A stereoselective cyclization to carbafuranose derivatives starting from 1,4-bis-epoxides

Article abstract of DOI:10.1021/ol052009h

(Chemical Equation Presented) A concise synthesis of highly functionalized cyclopentane derivatives via a Brook rearrangement mediated stereoselective linchpin cyclization reaction involving tert-butyldimethylsilyl-1,3- dithianyllithium and homochiral 1,4

Full text of DOI:10.1021/ol052009h

ORGANIC
LETTERS
2005
Vol. 7, No. 23
5183-5186
A Stereoselective Cyclization to
Carbafuranose Derivatives Starting from
1,4-Bis-epoxides
Leo M. H. Leung, A. James Boydell, Vicky Gibson,† Mark E. Light, and
Bruno Linclau*
School of Chemistry, UniVersity of Southampton, Highfield, Southampton SO17 1BJ,
United Kingdom, and CMS Chemicals Ltd., 9 Milton Park, Abingdon,
Oxfordshire OX14 4RR, United Kingdom
bruno.linclau@soton.ac.uk
Received August 19, 2005
ABSTRACT
A concise synthesis of highly functionalized cyclopentane derivatives via a Brook rearrangement mediated stereoselective linchpin cyclization
reaction involving tert-butyldimethylsilyl-1,3-dithianyllithium and homochiral 1,4-bis-epoxides is described.
Nucleosides are a pharmacologically important class of
compounds, with many examples currently in clinical use.1
Carbocyclic nucleosides have been the subject of intense
research.2 Successful examples of such nucleosides include
abacavir (Figure 1),1,3 approved as an anti-HIV drug, and
Many approaches have been reported for the synthesis of
carbafuranose sugars as precursors for carbanucleosides,2 and
as the synthesis of novel antiviral drugs is of prime
importance,6 there is a continuing interest for new synthetic
methodologies to access carbanucleosides in enantiomerically
pure form, ideally with both enantiomers equally accessible.
We wish to report a very short synthesis of highly
functionalized enantiopure cyclopentane derivatives 3 (Scheme
1), which would be suitable precursors for 6′-substituted
(2) (a) Amblard, F.; Nolan, S. P.; Agrofolio, L. A. Tetrahedron 2005,
61, 7067. (b) Rodriguez, J. B.; Comin, M. J. Mini ReV. Med. Chem. 2003,
3, 95. (c) Agrofolio, L. A.; Gillaizeau, I.; Saito, Y. Chem. ReV. 2003, 103,
1875. (d) Schneller, S. W. Curr. Top. Med. Chem. 2002, 2, 1807. (e) Ferrero,
M.; Gotor, V. Chem. ReV. 2000, 100, 4319. (f) Crimmins, M. T. Tetrahedron
1998, 54, 9229. (g) Borthwick, A. D.; Biggadike, K. Tetrahedron 1992,
48, 571.
Figure 1. Examples of carbanucleosides.
entecavir,1,4 which is currently undergoing phase III clinical
trials for the treatment of chronic hepatitis B virus infections.
In addition, antimalarial activity has been reported for certain
carbanucleoside analogues.5
(3) Hervey, P. S.; Perry, C. M. Drugs 2000, 60, 447.
(4) Levine, S.; Hernandez, D.; Yamanaka, G.; Zhang, S.; Rose, R.;
Weinheimer, S.; Colonno, R. J. Antimicrob. Agents Chemother. 2002, 46,
2525.
(5) Shuto, S.; Minakawa, N.; Niizuma, S.; Kim, H.-S.; Wataya, Y,;
Matsuda, A. J. Med. Chem. 2002, 45, 748.
(6) (a) El Kouni, M. H. Curr. Pharm. Design 2002, 8, 581. (b) Richman,
D. D. Nature 2001, 410, 995. (c) Wang, J.; Jin, Y.; Rapp, K. L.; Bennett,
M.; Schinazi, R. F.; Chu, C. K. J. Med. Chem. 2005, 48, 3736.
