Synthesis of aucubovir II, a new carbocyclic nucleoside analog

Article abstract of DOI:10.1002/1099-0690(200104)2001:7<1331::AID-EJOC1331>3.0.CO;2-M

The synthesis of a new carbocyclic nucleoside starting from aucubin, a natural methylcyclopentanoid monoterpene, has been performed, allowing the preparation of aucubovir II, a carbocyclic nucleoside analog with a highly functionalized cyclopentane ring. The stereocontrol of the coupling reaction was complete, utilizing the procedure described by Vorbrueggen with a purinic base.

Full text of DOI:10.1002/1099-0690(200104)2001:7<1331::AID-EJOC1331>3.0.CO;2-M

FULL PAPER
Synthesis of Aucubovir II, a New Carbocyclic Nucleoside Analog[‡]
Armandodoriano Bianco,*[a] Lucia Celletti,[a] Raffaele Antonio Mazzei,[a] and
Floriana Umani[a]
Keywords: Carbocycles / Nucleosides / Antiviral agents / Natural products / Asymmetric synthesis
The synthesis of a new carbocyclic nucleoside starting from
aucubin, a natural methylcyclopentanoid monoterpene, has
been performed, allowing the preparation of aucubovir II, a
carbocyclic nucleoside analog with a highly functionalized
cyclopentane ring. The stereocontrol of the coupling reaction
was complete, utilizing the procedure described by
Vorbrüggen with a purinic base.
Introduction
with a cyclopentane or cyclopentene moiety. The latter is
generally not functionalized.
There has been an increasing interest by synthetic organic
chemists in carbocyclic nucleoside analogs since the syn-
Results and Discussion
We decided to prepare a highly functionalized carbovir
analog to verify if, in the presence of several functions on
the cyclopentane ring, the nucleoside analog could be ac-
cepted as a substrate by the cellular kinase, and so could
inhibit the transcription process.
We have chosen iridoids, methyl-cyclopentanoid glucos-
ides belonging to the family of monoterpenoids, as the
starting compound for the synthesis of the enantiomerically
pure cyclopentanoid moiety. The choice of this class of
compounds arises because these products are characterized
by a cyclopentanoid residue. An easy elaboration of iridoids
allows generation of a wide collection of cyclopentanoid
derivatives, variously and generally highly substituted, and
enantiomerically pure. Starting from aucubin 1, a methyl-
cyclopentanoid glucoside present in large quantities in
plants of the Aucuba genus, we prepared a cyclopentanoid
synthon. After the insertion of uracil we obtained a new
carbocyclic nucleoside analog, functionalized at the C-5
carbon of the cyclopentane ring: we named it aucu-
bovir[16] (Figure 2).
Figure 1. Nucleoside analogs
thesis of carbovir (Figure 1) which a showed similar po-
tency as AZT in inhibiting viral reverse transcriptase of the
Human Immunodeficiency Virus (HIV) by acting as a
DNA chain terminator.[1Ϫ2] Furthermore carbovir appears
to be less toxic and has a longer half life than other nucleo-
side analogs such as AZT[3] (Figure 1). The higher metabolic
stability of carbovir, as well as of other carbocyclic nucleo-
side analogs, is due to the stability of the bond between the
base and the cyclopentanoid moiety. In fact, while the hy-
drolysis of the N-acetylic glycosidic bond in furanose nucle-
osides is a relatively facile process,[4] the cleavage of the am-
ino bond in carbocyclic nucleosides under the same condi-
tions is more difficult. The observed antiviral efficacy can
be explained by considering that the modification on the
five-atom ring is well tolerated by the cellular kinase, as the
methylene group is a bio-isostere with oxygen.
For all these reasons numerous syntheses of carbovir and
other carbocyclic nucleosides have been described.[5Ϫ15] The
most common approach to carbocyclic nucleosides is a con-
vergent synthesis which couples a purine or pyrimidine base
Figure 2. Synthesis of Aucubovir
[‡]
A preliminary communication appeared at 12th
International Conference on Organic Synthesis Ϫ ICOS 12.
