An unexpected elimination product leads to 4-alkyl-4-deoxy-4-epi-sialic acid derivatives

Article abstract of DOI:10.1139/V08-006

A useful, unexpected β,γ-unsaturated-α-keto ester (ethyl (E)-5-acetamido-3,4,5-trideoxy-6,7:8,9-di-O-isopropylidene-D-manno-non-3-en-2- ulosonate 5) was isolated in 91% yield following ozonolysis and chromatographic purification of its enoate ester precursor ethyl 5-acetamido-2,3,4,5-tetradeoxy- 6,7:8,9-di-O-isopropylidene-2-methylene-4-nitro-D-glycero-D-galacto-nononate (6). When the 4R enoate ester (ethyl 5-acetamido-2,3,4,5-tetradeoxy-6,7:8,9-di- O-isopropylidene-2-methylene-4-nitro-D-glycero-D-talo-nononate, 7) was subjected to the same conditions, enone 5 was a minor product (18%) while the major product did not eliminate HNO₂ but instead cyclized to form a five-membered ring containing a hemiaminal linkage between C-2 and the amide nitrogen on C-5 (9, 70%). Conjugate addition, to enone 5 opens up the potential to generate 4-substituted sialic acid derivatives, a general route to such compounds that has not been previously reported. In a preliminary investigation of such a route, diethylzinc and dimethylzinc were added to enone 5 resulting in generation of 4-alkyl-substituted cyclic hemiaminal structures 11 and 13, which could be deprotected to form 2,7-anhydrosialic acid analogues 14 and 15. These products could then be converted to peracetylated glycals 16 and 17, the 4-methyl-substituted compound 17 being finally deprotected to give a 4-methyl-substituted analogue of the glycal of sialic acid (5-acetamido-2,6- anhydro-3,4,5-trideoxy-4-methyl-D-glycero-D-talo-non-2-enonic acid 18).

Full text of DOI:10.1139/V08-006

238
An unexpected elimination product leads to
4-alkyl-4-deoxy-4-epi-sialic acid derivatives
Ivan Hemeon and Andrew J. Bennet
Abstract: A useful, unexpected β,γ-unsaturated-α-keto ester (ethyl (E)-5-acetamido-3,4,5-trideoxy-6,7:8,9-di-O-
isopropylidene-D-manno-non-3-en-2-ulosonate 5) was isolated in 91% yield following ozonolysis and chromatographic
purification of its enoate ester precursor ethyl 5-acetamido-2,3,4,5-tetradeoxy-6,7:8,9-di-O-isopropylidene-2-methylene-
4-nitro-D-glycero-D-galacto-nononate (6). When the 4R enoate ester (ethyl 5-acetamido-2,3,4,5-tetradeoxy-6,7:8,9-di-O-
isopropylidene-2-methylene-4-nitro-D-glycero-D-talo-nononate, 7) was subjected to the same conditions, enone 5 was a
minor product (18%) while the major product did not eliminate HNO2 but instead cyclized to form a five-membered
ring containing a hemiaminal linkage between C-2 and the amide nitrogen on C-5 (9, 70%). Conjugate addition to
enone 5 opens up the potential to generate 4-substituted sialic acid derivatives, a general route to such compounds that
has not been previously reported. In a preliminary investigation of such a route, diethylzinc and dimethylzinc were
added to enone 5 resulting in generation of 4-alkyl-substituted cyclic hemiaminal structures 11 and 13, which could be
deprotected to form 2,7-anhydrosialic acid analogues 14 and 15. These products could then be converted to
peracetylated glycals 16 and 17, the 4-methyl-substituted compound 17 being finally deprotected to give a 4-methyl-
substituted analogue of the glycal of sialic acid (5-acetamido-2,6-anhydro-3,4,5-trideoxy-4-methyl-D-glycero-D-talo-non-
2-enonic acid 18).
Key words: conjugate addition, dialkylzinc reagent, sialic acid, ozonolysis, inhibitors.
Résumé : À la suite de l’ozonolyse et la purification de son précurseur, l’ester énoate 5-acétamido-2,3,4,5-tétradésoxy-
6,7:89-di-O-isopropylidène-2-méthylène-4-nitro-D-glycéro-D-galactonononate d’éthyle (6), on a isolé avec un rendement
de 91% le (E)-5-acétamino-3,4,5-tridésoxy-6,7:8,9-di-O-isopropylidène-D-manno-non-3-én-2-ulosonate d’éthyle (5), un
α-cétoester β,γ-insaturé utile et inattendu. Lorsqu’on soumet l’ester énoate 4R [5-acétamino-2,3,4,5-tétradésoxy-6,7:8,9-
di-O-isopropylidène-2-méthylène-4-nitro-D-glycéro-D-tallo-nononate d’éthyle (7)], aux même conditions, l’énone 5 n’est
qu’un produit mineur (18%) alors que le produit majeur n’élimine pas de HNO2, mais conduit plutôt à un produit de
cyclisation comportant un cycle à cinq chaînons et une liaison hémiaminale entre le C-2 et l’azote de l’amide en C-5
(9, 70%). Une addition conjuguée sur l’énone 5 donne la possibilité de généré des dérivés de l’acide sialique substitués
en position 4 pour lesquels aucune voie de synthèse générale n’avait été proposée antérieurement. Dans une étude pré-
liminaire de cette route, l’addition de diéthylzinc et de diméthylzinc à l’énone 5 conduit à la formation des structures
hémiaminales cycliques 11 et 13 portant des substituants alkyles en position 4 qui peuvent être déprotégées pour
conduire aux analogues 14 et 15 de l’acide 2,7-anhydrosialique. Ces produits peuvent ensuite être convertis en glycals
peracétylés 16 et 17, le composé 17 portant un groupe méthyle en position 4 étant finalement déprotégé pour fournir
un analogue substitué en position 4 de l’acide sialique, l’acide 5-acétamido-2,6-anhydro-3,4,5-didésoxy-4-méthyl-D-gly-
céro-D-talo-non-2-énonique, 18.
Mots-clés : addition conjuguée, réactif dialkylzinc, acide sialique, ozonolyse, inhibiteurs.
[Traduit par la Rédaction] Hemeon and Bennet 247
Introduction
gates, sialic acid and the proteins that recognize it are impor-
tant in many biological events, including cell recognition
and pathogen infectivity (1, 2). For instance, influenza virus
sialidase activity is critical for viral proliferation (3). The vi-
rus first recognizes and binds to sialic acid residues on the
surfaces of host-cell glycoconjugates in the upper respiratory
tract, an event that initiates the infection cycle. After replica-
tion, the newly formed virus particles emerge from the in-
fected cell and become coated in cellular sialic acids that
cause the virus particles to clump together. The influenza
sialidase then cleaves bound sialic acid residues and thus al-
lows the virus particles to move freely to infect further cells.
As a consequence, inhibitors of this enzyme, which include
the glycal of sialic acid, 2-deoxy-2,3-didehydro-N-acetyl-
neuraminic acid (DANA, 2) and its 4-guanidinyl derivative
Sialic acid (N-acetylneuraminic acid, 1) is a nine-carbon
sugar that is often found α-ketosidically linked to the termini
of oligosaccharides on glycoconjugates (1). These linkages
can be hydrolyzed by sialidases (neuraminidases, EC
3.2.1.18). Because of their terminal locations on glycoconju-
Received 3 November 2007. Accepted 21 December 2007.
Published on the NRC Research Press Web site at
canjchem.nrc.ca on 19 February 2008.
I. Hemeon and A.J. Bennet.1 Department of Chemistry,
Simon Fraser University, 8888 University Drive, Burnaby,
BC V5A 1M5, Canada.
1Corresponding author (e-mail: bennet@sfu.ca).
Can. J. Chem. 86: 238–247 (2008)
doi:10.1139/V08-006
© 2008 NRC Canada

Hemeon and Bennet
239
Fig. 1. Sialic acid (Neu5Ac, 1) and structurally similar sialidase
inhibitors (2–4).
could be opened with various nucleophiles to generate
4-disubstituted sialic acids (13). As well, several modifica-
tions at the 4-position of DANA (2) have been made from a
common oxazoline intermediate resulting from the
BF3@OEt2-promoted displacement of a 4-acetyl group by the
carbonyl oxygen of the neighbouring 5-acetamido group.
This oxazoline intermediate has been hydrogenated to give
4-deoxy-DANA, hydrolyzed to give 4-epi-DANA, and
opened with azide to give 4-azido-4-deoxy-DANA that was
reduced to 4-amino-4-deoxy-DANA and then transformed
into 4-formamido- and 4-acetamido-substituted 4-deoxy-
DANA analogues (14). The oxazoline intermediate has also
been opened with halogens to generate allylic halides (15),
which have been used in palladium(0)-mediated couplings
with organotin reagents to install unsaturated functional
groups at the 4-position (16). Although these intermediates
are useful for synthesizing 4-modified sialic acid analogues,
the number of such analogues that can be generated is lim-
ited.
