Synthesis of griseolic acid B by pi-face-dependent radical cyclization.

Article abstract of DOI:10.1021/ol0167210

[reaction: see text]. Radical cyclization of 6 affords the bicyclic vinyl ether 9 with the appropriate stereochemistry for elaboration (seven steps) to griseolic acid B (1).

Full text of DOI:10.1021/ol0167210

ORGANIC
LETTERS
2001
Vol. 3, No. 22
3583-3585
Synthesis of Griseolic Acid B by
π-Face-Dependent Radical Cyclization
,†
Spencer Knapp,* Machender R. Madduru,† Zhijian Lu,† Gregori J. Morriello,‡
Thomas J. Emge,† and George A. Doss‡
Department of Chemistry & Chemical Biology, Rutgers The State UniVersity of
New Jersey, Piscataway, New Jersey 08854-8087, and Merck & Co., P.O. Box 2000,
Rahway, New Jersey 07065-0900
knapp@rutchem.rutgers.edu
Received September 7, 2001
ABSTRACT
Radical cyclization of 6 affords the bicyclic vinyl ether 9 with the appropriate stereochemistry for elaboration (seven steps) to griseolic acid
B (1).
Radical reactions have been heralded for application to the
synthesis of densely functionalized targets because polar
functional groups are often less conflictive in uncharged
radical intermediates than they might be in anionic or cationic
species, particularly during C-C bond formation. Further-
more, predicting and exploiting stereoselectivity in radical
reactions grows easier with each new application.1 We report
the synthesis of griseolic acid B (1) based upon a novel vinyl
radical cyclization reaction in which the acceptor alkene face
selectivity dictates both stereochemistry and ring size.
Griseolic acids are a family of complex nucleosides
isolated at Sankyo Co. from a cultured broth of Strepto-
mycetes griseoaurantiacus.2 Three representatives, griseolic
acids A, B, and C, exhibited comparable nanomolar inhibition
of cyclic adenosine 3′,5′-monophosphate phosphodiesterases.3
Spectroscopic and crystallographic analysis revealed nucleo-
side skeletons that broadly resemble cyclic AMP: A (2) and
B (1) feature unsaturated bicyclic sugar dicarboxylate gly-
† Rutgers The State University of New Jersey.
‡ Merck & Co.
(1) Some recent reviews and monographs: Curran, D. P. Aldrichimica
Acta 2000, 33, 104-110. A¨ıssa, C.; Delouvrie´, B.; Dhimane, A.-L.;
Fensterbank, L.; Malacria, M. Pure Appl. Chem. 2000, 72, 1605-1613.
Giese, B.; Kopping, B.; Go¨bel, T.; Dickhaut, J.; Thoma, G.; Kulicke, K. J.;
Trach, R. Org. React. 1996, 48, 301-856. Curran, D. P.; Porter, N. A.;
Giese, B. StereoselectiVity of Radical Reactions. Concepts, Guidelines, and
Synthetic Applications; VCH: Weinheim, 1995. Yet, L. Tetrahedron 1999,
55, 9349-9403. Sibi, M. P.; Ternes, T. R. In Modern Carbonyl Chemistry;
Otera, J., Ed.; Wiley-VCH: Weinheim, 2000; pp 507-538. Radicals in
Organic Synthesis; Renaud, P., Sibi, M. P., Eds.; Wiley-VCH: Weinheim,
2001.
cones, whereas C is a 4′(R),5′-dihydro version of B.
Extensive studies of griseolic analogues were carried out at
Sankyo4 and at Schering-Plough,5 and a synthesis of 2 has
been published by Tulshian and Czarniecki.6
Sporting almost as many functional groups as carbon
atoms, griseolic acid B (1) carries additional synthetic
challenges: a [3.3.0] ring system without obvious convex
facial bias, a quaternary stereogenic center at C-6′, a reactive
(2) Nakagawa, F.; Okazaki, T.; Naito, A. J. Antibiot. 1985, 38, 823-
829. Takahashi, S.; Nakagawa, F.; Kawazoe, K.; Furukawa, Y.; Sato, S.;
Tamura, C.; Naito, A. J. Antibiot. 1985, 38, 830-834. Takahashi, S.;
Nakagawa, F.; Sato, S. J. Antibiot. 1988, 41, 705-706.
(3) Iijima, Y.; Nakagawa, F.; Handa, S.; Oda, T.; Naito, A.; Yamazaki,
M. FEBS Lett. 1985, 192, 197-183.
10.1021/ol0167210 CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/04/2001

