Total synthesis of floresolide B and Δ6,7-Z-floresolide B

Article abstract of DOI:10.1039/b517385j

The total syntheses of the cytotoxin marine natural product floresolide B (1) and its Δ^6,7-Z isomer (2) have been achieved through an olefin metathesis-based strategy. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b517385j

View Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Total synthesis of floresolide B and D6,7-Z-floresolide B{
K. C. Nicolaou*ab and Hao Xua
Received (in Cambridge, UK) 7th December 2005, Accepted 6th January 2006
First published as an Advance Article on the web 20th January 2006
DOI: 10.1039/b517385j
guarantees success in producing a ring, or even if it did, with the
The total syntheses of the cytotoxin marine natural product
floresolide B (1) and its D6,7-Z isomer (2) have been achieved
through an olefin metathesis-based strategy.
proper size and stereochemistry with regard to double bond
geometry and atropisomerism. Rather, it was hoped that
experimental exploration of this olefin metathesis-based approach
would lead to some interesting observations from which a final
strategy towards floresolide B might emerge.
Floresolide B (1, Fig. 1) is a recently discovered natural product
possessing an intriguing molecular architecture.1 Isolated from an
extract of an ascidian of genus Aplidium collected in Pungu Besar,
Flores Island, floresolide exhibits cytotoxicity against KB tumor
cells and boasts in its structure an aromatic ring weaved within a
[10] metacyclophane bridged by a 7-membered lactone. Its
structure was deduced from its spectral data and from an X-ray
crystallographic analysis of a related natural product. The latter
structural analysis revealed only one atropisomer, that shown in
Fig. 1, despite the possibility for two such isomers. Although
critical, especially from the synthetic point of view, the atropi-
somerism of floresolide B and its siblings, floresolides A and C,
was not mentioned in the isolation and structural elucidation
report.1 In this communication, we disclose a total synthesis of
floresolide B (1) and its D6,7-Z isomer (2, Fig. 1) by an expedient
olefin metathesis-based strategy that delivers both as single
atropisomers.
With this anticipation in mind, we set out to first construct
substrate 3 as shown in Scheme 1. Thus, commercially available
2,5-dihydroxybenzaldehyde (5) was bis-benzoylated (Bz2O, Et3N)
and selectively mono-debenzoylated (K2CO3) by a modification of
a literature procedure3 to afford mono-benzoate hydroxy benzal-
dehyde derivative 6 in 70% overall yield for the two steps.
Sequential o-bromination (NBS, 95%), MOM-ether formation
(NaH, MOMCl, 90%) and debenzoylation (NaOMe, THF/
MeOH, 80%) furnished bromophenol 7. The latter compound
(7) was then protected with MOMCl in the presence of NaH, and
then reduced with NaBH4 to afford the corresponding benzylic
alcohol, which was converted to its TIPS ether derivative 8
(TIPSCl, imidazole) in 65% overall yield for the three steps.
Exposure of aryl bromide 8 to isopropenyl acetate and tri-
n-butyltin methoxide4 in the presence of catalytic amounts of
Pd2(dba)3 and Buchwald ligand A5 in toluene at 90 uC generated
aryl methyl ketone 9 in 85% yield.6 Alkylation7 of the thermo-
dynamic enolate derived from ketone 9 and NaH with 4-iodo-2-
methylbut-1-ene then afforded compound 10 (73% yield),8 whose
methylenation proceeded smoothly under standard Wittig condi-
tions to give the expected bis-olefin in 75% yield. Removal of the
TIPS group from this bis-olefin (TBAF, 80% yield) allowed its
conversion to benzylic bromide 12 (MsCl, Et3N, LiBr, 82% yield)9
through the corresponding benzylic alcohol 11. Extension of the
short aryl-bound chain of the growing molecule required vinyl
stannane B, whose construction from the corresponding vinyl
iodide relied on lithium–iodine exchange (tBuLi) followed by the
lithium–tin exchange (Me3SnCl). Thus 12 and B were united under
the influence of Pd2(dba)3 and AsPh3, through a Stille reaction,
leading to the olefin metathesis precursor 13 in 85% yield.10 Note,
the choice of the THP as a protecting group in the construction of
vinyl stannane B from the iodide was crucial for its success.
