A convenient and efficient synthetic approach to mono-, di-, and tri-O-mannosylated Fmoc amino acids

Article abstract of DOI:10.1016/j.tetlet.2013.02.060

A convenient and highly efficient synthesis of mono-, di-, and tri-O-mannosylated Fmoc-Ser and Thr is described. The short synthetic route and high overall yield highlight the synthetic utility of this unified approach.

Full text of DOI:10.1016/j.tetlet.2013.02.060

Tetrahedron Letters 54 (2013) 2190–2193
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
A convenient and efficient synthetic approach to mono-, di-,
and tri-O-mannosylated Fmoc amino acids
⇑
Liqun Chen, Zhongping Tan
Department of Chemistry and Biochemistry and BioFrontiers Institute, University of Colorado Boulder, Boulder, CO 80303, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A convenient and highly efficient synthesis of mono-, di-, and tri-O-mannosylated Fmoc-Ser and Thr is
described. The short synthetic route and high overall yield highlight the synthetic utility of this unified
approach.
Received 18 January 2013
Revised 14 February 2013
Accepted 19 February 2013
Available online 26 February 2013
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Tri-O-mannosylated
Glycoamino acid
Synthesis
Unified approach
Protein O-mannosylation is commonly found in industrially and
medically important eukaryotic microorganisms.1 The transfer of
the first O-linked mannose (Man) to serine (Ser) and threonine
protected glycoamino acids using Fmoc-based solid phase peptide
synthesis. We reasoned that the same approach could be extended
to the study of protein O-mannosylation. To generate protein
(Thr) by protein-O-
D
-mannosyltransferase occurs in the endoplas-
molecules with defined mannosylation patterns, a suitable
mic reticulum (ER) with dolichol-P-mannose as the donor. The
nascent chain is then linearly elongated in the Golgi apparatus by
the sequential addition of more mannose residues, each derived
from guanosine diphosphate mannose (GDP-mannose) by
mannosyltransferases.2 In yeast and filamentous fungi, the O-linked
glycan structures are predominantly based on common cores of
synthetic route that allows for the convenient preparation of
O-mannosylated glycoamino acid building blocks is required.
Several methods have been previously described for the synthe-
sis of O-mannosylated Fmoc amino acids.8,9 These methods have
generally relied on two strategies: one involves the use of
carboxyl-protected Ser and Thr and the other involves the use of
carboxyl-unprotected amino acids. Although the use of esters
makes the purification easier, the additional protection and depro-
tection steps lower the synthetic efficiency. To facilitate the produc-
tion of multigram scale quantities of the highly complex building
blocks, we elected to directly glycosylate the unprotected
Fmoc-Ser-OH and Fmoc-Thr-OH with the peracetylated mannose,
mannobiose, and mannotriose.10,11
Man1-Ser/Thr 1, Mana1,2Mana1-Ser/Thr 2, and Mana1,2Mana1,2-
Man 1-Ser/Thr 3 (Fig. 1). Further modifications of the structural
a
cores with glucose (Glc), galactopyranose (Galp), or galactofuranose
(Galf) residues also occur, leading to the formation of more complex
and branched structures 4.3
The O-linked mannose glycans have been increasingly impli-
cated in controlling the production and function of proteins in
yeast and fungi. Over the past several decades, genetic, biochemi-
cal, and computational studies have established the importance of
carbohydrates in protein secretion, localization, solubility, stabil-
ity, and biological activity.4 However, a key unresolved question
is how the structure of the O-mannosyl glycan is related to the
biophysical and biological properties of proteins.5,6
Chemical synthesis has emerged as a powerful approach to
study protein glycosylation.7 This method involves the systematic
comparison of the properties of differently glycosylated forms of
proteins (glycoforms). Each glycoform can be prepared from
We started with the synthesis of the mono-mannosylated Ser
and Thr. Although the direct condensation of the unprotected
Fmoc-Ser-OH with mannose pentaacetate has been reported, the
applicability of the same method for mannosyl Thr synthesis has
not been established.10,12 The added steric hindrance of the
b-methyl group of threonine may influence the condensation pro-
cess. To access the effect of the methyl group, we compared the
synthesis of
a
-O-mannosyl-Ser 7 and Thr 8.11 As illustrated in
Scheme 1, mannose pentaacetate 6 was generated from D-(+)-man-
nose 5 using acetic anhydride.13 Coupling of 6 with Fmoc-Ser-OH
and Fmoc-Thr-OH using BF3ÁOEt2 as promoter and CH3CN as
solvent afforded the
a-linked 7 and 8 in 44% and 41% yield,
⇑
Corresponding author. Tel.: +1 303 492 5160.
E-mail address: zhongping.tan@colorado.edu (Z. Tan).
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.02.060

