Automated synthesis of lipomannan backbone α(1-6) oligomannoside via glycosyl phosphates: Glycosyl tricyclic orthoesters revisited

Article abstract of DOI:10.1039/b804069a

Glycosyl tricyclic orthoesters provide a versatile basis for the efficient generation of glycosyl phosphates, which are used in the automated synthesis of lipomannan backbone α(1-6) hexa-mannoside. The Royal Society of Chemistry.

Full text of DOI:10.1039/b804069a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Automated synthesis of lipomannan backbone a(1-6) oligomannoside via
glycosyl phosphates: glycosyl tricyclic orthoesters revisitedwz
Xinyu Liu,ya Reiko Wada,a Siwarutt Boonyarattanakalin,ab Bastien Castagnera
and Peter H. Seeberger*a
Received (in Cambridge, UK) 10th March 2008, Accepted 18th April 2008
First published as an Advance Article on the web 27th May 2008
DOI: 10.1039/b804069a
Glycosyl tricyclic orthoesters provide a versatile basis for the
efficient generation of glycosyl phosphates, which are used in the
automated synthesis of lipomannan backbone a(1-6) hexa-
mannoside.
Mycobacterium tuberculosis (Mtb) cell surface lipomannan
Fig. 1 Retrosynthetic analysis of the automated synthesis of a(1-6)
(LM) and lipoarabinomannan (LAM) are considered to be
oligomannosides.
the virulent factors in tuberculosis pathogenesis.1 These oligo-
saccharides have become the target of chemical syntheses to
Due to the base-labile nature of the Fmoc group, a robust
generate the defined molecular tools for subsequent biological
route for its placement in 2 calls for the temporary protection
studies.2 We have recently completed the chemical syntheses of
of the C-6 hydroxyl with a triisopropylsilyl (TIPS) group,7
Mtb phosphatidylinositol mannosides (PIMs), as well as a
followed by benzylation of the C-3 and C-4 hydroxyls
LAM oligosaccharide in the solution phase.3 To facilitate
(Scheme 1, route a). Subsequent removal of the TIPS group
access to these molecules, an automated synthesis would be
is necessary before installation of the Fmoc. While such a
route is practically applicable,8 the temporary protection of
convenient. As the initial target for this synthetic campaign,
we selected a(1-6) oligomannosides,4 which constitute the
backbone of the LM and LAM complexes.
the C-6 hydroxyl is costly and not atom-efficient.
To develop a more efficient process, we turned our attention
To achieve the regio- and diastereoselective automated
to a glycosyl tricyclic orthoester where all three oxygen atoms
assembly of the a(1-6) oligomannosides in a rapid fashion, a
mannose building block such as 1 would be ideal (Fig. 1).
in the orthoester were derived from the same monosaccharide
unit. Although this class of glycosyl orthoester has been
Glycosyl trichloroacetimidates or dibutyl phosphates allow for
known for more than three decades,9 there are few examples
fast glycosylations by Lewis acid activation.5 A C-2 ester
of its application to oligosaccharide synthesis.10,11
serves as the stereo-directing group, and an orthogonal C-6
We planned to access glycosyl tricyclic orthoesters by the
simple acidic treatment of glycosyl 1,2-orthoesters.12 These
fluorenylmethoxycarbonyl (Fmoc) group can be rapidly
cleaved for glycan chain elongation events. Although a num-
suitably protected tricyclic orthoesters would be selectively
ber of approaches can be adopted for the preparation of 1, a
opened by dibutyl phosphate to furnish glycosyl phosphates,
synthetic route via mannosyl orthoester 2 should be the most
allowing for the direct protection of the remaining
efficient. The 1,2-O-orthoester would serve as a temporary
hydroxyl group.