The behavior of polyolefin 13 and its relative, ketoester 14, in
the presence of the two Grubbs olefin metathesis catalysts (Grubbs
cat. I and Grubbs cat. II) is summarized in Scheme 2. Thus, upon
treatment with Grubbs cat. I in refluxing CH2Cl2 for 48 h,
substrate 13 was converted to a 1 : 1 mixture of E : Z dimeric
olefins 15Z and 15E in 75% combined yield, despite the high
dilution conditions (0.5 mM) employed. When Grubbs cat. II was
employed under the same conditions, the reaction was complete
within 48 h, and the two dimeric olefins 15Z : 15E were obtained,
again in a 1 : 1 ratio, and in 35% combined yield. Along with them,
however, this time the two hydride-shift products 16Z : 16E were
also obtained (1 : 1 ratio, 20% combined yield).11 The same
A cursory inspection of the molecule of floresolide B (1) reveals
the ring closing olefin metathesis2 substrates 3 or 4 as potential
precursors to this considerably strained molecule, as shown
retrosynthetically in Fig. 1. A priori, however, neither substrate
Fig. 1 Molecular structures and retrosynthetic analysis of floresolide B
(1) and D6,7-Z-floresolide B (2).
aDepartment of Chemistry and The Skaggs Institute for Chemical
Biology, The Scripps Research Institute, 10550 North Torrey Pines
Road, La Jolla, California 92037, USA. E-mail: kcn@scripps.edu
bDepartment of Chemistry and Biochemistry, University of California
San Diego, 9500 Gilman Drive, La Jolla, California 92093, USA
{ Electronic supplementary information (ESI) available: Selected data for
synthetic 2. See DOI: 10.1039/b517385j
600 | Chem. Commun., 2006, 600–602
This journal is ß The Royal Society of Chemistry 2006

View Online
Scheme 1 Synthesis of advanced intermediate 13. Reagents and condi-
tions: (a) Bz2O (2.1 equiv), Et3N (2.3 equiv), 4-DMAP (0.1 equiv), CH2Cl2,
25 uC, 12 h, 85%; (b) K2CO3 (1.5 equiv), THF : MeOH (1 : 1), 0 uC, 2 h,
83%; (c) NBS (1.05 equiv), CH2Cl2, 25 uC, 1 h, 95%; (d) NaH (2.0 equiv),
MOMCl (1.0 equiv), DMF, 25 uC, 2 h, 90%; (e) NaOMe (1.1 equiv),
THF : MeOH (1 : 1), 240 uC, 2 h, 80%; (f) NaH (2.0 equiv), MOMCl
(1.0 equiv), DMF, 25 uC, 2 h, 85%; (g) NaBH4 (2.5 equiv), MeOH, 0 uC,
30 min, 90%; (h) TIPSCl (1.5 equiv), imidazole (2.0 equiv), 4-DMAP
(0.1 equiv), DMF, 25 uC, 12 h, 85%; (i) isopropenyl acetate (1.4 equiv),
nBu3SnOMe (1.4 equiv), Pd2(dba)3 (0.1 equiv), Buchwald ligand A
(0.4 equiv), toluene, 90 uC, 2 h, 85%; (j) NaH (1.2 equiv), THF, 0 uC, 0.5 h;
then 4-iodo-2-methylbut-1-ene (1.6 equiv), 0 uC, 4 h, 73%; (k) CH3P+Ph3
Br2 (1.5 equiv), NaHMDS (1.5 equiv), THF, 0 uC, 1 h; then 10 added to
the generated ylide, 0 uC, 0.5 h, 75%; (l) TBAF (1.1 equiv), THF, 0 uC,
15 min, 80%; (m) LiBr (20 equiv), Et3N (2.5 equiv), MsCl (1.5 equiv), THF,
210 uC, 30 min, 82%; (n) Pd2(dba)3 (0.1 equiv), AsPh3 (0.8 equiv), THF,
25 uC, 20 min; then vinyl tin compound B (1.2 equiv), 12 (1.0 equiv), 70 uC,
1 h, 85%. Bz = benzoyl, 4-DMAP = 4-dimethylaminopyridine, NBS =
N-bromosuccinimide, MOM = methoxymethyl, DMF = N,N-dimethyl-
Scheme 2 Attempted ring closing metathesis of open chain substrates 13
and 14. Reagents and conditions: (a) Grubbs cat. I (0.2 equiv), CH2Cl2,
0.5 mM, 40 uC, 48 h, 15Z : 15E = 1 : 1, 75% combined yield; (b) Grubbs
cat. II (0.2 equiv), CH2Cl2, 0.5 mM, 40 uC, 48 h, 15Z : 15E = 1 : 1, 35%
combined yield, 16Z : 16E = 1 : 1, 20% combined yield; (c) Grubbs cat. I
(0.2 equiv), CH2Cl2, 0.5 mM, 40 uC, 8 h, 17Z : 17E = 1 : 1, 90% combined
yield; (d) Grubbs cat. II (0.2 equiv), CH2Cl2, 0.5 mM, 40 uC, 8 h, 17Z :
17E = 1 : 1, 70% combined yield, 18, 9%.
alternative pathways for the reaction. In the event, Grubbs cat. I
(CH2Cl2 reflux, 8 h) resulted in the formation of only the dimeric
materials 17Z : 17E as a 1 : 1 mixture and in 90% combined yield.
