The leiodolide B puzzle

Article abstract of DOI:10.1002/anie.201005850

Out of options? Even though a systematic approach was chosen, which led to a set of four diastereomeric macrolides modeled around the proposed structure of leiodolide B (see picture), the puzzle concerning the stereostructure of this cytotoxic metabolite derived from a deep-sea sponge still remains unsolved. Copyright

Full text of DOI:10.1002/anie.201005850

Communications
DOI: 10.1002/anie.201005850
Natural Product Synthesis
The Leiodolide B Puzzle**
Alexandre Larivꢀe, John B. Unger, Mikaꢁl Thomas, Conny Wirtz, Christophe Dubost,
Shinya Handa, and Alois Fꢂrstner*
The quest for structurally novel and
biologically significant lead compounds
from marine sources is driving teams
involved in the isolation of natural prod-
ucts to explore ever more remote loca-
tions. While the collection of diversified
species in shallow waters by scuba diving
is now well established, investigations
into the deep sea environment are scarce
for logistical reasons, even though they
offer considerable promise.[1] During a
campaign in Palau, Fenical and co-work-
ers were able to collect sponges from the
rare genus Leiodermatium at a depth of
220 m by using a manned submersible.
Bioassay-guided fractionation of the
methanol extracts led to the isolation of
two unusual macrolides of apparently
mixed polyketide/nonribosomal peptide Scheme 1. Proposed structures of and biosynthetic relationship between leiodolide A (1) and
leiodolide B (2). Retrosynthetic analysis of 2 accounting for the unknown configuration at C13.
synthetase origin termed leiodolide A
(1) and B (2).[2,3] Both compounds, as
well as the derived methyl ester 3,
showed significant cytotoxicity and appreciable selectivity in
the NCI 60 tumor cell line assay.
hydroxy-a-methyl carboxylic acid terminus is S-configured in
both compounds, although this aspect was rigorously estab-
lished only for 1.[4]
Being the more abundant of the two metabolites, 1 (8 mg,
0.001% of the spongeꢀs dry weight) was the primary subject of
the structure-elucidation campaign. A combination of spec-
troscopic studies and degradation experiments led Fenical
and co-workers to propose the structure shown in Scheme 1,
in which the configuration of the stereogenic center at C13
remains unassigned.[2] Given the extremely limited supply of
leiodolide B (2, 0.8 mg, 0.0001% of the dry weight), its
structure was inferred from that of 1 according to a proposed
biosynthetic closure of the conspicuous brominated tetrahy-
drofuran ring.[2] This relationship also suggests that the a-
Intrigued by the architectural complexity and promising
bioactivity, we embarked on a total synthesis of these unusual
macrolides. In particular, our attention was caught by the
complex tetrahydrofuran segment embedded in the northern
sector of 2, which contains five contiguous stereogenic
centers, one of which is quaternary.[5] Moreover, the projected
synthesis must account for the unknown configuration of C13
(Scheme 1).
We planned to forge the polysubstituted THF ring by a
relay strategy which allows chiral information to be trans-
mitted from a readily available epoxide to the exigent tertiary
ether site at C23 via an axially chiral allene intermediate
(Scheme 2). To this end, epoxide 8 was prepared by a
Sonogashira coupling of 4 and 5, followed by selective
oxidation of the Z-alkene moiety in enyne 6 according to a
protocol developed by Katsuki and co-workers.[6] Specifically,
the reaction using catalytic amounts of Ti(OiPr)4 and the well
accessible salan ligand 7[7] in combination with H2O2 under
buffered biphasic conditions proceeded exceedingly well,
furnishing the desired oxirane with an enantiomeric excess of
97%. After protection of the primary alcohol, product 8 was
subjected to conjugate epoxide opening. Although iron
catalysis had previously been utilized to great advantage for
similar purposes,[5,8] it was plagued in this particular case by a
competitive direct opening of the oxirane ring. Gratifyingly
[*] Dr. A. Larivꢀe, Dr. J. B. Unger, Dr. M. Thomas, C. Wirtz, Dr. C. Dubost,
Dr. S. Handa, Prof. A. Fꢁrstner
Max-Planck-Institut fꢁr Kohlenforschung
45470 Mꢁlheim/Ruhr (Germany)
Fax: (+49)208-306-2994
E-mail: fuerstner@kofo.mpg.de
[**] Generous financial support by the MPG, the Fonds der Chemischen
Industrie, and the Japan Society for the Promotion of Science
(fellowship to S. H.) is gratefully acknowledged. We thank the
analytical departments of our Institute for excellent support
throughout the project, and Prof. W. Fenical and Dr. J. S. Sandler,
Scripps Institution of Oceanography, for an exchange of informa-
tion.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201005850.
