Synthesis of (+)-nocardione A - Use of formal radical cyclization onto a benzene ring

Article abstract of DOI:10.1039/b307937f

Juglone (7) was converted into enone 13; this underwent radical cyclization to afford 15, which was aromatized to 16 and elaborated into (+)-nocardione A (1), the enantiomer of the naturally-occurring tyrosine phosphatase inhibitor (-)-nocardione A (2).

Full text of DOI:10.1039/b307937f

Synthesis of (+)-nocardione A — use of formal radical cyclization onto a
benzene ring
Derrick L. J. Clive* and Stephen P. Fletcher
Chemistry Department, University of Alberta, Edmonton, Alberta, Canada T6G 2G2.
E-mail: derrick.clive@ualberta.ca
Received (in Cambridge, UK) 11th July 2003, Accepted 8th August 2003
First published as an Advance Article on the web 21st August 2003
Juglone (7) was converted into enone 13; this underwent
radical cyclization to afford 15, which was aromatized to 16
and elaborated into (+)-nocardione A (1), the enantiomer of
the naturally-occurring tyrosine phosphatase inhibitor
(2)-nocardione A (2).
following conditions. First, ethyl (2)-lactate6 was converted
into its trifluoromethanesulfonate by the literature method,7 and
a solution of this derivative (4 equiv.) was added to a stirred and
cooled (278 °C) solution of 9 in the presence of Cs2CO3.8 After
an 18-hour reaction period, 10 could be isolated in 50% yield
(88% after correction for recovered 9). Under these conditions,
stereochemical inversion of the trifluoromethanesulfonate oc-
curs and the alkylation is regioselective (9 ? 10). Ester
reduction (LiAlH4, 100%) gave alcohol 11 (ee > 95% by 19F
NMR on the derived Mosher ester) and replacement of the
hydroxyl by iodine (Ph3P, imidazole, I2, 89%) took the route as
far as phenolic ether 12. When this was oxidized with DDQ
(Scheme 3) in the presence of MeOH and K2CO3, the required
enone 13 was formed in good yield (87%) as a 1.1 : 1 mixture
of diastereoisomers. Radical cyclization under standard condi-
tions4 (Bu3SnH, AIBN, PhMe, 85 °C) led to the diaster-
eoisomeric furans 15 (82%) by way of the intermediate radicals
14. We found no evidence for cyclization of 14 through oxygen
onto the allyl pendant. Aromatization (15 ? 16) was easily
achieved, either by storage in CDCl3 (12 h, 97%) or by
treatment with AcOH in CHCl3 (88%); under these conditions
7-exo cyclization of the phenolic hydroxyl onto the allyl group
was not observed. With the phenolic dihydrofuran 16 in hand,
the o-quinone system was readily generated (96%, 82% after
crystallization from CHCl3–hexane) by the action of
[PhSe(O)]2O9 and, finally, palladium-mediated removal of the
allyl protecting group afforded (+)-nocardione A, which
crystallized from EtOAc–hexane as bright red needles (74%
yield from 17, 22% overall from juglone), mp 168–169 °C,10
[Lit. 115–120 °C;1 172.5–173.5 °C2]. Examination of the
compound by HPLC on a chiral column11 showed the material
to have an ee of 98.5%12
We report the synthesis of (+)-1, the enantiomer of the Cdc25B
tyrosine phosphatase inhibitor (2)-nocardione A (2).1,2 The
latter was isolated from the fermentation broth of a soil
microorganism tentatively identified as the Gram-positive
bacterium Nocardia sp. TP-A0248, and is of interest because
Cdc25B is a key enzyme in cell cycle regulation and is
overexpressed in several types of human cancer cells.3 No-
cardione A (2) shows moderate antifungal and cytotoxic
activity,1 and causes cell death with characteristics of apoptosis
in U937 human myeloid leukemia cell lines.1
Our synthetic route is based on a method for achieving formal
cyclization of a radical onto a benzene ring, along the lines
summarized in Scheme 1. The overall sequence 3 ? 6 is
general;4 our model studies4 have shown that it is an efficient
route to simple benzo-fused dihydrofurans, and its application
to the synthesis of 1 or 2 promised to be a significantly more
demanding test of the methodology. The prior synthesis2,5 of
nocardione A (2), revealed that the compound is rather sensitive
and its construction challenging; deprotection of the corre-
sponding O-methyl ether (a congener called nocardione B) was
accompanied by racemization, and even hydrogenolysis of the
corresponding O-benzyl ether was inefficient. Formation of the
dihydrofuran segment, generated by Mitsunobu displacement,
was also difficult. We chose to make the unnatural enantiomer
1 because ethyl (2)-lactate, a chiral building block used in our
approach, is much cheaper than the R-enantiomer.
We thank the Natural Sciences and Engineering Research
Council of Canada for financial support, and Dr X. Lu and H.
Application of the general method of Scheme 1 to the
particular case of (+)-nocardione A requires the substituted
naphthalenol 12 (see Scheme 2) as a key intermediate, and this
compound was prepared as shown. Juglone (7) was converted
(allyl bromide, Ag2O, 79%) into allyl ether 8 (Scheme 2), and
reduction with aqueous Na2S2O4 (100%) then gave bis-phenol
9. At this point, introduction of the chiral two-carbon chain
needed for the radical cyclization could be achieved under the
Scheme 2 Reagents and conditions: (i) allyl bromide, Ag2O, CH2Cl2,
reflux, 11 h, 79% ; (ii) Na2S2O4, ether–water, 40 min, 100%; (iii) triflate of
ethyl (2)-lactate, Cs2CO3, CH2Cl2, 278 °C, 18 h, 50%, 88% corrected for
recovered 9; (iv) LiAlH4, THF, 10 min, 100%; (v) Ph3P, imidazole, I2, THF,
12 h, 89%.
Scheme 1
2464
CHEM. COMMUN., 2003, 2464–2465
This journal is © The Royal Society of Chemistry 2003

