Nucleophilic substitution at the anomeric position of 1,2-O-isopropylidenefuranose derivatives. A novel stereoselective synthesis of cyclic phosphates analogous to cAMP

Article abstract of DOI:10.1016/j.carres.2006.10.015

1,2-O-Isopropylidenefuranose derivatives were treated with various nucleophiles in the presence of either BF₃·OEt₂ or trimethylsilyl trifluoromethanesulfonate (TMSOTf) leading to substitution products in a regio- and stereoselective

Full text of DOI:10.1016/j.carres.2006.10.015

Carbohydrate Research 341 (2006) 2883–2890
Nucleophilic substitution at the anomeric position of
1,2-O-isopropylidenefuranose derivatives. A novel
stereoselective synthesis of cyclic phosphates analogous to cAMP
*
´
´
Miriam Romero, Luıs Hernandez, Leticia Quintero and Fernando Sartillo-Piscil
Centro de Investigacio´n de la Facultad de Ciencias Quı´micas, Beneme´rita Universidad Auto´noma de Puebla,
14 Sur Esq. San Claudio, Col. San Manuel, 72570 Puebla, Mexico
Received 17 August 2006; received in revised form 8 September 2006; accepted 13 October 2006
Available online 20 October 2006
This work is dedicated to the memory of our friend Pilar Peregrina (Pily)
Abstract—1,2-O-Isopropylidenefuranose derivatives were treated with various nucleophiles in the presence of either BF3ÆOEt2 or
trimethylsilyl trifluoromethanesulfonate (TMSOTf) leading to substitution products in a regio- and stereoselective manner. In par-
ticular, nucleophilic substitution of 1,2-O-isopropylidenefuranose derivatives when treated with allyltrimethylsilane was controlled
by steric and electronic factors (similar to Woerpel’s stereoelectronic model). On the other hand, when 1,2-O-isopropylidenefuranose
derivatives were treated with trimethylsilane, in the presence of bis-O-trimethylsilyl-5-iodouracil or bis-O-trimethylsilyl-thymidine,
substitution products were generated in high regio- and stereoselectivities via an unusual nucleophilic substitution with opening of
the furanose ring. Based on these results, a stereoselective method for the synthesis of neutral cyclic phosphates analogous to cAMP
was developed.
Ó 2006 Elsevier Ltd. All rights reserved.
Keywords: Nucleosides; Analogous cAMP; Nucleophilic substitution; Oxacarbenium
1. Introduction
on the nucleophilic attack on the oxacarbenium ion,
which follows (mainly) the Felkin–Anh model.3a
Although this model has been used to explain several
nucleophilic substitution reactions,4,2d,g in some cases
it fails to explain the nucleophilic substitution at the
anomeric center of the carbohydrate acetals.5 More
recently, Woerpel has reported a more elegant stereo-
chemical model.6 Contrary to Reissig, the Woerpel mod-
el is based exclusively on the particular properties of the
five-membered ring oxacarbenium ion, while the sol-
vent, counterion effect and nucleophilic properties are
ignored.
The Woerpel model proposes that a cyclic five-mem-
bered oxacarbenium ion generated from the ionization
of an anomeric C–X bond (X = acetate or halogen) of
the corresponding furanose derivative, preferentially
adopts an envelope conformation, where the C@O+
segment resides in the flattened portion. There are two
Nucleophilic substitution by either carbon or nitrogen
nucleophiles at the anomeric position of carbohydrate
derivatives is a powerful method for the synthesis of tetra-
hydrofurans and nucleosides.1 A number of total syn-
thesis of important naturally occurring compounds have
been accomplished by using this reaction.2 Generally,
the nucleophilic substitution occurs via the formation
of a five-membered ring oxacarbenium ion followed by
the stereocontrolled addition of the respective nucleo-
phile to the cation center. There are two stereochemical
models that can explain the origin of this stereoselecti-
vity. The first model as reported by Reissig,3 is based
*
Corresponding author. Tel.: +52 2222 295500x7387; fax: +52 2222
454293; e-mail: fsarpis@siu.buap.mx
0008-6215/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2006.10.015

2884
M. Romero et al. / Carbohydrate Research 341 (2006) 2883–2890
inside
attack
outside
attack
R
R
O
X
O
X
R
Lewis acid
O
R
R
A
R
B
Ribofuranose
derivatives
inside
attack
outside
attack
X = acetates or halogens
R = Alkyls or alkoxys
inside
attack
outside
attack
R
R
R
O
X
O
X
R
Lewis acid
R
O
B'
R
A'
Xylofuranose
derivatives
inside
attack
outside
attack
Scheme 1. Woerpel model for the stereocontrolled attack of nucleophiles on cyclic five-membered oxacarbenium ions.
well-defined diastereotopic faces that can be attacked by
a suitable nucleophile, termed ‘the inside face’ or ‘the
outside face’. Nucleophilic attack on the ‘inside face’ is
the preferred pathway due to less steric congestion,
and having the lowest conformational energy of the final
product (Scheme 1).
