Studies toward the biomimetic total synthesis of (-)-PF-1018

Article abstract of DOI:10.1021/ol400508s

Pericyclic reaction cascades are unparalleled in their ability to quickly generate complex structures with excellent stereocontrol. Herein, the use of a biomimetic Stille/8π electrocyclization/Diels-Alder cascade to successfully assemble the core structure of (-)-PF-1018 is reported.

Full text of DOI:10.1021/ol400508s

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Studies toward the Biomimetic Total
Synthesis of (À)-PF-1018
Robert Webster,† Boris Gaspar,† Peter Mayer, and Dirk Trauner*
€
€
€
€
Fakultat fur Chemie und Pharmazie, Ludwig-Maximilians-Universitat Munchen, and
Munich Center for Integrated Protein Science, Butenandtstrasse 5-13, 81377 Mu€nchen,
Germany
dirk.trauner@lmu.de
Received February 23, 2013
ABSTRACT
Pericyclic reaction cascades are unparalleled in their ability to quickly generate complex structures with excellent stereocontrol. Herein, the use of
a biomimetic Stille/8π electrocyclization/DielsÀAlder cascade to successfully assemble the core structure of (À)-PF-1018 is reported.
Inadditiontobioactivity, themolecular architecture of a
natural product can provide a major impetus for its choice
as a target in total synthesis. This is especially true when a
beautiful structure can be coupled with efficient and
innovative construction methods. Pericyclic reaction cas-
cades have been shown to play an important role in the
biosynthesis of numerous natural product scaffolds,1 en-
abling chemists to borrow wisdom from nature by choos-
ing bold retrosynthetic disconnections that rapidly build
complexity.2
(À)-PF-1018 (1), an insecticidal polyketide isolated
from a fungal strain Humicola sp.,3 would certainly
appear attractive to many synthetic chemists (Scheme 1).
The molecule features a complex tricyclic core that con-
tains a total of two quaternary and four secondary
stereocenters, all of which are contiguous. This compact
nucleus is connected to a bicyclic tetramic acid moiety
bearing one additional stereocenter. At the time of its
isolation, the relative stereochemistry and overall structure
of (À)-PF-1018 (1) were confirmed by X-ray crystallogra-
phy, whereas its absolute stereochemistry was inferred by
degredation of the natural product and correlation with
L-proline.
Biosynthetically, (À)-PF-1018 (1) could arise from linear
polyene 4 (Scheme 1), itself assembled on a polyketide
synthase (PKS)4 from L-proline, four acetyl CoA units, and
four methylmalonyl CoA units. This hypothetical polyene 4,
contains five trisubstituted double bonds, four Z-configured
and one E-configured, as well as two disubstituted double
bonds, each with the E-configuration, in addition to the
bicyclotetramic acid headgroup. A thermal 8π conrotatory
electrocyclization of the terminal alkenes would yield a 1,3,5-
cyclooctatriene 2 that could subsequently engage in an
intramolecular DielsÀAlder reaction to afford (À)-PF-1018
(1). It is common, however, for cyclooctatrienes of type 2 to
rapidly undergo thermal 6π disrotatory electrocyclizations to
yield bicyclo[4.2.0]octadienes, such as compound 3.5 At least
(4) Aparicio, J. F.; Caffrey, P.; Gil, J. A.; Zotchev, S. B. Appl.
Microbiol. Biotechnol. 2003, 61, 179.
(5) For examples of biomimetic syntheses featuring 8π/6π electro-
cyclization cascades, see (endriantric acids): (a) Nicolaou, K. C.; Petasis,
N. A.; Zipkin, R. E. J. Am. Chem. Soc. 1982, 104, 5560. SNF4435 C/D:
(b) Parker, K. A.; Lim, Y.-H. J. Am. Chem. Soc. 2004, 126, 15968. (c)
Beaudry, C.; Trauner, D. Org. Lett. 2005, 7, 4475. (d) Jocobsen, M. F.;
Moses, J. E.; Adlington, R. M.; Baldwin, J. E. Tetrahedron 2006, 62,
1675. Shimalactones: (e) Sofiyev, V.; Navarro, G.; Trauner, D. Org.
Lett. 2008, 10, 149. Elisiapyrones A/B: (f) Barbarow, J. E.; Miller, A. K.;
Trauner, D. Org. Lett. 2005, 7, 2901. Ocellapyrones A/B: (g) Miller,
A. K.; Trauner, D. Angew. Chem., Int. Ed. 2005, 44, 4602. Prekingianin
A: (h) Sharma, P.; Ritson, D. J.; Burnley, J.; Moses, J. E. Chem Commun.
2011, 47, 10605. Kingianin A: (i) Lim, H. N.; Parker, K. A. Org. Lett.
2013, 15, 398.
† These authors contributed equally to this work.
(1) Beaudry, C.; Melarich, J.; Trauner, D. Chem. Rev. 2005, 105,
4757.
(2) (a) Nicolaou, K. C.; Sorensen, E. J. Classics in Total Synthesis;
VCH: Weinheim, Germany, 1996. (b) Bulger, P. G.; Bagal, S. K.; Marquez, R.
Nat. Prod. Rep. 2008, 25, 254.
(3) Gomi, S.; Imamura, K.; Yaguchi, T.; Kodama, Y.; Minowa, N.;
Koyama, M. J. Antibiot. 1994, 47, 571.
r
10.1021/ol400508s
XXXX American Chemical Society

