Spiroepoxytriazoles Are Fumagillin-like Irreversible Inhibitors of MetAP2 with Potent Cellular Activity

Article abstract of DOI:10.1021/acschembio.5b00755

Methionine aminopeptidases (MetAPs) are responsible for the cotranslational cleavage of initiator methionines from nascent proteins. The MetAP2 subtype is up-regulated in many cancers, and selective inhibition of MetAP2 suppresses both vascularization and growth of tumors in animal models. The natural product fumagillin is a selective and potent irreversible inhibitor of MetAP2, and semisynthetic derivatives of fumagillin have shown promise in clinical studies for the treatment of cancer, and, more recently, for obesity. Further development of fumagillin derivatives has been complicated, however, by their generally poor pharmacokinetics. In an attempt to overcome these limitations, we developed an easily diversifiable synthesis of a novel class of MetAP2 inhibitors that were designed to mimic fumagillin's molecular scaffold but have improved pharmacological profiles. These substances were found to be potent and selective inhibitors of MetAP2, as demonstrated in biochemical enzymatic assays against three MetAP isoforms. Inhibitors with the same relative and absolute stereoconfiguration as fumagillin displayed significantly higher activity than their diastereomeric and enantiomeric isomers. X-ray crystallographic analysis revealed that the inhibitors covalently modify His231 in the MetAP2 active site via ring-opening of a spiroepoxide. Biochemically active substances inhibited the growth of endothelial cells and a MetAP2-sensitive cancer cell line, while closely related inactive isomers had little effect on the proliferation of either cell type. These effects correlated with altered N-terminal processing of the protein 14-3-3-γ. Finally, selected substances were found to have improved stabilities in mouse plasma and microsomes relative to the clinically investigated fumagillin derivative beloranib.

Full text of DOI:10.1021/acschembio.5b00755

Articles
pubs.acs.org/acschemicalbiology
Spiroepoxytriazoles Are Fumagillin-like Irreversible Inhibitors of
MetAP2 with Potent Cellular Activity
Michael Morgen,†,⊥ Christian Jost,‡,⊥ Mona Malz,† Robert Janowski,§ Dierk Niessing,§,∥
̈
Christian D. Klein,‡ Nikolas Gunkel,† and Aubry K. Miller*,†
†Cancer Drug Development Group, German Cancer Research Center (DKFZ), Im Neunheimer Feld 280, D-69120 Heidelberg,
Germany
‡Medicinal Chemistry, Institute of Pharmacy and Molecular Biotechnology (IPMB), Heidelberg University, Im Neuenheimer Feld
364, D-69120 Heidelberg, Germany
§Institute of Structural Biology, Helmholtz Zentrum Munchen, German Research Center for Environmental Health (HMGU),
̈
D-85764 Neuherberg, Germany
∥Biomedical Center of the Ludwig-Maximilians-Universitat Munchen, D-82152 Planegg-Martinsried, Germany
̈
̈
S
* Supporting Information
ABSTRACT: Methionine aminopeptidases (MetAPs) are responsible for
the cotranslational cleavage of initiator methionines from nascent proteins.
The MetAP2 subtype is up-regulated in many cancers, and selective
inhibition of MetAP2 suppresses both vascularization and growth of tumors
in animal models. The natural product fumagillin is a selective and potent
irreversible inhibitor of MetAP2, and semisynthetic derivatives of fumagillin
have shown promise in clinical studies for the treatment of cancer, and,
more recently, for obesity. Further development of fumagillin derivatives has
been complicated, however, by their generally poor pharmacokinetics. In an
attempt to overcome these limitations, we developed an easily diversifiable
synthesis of a novel class of MetAP2 inhibitors that were designed to mimic
fumagillin’s molecular scaffold but have improved pharmacological profiles. These substances were found to be potent and
selective inhibitors of MetAP2, as demonstrated in biochemical enzymatic assays against three MetAP isoforms. Inhibitors with
the same relative and absolute stereoconfiguration as fumagillin displayed significantly higher activity than their diastereomeric
and enantiomeric isomers. X-ray crystallographic analysis revealed that the inhibitors covalently modify His231 in the MetAP2
active site via ring-opening of a spiroepoxide. Biochemically active substances inhibited the growth of endothelial cells and a
MetAP2-sensitive cancer cell line, while closely related inactive isomers had little effect on the proliferation of either cell type.
These effects correlated with altered N-terminal processing of the protein 14-3-3-γ. Finally, selected substances were found to
have improved stabilities in mouse plasma and microsomes relative to the clinically investigated fumagillin derivative beloranib.
Interest in the biological properties of 1 initially stemmed
atural products have been an excellent source of leads for
Ndrug development, particularly in the areas of antibiotic from Ingber and Folkman’s serendipitous discovery that it
selectively inhibits the growth of endothelial cells at
subnanomolar concentrations,2 propelling studies first into its
use as an antiangiogenic cancer therapy14 and, more recently, in
obesity indications, where beloranib (3) is currently inves-
tigated in a number of phase 2 and 3 trials.15 Studies to identify
fumagillin’s primary biological target16−18 eventually resulted in
an X-ray crystal structure of 1 bound to methionine
aminopeptidase 2 (MetAP2),19 a metalloprotease responsible
for the cotranslational cleavage of initiator methionines from
ribosomally synthesized proteins. Subsequent studies identified
the cell cycle regulator p21 as the furthest downstream effector
of fumagillin’s inhibitory activity on endothelial cells,20,21 but
and anticancer therapies where natural products (and their
derivatives) comprise a significant percentage of clinically used
drugs.1 Sometimes a natural product can be developed into a
drug, but often structural modifications to a natural product’s
scaffold are required to improve its pharmacological properties
such that it can successfully be administered to human patients.
Due to the structural complexity of many natural products,
semisynthetic approaches are often the method of choice for
optimization campaigns. A limitation to this approach is that
the resulting diversification strategies may be largely
opportunistic in nature, guided by issues of synthetic
convenience and practicality, i.e., where the molecule can
selectively be modified. A natural product where such an
approach has been applied is fumagillin (1), a metabolite of the
fungus Aspergillus fumigatus (Figure 1).2−13
Received: September 21, 2015
Accepted: December 18, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
instability in the isoprenoid side-chain.28,29 We chose to
address this problem through a de novo synthesis approach,
which would allow us to explore chemical space that is
inaccessible when using the natural product as a starting
material and designed a new class of spiro-epoxide containing
MetAP2 inhibitors that are structurally inspired by fumagil-
lin.30−35 The principle of our design was that these compounds
should (1) be stereochemically less complex than fumagillin,
(2) incorporate synthetically accessible and potentially drug-
like nitrogen-containing heterocycles, and (3) closely resemble
fumagillin’s three-dimensional structure so as to maintain its
highly optimized intermolecular fit into the target enzyme.
Taking into account previous SAR data, our own design
principles, and using molecular models, we hypothesized that
tetrahydrobenzotriazoles of the general structures 5 would be
fumagillin-like inhibitors, while cognizant of the risks associated
with such a scaffold jump (Figure 3). We report, herein, the
Figure 1. Fumagillin and clinically investigated derivatives. The
conserved fumagillol core is depicted in black, while side chain
variations are depicted in gray.
the mechanistic steps that link MetAP2 inhibition to p21
activation are still incompletely understood.22,23 Similarly, the
connection between fumagillin-based MetAP2 inhibition and
weight loss is still under investigation, and nonenzymatic
suppression of ERK1/2 phosphorylation has been impli-
cated.15,24
Figure 3. Retrosynthetic analysis of target compound 5.
