Scalable Synthesis of Anomerically Pure Orthogonal-Protected GlcN3 and GalN3 from d -Glucosamine

Article abstract of DOI:10.1021/acs.orglett.6b02241

An improved and scalable synthesis of orthogonally protected d-glucosamine and d-galactosamine building blocks from inexpensive d-glucosamine has been developed. The key reaction is an inversion/migration step providing access to a fully orthogonal protec

Full text of DOI:10.1021/acs.orglett.6b02241

Letter
pubs.acs.org/OrgLett
Scalable Synthesis of Anomerically Pure Orthogonal-Protected GlcN3
and GalN3 from D‑Glucosamine
Emil Glibstrup and Christian Marcus Pedersen*
Department of Chemistry, University of Copenhagen, Universitetsparken 5, 2100 Copenhagen Ø, Denmark
S
* Supporting Information
ABSTRACT: An improved and scalable synthesis of orthogonally protected D-glucosamine and D-galactosamine building blocks
from inexpensive D-glucosamine has been developed. The key reaction is an inversion/migration step providing access to a fully
orthogonal protecting group pattern, which is required for microbial oligosaccharide synthesis. The method can be carried out on
a multigram scale as several of the reactions can be purified by crystallization to give anomerically pure products.
nterest in the synthesis of microbial glycoconjugates and
oligosaccharides has increased dramatically during the past
have earlier been used as a precursor for 2-azido glucose
derivatives,8 wherein a nucleophilic displacement of a 2-OTf with
the azide ion yields the protected glucosamine derivative
(Scheme 1a). The starting material for this route is, however,
I
decade.1 Besides the common monosaccharide building blocks,
these natural products also contain rare carbohydrates, which are
less available and require elaborate synthesis. The D-glucosamine
and D-galactosamine moieties are found throughout glycospace,2
and easy access to substantial quantities of these building blocks
is therefore of great importance.3 Several methods have been
developed for the synthesis of galactosamine derivatives from
cheaper starting materials. An early example was the azidonitra-
tion protocol on the corresponding D-galactal by Lemieux and
Ratcliffe in 1979, which has the disadvantage of giving a mixture
of 1,2 isomers.4 This area has recently been reviewed.5 Several
methods have been refined in terms of orthogonal protective
group patterns for their use in synthesis of microbial
oligosaccharides. Despite these achievements, the synthesis of
D-galactosamine derivates remains a bottleneck and suffers from
lengthy reaction sequences, harsh reaction conditions, modest
yields, and anomeric mixtures, which cannot effectively be
separated by chromatography. As part of our ongoing work
toward the total synthesis of teichoic acids, easy access to
multiple grams of orthogonally protected D-galactosamine
derivatives was required. The requirements for the building
blocks were orthogonal access to the 3-, 4-, or 6-OH and, having
an anomeric protecting group able to tolerate the protection
group manipulations, required in a multistep synthesis. A
cornerstone in the protective group strategy would be the use
of azide as a protective group for the amine functionality, since
this would offer a unique reactivity; i.e., it would be stable under
acidic and basic conditions but could be selectively reduced
under various conditions. The azido moiety is furthermore not
participating in glycosylations, and hence, both anomers can be
obtained from the same donor depending on the exact reaction
conditions.6 The stability and versatility of the anomeric
protecting group is also important, and the use of thiophenyl
has proved highly efficient in this regard.7 β-Thiomannosides
Scheme 1. Previous Strategies for Obtaining Protected D-
Glucosamine and D-Galactosamine Derivatives
still a four-step procedure from commercially available D-
mannose.9 D-Glucosamine, on the other hand, is an ideal starting
material for this synthesis as it is readily available and inexpensive
