Synthesis and biological evaluation of novel fumagillin and ovalicin analogues

Article abstract of DOI:10.1039/b503163j

A promising approach among the numerous efforts to cure cancer is the interruption of the tumour-induced formation of new blood vessels (angiogenesis). By suppressing angiogenesis with drugs, the tumour can neither grow to a life threatening size, nor metastasize. The natural product fumagillin 1 and the structurally related ovalicin 2 are two of the most potent anti-angiogenic compounds. Here, we report the design and synthesis of novel fumagillin and ovalicin analogues lacking reactive epoxy functionalities, which were thought to be responsible for the severe toxic side-effects observed. We also report a new synthetic approach and the determination of the anti-angiogenic properties of these compounds in endothelial cells. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b503163j

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Synthesis and biological evaluation of novel fumagillin and ovalicin
analogues†
Ralph Mazitschek,*a Axel Huweb and Athanassios Giannis*b
a The Broad Institute, Harvard University and Massachusetts Institute of Technology,
Cambridge, Massachusetts, 02141, U. S. A.. E-mail: ralph_mazitschek@hms.harvard.edu;
Fax: +1 617 324 9601; Tel: +1 617 324 9622
b Institut fu¨r Organische Chemie, Universita¨t Leipzig, Johannisallee 29, D-04103, Leipzig,
Germany. E-mail: giannis@chemie.uni-leipzig.de; Fax: +49 (0)341 97 36 599;
Tel: +49 (0) 341 97 36 581
Received 3rd March 2005, Accepted 13th April 2005
First published as an Advance Article on the web 22nd April 2005
A promising approach among the numerous efforts to cure cancer is the interruption of the tumour-induced
formation of new blood vessels (angiogenesis). By suppressing angiogenesis with drugs, the tumour can neither grow
to a life threatening size, nor metastasize. The natural product fumagillin 1 and the structurally related ovalicin 2 are
two of the most potent anti-angiogenic compounds. Here, we report the design and synthesis of novel fumagillin and
ovalicin analogues lacking reactive epoxy functionalities, which were thought to be responsible for the severe toxic
side-effects observed. We also report a new synthetic approach and the determination of the anti-angiogenic
properties of these compounds in endothelial cells.
suggest that MetAP2 function is not required for endothelial
Introduction
cell proliferation. As we have shown recently, fumagillin inhibits
Claiming more than 2.5 million new victims every year, cancer
the expression of the transcription factor Ets-1, which is an
is one of the leading causes of death among the Western
essential positive regulator of angiogenesis.16
population.1 Consequently, there are numerous efforts to de-
velop effective therapies for treating this complex disease. One
promising approach is the interruption of the tumour induced
angiogenesis, the formation of new blood vessels. This strategy,
TNP-470 3 (Scheme 1), a semisynthetic derivative of fu-
magillin, reduces the proliferation of endothelial cells with an
IC50 value of 2.5 × 10−11 M, and has already entered several
clinical studies for the treatment of solid tumours including
which was initially proposed by J. Folkman, targets the nutrient
Kaposi sarcoma, prostate, kidney and cervix carcinomas.7,17–21
and oxygen supply of the tumour.2 Without initiating and
Apart from a very short physiological half-life, TNP-470 causes
maintaining a connection to the vascular system, a tumour
severe side effects such as ataxia, vertigo, drowsiness and
can not grow beyond the size of a few cubic millimetres. By
agitation.18,22–24 The side effects can be attributed to the reactive
suppressing angiogenesis with drugs, the tumour can neither
functional groups. Therefore, extensive efforts are being made
grow to a life threatening size, nor metastasize. In addition to
to develop new derivatives, especially those that lack the reactive
the application in oncology, anti-angiogenic drugs are in demand
for various diseases that are associated with pathological
epoxides as, e.g., in the recently synthesised Fumagalone.25
The present investigation and synthesis of fumagillin and
angiogenesis. These include rheumatoid arthritis and diabetic
ovalicin analogues is mostly based on semisynthetic strategies,
retinopathy, the major cause of blindness in Europe.3–5
The natural product fumagillin 1 and the structurally related
ovalicin 2 are two of the most potent anti-angiogenic compounds
starting from the natural products. This approach limited us
mainly to modifications of the C6–O position which can be
liberated by saponification of the ester moiety and derivatives
(Scheme 1).6–8 Both fumagillin and ovalicin bind covalently to
that are accessible by ring opening of the spiroepoxide with
the active site of the enzyme methionine-aminopeptidase type
suitable nucleophiles.14,26–28 The relevance of the other functional
2 (MetAP2) and irreversibly block its proteolytic activity.8–10
The exact relation between the inactivation of MetAP2 and the
anti-angiogenic effect of fumagillin is not understood, though
groups, especially the isoprenoid side chain, has been barely
studied. Recently, Eustache et al. presented a strategy that allows
modifications of the side chain and the access to ring-substituted
there seems to exist vast evidence for causal coherence.8,11–14
analogues.29 To date, eight and four total syntheses have been
However, Phillips and coworkers could recently demonstrate
reported for fumagillin and ovalicin, respectively.30–42 However,
that depletion of MetAP2 by siRNA does not inhibit endothelial
none seemed to be applicable or practicable for our intentions.