2 Synthesis: Y. Tanada and K. Mori, Eur. J. Org. Chem, 2001,
4313–4319.
3 (a) D. Gasparotto, R. Maestro, S. Piccinin, T. Vukosavljevic, L. Barzan,
S. Sulfaro and M. Boiocchi, Cancer Res., 1971, 57, 2366–2368; (b) K.
Galaktionov, A. K. Lee, J. Eckstein, G. Draetta, J. Meckler, M. Loda and
D. Beach, Science, 1995, 269, 1575–1577; (c) W. G. Dunphy and A.
Kumagai, Cell, 1991, 67, 189–196.
4 D. L. J. Clive, S. P. Fletcher and M. Zhao, J. Chem. Soc., Chem.
Commun., 2003, 526–527.
5 An o-quinone, structurally related to nocardione A, was produced as a
minor byproduct (5–10%) during oxidation of an o-naphthoquinone: C.
A. Merlic, C. C. Aldrich, J. Albaneze-Walker, A. Saghatelian and J.
Mammen, J. Org. Chem., 2001, 66, 1297–1309.
6 Commercial samples of ethyl (2)-lactate are probably ca. 97% ee, see:
L. E. Overman, K. L. Bell and F. Ito, J. Am. Chem. Soc., 1984, 106,
4192–4201and reference 39 therein.
7 (a) F. Effenberger, U. Burkard and J. Willfahrt, Liebigs Ann. Chem.,
1986, 314–333; (b) F. Effenberger, U. Burkard and J. Willfahrt, Angew.
Chem., Int. Ed. Engl., 1983, 95, 65–66.
8 Cf. W. H. Kruizinga, B. Strijtveen and R. M. Kellogg, J. Org. Chem.,
1981, 46, 4321–4323.
Scheme 3 Reagents and conditions: (i) DDQ, MeOH, K2CO3, 4 min, 87%;
(ii) Bu3SnH and AIBN (addition over 9 h), heat 6 h more, PhMe, 85 °C,
82%; (iii) AcOH, CHCl3, 2 h, 88%; (iv) [PhSe(O)]2O, THF, 12 min, 96%,
82% after recrystallization; (v) 10 mol% Pd(PPh3)4, dimedone, THF, 4 min,
74%.
9 D. H. R. Barton, A. G. Brewster, S. V. Ley, C. M. Read and M. N.
Rosenfeld, J. Chem. Soc., Perkin Trans. 1, 1981, 1473–1476.
10 There is a form change to needles at, or below, 158 °C if the original
material is a powder, as obtained by evaporation of a solution.
11 Chiralcel AD-RH column (0.46 cm 3 15 cm); eluant 2 : 1 water–MeCN;
flow rate 0.6 mL min21; detection at 230 nm, temperature 25 °C; sample
concentration and injection value 1 mL mg21 in MeCN 3 0.5 mL. Under
these conditions there is baseline separation between the nocardione A
enantiomers.
Lachance for HPLC measurements. SPF holds an NSERC
Postgraduate Scholarship.
12 Our [a]D values are unreliable; the high extinction coefficient of the
compound prevents adequate light transmission in our instrument:
found for 1, [a]2D4 +36.8, c = 1.0 CHCl3, 10 cm cell; [a]2D4 +49.5, c =
0.97 CHCl3, 1 cm cell. Lit. for 2, [a]2D1 256.0, c = 0.97 CHCl3 (ref. 2)
[a]2D6 285.4, c = 1.0 CHCl3 (ref. 1).
Notes and references
1 Isolation and biological properties: T. Otani, Y. Sugimoto, Y. Aoyagi,
Y. Igarashi, T. Furumai, N. Saito, Y. Yamada, T. Asao and T. Oki, J.
Antiobiot., 2000, 53, 337–344.
CHEM. COMMUN., 2003, 2464–2465
2465