Stereoselective synthesis of azanucleosides aza-stavudine (aza-D4T), aza-2′,3′-didehydro-3′-deoxyuridine (aza-D4U), and its hydrogenated analogues from an endocyclic enecarbamate

Article abstract of DOI:10.1016/S0040-4039(00)02343-1

The stereoselective synthesis of the aza-analogues of nucleosides stavudine (D4T) and dideoxyuridine (DDU) were accomplished in a very concise manner (3-4 step sequences) with high overall yields from a five-membered endocyclic enecarbamate employing phenylselenenyl bromide as an effective promoter.

Full text of DOI:10.1016/S0040-4039(00)02343-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 1599–1602
Stereoselective synthesis of azanucleosides aza-stavudine
(aza-D4T), aza-2%,3%-didehydro-3%-deoxyuridine (aza-D4U), and
its hydrogenated analogues from an endocyclic enecarbamate
Edson R. Costenaro,a Luis A. M. Fontoura,b Denilson F. Oliveiraa and Carlos R. D. Correiaa,*
aInstituto de Qu´ımica, Universidade Estadual de Campinas, UNICAMP, C.P. 6154, 13083-970, Campinas, Sa˜o Paulo, Brazil
bFundac¸a˜o de Cieˆncia e Tecnologia, Rua Washington Luiz 675, 90010-000, Porto Alegre, Rio Grande do Sul, Brazil
Received 21 December 2000; accepted 22 December 2000
Abstract—The stereoselective synthesis of the aza-analogues of nucleosides stavudine (D4T) and dideoxyuridine (DDU) were
accomplished in a very concise manner (3–4 step sequences) with high overall yields from a five-membered endocyclic
enecarbamate employing phenylselenenyl bromide as an effective promoter. © 2001 Elsevier Science Ltd. All rights reserved.
Azanucleosides show great promise as metabolites
capable of interfering with DNA processing enzymes in
a controlled manner.1 Among the key processes in
nucleic acid metabolism, the base-excision DNA repair
(BER) enzymes play a significant role by repairing
damage inflicted to DNA from spontaneous hydrolysis,
oxidation and errors in replication. The structural simi-
larity of aza-ribonucleosides to ribonucleosides makes
them suitable transition state analogues for fault incor-
poration into DNA (2-deoxy series) or RNA. A
remarkable example of the influence of an azanu-
cleoside incorporated into DNA was recently described
by Verdine.2 The DNA incorporated adenine pyrro-
lidine homonucleoside (phA) 1 (Fig. 1) was responsible
for a strong inhibition of DNA glycolylase MutY.
Moreover, azanucleosides such as 2, containing a CꢀN
glycosidic bond have been shown to stabilize antisense
oligonucleotides toward 3%-exonucleases.3
Our interest in azanucleosides stemmed from the possi-
bility of creating novel nucleoside analogues from chiral
five-membered endocyclic enecarbamates which would
be capable of interacting with reverse transcriptase in
an inhibitory mode, thus providing new therapeutic
leads against HIV viruses. Identification of new agents
against drug resistant strains of HIV-1 is a subject of
great medical relevance nowadays.4 Our initial objec-
tives were the synthesis of aza-analogues of stavudine 3
and dideoxyuridine 4 (DDU) (Fig. 1).5 Stavudine is an
anti-retroviral nucleoside drug currently employed as
part of a drug cocktail that has been effective in the
treatment of the acquired immunodeficiency syndrome
(AIDS).
In this communication we present the successful synthe-
sis of these compounds and of their saturated analogues
in a straightforward manner with good control of the
NH2
O
O
N
O
N
N
N
HN
Me
Me
HN
HN
N
N
O
H
N
DNA
N
Ac
O
O
O
O
O
HO
N
HO
HO
O
1
OH
2
4 (DDU)
3 (stavudine)
DNA
(D4T)
Figure 1.
