Directed hydrogenations and an Ireland-Claisen rearrangement linked to Evans-Tishchenko chemistry: The highly efficient total synthesis of the marine cyclodepsipeptide doliculide

Article abstract of DOI:10.1002/chem.201501252

Two new convergent total syntheses have been developed for the cytotoxic, actin microfilament-stabilizing marine cyclodepsipeptide doliculide (1). A key strategic element of both routes is the establishment of the central stereogenic center of the characteristic polydeoxypropionate stereotriad by means of a hydroxyl-directed catalytic hydrogenation of a trisubstituted double bond. The requisite olefin substrates were obtained through a modified Suzuki-Miyaura coupling or through Ireland-Claisen rearrangement of a propionate ester, respectively; the latter was the direct result of a highly selective Evans-Tishchenko reduction of a hydroxy ketone that had been obtained in a stereoselective Paterson aldol reaction. Doliculide (1) was finally obtained in a total number of 17 or 15 (14) linear steps, respectively, which represents a substantial improvement over previous syntheses of this highly bioactive natural product.

Full text of DOI:10.1002/chem.201501252

DOI: 10.1002/chem.201501252
Communication
&
Synthetic Methods
Directed Hydrogenations and an Ireland–Claisen Rearrangement
Linked to Evans–Tishchenko Chemistry: The Highly Efficient Total
Synthesis of the Marine Cyclodepsipeptide Doliculide
Tao Chen and Karl-Heinz Altmann*[a]
lide (1), which was isolated from the sea hare Dolabella auricu-
laria by Yamada and co-workers in 1994.[13]
Abstract: Two new convergent total syntheses have been
developed for the cytotoxic, actin microfilament-stabiliz-
ing marine cyclodepsipeptide doliculide (1). A key strate-
gic element of both routes is the establishment of the
central stereogenic center of the characteristic polydeoxy-
propionate stereotriad by means of a hydroxyl-directed
catalytic hydrogenation of a trisubstituted double bond.
The requisite olefin substrates were obtained through
a modified Suzuki–Miyaura coupling or through Ireland–
Claisen rearrangement of a propionate ester, respectively;
the latter was the direct result of a highly selective Evans–
Tishchenko reduction of a hydroxy ketone that had been
obtained in a stereoselective Paterson aldol reaction. Doli-
culide (1) was finally obtained in a total number of 17 or
15 (14) linear steps, respectively, which represents a sub-
stantial improvement over previous syntheses of this
highly bioactive natural product.
Natural products perturbing the structure and function of the
actin cytoskeleton are powerful molecular tools for the dissec-
tion of actin-dependent cellular processes.[1–3] Due to their pro-
nounced antiproliferative activity, they could also serve as lead
structures for anticancer drug discovery,[3–5] although the true
potential of actin as a viable drug target still remains to be es-
tablished. Actin-interacting natural products have been isolat-
ed from a variety of biological sources, including terrestrial
(plants, fungi, bacteria),[6–8] as well as marine (nudibranchs,
sponges, algae) organisms,[1] and they cover a diverse range of
molecular architectures. Although the oldest actin-interacting
natural product, phalloidin,[9,10] is a bicyclic peptide, the majori-
ty of currently known natural actin binders feature (mono)ma-
crocyclic core structures of varying ring size and complexi-
ty.[11,12] This group also includes the cyclic depsipeptide dolicu-
Doliculide (1) showed potent in vitro antiproliferative activity
against the human leukemia cell line HeLa-S3[13] (with an half
maximal inhibitory concentration (IC50) value of 1.5 nm) and
was later demonstrated to destroy actin stress fibers in cells
and to induce actin assembly into clump-like aggregates;[14]
the cellular effects of 1 were virtually identical with those of
jasplakinolide, a different microfilament-stabilizing cyclodepsi-
peptide of marine origin that has found widespread use as
a tool compound in cell biology.[15]
More recently, 1 was also shown to cause rapid immobiliza-
tion of G-actin and to induce apoptosis in cancer cells.[16] In
spite of its very favorable biochemical and cellular properties,
however, doliculide (1) has been studied much less extensively
than jasplakinolide, including the exploration of structure–ac-
tivity relationships (SAR).
