Synthesis of 7-Epi (+)-FR900482: An epimer of comparable anti-cancer activity

Article abstract of DOI:10.1021/ol800127a

(Chemical Equation Presented) FR900482 is a potent anti-tumor therapeutic that has been investigated as a replacement candidate for the clinically useful Mitomycin C. Herein, we report synthesis and biological testing of 7-Epi (+)-FR900482, which demonstrates equal potency relative to the natural product against several cancer cell lines. Highlights of this work include utilization of our palladium-catalyzed DYKAT methodology and development of a Polonovski oxidative ring expansion strategy to yield this equipotent epimer in 23 linear steps.

Full text of DOI:10.1021/ol800127a

ORGANIC
LETTERS
2008
Vol. 10, No. 7
1369-1372
Synthesis of 7-Epi (+)-FR900482: An
Epimer of Comparable Anti-Cancer
Activity
Barry M. Trost* and Brendan M. O’Boyle
Department of Chemistry, Stanford UniVersity, Stanford, California 94305-5080
bmtrost@stanford.edu
Received January 18, 2008
ABSTRACT
FR900482 is a potent anti-tumor therapeutic that has been investigated as a replacement candidate for the clinically useful Mitomycin C.
Herein, we report synthesis and biological testing of 7-Epi ( )-FR900482, which demonstrates equal potency relative to the natural product
+
against several cancer cell lines. Highlights of this work include utilization of our palladium-catalyzed DYKAT methodology and development
of a Polonovski oxidative ring expansion strategy to yield this equipotent epimer in 23 linear steps.
FR900482 (1)1 and mitomycin C (3)2 possess structural
similarities that suggest a common biosynthesis. These
compounds also share similar modes of action3,4 and
Scheme 1. General Approach to FR900482 and the
Mitomycins
medicinal properties. Specifically, mitomycin C is admin-
istered to patients suffering from a diverse range of cancers,5
whereas 1 and its triacetate FK973 (2) are replacement
candidates exhibiting superior activity and decreased toxic-
ity.6 We envisioned both natural products to be accessible
via a common synthetic route (Scheme 1). In practice,
preparation of the mitomycin skeleton en route to FR900482
would allow aziridine intermediate 4 to be a common
precursor for the preparation of both natural products.
Naturally occurring mitomycins of both C-9 stereochemistries
are active (mitomycin numbering), but no data regarding the
analogous FR900482 derivative was available prior to our
(1) Uchida, I.; Takase, S.; Kayakiri, H.; Kiyoto, S.; Hashimoto, M. J.
Am. Chem. Soc. 1987, 109, 4108.
(2) Gruschow, S.; Chang, L. C.; Mao, Y.; Sherman, D. H. J. Am. Chem.
Soc. 2007, 129, 6470 and references therein.
(3) Mitomycin, C.; Tomasz, M. Chem. Biol. 1995, 2, 575.
(4) FR900482: Fukuyama, T.; Goto, S. Tetrahedron Lett. 1989, 30, 6491.
See refs 25 and 26 for later studies on the mode of action.
(5) Bradner, W. T. Cancer Treat. ReV. 2001, 27, 35.
(6) Shimomura, K.; Manda, T.; Mukumoto, S.; Masuda, K.; Nakamura,
T.; Mizota, T.; Matsumoto, S.; Nishigaki, F.; Oku, T.; Mori, J.; Shibayama,
work. This report describes our synthesis and biological
F. Cancer Res. 1988, 48, 1166.
testing of 7-epi (+)-FR900482.
10.1021/ol800127a CCC: $40.75
© 2008 American Chemical Society
Published on Web 03/05/2008

