Remodelling of the natural product fumagillol employing a reaction discovery approach

Article abstract of DOI:10.1038/nchem.1178

In the search for new biologically active molecules, diversity-oriented synthetic strategies break through the limitation of traditional library synthesis by sampling new chemical space. Many natural products can be regarded as intriguing starting points for diversity-oriented synthesis, wherein stereochemically rich core structures may be reorganized into chemotypes that are distinctly different from the parent structure. Ideally, to be suited to library applications, such transformations should be general and involve few steps. With this objective in mind, the highly oxygenated natural product fumagillol has been successfully remodelled in several ways using a reaction-discovery-based approach. In reactions with amines, excellent regiocontrol in a bis-epoxide opening/cyclization sequence can be obtained by size-dependent interaction of an appropriate catalyst with the parent molecule, forming either perhydroisoindole or perhydroisoquinoline products. Perhydroisoindoles can be further remodelled by cascade processes to afford either morpholinone or bridged 4,1-benzoxazepine-containing structures.

Full text of DOI:10.1038/nchem.1178

ARTICLES
PUBLISHED ONLINE: 23 OCTOBER 2011 | DOI: 10.1038/NCHEM.1178
Remodelling of the natural product fumagillol
employing a reaction discovery approach
Bradley R. Balthaser, Meghan C. Maloney, Aaron B. Beeler, John A. Porco Jr and John K. Snyder
*
*
In the search for new biologically active molecules, diversity-oriented synthetic strategies break through the limitation of
traditional library synthesis by sampling new chemical space. Many natural products can be regarded as intriguing starting
points for diversity-oriented synthesis, wherein stereochemically rich core structures may be reorganized into chemotypes
that are distinctly different from the parent structure. Ideally, to be suited to library applications, such transformations
should be general and involve few steps. With this objective in mind, the highly oxygenated natural product fumagillol has
been successfully remodelled in several ways using a reaction-discovery-based approach. In reactions with amines,
excellent regiocontrol in a bis-epoxide opening/cyclization sequence can be obtained by size-dependent interaction of an
appropriate catalyst with the parent molecule, forming either perhydroisoindole or perhydroisoquinoline products.
Perhydroisoindoles can be further remodelled by cascade processes to afford either morpholinone or bridged
4,1-benzoxazepine-containing structures.
o identify pharmacological tools for biological processes, as both a synthetic target21 and for its anti-angiogenic proper-
compound discovery must expand beyond the sp2-dominated ties22–24. We envisioned a series of tandem processes that could
T
synthetic libraries common in biological screening1. Diversity- remodel fumagillol into novel chemotypes as dictated by either cat-
oriented libraries have been demonstrated to occupy areas of alyst or reaction partner choice. In this Article, we report our initial
chemical space not normally accessed by more traditional planar, studies, which were aimed at remodelling fumagillol through Lewis
heterocyclic libraries2–4. One approach to accessing increasingly acid-promoted addition of amines.
diverse libraries is to make use of natural products as starting scaf-
folds. A number of studies have used natural products such as Results and discussion
a-santonin5–7 and (–)-shikimic acid8,9 as starting materials to ident- A reaction screen15,16 was first carried out to explore sequential ami-
ify biologically active molecules including 5-lipoxygenase inhibitors nolyses of the 1,4-bis-epoxide. We anticipated that the sequence
and aurora A kinase ligands. Other methods have focused on alter- would be initiated at the spirocyclic epoxide, thereby mimicking
ing the core framework of natural products to create small collec- the reactivity of fumagillin with aminopeptidase MetAP-2, the puta-
tions of structurally unique compounds. A method involving tive mode of its antiangiogenic activity23. An initial reaction screen
catalytic, site-selective derivatization of complex natural products (Supplementary Figs S1–S5) with 12 Lewis acids and four amines
(for example, erythyromycin) has been demonstrated by the resulted in the conversion of the bis-epoxide motif into perhydroi-
Miller group at Yale10,11. In another study, the macrocyclic diterpe- soindole (3)25–27 and/or perhydroisoquinoline (4)28–30 compounds,
noid lathyrane was converted into a small collection of polycyclic which were identifiable through several characteristic signals in
structures using transannular reactions12. Meanwhile, the Miller 1H NMR spectra. Best results were obtained using p-anisidine and
group from Notre Dame has incorporated oxazine heterocycles a metal triflate catalyst. Preliminary optimization of this transform-
into natural products bearing 1,3-butadiene subunits by using ation demonstrated that the use of 2,6-di-tert-butyl-4-methylpyri-
iminonitroso Diels–Alder cycloadditions13,14
Further evolution of these ideas would involve the creation of a yields, presumably by buffering adventitious triflic acid31,32
.