2. Results and discussion
We first tested the nucleophilic substitution reaction of
1,2-O-isopropylidenefuranose derivative 3 using allyltri-
methylsilane as the nucleophile in the presence of
BF3ÆOEt2 as the Lewis acid.9 Thus, xylofuranose deri-
vative 3 was treated under these conditions yielding a
mixture of substitution products 4a (major) and 4b
(minor) in good yield and stereoselectivity (Scheme 2).
The stereochemical assignment of the major product
4a, was determined by 2D-NOESY spectroscopic stu-
dies, and was shown to be the cis product as predicted
by the Woerpel model.6 The stereoselective formation
of major product 4a is due not only to the stereocon-
trolled nucleophilic attack of allyltrimethylsilane on
the inside face of C, but also to the nucleophilic attack
on the outside face of D (Scheme 3). It seems that the
internal counterion Xꢀ in either C or D did not exert
additional influence, either on the stereochemical out-
come or reactivity. We postulate that the counterion
(2-O-isopropyltrifluoroxyborate)10 does not exert addi-
tional influence based on Woerpel’s observation under
similar conditions but different substrates (yields of
85%, and stereoselectivies of 90:10).6a
Both conformers A and B exist in a rapid equilibrium,
and the approach of the nucleophile on the inside face of
A and B would result in erosion of the overall stereo-
selectivity. Interestingly this conformational equilibrium
is controlled by the R group at C-3.7 When the R is an
alkyl group, nucleophilic attack occurs preferentially on
the inside face of A, and when R is an alkoxy group, the
nucleophilic approach occurs preferentially on the inside
face of B.6,7 It is also important to note that a small pref-
erence for the nucleophilic approach on the outside of
A0 or B0 also contributes to the overall stereoselectivity.
Consequently, the counterion (Xꢀ) and the solvent exert
a minimum effect on the stereochemical outcome.6 This
suggests that the cation center and the counterion are
found as free ions, however this is not common in non-
polar solvents, like toluene.8 Furthermore, there are few
reported cases describing the existence of radical cat-
ions, as contact ion pairs, in nonpolar media.8b,c We
present here results related to the nucleophilic substitu-
tion of 1,2-O-isopropylidenefuranose derivatives, and
their application to the synthesis of neutral cyclic phos-
phates analogous to cAMP.
We then treated cyclic phosphates 5 and 6 with both
Et3SiH and allyltrimethylsilane in the presence of
BF3ÆOEt2 as the Lewis acid (Scheme 4).11 We selected
these cyclic phosphates containing the 1,2-O-isopropyl-
OBn
O
BnO
BnO
SiMe3
O
O
O
BnO
+
O
BF3.OEt2
88%
4a/4b = 91/9
3
BnO
OH
BnO
OH
4b
4a
Scheme 2. Stereoselective allylation of 1,2-O-isopropylidenefuranose derivatives.

M. Romero et al. / Carbohydrate Research 341 (2006) 2883–2890
2885
corresponding substitution products (not shown). In
the case of 5, its lack of reactivity may be attributed to
the possible anchimeric assistance by the phosphoryl
group toward the cation center of E (Robins11 observed
similar anchimeric assistance with 3-O-benzoyl group
toward a cation center). On the other hand, the lack
of reactivity of 6 may be due to the steric hindrance
imposed by the phenoxy group on the favorable ‘inside
face’ of the oxacarbenium ion F. This steric hindrance is
increased by the chair M boat equilibrium (6c M 6b)
that which kind of six-membered ring phosphates have
been shown to exist in about equimolar amount in solu-
tion and in solid state (Scheme 5).12,13
Therefore, if cyclic phosphates 5 and 6 did not react
because of the reasons mentioned above, then cyclic
phosphates 7 and 8 (where neither the anchimeric assis-
tance nor the steric hindrance are present) should readily
undergo nucleophilic attack (mainly on the inside face of
the corresponding five-membered ring oxacarbenium
ion). To test this possibility, cyclic phosphates 7 and 8
were prepared by standard methods,14 and treated
with Et3SiH or allyltrimethylsilane in the presence of
BF3ÆOEt2. For the reaction of 7 and 8 with Et3SiH/
BF3ÆOEt2, an unusual furanose-ring opening occurred
in good and moderate yields, respectively (compounds
9 and 10). On the other hand, when 7 and 8 were
allowed to react with allyltrimethylsilane/BF3ÆOEt2,
Nu
Bn
O
OBn
O
outside attack
4b
inside attack
4a
minor
major
O
OBF3
Nu
C
BnO
O
Nu
Nu
O
O
Bn
OBF3
D
Scheme 3. Nucleophilic substitution model of 1,2-O-isopropylidene-
furanose derivative 3.
idene group to compare their reactivity toward nucleo-
philes with compound 3 and others as reported previ-
ously.10,11 This preliminary study would provide
valuable information for the design of a novel N-glyco-
sylation reaction.
After many attempts (including longer reaction times,
an excess of the nucleophiles, and different tempera-
tures) cyclic phosphates 5 and 6 failed to provide the
O
O
O
Nu
P
O
No reaction
O
OPh
O
BF3·OEt2
5
Nu = Et3SiH, Allyltrimethylsilane
O
O
O
Nu
P
No reaction
O
O
OPh
O
BF3·OEt2
6
Scheme 4. Nucleophilic substitution of cyclic phosphates.