C-4′ vinyl ether with a greater tendency to form a cation
than the C-1′ acetal, and a purine with multiple Lewis basic
sites. Radical cyclization to form the C-5′,6′ bond is an
interesting possibility. On the basis of our familiarity with
the synthesis of complex nucleoside antibiotics,7 we con-
sidered both early and late introduction of the purine but
found more success with late purine attachment even though
this necessitated installation of the reactive vinyl ether later
still.
7, but the most success was realized with 0.05 M n-Bu3SnH
in refluxing benzene. Under these conditions, two products,
8 and 9 (Scheme 2), were formed in a 3:2 ratio and 82%
Scheme 2. Radical Cyclization and CdC Protectiona
A C-5′ (griseolic acid numbering) vinyl radical precursor
was assembled from the commercial D-allo-furanose 3 as
shown in Scheme 1. Aldehyde 4 was converted to vinyl
Scheme 1. Synthesis of the Radical Cyclization Substratea
a Conditions: (j) PhSH, AIBN, CHCl3, NaHCO3, reflux; (k)
H2SO4, HOAc, Ac2O, CH2Cl2, 0 °C.
combined yield. The ratio was essentially unchanged at
n-Bu3SnH concentrations from 0.02 to 0.33 M. No other
product was discernible by TLC or NMR analysis. Charac-
terization of 8 and 9 as 6-endo-trig and 5-exo-trig cyclization
products, respectively, followed from 1H NMR analysis, and
the stereochemistry of each was established unambiguously
by X-ray crystallography. Thus, 9 matches the required
griseolic acid stereochemistry at C-6′.
The stereoselectivity of the cyclization of 6 is remarkable.
The presumed intermediate vinyl radical in 7 (Scheme 1)
must add, in the endo mode, to the re face of C-b in the
maleate alkene to set the stereochemistry at C-6′ of 8 (dashed
arrow), and in the exo mode to the opposite maleate face of
C-a (also re) to set the stereochemistry at C-6′ of 9 (solid
arrow). Subsequent hydrogen atom abstraction from the less
hindered â-face sets the stereochemistry at C-7′ of 8. In other
words, the maleate face that is exposed to the vinyl radical
determines both the stereochemistry and ring size of the
products. Possible interconversion of homoallylic radical
intermediates is ruled out by the concentration studies and
by stereoelectronic considerations.10
a Conditions: (a) t-BuPh2SiCl, imidazole, DMF; (b) aqueous
HOAc; (c) NaIO4, aqueous THF; (d) NH2NH2, EtOH; (e) I2, Et2O,
tetramethylguanidine; (f) DBU, toluene, 60 °C; (g) n-Bu4NF, THF;
(h) LiN(SiMe3)2 (0.2 equiv), Et2O, diethyl acetylenedicarboxylate;
(i) n-Bu3SnH, AIBN, benzene, 80 °C.
iodide 5 (95:5 mixture of isomers) by iodination of the
derived hydrazone,8 and then deprotection at O-3′ and
addition of this hydroxyl to diethyl acetylenedicarboxylate
under basic conditions afforded maleate derivative 6. The E
geometry is assigned on the basis of the vinyl-H chemical
shift (5.36 ppm).9 Iodine atom abstraction from 6 should give
vinyl radical 7.
For the synthesis to proceed from 9, temporary protection
of the CdC is required. This was accomplished by radical
addition of thiophenol to C-5′ (Scheme 2). Chloroform
proved to be a superior solvent for this addition, and NaHCO3
was added to neutralize any trace of acid that might promote
competing Markovnikov addition. In the best case the
reaction was run to half-completion, and unreacted 9 was
recycled. An easily separable mixture of thiophenol adducts
10 was obtained: the structure of 10-endo was proven
crystallographically, while 10-exo was taken on by acetolysis
to the anomeric acetate 11.
A variety of reducing reagents, initiators, solvents, con-
centrations, and temperatures were tried in order to generate
(4) Murofushi, Y.; Kimura, M.; Iijima, Y.; Yamazaki, M.; Kaneko, M.
Chem. Pharm. Bull. 1987, 35, 1036-1043. Murofushi, Y.; Kimura, M.;
Iijima, Y.; Yamazaki, M., Kaneko, M. Chem. Pharm. Bull. 1987, 35, 4442-
4453. Murofushi, Y.; Kimura, M.; Iijima, Y.; Yamazaki, M.; Kaneko, M.
Chem. Pharm. Bull. 1988, 36, 1309-1320. Murofushi, Y.; Kimura, M.;
Kuwano, H.; Iijima, Y.; Yamazaki, M.; Kaneko, M. Chem. Pharm. Bull.
1988, 36, 3760-3769.
(5) Tulshian, D.; Doll, R. J.; Stansberry, M. F. J. Org. Chem. 1991, 56,
6819-6822. Tulshian, D.; Czarniecki, M.; Doll, R. J.; Ahn, H.-S. J. Med.
Chem. 1993, 36, 1210-1220.
Vorbru¨ggen glycosylation11 of 11 (Scheme 3) with bis-
(trimethylsilyl)-N6-benzoyladenine in refluxing acetonitrile
(6) Tulshian, D. B.; Czarniecki, M. J. Am. Chem. Soc. 1995, 117, 7009-
7010.
(7) Knapp, S. Chem. ReV. 1995, 95, 1859-1876.
(8) Barton, D. H. R.; Bashiardes, G.; Fourrey, J.-L. Tetrahedron Lett.
1983, 24, 1605-1608.
(10) Beckwith, A. L. J.; O’Shea, D. M. Tetrahedron Lett. 1986, 27,
4525-4528. Stork, G.; Mook, R., Jr. Tetrahedron Lett. 1986, 27, 4529-
4532.
(9) Tobey, S. W. J. Org. Chem. 1969, 34, 1281-1298. Feit, B.-A.; Haag,
B.; Kast, J.; Schmidt, R. R. J. Chem. Soc., Perkin Trans. 1 1986, 2027-
2036.
(11) Vorbru¨ggen, H.; Ruh-Polenz, C. Org. React. 2000, 55, 1-630.
3584
Org. Lett., Vol. 3, No. 22, 2001