dine (DTBMP) as a proton scavenger significantly improved
.
diverse library of remodelled structures derived from a natural Several metal triflates were subsequently investigated in a second
product, each significantly different from the parent compound. screen, and a linear correlation was found between the atomic
The transformations that are used should allow for the incorpor- radius of the metal catalyst and the distribution of isomeric products
ation of new functionality and ideally be carried out in single-step (Fig. 1b). As metal size increased, perhydroisoindole product 3 was
or tandem processes. We therefore initiated a reaction discovery- increasingly favoured. Lanthanum triflate proved to be optimal for
based approach15–18 that meets these criteria, using a readily avail- the production of 3, with .95:5 regioselectivity (Fig. 1b, entry 2).
able natural product as a starting point for chemically diverse Conversely, the smaller bivalent-metal Zn(OTf)2 favoured the
library synthesis (Fig. 1a). The highly oxygenated natural product formation of perhydroisoquinoline 4 (Fig. 1b, entry 8), thereby
fumagillol (2) was chosen, because the reactivity and proximity of allowing access to either isomer simply by changing the catalyst33–35
.
the two epoxides present site(s) for potential chemistry, and the In the absence of a catalyst, no reaction occurred and fumagillol
hydroxyl and alkene groups offer additional functionality for was fully recovered.
further diversification. Crude fumagillin (1), a natural product
Further optimization using La(OTf)3 and Zn(OTf)2 catalysts
that is readily available from the fermentation broth of Aspergillus was next investigated. The transformations were robust and did
fumigatus and can be hydrolysed to fumagillol (2) (see not require an inert atmosphere nor special precautions with
Supplementary Information)19,20, has generated significant interest anhydrous solvent. Other nonpolar solvents provided similar
Department of Chemistry, Center for Chemical Methodology and Library Development (CMLD-BU), Boston University, 590 Commonwealth Avenue,
Boston, Massachusetts 02215, USA. e-mail: jsnyder@bu.edu; porco@bu.edu
*
NATURE CHEMISTRY | VOL 3 | DECEMBER 2011 | www.nature.com/naturechemistry
969
© 2011 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1178
ARTICLES
a
O
O
OH
Nu
O
HO–
Nu
Reaction
discovery-based
diversification
HO
O
O
HO
HO
H
O
H
O
2
H
O
MeO
MeO
MeO
1
2
Fumagillin
Fumagillol
Natural product scaffold
Common intermediate
,†
*
Atomic
radii (pm)
b
OH
Ratio
Entry
M(OTf)n
Metal-dependent selectivity
2
:
3
:
4
100
75
50
25
0
N
OMe
OMe
No catalyst
La(OTf)3
Pr(OTf)3
Er(OTf)3
Yb(OTf)3
Sc(OTf)3
Mg(OTf)2
Zn(OTf)2
–
:
:
:
1
2
3
4
5
6
7
8
100
0
:
0
0
O
HO
195
185
175
175
160
150
135
: 95
: 88
5
7
H
M(OTf)n (10 mol%)
p-anisidine (1.1 equiv.)
MeO
Isoindole
Isoquinoline
5
HO
HO
3
21 : 54 : 25
34 : 46 : 20
30 : 43 : 27
H
O
+
MeO
DTBMP (60 mol%)
toluene
OH
60 °C, 20 h
N
2
Zn
Mg Sc
Yb/Er Pr La
5
7
: 28 : 67
: 12 : 81
130 140 150 160 170 180 190 200
Atomic radii (pm)
HO
H
OH
MeO
*Ratios were determined by 1H NMR
†Average of three runs
4
Figure 1 | Natural product remodelling using fumagillol. a, Fumagillol can be transformed into multiple chemotypes by means of a panel of related reaction
conditions. Fumagillol was obtained by hydrolysis of crude fumagillin isolated from the fermentation broth of Aspergillus fumigatus. Nucleophiles add first to
the terminal epoxide to form a common intermediate that can be diversified by further reaction under a variety of conditions b, The second epoxide ring
opening produces either an isoindole or an isoquinoline as product depending on the regioselectivity of the reaction. There is a linear correlation between the
atomic radius of the metal in the metal triflate catalyst and the observed regioselectivity. DTBMP, 2,6-di-tert-butyl-4-methylpyridine.
Table 1 | Metal triflate catalysed reactions of fumagillol with primary and secondary amines.
