Synthesis of L-Oxetanocin

Article abstract of DOI:10.1021/ol025562x

matrix presented Hitherto unknown L-oxetanocin has been synthesized from L-xylose in 16 steps via a ribonolactone derivative.

Full text of DOI:10.1021/ol025562x

ORGANIC
LETTERS
2002
Vol. 4, No. 7
1147-1149
Synthesis of L-Oxetanocin
Giuseppe Gumina and Chung K. Chu*
Department of Pharmaceutical and Biomedical Sciences, College of Pharmacy,
The UniVersity of Georgia, Athens, Georgia 30602
dchu@rx.uga.edu
Received January 15, 2002
ABSTRACT
Hitherto unknown L-oxetanocin has been synthesized from L-xylose in 16 steps via a ribonolactone derivative.
Oxetanocin (Figure 1) was first isolated from the culture of
a strain of Bacillus megaterium.1 It has demonstrated good
antibacterial, antitumor, and antiviral activity.1,2
The most notable example regards oxathiolane analogues,
wherein 3TC (the L-isomer) is more potent and less toxic
than its D-counterpart [(+)-BCH-189].7 These findings,
together with the promising activity of the oxetanocin
nucleosides, prompted us to synthesize the L-isomer of
oxetanocin 1 (Figure 1).
As a result of its unusual structure, characterized by the
presence of an oxetane ring, oxetanocin has been the subject
of numerous synthetic studies,3,4 which led to the synthesis
of analogues where the adenine moiety is replaced by other
heterocyclic bases. Biologically, the most interesting com-
pound seems to be the thymine derivative, A-73209 (Figure
1), which shows potent in vitro and in vivo activity against
several thymidine kinase positive strains of VZV, HSV-1,
and HSV-2.5
As in all natural nucleosides, the absolute configuration
of the glycosyl moiety in oxetanocin is D. During the past
10 years, the interesting biological activities of several
“unnatural” L-nucleosides have been discovered.6,7
(1) Shimada, N.; Hasegawa, S.; Harada, T.; Tomisawa, T.; Fujii, A.;
Takita, T. J. Antibiot. 1986, 39, 1623-1625.
(2) Hoshino, H.; Shimizu, N.; Shimada, N.; Takita, T.; Takeuchi, T. J.
Antibiot. 1987, 40, 1077-1078.
(3) (a) Niitsuma, S.; Ichikawa, Y.; Kato, K.; Takita, T. Tetrahedron Lett.
1987, 28, 3967-3970. (b) Niitsuma, S.; Ichikawa, Y.; Kato, K.; Takita, T.
Tetrahedron Lett. 1987, 28, 4713-4714.
(4) (a) Nishiyama, S.; Yamamura, S.; Kato, K.; Takita, T. Tetrahedron
Lett. 1988, 29, 4743-4746. (b) Norbeck, D. W.; Kramer, J. B. J. Am. Chem.
Soc. 1988, 110, 7217-7218. (c) Norbeck, D. W.; Rosenbrook, W.; Plattner,
J. European Patent Application EP 433898, June 1991. (d) Wilson, F. X.;
Fleet, G. W. J.; Vogt, K.; Wang, Y.; Witty, D. R.; Choi, S.; Storer, R.;
Myers, P. L.; Wallis, C. J. Tetrahedron Lett. 1990, 31, 6931-6934.
(5) Alder, J.; Mitten, M.; Norbeck, D. W.; Marsh, K.; Kern, E. R.;
Clement, J. AntiViral Res. 1994, 23, 93-105.
Figure 1. Structures of D- and L-oxetanocin analogues.
10.1021/ol025562x CCC: $22.00 © 2002 American Chemical Society
Published on Web 03/05/2002

Scheme 1. Synthesis of L-Oxetanocin 1
a Reagents and conditions: (a) seven steps, ref 10; (b) NaH, THF, then TBAI, BnBr; (c) 10% HCl, 1,4-dioxane; (d) Br2, BaCO3, 3:1
H2O/1,4-dioxane; (e) MsCl, Py, CH2Cl2; (f) 1 N NaOH, MeOH; (g) i-BuOC(O)Cl, N-methylmorpholine, THF, then 2-mercaptopyridine-
N-oxide, Et3N, then hν; (h) N6-benzoyladenine, Br2, 4 Å molecular sieves, DMF; (i) MeONa, MeOH; (j) H2, Pd black, EtOH.
From a review of the current literature, none of the
synthetic schemes for the synthesis of natural oxetanocin
published so far seemed conveniently applicable to the
synthesis of L-oxetanocin. The first reported total synthesis
of D-oxetanocin,3 starting from D-ribose, led to the final
product in 19 steps with an overall yield of 0.008% and was
not applicable to the synthesis of other analogues. In another
approach,4a oxetanocin and its analogues were prepared in
26 steps from D-glucose with 0.15% overall yield. In this
procedure, the starting material was converted to an acyclic
structure that, eventually, was recyclized to afford the four-
membered ring.
Among the reported syntheses of D-oxetanocin, the method
that utilizes adenosine appears to be the most efficient, in
which adenosine was converted to oxetanocin in 12 steps,
with an overall yield of approximately 7%.4b By varying the
starting nucleoside, a series of analogues was successfully
obtained.4c The main advantage of this approach is that the
five-membered ring is contracted to an oxetane when it is
already bound to the base, thereby bypassing a difficult
condensation step involving an oxetane derivative. On the
other hand, this procedure seems to be less convenient in
the case of analogues bearing “unnatural” heterocyclic
moieties, as well as L-nucleosides, as a natural starting
material is not readily available.
In devising our scheme, we had to consider the reactivity
of oxetane systems. These are highly strained and tend to
open, particularly in the presence of acids, to give acyclic
compounds or, in the presence of particular substituents,
oxolane-type rearrangement products.8 This explains the low
yields observed in the early syntheses,3,4a where the base and
the sugar were coupled in the presence of a Lewis acid as
the catalyst. One way to avoid the use of an acidic catalyst
is to condense the base, free or activated as an anion, with
an oxetane substituted at the anomeric position with a good
leaving group. Unfortunately, such derivatives proved to be
very unstable, as reported in another low yield synthetic
procedure.4d
Our approach was based on some interesting syntheses in
which a base is coupled to a sugar, activated as thioglyco-
side.8,9 In most cases, the reaction is catalyzed by a Lewis
acid; however, in certain cases, a protected adenine is reacted
with a thiopyridyloxetane in the presence of bromine to give
the coupled product in good yield.8
As the starting material of our synthesis (Scheme 1), we
used protected 3-deoxy-3-(hydroxymethyl)-L-ribofuranose 2,
previously prepared from L-xylose in our laboratory.10
(8) Saksena, A. K.; Ganguly, A. K.; Girijavallabhan, V. M.; Pike, R. E.;
Chen, Y.-T.; Puar, M. S.; Tetrahedron Lett. 1992, 33, 7721-7724.
(9) (a) Pedersen, C.; Fletcher, H. G., Jr. J. Am. Chem. Soc. 1960, 82,
5210-5211. (b) Norton, D. Pure Appl. Chem. 1975, 42, 301-325. (c)
Hanessian, S.; Dixit, D. M.; Liak, T. J. Pure Appl. Chem. 1981, 53, 129-
148. (d) Hanessian, S.; Sato, K.; Liak, T. J.; Danh, N.; Dixit, D. J. Am.
Chem. Soc. 1984, 106, 6114-6115.
(6) Gumina, G.; Song, G.-Y.; Chu, C. K. FEMS Microbiol. Lett. 2001,
202, 9-15.
(7) Schinazi, R. F.; Chu, C. K.; Peck, A.; McMillan, A.; Mathis, R.;
Cannon, D.; Jeong, L.-S.; Beach, J. W.; Choi, W.-B.; Yeola, S.; Liotta, D.
C. Antimicrob. Agents Chemother. 1992, 36, 672-676.
(10) Cooperwood, J. S.; Boyd, V.; Gumina, G.; Chu, C. K. Nucleosides
Nucleotides 2000, 19, 219-236.
1148
Org. Lett., Vol. 4, No. 7, 2002

Alcohol 2 was benzylated in the presence of catalytic
tetrabutylammonium iodide (TBAI)11 to give the fully
protected derivative 3, which was deprotected to lactol 4.
The oxidation of 4 to lactone 5, accomplished by the
modification of a reported procedure,12 required a very
accurate control of pH, which needed to be maintained
between 4.5 and 4.8 in order to obtain satisfactory yields.
Lactone 5 was quantitatively mesylated to compound 6,
which underwent ring contraction, following the procedure
reported for the D-isomer,8 to afford acids 713 and 814 in 3:1
ratio in 73% yield. These acids, purified and characterized
separately, were reacted as an epimeric mixture in a three-
step, one-pot reaction to afford the epimeric thiopyridyl
oxetanes 9 in 30% yield. The epimeric mixture of compounds
9, not separable by column chromatography, was used in
the coupling reaction to yield fully protected oxetanocin
derivatives 10 and 11 in 2:3 ratio and 61% overall yield.
Compounds 10 and 11 could be separated by flash
chromatography, and each of them was debenzoylated to the
corresponding derivatives 12 and 13, which were in turn
catalytically hydrogenated over palladium black to yield
oxetanocin 115 and its R-isomer 1416 with overall yields of
90%.17
Figure 2. NOE correlations and stereochemistry of 7-13.
of â-isomers, whereas correlations between H2 and H3
characterized the spectra of R-isomers. In most cases,
correlations between H2 and H3′a and/or H3′b and between
H2 and H4′a and/or H4′b were also observed.
In the case of 9, these correlations could be clearly
observed in the spectra of the epimeric mixture and helped
1
correctly assign H signals in each isomer.
In summary, L-oxetanocin has been synthesized from
L-xylose in 16 steps and 2.8% overall yield. Our procedure
seems more convenient and versatile than the reported total
syntheses of D-oxetanocin, because the coupling between the
activated oxetane moiety and the nucleobase does not require
a Lewis acid, thereby proceeding smoothly and with minimal
side reactions. Optimization of the synthesis for other purine
and pyrimidine derivatives as well as biological evaluations
of L-oxetanocin are in progress.
Stereochemistry of epimers 7-13 has been unequivocally
determined by means of 2D-NOESY experiments on both
isomers; in all cases (Figure 2), very strong correlations
between protons H2 and H4 have been observed in the spectra
(11) Czernecki, S.; Georgoulis, C.; Provelenghiou, C. Tetrahedron Lett.
1976, 39, 3535-3536.
(12) Pudlo, J. S.; Townsend, L. B. Nucleic Acid Chemistry, part 4;
Townsend, L. B., Tipson, R. S., Eds.; Wiley: New York, 1978; p 1151.
(13) Colorless oil: Rf 0.14 (95:4.5:0.5 CHCl3/MeOH/AcOH); [R]25D 11.4
(c 0.92, CHCl3); 1H NMR (400 MHz, CDCl3) δ 7.38-7.26 (m, 10H), 5.08
(d, 1H, J ) 9.3 Hz), 4.98 (m, 1H, J ) 5.6, 4.1, 2.9), 4.67 (d, 1H, J )
11.9), 4.60 (d, 1H, J ) 11.9), 4.54 (d, 1H, J ) 11.7), 4.37 (d, 1H, J )
11.7), 3.72 (dd, 1H, J ) 11.4, 2.9), 3.66 (dd, 1H, J ) 11.4, 4.1), 3.60 (dd,
1H, J ) 10.2, 4.2), 3.55 (dd, 1H, J ) 10.2, 3.6), 3.43 (m, 1H, J ) 9.3, 5.6,
4.2, 3.6); 13C NMR (100 MHz, CDCl3) δ 172.66, 137.79, 137.56, 128.47,
128.37, 127.84, 127.82, 127.74, 127.64, 82.16, 75.68, 73.64, 73.31, 71.32,
66.01, 39.04; HRMS (FAB) m/z calcd for C20H23O5 [(M + H)+] 343.1545,
found 343.1531. Anal. Calcd for C20H22O5‚0.1MeOH: C, 69.86; H, 6.53.
Found: C, 69.48; H, 6.54.
Preliminary biological evaluation of L-oxetanocin showed
no activity up to 100 mM against HIV-1.
Acknowledgment. This research was supported by the
U.S. Public Health Service Research Grant AI 32351 from
the National Institute of Allergy and Infectious Diseases,
NIH. One of the Authors (G.G.) was partially supported by
the Italian Department of University and Scientific Research
(MURST). We thank Dr. Michael Bartlett of the College of
Pharmacy, The University of Georgia for performing the
high-resolution mass spectra.
(14) White solid: mp 80-81 °C; Rf 0.25 (95:4.5:0.5 CHCl3/MeOH/
AcOH); [R]25D -41.3 (c 0.35, CHCl3); 1H NMR (400 MHz, CDCl3) δ 7.36-
7.28 (m, 10H), 4.93 (d, 1H, J ) 6.5 Hz), 4.88 (m, 1H, J ) 6.1, 2.3, 1.6),
4.73 (d, 1H, J ) 12.0), 4.64 (d, 1H, J ) 12.0), 4.57 (d, 1H, J ) 12.1), 4.53
(d, 1H, J ) 12.1), 3.76 (dd, 1H, J ) 11.4, 2.3), 3.68 (dd, 1H, J ) 9.8, 5.7),
3.61 (dd, 1H, J ) 9.8, 4.3), 3.48 (dd, 1H, J ) 11.4, 1.6), 3.31 (m, 1H, J )
6.5, 6.1, 5.7, 4.3); 13C NMR (100 MHz, CDCl3) δ 172.34, 137.62, 136.39,
128.69, 128.50, 128.34, 128.15, 127.89, 127.67, 81.64, 76.78, 73.77, 73.26,
70.18, 68.19, 39.84; HRMS (FAB) m/z calcd for C20H23O5 [(M + H)+]
343.1545, found 343.1530. Anal. Calcd for C20H22O5: C, 70.16; H, 6.48.
Found: C, 70.39; H, 6.80.
OL025562X
(16) White solid: mp 170-172 °C.; Rf 0.07 (1:9 MeOH/CH2Cl2); [R]26
D
14.82 (c 0.22, MeOH); UV (H2O) λmax 257.0 (ꢀ 14685, pH 2), 259.0 (ꢀ
14738, pH 7), 254.0 (ꢀ 14280, pH 11); 1H NMR (400 MHz, CD3OD) δ
8.47 (s, 1H), 8.18 (s, 1H), 6.69 (d, 1H, J ) 6.2), 4.97 (m, 1H), 3.90 (dd,
1H, J ) 12.9, 2.7), 3.78 (dd, 1H, J ) 12.9, 3.9), 3.50-3.61 (m, 3H); 13C
NMR (100 MHz, CDCl3) δ 157.33, 153.86, 150.13, 141.14, 120.38, 86.31,
84.77, 64.48, 59.40, 43.72; HRMS (FAB) m/z calcd for C10H14N5O3 [(M
+ H)+] 252.1097, found 252.1092. Anal. Calcd for C10H13N5O3‚0.25CH2-
Cl2: C, 45.18; H, 4.99; N, 25.70. Found: C, 45.30; H, 5.22; N, 25.37.
(17) All the synthesized compounds showed satisfactory analytical and
spectroscopic data. When possible, comparison of their data with those of
known D-enantiomers confirmed their identity and optical purity.
(15) White solid: mp 189-190 °C; Rf 0.12 (1:9 MeOH/CH2Cl2); [R]27
D
20.38 (c 0.36, MeOH); UV (H2O) λmax 256.0 (ꢀ 18950, pH 2), 258.0 (ꢀ
19430, pH 7), 258.0 (ꢀ 19420, pH 11); 1H NMR (400 MHz, CD3OD) δ
8.66 (s, 1H), 8.20 (s, 1H), 6.50 (d, 1H, J ) 5.3), 4.67 (m, 1H), 3.91 (dd,
1H, J ) 13.2, 2.3), 3.82 (m, 3H), 3.77 (dd, 1H, J ) 13.2, 3.0); 13C NMR
(100 MHz, CDCl3) δ 157.48, 153.86, 150.27, 141.78, 120.26, 83.17, 80.08,
63.97, 60.31, 46.85; HRMS (FAB) m/z calcd for C10H14N5O3 [(M + H)+]:
252.1097, found: 252.1114. Anal. Calcd for C10H13N5O3‚H2O: C, 44.61;
H, 5.62; N, 26.01. Found: C, 44.84; H, 5.62; N, 26.12.
Org. Lett., Vol. 4, No. 7, 2002
1149