in contrast to its C4-epimer D-galactosamine.10
Conversion of glucosamine derivatives to the corresponding
galactosamine has been performed via inversion of the 4-OH:
Using the Lattrell−Dax8c,11 inversion with nitrite substituting a
triflate8c,d,12 or by a nucleophilic displacement with acetate13 or
benzoate14 (Scheme 1b). However, these reactions usually
require elevated temperatures. The guidelines for substitution of
glycoside nonanomeric triflates with external nucleophiles have
Received: July 30, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.6b02241
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
recently been updated.15 A third option is anchimeric assistance
leading to migration of an ester protecting group from the 3-O by
substituting the 4-OTf of a glucopyranoside, causing inversion of
stereochemistry; this is usually seen for pivaloyl esters16 but has
also been observed for the acetyl group17 (Scheme 1c). However,
pivaloyl is not an attractive protective group as it cannot be
removed selectively and under mild conditions. It has only been
observed for the benzoyl protecting group in special cases using
highly reactive 2,6-dideoxy sugars (Scheme 1d).18 With only a
few acetyl examples and no relevant benzoyl examples, the most
suitable ester protecting group for total synthesis purposes, we
were prompted to investigate this migration/inversion reaction
as a means to obtain fully orthogonally protected D-galactos-
amine derivatives. Starting from the inexpensive D-glucosamine−
HCl, the amine functionality was azidated using freshly prepared
TfN3, followed by peracetylation, as described by Yan et al.19
providing 1 in 95% yield (Scheme 2). The subsequent
Scheme 3. Derivatization of the Common Building Block To
Give Fully Orthogonal Protected D-Glucosamines
Scheme 2. Common Building Block Synthesis and
Purification
and 12 was first attempted on the 2-naphthylmethylidene to give
the 2-naphthylmethyl ether (Nap) as an orthogonal 6-O
protecting group, which can selectively be removed using CAN
or DDQ or mildly by cat. HCl in hexafluoroisopropanol
(HFIP).24 Opening of the acetal using BF3·OEt2, instead of the
more common TFA, was done to avoid acid catalyzed loss of the
acetal. The initial results were promising with 84% yield on a 0.1
mmol scale, but upon scale-up the yield dropped to 54% with
35% of the corresponding 4,6-diol isolated as byproduct. Cooling
the reaction to −78 °C and slowly reaching 5 °C over 3−4 h
increased the yield of 10 to 95% on a 1 mmol scale (Scheme 3).
The same procedure was used for the β-anomers. However, they
seemed to perform slightly worse than the α-anomer and gave 11
in 80% yield. The opening of the 4,6-O-benzylidene giving a
more permanent protecting group; i.e., the 6-OBn in 12
proceeded in 78% yield. A simple removal of the acetal
protecting group, using p-toluenesulfonic acid (TsOH·H2O) in
CH2Cl2/MeOH 1:1, gave the 4,6-diol 13 and 14 in 91% and 95%
yield, respectively. These could in turn be selectively protected as
the 6-OTBDPS ethers 15 and 16 in 99% and 97% yield,
respectively (Scheme 3). The stage was now set for investigation
of our proposed inversion/migration sequence (Table 1). The
initial conditions reported in the literature16e (entry 1) gave low
yields, although TLC analysis showed mainly desired product,
alongside a significant baseline spot. The conditions used for the
acetyl migration in the literature17b were also tested (entry 2).
This, however, gave a very complex mixture as judged by TLC,
and no HRMS traces of product could be observed. A reaction
performed solely with pyridine as solvent (entry 3), gave a low
yield as a mixture of the 3-OBz and 4-OBz. Analysis of the crude
reaction mixture by NMR and HRMS suggested the presence of
a pyridinium sugar. The substitution of carbohydrate triflates by
pyridine is known under similar conditions.25 To reduce the
nucleophilicity of the base, pyridine was replaced by 2,6-lutidine
(entry 4), but no conversion of the triflate was observed.
