What is amphidinolide V? Report on a likely conquest

Article abstract of DOI:10.1002/anie.200701640

(Chemical Equation Presented) The awesome power of metathesis is reflected in the synthesis of the proposed structure of the cytotoxic natural product amphidinolide V, as well as of all other stereomers containing a trans-epoxide unit. It can be concluded

Full text of DOI:10.1002/anie.200701640

Angewandte
Chemie
DOI: 10.1002/anie.200701640
Total Synthesis
What is Amphidinolide V? Report on a Likely Conquest**
Alois Fürstner,* Oleg Larionov, and Susanne Flügge
More than 30 macrolides have been isolated to date from
laboratory-cultured dinoflagellates of the Amphidinium sp.[1]
Despite their common origin and a uniformly high cytotox-
icity against various cancer cell lines, the individual “amphi-
dinolides” are structurally quite diverse and have therefore
inspired many synthetic approaches toward them.[2,3] One of
the rarest members of this series is amphidinolide V
(0.00005% of the wet weight of the harvested algae), for
which structure 1 (Scheme 1) was proposed on the basis of the
spectroscopic data of a sample of no more than 0.2 mg.[4]
A with ethylene,[7] the overall assembly process should gain
substantial flexibility (Scheme 1). At the outset, however, it
was by no means clear if the available alkyne metathesis
catalysts tolerate the labile trans-configured vinylepoxide and
the allylic alcohol residing in the cyclization precursor B,
while serious selectivity issues might plague the late-stage
enyne metathesis event proposed to complete the polyunsa-
turated framework.
The synthetic venture commenced with a copper-cata-
lyzed opening of epoxide 2 with the Grignard reagent 3,
followed by the addition of bromine for silyl exchange to
afford product 5 in excellent yield on a large scale (Scheme 2).
Scheme 1. Proposed structure and retrosynthetic analysis of amphidi-
nolide V.
Scheme 2. a) CuCN (10%), THF, 08C!RT, 99%; b) 1. Br2, CH2Cl2,
ꢀ788C; 2. NaOMe, MeOH, ꢀ208C; 3. HOAc, 91% (overall); c) 6,
Pd(OAc)2 (10%), dppf (10%), tBuNH2, THF, reflux (sealed tube),
84%; d) 4-hexynoic acid, EDC, 1-hydroxy-7-azabenzotriazole, (iPr)2NEt,
DMAP, CH2Cl2/DMF (4:1), 95%; e) PPTS (cat.), iPrOH, 708C, 98%;
f) d-(ꢀ)-DET (40%), Ti(OiPr)4 (40%), tBuOOH, MS (4 Š), CH2Cl2,
ꢀ258C, 77%; g) Dess–Martin periodinane, NaHCO3, CH2Cl2, 90%.
dppf=1,1’-bis(diphenylphosphanyl)ferrocene, EDC=N-(3-dimethylami-
nopropyl)-N’-ethyl-carbodiimide, DMAP=4-dimethylaminopyridine,
PPTS=pyridinium p-toluenesulfonate, DET=diethyl tartrate,
TBS=tert-butyldimethylsilyl, THP=tetrahydropyranyl, MS=molecular
sieves.
As part of our ongoing program on bioactive natural
products of marine origin,[5] amphidinolide V was chosen as a
target for further evaluation. To this end, a flexible entry was
sought that would provide meaningful amounts of this scarce
metabolite for testing, and allow for systematic modification
of its molecular edifice at a later stage. It was envisaged that
the distinctive s-trans diene unit of 1, spanned by the vicinal
exo-methylene branches at C-4 and C-5, would provide a
unique opportunity in this regard. If formed by a sequence of
ring-closing alkyne metathesis (RCAM)[6] and subsequent
intermolecular enyne metathesis of the resulting cycloalkyne
This compound was subjected to a Suzuki reaction with
trifluoroborate 6.[8] In contrast to the recommended protocol
that emphasizes the importance of aqueous media for
reactions of alkenyl trifluoroborates as Suzuki donors,[9] this
particular cross-coupling reaction would not proceed well
unless performed in anhydrous THF with tBuNH2 as the base.
