An expeditious route to N-Acetyl-D-galactosamine from N-Acetyl-D- glucosamine based on the selective protection of hydroxy groups

Article abstract of DOI:10.1246/bcsj.77.1181

GalNAc (1) was straightforwardly prepared from GlcNAc (2a) in six steps. Selective protection of the hydroxy groups on the C-1, C-3, and C-6 positions at the same time was performed by the treatment of TBDPS chloride (5.5 eq.) in DMF in 70% yield. The nuc

Full text of DOI:10.1246/bcsj.77.1181

ꢀ 2004 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 77, 1181–1186 (2004) 1181
An Expeditious Route to N-Acetyl-D-galactosamine from N-Acetyl-
D-glucosamine Based on the Selective Protection of Hydroxy Groups
ꢀ
Naoyuki Ito, Yoshihide Tokuda, Shigeru Ohba, and Takeshi Sugai
Department of Chemistry, Keio University, Hiyoshi, Yokohama 223-8522
Received November 10, 2003; E-mail: sugai@chem.keio.ac.jp
GalNAc (1) was straightforwardly prepared from GlcNAc (2a) in six steps. Selective protection of the hydroxy
groups on the C-1, C-3, and C-6 positions at the same time was performed by the treatment of TBDPS chloride (5.5
eq.) in DMF in 70% yield. The nucleophilic attack with CsOAc or KOBz on the chloromethylsulfonyloxy group at C-
4 worked well (69–77% yield), accompanied by an unexpected rearrangement, to give the furanose products (6). The de-
protection of all silyl and acyl groups under acidic conditions and the re-acetylation provided GalNAc (1) in 51% yield.
N-Acetyl-D-galactosamine (GalNAc, 1, Fig. 1) is a compo-
nent of mucin-type glycoprotein and of a blood type-determin-
ing oligosaccharide that plays important roles in physiological
events. There have been many longstanding demands from syn-
thetic carbohydrate chemistry and biochemistry researchers, for
its supply in large quantity. Natural resources, however, are
rather limited, the only main source being chondroitin sulfates1
from the cartilage of shark fins. In turn, chemical synthesis
from a protected form of galactal has been developed via azido-
nitration, sulfamidoglycosylation and related methods.2 Al-
though in its biosynthesis, the epimerization at the 4-position
occurs at the step of UDP-N-acetyl-D-glucosamine, the chemi-
cal synthesis of GalNAc from GlcNAc incurs considerable dif-
ficulties. Classical approaches3 have suffered from the multi-
step protection to provide the precursors whose hydroxy groups
are liberated at the proper positions. Crout and co-workers re-
ported an enzyme-catalyzed regioselective hydrolysis of pen-
taacetates toward a short-step synthesis;4 however, the avail-
ability of that specific hydrolase is limited. Here we describe
a short and direct approach by means of a one-step selective in-
troduction of the three protective groups, liberating only one
hydroxy group at the C-4 position, as the key-step.
anomeric mixture with one free hydroxy group at the C-3 and
C-4 positions and were not separable on chromatography. In
contrast, the increased steric hindrance of t-butyldiphenylsilyl
(TBDPS)5 over that of TBDMS was surprisingly effective.
The treatment with TBDPS chloride in DMF (room tempera-
ture, 47 h) provided two products. These were tri-TBDPS
ethers with the desired C-4 free hydroxy group (4a, Fig. 2,
70%) as the major product together with the regioisomeric C-
3 free hydroxy form (3a, 24%), obtained in an easily separable
fashion (ꢀRf /SiO2 = 0.33, hexane/EtOAc = 2/1) over con-
ventional silica-gel column chromatography. The position of
free hydroxy groups were confirmed after acetylation judged
from the downfield-shifted proton signals [3b: ꢀ ¼ 4:73
(J2;3 ¼ 10:3 Hz, J3;4 ¼ 7:6 Hz, H3); 4b: 5.03 (J3;4 ¼ 9:3 Hz,
J4;5 ¼ 9:3 Hz, H4)].
We became interested in the reason of selective formation of
3a and 4a when the reaction was quenched at the initial stage,
only 1 h after the commencement. The monoprotected form 2c
(Fig. 1) was isolated only in a low (11%) yield. In turn, the ac-
Our first attempt at the selective triphenylmethylation (trityl-
ation) of the less hindered hydroxy groups of GlcNAc (2a,
Fig. 1) by way of 2b only resulted in a complex mixture of par-
tially protected derivatives. Treatment with t-butyldimethyl-
silyl (TBDMS) chloride gave products on which three TBDMS
groups were introduced; however, those products were an
Fig. 1.
Fig. 2.
Published on the web June 9, 2004; DOI 10.1246/bcsj.77.1181

1182 Bull. Chem. Soc. Jpn., 77, No. 6 (2004)
Synthesis of N-Acetyl-D-galactosamine
cumulation of the bis-silylated intermediate (5a, Fig. 2, 53%)
was observed. In this case, the isolated yields of 4a and 3a were
11 and 2%, respectively. The structure of 5a was confirmed af-
1
ter the derivation to 5b by H NMR [ꢀ ¼ 4:89 (J2;3 ¼ 9:7 Hz,
J3;4 ¼ 9:8 Hz, H3) and 5.22 (J4;5 ¼ 9:6 Hz, H4)]. The preferen-
tial reaction at the sterically less hindered ꢁ-anomer at the C-1
position of 2c occurred, contrary to the oxygen anomeric effect.