Recently, the synthesis of 4-deoxy-4-nitrosialic acid was
reported (17), and during the course of that study, an unex-
pected β,γ-unsaturated-α-keto ester (5) was isolated follow-
ing the ozonolysis of enoate ester 6. The results described
herein report the investigation into this elimination reaction
and the use of this enone in conjugate addition reactions for
the installation of different substituents at the 4-position in
sialic acid analogues.
that is marketed as RelenzaTM (3), are useful biological tools
(Fig. 1) (3).
It is important to be able to synthesize analogues of sialic
acid to generate selective inhibitors of sialic acid-metabolizing
enzymes, which can also be used as probes of the biological
pathways. It is therefore advantageous to have access to general
synthetic routes that can lead to a number of derivatives for the
generation of compound libraries. While there are several posi-
tions that can be varied on sialic acid and many derivatives
have been made (3, 4), carbon 4 remains a target for
derivatization. This is partly due to the fact that variation at this
position is tolerated by influenza sialidase as shown by the suc-
cessful development of 4-substituted sialic acid analogues as
influenza sialidase inhibitors, such as RelenzaTM (3) and
TamifluTM (4) bearing 4-guanidinyl and 4-amino substituents,
respectively (Fig. 1) (5, 6).
Design of a general route to 4-modified sialic acid ana-
logues is synthetically challenging. Sialic acid analogues can
be made chemoenzymatically using sialic acid aldolase (N-
acylneuraminate pyruvate lyase, EC 4.1.3.3), which catalyzes
the nucleophilic attack of pyruvate on the aldehyde carbon of
N-acetylmannosamine to form sialic acid. This enzyme is
fairly promiscuous with respect to its specificity toward vari-
ous aldose units, and it can thus be used to make many sialic
acid analogues; however, a hydroxyl group is mandatorily
installed at the 4-position in the resulting addition product (7).
As well, many chemical synthetic routes involve nucleophilic
attack of a three-carbon moiety on an aldose carbonyl group,
also installing a 4-hydroxyl group (8–10).
While several 4-substituted sialic acid analogues have
been synthesized (4), many of these substituents had to be
incorporated early in the synthetic route giving intermediate
compounds with limited synthetic versatility. Several 4-
substituted sialic acid analogues had been synthesized from
a common 4-oxo intermediate by Zbiral and co-workers in
which the 4-hydroxyl group had been oxidized to a ketone.
This could be reduced to give 4-epi-sialic acid (11), undergo
addition of a nucleophilic methyl group to generate both
diastereomers of the 4-methyl addition product (12) and be
transformed into the 4-C-methylene derivative (12). The
4-C-methylene derivative could be reduced to give two
4-epimers of 4-deoxy-4-methylsialic acid (12) and was con-
verted into an epoxide by von Itzstein and co-workers that
Results and discussion
Isolation of ␤,␥-unsaturated ␣-keto ester 5
A diastereomeric mixture of isopropylidene-protected
enoate esters 6 and 7 was synthesized in six steps from
D-arabinose as previously reported (17). The intention was to
cleave the methylene group of 6 through ozonolysis to in-
stall the α-keto ester moiety (8), which would allow the mol-
ecule to cyclize to the pyranose structure of sialic acid upon
removal of the isopropylidene protecting groups. The initial
ozonolysis was performed solely on the 4S enoate ester 6
rather than on the diastereomeric mixture of 6 and 7 to re-
duce the complexity of product identification. However, the
only product isolated from the reaction, following chromato-
graphic purification, was the β,γ-unsaturated-α-keto ester 5
in 91% yield resulting from oxidative cleavage of the methy-
lene group and elimination of HNO2 across carbons 3 and 4
1
(Scheme 1). IR, H NMR, and 13C NMR spectra supported
the structure of enone 5 because of the presence of signals in
the vinylic regions of the NMR spectra, a ketone in the 13C
NMR spectrum, and the absence of nitro stretching bands in
the IR spectrum. The trans-geometry of the double bond was
assigned based on the large H3/H4 coupling constant (J3,4
=
15.9 Hz).
Isolation of enone 5 was an unexpected result, as no base
is present during the ozonolysis of 6 that could effect such
an elimination. It was reasoned that the elimination must be
occurring post-ozonolysis because if the elimination oc-
curred during ozonolysis, the newly-formed double bond
would also be cleaved. The ozonolysis was repeated in the
presence of acetic acid to neutralize any possible base
generated during the reaction, but enone 5 was again the sole
product isolated. Carrying out the ozonolysis in a 1:1
© 2008 NRC Canada

240
Can. J. Chem. Vol. 86, 2008
Scheme 1. Ozonolysis of 4S enoate ester 6 leads exclusively to unstable α-keto ester 8 that eliminates to enone 5 upon silica chroma-
tography or treatment with DBU, while ozonolysis of 4R enoate ester 7 leads predominantly to hemiaminal 9.
dichloromethane/methanol solvent mixture or using
triphenylphosphine as a reducing agent in place of methyl
sulfide also had no effect on the reaction outcome.
hemiaminal anomers (9, Scheme 1). These anomers were
observed to interconvert in solution by H NMR spectros-
1
copy, but only the major anomer could be isolated in suffi-
cient quantities for adequate characterization. The
hemiaminal structure was supported by the disappearance of
It was noticed that very little enone was present in the
crude ozonolysis product mixture as judged by 1H NMR
spectroscopy. In fact, the major compound present in the
crude product mixture appeared to be the α-keto ester 8 that
had been originally desired (Scheme 1). No elimination had
1
the amide-proton signal around 6.0 ppm from the H NMR
spectrum and by the appearance of a singlet at 4.8 ppm that
disappeared when a drop of D2O was added to the CDCl3
solution, corresponding to the new 2-hydroxyl group. In ad-
dition, an HMBC spectrum showed a correlation between H-
5 and C-2, which would not be possible unless the amide ni-
trogen was bonded to C-2. As well, the IR spectrum indi-
cated the nitro group was intact.
1
occurred owing to the absence of vinylic signals in the H
and 13C NMR spectra, and the IR spectrum showed the nitro
group was intact. Upon chromatographic purification on
silica gel, however, complete elimination to enone 5 was ob-
served, indicating that the elimination of HNO2 was silica-
induced (Scheme 1). Indeed, this silica-induced elimination
was not efficient enough to convert α-keto ester 8 into enone
5 completely on larger scales (5 and 8 are co-polar); thus,
the elimination could be induced by treating the crude α-
keto ester 8 with DBU (Scheme 1). It was crucial not to al-
low this elimination to proceed for extended periods, since
complete product decomposition was observed after 15 min.
Of note, Yao and co-workers noticed a similar silica-induced
elimination when synthesizing an α-keto ester similar to 8
that contained a 4-azido group en route to 4-azido-4-
deoxysialic acid; this ester eliminated HN3 upon chromato-
graphic purification (18). These researchers circumvented
the problem by using the 4-azido α-keto ester without purifi-
cation.
It therefore appears as though the ozonolysis product of
the 4S enoate ester 6 prefers to eliminate HNO2 to give
enone 5, while ozonolysis of the 4R enoate ester 7 gives rise
to a product that cyclizes to a five-membered ring
hemiaminal 9 (Scheme 1). Fortunately, the 4R enoate ester 7
could be epimerized to a 3:2 mixture of 6/7 that was separa-
ble by chromatography by treatment with DBU in THF for
40 min, allowing the conversion of 7 to the synthetically
useful 4S enoate ester 6. Interestingly, when enoate ester 6
does not have isopropylidene protecting groups, no elimina-
tion occurs after ozonolysis; the product spontaneously
cyclizes to the sialic acid pyranose (17). This, in addition to
the fact that the ozonolysis product of the 4R enoate ester 7
prefers to cyclize to a hemiaminal, suggests the possibility
that cyclization is faster than elimination for these com-
pounds and elimination is prohibited following cyclization.
While the reasons for these differences in reactivity are not
apparent, it appears as though something prevents or dramat-
ically slows down cyclization of 8 to a hemiaminal structure,
allowing it to undergo facile HNO2 elimination.
It was later noticed that ozonolysis of diastereomeric mix-
tures of the enoate esters 6 and 7 resulted in isolation of
varying yields of enone 5 that reflected the 6/7 ratio, indicat-
ing that only one diastereomer may have eliminated HNO2.
When the 4R enoate ester 7 was subjected to ozonolysis by
itself, enone 5 was isolated in 18% yield, but the major
product, isolated in 70% yield, was a saturated product in
which the 5-acetamido nitrogen cyclized onto the newly
generated 2-keto group to form two five-membered ring
Conjugate addition reactions with enone 5
Although isolation of enone 5 was initially disappointing,
© 2008 NRC Canada

Hemeon and Bennet
241
it was quickly realized that this enone could be a useful syn-
thetic intermediate, since a conjugate addition would afford
a sialic acid analogue bearing a new substituent in the 4-
position. Indeed, these types of enones have been deliber-
ately synthesized for various purposes. Vasella and co-
workers synthesized a similar enone from a protected enoate
ester similar to 6/7 bearing a 4-O-acetyl group, eliminating
HOAc with sodium bicarbonate during ozonolysis. They
then hydrogenated the resulting double bond, leading to
4-deoxysialic acid (19). Shing also synthesized a related 4-
deoxysugar by hydrogenation of such an enone resulting
from a Wittig reaction between an aldose and a phosphorus
ylide containing an α-keto ester moiety (20). As well, this
enone was used in alkoxymercuration reactions resulting in
installation of alkoxy groups in the 4-position after reduction
of the organomercury intermediate (21). However, literature
precedent could not be found for conjugate additions being
attempted on such enones.