O
OPh
O
O
Nu
P
PhO
P
O
O
O
O
O
O
BF3·OEt2
O
P
O
O
O
O
O
OPh
O
OBF3
O
5
E
O
Nu
O
O
OPh
O
P
OPh
O
O
O
P
O
O
O
O
O
O
BF3·OEt2
PhO
O
O
O
P
P
O
O
O
OPh
O
O
O
OBF3
O
6c
6c
6
F
Scheme 5. Cyclic phosphates 5 and 6, and their corresponding oxacarbenium ions E and F.

2886
M. Romero et al. / Carbohydrate Research 341 (2006) 2883–2890
O
O
OH
O
P
O
P
Et3SiH
BF3·OEt2
92%
O
O
O
O
PhO
O
O
PhO
7
OH
9
O
O
P
OH
O
P
O
Et3SiH
O
O
O
O
O
PhO
BF3·OEt2
O
PhO
OH
8
7
10
57%
O
O
O
SiMe3
O
O
P
O
P
O
P
O
+
O
O
PhO
O
O
O
O
BF3·OEt2
86%
11a/11b = 76/24
PhO
OH
PhO
OH
11b
11a
O
O
SiMe3
O
P
O
No reaction
O
PhO
O
BF3·OEt2
8
Scheme 6. Nucleophilic substitution at anomeric positions of 7 and 8.
only 7 reacted to give 11a and 11b with moderate stereo-
selectivity (11a and 11b: 76/24, respectively, see Scheme
6). This moderate stereoselectivity is very similar to that
observed by Woerpel in different six–five bicyclic
acetates.6b Based on these observations, the existence
of an additional oxacarbenium ion H is evident (Scheme
7).15
This unusual bond ionization in H may be promoted
by the energetically-favorable release of steric strain
from the trans-fused bicyclic system (e.g., Gerlt found
by thermochemical experiments that cAMP is 8 kcal/
mol more exothermic than its acyclic analogue, and
5 kcal/mol of the 8 kcal/mol can be explained by geo-
metric strain resulting from the trans-fusion).16 There-
fore, the difference in reactivity between 7 and 8
toward nucleophilic attack may depend not only on
the intrinsic properties of the oxacarbenium ions G or
H, but also on the steric properties of the nucleophiles.
In this sense, it is important to note that Et3SiH is steri-
cally more demanding than allyltrimethylsilane,
although steric and stereoelectronic factors certainly
contribute to explain the difference in reactivity of cyclic
OPh
OPh
OPh
P
P
O
O
P
O
O
O
O
O
O
BF3·OEt2
O
O
BF3·OEt2
O
OBF3
O
O
O
OBF3
O
O
G
7
or
8
H
Et
SiMe3
H
Si Et
Et
11a and 11b
9 and 10
O
P
O
OBF3
O
PhO
O
O
OBF3
O
O
P
OPh
O
O
O
H
H'
Scheme 7. Possible oxacarbenium ions generated from 7 or 8.

M. Romero et al. / Carbohydrate Research 341 (2006) 2883–2890
2887
phosphates 7 and 8. Additionally, it is also important to
note that 7 is considerably more stable than 8 due to the
axial orientation of the phenoxy group.12,13,17 Further
work is required to confirm these assumptions. This
interesting problem will be addressed in due course.
Hence, the difference in reactivity between 7 and 8
may also be attributed to a very significant difference
in the conformational stability of H. In the case of 8,
its respective oxacarbenium ion H is destabilized by
the spontaneous interconversion to H0 (the phenoxy
group tends to reach the axial position due to the ano-
meric effect,12,13,17 Scheme 7). It is important to note
that to obtain 9 and 10 from H (or may be from H0),
a second regioselective bond ionization followed by an-
other hydride addition is necessary. A similar regioselec-
tive ring-opening of a terminal 1,2-O-isopropylidene
group by a silane reagent to give 1,3-diols instead of
1,4-diols (not shown) has been recently reported by
our group.18
We next turned our attention to the glycosylation
reaction of the cyclic phosphates with nucleobases to
obtain neutral cyclic phosphates (analogues of cAMP
or cGMP). To this end, cyclic phosphates 5–8 were
allowed to react with bis-O-trimethylsilyl-5-iodouracil
or bis-O-trimethylsilyl-thymidine using TMSOTf as a
Lewis acid catalyst.19 Again, 5 and 6 were recovered
quantitatively, and 7 and 8 afforded neutral cyclic nucleo-
tides with the opening at the furanose ring and the
retention of the 1,2-O-isopropylidene group 12–15
(Table 1).
from the intensity of the anomeric hydrogen signals. Iso-
lation of this product in good yield from the complex
mixture reaction was not initially possible due to the
apparent instability of the product. However, repeated
chromatography on silica gel enabled isolation of the
major product in very small amount for characterization
(in the 1H spectrum, the anomeric hydrogen appears at d
6.29 ppm (3JH–H = 5.4 Hz), two singlets at d 1.55 and
1.49 ppm are attributed to the methyl of the 1,2-O-iso-
propylidene group, and in the 31P spectrum a singlet
at d ꢀ12.1 ppm is observed).
