Synthesis of the monomeric counterpart of marinomycin A

Article abstract of DOI:10.1021/jo900945x

(Chemical Equation Presented) An efficient and highly convergent synthesis of the monomeric counterpart of the antitumor-antibiotic marine natural product marinomycin A was achieved by using optically active titanium complexes to control the configuration of the stereogenic centers, a highly stereo- and regioselective cross-metathesis to generate the (E)-configured C20-C21 double bond, and a Horner-Wadsworth-Emmons olefination followed by a Pd-catalyzed Stille cross-coupling to construct the tetraene moiety.

Full text of DOI:10.1021/jo900945x

pubs.acs.org/joc
Synthesis of the Monomeric Counterpart of Marinomycin A
ꢀ
Dominique Amans, Laurianne Bareille, Veronique Bellosta, and Janine Cossy*
Laboratoire de Chimie Organique, ESPCI ParisTech, CNRS, 10 rue Vauquelin,
75231 Paris Cedex 05, France
janine.cossy@espci.fr
Received May 21, 2009
An efficient and highly convergent synthesis of the monomeric counterpart of the antitumor-
antibiotic marine natural product marinomycin A was achieved by using optically active titanium
complexes to control the configuration of the stereogenic centers, a highly stereo- and regioselec-
tive cross-metathesis to generate the (E)-configured C20-C21 double bond, and a Horner-
Wadsworth-Emmons olefination followed by a Pd-catalyzed Stille cross-coupling to construct
the tetraene moiety.
Introduction
et al. have also shown that marinomycin A, the most
abundant and most active of all, was photochemically con-
vertible into an equilibrium mixture of marinomycins A (1),
B (2), and C (3) upon exposure to ambient light (Scheme 1).1
Consequently, the total synthesis of marinomycin A would
constitute a total synthesis of the two others. Due to its
complex molecular architecture and its exceptional biologi-
cal properties, marinomycin A constitutes an ideal target for
total synthesis. To the best of our knowledge, only one total
synthesis has been reported to date from Nicolaou’s group.2
Herein, we wish to report the full account of our efforts that
culminated in a concise and highly convergent synthesis of
the monomeric counterpart of marinomycin A, and our
attempts at dimerization.3
Oceans provide an extraordinary source of biologically
active natural products. Marinomycins A-C are three
polyenic macrodiolides isolated in 2006 by Fenical et al.
from the saline culture of a new group of marine actinomy-
cetes, Marinispora strain CNQ-140, cultured from a sedi-
ment collected from the bottom of the ocean offshore of La
Jolla, CA (USA).1 These natural products display significant
antibiotic activities, with MIC values of 0.1-0.6 μM, against
methicillin-resistant Staphylococcus aureus and vancomycin-
resistant Enterococcus faecium. Further studies also revealed
that marinomycins A-C demonstrated impressive and se-
lective cancer cell cytotoxicities against six of the eight
melanoma cell lines in the National Cancer Institute’s 60
cell line panel (IC50 = 0.2-2.7 μM). Interestingly, only
marinomycin A, C2-symmetric, showed antifungal activity
against Candida albicans (MIC90 = 10 μM). The molecular
architecture of these three 44-membered dimeric macrolides
is characterized by the presence of a tetraene conjugated with
an aromatic unit derived from 2-hydroxybenzoic acid and
connected to a pentahydroxylated polyketide chain. Fenical
Results and Discussion
In our initial endeavor to synthesize marinomycin A (1), we
opted for a dimerization strategy as depicted in Scheme 2, and
we therefore concentrated our efforts toward the synthesis of
(2) (a) Nicolaou, K. C.; Nold, A. L.; Milburn, R. R.; Schindler, C. S.
Angew. Chem., Int. Ed. 2006, 45, 6527–6532. (b) Nicolaou, K. C.; Nold, A. L.;
Milburn, R. R.; Schindler, C. S.; Cole, K. P.; Yamaguchi, J. J. Am. Chem.
Soc. 2007, 129, 1760–1768.
(3) For a preliminary communication, see: Amans, D.; Bellosta, V.;
Cossy, J. Org. Lett. 2007, 9, 1453–1456.
(1) (a) Kwon, H. C.; Kauffman, C. A.; Jensen, P. R.; Fenical, W. J. Am.
Chem. Soc. 2006, 128, 1622–1632. (b) Erratum: Kwon, H. C.; Kauffman, C. A.;
Jensen, P. R.; Fenical, W. J. Am. Chem. Soc. 2006, 128, 16410.
DOI: 10.1021/jo900945x
r
Published on Web 09/21/2009
J. Org. Chem. 2009, 74, 7665–7674 7665
2009 American Chemical Society

Amans et al.
JOCArticle
SCHEME 1
SCHEME 2
SCHEME 3
a palladium-catalyzed Suzuki cross-coupling between aryl
triflate 4 and vinylboronic acid 5 (Scheme 3).
The coupling between aryl triflate 44 and vinyl boronic
acid 55 turned out to be not as trivial as it seemed, partly due
to significant decomposition of the R,β-unsaturated alde-
hyde generated in the course of the reaction. Nevertheless,
scrupulous optimization of the reaction conditions allowed
us to perform the envisaged Suzuki coupling using K3PO4 as
a base in 1,4-dioxane, in the presence of Pd(PPh3)4, under
microwave irradiation (130 °C, 15 min) to give the required
aldehyde 6 in 73% yield.
its monomericcounterpartA, which contains all the structural
and stereochemical features of the natural product.
Treatment of aldehyde 6 by the phosphonium ylide gener-
ated in situ upon treatment of commercially available (3-tri-
methylsilyl-2-propynyl)triphenylphosphonium bromide with
The monomer A could be accessed by performing a
palladium-catalyzed cross-coupling between the trienic
vinylmetal of type B and vinyl iodide C to create the
C13-C14 bond, while the addition of the lithium acetylide
of alkyne D onto β-hydroxyaldehyde C would allow the
formation of the C19-C20 bond.