L. Chen, Z. Tan / Tetrahedron Letters 54 (2013) 2190–2193
2191
Figure 1. Core structures of the O-linked glycans in yeast and fungi.
respectively, suggesting that Thr has similar reactivity with man-
nose acetate as Ser.
The next goal was to gauge the generality of this reaction for
preparing di-mannosylated building blocks. To this end, we first
of the reaction conditions did not improve the efficiency of the syn-
thetic process. This low yield would be a serious limitation in the
efficient conversion of mannose 5 to the peracetylated tri-manno-
side 21. According to the literature, the peracetylated tri-manno-
side can be synthesized by coupling 11 with a dimmanosyl
fluoride, which in turn, was also prepared from 11.8 Because of
the double use of a difficult-to-make monosaccharide 11, the pre-
vious synthesis of tri-O-mannosylated glycoamino acids was not
efficient.
To facilitate the preparation of di- and tri-mannosylated Ser and
Thr, we sought to develop a new approach for the synthesis of per-
acetylated oligomannoses. It is necessary that the approach should
avoid the use of 11 and be able to provide the access to the syn-
thetic precursors, including 19 and 21, in a unified, systematic
fashion. We began our efforts by taking note of a recent work by
Fraiser-Reid and co-workers, which described a one-pot synthesis
of methyl di- and trisaccharide 17 and 18 from a mannose-derived
methyl 1,2-orthoacetate 16.17 On the basis of this work, we envi-
sioned the possibility of utilizing a similar strategy to prepare
the peracetylated mannosyl saccharides.
Thus, we formulated the synthetic strategy outlined in Scheme
3, in which the methyl orthoacetate 16 precursor could be trans-
formed to both 19 and 21 in two steps. The synthesis of orthoace-
tate 16 started from mannose 9, which was prepared from
mannose 5 in two steps as described above. Compound 9 was
transformed to the orthoester 16 in the presence of 2,6-lutidine
and methanol. With 16 in hand, we then investigated its transfor-
mation to methyl di- and tri-mannoses 17 and 18. Treatment of
unpurified 16 with TMSOTf in CH2Cl2 only afforded the disaccha-
ride 17 in 56% yield. To our delight, reaction of purified 16 under
the same conditions gave a separable mixture containing both
the di- and trisaccharides 17 and 18 (58% and 16%, respectively).
We next sought to convert 17 and 18 into the 1-O-acteylated
saccharides 19 and 21, respectively, using acetic anhydride and a
catalytic amount of sulfuric acid. We were pleased to find that, un-
der previously reported conditions, both reactions proceeded
smoothly to give the desired peracetylated saccharides in good
yields.18 With 19 and 21 in hand, the stage was set to investigate
the feasibility of coupling the glycans with Fmoc-Ser-OH and
Fmoc-Thr-OH. As expected, coupling reactions using BF3ÁOEt2 as
promoter afforded the di-mannosylated Thr 20, the tri-mannosy-
lated Ser 22 and Thr 23 in 64%, 57%, and 58% yield, respectively.
In summary, we have developed a convenient and highly effi-
cient approach for the synthesis of mannosyl and oligomannosyl
Ser/Thr building blocks. Using this unified approach, all the key
intermediates, the peracetylated mono-, di-, and tri-mannoses,
can be obtained in only one to five steps through the same
synthetic route starting from mannose. Together with the direct
glycosylation of the carboxyl-unprotected Fmoc-Ser-OH and
Fmoc-Thr-OH, this synthetic approach allows us to shorten the
synthesis of mono- and di-mannosyl building blocks by two steps,
synthesized the peracetylated Mana1,2Man 12. There are a few
available methods for the preparation of this compound. In all
methods, 1,3,4,6-tetra-O-acetyl-b-D-mannopyranose 11 was used
as the glycosyl acceptor for mannosylation.8,9,14 Their differences
lie in their glycosyl donors. One method is based on the coupling
of mannosyl bromide 9 with 11, whereas other methods are based
on the use of peracetylated a-ethylthiomannoside or mannose tri-
chloroacetimidate, which has to be separately prepared from man-
nose. To simplify the synthesis, we chose to synthesize 12 from 9.
Since 9 can be obtained during the preparation of 11, by using the
synthetic route shown in Scheme 2, two reaction steps can be
spared.
The synthesis of 11 started from
Peracetylation of 5 followed by bromination gave 2,3,4,6-tetra-O-
acetyl- -mannopyranosyl bromide 9 in 88% yield over two
D-(+)-mannose 5 (Scheme 2A).
a-D
steps.13 Exposure to 2,4,6-collidine and EtOH converted 9 into
1,2-orthoester 10 in 84% yield. Regioselective opening of the ortho-
ester structure led to the desired product 11 in 25% yield.15,16
With compounds 9 and 11 in hand, we embarked on the synthe-
sis of 12. It has been reported that a highly efficient coupling could
be achieved in the presence of AgOTf and collidine.9 However, one-
portion addition of 11–9 did not provide 12 with the desired purity
and yield, presumably due to the instability of 11 under the reac-
tion conditions. Fortunately, dropwise addition of 11 to 9 led to a
much better outcome. Moreover, despite the presence of an insep-
arable disaccharide 13 (9%),17 we were able to generate the desired
di-mannosylated building block 15 from Fmoc-Ser-OH in 41%
yield, which is comparable to that for mono-mannosylated Ser.
While the results for the synthesis of di-O-mannosylated Ser
were promising, it is important to note that the yield for the
conversion of 10 into 11 was a disappointing 25%. Optimization
Scheme 1. Synthesis of mono-mannosylated Ser and Thr. Reagents and conditions:
(a) Ac2O, NaOAc, 110 °C, 91%; (b) Fmoc-Ser-OH, BF3ÁOEt2, CH3CN, 44% and (c) Fmoc-
Thr-OH, BF3ÁOEt2, CH3CN, 41%.

2192
L. Chen, Z. Tan / Tetrahedron Letters 54 (2013) 2190–2193
Scheme 2. Synthesis of di-mannosylated Ser. Reagents and conditions: (a) Ac2O, NaOAc, 110 °C; (b) HBr, AcOH, CH2Cl2, 88% over two steps; (c) 2,4,6-collidine, EtOH, CH3CN,
84%; (d) 1.0 N HCl, acetone, 25%; (e) AgOTf, 2,4,6-collidine, 4 Å molecular sieves, one-portion addition of 11–6, À40 °C, 42%; (f) AgOTf, 2,4,6-collidine, 4 Å molecular sieves,
dropwise addition of 11–6, À40 °C, 99% and (g) BF3ÁOEt2, CH3CN, 41%.
Scheme 3. Synthesis of di-mannosylated Thr and tri-mannosylated Ser and Thr. Reagents and conditions: (a) 2,6-lutidine, MeOH, CHCl3, 74%; (b) TMSOTf, CH2Cl2, À30 °C, 17,
58%, 18, 16%; (c) Ac2O, H2SO4, 0 °C, 89%; (d) BF3ÁOEt2, CH3CN, 64%; (e) Ac2O, H2SO4, 0 °C, 73%; (f) BF3ÁOEt2, CH3CN, 57% and (g) BF3ÁOEt2, CH3CN, 58%.
and the synthesis of tri-mannosyl building blocks by five steps.
Most notably, avoiding the use of a low-yield-intermediate multi-
ple times and the one-pot strategy helped improve the overall effi-
ciency of the synthesis of the di- and tri-mannosyl building blocks
by more than three times, from 6% in 8 steps to 22% in 6 steps for
20, and from 0.9% in 11 steps to 4.4% in 6 steps for 23. Overall, this
unified approach has enabled the quick and efficient preparation of
six glycosylated Ser/Thr building blocks, which can be used in the
synthesis of structurally defined glycoproteins containing multiple
mannose residues on their surface.