The coupling between the aucubin synthon and uracil
was performed using Vorbrüggen’s procedure,[17Ϫ18] which
allowed us to isolate the expected α-diastereoisomer in a
3:1 ratio with respect to the β-diastereoisomer.[16] The low
stereoselectivity of the coupling reaction was attributed to
Venezia, June 28ϪJuly 2, 1998. Book of Abstracts p 399.
[a]
`
Dipartimento di Chimica, Universita ‘‘La Sapienza’’, Centro di
Studio CNR per la Chimica delle Sostanze Organiche Naturali.
Piazzale Aldo Moro n. 5, 00185, Roma, Italy
Fax: (internat.) ϩ39-06/490-631
E-mail: Armandodoriano.Bianco@uniroma1.it
Eur. J. Org. Chem. 2001, 1331Ϫ1334
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/0407Ϫ1331 $ 17.50ϩ.50/0
1331

A. Bianco, L. Celletti, R. A. Mazzei, F. Umani
FULL PAPER
Scheme 1. R ϭ isobutyroyl; Bz ϭ benzoyl; (a) (i) Hg(OAc)2, NaBH4, or (ii) β-glucosidase, NaBH4, 90%; (b) isobutyric anhydride, pyridine,
0 °CǞ25 °C, 1.5 h, 90%; (c) Ac2O, pyridine, 1 h, 99%; (d) (i) N6-benzoyl-adenine, CH3CN, TMSCl, HMDS, (ii) CF3(CF2)3Si(CH3)3,
CH2Cl2, reflux, 8 h, 20%; (e) DIBAL, CH2Cl2, Ϫ78 °C, 95%
the presence of a cyclopentanoid moiety instead of the fur- acetylated with acetic anhydride and pyridine to give 4 in
anose one.
quantitative yield. The coupling of 4 was performed with
We therefore tried to use a different base, a purinic one, the previously silylated adenine derivative, according to the
to verify the stereoselectivity of the coupling reaction and procedure proposed by Vorbrüggen.[17Ϫ18]
to prepare a new carbocyclic nucleoside analog. The syn-
thetic strategy is described in Scheme 1 and is similar, in the was achieved by H NMR spectroscopy. A positive NOE
Assignment of the configuration at the C-1 center of 5
1
first steps, to that described for the preparation of aucu- effect is present between H-5 and H-1, demonstrating the
bovir.[16]
syn relationship between the two substituents. Because the
With adenine, although we did not get very satisfactory configuration of the C-5 center is not affected by the above
yields, as is typical of Vorbrüggen’s procedure (coupling re- reactions, its absolute configuration is the same as in the
action, about 20%), we observed complete stereoselectivity starting aucubin 1. In this way the absolute configuration
with formation, as the sole product, of the α-diastereo- of the C-1 center is also defined. The successive hydrolysis
isomer, in which the hydroxymethyl chain and adenine are of the protective groups in 5 was performed by a classical
syn with respect to the cyclopentane plane.
hydrolysis with DIBAL, obtaining the target carbocyclic
The new nucleoside 6 is characterized, as in the case of nucleoside 6.
carbovir, by a double bond in the cyclopentane moiety
which, besides the required hydroxymethyl function, shows
Experimental Section
additional hydroxylated functions. Compound 6, which we
named aucubovir II, is (1β,4β,5β)-1[5-hydroxyethyl-3,4-
di(hydroxymethyl)cyclopent-2-en-1-yl]adenine (carbocyclic
2Ј,3Ј-didehydro-5-Јhydroxyethyl-3Ј,4Ј-di(hydroxymethyl)-
adenine).
Two typical synthetic procedure were employed, starting
from aucubin 1, for the preparation of the first synthon 2.