To date, four successful total syntheses of doliculide (1) have
been reported in the literature[13,17–20] that are all based on ring
closure by macrolactamization, either between C14 and
N13[13,17] or between C1 and N16.[18–20] The requisite w-amino
acids 2 or 3 have been elaborated from different C1ꢀC9 pre-
cursors (doliculide numbering), including acid 4 (!2),[13,17]
ester 5 (!3),[18,20] and terminal olefin 6 (!3).[19] The major
conceptual differences between the existing syntheses of doli-
culide (1) relate to the construction of the respective C1ꢀC9
polyketide fragment, with its challenging all-syn 1,3,5-trimethyl
[a] T. Chen, Prof. K.-H. Altmann
Swiss Federal Institute of Technology (ETH) Zꢀrich
Department of Chemistry and Applied Biosciences
Institute of Pharmaceutical Sciences, HCI H405
Vladimir-Prelog-Weg 4, 8093 Zꢀrich (Switzerland)
Fax: (+41)44 633 13 69
E-mail: karl-heinz.altmann@pharma.ethz.ch
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201501252.
Chem. Eur. J. 2015, 21, 1 – 6
1
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
motif; even in the best case, however, 22 linear steps
were required to access the respective variant of this
key intermediate from readily available starting mate-
rials. Herein, we now describe two new approaches
to polyketide fragment 5 that are both significantly
more efficient than the existing syntheses. The
common conceptual feature for both approaches is
the establishment of the methyl-containing stereo-
genic center at C4 by the hydroxyl-directed stereose-
lective hydrogenation of a trisubstituted double
bond.[21] For the shorter of the two routes, this ulti-
mately enabled access to 5 in only 11 or 12 steps
(see below) and with almost perfect atom econo-
my.[22,23]
Scheme 1. Retrosynthesis of polyketide fragment 5. First-generation approach. Bn=ben-
zyl, PMB=p-methoxybenzyl, TBS=tert-butyldimethylsilyl.
As illustrated in Scheme 1, one possible option to
exploit the directing effect of a free hydroxyl group
for the establishment of the C4 stereocenter is the
change of the oxidation state of C1, thus revealing
homoallylic alcohol 8 as a suitable hydrogenation
substrate. Based on the conformational model pro-
posed by Evans for the rhodium-catalyzed hydroge-
nation of trisubstituted olefins related to 8,[21] the de-
sired isomer 7 was expected to be formed in this re-
action with high selectivity (via intermediate 9,
Scheme 1). Alcohol 7 would then be converted into 5
by oxidation of C1 and protection (C1-carboxyl
group)/deprotection (C9-hydroxyl group), whereas 8
itself was envisioned to be obtained by Negishi- or
Suzuki-type cross-coupling of vinyl iodide 10 and
alkyl iodide 11. The convergent construction of the
C1ꢀC9 segment of doliculide (1) by the coupling of
two fragments of comparable complexity is not part
of any of the previous syntheses of the compound;
in all four cases, intermediates 4, 5, or 6 were ob-
tained by stepwise linear chain extensions.
