Total synthesis and biological evaluation of rakicidin A and discovery of a simplified bioactive analogue

Article abstract of DOI:10.1002/anie.201509926

We report a concise asymmetric synthesis of rakicidin A, a macrocyclic depsipeptide that selectively inhibits the growth of hypoxic cancer cells and stem-like leukemia cells. Key transformations include a diastereoselective organocatalytic cross-aldol reaction to build the polyketide portion of the molecule, a highly hindered ester fragment coupling reaction, an efficient Helquist-type Horner - Wadsworth - Emmons (HWE) macrocyclization, and a new DSC-mediated elimination reaction to construct the sensitive APD portion of rakicidin A. We further report the preparation of a simplified structural analogue (WY1) with dramatically enhanced hypoxia-selective activity. Take my breath away: Rakicidin A, a depsipeptide natural product with hypoxia-selective antitumour activity, is comprised of a ring system containing sensitive and congested functionalities. A modular asymmetric synthesis and initial biological evaluation of the natural product, and the discovery of a simplified analogue displaying strongly enhanced hypoxia selectivity is reported.

Full text of DOI:10.1002/anie.201509926

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201509926
German Edition: DOI: 10.1002/ange.201509926
Natural Products
Total Synthesis and Biological Evaluation of Rakicidin A and
Discovery of a Simplified Bioactive Analogue
Michail Tsakos, Lise L. Clement, Eva S. Schaffert, Frank N. Olsen, Sebastiano Rupiani,
Rasmus Djurhuus, Wanwan Yu, Kristian M. Jacobsen, Nikolaj L. Villadsen, and
Thomas B. Poulsen*
Abstract: We report a concise asymmetric synthesis of
rakicidin A, a macrocyclic depsipeptide that selectively inhibits
the growth of hypoxic cancer cells and stem-like leukemia cells.
Key transformations include a diastereoselective organocata-
lytic cross-aldol reaction to build the polyketide portion of the
molecule, a highly hindered ester fragment coupling reaction,
an efficient Helquist-type Horner—Wadsworth—Emmons
(HWE) macrocyclization, and a new DSC-mediated elimina-
tion reaction to construct the sensitive APD portion of
rakicidin A. We further report the preparation of a simplified
structural analogue (WY1) with dramatically enhanced hypo-
xia-selective activity.
T
umor hypoxia is tightly associated with chemo- and radio-
resistance, increased metastatic potential, and poor prognosis
in a number of cancer sub-types.[1] Evidence connects hypoxia
to the capacity of cancer cell populations to form new
tumors—defined by the content of stem-like cancer cells—
through activation of a latent transcriptional program called
the epithelial–mesenchymal transition (EMT).[2] These
insights have elevated master hypoxic regulators such as
mediators of the ER stress response,[3,4a] the TSC1/2-mTOR-
RHEB axis,[4] and the hypoxia-inducible factors HIF-1/HIF-
2[5] to high-priority therapeutic targets. Mechanistic studies of
novel small molecules that specifically target hypoxic/stem-
like cancer cells may reveal new pathway addictions that are
critical in these cells. The natural product rakicidin A (1,
Figure 1A) is a macrocyclic depsipeptide from Micromono-
spora sp.[6] with selective toxicity towards hypoxic cancer
cells[7] and gleevec-resistant hypoxia-adapted chronic mye-
logenous leukemia (CML) cells.[8] The latter is a model system
for CML stem cells. The mechanism of action of rakicidin A
(1) is unknown, and preliminary data indicate that the
compound does not perturb the transcriptional activity of
HIF-1,[7] nor does it appear to undergo bioreductive activa-
tion.[7,9] Structurally, 1 is characterized by a complex archi-
tecture comprising a polyketide unit and three amino acids:
Sarcosine, b-hydroxyasparagine (b-HOAsn), and the rare 4-
amido-2,4-pentadienoate (APD) group. The combination of
Figure 1. A) Chemical structure and key features of rakicidin A. B) Ret-
rosynthetic analysis of rakicidin A. Colored numbers represent the
sequence of fragment couplings and macrocyclization.
interesting structure, potent biological activity, and unre-
solved mechanistic questions prompted us to investigate
rakicidin A.
