Nickel-catalyzed stereoselective glycosylation with C(2)- N-substituted benzylidene d-glucosamine and galactosamine trichloroacetimidates for the formation of 1,2-cis-2-amino glycosides. applications to the synthesis of heparin disaccharides, GPI anchor pseudodisaccharides, and α-GalNAc

Article abstract of DOI:10.1021/ja106682m

The 1,2-cis-2-amino glycosides are key components found within a variety of biologically important oligosaccharides and glycopeptides. Although there are remarkable advances in the synthesis of 1,2-cis-2-amino glycosides, disadvantages of the current state-of-the-art methods include limited substrate scope, low yields, long reaction times, and anomeric mixtures. We have developed a novel method for the synthesis of 1,2-cis-2-amino glycosides via nickel-catalyzed α-selective glycosylation with C(2)-N-substituted benzylidene d-glucosamine and galactosamine trichloroacetimidates. These glycosyl donors are capable of coupling to a wide variety of alcohols to provide glycoconjugates in high yields with excellent levels of α-selectivity. Additionally, only a substoichiometric amount of nickel (5-10 mol %) is required for the reaction to occur at 25 °C. The current nickel method relies on the nature of the nickel-ligand complex to control the α-selectivity. The reactive sites of the nucleophiles or the nature of the protecting groups have little effect on the α-selectivity. This methodology has also been successfully applied to both disaccharide donors and acceptors to provide the corresponding oligosaccharides in high yields and α-selectivity. The efficacy of the nickel procedure has been further applied toward the preparation of heparin disaccharides, GPI anchor pseudodisaccharides, and α-GluNAc/GalNAc. Mechanistic studies suggest that the presence of the substituted benzylidene functionality at the C(2)-amino position of glycosyl donors is crucial for the high α-selectivity observed in the coupling products. Additionally, the α-orientation of the C(1)-trichloroacetimidate group on glycosyl donors is necessary for the coupling process to occur.

Full text of DOI:10.1021/ja106682m

Published on Web 09/22/2010
Nickel-Catalyzed Stereoselective Glycosylation with
C(2)-N-Substituted Benzylidene D-Glucosamine and
Galactosamine Trichloroacetimidates for the Formation of
1,2-cis-2-Amino Glycosides. Applications to the Synthesis of
Heparin Disaccharides, GPI Anchor Pseudodisaccharides,
and r-GalNAc
Enoch A. Mensah, Fei Yu, and Hien M. Nguyen*
Department of Chemistry, UniVersity of Iowa, Iowa City, Iowa 52242
Received July 27, 2010; E-mail: hien-nguyen@uiowa.edu
Abstract: The 1,2-cis-2-amino glycosides are key components found within a variety of biologically important
oligosaccharides and glycopeptides. Although there are remarkable advances in the synthesis of 1,2-cis-
2-amino glycosides, disadvantages of the current state-of-the-art methods include limited substrate scope,
low yields, long reaction times, and anomeric mixtures. We have developed a novel method for the synthesis
of 1,2-cis-2-amino glycosides via nickel-catalyzed R-selective glycosylation with C(2)-N-substituted
benzylidene D-glucosamine and galactosamine trichloroacetimidates. These glycosyl donors are capable
of coupling to a wide variety of alcohols to provide glycoconjugates in high yields with excellent levels of
R-selectivity. Additionally, only a substoichiometric amount of nickel (5-10 mol %) is required for the reaction
to occur at 25 °C. The current nickel method relies on the nature of the nickel-ligand complex to control
the R-selectivity. The reactive sites of the nucleophiles or the nature of the protecting groups have little
effect on the R-selectivity. This methodology has also been successfully applied to both disaccharide donors
and acceptors to provide the corresponding oligosaccharides in high yields and R-selectivity. The efficacy
of the nickel procedure has been further applied toward the preparation of heparin disaccharides, GPI
anchor pseudodisaccharides, and R-GluNAc/GalNAc. Mechanistic studies suggest that the presence of
the substituted benzylidene functionality at the C(2)-amino position of glycosyl donors is crucial for the
high R-selectivity observed in the coupling products. Additionally, the R-orientation of the C(1)-trichloro-
acetimidate group on glycosyl donors is necessary for the coupling process to occur.
Introduction
The synthesis of 1,2-trans-2-amino glycosides can be ac-
complished by employing donors with C(2)-participatory pro-
tecting groups. In contrast, despite the variety of methods
available, the stereoselective construction of 1,2-cis-2-amino
glycosides remains challenging because it requires donors with
nonassisting neighboring groups at the C(2)-amino position. The
most commonly used method employs glycosyl donors with a
C(2)-azido functionality as the nonparticipatory group.3 Under
the C(2)-azido method, the anomeric selectivity at the newly
formed glycosidic bond can be difficult to predict, forming
mixtures of R- and ꢀ-isomers.4 In response to this problem,
Kerns reported an elegant strategy utilizing glycosyl donors
containing an oxazolidinone group spanning the C(2)- and C(3)-
positions.5 This approach requires at least 2 equiv of the
The C(2)-aminoglycosides are key components in glycopro-
teins, one of the most important classes of naturally occurring
oligosaccharides and glycoconjugates.1 Over half of biologically
important proteins are glycosylated in the form of glycoproteins.
Many of these aminosugars are found on cell surfaces and play
a crucial role as receptor ligands for macromolecules such as
lectins, antibodies, and enzymes.2 They also participate in
antibody-antigen interactions. Glycosides of 2-aminosugars are
linked to other sugar residues or serine/threonine amino acids
via either 1,2-cis- or 1,2-trans-2-amino O-glycosidic bonds
(Figure 1).
(1) (a) Petitou, M.; van Boeckel, C. A. A. Angew. Chem., Int. Ed. 2004,
43, 3118–3133. (b) Echardt, K. J. Nat. Prod. 1983, 46, 544–550. (c)
Magnet, S.; Blanchard, J. S. Chem. ReV. 2005, 105, 477–497. (d)
Danishefsky, S. J.; Allen, J. R. Angew. Chem., Int. Ed. 2000, 39, 836–
837. (e) Marcaurelle, L. A.; Bertozzi, C. R. Glycobiology 2002, 12,
69R–77R. (f) Ferguson, M. A. J.; Williams, A. F. Annu. ReV. Biochem.
1988, 57, 285–320. (g) Bongat, A. F. G.; Demchenko, A. V.
Carbohydr. Res. 2007, 342, 374–406.
(3) (a) Paulsen, H.; Kalar, C.; Stenzel, W. Chem. Ber. 1978, 111, 2358–
2369. (b) Lemieux, R. U.; Ratclife, R. M. Can. J. Chem. 1979, 57,
1244–1251. (c) Paulsen, H.; Lorentzen, J. P.; Kutschker, W. Carbo-
hydr. Res. 1985, 136, 153–176. (d) Seeberger, P. H.; Roehrig, S.;
Schell, P.; Wang, Y.; Christ, W. J. Carbohydr. Res. 2000, 328, 61–
69.
(2) (a) Lemieux, R. U. Chem. Soc. ReV. 1989, 18, 347–374. (b) Herzner,
H.; Reipen, T.; Schultz, M.; Kunz, H. Chem. ReV. 2000, 100, 4495–
4537. (c) Davis, B. G. Chem. ReV. 2002, 102, 579–601. (d) Lehle, L.;
Strahl, S.; Tanner, W. Angew. Chem., Int. Ed. 2006, 45, 6802–6818.
(4) Marcaurelle, L. A.; Bertozzi, C. R. Glycobiology 2002, 12, 69R–77R.
(5) (a) Benakli, K.; Zha, C.; Kerns, R. J. J. Am. Chem. Soc. 2001, 123,
9461–9462. (b) Manabe, S.; Ishii, K.; Ito, Y. J. Am. Chem. Soc. 2006,
128, 10666–10667.
9
14288 J. AM. CHEM. SOC. 2010, 132, 14288–14302
10.1021/ja106682m  2010 American Chemical Society