In the forward direction, the synthesis of building
block 11 commenced with the Noyori reduction[24] of
b-keto ester 12, which provided b-hydroxy ester 13
in high yield (94%) and with 93% enantiomeric
excess (ee; Scheme 2).[25] Protection of the hydroxyl
group in 13 as a benzyl ether followed by diisobuty-
laluminium hydride (DIBALH) reduction of the ester
moiety then furnished aldehyde 15 in 74% overall
yield. Aldehyde 15 underwent smooth aldol reaction
with the boron enolate of the propionylated Masa-
mune auxiliary 16,[26] to give the desired anti-aldol
product 17 in 83% isolated yield (the diastereomeric
ratio (d.r.) for the reaction was >11:1).[27] Silylation of
17 with TBS trifluoromethanesulfonate (TBSOTf) fol-
lowed by reductive removal of the auxiliary with
DIBALH gave a primary alcohol that was converted
Scheme 2. a) H2 (9 bar), (R)-RuIICl2[BINAP], MeOH, 1008C, 6 h, 94%, 93% ee; b) benzyl
2,2,2-trichloroacetimidate, TFMSA, CH2Cl2/cyclohexane 1:2, RT, 20 h, 93%; c) DIBALH,
CH2Cl2, ꢀ788C, 3 h, 79%; d) 16, Et3N, c-Hex2BOTf, ꢀ788C, 7 h, 83% (d.r.>11:1); e) TBSOTf,
2,6-lutidine, ꢀ788C!08C, 1 h, 97%; f) DIBALH, CH2Cl2, ꢀ788C, 3 h, 84%; g) I2, PPh3, imi-
dazole, Et2O/CH3CN 4:1, RT, 10 min, 86%; h) Ph3P, CBr4 CH2Cl2, ꢀ108C, 30 min, 78%;
i) nBuLi, MeI, THF, ꢀ788C!RT, 2.5 h, 85%; j) 1. (nBu)3Sn(nBu)CuCNLi2, THF, 08C, 3 h; 2. I2,
CH2Cl2, 08C, 49% over two steps. k) 11, tBuLi, 9-methoxy-BBN, [PdCl2(dppf)], ether/DMF
1:1, ꢀ788C!RT, 17 h, 83%; l) DDQ, pH 7 buffer, CH2Cl2, RT, 1 h, 82%; m) H2 (9 bar),
Rh[(nbd)dppb]BF4 (20 mol%), CH2Cl2, RT, 8 h, 86%; n) TEMPO, sodium phosphate buffer
(pH 6), 80% NaClO2, NaOCl (cat.), CH3CN, 358C, 7 h, 83%; o) (Boc)2O, DMAP, tBuOH, 308C,
5 h; p) H2 (1 atm), 20% Pd(OH)2/C, 1,4-dioxane, 308C, 5 h, 85% over two steps.
BINAP=(2,2’-bis(diphenylphosphino)-1,1’-binaphthyl, DIBALH=diisobutylaluminium hy-
dride, TFMSA=trifluoromethanesulfonic acid, TBSOTf=TBS trifluoromethanesulfonate,
BBN=9-borabicyclo[3.3.1]nonane, dppf=1,1’-bis(diphenyl-phosphino)ferrocene,
DDQ=2,3-dichlor-5,6-dicyano-1,4-benzoquinone, nbd=norbornadiene, dppb=1,4-bis(di-
phenylphosphino)butane, TEMPO=2,2,6,6-tetramethylpiperidinyloxyl, DMAP=4-dime-
thylaminopyridine.
into iodide 11 in an Appel reaction (Scheme 2). Building block
11 was obtained in 69% overall yield for the three-step se-
quence from 17.
Vinyl iodide 10 was prepared via aldehyde 19 (obtained in
two steps from R-Roche ester in 94% overall yield), according
to procedures previously developed by Smith and co-workers
in the context of their synthesis of tedanolide (Scheme 2).[28]
Specifically, 19 was converted into alkyne 20 by reaction with
Ph3P and CBr4 to produce a dibromo olefin intermediate, fol-
lowed by treatment of the latter with nBuLi and CH3I. Stannyl-
cupration/iodination of 20 then provided vinyl iodide 10 in
32% overall yield (based on 19).