FR900482 has been studied extensively, resulting in five
total,7-11 and several formal, syntheses,12-14 the shortest of
which was reported by Williams and requires 29 linear steps
(33 total). We reasoned that uniting a suitable aromatic frag-
ment with aziridine 4 would allow for convergent access to
both series, leaving only changes in oxidation state to com-
plete the synthesis (Scheme 1). Beginning with chemistry
reported in this group,15 we envisioned that aziridine fragment
4 could be assembled in an efficient manner. Finally, this
approach introduces several polar functional groups late in
the synthesis and avoids multiple protecting group manipula-
tions.
Scheme 3. Preparation of the Aziridine Fragment
Preparation of the aryl fragment 8 (Scheme 2) began with
the known triflate 5,12 which is prepared in three steps from
Scheme 2. Preparation of the Aryl Fragment
and Supporting Information). Thus, it appears that double
inversion occurs via neighboring group participation of the
tert-butyl carbamate placing the vinyl group cis to the
aziridine.
The single carbon-carbon bond formation of our synthesis
proceeds via Heck coupling17 to produce the exocyclic olefin
17 and construct the “mitomycin” skeleton of the natural
product. No detectable double bond isomerization is observed
in this step, even though such a process would lead to
aromatization, via formation of an indole.
Inspired by the pioneering works of Dmitrienko18 and
Ziegler,19 we envisioned that a Polonovski reaction, followed
by a subsequent oxidative ring expansion, might afford the
commercially available 5-nitrovanillin as reported previously.
The aryl triflate 5 is converted to an aryl iodide (6) via
nucleophilic aromatic substitution with sodium iodide in
DMF. Oxidation of aldehyde 6 with NIS and K2CO3 in
methanol affords the methyl benzoate 7. This substrate is
subjected to iron mediated reduction yielding aniline 8
without undesirable reduction of the aryl iodide.
(7) Fukuyama, T.; Xu, L.; Goto, S. J. Am. Chem. Soc. 1992, 114, 383.
(8) Schkeryantz, J. M.; Danishefsky, S. J. J. Am. Chem. Soc. 1995, 117,
4722.
The requisite aziridine 4 is prepared in eight steps from
divinyl carbonate 9 (Scheme 3). First, a palladium-catalyzed
DYKAT (dynamic catalytic asymmetric transformation)
reaction15 yields amino-alcohol 10 as a single stereoisomer.
Following TBS protection, diene 11 is submitted to Sharpless
dihydroxylation,16 yielding the diol 12 with excellent chemo-
and diastereoselectivity. The phthalimide group contained
in amino-alcohol 12 is exchanged for a tert-butyl carbamate
yielding diol 13. Selective silylation and mesylation, followed
by exposure of the crude mesylate to cesium carbonate in
warm DMF affords the aziridine 14. Finally, removal of the
TBDPS group and Dess-Martin oxidation yields the com-
pleted aziridine fragment 4.
Reductive amination between the aziridine 4 and aniline
8 affords the coupled product 15 (Scheme 4). Removal of
the allylic TBS group and activation of the resulting alcohol
with triflic anhydride yields the pyrrolidine 16. Simple
inversion at C-8 was initially expected; however, NOE
analysis of intermediate 18 and its epimer reveals strong
correlations between the C-8 and C-9 protons (Scheme 4
(9) (a) Katoh, T.; Itoh, E.; Yoshino, T.; Terashima, S. Tetrahedron 1997,
53, 10229. (b) Yoshino, T.; Nagata, Y.; Itoh, E.; Hashimoto, M.; Katoh,
T.; Terashima, S. Tetrahedron 1997, 53, 10239. (c) Katoh, T.; Nagata, Y.;
Yoshino, T.; Nakatani, S.; Terashima, S. Tetrahedron 1997, 53, 10253.
(10) (a) Judd, T. C.; Williams, R. M. Angew. Chem., Int. Ed. 2002, 41,
4683. (b) Judd, T. C.; Williams, R. M. J. Org. Chem. 2004, 69, 2825.
(11) (a) Suzuki, M.; Kambe, M.; Tokuyama, H.; Fukuyama, T. Angew.
Chem., Int. Ed. 2002, 41, 4686. (b) Suzuki, M.; Kambe, M.; Tokuyama,
H.; Fukuyama, T. J. Org. Chem. 2004, 69, 2831.
(12) Fellows, I. M.; Kaelin, D. E.; Martin, S. F. J. Am. Chem. Soc. 2000,
122, 10781.
(13) Paleo, M. R.; Aurrecoechea, N.; Jung, K. Y.; Rapoport, H. J. Org.
Chem. 2003, 68, 130.
(14) Ducray, R.; Ciufolini, M. A. Angew. Chem., Int. Ed. 2002, 41, 4688.
(15) Trost, B. M.; Aponick, A. J. Am. Chem. Soc. 2006, 128, 3931.
(16) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. ReV.
1994, 94, 2483.
(17) Abelman, M. M.; Overman, L. E.; Tran, V. D. J. Am. Chem. Soc.
1990, 112, 6959.
(18) (a) Dmitrienko, G. I.; Denhart, D.; Mithani, S.; Prasad, G. K. B.;
Taylor, N. J. Tetrahedron Lett. 1992, 33, 5705. (b) Mithani, S.; Drew, D.
M.; Rydberg, E. H.; Taylor, N. J.; Mooibroek, S.; Dmitrienko, G. I. J. Am.
Chem. Soc. 1997, 119, 1159.
(19) (a) Ziegler, F. E. Belema, M. J. Org. Chem. 1994, 59, 7962. (b)
Ziegler, F. E.; Belema, M. J. Org. Chem. 1997, 62, 1083.
1370
Org. Lett., Vol. 10, No. 7, 2008

