Multi gram-scale synthesis of galactothionolactam and its transformation into a galactonoamidine

Article abstract of DOI:10.1016/j.carres.2011.02.021

We recently proposed to conduct selective glycosylation reactions after in situ activation of a glycosyl donor promoted by a transition metal complex immobilized in a macromolecular matrix. In order to develop this catalytic entity, a feasible multi gram-

Full text of DOI:10.1016/j.carres.2011.02.021

Carbohydrate Research 346 (2011) 897–904
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Multi gram-scale synthesis of galactothionolactam and its transformation
into a galactonoamidine
⇑
Rami Kanso, Susanne Striegler
Auburn University, Department of Chemistry and Biochemistry, 179 Chemistry Building, Auburn, AL 36849, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
We recently proposed to conduct selective glycosylation reactions after in situ activation of a glyco-
syl donor promoted by a transition metal complex immobilized in a macromolecular matrix. In order
Received 13 December 2010
Received in revised form 18 February 2011
Accepted 24 February 2011
Available online 9 March 2011
to develop this catalytic entity,
-galactothionolactam, its transformation into galactonoamidines with aromatic aglycon, and subse-
quent debenzylation conditions were developed. The potential for epimerization reactions at C-2 of
the glycosidic ring during the transformations from the 2,3,4,6-tetra-O-benzyl- -galactonolactam
into the N-benzyl-2,3,4,6-tetra-O-benzyl- -galactonoamidines via the 2,3,4,6-tetra-O-benzyl-D-
a feasible multi gram-scale synthesis for 2,3,4,6-tetra-O-benzyl-
D
D
Keywords:
Galactonoamidine
Debenzylation
D
galactothionolactam are discussed and additionally characterized by using density functional theory
calculations.
Multi-gram scale synthesis
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
compound encompassing characteristics of the enzymatic galacto-
side hydrolysis (Scheme 1).3–6
A major obstacle to advances in glycobiology is the lack of pure
and structurally well-defined carbohydrates and glycoconjugates.1
Enzymatic approaches toward synthesizing these compounds are
often compromised by low regioselectivity,2 moderate thermal
stability, narrow pH profiles of enzyme activity and/or low yields;2
on the other hand, conventional multi-step syntheses involve
numerous and frequently time-consuming protection/deprotec-
tion steps as well as obligatory activation of glycosyl donors. The
usefulness of these synthetic methods is in addition compromised
by the fact that many glycosylation reactions give mixtures of the
two possible anomers.1 If these anomers are not separated satis-
factorily after each glycosylation step, complex mixtures of prod-
ucts are obtained in insufficient purity that cannot be used for
biological studies.1
To overcome these shortcomings, we proposed a unique ap-
proach for selective glycosylation reactions, which relies on the
in situ activation of a glycosyl donor promoted by a transition metal
complex immobilized in a macromolecular matrix. This catalytic
entity could be prepared by templating a suitable compound
resembling features of the transition state of the glycoside hydroly-
sis in a polymer matrix. To show proof of concept, we are in need of
such compound in gram amounts, and therefore directed initial fo-
cus on synthesizing (3S,4S,5S,6R,Z)-2-(benzylimino)-6-(hydroxy-
methyl) piperidine-3,4,5-triol, N-benzylgalactonoamidine 1, as
Gluco- and mannoamidines have been reported previously as
potent transition state analogs and competitive inhibitors of cor-
responding glycosidases.7–12 Comparable entities with the galac-
to-configuration are not described yet. However, the galacto-
configuration is preferable over a gluco-configuration for its in-
tended use as template due to the structural constraints imposed
by the axial hydroxyl group at C-4. While the benzyl group in 1
might not resemble the phenolate group liberated from a phe-
nylglycoside substrate during enzymatic hydrolysis, the methy-
lene group in the benzyl aglycon should provide additional
space during the envisaged template polymerization accounting
for the lengthening of the Cꢀ ꢀ ꢀO bond during the catalytic
reaction.
Although some procedures for the synthesis of glycosylami-
dine including aliphatic derivatives of glucoamidine and manno-
amidine were reported,9,12 we found several obstacles in our
efforts to adapt this work for a multi gram-scale synthesis of
galactothionolactam as a precursor for a galactonoamidine. Par-
ticularly, the previously suggested epimerization at C-2 during
the synthesis of the glycosylthionolactam intermediate appears
questionable for derivatives with the galacto-configuration.12,13
Along with the synthesis of 1, an optimized multi gram-scale
synthesis for its key intermediate, galactothionolactam, is re-
ported. In addition, debenzylation conditions with iron(III) chlo-
ride versus the use of hydrogen gas and palladium on charcoal
are explored, and computational data on putative epimerization
reactions are provided.
⇑
Corresponding author. Fax: +1 334 844 6954.
E-mail address: Susanne.Striegler@auburn.edu (S. Striegler).
0008-6215/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2011.02.021

898
R. Kanso, S. Striegler / Carbohydrate Research 346 (2011) 897–904
HA
Ph
O
OH
OH
HO
HO
HO
HO
OH
HO
HO
O
O
O
+
OH
O
H
Ph
PhO−
OH
OH
OH
O
OH
OH
HO
HO
H
N
R = benzyl
NR
OH
1
Scheme 1. Model reaction of the enzymatic galactoside hydrolysis indicating the first transition state and a putative structure for an enzyme inhibitor 1.
Analysis of the 1H NMR spectrum of the residue after chromato-
2. Results and discussion
graphic purification revealed the presence of 4 as a major product.
Its identity was confirmed by 1H and 13C NMR spectroscopy, and
high resolution ESI mass spectrometry (see Supplementary data).14
In contrast to literature reports,14 the ESI mass spectrum of the raw
material indicates the presence of another compound with the
same molar mass as 4, which is tentatively assigned to the C-5 epi-
2.1. Multi-gram scale synthesis of 2,3,4,6-tetra-O-benzyl-D-
galactothionolactam (6)
Following
2,3,4,6-tetra-O-benzyl-
2,3,4,6-tetra-O-benzyl-galactoamide (3) into 2,3,4,6-tetra-O-ben-
a
procedure previously outlined by Overkleeft,
D-galactose (2) was transformed via
mer of 4, 2,3,4,6-tetra-O-benzyl-L-altronolactam (5). The presence
zyl-D
-galactonolactam (4);14 lactam 4 was obtained as a colorless
of 5 is furthermore evident after the integration of two doublets
of doublets for the protons H-3 at 3.91 ppm (major product, 4)
and H-3 at 3.85 ppm (minor product, 5) in the 1H NMR spectrum.
