Selective deacylation of peracylated ribonucleosides

Article abstract of DOI:10.1016/j.tetlet.2009.01.147

A protocol for chemoselective deprotection of N,O-acylated ribonucleosides has been developed. Peracylated pyrimidine ribonucleosides subjected to guanidinium nitrate and NaOMe in MeOH/CH₂Cl₂ at 0 °C undergo high yielding O-deacylation, while even more pronounced chemoselectivity is observed with peracylated purine ribonucleosides as O5′-acyl groups are preserved. Nucleobase-protecting groups (A^Bz, C^Bz, G^iBu, and U^Bz) are stable to these conditions, rendering this reagent mixture as a valuable addition to the collection of protecting group protocols in nucleoside chemistry.

Full text of DOI:10.1016/j.tetlet.2009.01.147

Tetrahedron Letters 50 (2009) 1751–1753
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Selective deacylation of peracylated ribonucleosides
*
Jared W. Rigoli , Michael E. Østergaard , Kirsten M. Canady, Dale C. Guenther, Patrick J. Hrdlicka
Department of Chemistry, University of Idaho, Moscow, ID 83844-2343, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A protocol for chemoselective deprotection of N,O-acylated ribonucleosides has been developed. Peracy-
lated pyrimidine ribonucleosides subjected to guanidinium nitrate and NaOMe in MeOH/CH2Cl2 at 0 °C
undergo high yielding O-deacylation, while even more pronounced chemoselectivity is observed with
peracylated purine ribonucleosides as O50-acyl groups are preserved. Nucleobase-protecting groups
(ABz, CBz, GiBu, and UBz) are stable to these conditions, rendering this reagent mixture as a valuable addi-
tion to the collection of protecting group protocols in nucleoside chemistry.
Received 12 January 2009
Revised 28 January 2009
Accepted 28 January 2009
Available online 3 February 2009
Keywords:
Acylation
Published by Elsevier Ltd.
Deprotection
Guanidinium nitrate
LNA
Protecting groups
Chemically modified nucleosides and oligonucleotides are
widely explored for therapeutic and diagnostic purposes, for exam-
ple, as antiviral/anticancer agents,1 as building blocks within the
antigene/antisense/siRNA regimes,2–4 and for detection of single
nucleotide polymorphisms.5 Nucleosides are challenging synthetic
substrates as they contain several nucleophilic groups with compa-
rable reactivity (i.e., hydroxyl/amino groups) that must be chemi-
cally differentiated for successful multistep synthesis of target
nucleosides. Thus, development of methods for chemoselective
acylation/deacylation of nucleosides has been a subject of recent
interest, and includes direct O-benzoylation of free nucleosides,6,7
selective N-deacylation of peracylated ribonucleosides,8 and selec-
tive O20,O30-deacylation of O20,O30,O50-triacylated ribonucleo-
sides.9 Methods resulting in selective N-acylation of nucleobase
amines/imines are of particular interest since the base moieties
of adenosine, cytidine, and guanosine phosphoramidites must be
protected for successful automated oligonucleotide synthesis.
Selective N-acylation of these moieties has been realized by
(c) Peracylation followed by selective O-deacylation. Several
reagents have been developed toward this end including
(a) sodium methoxide used in the original Khorana
approach,13 (b) aqueous sodium hydroxide in pyridine,14,15
and (c) dilute methanolic ammonia.16 Most of these meth-
ods furnish the target nucleosides in moderate to good yield.
While exploring an alternative route toward N20-functionalized
20-amino- -L-LNA (Locked Nucleic Acid) monomers,17 we needed
a
facile access to thymine derivative 117b (Scheme 1). Chemoselec-
tive N3-benzoylation of O20-unprotected nucleoside 218 using di-
rect N-acylation,19 phase transfer conditions,20 or transient
protection protocols11 only resulted in low yields of 1 (<50%) due
to formation of O20-benzoylated and/or O20,N3-dibenzoylated
products.17b Alternatively, chemoselective O20-deacylation of fully
protected nucleoside 317b using the classic O-deacylation methods
outlined above (point c) was attempted, but was unsuccessful due
to the lability of the N3-benzoyl group and/or the formation of
undesired b-xylo-LNA nucleosides arising from intramolecular sub-
stitution reactions.17b Gratifyingly, subjection of 3 to a mixture of
guanidinium nitrate (GNO3) and sodium methoxide in methanol
and dichloromethane21 cleanly afforded O20-deacylated nucleoside
1 in 96% yield.17b This reagent mixture was also found to accom-
plish O20-deacetylation of the corresponding 6-N-benzoyladenine
and O50/O500-dibenzoylated thymine derivatives of 1 in similarly
high yields.22 These interesting findings prompted us to evaluate
the full potential of this reagent mixture to facilitate selective
deacylation of nucleoside substrates.
