Regioselective lactonization of naphthoquinones: Synthesis and antitumoral activity of the WS-5995 antibiotics

Article abstract of DOI:10.1016/S0040-4039(00)00329-4

An acid promoted quinolactonization of naphthoquinones has been developed, providing direct access to either ortho or para isomers as desired. Application of this methodology in syntheses of the antibiotics WS- 5995A, WS-5995C and functional analogs is demonstrated. Preliminary antitumoral activity of the analogs is presented together with electrochemical analysis. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00329-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 3007–3010
Regioselective lactonization of naphthoquinones: synthesis and
antitumoral activity of the WS-5995 antibiotics
Ghassan Qabaja,a Elisabeth M. Perchellet,b Jean-Pierre Perchellet b and Graham B. Jones a,∗
aBioorganic and Medicinal Chemistry Laboratories, Department of Chemistry, Northeastern University,
360 Huntington Avenue, Boston, MA, 02115-5000, USA
bAnti-Cancer Drug Laboratory, Kansas State University, Division of Biology, Ackert Hall, Manhattan, KS, 66506-4901, USA
Received 16 January 2000; revised 18 February 2000; accepted 21 February 2000
Abstract
An acid promoted quinolactonization of naphthoquinones has been developed, providing direct access to either
ortho or para isomers as desired. Application of this methodology in syntheses of the antibiotics WS-5995A, WS-
5995C and functional analogs is demonstrated. Preliminary antitumoral activity of the analogs is presented together
with electrochemical analysis. © 2000 Elsevier Science Ltd. All rights reserved.
The search for new antibiotics continues at a steady pace, and quinonoid derived systems have
been the subject of a number of promising leads. A central feature of conventional p-quinonoid based
cytotoxins is their ability to generate semi-quinone radicals 1 following bioreduction, accelerated under
hypoxic conditions.1 Though less studied, the potential for o-quinone isomers to participate in similar
chemistry via 2 also exists, providing the reduction potential lies within the window of opportunity
for cellular activation. As part of our program directed towards free radical based chemotherapies,2 we
became interested in scrutinizing the differential cytotoxicities of quinone isomers. We opted to study
the issue by preparing synthetic analogs of natural products, having the additional benefit of potentially
accessing deselected natural products, a key feature of current trends in chemical genetics.3 Though
several pyranonaphthoquinone antibiotics have been identified, including nanaomycin A, kalafungin
and eleutherin, relatively few members of the corresponding naphthoquinone lactones have been found.
∗
Corresponding author.
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00329-4
tetl 16609