Introduction of a methyl group was attempted via well-
known addition reactions using alkylcopper reagents (22).
Several conditions were tried for the conjugate addition of a
methyl group to enone 5, including the following:
(i) Me2CuLi from 5 equiv. CuI and 10 equiv. MeLi, warming
from –78 °C to room temperature while stirring overnight;
(ii) Me2CuLi with TMSCl and BF3·OEt2 as promoters;
(iii) Me2CuMgBr from MeMgBr instead of MeLi; and
(iv) MeCuBF3 prepared from 1 equiv. each of CuI, MeLi,
and BF3·OEt2. None of these conditions afforded appreciable
quantities of conjugate addition products. Also, 1 equiv. of
MeLi was added to enone 5 on one occasion, but no 1,2- or
1,4-addition products were observed, and most of the start-
ing material was recovered, indicating the low reactivity of 5
toward these nucleophiles.
Since the conditions for conjugate addition of alkylcopper
reagents to enone 5 seemed to be too harsh, conjugate
additions of alkyl halides under radical conditions were
attempted. These reactions are normally performed with
Bu3SnH/AIBN in refluxing benzene to generate radicals
thermally that are added to α,β-unsaturated carbonyl com-
pounds containing electron-withdrawing groups (23, 24). Initially,
benzyl bromide was used as an alkylating agent with
Bu3SnH and AIBN in refluxing toluene. This yielded a com-
plex mixture of products, none of which seemed to have in-
corporated a benzyl group. The same reagents were tried
using a 100 W incandescent light source to generate radicals
at room temperature but also gave no addition products.
These conditions using a 100 W lamp at room temperature
were tried with isopropyl iodide and allyl iodide as alkylating
agents to try to generate a stable alkyl radical but resulted in
decomposition of the enone. Slow addition of a solution
containing Bu3SnH and AIBN over 1 h via syringe pump to
a solution of allyl iodide and enone 5 in toluene at 60 °C
was attempted to generate radicals more slowly in hopes that
they would react with the enone, but this reaction also failed.
Milder conjugate additions were attempted using a
dialkylzinc reagent with a copper catalyst, as alkylzinc re-
agents are less reactive than their alkylcopper counterparts
(25). Alexakis and co-workers have screened a number of
copper catalysts for use in conjugate addition reactions with
dialkylzinc reagents and found that Cu(OTf)2 was one of the
most active, and these reactions were greatly accelerated by
the addition of a phosphorus ligand, usually PBu3 or P(OEt)3
(26, 27). Thus, Cu(OTf)2 and P(OEt)3 were mixed with
diethylzinc and enone 5 to give a 53% yield of conjugate ad-
dition products (Scheme 2). The initial product in the crude
reaction mixture appeared to be a single diastereomer of the
1
enol tautomer 10 as observed by H NMR spectroscopy. It
was clear that the ethyl group had added due to the presence
1
of two new signals in the H and 13C NMR spectra, and al-
1
though H-4 had shifted upfield to 2.8 ppm in the H NMR
spectrum, H-3 appeared as a doublet integrating to one pro
-
ton at 5.5 ppm, indicating that it was a vinylic proton. As
well, a singlet was observed at 6.5 ppm that disappeared
upon addition of a drop of D2O to the CDCl3 solution, indi-
cating that this was the enol hydroxyl proton. However, the
enol could not be purified for proper characterization, since
upon silica chromatography, the product tautomerized to
form an anomeric mixture of five-membered ring
hemiaminal structures, giving 11 in 53% yield (Scheme 2).
As with the 4-nitro hemiaminal 9, the structure of 4-ethyl
hemiaminal 11 was assigned based on the disappearance of
1
the amide proton signal in the H NMR and the appearance
of a signal for the new hydroxyl group proton around
4.9 ppm that disappeared when a drop of D2O was added to
the CDCl3 solution. An HMBC spectrum showed a correla-
tion between H-5 and C-2, which further supported the five-
membered ring structure. As well, the IR spectrum showed
absence of an NH stretching band but did show an OH
stretch. As with the 4-nitro hemiaminal 9, only the major
anomer of 11 could be purified for adequate characteriza-
tion, but the two anomers were observed to interconvert
when left for extended periods in CDCl3 solution, indicating
that they were indeed anomers and not C4-epimers. Since 11
was a five-membered ring, the stereochemistry of the ethyl
addition could be probed through a series of 1D NOE differ-
1
ence H NMR spectra. Upon irradiation of the methylene
protons of the newly-installed ethyl group, enhancement of
the signals for H-5 and H-3a (on the same face) was ob-
served, while irradiation of H-3b (on the opposite face) only
enhanced the signal for H-4. These spectra indicate that H-5
and the new ethyl group are on the same face of the ring,
which possesses the 4R-configuration. The exclusive forma-
tion of the 4R-diastereomer is unfortunate, since natural
sialosides possess the 4S-configuration; however, 4-epi-
sialic acid derivatives can still be useful inhibitors/probes for
various sialic acid processing enzymes (28). While there are
several possible explanations for this diastereoselectivity, it is
likely that the alkylzinc/alkylcopper complex is coordinating
to the acetamido group, thus directing nucleophilic attack
from the same face to give cis-addition.
Several variables were manipulated to try to increase the
yield of the 4-ethyl hemiaminal 11. Specifically, no reaction
was observed when THF was used as a coordinating solvent,
and when BF3·OEt2 was used to replace P(OEt)3, no reaction
products were afforded in either THF or toluene. Moreover,
longer reaction times resulted in the formation of increased
amounts of other unidentified products, and the use of larger
amounts of diethylzinc did not result in increased product
yields. Alexakis and co-workers have observed that α,β-
unsaturated esters are less reactive toward copper-catalyzed
dialkylzinc conjugate additions (26); thus, the 53% yield of
4-ethyl hemiaminal 11 was considered to be satisfactory.
© 2008 NRC Canada

242
Can. J. Chem. Vol. 86, 2008
Scheme 2. Conjugate addition of diethylzinc and dimethylzinc to
enone 5, initially affording enol products 10 and 12 that rear-
range to five-membered ring hemiaminals 11 and 13 upon chro-
matography.
with enone 5 but only afforded a very small amount of 4-
methyl hemiaminal 13, likely resulting from addition of
unreacted dimethylzinc. The synthesis of diphenylzinc
through transmetallation was accomplished by reacting 2
equiv. phenyl magnesium bromide with 1 equiv. ZnCl2 in
toluene at –30 °C for 2 h and gave a 68% yield of 3-
phenylcyclohexanone when reacted with 2-cyclohexen-1-
one. However, no conjugate addition products were obtained
when diphenylzinc was added to enone 5. The failure of
benzyl and phenyl groups to add to enone 5 in a conjugate
manner suggests that such groups may be too bulky to add
to the sterically congested 4-position of this enone. It may
also be that impurities generated during the synthesis of the
alkylzinc reagents impeded the conjugate addition to enone
5; thus, it may be useful to purify these alkylzinc reagents
prior to performing the addition reactions.
O
AcHN
EtO
O
O
O
O
O
5
i) 1.5 eq. R2Zn, 5% Cu(OTf)2
10% P(OEt)3, toluene
–20 oC, 15 min
ii) 5, –20 oC to RT, 1 d
O
AcHN
10 R=Et
12 R=Me
EtO
O
O
O
OH
R
R
O
Conversion of 4-alkylhemiaminals to sialic acid
analogues
chromatography
To convert the conjugate addition products 11 and 13 into
sialic acid analogues, removal of the isopropylidene protect-
ing groups was necessary to allow rearrangement to the six-
membered ring pyranose form of sialic acid. This was
achieved by stirring the 4-alkyl hemiaminals 11 and 13 in
90% aq. trifluoroacetic acid. These conditions resulted in re-
moval of the isopropylidene groups, allowing the com-
pounds to cyclize to the pyranose form, but also induced the
formation of a 2,7-anhydro linkage to give the 2,7-anhydro-
4-ethylsialoside 14 and the 2,7-anhydro-4-methylsialoside
15 (Scheme 3). These 2,7-anhydrosugars were moderately
unstable to silica gel chromatography and were normally
used in subsequent reactions without purification. The 2,7-
anhydro linkage in 15 was assigned based on NMR spectro-
scopic evidence, as H-5 and H-6 showed very little coupling
to any other protons, indicating they were not trans-diaxial,
and the frequencies of H-6 and C-2 were shifted downfield
from 4.2 to 4.6 ppm and from 94 to 105 ppm, respectively.
Zbiral and co-workers noticed similar spectral differences
when they made a 4-deoxy-4-epi-4-methylsialic acid by
deprotecting its precursor under acidic conditions, giving a
product that contained a 2,7-anhydro linkage (12). HMBC
spectra of 14 and 15 showed correlations between H-7 and
C-2, which would only be possible if a 2,7-anhydro linkage
existed. Incidentally, these spectra also showed correlations
between H-6 and C-2, indicating that the compounds were
indeed in their pyranose forms. It therefore seems as though
4-alkyl-4-deoxy-4-epi-sialic acid analogues preferentially
form such 2,7-anhydro linkages under acidic conditions.