protecting group for the C-1 and C-2 hydroxyl groups. Upon
We evaluated the formation of mannosyl tricyclic ortho-
esters from their corresponding 1,2-orthoesters (Table 1).13
differentiation of the remaining hydroxyls, the glycosyl 1,2-O-
orthoester can either be selectively hydrolysed to the hemi-
Camphorsulfonic acid (CSA) was found to be the optimal acid
acetal for subsequent installation of the anomeric trichloro-
acetimidate4 or ring opened by dibutyl phosphate to afford the
glycosylphosphate directly.6
catalyst for inducing the trans-orthoesterification of mannosyl
1,2-O-orthoesters. Interestingly, the substituents of mannosyl
orthoesters were found to have a profound effect on the
efficiency of the acid-induced cyclization. In the case of
mannosyl orthobenzoates 2a and 2b, the change of one
a Laboratory for Organic Chemistry, Swiss Federal Institute of
Technology, ETH Zurich, HCI F315, Wolfgang-Pauli Straße 10,
¨
Zurich 8093, Switzerland. E-mail: seeberger@org.chem.ethz.ch
¨
b Sirindhorn International Institute of Technology, Thammasat
University, PO Box 22 Thammasat-Rangsit Post Office,
Pathumthani 12121, Thailand
w Electronic supplementary information (ESI) available: Detailed
experimental procedures and compound characterization data. See
DOI: 10.1039/b804069a
z CCDC reference number 681095. For crystallographic data in CIF
or other electronic format see DOI: 10.1039/b804069a
y Current address: Department of Biological Chemistry and Molecu-
lar Pharmacology, Harvard Medical School, D1-608, 240 Longwood
Avenue, Boston, MA, 02115, USA.
Scheme 1 Plausible synthetic routes for the preparation of mannosyl
phosphates 4a and 4b via 1,2-O-orthoester 3 (route a) or tricyclic
orthoester 5 (route b).
ꢀc
This journal is The Royal Society of Chemistry 2008
3510 | Chem. Commun., 2008, 3510–3512

View Article Online
Table 1 Formation of mannosyl tricylic orthoesters via 1,2-O-
orthoestersa
Table 2 Generation of xylosyl and glucosyl phosphates via tricylic
orthoestersa
Entry
Mannosyl
1,2-orthoester
Tricyclic
orthoester
Yield
(%)
Entry Glycosyl
1,2-orthoester
Tricyclic
Glycosyl
orthoester (yield)c phosphate (yield)d
1
2
3
4
2a (R1 = Ph;
R2 = Me)
5a (R1 = Ph)
85
95
40
95
1
2b (R1 = Ph;
R2 = All)
5a
2c (R1 = Me;
R2 = Me)
5b (R1 = Me)
2
2d (R1 = Me;
R2 = All)
5b
a
All trans-orthoesterifications were performed in the presence of 4 A
MS with 10 mol% CSA. The completion of the reaction required 24 h
for entries 1 and 3, and 12 h for entries 2 and 4. The crude tricyclic
orthoester was subjected directly to benzylation without purification.
3b
orthoester substituent from O-methyl to O-allyl not only
shortened the reaction time, but also improved the chemical
yield (Table 1, entries 1 and 2). Similarly, in contrast to
mannosyl allyl orthoacetate 2d, the cyclization of mannosyl
methyl orthoacetate 2c was rather sluggish using CSA as the
acid catalyst. X-Ray crystallography clearly shows the
presence of the tricyclic system in 5a (Table 1).z
a
CSA (10 mol%) (entry 1), silica gel (entry 2) and 4 A AW-MS (entry
b
3) were used as acid catalysts. In entry 3, CH3CN was used instead of
c
CH2Cl2. Yields indicated for 5 are two-step yields, after benzyla-
d
tion. The ring openings of 5 to 6 by dibutyl phosphate were
performed as described for the conversion of 5a to 6a in Scheme 2.
With the desired mannosyl tricyclic orthoesters in hand, we
proceeded to investigate their ring opening with dibutyl phos-
phate (Scheme 2). In contrast to mannosyl 1,2-orthoesters,6
mannosyl tricyclic orthoesters were found to be less reactive
towards dibutyl phosphate.14 Although the addition of excess
dibutyl phosphate (6 equiv.) was required to complete the
conversion of 5a to 6a in 24 h using CH2Cl2 as the solvent, the
reaction proceeded cleanly with complete regioselectivity to
give mannosyl phosphate 6a in an 83% isolated yield. Surpris-
ingly, mannosyl tricyclic orthoacetate 5b was completely inert
to dibutyl phosphate at room temperature. Microwave
irradiation at 120 1C for 5 min was necessary to achieve the
conversion of 5b to 6b.
tive scale-up of these procedures in order to supply the
automation campaign (Scheme 3). Mannosyl allyl 1,2-O-
orthobenzoate 2b was readily prepared from D-mannose in 4
steps on a 50 mmol scale. After a short silica gel filtration to
remove the methylbenzoate by-product, 2b was subjected to a
CSA-induced cyclization. Without purification, the intermedi-
ate tricyclic orthoester was benzylated to give crystalline 5a in
70–80% yield in 6 steps. Ring opening of 5a with dibutyl
phosphate was routinely performed on a 10 mmol scale, and
the resulting C-6 hydroxyl group was capped with Fmoc to
give target building block 4a (Scheme 3).