Synthesis of the Boronic Ester B1 [M = B(OR)2]. The syn-
thesis of boronic ester B1 was envisioned via the hydrobora-
tion of enyne 8. The latter would be obtained from the
olefination of aldehyde 6, which in turn would result from
(4) Arylic triflate4 wassynthesized in two steps from commercially available
2,6-dihydroxybenzoic acid according to the following: (a) Hadfield, A.;
Schweitzer, H.; Trova, M. P.; Green, K. Synth. Commun. 1994, 24, 1025–1028.
(b) Dushin, R. G.; Danishefsky, S. J. J. Am. Chem. Soc. 1992, 114, 655–659.
(5) Boronic acid 5 was prepared in two steps via hydroboration of
commercially available propiolaldehyde dimethylacetal according to the
ꢀ
following: Toure, B. B.; Hoveyda, H. R.; Tailor, J.; Ulaczyk-Lesanko, A.;
Hall, D. G. Chem. Eur. J. 2003, 9, 466–474.
7666 J. Org. Chem. Vol. 74, No. 20, 2009

Amans et al.
JOCArticle
SCHEME 4
SCHEME 5
SCHEME 6
n-BuLi resulted in the formation of enyne 7 as a 75/25
inseparable mixture of E/Z geometric isomers, in 58% yield.
Replacing n-BuLi by sodium bis(trimethylsilyl)amide
(NaHMDS) slightly improved the E/Z ratio to 82/18 in favor
of the desired E-isomer and the reaction proceeded in similar
yield (Scheme 4).
At this stage, we believed that the E/Z ratio could be
improved by irradiation of compound 7 with visible light.
Against all expectations, exposure of enyne 7 to day light in
the presence of a catalytic amount of iodine led to an increase
of the proportion of the Z-isomer (E/Z ratio = 1:1). Being
aware of the fact that the isomerization of the polyenic moiety
couldin principle be conducted at a later stage ofthe synthesis,
we pursued the preparation of the targeted trienic vinyl
boronic ester 9 with the mixture of E/Z-isomers of enyne 7.
After treatment of the latter with tetra-n-butylammonium
fluoride (TBAF) at 0 °C in THF, the resulting terminal alkyne
8 underwent a chemo- and stereoselective hydroboration with
catecholborane in THF to furnish vinylboronic ester 9 as a
mixture ofE/Z-isomers(Scheme5). Unfortunately, numerous
attempts to purify this unstable boronic ester both by flash
chromatography on silica gel and by filtration over Florisil led
to the degradation of the product.
allylphosphonate 12 and the R,β-unsaturated aldehyde 13
(Scheme 6).
The preparation of phosphonate 12 was achieved
by performing a palladium-catalyzed Stille cross-coupling
between (E)-3-(tributylstannyl)prop-2-en-1-ol 106 and aryl
triflate 4 [PdCl2(PPh3)2, tri-2-furylphosphine, DMF,
60 °C], which furnished allylic alcohol 11 in 82% yield.7
Subsequent bromination of the latter by PBr3, followed by
exposure of the resulting allylic bromide to Arbuzov condi-
tions [P(OEt)3, toluene] provided the required diethyl allyl-
phosphonate 12 in excellent yield. Olefination between
Having experienced difficulties to obtain the required
vinylboronic ester 9 of type B1 in a satisfactory pure form,
we decided to explore another route toward the synthesis of
fragment B, using a vinyl stannane.
Synthesis of the Vinyl Stannane B2 [M = SnBu3]. We be-
lieved that trienic vinyl stannane B2 could result from
a Horner-Wadsworth-Emmons olefination between
(6) (E)-3-(Tributylstannyl)prop-2-en-1-ol 10 was synthesized by perform-
ing a stereo- and regioselective stannylcupration of commercially available
prop-2-yn-1-ol, using the mixed higher order cuprate [(Bu3Sn)BuCu(CN)Li2]
generated in situ according to the following: Lipshutz, B. H.; Ellsworth,
E. L.; Dimock, S. H.; Reuter, D. C. Tetrahedron Lett. 1989, 30, 2065–2068.
(7) Wang, X.; Bowman, E. J.; Bowman, B. J.; Porco, J. A. Jr. Angew.
Chem., Int. Ed. 2004, 43, 3601–3605.
J. Org. Chem. Vol. 74, No. 20, 2009 7667

Amans et al.
JOCArticle
SCHEME 7
conditions, the homoallylic alcohol 17 with high enantio
selectivity (95% ee).13 Regioselective oxidative cleavage of the
terminal double bond (OsO4, NMO then NaIO4) ultimately
furnished the desired aldehyde 18, which was not purified but
directly used in the subsequent alkynylation step (vide infra)
(Scheme 8).
The last fragment, alkyne D, needed to be synthesized. Its
preparation was first planned via opening of the optically active
terminal epoxide 22, which in turn could derive from the com-
mercially available (R)-(-)-3-hydroxybutyrate 19 (Scheme 9).
SCHEME 8
the previously obtained allylic phosphonate 12 and
R,β-unsaturated aldehyde 13 was attempted. Addition
of aldehyde 138 to a solution of the sodium salt of allylic
phosphonate 12 led to the desired trienic vinyl stannane
14 in 53% yield (Scheme 7).9
Having now efficiently obtained the required trienic stan-
nane 14 (of type B2), we turned our attention toward the
synthesis of vinyl iodide C.
