Fumagalone, a reversible inhibitor of type 2 methionine aminopeptidase and angiogenesis

Article abstract of DOI:10.1021/jm0341103

Fumagillin and ovalicin constitute a family of structurally related natural products that possess antiangiogenic activity. We report the synthesis of a new fumagillin analogue, fumagalone, in which the spiroepoxide group is replaced with an aldehyde. Fuma

Full text of DOI:10.1021/jm0341103

3452
J . Med. Chem. 2003, 46, 3452-3454
F u m a ga lon e, a Rever sible In h ibitor of
Typ e 2 Meth ion in e Am in op ep tid a se a n d
An giogen esis
Guochun Zhou, Chiawei W. Tsai, and J un O. Liu*
Department of Pharmacology and Molecular Sciences and
Department of Neuroscience, J ohns Hopkins University
School of Medicine, Baltimore, Maryland 21205
Received May 23, 2003
F igu r e 1. Structures of fumagillin, TNP-470, ovalicin, and
fumagalone.
Abstr a ct: Fumagillin and ovalicin constitute a family of
structurally related natural products that possess antiangio-
genic activity. We report the synthesis of a new fumagillin
analogue, fumagalone, in which the spiroepoxide group is
replaced with an aldehyde. Fumagalone inhibits type 2 me-
thionine aminopeptidase (MetAP2) with IC50 ) 8 µM and
endothelial cell proliferation with IC50 ) 52 nM. With dialysis
and competition assays, it was unambiguously demonstrated
that binding of fumagalone to MetAP2 is reversible.
Sch em e 1a
Fumagillin (1) and ovalicin (3) are structurally related
natural products of fungal origin1,2 that have been
shown to possess potent antiangiogenic activity.3,4 An
analogue of fumagillin, TNP-470 (2, also known as
AGM-1470), has undergone clinical trials for the treat-
ment of a variety of cancers. At the cellular level, these
compounds inhibit angiogenesis by blocking the cell
cycle progression of endothelial cells primarily in the
G1-to-S phase transition.5-8 At the molecular level, the
direct target for both fumagillin9,10 and ovalicin9 has
been identified as the type 2 methionine aminopeptidase
(MetAP2). Recent chemical, biochemical, and structural
studies have revealed that these natural products are
bound to MetAP2 irreversibly by covalently modifying
a histidine side chain at the active site of MetAP2 with
the spiroepoxide group.11-13 The binding of these natu-
ral products to MetAP2 leads, in a yet unknown mech-
anism, to the activation of the transcription factor p53,
causing the transcriptional up-regulation of p21 that
inhibits cyclin E/Cdk2 kinase activity required for cell
cycle progression.14,15
a
(a) NaSPh/DMF, room temp, 2.5 h; (b) mCPBA/CHCl3, -78
°C, 1 h; (c) TPAP/NMO, 1 h; (d) NaOAc/Ac2O, 155 °C (oil bath), 1
h; (e) K2CO3/MeOH/H2O, 15 min.
Because the spiroepoxide group in fumagillin and
ovalicin reacts with His231 imidazole side chain (eq
1),11,12 we reasoned that conversion of the spiroepoxide
into an aldehyde would allow for the retention of the
bonding distance between the C2 carbon in fumagillin
and the ꢀN of His231 in MetAP2. Since the carbonyl
group of an aldehyde remains electrophilic, it was
anticipated that the nucleophilic nitrogen of His231
would react with the aldehyde by forming an aminal
intermediate (eq 2). It was not certain, however, whether
the additional hydroxyl group formed as part of the
aminal could be accommodated by the active site of
MetAP2.
TNP-470 has encountered some problems in human
clinical trails. Among them are short serum half-life and
dose-limiting neurotoxicity.16,17 The spiroepoxide in
TNP-470 is prone to hydrolysis in a reaction catalyzed
by epoxide hydrolases in vivo. The irreversible binding
of TNP-470 to MetAP2 could also be responsible, at least
in part, for its dose-limiting side effects. A reversible
inhibitor of MetAP2 thus became not only an interesting
intellectual challenge but also a good candidate for
developing new angiogenesis inhibitors that are less
toxic than TNP-470. As a first step toward a reversible
MetAP2 inhibitor, we converted the spiroepoxide in
fumagillin responsible for covalent modification of
MetAP2 into an aldehyde. The new fumagillin analogue,
named fumagalone (4, Figure 1), was shown to be active,
albeit with decreased potency, as an inhibitor of MetAP2
in vitro and endothelial cell proliferation in vivo.