Since the stereochemistry of 1 was undetermined at the
onset of our studies, we conducted a combination of NOE and
conformational analyses of our previously developed model
system of the macrocyclic ring,[10] as well as surveys of the
literature (see the Supporting Information) in order to
prioritize stereoisomers for synthesis.
[*] Dr. M. Tsakos, Dr. L. L. Clement, Dr. E. S. Schaffert, F. N. Olsen,
S. Rupiani, R. Djurhuus, W. Yu, K. M. Jacobsen, N. L. Villadsen,
Prof. Dr. T. B. Poulsen
Department of Chemistry, Aarhus University
Langelandsgade 140, 8000 Aarhus C (Denmark)
E-mail: thpou@chem.au.dk
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201509926.
1030
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1030 –1035

Angewandte
Communications
Chemie
This analyses converged upon (2R,3R,14R,15R,16S) as
the most likely configuration of natural rakicidin A,[11,12] and
we initially targeted this particular isomer. Recently, two
reports appeared, both claiming assignment of the con-
figuration of rakicidin A.[13,14] Through synthesis of a com-
pound with spectroscopic data matching those of the
natural product, Chen and co-workers arrived at
a (2S,3S,14S,15S,16R) configuration,[13] whereas Igarashi
and co-workers paradoxically concluded the stereochemistry
to be (2R,3R,14S,15S,16R) by direct degradation of the
natural product and comparison with authentic standards
(Figure 1A).[14] This latter assignment, however, was cor-
rected earlier this year to corroborate a (2S,3S,14S,15S,16R)
configuration for rakicidin A, and it was thus clear that our
initial synthesis in fact had targeted ent-rakicidin A. Herein,
we report our unique route to the natural isomer of
rakicidin A, along with biological evaluation and the discov-
ery of a simplified bioactive rakicidin A analogue.
Although we have previously demonstrated in a model
system that the APD group can be constructed directly by
a Horner–Wadsworth–Emmons (HWE) reaction between
a b-amidophosphonate and a 2-amidoacrolein-derivative,[10] it
was deemed too challenging to handle this sensitive function-
ality during several synthetic steps. Consequently, dehydra-
tion would be delayed to the final step of the synthesis
(Figure 1B).[15] Considering macrocyclization sites to access 2,
we finally opted for pursuing an intramolecular HWE
reaction for building the trans-C9–C10 double bond because
of the relatively unhindered setup and the potential templat-
ing effects of the required metalloenolate intermediate. The
cyclization precursor would be constructed by three fragment
couplings, the first of which being an intermolecular ester
coupling reaction[16] between a b-hydroxyaspartate-building
block 5 and the secondary carbinol of b-hydroxy ester 4. The
(14S,15S,16R) isomer of 4 constitutes an anti-aldol-Felkin
motif that should be accessible through asymmetric aldol
methodology. The threo-(2S,3S) enantiomer of b-hydroxyas-
partic acid can be accessed through a known enantioresolu-
