in CH2Cl2 (2 mL) for 5 min, treated with 1 M NaOMe in
MeOH (82 µL), and shaken for 5 h at rt. The resin was filtered,
washed with 1:1 MeOH/CH2Cl2 (2 × 3 mL), and then subjected to
standard workup conditions to give the corresponding 3,6-diol. Op.
6: The resin was swollen in dry pyridine (3 mL) for 5 min, treated
with Et3N (0.6 mL), SO3 ·Py (218 mg, 1.37 mmol), and shaken for
2 day at 55 °C. Op. 7: The resin was filtered, washed with 1:2
MeOH/pyridine (2 × 3 mL), and then treated twice with Bu4NHSO4
(94 mg, 0.274 mmol) in 1:1 DMF/MeOH (2 × 2 mL, adjusted to
pH 7 with 1 M Bu4NOH in MeOH). The mixture was shaken for
1
white solid (9.5 mg, 53% over eight operations). H NMR (300
MHz, CD3OD): δ 7.44 (dd, 2 H, J ) 1.5, 7.8 Hz), 7.29 (m, 3 H),
5.19 (d, 1 H, J ) 10.2 Hz), 4.95 (d, 1 H, J ) 3.3 Hz), 4.76 (dd, 1
H, J ) 8.7, 11.1 Hz), 4.54 (d, 1 H, J ) 10.5 Hz), 4.98 (dd, 1 H,
J ) 3.3, 11.1 Hz), 3.71 (m, 6 H), 3.56 (t, 1 H, J ) 9 Hz), 3.47 (m,
1 H), 3.23 (m, 8 H), 1.99 (m, 8 H), 1.66 (m, 8 H), 1.42 (m, 8 H),
1.03 (t, 12 H, J ) 7.2 Hz). 13C NMR (100 MHz, CD3OD): δ 173.7,
139.9, 129.9, 129.1, 128.6, 98.6, 79.5, 78.0, 76.0, 73.3, 70.6, 62.2,
62.0, 59.4, 55.1, 24.8, 22.9, 20.7, 13.9. IR (neat): 3413, 2962, 2876,
1662, 1630, 1263, 1217, 1128, 1038 cm-1. [R]20D ) +57.1 (c 0.62,
6070 J. Org. Chem. Vol. 73, No. 16, 2008

Solid-Phase Synthesis of R-Glucosamine Sulfoforms
3 h each time at rt. The resin was filtered, washed with 8:1 MeOH/
pyridine (2 × 3 mL), and then subjected to standard workup
conditions. Op. 8,9: The sulfation and cation-exchange steps were
repeated one more time to give the resin-bound 3,6-di-O-sulfate.
Op. 10: The resin was cleaved with 0.3 M BF3 ·OEt2 in CH2Cl2
under standard conditions. Purification by reversed-phase HPLC
yielded 3,6-di-O-sulfate 5a as an amorphous white solid (11.5 mg,
(34 µL, 0.137 mmol). The mixture was shaken for 5 h at rt, filtered,
and then subjected to standard workup conditions to give the
intermediate N-ylide. Op. 3: The resin was treated with 95:5 DMF/
H2O (4 mL), shaken for 1 day at rt, washed with THF (2 × 3 mL),
and then subjected to standard workup conditions to give the resin-
bound 2,6-amino alcohol. Op. 4: The resin was swollen in dry DMF
(3 mL) for 5 min, treated with SO3 ·Py (21.8 mg, 0.137 mmol),
and shaken for 10 h at rt. Op. 5: The resin was filtered, washed
with DMF (2 × 3 mL), and then treated twice with Bu4NHSO4
(49 mg, 0.137 mmol) in 1:1 DMF/MeOH (2 × 2 mL, adjusted to
pH 7 with 1 M Bu4NOH in MeOH). The mixture was shaken for
3 h each time at rt. The resin was filtered, washed with 8:1 MeOH/
pyridine (2 × 3 mL), and then subjected to standard workup
conditions to give the resin-bound 6-O-sulfate. Op. 6: The resin
was swollen in dry pyridine (3 mL) for 5 min, treated with Et3N
(0.6 mL) and SO3 ·Py (88 mg, 0.548 mmol), and then shaken for
