Evolution of a Polyene Cyclization Cascade for the Total Synthesis of (?)-Cyclosmenospongine

Article abstract of DOI:10.1002/chem.201605029

We report a full account on the development of a unique cationic polyene cyclization for the total synthesis of the tetracyclic meroterpenoid (?)-cyclosmenospongine. A highly convergent three-component coupling strategy enabled rapid access to individual cyclization precursors that were tested for their reactivity. The successful transformation generates three rings and sets four consecutive stereocenters in a single operation proceeding in a highly efficient manner to give exclusively the trans-decalin framework. In addition, we found that the enol ether geometry and the relative configuration of C3 and C8 are crucial for the success of the polyene cyclization.

Full text of DOI:10.1002/chem.201605029

A Journal of
Accepted Article
Title: Evolution of a Polyene Cyclization Cascade for the Total
Synthesis of (-)-Cyclosmenospongine
Authors: Klaus Speck and Thomas Magauer
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201605029
Link to VoR: http://dx.doi.org/10.1002/chem.201605029
Supported by

10.1002/chem.201605029
Chemistry - A European Journal
FULL PAPER
Evolution of a Polyene Cyclization Cascade for the Total
Synthesis of (–)-Cyclosmenospongine
Klaus Speck and Thomas Magauer*[a]
Abstract: We report a full account on the development of a unique
cationic polyene cyclization for the total synthesis of the tetracyclic
meroterpenoid (–)-cyclosmenospongine. A highly convergent three-
component coupling strategy enabled rapid access to individual
cyclization precursors that were tested for their reactivity. The
successful transformation generates three rings and sets four
consecutive stereocenters in a single operation proceeding in a
highly efficient manner to give exclusively the trans-decalin
framework. In addition, we found that the enol ether geometry and
the relative configuration of C3 and C8 are crucial for the success of
the polyene cyclization.
Introduction
Throughout history, mankind has benefitted from the rich
chemical diversity found in nature. While natural products
isolated from terrestrial plants have always been an integral part
of traditional medicine, marine organisms were long unexplored
as a potential source of bioactive molecules.[1] During the past
decades, numerous sesquiterpenoids have been isolated from
various marine sponges.[2] Within this vast class of natural
products,
a
structurally unique family of tetracyclic
meroterpenoids (1–4) recently drew our interest (Scheme 1a).
Aureol (1), the first and eponymous member of this family, was
isolated from the Caribbean marine sponge Smenospongia
aurea by the group of Faulkner in 1980.[3a] Since then, several
congeners of this class of tetracyclic meroterpenoids, including
strongylin A (2), smenoqualone (3) and cyclosmenospongine (4)
have been isolated from various marine sources.[3b–d]
Meroterpenoids are hybrid natural products with a mixed
biosynthetic pathway that incorporates polyketide and terpenoid
building blocks (Scheme 1b).[4] The fusion of farnesyl
pyrophosphate (C15), which originates from the terpene pathway,
and a polyketide derived aromatic core gives linear polyene 5,
which undergoes a cationic cyclization cascade to initially
produce the key biosynthetic intermediate, drimane 6.[5]
Consecutive [1,2]-hydride and methyl shifts provide cation 7,
which can be either trapped by the adjacent phenol to directly
yield the cis-fused decalin ring system 9a (-5-H) of aureol (1)
and its congeners (2 and 3) or eliminate to give 8. The latter is
considered to be the direct biosynthetic precursor of the trans-
fused decalin system 9b (-5-H), which is present in
cyclosmenospongine (4).[6,7]
Scheme 1. a) Representative members of tetracyclic meroterpenoid natural
products and b) their proposed biosynthesis (adapted from George and
Faulkner).[3a,9c]
The so-formed tetracycle comprises four consecutive
stereocenters, two of which are tetrasubstituted. The aromatic
subunit, which can undergo further modifications is fused to
either a cis- (1–3) or a trans-decalin (4) ring system. Although
structurally closely related,
a broad range of biological
activities,[2] including antiviral,[3b,8a] anti-cancer[4d,8b–h] and anti-
bacterial,[8i] was reported for these natural products.[8i]
Owing to their intriguing structural features and a unique
biological profile, several groups have made efforts to develop
syntheses of 1–4. As shown in Scheme 2, previous studies
towards individual members hinged on the cationic cyclization of
biosynthetic-like precursors arenarol (10) or 11.[9a] In the seminal
work of Capon and van der Helm, it was shown that exposure of
10 or 11 to Lewis or Brønsted acids promotes cyclization to
aureol (1) and 5-epi-aureol (38).[1a,6] By simple variation of the
reaction conditions, either the kinetically favoured cis-decalin
framework (BF3·OEt2, dichloromethane, T < –10 °C) or the
thermodynamically more stable trans-decalin system (pTsOH,
benzene, 80 °C) were selectively obtained. While the groups of
[a]
Ms.Sc. K. Speck, Dr. T. Magauer
Department of Chemistry and Pharmacy
Ludwig Maximillians University Munich
Butenandtstrasse 5–13, 81377 Munich (Germany)
E-mail: thomas.magauer@lmu.de
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605029
Chemistry - A European Journal
FULL PAPER
Marcos[9b] and George[9c] developed chiral pool strategies to
synthesize 11, the synthesis of arenarol (10) described by the
group of Katoh, uses a Wieland–Miescher ketone derivative.[9d–g]
cyclization of a model substrate to access the tetracyclic core of
the aureol family of natural products.[9i]
As previous work has shown that formation of the tetracyclic ring
system via biomimetic polyene cyclization and Wagner–
Meerwein shifts can only occur in a stepwise manner in the
laboratory,[9h] we were searching for a strategy to mimic the
highly organized environment of an enzymatic active site more
efficiently. In order to overcome the current inability of synthetic
methodology to mimic these two independently occurring events
in a single flask, we recently investigated the cyclization of a
well-elaborated polyene.[10] Herein we report the evolution of our
idea to assemble the tetracyclic scaffold within one step through
an unprecedented polyene cyclization cascade.
Guided by the biosynthesis and inspired by the three-
dimensional drawing of aureol (1), we identified the
retrosynthetic bond disconnections as shown in Scheme 3a.
