Efficient synthesis of clitocine via 1,3-N (endo) to N (exo) migration: A revision to Kini's work

Article abstract of DOI:10.1055/s-2005-871564

Efficient synthesis of clitocine has been accomplished via unprecedented 1,3-N (endo) to N (exo) migration as a key transformation. Incorporation of p-chlorobenzoyl (PCB) group as a protecting group led to the easy solidification of the intermediate of 7,

Full text of DOI:10.1055/s-2005-871564

1942
LETTER
Efficient Synthesis of Clitocine via 1,3-N (endo) to N (exo) Migration:
A Revision to Kini’s Work
Efficient
S
ynthesis
y
of Clitocineeong-wook Choi,a Bo Seung Choi,a Jay Hyok Chang,a Kyu Woong Lee,a Do Hyun Nam,a Young Keun Kim,a
Jae Hoon Lee,a Taeho Heo,a Hyunik Shin,*a No-Soo Kimb
c
Chemical Development Division, LG Life Sciences, Ltd./R&D, 104-1 Moonji-dong, Yusong-gu, Daejon 305-380, Korea
Fax +82(42)8662471; E-mail: hisin@lgls.co.kr
d
Faculty of Life Science Engineering, Youngdong University, San 12-1 Sulkye-ri, Yeongdong-eup, Yeongdong-gun, Chungbuk, Korea
Received 30 April 2005
This paper is dedicated to Prof. James D. White.
Abstract: Efficient synthesis of clitocine has been accomplished
via unprecedented 1,3-N (endo) to N (exo) migration as a key trans-
formation. Incorporation of p-chlorobenzoyl (PCB) group as a pro-
tecting group led to the easy solidification of the intermediate of 7,
thus making the isolation very facile and to the minimization of the
epimerization of the anomeric center at the final deprotection stage.
Key words: 1,3-N (endo) to N (exo) migration, isomerizations,
clitocine
Clitocine [6-amino-5-nitro-4-(b-D-ribofuranosylamino)py-
rimidine], isolated from the mushroom Clitocybe inversa
by Kubo et al.,1 has potent insecticidal activity against
Pectinophora gossypiella. Later it was found that clito-
cine has also strong cytostatic effect towards several
leukemia cell lines and is an inhibitor of adenosine
Scheme 1
kinase.2 The interesting biological activity together with
biogenetically close relationship with adenosine triggered
intensive synthetic studies to lead to the completion of
two independent total syntheses of 1. Kamikawa’s group3
used the condensation reaction of ribofuranosylamine
derivatives with 4,6-dichloro-5-nitropyrimidine as a key
step (Scheme 1, route A), while Kini’s group2 employed a
direct coupling of silylated 4, 6-diamino-5-nitropyrimi-
dine (3) with 1-O-acetyl-2,3,5-tri-O-benzoyl-D-ribofura-
nose (4) as a critical step (Scheme 1, route B). Compared
to the former synthesis, the latter approach is more
efficient in terms of chemical yield and stereocontrol,
particularly at the anomeric center.
isomerization of the endo-product 6 to the exo-product 8
during silica gel chromatography, we further treated the
3:1 mixture with silica gel (100% wt/wt to the mixture) for
24 hours in dichloromethane to reach the equilibrated ra-
tio of 93:7. Thus, contrary to the reported result, the endo-
product 6 was formed exclusively as a kinetic product,
which was isomerized to the thermodynamically con-
trolled exo-product 8 through 1,3-N (endo) to N (exo) mi-
gration under acid catalysis. Another drawback of Kini’s
synthesis is the partial epimerization of the anomeric cen-
ter at the final deprotection stage: 98:2 of b:a anomer ratio
of 8 was deteriorated to 92:8 of b:a anomer ratio of 1
under the deprotection conditions (Scheme 2).
As a part of our interest towards further biological evalu-
ation of 1, we repeated Kini’s synthesis. In the repetition
of the coupling reaction of 34 and 4, we were surprised by
the observation that no desired product was formed on the
basis of the 1H NMR spectrum of the crude mixture: nei-
ther the doublet NH proton at d = 9.63 ppm nor the broad
NH2 protons at d = 8.62 ppm of the expected product 8
were observed.5 However, column chromatography of the
crude product provided pure 8 in 28% yield and a 3:1 mix-
ture of the desired 8 and the side product in 23% yield that
was later identified as the endo-isomer 6. Hinted by the
To minimize the partial epimerization of the anomeric
center, we changed the benzoyl protecting group to p-
chlorobenzoyl (PCB) group on the assumption that more
electron deficient aromatic ring would facilitate the meth-
anolysis reaction. In the same way as with the benzoate
derivatives, 1-acetoxy-PCB-protected ribofuranose 5 was
prepared without any incidence. Coupling of 3 and 5 un-
der the conditions used by Kini et al., provided exclusive-
ly the endo-product 7. Serendipitously, the kinetic endo-
product 7 was easily solidified in high purity simply by
triturating the crude reaction mixture with ethyl acetate–
diethyl ether co-solvent to make the isolation of the prod-
SYNLETT 2005, No. 12, pp 1942–1944
0
1
.0
8
.2
0
0
5
Advanced online publication: 04.07.2005
DOI: 10.1055/s-2005-871564; Art ID: U12805ST
© Georg Thieme Verlag Stuttgart · New York

