Total synthesis of (-)-zampanolide and structure-activity relationship studies on (-)-dactylolide derivatives

Article abstract of DOI:10.1002/chem.201202553

A new total synthesis of the marine macrolide (-)-zampanolide (1) and the structurally and stereochemically related non-natural levorotatory enantiomer of (+)-dactylolide (2), that is, ent-2, has been developed. The synthesis features a high-yielding, selective intramolecular Horner-Wadsworth-Emmons (HWE) reaction to close the 20-membered macrolactone ring of 1 and ent-2. The β-keto phosphonate/aldehyde precursor for the ring-closure reaction was obtained by esterification of a ω-diethylphosphono carboxylic acid fragment and a secondary alcohol fragment incorporating the THP ring that is embedded in the macrocyclic core structure of 1 and ent-2. THP ring formation was accomplished through a segment coupling Prins-type cyclization. Employing the same overall strategy, 13-desmethylene-ent-2 as well as the monocyclic desTHP derivatives of 1 and ent-2 were prepared. Synthetic 1 inhibited human cancer cell growth in vitro with nM IC₅₀ values, while ent-2, which lacks the diene-containing hemiaminal-linked side chain of 1, is 25- to 260-fold less active. 13-Desmethylene-ent-2 as well as the reduced versions of ent-2 and 13-desmethylene-ent-2 all showed similar cellular activity as ent-2 itself. The same activity level was attained by the monocyclic desTHP derivative of 1. Oxidation of the aldehyde functionality of ent-2 gave a carboxylic acid that was converted into the corresponding N-hexyl amide. The latter showed only μM antiproliferative activity, thus being several hundred-fold less potent than 1. It's the side chain that matters: The marine macrolide (-)-zampanolide (1) has been synthesized via (-)-dactylolide (ent-2), the non-natural enantiomer of the marine natural product (+)-dactylolide (2), employing a high-yielding intramolecular Horner-Wadsworth-Emmons reaction to close the macrolactone ring. While the hemiaminal-linked side chain in 1 is crucial for nanomolar antiproliferative activity, the methylene group and the aldehyde functionality in ent-2 are dispensable. A monocyclic destetrahydropyran derivative of 1 shows equal activity as ent-2. Copyright

Full text of DOI:10.1002/chem.201202553

DOI: 10.1002/chem.201202553
Total Synthesis of (ꢀ)-Zampanolide and Structure–Activity Relationship
Studies on (ꢀ)-Dactylolide Derivatives
Didier Zurwerra,[a] Florian Glaus,[a] Leo Betschart,[a] Julia Schuster,[a] Jꢀrg Gertsch,[b]
Walter Ganci,[c] and Karl-Heinz Altmann*[a]
Abstract: A new total synthesis of the
marine macrolide (ꢀ)-zampanolide (1)
and the structurally and stereochemi-
cally related non-natural levorotatory
enantiomer of (+)-dactylolide (2), that
is, ent-2, has been developed. The syn-
thesis features a high-yielding, selective
that is embedded in the macrocyclic
core structure of 1 and ent-2. THP ring
formation was accomplished through a
segment coupling Prins-type cycliza-
tion. Employing the same overall strat-
egy, 13-desmethylene-ent-2 as well as
the monocyclic desTHP derivatives of
1 and ent-2 were prepared. Synthetic 1
inhibited human cancer cell growth in
vitro with nm IC50 values, while ent-2,
which lacks the diene-containing hemi-
aminal-linked side chain of 1, is 25- to
260-fold less active. 13-Desmethylene-
ent-2 as well as the reduced versions of
ent-2 and 13-desmethylene-ent-2 all
showed similar cellular activity as ent-2
itself. The same activity level was at-
tained by the monocyclic desTHP de-
rivative of 1. Oxidation of the aldehyde
functionality of ent-2 gave a carboxylic
acid that was converted into the corre-
sponding N-hexyl amide. The latter
showed only mm antiproliferative activi-
ty, thus being several hundred-fold less
potent than 1.
intramolecular
Horner–Wadsworth–
Emmons (HWE) reaction to close the
20-membered macrolactone ring of 1
and ent-2. The b-keto phosphonate/al-
dehyde precursor for the ring-closure
reaction was obtained by esterification
Keywords: cancer
· dactylolide ·
of
a w-diethylphosphono carboxylic
natural product · total synthesis ·
tubulin · zampanolide
acid fragment and a secondary alcohol
fragment incorporating the THP ring
Introduction
have been developed into FDA-approved drugs, either di-
rectly or through appropriate structural modifications,[2] and
many other drug candidates derived from marine natural
products are currently at different stages of clinical develop-
ment. Among other modes of action, several marine natural
products have been shown recently to be microtubule-stabi-
lizing agents (MSA)[3] and thus to inhibit cancer cell prolif-
eration through the same mechanism as the taxane-based
anticancer drugs taxol, docetaxel, and cabazitaxel or the
epothilone derivative ixabepilone. MSA of marine origin in-
clude discodermolide, laulimalide and isolaulimalide, peloru-
sides A and B, dictyostatin, and (ꢀ)-zampanolide (1),[4] as
the latest addition to this group; with the exception of disco-
dermolide all of these compounds are macrolides of differ-
ent ring sizes.
Marine organisms are the source of a vast array of natural
products with diverse structural architectures, many of
which exhibit potent biological activity.[1] As such, marine
natural products have gained increasing importance as lead
structures for drug discovery research or as tool compounds
for chemical biology studies. So far, three such compounds
[a] Dr. D. Zurwerra, M. Sc. Chem. F. Glaus, M. Sc. Chem. L. Betschart,
J. Schuster, Prof. Dr. K.-H. Altmann
Department of Chemistry and Applied Biosciences
Institute of Pharmaceutical Sciences
Swiss Federal Institute of Technology (ETH) Zꢀrich
HCI H405
Wolfgang-Pauli-Str. 10, 8093 Zꢀrich (Switzerland)
Fax : (+41)44-6331369
E-mail: karl-heinz.altmann@pharma.ethz.ch
Homepage: http://www.pharma.ethz.ch/institute_groups/
institute_groups/pharmaceutical_biology/
[b] Prof. Dr. J. Gertsch
University of Bern
Institute of Biochemistry and Molecular Medicine
Bꢀhlstrasse 28, 3012 Bern (Switzerland)
[c] Dr. W. Ganci
University of Zꢀrich, Organic Chemistry Institute
Winterthurerstrasse 190, 8057 Zꢀrich (Switzerland)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202553.
16868
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16868 – 16883

FULL PAPER
While the microtubule-stabilizing properties of zampano-
lide were discovered only very recently,[4] this compound
had been reported to be a potent antiproliferative agent al-
ready several years prior. Thus, (ꢀ)-zampanolide (1) was
originally isolated by Higa and Tanaka from the marine
sponge Fasciospongia rimosa near Okinawa in 1996[5] and
found to inhibit human cancer cell growth in vitro with IC50
values in the low nanomolar range (2–10 nm). More recently,
1 was also isolated from the Togan sponge Cacospongia my-
cofijiensis by Northcote, Miller, and co-workers, who con-
firmed the in vitro antiproliferative activity of the com-
pound and, most importantly, revealed its microtubule-stabi-
lizing and tubulin-polymerizing activity.[4] In the meantime it
has been demonstrated that the binding of 1 to tubulin
occurs at the taxol binding site on b-tubulin and leads to the
formation of a covalent complex (by addition of His229
(and, to a lesser extent, also Asn228) to the enone system in
the eastern part of the macrocycle).[6] Retrospectively, it ap-
pears somewhat surprising that zampanolide was not recog-
nized as a MSA earlier, as material for biochemical testing
would have been available from the total synthesis work by
Smith[7] or Hoye[8] (see below) before re-isolation of the
compound by Northcote and co-workers in 2009; in fact, our
own synthetic work on zampanolide was partly driven by
the hypothesis that it might be an MSA, given its structural
resemblance to other MSA of marine origin.
Structurally, zampanolide is characterized by a highly un-
saturated 20-membered macrolactone core, which includes a
syn-2,6-disubstituted tetrahydropyran (THP) ring with an
exocyclic methylene group and an unusual hemiaminal-
linked side chain, a structural motif that is found only in a
limited number of other secondary metabolites.[9] Interest-
ingly, in 2001, a macrolactone structurally related to zampa-
nolide, that is, (+)-dactylolide (2), was isolated from the
sponge Dactylospongia sp. at Vanuatu Island by Cutignano
and co-workers.[10] In contrast to 1, 2 was reported to be
only a moderately potent inhibitor of human cancer cell
growth with IC50 values in the low mm range. While Cutigna-
no and co-workers did not establish the absolute configura-
tion of 2, it was initially assumed that it would be identical
with the configuration of the macrolactone core of zampa-
nolide. However, this premonition has been disproven by
Smith and co-workers as part of their elegant synthetic work
on zampanolide/dactylolide,[7] which revealed the absolute
configuration of the macrolactone core in 1 to be opposite
to that of natural 2; in fact, the levorotatory ent-2, the con-
figuration of which corresponds with that of the macrolac-
tone core in (ꢀ)-zampanolide (1), has not been isolated
from natural sources so far.[11] The discovery of the opposite
configuration of (+)-dactylolide and the macrolactone core
of (ꢀ)-zampanolide immediately raised the question, if the
difference in biological activity between 1 and 2 was related
to the difference in the absolute stereochemistry of the mac-
rolide ring or to the presence/absence of the hemiaminal-
linked side chain (or perhaps both), a question that had pro-
vided additional impetus on the work reported in this paper.
While this work was in progress, Ding and Jennings reported
synthetic ent-2 to be slightly more active than natural 2, al-
though a direct comparison of GI50 values is available only
for the SK-OV-3 cell line (GI50 of 1.8 mgmLꢀ1 for ent-2 vs
3.2 mgmLꢀ1 for 2).[12b] These findings clearly indicate that
the profound activity difference between 1 and 2 derives
from the presence of the hemiaminal side chain in the
former rather than the different configurations of their mac-
rolactone rings.
A number of stereoselective syntheses of 1[8,11,13]/ent-1[7a,b]
and of 2[7c,14,15]/ent-2[12,16,17] have been reported in the litera-
ture, together with approaches towards 1 that have not (yet)
been carried through to the natural product.[18] This work
has included different strategies for the closure of the mac-
rolide ring, such as intramolecular Horner–Wadsworth–
Emmons (HWE) reaction at C2/C3,[7,14,15] ring-closing meta-
thesis (RCM) at C8/C9[8,12,16] or C16/C17,[17] or ester forma-
tion by metal-mediated epoxy-acid coupling[8] or Kita–Trost
macrolactonization.[11] In contrast, little work has been re-
ported so far on analogue structures and their biological ac-
tivity.[19] Intrigued by the divergent stereochemistry of natu-
ral 1 and 2 and in light of the distinct lack of SAR data for
these structures, we had embarked on the total synthesis of
natural 1 and non-natural ent-2, and of analogue structures
for SAR studies, even before the recent discovery of the tu-
bulin-polymerizing activity of 1.[4] We have recently commu-
nicated some initial results from this work in relation to the
synthesis and biological evaluation of ent-2 and its C13-des-
methylene derivative.[20] In this paper we provide full details
on our total synthesis of (ꢀ)-zampanolide (1) and (ꢀ)-dacty-
lolide (ent-2); in addition, we describe the synthesis of 13-
desmethylene-ent-2 and a number of analogues of 1/ent-2
and the assessment of their antiproliferative activity.
Results and Discussion
Retrosynthetic analysis: When reflecting upon conceivable
strategies for the closure of the 20-membered macrocycle in
zampanolide/dactylolide, we recognized that one attractive,
novel option for this key step would be the formation of the
C8=C9 double bond through an intramolecular HWE reac-
tion. Quite surprisingly, this particular ring-closure had not
been part of any of the previous syntheses of dactylolide/
zampanolide, despite the fact that HWE-based macrocycli-
zations involving the formation of the C=C double bond in
a,b-unsaturated ketone units are well precedented in natural
product synthesis (even if they are not used extensively).[21]
Our retrosynthesis for 1/ent-2 was thus developed around a
ring-opening disconnection between C8 and C9 (Scheme 1);
1 was to be obtained from ent-2 and amide 42 employing an
aza-aldol reaction as had been reported by Hoye and co-
workers.[8] The requisite b-keto phosphonate/aldehyde pre-
cursor for the HWE-based macrocyclization (I-1) would be
obtained by esterification of acid I-2 with an appropriately
protected alcohol I-3, followed by protecting group manipu-
lations and oxidation. I-3 was envisioned to be accessible
from protected (R)-glycidol I-4 through regioselective epox-
Chem. Eur. J. 2012, 18, 16868 – 16883
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16869

K.-H. Altmann et al.
Scheme 2. Retrosynthesis of building block I-2. PG=protecting group.
action with PMB-trichloroacetimidate (Scheme 3).[23] Con-
version of the resulting Z vinyl iodide 4 into the correspond-
ing vinyl lithium species with nBuLi followed by treatment
Scheme 3. a) NaAlH2(OCH2CH2OMe)2, Et2O, 0 8C, then EtOAc, then I2
in THF, ꢀ788C, 88%; b) PMBO
hexane 2:1, RT, 90%; c) nBuLi, epichlorohydrin, BF3·OEt2, toluene,
(OEt)2, nBuLi,
BF3·OEt2, THF, ꢀ788C, 80%; f) TBSCl, ImH, DMAP, DMF, RT, 84%;
g) 1) DDQ, CH2Cl2/H2O 20:1, 08C; 2) (COCl)2, DMSO, NEt3, CH2Cl2,
ꢀ788C!RT, 88% (2 steps); h) 1) (EtO)2P(O)CH2COOEt, nBuLi, THF,
08C, 2) NaOH, EtOH, 08C, 94%.
ACHTUGNTERN(NUNG C=NH)CCl3, PPTS (cat.), CH2Cl2/cyclo-
Scheme 1. Retrosynthesis of (ꢀ)-zampanolide (1) and (ꢀ)-dactylolide
(ent-2). Only key disconnections are highlighted explicitly. PG=protect-
ing group; protecting groups may vary independently.
ꢀ858C, 50–70%; d) KOH, EtOH, 08C, 89%; e) HP(O)
ACHTUNGTRENNUNG
ide opening with lithiated vinyl iodide I-5. The latter could
be obtained by Prins-type reaction of alkyne I-6 to deliver a
4-iodo tetrahydropyran derivative; the iodo substituent
would then be elaborated into the desired methylene group
through displacement with an oxygen nucleophile, oxidation
to the ketone and methylenation. Finally, I-6 would be de-
rived from homoallylic alcohol I-7, which, in turn, was plan-
ned to be accessed from d-aspartic acid. The segment cou-
pling approach (to produce I-6) that we envisioned for the
construction of the THP-subring had not been employed in
any of the previous syntheses of 1/ent-1 or 2/ent-2.
As illustrated in Scheme 2, acid I-2 was envisaged to be
elaborated from aldehyde I-8 through HWE elongation. Al-
dehyde I-8 was to be obtained from vinyl iodide I-9 via two
epoxide opening reactions; this was to include the reaction
of I-9 with epichlorohydrin, conversion of the resulting
chlorohydrin to a new oxirane, and opening of the latter
with lithiated diethylphosphite to produce the desired b-hy-
droxy phosphonate.
with racemic epichlorohydrin and BF3·OEt2 then afforded
chlorohydrin 5 in 50–70% yield. Surprisingly, however, 5
was obtained only in toluene as the solvent;[24] no conver-
sion was observed in THF, although epoxide opening reac-
tions with lithiated vinyl species have been reported to pro-
ceed smoothly in THF or Et2O solution.[25] In addition, al-
though TLC analysis generally indicated clean conversion of
4 into 5 (in toluene), the reaction never went to completion
and yields for this transformation could not be improved
beyond 70%, neither by varying the temperature (ꢀ958C
up to RT) nor by the use of apolar solvent mixtures (such as
toluene/cyclohexane, toluene/Et2O, toluene/hexane, cyclo-
hexane/hexane, CH2Cl2, Et2O and THF). Excess epichloro-
hydrin led to higher conversion of 4, while the use of signifi-
cantly larger than stoichiometric amounts of BF3·OEt2 did
not produce any improvement in yield. Thus, the reaction
was best carried out with an excess of epichlorohydrin
(3 equiv) in the presence of 1.3 equiv of BF3·OEt2, which
gave 5 in at least 50% yield even on a 55 mmol scale. Treat-
ment of 5 with KOH/EtOH gave the epoxide 6, which could
be opened regioselectively with lithiated diethylphosphite in
THF,[26] to afford b-hydroxy phosphonate 7 in 71% overall
Synthesis of acid 10: The synthesis of building block 10 (cor-
responding to retron I-2 with PG=TBS; Schemes 1 and 2)
departed from 2-butynol (3), which was submitted to
Coreyꢂs reductive alumination/iodination[22] procedure, fol-
lowed by PMB-protection of the free hydroxyl group by re-
16870
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16868 – 16883