cell growth.15 Moreover, MetAP2-depleted endothelial cells
Here we report the design of novel fumagillin and ovalicin
remain responsive to inhibition by fumagillin. These findings
analogues lacking one or both reactive epoxides. We also report
a new synthetic route for the synthesis of these analogues, and the
determination of their anti-angiogenic properties in endothelial
cells.
† Electronic supplementary information (ESI) available: experimental
data. See http://www.rsc.org/suppdata/ob/b5/b503163j/
Previous findings show that the conversion of the spiroepoxide
to a methylene group results in a considerable reduction of the
biological activity (IC50 10−7 M).13,43 Thus, the spiroepoxide is
essential for the activity of fumagillin.
On the other hand the epoxide of the side chain has no
major impact on the biological activity. As previous studies
show the transformation to a carbon–carbon double bond,
however, reduces the potency (IC50 5 × 10−9 M for endothelial
Scheme 1 Fumagillin 1, ovalicin 2 and TNP470 3.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 1 5 0 – 2 1 5 4
cell proliferation).13 Therefore, we planned to retain the polar
2 1 5 0
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
©

View Article Online
character of this position. This can be achieved by the formal
scission of the carbon–carbon bond of the oxirane, which
is equivalent to the transformation to an ether (Scheme 2).
Furthermore, the oxygen can be replaced by other heteroatoms
or appropriate functional groups. In all cases, the incision
increases the total chain length by one methylene unit. Therefore,
the distal portion of the side chain needs to be shortened
accordingly. The simplest case yields an allyl ether. Apart
from the dimethylallyl moiety, other functional groups such as
aromatic substituents seem to be rational.44
The racemic primary alcohol 9 is accessible starting from
inexpensive and commercially available 2-methoxy benzoic acid
4 that was cleanly converted with quantitative yields to the
corresponding ethyl ester 5 (Scheme 4). The aromatic ester was
then reduced with sodium in liquid ammonia under modified
Birch conditions to yield the 2,5-cyclohexadiene 6.45 The enol
ether resembles an activated ketone that reacted under acidic
conditions in ethylene glycol–DMF to form the spiroketal 7.
Osmium-catalyzed bishydroxylation of the remaining double
bond yielded the desired diol as a single diastereomer in
excellent yields. Both newly generated hydroxyl functionalities
were then protected as acetonide 7 followed by the lithium
aluminium hydride reduction of the ethyl ester to afford the
central intermediate 9 in 7 steps and 46% overall yield.
Scheme 2 Novel fumagillin analogues.
Although recent studies had shown that the spiroepoxide
is crucial for the exceptional potency, the substitution of this
oxirane with an inert functionality seemed to be very attractive.
We reasoned that the replacement of the epoxide by a ketone was
interesting, since it would not extensively affect the geometry of
the central scaffold or the spatial requirements.