Since 2,7-anhydrosialic acids are of interest themselves as
conformationally restrained derivatives (34), this synthetic
route has the potential to lead to a number of such deriva-
tives.
HO
EtO2C
11 R=Et, 53%
13 R=Me, 39%
N
Ac
O
O
O
O
The scope of the copper-catalyzed conjugate addition to
enone 5 was examined by testing other dialkylzinc reagents.
Commercially available dimethylzinc was added, which also
afforded a single diastereomer of the conjugate addition
product as its enol tautomer 12 that rearranged to the five-
membered ring hemiaminal structure during silica chroma-
tography to give 13 in 39% yield (Scheme 2). Lower product
yields were expected, since dimethylzinc is known to be less
reactive than diethylzinc (29). The 4-methyl hemiaminal 13
exhibited similar spectral characteristics to its ethyl
homologue 11; notably, a series of 1D NOE difference
1H NMR spectra showed that 13 also had the 4R-configura-
tion. Also similarly, the 4-methyl hemiaminal product was
obtained as an interconverting mixture of anomers of which
only the major anomer could be purified for adequate char-
acterization although NMR spectra were obtained on the mi-
nor anomer.
Attempts to perform conjugate addition reactions were
made using other alkylzinc reagents that had been synthe-
sized. Benzyl zinc bromide was generated by stirring benzyl
bromide with zinc dust in toluene at room temperature for
6 h (30, 31), but it afforded no conjugate addition products
when reacted with enone 5 under the standard conditions
despite giving a 26% yield of 3-benzylcyclohexanone when
reacted with
a
model enone (2-cyclohexen-1-one).
Dibenzylzinc was synthesized by zinc-halogen exchange
from the reaction of 1 equiv. dimethylzinc with 2 equiv.
benzyl bromide in toluene at –30 °C for 45 min (25, 32).
When this was added to 2-cyclohexen-1-one, a 22% yield of
3-benzyl-2-methylcyclohexanone was isolated, as the enolate
resulting from the dibenzylzinc conjugate addition was
trapped by the in situ generated methyl bromide (33). De-
spite the generation of electrophilic methyl bromide during
the synthesis of dibenzylzinc, these conditions were used
It was decided to convert the 2,7-anhydro-4-alkyl-
sialosides 14 and 15 into glycals so as to develop a route to
give compounds that could be tested as sialidase inhibitors
against benchmark compounds such as DANA (2). This was
accomplished by stirring the crude 2,7-anhydrosugars in ace-
tic anhydride/acetic acid with a catalytic amount of H2SO4
followed by stirring in saturated aq. NaHCO3, a reaction se-
quence that opened the 2,7-anhydro linkage, peracetylated
all hydroxyl groups, and eliminated HOAc in one pot (35).
© 2008 NRC Canada

Hemeon and Bennet
243
Scheme 3. Conversion of 4-alkyl hemiaminals 11 and 13 into glycals 16 and 17 via their 2,7-anhydrosugars 14 and 15.
R
NHAc
O
O
HO
CF3CO2H (aq)
RT, 1 d
R
CO2Et
N
O
EtO2C
O
HO
O
Ac
O
OH
11 R=Et
13 R=Me
14 R=Et
15 R=Me
i) Ac2O, AcOH
cat. H2SO4, RT, 2 d
ii) NaHCO3 (aq), 2 h
OH
OAc
HO
i) 17, NaOEt, EtOH,
RT, 30 min
AcO
O
CO2H
O
CO2Et
HO
AcO
ii) LiOH, 3:2 THF/H2O
oC, 40 min
AcHN
AcHN
0
CH3
R
16 R=Et, 47% (2 steps)
17 R=Me, 59% (2 steps)
18, 82%
This afforded the peracetylated glycals 16 and 17, the 4-
ethyl-DANA analogue 16 being obtained in 47% yield over
two steps from 4-ethyl hemiaminal 11, and the 4-methyl-
DANA analogue being obtained in 59% yield over two steps
from 4-methyl hemiaminal 13 (Scheme 3). As with the 2,7-
anhydrosialoside precursors, these glycals could be purified
by silica chromatography, but their yields decreased signifi-
cantly despite the crude product mixtures being composed
mainly of the glycal products. It was attempted to convert
the 4-methyl hemiaminal 13 directly to glycal 17 by per-
forming the deprotection, acetylation, and elimination reac-
tions in one pot, but this did not result in the isolation of any
useful products.
16 and 17. Given the functional-group-tolerant nature of
dialkylzinc additions, this synthetic route should allow the
synthesis of a number of 4-modified sialic acid analogues,
including 2,7-anhydrosugars and DANA analogues. Unfor-
tunately, these conjugate addition reactions exclusively af-
forded the less desirable 4R-diastereomer. As well, addition
of small nucleophiles from commercially available diethyl-
zinc and dimethylzinc proceeded relatively smoothly, while
the addition of bulkier dibenzylzinc and diphenylzinc did
not proceed at all. Current work is focusing on attempts to
change the diastereoselectivity of the conjugate addition to
afford the more desirable 4S-diastereomer and on the search
for nucleophiles that will participate in conjugate addition
reactions with enone 5 to afford more interesting 4-modified
sialic acid analogues.
The deprotection of 4-deoxy-4-epi-4-methyl-DANA 17
was achieved first by removal of the acetyl groups using
NaOEt in ethanol, followed by hydrolysis of the ethyl ester
using LiOH in 3:2 THF/water to afford the deprotected 4-
deoxy-4-epi-4-methyl glycal of sialic acid 18 in 82% yield
over the two deprotection steps (Scheme 3). A series of 1D
Experimental
Thin-layer chromatography (TLC) was performed on
aluminum-backed TLC plates pre-coated with Merck silica
gel 60 F254. Compounds were visualized with UV light and
(or) staining with phosphomolybdic acid (5% solution in
EtOH). Flash chromatography was performed using Avanco
silica gel 60 (230–400 mesh). Melting points were recorded
on a Gallenkamp melting point apparatus and are uncor-
rected. Solvents used for anhydrous reactions were dried and
distilled immediately prior to use. Ethanol was dried and
distilled over magnesium ethoxide. Dichloromethane and to-
luene were dried and distilled over calcium hydride. Toluene
was degassed by repeated freezing, then thawing while un-
der vacuum. Glassware for anhydrous reactions was flame-
dried and cooled under a nitrogen atmosphere immediately
prior to use. The majority of the NMR spectra were recorded
on a Varian Unity 500 MHz spectrometer, while some were
recorded on Bruker AMX-400 MHz or Bruker TCI-
600 MHz spectrometers. Chemical shifts (δ) are listed in
ppm downfield from TMS using the residual solvent peak as
an internal reference. 1H and 13C NMR peak assignments are
1
NOE difference H NMR spectra again confirmed the 4R-
stereochemical assignment, as irradiation of the 4-methyl
group showed a strong contact to H-6 and a weak contact to
H-5, indicating that the 4-methyl group and H-6 were on the
same face of the six-membered ring. As well, H-4 showed a
strong contact to H-5 and virtually no contact to H-6, indi-
cating that H-4 and H-5 are on the same face.
Conclusions
Isolation of an unexpected β,γ-unsaturated α-keto ester 5
led to the development of a synthetic route leading to 4-
substituted sialic acid analogues. Ozonolysis of the 4S
enoate ester 6 led to enone 5 following elimination of HNO2
during silica chromatography or upon treatment of the crude
ozonolysis product with DBU, while ozonolysis of the 4R
enoate ester 7 led mainly to a five-membered ring hemi-
aminal structure 9. Conditions were developed to add alkyl
groups to enone 5 through copper-catalyzed conjugate addi-
tions of diethylzinc and dimethylzinc. The resulting 4-alkyl-
substituted conjugate addition products could be deprotected
to generate 2,7-anhydro-4-alkylsialosides 14 and 15 that
could then be converted into 4-modified DANA analogues
1
1
made based on H–1H COSY and H–13C HMQC experi-
ments. IR spectra were recorded on a Bomem IR spectrome-
ter, and samples were prepared as cast evaporative films on
NaCl plates from CH2Cl2. Optical rotations were measured
© 2008 NRC Canada

244
Can. J. Chem. Vol. 86, 2008
using a PerkinElmer 341 polarimeter and are reported in
units of deg cm2 g–1 (concentrations reported in units
of g/100 cm3). Shown in the supporting information are the
1H NMR and 13C NMR spectra for compounds 14–18.2
6.2 Hz, J9a,9b = 8.7 Hz, H-9b), 4.34 (q, 2H, J = 7.1 Hz,
CH3CH2O), 4.80–4.85 (m, 1H, H-5), 6.16 (d, 1H, JNH,5
7.9 Hz, NHCOCH3), 6.85 (dd, 1H, J3,4 = 15.9 Hz, J3,5
=
=
4
1.2 Hz, H-3), 7.15 (dd, 1H, J3,4 = 15.9 Hz, J4,5 = 6.4 Hz, H-
4). 13C NMR (CDCl3, 125 MHz) δ: 13.8 (CH3CH2O), 23.0
(NHCOCH3), 24.9, 26.5, 26.7, 27.0 (CMe2 × 2), 52.5 (C-5),
62.3 (CH3CH2O), 67.7 (C-9), 76.8 (C-8), 79.0 (C-7), 81.3
(C-6), 109.8, 110.3 (CMe2 × 2), 126.1 (C-3), 147.9 (C-4),
161.5 (C-1), 169.4 (NHCOCH3), 182.6 (C-2). Anal. calcd.
for C19H29NO8: C 57.13, H 7.32, N 3.51; found: C 56.82, H
7.17, N 3.76.