&
&
Chem. Eur. J. 2015, 21, 1 – 6
www.chemeurj.org
2
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
The conversion of 11 into a boronate complex by treatment
with nBuLi and 9-methoxy-BBN and in situ Suzuki coupling of
this complex with vinyl iodide 10 gave the desired trisubstitut-
ed olefin 21 in excellent yield (83%). Alternatively, 21 could
also be obtained from 10 and 11 by Negishi coupling, but the
yield of this reaction was substantially lower (51%; for details,
see the Supporting Information). Cleavage of the PMB ether in
21 with DDQ then provided the free homoallylic alcohol 8 as
the projected substrate for the hydroxyl-directed stereoselec-
tive hydrogenation step. In the event, treatment of 8 with hy-
drogen (9 bar) in CH2Cl2 in the presence of Rh[(nbd)dppb]BF4
gave the desired stereoisomer 7 in excellent yield (86% after
purification by flash chromatography) without the need for
any optimization of reaction conditions. Alcohol 7 was elabo-
rated into the polyketide fragment 5 by oxidation, esterifica-
tion with BOC2O/DMAP[18] and debenzylation of the C9 hydrox-
yl group with H2/20% Pd(OH)2/C at atmospheric pressure.
Overall, 5 was obtained from b-keto ester 12 in only 13 linear
steps and 20% overall yield.
different options evaluated for the construction of 22, an Ire-
land–Claisen rearrangement[29,30] of 23 appeared to be particu-
larly attractive, because the latter should be directly accessible
via Evans–Tishchenko reduction of hydroxy ketone 24
(Scheme 3);[31] 24 in turn was envisioned to result from the ste-
reoselective aldol reaction between methacrolein (25) and
ethyl ketone 26.
As illustrated in Scheme 4, ethyl ketone 26 was readily ac-
cessible from ester 14 (which had already been an intermedi-
ate in our first generation synthesis; Scheme 2) via formation
of the corresponding Weinreb amide and its in situ reaction
with ethylmagnesium bromide (Scheme 4).
Although the above-described approach delivered the poly-
ketide fragment of doliculide in significantly fewer steps than
any of the published syntheses, it appeared that a potentially
more effective way of exploiting the directing effect of a homo-
allylic hydroxyl group in a stereoselective hydrogenation reac-
tion would be the engagement of one of the hydroxyl groups
that are integral parts of the doliculide structure. By doing so,
the need for a subsequent adjustment of the oxidation state
of the hydroxyl-bearing carbon would be eliminated; in addi-
tion, the corresponding hydrogenation substrate(s) might be
accessible without having to rely on an auxiliary-based anti-
aldol reaction to establish the C6 and C7 stereocenters.
Scheme 4. a) d-proline, DMSO, RT, 48 h, 48%, 80% ee; b) benzyl 2,2,2-tri-
chloroacetimidate, TFMSA, CH2Cl2/cyclohexane 1:2, RT, 18 h, 83%;
c) Me(MeO)NH·HCl, EtMgBr, THF, ꢀ58C!RT, 15 h, 61%; d) (+)-(Ipc)2BOTf,
iPr2NEt, CH2Cl2, ꢀ788C, 2 h, then methacrolein (25), ꢀ788C, 2 h, ꢀ208C,
16 h, 63% (brsm 93%, 30% of 26 recovered); e) SmI2, C2H5CHO, THF, ꢀ25!