304
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 304 –309

protecting-group manipula-
tions, including the chemo-
selective cleavage of the pri-
mary TBS group in the trisi-
lylated intermediate 12, and
oxidation of the liberated
alcohol gave aldehyde 13 in
good overall yield.
With substantial amounts
of 13 in hand, the stage was
set for a chain extension with
a C3 building block embody-
ing the as yet undefined
stereocenter at C13 (leiodo-
lide numbering). Although
the addition of a nucleophile
under Cram-chelation con-
trol was expected to lead to
the correct configuration at
C15,[12] aldehyde 13 turned
out to be surprisingly
unreactive. Presumably it is
the bulky TIPS group which
gets in the way of the incom-
ing nucleophile (Nuꢀ) along
the Burgi–Dunitz trajectory
(see 13a, Scheme 2). After
extensive screening it was
found that the reagent
derived from bromide (R)-
14
by
metal–halogen
exchange with tBuLi in
Et2O followed by transmeta-
lation with freshly prepared
MgBr2 gave well reproduci-
ble results,[13] provided that
aldehyde 13 was adminis-
tered in a 4m solution of
LiBr in CH2Cl2. Under
these conditions, the iso-
meric alcohols 15 and 16
were obtained in 73% yield
in a 1:4 ratio.[14] Not only
were these compounds sepa-
rable by careful flash chro-
matography, but the minor
isomer 15 could be recycled
by oxidation/Luche reduc-
tion to further the material
throughput. The subsequent
O-methylation of 16 pro-
Scheme 2. a) [Pd(PPh3)4] (2.2 mol%), CuI (15 mol%), Et2NH, quant.; b) Ti(OiPr)4 (10 mol%), 7 (12 mol%),
aq H2O2, CH2Cl2, pH 7.4 buffer, 408C, 99%, 97% ee; c) TBSCl, imidazole, cat. DMAP, CH2Cl2, 08C!RT,
95%; d) MeMgBr, CuCN, P(OPh)3, THF, ꢀ408C!RT, 99%, d.r. >95:5; e) AgNO3, CaCO3, acetone, H2O,
91%; f) NBS, aq DMF, 108C, 64%; g) NaHCO3, MeOH/H2O, 85%; h) TIPSOTf, 2,6-lutidine, CH2Cl2, 93%;
i) Cl3CCOOH, THF/H2O, 86%; j) DMP, pyridine, CH2Cl2, 08C, 90%; k) (R)-14, tBuLi, Et2O, ꢀ788C, MgBr2,
then 13, LiBr, CH2Cl2, 73%, d.r. =4:1; l) DMP, pyridine, CH2Cl2, 08C!RT, 74%; m) NaBH4, CeCl3·(7H2O),
MeOH, ꢀ708C, 89%, d.r.=3.5:1; n) LiHMDS, THF, MeOTf, ꢀ788C!RT, 87%; o) DDQ, CH2Cl2/H2O, 99%;
p) I2, PPh3, imidazole, CH2Cl2, 92%; q) (S)-14, tBuLi, Et2O, ꢀ788C, MgBr2, then 13, LiBr, CH2Cl2, 85%,
d.r.=6.1:1; r) LiHMDS, THF, MeOTf, ꢀ788C!RT; s) DDQ, CH2Cl2/H2O, 08C!RT, 93% (over both steps);
t) CBr4, PPh3, benzene, 558C, 97%; u) NaI, acetone, 88%. DDQ=2,3-dichloro-5,6-dicyanobenzoquinone;
DMAP=4-dimethylaminopyridine; DMP=Dess–Martin periodinane; NBS=N-bromosuccinimide;
LiHMDS=lithium hexamethyldisilazide; PMB=para-methoxybenzyl; TBDPS=tert-butyldiphenylsilyl;
TBS=tert-butyldimethylsilyl; TIPS=tri(isopropyl)silyl; Tf=trifluoromethanesulfonyl.