Scheme 4. Completion of the Synthesis
desired [3.3.1] bicycle.20 To prevent aromatization during
that reduction of the o-quinone methide dominates to yield
Polonovski conditions, the benzylic position is blocked
by first dihydroxylating the exo-olefin and then protecting
the resulting diol as a cyclic carbonate. To our delight,
Polonovski oxidation of 18 affords 19 in good yield.
Exposure of the aminal to a second equivalent of m-CPBA
affords the hydroxylamine hemiketal 20. Here initial forma-
tion of an amine oxide leads to spontaneous ring-opening,
followed by recombination to yield the [3.3.1] bicycle, via
the open eight-membered ketone.21
At this time, our attention turned to the crucial C-7
deoxygenation. Unfortunately, hydrolysis of the carbonate
20 proved difficult due to unexpected sensitivity of the tert-
butyl carbamate toward base. This caused us to explore
sequential hydrogenolysis to remove the C-7 oxygen sub-
stituent. Initial hydrogenolysis of the benzyl ether yields the
phenol 21, which contains a handle for directed reduction
of the carbonate. Subsequent exposure of this phenol to
sodium borohydride in methanol results in removal of the
carbonate, but instead of the anticipated tetraol, a triol is
formed containing a methoxy group at C-7 (not shown).
Speculating that an ortho quinone methide intermediate is
responsible for this transformation, optimization focused on
suppressing the undesired solvolysis to facilitate capture of
the reactive intermediate (Scheme 5). Remarkably, replacing
methanol with ethanol inhibits solvolysis to such an extent
the triol 22.
The connectivity of 22 is confirmed by NMR analysis
(COSY, HMBC, HSQC), but the configuration at C-7
remained uncertain. Completion of the total synthesis is
accomplished in four additional steps. The primary alcohol
of 22 is converted to a primary carbamate via treatment with
trichloroacetyl isocyanate followed by in situ hydroylsis of
the trichloroacetyl group. Exchange of the methyl ester for
the ethyl thioester is followed by Fukuyama22 reduction with
palladium on carbon. Finally, the tert-butyl carbamate is
removed via hydrolysis with saturated aqueous potassium
carbonate in methanol, liberating the free aziridine. The
product is isolated as a 3:2 mixture of separable hemiketal
isomers, however, comparison to an authentic sample (NMR,
TLC) shows that neither isomer matches the natural product,
clearly demonstrating our product to be epimeric at C-7.23
Although we were somewhat disappointed not to have the
natural product in hand, we were quite excited by the
likelihood that this epimer might have comparable activity.
Both Hopkins24 and Williams25 have independently demon-
strated that Fe2+ mediated reductive activation of FR900482
(20) For a related oxidative ring expansion, see: Colandrea, V. J.;
Rajaraman, S.; Jimenez, L. S. Org. Lett. 2003, 5, 785.
(21) The stereochemistry obtained in hemiketal 20 is assigned by
comparison to its epimer, which could be obtained under slightly modified
conditions (see Supporting Information for details).
(22) Fukuyama, T.; Tokuyama, H. Aldrichim. Acta 2004, 37, 87.
(23) This stereochemical result is in contrast to that observed by
Danishefsky,8 who observed formation of an aldehyde leading to the C-7
hydroxymethyl group cis to the hydroxylamine bridge. In our case, this
pathway would be expected to give the natural stereochemistry. Formation
of 7-epi-FR900482 suggests the o-quinone methide is reduced directly by
sodium borohydride without the intermediacy of an aldehyde. Danishefsky
was also able to obtain the opposite stereochemistry at C-7 via a samerium
iodide mediated reduction of a benzylic epoxide. Unfortunately, reproduction
of these condidtions failed on carbonate 20 presumably due to the lower
reactivity for C-O homolytic cleavage of a cyclic carbonate compared to
an epoxide.
Scheme 5. Mechanism for o-Quinone Methide Reduction
Org. Lett., Vol. 10, No. 7, 2008
1371