MetAP2 is irreversibly inhibited by fumagillin via covalent
bond formation between the spiro-oxirane C2 and a histidine
residue (His231) in the active site of the enzyme (Figure 2).
synthesis of such a chemical series, and the characterization of
certain members as potent and selective MetAP2 inhibitors
which covalently modify His231 in the active site of the
enzyme. Moreover, active substances exhibit on-target growth
inhibition of endothelial cells, reduced proliferation of a
MetAP2-sensitive cancer cell line, and improved plasma and
microsomal stabilities relative to beloranib.36
RESULTS AND DISCUSSION
■
Figure 2. Covalent inhibition of MetAP2 by fumagillin.
Synthesis of Epoxytriazoles. At the outset of this project,
there were no published approaches to spiroepoxy tetrahy-
drobenzotriazoles 5 or their envisioned precursors 6. Moreover,
we were concerned that the intermediates 6, upon formation,
would spontaneously aromatize to the corresponding hydroxy-
benzotriazoles. Nevertheless, motivated by the potential to
rapidly synthesize a library of fumagillin analogs, we planned to
access compounds like 6 via Heck cyclization of iodoalkenes
like 7. In the forward sense, our synthetic route began with a
one-pot reaction (Scheme 1): Iodination and subsequent
deprotonation of ethynylmagnesium bromide (8) gave lithium
iodoacetylide (9) which reacted with Weinreb amide 1037 to
give iodoynone 11, a compound prone to decomposition and
While the isoprenoid side-chain of 1 is critical for MetAP2
binding by filling a hydrophobic pocket of the enzyme and
simultaneously positioning the oxirane moiety for reaction, the
polyunsaturated C6 side chain is solvent-exposed and not
required for activity.
The lack of a strict structural requirement at C6, coupled
with the ease by which the C6 ester moiety can be hydrolyzed
(to fumagillol) and the resulting secondary hydroxyl group
further modified, enabled a range of structure−activity
relationship (SAR) studies at this position.2−4,7,8,13 Indeed,
modifications to the fumagillin structure were required because
clinical studies with 1 and its derivative TNP-470 (2, Figure 1)
were hampered by the poor pharmacological properties of the
two substances, in particular extremely short plasma half-life
(∼2 min)25 and CNS-related side effects.26 More recently
developed derivatives like 3 and PPI-2458 (4) have improved
pharmacological profiles (no CNS effects) but still suffer from
relatively poor pharmacokinetics.27,28 For this reason, 3 is
currently administered as a subcutaneous suspension.
a
Scheme 1. Synthesis of Iodoynone 11
While compounds like 3 and 4 demonstrate that
modifications at C6 have the potential to alleviate some of
the pharmacological weaknesses associated with fumagillin,
such modifications will never be able to overcome liabilities
inherent to the fumagillol substructure, e.g., metabolic
a
Reagents and conditions: I2 (1.0 equiv), THF, −78 to 0 °C, 1 h, then
LDA (1.0 equiv), −78 to 0 °C, 30 min, then 10 (1.0 equiv), −78 °C to
−20 °C, 2 h, 60%.
B
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
that we were unable to prepare from the corresponding
terminal ynone, but which could be reliably prepared on a 50
mmol scale in this manner.38,39
Scheme 3. Representative Synthesis of Diasteromeric C6 O-
acyl Epoxides
a
Alkyne 11 could be coupled with a variety of azides in copper
catalyzed Huisgen cyclizations to generate iodotriazoles of the
general structure 7.40 For example, reaction with 4-methoxy-
benzyl azide (PMBN3) cleanly gave trisubstituted triazole 12 as
a single regioisomer (Scheme 2). Most attempts to form bicycle
a
Scheme 2. Representative Ketoepoxide Synthesis
a
Reagents and conditions: (a) DIBAL (1.6 equiv), CH2Cl2, −78 °C,
30 min, 85%; (b) benzoyl chloride (1.2 equiv), Et3N (1.4 equiv),
DMAP (10 mol %), CH2Cl2, RT, 18 h, 89%; c) DMDO (0.09 M in
acetone, 2.0 equiv), CH2Cl2, 45 min, 45% of 17a and 26% of 17b.
b
Thermal ellipsoid representation (50% probability) of 17a: C
(white); O (red); N (blue).
a
measured via HPLC quantification of the product GMM. All
compounds were found to be essentially inactive against both
EcMetAP and HsMetAP1 while some were effective inhibitors
of HsMetAP2. Dose−response IC50 values against HsMetAP2,
after a 20 min preincubation of inhibitor and enzyme, were
then determined for compounds with reasonable potency
(Table 1). While the alkyl side-chain derivatives (18−21) were
poorly active against HsMetAP2, the benzyl derivatives showed
much more promising activity, with many (14, 25−28) giving
submicromolar IC50 values. Both ortho (23) and meta (24)
substitutions were tolerated, as well as a larger 2-naphthyl
substitution (25), but para substitutions (14, 26−28) appeared
the most promising. Epoxide precursors 13 and 16 were found
to be inactive against HsMetAP2, indicating that the epoxide
moiety is required for activity and suggesting that the
compounds are covalent inhibitors.
Reagents and conditions: (a) PMBN3 (1.0 equiv), CuI (5 mol %),
TTTA (5 mol %), THF, RT, 79%; (b) Pd(OAc)2 (10 mol %), n-
Bu4NBr (1.0 equiv), Na2CO3 (2.5 equiv), MeCN/H2O (9:1), 70 °C,
24 h, 52%; (c) DMDO (0.09 M in acetone, 2.0 equiv), CH2Cl2, RT,
18 h, 69%.
13 from 12 via Heck cyclization resulted either in no reaction
or complete consumption of the starting material with very low
yields of the desired product.41 After extensive experimentation,
we found that Jeffery’s phase transfer method under rigorously
degassed conditions gave a reliable 45−52% yield of 13 (the
isomeric aromatized hydroxybenzotriazole was isolated in 15%
yield).42 Epoxidation of 13 with DMDO cleanly provided
racemic epoxide 14, an analog of the naturally occurring
fumagillin congeners ovalicin and RK-805.
Alternatively, DIBAL reduction of ketone 13 gave alcohol 15
(Scheme 3). This alcohol could be acylated to, for example, 16
and subsequently oxidized with DMDO to give the separable
racemic epoxides 17a/17b in equal amounts (determined by
1H NMR of the crude reaction mixture), indicating essentially
no diastereoselectivity for the epoxidation reaction. The relative
stereochemistry of the two compounds was confirmed via
crystallographic analysis of single crystals grown from 17a,
which revealed this diastereomer to have the fumagillin-like
“anti” relationship of the oxirane and carboxylate oxygens at C3
and C6, respectively (fumagillin numbering).
Biochemical MetAP2 Inhibition. We prepared a series of
alkyl and substituted benzyl triazoles in analogy to ketoepoxide
14 and measured their inhibitory activity against E. coli MetAP
(EcMetAP), human type 1 MetAP (HsMetAP1), and human
type 2 MetAP (HsMetAP2) at 25 μM inhibitor concentration.43
Using the tetrapeptide substrate MGMM, inhibition was
We then tested a series of compounds in the 17a/17b series
and found that many of these substances were also selective
HsMetAP2 inhibitors (Table 2, Supporting Information Table
3). Interestingly, compounds with a fumagillin-like “anti”
relationship between heteroatoms at C3 and C6 (17a, 29a,
30a, 31a) were significantly more active than their “syn”
counterparts (17b, 29b, 30b, 31b, respectively). The latter were
quite poor inhibitors, giving circumstantial evidence that the
binding mode is similar to fumagillin. Compounds with a
hydrolytically more stable carbamate, instead of ester, linkage
(30a, 31a, 32−35) were generally highly active compounds
with 31a having an IC50 value of 220 nM. Compound 35 was
also a submicromolar inhibitor, indicating that the carbamate
linkage could be modulated. The most active substance, 31a,
was separated into its component enantiomers, (+)-31a and
(−)-31a, via chiral chromatography, and (+)-31a was found to
be entirely responsible for the biochemical activity of the
C
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
a
Table 1. Biochemical Activities of Ketoepoxide Inhibitors
a
b
Each data set was measured in triplicate (mean standard deviation); n.i. = not inhibited (<15% inhibition value). Values are reported as percent
inhibition at 25 μM inhibitor concentration with respect to a solvent control.
a
Table 2. Biochemical Activities of C6-Stereogenic Epoxide Inhibitors
a
b
Each data set was measured in triplicate (mean standard deviation). 33 was an unseparable 1:1 mixture of diastereomers.
22
racemate. As expected, the component enantiomers of 31b,
(+)-31b, and (−)-31b were found to be inactive.
optical rotation ([α]D +46.5°). Moreover, this substance was
found to have an IC50 value against MetAP2 and a chiral HPLC
retention time identical to the chromatographically separated
active enantiomer (+)-31a. This indicates that the active
enantiomer and fumagillin share the same absolute config-
uration at C3 and C6. The optical rotation of (3S,6R)-31b was
Asymmetric Synthesis of (+)-31a. In order to determine
the absolute configuration of (+)-31a, an enantioselective
synthesis was developed as shown in Scheme 4. Starting with
known protected alcohol (R)-36 (>95% ee),37 iodination
followed by copper catalyzed Huisgen cyclization gave triazole
(R)-37. Heck cyclization under Jeffery conditions with (R)-37
was much slower than with 12 and its congeners, and while it
could be driven to completion over 96 h with repeated catalyst
loadings, yields were always well below 50%. Ultimately,
“solventless” conditions were found to be superior, the reaction
completing in a few hours with lower catalyst loading to give
(R)-38 in a yield of 47%.44 Silyl deprotection of (R)-38 gave
alcohol (R)-15, which could be converted in two steps to
carbamate (R)-39 and, ultimately, to the epoxides (3R,6R)-31a
and (3S,6R)-31b. The first product was found to have a positive
22
negative ([α]D −12.5°), thereby defining the absolute
stereochemistry of all four stereoisomers (Figure 4).
Crystal Structure Confirms Covalent Inhibition. As
stated above, the requirement of the epoxide moiety for
inhibitory activity is suggestive that the compounds are, like
fumagillin, covalent inhibitors. Time dependent inhibition
studies with 31a, 32, and 33 supported this hypothesis
(Supporting Information Figure 3). Moreover, the fact that
structurally similar substances exhibited very different bio-
chemical activities (e.g., ( )-31a vs ( )-31b and (+)-31a vs
(−)-31a) suggested that the substances are quiescent affinity
D
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
a
Scheme 4. Asymmetric Synthesis of (+)-31a
Figure 5. (A) X-ray crystal structure of (+)-31a at 1.75 Å resolution
(R = 16.0%, Rfree = 18.9%) covalently bound to HsMetAP2 at His231
(PDB code 5CLS). (B) Overlay of (+)-31a (tan carbon framework)
and fumagillin (PDB code 1BOA, light blue carbon framework) bound
to MetAP2.
analogy to fumagillin, and as per our design plan, His231, in the
active site of the enzyme, was found to be covalently linked to
(+)-31a at the (formerly) epoxide C2 carbon. The
4-methoxybenzyl group of (+)-31a fills the same hydrophobic
pocket as fumagillin’s prenyl side chain, and the carbamate side
chain is solvent exposed, like fumagillin’s polyene side chain.
Soaking MetAP2 crystals with (−)-31a at a concentration of 50
mM never produced any detectable electron density associated
with a bound inhibitor.
While the protein-inhibitor structures of (+)-31a and
fumagillin bound to MetAP2 show some similarities, an overlay
of the two structural models also reveals differences (Figure 5,
panel B). The imaginary planes defined by the central six-
membered rings of the two inhibitors intersect each other at an
angle of ∼30°. In addition, while fumagillin’s cyclohexane ring
adopts a chair configuration with the epoxide-derived hydroxyl
group oriented axially, the six-membered ring of (+)-31a adopts
a half-chair conformation with the analogous hydroxyl group
oriented in a pseudoequatorial position. Interestingly, while
both fumagillin and (+)-31a fill the same hydrophobic pocket,
their side chains take different “trajectories” to reach the back of
the pocket. This indicates that there may be room to “grow”
our inhibitors so as to improve their potencies further.
a
Reagents and conditions: (a) NIM (1.2 equiv), CuI (5 mol %), THF,
RT, 18 h, 95%; (b) PMBN3 (1.1 equiv), CuI (5 mol %), TTTA (5 mol
%), THF, RT, 98%; (c) Pd(OAc)2 (10 mol %), n-Bu4NBr (10.0 equiv)
n-Bu4NOAc (5.0 equiv), 100 °C, 2 h, 47%; (d) HF·Pyr, pyridine,
THF, RT, 3 h, 97%; (e) p-nitrophenyl chloroformate (2.0 equiv), Et3N
(2.0 equiv), DMAP (0.3 equiv), CH2Cl2, RT, 1.5 h, 62%; (f) i-PrNH2
(5.5 equiv), i-Pr2NEt (2.9 equiv), EtOH, RT, 20 h, 92%, (g) DMDO
(0.09 M in acetone, 1.6 equiv), CH2Cl2, 36% of (+)-31a and 44% of
(−)-31b.
In contrast to (−)-31a, individually soaking MetAP2 crystals
with (−)-31b and (+)-31b gave well-defined X-ray structures
showing electron density for each compound. The substance
(−)-31b (Figure 6, panel A; PDB code 5D6E) has a pose
similar to (+)-31a in the enzymatic pocket (Figure 6, panel B)
but was found to be covalently bound via Glu364, an amino
acid that normally chelates one of the two cobalt atoms in the
active site. This can be understood because (−)-31b is epimeric
to (+)-31a at C3 and, upon binding to MetAP2, presents the
oxirane oxygen, instead of the C2 methylene, toward His231.
While this orientation precludes SN2 ring-opening by His231,
Glu364, on the opposite side of the enzymatic pocket, is able to
do so. Compound (+)-31b, epimeric with (+)-31a at C6, was
found covalently bound to His231 but adopts a half-chair
conformation that is ring-flipped relative to (+)-31a (Figure 6,
panel C; Figure 6, panel D, PDB code 5D6F). Presumably, in
the case of (−)-31b and (+)-31b, the “forcing” conditions of
the crystallization experiments (high inhibitor concentrations,
covalent mechanism, long reaction times) drive the reaction of
the inhibitors with MetAP2, even though their noncovalent
affinity may be relatively low. Nevertheless, in spite of the low
Figure 4. Depiction of the four stereoisomers of compound 31.
labels with some reversible binding to the active site and only
make a covalent bond in a proximity-promoted second step.45
This supposition was supported by solution of the X-ray crystal
structure of one molecule of (+)-31a bound to HsMetAP2 at a
resolution of 1.75 Å (Figure 5, panel A; PDB code 5CLS). In
E
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
beloranib, are poor pharmacokinetics, in particular low
metabolic stability. In order to gain insight into the
pharmacokinetic properties of our compounds, stability tests
in mouse plasma were conducted in comparison to beloranib.
While the tested ketone and ester containing substances (14,
28, 29a) had relatively poor stability profiles, the carbamates
(30a, ( )-31a) had half-lives that were ∼2.0 times longer than
beloranib (Figure 8, panel A). In another experiment, beloranib
and (+)-31a were tested in mouse liver microsomes. Beloranib
was found to have a half-life of less than 5 min, while (+)-31a
had a significantly longer half-life of 40 min (Figure 8, panel B).
Conclusion. The development of MetAP2 inhibitors has
been, and continues to be, an area of intense research. While
interest in MetAP2 inhibitors was initially focused on oncology,
the potential for MetAP2 inhibitors as drugs targeting obesity
has recently expanded their therapeutic potential. MetAP2
inhibitors which are based on the fumagillin substructure have
been studied extensively, and their irreversible mode of action
distinguishes them from reversible inhibitor scaffolds, which
generally rely on metal chelating functional groups for
potency.14
Figure 6. (A) X-ray crystal structure of (−)-31b at 1.49 Å resolution
(R = 11.0%, Rfree = 15.6%) covalently bound to HsMetAP2 at Glu364
(PDB code 5D6E); (B) X-ray crystal structure overlay of (+)-31a
(tan) and (−)-31b (blue); (C) X-ray crystal structure of (+)-31b at
1.55 Å resolution (R = 14.7%, Rfree = 18.5%) covalently bound to
HsMetAP2 at His231 (PDB code 5D6F); (D) X-ray crystal structure
overlay of (+)-31a (tan) and (+)-31b (blue).
While nature has evolved fumagillin to be a near perfect
inhibitor in terms of potency and selectivity, attempts to use
fumagillin as a drug in clinical trials have failed due to
suboptimal ADME properties and poor pharmacokinetics.
Campaigns to optimize these properties through modulation
of the fumagillin C6 side chain have produced substances with
improved profiles (e.g., 3 and 4), but these compounds retain
liabilities inherent to the fumagillol scaffold.
Here, we introduce a new MetAP2 inhibitor scaffold that,
while inspired by the structure of fumagillin, lacks the fumagillol
core and, therefore, may have the potential for superior
pharmacological properties. Our successful route to the
compounds enabled the synthesis of a small but diverse set
of compounds. Initial SAR analysis revealed that benzyl
substitution on the triazole ring was optimal for effective
MetAP2 inhibition and low activity against two other MetAP
subtypes. Substances that are stereogenic at both C3 and C6
must display an “anti” relationship between the C3 oxygen and
the C6 side chain; the “syn” diasteromers are completely
inactive. Moreover, only one enantiomer of such an “anti”-
configured compound is responsible for the activity of the
inhibitors, and the absolute configuration of this enantiomer
matches that of the natural product fumagillin.
The X-ray crystal structure of (+)-31a bound to MetAP2
reveals a covalent bond to His231, a mode of action identical to
fumagillin. The binding pose is also similar to fumagillin, but
differences between the two structures indicate that there may
be additional space in the hydrophobic pocket to design further
derivatives with enhanced binding. An expanded SAR study of
allowed substitution patterns on the benzyl side chain could
help uncover more potent substances.
The differential biochemical activities of the stereoisomers of
compounds 31a and 31b were replicated in a cellular context.
In both HUVECs and HT1080 cells, the “anti”-configured
substance 31a was more potent than 31b. Similarly, the
enantiomer (+)-31a was far more active in both cell lines than
(−)-31a. An increase in the unprocessed form of 14-3-3-γ was
observed in a dose-dependent manner for (+)-31a in HUVECs,
while no change was observed with (−)-31a. Finally, carbamate
substituted analogs 30a and 31a showed longer half-lives,
relative to beloranib, in plasma and microsomal stability assays.
biochemical activity of (+)-31b and (−)-31b, no nucleophilic
solvent-exposed residues (e.g., Lys427 or Cys290) other than
His231 were modified, further underlining the selective binding
behavior of this compound class.
Cellular and Biochemical Data Correlate. The epoxides
( )-31a, ( )-31b, (+)-31a, and (−)-31a were tested for their
ability to arrest the growth of human umbilical vein endothelial
cells (HUVECs) and were found to have activities that
correlated with their biochemical MetAP2 activities. Whereas
( )-31a inhibited the growth of HUVECs with an EC50 of 165
nM, ( )-31b was much less effective, with an EC50 of 6 μM
(Figure 7, panel A). As expected, the active enantiomer (+)-31a
potently affected the growth of HUVECs with an EC50 of 94
nM, while (−)-31a showed no effect up to 10 μM. Similar
observations were made when HT1080 cells, a cell line known
to be sensitive to MetAP2 inhibition, were treated with these
four compounds (Figure 7, panel B).46 Consistent with
fumagillin-based inhibitors, an incomplete reduction of cell
number was observed in both cell lines. In HUVECs, it is well-
known that, at low concentrations, fumagillin-based inhibitors
have a cytostatic effect; i.e., cells no longer divide but remain
viable. Only at much higher concentrations do the inhibitors
become cytotoxic.47 A similar result was found when beloranib
was tested against HT1080 cells (Supporting Information
Figure 4).
The adaptor protein 14-3-3-γ has been shown to be a useful
cellular marker for MetAP2 activity.48 Proteolytic removal of
the N-terminal methionine from 14-3-3-γ was strongly altered
by (+)-31a in HUVECs, whereas little effect was observed with
(−)-31a (Figure 7C). Together with the biochemical and
cellular data, these results are consistent with the observed
cellular phenotype resulting from on-target MetAP2 inhibition.
Carbamates Are More Stable than Beloranib. The
primary clinical drawback of fumagillin-based substances, like
F
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
Figure 7. Dose−response curves for HUVECs (A) or HT1080 cells (B) when treated with inhibitor. Values represent triplicate means SD. (C)
Western blots for unprocessed 14-3-3-γ (only proteins with an N-terminal methionine are stained) versus total 14-3-3-γ levels when treated with
inhibitor at the indicated concentrations or DMSO control.
the potential for good pharmacokinetic properties, their cellular
activity remains lower than fumagillin-based derivatives, which
have picomolar IC50 growth-inhibition values against HUVECs.
The further development of this class of compounds would,
therefore, be enabled by the discovery of analogs that close the
gap with fumagillin-based substances in terms of cellular
potency. Additionally, a more thorough analysis of ADME
properties and, ultimately, pharmacokinetics in animal models
will be necessary to evaluate their real clinical potential.
In summary, we have reported the synthesis and preliminary
biological evaluation of a new chemical class of spiroepoxide
tetrahydrobenzotriazoles, inspired by the natural product
fumagillin. These compounds are potent and selective
inhibitors of MetAP2, as demonstrated in biochemical
Figure 8. (A) Compound stability in mouse plasma. (B) Compound
stability in mouse microsomes. (A and B) Each data set represents the
mean of two independent measurements made with both positive and
negative control substances.
enzymatic assays. Their mode of action (covalent modification
of the active site His231 residue) was confirmed to be identical
to fumagillin via X-ray crystallography, and they are effective
HUVEC growth inhibitors. With metabolic stabilities as good
as or superior to the clinical candidate beloranib, this novel
class of inhibitors potentially provides an alternative to
fumagillin-based inhibitors.
While these data indicate that the compounds exert their
cellular phenotype via irreversible MetAP2 inhibition and have
G
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
cell attachment. Compounds ( )-31a, ( )-31b, (+)-31a, and (−)-31a
were dissolved in DMSO as 10 mM stock solutions. Stock solutions
were diluted 1:1000 in cell-type specific medium to prepare the
treatment starting concentration (10 μM; 0.1% DMSO). This starting
concentration was diluted 1:2 in several diluting steps with cell-type
specific medium to obtain 16 different concentrations ranging from 10
μM to 0.3 nM. The seeding medium of the cells was aspirated and cells
were incubated with 100 μL/well (96-well plates) of the respective
drug solution. As a control, cells were treated with 0.1% DMSO. Each
drug concentration was performed in technical triplicates. The
compound treatment was repeated every 24 h on three consecutive
days until cell viability was measured. For Western immunoblotting,
HUVECs were treated on three consecutive days with six different
compound concentrations [(+)-31a and (−)-31a] prepared from the
10 mM stock solution (highest concentration: 5 μM; 0.05% DMSO,
lowest concentration: 1 nM). The seeding medium of the cells was
aspirated, and cells were incubated with 1 mL/well (6-well plates) of
the respective drug solution. As a control, cells were treated with
0.05% DMSO.
MTT Assay. Cell viability was analyzed 72 h after the first
compound treatment using the MTT (3-[4,5-dimethylthiazol-2-yl]-
2,5-diphenyltetrazolium bromide) assay according to the manufac-
turer’s instructions (Sigma-Aldrich). In brief, cells were incubated with
the MTT solution (0.5 mg mL−1 in medium) for 1 h at 37 °C.
Afterward, the solution was removed and the tetrazolium salt was
resolved in 100 μL of DMSO/ethanol solution (1:2). Colorimetric
measurement was performed at 570 nm using the FLUOstar Optima
multidetection microplate reader (BMG Labtech). Data are presented
METHODS
■
Synthetic Chemistry. Reaction mixtures were magnetically stirred
in oven-dried glassware under a blanket of argon. External bath
temperatures were used to record all reaction mixture temperatures.
Reagents were purchased at the highest level of commercial quality
and used without purification unless otherwise noted. THF was freshly
distilled from sodium/benzophenone ketyl prior to use. All other dry
solvents were bought from Aldrich and used without further
purification. N-iodomorpholine (NIM) was prepared according to
the procedure of Hein et al.40 Beloranib was prepared according to a
described method.49 The dimethyldioxirane (DMDO) solution was
prepared and stored according to the procedure described by Adam et
al.50 The concentration of DMDO was determined via iodometric
titration.
Enzyme Production for Biochemical Assays. C-terminal poly-
His-tagged EcMetAP was obtained by expression in E. coli BL21(DE3)
cells (Novagen) using an Arg-175-Gln mutant kindly provided by Prof.
W.T. Lowther and Prof. B. Matthews. A detailed description of the
preparation has been previously described.51 The gene for HsMetAP1
was synthesized in an optimized nucleotide sequence for E. coli
expression (Geneart) and cloned into the pET-28a(+) vector
(Novagen) using BamH1 and Sac1 cloning sites. The enzyme with a
N-terminal His-tag was overexpressed in E. coli BL21(DE3)
(Novagen) for 5 h at 37 °C. The enzyme was isolated and purified
by nickel affinity chromatography (Chelating Sepharose Fast Flow,
Amersham Biosciences; buffer: 50 mM Hepes pH 7.9, 500 mM KCl,
15 mM methionine, 10−250 mM imidazole gradient; removal of metal
ions by adding EDTA pH 8.0 to a final concentration of 5 mM). The
buffer was exchanged (storage buffer: 25 mM Hepes pH 7.9, 150 mM
KCl, 15 mM methionine) and the enzyme stored at −80 °C. The
HsMetAP2 was produced in Sf9 cells using the Bac-to-Bac Baculovirus
Expression System (Invitrogen). A detailed description of the cloning,
expression, and isolation procedures has been previously described.43
Biochemical Assays. Stock solutions for all assays were prepared
in DMSO (10 mM) and stored at −20 °C. Assays were performed in
96 well plates (Greiner) at a total reaction volume of 100 μL. First,
enzyme (final concentration: 50 nM) was incubated with inhibitor (25
μM) or positive control (DMSO, 0.25%) in a reaction buffer (final
concentration: 100 μM CoCl2·6H2O, 100 mM NaCl, 0.075% bovine
serum albumine, 50 mM Tris pH 7.5) for 20 min at 37 °C. Then, the
substrate Met-Gly-Met-Met was added to a final concentration of 400
μM. After an incubation period of 15 min (HsMetAP1), 20 min
(EcMetAP), or 60 min (HsMetAP2) at 37 °C, the enzyme reaction
was stopped by adding 10 μL 4% trifluoroacetic acid (TFA), and the
plate was centrifuged at 1000g for 10 min. Detection of the product
Gly-Met-Met was performed by HPLC: C18 column (Dr. Maisch,
ReproSil-Pur 120 ODS-3, 3 μm, 50 × 2 mm). An 80 μL injection
volume was eluted with mobile phases A (H2O and 0.1% TFA, filtered,
degassed) and B (acetonitrile and 0.1% TFA, degassed) with the
following profile: 10% → 58% B linear gradient (4 min), 95% B (2
min), 10% B (2 min, re-equilibration). The product was quantified by
UV absorption at 214 nm. Percentage inhibition was calculated as the
ratio of Gly-Met-Met produced in the inhibited to uninhibited
(positive control). IC50 values were determined by measuring percent
inhibition at a range of inhibitor concentrations and fitting to a
logarithmic dose−response curve.
as the mean
standard deviation. Dose−response curves were
established by plotting growth (values normalized against DMSO
control) against concentration (log) of compounds using the statistics
software GraphPad Prism 6. The EC50 was determined by
computerized curve fitting using nonlinear regression.
Protein Sample Preparation and Western Immunoblotting.
Cells were harvested 72 h after the first drug treatment. Total cell
protein extracts were prepared by lysing cells in 20 μL of cell lysis
buffer (New England BioLabs) supplemented with Protease inhibitor
mix (Serva Electrophoresis). Protein fractions were quantified (280
nm) and resuspended (80 μg protein/sample) in Laemmli buffer for
immunoblotting. Protein extracts were separated on a 15% sodium
dodecyl sulfate polyacrylamide gel electrophoresis and electro-
transferred to a nitrocellulose membrane. 14-3-3-γ (unprocessed
form; 1:1500; Novus Biologicals, Littleton CO, USA), 14-3-3-γ (total;
1:500; Cell Signaling Technology), and actin (1:10000; MP
Biomedicals) antibodies were diluted in TBST (Tris-buffered saline/
Tween) supplemented with 5% milk powder and incubated at 4 °C
overnight. The appropriate secondary antibody (LI-COR Biosciences)
was applied [IRDye 800CW Donkey Anti-Mouse IgG (H+L) -
1:10 000 for anti-14-3-3-γ unprocessed form; IRDye 800CW Donkey
Anti-Rabbit IgG (H+L) − 1:10 000 for anti-14-3-3-γ total; IRDye
680LT Donkey Anti-Mouse IgG (H+L) − 1:20 000 for antiactin] at
RT for 1 h. Visualization was performed by enhanced infrared
fluorescent with the Odyssey Sa Infrared Imaging System (LI-COR
Biosciences).
Enzyme Production for Crystallization. A cDNA fragment
corresponding to Gly108-Tyr478 was cloned into pFastBacHTa vector
containig an N-terminal His-Tag followed by a TEV protease cleavage
site. pFastBacHTa with the cloned HsMetAP2 fragment was then used
for transformation of DH10Bac E. coli cells to generate the
recombinant bacmid, which was then used for the insect cells
transfection and production of the recombinant baculovirus for the
further protein expression. A cell pellet from 1 to 2 L of the insect cells
was lysed in 40 mM Hepes (pH 7.4), 150 mM NaCl, 10% glycerol, 20
mM imidazole, +protease inhibitors, and DNase. The lysate was
centrifuged at 48 000g for 45 min. The supernatant after filtering was
applied to a HisTrap column equilibrated with 40 mM Hepes (pH
7.4), 150 mM NaCl, 10% glycerol, and 20 mM imidazole. The column
was washed with several volumes of 40 mM Hepes (pH 7.4), 150 mM
NaCl, 10% glycerol, and 20 mM imidazole, and additionally with 40
mM Hepes (pH 7.4), 1 M NaCl, 10% glycerol, and 20 mM imidazole.
Cell Culture and Drug Treatment. Human umbilical vein
endothelial cells (HUVECs; C-12203) were obtained from PromoCell
GmbH and cultured in Endothelial Cell Growth Medium (C-22010;
PromoCell GmbH). The fibrosarcoma cell line HT1080 was kindly
provided by Dr. Marcus Renner (Institute of Pathology, University
Hospital Heidelberg, Germany) and maintained in DMEM medium
(Sigma-Aldrich) supplemented with 10% fetal bovine serum, 1%
penicillin/streptomycin, 1 mM L-glutamine, and 0.5% Na-Pyruvate.
Cells were cultured at 37 °C in a 5% CO2 atmosphere. For viability
assays, cells were seeded in 96-well plates (HUVECs: 2000 cells/well;
HT1080: 2500 cells/well). For Western immunoblotting, HUVEC
cells were seeded in six-well plates (150 000 cells/well). Treatment of
cells with MetAP2 inhibitors was started 24 h after seeding to allow
H
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
The protein was eluted with a linear gradient from 20 to 500 mM
imidazole in the same buffer. Fractions with MetAP2 were combined
and dialyzed against 20 mM Hepes (pH 7.4), 150 mM NaCl, 5%
glycerol O/N at 4 °C with TEV protease added to remove the HisTag
(1:50 ratio). TEV protease, uncleaved MetAP2, and cleaved HisTag
were removed by applying everything on the HisTrap column. The
sample containing cleaved MetAP2 was concentrated to 5 mL and
applied on a Superdex200 16/60 column equilibrated with 10 mM
Hepes (pH 7.4), 150 mM NaCl and 10% glycerol. The selected
fraction from the size exclusion step was then pooled and concentrated
to 22 mg mL−1. The protein sample was supplemented with CoCl2 to
a final concentration of 1 mM.
Enzyme Crystallization. The native crystals were obtained from
similar conditions to those previously reported by Liu et al.: 50 mM
sodium citrate buffer at pH 5.4 and 15% (v/v) t-butanol.19
Crystallization was performed using the hanging-drop vapor-diffusion
method at 292 K in 24-well plates. The crystals appeared after a few
days and were then used for soaking with (−)-31a, (+)-31a, (−)-31b,
and (+)-31b: 0.5 μL of 50 mM inhibitor in 50% (v/v) DMSO, 50 mM
sodium citrate buffer at pH 5.4, and 15% (v/v) t-butanol were added
to the crystallization drop of 2−3 μL volume containing MetAP2
crystals.
Structure Determination and Refinement. Crystals were
mounted in a nylon fiber loop and flash-cooled to 100 K in liquid
nitrogen. The cryoprotection was performed for 2 s in reservoir
solution complemented with 20% (v/v) glycerol. Diffraction data for
MetAP2:(+)-31a and MetAP2:(−)-31a were collected on the ESRF
ID23−2 beamline; those for MetAP2:(−)-31b were collected on the
ID23−1 beamline and those for MetAP2:(+)-31b on the ESRF ID29
beamline. All data sets were indexed and integrated using XDS52 and
scaled using SCALA.53,54 Intensities were converted to structure-factor
amplitudes using the program TRUNCATE.55 The structure of
MetAP2 with all three compounds was solved by molecular
replacement using the native MetAP2 structure published by Liu et
al. (PDB: 1BN5).6 Model rebuilding was performed in COOT.56
(+)-31a, (−)-31b, and (+)-31b were modeled manually. The
refinement was done in REFMAC557 using the maximum-likelihood
target function. The stereochemical analysis of the final model was
done in PROCHECK58 and MolProbity.59
ACKNOWLEDGMENTS
■
We thank N. De Vries, E. Ghebrai, J. Hummel-Eisenbeiss, B.
Kraft, J. Lohbeck, H.-H. Schrenk, G. Schwebel, U. Wagner
(DKFZ), S. Calcagno, A. Marschner, M. Wacker (IPMB), and
V. Roman (HMGU) for technical assistance. We thank B. Hull
and K. Klika for assistance with NMR spectroscopy and S.
Barroso for suggesting the solvent-free Heck conditions. We
thank the Hashmi group for IR spectra, the Helmchen group
for use of their polarimeter, and Frank Rominger for the X-ray
data for compound 17a (all from the Organic Chemistry
Institute, Heidelberg University). We acknowledge the use of
the X-ray crystallography platform of the Helmholtz Zentrum
Munchen. This work was supported by the DKFZ Intramural
̈
Fund (A.K.M and C.D.K.), the DKFZ Development Fund
(A.K.M. and N.G.), and Deutsche Krebshilfe (C.D.K.).
Financial support from the Helmholtz Drug Initiative and the
German Consortium for Translational Research (DKTK) is
gratefully acknowledged. Protein X-ray graphics and analyses
were performed with the UCSF Chimera package. Chimera is
developed by the Resource for Biocomputing, Visualization,
and Informatics at the University of California, San Francisco
(supported by NIGMS P41-GM103311).
REFERENCES
■
(1) Li, J. W. H., and Vederas, J. C. (2009) Drug discovery and natural
products: end of an era or an endless frontier? Science 325, 161−165.
(2) Ingber, D., Fujita, T., Kishimoto, S., Sudo, K., Kanamaru, T.,
Brem, H., and Folkman, J. (1990) Synthetic analogues of fumagillin
that inhibit angiogenesis and suppress tumour growth. Nature 348,
555−557.
(3) Marui, S., Itoh, F., Kozai, Y., Sudo, K., and Kishimoto, S. (1992)
Chemical modification of fumagillin. I. 6-O-acyl, 6-O-sulfonyl, 6-O-
alkyl, and 6-O-(N-substituted-carbamoyl)fumagillols. Chem. Pharm.
Bull. 40, 96−101.
(4) Marui, S., and Kishimoto, S. (1992) Chemical modification of
fumagillin. II. 6-Amino-6-deoxyfumagillol and its derivatives. Chem.
Pharm. Bull. 40, 575−579.
(5) Marui, S., Yamamoto, T., Sudo, K., Akimoto, H., and Kishimoto,
S. (1995) Chemical modification of fumagillin. III. Modification of the
spiro-epoxide. Chem. Pharm. Bull. 43, 588−593.
(6) Turk, B. E., Su, Z., and Liu, J. O. (1998) Synthetic analogues of
TNP-470 and ovalicin reveal a common molecular basis for inhibition
of angiogenesis and immunosuppression. Bioorg. Med. Chem. 6, 1163−
1169.
(7) Han, C. K., Ahn, S. K., Choi, N. S., Hong, R. K., Moon, S. K.,
Chun, H. S., Lee, S. J., Kim, J. W., Hong, C. I., Kim, D., Yoon, J. H.,
and No, K. T. (2000) Design and synthesis of highly potent fumagillin
analogues from homology modeling for a human MetAP-2. Bioorg.
Med. Chem. Lett. 10, 39−43.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acschem-
bio.5b00755.
X-ray crystallographic data for 17a (CIF)
Synthetic chemistry procedures, compound character-
1
ization data, stability measurements, X-ray data, H and
13C NMR spectra (PDF)
Accession Codes
The crystallographic data for 17a has been deposited with the
Cambridge Crystallographic Data Centre as CCDC 894701.
The atomic model and structure factors for the MetAP2:
(+)-31a, MetAP2:(−)-31b, and MetAP2:(+)-31b complexes
have been deposited with the Protein Data Bank as entries
5CLS, 5D6E, and 5D6F, respectively.
(8) Fardis, M., Pyun, H.-J., Tario, J., Jin, H., Kim, C. U., Ruckman, J.,
Lin, Y., Green, L., and Hicke, B. (2003) Design, synthesis and
evaluation of a series of novel fumagillin analogues. Bioorg. Med. Chem.
11, 5051−5058.
(9) Zhou, G., Tsai, C. W., and Liu, J. O. (2003) Fumagalone, a
Reversible Inhibitor of Type 2 Methionine Aminopeptidase and
Angiogenesis. J. Med. Chem. 46, 3452−3454.
(10) Pyun, H.-J., Fardis, M., Tario, J., Yang, C. Y., Ruckman, J.,
Henninger, D., Jin, H., and Kim, C. U. (2004) Investigation of novel
fumagillin analogues as angiogenesis inhibitors. Bioorg. Med. Chem.
Lett. 14, 91−94.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: aubry.miller@dkfz.de.
(11) Lu, J., Chong, C. R., Hu, X., and Liu, J. O. (2006) Fumarranol, a
Rearranged Fumagillin Analogue That Inhibits Angiogenesis in Vivo. J.
Med. Chem. 49, 5645−5648.
(12) Lee, H. W., Cho, C. S., Kang, S. K., Yoo, Y. S., Shin, J. S., and
Ahn, S. K. (2007) Design, Synthesis, and Antiangiogenic Effects of a
Author Contributions
⊥M. Morgen and C. Jost contributed equally to this work.
̈
Notes
The authors declare no competing financial interest.
I
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
Series of Potent Novel Fumagillin Analogues. Chem. Pharm. Bull. 55,
1024−1029.
Gradhand, U., Griffin, S. T., Hill, S., Labenski, M. T., Morgan, B. A.,
O’Donovan, G., Prasad, K., Skinner, S., Taghizadeh, N., Thompson, C.
D., Wakefield, J., Westlin, W., and White, K. F. (2013) Metabolites of
PPI-2458, a selective, irreversible inhibitor of methionine amino-
peptidase-2: structure determination and in vivo activity. Drug Metab.
Dispos. 41, 814−826.
(29) Arico-Muendel, C. C., Blanchette, H., Benjamin, D. R., Caiazzo,
T. M., Centrella, P. A., DeLorey, J., Doyle, E. G., Johnson, S. R.,
Labenski, M. T., Morgan, B. A., O’Donovan, G., Sarjeant, A. A.,
Skinner, S., Thompson, C. D., Griffin, S. T., Westlin, W., and White, K.
F. (2013) Orally Active Fumagillin Analogues: Transformations of a
Reactive Warhead in the Gastric Environment. ACS Med. Chem. Lett.
4, 381−386.
(30) Baldwin, J. E., Bulger, P. G., and Marquez, R. (2002) Fast and
efficient synthesis of novel fumagillin analogues. Tetrahedron 58,
5441−5452.
(13) Arico-Muendel, C. C., Benjamin, D. R., Caiazzo, T. M.,
Centrella, P. A., Contonio, B. D., Cook, C. M., Doyle, E. G., Hannig,
G., Labenski, M. T., Searle, L. L., Lind, K., Morgan, B. A., Olson, G.,
Paradise, C. L., Self, C., Skinner, S. R., Sluboski, B., Svendsen, J. L.,
Thompson, C. D., Westlin, W., and White, K. F. (2009) Carbamate
Analogues of Fumagillin as Potent, Targeted Inhibitors of Methionine
Aminopeptidase-2. J. Med. Chem. 52, 8047−8056.
(14) Yin, S. Q., Wang, J. J., Zhang, C. M., and Liu, Z. P. (2012) The
development of MetAP-2 inhibitors in cancer treatment. Curr. Med.
Chem. 19, 1021−1035.
(15) Joharapurkar, A. A., Dhanesha, N. A., and Jain, M. R. (2014)
Inhibition of the methionine aminopeptidase 2 enzyme for the
treatment of obesity. Diabetes, Metab. Syndr. Obes.: Targets Ther. 7,
73−84.
(16) Griffith, E. C., Su, Z., Niwayama, S., Ramsay, C. A., Chang, Y.-
H., and Liu, J. O. (1998) Molecular recognition of angiogenesis
inhibitors fumagillin and ovalicin by methionine aminopeptidase 2.
Proc. Natl. Acad. Sci. U. S. A. 95, 15183−15188.
(17) Griffith, E. C., Su, Z., Turk, B. E., Chen, S., Chang, Y.-H., Wu,
Z., Biemann, K., and Liu, J. O. (1997) Methionine aminopeptidase
(type 2) is the common target for angiogenesis inhibitors AGM-1470
and ovalicin. Chem. Biol. 4, 461−471.
(18) Sin, N., Meng, L., Wang, M. Q. W., Wen, J. J., Bornmann, W. G.,
and Crews, C. M. (1997) The anti-angiogenic agent fumagillin
covalently binds and inhibits the methionine aminopeptidase, MetAP-
2. Proc. Natl. Acad. Sci. U. S. A. 94, 6099−6103.
(19) Liu, S., Widom, J., Kemp, C. W., Crews, C. M., and Clardy, J.
(1998) Structure of Human Methionine Aminopeptidase-2 Com-
plexed with Fumagillin. Science 282, 1324−1327.
(20) Yeh, J.-R. J., Mohan, R., and Crews, C. M. (2000) The
antiangiogenic agent TNP-470 requires p53 and p21CIP/WAF for
endothelial cell growth arrest. Proc. Natl. Acad. Sci. U. S. A. 97, 12782−
12787.
(21) Zhang, Y., Griffith, E. C., Sage, J., Jacks, T., and Liu, J. O. (2000)
Cell cycle inhibition by the anti-angiogenic agent TNP-470 is
mediated by p53 and p21WAF1/CIP1. Proc. Natl. Acad. Sci. U. S. A.
97, 6427−6432.
(22) Hines, J., Ju, R., Dutschman, G. E., Cheng, Y.-C., and Crews, C.
M. (2010) Reversal of TNP-470-induced endothelial cell growth arrest
by guanine and guanine nucleosides. J. Pharmacol. Exp. Ther. 334,
729−738.
(31) Jeong, B.-S., Choi, N. S., Ahn, S. K., Bae, H., Kim, H. S., and
Kim, D. (2005) Total synthesis and antiangiogenic activity of
cyclopentane analogues of fumagillol. Bioorg. Med. Chem. Lett. 15,
3580−3583.
(32) Mazitschek, R., Huwe, A., and Giannis, A. (2005) Synthesis and
biological evaluation of novel fumagillin and ovalicin analogues. Org.
Biomol. Chem. 3, 2150−2154.
(33) Rodeschini, V., Boiteau, J.-G., Van de Weghe, P., Tarnus, C., and
Eustache, J. (2004) MetAP-2 Inhibitors Based on the Fumagillin
Structure. Side-Chain Modification and Ring-Substituted Analogues. J.
Org. Chem. 69, 357−373.
(34) Rodeschini, V., de Weghe, P. V., Tarnus, C., and Eustache, J.
(2005) A simple spiroepoxide as methionine aminopeptidase-2
inhibitor: synthetic problems and solutions. Tetrahedron Lett. 46,
6691−6695.
(35) Rodeschini, V., Van de Weghe, P., Salomon, E., Tarnus, C., and
Eustache, J. (2005) Enantioselective Approaches to Potential MetAP-2
Reversible Inhibitors. J. Org. Chem. 70, 2409−2412.
(36) Miller, A.; Jost, C.; Klein C. New spiroepoxide tetrahydrobenzo-
̈
triazoles useful as MetAP-II inhibitors. PCT Pat. Appl.
WO2014044399 A1, 2014.