though, the reagent derived from MeMgBr, a stoichiometric
amount of CuCN, and P(OPh)3 cleanly afforded the desired
allenol 9.[9] The subsequent AgI-induced cyclization to the
dihydrofuran 10 was high yielding under the conditions
developed by Marshall and Pinney.[10] With the tertiary ether
being properly set by this efficient chirality transfer process, a
subsequent bromo-esterification with NBS in aqueous DMF
provided compound 11 as a single isomer.[11] Standard
ceeded well only with LiHMDS and precisely 1.05 equiva-
lents of MeOTf. Oxidative cleavage of the PMB ether in 17
followed by conversion of the resulting alcohol 18 into the
corresponding iodide 19 completed the preparation of the
northern sector of 2 in one of the two possible diastereomeric
forms.
Since the configuration of C13 in the leiodolides is
unknown, it was mandatory to prepare the epimeric building
Angew. Chem. Int. Ed. 2011, 50, 304 –309
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
305

Communications
block as well. To this end, bromide (S)-14 was added to
aldehyde 13 under the optimized conditions; the stereochem-
ical concord of this pair of reactants accounted for a higher
yield and improved diastereoselectivity (85%, 20/21 = 1:6.1).
Although the elaboration of 21 followed the tactics outlined
above, the ultimate step leading to the primary iodide 25 was
problematic. Whereas the use of I2 in combination with PPh3
and imidazole furnished substantial amounts of the tetrahy-
drofuran derivative 26, a halide exchange sequence turned
out to be viable.[15]
l-Malic acid (27) was the point of departure for the
preparation of the side-chain sector (Scheme 3). Protection
with pivaldehyde followed by diastereoselective methylation
Scheme 4. a) MnO2, CH2Cl2, 82%; b) Bu2BOTf, Et3N, CH2Cl2, ꢀ508C,
95%; c) LiBH4, THF, 08C, 97%; d) 4-methoxybenzaldehyde dimethyl-
acetal, cat. camphorsulfonic acid, DMF, 88%; e) [Pd(PPh3)4]
(70 mol%), copper thiophene-2-carboxylate, [Bu4N][Ph2PO2], DMF,
90%.
macrocyclic edifice. Exploratory studies, however, had shown
that the installation of the side chain should be given priority
Table 1: Comparison of the 13C NMR data of leiodolide B in [D4]MeOH
reported in the literature (75 MHz)[2] with the recorded data (150 MHz)
for acid (13R)-2 and the four diasteromeric esters 50 and 51 shown in
Scheme 6.[a]
Position Ref.
(13R)-2 (13R)-50 (13S)-50 (13R)-51 (13S)-51
Scheme 3. a) Pivaldehyde, H2SO4 cat., pentane, reflux, 50%;
b) LiHMDS, THF, MeI, ꢀ788C!ꢀ208C, 91%; c) BH3·SMe2, THF,
ꢀ208C!RT; d) I2, PPh3, imidazole, CH2Cl2, 41% (over two steps);
e) PPh3, MeCN, microwave, 1508C, quant.; f) nBuLi, THF, 3-phenyl-
propanal, ꢀ788C!RT, 80%.
1
166.6 166.7
166.7
121.5
153.2
44.7
166.7
121.5
153.5
44.5
166.6
121.8
153.3
45.6
71.9
128.9
129.1
143.7
134.8
166.6
25.2
34.5
29.2
36.7
78.3
80.9
166.9
121.7
153.5
46.2
71.6
129.5
128.4
143.5
135.5
167.1
26.0
35.6
31.5
39.9
79.6
83.7
77.8
16.7
14.2
21.1
2
3
122.8 121.5
153.1 153.1
4
5
45.1
71.7
44.7
72.0
72.0
72.0
6
7
8
9
131.2 128.3
125.5 128.4
143.3 143.4
134.5 135.2
166.4 166.5
128.3
128.4
143.1
135.2
166.5
25.8
33.9
29.4
37.0
77.9
80.9
78.5
15.5
14.1
20.9
58.0
55.9
84.5
37.3
127.9
128.4
39.2
75.7
128.0
129.2
143.6
135.1
166.5
25.1
34.0
30.4
38.7
79.0
83.7
79.2
15.6
of the dianion derived from 28[16] and subsequent borane
reduction gave alcohol 29, which had to be converted without
delay into the corresponding iodide 30. The seemingly simple
reaction of 30 with PPh3 required forcing conditions under
microwave irradiation. To avoid any surprise at a later stage
of the project, the performance of the Wittig reagent derived
from the phosphonium salt 31 was tested in a model reaction
with 3-phenylpropanal. As expected for a nonstabilized ylide,
the Z-alkene 32 was obtained as the only product in good
yield (Scheme 3).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
25.7
33.9
30.5
36.7
78.0
80.5
78.4
16.3
13.7
20.7
58.0
56.2
84.0
37.6
126.5 127.8
130.4 128.6
39.2
76.4
182.7 179.1
26.0
26.4
25.8
33.9
29.4
37.0
77.9
80.9
78.7
15.5
14.1
20.9
58.0
56.1
84.5
37.4
78.3
15.9
14.4
20.9
58.0
55.8
In parallel work, the known stannylated alcohol 33[17,18]
was oxidized with MnO2 and the resulting aldehyde 34
subjected to an Evans boron-aldol reaction[19] to give com-
pound 36 (Scheme 4). Reductive cleavage of the auxiliary and
protection of the resulting diol as the p-methoxybenzylidene
acetal gave product 37 in readiness for cross-coupling with the
known oxazolyl triflate 38.[20] This transformation failed
under a variety of experimental conditions. Pleasingly, how-
ever, the modified procedure recently developed in our
laboratory delivered product 39 in excellent yield and short
reaction times, despite a very high catalyst loading being
mandatory.[21,22]
14.3
20.2
59.1
56.0
84.7
59.9
55.4
84.4
84.2
37.8
127.8
128.5
39.2
75.7
37.3
37.5
127.9
128.5
39.2
75.8
127.9
128.4
39.2
75.7
39.1
75.4
177.5
25.9
25.9
177.5
26.1
25.9
177.5
26.0
25.9
177.5
26.4
25.9
25.9
26.0
The phase of fragment coupling commenced by alkylation
of 39 with iodide 19 using Et2NLi as the optimal base
(Scheme 5).[23] Reductive opening of the PMB acetal released
the primary alcohol 41 in preparation for the closure of the
[a] The heat map color-codes differences in chemical shift considered to
be beyond the experimental error as follows: red: Ddꢁ1 ppm; blue:
0.5 ppmꢂDd<1 ppm; green: 0.3 ppmꢂDd<0.5 ppm. For a compar-
ison of the 1H NMR data, see the Supporting Information.
306
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 304 –309

worth–Emmons olefination gave prod-
uct 47, which comprises the complete
carbon backbone of leiodolide B.
Buffered HF·pyridine at ambient
temperature was the only reagent that
would slowly deprotect the TIPS group
in 47, yet leave the fragile PMB group
intact and not induce epoxide ring
formation at the liberated bromohy-
drine site either. Palladium-catalyzed
cleavage of the allyl ester followed by
Yamaguchi lactonization[25] of the
resulting seco acid 48 then gave the
desired macrocycle 49 in 68% yield over
both steps. Removal of the residual
PMB group with DDQ completed the
total synthesis of the putative leiodoli-
de B methyl ester (13R)-50. The epi-
meric product (13S)-50 was prepared
analogously from 25 and 39 (for details
see the Supporting Information). We
were surprised, however, that neither of
them matched the reported data of 2,
with small but non-negligible deviations
being scattered over the entire frame-
work (Table 1). Our NMR experiments
indicated that different concentrations
of the samples were not responsible for
the observed deviances. Likewise, suc-
cessive additions of CD3COOD did not
induce noticeable changes. Even though
this experiment strongly suggested that
the differences in the spectra are not
caused by the fact that our synthetic
samples were methyl esters whereas the
natural product is a free carboxylic acid,
attempts were made to rigorously clarify
this point. Unfortunately, selective
cleavage of the methyl ester in the
presence of the macrolactone was trou-
blesome, despite exploring several enzy-
matic and chemical methods. Nonethe-
less, we were able to obtain a very small
quantity of the free acid (13R)-2 by
saponification of (13R)-50 with excess
Me3SnOH[26] and subsequent purifica-
tion by HPLC. In line with our expect-
ations, this free acid and its methyl ester
precursor had virtually identical spectra,
Scheme 5. a) 39, Et2NLi, THF, ꢀ788C, then 19, 79%; b) DIBAL-H, toluene, ꢀ408C, 75%;
c) benzoyl chloride, iPrNEt2, DMAP, CH2Cl2, 80%; d) HF·pyridine, THF, pyridine, 08C, 72%;
e) DMP, CH2Cl2, 08C!RT; f) 31, nBuLi, THF, ꢀ788C!RT, 71% (over two steps); g) NaOMe,
MeOH, 08C!RT, 75%; h) DMP, CH2Cl2, pyridine, 08C!RT; i) allyl diethyl phosphonoacetate,
LiHMDS, THF, 72% (over two steps); j) HF·pyridine, THF, pyridine, 76%; k) 4-MeC6H4SO2Na,
[Pd(PPh3)4] (16 mol%), MeOH, THF; l) 2,4,6-trichlorobenzoyl chloride, toluene, Et3N, then
DMAP, 458C, 68% (over two steps); m) DDQ, CH2Cl2, H2O, 08C!RT, 47%; h) Me3SnOH,
CH2Cl2, 608C, see text for further details. Bz=benzoyl; DIBAL-H=diisobutylaluminum hydride.
