Divergent syntheses of resorcylic acid lactones: L-783277, LL-Z1640-2, and hypothemycin

Article abstract of DOI:10.1002/chem.200901373

The resorcylic acid lactones (RAL) are endowed with diverse biological activity ranging from transcription factor modulators (zearalenone and zearalenol) to HSP90 inhibitors (radicicol and pochonin D) and reversible (aigialomycin D) as well as irreversibl

Full text of DOI:10.1002/chem.200901373

DOI: 10.1002/chem.200901373
Divergent Syntheses of Resorcylic Acid Lactones: L-783277, LL-Z1640-2, and
Hypothemycin
Pierre-Yves Dakas, Rajamalleswaramma Jogireddy, Gaꢀlle Valot, Sofia Barluenga, and
Nicolas Winssinger*[a]
Dedicated to Professor Jean-Marie Lehn on the occasion of his 70th birthday
Abstract: The resorcylic acid lactones
(RAL) are endowed with diverse bio-
logical activity ranging from transcrip-
tion factor modulators (zearalenone
and zearalenol) to HSP90 inhibitors
(radicicol and pochonin D) and revers-
enone). Our interest in broadening the
diversity of this family beyond natural-
ly occurring diversity has led us to seek
a general approach that could be used
to address the entire spectrum of func-
tionalities present within this family.
Herein, we present our efforts on ac-
cessing macrocycles bearing an alkane,
alkene, or epoxide at the benzylic posi-
tion from a common benzylic sulfide
intermediate to access L-783277, LL-
Z1640-2, and hypothemycin.
Keywords: kinases
resorcylic acid
synthesis · sulfur
·
lactones
solid-phase
·
ACHTUNGTRENNUNG
·
A
ACHTUNGTRENNUNG
cin and other RAL containing a cis-
Introduction
While a number of resorcylic acid lactones (RAL) have
been know for over 30 years, the more recent discovery that
several members of this family are potent kinase and
ATPase inhibitors have brought renewed attention to the
chemistry of this important family of natural products (a se-
lection of which are shown here).[1,2] Two members of this
family have now been co-crystallized with a targeted protein
(radicicol with HSP90[3] and LL-Z1640-2 with ERK2[4,5]) and
shown to be competitive ligands for the ATP-binding
pocket. We have been particularly interested in this subset
of RAL as a privileged starting point for the elaboration of
chemical probes based on the prevalence of ATP-binding
motifs across the proteome. Furthermore, resorcylides
appear to exploit a different area of chemical diversity
space than heterocyclic compounds typically pursued for
this purpose.[6] Thus far, over thirty natural products belong-
ing to this family have been reported and new sources[7,8]
continue to yield variants. In addition, a better understand-
ing of the biosynthesis[9,10] enabling manipulation of the po-
lyketide machinery will undoubtedly provide additional ana-
logues. Nevertheless, chemical synthesis remains important
in particular to access compounds that are chemically not
accessible through biosynthetic pathways.[11] Our first efforts
towards radicicol and pochonin C had led us to explore a
sulfide linker a to a carbonyl as a convenient means of
masking a sensitive conjugate alkene and providing numer-
ous chemical opportunities to construct adjacent bonds.[12,13]
[a] P.-Y. Dakas, Dr. R. Jogireddy, G. Valot, Dr. S. Barluenga,
Prof. Dr. N. Winssinger
Institut de Science et Ingꢀnierie Supramolꢀculaires (ISIS–UMR 7006)
Universitꢀ de Strasbourg—CNRS
8 allꢀe Gaspard Monge, 67000 Strasbourg (France)
Fax. (+33)368-855-113
E-mail: winssinger@isis.u-strasbg.fr
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901373.
11490
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 11490 – 11497

FULL PAPER
We envisioned that a sulfide at the benzylic position
(Scheme 1) could indeed provide access to all the other
functionalities present in the RAL family, namely, an
alkane, an alkene, and an epoxide. As shown in Scheme 1,
Scheme 1. Divergent synthesis of benzylic alkene 2, alkane 3, and epox-
ide 4 from common intermediate 1.
Scheme 2. a) H2O2 (4.0 equiv), HFIP/CH2Cl2 (1:1), 12 h, 238C, quant; b)
KOtBu (6.6 equiv), All-I (8.0 equiv), DMSO, 11 h, 238C, 60%; c) TFAA
(10 equiv), All-TMS (20 equiv), CH2Cl2, 12 h, À78 to 238C, 54%; d)
LDA (6.0 equiv), All-Br (3.0 equiv), THF/HMPA (10:1), À788C, 30 min,
92%; e) Me3OTf (1.5 equiv), CH2Cl2, 1.5 h, 238C; f) 3-phenylpropanal
(3.0 equiv), DBU (10 equiv), CH2Cl2, 12 h, 238C, 24%; All-Br=allylbro-
mide, All-I=allyliodide, All-TMS=allyltrimethylsilane, DBU=1,8-
the resorcylic ester 1 bearing a sulfide at the benzylic posi-
tion could be directly alkylated following deprotonation
with strong bases such as lithium diisopropylamide
(LDA).[14] Conversely, oxidation of the sulfide to a sulfoxide
should further acidify the benzylic position thus enabling al-
kylation under milder conditions. Elimination of the sulfox-
ide or reductive cleavage of the sulfide would afford the
benzylic alkene 2 and alkane 3, respectively. The sulfoxide
moiety could also provide an entry in Pummerer reactions.