Scission of the bonds highlighted in red (C4-C5, C9-C10 and
C15-C16) revealed the highly simplified cyclization precursor 12,
which was envisioned to undergo the cationic polyene
cyclization upon Lewis acid activation of the epoxide. This
remarkable transformation would generate three rings and three
consecutive stereocenters, two of which are tetrasubstituted, in
one step. Further dissection would give three building blocks of
equal complexity, which could be united by convergent fragment
coupling of phenol 13, bromoalkyne 14 and iodide 15.
Scheme 2. Key intermediates of previous syntheses.
In contrast to these approaches, where the crucial
stereochemistry of the decalin ring system is set in a stepwise
manner, Rosales applied
a very elegant Ti(III)-mediated
reductive epoxide cyclization cascade to construct 11.[9h] In
addition, Cramer reported a ruthenium(III)-catalyzed domino
f
Scheme 3. a) Retrosynthetic analysis of aureol (1) and b) hypothetical transition states for the cyclization of 12.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605029
Chemistry - A European Journal
FULL PAPER
This modular retrosynthetic approach would allow us to rapidly
access a library of both natural and fully synthetic analogues by
simple variation of the arene moiety. From a structural point of
view, the synthetically most challenging task is the selective
formation of the cis-decalin. At this point, we were uncertain
about the influence of the enol-ether geometry on the folding
and cyclization of 15 to either give the desired cis-decalin via
transition state 16a or the trans-fused decalin ring system
resulting from the alternative folding depicted for 16b (Scheme
3b). At the outset, we hypothesized that the cyclization would
proceed in a highly concerted fashion via a chair-like transition
state that forces the large aryl ether substituent into the
energetically favored pseudo-equatorial position. This would
exclusively produce cis-decalin framework 18, which is
conserved among aureol (1) and several other members of this
meroterpenoid family.
Epoxide 20 was synthesized in three steps from 1-bromo-1,5-
enyne 14.[12] Since the previously reported protocol for a base-
mediated nucleophilic addition of phenols to 1-bromoalkynes
proved to be low yielding and was plagued by side-product
formation, we were forced to further optimize the reaction
conditions.[13] After extensive screening we found that the
addition of 4-methoxyphenol (13) to alkyne 14 is best carried out
at 70 °C using excess phenol (10 equiv) and cesium carbonate
as base (3 equiv) in N,N-dimethylformamide (1 M). Epoxide 20
was then synthesized through
a
Sharpless asymmetric
dihydroxylation using AD-mix- and selective mesylation of the
secondary alcohol followed by intramolecular nucleophilic
displacement. The envisioned palladium-mediated sp2-sp3
cross-coupling of the key building blocks 15 and 20 to give the
cyclization precursor 12 proved to be more challenging than
anticipated and considerable experimentation was required to
establish a mild and efficient protocol that also tolerated the
epoxide functionality. As initial attempts to realize the coupling
between 15 and 20 using a Negishi cross-coupling were low
yielding (~5%), we decided to investigate B-alkyl Suzuki–
Miyaura coupling conditions. We chose to convert iodide 15 to
its corresponding borinate using the “B-OMe-9-BBN” variation.[14]
The borinate can be conveniently prepared by addition of tert-
butyl lithium to a premixed solution of 15 and B-OMe-9-BBN
(1 M in hexanes) in tetrahydrofuran at –78 °C.[15] We found that
the use of Buchwald´s Pd SPhos precatalyst system (5 mol%
SPhos Pd G2, 5 mol% SPhos, Cs2CO3, THF, DMF, H2O, 40 °C)
was crucial to obtain high yields, short reaction times (typically
~1 h) and to efficiently suppress unwanted -hydride
elimination.[16] The generality of these optimized conditions was
also demonstrated for a small array of bromoenol ethers.
Electron-rich and electron-poor bromoenol ethers could be
efficiently coupled with n-butyl lithium, an easily available alkyl
surrogate, in excellent yields (see Supporting Information for
further details).
Results and Discussion
We began our synthetic endeavor with the multi-gram synthesis
of iodide 15 and bromoenol ether 20 (Scheme 4a). Iodide 15
was prepared starting from the literature known ketone 19.[11]
Wittig olefination of 19, followed by cleavage of the silyl ether
and exposure of the alcohol to Appel´s conditions gave 15 in
excellent yield on gram-scale (1.6 g).
With our key intermediate 12 in hand, the stage was set to
investigate the intended cationic polyene cyclization (Scheme
4b).
Addition
of
either
tris(pentafluorophenyl)borane
(B(C6F5)3),[17] or diethylaluminum chloride (Et2AlCl) to a solution
of 12 in dichloromethane at –78 °C effected selective activation
of the epoxide and carbon-carbon bond formation between C4
and C5. Unfortunately, the putative oxonium ion 16 (compare
Scheme 3b) did not undergo further cyclization but was trapped
by the alkoxide to give acetal 21. All attempts to promote further
ring closure of 21 by means of Lewis acid activation resulted in
either hydrolysis of the acetal to the ketone 22 or decomposition
of the starting material. From this result, we concluded that
either steric hindrance around the 1,1-disubstituted olefin or a
stepwise mechanism, involving nucleophilic attack of the remote
C9-C15 olefin at C10 to generate an energetically unfavorable
primary carbocation at C15, prevent acetal 21 from further
cyclization. In order to account for these issues and to validate
our hypothesis, we decided to modify the C8-C15 substitution
pattern of 12 by synthesizing the cyclization precursors depicted
in Scheme 5a according to the chemistry developed before (see
Supporting Information for further details).