† CMS Chemicals Ltd.
(1) De Clercq, E. J. Clin. Virol. 2004, 30, 115.
10.1021/ol052009h CCC: $30.25
© 2005 American Chemical Society
Published on Web 10/15/2005

carbanucleoside analogues 4,7 via a stereoselective linchpin
cyclization reaction involving bis-epoxide 1 and dithianyl-
lithium 2. Bis-epoxide 1 is derived from arabitol in two
operations,8 and as the arabitol enantiomers are available at
similar cost, both D- and L-carbanucleoside derivatives would
be easily accessible.
Scheme 2. Mechanism of the Carbacyclization
Scheme 1. Synthesis Outline
would then be followed by 5-exo or 6-exo (7-endo)15
cyclization. Literature precedence for carbanionic cyclization
to δ-terminal epoxides suggest that the regioselectivity is
dependent on the metal ion, with lithiated species favoring
5-exo ring closure.16 Hence, the desired diastereomers 3R
and 3â were expected as main reaction products.
In the event, we were delighted to observe that reaction
of 2, obtained by tert-BuLi-mediated deprotonation of 2-tert-
butyldimethylsilyl-1,3-dithiane 9, with 1 proceeded as de-
picted above with high 5-exo selectivity (Table 1). The
double-addition product 11 was only formed in very small
amounts in all cases. As expected, a mixture of cyclopentane
diastereoisomers was formed, and 3â proved to be the major
diastereoisomer (see below).
Varying the temperature (entries 1-4) revealed -45 to
-18 °C to be the optimum temperature range for the yield
of 3 (entries 2-3). With increasing temperature, a decrease
in the diastereoselectivity 3â/3R was observed as well as an
increase in the formation of the six-membered cyclization
product 10.
Decreasing the concentration to 0.05 M led to an increased
yield of 3 (entry 5 vs 2). Reducing the concentration further
led to decreased yields of 3, but a greater amount of 1 was
recovered (entries 6 and 7). Replacing THF with Et2O as
solvent led to almost identical yields of 3, but an increased
selectivity toward 3â was observed (entry 8 vs 2). The
efficiency of carbacyclization was found to be significantly
hampered when the HMPA/THF ratio was reduced to 1:30
(entry 9).
When conducting the cyclization at -30 °C, a higher yield
was observed (entry 10 vs 7). In the end, addition of
molecular sieves led to a dramatic rise in the cyclization
yields (entries 11 vs 10, 12 vs 7). Increasing the HMPA/
THF ratio to 1:9 furnished a 77% yield of 3 (entry 13), and
the use of a slight excess of 2 completed the optimization in
which 3â and 3R were obtained in 80% combined yield (3.2:
1), together with 8% of 10 (entry 14). Increasing the
concentration further to 0.05 M did not lead to improved
results (entry 15).
The reaction of dithianyl anion with chiral epoxide
electrophiles enables the construction of partially protected
aldol linkages in a stereospecific manner.9 Fully protected
aldol linkages can be directly obtained by utilizing silylated
dithiane nucleophiles via a Brook rearrangement.10 The
rearrangement regenerates a dithianyl anion, which enables
subsequent transformations (see below).11 Judicious control
over the timing of the Brook rearrangement12 fully estab-
lished dithiane-based “linchpin” one-pot, multicomponent
coupling strategies for the construction of complex targets.13
This process has also been investigated with mannitol-derived
1,5-bis-epoxides14 to afford cyclohexane and cycloheptane
derivatives.
With 3-benzyl-1,2:4,5-dianhydroarabitol 1, it was envi-
sioned that the linchpin cyclization process would proceed
following the accepted mechanism as shown in Scheme 2.
