A new synthetic approach to 2,3-dideoxy-2-fluoro-β-D-threo- pentofuranose, the fluorofuranose unit of the anti-HIV-active nucleoside, β-FddA

Article abstract of DOI:10.1016/j.tetlet.2003.11.020

The fluorinated furanose unit of the anti-HIV-active nucleoside (2,3-dideoxy-2-fluoro-β-D-threo-pentofuranosyl)adenine, β-FddA, has been synthesized as its 5-trityl derivative with high stereocontrol from (S)-trityl glycidol and phenylthioacetic acid.

Full text of DOI:10.1016/j.tetlet.2003.11.020

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 863–865
A new synthetic approach to 2,3-dideoxy-2-fluoro-b-
pentofuranose, the fluorofuranose unit of the anti-HIV-active
D
-threo-
nucleoside, b-FddA
Jean-Claude Caille,a Hugues Miel,b Paul Armstrongb and M. Anthony McKerveyb,*
aPPG-SIPSY, B.P. 79, Route de Beaucouze, 49242 Avrilleꢀ Cedex, France
bCSS (Formerly QuChem Ltd.), David Keir Building, The QueenÕs University, Belfast BT9 5AG, Northern Ireland
Received 12 August 2003; revised 30 October 2003; accepted 6 November 2003
Abstract—The fluorinated furanose unit of the anti-HIV-active nucleoside (2,3-dideoxy-2-fluoro-b-D-threo-pentofuranosyl)adenine,
b-FddA, has been synthesized as its 5-trityl derivative with high stereocontrol from (S)-trityl glycidol and phenylthioacetic acid.
ꢀ 2003 Elsevier Ltd. All rights reserved.
The introduction of a fluorine atom into the sugar
moiety of some deoxynucleosides has led to the dis-
covery of a group of compounds with significant
antiviral activity. One such compound is (2,3-dideoxy-
at C-2 via an elimination (to a vinyl fluoride)-hydro-
genation sequence to produce the b-configuration.3
Despite the success of this and other approaches to the
synthesis of 1, there is still a need for a route capable of
producing multigram quantities of the product.
2-fluoro-b-
-threo-pentofuranosyl)adenine (b-FddA) 1.1
D
The efficacy of b-FddA has attracted considerable
interest, and consequently, there have been a number of
attempts to develop efficient synthetic routes to this
compound. Two broad strategies have been investi-
gated, one involving the construction of a 2-fluoro-
furanose unit with the correct stereochemistry prior to
coupling with a suitable adenine derivative and the
second involving the fluorination of a preformed fur-
anosyl adenine adduct.1;2 A crucial requirement of either
approach is the need to establish the correct (b) con-
figuration of the fluorine substituent. In the most recent
synthesis by Choudhury et al.3 the b-fluoro substituent
was introduced via an SN2 displacement of an a-hy-
droxy group at the C-2 position. In a contribution from
the Marquez group4 at NIH, who have been particularly
active in the area, the stereochemical requirement was
addressed by fluorination of a suitable furanosylade-
nosine intermediate on the preferred alpha face of the
sugar ring with a subsequent inversion of configuration
N
OH
NH2
O
N
N
N
F
FddA (1)
In seeking to develop a new route to a fluorofuranose
suitable for eventual conversion to b-FddA we had
identified a number of specific objectives. Firstly, the
route should be amenable to scale-up with the minimum
amount of chromatographic methods of purifications.
Secondly, we sought a route that was highly enantio-
selective, and thirdly, access to inexpensive starting
materials was highly desirable. These considerations led
us to examine a glycidol derivative as the principal
building block. Several glycidol derivatives are com-
mercially available in single enantiomer form. The
condensation of glycidol derivatives with malonate
esters under basic conditions is known to form
a-substituted-c-valerolactones (Eq. 1),5 which are ame-
nable to further transformations, notably decarboxyl-
ation. Diastereoisomeric mixtures are usually obtained
Keywords: Anti-viral; Fluorofuranose; Epoxide ring opening; Phenyl-
thioacetic acid; Stereospecific fluorination.
* Corresponding author. E-mail: info@css-almac.com
0040-4039/$ - see front matter ꢀ 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.11.020