Under these modified conditions, however, a yield of 84%
was secured in batches that produced up to 10 g of the desired
diene 7. Esterification with 4-hexynoic acid, unmasking of the
allylic alcohol terminus, and Sharpless epoxidation[10] with d-
(ꢀ)-DET furnished alcohol 10, which was oxidized under
[*] Prof. A. Fürstner, Dr. O. Larionov, Dipl.-Chem. S. Flügge
Max-Planck-Institut für Kohlenforschung
45470 Mülheim an der Ruhr (Germany)
Fax: (+49)208-306-2994
E-mail: fuerstner@mpi-muelheim.mpg.de
[**] Generous financial support from the MPG and the Fonds der
Chemischen Industrie (KekulØ fellowship to S.F.) is gratefully
acknowledged. We thank Prof. J.-I. Kobayashi for an exchange of
information.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2007, 46, 5545 –5548
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5545

Communications
buffered conditions to give epoxy-aldehyde 11 in high overall
yield.
Initial attempts to subject this labile aldehyde to chela-
base in DME/DMPU (50:1); this particular medium led to
substantially higher E/Z ratios (ca. 10:1) than were observed
in THF (ca. 3:1). The isomers were readily separable, thus
securing a good supply of 1 in pure form (Table 1).
tion-controlled additions of various organometallic reagents
derived from bromide 13 were unrewarding. Gratifyingly
though, it was found that 13 reacts with ZnBr2 and Li sand
under ultrasonication to produce the bis(alkenyl)zinc reagent
14.[11] Slow addition of aldehyde 11 to a salt-free solution of 14
in toluene in the presence of (+)-N-methylephedrine fur-
nished alcohol 15 in good yield and appreciable diastereose-
lectivity (up to 7:1) in favor of the desired syn adduct
(Scheme 3). Subsequent O-silylation provided diyne 16 as
the substrate for the envisaged ring closure by RCAM.
Table 1: Reference data set of compound 1.
[a]2D0 =ꢀ9.3 degcm3 gꢀ1 dmꢀ1 (CHCl3, c=0.6 gcmꢀ3); 1H NMR (CDCl3,
400 MHz): d=1.83 (3H, s), 2.12 (1H, d, J=5.4 Hz, OH), 2.39–2.41
(2H, m), 2.43–2.47 (2H, m), 2.50–2.67 (1H, m), 2.70–2.77 (1H, m),
2.79–2.82 (1H, m), 2.82–2.85 (2H, m), 3.11 (1H, d, J=16.4 Hz), 3.25
(1H, d, J=16.4 Hz), 3.46 (1H, brs), 4.00 (1H, dd, J=5.4, 5.8 Hz), 4.89
(2H, s), 4.93 (1H, s), 5.08 (1H, s), 5.10 (1H, s), 5.13 (1H, s), 5.16 (1H,
s), 5.19 (1H, s), 5.24 (1H, s), 5.40–5.43 (1H, m), 5.43–5.45 (1H, m),