Due to the abundance of biological data on TNP-470 and
the ease of its synthesis, we considered the substitution of
the polyunsaturated ester of fumagillin with the chloroacetyl
carbamate of TNP-470. Furthermore, we were interested in
utilizing the recently described 3,4,5-trimethoxy cinnamic acid
ester.14
Scheme 4 Synthesis of alcohol 9; (i) oxalyl chloride, DCM, 0 ◦C to
25 ◦C; (ii) EtOH, pyridine, DCM, 0 to 25 ◦C (100%, two steps); (iii) Na,
NH3–THF, −78 to −33 ◦C (84%); (iv) ethylene glycol, p-tosOH, DMF,
65 ◦C (70%); (v) N-methyl morpholine N-oxide, OsO4, acetone–water,
25 ◦C; (vi) 2,2-dimethoxy propane, p-tosOH, DMF, 65 ◦C (83%, two
steps); (vii) LAH, THF, 0 to 25 ◦C (95%).
For the synthesis of the fumagillin series, alcohol 9 was directly
alkylated with alkyl halides or oxidized with IBX in acetone to
yield the a-chiral aldehyde 10 in excellent yields, which was then
reacted with methyl magnesium bromide in THF at −78 ◦C
to afford the secondary alcohol 11 as a single diastereomer
(Scheme 5).46
Results and discussion
The retrosynthetic analysis is shown in Scheme 3. Both the
fumagillin and the ovalicin analogues, which feature an addi-
tional tertiary hydroxy group, can formally be dissected into four
basic elements that are linked by a central cyclohexane scaffold.
The individual residues comprise two vicinal, syn-oriented oxy
substituents, which can be ascribed to a bishydroxylation, an
adjacent anti-oriented alkyl substituent that can be utilized to
direct the stereospecific outcome of the bishydroxylation and an
adjoining spiroepoxide that can retrosynthetically be considered
as a ketone. The synthesis strategy should be designed in a way
that allows a wide variation of the alkyl side chain, a broad choice
of the acyl substituent and the modification and replacement
of the spiroepoxide, respectively. Both, the fumagillin and the
ovalicin series can be derived from one common precursor 9.
Scheme 5 Introduction of methyl branching; (i) IBX, acetone, reflux
(98%); (ii) methyl magnesium bromide, THF, −78◦C (96%).
The secondary alcohol 11 was analogously alkylated to alco-
hol 9 by treatment with NaH in DMF followed by addition of
alkylhalides to yield the desired ether 12. The vicinal diol could
selectively be deprotected by acid catalyzed transketalization
with ethylene glycol in DMF to afford compound 13 (Scheme 6).
Conversion of the diol 13 to the dibutyl tin ketal proved to be
a useful way to discriminate between the axial and equatorial
hydroxyl groups, since the tin ketal reacted with benzoyl chloride
in THF at low temperatures with a regioselectivity of 24 : 1 to
yield benzoate 14.47 The undesired regioisomer could easily be
separated by column chromatography and reisolated as starting
material 13 after saponification.
Diazomethane with catalytic amounts of boron trifluoride
etherate was the reagent of choice for the methylation of
the remaining alcohol functionality of 14. The spiroketal of
compound 15 that protected the carbonyl group and needed
to be removed at this step, proved to be exceptionally stable.
All standard methods for deprotection resulted in reisolation
or decomposition of starting material.48 However, the use of
catalytic amounts of cerium ammonium nitrate in aqueous
acetonitrile, that was recently described by Marko et al. cleanly
Scheme 3 Retrosynthetic analysis.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 1 5 0 – 2 1 5 4
2 1 5 1

View Article Online
yielded the desired ketone 16a and 16b, respectively.49 An
additional advantage of this procedure with regard to the
adjacent stereogenic center is the fact that the deprotection can
be carried out under neutral conditions (Scheme 6).
group was then acylated with either chloroacetyl isocyanate
or trimethoxy cinnamic acid chloride, followed by the CAN
catalyzed deprotection of the spiroketal to afford ketone 22 and
ketone 23, respectively (Scheme 7).