Ring opening of one of the diastereotopic epoxide groups
by 2 would lead to a diastereoisomeric mixture of alkoxides
5/7. HMPA-promoted Brook rearrangement to give 6/8
(7) Synthesis of 6′-substituted nucleosides: (a) Yang, Y.-Y.; Meng, W.-
D.; Qing, F.-L. Org. Lett. 2004, 6, 4257. (b) Hong, J. H.; Oh, C.-H.; Cho,
J.-H. Tetrahedron 2003, 59, 6103. (c) Bianco, A.; Celletti, L.; Mazzei, R.
A.; Umani, F. Eur. J. Org. Chem. 2001, 1331. (d) Wachtmeister, J.; Classon,
B.; Samuelsson, B.; Kvarnstro¨m, I. Tetrahedron 1997, 53, 1861. (e) Bisacchi,
G. S.; Chao, S. T.; Bachard, C.; Daris, J. P.; Innaimo, S.; Jacobs, G. A.;
Kocy, O.; Lapointe, P.; Martel, A.; Merchant, Z.; Slusarchyk, W. A.;
Sundeen, J. E.; Young, M. G.; Colonno, R.; Zahler, R. Bioorg. Med. Chem.
Lett. 1997, 7, 127. (f) Tanaka, M.; Norimine, Y.; Fujita, T.; Suemune, H.;
Sakai, K. J. Org. Chem. 1996, 61, 6952. (g) Altmann, K.-H.; Kesselring,
R.; Pieles, U. Tetrahedron 1996, 52, 12699. (h) Katagiri, N.; Nomura, M.;
Sato, H.; Kaneko, C.; Yusa, K.; Tsuruo, T. J. Med. Chem. 1992, 35, 1882.
(8) (a) Dreyer, G. B.; Boehm, J. C.; Chenera, B.; DesJarlais, R. L.;
Hassell, A. M.; Meek, T. D.; Tomaszek, T. A., Jr.; Lewis, M. Biochemistry
1993, 32, 937. (b) Schreiber, S. L.; Sammakia, T.; Uehling, D. E. J. Org.
Chem. 1989, 54, 15. (c) Leung, L. M. H.; Gibson, V.; Linclau, B.
Tetrahedron: Asymmetry 2005, 16, 2449.
(9) Yus, M.; Na´jera, C.; Foubelo, F. Tetrahedron 2003, 59, 6147.
(10) (a) Moser, W. H. Tetrahedron 2001, 57, 2065. (b) Brook, A. G.
Acc. Chem. Res. 1974, 7, 77.
(11) (a) Shinokubo, H.; Miura, K.; Oshima, K.; Utimoto, K. Tetrahedron
Lett. 1993, 34, 1951. (b) Tietze, L. F.; Geissler, H.; Gewert, J. A.; Jakobi,
U. Synlett 1994, 511. (c) Fischer, M.-R.; Kirschning, A.; Michel, T.;
Schaumann, E. Angew. Chem., Int. Ed. Engl. 1994, 33, 217.
(12) Smith, A. B., III; Boldi, A. M. J. Am. Chem. Soc. 1997, 119, 6925
(13) (a) Smith, A. B., III; Pitram, S. M.; Boldi, A. M.; Gaunt, M. J.;
Sfouggatakis, C.; Moser, W. H. J. Am. Chem. Soc. 2003, 125, 14435. (b)
Smith, A. B., III; Adams, C. M. Acc. Chem. Res. 2004, 37, 365.
(14) (a) Le Merrer, Y.; Gravier-Pelletier, C.; Maton, W.; Numa, M.;
Depezay, J.-C. Synlett 1999, 1322. (b) Bra¨uer, N.; Dreessen, S.; Schaumann,
E. Tetrahedron Lett. 1999, 40, 2921.
(15) For nomenclature, see: Narayan, R. S.; Sivakumar, M.; Bouhlel,
E.; Borhan, B. Org. Lett. 2001, 3, 2489 and references therein.
(16) (a) Babler, J. H.; Bauta, W. E. Tetrahedron Lett. 1984, 25, 4323.