On the other hand, methylation of the sulfide would yield a
sulfur ylide, which are known to participate in the formation
of epoxides such as 4. Indeed, some precedents for each of
these transformations existed starting with the pioneering
work of Takahashi and Tsuji using an alkylation of a benzyl-
ic sulfide for the macrocyclization of zearalenoneꢂs core.[15]
Gennari and co-workers had shown that stabilized sulfur
ylides could be used in cyclative cleavages from solid
phase[16] and the Pummerer reaction of sulfoxide had been
used on solid phase.[17] Of course, these transformations
could also be envisioned with the corresponding selenides
and a polymer-bound version of the most popular selenium
reagents[18] has been used extensively.
DiazabicycloACTHNUTRGNEUG[N 5.4.0]undec-7-ene, DMSO=dimethylsulfoxide, HFIP=hexa-
fluoroisopropanol, HMPA=hexamethylphosphoramide, LDA=lithium
diisopropylamide, TFAA=trifluoro acetic anhydride, THF=tetrahydro-
furan, Tf=triflate.
acid (mCPBA)). We had previously noted that alkylation of
sulfoxide proceed in highest yield in dimethylsulfoxide
(DMSO).[21] This was also the case for 6, which could be al-
kylated at room temperature with allyl iodide to obtain 7 in
66% yield (calculated based on the amount of product re-
covered following elimination from the resin). A similar
transformation using lithium bis(trimethylsilyl)amide
(LiHMDS) in THF/hexamethylphosphoramide (HMPA)
was previously reported to be unsuccessful[22] attesting to
the importance of the solvent in this reaction. The elimina-
tion of this sulfoxide did not proceed unless a temperature
of 60–808C was attained and thus allowed for washing of re-
agents prior to elimination in a different solvent (toluene).
The sulfoxide could also be allylated under Pummerer con-
ditions (trifluoro acetic anhydride (TFAA), allyl silane) to
afford 8 in moderate yield (54% based on recovered prod-
uct following oxidation/elimination). Deprotonation of sul-
fide 5 with an excess of LDA followed by addition of allyl
bromide afforded the product 8 in excellent yield (92%
based on the amount of product recovered following elimi-
nation from the resin). Methylation of sulfide 5 with MeOTf
(OTf=triflate) afforded 9 which upon treatment with 1,8-
Results and Discussion
As shown in Scheme 2, each of these proposed transforma-
tions was evaluated starting from sulfide 5. Thus, sulfide 5,
which was obtained from the 2-(chloromethyl)-4,6-dime-
thoxy benzoyl chloride[19] by esterification in methanol and
loading onto a thiophenol resin, could be oxidized to the
corresponding sulfoxide with H2O2 in hexafluoroisopropanol
(HFIP)/CH2Cl2 to afford 6.[20] This procedure was found to
be extremely practical as it can be driven to completion
with an excess of H2O2 (typically 4 equiv) without over oxi-
dation to the sulfone, as was observed with other procedures
(dimethyldioxirane (DMDO) or meta-chloroperoxybenzoic
diazabicycloACTHNUGTRNEUNG[5.4.0]undec-7-ene (DBU) and addition of an al-
dehyde afforded the epoxide 10 in modest yield (24%). It
should be noted that such benzylic epoxides were found to
be particularly sensitive to acid and were difficult to isolate
in good yield. While none of these reactions were optimized,
Chem. Eur. J. 2009, 15, 11490 – 11497
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11491

N. Winssinger et al.
they provided clear indication of the versatility of benzylic
sulfide.
by addition of diphenyl diselenide provided the desired
compound 12a in high yield. This reaction has proven ex-
tremely reliable with a number of different substrates. How-
ever, the analogous reaction with diphenyl disulfide to
obtain 13 proved to be less reliable, as it frequently was con-
taminated with the di-alkylated compound 14. Reactions
with different source of electrophilic selenium or sulfur that
are compatible with solid-phase processes proved to be less
productive. Indeed, reaction with phenyl selenium bromide
provided only modest yield of 12a, while reaction of 15 with
phenyl thiotosylate[23,24] was equally modest both in solution
and on solid phase (16). As can be expected, reaction with
nucleophilic selenium or thiol were found to be high yield-
ing (5, Scheme 2, was obtained quantitatively from the cor-
responding benzyl chloride and thiophenol resin[14] using
DBU in DMF); however, access to the benzyl chloride or
bromide with suitable protecting groups can be lengthy.