Scheme 1. Biosynthetic Analysis of PF-1018
Scheme 3. Asymmetric Synthethsis of the Tricyclic Core
conceivably, the chemical environment of the PKS may be
required for the DielsÀAlder reaction to outpace the 6π
electrocyclization, representing a potential design challenge in
our synthesis. Additionally, the potential reversibility of the
6π electrocyclization also has to be taken into account,
although the relevance of this process under biosynthetic
conditions is questionable given the relatively high tempera-
tures typically required for 6π cycloreversions.6
was to occur preferentially, we felt that the sole stereo-
center in the tetramic acid portion would be unlikely to
control its stereochemical outcome. Therefore, we opted to
dissect (À)-PF-1018 (1) into the known proline-derived
phosponate 57 and the “hydrated” polyene component of
type 6 (Scheme 2). The latter fragment would fulfill three
purposes: (a) to reduce the number of available electro-
cyclization modes, (b) to potentially suppress the 6π
electrocyclization in favor of the desired intramolecular
DielsÀAlder reaction, and (c) to enable an asymmetric
approach to the tricyclic core, independent from the
bicyclotetramic acid moiety. Further retrosynthetic dis-
connection led us to vinyl iodide 7 and vinyl stannane 8,8
using a strategy that we had successfully employed in the
synthesis of other highly unsaturated polyketides.5eÀg
Starting from aldehyde 9, available in five steps from
propargyl alcohol,5f a syn-selective Evans aldol employing
oxazolidinone 10 provided alcohol 11 as a single diaste-
reomer (Scheme 3).9 A two-step trans-amidation/silylation
protocol gave protected Weinreb amide 12, which was in
turn reduced to the corresponding aldehyde and subse-
quently olefinated to afford methyl ester 7. The key
domino Stille/8π electrocyclization/DielsÀAlder cascade
using Liebeskind-type Stille conditions,10 followed by
heating, furnished the desired tricycle 15 in an optimized
Scheme 2. Retrosynthetic Analysis of PF-1018
From a synthetic point of view, polyene 4 was deemed an
unsuitable intermediate for various reasons. We antici-
pated that this polyene would be highly unstable due to its
inbuilt acidic functionality. Second, polyene 4 could po-
tentially engage in multiple pericyclic reaction modes that
would be difficult to control. Finally, even if the 8π
electrocyclization involving the terminal double bonds
(7) Rosen, T.; Fernandes, P. B.; Marovich, M. A.; Shen, L.; Mao, J.;
Pernet, A. J. Med. Chem. 1989, 32, 1062.
(8) Vinyl stannane 8 was prepared in four steps from methyl angelate.
Detailed procedures are provided in the Supporting Information.
(9) Gage, J. R.; Evans, D. A. Org. Synth. 1990, 68, 83.
(10) Zhang, S.; Zhang, D.; Liebeskind, L. J. Org. Chem. 1997, 62,
2312.
(6) (a) Marvell, E. N. Thermal Electrocyclic Reactions; Academic
Press: New York, 1980. (b) Okamura, W. H.; de Lera, A. R. 1,3-Cyclohex-
adiene Formation Reactions. In Comprehensive Organic Synthesis; Trost,
B. M., Fleming, I., Paquette, L. A., Eds.; Pergamon Press; New York, 1991:
Vol. 5, section 6, pp 699À750.
B
Org. Lett., Vol. XX, No. XX, XXXX