As the major product in the initial stage was 5a, there was no
large difference between the rates of the silylation on the pri-
mary hydroxy group at C-6 and on the secondary hydroxy
group at C-1.
The reason for the predominant silylation of the C-3 position
(4a) was further examined. When the isolated 3a and 4a were
independently treated under strongly basic conditions6 in aq.
NaOH/EtOH, starting either from 3a or 4a, the product became
the same equilibrated mixture of 3a and 4a in a 3:2 ratio. In
contrast, under the reaction conditions of the silylation itself
(TBDPSCl/imidazole/DMF), there was negligible migration
of the protective group, although such a phenomenon had pre-
viously been reported.5b We concluded that the predominant
formation of 4a occurred due to the steric hindrance of the C-
6 TBDPSoxy group at the stage of 1,6-di-TBDPS ether, al-
though the isomer 3a is more thermodynamically stable. The
reaction proceeded in a kinetic manner, as judged from the
comparison of the ratio between 3a and 4a, 1:5.5 (1 h) and
1:2.9 (47 h), respectively. Alkaline treatment (aq. NaOH/
EtOH) of 3a enabled the recycling of the undesired isomer to
the further production of 4a (39%) with the recovery of 3a
(59%), in a preparative scale.
Fig. 3.
1
Based on its H NMR spectrum, the abnormal vicinal cou-
pling constants (J1;2 ¼ J2;3 ¼ 0 Hz) did not indicate the typical
4C1 pyranose chair conformation. First, we suspected that the
product occupied the boat or some unknown conformation of
the pyranose, but the downfield-shifted signal of H-5
(ꢀ ¼ 4:69) was inconsistent with the proposed structures. Final-
ly, the X-ray crystallographic analysis of the recrystallized
sample clarified the unexpected furanose structure of 6a, as
shown in Fig. 3. The identity between the crystalline sample
(recovery: 69%) and the mother liquor was confirmed.
The next and the most important task was the inversion of the
configuration at C-4. Around it, there is a severe steric hin-
drance of silyl groups but no apparent neighboring group par-
ticipation. Preliminary studies on the well-known three candi-
dates revealed that the reactivity of the mesylate (4c) was very
low. At elevated temperature, there was further decomposition
of the product. The triflate (4e), expected to show the highest
reactivity and on whose structurally similar substrates so far
the substitution reaction had been successful,4,7 however, was
too unstable for this attempted conversion, even at the stage
of formation and purification. In contrast, the reaction of chlo-
romethanesulfonate, which was developed by Shimizu and co-
workers,8 worked well to give the desired product. We then ela-
borated the reaction conditions for the efficient introduction of
the chloromethylsulfonyl group. As the initial attempts on the
use of pyridine as the base only resulted in the recovery of un-
reacted starting material, we realized the importance of the ef-
ficient generation of sulfene,9 a highly active species. Indeed,
under the conditions of the use of chloromethanesulfonyl chlo-
ride (McCl, 2 eq.) and triethylamine (Et3N, 6 eq.) at room tem-
perature, the reaction proceeded smoothly to give the desired
product in 72% yield, together with recovery of the starting ma-
terial (28%). Any change in the reaction temperature, the sol-
vent or the ratio of McCl/Et3N resulted only in an increase
of the unreacted starting material. The chloromethanesulfonate
(4d) was then treated with CsOAc (3 eq.) with 18-crown-6 in
toluene.7b The addition of the crown ether ensured the homoge-
neity of the reaction mixture to facilitate the inversion reaction.
The highest yield (69%) of the resulting acetate (6a) was re-
corded at reflux temperature.
A plausible mechanism of the rearrangement is shown in
Scheme 1. The inverting displacement of the chloromethylsul-
fonyloxy group of 4c with acetate is instantly followed by the
neighboring group participation of a ring oxygen atom (path
a), assisted by the electron-donating property of the 1-siloxy
substituent, to provide a stabilized oxonium intermediate. The
all trans-relationship of the substituents, which avoids the steric
repulsion, can explain the formation of the furanose structure of
6a, as had so far been experienced in the preferential furanose
formation of a galactosaminide.10 Neither of the reactions pro-
ceeded at low temperature, and we could not detect any trace of
the proposed intermediate. The possibility of direct attack of
the pyranose ring oxygen atom (path b) into a carbenium ion
species is negligible, due to the well-defined configuration at
C-5 with perfect retention of the stereochemistry without any
trace of the isomer 6b. The change of the nucleophile to the
benzoate anion,3a,11 could not prevent the rearrangement to
give 6c, although the low migratory nature of benzoate ester
had so far been reported.12 The higher yield (77%) and the easi-
er handling of the reagent were beneficial effects of the non-hy-
groscopic property of KOBz, instead of CsOAc.
Due to the rather strange furanose structure in 6a, the depro-
tection of the silyl protective group as well as the acyl group
was somewhat troublesome. Initial attempts at the deprotection
of 6a by either TBAF (tetrabutylammonium fluoride) or HF/
pyridine treatment, expecting the simultaneous deacylation un-
der basic conditions, resulted only in a complex mixture. This
situation was effectively solved by a stepwise transformation
and purification procedure, including hydrolysis under a strong-

N. Ito et al.
Bull. Chem. Soc. Jpn., 77, No. 6 (2004) 1183
Scheme 1.
ly acidic condition (6 M HCl) to give the fully deprotected gal-
actosamine hydrochloride, and the subsequent acetylation, to
and the variety of the regioselectively protected pure intermedi-
ates (3–5) are attractive.