Ethyl 5-acetamido-3,4,5-trideoxy-6,7:8,9-di-O-
isopropylidene-4-nitro-D-glycero-D-galacto-non-2-
ulosonate (8)
The 4S enoate ester 6 (494 mg, 1.11 mmol) was dissolved
in dry CH2Cl2 (50 mL) and cooled in a dry ice/acetone bath
to –78 °C. The flask was fitted with a CaCl2 drying tube and
ozone was bubbled through the solution until a blue color
persisted (10 min). The solution was purged with oxygen,
and methyl sulfide (1 mL) was added. The colorless solution
was allowed to warm to room temperature while stirring for
1.5 h, after which time it was concentrated under reduced
pressure to yield a colorless syrup (588 mg) that consisted
mainly of DMSO and α-keto ester 8. The crude material was
unstable and could not be fully characterized. IR (cm–1):
1373 (NO2), 1557 (NO2), 1673 (amide C=O), 1733 (α-keto
ester C=O, one band), 3267 (NH). 1H NMR (CDCl3,
500 MHz) δ: 1.35 (s, 3H, CMe2), 1.377 (t, 3H, J = 7.1 Hz,
CH3CH2O), 1.382 (s, 6H, CMe2), 1.43 (s, 3H, CMe2), 2.01
Ethyl N-acetyl-5-amino-3,4,5-trideoxy-6,7:8,9-di-O-
isopropylidene-4-nitro-D-glycero-D-talo-non-2-
ulofuranosonate (9)
The 4R enoate ester 7 (168 mg, 0.377 mmol) was con-
verted to the 4-nitro hemiaminal 9 (colorless syrup, 118 mg,
0.265 mmol, 70%) under the same conditions used to gener-
ate α-keto ester 8 as described earlier. A quantity of enone 5
was also obtained (26.5 mg, 0.066 mmol, 18%). IR (cm–1):
1372 (NO2), 1562 (NO2), 1658 (amide C=O), 1753 (ester
1
C=O), 3492 (OH). H NMR (CDCl3, 500 MHz) δ: 1.25 (t,
3H, J = 7.1 Hz, CH3CH2O), 1.33 (s, 3H, CMe2), 1.35 (s, 3H,
CMe2), 1.39 (s, 3H, CMe2), 1.45 (s, 3H, CMe2), 2.20 (s, 3H,
NHCOCH3), 2.97–2.98 (m, 2H, H-3a, H-3b), 3.58 (t, 1H,
J6,7 = J7,8 = 8.6 Hz, H-7), 3.99 (dd, 1H, J8,9a = 4.5 Hz,
J9a,9b = 8.9 Hz, H-9a), 4.06–4.11 (m, 1H, H-8), 4.14–4.26
(m, 3H, CH3CH2O, H-9b), 4.29 (d, 1H, J6,7 = 8.3 Hz, H-6),
4.81–4.83 (m, 2H, H-4, 2-OH), 5.40 (s, 1H, H-5). 13C NMR
(CDCl3, 125 MHz) δ: 13.9 (CH3CH2O), 22.5 (NHCOCH3),
25.0, 26.5, 26.6, 26.9 (CMe2 × 2), 41.5 (C-3), 62.2 (C-5),
62.3 (CH3CH2O), 68.0 (C-9), 77.3 (C-8), 78.4 (C-7), 81.8
(C-6), 84.1 (C-4), 89.2 (C-2), 110.2, 110.7 (CMe2 × 2),
169.9 (C=O), 171.1 (C=O). Anal. calcd. for C19H30N2O10: C
51.12, H 6.77, N 6.27; found: C 51.24, H 6.81, N 6.01.
(s, 3H, NHCOCH3), 3.46 (dd, 1H, J3a,3b = 19.2 Hz, J3a,4
=
6.8 Hz, H-3a), 3.63 (dd, 1H, J3a,3b = 19.2 Hz, J3b,4 = 7.3 Hz,
H-3b), 3.76 (dd, 1H, J6,7 = 5.2 Hz, J7,8 = 8.8 Hz, H-7), 3.81
(dd, 1H, J8,9a = 6.7 Hz, J9a,9b = 8.7 Hz, H-9a), 3.98 (dt, 1H,
J7,8 = 8.8 Hz, J8,9a + J8,9b = 12.9 Hz, H-8), 4.13 (dd, 1H,
J5,6 = 9.8 Hz, J6,7 = 5.2 Hz, H-6), 4.15 (dd, 1H, J8,9b
=
6.2 Hz, J9a,9b = 8.7 Hz, H-9b), 4.35 (q, 2H, J = 7.1 Hz,
CH3CH2O), 4.43 (dt, 1H, J4,5 = 2.3 Hz, J5,6 + JNH,5
18.5 Hz, H-5), 5.49 (dt, 1H, J3a,4 + J3b,4 = 14.1 Hz, J4,5
=
=
2.3 Hz, H-4), 5.96 (d, 1H, JNH,5 = 8.7 Hz, NHCOCH3). 13C
NMR (CDCl3, 125 MHz) δ: 13.2 (CH3CH2O), 22.2
(NHCOCH3), 24.7, 25.8, 26.4, 26.9 (CMe2 × 2), 38.4 (C-3),
52.3 (C-5), 62.1 (CH3CH2O), 66.4 (C-9), 75.9 (C-8), 77.5
(C-6), 80.5 (C-7), 80.7 (C-4), 109.0, 110.2 (CMe2 × 2),
158.9 (C-1), 170.3 (NHCOCH3), 188.7 (C-2).
Ethyl 5-acetamido-3,4,5-trideoxy-4-ethyl-6,7:8,9-di-O-
isopropylidene-D-glycero-D-talo-non-3-enonate (10)
Copper(II)
trifluoromethanesulfonate
(10
mg,
Ethyl (E)-5-acetamido-3,4,5-trideoxy-6,7:8,9-di-O-
isopropylidene-D-manno-non-3-en-2-ulosonate (5)
0.028 mmol) was suspended in dry, degassed toluene
(7 mL), and triethylphosphite (10 µL, 0.058 mmol) was
added slowly. The resulting colorless solution was stirred
under N2 at room temperature for 15 min, after which it was
cooled to –20 °C, and a solution of diethylzinc in hexanes
(1.0 mol/L, 0.90 mL, 0.90 mmol) was added dropwise. After
stirring under N2 for 10 min, a solution of enone 5 (236 mg,
0.59 mmol) in toluene (3 mL) was injected dropwise. The
solution was stirred under N2 overnight as it warmed to
room temperature. After 18 h, the reaction was quenched by
the addition of saturated aq. NH4Cl (5 mL). The mixture was
diluted with water (15 mL) and extracted with Et2O (2 ×
50 mL). The combined organic layers were dried (MgSO4),
filtered, and concentrated under reduced pressure to give a
mixture of compounds that contained mainly the enol tauto-
mer of the ethyl addition product 10 as colorless syrup
(217 mg). The enol addition product could not be purified
for characterization as it rearranged to the five-membered
The crude α-keto ester 8 from ozonolysis of enoate ester 6
was subjected to flash chromatography (hexanes–EtOAc gra-
dient solvent system from 1:1 v/v to 100% EtOAc), inducing
the elimination reaction to afford enone 5 as a colorless
syrup (403 mg, 1.01 mmol, 91%). The elimination could
also be effected by stirring 8 with 1 equiv. DBU in EtOAc
for 2 min, followed by quenching with saturated aq. ammo-
nium chloride and extraction with Et2O to afford crude
enone 5 that could be purified as described earlier. [α]20
+18.8 (c 1.10, CHCl3). IR (cm–1): 1073 (aliphatic ether CD–
O–C), 1654 (amide C=O), 1733 (α-keto ester C=O, one
1
band), 3292 (NH). H NMR (CDCl3, 500 MHz) δ: 1.36–
1.38 (m, 12H, CH3CH2O, CMe2 × 2), 1.44 (s, 3H, CMe2),
2.04 (s, 3H, NHCOCH3), 3.70 (dd, 1H, J6,7 = 7.2 Hz, J7,8
8.7 Hz, H-7), 3.89 (dd, 1H, J8,9a = 5.8 Hz, J9a,9b = 8.7 Hz, H-
9a), 3.99–4.04 (m, 2H, H-6, H-8), 4.17 (dd, 1H, J8,9b
=
=
2 Supplementary data for this article are available on the journal Web site (canjchem.nrc.ca) or may be purchased from the Depository of
Unpublished Data, Document Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0R6, Canada. DUD 3711. For more
information on obtaining material, refer to cisti-icist.nrc-cnrc.gc.ca/irm/unpub_e.shtml.