The first one consists of the opening of the dihydropyran
ring of 1 and was performed by enzymatic hydrolysis fol-
1
General: H NMR spectra were measured with a Bruker 500 MHz
spectrometer and chemical shifts are expressed in ppm relative to
TMS. Optical rotation was registered with a Jasco DIP-370 polari-
meter. Product purification was obtained by solid-liquid column
chromatography on Merck 0.063Ϫ0.20 nm silica gel; the elution
mixtures were determined case by case. Merck TLC plates coated
with Kiesel-Gel 60 F254 were employed to monitor the reactions
using 2  H2SO4, heating at 120 °C.
lowed by reduction of the hemiacetalic structure; the second 5-Hydroxyethyl-3,4-di(hydroxymethyl)cyclopenta-2-en-1-ol (2): En-
zymatic hydrolysis of aucubin 1 (100 mg, ca. 0.29 mmol) was per-
formed in a citrate buffer at pH 5.5 (3 mL) at 30 °C for 8 h. The
aglycone was then extracted with EtOAc (10 ϫ 10 mL) and, after
removal in vacuo of volatile material, dissolved in water (5 mL)
and treated with excess NaBH4 for 10 min. at 25 °C. Excess NaBH4
was destroyed by bubbling CO2 into the reaction mixture until it
reached pH ϭ 7 and decolorizing charcoal (500 mg) was added to
adsorb the organic material. The suspension was then stratified in
a Gooch funnel and the charcoal washed with water until elimina-
tion of the salts was complete. Successive elutions with methanol
one involves mercuriation/demercuriation. From the syn-
thetic point of view, the enzymatic route is preferable, af-
fording a yield of 92%, instead of 87% for the mercuriation/
demercuriation process.
The obtained intermediate 2 contains three primary alco-
holic functions, which were regioselectively protected with
an isobutyroyl residue to give the triester 3. The insertion
of a purinic base into the intermediate 3 has to be per-
formed with Vorbrüggen’s procedure on the allylic acetate
4; therefore the secondary alcoholic function of 3 was allowed the recovery of pure 2 (50 mg, 92% yield).
1332 Eur. J. Org. Chem. 2001, 1331Ϫ1334

Synthesis of Aucubovir II, a New Carbocyclic Nucleoside Analog
Alternatively, mercuriation/demercuriation was performed by dis-
FULL PAPER
lution was refluxed for 7 h. Volatile materials were evaporated un-
solving 1 in water (3 mL) together with Hg(OAc)2 (138 mg, ca. der reduced pressure (50 °C at 0.1 Torr) and to the residue was
0.44 mmol). After 10 min. an excess of NaBH4 (330 mg, ca.
added 4 (200 mg) dissolved in 1,2-dichloroethane (1.5 mL). After
8.7 mmol) was added over the course of 30 min. at 25 °C. The 5 min., (CH3)3SiOSO2C4F9 (trimethylsilylnonaflate) (12µL) was
successive workup was similar to that previously described. In this
case recovered crude 2 required a second chromatography on silica
gel with CHCl3/MeOH (8:2) to afford pure 2 (46 mg, ca. 87%
added whilst stirring and the solution was refluxed for 8 h until
complete disappearance of 4. The solution was neutralized with
NaHCO3 (sat. sol.), diluted with CH2Cl2 (100 mL) and the re-
sulting residue, after evaporation of volatile materials, was chroma-
yield).
1
2: H NMR (D2O): δ ϭ 1.90 (m, 2 H, 2H-5Ј), 2.13 (m, 1 H, H-5), tographed on silica gel with hexane/EtOAc (6:4 Ǟ1:1), furnishing
2.90 (m, 1 H, H-4), 3.71 (m, 2 H, 2H-4Ј), 3.75 (m, 2 H, 2H-5ЈЈ), the expected intermediates (39 mg, 20% yield) which consisted of
4.24 (m, 2 H, 2H-3Ј), 4.60 (m, 1 H, H-1), 5.80 (m, 1 H, H-2). Ϫ the α-diastereoisomer.