10 h at rt. Op. 7: The resin was filtered, washed with 1:2 MeOH/
pyridine (2 × 3 mL), and then treated twice with Bu4NHSO4 (94
mg, 0.274 mmol) in 1:1 DMF/MeOH (2 × 2 mL, adjusted to pH
7 with 1 M Bu4NOH in MeOH). The mixture was shaken for 3 h
each time at rt. The resin was filtered, washed with 8:1 MeOH/
pyridine (2 × 3 mL), and then subjected to standard workup
conditions to give the resin-bound 2,6-di-N,O-sulfate. Op. 8: The
resin was swollen in CH2Cl2 (2 mL) for 5 min, treated with 1 M
Bu4NOH in MeOH (137 µL), and shaken for 5 h at rt. The resin
was filtered, washed with 1:1 MeOH/CH2Cl2 (2 × 3 mL) and then
subjected to standard workup conditions to give the deprotected,
resin-bound 2,6-di-N,O-sulfate. Op. 9: The resin was treated with
0.3 M BF3 ·OEt2 in CH2Cl2 under standard cleavage conditions.
Purification by reversed-phase HPLC yielded glycol-linked 2,6-di-
N,O-sulfate 7a as an amorphous white solid (9.2 mg, 35% over
1
42% over 10 operations). H NMR (300 MHz, CD3OD): δ 7.54
(dd, 2 H, J ) 1.8, 8.1 Hz), 7.26 (m, 3 H), 5.12 (d, 1 H, J ) 9.3
Hz), 4.94 (d, 1 H, J ) 3.6 Hz), 4.76 (dd, 1 H, J ) 9.3, 11.1 Hz),
4.64 (d, 1 H, J ) 9.6 Hz), 4.26 (dd, 1 H, J ) 4.2, 10.8 Hz), 4.22
(dd, 1 H, J ) 2.4, 10.5 Hz), 3.98 (dd, 1 H, J ) 3.6, 11.1 Hz), 3.90
(ddd, 1 H, J ) 2.4, 4.4, 9.9 Hz), 3.59-3.80 (m, 4 H), 3.46 (m, 1
H), 3.24 (m, 16 H), 1.98 (s, 3 H), 1.66 (m, 16 H), 1.42 (m, 16 H),
1.03 (t, 24 H, J ) 7.4 Hz). 13C NMR (100 MHz, CD3OD): δ 173.6,
139.7, 130.5, 129.0, 128.6, 98.6, 79.4, 77.7, 76.0, 71.2, 70.6, 67.5,
62.0, 59.4, 50.0, 24.8, 22.9, 20.7, 13.9. IR (neat): 3410, 2962, 2876,
1671, 1629, 1459, 1258, 1223, 1024 cm-1; [R]20D ) +20.8 (c 0.52,
MeOH). HRESI-MS: m/z calcd for C17H23NO13S2Na [M - 2NBu4
+ Na]- 536.0509, found 536.0505.
2′-Hydroxyethyl 2-Amino-4-O-benzyl-2-deoxy-2,3-di-N,O-sul-
fonato-r-D-glucopyranoside Tetrabutylammonium Salt (6a). Op.
1: Resin-bound R-GlcN derivative 13 (50 mg, 0.0274 mmol sugar)
was swollen in CH2Cl2 (2 mL) for 5 min, treated with 1 M NaOMe
in MeOH (82 µL), and shaken for 5 h at rt. The resin was filtered,
washed with 1:1 MeOH/CH2Cl2 (2 × 3 mL), and then subjected to
standard workup conditions to give the C3 alcohol. Op. 2: The resin
was swollen in CH2Cl2 (3 mL) for 5 min and then treated with
Bu3P (34 µL, 0.137 mmol). The mixture was shaken for 5 h at rt,
filtered, and then subjected to standard workup conditions to give
the intermediate N-ylide. Op. 3: The resin was treated with 95:5
DMF/H2O (4 mL), shaken for 1 day at rt, washed with THF (2 ×
3 mL), and then subjected to standard workup conditions to give
the resin-bound 2,3-amino alcohol. Op. 4: The resin was swollen
in dry pyridine (3 mL) for 5 min, treated with Et3N (0.6 mL) and
SO3 ·Py (218 mg, 1.37 mmol), and shaken for 2 days at 55 °C. Op.