Formation of 1,2-cis-2-Amino Glycosides
A R T I C L E S
Figure 1. Representative structures of C(2)-aminosugars.
Scheme 1
Scheme 2a
a (i) p-Anisaldehyde, NaOH (1 M); (ii) Ac2O, DMAP, pyridine; (iii) NH3 in MeOH, THF; (iv) Cl3C-CN, DBU, CH2Cl2.
activating reagent (PhSOTf), and undesired N-glycosylation6
and N-sulfenylation5 are also observed in the reaction. Ad-
ditionally, the stereochemical outcome of the newly formed
glycosidic bond is dependent on reactivity of alcohol acceptors.7
Later, Schmidt reported the conjugate addition of serine/
threonine to C(2)-nitro-galactals in the presence of t-BuOK to
afford 1,2-cis-2-amino glycopeptides in good yields.8 Gin has
recently reported the opening of aziridines with C(1)-hemiacetal
nucleophiles to form the corresponding R-O-glycosyl serine
conjugates in good selectivity.9 These two latter methods are
limited in substrate scope.
Alternatively, use of a p-methoxy-benzylidene moiety as the
protecting group for the C(2)-nitrogen on a glycosyl bromide
to form 1,2-cis-2-amino glycosides was investigated over 40
years ago.10 Its use was, however, complicated by the multistep
synthesis required for the preparation of glycosyl bromide.
Additionally, the stereochemical outcome of the coupling
process depends on the nature of the promoters as well as the
alcohol acceptors. For instance, using stoichiometric amounts
of HgCN provided the desired glucosides either selectively as
the R-isomers11a or as the ꢀ-isomers12 depending on the nature
of the alcohol acceptors. In contrast, the use of AgOTf (2 equiv)
as a promoter provided exclusively ꢀ-glycosides.12 Use of
n-pentenyl donor provided 1,2-cis-2-amino glycosides in moder-
ate yields.11b On the other hand, no reaction was observed with
use of C(2)-p-methoxy-benzylidene D-glucosamine trichloro-
acetimidate 1 as the donor under traditional Lewis acid
conditions (Scheme 1).12
Given our own interests in merging transition-metal-catalyzed
reaction methodologies with carbohydrate synthesis to stereo-
selectively construct a variety of R- or ꢀ-glycosidic bonds,13
we explored the possibility for the synthesis of 1,2-cis-2-amino
glycosides 2 (Scheme 1). The plan was to exploit the ability of
cationic nickel to direct R-selective glycosylation with C(2)-
N-substituted benzylidene D-glucosamine and galactosamine
trichloroacetimidate donors 1 (Scheme 1).14 As a result of mild
reaction conditions (5-10 mol % of nickel, 25 °C, 3-12 h)
with the R-selectivity controlled by the nature of the cationic
nickel-ligand complex, this new and different departure from
the current state-of-the-art methods would overcome the prob-
lems associated with the preparation of the challenging 1,2-cis-
2-amino glycosides.
Results and Discussion
Initial Studies. We initially explored C(2)-p-methoxy-ben-
zylidene D-glucosamine 5 as the model donor to probe issues
of both reactivity and selectivity of the coupling process under
transition-metal-catalyzed reaction conditions because this glu-
cosamine donor had been previously studied under Lewis acid
conditions.12 Thus, our goal was to develop a simple and
efficient procedure for the synthesis of 5 (Scheme 2). The
preparation of 5 began with commercially available D-glu-
cosamine (3). Exposure of 3 to p-anisaldehyde and NaOH (1
M) followed by acetylation provided 4 in 83% yield over two
steps.15 Selective C(1)-O-deacetylation and subsequent treatment
of the resulting hemiacetals with Cl3C-CN and DBU afforded
(6) Kerns, R. J.; Zha, C.; Benakli, K.; Liang, Y.-Z. Tetrahedron Lett. 2003,
44, 8069–8072.
(12) Marra, A.; Sinay, P. Carbohydr. Res. 1990, 200, 319–337.
(13) (a) Schuff, B. P.; Mercer, G. J.; Nguyen, H. M. Org. Lett. 2007, 9,
3173–3176. (b) Yang, J.; Mercer, G. J.; Nguyen, H. M. Org. Lett.
2007, 9, 4231–4234. (c) Mercer, G. J.; Yang, J.; McKay, M. J.;
Nguyen, H. M. J. Am. Chem. Soc. 2008, 130, 11210–11218. (d) Yang,
J.; Cooper-Vanosdell, C.; Mensah, E. A.; Nguyen, H. M. J. Org. Chem.
2008, 73, 794–780. (e) Menash, E. A.; Azzarelli, J. M.; Nguyen, H. M.
J. Org. Chem. 2009, 74, 1650–1657. (f) McKay, M. J.; Naab, B. D.;
Mercer, G. J.; Nguyen, H. M. J. Org. Chem. 2009, 74, 4705–4711.
(14) Mensah, E. A.; Nguyen, H. M. J. Am. Chem. Soc. 2009, 131, 8778–
8780.
(7) Wei, P.; Kerns, R. J. J. Org. Chem. 2005, 70, 4195–4198.
(8) Winterfeld, G. A.; Schmidt, R. R. Angew. Chem., Int. Ed. 2001, 40,
2654–2657.
(9) Ryan, D. A.; Gin, D. Y. J. Am. Chem. Soc. 2008, 130, 15228–15229.
(10) Hardy, F. F.; Buchanan, J. G.; Baddiley, J. J. Chem. Soc. 1963, 3360–
3366.
(11) (a) Watanabe, I.; Tsuchiya, T.; Takase, T.; Umezawa, S.; Umezawa,
H. Bull. Chem. Soc. Jpn. 1977, 50, 2369–2374. (b) Mootoo, D. R.;
Konradsson, P.; Fraser-Reid, B. J. Am. Chem. Soc. 1989, 111, 8540–
8542.
9
J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010 14289

A R T I C L E S
Mensah et al.
Table 1. Optimization of Reaction Conditions for Selective
Formation of 1,2-cis-2-Amino Glycosidesa
% Ni(4-F-PhCN)4(OTf)2 as the catalyst provided the desired
disaccharide 10 in 93% yield with R:ꢀ ) 10:1 (entry 4).
Comparatively, the more electron-rich 4-methoxy-benzonitrile
ligand increased the reaction time (entry 5). A similar coupling
was also attempted with neutral Ni(PhCN)4Cl2 (entry 7), and
the desired disaccharide 10 was not observed in the reaction.
Because a stoichiometric amount of AgOTf had been used to
activate trichloroacetimidate donors,18 a control experiment was
performed to determine if trichloroacetimidate 5 was indeed
activated by 10 mol % of AgOTf. Less than 1% conversion
was detected from this experiment (entry 8). To determine if
triflic acid, which may be generated from cationic nickel, is
the source of catalysis, the coupling of 9 with 5 was performed
in the presence of 10 mol % of triflic acid (entry 9). The desired
disaccharide 10 was isolated in 10% yield with R:ꢀ ) 3:1. The
results obtained from these control experiments suggest that the
presence of the cationic nickel is important and it is not just
simply acting as a Lewis acid.
The efficacy of this nickel-catalyzed R-glycosylation reaction
was further explored with other N-substituted benzylidene
trichloroacetimidate donors, 6-819 (entries 10-12). Coupling
of 9 with C(2)-N-benzylidene donor 6 provided disaccharide
11 in 92% yield with R:ꢀ ) 10:1 (entry 10). On the other hand,
switching to the electron-withdrawing N-substituted benzylidene
donors 7 and 8 shortened the reaction time from 3 to 1 h (entries
11 and 12).20 Overall, cationic nickel ion acts as the efficient
catalyst to activate C(2)-N-substituted benzylidene D-glu-
cosamine donors 5-8, providing the desired disaccharides
10-13 in high yields and with excellent R-selectivity. In
contrast, coupling of glycosyl acceptor 9 with C(2)-azido
derivative of donors 5-8 yielded the desired disaccharide in a
2:1 R:ꢀ ratio.21
Substrate Scope. With the optimal conditions in hand, we
were able to examine the coupling of a number of primary,
secondary, and tertiary alcohols 14-21 with C(2)-N-substituted-
benzylidene D-glucosamine trichloroacetimidate donors 5-8.
For example, glycosylation of 14 with 5 proceeded smoothly
to provide disaccharide 22 in 77% yield with an R:ꢀ ratio of
20:1 (entry 1). In contrast, coupling of 14 with 5 did not occur
in the presence of TMSOTf as a promoter, and use of BF3 ·OEt2
resulted in only a trace amount of disaccharide 22 (entry 1).12
Furthermore, the nickel chemistry is more R-selective than other
methods. For example, coupling of 14 with a glycosyl bromide
derivative of donor 5 in the presence of AgOTf (1.5 equiv) as
the activating reagent provided the ꢀ-isomer of disaccharide 22
as the major product (R:ꢀ ) 1:9).13 On the other hand, coupling
of 14 with C(2)-oxazolidinone thioglycoside donor afforded the
product in 81% yield with moderate R-selectivity (R:ꢀ ) 3:1).22
The efficacy of the nickel method was further explored with an
yieldb
(%)
loading
(mol %)
time
(h)
c
entry donor
catalyst
R:ꢀ
1
2
3
4
5
6
7
8
5
5
5
5
5
5
5
5
5
6
7
8
Pd(PhCN)2(OTf)2
Ni(PhCN)4(OTf)2
Ni(4-CF3-PhCN)4(OTf)2
Ni(4-F-PhCN)4(OTf)2
Ni(4-MeOPhCN)4(OTf)2
Ni(dppe)(OTf)2
Ni(4-F-PhCN)4Cl2
AgOTf
TfOH
5
5
5
5
5
5
5
10
10
5
10
4
3
3
6
60
95
90
93
76
95
NR
<1%
10
4:1
8:1
10:1
10:1
10:1
8:1
7
10
10
5
3
1
9
3:1
10:1
9:1
10
11
12
Ni(4-F-PhCN)4(OTf)2
Ni(4-F-PhCN)4(OTf)2
Ni(4-F-PhCN)4(OTf)2
92
96
87
5
5
1
9:1
a The reactions were performed with 5 mol % LnNi(OTf)2, which was
generated in situ from 5 mol % LnNiCl2 and 10 mol % AgOTf.
c
b Isolated yield. 1H NMR ratio.
the corresponding trichloroacetimidate 5. With glycosyl donor
5 in hand, attention was focused on identifying conditions for
the stereoselective glycosylation reactions.
Preliminary studies examined the feasibility of the glycosy-
lation of galactose acceptor 9 with trichloroacetimidate donor
5 in the presence of 5 mol % of cationic palladium,
Pd(PhCN)2(OTf)2, as the activating reagent (Table 1, entry 1).13e
Although the coupling process was quite sluggish, the desired
disaccharide 10 was still isolated in 60% yield as a 4:1 mixture
of R- and ꢀ-isomers (entry 1). Despite the fact that the reaction
time, yield, and R-selectivity were not encouraging, we envi-
sioned that additional improvement would be accomplished
through the modification of the catalyst. Indeed, treatment of
the coupling partners 5 and 9 with 5 mol % of cationic nickel,
Ni(PhCN)4(OTf)2, generated in situ from Ni(PhCN)4Cl2 and
AgOTf in CH2Cl2, afforded disaccharide 10 with an improved
yield of 95% (entry 2) and a significantly higher R-selectivity
(R:ꢀ ) 4:1 f 8:1).16 Additional studies focused on the effect
of the nickel-ligand complex on the yields and anomeric
selectivity of the coupling products (entries 3-6).17 Gratifyingly,
it was found that the more electron-withdrawing substituted
benzonitrile ligands shortened the reaction time and increased
the R-selectivity (entries 3 and 4). For instance, use of 5 mol
(18) Douglas, S. P.; Whitfield, D. M.; Krepinsky, J. J. J. Carbohydr. Chem.
1993, 12, 131–136.
(15) (a) Silva, D. J.; Wang, H.; Allanson, N. M.; Jain, R. K.; Sofia, M. J.
J. Org. Chem. 1999, 64, 5926–5929. (b) Aly, M. R. E.; Schmidt, R. R.
Eur. J. Org. Chem. 2005, 4382–4392. (c) Perez, E. M. S.; Avalos,
M.; Babiano, R.; Cintas, P.; Light, M. E.; Jimenez, J. L.; Palacios,
J. C.; Sancho, A. Carbohydr. Res. 2010, 345, 23–32.
(19) Donor 6 was prepared using the same route for the preparation of
donor 5. See the Supporting Information for the preparation of donors
7 and 8.
(20) These results suggest that electron-withdrawing benzylidene donors
accelerate the rate of reaction. To further confirm this hypothesis, we
performed the control experiment with a 1:1 mixture of both electron-
donating 4-methoxy-benzylidene D-glucosamine donor 5 and electron-
withdrawing 4-fluoro-benzylidene D-glucosamine donor 7. In the
presence of galactose 9 as the nucleophilic acceptor, a 3:1 ratio of
disaccharide 12 to disaccharide 10 was observed under nickel
conditions.
(16) To determine if the R:ꢀ ratio of the coupling disaccharide 10 was
thermodynamically or kinetically derived, the ꢀ-isomer of 10 was
subjected to the similar nickel conditions. The ꢀ-isomer was not
converted to the corresponding R-isomer. This result supports our
hypothesis that the anomeric ratios of the 1,2-cis-2-amino glycoside
products are kinetically derived and are not reflective of a thermody-
namic product distribution arising from postcoupling anomeric
epimerization.
(21) Park, J.; Kawatkar, S.; Kim, J.-H.; Boons, G.-J. Org. Lett. 2007, 9,
1959–1962.
(17) Park, N. H.; Nguyen, H. M. Org. Lett. 2009, 11, 2433–2436.
(22) Geng, Y.; Zhang, L.-H.; Ye, X.-S. Chem. Commun. 2008, 597–599.
9
14290 J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010