Total Synthesis of (ꢀ)-Zampanolide
FULL PAPER
yield (from 5). Attempts to prepare 7 directly from 4
through reaction with epoxide 11 did not yield any of the
desired product, either with THF or toluene as the solvent
(Scheme 4).
Scheme 4. a) nBuLi, THF or toluene, BF3·OEt2, ꢀ788C.
TBS protection of 7 with TBSCl/imidazole in DMF re-
quired catalysis by DMAP (1 equiv), in order to achieve
complete conversion. Subsequent PMB removal under oxi-
dative conditions (DDQ) afforded a mixture of the unsatu-
rated aldehyde 9 and the corresponding primary alcohol.
This mixture was submitted to Swern oxidation, which pro-
vided the desired 9 in 74% overall yield from 7. HWE reac-
tion of 9 with triethyl phosphonoacetate then gave the
ꢀ
entire C1 C8 carbon fragment of zampanolide/dactylolide
as a carboxylic ester, which could be hydrolyzed to the de-
sired acid 10 with NaOH/EtOH in excellent yield (94% for
the two-step sequence from 9). The whole sequence leading
from 3 to 10 was amenable to scale-up and provided 10 in
multigram quantities.
Scheme 5. a) KBr, H2SO4, NaNO2, H2O, 0 8C, 90%; b) 1) BH3·THF or
BH3·DMS, THF, 08C!RT, 96%, 2) NaH, THF, then TBDPSCl, THF,
ꢀ108C, 90%; c) CH2=CHMgBr, CuI (cat.), THF, ꢀ55!ꢀ308C, 98%;
d) 2-butynoic acid, DCC, DMAP, CH2Cl2, 08C!RT, 85%; e) DIBAL-H,
then Ac2O, pyridine, DMAP, CH2Cl2, ꢀ788C, 92%; f) TMSI, 2,6-dime-
thylpyridine (0.2 equiv), CH2Cl2, ꢀ198C, 85%; g) CsOAc, [18]c-6, tol-
uene, 608C, 4 d, 72 %; h) 1) K2CO3, MeOH/H2O 20:1, RT, 2) DMP,
CH2Cl2, RT, 85%; i) CH3Ph3PBr, nBuLi, THF, 08C!508C, 94%;
j) 1) Bu3SnH, nBuLi, CuCN, THF, MeOH, ꢀ788C, 2) NIS, THF, ꢀ178C,
97%; k) tBuLi, 19, BF3·OEt2, toluene, ꢀ85!ꢀ788C, 61%.
Synthesis of alcohol 20: After having established a scalable
route to acid 10 we next addressed the synthesis of alcohol
20 (corresponding to retron I-3 with PG=PMB, TBDPS;
Scheme 1), which would be joined with 10 to produce the
fully elaborated linear precursor for the projected HWE-
based macrocyclization (after conversion of the terminal
TBDPS-ether into the requisite aldehyde functionality). As
discussed above, a key strategic element in the synthesis of
20 was to be the construction of the 2,6-syn-substituted THP
subring through a Prins-type cyclization reaction with an
acyl acetal of a chiral homoallylic alcohol as an oxonium ion
precursor. The implementation of this strategy is summar-
ized in Scheme 5; in the initial series of steps this involved
the elaboration of d-(ꢀ)-aspartic acid into epoxide 13
through conversion of the former into a-bromo acid 12, fol-
lowed by borane reduction and consecutive in situ treatment
of the ensuing diol with NaH and TBDPSCl, thus directly
producing the TBDPS-protected epoxy alcohol 13.[27] A re-
gioselective Cu-mediated epoxide opening with CH2=
CHMgBr[28] then afforded the desired homoallylic alcohol
14 in excellent yield (98%) and in multigram quantities. In
light of its reliance on readily available and cheap starting
materials, the synthesis of 14 from d-aspartic acid is an at-
tractive alternative to approaches based on asymmetric ally-
lation chemistry. Alcohol 14 was esterified with 2-butynoic
acid[29] and the ester was submitted to reductive acylation, to
provide the acid-sensitive acetylated acetal 15. The Prins-
type cyclization of 15 was effectively promoted by TMSI as
the Lewis acid (2.5 equiv),[30] to afford 16 in 85% yield, with
the substituents at positions 2 and 6 of the THP ring in a
syn orientation and the iodo substituent at position 4 occu-
pying an axial position (i.e., being anti to the 2- and 6-sub-
stituents). Thus, the configuration of the two stereocenters
formed in the course of the cyclization reaction was fully
controlled by the configuration of the chiral center originat-
ing from homoallylic alcohol 14.
The particular stereochemical outcome of the cyclization
reaction with 15 may be rationalized by a model originally
proposed by Rychnovsky and co-workers for TMSBr-in-
duced Prins-type cyclizations of a-acetoxy ethers that as-
sumes a least motion pathway for the attacking nucleo-
phile.[30] According to this model, 15 would be initially con-
verted into iodo ether 21 which would then undergo solvoly-
sis to a contact ion pair (Scheme 6); subsequent cyclization
to a chair intermediate (still as a contact ion pair) followed
by proximal attack of iodide ion at C4 would then result in
the axial product 16.
Cyclization of 15 could also be affected with TMSBr,
which gave axial bromide 22 in 69% yield (Scheme 7). Com-
pared to TMSI, a significantly larger excess of TMSBr had
to be employed (ca. 24-fold) and longer reaction times were
required to achieve full consumption of starting material. In
contrast to TMSI or TMSBr, the use of either TFA or SnBr4
to induce the cyclization of 15 gave only mixtures of 2,6-syn
and -anti isomers (Scheme 7).[31]
Chem. Eur. J. 2012, 18, 16868 – 16883
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16871

K.-H. Altmann et al.
promise between a practical reaction rate and the suppres-
sion of elimination side products, thus furnishing the desired
acetate in yields of about 70% after reaction times of 3–
4 days. At higher temperature (initial experiments were car-
ried out at 908C), the displacement reaction was accompa-
nied by substantial elimination to form both possible cyclo-
hexene isomers in about a 1:1 ratio (see Scheme 10 for a
later discussion). Among alternative sources of oxygen nu-
[35]
[36]
cleophiles investigated, AgOCOCF3
only elimination products, while no conversion was observed
with PhI
(OCOCF3)2.[37] Aqueous CuSO4 in DMSO[38] afford-
or AgClO4
gave
AHCTUNGTRENNUNG
Scheme 6. Model to explain the axial selectivity in the segment-coupling
Prins-type cyclization of acyl acetal 15.[30]
ed a multitude of products that were not further character-
ized. The acetate was then transformed into the desired
olefin 17 in 80% overall yield by base-mediated hydrolysis
of the ester group, oxidation of the ensuing free hydroxyl
group with DMP and finally Wittig methylenation
(Scheme 5). The subsequent conversion of 17 into vinyl
iodide 18 was accomplished by stannylcupration/iodination
[39]
with Bu3Sn(Bu)CuCNLi2
and iodine (or NIS) in THF/
CH2Cl2; these conditions provided 18 in good yields as a
single isomer. In contrast, attempted hydrozirconation with
Schwartz reagent[40] followed by treatment with iodine af-
forded only unchanged starting material. Similar to the reac-
tion of epichlorohydrin with lithiated vinyl iodide 4, the
elaboration of 18 into alcohol 20 via lithiation and subse-
quent reaction with PMB-protected (R)-glycidol (19) proved
to be a significant challenge, in spite of literature prece-
dence for a related transformation.[41] Thus, initial experi-
ments with the vinyl lithium species derived from iodide 18
either in THF, Et2O or mixtures thereof did not deliver any
of the desired product. In marked contrast, the extended al-
cohol 20 was obtained from vinyl iodide 18 in 61% yield
when the epoxide opening was performed in toluene as the
reaction solvent instead of THF or Et2O (with BF3·OEt2 as
the Lewis acid); while the efficiency of this transformation
is relatively moderate, the approach proved to be amenable
to reasonable scale-up and has allowed the preparation of
multigram quantities of the desired secondary alcohol 20.
Scheme 7. a) TMSBr (24 equiv), 2,6-dimethylpyridine (0.2 equiv), CH2Cl2,
08C!RT, 4 h, 69%; b) TFA, CH2Cl2, 40%, about 1:2 mixture of 23/23a;
c) SnBr4, CH2Cl2, ꢀ788C, 69%, 1:1.7 mixture of 24/24a.
It may be speculated that the dependence of the stereo-
chemical outcome of the cyclization reaction on the exact
reaction conditions reflects the small A-value of an alkyne
group (ca. 0.4 kcalmolꢀ1),[32] which leads only to a low pref-
erence for a pseudo-equatorial orientation in the transition
state depicted in Scheme 6. However, care should be exer-
cised when applying the A-value concept outside of a cyclo-
hexane structural framework (for an excellent review on the
concept of A-values see ref. [33]). All attempts to affect in-
termolecular Prins reactions with homoallylic alcohols 14 or
25 and diethoxy acetal 26 or aldehyde 27, respectively, met
with failure and did not provide any of the desired cycliza-
tion products.
Alternative synthesis of alcohol 20: In light of the difficul-
ties encountered in the reaction of epoxide 19 with metalat-
ed derivatives of vinyl iodide 18 and before discovering the
pronounced solvent dependence of the reaction (or the one
between lithiated vinyl iodide 4 and epichlorohydrin) we
had started to investigate an alternative approach to alcohol
20 that would not rely on epoxide opening by any metal-
vinyl intermediates. While these efforts finally did not come
to bear on the total synthesis of 1, as we were able eventual-
ly to overcome the initial problems in the synthesis of 20
from 18 and 19, they still led to a viable approach to 20,
whose most important features are outlined in this para-
graph (Scheme 8). Departing from l-malic acid, ester 29 was
prepared in three steps and 51% overall yield according to
literature procedures.[42,43] Reduction of 29 with DIBAL-H
followed by conversion of the ensuing aldehyde into the cor-
responding dibromoolefin according to the Corey–Fuchs
protocol,[44] treatment of the latter with nBuLi and finally
Elaboration of the iodo substituent in 16 into the required
C13 exo-methylene group began with the conversion of 16
into the corresponding acetate by reaction with CsOAc in
the presence of [18]crown-6.[34] This reaction was best con-
ducted at 55–608C; these conditions offered the best com-
16872
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16868 – 16883

Total Synthesis of (ꢀ)-Zampanolide
FULL PAPER
13C NMR analysis 33 was ob-
tained as a single isomer in
yields between 80 and 97%
after reaction times of only
15 min at ꢀ788C. The use of
the Duthaler–Hafner reagent
was clearly superior to other asymmetric allylation methods
investigated, with Brown allylation of 32 ((ꢀ)-DIPCl/al-
lylMgBr)[48] affording 33 with selectivities of up to 10:1, but
in highly variable yields (0–70%), while attempted Keck al-
lylation ((S)-BINOL/TiACHTNUTGRNEUNG
(OiPr)4)[49] never led to complete
conversion of starting material. The Duthaler–Hafner allyla-
tion protocol was amenable to scale-up (gram scale) and
NH4F-work-up allowed recovery of the (R)-taddol ligand.
Secondary alcohol 33 was then esterified with acid 38 (ob-
tained in three steps from propane-1,3-diol by mono-
TBDPS-protection followed by two-step oxidation to O-
TBDPS-3-hydroxy propanal (27) and finally 38), which was
followed by reductive acylation with Ac2O and DIBAL-H
to give acyl acetal 34 in good yield (85% from 33). Rather
surprisingly, the Prins-type cyclization of 34 worked only
with SnBr4 as the Lewis acid; unfortunately, under these
conditions THP ring formation was accompanied by cleav-
age of the acetonide moiety, which had to be re-installed
after the cyclization step, thus providing the desired cycliza-
tion product in 62% total yield. It is noteworthy that the
THP ring in 35 exhibits an all-syn configuration of the three
substituents at positions 2, 4, and 6 (based on NOE meas-
urements), that is the bromo substituent occupies an equato-
rial position. The stereochemical outcome of the SnBr4-
mediated cyclization of 34, thus, is distinctly different from
the one observed for the TMSI-induced cyclization of 15
(Schemes 5 and 6); this is in accordance with the results of
previous studies on the stereochemical course of segment
coupling Prins-type cyclizations by Rychnovsky and co-
workers.[30]
Scheme 8. a) DIBAL-H, CH2Cl2, ꢀ708C, 82%; b) CBr4, PPh3, 2,6-dime-
thylpyridine, THF, 08C, 86%; c) nBuLi, THF, ꢀ788C then (CHO)n, 64%;
d) NaAlH2(OCH2CH2OMe)2, THF, 08C!RT, then EtOAc, I2, THF,
ꢀ788C, 75%; e) 1) TMSCl, NEt3, 2) Me2Zn, [Pd
ACHTUNGTRNE(NUGN dppf)Cl2], THF, 808C,
3) K2CO3, MeOH, RT, 81% (3 steps); f) (COCl)2, DMSO, NEt3, CH2Cl2,
ꢀ788C, 84%; g) 37, Et2O, ꢀ788C, 15 min, then NH4F, ꢀ788C!RT, 80–
97%; h) 38, EDCI, DMAP, CH2Cl2, 08C, 94%; i) DIBAL-H, then Ac2O,
pyridine, DMAP, CH2Cl2, ꢀ788C, 91%; j) 1) SnBr4, CH2Cl2, ꢀ788C,
2) Me2CACHTUNGTRENNUNG(OMe)2, pTsOH·H2O, RT, 62% (2 steps); k) CsOAc, [18]c-6, tol-
uene, 1308C, 20 h, 88%; l) K2CO3, MeOH/H2O 10:1, RT; m) DMP,
CH2Cl2, RT, 94% (2 steps); n) MePh3PBr, nBuLi, THF, 0!458C, 92%;
o) CuCl2·2H2O, MeOH, 608C, 84%; p) Bu2SnO, toluene, Dean–Stark,
1408C, 1.5 d, then PMBCl, TBAI, 1208C, 1.5 h, 57%.
quenching of the resulting acetylide anion with paraformal-
dehyde gave propargylic alcohol 30 in 45% overall yield
(from 29).
Reductive iodination of 30 and subsequent Negishi cross-
coupling with Me2Zn and [PdACHTUNTRGENNUG(dppf)Cl2] gave the trisubsti-
No THP ring formation was observed with TMSI, instead
the major products formed in the reaction were homoallylic
alcohol 14 (stereochemistry unknown) and aldehyde 32
(Scheme 9), based on NMR and TLC analysis of crude reac-
tion products after extractive work-up and comparison with
authentic reference samples. The formation of 14 and 32
may be rationalized by an oxonia-Cope rearrangement[50] of
the oxonium ion initially formed from 34 as illustrated in
Scheme 9. Likewise, attempts to induce cyclization with
TMSBr, TMSOTf, BF3·OEt2/AcOH, TFA, or TFA/NaO-
COCF3 did not produce any of the desired cyclization prod-
uct.
The elaboration of bromide 35 into olefin 36 was based
on the same sequence of transformations that had been fol-
lowed for the conversion of 16 into 17 (see Scheme 5); thus,
displacement of the bromo substituent in 35 with CsOAc
followed by acetate hydrolysis and oxidation of the resulting
secondary alcohol gave a ketone that was converted into 36
by Wittig olefination (Scheme 8). Olefin 36 was obtained in
76% overall yield for the four-step sequence from 35. It is
worth pointing out that acetate formation from bromide 35
tuted olefin 31 in high yield as a single isomer. In order to
avoid reduction of the vinyl iodide moiety in the cross-cou-
pling step, temporary protection of the allylic alcohol
moiety as a TMS ether was required.[45] In contrast to aceto-
nide 30, reductive iodination of the corresponding bis-TBS
ether gave the desired vinyl iodide only in 34% yield (vs.
75% for 30); similar results were obtained when the pri-
mary and secondary hydroxyl groups were protected as
PMB and TBS ethers, respectively. Cross-coupling was also
possible with Me2CuLi[46] (64–77% yield), but this was ac-
companied by partial reduction of the vinyl iodide moiety.
Swern oxidation of allylic alcohol 31 then provided unsatu-
rated aldehyde 32, which was transformed into homoallylic
alcohol 33 by asymmetric allyltitanation with the tartrate-
derived Duthaler–Hafner reagent 37,[47] thus setting the ster-
1
eocenter at C15 (zampanolide numbering). Based on H and
Chem. Eur. J. 2012, 18, 16868 – 16883
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16873