ꢀ158C, 1 h, 87%; f) TBSOTf, 2,6-lutidine, CH2Cl2, ꢀ78!08C, 30 min, 91%;
g) TBSOTf, KHMDS, Et2O, ꢀ80!08C, 10 h, then HF/CH3CN, 1.5 h, RT, 85%;
h) iPrNHC(Ot-C4H9)=NiPr, tBuOH/CH2Cl2 1:2, 408C, 8 h, 84%; i) H2 (9 bar),
Rh[(nbd)dppb]BF4, CH2Cl2, RT, 8 h, 90%; j) TBSOTf, 2,6-lutidine, CH2Cl2,
ꢀ788C!08C, 1 h; k) H2 (9 bar), 20% Pd(OH)2/C, 1,4-dioxane, RT, 3 h, 86%
over two steps. Brsm=based on recovered starting material,
When analyzing the different possibilities, the most promis-
ing option was found to be the establishment of the C4 ste-
reocenter by exploiting the directing effect of the C7 hydroxyl
group in the hydrogenation of olefin 22 (Scheme 3). As for ho-
moallylic alcohol 8, following the conformational arguments
established by Evans,[21] the Rh-catalyzed hydrogenation of 22
was expected to deliver the desired isomer with high selectivi-
ty. Protecting-group reshuffling would then lead to 5. Of the
(Ipc)2BOTf=diisopinocampheylboron trifluoromethanesulfonate,
KHMDS=potassium hexamethyldisilazide.
Alternatively, 26 can also be obtained from 2-butanone (31)
and iso-butyraldehyde (32) by proline-catalyzed aldol reaction
followed by benzylation.[32] However, the ee of the organocata-
lytic aldol reaction did not exceed 80%, and no attempts were
made to improve the selectivity of this transformation.[32]
A Paterson aldol reaction between the diisopinocampheyl
enolborinate of 26 (obtained by treatment of 26 with freshly
prepared diisopinocampheylboron triflate and Hꢁnig base at
ꢀ788C for 2 h) and methacrolein (25) gave the desired aldol
product 24 as a single isomer in 63% yield.[33] Hydroxy ketone
24 underwent smooth Evans–Tishchenko reduction[31] with
propionaldehyde, to give the desired propionate ester 27 in
high yield (87%) and with excellent selectivity.[34]
Reaction of 27 with TBSOTf then produced TBS ether 23 as
the substrate for the crucial Ireland–Claisen rearrangement.
After a series of unsuccessful attempts to induce the rear-
rangement by treatment with TBSCl/LDA or TMSCl/LDA in
Scheme 3. Retrosynthesis of polyketide fragment 5. Second-generation ap-
proach. Bn=benzyl, tBu=tert-butyl.
Chem. Eur. J. 2015, 21, 1 – 6
www.chemeurj.org
3
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
THF/HMPA[29,30] at temperatures between ꢀ788C and
up to 608C, the difficulties with the reaction could be
overcome by the use of KHMDS and TBSOTf in Et2O/
THF.[35] This reagent/solvent combination induced
clean rearrangement of 23 and, after the in situ
cleavage of the TBS-ester/TBS-ether groups with
HF·pyridine, gave acid 28 as a single isomer in a re-
markable overall yield of 85%.
Scheme 5. a) DCC, DMAP, CH2Cl2, ꢀ208C, 25 h, 82%; b) 1. TFA, CH2Cl2, 08C!RT, 3 h; 2.
(NH3/MeOH aq.); c) BOP, DMAP, CH2Cl2, 08C!RT, 30 h, 82% over two steps; d) TBAF, THF,
08C, 10 min, 84%. DCC=dicyclohexyl carbodiimide, TFA=trifluoroacetic acid, BOP =
(benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexafluorophosphate, TBAF=te-
trabutylammonium fluoride.