5: The resin was filtered, washed with 1:2 MeOH/pyridine (2 × 3
mL), and then treated twice with Bu4NHSO4 (94 mg, 0.274 mmol)
in 1:1 DMF/MeOH (2 × 2 mL, adjusted to pH 7 with 1 M Bu4NOH
in MeOH). The mixture was shaken for 3 h each time at rt. The
resin was filtered, washed with 8:1 MeOH/pyridine (2 × 3 mL),
and then subjected to standard workup conditions to give the resin-
bound 2,3-di-N,O-sulfate. Op. 6: The resin was swollen in THF
(0.3 mL) for 5 min, treated with 1 M TBAF in THF (2.7 mL), and
shaken for 2 days at rt. The resin was filtered, washed with THF
(2 × 3 mL), and then subjected to standard workup conditions to
give the deprotected, resin-bound 2,3-di-N,O-sulfate. Op. 7: The
resin was treated with 0.3 M BF3 ·OEt2 in CH2Cl2 under standard
cleavage conditions. Purification by reversed-phase HPLC yielded
glycol-linked 2,3-di-N,O-sulfate 6a as an amorphous white solid
(12 mg, 46% over seven operations). 1H NMR (300 MHz, CD3OD):
δ 7.44 (dd, 2 H, J ) 1.8, 8.1 Hz), 7.28 (m, 3 H), 5.30 (d, 1 H, J
) 3.3 Hz), 5.19 (d, 1 H, J ) 10.8 Hz), 4.71 (dd, 1 H, J ) 9, 11.1
Hz), 4.40 (d, 1 H, J ) 10.2 Hz), 3.52-3.76 (m, 8 H), 3.47 (dd, 1
H, J ) 3, 11.1 Hz), 3.24 (m, 16 H), 1.66 (m, 16 H), 1.42 (m, 16
H), 1.03 (t, 24 H, J ) 7.5 Hz). 13C NMR (100 MHz, CD3OD): δ
139.9, 129.9, 129.1, 128.6, 98.9, 79.7, 78.3, 76.0, 72.9, 70.4, 62.3,
62.0, 59.4, 59.1, 24.8, 20.7, 13.9. IR (neat): 3402, 2961, 2875, 1630,
1
nine operations). H NMR (300 MHz, CD3OD): δ 7.50 (d, 2 H, J
) 8.1 Hz), 7.39 (m, 3 H), 5.10 (d, 1 H, J ) 4.2 Hz), 4.94 (d, 1 H,
J ) 10.5 Hz), 4.74 (d, 1 H, J ) 10.5 Hz), 4.27 (dd, 1 H, J ) 2.7,
10.5 Hz), 4.20 (dd, 1 H, J ) 4.8, 10.5 Hz), 3.69-3.88 (m, 5 H),
3.43-3.57 (m, 3 H), 3.24 (m, 16 H), 1.67 (m, 16 H), 1.42 (m, 16
H), 1.03 (t, 24 H, J ) 7.5 Hz). 13C NMR (100 MHz, CD3OD): δ
140.0, 129.5, 129.2, 128.5, 99.1, 79.7, 75.8, 74.5, 70.5, 67.9, 62.0,
59.9, 59.5, 24.8, 20.7, 13.9. IR (neat): 3401, 2962, 2876, 1631,
1459, 1251, 1226, 1055, 1027 cm-1. [R]20 ) +26.9 (c 0.52,
D
MeOH). HRESI-MS: m/z calcd for C15H21NO12S2Na [M - 2NBu4
+ Na]- 494.0403, found 494.0406.
2′-Hydroxyethyl 2-Amino-4-O-benzyl-2-deoxy-2,3,6-tri-N,O-
sulfonato-r-D-glucopyranoside Tetrabutylammonium Salt (8a).