Scheme 4. a) Synthesis of cyclization precursor 12 and b) attempted
cyclization of polyene 12. Reagents and conditions: a) PPh3CH3Br, KOtBu,
THF, 0 °C to 23 °C, 92%; b) NEt3∙3 HF, CH3CN, 23 °C, 95%; c) I2, PPh3, imH,
CH2Cl2, 0 °C, 81%; d) 4-methoxyphenol (13) (10 equiv), Cs2CO3, DMF, 70 °C,
56%; e) AD-mix-, tBuOH, H2O, 0 °C to 23 °C, 88%; f) MsCl, NEt3, CH2Cl2,
0 °C to 23 °C, then K2CO3, MeOH, 23 °C, 81%; g) 15, tBuLi, B-OMe-9-BBN,
THF, –78 °C to 23 °C; SPhos Pd G2 (5 mol%), SPhos (5 mol%), Cs2CO3,
DMF/H2O (9:1), 40 °C, 72%; h) LA = Et2AlCl, CH2Cl2, –78 °C, 81% of 21; LA =
B(C6F5)3, CH2Cl2, –78 °C, 54% of 22; B-OMe-9-BBN
= 9-methoxy-9-
borabicyclo[3.3.1]nonane, imH = imidazole, MsCl = methanesulfonyl chloride,
LA = Lewis acid.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605029
Chemistry - A European Journal
FULL PAPER
addition to steric hindrance might impede the cyclization. In
order to overcome the lack of cation stabilization without
sacrificing the required methyl substituents, we equipped our
substrate with a thioenol ether to give 24 (E:Z = 1:1.2).[18,19]
Since this measure was ineffective as well and, in analogy to the
activation of 12 and 23, no fully cyclized product could be
observed, we decided to further simplify our system. The first
breakthrough resulted from treating a solution of 25, which lacks
the C9 methyl substituent, in dichloromethane with
ethylaluminum dichloride at –78 °C. This allowed us to identify a
product whose 2D NMR spectra were in full agreement with a
tetracyclic structure (see Supporting Information for details).
Although, we were unable to increase the low yield (8%) for this
substrate, thioenol ether 26 finally emerged as a substrate that
underwent the intended cyclization in a highly efficient (91%)
and stereoselective manner to afford trans-decalin 28 as a
single product (Scheme 5b).[20] We believe that the intermediacy
of acetal 27 accounts for the selective trans-decalin formation by
blocking the top face of the molecule, thereby preventing a
pseudo-equatorial alignment of the aryl ether substituent
(compare Scheme 3b). The observed equilibration of the
phenylthio substituent provides evidence for a non-concerted,
stepwise mechanism for the formation of the C9-C10 (decalin)
and C15-C16 (tetrahydropyran) single bond.
Having developed a gram-scale synthesis of the full carbon
framework, we turned our attention to the selective
functionalization of 28. For the installation of the vicinal-cis-
dimethyl groups, we envisioned to convert sulfide 28 into enone
30. With enone 30 in hand, we hoped to install both methyl
substituents either through a sequence involving 1,4-addition of
a methyl cuprate followed by -methylation or a 1,3-dipolar
cycloaddition using a thiocarbonyl ylide.[21]
Scheme 5. a) Selectively modified cyclization precursors and b) successful
cyclization of 26. Reagents and conditions: a) EtAlCl2, CH2Cl2, –78 °C, 91%.
In a first attempt to rule out any possible steric hindrance around
the olefin, including the C8 stereocenter, we investigated
compound 23. Exposure of 23 to ethylaluminumdichloride led to
an inseparable mixture of several products, thus validating our
hypothesis that low nucleophilicity of the C9-C15 double bond in
Scheme 6. Formation of enone 30 via a) attempted Saegusa–Ito oxidation and b) styrene 30. Reagents and conditions: a) mCPBA (2.1 equiv), CH2Cl2, 0 °C,
70%; b) LiHMDS, MoOPH, THF, 0 °C to 23 °C, 74%; c) NEt3, TBSOTf, CH2Cl2, –78 °C to 23 °C, 96%; (d) mCPBA (0.95 equiv), CH2Cl2, 0 °C; e) NEt3, TBSOTf,
CH2Cl2, –78 °C to 23 °C, then CHCl3, 105 °C, 59% over 3 steps; g) K2OsO4, NMO, acetone/H2O, 0 °C to 23 °C, 69 °C%; h), NaH, THF, 0 °C to 23 °C, then CS2,
then MeI, 87%; i) sulfolane, 200 °C, 93%. mCPBA = meta-chloroperbenzoic acid, HMPA = hexamethylphosphoramide, MoOPH = MoO5•pyridine•HMPA, NMO =
N-methylmorpholin-N-oxide.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605029
Chemistry - A European Journal
FULL PAPER
For the elaboration of 30, we investigated two parallel routes.
The first approach was based on the conversion of sulfide 28 to
ketone 29 followed by a Saegusa–Ito oxidation (Scheme 6a).[22]
The second strategy was based on an elimination/oxidation
route via styrene 31 (Scheme 6b). Oxidation of sulfide 28 using
excess m-chloroperbenzoic acid (mCPBA, 2.1 equiv) gave the
corresponding sulfone in excellent yield. Modification of a
recently reported procedure for oxidative desulfonylation using
Vedej´s reagent (MoOPH) allowed us to access the
corresponding benzylic ketone.[23] Interestingly, under the
reported conditions (nBuLi (1.05 equiv), MoOPH (2.00 equiv),
–78 °C) the intermediate hydroxy sulfone did not collapse to give
the ketone but was left unreacted. We found that only the use of
excess lithium bis(trimethylsilyl)amide (LiHMDS) and prolonged
stirring of the reaction mixture at 23 °C effected complete
elimination. Protection of the secondary alcohol as its silyl ether
gave 29, which provided crystals suitable for X-ray diffraction.
This allowed us to unambiguously validate the relative
stereochemistry of this key tetracyclic intermediate.
Unfortunately, all attempts to further oxidize ketone 29 to enone
30 were unsuccessful. In most experiments, ketone 29 proved to
be unreactive and formation of a silyl enol ether, as required for
the Saegusa–Ito oxidation, could never be observed.[24] Using
more forcing conditions to effect -deprotonation (LiHMDS, THF,
70 °C) led to -elimination of the phenol substituent.
conditions recently reported by Bräse enabled a successful one-
pot oxidation to give an -hydroxy ketone.[25] For the elimination
of the tertiary alcohol, we examined a number of different
conditions (e.g. pTsOH, Martin´s sulfurane, Burgess reagent,
thionyl chloride).[26] We found that only conversion to its xanthate
32 followed by thermal elimination reproducibly afforded 30 in
good overall yield (81%) on large scale.[27]
Having succeeded in forming the enone, we investigated the
introduction of the vicinal methyl substituents. Although
conjugate addition of lithium dimethylcuprate (Me2CuLi) to enone
30 occurred with high diastereoselectivity, all attempts to trap
the
intermediate
enolate
were
unsuccessful.