OH
OH
O
R
N
M(OTf)n
amine
N
R
HO
HO
HO
+
H
O
H
H
DTBMP
toluene
60 °C
OH
MeO
MeO
MeO
HO
2
5
6
Entry
Amine
M(OTf)n
(reaction
conditions)
Reaction
time
% yield
5:6
Entry
Amine
M(OTf)n
(reaction
conditions)
Reaction
time
% yield
5:6
1
La(OTf)3*
Zn(OTf)†2
7 h
29 h
91:3
9:76
6
La(OTf)‡3
Zn(OTf)†2
Mg(OTf)§2
20 h
60 h
16 h
62:25
0:57
13:82
MeO
NH2
MeO
(a)
(f)
NH2
MeO
MeO
MeO
2
La(OTf)3*
Zn(OTf)†2
3 h
24 h
84:0
6:80
7
La(OTf)‡3
Zn(OTf)†2
Mg(OTf)§2
24 h
60 h
36 h
66:32
0:2
18:74
NH2
(g)
NH2
(b)
3
4
5
La(OTf)3*
Zn(OTf)†2
29 h
29 h
89:2
8:65
8
9
La(OTf)‡3
Zn(OTf)†2
Mg(OTf)§2
20 h
60 h
20 h
49:20
0:11
12:37
O
NH2
(h)
NH2
(c)
H
La(OTf)3*
Zn(OTf)†2
36 h
29 h
87:0
10:88
La(OTf)‡3
R ¼ PMP
7 (74%)
OTf
N
O
F
NH2
MeO
NH
(i)
R
(d)
La(OTf)‡3
R ¼ Bn
HO
La(OTf)3*
Zn(OTf)†2
4 d
48 h
40:0
10:61
10
OH
N
H
H
F3C
NH2
MeO
(j)
(e)
8 (75%)
*La(OTf)3 (10 mol%), DTBMP (60 mol%), amine (1.1 equiv.). †Zn(OTf)2 (10 mol%), DTBMP (60 mol%), amine (1.1 equiv.). ‡La(OTf)3 (50 mol%), DTBMP (1.5 equiv.), amine (2.0 equiv.). §Mg(OTf)2
(50 mol%), DTBMP (1.5 equiv.), amine (2.0 equiv.). DTBMP, 2,6-di-tert-butyl-4-methylpyridine; PMP, p-methoxyaniline.
regioselectivity, although toluene proved to be optimal, in which optimized using La(OTf)3, with 91% isolated yield (91:3 regioselec-
case catalyst loading could be reduced to 10 mol% while maintain- tivity), and 4 could be obtained with Zn(OTf)2 in 76% yield (9:76
ing reasonable reaction times. Production of 3 was ultimately regioselectivity) (Table 1, entry 1).
970
NATURE CHEMISTRY | VOL 3 | DECEMBER 2011 | www.nature.com/naturechemistry
© 2011 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1178
ARTICLES
perhydroisoindole formation. It was necessary, however, to increase
the catalyst loading to 50 mol% to obtain reasonable reaction times,
presumably due to the greater basicity of these amines leading
to tighter interaction with the catalyst. An even greater reduction
in reaction rate was observed with Zn(II) catalysis, rendering the
reaction unacceptably slow (60 h, ꢀ5–10% conversion). Use of
Mg(OTf)2 (50 mol%) as the catalyst, however, also led predomi-
nantly to the desired perhydroisoquinoline products (6f–6h) in
acceptable reaction times (Table 1, entries 6–8, 16–36 h). The
addition of aromatic and aliphatic secondary amines was also
carried out, providing highly substituted tetrahydrofuran products
7 and 8, both isolated as triflate salts (Table 1, entries 9 and 10).
Insights into the interaction of metal catalysts with fumagillol
were obtained by a series of 13C NMR experiments (Fig. 2a,
Supplementary Fig. S6). In the 13C NMR spectra of fumagillol
obtained with 2 mol% paramagnetic catalysts Yb(OTf)3 (r ¼
175 pm) or Fe(OTf)2 (r ¼ 140 pm)35,36, broadening of the C5 and
C6 resonances was observed, indicating that different-sized metals
are preferentially bound to the pocket formed by the hydroxyl
and methoxy groups of fumagillol (9). The same interaction was
not observed in the C6-silylated analogue 10, with only modest
broadening of the C2, C1′ and C2′ signals observed.
a
MLn
1
1
O
O
3
2
1
´
1
´
6
6
M = Yb or Fe
5
5
2
´
2
´
O
O
O
O
HO
TMSO
10
9
LnM
MLn
OH
OH
H
b
O
R
N
i
N
R
+
R O
R O
´
R O
´
´
H
O
H
OH
MeO
MeO
MeO
= H
HO
2 R =H
5
6
´
ii
11 R =TMS
12
13
´
R
´
R = TMS
´
Entry
Amine
M(OTf)n
5 : 6*
12 : 13†
La(OTf)3:
Zn(OTf)2:
1
2
1 : 9
91 :
3
MeO
F
NH2
(a)
(d)
1 : >20
9 : 76
La(OTf)3:
Zn(OTf)2:
3
4
87 :
0
1 : 13
NH2
10 : 88
1 : >20
To probe the importance of the interactions observed by 13C
NMR, the selectivity of silyl ether 11 under optimized reaction con-
ditions was evaluated. When 11 was reacted with several anilines,
the regioselectivity obtained with La(III) catalysis inverted, leading
predominantly to perhydroisoquinoline products as originally
found in the Zn(II)-catalysed reactions (Fig. 2b). In comparison,
reaction of 11 using Zn(II) became more selective, affording perhy-
droisoquinolines with greater than 20:1 selectivity. These results
suggest a mechanism in which coordination of the metal to the
C6 hydroxyl group of fumagillol with different-sized metals
greatly affects the regiochemical outcome, either through multiden-
tate ligand effects and/or conformational control.
The observed regioselectivity can be rationalized from a model of
metal-coordinated amino alcohol intermediate 14, obtained from
opening of the more labile spirocyclic epoxide (Fig. 2c)37,38. The
larger La(III) catalyst may more easily accommodate the C6 hydroxyl
group simultaneously with C1′ epoxide activation (cf. 15), thereby
leading to tridentate coordination to the substrate wherein the amine
is positioned closer and at a more optimal trajectory for addition to
C1′. In contrast, smaller metals such as Zn(II), which cannot as easily
accommodate the C6 hydroxyl while activating the second epoxide,
may adopt a looser bidentate coordination that places C2′ closer to
the amine (cf. 16). In the case of Zn(II) catalysis, activation of the
second epoxide through adoption of 16 may be rate-determining, as
perhydroisoquinoline formation appears to be largely independent of
aniline nucleophilicity. Further studies to understand the precise
*Isolated yield
†Ratio by 1H NMR
c
OH
15
OH
N
R
N
R
1
´
H
O
O
HO
HO
La
R
OH
H
H
La
H
MeO
N
HO
L
L
L
1
´
2
´
O
O
5
OH
HO
H
14
OH
R
R
N
2
Zn
N
O
´
H
Zn
HO
O
HO
H
L
H
OH
MeO
L
16
6
Figure 2 | Mechanistic studies of bis-epoxide ring opening. a, Metals of
different sizes are shown to preferentially coordinate to the C6 hydroxyl
and C5 methyl ether of fumagillol by 13C NMR. b, Epoxide opening of
TMS-protected fumagillol shows an inversion of regioselectivity: (i) amine
(1.1 equiv.), M(OTf)n (10 mol%), DTBMP (60 mol%), toluene, 60 8C;
(ii) TMSCl, imdazole, DMAP, CH2Cl2, 83%. c, Working model
demonstrating the role of the C6 hydroxyl towards regioselectivity. DMAP,
4-dimethylaminopyridine; DTBMP, 2,6-di-tert-butyl-4-methylpyridine;
TMS, trimethylsilyl.
Bis-epoxide opening and, in particular, catalyst-controlled mechanism leading to selectivity are currently under way.
regioselectivity proved to be quite general (Table 1). A variety of The methodology was further extended by the use of L- and
electron-rich and electron-deficient anilines produced either D-phenylalanine methyl esters, which underwent lactonization
heterocyclic motif (entries 1–5), with La(OTf)3 catalysis forming after initial epoxide opening (Fig. 3a). With Mg(OTf)2 as catalyst,
predominantly perhydroisoindoles 5 and Zn(OTf)2 yielding perhy- perhydroisoquinoline products 17 and 19, respectively, were pro-
droisoquinolines 6. In the case of La(III)-promoted reactions, there duced in ꢀ3:1 regioselectivity relative to the perhydroisoindole-
was a direct correlation of the nucleophilicity of the aniline with derived products (18 and 20/21, respectively) for each amino
reaction rate, with electron-rich amines reacting faster. The rate of acid. Further lactonization of the perhydroisoquinoline analogues
formation for perhydroisoquinoline 6 with Zn(II) catalysis was was not observed. In comparison, the perhydroisoindole formed
largely unaffected by the electronics of the aniline, until a suffi- with D-phenylalanine (catalysed by La(OTf)3) lactonized in situ to
ciently electron-deficient analogue (Table 1, entry 5) was used. afford the polycyclic morpholinone derivative 18. Reaction with L-
Thus, with p-trifluoromethylaniline, the reaction was significantly phenylalanine, however, formed isoindole 20 and morpholinone
slower than with the more electron-rich anilines, which all pro- 21 in a 4:1 ratio. The observed resistance to lactonization of 21
ceeded at approximately the same rate. Heteroaryl amines including can be rationalized from steric congestion caused by the additional
2-aminopyridine and 2-aminothiazole failed to react using either pseudoaxial prenyl substituent at C2′, which was calculated to be
La(OTf)3 or Zn(OTf)2, and fumagillol was fully recovered.
3.1 kcal mol21 higher in energy relative to diastereomer 18
More basic amines were also well tolerated in the reaction (Supplementary Fig. S7). Lactonization of 20 could eventually be
(Table 1, entries 6–8), with La(III) again proving to be optimal for accomplished under basic conditions to yield morpholinone 21 (85%).
NATURE CHEMISTRY | VOL 3 | DECEMBER 2011 | www.nature.com/naturechemistry
© 2011 Macmillan Publishers Limited. All rights reserved.
971

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1178
ARTICLES
a
CO2Me
Ph
Ph
O
b
O
OH
H
OH
H
D-Phe
N
+
N
2
i
HO
HO
O
H
O
HO
´
MeO
MeO
2
´
OH
MeO
2
17
18
18
73%
8%
27%
86%
Mg(OTf)2 (70 h):
La(OTf)3 (70 h):
CO2Me
Ph
O
OH
H
OH
H
OH
Ph
CO2Me
Ph
L-Phe
N
N
2
2
N
+
+
i
HO
O
2
´
HO
HO
´
H
MeO
OH
MeO
MeO
HO
21
19
20
21
Mg(OTf)2 (70 h):
La(OTf)3 (70 h):
67%
6%
20%
71%
3%
16%
ii
85%
Figure 3 | Use of amino-acid esters as reaction partners. a, Selective formation of perhydroisoindoles, perhydroisoquinolines or morpholinones with
phenylalanine as a reaction partner: (i) phenylalanine (2.0 equiv.), M(OTf)n (50 mol%), DTBMP (1.5 equiv.), toluene, 60 8C; (ii) NaOH (2.0 M),
tetrahydrofuran, room temperature, 6 h. DTBMP, 2,6-di-tert-butyl-4-methylpyridine; Phe, phenylalanine. b, Molecular models of phenylalanine-derived
morpholinones 18 and 21.
a
OH
OH
O
OH
H
2-ethynylaniline (2.0 equiv.)
M(OTf)n (50 mol%)
N
N
N
+
+
HO
HO
HO
H
H
DTBMP (1.5 equiv.)
toluene
H
H
O
HO
MeO
MeO
O
MeO
HO
OH
MeO
2
22
23
24
75%
1%
0%
73%
0%
11%
9%
Mg(OTf)2 (70 h, 60 °C):
La(OTf)3 (16 h, 60 °C):
La(OTf)3 (18 h, 90 °C):
17%
62%
b
OH
OH
OH
OH
N
N
N
[H+]
Prins
N
HO
HO
HO
H
H
H
HO
MeO
MeO
MeO
O
O
H
H
O
H
H
MeO
HO
23
25
26
24
Figure 4 | Metal triflate catalysed reactions of fumagillol with 2-ethynylaniline. a, Selective formation of a perhydroisoindole, perhydroisoquinoline or
4,1-benzoxazepine upon reaction with 2-ethynylaniline. DTBMP, 2,6-di-tert-butyl-4-methylpyridine. b, Working mechanism for the novel cascade process to
form a 4,1-benzoxazepine.
Reaction of 2 with 2-ethynylaniline with Mg(OTf)2 catalysis pro- substituted tetrahydrofurans. Perhydrosoindole products under-
duced perhydroisoquinoline 22, while La(III) afforded the expected went further reactions, including lactonizations using amino-acid
perhydroisoindole product 23 (Fig. 4a). With extended reaction esters as epoxide-opening nucleophiles and bridged 4,1-benzoxaze-
times (48 h) with La(OTf)3 or at elevated temperature (90 8C), the pines from an unexpected cascade sequence with 2-ethynylaniline.
novel 4,1-benzoxazepine39–41 24 bearing a [4.2.1] ring system was Remodelled structures produced in this study are currently being
formed from 23 in a highly efficient cascade process42–45
.
examined in a range of biological screens, including those as part
Benzoxazepine 24 is presumably formed by initial hydroalkoxylation of the Molecular Libraries Probe Production Centers Network
of the alkynyl alcohol of 23 to enol ether 2546,47, followed by protona- (MLPCN, http://mli.nih.gov/mli/) and the NIMH Psychoactive
tion to oxonium 26 (Fig. 4b). Subsequent Prins cyclization48 forms Drug Screening Program (PDSP, http://pdsp.med.unc.edu/indexR.
4,1-benzoxazepine 24, thereby providing a dramatic example of html). These studies should pave the way for work to remodel
natural product remodelling via an unanticipated cascade sequence.
other natural product scaffolds to access novel chemotypes and
In summary, the natural product fumagillol has been selectively pharmacological tools.
remodelled into a series of perhydroisoindoles and perhydroisoqui-
Received 1 July 2011; accepted 16 September 2011;
published online 23 October 2011
nolines through sequential ring-opening with amines. Regiocontrol
was achieved through the choice of metal triflate catalysts, with
smaller Zn(II) and Mg(II) catalysts leading to perhydroisoquinolines,
and the larger La(III) catalyst favouring the production of perhydro-
isoindoles. Addition of secondary amines provided highly
References
1. Lovering, F., Bikker, J. & Humblet, C. Escape from flatland: increasing saturation
as an approach to improving clinical success. J. Med. Chem. 52, 6752–6756 (2009).
972
NATURE CHEMISTRY | VOL 3 | DECEMBER 2011 | www.nature.com/naturechemistry
© 2011 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1178
ARTICLES
2. Galloway, W. R. J. D., Isidro-Llobet, A. & Spring, D. R. Diversity-oriented
synthesis as a tool for the discovery of novel biologically active small molecules.
Nature Commun. 1, 80 (2010).
30. Frankowski, K. J. et al. N-Alkyl-octhydroisoquinoline-1-one-8-carboxamides:
selective and nonbasic k-opioid receptor ligands. ACS Med. Chem. Lett. 1,
189–193 (2010).
3. Clemons, P. A. et al. Quantifying structure and performance diversity for sets of
small molecules comprising small-molecule screening collections. Proc. Natl
Acad. Sci. USA 108, 6817–6822 (2011).
31. Barrett, A. G. M., Braddock, D. C., Henschke, J. C. & Walker, E. R. Ytterbium(III)
triflate-catalyzed preparation of calix[4]resorcinarenes: Lewis-assisted Brønsted
acidity. J. Chem. Soc. Perkin Trans. 1 873–878 (1999).
´
4. Clemons, P. A. et al. Small molecules of different origins have distinct
distributions of structural complexity that correlate with protein-binding
profiles. Proc. Natl Acad. Sci. USA 107, 18787–18792 (2010).
5. Mang, C. et al. Natural products in combinatorial chemistry: an
andrographolide-based library. J. Comb. Chem. 8, 268–274 (2006).
6. Schwarz, O. et al. Natural products in parallel chemistry-novel 5-lipoxygenase
inhibitors from BIOS-based libraries starting from a-santonin. J. Comb. Chem.
9, 1104–1113 (2007).
32. Dumeunier, R. & Marko, I. E. On the role of triflic acid in the metal
triflate-catalyzed acylation of alcohols. Tetrahedron Lett. 45,
825–829 (2004).
33. Fujiwara, K., Tokiwano, T. & Murai, A. La(OTf)3-catalysed 6-endo epoxide
opening of 4,5-epoxy-4-methoxymethyl-1-hexanols. Tetrahedron Lett. 36,
8063–8066 (1995).
34. Fujiwara, K., Mishima, H., Amano, A., Tokiwano, T. & Murai, A. La(OTf)3-
catalyzed 7-endo and 8-endo selective cyclizations of hydroxy epoxides.
Tetrahedron Lett. 39, 393–396 (1998).
35. Marson, C. M. Oxygen-directed carbocyclizations of epoxides. Tetrahedron 56,
8779–8794 (2000).
7. Frank, L. et al. Identification of natural-product-derived inhibitors of
5-lipoxygenase activity by ligand-based virtual screening. J. Med. Chem. 50,
2640–2646 (2007).
8. Tan, D. S., Foley, M. A., Shair, M. D. & Schreiber, S. L. Stereoselective synthesis
of over two million compounds having structural features both reminiscent of
natural products and compatible with miniaturized cell-based assays. J. Am.
Chem. Soc. 120, 8565–8566 (1998).
9. Miao, H. et al. Ring-opening and ring-closing reactions of a shikimic
acid-derived substrate leading to diverse small molecules. J. Comb. Chem. 9,
245–253 (2007).
36. Goodell, J. R., Leng, B., Snyder, T. K., Beeler, A. B. & Porco, J. A. Jr.
Multidimensional screening and methodology development for condensations
involving complex 1,2-diketones. Synthesis 2254–2270 (2010).
37. Fardis, M. et al. Design, synthesis and evaluation of a series of novel fumagillin
analogues. Bioorg. Med. Chem. 11, 5051–5058 (2003).
38. Pyun, H.-J. et al. Investigation of novel fumagillin analogues as angiogenesis
inhibitors. Bioorg. Med. Chem. Lett. 14, 91–94 (2004).
10. Lewis, C. A. & Miller, S. J. Site-selective derivatization and remodeling of
erythromycin A by using simple peptide-based chiral catalysts. Angew. Chem.
Int. Ed. 45, 5616–5619 (2006).
11. Lewis, C. A., Longcore, K. E., Miller, S. J. & Wender, P. A. An approach to
site-selective diversification of apoptolidin A with peptide-based catalysts.
J. Nat. Prod. 72, 1864–1869 (2009).
39. Das, J. et al. Substituent activity relationship studies on new azolo
benzoxazepinyl oxazolidinones. Bioorg. Med. Chem. 14, 8032–8042 (2006).
´
´
40. Dıaz-Gavilan, M. et al. Anticancer activity of (1,2,3,5-tetrahydro-4,1-
benzoxazepine-3-yl)-pyrimidines and -purines against the MCF-7 cell line:
preliminary cDNA microarray studies. Bioorg. Med. Chem. Lett. 18,
1457–1460 (2008).
˚
´
12. Appendino, G., Tron, G. C., Jarevang, T. & Sterner, O. Unnatural natural
41. Lopez-Cara, L. C. et al. New (RS)-benzoxazepin-purines with antitumor activity:
products from the transannular cyclization of lathyrane diterpenes. Org. Lett. 3,
1609–1612 (2001).
13. Li, F. et al. Iminonitroso Diels–Alder reactions for efficient derivatization and
functionalization of complex diene-containing natural products. Org. Lett. 15,
2923–2926 (2007).
the chiral switch from (RS)-2,6-dichloro-9-[1-(p-nitrobenzenesulfonyl)-1,2,3,
5-tetrahydro-4,1-benzoxazepin-3-yl]-9H-purine. Eur. J. Med. Chem. 46,
249–258 (2011).
42. Bhunia, S., Wang, K.-C. & Liu, R.-S. PtII-catalyzed synthesis of
9-oxabicyclo[3.3.1]nona-2,6-dienes from 2-alkynyl-1-carbonylbenzenes and
allylsilanes by an allylation/annulation cascade. Angew. Chem. Int. Ed. 47,
5063–5066 (2008).
ˇ´
14. Krchnak, V. et al. Evolution of natural product scaffolds by acyl-arylnitroso
hetero-Diels–Alder reactions: new chemistry on piperine. J. Org. Chem. 73,
4559–4567 (2008).
15. Beeler, A. B., Su, S., Singleton, C. A. & Porco, J. A. Jr. Discovery of chemical
reactions through multidimensional screening. J. Am. Chem. Soc. 129,
1413–1419 (2007).
16. Han, C. et al. Reaction discovery employing macrocycles: transannular
cyclization of macrocyclic bis-lactams. Org. Lett. 11, 413–416 (2009).
17. Jones, A. L. & Snyder, J. K. Synthesis of unique scaffolds via Diels–Alder
cycloadditions of tetrasubstituted cyclohexadienes. Org. Lett. 12,
1592–1595 (2010).
43. Barluenga, J. et al. Tandem intramolecular hydroalkoxylation–hydroarylation
reactions: synthesis of enantiopure benzofused cyclic ethers from the chiral pool.
Chem. Eur J. 14, 4153–4156 (2008).
´
´
´
´
44. Barluenga, J., Fernandez, A., Dieguez, A., Rodrıguez, F. & Fan˜anas, F. J. Gold- or
platinum-catalyzed cascade processes of alkynol derivatives involving
hydroxylation reactions followed by Prins-type cyclizations. Chem. Eur. J. 15,
11660–11667 (2009).