Free-radical halogenations of the benzylic position was not
reliable in our hands with the protecting-group combination
sought (ethoxymethyl (EOM)-protected phenol, 2-trimethyl-
silylethanol (TMSE)-protected ester). We thus evaluated if
the phthalide ring could be opened with thioxides as there
are a number of precedents in the literature.[25] As shown in
Scheme 3, reaction of 17 with sodium thioxide in solution or
on a solid phase failed to provide the desired compound 18
in good yield and purity as significant level of demethylation
was observed. The most convenient access to polymer-
bound resorcylate with high flexibility in the choice of pro-
tecting groups on the phenol was to start from an acetonide
protection of the acid and ortho-phenol 19, which is ob-
tained in three steps from the readily available 3,5-dihydroxy-
benzyl alcohol (POCl3, DMF; NaClO2; acetone, trifluoro-
acetic acid (TFA), TFAA).[19,26] Substitution of the benzyl
chloride with a thiophenol resin afforded 20 in quantitative
yield (estimated based on the mass gain of the resin and
analogous reaction in solution). Protection of the para-
phenol followed by acetonide opening in the presence of
(trimethylsilyl)ethoxide afforded 21, which was readily con-
verted to 24. Alternatively, we reasoned that the ester
moiety could be introduced through palladium-mediated
carbonylation. The corresponding aryl iodide 22 being readi-
ly available from Ipy2BF4 iodination[27] of 3,5-dihydroxyben-
zyl alcohol. Tosylation of the benzyl alcohol 22 followed by
loading on the thiophenol resin afforded resin 23 in excel-
lent yield. While the carbonylation[28,29] worked well with
MeOH, it was not effective for secondary alcohols and af-
forded moderate yield with (trimethylsilyl)ethanol. In all re-
actions, partial or complete loss of the EOM group was ob-
served leading to 21 rather than 24. Alternatively, a Diels–
Alder reaction could be envisioned to construct the desired
aromatic ring from propargylic ester and dimedone diene, as
has been elegantly demonstrated by Danishefsky,[30] but the
high yield in the preparation of intermediate 20 and its ver-
satility did not warrant this exploration.
To access the required sulfide or selenoide such as 1
(Scheme 1), the sulfur or selenium can be envisioned as a
nucleophile or as an electrophile. For electrophilic sulfur or
selenium, the toluate position can readily be deprotonated
with LDA. Treatment of 11 (Scheme 3) with LDA followed
Scheme 3. a) LDA (2.0 equiv), (PhSe)2 (1.0 equiv), THF, À788C, 45 min,
90%; b) LDA (2.0 equiv), PhSeBr (1.0 equiv), THF, À788C, 45 min,
48%; c) LDA (2.0 equiv), (PhS)2 (1.0 equiv), THF, À788C, 45 min,
>90% for 13/14 1.8:1; d) 15 (3.0 equiv), LDA (6.0 equiv); PS-SSO2Tol
(1.0 equiv), THF, À788C, 1 h, 60%; e) PhSH (1.0 equiv), NaH
(1.0 equiv), DMF, 1508C, 12 h, 45%; f) in solution: Ph-SH (1.0 equiv), 19
(1.0 equiv), iPr2NEt (1.0 equiv), DMF, 608C, 12 h, 98%; on solid phase:
PS-SH (1.0 equiv, 0.8 mmolg)À1
, 19 (1.5 equiv), iPr2NEt (1.0 equiv),
DMF, 608C, 12 h, quantitative; g) in solution: EOM-Cl (2.0 equiv),
iPr2NEt (2.0 equiv), TBAI (cat), CH2Cl2, 238C, 12 h, 97%; on solid
phase: EOM-Cl (2.0 equiv), DBU (2.0 equiv), TBAI (cat), DMF, 238C,
12 h. quantitative; h) in solution: TMSEOH (1.1 equiv), NaHMDS
(1.1 equiv), THF, 08C, 2 h, 95%; on solid phase: TMSEOH (4.2 equiv),
NaHMDS (1.1 equiv), THF, 0 to 238C, 12 h; i) TsCl (1.5 equiv), pyridine
(2.2 equiv), CH2Cl2, 0 to 238C, 12 h, quantitative; j) In solution: [Pd-
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
idine, DMF=N,N-dimethylformamide, EOM=ethoxymethyl, LDA=
lithium diisopropylamide, PS=polystyrene, NaHMDS=sodium bis(tri-
methylsilyl)amide, THF=tetrahydrofuran, TBAI=tetrabutyl ammonium
iodide, TMSE=2-trimethylsilylethanol, Tol=toluene, Ts=tosyl.
A number of resorcylides of interest contain an alkene at
the b-position of the lactone, which provides an evident dis-
connection through metathesis. Conveniently, both stereo-
11492
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 11490 – 11497

Natural Products
FULL PAPER
chemical derivatives of 3-hy-
droxybutene are commercially
available. However, the me-
ACHTUNGTRENNUNGtathACHTUNGTRENNUNGesis cannot be used in a
straightforward manner to
access a RAL containing a cis-
enone (vide infra). To this end,
we needed access to a cis-vinyl
halide for transmetalation or
cross-coupling reactions. As
shown in Scheme 4, p-meth-
ACHTUNGTRENNUNGoxyACHTUNGTRENNUNGbenzyl (PMB)-protection
Scheme 4. a) PMBOCNHCCl3 (1.0 equiv), CSA (0.1 equiv), CH2Cl2, 238C, 12 h, 92%; b) O3, CH2Cl2, À788C;
Ph3P (2.0 equiv), À78 to 238C, 2 h, 92%; c) Ph3PCH2I (1.25 equiv), NaHMDS (1.25 equiv), THF/HMPA
(7.7:1), 238C, 2 h, 26%; d) vinyl borolane (1.0 equiv), Grubbs-Grella II (2.5 mol%), toluene, 808C, 12 h, 92%;
e) Br2 (1.0 equiv), Et2O, À208C, 15 min; NaOMe, À208C, 30 min, 89%; f) DIBAL-H, (1.1 equiv), toluene,
of 3-hydroxybutene afforded
26 in high yield, which was
converted to 27 by ozonolysis
followed by olefination with
methyl iodide ylide. This strat-
egy was appealing as immobi-
lized versions of all reagents
used in this sequence are avail-
able;[31,32] however, the Wittig
reaction could not be per-
formed in acceptable yield
(<30%). Alternatively, cross
metathesis of 26 with vinyl
borolane[33] by using the Grela
modification[34] of the Grubbs
second-generation catalyst af-
forded the trans-vinyl borolane
28 (>20:1 E:Z), which was
stereospecifically converted to
the cis-vinyl bromide using a
À788C, 2 h, 89%; g) CBr4 (4.0 equiv), Ph3P (8.0 equiv), CH2Cl2, 08C, 45 min, 90%; h) [Pd
ACHTUNGTRENUN(NG PPh3)4] (0.2 equiv),
nBu3SnH (5.1 equiv), benzene, 238C, 1.5 h, 95%. CSA=camphorsulfonic acid, DIBAL-H=diisobutylalumini-
um hydride, HMPA=hexamethylphosphoramide, NaHMDS=sodium bis(trimethylsilyl)amide, PMB=p-me-
thoxybenzyl, THF=tetrahydrofuran.