Returning to pyridine but decreasing the volume of pyridine as to
only have a large excess (entries 5−7), was attempted but still
baseline material was detected. Switching pyridine to a smaller
excess of DMAP caused the reaction mixture to become biphasic
upon addition of H2O. Therefore, a solvent change to THF in the
inversion step was considered necessary. The reaction was
performed at a lower temperature in the hope of avoiding the
nucleophilic nature of DMAP (entry 8); this, however, did not
improve the yield. Instead, a mixture of 3- and 4-OBz-protected
substitution with thiophenol, using an excess of BF3·OEt2 (10
equiv) and keeping the reaction time at 2 d, led to a 75% yield of 2
along with some starting material 1 corresponding to an 87%
yield based on recovered starting material (BRSM). After
recovery of starting material and removal of excess thiophenol by
short-plug SiO2 column chromatography, 80% of the α-anomer
could be crystallized from the α/β mixture, allowing for much
simpler purification and characterization in the later stages of the
synthesis. The crystals formed were of such quality that a crystal
structure of the α-anomer could be obtained (Figure S1). The α-
anomer has previously been crystallized, albeit with no
information on the starting α/β ratio and actual quantity
crystallized.20 Zemplen
́
conditions (cat. NaOMe in MeOH)
provided the deacetylated product 3 in 99% yield. With this
common building block in hand, the installation of the
orthogonal protecting groups began. The 4,6-O-benzylidene
was easily installed using benzaldehyde dimethylacetal and
catalytic camphorsulfonic acid (CSA) in MeCN to yield 4 in 94%
yield (Scheme 3). If a mixture of anomers was used, the pure α-
or β-anomer could be crystallized out at this stage. Installing the
2-naphthylmethylidene using the 2-naphthylaldehyde directly
along with VO(OTf)2 as both Lewis acid and hygroscopic
reagent, as described by Chen et al.,21 did not yield more than
15−29% of the desired acetal protected sugar. Preparing the
dimethyl acetal of 2-naphthylaldehyde in 94% yield, according to
a literature procedure,22 allowed the same conditions as the
benzylidene protection, which gave a 86% yield of 5. Anomeric
mixtures of the aryl acetal protected sugars were also possible to
separate by SiO2 column chromatography. To investigate the
migration of the benzoyl group, it was installed at the free 3-OH
using BzCl. Excess BzCl could easily be removed by quenching
the reaction with (N,N-dimethylamino)propylamine (DMAPA)
followed by acidic aqueous workup, as described by Andersen et
al.23 which produced pure 6, 7, 8, and 9 in 96%, 99%, 97%, and
90% yield, respectively, without the need for chromatography
(Scheme 3). Opening the 4,6-acetal to yield the 4-OH 10, 11,
B
DOI: 10.1021/acs.orglett.6b02241
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
a
Table 1. Intramolecular Inversion/Migration of the Benzoyl Protecting Group from C3 to C4
a
Standard conditions: (i) 2 equiv of Tf2O, 10 equiv of Py, concentration: 0.2 M. (ii) Conditions described in the table; for further details, see the
b
c
Supporting Information. A mixture of the 3-OBz and 4-OBz products was obtained. 2,6-lutidine used instead of Py in (i).
GalN3 derivatives, along with HRMS traces of the DMAP
substituted sugar, was obtained. Returning once again to
pyridine, at lower temperature (40 °C) in THF (entry 9−13),
generally increased yields, but reaction times were extended from
overnight to 1.5 d. These results suggested that a nucleophilic
catalyst was necessary when the intramolecular substitution has
to take place with the less nucleophilic benzoyl group, as
suggested by the pKa of the corresponding acids, i.e., 5.03, 4.76,
and 4.2 for the pivalic acid, acetic acid, and benzoic acid,
respectively. The lack of conversion when using the non-
nucleophilic 2,6-lutidine hinted toward a mechanism (Scheme
4), wherein the pyridine assists in forming the tetrahedral
challenge. The 6-OTBDPS, in 19 and 20, further adds to make
this a very bulky glycosyl acceptor. We therefore set out to
perform representative glycosylations resembling the terminal
disaccharide of the teichoic acid from S. pneumonia.27 As donors
we therefore used the known perbenzylated GalN3 trichlor-
acetimidate, 22,28 along with the 6-OTBDPS version, 23,29
which could of course also be synthesized from our building
blocks. The two donors could be α-selectively coupled with both
acceptors, 17 and 20, following literature procedure for similar
donor systems,28 but unoptimized for these more challenging
acceptors. The glycosylations were performed at room temper-
ature to achieve high α-selectivity with catalytic TfOH and 1.2
equiv of acceptor. Coupling donor 22 and 23 with acceptor 20
gave only the α-coupled dissacharide 24 and 25 in 47% and 36%
yield, respectively (Scheme 5). Substantial recovery of the
acceptors, 31% and 50%, was observed. Glycosylations using the
same conditions with acceptor 17 and donor 22 and 23 also
yielded only α-coupled product 26 and 27 in 41% and 46%,
respectively (Scheme 5). Again, recovered acceptor was
observed, this time in 10% and 30% yield, respectively. We
only achieved modest yields using these unoptimized conditions,
but compared to recent literature with similar donors and less
deactivated acceptors these yields and selectivities were
respectable.28,30
Scheme 4. Suggested Mechanism of the Intramolecular
Migration/Inversion Step
intermediate, which then undergoes hydrolysis. The hydrolysis
forms the axial benzoyl ester, as the equatorial alcohol is the
better leaving group.26 One application for these building blocks
would be to use them directly as acceptors in glycosylations.