Scheme 7 Synthesis of fumagillin analogues 22 and 23; (i) Bu2SnO,
MeOH, reflux; (ii) MeI, DMF, 25 ◦C (53%, two steps); (iii) chloroacetyl
isocyanate, DCM, 0 ◦C; (iv) cerium ammonium nitrate, MeCN–water,
65 ◦C (41%, two steps); (v) 3,4,5-trimethoxy cinnamic acid chloride,
DMAP, DCM, 25 ◦C; (vi) cerium ammonium nitrate, MeCN–water,
65 ◦C (46%, two steps).
The tertiary hydroxyl group in 4-position, which is charac-
teristic of the ovalicin series, was introduced starting from the
central intermediate alcohol 9 (Scheme 8). Mesitylation followed
by potassium tert-butoxide mediated elimination generated the
exocyclic olefin 24 in 85% yield. After deprotection of the
acetonide, we took advantage of the free allylic hydroxyl group
of 25 that served as a directing neighbor group for the vanadium
catalyzed epoxidation of the double bond to give compound
26 in good yields.53 Of special interest is the dibutyltinoxide
mediated epoxide opening by the alcohol that is used as solvent
for the tin ketal formation in order to selectively benzoylate the
axial hydroxyl group of diol 26. Selective methylation of the
secondary alcohol of compound 27 over the tertiary alcohol
was accomplished with diazomethane with moderate selectivity
to yield 28 in 45% yield with the bismethylated byproduct 29
in 23% yield. The subsequent transformations were carried out
with minor modifications according to the fumagillin series.
Scheme 6 Synthesis of fumagillin analogues 19a,b and 20a,b; (i)
BnBr, NaH, Bu4NI, 0 to 25◦C (95–98%); (ii) ethylene glycol, p-tosOH,
DMF, 65 ◦C (98–100%); (iii) Bu2SnO, MeOH, reflux; (iv) BzCl, THF,
−40 to 25 ◦C (91–92%, two steps); (v) diazomethane, BF3·Et2O,
DCM, −78◦C; (vi) cerium ammonium nitrate, MeCN–water, 65 ◦C
(94–95% two steps); (vii) Nysted reagent; TiCl4, −78 to 25 ◦C (65%);
(viii) mCPBA, DCM, 0 ◦C (96% R = H, 6%); (ix) NaOH, MeOH,
25 ◦C (quant.); (x) chloroacetyl isocyanate, DCM, 0 ◦C (86–92%); (xi)
trimethyl sulfoxonium iodide, NaH, 25 ◦C (22%); (xii) NaOH, MeOH,
25 ◦C (quant.); (xiii) chloroacetyl isocyanate, DCM, 0 ◦C (81–94%).
The direct epoxidation of ketone 16 with trimethylsulfoxo-
nium ylide was only moderately successful in case of the un-
branched side chain (R = H). The derivatives with a bulkier side
chain yielded, depending on the reaction time and equivalents
of sulfur ylide, only the a,b-unsaturated ketone and the cyclo-
propanated consecutive product, respectively.35,38,50,51 However,
a two step detour via olefin 17 with subsequent epoxidation of
the double bond with m-chloroperbenzoic acid, turned out to
be a valuable alternative. The derivative with the unbranched
side chain was converted to the desired epoxide 18a as the only
product, whereas compound 17b (R = Me) yielded a mixture
of both diastereomeric epoxides. However, both diastereomers
could be easily separated by column chromatography.52
After saponification of the benzoate, the free hydroxy group
readily reacted in the last step with chloroacetyl isocyanate to
form carbamate 19.27,28 In a similar approach, the olefin 17 was
converted to the epoxide-lacking derivative 20. Minor modifi-
cations of the synthetic route gave access to the cyclohexanone
derivatives 22 and 23. For this purpose, diol 13a was converted
to the dibutyl tin ketal as described above. Interestingly, the
equatorial hydroxyl group exhibits the higher reactivity with
alkyl halides. This feature was utilized in the reaction of
the tin ketal with methyl iodide to give the desired methyl
ether with a regioselectivity of 9 : 1. The remaining hydroxyl
Scheme 8 Synthesis of ovalicin analogues; (i) mesitoyl chloride, NEt3,
DCM; (ii) KOtBu, DMF 0 ◦C (85% two steps); (iii) ethylene glycol,
p-tosOH, DMF, 65◦C (90%); (iv) tBuOOH, VO(acac)2, DCM (67%); (v)
Bu2SnO, allyl alcohol, reflux; (vi) BzCl, THF, −40 to 25 ◦C (53%); (vii)
diazomethane, BF3·Et2O, DCM, −78◦C.