(b) Cooke, M. P., Jr.; Houpis, I. N. Tetrahedron Lett. 1985, 26, 3643. (c)
Rieke, R. D.; Wehmeyer, R. M.; Wu, T.-C.; Ebert, G. W. Tetrahedron 1989,
45, 443. (d) Fleming, I.; Mart´ınez de Marigorta, E. J. Chem. Soc., Perkin
Trans. 1 1999, 889.
5184
Org. Lett., Vol. 7, No. 23, 2005

Table 1. Optimization of the Linchpin Cyclization Process
9
concn
(M)
HMPA/
THF
time
(min)
mol
sieve
yield of
3âa (%)
yield of
3Ra (%)
combined yield
ratio
3â/3R
yield of
10b (%)
unreacted
1b (%)
entry
(equiv)
T (°C)
of 3 (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1.1
1.0
1.0
1.0
1.0
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.3d
1.3d
0.1
0.1
0.1
0.1
1:10
1:10
1:10
1:10
1:10
1:10
1:10
1:10c
1:30
1:10
1:10
1:10
1:9
-75
-45
-18
0
300
30
15
10
40
180
180
120
120
100
75
120
75
90
37
43
44
33
49
44
44
47
27
50
59
54
62
61
55
7
13
17
18
12
9
44
56
61
51
61
53
53
55
39
64
76
69
77
80
76
5.3:1
3.3:1
2.6:1
1.8:1
4.1:1
4.9:1
4.9:1
5.9:1
2.3:1
3.6:1
3.5:1
3.6:1
4.1:1
3.2:1
2.9:1
1
4
4
7
4
3
1
3
4
5
6
6
4
8e
7e
23
7
0
0
0.05
0.033
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.05
-45
-45
-45
-45
-45
-30
-30
-45
-30
-30
-30
13
24
17
2
0
0
0
0
0
0
9
8
12
14
17
15
15
19
19
x
x
x
x
x
1:9
1:9
90
0
a Isolated yields after separation by HPLC. b Compounds 10 and 1 were inseparable, and their ratio was estimated from the corresponding 1H NMR
spectra. The ratios and relative stereochemistry of 10R/10â could not be determined. c Et2O instead of THF. d 1.3 equiv of t-BuLi used. e Isolated yield.
Interestingly, 10 was isolated almost as a single diastereo-
isomer, but we were unable to establish the relative stereo-
chemistry of the major isomer.
The 3R/â diastereoisomers were only separable by tedious
preparative HPLC, which is clearly unsatisfactory from a
practical preparative point of view. In the event, we were
delighted to find that after tritylation (Scheme 3), the highly
epoxide moieties of the pseudo-C2-symmetric bis-epoxide 1
are expected to react at different rates, the obtained â/R ratio
of 3 appeared relatively large. However, epoxide opening
of 1 by 2 in the absence of HMPA in Et2O led to the adducts
13 and 14 in a 2:1 ratio (NMR), which unambiguously
proved that the diastereotopic epoxide groups reacted at
significantly different rates. The ratio observed here was
smaller than the ratio obtained in the linchpin cyclization
process, but this could be due to the absence of HMPA.
Reaction of nonsilylated dithianyl anion 16 with 1 in the
presence of the usual amount of HMPA (in THF) did lead
to a 2.4:1 diastereoselectivity for the epoxide opening (18/
19), despite the significantly reduced size of the nucleophile.
Scheme 3. Separation of the Diastereoisomers after Tritylation
crystalline derivative 12â could be easily separated from the
corresponding R-isomer by recrystallization from hot ethanol,
affording 51% yield from bis-epoxide 1. Single X-ray
crystallographic analysis confirmed the relative stereochem-
istry on the cyclopentane ring (Figure 2). Purification of the
remaining filtrate by HPLC afforded pure minor diastereomer
12R in 14% yield, with an additional 5% of 12â.
The recrystallization process greatly facilitated the puri-
fication procedure on a large scale.