5.46 (1H, s), 5.60 (1H, dt, J=6.7, 15.6 Hz), 5.67–5.77 (1H, m),
6.12 ppm (1H, d, J=15.6 Hz); 1H NMR (C6D6, 400 MHz): d=1.73 (3H,
s), 1.93 (1H, brs, -OH), 2.01 (1H, ddd, J=4.0, 5.6, 14.4 Hz), 2.08–2.18
(2H, m), 2.24 (1H, dd, J=2.6, 14.1 Hz), 2.34 (1H, dt, J=4.2, 14.1 Hz),
2.52 (1H, ddd, J=4.0, 12.1, 14.1 Hz), 2.64 (2H, t, J=6.5 Hz), 2.77 (1H,
dd, J=6.5, 2.1 Hz), 2.86 (1H, d, J=16.1 Hz), 3.01 (1H, J=16.1 Hz),
3.50 (1H, brs), 3.93 (1H, brs), 4.71 (1H, s), 4.89 (2H, brs), 4.92 (2H,
s), 4.95 (1H, s), 4.98 (1H, s), 5.04 (1H, s), 5.28 (1H, s), 5.42 (1H, dd,
J=6.6, 15.4 Hz), 5.52 (1H, dt, J=6.6, 15.6 Hz), 5.53–5.59 (1H, m), 5.62
(1H, s), 5.69 (1H, dt, J=6.6, 15.4 Hz), 6.15 ppm (1H, d, J=15.6 Hz);
13C NMR (CDCl3, 100 MHz): d=18.6, 30.5, 33.7, 35.1, 39.1, 39.2, 58.0,
63.3, 71.4, 74.3, 114.0, 114.8, 114.9, 115.1, 115.1, 127.5, 128.1, 132.0,
134.1, 140.8, 141.9, 142.1, 144.6, 145.0, 171.9 ppm; IR (NaCl): n˜ =3476,
3082, 2923, 2856, 1735, 1670, 1639, 1608, 1453, 1374, 1262, 1239, 1156,
1089, 969, 901, 679 cmꢀ1; MS (ESI+): 419.3 [M+Na]+; HRMS (ESI+)
calcd for [M+Na]+: 419.2193, found: 419.2192.
1
As can be seen from Table 2, all the H NMR data of
synthetic 1 recorded in [D6]benzene are in excellent agree-
ment with the reported spectra of the natural product;
Table 2: Deviation of selected NMR shifts of synthetic 1 and isomers
[a]
from the values reported for amphidinolide V.
ꢁ
Scheme 3. a) MeC CMgBr, CuBr·Me2S, Et2O, 99%; b) Li, ZnBr2, THF,
DdH
DdC
1[b]
DdC
28[b]
No.
1[b]
0.00
1[c]
24[b]
0.8
2.0
5.1
30[b]
3.0
08C, ultrasound; c) 11, toluene, (+)-N-methylephedrine (60%),
ꢀ258C, 69%; d) TBSCl, imidazole, CH2Cl2, 108C, 79%; e) 20 (20%),
CH2Cl2/toluene, 858C, 66%; f) 21 (2%), C2H4 (1.8 atm), toluene,
458C, 90%; g) PPTS (cat.), MeOH, 62%; h) Dess–Martin periodinane,
NaHCO3, CH2Cl2; i) 22, KHMDS, DME/DMPU, ꢀ788C!RT, 57%
(over both steps, E:Zꢂ10:1); j) TASF, DMF, ꢀ58C, 82%;
KHMDS=potassium hexamethyldisilazide; TASF=tris(dimethylami-
no)sulfonium difluorotrimethylsilicate, Mes=mesityl=2,4,6-trimethyl-
phenyl.
2
4
8
9
10
12
14
ꢀ0.04
–-
0.2
0.1
0.3
0.0
0.2
0.1
0.4
0.6
1.5
–-
ꢀ0.3
ꢀ0.50
0.00
0.00
0.00
0.00
ꢀ0.04
ꢀ0.05
ꢀ0.05
ꢀ0.05
ꢀ0.05
ꢀ0.7
ꢀ5.4
ꢀ1.3
ꢀ7.2
1.3
0.7
ꢀ3.4
ꢀ0.9
ꢀ2.1
ꢀ3.8
2.0
2.0
ꢀ5.7
1.1
[a] For the full data set, see the Supporting Information. [b] CDCl3.