The stereochemistry of the new stereogenic center
formed in 12–15 was determined by detailed 2D
NOESY experiments, which showed a correlation
between H-10 and H-20, each with one of the two isoprop-
ylidene methyl groups, respectively. Additionally, the
methyl group that interacts with H-20 shows a correla-
tion with H-6 of the base. Apparently this base glycosyl-
ation reaction followed the same route as the
nucleophilic attack of the Et3SiH on oxacarbenium ion
H. This is in agreement with the suggestion that steric
factors favor the opening of the furanose ring. Thus,
based on the stereochemistry of the nucleotides 12–15,
this novel highly stereoselective base-coupling reaction
can be rationalized in terms of a stereocontrolled nucleo-
philic addition of silylated bases on the more favorable
face of the oxacarbenium ion I (Scheme 8).
This novel highly stereoselective base-coupling reac-
tion, which occurs with the opening of the furanose ring
and retention of the 1,2-O-isopropylidene group (which
might be considered as a furanose mimic) represents a
very convenient way to synthesize novel cyclic phos-
phate analogues of cAMP or cGMP. To the best of
our knowledge, this is the first reported occurrence of
a 1,2-O-isopropylidene group that has been used in a
stereoselective nucleobase glycosylation reaction.
As shown in Table 1, exposure of phosphates 7 and 8
to the respective silylated nucleobases in the presence of
TMSOTf resulted in an efficient and highly stereoselec-
tive glycosylation reaction (entries 1–3). In contrast, a
dramatic decrease in yield as well as in stereoselectivity
was observed for 8 with thymine (entry 4). In this case,
1H NMR spectroscopy revealed the formation of a com-
plex reaction mixture whereby one nucleoside predomi-
nated to the extent of approximately 45% as judged
3. Conclusions
It was demonstrated herein that nucleophilic substitu-
tion at the anomeric position of 1,2-O-isopropylidene-
furanose derivatives occurs with high regioselectivity
and stereoselectivity. The product formed depends on
the nature of the nucleophile and the conformational
restrictions of the five-membered ring oxacarbenium
ion. In addition to being of mechanistic interest, the
development of this novel stereoselective N-glycosyl-
ation reaction offers a new access to novel cyclic phos-
phate analogues of cAMP and other important
nucleotides (e.g., prodrugs). In this regard, novel nucleo-
tides with biological activity are currently being synthe-
sized in our laboratory by applying the methodology
described herein. Although the experimental results
Table 1. Stereoselective coupling reaction with silylated nucleobasesa,b
B
OH
O
O
O
O
O
Silylated nucleobases
P
P
PhO
O
O
TMSOTf
CH3CN
PhO
O
O
O
O
7 or 8
12-15
Entry
Phosphate
Base
Nucleotide (yield, %)
1
2
3
4
7
7
8
8
5-Iodouracil
Thymine
5-Iodouracil
Thymine
12 (88)
13 (85)
14 (87)
15 (45)c
a Only one diastereoisomer was observed by 1H NMR and 31P NMR
spectroscopy.
b Yield of products after flash chromatography.
c Yield determined by NMR.
indicate
a similar behavior to that reported by
others, a number of interesting questions related to the

2888
M. Romero et al. / Carbohydrate Research 341 (2006) 2883–2890
Attack favoured
OTMS
R
R = I, Me
N
TMSO
N
OPh
P
TfO
H
O
O
12-15
O
OSiMe3
O
O
O
H
O
OSiMe3
O
TfO
O
P
O
I
PhO
I
Scheme 8. Stereoselective base-addition to oxacarbenium ion I.
mechanism of nulceophilic substitution reactions of the
cyclic phosphates remain unsolved, and further studies
on this matter are still needed.
(d, 1H, J = 3.3 Hz), 7.18–7.39 (m, 5H); 13C NMR
(75.4 MHz, CDCl3): d 26.1, 68.3 (JCP = 5.7 Hz), 70.0
(JCP = 9.2 Hz), 76.0 (JCP = 8.0 Hz), 80.1 (JCP = 5.7 Hz);
31P NMR (121 MHz, CDCl3): d ꢀ11.8 ppm.
4. Experimental
4.2.2. (SP)-1,2-O-Isopropylidene-3,5-O-phenoxyphosphor-
yl-a-D-ribofuranose (8). Syrup (41%); [a]D ꢀ8.1 (c 1.2,
1
4.1. General methods
CHCl3); H NMR (300 MHz, CDCl3): d 1.38 (s, 3H),
1.56 (s, 3H), 4.27–4.43 (m, 3H), 4.63 (m, 1H), 4.77 (t,
1H, J = 3.3 Hz), 5.90 (d, 1H, J = 3.3 Hz), 7.18–7.38 (m,
5H); 13C NMR (75.4 MHz, CDCl3): d 26.1, 68.0
Reagents were obtained from commercial sources and
used without purification. Solvents of technical grade
were used and freshly distilled prior to use. NMR stud-
ies were carried out on 400 and 300 MHz spectrometers;
(JCP = 7.95 Hz), 69.8 (JCP = 7.95 Hz), 76.1 (JCP
=
6.8 Hz), 79.1 (JCP = 4.6 Hz), 106.0, 114.6, 120.3, 125.7,
129.8; 31P NMR (121 MHz, CDCl3): d ꢀ9.7 ppm; FAB-
HRMS m/z Calcd for C14H17O7P [M+H]+: 329.07196.