Formation of 1,2-cis-2-Amino Glycosides
A R T I C L E S
Table 2. R-Selective Coupling with C(2)-N-Substituted Benzylidene D-Glucosamine Donorsa
c
a The reactions were performed with 5-10 mol % of Ni(4-F-PhCN)4(OTf)2 in CH2Cl2 at 25 °C. b Isolated yield. 1H NMR ratio.
electron-withdrawing nucleophilic acceptor 15 (entry 2), and
the desired disaccharide 24 was isolated in 78% yield with R:ꢀ
) 15:1. In contrast, glycosylation of 15 with a C(2)-azido
derivative of donor 5 under BF3 ·OEt2-mediated conditions
yielded the coupling product in an 8:1 mixture of R- and
ꢀ-isomers.21 We next surveyed the possibility that secondary
alcohols 17-19 might be the viable nucleophilic acceptors
(entries 4-7). Accordingly, coupling of C(2)-hydroxyl group
of mannose derivative 18 with C(2)-N-substituted benzylidene
donors 5-7 provided disaccharides 28-30 in good yields
(87-97%) and exclusively as the R-isomers (entry 5). With
Kochetkov’s C(2)-azido thiocyanate donor, coupling of 18 also
afforded the R-isomer as the major product, albeit in lower yield
(72%).23 When dihydrocholesterol 19 was employed as a
nucleophile, the corresponding glycoconjugate 31 was obtained
in 85% yield with an 11:1 R:ꢀ ratio (entry 6). Under the
oxazolidinone method, the glycoconjugate product was formed
exclusively as the ꢀ-isomer using dihydrocholesterol as the
acceptor.24 The R-isomer, however, could be obtained when a
large quantity of AgOTf (0.4 equiv) was employed to promote
the anomeric epimerization of the kinetically formed ꢀ-isomer
into the corresponding thermodynamically favored R-isomer.24
The sterically hindered alcohol acceptor 20 (Table 2) is known
to provide the coupling products in poor yields and R-selecti-
vity.21,25 For instance, coupling of 20 with C(2)-oxazolidinone
thioglycoside donor afforded the desired disaccharide in 82%
yield with R:ꢀ ) 2:1.25 Under the C(2)-azido method, coupling
of 20 provided the expected disaccharide exclusively as the
R-isomer, albeit in lower yield (40%).21 To demonstrate that
the C(4)-hydroxyl functionality of D-glucopyranoside 20 (Table
(24) Boysen, M.; Gemma, E.; Lahmann, M.; Oscarson, S. Chem. Commun.
2005, 3044–3046.
(23) Kochetkov, N. K.; Klimov, E. M.; Malysheva, N. N.; Demchenko,
A. V. Carbohydr. Res. 1993, 242, C7–C10.
(25) Geng, Y.; Zhang, L.-H.; Ye, X.-S. Tetrahedron 2008, 64, 4949–4958.
9
J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010 14291

A R T I C L E S
Mensah et al.
Table 3. R-Selective Coupling with D-Galactosamine
2, entry 7) can be employed as a viable nucleophile under our
nickel conditions, it was coupled with glycosyl donors 5, 7,
and 8 (Table 2, entry 7). The desired disaccharides 32-34 were
obtained in good yields and with moderate to excellent
R-selectivity. These results clearly show that the nature of the
benzylidene groups on glycosyl donors plays an important role
in the anomeric selectivity of the coupling products. For
instance, use of an electron-donating 4-methoxy-benzylidene
donor 5 provided disaccharide 32 with R:ꢀ ) 6:1 (entry 7). On
the other hand, coupling of glycosyl acceptor 20 with an
electron-withdrawing 4-trifluoromethyl-benzylidene donor 8
afforded the corresponding disaccharide 34 exclusively as the
R-isomer (entry 7). The efficiency of the nickel chemistry was
further explored with tertiary alcohol 21 (entry 8). The desired
glycoconjugate 35 was obtained in 96% yield and with excellent
R-selectivity (R:ꢀ ) 17:1). Overall, the nickel-catalyzed R-se-
lective coupling with C(2)-N-substituted benzylidene D-glu-
cosamine donors 5-8 provided a practical approach to get
access to a variety of disaccharides and glycoconjugates in high
yields and excellent levels of R-selectivity. Additionally, it does
not require PhSEt as a putative nucleophile (the azide method)21
or a large quantity of AgOTf (the oxazolidinone method)24 in
order to improve the R-selectivity. It requires only a substo-
ichiometric amount of cationic nickel (5-10 mol %) to activate
glycosyl donors 5-8.
Trichloroacetimidatea
Similarly, C(2)-N-4-methoxy-benzylidene D-galactosamine
3626 was found to be a viable donor in the reaction with primary
alcohol 15, secondary alcohols 17-19, and tertiary alcohol 21
(Table 3). Each coupling reaction was high yielding (74-93%)
and R-selective (R:ꢀ ) 10:1 to <20:1) regardless of the position
of the hydroxyl moiety and the nature of the protecting groups
on the nucleophilic acceptors.
To demonstrate that the current nickel method can be
employed for oligosaccharide synthesis, a number of trisaccha-
rides and tetrasaccharides were prepared from disaccharide
acceptors (Table 4) and donors (Table 5). The prospect of
trisaccharide synthesis was first evaluated with disaccharide
nucleophiles 42-44 (Table 4). In the first [1 + 2] convergent
approach, reaction of the disaccharide acceptor 4227 with C(2)-
N-4-fluoro-benzylidene D-glucosamine donor 7 provided trisac-
charide 45 (57%) with a 13:1 R:ꢀ ratio (entry 1).28 Similarly,
coupling of 42 with C(2)-N-4-trifluoromethyl-benzylidene elec-
trophile 8 afforded a 56% yield of 46 with a 14:1 R:ꢀ ratio.
Switching to a more electron-donating disaccharide acceptor
4327 provided higher yielding trisaccharides 47 and 48 (entry
2). Secondary alcohol of disaccharide acceptor 4429 was also a
viable nucleophile, and the desired trisaccharide 49 was obtained
in 76% yield with an 11:1 R:ꢀ ratio (entry 3).
a The reactions were performed with 5-10 mol
%
of
c
Ni(4-F-PhCN)4(OTf)2 in CH2Cl2 at 25 °C. b Isolated yield. 1H NMR
ratio.
R-selectivity (R:ꢀ >20:1). Use of the more sterically hindered
secondary alcohol 20 provided trisaccharide 53 in lower yield
and R-selectivity than that of 18 (entry 2). In a [2 + 2] approach,
coupling of disaccharide acceptor 43 with disaccharide donor
50 proceeded smoothly to provide tetrasaccharide 55 in 72%
yield with R:ꢀ ) 11: 1 (entry 3).
In [2 + 1] and [2 + 2] convergent approaches, disaccharide
trichloroacetimidates 50 and 5130 were investigated as feasible
donors (Table 5).28 In a [2 + 1] strategy, coupling of mannose
acceptor 18 with both donors 50 and 51 provided trisaccharides
52 and 53 (entry 1) in good yields (>70%) and with excellent
Synthetic Applications
(26) D-Galactosamine trichloroacetimidate donor 36 was prepared using
the same route for the preparation of D-glucosamine donor 5.
(27) (a) Yi, H.; Maisonneuve, S.; Xie, J. Org.Biol.Chem. 2009, 7, 3847–
3854. (b) Pan, Q.; Du, Y.; Kong, F.; Pan, J.; Lu, M. J. Carbohydr.
Chem. 2001, 20, 297–306.
Synthesis of Glucosamine-r-(1f4)-Linked-Glucuronic Acid
Disaccharide of Heparin. During the course of our methodology
studies, we became interested in the synthesis of heparin, which
is a sulfated and anionic polysaccharide consisting of alternating
R-(1f4)-linked-disaccharide units of L-iduronic acid or D-
glucuronic acid and D-glucosamine components.31 This polysac-
charide was discovered in 1916 and introduced into clinical use
in 1935.24 It has been widely recognized to participate in a
(28) We only explored with the electron-withdrawing donors 7 and 8
because they would undergo glycosylation at a rate faster than that of
4-methoxy-benzylidene donor 5.
(29) Meoncelli, P.; Lowary, T. L. Aus. J. Chem. 2009, 62, 558–574.
(30) The procedures for preparation of disaccharide trichloroacetimidate
donors 50 and 51 are described in the Supporting Information.
9
14292 J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010

