Access to the pactamycin core via an epoxide opening cascade

Article abstract of DOI:10.1021/ol301461e

A synthetic strategy to establish five contiguous stereocenters, in a stereocontrolled manner, on the core structure of pactamycin is described. This sequence exploits the use of a Lewis acid mediated epoxide opening cascade to set the relative configuration of the C4-C5 diol while reversing the configuration at C7. This sequence provides the oxygenated core of pactamycin in just 11 steps.

Full text of DOI:10.1021/ol301461e

ORGANIC
LETTERS
2012
Vol. 14, No. 14
3632–3635
Access to the Pactamycin Core via an
Epoxide Opening Cascade
Travis J. Haussener and Ryan E. Looper*
Department of Chemistry, University of Utah, Salt Lake City, Utah 84112,
United States
r.looper@utah.edu
Received May 28, 2012
ABSTRACT
A synthetic strategy to establish five contiguous stereocenters, in a stereocontrolled manner, on the core structure of pactamycin is described.
This sequence exploits the use of a Lewis acid mediated epoxide opening cascade to set the relative configuration of the C4ꢀC5 diol while
reversing the configuration at C7. This sequence provides the oxygenated core of pactamycin in just 11 steps.
Isolated from the fermentation broth of Streptomyces
pactum var. pactum,1 pactamycin (1) is a highly potent
antiproliferative agent, displaying activity against all three
phylogenetic domains (eukarya, bacteria, and archaea).
When tested against eukaryotic cell lines pactamycin gen-
erally displays nanomolar cytotoxicity (KB, IC50 = 5.3 nM;
MCR-5, IC50 = 53 nM; HCT116 IC50 ≈ 100 nM).2,3
Pactamycin is active against both gram-negative and
gram-positive bacteria4 and has antiproliferative effects
against the protozoan parasites Plasmodium falciparum
(IC50 = 14 nm) and Trypanosoma brucei (IC50 = 7 nm).2
Pactamycin’s activity stems from its ability to inhibit
protein synthesis.5,6 Given its broad-spectrum activity, it
was hypothesized that it must interact with a conserved
region of the ribosome. Originally thought to bind the
P-site,7,8 it was later shown that it most likely binds the
highly conserved E site of the 30S ribosomal subunit.9 This
was further supported by the identification of functional
mutants of H. halobium resistant to pactamycin.10 It was
ultimately shown, with atomic resolution, that pactamycin
does bind a single site on the 30S ribosome.11 In doing so, it
distorts the position of the mRNA such that it cannot
interact with a neighboring E-site bound tRNA. This
prevents the initiation of protein synthesis ultimately
leading to cell death.
Small structural variations, as seen in the related natural
products 7-deoxypactamycin (2)12 and jogyamycin (3)13
(Figure 1), suggest that functional group changes can have
dramatic effects on both potency and selectivity in regard
to antiproliferative effects between kingdoms.13 Recent
work by Mahmud et al. has shown that biosynthetic
(1) Argoudelis, A. D.; Jahnke, H. K.; Fox, J. A. Antimicrob. Agents
Chemother. 1962, 191.
(2) Otogourol, K.; Iwatsuki, M.; Ishiyama, A.; Namatame, M.;
Nishihara-Tukashima, A.; Shibahara, S.; Kondo, S.; Yamada, H.;
Omura, S. J. Antibiot. 2010, 63, 1.
(3) Lu, W.; Roongsawang, N.; Mahmud, T. Chem. Biol. 2011, 18,
425.
(9) Carter, A. P.; Clemons, W. M., Jr.; Brodersen, D. E.; Morgan-
Warren, R. J.; Wimberly, B. T.; Ramakrishnan, V. Nature 2000, 407,
340.
(10) Mankin, A. S. J. Mol. Biol. 1997, 274, 8.
(4) Bhuyan, B. K.; Dietz, A.; Smith, C. G. Antimicrob. Agents Che-
mother. 1962, 184.
(5) Cohen, L. B.; Goldberg, I. H.; Herner, A. E. Biochemistry 1969, 8,
1327.
(6) Cohen, L. B.; Herner, A. E.; Goldberg, I. H. Biochemistry 1969, 8,
1312.
(7) De Stasio, E. A.; Dahlberg, A. E. J. Mol. Biol. 1990, 212, 127.
(8) Tejedor, F.; Amils, R.; Ballesta, J. P. G. Biochemistry 1987, 26,
653.
(11) Brodersen, D. E.; Clemons, W. M., Jr.; Carter, A. P.; Morgan-
Warren, R. J.; Wimberly, B. T.; Ramakrishnan, V. Cell 2000, 103, 1143.
(12) Dobashi, K.; Isshiki, K.; Sawa, T.; Obata, T.; Hamada, M.;
Naganawa, H.; Takita, T.; Takeuchi, T.; Umezawa, H.; Bei, H. S.
J. Antibiot. 1986, 39, 1779.
(13) Iwatsuki, M.; Nishihara-Tsukashima, A.; Ishiyama, A.; Namatame,
M.; Watanabe, Y.; Handasah, S.; Pranamuda, H.; Marwoto, B.;
Matsumoto, A.; Takahashi, Y.; Otoguro, K.; Omura, S. J. Antibiot.
2012, 65, 169.
r
10.1021/ol301461e
Published on Web 07/03/2012
2012 American Chemical Society