1
13C NMR (D2O): δ ϭ 147.3 (C-3), 130.8 (C-2), 81.8 (C-1), 62.0 (C-
5: H NMR (CDCl3): δ ϭ 1.11, 1.12, 1.14 [d, J ϭ 7.5 Hz, 3 ϫ 6
3Ј), 60.4 (C-4Ј), 60.2 (C-5ЈЈ), 48.7 (C-4), 48.5 (C-5), 31.1 (C-5Ј). See H, 3 ϫ (CH3)2CH], 1.84 (m, 2 H, 2H-5Ј), 2.40 (m, 1 H, H-5), 2.49
also ref.[19]
[septulet, J ϭ 7.5 Hz, 3 ϫ 1 H, 3 ϫ (CH3)2CH], 2.89 (m, 1 H, H-
4), 3.98, 4.33 (m, 2 H, 2H-4Ј), 4.22 (m, 2 H, 2H-5ЈЈ), 4.42 (m, 2 H,
2H-3Ј) 4.44 (m, 1 H, H-1), 5.99 (m, 1 H, H-2), 8.66 (s, 1 H, H-2
adenine), 8.14 (s, 1 H, H-8 adenine), 8.06 (m, 2 H, aromatics 2,6),
7.72Ϫ7.31 (m, 3 H, aromatics 3, 4, 5). Ϫ 13C NMR (CDCl3): δ ϭ
19.0 [(CH3)2CHCO], 24.7 [(CH3)2CHCO], 34.2 (C-5Ј), 42.2 (C-5),
46.0 (C-4), 61.6, 61.8 (C-4Ј, C-5ЈЈ), 63.6 (C-3Ј), 68.4 (C-1), 131.9
(C-2), 128.7 (CH-aromatics, C-5 adenine), 129.4, 132.9, 134.1 (C-
H-aromatics), 143.4 (C-8 adenine), 151.6 (C-4 adenine), 151.8 (C-
2 adenine), 153.6 (C-6 adenine). Ϫ [α]2D5 ϭ Ϫ77 (MeOH, c ϭ 0.1).
Ϫ C33H41N5O7 (619.71): C 63.96, H 6.67, N 11.30; found C 63.80,
H 6.70, N 11.24.
5-Hydroxyethyl-3,4-di(hydroxymethyl)cyclopenta-2-en-1-ol 3Ј,4Ј,5ЈЈ-
Triisobutyroyl Ester (3): The reaction was accomplished by dissol-
ving 2 (50 mg) in pyridine (0.7 mL) followed by the addition of
isobutyric anhydride (0.2 mL) at 0 °C. After 2.5 h the starting prod-
uct was no longer present in the reaction mixture, which was then
diluted with EtOAc (100 mL). After washing with 2  HCl and
successively with brine solution until neutrality, the residue ob-
tained after evaporation in vacuo of volatile material, was chroma-
tographed on silica gel with hexane/EtOAc (9:1), affording pure 3
(96 mg, ca. 91% yield), the rest being a tetrabutyroyl derivative.
1
3: H NMR (CDCl3): δ ϭ 1.10, 1.11, 1.13 [d, J ϭ 7.5 Hz, 3 ϫ 6
H, 3 ϫ (CH3)2CH], 1.89 (m, 2 H, 2H-5Ј), 2.13 (m, 1 H, H-5), 2.45,
2.50, 2.51 [septuplet, J ϭ 7.5 Hz, 3 ϫ 1 H, 3 ϫ (CH3)2CH], 2.90
(m, 1 H, X part of an ABX system, H-4), 3.92, 4.34 (AB part of
an ABX system, JAB ϭ 11.0, JAX ϭ 2.5, JBX ϭ 4.0 Hz, 2 H, 2H-
4Ј), 4.20 (m, 2 H, 2H-5ЈЈ), 4.56 (m, 1 H, H-1), 4.62 (m, 2 H, 2H-
3Ј), 5.76 (m, 1 H, H-2). Ϫ 13C NMR (CDCl3): δ ϭ 19.2
[(CH3)2CHCO], 27.2 [(CH3)2CHCO], 34.8 (C-5Ј), 44.9 (C-5), 45.6
(C-4), 61.6 (C-3Ј), 62.0 (C-5ЈЈ), 63.2 (C-4Ј), 82.1 (C-1), 129.2 (C-2),
143.7 (C-3), 177.1, 177.2, 177.4 [3 ϫ (CH3)2CHCO]. Ϫ [α]2D5 ϭ Ϫ87
(MeOH, c ϭ 0.1). Ϫ C21H34O7 (398.49): C 63.30, H 8.60; found C
63.18, H 8.68.