LETTER
Efficient Synthesis of Clitocine
1943
Figure 2 ORTEP plot of 7.
anomeric center. Simple recrystallization of the crude re-
action mixture in methanol provided 1 having 98:2 ratio
of b/a anomer in 51% yield. One more recrystallization
cycle of the filtrate provided additional 32% of 1 having
the same purity of the first crop.8
Scheme 2 Reagents and conditions: i) HMDS, pyridine, H2SO4,
(NH4)2SO4; ii) TMSOTf; iii) silica gel or HOAc; iv) NaOMe, MeOH–
dioxane.
1
In conclusion, we have discovered that the endo-product
6 or 7 was kinetically formed in the glycosylation stage,
which was isomerized to the thermodynamic exo-product
8 or 9, respectively, under acid catalysis. Fortunately, the
incorporation of the PCB group facilitated the solidifica-
tion of the endo-product 7, thus making the isolation and
purification very facile. Isomerization and subsequent
deprotection underwent without any sign of the deteriora-
tion of the anomeric purity. The established synthesis pro-
vides highly pure clitocine (1) without any column
separation and delineates the new reaction pathway, 1,3-
N (endo) to N (exo) migration.
uct very facile (for the H NMR spectra of 7 and 9, see
Figure 1). Furthermore, X-ray analysis of a single crystal
of 7 unambiguously determined the structure of the endo-
product (Figure 2).6 Subsequent isomerization of the
endo-product 7 by silica gel for 24 hours gave exclusively
the exo-product 9 in high yield and purity.7 Besides silica
gel, Brønsted acid such as acetic acid also caused isomer-
ization to reach the exo:endo ratio of 96:4 after 24 hours
but Lewis acids like lithium chloride or magnesium
chloride did not trigger isomerization.
As expected, the deprotection of the PCB group proceed-
ed much faster (in 2 h at 0 °C) than the benzoyl group (in
17 h at 0–8 °C) to result in no epimerization of the
Figure 1 1H NMR spectra of 7 and 9.
Synlett 2005, No. 12, 1942–1944 © Thieme Stuttgart · New York