tion of the commercial racemate.[17]
We first devised an efficient route from 1,12-dodecanediol
to chiral aldehyde 13 (Scheme 1). The first stereogenic center
at the C4 position of fragment 4 (C16 in 1) was installed
through an asymmetric methylation employing Evans (R)-
oxazolidinone.[18] This was followed by reductive cleavage of
the auxiliary and Parikh–Doering oxidation[19] to give enan-
tiopure aldehyde 13. Working initially on ent-4, we found that
an Oppolzer aldol reaction[20] using superstoichiometric TiCl4
could be employed to selectively prepare the desired anti-syn
stereotriad. In our hands, however, this method lacked
reproducibility, notably on a large scale, resulting occasionally
in drastically reduced yields and selectivity. Poor solubility of
13 at low temperature was a major issue. We therefore
considered alternatives,[21] and to our delight, we found that
the proline-catalyzed cross-aldol reaction[22] between propa-
nal and 13, despite the double stereocontrol, proceeded with
high diastereoselectivity. The typical yields ranged from 30–
40% after chromatography, however, owing to its operational
simplicity, relative step-economy, and cost-effectiveness, this
solution is superior to the Oppolzer aldol reaction sequence,
Scheme 1. Synthesis of b-hydroxyester 4. Reagents and conditions:
a) 48% aq. HBr (1.7 equiv), toluene, reflux, 65 h; b) iBuMgBr
(3.0 equiv), Li2CuCl4 (0.013 equiv), THF, À788C to RT, 100 min;
c) H5IO6 (3.0 equiv), CrO3 (0.012 equiv), aq. CH3CN, 08C to RT, 8 h;
d) PivCl (1.1 equiv), NEt3 (3.0 equiv), THF, 08C to RT, 50 min; then
DMAP (0.3 equiv), (R)-4-Benzyl-2-oxazolidinone (1.2 equiv), THF,
508C, 5 h; e) NaHMDS (1.2 equiv), THF, À788C, 1 h; then MeI
(5.0 equiv), À788C, 3 h; f) LiBH4 (2.0 equiv), MeOH (1.5 equiv), THF,
08C to RT, 7 h; g) SO3-Pyr (4.0 equiv), NEt3 (5.0 equiv), CH2Cl2/DMSO,
08C 1 h; h) propanal (2.0 equiv, syringe pump), l-proline (0.2 equiv),
DMF, 08C, 64 h; i) NaClO2 (4.0 equiv), isoamylene (10 equiv),
t-BuOH/H2O, 08C, 17 h; j) BnBr (2.5 equiv), K2CO3 (3.0 equiv), DMF/
CH2Cl2, RT, 48 h. iBuMgBr=isobutyl magnesium bromide, PivCl=pi-
valoyl chloride, DMAP=4-dimethylaminopyridine, NaHMDS=sodium
hexamethyldisilazide, Pyr=pyridine, THF=tetrahydrofuran, DMF=di-
methylformamide, DMSO=dimethylsulfoxide.
at least for this particular target structure. Pinnick oxidation
and benzyl ester formation completed an efficient route to b-
hydroxy ester 4.
Proceeding with the key ester coupling, we encountered
serious problems and came up short trying to achieve the
coupling between various protected forms of (2S,3S)-b-
HOAsn or (2S,3S)-b-HOAsp with hindered alcohol 4. We
had anticipated challenges in this reaction,[16] but not even
traces of product could be isolated. Searching for ways to
reduce steric crowding as much as possible, we discovered
that tying together the b-OH and b-COOH groups as a lactone
ketal[23] uniquely enabled reactivity in this coupling
(Scheme 2).