The splitting of the signal for H-3 of 4 indicates a cis-coupling to
H-4 at 4.05 ppm (J3,4 = 1.9 Hz), and a trans-coupling to H-2 at
oil after flash chromatography (Scheme 2).14 The overall yield for
the multi gram-scale synthesis of 4 from 2 is 31%, and thus only
slightly lower than the transformation described for a small-scale
synthesis (43% overall yield of 4).14
OBn
O
OBn
O
OBn
BnO
BnO
BnO
BnO
BnO
DMSO, Ac2O
r.t., 16 h
NH3/MeOH
DMSO, Ac2O
r.t., 16 h
OH
NH2
O
BnO
r.t., 16 h, 45 %
over two steps
OH
O
OBn
OBn
OBn
2
3
OBn
BnO
BnO
OBn
BnO
BnO
H
N
6
H
N
4
S
O
OBn
O
OBn
BnO
BnO
OBn
NaCNBH4, HCOOH
Lawesson's reagent, C6H6
reflux, 2h, 93 %
+
+
NH2
O
CH3CN, reflux, 2 h,
71 % over two steps
BnO
OBn
BnO
H
N
H
N
7
5
BnO
BnO
BnO
HO
S
O
OBn
OBn
OBn
BnO
OBn
BnO
Et3O+BF4 , CH2Cl2
0 °C, 2h
-
H
N
C6H5CH2-NH2 (9), CH2Cl2
N
BnO
BnO
SEt
0 °C to r.t., 16 h
80 % for 2 steps
N
OBn
OBn
10
8
OH
OH
HO
HO
HO
H
H2/Pd, TFA/EtOH
N
N
r.t., 1 atm, 16 h, quantitative
HO
N
N
H
OH
OH
1
Scheme 2. Synthesis of N-benzylgalactonoamidine 1.

R. Kanso, S. Striegler / Carbohydrate Research 346 (2011) 897–904
899
4.42 ppm (J3,2 = 9.1 Hz); identical shifts and coupling constants be-
tween the protons H-2, H-3 and H-4 have been previously assigned
to 4.15,14 The coupling constants for the signal of H-3 at 3.85 ppm
(dd, 8.5, 2.1 Hz) for the minor product 5 reveal the same configura-
tion of the protons at H-2, H-3 and H-4 as in 4. The proton H-3 in
apparent epimerization of the product 10 nor was the product
yield significantly increased. This finding opposes previous reports
for the transformation of iminothioethers with the gluco-configu-
ration describing epimerization of the reaction products and
increasing yields in dependence of the molar amounts of aliphatic
amine used.11,12 To explore our findings further and to rationalize
the lack of epimerization at C-2 during the reaction of 8 with 9, we
examined the stability of the D- and L-lactams, and of the 2-epimers
of the corresponding thiolactams with a computational approach
using density functional theory (see below).
2,3,4-6-tetra-O-benzyl-L-altrose was previously described with a
similar chemical shift and related coupling constants (dH-3
3.99 ppm; dd, 10.1, 2.7 Hz) as H-3 in 5.16 Lactam 5 is observed as
the minor product in a 3:2 molar ratio before, and in a 10:1 molar
ratio after chromatographic purification.
Hypothetical side products resulting from partial benzylation of
4 that might cause comparable chemical shifts of the proton sig-
nals in the 1H NMR spectrum were not considered due to the mass
identity for the compounds found in the raw material of the reac-
2.3. Synthesis of N-benzyl-D-galactonoamidine
2.3.1. Debenzylation reactions with hydrogen and Pd on
charcoal
tion mixture by ESI-MS. Although the formation of the L-epimer 5
Our early attempts for the debenzylation of 10 to yield 1 relied
on a procedure described for the debenzylation of tetrabenzylg-
luco- and manno-derivatives in the presence of 30% palladium on
charcoal at 1 atm;12 however, the procedure failed to provide 1
from 10 even when using 1 or 2 wt equiv of the catalyst nor did
the use of higher pressures of hydrogen gas (10 and 25 atm) afford
1. Nevertheless, the debenzylation of galactonolactam (4) proceeds
without difficulty under the described conditions.12
would be expected during the ring closure reaction starting from 3,
it was not described for the small-scale synthesis of 4.14 Complete
elimination of the diasteriomeric side product 5 during the purifi-
cation of 4 by repetitive flash column chromatography over silica
gel was not possible despite all efforts. A very small amount of
an additional unidentified impurity also remained.
Subsequently, galactonolactam 4 was transformed into 2,3,4,6-
tetra-O-benzyl-D-galactothionolactam (6) with Lawesson’s reagent
The failure of the debenzylation reaction of 10 is presumably
due to coordination of palladium to the amidine site in 10. The for-
mation of a Pd-10 complex (Chart 1) was evident from the analysis
of 1H and 13C NMR spectra of a sample aliquot taken from the reac-
tion mixture after 24 h. The 13C signals for the imino carbon and
the aliphatic carbon atom of the N-benzyl group are shifted down-
field by 4.5 ppm and 2.9 ppm, respectively, in comparison to the
corresponding 13C signals of the perbenzylated amidine 10 (see
Supplementary data); the proton signals of H-2, H-3, H-4 and H-
5 appear as broad singlets (spectrum not shown). The integration
of the proton signals in the aromatic region indicates furthermore
the presence of O-benzyl groups and thus an unsuccessful deben-
zylation of 10. The residue obtained after evaporation of the sol-
vent was not soluble in deuterium oxide, and the NMR spectra
were taken in CDCl3 instead. Last, the TLC analysis of a sample ali-
quot on silica gel with cyclohexane–EtOAc (1:1, v/v) as eluent indi-
cated complete transformation of the starting material 10
(Rf = 0.55) toward a slightly less polar compound (Rf = 0.6). These
combined observations provide evidence for the coordination of
the palladium catalyst to the amidine function in 10, which pre-
vents the catalytic hydrogenolysis to proceed.
in 93% yield for a 10 g-scale synthesis following the small-scale
transformation of 4 into 6 as described by Vasella and co-workers
(Scheme 2).17 Likewise, lactams with gluco- and manno-configura-
tion were transformed into corresponding thiolactams.12,13 How-
ever, the latter reports imply a possible epimerization at the
carbon C-2 for derivatives with gluco-configuration during the
reaction,12,13 for which we did not obtain any evidence when
transforming 4 into 6. Instead, the side product 5, obtained during
the synthesis of 4, is converted in the corresponding 2,3,4,6-tetra-
O-benzyl-L
-altrothionolactam (7) as concluded after analysis of 1H
NMR and ESI MS spectra. Attempts to purify 6 on silica gel follow-
ing literature procedures resulted in compound decomposition,13
and were consequently not further pursued. Instead, galactothion-
olactam 6 was efficiently purified by flash chromatography on ba-
sic aluminum oxide applying gradient elution using cyclohexane
and EtOAc. Although the procedure removed some colored, unpo-
lar impurities, the side product 7 was not eliminated. The identity
of 6 with 2,3,4,6-tetra-O-benzyl-D-galactothionolactam described
by Vasella and co-workers,17 was confirmed by NMR and ESI mass
spectrometry (see Supplementary data).