The synthesis of fumagalone commences with fum-
agillol, a fumagillin analogue in which the lower polyene
side chain has been removed by hydrolysis (Scheme 1).18
Opening of the spiroepoxide with sodium benzenethiol
in DMF afforded adduct 6, which was then oxidized by
mCPBA to yield the sulfoxide 7. The secondary alcohol
in 7 was converted into a ketone (8) by treatment with
TPAP/NMO. Intermediate 8 underwent a Pummerer
* To whom correspondence should be addressed. Phone: 410-955-
4619. Fax: 410-955-4620. E-mail: joliu@jhu.edu.
10.1021/jm0341103 CCC: $25.00 © 2003 American Chemical Society
Published on Web 07/09/2003

Letters
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 16 3453
rearrangement upon treatment with NaOAc and acetic
anhydride to give rise to an intermediate that, upon
hydrolysis in the presence of potassium carbonate,
yielded fumagalone.19,20 The overall yield of this five-
step process is 16%.
Fumagalone is active in both the MetAP2 enzymatic
assay in vitro and in endothelial cell proliferation assay.
By use of 100 nM of MetAP2 enzyme and the tetrapep-
tide Met-Gly-Met-Met as a substrate, the IC50 for
fumagalone was determined to be 8 µM. For the bovine
aortic endothelial cell proliferation assay, fumagalone
had an IC50 of about 52 nM. In comparison with its
natural precursor fumagillin (IC50 of 200 nM for MetAP2
and 1 nM for bovine aortic endothelial cell proliferation),
the potency of fumagalone is decreased by 40- to 50-
fold. We have also determined the effect of fumagalone
on type 1 human methionine aminopeptidase. Like
fumagillin and ovalicin, fumagalone has little effect on
human MetAP1 at the highest concentration tested (50
µM), indicating that fumagalone remains highly specific
for human MetAP2.
F igu r e 2. Reversible binding of fumagalone to MetAP2. (A)
Relative MetAP2 enzymatic activity upon preincubation with
DMSO (bar 1) or 50 µM fumagalone (bar3), followed by
dialysis. As a control, MetAP2 activity was determined in the
presence of 50 µM fumagalone without dialysis (bar 2). (B)
Commassie blue stained polyacrylamide gel for MetAP2 bound
by biotin-fumagillin after preincubation with solvent (lane 1),
TNP-470 (lane 2), or fumagalone (lane 3).
It is noteworthy that the IC50 for inhibition of MetAP2
in vitro by fumagalone is over 100-fold higher than that
for inhibition of endothelial cell proliferation in cell
culture. The same is true for fumagillin and most other
fumagallin analogues tested to date.9 This difference in
IC50 for MetAP2 and endothelial cell inhibition may be
attributable to two factors. First, the sensitivity of the
MetAP2 assay in vitro is such that it requires 20-100
nM of recombinant MetAP2 to detect the product signal.
And the minimum IC50 for an inhibitor as determined
by the in vitro MetAP2 enzyme assay is at least half of
the enzyme concentration used. In contrast, the total
MetAP2 concentration in a cellular assay is determined
by the level of expression of MetAP2 in endothelial cells,
which could be much lower. Second, because both
fumagillin and fumagalone (presumably) form a cova-
lent adduct with MetAP2, intracellular MetAP2 could
serve as a trap for both inhibitors in a cellular assay,
significantly raising intracellular concentration of each
inhibitor in comparison to the concentrations of inhibi-
tors in the culture media.
is followed by the addition of the biotin-fumagillin
conjugate, which binds to MetAP2 irreversibly. The
MetAP2 protein bound by the biotin-fumagillin conju-
gate is captured using streptavidin-sepharose, which
binds to the biotin moiety of the conjugate,9 followed
by SDS-polyacrylamide gel electrophoresis and subse-
quent staining with commassie blue to visualize the
bound MetAP2 protein. Biotin-fumagillin was able to
compete effectively for MetAP2 after its preincubation
with fumagalone (Figure 2B, lane 3) but not with TNP-
470 (Figure 2B, lane 2). These results unambiguously
demonstrate that fumagalone, unlike TNP-470, is bound
to MetAP2 reversibly.
Although we have shown that fumagalone (4) inhibits
MetAP2 reversibly, in contrast to fumagillin or its
synthetic derivative TNP-470, it remains to be proven
that fumagalone forms the tetrahedral adducts as
envisioned. The final proof will have to await the X-ray
structure of the complex between fumagalone and
human MetAP2.