* Corresponding author.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)02343-1

1600
E. R. Costenaro et al. / Tetrahedron Letters 42 (2001) 1599–1602
I
HO
N
O
O
a
HN
c
HN
N
N
N
N
OTBDPS
+
OTBDPS
Boc
Boc
OTBDPS
O
O
Boc
HO
CH3
CH3
5a
O
6
9
HN
N
N
b
OTBDPS
Boc
O
c
CH3
10
O
O
N
N
O
O
+
OTBDPS
N
N
N
N
H3C
H3C
OTBDPS
Boc
Boc
8
7
Scheme 1. Reagents and conditions: (a) (TMS)2-thymine, NIS, CH2Cl2, 0°C; (b) DBU, CH2Cl2, 0°C; (c) NaOH/H2O/THF, rt, 30
min.
PhSe
O
O
OH
OR
a or b
OTr
c
HN
N
HN
N
N
N
N
Boc
Boc
Boc
O
O
CH3
5b
CH3
13
11, R = Tr
12, R = H
d
O
OH
HN
N
N
Boc
O
CH3
14
Scheme 2. Reagents and conditions: (a) (TMS)2-thymine, PhSeBr, CH3CN, −23°C (b) i. (TMS)2-thymine, PhSeBr, CH3CN, −23°C;
ii. ZnBr2, MeOH, CH2Cl2 (85%); (c) H2O2, dioxane, NaHCO3 (96%) (d) H2, Pd/C, EtOAc (92%).
stereoselectivity of the process and in very good overall
yields.
of b and a anomers. This 3:1 stereoselectivity was
confirmed by the conversion of azatricyclic nucleoside 7
into the 3%-deoxyazanucleoside 9 in 73% yield. Attempts
to perform dehydrohalogenation of iodonucleoside 6
with sodium methoxide resulted only in a mixture of
tricyclic products 7 and 8, together with small quanti-
ties of the nucleosides 9 and 10; the desired unsaturated
azanucleoside was not observed in this case. Prompted
by these results and by the low stereoselectivity
observed for the addition of (TMS)2-thymine to the
five-membered endocyclic enecarbamate, we looked for
other promoters for the nucleophilic addition of pyrim-
idine and/or purine bases to the endocyclic enecarba-
mates.
Addition of pyrimidine bases promoted by N-iodosuccin-
imide (NIS)6a,b
Addition of bis(trimethylsilyl)thymine to endocyclic
enecarbamate 5a promoted by NIS occurred quickly
and cleanly to provide the expected iodonucleoside 6 as
a single spot on TLC (Scheme 1).7 Although compound
6 is stable it was not isolated. Instead, the crude
iodonucleoside 6 was directly reacted with DBU to give
a
3:1 diastereomeric mixture of the azatricyclic
nucleosides 7 and 8 in 75% overall yield demonstrating
moderated stereoselectivity for the electrophilic addi-
tion process. The diastereomers were separated by
column chromatography and analysed spectroscopically
(1H, 13C NMR and COSY spectroscopy) confirming the
b anomer as the major reaction product.
Use of phenylselenenyl bromide as promoter6c–e
Addition of silylated thymine to the five-membered
endocyclic enecarbamate 5b7 promoted by phenylsele-
nenyl bromide at −23°C (Scheme 2) occurred smoothly
to give the seleno azanucleoside 11 as a major product.
Interestingly, contrary to glycals, no Lewis acid catalyst
was necessary to perform the addition of the silylated
Reaction of iodonucleoside 6 with aqueous sodium
hydroxide provided the 3%-deoxy azanucleosides 9 and
10 in 87% yield from enecarbamate 5a as a 3:1 mixture

E. R. Costenaro et al. / Tetrahedron Letters 42 (2001) 1599–1602
1601
O
O
OH
OH
HN
N
a, b
HN
N
OTr
c
N
N
N
Boc
Boc
O
O
Boc
16; aza-DDU
5b
15; aza-D4U
Scheme 3. Reagents and conditions: (a) i. (TMS)2-uracil, PhSeBr, CH3CN, −23°C; ii. ZnBr2, MeOH, CH2Cl2; (b) H2O2, dioxane,
NaHCO3; (c) H2, Pd/C, EtOAc.