The
stereochemistry of 33 could be confirmed by silyl ether cleavage
and single-crystal X-ray diffraction of alcohol 34. Due to the
difficulties encountered in the two-step methylation procedure,
we searched for alternative alkylation methods that would both
avoid possible -elimination of the phenoxide and allow for cis-
selectivity. For this purpose, we selected sulfoxide 35, which
was previously developed as an efficient source of thiocarbonyl
ylide for the [3+2] cycloaddition reaction with sterically
unhindered, electron-poor alkenes.[21a,28] Unfortunately, initial
attempts to accomplish the cycloaddition between enone 30 and
35 or modifications thereof failed under a variety of standard
conditions. Based on our ongoing efforts to establish high-
pressure as a powerful external stimulus to realize otherwise
infeasible transformations, we investigated the cycloaddition at
14 kbar.[29] Pleasingly, pressurizing a solution of 30 in a 1:1
mixture of dichloromethane and N,N’-dimethylpropylene urea
(DMPU) in the presence of excess sulfoxide 35 at 23 °C cleanly
afforded the tetrahydrothiophene 36 (68%) along with recovered
starting material 30 (16%).
Thus, we turned to an alternative strategy that involved
elimination of the phenyl thioether via its sulfoxide. Selective
oxidation of the phenyl thioether to the sulfoxide could be
accomplished using equimolar amounts of mCPBA. Protection
of the secondary alcohol as its silyl ether followed by thermally-
induced elimination (105 °C, CHCl3) furnished styrene 31 in
good yield. We were delighted to see that dihydroxylation
Scheme 7. Installation of the vicinal cis-dimethyl group via a) cuprate addition and attempted -methylation and b) high-pressure [3+2] cycloaddition. Reagents
and conditions: a) Me2CuLi, Et2O, 0 °C, 98%; b) NEt3∙3HF, CH3CN, 40 °C, 59%; c) 35, DMPU/CH2Cl2 (1:1), 23 °C, 14 kbar, 68% (16% of 30); d) Raney®-Ni, THF,
65 °C, 87%; e) LiAlH4, THF, 23 °C; then BF3∙OEt2, Et3SiH, 0 °C to 23 °C, 88%; f) (C6F5)OC(S)Cl, DMAP, CH2Cl2, 0 to 23 °C, then AIBN, nBu3SnH, benzene, 80 °C,
83%; e) BBr3, CH2Cl2, –78 °C to 23 °C, 86%. DMPU = N,N’-dimethylpropylene urea, AIBN = azobisisobutyronitrile, DMAP = 4-(dimethyl)aminopyridine.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605029
Chemistry - A European Journal
FULL PAPER
The best results were obtained when 35 was added in two
portions (2 × 5 equiv) over the course of four hours. In this
context, it is important to note that the purity of sulfoxide 35 was
crucial to observe high conversion. Traces of m-chlorobenzoic
acid that are formed during the generation of the sulfoxide have
a detrimental effect and inhibit the progress of the reaction. For
the completion of the synthesis, we first unmasked the vicinal
cis-dimethyl group with the aid of Raney®-nickel.[30] The ketone
function was then fully reduced upon exposure to lithium
aluminum hydride (LiAlH4) followed by treatment with
triethylsilane (Et3SiH) in the presence of boron trifluoride
etherate (BF3∙OEt2) to give 37.[31] Deoxygenation according to a
recently described protocol and cleavage of the methyl ether
with boron tribromide afforded (–)-5-epi-aureol (38), which has
yet to be isolated from natural sources.[6a, 9b]
At this stage, we focused on the isomerization of (–)-5-epi-aureol
(38) to (–)-aureol (1). Guided by previous results of Faulkner and
coworkers for the conversion of 1 to 11, we initially tried to adopt
the reported elimination conditions (Ac2O, BF3∙OEt2) for the
formation of 11 from 38 (Scheme 8).[3a] Surprisingly, elimination
of the phenol of the trans-fused decalin ring system of 5-epi-
aureol (38), which is perfectly oriented for an E2-elimination,
could not be promoted. Thus far, only monoacetylated 38 could
be isolated, which upon heating to 40 °C underwent Fries
rearrangement and simultaneous decomposition. The use of
trifluoroacetic anhydride in order to suppress the Fries
rearrangement and facilitate the elimination step was equally
ineffective.[32] Although we conducted an extensive screen of
Lewis (BF3∙OEt2, BCl3, AlCl3, B(C6F5)3, TiCl4, Er(OTf)3, CeCl3)[33]
and Brønsted acids (AcOH, TFA, HI, pTsOH, H2SO4,
H3PO4),[9b,34] an array of bases (KOtBu, DBU, pyridine)[35] with
and without additional electrophiles to trap the generate
phenolate (Ac2O, TFAA, MeI), and several oxidative protocols
(CAN, DDQ, NaIO4 supported on silica, PIDA, Fetizon´s
reagent),[32a,36] we were unable to observe formation of the
elimination product.
Scheme 8. Attempted elimination of 5-epi-aureol (38) to known intermediate
11.
Despite the success of promoting this remarkable cyclization
and having shown that tetracycle 28 can be advanced to 5-epi-
aureol (38), we faced serious challenges. Firstly, multiple steps
were necessary to install the missing vicinal cis-dimethyl
substitution pattern of C8/C9. Secondly, the overall synthesis
was not amenable to large-scale synthesis as the high-pressure
cycloaddition reaction was limited to a reaction volume of 8 ml
(~80 mg) per batch.
To address these limitations, we revised the design of our
cyclization precursor (Scheme 9). From the attempted
cyclization of 13 and 23–26, we learned that the cyclization
proceeds via the intermediacy of an acetal that is further
activated. The formation of this discrete species presumably
prevents any top-face approach and forces the vinyl sulfide to
adopt an alignment as depicted for 39 (Scheme 9a). From the
inspection of 39, we concluded that the axial aligned C8 methyl
group avoids any further cyclization to 40.