In addition to fumagalone, we have also synthesized
several other aldehyde-containing analogues of fumag-
illin. There appears to be a stringent structural require-
ment for the activity of the reversible inhibitor because
reduction of the ketone in fumagalone to an alcohol
(Table 1, 9) significantly decreased its activity against
MetAP2 and almost completely abrogated its activity
in endothelial cells. Inversion of stereochemistry at C6
(Table 1, 10) did not recover any activity. It is possible
that an intramolecular hydrogen bond is formed be-
tween the newly formed hydroxyl group of the aminal
and a yet-to-be-determined residue in MetAP2, which
may place a severe conformational constraint on the
juxtaposition of the cyclohexyl ring within the active site
of MetAP2.
Two distinct biochemical assays were performed to
determine whether fumagalone is bound to MetAP2
reversibly. In the first assay, fumagalone was preincu-
bated with purified recombinant human MetAP2.11 The
reaction mixture containing fumagalone and MetAP2
was then subjected to dialysis in 1000 volumes of buffer
at 4 °C, with the dialysis buffer changed twice during a
24 h period. The dialyzed mixture was then checked for
the recovery of the enzymatic activity of MetAP2, in
comparison with MetAP2 samples that are preincubated
with DMSO as a solvent carrier control (normalized to
100%). As shown in Figure 2A, the presence of 50 µM
fumagalone inhibited MetAP2 activity by over 96%
(Figure 2A, bar 2). Upon dialysis, close to 80% of the
activity of MetAP2 was recovered (Figure 2A, bar 3) in
comparison to the solvent control (Figure 2A, bar 1),
strongly suggesting that fumagalone is bound to MetAP2
reversibly. In an alternative assay, we determined
whether fumagalone is capable of competing with the
biotin-fumagillin conjugate for binding to MetAP2 in
a reversible manner. Thus, recombinant MetAP2 was
preincubated with fumagalone or TNP-470 for 1 h. This
The synthesis of fumagalone as a reversible inhibitor
of MetAP2 paves the way to designing, synthesizing,
and identifying more potent reversible inhibitors of this
enzyme using the fumagillin core structure as a leading

3454 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 16
Letters
(3) Ingber, D.; Fujita, T.; Kishmoto, S.; Sudo, K.; Kanamaru, T.; et
al. Synthetic Analogues of Fumagillin That Inhibit Angiogenesis
and Suppress Tumour Growth. Nature 1990, 348, 555-557.
(4) Corey, E. J .; Guzman-Perez, A.; Noe, M. C. Short enantioselective
synthesis of (-)-ovalicin, a potent inhibitor of angiogenesis, using
substrate-enhanced catalytic asymmetric dihydroxylation. J .
Am. Chem. Soc. 1994, 116, 12109-12110.
Ta ble 1. Activity of Fumgalone and Its Structurally Related
Analogues
(5) Kusaka, M.; Sudo, K.; Matsutani, E.; Kozai, Y.; Marui, S.; et al.
Cytostatic inhibition of endothelial cell growth by the angiogen-
esis inhibitor TNP-470. Br. J . Cancer 1993, 69, 212-216.
(6) Antoine, N.; Greimers, R.; De Roanne, C.; Kusaka, M.; Heinen,
E.; et al. AGM-1470, a potent angiogenesis inhibitor, prevents
the entry of normal but not transformed endothelial cells into
the G1 phase of the cell cycle. Cancer Res. 1994, 54, 2073-2076.
(7) Abe, J .; Zhou, W.; Takuwa, N.; Taguchi, J .; Kurokawa, K.; et al.
A Fumagillin Derivative Angiogenesis Inhibitor, AGM-1470,
Inhibits Activation of Cyclin-Dependent Kinases and Phospho-
rylation of Retinoblastoma Gene Product but Not Protein Tyrosyl
Phosphorylation or Protooncogene Expression in Vascular En-
dothelial Cells. Cancer Res. 1994, 54, 3407-3412.
(8) Hori, A.; Ikeyama, S.; Sudo, K. Suppression of cyclin D1 mRNA
expression by the angiogenesis inhibitor TNP-470 (AGM-1470)
in vascular endothelial cells. Biochem. Biophys. Res. Commun.
1994, 204, 1067-1073.
(9) Griffith, E. C.; Su, Z.; Turk, B. E.; Chen, S.; Chang, Y.-W.; et al.
Methionine aminopeptidase (type 2) is the common target for
angiogenesis inhibitors AGM-1470 and ovalicin. Chem. Biol.