Angew. Chem. Int. Ed. 2016, 55, 1030 –1035
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1031

Angewandte
Communications
Chemie
Scheme 2. Amide fragment couplings and completion of synthesis. Reagents and conditions: a) DCC (2.4 equiv), DMAP (0.3 equiv), DMF/CH2Cl2
(9:1), 08C, 5 h, b) MeCN/DEA (1:1), RT, 30 min; c) 3 (2.5 equiv), EDCI (2.6 equiv), HOAt (2.6 equiv), NEt(iPr)2 (3.0 equiv), RT, 16 h; d) 10% Pd/
C, H2, EtOH, RT, 2 h; e) 6 (2.0 equiv), EDCI (1.5 equiv), HOAt (1.5 equiv), NEt(iPr)2 (2.0 equiv), rt, 16 h; f) DMP (2.0 equiv), CH2Cl2, RT, 1 h;
g) Zn(OTf)2 (4.0 equiv), TMEDA (2.0 equiv), NEt3 (5.0 equiv), THF, 458C, 16 h; h) AcOH/H2O/THF (2:2:1), 458C, 24 h; i) PyBOP (1.2 equiv),
NMM (4.0 equiv), NH4HCO3 (3.0 equiv), THF, 08C to RT, 24 h; j) DSC (1.5 equiv), NEt(iPr)2 (2.0 equiv), MeCN/THF, 458C, 24 h. DCC=Dicyclo-
hexylcarbodiimide, DMAP=4-dimethylaminopyridine, EDCI=1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide, HOAt=1-Hydroxy-7-azabenzotri-
azole, DMP=Dess–Martin periodinane, TMEDA=N,N,N’,N’-Tetramethylethane-1,2-diamine, PyBOP=(Benzotriazol-1-yloxy)tripyrrolidinophospho-
nium hexafluorophosphate, NMM=N-methylmorpholine, DSC=N,N’-Disuccinimidyl carbonate, DEA=Diethylamine.
After extensive optimization, we converged on a DCC-
coupling method that reproducibly generated the required
ester 16 in yields of around 60%, typically with recovery of
alcohol 4. The lactone ketal protecting group may find use in
other syntheses of natural products containing b-HOAsn or b-
HOAsp functionalities. With fragment 16 in hand, we could
proceed with the construction of the two amide bonds
(Scheme 2) and in both instances, efficient EDCI-mediated
couplings were achieved to give the phosphonate 20 in an
excellent yield (79%, 4 steps). The macrocyclization sequence
was initiated with DMP oxidation of the primary alcohol,
followed by exposure of aldehyde phosphonate 21 to con-
ditions that would enable the HWE cyclization. We finally
achieved an efficient (E)-selective macrocyclization under the
soft-enolization conditions originally described by Helquist in
the presence of TMEDA/Zn(OTf)2 and triethylamine.[24]
These conditions also reproducibly resulted in removal of
the TBS ether, delivering advanced macrocycle 2 as a 4:1
mixture of diastereomers[25] in 78% yield over the three
transformations. At this stage, the lactone ketal could be
unmasked to install the primary amide functionality. Expo-
sure of macrocycle 2 to 2m NH3 in 2-propanol allowed
selective opening of the ketal to generate the b-HOAsn group
but, unfortunately, it was generated as a 1:1 mixture of
epimers most likely at C3. We attribute this stereochemical
scrambling to the strongly basic conditions. Scouting for an
alternative sequence for efficient introduction of the primary
amide, we found that deprotection to the a-hydroxy acid
could be carried out cleanly by using aqueous AcOH in THF,
and when this material was directly subjected to a PyBOP-
mediated coupling with ammonium bicarbonate, b-hydroxy
amide 22 was formed in high yield.
The final dehydration of the primary alcohol also turned
out to be difficult. Mesylation with MsCl curiously resulted in
a non-mesylate derivative that still contained an sp3 center at
C11. We tentatively assign this material as the corresponding
chloride, although subjection of the compound to HRMS
analysis paradoxically only returned the mass matching that
of rakicidin A. Elimination from this intermediate under
basic conditions was not successful. Finally, we discovered
1032
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1030 –1035

Angewandte
Communications
Chemie
that dehydration to the sensitive APD functionality could be
din A causes potent and selective growth inhibition under
achieved in a one-pot reaction employing N,N’-disuccinimidyl
carbonate (DSC)[26] at slightly elevated temperature, likely
through the intermediacy of a mixed NHS carbonate
(Scheme 2).