Synthesis of Vinyl Iodide C. The synthesis of vinyl iodide C
commenced from the regio- and stereoselective carboalumina-
tion of homopropargyl alcohol 15 catalyzed by zirconium
(Cp2ZrCl2, Me3Al, H2O)10,11 followed by a metal-halogen
exchange with iodine to give, after oxidation with Dess-Martin
periodinane (DMP), the E-configured trisubstituted vinyl
iodide 16 (77% yield over three steps). Aldehyde 16 was
subsequently treated by the highly face-selective Hafner-
Duthaler complex12 (S,S)-I to provide, after protection of the
resulting secondary alcohol as a PMB-ether under acidic
SCHEME 9
The (R)-3-hydroxybutyrate 19 was first protected as
a TBS-ether and subsequently reduced to provide the
corresponding aldehyde,14 which was directly submitted to
enantioselective allyltitanation by using the (S,S)-I allyltita-
nium complex to furnish, after protection as a TES-ether
(TESCl, Et3N, DMAP, CH2Cl2), the optically active homo-
allylic alcohol 20 with excellent diastereomeric purity
(dr >95/5) and in good overall yield (67% over four steps).
Sharpless asymmetric dihydroxylation15 of alkene 20 with
AD-mix-β proceeded uneventfully to furnish the 1,2-diol 21
in 56% yield as a 75/25 mixture of inseparable diastereomers
in favor of the desired diastereoisomer 21. Subsequent
treatment of this mixture of 1,2-diols by NaH and tosylimi-
dazole successfully provided the optically active epoxide 22
in 83% yield as an inseparable 75/25 mixture of diastereo-
isomers. Oxirane 22 was then exposed to the lithium salt of
(8) β-Stannylacrolein 13 was synthesized by performing a regio- and
stereoselective stannylcupration of commercially available propiolaldehyde
diethylacetal followed by acetal hydrolysis according to the following:
(a) Lipshutz, B. H.; Lindsley, C. J. Am. Chem. Soc. 1997, 119, 4555–4556.
(b) Lipshutz, B. H.; Ullman, B.; Lindsley, C.; Pecchi, S.; Buzard, D. J.;
Dickson, D. J. Org. Chem. 1998, 63, 6092–6093. (c) Beaudet, I.; Parrain,
J.-L.; Quintard, J.-P. Tetrahedron Lett. 1991, 32, 6333–6336.
(9) It should be noted that aldehyde 13 needs to be added right away after
the deprotonation of phosphonate 12 by NaH, to prevent the competitive
formation of the gem-dimethylated olefin 42, hardly separable from the
desired trienic vinylstannane 14 by flash chromatography on silica gel. The
diene 42 is the result of the condensation between the sodium salt of
phosphonate 12 with acetone, presumably generated from the attack of an
unidentified nucleophile in the reaction mixture (maybe the sodium salt of
phosphonate 12 itself), on the carbonyl group of the aryl ester 12.
trimethylsilylacetylene in the presence of BF3 Et2O in THF,
3
yielding the expected homopropargylic alcohol 23 (77%
yield). Protection of the latter as a TBS-ether followed by
(10) (a) Van Horn, D. E.; Negishi, E. J. Am. Chem. Soc. 1978, 100, 2252–
2254. (b) Negishi, E.; Van Horn, D. E.; Yoshida, T. J. Am. Chem. Soc. 1985,
107, 6639–6647.
(13) The (R) absolute configuration of 17 was confirmed by examination
of the 1H NMR spectra of the two corresponding mandelates, following the
~
procedure described by the following: Seco, J. M.; Quinoa, E.; Riguera, R.
(11) (a) Wipf, P.; Lim, S. Angew. Chem., Int. Ed. Engl. 1993, 32, 1068–
1071. (b) Marshall, J. A.; Eidam, P. Org. Lett. 2004, 6, 445–448.
(12) (a) Hafner, A.; Duthaler, R. O.; Marti, R.; Rihs, J.; Rothe-Streit, P.;
Schwarzenbach, F. J. Am. Chem. Soc. 1992, 114, 2321–2336. (b) Duthaler,
R. O.; Hafner, A. Chem. Rev. 1992, 92, 807–832. (c) Cossy, J.; BouzBouz, S.;
Pradaux, F.; Willis, C.; Bellosta, V. Synlett 2002, 1595–1606.
Tetrahedron: Asymmetry 2001, 12, 2915–2925.
(14) Wattanasereekul, S.; Maier, M. E. Adv. Synth. Catal. 2004, 346, 855–
861.
(15) For a review on Sharpless catalytic asymmetric dihydroxylation, see:
Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. Rev. 1994,
2483–2547.
7668 J. Org. Chem. Vol. 74, No. 20, 2009

Amans et al.
JOCArticle
SCHEME 10
SCHEME 12
SCHEME 13
SCHEME 11
acetal 25 and the β-hydroxyaldehyde derived from ester 19,
while a diasteroselective allyltitanation of the latter fol-
lowed by oxidative cleavage of the resulting homoallylic
alcohol 20 would also provide the required aldehyde 28
(Scheme 11).
The Mukaiyama-aldol strategy was first explored. Thus,
after protection and reduction of commercially available
β-hydroxyester 19 of (R)-configuration with DIBAL-H,
the resulting β-hydroxyaldehyde was subjected to the
addition of the silylketene acetal derived from tert-butyl
thioacetate 25, in the presence of 2.5 equiv of MeAlCl2
(-78 °C in CH2Cl2), which gave rise to 1,3-diol 26 as a
80/20 mixture of inseparable diastereoisomers in favor of the
desired anti 1,3-diol (82% yield).16 Subsequent protection of
the secondary alcohol as a TES-ether followed by reduction
of the tert-butyl thioester upon treatment with DIBAL-H in
toluene led to aldehyde 28 in 99% yield (anti/syn = 80/20).