(37) Goncalves-Martin, M. G., Saxer, A., and Renaud, P. (2009) A
practical synthesis of (S)-cyclopent-2-enol. Synlett 2009, 2801−2802.
(38) Li, J., Yan, W., and Kishi, Y. (2015) Unified Synthesis of C1−
C19 Building Blocks of Halichondrins via Selective Activation/
Coupling of Polyhalogenated Nucleophiles in (Ni)/Cr-Mediated
Reactions. J. Am. Chem. Soc. 137, 6226−6231.
(23) Sundberg, T. B., Darricarrere, N., Cirone, P., Li, X., McDonald,
L., Mei, X., Westlake, C. J., Slusarski, D. C., Beynon, R. J., and Crews,
C. M. (2011) Disruption of Wnt planar cell polarity signaling by
aberrant accumulation of the MetAP-2 substrate Rab37. Chem. Biol.
18, 1300−1311.
(24) Datta, B., Majumdar, A., Datta, R., and Balusu, R. (2004)
Treatment of Cells with the Angiogenic Inhibitor Fumagillin Results in
Increased Stability of Eukaryotic Initiation Factor 2-Associated
Glycoprotein, p67, and Reduced Phosphorylation of Extracellular
Signal-Regulated Kinases. Biochemistry 43, 14821−14831.
(25) Bhargava, P., Marshall, J. L., Rizvi, N., Dahut, W., Yoe, J.,
Figuera, M., Phipps, K., Ong, V. S., Kato, A., and Hawkins, M. J.
(1999) A Phase I and pharmacokinetic study of TNP-470
administered weekly to patients with advanced cancer. Clin. Cancer
Res. 5, 1989−1995.
(39) Witulski, B., and Alayrac, C. (2006) Product subclass 1:1-
haloalk-1-ynes and alk-1-yn-1-ols. Sci. Synth. 24, 905−932.
(40) Hein, J. E., Tripp, J. C., Krasnova, L. B., Sharpless, K. B., and
Fokin, V. V. (2009) Copper(I)-catalyzed cycloaddition of organic
azides and 1-iodoalkynes. Angew. Chem., Int. Ed. 48, 8018−8021.
(41) Schulman, J. M., Friedman, A. A., Panteleev, J., and Lautens, M.
(2012) Synthesis of 1,2,3-triazole-fused heterocycles via Pd-catalyzed
cyclization of 5-iodotriazoles. Chem. Commun. 48, 55−57.
(42) Jeffery, T. (1996) On the efficiency of tetraalkylammonium salts
in Heck type reactions. Tetrahedron 52, 10113−10130.
(43) Altmeyer, M. A., Marschner, A., Schiffmann, R., and Klein, C. D.
(2010) Subtype-selectivity of metal-dependent methionine amino-
peptidase inhibitors. Bioorg. Med. Chem. Lett. 20, 4038−4044.
(44) Calo, V., Nacci, A., Monopoli, A., and Cotugno, P. (2009) Heck
reactions with palladium nanoparticles in ionic liquids: coupling of aryl
chlorides with deactivated olefins. Angew. Chem., Int. Ed. 48, 6101−
6103.
(45) Copeland, R. A. (2013) Evaluation of enzyme inhibitors in drug
discovery: a guide for medicinal chemists and pharmacologists, 2nd ed.,
Wiley, Hoboken, NJ.
(46) Tucker, L. A., Zhang, Q., Sheppard, G. S., Lou, P., Jiang, F.,
McKeegan, E., Lesniewski, R., Davidsen, S. K., Bell, R. L., and Wang, J.
(2008) Ectopic expression of methionine aminopeptidase-2 causes cell
transformation and stimulates proliferation. Oncogene 27, 3967−3976.
(47) Kusaka, M., Sudo, K., Matsutani, E., Kozani, Y., Marui, S., Fujita,
T., Inger, D., and Folkman, J. (1994) Cytostatic inhibition of
(26) Stadler, W. M., Kuzel, T., Shapiro, C., Sosman, J., Clark, J., and
Vogelzang, N. J. (1999) Multi-Institutional Study of the Angiogenesis
Inhibitor TNP-470 in Metastatic Renal Carcinoma. J. Clin. Oncol. 17,
2541−2545.
(27) Shin, S. J., Jeung, H.-C., Ahn, J. B., Rha, S. Y., Roh, J. K., Park, K.
S., Kim, D.-H., Kim, C., and Chung, H. C. (2010) A phase I
pharmacokinetic and pharmacodynamic study of CKD-732, an
antiangiogenic agent, in patients with refractory solid cancer. Invest.
New Drugs 28, 650−658.
(28) Arico-Muendel, C. C., Belanger, B., Benjamin, D., Blanchette, H.
S., Caiazzo, T. M., Centrella, P. A., DeLorey, J., Doyle, E. G.,
J
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Articles
endothelial cell growth by the angiogenesis inhibitor TNP-470 (AGM-
1470). Br. J. Cancer 69, 212−216.
(48) Towbin, H., Bair, K. W., DeCaprio, J. A., Eck, M. J., Kim, S.,
Kinder, F. R., Morollo, A., Mueller, D. R., Schindler, P., Song, H. K.,
van, O. J., Versace, R. W., Voshol, H., Wood, J., Zabludoff, S., and
Phillips, P. E. (2003) Proteomics-based Target Identification:
Bengamides as a new class of methionine aminopeptidase inhibitors.
J. Biol. Chem. 278, 52964−52971.
(49) Crawford, T., Reece H. A. (2012) Crystalline Solids of a MetAP-
2 Inhibitor and Methods of Making and Using Same. PCT Pat. Appl.
WO2012064838 A1.
(50) Adam, W., Bialas, J., and Hadjiarapoglou, L. (1991) A
Convenient Preparation of Acetone Solutions of Dimethyldioxirane.
Chem. Ber. 124, 2377.
(51) Lowther, W. T., McMillen, D. A., Orville, A. M., and Matthews,
B. W. (1998) 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. 95, 12153−12157.
(52) Kabsch, W. (2010) XDS. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 66, 125−132.
(53) Evans, P. (2006) Scaling and assessment of data quality. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 62, 72−82.
(54) Winn, M. D., Ballard, C. C., Cowtan, K. D., Dodson, E. J.,
Emsley, P., Evans, P. R., Keegan, R. M., Krissinel, E. B., Leslie, A. G.,
McCoy, A., McNicholas, S. J., Murshudov, G. N., Pannu, N. S.,
Potterton, E. A., Powell, H. R., Read, R. J., Vagin, A., and Wilson, K. S.
(2011) Overview of the CCP4 suite and current developments. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 67, 235−242.
(55) French, S., and Wilson, K. (1978) On the treatment of negative
intensity observations. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr.,
Theor. Gen. Crystallogr. 34, 517−525.
(56) Emsley, P., Lohkamp, B., Scott, W. G., and Cowtan, K. (2010)
Features and development of Coot. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 66, 486−501.
(57) Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997)
Refinement of macromolecular structures by the maximum-likelihood
method. Acta Crystallogr., Sect. D: Biol. Crystallogr. 53, 240−255.
(58) Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton,
J. M. (1993) PROCHECK: a program to check the stereochemical
quality of protein structures. J. Appl. Crystallogr. 26, 283−291.
(59) Chen, V. B., Arendall, W. B., 3rd, Headd, J. J., Keedy, D. A.,
Immormino, R. M., Kapral, G. J., Murray, L. W., Richardson, J. S., and
Richardson, D. C. (2010) MolProbity: all-atom structure validation for
macromolecular crystallography. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 66, 12−21.
K
DOI: 10.1021/acschembio.5b00755
ACS Chem. Biol. XXXX, XXX, XXX−XXX