Total synthesis and biological evaluation of pederin, psymberin, and highly potent analogs

Article abstract of DOI:10.1021/ja207331m

The potent cytotoxins pederin and psymberin have been prepared through concise synthetic routes (10 and 14 steps in the longest linear sequences, respectively) that proceed via a late-stage multicomponent approach to construct the N-acyl aminal linkages.

Full text of DOI:10.1021/ja207331m

ARTICLE
pubs.acs.org/JACS
Total Synthesis and Biological Evaluation of Pederin, Psymberin,
and Highly Potent Analogs
||
Shuangyi Wan,†,‡ Fanghui Wu,†,‡ Jason C. Rech,†,§ Michael E. Green,†, Raghavan Balachandran,^
W. Seth Horne,*,† Billy W. Day,*,†,^ and Paul E. Floreancig*,†
† Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260, United States
^Department of Pharmaceutical Sciences, University of Pittsburgh, Pittsburgh, Pennsylvania 15260, United States
S Supporting Information
b
ABSTRACT: The potent cytotoxins pederin and psymberin
have been prepared through concise synthetic routes (10 and 14
steps in the longest linear sequences, respectively) that proceed
via a late-stage multicomponent approach to construct the N-
acyl aminal linkages. This route allowed for the facile prepara-
tion of a number of analogs that were designed to explore the
importance of the alkoxy group in the N-acyl aminal and
functional groups in the two major subunits on biological
activity. These analogs, including a pederin/psymberin chimera,
were analyzed for their growth inhibitory effects, revealing
several new potent cytotoxins and leading to postulates regard-
ing the molecular conformational and hydrogen bonding
patterns that are required for biological activity. Second generation analogs have been prepared based on the results of the initial
assays and a structure-based model for the binding of these compounds to the ribosome. The growth inhibitory properties of these
compounds are reported. These studies show the profound role that organic chemistry in general and specifically late-stage
multicomponent reactions can play in the development of unique and potent effectors for biological responses.
’ INTRODUCTION
these compounds induce apoptosis in a number of cell lines by
activating the c-Jun kinase (JNK) and the p31 mitogen-activated
protein kinase.11 Usui and co-workers showed12 that psymberin
induces a similar JNK activation, that the activation is a response
to the accumulation of reactive oxygen species in the mitochon-
dria, and that apoptosis is at least partially induced by caspase-8.
Ribosome binding was demonstrated13 by the displacement
of radiolabeled 13-deoxytedanolide (independently shown to
be a ribosome binding molecule) from the 60S subunit by
pederin, and by the crystal structure of mycalamide A in the
ribosomal binding site for the CCA end of tRNAs that occupy
the E-site.14
Pederin1 (1) and psymberin,2 also known as irciniastatin A,3
(2) are potent cytotoxins that contain a densely functionalized
tetrahydropyran subunit and an N-acyl aminal linkage. These
structural features are also found in the mycalamide/theopeder-
in/onnamide class of molecules, represented in Figure 1 by
mycalamide A (3) and onnamide A (4).4 While the mycalamides,
theopederins, onnamides and psymberin were isolated from
marine sponges, pederin was originally isolated from the Paederus
beetle. The Piel group has explained this curious observation by
demonstrating that these compounds are actually synthesized by
symbiotic bacteria.5
Pederin was shown6 to be a mitotic poison in 1966, but further
studies on the biological activity of this family of compounds
were not reported until 1989,7 when the mycalamides were
identified as potent cytotoxins that showed efficacy in vivo
against leukemia and solid tumor models. This study also
demonstrated that these compounds are protein synthesis in-
hibitors. Ogawara and co-workers showed8 that the mycalamides
change ras-transformed cells to normal morphology and corre-
lated this activity to the inhibition of the synthesis of p21, a
cyclin-dependent kinase inhibitor that regulates cell growth. The
Kocienski group reported9 that pederin and the mycalamides
induce necrosis in squamous carcinoma cells but not in fibro-
blasts. Further mechanistic work10 led to the conclusion that
The interesting biological activity and unique structures of
these compounds, coupled with their scarcity from natural
sources, has resulted in a number of total, formal, and partial
syntheses of pederin,15 the mycalamides,16 and psymberin.17 Our
interest in this molecular class arose from our efforts18 in
oxidative approaches to cyclic acyl aminal formation. While this
work culminated in the total synthesis of theopederin D,16k we
were cognizant that the approach we developed to prepare the
amido trioxadecalin unit in theopederin D would be ineffective
for constructing the N-acyl aminal groups in pederin and
Received: August 4, 2011
Published: September 08, 2011
r
2011 American Chemical Society
16668
dx.doi.org/10.1021/ja207331m J. Am. Chem. Soc. 2011, 133, 16668–16679
|

Journal of the American Chemical Society
ARTICLE
Scheme 2. Synthesis of the Right Fragments for Pederin and
Analogs
Figure 1. Cytotoxic N-acyl aminal-containing natural products.
Scheme 1. Matsuda’s Epimerization Studies and a Multi-
component Approach to N-Acyl Aminal Construction
’ RESULTS AND DISCUSSION
Subunit Syntheses. The synthesis of the right fragment of
pederin and its analogs is shown in Scheme 2.22 The sequence
began with the asymmetric allylation of keto aldehyde 5,23
prepared on multigram scale from the condensation of acetyl
chloride with the morpholine enamine of isobutyraldehyde.
Several methods were shown to be successful for this transfor-
mation. The Leighton allylation24 provided superb enantioselec-
tivity and yield, though accessing sufficient quantities of the
pseudoephedrine-based reagent for large scale studies proved to
be difficult. The abundance of tartaric acid derivatives rendered
the Roush allylation25 a more practical alternative for large scale
reactions, though the enantioselectivity was lower and reagent
preparation required significant effort. The Krische allylation26
ultimately provided an ideal solution with respect to reagent
availability and reaction enantioselectivity. Exposing 5 to allyl
acetate, under reductive conditions (iPrOH, [Ir(cod)Cl]2, (R)-
Cl,MeO-BIPHEP, Cs2CO3, m-NO2BzOH, THF) provided al-
cohol 6 in 71% yield and 93% ee. Thus a key early stage
intermediate that is required for the syntheses of all molecules
in this study can be accessed from inexpensive precursors on
multigram scale by using asymmetric catalysis. Allylation of the
primary alcohol corresponding to 4 was possible under these
conditions in the absence of iPrOH, though the overall efficiency
from the aldehyde proved to be superior. Conversion of the
alcohol to silyl ether 7a, the required intermediate for the total
synthesis, and to methyl ether 7b, an intermediate for analog
psymberin. Inspired by Matsuda’s studies15c on acidic MeOH-
mediated N-acyl aminal stereochemical inversion in advanced
pederin intermediates, we envisioned an approach to these
compounds that would proceed through a late stage addition
of MeOH to an acylimine that would arise from the union of the
two major subunits of the natural products. The acylimine would
arise through a sequence of nitrile hydrozirconation19 and
acylation (Scheme 1).20 We have applied21 the multicomponent
sequence of nitrile hydrozirconation, acylation, and nucleophilic
addition to the synthesis of several amide structural classes. In
this manuscript we provide a full account of our application of the
multicomponent acyl aminal construction to the total syntheses
of pederin, psymberin, and analogs. The abilities of these
compounds to inhibit cell growth have been determined, and
many of the analogs are extremely potent cytotoxins. The results
of these assays were used to generate postulates regarding the
roles that several structural features in these molecules play in
causing the biological response. These postulates guide the
design of second generation analogs that were synthesized and
evaluated for activity.
16669
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679