base to the enecarbamate.6c,d A minor secondary
product was identified as the free alcohol 12 resulting
from deprotection of the trityl group. Since detrityla-
tion was a planned step in the synthesis we decided to
remove the Tr group immediately after thymidine addi-
tion, without purification. Therefore, the products
obtained from addition of (TMS)2-thymine promoted
Scheme 3). Oxidation of the seleno azanucleoside fol-
lowed by an easy syn-selenoxy elimination provided the
aza-D4U 15 in 87% yield (ꢀ10:1 diastereomeric ratio
by 1H NMR at 60°C). Catalytic hydrogenation of
aza-D4U then provided aza-DDU 16 in 86% yield.
In conclusion, the azanucleosides aza-D4T, aza-DDT,
aza-D4U and aza-DDU were prepared in a concise
manner and in high overall yields starting from the
five-membered endocyclic enecarbamate 5b. The
method employing phenylselenenyl bromide was more
stereoselective than that employing NIS as promoter.
Moreover, it is pertinent to note that in contrast to
similar additions of silylated bases to glycals, no Lewis
acid catalysis was necessary to promote addition of the
silylated bases to the endocyclic enecarbamate.6c,d The
methodology described herein to prepare cis-azanu-
cleosides complements those available to the synthesis
of trans-azanucleosides, which make use of N-acyli-
minium intermediates.10 The methodology is also
straightforward and amenable to the synthesis of other
azanucleosides of pharmacological and therapeutic
interest.
8
by PhSeBr were treated with ZnBr2 to afford a single
product by TLC corresponding to the thymidine
derivative 12 in 85% isolated yield. Elimination of the
phenylselenenyl group to generate the desired olefin
was carried out under standard conditions with hydro-
gen peroxide to give aza-stavudine 13 (aza-2%,3%-didehy-
dro-3%-deoxythymidine; aza-D4T) in 96% yield. The
entire process to aza-stavudine from endocyclic enecar-
bamate comprised only three steps involving two sepa-
rate operations with an overall yield of 81% yield. The
simplicity and high yields associated with these trans-
formations make the process suitable for a considerable
increase of scale (we easily prepared batches of 0.5–1.0
g of aza-stavudine 13). Starting from aza-stavudine
preparation of dideoxythymidine 14 (aza-DDT) was
straightforward. Catalytic hydrogenation of aza-
stavudine 13 cleanly provided aza-DDT 14 in 92%
yield. The stereoselectivity of thymine addition employ-
ing phenylselenenyl bromide was much higher than that
observed for thymine addition promoted by NIS. Aza-
stavudine 13† was obtained as a 9:1 ratio mixture of the
b/a anomers as estimated by 1H NMR analysis at
60°C.9
Acknowledgements
This work was supported by a grant from the Research
Supporting Foundation of the State of Sa˜o Paulo
(FAPESP 99/06566-7). We also thank CNPq and
CAPES for fellowships.
The synthesis of the aza-uridine analogues (aza-DDU)
was carried out using the same sequence of reactions
depicted above for aza-stavudine and aza-DDT
(Scheme 3). Thus, addition of bis(trimethylsilyl)uracil
to endocyclic enecarbamate 5b, followed by deprotec-
tion of the uracil adduct with zinc bromide provided
the seleno azanucleoside in 73% yield (not shown in
References
1. For a review, see: Yokoyama, M.; Momotake, A. Synthe-
sis 1999, 1541–1554.
2. Deng, L.; Scharer, O. D.; Verdine, G. L. J. Am. Chem.
Soc. 1997, 119, 7865–7866.