As a consequence, the hydrogenation of 10 with 30% Pd on char-
coal was explored in the presence of anhydrous hydrochloric acid
in ethanol to preclude Pd binding to the amidine function after
protonation.18 However, the reaction failed under these conditions
as well, and only the formation of a Pd-10 complex was observed.
The debenzylation of 10 was eventually achieved by using 2 wt
equiv of Pd on charcoal in the presence of trifluoroacetic acid
obtaining 1 quantitatively for reactions in a 30–100 mg scale.
Surprisingly, 1 is stable under these strongly acidic conditions.
Attempts to use stoichiometric or catalytic amounts of palladium
on charcoal in the presence of TFA were futile. Consequently, the
use of neat TFA for the debenzylation of 10 was not explored.19
2.2. Synthesis of N-benzyl-2,3,4,6-tetra-O-benzyl-D-galactonoa-
midine (10)
In situ activation of 6 with Meerwein’s salt afforded 2,3,4,6-tet-
ra-O-benzyl-D-galactoiminothioether (8) as an intermediate that
was not isolated from the solution,12 but characterized by ESI mass
spectrometry (see Supplementary data). The reaction of 8 with
dried benzylamine (9) in the cold provided N-benzyl-2,3,4,6-tet-
ra-O-benzyl-D-galactonoamidine (10) after purification of the resi-
due of the reaction mixture by flash chromatography over
aluminum oxide. Efforts to follow the procedures described for
the purification of the corresponding aliphatic gluconoamidines
over silica gel resulted in compound decomposition,12 and were
thus not further pursued.
Surprisingly, galactonoamidine 10 was isolated as the sole
product of the reaction. The conclusion is based on the observation
of only one signal for the proton H-3 appearing as a doublet of dou-
blets at 4.04 ppm with a coupling constant of 9.7 Hz in the 1H NMR
spectrum of 10 prior to and after chromatographic purification.
Evidence for the presence of C-5 or C-2 epimers was not obtained
at any time indicating that 7 might have not reacted. Furthermore,
the use of up to 4 mol equiv of 9 in respect to 8 did not cause any
X
Pd
OBn
OBn
N
X
Pd
X
BnO
BnO
BnO
BnO
X
H
N
N
N
H
OBn
OBn
10-Pd
X = solvent or TFA
Chart 1. Putative structure of the Pd-10 complex.

900
R. Kanso, S. Striegler / Carbohydrate Research 346 (2011) 897–904
Partial O-debenzylation of 10, N-debenzylation or a cyclization at
O-4 with loss of the iminopiperidine substructure of 1 could come
into mind considering the harsh reaction conditions (Chart 2).
However, none of the resulting hypothetical structures 1a–c is sup-
ported by the obtained NMR data for 1.
conditions for the encompassing deprotection of a perbenzylated
carbohydrate with an amidine functionality, we explored the use
of FeC13 in dichloromethane.25 The FeC13 reagent has previously
been utilized for the anomerization of glycosides,26 and for the
debenzylation of monosaccharides.27
The 13C signals for the methylene carbon atoms of the O-benzyl
groups in 10 are identified between 70 and 75 ppm, and are thus
significantly different from the 13C signal of the methylene carbon
atom of the N-benzyl group at 45.0 ppm. This signal assignment
was based on a 13C DEPT experiment to distinguish the methylene
carbon atoms from ring carbon atoms. Although the removal of the
N-benzyl group of 10 might appear facile under the applied harsh
conditions, the 13C signal at 45.4 ppm in the NMR spectrum of 1
strongly supports the presence of the N-benzyl group. The integra-
tion of the proton signals in the aromatic region of the 1H NMR
spectrum of 1, and the absence of the methylene carbon atoms
of the O-benzyl groups in the 13C NMR spectrum of 1 furthermore
indicate the complete removal of all O-benzyl groups (see Supple-
mentary data). Consequently, any compound representing partial
O-debenzylation of 10, such as 1a, is not considered any further.
Likewise, the iminopiperidine substructure of 1 has to be con-
sidered intact due to the identified 13C signal of the imino group
at 155.9 ppm. If a bicyclic substructure 1b with cyclization at O-4
was formed after the debenzylation of 10, the hybridization of C-
1 would change, and the 13C signal for the imino carbon atom
would disappear. Again, the obtained 13C NMR data do not support
this hypothesis. The 13C signal of the carbon atom in the C@N
group in a furanolactam derivative with a structure related to 1c
While the debenzylation of 10 proceeds in the presence of
16 equiv of iron(III) chloride without difficulty,25 the isolation
of 1 was found to be tedious in the presence of such large excess
of ferric ions. ESI mass spectrometry on both the aqueous and
the organic layer confirmed the presence of 1 in either layer after
aqueous workup of the reaction mixture. As only minor amounts
of 1 were recovered from the organic layer, extraction and isolation
of 1 from the aqueous layer was undertaken.
Attempts to completely remove the iron(III) ions as iron sulfide
by repetitive addition of sodium sulfide to acidified or neutral
aqueous solutions were futile, leading predominantly to the recov-
ery of excess sodium sulfide, and were consequently not further
pursued. The addition of aqueous sodium hydroxide solution to
the reaction mixture, allowed the complete removal of the ferric
ions as ferric oxide by precipitation. Further purification of 1 and
removal of the sodium chloride byproduct was achieved by ion ex-
change yielding 1 as a yellow oil in 18% overall yield. Initial at-
tempts to use ion exchange for a complete removal of the ferric
ions from the reaction mixture failed most likely due to the large
excess of metal ions relative to the amount of product formed
and the strong coordination of the ferric ions to the amidine func-
tion in 1. Thus, the tedious work-up and the low yield of the iso-
lated product compromise the use of anhydrous iron(III) chloride
for the debenzylation of 10. Consequently, the elaborated condi-
tions with hydrogen gas and Pd on charcoal constitute—despite
the costs and the amount of the catalyst required—the only practi-
cal synthetic route toward galactonoamidine 1. Computational ef-
forts are summarized below to rationalize the lack of epimerization
reactions along the way to synthesize 1 from 4 via 6 (vide infra).
has been previously described by
a
chemical shift of
175.4 ppm,20 which is about 20 ppm downfield of the chemical
shift observed for the imino carbon in 1. The hypothetical structure
1c is therefore not considered any further.