procedure
reported
by
Brown[35] to obtain vinyl bro-
mide 29 in good yield (82%
for two steps). Conversely, the
same vinyl bromide 29 could
be obtained in four steps from
the less expensive methyl 3-hy-
droxybutyrate (30), which is
also available in both stereo-
chemistries. Thus, PMB protec-
tion followed by a diisobutyl-
aluminium hydride (DIBAL-
H) reduction and Corey–Fuchs
reaction afforded compound
31, in which the more reactive
trans-vinyl bromide was re-
duced with tributyltin hydride
under the action of palladium
tetrakis to afford 29 in excel-
lent yield.[36]
Scheme 5. a) 2-Buten-1,4-diol (2.0 equiv), Grubbs-Hoveyda II (0.01 equiv), CH2Cl2, 238C, 4 h, 85–97%; b) l-
(+) diethyltartrate (0.12 equiv), Ti
(OiPr)4 (0.1 equiv), tBuOOH (1.52 equiv), CH2Cl2, À408C,12 h, 85%; c)
SO3·py (3.5 equiv), Et3N (4.9 equiv), CH2Cl2/DMSO (4:1), 0 to 238C, 30 min; d) Ph3P=CH2 (1.9 equiv),
NaHMDS (1.8 equiv), THF, À108C, 40 min, 70% (2 steps); e) PS-SO3H (0.35 equiv), acetone, 238C, 12 h,
40%; f) ScACHTNUGTRNEG(UN OTf)3 (0.2 equiv), THF/H2O (10:1), 238C, 2.5–12 h, 50–99%; dimethoxypropane (10 equiv),
TsOH·H2O (0.05 equiv), CH2Cl2, 12 h, 238C, 70–90%; g) O3, Ph3P (2.0 equiv), CH2Cl2, À78 to 238C, 2 h, 92 %;
h) LiAlH4 (1.4 equiv), THF, 0 to 238C, 2 h, 95%; i) TBDPSCl (1.0 equiv), imidazole (1.5 equiv), DMF, 238C,
2 h, 66%; j) PS-IBX (3.0 equiv), CH2Cl2, 238C, 2 h, quantitative; k) vinylMgBr (1.5 equiv), THF, 08C, 1 h,
92%; l) EOM-Cl (8.0 equiv), iPr2NEt (8.0 equiv), TBAI (cat), CH2Cl2, 238C, 12 h, 99%; m) TBAF (2.0 equiv),
THF, 238C, 6 h, quantitative; I2 (1.5 equiv), Ph3P (1.5 equiv), imidazole (2.5 equiv), THF, 08C, 30 min, 91%; n)
29 (1.0 equiv), tBuLi (2.0 equiv), Et2O, À1008C, 30 min, 88%; o) EOM-Cl (8.0 equiv), iPr2EtN (8.0 equiv),
TBAI (cat), CH2Cl2, 238C, 6 h, 98%; p) BzCl (2.5 equiv), pyridine (2.5 equiv), TBAI (cat), CH2Cl2, 0 to 238C,
6 h, 90%. Bz=benzoyl, DMF=N,N-dimethylformamide, EOM=ethoxymethyl, IBX=2-iodobenzoicacid,
Imid=imidazole, NaHMDS=sodium bis(trimethylsilyl)amide, SAE=Sharpless asymmetric epoxidation, PS=
polymer supported, Py=pyridine, TBAF=tetrabutylammonium fluoride, TBAI=Tetrabutylammonium
iodide, TBDPS=tert-butyldiphenylsilyl, THF=tetrahydrofuran, Tos=Tosyl.
A number of resorcylides contain an allylic or homoallylic
anti-diol moiety in the macrocycle. We had previously
shown[37] that such acetonide-protected diols could be con-
veniently obtained starting from the alkene bromide 32
(Scheme 5) in six steps via allylic epoxide 35 through using a
sequence involving cross metathesis with 1,4-butenediol,
Sharpless epoxidation, oxidation, olefination, and stereose-
lective opening of the epoxide 35.[37] This epoxide could be
opened by using protic or Lewis acid conditions, such as Sc-
AHCTUNGRTEG(NNUN OTf)3, followed by acetonide protection to obtain 38. Most
Chem. Eur. J. 2009, 15, 11490 – 11497
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11493

N. Winssinger et al.
conveniently, the epoxide opening and acetonide protection
can be carried out in one-pot by using sulfonic acid resin in
acetone. However, using the shorter alkene chain, the
chemistry proved less productive. While 36 could be ob-
tained in good yield, its conversion to 39 was accompanied
by the formation of an inseparable side product. Alterna-
tively, the bromide could be masked in the form of a pro-
tected hydroxyl. Starting with tert-butyldiphenylsilyl
(TBDPS)-protected butenol 34, allylic epoxide 37 was ob-
tained in good yield; however, its conversion to 40 was ac-
companied by partial desilylation under all conditions tried.