Formation of 1,2-cis-2-Amino Glycosides
A R T I C L E S
Table 4. R-Selective Glycosylation of Disaccharide Acceptorsa
c
a The reactions were performed with 5-10 mol % of Ni(4-F-PhCN)4(OTf)2 in CH2Cl2 at 25 °C. b Isolated yield. 1NMR ratio.
variety of biological functions including blood anticoagulation,
cell differentiation, cell growth, inflammation, and pathogen
infection.32 Currently, more than 100 heparin binding proteins
have been discovered.33 However, for most heparin binding
proteins there is a lack of detailed knowledge of the ligand
requirements for binding and mediating biological activity.34
This is due to the microhetereogeneity of heparins and the
difficulties associated with preparation of well-defined heparin
oligosaccahrides.35 Thus, to understand the effect of heparin
structure on its biological activity, well-defined heparin se-
quences are required. One of the powerful tools for obtaining
well-defined low molecular weight heparins (LMWH) is through
chemical synthesis.36 Although there have been remarkable
advances in the synthesis of heparin oligosaccharides, limitations
include long synthetic steps, low yields, and anomeric mix-
tures.37 One of the major challenges in the preparation of well-
defined heparin oligosaccharides is to achieve high selectivity
in the synthesis of R-(1f4)-linked-disaccharide unit of D-
glucuronic acid and D-glucosamine components.
Thus, our goal was to determine how D-glucuronic acid
acceptors 56-5836 would behave in the nickel-catalyzed
R-selective glycosylation with C(2)-N-substituted benzylidene
D-glucosamine donors 5-8 (Table 6). The key model investiga-
tions in establishing the feasibility of this method involved a
series of couplings of glucuronic acid acceptors 56-58 with
(31) (a) Karst, N. A.; Linhardt, R. J. Curr. Med. Chem. 2003, 10, 1993–
2031. (b) Avci, F. Y.; Karst, N. A.; Linhardt, R. J. Curr. Pharm. Des.
2003, 9, 2323–2335. (c) Petitou, M.; van Boeckel, C. A. A. Angew.
Chem., Int. Ed. 2004, 43, 3118–3133. (d) Liu, H.; Zhang, Z.; Linhardt,
R. J. Nat. Prod. Rep. 2009, 26, 313–321.
(36) (a) Petitou, M.; van Boeckel, C. A. A. Prog. Chem. Org. Nat. Prod.
1992, 60, 143–210. (b) van Boeckel, C. A. A.; Petitou, M. Angew.
Chem., Int. Ed. 1993, 32, 1671–1690. (c) Arungundram, S.; Al-Mafraji,
K.; Asong, J.; Leach, F. E., III; Amster, I. J.; Venot, A.; Turnbull,
J. E.; Boons, G.-J. J. Am. Chem. Soc. 2009, 131, 17394–17405.
(37) (a) Orgueira, H. A.; Bartolozzi, A.; Schell, P.; Litjens, R. E. J. N.;
Palmacci, E. R.; Seeberger, P. H. Chem.sEur. J. 2003, 9, 140–169.
(b) Codee, J. D. C.; Stubba, B.; Schiattarella, M.; Overkleef, H. S.;
van Boeckel, C. A. A.; van Boom, J. H.; van der Marel, G. A. J. Am.
Chem. Soc. 2005, 127, 3767–3773. (c) Haller, M.; Boons, G.-J.
J. Chem. Soc., Perkin Trans. 1 2001, 814–822.
(32) (a) Lindahl, U.; Kusche-Gullberg, M.; Kjellen, L. J. Biol. Chem. 1998,
273, 24979–24982. (b) Bernfiled, M.; Gotte, M.; Park, P. W.; Reizes,
O.; Fitzgerald, M. L.; Lincecum, J.; Zako, M. Annu. ReV. Chem. 1999,
68, 729–777. (c) Capila, I.; Lindhardt, R. J. Angew. Chem., Int. Ed.
2002, 41, 390–412.
(33) Ori, A.; Wilkinson, M. C.; Ferning, D. G. Front. Biosci. 2008, 13,
4309–4338.
(34) Kreuger, J.; Spillman, D.; Li, J. P.; Lindahl, U. J. Cell. Biol. 2006,
174, 323–327.
(35) Rabenstein, D. Nat. Prod. Rep. 2002, 19, 312–331.
9
J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010 14293

A R T I C L E S
Mensah et al.
Table 5. R-Selective Glycosylation with Disaccharide Donorsa
c
a The reactions were performed with 5-10 mol % of Ni(4-F-PhCN)4(OTf)2 in CH2Cl2 at 25 °C. b Isolated yield. 1NMR ratio.
C(2)-p-methoxy-benzylidene glucosamine donor 5 (Table 6,
entries 1-3). Use of both glucuronic acid benzyl ester 56
and allyl ester 57 (entries 1 and 2) provided disaccharides
59 and 60, respectively, in moderate to good yields (55-84%)
and with good R-selectivity (R:ꢀ ) 9:1-14:1). Gratifyingly,
it was found that use of glucuronic acid methyl ester 58
provided the desired disaccharide 61 exclusively as the
R-isomer (entry 3). This result prompted us to further
investigate the glycosylation of methyl ester D-glucuronic acid
acceptor 58 with other C(2)-N-substituted benzylidene donors
6-8 (entry 4). The C(2)-N-p-trifluoromethyl-benzylidene
derivative 8 was found to be the most effective donor, and
disaccharide 64 was isolated in 87% yield and exclusively
as the R-isomer. Compared to other systeyms,37 our nickel
method is much more R-selective. For instance, Seeberger
and co-workers have reported in a modular synthesis of
heparin oligosaccharides that coupling of D-glucuronic acid
acceptor with a C(2)-azido trichloroacetimidate donor pro-
vided the disaccharide product in 57% yield with R:ꢀ )
3:1.37a Under dehydrative glycosylation conditions,38 cou-
pling of D-glucuronic acid acceptor with a C(1)-hydroxyl
glucoazido donor afforded the desired disaccharide in 76%
yield with an 7:1 R:ꢀ ratio.37b Because of its low reactivity,
D-glucuronic acid derivatives are often masked as D-glu-
copyranosides that undergo selective oxidation of the C(6)-
hydroxyl group to provide glucuronic acid after the assembly
of oligosaccharides is complete. Even with this strategy,
glycosylation of D-glucopyranoside acceptor with a C(2)-
azido trichloroacetimidate still provided the coupling product
in 78% yield with R:ꢀ ) 3:1.37c
Synthesis of the Pseudodisaccharide Core of GPI Anchors.
Glycosylphosphatidylinositol (GPI) anchors are a large family
of glycolipids that serve to attach many eukaryotic proteins onto
the outer leaflet of the cell membrane.39 Proteins containing GPI
anchors are functionally diverse and play significant roles in a
variety of biological processes such as signal transduction, prion
disease pathogenesis, immune response, and the patho-biology
(38) (a) Garcia, B. A.; Poole, J. L.; Gin, D. Y. J. Am. Chem. Soc. 1997,
119, 7597–7598. (b) Garcia, B. A.; Gin, D. Y. J. Am. Chem. Soc.
2000, 122, 4269–4279.
(39) (a) Paulick, M. G.; Bertozzi, C. R. Biochemistry 2008, 47, 6991–
7000. (b) Ferguson, M. A. J. Cell. Sci. 1999, 112, 2799–2809. (c)
Tiede, A.; Bastsch, I.; Schubert, J.; Orlean, P.; Schmidt, R. E. Biol.
Chem. 1999, 380, 503–523.
9
14294 J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010