Journal of the American Chemical Society
ARTICLE
synthesis, proceeded readily under standard conditions. The
resulting ketones are well-suited for chain extension through
aldol reactions. Numerous studies have shown that boron
enolates of β-alkoxy methyl ketones react with aldehydes to
provide good levels of 1,5-asymmetric induction,27 as required
for pederin synthesis. Unfortunately the diastereomeric ratios
that we observed for reactions between the dibutylboron or the
dicyclohexylboron enolates of 7a and 7b never exceeded 3:1. We
therefore explored the pinene-derived boron enolates that have
been reported by the Paterson group.28 These readily prepared
compounds substantially enhanced the stereoselectivity of the
process to provide product diastereomeric ratios in excess of
15:1. Purification of the product from the pinene-derived by-
product proved to be difficult, however, so we exploited the
stability of chelated structure 8, the initial product of the aldol
reaction, to effect a stereoselective reduction with LiBH4.29
These one flask aldol/reduction reactions provided readily
purified diols 9a and 9b in 80% yield as a single stereoisomer
and 78% yield as a 13:1 mixture of stereoisomers, respectively.
Although we observed low levels of stereoinduction from the β-
substituent and enol diiopinocamphenylborinates from methyl
ketones have been shown to react with moderate levels of
control, the reenforcing effects led to very satisfying levels of
stereocontrol. Ozonolytic alkene cleavage provided lactols 10a
and 10b in excellent yield. Cyano group incorporation proceeded
without converting the anomeric hydroxyl group to a leaving
group or protecting the C17 by ionizing the lactol with BiBr330 in
the presence of TMSCN. Yields of nitriles 11a and 11b were
low, however, because of the competitive formation of bis-silyl
ether 12. No reaction was observed when 12 was isolated and
resubjected to the reaction conditions, indicating that the
efficiency of the process was compromised by forming the bis-
silyl ether or that anhydrous HBr, formed through the reaction of
BiBr3 with the hydroxy groups,31 is the relevant catalyst in the
early stages of the reaction. We reasoned, however, that a
stronger Lewis acid would be able to induce ionization of 12.
Scheme 3. Synthesis of the Right Fragment of Psymberin
protected as silyl ethers under standard conditions and the
sterically less hindered aliphatic ester group reacted with one
equiv of DIBAL-H to provide aldehyde 17. A Brown crotyla-
tion37 to yield 18 in suitable yield and enantioselectivity followed
by protection of the resulting hydroxyl group as a TBS ether and
ozonlytic cleavage of the alkene produced aldehyde 19. The
boron enolate of 7a did not react with 19 smoothly, presumably
as a result of enhanced steric hindrance in comparison to the
reaction with methoxy acetaldehyde. Therefore we pursued an
approach in which the chirality of the aldehyde controls the
approach of an enolsilane nucleophile. Applying the Felkin-Anh
model38 to 19 leads to the prediction that nucleophilic addition
should provide the desired 1,2-syn-stereochemical relationship.
Evans’ polar extended Felkin model,39 however, predicts that the
silyloxy group at the β-position of the aldehyde should guide
nucleophiles to produce the undesired 1,3-anti-stereoisomer.
Adding enolsilane 20, prepared in quantitative yield from 7a
Therefore, we added BF3 OEt2 to the reaction mixture after the
3
initial process had occurred. This led to suitable yields of nitriles
11a and 11b. Directly subjecting 10a or 10b to BF3 OEt2 and
3
TMSCN led to decomposition, indicating that C13 silyl ether
formation through the BiBr3-mediated pathway is a prerequisite
for efficient cyanation. The stereochemical outcomes of thse
reactions are consistent with Woerpel’s studies32 of cyclic
oxocarbenium ion cyanation in acetonitrile. Methylation of the
secondary alcohols completed the syntheses of multicomponent
reaction substrates 13a and 13b. Competitive silyl transfer
plagued the methylation of 13a under standard Williamson
etherification conditions (NaH, DMF, then MeI), causing us
to employ MeOTf and di-tert-butylpyridine in CH2Cl2 for this
transformation. The Williamson conditions were appropriate for
the more robust substrate 13b.
under standard conditions, to 19 in the presence of BF3 OEt2
3
produced aldol product 21 in excellent yield as an inseparable 6:1
mixture of diastereomers that favored the syn, syn-stereoisomer.
This result was consistent with observations from the Evans
group40 that Felkin-Anh selectivity overrides β-alkoxy induction
when sterically hindered enolsilanes are used as nucleophiles.
The resulting hydroxy ketone was reduced with NaBH4 and
Et2BOMe41 to provide the expected diol, which could be isolated
as a single stereoisomer with high syn-control being confirmed by
forming the acetonide and analyzing its 13C NMR spectrum.42
As in Scheme 2, ozonolysis was employed to form lactol 22.
The BiBr3-mediated lactol cyanation conditions from Scheme 2,
however, proved to be ineffective in this system and led to sub-
stantial decomposition. The strategy of acylating both alcohols
The synthesis of the right fragment for psymberin is shown in
Scheme 3.33 An efficient synthesis of this unit requires a rapid
construction of the pentasubstituted arene. Rather than modify
commercially available arenes, we chose to employ a cycloaddi-
tion-based approach to this unit based on work from the Langer
group.34 Thus, allene 1435 and diene 15,36 each available in one
step from commercially available materials, were mixed in the
absence of solvent to form a cycloadduct that, upon exposure to
Et3N HF, directly yielded 16 in 70% yield. This procedure was
3
very attractive at this point in the synthesis because it could
readily be run on large (30 g) scale. The hydroxyl groups were
16670
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679

Journal of the American Chemical Society
ARTICLE
Scheme 4. Synthesis of Pederic Acid and Des-methylene
Pederic Acid
Scheme 5. Synthesis of Psymberic Acid
Scheme 6. Synthesis of Pederin and Acyl Aminal Analogs
and selectively ionizing to yield the tetrahydropyranyl cation33
worked well, but the subsequent multicomponent reaction was
not compatible with the presence of the acetoxy group at C15.
The observation that the anomeric hydroxyl group undergoes
acylation much faster than the C15 hydoxyl group led us to
conduct a selective acylation followed by trimethylsilyl ether
formation at C15 to form 23. Cyanation of 23 with TMSOTf
provided 24, the key fragment for the multicomponent reaction,
in 88% yield.
Numerous approaches to the acyl fragment of pederin, in
addition to those that were developed for the total syntheses,15,16
have been reported.43 Asymmetric induction in these sequences
invariably arose from the use of stoichiometric chiral reagents or
auxiliaries, or from chiral pool starting materials. Our objective
for the synthesis of this subunit was to develop a route that relies
upon asymmetric catalysis to set the absolute stereochemistry.44
We based our approach on the excellent route from the Nakata
group,43c in which asymmetric induction arose from the use of an
Evans aldol reaction. β-Lactones are effective surrogates for aldol
products that can be prepared enantio- and diastereoselectively
through the condensation of aldehydes with acid chlorides in the
presence of cinchona alkaloid derivatives.45 Thus acetaldehyde
and propionyl chloride were combined in the presence of Et3N,
trimethylsilyl quinidine, and LiClO4 to yield a volatile β-lactone
that was exposed to the lithium enolate of tert-butyl acetate to
provide keto ester 25 in 76% yield as a single stereoisomer to the
limits of gas chromatographic detection. In accord with Nakata’s
route, 25 was converted to ester 26 through a sequence of
thioacetal formation, stereoselective Claisen condensation with
enolate 2743a in the presence of ZnCl2, and acidic methanol
treatment. Benzoylation, thioacetal cleavage, and methylenation
provided benzoylated pederic ester 28. Thiolate-mediated clea-
vage produced acid 29. The presence of the exocyclic alkene in
the pederic acid unit causes instability toward a number of
common reagents. Therefore, we prepared an analog that lacks
the alkene as a potentially more versatile subunit. Desulfurization
of 26 with Raney nickel followed by benzoylation provided ester
30, which was treated with LiSPr to yield des-methylene pederic
acid derivative 31 (Scheme 4).
Pietruszka’s exceptional route46c (Scheme 5) that is based on a
stereoselective aldol reaction between β,γ-unsaturated aldehyde
32 and glycolic acid acetal 33.47 This route provided abundant
quantities of acid 34 in which the hydroxyl group at C5 was
protected as a benzoate ester, in accord with our projected final
deprotection strategy.
Fragment coupling and total synthesis. The fragment
coupling in the pederin series is shown in Scheme 6. Hydro-
zirconation of nitrile 13a with Cp2Zr(H)Cl followed by the
addition of acid chloride 35, freshly prepared from 29 prior to the
reaction and used without purification, resulted in the transient
formation of acylimine 36. The addition was conducted at ꢀ78 ꢀC
due to the propensity of 36 to undergo tautomerization to form a
stable enamide product. Competitive tautomerization was not
observed in simpler systems, even when a tetrahydropyranyl
nitrile was used as the substrate.21a Adding structurally simpler acid
chlorides to the hydrozirconation product of 13a also promoted
the formation enamide products, indicating that tautomerization
can be attributed to structural elements in the nitrile component.
Mg(ClO4)2 was added to lock the acylimine conformation by
chelation between the nitrogen of the imine and the oxygen of
the tetrahydropyran, thereby promoting nucleophilic attack through
the desired, less hindered trajectory. Alcohol addition led to the
The acyl fragment of psymberin (psymberic acid) has also been
prepared through several approaches.17,46 In consideration of our
desire to minimize late stage deprotection efforts we employed
16671
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679