K.-H. Altmann et al.
Assembly of building blocks and completion of the total
synthesis: The esterification of alcohol 20 with acid 10, as
the first step in the elaboration of the macrocyclic core
structure of (ꢀ)-dactylolide (ent-2) and (ꢀ)-zampanolide
(1), was best accomplished under Yamaguchi conditions,[53]
which provided the desired ester in 85% yield (Scheme 11).
In contrast, the use of DCC or EDCI as condensing agents
gave only trace amounts of product. Esterification was fol-
lowed by simultaneous cleavage of both silyl ethers with
HF·py and oxidation of the resulting diol to the b-keto phos-
phonate/aldehyde 39 in 63% overall yield.
Scheme 9. Formation of 32 and 14 from 34 via oxonia-Cope rearrange-
ment. LA=Lewis acid.
was less problematic (no elimination side products) and thus
higher yielding than from iodide 16 (Scheme 5). As illustrat-
ed in Scheme 10, this is a direct consequence of the equato-
Scheme 11. a) 2,4,6-Trichlorobenzoyl chloride, NEt3, DMAP, 10, toluene,
RT, 85%; b) HF·py, THF, 08C!RT, 85%; c) DMP, CH2Cl2, RT, 74%;
d) Ba(OH)2·0.8H2O, THF/H2O 40:1, 08C!RT, 81%; e) DDQ, CH2Cl2/
H2O 5:1, RT, 82%; f) DMP, CH2Cl2, RT, 78%; g) 42, DIBAL-H then ent-
2, THF, RT, for (ꢀ)-1: 18%; (ꢀ)-epi-1: 12% (see text).
Scheme 10. Unfavorable configuration of bromide 35 for dehydrohaloge-
nation and dehydrohalogenation of iodide 16.
rial orientation of the bromo substituent in 35, which effec-
tively precludes elimination of HBr by an E2 mechanism,
due to the gauche arrangement of the bromine atom and
the axial hydrogens on carbon atoms 3 and 5 of the THP
ring. In contrast, the axial orientation of the iodo substituent
in 16 leads to an antiperiplanar arrangement with the axial
hydrogens on C3 and C5, which allows for facile elimination
of HI (which is what is observed experimentally; see
above).
Cleavage of the acetonide moiety in 36 with
CuCl2·2H2O[51] followed by regioselective PMB protection
of the primary hydroxyl group in the resulting diol via a
cyclic Sn-acetal[52] then concluded the alternative synthesis
of secondary alcohol 20. In comparison to the d-aspartic
acid-based route depicted in Scheme 5, the malic acid-based
approach to 20 involved more steps (22 steps for the longest
linear sequence vs 14 from d-aspartic acid) and provided
the target alcohol in lower overall yield (2.1 vs 17%). Mate-
rial for the progression of the total synthesis was thus gener-
ally produced via the d-aspartic acid route. Samples of 20
obtained from both approaches displayed fully identical
spectral and chiro-optical properties.
With 39 in hand, the stage was set for the exploration of
the intramolecular HWE reaction as a means to affect mac-
rocyclization to the dactylolide and zampanolide core struc-
ture (Scheme 1). Initial experiments to this end involved the
use of NaHMDS as a base in THF; ring-closure to the de-
sired macrocycle 40 (as single isomer) was indeed observed
under these conditions, but extended reaction times were re-
quired (up to 4 d for complete consumption of starting ma-
terial) and product yields were highly variable. Gratifyingly,
these problems could be eliminated by the use of
Ba(OH)2,[54] which led to significantly reduced reaction
times of 0.5–1 h and afforded macrocycle 40 in very good
yields (ca. 80%) in a reproducible fashion. In the largest
single preparation performed so far, 430 mg of 39 were suc-
cessfully converted into 40 in 78% yield. Oxidative PMB re-
moval with DDQ and oxidation of the resulting free alcohol
41 with DMP[55] then provided (ꢀ)-dactylolide (ent-2) in
64% overall yield.
The elaboration of ent-2 into (ꢀ)-zampanolide (1) was ac-
complished by aza-aldol reaction with (Z,E)-sorbamide (42)
(obtained in two steps from crotonaldehyde; see Supple-
16874
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16868 – 16883

Total Synthesis of (ꢀ)-Zampanolide
FULL PAPER
mentary Information) as had been reported by Hoye and
co-workers[8] (and also based on the more general work on
aza-aldol reactions by Maier and co-workers).[56] Thus, treat-
ment of 42 with DIBAL-H in CH2Cl2 followed by reaction
with ent-2 furnished 1 as a 1.1:1 mixture with its C20 epimer
(epi-1) in 46% yield after flash chromatography on deacti-
vated silica gel.[57] The isomers could only be separated by
normal phase HPLC; both 1 and epi-1 were subsequently
submitted to final purification by RP-HPLC, which gave
both compounds in analytically pure form in 18 and 12%
yield, respectively (based on ent-2). While these final yields
are clearly unsatisfactory, they have to be ascribed largely to
the lack of selectivity in the aza-aldol step and the resulting
need for isomer separation by HPLC. After completion of
our own work Ghosh and co-workers have shown that the
reaction of ent-2 and 42 in the presence of a matched chiral
phosphoric acid catalyst ((S)-TRIP) proceeds with about 3:1
selectivity, thus providing 1 in 51% yield after isomer sepa-
ration by HPLC.[13] However, no fully asymmetric synthesis
of 1 has been reported to date. Overall, our synthesis
proved to be highly reliable and notwithstanding the modest
overall yield in the aza-aldol step, it has allowed us to pro-
duce sufficient amounts of material for extensive biological
and biochemical profiling of 1;[6] in addition, this material
has been used to prepare a tubulin-bound complex for X-
ray crystallographic studies.[58]
Scheme 12. a) DMP, CH2Cl2, RT, 78%; b) NaClO2, NaH2PO4·H2O,
tBuOH/H2O, 2-methyl-2-butene, RT, 97%; c) n-hexylamine, HATU,
DIPEA, DMF, RT, 13%; d) Me3OBF4, Proton Sponge, CH2Cl2, RT,
83%.
Synthesis of analogues: As indicated in the introductory sec-
tion, the SAR of dactylolide/zampanolide to this date has
remained largely unexplored. Building on the chemistry that
we had developed for the synthesis of natural 1 and non-nat-
ural ent-2 we have thus started to explore the importance of
individual structural features of 1/ent-2 for microtubule sta-
bilization and antiproliferative activity.
In a first step these SAR inquiries involved the investiga-
tion of side chain-modified zampanolide analogue 45, of al-
cohol 41, which was an intermediate in the synthesis of 1
and ent-2, and of methyl ether 43 (Scheme 12); the latter
was obtained from 41 by reaction with Meerwein salt in
83% yield (Scheme 12). Amide-based zampanolide ana-
logue 45 was obtained from ent-2 by Pinnick–Kraus oxida-
tion[59] (to give acid 44 in almost quantitative yield) followed
by HATU-mediated coupling with n-hexylamine. For rea-
sons unknown, the efficiency of the coupling reaction was
low and provided 45 only in 13% yield (from 44); however,
no attempts were made to optimize this transformation.
As part of our efforts to identify highly active, but struc-
turally simplified analogues of zampanolide/dactylolide with
improved synthetic accessibility, we have also assessed the
significance of the exo-methylene group on the THP ring for
biological activity. The corresponding analogue of ent-2, that
is 13-desmethylene-(ꢀ)-dactylolide (51), was accessed from
iodide 16, which could be converted into the 4-unsubstituted
THP derivative 46 by radical reduction with Bu3SnH/AIBN
in excellent yield (88%; Scheme 13) and without affecting
the alkyne moiety.[39a] Attempts to transform 16 into 46 by
iodide/lithium exchange with tBuLi followed by protonolysis
Scheme 13. a) Bu3SnH, AIBN (cat.), toluene, 608C, 88%; b) 1) Bu3SnH,
nBuLi, CuCN, THF, MeOH, ꢀ78!ꢀ108C, 2) NIS, THF, ꢀ788C, 73%;
c) tBuLi, 19, BF3·OEt2, toluene, ꢀ788C, 31%; d) 2,4,6-trichlorobenzoyl
chloride, NEt3, DMAP, 10, toluene, RT, 74%; e) HF·py, THF, 08C!RT,
80%; f) DMP, CH2Cl2, RT, 72%; g) NaHMDS, THF, ꢀ788C!RT, 2 d,
49%; h) DDQ, CH2Cl2/H2O 5:1, RT, 72%; i) DMP, CH2Cl2, RT, 77%.
resulted in significantly lower yields that did not exceed
35%.
As for alkyne 17, reductive iodination of 46 was achieved
[39a]
by stannylcupration/iodination with Bu3Sn(Bu)CuCNLi2
and NIS to provide the trisubstituted vinyl iodide 47 in 73%
yield. Other methods investigated for the conversion of 46
into 47, such as hydrozirconation/iodination,[40] silylcupra-
tion/iodination[60] or Pd-mediated hydrostannylation[39]
either gave lower yields or were (additionally) plagued by
the formation of the regioisomeric iodination product. Sub-
sequent lithiation of 47 and reaction with epoxide 19 in tol-
uene followed by esterification of the resulting secondary al-
cohol with acid 10, simultaneous cleavage of TBS and
TBDPS ethers, and oxidation of the free hydroxyl groups
with DMP gave aldehyde-phosphonate 48 in 13% overall
yield (based on 47); macrocyclization with NaHMDS as the
Chem. Eur. J. 2012, 18, 16868 – 16883
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16875

K.-H. Altmann et al.
base then gave the desired protected macrolactone 49 in
49% yield. This experiment was carried out before the in-
vestigation of Ba(OH)2-mediated macrocyclizations (see
above); as it provided sufficient material of the correspond-
ing free alcohol 50 and of 13-desmethylene-ent-2 (51) for bi-
ological testing, no attempts were made to improve the
yield of the cyclization, although we assume that the use of
Ba(OH)2 in place of NaHMDS would have provided the
cyclization product in superior yield. In analogy to the syn-
thesis of ent-2, 51 was obtained from 49 by oxidative PMB
removal and DMP oxidation in 55% overall yield.
Encouraged by the cellular data obtained for dactylolide
analogues 50 and 51 (see below), we have also followed a
more radical approach to simplified zampanolide/dactylolide
analogues that involved complete removal of the THP sub-
ring from the macrolactone core. As illustrated in
Scheme 14, the synthesis of the corresponding desTHP de-
rivatives 60 and 59, respectively, was based on the same
overall strategy that had led to the successful synthesis of 1,
ent-2, and THP ring-containing analogues thereof, with sec-
ondary alcohol 55 substituting for intermediate 20 (or its
desmethylene variant) in the esterification with acid 10. The
synthesis of 55 departed from propargylic alcohol (3), which
was converted into vinyl iodide 52 according to literature
procedures.[39] The latter was then elaborated into ether 54
by reaction with allyl bromide in the presence of NaH (to
produce 53) followed by hydroboration and TBDPS protec-
tion of the resulting terminal hydroxyl group in 21% overall
yield. In contrast to allylation with allyl bromide, all at-
tempts at the direct alkylation of 52 with TBDPS-protected
3-bromo-1-propanol proved to be unsuccessful. The latter
could be used to alkylate 3 in 45% yield, but we were
unable to convert the resulting propargylic ether 61 into the
Scheme 14. a) 1) Bu3SnH, nBuLi, CuCN, THF, MeOH, ꢀ78!ꢀ158C,
74%, 2) I2, THF, ꢀ178C, 94%; b) CH2=CHCH2Br, NaH, THF, 08C,
42%; c) BH3·THF, THF, 08C, NaOH/H2O2, 08C, 52%; d) TBDPSCl,
DMAP (cat.), NEt3, CH2Cl2, RT, 94%; e) tBuLi, 19, BF3·OEt2, toluene,
ꢀ788C, 61%; f) 2,4,6-trichlorobenzoyl chloride, NEt3, DMAP, 10, tol-
uene, RT, 81%; g) HF·py, THF, 08C!RT, 86%; h) DMP, CH2Cl2, RT,
73%; i) Ba(OH)2·0.8H2O, THF/H2O 40:1, 08C!RT, 85%; j) DDQ,
CH2Cl2/H2O 5:1, RT, 77%; k) DMP, CH2Cl2, RT, 75%; l) 42, DIBAL-H,
then 59, THF, RT, 72% for the mixture of epimers (ca. 1.6:1).
Ba(OH)2 to furnish macrolactone 57 as a single isomer in
85% yield. PMB removal from 57 under oxidative condi-
tions (DDQ) followed by DMP oxidation then gave (ꢀ)-
dactylolide analogue 59 (77%). Finally, aza-aldol reaction
between 59 and 42 gave desTHP-zampanolide (60) in 28%
yield as a ca. 1.6:1 mixture of isomers at C18 after HPLC
purification (see Scheme 14 for atom numbering).[62] As the
isomers were very difficult to separate, initial biological test-
ing was performed with the isomeric mixture.
required vinyl iodide 54 under the conditions previously es-
tablished for the conversion of 17 into 18 and 46 into 47, re-
spectively. The moderate yield in the allylation step (42%)
is the result of the formation of a side product we assume to
be allene 62 and that was difficult to separate from 53;[61]
while the formation of this side product could be minimized
under optimized conditions (1.7 equiv NaH), it could not be
suppressed completely.
Hydroboration of 53 was best performed with BH3·THF,
while neither (Sia)2BH nor 9-BBN gave any of the desired
alcohol. Reaction of lithiated 54 with epoxide 19 in toluene
in the presence of BF3·OEt2 then afforded secondary alcohol
55 in 61% yield, that is with similar efficiency as for the
conversion of 18 into 20 (Scheme 5). Esterification of 55
with acid 10 under Yamaguchi conditions followed by global
desilylation and subsequent oxidation of the liberated hy-
droxyl groups with DMP provided b-keto phosphonate/alde-
hyde 56; the latter underwent smooth macrocyclization with
Antiproliferative activity: As illustrated by the data sum-
marized in Table 1, synthetic (ꢀ)-zampanolide (1) was
found to inhibit human cancer cell proliferation with nm
IC50 values, which is in perfect agreement with previous lit-
erature data for natural[4,5] as well as synthetic[11] material;
likewise, our data confirm the previous observation by Ue-
nishi et al. that epi-1 is about one order of magnitude less
active than the natural product 1.[11] (ꢀ)-Dactylolide (ent-2)
was found to be even less active than epi-1, with IC50 values
being similar to those reported for natural 2, as had been
observed previously by Ding and Jennings.[12b] These find-
ings identify the hemiaminal-linked side chain of 1 as the
major determinant of the activity difference between 1 and
2, rather than the absolute configuration of the macrocycle.
Analogues 41, 50, and 51 showed comparable activity
with ent-2; in all three cases a slight trend towards higher
potency relative to ent-2 was observed, but the true signifi-
16876
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16868 – 16883