When Grubbs cat. II was used under the same conditions,
however, the cyclic product 18 was obtained, albeit in low yield
(9%, single isomer), in addition to 17Z : 17E (1 : 1 ratio, 70%
combined yield). Exposure of the dimeric mixture 17Z : 17E to
Grubbs cat. II under the same conditions led to the same ratio of
products 17Z : 17E and 18.
formamide, TIPS
=
triisopropylsilyl, dba
=
dibenzylideneacetone,
HMDS = hexamethyldisilazane, TBAF
= tetra-n-butylammonium
fluoride, Ms = methanesulfonyl, THP = tetrahydropyranyl.
The failure of substrates 13 and 14 to advance to cyclic products
in a synthetically useful way prompted us to modify our approach
to floresolide B (1), so as to enforce the desired olefin metathesis-
based cyclization reactions. It was reasoned that such enforcement
would, perhaps, emerge from a lactone bridge installed within the
molecule prior to the olefin metathesis step,12 and thus the design
of precursor 4 (Fig. 1). Removal of the THP group from
compound 13 (PPTS, MeOH, 75% yield) proceeded smoothly as
shown in Scheme 3 to afford the corresponding alcohol, which was
distribution of products 15Z : 15E and 16Z : 16E was obtained
from 15Z : 15E (1 : 1 ratio) under the same conditions. Faced with
the reluctance of 13 to cyclize, and given the hydride shift side-
products arising from the action of the Grubbs cat. II on its
activated olefin, we chose ketoester 14 as the next olefin metathesis
precursor to be explored. It was reasoned that the absence of the
bulky THP group and of the isopropenyl moiety next to the
aromatic ring would perhaps allow ring closure and avoid
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 600–602 | 601

View Online
It is of interest that both products of the ring olefin metathesis
reaction were formed exclusively as single atropisomers and of the
same form as the natural product (1). This observation is in line
with severe steric congestion for the transition states leading to the
unnatural atropisomers for both E and Z isomers, whereas those
leading to the natural isomers are relatively free of such
interactions (manual molecular models). It is also of interest that
several modifications of the experimental conditions (solvent,
temperature, catalyst) thus far have failed to change the ratio of
the geometrical isomers of the metathesis ring closing reaction.
The described chemistry demonstrates the power of the olefin
metathesis reaction in total synthesis, exposing at the same time its
dependence on substrate structure with regard to product geometry.
Further studies to improve the stereochemical outcome of the
ring closing reaction in this case and in pursuit of other members
of the floresolide family of natural products are in progress.
We thank Drs D. H. Huang, G. Siuzdak, and R. Chadha for
NMR spectroscopic, mass spectrometric, X-ray crystallographic
assistance, respectively. Financial support for this work was
provided by The Skaggs Institute for Chemical Biology and
fellowships from The Skaggs Institute for Research (to H. X.) and
Bristol-Myers Squibb (to H. X.).
Notes and references
1 H. H. Issa, J. Tanaka, R. Rachmat and T. Higa, Tetrahedron Lett.,
2003, 44, 1243.
2 For a recent review on metathesis reactions in total synthesis, see:
K. C. Nicolaou, P. G. Bulger and D. Sarlah, Angew. Chem., Int. Ed.,
2005, 44, 4490.
3 J. G. Bruno, M. N. Chang, Y. M. Choi-Sledeski, D. M. Green, D. G.
McGarry, J. R. Regan and F. A. Volz, J. Org. Chem., 1997, 62, 5174.
4 M. Kosugi, I. Hagiwara, T. Sumiya and T. Migita, Bull. Chem. Soc.
Jpn., 1984, 57, 242.
5 M. Palucki and S. L. Buchwald, J. Am. Chem. Soc., 1997, 119, 11108.
6 For a discussion of palladium-catalyzed direct arylation of ketones or
their equivalent: (a) B. C. Hamann and J. F. Hartwig, J. Am. Chem.
Soc., 1997, 119, 12382; (b) P. Liu, T. J. Lanza, Jr, J. P. Jewell, C. P. Jones,
W. K. Hagmann and L. S. Lin, Tetrahedron Lett., 2003, 44, 8869.
7 G. Majetich, S. Liu, J. Fang, D. Siesel and Y. Zhang, J. Org. Chem.,
1997, 62, 6928.