yield of 32% as a single diastereomer. Compound 15 was
the only isomer, among many conceivable ones, that could
be isolated under these conditions.11 Its structure was
verified by X-ray crystallography which also confirmed
its absolute configuration by virtue of the heavy silicon
atom (Figure 1). It is noteworthy that employing the
diastereomeric vinyl iodide anti-7, as well variants of the
vinyl iodide 7 with the alcohol unprotected, yielded no
isolable DielsÀAlder products under the same conditions
and only complex mixtures that contained isomeric
bicyclo[4.2.0]octadienes were obtained.
Unfortunately, despite extensive efforts,12 dehydration of
alcohol 18 by syn-elimination or E1-elimination was also
problematic and could never be effected in synthetically
useful yields.
Scheme 4. Derivitization of the Tricyclic Core and Successful
Inversion of the Alcohol Functionality
Figure 1. X-ray structure of tricyclic ester 15.
With the tricyclic diene 15 now in hand, we were faced
with the challenge of converting it to the corresponding
triene in order to complete the core structure of (À)-PF-
1018 (1). We anticipated that removing the ester function-
ality would improve thermal stability toward homolytic or
heterolytic bond cleavageof the carbon skeleton, and diene
15 was first desilylated and then reduced to give diol 16
(Scheme 4). After selective protection of the primary
alcohol moiety, it was possible to obtain hindered ketone
17 via oxidation with DMP. However, further transforma-
tion into the corresponding enol triflate or hydrazone, as
well as epimerization of the R-methyl group using base,
proved unsuccessful, even under forcing conditions.
We reasoned that the hindered, neopentylic alcohol 18
may possess an unfavorable geometry for the requisite
stereoelectronics of a syn-elimination, prompting us to
investigate methods for its stereochemical inversion. Re-
duction of ketone 17proceeded smoothly under Birch-type
conditions (Li, NH3/THF, À78 °C) but was completely
unselective, yielding a 1:1 dr. By contrast, conversion of
alcohol 18 into its chloromethyl sulfonate derivative 19,
followed by treatment with CsOAc at elevated tempera-
tures and prolonged reaction times,13 cleanly yielded the
inverted acetate 20. Solvolysis of acetate 20 delivered the
inverted alcohol 21, which was accompanied by a small
amount desilylation to give diol 22, the structure of which
was confirmed by X-ray crystallography (Figure 2). We
were encouraged by the axial orientation of the alcohol
evident from the crystal structure, the geometry of which
appears well-suited for anti-elimination. However, expo-
sure of alcohol 21 to an array of standard elimination
conditions has not yet successfully introduced the requisite
alkene for the core structure of (À)-PF-1018 (1). All
attempts to convert the secondary alcohol function in 21
into a leaving group led to its decomposition, presumably
via interference of the proximate, highly nucleophilic
alkene functionality.
(11) The majority of the remaining mass balance from the key
reaction cascade consisted of a nonpolar fraction containing a complex
mixture of inseparable products. The various components of this
intractable mixture likely originate from tetraene 6 via alkene isomer-
izations and/or alternate available pericyclic reaction modes. Despite
attempted purification, no additional products could be successfully
isolated or identified.
(12) Including, but not limited to, Burgess reagent: (a) Atkins, G. M.,
Jr.; Burgess, E. M. J. Am. Chem. Soc. 1968, 90, 4744. (b) Burgess, E. M.;
Penton, H. R., Jr.; Taylor, E. A. J. Org. Chem. 1973, 38, 26. (c) Ishikawa,
H.; Elliott, G.; Velcicky, J.; Choi, Y.; Boger, D. J. Am. Chem. Soc. 2006,
128, 10596. Martin’s sulfurane: (d) Arhart, R. J.; Martin, J. C. J. Am.
Chem. Soc. 1972, 94, 5003. (e) Martin, J. C.; Franz, J. A.; Arhart, R. J.
J. Am. Chem. Soc. 1974, 96, 4604. Chugaev elimination: (f) Tschugaeff,
L. Ber. Dtsch. Chem. Ges. 1900, 33, 3118. (g) Nace, H. R. Org. React.
2011, 57–100. Re2O7-mediated elimination: (h) Korstanje, T. J.;
de Waard, E. F.; Jastrzebski, J. T.; Klein Gebbink, R. J. M. ACS Catal.
2012, 2, 2173. Tf2O, 2,6-lutidine: (i) Tanino, K.; Onuki, K.; Asano, K.;
Miyashita, M.; Nakamura, T.; Takahashi, Y.; Kuwajima, I. J. Am.
Chem. Soc. 2003, 125, 1498. POCl3, pyridine: (j) Birch, A. M.; Pattenden,
G. J. Chem Soc., Chem. Commun. 1980, 24, 1195. SOCl2, pyridine: (k)
Watanabe, H.; Takano, M.; Umino, A.; Ito, T.; Ishikawa, H.; Nakada,
M. Org. Lett. 2007, 9, 359. MsCl, Et3N: (l) Bonikowski, R.; Kula, J.;
Bujacz, A.; Wajs-Bonikowska, A.; Zakzos-Szyda, M.; Wysocki, S.
Tetrahedron: Asymmetry 2012, 23, 1038. Mitsonobu-type elimination:
(m) Izuhara, T.; Katoh, T. Tetrahedron Lett. 2000, 41, 7651.
(13) (a) Shimizu, T.; Hiranuma, S.; Nakata, T. Tetrahedron Lett. 1996,
37, 6145. (b) Shirahata, T.; Sunazuka, T.; Yoshida, K.; Yamamoto, D.;
Harigaya, Y.; Kuwajima, I.; Nagai, T.; Kiyohara, H.; Yamada, H.; Omura,
S. Tetrahedron Lett. 2006, 62, 9483.
Org. Lett., Vol. XX, No. XX, XXXX
C