Formation of 1,2-cis-2-Amino Glycosides
A R T I C L E S
Table 6. R-Selective Glycosylation of Glucuronic Acid Acceptorsa
c
a The couplings were performed with 5 mol % of Ni(4-F-PhCN)4(OTf)2 in CH2Cl2 at 25 °C. b Isolated yield. 1NMR ratio.
of trypanosomal parasites.40 Because of structural complexity
and limited quantities, the connection between the structure of
GPI anchors and their biological function in mammalian cells
is difficult to study. In an effort to define the functional
importance of GPI anchors, several total syntheses of GPI
anchors have been reported.41 One of the major challenges in
the total synthesis of native GPI anchors is the stereoselective
construction of the pseudodisaccharide unit of D-glucosamine
and inositol components with high R-selectivity.41 For instance,
it has been reported that coupling of inositol with C(2)-azido
glycosyl bromide in the presence of AgClO4 as the activating
reagent provided the desired pseudodisaccharide as a 3:1 mixture
of R- and ꢀ-isomers.42 Although switching to n-Bu4NBr as the
activating reagent provided the coupling product exclusively
as the R-isomer, it took 72 h for the reaction to go to
completion.43 In the total synthesis of the P. falciparum GPI
anchor, coupling of inositol with a C(2)-azido trichloroacetimi-
date provided pseudodisaccharide with an 4:1 R:ꢀ ratio.44 In
the synthesis of GPI anchor bearing unsaturated lipid chains,
the desired pseudodisaccharide was formed as a 1.2:1 mixture
of R- and ꢀ-isomers.45
To address this problem, we investigated the coupling of
inositol nucleophile 6546 with C(2)-N-substituted benzylidene
D-glucosamine trichloroacetimidate donors 5-8 (Table 7). To
our delight, the coupling products 66-69 were formed in good
yields and with excellent R-selectivity (entries 1-4). The best
result was with C(2)-N-4-fluoro-benzylidene D-glucosamine
donor 7 (entry 3), and the desired pseudodisaccharide 68 was
isolated in 70% yield and exclusively as the R-anomer.
(40) (a) Nosjean, O.; Briolay, A.; Roux, B. Biochim. Biophys. Acta 1997,
1331, 153–186. (b) Chesebro, B.; Trifilo, M.; Race, R.; Meade-White,
K.; Teng, C.; LaCasse, R.; Raymond, L.; Favara, C.; Baron, G.; Priola,
S.; Caughey, B.; Masliah, E.; Oldstone, M. Science 2005, 308, 1435–
1439.
(43) Baeschlin, D. K.; Chaperon, A. R.; Green, L. G.; Hahn, M. G.; Ince,
S. J.; Ley, S. V. Chem.sEur. J. 2000, 6, 172–186.
(41) For a review, see: Guo, Z.; Bishop, L. Eur. J. Org. Chem. 2004, 358,
5–3596.
(44) Liu, X.; Kwon, Y.-U.; Seeberger, P. H. J. Am. Chem. Soc. 2005, 127,
5004–5005.
(42) Udodong, U. E.; Madsen, R.; Roberts, C.; Fraser-Reid, B. J. Am. Chem.
Soc. 1993, 115, 7886–7887.
(45) Swarts, B. M.; Guo, Z. J. Am. Chem. Soc. 2010, 132, 6648–6650.
(46) Yu, F.; Guo, Z. Bioorg. Med. Chem. Lett. 2009, 19, 3852–3855.
9
J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010 14295

A R T I C L E S
Mensah et al.
Table 7. R-Selective Glycosylation of Inositol Acceptora
c
a The couplings were performed with 5 mol % of Ni(4-F-PhCN)4(OTf)2 in CH2Cl2 at 25 °C. b Isolated yield. 1NMR ratio.
Synthesis of r-Glc- and r-GalNAc-Serine/Threonine Deriva-
tives. Mucin-type glycoproteins carry many of the Lewis and
blood group antigens and serve as ligands for the endothelial
cell-surface receptors L- and P-selectin.47 As a result, they are
important mediators in normal and disease processes and have
received considerable attention in cancer vaccine therapies.48
In mucin O-linked glycoproteins, a GalNAc residue is R-linked
to the oxygen atom on the side chain of serine (Ser) or threonine
(Thr) amino acid. Thus, all mucin O-linked glycans contain a
1,2-cis-2-amino glycosidic bond. One of the major challenges
in the synthesis of tumor-associated mucin antigens is to achieve
high R-selectivity in the connection of a N-acetylgalactosamine
(GalNAc) residue to the oxygen atom on the side chain of serine
or threonine amino acid. Even with simple monosaccharide
galactosamine donors, the stereochemical outcome of the newly
formed glycosidic bond can be difficult to predict and often
results in moderate R-selectivity.49 For instance, coupling of
threonine with C(2)-oxazolidinone thiogalactoside donor has
been reported to provide glycopeptides as a 1:1 mixture of R-
and ꢀ-isomers.6 In the synthesis of mucin-related TN and TF
O-linked antigens, it had been reported that coupling of
threonine with C(2)-azido galactosamine donor afforded the
desired glycopeptide as a 4:1 mixture of R- and ꢀ-isomers.49a
Our goal is to establish the ability of C(2)-N-substituted
benzylidene D-glucosamine and galactosamine trichloroacetimi-
dates to serve as viable donors for the stereoselective synthesis
of both the R-GlcNAc- and R-GalNAc-Ser/Thr derivatives
(Table 8). Under our nickel conditions, coupling of serine 71
and threonine 72 with D-glucosamine trichloroacetimidate donor
5 provided the corresponding glycopeptides 73 and 74 (entry
1), respectively, in good yields (65-73%) and R-selectivity (R:ꢀ
) 8:1-10:1). Use of an electron-withdrawing C(2)-N-4-fluoro-
benzylidene D-glucosamine 7 improved both the yields (75-83%)
and R-selectivity (R:ꢀ ) 9:1-14:1) of the resulting glycopep-
tides 75 and 76 (entry 2). We also explored the coupling of
threonine 72 with both D-galactosamine donors 36 and 70 (entry
3). The desired glycopeptides 77 and 78 were formed, respec-
tively, in good yields (74-81%) and with excellent R-selectivity
(R:ꢀ ) 14:1-15:1). Although both 36 and 70 provided the
(49) (a) Kuduk, S. D.; Schwarz, J. B.; Chen, X.-T.; Glunz, P. W.; Sames,
D.; Ragupathi, G.; Livingston, P. O.; Danishefsky, S. J. J. Am. Chem.
Soc. 1998, 120, 12474–12485. (b) George, S. K.; Schwientek, T.;
Holm, B.; Reis, C. A.; Clausen, H.; Kihlberg, J. J. Am. Chem. Soc.
2001, 123, 11117–11125.
(47) Danishefsky, S. J.; Behar, B.; Randolph, J. T.; Lloyd, K. O. J. Am.
Chem. Soc. 1995, 117, 5701–5711.
(48) Danishefsky, S. J.; Allen, J. R. Angew. Chem., Int. Ed. 2000, 39, 836–
863.
9
14296 J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010

Formation of 1,2-cis-2-Amino Glycosides
A R T I C L E S
Table 8. R-Selective Coupling of Serine and Threonine Amino Acidsa
c
a The couplings were performed with 5-10 mol % of Ni(4-F-PhCN)4(OTf)2 in CH2Cl2 at 25 °C. b Isolated yield. 1H NMR ratio.
products with comparable yield and R-selectivity, the reaction
was faster with an electron-withdrawing benzylidene donor 70.
Overall, our chemistry is much more R-selective than both the
current state-of-the-art methods (e.g., the C(2)-azido (R:ꢀ )
4:1)49a and C(2)-oxazolidinone (R:ꢀ ) 1:1)6 derived donors)
at accessing this family of R-glycosides.
Mechanistic Proposal. Although the exact mechanism for the
nickel-catalyzed selective R-coupling of alcohol nucleophiles
with C(2)-N-substituted benzylidene D-glucosamine and galac-
tosamine trichloroacetimidates to form 1,2-cis-2-amino glyco-
sides awaits further study, we suggest Figure 2 as a working
hypothesis. In pathway A, nickel reversibly coordinates to both
the C(1)-trichloroacetimidate nitrogen and C(2)-benzylidene
group of donor 83 to form a seven-membered ring complex
84. We reasoned that hydrogen bonding between the external
oxygen nucleophile and trichloroacetamide could facilitate
ionization of 84 to form the corresponding complex 85. Ligand
exchange followed by dissociation of trichloroacetamide would
provide ion pair 88. The resulting intermediate 88 then
recombines in a stereoelectronically favored mode to form a
five-membered ring intermediate 89. Dissociation of the nickel
species from 89 would provide 1,2-cis-2-amino glycoside 90.
Alternatively, nickel can act as a mild Lewis acid (pathway
B). This pathway could involve reversible coordination of
LnNi(OTf)2 to the C(1)-trichloroacetimidate nitrogen of 83 to
form the corresponding complex 86. Hydrogen bonding between
the external oxygen nucleophile and trichloroacetamide would
promote ionization of 86, leading to the formation of the
oxocarbenium intermediate 87. Ligand exchange between the
external oxygen nucleophile and trichloroacetamide followed
by coordination of LnNi(OR) to the C(2)-benzylidene nitrogen
would provide the ion pair 88, which ultimately leads to the
desired 1,2-cis-2-amino glycoside 90.
Removal of the C(2)-N-Substituted Benzylidene Protecting
Groups. Although it will be of interest to study the biological
properties of these benzylidenes, it was essential to demonstrate
that these benzylidenes could be converted to the corresponding
N-acetyl or other functionalities. Because the reported condi-
tions15 (5 N HCl) for removal of the benzylidene protecting
groups would not work well with certain acid-sensitive oli-
gosaccharides and glycoconjugates, we screened a number of
acids (TsOH, TfOH, 1 N and 2 N HCl) and found that the
optimal conditions for removing the benzylidene functionality
were with use of 2 N HCl (1.1 equiv), acetone/CH2Cl2, 25 °C,
5 min. For instance, treatment of glycopeptides 78 with 2 N
HCl, acetone/CH2Cl2 at 25 °C for less than 3 min followed by
acetylation of the resulting amine salt intermediate 79 afforded
the fully protected R-GalNAc 80 in 81% yield over two steps
(Scheme 3). The glycoconjugate 80 is a fully protected structure
of a TN-tumor associated mucin antigen.49a Similarly, removal
of N-benzylidene functionality of 31 and 29 followed by
acetylation or sulfation of the resulting amine salt intermediates
provided glycoconjugates 81 and 82 in overall good yields
(Scheme 3).
9
J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010 14297