All individual fumagillin and ovalicin derivatives were tested
for their anti-angiogenic properties in an endothelial cell pro-
liferation assay (Scheme 9, Table 1). For active compounds,
the IC50 values were determined (as shown for compound 23
in Scheme 9). We were pleased to find that compound 23,
which lacks both reactive epoxides, inhibits the proliferation
of endothelial cells in a dose dependent manner down to high
nanomolar concentrations.
The fact that the presence of the spiroepoxide has no major
impact on the activity in comparison to the corresponding
ketones or methylene derivatives, suggests that these compounds
might bind with a modified geometry their target, resulting in
an incapacitated orientation of the epoxide.
2 1 5 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 1 5 0 – 2 1 5 4

View Article Online
Table 1 Biological activity: inhibition of cell proliferation (HUVEC)
Compound
IC50/lM
19a
19b
20a
20b
22
1.5
20
11
12
3.6
0.8
3
20
9
35
27
15
53
23
24a
24b
30
31
32
33
34
Scheme 9 Dose dependent inhibition of cell proliferation (HUVEC)
by compound 23.
Scheme 10 Synthesised fumagillin and ovalicin analogues.
Acknowledgements
The comparison of compounds 19 and 24 shows that the
benzyl group is a suitable substitute for the isoprenoid side chain.
However, the introduction of the methyl branch results in a
reduction of activity.
The research was supported by the Baden-Wuerttemberg Lan-
desgraduiertenfo¨rdung and the Fonds der Chemischen Indus-
trie.
References
Conclusion
1 WHO, National cancer control programmes: policies and managerial
guidelines, 2nd ed., World Health Organization, Geneva, 2002.
2 J. Folkman and N. Engl, J. Med., 1971, 285, 1182–1186.
3 J. Folkman, Nat. Med., 1995, 1, 27–31.
Our approach to novel analogues of fumagillin and ovalicin
has proved to be very efficient and versatile. It offers also
access to structure–activity relationship studies. Apart from
modifications of the C6–O-acyl moiety, the synthetic approach
enables an extensive variation of the alkyl side chain as well as
the substitution of the spiroepoxide by alternative functionalities
(see Scheme 10). Starting from 2-methoxybenzoic acid the
desired target compounds were synthesised in 13 to 19 steps.
All reactions afforded high yields and can be easily carried out
on a multi-gram scale.
4 P. Carmeliet, Nat. Med., 2000, 6, 389–395.
5 D. W. Foster, in Harrison’s Principles of Internal Medicine, 14th
ed., ed. A. S. Fauci, E. Braunwald, K. J. Isselbacher, J. D.
Wilson, J. B. Martin, D. L. Kasper, S. L. Hauser and D. L. Longo,
McGraw Hill, New York, 1998, p. 2075.
6 T. E. Eble and F. R. Hanson, Antibiot. Chemother., 1951, 1, 54–58.
7 D. Ingber, T. Fujita, S. Kishimoto, K. Sudo, T. Kanamaru, H. Brem
and J. Folkman, Nature, 1990, 348, 555–557.
Compound 23 is of particular interest. Exhibiting an IC50
value of 0.5–1 lM, it was the most potent inhibitor of endothe-
lial cell proliferation among the tested fumagillin analogues.
This derivative has no reactive functionality and represents a
promising lead structure for the development of potent anti-
angiogenic drugs lacking the reactivity of fumagillin.
For experimental data refer to the electronic supplementary
information.
8 S. Liu, J. Widom, C. W. Kemp, C. M. Crews and J. Clardy, Science,
1998, 282, 1324–1327.