at this point. Therefore, 41 was temporarily protected as
benzoate ester 42, followed by selective removal of the
TBDPS ether at C25 with buffered HF·pyridine.[24] Oxidation
of 43 followed by reaction of the resulting aldehyde 44 with
the ylide derived from 31 gave alkene 45 in respectable
overall yield. Exposure to NaOMe in MeOH removed the
benzoate at C3 and concomitantly converted the cyclic acetal
at the acid terminus into the corresponding methyl ester.
Dess–Martin oxidation of 46 followed by a Horner–Wads-
differing only in the carboxylate terminus region (see Table 1
and the Supporting Information).
A closer inspection of the data showed stunning differ-
ences in the chemical shifts in the C2–C9 region of the
molecule. We were apprehensive that the isolation team had
been reasonably confident in assigning the 4S,5R configura-
tion to the stereogenic centers in this sector, but had
expressed a caveat that the recorded data were not totally
unambiguous.[2] Therefore, we next rushed to prepare the
enantiomeric oxazole building block ent-39 (via ent-35). Its
Angew. Chem. Int. Ed. 2011, 50, 304 –309
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
307

Communications
[3] The original publication contains two inconsistent spellings,
“leiodelide” and “leiodolide”. We use the “olide” nomenclature
to denote the macrolide character of these metabolites.
[4] Degradation studies proved the S configuration at C28 in
leiodolide A. Interestingly, however, small
amounts of the 28R epimer were also detected,
suggesting that the biosynthetic formation of this
remote center may not be stereospecific.
incorporation into the macrocyclic frame then followed the
assembly process outlined above for (13R)-50. Disappoint-
ingly, however, the data of the resulting products (13R)-51
and (13S)-51 (Scheme 6) did not match either, with even
[5] The THF ring of 2 shows remote resemblance to
the core region of amphidinolide X and Y, two
macrolides that were previously the subjects of
total synthesis and a molecular editing exercise
in our laboratory: a) O. Lepage, E. Kattnig, A.
Fꢁrstner, J. Am. Chem. Soc. 2004, 126, 15970 –
15971; b) A. Fꢁrstner, E. Kattnig, O. Lepage, J.
Am. Chem. Soc. 2006, 128, 9194 – 9204; c) A.
Fꢁrstner, E. Kattnig, G. Kelter, H.-H. Fiebig,
Chem. Eur. J. 2009, 15, 4030 – 4043.
[6] a) Y. Sawada, K. Matsumoto, S. Kondo, H.
Watanabe, T. Ozawa, K. Suzuki, B. Saito, T.
Katsuki, Angew. Chem. 2006, 118, 3558 – 3560;
Angew. Chem. Int. Ed. 2006, 45, 3478 – 3480;
b) Y. Shimada, S. Kondo, Y. Ohara, K. Matsu-
moto, T. Katsuki, Synlett 2007, 2445 – 2447.
[7] Prepared by NaBH4 reduction of the corre-
Scheme 6. Panel of diastereomeric products formed by total synthesis. The preparation
of the individual isomers follows the sequence outlined for (13R)-50 in the text; details
are given in the Supporting Information. Arbitrary numbering Scheme used in Table 1.
sponding salen ligand, which in turn was pre-
pared according to: J. F. Larrow, E. N. Jacobsen,
Org. Synth. 1998, 75, 1 – 11.
[8] a) A. Fꢁrstner, M. Mꢂndez, Angew. Chem. 2003,
115, 5513 – 5515; Angew. Chem. Int. Ed. 2003, 42,
5355 – 5357; b) B. D. Sherry, A. Fꢁrstner, Acc. Chem. Res. 2008,
41, 1500 – 1511.
larger discrepancies encountered throughout the carbon
skeleton (Table 1).
To our dismay, we must hence conclude that none of the
four isomers prepared during this investigation reproduces
the published data of the natural product well enough to claim
identity. The surprising scattering of the observed deviances
makes it impossible at this point to localize the cause for the
mismatch. Since the observed shift differences are small, the
constitution originally assigned to leiodolide B is most likely
correct and the variances must stem from a rather subtle
structural detail.[27] We cannot even exclude that they are
artifacts caused by the possible low resolution of the original
spectra recorded on minute amounts of a not perfectly pure
sample (the published 13C NMR data were recorded at
75 MHz with 0.8 mg of material).[28] Since neither the original
13C NMR spectra nor authentic leiodolide B could be made
available to us, we have to leave this aspect open.[29,30] Yet, we
are hopeful that the conquest of the more carefully charac-
terized sister compound leiodolide A and its conversion into
leiodolide B by a biomimetic path will resolve the puzzle.
[9] a) F. Volz, N. Krause, Org. Bioorg. Chem. 2007, 5, 1519 – 1521;
b) C. Deutsch, A. Hoffmann-Rꢃder, A. Domke, N. Krause,
Synlett 2007, 737 – 740.
[10] J. A. Marshall, K. G. Pinney, J. Org. Chem. 1993, 58, 7180 – 7184.
[11] P. Crotti, V. Di Bussolo, L. Favero, F. Macchia, M. Pineschi, Eur.
J. Org. Chem. 1998, 1675 – 1686.
[12] A. Mengel, O. Reiser, Chem. Rev. 1999, 99, 1191 – 1224.
[13] D. Muri, N. Lohse-Fraefel, E. M. Carreira, Angew. Chem. 2005,
117, 4104 – 4106; Angew. Chem. Int. Ed. 2005, 44, 4036 – 4038.
[14] The configuration of the newly formed alcohol center was
confirmed by applying the modified Mosher method to both
isomers; see I. Ohtani, T. Kusumi, Y. Kashman, H. Kakisawa, J.
Am. Chem. Soc. 1991, 113, 4092 – 4096.
[15] Bromide 24 was not sufficiently reactive to alkylate the benzyl
lithium reagent derived from oxazole 39.
[16] D. Seebach, R. Naef, G. Calderari, Tetrahedron 1984, 40, 1313 –
1324.
[17] a) J.-F. Betzer, F. Delaloge, B. Muller, A. Pancrazi, J. Prunet, J.
Org. Chem. 1997, 62, 7768 – 7780; b) B. H. Lipshutz, E. L.
Ellsworth, S. H. Dimock, D. C. Reuter, Tetrahedron Lett. 1989,
30, 2065 – 2068.
[18] A. Fꢁrstner, C. Nevado, M. Waser, M. Tremblay, C. Chevrier, F.
Received: September 17, 2010
Published online: November 16, 2010
´
Teply, C. Aissa, E. Moulin, O. Mꢁller, J. Am. Chem. Soc. 2007,
129, 9150 – 9161.
[19] D. A. Evans, J. Bartroli, T. L. Shih, J. Am. Chem. Soc. 1981, 103,
2127 – 2129.
[20] A. B. Smith III, K. B. Minbiole, S. Freeze, Synlett 2001, 1739 –
1742.
[21] A. Fꢁrstner, J.-A. Funel, M. Tremblay, L. C. Bouchez, C.
Nevado, M. Waser, J. Ackerstaff, C. C. Stimson, Chem.
Commun. 2008, 2873 – 2875.
[22] For applications, see a) A. Fꢁrstner, J. Ackerstaff, Chem.