Aucubovir II (6): Compound 5 (30 mg) was dissolved in CH2Cl2
(2 mL) and treated with DIBAL in hexane (1.3 mL of a 1  solu-
tion) at Ϫ78 °C for 2 h. After bubbling CO2 through the mixture
for a few minutes volatile materials were evaporated in vacuo and
the residue was chromatographed on silica gel with CHCl3/MeOH
(6:4) to give pure aucubovir II 6 (16 mg, 91%) as a colorless pow-
der.
Aucubovir II 6: 1H NMR ([D6]DMSO): δ ϭ 1. 81 (m, 2 H, 2H-
5Ј), 2.38 (m, 1 H, H-5), 3.81 (m, 2 H, H-4Ј), 3.98, 4.33 (m, 2 H,
2H-4Ј), 3.76 (m, 2 H, 2H-5ЈЈ), 4.04 (m, 2 H, 2H-3Ј) 4.40 (m, 1 H,
H-1), 6.08 (m, 1 H, H-2), 8.18 (s, 1 H, H-2 adenine), 8.08 (s, 1 H,
H-8 adenine). 13C NMR ([D6]DMSO): δ ϭ 32.0 (C-5Ј), 43.6 (C-5),
44.7 (C-4), 60.1, 60.2 (C-4Ј, C-5ЈЈ), 61.2 (C-3Ј), 69.4 (C-1), 131.7
(C-2), 128.7 (C-5 adenine), 143.7 (C-8 adenine), 152.6 (C-4 aden-
ine), 152.0 (C-2 adenine), 151.4 (C-6 adenine). Ϫ [α]2D5 ϭ Ϫ69
(MeOH, c ϭ 0.1). Ϫ C14H19N5O3 (305.33): C 55.07, H 6.27, N
22.94; found C 54.96, H 6.33, N 22.80.
5-Hydroxyethyl-3,4-di(hydroxymethyl)cyclopenta-2-en-1-ol 3Ј,4Ј,5ЈЈ-
Triisobutyroyl-1-acetyl Ester (4): The reaction was accomplished by
dissolving 3 (50 mg) in pyridine (0.7 mL) followed by the addition
of acetic anhydride (0.35 mL) at room temperature. After 2 h the
starting product was no longer present in the reaction mixture,
which was then diluted with EtOAc (100 mL). After a simple
workup (washing with 2  HCl and successively with brine solution
until neutrality) the residue, obtained after evaporation in vacuo of
volatile material, was chromatographed on silica gel with CHCl3/
Et2O (9:1) to afford pure 4 (66 mg, about 100% yield).
Acknowledgments
Financial support from CNR and MURST.
1
4: H NMR (CDCl3): δ ϭ 1.11, 1.12, 1.13 [d, J ϭ 7.5 Hz, 3 ϫ 6
[1]
H, 3 ϫ (CH3)2CH], 1.87 (m, 2 H, 2H-5Ј), 2.02 (s, 3 H, CH3COO),
2.4Ϫ2.6 [4 H, H-5 superimposed to 3 ϫ (CH3)2CH], 2.93 (m, 1 H,
X part of an ABX system, H-4), 3.96, 4.40 (AB part of an ABX
system, JAB ϭ 11.0, JAX ϭ 2.5, JBX ϭ 4.0 Hz, 2 H, 2H-4Ј), 4.10
(m, 2 H, 2H-5ЈЈ), 4.64 (m, 2 H, 2H-3Ј), 5.49 (m, 1 H, H-1), 5.79
(m, 1 H, H-2). Ϫ 13C NMR (CDCl3): δ ϭ 19.1 [(CH3)2CHCO],
21.4 (CH3COO), 27.5 [(CH3)2CHCO], 34.1 (C-5Ј), 44.4 (C-5), 45.7
(C-4), 61.6, 62.0 (C-3Ј, C-5ЈЈ), 63.2 (C-4Ј), 83.3 (C-1), 129.5 (C-2),
R. Vince, M. Hua, J. Brownell, S. Daluge, F. Lee, W. M. Shan-
non, G. C. Lavelle, J. Qualls, O. S. Weislow, R. Kiser, P. G.