hypoxic conditions (GI50 = 36 Æ 1.4 nm, SI = 9.8 Æ 1.0), in
accord with the data initially reported for the natural
product.[7] Interestingly, the hypoxia selectivity of rakicidin A
exceeds that of tirapazamine (GI50 = 17 Æ 1.3 mm, SI = 6.4 Æ
1.3), an established bioreductively triggered cytotoxin used as
a positive control. The combination of a macrocycle-embed-
ded weak electrophile (APD)[15b] and a pendent long lip-
ophilic chain means that rakicidin A shows significant struc-
tural similarity to the syrbactins (e.g. cepafungin I, Figure S2
in the Supporting Information), a family of potent inhibitors
of the eukaryotic 20S proteasome.[28] However, rakicidin A
(1) does not act through this mechanism, as demonstrated by
the fact that none of the proteolytic activities of the
proteasome were inhibited by 1 (10 mm) in direct biochemical
assays (Figure 2B and Figure S2). In further support of this
conclusion, we found that bortezomib, an FDA-approved
proteasome inhibitor, did not display hypoxia-selective
growth inhibition (Figure 2B). Indeed, specific small-mole-
cule inhibitors of several additional pathways known to be
important in the hypoxic state failed to phenocopy the activity
of rakicidin A in PANC-1 cells (Figure S3). We found that the
APD unit is critical since its removal in compound 22 results
in a complete loss of activity (Figure 2D). We next generated
a structure (WY1) encompassing only the APD unit and the
lipopeptide chain. Despite the increased conformational
freedom of this compound, the synthesis and notably the
macrocyclization proceed with high efficiency (Figure 2C and
Purification and isolation of the natural product proved to
be very challenging and resulted in low yields of the analyti-
cally pure material even though the dehydration appeared to
proceed smoothly (by TLC analysis), with full conversion of
the starting material after 24 hours. In accord with the
observations of Chen and co-workers,[13] rakicidin A appears
not to be stable in the organic solvents used for standard
purifications (see the Supporting Information), and we
observed extensive degradation after flash column chroma-
tography and subsequent concentration (or lyophilization)
from these solvents.[27] Final purification by RP–HPLC
(eluting with MeCN/H2O) and subsequent lyophilization
delivered pure rakicidin A as a white powder, which could
be readily manipulated. The synthetic material matched all
spectroscopic data reported for the natural product (see the
Supporting Information). Attempted DSC dehydration of
epimer (3R)-22, generated as a side product in our initial
NH3-mediated lactone ketal opening, resulted in complete
decomposition.
We investigated the performance of rakicidin A (1) in
comparative cell viability assays with PANC-1 pancreatic
carcinoma cells under normoxic (20% O2) and hypoxic (0–
0.5% O2) growth conditions (Figure 2A). Synthetic rakici-
Figure 2. Biological evaluation of synthetic rakicidin A and analogues. A) Representative growth inhibition curves in PANC-1 cells evaluated under
normoxic (20% O2, red) and hypoxic (0.1–0.5% O2, blue) culture conditions for 48 h. Bioreductively activated cytotoxin tirapazamine is the
positive control. B) Hypoxia-selective activity of bortezomib and in vitro 20S proteasome inhibition assay with an AMC-conjugated substrate
specific for the chymotrypsin-like activity. C) Synthetic route to WY1, a rakicidin A analogue comprising only the APD and lipid functionalities (for
details see Supporting Information and Scheme S1). D) Growth inhibition by compound 22 (Scheme 2) and WY1 in PANC-1 cells. E) Effect of
N-acetylcysteine (NAC) on growth inhibition by rakicidin A and tirapazamine under hypoxic conditions. SI=Selectivity index.
Angew. Chem. Int. Ed. 2016, 55, 1030 –1035
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1033

Angewandte
Communications
Chemie
Cuda, W. Court, K. R. Fox, P. A. Townsend, G. K. Packham,
et al., J. Am. Chem. Soc. 2013, 135, 10418 – 10425.
[6] K. D. McBrien, R. L. Berry, S. E. Lowe, K. M. Neddermann, I.