Total Synthesis of (ꢀ)-Zampanolide
FULL PAPER
Table 1. Antiproliferative activity of (ꢀ)-zampanolide (1), epi-1, ent-2,
and derivatives of 1 and ent-2 in four human cancer cell lines (IC50 values
[nm]).[a]
an interesting lead structure in its own right. We are now in-
vestigating whether improvements in the activity of 60 are
possible without any undue (re)increase in structural com-
plexity. The results of these efforts will be reported in due
course.
Compound
A549
(lung)
MCF-7
(breast)
HCT116
(colon)
PC-3
ACHTUNGTRENUN(NG prostate)
G
E
U
1
3.2ꢁ0.4
53ꢁ5.9
6.5ꢁ0.7
7.2ꢁ0.8
88ꢁ5.1
2.9ꢁ0.4
50ꢁ11.7
751ꢁ69
320ꢁ26
104ꢁ4.1
n.d.
1274ꢁ117
9338ꢁ242
829ꢁ27
3051ꢁ178
4021ꢁ102
218ꢁ7
epi-1
ent-2[b]
41
50
51
43
44
45
58
42ꢁ9.3
301ꢁ4.3
127ꢁ2.9
189ꢁ19.3
149ꢁ12.8
1072ꢁ103
9732ꢁ260
973ꢁ90
2378ꢁ70
3921ꢁ216
n.d.
247ꢁ2.6
106ꢁ3.6
114ꢁ10.2
68ꢁ5.6
210ꢁ4.7
155ꢁ2.1
74ꢁ1.5
Conclusion
249ꢁ28
1603ꢁ122
12733ꢁ379
1204ꢁ63
1846ꢁ92
2653ꢁ68
309ꢁ47
We have established a new total synthesis of the marine nat-
ural product (ꢀ)-zampanolide (1) which is based on the con-
struction of its 20-membered macrolide core structure by
means of a high-yielding intramolecular HWE reaction and
the equally effective formation of the embedded THP ring
in a Prins-type cyclization. This strategy gave efficient
access to the non-natural, levorotatory enantiomer of
(+)-dactylolide (2), that is ent-2, which served as the imme-
diate precursor for 1. The same HWE-based macrocycliza-
tion approach was then followed to prepare 13-desmethy-
lene-(ꢀ)-dactylolide (51), based on THP-derivative 16 as a
common advanced intermediate; likewise, this strategy has
also allowed the synthesis of monocyclic desTHP-1 (60) and
desTHP-ent-2 (59). Simple manipulations of the aldehyde
moiety in ent-2 gave access to amide 45, which may be con-
sidered a simplified variant of (ꢀ)-zampanolide (1).
1489ꢁ83
7624ꢁ303
1138ꢁ72
3891ꢁ102
2894ꢁ144
165ꢁ13
59
60[c]
[a] Cells were exposed to compounds for 72 h. n.d. = not determined.
[b] IC50 values of 198–346 nm have been reported for ent-2 in ref. [12b].
[c] Mixture of diastereoisomers at C18.
cance of these differences remains to be determined. Inde-
pendent of this, the data clearly suggest that neither the al-
dehyde functionality as such nor the 13-methylene group
are required for the antiproliferative activity of ent-2. How-
ever, methylation of the C20-hydroxyl group in 41 leads to a
clear loss in antiproliferative potency (up to 14-fold in the
MCF-7 cell line; methyl ether 43). Analogues 41, 50, and 51
(like 1,[4,6] and ent-2[6,20]) all promote tubulin polymeriza-
tion.[20] Details on the interactions of these and other com-
pounds described in this paper with the tubulin/microtubule
system will be reported elsewhere; for (ꢀ)-zampanolide (1)
and ent-2 a detailed study on their binding to dimeric tubu-
lin and microtubules has been published recently.[6]
Compared to alcohol 41, the corresponding carboxylic
acid 44 exhibits significantly reduced cellular activity. This
may be a consequence of poor cell penetration, due to the
negatively charged carboxyl group, although this conclusion
is purely hypothetical at this point. Interestingly, the activity
of amide 45 is about 10-fold enhanced over that of 44; while
this makes the compound several hundred-fold less potent
than 1, the activity of 45 is still encouraging, as it is well con-
ceivable that fine-tuning of the substituent moiety on the
amide nitrogen could lead to improved potency. The synthe-
sis of such analogues is currently ongoing in our laboratory.
Perhaps the most intriguing finding that has emerged
from the cellular profiling experiments is the clearly sub-mm
activity of desTHP-zampanolide (60). Although the com-
pound is 25- to 80-fold less active than the parent zampano-
lide (for the cell lines investigated), the retention of signifi-
cant antiproliferative activity appears truly surprising, in
light of the removal of two (out of four) chiral centers and a
rigidifying structural element. As for 1 and ent-2, desTHP-
(ꢀ)-dactylolide (59) is less active than desTHP-(ꢀ)-zampa-
nolide (60). It is important to emphasize that 60 is a ca.
1.6:1 mixture of diastereoisomers at C18; in light of the re-
sults obtained for 1 and epi-1 it is tempting to speculate that
the IC50 values for 18S-60 could in fact be lower than those
observed for the mixture. Overall, the activity of 60 makes it
Natural 1 was confirmed to be a highly potent cancer cell
growth inhibitor, while ent-2 was 30- to 260-fold less potent
than 1; these findings highlight the importance of the zam-
panolide side chain for nanomolar antiproliferative activity.
Removal of the C13-methylene group and/or the reduction
of the C20-aldehyde moiety in ent-2 did not significantly
affect cellular potency. Intriguingly, analogue 60, which lacks
the entire THP ring and is thus based on a monocyclic scaf-
fold, retains significant antiproliferative activity. It will be
interesting to investigate whether modification of this com-
pound can lead to even more potent analogues with similar-
ly reduced structural complexity (relative to the natural
product lead 1). Amide 45 was 3- to 6-fold less potent than
ent-2 and several hundred-fold less active than (ꢀ)-zampa-
nolide (1); nevertheless, given the micromolar activity of 45,
amide-based analogues of (ꢀ)-zampanolide (1) may still be
an interesting group of analogues for further exploration.
Experimental Section
General: All solvents used for reactions were purchased as anhydrous
grade from Fluka (puriss.; dried over molecular sieves; H2O <0.005%)
and used without further purification. Solvents for extractions, FC and
thin-layer chromatography (TLC) were purchased as commercial grade
and distilled prior to use. All non-aqueous reactions were performed
under an argon atmosphere using flame-dried glassware and standard sy-
ringe/septa techniques. All other commercially available reagents were
used without further purification, unless otherwise noted. In general, re-
actions were magnetically stirred and monitored by TLC on Merck TLC
aluminum sheets (silica gel 60 F254). Spots were visualized with UV light
(l=254 nm) or through staining with Ce
2ACHTNUGTREN(NUGN SO4)3/phosphomolybdic acid/
H2SO4 (CPS) or KMnO4/K2CO3. Purification of products by flash chro-
Chem. Eur. J. 2012, 18, 16868 – 16883
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16877

K.-H. Altmann et al.
matography (FC) was performed using Fluka silica gel 60 for preparative
column chromatography (particle size 40–63 mm).
129.7, 129.7, 129.7, 127.8, 127.8, 118.0, 117.2, 87.2, 86.2, 83.1, 82.9, 76.4,
74.9, 74.4, 74.1, 60.6, 60.3, 39.7, 39.2, 37.4, 37.4, 27.0, 26.9, 21.2, 21.2, 19.3,
19.3, 3.7, 3.7 ppm; IR (thin film): n˜ = 3072, 2956, 2857, 2259, 1740, 1472,
Melting points were obtained in open capillary tubes using a Bꢀchi melt-
ing point apparatus B-540 and are uncorrected. 1H and 13C NMR spectra
were recorded in CDCl3 or [D4]MeOH (unless otherwise noted) on
Bruker AV-400 400 MHz and AV-500 500 MHz instruments at room tem-
perature. Chemical shifts (d) are reported in ppm and are referenced to
the solvent signal as an internal standard (chloroform d=7.26 ppm for
1H, d=77.16 ppm for 13C and [D4]MeOH d=3.34 ppm for 1H, d=
49.00 ppm for 13C). All 13C NMR spectra were measured with complete
proton decoupling. Data for NMR spectra are reported as follows: s=
singlet, d=doublet, t=triplet, q=quartet, quint=quintet, sext=sextet,
m=multiplet, br=broad signal, J=coupling constant in Hz. Infrared
spectra (IR) were recorded on a Jasco FT/IR-6200 instrument as thin
film. Resonance frequencies are given as wavenumbers in cmꢀ1. Optical
rotations were measured on a Jasco P-1020 polarimeter operating at the
sodium D line with a 10 mm or 100 mm path length cell and are reported
as follows: [a]TD, concentration (g per 100 mL), and solvent. Mass spectra
were recorded by the ETH Zꢀrich MS service; HRMS (ESI) spectra
were obtained on a Bruker Daltonics maxis (UHR-TOF) and HRMS
(EI) on a Waters Micromass AutoSpec Ultima instrument.
1370, 1228, 1082, 903, 822, 701 cmꢀ1
; HRMS (ESI): m/z: calcd for
C28H36NaO4Si [M+Na+], 487.2281, found 487.2265.
tert-Butyl-(2-((2S,4S,6S)-4-iodo-6-(prop-1-ynyl)tetrahydro-2H-pyran-2-
yl)ethoxy)diphenylsilane (16): To a solution of 15 (1.79 g, 3.85 mmol,
1.00 equiv) in CH2Cl2 (55 mL) at ꢀ198C (NaCl/ice) was added 2,6-dime-
thylpyridine (0.09 mL, 0.77 mmol, 0.20 equiv) followed by slow addition
of TMSI (1.37 mL, 9.62 mmol, 2.50 equiv). The cooling bath was removed
after 10 min and the yellow solution was allowed to warm to RT. After a
total of 45 min sat aq NaHCO3 (20 mL) was carefully added, the phases
were separated, and the aqueous phase was extracted with CH2Cl2 (3ꢃ
10 mL). The combined organic extracts were dried over MgSO4 and con-
centrated under reduced pressure. Purification of the residue by FC
(EtOAc/Hex 1:30!1:20) afforded 16 (1.78 g, 3.34 mmol, 85%) as a pale-
yellow, viscous oil. Material obtained in this way was generally contami-
nated by 2–3% of aldehyde 27. Spectroscopic data for 16 were acquired
with a pure sample. Rf =0.45 (EtOAc/Hex 1:10, UV, CPS); [a]2D4 =ꢀ3.308
(c = 1.00, CHCl3); 1H NMR (400 MHz, CDCl3): d=7.71–7.67 (m, 4H),
7.44–7.36 (m, 6H), 4.83 (quin, J=3.1; 1H), 4.57 (brdquin, J=10.8; 2.1,
1H), 4.21–4.15 (m, 1H), 3.86 (ddd, J=10.5, 8.2, 4.9; 1H), 3.73 (dt, J=
10.3, 5.4; 1H), 2.16 (dq, J=14.8, 2.3; 1H), 1.99 (ddd, J=14.7, 2.4, 2.1;
1H), 1.93–1.82 (m, 2H), 1.87 (d, J=2.1; 3H), 1.69 (ddt, J=13.7, 8.3, 5.3;
1H), 1.57–1.50 (m, 1H), 1.07 ppm (s, 9H); 13C NMR (100 MHz, CDCl3):
d=135.7, 135.7, 134.1, 133.9, 129.7, 127.8, 127.8, 81.5, 77.9, 71.1, 65.1,
60.1, 41.8, 40.3, 38.3, 29.3, 27.0, 19.4, 3.8 ppm; IR (thin film): n˜ =3070,
2953, 2856, 2360, 2341, 1472, 1427, 1389, 1232, 1107, 1095, 1049, 822, 737,
1-((R)-1-(tert-Butyldiphenylsilyloxy)hex-5-en-3-yloxy)but-2-ynyl acetate
(15): To
a solution of homoallylic alcohol 14 (1.84 g, 5.2 mmol,
1.00 equiv) in dry CH2Cl2 (15 mL) were added sequentially DMAP
(65 mg, 0.52 mmol, 0.10 equiv), 2-butynoic acid (0.49 g, 5.70 mmol,
1.10 equiv), and a solution of DCC (1.30 g, 6.30 mmol, 1.20 equiv) in
CH2Cl2 (15 mL) at 08C. The suspension formed was allowed to warm to
RT and stirring was continued for 16 h. Et2O (100 mL) was then added,
the mixture was filtered, and the filter cake was washed with Et2O
(50 mL). The filtrate was concentrated under reduced pressure to give a
brown-red oil that was again treated with Et2O (100 mL) followed by re-
filtration of the mixture and washing of the precipitate with Et2O
(50 mL). The combined filtrates and washings were concentrated under
reduced pressure and the residue was purified by FC (EtOAc/Hex 1:30!
1:20), to give the desired ester as a colorless oil (1.87 g, 4.44 mmol,
85%). Rf =0.44 (EtOAc/Hex 1:10, UV, CPS); [a]2D4 =ꢀ18.048 (c = 0.93,
CHCl3); 1H NMR (400 MHz, CDCl3): d = 7.67–7.65 (m, 4H), 7.46–7.37
(m, 6H), 5.77 (ddt, J=17.4, 9.7; 7.0, 1H), 5.29–5.23 (m, 1H), 5.13–5.07
(m, 2H), 3.77–3.68 (m, 2H), 2.46–2.34 (m, 2H), 1.98 (s, 3H), 1.89–1.83
(m, 2H), 1.07 ppm (s, 9H); 13C NMR (100 MHz, CDCl3): d = 153.3,
135.7, 135.6, 133.8, 133.6, 133.2, 129.7, 127.8, 118.2, 85.3, 72.8, 72.6, 60.0,
38.6, 36.1, 26.9, 19.2, 3.9 ppm; IR (thin film): n˜ =3071, 3050, 2957, 2359,
2342, 2243, 1706, 1472, 1428, 1389, 1249, 1110, 1063, 700 cmꢀ1. HRMS
(ESI): m/z: calcd for C23H27O3Si [MꢀCH3H5+] 379.1730, found 379.1724.
To a solution of the above ester (1.77 g, 4.20 mmol, 1.00 equiv) in CH2Cl2
(40 mL) at ꢀ788C was added slowly DIBAL-H (1m in toluene, 8.40 mL,
8.40 mmol, 2.00 equiv) such that the temperature did not exceed ꢀ708C;
after 30 min pyridine (1 mL, 12.60 mmol, 3.00 equiv), DMAP (1.54 g,
12.60 mmol, 3.00 equiv), and Ac2O (2.37 mL, 25.20 mmol, 6.00 equiv)
were added sequentially at ꢀ788C and the mixture was stirred at this
temperature for 22 h. Sat aq NH4Cl (20 mL) and sat aq Rochelle salt
(40 mL) were added at ꢀ788C and the mixture was allowed to warm to
RT. Vigorous stirring was continued for 90 min in a beaker, resulting in
the formation of two clear phases that were readily separable. The aque-
ous phase was extracted with CH2Cl2 (3ꢃ40 mL) and the combined or-
ganic phases were washed with sat aq NaHCO3 (2ꢃ20 mL) and brine
(10 mL), and then dried over MgSO4. Concentration of the solution
under reduced pressure and purification of the residue by FC (EtOAc/
Hex 1:30!1:20, 2% NEt3 v/v) afforded 15 (1.80 g, 3.87 mmol, 92%) as a
1.6:1 mixture of diastereomers as a colorless, viscous oil. Spectroscopic
data are for the diastereomeric mixture. Rf =0.40 (EtOAc/Hex 1:10, UV,
CPS); [a]2D4 =ꢀ19.248 (c = 0.99, CHCl3); 1H NMR (400 MHz, CDCl3):
d=7.72–7.66 (m, 4H), 7.45–7.36 (m, 6H), 6.45 (q, J=1.8; 0.37H), 6.44
(q, J=1.8; 0.63H), 5.87–5.75 (m, 1H), 5.12–5.02 (m, 2H), 4.17–4.06 (m,
1H), 3.85–3.71 (m, 2H), 2.41–2.30 (m, 2H), 2.05 (s, 1.12H), 1.99 (s,
1.88H), 1.86 (d, J=1.8; 1.15H), 1.84 (d, J=1.8; 1.85H), 1.83–1.71 (m,
2H), 1.07 (s, 3.38H), 1.06 ppm (s, 5.62H); 13C NMR (100 MHz, CDCl3):
d=169.9, 169.7, 135.7, 135.7, 135.6, 135.6, 134.5, 134.1, 134.0, 133.9, 133.9,
702 cmꢀ1
475.0590, found 475.0585.
;
HRMS (ESI): m/z: calcd for C22H24IO2Si [MꢀC4H9+],
tert-Butyl-(2-((2R,6S)-6-((E)-2-iodoprop-1-enyl)-4-methylenetetrahydro-
2H-pyran-2-yl)ethoxy) diphenylsilane (18): To a suspension of CuCN
(668 mg, 7.42 mmol, 5.00 equiv) in THF (16 mL) at ꢀ788C was added a
solution of nBuLi (1.6m in hexane, 9.30 mL, 14.82 mmol, 10.00 equiv).
After 5 min the flask was immersed in a cooling bath at ꢀ408C, resulting
in the formation of a pale-yellow, almost clear solution. The mixture was
cooled back to ꢀ788C after 10 min, which made it become slightly het-
erogenous. Neat Bu3SnH (4.00 mL, 14.82 mmol, 10.00 equiv) was then
added dropwise, immediately leading to a turbid yellow solution with lib-
eration of gas. After 20 min at ꢀ788C the mixture was stirred for 5 min
at ꢀ408C, giving an almost clear golden-yellow solution. After 10 min at
ꢀ408C the solution was cooled back to ꢀ788C followed by addition of
MeOH (6.60 mL, 163.00 mmol, 110.00 equiv) under vigorous stirring.
After 10 min at ꢀ788C the flask was immersed in a cooling bath at
ꢀ408C; the reaction mixture now was a clear red solution. After 10 min
at ꢀ408C this solution was cooled back to ꢀ788C and a solution of 17
(0.62 g, 1.48 mmol, 1.00 equiv) in THF (10 mL) was added. The mixture
was stirred for 15 h, during which period the temperature was allowed to
rise to ꢀ158C. Sat aq NH4Cl (30 mL) and 25% aq NH4OH (6 mL) were
then added together with EtOAc (20 mL). Stirring was continued for
30 min, the two almost clear phases were separated, and the aqueous
phase was extracted with EtOAc (3ꢃ10 mL). The combined organic ex-
tracts were dried over MgSO4 and the solution was concentrated under
reduced pressure. Purification of the residue by FC (Hex!EtOAc/Hex
1:100!1:50, 1% (v/v) NEt3) gave the vinylstannane (1.02 g, 1.43 mmol,
97%) as a pale-yellow oil that was used immediately in the next step.
A solution of the above vinylstannane in THF (11 mL) was cooled to
ꢀ178C (NaCl/ice) followed by addition of N-iodosuccinimide (0.49 g,
2.10 mmol, 1.50 equiv) in THF (2 mL), to give an almost clear yellow sol-
ution. After 20 min a mixture of sat aq Na2S2O3 (5 mL) and sat aq
NaHCO3 (5 mL) was added followed by EtOAc (5 mL). Stirring was con-
tinued for 2 min when two clear, colorless phases had formed. The
phases were separated and the aqueous phase was extracted with EtOAc
(3ꢃ5 mL). The combined organic extracts were dried over MgSO4 and
then concentrated under reduced pressure. The residue was purified by
FC (Hex/EtOAc 1:100) to afford the desired product 18 (0.79 g,
1.44 mmol, quant.) as a pale yellow oil. Rf =0.64 (EtOAc/Hex 1:20, UV,
CPS); [a]2D4 = +4.908 (c = 1.81, CHCl3); 1H NMR (400 MHz, CDCl3): d =
16878
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16868 – 16883