A R T I C L E S
Mensah et al.
Scheme 3
Our first goal is to determine if the R-orientation of the
trichloroacetimidate leaving group is crucial for ionization and
subsequent formation of the coupling disaccharide under nickel
conditions (Scheme 4). Accordingly, coupling of galactose
acceptor 9 with ꢀ-trichloroacetimidate donor 5 was attempted
in the presence of 5 mol % of Ni(4-F-PhCN)4(OTf)2. The
coupling reaction did not proceed even after stirring for 12 h at
25 °C.
As shown in Figure 2, hydrogen bonding between the external
alcohol nucleophile and the trichloroacetamide is presumed to
be necessary for the facile ionization of a seven-membered ring
intermediate 84 (pathway A) or a Lewis acid complex 87
(pathway B). To verify this hypothesis, the glycosylation of
TMSN3 or NaN3, which lacks an acidic hydrogen, was
performed with the C(2)-N-4-methoxyl-benzylidene D-glu-
cosamine trichloroacetimidate 5 (Scheme 5a). In this reaction,
no coupling products were observed. A second control experi-
ment was attempted in the absence of an external alcohol
nucleophile (Scheme 5b). Trichloroacetamide 91, which was
the result of the [1,3]-rearrangement of 5, was obtained in 71%
yield. The results obtained in Schemes 4 and 5 support our
hypothesis that both the R-orientation of the trichloroacetimidate
leaving group and hydrogen bonding are the two important
factors to promote ionization of the glycosyl trichloroacetimidate
donor.
Our next set of experiments were to determine if the presence
of the benzylidene functionality at the C(2)-position of glycosyl
trichloroacetimidate is necessary for the high R-selectivity
observed in the coupling products. A first control experiment
was attempted with C(2)-azido trichloroacetimidate 92 (Scheme
6a) because this donor is the most commonly used substrate
under traditional Lewis acid conditions to form 1,2-cis-2-amino
glycosides. Coupling of nucleophilic acceptor 9 with donor 92
in the presence of 5 mol % of Ni(4-F-PhCN)4(OTf)2 at 25 °C
afforded disaccharide 94 as a 1:1 mixture of R- and ꢀ-isomer.
We also explored the use of C(2)-N-phenylsulfonamide trichlo-
roacetimidate 93 (Scheme 6b) as the electrophilic donor because
phenylsulfonamide had been used as the directing group for
Figure 2. Proposed mechanism for nickel-catalyzed selective formation of 1,2-cis-2-amino glycosides.
9
14298 J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010

Formation of 1,2-cis-2-Amino Glycosides
A R T I C L E S
Scheme 4
If the presence of the oxygen functionality at the ortho-
position of the benzylidene group on glycosyl donor will shut
down the coupling process, glycosylation of acceptor 9 with
C(2)-o-trifluoromethyl-benzylidene D-glucosamine trichloro-
acetimidate donor 99 would validate this hypothesis (Scheme
8). Indeed, the coupling process proceeded smoothly with 5 mol
% of Ni(4-F-PhCN)4(OTf)2 to provide disaccharide 100 in 81%
yield and with good R-selectivity (R:ꢀ ) 10:1).
Conclusions
palladium-catalyzed C-H activation reactions.50 The ꢀ-isomer
95 was isolated as the major product (R:ꢀ ) 1:5) under nickel
conditions (Scheme 6b). These two experiments highlight the
critical role of the C(2)-N-substituted benzylidene functionality
in the nickel-catalyzed R-selective glycosylation reaction.
To further investigate whether coordination of nickel to both
the C(1)-trichloroacetimidate nitrogen and C(2)-benzyldidene
nitrogen of the glycosyl donor is essential for the coupling
process to occur, coupling of galactose acceptor 9 with C(2)-
o-methoxy-benzylidene D-glucosamine donor 96 was attempted
under nickel conditions (Scheme 7). The coupling products were
not observed in the reaction. This is likely due to the preferential
coordination of nickel to both C(2)-benzylidene nitrogen and
o-methoxy functionality of 96 to form a six-membered nickel
complex 97. Formation of 97 prevents nickel from coordinating
to the C(1)-trichloroacetimidate nitrogen of glycosyl donor 96.
A control experiment was performed in the presence of 10 mol
% of triflic acid (Scheme 7). In this coupling reaction, the desired
disaccharide 97 was isolated in 15% yield with a 1:1 R:ꢀ ratio.
Scheme 5
In summary, we have developed a novel method for the
stereoselective synthesis of 1,2-cis-2-amino glycosides via
nickel-catalyzed R-selective glycosylation with C(2)-N-substi-
tuted benzylidene D-glucosamine and galactosamine trichloro-
acetimidate donors. These glycosyl donors are able to couple
to a number of primary, secondary, and tertiary alcohol acceptors
to provide glycoconjugates in good yields and with excellent
R-selectivity. The current nickel method relies on the nature of
the nickel-ligand complex to control the R-selectivity. The
reactive sites of the nucleophiles or the nature of the protecting
groups have little effect on the selectivity. Additionally, only a
substoichiometric amount of nickel (5-10 mol %) is required
for the coupling reaction to occur at room temperature. This
method has also been applied to both disaccharide donors and
acceptors to provide the corresponding oligosaccharides in high
yields and with excellent levels of R-selectivity. The efficiency
of the nickel chemistry has been further utilized in the
preparation of the high-yielding and R-selective heparin disac-
charides, GPI anchor pseudodisaccharides, and R-GluNAc/
Scheme 6
9
J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010 14299

A R T I C L E S
Mensah et al.
Scheme 7
3H). 13C NMR (CDCl3, 125 MHz): δ 170.6, 170.0, 169.7, 163.9,
162.3, 160.8, 130.2, 128.6, 113.9, 95.8, 91.1, 71.0, 70.8, 70.4, 68.3,
61.8, 55.4, 20.7, 20.6. IR (film, cm-1): ν 3334, 2961, 2839, 1752,
1674, 1642, 1605, 1579, 1512, 1246, 1065, 1021. HRMS (ESI):
calcd for C22H25Cl3N2O9 (M + Na) 589.0518, found 589.0525.
Preparation of Ni(4-F-PhCN)4Cl2.51 A 10 mL oven-dried
Schlenk flask was charged with NiCl2 (129.59 mg, 1 mmol, 1 equiv)
and CH2Cl2 (2 mL). To the resulting yellow solution was added
4-fluoro-benzonitrile (3.03 g, 25 mmol, 25 equiv). The resulting
mixture was stirred at 25 °C for 24 h and then poured into hexane
(40 mL). The hexane layer was decanted. The resulting yellow solid
was further washed with hexane (2 × 20 mL), transferred into a
preweighed vial, and dried under vacuum overnight.
Scheme 8
GalNAc derivatives. Mechanistic studies suggest that the
presence of the substituted benzylidene functionality at the C(2)-
amino position of glycosyl donors is crucial for the high
R-selectivity observed in the coupling products. Furthermore,
we have demonstrated that the R-orientation of the C(1)-
trichloroacetimidate group as well as the presence of the external
alcohol nucleophile are necessary for the facile ionization of
glycosyl trichloroacetimidate donors. We expect that insights
gained from our present studies will help future advances in
the stereoselective formation of 1,2-cis-2-amino glycosides,
including developing new glycosyl donors, activating agents,
and strategies for controlling R-selectivity.
Preparation of Disaccharide 10. A 10 mL oven-dried Schlenk
flask was charged with D-glucosamine trichloroacetimidate donor
5 (85.2 mg, 0.15 mmol, 1 equiv), galactose acceptor 9 (50.7 mg,
0.19 mmol, 1.3 equiv), and CH2Cl2 (1.1 mL). The resulting solution
was cooled to 0 °C, and a preformed solution of Ni(4-F-
PhCN)4(OTf)2, which was generated in situ from a reaction of Ni(4-
F-PhCN)4Cl2 (4.61 mg, 0.0075 mmol, 5 mol %) and AgOTf (3.85
mg, 0.015 mmol, 10 mol %) in dichloromethane (1 mL) for 30
min, was then added to the solution. The resulting mixture was
stirred at 0 °C for 5 min and then warmed to 25 °C. The reaction
mixture was stirred at the ambient temperature for 3 h, diluted with
benzene (2 mL), and purified by silica gel flash chromatography
(3/1, benzene/acetonitrile + 1% triethylamine) to give disaccharide
10 (93 mg, 93%, R:ꢀ ) 10:1). Rf ) 0.33 (benzene/acetonitrile, 3/1
+ 1% TEA). 1H NMR (CDCl3, 500 MHz): δ 8.17 (s, 1H), 7.64 (d,
J ) 10.0 Hz, 2H), 6.86 (d, J ) 10.0 Hz, 2H), 5.61 (t, J ) 10.0 Hz,
1H), 5.46 (d, J ) 5.0 Hz, 1H), 5.08 (t, J ) 10.0 Hz, 1H), 4.89 (d,
J ) 5.0 Hz, 1H), 4.46 (d, J ) 10.0 Hz, 1H), 4.34 (dd, J ) 10.0,
4.0 Hz, 1H), 4.28-4.23 (m, 3H), 4.07 (d, J ) 10.0 Hz, 1H), 4.01
(t, J ) 5.0 Hz 1H), 3.80 (s, 3H), 3.80 (m, 1H), 3.72-3.70 (m,
1H), 3.54 (dd, J ) 10.2, 3.2 Hz, 1H), 2.07 (s, 3H), 2.00 (s, 3H),
1.83 (s, 3H), 1.52 (s, 3H), 1.32 (s, 3H), 1.28 (s, 3H), 1.00 (s, 3H).
13C NMR (CDCl3, 125 MHz): δ 170.9, 170.1, 169.9, 163.6, 163.5,
162.1, 130.2, 128.6, 128,3, 114.0, 108.9, 108.7, 100.5, 96.2, 72.2,
71.5, 70.7, 70.4, 68.9, 67.8, 67.7, 66.4, 62.3, 55.4, 26.2, 25.9, 24.9,
24.0, 20.8, 20.76, 20.6. IR (film, cm-1): ν 3427, 3349, 2987, 2935,
1745, 1642, 1608, 1513, 1376, 1251, 1165, 1110, 1069, 1029.
J(13CH) ) 171 Hz (100.5 Hz). HRMS (ESI): calcd for C32H43 NO14
(M + H) 666.2756, found 666.2753.
Experimental Section
Representative experimental procedures are listed here. Full
experimental details and spectral data for all new compounds can
be found within Supporting Information.
Preparation of D-Glucosamine Trichloroacetimidate Donor
5. A 100 mL oven-dried Schlenk was charged with hemiacetal (2.4
g, 5.67 mmol, 1 equiv) and dichloromethane (30 mL). The solution
was cooled to 0 °C, and trichloroacetonitrile (1.7 mL, 17.01 mmol,
3 equiv) was then added to the reaction mixture followed by DBU
(0.42 mL, 2.84 mmol, 0.5 equiv). The resulting mixture was stirred
at this temperature for 4 h and then concentrated in vacuo. The
residue was purified by silica gel flash chromatography (2/1, hexane/
ethyl acetate + 1% triethylamine) to provide trichloroacetimidate
5 (2.89 g, 90%) as a yellow solid. 1H NMR (CDCl3, 500 MHz): δ
8.56 (s, 1H), 8.22 (s, 1H), 7.61 (d, J ) 10.0 Hz, 2H), 6.86 (d, J )
10.0 Hz, 2H), 6.38 (s, 1H), 5.67 (t, J ) 10.0 Hz, 1H), 5.20 (t, J )
10.0 Hz, 1H), 4.34-4.32 (m, 2H), 4.13 (d, J ) 10.0 Hz, 1H), 3.80
(s, 3H), 3.80-3.77 (m, 1H), 2.07 (s, 3H), 2.03 (s, 3H), 1.86 (s,
Preparation of Trisaccharide 47. A 10 mL oven-dried Schlenk
flask was charged with D-glucosamine trichloroacetimidate donor
(51) (a) Suzuki, H.; Ishiguro, S.-I. Bull. Chem. Soc. Jpn. 1993, 66, 83–88.
(b) Yamamoto, T.; Yamamoto, A.; Ikeda, S. J. Am. Chem. Soc. 1971,
93, 3360–3364. (c) Qian, H.; Pei, T.; Widenhoefer, R. A. Organo-
metallics 2005, 24, 287–301.
(50) (a) Miura, M.; Tsuda, T.; Satoh, T.; Pivsa-Art, S.; Nomura, M. J. Org.
Chem. 1998, 63, 5211–5215. (b) Li, J.-J.; Mei, T.-S.; Yu, J.-Q. Angew.
Chem., Int. Ed. 2008, 47, 6452–6455.
9
14300 J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010

