Multicomponent Coupling Approach to (±)-Frondosin B and a Ring-Expanded Analogue

Article abstract of DOI:10.1021/ol035822q

(Equation presented) A recently discovered multicomponent coupling reaction is used to give direct access to a late intermediate in the synthesis of frondosin B. This intermediate can also be efficiently converted to a ring-expanded analogue of frondosin

Full text of DOI:10.1021/ol035822q

ORGANIC
LETTERS
2004
Vol. 6, No. 4
457-460
Multicomponent Coupling Approach to
(±)-Frondosin B and a Ring-Expanded
Analogue
,†,‡
Daniel J. Kerr,†,‡ Anthony C. Willis,§ and Bernard L. Flynn*
Department of Chemistry and The Research School of Chemistry,
Australian National UniVersity, Canberra, ACT, 0200, Australia
bernard.flynn@Vcp.monash.edu.au
Received September 22, 2003 (Revised Manuscript Received December 14, 2003)
ABSTRACT
A recently discovered multicomponent coupling reaction is used to give direct access to a late intermediate in the synthesis of frondosin B.
This intermediate can also be efficiently converted to a ring-expanded analogue of frondosin B by sustained heating of the reaction mixture.
An unprecedented tandem 1,7-hydrogen shift, 8π-electrocyclization is proposed to explain the formation of this ring-expanded species.
Multicomponent reactions (MCRs) are, by definition, both
convergent and multibond forming and, as such, are capable
of synthetically addressing structural complexity in a manner
that is both intellectually satisfying and practical.1 Recently,
we reported a palladium-catalyzed multicomponent coupling
(MCC) approach to indoles and benzofurans (Scheme 1),
which we have subsequently applied to a one-step synthesis
of some potent analogues of the anticancer agent com-
bretastatin A-4.2,3 Herein, we report the application of this
reaction to the concise synthesis of (()-frondosin B and a
ring-expanded analogue through a remarkable reaction
cascade.
(+)-Frondosin B (2) belongs to a family of related marine
sesquiterpenoids, the frondosins 1-5, isolated from a marine
sponge Dysidea frondosa (Figure 2).4 The activity of (+)-
frondosin B (2) as an interleukin-8 (IL-8) receptor antagonist
and its novel molecular architecture have inspired the groups
of Danishefsky and Trauner to pursue its total synthesis.5,6
Danishefsky and co-workers prepared 2 in 18 steps in overall
0.8% yield and 88% ee.5 Hughes and Trauner prepared 2 in
Scheme 1. Multicomponent Coupling (MCC) Approach to
Benzofurans and Indoles
† Department of Chemistry.
‡ New Address: Department of Medicinal Chemistry, Victorian College
of Pharmacy, Monash University, 381 Royal Pde, Parkville, Victoria, 3052,
Australia.
§ The Research School of Chemistry.
(1) (a) Domling, D.; Ugi, I. Angew. Chem., Int. Ed. 2000, 39, 3168. (b)
Lee, D.; Sello, J. K.; Schreiber; S. L. Org. Lett. 2000, 2, 709. (c) Orru, R.
V. A.; de Greef, M. Synthesis 2003, 1471. (d) Weber, L. Curr. Med. Chem.
2002, 9, 2085. (e) Weber, L. Curr. Opin. Chem. Biol. 2000, 4, 295. (f)
Hulme, C.; Gore, V. Curr. Med. Chem. 2003, 10, 51. (g) Hughes, B.; Hulme,
C.; Oddon G.; Schmitt, P. Chem. Eur. J. 2000, 6, 3321. (h) Balme, G.;
Bossharth, E.; Monteiro, N. Eur. J. Org. Chem. 2003, 4101.
(2) (a) Chaplin, J. H.; Flynn, B. L. Chem. Commun. 2001, 1594. (b)
Flynn, B. L.; Hamel, E.; Jung, M. K. J. Med. Chem. 2002, 45, 2670.
(3) This MCC was based on some earlier studies performed by the group
of Cacchi and Arcadi: Arcadi, A.; Cacchi, S.; Del Rosario, M.; Fabrizi,
G.; Marinelli, F. J. Org. Chem. 1996, 61, 9280
10.1021/ol035822q CCC: $27.50 © 2004 American Chemical Society
Published on Web 01/27/2004