Op. 1: Resin-bound R-GlcN derivative 13 (50 mg, 0.0274 mmol
sugar) was swollen in CH2Cl2 (2 mL) for 5 min, treated with 1 M
NaOMe in MeOH (82 µL), and shaken for 5 h at rt. The resin was
filtered, washed with 1:1 MeOH/CH2Cl2 (2 × 3 mL), and then
subjected to standard workup conditions to give the C3 alcohol.
Op. 2: The resin was swollen in CH2Cl2 (3 mL) for 5 min and then
treated with Bu3P (34 µL, 0.137 mmol). The mixture was shaken
for 5 h at rt, filtered, and then subjected to standard workup
conditions to give the intermediate N-ylide. Op. 3: The resin was
treated with 95:5 DMF/H2O (4 mL), shaken for 1 day at rt, washed
with THF (2 × 3 mL), and then subjected to standard workup
conditions to give the resin-bound 2,3-amino alcohol. Op. 4: The
resin was swollen in THF (0.3 mL) for 5 min, treated with 1 M
TBAF in THF (2.7 mL), and shaken for 10 h at rt. The resin was
filtered, washed with THF (2 × 3 mL), and then subjected to
standard workup conditions to give the resin-bound, fully depro-
tected R-GlcN. Op. 5: The resin was swollen in dry pyridine (3
mL) for 5 min, treated with Et3N (0.6 mL) and SO3 ·Py (218 mg,
1.37 mmol), and shaken for 2 days at 55 °C. Op. 6: The resin was
filtered, washed with 1:2 MeOH/pyridine (2 × 3 mL), and then
treated twice with Bu4NHSO4 (140 mg, 0.411 mmol) in 1:1 DMF/
MeOH (2 × 2 mL, adjusted to pH 7 with 1 M Bu4NOH in MeOH).
The mixture was shaken for 3 h each time at rt. The resin was
filtered, washed with 8:1 MeOH/pyridine (2 × 3 mL), and then
subjected to standard workup conditions. Op. 7,8: The sulfation
1460, 1382, 1225, 1035 cm-1. [R]20 ) +30.2 (c 0.58, MeOH);
D
HRESI-MS: m/z calcd for C15H22NO12S2 [M - 2NBu4 + H]-
472.0583, found 472.0585.
2′-Hydroxyethyl 2-Amino-4-O-benzyl-2-deoxy-2,6-di-N,O-sul-
fonato-r-D-glucopyranoside Tetrabutylammonium Salt (7a). Op.
1: Resin-bound R-GlcN derivative 13 (50 mg, 0.0274 mmol sugar)
was swollen in THF (3 mL) for 5 min, treated with 1 M TBAF in
THF (164 µL), and shaken for 10 h at rt. The resin was filtered,
washed with THF (2 × 3 mL), and then subjected to standard
workup conditions to give the C6 alcohol. Op. 2: The resin was
swollen in CH2Cl2 (3 mL) for 5 min and then treated with Bu3P
J. Org. Chem. Vol. 73, No. 16, 2008 6071

Liu et al.
and cation-exchange steps were repeated one more time to yield
the resin-bound, 2,3,6-tri-N,O-sulfate. Op. 9: The resin was treated
with 0.3 M BF3 ·OEt2 in CH2Cl2 under standard conditions.
Purification by reversed-phase HPLC yielded glycol-linked 2,3,6-
tri-N,O-sulfate 8a as an amorphous white solid (13.3 mg, 38% over
72.4, 71.9, 70.7, 68.9, 61.3, 60.6, 58.2. [R]20 ) +131.8 (c 0.09,
D
MeOH). HRESI-MS: m/z calcd for C8H16NO9S [M - Na]-
302.0546, found 302.0548.