Based on these insights, we hypothesized that generation of the
fully-substituted cyclosmenospongine-like framework might be
feasible by using polyene 41. Simple inversion of the relative
configuration at C8 should minimize unwanted 1,3-diaxial
interactions with the aryl ether and therefore enable the attack of
the vinylsulfide to yield tetracycle 43 (Scheme 9b).
Scheme 9. Cyclization precursors and hypothetical transition states (bonds formed during the cyclization are highlighted in red). a) Formation of acetal
intermediate 39 from 24 (1,3-diaxial interactions are highlighted in blue). b) Formation of acetal 42 from 41 and further cyclization to tetracycle 43.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605029
Chemistry - A European Journal
FULL PAPER
To put this idea in practice, we had two options at hand. Either
we could exchange the configuration at C8 or adjust the
orientation of the epoxide at C3. For practical reasons, we
decided to alter the epoxide stereocenter. Simple exchange of
AD-mix- for AD-mix- allowed us to synthesize epoxide ent-41
(E:Z = 1:1.2 at C15) on gram scale (5.6 g) according to the
chemistry developed before.[10] Upon exposure of ent-41 to the
previously optimized cyclization conditions (EtAlCl2, CH2Cl2,
–78 °C), we were delighted to observe full and clean conversion
to the fully cyclized product ent-43. The occurance of an acetal
intermediate could be validiated through the use of less reactive
diethylaluminum chloride (Et2AlCl), which allowed us to
selectively discontinue the reaction cascade on the acetal level
and isolate ent-42 in pure form. Further cyclization of ent-42 was
achieved by treatment with ethylaluminum dichloride (EtAlCl2) as
before to yield ent-43 in excellent yield (see Supporting
Information for further details). This remarkable cyclization
generates three rings and sets four stereocenters, three of which
are part of the final meroterpenoid skeleton, in a highly
stereoselective manner. While the relative configuration proved
crucial to enable successful cyclization of the fully substituted
epoxide precursor, we were able to control the absolute
stereochemistry of the product by simple adjustment of the
epoxide stereocenter.
sequential functionalization of the aromatic core. Selective
bromination and methylation gave 44 in excellent yield.[3a]
Conversion of the bromine substituent to the corresponding
phenol was best carried out via a boronation-oxidation sequence.
The final oxidation to the p-quinone moiety of 5-epi-
smenoqualone (45) could be perfomed by exposure of the
phenol to salcomine under an oxygen atmosphere.[9d]
(–)-Cyclosmenospongine (4)
was then
obtained
via
aminolysis.[3d,7] Although we could unambiguously validate our
synthesized structure by means of single crystal X-ray analysis,
1
we were surprised to observe remarkable deviations of the H
and 13C NMR data (CDCl3) of synthetic and natural
cyclosmenospongine (4).[3d] While concentration effects could be
excluded, we hypothesized whether residual hydrogen chloride
could account for the observed inconsistencies. Indeed,
successive addition of a CDCl3 solution saturated with hydrogen
chloride to 4 led to formation of a deep-purple solution whose 1H
and 13C NMR shifts were now in better agreement with the
literature values. It was interesting to note that formation of the
hydrogen chloride adduct could be reversed by simply
concentrating the NMR sample (see Supporting Information for
NMR studies).
Having successfully investigated the impact of the relative
configuration of the epoxide at C3 and the C8 methyl substituent
on the cyclization cascade, we finally wanted to extend our
strategy to the synthesis of cis-fused decalins. According to the
Stork–Eschenmoser hypothesis, which postulates that adjacent
olefins attack via an anti-periplanar fashion, and its beautiful
application by Snyder in the synthesis of peyssonoic acid A,[37,38]
we anticipated that simple alteration of the enolether geometry
(C5-C10) might selectively yield the desired cis-decalin (Scheme
11).
Desulfurization of ent-43 intersected the previously route to
(–)-5-epi-aureol (38) and afforded ent-37 in good yield on gram-
scale (3.3 g). By adjusting the configuration of the epoxide at C3
in order to incorporate both methyl substituents, we could fully
bypass the previously developed eight-step sequence required
for the conversion of 28 to 37 and thus produce (+)-5-epi-(38) on
a 1.6-gram scale. Moreover, we had enough material to advance
5-epi-aureol (38) to (–)-cyclosmenospongine (4) through
Scheme 10. Total synthesis of (–)-cyclosmenospongine (4) via the optimized polyene cyclization. Reagents and conditions: a) EtAlCl2, CH2Cl2, –78 °C, 83%; b)
Et3SiH, CH2Cl2, BF3∙OEt2, 0 °C, 97%; c) (C6F5)OC(S)Cl, DMAP, CH2Cl2, then AIBN, nBu3SnH, benzene, 64%; d) BBr3, CH2Cl2, –78 °C to 23 °C, 85%; e) Br2,
CH2Cl2, –55 °C, 93%; f) Me2SO4, K2CO3, acetone, 23 °C, 87%; g) tBuLi, (iPrO)Bpin, THF, –78 °C to 0 °C, then H2O2, NaOH, 0 °C to 23 °C, 75%; h) salcomine, O2,
DMF, 23 °C, 77%; i) NH3, MeOH, H2O, pyridine, 23 °C, 60%. AIBN = azobisisobutyronitrile, B-OMe-9-BBN = B-methoxy-9-borabicyclo[3.3.1]nonane, DMAP = 4-
dimethylaminopyridine, pin = pinacol.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605029
Chemistry - A European Journal
FULL PAPER
Scheme 11. Preparation of cyclization precursor 47 and attempted synthesis of the cis-decalin framework 49. Reagents and conditions: a) CuI, Cs2CO3, N,N-
dimethylglycine hydrochloride, 1,4-dioxane, 100 °C, 61%; b) EtAlCl2, CH2Cl2, –78 °C, 44% of 50.
These considerations inspired us to investigate the simplified
model cyclization precursor 47, which was synthesized by
Ullmann coupling of vinyl iodide 46 with 4-methoxyphenol
Experimental Section
(13).[39,40] Unfortunately, all efforts to realize the cyclization of 47
failed and only bisketone 50, resulting from hydrolysis of the
enol ether and Meinwald rearrangement of the epoxide, could be
isolated. We believe that the pseudo-axial alignment of the C5-
substituent, as depicted for 48, is highly unfavorable and
therefore might prevent formation of acetal 48 or tetracycle 49.
Experimental and crystallographic details, as well as compound
characterization data and copies of 1H and 13C NMR spectra, are
available in the Supporting Information.
Acknowledgements
Conclusions
We gratefully acknowledge financial support from the Funds of
the Chemical Industry (Sachkostenzuschuss and Dozentenpreis
to T.M., Doktorandenstipendium to K.S.) and the German
Research Foundation (DFG Emmy Noether Fellowship to T.M.
and SFB TRR 152). We thank Dr. Peter Mayer (LMU Munich) for
X-ray crystal structure determination, Raphael Wildermuth (LMU
Munich), Benjamin Williams (LMU Munich) and Tatjana Huber
(LMU Munich) for helpful discussions, Michael Breunig
(University of Konstanz) for experimental assistance during the
investigation of the Suzuki–Miyaura cross-coupling scope and
Simone Schmidt (LMU Munich) for initial work on the Ullmann-
coupling approach.
In summary, we were able to develop a highly modular and
efficient
synthesis
of
the
tetracyclic
meroterpenoid
cyclosmenspongine (4). The establishment of a highly modular
three component coupling strategy, which is based on a
base-mediated phenol-alkyne addition and a highly efficient
sp2-sp3 B-alkyl Suzuki–Miyaura cross-coupling, allowed us to
rapidly modify the cyclization precursor in order to investigate
the cyclization cascade. Our initial efforts led to the development
of an unprecedented cationic polyene cyclization cascade of the
simplified polyene 26. Exposure of 26 to ethylaluminum
dichloride initiated the unique cascade, which forged the carbon
skeleton of aureol (1) in a single step. Further elaboration to 5-
epi-aureol (38) could be achieved by utilizing a Tschugaeff
elimination to access the key enone and by implementing an
unprecedented high-pressure 1,3-dipolar cycloaddition of a
thiocarbonylylide to install the vicinal cis-dimethyl group. Careful
analysis of the impact of stereochemistry on the cyclization
cascade allowed us to further optimize the precursor, thereby
shortening the synthesis of (–)-cyclosmenospongine (4) by eight
steps. The operationally simple and robust cyclization cascade
can be performed on multi-gram scale and allowed for the
synthesis of more than 400 mg of (–)-cyclosmenospongine (4).
Although we were unable to adapt the polyene cyclization to the
formation of the cis-decalin ring system, we hope to stimulate
further research in this area and advance polyene cyclizations to
the next level of complexity. Application of this concept to other
trans-decalin natural products and an alternative synthesis of
tetracyclic meroterpenoids bearing the cis-fused decalin ring
system are currently underway in our laboratories and will be
reported in due course.
Keywords: total synthesis • cyclization • high-pressure
chemistry • natural products • terpenoids
[1]
a) J. W. Blunt, B. R. Copp, R. A. Keyzers, M. H. G. Munro, M. R.
Prinsep, Nat. Prod. Rep. 2016, 33, 382–431; b) G. M. Cragg, P. G.
Grothaus, D. J. Newman, J. Nat. Prod. 2014, 77, 703–723.
M. Gordaliza, Mar. Drugs 2010, 8, 2849–2870.
[2]
[3]
Aureol (1): a) P. Djura, D. B. Stierle, B. Sullivan, D. J. Faulkner, J. Org.
Chem. 1980, 45, 1435–1441; Strongylin A (2): b) A. E. Wright, S. A.
Rueth, S. S. Cross, J. Nat. Prod. 1991, 54, 1108–1111; Smenoqualone
(3): c) M. L. Bourguet-Kondracki, M. T. Martin, M. Guyot, Tetrahedron
Lett. 1992, 33, 8079–8080; Cylosmenospongine (4): d) N. K. Utkina, V.
A. Denisenko, O. V. Scholokova, M. V. Virovaya, N. G. Prokof'eva,
Tetrahedron Lett. 2003, 44, 101–102.
[4]
[5]
J. W. Cornforth, Chem. Brit. 1968, 4, 102–106.
a) B. J. M. Jansen, Ae. De Groot, Nat. Prod. Rep. 2004, 21, 449–477.
b) P. M. Dewick in Medicinal Natural Products: A Biosynthetic Approach,
3rd Edition, John Wiley & Sons Ltd, 2009.
[6]
a) V. Lakshmi, S. P. Gunasekera, F. J. Schmitz, X. Ji, D. v. d. Helm, J.
Org. Chem. 1990, 55, 4709–4711; b) R. J. Capon, J. Nat. Prod. 1990,
This article is protected by copyright. All rights reserved.

10.1002/chem.201605029
Chemistry - A European Journal
FULL PAPER
53, 753–756; c) S. Urban, R. J. Capon, Aust. J. Chem. 1994, 47, 1023–
[20] The intermediate acetal 29 was not isolated as it is short-lived under
the reaction conditions (EtAlCl2, CH2Cl2, –78 °C).
1029.
[7]
[8]
N. K. Utkina, V. A. Denisenko, O. V. Scholokova, A. E. Makarchenko, J.
Nat. Prod. 2003, 66, 1263–1265.
[21] a) Y. Terao, M. Aono, N. Imai, K. Achiwa, Chem. Pharm. Bull. 1987, 35,
1734–1740; b) A. Hosomi, Y. Matsuyama, H. Sakurai, J. Chem. Soc.
Chem. Comm. 1986, 1073–1074.
Anti-viral activity for aureol (1): a) A. E. Wright, S. S. Cross, N. S.
Burres, F. Koehn, (Harbor Branch Oceanographics Institution, Inc.,
USA). PCT WO 9112250 A1, August 22, 1991. Anti-cancer activity for
aureol (1): b) L. Mani, V. Jullian, B. Mourkazel, A. Valentin, J. Dubois, T.
Cresteil, E. Folcher, J. N. A. Hooper, D. Erpenbeck, W. Aalbersberg, C.
Debitus, Chem Biodivers. 2012, 9, 1436–1451; c) H. Prawat, C.
Mahidol, W. Kaweetripob, S. Wittayalai, S. Ruchirawat, Tetrahedron
2012, 68, 6881–6886; d) R. E. Longley, O. J. Mcconnell, E. Essich, D.
Harmody, J. Nat. Prod. 1993, 56, 915–920; e) D. Tasdemir, T. S. Bugni,
G. C. Mangalindan, G. P. Concepcio, M. K. Harper, C. M. Ireland, Z.
Naturforsch. 2002, 57c, 914–922; for strongylin A (2): f) S. J. Coval, M.
A. Conover, R. Mierzwa, A. King, M. S. Puar, Bioorg. Med. Chem. Lett.
1995, 5, 605–610; for cyclosmenospongine (4): g) N. G. Prokof'eva, N.
K. Utkina, E. L. Chaikina, A. E. Makarchenko, Comp. Biochem. Phys. B.
2004, 139, 169–173; h) N. K. Utkina, A. E. Makarchenko, O. V.
Shchelokova, M. V. Virovaya, Chem. Nat. Compd. 2004, 40, 373–377.
Anti-bacterial activity for aureol (1): i) J. Hu, J. A. Schetz, M. Kelly, J.
Peng, K. K. H. Ang, H. Flotow, C. Y. Leong, S. B. Ng, A. D. Buss, S. P.
Wilkins, M. T. Hamann, J. Nat. Prod. 2002, 65, 476–480.
[22] Y. Ito, T. Hirao, T. Saegusa, J. Org. Chem. 1978, 43, 1011–1013.
[23] a) E. Vedejs, J. E. Telschow, J. Org. Chem. 1976, 41, 740–741; b) J.
Clayden, M. N. Kenworthy, M. Helliwell, Org. Lett. 2003, 5, 831–834.
[24] a) T. N. Diao, S. S. Stahl, J. Am. Chem. Soc. 2011, 133, 14566–14569;
b) K. C. Nicolaou, Y. L. Zhong, P. S. Baran, J. Am. Chem. Soc. 2000,
122, 7596–7597; c) Y. Chen, J. P. Romaire, T. R. Newhouse, J. Am.
Chem. Soc. 2015, 137, 5875–5878; d) A. Turlik, Y. F. Chen, T. R.
Newhouse, Synlett 2016, 27, 331–336.
[25] B. Lesch, S. Bräse, Angew. Chem, 2004, 116, 118–120; Angew. Chem.
Int. Ed. 2004, 43, 115–118.
[26] a) G. M. Atkins, E. M. Burgess, J. Am. Chem. Soc. 1968, 90, 4744–
4745; b) R. J. Arhart, J. C. Martin, J. Am. Chem. Soc. 1972, 94, 5003–
5010; c) H. Watanabe, M. Takano, A. Umino, T. Ito, H. Ishikawa, M.
Nakada, Org. Lett. 2007, 9, 359–362.
[27] a) L. Tschugaeff, Ber. Dtsch. Chem. Ges. 1900, 33, 3118–3126; b) K.
Lee, D. L. Boger, J. Am. Chem. Soc. 2014, 136, 3312–3317.
[28] a) A. Magyarosy, R. M. Mohareb, J. Z. Ho, Heteroatom Chem. 2006, 17,
648–652; b) D. B. Li, M. Rogers-Evans, E. M. Carreira, Org. Lett. 2013,
15, 4766–4769.
[9]
For a review on published syntheses of tetraycyclic meroterpenoids,
see: a) T. Katoh, Heterocycles 2013, 87, 2199–2224; b) I. S. Marcos, A.
Conde, R. F. Moro, P. Basabe, D. Díez, J. G. Urones, Tetrahedron
2010, 66, 8280–8290; c) K. K. W. Kuan, H. P. Pepper, W. M. Bloch, J.
H. George, Org. Lett. 2012, 14, 4710–4713; d) M. Nakamura, A. Suzuki,
M. Nakatani, T. Fuchikami, M. Inoue, T. Katoh, Tetrahedron Lett. 2002,
43, 6929–6932; e) M. Nakatani, M. Nakamura, A. Suzuki, M. Inoue, T.
Katoh, Org. Lett. 2002, 4, 4483–4486; f) J. Sakurai, T. Kikuchi, O.
Takahashi, K. Watanabe, T. Katoh, Eur. J. Org. Chem. 2011, 2011,
2948–2957; g) T. Kamishima, T. Kikuchi, T. Katoh, Eur. J. Org. Chem.
2013, 4558–4563; h) A. Rosales, J. Muñoz-Bascón, E. Roldan-Molina,
N. Rivas-Bascón, N. M. Padial, R. Rodríguez-Maecker, I. Rodríguez-
García, J. E. Oltra, J. Org. Chem. 2015, 80, 1866–1870; i) D. T. Ngoc,
M. Albicker, L. Schneider, N. Cramer, Org. Biomol. Chem. 2010, 8,
1781–1784.
[29] C. L. Hugelshofer, T. Magauer, Synthesis 2014, 46, 1279–1296.
[30] M. W. Smith, S. A. Snyder, J. Am. Chem. Soc. 2013, 135, 12964–
12967.
[31] a) R. M. Letcher, T.-Y. Yue, K.-F. Chiu, A. S. Kelkar, K.-K. Cheung, J.
Chem. Soc. Perkin Trans. 1 1998, 3267–3276; b) C. S. Mizuno, A. G.
Chittiboyina, F. H. Shah, A. Patny, T. W. Kurtz, H. A. Pershadsingh, R.
C. Speth, V. T. Karamyan, P. B. Carvalho, M. A. Avery, J. Med. Chem.
2010, 53, 1076–1085.
[32] D. S. Tarbell, P. E. Fanta, J. Am. Chem. Soc. 1943, 65, 2169–2174.
[33] a) D. D. Dixon, J. W. Lockner, Q. Zhou, P. S. Baran, J. Am. Chem. Soc.
2012, 134, 8432–8435; b) D. S. Treitler, Z. F. Li, M. Krystal, N. A.