However, the axial electronically deactivating benzoyl ester along
with the electron-withdrawing azide at C-2 could prove a
In conclusion, we have improved the synthesis of fully
orthogonal protected D-glucosamine derivatives to give 47−59%
overall yield in seven or eight steps starting from inexpensive D-
glucosamine hydrochloride. A major advantage is the possibility
C
DOI: 10.1021/acs.orglett.6b02241
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(8) (a) Pozsgay, V. J. Org. Chem. 1999, 64, 7277−7280. (b) Pozsgay, V.
J. Carbohydr. Chem. 2001, 20, 659−665. (c) Emmadi, M.; Kulkarni, S. S.
J. Org. Chem. 2011, 76, 4703−4709. (d) Emmadi, M.; Kulkarni, S. S. Org.
Biomol. Chem. 2013, 11, 4825−4830.
Scheme 5. Glycosylations Using 17 and 20 as Acceptors
̀
(9) Pedretti, V.; Veyrieres, A.; Sinay, P. Tetrahedron 1990, 46, 77−88.
̈
(10) A price comparison in February 2016 showed galactosamine
hydrochloride to be 27−49 times more expensive.
(11) (a) Lattrell, R.; Lohaus, G. Justus Liebigs Ann. Chem. 1974, 1974,
901−920. (b) Albert, R.; Dax, K.; Link, R. W.; Stutz, A. E. Carbohydr.
̈
Res. 1983, 118, C5−C6.
(12) Hederos, M.; Konradsson, P. J. Carbohydr. Chem. 2005, 24, 297−
320.
(13) Gross, P. H.; Du Bois, F.; Jeanloz, R. W. Carbohydr. Res. 1967, 4,
244−248.
(14) (a) Nashed, M. A.; El-Sokkary, R. I.; Rateb, L. Carbohydr. Res.
1984, 131, 47−52. (b) McGeary, R. P.; Wright, K.; Toth, I. J. Org. Chem.
2001, 66, 5102−5105.
(15) Hale, K. J.; Hough, L.; Manaviazar, S.; Calabrese, A. Org. Lett.
2014, 16, 4838−4841. Hale, K. J.; Hough, L.; Manaviazar, S.; Calabrese,
A. Org. Lett. 2015, 17, 1738−1741.
to purify the products by crystallizations at several stages, and
hence, the procedure is scalable. Furthermore, access to the
corresponding D-galactosamine derivatives was achieved via an
intramolecular migration/inversion using a benzoyl ester as the
migrating group, ensuring orthogonality and sufficient stability
for complex oligosaccharide synthesis. Both are key building
blocks in oligosaccharide synthesis, and we have shown how they
can be used directly as acceptors in glycosylations with α-
selectivity.
(16) (a) Lay, L.; Nicotra, F.; Panza, L.; Russo, G. Helv. Chim. Acta
1994, 77, 509−514. (b) Belot, F.; Jacquinet, J.-C. Carbohydr. Res. 1996,
290, 79−86. (c) Rochepeau-Jobron, L.; Jacquinet, J.-C. Carbohydr. Res.
1997, 305, 181−191. (d) Cai, Y.; Ling, C.-C.; Bundle, D. R. J. Org. Chem.
2009, 74, 580−589. (e) Feng, J.; Ling, C.-C. Carbohydr. Res. 2010, 345,
2450−2457. (f) Rasmussen, M. R.; Marqvorsen, M. H. S.; Kristensen, S.
K.; Jensen, H. H. J. Org. Chem. 2014, 79, 11011−11019.
(17) (a) Takeda, Y.; Horito, S. Carbohydr. Res. 2005, 340, 211−220.
(b) Behera, A.; Emmadi, M.; Kulkarni, S. S. RSC Adv. 2014, 4, 58573−
58580.
ASSOCIATED CONTENT
* Supporting Information
■
S
(18) (a) Binkley, R. W.; Sivik, M. R. J. Org. Chem. 1986, 51, 2619−
2621. (b) Binkley, R. W.; Abdulaziz, M. A. J. Org. Chem. 1987, 52, 4713−
4717. (c) Binkley, R. W. J. Org. Chem. 1991, 56, 3892−3896.
(19) Yan, R.-B.; Yang, F.; Wu, Y.; Zhang, L.-H.; Ye, X.-S. Tetrahedron
Lett. 2005, 46, 8993−8995.
