Catalytic asymmetric synthesis of a 1-deoxy-1,1-difluoro-D-xylulose.

Article abstract of DOI:10.1021/ol0273553

[reaction: see text] A new route, of potential strategic importance, to a difluorosugar analogue has been developed. Key steps included a Stille coupling and a highly regio- and enantioselective dihydroxylation of a highly substituted diene. Protecting gr

Full text of DOI:10.1021/ol0273553

ORGANIC
LETTERS
2003
Vol. 5, No. 3
337-339
Catalytic Asymmetric Synthesis of a
1-Deoxy-1,1-difluoro-D-xylulose
,§
Liam R. Cox,† Gareth A. DeBoos,‡ Jeremy J. Fullbrook,† Jonathan M. Percy,*
Neil S. Spencer,† and Malcolm Tolley†
School of Chemical Sciences, UniVersity of Birmingham, Edgbaston,
Birmingham B15 2TT, United Kingdom, AVecia, Ltd., P.O. Box 42, Hexagon House,
Blackley, Manchester M9 8ZS, United Kingdom, and Department of Chemistry,
UniVersity of Leicester, UniVersity Road, Leicester, LE1 7RH, United Kingdom
jmp29@le.ac.uk
Received November 26, 2002
ABSTRACT
A new route, of potential strategic importance, to a difluorosugar analogue has been developed. Key steps included a Stille coupling and a
highly regio- and enantioselective dihydroxylation of a highly substituted diene. Protecting groups were chosen to enhance the reactivity of
the disubstituted allylic fragment in the AD reaction and allow deprotection under orthogonal conditions.
Fluorosugars1 are of considerable current interest because
they retain much of the shape, electron distribution, and
function of natural saccharides while lacking the ability to
enter, as hydrogen bond donors, into critical hydrogen
bonding interactions with nucleic acids or proteins. This
makes them very useful molecular tools for identifying key
interactions used by receptors in binding saccharide ligands.2
Syntheses of difluorosugars using fluorination methods to
transform a ketonic carbonyl group are common,3 though
not without difficulties caused by side reactions4 associated
with high electron demand (elimination, neighboring group
participation, 1,2-shifts).5 There are still few routes to
fluorosugars that are based upon inexpensive and readily
available fluorinated feedstocks or building blocks.6 One
important method for the de novo synthesis of monosaccha-
rides7 uses the Sharpless asymmetric dihydroxylation (AD)
reaction of dienes.8 We wished to develop the chemistry of
fluorinated building block trifluoroethanol so that catalytic
asymmetric9 syntheses of fluorosugars would become avail-
able from dihydroxylation reactions of fluorinated dienes,10
and we selected 1-deoxy-1,1-difluoro-D-xylulose as an il-
lustrative target.
There is a high level of interest in 1-deoxy-D-xylulose
currently due to its pivotal role in the biosyntheses of
pyridoxol phosphate, thiamine pyrophosphate, and the phytyl
chain of ubiquinone in E. coli (Scheme 1).11
† University of Birmingham.
‡ Avecia, Ltd.
§ University of Leicester.
(6) (a) For the large-scale route to Gemcitabine, see: Chou, T. S.; Heath,
P. C.; Patterson, L. E.; Poteet, L. M.; Lakin, R. E.; Hunt, A. H. Synthesis
1992, 565. (b) For a recent review, see: Plantier-Royon, R.; Portella, C.
Carbohydr. Res. 2000, 327, 119.
(7) (a) Guzman-Perez, A.; Corey, E. J. Tetrahedron Lett. 1997, 38, 5941.
(b) Lindstrom, U. M.; Somfai, P. Tetrahedron Lett. 1998, 39, 7173.
(8) Becker, H.; Soler, M. A.; Sharpless, K. B. Tetrahedron 1995, 51,
1345.
(9) For a series of reviews in the area, see: Ramachandran, P. V.
Asymmetric Fluoroorganic Chemistry; ACS Symposium Series 746;
American Chemical Society: Washington, DC, 2000.
(10) (a) Herrmann, W. A.; Eder, S. J. Chem. Ber. 1993, 126, 31. (b)
Vanhessche, K. P. M.; Sharpless, K. B. Chem. Eur. J. 1997, 3, 517.
(1) Dax, K.; Albert, M.; Ortner, J.; Paul, B. J. Carbohydr. Res. 2000,
327, 47.
(2) For a recent example, see: Eguchi, T.; Sasaki, S.; Huang, Z.;
Kakinuma, K. J. Org. Chem. 2002, 67, 3979.
(3) Barrena, M. I.; Matheu, M. I.; Castillon, S. J. Org. Chem. 1998, 63,
2184.
(4) For recent examples, see: (a) Aghmiz, M. L.; Diaz, Y.; Jana, G. H.;
Matheu, M. I.; Echarri, R.; Castillon, S.; Jimeno, M. L. Tetrahedron 2001,
57, 6733. (b) Dax, K.; Albert, M.; Hammond, D.; Illaszewicz, C.;
Purkarthofer, T.; Tscherner, M.; Weber, H. Monatsh. Chem. 2002, 133,
427.
(5) Dax, K.; Albert, M. Top. Curr. Chem. 2001, 215, 193.
10.1021/ol0273553 CCC: $25.00 © 2003 American Chemical Society
Published on Web 01/10/2003

Scheme 1. 1-Deoxy-d-xylulose 1 and Putative Inhibitors 2
and 3 of Downstream Enzymes
Scheme 2. Diene Synthesis and Elaboration to Xylulose
Analoguesa
Bouvet and O’Hagan12 proposed 2 and 3 as potential
inhibitors of enzymes involved in the metabolism of 1-deoxy-
D-xylulose and the 5-phosphate 4. The compounds were
inactive; however, both exist almost exclusively in cyclic
forms 2a and 3a, preventing the key phosphorylation step.
Dihydroxylation of enone 5 was attempted during their
synthesis, but the diol product could not be isolated cleanly
from the reaction.
The same fluoroketone reactivity that prevents processing
of 2 and 3 creates potential handling difficulties, so we
planned a route in which the ketone is masked until a late
stage, and significantly, one in which phosphorylation of the
5-hydroxyl group is possible.
Stannane 6 was synthesized on a 0.5 mol scale according
to our published procedure13 and coupled with protected
iodoallylic alcohol 7 (synthesized from propiolic acid in four
steps; hydroiodination,14 esterification, selective reduction,15
and esterification) under modified Farina-Liebeskind condi-
tions16 with full conversion to afford the crude diene product
8 with a clean 19F NMR spectrum (Scheme 2).17 The diene
was freed from tin residues by KF treatment and could be
chromatographed on basic alumina, an essential precaution
considering the sensitivity of similar products18 to even
mildly acidic conditions. Overall, a moderate yield of diene
(which should be used soon after synthesis) was obtained.
Dihydroxylation occurred very slowly under standard condi-
tions; typical reactions required 1 week to reach 70-85%
conversion, and a significant fall in the pH of the medium
was observed over this period. However, when the pH was
maintained19 in the range 11.0-12.0, the reaction reached
a Reaction conditions: (i) Pd(OAc)2 (2.6 mol %), CuI (20.8 mol
%), Ph3P (10.4 mol %), DMF, 30-50 °C, 41%; (ii) K2OsO4‚2H2O
(2 mol %), (DHQD)2PHAL (4 mol %), K3Fe(CN)6 (3 equiv), K2CO3
(3 equiv), 1:1 t-BuOH/H2O, pH 11.0-12.0, rt, 54%; (iii) anhydrous
CuSO4 (2 equiv), PTSA (1 mol %), acetone, rt, 68%; (iv) 30%
H2O2 (4.2 equiv), LiOH‚H2O (2.2 equiv), 3:1 THF/H2O, 0 °C, 61%;
(v) Me3SiCl (1.2 equiv), MeOH, rt, 88%; (vi) HCl (12 M), THF,
rt.
completion in only 1 h and we were able to isolate diol 9 in
55% yield. The effect of pH maintenance is spectacular and
demonstrates the critical importance of rapid osmate ester
turnover for efficient catalysis. The use of the PMBz
protecting group is also relatively unusual;20 ester hydrolysis
probably prevents recovery of the diol in higher yield, and
it is possible that a PMB ether7b or PMP ether7a would offer
similar reactivity and greater stability under the reaction
conditions. However, we wished to be able to cleave the
primary hydroxyl group under mild nucleophilic conditions,
so these possibilities were not investigated. The dihydroxy-
lation failed completely when THP-protected dienol was
exposed to the AD conditions, demonstrating clearly that the
electronic complementarity between protecting group and
catalyst is a requirement for this reaction to proceed
efficiently.
The diol was protected in acetonide 10, and then the
synthesis was completed by hydrolysis of the ester (present-
ing an opportunity for phosphorylation of alcohol 11)
followed by removal of both acetals, to afford deoxy
difluorosugar 3a as a 1:3 mixture of R- and â-anomers.
Resonances in the 1H NMR were assigned fully (by gradient
COSY/HSQC/HMBC experiments), allowing the unambigu-
ous identification of the H-3 methine proton. The configu-
ration of the major anomer (3aâ) was assigned by a NOE
experiment that showed a strong transfer of magnetization
from the H-3 methine to the CF2H proton in the major
anomer. Both anomers converged on 12 when the mixture
was treated with acetone and an acid catalyst.
(11) (a) Gupta, R. N.; Hemscheidt, T.; Sayer, B. G.; Spenser, I. D. J.
Am. Chem. Soc. 2001, 123, 11353. (b) Zhao, S.; Petrus, L.; Serianni, A. S.
Org. Lett. 2001, 3, 3819. (c) Reuter, K.; Sanderbrand, S.; Jomaa, H.;
Wiesner, J.; Steinbrecher, I.; Beck, E.; Hintz, M.; Klebe, G.; Stubbs, M. T.
J. Biol. Chem. 2002, 277, 5378.
(12) Bouvet, D.; O’Hagan, D. Tetrahedron 1999, 55, 10481.
(13) Patel, S. T.; Percy, J. M.; Wilkes, R. D. Tetrahedron 1995, 51, 9201.
(14) (a) Abarbri, M.; Parrain, J. L.; Cintrat, J. C.; Ducheˆne, A. Synthesis
1996, 82. (b) Takeuchi, R.; Tanabe, K.; Tanaka, S. J. Org. Chem. 2000,
65, 1558.
(15) Marek, I.; Meyer, C.; Normant, J.-F. Org. Synth. 1996, 74, 194.
(16) Farina, V.; Kapadia, S.; Krishnan, B.; Wang, C.; Liebeskind, L. S.
J. Org. Chem. 1994, 59, 5905.
(17) For related couplings, see: (a) DeBoos, G. A.; Fullbrook, J. J.;
Owton, W. M.; Percy, J. M.; Thomas, A. C. Synlett 2000, 963. (b) DeBoos,
G. A.; Fullbrook, J. J.; Percy, J. M. Org. Lett. 2001, 3, 2859.
(18) Broadhurst, M. J.; Brown, S. J.; Percy, J. M.; Prime, M. E. J. Chem.
Soc., Perkin Trans. 1 2000, 3217.
(19) Mehltretter, G. M.; Dobler, C.; Sundermeier, U.; Beller, M.
Tetrahedron Lett. 2000, 41, 8083.
The material exhibited the same magnitude and sign of
rotation as that reported by Bouvet and O’Hagan, confirming
(20) (a) Ref 7. (b) Report on slow AD reactions of a related alkenyl
substrate: Yokomatsu, T.; Yamagishi, T.; Suemune, K.; Yoshida, Y.;
Shibuya, S. Tetrahedron 1998, 54, 767.
338
Org. Lett., Vol. 5, No. 3, 2003