Bursuker, S. Huang, S. E. Klohr, J. E. Leet, J. Antibiot. 1995, 48,
1446 – 1452.
[7] Y. Yamazaki, S. Kunimoto, D. Ikeda, Biol. Pharm. Bull. 2007, 30,
261 – 265.
[8] M. Takeuchi, E. Ashihara, Y. Yamazaki, S. Kimura, Y. Naka-
gawa, R. Tanaka, H. Yao, R. Nagao, Y. Hayashi, H. Hirai, et al.,
Cancer Sci. 2011, 102, 591 – 596.
[9] J. M. Brown, W. R. Wilson, Nat. Rev. Cancer 2004, 4, 437 – 447.
[10] T. B. Poulsen, Chem. Commun. 2011, 47, 12837 – 12839.
[11] Lipodepsipeptide natural products featuring a stereotriad sim-
ilar to rakicidin A. Emericellamide A/B: a) D. C. Oh, C. A.
Kauffman, P. R. Jensen, W. Fenical, J. Nat. Prod. 2007, 70, 515 –
520; stevastelin C3: b) K. Kurosawa, K. Matsuura, N. Chida,
Tetrahedron Lett. 2005, 46, 389 – 392; c) N. Bisek, S. Wetzel, H.-
D. Arndt, H. Waldmann, Chem. Eur. J. 2008, 14, 8847 – 8860;
microtermolides/vinylamycin: d) G. Carr, M. Poulsen, J. L.
Klassen, Y. Hou, T. P. Wyche, T. S. Bugni, C. R. Currie, J.
Clardy, Org. Lett. 2012, 14, 2822 – 2825; LL15G256g: e) S. Li, S.
Liang, Z. Xu, T. Ye, Synlett 2008, 569 – 574.
[12] (2R,3S)-HOAsn: a) Z. Hojati, C. Milne, B. Harvey, L. Gordon,
M. Borg, F. Flett, B. Wilkinson, P. J. Sidebottom, B. A. M. Rudd,
M. A. Hayes, C. P. Smith, J. Micklefield, Chem. Biol. 2002, 9,
1175 – 1187; (2R,3R)-HOAsn: b) I. Bonnard, M. Rolland, J.-M.
Salmon, E. Debiton, C. Barthomeuf, B. Banaigs, J. Med. Chem.
2007, 50, 1266 – 1279; c) A. Kato, S. Nakaya, Y. Ohashi, H.
Hirata, J. Am. Chem. Soc. 1997, 119, 6680 – 6681.
Scheme S1 in the Supporting Information). Cell-based eval-
uation of WY1 revealed that this minimal APD-cyclolipo-
depsipeptide (APD-CLD) is sufficient to afford hypoxia-
selective growth inhibition. Although the potency is approx-
imately 10-fold reduced during hypoxia (GI50 = 352 Æ 21 nm,
Figure 2D) compared to rakicidin A, WY1 is completely
inactive under normoxic conditions up to 25 mm (SI > 70 Æ
3.8). This remarkable increase in selectivity clearly indicates
that rakicidin A has at least two separable mechanisms of
toxicity. Finally, we evaluated the effect of augmenting the
hypoxia-induced reductive stress through co-treatment with
the glutathione-precursor N-acetylcysteine (NAC, Fig-
ure 2E). We found that the activity of tirapazamine in
hypoxic PANC-1 cells was not further enhanced by NAC
but that the activity of rakicidin A was dramatically reduced.
The mechanistic basis for this finding remains to be deter-
mined.
In conclusion, we have developed a modular asymmetric
synthesis of rakicidin A and have discovered a simplified
analogue with enhanced hypoxia-selective activity. Our
synthetic route includes novel solutions to the preparation
of the key functionalities. Rakicidin A appears to act through
a non-canonical pathway upon which hypoxic cancer cells
show an increased dependence.
[13] F. Sang, D. Li, X. Sun, X. Cao, L. Wang, J. Sun, B. Sun, L. Wu, G.
