Investigation of a convergent route to purpuromycin: Benzofuran formation vs spiroketalization

Article abstract of DOI:10.1021/ol061112j

A mild and efficient [3+2] nitrile oxide/olefin cycloaddition allows coupling of the highly functionalized naphthalene and isocoumarin hemispheres of purpuromycin. A rationale of the inability of advanced keto alcohols to spirocyclize is presented based u

Full text of DOI:10.1021/ol061112j

ORGANIC
LETTERS
2006
Vol. 8, No. 15
3243-3246
Investigation of a Convergent Route to
Purpuromycin: Benzofuran Formation
vs Spiroketalization
Stephen P. Waters, Michael W. Fennie, and Marisa C. Kozlowski*
Department of Chemistry, Roy and Diana Vagelos Laboratories, UniVersity of
PennsylVania, Philadelphia, PennsylVania 19104-6323
marisa@sas.upenn.edu
Received May 5, 2006
ABSTRACT
A mild and efficient [3+2] nitrile oxide/olefin cycloaddition allows coupling of the highly functionalized naphthalene and isocoumarin hemispheres
of purpuromycin. A rationale of the inability of advanced keto alcohols to spirocyclize is presented based upon a systematic examination of
the electronic factors present in these systems and suggests that the biosynthesis of purpuromycin does not proceed through open-chain
intermediates.
Purpuromycin (1),1 a member of the rubromycin family of
natural products that also includes heliquinomycin and the
griseorhodins,2-4 is a polyketide consisting of highly func-
tionalized naphthazarin and isocoumarin ring systems linked
through a bisbenzannelated 5,6-spiroketal (above). Our
particular interest in purpuromycin, isolated from the soil
bacterium A. ianthinogenes, followed reports of its potent
antimicrobial5 and human telomerase inhibitory properties.6
Moreover, its unique ability to bind with high affinity to all
tRNAs, thereby inhibiting their acceptor capacity and
disrupting further protein synthesis, represents a novel mode
of action.7
While a formidable array of oxidation is present within
the hexacyclic ring system, the most striking feature is the
highly unusual 5,6-bisbenzannelated spiroketal core. Indeed,
the only reported synthesis of a natural product featuring
this moiety is that of heliquinomycin aglycon by Danishefsky
and co-workers, which was realized through coupling of a
lithiated naphthofuran with an arylacetaldehyde and Mit-
sunobu-like ring closure to form the spiroketal.8 In addition,
syntheses of simple 5,6-bisbenzannulated spiroketals have
been disclosed by both de Koning,9 and Brimble.10
(1) (a) Coronelli, C.; Pagani, H.; Bardone, M. R.; Lancini, G. C. J.
Antibiot. 1974, 27, 161. (b) Bardone, M. R.; Martinelli, E.; Zerilli, E. F.;
Coronelli, C. Tetrahedron 1974, 30, 2747.
(2) Brockmann, H.; Zeeck, A. Chem. Ber. 1970, 103, 1709.
(3) (a) Chino, M.; Nishikawa, K.; Umekita, M.; Hayashi, C.; Yamazaki,
T.; Tsuchida, T.; Sawa, R.; Hamada M.; Takeuchi, T. J. Antibiot. 1996,
49, 752. (b) Chino, M.; Nishikawa, K.; Tsuchida, T.; Sawa, R.; Nakamura,
H.; Nakamura, K. T.; Muraoka, M. Y.; Ikeda, D.; Naganawa, H.; Sawa,
T.; Takeuchi, T. J. Antibiot. 1997, 50, 143.
(4) (a) Yang, J.; Fan, S.; Pei, H.; Zhu, B.; Xu, W.; Naganawa, H.;
Hamada, M.; Takeuchi, T. J. Antibiot. 1991, 44, 1277. (b) Panzone, G.;
Trani, A.; Ferrari, P.; Gastaldo, L.; Colombo, L. J. Antibiot. 1997, 50, 665.
(5) (a) Trani, A.; Dallanoce, C.; Ferarri, P.; Goldstein, B. P.; Ripamonti,
F.; Ciabatti, R. Il Farmaco 1996, 51, 503. (b) Trani, A.; Kettenring, J.;
Ripamonti, F.; Goldstein, B.; Ciabatti, R. Il Farmaco 1993, 48, 637.
(6) Ueno, T.; Takahashi, H.; Oda, M.; Mizunuma, M.; Yokoyama, A.;
Goto, Y.; Mizushina, Y.; Sakaguchi, K.; Hayashi, H. Biochemistry 2000,
39, 5995.
(7) Krillov, S.; Vitali, L. A.; Goldstein, B. P.; Monti, F.; Semenkov, Y.;
Makhno, V.; Ripa, S.; Pon, C. L.; Gualerzi, C. O. RNA 1997, 3, 905.
(8) (a) Qin, D.; Ren, R. X.; Siu, T.; Zheng, C.; Danishefsky, S. J. Angew.
Chem., Int. Ed. 2001, 40, 4709. (b) Siu, T.; Qin, D.; Danishefsky, S. J.
Angew. Chem., Int. Ed. 2001, 40, 4713.
(9) Capecchi, T.; de Koning, C. B.; Michael, J. P. J. Chem. Soc., Perkin
Trans. 1 2000, 2681.
(10) Tsang, K. T.; Brimble, M. A.; Bremner, J. B. Org. Lett. 2003, 5,
4425.
10.1021/ol061112j CCC: $33.50
© 2006 American Chemical Society
Published on Web 06/30/2006