© 2008 NRC Canada

Hemeon and Bennet
245
ring hemiaminal 11 upon chromatography. 1H NMR
(CDCl3, 500 MHz) δ: 0.86 (t, 3H, J = 7.3 Hz, CH2CH3),
1.29–1.41 (m, 16H, CMe2 × 2, OCH2CH3, CHaHbCH3),
1.51–1.58 (m, 1H, CHaCHbCH3), 1.95 (s, 3H, COCH3),
2.82–2.88 (m, 1H, H-4), 3.78–3.83 (m, 3H, H-6, H-7, H-9a),
3.89–3.95 (m, 1H, H-8), 4.06–4.09 (m, 1H, H-9b), 4.23–
4.28 (m, 3H, OCH2CH3, H-5), 5.50 (d, 1H, J3,4 = 10.2 Hz,
H-3), 5.73 (d, 1H, J5,NH = 9.4 Hz, NH), 6.54 (s, 1H, OH).
13C NMR (CDCl3, 125 MHz) δ: 12.0 (CH2CH3), 14.1
(OCH2CH3), 23.4 (COCH3), 24.6 (CH2CH3), 25.3, 26.5,
27.2, 27.5 (C(CH3)2 × 2), 39.1 (C-4), 54.1 (OCH2CH3/C-5),
61.8 (OCH2CH3/C-5), 67.8 (C-9), 77.0 (C-8), 80.5 (C-6/C-
7), 80.7 (C-6/C-7), 109.6, 110.3 (C(CH3)2 × 2), 113.1 (C-3),
142.3 (C-2), 164.4, 170.1 (C=O × 2).
(CDCl3, 125 MHz) δ: 14.0 (CH3CH2O), 21.3 (4-CH3), 22.6
(COCH3), 25.1, 26.8 (× 2), 27.0 (C(CH3)2 × 2), 33.0 (C-4),
43.0 (C-3), 61.8 (CH3CH2O), 67.6 (C-5), 68.2 (C-9), 77.5
(C-8), 78.8 (C-7), 83.0 (C-6), 90.8 (C-2), 109.8, 110.1
(C(CH3)2 × 2), 171.4, 171.7 (C=O × 2). Anal. calcd. for
C20H33NO8: C 57.82, H 8.01, N 3.37; found: C 57.52, H
7.90, N 3.59.
Minor anomer
1H NMR (CDCl3, 500 MHz) δ: 1.23 (d, 3H, JCH3,4
=
7.5 Hz, 4-CH3), 1.30 (t, 3H, J = 7.1 Hz, CH3CH2O), 1.34,
1.37, 1.42, 1.44 (s, 12H, C(CH3)2 × 2), 1.97 (d, 1H, J3a,3b
=
12.5 Hz, H-3a), 2.14 (s, 3H, COCH3), 2.50–2.60 (m, 2H, H-
3b, H-4), 3.57 (t, 1H, J6,7 + J7,8 = 17.0 Hz, H-7), 3.95 (s, 1H,
H-5), 3.98 (dd, 1H, J8,9a = 4.9 Hz, J9a,9b = 8.8 Hz, H-9a),
4.04–4.08 (m, 1H, H-8), 4.17–4.30 (m, 4H, H-6, H-9b,
CH3CH2O), 4.47 (s, 1H, OH). 13C NMR (CDCl3, 125 MHz)
δ: 14.0 (CH3CH2O), 21.4 (4-CH3), 22.1 (COCH3), 25.0,
26.8, 26.90, 26.93 (C(CH3)2 × 2), 31.9 (C-4), 45.0 (C-3),
62.0 (CH3CH2O), 67.6 (C-5), 68.1 (C-9), 77.6 (C-8), 78.8
(C-7), 83.1 (C-6), 90.0 (C-2), 109.8, 110.6 (C(CH3)2 × 2),
170.5 (COCH3), 171.8 (C-1).
Ethyl N-acetyl-5-amino-3,4,5-trideoxy-4-ethyl-6,7:8,9-di-
O-isopropylidene-D-glycero-D-talo-non-2-ulofuranosonate
(11)
The crude enol 10 was purified via flash chromatography
(hexanes–EtOAc gradient solvent system from 2:1 to 1:2
v/v) to afford the hemiaminal 11 as a colorless syrup
(134 mg, 0.31 mmol, 53%). IR (cm–1): 1648 (amide C=O),
1
1747 (ester C=O), 3490 (OH). H NMR (CDCl3, 500 MHz)
Ethyl 5-acetamido-2,7-anhydro-3,4,5-trideoxy-4-ethyl-D-
glycero-D-talo-non-2-ulopyranosonate (14)
δ: 0.97 (t, 3H, J = 7.4 Hz, CH2CH3), 1.25 (t, 3H, J = 7.1 Hz
OCH2CH3), 1.34, 1.35, 1.40, 1.41 (s, 12H, C(CH3)2 × 2),
1.65–1.71 (m, 2H, CH2CH3), 1.89 (d, 1H, J3a,3b = 13.2 Hz,
H-3a), 2.14–2.19 (m, 4H, COCH3, H-4), 2.62 (ddd, 1H,
J3a,3b = 13.2 Hz, J3b,4 = 7.6 Hz, J3b,OH = 1.6 Hz, H-3b), 3.47
4-Ethyl hemiaminal 11 (239 mg, 0.557 mmol) was dis-
solved in trifluoroacetic acid (4.0 mL) to which water
(0.4 mL) was added. The resulting solution was stirred at
room temperature for 15 h and was then concentrated under
reduced pressure to give a dark brown syrup (249 mg) that
was used directly in the next reaction. A sample for charac-
terization could be purified by flash chromatography
(EtOAc–MeOH–H2O gradient solvent system from 20:3:1 to
10:3:1 v/v/v), giving the 2,7-anhydro-4-ethylsialoside 14 as a
light brown foamy syrup. IR (cm–1): 1645 (amide C=O),
(t, 1H, J6,7 = J7,8 = 8.5 Hz, H-7), 3.93 (dd, 1H, J8,9a
5.8 Hz, J9a,9b = 8.7 Hz, H-9a), 4.04 (dt, 1H, J7,8 + J8,9a
=
+
J8,9b = 20.8 Hz, H-8), 4.12 (s, 1H, H-5), 4.14–4.26 (m, 4H,
H-9b, H-6, OCH2CH3), 4.87 (d, 1H, J3b.OH = 1.5 Hz, OH).
13C NMR (CDCl3, 100 MHz) δ: 12.6 (CH2CH3), 14.0
(OCH2CH3), 22.5 (COCH3), 25.1, 26.6, 26.8, 27.0 (C(CH3)2 ×
2), 27.3 (CH2CH3), 40.5 (C-4), 41.1 (C-3), 61.7 (OCH2CH3),
65.3 (C-5), 68.5 (C-9), 77.5 (C-8), 78.9 (C-7), 83.4 (C-6),
90.6 (C-2), 109.8, 110.2 (C(CH3)2 × 2), 171.4, 171.8 (C=O
× 2).
1
1746 (ester C=O), 3328 (OH). H NMR (D2O, 600 MHz) δ:
0.82 (t, 3H, J = 7.4 Hz, CH2CH3), 1.24–1.33 (m, 5 H,
CH2CH3, OCH2CH3), 1.65 (t, 1H, J3a,3b = J3a,4 = 13.1 Hz, H-
3a), 2.02–2.06 (m, 4H, H-3b, COCH3), 2.12–2.17 (m, 1H,
H-4), 3.52–3.58 (m, 2H, H-8, H-9a), 3.71 (br d, 1H, J9a,9b
=
Ethyl N-acetyl-5-amino-3,4,5-trideoxy-6,7:8,9-di-O-
isopropylidene-4-methyl-D-glycero-D-talo-non-2-
ulofuranosonate (13)
Enone 5 (1.02 g, 2.56 mmol) was reacted with dimethyl-
zinc (2.0 mol/L in toluene, 1.9 mL, 3.8 mmol) to generate
the 4-methyl hemiaminal 13 (1.4:1.0 anomeric ratio,
373 mg, 0.898 mmol, 35%) using the methods reported ear-
lier for the synthesis of the ethyl-addition product 10. The
major anomer of 13 was purified to analytical standards for
complete characterization.
11.5 Hz, H-9b), 4.05 (dd, 1H, J4,5 = 2.0 Hz, J5,6 = 4.7 Hz, H-
5), 4.21 (d, 1H, J7,8 = 7.4 Hz, H-7), 4.29 (q, 2H, J = 7.1 Hz,
OCH2CH3), 4.59 (d, 1H, J5,6 = 1.9 Hz, H-6). 13C NMR
(D2O, 150 MHz) δ: 10.2 (CH2CH3), 13.1 (OCH2CH3), 21.7
(COCH3), 23.5 (CH2CH3), 32.3 (C-4), 34.3 (C-3), 47.9 (C-
5), 61.9 (C-9), 63.5 (OCH2CH3), 71.3 (C-8), 77.9 (C-7),
79.9 (C-6), 105.0 (C-2), 168.2 (C-1), 173.9 (COCH3). HR–
MS (ESI) for C15H25NO7: (MNa+) calcd: 354.1528; found:
354.1532.