Figure 1. Frondosins. The stereochemistry of frondosins A and
C-E (1 and 3-5, respectively) is based on that determined for B
(2) and has not been independently determined.
20 steps in an overall 7.3% yield and 91% ee from
commercially available substrates.6 While these syntheses
provide enantioselective access to the natural product, they
are somewhat lengthy, so we sought to utilize our MCC
protocol to provide a more concise, convergent access to
this system and its analogues in order to aid structure-
activity relationship (SAR) studies.
Figure 2. ORTEP representation of a molecule of 15 derived from
a crystallographic study (arbitrary numbering).
providing a three-step (longest linear sequence) access to
the basic core of frondosin B, 17, from commercially
available substrates.
Commercially available 4-methoxyphenol (6) and 2-allyl-
1,3-cyclohexandione (7) were converted to the bromides 8
and 10, respectively, in excellent yields following standard
procedures (Scheme 2).7,8 These two bromides and the
commercially available 3-methylbutenyne (9) were subjected
to our MCC protocol. This involved deprotonating 8 and 9
with MeMgBr and coupling the resultant o-bromophenolate
and acetylide (not shown) using palladium to give the
o-alkynylphenolate 11, which undergoes heteroannulative
coupling with 10 at 80 °C giving the desired product 12 in
an acceptable yield (48%). Other products resulting from the
MCC included the protocyclized material 13 the nucleophilic
addition-elimination product 14 and an unusual polycyclic
product 15. It was demonstrated that the latter product, 15,
could be selectively formed by extended heating of the
reaction mixture at 100 °C for 48 h (61%), the structure of
this material was determined by spectral methods and
confirmed using X-ray crystallography (Figure 2).9
Reetz and co-workers have described a method for
converting ketones to gem-dimethyl groups using 2 equiv
of Me2TiCl2.10 In our attempts to convert 17 to 18 using the
Reetz method, we observed very rapid reaction of 17 with
Me2TiCl2 to give 18 at 0 °C but only to the point of
approximately 50% conversion (41% 17 and 40% 18,
isolated), without any observable improvement after standing
at room temperature for several hours (Scheme 3).11 Interest-
ingly, direct TLC analysis of the supernatant of the reaction
mixture revealed almost complete consumption of the starting
material 17 after 1 h but after hydrolytic workup, TLC
analysis of the organic (EtOAc) extract revealed considerable
starting material. This situation remained the same even in
the presence of large excesses of Reetz reagent. Further
experimentation revealed that considerable increases in yield
could be obtained when the reaction was left to stir at room
temperature for several days and that heating the reaction
mixture to 85 °C (1,2-dichloroethane used in place of
dichloromethane) for 24 h would give the product 18 in high
The desired MCC product 12 was efficiently cyclized
using catalyst 16 in a ring-closing metathesis (RCM) reaction,
(9) Crystal data: C21H22O3, M ) 322.40, monoclinic, a ) 10.5169(2),
b ) 9.5041(2), c ) 16.8671(3) Å, â ) 100.0201(7) °, U ) 1660.21(6) Å3,
(4) (a) Patil, A. D.; Freyer, A. J.; Killmer, L.; Offen, P.; Carte, B.;
Jurwicz, A. J.; Johnson, R. K. Tetrahedron 1997, 57, 5047. (b) Hallock, Y.
F.; Cerdenellina. J. H.; Boyd, M. R. Nat. Prod. Lett. 1998, 11, 153.
(5) (a) Danishefsky, S. J.; Inoue, M.; Frontier, A. J. Angew. Chem., Int.
Ed. 2000, 39, 761-764. (b) Inoue, M.; Carson, M. W.; Frontier, A. J.;
Danishefsky, S. J. J. Am. Chem. Soc. 2001, 123, 1878.
(6) Hughes, C. C.; Trauner, D. Angew. Chem., Int. Ed. 2002, 41, 1569-
1572. (Erratum: Angew. Chem., Int. Ed. 2002, 41, 2227).
(7) (a) Kajigaeshi, S.; Kakinami, T.; Okamoto, T.; Nakamura, H.;
Fujikawa, M. Bull. Chem. Soc. Jpn. 1987, 60, 4187. (b) Dodsworth D. J.;
Calcagno, M. P.; Ehrmann, E. U.; Devadas, B.; Sammes, P. G. J. Chem
Soc., Perkin Trans. 1 1981, 2120.
T ) 200 K, space group P21/a (no. 14), Z ) 4, µ(Mo KR) ) 0.085 mm-1
,
37 199 reflections measured, 3811 unique (Rint ) 0.050), 1811 with I >
3σ(I) used in refinement. The final R ) 0.029 and wR ) 0.034 for the
reflections used in the refinement. X-ray diffraction data were collected on
a Nonius Kappa CCD diffractometer (graphite monochromator, λ ) 0.71073
Å). Structure solution was by direct methods and refinement completed by
full-matrix least-squares on F. Non-hydrogen atoms were refined with
anisotropic displacement parameters, and hydrogen atoms were included
at calculated positions and ride on the atoms to which they are bonded.
The crystallographic data have been deposited with the Cambridge
Crystallographic Data Center (CCDC 227359).
(8) Piers, E.; Grierson, J. R.; Lau, C. K.; Nagakura, I. Can. J. Chem.
1982, 60, 210.
(10) Reetz, M. T.; Westermann, J.; Kyung, S. Chem. Ber. 1985, 118,
1050.
458
Org. Lett., Vol. 6, No. 4, 2004