3008
The structure of one member of this class, WS-5995A, isolated from Streptomyces auranticolor was
determined in 1983,4 which led to the subsequent discovery of the structural analogs including its open
form WS-5995C.5 These agents have demonstrated chemoprotective activity against Eimeria tenella
infection.6 Though the origin of their activity remains to be clarified, in an unrelated study it was shown
that the bacterium Streptomyces acidiscabies, present in acidic soils, produces these compounds, and
that they likely contribute to the observed pathogenicicty of this organism.7 Our interest in this family
was strengthened by their structural similarities with the gilvocarcin class of antitumor antibiotics, which
participate in DNA intercalation, and may function as topoisomerase inhibitors.8 We opted to develop an
expeditious route to the core structure of the WS-5995 family, focusing on novel lactonization chemistry.
Accordingly, an appropriate model compound 3 was assembled to test the hypothesis.9 A variety of
acid catalyzed quinolactonizations were attempted using conventional methods, which would be expected
to yield 4 via the mixed anhydride. Though unsuccessful with acetic anhydride, TFAA gave the product
cleanly and in high yield (Scheme 1). More significantly, under the influence of Lewis acids, the alternate
ortho quinone product 5 could be produced. This is presumably the result of enolization of the p-quinone
carbonyl, encouraged by chelation of the phenol with the Lewis acid. Following attempts to optimize the
yield of this isomer, TMS triflate was found to be most effective, giving good yields of the o-isomer,
together with traces of the p-isomer. As expected, extended exposure of 5 to various acids resulted
in isomerization to the more thermodynamically stable 4, methanesulfonic acid proving most efficient
(−78°C, CH2Cl2, 2 h, 95%).
Scheme 1. Quinolactonization of benzoquinone model substrate
Encouraged by the regiocontrol observed in the model study, we sought to incorporate this metho-
dology in the synthesis of natural and unnatural isomers of the WS-5995 family. Accordingly, substituted
substrate 10 was prepared; methyl juglone 6 was reduced and esterified with acyl chloride 7, prepared
from dimethylanisole acid in five steps (Scheme 2).10 The resulting iodo ester 8 was methylated,
subjected to Pd catalyzed annulation, to give tetracycle 9, then the quinone revealed to give 10. As was
found in the model series, exposure of the hydroxy acid to methane sulfonic acid gave the p-quinone 11,
however, in stark contrast to the model study, the alternate o-quinone 12 was produced exclusively using
TFAA (Scheme 3). Interconversion of this isomer to 11 was effected either using MSA or TMSOTf,
but extended exposure of 12 with TFAA did not result in isomerization. The origins of this apparent
stabilizing effect are presumably related to the pendant methoxy functionality, and are the subject of an
ongoing investigation. Regioselective demethylation of either 12 or 11 could be effected using lithium
iodide, in the case of 11 giving the natural product WS-5995A, and in the case of 12, the interesting
analog 13. Hydrolysis of the lactone in 13 resulted in concomitant formation of the natural product WS-
5995C, which itself can be converted to WS-5995A using established methods.5
The described route allowed us to now scrutinize the redox chemistry of this intriguing class of
natural products, and additionally explore the relative contribution to cytotoxicity this imparts. The
electrochemical reduction potential of the quinone isomers was determined, and as expected marked
differences exist, which could be expected to have an impact under biological conditions (Table 1).11
Initial cytotoxicity assays against a murine cell line reveal that in addition to WS-5995, ortho quinone 12

3009
Scheme 2. Preparation of functionalized hydroxyquinone
Scheme 3. Synthesis of WS-5995 antibiotics and quinone isomers
and model quinone 4 are appreciably cytotoxic (Table 1). Additionally, comparison of 12 and 13 suggests
that the hydrogen bonding phenolic group may attenuate the function of the o-quinone moiety.
It is well known that the differing reduction potentials of o- and p-quinones can play a dramatic role in
electron transfer chemistry, and we are now investigating the antiproliferative capacity of the entire family
of analogs under a variety of hypoxic and aerobic screens. Also of interest is the structural similarity of
this class of agents to the ellagic acids, whose antiproliferative and chemoprotective ability has been
noted.12 Since ellagic acid had an IC50 of 10 µM in the current assay, additional efforts to determine the
molecular mechanism of action of these agents is also underway.
In summary, substrate dependent regioselective acid catalyzed quinolactonization procedures have
been developed and applied to the total synthesis of WS-5995A and C, and their quinone isomers. In-
depth studies on the antitumoral and chemoprotective potential of these compounds, together with the
scope of the quinolactonization procedure itself, will be reported in due course.