Found: 329.07198.
1
tetramethylsilane was used as the reference for the H
and 13C NMR spectra and chemical shifts are stated
in parts per million. COSY, HSQC, and NOESY exper-
1
iments were carried out to assign fully the H and 13C
NMR. High resolution mass spectra were obtained by
fast atom bombardment ionization.
4.3. General procedure for the deprotection/reduction
reaction of 1,2-O-isopropylidene xylo- and ribofuranose
derivatives
4.2. Phosphorylation reactions
A solution of 1,2-O-isopropylidene xylo- and ribofura-
nose derivatives (2.0 mmol) in 50 mL of dry CH2Cl2 at
0 °C was treated with Et3SiH (12.0 mmol) and BF3ÆOEt2
(12.0 mmol) before it was warmed to room temperature,
and stirred for 2 h. The reaction mixture was treated
with a satd aq NaHCO3 (50 mL), the aq phase was ex-
tracted three times with CH2Cl2, dried with MgSO4
and the residue purified by column chromatography
on silica gel (neutral pH) with ether–EtOAc (v/v = 3/1).
For the six-membered ring phosphates 8 and 9, the
neutralization was carried out using a dilute aq NaH-
CO3 at 0 °C; otherwise, a messy reaction occurred.
To a solution of 1,2-O-isopropylidene-ribofuranose
(0.37 g, 1.96 mmol) and Et3N (0.81 mL, 4.9 mmol) was
added dropwise phenyldichlorophosphate (0.36 mL,
2.35 mmol) dissolved in 20 mL CH2Cl2 at 0 °C. The
reaction mixture was allowed to react for 4 h before it
was poured into water (50 mL). The aqueous layer was
then extracted three times with CH2Cl2 (50 mL) and
the combined organic phases were dried over MgSO4
and concentrated in vacuo, and the residue was purified
by column chromatography on silica gel (230–400 mesh)
with ether–EtOAc (v/v = 2/1).
4.2.1. (RP)-1,2-O-Isopropylidene-3,5-O-phenoxyphosphor-
yl-a-D-ribofuranose (7).14 White powder (51%);
decomposition temp = 172 °C; [a]D +50 (c 1.4, CHCl3);
1H NMR (300 MHz, CDCl3): d 1.37 (s, 3H), 1.67 (s,
3H), 4.25 (m, 1H), 4.35 (m, 2H), 4.64 (ddd, 1H,
J = 22.8, 10.2, 4.5 Hz), 4.76 (t, 1H, J = 3.6 Hz), 5.88
4.3.1. (RP,4S,5R,10S)-4-(1-Hydroxy-2-isopropoxy-ethyl)-
5-hydroxy-2-oxo-2-phenoxy-1,3,2-dioxaphosphorinane (9).
Syrup (92%); [a]D ꢀ17.1 (c 1.0, CHCl3); 1H NMR
(400 MHz, CDCl3): d 1.12 (d, 3H, J = 6.0 Hz), 1.17 (d,
3H, J = 6.4 Hz), 3.57–3.65 (m, 3H), 4.01 (dt, 1H,
J = 6.4, 3.5 Hz), 4.23 (m, 2H), 4.31 (m, 1H), 4.45 (m,

M. Romero et al. / Carbohydrate Research 341 (2006) 2883–2890
2889
1H), 7.18–7.38 (m, 5H); 13C NMR (100 MHz, CDCl3): d
21.8, 22.0, 64.6 (JCP = 6.1 Hz), 66.6, 70.2 (JCP
7.6 Hz), 72.8, 73.0 (JCP = 10.6), 80.5 (JCP = 7.6 Hz),
119.3, 119.2 (JCP = 6.1 Hz), 125.2, 129.8; 31P NMR
(121 MHz, CDCl3): d ꢀ12.0 ppm; MS (EI-mode) m/z
332, (M+ 26%); FAB-HRMS m/z Calcd for
C14H21O7P [M+H]+: 333.1103. Found: 333.1101.
J = 17.6, 2.8 Hz), 5.78 (m, 2H), 7.18–7.41 (m, 10H);
13C NMR (100 MHz, CDCl3): d 36.4, 38.6, 61.6, 62.9
(JCP = 6.1 Hz), 68.8, 70.3 (JCP = 10.6 Hz), 70.5
(JCP = 9.1 Hz), 70.9, 74.2 (JCP = 11.7 Hz), 82.3
(JCP = 10.7 Hz), 82.7 (10.7 Hz); 115.3, 118.4, 118.8,
119.5, 125.5, 130.0, 133.6; 31P NMR (121 MHz, CDCl3):
d ꢀ11.6, ꢀ11.3; MS m/z 313, (M+H, 4%); FAB-HRMS
m/z Calcd for C14H17O6P [M+H]+: 313.0841. Found:
313.0851.