At this stage, the two epimers could be easily separated
selective desilylation of the alkyne upon treatment with
potassium carbonate in MeOH ultimately furnished the
desired alkyne 24 (74% yield), unfortunately contaminated
with 25% of its epimer at C23 (Scheme 10).
As we were not able to separate the mixture of epimers
at C23 of alkyne 24, another synthetic route was investi-
gated for the diastereoselective preparation of the required
1,3,5-triol of type D.
Alkyne D could alternatively be accessed from aldehyde
28, which could be obtained via two distinct synthetic
pathways. A first possibility would involve a diastereoselec-
tive Mukaiyama-type aldol reaction between silylketene
(16) Evans, D. A.; Allison, B. D.; Yang, M. G.; Masse, C. E. J. Am. Chem.
Soc. 2001, 123, 10840–10852.
J. Org. Chem. Vol. 74, No. 20, 2009 7669

Amans et al.
JOCArticle
SCHEME 14
SCHEME 15
SCHEME 16
SCHEME 17
by flash chromatography on silica gel and the desired
optically pure anti 1,3-diol 28 was isolated in 70% yield
(Scheme 12).
The second approach, which turned out to be equally
efficient in providing the anti 1,3-diol 28, involved an oxida-
tive cleavage of the terminal double bond of the previously
obtained diol 20 to give the desired aldehyde 28. The latter
was directly treated with allyltitanium complex (R,R)-I to
furnish, after protection of the resulting homoallylic alcohol
as a TBS-ether (TBSCl, imidazole, DMF),17 the protected
1,3,5-triol 29 (65% yield over three steps). Oxidative cleavage
of the terminal double bond (OsO4, NaIO4, 2,6-lutidine)18
provided the corresponding aldehyde, which was subse-
quently transformed into the corresponding gem-dibromo
olefin under Corey-Fuchs conditions (PPh3, CBr4, Zn, py).
Subjection of the obtained gem-dibromoalkene to 2 equiv of
n-BuLi allowed for the Fritsch-Buttemberg-Wiechell re-
arrangement to proceed and ultimately provided the desired
alkyne 24 in 87% yield (Scheme 13).
by n-BuLi and the resulting lithium acetylide was then
condensed onto aldehyde 18, in THF at -78 °C, to give
propargylic alcohol 30 as a mixture of two epimers at C19 in
18% yield. Several experiments were performed with the aim
of improving the yield, to no avail.19
To lower the basicity of the lithium acetylide while at the
same time increasing its reactivity, the alkynyl lithium species
generated in situ was transmetalated to the corresponding
organocerium reagent upon exposure to cerium trichloride.
Unfortunately, subsequent addition of aldehyde 18 (-78 °C,
THF) failed to provide the desired propargylic alcohol 30
(Scheme 14).
With the three fragments 14 (of type B), 18 (of type C), and
24 (of type D) in hand, the coupling reactions could be
implemented to complete the synthesis of the monomeric
counterpart of marinomycin A. Alkyne 24 was deprotonated
(17) It should be noted that the utilization of TBSOTf and 2,6-lutidine led
to migrations of the silyl groups from one oxygen to another.
(18) Yu, W.; Mei, Y.; Kang, Y.; Hua, Z.; Jin, Z. Org. Lett. 2004, 6, 3217–
3219.
(19) This coupling was also attempted by replacing aldehyde 18 by the
corresponding Weinreb amide, but under these conditions no trace of the
desired acetylenic ketone could be detected.
7670 J. Org. Chem. Vol. 74, No. 20, 2009

Amans et al.
JOCArticle
SCHEME 18
The lack of success associated with this strategy led us to
investigate a revised approach for the synthesis of the
monomeric counterpart of marinomycin A.
SCHEME 19
Revised Retrosynthetic Analysis. We believed that the
monomeric counterpart of marinomycin A could alterna-
tively be accessed via a stereo- and chemoselective cross-
metathesis between β-hydroxy vinyl ketone 31 and olefin 29
to form the E-configured C20-C21 double bond. As pre-
viously, the vinyl iodide 31 would be connected to vinyl
stannane 14 through a Pd-catalyzed Stille cross-coupling
(Scheme 15).
Having previously obtained the trienic vinylstananne 14 as
well as 1,3,5-triol 29, we concentrated our efforts on the
enantioselective synthesis of β-hydroxy vinyl ketone 31. The
most direct access to this compound would be to perform an
enantioselective addition of methyl vinyl ketone (MVK)
enolate to aldehyde 16. This transformation, although very
attractive, is nevertheless extremely challenging due to the base
instability of MVK and the aldol product generated through-
out the reaction under basic conditions. To the best of our
knowledge, only Trost et al. were able to perform such a
transformation by using their dinuclear zinc catalyst II.20,21
We consequently decided to attempt this methodology to
synthesize the targeted β-hydroxy ketone 31. Thus, aldehyde
16 was treated with a large excess of MVK (24 equiv), in the
presence of dinuclear zinc catalyst II of (S,S)-configuration
˚
and molecular sieves 4 A in THF at -35 °C, which unfortu-
nately resulted in the obtention of a complex mixture of
products in which the desired β-hydroxy ketone 31 could
not be observed. Presumably, the relatively basic conditions
employed led to the formation of elimination products
and prevented the isolation of the targeted enone 31
(Scheme 16).
(20) Trost, B. M.; Ito, H. J. Am. Chem. Soc. 2000, 122, 12003–12004.
(21) Trost, B. M.; Shin, S.; Sclafani, J. A. J. Am. Chem. Soc. 2005, 127,
8602–8603.
J. Org. Chem. Vol. 74, No. 20, 2009 7671

Amans et al.