Journal of the American Chemical Society
ARTICLE
Scheme 7. Variations in the Left and Right Fragments of
Pederin
converted to methyl ether analog 43. The syntheses of these
analogs demonstrates late stage that the multicomponent assem-
bly of the acyl aminal group can be used to introduce structural
variations in all sections of the natural product, thereby greatly
facilitating structureꢀactivity relationship studies.
The multicomponent approach to the acyl aminal of psymber-
in proved to be much more challenging than the corresponding
reaction in the pederin series. Difficulties in controlling the
stereochemical outcome of the reaction could be predicted based
on De Brabander’s approach to a pederin/psymberin hybrid
strucure48 in which the approach of NaBH4 to acylated imidate
intermediates that are structurally related to the acylimines in our
procedure proceeded with opposite trajectories for the pederin
and psymberin subunits. This study provided further evidence49
that remote interactions can exert a strong influence over the
reactivity at the acyl aminal site. Exposing nitrile 24 to Cp2Zr-
(H)Cl, acyating with acid chloride 44, and adding MeOH
provided diastereomeric acyl aminals 45 and 46, in a 1:3 ratio
in which the undesired stereoisomer was the major product.
Adding one equivalent of Mg(ClO4)2 improved the ratio to 1:2.
Increasing the equivalents of Mg(ClO4)2 incrementally in-
creased the stereocontrol, with 2 equivalents providing a 1:1
ratio and 10 equivalents leading to a 3:1 ratio, albeit at the
expense of the overall efficiency of the reaction. We hypothesized
that replacing MeOH with a less reactive or bulkier surrogate
would improve the stereoselectivity. Conducting the addition
with 2 eq Mg(ClO4)2 and (MeO)3CH as the source of the
methoxy group provided a 3:1 mixture of 45 and 46, as
determined by crude 1H NMR, though the yield for the
transformation was low (∼20%). Changing the Lewis acid to
Zn(OTf)2 resulted in lower stereocontrol (∼1.5:1) but higher
overall efficiency. For simplicity the crude mixture was carried to
the next step without purification. Exposing the acyl aminal
mixture to Bu4NF in DMF led to the cleavage of all silyl groups,
formation of the dihydroisocoumarin, and removal of the
benzoate to yield psymberin in 27% yield along with 8-epi-
psymberin in 12% yield for the two steps. The benzoate cleavage
most likely resulted from the generation of Bu4NOH during the
cleavage of the silyl ethers. The identity of the protecting group at
C15 proved to be critical for the success of this transformation.
Cleavage at of the silyl ether at C17 was not observed when the
C15 oxygen was protected as an acetate or a bulkier silyl ether.
We postulate that the cleavage of the labile C15 TMS ether
allows for intramolecular silyl transfer form C17. Deprotection
of the resulting C15 silyl ether then leads to the final pro-
duct. Thus psymberin is available from commercially available
materials through a route that is only 14 steps in its longest
linear sequence (28 steps total) and proceeds in an overall yield
of 4.4% (Scheme 8). As with the pederin synthesis the modest
efficiency of the multicomponent reaction did not impede
further studies since tens of milligrams of the natural product
can be readily accessed.
desired acyl aminals 37aꢀd. Ethanol and trifluoroethanol were
selected as nucleophiles to determine whether the ease of acyl
aminal ionization, a proposed mechanism of action for these
species,16a would cause a difference in biological activity for
structurally similar species. Dimethoxybenzyl alcohol was se-
lected because it provides potential access to an acyl hemiaminal
analog that should be a biosynthetic precursor to pederin. Water
could be used directly as a nucleophile in the multocomponent
reaction but the resulting N-acyl hemiaminal was not stable
toward the final deprotection conditions. An additional benefit of
the dimethyoxybenzyl analog is that the final product will provide
information on the tolerance of the binding site for larger groups
that could provide useful handles for probe development. The
yields for these reactions were moderate but the stereocontrol
was good to excellent and products were readily isolated in
sufficient quantities to complete the syntheses and conduct
biological evaluations. Reduction of acylimine 36 by residual
Cp2Zr(H)Cl to form amide 37e was a side reaction in each of
these reactions, allowing us to accrue sufficient quantities of
the amide for subsequent studies. The syntheses of pederin and
analogs were completed by a one flask protocol that was inspired
by Rawal’s synthesis15h whereby silyl group cleavage was effected
by Bu4NF in THF followed by the addition of aqueous LiOH
and MeOH to cleave the benzoate ester. Pederin (1) and analogs
38bꢀe were prepared through this route. The efficiency of the
process was high for all substrates except 37c, in which the basic
conditions promoted the elimination of the trifluoroethoxy
group to form enamide analog 39 in 29% yield as a side product.
The geometry of the enamide group was confirmed by NOESY
experiments.22 We isolated silyl ether 40, which was resistant
toward cleavage, when the sequence was conducted with a
TBDMS ether at C13 rather than a TES ether. The longest
linear sequence for these syntheses was 10 steps, and the overall
sequence required 18 steps. Pederin was prepared in 10.4%
overall yield from keto aldehyde 5.
The availability of the left and right fragments for pederin and
psymberin led us to undertake the synthesis of chimeric struc-
tures. Psympederin, a pederin/psymberin chimera that contains
the left fragment of psymberin and the right fragment of pederin,
was prepared and shown to be far less cytotoxic than either
pederin or psymberin.48 This led the authors to conclude that
pederin and psymberin do not share a common biological target.
While this conclusion is perfectly reasonable, we posited an alter-
native explanation in which the dihydroisocoumarin group in
psymberin contributes additional binding affinity to compensate
Through related sequences (Scheme 7) 13a was coupled with
acid chloride 41, prepared from 31 immediately prior to cou-
pling, to form des-methylene pederin analog 42. Nitrile 13b was
16672
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679