The cyclization was further investigated by a set of control
experiments (Scheme 4). Though the two diastereotopic
Figure 2. ORTEP representation of the crystal structure of 12â.
Org. Lett., Vol. 7, No. 23, 2005
5185

Scheme 4. Investigation of the Diastereoselectivity by Preventing the Brook Rearrangement
Under the conditions as shown in Scheme 4, when the
2 proceeds in excellent yield with high 5-exo selectivity,
giving access to highly substituted carbafuranose-type struc-
tures. As both enantiomers of both 1,4-bis-epoxides are
available, and both diastereomers 12R/â could be efficiently
separated, the methodology is undoubtedly an attractive
approach toward the synthesis of 6′-substituted carbanucleo-
sides. The configuration of the OTBDMS substituent in 12/
23 is ideally set up for subsequent nucleobase introduction.
Current efforts in our laboratory are focusing on the
transformation of 12R/â, and of 23, to carbanucleoside
analogues.
cyclization process is not possible, a relatively large amount
of the diaddition products 15 and 17 was formed despite the
use of only 1 equiv of 2. In contrast, only very small amounts
of the diaddition product 11 have been isolated from the
cyclization experiments (Table 1), with no 11 detected at
all under the optimized conditions (entry 15). This confirms
that the HMPA-mediated Brook rearrangement is a very fast
process12 and that in the concentration range investigated,
the actual 5-exo (and 6-exo) cyclization reaction is much
faster than the epoxide opening by 2. In addition, the absence
of 20 and 21, which were independently synthesized from
18/19 (Scheme 4), in the optimization experiments (Table
1) further suggests that the cyclization is a facile process
and that the 3â/3R ratio relates to the epoxide opening and
not to an incomplete cyclization of 8.
Scheme 6. Further Conversion of the Dithianyl Group
Finally, the optimum conditions for the linchpin cyclization
were applied with (2S,4S)-1,2:4,5-diepoxypentane 22 as
substrate, which is also available in both enantiomeric forms
at similar cost,17 with the resulting carbocycle 23 obtained
in equally excellent yield (Scheme 5). Interestingly, no 6-exo
cyclization product was observed in this case.
Scheme 5. Linchpin Cyclization Using Bis-epoxide 22
Acknowledgment. We thank the EPSRC and CMS
Chemicals for financial support. A.J.B. is indebted to Pfizer
for a summer studentship stipend. Further unrestricted
funding from Pfizer is gratefully acknowledged. We thank
Prof. Mike Hursthouse for access to the crystallographic
facilities of the EPSRC National Crystallography Service at
Southampton. We also wish to acknowledge the use of the
EPSRC’s Chemical Database service at Daresbury.18
The preliminary results for the further conversion of the
dithianyl group are given in Scheme 6. Reduction with Raney
nickel lead to the cyclopentane derivative 24 in excellent
yield. For the hydrolysis to the ketone 25, a yield of 45%
has been achieved, and further optimization of this trans-
formation is underway.
Supporting Information Available: Experimental pro-
1
cedures and spectral data, including copies of H and 13C
NMR spectra, for 3R/â, 10, 11, 12R/â, and 23-25, and the
crystal information file (CIF) for compound 12â. This
material is available free of charge via the Internet at
http://pubs.acs.org.
In summary, it was demonstrated that a linchpin cycliza-
tion sequence involving 1,4-bis-epoxides and dithianyllithium
OL052009H
(17) (a) Rychnovsky, S. D.; Griesgraber, G.; Powers, J. P. Org. Synth.
2000, 77, 1. (b) Boydell, A. J.; Jeffery, M. J.; Bu¨rkstu¨mmer, E.; Linclau,
B. J. Org. Chem. 2003, 68, 8252.
(18) The United Kingdom Database Service. Fletcher, D. A.; McMeeking,
R. F.; Parkin, D. J. Chem. Inf. Comput. Sci. 1996, 36, 746.
5186
Org. Lett., Vol. 7, No. 23, 2005