JOCArticle
SCHEME 20
We consequently resorted to more classical transforma-
tions that involved manipulation of the previously obtained
optically active β-hydroxyaldehyde 18 as depicted in
Scheme 17. Thus, addition of vinylmagnesium bromide onto
aldehyde 18, followed by oxidation of the resulting epimeric
mixture by pyridinium chlorochromate (PCC) allowed the
formation of vinyl ketone 32 in 56% yield (over two steps).
Exposure of the latter to 2,3-dichloro-5,6-dicyano-1,4-ben-
zoquinone (DDQ) ultimately furnished the desired β-hydro-
xy ketone 31 in 71% yield (Scheme 17).
Completion of the Synthesis of the Monomeric Counterpart
of Marinomycin A. At this stage, we attempted to couple
alkenyl ketone 31 with terminal alkene 29 by using a ru-
thenium-catalyzed cross-metathesis reaction. Gratifyingly,
addition of 10 mol % of the Grubbs-Hoveyda second
generation catalyst Ru-III22 to an equimolar mixture of vinyl
ketone 31 and olefin 29 successfully generated enone 33 in
66% yield and with total E-selectivity. Subsequent syn-
reduction23 of β-hydroxy ketone 33 occurred uneventfully
to produce 1,3-syn-diol 34 as a single diastereoisomer
(dr > 98/2) in 91% yield. Protection of the two hydroxy
groups at C17 and C19 as tert-butyldimethylsilyl ethers
(TBSOTf, 2,6-lutidine) gave rise to the fully protected pen-
ta-alkoxylated alkenyliodide35 almost quantitatively. Finally,
a Pd-catalyzed Stille cross-coupling24 between trienic vinyl
stannane 14 and alkenyl iodide 35 (Pd2dba3, triphenylarsine,
DMF) successfully furnished the monomeric counterpart of
marinomycin A 36 (of type A) in 59% yield (Scheme 18).
To perform the envisaged dimerization reaction, the hy-
droxy group at C25 protected as a TES-ether needed to be
selectively removed.
pyridinium p-toluenesulfonate (PPTS) in EtOH, which ulti-
mately furnished the desired alcohol 39, albeit in low yield
due to significant decomposition (Scheme 19).
Several attempts at saponification of the methyl ester 39
were tested in order to provide the targeted carboxylic acid
40. The presence of silyl ethers required the utilization of
mild basic conditions, to avoid any fortuitous deprotection.
Thus, methyl ester 39 was treated with potassium trimethyl-
silanolate, in ether at rt, which unfortunately failed in giving
the desired carboxylic acid 40 (the starting material was
recovered). Using an excess of lithium hydroxide (20 equiv)
also met with failure, since it led to the deprotection of the
phenol, while the methyl ester remained intact (Scheme 20).
Unfortunately, despite numerous efforts, we were not able to
obtain the targeted hydroxyacid 40. It may be necessary to
change the nature of the protecting groups to achieve the
proposed dimerization.
Conclusion
In conclusion, the monomeric counterpart of marinomy-
cin A 36, which incorporates all the structural and stereo-
chemical features of the natural product, was efficiently
synthesized by using a highly convergent approach, in 11
steps from commercially available ethyl (R)-3-hydroxybuty-
rate 19 for the longest linear sequence and in 15% overall
yield.
Major highlights of this synthetic endeavor include highly
enantioselective allyltitanations to introduce the stereogenic
centers at C17, C23, and C25, a stereo- and chemoselective
cross-metathesis to form the C20-C21 double bond of
E-configuration and a Pd-catalyzed Stille cross-coupling to
elaborate the very sensitive conjugated tetraenic moiety. A di-
merization should in principle lead to the extremely scarce but
exceptionally potent natural product marinomycin A.
Due to the intrinsic instability of the acetonide functiona-
lity under the acidic conditions required to perform the
aforementioned deprotection, compound 36 was subjected
to methanolysis by using sodium methanolate in MeOH, to
afford the corresponding methyl ester 37 in 86% yield.
Subsequent protection of the resulting phenol as a triisopro-
pylsilyl ether (TIPS) furnished compound 38. To achieve the
selective deprotection of the TES-ether at C25 in view of
the subsequent dimerization, polyol 38 was treated by
Experimental Section
5-[(1E,3E,5E)-6-(Tributylstannyl)hexa-1,3,5-trienyl]-2,2-di-
methyl-4H benzo[1,3]dioxin-4-one (14). To a stirred suspension of
sodium hydride (60% in oil, 46 mg, 1.16 mmol, 4.0 equiv) in
anhydrous THF (2 mL) at 0 °C was added a solution of
phosphonate 12 (205 mL, 0.58 mmol, 2.0 equiv) in THF (2 mL)
dropwise. The resulting deep purple solution was cooled to
-20 °C and a solution of aldehyde 6 (100 mg, 0.29 mmol, 1.0
equiv) in THF (1 mL) was added dropwise. The resulting mixture
was warmed to 0 °C and stirred for 15 min, and was then allowed
to warm to rt. After 30 min of stirring, the resulting black solution
was hydrolyzed by adding a saturated aqueous solution of
(22) (a) Kingsbury, J. S.; Harrity, J. P. A.; Bonitatebus, P. J. Jr.; Hoveyda,
A. H. J. Am. Chem. Soc. 1999, 121, 791–799. (b) Chatterjee, A. K.; Choi,
T.-L.; Sanders, D. P.; Grubbs, R. H. J. Am. Chem. Soc. 2003, 125, 11360–
11370.
ꢁ
(23) Chen, K.-M.; Hardtmann, G. E.; Prasad, K.; Repic, O.; Shapiro,
M. J. Tetrahedron Lett. 1987, 28, 155–158.
(24) (a) Milstein, D.; Stille, J. K. J. Am. Chem. Soc. 1978, 100, 3636–3638.
(b) Stille, J. K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508–524. (c) Coleman,
R. S.; Walczak, M. C. Org. Lett. 2005, 7, 2289–2291.
7672 J. Org. Chem. Vol. 74, No. 20, 2009

Amans et al.
JOCArticle
NaHCO3 (10 mL). The layers were separated and the aqueous
layerwas extractedwithEt2O (3 ꢀ 15 mL). The combined organic
layers were washed with brine (15 mL), dried over MgSO4,
filtered, and concentrated in vacuo. Purification of the residue
by flash chromatography on silica gel (petroleum ether/EtOAc
98/2) afforded the trienic stannane 14 (83 mg, 53%) as a viscous
yellow oil that was prone to decomposition upon storage at
-20 °C: Rf 0.5 (petroleum ether/EtOAc 98/2); IR (neat) 2920,
2847, 1734, 1593, 1573, 1473, 1316, 1265, 1209, 1043, 1003, 688,
666 cm-1; 1H NMR (400 MHz, CDCl3) δ 7.75 (d, 1H, J = 15.5
Hz), 7.43 (tapp, 1H, J = 8.0 Hz), 7.33 (d, 1H, J = 7.8 Hz),
6.87-6.80(m, 2H), 6.65(dd, 1H, J = 18.3, 9.8Hz), 6.50-6.32(m,
3H), 1.71 (s, 6H), 1.57-1.47 (m, 6H), 1.35-1.25 (m, 6H),
0.95-0.86 (m, 15H); 13C NMR (100 MHz, CDCl3) δ 160.4 (s),
157.0 (s), 146.8 (d), 141.7 (s), 137.7 (d), 137.2 (d), 135.0 (d), 133.5
(d), 133.0 (d), 130.3 (d), 120.6 (d), 115.9 (d), 110.5 (s), 105.2 (s),
29.1 (3t), 27.3 (3t), 25.6 (2q), 13.7 (3q), 9.7 (3t); HRMS (ESI)
calcd for C28H42NaO3Sn þ Naþ 569.2054, found 569.2050.
(E)-(S)-5-Hydroxy-8-iodo-7-methylocta-1,7-dien-3-one (31).
To a stirred solution of vinyl ketone 32 (1.0 g, 2.49 mmol, 1.0
equiv) in a CH2Cl2/H2O mixture (9/1, 35 mL) was added
2,3-dichloro-5,6-dicyanoquinone (680 mg, 2.99 mmol, 1.2
equiv) at rt. After 45 min of stirring, the resulting deep orange
heterogeneous solution was quenched by adding a saturated
solution of Na2SO3 (20 mL). The layers were separated and the
aqueous phase was extracted with EtOAc (3 ꢀ 30 mL). The
combined organic layers were washed with brine (50 mL), dried
over MgSO4, filtered, and concentrated in vacuo. Purification of
the residue by flash chromatography on silica gel (petroleum
ether/AcOEt 95/5 to 80/20) provided the desired β-hydroxy
vinyl ketone 31 (498 mg, 71%) as a colorless oil: Rf 0.2
(1E,7E)-(4S,6S,10S,12S,14R)-10,14-Bis(tert-butyldimethylsi-
lyloxy)-1-iodo-2-methyl-12-triethylsiloxypentadeca-1,7-dien-4,
6-diol (34). To a stirred solution of β-hydroxy ketone 33 (75 mg,
0.10 mmol, 1.0 equiv) in a THF/MeOH mixture (4/1, 2 mL) at
-78 °C under argon was added dropwise methoxydiethylborane
(1 M in THF, 111 μL, 0.11 mmol, 1.1 equiv) and the resulting
mixture was stirred for 15 min. Sodium borohydride (5 mg, 0.11
mmol, 1.1 equiv) was added in one portion, and the mixture was
stirred for 6 h at -78 °C. The reaction was quenched by adding
acetic acid (1 mL), diluted with ethyl acetate (10 mL), and the
resulting mixture was successively washed with a saturated
aqueous Na2CO3 solution (10 mL) and brine (10 mL). The
organic layer was dried over MgSO4, filtered, and evaporated
to dryness under reduced pressure. The residue thus obtained was
azeotroped 5 times with methanol until the hydrolysis of the
boronate was complete, and then purified by flash chromatogra-
phy on silica gel (petroleum ether/EtOAc 95/5 to 90/10), which
provided the desired syn-1,3-diol 34 (69 mg, 91%) as a colorless
oil: Rf 0.2 (petroleum ether/EtOAc 90/10); [R]20 - 4.5 (c 1.1,
D
CHCl3); IR (neat) 3352, 2952, 2928, 2856, 1616, 1462, 1377, 1252,
1067, 1004, 833, 772, 725, 668 cm-1; 1H NMR (400 MHz, CDCl3)
δ 5.94 (s, 1H), 5.63 (dt, 1H, J = 15.5, 7.0 Hz), 5.45 (dd, 1H, J =
15.5, 6.9 Hz), 4.27 (qapp, 1H, J = 6.9 Hz), 3.95 (quintapp, 1H, J =
7.0 Hz), 3.83 (sextapp, 1H, J = 6.0 Hz), 3.78-3.68 (m, 2H), 2.35
(dd, 1H, JsystAB = 13.5, 7.4 Hz), 2.26 (dd, 1H, JsystAB = 13.5, 5.0
Hz), 2.21-2.07 (m, 2H), 1.82 (br s, 3H), 1.59-1.48 (m, 5H), 1.43
(m, 1H), 1.08 (d, 3H, J = 6.1 Hz), 0.89 (t, 9H, J = 8.0 Hz), 0.82(s,
9H), 0.81 (s, 9H), 0.53 (q, 6H, J = 8.0 Hz), 0.00 (s, 12H); 13C
NMR (100 MHz, CDCl3) δ 144.6(s), 134.7 (d), 128.3 (d), 77.5(d),
73.6 (d), 69.5 (2d), 67.6 (d), 66.0 (d), 48.4 (t), 47.8 (t), 45.9 (t),
42.7 (t), 40.3 (t), 25.9 (6q), 24.4 (q), 24.2 (q), 18.1 (2s), 7.1 (3q),
5.7 (3t), -3.9, -4.1, -4.3, -4.5 (4q); HRMS (ESI) calcd for
C34H71IO5Si3 þ Naþ 793.3552, found 793.3546.