=
4.3.2. (SP,4S,5R,10S)-4-(1-Hydroxy-2-isopropoxy-ethyl)-
5-hydroxy-2-oxo-2-phenoxy-1,3,2-dioxaphosphorinane (10).
Syrup (57%); [a]D ꢀ29.7 (c 0.8, CHCl3); 1H NMR
(400 MHz, CDCl3): d 1.16 (d, 3H, J = 6.0 Hz), 1.18 (s,
3H, J = 6.0 Hz), 2.91 (br, 1H), 3.56 (dd, 1H, J = 9.9,
4.2 Hz), 3.64 (m, 2H), 3.99 (m, 1H), 4.24 (m, 1H), 4.33–
4.57 (m, 3H), 7.16–7.37 (m, 5H); 13C NMR (100 MHz,
CDCl3): d 22.0, 22.1, 64.7 (JCP = 6.8 Hz), 67.4, 70.5
(JCP = 5.7 Hz), 71.4 (JCP = 5.7 Hz), 73.0, 83.2
(JCP = 5.7 Hz), 120.0, 125.7, 130.1; 31P NMR (121 MHz,
CDCl3): d ꢀ11.3 ppm; MS (EI-mode) m/z 332 (M+
15%); HRMS (FAB-mode) m/z Calcd for C14H21O7P
[M+H]+: 333.1103. Found: 333.1109.
4.5. General procedure for the nucleobase-coupling
reaction
The nucleobase (0.72 mmol) and 1,1,1,3,3,3-hexamethyl-
disilazane (2 mL) were stirred at 100 °C for 8 h. The
reaction mixture was cooled to room temperature and
cyclic phosphate (0.6 mmol) and TMSOTf (0.92 mmol)
in freshly distilled acetonitrile were added. The reaction
mixture was then stirred until the disappearance of the
respective starting material. The reaction mixture was
treated with a diluted aq soln NaHCO3 (30 mL). The
aq layer was extracted three times with CH2Cl2
(50mL), the organic phase was dried with MgSO4, con-
centrated in vacuo and the residue was purified by col-
umn chromatography on silica gel (230–400 mesh).
4.4. General procedure for the allylation reaction of 1,2-
O-isopropylidene xylo- and ribofuranose derivatives
A solution of 1,2-O-isopropylidene xylo- and ribofura-
nose derivatives (2.0 mmol) in 50 mL of dry CH2Cl2 at
0 °C was treated with allyltrimethylsilane (3.0 mmol)
and BF3ÆOEt2 (3.0 mmol) (for the six-membered ring
phosphates, 10.0 mmol of allyltrimethylsilane and
BF3ÆOEt2 were used). The reaction mixture was warmed
to room temperature over 4 h. The reaction mixture was
treated with satd aq NaHCO3 (50 mL). The aq layer was
extracted three times with CH2Cl2 (50 mL). The organic
phase was dried with MgSO4, concentrated in vacuo and
the residue was purified by column chromatography on
silica gel (230–400 mesh).
4.5.1. (RP,4R,5R,10S,2R)-4[1-(5-Iodouracil)-1,2-O-iso-
propylidene-ethyl]-5-hydroxy-2-oxo-2-phenoxy-1,3,2-di-
oxaphosphorinane (12). Syrup (88%); [a]D +10.7 (c 0.9,
1
CHCl3); H NMR (400 MHz, CDCl3/CD3OD): d 1.51
(s, 3H), 1.58 (s, 3H), 4.17 (td, 1H, J = 10.4, 4.4 Hz),
4.22 (td, 1H, J = 10.0, 2.8 Hz), 4.37 (ddd, 1H,
J = 24.0, 10.0, 4.4 Hz), 4.57 (ddd, 1H, J = 5.6, 3.2,
2.0 Hz), 4.78 (dt, 1H, J = 9.6, 2.0 Hz), 6.33 (d, 1H,
J = 5.6 Hz), 7.24–37 (m, 5H), 7.50 (s, 1H), 7.91 (s,
1H); 13C NMR (75 MHz, CDCl3/CD3OD): d 27.8,
27.1, 61.6 (JCP = 4.6 Hz), 69.0, 70.6 (JCP = 6.8 Hz),
80.9 (JCP = 7.9 Hz), 81.2 (JCP = 6.8 Hz), 83.4, 131.1,
119.5 (JCP = 4.5 Hz), 125.4, 129.7, 143.8, 149.8, 150.4,
160.5; 31P NMR (121 MHz, CDCl3): d ꢀ11.7 ppm;
FAB-HRMS m/z Calcd for C18H21IN2O9P [M+H]+:
567.0029. Found: 567.0027.
4.4.1. (2S,3R,4R,5R)-2-Allyl-4-(benzyloxy)-5-((benzyl-
oxy)methyl)-tetrahydrofuran-3-ol (4a, major stereoiso-
mer). Syrup (80%); [a]D ꢀ20.2 (c 1.0, CHCl3); 1H
NMR (300 MHz, CDCl3): d 2.38 (m, 1H), 2.48 (m,
1H), 3.70 (m, 3H), 3.87 (dd, 1H, J = 5.2, 2.4 Hz), 3.99
(dd, 1H, J = 4.4, 2.2 Hz), 4.23 (m, 1H), 4.49–4.62 (m,
4H), 5.08 (2H, m), 5.84 (1H, m), 7.18–7.41 (10H, m);
13C NMR (75 MHz, CDCl3): d 38.0, 68.7, 71.7, 73.3,
79.0, 79.1, 83.8, 85.2, 117.3, 127.3, 127.4, 127.5, 127.6,
128.1, 128.2, 134.1, 137.7. Anal. Calcd for C22H26O4:
C, 74.55; H, 7.39. Found: C, 74.51; H, 7.38.