Scheme 2 a
a Reagents and conditions: (a) Br2, CH2Cl2. (b) PPh3.Br2, Et3N, CH2Cl2. (c) 8 and 9, 2.1 x MeMgBr, THF 0 °C then 5 mol % Pd(PPh3)2Cl2,
65 °C for 24 h and then 10 and DMSO, 80 °C for 9 h (for selective formation of 12) or 100 °C for 48 h (for selective formation of 15).
(d) Grubbs’ catalyst 16, 1,2-dichloroethane 55 °C for 1 h. (e) 3 x Me2TiCl2 (formed in situ from Me2Zn and TiCl4), 1,2-dichloroethane,
-30 to 18°C over 2 h, 83 °C for 20 h. (f) 10 mol % 10% Pd/C, MeOH, H2 1 atm, 9 h. (g) as for e in CH2Cl2, 0-18°C over 2 h, and then
40 °C for 19 h.
yield (Scheme 3). In short, the gem-dimethylation reaction
appeared to proceed in two stages, a rapid initial stage to
the point of approximately 50% conversion, with the
remainder of substrate 17 evidently tied up as a hydrolytically
labile species, which converts to 18 in a slow second stage.
We have tentatively explained these observations as resulting
from an initial reaction of 17 with the requisite 2 equiv of
Me2TiCl2, giving 18 and a byproduct complex 21 (Scheme
3). As this byproduct 21 is formed, it reacts with the starting
material 17 at faster a rate than Me2TiCl2 to give an insoluble
complex of unknown structure (possibly of the structure 22)
that hydrolyses upon aqueous workup to return 17. This
complex 22 may react directly but slowly with the Me2TiCl2
to give 18 or through reversible formation of 17 where the
overall reaction rate is reduced by virtue of the low
equilibrium concentration of 17 present.
A similar two-stage reaction profile was also observed in
the gem-dimethylation of 15 to give 20, but again, a good
Scheme 3 a
(11) During the course of our work in this area, Hughes and Trauner
reported their synthesis of frondosin B (ref 5), in which they attempted to
use the Reetz reaction to convert a ketone very similar to 17, I, to the gem-
dimethyl product III. They reported that “the low electrophilicity of the
vinylogous aryl ketone (I)” was responsible for the initial failure of the
Reetz reaction, which they then conducted in two steps, first converting
the ketone I to the alcohol II with MeMgBr and then reacting III with
Me2TiCl2 to give III. Based on the results of this study, such a modification
of the Reetz protocol may not be necessary.
b
a In CH2Cl2 with 2 equiv of Me2TiCl2. In 1,2-dichloroethane
with 2 equiv of Me2TiCl2.
Org. Lett., Vol. 6, No. 4, 2004
459