Yang, X. Chu, J. Wang, C. Dong, Y. Geng, H. Jiang, H. Long, S.
Chen, G. Wang, S. Zhang, Q. Zhang, Y. Chen, J. Am. Chem. Soc.
2014, 136, 15787 – 15791.
[14] N. Oku, S. Matoba, Y. M. Yamazaki, R. Shimasaki, S. Miyanaga,
Y. Igarashi, J. Nat. Prod. 2014, 77, 2561 – 2565.
[15] The APD functionality is likely biosynthesized from serine or
cysteine: a) J. C. Albright, A. W. Goering, J. R. Doroghazi,
W. M. Metcalf, N. L. Kelleher, J. Ind. Microbiol. Biotechnol.
2014, 41, 451 – 459. Construction of the APD functionality in
a simplified model system by elimination from a cysteine
precursor: b) L. L. Clement, M. Tsakos, E. S. Schaffert, C.
Scavenius, J. J. Enghild, T. B. Poulsen, Chem. Commun. 2015, 51,
12427 – 12430.
[16] M. Tsakos, E. S. Schaffert, L. L. Clement, N. L. Villadsen, T. B.
Poulsen, Nat. Prod. Rep. 2015, 32, 605 – 632.
[17] N. Bionda, M. Cudic, L. Barisic, M. Stawikowski, R. Stawikow-
ska, D. Binetti, P. Cudic, Amino Acids 2012, 42, 285 – 293.
[18] D. A. Evans, M. D. Ennis, D. J. Mathre, J. Am. Chem. Soc. 1982,
104, 1737 – 1739.
Acknowledgements
This work was generously supported by the Lundbeck
foundation, the Danish Cancer Society and the Carlsberg-
foundation. Christopher Loveridge is acknowledged for
technical assistance. Access to the 950 MHz NMR spectrom-
eter at the Danish Center for Ultrahigh-Field NMR Spec-
troscopy (Ministry of Higher Education and Science grant
AU-2010-612-181) is acknowledged.
Keywords: antitumor agents · cyclolipodepsipeptides ·
natural products · total synthesis · tumor hypoxia
How to cite: Angew. Chem. Int. Ed. 2016, 55, 1030–1035
Angew. Chem. 2016, 128, 1042–1047
[1] P. Vaupel, A. Mayer, Cancer Metastasis Rev. 2007, 26, 225 – 239.
[2] a) E. H. Gort, G. van Haaften, I. Verlaan, A. J. Groot, R. H. A.
Plasterk, A. Shvarts, K. P. M. Suijkerbuijk, T. van Laar, E.
van der Wall, V. Raman, et al., Oncogene 2008, 27, 1501 – 1510;
b) M.-H. Yang, M.-Z. Wu, S.-H. Chiou, P. -M. Chen, S.-Y. Chang,
C.-J. Liu, S.-C. Teng, K.-J. Wu, Nat. Cell Biol. 2008, 10, 295 – 305.
[3] a) C. Koumenis, C. Naczki, M. Koritzinsky, S. Rastani, A. Diehl,
N. Sonenberg, A. Koromilas, B. G. Wouters, Mol. Cell. Biol.
2002, 22, 7405 – 7416; b) D. Cojocari, R. N. Vellanki, B. Sit, D.
Uehling, M. Koritzinsky, B. G. Wouters, Radiother. Oncol. 2013,
108, 541 – 547.
[19] J. R. Parikh, W. v. E. Doering, J. Am. Chem. Soc. 1967, 89, 5505 –
5507.
[20] a) W. Oppolzer, C. Starkemann, I. Rodriguez, G. Bernardinelli,
Tetrahedron Lett. 1991, 32, 61 – 64; b) W. Oppolzer, P. Lienard,
Tetrahedron Lett. 1993, 34, 4321 – 4324. Earlier pivotal studies:
c) H. Danda, M. M. Hansen, C. H. Heathcock, J. Org. Chem.