With the critical rearrangement successfully accom-
plished, esterification of 28 with N,N’-diisopropyl-O-
tert-butylisourea at 408C gave unsaturated ester 22
in 84% yield.[36] The subsequent hydrogenation of 22, which
was to be the second key transformation in our second gener-
ation approach to 5, was conducted under the same condi-
tions that had been successfully employed for the conversion
of 8 into 7 (i.e., Rh[(nbd)dppb]BF4, 9 bar H2, CH2Cl2, RT). Re-
markably, the reaction gave the desired product 29 as a single
isomer in excellent yield (89%), attesting to the power of the
directing effect of the free hydroxyl group; as for the stereose-
lective reduction of 8, no further optimization of reaction con-
ditions was required. Ester 29 was then elaborated into 5 by
reaction with TBS triflate and subsequent hydrogenolytic re-
moval of the benzyl group in 77% overall yield. The polyketide
unit 5 was thus obtained from b-keto ester 12 in only 11 steps
and in 15% overall yield (Scheme 2). In principle, the route can
be further shortened by one step, if ethyl ketone 26 is pro-
duced from 2-butanone (31) and iso-butyraldehyde (32) (see
above). Polyketide fragment 5 prepared by either of the two
routes described above had identical spectroscopic properties
that were also in agreement with the data previously reported
for this intermediate in the literature.[18,37]
teresting bioactive natural product. Meanwhile, the chemistry
described herein has also been exploited for the synthesis of
a series of analogues with modifications in the dipeptide part
of 1. The biological effects of these compounds will be report-
ed in a separate publication.
Acknowledgements
This work was supported by the Chinese Scientific Council
(graduate fellowship to T.C.) We thank Dr. Bernhard Pfeiffer and
Leo Betschart for NMR support. We are grateful to Dr. Michael
Wçrle, Dr. Nils Trapp, and Michael Solar for crystallographic
analysis. Dr. Xingyang Zhang, Louis Bertschi, Renꢂ Hꢃfliger, and
Oswald Greter are acknowledged for HRMS spectra.
Keywords: actin
· Claisen rearrangement · doliculide ·
hydrogenation · total synthesis
[1] S.-Y. Saito in Progress in Molecular and Subcellular Biology, Marine Molec-
ular Biotechnology (Eds.: N. Fusetani and W. Kem), Springer, Heidelberg,
2009, pp. 187–219.
[2] I. Spector, F. Braet, N. R. Shochet, M. R. Bubb, Microsc. Res. Tech. 1999,
47, 18–37.
Polyketide fragment 5 was then elaborated into 1 in 4 steps
and 56% overall yield according to Ghosh and Liu
(Scheme 5).[18] The analytical data for the material thus ob-
tained were in excellent agreement with those previously re-
ported for the compound in the literature.[17,18] Likewise, the
biological profile of synthetic doliculide (1)[16] was found to be
highly comparable with the one reported for the natural prod-
uct.[14]
[3] J. S. Allingham, V. A. Klenchin, I. Rayment, Cell. Mol. Life Sci. 2006, 63,
2119–2134.
[4] G.-Y. Kim, W.-J. Kim, Y. H. Choi, Mar. Drugs 2011, 9, 2176–2187.
[5] G. S. B. Andavan, R. Lemmens-Gruber, Mar. Drugs 2010, 8, 810–834.
[6] P. M. Sçrensen, R. E. Iacob, M. Fritzsche, J. R. Engen, R. M. Brieher, G.
Charras, U. S. Eggert, ACS Chem. Biol. 2012, 7, 1502–1508.
[7] K. Scherlach, D. Boettger, N. Remme, C. Hertweck, Nat. Prod. Rep. 2010,
27, 869–886.
[8] L. A. Smyth, T. Brodmann, M. Kalesse in Drug Discovery Series, Vol. 25:
Drug Discovery from Natural Products, RSC, Cambridge (UK), 2012,
pp. 17–41; and references cited therein.
In conclusion, we have developed two new highly efficient,
convergent approaches towards the total synthesis of the
marine macrolide doliculide (1) that are both centered on the
establishment of the C4 chiral center by the hydroxyl-directed
catalytic hydrogenation of a trisubstituted olefin. While the
first generation route, with the directing hydroxyl group at C1,
gave 1 in higher overall yield (20% vs. 15%), the second gen-
eration approach, which exploited the directing effect of the
native C7 hydroxyl group, is shorter (15 (14) linear steps vs. 17
for the first generation approach); the second-generation ap-
proach also critically relied on the exploitation of the ester
product obtained in an Evans–Tishchenko reduction as a sub-
strate for an Ireland–Claisen rearrangement. To the best of our
knowledge, such a strategy is unprecedented in the literature
and contributes to the enhanced step economy of our second-
generation approach towards doliculide (1).