864
J.-C. Caille et al. / Tetrahedron Letters 45 (2004) 863–865
OR
O
O
O
R1CH(CO2Et)2
(eq. 1)
OR
+
CO2Et
R1
OCPh3
OCPh3
OCPh3
O
FCH(CO2Et)
2 (3)
O
O
O
O
O
O
OCPh3
NaH, THF
CO2Et
2
F
F
F
4
5
6
Scheme 1.
after decarboxylation when R1 ¼ H. Examples are
known with R1 ¼ a benzenoid substituent where sub-
sequent deprotonation to form the enolate followed by
kinetic reprotonation at low temperature has been
employed to transform a diastereoisomeric mixture into
a single isomer, the cis-lactone.6 On the strengths of
these precedents the route summarized in Scheme 1 was
designed. We chose (S)-tritylglycidol 2 and diethyl
summarized in Scheme 2. The new route commenced
with addition of the dianion of phenylthioacetic acid,
efficiently generated in THF with 2.4 equiv of LDA, to
(S)-tritylglycidol.
After acidification, but without purification, the result-
ing adduct 7 was taken up in toluene and heated under
reflux for 2 h to form the phenylsulfenyl lactone 8 in 70%
yield over the two steps. Oxidation of 8 to the corre-
sponding sulfoxides 9 (a mixture of diastereoisomers)
proceeded smoothly in quantitative yield through the
action of urea–hydrogen peroxide (UHP) in the presence
of methyltrioxorhenium (MTO) (3 mol %) in dichlo-
romethane at room temperature.
fluoromalonate
3 as the reaction partners. (S)-
Tritylglycidol is readily available in bulk with high
enantiopurity (ꢀ98% ee). Step 1 involves condensation of
the epoxide with the fluoromalonate anion to form the
fluorinated lactone 4. Step 2 requires decarboxylation to
form 5, and step 3, epimerization of the latter, if nec-
essary, to produce the cis-fluorolactone 6. A literature
survey revealed that whereas there are many examples of
the reaction of malonate ions with epoxides, the corre-
sponding process with fluoromalonate was unknown,
though this anion has been alkylated with other elec-
trophiles. In the event, treatment of (S)-tritylglycidol
with diethyl fluoromalonate in the presence of sodium
hydride in THF furnished a mixture of products, which
contained the cis- and trans-fluorolactones 5 to the
extent of 20%.
Introduction of the 2-fluoro substituent was achieved in
70% yield by treatment of the sulfoxides 9 with Select-
fluor in the presence of sodium hydride in THF at 0 ꢁC
to form fluorosulfoxide 10 (also a mixture of diastereo-
isomers). Thermolysis of 10 for 10 min in hot toluene
containing sodium bicarbonate caused a smooth elimi-
nation reaction, furnishing the unsaturated fluorolac-
tone 11 in quantitative yield. To complete the synthesis,
hydrogenation of 11 over palladium on carbon afforded
the target molecule 6 in quantitative yield with a
diastereoselectivity of >98%.7
Evidently, decarboxylation, or its equivalent, had
occurred in the course of the reaction.
In summary, an efficient synthesis of a key intermediate
for b-FddA has been developed with an overall yield of
ca. 50% via readily accessible starting materials and
easily purified intermediates. The main advantage over
the approach of Confalone and co-workers3 is the use of
Selectfluor at room temperature to introduce the fluoro
substituent in place of the less convenient DAST at low
temperature. Our approach also avoids the Horner–
Emmons chemistry, which forms the basis of the Patrick
and Wei7 approach. Another significant feature of our
synthesis, which enhances its scalability is the minimal
use of chromatography with only one intermediate, 11,
requiring purification in this way to remove a trace
amount of precursor 9.
Although we had succeeded in preparing the cis-flu-
orolactone in a one-pot process, there were significant
shortcomings in the chemistry, principally the reaction
yield, which could not be improved beyond 40% and an
inability to raise the cis-content of the isomer mixture
above 75% via low temperature deprotonation–repro-
tonation treatment.
Repeated recrystallizations of the mixture did not fur-
ther enhance the cis content. The cis and trans isomers
could readily be separated by chromatography, though
this was not considered an acceptable solution to the
problem.
In an attempt to address these limitations we devised a
modified route based on (S)-trityl glycidol, which is
Fluorination of phenylsulfinyl lactone 9 with Selectfluor.
To a three-necked round-bottomed flask fitted with a