2′-Hydroxyethyl 2-Acetamido-2-deoxy-3,6-di-O-sulfonato-r-
D-glucopyranoside Sodium Salt (5b). A mixture of 5a (5.7 mg)
was dissolved in MeOH, subjected to the hydrogenation conditions
and ion-exchange chromatography described above, and then
concentrated and purified by reversed-phase HPLC to yield 5b as
1
nine operations). H NMR (300 MHz, CD3OD): δ 7.54 (dd, 2 H,
J ) 1.2, 8.1 Hz), 7.26 (m, 3 H), 5.27 (d, 1 H, J ) 3.3 Hz), 5.16 (d,
1 H, J ) 9.9 Hz), 4.71 (d, 1 H, J ) 10.8 Hz), 4.66 (t, 1 H, J ) 9.3
Hz), 4.26 (dd, 1 H, J ) 3.9, 10.5 Hz), 4.20 (dd, 1 H, J ) 2.1, 10.2
Hz), 3.84 (ddd, 1 H, J ) 2.4, 3.9, 9.6 Hz), 3.53-3.80 (m, 5 H),
3.47 (dd, 1 H, J ) 3.3, 10.8 Hz), 3.24 (m, 24 H), 1.66 (m, 24 H),
1.42 (m, 24 H), 1.03 (t, 36 H, J ) 7.5 Hz). 13C NMR (100 MHz,
CD3OD): δ 139.8, 130.5, 129.0, 128.5, 98.9, 79.6, 78.0, 76.0, 70.9,
70.4, 67.6, 61.9, 59.5, 59.1, 24.8, 20.7, 13.9. IR (neat): 3401, 2962,
2875, 1489, 1460, 1257, 1227, 1035 cm-1. [R]20D ) +22.0 (c 0.65,
MeOH). HRESI-MS: m/z calcd for C15H20NO15S3Na2 [M - 3NBu4
+ 2Na]- 595.9791, found 595.9795.
1
an amorphous white solid (2.1 mg, 82%). H NMR (500 MHz,
CD3OD): δ 4.83 (d, 1 H, J ) 3.4 Hz), 4.50 (dd, 1 H, J ) 8.8, 10.7
Hz), 4.34 (dd, 1 H, J ) 1.9, 10.8 Hz), 4.10 (dd, 1 H, J ) 6.5, 10.1
Hz), 4.03 (dd, 1 H, J ) 3.5, 10.7 Hz), 3.91 (ddd, 1 H, J ) 1.8, 6.5,
11.7 Hz), 3.81 (ddd, 1H, J ) 3.1, 4.8, 10.6 Hz), 3.36-3.78 (m,
2H), 3.54 (dd, 1H, J ) 9.0, 9.8 Hz), 3.47 (ddd, 1H, J ) 3.4, 7.2,
8.8 Hz). 13C NMR (125 MHz, CD3OD): δ 173.6, 98.7, 79.9, 71.9,
71.2, 70.5, 68.3, 62.1, 53.7, 22.8. [R]20D ) +77.1 (c 0.14, MeOH).
HRESI-MS: m/z calcd for C10H17NO13S2Na [M - Na]- 446.0039,
found 446.0042.
2′-Hydroxyethyl 2-Amino-2-deoxy-2,3-di-N,O-sulfonato-r-D-
glucopyranoside Sodium Salt (6b). A mixture of 6a (3.1 mg) was
dissolved in MeOH, subjected to the hydrogenation conditions and
ion-exchange chromatography described above, and then concen-
trated and purified by reversed-phase HPLC to yield 6b as an
amorphous white solid (0.8 mg, 62%). 1H NMR (500 MHz,
CD3OD): δ 5.27 (d, 1 H, J ) 3.4 Hz), 4.43 (dd, 1 H, J ) 8.6, 10.6
Hz), 3.78 (m, 2H), 3.65-3.72 (m, 4H), 3.58 (m, 2H), 3.39 (dd, 1
H, J ) 3.4, 10.5 Hz). 13C NMR (125 MHz, CD3OD): δ 97.3, 78.5,
72.1, 69.7, 68.9, 61.0, 60.5, 56.5. [R]20D ) +65.6 (c 0.05, MeOH).
HRESI-MS: m/z calcd for C8H15NO12S2Na [M - Na]- 403.9933,
found 403.9931.