Journal of the American Chemical Society
ARTICLE
Scheme 8. Completion of the Psymberin Synthesis
Table 1. GI50 Values of the Natural Products and Analogs
against HCT116 Cells
entry
compound
description
GI50 (nM)
1
1
pederin
0.6 ( 0.1
0.34 ( 0.1
0.55 ( 0.02
0.32 ( 0.09
7.7 ( 0.2
2
38b
38c
38d
38e
39
42
43
40
2
10-ethoxy pederin
3
10-trifluoroethoxy pederin
10-dimethoxybenzyloxy pederin
10-desmethoxy pederin
pederin enamide
4
5
6
27 ( 0.9
7
C4-desmethylene pederin
13-O-methyl pederin
13-OTBS pederin
6.5 ( 0.8
8
0.084 ( 0.01
3.1 ( 0.1
9
10
11
psymberin
0.052 ( 0.02
0.004 ( 0.003a
48
pedestatin
a Result of averaging two independent experiments.
pedestatin (pederin + irciniastatin).50 The statin suffix was selected
as an homage to the remarkably influential contributions from the
Pettit group to the isolation and development of antineoplastic
agents.
Scheme 9. Synthesis of the PederinꢀPsymberin Chimera
Pedestatin
Cytotoxicity Studies. The potency of these compounds was
measured in cell viability studies using HCT-116 colon cancer
cells by the MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxy-
methoxy-phenyl)-2-(4-sulfophenyl-2H-tetrazolium) dye reduc-
tion assay with PMS (phenozine methosulfate) as an electron
acceptor. The assays were conducted over a three day incubation
period at 37 ꢀC and were run in triplicate or quintuplicate.
Growth inhibition (GI) was calculated as defined by the National
Cancer Institute (GI = 100(T ꢀ T0)/(C ꢀ T0), where T0 = cell
density at time zero, T = cell density after 72 h with psymberin,
and C = cell density at 72 h with vehicle). The results are shown
in Table 1.
These studies showed that all of the compounds in this series
are potent antiproliferative agents. The identity of the alkoxy
group in the acyl aminal subunit appears to exert little influence
over the biological activity, but the presence of the alkoxy group
enhances potency by at least 1 order of magnitude (entries 1ꢀ4
vs entry 5). The reasonable potency of desmethoxy analog 38e
suggests that acyl aminal ionization is not a requirement for
biological activity. The most potent pederin analog contains a
methyl ether at the C13 position rather than a hydroxyl group,
suggesting that reducing hydrophilicty at that site is beneficial to
activity. This has also been observed in psymberin analogs that
lack a substituent at that position.51 Dramatic increases to the size
and hydrophobicity at the C13 position proved to be modestly
detrimental to activity, as seen in the TBS ether analog 40 in
entry 9. Remarkably the enamide analog also showed reasonable
activity (entry 6) despite the structural changes that result from
the elimination reaction. The approximately 1 order of magni-
tude potency reduction of the des-methylene analog (entry 7),
though consistent with results from the Nakata group,52 was
somewhat surprising based on the minimal structural perturba-
tion that this change effects on the hydrogen bond donors and
acceptors in the molecule. Psymberin is more potent than
pederin (entries 1 and 10). While the GI50 value that we observed
was somewhat lower than the value in the literature for this cell
line (0.16 nM)53 the values were within a factor of 3 and were con-
sistent over several assays. The primary data are shown in Figure 2.
Chimera 48 is even more potent than psymberin (entry 11),
for a reduced binding contribution from the less-organized acyl
fragment. Therefore we prepared a new analog that contains the
left fragment of pederin and the right fragment of psymberin
(Scheme 9). The multicomponent reaction employed nitrile 24
and acid chloride 35. MeOH was a suitable nucleophile in this
reaction since stereocontrol did not prove to be a problem when
2 equivalents of Mg(ClO4)2 were employed. Crude product 47
was subjected to Bu4NF in DMF to cleave the silyl ethers and
form the dihydroisocoumarin unit. The benzoate ester was only
partially cleaved under these conditions, in contrast to our
observations with the less hindered ester in psymberin. The
crude mixture was treated with LiOH in MeOH and H2O to
complete the synthesis of chimera 48 that we have named
16673
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679

Journal of the American Chemical Society
ARTICLE
Figure 3. Binding models of mycalamide A and pederin bound to the
ribosome. (A) Revised structure of mycalamide A based on a reinterpreta-
tion of the electron density from the crystal structure. (B) Modeled structure
of pederin derived from mutating the structure of mycalamide A.
at C10. Since the prior study reported that material of natural
origin was used in crystallization,14 we reinterpreted the pub-
lished electron density using coordinates for the correct
diastereomer (see Supporting Information). With the refined
structure of the mycalamide A/ribosome complex in hand
(Figure 3A), pederin was docked into the binding site by
superposition of the backbone atoms shared between the two
molecules (Figure 3B). Analysis of the resulting model pro-
vides several insights into the structural basis for the biological
activity of a variety of natural N-acyl aminal-containing
cytotoxins and their synthetic analogs.
The potency of 10-desmethoxy pederin (38e) was somewhat
surprising in consideration of reports that related compounds
that lack oxygenation55 or are epimeric at C109 show significantly
diminished activity. An examination of the bioactive conforma-
tion of pederin in the bound structure suggests multiple roles for
oxygenation at C10 (Figure 4). Foremost, the C10 alkoxy group
promotes a conformation about the C10ꢀC11 bond that
matches the ribosome-bound structure by forcing the less bulky
C10-hydrogen to occupy the position over the tetrahydropyran
ring (Figure 4A). Evidence for this conformational preorganiza-
tion in solution is provided by the large coupling constant (∼8 Hz)
between the C10 and C11 hydrogens in the 1H NMR spectrum
of pederin. Epimerization at C10 forces the amide group into an
unproductive binding orientation to relieve the energetic penalty
of placing the methoxy group over the tetrahydropyran ring. The
10-deoxy compound lacks a conformational bias around the
C10ꢀC11 bond. This loss of conformational preorganization is a
Figure 2. Growth inhibition data for psymberin against HCT-116 cells.
(Top) Growth inhibition vs log[psymberin], where [psymberin] is
reported as nM. (Bottom) Growth inhibition in the responsive range
of concentrations.
showing a 4 pM GI50 value. In addition to the remarkable potency,
this result is quite significant in consideration of De Brabander’s
results47 showing that a chimera containing the left fragment of
psymberin and the right fragment of pederin is not a potent
cytotoxin. Our results suggest that pederin and psymberin share a
common binding site on the ribosome and that the left fragment
of pederin and the right fragment of psymberin are the essential
components for the biological activity of these natural products.
Compounds 1, 38bꢀe, 39, 42, and 43 were tested against a p53
knockout variant of the HCT116 cell line54 and showed essen-
tially identical activity. These results indicate that the apoptotic
pathway is not dependent on p53.
Structural Basis for Biological Activity. The published
crystal structure of the complex between mycalamide A (3)
and the large ribosomal subunit14 provides an opportunity to
identify structural bases for the differences in biological activity
among the molecules reported here. Such analysis should prove
to be valuable in the design of second-generation analogs with
superior activity or greater accessibility. Given the fact that the
structures of mycalamide A and pederin are closely related, we
reasoned that the two natural products share similar binding modes.
The coordinates published for mycalamide A in complex with
the ribosome14 differ from the natural product by epimerization
16674
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679