Canonico, R. H. Schultz, V. L. Narayanan, J. C. Mayo, R. H.
Shoemaker, M. R. Boyd, Biochem. Biophys. Res. Comm. 1988,
156, 1046Ϫ1053.
E. L. White, W. B. Parker, L. J. Macy, S. C. Shaddix, G. McCa-
leb, J. A. Secrist, R. Vince, W. M. Shannon, Biochem. Biophys.
Res. Comm. 1989, 161, 393Ϫ98.
R. Vince, M. Hua, J. Med. Chem. 1990, 31, 17Ϫ21.
M. F. Jones, Chem. Br. 1988, 1122Ϫ26.
D. M. Huryn, M. Okabe, Chem. Rev. 1992, 92, 1745Ϫ1768.
A. D. Borthwick, K. Biggadike, Tetrahedron 1992, 48,
571Ϫ623.
[2]
[3]
[4]
143.7 (C-3), 171.7 (CH3COO), 177.2, 177.3, 177.6 [3
ϫ
[5]
(CH3)2CHCO]. Ϫ [α]2D5 ϭ Ϫ81 (MeOH, c ϭ 0.1). Ϫ C23H36O8
[6]
(440.53): C 62.71, H 8.24; found C 62.65, H 8.33.
[7]
L. Agrofoglio, E. Suhas, A. Farese, R. Condom, S. R. Chal-
Coupling Reaction with 4: Under an argon atmosphere, N-benzoyl-
adenine (108 mg) was dissolved in anhydrous CH3CN (10 mL).
Subsequently, 1,1,1,3,3,3-hexamethyldisilazane (HMDS) (1.6 mL)
and trimethylsilyl chloride (TMSCl) (0.5 mL) were added. The so-
land, R. A. Earl, R. Guedj, Tetrahedron 1994, 50, 10611Ϫ70,
and references therein.
[8]
M. Tanaka, Y. Norimine, T. Fujita, H. Suemune, K. Sakai, J.
Org. Chem. 1996, 61, 6952Ϫ57.
Eur. J. Org. Chem. 2001, 1331Ϫ1334
1333

A. Bianco, L. Celletti, R. A. Mazzei, F. Umani
X.-F. Zhu, Nucleosides Nucleotides 2000, 19, 651.
A. Bianco, R. A. Mazzei, Tetrahedron Lett. 1997, 38, 6433Ϫ36.
H. Vorbrüggen, G. Höfle, Chem. Ber. 1981, 114, 1256Ϫ68.
FULL PAPER
[9]
[15]
[16]
[17]
[18]
R. Vince, P. Pham, Nucleosides Nucleotides 1995, 14, 2051.
[10]
R. Vince, J. Kilama, P. Pham, S. Beers, Nucleosides Nucleotides
1995, 14, 1703.
[11]
A. Grumann, H. Marley, R. Taylor, Tetrahedron Lett. 1995,
H. Vorbrüggen, B. Bennua, Chem. Ber. 1981, 114, 1279Ϫ86,
and references therein.
7767Ϫ68.
[12]
[19]
T. Berranger, Y. Langlois, Tetrahedron Lett. 1995, 5523Ϫ26.
A. Bianco, M. Guiso, C. Iavarone, P. Passacantilli, C. Trogolo,
Tetrahedron 1977, 33, 851Ϫ54.
Received July 9, 2000
[13]
S. Handa, G. Earlam, P. Geary, J. Hawes, G. Phillips, R. Pryce,
G. Ryback, J. Shears, J. Chem. Soc., Perkin Trans. 1 1994,
1885Ϫ86.
[O00339]
[14]
M. T. Crimmins, Tetrahedron 1998, 54, 9229Ϫ72.
1334
Eur. J. Org. Chem. 2001, 1331Ϫ1334