(petroleum ether/EtOAc 80/20); [R]20 þ27.8 (c 1.0, CHCl3);
D
IR (neat) 3423, 2909, 1672, 1612, 1401, 1272, 1047, 964, 766, 668
cm-1; 1H NMR (400 MHz, CDCl3) δ 6.29 (dd, 1H, J = 17.7,
10.3 Hz), 6.18 (dd, 1H, J = 17.7, 1.0 Hz), 5.97 (br s, 1H), 5.85
(dd, 1H, J = 10.3, 1.0 Hz), 4.21 (m, 1H), 2.92 (br s, 1H, OH),
2.73-2.58 (m, 2H), 2.40 (dd, 1H, JsystAB = 13.5, 7.5 Hz), 3.29
(dd, 1H, JsystAB = 13.5, 5.4 Hz), 1.82 (br s, 3H); 13C NMR (100
MHz, CDCl3) δ 200.7 (s), 144.5 (s), 136.6 (d), 129.5 (t), 77.7 (d),
65.5 (d), 46.3 (t), 45.1 (t), 24.2 (q); HRMS (ESI) calcd for
C9H13O2I þ Naþ 302.9858, found 302.9852.
(1E,7E)-(4S,6S,10S,12S,14R)-4,6,10,14-Tetrakis(tert-butyl-
dimethylsilyloxy)-1-iodo-2-methyl-12-triethylsiloxypentadeca-1,
7-diene (35). To a stirred solution of 34 (153 mg, 0.20 mmol, 1.0
equiv) in CH2Cl2 (5 mL) at -78 °C were added 2,6-lutidine
(180 μL, 0.79 mmol, 4 equiv) and tert-butyldimethylsilyl trifluoro-
methanesulfonate (115 μL, 0.99 mmol, 5 equiv). After 30 min of
stirring at -78 °C, a saturated aqueous solution of NH4Cl (3 mL)
was added. The layers were separated and the aqueous layer was
extracted with Et2O (3 ꢀ 10 mL). The combined organic layers
were washed with brine (10 mL), dried over MgSO4, filtered, and
concentrated in vacuo. Purification of the residue by flash
chromatography on silica gel (petroleum ether/EtOAc 99/1)
provided the desired fully protected polyol 35 (193 mg, 98%) as
a viscous colorless oil: Rf 0.9 (petroleum ether/EtOAc 95/5);
(1E,7E)-(4S,10S,12S,14R)-10,14-Bis(tert-butyldimethylsilyl-
oxy)-4-hydroxy-1-iodo-2-methyl-12-triethylsilyloxypentadeca-1,
7-dien-6-one (33). To a stirred solution of olefin 29 (90 mg, 0.17
mmol, 1.0 equiv) and β-hydroxy vinyl ketone 31 (49 mg, 0.17
mmol, 1.0 equiv) in anhydrous CH2Cl2 (2 mL) at rt was added
Grubbs-Hoveyda catalyst Ru-III (11 mg, 0.017 mmol, 0.1
equiv). The resulting reaction mixture was warmed to 30 °C
and stirred at this temperature for 15 h. The reaction mixture was
cooled to rt and concentrated in vacuo. Purification of the residue
by flash chromatography on silica gel (petroleum ether/EtOAc
95/5 to 90/10) provided the desired enone 33 (89 mg, 66%) as a
colorless oil: Rf 0.5 (petroleum ether/EtOAc 90/10); [R]20D þ 7.1
(c 0.60, CHCl3); IR (neat) 3460, 2952, 2928, 2879, 2856, 1666,
1627, 1471, 1462, 1376, 1361, 1253, 1068, 1005, 834, 773, 739, 725,
669 cm1; 1H NMR (400 MHz, CDCl3) δ 6.82 (dt, 1H, J = 16.0,
7.0 Hz), 6.05 (d, 1H, J = 16.0 Hz), 5.96 (br s, 1H), 4.19 (m, 1H),
3.83 (m, 2H), 3.74 (quintapp, 1H, J = 7.0 Hz), 3.12 (br s, 1H, OH),
2.65 (dd, 1H, JsystAB = 17.5, 3.1 Hz), 2.53(dd, 1H, JsystAB = 17.5,
8.6 Hz), 2.41 (m, 2H), 2.26 (m, 2H), 1.82 (br s, 3H), 1.61-1.39 (m,
4H), 1.09 (d, 3H, J = 6.1 Hz), 0.89 (t, 9H, J = 8.0 Hz), 0.82 (s,
[R]20 - 8.3 (c 0.72, CHCl3); IR (neat) 2953, 2928, 2885, 2856,
D
1471, 1462, 1377, 1361, 1252, 1068, 1004, 833, 772 cm-1 1H
;
NMR (400 MHz, CDCl3) δ 5.84 (s, 1H), 5.50 (m, 1H), 5.35 (dd,
1H, J = 15.5, 6.9 Hz), 4.07 (m, 1H), 3.86-3.76 (m, 4H), 2.35 (dd,
1H, J = 13.4, 4.2 Hz), 2.24 (m, 1H), 2.13 (m, 2H), 1.77 (s, 3H),
1.69-1.41 (m, 6H), 1.08 (d, 3H, J = 6.1 Hz), 0.90 (t, 9H, J = 8.0
Hz), 0.84-0.79 (m, 36H), 0.54 (q, 6H, J = 8.0 Hz), 0.01-0.05 (m,
24H); 13C NMR (100 MHz, CDCl3) δ 145.0 (s), 135.8 (d), 126.6
(d), 77.5 (d), 70.9 (d), 69.4 (d), 67.7 (d), 67.4 (d), 66.0 (d), 48.6 (t),
47.2 (t), 46.5 (t), 45.6 (t), 40.6 (t), 26.1-25.7 (12q), 24.5 (q), 24.1
(q), 18.1 (4s), 7.1 (3q), 5.7 (3t), -2.9 to -4.4 (8q); HRMS (ESI)
calcd for C46H99IO5Si5 þ Naþ 1021.5281, found 1021.5287.