Journal of the American Chemical Society
ARTICLE
Figure 5. Exocyclic alkene at C4 in the left-hand fragment of pederin
intercalates between two sequential bases in the ribosome.
Figure 6. Geminal methyl groups in the right-hand tetrahydropyran
ring in pederin restrict rotation about C15ꢀC16 and thereby the
orientation of the branched side chain (R0).
removal of ring conformational restraints (as in psymberic acid)
attenuates the binding contribution of the acyl fragment and
reduces activity. The pocket between the nucleobases appears
sufficiently large to accommodate the products of exocyclic
alkene reduction, which have been shown56 to retain cytotoxic
activity.
The lack of hydrogen-bonding contacts involving the C13
hydroxyl group suggests that it serves no role in ribosome
binding. This is consistent with results from the Schering
group51 showing that deoxygenation at the corresponding posi-
tion of psymberin actually increased activity. Replacement of the
C13 hydroxyl group in pederin with a methyl ether led to the
highly cytotoxic analog 43. Taken together, these observations
suggest that increasing the hydrophobicity at C13 could be a
general strategy to enhance potency in this family of natural
products.
Figure 4. C10 methoxy group in pederin plays multiple roles in
ribosome binding. (A) C10 substituent acts as a conformational control
element for rotation about the C10ꢀC11 bond; (B) pederin binding
pocket has space to accommodate larger substituents at C10; (C) C10
oxygen forms a hydrogen-bond to a nearby ribosome nucleobase.
likely contributor to the weaker activity observed for 10-des-
methoxy pederin compared to the natural product.
Two other features around C10 in the model for pederin
bound to the ribosome are noteworthy. The binding region for
the C10 methoxy group appears capable of accommodating
larger substituents (Figure 4B), in accord with our observation
that larger alkoxy groups do not reduce activity. The structure
also shows a hydrogen bond between the oxygen at C10 and the
amino group of a nearby guanine residue (Figure 4C). Thus, the
role of the alkoxy group may transcend simple conformational
restriction by providing direct contacts that enhance binding
affinity. Collectively, the results of our structural analysis provide
a picture of the role of the C10 substituent that is consistent with
available biological data for pederin and its analogs. Closer
examinations of the known natural products that lack oxygen at
the C10 position reveal the presence of other structural perturba-
tions that are the likely source of their diminished biological activity.
The origin of the strong influence of the pederic acid compo-
nent on pederin bioactivity, and the exocyclic alkene in particular,
can be proposed based on the bound structure. In the model of
the pederin/ribosome complex, the exocyclic alkene of the
pederic acid ring intercalates between adjacent adenine and
guanine residues in the binding pocket (Figure 5). We propose
that this interaction acts as an anchor that organizes other
functional groups of the acyl fragment into a suitable binding
orientation. Deletion of the exocyclic methylene (as in 42) or
We propose that the geminal methyl groups at C14 set the
orientation of the branched side chain projecting off the right
fragment of pederin (Figure 6). Conformational constraint of the
C15ꢀC16 bond arises from the minimization of the syn-pentane
interactions that would be present in conformations that are not
relevant for binding. The existence of this conformational con-
1
straint in solution is supported by observed H NMR coupling
constants of 10.4 and 1.8 Hz for the C15 hydrogen that indicate a
rigid conformation with one anti and one gauche relationship to
the hydrogens on C16. Thus, the tetrahydropyran ring appears to
serve as a scaffold to align the left and right arms into the proper
binding orientation.
Although we have no information regarding the bound con-
formation of psymberin, we note the presence of a sizable cavity
adjacent to the end of the right fragment of pederin (Figure 7)
that could accommodate the dihydroisocoumarin group of
psymberin and the long chain of the onnamides.4b,e Further
crystallographic studies are required to determine the precise
structural basis for the potency enhancement that the dihydro-
isocoumarin group effects.
16675
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679

Journal of the American Chemical Society
ARTICLE
Scheme 11. Synthesis of Alkyl Branched Amide Analogs
Figure 7. Unoccupied pocket adjacent to the right-hand fragment of
pederin suggests a possible basis for the enhanced efficacy of psymberin
and pedestatin.
three step sequence in which only one quick filtration through
silica gel was required. A similar sequence was executed with
pederic acid derivative 30. An additional step was required to
complete the cleavage of the sterically hindered benzoate group,
similar to the synthesis of chimera 48. Through this route 10-
desmethoxy pedestatin 50 was prepared, again in 47% overall
yield and with minimal purification. We also constructed a
pederin analog that lacks the exocyclic alkene at C4 and the
C10 methoxy group to facilitate synthesis and handling, and
contains the C13 methyl ether to compensate for the potency
loss that these changes would accrue. Reduction of nitrile 12b
followed by coupling of the resulting amine with desmethylene
pederic acid derivative 31 in the presence of EDC and HOBt
provided the desired amide in 65% yield. Cleavage of the
benzoate group provided the desired analog in 87% yield.
The final set of analogs was designed to test our hypotheses
regarding the importance of conformation constraint and a
hydrogen bonding interaction between the C10 methoxy group
and the ribosome. This was achieved by replacing the methoxy
group with a methyl group (Scheme 11). Exposing nitrile 13b to
MeMgBr followed by quenching with MeOH and reducing the
resulting imine with NaBH458 provided amine 52 as an insepar-
able 1.5:1 mixture of diastereomers (as determined by NMR).
Acylation of the mixture with acid 31 followed by benzoate
cleavage with LiOH provided methyl analogs 53 and 54 in 31 and
26% yields, respectively, for the three-step sequence. The
stereochemical assignments for these structures were based on
the observation of a cross peak between the C10 methyl group
and the equatorial hydrogen of C12 in the NOESY spectrum of
53 and between the C10 methyl group and the C15 hydrogen in
the NOESY spectrum of 54.
Biological Evaluation of Second Generation Analogs. The
second generation analogs were evaluated for their ability to
inhibit HCT116 cell growth through the same protocol that was
used to evaluate the initial set of compounds. The results are
shown in Table 2.
The results of these studies revealed some highly valuable informa-
tion. The potency of the psymberin and pedestatin cores is
sufficient to provide extremely strong cytotoxicity even in the
absence of the N-acyl aminal subunit. These results could be
useful if large scale preparation of these agents is desired since
catalytic nitrile hydrogenation is more economical and experi-
mentally facile than the nitrile hydrozirconation sequence. The
significantly improved activity of pederic acid-containing analog
50 in comparison to psymberic acid-containing analog 49 again
Scheme 10. Synthesis of Desmethoxy Analogs
Second Generation Analogs. The results from the initial
round of biological assays and the structural details that were
gleaned from the modeling studies provide a basis for the design
and synthesis of a second wave of analogs. These compounds
were developed to test hypotheses regarding ribosome binding
and/or to facilitate synthesis.
The respectable activity of C10-desmethoxy analog 38e led us
to consider other analogs that lack oxygenation at the C10 site
(C8 site of psymberin). Since 38e was formed from side-
products of the multicomponent reaction we changed the route
so that the nitrile was reduced to a primary amine through
catalytic hydrogenation followed by conventional amide cou-
pling (Scheme 10). The reduction of nitrile 24 proceeded with
H2 (1 atm) in the presence of a mixture of Pd/C and PtO2 in
HOAc and EtOAc57 to provide the corresponding amine. Coupling
the crude amine to psymberic acid derivative 34 with EDC and
HOBt led to the amide, which was treated with Bu4NF in DMF
to yield 8-desmethoxy psymberin, 49. The yield was 47% for the
16676
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679