Nevertheless, ozonolysis of 40 afforded aldehyde 42, which
was more conveniently obtained from the known acetonide
protected deoxyribose 41 in three steps.[37] Vinyl Grignard
addition onto the aldehyde 42 followed by EOM-Cl protec-
tion of the resulting alcohol afforded compound 43, which
was converted to the required iodide 44 through convention-
al procedure (tetrabutylammonium fluoride (TBAF) desily-
lation and iodination). Conversely, treatment of 42 with the
vinyl lithium obtained upon treatment of 29 with tBuLi af-
forded 45 in good yield as a diastereomeric mixture. This in-
termediate can be converted to the fragment bearing acid
labile protecting groups on the triol (46) or compound 47
bearing an orthogonal protecting group on the allylic alco-
hol (Bz).
Scheme 6. a) LDA (2.0 equiv), 44 (1.0 equiv), THF/HMPA (10:1),
À788C, 20 min, 80%; b) Grubbs II (0.1 equiv), toluene, 608C, 3 h, 94%;
c) H2O2 (2.0 equiv), THF, 238C, 3 h, quantitative; d) PS-SO3H (10 equiv),
MeOH, 508C, 12 h, quantitative. Grubbs II=benzylidene [1,3-bis(2,4,6-
trimethylphenyl)-2-imidazolidinylidene]dichloro(tricyclohexylphosphi-
ne)ruthenium, HMPA=hexamethylphosphoramide, LDA=lithium diiso-
propylamide, PS-SO3H=polysmer supported sulfonic acid, THF=tetra-
hydrofuran.
thesis, we envisioned starting from 18 (Scheme 7) and 51
(obtained by PMB-deprotection of 46) through Mitsunobu
reaction followed by a macrocyclization. Macrocyclization
directly from the sulfide (not shown) was low yielding due
to decomposition of the product over prolonged exposure to
LDA; however, excellent yields were obtained using the
sulfoxide 52 in DMSO. While the sulfoxide may be isolated
by flash chromatography, the crude product was redissolved
directly in toluene and irradiated with microwave to 1208C
to recover the desired alkene in 98% yield (two steps). This
chemistry also proved equally productive on solid phase af-
fording macrocycle 53 in 34% overall yield (based on the
loading of thiophenol resin). Compound 53 was treated with
BCl3 to deprotect selectively the ortho-phenol as well as the
acetonide and EOM to afford the triol, which was selective-
ly oxidized at the allylic position to yield LL-Z1640-2. We
While our experience with aigialomycin D suggested that
a metathesis should provide the trans-alkene,[14] this specula-
tion was assessed experimentally. Starting from readily avail-
able 12b, deprotonation of the toluate with LDA and al
ACHTUNGTNERkNUNG yl-
ACHTUNGTRENNUNGation with 44 afforded the metathesis precursor 48 in good
yield (Scheme 6). Treatment with the second-generation
Grubbs catalyst[38] at 608C afforded a complete conversion
to 49 within three hours. While the product obtained had a
complex NMR spectrum due to the presence of multiple
diastereoisomers, oxidation/elimination of the selenide fol-
lowed by removal of the protecting groups led to compound
50, which was assigned as the product of an E-selective
metathesis. While a number of
RAL contain
a trans-enone
system, including some aigialo-
mycins[39] and the more recent-
ly isolated caryospomycins,[8]
none of them has been report-
ed to be potent modulators of
a protein function. Of course,
although introduction of con-
formation constrains with dif-
ferent protecting groups may
lead to a Z-selective metathe-
sis reaction (as is the case of
radicicol),[12,40] the strategy was
not deemed general enough
for the purposes of library syn-
thesis.
Scheme 7. a) In solution: 51 (1.0 equiv), DIAD (2.0 equiv), Ph3P (2.0 equiv), toluene, 08C, 1 h, 75%; on solid
phase: 51 (2.0 equiv), DIAD (4.0 equiv), Ph3P (4.0 equiv), toluene, 238C, 12 h; b) in solution: H2O2
(1.0 equiv), HFIP, 238C, 1 h, quantitative; on solid phase: H2O2 (4.0 equiv), HFIP/CH2Cl2 (1:1), 238C, 12 h; c)
in solution: KOtBu (1.5 equiv), DMSO, 238C, 2 h, then toluene, 1208C, mwave, 25 min, 98%; on solid phase:
KOtBu (10 equiv), DMSO, 238C, 12 h, then toluene, 1208C, mwave, 25 min, 34% overall yield based on the
loading of thiophenol resin; d) BCl3 (6.0 equiv), CH2Cl2, 08C, 15 min, 82%; e) PS-IBX (3.0 equiv), CH2Cl2,
238C, 1 h, 86%; DIAD=diisopropyl azodicarboxylate, DMSO=dimethylsulfoxide, HFIP=hexafluoroisopro-
panol, IBX=2-iodobenzoicacid, PS=polystyrene.