ꢁ
give 713 (Fig. 2) in 61% yield; mp 170–173 C (lit.4 169–171
ꢁC); ½ꢂꢂD þ96:3ꢁ (CHCl3) [lit.4 ½ꢂꢂD þ97ꢁ (CHCl3)]. Finally,
alkaline hydrolysis of the O-acetyl protective groups in 7 was
successful and the pure GalNAc (1)1c,4 was obtained in 84%
yield; mp 154.5–155.5 ꢁC (lit.4 156–158 ꢁC); ½ꢂꢂD þ86:1ꢁ
(H2O, equilibrated for 3 h) [lit.1c ½ꢂꢂD þ86:1ꢁ (H2O, equilibrat-
ed)]. When the same procedure was attempted to deprotect the
benzoate 6c, however, the reaction resulted in a complex mix-
ture.
In conclusion, GalNAc (1) was straightforwardly prepared
from GlcNAc (2) in six steps, based on the selective protection
with the TBDPS groups and the nucleophilic inversion at C-4
on the chloromethylsulfonyloxy group to provide a furanose
key intermediate (6), by a rearrangement of the pyranose pre-
cursor (4c). The five steps to the pentaacetate (7) in the present
work were more than three steps, demonstrated in the well-
known azidonitration approach2a from tri-O-acetyl-D-galactal,
which is available from D-galactose in three steps.14 However,
the inexpensiveness of the starting material (GlcNAc), an in-
creased total yield (18% from GlcNAc/11% from galactose),
Experimental
Analytical and preparative thin-layer chromatography (TLC)
procedures were developed on E. Merck Silica-Gel 60 F256 plates
(No. 5715; 0.25 mm and 5744; 0.50 mm), respectively. Column
chromatography was performed on Kanto Chemical Co., Inc. Sili-
ca-Gel 60 (spherical; 100–210 mm, 37558-79). NMR spectra were
measured on a JEOL EX-270, GX-400 spectrometer (1H at 270 and
400 MHz, and 13C at 100 MHz). 1H chemical shifts are referenced
with CHCl3 at 7.26 ppm or HDO at 4.80 ppm and 13C chemical
shifts with CDCl3 at 77.0 ppm. IR spectra were determined on a
Jasco FT/IR-410 spectrometer. Optical rotations were measured
on a Jasco DIP 360 polarimeter. Melting points were measured
by Yanaco MP-S3.
2-Acetamido-1,4,6-tris-O-t-butyldiphenylsilyl-2-deoxy-ꢀ-D-
glucopyranose (3a) and 2-Acetamido-1,3,6-tris-O-t-butyldi-
phenylsilyl-2-deoxy-ꢀ-D-glucopyranose (4a). GlcNAc 2 (200.7
mg, 0.91 mmol) was dissolved in DMF (2.2 mL) under an argon
atmosphere. Imidazole (307.0 mg, 4.96 mmol) and t-butyldiphen-
ylsilyl chloride (1.16 mL, 4.97 mmol, 5.5 eq.) were added to the

1184 Bull. Chem. Soc. Jpn., 77, No. 6 (2004)
Synthesis of N-Acetyl-D-galactosamine
solution with stirring. The reaction was monitored by silica-gel
TLC (hexane–ethyl acetate, 2:1). After stirring for 47 h at room
temperature, the mixture was quenched by the addition of water,
and the aqueous layer was extracted with a mixture of hexane–eth-
yl acetate (1:1). The organic layers were combined, washed with a
saturated aqueous NaHCO3 solution, water and brine, dried over
Na2SO4, and concentrated in vacuo. The residue was purified by
silica-gel column chromatography (35 g). Elution with hexane–
ethyl acetate (5:1) afforded 3a (207.9 mg, 24%) and 4a (596.5
mg, 70%). 3a: Rf ¼ 0:18 (hexane–ethyl acetate, 2:1); mp 217–
2-deoxy-ꢀ-D-glucopyranose (5b). 1H NMR (CDCl3, 400 MHz) ꢀ
1.03 (9H, s, t-Bu), 1.09 (9H, s, t-Bu), 1.80 (3H, s, Ac), 1.82 (3H, s,
Ac), 2.00 (3H, s, Ac), 3.17 (1H, ddd, J4;5 ¼ 9:6 Hz, J5,6a ¼ 2:8 Hz,
J5,6b ¼ 2:8 Hz, H5), 3.55 (2H, m, H6a, H6b), 4.24 (1H, ddd, J1;2
¼
8:3 Hz, J2,NH ¼ 9:8 Hz, J2;3 ¼ 9:7 Hz, H2), 4.50 (1H, d, H1), 4.89
(1H, dd, J3;4 ¼ 9:8 Hz, H3), 5.07 (1H, d, NH), 5.22 (1H, dd, H4),
7.29–7.70 (20H, m, Ph).
2-Acetamido-1,3,6-tris-O-t-butyldiphenylsilyl-2-deoxy-4-
methylsulfonyl-ꢀ-D-glucopyranose (4c). The starting material
4a (50.0 mg, 0.05 mmol) was dissolved in diethyl ether (0.3 mL)
under an argon atmosphere. Methanesulfonyl chloride (16.4 mL,
0.21 mmol, 4.2 eq.) and triethylamine (29.8 mL, 0.21 mmol, 4.2
eq.) were added to the solution with stirring at room temperature.