9 E. C. Griffith, Z. Su, B. E. Turk, S. Chen, Y. H. Chang, Z. Wu, K.
Biemann and J. O. Liu, Chem. Biol., 1997, 4, 461–471.
10 B. E. Turk, Z. Su and J. O. Liu, Bioorg. Med. Chem., 1998, 6, 1163–
1169.
11 B. E. Turk, E. C. Griffith, S. Wolf, K. Biemann, Y. H. Chang and
J. O. Liu, Chem. Biol., 1999, 6, 823–833.
12 H. U. Wang, Z. F. Chen and D. J. Anderson, Cell, 1998, 93, 741–753.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 1 5 0 – 2 1 5 4
2 1 5 3

View Article Online
13 E. C. Griffith, Z. Su, S. Niwayama, C. A. Ramsay, Y. H. Chang and
J. O. Liu, Proc. Natl. Acad. Sci. U. S. A., 1998, 95, 15183–15188. The
derivative was synthesised starting from fumagillin by deoxygenation
with WCl6–nBuLi.
14 C. K. Han, S. K. Ahn, N. S. Choi, R. K. Hong, S. K. Moon, H. S.
Chun, S. J. Lee, J. W. Kim, C. I. Hong, D. Kim, J. H. Yoon and K. T.
No, Bioorg. Med. Chem. Lett., 2000, 10, 39–43.
15 S. Kim, K. LaMontagne, M. Sabio, S. Sharma, R. W. Versace, N.
Yusuff and P. E. Phillips, Cancer Res., 2004, 64, 2984–2987.
16 N. Wernert, A. Stanjek, S. Kiriakidis, A. Hugel, H. C. Jha, R.
Mazitschek and A. Giannis, Angew. Chem., Int. Ed., 1999, 38, 3228–
3231.
33 D. A. Vosburg, S. Weiler and E. J. Sorensen, Angew. Chem., Int. Ed.,
1999, 38, 971–974.
34 W. Picoul, R. Urchegui, A. Haudrechy and Y. Langlois, Tetrahedron
Lett., 1999, 40, 4797–4800.
35 J.-G. Boiteau, P. Van de Weghe and J. Eustache, Org. Lett., 2001, 3,
2737–2740.
36 D. A. Vosburg, S. Weiler and E. J. Sorensen, Chirality, 2003, 15,
156–166.
37 W. Picoul, O. Bedel, A. Haudrechy and Y. Langlois, Pure Appl. Chem.,
2003, 75, 235–249.
38 M. Hutchings, D. Moffat and N. S. Simpkins, SYNLETT, 2001,
661–663.
17 B. J. Dezube, J. H. Von Roenn, J. Holden-Wiltse, T. W. Cheung, S. C.
Remick, T. P. Cooley, J. Moore, J. P. Sommadossi, S. L. Shriver, C. W.
Suckow and P. S. Gill, J. Clin. Oncol., 1998, 16, 1444–1449.
18 C. J. Logothetis, K. K. Wu, L. D. Finn, D. Daliani, W. Figg, H.
Ghaddar and J. U. Gutterman, Clin. Cancer Res., 2001, 7, 1198–
1203.
19 W. M. Stadler, T. Kuzel, C. Shapiro, J. Sosman, J. Clark and N. J.
Vogelzang, J. Clin. Oncol., 1999, 17, 2541–2545.
20 A. P. Kudelka, C. F. Verschraegen and E. Loyer, J. Med., 1998, 338,
991–992.
39 A. Barco, S. Benetti, C. De Risi, P. Marchetti, G. P. Pollini and V.
Zanirato, Tetrahedron: Asymmetry, 1998, 9, 2857–2864.
40 E. J. Corey, A. Guzman-Perez and M. C. Noe, J. Am. Chem. Soc.,
1994, 116, 12109–12110.
41 S. Bath, D. C. Billington, S. D. Gero, B. Quiclet-Sire and M. Samadi,
J. Chem. Soc., Chem. Commun., 1994, 1495–1496.
42 E. J. Corey and J. P. Dittami, J. Am. Chem. Soc., 1985, 107, 256–257.
43 The derivative was synthesised starting from fumagillin by deoxy-
genation with WCl6–nBuLi.