The structure of 1 is proposed as iminopiperidine substructure
based on the evidence provided above, which should encompass
equilibrium structures of an exo- and an endocyclic imino group
(Scheme 2). Vasella and co-workers have previously discussed
such equilibria in depth using 15N NMR spectroscopy and X-ray
analysis concluding the presence of an exocyclic C@N group for
hydroximolactams.13 A similar arrangement for 1 is likely, but will
have to be verified in future studies. An exocyclic C@N group has
been previously assigned to aliphatic gluco- and mannoamidines
referring to the studies by Vasella and co-workers.12
2.4. Computational characterization of putative epimerization
reactions
All computations were performed using the GAUSSIAN 09 suite.28
The geometry of all structures was optimized using Becke’s
three-parameter functional coupled with the correlation function
of Lee at al. (B3LYP) and the basis set 6-31+G(d).29,30 This level of
theory was previously proven to be applicable for the characteriza-
2.3.2. Debenzylation reactions with Fe(III) chloride
tion of the 4C1 and 1C4 conformers of -glucopyranose.31 The use
a-D
In addition to catalytic hydrogenations with Pd or Pt catalysts,
Lewis acid-mediated cleavage of benzyl ethers is widely applied
in multi-step organic syntheses.21 Among others, debenzylation
reactions of aromatic dibenzyloxy aldehydes and the regioselective
debenzylation of monosaccharides and complex oligosaccharides
have been achieved by using magnesium bromide,22 chromium(II)
chloride,23 or iron(III) chloride.24,25 While in need for mild reaction
of the B3-based hybrid density functional together with a split-va-
lence basis set has been shown to perform better than Hartree–
Fock (HF) or Møeller–Plesset (MP2) procedures.31
The Gibbs free energies
DG
298 of the 3H4 and 4H3 half-chair con-
formers of galactonolactam 4, galactothionolactam 6, galactono-
amidine 10, and of their corresponding C-5 epimers (5, 7 and
OH
RO
H
N
1a
HO
R = CH2-C5H6
N
OH
OH
OH
H
OH
OH
HO
N
NH2
OH
H
O
O
N
OH
OH
NHBn
NBn
N
OH
OH
1
1b
1c
Chart 2. Hypothetical structures 1a–c.

R. Kanso, S. Striegler / Carbohydrate Research 346 (2011) 897–904
901
mer 12 in benzene (Table 2). The Lawesson’s reagent used during
the experimental setup is substituted by hydrogen sulfide during
the computation for simplicity reasons.
BnO
OR
H
N
RO
RO
H
12
N
BnO
BnO
OR
N
S
The relative Gibbs free energies for all reactions are positive
indicating non-spontaneous reactions. The Gibbs free energy for
the transformation of the 4H3 half-chair conformer of 4 into the
4H3 half-chair conformer of 6 is about 1 kcal/mol smaller than
the transformation of the same compound in a 3H4 half-chair.
Assuming 4 predominantly in the 4H3 half-chair conformation
(see above), an epimerization at C-2 leading to 12 in either half-
chair conformation is thermodynamically less favored by 0.3–
1.6 kcal/mol. However, lactam 4 in the 3H4 half-chair conformation
requires about 2–3.3 kcal/mol less than 4 in its 4H3 half-chair con-
formation for the transformation into the C-2 epimer 12. As we did
not find any experimental evidence for such epimerization at C-2,
we conclude that 4 predominantly exists and reacts in its 4H3 half-
chair conformation, which was characterized as the conformer
with the lowest energy before.
OBn
OR
11
RO
RO
H
13
N
R = CH2-C5H6
OR
N
Chart 3. Structures of the compounds 11, 12 and 13 (C-2 and C-5 epimers of 6 and
10, respectively).
N-benzyl-2,3,4,6-tetra-O-benzyl-
tially calculated in the gas phase using the B3LYP/6-31G(d) level
of theory. Additionally, the Gibbs free energies
G298 of the C-2
epimers of 6 and 10, namely 2,3,4,6-tetra-O-benzyl- -talothiono-
lactam (12) and N-benzyl-2,3,4,6-tetra-O-benzyl- -talonoamidine
13 were determined (Chart 3). Vibrational frequencies were com-
puted at the same level of theory in order to define optimized
geometries as minima without imaginary frequencies. Solvation
free energies were subsequently computed as single point calcula-
tions of the geometry-optimized structures at the SMD/B3LYP/6-
31+G(d) level of theory in acetonitrile, benzene and
dichloromethane.
L-altronoamidine 11) were ini-
D
D
D
2.4.3. Stability of amidines
Thiolactams with gluco-configuration were experimentally ob-
served to epimerize during the reaction with amines by others pre-
viously.12 In contrast, we isolated 10 as the sole product after the
reaction of 6 with benzylamine and chromatographic purification
(vide infra). We consequently computed the relative Gibbs free
energies during the reactions of the thiolactam 6 with benzylamine
9 in dichloromethane resulting in galactonoamidines 10 and its C-2
epimer 13 (Table 3)
2.4.1. Stability of lactams
The computational data suggest that the 4H3 half-chair con-
former is the predominant form of lactam 4 in acetonitrile and
3.4 kcal/mol more stable than its corresponding 3H4 half-chair con-
former (Table 1). A similar trend is observed upon comparing the
stability of the half-chair conformers for 4 in benzene, for its C-5
epimer 5 in acetonitrile and for thiolactam 6 in benzene (see Sup-
plementary data). These results are in agreement with experimen-
tal results and computational studies for related pyranoses.32 We
consequently focus on the 4H3 half-chair conformers of 4 and 6
during the following discussions.
All computed reactions are non-spontaneous. Like 4, the 4H3
half-chair conformer of 6 is about 4 kcal/mol more stable than its
3H4 half-chair conformer (see Supplementary data), and is there-
fore considered to be the predominant form of 6 during the follow-
ing discussion. The Gibbs free energy for the transformation of the
4H3 half-chair conformer of 6 into the same half-chair conformer of
10 requires 3.9 kcal/mol (Table 3, entry 1), while the transforma-
tion of 6 into either half-chair conformer of it’s C-2 epimer 13 is
additional 1.3–2.1 kcal/mol higher in energy and consequently less
favored (Table 3, entries 3 and 4). The computation thus suggests
that the C-2 epimerization of the major conformer of 6 is thermo-
dynamically not favored.