The reaction was monitored by silica-gel TLC, developed with
hexane–ethyl acetate (2:1). After stirring for 4 h at room tempera-
ture, the mixture was quenched by the addition of water, and the
aqueous layer was extracted with ethyl acetate. The organic layers
were combined, washed with water and brine, dried over Na2SO4,
and concentrated in vacuo. The residue was purified by silica-gel
column chromatography. Elution with hexane–ethyl acetate (8:1)
20
219 ꢁC; ½ꢂꢂD þ9:2ꢁ (c 1.00, CHCl3); IR 3579 (OH), 3288
(NH), 1651 cmꢃ1 (C=O); 1H NMR (CDCl3, 400 MHz) ꢀ 0.83
(9H, s, t-Bu), 0.97 (9H, s, t-Bu), 1.07 (9H, s, t-Bu), 1.63 (3H, s,
Ac), 3.34–3.62 (6H, m, H2-6), 3.85 (1H, d, J4,OH ¼ 10:7 Hz,
OH), 4.71 (1H, d, J1;2 ¼ 7:3 Hz, H1), 5.09 (1H, d, J2,NH ¼ 6:9
Hz, NH), 7.16–7.70 (30H, m, Ph); 13C NMR (CDCl3, 100 MHz)
ꢀ 19.2, 19.6, 23.3, 26.8, 26.9, 27.1, 58.5, 64.2, 73.5, 75.7, 77.8,
95.3, 127.5, 127.7, 127.8, 129.3, 129.4, 129.5, 129.9, 132.5,
132.8, 133.1, 133.3, 133.4, 133.5, 135.6, 135.7, 135.8, 170.9.
Found: C, 71.66; H, 7.39; N, 1.45%. Calcd for C56H69NO6Si3:
C, 71.83; H, 7.43; N, 1.50%. The structure was further confirmed
by the derivatization to the corresponding acetate 3b. 1H NMR
(CDCl3, 270 MHz) ꢀ 0.69 (9H, s, t-Bu), 0.98 (9H, s, t-Bu), 1.05
(9H, s, t-Bu), 1.13 (3H, s, Ac), 1.56 (3H, s, Ac), 3.43 (1H, ddd,
J4;5 ¼ 7:8 Hz, J5,6a ¼ 7:8 Hz, J5,6b ¼ 2:1 Hz, H5), 3.71 (2H, m,
H4, H6a), 3.98 (2H, m, H2, H6b), 4.54 (1H, d, J1;2 ¼ 8:1 Hz,
H1), 4.73 (1H, dd, J2;3 ¼ 10:3 Hz, J3;4 ¼ 7:6 Hz, H3), 4.99 (1H,
d, J2,NH ¼ 9:9 Hz, NH), 7.20–7.43 (20H, m, Ph), 7.52–7.68
(10H, m, Ph). 4a: Rf ¼ 0:51 (hexane–ethyl acetate, 2:1); mp
1
afforded 4c (46.4 mg, 86% yield). H NMR (CDCl3, 270 MHz) ꢀ
0.98 (9H, s, t-Bu), 1.02 (9H, s, t-Bu), 1.08 (9H, s, t-Bu), 1.11
(3H, s, Ac), 2.56 (3H, s, Me), 3.40 (1H, ddd, J4;5 ¼ 6:8 Hz, J5,6a
4:5 Hz, J5,6b ¼ 4:4 Hz, H5), 3.60 (1H, ddd, J1;2 ¼ 8:6 Hz, J2,NH
¼
¼
6:6 Hz, J2;3 ¼ 6:7 Hz, H2), 3.81 (1H, dd, J6a,6b ¼ 11:1 Hz, H6a),
3.93 (1H, dd, H6b), 4.55 (1H, dd, J3;4 ¼ 6:6 Hz, H3), 4.69 (1H,
d, H1), 4.88 (1H, d, NH), 4.91 (1H, dd, H4), 7.16–7.83 (30H, m,
Ph).
2-Acetamido-1,3,6-tris-O-t-butyldiphenylsilyl-4-chlorometh-
ylsulfonyl-2-deoxy-ꢀ-D-glucopyranose (4d). In the same man-
ner as described for the methylsulfonylation, 4a (200.0 mg, 0.21
mmol) was treated with chloromethanesulfonyl chloride (38.1
mL, 0.43 mmol, 2.0 eq.) for 3 h. The work-up and the purification
22
73.5–75.5 ꢁC; ½ꢂꢂD ꢃ9:1ꢁ (c 1.00, CHCl3); IR 3571 (OH),
1664 cmꢃ1 (C=O); 1H NMR (CDCl3, 270 MHz) ꢀ 1.02 (9H, s,
t-Bu), 1.03 (9H, s, t-Bu), 1.05 (9H, s, t-Bu), 1.49 (3H, s, Ac),
2.83 (1H, ddd, J4;5 ¼ 9:2 Hz, J5,6a ¼ 3:6 Hz, J5,6b ¼ 3:1 Hz,
H5), 3.51 (1H, dd, J2;3 ¼ 10:1 Hz, J3;4 ¼ 8:7 Hz, H3), 3.61 (1H,
dd, J6a,6b ¼ 10:7 Hz, H6a), 3.68 (1H, dd, H6b), 3.89 (1H, dd,
H4), 4.03 (1H, ddd, J1;2 ¼ 8:1 Hz, J2,NH ¼ 9:4 Hz, H2), 4.36
(1H, d, H1), 4.65 (1H, d, NH), 7.16–7.43 (15H, m, Ph), 7.55–
7.71 (15H, m, Ph); 13C NMR (CDCl3, 100 MHz) ꢀ 19.1, 19.3,
19.7, 23.5, 26.8, 26.9, 58.2, 63.2, 71.8, 74.7, 76.8, 96.1, 127.2,
127.3, 127.6, 127.6, 127.7, 127.9, 129.4, 129.5, 129.7, 129.9,
132.7, 132.9, 133.1, 133.2, 133.3, 133.9, 134.7, 135.4, 135.5,
135.7, 135.8, 135.9, 169.4. Found: C, 71.79; H, 7.43; N, 1.46%.