1990, 55, 173 – 181; d) M. A. Walker, C. H. Heathcock, J. Org.
Chem. 1991, 56, 5747 – 5750. An alternative reductive aldol
procedure: e) C. Allais, P. Nuhant, W. R. Roush, Org. Lett. 2013,
15, 3922 – 3925.
[21] Alternative methods that enable the stereocontrolled prepara-
tion of the anti-syn stereotriad. Crotyl boranes: a) H. C. Brown,
K. S. Bhat, R. S. Randad, J. Org. Chem. 1989, 54, 1570; b) W. R.
Roush, A. D. Palkowitz, K. Ando, J. Am. Chem. Soc. 1990, 112,
6348 – 6359. See also: c) W. R. Roush, J. Org. Chem. 1991, 56,
4151 – 4157.
[4] a) B. G. Wouters, M. Koritzinsky, Nat. Rev. Cancer 2008, 8, 851 –
864; b) A. M. Arsham, J. J. Howell, M. C. Simon, J. Biol. Chem.
2003, 278, 29655 – 29660.
[5] a) D. T. Jones, A. L. Harris, Expert Opin. Ther. Targets 2012, 16,
463 – 480; b) T. H. Scheuermann, Q. Li, H.-W. Ma, J. Key, L.
Zhang, R. Chen, J. A. Garcia, J. Naidoo, J. Longgood, D. E.
Frantz, et al., Nat. Chem. Biol. 2013, 9, 271 – 276; c) E. Miranda,
I. K. Nordgren, A. L. Male, C. E. Lawrence, F. Hoakwie, F.
[22] a) A. B. Northrup, D. W. C. MacMillan, J. Am. Chem. Soc. 2002,
124, 6798 – 6799; b) A. B. Northrup, I. K. Mangion, F. Hettche,
1034
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1030 –1035

Angewandte
Communications
Chemie
D. W. C. MacMillan, Angew. Chem. Int. Ed. 2004, 43, 2152 –
and co-workers (Ref. [13]) for their mesylation-elimination
2154; Angew. Chem. 2004, 116, 2204 – 2206.
[23] B. France, V. Bruno, I. Nicolas, Tetrahedron Lett. 2013, 54, 158 –
161.
[24] a) D. J. Schauer, P. Helquist, Synthesis 2006, 3654 – 3660;
b) M. C. Pirrung, G. Biswas, T. R. Ibarra-Rivera, Org. Lett.
2010, 12, 2402 – 2405; c) B. J. Larsen, Z. Sun, P. Nagorny, Org.
Lett. 2013, 15, 2998 – 3001.
[25] We attribute the minor diastereomer in compounds 2 and 22 to
partial epimerization of the a-amidoaldehyde (ultimately C11)
during the macrocyclization, although this was not directly
verified.
sequence on the C11 epimer of alcohol 22. In our hands,
attempted dissolution of this TLC-pure material in [D6]DMSO
afforded a > 90% pure NMR spectrum of rakicidin A but nearly
3/4 of the material (by mass) did not dissolve and thus very likely
represents a degraded compound.
[28] a) M. Groll, B. Schellenberg, A. S. Bachmann, C. R. Archer, R.
Huber, T. K. Powell, S. Lindow, M. Kaiser, R. Dudler, Nature
2008, 452, 755 – 758; b) D. Krahn, C. Ottmann, M. Kaiser, Nat.
Prod. Rep. 2011, 28, 1854 – 1867.
[26] H. Ogura, O. Sato, K. Takeda, Tetrahedron Lett. 1981, 22, 4817 –
4818.
[27] We have isolated material by FC (pure by TLC) that would
equate a yield of 47%, which is similar to that reported by Chen
Received: October 23, 2015
Published online: December 4, 2015
Angew. Chem. Int. Ed. 2016, 55, 1030 –1035
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1035