Meanwell, S. A. Snyder, Bioorg. Med. Chem. Lett. 2013, 23, 2192–
2196; c) A. Procopio, P. Costanzo, M. Curini, M. Nardi, M. Oliverio, R.
Paonessa, Synthesis 2011, 1, 73–78; d) G. Bartoli, M. Bosco, A.
Carlone, M. Locatelli, E. Marcantoni, P. Melchiorre, L. Sambri, Adv.
Synth. Catal. 2006, 348, 905–910.
[10] K. Speck, R. Wildermuth, T. Magauer, Angew. Chem, 2016, 128,
14337–14341; Angew. Chem. Int. Ed. 2016, 55, 14131–14135.
[11] A. G. Myers, L. McKinstry, J. Org. Chem. 1996, 61, 2428–2440
[12] K. Speck, K. Karaghiosoff, T. Magauer, Org. Lett. 2015, 17, 1982–1985.
[13] S. H. Wang, P. H. Li, L. Yu, L. Wang, Org. Lett. 2011, 13, 5968–5971.
[14] a) N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457–2483; b) G.
Seidel, A. Fürstner, Chem. Commun. 2012, 48, 2055–2070; c) Z. Lu, G.
C. Fu, Angew. Chem, 2010, 122, 6826–6828; Angew. Chem. Int. Ed.
2010, 49, 6676–6678; d) S. R. Chemler, D. Trauner, S. J. Danishefsky,
Angew. Chem, 2001, 113, 4676–4701; Angew. Chem. Int. Ed. 2001, 40,
4544–4568; e) A. J. J. Lennox, G. C. Lloyd-Jones, Chem. Soc. Rev.
2014, 43, 412–443.
[34] a) Y. Fukuyama, Y. Otoshi, M. Kodama, T. Hasegawa, H. Okazaki, M.
Nagasawa, Tetrahedron Lett. 1989, 30, 5907–5910; b) C. Calderon-
Higginson, L. Crombie, S. D. Redshaw, D. A. Whiting, J. Chem. Soc.,
Perkin Trans. 1 2000, 2491–2494; c) B. Li, M. Berliner, R. Buzon, C. K.
F. Chiu, S. T. Colgan, T. Kaneko, N. Keene, W. Kissel, T. Le, K. R.
Leeman, B. Marquez, R. Morris, L. Newell, S. Wunderwald, M. Witt, J.
Weaver, Z. Zhang, Z. Zhang, J. Org. Chem. 2006, 71, 9045–9050.
[35] a) S. Ramakanth, K. Narayanan, K. K. Balasubramanian, K.
Rajagopalan, Tetrahedron 1986, 42, 863–876; b) F. González-López
de Turiso, D. P. Curran, Org. Lett. 2005, 7, 151–154.
[15] K. Ishigai, H. Fuwa, K. Hashizume, R. Fukazawa, Y. Cho, M. Yotsu-
Yamashita, M. Sasaki, Chem. Eur. J. 2013, 19, 5276–5288.
[36] a) J.-P. Lumb, K. C. Choong, D. Trauner, J. Am. Chem. Soc. 2008, 130,
9230–9231; b) M. Oikawa, T. Ueno, H. Oikawa, A. Ichihara, J. Org.
Chem. 1995, 60, 5048–5068; c) S. Böhmdorfer, A. Patel, A. Hofinger, T.
Netscher, L. Gille, T. Rosenau, Eur. J. Org. Chem. 2011, 3036–3049.
[37] a) G. Stork, A. W. Burgstahler, J. Am. Chem. Soc. 1955, 77, 5068–
5077; b) A. Eschenmoser, L. Ruzicka, O. Jeger, D. Arigoni, Helv. Chim.
Acta 1955, 38, 1890–1904.
[16] a) T. E. Barder, S. D. Walker, J. R. Martinelli, S. L. Buchwald, J. Am.
Chem. Soc. 2005, 127, 4685–4696; b) N. C. Bruno, M. T. Tudge, S. L.
Buchwald, Chem. Sci. 2013, 4, 916–920.
[17] Tris(pentafluorophenyl)borane B(C6F5)3 was used instead of BF3·OEt2
to avoid fluorohydrine formation.
[18] For related thioenol ether cyclizations, see: a) T. A. Blumenkopf, M.
Bratz, A. Castaneda, G. C. Look, L. E. Overman, D. Rodriguez, A. S.
Thompson, J. Am. Chem. Soc. 1990, 112, 4386–4399; b) H. Kusama,
R. Hara, S. Kawahara, T. Nishimori, H. Kashima, N. Nakamura, K.
Morihira, I. Kuwaijima, J. Am. Chem. Soc. 2000, 122, 3811–3820; c) P.
K. Sasmal, M. E. Maier, Org. Lett. 2002, 4, 1271–1274; d) P. K. Sasmal,
M. E. Maier, J. Org. Chem. 2003, 68, 824–831.
[38] a) S. A. Snyder, D. S. Treitler, Angew. Chem, 2009, 121, 8039–8043;
Angew. Chem. Int. Ed. 2009, 48, 7899–7903; b) S. A. Snyder, D. S.
Treitler, A. P. Brucks, J. Am. Chem. Soc. 2010, 132, 14303–14314.
[39] For the preparation of vinyliodide 46, see Supporting Information.
[40] D. Ma, Q. Cai, X. Xie, Synlett, 2005, 1767–1770.
[19] D. J. Ager, J. Chem. Soc. Perkin Trans. 1 1983, 1131–1136.
This article is protected by copyright. All rights reserved.

10.1002/chem.201605029
Chemistry - A European Journal
FULL PAPER
FULL PAPER
Klaus Speck and Thomas Magauer*
Page No. – Page No.
Evolution of a Polyene Cyclization
Cascade for the Total Synthesis of
(–)-Cyclosmenospongine
We report a full account on the development of a unique polyene cyclization for the
total synthesis of (–)-cyclosmenospongine that generates three rings and sets four
consecutive stereocenters in a single step. A highly convergent three-component
coupling strategy enabled rapid access to individual cyclization precursors which
were tested for their reactivity. We found that the enol ether geometry and the
relative configuration of C3 and C8 are crucial for the success of the cyclization.
This article is protected by copyright. All rights reserved.