manipulations can generate alternate analogues of 1,3
more specifically TM-025 (4) and TM-026 (5) which dis-
play activity against both chloroquine resistant and chloro-
quine sensitive P. falciparum (IC50 = 25ꢀ30 nm), but
display significantly reduced cytotoxicity toward mamma-
lian cells (HCT116, IC50 > 1 mM). While it is unclear if
these analogues maintain pactamycin’s mechanism of ac-
tion, it does suggest that this scaffold holds promise for the
development of efficacious agents against biomedically
significant pathogens.
carbonate group on O70 might serve well to establish the
relative stereochemistry at C4ꢀC5 as depicted in Figure
2A.17 A hypothetical transition state leading to the ring-
opened epoxide, however, reveals a highly destabilizing
syn-pentaneinteraction if the correct stereochemistry atC7
is present.
This led us to consider a ring-opening cascade in which
the incorrect stereochemistry at C7 would be introduced
but invertedby acascade cyclization of apendantamideon
N1 (Figure 2B). Herein we report our progress in imple-
menting this strategy.
Figure 1. Pactamycin and its structural relatives.
Pactamycin’sstructure, elucidated in1972, ishighlighted
by six contiguous stereogenic centers, three of which are
quaternary, with all five positions of the cyclopentane core
functionalized.14 Being the most complex of the aminocy-
lopentitol antibiotics, it is not surprising that the chemical
synthesis of pactamycin presents a significant challenge.
Only two approaches to 1 had been reported prior to the
recent total synthesis by Hanessian’s group.15,16 Even with
this monumental accomplishment, structural manipula-
tions to the scaffold are hampered by its extreme complex-
ity, and its synthetic difficulty has prevented any noteworthy
structureꢀactivity relationship studies.
Figure 2. O7-assisted epoxide ring-opening cascade design.
Installation of the opposite stereochemistry at C7 is
easily accomplished by formation of the oxazoline (6) from
L-threonine as described by Balavoine.18 The enolate of 6
can be alkylated with 5-iodo-1-pentene to give 7 (Scheme 1).19
Significant optimization was required to inhibit β-elimina-
tion of the enolate, identifying HMPA as a key additive.
Under optimized conditions, 7 could be obtained as a
single diastereomer in 80% yield.20 This reaction is note-
worthy in that it can be done on a 100 g scale requiring only
a water workup to remove HMPA and no chromatogra-
phy. The methyl ester was then converted to the methyl
ketone 8 in a single operation utilizing conditions reported
by Grabowski and co-workers.21 The terminal olefin was
then oxidatively cleaved to give the aldehyde 8. An intramo-
Concurrently with Hanessian’s efforts, we were inter-
ested in designing a modular synthesis of pactamycin to
address these issues. Similar in approach, we had recog-
nized threonine as an obvious synthon from which to begin
our campaign as it contains the stereochemcial informa-
tion needed at C7 and C1. However, a key feature in our
approach was to relay the stereochemistry at C7 to the
antidiol functionality present at C4/C5. We envisioned
that an intramolecular epoxide opening by a pendant
€
lecular aldol condensation promoted by SiO2 and Hunig’s
base in toluene at 70 °C afforded the spirocyclic cyclopente-
nal 9 in 75% yield. Oxidation of the unfunctionalized
cyclopentane carbon at C3 was accomplished by radical
bromination.22 To our delight this provided 10 as a single
diastereomer, albeit in modest yield. As long as reaction
(14) Duchamp, D. J. Abstracts J. Am. Crystal. Assoc. Winter Meet-
ing; Albuquerque, 1972, April, 23.
(15) For partial syntheses, see: (a) Knapp, S.; Younong, Y. Org. Lett.
2007, 9, 1359. (b) Tsujimoto, T.; Nishikawa, T.; Urabe, T.; Isobe, M.
Synlett. 2005, 433. (c) Malinowski, J. T.; McCarver, S. J.; Johnson, J. S.
Org. Lett. 2012, 14, 2878.
(16) Hanessian, S.; Vakiti, R. R.; Dorich, S.; Banerjee, S.; Lecomte,
^
F.; DelValle, J. R.; Zhang, J.; Deschenes-Simard, B. Angew. Chem., Int.
Ed. 2011, 50, 3497.
(17) For selected epoxide cascade reactions, see: (a) Vilotijevic, I.;
Jamison, T. F. Angew. Chem., Int. Ed. 2009, 48, 5250. (b) Feng, X.; Shu,
L.; Shi, Y. J. Am. Chem. Soc. 1999, 121, 11002. (c) Molander, G. A.;
Pozo Losada, C. D. J. Org. Chem. 1997, 62, 2935. (d) Minami, A.;
Migita, A.; Inada, D.; Hotta, K.; Watanabe, K.; Oguri, H.; Oikawa, H.
Org. Lett. 2011, 13, 1638.
ꢀ
(18) Aıt-Haddou, H.; Hoarau, O.; Cramailere, D.; Pezet, F.; Daren,
¨
J.; Balavoine, G. G. A. Chem.;Eur. J. 2004, 10, 699.
(19) For the synthesis of 5-iodo-pentene from 5-bromo-pentene, see:
Shi, B.; Hawryluk, N. A.; Snider, B. B. J. Org. Chem. 2003, 68, 1030.
(20) Reddy, L. R.; Fournier, J.; Subba Reddy, B. V.; Corey, E. J.
J. Am. Chem. Soc. 2005, 127, 8974.
(21) Williams, J. M.; Jobson, R. B.; Nobuyoshi, Y.; Marchesini, G.;
Dolling, U.; Grabowski, E. J. Tetrahedron. Lett. 1995, 36, 5461.
(22) Hattori, K.; Kido, Y.; Yamamoto, H.; Ishida, J.; Iwashita, A.;
Mihara, K. Bioorg. Med. Chem. Lett. 2007, 17, 5577.
Org. Lett., Vol. 14, No. 14, 2012
3633