With a large quantity of mannosyl phosphate 4a secured, we
assembled a(1-6) oligomannosides by an automated synthesis.
For each cycle, a single glycosylation was performed using 5
equiv. of trimethylsilyl trifluoromethanesulfonate (TMSOTf)
as an activator and 5 equiv. of building block 4a at ꢁ15 1C for
45 min. Removal of the Fmoc group was efficiently conducted
using piperidine. Six-fold repetition of the glycosylation–
deprotection sequence, followed by cleavage of the oligosac-
charide from the resin,15 furnished the desired hexasaccharide,
7, in an 80% isolated yield (Scheme 4). The HPLC spectrum of
the crude material after the cleavage showed oligosaccharide 7
as a major peak, with a negligible amount of impurities
(Fig. 2). Purification was performed by simple column
chromatography.
We further expanded this approach to generate xylosyl and
glucosyl phosphates via their corresponding tricylic ortho-
esters (Table 2).13 The direct formation of xylosyl tricylic
orthoester 5c from 1,2-O-orthoester 2e was effectively induced
by CSA, although 2f and glucosyl 1,2-O-orthoester 2g re-
quired milder acidic agents such as silica gel or acid-washed
molecular sieves (AW-MS). The subsequent ring openings of
tricyclic orthoesters 5c–5e by dibutyl phosphate proceeded
regioselectively to give the corresponding glycosyl phosphates,
6c–6e, in good to excellent yield.
Having established suitable procedures for the formation of
tricyclic mannosyl orthoesters and their transformation into
mannosyl phosphates, we focused our efforts on the prepara-
Scheme 2 Ring opening of mannosyl tricylic orthoesters with dibutyl
phosphate.
Scheme 3 Scaled-up preparation of 4a via a tricyclic mannosyl
orthoester.
ꢀc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 3510–3512 | 3511

View Article Online
3. (a) X. Liu, B. L. Stocker and P. H. Seeberger, J. Am. Chem. Soc.,
2006, 128, 3638; (b) A. Holemann, B. L. Stocker and
P. H. Seeberger, J. Org. Chem., 2006, 71, 8071.
4. Synthesis of a(1-6) tetra-mannoside using ionic liquid-supported
glycosylation was recently reported: A. K. Pathak, C. K. Yerneni,
Z. Young and V. Pathak, Org. Lett., 2008, 10, 145.
5. (a) R. R. Schmidt and W. Kinzy, Adv. Carbohydr. Chem. Bio-
chem., 1994, 50, 21; (b) J. D. C. Codee and P. H. Seeberger, ACS
Symp. Ser., 2007, 960, 150.
Scheme 4 Automated assembly of a-(1,6) hexa-mannoside 7.
6. A. Ravida, X. Liu, L. Kovacs and P. H. Seeberger, Org. Lett.,
2006, 8, 1815.
7. We are aware that the TIPS group is not the only protecting group
that could be chosen for this purpose. However, the steric bulk
and base-stable nature of TIPS make it the ideal choice in this
case.
Fig. 2 HPLC trace of crude hexasaccharide 7.
8. F. R. Carrel and P. H. Seeberger, J. Carbohydr. Chem., 2007, 26,
125.
In summary, the pursuit of an automated synthesis of LM
backbone a(1-6) hexa-mannoside led us to identify glycosyl
tricyclic orthoesters as unique synthetic intermediates for the
preparation of functional carbohydrate building blocks. The
simple acid-catalyzed trans-orthoesterification of glycosyl 1,2-
orthoesters allows the direct formation of glycosyl tricyclic
orthoesters. These orthoesters can be efficiently converted to
glycosyl dibutyl phosphates, as exemplified by mannose,
xylose and glucose.
9. (a) A. F. Bochkov, I. V. Obruchnikov and N. K. Kochetkov, Izv.
Akad. Nauk. SSSR, Ser. Khim., 1971, 1291; (b) A. F. Bochkov,
V. I. Snyatkova, Y. V. Voznyi and N. K. Kochetkov, Zh. Obshch.
Khim., 1971, 41, 2776; (c) A. F. Bochkov, Y. V. Voznyi,
V. N. Chernetskii, V. M. Dashunin and A. V. Rodionov, Izv.
Akad. Nauk. SSSR, Ser. Khim., 1975, 420.