Major anomer
Ethyl 5-acetamido-2,7-anhydro-3,4,5-trideoxy-4-methyl-
D-glycero-D-talo-non-2-ulopyranosonate (15)
[α]20 +16.5 (c 1.95, CHCl3). IR (cm–1): 1648 (amide
D
C=O), 1746 (ester C=O), 3482 (OH). 1H NMR (CDCl3,
500 MHz) δ: 1.25 (t, 3H, J = 7.1 Hz, CH3CH2O), 1.32 (d,
3H, JCH3,4 = 7.3 Hz, 4-CH3), 1.33, 1.34, 1.39, 1.42 (s, 12H,
C(CH3)2 × 2), 1.78 (d, 1H, J3a,3b = 13.1 Hz, H-3a), 2.16 (s,
4-Methyl hemiaminal 13 (228 mg, 0.548 mmol) was con-
verted into the 2,7-anhydro-4-methylsialoside 15 (263 mg)
using the method described earlier for synthesis of the ethyl
analogue 14. The light brown foamy syrup was of adequate
purity to be used directly in the next reaction. An analytical
sample was purified via flash chromatography (20:3:1
EtOAc–methanol–water v/v/v) to afford the product as a tan
3H, COCH3), 2.44–2.53 (m, 1H, H-4), 2.66 (dd, 1H, J3a,3b
13.0 Hz, J3b,4 = 7.6 Hz, H-3b), 3.49 (t, 1H, J6,7 + J7,8
17.2 Hz, H-7), 3.99 (dd, 1H, J8,9a = 4.9 Hz, J9a,9b = 8.7 Hz,
H-9a), 4.03 (s, 1H, H-5), 4.03–4.07 (m, 1H, H-8), 4.12–4.28
(m, 4H, H-6, H-9b, CH3CH2O), 4.91 (s, 1H, OH). 13C NMR
=
=
solid; mp 117–120 °C (dec). [α]20 +48 (c 0.13, H2O). IR
D
(cm–1): 1647 (amide C=O), 1741 (ester C=O), 3333 (OH).
© 2008 NRC Canada

246
Can. J. Chem. Vol. 86, 2008
1H NMR (D2O, 500 MHz) δ: 0.87 (d, 3H, JMe,4 = 6.8 Hz, 4-
CH3), 1.28 (t, 3H, J = 7.1 Hz, CH3CH2O), 1.65 (dd, 1H,
[α]20D +1.2 (c 0.995, CHCl3). IR (cm–1): 1653 (amide C=O),
1
1683 (trisubstituted C=C), 1748 (ester C=O), 3291 (NH). H
J3a,3b = 14.0 Hz, J3a,4 = 12.3 Hz, H-3a), 1.97 (dd, 1H, J3a,3b
=
NMR (CDCl3) δ: 1.08 (d, 3H, JMe,4 = 7.1 Hz, 4-CH3), 1.30
(t, 3H, J = 7.1 Hz, CH3CH2O), 1.98 (s, 3H, Ac), 2.05 (s, 3H,
Ac), 2.06 (s, 3H, Ac), 2.10 (s, 3H, Ac), 2.77–2.83 (m, 1H,
H-4), 4.20 (dd, 1H, J5,6 = 5.0 Hz, J6,7 = 7.1 Hz, H-6), 4.20–
4.27 (m, 2H, CH3CH2O), 4.28 (dd, 1H, J8,9a = 7.1 Hz,
14.1 Hz, J3b,4 = 5.6 Hz, H-3b), 2.06 (s, 3H, COCH3), 2.34–
2.43 (m, 1H, H-4), 3.52–3.59 (m, 2H, H-8, H-9a), 3.71 (dd,
1H, J8,9b = 2.4 Hz, J9a,9b = 11.5 Hz, H-9b), 3.99 (dd, 1H,
J4,5 = 4.7 Hz, J5,6 = 2.0 Hz, H-5), 4.22 (d, 1H, J7,8 = 7.4 Hz,
H-7), 4.29 (q, 2H, J = 7.1 Hz, CH3CH2O), 4.60 (d, 1H,
J5,6 = 1.8 Hz, H-6). 13C NMR (D2O, 125 MHz) δ: 13.2
(CH3CH2O), 15.5 (4-CH3), 21.9 (COCH3), 26.0 (C-4), 36.0
(C-3), 49.7 (C-5), 62.1 (C-9), 63.7 (CH3CH2O), 71.5 (C-8),
78.2 (C-7), 80.0 (C-6), 105.1 (C-2), 168.4 (C-1), 174.2
(COCH3). HR–MS (ESI) for C14H23NO7: (MH+) calcd:
318.1552; found: 318.1564.
J9a,9b = 12.1 Hz, H-9a), 4.48 (dd, 1H, J8,9b = 4.3 Hz, J9a,9b
12.0 Hz, H-9b), 4.48–4.51 (m, 1H, H-5), 5.21 (br quintet,
1H, J7,8 + J8,9a + J8,9b = 14.5 Hz, H-8), 5.43 (dd, 1H, J6,7
=
=
7.1 Hz, J7,8 = 3.3 Hz, H-7), 5.55 (d, 1H, JNH,5 = 9.6 Hz,
NHCOCH3), 5.96 (d, 1H, J3,4 = 3.2 Hz, H-3). 13C NMR
(CDCl3) δ: 14.1 (CH3CH2O), 15.6 (4-CH3), 20.61, 20.67,
20.8, 23.2 (COCH3 × 4), 27.8 (C-4), 45.1 (C-5), 61.36
(CH3CH2O), 61.38 (C-9), 68.6 (C-7), 70.0 (C-8), 75.8 (C-6),
114.7 (C-3), 141.3 (C-2), 161.8, 169.7, 170.0, 170.3, 170.5
(C=O × 5). HR–MS (ESI) for C20H29NO10: (MH+) calcd:
444.1869; found: 444.1870.
Ethyl 5-acetamido-7,8,9-tri-O-acetyl-2,6-anhydro-3,4,5-
trideoxy-4-ethyl-D-glycero-D-talo-non-2-enonate (16)
The crude preparation of 2,7-anhydro-4-ethylsialoside 14
from the previous reaction (249 mg) was dissolved in acetic
anhydride (2.0 mL) and glacial acetic acid (2.0 mL). Con-
centrated H2SO4 (15 drops) was added, and the resulting so-
lution was stirred at room temperature. After 3 d, the
reaction was poured into saturated aq. NaHCO3 (100 mL),
which bubbled vigorously, and the resulting basic mixture
was stirred for 2 h. The mixture was then extracted with
EtOAc (2 × 100 mL), and the combined organic layers were
dried (MgSO4), filtered, and concentrated under reduced
pressure to give the 4-ethyl-substituted glycal 16 as a light
brown syrup (119 mg, 0.261 mmol, 47%). The glycal could
be further purified via flash chromatography (hexanes–
EtOAc gradient solvent system from 1:2 v/v to 100%
5-Acetamido-2,6-anhydro-3,4,5-trideoxy-4-methyl-D-
glycero-D-talo-non-2-enonic acid (18)
A 1.2 mol/L solution of sodium ethoxide in ethanol
(0.080 mL, 0.096 mmol) was added to a solution of per-
acetylated 4-methyl-substituted glycal 17 (29 mg,
0.065 mmol) in dry ethanol (3.0 mL). This solution was
stirred at room temperature for 30 min, after which time
TLC analysis indicated the absence of starting material. The
reaction was neutralized with Amberlite™ H+ resin and fil-
tered. The resin was washed with methanol (30 mL), and the
filtrate was concentrated under reduced pressure to give a
light yellow syrup (21.1 mg, 0.065 mmol). The de-acetylated
product was then dissolved in THF/water (3:2 v/v, 5 mL) and
cooled in an ice bath. Lithium hydroxide monohydrate was
added (23.7 mg, 0.565 mmol), and the solution stirred in ice
for 30 min, after which time TLC analysis indicated com-
plete de-esterification. The reaction was removed from the
cooling bath, diluted with water (5 mL), and neutralized
with Amberlite™ H+ resin to pH 5, which was then filtered
and washed with water (30 mL). The combined filtrates
were concentrated under reduced pressure to afford the lith-
ium salt of the deprotected glycal 18 as a colorless syrup
EtOAc) to afford a sample for characterization. [α]20 +3 (c
D
0.24, CHCl3). IR (cm–1): 1656 (amide C=O), 1747 (ester
C=O). 1H NMR (CDCl3, 600 MHz) δ: 1.04 (t, 3H, J =
7.4 Hz, CH2CH3), 1.31 (t, 3H, J = 7.1 Hz, OCH2CH3), 1.34–
1.39 (m, 1H, CHaHbCH3), 1.48–1.56 (m, 1H, CHaHbCH3),
1.97 (s, 3H, COCH3), 2.06 (s, 3H, COCH3), 2.07 (s, 3H,
COCH3), 2.11 (s, 3H, COCH3), 2.56–2.60 (m, 1H, H-4),
4.20–4.31 (m, 4H, H-6, H-9a, OCH2CH3), 4.40 (dd, 1H,
J8,9b = 4.4 Hz, J9a,9b = 12.0 Hz, H-9b), 4.55–4.58 (m, 1H, H-
5), 5.19 (ddd, 1H, J7,8 + J8,9b = 11.4 Hz, J8,9a = 3.2 Hz, H-8),
5.40 (dd, 1H, J6,7 = 8.5 Hz, J7,8 = 3.1 Hz, H-7), 5.54 (d, 1H,
J5,NH = 9.4 Hz, NH), 6.03 (br d, 1H, J3,4 = 2.6 Hz, H-3). 13C
NMR (CDCl3, 150 MHz) δ: 11.4 (CH2CH3), 14.1
(OCH2CH3), 20.66 (COCH3), 20.72 (COCH3), 20.9 (COCH3),
23.2 (CH2CH3), 23.3 (COCH3), 33.9 (C-4), 43.7 (C-5), 61.1
(C-9), 61.5 (OCH2CH3), 68.7 (C-7), 69.3 (C-8), 76.7 (C-6),
112.8 (C-3), 141.1 (C-2), 161.8, 169.7, 170.0, 170.3, 170.6
(C=O × 5). HR–MS (ESI) for C21H31NO10: (MH+) calcd:
458.2026; found: 458.2020.