The 4H3 half-chair conformer of 5 is thermodynamically about
4.6 kcal/mol more stable in acetonitrile than the lowest energy
conformer identified for 4. This finding suggests a considerable
amount of epimerization at C-5 during the ring closing reaction
in acetonitrile, and the formation of the lactams 4 and 5 in this sol-
vent confirming our experimental observations (vide infra). Never-
theless, lactam 4 is the main product of the ring closing reaction,
and the following computational investigation consequently fo-
cuses on its transformation and putative epimerization at C-2.
Nevertheless, after conformational change of the 4H3 half-chair
conformer of 6 into the 3H4 half-chair conformer, epimerization
reactions at C-2 and the formation of 13 would be favored over
the transformation of 6 into 10 by 1.7–2.6 kcal/mol. However, as
there is no experimental evidence for such epimerization products,
we conclude that thiolactam 6, like lactam 4, exists and reacts
predominantly as its 4H3 half-chair conformer during the discussed
reactions.
2.4.2. Stability of thiolactams
Others previously noted unwanted epimerizations at C-2 of
glycolactams with gluco- and manno-configuration during the
transformation into thiolactams with Lawesson’s reagent.12,13 In
contrast, our experimental results do not indicate any evidence
for such a reaction. We therefore calculated the relative Gibbs free
energies for the reaction of 4 into the thiolactam 6 and its C-2 epi-
Table 2
Relative free energies DðDG2f 98Þ (in kcal/mol) for the transformation of lactam 4 into
the thiolactams 6 and its C-2 epimer 12 at the SMD/B3LYP/6-31+G(d) level of theory
in benzene (
Entry
e
= 2.27) at 298.15 K, 1 atm
Table 1
Relative free energies DðDGf298Þ (in kcal/mol) of the lactams 4 and 5 in 4H3 and 3H4
+ H2S
O
+ H2O
half-chair conformation at the SMD/B3LYP/6-31+G(d) level of theory in acetonitrile
S
(e
= 35.69) and benzene (e = 2.27) at 298.15 K, 1 atm
G2f 98
Þ
D
ð
D
G2f 98
Þ
Entry Half-chair
conformation
lactam
D
G2f 98
D
ð
D
G2f 98
Þ
Lactam
Thiolactam
D
ð
D
4 (4H3)
4 (3H4)
6 (4H3)
6 (3H4)
16.0
16.8
0
0.8
(CH3CN)
(C6H6)
1
2
1
2
4H3
3H4
4
4
0
3.4
0
4.4
Epimerization at C-2:
3
4
5
6
4 (4H3)
4 (3H4)
12 (4H3)
12 (3H4)
12 (4H3)
12 (3H4)
17.6
16.3
14.0
12.7
1.6
0.3
C-5 epimer:
3
4
a
4H3
3H4
5
5
ꢁ4.6
ꢁ0.7
–
a
ꢁ2.0
–
ꢁ3.3
a
Not determined.

902
R. Kanso, S. Striegler / Carbohydrate Research 346 (2011) 897–904
0.5 cmꢁ1 using a Shimadzu IR Prestige-21 FT-Infrared Spectrome-
ter;
in cmꢁ1. MS analysis of samples was performed using an Ul-
tra Performance LC Systems (ACQUITY, Waters Corp., Milford, MA,
USA) coupled with a quadrupole time-of-flight mass spectrometer
(Q-TOF Premier, Waters) with electrospray ionization (ESI) in both
ESI-MS and ESI-MS/MS modes operated by the Masslynx software
(V4.1). Each sample in 50% aqueous acetonitrile was directly in-
Table 3
Relative Gibbs free energies DðDG2f 98Þ (in kcal/mol) for the transformation of the
m
thiolactam 6 into the N-benzylamidines 10 and its C-2 epimer 13 at the SMD/B3LYP/
6-31+G(d) level of theory in dichloromethane (
e
= 8.93) at 298.15 K, 1 atm
Bn
+ H2S
N
+ H2N-Bn
S
Entry
Thiolactam
Amidine
D
G2f 98
Dð
D
G2f 98
Þ
jected into the ESI source at a flow rate of 50 lL/min. The ion
1
2
6 (4H3)
6 (3H4)
10 (4H3)
10 (3H4)
3.9
4.3
0
0.4
source voltages were set at 3 kV for positive and negative ion mode
acquisitions, respectively. The sampling cone was set at 37 V and
the extraction cone was at 3 V. In both modes the source and
desolvation temperature was maintained at 120 and 225 °C,
respectively, with the desolvation gas flow at 200 L/h.
Epimerization at C-2:
3
4
5
6
6 (4H3)
6 (3H4)
13 (4H3)
13 (3H4)
13 (4H3)
13 (3H4)
6.0
5.2
2.2
1.3
2.1
1.3
ꢁ1.7
ꢁ2.6
Thin layer chromatography (TLC) was performed using silica gel
TLC plates from SORBENT Technologies, 200
l
m, 4 ꢂ 8 cm, alumi-
num backed, with fluorescence indicator F254 and detection by
UV light or by charring with an aqueous vanillin-sulfuric acid re-
agent and subsequent heating of the TLC plate. The vanillin reagent
is prepared from vanillin (10 g, 0.066 mol) and concentrated sulfu-
ric acid (20 mL) in 0.5 L of water. After stirring of this mixture at
ambient temperature for 2 h, the filtrate of the suspension is used
for charring. Column chromatography was carried out using Silica
3. Conclusions and summary of results
A feasible synthesis of galactonoamidines resembling features
of the transition state for the cleavage of glycosidic bonds is a pre-
requisite for a future development of templated macromolecular
catalysts that might allow selective glycosylation reactions. With
this goal in mind, a practical multi gram-scale synthesis for
Gel 60 from SilicycleÒ (40–63
phase or basic aluminum oxide (pH range 9.7 0.4, activity I),
32–63
m, surface area 150 m2 gꢁ1 from Sorbent Technologies.
lm, 230–240 mesh) as stationary
2,3,4,6-tetra-O-benzyl-D-galactothionolactam (6) was developed.