44 M. Fardis, H. J. Pyun, J. Tario, H. Jin, C. U. Kim, J. Ruckman, Y. Lin,
L. Green and B. Hicke, Bioorg. Med. Chem., 2003, 11, 5051–5058.
45 P. W. Rabideau, D. M. Wetzel and D. M. Young, J. Org. Chem., 1984,
49, 1544–1549.
46 Whereas employing methyl lithium in lieu of Grignard reagents yields
a mixture of both diastereomers. We assume that the complexation of
the aldehyde, apart from steric factors and its spatial fixation is crucial
for the excellent stereoselectivity. The configuration of the newly
generated stereocenter was assigned by transformation of compound
10 into a dioxydecaline derivative and comparison of the chemical
shifts to literature data; see: K. Pihlaja, J. Mattinen and J. Czombos,
J. Prakt. Chem./Chem.-Ztg., 1997, 339, 77–78.
47 T. B. Grindley, Adv. Carbohydr. Chem., 1998, 53, 17–142.
48 T. W. Greene, P. G. M. Wuts, Protective Groups in Org. Synth., 2nd
edn., 1991, p. 317.
49 I. E. Marko, A. Ates, A. Gautier, B. Leroy, J.-M. Plancher, Y. Quesnel
and J.-C. Vanherck, Angew. Chem., Int. Ed., 1999, 38, 3207–3209.
50 E. J. Corey and M. Chaykovsky, J. Am. Chem. Soc., 1965, 87, 1353–
1364.
51 Depending on reaction time, conditions and number of equivalents of
ylide, the major side products were either the a,b-unsaturated ketone
or the cyclopropylated consecutive product.
52 The relative configuration of both epoxides was assigned by NOE.
53 K. B. Sharpless and R. C. Michaelson, J. Am. Chem. Soc., 1973, 95,
6136–6137.
21 V. Castronovo and D. Belotti, Eur. J. Cancer, 1996, 32A, 2520–
2527.
22 W. D. Figg, J. M. Pluda, R. M. Lush, M. W. Saville, K. Wyvill, E.
Reed and R. Yarchoan, Pharmacotherapy, 1997, 17, 91–97.
23 P. Bhargava, J. L. Marshall, N. Rizvi, W. Dahut, J. Yoe, M. Figuera,
K. Phipps, V. S. Ong, A. Kato and M. J. Hawkins, Clin. Cancer Res.,
1999, 5, 1989–1995.
24 L. Placidi, E. Cretton-Scott, G. de Sousa, R. Rahmani, M. Placidi
and J. P. Sommadossi, Cancer Res., 1995, 55, 3036–3042.
25 G. Zhou, C. W. Tsai and J. O. Liu, J. Med. Chem., 2003, 46, 3452–
3454.
26 S. Marui, F. Itoh, Y. Kozai, K. Sudo and S. Kishimoto, Chem. Pharm.
Bull., 1992, 40, 96–101.
27 S. Marui and S. Kishimoto, Chem. Pharm. Bull., 1992, 40, 575–
579.
28 S. Marui, T. Yamamoto, K. Sudo, H. Akimoto and S. Kishimoto,
Chem. Pharm. Bull., 1995, 43, 588–593.
29 V. Rodeschini, J.-G. Boiteau, P. Van de Weghe, C. Tarnus and J.
Eustache, J. Org. Chem., 2004, 69, 357–373.
30 E. J. Corey and B. B. Snider, J. Am. Chem. Soc., 1972, 94, 2549–2550.
31 D. Kim, S. K. Ahn, H. Bae, W. J. Choi and H. S. Kim, Tetrahedron
Lett., 1997, 38, 4437–4440.
32 D. F. Taber, T. E. Christos, A. L. Rheingold and I. A. Guzei, J. Am.
Chem. Soc., 1999, 121, 5589–5590.
2 1 5 4
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 1 5 0 – 2 1 5 4