the fact that tricycle 15 could be reproducibly isolated as a
single diastereomer in synthetically useful yields is remark-
able. Extensive effort was channeled into dehydration of this
adduct, and studies directed at incorporation of the final
alkene and completion of the total synthesis are currently
underway.
Acknowledgment. This work was supported by SFB 749
“Dynamics and Intermediates of Molecular Transforma-
€
€
tions”. R.W. (LMU Munchen) and B.G. (LMU Munchen)
are each grateful to the Alexander von Humboldt founda-
tion for postdoctoral fellowships. Additionally, the authors
would like to acknowledge Dr. Tatsuya Urushima (LMU
€
Munchen) for helpful discussions and assistance with scale-
up experiments.
Figure 2. X-ray structure of diol 22.
Supporting Information Available. Experimental pro-
cedures and characterization of all new compounds,
including X-ray data for compounds 15 and 21. This
material is available free of charge via the Internet at
http://pubs.acs.org.
In summary, we have completed a concise asymmetric
synthesis of the tricyclic core of (À)-PF-1018 (1), utilizing a
three-fold domino Stille/8π electrocyclization/DielsÀAlder
sequence. Considering the number of pericyclic reaction
modes available to the intermediate polyene 6 from the key
cascade, as well as the potential formation of diastereomers,
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX