Total synthesis and biological evaluation of ustiloxin natural products and two analogs

Article abstract of DOI:10.1016/j.bmcl.2006.06.071

Synthetic investigations of ustiloxin natural products are described. The first total synthesis of ustiloxin F was completed in 15 steps via ethynyl aziridine ring-opening by a phenol derivative. The results of biological tests of synthetic ustiloxins D a

Full text of DOI:10.1016/j.bmcl.2006.06.071

Bioorganic & Medicinal Chemistry Letters 16 (2006) 4804–4807
Total synthesis and biological evaluation of ustiloxin
natural products and two analogs
Pixu Li,a, Cory D. Evans,a Erin M. Forbeck,a Haengsoon Park,a,à Ruoli Bai,b
a,
Ernest Hamelb, and M. M. Joullie
*
*
´
aDepartment of Chemistry, University of Pennsylvania, 231 S. 34th Street, Philadelphia, PA 19104, USA
bToxicology and Pharmacology Branch, Developmental Therapeutics Program, Division of Cancer Treatment and Diagnosis,
National Cancer Institute at Frederick, National Institutes of Health, Frederick, MD 21702, USA
Received 24 May 2006; revised 21 June 2006; accepted 22 June 2006
Available online 11 July 2006
Abstract—Synthetic investigations of ustiloxin natural products are described. The first total synthesis of ustiloxin F was completed
in 15 steps via ethynyl aziridine ring-opening by a phenol derivative. The results of biological tests of synthetic ustiloxins D and F,
and two analogs, O-Me-ustiloxin D and 6-Ile-ustiloxin, demonstrated that the free hydroxyl group ortho to the ether linkage is
critical for activity and variations at the Val/Ala site produce changes in the biological activity suggesting the need for further
perturbations at this site to more extensively study the tubulin binding.
Ó 2006 Elsevier Ltd. All rights reserved.
Ustiloxins and phomopsins (Fig. 1) are antimitotic het-
erodetic peptides isolated from distinctly different sourc-
es. Ustiloxins A–F (1–5) were isolated from the water
extracts of false smut balls on rice plant panicles caused
by the fungus Ustilaginoidia virens.1 Phomopsins A (6)
and B (7) were isolated from cultures of Phomopsis lepto-
stromiformis.2 Phomopsin A (6) is the natural product
responsible for lupinosis, a potentially fatal liver disease
occurring in livestock in Australia and other countries.3
and inhibit polymerization of tubulin but have no
growth inhibitory effect on bacteria and fungi.4 Both
ustiloxin A (1) and phomopsin A (6) are believed to
interact with tubulin in the Vinca domain, a region
which also binds to vinca alkaloids, rhizoxin, dolasta-
tin 10, spongistatin, and several other antimitotic
agents.5,6 However, the mechanism of action of pep-
tide-based antimitotic agents, including dolastatin 10
and cryptophycin 1 as well as ustiloxin A (1) and pho-
mopsin A (6), is poorly understood in terms of the
structural aspects of their interaction with tubulin.
Interpretation of the actions and toxicities of these
The biological properties of ustiloxins and phomop-
sins are similar in that they both cause mycotoxicosis
O
CO2H
CO2H
R1
R2
CH(CH3)2
H
Compounds
O
N
O
O
S
NH2 OH
OH
N
O
OH
H
R3
Ustiloxin A (1)
Ustiloxin B (2)
Ustiloxin C (3)
O
O
HO2C
HO2C
N
O
O
O
S
2
N
CO2H
NH2 OH
3
HN
H
CH3
R3
HN
O
R1
HO
O
S
O
*
Phomopsin A (6)
Phomopsin B (7)
Cl
HO
N
CH3
*
HO
H
R2
N
10
9
HN
H
H
HN
H
H
CH(CH3)2
CH3
Ustiloxin D (4)
Ustiloxin F (5)
Phomopsin
Ustiloxin
Figure 1. Structures of ustiloxins and phomopsins.
Keywords: Ustiloxins D and F; Tubulin inhibitor; Phomopsin; Total synthesis; Biological evalution; Antimitotic agent.
*
Corresponding authors. Tel.: +1 215 898 3158 (M.M.J.); e-mail: mjoullie@sas.upenn.edu
Present address: Wyeth Research, 401 N. Middletown Road, Pearl River, NY 10965, USA.
Present address: Pharmacopeia Drug Discovery, Inc., 3000 Eastpark Boulevard, Cranbury, NJ 08512, USA.
à
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.06.071

P. Li et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4804–4807
4805
natural products suffers from a lack of receptor map-
ping and structural detail.
in our laboratory in 2002.9,10 The possibility of atrop-
isomerism was investigated by NOE experiments, which
showed that ustiloxin D existed in only one isomeric
form at room temperature.9
One of the most striking aspects of the biology of usti-
loxin A (1) and phomopsin A (6) is that these compounds
are only modestly cytotoxic (IC50 = 2–17 lM) to virtual-
ly all tumor cells in vitro, even though they are potent
inhibitors of tubulin polymerization.6 Dolastatin 10,
for example, is much more toxic to cells (IC50 = 0.5 nM)
than 1 or 6, although they are all comparable inhibitors
of tubulin polymerization and phomopsin A (6) is a com-
petitive inhibitor of dolastatin 10 binding to tubulin.5–7
The reasons for the variation in toxicity of tubulin-bind-
ing natural products are not well understood. A detailed
SAR study would help to determine whether the differ-
ence derives from the overall structure of the molecules
or from specific substituents.
After the completion of the first total synthesis of usti-
loxin D, efforts to make the route more convergent were
undertaken to facilitate access to other ustiloxin and
phomopsin congeners and their analogs. The strategy
of the approach was based on an intramolecular SNAr
reaction to form the 13-membered macrocycle (Fig. 2).
Disconnection of the amide bond of 8 gave two key
fragments, b-hydroxyphenylalanine derivative 9 and a
tripeptide 10 containing a noncoded b-hydroxyisoleu-
cine residue. Regioreversed asymmetric aminohydroxyl-
ation could be used to install the requisite
stereochemistry of the vicinal amino alcohol functional-
ities in 9.11 Disconnection of amide bonds of compound
10 provides diol 11, which was an intermediate of the
first total synthesis of ustiloxin D.10
Further complicating matters is the finding that phomop-
sin A (6) is quite toxic to animals (LD50 = 20 mg/kg in
sheep) in spite of its moderate toxicity against cell lines.
This activity was later attributed to the action of 6 on liver
cells,8 suggesting a possible application in the treatment
of liver cancer. The reason for the discrepancy between
cellular and animal toxicity remains unexplained. One
possibility is that a metabolite of 6 is more toxic than
the natural product. Alternatively, the differential toxici-
ty within this family of natural products might be due to
widely differing sensitivity of various cell types.
The synthesis began with sequential treatment of the pri-
mary hydroxyl group in 11 with Dess–Martin periodinane
and then with NaClO2 to afford the corresponding car-
boxylic acid, which was coupled with glycine benzyl ester
to give 12. After acid deprotection of the N-Boc carba-
mate, the free amine was coupled with N-Boc-valine, fol-
lowed by a second acid cleavage of the Boc protecting
group to afford the tripeptide 10 with a free amino group
necessary for the next coupling reaction.
Therefore, ustiloxins and phomopsins provide an attrac-
tive entry to the study of antimitotic natural products
and their biochemical effects at the molecular and
cellular levels. Chemical synthesis would be a suitable
approach to access a diverse class of compounds careful-
ly designed to critically examine the biological functions
of these natural products. In our efforts to achieve the
total synthesis of ustiloxin and phomopsin natural
products, ustiloxin D (4) was chosen as the lead com-
pound because it is one of the simplest congeners of
the ustiloxin and phomopsin natural product families,
and it retains fairly high biological activity. The total
synthesis of ustiloxin D would provide a scaffold for
the other congeners and analogs.
The amino group of the b-hydroxytyrosine derivative
1311 was methylated by oxazolidine formation. Sequen-
tial reduction12 was followed by protection of the amino
group as its carbobenzyloxy (Z) carbamate. The methyl
ester of compound 14 was converted to the correspond-
ing carboxylic acid 9 by AlBr3 and tetrahydrothioph-
ene13 because basic hydrolysis induced the formation
of an oxazolidinone between the benzylic hydroxyl
group and the Z carbamate. The coupling of 9 and 10
using EDCI–HOBt provided the macrocyclization
precursor 8 in high yield (Scheme 1). The longest linear
sequence of the synthesis of 8 was 13 steps.
One of the main challenges of the total synthesis of usti-
loxin and phomopsin natural products is the stereoselec-
tive formation of the chiral tertiary alkyl–aryl ether.
Extensive investigations to identify a method to form
the challenging tertiary alkyl–aryl ether were carried
out since the project began in our laboratory in 1999.9
The first total synthesis of ustiloxin D was completed
At this point, if the intramolecular reaction had provid-
ed the desired macrocycle 15, manipulation of the nitro
group to the requisite hydroxyl group ortho to the ether
linkage and global hydrogenation would complete the
total synthesis. However, attempts using basic condi-
tions (TBAF, molecular sieves in DMF) as reported
by Zhu or under microwave conditions did not afford
NO2
NO2
O
OH
O
O
F
F
O
O
HO
N
CO2H
HO
N
H
O
CO2Bn
N
H
CO2Bn
HN
H
HN
O
HN
+
HO
OH
NHBoc
11
O
O
CO2H
N
HO
HO
HO
N
H
H2N
ZN
H
ZN
HN
ustiloxin D (4)
9
10
8
Figure 2. Retrosynthetic analysis of ustiloxin D (4).

4806
P. Li et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4804–4807
1) i. Dess-Martin
ii. NaClO2
O
1) TFA, CH2Cl2 94%
2) N-Boc-Val
EDCI, HOBt 89%
O
NaH2PO4 88%
HO
N
CO2Bn
NO2
HO
OH
HO
BocHN
N
CO2Bn
H
O
H
HN
F
2) Gly-OBn
EDCI, HOBt 98%
O
NHBoc
3) TFA, CH2Cl2 quant.
HO
N
12
11
CO2Bn
HN
H
H2N
O
EDCI
HOBt
O
10
N
H
HO
+
ZN
NO2
F
86%
NO2
NO2
8
F
F
1) i. 37% HCHO
ii. Et3SiH, TFA 89%
X
S
AlBr3
quant.
O
NO2
2) CBzCl, NaHCO3 quant.
CO2Me
CO2Me
CO2H
HO
HO
O
HO
N
CO2Bn
NH2
13
ZN
14
H
ZN
HN
O
9
O
HO
N
H
ZN
15
Scheme 1. Synthesis of linear precursor for SNAr ring closure.
the macrocycle although a similar strategy was able to
provide simplified ustiloxin analogs.14 These results sug-
gested that the intramolecular SNAr reaction is not suit-
able for the synthesis of tertiary alkyl–aryl ethers and
prompted us to evaluate alternative approaches to syn-
thesize this motif.
uct family. To take advantage of the new approach in
which the C-6 amino acid residue was installed at a late
stage, ustiloxin F (5) was synthesized via a slight modi-
fication (Scheme 2). Amine 19 was synthesized from the
ring-opening of ethynyl aziridine 17 by b-hydroxytyro-
sine derivative 16, followed by o-nosyl deprotection.
The coupling of 19 and N-Z-alanine provided the pre-
cursor 21 for ustiloxin F. Following the simultaneous
removal of the benzyl carbamate, benzyl ester, benzyl
ether, and reduction of the ethynyl group to the requisite
ethyl group by hydrogenation, the resulting linear pre-
cursor was cyclized to provide macrocycle 23 in 13%
for two steps. We believe that difficulties in the macro-
lactamization step are responsible for the moderate yield
as hydrogenation completely converted 23 to the desired
linear precursor. Global acidic deprotection of the Boc
While we were investigating a new method to build the
unusual chiral tertiary alkyl–aryl ether motif under mild
conditions, Wandless and co-workers reported a second
total synthesis of ustiloxin D using the Pd-catalyzed
asymmetric allylic O-alkylation (AAA) reaction to form
the challenging ether linkage. Although the synthesis
significantly reduced the linear sequence to 20 steps, it
suffered from a diastereoselectivity issue brought by
the AAA reaction.15 In the meantime, an unprecedented
copper-catalyzed ethynyl aziridine ring-opening reaction
by phenol derivatives was discovered and optimized in
our group. This reaction was the key to a convergent ap-
proach to the ustiloxin and phomopsin natural products
because of its high stereo- and regioselectivity and broad
functional group tolerance. Based on the new ethynyl
aziridine ring-opening reaction, a convergent total syn-
thesis of ustiloxin D was completed in 15 steps.16
O
O
OMe
OH
O
O
O
O
N
CO2H
N
CO2H
H
H
HN
HN
O
O
HO
N
HO
N
H
H
HN
HN
6-Ile-ustiloxin (24)
O-Me-ustiloxin D (25)
As shown in Figure 1, the C-6 amino acid residue is one
of the two variations among the ustiloxin natural prod-
Figure 3. Structures of O-Me-ustiloxin and 6-Ile-ustiloxin analogs.
OBn
OBn
O
OBn
O
OH
O
CuOAc
DBU
O
CO2t-Bu
O
N
CO2t-Bu
PhSH
Cs2CO3
N
CO2t-Bu
H
NH
H
+
NHNs
NH2
Toluene
90%
CO2Bn
DMF
78%
N
HO
CO2Bn
CO2Bn
Ns
HO
HO
BocN
BocN
BocN
18
17
16
19
O
O
OBn
O
OH
OH
ZHN
CO2H
O
O
O
O
N
N
H
CO2t-Bu
CO2t-Bu
N
CO2H
H
O
TFA
Et3SiH
H
1) H2, Pd black
R
HN
HN
HN
O
O
EDCI, HOBt
NaHCO3, DMF
O
2) EDCI, HOBt
NaHCO3, DMF
CH2Cl2
CO2Bn
ZHN
HO
HO
N
HO
N
R
R
R
H
H
BocN
BocN
HN
20: R=i-Pr, 97%
21: R=Me, 85%
ustiloxin D (4): R=i-Pr, 50%
ustiloxin F (5): R=Me, 80%
22: R=i-Pr, 18% for 2 steps
23: R=Me, 13%
Scheme 2. Synthesis of ustiloxin D (4) and ustiloxin F (5).

P. Li et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4804–4807
Table 1. Inhibitory effects of ustiloxins on purified tubulin polymerization
4807
Compound
Ustiloxin A (1)
Ustiloxin D (4) (synthetic)
Ustiloxin F (5) (synthetic)
8.2
6-Ile-ustiloxin (24)
O-Me-ustiloxin D (25)
>40
IC50 (lM)
1.1 0.2
1.5 0.06
1
4.8 0.4
carbamate and tert-butyl ester completed the synthesis
of ustiloxin F (5) in 80% yield. The first total synthesis
of ustiloxin F was completed in 15 steps with a 1.8%
overall yield.
Acknowledgments
We thank the NIH (CA-40081), NSF (CHEM-0130958),
Wyeth Research, and the University of Pennsylvania
(Research Foundation) for financial support. We thank
Professor Iwasaki for a sample of ustiloxin A.
Recently, two ustiloxin analogs, 6-Ile-ustiloxin (24) and
O-Me-ustiloxin D (25), were synthesized in our labora-
tory by slight modification of the newly developed route
(Fig. 3).
Supplementary data
Together with ustiloxin A (1) (provided by Professor
Iwasaki), synthetic ustiloxin D (4), ustiloxin F (5), 6-
Ile-ustiloxin (24), and O-Me-ustiloxin D (25) were eval-
uated for their inhibitory effects on the polymerization
of purified tubulin.17 The biological evaluation provided
information on the importance of the Ala/Val variant
site and the free hydroxyl group ortho to the ether link-
age to the biological activity of the ustiloxin natural
product family.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.06.071.
References and notes
1. (a) Koiso, Y.; Li, Y.; Iwasaki, S.; Hanaoka, K.; Kobay-
ashi, T.; Sonoda, R.; Fujita, Y.; Yaegashi, H.; Sato, Z.
J. Antibiot. 1994, 47, 765; (b) Koiso, Y.; Morisaki, N.;
Yamashita, Y.; Mitsui, Y.; Shirai, R.; Hashimoto, Y.;
Iwasaki, S. J. Antibiot. 1998, 51, 418.
2. Culvenor, C. C. J.; Beck, A. B.; Clarke, M.; Cockrum, P.
A.; Edgar, J. A.; Frahn, J. L.; Jago, M. V.; Lanigan, G.
W.; Payne, A. L.; Peterson, J. E.; Petterson, D. S.; Smith,
L. W.; White, R. R. Aust. J. Biol. Sci. 1977, 30, 269.
3. (a) van Warmelo, K. T.; Marasas, W. F. O.; Adelaar, T.
F.; Kellerman, T. S.; van Rensburg, I. B. J.; Minne, J. A.
J. S. Afr. Vet. Med. Assoc. 1970, 41, 235; (b) Gardiner, M.
R.; Petterson, D. S. J. Comp. Pathol. 1972, 82, 5.
4. Iwasaki, S. Med. Res. Rev. 1993, 13, 183.
The IC50 values of the inhibition of several ustiloxins
on purified tubulin polymerization were obtained (Ta-
ble 1). The activities of our synthetic ustiloxins D and
F, relative to the natural ustiloxin A, are similar to
those previously reported for the natural products,
although we find greater relative activity in ustiloxin
D (4) as compared with ustiloxin A (1).2–4 The great-
er than 3-fold and 5-fold reduced activity of 24 and 5,
respectively, relative to 4 indicates a limited tolerance
for size changes, either smaller or larger, for the alkyl
substituent at C-6. Even more important for tubulin
inhibition is the unsubstituted hydroxyl group on
the phenyl ring, since methylation at this position
resulted in an almost inactive compound (25),
although slight inhibitory activity was observed when
compound 25 was present in the reaction mixture at
40 lM.
5. Hamel, E. Med. Res. Rev. 1996, 16, 207.
6. Hamel, E.; Covell, D. G. Curr. Med. Chem.: Anti-Cancer
Agents 2002, 2, 19.
7. Hamel, E. Biopolymers 2002, 66, 142.
8. (a) Culvenor, C. C. J.; Edgar, J. A.; Mackay, M. F.;
Gorst-Allman, C. P.; Marasas, W. O. F.; Steyn, P. S.;
Vleggar, R.; Wessels, P. L. Tetrahedron 1989, 45, 2351; (b)
Petterson, D. S.; Howlett, R. M.; Robertson, T. A.;
Papadimitriou, J. M. Aust. J. Exp. Biol. 1979, 57, 211; (c)
Toensing, E. M.; Steyn, P. S.; Osborn, M.; Weber, K. Eur.
J. Cell. Biol. 1984, 35, 156.
In addition, we performed a cytotoxicity evaluation of
these five compounds with the human Burkitt lympho-
ma cell line CA46. The highest drug concentration
examined was 2.5 lM, and only ustiloxin A (1) showed
any activity, with an IC50 value of 2.5 lM. For compar-
ison, the potent antimitotic peptide dolastatin 1018 was
examined simultaneously and yielded an IC50 value of
30 nM. The limited cytotoxicity observed here with the
ustiloxins is also in agreement with previous results.2
9. Cao, B. Ph.D. Thesis, University of Pennsylvania, Phila-
delphia, PA, USA, 2002.
´
10. Cao, B.; Park, H.; Joullie, M. M. J. Am. Chem. Soc. 2002,
124, 520.
11. Park, H.; Cao, B.; Joullie, M. M. J. Org. Chem. 2001, 66,
´
7223.
12. Beulshausen, T.; Groth, U.; Schoellkopf, U. Liebigs Ann.
Chem. 1992, 523.
13. Node, M.; Nishide, K.; Sai, M.; Fuji, K.; Fujita, E. J. Org.
Chem. 1981, 46, 1991.
14. Laib, T.; Zhu, J. Synlett 2000, 1363.
15. (a) Tanaka, H.; Sawayama, A. M.; Wandless, T. J. J. Am.
Chem. Soc. 2003, 125, 6864; (b) Sawayama, A. M.;
Tanaka, H.; Wandless, T. J. J. Org. Chem. 2004, 69, 8810.
´
16. Li, P.; Evans, C. D.; Joullie, M. M. Org. Lett. 2005, 7,
5325.
17. Hamel, E. Cell Biochem. Biophys. 2003, 38, 1.
18. Bai, R.; Pettit, G. R.; Hamel, E. Biochem. Pharmacol.
1990, 39, 194.
In conclusion, synthetic studies of a convergent ap-
proach to ustiloxin D were investigated. The first total
synthesis of ustiloxin F was completed in 15 steps.
Tubulin inhibition studies of natural ustiloxin A, syn-
thetic ustiloxins D and F, 6-Ile-ustiloxin and O-Me-usti-
loxin D showed that tolerance of substitution at the C-6
position is limited and that the free hydroxyl group
ortho to the ether linkage is essential for biological
activity.