´
´
´
45. Fan˜anas, F. J., Fernandez, A., Çevic, D. & Rodrıguez, F. An expeditious synthesis
of bruguierol A. J. Org. Chem. 74, 932–934 (2009).
18. Medeiros, M. R., Narayan, R. S., McDougal, N. T., Schaus, S. E. & Porco, J. A. Jr.
Skeletal diversity via cationic rearrangements of substituted dihydropyrans. Org.
Lett. 12, 3222–3225 (2010).
46. Yu, X., Seo, S. Y. & Marks, T. J. Effective, selective hydroalkoxylation/cyclization
of alkynyl and allenyl alcohols mediated by lanthanide catalysts. J. Am. Chem.
Soc. 129, 7244–7245 (2007).
`
19. Hanson, F. R. & Eble, T. E. Fumagillin and preparation. US patent
2,652,356 (1950).
20. Tarbell, D. S. et al. The chemistry of fumagillin. J. Am. Chem. Soc. 83,
3096–3113 (1961).
47. Motto, A., Fragala, I. L. & Marks, T. J. Atom-efficient carbon–oxygen bond
formation process. DFT analysis of the intramolecular
hydroalkoxylation/cyclization of alkynyl alcohols mediated by lanthanide
catalysis. Organometallics 29, 2004–2012 (2010).
21. Yamaguchi, J. & Hayashi, Y. Syntheses of fumagillin and ovalicin. Chem. Eur. J.
16, 3884–3901 (2010).
22. Ingber, D. et al. Synthetic analogues of fumagillin that inhibit angiogenesis and
suppress tumor growth. Nature 348, 555–557 (1990).
48. Olier, C., Kaafarani, M., Gastaldi, S. & Bertrand, M. P. Synthesis of
tetrahydropyrans and related heterocycles via Prins cyclization; extension to
aza-Prins cyclization. Tetrahedron 66, 413–445 (2010).
23. Liu, S., Widom, J., Kemp, C. W., Crews, C. M. & Clardy, J. Structure of human
methionine aminopeptidase-2 complexed with fumagillin. Science 282,
1324–1327 (1998).
Acknowledgements
The authors acknowledge the NIGMS CMLD initiative (P50 GM067041) for financial
support, the National Science Foundation for supporting the purchase of NMR (CHE
0619339) and high-resolution mass spectrometry (CHE 0443618) spectrometers, and the
Boston University Undergraduate Research Opportunities Program for support to M.C.M.
The authors also thank Jia-He Li of Sinova Inc. for the generous donation of fumagillin.
24. Lu, J., Chong, C. R., Hu, X. & Liu, J. O. Fumarranol, a rearranged fumagillin
analogue that inhibits angiogenesis in vivo. J. Med. Chem. 49, 5645–5648 (2006).
25. Lins, L. et al. Importance of hydrophobic energy: structural determination of a
hypoglycemic drug of the meglitinide family by nuclear magnetic resonance and
molecular modeling. Biochem. Pharmacol. 52, 1155–1168 (1996).
26. Giraud, E. et al. Multivariate data analysis using D-optimal designs, partial least
squares, and response surface modeling: a directional approach for the analysis
of farnesyltransferase inhibitors. J. Med. Chem. 43, 1807–1816 (2000).
27. Jiang, J. et al. Potent, brain-penetrant, hydroisoindoline-based human
neurokinin-1 receptor antagonists. J. Med. Chem. 52, 3039–3046 (2009).
Author contributions
B.R.B. and M.C.M. carried out the experimental work. A.B.B., J.A.P. and J.S.K. provided
oversight. B.R.B, J.A.P. and J.S.K. conceived the experiments and wrote the manuscript.
28. Hansen, M. M. et al. An enantioselective synthesis of cis-perhydroisoquinoline Additional information
LY235959. J. Org. Chem. 63, 775–785 (1998).
The authors declare no competing financial interests. Supplementary information and
29. Rennison, D. et al. Cinnamoyl derivatives of 7a-aminomethyl-6,14-endo-
ethanotetrahydrothebaine and 7a-aminomethyl-6,14-endo-
ethanotetrahydrooripavine and related opioid ligands. J. Med. Chem. 50,
5176–5182 (2007).
chemical compound information accompany this paper at www.nature.com/
naturechemistry. Reprints and permission information is available online at http://www.
nature.com/reprints. Correspondence and requests for materials should be addressed
to J.A.P. and J.K.S.
NATURE CHEMISTRY | VOL 3 | DECEMBER 2011 | www.nature.com/naturechemistry
973
© 2011 Macmillan Publishers Limited. All rights reserved.