During the course of the study, the
D-galacto configuration of the
l
carbohydrate was found to be a key element to prevent epimeriza-
tion at C-2 during the formation of galactonolactam 4, galactoth-
ionolactam 6 and galactonoamidine 10. The rational for this
observation was ascribed to the steric hindrance imposed by the
hydroxyl groups at C-2 and C-4 that would be axial to each other
after such epimerization. This hypothesis was further supported
by computational data suggesting that the 4H3 half-chair con-
former is a major conformer for galactonolactam 4 and galactoth-
ionolactam (6). The computational data support the experimental
results and indicate that the epimerization at C-2 of 4 or 6 is ther-
modynamically not favored for the identified major 4H3 half-chair
conformers.
Ion exchange experiments were performed on Dowex 50WX2 ion
exchange resin from Sigma–Aldrich.
Benzylamine (Acros) was distilled in vacuum and stored over
molecular sieves 3 Å prior to use;33 dichloromethane (Pharmco-
Aaper) and benzene (Sigma–Aldrich) were dried dynamically over
neutral aluminum oxide from Acros or basic aluminum oxide from
EMD; DMSO (Fisher) was dried over molecular sieves 4 Å; acetic
anhydride (Fisher) was distilled prior to use. All other chemicals
were reagent grade or better, and were used as received.
4.2. Calculations
The debenzylation of N-benzyl-2,3,4,6-tetra-O-benzyl-
D-galac-
All electronic structure calculations in gas phase and solution
were performed with the GAUSSIAN 09 suite at the Alabama Super-
computer Center.28 The geometry of all structures was optimized
using Becke’s three-parameter functional coupled with the correla-
tion function of Lee at al. (B3LYP) and the basis set 6-31G(d).29,30
The Gibbs free energies
using the B3LYP/6-31G(d) level of theory. Vibrational frequencies
were computed at the same level of theory in order to define opti-
tonoamidine (10) yielding the N-benzyl- -galactonoamidine (1)
D
was explored by using hydrogenation with palladium on charcoal,
and ferric chloride in dichloromethane as reagents. Although the
hydrogenolysis of 10 requires 2 wt equiv of palladium on charcoal
in the presence of a strong acid, the cleavage of the benzyl ether
groups affording 1 was accomplished in significantly higher iso-
lated yields than comparable debenzylation reactions with ferric
chloride. The latter reaction required a very tedious workup proce-
dure for complete ferric ion removal. Current efforts are directed
toward the characterization of the inhibitory activity of 1 during
the enzymatic cleavage of phenylgalactosides, and the study of
its coordination ability to metal complexes, which will be reported
in due course.
D
G298 for all structures were calculated
mized geometries as minima without imaginary frequency. The
298
sol
Gibbs free energies in solution
DG
are computed as single point
calculations at the B3LYP/6-31G(d) level of theory in acetonitrile,
benzene or dichloromethane at 298.15 K and 1 atm using the
SMD solvation model for absolute solvation energies. The SMD
model is implemented in the GAUSSIAN 09 program,34 considers
non-electrostatic solvation terms, and achieves mean unsigned er-
rors of 0.6–1.0 kcal/mol in the solvation free energies with the 6-
31G(d) basis set.34
4. Experimental
4.1. General methods
4.3. Syntheses
1H and 13C NMR spectra were recorded on a Bruker AV400
(400.2 MHz for 1H, and 100.6 MHz for 13C). Chemical shifts (d) in
1H NMR are expressed in ppm and coupling constants (J) in Hz. Sig-
nal multiplicities are denoted as s (singlet), d (doublet), t (triplet), q
(quartet) and m (multiplet). Deuterated chloroform and deuterium
oxide were used as solvents, and chemical shift values are reported
relative to the residual signals of these solvents (CDCl3: d = 7.29 for
1H and d = 77.0 for 13C; D2O: d = 4.80 for 1H and d = 49.0 for 13C
after addition of a few drops of CH3OD). IR spectra of the obtained
oils were recorded as thin films on KBr disks with a resolution of
4.3.1. 2,3,4,6-Tetra-O-benzyl-D-galactonolactam (4)
The title compound 4 and its epimer 5 were obtained from 2
(95.5 g, 0.18 mol) following a procedure disclosed by Overkleft
et al.14 via 2,3,4,6-tetra-O-benzyl-galactoamide
3
(42.5 g,
76.6 mmol), and obtained as yellow oil after purification by flash
chromatography on silica gel using cyclohexane–EtOAc (4:1, v/v)
as eluent. ESI MS analysis revealed
z = 538.2628 (calcd m/z = 538.2593) for the epimer mixture. The
repeated purification of a small sample aliquot by flash chromatog-
a molar mass of m/

R. Kanso, S. Striegler / Carbohydrate Research 346 (2011) 897–904
903
raphy over silica gel using cyclohexane–EtOAc (1:1, v/v) allowed
the isolation of 4 as major product; colorless oil; 71% yield
(29.1 g, 54.2 mmol); Rf = 0.6 (SiO2, cyclohexane–EtOAc (1:1, v/v)).
The compound identity to previously described 2,3,4,6-tetra-O-
–CH2), 1.34 (br s, 1H, NH); dC (CDCl3) 155.9, 139.2, 139.0,
138.4, 138.1, 137.7, 128.5–126.9, 81.9, 75.3, 74.5, 74.0, 73.7,
73,3, 71.2, 71.0, 59.6, 45.0; mmax (KBr, thin film) 3439, 3029,
2922, 2860, 1654, 1495, 1453, 1091, 737, 696 cmꢁ1; ESI-TOF
MS m/z calcd for C41H43N2O4 [M+H]+: 627.3223; found: 627.3216.