2′-Hydroxyethyl 2-Amino-2-deoxy-2,6-di-N,O-sulfonato-r-D-
glucopyranoside Sodium Salt (7b). A mixture of 7a (6.9 mg) was
dissolved in MeOH, subjected to the hydrogenation conditions and
ion-exchange chromatography described above, and then concen-
trated and purified by reversed-phase HPLC to yield 7b as an
amorphous white solid (1.9 mg, 65%). 1H NMR (500 MHz,
CD3OD): δ 5.07 (d, 1 H, J ) 3.6 Hz), 4.27 (dd, 1 H, J ) 2.0, 10.9
Hz), 4.12 (dd, 1 H, J ) 5.9, 10.9 Hz), 3.79 (m, 2H), 3.71 (m, 2H),
3.59 (dd, 1 H, J ) 9.1, 10.2 Hz), 3.52 (ddd, 1 H, J ) 4.1, 6.2, 10.2
Hz), 3.25 (dd, 1 H, J ) 3.5, 10.2 Hz). 13C NMR (125 MHz, D2O):
δ 95.8, 69.7, 68.1, 67.6, 65.5, 59.0, 56.1. [R]20D ) +31.6 (c 0.13,
MeOH). HRESI-MS: m/z calcd for C8H15NO12S2Na [M - Na]-
403.9933, found 403.9940.
2′-Hydroxyethyl 2-Acetamido-2-deoxy-r-D-glucopyranoside
(1b). Compound 1a (7 mg) was dissolved in MeOH (3 mL), treated
with Pd(OH)2 on charcoal (5 mg), and then stirred under positive
H2 pressure at rt for 20 h. The reaction mixture was filtered to
remove catalyst, concentrated, and purified by reversed-phase HPLC
1
to yield 1b as an amorphous white solid (5.5 mg, 90%). H NMR
(500 MHz, D2O): δ 4.85 (d, 1 H, J ) 3.6 Hz), 3.92 (dd, 1 H, J )
3.6, 10.6 Hz), 3.85 (dd, 1 H, J ) 2.0, 12.0 Hz), 3.69-3.80 (m,
6H), 3.53 (m, 1H), 3.46 (t, 1H, J ) 9.5 Hz). 13C NMR (125 MHz,
D2O): δ 174.4, 97.1, 71.8, 71.1, 69.9, 68.8, 60.5, 60.4, 53.5, 21.8.
[R]20 ) +75.7 (c 0.46, MeOH). HRESI-MS: m/z calcd for
D
C10H19NO7Na [M + Na]+ 288.1059, found 288.1061.
2′-Hydroxyethyl 2-Acetamido-2-deoxy-6-O-sulfonato-r-D-glu-
copyranoside Sodium Salt (2b). Compound 2a (6.2 mg) was
dissolved in MeOH, subjected to the hydrogenation conditions
described above, and then filtered and loaded onto a column packed
with a cation-exchange resin (Dowex Marathon MSC, Na form)
and eluted with MeOH. The fractions were concentrated and
purified by reversed-phase HPLC to yield 2b as an amorphous white
1
solid (2.6 mg, 82%). H NMR (500 MHz, CD3OD): δ 4.72 (d, 1
H, J ) 3.5 Hz), 4.29 (dd, 1 H, J ) 1.8, 10.8 Hz), 4.11 (dd, 1 H,
J ) 6.1, 10.9 Hz), 3.95 (dd, 1 H, J ) 3.6, 10.6 Hz), 3.80 (m, 2H),
3.70 (m, 2H), 3.64 (dd, 1 H, J ) 9.0, 10.4 Hz), 3.44 (ddd, 1 H, J
) 3.7, 7.1, 10.8 Hz), 3.36 (t, 1H, J ) 9.7 Hz). 13C NMR (125
MHz, CD3OD): δ 172.1, 97.4, 71.6, 70.6, 70.4, 68.7, 66.9, 60.6,
53.6, 21.1. [R]20D ) +81.9 (c 0.17, MeOH). HRESI-MS: m/z calcd
for C10H18NO10S [M - Na]- 344.0651, found 344.0648.