Calcd for C56H69NO6Si3: C, 71.83; H, 7.43; N, 1.50%. The struc-
ture was further confirmed by the derivatization to the correspond-
20
as above afforded 4d (161.4 mg, 72% yield). ½ꢂꢂD ꢃ1:1ꢁ (c 1.00,
CHCl3); IR 3442 (NH), 1685 cmꢃ1 (C=O); 1H NMR (CDCl3, 400
MHz) ꢀ 0.98 (9H, s, t-Bu), 1.03 (9H, s, t-Bu), 1.10 (9H, s, t-Bu),
1.11 (3H, s, Ac), 3.42 (1H, ddd, J4;5 ¼ 6:3 Hz, J5,6a ¼ 3:9 Hz,
J5,6b ¼ 4:6 Hz, H5), 3.65 (1H, ddd, J1;2 ¼ 8:3 Hz, J2,NH ¼ 6:3
Hz, J2;3 ¼ 6:6 Hz, H2), 3.79 (1H, dd, J6a,6b ¼ 11:1 Hz, H6a),
3.96 (1H, d, JClCH ¼ 12:7 Hz, ClCHa), 4.00 (1H, dd, H6b), 4.14
2
(1H, d, ClCHb), 4.55 (1H, dd, J3;4 ¼ 6:5 Hz, H3), 4.71 (1H, d,
H1), 4.90 (1H, d, NH), 5.02 (1H, dd, H4), 7.16–7.85 (30H, m,
Ph); 13C NMR (CDCl3, 100 MHz) ꢀ 19.1, 19.2, 19.5, 23.0, 26.6,
26.8, 26.9, 53.8, 58.1, 63.0, 72.9, 74.5, 80.7, 94.5, 127.4, 127.6,
127.7, 127.9, 129.6, 129.7, 129.8, 129.9, 130.0, 132.6, 132.6,
132.7, 132.8, 133.0, 134.2, 135.2, 135.5, 135.6, 135.7, 169.1.
Found: C, 65.17; H, 6.76; N, 1.28%. Calcd for C57H70ClNO8SSi3:
C, 65.27; H, 6.73; N, 1.34%.
1
ing acetate 4b. H NMR (CDCl3, 400 MHz) ꢀ 0.91 (9H, s, t-Bu),
0.99 (9H, s, t-Bu), 1.02 (9H, s, t-Bu), 1.15 (3H, s, Ac), 1.47 (3H,
s, Ac), 3.15 (1H, ddd, J4;5 ¼ 9:3 Hz, J5,6a ¼ 2:4 Hz, J5,6b ¼ 5:6
Hz, H5), 3.45 (1H, dd, J6a,6b ¼ 11:6 Hz, H6a), 3.55 (1H, dd,
H6b) 3.71 (1H, ddd, J1;2 ¼ 8:8 Hz, J2,NH ¼ 8:8 Hz, J2;3 ¼ 9:3
Hz, H2), 4.05 (1H, dd, J3;4 ¼ 9:3 Hz, H3), 4.54 (1H, d, H1),
4.66 (1H, d, NH), 5.03 (1H, dd, H4), 7.19–7.72 (30H, m, Ph).
2-Acetamido-1,3,4-tris-O-acetyl-6-O-t-butyldiphenylsilyl-2-
deoxy-ꢁ-D-glucopyranose (2d). 1H NMR (CDCl3, 270 MHz) ꢀ
1.04 (9H, s, t-Bu), 1.92 (3H, s, Ac), 1.95 (3H, s, Ac), 2.06 (3H,
s, Ac), 2.16 (3H, s, Ac), 3.69 (2H, m, H6a, H6b), 3.83 (1H, ddd,
J4;5 ¼ 9:6 Hz, J5,6a ¼ 3:2 Hz, J5,6b ¼ 3:2 Hz, H5), 4.47 (1H,
ddd, J1;2 ¼ 3:5 Hz, J2,NH ¼ 8:9 Hz, J2;3 ¼ 9:7 Hz, H2), 5.21 (1H,
dd, J3;4 ¼ 9:9 Hz, H3), 5.30 (1H, dd, H4), 5.51 (1H, d, NH),
6.23 (1H, d, H1), 7.36–7.66 (10H, m, Ph). The anomeric proton
of the minor (3.8:1) of ꢁ-isomer appeared at ꢀ ¼ 5:65
(J1;2 ¼ 8:7 Hz).