3010
Table 1
Cytotoxicity of o/p quinones#
Acknowledgements
This work was supported by grants from the National Institute of General Medical Sciences
(RO1GM57123), The Petroleum Research Fund (33920-AC1) and The American Cancer Society (to
G.B.J.), and BioServe Space Technologies (NASA#NAGW-1197) and Kansas State University (to J.-
P.P.).
References
1. Brown, J. M.; Koong, A. J. Natl. Cancer Inst. 1991, 83, 178.
2. Jones, G. B.; Huber, R. S. Tetrahedron Lett. 1994, 35, 2655; Jones, G. B.; Huber, R. S.; Mathews, J. E. J. Chem. Soc., Chem.
Commun. 1995, 1791; Jones, G. B.; Huber, R. S.; Mathews, J. E.; Li, A. Tetrahedron Lett. 1996, 37, 3643; Jones, G. B.;
Kilgore, M. W.; Pollenz, R. S.; Li, A.; Mathews, J. E.; Wright, J. M.; Huber, R. S.; Tate, P. L.; Price, T. L.; Sticca, R. P.
Bioorg. Med. Chem. Lett. 1996, 6, 1971; Jones, G. B.; Mathews, J. E. Bioorg. Med. Chem. Lett. 1997, 7, 745; Jones, G.
B.; Mathews, J. E. Tetrahedron 1997, 53, 14599; Jones, G. B.; Wright, J. M.; Rush, T. M.; Plourde, G. W.; Kelton, T. F.;
Mathews, J. E.; Huber, R. S.; Davidson, J. P. J. Org. Chem. 1997, 62, 9379; Hynd, G.; Jones, G. B.; Plourde, G. W.; Wright,
J. M. Tetrahedron Lett. 1999, 40, 4481; Jones, G. B.; Wright, J. M. Tetrahedron Lett. 1999, 40, 7605; Xi, Z.; Jones, G. B.;
Qabaja, G.; Wright, J. M.; Johnson, F.; Goldberg, I. H. Org. Lett. 1999, 1, 1375.
3. Schreiber, S. L. Bioorg. Med. Chem. 1998, 6, 1127.
4. Ikushima, H.; Takase, S.; Kawai, Y.; Itoh, Y.; Okamoto, M.; Tanaka, H.; Imanaka, H. Agric. Biol. Chem. 1983, 47, 2231.
5. Syntheses of WS-5995 antibiotics: Watanabe, M.; Date, M.; Furukawa, S. Chem. Pharm. Bull. 1989, 37, 292; de Frutos, O.;
Echavarren, A. M. Tetrahedron Lett. 1996, 37, 8953; Tamayo, N.; Echavarren, A. M.; Paredes, M. C. J. Org. Chem. 1991,
56, 6488; Echavarren, A. M.; Tamayo, N.; Cardenas, D. J. J. Org. Chem. 1994, 59, 6075; McKenzie, T. C.; Choi, W.-B.
Synth. Commun. 1989, 19, 1523.
6. Ikushima, H.; Okamoto, M.; Tanaka, H.; Ohe, O.; Kohsaka, M.; Aoki, H.; Imanaka, H. J. Antibiot. 1980, 33, 1107 and
references cited therein.
7. King, R. R.; Lawrence, C. H. J. Agric. Food Chem. 1996, 44, 2849.
8. Matson, J. A.; Rose, W. C.; Bush, J. A.; Myllymaki, R.; Bradner, W. T.; Doyle, T. W. J. Antibiot. 1989, 42, 1446.
9. Commercially available methoxynaphthol, was coupled with the acyl chloride derived from 2-iodobenzoic acid, giving the
corresponding ester (90%), followed by Pd-mediated cyclization (PdCl2(PPh3)2, NaOAc, DMA, 65%) and hydrolysis (KOH,
THF 70%). All new compounds gave satisfactory spectroscopic and analytical data.
10. The desired acid was prepared from 3,5 dimethylanisole via (i) bromination (NBS, CCl4) followed by (ii) in situ hydrolysis
(K2CO3, dioxane, 40% over two steps) (iii) iodination (2 nBuLi, I2, −78°C, 67%), and (iv) oxidation (Bu4NMnO4, Py, 95%).
11. We thank Sivashankar G. Sivakolundu for these determinations.
12. Perchellet, J. P.; Gali, H. U.; Perchellet, E. M.; Klish, D. S.; Armbrust, A. D. In Plant Polyphenols: Synthesis, Properties,
Significance; Hemingway, R. W.; Laks, P. E., Eds. Plenum Press: New York, 1992; pp. 783–801.