10. Prandi et al. have reported the CSA-catalyzed trans-orthoester-
ification of 1,2-orthoarabinose to the tricyclic 1,2,5-orthoester.
These tricyclic orthoesters were reacted with thiols to give thio-
glycosides, which were used in the syntheses of oligoarabinofur-
anosides. See: (a) S. Sanchez, T. Bamhaoud and J. Prandi,
Tetrahedron Lett., 2000, 41, 7447; (b) T. Bamhaoud, S. Sanchez
and J. Prandi, Chem. Commun., 2000, 659. Glycosyl phosphates
are currently easier to use for automated synthesis than thioglyco-
sides because glycosyl phosphates can be activated under homo-
geneous reaction conditions with a single activator, such as
TMSOTf.
This research was supported by ETH Zurich, the Swiss
¨
National Science Foundation (SNF Grant 200121-101593), the
Japan Society for the Promotion of Science (JSPS postdoctoral
fellowship for research abroad, for R. W.) and the Roche
Research Foundation (postdoctoral fellowship, for S. B.). We
thank P. Seiler for the X-ray analysis of 5a. We thank CEM
Corp. for providing microwave reactors in our laboratory.
11. (a) S. Hiranuma, O. Kanie and C.-H. Wong, Tetrahedron Lett.,
1999, 40, 6423; (b) M. Hori and F. Nakatsubo, Macromolecules,
2000, 33, 1148.
12. The acid-induced trans-orthoesterification of glycosyl 1,2-ortho-
esters has been described in the literature; see refs. 9, 10 and 11a.
However, we were unable to reproduce the results of Bochkov
et al. (see ref. 9c), reporting that para-toluenesulfonic acid can
induce the trans-orthoesterification of 2g to the corresponding
tricyclic orthoester in 48% yield (our result: 26%). Hiranuma et al.
(ref. 11a) reported the formation of a mannosyl tricyclic orthoester
using pyridine as the solvent and pyridium trifluoromethanesul-
fonate as the acid catalyst. However, these conditions are less
practical for scale-up.
Notes and references
1. (a) P. J. Brennan, Annu. Rev. Biochem., 1995, 64, 29;
(b) D. Chatterjee and K. H. Khoo, Glycobiology, 1998, 8, 113.
2. For selected examples: (a) J. D. Ayers, T. L. Lowary,
C. B. Morehouse and G. S. Besra, Bioorg. Med. Chem. Lett.,
1998, 8, 437; (b) F. W. D’Souza, J. D. Ayers, P. R. McCarren and
T. L. Lowary, J. Am. Chem. Soc., 2000, 122, 1251;
(c) A. Stadelmaier, M. B. Biskup and R. R. Schmidt, Eur. J.
Org. Chem., 2004, 3292; (d) K. N. Jayaprakash, J. Lu and
B. Fraser-Reid, Bioorg. Med. Chem. Lett., 2004, 14, 3815;
(e) J. A. Watt and S. J. Williams, Org. Biomol. Chem., 2005, 3,
1982; (f) B. S. Dyer, J. D. Jones, G. D. Ainge, M. Denis,
D. S. Larsen and G. F. Painter, J. Org. Chem., 2007, 72, 3282;
(g) M. Joe, Y. Bai, R. C. Nacario and T. L. Lowary, J. Am. Chem.
Soc., 2007, 129, 9885; (h) C. Rademacher, G. K. Shoemaker,
H. S. Kim, R. B. Zheng, H. Taha, C. Liu, R. C. Nacario,
D. C. Schriemer, J. S. Klassen, T. Peters and T. L. Lowary,
J. Am. Chem. Soc., 2007, 129, 10489.
13. 1,2-Orthoesters were prepared in four steps from D-mannose,
D-glucose or D-xylose in good yield. See ESIw.
14. Tricyclic orthoesters have been used as protecting groups for
carboxylic acids, and they require fairly vigorous conditions for
cleavage. See: (a) E. J. Corey and N. Raju, Tetrahedron Lett.,
1983, 24, 5571; (b) M. A. Blaskovich and G. A. Lajoie, J. Am.
Chem. Soc., 1993, 115, 5021.
15. R. B. Andrade, O. J. Plante, L. G. Melean and P. H. Seeberger,
Org. Lett., 1999, 11, 1811.
ꢀc
This journal is The Royal Society of Chemistry 2008
3512 | Chem. Commun., 2008, 3510–3512