[9] T. Wieland, Science 1968, 159, 946–952.
[10] For a review see: T. Wieland, Adv. Enz. Reg. 1977, 15, 285–299.
[11] K.-S. Yeung, I. Paterson, Angew. Chem. Int. Ed. 2002, 41, 4632–4653;
Angew. Chem. 2002, 114, 4826–4847.
[12] M. Kita, H. Kigoshi, Nat. Prod. Rep. 2015, 32, 534–542.
[13] H. Ishiwata, T. Nemoto, M. Ojika, K. Yamada, J. Org. Chem. 1994, 59,
4710–4711.
[14] R. Bai, D. G. Covell, C. Liu, A. K. Ghosh, E. Hamel, J. Biol. Chem. 2002,
277, 32165–32171.
[15] M. R. Bubb, A. M. J. Senderowicz, E. A. Sausville, K. L. K. Duncan, E. D.
Korn, J. Biol. Chem. 1994, 269, 14869–14871.
[16] F. Fçrster, S. Braig, T. Chen, K.-H. Altmann, A. M. Vollmar, Bioorg. Med.
Chem. 2014, 22, 5117–5122.
[17] H. Ishiwata, H. Sone, H. Kigoshi, K. Yamada, Tetrahedron 1994, 50,
12853–12882.
[18] A. K. Ghosh, C. Liu, Org. Lett. 2001, 3, 635–638.
[19] S. Hanessian, V. Mascitti, S. Giroux, Proc. Natl. Acad. Sci. USA 2004, 101,
11996–12001.
Both syntheses are significantly shorter than any of the pre-
vious syntheses of 1, thus greatly facilitating access to this in-
&
&
Chem. Eur. J. 2015, 21, 1 – 6
www.chemeurj.org
4
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
[20] K. Matcha, A. V. R. Madduri, S. Roy, S. Ziegler, H. Waldmann, A. K. H.
Hirsch, A. J. Minnaard, ChemBioChem 2012, 13, 2537–2548.
[21] D. A. Evans, M. M. Morrissey, R. L. Dow, Tetrahedron Lett. 1985, 26,
6005–6008.
[28] A. B. Smith III, D. Lee, J. Am. Chem. Soc. 2007, 129, 10957–10962.
[29] a) R. E. Ireland, R. H. Mueller, J. Am. Chem. Soc. 1972, 94, 5897–5898;
b) R. E. Ireland, R. H. Mueller, A. K. Willard, J. Am. Chem. Soc. 1976, 98,
2868–2877; c) R. E. Ireland, P. Wipf, J. D. Armstrong III, J. Org. Chem.
1991, 56, 650–657.
[30] For reviews see: a) S. Blechert, Synthesis 1989, 71–82; b) A. M. M.
Castro, Chem. Rev. 2004, 104, 2939–3002.
[31] D. A. Evans, A. H. Hoveyda, J. Am. Chem. Soc. 1990, 112, 6447–6449.
[32] The same ee has been reported for this reaction by Barbas and co-
workers, who obtained the aldol product in 60% yield: K. Sakthivel, W.
Notz, T. Bui, C. F. Barbas, III, J. Am. Chem. Soc. 2001, 123, 5260–5267. In
our synthesis of doliculide (1), only 26 obtained from 14 was processed
in the aldol reaction with methacrolein (25).
[33] I. Paterson, J. M. Goodman, Tetrahedron Lett. 1989, 30, 997–1000.
[34] CCDC-1047307 and CCDC-1047308 contain the supplementary crystal-
lographic data for this paper (compund 24). These data can be ob-
tained free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
[22] For a recent report on the convergent synthesis of deoxypropionates
via metallacycle-mediated allylic alcohol-alkyne cross-coupling and sub-
sequent hydroxyl-directed stereoselective hydrogenation, see: P. S. Diez,
G. C. Micalizio, Angew. Chem. Int. Ed. 2012, 51, 5152–5156; Angew.
Chem. 2012, 124, 5242–5246.
[23] Pfaltz and co-workers have recently presented a strategy for the con-
vergent synthesis of long-chain polydeoxypropionates that employs
the stereoselective catalytic hydrogenation of trisubstituted internal
double bonds as key steps: M. C. Pischl, C. F. Weise, S. Haseloff, M.-A.
Mꢁller, A. Pfaltz, C. Schneider, Chem. Eur. J. 2014, 20, 17360–17374.
[24] a) R. Noyori, T. Ikeda, T. Ohkuma, M. Widhalm, M. Kitamura, H. Takaya, S.
Akutagawa, N. Sayo, T. Saito, J. Am. Chem. Soc. 1989, 111, 9134–9135;
b) R. Noyori, T. Ohkuma, M. Kitamura, H. Takaya, N. Sayo, H. Kumobaya-
shi, S. Akutagawa, J. Am. Chem. Soc. 1987, 109, 5856–5858.
[25] The absolute configuration of the newly formed stereocenter was de-
termined by Mosher ester analysis as described in: T. R. Hoye, C. S. Jef-
frey, F. Shao, Nat. Protoc. 2007, 2, 2451–2458.
[35] S.-p. Hong, H. A. Lindsay, T. Yaramasu, X. Zhang, M. C. McIntosh, J. Org.
Chem. 2002, 67, 2042–2055.
[36] L. J. Mathias, Synthesis 1979, 561–576.
[26] A. Abiko, J. F. Liu, S. Masamune, J. Am. Chem. Soc. 1997, 119, 2586–
2587.
[37] The all-syn arrangement of the methyl groups on C2, C4, and C6 was
further confirmed by an NMR spectroscopy method recently developed
by Breit and co-workers that is based on the differences in chemical
shifts between the vicinal protons attached to the carbon atoms con-
necting the chiral centers in polydeoxypropionates: Y. Schmidt, K. Lehr,
L. Colas, B. Breit, Chem. Eur. J. 2012, 18, 7071–7081.
[27] The absolute configuration of the newly formed hydroxyl-bearing ste-
reocenter in 17 was confirmed by conversion into the corresponding
acetonide 18 (after debenzylation) and analysis of its 13C NMR spec-
trum. The 13C chemical shifts of the two ketal methyl groups in 18 were
observed at d=24.34 and 24.41 ppm, whereas the quaternary ketal
carbon appeared at 100.8 ppm. These data are in excellent agreement
with the predicted shifts for acetonides of anti-1,3-diols: S. D. Rychnov-
sky, B. N. Rogers, T. I. Richardson, Acc. Chem. Res. 1998, 31, 9–17.
Received: March 30, 2015
Published online on && &&, 0000
Chem. Eur. J. 2015, 21, 1 – 6
www.chemeurj.org
5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
COMMUNICATION
Two concise total syntheses have been
developed for the marine natural prod-
uct doliculide (1) from simple b-keto
ester 12. A key strategic element of
both syntheses is the high-yielding and
highly selective hydroxyl-directed Rh-
catalyzed hydrogenation of a homoallylic
alcohol intermediate, which was ob-
tained from 12 either by Suzuki cou-
pling (9) or Ireland–Claisen rearrange-
ment (22) as additional key steps. Both
syntheses are substantially shorter than
existing routes to 1.
&
Synthetic Methods
T. Chen, K.-H. Altmann*
&& – &&
Directed Hydrogenations and an
Ireland–Claisen Rearrangement Linked
to Evans–Tishchenko Chemistry: The
Highly Efficient Total Synthesis of the
Marine Cyclodepsipeptide Doliculide
&
&
Chem. Eur. J. 2015, 21, 1 – 6
www.chemeurj.org
6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!