To access the cis-enone re-
sorcylides with a strategy com-
patible with solid-phase syn-
11494
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 11490 – 11497

Natural Products
FULL PAPER
had previously reported in the context of radicicol A synthe-
sis[37] that this allylic oxidation could be conveniently carried
out with immobilized 2-iodobenzoicacid (IBX).[41] Indeed,
this transformation had precedent in the literature with
Dess–Martin periodinane (DMP).[42] However, for com-
pounds that did not possess a substituent on the aryl ring
ortho to the alkene (such as radicicol A), the oxidation of
the two different diastereoisomers of the allylic alcohol af-
forded different products. While the less polar isomer led
cleanly to the desired compound (LL-Z1640-2) in 86%
yield, the other isomer afforded another mono-oxidation
product with an NMR spectrum consistent with the oxida-
tion of the hydroxyl group g to the alkene. Several other an-
alogues lacking such aryl substituents were also found to
afford different oxidation products for the different diaste-
reoisomers of the allylic alcohol. Similar observations were
noted by Altmann for the synthesis of L-783277, which is
fully saturated at the benzylic position.[43] We then turned
our attention to a selective protection of this allylic hydroxyl
using a strategy that could lead to both LL-Z1640-2 and L-
783277.
As shown in Scheme 8, starting from resin 24, the toluate
position was alkylated with iodide 47 to obtain polymer-
bound intermediate 54 following removal of the PMB (2,3-
dichloro-5,6-dicyanobenzoquinone (DDQ)) and the TMSE
(TBAF). Macrolactonization under Mitsunobu conditions
afforded the macrocycle 55, which was released from the
resin under oxidative condition to afford, after benzoate de-
protection, macrocycle 56. Conversely, polymer-bound mac-
rocycle 55 could be released under reductive conditions
(R3SnH, azobisisobutyronitrile (AIBN)) to get macrocycle
59 after benzoate hydrolysis. For the reductive cleavage, the
use of a fluorous-tagged tin hydride greatly facilitate prod-
uct isolation.[44,45] While the benzoate hydrolysis proceeded
smoothly in solution phase by using sodium hydroxide with-
out any lactone opening, attempts to perform the same reac-
tion on solid phase failed with a variety of nucleophiles
(LiOH, NaOH, NaOMe, NaOBu, NaSMe, hydrazine). A
number of alternative protecting groups for the allylic hy-
droxyl were investigated, such as a 1) PMB, which would re-
quire a selective 14-membered versus ten-membered lacto-
nization, and 2) cinnamoyl, which should allow hydrolysis
under milder conditions, but none of these resulted in pure
final products. Both diastereoisomers of 56 and 59 could
now be oxidized to the desired cis-enone under slightly
more forceful conditions by using DMP in CH2Cl2 under
reflux. There remained to find appropriate conditions to de-
protect the EOM and the acetonide without isomerization
of the cis-enone. While Lett reported the use of TsOH with
moderate success,[46] the outcome of this reaction was too
finicky to be used in the context of a library. Other acids
such as TFA or HFIP were not suitable either; however,
aqueous HF in acetonitrile[47] was found to give very clean
deprotection without any isomerization and had the virtue
that the crude mixture could be directly lyophilized at the
end of the reaction. The para-EOM group was found to be
the most resistant to acidic cleavage and it was found best
Scheme 8. a) 47 (3.0 equiv), LDA (6.0 equiv), THF/HMPA (10:1),
À788C, 20 min; b) DDQ (2.4 equiv), CH2Cl2/H2O (2:1), 238C, 4 h; c)
TBAF (10 equiv), THF, 238C, 6 h; d) DIAD (3.0 equiv), Ph3P (3.0 equiv),
toluene, 238C, 12 h; e) H2O2 (4.0 equiv), CH2Cl2/HFIP (1:1), 238C, 12 h;
toluene, 808C, 12 h, 62% over 6 steps; f) 1% NaOH/MeOH, 238C, 12 h,
76–80%; g) DMP (3.0 equiv), CH2Cl2, 658C, 6 h, 85–87%; h) 40% HF in
CH3CN (1:10), 238C, 7 h, 50%; i) CH2N2 (5.0–10 equiv), Et2O, 23 8C, 6 h,
63–74%; j) F-OctSnH (5.0 equiv), AIBN (cat), toluene, 1508C, mwave,
10 min, 52% over 5 steps; k) DMDO (5.0 equiv), CH3CN, 08C, 1 h,
25%; AIBN=azobisisobutyronitrile, DDQ=2,3-dichloro-5,6-dicyano-
benzoquinone, DIAD=diisopropyl azodicarboxylate, DMDO=dime-
thyldioxirane,
DMP=Dess
Martin
periodinane,
F-Oct3SnH=
tris(1H,1H,2H,2H-perfluorooctyl)tin hydride, HFIP=hexafluoroisopro-
panol, HMPA=hexamethylphosphoramide, LDA=lithium diisopropyl-
amide, TBAF=tetrabutylammonium fluoride, THF=tetrahydrofuran.
not to drive the reaction to completion, but rather stop it
after 7 h which typically afforded 50% of the desired prod-
uct 58 and 61 along with 50% of the corresponding mono-
EOM analogue (not shown). The natural products LL-
Z1640-2 and L-783277 could be obtained by straightforward
diazomethane treatment of compounds 58 and 61, respec-
tively[48] (>90% conversion based on LC/MS, 63-74% iso-
lated yield after HPLC purification). It should be noted that
the final product isolated by HPLC were typically contami-
nated with small amounts (>5%) of a side product that is
tentatively ascribed to the trans-enone isomer. It had been
reported by Lett that LL-Z1640-2 could be regio- and ste-
Chem. Eur. J. 2009, 15, 11490 – 11497
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11495

N. Winssinger et al.
[11] E. Moulin, S. Barluenga, F. Totzke, N. Winssinger, Chem. Eur. J.