2-Acetamido-5-O-acetyl-1,3,6-tris-O-t-butyldiphenylsilyl-2-
deoxy-ꢀ-D-glucofuranose (6a). The chloromethanesulfonate 4d
(457.5 mg, 0.44 mmol) was dissolved in toluene (6.5 mL) under an
argon atmosphere. 18-Crown-6 (345.8 mg, 1.31 mmol) and cesium
acetate (251.1 mg, 1.31 mmol) were added to the solution with stir-
ring, and the mixture became homogeneous. The reaction was
monitored by TLC, developed with hexane–ethyl acetate (2:1).
The mixture was stirred for 4 h at 120 ꢁC. After cooling, water
was added, and the aqueous layer was extracted with ethyl acetate.
The combined organic layer was washed with water and brine,
dried over Na2SO4, and concentrated in vacuo. The residue was
purified by silica-gel column chromatography. Elution with hex-
ane–ethyl acetate (4:1) afforded 6a (293.0 mg, 69% yield). A small
amount was recrystallized from hexane to give an analytical sam-
2-Acetamido-3,4-di-O-acetyl-1,6-di-O-t-butyldiphenylsilyl-

N. Ito et al.
Bull. Chem. Soc. Jpn., 77, No. 6 (2004) 1185
22
ple of 6a. Mp 119.4–120.0 ꢁC; ½ꢂꢂD ꢃ28:2 (c 1.00, CHCl3); IR
3284 (NH), 1747 (C=O), 1651 cmꢃ1 (C=O); 1H NMR (CDCl3,
270 MHz) ꢀ 0.99 (9H, s, t-Bu), 1.11 (9H, s, t-Bu), 1.12 (9H, s, t-
Bu), 1.25 (3H, s, Ac), 1.79 (3H, s, Ac), 3.46 (1H, dd, J5,6a ¼ 6:4
Hz, J6a,6b ¼ 10:1 Hz, H6a), 3.54 (1H, dd, J5,6b ¼ 6:6 Hz, H6b),
3.71 (1H, d, J3;4 ¼ 3:5 Hz, H3), 4.51 (1H, dd, J4;5 ¼ 3:3 Hz,
H4), 4.57 (1H, d, J2,NH ¼ 9:1 Hz, H2), 4.69 (1H, ddd, H5), 5.12
(1H, s, H1), 5.23 (1H, d, NH), 7.21–7.46 (15H, m, Ph), 7.57–
7.73 (15H, m, Ph); 13C NMR (CDCl3, 100 MHz) ꢀ 19.2, 19.4,
20.8, 23.2, 26.7, 26.9, 27.0, 61.8, 63.5, 72.8, 79.6, 83.4, 102.5,
127.4, 127.5, 127.6, 127.7, 127.8, 129.5, 129.6, 129.7, 129.8,
129.9, 132.7, 132.8, 133.0, 133.1, 134.7, 135.4, 135.5, 135.7,
135.8, 136.2, 168.0, 169.2. Found: C, 71.10; H, 7.24; N, 1.38%.
Calcd for C58H71NO7Si3: C, 71.20; H, 7.31; N, 1.43%.
centrated in vacuo with EtOH. The residue was diluted with water
(10 mL), and desalted by AC-220-10 on Asahi Chemical Micro
Acylyzer S1. At the initial stage, the conductivity was 23 mS
and after the desaltation at 4.3 V (0.69 A), it reached 0.1 mS.
The mixture was concentrated in vacuo and the residue was puri-
fied by preparative TLC. Elution with ethyl acetate–EtOH–water
(63:25:12) afforded 1 (20.7 mg, 84% yield) as an oil. This was sol-
idified upon standing, and an analytical sample was recrystallized
from EtOH to give pure 1. Mp 154.5–155.5 ꢁC (lit.4 156–158 ꢁC);
22
½ꢂꢂD þ86:1ꢁ (c 1.00, H2O, equilibrated for 3 h) [lit.1c ½ꢂꢂD
þ86:1ꢁ (c 1.0, H2O, final)]; IR 3332 (OH), 1641 cmꢃ1 (C=O);
1H NMR (CDCl3, 400 MHz) ꢀ 2.05 (3H, s, Ac), 3.66–3.94 (m),
4.00 (1H, d, J ¼ 2:9 Hz, H4ꢂ), 4.09–4.15 (2H, m, H2ꢂ, H5ꢂ),
4.64 (1H, d, J1;2 ¼ 8:3 Hz, H1ꢁ), 5.23 (1H, d, J1;2 ¼ 3:4 Hz,
H1ꢂ); 13C NMR (CDCl3, 100 MHz) ꢀ 22.7, 23.0, 51.0, 54.4,
61.7, 61.9, 68.1, 68.6, 69.3, 71.2, 71.8, 75.9, 91.7, 96.1, 175.3,
175.5. The NMR spectra were identical with the spectrum of an au-
thentic mixture of the anomers of 2-acetamido-2-deoxy-D-galacto-
pyranose (Wako Pure Chemical Industries, Ltd., 013-12821).
Found: C, 43.14; H, 6.85; N, 6.05%. Calcd for C8H15NO6: C,
43.44; H, 6.83; N, 6.33%.