Formation of 1,2-cis-2-Amino Glycosides
A R T I C L E S
7 (55.6 mg, 0.10 mmol, 1 equiv), disaccharide acceptor 43 (90.1
mg, 0.13 mmol, 1.3 equiv), and CH2Cl2 (0.7 mL). The resulting
solution was cooled to 0 °C, and a preformed solution of Ni(4-F-
PhCN)4(OTf)2, which was generated in situ from a reaction of Ni(4-
F-PhCN)4Cl2 (3.07 mg, 0.005 mmol, 5 mol %) and AgOTf (2.57
mg, 0.01 mmol, 10 mol %) in dichloromethane (0.5 mL) for 30
min, was then added to the solution. The resulting mixture was
stirred at 0 °C for 5 min and then warmed to 25 °C. The reaction
mixture was stirred at the ambient temperature for 3 h, diluted with
benzene (2 mL), and purified by silica gel flash chromatography
(10/1, toluene/acetonitrile + 1% triethylamine) to give the desired
trisaccharide 47 (77 mg, 70%, R:ꢀ ) 20:1). 1H NMR (CDCl3, 400
MHz): δ 8.26 (s, 1H), 7.78 (dd, J ) 8.6, 5.5 Hz, 2H), 7.41-7.27
(m, 15H), 6.85 (t, J ) 8.6 Hz, 2H), 5.74 (t, J ) 9.8 Hz, 1H), 5.55
(d, J ) 5.0 Hz, 1H), 5.28 (d, J ) 3.5 Hz, 1H), 5.13 (t, J ) 9.8 Hz,
1H), 4.99 (d, J ) 11.6 Hz, 1H), 4.88 (d, J ) 11.2 Hz, 1H), 4.83
(d, J ) 10.6 Hz, 1H), 4.76 (d, J ) 10.7 Hz, 1H), 4.70 (d, J ) 10.9
Hz, 1H), 4.60 (dd, J ) 7.9, 2.3 Hz, 1H), 4.38-4.31 (m, 3H),
4.26-4.19 (m, 3H), 4.13 (dd, J ) 12.2, 2.0 Hz, 1H), 4.03 (t, J )
9.5 Hz, 1H), 3.98-3.94 (m, 4H), 3.65-3.59 (m, 2H), 3.54 (t, J )
9.2 Hz, 1H), 3.39-3.36 (m, 1H), 2.81 (t, J ) 8.1 Hz, 1H), 2.13 (s,
3H), 2.06 (s, 3H), 1.89 (s, 3H), 1.49 (s, 3H), 1.46 (s, 3H), 1.34 (s,
3H), 1.33 (s, 3H). 13C NMR (CDCl3, 100 MHz): δ 171.0, 170.3,
170.0, 163.1, 139.1, 138.9, 138.8, 131.0, 130.9, 128.7, 128.6, 128.5,
128.4, 128.3, 128.2, 127.9, 127.8, 127.7, 116.2, 116.0, 109.6, 108.7,
104.0, 99.7, 96.5, 84.4, 82.1, 75.9, 75.8, 75.2, 74.7, 72.2, 71.5, 71.4,
70.9, 70.7, 69.1, 68.8, 67.9, 67.1, 65.4, 62.5, 26.3, 26.2, 25.2, 24.7,
21.1, 21.0, 20.9. IR (film, cm-1): ν 2983, 2935, 1748, 1643, 1603,
1509, 1455, 1364, 1231, 1149, 1067, 1022. J(13CH) ) 173.3 Hz.
HRMS (ESI): calcd for C59H69NO18F3 (M + H) 1136.4467, found
1136.4467.
donor 8 (90.87 mg, 0.15 mmol, 1 equiv), D-glucuronic acid methyl
ester 58 (78.5 mg, 0.195 mmol, 1.3 equiv), and CH2Cl2 (1 mL).
The resulting solution was cooled to 0 °C, and a preformed solution
of Ni(4-F-PhCN)4(OTf)2, which was generated in situ from a
reaction of Ni(4-F-PhCN)4Cl2 (4.61 mg, 0.0075 mmol, 5 mol %)
and AgOTf (3.85 mg, 0.015 mmol, 10 mol %) in dichloromethane
(1 mL) for 30 min, was then added to the solution. The resulting
mixture was stirred at 0 °C for 5 min and then warmed to 25 °C.
The reaction mixture was stirred at the ambient temperature for
3 h, diluted with benzene (2 mL), and purified by silica gel flash
chromatography (2/1, hexane/ethyl acetate + 1% triethylamine) to
give the desired disaccharide 64 (110.8 mg, 87%, R only). 1H NMR
(CDCl3, 500 MHz): δ 7.96 (s, 1H), 7.63 (d, J ) 8.5 Hz, 2H), 7.56
(d, J ) 8.0 Hz, 2H), 7.24-7.10 (m, 8H), 6.75 (d, J ) 7.0 Hz, 2H),
5.64 (d, J ) 3.5 Hz, 1H), 5.53 (t, J ) 10.0 Hz, 1H), 5.03 (t, J )
10.0 Hz, 1H), 4.82 (d, J ) 12.0 Hz, 1H), 4.61 (d, J ) 12.0 Hz,
1H), 4.56 (d, J ) 3.0 Hz, 1H), 4.48 (d, J ) 12.0 Hz, 1H), 4.31 (d,
J ) 11.5 Hz, 1H), 4.30 (d, J ) 10.0 Hz, 2H), 4.11-4.08 (m, 2H),
3.98 (t, J ) 9.0 Hz, 1H), 3.85 (d, J ) 10.0 Hz, 1H), 3.78 (s, 3H),
3.53 (dd, J ) 9.5, 3.0 Hz, 1H), 3.46 (dd, J ) 10.0, 3.0 Hz, 1H),
3.41 (s, 3H), 2.10 (s, 3H), 1.99 (s, 3H), 1.77 (s, 3H). 13C NMR
(CDCl3, 125 MHz): δ 170.8, 170.0, 169.8, 169.5, 162.9, 138.4,
138.1, 137.4, 128.6, 128.4, 128.2, 128.1, 128.0, 126.7, 125.6, 125.5,
98.8, 98.4, 81.2, 79.4, 75.4, 74.3, 73.5, 72.5, 70.5, 69.6, 68.3, 68.1,
61.6, 55.7, 52.5, 20.8, 20.6, 20.5. IR (film, cm-1): ν 2942, 1749,
1645, 1453, 1438, 1368, 1323, 1231, 1164, 1128, 1047, 1028.
J(13CH) ) 172.7 Hz. HRMS (ESI): calcd for C42H46F3NO14 (M +
H) 846.2943, found 846.2905.
Preparation of GPI Anchor Pseudodisaccharide 68. A 10 mL
oven-dried Schlenk flask was charged with D-glucosamine trichlo-
roacetimidate donor 7 (37.8 mg, 0.068 mmol, 1 equiv), inositol
acceptor 65 (40 mg, 0.082 mmol, 1.2 equiv), and CH2Cl2 (0.7 mL).
The resulting solution was cooled to 0 °C, and a preformed solution
of Ni(4-F-PhCN)4(OTf)2, which was generated in situ from a
reaction of Ni(4-F-PhCN)4Cl2 (2.09 mg, 0.0034 mmol, 5 mol %)
and AgOTf (1.75 mg, 0.0068 mmol, 10 mol %) in dichloromethane
(0.3 mL) for 30 min, was then added to the solution. The resulting
mixture was stirred at 0 °C for 5 min and then warmed to 25 °C.
The reaction mixture was stirred at the ambient temperature for
4 h, diluted with benzene (2 mL), and purified by silica gel flash
chromatography (2/1, hexane/ethyl acetate + 1% triethylamine) to
give the desired pseudodisaccharide 68 (42.2 mg, 70%, R only).
1H NMR (CDCl3, 400 MHz): δ 8.27 (s, 1H), 7.74 (dd, J ) 8.6, 5.5
Hz, 2H), 7.40-7.27 (m, 15H), 7.11 (t, J ) 11.2 Hz, 2H), 5.69 (t,
J ) 9.8 Hz, 1H), 5.49 (d, J ) 3.5 Hz, 1H), 5.07 (t, J ) 10.1 Hz,
1H), 4.95 (d, J ) 11.1 Hz, 1H), 4.87-4.73 (m, 5H), 4.38 (ddd, J
) 10.3, 5.3, 3.0 Hz, 1H), 4.18 (t, J ) 5.0 Hz, 1H), 4.14-4.05 (m,
2H), 3.97-3.90 (m, 2H), 3.81 (dd, J ) 12.4, 1.8 Hz, 1H), 3.72
(dd, J ) 8.1, 3.6 Hz, 1H), 3.59 (dd, J ) 10.2, 3.5 Hz, 1H), 3.51 (t,
J ) 8.6 Hz, 1H), 2.05 (s, 3H), 1.91 (s, 3H), 1.89 (s, 3H), 1.52 (s,
3H), 1.18 (s, 3H). 13C NMR (CDCl3, 100 MHz): δ 170.9, 170.2,
170.0, 163.1, 138.6, 138.5, 138.3, 132.4, 132.3, 130.7, 130.6, 128.7,
128.5, 128.3, 128.2, 128.1, 128.0, 127.8, 127.6, 116.0, 115.8, 110.1,
97.4, 81.2, 80.5, 79.4, 78.4, 75.2, 74.8, 73.6, 71.9, 71.6, 68.7, 67.6,
62.0, 27.9, 25.8, 21.1, 20.1, 20.9. IR (film, cm-1): ν 2936, 2870,
1745, 1644, 1601, 1509, 1454, 1366, 1222, 1152, 1125, 1025.
J(13CH) ) 175.0 Hz. HRMS (ESI): calcd for C49H55NO13F (M +
H) 884.3669, found 884.3657.
Preparation of Tetrasaccharide 55. A 10 mL oven-dried
Schlenk flask was charged with disaccharide trichloroacetimidate
donor 50 (42 mg, 0.05 mmol, 1 equiv), disaccharide acceptor 43
(42 mg, 0.0.6 mmol, 1.2 equiv), and CH2Cl2 (0.45 mL). The
resulting solution was cooled to 0 °C, and a preformed solution of
Ni(4-F-PhCN)4(OTf)2, which was generated in situ from a reaction
of Ni(4-F-PhCN)4Cl2 (3.07 mg, 0.005 mmol, 10 mol %) and AgOTf
(2.57 mg, 0.01 mmol, 20 mol %) in dichloromethane (0.1 mL) for
30 min, was then added to the solution. The resulting mixture was
stirred at 0 °C for 5 min and then warmed to 25 °C. The reaction
mixture was stirred at the ambient temperature for 7 h, diluted with
benzene (2 mL), and purified by silica gel flash chromatography
(9/1, toluene/acetonitrile + 1% triethylamine) to give the desired
1
tetrasaccharide 55 (42 mg, 72%, R:ꢀ ) 11:1). H NMR (CDCl3,
400 MHz): δ 8.24 (s, 1H), 7.77 (d, J ) 8.4 Hz, 1H), 7.76 (d, J )
8.4 Hz, 1H), 7.39-7.10 (m, 15H), 6.82 (d, J ) 8.4 Hz, 1H), 6.80
(d, J ) 8.4 Hz, 1H), 5.70 (t, J ) 10.0 Hz, 1H), 5.53 (d, J ) 5.2
Hz, 1H), 5.27-5.20 (m, 2H), 5.11-4.95 (m, 5H), 4.87 (d, J )
10.8 Hz, 1H), 4.79 (d, J ) 10.8 Hz, 1H), 4.73 (d, J ) 10.8 Hz,
1H), 4.66 (d, J ) 10.8 Hz, 1H), 4.57-4.52 (m, 2H), 4.31-4.24
(m, 3H), 4.20-4.10 (m, 4H), 4.08-4.02 (m, 2H), 3.97-3.89 (m,
4H), 3.72-3.67 (m, 1H), 3.60-3.50 (m, 3H), 3.33 (d, J ) 9.6 Hz,
1H), 2.75 (t, J ) 8.8 Hz, 1H), 2.10 (s, 3H), 2.09 (s, 3H), 2.04 (s,
3H), 2.03 (s, 3H), 2.01 (s, 3H), 1.85 (s, 3H), 1.48 (s, 3H), 1.44 (s,
3 H), 1.32 (s, 3 H), 1.31 (s, 3 H). 13C NMR (CDCl3, 100 MHz): δ
170.7, 170.3, 169.9, 169.8, 169.6, 169.4, 162.8, 138.9, 138.63,
138.62, 130.7 (d, JCF ) 8.2 Hz), 128.4, 128.3, 128.24, 128.20,
128.17, 128.03, 128.00, 127.7, 127.62, 127.58, 127.5, 127.4, 115.8
(d, JCF ) 21.5 Hz), 109.3, 108.5, 103.7, 100.9, 99.4, 96.3, 84.2,
81.9, 75.7, 75.5, 74.9, 74.4, 72.7, 71.8, 71.7, 71.4, 71.1, 70.9, 70.6,
70.4, 69.0, 68.5, 68.3, 67.8, 66.8, 64.9, 61.8, 26.03, 25.96, 24.9,
24.4, 20.79, 20.76, 20.69, 20.67, 20.6. IR (film, cm-1): ν 2932,
1752, 1644, 1600, 1508, 1455, 1368, 1215, 1067, 1035. J(13CH)
) 179 Hz (103.7 Hz), 159 (100.9 Hz), 176 (99.4 Hz), 165 (96.3
Hz). HRMS (ESI): calcd for C70H85NO26F (M + H) 1374.5363,
found 1374.5344.
Preparation of Glycopeptide 78. A 10 mL oven-dried Schlenk
flask was charged with D-galactosamine trichloroacetimidate donor
70 (90.87 mg, 0.150 mmol, 1 equiv), threonine amino acid 72 (67
mg, 0.195 mmol, 1.2 equiv), and CH2Cl2 (1.1 mL). The resulting
solution was cooled to 0 °C, and a preformed solution of Ni(4-F-
PhCN)4(OTf)2, which was generated in situ from a reaction of Ni(4-
F-PhCN)4Cl2 (9.21 mg, 0.015 mmol, 10 mol %) and AgOTf (7.71
mg, 0.030 mmol, 20 mol %) in dichloromethane (1 mL) for 30
min, was then added to the solution. The resulting mixture was
stirred at 0 °C for 5 min and then warmed to 25 °C. The reaction
mixture was stirred at the ambient temperature for 4 h, diluted with
Preparation of Heparin Disaccharide 64. A 10 mL oven-dried
Schlenk flask was charged with D-glucosamine trichloroacetimidate
9
J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010 14301