4.5.2. (RP,4R,5R,10S,2R)-4[1-(Thymidine)-1,2-O-isoprop-
ylidene-ethyl]-5-hydroxy-2-oxo-2-phenoxy-1,3,2-dioxaphos-
phorinane (13). Syrup (85%); [a]D +19.2 (c 1.0,
1
CHCl3); H NMR (400 MHz, CDCl3): d 1.48 (s, 3H),
1.52 (s, 3H), 1.91 (s, 3H), 4.21 (td, 1H, J = 13.0,
1.2 Hz), 4.41 (ddd, 1H, J = 24.2, 10.4, 4.8 Hz), 4.49
(m, 1H), 4.59 (dt, 1H, J = 10.4, 1.2 Hz), 4.91 (m, 1H),
4.96 (d, 1H, J = 4.0 Hz), 6.13 (d, 1H, J = 3.2 Hz),
7.16–7.35 (m, 2H), 7.42 (d, 1H, J = 1.2 Hz), 10.61 (s,
1H); 13C NMR (75 MHz, CDCl3): d 12.6, 27.3, 28.0,
62.1 (JCP = 4.6 Hz), 70.9 (JCP = 6.9 Hz), 81.8
(JCP = 9.0 Hz), 83.3 (JCP = 7.9 Hz), 85.2, 111.3, 114.6,
119.7, 125.4, 129.9, 135.3, 151.7, 164.3; 31P NMR
4.4.2. (RP)-1-Allyl-1-desoxy-2-O-isopropoxy-3,5-O-phen-
oxyphosphoryl-a-D-ribofuranose (11a and 11b 76/24,
respectively). Data reported as a mixture of stereoiso-
mers: Syrup (86%); 1H NMR (400 MHz, CDCl3): d
2.29–2.46 (m, 4H), 3.84 (m, 2H), 3.95 (m, 2H), 4.19–
4.49 (m, 8H), 5.01 (d, 2H, J = 17.6 Hz), 5.15 (dd, 2H,

2890
M. Romero et al. / Carbohydrate Research 341 (2006) 2883–2890
D. J.; Ley, S. V. Angew. Chem., Int. Ed. 2005, 44, 580–584;
(g) Colby, E. A.; O’Brien, K. C.; Jamison, T. F. J. Am.
Chem. Soc. 2005, 127, 4297–4307.
(121 MHz, CDCl3): d ꢀ11.8 ppm; FAB-HRMS m/z
Calcd for C19H24N2O9P [M+H]+: 455.1219. Found:
455.1211.
3. (a) Schmitt, A.; Reissig, H.-U. Synlett 1990, 40–42; (b)
Schmitt, A.; Reissig, H.-U. Chem. Ber. 1995, 128, 871–876;
(c) Schmitt, A.; Reissig, H.-U. Eur. J. Org. Chem. 2000,
3893–3901; (d) Schmitt, A.; Reissig, H.-U. Eur. J. Org.
Chem. 2001, 1169–1174.
4. Alonso, E.; Ramon, D. J.; Yus, M. Tetrahedron 1997, 53,
2641–2652.
5. Araki, Y.; Kobayashi, N.; Ishido, Y.; Nagasawa, J.
Carbohydr. Res. 1987, 145–148.
6. (a) Larsen, C. H.; Ridgway, B. H.; Shaw, T. J.; Woerpel,
K. A. J. Am. Chem. Soc. 1999, 121, 12208–12209;
(b) Smith, D. M.; Tran, M. B.; Woerpel, K. A. J. Am.
Chem. Soc. 2003, 125, 14149–14152.
7. (a) Larsen, C. H.; Ridgway, B. H.; Shaw, J. T.; Smith, D.
M.; Woerpel, K. A. J. Am. Chem. Soc. 2005, 127, 10879–
10884; (b) Smith, D. M.; Woerpel, K. A. Org. Lett. 2004,
6, 2063–2066.
8. (a) Raber, D. J.; Harris, J. M.; Schleyer, P. v. R. In Ions
and Ions Pairs in Organic Reactions; Szwarc, M., Ed.;
Wiley: New York, 1974; Vol. 2; (b) Crich, D.; Ranga-
nathan, K. J. Am. Chem. Soc. 2005, 127, 9924–9929;
(c) Sartillo-Piscil, F.; Vargas, M.; Anaya de Parrodi, C.;
Quintero, L. Tetrahedron Lett. 2003, 44, 3919–3921.