2′-Hydroxyethyl 2-Amino-2-deoxy-2,3,6-tri-N,O-sulfonato-r-
D-glucopyranoside Sodium Salt (8b). A mixture of 8a (9.8 mg)
was dissolved in MeOH, subjected to the hydrogenation conditions
and ion-exchange chromatography described above, and then
concentrated and purified by reversed-phase HPLC to yield 8b as
an amorphous white solid (3 mg, 81%). 1H NMR (500 MHz,
CD3OD): δ 5.23 (d, 1 H, J ) 3.4 Hz), 4.42 (dd, 1 H, J ) 8.8, 10.6
Hz), 4.33 (dd, 1 H, J ) 1.9, 10.7 Hz), 4.10 (dd, 1 H, J ) 6.3, 10.7
Hz), 3.88 (m, 1H), 3.80 (m, 1H), 3.72 (dd, 2 H, J ) 3.8, 6.2 Hz),
3.54 (m, 1H), 3.41 (dd, 1 H, J ) 3.4, 10.6 Hz. 13C NMR (125
MHz, CD3OD): δ 97.2, 78.3, 70.1, 69.6, 68.9, 66.9, 60.4, 56.6.
2′-Hydroxyethyl 2-Acetamido-2-deoxy-3-O-sulfonato-r-D-glu-
copyranoside Sodium Salt (3b). Compound 3a (5.1 mg) was
dissolved in MeOH, subjected to the hydrogenation conditions and
ion-exchange chromatography described above, and then concen-
trated and purified by reversed-phase HPLC to yield 3b as an
amorphous white solid (2.1 mg, 81%). 1H NMR (500 MHz,
CD3OD): δ 4.84 (d, 1 H, J ) 3.5 Hz), 4.50 (dd, 1 H, J ) 8.7, 10.7
Hz), 4.02 (dd, 1 H, J ) 3.5, 10.7 Hz), 3.75-3.80 (m, 2H),
3.62-3.72 (m, 4H), 3.57 (t, 1H, J ) 8.9 Hz), 3.47 (ddd, 1 H, J )
3.5, 6.8, 10.7 Hz). 13C NMR (125 MHz, CD3OD): δ 172.1, 97.3,
[R]20 ) +23.5 (c 0.2, MeOH). HRESI-MS: m/z calcd for
78.4, 72.4, 69.6, 69.0, 60.9, 60.5, 52.1, 21.2. [R]20 ) +96.4 (c
D
D
C8H16NO15S3 [M - 3Na + 2H]- 461.9682, found 461.9670.
0.14, MeOH). HRESI-MS: m/z calcd for C10H18NO10S [M - Na]-
344.0651, found 344.0652.
Acknowledgment. This work was supported by the National
Institutes of Health under Grant Nos. GM06982 (A.W.) and
GM45372 (S.A.M.). We also gratefully acknowledge support
from the Purdue Cancer Center.
2′-Hydroxyethyl 2-Amino-2-deoxy-2-N-sulfonato-r-D-glu-
copyranoside Sodium Salt (4b). Compound 4a (3.8 mg) was
dissolved in MeOH, subjected to the hydrogenation conditions and
ion-exchange chromatography described above, and then concen-
trated and purified by reversed-phase HPLC to yield 4b as an
amorphous white solid (1.4 mg, 78%). 1H NMR (500 MHz,
CD3OD): δ 3.78 (dd, 1 H, J ) 2.3, 11.8 Hz), 3.75 (m, 1H),
3.68-3.72 (m, 1H), 3.65 (dd, 1 H, J ) 5.8, 11.8 Hz), 3.58 (dd, 1
H, J ) 8.8, 10.2 Hz), 3.51-3.57 (m, 2H), 3.33 (m, 2H), 3.22 (dd,
1 H, J ) 3.7, 10.3 Hz). 13C NMR (125 MHz, CD3OD): δ 97.6,
Supporting Information Available: 1H and 13C NMR
spectra of compounds 1-12. This material is available free of
charge via the Internet at http://pubs.acs.org.
JO800713M
6072 J. Org. Chem. Vol. 73, No. 16, 2008