benzyl-D
-galactonolactam14 was confirmed by 1H and 13C NMR
spectroscopy and high resolution ESI mass spectrometry; dH
(CDCl3) 7.52–7.30 (m, 20H, PhH), 6.04 (br s, 1H, NH), 5.31 (d,
11.2, 1H, PhCH2), 4.99 (d, 11.6, 1H, PhCH2), 4.90 (d, 11.3, 1H,
PhCH2), 4.86 (d, 11.9, 1H, PhCH2), 4.76 (d, 12.1, 1H, PhCH2), 4.64
(d, 11.5, 1H, PhCH2), 4.53 (q, 11.6, 2H, PhCH2), 4.42 (d, 9.1, 1H, H-
2), 4.05 (br s, 1H, H-4), 3.91 (dd, 1.9, 9.2, 1H, H-3), 3.68–3.59 (m,
2H, H-5, H-6b), 3.51 (dd, 3.2, 7.8, 1H, H-6a); dC (CDCl3) 170.9,
138.2, 138.0, 137.8, 137.3, 128.5–127.5, 80.6, 77.3, 75.3, 74.0,
73.50, 73.0, 72.9, 70.2, 53.5; HRMS, ESI-TOF+ m/z calcd for
4.3.5. N-Benzyl-D-galactonoamidine (1)
Procedure A. Palladium on charcoal (160 mg, 200%; w/w) and
trifluoroacetic acid (1 mL, 13.5 mmol) are added to a solution of
N-benzyl-2,3,4,6-tetra-O-benzyl-D-galactonoamidine (10) (80 mg,
0.128 mmol) in 5 mL of EtOH. The resulting suspension was stir-
red at room temperature in the presence of hydrogen gas at
1 atm. After 16 h, a TLC analysis (SiO2, cyclohexane–EtOAc (1:1,
v/v)) indicated complete transformation of the starting material
(Rf = 0.55) into a more polar spot at the baseline. The reaction
mixture was then diluted with 10 mL of EtOH, centrifuged at
6000 rpm for 6 min and filtered over Celite. The filtrate was con-
centrated under reduced pressure at 38 °C to afford 1 as a light
yellow oil (34 mg, quantitative); Rf = 0.8 (SiO2, MeOH); dH (D2O)
7.50–7.35 (m, 5H, PhH), 4.72–4.62 (m, 3H, –NCH2, –CH), 4.34–
4.30 (m, 1H, –CH), 4.01 (dd, 9.2, 2.4, 1H, –CH), 3.84–3.65 (m,
3H, –C H, –CH2); dC (D2O/CH3OD) 164.6, 133.9, 129.3, 128.6,
127.5, 70.9, 67.2, 66.7, 60.2, 57.7, 45.3; mmax (KBr, thin film)
3363, 1672, 1438, 1206, 1134, 711 cmꢁ1; ESI-TOF MS m/z calcd
for C13H18N2O4 [M+H]+: 267.1345; found: 267.1347.
Procedure B. Anhydrous FeCl3 (332 mg, 2.04 mmol, 16 equiv)
was added to a solution of 10 (80 mg, 0.127 mmol) in 5 mL of
dry dichloromethane at 0 °C under argon, and the resulting reac-
tion mixture was continued to stir in the cold. After 2 h, the reac-
tion was quenched by addition of 1 mL of 0.1 N aqueous
hydrochloric acid. After 10 min of continued stirring, the reaction
mixture was treated with aqueous sodium hydroxide solution to
adjust the pH of the reaction mixture to 6. The immediate forma-
tion of a brownish precipitate is observed presumably due to the
precipitation of iron(III) ions as Fe2O3. The resulting suspension
was centrifuged, and the colorless supernatant concentrated under
reduced pressure to yield a colorless residue.
This residue was triturated with 10 mL of ethanol, centrifuged,
and the supernatant was concentrated under reduced pressure.
The remaining residue was separated in 7.40 mL fractions on a
Dowex 50WX2 ion exchange resin column (2 ꢂ 5 cm). The first
three fractions containing predominantly sodium chloride were
eluted with a solution of 120 mL of 0.1 M aqueous hydrochloric
acid. The following four fractions containing the product were then
eluted with 160 mL of 0.5 M aqueous hydrochloric acid. The ami-
dine-containing fractions were identified by ESI MS analysis, col-
lected and evaporated under reduced pressure to yield 1 as a
yellow oil (6 mg, 18%). The product is identical to one obtained with
procedure A.
C
34H36NO5 [M+H]+: 538.2593; found: 538.2598.
4.3.2. 2,3,4,6-Tetra-O-benzyl- -galactothionolactam (6)
D
A solution of Lawesson’s reagent (4.5 g, 12 mmol) and 4 (9 g,
16.8 mmol) in 200 mL of dry benzene was heated to reflux under
argon for 2 h. The TLC analysis (SiO2, cyclohexane–EtOAc (1:1, v/
v) of an aliquot of the reaction mixture showed complete trans-
formation of the starting material (Rf = 0.6) after that time peri-
od, and the formation of the target compound with an Rf value
of 0.85 along with the presence of a minor side product with
an Rf value of 0.9. The solvent and any other remaining volatile
compounds were subsequently evaporated in vacuum at 40 °C
yielding an orange residue. Purification of this residue by flash
chromatography on aluminum oxide using a solvent gradient
from cyclohexane (100%) to cyclohexane–EtOAc (4:1, v/v) and
cyclohexane–EtOAc (1:1, v/v) allowed the isolation of 6 as a yel-
low oil (8.6 g, 93%); dH (CDCl3) 8.25 (br s, 1H, NH), 7.54–7.30 (m,
20H, PhH), 5.43 (d, 10.6, 1H, PhCH2), 4.94–4.47 (m, 8H, H-2,
PhCH2), 4.12–4.09 (br s, 1H, H-4), 3.86 (dd, 1.9, 7.7, 1H, H-3),
3.78–3.71 (m, 2H, H-5, H-6b), 3.57–3.54 (m, 1H, H-6a); dC
(CDCl3) 201.6, 137.9, 137.8, 137.5, 137.1, 128.6–127.6, 81.0,
78.8, 75.7, 73.5, 73.4, 72.9, 72.2, 69.5, 57.5; mmax (KBr, thin film):
3064, 3033, 2923, 2865, 1517, 1451, 1363, 1315, 1209, 1103,
1030, 738, 697 cmꢁ1; HRMS (+TOF-MS) m/z calcd for C34H35NO4S
[M+H]+: 554.2365; found: 554.2396.
4.3.3. 2,3,4,6-Tetra-O-benzyl-D-galactoiminothioether (8)
Meerwein’s salt (320 mg, 1.68 mmol, 1.5 equiv) is added to a
solution of 6 (620 mg, 1.12 mmol) in 6 mL of dry CH2Cl2 at 0 °C.
The green-yellow solution is stirred at 0 °C for 2 h, and used di-
rectly for the next step without any further purification. A (+)
TOF MS-ESI analysis of the reaction solution indicates the pres-
ence of 8; HRMS m/z calcd for C36H40O4NS [M+H]+: 582.2678;
found: 582.2728.
4.3.4. N-Benzyl-2,3,4,6-tetra-O-benzyl-
solution of distilled and dried benzylamine (245
2.24 mmol) in 2 mL of dry CH2Cl2 is added to a solution of
2,3,4,6-tetra-O-benzyl- -galactoiminothioether (8) under argon
D
-galactonoamidine (10)
Acknowledgment
A
lL,
A CAREER award from the National Science Foundation (CHE-
0746635) to S.S. is gratefully acknowledged. The work was made
possible in part by a grant of high performance computing re-
sources and technical support from the Alabama Supercomputer
Authority.