2-Acetamido-5-O-benzoyl-1,3,6-tris-O-t-butyldiphenylsilyl-
2-deoxy-ꢀ-D-glucofuranose (6c). In the same manner as describ-
ed for 6a, the chloromethanesulfonate 4d (110.9 mg, 0.11 mmol)
was treated with 18-crown-6 (150.7 mg, 0.57 mmol) and potassium
benzoate (84.7 mg, 0.53 mmol) in toluene (1.8 mL) for 18 h at
ꢁ
room temperature to 120 C. The work-up and the purification as
22
above afforded 6c (84.5 mg, 77% yield). ½ꢂꢂD ꢃ32:1 (c 1.00,
CHCl3); IR 3429 (NH), 1728 (C=O), 1685 cmꢃ1 (C=O); 1H NMR
(CDCl3, 400 MHz) ꢀ 0.95 (9H, s, t-Bu), 1.12 (18H, s, t-Bu), 1.23
(3H, s, Ac), 3.59 (1H, dd, J5,6a ¼ 5:9 Hz, J6a,6b ¼ 10:4 Hz, H6a),
3.68 (1H, dd, J5,6b ¼ 7:1 Hz, H6b), 3.74 (1H, d, J ¼ 3:9 Hz,
X-ray Crystal Structure Analysis of 6a. Crystal data: ortho-
rhombic, P212121, a ¼ 22:847ð4Þ, b ¼ 24:910ð3Þ, c ¼ 9:9146ð13Þ
A, V ¼ 5642:5ð13Þ A , Z ¼ 4, Dx ¼ 1:152 g cmꢃ3. A colorless
needle crystal grown from hexane was used. The X-ray intensities
were measured on a Rigaku AFC-7R four-circle diffractometer
with Mo Kꢂ radiation. The positional and anisotropic thermal pa-
rameters of non-H atoms were refined, and all the H atoms were
positioned geometrically (R ¼ 0:047). There is an intermolecular
ꢁ
ꢁ 3
H3), 4.53 (2H, m, H2, H4), 4.92 (1H, d, NH), 5.09 (1H, ddd, J4;5
¼
2:1 Hz, H5), 5.16 (1H, d, H1), 7.19–7.45 (15H, m, Ph), 7.50–7.82
(15H, m, Ph); 13C NMR (CDCl3, 100 MHz) ꢀ 19.2, 19.4, 22.5,
26.7, 26.9, 27.0, 62.5, 63.5, 73.2, 80.1, 84.0, 102.6, 127.4, 127.5,
127.6, 127.8, 128.4, 129.4, 129.5, 129.6, 129.7, 129.8, 132.7,
132.8, 133.0, 133.1, 133.2, 133.3, 135.4, 135.5, 135.7, 135.8,
135.9, 165.1, 168.1. Found: C, 72.72; H, 7.07; N, 1.35%. Calcd
for C63H73NO7Si3: C, 72.47; H, 7.34; N, 1.30%.
N–H O hydrogen bond between the acetamido groups, forming
ꢄꢄꢄ
one-dimensional molecular chains along the c axis. Calculations
were performed using the TEXSAN crystallographic software
package (Version 1.11) of Molecular Structure Corporation.
Crystallographic data have been deposited at the CCDC, 12
Union Road, Cambridge CB2 1EZ, UK and copies can be obtained
on request, free of charge, by quoting the publication citation and
the deposition number CCDC-221140.
2-Acetamido-1,3,4,6-tetra-O-acetyl-2-deoxy-ꢁ-D-galactopy-
ranose (7). A mixture of 6a (122.6 mg, 0.13 mmol) and 6 M HCl
(6 mL) was heated under reflux for 24 h. After cooling to room
temperature, the mixture was azeotropically concentrated in vacuo
with EtOH. The residue was dissolved in pyridine (0.5 mL) and
acetic anhydride (0.5 mL) was added. After stirring for 23 h at
room temperature, the mixture was quenched by the addition of
MeOH. After the mixture was concentrated in vacuo water was
added, and the aqueous layer was extracted with ethyl acetate.
The combined organic solution was dried over Na2SO4, and con-
centrated in vacuo. The residue was purified by silica-gel column
chromatography. Elution with hexane–ethyl acetate (8:1) afforded
7 (29.8 mg, 61% yield). A small amount was solidified upon stand-
ing, and the analytical sample was recrystallized from EtOH to
The authors thank Professors Yoshiaki Nakahara of Tokai
University, Teruo Yoshino of International Christian Universi-
ty, and Jun-ichi Tamura of Tottori University, for their helpful
discussions. This work was accomplished as part of the 21st
Century COE Program (KEIO LCC) from the Ministry of Ed-
ucation, Culture, Sports, Science and Technology. Part of this
work was supported by ‘‘Science and Technology Program
on Molecules, Supra-Molecules and Supra-Structured Materi-
als’’ of an Academic Frontier Promotional Project and Grant-
in-Aid for Scientific Research (No. 14560084).
25
give pure 7. Mp 170–173 ꢁC (lit.4 169–171 ꢁC); ½ꢂꢂD þ96:3ꢁ
(c 1.00, CHCl3) [lit.4 ½ꢂꢂD þ97ꢁ (c 1.0, CHCl3)]; IR 3296 (NH),
1749 cmꢃ1 (C=O); 1H NMR (CDCl3, 400 MHz) ꢀ 1.96 (3H, s,
Ac), 2.04 (6H, s, Ac), 2.18 (6H, s, Ac), 4.06 (1H, dd, J5,6a ¼ 6:2
Hz, J6a,6b ¼ 11:2 Hz, H6a), 4.12 (1H, dd, J5,6b ¼ 6:5 Hz, H6b),
4.24 (1H, dd, H5), 4.74 (1H, ddd, J1;2 ¼ 3:5 Hz, J2,NH ¼ 10:4
Hz, J2;3 ¼ 11:7 Hz, H2), 5.23 (1H, dd, J3;4 ¼ 3:4 Hz, H3), 5.39
(1H, d, NH), 5.43 (1H, d, H4), 6.22 (1H, d, H1); 13C NMR (CDCl3,
100 MHz) ꢀ 20.6, 20.7, 20.9, 23.1, 46.9, 61.2, 66.6, 67.7, 68.4,
91.2, 168.7, 170.0, 170.1, 170.9. Found: C, 49.21; H, 6.09; N,
3.49%. Calcd for C16H23NO10: C, 49.36; H, 5.95; N, 3.60%.