Total Synthesis of (ꢀ)-Zampanolide
FULL PAPER
7.70–7.67 (m, 4H), 7.47–7.38 (m, 6H), 6.24 (dq, J=7.7, 1.5; 1H), 4.81–
4.77 (m, 2H), 3.99 (ddd, J=10.8, 7.7, 2.6; 1H), 3.87 (ddd, J=10.1, 8.1,
5.4; 1H), 3.76 (dt, J=10.1, 5.6; 1H), 3.64–3.57 (m, 1H), 2.44 (d, J=1.5;
3H), 2.27–2.20 (m, 2H), 2.13–2.06 (m, 1H), 2.00–1.94 (m, 1H), 1.87–1.73
(m, 2H), 1.08 ppm (s, 9H); 13C NMR (100 MHz, CDCl3): d=141.6, 135.6,
135.6, 134.0, 133.9, 129.7, 127.7, 109.3, 98.5, 76.3, 75.3, 60.2, 40.6, 40.4,
39.0, 28.8, 26.9, 19.3 ppm; IR (thin film): n˜ =3070, 2931, 2890, 2856, 1651,
1472, 1427, 1360, 1105, 1087, 998, 858, 700 cmꢀ1; HRMS (ESI): m/z: calcd
for C27H36IO2Si [M+H+], 547.1524, found 547.1503. The 1H NMR spec-
trum indicated the presence of ca. 3% of the undesired regioisomer.
(100 MHz, CDCl3): d=198.3, 166.9, 159.4, 146.5, 143.9, 142.9, 139.4,
132.6, 131.6, 130.2, 129.5, 129.5, 125.6, 121.3, 114.0, 109.1, 76.7, 76.1, 73.1,
71.6, 69.6, 55.4, 45.2, 42.8, 41.1, 40.9, 40.4, 23.6, 16.8 ppm; IR (thin film):
n˜ =3016, 2923, 2852, 1713, 1668, 1635, 1614, 1513, 1463, 1360, 1281, 1249,
1215, 1176, 1152, 1086, 1035, 978 cmꢀ1; HRMS (ESI): m/z: calcd for
C31H38NaO6 [M+Na+], 529.2561, found 529.2571.
(1S,2E,5S,8E,10Z,14E,17S)-5-(Hydroxymethyl)-3,11-dimethyl-19-methyl-
ene-6,21-dioxabicycloACTHNUTRGNEU[GN 15.3.1]henicosa-2,8,10,14-tetraene-7,13-dione (41):
To a solution of PMB ether 40 (4 mg, 0.008 mmol, 1.00 equiv) in CH2Cl2
(0.5 mL) was added H2O (0.1 mL) followed by DDQ (5.4 mg,
0.024 mmol, 3.50 equiv) at RT. The mixture was vigorously stirred for
3 h. Then sat aq NaHCO3 (5 mL) and CH2Cl2 (5 mL) were added and the
phases were separated. The aqueous phase was extracted with CH2Cl2
(3ꢃ5 mL) and the combined organic phases were dried over MgSO4 and
concentrated under reduced pressure. Purification by FC (EtOAc/Hex
1:3!1:2) gave 41 (2.49 mg, 0.0064 mmol, 82%) as a colorless solid. Rf =
0.30 (EtOAc/Hex 1:1, UV, CPS); [a]2D4 =ꢀ136.268 (c 0.11, CHCl3);
1H NMR (400 MHz, CDCl3): d=7.64 (dd, J=15.1, 11.6; 1H), 6.84 (ddd,
J=16.2, 9.6, 4.6; 1H), 6.11 (d, J=11.7; 1H), 5.94 (d, J=15.1; 1H), 5.93
(d, J=16.5; 1H), 5.28 (dddd, J=10.8, 5.9, 4.1, 2.1; 1H), 5.19 (d, J=8.0,
1H), 4.73 (d, J=1.6; 1H), 4.73 (d, J=1.6; 1H), 4.14 (d, J=13.7; 1H),
3.97 (ddd, J=11.2, 8.2, 2.7, 1H), 3.77–3.70 (m, 2H), 3.29 (ddt, J=11.8,
9.5, 2.1, 1H), 3.04 (d, J=13.7, 1H), 2.38 (dddd, J=15.1, 10.1, 4.6, 2.0,
1H), 2.30–2.08 (m, 5H), 1.98–1.91 (m, 2H), 1.81 (s, 3H), 1.73 ppm (d, J=
1.2, 3H); 13C NMR (100 MHz, CDCl3): d=198.1, 167.1, 146.5, 143.9,
143.3, 139.8, 132.6, 131.6, 129.6, 125.6, 121.0, 109.2, 76.7, 76.1, 71.9, 65.4,
45.2, 42.1, 41.1, 40.8, 40.3, 23.7, 16.8 ppm; IR (thin film): n˜ =3389, 2925,
2853, 1715, 1669, 1634, 1553, 1449, 1436, 1357, 1280, 1259, 1148, 1086,
1049, 1019, 976, 799 cmꢀ1; HRMS (ESI): m/z: calcd for C23H30NaO5 [M+
Na+]: 409.1985; found: 409.1983.
ACHTUNGTRENNUNG(S,E)-5-((2S,6R)-6-(2-(tert-Butyldiphenylsilyloxy)ethyl)-4-methylenetetra-
hydro-2H-pyran-2-yl)-1-(4-methoxybenzyloxy)-4-methylpent-4-en-2-ol
(20): Vinyl iodide 18 (385 mg, 0.70 mmol, 1.00 equiv, azeotropically dried
once with 2 mL of acetonitrile or toluene immediately before use) was
dissolved in dry toluene (7 mL) and the solution was cooled to ꢀ788C.
tBuLi (1.6m in pentane, 0.88 mL, 1.41 mmol, 2.00 equiv) was then added
and the near colorless solution was stirred for 30 min; it was then cooled
to around ꢀ85–ꢀ908C with liquid nitrogen and a solution of 19 (342 mg,
1.76 mmol, 2.50 equiv, azeotropically dried once with 2 mL of acetonitrile
or toluene immediately before use) in dry toluene (2 mL) was added fol-
lowed by BF3·OEt2 (0.22 mL, 1.76 mmol, 2.50 equiv; addition of
BF3·OEt2 about 1 min after the addition of 19) giving a pale yellow solu-
tion. Stirring was continued at ꢀ788C for 1 h; then the cooling bath was
removed and sat aq NaHCO3 (10 mL) and 10 mL of EtOAc were added.
After the mixture had reached RT, the phases were separated and the
aqueous phase was extracted with EtOAc (3ꢃ5 mL). The combined or-
ganic extracts were dried over MgSO4, concentrated under reduced pres-
sure, and the residue purified by FC (EtOAc/Hex 1:5!1:4) to give 20
(264.2 mg, 0.43 mmol, 61%) as a colorless oil. Note: Chromatographic
separation was difficult and two FC runs were needed in order to remove
the iodohydrin derived from competing epoxide opening by iodide. Rf =
0.17 (EtOAc/Hex 1:5, UV, CPS); [a]2D4 = +5.978 (c = 0.88, CHCl3);
1H NMR (400 MHz, CDCl3): d=7.68–7.65 (m, 4H), 7.44–7.34 (m, 6H),
7.28–7.24 (m, 2H), 6.90–6.87 (m, 2H), 5.29 (dq, J=7.7, 1.2; 1H), 4.75–
4.73 (m, 2H), 4.49 (s, 2H), 3.99 (ddd, J=10.9, 7.7, 2.7; 1H), 3.98–3.91 (m,
1H), 3.84 (ddd, J=10.1, 8.0, 5.5; 1H), 3.80 (s, 3H), 3.74 (dt, J=10.1, 5.7;
1H), 3.60–3.54 (m, 1H), 3.46 (dd, J=9.5, 3.5; 1H), 3.33 (dd, J=9.5, 7.1;
1H), 2.31 (d, J=3.5; 1H), 2.25–2.22 (m, 1H), 2.20 (d, J=6.8; 2H), 2.16–
2.12 (m, 1H), 2.04–2.00 (m, 1H), 1.97–1.90 (m, 1H), 1.89–1.80 (m, 1H),
1.77–1.71 (m, 1H), 1.69 (d, J=1.2; 3H), 1.05 ppm (s, 9H); 13C NMR
(100 MHz, CDCl3): d=159.4, 144.7, 135.7, 135.7, 135.4, 134.1, 134.0,
130.2, 129.7, 129.5, 129.0, 127.7, 127.7, 114.0, 108.7, 75.5, 75.3, 73.7, 73.2,
68.6, 60.4, 55.4, 43.7, 41.0, 40.7, 39.2, 27.0, 19.4, 17.3 ppm; IR (thin film):
n˜ =3070, 2932, 2857, 1612, 1513, 1471, 1427, 1247, 1106, 1087, 1058, 1036,
998, 821, 702 cmꢀ1; HRMS (ESI): m/z: calcd for C38H50NaO5Si [M+Na+
], 637.3320, found 637.3322.
(ꢀ)-Dactylolide (ent-2): To a solution of alcohol 41 (2.33 mg, 0.006 mmol,
1.00 equiv) in CH2Cl2 (0.5 mL) was added DMP (15 mg, 0.036 mmol,
6.00 equiv; added in 3 equal portions in 20 min intervals) and stirring was
continued for 60 min. A mixture of sat aq NaHCO3 (5 mL) and sat aq
Na2S2O3 (5 mL) was then added and stirring was continued for 10 min,
when two clear phases had formed. The phases were separated, and the
aqueous phase was extracted with CH2Cl2 (3ꢃ10 mL). The combined or-
ganic extracts were dried over MgSO4, concentrated under reduced pres-
sure and the residue was purified by FC (EtOAc/Hex 1:3) to provide ent-
2 (1.8 mg, 0.0048 mmol, 78%) as a colorless solid. Rf =0.57 (EtOAc/Hex
1:1, UV, CPS or KMnO4); [a]D24 =ꢀ258.338 (c = 0.11, MeOH); 1H NMR
(400 MHz, CDCl3): d=9.67 (s, 1H), 7.63 (dd, J=15.1, 11.6; 1H), 6.85
(ddd, J=16.2, 8.6, 6.0; 1H), 6.16 (d, J=11.7; 1H), 6.03–5.94 (m, 2H),
5.32 (dd, J=11.3, 2.5; 1H), 5.24 (d, J=8.0; 1H), 4.75 (d, J=1.6; 1H),
4.75 (d, J=1.6; 1H), 3.97 (ddd, J=11.5, 8.1, 2.7; 1H), 3.94 (d, J=14.3;
1H), 3.33 (ddt, J=11.1, 8.7, 2.7; 1H), 3.24 (d, J=14.5; 1H), 2.55 (d, J=
14.3; 1H), 2.36–2.28 (m, 3H), 2.19–2.15 (m, 1H), 2.14–2.09 (m, 1H),
1.99–1.93 (m, 2H), 1.87 (s, 3H), 1.72 ppm (d, J=0.9; 3H); 13C NMR
(100 MHz, CDCl3): d=199.2, 197.6, 166.4, 146.1, 144.2, 143.6, 140.6,
131.6, 131.1, 130.7, 125.7, 119.9, 109.5, 76.6, 75.9, 75.5, 45.0, 40.9, 40.6,
39.9, 39.8, 24.3, 16.2 ppm; IR (thin film): n˜ =2936, 2858, 1733, 1716, 1706,
(1S,2E,5S,8E,10Z,14E,17S)-5-((4-Methoxybenzyloxy)methyl)-3,11-di-
methyl-19-methylene-6,21-dioxabicycloACTHUNTGRNEUNG[15.3.1]henicosa-2,8,10,14-tet-
raene-7,13-dione (40): To a stirred solution of 39 (62.2 mg, 0.094 mmol,
1.00 equiv, co-evaporated with 3 mL of toluene immediately before use)
in THF (31 mL) was added H2O (0.8 mL) followed by freshly activated
Ba(OH)2·0.8H2O at 08C. (Commercial Ba(OH)2 was activated by heating
to 100–1408C for 1–2 h before use)).[54d] The cooling bath was removed
after 30 min and stirring was continued at RT for additional 30 min;
30 mL of Et2O were then added and the solution was washed first with
sat aq NaHCO3 (2ꢃ10 mL) and then with brine (1ꢃ10 mL). The clear or-
ganic phase was dried over MgSO4 and concentrated. The resulting
yellow oil was purified by FC (EtOAc/Hex 1:3) to afford 40 (38.6 mg,
0.076 mmol, 81%) as a colorless oil. Rf =0.40 (EtOAc/Hex 1:3, UV,
KMnO4, CPS); [a]2D4 =ꢀ158.798 (c = 0.25, CHCl3); 1H NMR (400 MHz,
CDCl3): d=7.62 (dd, J=15.1, 11.6; 1H), 7.27–7.24 (m, 2H), 6.90–6.86
(m, 2H), 6.83 (ddd, J=16.2, 9.8, 4.4; 1H), 6.10 (d, J=11.6; 1H), 5.94 (d,
J=15.1; 1H), 5.92 (d, J=16.4; 1H), 5.40–5.34 (m, 1H), 5.17 (dd, J=8.1,
0.9; 1H), 4.74–4.70 (m, 2H), 4.52 (d, J=11.8; 1H), 4.48 (d, J=11.8; 1H),
4.17 (d, J=13.6; 1H), 3.96 (ddd, J=11.3, 8.1, 2.5; 1H), 3.81 (s, 3H), 3.58
(dd, J=10.4, 6.0; 1H), 3.51 (dd, J=10.4, 4.9; 1H), 3.30–3.24 (m, 1H),
3.00 (d, J=13.5; 1H), 2.37 (dddd, J=15.0, 10.1, 4.4, 2.0; 1H), 2.26–2.20
(m, 1H), 2.20 (d, J=6.7; 2H), 2.16–2.11 (m, 1H), 2.11–2.05 (m, 1H),
1.97–1.89 (m, 2H), 1.79 (s, 3H), 1.70 ppm (d, J=1.1; 3H); 13C NMR
1670, 1635, 1438, 1355, 1278, 1256, 1144, 1086, 1050, 978, 890 cmꢀ1
;
HRMS (ESI): m/z: calcd for C23H28NaO5 [M+Na+]: 407.1829; found:
407.1820.
(ꢀ)-Zampanolide (1): To a solution of amide 42 (36.6 mg, 0.33 mmol,
4.6 equiv) in THF (2 mL) was added DIBAL-H (1m in CH2Cl2, 0.27 mL,
0.27 mmol, 3.76 equiv) and the mixture was stirred at RT for 45 min.
After that time a solution of ent-2 (27.6 mg, 0.072 mmol, 1 equiv) in THF
(1 mL, flask rinsed twice with 0.5 mL of THF) was added and stirring
was continued for a total of 3 h. Sat aq Rochelle salt (10 mL) was then
added together with EtOAc (10 mL) and the mixture was stirred for
15 min. After addition of brine (10 mL) the phases were separated and
the aqueous phase was further extracted with EtOAc (3ꢃ5 mL). The
combined organic extracts were dried over MgSO4, the solvent was
evaporated and the residue was purified by FC (EtOAc/Hex 1:3!1:1,
2% NEt3 v/v) to give a 1.1:1 mixture of 1 and epi-1 (16.4 mg, 0.033 mmol,
46%) as a pale-yellow foam. The epimers were separated by semiprepar-
ative normal phase HPLC (Phenomenex Luna 5 mm NH2 100 ꢄ, 150ꢃ
Chem. Eur. J. 2012, 18, 16868 – 16883
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16879

K.-H. Altmann et al.
10 mm, EtOH/Hex 1:9) followed by the purification of the individual iso-
mers by reverse phase HPLC (Waters Symmetry C18, 5 m, 100ꢃ7.8 mm,
CH3CN/H2O 1:1). After lyophilization, 6.4 mg (0.013 mmol, 18%) of
(ꢀ)-zampanolide (1) and 4.42 mg (0.0089 mmol, 12%) of epi-(ꢀ)-1 were
obtained. Rf =0.40 (EtOAc/Hex 1:1, UV, CPS); [a]2D4 =ꢀ241.338 (c =
0.18, CHCl3, deactivated over basic Alox); 1H NMR (500 MHz, CDCl3):
d=8.35 (d, J=8.9; 1H). 7.51 (dd, J=14.9, 11.8; 1H). 7.45 (dd, J=14.9,
11.8; 1H), 6.75 (ddd, J=16.3, 8.6, 5.7; 1H), 6.36 (t, J=11.3; 1H), 6.20 (d,
J=11.9; 1H), 6.18 (brs, 1H), 6.00–5.94 (m, 1H), 5.95 (d, J=15.9; 1H),
5.93 (d, J=15.1; 1H), 5.65 (d, J=11.4; 1H), 5.32 (dd, J=8.4, 6.4; 1H),
5.10 (d, J=7.7; 1H), 4.96 (dd, J=10.2, 6.2; 1H), 4.73 (brs, 2H), 4.13 (d,
J=14.2; 1H), 3.86 (ddd, J=11.4, 7.7, 1.8; 1H), 3.26 (t, J=10.1, 1H), 3.00
(d, J=14.3; 1H), 2.35–2.26 (m, 3H), 2.17 (d, J=12.7; 1H), 2.11–2.05 (m,
2H), 1.89–1.86 (m, 1H), 1.85–1.82 (m, 1H), 1.79 (d, J=6.7; 3H), 1.74 (s,
3H), 1.61 ppm (s, 3H); 13C NMR (125 MHz, CDCl3): d=197.3, 165.6,
165.3, 145.9, 143.8, 143.0, 140.6, 139.5, 137.2, 132.5, 130.9, 129.0, 128.6,
125.1, 120.7, 119.2, 109.0, 76.0, 75.1, 72.9, 72.0, 44.9, 40.9, 40.3, 40.3, 39.3,
23.6, 18.3, 16.7 ppm; IR (thin film): n˜ =3325, 3015, 2960, 2924, 2853,
1708, 1664, 1634, 1604, 1520, 1431, 1355, 1281, 1259, 1213, 1147, 1085,
1050, 1034, 1025, 802 cmꢀ1; HRMS (ESI): m/z: calcd for C29H38NO6 [M+
H+]: 496.2694; found: 496.2681.
10 min hexylamine (0.026 mL, 0.195 mmol, 3 equiv; freshly distilled im-
mediately before use) was added and the mixture was stirred for a total
of 16 h. Water (5 mL) was then added followed by Et2O (5 mL). The
phases were separated and the aqueous phase was further extracted with
Et2O (3ꢃ5 mL). The combined organic extracts were washed with H2O
(2ꢃ5 mL) and the washing solutions were re-extracted with CH2Cl2 (2ꢃ
5 mL). The combined organic extracts were dried over MgSO4, concen-
trated under reduced pressure and the residue was purified by FC
(EtOAc/Hex 1:5!1:3) to give amide 45 (4.1 mg, 0.0085 mmol, 13%) as a
yellow oil. Rf =0.15 (EtOAc/Hex 1:3, UV, CPS); [a]2D4 =ꢀ166.228 (c =
0.82 in CHCl3); 1H NMR (400 MHz, CDCl3): d=7.70 (dd, J=15.1, 11.6;
1H), 6.82 (ddd, J=16.4, 9.5, 4.7; 1H), 6.23–6.19 (m, 1H), 6.14 (d, J=
11.5; 1H), 5.97 (d, J=14.9; 1H), 5.93 (d, J=16.0; 1H), 5.56 (dd, J=11.2,
2.1; 1H), 5.18 (d, J=8.1; 1H), 4.75–4.71 (m, 2H), 4.19 (d, J=13.8; 1H),
3.97 (ddd, J=11.2, 8.1, 2.6; 1H), 3.32–3.24 (m, 3H), 3.01 (d, J=13.7;
1H), 2.74 (d, J=13.7; 1H), 2.34 (dddd, J=14.8, 10.0, 4.8, 1.8; 1H), 2.28–
2.25 (m, 1H), 2.24–2.21 (m, 1H), 2.17–2.13 (m, 1H), 2.12–2.07 (m, 1H),
1.99–1.88 (m, 2H), 1.83 (s, 3H), 1.74 (d, J=0.9; 3H), 1.55–1.48 (m, 3H),
1.35–1.28 ppm (m, 8H); 13C NMR (100 MHz, CDCl3): d = 197.8, 169.6,
165.5, 146.6, 144.6, 143.7, 140.9, 132.6, 131.7, 130.1, 125.3, 119.7, 109.3,
76.7, 76.0, 71.1, 45.1, 43.7, 41.1, 40.8, 40.3, 39.5, 31.6, 29.7, 26.7, 23.9, 22.7,
16.4, 14.1 ppm; IR (thin film): n˜ =3336, 2929, 2859, 1717, 1668, 1635,
1533, 1436, 1355, 1277, 1256, 1206, 1176, 1141, 1117, 1086, 1053, 977,
888 cmꢀ1; HRMS (ESI): m/z: calcd for C29H42NO5 [M+H+]: 484.3057;
found: 484.3057.
(ꢀ)-epi-Zampanolide (epi-1): Rf =0.40 (EtOAc/Hex 1:1, UV, CPS);
[a]2D4 =ꢀ172.928 (c = 0.65, CHCl3, deactivated over Alox); 1H NMR
(400 MHz, CDCl3): d=8.39 (d, J=9.0; 1H), 7.48 (dd, J=15.1, 11.5; 1H),
7.47–7.42 (m, 1H), 6.74 (ddd, J=16.1, 8.3, 5.7; 1H), 6.41 (t, J=11.3; 1H),
6.22 (d, J=11.5; 1H), 6.03–6.01 (m, 1H), 6.01–5.97 (m, 1H), 5.98 (d, J=
14.8; 1H), 5.93 (d, J=16.2; 1H), 5.65 (d, J=11.4; 1H), 5.33 (dd, J=8.9,
6.0; 1H), 5.07 (d, J=7.9; 1H), 5.02 (ddd, J=9.8, 5.9, 2.9; 1H), 4.72 (brs,
2H), 4.16 (d, J=14.1; 1H), 3.87 (ddd, J=11.0, 8.2, 2.4; 1H), 2.91 (d, J=
14.2; 1H), 2.33–2.27 (m, 2H), 2.18–2.09 (m, 3H), 2.07–2.03 (m, 1H),
1.87–1.79 (m, 2H), 1.80 (dd, J=6.8, 1.1; 3H), 1.75 (s, 3H), 1.62 ppm (s,
3H) (one signal overlapping with the solvent peak); 13C NMR (125 MHz,
CDCl3): d=197.4, 165.9, 165.2, 146.0, 143.7, 142.7, 140.9, 138.9, 137.4,
132.0, 130.9, 129.1, 128.6, 125.2, 121.3, 119.1, 109.0, 75.9, 75.2, 72.8, 71.8,
44.9, 40.6, 40.3, 40.3, 39.3, 23.5, 18.4, 16.4 ppm; IR (thin film): n˜ =3325,
2962, 2927, 2853, 1714, 1654, 1634, 1520, 1431, 1355, 1280, 1259, 1213,
(S,3E,9E,11Z,15E)-6-((4-Methoxybenzyloxy)methyl)-4,12-dimethyl-1,7-
dioxacyclooctadeca-3,9,11,15-tetraene-8,14-dione (57): To a solution of
phosphonate 56 (216.1 mg, 0.355 mmol, 1.0 equiv; co-evaporated once
with 2 mL of toluene immediately before use) in THF (300 mL) and H2O
(7.5 mL) was added freshly activated Ba(OH)2·0.8H2O[53d] (53 mg,
0.284 mmol, 0.8 equiv) at 08C. After 30 min the cooling bath was re-
moved and stirring of the orange mixture was continued for a total of
3 h. Et2O (50 mL) was then added followed by sat aq NaHCO3 (50 mL),
the phases were separated, and the organic phase was washed with sat aq
NaHCO3 (50 mL) and with brine (50 mL). The combined aqueous ex-
tracts were washed once with Et2O (20 mL) and the combined organic
extracts were dried over MgSO4 and concentrated under reduced pres-
sure. The remaining yellow oil was purified by FC (EtOAc/Hex 1:3!1:1)
to afford 136.9 mg of macrolactone 57 (0.30 mmol, 85%) as a pale-yellow
oil. Rf =0.50 (EtOAc/Hex 1:1, UV, CPS); [a]2D4 =ꢀ76.058 (c = 0.61,
CHCl3); 1H NMR (400 MHz, CDCl3): d=7.60 (dd, J=15.2, 11.6; 1H),
7.26–7.24 (m, 2H), 6.89–6.80 (m, 3H), 6.12 (d, J=11.6; 1H), 6.04 (d, J=
16.2; 1H), 5.90 (d, J=15.2; 1H), 5.41–5.35 (m, 1H), 5.30–5.27 (m, 1H),
4.53 (d, J=11.8; 1H), 4.47 (d, J=11.8; 1H), 4.01 (dd, J=12.1, 8.0; 1H),
3.88 (dd, J=12.1, 4.7; 1H), 3.80 (s, 3H), 3.76 (d, J=12.8; 1H), 3.59–3.37
(m, 4H), 3.26 (d, J=12.8; 1H), 2.48–2.38 (m, 2H), 2.35–2.24 (m, 2H),
1.83 (s, 3H), 1.69 ppm (s, 3H); 13C NMR (100 MHz, CDCl3): d=197.1,
166.7, 159.4, 146.8, 142.4, 139.5, 134.6, 130.3, 130.2, 129.5, 125.9, 124.9,
121.3, 114.0, 73.0, 71.6, 69.7, 67.8, 67.8, 55.4, 45.9, 42.0, 33.0, 24.1,
16.7 ppm; IR (thin film): n˜ =3009, 2999, 2959, 2916, 2857, 1708, 1667,
1633, 1613, 1586, 1513, 1456, 1441, 1360, 1301, 1279, 1247, 1208, 1173,
1148, 1089, 1033, 976, 890, 846, 819 cmꢀ1; HRMS (ESI): m/z: calcd for
C27H34NaO6 [M+Na+]: 477.2248; found: 477.2230.
1147, 1085, 1048, 1034, 1024 cmꢀ1
; HRMS (ESI): m/z: calcd for
C29H37NNaO6 [M+Na+]: 518.2513; found: 518.2518.
(1S,2E,5S,8E,10Z,14E,17S)-3,11-Dimethyl-19-methylene-7,13-dioxo-6,21-
dioxabicycloACHTUNGTRENNUNG[15.3.1]henicosa-2,8,10,14-tetraene-5-carboxylic acid (44): To
a solution of ent-2 (9.5 mg, 0.0247 mmol, 1 equiv) in tBuOH (3 mL) and
2-methyl-2-butene (2 mL, 18.88 mmol, 764 equiv) was added a solution of
NaClO2 (22.3 mg, 0.247 mmol, 10 equiv) and NaH2PO4·H2O (27.3 mg,
0.198 mmol, 8 equiv) dissolved in H2O (1.2 mL) slowly at RT. After
40 min stirring the reaction mixture was diluted with brine (10 mL) and
the phases were separated. The aqueous phase was extracted with
EtOAc (3ꢃ10 mL) and the combined organic extracts were dried over
MgSO4 and concentrated under reduced pressure. Purification of the resi-
due by FC (EtOAc, 0.5% AcOH) gave 44 (9.6 mg, 0.024 mmol, 97%;
after coevaporation with toluene (1ꢃ2 mL)). Rf =0.31 (EtOAc, 0.5%
AcOH, UV, CPS or KMnO4); [a]2D4 =ꢀ68.828 (c
= 0.49, CHCl3);
1H NMR (400 MHz, CDCl3): d=7.56 (dd, J=15.1, 11.7; 1H), 6.85 (dt,
J=16.2, 7.2; 1H), 6.13 (d, J=11.8; 1H), 6.04 (d, J=16.1; 1H), 5.91 (d,
J=15.5; 1H), 5.43 (dd, J=11.3, 2.6; 1H), 5.30 (d, J=7.9; 1H), 4.75 (s,
2H), 3.96 (ddd, J=11.1, 7.9, 2.5; 1H), 3.73 (d, J=14.8; 1H), 3.41 (d, J=
14.8; 1H), 3.39–3.32 (m, 1H), 2.63 (brd, J=13.5; 1H), 2.53 (dd, J=14.1,
11.3; 1H), 2.35–2.30 (m, 2H), 2.19–2.15 (m, 1H), 2.14–2.09 (m, 1H),
2.00–1.93 (m, 2H), 1.87 (s, 3H), 1.71 ppm (s, 3H); 13C NMR (100 MHz,
CDCl3): d=197.4, 174.6, 166.2, 145.8, 143.8, 143.7, 140.3, 131.7, 131.2,
130.7, 126.0, 120.2, 109.5, 76.5, 75.8, 69.1, 45.0, 41.8, 40.8, 40.5, 39.4, 24.5,
16.1 ppm; IR (thin film): n˜ =3020, 2936, 1714, 1711, 1635, 1436, 1355,
1258, 976, 889, 752 cmꢀ1; HRMS (ESI): m/z: calcd for C23H28NaO8 [M+
Na+]: 423.1778; found: 423.1767.
(S,3E,9E,11Z,15E)-6-(Hydroxymethyl)-4,12-dimethyl-1,7-dioxacyclooc-
tadeca-3,9,11,15-tetraene-8,14-dione (58): To a solution of macrolactone
57 (72 mg, 0.158 mmol, 1.0 equiv) in CH2Cl2 (4 mL) was added H2O
(0.8 mL) followed by DDQ (72 mg, 0.32 mmol, 2.0 equiv) and the mix-
ture was stirred vigorously at room temperature. After 60 min the reac-
tion mixture was added to sat aq NaHCO3 (10 mL) and CH2Cl2 (5 mL),
the phases were separated, and the aqueous phase was extracted with
CH2Cl2 (3ꢃ10 mL). The combined organic extracts were dried over
MgSO4, filtered and the filtrate concentrated under reduced pressure. Pu-
rification of the residue by FC (EtOAc/Hex 1:1) gave alcohol 58
(40.7 mg, 0.122 mmol, 77%) as a pale-yellow oil. Rf =0.19 (EtOAc/Hex
1:1, UV, CPS); [a]2D4 =ꢀ74.678 (c= 0.29, CHCl3); 1H NMR (400 MHz,
CDCl3): d=7.62 (dd, J=15.1, 11.6; 1H), 6.85 (dt, J=16.2, 6.6; 1H), 6.13
(d, J=11.5; 1H), 6.06 (dt, J=16.2, 1.5; 1H), 5.91 (d, J=15.2; 1H), 5.32–
5.25 (m, 2H), 4.02 (dd, J=12.0, 7.9; 1H), 3.91–3.86 (m, 1H), 3.79–3.70
(m, 3H), 3.51–3.38 (m, 2H), 3.29 (d, J=12.9; 1H), 2.49–2.33 (m, 3H),
(1S,2E,5S,8E,10Z,14E,17S)-N-Hexyl-3,11-dimethyl-19-methylene-7,13-
dioxo-6,21-dioxabicycloACHTUNGTRENNUNG[15.3.1]henicosa-2,8,10,14-tetraene-5-carboxamide
(45): To a solution of 44 (26 mg, 0.065 mmol, 1 equiv) in dry DMF
(2 mL) was added HATU (27.4 mg, 0.072 mmol, 1.1 equiv) and DIEA
(0.023 mL, 0.13 mmol, 2 equiv), producing
a yellow solution. After
16880
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16868 – 16883