(a) Direct chemoselective N-acylation of unprotected nucleo-
sides using mild acylating agents such as aqueous benzoic
anhydride.10 This approach is restricted to cytidine nucleo-
sides, as O-acylation typically accompanies N-acylation with
other nucleosides.
(b) Transient protection methodology, that is, a three-step pro-
cess involving protection of free alcohol groups as silyl
ethers, N-acylation and desilylation using a weak base.11,12
Peracylated cytosine derivative 4a was chosen as an easily
accessible23 model compound for qualitative optimization of the
deacylation conditions (Table 1). Subjection of 4a to conditions
* Corresponding author. Tel.: +1 208 885 0108; fax: +1 208 885 6173.
E-mail address: hrdlicka@uidaho.edu (P.J. Hrdlicka).
To be considered joint first authors.
0040-4039/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.tetlet.2009.01.147

1752
J. W. Rigoli et al. / Tetrahedron Letters 50 (2009) 1751–1753
that mimic the original deacylation conditions very closely17b,21
surprisingly resulted in complete deacylation to exclusively afford
cytidine after 1 h (entry 1). Inexpensive analogs of guanidine such
as diphenylguanidine (DPG), creatinine, and tetramethylguanidine
(TMG) were evaluated as alternative reagents since their altered
steric requirements, nucleophilicity and/or basicity may affect
the rate and selectivity of the deacylation reactions. Deacylations
with DPG and creatinine were slow and non-selective (entries 2
and 3), while the use of TMG resulted in complete deacylation to
afford cytidine (entry 4). A decrease in the solvent polarity slowed
down the deacylation reaction with TMG and allowed isolation of
N4-benzoylcytidine 4b23 although only in 40% yield (entry 5).
Exchanging the methanol component of the solvent mixture with
water, ethanol, iso-propanol, acetonitrile, or THF was not advanta-
geous (results not shown). Accordingly, we turned our attention
back to the original guanidinium nitrate reagent. If sodium meth-
oxide is excluded from the reagent mixture, the deacylations do
not occur (entry 6). Thus, sodium methoxide is necessary to initiate
deacylations by GNO3, most likely by deprotonating the guanidi-
nium to form the active guanidine in situ. Decreasing the relative
amount of either guanidinium nitrate or sodium methoxide slowed
down deacylations, but did not improve chemoselectivity (entries
7 and 8). Gratifyingly, subjection of peracylated 4a to the original
conditions at lower temperatures (0 °C) successfully resulted in
fast and selective O-deacylation and formation of desired 4-N-ben-
zoylcytidine 4b23 in excellent yield (95%, entry 9).
O
NH
MsO
MsO
N
O
O
N3-Benzoylation
<50%
OBn
O
N
OH
NBz
O
2
MsO
MsO
O
OBn
O
N
NBz
O
OH
To investigate the scope of the most promising deacylation con-
dition (Table 1, entry 9), a representative set of fully protected
ribonucleosides 5a–10a was selected and synthesized using either
established protocols (5a,8 9a,24 and 10a25) or simple protecting
group manipulations of readily available nucleosides.23 The iden-
tity of known model nucleosides and deacylated products was ver-
ified by comparison of known spectral data (e.g., 1H NMR), while all
new compounds reported in this study were characterized by a
1
MsO
MsO
O
GNO3/NaOMe
MeOH/CH2Cl2
96%
OBn
OAc
3
Scheme 1. Synthesis of N3-benzoylated nucleoside 1.17b
Table 1
Qualitative optimization of reaction conditions for selective O-deacylation of 4aa
NHBz
N
NHBz
N
AcO
N
O
O
HO
NH
N
O
O
OAc OAc
OH OH
4a
4b
-
NH
NH
NO3
NH2
NH2
Ph
Ph
NH
N
N
N
N
H
N
H2N
H
O
Creatinine
GNO3
DPG
TMG
Entry
Reagent (equiv)
MeOH:CH2Cl2 (v:v)
NaOMe (equiv)
Product
1h
1d
1
2
3
4
5b
6
7
8
9
GNO3 (5.0)
DPG (1.0)
Creatinine (1.0)
TMG (1.0)
90:10
50:50
90:10
90:10
50:50
90:10
90:10
90:10
90:10
1.0
—
—
—
—
C
Mix
NR
C
—
Mix
Mix
—
C
NR
C
TMG (1.0)
4b (40%)c
NR
GNO3 (5.0)
GNO3 (0.33)
GNO3 (5.0)
GNO3 (5.0), 0 °C
—
1.0
0.2
1.0
Mix
Mix
C
—
4b (95%)c
a
Reactions were performed at room temperature using 1.0 equiv of nucleoside 4a (0.02 M) unless otherwise noted. C = cytidine, Mix = formation of complex mixture with
multiple products, NR = no reaction.
b
reaction performed using a 0.1 M solution of 4a. Nucleoside 4b was formed after 20 min.
Isolated yield.
c

J. W. Rigoli et al. / Tetrahedron Letters 50 (2009) 1751–1753
1753
Table 2
2.0 mmol) in MeOH (90 mL) and CH2Cl2 (10 mL) was prepared. Pro-
tected nucleoside (0.1 mmol) was dissolved in 5 mL of this solution
at 0 °C. Upon completion of the reaction, 5% aq. citric acid was
added and the aqueous phase was extracted with CH2Cl2. After
evaporation of solvents, the resulting crude was purified by silica
gel chromatography using methanol in CH2Cl2 as eluent.
Deacylation of model nucleosides using guanidinium nitrate (GNO3) and sodium
methoxidea
RO
B
RO
B
5.0 eq GNO3
O
O
1.0 eq NaOMe
MeOH/CH2Cl2, 0 ºC
Acknowledgments
AcylO OAcyl
OH OH
The assistance of Michael Chavez, Sarah Doornbos, T. Santhosh
Kumar, and Richa Tungal during preparation of model nucleosides
is appreciated. The input of Andreas S. Madsen, Nicolai K. Ander-
sen, and Professor Jesper Wengel (Nucleic Acid Center, University
of Southern Denmark) during early phases of this study is valued.
We greatly appreciate financial support from Idaho NSF EPSCoR
and Grant No. [UI08687] from the University of Idaho Student
Grant Program at the Univ. of Idaho.
4a-10a
4b-10b
Substrate
Product
Time (min)
Yield (%)
R
Acyl
B
R
B
4a
5a
6a
7a
8a
9a
10a
Ac
Ac
DMTr
Ac
Ac
Bz
Bz
Ac
Ac
Ac
Ac
Ac
Bz
Bz
CBz
CAc
UBz
AAc,Bz
GiBu
U
4b
5b
6b
7b
8b
9b
10b
H
H
DMTr
Ac
Ac
Bz
CBz
C
90
10
30
15
180
24 h
20
95
95
85
60
80
20
75
UBz
ABz
GiBu
U
Supplementary data
ABz
Bz
ABz
a
Reactions were performed as described for entry 9 in Table 1. CBz = 4-N-ben-
Experimental description and characterization data of com-
pounds 4a–10a and 4b–10b are available. Supplementary data
associated with this article can be found, in the online version, at
doi:10.1016/j.tetlet.2009.01.147.
zoyl-cytosin-1-yl, CAc = 4-N-acetyl-cytosin-1-yl, C = cytosin-1-yl, GiBu = 2-N-isobu-
tyryl-guanin-9-yl,
A
Ac,Bz = 6-N-acetyl,6-N-benzoyl-adenin-9-yl, ABz = 6-N-benzoyl-
adenin-9-yl, UBz = 3-N-benzoyl-uracil-1-yl, DMT = 4,40-dimethoxytrityl.
combination of 1H NMR (including D2O exchange studies), COSY, or
FAB-MS. The selected model nucleosides 4a–10a feature alcohol-
protecting groups that are commonly encountered in nucleoside
chemistry (–OAc, –OBz, ODMTr). The nucleobase moieties were
protected with acyl groups that are commonly used in standard so-
lid phase oligonucleotide synthesis (ABz, CBz, CAc, GiBu).
References and notes
1. Herdewijn, P. In Modified Nucleosides in Biochemistry, Biotechnology and
Medicine, 1st ed.; Wiley, 2008.
2. Duca, M.; Vekhoff, P.; Oussedik, K.; Halby, L.; Arimondo, P. B. Nucleic Acids Res.
2008, 36, 5123–5138.
3. Kurreck, J. Eur. J. Biochem. 2003, 270, 1628–1644.
4. Bumcrot, D.; Manoharan, M.; Koteliansky, V.; Sah, D. W. Y. Nat. Chem. Biol.
2006, 2, 711–719.
5. Asseline, U. Curr. Opin. Chem. 2006, 10, 491–518.
6. Prasad, A. K.; Kumar, V.; Malhotra, S.; Ravikumar, V. T.; Sanghvi, Y. S.; Parmar,
V. S. Bioorg. Med. Chem. 2005, 13, 4467–4472.
7. Kumar, V.; Parmar, V. S.; Malhotra, S. V. Tetrahedron Lett. 2007, 48, 809–812.
8. Nowak, I.; Conda-Sheridan, M.; Robins, M. J. J. Org. Chem. 2005, 70, 7455–7458.
9. Nowak, I.; Jones, C. T.; Robins, M. J. J. Org. Chem. 2006, 71, 3077–3081. and Ref. 9
therein.