[c] C6D6; Dd=d(synthetic isomer)ꢀd(natural product).[4]
This transformation proceeded nicely, even though the
resulting product is fairly strained, thus attesting to the
excellent application profile of the molybdenum-based cata-
lyst formed in situ upon activation of complex 20 with CH2Cl2
as previously described by our research group.[12] Equally
gratifying was the outcome of the subsequent enyne meta-
thesis between cycloalkyne 17 thus formed and ethylene gas,[7]
which installed the vicinal one-carbon branches characteristic
of amphidinolide V without any appreciable interference
from the preexisting double bonds. Attachment of the lateral
chain involved a routine protecting-group and oxidation-state
management as well as a Julia–Kocienski olefination[13] with
sulfone 22,[8] which was best performed with KHMDS as the
likewise, the 13C NMR data recorded in CDCl3 are also in
good accord.[4,8] Surprisingly, however, a single resonance in
the 1H NMR spectrum recorded in CDCl3, assigned to H-8, is
displaced by d = 0.50 ppm (!), whereas all the other shifts and
coupling constants show an almost perfect match in this
particular medium. Unfortunately, we were unable to resolve
this rather suspicious singular discrepancy at this point
because neither an authentic sample nor the original spectra
of amphidinolide V could be made available to us.
Therefore, it was necessary to obtain a more comprehen-
sive data set for further comparison, which required prepa-
5546
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 5545 –5548

Angewandte
Chemie
ration of all other conceivable diastereomers that incorporate
a trans-epoxide entity but differ in the configuration of the
individual tetrahedral centers.[14] To this end, it sufficed to
replace d-(ꢀ)-DET by l-(+)-DET in the Sharpless epoxida-
tion of alcohol 9, and to pursue an analogous assembly
process from there on to obtain isomer 24 (Scheme 4).[8]
is important to note, however, that the RCAM reactions (66–
93%) and the ensuing enyne metatheses (73–96%) were
highly efficient for all isomers prepared during this venture.[8]
The spectral characteristics of 24, 28, as well as 30 showed
a clear mismatch with the published data and hence none of
them represents the actual target (Table 2). Therefore, we
must conclude that the constitution and relative stereochem-
istry of amphidinolide V is most likely represented by
structure 1, as originally proposed.[4] Whether the conspicu-
ous mismatch of a single resonance in only one of the two
1
reported H NMR spectra is due to a typographical error in
the original data set can only be clarified after re-isolation of
the natural product from the producing Amphidinium strain,
which is beyond our possibility.
Received: April 13, 2007
Published online: June 20, 2007
Keywords: alkynes · macrolides · metathesis · natural products ·
.
structure elucidation
[1] J. Kobayashi, M. Tsuda, Nat. Prod. Rep. 2004, 21, 77 – 93.
[2] For the synthesis of various amphidinolides, see Ref. [3] and the
following for leading references: a) D. R. Williams, W. S. Kissel,
J. Am. Chem. Soc. 1998, 120, 11198 – 11199; b) D. R. Williams,
B. J. Myers, L. Mi, Org. Lett. 2000, 2, 945 – 948; c) D. R. Williams,
K. G. Meyer, J. Am. Chem. Soc. 2001, 123, 765 – 766; d) H. W.
Lam, G. Pattenden, Angew. Chem. 2002, 114, 526 – 529; Angew.
Chem. Int. Ed. 2002, 41, 508 – 511; e) R. E. Maleczka, Jr., L. R.
Terrell, F. Geng, J. S. Ward, Org. Lett. 2002, 4, 2841 – 2844;
f) B. M. Trost, P. E. Harrington, J. D. Chisholm, S. T. Wrobleski,
J. Am. Chem. Soc. 2005, 127, 13598 – 13610; g) B. M. Trost, S. T.
Wrobelski, J. D. Chisholm, P. E. Harrington, M. Jung, J. Am.
Chem. Soc. 2005, 127, 13589 – 13597; h) P. Va, W. R. Roush, J.