Commun. 2008, 2870 – 2872; b) A. Fꢁrstner, L. C. Bouchez, J.-
A. Funel, V. Liepins, F.-H. Porꢂe, R. Gilmour, F. Beaufils, D.
Laurich, M. Tamiya, Angew. Chem. 2007, 119, 9425 – 9430;
Keywords: Homogeneous catalysis · marine natural products ·
NMR spectroscopy · structure elucidation · total synthesis
.
[1] Halichondrin and discodermolide are examples of anticancer
agents derived from deep water species; for a general survey, see
Drugs from the Sea (Ed.: N. Fusetani), Karger, Basel, 2000.
[2] J. S. Sandler, P. L. Colin, M. Kelly, W. Fenical, J. Org. Chem.
2006, 71, 7245 – 7251; correction: J. S. Sandler, P. L. Colin, M.
Kelly, W. Fenical, J. Org. Chem. 2006, 71, 8684.
308
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 304 –309

Angew. Chem. Int. Ed. 2007, 46, 9265 – 9270; c) A. Francais, A.
Leyva, G. Etxebarria-Jardi, S. V. Ley, Org. Lett. 2010, 12, 340 –
343; d) R. S. Paley, K. E. Berry, J. M. Liu, T. T. Sanan, J. Org.
Chem. 2009, 74, 1611 – 1620.
[28] The 1H NMR spectrum of leiodolide B depicted in the Support-
ing Information of the isolation paper clearly shows the presence
of an impurity (ca. 5–10%).
[29] For a review on natural products of mistaken identity, see K. C.
Nicolaou, S. A. Snyder, Angew. Chem. 2005, 117, 1036 – 1069;
Angew. Chem. Int. Ed. 2005, 44, 1012 – 1044.
[23] D. A. Evans, D. M. Fitch, T. E. Smith, V. Cee, J. Am. Chem. Soc.
2000, 122, 10033 – 10046.
[24] More basic fluoride sources also cleave the TIPS ether and cause
epoxide formation by intramolecular substitution of the adjacent
bromide.
[25] a) J. Inanaga, K. Hirata, H. Saeki, T. Katsuki, M. Yamaguchi,
Bull. Chem. Soc. Jpn. 1979, 52, 1989 – 1993; b) A. Parenty, X.
Moreau, J.-M. Campagne, Chem. Rev. 2006, 106, 911 – 939.
[26] K. C. Nicolaou, A. A. Estrada, M. Zak, S. H. Lee, B. S. Safina,
Angew. Chem. 2005, 117, 1402 – 1406; Angew. Chem. Int. Ed.
2005, 44, 1378 – 1382.
[27] One possibility is that the a-hydroxy-a-methyl carboxylic acid
terminus in 2 is R rather than S configured (see Ref. [4]),
although it is not clear that this change would have an impact on
the chemical shift of so many remote C atoms. Likewise, Table 1
shows both olefinic regions as “hotspots”, but the chemical shift
differences are not large enough to suggest misassigned geo-
metries.
[30] For other total syntheses from our research group which led to
the revision of the originally proposed structures of bioactive
natural products, see a) P. Buchgraber, T. N. Snaddon, C. Wirtz,
R. Mynott, R. Goddard, A. Fꢁrstner, Angew. Chem. 2008, 120,
8578 – 8582; Angew. Chem. Int. Ed. 2008, 47, 8450 – 8454; b) A.
Fꢁrstner, L. C. Bouchez, L. Morency, J.-A. Funel, V. Liepins, F.-
H. Porꢂe, R. Gilmour, D. Laurich, F. Beaufils, M. Tamiya, Chem.
Eur. J. 2009, 15, 3983 – 4010; c) A. Fꢁrstner, K. Radkowski, H.
Peters, G. Seidel, C. Wirtz, R. Mynott, C. W. Lehmann, Chem.
Eur. J. 2007, 13, 1929 – 1945; d) A. Fꢁrstner, M. Albert, J.
Mlynarski, M. Matheu, E. DeClercq, J. Am. Chem. Soc. 2003,
125, 13132 – 13142; e) A. Fꢁrstner, K. Radkowski, C. Wirtz, R.
Goddard, C. W. Lehmann, R. Mynott, J. Am. Chem. Soc. 2002,
124, 7061 – 7069.
Angew. Chem. Int. Ed. 2011, 50, 304 –309
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
309