J.-C. Caille et al. / Tetrahedron Letters 45 (2004) 863–865
865
HO
PhS
O
1. PhSCH2CO2H (1.2 equiv.)
OH
O
LDA (2.4 equiv.), THF, 0 °C, 1 h
OCPh3
OCPh3
2. HCl
3. Extraction into toluene
2
7
toluene
reflux, 2 h
OCPh3
OCPh3
O
O
MTO (3 mol%), UHP, CH2Cl2, rt, 3 h
O
O
Ph
S
SPh
O
9
8
1. NaH (1.0 equiv.), THF, 0 °C
2. Selectfluor (1.2 equiv.), DMF, 30 min, then rt, 3 h
OCPh3
OCPh3
OCPh3
O
NaHCO3
O
F
O
O
O
H2, Pd/C
rt, 15 h
O
toluene, reflux,
10 min
S
F
Ph
F
O
6
11
10
Scheme 2.
Driscoll, J. S. Biochem. Pharmacol. 1987, 36, 2719; (c)
Marquez, V. E.; Tseng, C. K.-H.; Mitsuya, H.; Aoki, S.;
Kelly, J. A.; Ford, H., Jr.; Roth, J. S.; Broder, S.; Johns,
D. G.; Driscoll, J. S. J. Med. Chem. 1990, 33, 978.
septum and a nitrogen inlet was added sodium hydride
(60% in mineral oil, 0.5 g, 12.4 mmol) followed by dry
THF (20 mL).The suspension was cooled to 0 ꢁC. A
solution of lactone 9 (4.0 g, 8.2 mmol) in dry THF
(40 mL) was then added via a syringe over 10 min. The
resulting mixture was stirred at 0 ꢁC for 1 h then at room
temperature for 1/2 h. The color had meanwhile turned
from light yellow to orange to yellow. It was then
transferred in portions via a cannula over 10 min to a
cooled (ice bath) solution of Selectfluor (2.84 g,
8.7 mmol) in dry DMF. When the addition was com-
pleted, the solution was stirred at room temperature for
3 h. It was then poured into a 500 mL separating funnel
containing aqueous saturated ammonium chloride
(200 mL) and extracted with ether (2 · 150 mL). The
combined ethereal extracts were washed with water
(6 · 150 mL), dried over magnesium sulfate, filtered, and
concentrated at reduced pressure to yield the fluorinated
lactone 10, a white foam (3.8 g), which was used without
purification. The 300 MHz NMR spectrum measured
immediately after completion of the reaction workup
revealed the presence of a small amount of starting
material, which was not removed prior to the next step.
The NMR spectrum measured 24 h later revealed the
presence of a new component, indicating that elimina-
tion to form unsaturated lactone 10 was already taking
place to a small extent.
2. For accounts of earlier synthetic approaches to b-FddA,
see Ref. 1 and: (a) Herdewijn, P.; Pauwels, R.; Baba, M.;
Balzarini, J.; De Clercq, E. J. Med. Chem. 1987, 30, 2131;
(b) Marquez, V. E.; Tseng, C. K.-H.; Misty, H.; Okay, S.;
Kelly, J. A.; Ford, H., Jr.; Roth, J. S.; Johns, D. G.;
Driscoll, J. S. J. Med. Chem. 1990, 33, 978; (c) Shiragami,
H.; Tanaka, Y.; Uchida, Y.; Iwagami, H.; Isawa, K.;
Yukawa, T. Nucleos. Nucleot. 1992, 11, 391; (d) Siddiqui,
M. A.; Marquez, V. E.; Driscoll, J. S.; Barchi, J. J.
Tetrahedron Lett. 1994, 35, 3263; (e) Takamatsu, S.;
Maruyama, T.; Katayama, S.; Hirose, N.; Naito, M.;
Izawa, K. Tetrahedron Lett. 2001, 42, 2325; (f) Okabe, M.;
Sun, R.-C.; Zenchoff, G. B. J. Org. Chem. 1991, 56, 4392;
(g) Wysocki, R. J., Jr.; Siddiqui, M. A.; Barchi, J. J., Jr.;
Driscoll, J. S.; Marquez, V. E. Synthesis 1991, 1005; (h)
Jin, F.; Wang, D.; Confalone, P. N.; Pierce, M. E.; Wang,
Z.; Xu, G.; Choudhury, A.; Nguyen, D. D. Tetrahedron
Lett. 2001, 42, 4787.
3. Choudhury, A.; Jin, F.; Wang, Z.; Xu, G.; Nguyen, D.;
Castro, J.; Pierce, M. E.; Confalone, P. N. Tetrahedron
Lett. 2003, 44, 247.
4. Siddiqui, M. A.; Driscoll, J. S.; Marquez, V. E. Tetrahe-
dron Lett. 1998, 39, 1657.
5. (a) Van Zyl, G.; Zuidema, G. D.; Zack, J. F.; Kroman, P.
B. J. Am. Chem. Soc. 1953, 75, 5002; (b) Chatterjee, A.;
Banerjee, D.; Banerjee, B.; Mallik, R. Tetrahedron 1983,
18, 2965; (c) Chatterjee, A.; Banerjee, D.; Mallik, R.
Tetrahedron 1983, 18, 2965.
References and notes
6. Takano, S.; Tanaka, M.; Seo, K.; Hirama, M.; Ogosa-
wara, K. J. Org. Chem. 1985, 50, 931.
7. The hydrogenation of a similar fluorobutenolide, which
1. For the background to the discovery of this active
compound see: (a) Graul, A.; Silvestre, J.; Castaner, J.
Drugs Future 1998, 23, 1176; (b) Marquez, V. E.; Tseng, C.
K.-H.; Kelly, J. A.; Mitsuya, H.; Broder, S.; Roth, J. S.;
had been prepared from D-mannose, has been des-
cribed by Patrick, T. B.; Wei, Y. J. Fluorine Chem. 1998,
90, 53.