† Selected spectroscopic data: 13 (major isomer). 1H NMR (300 MHz,
CDCl3, 60°C): l 8.97 (br s, 1H), 7.55 (br s, 1H), 6.89 (br s, 1H),
6.10 (dt, 1H, J=6.2, 1.8 Hz), 5.73 (dt, 1H, J=6.2, 1.8 Hz), 4.63 (br
s, 1H), 4.12 (dd, 1H, J=11.4, 2.9 Hz), 3.86 (dd, 1H, J=11.4, 4.2
Hz), 1.87 (s, 3H), 1.42 (s, 9H). 13C NMR (75 MHz, CDCl3) l 163.9,
154.7, 150.5, 136.0, 132.5, 126.7, 82.4, 73.7, 66.9, 63.9, 28.2, 12.5. 15
(major isomer): 1H NMR (300 MHz, CDCl3, 55°C) l 8.92 (br s,
1H), 7.80 (d, 1H, J=8.1 Hz), 6.91 (br s, 1H), 6.10 (dt, 1H, J=6.25,
1.6 Hz), 5.76 (dt, 1H, J=6.25, 2.1 Hz), 5.67 (d, 1H, J=8.1 Hz),
4.64 (br s, 1H), 4.15 (dd, 1H, J=11.4, 1.8 Hz), 3.84 (dd, 1H,
J=11.4, 4.0 Hz), 3.02 (br s, 0.6H), 1.76 (br s, 0.4H), 1.44 (s, 9H).
13C NMR (75 MHz, CDCl3): l 163.6, 154.3, 150.5, 140.6, 132.8,
126.4, 102.3, 82.4, 73.8, 66.7, 63.3, 28.1.
3. Altmann, K. H.; Freier, S. M.; Pieles, U.; Winkler, T.
Angew. Chem., Int. Ed. Engl. 1994, 33, 1654–1657.
4. Rando, R. F.; Nguyen-Ba, N. Drug Discovery Today
2000, 5, 465–476.
5. During the course of these studies the synthesis of L-aza-
stavudine was reported by Ramasamy (Varaprasad, C.;
Averett, D.; Ramasamy, K. S. Tetrahedron 1999, 55,
13345–13368).
6. For similar strategies, see: (a) McDonald, F. E.; Gleason,
M. M. J. Am. Chem. Soc. 1996, 118, 6648–6659; (b) Kim,
C. U.; Misco, P. F. Tetrahedron Lett. 1992, 33, 5733–
5736; (c) Diaz, Y.; Laghdach-E., A.; Matheu, M. I.;

1602
E. R. Costenaro et al. / Tetrahedron Letters 42 (2001) 1599–1602
Castillo´n, S. J. Org. Chem. 1997, 62, 1501–1505; (d)
3029–3042; (b) Kohli, V.; Blo¨ecker, H.; Ko¨ester, H. Tet-
rahedron Lett. 1980, 21, 2683–2686.
Kassou, M.; Castillo´n, S. J. Org. Chem. 1997, 62, 3696–
3701; (e) Beach, J. W.; Kim, H. O.; Jeong, L. S.; Nam-
palli, S.; Islam, Q.; Ahn, S. K.; Babu, J. R.; Chu, C. K.
J. Org. Chem. 1992, 57, 3887–3894.
9. The presence of rotamers greatly complicated NMR
analysis, therefore high temperature 1H NMR of com-
pound 13 was necessary to observe the two double
triplets at 6.10 ppm corresponding to H-3%. The b
anomer of compound 14 also displays a small but sig-
nificant NOE between H-6 (vinylic hydrogen on the
thymine moiety) and H-5% (hydroxymethylene on the
pyrrolidine moiety).
7. For a general procedure that includes preparation of 5a,
see: Oliveira, D. F.; Miranda, P. C. M. L.; Correia, C. R.
D. J. Org. Chem. 1999, 64, 6646–6652. Enecarbamate 5b
was prepared in three steps in 85% overall yield from the
commercially available (S)-(+)-5-(trityloxymethyl)-2-
pyrrolidinone.
10. Rassu, G.; Pinna, L.; Spanu, P.; Ulgheri, F.; Casiraghi,
G. Tetrahedron Lett. 1994, 35, 4019–4022.
8. (a) Severino, E. A.; Correia, C. R. D. Org. Lett. 2000, 2,
.
.