Figure 1. Enone AD substrate used by Bouvet and O’Hagan.
that the sense of enantioselection conforms with the Sharpless
model. Use of the (DHQD)2PHAL ligand predicts that 9 is
the product; we determined an ee for the AD reaction by
preparing the bis-(S)-Mosher ester (13) of 9 and the bis-(S)-
Mosher ester of racemic diol 14 and examining their 19F
NMR spectra. Whereas the CF3 region of the spectrum
contained a number of peaks from esters and unreacted
derivatizing agent that could not be deconvoluted, the CF2
region of the spectrum of 13 contained only a single doublet
for each of the vinylic fluorine atoms, even when a large
number of data points (1579 between -90.0 and -95.0 ppm)
and scans were recorded to optimize the signal-to-noise ratio
(Figure 2a, Figure 3).
Figure 3. Mosher esters used for determination of ee.
racemic diol (Figure 2b); the ee is calculated as g99.5% on
this basis.
Our method relies on Stille coupling to synthesize the key
diene; obviously, methods that avoid the use of tin reagents
and transition metal-catalyzed coupling would be preferable,
and we are examining alternative routes currently. These
findings do show that dienes such as 8 are useful substrates
for Sharpless catalytic AD reactions, despite their relatively
high lability. The dihydroxylation reaction is completely
regioselective with the highly substituted alkenyl bond
emerging untouched from the reaction. Sharpless and others
have reported successful AD reactions of polyenes, but there
are few examples where both alkenes bear deactivating
substituents. Even with the use of a protecting group known
to enhance AD reactivity, the dihydroxylation reaction was
slow and pH control was essential, suggesting that 8 lies
close to the lower limit of substrate reactivity. The AD
products possess great potential for the synthesis of a wider
range of difluorosugar analogues.
In contrast, two doublets were observed for each of the
vinylic fluorine atoms in the bis-(S)-MTPA ester 14 of the
Acknowledgment. We thank the EPSRC and Avecia,
Ltd., for a CASE Award (to J.J.F.).
Supporting Information Available: Full experimental
procedures for 7-14, characterization data for 7-12, partial
1
19F NMR data for 13 and 14, HRMS for 13, and H, 19F,
Figure 2. Partial 19F NMR spectra [282 MHz, 300K, CDCl3] of
(a) 13 with enhanced signal-to-noise (1579 data points between
-90.0 and -95.0 ppm) and (b) bis-(S)-Mosher ester of racemic
diol 14.
and 13C NMR spectra for 3aR/3aâ. This material is available
free of charge via the Internet at http://pubs.acs.org.
OL0273553
Org. Lett., Vol. 5, No. 3, 2003
339