1997, 4, 461-471.
a
b
BAEC ) bovine aortic endothelial cell. No inhibition at 50
µM or above.
pharmacophore. While the decrease in potency upon
conversion of an irreversible inhibitor into a reversible
one may be an inevitable price to pay, there is relatively
large room to accommodate such a loss in activity, given
the extremely high potency of the epoxide-containing
precursors. The ready availability of fumagillin as a
fermentation product and the relatively short synthetic
route from fumagillin to fumagalone should facilitate
large-scale production of related analogues. On the basis
of the results reported in this manuscript, we have
begun to synthesize the second generation of derivatives
to further probe the structure-activity relationship of
this new class of inhibitors of MetAP2. It will be
interesting to see whether fumagalone-based reversible
inhibitors of MetAP2 will have improved pharmacologi-
cal properties and decreased side effects in comparison
with TNP-470. Progress along these lines will be
reported in due course.
(10) Sin, N.; Meng, L.; Wang, M. Q. W.; Wen, J . J .; Bornmann, W.
G.; et al. The anti-angiogenic agent fumagillin covalently binds
and inhibits the methionine aminopeptidase, MetAP-2. Proc.
Natl. Acad. Sci. U.S.A. 1997, 94, 6099-6103.
(11) Griffith, E. C.; Su, Z.; Niwayama, S.; Ramsay, C. A.; Chang, Y.
H.; et al. Molecular recognition of angiogenesis inhibitors
fumagillin and ovalicin by methionine aminopeptidase 2. Proc.
Natl. Acad. Sci. U.S.A. 1998, 95, 15183-15188.
(12) Liu, S.; Widom, J .; Kemp, C. W.; Crews, C. M.; Clardy, J .
Structure of human methionine aminopeptidase-2 complexed
with fumagillin. Science 1998, 282, 1324-1327.
(13) Lowther, W. T.; McMillen, D. A.; Orville, A. M.; Matthews, B.
W. The anti-angiogenic agent fumagillin covalently modifies a
conserved active-site histidine in the Escherichia coli methionine
aminopeptidase. Proc Natl Acad Sci U.S.A. 1998, 95, 12153-
12157.
(14) Zhang, Y.; Griffith, E. C.; Sage, J .; J acks, T.; Liu, J . O. Cell cycle
inhibition by the anti-angiogenic agent TNP-470 is mediated by
p53 and p21WAF1/CIP1. Proc. Natl. Acad. Sci. U.S.A. 2000, 97,
6427-6432.
(15) Yeh, J . R.; Mohan, R.; Crews, C. M. The antiangiogenic agent
TNP-470 requires p53 and p21CIP/WAF for endothelial cell
growth arrest. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 12782-
12787.
(16) Placidi, L.; Cretton-Scott, E.; de Sousa, G.; Rahmani, R.; Placidi,
M.; et al. Disposition and metabolism of the angiogenic modera-
tor O-(chloroacetyl-carbamoyl) fumagillol (TNP-470; AGM-1470)
in human hepatocytes and tissue microsomes. Cancer Res. 1995,
55, 3036-3042.
Ack n ow led gm en t. This work was supported by the
NCI and the J ohns Hopkins School of Medicine. We
thank Dr. Yi Zhang for technical advice on the bio-
chemical and endothelial cell proliferation assays.
(17) Bhargava, P.; Marshall, J . L.; Rizvi, N.; Dahut, W.; Yoe, J .; et
al. A phase I and pharmacokinetic study of TNP-470 adminis-
tered weekly to patients with advanced cancer. Clin. Cancer Res.
1999, 5, 1989-1995.
(18) Tarbell, D. S.; Carman, R. M.; Chapman, D. D.; Cremer, S. E.;
Cross, A. D.; et al. The chemistry of fumagillin. J . Am. Chem.
Soc. 1961, 83, 3096-3113.
(19) Pummerer, R. Phenylsulphoxyacetic acid. II. Ber. Dtsch. Chem.
Ges. 1910, 43, 1401-1412.
(20) Horner, L. The reaction of carboxylic acid anhydrides with
sulfoxides. Ann. Phys. Chem. Poggendorf 1960, 631, 198-199.
Su p p or tin g In for m a tion Ava ila ble: Synthetic proce-
dures for and characterization of fumagalone and intermedi-
ates 6-8 as well as the protocols for the biochemical assays.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Refer en ces
(1) Hanson, F. R.; Eble, T. E. An antiphage agent isolated from
aspergillus sp. J . Bacteriol. 1949, 58, 527-529.
(2) Sigg, H. P.; Weber, H. P. Isolation and structure elucidation of
ovalicin. Helv. Chim. Acta 1968, 51, 1395-1408.
J M0341103