Total Synthesis of (ꢀ)-Zampanolide
FULL PAPER
2.22 (d, J=13.8; 1H), 1.84 (s, 3H), 1.71 ppm (s, 3H); 13C NMR
(100 MHz, CDCl3): d=197.0, 167.1, 146.8, 142.8, 139.8, 134.3, 130.3,
125.9, 125.0, 121.0, 72.2, 68.0, 67.8, 65.5, 46.0, 41.5, 33.0, 24.1, 16.8 ppm;
IR (thin film): n˜ =3442, 2929, 2855, 1703, 1693, 1667, 1631, 1437, 1380,
1359, 1279, 1258, 1208, 1174, 1148, 1113, 1088, 1059, 1038, 976, 936,
891 cmꢀ1; HRMS (ESI): m/z: calcd for C19H26NaO5 [M+Na+]: 357.1672;
found: 357.1666.
Cytotoxicity: Inhibition of cell proliferation was determined in the MCF-
7, (breast), A549 (lung), HCT-116 (colon), and PC-3 (prostate) cell lines,
which were obtained as a kind gift from Markus Wartmann (Novartis In-
stitute for Biomedical Research (NIBR) Basel, Switzerland). Cells were
maintained in a 5% CO2 humidified atmosphere at 378C in RPMI
medium 1640 (Gibco BRL) containing 10% fetal bovine serum, penicil-
lin (100 UmLꢀ1) and streptomycin (100 mgmLꢀ1) (Gibco BRL). Cells
were seeded at 1.5ꢃ103 per well into 96-well microtiter plates and incu-
bated overnight. Compounds were added in serial dilutions on day 1.
Subsequently, the plates were incubated for two population doublings
(72 h) and then fixed with 3.3% v/v glutaraldehyde, washed with water
and stained with 0.05% methylene blue. After washing, the dye was
eluted with 3% v/v HCl and the optical density (OD) measured at
665 nm with a TECAN GeniosPro (Switzerland). IC50 values were deter-
mined with Graphpad Prism 4 using the formula (ODtreatedꢀODstart)/
(ODcontrolꢀODstart)ꢃ100. The IC50 is the drug concentration for which the
total cell number per well corresponds to 50% of the cell number in un-
treated control cultures (100%) at the end of the incubation period.
Data shown in Table 1 represent the mean of three independent experi-
ments.
desTHP-(ꢀ)-dactylolide (59): To
a solution of alcohol 58 (40.7 mg,
0.122 mmol, 1.0 equiv) in CH2Cl2 (1.5 mL) was added solid DMP
(156 mg, 0.37 mmol, 3.0 equiv; added in two equal portions, with the
second portion added after 10 min). The mixture was stirred for a total of
30 min at room temperature. Sat aq NaHCO3 (5 mL) and sat aq Na2S2O3
(5 mL) were then added and stirring was continued for 15 min, leading to
a clear organic phase and a turbid aqueous phase. The phases were sepa-
rated and the aqueous phase was extracted with CH2Cl2 (3ꢃ5 mL). The
combined organic extracts were washed with sat aq NaHCO3 (2ꢃ5 mL),
dried over MgSO4, and concentrated under reduced pressure. Purification
of the residue by FC (EtOAc/Hex 1:2!1:1) gave alcohol 59 (30.5 mg,
0.092 mmol, 75%) as a pale-yellow semisolid. Rf =0.37 (EtOAc/Hex 1:1,
UV, CPS); [a]2D4 =ꢀ50.498 (c
=
0.44, CHCl3); 1H NMR (400 MHz,
Experimental details for all other compounds can be found in the Sup-
porting Information, including synthetic procedures, full analytical data,
CDCl3): d=9.65 (s, 1H), 7.66 (dd, J=15.2, 11.6; 1H), 6.86 (dt, J=16.2,
6.7; 1H), 6.19–6.15 (m, 1H), 6.10 (dt, J=16.2, 1.5; 1H), 5.94 (d, J=15.2;
1H), 5.43–5.37 (m, 1H), 5.33–5.30 (m, 1H), 4.01 (dd, J=12.0, 7.8; 1H),
3.93–3.89 (m, 1H), 3.58–3.44 (m, 4H), 2.61–2.58 (m, 1H), 2.47–2.37 (m,
3H), 1.89 (s, 3H), 1.71 ppm (s, 3H); 13C NMR (100 MHz, CDCl3): d=
199.0, 196.6, 166.3, 146.8, 143.8, 140.5, 133.1, 130.5, 126.0, 125.9, 120.0,
75.8, 67.9, 67.5, 45.7, 39.2, 32.9, 24.5, 16.4 ppm; IR (thin film): n˜ =3424,
2957, 2921, 2853, 1732, 1706, 1668, 1632, 1456, 1437, 1377, 1356, 1317,
1258, 1206, 1174, 1143, 1112, 1080, 1026, 976, 888, 800 cmꢀ1; HRMS
(ESI): m/z: calcd for C19H24NaO5 [M+Na+]: 355.1516; found: 355.1523.
1
and H/13C NMR spectra.
Acknowledgements
We thank the Roche Research Foundation for a PhD fellowship to D.Z.
We are indebted to Dr. Bernhard Pfeiffer and Fabienne Gaugaz (ETHZ)
for NMR support, to Kurt Hauenstein (ETHZ) for technical advice and
to Louis Bertschi from the ETHZ-LOC MS-Service for HRMS spectra
acquisition. This work was conducted within the framework of COST
Action CM0804 “Chemical Biology with Natural Product”.
desTHP-zampanolide (60): To a solution of amide 42 (6.0 mg, 54.2 mmol,
6.0 equiv; co-evaporated with 0.5 mL toluene immediately before use) in
THF (0.5 mL) was added dropwise DIBAL-H (1.2m in toluene, 38 mL,
45.1 mmol, 5.0 equiv) at 08C under argon. The colorless solution was stir-
red at 08C for 30 min before a solution of 59 (3.0 mg, 9.0 mmol, 1.0 equiv;
vial flushed with argon) in THF (0.3 mL) was added dropwise. After
30 min the cooling bath was removed and the reaction was stirred at RT
overnight. Sat aq Rochelle salt (3 mL) and EtOAc (5 mL) were then
added. The phases were separated and the aqueous layer was extracted
with EtOAc (3ꢃ5 mL). The combined organic extracts were dried over
MgSO4, filtered and concentrated under reduced pressure. The crude
product was filtered through a short plug of silica (hexane/EtOAc 2:3,
1% NEt3 v/v) and then further purified by preparative HPLC to give 60
(1.1 mg, 28%) as a ca. 1.6:1 mixture of C18 isomers. Preparative HPLC
was carried out on a Gilson system equipped with a photodiode array de-
tector and employing a Waters SymmetryPrepTM C-18 column (5 mm,
19ꢃ100 mm). l=215 nm and 230 nm. Flow rate 25 mLminꢀ1. Eluent: 0–
2 min: 10% acetonitrile in water, 2–17 min: gradient from 10% to 100%
acetonitrile; tR = 9.3–9.9 min). Rf =0.24 (EtOAc/Hex 1:1, UV, CPS);
1H NMR (500 MHz, [D6]DMSO): d=8.40 (d, J=9.0; 1H (isomer 1)),
8.29 (d, J=9.2; 1H (isomer2)), 7.58–7.52 (m, 1H), 7.48–7.43 (m, 1H),
6.83–6.73 (m, 1H), 6.42–6.35 (m, 1H), 6.25–6.18 (m, 1H+1H (isomer
1)), 6.14 (d, J=5.2; 1H (isomer 2)), 6.05–5.93 (m, 2H+1H (isomer 1)),
5.92 (d, J=15.0; 1H (isomer 2)), 5.67–5.64 (m, 1H), 5.39–5.33 (m, 1H),
5.23–5.18 (m, 1H), 5.08–5.04 (m, 1H (isomer 1)), 5.00–4.96 (m, 1H
(isomer 2)), 3.97–3.92 (m, 1H), 3.90–3.87 (m, 1H (isomer 1)), 3.83–3.77
(m, 1H+1H (isomer 2)), 3.42–3.29 (m, 2H), 3.23 (d, J=12.9; 1H (isomer
2)), 3.12 (d, J=12.7; 1H (isomer 1)), 2.42–2.32 (m, 3H), 2.22–2.16 (m,
1H), 1.80–1.76 (m, 6H), 1.63 (s, 3H (isomer 2)), 1.62 ppm (s, 3H (isomer
1)); 13C NMR (125 MHz, [D6]DMSO): d=196.4, 196.4, 165.8, 165.6,
165.4, 165.2, 146.5, 146.4, 142.4, 142.0, 140.9, 140.8, 139.8, 139.4, 137.5,
137.3, 134.4, 134.1, 130.0, 129.8, 128.6, 125.6, 125.6, 124.6, 124.5, 121.2,
120.7, 119.1, 119.1, 72.9, 72.9, 72.2, 71.8, 67.1, 67.0, 66.7, 66.6, 45.4, 45.1,
32.3, 32.2, 24.0, 23.9, 18.4, 16.7, 16.4 ppm (one set of signals is hidden un-
derneath the DMSO signal); IR (thin film): n˜ =3326, 2960, 2854, 1691,
1661, 1633, 1604, 1518, 1434, 1358, 1278, 1259, 1208, 1147, 1085, 1047,
1037, 1019, 977, 928, 835, 828, 801 cmꢀ1; HRMS (ESI): m/z: calcd for
C25H34NO6 [M+H+]: 444.2381; found: 444.2380.
[1] a) J. W. Blunt, B. R. Copp, W. P. Hu, M. H. G. Munro, P. T. North-
cote, M. R. Prinsep, Nat. Prod. Rep. 2009, 26, 170–244; b) J. W.
Blunt, B. R. Copp, W. P. Hu, M. H. G. Munro, P. T. Northcote, M. R.
Prinsep, Nat. Prod. Rep. 2007, 24, 31–86.
[2] a) C. Nastrucci, A. Cesario, P. Russo, Recent Pat. Anti-Cancer Drug
Discovery 2012, 7, 218–232; b) N. L. Daly, D. J. Craik, Drugs Future
2011, 36, 25–32. If one includes the antibody drug conjugate (ADC)
brentuximab vedotin, which incorporates a synthetic analogue of the
marine natural product family of dolastatins (i.e., monomethyl auri-
statin E), the number is actually four: c) A. P. Z. Skarbnik, M. R.
Smith, Expert Opin. Biol. Ther. 2012, 12, 633–639.
[3] For a review on microtubule-stabilizing natural products see: K.-H.
Altmann, J. Gertsch, Nat. Prod. Rep. 2007, 24, 327–357.
[4] J. J. Field, A. J. Singh, A. Kanakkanthara, T. Halafihi, P. T. North-
cote, J. H. Miller, J. Med. Chem. 2009, 52, 7328–7332.
[5] J.-i. Tanaka, T. Higa, Tetrahedron Lett. 1996, 37, 5535–5538.
[6] J. J. Field, B. Pera, E. Calvo, A. Canales, D. Zurwerra, C. Trigili, J.
Rodrꢅguez-Salarichs, R. Matesanz, A. Kanakkanthara, S. J. Wake-
field, A. J. Singh, J. Jimꢆnez-Barbero, P. Northcote, J. H. Miller, J. A.
Lꢇpez, E. Hamel, I. Barasoain, K.-H. Altmann, J. F. Dꢅaz, Chem.
Biol. 2012, 19, 686–698.
[7] a) A. B. Smith III, I. G. Safonov, R. M. Corbett, J. Am. Chem. Soc.
2001, 123, 12426–12427; b) A. B. Smith III, I. G. Safonov, R. M. Cor-
bett, J. Am. Chem. Soc. 2002, 124, 11102–11113; c) A. B. Smith III,
I. G. Safonov, Org. Lett. 2002, 4, 635–637.
[8] T. R. Hoye, M. Hu, J. Am. Chem. Soc. 2003, 125, 9576–9577.
[9] a) R. Traber, C. Keller-Juslꢆn, H.-R. Loosli, M. Kuhn, A. Von Wart-
burg, Helv. Chim. Acta 1979, 62, 1252–1267; b) T. Takeuchi, H.
Iinuma, S. Kunimoto, T. Masada, M. Ishizuka, M. Takeuchi, M.
Hamada, H. Naganawa, S. Kondo, H. Umezawa, J. Antibiot. 1981,
34, 1619–1621; c) H. Umezawa, S. Kondo, H. Iinuma, S. Kunimoto,
Y. Ikeda, H. Iwazawa, D. Ikeda, T. Takeuchi, J. Antibiot. 1981, 34,
Chem. Eur. J. 2012, 18, 16868 – 16883
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16881