Our lab11 has recently reported a convergent and mild
coupling strategy involving the [3+2] cycloaddition of nitrile
oxides and olefins to afford isoxazolines, which are converted
to 5,6-bisbenzannulated spiroketals in three steps. Herein we
report the application of that strategy to the requisite
naphthalene and isocoumarin coupling partners (Scheme 1).
Scheme 3. Synthesis of Naphthalene Fragment
Scheme 1. Retrosynthesis of Purpuromycin
naphthalene 11. Methylation, ester reduction, and silylation
afforded naphthalene 12. Selective oxidation with DDQ at
the more electron-rich benzylic carbon14 afforded an inter-
mediate aldehyde, which was subjected to Henry condensa-
tion and reduction to provide nitroalkane 13. Removal of
the remaining silyl group, oxidation to the aldehyde, Dakin
oxidation, and protection of the resulting phenol provided
coupling partner 14.
The isocoumarin coupling partner was synthesized as
shown in Scheme 2. Oxidation of the primary alcohol in
Attention turned to utilization of the [3+2] cycloaddition
method for the complex fragment coupling. Fully function-
alized isocoumarin 8 and naphthalene 14 underwent facile
cycloaddition to afford isoxazoline 15 in good yield. Hy-
drolysis of the TES group in acidic methanol, reduction of
the isoxazoline with Raney Ni, and hydrogenolysis of the
benzyl ethers efficiently provided keto alcohol 18 in three
steps (Scheme 4).
Scheme 2. Synthesis of Isocoumarin Fragment
Scheme 4. Coupling of the Purpuromycin Fragments
previously reported 612 with Dess-Martin periodinane,
followed by protecting group interchange provided aldehyde
7. Since attempts to homologate aldehyde 7 to styrene 8
failed with conventional Wittig reagents, a two-step protocol
was implemented. Accordingly, dimethylzinc addition fol-
lowed by TsOH mediated elimination afforded the desired
alkene coupling partner.
For the naphthalene portion (Scheme 3), synthesis started
with a [4+2] cycloaddition between furan 9 and dimethyl
acetylene dicarboxylate (DMAD). Methylation and double
Claisen condensation13 with dimethyl succinate provided
(11) Waters, S. P.; Fennie, M. W.; Kozlowski, M. C. Tetrahedron Lett.
2006, 47, 5409-5413.
(12) Waters, S. P.; Kozlowski, M. C. Tetrahedron Lett. 2001, 42, 3567.
For another isocoumarin synthesis see also: Thrash, T. P.; Welton, T. D.;
Behar, V. Tetrahedron Lett. 2000, 41, 29.
(13) Kelly, T. R.; Bell, S. H.; Ohashi, N.; Armstrong-Chong, R. J. J.
Am. Chem. Soc. 1988, 110, 6471.
3244
Org. Lett., Vol. 8, No. 15, 2006