2006, 12, 8819–8834.
[12] S. Barluenga, E. Moulin, P. Lopez, N. Winssinger, Chem. Eur. J.
2005, 11, 4935–4952.
[13] S. Barluenga, P. Lopez, E. Moulin, N. Winssinger, Angew. Chem.
2004, 116, 3549–3552; Angew. Chem. Int. Ed. 2004, 43, 3467–3470.
[14] S. Barluenga, P.-Y. Dakas, Y. Ferandin, L. Meijer, N. Winssinger,
Angew. Chem. 2006, 118, 4055–4058; Angew. Chem. Int. Ed. 2006,
45, 3951–3954.
[15] T. Takahashi, K. Kasuga, J. Tsuji, Tetrahedron Lett. 1978, 19, 4917–
4920.
[16] E. La Porta, U. Piarulli, F. Cardullo, A. Paio, S. Provera, P. Seneci,
C. Gennari, Tetrahedron Lett. 2002, 43, 761–766.
[17] L. A. McAllister, S. Brand, R. de Gentile, D. J. Procter, Chem.
Commun. 2003, 2380–2381.
reoselectively converted to hypothemycin by using
mCPBA.[46] Attracted by the fact that DMDO should pro-
vide the same selectivity while not requiring a work-up, we
attempted the selective epoxidation of LL-Z1640-2 with
DMDO, which afforded hypothemycin in 25% isolated
yield at 50% conversion.[48] While the reaction appeared
very clean when the crude reaction mixture was monitored
by LC-MS and NMR spectroscopy, attempts to drive it to
completion resulted in decomposition. Furthermore, it was
found that the epoxide was quite sensitive to acid and par-
tial degradation of the final product was observed upon iso-
lation.
In conclusion, we have extended the scope of the sulfide
linker to access all the functionalities present in the natural
members of the resorcylic acid lactones. The use of the ben-
zylic sulfide was efficient in providing access to both LL-
Z1640-2 and L-783277 by alkylation followed by oxidative
elimination or reductive cleavage respectively. LL-Z1640-2
was converted to hypothemycin with excellent regio and ste-
reoselectivity by using DMDO. The benzylic sulfoxide af-
forded extremely efficient macrocyclization; however, this
strategy was only compatible with the synthesis of macrocy-
cles bearing an alkene at the benzylic position and thus pre-
cluded access to L-783277.
[18] K. C. Nicolaou, J. Pastor, S. Barluenga, N. Winssinger, Chem.
Commun. 1998, 1947–1948.
[19] R. M. Garbaccio, S. J. Stachel, D. K. Baeschlin, S. J. Danishefsky, J.
Am. Chem. Soc. 2001, 123, 10903–10908.
[20] K. Ravikumar, J. Begue, D. Bonnet-Delpon, Tetrahedron Lett. 1998,
39, 3141–3144.
[21] E. Moulin, V. Zoete, S. Barluenga, M. Karplus, N. Winssinger, J.
Am. Chem. Soc. 2005, 127, 6999–7004.
[22] Previously reported not to work, G. Solladie, M. C. Maestro, A.
Rubio, C. Pedregal, M. C. Carreno, J. L. Garcia Ruano, J. Org.
Chem. 1991, 56, 2317–2322.
[23] S. Hayashi, M. Furukawa, Y. Fujino, H. Matsukura, Chem. Pharm.
Bull. 1969, 17, 419–424.
[24] M. L. Quesada, R. H. Schlessinger, J. Org. Chem. 1978, 43, 346–347.
[25] M. Protiva, M. Rajsner, E. Adlerova, V. Seidlova, Z. J. Vejdelek,
Collect. Czech. Chem. Commun. 1964, 29, 2161–2181.
[26] R. G. Dushin, S. J. Danishefsky, J. Am. Chem. Soc. 1992, 114, 655–
659.
Experimental Section
[27] J. Barluenga, J. Gonzalez, M. Garciamartin, P. Campos, G. Asensio,
J. Chem. Soc. Chem. Commun. 1992, 1016–1017.
[28] T. Takahashi, S. Kusaka, T. Doi, T. Sunazuka, S. Omura, Angew.
Chem. 2003, 115, 5388–5392; Angew. Chem. Int. Ed. 2003, 42, 5230–
5234.
The physical characterization and experimental procedures for all new
compounds are given in the Supporting Information.
[29] J. Georgsson, A. Hallberg, M. Larhed, J. Comb. Chem. 2003, 5, 350–
352.
[30] Z. Q. Yang, X. Geng, D. Solit, C. A. Pratilas, N. Rosen, S. J. Dani-
shefsky, J. Am. Chem. Soc. 2004, 126, 7881–7889.
[31] M. H. Bolli, S. V. Ley, J. Chem. Soc. Perkin Trans. 1 1998, 2243–
2246.