Journal of the American Chemical Society
ARTICLE
Table 2. GI50 Values for Second Generation Analogs against HCT116 Cells
entry
compound
description
GI50 (nM)
1
2
3
4
5
49
50
51
53
54
8-desmethoxy psymberin
0.83 ( 0.1
0.068 ( 0.02
79 ( 8
10-desmethoxy pedestatin
4-desmethylene-10-desmethoxy 13-O-methyl pederin
4-desmethylene-10-(S)-methyl 13-O-methyl pederin
4-desmethylene-10-(R)-methyl 13-O-methyl pederin
42 ( 5
>9000
highlights the importance of structural organizization in this part
of the molecule. Additional evidence for this point is the loss of
activity for 51. Removing two structural elements that promote
ribosome binding amplified the detrimental effects, even in the
presence of the activity-enhancing C13 methyl ether. C10-
Methyl ether analogs 53 and 54 show the essential role that
the conformational organization that was illustrated in Figure 4
plays in biological activity. The 10-(S) isomer 53 shows slightly
enhanced activity in comparison to the desmethoxy analog 51
while the 10-(R) isomer 54 is completely inactive because of the
energetic penalty that results from placing the methyl group over
the tetrahydropyran ring in the biologically active conformation.
The diminished activity of 53 in comparison to C4-desmethylene
pederin (42) also shows that either the hydrogen bond to the
guanine residue in the ribosome (Figure 4) adds to the activity,
that a steric clash between the alkyl group and this residue
diminishes activity, or both. Regardless of the exact origin of the
differential activity, these results suggest that the indirect biosyn-
thetic sequence of oxidizing a glycine subunit to an N-acyl aminal
as opposed to using an α-alkyl amino acid subunit provides the
producing organism with a superior defensive agent. The ques-
tion of whether the enhanced activity resulted from evolutionary
pressure or through chance must remain open to speculation.
have been recorded, including the natural product spongistatin-
1
60 and the synthetic agent meayamycin B.61
The biological results were analyzed in conjunction with a
model for the complex between pederin and the ribosome that
was derived from a crystal structure between the structurally
related natural product mycalamide A and the ribosome. This
analysis identified the structural basis for several of our observa-
tions and showed that the main function of the central tetra-
hydropyran ring is to act as a scaffold to project the left and right
fragments of the molecules into the proper binding orientations.
Second generation analogs were designed from our initial con-
clusions and showed that compounds in which the C10/C8
alkoxy group is deleted are still quite potent when they contain
the appropriate left and right fragments. Additionally replacing
the C10 alkoxy group of pederin with a methyl group causes a
reduction in activity as a result of losing a hydrogen-bonding
interaction, accruing a steric interaction, or both. This study
highlights the uniquely powerful role that chemical synthesis in
general and specifically late stage multicomponent reactions can
play in understanding the structural basis for and improving upon
the biological activity of natural products and other medicinal
agents.
’ ASSOCIATED CONTENT
’ CONCLUSIONS
S
Supporting Information. Experimental procedures, com-
b
The hydrozirconation-initiated multicomponent assembly of
nitriles, acid chlorides, and alcohols into acyl aminals serves as a
versatile reaction that allows for the synthesis of pederin,
psymberin, and several analogs. The linear sequences for for
pederin (10 steps) and psymberin (14 steps) are the shortest
reported synthetic routes to these well-studied natural products.
The brevity of these sequences can in part be attributed to the
facility by which the nitrile intermediates can be prepared.
Employing the multicomponent reaction as the penultimate step
in diverted total synthesis59 sequences provides unprecedented
access to analogs where the effects of variations in the acyl-,
nitrile-, and alcohol-derived fragments on biological activity
could be studied. These analogs include the pederin/psymberin
chimeric structures that we have named the pedestatins.
pound characterization data, and spectra for all new compounds,
experimental procedures for cytotoxicity studies, and protocols for
molecular modeling. This material is available free of charge via the
Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
florean@pitt.edu, bday@pitt.edu, horne@pitt.edu.
Present Addresses
§Johnson & Johnson, Pharmaceutical Research and Develop-
ment, 3210 Merryfield Row, San Diego, CA 92121
Pfizer Global Research and Development, Groton, CT 06340
Biological assays showed several important structureꢀactivity
trends including: (1) the pederic acid subunit confers greater
activity on compounds than the psymberic acid subunit, (2) the
identity of the alkoxy group in the N-acyl aminal linkage can be
altered significantly, but activity is diminished when the alkoxy
group is replaced by a hydrogen atom, (3) enhancing the
hydrophobicity at the C13 site of pederin (C11 site of psy-
mberin) increases activity, (4) the dihydroisocoumarin group of
psymberin yields stronger activity than the structurally simpler
righthand fragment of pederin, and (5) combining the most potent
fragments of pederin and psymberin produces exceedingly powerful
antineoplastic agents. Initial studies show that the chimeric
structure pedestatin is among the most potent cytotoxins that
Author Contributions
‡These authors contributed equally to this work.
’ ACKNOWLEDGMENT
This work was supported by generous funding from the
National Science Foundation (CHE-0848299, P.E.F.) and the
University of Pittsburgh (W.S.H.) We thank Professor Kazunori
Koide and Mr. Sami Osman (University of Pittsburgh) for
assistance in the cytotoxicity assays and the National Institutes
of Health (S10RR023404) for the 700 MHz NMR at the
University of Pittsburgh.
16677
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679

Journal of the American Chemical Society
ARTICLE
’ REFERENCES
(17) (a) Jiang, X.; Garcia-Fortanet, J.; De Brabander, J. K. J. Am.
Chem. Soc. 2005, 127, 11254. (b) Shangguan, N.; Kiren, S.; Williams, L. J.
Org. Lett. 2007, 9, 1093. (c) Huang, X.; Shao, N.; Palani, A.; Aslanian, R.;
Buevich, A. Org. Lett. 2007, 9, 2597. (d) Lachance, H.; Marion, O.; Hall,
D. G. Tetrahedron Lett. 2008, 49, 6061. (e) Smith, A. B., III; Jurica, J. A.;
Walsh, S. P. Org. Lett. 2008, 10, 5625. (f) Crimmins, M. T.; Stevens,
J. M.; Schaaf, G. M. Org. Lett. 2009, 11, 3990.
(18) (a) Aubele, D. L.; Floreancig, P. E. Org. Lett. 2002, 4, 3443. (b)
Rech, J. C.; Floreancig, P. E. Org. Lett. 2003, 5, 1495. (c) Aubele, D. L.;
Rech, J. C.; Floreancig, P. E. Adv. Synth. Catal. 2004, 346, 359.
(19) (a) Erker, G.; Fr€omberg, W.; Atwood, J. L.; Hunter, W. E.
Angew. Chem., Int. Ed. 1984, 23, 68. (b) Fr€omberg, W.; Erker, G.
J. Organomet. Chem. 1985, 280, 343. (c) Anbhaikar, N. B.; Herold, M.;
Liotta, D. C. Heterocycles 2004, 62, 217.
(1) (a) Pavan, M.; Bo, G. Physiol. Comp. Oecol. 1953, 3, 307. (b)
Cardani, C.; Ghiringhelli, D.; Mondelli, R.; Quilico, A. Tetrahedron Lett.
1965, 6, 2537. (c) Furusaki, A.; Watanabꢀe, T.; Matsumoto, T.; Yanagiya,
M. Tetrahedron Lett. 1968, 9, 6301.
(2) Cichewicz, R. H.; Valeriote, F. A.; Crews, P. Org. Lett. 2004, 6,
1951.
(3) Petit, G. R.; Xu, J.; Chapuis, J.; Petit, R. K.; Tackett, L. P.;
Doubek, D. L.; Hooper, J. N. A.; Schmidt, J. M. J. Med. Chem. 2004,
47, 1149.
(4) (a) Perry, N. B.; Blunt, J. W.; Munro, M. H. G.; Pannell, L. K.
J. Am. Chem. Soc. 1988, 110, 4850. (b) Sakemi, S.; Ichiba, T.; Kohmoto,
S.; Saucy, G.; Higa, T. J. Am. Chem. Soc. 1988, 110, 4851. (c) Perry, N. B.;
Blunt, J. W.; Munro, M. H. G.; Thompson, A. M. J. Org. Chem. 1990,
55, 223. (d) Fusetani, N.; Sugawara, T.; Matsunaga, S. J. Org. Chem.
1992, 57, 3828. (e) Kobayashi, J.; Itagaki, F.; Shigemori, H.; Sasaki, T.
J. Nat. Prod. 1993, 56, 976. (f) Tsukamoto, S.; Matsunaga, S.; Fusetani,
N.; Toh-e, A. Tetrahedron 1999, 55, 13697.
(5) (a) Piel, J.; Butzke, D.; Fusetani, N.; Hui, D.; Platzer, M.; Wen,
G.; Matsunaga, S. J. Nat. Prod. 2005, 68, 472. (b) Piel, J.; Hui, D.; Wen,
G.; Butzke, D.; Platzer, M.; Fusetani, N.; Matsunaga, S. Proc. Natl. Acad.
Sci. U.S.A. 2004, 101, 16222. (c) Piel, J.; Wen, G.; Platzer, M.; Hui, D.
Chembiochem 2004, 5, 93. (d) Fisch, K. M.; Gurgui, C.; Heycke, N.; van
der Sar, S. A.; Anderson, S. A.; Webb, V. L.; Taudien, S.; Platzer, M.;
Rubio, B. K.; Robinson, S. J.; Crews, P.; Piel, J. Nat. Chem. Biol. 2009,
5, 494.
(6) Soldati, M.; Fioretti, A.; Ghione, M. Experimentia 1966, 22, 176.
(7) Burres, N. S.; Clement, J. J. Cancer Res. 1989, 49, 2935.
(8) Ogawara, H.; Higashi, K.; Uchino, K.; Perry, N. B. Chem. Pharm.
Bull. 1991, 39, 2152.
(9) Richter, A.; Kocienski, P.; Raubo, P.; Davies, D. E. Anti-Cancer
Drug Des. 1997, 12, 217.
(10) Hood, K. A.; West, L. M.; Northcote, P. T.; Berridge, M. V.;
Miller, J. H. Apoptosis 2001, 6, 207.
(11) Lee, K.-H.;Nishimura, S.;Matsunaga, S.;Fusetani, N.;Horinouchi,
S.; Yoshida, M. Cancer Sci. 2005, 96, 357.
(12) Chinen, T.; Nagumo, Y.; Watanabe, T.; Imaizumi, T.; Shibuya,
M.; Kataoka, T.; Kanoh, N.; Iwabuchi, Y.; Usui, T. Toxicol. Lett. 2010,
199, 341.
(13) Nishimura, S.; Matsunaga, S.; Yoshida, M.; Hirota, H.; Yokoyama,
S.; Fusetani, N. Bioorg. Med. Chem. 2005, 13, 449.
(20) Maraval, A.; Igau, A.; Donnadieu, B.; Majoral, J. P. Eur. J. Org.
Chem. 2003, 385.
(21) (a) Wan, S.; Green, M. E.; Park, J.-H.; Floreancig, P. E. Org. Lett.
2007, 9, 5385. (b) Xiao, Q.; Floreancig, P. E. Org. Lett. 2008, 10, 1139.
(c) DeBenedetto, M. V.; Green, M. E.; Wan, S.; Park, J.-H.; Floreancig,
P. E. Org. Lett. 2009, 11, 835. (d) Lu, C.; Xiao, Q.; Floreancig, P. E. Org.
Lett. 2010, 12, 5112.
(22) Portions of this material were previously reported. Please see: Wu,
F.; Green, M. E.; Floreancig, P. E. Angew. Chem., Int. Ed. 2011, 50, 1131.
(23) Inukai, T.; Yoshizawa, R. J. Org. Chem. 1967, 32, 404.
(24) Kinnaird, J. W. A.; Ng, P. Y.; Kubota, K.; Wang, X.; Leighton,
J. A. J. Am. Chem. Soc. 2002, 124, 7920.
(25) Roush, W. R.; Palkowitz, A. D.; Ando, K. J. Am. Chem. Soc. 1990,
112, 6348.
(26) Kim, I. S.; Ngai, M.-Y.; Krische, M. J. J. Am. Chem. Soc. 2008,
130, 14891.
(27) (a) Paterson, I.; Gibson, K. R.; Oballa, R. M. Tetrahedron Lett.
1996, 37, 8585. (b) Evans, D. A.; Coleman, P. J.; C^otꢀe, B. J. Org. Chem.
1997, 62, 788. (c) Evans, D. A.; C^otꢀe, B.; Coleman, P. J.; Connell, B. T.
J. Am. Chem. Soc. 2003, 125, 10893. (d) Paton, R. S.; Goodman, J. M.
J. Org. Chem. 2008, 73, 1253. (e) Dias, L. C.; Pinheiro, S. M.; de Oliveira,
V. M.; Ferreira, M. A. B.; Tormena, C. F.; Aguilar, A. M.; Zukerman-
Schpector, J.; Tiekink, E. R. T. Tetrahedron 2009, 65, 8714 and
references therein.
(28) Paterson, I.; Goodman, J. M.; Lister, M. A.; Schumann, R. C.;
McClure, C. K.; Norcross, R. D. Tetrahedron 1990, 46, 4663.
(29) Paterson, I.; Perkins, M. V. Tetrahedron Lett. 1992, 33, 801.
(30) (a) Komatsu, N.; Uda, M.; Suzuki, H.; Takahashi, T.; Domae,
T.; Wada, M. Tetrahedron Lett. 1997, 38, 7215. (b) Jung, H. H.; Seiders,
J. R., II; Floreancig, P. E. Angew. Chem., Int. Ed. 2007, 46, 8464.
(31) Evans, P. A.; Cui, J.; Gharpure, S. J.; Hinkle, R. J. J. Am. Chem.
Soc. 2003, 125, 11456.
(14) G€urel, G.; Blaha, G.; Steitz, T. A.; Moore, P. B. Antimicrob.
Agents Chemother. 2009, 53, 5010.
(15) (a) Meinwald, J. Pure Appl. Chem. 1977, 49, 1275. (b) Nakata,
T.; Nagao, S.; Oishi, T. Tetrahedron Lett. 1985, 26, 6465. (c) Matsuda,
F.; Tomiyoshi, N.; Yanagiya, M.; Matsumoto, T. Tetrahedron 1988,
44, 7063. (d) Jarowicki, K.; Kocienski, P.; Marczak, S.; Willson, T.
Tetrahedron Lett. 1990, 31, 3433. (e) Hoffmann, R. W.; Schlapbach, A.
Tetrahedron 1992, 48, 1959. (f) Takemura, T.; Nishii, Y.; Takahashi, S.;
Kobayashi, J.; Nakata, T. Tetrahedron 2002, 58, 6359. (g) Rolle, T.;
Hoffmann, R. W. Helv. Chim. Acta 2004, 87, 1214. (h) Jewett, J. C.;
Rawal, V. H. Angew. Chem., Int. Ed. 2007, 46, 6502.
(32) Shenoy, S. R.; Smith, D. M.; Woerpel, K. A. J. Am. Chem. Soc.
2006, 128, 8671.
(33) A portion of this work has been reported previously. Please see:
Rech, J. C.; Floreancig, P. E. Org. Lett. 2005, 7, 5175.
(34) (a) Langer, P.; Kracke, B. Tetrahedron Lett. 2000, 41, 4545. (b)
Hussain, I.; Yawer, M. A.; Appel, B.; Sher, M.; Mahal, A.; Villinger, A.;
Fischer, C.; Langer, P. Tetrahedron 2008, 64, 8003.
(35) Node, M.; Fujiwara, T.; Ichihashi, S.; Nishide, K. Tetrahedron
Lett. 1998, 39, 6331.
(36) Barker, D.; Brimble, M.; Do, P.; Turner, P. Tetrahedron 2003,
59, 2441.
(37) Brown, H. C.; Bhat, K. S. J. Am. Chem. Soc. 1986, 108, 5919.
(38) (a) Cherest, M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968,
9, 2199. (b) Anh, N. T.; Eisenstein, O. Nouv. J. Chem. 1977, 1, 61. (c)
Lodge, E. P.; Heathcock, C. H. J. Am. Chem. Soc. 1987, 109, 3353.
(39) Evans, D. A.; Duffy, J. L.; Dart, M. J. Tetrahedron Lett. 1994,
35, 8537.
(40) (a) Evans, D. A.; Allison, B. D.; Yang, M. G.; Masse, C. E. J. Am.
Chem. Soc. 2001, 123, 10840. (b) Evans, D. A.; Dart, M. J.; Duffy, J. L.;
Yang, M. G. J. Am. Chem. Soc. 1996, 118, 4322.
(16) (a) Hong, C. Y.; Kishi, Y. J. Org. Chem. 1990, 55, 4242. (b)
Hong, C. Y.; Kishi, Y. J. Am. Chem. Soc. 1991, 113, 9693. (c) Hoffmann,
R. W.; Schlapbach, A. Tetrahedron Lett. 1993, 34, 7903. (d) Nakata, T.;
Matsukura, H.; Jian, D.; Nagashima, H. Tetrahedron Lett. 1994, 35, 8229.
(e) Marron, T. G.; Roush, W. R. Tetrahedron Lett. 1995, 36, 1581. (f)
Roush, W. A.; Pfeifer, L. A. Org. Lett. 2000, 2, 859. (g) Kocienski, P.;
Narquizian, R.; Raubo, P.; Smith, C.; Farrugia, L. J.; Muir, K.; Boyle, F. T.
J. Chem. Soc., Perkin Trans. 1 2000, 2357. (h) Trost, B. M.; Yang, H.;
Probst, G. D. J. Am. Chem. Soc. 2004, 126, 48. (i) Sohn, J.-H.; Waizumi,
N.; Zhong, H. M.; Rawal, V. H. J. Am. Chem. Soc. 2005, 127, 7290. (j)
Kagawa, N.; Ihara, M.; Toyota, M. J. Org. Chem. 2006, 71, 6796. (k)
Green, M. E.; Rech, J. C.; Floreancig, P. E. Angew. Chem., Int. Ed. 2008,
47, 7317. (l) Nishii, Y.; Higa, T.; Takahashi, S.; Nakata, T. Tetrahedron
Lett. 2009, 50, 3597. (m) Jewett, J. C.; Rawal, V. H. Angew. Chem., Int. Ed.
2010, 49, 8682.
(41) Chen, K.-M.; Hardtmann, G. E.; Prasad, K.; Repic, O.; Shapiro,
M. J. Tetrahedron Lett. 1987, 28, 155.
16678
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679