Despite an extensive body of precedent,11,15 all attempts
to effect acid-induced spirocyclization of 18 were unsuc-
cessful (Scheme 5). Reasoning that undesirable transforma-
Scheme 8. Benzofuran vs Spirocyclization
Scheme 5. Attempted Spirocyclization of Naphthalene 18
tions were occurring at the sensitive electron-rich naphthalene
under the conditions required for spiroketalization, the
corresponding naphthoquinone keto alcohol 20 was also
synthesized from isoxazoline 17 (Scheme 6). Again, spiroket-
Scheme 6. Attempted Spirocyclization of Naphthoquinone
However, the keto alcohol from isocoumarin fragment 8
and model nitroalkane 2511 did not form a spiroketal upon
treatment with catalytic acid; instead, only the aromatic
benzofuran 27 was obtained (Scheme 8). These findings
suggest that the nucleophilicity of each phenolic participant
is crucial. The inductive effects of the two carbonyl substit-
uents in 8 diminish the nucleophilicity of the relevant phenol
such that irreversible formation of the stable aromatic
benzofuran becomes the dominant pathway (Scheme 9, path
alization conditions gave rise to complex reaction mixtures.
Clearly, a more precise understanding of the factors
inhibiting spirocyclization was imperative. The [3+2] cou-
pling strategy was ideal for this purpose as different left-
hand and right-hand fragments could be examined in a
combinatorial fashion (Schemes 7 and 8). Following the
Scheme 9. Spirocyclization and Elimination Pathways
Scheme 7. Deconvoluting the Spirocyclization Factors
b vs path a).16 With catechol ester 28 (Scheme 8), formation
of a similar benzofuran 30 was observed.17 From this result,
even the presence of one electron-withdrawing group pre-
disposes the system toward benzofuran formation (Scheme
9, path b).
Due to such facile formation, the conversion of benzo-
furans into the corresponding spiroketals is appealing.
protocol outlined above, 14 and model styrene 2211 thus
provided a keto alcohol which, upon treatment with catalytic
acid, underwent smooth spirocyclization to 24a and 24b (1:1
mixture of diastereomers).
(14) Xie, X.; Kozlowski, M. C. Org. Lett. 2001, 3, 2661.
(15) Perron, F.; Albizati, K. F. Chem. ReV. 1989, 89, 1617.
(16) For a discussion, see: Liptak, M. D.; Gross, K. C.; Seybold, P. G.;
Feldgus, S.; Shields, G. C. J. Am. Chem. Soc. 2002, 124, 6421.
(17) Raney Ni also removed the iodide.
Org. Lett., Vol. 8, No. 15, 2006
3245

Degradation studies of â-rubromycin (35) have shown that
furan 36 is obtained from refluxing pyridine (Scheme 10).1b
Scheme 11. Previously Proposed Biosynthesis Involving
Spirocyclization from an Open Chain Intermediate
Scheme 10. Degradation of â-Rubromycin
multiple bisbenzannulated spiroketal precursors in the pres-
ence of nucleophile and electrophile sensitive functionality.
This approach permits the generation of multiple analogues
to explore the bioactivity of the 5,6-bisbenzannulated
spiroketals. The electronic factors required for spirocycliza-
tion of the rubromycins were identified through a systematic
analysis of the individual fragments. These results have
important implications to any synthesis of this class of
compounds.19 Further synthetic efforts toward this family
of natural products and exploration of alternative biomimetic
sequences are underway.
However, we found that this electron-deficient furan fails
to spirocyclize to the natural product under a variety of
conditions. These results complement the finding of Dan-
ishefsky and co-workers that an electron-rich naphthofuran
intermediate in the heliquinomycin aglycon synthesis was
unreactive.8
Recent studies by Piel, who mapped the gene cluster for
griseorhodin biosynthesis and found no less than 11 distinct
redox-tailoring events along the pathway, suggest that the
spiroketal cores are the product of a series of oxidative bond
fragmentations from a common all-carbon precursor (37-
40, Scheme 11).18 From our studies, however, it appears that
any electron-withdrawing group on the eastern hemisphere
complicates conventional spirocyclization. This suggests that
spirocyclization from an open-chain ketone precursor (18,
20) may not play any direct role in the biosynthesis of
purpuromycin. On the whole, these findings point to an as
yet undefined biogenic pathway.
Acknowledgment. Support for this work was provided
by the American Cancer Society (RSG-02-137-01-CDD). We
thank Dr. Rakeshwar Bandichhor for the synthesis of
precursors. â-Rubromycin was from Galileaus.
Supporting Information Available: Experimental pro-
cedures and characterization data. This material is available
free of charge via the Internet at http://pubs.acs.org.
We have demonstrated that the [3+2] cycloaddition is a
mild and effective means for rapid, convergent assembly of
OL061112J
(19) For example, see the following recent report: Lindsey, C. C.; Wu,
K. L.; Pettus, T. R. R. Org. Lett. 2006, 8, 2365.
(18) Li, A.; Piel, J. Chem. Biol. 2002, 9, 1017.
3246
Org. Lett., Vol. 8, No. 15, 2006