(16.0 mg, 0.0542 mmol, 82%). [α]20 –100 (c 0.48, H2O).
D
IR (cm–1): 1645 (amide C=O), 1712 (acid C=O), 3287 (NH),
1
3340 (OH). H NMR (D2O) δ: 1.04 (d, 3H, JMe,4 = 7.2 Hz,
4-CH3), 2.02 (s, 3H, NHCOCH3), 2.70–2.76 (m, 1H, H-4),
3.64 (dd, 1H, J8,9a = 6.3 Hz, J9a,9b = 11.9 Hz, H-9a), 3.67
(dd, 1H, J6,7 = 2.2 Hz, J7,8 = 8.8 Hz, H-7), 3.86 (dd, 1H,
J8,9b = 2.8 Hz, J9a,9b = 11.9 Hz, H-9b), 3.90 (ddd, 1H, J7,8
8.9 Hz, J8,9a = 6.3 Hz, J8,9b = 2.8 Hz, H-8), 4.20 (dd, 1H,
J5,6 = 8.8 Hz, J6,7 = 2.2 Hz, H-6), 4.30 (dd, 1H, J4,5
=
=
5.8 Hz, J5,6 = 8.8 Hz, H-5), 6.14 (d, 1H, J3,4 = 5.0 Hz, H-3).
13C NMR (D2O) δ: 15.1 (4-CH3), 21.9 (NHCOCH3), 28.9
(C-4), 46.4 (C-5), 63.0 (C-9), 69.0 (C-7), 70.5 (C-8), 72.6
(C-6), 116.7 (C-3), 141.8 (C-2), 166.6 (C=O), 174.1 (C=O).
HR–MS (ESI) for C12H18NO7: (M–) calcd: 288.1083; found:
288.1091.
Ethyl 5-acetamido-7,8,9-tri-O-acetyl-2,6-anhydro-3,4,5-
trideoxy-4-methyl-D-glycero-D-talo-non-2-enonate (17)
The crude 2,7-anhydro-4-methylsialoside 15 from the pre-
vious reaction (263 mg) was converted into glycal 17 (light
brown syrup, 0.144 mg, 0.324 mmol, 59% over 2 steps) us-
ing the same conditions described earlier for synthesis of the
ethyl analogue 16. This product could be further purified via
flash chromatography (hexanes–EtOAc gradient solvent sys-
tem from 1:2 v/v to 100% EtOAc) giving the pure glycal as a
colorless syrup (58 mg, 0.13 mmol, 24% over 2 steps).
Acknowledgements
The authors gratefully acknowledge the Natural Sciences
and Engineering Research Council of Canada (NSERC) and
© 2008 NRC Canada

Hemeon and Bennet
247
16. G.B. Kok and M. Von Itzstein. J. Chem. Soc., Perkin Trans. 1,
905 (1998).
17. I. Hemeon and A.J. Bennet. Org. Biomol. Chem. 4, 2986 (2006).
18. K.G. Liu, S. Yan, Y.L. Wu, and Z.J. Yao. Org. Lett. 6, 2269
(2004).
Simon Fraser University for financial support of this work.
IH thanks the Michael Smith Foundation for Health Re-
search for a Trainee Scholarship. The authors also thank the
Saskatchewan Structural Sciences Centre for measuring HR-
MS data in a timely manner.
19. F. Baumberger and A. Vasella. Helv. Chim. Acta, 69, 1535
(1986).
References
20. T.K.M. Shing. Tetrahedron, 48, 6777 (1992).
21. Z. Pakulski and A. Zamojski. Tetrahedron, 53, 3723 (1997).
22. B. H. Lipshutz. In Organometallics in synthesis: a manual. 2nd
ed. Edited by M. Schlosser. J. Wiley Sons, Chichester, NY.
2004. p. 665.
23. H. Urabe, K. Kobayashi, and F. Sato. J. Chem. Soc., Chem.
Commun. 1043 (1995).
24. H. Urabe, K. Yamashita, K. Suzuki, K. Kobayashi, and F. Sato.
1. R. Schauer. Glycoconjugate J. 17, 485 (2000).
2. G. Taylor. Curr. Opin. Struct. Biol. 6, 830 (1996).
3. R. Thomson and M. von Itzstein. In Carbohydrate-based drug
discovery. Vol. 2. Edited by C.-H. Wong. Wiley-VCH,
Weinheim, Germany. 2003. p. 831.
4. I. Hemeon and A.J. Bennet. Synthesis, 1899 (2007).
5. M. von Itzstein, W.Y. Wu, G.B. Kok, M.S. Pegg, J.C. Dyason,
B. Jin, T.V. Phan, M.L. Smythe, H.F. White, S.W. Oliver, P.M.
Colman, J.N. Varghese, D.M. Ryan, J.M. Woods, R.C. Bethell,
V.J. Hotham, J.M. Cameron, and C.R. Penn. Nature (London),
363, 418 (1993).
J. Org. Chem. 60, 3576 (1995).
25. E. Nakamura. In Organometallics in synthesis: a manual. 2nd
ed. Edited by M. Schlosser. J. Wiley Sons, Chichester, NY.
2004. p. 579.
26. A. Alexakis, J. Vastra, and P. Mangeney. Tetrahedron Lett. 38,
7745 (1997).
6. C.U. Kim, W. Lew, M.A. Williams, H. Liu, L. Zhang, S.
Swaminathan, N. Bischofberger, M. S. Chen, D.B. Mendel,
C.Y. Tai, W.G. Laver, and R.C. Stevens. J. Am. Chem. Soc.
119, 681 (1997).
27. A. Alexakis, C. Benhaim, S. Rosset, M. Humam. J. Am.
Chem. Soc. 124, 5262 (2002).
28. R. Schauer, S. Stoll, E. Zbiral, E. Schreiner, H.H. Brandstetter,
A. Vasella, and F. Baumberger. Glycoconjugate J. 4, 361
(1987).
7. M.-J. Kim, W.J. Hennen, H.M. Sweers, C.-H. Wong. J. Am.
Chem. Soc. 110, 6481 (1988).
8. T.-H. Chan and M.-C. Lee. J. Org. Chem. 60, 4228 (1995).
9. S. Vorwerk and A. Vasella. Angew. Chem., Int. Ed. Eng. 37,
1732 (1998).
10. I.A. Kozlov, S. Mao, Y. Xu, X. Huang, L. Lee, P.S. Sears, C.
Gao, A.R. Coyle, K.D. Janda, and C.-H. Wong.
ChemBioChem, 2, 741 (2001).
11. M. Hartmann and E. Zbiral. Monatsh. Chem. 120, 899 (1989).
12. M. Hartmann, R. Christian, and E. Zbiral. Monatsh. Chem.
122, 111 (1991).
29. H. Mizutani, S.J. Degrado, A.H. Hoveyda. J. Am. Chem. Soc.
124, 779 (2002).
30. P. Knochel, M.C.P. Yeh, S.C. Berk, and J. Talbert. J. Org.
Chem. 53, 2390 (1988).
31. S.C. Berk, M.C.P. Yeh, N. Jeong, and P. Knochel.
Organometallics, 9, 3053 (1990).
32. P. Knochel and R.D. Singer. Chem. Rev. 93, 2117 (1993).
33. X. Rathgeb, S. March, and A. Alexakis. J. Org. Chem. 71,
5737 (2006).
13. D.R. Groves and M. von Itzstein. J. Chem. Soc., Perkin Trans.
34. K.G. Liu, H.B. Zhou, Y.L. Wu, and Z.L. Yao. J. Org. Chem.
1, 2817 (1996).
68, 9528 (2003).
35. G.B. Kok, D. Groves, and M. Von Itzstein. J. Chem. Soc.,
Perkin Trans. 1, 2109 (1999).
14. E. Schreiner, E. Zbiral, R.G. Kleineidam, and R. Schauer.
Liebigs Ann. Chem. 129 (1991).
15. G.B. Kok, A.K. Norton, and M. Von Itzstein. Synthesis, 1185
(1997).
© 2008 NRC Canada