Journal of the American Chemical Society
ARTICLE
(42) (a) Rychnovsky, S. D.; Skalitzky, D. J. Tetrahedron Lett. 1990,
31, 945. (b) Evans, D. A.; Rieger, D. L.; Gage, J. R. Tetrahedron Lett.
1990, 31, 7099.
(43) (a) Roush, W. R.; Marron, T. G.; Pfeifer, L. A. J. Org. Chem.
1997, 62, 474. (b) Trotter, N. S.; Takahashi, S.; Nakata, T. Org. Lett.
1999, 1, 957. (c) Breitfelder, S.; Schuemacher, A. C.; Rolle, T.; Kikuchi,
M.; Hoffmann, R. W. Helv. Chim. Acta 2004, 87, 1202.
(44) A portion of this work has been reported in ref 16k.
(45) Zhu, C.; Shen, X.; Nelson, S. G. J. Am. Chem. Soc. 2004, 126, 5352.
(46) (a) Kiren, S.; Williams, L. J. Org. Lett. 2005, 7, 2905. (b) Green,
M. E.; Rech, J. C.; Floreancig, P. E. Org. Lett. 2005, 7, 4117. (c)
Pietruszka, J.; Simon, R. C. Eur. J. Org. Chem. 2009, 3628.
(47) Ley, S. V.; Dixon, D. J.; Guy, R. T.; Palomero, M. A.; Polara, A.;
Rodriguez, F.; Sheppard, T. D. Org. Biomol. Chem. 2004, 2, 3618.
(48) Jiang, X.; Williams, N.; De Brabander, J. K. Org. Lett. 2007,
9, 227.
(49) For other examples please see refs 16a, 16k and 17f.
(50) The structure of pedestatin was reported in a patent by the
Schering-Plough group though it has yet to appear in the peer-reviewed
literature. See: Huang, X.; Shao, N.; Seidel-Dugan, C.; Palani, A.; Aslanian,
R. G.; Huryk, R. PCT Int. Appl., WO 2009158381 A1 20091230, 2009. We
thank an anonymous referee for bringing this to our attention.
(51) Huang, X.; Shao, N.; Huryk, R.; Palani, A.; Aslanian, R.; Seidel-
Dugan, C. Org. Lett. 2009, 11, 867.
(52) Fukui, H.; Tsuchiya, Y.; Fujita, K.; Nakagawa, T.; Koshino, H.;
Nakata, T. Bioorg. Med. Chem. Lett. 1997, 7, 2081.
(53) Robinson, S. J.; Tenney, K.; Yee, D. F.; Martinez, L.; Media,
J. E.; Valeriote, F. A.; van Soest, R. W. M.; Crews, P. J. Nat. Prod. 2007,
70, 1002.
(54) We thank Dr. Bert Vogelstein from Johns Hopkins University
for supplying this cell line.
(55) (a) Matsunaga, S.; Fusetani, N.; Nakao, Y. Tetrahedron 1992,
48, 8369. (b) Simpson, J. S.; Garson, M. J.; Blunt, J. W.; Munro, M. H. G.;
Hooper, J. N. A. J. Nat. Prod. 2000, 63, 704.
(56) Thomson, A. M.; Blunt, J. W.; Munro, M. H. G.; Perry, N. B.
J. Chem. Soc., Perkin Trans. 1 1995, 1233.
(57) Bergeron, R. J.; Ludin, C.; M€uller, R.; Smith, R. E.; Phanstiel, O.,
IV J. Org. Chem. 1997, 62, 3285.
(58) Krepski, L. R.; Jensen, K. M.; Heilmann, S. M.; Rasmussen, J. K.
Synthesis 1986, 301.
(59) (a) Gaul, C.; Njardarson, J.; Danishefsky, S. J. J. Am. Chem. Soc.
2003, 125, 6042. (b) Szpilman, A. M.; Carreira, E. M. Angew. Chem., Int.
Ed. 2010, 49, 9592.
(60) Pettit, G. R.; Cichacz, Z. A.; Gao, F.; Herald, C. L.; Boyd, M. R.;
Schmidt, J. M.; Hooper, J. N. A. J. Org. Chem. 1993, 58, 1302.
(61) Osman, S.; Albert, B. J.; Wang, Y.; Li, M.; Czaicki, N. L.; Koide,
K. Chem.—Eur. J. 2011, 17, 895.
16679
dx.doi.org/10.1021/ja207331m |J. Am. Chem. Soc. 2011, 133, 16668–16679