yield of product could be achieved upon overnight heating
(Scheme 2).
to giffordene 26 has been shown to involve a similar 1,7-
hydrogen shift from precursor 25, this reaction is selective
for the (6E)- arrangement of giffordene 26, preventing what
would otherwise be a very rapid 8π-electrocyclization from
occurring (eq 1).13 We propose that unlike 25, which
undergoes complete rearrangement through to the lower
energy (more fully conjugated) tetraene system of giffordene
26, the 1,7-hydrogen shift of 12 to 23 or 24 is reversible
with the equilibrium lying to the left, in favor of the
aromatized benzofuran. Even though 24 may be the highest
energy arrangement of the three interconvertible species, 12,
23, and 24, the fact that it can be rapidly and irreversibly
converted to 15 through 8π-electrocyclization provides the
necessary driving force for directing the total reaction flux
through this intermediate. 8π-Electrocyclizations are gener-
ally quite facile,13 and the cyclization of 24 to give 15 is
expected to be particularly rapid since the benzofuran and
cyclohexyl rings in 24 reduce the degrees of freedom of the
molecule and because it restores aromaticity to the furan ring.
In summary, we have provided a concise (six-step),
protection-groupless access to (()-frondosin B 2 from
commercially available substrates 6, 7, and 9 in an overall
32% yield. The two key steps, MCC and RCM, are all that
are required to convert any given set of 2-bromophenols,
butenynes, and 1-bromo-2-allyl-1-cyclohexenes to a variety
of frondosin B analogues, assisting SAR studies on these
systems as IL-8 inhibitors. Furthermore, the MCC can be
placed in tandem with a pericyclic cascade to provide direct
access to a series of eight-membered ring analogues of the
type 15. Such concise, divergent access to complex poly-
cyclic cores may have broader implications for the diversity-
orientated synthesis of natural product-like compounds in
other chemical biology studies.
Selective hydrogenation of the sterically least con-
gested C8-C9 double bond in 18 was achieved in ex-
cellent yield with a palladium on charcoal catalyst giving
19 (90%). Attempts to hydrogenate 18 enantioselectively
using Pfaltz-type catalysts have thus far proven unsuccess-
ful.12
Cleavage of the methyl ether in racemic 19 to give (()-
frondosin B (2) has been accomplished by others using
sodium methylthiolate.5,6
Albeit racemic, this approach to (()-frondosin B does
nonetheless provide ready access to analogues of this material
for SAR studies. Further, studies into the enantioselective
hydrogenation of 18 are ongoing.
Most remarkable is the single-step formation of 15. This
step brings three substrates together, forming four new bonds
and two new rings. The most likely mechanism of formation
of this product seems to be through a tandem 1,7-hydrogen
shift, 8π-electrocyclization process, converting 12 into 15
(Scheme 4). That 15 results from simple thermal rearrange-
Scheme 4. Proposed Mechanism for the Thermal
Rearrangement of 12 to 15
Acknowledgment. The authors thank Professor Pfaltz and
Mr. Frederick Menges (ETH, Switzerland) for providing a
sample of one of their chiral hydrogenation catalysts and
the Australian Research Council for providing B.L.F. with
an Australian Research Fellowship.
ment of 12 and is not mediated by any other reagent or
catalyst used in the MCC was evidenced by the fact that
heating isolated 12 in toluene at 100 °C for 48 h gave 15 in
high yield (95%). To the best of the authors knowledge, such
tandem 1,7-hydrogen shift, 8π-electrocyclization processes
are unprecedented. While the biosynthetic pathway leading
Supporting Information Available: Full experimental
and spectroscopic details on all compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
(12) Menges, F.; Pfaltz, A. AdV. Synth. Catal. 2002, 344, 40. A sample
of a chiral N,P-chelated iridium catalyst was kindly supplied to us by this
group. However, this catayst appeared to be inhibited by 18 since even
known substrates were not able to be hydrogenated in the presence of a
small amount of 18.
OL035822Q
(13) Pohnert, G.; Boland, W. Tetrahedron 1994, 50, 10235.
460
Org. Lett., Vol. 6, No. 4, 2004