2,2-Dimethyl-5-[(1E,3E,5E,7E,13E)-(10S,12S,16S,18S,20R)-
10,12,16,20-tetrakis(tert-butyldimethylsilyloxy)-8-methyl-18-tri-
ethylsilyloxyhenicosa-1,3,5,7,13-pentaenyl]-1,3-benzodixin-4-one
(36). To a stirred solution of vinyl iodide 35 (57 mg, 57 μmol,
1.0 equiv) and trienic stannane 14 in anhydrous and degassed
DMF (2 mL) were successively added triphenylarsine (2 mg,
6 μmol, 0.1 equiv) and Pd2(dba)3 (4 mg, 3 μmol, 0.05 equiv) at rt.
After 15 h of stirring in the dark at rt, the reaction mixture was
diluted in Et2O (5 mL) and poured in a saturated aqueous NH4Cl
18H), 0.53 (q, 6H, J = 8.0 Hz), 0.01, 0.00, -0.01 (s, 12H); 13
C
NMR (100 MHz, CDCl3) δ 200.1 (s), 145.6 (d), 144.6 (s), 132.5
(d), 77.5 (d), 68.8 (d), 67.5 (d), 65.9 (d), 65.6 (d), 48.4 (t), 46.3 (t),
46.2 (t), 45.0 (t), 40.5 (t), 26.0, 25.8 (6q), 24.4 (q), 24.2 (q),
18.1 (2s), 7.0 (3q), 5.6 (3t), -3.9 (q), -4.3 (2q), -4.5 (4q);
HRMS (ESI) calcd for C34H69IO5Si3 þ Naþ 791.3395, found
791.3386.
J. Org. Chem. Vol. 74, No. 20, 2009 7673

Amans et al.
JOCArticle
solution (5 mL). The aqueous layer was extracted with Et2O (3 ꢀ
5 mL) and the combined organic layers were washed with brine
(5 mL), dried over MgSO4, filtered and concentrated in vacuo.
Purification of the residue by chromatography on a preparative
TLC plate (hexanes/EtOAc 95/5) provided the desired protected
monomer 36 (38 mg, 59%) as a yellow oil. Rf 0.5 (petroleum
(m, 24H); 13C NMR (100 MHz, CDCl3) δ 160.5 (s), 157.0 (s),
141.9 (s), 137.9 (s), 136.0 (d), 135.9 (d), 134.9 (d), 133.7 (d), 132.2
(d), 130.9 (d), 130.8 (d), 129.1 (d), 128.3 (d), 126.4 (d), 120.5 (d),
115.7 (d), 110.5 (s), 105.1 (s), 70.9 (d), 69.4 (d), 68.2 (d), 67.7 (d),
66.0 (d), 48.6 (t), 48.1 (t), 46.6 (t), 45.6 (t), 40.6 (t), 25.9 (12q), 25.6
(2q), 24.1 (q), 18.1 (4s), 17.8 (q), 7.1 (3q), 5.7 (3t), -3.8 to -4.7
þ
ether/EtOAc 90/10); [R]20 -12.0 (c 0.68, CHCl3); IR (neat)
(8q); HRMS (ESI) calcd for C62H114O8Si5 þ Na 1149.7258,
D
2954, 2929, 2857, 1739, 1576, 1473, 1253, 1079, 1046, 1005, 836,
775 cm-1; 1H NMR (400 MHz, CDCl3) δ 7.66 (d, 1H, J = 15.4
Hz), 7.37 (t, 1H, J = 8.0 Hz), 7.27 (d, 1H, J = 7.3 Hz), 6.81 (dd,
1H, J = 15.4, 10.0 Hz), 6.74 (dd, 1H, J = 8.0, 1.0 Hz), 6.47-6.37
(m, 3H), 6.17 (dd, 1H, J = 14.7, 10.0 Hz), 6.88 (d, 1H, J = 14.7
Hz), 5.50 (dt, 1H, J = 15.4, 6.8 Hz), 5.36 (dd, 1H, J = 15.4, 6.8
Hz), 4.10 (br q, 1H, J = 6.5 Hz), 3.86-3.74 (m, 4H), 2.23 (dd, 1H,
J = 13.3, 4.8 Hz), 2.14 (m, 3H), 1.74 (s, 3H), 1.65 (s, 6H),
1.62-1.43 (m, 6H), 1.07 (d, 3H, J = 6.1 Hz), 0.90 (t, 9H, J = 8.0
Hz), 0.84-0.79 (m, 36H), 0.54 (q, 6H, J = 13.3 Hz), 0.00 to -0.06
found 1149.7248.
ꢀ ꢀ
Acknowledgment. We wish to thank the Societe de Chimie
ꢀ
Therapeutique/SERVIER for financial support of this work
and for a grant to one of us (D.A.).
Supporting Information Available: 1H and 13C NMR spec-
tra for compounds 6, 8, 11, 12, 14, 17, 20, 22-24, 26-29, and
31-38. This material is available free of charge via the Internet
at http://pubs.acs.org.
7674 J. Org. Chem. Vol. 74, No. 20, 2009