A R T I C L E S
Mensah et al.
benzene (2 mL), and purified by silica gel flash chromatography
(9/1, benzene/acetonitrile + 1% triethylamine) to give the desired
glycopeptide 78 (95 mg, 80%, R:ꢀ ) 15:1). 1H NMR (CDCl3, 500
MHz): δ 8.18 (s, 1H), 7.76 (d, J ) 8.5 Hz, 2H), 7.36-7.20 (m,
12H), 7.07 (d, J ) 6.5 Hz, 2H), 6.11 (d, J ) 9.0 Hz, 1H), 5.45 (bs,
1H), 5.38 (dd, J ) 10.5, 2.5 Hz, 1H), 5.18 (d, J ) 12.0 Hz, 1H),
5.09 (d, J ) 12.0 Hz, 1H), 4.91 (d, J ) 12.5 Hz 1H), 4.81-4.76
(m, 2H), 4.46 (d, J ) 6.0 Hz, 1H), 4.39-4.33 (m, 2H), 4.11-4.10
(m, 2H), 3.62 (dd, J ) 11.0, 3.5 Hz, 1H), 2.13 (s, 3H), 2.03 (s,
3H), 1.81 (s, 3H), 1.35 (d, J ) 6 Hz, 3H). 13C NMR (CDCl3, 125
MHz): δ 170.6, 170.1, 169.8, 164.3, 164.1, 156.9, 138.0, 136.1,
134.8, 128.8, 128.7, 128.5, 128.4, 128.34, 128.3, 128.1, 127.9,
125.9, 125.5, 99.6, 74.7, 68.4, 67.9, 67.2, 67.1, 66.9, 62.2, 58.6,
20.8, 20.5, 19.4. IR (film, cm-1): ν 3353, 2938, 1745, 1645, 1372,
1327, 1235, 1168, 1132. J(13CH) ) 170.34 Hz. HRMS (ESI): calcd
for C39H41FN2O13 (M + H) 787.2684, found 787.2709.
Acknowledgment. We thank both University of Iowa and
Montana State University for financial support. We also thank
Joseph M. Azzarelli for preparing disaccharide 32. A portion of
this work was performed at Montana State University.
Supporting Information Available: Experimental procedures
and 1H NMR and 13C NMR spectra of all new compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
JA106682M
9
14302 J. AM. CHEM. SOC. VOL. 132, NO. 40, 2010