Acknowledgements
This work was funded by a grant from the Agence National de la Recher-
che (ANR) and Conectus. An MNRT fellowship (G.V.) and fellowship
from ARC (P.-Y.D.) are also gratefully acknowledged. The authors thank
Monica Trincado, Kerstin Knepper, Mehdi Boulifa and Sylvain Jacques
for preliminary work in the area of resorcylic acid lactones synthesis.
[32] S. V. Ley, I. R. Baxendale, Nat. Rev. Drug Discovery 2002, 1, 573–
586.
[33] C. Morrill, R. H. Grubbs, J. Org. Chem. 2003, 68, 6031–6034.
[34] A. Michrowska, R. Bujok, S. Harutyunyan, V. Sashuk, G. Dolgonos,
K. Grela, J. Am. Chem. Soc. 2004, 126, 9318–9325.
[35] H. C. Brown, C. Subrahmanyam, T. Hamaoka, N. Ravindran, D. H.
Bowman, S. Misumi, M. K. Unni, V. Somayaji, N. G. Bhat, J. Org.
Chem. 1989, 54, 6068–6075.
[36] J. i. Uenishi, R. Kawahama, O. Yonemitsu, J. Tsuji, J. Org. Chem.
1996, 61, 5716–5717.
[37] P. Y. Dakas, S. Barluenga, F. Totzke, U. Zirrgiebel, N. Winssinger,
Angew. Chem. 2007, 119, 7023–7026; Angew. Chem. Int. Ed. 2007,
46, 6899–6902.
[38] M. Scholl, S. Ding, C. W. Lee, R. H. Grubbs, Org. Lett. 1999, 1, 953–
956.
[39] M. Isaka, C. Suyarnsestakorn, M. Tanticharoen, P. Kongsaeree, Y.
Thebtaranonth, J. Org. Chem. 2002, 67, 1561–1566.
[40] K. Yamamoto, K. Biswas, C. Gaul, S. J. Danishefsky, Tetrahedron
Lett. 2003, 44, 3297–3299.
[41] G. Sorg, A. Mengei, G. Jung, J. Rademann, Angew. Chem. 2001, 113,
4532–4535; Angew. Chem. Int. Ed. 2001, 40, 4395–4397.
[42] K. Tatsuta, S. Takano, T. Sato, S. Nakano, Chem. Lett. 2001, 172–
173.
[1] S. Barluenga, P. Dakas, M. Boulifa, E. Moulin, N. Winssinger, C. R.
Chim. 2008, 11, 1306–1317; T. Hofmann, K.-H. Altmann, C. R.
Chim. 2008, 11, 1318–1335.
[2] N. Winssinger, S. Barluenga, Chem. Commun. 2007, 22–36.
[3] S. M. Roe, C. Prodromou, R. OꢂBrien, J. E. Ladbury, P. W. Piper,
L. H. Pearl, J. Med. Chem. 1999, 42, 260–266.
[4] G. Rastelli, R. Rosenfeld, R. Reid, D. V. Santi, J. Struct. Biol. 2008,
164, 18–23.
[5] M. Ohori, T. Kinoshita, S. Yoshimura, M. Warizaya, H. Nakajima,
H. Miyake, Biochem. Biophys. Res. Commun. 2007, 353, 633–637.
[6] J. Zhang, P. L. Yang, N. S. Gray, Nat. Rev. Cancer 2009, 9, 28–39.
[7] T. J. Turbyville, E. M. Wijeratne, M. X. Liu, A. M. Burns, C. J.
Seliga, L. A. Luevano, C. L. David, S. H. Faeth, L. Whitesell, A. A.
Gunatilaka, J. Nat. Prod. 2006, 69, 178–184.
[8] J. Dong, Y. Zhu, H. Song, R. Li, H. He, H. Liu, R. Huang, Y. Zhou,
L. Wang, Y. Cao, K. Zhang, J. Chem. Ecol. 2007, 33, 1115–1126.
[9] S. Wang, Y. Xu, E. A. Maine, E. M. Wijeratne, P. Espinosa-Artiles,
A. A. Gunatilaka, I. Molnar, Chem. Biol. 2008, 15, 1328–1338.
[10] H. Zhou, J. Zhan, K. Watanabe, X. Xie, Y. Tang, Proc. Natl. Acad.
Sci. USA 2008, 105, 6249–6254.
[43] T. Hofmann, K.-H. Altmann, Synlett 2008, 1500–1504.
11496
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 11490 – 11497

Natural Products
FULL PAPER
[44] D. P. Curran, Handbook of Fluorous Chemistry, Wiley-VCH, Weine-
him, 2004, pp. 101–127.
[45] W. Zhang, D. P. Curran, Tetrahedron 2006, 62, 11837–11865.
[46] P. Selles, R. Lett, Tetrahedron Lett. 2002, 43, 4627–4631.
[47] M. Kinoshita, M. Arai, N. Ohsawa, M. Nakata, Tetrahedron Lett.
1986, 27, 1815–1818.
Isaka, NSTDA, Thailand). The spectral characteristics of LL-Z1640-
2 and L-783277 were consistent with the previously reported NMR
data and spectral characteristic of closely related radicicol A.
[48] The spectral characteristics of hypothemycin were identical to those
of natural samples (NMR spectra were kindly provided by Dr.
Received: May 22, 2009
Published online: October 9, 2009
Chem. Eur. J. 2009, 15, 11490 – 11497
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11497