2-Acetamido-2-deoxy-D-galactopyranose (1). A mixture of 7
(43.3 mg, 0.11 mmol), 2 M NaOH (1.2 mL), and EtOH (0.3 mL)
was stirred at room temperature. After stirring for 1 h, the mixture
was quenched by the addition of 1 M HCl and azeotropically con-
References
1
a) P. A. Levene and F. B. La Forge, J. Biol. Chem., 20, 433
(1915). b) C. S. Hudson and J. K. Dale, J. Am. Chem. Soc., 38, 1431
(1916). c) S. Roseman and J. Ludowieg, J. Am. Chem. Soc., 76, 301
(1954). d) S. Suzuki, J. Biol. Chem., 253, 3580 (1960).
2
a) R. U. Lemieux and R. M. Ratcliffe, Can. J. Chem., 57,
1244 (1979). b) D. A. Griffith and S. J. Danishefsky, J. Am. Chem.
Soc., 112, 5811 (1990). c) J. Liu and D. Y. Gin, J. Am. Chem. Soc.,
124, 9789 (2002).
3
a) P. H. Gross, F. Du Bois, and R. W. Jeanloz, Carbohydr.
Res., 4, 244 (1967). b) J. Hill and L. Hough, Carbohydr. Res., 8,
398 (1968). c) Y. Fukuda, H. Sasai, and T. Suami, Bull. Chem.

1186 Bull. Chem. Soc. Jpn., 77, No. 6 (2004)
Synthesis of N-Acetyl-D-galactosamine
Soc. Jpn., 55, 1574 (1982). d) R. P. McGeary, K. Wright, and I.
Toth, J. Org. Chem., 66, 5102 (2001).
Habuchi, 23rd Japanese Carbohydrate Symposium, PI-27, August
22, Yokohama (2002).
4 D. Chaplin, D. H. G. Crout, S. Bornemann, D. W.
Hutchinson, and R. Khan, J. Chem. Soc., Perkin Trans. 1, 1992,
8 T. Shimizu, S. Hiranuma, and T. Nakata, Tetrahedron Lett.,
37, 6145 (1996).
235.
5 a) Y. Ikeda, K. Furukawa, and H. Yamada, Carbohydr.
Res., 337, 1499 (2002). b) M. S. Arias-Perez, M. S. Lopez, and
9
Chem. Soc., Perkin Trans. 2, 1993, 1153.
10 R. T. Lee, T. C. Wong, and Y. C. Lee, J. Carbohydr. Chem.,
5, 343 (1986).
S. N. Lyashchuk, Y. G. Skrypnik, and V. P. Besrodnyi, J.
´
´
M. J. Santos, J. Chem. Soc., Perkin Trans. 2, 2002, 1549.
6 Y. Torisawa, M. Shibasaki, and S. Ikegami, Chem. Pharm.
Bull., 31, 2607 (1983).
11 A. J. H. Houssa, J. Kenyon, and H. Phillips, J. Chem. Soc.,
1929, 1700.
7
a) J. Alais and S. David, Carbohydr. Res., 201, 69 (1990). b)
12 S. Akai, T. Naka, Y. Takebe, and Y. Kita, Chem. Pharm.
Bull., 48, 1519 (2000).
13 a) M. Stacey, J. Chem. Soc., 1944, 272. b) B. Weismann, J.
Org. Chem., 31, 2505 (1966). c) T. C. Wong, R. R. Townsend, and
Y. C. Lee, Carbohydr. Res., 170, 27 (1987). d) M. A. Findeis, Int.
J. Pept. Protein Res., 43, 477 (1994).
K. Sato, A. Yoshimoto, and Y. Takai, Bull. Chem. Soc. Jpn., 70,
885 (1997). c) K. Sato, H. Seki, A. Yoshitomo, H. Nanaumi, Y.
Takai, and Y. Ishido, J. Carbohydr. Chem., 17, 703 (1998). d) I.
Matsuo, M. Isomura, R. Walton, and K. Ajisaka, Tetrahedron Lett.,
37, 8795 (1996). e) I. Matsuo, M. Isomura, and K. Ajisaka, J.
Carbohydr. Chem., 18, 841 (1999). f) A. Furstner and I. Konetzki,
14 a) P. A. Levene and R. S. Tipson, J. Biol. Chem., 93, 631
(1931). b) A. P. Kozikowski and J. Lee, J. Org. Chem., 55, 863
(1990). c) R. Bukowski, L. M. Morris, R. J. Woods, and T.
Weimar, Eur. J. Org. Chem., 2001, 2697.
¨
Tetrahedron Lett., 39, 5721 (1998). g) L. Cipolla, B. L. Ferla, L.
Lay, F. Peri, and F. Nicotra, Tetrahedron: Asymmetry, 11, 295
(2000). h) T. Sawada, S. Fujii, H. Nakano, S. Ohtake, and O.