D
at 0 °C. The resulting solution is stirred over night and allowed
to warm to ambient temperature. After 8 h, the TLC analysis
(SiO2, cyclohexane–EtOAc (1:1, v/v)) of a sample aliquot revealed
complete transformation of the galactoiminothioether 8 (Rf = 0.8)
and appearance of a more polar spot at Rf = 0.55. The reaction
was subsequently stopped by evaporation of all volatile material
in vacuum at ambient temperature; the remaining residue was
purified by flash chromatography on basic aluminum oxide using
gradient elution from cyclohexane to cyclohexane–ethyl cyclo-
hexane (1:1, v/v) yielding 10 as a yellow oil (561 mg, 0.90 mmol)
in 80%; dH (CDCl3) 7.48–7.25 (m, 28H, PhH), 5.05–4.84 (m, 4H,
OCH2), 4.79–4.61 (m, 5H, –CH, OCH2), 4.41–4.27 (m, 3H, –CH,
NCH2), 4.04 (dd, 1.72, 9.7, 1H, –CH), 3.88–3.72 (m, 3H, –CH,
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.carres.2011.02.021.
References
1. Kim, J.-H.; Yang, H.; Boons, G.-J. Angew. Chem., Int. Ed. 2005, 44, 947–949. S947/
1–S947/16.

904
R. Kanso, S. Striegler / Carbohydrate Research 346 (2011) 897–904
2. Kren, V.; Thiem, J. Chem. Soc. Rev. 1997, 26, 463–473.
19. Fletcher, S.; Gunning, P. T. Tetrahedron Lett. 2008, 49, 4817–4819.
20. El Khadem, H. S.; Crossman, A., Jr.; Bensen, D.; Allen, A. J. Org. Chem. 1991, 56,
6944–6946.
21. Greene, T. W.; Wuts, P. G. M. Protective Groups in Organic Synthesis, 4th ed.; John
Wiley & Sons: New York, 2006.
22. Haraldsson, G. G.; Baldwin, J. E. Tetrahedron 1997, 53, 215–224.
23. Falck, J. R.; Barma, D. K.; Venkataraman, S. K.; Baati, R.; Mioskowski, C.
Tetrahedron Lett. 2002, 43, 963–966.
24. Verchere, J.-F.; Chapelle, S.; Xin, F.; Crans, D. C. Prog. Inorg. Chem. 1998, 47, 837–
945.
25. Rodebaugh, R.; Debenham, J. S.; Fraser-Reid, B. Tetrahedron Lett. 1996, 37,
5477–5478.
26. Ikemoto, N.; Kim, O. K.; Lo, L. C.; Satyanarayana, V.; Chang, M.; Nakanishi, K.
Tetrahedron Lett. 1992, 33, 4295–4298.
27. Diaz Diaz, D.; Miranda, P. O.; Padron, J. I.; Martin, V. S. Curr. Org. Chem. 2006, 10,
457–476.
28. Frisch, M. J.; Gaussian: Wallingford CT, 2009 (full reference given in the
Supplementary data).
29. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. 1988, 37, 785–789.
30. Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
31. Klein, R. A. J. Am. Chem. Soc. 2002, 124, 13931–13937.
32. Collins, P. M.; Ferrier, R. J. Monosaccharides—Their Chemistry and Their Role in
Natural Products; John Wiley & Sons: Chichester, England, 1995.
33. Burfield, D. R.; Smithers, R. H.; Tan, A. S. C. J. Org. Chem. 1981, 46, 629–631.
34. Marenich Aleksandr, V.; Cramer Christopher, J.; Truhlar Donald, G. J. Phys.
Chem. 2009, 113, 6378–6396.
3. Vasella, A.; Davies, G. J.; Böhm, M. Curr. Opin. Chem. Biol. 2002, 6, 619–629.
4. Lillelund, V. H.; Jensen, H. H.; Lian, X.; Bols, M. Chem. Rev. 2002, 102, 515–553.
5. Zechel, D. L.; Withers, S. G. Acc. Chem. Res. 2000, 33, 11–18.
6. Heightman, T. D.; Vasella, A. T. Angew. Chem., Int. Ed. 1999, 38, 750–770.
7. Bleriot, Y.; Dintinger, T.; Genre-Grandpierre, A.; Padrines, M.; Tellier, C. Bioorg.
Med. Chem. Lett. 1995, 5, 2655–2660.
8. Bleriot, Y.; Dintinger, T.; Guillo, N.; Tellier, C. Tetrahedron Lett. 1995, 36, 5175–
5178.
9. Bleriot, Y.; Genre-Grandpierre, A.; Tellier, C. Tetrahedron Lett. 1994, 35, 1867–
1870.
10. Guo, W.; Hiratake, J.; Ogawa, K.; Yamamoto, M.; Ma, S. J.; Sakata, K. Bioorg. Med.
Chem. Lett. 2001, 11, 467–470.
11. Papandreou, G.; Tong, M. K.; Ganem, B. J. Am. Chem. Soc. 1993, 115, 11682–
11690.
12. Heck, M.-P.; Vincent, S. P.; Murray, B. W.; Bellamy, F.; Wong, C.-H.;
Mioskowski, C. J. Am. Chem. Soc. 2004, 126, 1971–1979.
13. Hoos, R.; Naughton, A. B.; Thiel, W.; Vasella, A.; Weber, W.; Rupitz, K.; Withers,
S. G. Helv. Chim. Acta 1993, 76, 2666–2686.
14. Overkleeft, H. S.; van Wiltenburg, J.; Pandit, U. K. Tetrahedron 1994, 50, 4215–
4224.
15. Nishimura, Y.; Adachi, H.; Satoh, T.; Shitara, E.; Nakamura, H.; Kojima, F.;
Takeuchi, T. J. Org. Chem. 2000, 65, 4871–4882.
16. Takahashi, H.; Hitomi, Y.; Iwai, Y.; Ikegami, S. J. Am. Chem. Soc. 2000, 122, 2995–
3000.
17. Vonhoff, S.; Heightman, T. D.; Vasella, A. Helv. Chim. Acta 1998, 81, 1710–1725.
18. Racine, E.; Bello, C.; Gerber-Lemaire, S.; Vogel, P.; Py, S. J. Org. Chem. 2009, 74,
1766–1769.