4.5.3.
(SP,4R,5R,10S,2R)-4[1-(5-Iodouracil)-1,2-O-iso-
propylidene-ethyl]-5-hydroxy-2-oxo-2-phenoxy-1,3,2-di-
oxaphosphorinane (14). Syrup (87%); [a]D +11.2 (c 1.0,
CHCl3); 1H NMR (400 MHz, DMSO-d6): d 1.43 (s, 3H),
1.50 (s, 3H), 3.90 (br s, 1H), 4.11 (m, 1H), 4.27 (dd, 1H,
J = 21.6, 12.6 Hz), 4.68 (dd, 1H, J = 17.2, 5.2 Hz), 5.91
(d, 1H, J = 4.4 Hz), 6.27 (d, 1H, J = 5.2 Hz), 7.08–7.29
(m, 5H), 7.97 (s, 1H); 1H NMR (100 MHz, DMSO-
d6): d 27.9, 63.1, 68.1, 70.5, 71.9, 76.2, 85.1, 86.0,
111.9, 120.3, 125.6, 130.2, 145.4, 147.4, 150.2; 31P
NMR (121 MHz, CDCl3): d ꢀ13.3 ppm; FAB-HRMS
m/z Calcd for C18H21IN2O9P [M+H]+: 567.0029.
Found: 567.0025.
Acknowledgements
We thank CONACyT and BUAP-PROMEP for finan-
cial support. We also thank Dr. Xianhai Huang and
Dr. David Kim for helpful discussions.
´
9. Garcıa-Tellado, F.; De Armas, P.; Marrero-Tellado, J. J.
Angew. Chem., Int. Ed. 2000, 39, 2727–2729.
´
10. Benito, J. M.; Gomez-Garcıa, M.; Ortiz-Mellet, C.;
´
´
Garcıa-Fernandes, J. M.; Defaye, J. Org. Lett. 2001, 3,
549–552.
11. Ewing, G. J.; Robins, M. J. Org. Lett. 1999, 1, 635–636.
Supplementary data
´
12. Sartillo-Piscil, F.; Cruz-Gregorio, S.; Sanchez, M.; Ho¨pfl,
H.; Anaya de Parrodi, C.; Quintero, L. Tetrahedron 2003,
59, 4077–4083.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.carres.2006.10.015.
´
13. Sartillo-Piscil, F.; Cruz-Gregorio, S.; Sanchez, M.; Quin-
tero, L. Tetrahedron 2004, 60, 3001–3008.
14. Neeser, J. R.; Tronchet, J. M. J.; Charollais, E. J. Can. J.
Chem. 1983, 61, 1387–1395.
15. Valdivia, V.; Hernandez, A.; Rivera, A.; Sartillo-Piscil, F.;
Loukaci, A.; Fourrey, J.-L.; Quintero, L. Tetrahedron
Lett. 2005, 46, 6511–6514.
References
16. (a) Gerlt, J. A.; Gutterson, N. I.; Datta, P.; Belleau, B.;
Penney, C. I. J. Am. Chem. Soc. 1980, 102, 1655–1665;
(b) Gerlt, J. A.; Gutterson, N. I.; Drews, R. E.; Sokolow,
J. A. J. Am. Chem. Soc. 1980, 102, 1655–1670.
17. Bentrude, W. G. Steric and Stereoelectrinic Effects in
1,3,2-Dioxaphosphorinanes. In Conformational Behavior
of Six-Membered Rings; Juaristi, E., Ed.; VCH: New
York, 1995.
18. Cruz-Gregorio, S.; Sanchez, M.; Clara-Sosa, A.; Bernes,
S.; Quintero, L.; Sartillo-Piscil, F. J. Org. Chem. 2005, 70,
7107–7113.
19. Vorbrueggen, H.; Krolikiewicz, K.; Bennua, B. Chem. Ber.
1981, 114, 1234–1255.
1. (a) Postema, M. H. D. C-Glycoside Synthesis; CRC Press:
Boca Raton, FL, 1995; (b) Vorbruggen, H.; Ruh-Pohlenz,
¨
C. Org. React. 2000, 55, 1.
2. See interesting applications in total synthesis: (a) Hori, K.;
Hikage, N.; Inagaki, A.; Mori, S.; Nombra, K.; Yoshii, E.
J. Org. Chem. 1992, 57, 2888–2902; (b) Zeng, L.; Ye, Q.;
Oberlies, N. H.; Shi, G.; Gu, Z.-M.; He, K.; McLaughlin,
J. L. Nat. Prod. Rep. 1996, 13, 275–306; (c) Ghosh, A. K.;
Liu, C. J. Am. Chem. Soc. 2003, 125, 2374–2375; (d) A¨ıssa,
C.; Riveiros, R.; Ragot, J.; Furstner, A. J. Am. Chem. Soc.
¨
2003, 125, 15512–15520; (e) Colby, E. A.; O’Brien, K. C.;
Jamison, T. F. J. Am. Chem. Soc. 2004, 126, 998–999;
(f) Nattrass, G. L.; Diez, E.; McLachlan, M. M.; Dixon,