(20) Gammon, D. W.; Steenkamp, D. J.; Mavumengwana, V.;
Marakalala, M. J.; Mudzunga, T. T.; Hunter, R.; Munyololo, M. Bioorg.
Med. Chem. 2010, 18, 2501−2514.
(21) Chen, C.-T.; Weng, S.-S.; Kao, J.-Q.; Lin, C.-C.; Jan, M.-D. Org.
Lett. 2005, 7, 3343−3346.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.6b02241.
Crystal structure of the α-anomer of 2 (CCDC no.
1477833) (CIF)
Experimental procedures, characterization data, crystallo-
graphic information, and NMR spectra (PDF)
AUTHOR INFORMATION
Corresponding Author
■
(22) Szabo,
Asymmetry 2005, 16, 83−95.
́ ́ ́
Z. B.; Borbas, A.; Bajza, I.; Liptak, A. Tetrahedron:
*E-mail: cmp@chem.ku.dk.
Notes
(23) Andersen, S. M.; Heuckendorff, M.; Jensen, H. H. Org. Lett. 2015,
17, 944−947.
(24) Volbeda, A. G.; Kistemaker, H. A. V.; Overkleeft, H. S.; van der
The authors declare no competing financial interest.
Marel, G. A.; Filippov, D. V.; Codee
8796−8806.
́
, J. D. C. J. Org. Chem. 2015, 80,
ACKNOWLEDGMENTS
We thank CHEM, UCPH, for funding and Rikke Munch Gelardi,
CHEM, UCPH, for solving the X-ray crystallographic data.
(25) El Nemr, A.; Tsuchiya, T. Carbohydr. Res. 2001, 330, 205−214.
(26) Pedersen, C. M.; Olsen, J.; Brka, A. B.; Bols, M. Chem. - Eur. J.
2011, 17, 7080−7086.
■
(27) Gisch, N.; Kohler, T.; Ulmer, A. J.; Muthing, J.; Pribyl, T.; Fischer,
̈
K.; Lindner, B.; Hammerschmidt, S.; Zahringer, U. J. Biol. Chem. 2013,
288, 15654−15667.
̈
REFERENCES
■
(1) Dube, D. H.; Champasa, K.; Wang, B. Chem. Commun. 2011, 47,
87−101.
(28) Kalikanda, J.; Li, Z. J. Org. Chem. 2011, 76, 5207−5218.
(29) Dere, R. T.; Kumar, A.; Kumar, V.; Zhu, X.; Schmidt, R. R. J. Org.
Chem. 2011, 76, 7539−7545.
(2) (a) Adibekian, A.; Stallforth, P.; Hecht, M.-L.; Werz, D. B.;
Gagneux, P.; Seeberger, P. H. Chem. Sci. 2011, 2, 337−344. (b) Werz, D.
B.; Ranzinger, R.; Herget, S.; Adibekian, A.; von der Lieth, C.-W.;
Seeberger, P. H. ACS Chem. Biol. 2007, 2, 685−691.
(3) Moran, A. P.; Holst, O.; Brennan, P. J.; von Itzstein, M. Part I:
Microbial Glycolipids, Glycoproteins and Glycopolymers. In Microbial
Glycobiology: Structures, Relevance and Applications; Academic Press:
London, 2009; pp 3−231.
(30) van der Es, D.; Groenia, N. A.; Laverde, D.; Overkleeft, H. S.;
Huebner, J.; van der Marel, G.; Codee
24, 3893−3907.
́
, J. D. C. Bioorg. Med. Chem. 2016,
(4) Lemieux, R. U.; Ratcliffe, R. M. Can. J. Chem. 1979, 57, 1244−1251.
(5) Mirabella, S.; Cardona, F.; Goti, A. Org. Biomol. Chem. 2016, 14,
5186−5204.
(6) Enugala, R.; Carvalho, L. C. R.; Dias Pires, M. J.; Marques, M. M. B.
Chem. - Asian J. 2012, 7, 2482−2501.
́
(7) Codee, J. D. C.; Litjens, R. E. J. N.; van den Bos, L. J.; Overkleeft, H.
S.; van der Marel, G. A. Chem. Soc. Rev. 2005, 34, 769−782.
D
DOI: 10.1021/acs.orglett.6b02241
Org. Lett. XXXX, XXX, XXX−XXX