in the presence of a thiol reducing agent affords the reactive
“mitosene” intermediate capable of cross-linking DNA
(Scheme 6). Obviously, the stereochemistry at C-7 has been
To our delight, comparison of our epimer to the natural
product does demonstrate that the C-7 stereochemistry has
only negligible effect on cytotoxicity against two human
breast cancer cell lines (Table 1).
Scheme 6. Mechanism of Activation
Table 1. Comparitive Cytotoxicity Dataa
IC50 mM
FR900482
7-Epi-FR900482
SKBr3
MCF-7
0.980
1.000
1.020
1.100
a Preparation of the triacetate FK973 is known to increase cytotoxicity
up to 50 fold.6
From a synthetic perspective, the fact that both epimers
have equal activity also means that both are equally valuable
synthetic targets. In this sense, our synthesis which requires
23 linear steps (29 total) and proceeds in 1.4% yield from
commercially available 1,5-hexadiene-3,4-diol is the shortest
route to this family yet disclosed.27 Highlights of this
synthesis include: novel formation of a chiral building block
4 via our DYKAT methodology, a chemo- and diastereo-
selective Sharpless dihydroxylation of 11, effective utilization
of the Polonovski reaction, as well as sequential hydro-
genolysis via a putative o-quinone methide intermediate. As
this synthesis efficiently delivers the mitomycin skeleton
prior to oxidative expansion to the FR900482 series, this
synthesis may also prove amenable to the preparation of
mitomycins or analogues of either class. Efforts to expand
upon this synthesis are underway and will be reported in
due course.
lost and both 7-epi-FR900482 and the natural product must
reach a common intermediate prior to alkylation of DNA.
However, it remained unclear if reductive activation would
function in ViVo to yield the same “mitosene” intermediate,
since, if it is an enzyme mediated process, it should, a priori,
be sensitive to stereochemistry.
Naoe26 has shown a connection between the metallo-
enzyme DT-diaphorase and cytotoxicity for FK317 (Scheme
1). In this case, exposure of cells to dicumarol, an inhibitor
of this enzyme significantly lowers cytotoxicity; furthermore,
FK317 is more active against cells expressing this enzyme.
Given the selectivity of this family for inhibiting the growth
of cancer cells, the reductive activation is expected to be
enzyme dependent, as indiscriminate reduction would lead
to a nonselective DNA alkylating reagent.
Acknowledgment. We are grateful to the NIH (GM
33049) for their generous support of our program. Many
thanks to David Barrett of Astellas for providing authentic
material. We thank Dr. Yong Li and Dr. Hugo Menzella from
Kosan Biosciences for the comparative biological testing of
natural FR900482 against 7-epi FR900482. Mass spectra
were provided by the Mass Spectrometry Regional Center
of the UCSF, supported by the NIH Division of Research
Resources.
Supporting Information Available: Experimental and
spectroscopic data are provided for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(24) (a) Woo, J.; Sigurdsson, S. T.; Hopkins, P. B. J. Am. Chem. Soc.
1993, 115, 1199. (b) Huang, H.; Pratum, T. K.; Hopkins, P. B. J. Am. Chem.
Soc. 1994, 116, 2703. (c) Paz, M. M.; Hopkins, P. B. J. Am. Chem. Soc.
1997, 119, 5999. (d) Paz, M. M.; Hopkins, P. B. Tetrahedron Lett. 1997,
38, 343.
(25) (a) Williams, R. M.; Rajski, S. R. Tetrahedron Lett. 1992, 33, 2929.
(b) Williams, R. M.; Rajski, S. R. Tetrahedron Lett. 1993, 34, 7023. (c)
Williams, R. M.; Ducept, P. Biochemistry 2003, 42, 14696.
(26) Naoe, Y.; Inami, M.; Kawamura, I.; Nishigaki, F.; Tsujimoto, S.;
Matsumoto, S.; Manda, T.; Shimomura, K. Jpn. J. Cancer Res. 1998, 89,
666.
OL800127A
(27) Fukuyama 1st synthesis:7 0.20% yield over 40 linear steps (40 total,
racemic); Danishefsky:8 30 linear steps and 3.3% yield over the final 23
steps (yield of first 7 steps and number of off-line steps are not reported,
racemic); Terashima:9 0.20% yield over 42 linear steps (57 total, +);
Williams:10 0.08% over 29 linear steps (33 total, +); Fukuyama 2nd
synthesis:11 1.7% over 31 linear steps (37 total, +).
1372
Org. Lett., Vol. 10, No. 7, 2008