10. Bhat, V.; Ugarkar, B. G.; Sayeed, V. A.; Grimm, K.; Kosora, N.; Domenico, P. A.;
Stocker, E. Nucleosides Nucleotides 1989, 8, 179–183.
11. Ti, G. S.; Gaffney, B. L.; Jones, R. A. J. Am. Chem. Soc. 1982, 104, 1316–1319.
12. Zhu, X.-F.; Williams, H. J.; Scott, A. I. Synth. Commun. 2003, 33, 1233–1243.
13. Rammler, D. H.; Khorana, H. G. J. Am. Chem. Soc. 1962, 84, 3112–3122.
14. Saito, Y.; Nyilas, A.; Agrofoglio, L. A. Carbohydr. Res. 2001, 331, 83–90.
15. Maguire, A. R.; Hladezuk, I.; Ford, A. Carbohydr. Res. 2002, 337, 369–372.
16. Neilson, T.; Werstjuk, E. S. Can. J. Chem. 1971, 49, 493–499.
17. (a) Hrdlicka, P. J.; Kumar, T. S.; Wengel, J. Chem. Commun. 2005, 4279–4281; (b)
Kumar, T. S.; Madsen, A. S.; Wengel, J.; Hrdlicka, P. J. J. Org. Chem. 2006, 71,
4188–4201; (c) Kumar, T. S.; Wengel, J.; Hrdlicka, P. J. Chem. Biol. Chem. 2007, 8,
1122–1125; (d) Kumar, T. S.; Madsen, A. S.; Østergaard, M. E.; Wengel, J.;
Hrdlicka, P. J. J. Org. Chem. 2008, 73, 7060–7066; e Kumar, T. S.; Madsen, A. S.;
Østergaard, M. E.; Sau, S. P.; Wengel, J.; Hrdlicka, P. J. J. Org. Chem. 2009, 74,
1070–1081.
18. Babu, B. R.; Raunak; Poopeiko, N. E.; Juhl, M.; Bond, A. D.; Parmar, V. S.; Wengel,
J. Eur. J. Org. Chem. 2005, 2297–2321.
19. Matsuda, A.; Yasuoka, J.; Sasaki, T.; Ueda, T. J. Med. Chem. 1991, 34, 999–1002.
20. Sekine, M. J. Org. Chem. 1989, 54, 2321–2326.
21. Ellervik, U.; Magnusson, G. Tetrahedron Lett. 1997, 38, 1627–1628.
22. See note 37 in Ref. 17b..
23. See Supplementary data..
The O20/O30-acyl groups (OAc/OBz) of all studied nucleosides
were cleaved using the optimized conditions, while most nucleo-
base-protecting groups, including ABz, CBz, GiBu, and the challenging
U
Bz, were stable to these conditions (Table 2). Only the highly labile
N-acetyl groups of cytidine and adenosine derivatives 5a and 7a,
respectively, were cleaved under these conditions. Interestingly,
O50-acyl groups (OAc/OBz) of purines were left unharmed (7a,
8a, and 10a), while they were cleaved in pyrimidine substrates
(4a and 5a). It was possible, however, to isolate O50-benzoylated
uridine derivative 9b albeit in very low yield. The commonly used
DMTr-protecting group was fully stable to these deacylation condi-
tions. Thus, selective O-deacylation of suitably protected pyrimi-
dines proceeds in excellent yields (85–95%, Table 2), while
selective O20,O30-deacylation of peracylated purines proceeded in
good yield (60–80%, Table 2).
To sum up, the guanidinium nitrate/sodium methoxide reagent
mixture presented herein facilitates selective O-deacylation of
ribonucleosides with N-acylated nucleobase moieties (ABz, CBz
G
,
iBu, TBz, and UBz). A special selectivity is observed with peracylat-
ed purines as O20,O30-deacylated products are exclusively formed.
The examples presented herein (Scheme 1 and Table 2) underline
that the guanidinium nitrate/sodium methoxide reagent mixture
is a valuable addition to the collection of protecting group proto-
cols in nucleoside chemistry.
24. Kamaike, K.; Uemura, F.; Yamakage, S.; Nishino, S.; Ishido, Y. Nucleosides
Nucleotides 1987, 6, 699–736.
25. Nishino, S.; Takamura, H.; Ishido, Y. Tetrahedron 1986, 42, 1995–2004.
Representative deacylation protocol: A stock solution of guanidi-
nium nitrate (1.24 g, 10 mmol) and sodium methoxide (108 mg,