times were kept short, a significant amount of starting
material could later be reisolated and recycled. Longer
reaction times led to substantial amounts of decomposition.
which we presumed to be a product arising from migration
of the benzoyl group. It was hypothesized that 1,3-acyl
migration had occurred to the primary alcohol at C9 as
evidenced by significant deshielding of the AB quartet
signal from the C9 methylene protons in the H NMR
spectrum.
1
Scheme 1. Synthesis of the Key Spirocyclic Cyclopentenal
Scheme 2. Key Cascade Sequence with Azide 11
While the allylic bromide 10 was produced as a single
diastereomer, the relative stereochemistry at C3 could not
be unambiguously assigned. Alternative reactions on sub-
strates similar to 9 (e.g., epoxidation of the C4ꢀC5 alkene)
had demonstrated that addition to the spirocyclic cyclo-
pentene occurs from the same face as N1.23 This is
presumably a consequence of the C7 methyl group being
positioned directly under the cyclopentene. This led us to
believe that the bromide might produce the correct relative
stereochemistry at C3 after substitution with a nitrogen
based nucleophile.
Attempts to reduce the aldehyde in 10 with borohydride
reagents in alcoholic solutions led to either alcoholysis or
reduction of the highly reactive allylic bromide. The
bromide could be cleanly displaced, however, with sodium
azide first, and then the aldehyde cleanly reduced to
provide 11 (Scheme 2).
With the azido alcohol in hand, hydrolysis of the oxazo-
line afforded the free primary amine at C1 which was
reacylated with p-MeOBzCl to give 12. Epoxidation of the
alkene with m-CPBA produced 13 in good yield and as a
single diastereomer. Again, epoxidation was presumed to
have occurred from the top face as shown, due to the C7-
substituents shielding the bottom face of the cyclopentene
core. Screening a variety of Lewis acids identified
BF3•OEt2 as an active catalyst to trigger the cascade
rearrangement to provide 14. NOE studies on 14 suggested
that the rearrangement had proceeded with inversion at C7
as evidenced by a strong correlation between the proton on
C7 and those on C6. Further structural studies were
hampered by the presence of a structural isomer of 14
To simplify our investigations we attempted a global
hydrolysis of the benzoate esters which might be present.
Treatment of 14 with LiOH afforded a single compound
(15), confirming that these were indeed structural isomers
(Figure 3). Compound 15 was a crystalline solid, with
which we were able to confirm the structure by X-ray
crystallography.24 This confirmed that we had successfully
inverted the stereochemistry at C7 and had established the
correct relative stereochemistry of the C4ꢀC5 diol. How-
ever, this revealed that the stereochemistry of the azide at
C3 was incorrect.
With the improper stereochemistry at C3 we sought an
alternative strategy that would permit double inversion of
this center to access the correct diastereomer. To accom-
plish this we first treated 10 with sodium acetate under
solvolysis conditions.
This cleanly delivered the C3-alcohols 16 and 17 as an
∼1:1 mixture of diastereomers (Scheme 3); presumably the
intermediate acetates are cleaved during the reaction.
1
Analysis of the H NMR spectrum of the diastereomer
16 revealed an identical 3J splitting pattern as the azide 15
(dd, 3J = 6.0, 6.0 Hz) suggesting that it carried the inverted
center at C3. The other diastereomer 17 lacks coupling to
3
one of the methylene protons at C2 (d, J = 7.5 Hz)
consistent with the bromide 10, indicating it retained the
original stereochemistry at C3.
Able to identify both diastereomers, we endeavored to
stereoselectively introduce an oxygen nucleophile at C3.
To our surprise, substitution with sodium benzoate produced
(23) See “An explanation of oxazoline directed stereocontrol” in the
Supporting Information for details.
3634
Org. Lett., Vol. 14, No. 14, 2012