K.-H. Altmann et al.
1622–1624; d) N. B. Perry, J. W. Blunt, M. H. G. Munro, L. K. Pan-
nell, J. Am. Chem. Soc. 1988, 110, 4850–4851; e) N. B. Perry, J. W.
Blunt, M. H. G. Munro, A. M. Thompson, J. Org. Chem. 1990, 55,
223–227; f) J. I. Jimꢆnez, G. Goetz, C. M. S. Mau, W. Y. Yoshida,
P. J. Scheuer, R. T. Williamson, M. Kelly, J. Org. Chem. 2000, 65,
8465–8469.
[34] J. P. Vitale, S. A. Wolckenhauer, N. M. Do, S. D. Rychnovsky, Org.
Lett. 2005, 7, 3255–3258.
[35] X. Liang, A. Lohse, M. Bols, J. Org. Chem. 2000, 65, 7432–7437.
[36] P. Kocovsky, J. Org. Chem. 1988, 53, 5816–5819.
[37] T. L. Macdonald, N. Narasimhan, J. Org. Chem. 1985, 50, 5000–
5001.
[10] A. Cutignano, I. Bruno, G. Bifulco, A. Casapullo, C. Debitus, L.
Gomez-Paloma, R. Riccio, Eur. J. Org. Chem. 2001, 775–778.
[11] J. Uenishi, T. Iwamoto, J. Tanaka, Org. Lett. 2009, 11, 3262–3265.
[12] a) F. Ding, M. P. Jennings, Org. Lett. 2005, 7, 2321–2324; b) F. Ding,
M. P. Jennings, J. Org. Chem. 2008, 73, 5965–5976.
[13] a) A. Ghosh, X. Cheng, R. Bai, E. Hamel, Eur. J. Org. Chem. 2012,
4130–4139; b) A. Ghosh, X. Cheng, Org. Lett. 2011, 13, 4108–4111.
[14] D. L. Aubele, S. Wan, P. E. Floreancig, Angew. Chem. 2005, 117,
3551–3554; Angew. Chem. Int. Ed. 2005, 44, 3485–3488.
[15] C. C. Sanchez, G. E. Keck, Org. Lett. 2005, 7, 3053–3056.
[16] I. Louis, N. L. Hungerford, E. J. Humphries, M. D. McLeod, Org.
Lett. 2006, 8, 1117–1120.
[38] L. G. Menchikov, A. V. Vorogushin, O. S. Korneva, O. M. Nefedov,
Mendeleev Commun. 1995, 5, 223–224.
[39] a) J.-F. Betzer, J. Ardisson, J.-Y. Lallemand, A. Pancrazi, Tetrahe-
dron Lett. 1997, 38, 2279–2282; b) J.-F. Betzer, F. Delaloge, B.
Muller, A. Pancrazi, J. Prunet, J. Org. Chem. 1997, 62, 7768–7780.
[40] J. Schwartz, J. A. Labinger, Angew. Chem. 1976, 88, 402–409;
Angew. Chem. Int. Ed. Engl. 1976, 15, 333–340.
[41] Uenishi et al. (ref. [11]) have reported the opening of 19 with vinyl
lithium 28 (derived from the corresponding stannane by Sn–Li ex-
change with nBuLi) in THF in 63% yield. In preliminary experi-
ments we have not been able to reproduce this result.
[17] S. Y. Yun, E. C. Hansen, I. Volchkov, E. J. Cho, W. Y. Lo, D. Lee,
Angew. Chem. 2010, 122, 4357–4359; Angew. Chem. Int. Ed. 2010,
49, 4261–4263.
[18] a) D. M. Troast, J. A. Porco, Jr., Org. Lett. 2002, 4, 991–994;
b) D. M. Troast, J. Yuan, J. A. Porco, Adv. Synth. Catal. 2008, 350,
1701–1711; c) M. R. Wilson, R. E. Taylor, Org. Lett. 2012, 14, 3408–
3411.
[19] A notable exception is the work by Uenishi et al. (ref. [11]), who
have shown the C20 epimer of 1 to be about 10-fold less active than
1. See also ref. [13a].
[42] V. I. Tararov, G. Kçnig, A. Bçrner, Adv. Synth. Catal. 2006, 348,
2633–2644.
[43] a) S. Saito, T. Hasegawa, M. Inaba, R. Nishida, T. Fujii, S. Nomizu,
T. Moriwake, Chem. Lett. 1984, 1389–1392; b) S. Saito, T. Ishikawa,
A. Kuroda, K. Koga, T. Moriwake, Tetrahedron 1992, 48, 4067–
4086.
[44] E. J. Corey, P. L. Fuchs, Tetrahedron Lett. 1972, 13, 3769.
[45] a) K. C. Nicolaou, A. L. Nold, R. R. Milburn, C. S. Schindler,
Angew. Chem. 2006, 118, 6677–6682; Angew. Chem. Int. Ed. 2006,
45, 6527–6532; b) K. C. Nicolaou, A. L. Nold, R. R. Milburn, C. S.
Schindler, K. P. Cole, J. Yamaguchi, J. Am. Chem. Soc. 2007, 129,
1760–1768.
[20] D. Zurwerra, J. Gertsch, K.-H. Altmann, Org. Lett. 2010, 12, 2302–
2305.
[21] For examples see: a) K. C. Nicolaou, S. P. Seitz, M. R. Pavia, J. Am.
Chem. Soc. 1982, 104, 2030–2031; b) I. Kadota, Y. Hu, G. K. Pack-
ard, S. D. Rychnovsky, Proc. Natl. Acad. Sci. USA 2004, 101, 11192–
11195; c) G. O. Berger, M. O. Tius, J. Org. Chem. 2007, 72, 6473–
6480.
[22] E. J. Corey, J. A. Katzenellenbogen, G. H. Posner, J. Am. Chem.
Soc. 1967, 89, 4245–4247.
[23] A. Chau, J.-F. Paquin, M. Lautens, J. Org. Chem. 2006, 71, 1924–
1933.
[24] The investigation of toluene as the reaction medium was based on
prior work by Okano et al. in the context of their total synthesis of
(+)-yatakemycin: K. Okano, H. Tokuyama, T. Fukuyama, J. Am.
Chem. Soc. 2006, 128, 7136–7137.
[25] a) A. J. Barlow, B. J. Compton, R. T. Weavers, J. Org. Chem. 2005,
70, 2470–2475; b) L.-W. Hsin, C. M. Dersch, M. H. Baumann, D.
Stafford, J. R. Glowa, R. B. Rothman, A. E. Jacobson, K. C. Rice, J.
Med. Chem. 2002, 45, 1321–1329.
[46] B. Ganem, Y. Dong, Y. F. Zheng, G. D. Prestwich, J. Org. Chem.
1999, 64, 5441–5446.
[47] A. Hafner, R. O. Duthaler, R. Marti, G. Rihs, P. Rothe-Streit, F.
Schwarzenbach, J. Am. Chem. Soc. 1992, 114, 2321–2336.
[48] a) H. C. Brown, P. K. Jadhav, J. Am. Chem. Soc. 1983, 105, 2092–
2093; b) U. S. Racherla, H. C. Brown, J. Org. Chem. 1991, 56, 401–
404; c) H. C. Brown, U. S. Racherla, Y. Liao, V. V. Khanna, J. Org.
Chem. 1992, 57, 6608–6614.
[49] a) G. E. Keck, K. H. Tarbet, L. S. Geraci, J. Am. Chem. Soc. 1993,
115, 8467–8468; b) G. E. Keck, D. S. Welch, P. K. Vivian, Org. Lett.
2006, 8, 3667–3670.
[50] a) S. R. Crosby, J. R. Harding, C. D. King, G. D. Parker, C. L. Willis,
Org. Lett. 2002, 4, 577–580. For a more detailed discussion of side
reactions in Prins cyclizations see: b) R. Jasti, S. D. Rychnovsky, J.
Am. Chem. Soc. 2006, 128, 13640–13648.
[51] A. S. Kende, K. Liu, I. Kaldor, G. Dorey, K. Koch, J. Am. Chem.
Soc. 1995, 117, 8258–8270.
[26] a) S. Racha, Z. Li, H. El-Subbagh, E. Abushanab, Tetrahedron Lett.
1992, 33, 5491–5494; b) Z. Li, S. Racha, L. Dan, H. El-Subbagh, E.
Abushanab, J. Org. Chem. 1993, 58, 5779–5783.
[27] J. A. Frick, J. B. Klassen, A. Bathe, J. M. Abramson, H. Rapoport,
Synthesis 1992, 621–623.
[52] K. Miyashita, M. Ikejiri, H. Kawasaki, S. Maemura, T. Imanishi, J.
Am. Chem. Soc. 2003, 125, 8238–8243.
[28] D. L. J. Clive, K. S. K. Murthy, A. G. H. Wee, J. S. Prasad, G. V. J. Da
Silva, M. Majewski, P. C. Anderson, C. F. Evans, R. D. Haugen, J.
Am. Chem. Soc. 1990, 112, 3018–3028.
[29] B. M. Trost, M. T. Rudd, J. Am. Chem. Soc. 2005, 127, 4763–4776.
[30] R. Jasti, J. Vitale, S. D. Rychnovsky, J. Am. Chem. Soc. 2004, 126,
9904–9905.
[31] S. Marumoto, J. J. Jaber, J. P. Vitale, S. D. Rychnovsky, Org. Lett.
2002, 4, 3919–3922. The relative configuration at positions 2 and 6
of the THP ring in 23/23a was established by NMR spectroscopy
after conversion of the mixture to the corresponding mixture of 4-
keto derivatives through base-catalyzed cleavage of the trifluoroace-
tate followed by oxidation with DMP. For details see Supplementary
Information.
[32] H. J. Schneider, V. Hoppen, J. Org. Chem. 1978, 43, 3866–3873.
[33] E. L. Eliel, S. H. Wilen, Stereochemistry of Organic Compounds,
Wiley, New York, 1994, p. 695.
[53] J. Inanaga, K. Hirata, H. Saeki, T. Katsuki, M. Yamaguchi, Bull.
Chem. Soc. Jpn. 1979, 52, 1989–1993.
[54] a) I. Paterson, K. S. Yeung, J. B. Smaill, Synlett 1993, 774–776; b) I.
Paterson, K.-S. Yeung, Tetrahedron Lett. 1993, 34, 5347–5350; c) I.
Paterson, K.-S. Yeung, C. Watson, R. A. Ward, P. A. Wallace, Tetra-
hedron 1998, 54, 11935–11954; d) J. Barrios, J. M. Marinas, J. V. Sin-
isterra, Bull. Soc. Chim. Belg. 1986, 95, 107–117.
[55] D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155–4156.
[56] A. Bayer, M. E. Maier, Tetrahedron 2004, 60, 6665–6677.
[57] This yield was obtained on a 0.072 mmol scale of 2 (27.6 mg). Yields
of up to 83% were obtained on smaller scales.
[58] A. Prota, manuscript in preparation.
[59] a) B. O. Lindgren, H. Nilsson, Acta Chem. Scand. 1973, 27, 888–890;
b) G. A. Kraus, B. Roth, J. Org. Chem. 1980, 45, 4825–4830; c) B. S.
Bal, W. E. Childers, H. W. Pinnick, Tetrahedron 1981, 37, 2091–
2096.
16882
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 16868 – 16883

Total Synthesis of (ꢀ)-Zampanolide
FULL PAPER
[60] For a recent example see: E. A. Ilardi, C. E. Stivala, A. Zakarian,
Org. Lett. 2008, 10, 1727–1730.
[61] The tentative 62 was not isolated and characterized. However, we
have isolated and characterized allene 63 that was formed upon at-
tempted alkylation of 52 with TBDPS-protected 3-bromo-1-propa-
nol.
[62] In a different experiment 60 was obtained in 45% yield and as a ca.
1:1 isomeric mixture after preparative HPLC purification. The dif-
ferences in yield and isomer ratio are likely due to the more conser-
vative pooling of product fractions in the purification step for the
lower yielding experiment. The latter material (ca. 1.6:1 mixture of
isomers) was used in the biological experiments.
Received: July 17, 2012
Published online: November 7, 2012
Chem. Eur. J. 2012, 18, 16868 – 16883
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16883