these results; the small amount of 9 that formed after 5 h at
137 °C was also destroyed after 12 h, possibly by action of
PhSOH. Separate thermolysis of N6-benzyl-2′,3′,5′-tri-O-
acetyladenosine (which cannot eliminate PhSOH) in reflux-
ing toluene gave some depurinylation after 2 h. The
inescapable conclusion is that depurinylation of these ad-
enosines is at least as fast as sulfoxide elimination.
Scheme 3. N-Glycosylation and Deprotectiona
The depurinylation problem was solved by conducting the
thermolysis in refluxing Ph2O (259 °C) for 2.0 min in the
presence of 5 equiv of N6-benzoyladenine. Presumably the
added purine scavenged PhSOH and allowed the elimination
to compete successfully at the higher temperature.
Attempted deprotection of 14 with 1 M NaOH destroyed
the nucleoside; however, brief treatment with 0.5 M LiOH
cleaved the esters (the N6 is deprotonated12), and subsequent
ammonolysis13 removed the N-benzoyl to afford 1. Its
1
identity was secured by H NMR comparison (D2O, 600
MHz) with the authentic natural product and with a 1:1
admixture.
Acknowledgment. We thank Merck & Co. and SynChem
Research Inc. for financial support, Dr. Toshio Takatsu
(Sankyo Co.) for supplying a sample of 1, Lyudmila
Goncharova for assistance with the crystallography, and
Profs. Samir Z. Zard, Roger A. Jones, and Philip L. Fuchs
for helpful discussions.
a Conditions: (l) bis(trimethylsilyl)-N6-benzoyladenine, Me3SiOTf,
CH3CN, 80 °C; (m) m-CPBA, CH2Cl2, NaHCO3, 0 °C; (n) Ph2O,
N-benzoyladenine, 270 °C bath, 2.0 min; (o) 0.5 M LiOH, EtOH,
1.5 h; (p) concentrated NH4OH, 18 h.
gave the nucleoside 12 apparently uncontaminated with other
stereo- or regioisomers (NMR, TLC analysis). Only regen-
eration of the alkene and final deprotection remained for
conversion to 1.
Oxidation of 12 to the sulfoxides 13 with exactly 1.0 equiv
of m-CPBA was accomplished without incident. However,
the thermal elimination proved to be particularly troublesome.
No elimination of PhSOH occurred in refluxing toluene or
refluxing xylenes (137 °C). At the latter temperature,
significant depurinylation occurred, and 13 was destroyed
over several hours. Control experiments on the sulfoxides
derived from 10-exo (which cannot depurinylate) supported
Supporting Information Available: Experimental details
and spectral characterization for all new compounds, H
NMR comparison for 1, and details of crystallographic
analyses of 8, 9, and 10-endo. This material is available free
of charge via the Internet at http://pubs.acs.org.
1
OL0167210
(12) For discussion of adenosine deprotection, see: Sonveaux, E. In
Methods in Molecular Biology 26. Protocols for Oligonucleotide Conjugates;
Agrawal, S., Ed.; Humana Press: Totawa, NJ, 1994; pp 6-15.
(13) For example, see: Johnson, C. R.; Esker, J. E.; Van Zandt, M. C.
J. Org. Chem. 1994, 59, 5854-5855.
Org. Lett., Vol. 3, No. 22, 2001
3585