Scheme 4. Key Cascade Sequence and 1,3-Acyl Migration
Figure 3. Confirmation of the stereochemistry at C3.
Scheme 3. Solvolysis of the Allylic Bromide
18 as a single diastereomer consistent with complete
inversion, and with the benzoate intact (Scheme 4). Fol-
lowing introduction of the benzoate, simultaneous reduc-
tion of the aldehyde and the C3 ester withLiAlH4 provided
the diol 19. Hydrolysis of the oxazoline and reacylation of
the N1 amine provided 20. As before, epoxidation of the
alkene gave 21 as a single diastereomer. Treatment of 21
with BF3•OEt2 successfully triggered the cascade reaction,
to again give a mixture of benzoate isomers. It was found
that protic acids were effective at completing the 1,3-acyl
migration. Thus, treatment of the reaction mixture with
excess benzoic acid gave 22 as a single isomer in 90% yield.
In summary, an efficient 11-step sequence to access the
oxygenated pactamycin core is described. This approach is
highlighted by an epoxide opening cascade which ulti-
mately delivers five of the six contiguous stereocenters in
a stereocontrolled manner. Studies to install the C2ꢀC3
antidiamine functionality from the C3 alcohol are cur-
rently underway. The brevity of this approach should
strengthen our ability to identify functional groups critical
to achieving selectivity between organisms with this com-
plex antibiotic scaffold.
Acknowledgment. We are grateful to the University of
Utah for financial support. We thank Dr. Atta Arif
(University of Utah) for help in determining the structure
of 15 and Dr. Jim Muller (University of Utah) for assis-
tance with mass spectral analyses.
Supporting Information Available. Experimental and
characterization details, including 1H and 13C NMR
spectra for new compounds 6ꢀ10 and 16ꢀ22. This ma-
terial is available free of charge via the Internet at http://
pubs.acs.org.
(24) This structure has been deposited in the Cambridge Crystal-
lographic Data Centre (CCDC No. 887952).
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 14, 2012
3635