L. Chen, Z. Tan / Tetrahedron Letters 54 (2013) 2190–2193
2193
4. Willer, T.; Valero, M. C.; Tanner, W.; Cruces, J.; Strahl, S. Curr. Opin. Struct. Biol.
2003, 13, 621.
Acknowledgments
5. Espitia, C.; Servin-Gonzalez, L.; Mancilla, R. Mol. Biosyst. 2010, 6, 775.
6. Tsuboi, S.; Hatakeyama, S.; Ohyama, C.; Fukuda, M. Trends Mol. Med. 2012, 18,
224.
7. Price, J. L.; Shental-Bechor, D.; Dhar, A.; Turner, M. J.; Powers, E. T.; Gruebele,
M.; Levy, Y.; Kelly, J. W. J. Am. Chem. Soc. 2010, 132, 15359.
8. Varon, D.; Lioy, E.; Patarroyo, M. E.; Schratt, X.; Unverzagt, C. Aust. J. Chem.
2002, 55, 161.
We are grateful to Patrick Chaffey for assistance with the syn-
thesis. We thank Matthew Drake for assistance with the prepara-
tion of the manuscript. We also thank the University of Colorado
Boulder for the financial support.
9. Jansson, A. M.; Meldal, M.; Bock, K. J. Chem. Soc., Perkin Trans. 1 1992, 1699.
10. Fahmi, N. E.; Golovine, S.; Wang, B.; Hecht, S. M. Carbohydr. Res. 2001, 330, 149.
11. Salvador, L. A.; Elofsson, M.; Kihlberg, J. Tetrahedron 1995, 51, 5643.
12. Yao, N.; Fung, G.; Malekan, H.; Ye, L.; Kurth, M. J.; Lam, K. S. Carbohydr. Res.
2010, 345, 2277.
13. Nobrega, C.; Vazquez, J. T. Tetrahedron: Asymmetry 2003, 14, 2793.
14. Upreti, M.; Ruhela, D.; Vishwakarma, R. A. Tetrahedron 2000, 56, 6577.
15. Deferrari, J. O.; Gros, E. G.; Mastronardi, I. O. Carbohydr. Res. 1967, 4, 432.
16. Toyokuni, T.; Kumar, J. S. D.; Gunawan, P.; Basarah, E. S.; Liu, J.; Barrio, J. R.;
Satyamurthy, N. Mol. Imaging Biol. 2004, 6, 324.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tetlet.2013.02.
060.
References and notes
17. Uriel, C.; Ventura, J.; Gomez, A. M.; Lopez, J. C.; Fraser-Reid, B. J. Org. Chem.
2012, 77, 795.
18. Meloncelli, P. J.; Williams, T. M.; Hartmann, P. E.; Stick, R. V. Carbohydr. Res.
2007, 342, 1793.
1. Lommel, M.; Strahl, S. Glycobiology 2009, 19, 816.
2. Gemmill, T. R.; Trimble, R. B. Biochim. Biophys. Acta 1999, 1426, 227.
3. Goto, M. Biosci. Biotechnol. Biochem. 2007, 71, 1415.