Am. Chem. Soc. 2006, 128, 15960 – 15961; i) C. H. Kim, H. J. An,
W. K. Shin, W. Yu, S. K. Woo, S. K. Jung, E. Lee, Angew. Chem.
2006, 118, 8187 – 8189; Angew. Chem. Int. Ed. 2006, 45, 8019 –
8021; j) K. C. Nicolaou, W. E. Brenzovich, P. G. Bulger, T. M.
Francis, Org. Biomol. Chem. 2006, 4, 2119 – 2157; k) K. C.
Nicolaou, P. G. Bulger, W. E. Brenzovich, Org. Biomol. Chem.
2006, 4, 2158 – 2183; l) B. M. Trost, J. P. N. Papillon, T. Nuss-
baumer, J. Am. Chem. Soc. 2005, 127, 17921 – 17937; m) A. K.
Ghosh, G. Gong, J. Org. Chem. 2006, 71, 1085 – 1093; n) E. A.
Colby, K. C. OꢀBrien, T. F. Jamison, J. Am. Chem. Soc. 2005, 127,
4297 – 4307; o) E. A. Colby, T. F. Jamison, Org. Biomol. Chem.
2005, 3, 2675 – 2684; p) L.-S. Deng, X.-P. Huang, G. Zhao, J. Org.
Chem. 2006, 71, 4625 – 4635.
Scheme 4. a) l-(+)-DET (40%), Ti(OiPr)4 (40%), tBuOOH, MS (4 Š),
CH2Cl2, ꢀ208C, 92%; b) Dess–Martin periodinane, NaHCO3, CH2Cl2;
c) Grignard reagent 25, THF, ꢀ158C, 86%; d) Dess–Martin period-
inane, CH2Cl2; e) NaBH4, CaCl2, MeOH, 08C, 71% (over 2 steps,
d.r.=11:1); for further details, see the Supporting Information.
Access to isomer 28 containing an anti-configured hydroxy-
epoxide unit was gained by a chelation-controlled reduction
of ketone 26. The combination of NaBH4 and CaCl2 served
well for this purpose,[15] which is thought to lock a transition
state of type F without damaging the fragile epoxide unit. The
completion of the synthesis then followed the blueprint
outlined above for the diastereomeric series.[8] Chelate-
controlled reduction of the diastereomeric epoxyketone 29
gave access to the last isomer 30 (Scheme 5); this latter
compound turned out to be much less stable than its
congeners, and readily decomposed even when kept cold. It
[3] a) A. Fürstner, E. Kattnig, O. Lepage, J. Am. Chem. Soc. 2006,
128, 9194 – 9204; b) C. Aïssa, R. Riveiros, J. Ragot, A. Fürstner,
J. Am. Chem. Soc. 2003, 125, 15512 – 15520; c) O. Lepage, E.
Kattnig, A. Fürstner, J. Am. Chem. Soc. 2004, 126, 15970 –
15971; d) A. Fürstner, C. Aïssa, R. Riveiros, J. Ragot, Angew.
Chem. 2002, 114, 4958 – 4960; Angew. Chem. Int. Ed. 2002, 41,
4763 – 4766.
[4] T. Kubota, M. Tsuda, J. Kobayashi, Tetrahedron Lett. 2000, 41,
713 – 716.
´
[5] a) A. Fürstner, C. Nevado, M. Tremblay, C. Chevrier, F. Teply, C.
Aïssa, M. Waser, Angew. Chem. 2006, 118, 5969 – 5974; Angew.
Chem. Int. Ed. 2006, 45, 5837 – 5842; b) A. Fürstner, C. Aïssa, C.
´
Chevrier, F. Teply, C. Nevado, M. Tremblay, Angew. Chem. 2006,
Scheme 5. a) Grignard reagent 25, THF, ꢀ158C, 70%; b) Dess–Martin
periodinane, CH2Cl2; 97%; for further details, see the Supporting
Information.
118, 5964 – 5969; Angew. Chem. Int. Ed. 2006, 45, 5832 – 5837;
c) A. Fürstner, M. M. Domostoj, B. Scheiper, J. Am. Chem. Soc.