Scheme 3 Total synthesis of floresolide B (1) and D6,7-Z-floresolide B (2)
and ORTEP drawing of D6,7-Z-floresolide B (2). (Only the (S)-enantiomer
is shown from the two enantiomers present in the crystal.) Reagents and
conditions: (a) PPTS (5.0 equiv), 0.2 M in MeOH, 25 uC, 6 h, 75%; (b)
DMP (1.2 equiv), NaHCO3 (20 equiv), CH2Cl2, 25 uC, 92%; (c) NaClO2
(4.0 equiv), NaH2PO4 (2.0 equiv), 2-methyl-2-butene (5.0 equiv), tBuOH/
H2O (4 : 1), 25 uC, 4 h, 95%; (d) HCl in MeOH (0.1 M), 25 uC, 4 h, 75%;
(e) 2-nitro-6-methyl-benzoic anhydride (2.0 equiv), 4-DMAP (4.0 equiv),
0.5 mM in CH2Cl2, then 20 was added via syringe pump over 12 h, 25 uC,
66%; (f) Grubbs cat. II (0.1 equiv), 0.5 mM in CH2Cl2, 40 uC, 15 min, 22E
(23%), 22Z (66%); (g) K2CO3 (10.0 equiv), MeOH/H2O (1 : 1), 25 uC, 2 h,
90%. PPTS = pyridinium p-toluenesulfonate, DMP = Dess–Martin
periodinane.
8 Compound 10 and its descendents are racemic even though only one
enantiomer is shown.
9 M. B. Andrus, E. J. Hicken, E. L. Meredith, B. L. Simmons and
J. F. Cannon, Org. Lett., 2003, 5, 3859.
10 S. Kamlage, M. Sefkow and M. G. Peter, J. Org. Chem., 1999, 64, 2938;
D. Milstein and J. K. Stille, J. Am. Chem. Soc., 1978, 100, 3636.
11 For ruthenium-catalyzed isomerizations of activated olefins, see:
S. H. Hong, M. W. Day and R. H. Grubbs, J. Am. Chem. Soc.,
2004, 126, 7414 and references therein.
12 For recent examples of templated olefin metathesis reactions, see: (a)
J. D. Badjic, S. J. Cantrill, R. H. Grubbs, E. N. Guidry, R. Orenes and
J. F. Stoddart, Angew. Chem., Int. Ed., 2004, 43, 3273; (b) E. N. Guidry,
S. J. Cantrill and J. F. Stoddart, Org. Lett., 2005, 7, 2129.
13 I. Shinna, M. Kubota, H. Oshiumi and M. Hashizume, J. Org. Chem.,
2004, 69, 1822.
converted to carboxylic acid 19 through a two-step oxidation
procedure (DMP, NaClO2, 66% overall yield). Cleavage of the
MOM group from the latter compound (19) was effected under
acidic conditions (aqueous HCl, MeOH, 75% yield), furnishing
carboxy dihydroquinone 20, which was reacted with 2-nitro-6-
methyl-benzoic anhydride in the presence of 4-DMAP to afford
lactone 21 in 66% overall yield.13 Exposure of 21 to Grubbs cat. II
in refluxing CH2Cl2 pleasantly resulted, within 15 min, in the
formation of cyclic olefins 22Z and 22E in ca. 2.9 : 1 ratio, and
89% combined yield. Finally, cleavage of the nitrobenzoate group
from these compounds with K2CO3 in MeOH furnished, after
HPLC separation, floresolide B (1)14 (20% overall yield from 21)
and its D6,7-Z isomer (2) (59% overall yield from 21).{
The structure of the latter compound (2) was further confirmed
by X-ray crystallographic analysis (see ORTEP drawing,
Scheme 3).15,16
14 We thank T. Higa and J. Tanaka (University of the Ryukyus, Japan)
for 1H- and 13C-NMR spectra of natural floresolide B.
15 CCDC 290049 (2). For crystallographic data in CIF or other electronic
format see DOI: 10.1039/b517385j.
16 Chemical formula (C21H24O3), formula weight (324.40), crystal system
˚
(monoclinic), unit cell dimensions (a = 27.321 (6) A, b = 8.7467 (17) A,
˚
3
˚
˚
c = 15.029 (3) A, b = 97.75 (3)u), unit cell volume (3558.5 (12) A ), Z = 8,
temperature (296 (2) K), space group (C2/c), absorption coefficient
(0.080 mm21), reflections collected (12624), independent reflections
(3134), Rint = 0.0506, R1 = 0.0813, wR2 = 0.1693 for observed data, I >
2s (I).
602 | Chem. Commun., 2006, 600–602
This journal is ß The Royal Society of Chemistry 2006