1944
H.-w. Choi et al.
LETTER
H, d, J = 2.4 Hz), 6.12 (1 H, dd, J = 6.4, 2.4 Hz), 6.06 (1 H,
dd, J = 7.2, 6.4 Hz), 4.60–4.80 (3 H, m). 13C NMR (100
MHz, DMSO-d6): d = 164.8, 164.0, 163.9, 158.5, 152.3,
148.5, 139.0, 138.9, 138.7, 131.4, 131.3, 131.2, 129.2,
129.1, 128.2, 127.7, 127.5, 111.3, 91.2, 78.7, 74.1, 70.4,
63.7. IR: 3424, 3300, 1713, 1632, 1589, 1478, 1396, 1252,
1087, 1009, 752 cm–1. UV/Vis (CH2Cl2): lmax = 243, 374
nm.
Acknowledgment
We thank Prof. M. S. Lah of Hanyang University for the X-ray
analysis of the compound 7.
References
(1) Kubo, I.; Kim, M.; Wood, W. F.; Naoki, H. Tetrahedron
Lett. 1986, 27, 4277.
(2) Moss, R. J.; Petrie, C. R.; Meyer, R. B. Jr.; Nord, L. D.;
Willis, R. C.; Smith, R. A.; Larson, S. B.; Kini, G. D.;
Robins, R. K. J. Med. Chem. 1988, 31, 786.
6-Amino-5-nitro-4-[(2,3,5-tri-O-para-chlorobenzoyl-b-D-
ribofuranosyl)amino]pyrimidine (9).
A solution of 7 (2.166 g, 3.08 mmol) in methylene chloride
(60 mL) was stirred with silica gel (2.2 g) for 20 h. Silica gel
was filtered off and washed with CH2Cl2. The filtrate was
concentrated to give 9 (2.06 g, 95%) as an yellow foam. 1H
NMR (400 MHz, DMSO-d6): d = 9.60 (1 H, d, J = 8.0 Hz),
8.58 (2 H, br s), 8.05 (1 H, s), 8.00 (2 H, dd, J = 6.8, 2.0 Hz),
7.90 (2 H, dd, J = 6.8, 2.0 Hz), 7.84 (2 H, dd, J = 6.8, 2.0
Hz), 7.55 (6 H, m), 6.35 (1 H, dd, J = 8.4, 4.8 Hz), 5.97 (1 H,
dd, J = 5.6, 4.8 Hz), 5.89 (1 H, dd, J = 6.0, 5.2 Hz), 4.65 (2
H, m), 4.55 (1 H, m). 13C NMR (100 MHz, DMSO-d6): d =
165.5, 164.8, 164.7, 160.0, 159.4, 157.4, 139.8, 139.7,
139.4, 132.0, 131.9, 129.9, 129.8, 129.7, 128.9, 128.3,
128.2, 113.4, 84.8, 78.7, 75.1, 72.1, 64.8. UV/Vis (CH2Cl2):
lmax = 240, 331 nm.
(3) Kamikawa, T.; Fujie, S.; Yamagiwa, Y.; Kim, M.;
Kawaguchi, H. J. Chem. Soc., Chem. Commun. 1988, 195.
(4) (a) Boon, W. R.; Jones, W. G. M.; Ramage, G. R. J. Chem.
Soc. 1951, 96. (b) Robins, R. K.; Dille, K. J.; Willits, C. H.;
Christensen, B. E. J. Am. Chem. Soc. 1953, 75, 263.
(5) We observed peaks at d = 9.61 (s, 1 H), 9.43 (s, 1 H), and
9.35 (s, 1 H) in DMSO-d6, which are typical peaks of the
endo-product 6. Since 6 was not solidified under various
conditions, it was difficult to purify the product thoroughly.
(6) X-ray data for the compound 7: C30H22Cl3N5O9, monoclinic,
space group P21 with cell parameters: a = 14.6041 (11) Å,
b = 6.8064 (5) Å, c = 16.1106 (12) Å; a = 90°, b = 108.0770
(10)°, g = 90°; V = 1522.4 (2) Å3; Dc = 1.533 mg/m3; Z = 2.
(7) 4-Amino-6-imino-5-nitro-1-N-(2,3,5-tri-O-para-
chlorobenzoyl-b-D-ribofuranosyl)pyrimidine (7).
A suspension of 4, 6-diamino-5-nitropyrimidine (1.0 g, 6.45
mmol) in HMDS (30 mL) was treated with pyridine (4.8
mL), H2SO4 (0.2 mL) and ammonium sulfate (60 mg). The
mixture was refluxed at 130 °C for 18 h, and HMDS was
evaporated in vacuo. The residual solid was dissolved in
MeCN (50 mL), cooled to 0 °C, and treated with 1-O-acetyl-
2,3,5-tri-O-para-chlorobenzoyl-D-ribofuranose (5, 4.5 g,
7.42 mmol) and TMSOTf (2.0 mL, 11.0 mmol). The mixture
was stirred at 0 °C for 23 h, and it turned into dark clear
solution. Then, MeCN was evaporated in vacuo, and the
residue was treated with sat. NaHCO3 (100 mL). The
aqueous layer was extracted with CH2Cl2 (100 mL × 2), and
the combined organic phase was dried over anhyd MgSO4.
After concentration, the residue was triturated with EtOAc
(4 mL) and enough amount of Et2O. The resulting solid was
filtered and washed with Et2O to give the endo-product 7
(2.44 g, 55%) as a yellow powder. 1H NMR (400 MHz,
DMSO-d6): d = 9.61 (1 H, s), 9.44 (1 H, s), 9.36 (1 H, s), 8.33
(1 H, s), 7.98 (2 H, dd, J = 6.8, 2.0 Hz), 7.94 (2 H, dd,
J = 6.8, 2.0 Hz), 7.81 (2 H, dd, J = 6.8, 2.0 Hz), 7.57 (4 H, 2
dd, J = 6.8, 2.0 Hz), 7.50 (2 H, dd, J = 6.8, 2.0 Hz), 6.33 (1
(8) Clitocine (1).
6-Amino-5-nitro-4-[(2,3,5-tri-O-para-chlorobenzoyl-b-D-
ribofuranosyl)amino]pyrimidine (9, 1.93 g, 2.75 mmol) was
dissolved in 1,4-dioxane (26 mL) and diluted with MeOH
(65 mL). After cooling to 0 °C, the solution was treated with
MeONa (0.55 mmol, 0.1 mL of 28% MeOH soln was diluted
with 0.4 mL of MeOH) and stirred at 0 °C for 2 h. The
resulting solution was stirred with Amberlite IR-120 (900
mg, 3.96 mol equiv, acidic resin was washed with distilled
H2O and MeOH before use) for 30 min. The resin was
filtered off and the filtrate was concentrated in vacuo. The
residue was triturated with Et2O and washed with Et2O (10
mL × 3). The solid was recrystallized from hot MeOH (135
mL) to give 1 (≥ 98% of b-anomer, 400 mg, 51%) as a
slightly yellow solid. More clitocine (1, ≥ 98% of b-anomer,
255 mg, 32%) was obtained from the mother liquor. 1H
NMR (400 MHz, DMSO-d6): d = 9.28 (1 H, d, J = 7.6 Hz),
8.56 (2 H, s), 8.00 (1 H, s), 5.79 (1 H, dd, J = 7.6, 3.6 Hz),
5.20 (1 H, d, J = 5.2 Hz), 5.06 (1 H, t, J = 4.8 Hz), 4.92 (1 H,
d, J = 6.0 Hz), 4.05 (1 H, m), 3.93 (1 H, m), 3.79 (1 H, m),
3.53 (1 H, m), 3.47 (1 H, m). 13C NMR (100 MHz, DMSO-
d6): d = 160.2, 159.5, 156.8, 112.8, 87.0, 84.8, 75.9, 70.8,
61.3.
Synlett 2005, No. 12, 1942–1944 © Thieme Stuttgart · New York