2006, 128, 8087 – 8094; d) A. Fürstner, L. Turet, Angew. Chem.
Angew. Chem. Int. Ed. 2007, 46, 5545 –5548
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5547

Communications
2005, 117, 3528 – 3532; Angew. Chem. Int. Ed. 2005, 44, 3462 –
[10] R. A. Johnson, K. B. Sharpless in Catalytic Asymmetric Syn-
thesis, 2nd ed. (Ed.: I. Ojima), Wiley-VCH, New York, 2000,
pp. 231 – 280.
[11] T. Shibata, K. Nakatsui, K. Soai, Inorg. Chim. Acta 1999, 296,
33 – 36.
[12] a) A. Fürstner, C. Mathes, C. W. Lehmann, J. Am. Chem. Soc.
1999, 121, 9453 – 9454; b) A. Fürstner, C. Mathes, C. W. Leh-
mann, Chem. Eur. J. 2001, 7, 5299 – 5317; c) A. Fürstner, K.
Grela, C. Mathes, C. W. Lehmann, J. Am. Chem. Soc. 2000, 122,
11799 – 11805.
3466; e) A. Fürstner, M. D. B. Fenster, B. Fasching, C. Godbout,
K. Radkowski, Angew. Chem. 2006, 118, 5632 – 5636; Angew.
Chem. Int. Ed. 2006, 45, 5506 – 5510; f) A. Fürstner, M. D. B.
Fenster, B. Fasching, C. Godbout, K. Radkowski, Angew. Chem.
2006, 118, 5636 – 5641; Angew. Chem. Int. Ed. 2006, 45, 5510 –
5515; g) A. Fürstner, D. De Souza, L. Turet, M. D. B. Fenster, L.
Parra-Rapado, C. Wirtz, R. Mynott, Chem. Eur. J. 2007, 13, 115 –
134; h) A. Fürstner, D. Kirk, M. D. B. Fenster, C. Aïssa, D.
De Souza, C. Nevado, C. T. T. Tuttle, W. Thiel, O. Müller, Chem.
Eur. J. 2007, 13, 135 – 149, and references therein.
[13] P. R. Blakemore, J. Chem. Soc. Perkin Trans. 1 2002, 2563 – 2585.
[14] Not considered were isomers with different double-bond geo-
metries as well as isomers containing a cis-epoxide unit. The
[6] a) A. Fürstner, G. Seidel, Angew. Chem. 1998, 110, 1758 – 1760;
Angew. Chem. Int. Ed. 1998, 37, 1734 – 1736; b) A. Fürstner, O.
Guth, A. Rumbo, G. Seidel, J. Am. Chem. Soc. 1999, 121, 11108 –
11113; c) for a review, see A. Fürstner, P. W. Davies, Chem.
Commun. 2005, 2307 – 2320.
[7] a) K. Tonogaki, M. Mori, Tetrahedron Lett. 2002, 43, 2235 – 2238;
b) M. Mori, Adv. Synth. Catal. 2007, 349, 121 – 135; c) S. T. Diver,
A. J. Giessert, Chem. Rev. 2004, 104, 1317 – 1382.
3
presence of the latter was ruled out on the basis of the JH9,H10
coupling constant of 2.2 Hz reported for natural amphidinoli-
de V, which is characteristic of a trans-epoxide, see Ref. [4]. This
assignment was corroborated by the NMR data of the synthetic
trans epoxides prepared during this investigation, the corre-
sponding coupling constants of which were invariably about
2 Hz.
[8] For details, see the Supporting Information.
[9] a) G. A. Molander, L. A. Felix, J. Org. Chem. 2005, 70, 3950 –
3956; b) G. A. Molander, N. Ellis, Acc. Chem. Res. 2007, 40, 275 –
286.
[15] M. Taniguchi, H. Fujii, K. Oshima, K. Utimoto, Tetrahedron
1995, 51, 679 – 686.
5548
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 5545 –5548